135 files changed, 250639 insertions, 0 deletions

diff --git a/generic/README b/generic/README
new file mode 100644
index 0000000..d1c078e
--- /dev/null
+++ b/generic/README
@@ -0,0 +1,3 @@
+This directory contains Tcl source files that work on all the platforms
+where Tcl runs (e.g. UNIX, PCs, and MacOSX).  Platform-specific
+sources are in the directories ../unix, ../win, and ../macosx.
diff --git a/generic/regc_color.c b/generic/regc_color.c
new file mode 100644
index 0000000..92e0aad
--- /dev/null
+++ b/generic/regc_color.c
@@ -0,0 +1,856 @@
+/*
+ * colorings of characters
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer. All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * Note that there are some incestuous relationships between this code and NFA
+ * arc maintenance, which perhaps ought to be cleaned up sometime.
+ */
+
+#define	CISERR()	VISERR(cm->v)
+#define	CERR(e)		VERR(cm->v, (e))
+
+/*
+ - initcm - set up new colormap
+ ^ static void initcm(struct vars *, struct colormap *);
+ */
+static void
+initcm(
+    struct vars *v,
+    struct colormap *cm)
+{
+    int i;
+    int j;
+    union tree *t;
+    union tree *nextt;
+    struct colordesc *cd;
+
+    cm->magic = CMMAGIC;
+    cm->v = v;
+
+    cm->ncds = NINLINECDS;
+    cm->cd = cm->cdspace;
+    cm->max = 0;
+    cm->free = 0;
+
+    cd = cm->cd;		/* cm->cd[WHITE] */
+    cd->sub = NOSUB;
+    cd->arcs = NULL;
+    cd->flags = 0;
+    cd->nchrs = CHR_MAX - CHR_MIN + 1;
+
+    /*
+     * Upper levels of tree.
+     */
+
+    for (t=&cm->tree[0], j=NBYTS-1 ; j>0 ; t=nextt, j--) {
+	nextt = t + 1;
+	for (i=BYTTAB-1 ; i>=0 ; i--) {
+	    t->tptr[i] = nextt;
+	}
+    }
+
+    /*
+     * Bottom level is solid white.
+     */
+
+    t = &cm->tree[NBYTS-1];
+    for (i=BYTTAB-1 ; i>=0 ; i--) {
+	t->tcolor[i] = WHITE;
+    }
+    cd->block = t;
+}
+
+/*
+ - freecm - free dynamically-allocated things in a colormap
+ ^ static void freecm(struct colormap *);
+ */
+static void
+freecm(
+    struct colormap *cm)
+{
+    size_t i;
+    union tree *cb;
+
+    cm->magic = 0;
+    if (NBYTS > 1) {
+	cmtreefree(cm, cm->tree, 0);
+    }
+    for (i=1 ; i<=cm->max ; i++) {	/* skip WHITE */
+	if (!UNUSEDCOLOR(&cm->cd[i])) {
+	    cb = cm->cd[i].block;
+	    if (cb != NULL) {
+		FREE(cb);
+	    }
+	}
+    }
+    if (cm->cd != cm->cdspace) {
+	FREE(cm->cd);
+    }
+}
+
+/*
+ - cmtreefree - free a non-terminal part of a colormap tree
+ ^ static void cmtreefree(struct colormap *, union tree *, int);
+ */
+static void
+cmtreefree(
+    struct colormap *cm,
+    union tree *tree,
+    int level)			/* level number (top == 0) of this block */
+{
+    int i;
+    union tree *t;
+    union tree *fillt = &cm->tree[level+1];
+    union tree *cb;
+
+    assert(level < NBYTS-1);	/* this level has pointers */
+    for (i=BYTTAB-1 ; i>=0 ; i--) {
+	t = tree->tptr[i];
+	assert(t != NULL);
+	if (t != fillt) {
+	    if (level < NBYTS-2) {	/* more pointer blocks below */
+		cmtreefree(cm, t, level+1);
+		FREE(t);
+	    } else {		/* color block below */
+		cb = cm->cd[t->tcolor[0]].block;
+		if (t != cb) {	/* not a solid block */
+		    FREE(t);
+		}
+	    }
+	}
+    }
+}
+
+/*
+ - setcolor - set the color of a character in a colormap
+ ^ static color setcolor(struct colormap *, pchr, pcolor);
+ */
+static color			/* previous color */
+setcolor(
+    struct colormap *cm,
+    pchr c,
+    pcolor co)
+{
+    uchr uc = c;
+    int shift;
+    int level;
+    int b;
+    int bottom;
+    union tree *t;
+    union tree *newt;
+    union tree *fillt;
+    union tree *lastt;
+    union tree *cb;
+    color prev;
+
+    assert(cm->magic == CMMAGIC);
+    if (CISERR() || co == COLORLESS) {
+	return COLORLESS;
+    }
+
+    t = cm->tree;
+    for (level=0, shift=BYTBITS*(NBYTS-1) ; shift>0; level++, shift-=BYTBITS){
+	b = (uc >> shift) & BYTMASK;
+	lastt = t;
+	t = lastt->tptr[b];
+	assert(t != NULL);
+	fillt = &cm->tree[level+1];
+	bottom = (shift <= BYTBITS) ? 1 : 0;
+	cb = (bottom) ? cm->cd[t->tcolor[0]].block : fillt;
+	if (t == fillt || t == cb) {	/* must allocate a new block */
+	    newt = (union tree *) MALLOC((bottom) ?
+		    sizeof(struct colors) : sizeof(struct ptrs));
+	    if (newt == NULL) {
+		CERR(REG_ESPACE);
+		return COLORLESS;
+	    }
+	    if (bottom) {
+		memcpy(newt->tcolor, t->tcolor, BYTTAB*sizeof(color));
+	    } else {
+		memcpy(newt->tptr, t->tptr, BYTTAB*sizeof(union tree *));
+	    }
+	    t = newt;
+	    lastt->tptr[b] = t;
+	}
+    }
+
+    b = uc & BYTMASK;
+    prev = t->tcolor[b];
+    t->tcolor[b] = (color) co;
+    return prev;
+}
+
+/*
+ - maxcolor - report largest color number in use
+ ^ static color maxcolor(struct colormap *);
+ */
+static color
+maxcolor(
+    struct colormap *cm)
+{
+    if (CISERR()) {
+	return COLORLESS;
+    }
+
+    return (color) cm->max;
+}
+
+/*
+ - newcolor - find a new color (must be subject of setcolor at once)
+ * Beware: may relocate the colordescs.
+ ^ static color newcolor(struct colormap *);
+ */
+static color			/* COLORLESS for error */
+newcolor(
+    struct colormap *cm)
+{
+    struct colordesc *cd;
+    size_t n;
+
+    if (CISERR()) {
+	return COLORLESS;
+    }
+
+    if (cm->free != 0) {
+	assert(cm->free > 0);
+	assert((size_t) cm->free < cm->ncds);
+	cd = &cm->cd[cm->free];
+	assert(UNUSEDCOLOR(cd));
+	assert(cd->arcs == NULL);
+	cm->free = cd->sub;
+    } else if (cm->max < cm->ncds - 1) {
+	cm->max++;
+	cd = &cm->cd[cm->max];
+    } else {
+	struct colordesc *newCd;
+
+	/*
+	 * Oops, must allocate more.
+	 */
+
+	if (cm->max == MAX_COLOR) {
+	    CERR(REG_ECOLORS);
+	    return COLORLESS;		/* too many colors */
+	}
+	n = cm->ncds * 2;
+	if (n > MAX_COLOR + 1) {
+	    n = MAX_COLOR + 1;
+	}
+	if (cm->cd == cm->cdspace) {
+	    newCd = (struct colordesc *) MALLOC(n * sizeof(struct colordesc));
+	    if (newCd != NULL) {
+		memcpy(newCd, cm->cdspace,
+			cm->ncds * sizeof(struct colordesc));
+	    }
+	} else {
+	    newCd = (struct colordesc *)
+		    REALLOC(cm->cd, n * sizeof(struct colordesc));
+	}
+	if (newCd == NULL) {
+	    CERR(REG_ESPACE);
+	    return COLORLESS;
+	}
+	cm->cd = newCd;
+	cm->ncds = n;
+	assert(cm->max < cm->ncds - 1);
+	cm->max++;
+	cd = &cm->cd[cm->max];
+    }
+
+    cd->nchrs = 0;
+    cd->sub = NOSUB;
+    cd->arcs = NULL;
+    cd->flags = 0;
+    cd->block = NULL;
+
+    return (color) (cd - cm->cd);
+}
+
+/*
+ - freecolor - free a color (must have no arcs or subcolor)
+ ^ static void freecolor(struct colormap *, pcolor);
+ */
+static void
+freecolor(
+    struct colormap *cm,
+    pcolor co)
+{
+    struct colordesc *cd = &cm->cd[co];
+    color pco, nco;		/* for freelist scan */
+
+    assert(co >= 0);
+    if (co == WHITE) {
+	return;
+    }
+
+    assert(cd->arcs == NULL);
+    assert(cd->sub == NOSUB);
+    assert(cd->nchrs == 0);
+    cd->flags = FREECOL;
+    if (cd->block != NULL) {
+	FREE(cd->block);
+	cd->block = NULL;	/* just paranoia */
+    }
+
+    if ((size_t) co == cm->max) {
+	while (cm->max > WHITE && UNUSEDCOLOR(&cm->cd[cm->max])) {
+	    cm->max--;
+	}
+	assert(cm->free >= 0);
+	while ((size_t) cm->free > cm->max) {
+	    cm->free = cm->cd[cm->free].sub;
+	}
+	if (cm->free > 0) {
+	    assert((size_t)cm->free < cm->max);
+	    pco = cm->free;
+	    nco = cm->cd[pco].sub;
+	    while (nco > 0) {
+		if ((size_t) nco > cm->max) {
+		    /*
+		     * Take this one out of freelist.
+		     */
+
+		    nco = cm->cd[nco].sub;
+		    cm->cd[pco].sub = nco;
+		} else {
+		    assert((size_t)nco < cm->max);
+		    pco = nco;
+		    nco = cm->cd[pco].sub;
+		}
+	    }
+	}
+    } else {
+	cd->sub = cm->free;
+	cm->free = (color) (cd - cm->cd);
+    }
+}
+
+/*
+ - pseudocolor - allocate a false color, to be managed by other means
+ ^ static color pseudocolor(struct colormap *);
+ */
+static color
+pseudocolor(
+    struct colormap *cm)
+{
+    color co;
+
+    co = newcolor(cm);
+    if (CISERR()) {
+	return COLORLESS;
+    }
+    cm->cd[co].nchrs = 1;
+    cm->cd[co].flags = PSEUDO;
+    return co;
+}
+
+/*
+ - subcolor - allocate a new subcolor (if necessary) to this chr
+ ^ static color subcolor(struct colormap *, pchr c);
+ */
+static color
+subcolor(
+    struct colormap *cm,
+    pchr c)
+{
+    color co;			/* current color of c */
+    color sco;			/* new subcolor */
+
+    co = GETCOLOR(cm, c);
+    sco = newsub(cm, co);
+    if (CISERR()) {
+	return COLORLESS;
+    }
+    assert(sco != COLORLESS);
+
+    if (co == sco) {		/* already in an open subcolor */
+	return co;		/* rest is redundant */
+    }
+    cm->cd[co].nchrs--;
+    cm->cd[sco].nchrs++;
+    setcolor(cm, c, sco);
+    return sco;
+}
+
+/*
+ - newsub - allocate a new subcolor (if necessary) for a color
+ ^ static color newsub(struct colormap *, pcolor);
+ */
+static color
+newsub(
+    struct colormap *cm,
+    pcolor co)
+{
+    color sco;			/* new subcolor */
+
+    sco = cm->cd[co].sub;
+    if (sco == NOSUB) {		/* color has no open subcolor */
+	if (cm->cd[co].nchrs == 1) {	/* optimization */
+	    return co;
+	}
+	sco = newcolor(cm);	/* must create subcolor */
+	if (sco == COLORLESS) {
+	    assert(CISERR());
+	    return COLORLESS;
+	}
+	cm->cd[co].sub = sco;
+	cm->cd[sco].sub = sco;	/* open subcolor points to self */
+    }
+    assert(sco != NOSUB);
+
+    return sco;
+}
+
+/*
+ - subrange - allocate new subcolors to this range of chrs, fill in arcs
+ ^ static void subrange(struct vars *, pchr, pchr, struct state *,
+ ^ 	struct state *);
+ */
+static void
+subrange(
+    struct vars *v,
+    pchr from,
+    pchr to,
+    struct state *lp,
+    struct state *rp)
+{
+    uchr uf;
+    int i;
+
+    assert(from <= to);
+
+    /*
+     * First, align "from" on a tree-block boundary
+     */
+
+    uf = (uchr) from;
+    i = (int) (((uf + BYTTAB - 1) & (uchr) ~BYTMASK) - uf);
+    for (; from<=to && i>0; i--, from++) {
+	newarc(v->nfa, PLAIN, subcolor(v->cm, from), lp, rp);
+    }
+    if (from > to) {		/* didn't reach a boundary */
+	return;
+    }
+
+    /*
+     * Deal with whole blocks.
+     */
+
+    for (; to-from>=BYTTAB ; from+=BYTTAB) {
+	subblock(v, from, lp, rp);
+    }
+
+    /*
+     * Clean up any remaining partial table.
+     */
+
+    for (; from<=to ; from++) {
+	newarc(v->nfa, PLAIN, subcolor(v->cm, from), lp, rp);
+    }
+}
+
+/*
+ - subblock - allocate new subcolors for one tree block of chrs, fill in arcs
+ ^ static void subblock(struct vars *, pchr, struct state *, struct state *);
+ */
+static void
+subblock(
+    struct vars *v,
+    pchr start,			/* first of BYTTAB chrs */
+    struct state *lp,
+    struct state *rp)
+{
+    uchr uc = start;
+    struct colormap *cm = v->cm;
+    int shift;
+    int level;
+    int i;
+    int b;
+    union tree *t;
+    union tree *cb;
+    union tree *fillt;
+    union tree *lastt;
+    int previ;
+    int ndone;
+    color co;
+    color sco;
+
+    assert((uc % BYTTAB) == 0);
+
+    /*
+     * Find its color block, making new pointer blocks as needed.
+     */
+
+    t = cm->tree;
+    fillt = NULL;
+    for (level=0, shift=BYTBITS*(NBYTS-1); shift>0; level++, shift-=BYTBITS) {
+	b = (uc >> shift) & BYTMASK;
+	lastt = t;
+	t = lastt->tptr[b];
+	assert(t != NULL);
+	fillt = &cm->tree[level+1];
+	if (t == fillt && shift > BYTBITS) {	/* need new ptr block */
+	    t = (union tree *) MALLOC(sizeof(struct ptrs));
+	    if (t == NULL) {
+		CERR(REG_ESPACE);
+		return;
+	    }
+	    memcpy(t->tptr, fillt->tptr, BYTTAB*sizeof(union tree *));
+	    lastt->tptr[b] = t;
+	}
+    }
+
+    /*
+     * Special cases: fill block or solid block.
+     */
+    co = t->tcolor[0];
+    cb = cm->cd[co].block;
+    if (t == fillt || t == cb) {
+	/*
+	 * Either way, we want a subcolor solid block.
+	 */
+
+	sco = newsub(cm, co);
+	t = cm->cd[sco].block;
+	if (t == NULL) {	/* must set it up */
+	    t = (union tree *) MALLOC(sizeof(struct colors));
+	    if (t == NULL) {
+		CERR(REG_ESPACE);
+		return;
+	    }
+	    for (i=0 ; i<BYTTAB ; i++) {
+		t->tcolor[i] = sco;
+	    }
+	    cm->cd[sco].block = t;
+	}
+
+	/*
+	 * Find loop must have run at least once.
+	 */
+
+	lastt->tptr[b] = t;
+	newarc(v->nfa, PLAIN, sco, lp, rp);
+	cm->cd[co].nchrs -= BYTTAB;
+	cm->cd[sco].nchrs += BYTTAB;
+	return;
+    }
+
+    /*
+     * General case, a mixed block to be altered.
+     */
+
+    i = 0;
+    while (i < BYTTAB) {
+	co = t->tcolor[i];
+	sco = newsub(cm, co);
+	newarc(v->nfa, PLAIN, sco, lp, rp);
+	previ = i;
+	do {
+	    t->tcolor[i++] = sco;
+	} while (i < BYTTAB && t->tcolor[i] == co);
+	ndone = i - previ;
+	cm->cd[co].nchrs -= ndone;
+	cm->cd[sco].nchrs += ndone;
+    }
+}
+
+/*
+ - okcolors - promote subcolors to full colors
+ ^ static void okcolors(struct nfa *, struct colormap *);
+ */
+static void
+okcolors(
+    struct nfa *nfa,
+    struct colormap *cm)
+{
+    struct colordesc *cd;
+    struct colordesc *end = CDEND(cm);
+    struct colordesc *scd;
+    struct arc *a;
+    color co;
+    color sco;
+
+    for (cd=cm->cd, co=0 ; cd<end ; cd++, co++) {
+	sco = cd->sub;
+	if (UNUSEDCOLOR(cd) || sco == NOSUB) {
+	    /*
+	     * Has no subcolor, no further action.
+	     */
+	} else if (sco == co) {
+	    /*
+	     * Is subcolor, let parent deal with it.
+	     */
+	} else if (cd->nchrs == 0) {
+	    /*
+	     * Parent empty, its arcs change color to subcolor.
+	     */
+
+	    cd->sub = NOSUB;
+	    scd = &cm->cd[sco];
+	    assert(scd->nchrs > 0);
+	    assert(scd->sub == sco);
+	    scd->sub = NOSUB;
+	    while ((a = cd->arcs) != NULL) {
+		assert(a->co == co);
+		uncolorchain(cm, a);
+		a->co = sco;
+		colorchain(cm, a);
+	    }
+	    freecolor(cm, co);
+	} else {
+	    /*
+	     * Parent's arcs must gain parallel subcolor arcs.
+	     */
+
+	    cd->sub = NOSUB;
+	    scd = &cm->cd[sco];
+	    assert(scd->nchrs > 0);
+	    assert(scd->sub == sco);
+	    scd->sub = NOSUB;
+	    for (a=cd->arcs ; a!=NULL ; a=a->colorchain) {
+		assert(a->co == co);
+		newarc(nfa, a->type, sco, a->from, a->to);
+	    }
+	}
+    }
+}
+
+/*
+ - colorchain - add this arc to the color chain of its color
+ ^ static void colorchain(struct colormap *, struct arc *);
+ */
+static void
+colorchain(
+    struct colormap *cm,
+    struct arc *a)
+{
+    struct colordesc *cd = &cm->cd[a->co];
+
+    if (cd->arcs != NULL) {
+	cd->arcs->colorchainRev = a;
+    }
+    a->colorchain = cd->arcs;
+    a->colorchainRev = NULL;
+    cd->arcs = a;
+}
+
+/*
+ - uncolorchain - delete this arc from the color chain of its color
+ ^ static void uncolorchain(struct colormap *, struct arc *);
+ */
+static void
+uncolorchain(
+    struct colormap *cm,
+    struct arc *a)
+{
+    struct colordesc *cd = &cm->cd[a->co];
+    struct arc *aa = a->colorchainRev;
+
+    if (aa == NULL) {
+	assert(cd->arcs == a);
+	cd->arcs = a->colorchain;
+    } else {
+	assert(aa->colorchain == a);
+	aa->colorchain = a->colorchain;
+    }
+    if (a->colorchain != NULL) {
+	a->colorchain->colorchainRev = aa;
+    }
+    a->colorchain = NULL;	/* paranoia */
+    a->colorchainRev = NULL;
+}
+
+/*
+ - rainbow - add arcs of all full colors (but one) between specified states
+ ^ static void rainbow(struct nfa *, struct colormap *, int, pcolor,
+ ^ 	struct state *, struct state *);
+ */
+static void
+rainbow(
+    struct nfa *nfa,
+    struct colormap *cm,
+    int type,
+    pcolor but,			/* COLORLESS if no exceptions */
+    struct state *from,
+    struct state *to)
+{
+    struct colordesc *cd;
+    struct colordesc *end = CDEND(cm);
+    color co;
+
+    for (cd=cm->cd, co=0 ; cd<end && !CISERR(); cd++, co++) {
+	if (!UNUSEDCOLOR(cd) && (cd->sub != co) && (co != but)
+		&& !(cd->flags&PSEUDO)) {
+	    newarc(nfa, type, co, from, to);
+	}
+    }
+}
+
+/*
+ - colorcomplement - add arcs of complementary colors
+ * The calling sequence ought to be reconciled with cloneouts().
+ ^ static void colorcomplement(struct nfa *, struct colormap *, int,
+ ^ 	struct state *, struct state *, struct state *);
+ */
+static void
+colorcomplement(
+    struct nfa *nfa,
+    struct colormap *cm,
+    int type,
+    struct state *of,		/* complements of this guy's PLAIN outarcs */
+    struct state *from,
+    struct state *to)
+{
+    struct colordesc *cd;
+    struct colordesc *end = CDEND(cm);
+    color co;
+
+    assert(of != from);
+    for (cd=cm->cd, co=0 ; cd<end && !CISERR() ; cd++, co++) {
+	if (!UNUSEDCOLOR(cd) && !(cd->flags&PSEUDO)) {
+	    if (findarc(of, PLAIN, co) == NULL) {
+		newarc(nfa, type, co, from, to);
+	    }
+	}
+    }
+}
+
+#ifdef REG_DEBUG
+/*
+ ^ #ifdef REG_DEBUG
+ */
+
+/*
+ - dumpcolors - debugging output
+ ^ static void dumpcolors(struct colormap *, FILE *);
+ */
+static void
+dumpcolors(
+    struct colormap *cm,
+    FILE *f)
+{
+    struct colordesc *cd;
+    struct colordesc *end;
+    color co;
+    chr c;
+    char *has;
+
+    fprintf(f, "max %ld\n", (long) cm->max);
+    if (NBYTS > 1) {
+	fillcheck(cm, cm->tree, 0, f);
+    }
+    end = CDEND(cm);
+    for (cd=cm->cd+1, co=1 ; cd<end ; cd++, co++) {	/* skip 0 */
+	if (!UNUSEDCOLOR(cd)) {
+	    assert(cd->nchrs > 0);
+	    has = (cd->block != NULL) ? "#" : "";
+	    if (cd->flags&PSEUDO) {
+		fprintf(f, "#%2ld%s(ps): ", (long) co, has);
+	    } else {
+		fprintf(f, "#%2ld%s(%2d): ", (long) co, has, cd->nchrs);
+	    }
+
+	    /*
+	     * Unfortunately, it's hard to do this next bit more efficiently.
+	     *
+	     * Spencer's original coding has the loop iterating from CHR_MIN
+	     * to CHR_MAX, but that's utterly unusable for 32-bit chr, or
+	     * even 16-bit.  For debugging purposes it seems fine to print
+	     * only chr codes up to 1000 or so.
+	     */
+
+	    for (c=CHR_MIN ; c<1000 ; c++) {
+		if (GETCOLOR(cm, c) == co) {
+		    dumpchr(c, f);
+		}
+	    }
+	    fprintf(f, "\n");
+	}
+    }
+}
+
+/*
+ - fillcheck - check proper filling of a tree
+ ^ static void fillcheck(struct colormap *, union tree *, int, FILE *);
+ */
+static void
+fillcheck(
+    struct colormap *cm,
+    union tree *tree,
+    int level,			/* level number (top == 0) of this block */
+    FILE *f)
+{
+    int i;
+    union tree *t;
+    union tree *fillt = &cm->tree[level+1];
+
+    assert(level < NBYTS-1);	/* this level has pointers */
+    for (i=BYTTAB-1 ; i>=0 ; i--) {
+	t = tree->tptr[i];
+	if (t == NULL) {
+	    fprintf(f, "NULL found in filled tree!\n");
+	} else if (t == fillt) {
+	    /* empty body */
+	} else if (level < NBYTS-2) {	/* more pointer blocks below */
+	    fillcheck(cm, t, level+1, f);
+	}
+    }
+}
+
+/*
+ - dumpchr - print a chr
+ * Kind of char-centric but works well enough for debug use.
+ ^ static void dumpchr(pchr, FILE *);
+ */
+static void
+dumpchr(
+    pchr c,
+    FILE *f)
+{
+    if (c == '\\') {
+	fprintf(f, "\\\\");
+    } else if (c > ' ' && c <= '~') {
+	putc((char) c, f);
+    } else {
+	fprintf(f, "\\u%04lx", (long) c);
+    }
+}
+
+/*
+ ^ #endif
+ */
+#endif				/* ifdef REG_DEBUG */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regc_cvec.c b/generic/regc_cvec.c
new file mode 100644
index 0000000..d450d3e
--- /dev/null
+++ b/generic/regc_cvec.c
@@ -0,0 +1,147 @@
+/*
+ * Utility functions for handling cvecs
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * Notes:
+ * Only (selected) functions in _this_ file should treat chr* as non-constant.
+ */
+
+/*
+ - newcvec - allocate a new cvec
+ ^ static struct cvec *newcvec(int, int);
+ */
+static struct cvec *
+newcvec(
+    int nchrs,			/* to hold this many chrs... */
+    int nranges)		/* ... and this many ranges... */
+{
+    size_t nc = (size_t)nchrs + (size_t)nranges*2;
+    size_t n = sizeof(struct cvec) + nc*sizeof(chr);
+    struct cvec *cv = (struct cvec *) MALLOC(n);
+
+    if (cv == NULL) {
+	return NULL;
+    }
+    cv->chrspace = nchrs;
+    cv->chrs = (chr *)(((char *)cv)+sizeof(struct cvec));
+    cv->ranges = cv->chrs + nchrs;
+    cv->rangespace = nranges;
+    return clearcvec(cv);
+}
+
+/*
+ - clearcvec - clear a possibly-new cvec
+ * Returns pointer as convenience.
+ ^ static struct cvec *clearcvec(struct cvec *);
+ */
+static struct cvec *
+clearcvec(
+    struct cvec *cv)		/* character vector */
+{
+    assert(cv != NULL);
+    cv->nchrs = 0;
+    cv->nranges = 0;
+    return cv;
+}
+
+/*
+ - addchr - add a chr to a cvec
+ ^ static void addchr(struct cvec *, pchr);
+ */
+static void
+addchr(
+    struct cvec *cv,		/* character vector */
+    pchr c)			/* character to add */
+{
+    assert(cv->nchrs < cv->chrspace);
+    cv->chrs[cv->nchrs++] = (chr)c;
+}
+
+/*
+ - addrange - add a range to a cvec
+ ^ static void addrange(struct cvec *, pchr, pchr);
+ */
+static void
+addrange(
+    struct cvec *cv,		/* character vector */
+    pchr from,			/* first character of range */
+    pchr to)			/* last character of range */
+{
+    assert(cv->nranges < cv->rangespace);
+    cv->ranges[cv->nranges*2] = (chr)from;
+    cv->ranges[cv->nranges*2 + 1] = (chr)to;
+    cv->nranges++;
+}
+
+/*
+ - getcvec - get a cvec, remembering it as v->cv
+ ^ static struct cvec *getcvec(struct vars *, int, int);
+ */
+static struct cvec *
+getcvec(
+    struct vars *v,		/* context */
+    int nchrs,			/* to hold this many chrs... */
+    int nranges)		/* ... and this many ranges... */
+{
+    if ((v->cv != NULL) && (nchrs <= v->cv->chrspace) &&
+	    (nranges <= v->cv->rangespace)) {
+	return clearcvec(v->cv);
+    }
+
+    if (v->cv != NULL) {
+	freecvec(v->cv);
+    }
+    v->cv = newcvec(nchrs, nranges);
+    if (v->cv == NULL) {
+	ERR(REG_ESPACE);
+    }
+
+    return v->cv;
+}
+
+/*
+ - freecvec - free a cvec
+ ^ static void freecvec(struct cvec *);
+ */
+static void
+freecvec(
+    struct cvec *cv)		/* character vector */
+{
+    FREE(cv);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regc_lex.c b/generic/regc_lex.c
new file mode 100644
index 0000000..4c8f15f
--- /dev/null
+++ b/generic/regc_lex.c
@@ -0,0 +1,1195 @@
+/*
+ * lexical analyzer
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/* scanning macros (know about v) */
+#define	ATEOS()		(v->now >= v->stop)
+#define	HAVE(n)		(v->stop - v->now >= (n))
+#define	NEXT1(c)	(!ATEOS() && *v->now == CHR(c))
+#define	NEXT2(a,b)	(HAVE(2) && *v->now == CHR(a) && *(v->now+1) == CHR(b))
+#define	NEXT3(a,b,c) \
+	(HAVE(3) && *v->now == CHR(a) && \
+		*(v->now+1) == CHR(b) && \
+		*(v->now+2) == CHR(c))
+#define	SET(c)		(v->nexttype = (c))
+#define	SETV(c, n)	(v->nexttype = (c), v->nextvalue = (n))
+#define	RET(c)		return (SET(c), 1)
+#define	RETV(c, n)	return (SETV(c, n), 1)
+#define	FAILW(e)	return (ERR(e), 0)	/* ERR does SET(EOS) */
+#define	LASTTYPE(t)	(v->lasttype == (t))
+
+/* lexical contexts */
+#define	L_ERE	1	/* mainline ERE/ARE */
+#define	L_BRE	2	/* mainline BRE */
+#define	L_Q	3	/* REG_QUOTE */
+#define	L_EBND	4	/* ERE/ARE bound */
+#define	L_BBND	5	/* BRE bound */
+#define	L_BRACK	6	/* brackets */
+#define	L_CEL	7	/* collating element */
+#define	L_ECL	8	/* equivalence class */
+#define	L_CCL	9	/* character class */
+#define	INTOCON(c)	(v->lexcon = (c))
+#define	INCON(con)	(v->lexcon == (con))
+
+/* construct pointer past end of chr array */
+#define	ENDOF(array)	((array) + sizeof(array)/sizeof(chr))
+
+/*
+ - lexstart - set up lexical stuff, scan leading options
+ ^ static void lexstart(struct vars *);
+ */
+static void
+lexstart(
+    struct vars *v)
+{
+    prefixes(v);		/* may turn on new type bits etc. */
+    NOERR();
+
+    if (v->cflags&REG_QUOTE) {
+	assert(!(v->cflags&(REG_ADVANCED|REG_EXPANDED|REG_NEWLINE)));
+	INTOCON(L_Q);
+    } else if (v->cflags&REG_EXTENDED) {
+	assert(!(v->cflags&REG_QUOTE));
+	INTOCON(L_ERE);
+    } else {
+	assert(!(v->cflags&(REG_QUOTE|REG_ADVF)));
+	INTOCON(L_BRE);
+    }
+
+    v->nexttype = EMPTY;	/* remember we were at the start */
+    next(v);			/* set up the first token */
+}
+
+/*
+ - prefixes - implement various special prefixes
+ ^ static void prefixes(struct vars *);
+ */
+static void
+prefixes(
+    struct vars *v)
+{
+    /*
+     * Literal string doesn't get any of this stuff.
+     */
+
+    if (v->cflags&REG_QUOTE) {
+	return;
+    }
+
+    /*
+     * Initial "***" gets special things.
+     */
+
+    if (HAVE(4) && NEXT3('*', '*', '*')) {
+	switch (*(v->now + 3)) {
+	case CHR('?'):		/* "***?" error, msg shows version */
+	    ERR(REG_BADPAT);
+	    return;		/* proceed no further */
+	    break;
+	case CHR('='):		/* "***=" shifts to literal string */
+	    NOTE(REG_UNONPOSIX);
+	    v->cflags |= REG_QUOTE;
+	    v->cflags &= ~(REG_ADVANCED|REG_EXPANDED|REG_NEWLINE);
+	    v->now += 4;
+	    return;		/* and there can be no more prefixes */
+	    break;
+	case CHR(':'):		/* "***:" shifts to AREs */
+	    NOTE(REG_UNONPOSIX);
+	    v->cflags |= REG_ADVANCED;
+	    v->now += 4;
+	    break;
+	default:		/* otherwise *** is just an error */
+	    ERR(REG_BADRPT);
+	    return;
+	    break;
+	}
+    }
+
+    /*
+     * BREs and EREs don't get embedded options.
+     */
+
+    if ((v->cflags&REG_ADVANCED) != REG_ADVANCED) {
+	return;
+    }
+
+    /*
+     * Embedded options (AREs only).
+     */
+
+    if (HAVE(3) && NEXT2('(', '?') && iscalpha(*(v->now + 2))) {
+	NOTE(REG_UNONPOSIX);
+	v->now += 2;
+	for (; !ATEOS() && iscalpha(*v->now); v->now++) {
+	    switch (*v->now) {
+	    case CHR('b'):	/* BREs (but why???) */
+		v->cflags &= ~(REG_ADVANCED|REG_QUOTE);
+		break;
+	    case CHR('c'):	/* case sensitive */
+		v->cflags &= ~REG_ICASE;
+		break;
+	    case CHR('e'):	/* plain EREs */
+		v->cflags |= REG_EXTENDED;
+		v->cflags &= ~(REG_ADVF|REG_QUOTE);
+		break;
+	    case CHR('i'):	/* case insensitive */
+		v->cflags |= REG_ICASE;
+		break;
+	    case CHR('m'):	/* Perloid synonym for n */
+	    case CHR('n'):	/* \n affects ^ $ . [^ */
+		v->cflags |= REG_NEWLINE;
+		break;
+	    case CHR('p'):	/* ~Perl, \n affects . [^ */
+		v->cflags |= REG_NLSTOP;
+		v->cflags &= ~REG_NLANCH;
+		break;
+	    case CHR('q'):	/* literal string */
+		v->cflags |= REG_QUOTE;
+		v->cflags &= ~REG_ADVANCED;
+		break;
+	    case CHR('s'):	/* single line, \n ordinary */
+		v->cflags &= ~REG_NEWLINE;
+		break;
+	    case CHR('t'):	/* tight syntax */
+		v->cflags &= ~REG_EXPANDED;
+		break;
+	    case CHR('w'):	/* weird, \n affects ^ $ only */
+		v->cflags &= ~REG_NLSTOP;
+		v->cflags |= REG_NLANCH;
+		break;
+	    case CHR('x'):	/* expanded syntax */
+		v->cflags |= REG_EXPANDED;
+		break;
+	    default:
+		ERR(REG_BADOPT);
+		return;
+	    }
+	}
+	if (!NEXT1(')')) {
+	    ERR(REG_BADOPT);
+	    return;
+	}
+	v->now++;
+	if (v->cflags&REG_QUOTE) {
+	    v->cflags &= ~(REG_EXPANDED|REG_NEWLINE);
+	}
+    }
+}
+
+/*
+ - lexnest - "call a subroutine", interpolating string at the lexical level
+ * Note, this is not a very general facility.  There are a number of
+ * implicit assumptions about what sorts of strings can be subroutines.
+ ^ static void lexnest(struct vars *, const chr *, const chr *);
+ */
+static void
+lexnest(
+    struct vars *v,
+    const chr *beginp,		/* start of interpolation */
+    const chr *endp)		/* one past end of interpolation */
+{
+    assert(v->savenow == NULL);	/* only one level of nesting */
+    v->savenow = v->now;
+    v->savestop = v->stop;
+    v->now = beginp;
+    v->stop = endp;
+}
+
+/*
+ * string constants to interpolate as expansions of things like \d
+ */
+
+static const chr backd[] = {	/* \d */
+    CHR('['), CHR('['), CHR(':'),
+    CHR('d'), CHR('i'), CHR('g'), CHR('i'), CHR('t'),
+    CHR(':'), CHR(']'), CHR(']')
+};
+static const chr backD[] = {	/* \D */
+    CHR('['), CHR('^'), CHR('['), CHR(':'),
+    CHR('d'), CHR('i'), CHR('g'), CHR('i'), CHR('t'),
+    CHR(':'), CHR(']'), CHR(']')
+};
+static const chr brbackd[] = {	/* \d within brackets */
+    CHR('['), CHR(':'),
+    CHR('d'), CHR('i'), CHR('g'), CHR('i'), CHR('t'),
+    CHR(':'), CHR(']')
+};
+static const chr backs[] = {	/* \s */
+    CHR('['), CHR('['), CHR(':'),
+    CHR('s'), CHR('p'), CHR('a'), CHR('c'), CHR('e'),
+    CHR(':'), CHR(']'), CHR(']')
+};
+static const chr backS[] = {	/* \S */
+    CHR('['), CHR('^'), CHR('['), CHR(':'),
+    CHR('s'), CHR('p'), CHR('a'), CHR('c'), CHR('e'),
+    CHR(':'), CHR(']'), CHR(']')
+};
+static const chr brbacks[] = {	/* \s within brackets */
+    CHR('['), CHR(':'),
+    CHR('s'), CHR('p'), CHR('a'), CHR('c'), CHR('e'),
+    CHR(':'), CHR(']')
+};
+
+#define PUNCT_CONN \
+	CHR('_'), \
+	0x203f /* UNDERTIE */, \
+	0x2040 /* CHARACTER TIE */,\
+	0x2054 /* INVERTED UNDERTIE */,\
+	0xfe33 /* PRESENTATION FORM FOR VERTICAL LOW LINE */, \
+	0xfe34 /* PRESENTATION FORM FOR VERTICAL WAVY LOW LINE */, \
+	0xfe4d /* DASHED LOW LINE */, \
+	0xfe4e /* CENTRELINE LOW LINE */, \
+	0xfe4f /* WAVY LOW LINE */, \
+	0xff3f /* FULLWIDTH LOW LINE */
+
+static const chr backw[] = {	/* \w */
+    CHR('['), CHR('['), CHR(':'),
+    CHR('a'), CHR('l'), CHR('n'), CHR('u'), CHR('m'),
+    CHR(':'), CHR(']'), PUNCT_CONN, CHR(']')
+};
+static const chr backW[] = {	/* \W */
+    CHR('['), CHR('^'), CHR('['), CHR(':'),
+    CHR('a'), CHR('l'), CHR('n'), CHR('u'), CHR('m'),
+    CHR(':'), CHR(']'), PUNCT_CONN, CHR(']')
+};
+static const chr brbackw[] = {	/* \w within brackets */
+    CHR('['), CHR(':'),
+    CHR('a'), CHR('l'), CHR('n'), CHR('u'), CHR('m'),
+    CHR(':'), CHR(']'), PUNCT_CONN
+};
+
+/*
+ - lexword - interpolate a bracket expression for word characters
+ * Possibly ought to inquire whether there is a "word" character class.
+ ^ static void lexword(struct vars *);
+ */
+static void
+lexword(
+    struct vars *v)
+{
+    lexnest(v, backw, ENDOF(backw));
+}
+
+/*
+ - next - get next token
+ ^ static int next(struct vars *);
+ */
+static int			/* 1 normal, 0 failure */
+next(
+    struct vars *v)
+{
+    chr c;
+
+    /*
+     * Errors yield an infinite sequence of failures.
+     */
+
+    if (ISERR()) {
+	return 0;		/* the error has set nexttype to EOS */
+    }
+
+    /*
+     * Remember flavor of last token.
+     */
+
+    v->lasttype = v->nexttype;
+
+    /*
+     * REG_BOSONLY
+     */
+
+    if (v->nexttype == EMPTY && (v->cflags&REG_BOSONLY)) {
+	/* at start of a REG_BOSONLY RE */
+	RETV(SBEGIN, 0);	/* same as \A */
+    }
+
+    /*
+     * If we're nested and we've hit end, return to outer level.
+     */
+
+    if (v->savenow != NULL && ATEOS()) {
+	v->now = v->savenow;
+	v->stop = v->savestop;
+	v->savenow = v->savestop = NULL;
+    }
+
+    /*
+     * Skip white space etc. if appropriate (not in literal or [])
+     */
+
+    if (v->cflags&REG_EXPANDED) {
+	switch (v->lexcon) {
+	case L_ERE:
+	case L_BRE:
+	case L_EBND:
+	case L_BBND:
+	    skip(v);
+	    break;
+	}
+    }
+
+    /*
+     * Handle EOS, depending on context.
+     */
+
+    if (ATEOS()) {
+	switch (v->lexcon) {
+	case L_ERE:
+	case L_BRE:
+	case L_Q:
+	    RET(EOS);
+	    break;
+	case L_EBND:
+	case L_BBND:
+	    FAILW(REG_EBRACE);
+	    break;
+	case L_BRACK:
+	case L_CEL:
+	case L_ECL:
+	case L_CCL:
+	    FAILW(REG_EBRACK);
+	    break;
+	}
+	assert(NOTREACHED);
+    }
+
+    /*
+     * Okay, time to actually get a character.
+     */
+
+    c = *v->now++;
+
+    /*
+     * Deal with the easy contexts, punt EREs to code below.
+     */
+
+    switch (v->lexcon) {
+    case L_BRE:			/* punt BREs to separate function */
+	return brenext(v, c);
+	break;
+    case L_ERE:			/* see below */
+	break;
+    case L_Q:			/* literal strings are easy */
+	RETV(PLAIN, c);
+	break;
+    case L_BBND:		/* bounds are fairly simple */
+    case L_EBND:
+	switch (c) {
+	case CHR('0'): case CHR('1'): case CHR('2'): case CHR('3'):
+	case CHR('4'): case CHR('5'): case CHR('6'): case CHR('7'):
+	case CHR('8'): case CHR('9'):
+	    RETV(DIGIT, (chr)DIGITVAL(c));
+	    break;
+	case CHR(','):
+	    RET(',');
+	    break;
+	case CHR('}'):		/* ERE bound ends with } */
+	    if (INCON(L_EBND)) {
+		INTOCON(L_ERE);
+		if ((v->cflags&REG_ADVF) && NEXT1('?')) {
+		    v->now++;
+		    NOTE(REG_UNONPOSIX);
+		    RETV('}', 0);
+		}
+		RETV('}', 1);
+	    } else {
+		FAILW(REG_BADBR);
+	    }
+	    break;
+	case CHR('\\'):		/* BRE bound ends with \} */
+	    if (INCON(L_BBND) && NEXT1('}')) {
+		v->now++;
+		INTOCON(L_BRE);
+		RET('}');
+	    } else {
+		FAILW(REG_BADBR);
+	    }
+	    break;
+	default:
+	    FAILW(REG_BADBR);
+	    break;
+	}
+	assert(NOTREACHED);
+	break;
+    case L_BRACK:		/* brackets are not too hard */
+	switch (c) {
+	case CHR(']'):
+	    if (LASTTYPE('[')) {
+		RETV(PLAIN, c);
+	    } else {
+		INTOCON((v->cflags&REG_EXTENDED) ? L_ERE : L_BRE);
+		RET(']');
+	    }
+	    break;
+	case CHR('\\'):
+	    NOTE(REG_UBBS);
+	    if (!(v->cflags&REG_ADVF)) {
+		RETV(PLAIN, c);
+	    }
+	    NOTE(REG_UNONPOSIX);
+	    if (ATEOS()) {
+		FAILW(REG_EESCAPE);
+	    }
+	    (void)lexescape(v);
+	    switch (v->nexttype) {	/* not all escapes okay here */
+	    case PLAIN:
+		return 1;
+		break;
+	    case CCLASS:
+		switch (v->nextvalue) {
+		case 'd':
+		    lexnest(v, brbackd, ENDOF(brbackd));
+		    break;
+		case 's':
+		    lexnest(v, brbacks, ENDOF(brbacks));
+		    break;
+		case 'w':
+		    lexnest(v, brbackw, ENDOF(brbackw));
+		    break;
+		default:
+		    FAILW(REG_EESCAPE);
+		    break;
+		}
+
+		/*
+		 * lexnest() done, back up and try again.
+		 */
+
+		v->nexttype = v->lasttype;
+		return next(v);
+		break;
+	    }
+
+	    /*
+	     * Not one of the acceptable escapes.
+	     */
+
+	    FAILW(REG_EESCAPE);
+	    break;
+	case CHR('-'):
+	    if (LASTTYPE('[') || NEXT1(']')) {
+		RETV(PLAIN, c);
+	    } else {
+		RETV(RANGE, c);
+	    }
+	    break;
+	case CHR('['):
+	    if (ATEOS()) {
+		FAILW(REG_EBRACK);
+	    }
+	    switch (*v->now++) {
+	    case CHR('.'):
+		INTOCON(L_CEL);
+
+		/*
+		 * Might or might not be locale-specific.
+		 */
+
+		RET(COLLEL);
+		break;
+	    case CHR('='):
+		INTOCON(L_ECL);
+		NOTE(REG_ULOCALE);
+		RET(ECLASS);
+		break;
+	    case CHR(':'):
+		INTOCON(L_CCL);
+		NOTE(REG_ULOCALE);
+		RET(CCLASS);
+		break;
+	    default:		/* oops */
+		v->now--;
+		RETV(PLAIN, c);
+		break;
+	    }
+	    assert(NOTREACHED);
+	    break;
+	default:
+	    RETV(PLAIN, c);
+	    break;
+	}
+	assert(NOTREACHED);
+	break;
+    case L_CEL:			/* collating elements are easy */
+	if (c == CHR('.') && NEXT1(']')) {
+	    v->now++;
+	    INTOCON(L_BRACK);
+	    RETV(END, '.');
+	} else {
+	    RETV(PLAIN, c);
+	}
+	break;
+    case L_ECL:			/* ditto equivalence classes */
+	if (c == CHR('=') && NEXT1(']')) {
+	    v->now++;
+	    INTOCON(L_BRACK);
+	    RETV(END, '=');
+	} else {
+	    RETV(PLAIN, c);
+	}
+	break;
+    case L_CCL:			/* ditto character classes */
+	if (c == CHR(':') && NEXT1(']')) {
+	    v->now++;
+	    INTOCON(L_BRACK);
+	    RETV(END, ':');
+	} else {
+	    RETV(PLAIN, c);
+	}
+	break;
+    default:
+	assert(NOTREACHED);
+	break;
+    }
+
+    /*
+     * That got rid of everything except EREs and AREs.
+     */
+
+    assert(INCON(L_ERE));
+
+    /*
+     * Deal with EREs and AREs, except for backslashes.
+     */
+
+    switch (c) {
+    case CHR('|'):
+	RET('|');
+	break;
+    case CHR('*'):
+	if ((v->cflags&REG_ADVF) && NEXT1('?')) {
+	    v->now++;
+	    NOTE(REG_UNONPOSIX);
+	    RETV('*', 0);
+	}
+	RETV('*', 1);
+	break;
+    case CHR('+'):
+	if ((v->cflags&REG_ADVF) && NEXT1('?')) {
+	    v->now++;
+	    NOTE(REG_UNONPOSIX);
+	    RETV('+', 0);
+	}
+	RETV('+', 1);
+	break;
+    case CHR('?'):
+	if ((v->cflags&REG_ADVF) && NEXT1('?')) {
+	    v->now++;
+	    NOTE(REG_UNONPOSIX);
+	    RETV('?', 0);
+	}
+	RETV('?', 1);
+	break;
+    case CHR('{'):		/* bounds start or plain character */
+	if (v->cflags&REG_EXPANDED) {
+	    skip(v);
+	}
+	if (ATEOS() || !iscdigit(*v->now)) {
+	    NOTE(REG_UBRACES);
+	    NOTE(REG_UUNSPEC);
+	    RETV(PLAIN, c);
+	} else {
+	    NOTE(REG_UBOUNDS);
+	    INTOCON(L_EBND);
+	    RET('{');
+	}
+	assert(NOTREACHED);
+	break;
+    case CHR('('):		/* parenthesis, or advanced extension */
+	if ((v->cflags&REG_ADVF) && NEXT1('?')) {
+	    NOTE(REG_UNONPOSIX);
+	    v->now++;
+	    switch (*v->now++) {
+	    case CHR(':'):	/* non-capturing paren */
+		RETV('(', 0);
+		break;
+	    case CHR('#'):	/* comment */
+		while (!ATEOS() && *v->now != CHR(')')) {
+		    v->now++;
+		}
+		if (!ATEOS()) {
+		    v->now++;
+		}
+		assert(v->nexttype == v->lasttype);
+		return next(v);
+		break;
+	    case CHR('='):	/* positive lookahead */
+		NOTE(REG_ULOOKAHEAD);
+		RETV(LACON, 1);
+		break;
+	    case CHR('!'):	/* negative lookahead */
+		NOTE(REG_ULOOKAHEAD);
+		RETV(LACON, 0);
+		break;
+	    default:
+		FAILW(REG_BADRPT);
+		break;
+	    }
+	    assert(NOTREACHED);
+	}
+	if (v->cflags&REG_NOSUB) {
+	    RETV('(', 0);	/* all parens non-capturing */
+	} else {
+	    RETV('(', 1);
+	}
+	break;
+    case CHR(')'):
+	if (LASTTYPE('(')) {
+	    NOTE(REG_UUNSPEC);
+	}
+	RETV(')', c);
+	break;
+    case CHR('['):		/* easy except for [[:<:]] and [[:>:]] */
+	if (HAVE(6) &&	*(v->now+0) == CHR('[') &&
+		*(v->now+1) == CHR(':') &&
+		(*(v->now+2) == CHR('<') || *(v->now+2) == CHR('>')) &&
+		*(v->now+3) == CHR(':') &&
+		*(v->now+4) == CHR(']') &&
+		*(v->now+5) == CHR(']')) {
+	    c = *(v->now+2);
+	    v->now += 6;
+	    NOTE(REG_UNONPOSIX);
+	    RET((c == CHR('<')) ? '<' : '>');
+	}
+	INTOCON(L_BRACK);
+	if (NEXT1('^')) {
+	    v->now++;
+	    RETV('[', 0);
+	}
+	RETV('[', 1);
+	break;
+    case CHR('.'):
+	RET('.');
+	break;
+    case CHR('^'):
+	RET('^');
+	break;
+    case CHR('$'):
+	RET('$');
+	break;
+    case CHR('\\'):		/* mostly punt backslashes to code below */
+	if (ATEOS()) {
+	    FAILW(REG_EESCAPE);
+	}
+	break;
+    default:		/* ordinary character */
+	RETV(PLAIN, c);
+	break;
+    }
+
+    /*
+     * ERE/ARE backslash handling; backslash already eaten.
+     */
+
+    assert(!ATEOS());
+    if (!(v->cflags&REG_ADVF)) {/* only AREs have non-trivial escapes */
+	if (iscalnum(*v->now)) {
+	    NOTE(REG_UBSALNUM);
+	    NOTE(REG_UUNSPEC);
+	}
+	RETV(PLAIN, *v->now++);
+    }
+    (void)lexescape(v);
+    if (ISERR()) {
+	FAILW(REG_EESCAPE);
+    }
+    if (v->nexttype == CCLASS) {/* fudge at lexical level */
+	switch (v->nextvalue) {
+	case 'd':	lexnest(v, backd, ENDOF(backd)); break;
+	case 'D':	lexnest(v, backD, ENDOF(backD)); break;
+	case 's':	lexnest(v, backs, ENDOF(backs)); break;
+	case 'S':	lexnest(v, backS, ENDOF(backS)); break;
+	case 'w':	lexnest(v, backw, ENDOF(backw)); break;
+	case 'W':	lexnest(v, backW, ENDOF(backW)); break;
+	default:
+	    assert(NOTREACHED);
+	    FAILW(REG_ASSERT);
+	    break;
+	}
+	/* lexnest done, back up and try again */
+	v->nexttype = v->lasttype;
+	return next(v);
+    }
+
+    /*
+     * Otherwise, lexescape has already done the work.
+     */
+
+    return !ISERR();
+}
+
+/*
+ - lexescape - parse an ARE backslash escape (backslash already eaten)
+ * Note slightly nonstandard use of the CCLASS type code.
+ ^ static int lexescape(struct vars *);
+ */
+static int			/* not actually used, but convenient for RETV */
+lexescape(
+    struct vars *v)
+{
+    chr c;
+    int i;
+    static const chr alert[] = {
+	CHR('a'), CHR('l'), CHR('e'), CHR('r'), CHR('t')
+    };
+    static const chr esc[] = {
+	CHR('E'), CHR('S'), CHR('C')
+    };
+    const chr *save;
+
+    assert(v->cflags&REG_ADVF);
+
+    assert(!ATEOS());
+    c = *v->now++;
+    if (!iscalnum(c)) {
+	RETV(PLAIN, c);
+    }
+
+    NOTE(REG_UNONPOSIX);
+    switch (c) {
+    case CHR('a'):
+	RETV(PLAIN, chrnamed(v, alert, ENDOF(alert), CHR('\007')));
+	break;
+    case CHR('A'):
+	RETV(SBEGIN, 0);
+	break;
+    case CHR('b'):
+	RETV(PLAIN, CHR('\b'));
+	break;
+    case CHR('B'):
+	RETV(PLAIN, CHR('\\'));
+	break;
+    case CHR('c'):
+	NOTE(REG_UUNPORT);
+	if (ATEOS()) {
+	    FAILW(REG_EESCAPE);
+	}
+	RETV(PLAIN, (chr)(*v->now++ & 037));
+	break;
+    case CHR('d'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 'd');
+	break;
+    case CHR('D'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 'D');
+	break;
+    case CHR('e'):
+	NOTE(REG_UUNPORT);
+	RETV(PLAIN, chrnamed(v, esc, ENDOF(esc), CHR('\033')));
+	break;
+    case CHR('f'):
+	RETV(PLAIN, CHR('\f'));
+	break;
+    case CHR('m'):
+	RET('<');
+	break;
+    case CHR('M'):
+	RET('>');
+	break;
+    case CHR('n'):
+	RETV(PLAIN, CHR('\n'));
+	break;
+    case CHR('r'):
+	RETV(PLAIN, CHR('\r'));
+	break;
+    case CHR('s'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 's');
+	break;
+    case CHR('S'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 'S');
+	break;
+    case CHR('t'):
+	RETV(PLAIN, CHR('\t'));
+	break;
+    case CHR('u'):
+	c = (uchr) lexdigits(v, 16, 1, 4);
+	if (ISERR()) {
+	    FAILW(REG_EESCAPE);
+	}
+	RETV(PLAIN, c);
+	break;
+    case CHR('U'):
+	i = lexdigits(v, 16, 1, 8);
+	if (ISERR()) {
+	    FAILW(REG_EESCAPE);
+	}
+	if (i > 0xFFFF) {
+	    /* TODO: output a Surrogate pair
+	     */
+	    i = 0xFFFD;
+	}
+	RETV(PLAIN, (uchr) i);
+	break;
+    case CHR('v'):
+	RETV(PLAIN, CHR('\v'));
+	break;
+    case CHR('w'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 'w');
+	break;
+    case CHR('W'):
+	NOTE(REG_ULOCALE);
+	RETV(CCLASS, 'W');
+	break;
+    case CHR('x'):
+	NOTE(REG_UUNPORT);
+	c = (uchr) lexdigits(v, 16, 1, 2);
+	if (ISERR()) {
+	    FAILW(REG_EESCAPE);
+	}
+	RETV(PLAIN, c);
+	break;
+    case CHR('y'):
+	NOTE(REG_ULOCALE);
+	RETV(WBDRY, 0);
+	break;
+    case CHR('Y'):
+	NOTE(REG_ULOCALE);
+	RETV(NWBDRY, 0);
+	break;
+    case CHR('Z'):
+	RETV(SEND, 0);
+	break;
+    case CHR('1'): case CHR('2'): case CHR('3'): case CHR('4'):
+    case CHR('5'): case CHR('6'): case CHR('7'): case CHR('8'):
+    case CHR('9'):
+	save = v->now;
+	v->now--;		/* put first digit back */
+	c = (uchr) lexdigits(v, 10, 1, 255);	/* REs >255 long outside spec */
+	if (ISERR()) {
+	    FAILW(REG_EESCAPE);
+	}
+
+	/*
+	 * Ugly heuristic (first test is "exactly 1 digit?")
+	 */
+
+	if (v->now - save == 0 || ((int) c > 0 && (int)c <= v->nsubexp)) {
+	    NOTE(REG_UBACKREF);
+	    RETV(BACKREF, (chr)c);
+	}
+
+	/*
+	 * Oops, doesn't look like it's a backref after all...
+	 */
+
+	v->now = save;
+
+	/*
+	 * And fall through into octal number.
+	 */
+
+    case CHR('0'):
+	NOTE(REG_UUNPORT);
+	v->now--;		/* put first digit back */
+	c = (uchr) lexdigits(v, 8, 1, 3);
+	if (ISERR()) {
+	    FAILW(REG_EESCAPE);
+	}
+	if (c > 0xff) {
+	    /* out of range, so we handled one digit too much */
+	    v->now--;
+	    c >>= 3;
+	}
+	RETV(PLAIN, c);
+	break;
+    default:
+	assert(iscalpha(c));
+	FAILW(REG_EESCAPE);	/* unknown alphabetic escape */
+	break;
+    }
+    assert(NOTREACHED);
+}
+
+/*
+ - lexdigits - slurp up digits and return chr value
+ ^ static int lexdigits(struct vars *, int, int, int);
+ */
+static int			/* chr value; errors signalled via ERR */
+lexdigits(
+    struct vars *v,
+    int base,
+    int minlen,
+    int maxlen)
+{
+    int n;
+    int len;
+    chr c;
+    int d;
+    const uchr ub = (uchr) base;
+
+    n = 0;
+    for (len = 0; len < maxlen && !ATEOS(); len++) {
+	if (n > 0x10fff) {
+	    /* Stop when continuing would otherwise overflow */
+	    break;
+	}
+	c = *v->now++;
+	switch (c) {
+	case CHR('0'): case CHR('1'): case CHR('2'): case CHR('3'):
+	case CHR('4'): case CHR('5'): case CHR('6'): case CHR('7'):
+	case CHR('8'): case CHR('9'):
+	    d = DIGITVAL(c);
+	    break;
+	case CHR('a'): case CHR('A'): d = 10; break;
+	case CHR('b'): case CHR('B'): d = 11; break;
+	case CHR('c'): case CHR('C'): d = 12; break;
+	case CHR('d'): case CHR('D'): d = 13; break;
+	case CHR('e'): case CHR('E'): d = 14; break;
+	case CHR('f'): case CHR('F'): d = 15; break;
+	default:
+	    v->now--;		/* oops, not a digit at all */
+	    d = -1;
+	    break;
+	}
+
+	if (d >= base) {	/* not a plausible digit */
+	    v->now--;
+	    d = -1;
+	}
+	if (d < 0) {
+	    break;		/* NOTE BREAK OUT */
+	}
+	n = n*ub + (uchr)d;
+    }
+    if (len < minlen) {
+	ERR(REG_EESCAPE);
+    }
+
+    return n;
+}
+
+/*
+ - brenext - get next BRE token
+ * This is much like EREs except for all the stupid backslashes and the
+ * context-dependency of some things.
+ ^ static int brenext(struct vars *, pchr);
+ */
+static int			/* 1 normal, 0 failure */
+brenext(
+    struct vars *v,
+    pchr pc)
+{
+    chr c = (chr)pc;
+
+    switch (c) {
+    case CHR('*'):
+	if (LASTTYPE(EMPTY) || LASTTYPE('(') || LASTTYPE('^')) {
+	    RETV(PLAIN, c);
+	}
+	RET('*');
+	break;
+    case CHR('['):
+	if (HAVE(6) &&	*(v->now+0) == CHR('[') &&
+		*(v->now+1) == CHR(':') &&
+		(*(v->now+2) == CHR('<') || *(v->now+2) == CHR('>')) &&
+		*(v->now+3) == CHR(':') &&
+		*(v->now+4) == CHR(']') &&
+		*(v->now+5) == CHR(']')) {
+	    c = *(v->now+2);
+	    v->now += 6;
+	    NOTE(REG_UNONPOSIX);
+	    RET((c == CHR('<')) ? '<' : '>');
+	}
+	INTOCON(L_BRACK);
+	if (NEXT1('^')) {
+	    v->now++;
+	    RETV('[', 0);
+	}
+	RETV('[', 1);
+	break;
+    case CHR('.'):
+	RET('.');
+	break;
+    case CHR('^'):
+	if (LASTTYPE(EMPTY)) {
+	    RET('^');
+	}
+	if (LASTTYPE('(')) {
+	    NOTE(REG_UUNSPEC);
+	    RET('^');
+	}
+	RETV(PLAIN, c);
+	break;
+    case CHR('$'):
+	if (v->cflags&REG_EXPANDED) {
+	    skip(v);
+	}
+	if (ATEOS()) {
+	    RET('$');
+	}
+	if (NEXT2('\\', ')')) {
+	    NOTE(REG_UUNSPEC);
+	    RET('$');
+	}
+	RETV(PLAIN, c);
+	break;
+    case CHR('\\'):
+	break;			/* see below */
+    default:
+	RETV(PLAIN, c);
+	break;
+    }
+
+    assert(c == CHR('\\'));
+
+    if (ATEOS()) {
+	FAILW(REG_EESCAPE);
+    }
+
+    c = *v->now++;
+    switch (c) {
+    case CHR('{'):
+	INTOCON(L_BBND);
+	NOTE(REG_UBOUNDS);
+	RET('{');
+	break;
+    case CHR('('):
+	RETV('(', 1);
+	break;
+    case CHR(')'):
+	RETV(')', c);
+	break;
+    case CHR('<'):
+	NOTE(REG_UNONPOSIX);
+	RET('<');
+	break;
+    case CHR('>'):
+	NOTE(REG_UNONPOSIX);
+	RET('>');
+	break;
+    case CHR('1'): case CHR('2'): case CHR('3'): case CHR('4'):
+    case CHR('5'): case CHR('6'): case CHR('7'): case CHR('8'):
+    case CHR('9'):
+	NOTE(REG_UBACKREF);
+	RETV(BACKREF, (chr)DIGITVAL(c));
+	break;
+    default:
+	if (iscalnum(c)) {
+	    NOTE(REG_UBSALNUM);
+	    NOTE(REG_UUNSPEC);
+	}
+	RETV(PLAIN, c);
+	break;
+    }
+
+    assert(NOTREACHED);
+}
+
+/*
+ - skip - skip white space and comments in expanded form
+ ^ static void skip(struct vars *);
+ */
+static void
+skip(
+    struct vars *v)
+{
+    const chr *start = v->now;
+
+    assert(v->cflags&REG_EXPANDED);
+
+    for (;;) {
+	while (!ATEOS() && iscspace(*v->now)) {
+	    v->now++;
+	}
+	if (ATEOS() || *v->now != CHR('#')) {
+	    break;		/* NOTE BREAK OUT */
+	}
+	assert(NEXT1('#'));
+	while (!ATEOS() && *v->now != CHR('\n')) {
+	    v->now++;
+	}
+
+	/*
+	 * Leave the newline to be picked up by the iscspace loop.
+	 */
+    }
+
+    if (v->now != start) {
+	NOTE(REG_UNONPOSIX);
+    }
+}
+
+/*
+ - newline - return the chr for a newline
+ * This helps confine use of CHR to this source file.
+ ^ static chr newline(void);
+ */
+static chr
+newline(void)
+{
+    return CHR('\n');
+}
+
+/*
+ - chrnamed - return the chr known by a given (chr string) name
+ * The code is a bit clumsy, but this routine gets only such specialized
+ * use that it hardly matters.
+ ^ static chr chrnamed(struct vars *, const chr *, const chr *, pchr);
+ */
+static chr
+chrnamed(
+    struct vars *v,
+    const chr *startp,		/* start of name */
+    const chr *endp,		/* just past end of name */
+    pchr lastresort)		/* what to return if name lookup fails */
+{
+    celt c;
+    int errsave;
+    int e;
+    struct cvec *cv;
+
+    errsave = v->err;
+    v->err = 0;
+    c = element(v, startp, endp);
+    e = v->err;
+    v->err = errsave;
+
+    if (e != 0) {
+	return (chr)lastresort;
+    }
+
+    cv = range(v, c, c, 0);
+    if (cv->nchrs == 0) {
+	return (chr)lastresort;
+    }
+    return cv->chrs[0];
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regc_locale.c b/generic/regc_locale.c
new file mode 100644
index 0000000..d781212
--- /dev/null
+++ b/generic/regc_locale.c
@@ -0,0 +1,1267 @@
+/*
+ * regc_locale.c --
+ *
+ *	This file contains the Unicode locale specific regexp routines.
+ *	This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/* ASCII character-name table */
+
+static const struct cname {
+    const char *name;
+    const char code;
+} cnames[] = {
+    {"NUL",		'\0'},
+    {"SOH",		'\001'},
+    {"STX",		'\002'},
+    {"ETX",		'\003'},
+    {"EOT",		'\004'},
+    {"ENQ",		'\005'},
+    {"ACK",		'\006'},
+    {"BEL",		'\007'},
+    {"alert",		'\007'},
+    {"BS",		'\010'},
+    {"backspace",	'\b'},
+    {"HT",		'\011'},
+    {"tab",		'\t'},
+    {"LF",		'\012'},
+    {"newline",		'\n'},
+    {"VT",		'\013'},
+    {"vertical-tab",	'\v'},
+    {"FF",		'\014'},
+    {"form-feed",	'\f'},
+    {"CR",		'\015'},
+    {"carriage-return",	'\r'},
+    {"SO",		'\016'},
+    {"SI",		'\017'},
+    {"DLE",		'\020'},
+    {"DC1",		'\021'},
+    {"DC2",		'\022'},
+    {"DC3",		'\023'},
+    {"DC4",		'\024'},
+    {"NAK",		'\025'},
+    {"SYN",		'\026'},
+    {"ETB",		'\027'},
+    {"CAN",		'\030'},
+    {"EM",		'\031'},
+    {"SUB",		'\032'},
+    {"ESC",		'\033'},
+    {"IS4",		'\034'},
+    {"FS",		'\034'},
+    {"IS3",		'\035'},
+    {"GS",		'\035'},
+    {"IS2",		'\036'},
+    {"RS",		'\036'},
+    {"IS1",		'\037'},
+    {"US",		'\037'},
+    {"space",		' '},
+    {"exclamation-mark",'!'},
+    {"quotation-mark",	'"'},
+    {"number-sign",	'#'},
+    {"dollar-sign",	'$'},
+    {"percent-sign",	'%'},
+    {"ampersand",	'&'},
+    {"apostrophe",	'\''},
+    {"left-parenthesis",'('},
+    {"right-parenthesis", ')'},
+    {"asterisk",	'*'},
+    {"plus-sign",	'+'},
+    {"comma",		','},
+    {"hyphen",		'-'},
+    {"hyphen-minus",	'-'},
+    {"period",		'.'},
+    {"full-stop",	'.'},
+    {"slash",		'/'},
+    {"solidus",		'/'},
+    {"zero",		'0'},
+    {"one",		'1'},
+    {"two",		'2'},
+    {"three",		'3'},
+    {"four",		'4'},
+    {"five",		'5'},
+    {"six",		'6'},
+    {"seven",		'7'},
+    {"eight",		'8'},
+    {"nine",		'9'},
+    {"colon",		':'},
+    {"semicolon",	';'},
+    {"less-than-sign",	'<'},
+    {"equals-sign",	'='},
+    {"greater-than-sign", '>'},
+    {"question-mark",	'?'},
+    {"commercial-at",	'@'},
+    {"left-square-bracket", '['},
+    {"backslash",	'\\'},
+    {"reverse-solidus",	'\\'},
+    {"right-square-bracket", ']'},
+    {"circumflex",	'^'},
+    {"circumflex-accent", '^'},
+    {"underscore",	'_'},
+    {"low-line",	'_'},
+    {"grave-accent",	'`'},
+    {"left-brace",	'{'},
+    {"left-curly-bracket", '{'},
+    {"vertical-line",	'|'},
+    {"right-brace",	'}'},
+    {"right-curly-bracket", '}'},
+    {"tilde",		'~'},
+    {"DEL",		'\177'},
+    {NULL,		0}
+};
+
+/*
+ * Unicode character-class tables.
+ */
+
+typedef struct {
+    chr start;
+    chr end;
+} crange;
+
+/*
+ *	Declarations of Unicode character ranges.  This code
+ *	is automatically generated by the tools/uniClass.tcl script
+ *	and used in generic/regc_locale.c.  Do not modify by hand.
+ */
+
+/*
+ * Unicode: alphabetic characters.
+ */
+
+static const crange alphaRangeTable[] = {
+    {0x41, 0x5a}, {0x61, 0x7a}, {0xc0, 0xd6}, {0xd8, 0xf6},
+    {0xf8, 0x2c1}, {0x2c6, 0x2d1}, {0x2e0, 0x2e4}, {0x370, 0x374},
+    {0x37a, 0x37d}, {0x388, 0x38a}, {0x38e, 0x3a1}, {0x3a3, 0x3f5},
+    {0x3f7, 0x481}, {0x48a, 0x52f}, {0x531, 0x556}, {0x561, 0x587},
+    {0x5d0, 0x5ea}, {0x5f0, 0x5f2}, {0x620, 0x64a}, {0x671, 0x6d3},
+    {0x6fa, 0x6fc}, {0x712, 0x72f}, {0x74d, 0x7a5}, {0x7ca, 0x7ea},
+    {0x800, 0x815}, {0x840, 0x858}, {0x860, 0x86a}, {0x8a0, 0x8b4},
+    {0x8b6, 0x8bd}, {0x904, 0x939}, {0x958, 0x961}, {0x971, 0x980},
+    {0x985, 0x98c}, {0x993, 0x9a8}, {0x9aa, 0x9b0}, {0x9b6, 0x9b9},
+    {0x9df, 0x9e1}, {0xa05, 0xa0a}, {0xa13, 0xa28}, {0xa2a, 0xa30},
+    {0xa59, 0xa5c}, {0xa72, 0xa74}, {0xa85, 0xa8d}, {0xa8f, 0xa91},
+    {0xa93, 0xaa8}, {0xaaa, 0xab0}, {0xab5, 0xab9}, {0xb05, 0xb0c},
+    {0xb13, 0xb28}, {0xb2a, 0xb30}, {0xb35, 0xb39}, {0xb5f, 0xb61},
+    {0xb85, 0xb8a}, {0xb8e, 0xb90}, {0xb92, 0xb95}, {0xba8, 0xbaa},
+    {0xbae, 0xbb9}, {0xc05, 0xc0c}, {0xc0e, 0xc10}, {0xc12, 0xc28},
+    {0xc2a, 0xc39}, {0xc58, 0xc5a}, {0xc85, 0xc8c}, {0xc8e, 0xc90},
+    {0xc92, 0xca8}, {0xcaa, 0xcb3}, {0xcb5, 0xcb9}, {0xd05, 0xd0c},
+    {0xd0e, 0xd10}, {0xd12, 0xd3a}, {0xd54, 0xd56}, {0xd5f, 0xd61},
+    {0xd7a, 0xd7f}, {0xd85, 0xd96}, {0xd9a, 0xdb1}, {0xdb3, 0xdbb},
+    {0xdc0, 0xdc6}, {0xe01, 0xe30}, {0xe40, 0xe46}, {0xe94, 0xe97},
+    {0xe99, 0xe9f}, {0xea1, 0xea3}, {0xead, 0xeb0}, {0xec0, 0xec4},
+    {0xedc, 0xedf}, {0xf40, 0xf47}, {0xf49, 0xf6c}, {0xf88, 0xf8c},
+    {0x1000, 0x102a}, {0x1050, 0x1055}, {0x105a, 0x105d}, {0x106e, 0x1070},
+    {0x1075, 0x1081}, {0x10a0, 0x10c5}, {0x10d0, 0x10fa}, {0x10fc, 0x1248},
+    {0x124a, 0x124d}, {0x1250, 0x1256}, {0x125a, 0x125d}, {0x1260, 0x1288},
+    {0x128a, 0x128d}, {0x1290, 0x12b0}, {0x12b2, 0x12b5}, {0x12b8, 0x12be},
+    {0x12c2, 0x12c5}, {0x12c8, 0x12d6}, {0x12d8, 0x1310}, {0x1312, 0x1315},
+    {0x1318, 0x135a}, {0x1380, 0x138f}, {0x13a0, 0x13f5}, {0x13f8, 0x13fd},
+    {0x1401, 0x166c}, {0x166f, 0x167f}, {0x1681, 0x169a}, {0x16a0, 0x16ea},
+    {0x16f1, 0x16f8}, {0x1700, 0x170c}, {0x170e, 0x1711}, {0x1720, 0x1731},
+    {0x1740, 0x1751}, {0x1760, 0x176c}, {0x176e, 0x1770}, {0x1780, 0x17b3},
+    {0x1820, 0x1877}, {0x1880, 0x1884}, {0x1887, 0x18a8}, {0x18b0, 0x18f5},
+    {0x1900, 0x191e}, {0x1950, 0x196d}, {0x1970, 0x1974}, {0x1980, 0x19ab},
+    {0x19b0, 0x19c9}, {0x1a00, 0x1a16}, {0x1a20, 0x1a54}, {0x1b05, 0x1b33},
+    {0x1b45, 0x1b4b}, {0x1b83, 0x1ba0}, {0x1bba, 0x1be5}, {0x1c00, 0x1c23},
+    {0x1c4d, 0x1c4f}, {0x1c5a, 0x1c7d}, {0x1c80, 0x1c88}, {0x1ce9, 0x1cec},
+    {0x1cee, 0x1cf1}, {0x1d00, 0x1dbf}, {0x1e00, 0x1f15}, {0x1f18, 0x1f1d},
+    {0x1f20, 0x1f45}, {0x1f48, 0x1f4d}, {0x1f50, 0x1f57}, {0x1f5f, 0x1f7d},
+    {0x1f80, 0x1fb4}, {0x1fb6, 0x1fbc}, {0x1fc2, 0x1fc4}, {0x1fc6, 0x1fcc},
+    {0x1fd0, 0x1fd3}, {0x1fd6, 0x1fdb}, {0x1fe0, 0x1fec}, {0x1ff2, 0x1ff4},
+    {0x1ff6, 0x1ffc}, {0x2090, 0x209c}, {0x210a, 0x2113}, {0x2119, 0x211d},
+    {0x212a, 0x212d}, {0x212f, 0x2139}, {0x213c, 0x213f}, {0x2145, 0x2149},
+    {0x2c00, 0x2c2e}, {0x2c30, 0x2c5e}, {0x2c60, 0x2ce4}, {0x2ceb, 0x2cee},
+    {0x2d00, 0x2d25}, {0x2d30, 0x2d67}, {0x2d80, 0x2d96}, {0x2da0, 0x2da6},
+    {0x2da8, 0x2dae}, {0x2db0, 0x2db6}, {0x2db8, 0x2dbe}, {0x2dc0, 0x2dc6},
+    {0x2dc8, 0x2dce}, {0x2dd0, 0x2dd6}, {0x2dd8, 0x2dde}, {0x3031, 0x3035},
+    {0x3041, 0x3096}, {0x309d, 0x309f}, {0x30a1, 0x30fa}, {0x30fc, 0x30ff},
+    {0x3105, 0x312e}, {0x3131, 0x318e}, {0x31a0, 0x31ba}, {0x31f0, 0x31ff},
+    {0x3400, 0x4db5}, {0x4e00, 0x9fea}, {0xa000, 0xa48c}, {0xa4d0, 0xa4fd},
+    {0xa500, 0xa60c}, {0xa610, 0xa61f}, {0xa640, 0xa66e}, {0xa67f, 0xa69d},
+    {0xa6a0, 0xa6e5}, {0xa717, 0xa71f}, {0xa722, 0xa788}, {0xa78b, 0xa7ae},
+    {0xa7b0, 0xa7b7}, {0xa7f7, 0xa801}, {0xa803, 0xa805}, {0xa807, 0xa80a},
+    {0xa80c, 0xa822}, {0xa840, 0xa873}, {0xa882, 0xa8b3}, {0xa8f2, 0xa8f7},
+    {0xa90a, 0xa925}, {0xa930, 0xa946}, {0xa960, 0xa97c}, {0xa984, 0xa9b2},
+    {0xa9e0, 0xa9e4}, {0xa9e6, 0xa9ef}, {0xa9fa, 0xa9fe}, {0xaa00, 0xaa28},
+    {0xaa40, 0xaa42}, {0xaa44, 0xaa4b}, {0xaa60, 0xaa76}, {0xaa7e, 0xaaaf},
+    {0xaab9, 0xaabd}, {0xaadb, 0xaadd}, {0xaae0, 0xaaea}, {0xaaf2, 0xaaf4},
+    {0xab01, 0xab06}, {0xab09, 0xab0e}, {0xab11, 0xab16}, {0xab20, 0xab26},
+    {0xab28, 0xab2e}, {0xab30, 0xab5a}, {0xab5c, 0xab65}, {0xab70, 0xabe2},
+    {0xac00, 0xd7a3}, {0xd7b0, 0xd7c6}, {0xd7cb, 0xd7fb}, {0xdc00, 0xdc3e},
+    {0xdc40, 0xdc7e}, {0xdc80, 0xdcbe}, {0xdcc0, 0xdcfe}, {0xdd00, 0xdd3e},
+    {0xdd40, 0xdd7e}, {0xdd80, 0xddbe}, {0xddc0, 0xddfe}, {0xde00, 0xde3e},
+    {0xde40, 0xde7e}, {0xde80, 0xdebe}, {0xdec0, 0xdefe}, {0xdf00, 0xdf3e},
+    {0xdf40, 0xdf7e}, {0xdf80, 0xdfbe}, {0xdfc0, 0xdffe}, {0xf900, 0xfa6d},
+    {0xfa70, 0xfad9}, {0xfb00, 0xfb06}, {0xfb13, 0xfb17}, {0xfb1f, 0xfb28},
+    {0xfb2a, 0xfb36}, {0xfb38, 0xfb3c}, {0xfb46, 0xfbb1}, {0xfbd3, 0xfd3d},
+    {0xfd50, 0xfd8f}, {0xfd92, 0xfdc7}, {0xfdf0, 0xfdfb}, {0xfe70, 0xfe74},
+    {0xfe76, 0xfefc}, {0xff21, 0xff3a}, {0xff41, 0xff5a}, {0xff66, 0xffbe},
+    {0xffc2, 0xffc7}, {0xffca, 0xffcf}, {0xffd2, 0xffd7}, {0xffda, 0xffdc}
+#if TCL_UTF_MAX > 4
+    ,{0x10000, 0x1000b}, {0x1000d, 0x10026}, {0x10028, 0x1003a}, {0x1003f, 0x1004d},
+    {0x10050, 0x1005d}, {0x10080, 0x100fa}, {0x10280, 0x1029c}, {0x102a0, 0x102d0},
+    {0x10300, 0x1031f}, {0x1032d, 0x10340}, {0x10342, 0x10349}, {0x10350, 0x10375},
+    {0x10380, 0x1039d}, {0x103a0, 0x103c3}, {0x103c8, 0x103cf}, {0x10400, 0x1049d},
+    {0x104b0, 0x104d3}, {0x104d8, 0x104fb}, {0x10500, 0x10527}, {0x10530, 0x10563},
+    {0x10600, 0x10736}, {0x10740, 0x10755}, {0x10760, 0x10767}, {0x10800, 0x10805},
+    {0x1080a, 0x10835}, {0x1083f, 0x10855}, {0x10860, 0x10876}, {0x10880, 0x1089e},
+    {0x108e0, 0x108f2}, {0x10900, 0x10915}, {0x10920, 0x10939}, {0x10980, 0x109b7},
+    {0x10a10, 0x10a13}, {0x10a15, 0x10a17}, {0x10a19, 0x10a33}, {0x10a60, 0x10a7c},
+    {0x10a80, 0x10a9c}, {0x10ac0, 0x10ac7}, {0x10ac9, 0x10ae4}, {0x10b00, 0x10b35},
+    {0x10b40, 0x10b55}, {0x10b60, 0x10b72}, {0x10b80, 0x10b91}, {0x10c00, 0x10c48},
+    {0x10c80, 0x10cb2}, {0x10cc0, 0x10cf2}, {0x11003, 0x11037}, {0x11083, 0x110af},
+    {0x110d0, 0x110e8}, {0x11103, 0x11126}, {0x11150, 0x11172}, {0x11183, 0x111b2},
+    {0x111c1, 0x111c4}, {0x11200, 0x11211}, {0x11213, 0x1122b}, {0x11280, 0x11286},
+    {0x1128a, 0x1128d}, {0x1128f, 0x1129d}, {0x1129f, 0x112a8}, {0x112b0, 0x112de},
+    {0x11305, 0x1130c}, {0x11313, 0x11328}, {0x1132a, 0x11330}, {0x11335, 0x11339},
+    {0x1135d, 0x11361}, {0x11400, 0x11434}, {0x11447, 0x1144a}, {0x11480, 0x114af},
+    {0x11580, 0x115ae}, {0x115d8, 0x115db}, {0x11600, 0x1162f}, {0x11680, 0x116aa},
+    {0x11700, 0x11719}, {0x118a0, 0x118df}, {0x11a0b, 0x11a32}, {0x11a5c, 0x11a83},
+    {0x11a86, 0x11a89}, {0x11ac0, 0x11af8}, {0x11c00, 0x11c08}, {0x11c0a, 0x11c2e},
+    {0x11c72, 0x11c8f}, {0x11d00, 0x11d06}, {0x11d0b, 0x11d30}, {0x12000, 0x12399},
+    {0x12480, 0x12543}, {0x13000, 0x1342e}, {0x14400, 0x14646}, {0x16800, 0x16a38},
+    {0x16a40, 0x16a5e}, {0x16ad0, 0x16aed}, {0x16b00, 0x16b2f}, {0x16b40, 0x16b43},
+    {0x16b63, 0x16b77}, {0x16b7d, 0x16b8f}, {0x16f00, 0x16f44}, {0x16f93, 0x16f9f},
+    {0x17000, 0x187ec}, {0x18800, 0x18af2}, {0x1b000, 0x1b11e}, {0x1b170, 0x1b2fb},
+    {0x1bc00, 0x1bc6a}, {0x1bc70, 0x1bc7c}, {0x1bc80, 0x1bc88}, {0x1bc90, 0x1bc99},
+    {0x1d400, 0x1d454}, {0x1d456, 0x1d49c}, {0x1d4a9, 0x1d4ac}, {0x1d4ae, 0x1d4b9},
+    {0x1d4bd, 0x1d4c3}, {0x1d4c5, 0x1d505}, {0x1d507, 0x1d50a}, {0x1d50d, 0x1d514},
+    {0x1d516, 0x1d51c}, {0x1d51e, 0x1d539}, {0x1d53b, 0x1d53e}, {0x1d540, 0x1d544},
+    {0x1d54a, 0x1d550}, {0x1d552, 0x1d6a5}, {0x1d6a8, 0x1d6c0}, {0x1d6c2, 0x1d6da},
+    {0x1d6dc, 0x1d6fa}, {0x1d6fc, 0x1d714}, {0x1d716, 0x1d734}, {0x1d736, 0x1d74e},
+    {0x1d750, 0x1d76e}, {0x1d770, 0x1d788}, {0x1d78a, 0x1d7a8}, {0x1d7aa, 0x1d7c2},
+    {0x1d7c4, 0x1d7cb}, {0x1e800, 0x1e8c4}, {0x1e900, 0x1e943}, {0x1ee00, 0x1ee03},
+    {0x1ee05, 0x1ee1f}, {0x1ee29, 0x1ee32}, {0x1ee34, 0x1ee37}, {0x1ee4d, 0x1ee4f},
+    {0x1ee67, 0x1ee6a}, {0x1ee6c, 0x1ee72}, {0x1ee74, 0x1ee77}, {0x1ee79, 0x1ee7c},
+    {0x1ee80, 0x1ee89}, {0x1ee8b, 0x1ee9b}, {0x1eea1, 0x1eea3}, {0x1eea5, 0x1eea9},
+    {0x1eeab, 0x1eebb}, {0x20000, 0x2a6d6}, {0x2a700, 0x2b734}, {0x2b740, 0x2b81d},
+    {0x2b820, 0x2cea1}, {0x2ceb0, 0x2ebe0}, {0x2f800, 0x2fa1d}
+#endif
+};
+
+#define NUM_ALPHA_RANGE (sizeof(alphaRangeTable)/sizeof(crange))
+
+static const chr alphaCharTable[] = {
+    0xaa, 0xb5, 0xba, 0x2ec, 0x2ee, 0x376, 0x377, 0x37f, 0x386,
+    0x38c, 0x559, 0x66e, 0x66f, 0x6d5, 0x6e5, 0x6e6, 0x6ee, 0x6ef,
+    0x6ff, 0x710, 0x7b1, 0x7f4, 0x7f5, 0x7fa, 0x81a, 0x824, 0x828,
+    0x93d, 0x950, 0x98f, 0x990, 0x9b2, 0x9bd, 0x9ce, 0x9dc, 0x9dd,
+    0x9f0, 0x9f1, 0x9fc, 0xa0f, 0xa10, 0xa32, 0xa33, 0xa35, 0xa36,
+    0xa38, 0xa39, 0xa5e, 0xab2, 0xab3, 0xabd, 0xad0, 0xae0, 0xae1,
+    0xaf9, 0xb0f, 0xb10, 0xb32, 0xb33, 0xb3d, 0xb5c, 0xb5d, 0xb71,
+    0xb83, 0xb99, 0xb9a, 0xb9c, 0xb9e, 0xb9f, 0xba3, 0xba4, 0xbd0,
+    0xc3d, 0xc60, 0xc61, 0xc80, 0xcbd, 0xcde, 0xce0, 0xce1, 0xcf1,
+    0xcf2, 0xd3d, 0xd4e, 0xdbd, 0xe32, 0xe33, 0xe81, 0xe82, 0xe84,
+    0xe87, 0xe88, 0xe8a, 0xe8d, 0xea5, 0xea7, 0xeaa, 0xeab, 0xeb2,
+    0xeb3, 0xebd, 0xec6, 0xf00, 0x103f, 0x1061, 0x1065, 0x1066, 0x108e,
+    0x10c7, 0x10cd, 0x1258, 0x12c0, 0x17d7, 0x17dc, 0x18aa, 0x1aa7, 0x1bae,
+    0x1baf, 0x1cf5, 0x1cf6, 0x1f59, 0x1f5b, 0x1f5d, 0x1fbe, 0x2071, 0x207f,
+    0x2102, 0x2107, 0x2115, 0x2124, 0x2126, 0x2128, 0x214e, 0x2183, 0x2184,
+    0x2cf2, 0x2cf3, 0x2d27, 0x2d2d, 0x2d6f, 0x2e2f, 0x3005, 0x3006, 0x303b,
+    0x303c, 0xa62a, 0xa62b, 0xa8fb, 0xa8fd, 0xa9cf, 0xaa7a, 0xaab1, 0xaab5,
+    0xaab6, 0xaac0, 0xaac2, 0xfb1d, 0xfb3e, 0xfb40, 0xfb41, 0xfb43, 0xfb44
+#if TCL_UTF_MAX > 4
+    ,0x1003c, 0x1003d, 0x10808, 0x10837, 0x10838, 0x1083c, 0x108f4, 0x108f5, 0x109be,
+    0x109bf, 0x10a00, 0x11176, 0x111da, 0x111dc, 0x11288, 0x1130f, 0x11310, 0x11332,
+    0x11333, 0x1133d, 0x11350, 0x114c4, 0x114c5, 0x114c7, 0x11644, 0x118ff, 0x11a00,
+    0x11a3a, 0x11a50, 0x11c40, 0x11d08, 0x11d09, 0x11d46, 0x16f50, 0x16fe0, 0x16fe1,
+    0x1d49e, 0x1d49f, 0x1d4a2, 0x1d4a5, 0x1d4a6, 0x1d4bb, 0x1d546, 0x1ee21, 0x1ee22,
+    0x1ee24, 0x1ee27, 0x1ee39, 0x1ee3b, 0x1ee42, 0x1ee47, 0x1ee49, 0x1ee4b, 0x1ee51,
+    0x1ee52, 0x1ee54, 0x1ee57, 0x1ee59, 0x1ee5b, 0x1ee5d, 0x1ee5f, 0x1ee61, 0x1ee62,
+    0x1ee64, 0x1ee7e
+#endif
+};
+
+#define NUM_ALPHA_CHAR (sizeof(alphaCharTable)/sizeof(chr))
+
+/*
+ * Unicode: control characters.
+ */
+
+static const crange controlRangeTable[] = {
+    {0x0, 0x1f}, {0x7f, 0x9f}, {0x600, 0x605}, {0x200b, 0x200f},
+    {0x202a, 0x202e}, {0x2060, 0x2064}, {0x2066, 0x206f}, {0xe000, 0xf8ff},
+    {0xfff9, 0xfffb}
+#if TCL_UTF_MAX > 4
+    ,{0x1bca0, 0x1bca3}, {0x1d173, 0x1d17a}, {0xe0020, 0xe007f}, {0xf0000, 0xffffd},
+    {0x100000, 0x10fffd}
+#endif
+};
+
+#define NUM_CONTROL_RANGE (sizeof(controlRangeTable)/sizeof(crange))
+
+static const chr controlCharTable[] = {
+    0xad, 0x61c, 0x6dd, 0x70f, 0x8e2, 0x180e, 0xfeff
+#if TCL_UTF_MAX > 4
+    ,0x110bd, 0xe0001
+#endif
+};
+
+#define NUM_CONTROL_CHAR (sizeof(controlCharTable)/sizeof(chr))
+
+/*
+ * Unicode: decimal digit characters.
+ */
+
+static const crange digitRangeTable[] = {
+    {0x30, 0x39}, {0x660, 0x669}, {0x6f0, 0x6f9}, {0x7c0, 0x7c9},
+    {0x966, 0x96f}, {0x9e6, 0x9ef}, {0xa66, 0xa6f}, {0xae6, 0xaef},
+    {0xb66, 0xb6f}, {0xbe6, 0xbef}, {0xc66, 0xc6f}, {0xce6, 0xcef},
+    {0xd66, 0xd6f}, {0xde6, 0xdef}, {0xe50, 0xe59}, {0xed0, 0xed9},
+    {0xf20, 0xf29}, {0x1040, 0x1049}, {0x1090, 0x1099}, {0x17e0, 0x17e9},
+    {0x1810, 0x1819}, {0x1946, 0x194f}, {0x19d0, 0x19d9}, {0x1a80, 0x1a89},
+    {0x1a90, 0x1a99}, {0x1b50, 0x1b59}, {0x1bb0, 0x1bb9}, {0x1c40, 0x1c49},
+    {0x1c50, 0x1c59}, {0xa620, 0xa629}, {0xa8d0, 0xa8d9}, {0xa900, 0xa909},
+    {0xa9d0, 0xa9d9}, {0xa9f0, 0xa9f9}, {0xaa50, 0xaa59}, {0xabf0, 0xabf9},
+    {0xff10, 0xff19}
+#if TCL_UTF_MAX > 4
+    ,{0x104a0, 0x104a9}, {0x11066, 0x1106f}, {0x110f0, 0x110f9}, {0x11136, 0x1113f},
+    {0x111d0, 0x111d9}, {0x112f0, 0x112f9}, {0x11450, 0x11459}, {0x114d0, 0x114d9},
+    {0x11650, 0x11659}, {0x116c0, 0x116c9}, {0x11730, 0x11739}, {0x118e0, 0x118e9},
+    {0x11c50, 0x11c59}, {0x11d50, 0x11d59}, {0x16a60, 0x16a69}, {0x16b50, 0x16b59},
+    {0x1d7ce, 0x1d7ff}, {0x1e950, 0x1e959}
+#endif
+};
+
+#define NUM_DIGIT_RANGE (sizeof(digitRangeTable)/sizeof(crange))
+
+/*
+ * no singletons of digit characters.
+ */
+
+/*
+ * Unicode: punctuation characters.
+ */
+
+static const crange punctRangeTable[] = {
+    {0x21, 0x23}, {0x25, 0x2a}, {0x2c, 0x2f}, {0x5b, 0x5d},
+    {0x55a, 0x55f}, {0x66a, 0x66d}, {0x700, 0x70d}, {0x7f7, 0x7f9},
+    {0x830, 0x83e}, {0xf04, 0xf12}, {0xf3a, 0xf3d}, {0xfd0, 0xfd4},
+    {0x104a, 0x104f}, {0x1360, 0x1368}, {0x16eb, 0x16ed}, {0x17d4, 0x17d6},
+    {0x17d8, 0x17da}, {0x1800, 0x180a}, {0x1aa0, 0x1aa6}, {0x1aa8, 0x1aad},
+    {0x1b5a, 0x1b60}, {0x1bfc, 0x1bff}, {0x1c3b, 0x1c3f}, {0x1cc0, 0x1cc7},
+    {0x2010, 0x2027}, {0x2030, 0x2043}, {0x2045, 0x2051}, {0x2053, 0x205e},
+    {0x2308, 0x230b}, {0x2768, 0x2775}, {0x27e6, 0x27ef}, {0x2983, 0x2998},
+    {0x29d8, 0x29db}, {0x2cf9, 0x2cfc}, {0x2e00, 0x2e2e}, {0x2e30, 0x2e49},
+    {0x3001, 0x3003}, {0x3008, 0x3011}, {0x3014, 0x301f}, {0xa60d, 0xa60f},
+    {0xa6f2, 0xa6f7}, {0xa874, 0xa877}, {0xa8f8, 0xa8fa}, {0xa9c1, 0xa9cd},
+    {0xaa5c, 0xaa5f}, {0xfe10, 0xfe19}, {0xfe30, 0xfe52}, {0xfe54, 0xfe61},
+    {0xff01, 0xff03}, {0xff05, 0xff0a}, {0xff0c, 0xff0f}, {0xff3b, 0xff3d},
+    {0xff5f, 0xff65}
+#if TCL_UTF_MAX > 4
+    ,{0x10100, 0x10102}, {0x10a50, 0x10a58}, {0x10af0, 0x10af6}, {0x10b39, 0x10b3f},
+    {0x10b99, 0x10b9c}, {0x11047, 0x1104d}, {0x110be, 0x110c1}, {0x11140, 0x11143},
+    {0x111c5, 0x111c9}, {0x111dd, 0x111df}, {0x11238, 0x1123d}, {0x1144b, 0x1144f},
+    {0x115c1, 0x115d7}, {0x11641, 0x11643}, {0x11660, 0x1166c}, {0x1173c, 0x1173e},
+    {0x11a3f, 0x11a46}, {0x11a9a, 0x11a9c}, {0x11a9e, 0x11aa2}, {0x11c41, 0x11c45},
+    {0x12470, 0x12474}, {0x16b37, 0x16b3b}, {0x1da87, 0x1da8b}
+#endif
+};
+
+#define NUM_PUNCT_RANGE (sizeof(punctRangeTable)/sizeof(crange))
+
+static const chr punctCharTable[] = {
+    0x3a, 0x3b, 0x3f, 0x40, 0x5f, 0x7b, 0x7d, 0xa1, 0xa7,
+    0xab, 0xb6, 0xb7, 0xbb, 0xbf, 0x37e, 0x387, 0x589, 0x58a,
+    0x5be, 0x5c0, 0x5c3, 0x5c6, 0x5f3, 0x5f4, 0x609, 0x60a, 0x60c,
+    0x60d, 0x61b, 0x61e, 0x61f, 0x6d4, 0x85e, 0x964, 0x965, 0x970,
+    0x9fd, 0xaf0, 0xdf4, 0xe4f, 0xe5a, 0xe5b, 0xf14, 0xf85, 0xfd9,
+    0xfda, 0x10fb, 0x1400, 0x166d, 0x166e, 0x169b, 0x169c, 0x1735, 0x1736,
+    0x1944, 0x1945, 0x1a1e, 0x1a1f, 0x1c7e, 0x1c7f, 0x1cd3, 0x207d, 0x207e,
+    0x208d, 0x208e, 0x2329, 0x232a, 0x27c5, 0x27c6, 0x29fc, 0x29fd, 0x2cfe,
+    0x2cff, 0x2d70, 0x3030, 0x303d, 0x30a0, 0x30fb, 0xa4fe, 0xa4ff, 0xa673,
+    0xa67e, 0xa8ce, 0xa8cf, 0xa8fc, 0xa92e, 0xa92f, 0xa95f, 0xa9de, 0xa9df,
+    0xaade, 0xaadf, 0xaaf0, 0xaaf1, 0xabeb, 0xfd3e, 0xfd3f, 0xfe63, 0xfe68,
+    0xfe6a, 0xfe6b, 0xff1a, 0xff1b, 0xff1f, 0xff20, 0xff3f, 0xff5b, 0xff5d
+#if TCL_UTF_MAX > 4
+    ,0x1039f, 0x103d0, 0x1056f, 0x10857, 0x1091f, 0x1093f, 0x10a7f, 0x110bb, 0x110bc,
+    0x11174, 0x11175, 0x111cd, 0x111db, 0x112a9, 0x1145b, 0x1145d, 0x114c6, 0x11c70,
+    0x11c71, 0x16a6e, 0x16a6f, 0x16af5, 0x16b44, 0x1bc9f, 0x1e95e, 0x1e95f
+#endif
+};
+
+#define NUM_PUNCT_CHAR (sizeof(punctCharTable)/sizeof(chr))
+
+/*
+ * Unicode: white space characters.
+ */
+
+static const crange spaceRangeTable[] = {
+    {0x9, 0xd}, {0x2000, 0x200b}
+};
+
+#define NUM_SPACE_RANGE (sizeof(spaceRangeTable)/sizeof(crange))
+
+static const chr spaceCharTable[] = {
+    0x20, 0x85, 0xa0, 0x1680, 0x180e, 0x2028, 0x2029, 0x202f, 0x205f,
+    0x2060, 0x3000, 0xfeff
+};
+
+#define NUM_SPACE_CHAR (sizeof(spaceCharTable)/sizeof(chr))
+
+/*
+ * Unicode: lowercase characters.
+ */
+
+static const crange lowerRangeTable[] = {
+    {0x61, 0x7a}, {0xdf, 0xf6}, {0xf8, 0xff}, {0x17e, 0x180},
+    {0x199, 0x19b}, {0x1bd, 0x1bf}, {0x233, 0x239}, {0x24f, 0x293},
+    {0x295, 0x2af}, {0x37b, 0x37d}, {0x3ac, 0x3ce}, {0x3d5, 0x3d7},
+    {0x3ef, 0x3f3}, {0x430, 0x45f}, {0x561, 0x587}, {0x13f8, 0x13fd},
+    {0x1c80, 0x1c88}, {0x1d00, 0x1d2b}, {0x1d6b, 0x1d77}, {0x1d79, 0x1d9a},
+    {0x1e95, 0x1e9d}, {0x1eff, 0x1f07}, {0x1f10, 0x1f15}, {0x1f20, 0x1f27},
+    {0x1f30, 0x1f37}, {0x1f40, 0x1f45}, {0x1f50, 0x1f57}, {0x1f60, 0x1f67},
+    {0x1f70, 0x1f7d}, {0x1f80, 0x1f87}, {0x1f90, 0x1f97}, {0x1fa0, 0x1fa7},
+    {0x1fb0, 0x1fb4}, {0x1fc2, 0x1fc4}, {0x1fd0, 0x1fd3}, {0x1fe0, 0x1fe7},
+    {0x1ff2, 0x1ff4}, {0x2146, 0x2149}, {0x2c30, 0x2c5e}, {0x2c76, 0x2c7b},
+    {0x2d00, 0x2d25}, {0xa72f, 0xa731}, {0xa771, 0xa778}, {0xa793, 0xa795},
+    {0xab30, 0xab5a}, {0xab60, 0xab65}, {0xab70, 0xabbf}, {0xfb00, 0xfb06},
+    {0xfb13, 0xfb17}, {0xff41, 0xff5a}
+#if TCL_UTF_MAX > 4
+    ,{0x10428, 0x1044f}, {0x104d8, 0x104fb}, {0x10cc0, 0x10cf2}, {0x118c0, 0x118df},
+    {0x1d41a, 0x1d433}, {0x1d44e, 0x1d454}, {0x1d456, 0x1d467}, {0x1d482, 0x1d49b},
+    {0x1d4b6, 0x1d4b9}, {0x1d4bd, 0x1d4c3}, {0x1d4c5, 0x1d4cf}, {0x1d4ea, 0x1d503},
+    {0x1d51e, 0x1d537}, {0x1d552, 0x1d56b}, {0x1d586, 0x1d59f}, {0x1d5ba, 0x1d5d3},
+    {0x1d5ee, 0x1d607}, {0x1d622, 0x1d63b}, {0x1d656, 0x1d66f}, {0x1d68a, 0x1d6a5},
+    {0x1d6c2, 0x1d6da}, {0x1d6dc, 0x1d6e1}, {0x1d6fc, 0x1d714}, {0x1d716, 0x1d71b},
+    {0x1d736, 0x1d74e}, {0x1d750, 0x1d755}, {0x1d770, 0x1d788}, {0x1d78a, 0x1d78f},
+    {0x1d7aa, 0x1d7c2}, {0x1d7c4, 0x1d7c9}, {0x1e922, 0x1e943}
+#endif
+};
+
+#define NUM_LOWER_RANGE (sizeof(lowerRangeTable)/sizeof(crange))
+
+static const chr lowerCharTable[] = {
+    0xb5, 0x101, 0x103, 0x105, 0x107, 0x109, 0x10b, 0x10d, 0x10f,
+    0x111, 0x113, 0x115, 0x117, 0x119, 0x11b, 0x11d, 0x11f, 0x121,
+    0x123, 0x125, 0x127, 0x129, 0x12b, 0x12d, 0x12f, 0x131, 0x133,
+    0x135, 0x137, 0x138, 0x13a, 0x13c, 0x13e, 0x140, 0x142, 0x144,
+    0x146, 0x148, 0x149, 0x14b, 0x14d, 0x14f, 0x151, 0x153, 0x155,
+    0x157, 0x159, 0x15b, 0x15d, 0x15f, 0x161, 0x163, 0x165, 0x167,
+    0x169, 0x16b, 0x16d, 0x16f, 0x171, 0x173, 0x175, 0x177, 0x17a,
+    0x17c, 0x183, 0x185, 0x188, 0x18c, 0x18d, 0x192, 0x195, 0x19e,
+    0x1a1, 0x1a3, 0x1a5, 0x1a8, 0x1aa, 0x1ab, 0x1ad, 0x1b0, 0x1b4,
+    0x1b6, 0x1b9, 0x1ba, 0x1c6, 0x1c9, 0x1cc, 0x1ce, 0x1d0, 0x1d2,
+    0x1d4, 0x1d6, 0x1d8, 0x1da, 0x1dc, 0x1dd, 0x1df, 0x1e1, 0x1e3,
+    0x1e5, 0x1e7, 0x1e9, 0x1eb, 0x1ed, 0x1ef, 0x1f0, 0x1f3, 0x1f5,
+    0x1f9, 0x1fb, 0x1fd, 0x1ff, 0x201, 0x203, 0x205, 0x207, 0x209,
+    0x20b, 0x20d, 0x20f, 0x211, 0x213, 0x215, 0x217, 0x219, 0x21b,
+    0x21d, 0x21f, 0x221, 0x223, 0x225, 0x227, 0x229, 0x22b, 0x22d,
+    0x22f, 0x231, 0x23c, 0x23f, 0x240, 0x242, 0x247, 0x249, 0x24b,
+    0x24d, 0x371, 0x373, 0x377, 0x390, 0x3d0, 0x3d1, 0x3d9, 0x3db,
+    0x3dd, 0x3df, 0x3e1, 0x3e3, 0x3e5, 0x3e7, 0x3e9, 0x3eb, 0x3ed,
+    0x3f5, 0x3f8, 0x3fb, 0x3fc, 0x461, 0x463, 0x465, 0x467, 0x469,
+    0x46b, 0x46d, 0x46f, 0x471, 0x473, 0x475, 0x477, 0x479, 0x47b,
+    0x47d, 0x47f, 0x481, 0x48b, 0x48d, 0x48f, 0x491, 0x493, 0x495,
+    0x497, 0x499, 0x49b, 0x49d, 0x49f, 0x4a1, 0x4a3, 0x4a5, 0x4a7,
+    0x4a9, 0x4ab, 0x4ad, 0x4af, 0x4b1, 0x4b3, 0x4b5, 0x4b7, 0x4b9,
+    0x4bb, 0x4bd, 0x4bf, 0x4c2, 0x4c4, 0x4c6, 0x4c8, 0x4ca, 0x4cc,
+    0x4ce, 0x4cf, 0x4d1, 0x4d3, 0x4d5, 0x4d7, 0x4d9, 0x4db, 0x4dd,
+    0x4df, 0x4e1, 0x4e3, 0x4e5, 0x4e7, 0x4e9, 0x4eb, 0x4ed, 0x4ef,
+    0x4f1, 0x4f3, 0x4f5, 0x4f7, 0x4f9, 0x4fb, 0x4fd, 0x4ff, 0x501,
+    0x503, 0x505, 0x507, 0x509, 0x50b, 0x50d, 0x50f, 0x511, 0x513,
+    0x515, 0x517, 0x519, 0x51b, 0x51d, 0x51f, 0x521, 0x523, 0x525,
+    0x527, 0x529, 0x52b, 0x52d, 0x52f, 0x1e01, 0x1e03, 0x1e05, 0x1e07,
+    0x1e09, 0x1e0b, 0x1e0d, 0x1e0f, 0x1e11, 0x1e13, 0x1e15, 0x1e17, 0x1e19,
+    0x1e1b, 0x1e1d, 0x1e1f, 0x1e21, 0x1e23, 0x1e25, 0x1e27, 0x1e29, 0x1e2b,
+    0x1e2d, 0x1e2f, 0x1e31, 0x1e33, 0x1e35, 0x1e37, 0x1e39, 0x1e3b, 0x1e3d,
+    0x1e3f, 0x1e41, 0x1e43, 0x1e45, 0x1e47, 0x1e49, 0x1e4b, 0x1e4d, 0x1e4f,
+    0x1e51, 0x1e53, 0x1e55, 0x1e57, 0x1e59, 0x1e5b, 0x1e5d, 0x1e5f, 0x1e61,
+    0x1e63, 0x1e65, 0x1e67, 0x1e69, 0x1e6b, 0x1e6d, 0x1e6f, 0x1e71, 0x1e73,
+    0x1e75, 0x1e77, 0x1e79, 0x1e7b, 0x1e7d, 0x1e7f, 0x1e81, 0x1e83, 0x1e85,
+    0x1e87, 0x1e89, 0x1e8b, 0x1e8d, 0x1e8f, 0x1e91, 0x1e93, 0x1e9f, 0x1ea1,
+    0x1ea3, 0x1ea5, 0x1ea7, 0x1ea9, 0x1eab, 0x1ead, 0x1eaf, 0x1eb1, 0x1eb3,
+    0x1eb5, 0x1eb7, 0x1eb9, 0x1ebb, 0x1ebd, 0x1ebf, 0x1ec1, 0x1ec3, 0x1ec5,
+    0x1ec7, 0x1ec9, 0x1ecb, 0x1ecd, 0x1ecf, 0x1ed1, 0x1ed3, 0x1ed5, 0x1ed7,
+    0x1ed9, 0x1edb, 0x1edd, 0x1edf, 0x1ee1, 0x1ee3, 0x1ee5, 0x1ee7, 0x1ee9,
+    0x1eeb, 0x1eed, 0x1eef, 0x1ef1, 0x1ef3, 0x1ef5, 0x1ef7, 0x1ef9, 0x1efb,
+    0x1efd, 0x1fb6, 0x1fb7, 0x1fbe, 0x1fc6, 0x1fc7, 0x1fd6, 0x1fd7, 0x1ff6,
+    0x1ff7, 0x210a, 0x210e, 0x210f, 0x2113, 0x212f, 0x2134, 0x2139, 0x213c,
+    0x213d, 0x214e, 0x2184, 0x2c61, 0x2c65, 0x2c66, 0x2c68, 0x2c6a, 0x2c6c,
+    0x2c71, 0x2c73, 0x2c74, 0x2c81, 0x2c83, 0x2c85, 0x2c87, 0x2c89, 0x2c8b,
+    0x2c8d, 0x2c8f, 0x2c91, 0x2c93, 0x2c95, 0x2c97, 0x2c99, 0x2c9b, 0x2c9d,
+    0x2c9f, 0x2ca1, 0x2ca3, 0x2ca5, 0x2ca7, 0x2ca9, 0x2cab, 0x2cad, 0x2caf,
+    0x2cb1, 0x2cb3, 0x2cb5, 0x2cb7, 0x2cb9, 0x2cbb, 0x2cbd, 0x2cbf, 0x2cc1,
+    0x2cc3, 0x2cc5, 0x2cc7, 0x2cc9, 0x2ccb, 0x2ccd, 0x2ccf, 0x2cd1, 0x2cd3,
+    0x2cd5, 0x2cd7, 0x2cd9, 0x2cdb, 0x2cdd, 0x2cdf, 0x2ce1, 0x2ce3, 0x2ce4,
+    0x2cec, 0x2cee, 0x2cf3, 0x2d27, 0x2d2d, 0xa641, 0xa643, 0xa645, 0xa647,
+    0xa649, 0xa64b, 0xa64d, 0xa64f, 0xa651, 0xa653, 0xa655, 0xa657, 0xa659,
+    0xa65b, 0xa65d, 0xa65f, 0xa661, 0xa663, 0xa665, 0xa667, 0xa669, 0xa66b,
+    0xa66d, 0xa681, 0xa683, 0xa685, 0xa687, 0xa689, 0xa68b, 0xa68d, 0xa68f,
+    0xa691, 0xa693, 0xa695, 0xa697, 0xa699, 0xa69b, 0xa723, 0xa725, 0xa727,
+    0xa729, 0xa72b, 0xa72d, 0xa733, 0xa735, 0xa737, 0xa739, 0xa73b, 0xa73d,
+    0xa73f, 0xa741, 0xa743, 0xa745, 0xa747, 0xa749, 0xa74b, 0xa74d, 0xa74f,
+    0xa751, 0xa753, 0xa755, 0xa757, 0xa759, 0xa75b, 0xa75d, 0xa75f, 0xa761,
+    0xa763, 0xa765, 0xa767, 0xa769, 0xa76b, 0xa76d, 0xa76f, 0xa77a, 0xa77c,
+    0xa77f, 0xa781, 0xa783, 0xa785, 0xa787, 0xa78c, 0xa78e, 0xa791, 0xa797,
+    0xa799, 0xa79b, 0xa79d, 0xa79f, 0xa7a1, 0xa7a3, 0xa7a5, 0xa7a7, 0xa7a9,
+    0xa7b5, 0xa7b7, 0xa7fa
+#if TCL_UTF_MAX > 4
+    ,0x1d4bb, 0x1d7cb
+#endif
+};
+
+#define NUM_LOWER_CHAR (sizeof(lowerCharTable)/sizeof(chr))
+
+/*
+ * Unicode: uppercase characters.
+ */
+
+static const crange upperRangeTable[] = {
+    {0x41, 0x5a}, {0xc0, 0xd6}, {0xd8, 0xde}, {0x189, 0x18b},
+    {0x18e, 0x191}, {0x196, 0x198}, {0x1b1, 0x1b3}, {0x1f6, 0x1f8},
+    {0x243, 0x246}, {0x388, 0x38a}, {0x391, 0x3a1}, {0x3a3, 0x3ab},
+    {0x3d2, 0x3d4}, {0x3fd, 0x42f}, {0x531, 0x556}, {0x10a0, 0x10c5},
+    {0x13a0, 0x13f5}, {0x1f08, 0x1f0f}, {0x1f18, 0x1f1d}, {0x1f28, 0x1f2f},
+    {0x1f38, 0x1f3f}, {0x1f48, 0x1f4d}, {0x1f68, 0x1f6f}, {0x1fb8, 0x1fbb},
+    {0x1fc8, 0x1fcb}, {0x1fd8, 0x1fdb}, {0x1fe8, 0x1fec}, {0x1ff8, 0x1ffb},
+    {0x210b, 0x210d}, {0x2110, 0x2112}, {0x2119, 0x211d}, {0x212a, 0x212d},
+    {0x2130, 0x2133}, {0x2c00, 0x2c2e}, {0x2c62, 0x2c64}, {0x2c6d, 0x2c70},
+    {0x2c7e, 0x2c80}, {0xa7aa, 0xa7ae}, {0xa7b0, 0xa7b4}, {0xff21, 0xff3a}
+#if TCL_UTF_MAX > 4
+    ,{0x10400, 0x10427}, {0x104b0, 0x104d3}, {0x10c80, 0x10cb2}, {0x118a0, 0x118bf},
+    {0x1d400, 0x1d419}, {0x1d434, 0x1d44d}, {0x1d468, 0x1d481}, {0x1d4a9, 0x1d4ac},
+    {0x1d4ae, 0x1d4b5}, {0x1d4d0, 0x1d4e9}, {0x1d507, 0x1d50a}, {0x1d50d, 0x1d514},
+    {0x1d516, 0x1d51c}, {0x1d53b, 0x1d53e}, {0x1d540, 0x1d544}, {0x1d54a, 0x1d550},
+    {0x1d56c, 0x1d585}, {0x1d5a0, 0x1d5b9}, {0x1d5d4, 0x1d5ed}, {0x1d608, 0x1d621},
+    {0x1d63c, 0x1d655}, {0x1d670, 0x1d689}, {0x1d6a8, 0x1d6c0}, {0x1d6e2, 0x1d6fa},
+    {0x1d71c, 0x1d734}, {0x1d756, 0x1d76e}, {0x1d790, 0x1d7a8}, {0x1e900, 0x1e921}
+#endif
+};
+
+#define NUM_UPPER_RANGE (sizeof(upperRangeTable)/sizeof(crange))
+
+static const chr upperCharTable[] = {
+    0x100, 0x102, 0x104, 0x106, 0x108, 0x10a, 0x10c, 0x10e, 0x110,
+    0x112, 0x114, 0x116, 0x118, 0x11a, 0x11c, 0x11e, 0x120, 0x122,
+    0x124, 0x126, 0x128, 0x12a, 0x12c, 0x12e, 0x130, 0x132, 0x134,
+    0x136, 0x139, 0x13b, 0x13d, 0x13f, 0x141, 0x143, 0x145, 0x147,
+    0x14a, 0x14c, 0x14e, 0x150, 0x152, 0x154, 0x156, 0x158, 0x15a,
+    0x15c, 0x15e, 0x160, 0x162, 0x164, 0x166, 0x168, 0x16a, 0x16c,
+    0x16e, 0x170, 0x172, 0x174, 0x176, 0x178, 0x179, 0x17b, 0x17d,
+    0x181, 0x182, 0x184, 0x186, 0x187, 0x193, 0x194, 0x19c, 0x19d,
+    0x19f, 0x1a0, 0x1a2, 0x1a4, 0x1a6, 0x1a7, 0x1a9, 0x1ac, 0x1ae,
+    0x1af, 0x1b5, 0x1b7, 0x1b8, 0x1bc, 0x1c4, 0x1c7, 0x1ca, 0x1cd,
+    0x1cf, 0x1d1, 0x1d3, 0x1d5, 0x1d7, 0x1d9, 0x1db, 0x1de, 0x1e0,
+    0x1e2, 0x1e4, 0x1e6, 0x1e8, 0x1ea, 0x1ec, 0x1ee, 0x1f1, 0x1f4,
+    0x1fa, 0x1fc, 0x1fe, 0x200, 0x202, 0x204, 0x206, 0x208, 0x20a,
+    0x20c, 0x20e, 0x210, 0x212, 0x214, 0x216, 0x218, 0x21a, 0x21c,
+    0x21e, 0x220, 0x222, 0x224, 0x226, 0x228, 0x22a, 0x22c, 0x22e,
+    0x230, 0x232, 0x23a, 0x23b, 0x23d, 0x23e, 0x241, 0x248, 0x24a,
+    0x24c, 0x24e, 0x370, 0x372, 0x376, 0x37f, 0x386, 0x38c, 0x38e,
+    0x38f, 0x3cf, 0x3d8, 0x3da, 0x3dc, 0x3de, 0x3e0, 0x3e2, 0x3e4,
+    0x3e6, 0x3e8, 0x3ea, 0x3ec, 0x3ee, 0x3f4, 0x3f7, 0x3f9, 0x3fa,
+    0x460, 0x462, 0x464, 0x466, 0x468, 0x46a, 0x46c, 0x46e, 0x470,
+    0x472, 0x474, 0x476, 0x478, 0x47a, 0x47c, 0x47e, 0x480, 0x48a,
+    0x48c, 0x48e, 0x490, 0x492, 0x494, 0x496, 0x498, 0x49a, 0x49c,
+    0x49e, 0x4a0, 0x4a2, 0x4a4, 0x4a6, 0x4a8, 0x4aa, 0x4ac, 0x4ae,
+    0x4b0, 0x4b2, 0x4b4, 0x4b6, 0x4b8, 0x4ba, 0x4bc, 0x4be, 0x4c0,
+    0x4c1, 0x4c3, 0x4c5, 0x4c7, 0x4c9, 0x4cb, 0x4cd, 0x4d0, 0x4d2,
+    0x4d4, 0x4d6, 0x4d8, 0x4da, 0x4dc, 0x4de, 0x4e0, 0x4e2, 0x4e4,
+    0x4e6, 0x4e8, 0x4ea, 0x4ec, 0x4ee, 0x4f0, 0x4f2, 0x4f4, 0x4f6,
+    0x4f8, 0x4fa, 0x4fc, 0x4fe, 0x500, 0x502, 0x504, 0x506, 0x508,
+    0x50a, 0x50c, 0x50e, 0x510, 0x512, 0x514, 0x516, 0x518, 0x51a,
+    0x51c, 0x51e, 0x520, 0x522, 0x524, 0x526, 0x528, 0x52a, 0x52c,
+    0x52e, 0x10c7, 0x10cd, 0x1e00, 0x1e02, 0x1e04, 0x1e06, 0x1e08, 0x1e0a,
+    0x1e0c, 0x1e0e, 0x1e10, 0x1e12, 0x1e14, 0x1e16, 0x1e18, 0x1e1a, 0x1e1c,
+    0x1e1e, 0x1e20, 0x1e22, 0x1e24, 0x1e26, 0x1e28, 0x1e2a, 0x1e2c, 0x1e2e,
+    0x1e30, 0x1e32, 0x1e34, 0x1e36, 0x1e38, 0x1e3a, 0x1e3c, 0x1e3e, 0x1e40,
+    0x1e42, 0x1e44, 0x1e46, 0x1e48, 0x1e4a, 0x1e4c, 0x1e4e, 0x1e50, 0x1e52,
+    0x1e54, 0x1e56, 0x1e58, 0x1e5a, 0x1e5c, 0x1e5e, 0x1e60, 0x1e62, 0x1e64,
+    0x1e66, 0x1e68, 0x1e6a, 0x1e6c, 0x1e6e, 0x1e70, 0x1e72, 0x1e74, 0x1e76,
+    0x1e78, 0x1e7a, 0x1e7c, 0x1e7e, 0x1e80, 0x1e82, 0x1e84, 0x1e86, 0x1e88,
+    0x1e8a, 0x1e8c, 0x1e8e, 0x1e90, 0x1e92, 0x1e94, 0x1e9e, 0x1ea0, 0x1ea2,
+    0x1ea4, 0x1ea6, 0x1ea8, 0x1eaa, 0x1eac, 0x1eae, 0x1eb0, 0x1eb2, 0x1eb4,
+    0x1eb6, 0x1eb8, 0x1eba, 0x1ebc, 0x1ebe, 0x1ec0, 0x1ec2, 0x1ec4, 0x1ec6,
+    0x1ec8, 0x1eca, 0x1ecc, 0x1ece, 0x1ed0, 0x1ed2, 0x1ed4, 0x1ed6, 0x1ed8,
+    0x1eda, 0x1edc, 0x1ede, 0x1ee0, 0x1ee2, 0x1ee4, 0x1ee6, 0x1ee8, 0x1eea,
+    0x1eec, 0x1eee, 0x1ef0, 0x1ef2, 0x1ef4, 0x1ef6, 0x1ef8, 0x1efa, 0x1efc,
+    0x1efe, 0x1f59, 0x1f5b, 0x1f5d, 0x1f5f, 0x2102, 0x2107, 0x2115, 0x2124,
+    0x2126, 0x2128, 0x213e, 0x213f, 0x2145, 0x2183, 0x2c60, 0x2c67, 0x2c69,
+    0x2c6b, 0x2c72, 0x2c75, 0x2c82, 0x2c84, 0x2c86, 0x2c88, 0x2c8a, 0x2c8c,
+    0x2c8e, 0x2c90, 0x2c92, 0x2c94, 0x2c96, 0x2c98, 0x2c9a, 0x2c9c, 0x2c9e,
+    0x2ca0, 0x2ca2, 0x2ca4, 0x2ca6, 0x2ca8, 0x2caa, 0x2cac, 0x2cae, 0x2cb0,
+    0x2cb2, 0x2cb4, 0x2cb6, 0x2cb8, 0x2cba, 0x2cbc, 0x2cbe, 0x2cc0, 0x2cc2,
+    0x2cc4, 0x2cc6, 0x2cc8, 0x2cca, 0x2ccc, 0x2cce, 0x2cd0, 0x2cd2, 0x2cd4,
+    0x2cd6, 0x2cd8, 0x2cda, 0x2cdc, 0x2cde, 0x2ce0, 0x2ce2, 0x2ceb, 0x2ced,
+    0x2cf2, 0xa640, 0xa642, 0xa644, 0xa646, 0xa648, 0xa64a, 0xa64c, 0xa64e,
+    0xa650, 0xa652, 0xa654, 0xa656, 0xa658, 0xa65a, 0xa65c, 0xa65e, 0xa660,
+    0xa662, 0xa664, 0xa666, 0xa668, 0xa66a, 0xa66c, 0xa680, 0xa682, 0xa684,
+    0xa686, 0xa688, 0xa68a, 0xa68c, 0xa68e, 0xa690, 0xa692, 0xa694, 0xa696,
+    0xa698, 0xa69a, 0xa722, 0xa724, 0xa726, 0xa728, 0xa72a, 0xa72c, 0xa72e,
+    0xa732, 0xa734, 0xa736, 0xa738, 0xa73a, 0xa73c, 0xa73e, 0xa740, 0xa742,
+    0xa744, 0xa746, 0xa748, 0xa74a, 0xa74c, 0xa74e, 0xa750, 0xa752, 0xa754,
+    0xa756, 0xa758, 0xa75a, 0xa75c, 0xa75e, 0xa760, 0xa762, 0xa764, 0xa766,
+    0xa768, 0xa76a, 0xa76c, 0xa76e, 0xa779, 0xa77b, 0xa77d, 0xa77e, 0xa780,
+    0xa782, 0xa784, 0xa786, 0xa78b, 0xa78d, 0xa790, 0xa792, 0xa796, 0xa798,
+    0xa79a, 0xa79c, 0xa79e, 0xa7a0, 0xa7a2, 0xa7a4, 0xa7a6, 0xa7a8, 0xa7b6
+#if TCL_UTF_MAX > 4
+    ,0x1d49c, 0x1d49e, 0x1d49f, 0x1d4a2, 0x1d4a5, 0x1d4a6, 0x1d504, 0x1d505, 0x1d538,
+    0x1d539, 0x1d546, 0x1d7ca
+#endif
+};
+
+#define NUM_UPPER_CHAR (sizeof(upperCharTable)/sizeof(chr))
+
+/*
+ * Unicode: unicode print characters excluding space.
+ */
+
+static const crange graphRangeTable[] = {
+    {0x21, 0x7e}, {0xa1, 0xac}, {0xae, 0x377}, {0x37a, 0x37f},
+    {0x384, 0x38a}, {0x38e, 0x3a1}, {0x3a3, 0x52f}, {0x531, 0x556},
+    {0x559, 0x55f}, {0x561, 0x587}, {0x58d, 0x58f}, {0x591, 0x5c7},
+    {0x5d0, 0x5ea}, {0x5f0, 0x5f4}, {0x606, 0x61b}, {0x61e, 0x6dc},
+    {0x6de, 0x70d}, {0x710, 0x74a}, {0x74d, 0x7b1}, {0x7c0, 0x7fa},
+    {0x800, 0x82d}, {0x830, 0x83e}, {0x840, 0x85b}, {0x860, 0x86a},
+    {0x8a0, 0x8b4}, {0x8b6, 0x8bd}, {0x8d4, 0x8e1}, {0x8e3, 0x983},
+    {0x985, 0x98c}, {0x993, 0x9a8}, {0x9aa, 0x9b0}, {0x9b6, 0x9b9},
+    {0x9bc, 0x9c4}, {0x9cb, 0x9ce}, {0x9df, 0x9e3}, {0x9e6, 0x9fd},
+    {0xa01, 0xa03}, {0xa05, 0xa0a}, {0xa13, 0xa28}, {0xa2a, 0xa30},
+    {0xa3e, 0xa42}, {0xa4b, 0xa4d}, {0xa59, 0xa5c}, {0xa66, 0xa75},
+    {0xa81, 0xa83}, {0xa85, 0xa8d}, {0xa8f, 0xa91}, {0xa93, 0xaa8},
+    {0xaaa, 0xab0}, {0xab5, 0xab9}, {0xabc, 0xac5}, {0xac7, 0xac9},
+    {0xacb, 0xacd}, {0xae0, 0xae3}, {0xae6, 0xaf1}, {0xaf9, 0xaff},
+    {0xb01, 0xb03}, {0xb05, 0xb0c}, {0xb13, 0xb28}, {0xb2a, 0xb30},
+    {0xb35, 0xb39}, {0xb3c, 0xb44}, {0xb4b, 0xb4d}, {0xb5f, 0xb63},
+    {0xb66, 0xb77}, {0xb85, 0xb8a}, {0xb8e, 0xb90}, {0xb92, 0xb95},
+    {0xba8, 0xbaa}, {0xbae, 0xbb9}, {0xbbe, 0xbc2}, {0xbc6, 0xbc8},
+    {0xbca, 0xbcd}, {0xbe6, 0xbfa}, {0xc00, 0xc03}, {0xc05, 0xc0c},
+    {0xc0e, 0xc10}, {0xc12, 0xc28}, {0xc2a, 0xc39}, {0xc3d, 0xc44},
+    {0xc46, 0xc48}, {0xc4a, 0xc4d}, {0xc58, 0xc5a}, {0xc60, 0xc63},
+    {0xc66, 0xc6f}, {0xc78, 0xc83}, {0xc85, 0xc8c}, {0xc8e, 0xc90},
+    {0xc92, 0xca8}, {0xcaa, 0xcb3}, {0xcb5, 0xcb9}, {0xcbc, 0xcc4},
+    {0xcc6, 0xcc8}, {0xcca, 0xccd}, {0xce0, 0xce3}, {0xce6, 0xcef},
+    {0xd00, 0xd03}, {0xd05, 0xd0c}, {0xd0e, 0xd10}, {0xd12, 0xd44},
+    {0xd46, 0xd48}, {0xd4a, 0xd4f}, {0xd54, 0xd63}, {0xd66, 0xd7f},
+    {0xd85, 0xd96}, {0xd9a, 0xdb1}, {0xdb3, 0xdbb}, {0xdc0, 0xdc6},
+    {0xdcf, 0xdd4}, {0xdd8, 0xddf}, {0xde6, 0xdef}, {0xdf2, 0xdf4},
+    {0xe01, 0xe3a}, {0xe3f, 0xe5b}, {0xe94, 0xe97}, {0xe99, 0xe9f},
+    {0xea1, 0xea3}, {0xead, 0xeb9}, {0xebb, 0xebd}, {0xec0, 0xec4},
+    {0xec8, 0xecd}, {0xed0, 0xed9}, {0xedc, 0xedf}, {0xf00, 0xf47},
+    {0xf49, 0xf6c}, {0xf71, 0xf97}, {0xf99, 0xfbc}, {0xfbe, 0xfcc},
+    {0xfce, 0xfda}, {0x1000, 0x10c5}, {0x10d0, 0x1248}, {0x124a, 0x124d},
+    {0x1250, 0x1256}, {0x125a, 0x125d}, {0x1260, 0x1288}, {0x128a, 0x128d},
+    {0x1290, 0x12b0}, {0x12b2, 0x12b5}, {0x12b8, 0x12be}, {0x12c2, 0x12c5},
+    {0x12c8, 0x12d6}, {0x12d8, 0x1310}, {0x1312, 0x1315}, {0x1318, 0x135a},
+    {0x135d, 0x137c}, {0x1380, 0x1399}, {0x13a0, 0x13f5}, {0x13f8, 0x13fd},
+    {0x1400, 0x167f}, {0x1681, 0x169c}, {0x16a0, 0x16f8}, {0x1700, 0x170c},
+    {0x170e, 0x1714}, {0x1720, 0x1736}, {0x1740, 0x1753}, {0x1760, 0x176c},
+    {0x176e, 0x1770}, {0x1780, 0x17dd}, {0x17e0, 0x17e9}, {0x17f0, 0x17f9},
+    {0x1800, 0x180d}, {0x1810, 0x1819}, {0x1820, 0x1877}, {0x1880, 0x18aa},
+    {0x18b0, 0x18f5}, {0x1900, 0x191e}, {0x1920, 0x192b}, {0x1930, 0x193b},
+    {0x1944, 0x196d}, {0x1970, 0x1974}, {0x1980, 0x19ab}, {0x19b0, 0x19c9},
+    {0x19d0, 0x19da}, {0x19de, 0x1a1b}, {0x1a1e, 0x1a5e}, {0x1a60, 0x1a7c},
+    {0x1a7f, 0x1a89}, {0x1a90, 0x1a99}, {0x1aa0, 0x1aad}, {0x1ab0, 0x1abe},
+    {0x1b00, 0x1b4b}, {0x1b50, 0x1b7c}, {0x1b80, 0x1bf3}, {0x1bfc, 0x1c37},
+    {0x1c3b, 0x1c49}, {0x1c4d, 0x1c88}, {0x1cc0, 0x1cc7}, {0x1cd0, 0x1cf9},
+    {0x1d00, 0x1df9}, {0x1dfb, 0x1f15}, {0x1f18, 0x1f1d}, {0x1f20, 0x1f45},
+    {0x1f48, 0x1f4d}, {0x1f50, 0x1f57}, {0x1f5f, 0x1f7d}, {0x1f80, 0x1fb4},
+    {0x1fb6, 0x1fc4}, {0x1fc6, 0x1fd3}, {0x1fd6, 0x1fdb}, {0x1fdd, 0x1fef},
+    {0x1ff2, 0x1ff4}, {0x1ff6, 0x1ffe}, {0x2010, 0x2027}, {0x2030, 0x205e},
+    {0x2074, 0x208e}, {0x2090, 0x209c}, {0x20a0, 0x20bf}, {0x20d0, 0x20f0},
+    {0x2100, 0x218b}, {0x2190, 0x2426}, {0x2440, 0x244a}, {0x2460, 0x2b73},
+    {0x2b76, 0x2b95}, {0x2b98, 0x2bb9}, {0x2bbd, 0x2bc8}, {0x2bca, 0x2bd2},
+    {0x2bec, 0x2bef}, {0x2c00, 0x2c2e}, {0x2c30, 0x2c5e}, {0x2c60, 0x2cf3},
+    {0x2cf9, 0x2d25}, {0x2d30, 0x2d67}, {0x2d7f, 0x2d96}, {0x2da0, 0x2da6},
+    {0x2da8, 0x2dae}, {0x2db0, 0x2db6}, {0x2db8, 0x2dbe}, {0x2dc0, 0x2dc6},
+    {0x2dc8, 0x2dce}, {0x2dd0, 0x2dd6}, {0x2dd8, 0x2dde}, {0x2de0, 0x2e49},
+    {0x2e80, 0x2e99}, {0x2e9b, 0x2ef3}, {0x2f00, 0x2fd5}, {0x2ff0, 0x2ffb},
+    {0x3001, 0x303f}, {0x3041, 0x3096}, {0x3099, 0x30ff}, {0x3105, 0x312e},
+    {0x3131, 0x318e}, {0x3190, 0x31ba}, {0x31c0, 0x31e3}, {0x31f0, 0x321e},
+    {0x3220, 0x32fe}, {0x3300, 0x4db5}, {0x4dc0, 0x9fea}, {0xa000, 0xa48c},
+    {0xa490, 0xa4c6}, {0xa4d0, 0xa62b}, {0xa640, 0xa6f7}, {0xa700, 0xa7ae},
+    {0xa7b0, 0xa7b7}, {0xa7f7, 0xa82b}, {0xa830, 0xa839}, {0xa840, 0xa877},
+    {0xa880, 0xa8c5}, {0xa8ce, 0xa8d9}, {0xa8e0, 0xa8fd}, {0xa900, 0xa953},
+    {0xa95f, 0xa97c}, {0xa980, 0xa9cd}, {0xa9cf, 0xa9d9}, {0xa9de, 0xa9fe},
+    {0xaa00, 0xaa36}, {0xaa40, 0xaa4d}, {0xaa50, 0xaa59}, {0xaa5c, 0xaac2},
+    {0xaadb, 0xaaf6}, {0xab01, 0xab06}, {0xab09, 0xab0e}, {0xab11, 0xab16},
+    {0xab20, 0xab26}, {0xab28, 0xab2e}, {0xab30, 0xab65}, {0xab70, 0xabed},
+    {0xabf0, 0xabf9}, {0xac00, 0xd7a3}, {0xd7b0, 0xd7c6}, {0xd7cb, 0xd7fb},
+    {0xdc00, 0xdc3e}, {0xdc40, 0xdc7e}, {0xdc80, 0xdcbe}, {0xdcc0, 0xdcfe},
+    {0xdd00, 0xdd3e}, {0xdd40, 0xdd7e}, {0xdd80, 0xddbe}, {0xddc0, 0xddfe},
+    {0xde00, 0xde3e}, {0xde40, 0xde7e}, {0xde80, 0xdebe}, {0xdec0, 0xdefe},
+    {0xdf00, 0xdf3e}, {0xdf40, 0xdf7e}, {0xdf80, 0xdfbe}, {0xdfc0, 0xdffe},
+    {0xf900, 0xfa6d}, {0xfa70, 0xfad9}, {0xfb00, 0xfb06}, {0xfb13, 0xfb17},
+    {0xfb1d, 0xfb36}, {0xfb38, 0xfb3c}, {0xfb46, 0xfbc1}, {0xfbd3, 0xfd3f},
+    {0xfd50, 0xfd8f}, {0xfd92, 0xfdc7}, {0xfdf0, 0xfdfd}, {0xfe00, 0xfe19},
+    {0xfe20, 0xfe52}, {0xfe54, 0xfe66}, {0xfe68, 0xfe6b}, {0xfe70, 0xfe74},
+    {0xfe76, 0xfefc}, {0xff01, 0xffbe}, {0xffc2, 0xffc7}, {0xffca, 0xffcf},
+    {0xffd2, 0xffd7}, {0xffda, 0xffdc}, {0xffe0, 0xffe6}, {0xffe8, 0xffee}
+#if TCL_UTF_MAX > 4
+    ,{0x10000, 0x1000b}, {0x1000d, 0x10026}, {0x10028, 0x1003a}, {0x1003f, 0x1004d},
+    {0x10050, 0x1005d}, {0x10080, 0x100fa}, {0x10100, 0x10102}, {0x10107, 0x10133},
+    {0x10137, 0x1018e}, {0x10190, 0x1019b}, {0x101d0, 0x101fd}, {0x10280, 0x1029c},
+    {0x102a0, 0x102d0}, {0x102e0, 0x102fb}, {0x10300, 0x10323}, {0x1032d, 0x1034a},
+    {0x10350, 0x1037a}, {0x10380, 0x1039d}, {0x1039f, 0x103c3}, {0x103c8, 0x103d5},
+    {0x10400, 0x1049d}, {0x104a0, 0x104a9}, {0x104b0, 0x104d3}, {0x104d8, 0x104fb},
+    {0x10500, 0x10527}, {0x10530, 0x10563}, {0x10600, 0x10736}, {0x10740, 0x10755},
+    {0x10760, 0x10767}, {0x10800, 0x10805}, {0x1080a, 0x10835}, {0x1083f, 0x10855},
+    {0x10857, 0x1089e}, {0x108a7, 0x108af}, {0x108e0, 0x108f2}, {0x108fb, 0x1091b},
+    {0x1091f, 0x10939}, {0x10980, 0x109b7}, {0x109bc, 0x109cf}, {0x109d2, 0x10a03},
+    {0x10a0c, 0x10a13}, {0x10a15, 0x10a17}, {0x10a19, 0x10a33}, {0x10a38, 0x10a3a},
+    {0x10a3f, 0x10a47}, {0x10a50, 0x10a58}, {0x10a60, 0x10a9f}, {0x10ac0, 0x10ae6},
+    {0x10aeb, 0x10af6}, {0x10b00, 0x10b35}, {0x10b39, 0x10b55}, {0x10b58, 0x10b72},
+    {0x10b78, 0x10b91}, {0x10b99, 0x10b9c}, {0x10ba9, 0x10baf}, {0x10c00, 0x10c48},
+    {0x10c80, 0x10cb2}, {0x10cc0, 0x10cf2}, {0x10cfa, 0x10cff}, {0x10e60, 0x10e7e},
+    {0x11000, 0x1104d}, {0x11052, 0x1106f}, {0x1107f, 0x110bc}, {0x110be, 0x110c1},
+    {0x110d0, 0x110e8}, {0x110f0, 0x110f9}, {0x11100, 0x11134}, {0x11136, 0x11143},
+    {0x11150, 0x11176}, {0x11180, 0x111cd}, {0x111d0, 0x111df}, {0x111e1, 0x111f4},
+    {0x11200, 0x11211}, {0x11213, 0x1123e}, {0x11280, 0x11286}, {0x1128a, 0x1128d},
+    {0x1128f, 0x1129d}, {0x1129f, 0x112a9}, {0x112b0, 0x112ea}, {0x112f0, 0x112f9},
+    {0x11300, 0x11303}, {0x11305, 0x1130c}, {0x11313, 0x11328}, {0x1132a, 0x11330},
+    {0x11335, 0x11339}, {0x1133c, 0x11344}, {0x1134b, 0x1134d}, {0x1135d, 0x11363},
+    {0x11366, 0x1136c}, {0x11370, 0x11374}, {0x11400, 0x11459}, {0x11480, 0x114c7},
+    {0x114d0, 0x114d9}, {0x11580, 0x115b5}, {0x115b8, 0x115dd}, {0x11600, 0x11644},
+    {0x11650, 0x11659}, {0x11660, 0x1166c}, {0x11680, 0x116b7}, {0x116c0, 0x116c9},
+    {0x11700, 0x11719}, {0x1171d, 0x1172b}, {0x11730, 0x1173f}, {0x118a0, 0x118f2},
+    {0x11a00, 0x11a47}, {0x11a50, 0x11a83}, {0x11a86, 0x11a9c}, {0x11a9e, 0x11aa2},
+    {0x11ac0, 0x11af8}, {0x11c00, 0x11c08}, {0x11c0a, 0x11c36}, {0x11c38, 0x11c45},
+    {0x11c50, 0x11c6c}, {0x11c70, 0x11c8f}, {0x11c92, 0x11ca7}, {0x11ca9, 0x11cb6},
+    {0x11d00, 0x11d06}, {0x11d0b, 0x11d36}, {0x11d3f, 0x11d47}, {0x11d50, 0x11d59},
+    {0x12000, 0x12399}, {0x12400, 0x1246e}, {0x12470, 0x12474}, {0x12480, 0x12543},
+    {0x13000, 0x1342e}, {0x14400, 0x14646}, {0x16800, 0x16a38}, {0x16a40, 0x16a5e},
+    {0x16a60, 0x16a69}, {0x16ad0, 0x16aed}, {0x16af0, 0x16af5}, {0x16b00, 0x16b45},
+    {0x16b50, 0x16b59}, {0x16b5b, 0x16b61}, {0x16b63, 0x16b77}, {0x16b7d, 0x16b8f},
+    {0x16f00, 0x16f44}, {0x16f50, 0x16f7e}, {0x16f8f, 0x16f9f}, {0x17000, 0x187ec},
+    {0x18800, 0x18af2}, {0x1b000, 0x1b11e}, {0x1b170, 0x1b2fb}, {0x1bc00, 0x1bc6a},
+    {0x1bc70, 0x1bc7c}, {0x1bc80, 0x1bc88}, {0x1bc90, 0x1bc99}, {0x1bc9c, 0x1bc9f},
+    {0x1d000, 0x1d0f5}, {0x1d100, 0x1d126}, {0x1d129, 0x1d172}, {0x1d17b, 0x1d1e8},
+    {0x1d200, 0x1d245}, {0x1d300, 0x1d356}, {0x1d360, 0x1d371}, {0x1d400, 0x1d454},
+    {0x1d456, 0x1d49c}, {0x1d4a9, 0x1d4ac}, {0x1d4ae, 0x1d4b9}, {0x1d4bd, 0x1d4c3},
+    {0x1d4c5, 0x1d505}, {0x1d507, 0x1d50a}, {0x1d50d, 0x1d514}, {0x1d516, 0x1d51c},
+    {0x1d51e, 0x1d539}, {0x1d53b, 0x1d53e}, {0x1d540, 0x1d544}, {0x1d54a, 0x1d550},
+    {0x1d552, 0x1d6a5}, {0x1d6a8, 0x1d7cb}, {0x1d7ce, 0x1da8b}, {0x1da9b, 0x1da9f},
+    {0x1daa1, 0x1daaf}, {0x1e000, 0x1e006}, {0x1e008, 0x1e018}, {0x1e01b, 0x1e021},
+    {0x1e026, 0x1e02a}, {0x1e800, 0x1e8c4}, {0x1e8c7, 0x1e8d6}, {0x1e900, 0x1e94a},
+    {0x1e950, 0x1e959}, {0x1ee00, 0x1ee03}, {0x1ee05, 0x1ee1f}, {0x1ee29, 0x1ee32},
+    {0x1ee34, 0x1ee37}, {0x1ee4d, 0x1ee4f}, {0x1ee67, 0x1ee6a}, {0x1ee6c, 0x1ee72},
+    {0x1ee74, 0x1ee77}, {0x1ee79, 0x1ee7c}, {0x1ee80, 0x1ee89}, {0x1ee8b, 0x1ee9b},
+    {0x1eea1, 0x1eea3}, {0x1eea5, 0x1eea9}, {0x1eeab, 0x1eebb}, {0x1f000, 0x1f02b},
+    {0x1f030, 0x1f093}, {0x1f0a0, 0x1f0ae}, {0x1f0b1, 0x1f0bf}, {0x1f0c1, 0x1f0cf},
+    {0x1f0d1, 0x1f0f5}, {0x1f100, 0x1f10c}, {0x1f110, 0x1f12e}, {0x1f130, 0x1f16b},
+    {0x1f170, 0x1f1ac}, {0x1f1e6, 0x1f202}, {0x1f210, 0x1f23b}, {0x1f240, 0x1f248},
+    {0x1f260, 0x1f265}, {0x1f300, 0x1f6d4}, {0x1f6e0, 0x1f6ec}, {0x1f6f0, 0x1f6f8},
+    {0x1f700, 0x1f773}, {0x1f780, 0x1f7d4}, {0x1f800, 0x1f80b}, {0x1f810, 0x1f847},
+    {0x1f850, 0x1f859}, {0x1f860, 0x1f887}, {0x1f890, 0x1f8ad}, {0x1f900, 0x1f90b},
+    {0x1f910, 0x1f93e}, {0x1f940, 0x1f94c}, {0x1f950, 0x1f96b}, {0x1f980, 0x1f997},
+    {0x1f9d0, 0x1f9e6}, {0x20000, 0x2a6d6}, {0x2a700, 0x2b734}, {0x2b740, 0x2b81d},
+    {0x2b820, 0x2cea1}, {0x2ceb0, 0x2ebe0}, {0x2f800, 0x2fa1d}, {0xe0100, 0xe01ef}
+#endif
+};
+
+#define NUM_GRAPH_RANGE (sizeof(graphRangeTable)/sizeof(crange))
+
+static const chr graphCharTable[] = {
+    0x38c, 0x589, 0x58a, 0x85e, 0x98f, 0x990, 0x9b2, 0x9c7, 0x9c8,
+    0x9d7, 0x9dc, 0x9dd, 0xa0f, 0xa10, 0xa32, 0xa33, 0xa35, 0xa36,
+    0xa38, 0xa39, 0xa3c, 0xa47, 0xa48, 0xa51, 0xa5e, 0xab2, 0xab3,
+    0xad0, 0xb0f, 0xb10, 0xb32, 0xb33, 0xb47, 0xb48, 0xb56, 0xb57,
+    0xb5c, 0xb5d, 0xb82, 0xb83, 0xb99, 0xb9a, 0xb9c, 0xb9e, 0xb9f,
+    0xba3, 0xba4, 0xbd0, 0xbd7, 0xc55, 0xc56, 0xcd5, 0xcd6, 0xcde,
+    0xcf1, 0xcf2, 0xd82, 0xd83, 0xdbd, 0xdca, 0xdd6, 0xe81, 0xe82,
+    0xe84, 0xe87, 0xe88, 0xe8a, 0xe8d, 0xea5, 0xea7, 0xeaa, 0xeab,
+    0xec6, 0x10c7, 0x10cd, 0x1258, 0x12c0, 0x1772, 0x1773, 0x1940, 0x1f59,
+    0x1f5b, 0x1f5d, 0x2070, 0x2071, 0x2d27, 0x2d2d, 0x2d6f, 0x2d70, 0xfb3e,
+    0xfb40, 0xfb41, 0xfb43, 0xfb44, 0xfffc, 0xfffd
+#if TCL_UTF_MAX > 4
+    ,0x1003c, 0x1003d, 0x101a0, 0x1056f, 0x10808, 0x10837, 0x10838, 0x1083c, 0x108f4,
+    0x108f5, 0x1093f, 0x10a05, 0x10a06, 0x11288, 0x1130f, 0x11310, 0x11332, 0x11333,
+    0x11347, 0x11348, 0x11350, 0x11357, 0x1145b, 0x1145d, 0x118ff, 0x11d08, 0x11d09,
+    0x11d3a, 0x11d3c, 0x11d3d, 0x16a6e, 0x16a6f, 0x16fe0, 0x16fe1, 0x1d49e, 0x1d49f,
+    0x1d4a2, 0x1d4a5, 0x1d4a6, 0x1d4bb, 0x1d546, 0x1e023, 0x1e024, 0x1e95e, 0x1e95f,
+    0x1ee21, 0x1ee22, 0x1ee24, 0x1ee27, 0x1ee39, 0x1ee3b, 0x1ee42, 0x1ee47, 0x1ee49,
+    0x1ee4b, 0x1ee51, 0x1ee52, 0x1ee54, 0x1ee57, 0x1ee59, 0x1ee5b, 0x1ee5d, 0x1ee5f,
+    0x1ee61, 0x1ee62, 0x1ee64, 0x1ee7e, 0x1eef0, 0x1eef1, 0x1f250, 0x1f251, 0x1f9c0
+#endif
+};
+
+#define NUM_GRAPH_CHAR (sizeof(graphCharTable)/sizeof(chr))
+
+/*
+ *	End of auto-generated Unicode character ranges declarations.
+ */
+
+#define	CH	NOCELT
+
+/*
+ - element - map collating-element name to celt
+ ^ static celt element(struct vars *, const chr *, const chr *);
+ */
+static celt
+element(
+    struct vars *v,		/* context */
+    const chr *startp,		/* points to start of name */
+    const chr *endp)		/* points just past end of name */
+{
+    const struct cname *cn;
+    size_t len;
+    Tcl_DString ds;
+    const char *np;
+
+    /*
+     * Generic: one-chr names stand for themselves.
+     */
+
+    assert(startp < endp);
+    len = endp - startp;
+    if (len == 1) {
+	return *startp;
+    }
+
+    NOTE(REG_ULOCALE);
+
+    /*
+     * Search table.
+     */
+
+    Tcl_DStringInit(&ds);
+    np = Tcl_UniCharToUtfDString(startp, (int)len, &ds);
+    for (cn=cnames; cn->name!=NULL; cn++) {
+	if (strlen(cn->name)==len && strncmp(cn->name, np, len)==0) {
+	    break;			/* NOTE BREAK OUT */
+	}
+    }
+    Tcl_DStringFree(&ds);
+    if (cn->name != NULL) {
+	return CHR(cn->code);
+    }
+
+    /*
+     * Couldn't find it.
+     */
+
+    ERR(REG_ECOLLATE);
+    return 0;
+}
+
+/*
+ - range - supply cvec for a range, including legality check
+ ^ static struct cvec *range(struct vars *, celt, celt, int);
+ */
+static struct cvec *
+range(
+    struct vars *v,		/* context */
+    celt a,			/* range start */
+    celt b,			/* range end, might equal a */
+    int cases)			/* case-independent? */
+{
+    int nchrs;
+    struct cvec *cv;
+    celt c, lc, uc, tc;
+
+    if (a != b && !before(a, b)) {
+	ERR(REG_ERANGE);
+	return NULL;
+    }
+
+    if (!cases) {		/* easy version */
+	cv = getcvec(v, 0, 1);
+	NOERRN();
+	addrange(cv, a, b);
+	return cv;
+    }
+
+    /*
+     * When case-independent, it's hard to decide when cvec ranges are usable,
+     * so for now at least, we won't try. We allocate enough space for two
+     * case variants plus a little extra for the two title case variants.
+     */
+
+    nchrs = (b - a + 1)*2 + 4;
+
+    cv = getcvec(v, nchrs, 0);
+    NOERRN();
+
+    for (c=a; c<=b; c++) {
+	addchr(cv, c);
+	lc = Tcl_UniCharToLower((chr)c);
+	uc = Tcl_UniCharToUpper((chr)c);
+	tc = Tcl_UniCharToTitle((chr)c);
+	if (c != lc) {
+	    addchr(cv, lc);
+	}
+	if (c != uc) {
+	    addchr(cv, uc);
+	}
+	if (c != tc && tc != uc) {
+	    addchr(cv, tc);
+	}
+    }
+
+    return cv;
+}
+
+/*
+ - before - is celt x before celt y, for purposes of range legality?
+ ^ static int before(celt, celt);
+ */
+static int			/* predicate */
+before(
+    celt x, celt y)		/* collating elements */
+{
+    if (x < y) {
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ - eclass - supply cvec for an equivalence class
+ * Must include case counterparts on request.
+ ^ static struct cvec *eclass(struct vars *, celt, int);
+ */
+static struct cvec *
+eclass(
+    struct vars *v,		/* context */
+    celt c,			/* Collating element representing the
+				 * equivalence class. */
+    int cases)			/* all cases? */
+{
+    struct cvec *cv;
+
+    /*
+     * Crude fake equivalence class for testing.
+     */
+
+    if ((v->cflags&REG_FAKE) && c == 'x') {
+	cv = getcvec(v, 4, 0);
+	addchr(cv, (chr)'x');
+	addchr(cv, (chr)'y');
+	if (cases) {
+	    addchr(cv, (chr)'X');
+	    addchr(cv, (chr)'Y');
+	}
+	return cv;
+    }
+
+    /*
+     * Otherwise, none.
+     */
+
+    if (cases) {
+	return allcases(v, c);
+    }
+    cv = getcvec(v, 1, 0);
+    assert(cv != NULL);
+    addchr(cv, (chr)c);
+    return cv;
+}
+
+/*
+ - cclass - supply cvec for a character class
+ * Must include case counterparts on request.
+ ^ static struct cvec *cclass(struct vars *, const chr *, const chr *, int);
+ */
+static struct cvec *
+cclass(
+    struct vars *v,		/* context */
+    const chr *startp,		/* where the name starts */
+    const chr *endp,		/* just past the end of the name */
+    int cases)			/* case-independent? */
+{
+    size_t len;
+    struct cvec *cv = NULL;
+    Tcl_DString ds;
+    const char *np;
+    const char *const *namePtr;
+    int i, index;
+
+    /*
+     * The following arrays define the valid character class names.
+     */
+
+    static const char *const classNames[] = {
+	"alnum", "alpha", "ascii", "blank", "cntrl", "digit", "graph",
+	"lower", "print", "punct", "space", "upper", "xdigit", NULL
+    };
+
+    enum classes {
+	CC_ALNUM, CC_ALPHA, CC_ASCII, CC_BLANK, CC_CNTRL, CC_DIGIT, CC_GRAPH,
+	CC_LOWER, CC_PRINT, CC_PUNCT, CC_SPACE, CC_UPPER, CC_XDIGIT
+    };
+
+
+    /*
+     * Extract the class name
+     */
+
+    len = endp - startp;
+    Tcl_DStringInit(&ds);
+    np = Tcl_UniCharToUtfDString(startp, (int)len, &ds);
+
+    /*
+     * Map the name to the corresponding enumerated value.
+     */
+
+    index = -1;
+    for (namePtr=classNames,i=0 ; *namePtr!=NULL ; namePtr++,i++) {
+	if ((strlen(*namePtr) == len) && (strncmp(*namePtr, np, len) == 0)) {
+	    index = i;
+	    break;
+	}
+    }
+    Tcl_DStringFree(&ds);
+    if (index == -1) {
+	ERR(REG_ECTYPE);
+	return NULL;
+    }
+
+    /*
+     * Remap lower and upper to alpha if the match is case insensitive.
+     */
+
+    if (cases && ((index == CC_LOWER) || (index == CC_UPPER))) {
+	index = CC_ALNUM;
+    }
+
+    /*
+     * Now compute the character class contents.
+     */
+
+    switch((enum classes) index) {
+    case CC_ALNUM:
+	cv = getcvec(v, NUM_ALPHA_CHAR, NUM_DIGIT_RANGE + NUM_ALPHA_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_ALPHA_CHAR ; i++) {
+		addchr(cv, alphaCharTable[i]);
+	    }
+	    for (i=0 ; (size_t)i<NUM_ALPHA_RANGE ; i++) {
+		addrange(cv, alphaRangeTable[i].start,
+			alphaRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_DIGIT_RANGE ; i++) {
+		addrange(cv, digitRangeTable[i].start,
+			digitRangeTable[i].end);
+	    }
+	}
+	break;
+    case CC_ALPHA:
+	cv = getcvec(v, NUM_ALPHA_CHAR, NUM_ALPHA_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_ALPHA_RANGE ; i++) {
+		addrange(cv, alphaRangeTable[i].start,
+			alphaRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_ALPHA_CHAR ; i++) {
+		addchr(cv, alphaCharTable[i]);
+	    }
+	}
+	break;
+    case CC_ASCII:
+	cv = getcvec(v, 0, 1);
+	if (cv) {
+	    addrange(cv, 0, 0x7f);
+	}
+	break;
+    case CC_BLANK:
+	cv = getcvec(v, 2, 0);
+	addchr(cv, '\t');
+	addchr(cv, ' ');
+	break;
+    case CC_CNTRL:
+	cv = getcvec(v, NUM_CONTROL_CHAR, NUM_CONTROL_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_CONTROL_RANGE ; i++) {
+		addrange(cv, controlRangeTable[i].start,
+			controlRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_CONTROL_CHAR ; i++) {
+		addchr(cv, controlCharTable[i]);
+	    }
+	}
+	break;
+    case CC_DIGIT:
+	cv = getcvec(v, 0, NUM_DIGIT_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_DIGIT_RANGE ; i++) {
+		addrange(cv, digitRangeTable[i].start,
+			digitRangeTable[i].end);
+	    }
+	}
+	break;
+    case CC_PUNCT:
+	cv = getcvec(v, NUM_PUNCT_CHAR, NUM_PUNCT_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_PUNCT_RANGE ; i++) {
+		addrange(cv, punctRangeTable[i].start,
+			punctRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_PUNCT_CHAR ; i++) {
+		addchr(cv, punctCharTable[i]);
+	    }
+	}
+	break;
+    case CC_XDIGIT:
+	/*
+	 * This is a 3 instead of (NUM_DIGIT_RANGE+2) because I've no idea how
+	 * to define the digits 'a' through 'f' in non-western locales. The
+	 * concept is quite possibly non portable, or only used in contextx
+	 * where the characters used would be the western ones anyway!
+	 * Whatever is actually the case, the number of ranges is fixed (until
+	 * someone comes up with a better arrangement!)
+	 */
+
+	cv = getcvec(v, 0, 3);
+	if (cv) {
+	    addrange(cv, '0', '9');
+	    addrange(cv, 'a', 'f');
+	    addrange(cv, 'A', 'F');
+	}
+	break;
+    case CC_SPACE:
+	cv = getcvec(v, NUM_SPACE_CHAR, NUM_SPACE_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_SPACE_RANGE ; i++) {
+		addrange(cv, spaceRangeTable[i].start,
+			spaceRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_SPACE_CHAR ; i++) {
+		addchr(cv, spaceCharTable[i]);
+	    }
+	}
+	break;
+    case CC_LOWER:
+	cv  = getcvec(v, NUM_LOWER_CHAR, NUM_LOWER_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_LOWER_RANGE ; i++) {
+		addrange(cv, lowerRangeTable[i].start,
+			lowerRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_LOWER_CHAR ; i++) {
+		addchr(cv, lowerCharTable[i]);
+	    }
+	}
+	break;
+    case CC_UPPER:
+	cv  = getcvec(v, NUM_UPPER_CHAR, NUM_UPPER_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_UPPER_RANGE ; i++) {
+		addrange(cv, upperRangeTable[i].start,
+			upperRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_UPPER_CHAR ; i++) {
+		addchr(cv, upperCharTable[i]);
+	    }
+	}
+	break;
+    case CC_PRINT:
+    	cv  = getcvec(v, NUM_SPACE_CHAR + NUM_GRAPH_CHAR, NUM_SPACE_RANGE + NUM_GRAPH_RANGE  - 1);
+    	if (cv) {
+    	    for (i=1 ; (size_t)i<NUM_SPACE_RANGE ; i++) {
+    		addrange(cv, spaceRangeTable[i].start,
+    				spaceRangeTable[i].end);
+    	    }
+    	    for (i=0 ; (size_t)i<NUM_SPACE_CHAR ; i++) {
+    		addchr(cv, spaceCharTable[i]);
+    	    }
+    	    for (i=0 ; (size_t)i<NUM_GRAPH_RANGE ; i++) {
+    		addrange(cv, graphRangeTable[i].start,
+    				graphRangeTable[i].end);
+    	    }
+    	    for (i=0 ; (size_t)i<NUM_GRAPH_CHAR ; i++) {
+    		addchr(cv, graphCharTable[i]);
+    	    }
+    	}
+    	break;
+    case CC_GRAPH:
+	cv  = getcvec(v, NUM_GRAPH_CHAR, NUM_GRAPH_RANGE);
+	if (cv) {
+	    for (i=0 ; (size_t)i<NUM_GRAPH_RANGE ; i++) {
+		addrange(cv, graphRangeTable[i].start,
+			graphRangeTable[i].end);
+	    }
+	    for (i=0 ; (size_t)i<NUM_GRAPH_CHAR ; i++) {
+		addchr(cv, graphCharTable[i]);
+	    }
+	}
+	break;
+    }
+    if (cv == NULL) {
+	ERR(REG_ESPACE);
+    }
+    return cv;
+}
+
+/*
+ - allcases - supply cvec for all case counterparts of a chr (including itself)
+ * This is a shortcut, preferably an efficient one, for simple characters;
+ * messy cases are done via range().
+ ^ static struct cvec *allcases(struct vars *, pchr);
+ */
+static struct cvec *
+allcases(
+    struct vars *v,		/* context */
+    pchr pc)			/* character to get case equivs of */
+{
+    struct cvec *cv;
+    chr c = (chr)pc;
+    chr lc, uc, tc;
+
+    lc = Tcl_UniCharToLower((chr)c);
+    uc = Tcl_UniCharToUpper((chr)c);
+    tc = Tcl_UniCharToTitle((chr)c);
+
+    if (tc != uc) {
+	cv = getcvec(v, 3, 0);
+	addchr(cv, tc);
+    } else {
+	cv = getcvec(v, 2, 0);
+    }
+    addchr(cv, lc);
+    if (lc != uc) {
+	addchr(cv, uc);
+    }
+    return cv;
+}
+
+/*
+ - cmp - chr-substring compare
+ * Backrefs need this.  It should preferably be efficient.
+ * Note that it does not need to report anything except equal/unequal.
+ * Note also that the length is exact, and the comparison should not
+ * stop at embedded NULs!
+ ^ static int cmp(const chr *, const chr *, size_t);
+ */
+static int			/* 0 for equal, nonzero for unequal */
+cmp(
+    const chr *x, const chr *y,	/* strings to compare */
+    size_t len)			/* exact length of comparison */
+{
+    return memcmp((void*)(x), (void*)(y), len*sizeof(chr));
+}
+
+/*
+ - casecmp - case-independent chr-substring compare
+ * REG_ICASE backrefs need this.  It should preferably be efficient.
+ * Note that it does not need to report anything except equal/unequal.
+ * Note also that the length is exact, and the comparison should not
+ * stop at embedded NULs!
+ ^ static int casecmp(const chr *, const chr *, size_t);
+ */
+static int			/* 0 for equal, nonzero for unequal */
+casecmp(
+    const chr *x, const chr *y,	/* strings to compare */
+    size_t len)			/* exact length of comparison */
+{
+    for (; len > 0; len--, x++, y++) {
+	if ((*x!=*y) && (Tcl_UniCharToLower(*x) != Tcl_UniCharToLower(*y))) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regc_nfa.c b/generic/regc_nfa.c
new file mode 100644
index 0000000..240fcfe
--- /dev/null
+++ b/generic/regc_nfa.c
@@ -0,0 +1,3213 @@
+/*
+ * NFA utilities.
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer. All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * One or two things that technically ought to be in here are actually in
+ * color.c, thanks to some incestuous relationships in the color chains.
+ */
+
+#define	NISERR()	VISERR(nfa->v)
+#define	NERR(e)		VERR(nfa->v, (e))
+#define STACK_TOO_DEEP(x) (0)
+#define CANCEL_REQUESTED(x) (0)
+#define REG_CANCEL 777
+
+/*
+ - newnfa - set up an NFA
+ ^ static struct nfa *newnfa(struct vars *, struct colormap *, struct nfa *);
+ */
+static struct nfa *		/* the NFA, or NULL */
+newnfa(
+    struct vars *v,
+    struct colormap *cm,
+    struct nfa *parent)		/* NULL if primary NFA */
+{
+    struct nfa *nfa;
+
+    nfa = (struct nfa *) MALLOC(sizeof(struct nfa));
+    if (nfa == NULL) {
+	ERR(REG_ESPACE);
+	return NULL;
+    }
+
+    nfa->states = NULL;
+    nfa->slast = NULL;
+    nfa->free = NULL;
+    nfa->nstates = 0;
+    nfa->cm = cm;
+    nfa->v = v;
+    nfa->bos[0] = nfa->bos[1] = COLORLESS;
+    nfa->eos[0] = nfa->eos[1] = COLORLESS;
+    nfa->parent = parent;	/* Precedes newfstate so parent is valid. */
+    nfa->post = newfstate(nfa, '@');	/* number 0 */
+    nfa->pre = newfstate(nfa, '>');	/* number 1 */
+
+    nfa->init = newstate(nfa);	/* May become invalid later. */
+    nfa->final = newstate(nfa);
+    if (ISERR()) {
+	freenfa(nfa);
+	return NULL;
+    }
+    rainbow(nfa, nfa->cm, PLAIN, COLORLESS, nfa->pre, nfa->init);
+    newarc(nfa, '^', 1, nfa->pre, nfa->init);
+    newarc(nfa, '^', 0, nfa->pre, nfa->init);
+    rainbow(nfa, nfa->cm, PLAIN, COLORLESS, nfa->final, nfa->post);
+    newarc(nfa, '$', 1, nfa->final, nfa->post);
+    newarc(nfa, '$', 0, nfa->final, nfa->post);
+
+    if (ISERR()) {
+	freenfa(nfa);
+	return NULL;
+    }
+    return nfa;
+}
+
+/*
+ - freenfa - free an entire NFA
+ ^ static void freenfa(struct nfa *);
+ */
+static void
+freenfa(
+    struct nfa *nfa)
+{
+    struct state *s;
+
+    while ((s = nfa->states) != NULL) {
+	s->nins = s->nouts = 0;	/* don't worry about arcs */
+	freestate(nfa, s);
+    }
+    while ((s = nfa->free) != NULL) {
+	nfa->free = s->next;
+	destroystate(nfa, s);
+    }
+
+    nfa->slast = NULL;
+    nfa->nstates = -1;
+    nfa->pre = NULL;
+    nfa->post = NULL;
+    FREE(nfa);
+}
+
+/*
+ - newstate - allocate an NFA state, with zero flag value
+ ^ static struct state *newstate(struct nfa *);
+ */
+static struct state *		/* NULL on error */
+newstate(
+    struct nfa *nfa)
+{
+    struct state *s;
+
+    if (nfa->free != NULL) {
+	s = nfa->free;
+	nfa->free = s->next;
+    } else {
+	if (nfa->v->spaceused >= REG_MAX_COMPILE_SPACE) {
+	    NERR(REG_ETOOBIG);
+	    return NULL;
+	}
+	s = (struct state *) MALLOC(sizeof(struct state));
+	if (s == NULL) {
+	    NERR(REG_ESPACE);
+	    return NULL;
+	}
+	nfa->v->spaceused += sizeof(struct state);
+	s->oas.next = NULL;
+	s->free = NULL;
+	s->noas = 0;
+    }
+
+    assert(nfa->nstates >= 0);
+    s->no = nfa->nstates++;
+    s->flag = 0;
+    if (nfa->states == NULL) {
+	nfa->states = s;
+    }
+    s->nins = 0;
+    s->ins = NULL;
+    s->nouts = 0;
+    s->outs = NULL;
+    s->tmp = NULL;
+    s->next = NULL;
+    if (nfa->slast != NULL) {
+	assert(nfa->slast->next == NULL);
+	nfa->slast->next = s;
+    }
+    s->prev = nfa->slast;
+    nfa->slast = s;
+    return s;
+}
+
+/*
+ - newfstate - allocate an NFA state with a specified flag value
+ ^ static struct state *newfstate(struct nfa *, int flag);
+ */
+static struct state *		/* NULL on error */
+newfstate(
+    struct nfa *nfa,
+    int flag)
+{
+    struct state *s;
+
+    s = newstate(nfa);
+    if (s != NULL) {
+	s->flag = (char) flag;
+    }
+    return s;
+}
+
+/*
+ - dropstate - delete a state's inarcs and outarcs and free it
+ ^ static void dropstate(struct nfa *, struct state *);
+ */
+static void
+dropstate(
+    struct nfa *nfa,
+    struct state *s)
+{
+    struct arc *a;
+
+    while ((a = s->ins) != NULL) {
+	freearc(nfa, a);
+    }
+    while ((a = s->outs) != NULL) {
+	freearc(nfa, a);
+    }
+    freestate(nfa, s);
+}
+
+/*
+ - freestate - free a state, which has no in-arcs or out-arcs
+ ^ static void freestate(struct nfa *, struct state *);
+ */
+static void
+freestate(
+    struct nfa *nfa,
+    struct state *s)
+{
+    assert(s != NULL);
+    assert(s->nins == 0 && s->nouts == 0);
+
+    s->no = FREESTATE;
+    s->flag = 0;
+    if (s->next != NULL) {
+	s->next->prev = s->prev;
+    } else {
+	assert(s == nfa->slast);
+	nfa->slast = s->prev;
+    }
+    if (s->prev != NULL) {
+	s->prev->next = s->next;
+    } else {
+	assert(s == nfa->states);
+	nfa->states = s->next;
+    }
+    s->prev = NULL;
+    s->next = nfa->free;	/* don't delete it, put it on the free list */
+    nfa->free = s;
+}
+
+/*
+ - destroystate - really get rid of an already-freed state
+ ^ static void destroystate(struct nfa *, struct state *);
+ */
+static void
+destroystate(
+    struct nfa *nfa,
+    struct state *s)
+{
+    struct arcbatch *ab;
+    struct arcbatch *abnext;
+
+    assert(s->no == FREESTATE);
+    for (ab=s->oas.next ; ab!=NULL ; ab=abnext) {
+	abnext = ab->next;
+	FREE(ab);
+	nfa->v->spaceused -= sizeof(struct arcbatch);
+    }
+    s->ins = NULL;
+    s->outs = NULL;
+    s->next = NULL;
+    FREE(s);
+    nfa->v->spaceused -= sizeof(struct state);
+}
+
+/*
+ - newarc - set up a new arc within an NFA
+ ^ static void newarc(struct nfa *, int, pcolor, struct state *,
+ ^	struct state *);
+ */
+/*
+ * This function checks to make sure that no duplicate arcs are created.
+ * In general we never want duplicates.
+ */
+static void
+newarc(
+    struct nfa *nfa,
+    int t,
+    pcolor co,
+    struct state *from,
+    struct state *to)
+{
+    struct arc *a;
+
+    assert(from != NULL && to != NULL);
+
+    /* check for duplicate arc, using whichever chain is shorter */
+    if (from->nouts <= to->nins) {
+	for (a = from->outs; a != NULL; a = a->outchain) {
+	    if (a->to == to && a->co == co && a->type == t) {
+		return;
+	    }
+	}
+    } else {
+	for (a = to->ins; a != NULL; a = a->inchain) {
+	    if (a->from == from && a->co == co && a->type == t) {
+		return;
+	    }
+	}
+    }
+
+    /* no dup, so create the arc */
+    createarc(nfa, t, co, from, to);
+}
+
+/*
+ * createarc - create a new arc within an NFA
+ *
+ * This function must *only* be used after verifying that there is no existing
+ * identical arc (same type/color/from/to).
+ */
+static void
+createarc(
+    struct nfa * nfa,
+    int t,
+    pcolor co,
+    struct state * from,
+    struct state * to)
+{
+    struct arc *a;
+
+    /* the arc is physically allocated within its from-state */
+    a = allocarc(nfa, from);
+    if (NISERR()) {
+	return;
+    }
+    assert(a != NULL);
+
+    a->type = t;
+    a->co = (color) co;
+    a->to = to;
+    a->from = from;
+
+    /*
+     * Put the new arc on the beginning, not the end, of the chains; it's
+     * simpler here, and freearc() is the same cost either way.  See also the
+     * logic in moveins() and its cohorts, as well as fixempties().
+     */
+    a->inchain = to->ins;
+    a->inchainRev = NULL;
+    if (to->ins) {
+	to->ins->inchainRev = a;
+    }
+    to->ins = a;
+    a->outchain = from->outs;
+    a->outchainRev = NULL;
+    if (from->outs) {
+	from->outs->outchainRev = a;
+    }
+    from->outs = a;
+
+    from->nouts++;
+    to->nins++;
+
+    if (COLORED(a) && nfa->parent == NULL) {
+	colorchain(nfa->cm, a);
+    }
+}
+
+/*
+ - allocarc - allocate a new out-arc within a state
+ ^ static struct arc *allocarc(struct nfa *, struct state *);
+ */
+static struct arc *		/* NULL for failure */
+allocarc(
+    struct nfa *nfa,
+    struct state *s)
+{
+    struct arc *a;
+
+    /*
+     * Shortcut
+     */
+
+    if (s->free == NULL && s->noas < ABSIZE) {
+	a = &s->oas.a[s->noas];
+	s->noas++;
+	return a;
+    }
+
+    /*
+     * if none at hand, get more
+     */
+
+    if (s->free == NULL) {
+	struct arcbatch *newAb;
+	int i;
+
+	if (nfa->v->spaceused >= REG_MAX_COMPILE_SPACE) {
+	    NERR(REG_ETOOBIG);
+	    return NULL;
+	}
+	newAb = (struct arcbatch *) MALLOC(sizeof(struct arcbatch));
+	if (newAb == NULL) {
+	    NERR(REG_ESPACE);
+	    return NULL;
+	}
+	nfa->v->spaceused += sizeof(struct arcbatch);
+	newAb->next = s->oas.next;
+	s->oas.next = newAb;
+
+	for (i=0 ; i<ABSIZE ; i++) {
+	    newAb->a[i].type = 0;
+	    newAb->a[i].freechain = &newAb->a[i+1];
+	}
+	newAb->a[ABSIZE-1].freechain = NULL;
+	s->free = &newAb->a[0];
+    }
+    assert(s->free != NULL);
+
+    a = s->free;
+    s->free = a->freechain;
+    return a;
+}
+
+/*
+ - freearc - free an arc
+ ^ static void freearc(struct nfa *, struct arc *);
+ */
+static void
+freearc(
+    struct nfa *nfa,
+    struct arc *victim)
+{
+    struct state *from = victim->from;
+    struct state *to = victim->to;
+    struct arc *predecessor;
+
+    assert(victim->type != 0);
+
+    /*
+     * Take it off color chain if necessary.
+     */
+
+    if (COLORED(victim) && nfa->parent == NULL) {
+	uncolorchain(nfa->cm, victim);
+    }
+
+    /*
+     * Take it off source's out-chain.
+     */
+
+    assert(from != NULL);
+    predecessor = victim->outchainRev;
+    if (predecessor == NULL) {
+	assert(from->outs == victim);
+	from->outs = victim->outchain;
+    } else {
+	assert(predecessor->outchain == victim);
+	predecessor->outchain = victim->outchain;
+    }
+    if (victim->outchain != NULL) {
+	assert(victim->outchain->outchainRev == victim);
+	victim->outchain->outchainRev = predecessor;
+    }
+    from->nouts--;
+
+    /*
+     * Take it off target's in-chain.
+     */
+
+    assert(to != NULL);
+    predecessor = victim->inchainRev;
+    if (predecessor == NULL) {
+	assert(to->ins == victim);
+	to->ins = victim->inchain;
+    } else {
+	assert(predecessor->inchain == victim);
+	predecessor->inchain = victim->inchain;
+    }
+    if (victim->inchain != NULL) {
+	assert(victim->inchain->inchainRev == victim);
+	victim->inchain->inchainRev = predecessor;
+    }
+    to->nins--;
+
+    /*
+     * Clean up and place on from-state's free list.
+     */
+
+    victim->type = 0;
+    victim->from = NULL;	/* precautions... */
+    victim->to = NULL;
+    victim->inchain = NULL;
+    victim->inchainRev = NULL;
+    victim->outchain = NULL;
+    victim->outchainRev = NULL;
+    victim->freechain = from->free;
+    from->free = victim;
+}
+
+/*
+ * changearctarget - flip an arc to have a different to state
+ *
+ * Caller must have verified that there is no pre-existing duplicate arc.
+ *
+ * Note that because we store arcs in their from state, we can't easily have
+ * a similar changearcsource function.
+ */
+static void
+changearctarget(struct arc * a, struct state * newto)
+{
+    struct state *oldto = a->to;
+    struct arc *predecessor;
+
+    assert(oldto != newto);
+
+    /* take it off old target's in-chain */
+    assert(oldto != NULL);
+    predecessor = a->inchainRev;
+    if (predecessor == NULL) {
+	assert(oldto->ins == a);
+	oldto->ins = a->inchain;
+    } else {
+	assert(predecessor->inchain == a);
+	predecessor->inchain = a->inchain;
+    }
+    if (a->inchain != NULL) {
+	assert(a->inchain->inchainRev == a);
+	a->inchain->inchainRev = predecessor;
+    }
+    oldto->nins--;
+
+    a->to = newto;
+
+    /* prepend it to new target's in-chain */
+    a->inchain = newto->ins;
+    a->inchainRev = NULL;
+    if (newto->ins) {
+	newto->ins->inchainRev = a;
+    }
+    newto->ins = a;
+    newto->nins++;
+}
+
+/*
+ - hasnonemptyout - Does state have a non-EMPTY out arc?
+ ^ static int hasnonemptyout(struct state *);
+ */
+static int
+hasnonemptyout(
+    struct state *s)
+{
+    struct arc *a;
+
+    for (a = s->outs; a != NULL; a = a->outchain) {
+	if (a->type != EMPTY) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ - findarc - find arc, if any, from given source with given type and color
+ * If there is more than one such arc, the result is random.
+ ^ static struct arc *findarc(struct state *, int, pcolor);
+ */
+static struct arc *
+findarc(
+    struct state *s,
+    int type,
+    pcolor co)
+{
+    struct arc *a;
+
+    for (a=s->outs ; a!=NULL ; a=a->outchain) {
+	if (a->type == type && a->co == co) {
+	    return a;
+	}
+    }
+    return NULL;
+}
+
+/*
+ - cparc - allocate a new arc within an NFA, copying details from old one
+ ^ static void cparc(struct nfa *, struct arc *, struct state *,
+ ^ 	struct state *);
+ */
+static void
+cparc(
+    struct nfa *nfa,
+    struct arc *oa,
+    struct state *from,
+    struct state *to)
+{
+    newarc(nfa, oa->type, oa->co, from, to);
+}
+
+/*
+ * sortins - sort the in arcs of a state by from/color/type
+ */
+static void
+sortins(
+    struct nfa * nfa,
+    struct state * s)
+{
+    struct arc **sortarray;
+    struct arc *a;
+    int n = s->nins;
+    int i;
+
+    if (n <= 1) {
+	return;		/* nothing to do */
+    }
+    /* make an array of arc pointers ... */
+    sortarray = (struct arc **) MALLOC(n * sizeof(struct arc *));
+    if (sortarray == NULL) {
+	NERR(REG_ESPACE);
+	return;
+    }
+    i = 0;
+    for (a = s->ins; a != NULL; a = a->inchain) {
+	sortarray[i++] = a;
+    }
+    assert(i == n);
+    /* ... sort the array */
+    qsort(sortarray, n, sizeof(struct arc *), sortins_cmp);
+    /* ... and rebuild arc list in order */
+    /* it seems worth special-casing first and last items to simplify loop */
+    a = sortarray[0];
+    s->ins = a;
+    a->inchain = sortarray[1];
+    a->inchainRev = NULL;
+    for (i = 1; i < n - 1; i++) {
+	a = sortarray[i];
+	a->inchain = sortarray[i + 1];
+	a->inchainRev = sortarray[i - 1];
+    }
+    a = sortarray[i];
+    a->inchain = NULL;
+    a->inchainRev = sortarray[i - 1];
+    FREE(sortarray);
+}
+
+static int
+sortins_cmp(
+    const void *a,
+    const void *b)
+{
+    const struct arc *aa = *((const struct arc * const *) a);
+    const struct arc *bb = *((const struct arc * const *) b);
+
+    /* we check the fields in the order they are most likely to be different */
+    if (aa->from->no < bb->from->no) {
+	return -1;
+    }
+    if (aa->from->no > bb->from->no) {
+ 	return 1;
+    }
+    if (aa->co < bb->co) {
+ 	return -1;
+    }
+    if (aa->co > bb->co) {
+ 	return 1;
+    }
+    if (aa->type < bb->type) {
+ 	return -1;
+    }
+    if (aa->type > bb->type) {
+ 	return 1;
+    }
+    return 0;
+}
+
+/*
+ * sortouts - sort the out arcs of a state by to/color/type
+ */
+static void
+sortouts(
+    struct nfa * nfa,
+    struct state * s)
+{
+    struct arc **sortarray;
+    struct arc *a;
+    int	n = s->nouts;
+    int	i;
+
+    if (n <= 1) {
+	return;					/* nothing to do */
+    }
+    /* make an array of arc pointers ... */
+    sortarray = (struct arc **) MALLOC(n * sizeof(struct arc *));
+    if (sortarray == NULL) {
+	NERR(REG_ESPACE);
+	return;
+    }
+    i = 0;
+    for (a = s->outs; a != NULL; a = a->outchain) {
+	sortarray[i++] = a;
+    }
+    assert(i == n);
+    /* ... sort the array */
+    qsort(sortarray, n, sizeof(struct arc *), sortouts_cmp);
+    /* ... and rebuild arc list in order */
+    /* it seems worth special-casing first and last items to simplify loop */
+    a = sortarray[0];
+    s->outs = a;
+    a->outchain = sortarray[1];
+    a->outchainRev = NULL;
+    for (i = 1; i < n - 1; i++) {
+	a = sortarray[i];
+	a->outchain = sortarray[i + 1];
+	a->outchainRev = sortarray[i - 1];
+    }
+    a = sortarray[i];
+    a->outchain = NULL;
+    a->outchainRev = sortarray[i - 1];
+    FREE(sortarray);
+}
+
+static int
+sortouts_cmp(
+    const void *a,
+    const void *b)
+{
+    const struct arc *aa = *((const struct arc * const *) a);
+    const struct arc *bb = *((const struct arc * const *) b);
+
+    /* we check the fields in the order they are most likely to be different */
+    if (aa->to->no < bb->to->no) {
+	return -1;
+    }
+    if (aa->to->no > bb->to->no) {
+	return 1;
+    }
+    if (aa->co < bb->co) {
+	return -1;
+    }
+    if (aa->co > bb->co) {
+	return 1;
+    }
+    if (aa->type < bb->type) {
+	return -1;
+    }
+    if (aa->type > bb->type) {
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ * Common decision logic about whether to use arc-by-arc operations or
+ * sort/merge.  If there's just a few source arcs we cannot recoup the
+ * cost of sorting the destination arc list, no matter how large it is.
+ * Otherwise, limit the number of arc-by-arc comparisons to about 1000
+ * (a somewhat arbitrary choice, but the breakeven point would probably
+ * be machine dependent anyway).
+ */
+#define BULK_ARC_OP_USE_SORT(nsrcarcs, ndestarcs) \
+	((nsrcarcs) < 4 ? 0 : ((nsrcarcs) > 32 || (ndestarcs) > 32))
+
+/*
+ - moveins - move all in arcs of a state to another state
+ * You might think this could be done better by just updating the
+ * existing arcs, and you would be right if it weren't for the need
+ * for duplicate suppression, which makes it easier to just make new
+ * ones to exploit the suppression built into newarc.
+ *
+ * However, if we have a whole lot of arcs to deal with, retail duplicate
+ * checks become too slow.  In that case we proceed by sorting and merging
+ * the arc lists, and then we can indeed just update the arcs in-place.
+ *
+ ^ static void moveins(struct nfa *, struct state *, struct state *);
+ */
+static void
+moveins(
+    struct nfa *nfa,
+    struct state *oldState,
+    struct state *newState)
+{
+    assert(oldState != newState);
+
+    if (!BULK_ARC_OP_USE_SORT(oldState->nins, newState->nins)) {
+	/* With not too many arcs, just do them one at a time */
+	struct arc *a;
+
+	while ((a = oldState->ins) != NULL) {
+	    cparc(nfa, a, a->from, newState);
+	    freearc(nfa, a);
+	}
+    } else {
+	/*
+	 * With many arcs, use a sort-merge approach.  Note changearctarget()
+	 * will put the arc onto the front of newState's chain, so it does not
+	 * break our walk through the sorted part of the chain.
+	 */
+	struct arc *oa;
+	struct arc *na;
+
+	/*
+	 * Because we bypass newarc() in this code path, we'd better include a
+	 * cancel check.
+	 */
+	if (CANCEL_REQUESTED(nfa->v->re)) {
+	    NERR(REG_CANCEL);
+	    return;
+	}
+
+	sortins(nfa, oldState);
+	sortins(nfa, newState);
+	if (NISERR()) {
+	    return;		/* might have failed to sort */
+	}
+	oa = oldState->ins;
+	na = newState->ins;
+	while (oa != NULL && na != NULL) {
+	    struct arc *a = oa;
+
+	    switch (sortins_cmp(&oa, &na)) {
+		case -1:
+		    /* newState does not have anything matching oa */
+		    oa = oa->inchain;
+
+		    /*
+		     * Rather than doing createarc+freearc, we can just unlink
+		     * and relink the existing arc struct.
+		     */
+		    changearctarget(a, newState);
+		    break;
+		case 0:
+		    /* match, advance in both lists */
+		    oa = oa->inchain;
+		    na = na->inchain;
+		    /* ... and drop duplicate arc from oldState */
+		    freearc(nfa, a);
+		    break;
+		case +1:
+		    /* advance only na; oa might have a match later */
+		    na = na->inchain;
+		    break;
+		default:
+		    assert(NOTREACHED);
+	    }
+	}
+	while (oa != NULL) {
+	    /* newState does not have anything matching oa */
+	    struct arc *a = oa;
+
+	    oa = oa->inchain;
+	    changearctarget(a, newState);
+	}
+    }
+
+    assert(oldState->nins == 0);
+    assert(oldState->ins == NULL);
+}
+
+/*
+ - copyins - copy in arcs of a state to another state
+ ^ static void copyins(struct nfa *, struct state *, struct state *, int);
+ */
+static void
+copyins(
+    struct nfa *nfa,
+    struct state *oldState,
+    struct state *newState)
+{
+    assert(oldState != newState);
+
+    if (!BULK_ARC_OP_USE_SORT(oldState->nins, newState->nins)) {
+	/* With not too many arcs, just do them one at a time */
+	struct arc *a;
+
+	for (a = oldState->ins; a != NULL; a = a->inchain) {
+	    cparc(nfa, a, a->from, newState);
+	}
+    } else {
+	/*
+	 * With many arcs, use a sort-merge approach.  Note that createarc()
+	 * will put new arcs onto the front of newState's chain, so it does
+	 * not break our walk through the sorted part of the chain.
+	 */
+	struct arc *oa;
+	struct arc *na;
+
+	/*
+	 * Because we bypass newarc() in this code path, we'd better include a
+	 * cancel check.
+	 */
+	if (CANCEL_REQUESTED(nfa->v->re)) {
+	    NERR(REG_CANCEL);
+	    return;
+	}
+
+	sortins(nfa, oldState);
+	sortins(nfa, newState);
+	if (NISERR()) {
+	    return;		/* might have failed to sort */
+	}
+	oa = oldState->ins;
+	na = newState->ins;
+	while (oa != NULL && na != NULL) {
+	    struct arc *a = oa;
+
+	    switch (sortins_cmp(&oa, &na)) {
+		case -1:
+		    /* newState does not have anything matching oa */
+		    oa = oa->inchain;
+		    createarc(nfa, a->type, a->co, a->from, newState);
+		    break;
+		case 0:
+		    /* match, advance in both lists */
+		    oa = oa->inchain;
+		    na = na->inchain;
+		    break;
+		case +1:
+		    /* advance only na; oa might have a match later */
+		    na = na->inchain;
+		    break;
+		default:
+		    assert(NOTREACHED);
+	    }
+	}
+	while (oa != NULL) {
+	    /* newState does not have anything matching oa */
+	    struct arc *a = oa;
+
+	    oa = oa->inchain;
+	    createarc(nfa, a->type, a->co, a->from, newState);
+	}
+    }
+}
+
+/*
+ * mergeins - merge a list of inarcs into a state
+ *
+ * This is much like copyins, but the source arcs are listed in an array,
+ * and are not guaranteed unique.  It's okay to clobber the array contents.
+ */
+static void
+mergeins(
+    struct nfa * nfa,
+    struct state * s,
+    struct arc ** arcarray,
+    int arccount)
+{
+    struct arc *na;
+    int	i;
+    int	j;
+
+    if (arccount <= 0) {
+	return;
+    }
+
+    /*
+     * Because we bypass newarc() in this code path, we'd better include a
+     * cancel check.
+     */
+    if (CANCEL_REQUESTED(nfa->v->re)) {
+	NERR(REG_CANCEL);
+	return;
+    }
+
+    /* Sort existing inarcs as well as proposed new ones */
+    sortins(nfa, s);
+    if (NISERR()) {
+	return;			/* might have failed to sort */
+    }
+
+    qsort(arcarray, arccount, sizeof(struct arc *), sortins_cmp);
+
+    /*
+     * arcarray very likely includes dups, so we must eliminate them.  (This
+     * could be folded into the next loop, but it's not worth the trouble.)
+     */
+    j = 0;
+    for (i = 1; i < arccount; i++) {
+	switch (sortins_cmp(&arcarray[j], &arcarray[i])) {
+	    case -1:
+		/* non-dup */
+		arcarray[++j] = arcarray[i];
+		break;
+	    case 0:
+		/* dup */
+		break;
+	    default:
+		/* trouble */
+		assert(NOTREACHED);
+	}
+    }
+    arccount = j + 1;
+
+    /*
+     * Now merge into s' inchain.  Note that createarc() will put new arcs
+     * onto the front of s's chain, so it does not break our walk through the
+     * sorted part of the chain.
+     */
+    i = 0;
+    na = s->ins;
+    while (i < arccount && na != NULL) {
+	struct arc *a = arcarray[i];
+
+	switch (sortins_cmp(&a, &na)) {
+	    case -1:
+		/* s does not have anything matching a */
+		createarc(nfa, a->type, a->co, a->from, s);
+		i++;
+		break;
+	    case 0:
+		/* match, advance in both lists */
+		i++;
+		na = na->inchain;
+		break;
+	    case +1:
+		/* advance only na; array might have a match later */
+		na = na->inchain;
+		break;
+	    default:
+		assert(NOTREACHED);
+	}
+    }
+    while (i < arccount) {
+	/* s does not have anything matching a */
+	struct arc *a = arcarray[i];
+
+	createarc(nfa, a->type, a->co, a->from, s);
+	i++;
+    }
+}
+
+/*
+ - moveouts - move all out arcs of a state to another state
+ ^ static void moveouts(struct nfa *, struct state *, struct state *);
+ */
+static void
+moveouts(
+    struct nfa *nfa,
+    struct state *oldState,
+    struct state *newState)
+{
+    assert(oldState != newState);
+
+    if (!BULK_ARC_OP_USE_SORT(oldState->nouts, newState->nouts)) {
+	/* With not too many arcs, just do them one at a time */
+	struct arc *a;
+
+	while ((a = oldState->outs) != NULL) {
+	    cparc(nfa, a, newState, a->to);
+	    freearc(nfa, a);
+	}
+    } else {
+	/*
+	 * With many arcs, use a sort-merge approach.  Note that createarc()
+	 * will put new arcs onto the front of newState's chain, so it does
+	 * not break our walk through the sorted part of the chain.
+	 */
+	struct arc *oa;
+	struct arc *na;
+
+	/*
+	 * Because we bypass newarc() in this code path, we'd better include a
+	 * cancel check.
+	 */
+	if (CANCEL_REQUESTED(nfa->v->re)) {
+	    NERR(REG_CANCEL);
+	    return;
+	}
+
+	sortouts(nfa, oldState);
+	sortouts(nfa, newState);
+	if (NISERR()) {
+	    return;	/* might have failed to sort */
+	}
+	oa = oldState->outs;
+	na = newState->outs;
+	while (oa != NULL && na != NULL) {
+	    struct arc *a = oa;
+
+	    switch (sortouts_cmp(&oa, &na)) {
+		case -1:
+		    /* newState does not have anything matching oa */
+		    oa = oa->outchain;
+		    createarc(nfa, a->type, a->co, newState, a->to);
+		    freearc(nfa, a);
+		    break;
+		case 0:
+		    /* match, advance in both lists */
+		    oa = oa->outchain;
+		    na = na->outchain;
+		    /* ... and drop duplicate arc from oldState */
+		    freearc(nfa, a);
+		    break;
+		case +1:
+		    /* advance only na; oa might have a match later */
+		    na = na->outchain;
+		    break;
+		default:
+		    assert(NOTREACHED);
+	    }
+	}
+	while (oa != NULL) {
+	    /* newState does not have anything matching oa */
+	    struct arc *a = oa;
+
+	    oa = oa->outchain;
+	    createarc(nfa, a->type, a->co, newState, a->to);
+	    freearc(nfa, a);
+	}
+    }
+
+    assert(oldState->nouts == 0);
+    assert(oldState->outs == NULL);
+}
+
+/*
+ - copyouts - copy out arcs of a state to another state
+ ^ static void copyouts(struct nfa *, struct state *, struct state *, int);
+ */
+static void
+copyouts(
+    struct nfa *nfa,
+    struct state *oldState,
+    struct state *newState)
+{
+    assert(oldState != newState);
+
+    if (!BULK_ARC_OP_USE_SORT(oldState->nouts, newState->nouts)) {
+	/* With not too many arcs, just do them one at a time */
+	struct arc *a;
+
+	for (a = oldState->outs; a != NULL; a = a->outchain) {
+	    cparc(nfa, a, newState, a->to);
+	}
+    } else {
+ 	/*
+	 * With many arcs, use a sort-merge approach.  Note that createarc()
+	 * will put new arcs onto the front of newState's chain, so it does
+	 * not break our walk through the sorted part of the chain.
+	 */
+	struct arc *oa;
+	struct arc *na;
+
+	/*
+	 * Because we bypass newarc() in this code path, we'd better include a
+	 * cancel check.
+	 */
+	if (CANCEL_REQUESTED(nfa->v->re)) {
+	    NERR(REG_CANCEL);
+	    return;
+	}
+
+	sortouts(nfa, oldState);
+	sortouts(nfa, newState);
+	if (NISERR()) {
+	    return;		/* might have failed to sort */
+	}
+	oa = oldState->outs;
+	na = newState->outs;
+	while (oa != NULL && na != NULL) {
+	    struct arc *a = oa;
+
+	    switch (sortouts_cmp(&oa, &na)) {
+		case -1:
+		    /* newState does not have anything matching oa */
+		    oa = oa->outchain;
+		    createarc(nfa, a->type, a->co, newState, a->to);
+		    break;
+		case 0:
+		    /* match, advance in both lists */
+		    oa = oa->outchain;
+		    na = na->outchain;
+		    break;
+		case +1:
+		    /* advance only na; oa might have a match later */
+		    na = na->outchain;
+		    break;
+		default:
+		    assert(NOTREACHED);
+	    }
+	}
+	while (oa != NULL) {
+	    /* newState does not have anything matching oa */
+	    struct arc *a = oa;
+
+	    oa = oa->outchain;
+	    createarc(nfa, a->type, a->co, newState, a->to);
+	}
+    }
+}
+
+/*
+ - cloneouts - copy out arcs of a state to another state pair, modifying type
+ ^ static void cloneouts(struct nfa *, struct state *, struct state *,
+ ^ 	struct state *, int);
+ */
+static void
+cloneouts(
+    struct nfa *nfa,
+    struct state *old,
+    struct state *from,
+    struct state *to,
+    int type)
+{
+    struct arc *a;
+
+    assert(old != from);
+
+    for (a=old->outs ; a!=NULL ; a=a->outchain) {
+	newarc(nfa, type, a->co, from, to);
+    }
+}
+
+/*
+ - delsub - delete a sub-NFA, updating subre pointers if necessary
+ * This uses a recursive traversal of the sub-NFA, marking already-seen
+ * states using their tmp pointer.
+ ^ static void delsub(struct nfa *, struct state *, struct state *);
+ */
+static void
+delsub(
+    struct nfa *nfa,
+    struct state *lp,		/* the sub-NFA goes from here... */
+    struct state *rp)		/* ...to here, *not* inclusive */
+{
+    assert(lp != rp);
+
+    rp->tmp = rp;		/* mark end */
+
+    deltraverse(nfa, lp, lp);
+    assert(lp->nouts == 0 && rp->nins == 0);	/* did the job */
+    assert(lp->no != FREESTATE && rp->no != FREESTATE);	/* no more */
+
+    rp->tmp = NULL;		/* unmark end */
+    lp->tmp = NULL;		/* and begin, marked by deltraverse */
+}
+
+/*
+ - deltraverse - the recursive heart of delsub
+ * This routine's basic job is to destroy all out-arcs of the state.
+ ^ static void deltraverse(struct nfa *, struct state *, struct state *);
+ */
+static void
+deltraverse(
+    struct nfa *nfa,
+    struct state *leftend,
+    struct state *s)
+{
+    struct arc *a;
+    struct state *to;
+
+    if (s->nouts == 0) {
+	return;			/* nothing to do */
+    }
+    if (s->tmp != NULL) {
+	return;			/* already in progress */
+    }
+
+    s->tmp = s;			/* mark as in progress */
+
+    while ((a = s->outs) != NULL) {
+	to = a->to;
+	deltraverse(nfa, leftend, to);
+	assert(to->nouts == 0 || to->tmp != NULL);
+	freearc(nfa, a);
+	if (to->nins == 0 && to->tmp == NULL) {
+	    assert(to->nouts == 0);
+	    freestate(nfa, to);
+	}
+    }
+
+    assert(s->no != FREESTATE);	/* we're still here */
+    assert(s == leftend || s->nins != 0);	/* and still reachable */
+    assert(s->nouts == 0);	/* but have no outarcs */
+
+    s->tmp = NULL;		/* we're done here */
+}
+
+/*
+ - dupnfa - duplicate sub-NFA
+ * Another recursive traversal, this time using tmp to point to duplicates as
+ * well as mark already-seen states. (You knew there was a reason why it's a
+ * state pointer, didn't you? :-))
+ ^ static void dupnfa(struct nfa *, struct state *, struct state *,
+ ^ 	struct state *, struct state *);
+ */
+static void
+dupnfa(
+    struct nfa *nfa,
+    struct state *start,	/* duplicate of subNFA starting here */
+    struct state *stop,		/* and stopping here */
+    struct state *from,		/* stringing duplicate from here */
+    struct state *to)		/* to here */
+{
+    if (start == stop) {
+	newarc(nfa, EMPTY, 0, from, to);
+	return;
+    }
+
+    stop->tmp = to;
+    duptraverse(nfa, start, from, 0);
+    /* done, except for clearing out the tmp pointers */
+
+    stop->tmp = NULL;
+    cleartraverse(nfa, start);
+}
+
+/*
+ - duptraverse - recursive heart of dupnfa
+ ^ static void duptraverse(struct nfa *, struct state *, struct state *);
+ */
+static void
+duptraverse(
+    struct nfa *nfa,
+    struct state *s,
+    struct state *stmp,		/* s's duplicate, or NULL */
+    int depth)
+{
+    struct arc *a;
+
+    if (s->tmp != NULL) {
+	return;			/* already done */
+    }
+
+    s->tmp = (stmp == NULL) ? newstate(nfa) : stmp;
+    if (s->tmp == NULL) {
+	assert(NISERR());
+	return;
+    }
+
+    /*
+     * Arbitrary depth limit. Needs tuning, but this value is sufficient to
+     * make all normal tests (not reg-33.14) pass.
+     */
+#ifndef DUPTRAVERSE_MAX_DEPTH
+#define DUPTRAVERSE_MAX_DEPTH 15000
+#endif
+
+    if (depth++ > DUPTRAVERSE_MAX_DEPTH) {
+	NERR(REG_ESPACE);
+    }
+
+    for (a=s->outs ; a!=NULL && !NISERR() ; a=a->outchain) {
+	duptraverse(nfa, a->to, NULL, depth);
+	if (NISERR()) {
+	    break;
+	}
+	assert(a->to->tmp != NULL);
+	cparc(nfa, a, s->tmp, a->to->tmp);
+    }
+}
+
+/*
+ - cleartraverse - recursive cleanup for algorithms that leave tmp ptrs set
+ ^ static void cleartraverse(struct nfa *, struct state *);
+ */
+static void
+cleartraverse(
+    struct nfa *nfa,
+    struct state *s)
+{
+    struct arc *a;
+
+    if (s->tmp == NULL) {
+	return;
+    }
+    s->tmp = NULL;
+
+    for (a=s->outs ; a!=NULL ; a=a->outchain) {
+	cleartraverse(nfa, a->to);
+    }
+}
+
+/*
+ - specialcolors - fill in special colors for an NFA
+ ^ static void specialcolors(struct nfa *);
+ */
+static void
+specialcolors(
+    struct nfa *nfa)
+{
+    /*
+     * False colors for BOS, BOL, EOS, EOL
+     */
+
+    if (nfa->parent == NULL) {
+	nfa->bos[0] = pseudocolor(nfa->cm);
+	nfa->bos[1] = pseudocolor(nfa->cm);
+	nfa->eos[0] = pseudocolor(nfa->cm);
+	nfa->eos[1] = pseudocolor(nfa->cm);
+    } else {
+	assert(nfa->parent->bos[0] != COLORLESS);
+	nfa->bos[0] = nfa->parent->bos[0];
+	assert(nfa->parent->bos[1] != COLORLESS);
+	nfa->bos[1] = nfa->parent->bos[1];
+	assert(nfa->parent->eos[0] != COLORLESS);
+	nfa->eos[0] = nfa->parent->eos[0];
+	assert(nfa->parent->eos[1] != COLORLESS);
+	nfa->eos[1] = nfa->parent->eos[1];
+    }
+}
+
+/*
+ - optimize - optimize an NFA
+ ^ static long optimize(struct nfa *, FILE *);
+ */
+
+ /*
+  * The main goal of this function is not so much "optimization" (though it
+  * does try to get rid of useless NFA states) as reducing the NFA to a form
+  * the regex executor can handle.  The executor, and indeed the cNFA format
+  * that is its input, can only handle PLAIN and LACON arcs.  The output of
+  * the regex parser also includes EMPTY (do-nothing) arcs, as well as
+  * ^, $, AHEAD, and BEHIND constraint arcs, which we must get rid of here.
+  * We first get rid of EMPTY arcs and then deal with the constraint arcs.
+  * The hardest part of either job is to get rid of circular loops of the
+  * target arc type.  We would have to do that in any case, though, as such a
+  * loop would otherwise allow the executor to cycle through the loop endlessly
+  * without making any progress in the input string.
+  */
+static long			/* re_info bits */
+optimize(
+    struct nfa *nfa,
+    FILE *f)			/* for debug output; NULL none */
+{
+    int verbose = (f != NULL) ? 1 : 0;
+
+    if (verbose) {
+	fprintf(f, "\ninitial cleanup:\n");
+    }
+    cleanup(nfa);		/* may simplify situation */
+    if (verbose) {
+	dumpnfa(nfa, f);
+    }
+    if (verbose) {
+	fprintf(f, "\nempties:\n");
+    }
+    fixempties(nfa, f);		/* get rid of EMPTY arcs */
+    if (verbose) {
+	fprintf(f, "\nconstraints:\n");
+    }
+    fixconstraintloops(nfa, f);	/* get rid of constraint loops */
+    pullback(nfa, f);		/* pull back constraints backward */
+    pushfwd(nfa, f);		/* push fwd constraints forward */
+    if (verbose) {
+	fprintf(f, "\nfinal cleanup:\n");
+    }
+    cleanup(nfa);		/* final tidying */
+#ifdef REG_DEBUG
+    if (verbose) {
+	dumpnfa(nfa, f);
+    }
+#endif
+    return analyze(nfa);	/* and analysis */
+}
+
+/*
+ - pullback - pull back constraints backward to eliminate them
+ ^ static void pullback(struct nfa *, FILE *);
+ */
+static void
+pullback(
+    struct nfa *nfa,
+    FILE *f)			/* for debug output; NULL none */
+{
+    struct state *s;
+    struct state *nexts;
+    struct arc *a;
+    struct arc *nexta;
+    struct state *intermediates;
+    int progress;
+
+    /*
+     * Find and pull until there are no more.
+     */
+
+    do {
+	progress = 0;
+	for (s=nfa->states ; s!=NULL && !NISERR() ; s=nexts) {
+	    nexts = s->next;
+	    intermediates = NULL;
+	    for (a=s->outs ; a!=NULL && !NISERR() ; a=nexta) {
+		nexta = a->outchain;
+		if (a->type == '^' || a->type == BEHIND) {
+		    if (pull(nfa, a, &intermediates)) {
+			progress = 1;
+		    }
+		}
+		assert(nexta == NULL || s->no != FREESTATE);
+	    }
+	    /* clear tmp fields of intermediate states created here */
+	    while (intermediates != NULL) {
+		struct state *ns = intermediates->tmp;
+
+		intermediates->tmp = NULL;
+		intermediates = ns;
+	    }
+	    /* if s is now useless, get rid of it */
+	    if ((s->nins == 0 || s->nouts == 0) && !s->flag) {
+		dropstate(nfa, s);
+	    }
+	}
+	if (progress && f != NULL) {
+	    dumpnfa(nfa, f);
+	}
+    } while (progress && !NISERR());
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * Any ^ constraints we were able to pull to the start state can now be
+     * replaced by PLAIN arcs referencing the BOS or BOL colors.  There should
+     * be no other ^ or BEHIND arcs left in the NFA, though we do not check
+     * that here (compact() will fail if so).
+     */
+    for (a=nfa->pre->outs ; a!=NULL ; a=nexta) {
+	nexta = a->outchain;
+	if (a->type == '^') {
+	    assert(a->co == 0 || a->co == 1);
+	    newarc(nfa, PLAIN, nfa->bos[a->co], a->from, a->to);
+	    freearc(nfa, a);
+	}
+    }
+}
+
+/*
+ - pull - pull a back constraint backward past its source state
+ *
+ * Returns 1 if successful (which it always is unless the source is the
+ * start state or we have an internal error), 0 if nothing happened.
+ *
+ * A significant property of this function is that it deletes no pre-existing
+ * states, and no outarcs of the constraint's from state other than the given
+ * constraint arc.  This makes the loops in pullback() safe, at the cost that
+ * we may leave useless states behind.  Therefore, we leave it to pullback()
+ * to delete such states.
+ *
+ * If the from state has multiple back-constraint outarcs, and/or multiple
+ * compatible constraint inarcs, we only need to create one new intermediate
+ * state per combination of predecessor and successor states.  *intermediates
+ * points to a list of such intermediate states for this from state (chained
+ * through their tmp fields).
+ ^ static int pull(struct nfa *, struct arc *);
+ */
+static int
+pull(
+    struct nfa *nfa,
+    struct arc *con,
+    struct state **intermediates)
+{
+    struct state *from = con->from;
+    struct state *to = con->to;
+    struct arc *a;
+    struct arc *nexta;
+    struct state *s;
+
+    assert(from != to);		/* should have gotten rid of this earlier */
+    if (from->flag) {		/* can't pull back beyond start */
+	return 0;
+    }
+    if (from->nins == 0) {	/* unreachable */
+	freearc(nfa, con);
+	return 1;
+    }
+
+    /*
+     * First, clone from state if necessary to avoid other outarcs.  This may
+     * seem wasteful, but it simplifies the logic, and we'll get rid of the
+     * clone state again at the bottom.
+     */
+
+    if (from->nouts > 1) {
+	s = newstate(nfa);
+	if (NISERR()) {
+	    return 0;
+	}
+	copyins(nfa, from, s);	/* duplicate inarcs */
+	cparc(nfa, con, s, to);		/* move constraint arc */
+	freearc(nfa, con);
+	if (NISERR()) {
+	    return 0;
+	}
+	from = s;
+	con = from->outs;
+    }
+    assert(from->nouts == 1);
+
+    /*
+     * Propagate the constraint into the from state's inarcs.
+     */
+
+    for (a=from->ins ; a!=NULL && !NISERR(); a=nexta) {
+	nexta = a->inchain;
+	switch (combine(con, a)) {
+	case INCOMPATIBLE:	/* destroy the arc */
+	    freearc(nfa, a);
+	    break;
+	case SATISFIED:		/* no action needed */
+	    break;
+	case COMPATIBLE:	/* swap the two arcs, more or less */
+	    /* need an intermediate state, but might have one already */
+	    for (s = *intermediates; s != NULL; s = s->tmp) {
+		assert(s->nins > 0 && s->nouts > 0);
+		if (s->ins->from == a->from && s->outs->to == to) {
+		    break;
+		}
+	    }
+	    if (s == NULL) {
+		s = newstate(nfa);
+		if (NISERR()) {
+		    return 0;
+		}
+		s->tmp = *intermediates;
+		*intermediates = s;
+	    }
+  	    cparc(nfa, con, a->from, s);
+	    cparc(nfa, a, s, to);
+ 	    freearc(nfa, a);
+  	    break;
+	default:
+	    assert(NOTREACHED);
+	    break;
+	}
+    }
+
+    /*
+     * Remaining inarcs, if any, incorporate the constraint.
+     */
+
+    moveins(nfa, from, to);
+    freearc(nfa, con);
+    /* from state is now useless, but we leave it to pullback() to clean up */
+    return 1;
+}
+
+/*
+ - pushfwd - push forward constraints forward to eliminate them
+ ^ static void pushfwd(struct nfa *, FILE *);
+ */
+static void
+pushfwd(
+    struct nfa *nfa,
+    FILE *f)			/* for debug output; NULL none */
+{
+    struct state *s;
+    struct state *nexts;
+    struct arc *a;
+    struct arc *nexta;
+    struct state *intermediates;
+    int progress;
+
+    /*
+     * Find and push until there are no more.
+     */
+
+    do {
+	progress = 0;
+	for (s=nfa->states ; s!=NULL && !NISERR() ; s=nexts) {
+	    nexts = s->next;
+	    intermediates = NULL;
+	    for (a = s->ins; a != NULL && !NISERR(); a = nexta) {
+		nexta = a->inchain;
+		if (a->type == '$' || a->type == AHEAD) {
+		    if (push(nfa, a, &intermediates)) {
+			progress = 1;
+		    }
+		}
+	    }
+	    /* clear tmp fields of intermediate states created here */
+	    while (intermediates != NULL) {
+		struct state *ns = intermediates->tmp;
+
+		intermediates->tmp = NULL;
+		intermediates = ns;
+	    }
+	    /* if s is now useless, get rid of it */
+	    if ((s->nins == 0 || s->nouts == 0) && !s->flag) {
+		dropstate(nfa, s);
+	    }
+	}
+	if (progress && f != NULL) {
+	    dumpnfa(nfa, f);
+	}
+    } while (progress && !NISERR());
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * Any $ constraints we were able to push to the post state can now be
+     * replaced by PLAIN arcs referencing the EOS or EOL colors.  There should
+     * be no other $ or AHEAD arcs left in the NFA, though we do not check
+     * that here (compact() will fail if so).
+     */
+    for (a = nfa->post->ins; a != NULL; a = nexta) {
+	nexta = a->inchain;
+	if (a->type == '$') {
+	    assert(a->co == 0 || a->co == 1);
+	    newarc(nfa, PLAIN, nfa->eos[a->co], a->from, a->to);
+	    freearc(nfa, a);
+	}
+    }
+}
+
+/*
+ - push - push a forward constraint forward past its destination state
+ *
+ * Returns 1 if successful (which it always is unless the destination is the
+ * post state or we have an internal error), 0 if nothing happened.
+ *
+ * A significant property of this function is that it deletes no pre-existing
+ * states, and no inarcs of the constraint's to state other than the given
+ * constraint arc.  This makes the loops in pushfwd() safe, at the cost that
+ * we may leave useless states behind.  Therefore, we leave it to pushfwd()
+ * to delete such states.
+ *
+ * If the to state has multiple forward-constraint inarcs, and/or multiple
+ * compatible constraint outarcs, we only need to create one new intermediate
+ * state per combination of predecessor and successor states.  *intermediates
+ * points to a list of such intermediate states for this to state (chained
+ * through their tmp fields).
+ ^ static int push(struct nfa *, struct arc *);
+ */
+static int
+push(
+    struct nfa *nfa,
+    struct arc *con,
+    struct state **intermediates)
+{
+    struct state *from = con->from;
+    struct state *to = con->to;
+    struct arc *a;
+    struct arc *nexta;
+    struct state *s;
+
+    assert(to != from);		/* should have gotten rid of this earlier */
+    if (to->flag) {		/* can't push forward beyond end */
+	return 0;
+    }
+    if (to->nouts == 0) {	/* dead end */
+	freearc(nfa, con);
+	return 1;
+    }
+
+    /*
+     * First, clone to state if necessary to avoid other inarcs.  This may
+     * seem wasteful, but it simplifies the logic, and we'll get rid of the
+     * clone state again at the bottom.
+     */
+
+    if (to->nins > 1) {
+	s = newstate(nfa);
+	if (NISERR()) {
+	    return 0;
+	}
+	copyouts(nfa, to, s);		/* duplicate outarcs */
+	cparc(nfa, con, from, s);	/* move constraint arc */
+	freearc(nfa, con);
+	if (NISERR()) {
+	    return 0;
+	}
+	to = s;
+	con = to->ins;
+    }
+    assert(to->nins == 1);
+
+    /*
+     * Propagate the constraint into the to state's outarcs.
+     */
+
+    for (a = to->outs; a != NULL && !NISERR(); a = nexta) {
+	nexta = a->outchain;
+	switch (combine(con, a)) {
+	case INCOMPATIBLE:	/* destroy the arc */
+	    freearc(nfa, a);
+	    break;
+	case SATISFIED:		/* no action needed */
+	    break;
+	case COMPATIBLE:	/* swap the two arcs, more or less */
+	    /* need an intermediate state, but might have one already */
+	    for (s = *intermediates; s != NULL; s = s->tmp) {
+		assert(s->nins > 0 && s->nouts > 0);
+		if (s->ins->from == from && s->outs->to == a->to) {
+		    break;
+		}
+	    }
+	    if (s == NULL) {
+		s = newstate(nfa);
+		if (NISERR()) {
+		    return 0;
+		}
+		s->tmp = *intermediates;
+		*intermediates = s;
+	    }
+	    cparc(nfa, con, s, a->to);
+  	    cparc(nfa, a, from, s);
+  	    freearc(nfa, a);
+  	    break;
+	default:
+	    assert(NOTREACHED);
+	    break;
+	}
+    }
+
+    /*
+     * Remaining outarcs, if any, incorporate the constraint.
+     */
+
+    moveouts(nfa, to, from);
+    freearc(nfa, con);
+    /* to state is now useless, but we leave it to pushfwd() to clean up */
+    return 1;
+}
+
+/*
+ - combine - constraint lands on an arc, what happens?
+ ^ #def	INCOMPATIBLE	1	// destroys arc
+ ^ #def	SATISFIED	2	// constraint satisfied
+ ^ #def	COMPATIBLE	3	// compatible but not satisfied yet
+ ^ static int combine(struct arc *, struct arc *);
+ */
+static int
+combine(
+    struct arc *con,
+    struct arc *a)
+{
+#define CA(ct,at)	(((ct)<<CHAR_BIT) | (at))
+
+    switch (CA(con->type, a->type)) {
+    case CA('^', PLAIN):	/* newlines are handled separately */
+    case CA('$', PLAIN):
+	return INCOMPATIBLE;
+	break;
+    case CA(AHEAD, PLAIN):	/* color constraints meet colors */
+    case CA(BEHIND, PLAIN):
+	if (con->co == a->co) {
+	    return SATISFIED;
+	}
+	return INCOMPATIBLE;
+	break;
+    case CA('^', '^'):		/* collision, similar constraints */
+    case CA('$', '$'):
+    case CA(AHEAD, AHEAD):
+    case CA(BEHIND, BEHIND):
+	if (con->co == a->co) {	/* true duplication */
+	    return SATISFIED;
+	}
+	return INCOMPATIBLE;
+	break;
+    case CA('^', BEHIND):	/* collision, dissimilar constraints */
+    case CA(BEHIND, '^'):
+    case CA('$', AHEAD):
+    case CA(AHEAD, '$'):
+	return INCOMPATIBLE;
+	break;
+    case CA('^', '$'):		/* constraints passing each other */
+    case CA('^', AHEAD):
+    case CA(BEHIND, '$'):
+    case CA(BEHIND, AHEAD):
+    case CA('$', '^'):
+    case CA('$', BEHIND):
+    case CA(AHEAD, '^'):
+    case CA(AHEAD, BEHIND):
+    case CA('^', LACON):
+    case CA(BEHIND, LACON):
+    case CA('$', LACON):
+    case CA(AHEAD, LACON):
+	return COMPATIBLE;
+	break;
+    }
+    assert(NOTREACHED);
+    return INCOMPATIBLE;	/* for benefit of blind compilers */
+}
+
+/*
+ - fixempties - get rid of EMPTY arcs
+ ^ static void fixempties(struct nfa *, FILE *);
+ */
+static void
+fixempties(
+    struct nfa *nfa,
+    FILE *f)			/* for debug output; NULL none */
+{
+    struct state *s;
+    struct state *s2;
+    struct state *nexts;
+    struct arc *a;
+    struct arc *nexta;
+    int totalinarcs;
+    struct arc **inarcsorig;
+    struct arc **arcarray;
+    int arccount;
+    int prevnins;
+    int nskip;
+
+    /*
+     * First, get rid of any states whose sole out-arc is an EMPTY,
+     * since they're basically just aliases for their successor.  The
+     * parsing algorithm creates enough of these that it's worth
+     * special-casing this.
+     */
+    for (s = nfa->states; s != NULL && !NISERR(); s = nexts) {
+	nexts = s->next;
+	if (s->flag || s->nouts != 1) {
+	    continue;
+	}
+	a = s->outs;
+	assert(a != NULL && a->outchain == NULL);
+	if (a->type != EMPTY) {
+	    continue;
+	}
+	if (s != a->to) {
+	    moveins(nfa, s, a->to);
+	}
+	dropstate(nfa, s);
+    }
+
+    /*
+     * Similarly, get rid of any state with a single EMPTY in-arc, by
+     * folding it into its predecessor.
+     */
+    for (s = nfa->states; s != NULL && !NISERR(); s = nexts) {
+	nexts = s->next;
+	/* Ensure tmp fields are clear for next step */
+	assert(s->tmp == NULL);
+	if (s->flag || s->nins != 1) {
+	    continue;
+	}
+	a = s->ins;
+	assert(a != NULL && a->inchain == NULL);
+	if (a->type != EMPTY) {
+	    continue;
+	}
+	if (s != a->from) {
+	    moveouts(nfa, s, a->from);
+	}
+	dropstate(nfa, s);
+    }
+
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * For each remaining NFA state, find all other states from which it is
+     * reachable by a chain of one or more EMPTY arcs.  Then generate new arcs
+     * that eliminate the need for each such chain.
+     *
+     * We could replace a chain of EMPTY arcs that leads from a "from" state
+     * to a "to" state either by pushing non-EMPTY arcs forward (linking
+     * directly from "from"'s predecessors to "to") or by pulling them back
+     * (linking directly from "from" to "to"'s successors).  We choose to
+     * always do the former; this choice is somewhat arbitrary, but the
+     * approach below requires that we uniformly do one or the other.
+     *
+     * Suppose we have a chain of N successive EMPTY arcs (where N can easily
+     * approach the size of the NFA).  All of the intermediate states must
+     * have additional inarcs and outarcs, else they'd have been removed by
+     * the steps above.  Assuming their inarcs are mostly not empties, we will
+     * add O(N^2) arcs to the NFA, since a non-EMPTY inarc leading to any one
+     * state in the chain must be duplicated to lead to all its successor
+     * states as well.  So there is no hope of doing less than O(N^2) work;
+     * however, we should endeavor to keep the big-O cost from being even
+     * worse than that, which it can easily become without care.  In
+     * particular, suppose we were to copy all S1's inarcs forward to S2, and
+     * then also to S3, and then later we consider pushing S2's inarcs forward
+     * to S3.  If we include the arcs already copied from S1 in that, we'd be
+     * doing O(N^3) work.  (The duplicate-arc elimination built into newarc()
+     * and its cohorts would get rid of the extra arcs, but not without cost.)
+     *
+     * We can avoid this cost by treating only arcs that existed at the start
+     * of this phase as candidates to be pushed forward.  To identify those,
+     * we remember the first inarc each state had to start with.  We rely on
+     * the fact that newarc() and friends put new arcs on the front of their
+     * to-states' inchains, and that this phase never deletes arcs, so that
+     * the original arcs must be the last arcs in their to-states' inchains.
+     *
+     * So the process here is that, for each state in the NFA, we gather up
+     * all non-EMPTY inarcs of states that can reach the target state via
+     * EMPTY arcs.  We then sort, de-duplicate, and merge these arcs into the
+     * target state's inchain.  (We can safely use sort-merge for this as long
+     * as we update each state's original-arcs pointer after we add arcs to
+     * it; the sort step of mergeins probably changed the order of the old
+     * arcs.)
+     *
+     * Another refinement worth making is that, because we only add non-EMPTY
+     * arcs during this phase, and all added arcs have the same from-state as
+     * the non-EMPTY arc they were cloned from, we know ahead of time that any
+     * states having only EMPTY outarcs will be useless for lack of outarcs
+     * after we drop the EMPTY arcs.  (They cannot gain non-EMPTY outarcs if
+     * they had none to start with.)  So we need not bother to update the
+     * inchains of such states at all.
+     */
+
+    /* Remember the states' first original inarcs */
+    /* ... and while at it, count how many old inarcs there are altogether */
+    inarcsorig = (struct arc **) MALLOC(nfa->nstates * sizeof(struct arc *));
+    if (inarcsorig == NULL) {
+	NERR(REG_ESPACE);
+	return;
+    }
+    totalinarcs = 0;
+    for (s = nfa->states; s != NULL; s = s->next) {
+	inarcsorig[s->no] = s->ins;
+	totalinarcs += s->nins;
+    }
+
+    /*
+     * Create a workspace for accumulating the inarcs to be added to the
+     * current target state.  totalinarcs is probably a considerable
+     * overestimate of the space needed, but the NFA is unlikely to be large
+     * enough at this point to make it worth being smarter.
+     */
+    arcarray = (struct arc **) MALLOC(totalinarcs * sizeof(struct arc *));
+    if (arcarray == NULL) {
+	NERR(REG_ESPACE);
+	FREE(inarcsorig);
+	return;
+    }
+
+    /* And iterate over the target states */
+    for (s = nfa->states; s != NULL && !NISERR(); s = s->next) {
+	/* Ignore target states without non-EMPTY outarcs, per note above */
+	if (!s->flag && !hasnonemptyout(s)) {
+	    continue;
+	}
+
+	/* Find predecessor states and accumulate their original inarcs */
+	arccount = 0;
+	for (s2 = emptyreachable(nfa, s, s, inarcsorig); s2 != s; s2 = nexts) {
+	    /* Add s2's original inarcs to arcarray[], but ignore empties */
+	    for (a = inarcsorig[s2->no]; a != NULL; a = a->inchain) {
+		if (a->type != EMPTY) {
+		    arcarray[arccount++] = a;
+		}
+	    }
+
+  	    /* Reset the tmp fields as we walk back */
+  	    nexts = s2->tmp;
+  	    s2->tmp = NULL;
+  	}
+  	s->tmp = NULL;
+	assert(arccount <= totalinarcs);
+
+	/* Remember how many original inarcs this state has */
+	prevnins = s->nins;
+
+	/* Add non-duplicate inarcs to target state */
+	mergeins(nfa, s, arcarray, arccount);
+
+	/* Now we must update the state's inarcsorig pointer */
+	nskip = s->nins - prevnins;
+	a = s->ins;
+	while (nskip-- > 0) {
+	    a = a->inchain;
+	}
+	inarcsorig[s->no] = a;
+    }
+
+    FREE(arcarray);
+    FREE(inarcsorig);
+
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * Remove all the EMPTY arcs, since we don't need them anymore.
+     */
+    for (s = nfa->states; s != NULL; s = s->next) {
+	for (a = s->outs; a != NULL; a = nexta) {
+	    nexta = a->outchain;
+	    if (a->type == EMPTY) {
+		freearc(nfa, a);
+	    }
+	}
+    }
+
+    /*
+     * And remove any states that have become useless.  (This cleanup is
+     * not very thorough, and would be even less so if we tried to
+     * combine it with the previous step; but cleanup() will take care
+     * of anything we miss.)
+     */
+    for (s = nfa->states; s != NULL; s = nexts) {
+	nexts = s->next;
+	if ((s->nins == 0 || s->nouts == 0) && !s->flag) {
+	    dropstate(nfa, s);
+	}
+    }
+
+    if (f != NULL) {
+	dumpnfa(nfa, f);
+    }
+}
+
+/*
+ - emptyreachable - recursively find all states that can reach s by EMPTY arcs
+ * The return value is the last such state found.  Its tmp field links back
+ * to the next-to-last such state, and so on back to s, so that all these
+ * states can be located without searching the whole NFA.
+ *
+ * Since this is only used in fixempties(), we pass in the inarcsorig[] array
+ * maintained by that function.  This lets us skip over all new inarcs, which
+ * are certainly not EMPTY arcs.
+ *
+ * The maximum recursion depth here is equal to the length of the longest
+ * loop-free chain of EMPTY arcs, which is surely no more than the size of
+ * the NFA, and in practice will be less than that.
+ ^ static struct state *emptyreachable(struct state *, struct state *);
+ */
+static struct state *
+emptyreachable(
+    struct nfa *nfa,
+    struct state *s,
+    struct state *lastfound,
+    struct arc **inarcsorig)
+{
+    struct arc *a;
+
+    s->tmp = lastfound;
+    lastfound = s;
+    for (a = inarcsorig[s->no]; a != NULL; a = a->inchain) {
+	if (a->type == EMPTY && a->from->tmp == NULL) {
+	    lastfound = emptyreachable(nfa, a->from, lastfound, inarcsorig);
+	}
+    }
+    return lastfound;
+}
+
+/*
+ * isconstraintarc - detect whether an arc is of a constraint type
+ */
+static inline int
+isconstraintarc(struct arc * a)
+{
+    switch (a->type)
+    {
+	case '^':
+	case '$':
+	case BEHIND:
+	case AHEAD:
+	case LACON:
+	    return 1;
+    }
+    return 0;
+}
+
+/*
+ * hasconstraintout - does state have a constraint out arc?
+ */
+static int
+hasconstraintout(struct state * s)
+{
+    struct arc *a;
+
+    for (a = s->outs; a != NULL; a = a->outchain) {
+	if (isconstraintarc(a)) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ * fixconstraintloops - get rid of loops containing only constraint arcs
+ *
+ * A loop of states that contains only constraint arcs is useless, since
+ * passing around the loop represents no forward progress.  Moreover, it
+ * would cause infinite looping in pullback/pushfwd, so we need to get rid
+ * of such loops before doing that.
+ */
+static void
+fixconstraintloops(
+    struct nfa * nfa,
+    FILE *f)		/* for debug output; NULL none */
+{
+    struct state *s;
+    struct state *nexts;
+    struct arc *a;
+    struct arc *nexta;
+    int hasconstraints;
+
+    /*
+     * In the trivial case of a state that loops to itself, we can just drop
+     * the constraint arc altogether.  This is worth special-casing because
+     * such loops are far more common than loops containing multiple states.
+     * While we're at it, note whether any constraint arcs survive.
+     */
+    hasconstraints = 0;
+    for (s = nfa->states; s != NULL && !NISERR(); s = nexts) {
+	nexts = s->next;
+	/* while we're at it, ensure tmp fields are clear for next step */
+	assert(s->tmp == NULL);
+	for (a = s->outs; a != NULL && !NISERR(); a = nexta) {
+	    nexta = a->outchain;
+	    if (isconstraintarc(a)) {
+		if (a->to == s) {
+		    freearc(nfa, a);
+		} else {
+		    hasconstraints = 1;
+ 		}
+	    }
+	}
+ 	/* If we removed all the outarcs, the state is useless. */
+ 	if (s->nouts == 0 && !s->flag) {
+ 	    dropstate(nfa, s);
+	}
+    }
+
+    /* Nothing to do if no remaining constraint arcs */
+    if (NISERR() || !hasconstraints) {
+	return;
+    }
+
+    /*
+     * Starting from each remaining NFA state, search outwards for a
+     * constraint loop.  If we find a loop, break the loop, then start the
+     * search over.  (We could possibly retain some state from the first scan,
+     * but it would complicate things greatly, and multi-state constraint
+     * loops are rare enough that it's not worth optimizing the case.)
+     */
+  restart:
+    for (s = nfa->states; s != NULL && !NISERR(); s = s->next) {
+	if (findconstraintloop(nfa, s)) {
+	    goto restart;
+	}
+    }
+
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * Now remove any states that have become useless.  (This cleanup is not
+     * very thorough, and would be even less so if we tried to combine it with
+     * the previous step; but cleanup() will take care of anything we miss.)
+     *
+     * Because findconstraintloop intentionally doesn't reset all tmp fields,
+     * we have to clear them after it's done.  This is a convenient place to
+     * do that, too.
+     */
+    for (s = nfa->states; s != NULL; s = nexts) {
+	nexts = s->next;
+	s->tmp = NULL;
+	if ((s->nins == 0 || s->nouts == 0) && !s->flag) {
+	    dropstate(nfa, s);
+	}
+    }
+
+    if (f != NULL) {
+ 	dumpnfa(nfa, f);
+    }
+}
+
+/*
+ * findconstraintloop - recursively find a loop of constraint arcs
+ *
+ * If we find a loop, break it by calling breakconstraintloop(), then
+ * return 1; otherwise return 0.
+ *
+ * State tmp fields are guaranteed all NULL on a success return, because
+ * breakconstraintloop does that.  After a failure return, any state that
+ * is known not to be part of a loop is marked with s->tmp == s; this allows
+ * us not to have to re-prove that fact on later calls.  (This convention is
+ * workable because we already eliminated single-state loops.)
+ *
+ * Note that the found loop doesn't necessarily include the first state we
+ * are called on.  Any loop reachable from that state will do.
+ *
+ * The maximum recursion depth here is one more than the length of the longest
+ * loop-free chain of constraint arcs, which is surely no more than the size
+ * of the NFA, and in practice will be a lot less than that.
+ */
+static int
+findconstraintloop(struct nfa * nfa, struct state * s)
+{
+    struct arc *a;
+
+    /* Since this is recursive, it could be driven to stack overflow */
+    if (STACK_TOO_DEEP(nfa->v->re)) {
+	NERR(REG_ETOOBIG);
+	return 1;		/* to exit as quickly as possible */
+    }
+
+    if (s->tmp != NULL) {
+	/* Already proven uninteresting? */
+	if (s->tmp == s) {
+	    return 0;
+	}
+	/* Found a loop involving s */
+	breakconstraintloop(nfa, s);
+	/* The tmp fields have been cleaned up by breakconstraintloop */
+	return 1;
+    }
+    for (a = s->outs; a != NULL; a = a->outchain) {
+	if (isconstraintarc(a)) {
+	    struct state *sto = a->to;
+
+	    assert(sto != s);
+	    s->tmp = sto;
+	    if (findconstraintloop(nfa, sto)) {
+		return 1;
+	    }
+	}
+    }
+
+    /*
+     * If we get here, no constraint loop exists leading out from s.  Mark it
+     * with s->tmp == s so we need not rediscover that fact again later.
+     */
+    s->tmp = s;
+    return 0;
+}
+
+/*
+ * breakconstraintloop - break a loop of constraint arcs
+ *
+ * sinitial is any one member state of the loop.  Each loop member's tmp
+ * field links to its successor within the loop.  (Note that this function
+ * will reset all the tmp fields to NULL.)
+ *
+ * We can break the loop by, for any one state S1 in the loop, cloning its
+ * loop successor state S2 (and possibly following states), and then moving
+ * all S1->S2 constraint arcs to point to the cloned S2.  The cloned S2 should
+ * copy any non-constraint outarcs of S2.  Constraint outarcs should be
+ * dropped if they point back to S1, else they need to be copied as arcs to
+ * similarly cloned states S3, S4, etc.  In general, each cloned state copies
+ * non-constraint outarcs, drops constraint outarcs that would lead to itself
+ * or any earlier cloned state, and sends other constraint outarcs to newly
+ * cloned states.  No cloned state will have any inarcs that aren't constraint
+ * arcs or do not lead from S1 or earlier-cloned states.  It's okay to drop
+ * constraint back-arcs since they would not take us to any state we've not
+ * already been in; therefore, no new constraint loop is created.  In this way
+ * we generate a modified NFA that can still represent every useful state
+ * sequence, but not sequences that represent state loops with no consumption
+ * of input data.  Note that the set of cloned states will certainly include
+ * all of the loop member states other than S1, and it may also include
+ * non-loop states that are reachable from S2 via constraint arcs.  This is
+ * important because there is no guarantee that findconstraintloop found a
+ * maximal loop (and searching for one would be NP-hard, so don't try).
+ * Frequently the "non-loop states" are actually part of a larger loop that
+ * we didn't notice, and indeed there may be several overlapping loops.
+ * This technique ensures convergence in such cases, while considering only
+ * the originally-found loop does not.
+ *
+ * If there is only one S1->S2 constraint arc, then that constraint is
+ * certainly satisfied when we enter any of the clone states.  This means that
+ * in the common case where many of the constraint arcs are identically
+ * labeled, we can merge together clone states linked by a similarly-labeled
+ * constraint: if we can get to the first one we can certainly get to the
+ * second, so there's no need to distinguish.  This greatly reduces the number
+ * of new states needed, so we preferentially break the given loop at a state
+ * pair where this is true.
+ *
+ * Furthermore, it's fairly common to find that a cloned successor state has
+ * no outarcs, especially if we're a bit aggressive about removing unnecessary
+ * outarcs.  If that happens, then there is simply not any interesting state
+ * that can be reached through the predecessor's loop arcs, which means we can
+ * break the loop just by removing those loop arcs, with no new states added.
+ */
+static void
+breakconstraintloop(struct nfa * nfa, struct state * sinitial)
+{
+    struct state *s;
+    struct state *shead;
+    struct state *stail;
+    struct state *sclone;
+    struct state *nexts;
+    struct arc *refarc;
+    struct arc *a;
+    struct arc *nexta;
+
+    /*
+     * Start by identifying which loop step we want to break at.
+     * Preferentially this is one with only one constraint arc.  (XXX are
+     * there any other secondary heuristics we want to use here?)  Set refarc
+     * to point to the selected lone constraint arc, if there is one.
+     */
+    refarc = NULL;
+    s = sinitial;
+    do {
+	nexts = s->tmp;
+	assert(nexts != s);	/* should not see any one-element loops */
+	if (refarc == NULL) {
+	    int narcs = 0;
+
+	    for (a = s->outs; a != NULL; a = a->outchain) {
+		if (a->to == nexts && isconstraintarc(a)) {
+		    refarc = a;
+		    narcs++;
+		}
+	    }
+	    assert(narcs > 0);
+	    if (narcs > 1) {
+		refarc = NULL;	/* multiple constraint arcs here, no good */
+	    }
+	}
+	s = nexts;
+    } while (s != sinitial);
+
+    if (refarc) {
+	/* break at the refarc */
+	shead = refarc->from;
+	stail = refarc->to;
+	assert(stail == shead->tmp);
+    } else {
+	/* for lack of a better idea, break after sinitial */
+	shead = sinitial;
+	stail = sinitial->tmp;
+    }
+
+    /*
+     * Reset the tmp fields so that we can use them for local storage in
+     * clonesuccessorstates.  (findconstraintloop won't mind, since it's just
+     * going to abandon its search anyway.)
+     */
+    for (s = nfa->states; s != NULL; s = s->next) {
+	s->tmp = NULL;
+    }
+
+    /*
+     * Recursively build clone state(s) as needed.
+     */
+    sclone = newstate(nfa);
+    if (sclone == NULL) {
+	assert(NISERR());
+	return;
+    }
+
+    clonesuccessorstates(nfa, stail, sclone, shead, refarc,
+	    NULL, NULL, nfa->nstates);
+
+    if (NISERR()) {
+	return;
+    }
+
+    /*
+     * It's possible that sclone has no outarcs at all, in which case it's
+     * useless.  (We don't try extremely hard to get rid of useless states
+     * here, but this is an easy and fairly common case.)
+     */
+    if (sclone->nouts == 0) {
+	freestate(nfa, sclone);
+	sclone = NULL;
+    }
+
+    /*
+     * Move shead's constraint-loop arcs to point to sclone, or just drop them
+     * if we discovered we don't need sclone.
+     */
+    for (a = shead->outs; a != NULL; a = nexta) {
+	nexta = a->outchain;
+	if (a->to == stail && isconstraintarc(a)) {
+	    if (sclone) {
+		cparc(nfa, a, shead, sclone);
+	    }
+	    freearc(nfa, a);
+	    if (NISERR()) {
+		break;
+	    }
+	}
+    }
+}
+
+/*
+ * clonesuccessorstates - create a tree of constraint-arc successor states
+ *
+ * ssource is the state to be cloned, and sclone is the state to copy its
+ * outarcs into.  sclone's inarcs, if any, should already be set up.
+ *
+ * spredecessor is the original predecessor state that we are trying to build
+ * successors for (it may not be the immediate predecessor of ssource).
+ * refarc, if not NULL, is the original constraint arc that is known to have
+ * been traversed out of spredecessor to reach the successor(s).
+ *
+ * For each cloned successor state, we transiently create a "donemap" that is
+ * a boolean array showing which source states we've already visited for this
+ * clone state.  This prevents infinite recursion as well as useless repeat
+ * visits to the same state subtree (which can add up fast, since typical NFAs
+ * have multiple redundant arc pathways).  Each donemap is a char array
+ * indexed by state number.  The donemaps are all of the same size "nstates",
+ * which is nfa->nstates as of the start of the recursion.  This is enough to
+ * have entries for all pre-existing states, but *not* entries for clone
+ * states created during the recursion.  That's okay since we have no need to
+ * mark those.
+ *
+ * curdonemap is NULL when recursing to a new sclone state, or sclone's
+ * donemap when we are recursing without having created a new state (which we
+ * do when we decide we can merge a successor state into the current clone
+ * state).  outerdonemap is NULL at the top level and otherwise the parent
+ * clone state's donemap.
+ *
+ * The successor states we create and fill here form a strict tree structure,
+ * with each state having exactly one predecessor, except that the toplevel
+ * state has no inarcs as yet (breakconstraintloop will add its inarcs from
+ * spredecessor after we're done).  Thus, we can examine sclone's inarcs back
+ * to the root, plus refarc if any, to identify the set of constraints already
+ * known valid at the current point.  This allows us to avoid generating extra
+ * successor states.
+ */
+static void
+clonesuccessorstates(
+    struct nfa * nfa,
+    struct state * ssource,
+    struct state * sclone,
+    struct state * spredecessor,
+    struct arc * refarc,
+    char *curdonemap,
+    char *outerdonemap,
+    int nstates)
+{
+    char *donemap;
+    struct arc *a;
+
+    /* Since this is recursive, it could be driven to stack overflow */
+    if (STACK_TOO_DEEP(nfa->v->re)) {
+	NERR(REG_ETOOBIG);
+	return;
+    }
+
+    /* If this state hasn't already got a donemap, create one */
+    donemap = curdonemap;
+    if (donemap == NULL) {
+	donemap = (char *) MALLOC(nstates * sizeof(char));
+	if (donemap == NULL) {
+	    NERR(REG_ESPACE);
+	    return;
+	}
+
+	if (outerdonemap != NULL) {
+	    /*
+	     * Not at outermost recursion level, so copy the outer level's
+	     * donemap; this ensures that we see states in process of being
+	     * visited at outer levels, or already merged into predecessor
+	     * states, as ones we shouldn't traverse back to.
+	     */
+	    memcpy(donemap, outerdonemap, nstates * sizeof(char));
+	} else {
+	    /* At outermost level, only spredecessor is off-limits */
+	    memset(donemap, 0, nstates * sizeof(char));
+	    assert(spredecessor->no < nstates);
+	    donemap[spredecessor->no] = 1;
+	}
+    }
+
+    /* Mark ssource as visited in the donemap */
+    assert(ssource->no < nstates);
+    assert(donemap[ssource->no] == 0);
+    donemap[ssource->no] = 1;
+
+    /*
+     * We proceed by first cloning all of ssource's outarcs, creating new
+     * clone states as needed but not doing more with them than that.  Then in
+     * a second pass, recurse to process the child clone states.  This allows
+     * us to have only one child clone state per reachable source state, even
+     * when there are multiple outarcs leading to the same state.  Also, when
+     * we do visit a child state, its set of inarcs is known exactly, which
+     * makes it safe to apply the constraint-is-already-checked optimization.
+     * Also, this ensures that we've merged all the states we can into the
+     * current clone before we recurse to any children, thus possibly saving
+     * them from making extra images of those states.
+     *
+     * While this function runs, child clone states of the current state are
+     * marked by setting their tmp fields to point to the original state they
+     * were cloned from.  This makes it possible to detect multiple outarcs
+     * leading to the same state, and also makes it easy to distinguish clone
+     * states from original states (which will have tmp == NULL).
+     */
+    for (a = ssource->outs; a != NULL && !NISERR(); a = a->outchain) {
+	struct state *sto = a->to;
+
+	/*
+	 * We do not consider cloning successor states that have no constraint
+	 * outarcs; just link to them as-is.  They cannot be part of a
+	 * constraint loop so there is no need to make copies.  In particular,
+	 * this rule keeps us from trying to clone the post state, which would
+	 * be a bad idea.
+	 */
+	if (isconstraintarc(a) && hasconstraintout(sto)) {
+	    struct state *prevclone;
+	    int canmerge;
+	    struct arc *a2;
+
+	    /*
+	     * Back-link constraint arcs must not be followed.  Nor is there a
+	     * need to revisit states previously merged into this clone.
+	     */
+	    assert(sto->no < nstates);
+	    if (donemap[sto->no] != 0) {
+		continue;
+	    }
+
+	    /*
+	     * Check whether we already have a child clone state for this
+	     * source state.
+	     */
+	    prevclone = NULL;
+	    for (a2 = sclone->outs; a2 != NULL; a2 = a2->outchain) {
+		if (a2->to->tmp == sto) {
+		    prevclone = a2->to;
+		    break;
+		}
+	    }
+
+	    /*
+	     * If this arc is labeled the same as refarc, or the same as any
+	     * arc we must have traversed to get to sclone, then no additional
+	     * constraints need to be met to get to sto, so we should just
+	     * merge its outarcs into sclone.
+	     */
+	    if (refarc && a->type == refarc->type && a->co == refarc->co) {
+		canmerge = 1;
+	    } else {
+		struct state *s;
+
+		canmerge = 0;
+		for (s = sclone; s->ins; s = s->ins->from) {
+		    if (s->nins == 1 &&
+			    a->type == s->ins->type && a->co == s->ins->co) {
+			canmerge = 1;
+			break;
+		    }
+		}
+	    }
+
+	    if (canmerge) {
+		/*
+		 * We can merge into sclone.  If we previously made a child
+		 * clone state, drop it; there's no need to visit it.  (This
+		 * can happen if ssource has multiple pathways to sto, and we
+		 * only just now found one that is provably a no-op.)
+		 */
+		if (prevclone) {
+		    dropstate(nfa, prevclone);	/* kills our outarc, too */
+		}
+
+		/* Recurse to merge sto's outarcs into sclone */
+		clonesuccessorstates(nfa, sto, sclone, spredecessor, refarc,
+			donemap, outerdonemap, nstates);
+		/* sto should now be marked as previously visited */
+		assert(NISERR() || donemap[sto->no] == 1);
+	    } else if (prevclone) {
+		/*
+		 * We already have a clone state for this successor, so just
+		 * make another arc to it.
+		 */
+		cparc(nfa, a, sclone, prevclone);
+	    } else {
+		/*
+		 * We need to create a new successor clone state.
+		 */
+		struct state *stoclone;
+
+		stoclone = newstate(nfa);
+		if (stoclone == NULL) {
+		    assert(NISERR());
+		    break;
+		}
+		/* Mark it as to what it's a clone of */
+		stoclone->tmp = sto;
+		/* ... and add the outarc leading to it */
+		cparc(nfa, a, sclone, stoclone);
+	    }
+	} else {
+	    /*
+	     * Non-constraint outarcs just get copied to sclone, as do outarcs
+	     * leading to states with no constraint outarc.
+	     */
+	    cparc(nfa, a, sclone, sto);
+	}
+    }
+
+    /*
+     * If we are at outer level for this clone state, recurse to all its child
+     * clone states, clearing their tmp fields as we go.  (If we're not
+     * outermost for sclone, leave this to be done by the outer call level.)
+     * Note that if we have multiple outarcs leading to the same clone state,
+     * it will only be recursed-to once.
+     */
+    if (curdonemap == NULL) {
+	for (a = sclone->outs; a != NULL && !NISERR(); a = a->outchain) {
+	    struct state *stoclone = a->to;
+	    struct state *sto = stoclone->tmp;
+
+	    if (sto != NULL) {
+		stoclone->tmp = NULL;
+		clonesuccessorstates(nfa, sto, stoclone, spredecessor, refarc,
+			NULL, donemap, nstates);
+	    }
+	}
+
+	/* Don't forget to free sclone's donemap when done with it */
+	FREE(donemap);
+    }
+}
+
+/*
+ - cleanup - clean up NFA after optimizations
+ ^ static void cleanup(struct nfa *);
+ */
+static void
+cleanup(
+    struct nfa *nfa)
+{
+    struct state *s;
+    struct state *nexts;
+    int n;
+
+    /*
+     * Clear out unreachable or dead-end states. Use pre to mark reachable,
+     * then post to mark can-reach-post.
+     */
+
+    markreachable(nfa, nfa->pre, NULL, nfa->pre);
+    markcanreach(nfa, nfa->post, nfa->pre, nfa->post);
+    for (s = nfa->states; s != NULL; s = nexts) {
+	nexts = s->next;
+	if (s->tmp != nfa->post && !s->flag) {
+	    dropstate(nfa, s);
+	}
+    }
+    assert(nfa->post->nins == 0 || nfa->post->tmp == nfa->post);
+    cleartraverse(nfa, nfa->pre);
+    assert(nfa->post->nins == 0 || nfa->post->tmp == NULL);
+    /* the nins==0 (final unreachable) case will be caught later */
+
+    /*
+     * Renumber surviving states.
+     */
+
+    n = 0;
+    for (s = nfa->states; s != NULL; s = s->next) {
+	s->no = n++;
+    }
+    nfa->nstates = n;
+}
+
+/*
+ - markreachable - recursive marking of reachable states
+ ^ static void markreachable(struct nfa *, struct state *, struct state *,
+ ^ 	struct state *);
+ */
+static void
+markreachable(
+    struct nfa *nfa,
+    struct state *s,
+    struct state *okay,		/* consider only states with this mark */
+    struct state *mark)		/* the value to mark with */
+{
+    struct arc *a;
+
+    if (s->tmp != okay) {
+	return;
+    }
+    s->tmp = mark;
+
+    for (a = s->outs; a != NULL; a = a->outchain) {
+	markreachable(nfa, a->to, okay, mark);
+    }
+}
+
+/*
+ - markcanreach - recursive marking of states which can reach here
+ ^ static void markcanreach(struct nfa *, struct state *, struct state *,
+ ^ 	struct state *);
+ */
+static void
+markcanreach(
+    struct nfa *nfa,
+    struct state *s,
+    struct state *okay,		/* consider only states with this mark */
+    struct state *mark)		/* the value to mark with */
+{
+    struct arc *a;
+
+    if (s->tmp != okay) {
+	return;
+    }
+    s->tmp = mark;
+
+    for (a = s->ins; a != NULL; a = a->inchain) {
+	markcanreach(nfa, a->from, okay, mark);
+    }
+}
+
+/*
+ - analyze - ascertain potentially-useful facts about an optimized NFA
+ ^ static long analyze(struct nfa *);
+ */
+static long			/* re_info bits to be ORed in */
+analyze(
+    struct nfa *nfa)
+{
+    struct arc *a;
+    struct arc *aa;
+
+    if (nfa->pre->outs == NULL) {
+	return REG_UIMPOSSIBLE;
+    }
+    for (a = nfa->pre->outs; a != NULL; a = a->outchain) {
+	for (aa = a->to->outs; aa != NULL; aa = aa->outchain) {
+	    if (aa->to == nfa->post) {
+		return REG_UEMPTYMATCH;
+	    }
+	}
+    }
+    return 0;
+}
+
+/*
+ - compact - construct the compact representation of an NFA
+ ^ static void compact(struct nfa *, struct cnfa *);
+ */
+static void
+compact(
+    struct nfa *nfa,
+    struct cnfa *cnfa)
+{
+    struct state *s;
+    struct arc *a;
+    size_t nstates;
+    size_t narcs;
+    struct carc *ca;
+    struct carc *first;
+
+    assert(!NISERR());
+
+    nstates = 0;
+    narcs = 0;
+    for (s = nfa->states; s != NULL; s = s->next) {
+	nstates++;
+	narcs += s->nouts + 1;	/* need one extra for endmarker */
+    }
+
+    cnfa->stflags = (char *) MALLOC(nstates * sizeof(char));
+    cnfa->states = (struct carc **) MALLOC(nstates * sizeof(struct carc *));
+    cnfa->arcs = (struct carc *) MALLOC(narcs * sizeof(struct carc));
+    if (cnfa->stflags == NULL || cnfa->states == NULL || cnfa->arcs == NULL) {
+	if (cnfa->stflags != NULL) {
+	    FREE(cnfa->stflags);
+	}
+	if (cnfa->states != NULL) {
+	    FREE(cnfa->states);
+	}
+	if (cnfa->arcs != NULL) {
+	    FREE(cnfa->arcs);
+	}
+	NERR(REG_ESPACE);
+	return;
+    }
+    cnfa->nstates = nstates;
+    cnfa->pre = nfa->pre->no;
+    cnfa->post = nfa->post->no;
+    cnfa->bos[0] = nfa->bos[0];
+    cnfa->bos[1] = nfa->bos[1];
+    cnfa->eos[0] = nfa->eos[0];
+    cnfa->eos[1] = nfa->eos[1];
+    cnfa->ncolors = maxcolor(nfa->cm) + 1;
+    cnfa->flags = 0;
+
+    ca = cnfa->arcs;
+    for (s = nfa->states; s != NULL; s = s->next) {
+	assert((size_t) s->no < nstates);
+	cnfa->stflags[s->no] = 0;
+	cnfa->states[s->no] = ca;
+	first = ca;
+	for (a = s->outs; a != NULL; a = a->outchain) {
+	    switch (a->type) {
+	    case PLAIN:
+		ca->co = a->co;
+		ca->to = a->to->no;
+		ca++;
+		break;
+	    case LACON:
+		assert(s->no != cnfa->pre);
+		ca->co = (color) (cnfa->ncolors + a->co);
+		ca->to = a->to->no;
+		ca++;
+		cnfa->flags |= HASLACONS;
+		break;
+	    default:
+		NERR(REG_ASSERT);
+		break;
+	    }
+	}
+	carcsort(first, ca - first);
+	ca->co = COLORLESS;
+	ca->to = 0;
+	ca++;
+    }
+    assert(ca == &cnfa->arcs[narcs]);
+    assert(cnfa->nstates != 0);
+
+    /*
+     * Mark no-progress states.
+     */
+
+    for (a = nfa->pre->outs; a != NULL; a = a->outchain) {
+	cnfa->stflags[a->to->no] = CNFA_NOPROGRESS;
+    }
+    cnfa->stflags[nfa->pre->no] = CNFA_NOPROGRESS;
+}
+
+/*
+ - carcsort - sort compacted-NFA arcs by color
+ ^ static void carcsort(struct carc *, struct carc *);
+ */
+static void
+carcsort(
+    struct carc *first,
+    size_t n)
+{
+    if (n > 1) {
+	qsort(first, n, sizeof(struct carc), carc_cmp);
+    }
+}
+
+static int
+carc_cmp(
+    const void *a,
+    const void *b)
+{
+    const struct carc *aa = (const struct carc *) a;
+    const struct carc *bb = (const struct carc *) b;
+
+    if (aa->co < bb->co) {
+	return -1;
+    }
+    if (aa->co > bb->co) {
+	return +1;
+    }
+    if (aa->to < bb->to) {
+	return -1;
+    }
+    if (aa->to > bb->to) {
+	return +1;
+    }
+    return 0;
+}
+
+/*
+ - freecnfa - free a compacted NFA
+ ^ static void freecnfa(struct cnfa *);
+ */
+static void
+freecnfa(
+    struct cnfa *cnfa)
+{
+    assert(cnfa->nstates != 0);	/* not empty already */
+    cnfa->nstates = 0;
+    FREE(cnfa->stflags);
+    FREE(cnfa->states);
+    FREE(cnfa->arcs);
+}
+
+/*
+ - dumpnfa - dump an NFA in human-readable form
+ ^ static void dumpnfa(struct nfa *, FILE *);
+ */
+static void
+dumpnfa(
+    struct nfa *nfa,
+    FILE *f)
+{
+#ifdef REG_DEBUG
+    struct state *s;
+    int nstates = 0;
+    int narcs = 0;
+
+    fprintf(f, "pre %d, post %d", nfa->pre->no, nfa->post->no);
+    if (nfa->bos[0] != COLORLESS) {
+	fprintf(f, ", bos [%ld]", (long) nfa->bos[0]);
+    }
+    if (nfa->bos[1] != COLORLESS) {
+	fprintf(f, ", bol [%ld]", (long) nfa->bos[1]);
+    }
+    if (nfa->eos[0] != COLORLESS) {
+	fprintf(f, ", eos [%ld]", (long) nfa->eos[0]);
+    }
+    if (nfa->eos[1] != COLORLESS) {
+	fprintf(f, ", eol [%ld]", (long) nfa->eos[1]);
+    }
+    fprintf(f, "\n");
+    for (s = nfa->states; s != NULL; s = s->next) {
+	dumpstate(s, f);
+	nstates++;
+	narcs += s->nouts;
+    }
+    fprintf(f, "total of %d states, %d arcs\n", nstates, narcs);
+    if (nfa->parent == NULL) {
+	dumpcolors(nfa->cm, f);
+    }
+    fflush(f);
+#endif
+}
+
+#ifdef REG_DEBUG		/* subordinates of dumpnfa */
+/*
+ ^ #ifdef REG_DEBUG
+ */
+
+/*
+ - dumpstate - dump an NFA state in human-readable form
+ ^ static void dumpstate(struct state *, FILE *);
+ */
+static void
+dumpstate(
+    struct state *s,
+    FILE *f)
+{
+    struct arc *a;
+
+    fprintf(f, "%d%s%c", s->no, (s->tmp != NULL) ? "T" : "",
+	    (s->flag) ? s->flag : '.');
+    if (s->prev != NULL && s->prev->next != s) {
+	fprintf(f, "\tstate chain bad\n");
+    }
+    if (s->nouts == 0) {
+	fprintf(f, "\tno out arcs\n");
+    } else {
+	dumparcs(s, f);
+    }
+    fflush(f);
+    for (a = s->ins; a != NULL; a = a->inchain) {
+	if (a->to != s) {
+	    fprintf(f, "\tlink from %d to %d on %d's in-chain\n",
+		    a->from->no, a->to->no, s->no);
+	}
+    }
+}
+
+/*
+ - dumparcs - dump out-arcs in human-readable form
+ ^ static void dumparcs(struct state *, FILE *);
+ */
+static void
+dumparcs(
+    struct state *s,
+    FILE *f)
+{
+    int pos;
+    struct arc *a;
+
+    /* printing oldest arcs first is usually clearer */
+    a = s->outs;
+    assert(a != NULL);
+    while (a->outchain != NULL) {
+	a = a->outchain;
+    }
+    pos = 1;
+    do {
+	dumparc(a, s, f);
+	if (pos == 5) {
+	    fprintf(f, "\n");
+	    pos = 1;
+	} else {
+	    pos++;
+	}
+	a = a->outchainRev;
+    } while (a != NULL);
+    if (pos != 1) {
+	fprintf(f, "\n");
+    }
+}
+
+/*
+ - dumparc - dump one outarc in readable form, including prefixing tab
+ ^ static void dumparc(struct arc *, struct state *, FILE *);
+ */
+static void
+dumparc(
+    struct arc *a,
+    struct state *s,
+    FILE *f)
+{
+    struct arc *aa;
+    struct arcbatch *ab;
+
+    fprintf(f, "\t");
+    switch (a->type) {
+    case PLAIN:
+	fprintf(f, "[%ld]", (long) a->co);
+	break;
+    case AHEAD:
+	fprintf(f, ">%ld>", (long) a->co);
+	break;
+    case BEHIND:
+	fprintf(f, "<%ld<", (long) a->co);
+	break;
+    case LACON:
+	fprintf(f, ":%ld:", (long) a->co);
+	break;
+    case '^':
+    case '$':
+	fprintf(f, "%c%d", a->type, (int) a->co);
+	break;
+    case EMPTY:
+	break;
+    default:
+	fprintf(f, "0x%x/0%lo", a->type, (long) a->co);
+	break;
+    }
+    if (a->from != s) {
+	fprintf(f, "?%d?", a->from->no);
+    }
+    for (ab = &a->from->oas; ab != NULL; ab = ab->next) {
+	for (aa = &ab->a[0]; aa < &ab->a[ABSIZE]; aa++) {
+	    if (aa == a) {
+		break;		/* NOTE BREAK OUT */
+	    }
+	}
+	if (aa < &ab->a[ABSIZE]) {	/* propagate break */
+	    break;		/* NOTE BREAK OUT */
+	}
+    }
+    if (ab == NULL) {
+	fprintf(f, "?!?");	/* not in allocated space */
+    }
+    fprintf(f, "->");
+    if (a->to == NULL) {
+	fprintf(f, "NULL");
+	return;
+    }
+    fprintf(f, "%d", a->to->no);
+    for (aa = a->to->ins; aa != NULL; aa = aa->inchain) {
+	if (aa == a) {
+	    break;		/* NOTE BREAK OUT */
+	}
+    }
+    if (aa == NULL) {
+	fprintf(f, "?!?");	/* missing from in-chain */
+    }
+}
+
+/*
+ ^ #endif
+ */
+#endif				/* ifdef REG_DEBUG */
+
+/*
+ - dumpcnfa - dump a compacted NFA in human-readable form
+ ^ static void dumpcnfa(struct cnfa *, FILE *);
+ */
+static void
+dumpcnfa(
+    struct cnfa *cnfa,
+    FILE *f)
+{
+#ifdef REG_DEBUG
+    int st;
+
+    fprintf(f, "pre %d, post %d", cnfa->pre, cnfa->post);
+    if (cnfa->bos[0] != COLORLESS) {
+	fprintf(f, ", bos [%ld]", (long) cnfa->bos[0]);
+    }
+    if (cnfa->bos[1] != COLORLESS) {
+	fprintf(f, ", bol [%ld]", (long) cnfa->bos[1]);
+    }
+    if (cnfa->eos[0] != COLORLESS) {
+	fprintf(f, ", eos [%ld]", (long) cnfa->eos[0]);
+    }
+    if (cnfa->eos[1] != COLORLESS) {
+	fprintf(f, ", eol [%ld]", (long) cnfa->eos[1]);
+    }
+    if (cnfa->flags&HASLACONS) {
+	fprintf(f, ", haslacons");
+    }
+    fprintf(f, "\n");
+    for (st = 0; st < cnfa->nstates; st++) {
+	dumpcstate(st, cnfa, f);
+    }
+    fflush(f);
+#endif
+}
+
+#ifdef REG_DEBUG		/* subordinates of dumpcnfa */
+/*
+ ^ #ifdef REG_DEBUG
+ */
+
+/*
+ - dumpcstate - dump a compacted-NFA state in human-readable form
+ ^ static void dumpcstate(int, struct cnfa *, FILE *);
+ */
+static void
+dumpcstate(
+    int st,
+    struct cnfa *cnfa,
+    FILE *f)
+{
+    struct carc *ca;
+    int pos;
+
+    fprintf(f, "%d%s", st, (cnfa->stflags[st] & CNFA_NOPROGRESS) ? ":" : ".");
+    pos = 1;
+    for (ca = cnfa->states[st]; ca->co != COLORLESS; ca++) {
+	if (ca->co < cnfa->ncolors) {
+	    fprintf(f, "\t[%ld]->%d", (long) ca->co, ca->to);
+	} else {
+	    fprintf(f, "\t:%ld:->%d", (long) (ca->co - cnfa->ncolors), ca->to);
+	}
+	if (pos == 5) {
+	    fprintf(f, "\n");
+	    pos = 1;
+	} else {
+	    pos++;
+	}
+    }
+    if (ca == cnfa->states[st] || pos != 1) {
+	fprintf(f, "\n");
+    }
+    fflush(f);
+}
+
+/*
+ ^ #endif
+ */
+#endif				/* ifdef REG_DEBUG */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regcomp.c b/generic/regcomp.c
new file mode 100644
index 0000000..58d55fb
--- /dev/null
+++ b/generic/regcomp.c
@@ -0,0 +1,2225 @@
+/*
+ * re_*comp and friends - compile REs
+ * This file #includes several others (see the bottom).
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer. All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#include "regguts.h"
+
+/*
+ * forward declarations, up here so forward datatypes etc. are defined early
+ */
+/* =====^!^===== begin forwards =====^!^===== */
+/* automatically gathered by fwd; do not hand-edit */
+/* === regcomp.c === */
+int compile(regex_t *, const chr *, size_t, int);
+static void moresubs(struct vars *, int);
+static int freev(struct vars *, int);
+static void makesearch(struct vars *, struct nfa *);
+static struct subre *parse(struct vars *, int, int, struct state *, struct state *);
+static struct subre *parsebranch(struct vars *, int, int, struct state *, struct state *, int);
+static void parseqatom(struct vars *, int, int, struct state *, struct state *, struct subre *);
+static void nonword(struct vars *, int, struct state *, struct state *);
+static void word(struct vars *, int, struct state *, struct state *);
+static int scannum(struct vars *);
+static void repeat(struct vars *, struct state *, struct state *, int, int);
+static void bracket(struct vars *, struct state *, struct state *);
+static void cbracket(struct vars *, struct state *, struct state *);
+static void brackpart(struct vars *, struct state *, struct state *);
+static const chr *scanplain(struct vars *);
+static void onechr(struct vars *, pchr, struct state *, struct state *);
+static void dovec(struct vars *, struct cvec *, struct state *, struct state *);
+static void wordchrs(struct vars *);
+static struct subre *subre(struct vars *, int, int, struct state *, struct state *);
+static void freesubre(struct vars *, struct subre *);
+static void freesrnode(struct vars *, struct subre *);
+static void optst(struct vars *, struct subre *);
+static int numst(struct subre *, int);
+static void markst(struct subre *);
+static void cleanst(struct vars *);
+static long nfatree(struct vars *, struct subre *, FILE *);
+static long nfanode(struct vars *, struct subre *, FILE *);
+static int newlacon(struct vars *, struct state *, struct state *, int);
+static void freelacons(struct subre *, int);
+static void rfree(regex_t *);
+static void dump(regex_t *, FILE *);
+static void dumpst(struct subre *, FILE *, int);
+static void stdump(struct subre *, FILE *, int);
+static const char *stid(struct subre *, char *, size_t);
+/* === regc_lex.c === */
+static void lexstart(struct vars *);
+static void prefixes(struct vars *);
+static void lexnest(struct vars *, const chr *, const chr *);
+static void lexword(struct vars *);
+static int next(struct vars *);
+static int lexescape(struct vars *);
+static int lexdigits(struct vars *, int, int, int);
+static int brenext(struct vars *, pchr);
+static void skip(struct vars *);
+static chr newline(void);
+static chr chrnamed(struct vars *, const chr *, const chr *, pchr);
+/* === regc_color.c === */
+static void initcm(struct vars *, struct colormap *);
+static void freecm(struct colormap *);
+static void cmtreefree(struct colormap *, union tree *, int);
+static color setcolor(struct colormap *, pchr, pcolor);
+static color maxcolor(struct colormap *);
+static color newcolor(struct colormap *);
+static void freecolor(struct colormap *, pcolor);
+static color pseudocolor(struct colormap *);
+static color subcolor(struct colormap *, pchr c);
+static color newsub(struct colormap *, pcolor);
+static void subrange(struct vars *, pchr, pchr, struct state *, struct state *);
+static void subblock(struct vars *, pchr, struct state *, struct state *);
+static void okcolors(struct nfa *, struct colormap *);
+static void colorchain(struct colormap *, struct arc *);
+static void uncolorchain(struct colormap *, struct arc *);
+static void rainbow(struct nfa *, struct colormap *, int, pcolor, struct state *, struct state *);
+static void colorcomplement(struct nfa *, struct colormap *, int, struct state *, struct state *, struct state *);
+#ifdef REG_DEBUG
+static void dumpcolors(struct colormap *, FILE *);
+static void fillcheck(struct colormap *, union tree *, int, FILE *);
+static void dumpchr(pchr, FILE *);
+#endif
+/* === regc_nfa.c === */
+static struct nfa *newnfa(struct vars *, struct colormap *, struct nfa *);
+static void freenfa(struct nfa *);
+static struct state *newstate(struct nfa *);
+static struct state *newfstate(struct nfa *, int flag);
+static void dropstate(struct nfa *, struct state *);
+static void freestate(struct nfa *, struct state *);
+static void destroystate(struct nfa *, struct state *);
+static void newarc(struct nfa *, int, pcolor, struct state *, struct state *);
+static void createarc(struct nfa *, int, pcolor, struct state *, struct state *);
+static struct arc *allocarc(struct nfa *, struct state *);
+static void freearc(struct nfa *, struct arc *);
+static void changearctarget(struct arc *, struct state *);
+static int hasnonemptyout(struct state *);
+static struct arc *findarc(struct state *, int, pcolor);
+static void cparc(struct nfa *, struct arc *, struct state *, struct state *);
+static void sortins(struct nfa *, struct state *);
+static int sortins_cmp(const void *, const void *);
+static void sortouts(struct nfa *, struct state *);
+static int sortouts_cmp(const void *, const void *);
+static void moveins(struct nfa *, struct state *, struct state *);
+static void copyins(struct nfa *, struct state *, struct state *);
+static void mergeins(struct nfa *, struct state *, struct arc **, int);
+static void moveouts(struct nfa *, struct state *, struct state *);
+static void copyouts(struct nfa *, struct state *, struct state *);
+static void cloneouts(struct nfa *, struct state *, struct state *, struct state *, int);
+static void delsub(struct nfa *, struct state *, struct state *);
+static void deltraverse(struct nfa *, struct state *, struct state *);
+static void dupnfa(struct nfa *, struct state *, struct state *, struct state *, struct state *);
+static void duptraverse(struct nfa *, struct state *, struct state *, int);
+static void cleartraverse(struct nfa *, struct state *);
+static void specialcolors(struct nfa *);
+static long optimize(struct nfa *, FILE *);
+static void pullback(struct nfa *, FILE *);
+static int pull(struct nfa *, struct arc *, struct state **);
+static void pushfwd(struct nfa *, FILE *);
+static int push(struct nfa *, struct arc *, struct state **);
+#define	INCOMPATIBLE	1	/* destroys arc */
+#define	SATISFIED	2	/* constraint satisfied */
+#define	COMPATIBLE	3	/* compatible but not satisfied yet */
+static int combine(struct arc *, struct arc *);
+static void fixempties(struct nfa *, FILE *);
+static struct state *emptyreachable(struct nfa *, struct state *,
+			struct state *, struct arc **);
+static int	isconstraintarc(struct arc *);
+static int	hasconstraintout(struct state *);
+static void fixconstraintloops(struct nfa *, FILE *);
+static int	findconstraintloop(struct nfa *, struct state *);
+static void breakconstraintloop(struct nfa *, struct state *);
+static void clonesuccessorstates(struct nfa *, struct state *, struct state *,
+		 struct state *, struct arc *, char *, char *, int);
+static void cleanup(struct nfa *);
+static void markreachable(struct nfa *, struct state *, struct state *, struct state *);
+static void markcanreach(struct nfa *, struct state *, struct state *, struct state *);
+static long analyze(struct nfa *);
+static void compact(struct nfa *, struct cnfa *);
+static void carcsort(struct carc *, size_t);
+static int carc_cmp(const void *, const void *);
+static void freecnfa(struct cnfa *);
+static void dumpnfa(struct nfa *, FILE *);
+#ifdef REG_DEBUG
+static void dumpstate(struct state *, FILE *);
+static void dumparcs(struct state *, FILE *);
+static void dumparc(struct arc *, struct state *, FILE *);
+#endif
+static void dumpcnfa(struct cnfa *, FILE *);
+#ifdef REG_DEBUG
+static void dumpcstate(int, struct cnfa *, FILE *);
+#endif
+/* === regc_cvec.c === */
+static struct cvec *clearcvec(struct cvec *);
+static void addchr(struct cvec *, pchr);
+static void addrange(struct cvec *, pchr, pchr);
+static struct cvec *newcvec(int, int);
+static struct cvec *getcvec(struct vars *, int, int);
+static void freecvec(struct cvec *);
+/* === regc_locale.c === */
+static celt element(struct vars *, const chr *, const chr *);
+static struct cvec *range(struct vars *, celt, celt, int);
+static int before(celt, celt);
+static struct cvec *eclass(struct vars *, celt, int);
+static struct cvec *cclass(struct vars *, const chr *, const chr *, int);
+static struct cvec *allcases(struct vars *, pchr);
+static int cmp(const chr *, const chr *, size_t);
+static int casecmp(const chr *, const chr *, size_t);
+/* automatically gathered by fwd; do not hand-edit */
+/* =====^!^===== end forwards =====^!^===== */
+
+/* internal variables, bundled for easy passing around */
+struct vars {
+    regex_t *re;
+    const chr *now;		/* scan pointer into string */
+    const chr *stop;		/* end of string */
+    const chr *savenow;		/* saved now and stop for "subroutine call" */
+    const chr *savestop;
+    int err;			/* error code (0 if none) */
+    int cflags;			/* copy of compile flags */
+    int lasttype;		/* type of previous token */
+    int nexttype;		/* type of next token */
+    chr nextvalue;		/* value (if any) of next token */
+    int lexcon;			/* lexical context type (see lex.c) */
+    int nsubexp;		/* subexpression count */
+    struct subre **subs;	/* subRE pointer vector */
+    size_t nsubs;		/* length of vector */
+    struct subre *sub10[10];	/* initial vector, enough for most */
+    struct nfa *nfa;		/* the NFA */
+    struct colormap *cm;	/* character color map */
+    color nlcolor;		/* color of newline */
+    struct state *wordchrs;	/* state in nfa holding word-char outarcs */
+    struct subre *tree;		/* subexpression tree */
+    struct subre *treechain;	/* all tree nodes allocated */
+    struct subre *treefree;	/* any free tree nodes */
+    int ntree;			/* number of tree nodes, plus one */
+    struct cvec *cv;		/* interface cvec */
+    struct cvec *cv2;		/* utility cvec */
+    struct subre *lacons;	/* lookahead-constraint vector */
+    int nlacons;		/* size of lacons */
+    size_t spaceused;		/* approx. space used for compilation */
+};
+
+/* parsing macros; most know that `v' is the struct vars pointer */
+#define	NEXT()	(next(v))		/* advance by one token */
+#define	SEE(t)	(v->nexttype == (t))	/* is next token this? */
+#define	EAT(t)	(SEE(t) && next(v))	/* if next is this, swallow it */
+#define	VISERR(vv)	((vv)->err != 0)/* have we seen an error yet? */
+#define	ISERR()	VISERR(v)
+#define VERR(vv,e)	((vv)->nexttype = EOS, \
+			 (vv)->err = ((vv)->err ? (vv)->err : (e)))
+#define	ERR(e)	VERR(v, e)		/* record an error */
+#define	NOERR()	{if (ISERR()) return;}	/* if error seen, return */
+#define	NOERRN()	{if (ISERR()) return NULL;}	/* NOERR with retval */
+#define	NOERRZ()	{if (ISERR()) return 0;}	/* NOERR with retval */
+#define INSIST(c, e) do { if (!(c)) ERR(e); } while (0)	/* error if c false */
+#define	NOTE(b)	(v->re->re_info |= (b))		/* note visible condition */
+#define	EMPTYARC(x, y)	newarc(v->nfa, EMPTY, 0, x, y)
+
+/* token type codes, some also used as NFA arc types */
+#define	EMPTY	'n'		/* no token present */
+#define	EOS	'e'		/* end of string */
+#define	PLAIN	'p'		/* ordinary character */
+#define	DIGIT	'd'		/* digit (in bound) */
+#define	BACKREF	'b'		/* back reference */
+#define	COLLEL	'I'		/* start of [. */
+#define	ECLASS	'E'		/* start of [= */
+#define	CCLASS	'C'		/* start of [: */
+#define	END	'X'		/* end of [. [= [: */
+#define	RANGE	'R'		/* - within [] which might be range delim. */
+#define	LACON	'L'		/* lookahead constraint subRE */
+#define	AHEAD	'a'		/* color-lookahead arc */
+#define	BEHIND	'r'		/* color-lookbehind arc */
+#define	WBDRY	'w'		/* word boundary constraint */
+#define	NWBDRY	'W'		/* non-word-boundary constraint */
+#define	SBEGIN	'A'		/* beginning of string (even if not BOL) */
+#define	SEND	'Z'		/* end of string (even if not EOL) */
+#define	PREFER	'P'		/* length preference */
+
+/* is an arc colored, and hence on a color chain? */
+#define	COLORED(a) \
+	((a)->type == PLAIN || (a)->type == AHEAD || (a)->type == BEHIND)
+
+/* static function list */
+static const struct fns functions = {
+    rfree,			/* regfree insides */
+};
+
+/*
+ - compile - compile regular expression
+ * Note: on failure, no resources remain allocated, so regfree()
+ * need not be applied to re.
+ ^ int compile(regex_t *, const chr *, size_t, int);
+ */
+int
+compile(
+    regex_t *re,
+    const chr *string,
+    size_t len,
+    int flags)
+{
+    AllocVars(v);
+    struct guts *g;
+    int i;
+    size_t j;
+    FILE *debug = (flags&REG_PROGRESS) ? stdout : NULL;
+#define	CNOERR()	{ if (ISERR()) return freev(v, v->err); }
+
+    /*
+     * Sanity checks.
+     */
+
+    if (re == NULL || string == NULL) {
+	FreeVars(v);
+	return REG_INVARG;
+    }
+    if ((flags&REG_QUOTE) && (flags&(REG_ADVANCED|REG_EXPANDED|REG_NEWLINE))) {
+	FreeVars(v);
+	return REG_INVARG;
+    }
+    if (!(flags&REG_EXTENDED) && (flags&REG_ADVF)) {
+	FreeVars(v);
+	return REG_INVARG;
+    }
+
+    /*
+     * Initial setup (after which freev() is callable).
+     */
+
+    v->re = re;
+    v->now = string;
+    v->stop = v->now + len;
+    v->savenow = v->savestop = NULL;
+    v->err = 0;
+    v->cflags = flags;
+    v->nsubexp = 0;
+    v->subs = v->sub10;
+    v->nsubs = 10;
+    for (j = 0; j < v->nsubs; j++) {
+	v->subs[j] = NULL;
+    }
+    v->nfa = NULL;
+    v->cm = NULL;
+    v->nlcolor = COLORLESS;
+    v->wordchrs = NULL;
+    v->tree = NULL;
+    v->treechain = NULL;
+    v->treefree = NULL;
+    v->cv = NULL;
+    v->cv2 = NULL;
+    v->lacons = NULL;
+    v->nlacons = 0;
+    v->spaceused = 0;
+    re->re_magic = REMAGIC;
+    re->re_info = 0;		/* bits get set during parse */
+    re->re_csize = sizeof(chr);
+    re->re_guts = NULL;
+    re->re_fns = (void*)(&functions);
+
+    /*
+     * More complex setup, malloced things.
+     */
+
+    re->re_guts = (void*)(MALLOC(sizeof(struct guts)));
+    if (re->re_guts == NULL) {
+	return freev(v, REG_ESPACE);
+    }
+    g = (struct guts *) re->re_guts;
+    g->tree = NULL;
+    initcm(v, &g->cmap);
+    v->cm = &g->cmap;
+    g->lacons = NULL;
+    g->nlacons = 0;
+    ZAPCNFA(g->search);
+    v->nfa = newnfa(v, v->cm, NULL);
+    CNOERR();
+    v->cv = newcvec(100, 20);
+    if (v->cv == NULL) {
+	return freev(v, REG_ESPACE);
+    }
+
+    /*
+     * Parsing.
+     */
+
+    lexstart(v);		/* also handles prefixes */
+    if ((v->cflags&REG_NLSTOP) || (v->cflags&REG_NLANCH)) {
+	/*
+	 * Assign newline a unique color.
+	 */
+
+	v->nlcolor = subcolor(v->cm, newline());
+	okcolors(v->nfa, v->cm);
+    }
+    CNOERR();
+    v->tree = parse(v, EOS, PLAIN, v->nfa->init, v->nfa->final);
+    assert(SEE(EOS));		/* even if error; ISERR() => SEE(EOS) */
+    CNOERR();
+    assert(v->tree != NULL);
+
+    /*
+     * Finish setup of nfa and its subre tree.
+     */
+
+    specialcolors(v->nfa);
+    CNOERR();
+    if (debug != NULL) {
+	fprintf(debug, "\n\n\n========= RAW ==========\n");
+	dumpnfa(v->nfa, debug);
+	dumpst(v->tree, debug, 1);
+    }
+    optst(v, v->tree);
+    v->ntree = numst(v->tree, 1);
+    markst(v->tree);
+    cleanst(v);
+    if (debug != NULL) {
+	fprintf(debug, "\n\n\n========= TREE FIXED ==========\n");
+	dumpst(v->tree, debug, 1);
+    }
+
+    /*
+     * Build compacted NFAs for tree and lacons.
+     */
+
+    re->re_info |= nfatree(v, v->tree, debug);
+    CNOERR();
+    assert(v->nlacons == 0 || v->lacons != NULL);
+    for (i = 1; i < v->nlacons; i++) {
+	if (debug != NULL) {
+	    fprintf(debug, "\n\n\n========= LA%d ==========\n", i);
+	}
+	nfanode(v, &v->lacons[i], debug);
+    }
+    CNOERR();
+    if (v->tree->flags&SHORTER) {
+	NOTE(REG_USHORTEST);
+    }
+
+    /*
+     * Build compacted NFAs for tree, lacons, fast search.
+     */
+
+    if (debug != NULL) {
+	fprintf(debug, "\n\n\n========= SEARCH ==========\n");
+    }
+
+    /*
+     * Can sacrifice main NFA now, so use it as work area.
+     */
+
+    (void) optimize(v->nfa, debug);
+    CNOERR();
+    makesearch(v, v->nfa);
+    CNOERR();
+    compact(v->nfa, &g->search);
+    CNOERR();
+
+    /*
+     * Looks okay, package it up.
+     */
+
+    re->re_nsub = v->nsubexp;
+    v->re = NULL;		/* freev no longer frees re */
+    g->magic = GUTSMAGIC;
+    g->cflags = v->cflags;
+    g->info = re->re_info;
+    g->nsub = re->re_nsub;
+    g->tree = v->tree;
+    v->tree = NULL;
+    g->ntree = v->ntree;
+    g->compare = (v->cflags&REG_ICASE) ? casecmp : cmp;
+    g->lacons = v->lacons;
+    v->lacons = NULL;
+    g->nlacons = v->nlacons;
+
+    if (flags&REG_DUMP) {
+	dump(re, stdout);
+    }
+
+    assert(v->err == 0);
+    return freev(v, 0);
+}
+
+/*
+ - moresubs - enlarge subRE vector
+ ^ static void moresubs(struct vars *, int);
+ */
+static void
+moresubs(
+    struct vars *v,
+    int wanted)			/* want enough room for this one */
+{
+    struct subre **p;
+    size_t n;
+
+    assert(wanted > 0 && (size_t)wanted >= v->nsubs);
+    n = (size_t)wanted * 3 / 2 + 1;
+    if (v->subs == v->sub10) {
+	p = (struct subre **) MALLOC(n * sizeof(struct subre *));
+	if (p != NULL) {
+	    memcpy(p, v->subs, v->nsubs * sizeof(struct subre *));
+	}
+    } else {
+	p = (struct subre **) REALLOC(v->subs, n*sizeof(struct subre *));
+    }
+    if (p == NULL) {
+	ERR(REG_ESPACE);
+	return;
+    }
+
+    v->subs = p;
+    for (p = &v->subs[v->nsubs]; v->nsubs < n; p++, v->nsubs++) {
+	*p = NULL;
+    }
+    assert(v->nsubs == n);
+    assert((size_t)wanted < v->nsubs);
+}
+
+/*
+ - freev - free vars struct's substructures where necessary
+ * Optionally does error-number setting, and always returns error code (if
+ * any), to make error-handling code terser.
+ ^ static int freev(struct vars *, int);
+ */
+static int
+freev(
+    struct vars *v,
+    int err)
+{
+    register int ret;
+
+    if (v->re != NULL) {
+	rfree(v->re);
+    }
+    if (v->subs != v->sub10) {
+	FREE(v->subs);
+    }
+    if (v->nfa != NULL) {
+	freenfa(v->nfa);
+    }
+    if (v->tree != NULL) {
+	freesubre(v, v->tree);
+    }
+    if (v->treechain != NULL) {
+	cleanst(v);
+    }
+    if (v->cv != NULL) {
+	freecvec(v->cv);
+    }
+    if (v->cv2 != NULL) {
+	freecvec(v->cv2);
+    }
+    if (v->lacons != NULL) {
+	freelacons(v->lacons, v->nlacons);
+    }
+    ERR(err);			/* nop if err==0 */
+
+    ret = v->err;
+    FreeVars(v);
+    return ret;
+}
+
+/*
+ - makesearch - turn an NFA into a search NFA (implicit prepend of .*?)
+ * NFA must have been optimize()d already.
+ ^ static void makesearch(struct vars *, struct nfa *);
+ */
+static void
+makesearch(
+    struct vars *v,
+    struct nfa *nfa)
+{
+    struct arc *a, *b;
+    struct state *pre = nfa->pre;
+    struct state *s, *s2, *slist;
+
+    /*
+     * No loops are needed if it's anchored.
+     */
+
+    for (a = pre->outs; a != NULL; a = a->outchain) {
+	assert(a->type == PLAIN);
+	if (a->co != nfa->bos[0] && a->co != nfa->bos[1]) {
+	    break;
+	}
+    }
+    if (a != NULL) {
+	/*
+	 * Add implicit .* in front.
+	 */
+
+	rainbow(nfa, v->cm, PLAIN, COLORLESS, pre, pre);
+
+	/*
+	 * And ^* and \A* too -- not always necessary, but harmless.
+	 */
+
+	newarc(nfa, PLAIN, nfa->bos[0], pre, pre);
+	newarc(nfa, PLAIN, nfa->bos[1], pre, pre);
+    }
+
+    /*
+     * Now here's the subtle part. Because many REs have no lookback
+     * constraints, often knowing when you were in the pre state tells you
+     * little; it's the next state(s) that are informative. But some of them
+     * may have other inarcs, i.e. it may be possible to make actual progress
+     * and then return to one of them. We must de-optimize such cases,
+     * splitting each such state into progress and no-progress states.
+     */
+
+    /*
+     * First, make a list of the states.
+     */
+
+    slist = NULL;
+    for (a=pre->outs ; a!=NULL ; a=a->outchain) {
+	s = a->to;
+	for (b=s->ins ; b!=NULL ; b=b->inchain) {
+	    if (b->from != pre) {
+		break;
+	    }
+	}
+
+	/*
+	 * We want to mark states as being in the list already by having non
+	 * NULL tmp fields, but we can't just store the old slist value in tmp
+	 * because that doesn't work for the first such state.  Instead, the
+	 * first list entry gets its own address in tmp.
+	 */
+	if (b != NULL && s->tmp == NULL) {
+	    s->tmp = (slist != NULL) ? slist : s;
+	    slist = s;
+	}
+    }
+
+    /*
+     * Do the splits.
+     */
+
+    for (s=slist ; s!=NULL ; s=s2) {
+	s2 = newstate(nfa);
+	NOERR();
+	copyouts(nfa, s, s2);
+	NOERR();
+	for (a=s->ins ; a!=NULL ; a=b) {
+	    b = a->inchain;
+
+	    if (a->from != pre) {
+		cparc(nfa, a, a->from, s2);
+		freearc(nfa, a);
+	    }
+	}
+	s2 = (s->tmp != s) ? s->tmp : NULL;
+	s->tmp = NULL;		/* clean up while we're at it */
+    }
+}
+
+/*
+ - parse - parse an RE
+ * This is actually just the top level, which parses a bunch of branches tied
+ * together with '|'. They appear in the tree as the left children of a chain
+ * of '|' subres.
+ ^ static struct subre *parse(struct vars *, int, int, struct state *,
+ ^ 	struct state *);
+ */
+static struct subre *
+parse(
+    struct vars *v,
+    int stopper,		/* EOS or ')' */
+    int type,			/* LACON (lookahead subRE) or PLAIN */
+    struct state *init,		/* initial state */
+    struct state *final)	/* final state */
+{
+    struct state *left, *right;	/* scaffolding for branch */
+    struct subre *branches;	/* top level */
+    struct subre *branch;	/* current branch */
+    struct subre *t;		/* temporary */
+    int firstbranch;		/* is this the first branch? */
+
+    assert(stopper == ')' || stopper == EOS);
+
+    branches = subre(v, '|', LONGER, init, final);
+    NOERRN();
+    branch = branches;
+    firstbranch = 1;
+    do {	/* a branch */
+	if (!firstbranch) {
+	    /*
+	     * Need a place to hang the branch.
+	     */
+
+	    branch->right = subre(v, '|', LONGER, init, final);
+	    NOERRN();
+	    branch = branch->right;
+	}
+	firstbranch = 0;
+	left = newstate(v->nfa);
+	right = newstate(v->nfa);
+	NOERRN();
+	EMPTYARC(init, left);
+	EMPTYARC(right, final);
+	NOERRN();
+	branch->left = parsebranch(v, stopper, type, left, right, 0);
+	NOERRN();
+	branch->flags |= UP(branch->flags | branch->left->flags);
+	if ((branch->flags &~ branches->flags) != 0) {	/* new flags */
+	    for (t = branches; t != branch; t = t->right) {
+		t->flags |= branch->flags;
+	    }
+	}
+    } while (EAT('|'));
+    assert(SEE(stopper) || SEE(EOS));
+
+    if (!SEE(stopper)) {
+	assert(stopper == ')' && SEE(EOS));
+	ERR(REG_EPAREN);
+    }
+
+    /*
+     * Optimize out simple cases.
+     */
+
+    if (branch == branches) {	/* only one branch */
+	assert(branch->right == NULL);
+	t = branch->left;
+	branch->left = NULL;
+	freesubre(v, branches);
+	branches = t;
+    } else if (!MESSY(branches->flags)) {	/* no interesting innards */
+	freesubre(v, branches->left);
+	branches->left = NULL;
+	freesubre(v, branches->right);
+	branches->right = NULL;
+	branches->op = '=';
+    }
+
+    return branches;
+}
+
+/*
+ - parsebranch - parse one branch of an RE
+ * This mostly manages concatenation, working closely with parseqatom().
+ * Concatenated things are bundled up as much as possible, with separate
+ * ',' nodes introduced only when necessary due to substructure.
+ ^ static struct subre *parsebranch(struct vars *, int, int, struct state *,
+ ^ 	struct state *, int);
+ */
+static struct subre *
+parsebranch(
+    struct vars *v,
+    int stopper,		/* EOS or ')' */
+    int type,			/* LACON (lookahead subRE) or PLAIN */
+    struct state *left,		/* leftmost state */
+    struct state *right,	/* rightmost state */
+    int partial)		/* is this only part of a branch? */
+{
+    struct state *lp;		/* left end of current construct */
+    int seencontent;		/* is there anything in this branch yet? */
+    struct subre *t;
+
+    lp = left;
+    seencontent = 0;
+    t = subre(v, '=', 0, left, right);	/* op '=' is tentative */
+    NOERRN();
+    while (!SEE('|') && !SEE(stopper) && !SEE(EOS)) {
+	if (seencontent) {	/* implicit concat operator */
+	    lp = newstate(v->nfa);
+	    NOERRN();
+	    moveins(v->nfa, right, lp);
+	}
+	seencontent = 1;
+
+	/* NB, recursion in parseqatom() may swallow rest of branch */
+	parseqatom(v, stopper, type, lp, right, t);
+	NOERRN();
+    }
+
+    if (!seencontent) {		/* empty branch */
+	if (!partial) {
+	    NOTE(REG_UUNSPEC);
+	}
+	assert(lp == left);
+	EMPTYARC(left, right);
+    }
+
+    return t;
+}
+
+/*
+ - parseqatom - parse one quantified atom or constraint of an RE
+ * The bookkeeping near the end cooperates very closely with parsebranch(); in
+ * particular, it contains a recursion that can involve parsing the rest of
+ * the branch, making this function's name somewhat inaccurate.
+ ^ static void parseqatom(struct vars *, int, int, struct state *,
+ ^ 	struct state *, struct subre *);
+ */
+static void
+parseqatom(
+    struct vars *v,
+    int stopper,		/* EOS or ')' */
+    int type,			/* LACON (lookahead subRE) or PLAIN */
+    struct state *lp,		/* left state to hang it on */
+    struct state *rp,		/* right state to hang it on */
+    struct subre *top)		/* subtree top */
+{
+    struct state *s;		/* temporaries for new states */
+    struct state *s2;
+#define	ARCV(t, val)	newarc(v->nfa, t, val, lp, rp)
+    int m, n;
+    struct subre *atom;		/* atom's subtree */
+    struct subre *t;
+    int cap;			/* capturing parens? */
+    int pos;			/* positive lookahead? */
+    int subno;			/* capturing-parens or backref number */
+    int atomtype;
+    int qprefer;		/* quantifier short/long preference */
+    int f;
+    struct subre **atomp;	/* where the pointer to atom is */
+
+    /*
+     * Initial bookkeeping.
+     */
+
+    atom = NULL;
+    assert(lp->nouts == 0);	/* must string new code */
+    assert(rp->nins == 0);	/* between lp and rp */
+    subno = 0;			/* just to shut lint up */
+
+    /*
+     * An atom or constraint...
+     */
+
+    atomtype = v->nexttype;
+    switch (atomtype) {
+	/* first, constraints, which end by returning */
+    case '^':
+	ARCV('^', 1);
+	if (v->cflags&REG_NLANCH) {
+	    ARCV(BEHIND, v->nlcolor);
+	}
+	NEXT();
+	return;
+    case '$':
+	ARCV('$', 1);
+	if (v->cflags&REG_NLANCH) {
+	    ARCV(AHEAD, v->nlcolor);
+	}
+	NEXT();
+	return;
+    case SBEGIN:
+	ARCV('^', 1);		/* BOL */
+	ARCV('^', 0);		/* or BOS */
+	NEXT();
+	return;
+    case SEND:
+	ARCV('$', 1);		/* EOL */
+	ARCV('$', 0);		/* or EOS */
+	NEXT();
+	return;
+    case '<':
+	wordchrs(v);		/* does NEXT() */
+	s = newstate(v->nfa);
+	NOERR();
+	nonword(v, BEHIND, lp, s);
+	word(v, AHEAD, s, rp);
+	return;
+    case '>':
+	wordchrs(v);		/* does NEXT() */
+	s = newstate(v->nfa);
+	NOERR();
+	word(v, BEHIND, lp, s);
+	nonword(v, AHEAD, s, rp);
+	return;
+    case WBDRY:
+	wordchrs(v);		/* does NEXT() */
+	s = newstate(v->nfa);
+	NOERR();
+	nonword(v, BEHIND, lp, s);
+	word(v, AHEAD, s, rp);
+	s = newstate(v->nfa);
+	NOERR();
+	word(v, BEHIND, lp, s);
+	nonword(v, AHEAD, s, rp);
+	return;
+    case NWBDRY:
+	wordchrs(v);		/* does NEXT() */
+	s = newstate(v->nfa);
+	NOERR();
+	word(v, BEHIND, lp, s);
+	word(v, AHEAD, s, rp);
+	s = newstate(v->nfa);
+	NOERR();
+	nonword(v, BEHIND, lp, s);
+	nonword(v, AHEAD, s, rp);
+	return;
+    case LACON:			/* lookahead constraint */
+	pos = v->nextvalue;
+	NEXT();
+	s = newstate(v->nfa);
+	s2 = newstate(v->nfa);
+	NOERR();
+	t = parse(v, ')', LACON, s, s2);
+	freesubre(v, t);	/* internal structure irrelevant */
+	assert(SEE(')') || ISERR());
+	NEXT();
+	n = newlacon(v, s, s2, pos);
+	NOERR();
+	ARCV(LACON, n);
+	return;
+
+	/*
+	 * Then errors, to get them out of the way.
+	 */
+
+    case '*':
+    case '+':
+    case '?':
+    case '{':
+	ERR(REG_BADRPT);
+	return;
+    default:
+	ERR(REG_ASSERT);
+	return;
+
+	/*
+	 * Then plain characters, and minor variants on that theme.
+	 */
+
+    case ')':			/* unbalanced paren */
+	if ((v->cflags&REG_ADVANCED) != REG_EXTENDED) {
+	    ERR(REG_EPAREN);
+	    return;
+	}
+
+	/*
+	 * Legal in EREs due to specification botch.
+	 */
+
+	NOTE(REG_UPBOTCH);
+	/* fallthrough into case PLAIN */
+    case PLAIN:
+	onechr(v, v->nextvalue, lp, rp);
+	okcolors(v->nfa, v->cm);
+	NOERR();
+	NEXT();
+	break;
+    case '[':
+	if (v->nextvalue == 1) {
+	    bracket(v, lp, rp);
+	} else {
+	    cbracket(v, lp, rp);
+	}
+	assert(SEE(']') || ISERR());
+	NEXT();
+	break;
+    case '.':
+	rainbow(v->nfa, v->cm, PLAIN,
+		(v->cflags&REG_NLSTOP) ? v->nlcolor : COLORLESS, lp, rp);
+	NEXT();
+	break;
+
+	/*
+	 * And finally the ugly stuff.
+	 */
+
+    case '(':			/* value flags as capturing or non */
+	cap = (type == LACON) ? 0 : v->nextvalue;
+	if (cap) {
+	    v->nsubexp++;
+	    subno = v->nsubexp;
+	    if ((size_t)subno >= v->nsubs) {
+		moresubs(v, subno);
+	    }
+	    assert((size_t)subno < v->nsubs);
+	} else {
+	    atomtype = PLAIN;	/* something that's not '(' */
+	}
+	NEXT();
+
+	/*
+	 * Need new endpoints because tree will contain pointers.
+	 */
+
+	s = newstate(v->nfa);
+	s2 = newstate(v->nfa);
+	NOERR();
+	EMPTYARC(lp, s);
+	EMPTYARC(s2, rp);
+	NOERR();
+	atom = parse(v, ')', PLAIN, s, s2);
+	assert(SEE(')') || ISERR());
+	NEXT();
+	NOERR();
+	if (cap) {
+	    v->subs[subno] = atom;
+	    t = subre(v, '(', atom->flags|CAP, lp, rp);
+	    NOERR();
+	    t->subno = subno;
+	    t->left = atom;
+	    atom = t;
+	}
+
+	/*
+	 * Postpone everything else pending possible {0}.
+	 */
+
+	break;
+    case BACKREF:		/* the Feature From The Black Lagoon */
+	INSIST(type != LACON, REG_ESUBREG);
+	INSIST(v->nextvalue < v->nsubs, REG_ESUBREG);
+	INSIST(v->subs[v->nextvalue] != NULL, REG_ESUBREG);
+	NOERR();
+	assert(v->nextvalue > 0);
+	atom = subre(v, 'b', BACKR, lp, rp);
+	NOERR();
+	subno = v->nextvalue;
+	atom->subno = subno;
+	EMPTYARC(lp, rp);	/* temporarily, so there's something */
+	NEXT();
+	break;
+    }
+
+    /*
+     * ...and an atom may be followed by a quantifier.
+     */
+
+    switch (v->nexttype) {
+    case '*':
+	m = 0;
+	n = DUPINF;
+	qprefer = (v->nextvalue) ? LONGER : SHORTER;
+	NEXT();
+	break;
+    case '+':
+	m = 1;
+	n = DUPINF;
+	qprefer = (v->nextvalue) ? LONGER : SHORTER;
+	NEXT();
+	break;
+    case '?':
+	m = 0;
+	n = 1;
+	qprefer = (v->nextvalue) ? LONGER : SHORTER;
+	NEXT();
+	break;
+    case '{':
+	NEXT();
+	m = scannum(v);
+	if (EAT(',')) {
+	    if (SEE(DIGIT)) {
+		n = scannum(v);
+	    } else {
+		n = DUPINF;
+	    }
+	    if (m > n) {
+		ERR(REG_BADBR);
+		return;
+	    }
+
+	    /*
+	     * {m,n} exercises preference, even if it's {m,m}
+	     */
+
+	    qprefer = (v->nextvalue) ? LONGER : SHORTER;
+	} else {
+	    n = m;
+	    /*
+	     * {m} passes operand's preference through.
+	     */
+
+	    qprefer = 0;
+	}
+	if (!SEE('}')) {	/* catches errors too */
+	    ERR(REG_BADBR);
+	    return;
+	}
+	NEXT();
+	break;
+    default:			/* no quantifier */
+	m = n = 1;
+	qprefer = 0;
+	break;
+    }
+
+    /*
+     * Annoying special case: {0} or {0,0} cancels everything.
+     */
+
+    if (m == 0 && n == 0) {
+	if (atom != NULL) {
+	    freesubre(v, atom);
+	}
+	if (atomtype == '(') {
+	    v->subs[subno] = NULL;
+	}
+	delsub(v->nfa, lp, rp);
+	EMPTYARC(lp, rp);
+	return;
+    }
+
+    /*
+     * If not a messy case, avoid hard part.
+     */
+
+    assert(!MESSY(top->flags));
+    f = top->flags | qprefer | ((atom != NULL) ? atom->flags : 0);
+    if (atomtype != '(' && atomtype != BACKREF && !MESSY(UP(f))) {
+	if (!(m == 1 && n == 1)) {
+	    repeat(v, lp, rp, m, n);
+	}
+	if (atom != NULL) {
+	    freesubre(v, atom);
+	}
+	top->flags = f;
+	return;
+    }
+
+    /*
+     * hard part: something messy
+     * That is, capturing parens, back reference, short/long clash, or an atom
+     * with substructure containing one of those.
+     */
+
+    /*
+     * Now we'll need a subre for the contents even if they're boring.
+     */
+
+    if (atom == NULL) {
+	atom = subre(v, '=', 0, lp, rp);
+	NOERR();
+    }
+
+    /*
+     * Prepare a general-purpose state skeleton.
+     *
+     * In the no-backrefs case, we want this:
+     *
+     * [lp] ---> [s] ---prefix---> [begin] ---atom---> [end] ---rest---> [rp]
+     *
+     * where prefix is some repetitions of atom.  In the general case we need
+     *
+     * [lp] ---> [s] ---iterator---> [s2] ---rest---> [rp]
+     *
+     * where the iterator wraps around [begin] ---atom---> [end]
+     *
+     * We make the s state here for both cases; s2 is made below if needed
+     */
+
+    s = newstate(v->nfa);	/* first, new endpoints for the atom */
+    s2 = newstate(v->nfa);
+    NOERR();
+    moveouts(v->nfa, lp, s);
+    moveins(v->nfa, rp, s2);
+    NOERR();
+    atom->begin = s;
+    atom->end = s2;
+    s = newstate(v->nfa);	/* set up starting state */
+    NOERR();
+    EMPTYARC(lp, s);
+    NOERR();
+
+    /*
+     * Break remaining subRE into x{...} and what follows.
+     */
+
+    t = subre(v, '.', COMBINE(qprefer, atom->flags), lp, rp);
+    NOERR();
+    t->left = atom;
+    atomp = &t->left;
+
+    /*
+     * Here we should recurse... but we must postpone that to the end.
+     */
+
+    /*
+     * Split top into prefix and remaining.
+     */
+
+    assert(top->op == '=' && top->left == NULL && top->right == NULL);
+    top->left = subre(v, '=', top->flags, top->begin, lp);
+    NOERR();
+    top->op = '.';
+    top->right = t;
+
+    /*
+     * If it's a backref, now is the time to replicate the subNFA.
+     */
+
+    if (atomtype == BACKREF) {
+	assert(atom->begin->nouts == 1);	/* just the EMPTY */
+	delsub(v->nfa, atom->begin, atom->end);
+	assert(v->subs[subno] != NULL);
+
+	/*
+	 * And here's why the recursion got postponed: it must wait until the
+	 * skeleton is filled in, because it may hit a backref that wants to
+	 * copy the filled-in skeleton.
+	 */
+
+	dupnfa(v->nfa, v->subs[subno]->begin, v->subs[subno]->end,
+		atom->begin, atom->end);
+	NOERR();
+    }
+
+    /*
+     * It's quantifier time.  If the atom is just a backref, we'll let it deal
+     * with quantifiers internally.
+     */
+
+    if (atomtype == BACKREF) {
+	/*
+	 * Special case: backrefs have internal quantifiers.
+	 */
+
+	EMPTYARC(s, atom->begin);	/* empty prefix */
+
+	/*
+	 * Just stuff everything into atom.
+	 */
+
+	repeat(v, atom->begin, atom->end, m, n);
+	atom->min = (short) m;
+	atom->max = (short) n;
+	atom->flags |= COMBINE(qprefer, atom->flags);
+	/* rest of branch can be strung starting from atom->end */
+	s2 = atom->end;
+    } else if (m == 1 && n == 1) {
+	/*
+	 * No/vacuous quantifier: done.
+	 */
+
+	EMPTYARC(s, atom->begin);	/* empty prefix */
+	/* rest of branch can be strung starting from atom->end */
+	s2 = atom->end;
+    } else if (m > 0 && !(atom->flags & BACKR)) {
+	/*
+	 * If there's no backrefs involved, we can turn x{m,n} into
+	 * x{m-1,n-1}x, with capturing parens in only the second x.  This
+	 * is valid because we only care about capturing matches from the
+	 * final iteration of the quantifier.  It's a win because we can
+	 * implement the backref-free left side as a plain DFA node, since
+	 * we don't really care where its submatches are.
+	 */
+
+	dupnfa(v->nfa, atom->begin, atom->end, s, atom->begin);
+	assert(m >= 1 && m != DUPINF && n >= 1);
+	repeat(v, s, atom->begin, m-1, (n == DUPINF) ? n : n-1);
+	f = COMBINE(qprefer, atom->flags);
+	t = subre(v, '.', f, s, atom->end);	/* prefix and atom */
+	NOERR();
+	t->left = subre(v, '=', PREF(f), s, atom->begin);
+	NOERR();
+	t->right = atom;
+	*atomp = t;
+	/* rest of branch can be strung starting from atom->end */
+	s2 = atom->end;
+    } else {
+	/* general case: need an iteration node */
+	s2 = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, atom->end, s2);
+	NOERR();
+	dupnfa(v->nfa, atom->begin, atom->end, s, s2);
+	repeat(v, s, s2, m, n);
+	f = COMBINE(qprefer, atom->flags);
+	t = subre(v, '*', f, s, s2);
+	NOERR();
+	t->min = (short) m;
+	t->max = (short) n;
+	t->left = atom;
+	*atomp = t;
+	/* rest of branch is to be strung from iteration's end state */
+    }
+
+    /*
+     * And finally, look after that postponed recursion.
+     */
+
+    t = top->right;
+    if (!(SEE('|') || SEE(stopper) || SEE(EOS))) {
+	t->right = parsebranch(v, stopper, type, s2, rp, 1);
+    } else {
+	EMPTYARC(s2, rp);
+	t->right = subre(v, '=', 0, s2, rp);
+    }
+    NOERR();
+    assert(SEE('|') || SEE(stopper) || SEE(EOS));
+    t->flags |= COMBINE(t->flags, t->right->flags);
+    top->flags |= COMBINE(top->flags, t->flags);
+}
+
+/*
+ - nonword - generate arcs for non-word-character ahead or behind
+ ^ static void nonword(struct vars *, int, struct state *, struct state *);
+ */
+static void
+nonword(
+    struct vars *v,
+    int dir,			/* AHEAD or BEHIND */
+    struct state *lp,
+    struct state *rp)
+{
+    int anchor = (dir == AHEAD) ? '$' : '^';
+
+    assert(dir == AHEAD || dir == BEHIND);
+    newarc(v->nfa, anchor, 1, lp, rp);
+    newarc(v->nfa, anchor, 0, lp, rp);
+    colorcomplement(v->nfa, v->cm, dir, v->wordchrs, lp, rp);
+    /* (no need for special attention to \n) */
+}
+
+/*
+ - word - generate arcs for word character ahead or behind
+ ^ static void word(struct vars *, int, struct state *, struct state *);
+ */
+static void
+word(
+    struct vars *v,
+    int dir,			/* AHEAD or BEHIND */
+    struct state *lp,
+    struct state *rp)
+{
+    assert(dir == AHEAD || dir == BEHIND);
+    cloneouts(v->nfa, v->wordchrs, lp, rp, dir);
+    /* (no need for special attention to \n) */
+}
+
+/*
+ - scannum - scan a number
+ ^ static int scannum(struct vars *);
+ */
+static int			/* value, <= DUPMAX */
+scannum(
+    struct vars *v)
+{
+    int n = 0;
+
+    while (SEE(DIGIT) && n < DUPMAX) {
+	n = n*10 + v->nextvalue;
+	NEXT();
+    }
+    if (SEE(DIGIT) || n > DUPMAX) {
+	ERR(REG_BADBR);
+	return 0;
+    }
+    return n;
+}
+
+/*
+ - repeat - replicate subNFA for quantifiers
+ * The sub-NFA strung from lp to rp is modified to represent m to n
+ * repetitions of its initial contents.
+ * The duplication sequences used here are chosen carefully so that any
+ * pointers starting out pointing into the subexpression end up pointing into
+ * the last occurrence. (Note that it may not be strung between the same left
+ * and right end states, however!) This used to be important for the subRE
+ * tree, although the important bits are now handled by the in-line code in
+ * parse(), and when this is called, it doesn't matter any more.
+ ^ static void repeat(struct vars *, struct state *, struct state *, int, int);
+ */
+static void
+repeat(
+    struct vars *v,
+    struct state *lp,
+    struct state *rp,
+    int m,
+    int n)
+{
+#define	SOME		2
+#define	INF		3
+#define	PAIR(x, y)	((x)*4 + (y))
+#define	REDUCE(x)	( ((x) == DUPINF) ? INF : (((x) > 1) ? SOME : (x)) )
+    const int rm = REDUCE(m);
+    const int rn = REDUCE(n);
+    struct state *s, *s2;
+
+    switch (PAIR(rm, rn)) {
+    case PAIR(0, 0):		/* empty string */
+	delsub(v->nfa, lp, rp);
+	EMPTYARC(lp, rp);
+	break;
+    case PAIR(0, 1):		/* do as x| */
+	EMPTYARC(lp, rp);
+	break;
+    case PAIR(0, SOME):		/* do as x{1,n}| */
+	repeat(v, lp, rp, 1, n);
+	NOERR();
+	EMPTYARC(lp, rp);
+	break;
+    case PAIR(0, INF):		/* loop x around */
+	s = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, lp, s);
+	moveins(v->nfa, rp, s);
+	EMPTYARC(lp, s);
+	EMPTYARC(s, rp);
+	break;
+    case PAIR(1, 1):		/* no action required */
+	break;
+    case PAIR(1, SOME):		/* do as x{0,n-1}x = (x{1,n-1}|)x */
+	s = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, lp, s);
+	dupnfa(v->nfa, s, rp, lp, s);
+	NOERR();
+	repeat(v, lp, s, 1, n-1);
+	NOERR();
+	EMPTYARC(lp, s);
+	break;
+    case PAIR(1, INF):		/* add loopback arc */
+	s = newstate(v->nfa);
+	s2 = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, lp, s);
+	moveins(v->nfa, rp, s2);
+	EMPTYARC(lp, s);
+	EMPTYARC(s2, rp);
+	EMPTYARC(s2, s);
+	break;
+    case PAIR(SOME, SOME):		/* do as x{m-1,n-1}x */
+	s = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, lp, s);
+	dupnfa(v->nfa, s, rp, lp, s);
+	NOERR();
+	repeat(v, lp, s, m-1, n-1);
+	break;
+    case PAIR(SOME, INF):		/* do as x{m-1,}x */
+	s = newstate(v->nfa);
+	NOERR();
+	moveouts(v->nfa, lp, s);
+	dupnfa(v->nfa, s, rp, lp, s);
+	NOERR();
+	repeat(v, lp, s, m-1, n);
+	break;
+    default:
+	ERR(REG_ASSERT);
+	break;
+    }
+}
+
+/*
+ - bracket - handle non-complemented bracket expression
+ * Also called from cbracket for complemented bracket expressions.
+ ^ static void bracket(struct vars *, struct state *, struct state *);
+ */
+static void
+bracket(
+    struct vars *v,
+    struct state *lp,
+    struct state *rp)
+{
+    assert(SEE('['));
+    NEXT();
+    while (!SEE(']') && !SEE(EOS)) {
+	brackpart(v, lp, rp);
+    }
+    assert(SEE(']') || ISERR());
+    okcolors(v->nfa, v->cm);
+}
+
+/*
+ - cbracket - handle complemented bracket expression
+ * We do it by calling bracket() with dummy endpoints, and then complementing
+ * the result. The alternative would be to invoke rainbow(), and then delete
+ * arcs as the b.e. is seen... but that gets messy.
+ ^ static void cbracket(struct vars *, struct state *, struct state *);
+ */
+static void
+cbracket(
+    struct vars *v,
+    struct state *lp,
+    struct state *rp)
+{
+    struct state *left = newstate(v->nfa);
+    struct state *right = newstate(v->nfa);
+
+    NOERR();
+    bracket(v, left, right);
+    if (v->cflags&REG_NLSTOP) {
+	newarc(v->nfa, PLAIN, v->nlcolor, left, right);
+    }
+    NOERR();
+
+    assert(lp->nouts == 0);	/* all outarcs will be ours */
+
+    /*
+     * Easy part of complementing, and all there is to do since the MCCE code
+     * was removed.
+     */
+
+    colorcomplement(v->nfa, v->cm, PLAIN, left, lp, rp);
+    NOERR();
+    dropstate(v->nfa, left);
+    assert(right->nins == 0);
+    freestate(v->nfa, right);
+    return;
+}
+
+/*
+ - brackpart - handle one item (or range) within a bracket expression
+ ^ static void brackpart(struct vars *, struct state *, struct state *);
+ */
+static void
+brackpart(
+    struct vars *v,
+    struct state *lp,
+    struct state *rp)
+{
+    celt startc, endc;
+    struct cvec *cv;
+    const chr *startp, *endp;
+    chr c;
+
+    /*
+     * Parse something, get rid of special cases, take shortcuts.
+     */
+
+    switch (v->nexttype) {
+    case RANGE:			/* a-b-c or other botch */
+	ERR(REG_ERANGE);
+	return;
+	break;
+    case PLAIN:
+	c = v->nextvalue;
+	NEXT();
+
+	/*
+	 * Shortcut for ordinary chr (not range).
+	 */
+
+	if (!SEE(RANGE)) {
+	    onechr(v, c, lp, rp);
+	    return;
+	}
+	startc = element(v, &c, &c+1);
+	NOERR();
+	break;
+    case COLLEL:
+	startp = v->now;
+	endp = scanplain(v);
+	INSIST(startp < endp, REG_ECOLLATE);
+	NOERR();
+	startc = element(v, startp, endp);
+	NOERR();
+	break;
+    case ECLASS:
+	startp = v->now;
+	endp = scanplain(v);
+	INSIST(startp < endp, REG_ECOLLATE);
+	NOERR();
+	startc = element(v, startp, endp);
+	NOERR();
+	cv = eclass(v, startc, (v->cflags&REG_ICASE));
+	NOERR();
+	dovec(v, cv, lp, rp);
+	return;
+	break;
+    case CCLASS:
+	startp = v->now;
+	endp = scanplain(v);
+	INSIST(startp < endp, REG_ECTYPE);
+	NOERR();
+	cv = cclass(v, startp, endp, (v->cflags&REG_ICASE));
+	NOERR();
+	dovec(v, cv, lp, rp);
+	return;
+	break;
+    default:
+	ERR(REG_ASSERT);
+	return;
+	break;
+    }
+
+    if (SEE(RANGE)) {
+	NEXT();
+	switch (v->nexttype) {
+	case PLAIN:
+	case RANGE:
+	    c = v->nextvalue;
+	    NEXT();
+	    endc = element(v, &c, &c+1);
+	    NOERR();
+	    break;
+	case COLLEL:
+	    startp = v->now;
+	    endp = scanplain(v);
+	    INSIST(startp < endp, REG_ECOLLATE);
+	    NOERR();
+	    endc = element(v, startp, endp);
+	    NOERR();
+	    break;
+	default:
+	    ERR(REG_ERANGE);
+	    return;
+	    break;
+	}
+    } else {
+	endc = startc;
+    }
+
+    /*
+     * Ranges are unportable. Actually, standard C does guarantee that digits
+     * are contiguous, but making that an exception is just too complicated.
+     */
+
+    if (startc != endc) {
+	NOTE(REG_UUNPORT);
+    }
+    cv = range(v, startc, endc, (v->cflags&REG_ICASE));
+    NOERR();
+    dovec(v, cv, lp, rp);
+}
+
+/*
+ - scanplain - scan PLAIN contents of [. etc.
+ * Certain bits of trickery in lex.c know that this code does not try to look
+ * past the final bracket of the [. etc.
+ ^ static const chr *scanplain(struct vars *);
+ */
+static const chr *		/* just after end of sequence */
+scanplain(
+    struct vars *v)
+{
+    const chr *endp;
+
+    assert(SEE(COLLEL) || SEE(ECLASS) || SEE(CCLASS));
+    NEXT();
+
+    endp = v->now;
+    while (SEE(PLAIN)) {
+	endp = v->now;
+	NEXT();
+    }
+
+    assert(SEE(END) || ISERR());
+    NEXT();
+
+    return endp;
+}
+
+/*
+ - onechr - fill in arcs for a plain character, and possible case complements
+ * This is mostly a shortcut for efficient handling of the common case.
+ ^ static void onechr(struct vars *, pchr, struct state *, struct state *);
+ */
+static void
+onechr(
+    struct vars *v,
+    pchr c,
+    struct state *lp,
+    struct state *rp)
+{
+    if (!(v->cflags&REG_ICASE)) {
+	newarc(v->nfa, PLAIN, subcolor(v->cm, c), lp, rp);
+	return;
+    }
+
+    /*
+     * Rats, need general case anyway...
+     */
+
+    dovec(v, allcases(v, c), lp, rp);
+}
+
+/*
+ - dovec - fill in arcs for each element of a cvec
+ ^ static void dovec(struct vars *, struct cvec *, struct state *,
+ ^ 	struct state *);
+ */
+static void
+dovec(
+    struct vars *v,
+    struct cvec *cv,
+    struct state *lp,
+    struct state *rp)
+{
+    chr ch, from, to;
+    const chr *p;
+    int i;
+
+    for (p = cv->chrs, i = cv->nchrs; i > 0; p++, i--) {
+	ch = *p;
+	newarc(v->nfa, PLAIN, subcolor(v->cm, ch), lp, rp);
+    }
+
+    for (p = cv->ranges, i = cv->nranges; i > 0; p += 2, i--) {
+	from = *p;
+	to = *(p+1);
+	if (from <= to) {
+	    subrange(v, from, to, lp, rp);
+	}
+    }
+
+}
+
+/*
+ - wordchrs - set up word-chr list for word-boundary stuff, if needed
+ * The list is kept as a bunch of arcs between two dummy states; it's disposed
+ * of by the unreachable-states sweep in NFA optimization. Does NEXT(). Must
+ * not be called from any unusual lexical context. This should be reconciled
+ * with the \w etc. handling in lex.c, and should be cleaned up to reduce
+ * dependencies on input scanning.
+ ^ static void wordchrs(struct vars *);
+ */
+static void
+wordchrs(
+    struct vars *v)
+{
+    struct state *left, *right;
+
+    if (v->wordchrs != NULL) {
+	NEXT();		/* for consistency */
+	return;
+    }
+
+    left = newstate(v->nfa);
+    right = newstate(v->nfa);
+    NOERR();
+
+    /*
+     * Fine point: implemented with [::], and lexer will set REG_ULOCALE.
+     */
+
+    lexword(v);
+    NEXT();
+    assert(v->savenow != NULL && SEE('['));
+    bracket(v, left, right);
+    assert((v->savenow != NULL && SEE(']')) || ISERR());
+    NEXT();
+    NOERR();
+    v->wordchrs = left;
+}
+
+/*
+ - subre - allocate a subre
+ ^ static struct subre *subre(struct vars *, int, int, struct state *,
+ ^	struct state *);
+ */
+static struct subre *
+subre(
+    struct vars *v,
+    int op,
+    int flags,
+    struct state *begin,
+    struct state *end)
+{
+    struct subre *ret = v->treefree;
+
+    if (ret != NULL) {
+	v->treefree = ret->left;
+    } else {
+	ret = (struct subre *) MALLOC(sizeof(struct subre));
+	if (ret == NULL) {
+	    ERR(REG_ESPACE);
+	    return NULL;
+	}
+	ret->chain = v->treechain;
+	v->treechain = ret;
+    }
+
+    assert(strchr("=b|.*(", op) != NULL);
+
+    ret->op = op;
+    ret->flags = flags;
+    ret->id = 0;		/* will be assigned later */
+    ret->subno = 0;
+    ret->min = ret->max = 1;
+    ret->left = NULL;
+    ret->right = NULL;
+    ret->begin = begin;
+    ret->end = end;
+    ZAPCNFA(ret->cnfa);
+
+    return ret;
+}
+
+/*
+ - freesubre - free a subRE subtree
+ ^ static void freesubre(struct vars *, struct subre *);
+ */
+static void
+freesubre(
+    struct vars *v,		/* might be NULL */
+    struct subre *sr)
+{
+    if (sr == NULL) {
+	return;
+    }
+
+    if (sr->left != NULL) {
+	freesubre(v, sr->left);
+    }
+    if (sr->right != NULL) {
+	freesubre(v, sr->right);
+    }
+
+    freesrnode(v, sr);
+}
+
+/*
+ - freesrnode - free one node in a subRE subtree
+ ^ static void freesrnode(struct vars *, struct subre *);
+ */
+static void
+freesrnode(
+    struct vars *v,		/* might be NULL */
+    struct subre *sr)
+{
+    if (sr == NULL) {
+	return;
+    }
+
+    if (!NULLCNFA(sr->cnfa)) {
+	freecnfa(&sr->cnfa);
+    }
+    sr->flags = 0;
+
+    if (v != NULL && v->treechain != NULL) {
+	/* we're still parsing, maybe we can reuse the subre */
+	sr->left = v->treefree;
+	v->treefree = sr;
+    } else {
+	FREE(sr);
+    }
+}
+
+/*
+ - optst - optimize a subRE subtree
+ ^ static void optst(struct vars *, struct subre *);
+ */
+static void
+optst(
+    struct vars *v,
+    struct subre *t)
+{
+    /*
+     * DGP (2007-11-13): I assume it was the programmer's intent to eventually
+     * come back and add code to optimize subRE trees, but the routine coded
+     * just spends effort traversing the tree and doing nothing. We can do
+     * nothing with less effort.
+     */
+
+    return;
+}
+
+/*
+ - numst - number tree nodes (assigning "id" indexes)
+ ^ static int numst(struct subre *, int);
+ */
+static int			/* next number */
+numst(
+    struct subre *t,
+    int start)			/* starting point for subtree numbers */
+{
+    int i;
+
+    assert(t != NULL);
+
+    i = start;
+    t->id = (short) i++;
+    if (t->left != NULL) {
+	i = numst(t->left, i);
+    }
+    if (t->right != NULL) {
+	i = numst(t->right, i);
+    }
+    return i;
+}
+
+/*
+ - markst - mark tree nodes as INUSE
+ * Note: this is a great deal more subtle than it looks.  During initial
+ * parsing of a regex, all subres are linked into the treechain list;
+ * discarded ones are also linked into the treefree list for possible reuse.
+ * After we are done creating all subres required for a regex, we run markst()
+ * then cleanst(), which results in discarding all subres not reachable from
+ * v->tree.  We then clear v->treechain, indicating that subres must be found
+ * by descending from v->tree.  This changes the behavior of freesubre(): it
+ * will henceforth FREE() unwanted subres rather than sticking them into the
+ * treefree list.  (Doing that any earlier would result in dangling links in
+ * the treechain list.)  This all means that freev() will clean up correctly
+ * if invoked before or after markst()+cleanst(); but it would not work if
+ * called partway through this state conversion, so we mustn't error out
+ * in or between these two functions.
+ ^ static void markst(struct subre *);
+ */
+static void
+markst(
+    struct subre *t)
+{
+    assert(t != NULL);
+
+    t->flags |= INUSE;
+    if (t->left != NULL) {
+	markst(t->left);
+    }
+    if (t->right != NULL) {
+	markst(t->right);
+    }
+}
+
+/*
+ - cleanst - free any tree nodes not marked INUSE
+ ^ static void cleanst(struct vars *);
+ */
+static void
+cleanst(
+    struct vars *v)
+{
+    struct subre *t;
+    struct subre *next;
+
+    for (t = v->treechain; t != NULL; t = next) {
+	next = t->chain;
+	if (!(t->flags&INUSE)) {
+	    FREE(t);
+	}
+    }
+    v->treechain = NULL;
+    v->treefree = NULL;		/* just on general principles */
+}
+
+/*
+ - nfatree - turn a subRE subtree into a tree of compacted NFAs
+ ^ static long nfatree(struct vars *, struct subre *, FILE *);
+ */
+static long			/* optimize results from top node */
+nfatree(
+    struct vars *v,
+    struct subre *t,
+    FILE *f)			/* for debug output */
+{
+    assert(t != NULL && t->begin != NULL);
+
+    if (t->left != NULL) {
+	(void) nfatree(v, t->left, f);
+    }
+    if (t->right != NULL) {
+	(void) nfatree(v, t->right, f);
+    }
+
+    return nfanode(v, t, f);
+}
+
+/*
+ - nfanode - do one NFA for nfatree
+ ^ static long nfanode(struct vars *, struct subre *, FILE *);
+ */
+static long			/* optimize results */
+nfanode(
+    struct vars *v,
+    struct subre *t,
+    FILE *f)			/* for debug output */
+{
+    struct nfa *nfa;
+    long ret = 0;
+    char idbuf[50];
+
+    assert(t->begin != NULL);
+
+    if (f != NULL) {
+	fprintf(f, "\n\n\n========= TREE NODE %s ==========\n",
+		stid(t, idbuf, sizeof(idbuf)));
+    }
+    nfa = newnfa(v, v->cm, v->nfa);
+    NOERRZ();
+    dupnfa(nfa, t->begin, t->end, nfa->init, nfa->final);
+    if (!ISERR()) {
+	specialcolors(nfa);
+	ret = optimize(nfa, f);
+    }
+    if (!ISERR()) {
+	compact(nfa, &t->cnfa);
+    }
+
+    freenfa(nfa);
+    return ret;
+}
+
+/*
+ - newlacon - allocate a lookahead-constraint subRE
+ ^ static int newlacon(struct vars *, struct state *, struct state *, int);
+ */
+static int			/* lacon number */
+newlacon(
+    struct vars *v,
+    struct state *begin,
+    struct state *end,
+    int pos)
+{
+    int n;
+    struct subre *newlacons;
+    struct subre *sub;
+
+    if (v->nlacons == 0) {
+	n = 1;		/* skip 0th */
+	newlacons = (struct subre *) MALLOC(2 * sizeof(struct subre));
+    } else {
+	n = v->nlacons;
+	newlacons = (struct subre *) REALLOC(v->lacons,
+					     (n + 1) * sizeof(struct subre));
+    }
+
+    if (newlacons == NULL) {
+	ERR(REG_ESPACE);
+	return 0;
+    }
+
+    v->lacons = newlacons;
+    v->nlacons = n + 1;
+    sub = &v->lacons[n];
+    sub->begin = begin;
+    sub->end = end;
+    sub->subno = pos;
+    ZAPCNFA(sub->cnfa);
+    return n;
+}
+
+/*
+ - freelacons - free lookahead-constraint subRE vector
+ ^ static void freelacons(struct subre *, int);
+ */
+static void
+freelacons(
+    struct subre *subs,
+    int n)
+{
+    struct subre *sub;
+    int i;
+
+    assert(n > 0);
+    for (sub=subs+1, i=n-1; i>0; sub++, i--) {	/* no 0th */
+	if (!NULLCNFA(sub->cnfa)) {
+	    freecnfa(&sub->cnfa);
+	}
+    }
+    FREE(subs);
+}
+
+/*
+ - rfree - free a whole RE (insides of regfree)
+ ^ static void rfree(regex_t *);
+ */
+static void
+rfree(
+    regex_t *re)
+{
+    struct guts *g;
+
+    if (re == NULL || re->re_magic != REMAGIC) {
+	return;
+    }
+
+    re->re_magic = 0;	/* invalidate RE */
+    g = (struct guts *) re->re_guts;
+    re->re_guts = NULL;
+    re->re_fns = NULL;
+    if (g != NULL) {
+	g->magic = 0;
+	freecm(&g->cmap);
+	if (g->tree != NULL) {
+	    freesubre(NULL, g->tree);
+	}
+	if (g->lacons != NULL) {
+	    freelacons(g->lacons, g->nlacons);
+	}
+	if (!NULLCNFA(g->search)) {
+	    freecnfa(&g->search);
+	}
+	FREE(g);
+    }
+}
+
+/*
+ - dump - dump an RE in human-readable form
+ ^ static void dump(regex_t *, FILE *);
+ */
+static void
+dump(
+    regex_t *re,
+    FILE *f)
+{
+#ifdef REG_DEBUG
+    struct guts *g;
+    int i;
+
+    if (re->re_magic != REMAGIC) {
+	fprintf(f, "bad magic number (0x%x not 0x%x)\n",
+		re->re_magic, REMAGIC);
+    }
+    if (re->re_guts == NULL) {
+	fprintf(f, "NULL guts!!!\n");
+	return;
+    }
+    g = (struct guts *) re->re_guts;
+    if (g->magic != GUTSMAGIC) {
+	fprintf(f, "bad guts magic number (0x%x not 0x%x)\n",
+		g->magic, GUTSMAGIC);
+    }
+
+    fprintf(f, "\n\n\n========= DUMP ==========\n");
+    fprintf(f, "nsub %d, info 0%lo, csize %d, ntree %d\n",
+	    (int) re->re_nsub, re->re_info, re->re_csize, g->ntree);
+
+    dumpcolors(&g->cmap, f);
+    if (!NULLCNFA(g->search)) {
+	fprintf(f, "\nsearch:\n");
+	dumpcnfa(&g->search, f);
+    }
+    for (i = 1; i < g->nlacons; i++) {
+	fprintf(f, "\nla%d (%s):\n", i,
+		(g->lacons[i].subno) ? "positive" : "negative");
+	dumpcnfa(&g->lacons[i].cnfa, f);
+    }
+    fprintf(f, "\n");
+    dumpst(g->tree, f, 0);
+#endif
+}
+
+/*
+ - dumpst - dump a subRE tree
+ ^ static void dumpst(struct subre *, FILE *, int);
+ */
+static void
+dumpst(
+    struct subre *t,
+    FILE *f,
+    int nfapresent)		/* is the original NFA still around? */
+{
+    if (t == NULL) {
+	fprintf(f, "null tree\n");
+    } else {
+	stdump(t, f, nfapresent);
+    }
+    fflush(f);
+}
+
+/*
+ - stdump - recursive guts of dumpst
+ ^ static void stdump(struct subre *, FILE *, int);
+ */
+static void
+stdump(
+    struct subre *t,
+    FILE *f,
+    int nfapresent)		/* is the original NFA still around? */
+{
+    char idbuf[50];
+
+    fprintf(f, "%s. `%c'", stid(t, idbuf, sizeof(idbuf)), t->op);
+    if (t->flags&LONGER) {
+	fprintf(f, " longest");
+    }
+    if (t->flags&SHORTER) {
+	fprintf(f, " shortest");
+    }
+    if (t->flags&MIXED) {
+	fprintf(f, " hasmixed");
+    }
+    if (t->flags&CAP) {
+	fprintf(f, " hascapture");
+    }
+    if (t->flags&BACKR) {
+	fprintf(f, " hasbackref");
+    }
+    if (!(t->flags&INUSE)) {
+	fprintf(f, " UNUSED");
+    }
+    if (t->subno != 0) {
+	fprintf(f, " (#%d)", t->subno);
+    }
+    if (t->min != 1 || t->max != 1) {
+	fprintf(f, " {%d,", t->min);
+	if (t->max != DUPINF) {
+	    fprintf(f, "%d", t->max);
+	}
+	fprintf(f, "}");
+    }
+    if (nfapresent) {
+	fprintf(f, " %ld-%ld", (long)t->begin->no, (long)t->end->no);
+    }
+    if (t->left != NULL) {
+	fprintf(f, " L:%s", stid(t->left, idbuf, sizeof(idbuf)));
+    }
+    if (t->right != NULL) {
+	fprintf(f, " R:%s", stid(t->right, idbuf, sizeof(idbuf)));
+    }
+    if (!NULLCNFA(t->cnfa)) {
+	fprintf(f, "\n");
+	dumpcnfa(&t->cnfa, f);
+    }
+    fprintf(f, "\n");
+    if (t->left != NULL) {
+	stdump(t->left, f, nfapresent);
+    }
+    if (t->right != NULL) {
+	stdump(t->right, f, nfapresent);
+    }
+}
+
+/*
+ - stid - identify a subtree node for dumping
+ ^ static const char *stid(struct subre *, char *, size_t);
+ */
+static const char *			/* points to buf or constant string */
+stid(
+    struct subre *t,
+    char *buf,
+    size_t bufsize)
+{
+    /*
+     * Big enough for hex int or decimal t->id?
+     */
+
+    if (bufsize < sizeof(void*)*2 + 3 || bufsize < sizeof(t->id)*3 + 1) {
+	return "unable";
+    }
+    if (t->id != 0) {
+	sprintf(buf, "%d", t->id);
+    } else {
+	sprintf(buf, "%p", t);
+    }
+    return buf;
+}
+
+#include "regc_lex.c"
+#include "regc_color.c"
+#include "regc_nfa.c"
+#include "regc_cvec.c"
+#include "regc_locale.c"
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regcustom.h b/generic/regcustom.h
new file mode 100644
index 0000000..c4dbc73
--- /dev/null
+++ b/generic/regcustom.h
@@ -0,0 +1,154 @@
+/*
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms - with or without
+ * modification - are permitted for any purpose, provided that redistributions
+ * in source form retain this entire copyright notice and indicate the origin
+ * and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * Headers if any.
+ */
+
+#include "regex.h"
+
+/*
+ * Overrides for regguts.h definitions, if any.
+ */
+
+#define	MALLOC(n)		(void*)(attemptckalloc(n))
+#define	FREE(p)			ckfree((void*)(p))
+#define	REALLOC(p,n)		(void*)(attemptckrealloc((void*)(p),n))
+
+/*
+ * Do not insert extras between the "begin" and "end" lines - this chunk is
+ * automatically extracted to be fitted into regex.h.
+ */
+
+/* --- begin --- */
+/* Ensure certain things don't sneak in from system headers. */
+#ifdef __REG_WIDE_T
+#undef __REG_WIDE_T
+#endif
+#ifdef __REG_WIDE_COMPILE
+#undef __REG_WIDE_COMPILE
+#endif
+#ifdef __REG_WIDE_EXEC
+#undef __REG_WIDE_EXEC
+#endif
+#ifdef __REG_REGOFF_T
+#undef __REG_REGOFF_T
+#endif
+#ifdef __REG_NOFRONT
+#undef __REG_NOFRONT
+#endif
+#ifdef __REG_NOCHAR
+#undef __REG_NOCHAR
+#endif
+/* Interface types */
+#define	__REG_WIDE_T	Tcl_UniChar
+#define	__REG_REGOFF_T	long	/* Not really right, but good enough... */
+/* Names and declarations */
+#define	__REG_WIDE_COMPILE	TclReComp
+#define	__REG_WIDE_EXEC		TclReExec
+#define	__REG_NOFRONT		/* Don't want regcomp() and regexec() */
+#define	__REG_NOCHAR		/* Or the char versions */
+#define	regfree		TclReFree
+#define	regerror	TclReError
+/* --- end --- */
+
+/*
+ * Internal character type and related.
+ */
+
+typedef Tcl_UniChar chr;	/* The type itself. */
+typedef int pchr;		/* What it promotes to. */
+typedef unsigned uchr;		/* Unsigned type that will hold a chr. */
+typedef int celt;		/* Type to hold chr, or NOCELT */
+#define	NOCELT (-1)		/* Celt value which is not valid chr */
+#define	CHR(c) (UCHAR(c))	/* Turn char literal into chr literal */
+#define	DIGITVAL(c) ((c)-'0')	/* Turn chr digit into its value */
+#if TCL_UTF_MAX > 4
+#define	CHRBITS	32		/* Bits in a chr; must not use sizeof */
+#define	CHR_MIN	0x00000000	/* Smallest and largest chr; the value */
+#define	CHR_MAX	0xffffffff	/* CHR_MAX-CHR_MIN+1 should fit in uchr */
+#else
+#define	CHRBITS	16		/* Bits in a chr; must not use sizeof */
+#define	CHR_MIN	0x0000		/* Smallest and largest chr; the value */
+#define	CHR_MAX	0xffff		/* CHR_MAX-CHR_MIN+1 should fit in uchr */
+#endif
+
+/*
+ * Functions operating on chr.
+ */
+
+#define	iscalnum(x)	Tcl_UniCharIsAlnum(x)
+#define	iscalpha(x)	Tcl_UniCharIsAlpha(x)
+#define	iscdigit(x)	Tcl_UniCharIsDigit(x)
+#define	iscspace(x)	Tcl_UniCharIsSpace(x)
+
+/*
+ * Name the external functions.
+ */
+
+#define	compile		TclReComp
+#define	exec		TclReExec
+
+/*
+& Enable/disable debugging code (by whether REG_DEBUG is defined or not).
+*/
+
+#if 0				/* No debug unless requested by makefile. */
+#define	REG_DEBUG	/* */
+#endif
+
+/*
+ * Method of allocating a local workspace. We used a thread-specific data
+ * space to store this because the regular expression engine is never
+ * reentered from the same thread; it doesn't make any callbacks.
+ */
+
+#if 1
+#define AllocVars(vPtr) \
+    static Tcl_ThreadDataKey varsKey; \
+    register struct vars *vPtr = (struct vars *) \
+	    Tcl_GetThreadData(&varsKey, sizeof(struct vars))
+#else
+/*
+ * This strategy for allocating workspace is "more proper" in some sense, but
+ * quite a bit slower. Using TSD (as above) leads to code that is quite a bit
+ * faster in practice (measured!)
+ */
+#define AllocVars(vPtr) \
+    register struct vars *vPtr = (struct vars *) MALLOC(sizeof(struct vars))
+#define FreeVars(vPtr) \
+    FREE(vPtr)
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/rege_dfa.c b/generic/rege_dfa.c
new file mode 100644
index 0000000..e5f22c4
--- /dev/null
+++ b/generic/rege_dfa.c
@@ -0,0 +1,805 @@
+/*
+ * DFA routines
+ * This file is #included by regexec.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ - longest - longest-preferred matching engine
+ ^ static chr *longest(struct vars *, struct dfa *, chr *, chr *, int *);
+ */
+static chr *			/* endpoint, or NULL */
+longest(
+    struct vars *const v,	/* used only for debug and exec flags */
+    struct dfa *const d,
+    chr *const start,		/* where the match should start */
+    chr *const stop,		/* match must end at or before here */
+    int *const hitstopp)	/* record whether hit v->stop, if non-NULL */
+{
+    chr *cp;
+    chr *realstop = (stop == v->stop) ? stop : stop + 1;
+    color co;
+    struct sset *css, *ss;
+    chr *post;
+    int i;
+    struct colormap *cm = d->cm;
+
+    /*
+     * Initialize.
+     */
+
+    css = initialize(v, d, start);
+    cp = start;
+    if (hitstopp != NULL) {
+	*hitstopp = 0;
+    }
+
+    /*
+     * Startup.
+     */
+
+    FDEBUG(("+++ startup +++\n"));
+    if (cp == v->start) {
+	co = d->cnfa->bos[(v->eflags&REG_NOTBOL) ? 0 : 1];
+	FDEBUG(("color %ld\n", (long)co));
+    } else {
+	co = GETCOLOR(cm, *(cp - 1));
+	FDEBUG(("char %c, color %ld\n", (char)*(cp-1), (long)co));
+    }
+    css = miss(v, d, css, co, cp, start);
+    if (css == NULL) {
+	return NULL;
+    }
+    css->lastseen = cp;
+
+    /*
+     * Main loop.
+     */
+
+    if (v->eflags&REG_FTRACE) {
+	while (cp < realstop) {
+	    FDEBUG(("+++ at c%d +++\n", (int) (css - d->ssets)));
+	    co = GETCOLOR(cm, *cp);
+	    FDEBUG(("char %c, color %ld\n", (char)*cp, (long)co));
+	    ss = css->outs[co];
+	    if (ss == NULL) {
+		ss = miss(v, d, css, co, cp+1, start);
+		if (ss == NULL) {
+		    break;	/* NOTE BREAK OUT */
+		}
+	    }
+	    cp++;
+	    ss->lastseen = cp;
+	    css = ss;
+	}
+    } else {
+	while (cp < realstop) {
+	    co = GETCOLOR(cm, *cp);
+	    ss = css->outs[co];
+	    if (ss == NULL) {
+		ss = miss(v, d, css, co, cp+1, start);
+		if (ss == NULL) {
+		    break;	/* NOTE BREAK OUT */
+		}
+	    }
+	    cp++;
+	    ss->lastseen = cp;
+	    css = ss;
+	}
+    }
+
+    /*
+     * Shutdown.
+     */
+
+    FDEBUG(("+++ shutdown at c%d +++\n", (int) (css - d->ssets)));
+    if (cp == v->stop && stop == v->stop) {
+	if (hitstopp != NULL) {
+	    *hitstopp = 1;
+	}
+	co = d->cnfa->eos[(v->eflags&REG_NOTEOL) ? 0 : 1];
+	FDEBUG(("color %ld\n", (long)co));
+	ss = miss(v, d, css, co, cp, start);
+
+	/*
+	 * Special case: match ended at eol?
+	 */
+
+	if (ss != NULL && (ss->flags&POSTSTATE)) {
+	    return cp;
+	} else if (ss != NULL) {
+	    ss->lastseen = cp;	/* to be tidy */
+	}
+    }
+
+    /*
+     * Find last match, if any.
+     */
+
+    post = d->lastpost;
+    for (ss = d->ssets, i = d->nssused; i > 0; ss++, i--) {
+	if ((ss->flags&POSTSTATE) && (post != ss->lastseen) &&
+		(post == NULL || post < ss->lastseen)) {
+	    post = ss->lastseen;
+	}
+    }
+    if (post != NULL) {		/* found one */
+	return post - 1;
+    }
+
+    return NULL;
+}
+
+/*
+ - shortest - shortest-preferred matching engine
+ ^ static chr *shortest(struct vars *, struct dfa *, chr *, chr *, chr *,
+ ^ 	chr **, int *);
+ */
+static chr *			/* endpoint, or NULL */
+shortest(
+    struct vars *const v,
+    struct dfa *const d,
+    chr *const start,		/* where the match should start */
+    chr *const min,		/* match must end at or after here */
+    chr *const max,		/* match must end at or before here */
+    chr **const coldp,		/* store coldstart pointer here, if nonNULL */
+    int *const hitstopp)	/* record whether hit v->stop, if non-NULL */
+{
+    chr *cp;
+    chr *realmin = (min == v->stop) ? min : min + 1;
+    chr *realmax = (max == v->stop) ? max : max + 1;
+    color co;
+    struct sset *css, *ss;
+    struct colormap *cm = d->cm;
+
+    /*
+     * Initialize.
+     */
+
+    css = initialize(v, d, start);
+    cp = start;
+    if (hitstopp != NULL) {
+	*hitstopp = 0;
+    }
+
+    /*
+     * Startup.
+     */
+
+    FDEBUG(("--- startup ---\n"));
+    if (cp == v->start) {
+	co = d->cnfa->bos[(v->eflags&REG_NOTBOL) ? 0 : 1];
+	FDEBUG(("color %ld\n", (long)co));
+    } else {
+	co = GETCOLOR(cm, *(cp - 1));
+	FDEBUG(("char %c, color %ld\n", (char)*(cp-1), (long)co));
+    }
+    css = miss(v, d, css, co, cp, start);
+    if (css == NULL) {
+	return NULL;
+    }
+    css->lastseen = cp;
+    ss = css;
+
+    /*
+     * Main loop.
+     */
+
+    if (v->eflags&REG_FTRACE) {
+	while (cp < realmax) {
+	    FDEBUG(("--- at c%d ---\n", (int) (css - d->ssets)));
+	    co = GETCOLOR(cm, *cp);
+	    FDEBUG(("char %c, color %ld\n", (char)*cp, (long)co));
+	    ss = css->outs[co];
+	    if (ss == NULL) {
+		ss = miss(v, d, css, co, cp+1, start);
+		if (ss == NULL) {
+		    break;	/* NOTE BREAK OUT */
+		}
+	    }
+	    cp++;
+	    ss->lastseen = cp;
+	    css = ss;
+	    if ((ss->flags&POSTSTATE) && cp >= realmin) {
+		break;		/* NOTE BREAK OUT */
+	    }
+	}
+    } else {
+	while (cp < realmax) {
+	    co = GETCOLOR(cm, *cp);
+	    ss = css->outs[co];
+	    if (ss == NULL) {
+		ss = miss(v, d, css, co, cp+1, start);
+		if (ss == NULL) {
+		    break;	/* NOTE BREAK OUT */
+		}
+	    }
+	    cp++;
+	    ss->lastseen = cp;
+	    css = ss;
+	    if ((ss->flags&POSTSTATE) && cp >= realmin) {
+		break;		/* NOTE BREAK OUT */
+	    }
+	}
+    }
+
+    if (ss == NULL) {
+	return NULL;
+    }
+
+    if (coldp != NULL) {	/* report last no-progress state set, if any */
+	*coldp = lastCold(v, d);
+    }
+
+    if ((ss->flags&POSTSTATE) && cp > min) {
+	assert(cp >= realmin);
+	cp--;
+    } else if (cp == v->stop && max == v->stop) {
+	co = d->cnfa->eos[(v->eflags&REG_NOTEOL) ? 0 : 1];
+	FDEBUG(("color %ld\n", (long)co));
+	ss = miss(v, d, css, co, cp, start);
+
+	/*
+	 * Match might have ended at eol.
+	 */
+
+	if ((ss == NULL || !(ss->flags&POSTSTATE)) && hitstopp != NULL) {
+	    *hitstopp = 1;
+	}
+    }
+
+    if (ss == NULL || !(ss->flags&POSTSTATE)) {
+	return NULL;
+    }
+
+    return cp;
+}
+
+/*
+ - lastCold - determine last point at which no progress had been made
+ ^ static chr *lastCold(struct vars *, struct dfa *);
+ */
+static chr *			/* endpoint, or NULL */
+lastCold(
+    struct vars *const v,
+    struct dfa *const d)
+{
+    struct sset *ss;
+    chr *nopr = d->lastnopr;
+    int i;
+
+    if (nopr == NULL) {
+	nopr = v->start;
+    }
+    for (ss = d->ssets, i = d->nssused; i > 0; ss++, i--) {
+	if ((ss->flags&NOPROGRESS) && nopr < ss->lastseen) {
+	    nopr = ss->lastseen;
+	}
+    }
+    return nopr;
+}
+
+/*
+ - newDFA - set up a fresh DFA
+ ^ static struct dfa *newDFA(struct vars *, struct cnfa *,
+ ^ 	struct colormap *, struct smalldfa *);
+ */
+static struct dfa *
+newDFA(
+    struct vars *const v,
+    struct cnfa *const cnfa,
+    struct colormap *const cm,
+    struct smalldfa *sml)	/* preallocated space, may be NULL */
+{
+    struct dfa *d;
+    size_t nss = cnfa->nstates * 2;
+    int wordsper = (cnfa->nstates + UBITS - 1) / UBITS;
+    struct smalldfa *smallwas = sml;
+
+    assert(cnfa != NULL && cnfa->nstates != 0);
+
+    if (nss <= FEWSTATES && cnfa->ncolors <= FEWCOLORS) {
+	assert(wordsper == 1);
+	if (sml == NULL) {
+	    sml = (struct smalldfa *) MALLOC(sizeof(struct smalldfa));
+	    if (sml == NULL) {
+		ERR(REG_ESPACE);
+		return NULL;
+	    }
+	}
+	d = &sml->dfa;
+	d->ssets = sml->ssets;
+	d->statesarea = sml->statesarea;
+	d->work = &d->statesarea[nss];
+	d->outsarea = sml->outsarea;
+	d->incarea = sml->incarea;
+	d->cptsmalloced = 0;
+	d->mallocarea = (smallwas == NULL) ? (char *)sml : NULL;
+    } else {
+	d = (struct dfa *) MALLOC(sizeof(struct dfa));
+	if (d == NULL) {
+	    ERR(REG_ESPACE);
+	    return NULL;
+	}
+	d->ssets = (struct sset *) MALLOC(nss * sizeof(struct sset));
+	d->statesarea = (unsigned *)
+		MALLOC((nss+WORK) * wordsper * sizeof(unsigned));
+	d->work = &d->statesarea[nss * wordsper];
+	d->outsarea = (struct sset **)
+		MALLOC(nss * cnfa->ncolors * sizeof(struct sset *));
+	d->incarea = (struct arcp *)
+		MALLOC(nss * cnfa->ncolors * sizeof(struct arcp));
+	d->cptsmalloced = 1;
+	d->mallocarea = (char *)d;
+	if (d->ssets == NULL || d->statesarea == NULL ||
+		d->outsarea == NULL || d->incarea == NULL) {
+	    freeDFA(d);
+	    ERR(REG_ESPACE);
+	    return NULL;
+	}
+    }
+
+    d->nssets = (v->eflags&REG_SMALL) ? 7 : nss;
+    d->nssused = 0;
+    d->nstates = cnfa->nstates;
+    d->ncolors = cnfa->ncolors;
+    d->wordsper = wordsper;
+    d->cnfa = cnfa;
+    d->cm = cm;
+    d->lastpost = NULL;
+    d->lastnopr = NULL;
+    d->search = d->ssets;
+
+    /*
+     * Initialization of sset fields is done as needed.
+     */
+
+    return d;
+}
+
+/*
+ - freeDFA - free a DFA
+ ^ static void freeDFA(struct dfa *);
+ */
+static void
+freeDFA(
+    struct dfa *const d)
+{
+    if (d->cptsmalloced) {
+	if (d->ssets != NULL) {
+	    FREE(d->ssets);
+	}
+	if (d->statesarea != NULL) {
+	    FREE(d->statesarea);
+	}
+	if (d->outsarea != NULL) {
+	    FREE(d->outsarea);
+	}
+	if (d->incarea != NULL) {
+	    FREE(d->incarea);
+	}
+    }
+
+    if (d->mallocarea != NULL) {
+	FREE(d->mallocarea);
+    }
+}
+
+/*
+ - hash - construct a hash code for a bitvector
+ * There are probably better ways, but they're more expensive.
+ ^ static unsigned hash(unsigned *, int);
+ */
+static unsigned
+hash(
+    unsigned *const uv,
+    const int n)
+{
+    int i;
+    unsigned h;
+
+    h = 0;
+    for (i = 0; i < n; i++) {
+	h ^= uv[i];
+    }
+    return h;
+}
+
+/*
+ - initialize - hand-craft a cache entry for startup, otherwise get ready
+ ^ static struct sset *initialize(struct vars *, struct dfa *, chr *);
+ */
+static struct sset *
+initialize(
+    struct vars *const v,	/* used only for debug flags */
+    struct dfa *const d,
+    chr *const start)
+{
+    struct sset *ss;
+    int i;
+
+    /*
+     * Is previous one still there?
+     */
+
+    if (d->nssused > 0 && (d->ssets[0].flags&STARTER)) {
+	ss = &d->ssets[0];
+    } else {			/* no, must (re)build it */
+	ss = getVacantSS(v, d, start, start);
+	for (i = 0; i < d->wordsper; i++) {
+	    ss->states[i] = 0;
+	}
+	BSET(ss->states, d->cnfa->pre);
+	ss->hash = HASH(ss->states, d->wordsper);
+	assert(d->cnfa->pre != d->cnfa->post);
+	ss->flags = STARTER|LOCKED|NOPROGRESS;
+
+	/*
+	 * lastseen dealt with below
+	 */
+    }
+
+    for (i = 0; i < d->nssused; i++) {
+	d->ssets[i].lastseen = NULL;
+    }
+    ss->lastseen = start;	/* maybe untrue, but harmless */
+    d->lastpost = NULL;
+    d->lastnopr = NULL;
+    return ss;
+}
+
+/*
+ - miss - handle a cache miss
+ ^ static struct sset *miss(struct vars *, struct dfa *, struct sset *,
+ ^ 	pcolor, chr *, chr *);
+ */
+static struct sset *		/* NULL if goes to empty set */
+miss(
+    struct vars *const v,	/* used only for debug flags */
+    struct dfa *const d,
+    struct sset *const css,
+    const pcolor co,
+    chr *const cp,		/* next chr */
+    chr *const start)		/* where the attempt got started */
+{
+    struct cnfa *cnfa = d->cnfa;
+    unsigned h;
+    struct carc *ca;
+    struct sset *p;
+    int i, isPost, noProgress, gotState, doLAConstraints, sawLAConstraints;
+
+    /*
+     * For convenience, we can be called even if it might not be a miss.
+     */
+
+    if (css->outs[co] != NULL) {
+	FDEBUG(("hit\n"));
+	return css->outs[co];
+    }
+    FDEBUG(("miss\n"));
+
+    /*
+     * First, what set of states would we end up in?
+     */
+
+    for (i = 0; i < d->wordsper; i++) {
+	d->work[i] = 0;
+    }
+    isPost = 0;
+    noProgress = 1;
+    gotState = 0;
+    for (i = 0; i < d->nstates; i++) {
+	if (ISBSET(css->states, i)) {
+	    for (ca = cnfa->states[i]; ca->co != COLORLESS; ca++) {
+		if (ca->co == co) {
+		    BSET(d->work, ca->to);
+		    gotState = 1;
+		    if (ca->to == cnfa->post) {
+			isPost = 1;
+		    }
+		    if (!(cnfa->stflags[ca->to] & CNFA_NOPROGRESS)) {
+			noProgress = 0;
+		    }
+		    FDEBUG(("%d -> %d\n", i, ca->to));
+		}
+	    }
+	}
+    }
+    doLAConstraints = (gotState ? (cnfa->flags&HASLACONS) : 0);
+    sawLAConstraints = 0;
+    while (doLAConstraints) {		/* transitive closure */
+	doLAConstraints = 0;
+	for (i = 0; i < d->nstates; i++) {
+	    if (ISBSET(d->work, i)) {
+		for (ca = cnfa->states[i]; ca->co != COLORLESS; ca++) {
+		    if (ca->co < cnfa->ncolors) {
+			continue;	/* NOTE CONTINUE */
+		    }
+		    sawLAConstraints = 1;
+		    if (ISBSET(d->work, ca->to)) {
+			continue;	/* NOTE CONTINUE */
+		    }
+		    if (!checkLAConstraint(v, cnfa, cp, ca->co)) {
+			continue;	/* NOTE CONTINUE */
+		    }
+		    BSET(d->work, ca->to);
+		    doLAConstraints = 1;
+		    if (ca->to == cnfa->post) {
+			isPost = 1;
+		    }
+		    if (!(cnfa->stflags[ca->to] & CNFA_NOPROGRESS)) {
+			noProgress = 0;
+		    }
+		    FDEBUG(("%d :> %d\n", i, ca->to));
+		}
+	    }
+	}
+    }
+    if (!gotState) {
+	return NULL;
+    }
+    h = HASH(d->work, d->wordsper);
+
+    /*
+     * Next, is that in the cache?
+     */
+
+    for (p = d->ssets, i = d->nssused; i > 0; p++, i--) {
+	 if (HIT(h, d->work, p, d->wordsper)) {
+	     FDEBUG(("cached c%d\n", (int) (p - d->ssets)));
+	     break;			/* NOTE BREAK OUT */
+	 }
+    }
+    if (i == 0) {		/* nope, need a new cache entry */
+	p = getVacantSS(v, d, cp, start);
+	assert(p != css);
+	for (i = 0; i < d->wordsper; i++) {
+	    p->states[i] = d->work[i];
+	}
+	p->hash = h;
+	p->flags = (isPost ? POSTSTATE : 0);
+	if (noProgress) {
+	    p->flags |= NOPROGRESS;
+	}
+
+	/*
+	 * lastseen to be dealt with by caller
+	 */
+    }
+
+    if (!sawLAConstraints) {	/* lookahead conds. always cache miss */
+	FDEBUG(("c%d[%d]->c%d\n",
+		(int) (css - d->ssets), co, (int) (p - d->ssets)));
+	css->outs[co] = p;
+	css->inchain[co] = p->ins;
+	p->ins.ss = css;
+	p->ins.co = (color) co;
+    }
+    return p;
+}
+
+/*
+ - checkLAConstraint - lookahead-constraint checker for miss()
+ ^ static int checkLAConstraint(struct vars *, struct cnfa *, chr *, pcolor);
+ */
+static int			/* predicate:  constraint satisfied? */
+checkLAConstraint(
+    struct vars *const v,
+    struct cnfa *const pcnfa,	/* parent cnfa */
+    chr *const cp,
+    const pcolor co)		/* "color" of the lookahead constraint */
+{
+    int n;
+    struct subre *sub;
+    struct dfa *d;
+    struct smalldfa sd;
+    chr *end;
+
+    n = co - pcnfa->ncolors;
+    assert(n < v->g->nlacons && v->g->lacons != NULL);
+    FDEBUG(("=== testing lacon %d\n", n));
+    sub = &v->g->lacons[n];
+    d = newDFA(v, &sub->cnfa, &v->g->cmap, &sd);
+    if (d == NULL) {
+	ERR(REG_ESPACE);
+	return 0;
+    }
+    end = longest(v, d, cp, v->stop, NULL);
+    freeDFA(d);
+    FDEBUG(("=== lacon %d match %d\n", n, (end != NULL)));
+    return (sub->subno) ? (end != NULL) : (end == NULL);
+}
+
+/*
+ - getVacantSS - get a vacant state set
+ * This routine clears out the inarcs and outarcs, but does not otherwise
+ * clear the innards of the state set -- that's up to the caller.
+ ^ static struct sset *getVacantSS(struct vars *, struct dfa *, chr *, chr *);
+ */
+static struct sset *
+getVacantSS(
+    struct vars *const v,	/* used only for debug flags */
+    struct dfa *const d,
+    chr *const cp,
+    chr *const start)
+{
+    int i;
+    struct sset *ss, *p;
+    struct arcp ap, lastap = {NULL, 0}; /* silence gcc 4 warning */
+    color co;
+
+    ss = pickNextSS(v, d, cp, start);
+    assert(!(ss->flags&LOCKED));
+
+    /*
+     * Clear out its inarcs, including self-referential ones.
+     */
+
+    ap = ss->ins;
+    while ((p = ap.ss) != NULL) {
+	co = ap.co;
+	FDEBUG(("zapping c%d's %ld outarc\n", (int) (p - d->ssets), (long)co));
+	p->outs[co] = NULL;
+	ap = p->inchain[co];
+	p->inchain[co].ss = NULL; /* paranoia */
+    }
+    ss->ins.ss = NULL;
+
+    /*
+     * Take it off the inarc chains of the ssets reached by its outarcs.
+     */
+
+    for (i = 0; i < d->ncolors; i++) {
+	p = ss->outs[i];
+	assert(p != ss);	/* not self-referential */
+	if (p == NULL) {
+	    continue;		/* NOTE CONTINUE */
+	}
+	FDEBUG(("del outarc %d from c%d's in chn\n", i, (int) (p - d->ssets)));
+	if (p->ins.ss == ss && p->ins.co == i) {
+	    p->ins = ss->inchain[i];
+	} else {
+	    assert(p->ins.ss != NULL);
+	    for (ap = p->ins; ap.ss != NULL && !(ap.ss == ss && ap.co == i);
+		    ap = ap.ss->inchain[ap.co]) {
+		lastap = ap;
+	    }
+	    assert(ap.ss != NULL);
+	    lastap.ss->inchain[lastap.co] = ss->inchain[i];
+	}
+	ss->outs[i] = NULL;
+	ss->inchain[i].ss = NULL;
+    }
+
+    /*
+     * If ss was a success state, may need to remember location.
+     */
+
+    if ((ss->flags&POSTSTATE) && ss->lastseen != d->lastpost &&
+	    (d->lastpost == NULL || d->lastpost < ss->lastseen)) {
+	d->lastpost = ss->lastseen;
+    }
+
+    /*
+     * Likewise for a no-progress state.
+     */
+
+    if ((ss->flags&NOPROGRESS) && ss->lastseen != d->lastnopr &&
+	    (d->lastnopr == NULL || d->lastnopr < ss->lastseen)) {
+	d->lastnopr = ss->lastseen;
+    }
+
+    return ss;
+}
+
+/*
+ - pickNextSS - pick the next stateset to be used
+ ^ static struct sset *pickNextSS(struct vars *, struct dfa *, chr *, chr *);
+ */
+static struct sset *
+pickNextSS(
+    struct vars *const v,	/* used only for debug flags */
+    struct dfa *const d,
+    chr *const cp,
+    chr *const start)
+{
+    int i;
+    struct sset *ss, *end;
+    chr *ancient;
+
+    /*
+     * Shortcut for cases where cache isn't full.
+     */
+
+    if (d->nssused < d->nssets) {
+	i = d->nssused;
+	d->nssused++;
+	ss = &d->ssets[i];
+	FDEBUG(("new c%d\n", i));
+
+	/*
+	 * Set up innards.
+	 */
+
+	ss->states = &d->statesarea[i * d->wordsper];
+	ss->flags = 0;
+	ss->ins.ss = NULL;
+	ss->ins.co = WHITE;	/* give it some value */
+	ss->outs = &d->outsarea[i * d->ncolors];
+	ss->inchain = &d->incarea[i * d->ncolors];
+	for (i = 0; i < d->ncolors; i++) {
+	    ss->outs[i] = NULL;
+	    ss->inchain[i].ss = NULL;
+	}
+	return ss;
+    }
+
+    /*
+     * Look for oldest, or old enough anyway.
+     */
+
+    if (cp - start > d->nssets*2/3) {	/* oldest 33% are expendable */
+	ancient = cp - d->nssets*2/3;
+    } else {
+	ancient = start;
+    }
+    for (ss = d->search, end = &d->ssets[d->nssets]; ss < end; ss++) {
+	if ((ss->lastseen == NULL || ss->lastseen < ancient)
+		&& !(ss->flags&LOCKED)) {
+	    d->search = ss + 1;
+	    FDEBUG(("replacing c%d\n", (int) (ss - d->ssets)));
+	    return ss;
+	}
+    }
+    for (ss = d->ssets, end = d->search; ss < end; ss++) {
+	if ((ss->lastseen == NULL || ss->lastseen < ancient)
+		&& !(ss->flags&LOCKED)) {
+	    d->search = ss + 1;
+	    FDEBUG(("replacing c%d\n", (int) (ss - d->ssets)));
+	    return ss;
+	}
+    }
+
+    /*
+     * Nobody's old enough?!? -- something's really wrong.
+     */
+
+    FDEBUG(("can't find victim to replace!\n"));
+    assert(NOTREACHED);
+    ERR(REG_ASSERT);
+    return d->ssets;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regerror.c b/generic/regerror.c
new file mode 100644
index 0000000..49d93ed
--- /dev/null
+++ b/generic/regerror.c
@@ -0,0 +1,129 @@
+/*
+ * regerror - error-code expansion
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#include "regguts.h"
+
+/*
+ * Unknown-error explanation.
+ */
+
+static const char unk[] = "*** unknown regex error code 0x%x ***";
+
+/*
+ * Struct to map among codes, code names, and explanations.
+ */
+
+static const struct rerr {
+    int code;
+    const char *name;
+    const char *explain;
+} rerrs[] = {
+    /* The actual table is built from regex.h */
+#include "regerrs.h"
+    { -1, "", "oops" },		/* explanation special-cased in code */
+};
+
+/*
+ - regerror - the interface to error numbers
+ */
+/* ARGSUSED */
+size_t				/* Actual space needed (including NUL) */
+regerror(
+    int code,			/* Error code, or REG_ATOI or REG_ITOA */
+    const regex_t *preg,	/* Associated regex_t (unused at present) */
+    char *errbuf,		/* Result buffer (unless errbuf_size==0) */
+    size_t errbuf_size)		/* Available space in errbuf, can be 0 */
+{
+    const struct rerr *r;
+    const char *msg;
+    char convbuf[sizeof(unk)+50]; /* 50 = plenty for int */
+    size_t len;
+    int icode;
+
+    switch (code) {
+    case REG_ATOI:		/* Convert name to number */
+	for (r = rerrs; r->code >= 0; r++) {
+	    if (strcmp(r->name, errbuf) == 0) {
+		break;
+	    }
+	}
+	sprintf(convbuf, "%d", r->code); /* -1 for unknown */
+	msg = convbuf;
+	break;
+    case REG_ITOA:		/* Convert number to name */
+	icode = atoi(errbuf);	/* Not our problem if this fails */
+	for (r = rerrs; r->code >= 0; r++) {
+	    if (r->code == icode) {
+		break;
+	    }
+	}
+	if (r->code >= 0) {
+	    msg = r->name;
+	} else {		/* Unknown; tell him the number */
+	    sprintf(convbuf, "REG_%u", (unsigned)icode);
+	    msg = convbuf;
+	}
+	break;
+    default:			/* A real, normal error code */
+	for (r = rerrs; r->code >= 0; r++) {
+	    if (r->code == code) {
+		break;
+	    }
+	}
+	if (r->code >= 0) {
+	    msg = r->explain;
+	} else {		/* Unknown; say so */
+	    sprintf(convbuf, unk, code);
+	    msg = convbuf;
+	}
+	break;
+    }
+
+    len = strlen(msg) + 1;	/* Space needed, including NUL */
+    if (errbuf_size > 0) {
+	if (errbuf_size > len) {
+	    strcpy(errbuf, msg);
+	} else {		/* Truncate to fit */
+	    strncpy(errbuf, msg, errbuf_size-1);
+	    errbuf[errbuf_size-1] = '\0';
+	}
+    }
+
+    return len;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regerrs.h b/generic/regerrs.h
new file mode 100644
index 0000000..ee203d5
--- /dev/null
+++ b/generic/regerrs.h
@@ -0,0 +1,20 @@
+{ REG_OKAY,	"REG_OKAY",	"no errors detected" },
+{ REG_NOMATCH,	"REG_NOMATCH",	"failed to match" },
+{ REG_BADPAT,	"REG_BADPAT",	"invalid regexp (reg version 0.8)" },
+{ REG_ECOLLATE,	"REG_ECOLLATE",	"invalid collating element" },
+{ REG_ECTYPE,	"REG_ECTYPE",	"invalid character class" },
+{ REG_EESCAPE,	"REG_EESCAPE",	"invalid escape \\ sequence" },
+{ REG_ESUBREG,	"REG_ESUBREG",	"invalid backreference number" },
+{ REG_EBRACK,	"REG_EBRACK",	"brackets [] not balanced" },
+{ REG_EPAREN,	"REG_EPAREN",	"parentheses () not balanced" },
+{ REG_EBRACE,	"REG_EBRACE",	"braces {} not balanced" },
+{ REG_BADBR,	"REG_BADBR",	"invalid repetition count(s)" },
+{ REG_ERANGE,	"REG_ERANGE",	"invalid character range" },
+{ REG_ESPACE,	"REG_ESPACE",	"out of memory" },
+{ REG_BADRPT,	"REG_BADRPT",	"quantifier operand invalid" },
+{ REG_ASSERT,	"REG_ASSERT",	"\"can't happen\" -- you found a bug" },
+{ REG_INVARG,	"REG_INVARG",	"invalid argument to regex function" },
+{ REG_MIXED,	"REG_MIXED",	"character widths of regex and string differ" },
+{ REG_BADOPT,	"REG_BADOPT",	"invalid embedded option" },
+{ REG_ETOOBIG,	"REG_ETOOBIG",	"regular expression is too complex" },
+{ REG_ECOLORS,	"REG_ECOLORS",	"too many colors" },
diff --git a/generic/regex.h b/generic/regex.h
new file mode 100644
index 0000000..8845f72
--- /dev/null
+++ b/generic/regex.h
@@ -0,0 +1,305 @@
+#ifndef _REGEX_H_
+#define	_REGEX_H_	/* never again */
+
+#include "tclInt.h"
+
+/*
+ * regular expressions
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ *
+ * Prototypes etc. marked with "^" within comments get gathered up (and
+ * possibly edited) by the regfwd program and inserted near the bottom of this
+ * file.
+ *
+ * We offer the option of declaring one wide-character version of the RE
+ * functions as well as the char versions. To do that, define __REG_WIDE_T to
+ * the type of wide characters (unfortunately, there is no consensus that
+ * wchar_t is suitable) and __REG_WIDE_COMPILE and __REG_WIDE_EXEC to the
+ * names to be used for the compile and execute functions (suggestion:
+ * re_Xcomp and re_Xexec, where X is a letter suggestive of the wide type,
+ * e.g. re_ucomp and re_uexec for Unicode). For cranky old compilers, it may
+ * be necessary to do something like:
+ * #define	__REG_WIDE_COMPILE(a,b,c,d)	re_Xcomp(a,b,c,d)
+ * #define	__REG_WIDE_EXEC(a,b,c,d,e,f,g)	re_Xexec(a,b,c,d,e,f,g)
+ * rather than just #defining the names as parameterless macros.
+ *
+ * For some specialized purposes, it may be desirable to suppress the
+ * declarations of the "front end" functions, regcomp() and regexec(), or of
+ * the char versions of the compile and execute functions. To suppress the
+ * front-end functions, define __REG_NOFRONT. To suppress the char versions,
+ * define __REG_NOCHAR.
+ *
+ * The right place to do those defines (and some others you may want, see
+ * below) would be <sys/types.h>. If you don't have control of that file, the
+ * right place to add your own defines to this file is marked below. This is
+ * normally done automatically, by the makefile and regmkhdr, based on the
+ * contents of regcustom.h.
+ */
+
+/*
+ * voodoo for C++
+ */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Add your own defines, if needed, here.
+ */
+
+/*
+ * Location where a chunk of regcustom.h is automatically spliced into this
+ * file (working from its prototype, regproto.h).
+ */
+
+/* --- begin --- */
+/* ensure certain things don't sneak in from system headers */
+#ifdef __REG_WIDE_T
+#undef __REG_WIDE_T
+#endif
+#ifdef __REG_WIDE_COMPILE
+#undef __REG_WIDE_COMPILE
+#endif
+#ifdef __REG_WIDE_EXEC
+#undef __REG_WIDE_EXEC
+#endif
+#ifdef __REG_REGOFF_T
+#undef __REG_REGOFF_T
+#endif
+#ifdef __REG_NOFRONT
+#undef __REG_NOFRONT
+#endif
+#ifdef __REG_NOCHAR
+#undef __REG_NOCHAR
+#endif
+/* interface types */
+#define	__REG_WIDE_T	Tcl_UniChar
+#define	__REG_REGOFF_T	long	/* not really right, but good enough... */
+/* names and declarations */
+#define	__REG_WIDE_COMPILE	TclReComp
+#define	__REG_WIDE_EXEC		TclReExec
+#define	__REG_NOFRONT		/* don't want regcomp() and regexec() */
+#define	__REG_NOCHAR		/* or the char versions */
+#define	regfree		TclReFree
+#define	regerror	TclReError
+/* --- end --- */
+
+/*
+ * interface types etc.
+ */
+
+/*
+ * regoff_t has to be large enough to hold either off_t or ssize_t, and must
+ * be signed; it's only a guess that long is suitable, so we offer
+ * <sys/types.h> an override.
+ */
+#ifdef __REG_REGOFF_T
+typedef __REG_REGOFF_T regoff_t;
+#else
+typedef long regoff_t;
+#endif
+
+/*
+ * other interface types
+ */
+
+/* the biggie, a compiled RE (or rather, a front end to same) */
+typedef struct {
+    int re_magic;		/* magic number */
+    size_t re_nsub;		/* number of subexpressions */
+    long re_info;		/* information about RE */
+#define	REG_UBACKREF		000001
+#define	REG_ULOOKAHEAD		000002
+#define	REG_UBOUNDS		000004
+#define	REG_UBRACES		000010
+#define	REG_UBSALNUM		000020
+#define	REG_UPBOTCH		000040
+#define	REG_UBBS		000100
+#define	REG_UNONPOSIX		000200
+#define	REG_UUNSPEC		000400
+#define	REG_UUNPORT		001000
+#define	REG_ULOCALE		002000
+#define	REG_UEMPTYMATCH		004000
+#define	REG_UIMPOSSIBLE		010000
+#define	REG_USHORTEST		020000
+    int re_csize;		/* sizeof(character) */
+    char *re_endp;		/* backward compatibility kludge */
+    /* the rest is opaque pointers to hidden innards */
+    char *re_guts;		/* `char *' is more portable than `void *' */
+    char *re_fns;
+} regex_t;
+
+/* result reporting (may acquire more fields later) */
+typedef struct {
+    regoff_t rm_so;		/* start of substring */
+    regoff_t rm_eo;		/* end of substring */
+} regmatch_t;
+
+/* supplementary control and reporting */
+typedef struct {
+    regmatch_t rm_extend;	/* see REG_EXPECT */
+} rm_detail_t;
+
+/*
+ * compilation
+ ^ #ifndef __REG_NOCHAR
+ ^ int re_comp(regex_t *, const char *, size_t, int);
+ ^ #endif
+ ^ #ifndef __REG_NOFRONT
+ ^ int regcomp(regex_t *, const char *, int);
+ ^ #endif
+ ^ #ifdef __REG_WIDE_T
+ ^ int __REG_WIDE_COMPILE(regex_t *, const __REG_WIDE_T *, size_t, int);
+ ^ #endif
+ */
+#define	REG_BASIC	000000	/* BREs (convenience) */
+#define	REG_EXTENDED	000001	/* EREs */
+#define	REG_ADVF	000002	/* advanced features in EREs */
+#define	REG_ADVANCED	000003	/* AREs (which are also EREs) */
+#define	REG_QUOTE	000004	/* no special characters, none */
+#define	REG_NOSPEC	REG_QUOTE	/* historical synonym */
+#define	REG_ICASE	000010	/* ignore case */
+#define	REG_NOSUB	000020	/* don't care about subexpressions */
+#define	REG_EXPANDED	000040	/* expanded format, white space & comments */
+#define	REG_NLSTOP	000100	/* \n doesn't match . or [^ ] */
+#define	REG_NLANCH	000200	/* ^ matches after \n, $ before */
+#define	REG_NEWLINE	000300	/* newlines are line terminators */
+#define	REG_PEND	000400	/* ugh -- backward-compatibility hack */
+#define	REG_EXPECT	001000	/* report details on partial/limited matches */
+#define	REG_BOSONLY	002000	/* temporary kludge for BOS-only matches */
+#define	REG_DUMP	004000	/* none of your business :-) */
+#define	REG_FAKE	010000	/* none of your business :-) */
+#define	REG_PROGRESS	020000	/* none of your business :-) */
+
+/*
+ * execution
+ ^ #ifndef __REG_NOCHAR
+ ^ int re_exec(regex_t *, const char *, size_t,
+ ^				rm_detail_t *, size_t, regmatch_t [], int);
+ ^ #endif
+ ^ #ifndef __REG_NOFRONT
+ ^ int regexec(regex_t *, const char *, size_t, regmatch_t [], int);
+ ^ #endif
+ ^ #ifdef __REG_WIDE_T
+ ^ int __REG_WIDE_EXEC(regex_t *, const __REG_WIDE_T *, size_t,
+ ^				rm_detail_t *, size_t, regmatch_t [], int);
+ ^ #endif
+ */
+#define	REG_NOTBOL	0001	/* BOS is not BOL */
+#define	REG_NOTEOL	0002	/* EOS is not EOL */
+#define	REG_STARTEND	0004	/* backward compatibility kludge */
+#define	REG_FTRACE	0010	/* none of your business */
+#define	REG_MTRACE	0020	/* none of your business */
+#define	REG_SMALL	0040	/* none of your business */
+
+/*
+ * misc generics (may be more functions here eventually)
+ ^ void regfree(regex_t *);
+ */
+
+/*
+ * error reporting
+ * Be careful if modifying the list of error codes -- the table used by
+ * regerror() is generated automatically from this file!
+ *
+ * Note that there is no wide-char variant of regerror at this time; what kind
+ * of character is used for error reports is independent of what kind is used
+ * in matching.
+ *
+ ^ extern size_t regerror(int, const regex_t *, char *, size_t);
+ */
+#define	REG_OKAY	 0	/* no errors detected */
+#define	REG_NOMATCH	 1	/* failed to match */
+#define	REG_BADPAT	 2	/* invalid regexp */
+#define	REG_ECOLLATE	 3	/* invalid collating element */
+#define	REG_ECTYPE	 4	/* invalid character class */
+#define	REG_EESCAPE	 5	/* invalid escape \ sequence */
+#define	REG_ESUBREG	 6	/* invalid backreference number */
+#define	REG_EBRACK	 7	/* brackets [] not balanced */
+#define	REG_EPAREN	 8	/* parentheses () not balanced */
+#define	REG_EBRACE	 9	/* braces {} not balanced */
+#define	REG_BADBR	10	/* invalid repetition count(s) */
+#define	REG_ERANGE	11	/* invalid character range */
+#define	REG_ESPACE	12	/* out of memory */
+#define	REG_BADRPT	13	/* quantifier operand invalid */
+#define	REG_ASSERT	15	/* "can't happen" -- you found a bug */
+#define	REG_INVARG	16	/* invalid argument to regex function */
+#define	REG_MIXED	17	/* character widths of regex and string differ */
+#define	REG_BADOPT	18	/* invalid embedded option */
+#define	REG_ETOOBIG	19	/* regular expression is too complex */
+#define	REG_ECOLORS	20	/* too many colors */
+/* two specials for debugging and testing */
+#define	REG_ATOI	101	/* convert error-code name to number */
+#define	REG_ITOA	102	/* convert error-code number to name */
+
+/*
+ * the prototypes, as possibly munched by regfwd
+ */
+/* =====^!^===== begin forwards =====^!^===== */
+/* automatically gathered by fwd; do not hand-edit */
+/* === regproto.h === */
+#ifndef __REG_NOCHAR
+int re_comp(regex_t *, const char *, size_t, int);
+#endif
+#ifndef __REG_NOFRONT
+int regcomp(regex_t *, const char *, int);
+#endif
+#ifdef __REG_WIDE_T
+MODULE_SCOPE int __REG_WIDE_COMPILE(regex_t *, const __REG_WIDE_T *, size_t, int);
+#endif
+#ifndef __REG_NOCHAR
+int re_exec(regex_t *, const char *, size_t, rm_detail_t *, size_t, regmatch_t [], int);
+#endif
+#ifndef __REG_NOFRONT
+int regexec(regex_t *, const char *, size_t, regmatch_t [], int);
+#endif
+#ifdef __REG_WIDE_T
+MODULE_SCOPE int __REG_WIDE_EXEC(regex_t *, const __REG_WIDE_T *, size_t, rm_detail_t *, size_t, regmatch_t [], int);
+#endif
+MODULE_SCOPE void regfree(regex_t *);
+MODULE_SCOPE size_t regerror(int, const regex_t *, char *, size_t);
+/* automatically gathered by fwd; do not hand-edit */
+/* =====^!^===== end forwards =====^!^===== */
+
+/*
+ * more C++ voodoo
+ */
+#ifdef __cplusplus
+}
+#endif
+
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regexec.c b/generic/regexec.c
new file mode 100644
index 0000000..128d439
--- /dev/null
+++ b/generic/regexec.c
@@ -0,0 +1,1335 @@
+/*
+ * re_*exec and friends - match REs
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "regguts.h"
+
+/*
+ * Lazy-DFA representation.
+ */
+
+struct arcp {			/* "pointer" to an outarc */
+    struct sset *ss;
+    color co;
+};
+
+struct sset {			/* state set */
+    unsigned *states;		/* pointer to bitvector */
+    unsigned hash;		/* hash of bitvector */
+#define	HASH(bv, nw)	(((nw) == 1) ? *(bv) : hash(bv, nw))
+#define	HIT(h,bv,ss,nw)	((ss)->hash == (h) && ((nw) == 1 || \
+	memcmp((void*)(bv), (void*)((ss)->states), (nw)*sizeof(unsigned)) == 0))
+    int flags;
+#define	STARTER		01	/* the initial state set */
+#define	POSTSTATE	02	/* includes the goal state */
+#define	LOCKED		04	/* locked in cache */
+#define	NOPROGRESS	010	/* zero-progress state set */
+    struct arcp ins;		/* chain of inarcs pointing here */
+    chr *lastseen;		/* last entered on arrival here */
+    struct sset **outs;		/* outarc vector indexed by color */
+    struct arcp *inchain;	/* chain-pointer vector for outarcs */
+};
+
+struct dfa {
+    int nssets;			/* size of cache */
+    int nssused;		/* how many entries occupied yet */
+    int nstates;		/* number of states */
+    int ncolors;		/* length of outarc and inchain vectors */
+    int wordsper;		/* length of state-set bitvectors */
+    struct sset *ssets;		/* state-set cache */
+    unsigned *statesarea;	/* bitvector storage */
+    unsigned *work;		/* pointer to work area within statesarea */
+    struct sset **outsarea;	/* outarc-vector storage */
+    struct arcp *incarea;	/* inchain storage */
+    struct cnfa *cnfa;
+    struct colormap *cm;
+    chr *lastpost;		/* location of last cache-flushed success */
+    chr *lastnopr;		/* location of last cache-flushed NOPROGRESS */
+    struct sset *search;	/* replacement-search-pointer memory */
+    int cptsmalloced;		/* were the areas individually malloced? */
+    char *mallocarea;		/* self, or master malloced area, or NULL */
+};
+
+#define	WORK	1		/* number of work bitvectors needed */
+
+/*
+ * Setup for non-malloc allocation for small cases.
+ */
+
+#define	FEWSTATES	20	/* must be less than UBITS */
+#define	FEWCOLORS	15
+struct smalldfa {
+    struct dfa dfa;
+    struct sset ssets[FEWSTATES*2];
+    unsigned statesarea[FEWSTATES*2 + WORK];
+    struct sset *outsarea[FEWSTATES*2 * FEWCOLORS];
+    struct arcp incarea[FEWSTATES*2 * FEWCOLORS];
+};
+#define	DOMALLOC	((struct smalldfa *)NULL)	/* force malloc */
+
+/*
+ * Internal variables, bundled for easy passing around.
+ */
+
+struct vars {
+    regex_t *re;
+    struct guts *g;
+    int eflags;			/* copies of arguments */
+    size_t nmatch;
+    regmatch_t *pmatch;
+    rm_detail_t *details;
+    chr *start;			/* start of string */
+    chr *stop;			/* just past end of string */
+    int err;			/* error code if any (0 none) */
+    struct dfa **subdfas;	/* per-subre DFAs */
+    struct smalldfa dfa1;
+    struct smalldfa dfa2;
+};
+#define	VISERR(vv) ((vv)->err != 0)	/* have we seen an error yet? */
+#define	ISERR()	VISERR(v)
+#define VERR(vv,e) ((vv)->err = ((vv)->err ? (vv)->err : (e)))
+#define	ERR(e)	VERR(v, e)	/* record an error */
+#define	NOERR()	{if (ISERR()) return v->err;}	/* if error seen, return it */
+#define	OFF(p)	((p) - v->start)
+#define	LOFF(p)	((long)OFF(p))
+
+/*
+ * forward declarations
+ */
+/* =====^!^===== begin forwards =====^!^===== */
+/* automatically gathered by fwd; do not hand-edit */
+/* === regexec.c === */
+int exec(regex_t *, const chr *, size_t, rm_detail_t *, size_t, regmatch_t [], int);
+static struct dfa *getsubdfa(struct vars *, struct subre *);
+static int simpleFind(struct vars *const, struct cnfa *const, struct colormap *const);
+static int complicatedFind(struct vars *const, struct cnfa *const, struct colormap *const);
+static int complicatedFindLoop(struct vars *const, struct cnfa *const, struct colormap *const, struct dfa *const, struct dfa *const, chr **const);
+static void zapallsubs(regmatch_t *const, const size_t);
+static void zaptreesubs(struct vars *const, struct subre *const);
+static void subset(struct vars *const, struct subre *const, chr *const, chr *const);
+static int cdissect(struct vars *, struct subre *, chr *, chr *);
+static int ccondissect(struct vars *, struct subre *, chr *, chr *);
+static int crevcondissect(struct vars *, struct subre *, chr *, chr *);
+static int cbrdissect(struct vars *, struct subre *, chr *, chr *);
+static int caltdissect(struct vars *, struct subre *, chr *, chr *);
+static int citerdissect(struct vars *, struct subre *, chr *, chr *);
+static int creviterdissect(struct vars *, struct subre *, chr *, chr *);
+/* === rege_dfa.c === */
+static chr *longest(struct vars *const, struct dfa *const, chr *const, chr *const, int *const);
+static chr *shortest(struct vars *const, struct dfa *const, chr *const, chr *const, chr *const, chr **const, int *const);
+static chr *lastCold(struct vars *const, struct dfa *const);
+static struct dfa *newDFA(struct vars *const, struct cnfa *const, struct colormap *const, struct smalldfa *);
+static void freeDFA(struct dfa *const);
+static unsigned hash(unsigned *const, const int);
+static struct sset *initialize(struct vars *const, struct dfa *const, chr *const);
+static struct sset *miss(struct vars *const, struct dfa *const, struct sset *const, const pcolor, chr *const, chr *const);
+static int checkLAConstraint(struct vars *const, struct cnfa *const, chr *const, const pcolor);
+static struct sset *getVacantSS(struct vars *const, struct dfa *const, chr *const, chr *const);
+static struct sset *pickNextSS(struct vars *const, struct dfa *const, chr *const, chr *const);
+/* automatically gathered by fwd; do not hand-edit */
+/* =====^!^===== end forwards =====^!^===== */
+
+/*
+ - exec - match regular expression
+ ^ int exec(regex_t *, const chr *, size_t, rm_detail_t *,
+ ^					size_t, regmatch_t [], int);
+ */
+int
+exec(
+    regex_t *re,
+    const chr *string,
+    size_t len,
+    rm_detail_t *details,
+    size_t nmatch,
+    regmatch_t pmatch[],
+    int flags)
+{
+    AllocVars(v);
+    int st, backref;
+    size_t n;
+    size_t i;
+#define	LOCALMAT	20
+    regmatch_t mat[LOCALMAT];
+#define LOCALDFAS	40
+    struct dfa *subdfas[LOCALDFAS];
+
+    /*
+     * Sanity checks.
+     */
+
+    if (re == NULL || string == NULL || re->re_magic != REMAGIC) {
+	FreeVars(v);
+	return REG_INVARG;
+    }
+    if (re->re_csize != sizeof(chr)) {
+	FreeVars(v);
+	return REG_MIXED;
+    }
+
+    /*
+     * Setup.
+     */
+
+    v->re = re;
+    v->g = (struct guts *)re->re_guts;
+    if ((v->g->cflags&REG_EXPECT) && details == NULL) {
+	FreeVars(v);
+	return REG_INVARG;
+    }
+    if (v->g->info&REG_UIMPOSSIBLE) {
+	FreeVars(v);
+	return REG_NOMATCH;
+    }
+    backref = (v->g->info&REG_UBACKREF) ? 1 : 0;
+    v->eflags = flags;
+    if (v->g->cflags&REG_NOSUB) {
+	nmatch = 0;		/* override client */
+    }
+    v->nmatch = nmatch;
+    if (backref) {
+	/*
+	 * Need work area.
+	 */
+
+	if (v->g->nsub + 1 <= LOCALMAT) {
+	    v->pmatch = mat;
+	} else {
+	    v->pmatch = (regmatch_t *)
+		    MALLOC((v->g->nsub + 1) * sizeof(regmatch_t));
+	}
+	if (v->pmatch == NULL) {
+	    FreeVars(v);
+	    return REG_ESPACE;
+	}
+	v->nmatch = v->g->nsub + 1;
+    } else {
+	v->pmatch = pmatch;
+    }
+    v->details = details;
+    v->start = (chr *)string;
+    v->stop = (chr *)string + len;
+    v->err = 0;
+    assert(v->g->ntree >= 0);
+    n = (size_t) v->g->ntree;
+    if (n <= LOCALDFAS)
+	v->subdfas = subdfas;
+    else
+	v->subdfas = (struct dfa **) MALLOC(n * sizeof(struct dfa *));
+    if (v->subdfas == NULL) {
+	if (v->pmatch != pmatch && v->pmatch != mat)
+	    FREE(v->pmatch);
+	FreeVars(v);
+	return REG_ESPACE;
+    }
+    for (i = 0; i < n; i++)
+	v->subdfas[i] = NULL;
+
+    /*
+     * Do it.
+     */
+
+    assert(v->g->tree != NULL);
+    if (backref) {
+	st = complicatedFind(v, &v->g->tree->cnfa, &v->g->cmap);
+    } else {
+	st = simpleFind(v, &v->g->tree->cnfa, &v->g->cmap);
+    }
+
+    /*
+     * Copy (portion of) match vector over if necessary.
+     */
+
+    if (st == REG_OKAY && v->pmatch != pmatch && nmatch > 0) {
+	zapallsubs(pmatch, nmatch);
+	n = (nmatch < v->nmatch) ? nmatch : v->nmatch;
+	memcpy((void*)(pmatch), (void*)(v->pmatch), n*sizeof(regmatch_t));
+    }
+
+    /*
+     * Clean up.
+     */
+
+    if (v->pmatch != pmatch && v->pmatch != mat) {
+	FREE(v->pmatch);
+    }
+    n = (size_t) v->g->ntree;
+    for (i = 0; i < n; i++) {
+	if (v->subdfas[i] != NULL)
+	    freeDFA(v->subdfas[i]);
+    }
+    if (v->subdfas != subdfas)
+	FREE(v->subdfas);
+    FreeVars(v);
+    return st;
+}
+
+/*
+ - getsubdfa - create or re-fetch the DFA for a subre node
+ * We only need to create the DFA once per overall regex execution.
+ * The DFA will be freed by the cleanup step in exec().
+ */
+static struct dfa *
+getsubdfa(struct vars * v,
+	  struct subre * t)
+{
+    if (v->subdfas[t->id] == NULL) {
+	v->subdfas[t->id] = newDFA(v, &t->cnfa, &v->g->cmap, DOMALLOC);
+	if (ISERR())
+	    return NULL;
+    }
+    return v->subdfas[t->id];
+}
+
+/*
+ - simpleFind - find a match for the main NFA (no-complications case)
+ ^ static int simpleFind(struct vars *, struct cnfa *, struct colormap *);
+ */
+static int
+simpleFind(
+    struct vars *const v,
+    struct cnfa *const cnfa,
+    struct colormap *const cm)
+{
+    struct dfa *s, *d;
+    chr *begin, *end = NULL;
+    chr *cold;
+    chr *open, *close;		/* Open and close of range of possible
+				 * starts */
+    int hitend;
+    int shorter = (v->g->tree->flags&SHORTER) ? 1 : 0;
+
+    /*
+     * First, a shot with the search RE.
+     */
+
+    s = newDFA(v, &v->g->search, cm, &v->dfa1);
+    assert(!(ISERR() && s != NULL));
+    NOERR();
+    MDEBUG(("\nsearch at %ld\n", LOFF(v->start)));
+    cold = NULL;
+    close = shortest(v, s, v->start, v->start, v->stop, &cold, NULL);
+    freeDFA(s);
+    NOERR();
+    if (v->g->cflags&REG_EXPECT) {
+	assert(v->details != NULL);
+	if (cold != NULL) {
+	    v->details->rm_extend.rm_so = OFF(cold);
+	} else {
+	    v->details->rm_extend.rm_so = OFF(v->stop);
+	}
+	v->details->rm_extend.rm_eo = OFF(v->stop);	/* unknown */
+    }
+    if (close == NULL) {	/* not found */
+	return REG_NOMATCH;
+    }
+    if (v->nmatch == 0) {	/* found, don't need exact location */
+	return REG_OKAY;
+    }
+
+    /*
+     * Find starting point and match.
+     */
+
+    assert(cold != NULL);
+    open = cold;
+    cold = NULL;
+    MDEBUG(("between %ld and %ld\n", LOFF(open), LOFF(close)));
+    d = newDFA(v, cnfa, cm, &v->dfa1);
+    assert(!(ISERR() && d != NULL));
+    NOERR();
+    for (begin = open; begin <= close; begin++) {
+	MDEBUG(("\nfind trying at %ld\n", LOFF(begin)));
+	if (shorter) {
+	    end = shortest(v, d, begin, begin, v->stop, NULL, &hitend);
+	} else {
+	    end = longest(v, d, begin, v->stop, &hitend);
+	}
+	if (ISERR()) {
+	    freeDFA(d);
+	    return v->err;
+	}
+	if (hitend && cold == NULL) {
+	    cold = begin;
+	}
+	if (end != NULL) {
+	    break;		/* NOTE BREAK OUT */
+	}
+    }
+    assert(end != NULL);	/* search RE succeeded so loop should */
+    freeDFA(d);
+
+    /*
+     * And pin down details.
+     */
+
+    assert(v->nmatch > 0);
+    v->pmatch[0].rm_so = OFF(begin);
+    v->pmatch[0].rm_eo = OFF(end);
+    if (v->g->cflags&REG_EXPECT) {
+	if (cold != NULL) {
+	    v->details->rm_extend.rm_so = OFF(cold);
+	} else {
+	    v->details->rm_extend.rm_so = OFF(v->stop);
+	}
+	v->details->rm_extend.rm_eo = OFF(v->stop);	/* unknown */
+    }
+    if (v->nmatch == 1) {	/* no need for submatches */
+	return REG_OKAY;
+    }
+
+    /*
+     * Find submatches.
+     */
+
+    zapallsubs(v->pmatch, v->nmatch);
+    return cdissect(v, v->g->tree, begin, end);
+}
+
+/*
+ - complicatedFind - find a match for the main NFA (with complications)
+ ^ static int complicatedFind(struct vars *, struct cnfa *, struct colormap *);
+ */
+static int
+complicatedFind(
+    struct vars *const v,
+    struct cnfa *const cnfa,
+    struct colormap *const cm)
+{
+    struct dfa *s, *d;
+    chr *cold = NULL; /* silence gcc 4 warning */
+    int ret;
+
+    s = newDFA(v, &v->g->search, cm, &v->dfa1);
+    NOERR();
+    d = newDFA(v, cnfa, cm, &v->dfa2);
+    if (ISERR()) {
+	assert(d == NULL);
+	freeDFA(s);
+	return v->err;
+    }
+
+    ret = complicatedFindLoop(v, cnfa, cm, d, s, &cold);
+
+    freeDFA(d);
+    freeDFA(s);
+    NOERR();
+    if (v->g->cflags&REG_EXPECT) {
+	assert(v->details != NULL);
+	if (cold != NULL) {
+	    v->details->rm_extend.rm_so = OFF(cold);
+	} else {
+	    v->details->rm_extend.rm_so = OFF(v->stop);
+	}
+	v->details->rm_extend.rm_eo = OFF(v->stop);	/* unknown */
+    }
+    return ret;
+}
+
+/*
+ - complicatedFindLoop - the heart of complicatedFind
+ ^ static int complicatedFindLoop(struct vars *, struct cnfa *, struct colormap *,
+ ^	struct dfa *, struct dfa *, chr **);
+ */
+static int
+complicatedFindLoop(
+    struct vars *const v,
+    struct cnfa *const cnfa,
+    struct colormap *const cm,
+    struct dfa *const d,
+    struct dfa *const s,
+    chr **const coldp)		/* where to put coldstart pointer */
+{
+    chr *begin, *end;
+    chr *cold;
+    chr *open, *close;		/* Open and close of range of possible
+				 * starts */
+    chr *estart, *estop;
+    int er, hitend;
+    int shorter = v->g->tree->flags&SHORTER;
+
+    assert(d != NULL && s != NULL);
+    cold = NULL;
+    close = v->start;
+    do {
+	MDEBUG(("\ncsearch at %ld\n", LOFF(close)));
+	close = shortest(v, s, close, close, v->stop, &cold, NULL);
+	if (close == NULL) {
+	    break;		/* NOTE BREAK */
+	}
+	assert(cold != NULL);
+	open = cold;
+	cold = NULL;
+	MDEBUG(("cbetween %ld and %ld\n", LOFF(open), LOFF(close)));
+	for (begin = open; begin <= close; begin++) {
+	    MDEBUG(("\ncomplicatedFind trying at %ld\n", LOFF(begin)));
+	    estart = begin;
+	    estop = v->stop;
+	    for (;;) {
+		if (shorter) {
+		    end = shortest(v, d, begin, estart, estop, NULL, &hitend);
+		} else {
+		    end = longest(v, d, begin, estop, &hitend);
+		}
+		if (hitend && cold == NULL) {
+		    cold = begin;
+		}
+		if (end == NULL) {
+		    break;	/* NOTE BREAK OUT */
+		}
+
+		MDEBUG(("tentative end %ld\n", LOFF(end)));
+		zapallsubs(v->pmatch, v->nmatch);
+		er = cdissect(v, v->g->tree, begin, end);
+		if (er == REG_OKAY) {
+		    if (v->nmatch > 0) {
+			v->pmatch[0].rm_so = OFF(begin);
+			v->pmatch[0].rm_eo = OFF(end);
+		    }
+		    *coldp = cold;
+		    return REG_OKAY;
+		}
+		if (er != REG_NOMATCH) {
+		    ERR(er);
+		    *coldp = cold;
+		    return er;
+		}
+		if ((shorter) ? end == estop : end == begin) {
+		    break;
+		}
+
+		/*
+		 * Go around and try again
+		 */
+
+		if (shorter) {
+		    estart = end + 1;
+		} else {
+		    estop = end - 1;
+		}
+	    }
+	}
+    } while (close < v->stop);
+
+    *coldp = cold;
+    return REG_NOMATCH;
+}
+
+/*
+ - zapallsubs - initialize all subexpression matches to "no match"
+ ^ static void zapallsubs(regmatch_t *, size_t);
+ */
+static void
+zapallsubs(
+    regmatch_t *const p,
+    const size_t n)
+{
+    size_t i;
+
+    for (i = n-1; i > 0; i--) {
+	p[i].rm_so = -1;
+	p[i].rm_eo = -1;
+    }
+}
+
+/*
+ - zaptreesubs - initialize subexpressions within subtree to "no match"
+ ^ static void zaptreesubs(struct vars *, struct subre *);
+ */
+static void
+zaptreesubs(
+    struct vars *const v,
+    struct subre *const t)
+{
+    if (t->op == '(') {
+	int n = t->subno;
+	assert(n > 0);
+	if ((size_t) n < v->nmatch) {
+	    v->pmatch[n].rm_so = -1;
+	    v->pmatch[n].rm_eo = -1;
+	}
+    }
+
+    if (t->left != NULL) {
+	zaptreesubs(v, t->left);
+    }
+    if (t->right != NULL) {
+	zaptreesubs(v, t->right);
+    }
+}
+
+/*
+ - subset - set subexpression match data for a successful subre
+ ^ static void subset(struct vars *, struct subre *, chr *, chr *);
+ */
+static void
+subset(
+    struct vars *const v,
+    struct subre *const sub,
+    chr *const begin,
+    chr *const end)
+{
+    int n = sub->subno;
+
+    assert(n > 0);
+    if ((size_t)n >= v->nmatch) {
+	return;
+    }
+
+    MDEBUG(("setting %d\n", n));
+    v->pmatch[n].rm_so = OFF(begin);
+    v->pmatch[n].rm_eo = OFF(end);
+}
+
+/*
+ - cdissect - check backrefs and determine subexpression matches
+ * cdissect recursively processes a subre tree to check matching of backrefs
+ * and/or identify submatch boundaries for capture nodes.  The proposed match
+ * runs from "begin" to "end" (not including "end"), and we are basically
+ * "dissecting" it to see where the submatches are.
+ * Before calling any level of cdissect, the caller must have run the node's
+ * DFA and found that the proposed substring satisfies the DFA.  (We make
+ * the caller do that because in concatenation and iteration nodes, it's
+ * much faster to check all the substrings against the child DFAs before we
+ * recurse.)  Also, caller must have cleared subexpression match data via
+ * zaptreesubs (or zapallsubs at the top level).
+ ^ static int cdissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+cdissect(
+    struct vars *v,
+    struct subre *t,
+    chr *begin,		/* beginning of relevant substring */
+    chr *end)		/* end of same */
+{
+    int er;
+
+    assert(t != NULL);
+    MDEBUG(("cdissect %ld-%ld %c\n", LOFF(begin), LOFF(end), t->op));
+
+    switch (t->op) {
+    case '=':			/* terminal node */
+	assert(t->left == NULL && t->right == NULL);
+	er = REG_OKAY;		/* no action, parent did the work */
+	break;
+    case 'b':			/* back reference */
+	assert(t->left == NULL && t->right == NULL);
+	er = cbrdissect(v, t, begin, end);
+	break;
+    case '.':			/* concatenation */
+	assert(t->left != NULL && t->right != NULL);
+	if (t->left->flags & SHORTER) /* reverse scan */
+	    er = crevcondissect(v, t, begin, end);
+	else
+	    er = ccondissect(v, t, begin, end);
+	break;
+    case '|':			/* alternation */
+	assert(t->left != NULL);
+	er = caltdissect(v, t, begin, end);
+	break;
+    case '*':			/* iteration */
+	assert(t->left != NULL);
+	if (t->left->flags & SHORTER) /* reverse scan */
+	    er = creviterdissect(v, t, begin, end);
+	else
+	    er = citerdissect(v, t, begin, end);
+	break;
+    case '(':			/* capturing */
+	assert(t->left != NULL && t->right == NULL);
+	assert(t->subno > 0);
+	er = cdissect(v, t->left, begin, end);
+	if (er == REG_OKAY) {
+	    subset(v, t, begin, end);
+	}
+	break;
+    default:
+	er = REG_ASSERT;
+	break;
+    }
+
+    /*
+     * We should never have a match failure unless backrefs lurk below;
+     * otherwise, either caller failed to check the DFA, or there's some
+     * inconsistency between the DFA and the node's innards.
+     */
+    assert(er != REG_NOMATCH || (t->flags & BACKR));
+
+    return er;
+}
+
+/*
+ - ccondissect - dissect match for concatenation node
+ ^ static int ccondissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+ccondissect(
+    struct vars *v,
+    struct subre *t,
+    chr *begin,		/* beginning of relevant substring */
+    chr *end)		/* end of same */
+{
+    struct dfa *d, *d2;
+    chr *mid;
+
+    assert(t->op == '.');
+    assert(t->left != NULL && t->left->cnfa.nstates > 0);
+    assert(t->right != NULL && t->right->cnfa.nstates > 0);
+    assert(!(t->left->flags & SHORTER));
+
+    d = getsubdfa(v, t->left);
+    NOERR();
+    d2 = getsubdfa(v, t->right);
+    NOERR();
+
+    MDEBUG(("cConcat %d\n", t->id));
+
+    /*
+     * Pick a tentative midpoint.
+     */
+    mid = longest(v, d, begin, end, (int *) NULL);
+    if (mid == NULL) {
+	return REG_NOMATCH;
+    }
+    MDEBUG(("tentative midpoint %ld\n", LOFF(mid)));
+
+    /*
+     * Iterate until satisfaction or failure.
+     */
+
+    for (;;) {
+	/*
+	 * Try this midpoint on for size.
+	 */
+
+	if (longest(v, d2, mid, end, NULL) == end) {
+	    int er = cdissect(v, t->left, begin, mid);
+
+	    if (er == REG_OKAY) {
+		er = cdissect(v, t->right, mid, end);
+		if (er == REG_OKAY) {
+		    /*
+		     * Satisfaction.
+		     */
+
+		    MDEBUG(("successful\n"));
+		    return REG_OKAY;
+		}
+	    }
+	    if (er != REG_NOMATCH) {
+		return er;
+	    }
+	}
+
+	/*
+	 * That midpoint didn't work, find a new one.
+	 */
+
+	if (mid == begin) {
+	    /*
+	     * All possibilities exhausted.
+	     */
+
+	    MDEBUG(("%d no midpoint\n", t->id));
+	    return REG_NOMATCH;
+	}
+	mid = longest(v, d, begin, mid-1, NULL);
+	if (mid == NULL) {
+	    /*
+	     * Failed to find a new one.
+	     */
+
+	    MDEBUG(("%d failed midpoint\n", t->id));
+	    return REG_NOMATCH;
+	}
+	MDEBUG(("%d: new midpoint %ld\n", t->id, LOFF(mid)));
+	zaptreesubs(v, t->left);
+	zaptreesubs(v, t->right);
+    }
+}
+
+/*
+ - crevcondissect - dissect match for concatenation node, shortest-first
+ ^ static int crevcondissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+crevcondissect(
+    struct vars *v,
+    struct subre *t,
+    chr *begin,		/* beginning of relevant substring */
+    chr *end)		/* end of same */
+{
+    struct dfa *d, *d2;
+    chr *mid;
+
+    assert(t->op == '.');
+    assert(t->left != NULL && t->left->cnfa.nstates > 0);
+    assert(t->right != NULL && t->right->cnfa.nstates > 0);
+    assert(t->left->flags&SHORTER);
+
+    d = getsubdfa(v, t->left);
+    NOERR();
+    d2 = getsubdfa(v, t->right);
+    NOERR();
+
+    MDEBUG(("crevcon %d\n", t->id));
+
+    /*
+     * Pick a tentative midpoint.
+     */
+
+    mid = shortest(v, d, begin, begin, end, (chr **) NULL, (int *) NULL);
+    if (mid == NULL) {
+	return REG_NOMATCH;
+    }
+    MDEBUG(("tentative midpoint %ld\n", LOFF(mid)));
+
+    /*
+     * Iterate until satisfaction or failure.
+     */
+
+    for (;;) {
+	/*
+	 * Try this midpoint on for size.
+	 */
+
+	if (longest(v, d2, mid, end, NULL) == end) {
+	    int er = cdissect(v, t->left, begin, mid);
+
+	    if (er == REG_OKAY) {
+		er = cdissect(v, t->right, mid, end);
+		if (er == REG_OKAY) {
+		    /*
+		     * Satisfaction.
+		     */
+
+		    MDEBUG(("successful\n"));
+		    return REG_OKAY;
+		}
+	    }
+	    if (er != REG_NOMATCH) {
+		return er;
+	    }
+	}
+
+	/*
+	 * That midpoint didn't work, find a new one.
+	 */
+
+	if (mid == end) {
+	    /*
+	     * All possibilities exhausted.
+	     */
+
+	    MDEBUG(("%d no midpoint\n", t->id));
+	    return REG_NOMATCH;
+	}
+	mid = shortest(v, d, begin, mid+1, end, NULL, NULL);
+	if (mid == NULL) {
+	    /*
+	     * Failed to find a new one.
+	     */
+
+	    MDEBUG(("%d failed midpoint\n", t->id));
+	    return REG_NOMATCH;
+	}
+	MDEBUG(("%d: new midpoint %ld\n", t->id, LOFF(mid)));
+	zaptreesubs(v, t->left);
+	zaptreesubs(v, t->right);
+    }
+}
+
+/*
+ - cbrdissect - dissect match for backref node
+ ^ static int cbrdissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+cbrdissect(
+    struct vars *v,
+    struct subre *t,
+    chr *begin,		/* beginning of relevant substring */
+    chr *end)		/* end of same */
+{
+    int n = t->subno, min = t->min, max = t->max;
+    size_t numreps;
+    size_t tlen;
+    size_t brlen;
+    chr *brstring;
+    chr *p;
+
+    assert(t != NULL);
+    assert(t->op == 'b');
+    assert(n >= 0);
+    assert((size_t)n < v->nmatch);
+
+    MDEBUG(("cbackref n%d %d{%d-%d}\n", t->id, n, min, max));
+
+    /* get the backreferenced string */
+    if (v->pmatch[n].rm_so == -1) {
+	return REG_NOMATCH;
+    }
+    brstring = v->start + v->pmatch[n].rm_so;
+    brlen = v->pmatch[n].rm_eo - v->pmatch[n].rm_so;
+
+    /* special cases for zero-length strings */
+    if (brlen == 0) {
+	/*
+	 * matches only if target is zero length, but any number of
+	 * repetitions can be considered to be present
+	 */
+	if (begin == end && min <= max) {
+	    MDEBUG(("cbackref matched trivially\n"));
+	    return REG_OKAY;
+	}
+	return REG_NOMATCH;
+    }
+    if (begin == end) {
+	/* matches only if zero repetitions are okay */
+	if (min == 0) {
+	    MDEBUG(("cbackref matched trivially\n"));
+	    return REG_OKAY;
+	}
+	return REG_NOMATCH;
+    }
+
+    /*
+     * check target length to see if it could possibly be an allowed number of
+     * repetitions of brstring
+     */
+
+    assert(end > begin);
+    tlen = end - begin;
+    if (tlen % brlen != 0)
+	return REG_NOMATCH;
+    numreps = tlen / brlen;
+    if (numreps < (size_t)min || (numreps > (size_t)max && max != DUPINF))
+	return REG_NOMATCH;
+
+    /* okay, compare the actual string contents */
+    p = begin;
+    while (numreps-- > 0) {
+	if ((*v->g->compare) (brstring, p, brlen) != 0)
+	    return REG_NOMATCH;
+	p += brlen;
+    }
+
+    MDEBUG(("cbackref matched\n"));
+    return REG_OKAY;
+}
+
+/*
+ - caltdissect - dissect match for alternation node
+ ^ static int caltdissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+caltdissect(
+    struct vars *v,
+    struct subre *t,
+    chr *begin,		/* beginning of relevant substring */
+    chr *end)		/* end of same */
+{
+    struct dfa *d;
+    int er;
+
+    /* We loop, rather than tail-recurse, to handle a chain of alternatives */
+    while (t != NULL) {
+	assert(t->op == '|');
+	assert(t->left != NULL && t->left->cnfa.nstates > 0);
+
+	MDEBUG(("calt n%d\n", t->id));
+
+	d = getsubdfa(v, t->left);
+	NOERR();
+	if (longest(v, d, begin, end, (int *) NULL) == end) {
+	    MDEBUG(("calt matched\n"));
+	    er = cdissect(v, t->left, begin, end);
+	    if (er != REG_NOMATCH) {
+		return er;
+	    }
+	}
+
+	t = t->right;
+    }
+
+    return REG_NOMATCH;
+}
+
+/*
+ - citerdissect - dissect match for iteration node
+ ^ static int citerdissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+citerdissect(struct vars * v,
+	     struct subre * t,
+	     chr *begin,	/* beginning of relevant substring */
+	     chr *end)		/* end of same */
+{
+    struct dfa *d;
+    chr	  **endpts;
+    chr	   *limit;
+    int		min_matches;
+    size_t	max_matches;
+    int		nverified;
+    int		k;
+    int		i;
+    int		er;
+
+    assert(t->op == '*');
+    assert(t->left != NULL && t->left->cnfa.nstates > 0);
+    assert(!(t->left->flags & SHORTER));
+    assert(begin <= end);
+
+    /*
+     * If zero matches are allowed, and target string is empty, just declare
+     * victory.  OTOH, if target string isn't empty, zero matches can't work
+     * so we pretend the min is 1.
+     */
+    min_matches = t->min;
+    if (min_matches <= 0) {
+	if (begin == end)
+	    return REG_OKAY;
+	min_matches = 1;
+    }
+
+    /*
+     * We need workspace to track the endpoints of each sub-match.  Normally
+     * we consider only nonzero-length sub-matches, so there can be at most
+     * end-begin of them.  However, if min is larger than that, we will also
+     * consider zero-length sub-matches in order to find enough matches.
+     *
+     * For convenience, endpts[0] contains the "begin" pointer and we store
+     * sub-match endpoints in endpts[1..max_matches].
+     */
+    max_matches = end - begin;
+    if (max_matches > (size_t)t->max && t->max != DUPINF)
+	max_matches = t->max;
+    if (max_matches < (size_t)min_matches)
+	max_matches = min_matches;
+    endpts = (chr **) MALLOC((max_matches + 1) * sizeof(chr *));
+    if (endpts == NULL)
+	return REG_ESPACE;
+    endpts[0] = begin;
+
+    d = getsubdfa(v, t->left);
+    if (ISERR()) {
+	FREE(endpts);
+	return v->err;
+    }
+    MDEBUG(("citer %d\n", t->id));
+
+    /*
+     * Our strategy is to first find a set of sub-match endpoints that are
+     * valid according to the child node's DFA, and then recursively dissect
+     * each sub-match to confirm validity.  If any validity check fails,
+     * backtrack the last sub-match and try again.  And, when we next try for
+     * a validity check, we need not recheck any successfully verified
+     * sub-matches that we didn't move the endpoints of.  nverified remembers
+     * how many sub-matches are currently known okay.
+     */
+
+    /* initialize to consider first sub-match */
+    nverified = 0;
+    k = 1;
+    limit = end;
+
+    /* iterate until satisfaction or failure */
+    while (k > 0) {
+	/* try to find an endpoint for the k'th sub-match */
+	endpts[k] = longest(v, d, endpts[k - 1], limit, (int *) NULL);
+	if (endpts[k] == NULL) {
+	    /* no match possible, so see if we can shorten previous one */
+	    k--;
+	    goto backtrack;
+	}
+	MDEBUG(("%d: working endpoint %d: %ld\n",
+		t->id, k, LOFF(endpts[k])));
+
+	/* k'th sub-match can no longer be considered verified */
+	if (nverified >= k)
+	    nverified = k - 1;
+
+	if (endpts[k] != end) {
+	    /* haven't reached end yet, try another iteration if allowed */
+	    if ((size_t)k >= max_matches) {
+		/* must try to shorten some previous match */
+		k--;
+		goto backtrack;
+	    }
+
+	    /* reject zero-length match unless necessary to achieve min */
+	    if (endpts[k] == endpts[k - 1] &&
+		(k >= min_matches || min_matches - k < end - endpts[k]))
+		goto backtrack;
+
+	    k++;
+	    limit = end;
+	    continue;
+	}
+
+	/*
+	 * We've identified a way to divide the string into k sub-matches
+	 * that works so far as the child DFA can tell.  If k is an allowed
+	 * number of matches, start the slow part: recurse to verify each
+	 * sub-match.  We always have k <= max_matches, needn't check that.
+	 */
+	if (k < min_matches)
+	    goto backtrack;
+
+	MDEBUG(("%d: verifying %d..%d\n", t->id, nverified + 1, k));
+
+	for (i = nverified + 1; i <= k; i++) {
+	    zaptreesubs(v, t->left);
+	    er = cdissect(v, t->left, endpts[i - 1], endpts[i]);
+	    if (er == REG_OKAY) {
+		nverified = i;
+		continue;
+	    }
+	    if (er == REG_NOMATCH)
+		break;
+	    /* oops, something failed */
+	    FREE(endpts);
+	    return er;
+	}
+
+	if (i > k) {
+	    /* satisfaction */
+	    MDEBUG(("%d successful\n", t->id));
+	    FREE(endpts);
+	    return REG_OKAY;
+	}
+
+	/* match failed to verify, so backtrack */
+
+    backtrack:
+	/*
+	 * Must consider shorter versions of the current sub-match.  However,
+	 * we'll only ask for a zero-length match if necessary.
+	 */
+	while (k > 0) {
+	    chr	   *prev_end = endpts[k - 1];
+
+	    if (endpts[k] > prev_end) {
+		limit = endpts[k] - 1;
+		if (limit > prev_end ||
+		    (k < min_matches && min_matches - k >= end - prev_end)) {
+		    /* break out of backtrack loop, continue the outer one */
+		    break;
+		}
+	    }
+	    /* can't shorten k'th sub-match any more, consider previous one */
+	    k--;
+	}
+    }
+
+    /* all possibilities exhausted */
+    MDEBUG(("%d failed\n", t->id));
+    FREE(endpts);
+    return REG_NOMATCH;
+}
+
+/*
+ - creviterdissect - dissect match for iteration node, shortest-first
+ ^ static int creviterdissect(struct vars *, struct subre *, chr *, chr *);
+ */
+static int			/* regexec return code */
+creviterdissect(struct vars * v,
+		struct subre * t,
+		chr *begin,	/* beginning of relevant substring */
+		chr *end)	/* end of same */
+{
+    struct dfa *d;
+    chr	  **endpts;
+    chr	   *limit;
+    int		min_matches;
+    size_t	max_matches;
+    int		nverified;
+    int		k;
+    int		i;
+    int		er;
+
+    assert(t->op == '*');
+    assert(t->left != NULL && t->left->cnfa.nstates > 0);
+    assert(t->left->flags & SHORTER);
+    assert(begin <= end);
+
+    /*
+     * If zero matches are allowed, and target string is empty, just declare
+     * victory.  OTOH, if target string isn't empty, zero matches can't work
+     * so we pretend the min is 1.
+     */
+    min_matches = t->min;
+    if (min_matches <= 0) {
+	if (begin == end)
+	    return REG_OKAY;
+	min_matches = 1;
+    }
+
+    /*
+     * We need workspace to track the endpoints of each sub-match.  Normally
+     * we consider only nonzero-length sub-matches, so there can be at most
+     * end-begin of them.  However, if min is larger than that, we will also
+     * consider zero-length sub-matches in order to find enough matches.
+     *
+     * For convenience, endpts[0] contains the "begin" pointer and we store
+     * sub-match endpoints in endpts[1..max_matches].
+     */
+    max_matches = end - begin;
+    if (max_matches > (size_t)t->max && t->max != DUPINF)
+	max_matches = t->max;
+    if (max_matches < (size_t)min_matches)
+	max_matches = min_matches;
+    endpts = (chr **) MALLOC((max_matches + 1) * sizeof(chr *));
+    if (endpts == NULL)
+	return REG_ESPACE;
+    endpts[0] = begin;
+
+    d = getsubdfa(v, t->left);
+    if (ISERR()) {
+	FREE(endpts);
+	return v->err;
+    }
+    MDEBUG(("creviter %d\n", t->id));
+
+    /*
+     * Our strategy is to first find a set of sub-match endpoints that are
+     * valid according to the child node's DFA, and then recursively dissect
+     * each sub-match to confirm validity.  If any validity check fails,
+     * backtrack the last sub-match and try again.  And, when we next try for
+     * a validity check, we need not recheck any successfully verified
+     * sub-matches that we didn't move the endpoints of.  nverified remembers
+     * how many sub-matches are currently known okay.
+     */
+
+    /* initialize to consider first sub-match */
+    nverified = 0;
+    k = 1;
+    limit = begin;
+
+    /* iterate until satisfaction or failure */
+    while (k > 0) {
+	/* disallow zero-length match unless necessary to achieve min */
+	if (limit == endpts[k - 1] &&
+	    limit != end &&
+	    (k >= min_matches || min_matches - k < end - limit))
+	    limit++;
+
+	/* if this is the last allowed sub-match, it must reach to the end */
+	if ((size_t)k >= max_matches)
+	    limit = end;
+
+	/* try to find an endpoint for the k'th sub-match */
+	endpts[k] = shortest(v, d, endpts[k - 1], limit, end,
+			     (chr **) NULL, (int *) NULL);
+	if (endpts[k] == NULL) {
+	    /* no match possible, so see if we can lengthen previous one */
+	    k--;
+	    goto backtrack;
+	}
+	MDEBUG(("%d: working endpoint %d: %ld\n",
+		t->id, k, LOFF(endpts[k])));
+
+	/* k'th sub-match can no longer be considered verified */
+	if (nverified >= k)
+	    nverified = k - 1;
+
+	if (endpts[k] != end) {
+	    /* haven't reached end yet, try another iteration if allowed */
+	    if ((size_t)k >= max_matches) {
+		/* must try to lengthen some previous match */
+		k--;
+		goto backtrack;
+	    }
+
+	    k++;
+	    limit = endpts[k - 1];
+	    continue;
+	}
+
+	/*
+	 * We've identified a way to divide the string into k sub-matches
+	 * that works so far as the child DFA can tell.  If k is an allowed
+	 * number of matches, start the slow part: recurse to verify each
+	 * sub-match.  We always have k <= max_matches, needn't check that.
+	 */
+	if (k < min_matches)
+	    goto backtrack;
+
+	MDEBUG(("%d: verifying %d..%d\n", t->id, nverified + 1, k));
+
+	for (i = nverified + 1; i <= k; i++) {
+	    zaptreesubs(v, t->left);
+	    er = cdissect(v, t->left, endpts[i - 1], endpts[i]);
+	    if (er == REG_OKAY) {
+		nverified = i;
+		continue;
+	    }
+	    if (er == REG_NOMATCH)
+		break;
+	    /* oops, something failed */
+	    FREE(endpts);
+	    return er;
+	}
+
+	if (i > k) {
+	    /* satisfaction */
+	    MDEBUG(("%d successful\n", t->id));
+	    FREE(endpts);
+	    return REG_OKAY;
+	}
+
+	/* match failed to verify, so backtrack */
+
+    backtrack:
+	/*
+	 * Must consider longer versions of the current sub-match.
+	 */
+	while (k > 0) {
+	    if (endpts[k] < end) {
+		limit = endpts[k] + 1;
+		/* break out of backtrack loop, continue the outer one */
+		break;
+	    }
+	    /* can't lengthen k'th sub-match any more, consider previous one */
+	    k--;
+	}
+    }
+
+    /* all possibilities exhausted */
+    MDEBUG(("%d failed\n", t->id));
+    FREE(endpts);
+    return REG_NOMATCH;
+}
+
+#include "rege_dfa.c"
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regfree.c b/generic/regfree.c
new file mode 100644
index 0000000..b0aaa70
--- /dev/null
+++ b/generic/regfree.c
@@ -0,0 +1,60 @@
+/*
+ * regfree - free an RE
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * You might think that this could be incorporated into regcomp.c, and that
+ * would be a reasonable idea... except that this is a generic function (with
+ * a generic name), applicable to all compiled REs regardless of the size of
+ * their characters, whereas the stuff in regcomp.c gets compiled once per
+ * character size.
+ */
+
+#include "regguts.h"
+
+/*
+ - regfree - free an RE (generic function, punts to RE-specific function)
+ *
+ * Ignoring invocation with NULL is a convenience.
+ */
+void
+regfree(
+    regex_t *re)
+{
+    if (re == NULL) {
+	return;
+    }
+    (*((struct fns *)re->re_fns)->free)(re);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regfronts.c b/generic/regfronts.c
new file mode 100644
index 0000000..088a640
--- /dev/null
+++ b/generic/regfronts.c
@@ -0,0 +1,91 @@
+/*
+ * regcomp and regexec - front ends to re_ routines
+ *
+ * Mostly for implementation of backward-compatibility kludges. Note that
+ * these routines exist ONLY in char versions.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "regguts.h"
+
+/*
+ - regcomp - compile regular expression
+ */
+int
+regcomp(
+    regex_t *re,
+    const char *str,
+    int flags)
+{
+    size_t len;
+    int f = flags;
+
+    if (f&REG_PEND) {
+	len = re->re_endp - str;
+	f &= ~REG_PEND;
+    } else {
+	len = strlen(str);
+    }
+
+    return re_comp(re, str, len, f);
+}
+
+/*
+ - regexec - execute regular expression
+ */
+int
+regexec(
+    regex_t *re,
+    const char *str,
+    size_t nmatch,
+    regmatch_t pmatch[],
+    int flags)
+{
+    const char *start;
+    size_t len;
+    int f = flags;
+
+    if (f & REG_STARTEND) {
+	start = str + pmatch[0].rm_so;
+	len = pmatch[0].rm_eo - pmatch[0].rm_so;
+	f &= ~REG_STARTEND;
+    } else {
+	start = str;
+	len = strlen(str);
+    }
+
+    return re_exec(re, start, len, nmatch, pmatch, f);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/regguts.h b/generic/regguts.h
new file mode 100644
index 0000000..ad9d5b9
--- /dev/null
+++ b/generic/regguts.h
@@ -0,0 +1,427 @@
+/*
+ * Internal interface definitions, etc., for the reg package
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * Environmental customization. It should not (I hope) be necessary to alter
+ * the file you are now reading -- regcustom.h should handle it all, given
+ * care here and elsewhere.
+ */
+#include "regcustom.h"
+
+/*
+ * Things that regcustom.h might override.
+ */
+
+/* assertions */
+#ifndef assert
+#ifndef REG_DEBUG
+#ifndef NDEBUG
+#define	NDEBUG		/* no assertions */
+#endif
+#endif /* !REG_DEBUG */
+#include <assert.h>
+#endif
+
+/* memory allocation */
+#ifndef MALLOC
+#define	MALLOC(n)	malloc(n)
+#endif
+#ifndef REALLOC
+#define	REALLOC(p, n)	realloc(p, n)
+#endif
+#ifndef FREE
+#define	FREE(p)		free(p)
+#endif
+
+/* want size of a char in bits, and max value in bounded quantifiers */
+#ifndef _POSIX2_RE_DUP_MAX
+#define	_POSIX2_RE_DUP_MAX 255	/* normally from <limits.h> */
+#endif
+
+/*
+ * misc
+ */
+
+#define	NOTREACHED	0
+#define	xxx		1
+
+#define	DUPMAX	_POSIX2_RE_DUP_MAX
+#define	DUPINF	(DUPMAX+1)
+
+#define	REMAGIC	0xfed7		/* magic number for main struct */
+
+/*
+ * debugging facilities
+ */
+#ifdef REG_DEBUG
+/* FDEBUG does finite-state tracing */
+#define	FDEBUG(arglist)	{ if (v->eflags&REG_FTRACE) printf arglist; }
+/* MDEBUG does higher-level tracing */
+#define	MDEBUG(arglist)	{ if (v->eflags&REG_MTRACE) printf arglist; }
+#else
+#define	FDEBUG(arglist)	{}
+#define	MDEBUG(arglist)	{}
+#endif
+
+/*
+ * bitmap manipulation
+ */
+#define	UBITS	(CHAR_BIT * sizeof(unsigned))
+#define	BSET(uv, sn)	((uv)[(sn)/UBITS] |= (unsigned)1 << ((sn)%UBITS))
+#define	ISBSET(uv, sn)	((uv)[(sn)/UBITS] & ((unsigned)1 << ((sn)%UBITS)))
+
+/*
+ * We dissect a chr into byts for colormap table indexing. Here we define a
+ * byt, which will be the same as a byte on most machines... The exact size of
+ * a byt is not critical, but about 8 bits is good, and extraction of 8-bit
+ * chunks is sometimes especially fast.
+ */
+
+#ifndef BYTBITS
+#define	BYTBITS	8		/* bits in a byt */
+#endif
+#define	BYTTAB	(1<<BYTBITS)	/* size of table with one entry per byt value */
+#define	BYTMASK	(BYTTAB-1)	/* bit mask for byt */
+#define	NBYTS	((CHRBITS+BYTBITS-1)/BYTBITS)
+/* the definition of GETCOLOR(), below, assumes NBYTS <= 4 */
+
+/*
+ * As soon as possible, we map chrs into equivalence classes -- "colors" --
+ * which are of much more manageable number.
+ */
+
+typedef short color;		/* colors of characters */
+typedef int pcolor;		/* what color promotes to */
+#define MAX_COLOR	SHRT_MAX /* max color value */
+#define	COLORLESS	(-1)	/* impossible color */
+#define	WHITE		0	/* default color, parent of all others */
+
+/*
+ * A colormap is a tree -- more precisely, a DAG -- indexed at each level by a
+ * byt of the chr, to map the chr to a color efficiently. Because lower
+ * sections of the tree can be shared, it can exploit the usual sparseness of
+ * such a mapping table. The tree is always NBYTS levels deep (in the past it
+ * was shallower during construction but was "filled" to full depth at the end
+ * of that); areas that are unaltered as yet point to "fill blocks" which are
+ * entirely WHITE in color.
+ */
+
+/* the tree itself */
+struct colors {
+    color ccolor[BYTTAB];
+};
+struct ptrs {
+    union tree *pptr[BYTTAB];
+};
+union tree {
+    struct colors colors;
+    struct ptrs ptrs;
+};
+#define	tcolor	colors.ccolor
+#define	tptr	ptrs.pptr
+
+/* Internal per-color descriptor structure for the color machinery */
+struct colordesc {
+    uchr nchrs;			/* number of chars of this color */
+    color sub;			/* open subcolor (if any); free chain ptr */
+#define	NOSUB	COLORLESS
+    struct arc *arcs;		/* color chain */
+    int flags;
+#define	FREECOL	01		/* currently free */
+#define	PSEUDO	02		/* pseudocolor, no real chars */
+#define	UNUSEDCOLOR(cd)	((cd)->flags&FREECOL)
+    union tree *block;		/* block of solid color, if any */
+};
+
+/*
+ * The color map itself
+ *
+ * Much of the data in the colormap struct is only used at compile time.
+ * However, the bulk of the space usage is in the "tree" structure, so it's
+ * not clear that there's much point in converting the rest to a more compact
+ * form when compilation is finished.
+ */
+struct colormap {
+    int magic;
+#define	CMMAGIC	0x876
+    struct vars *v;		/* for compile error reporting */
+    size_t ncds;		/* number of colordescs */
+    size_t max;			/* highest in use */
+    color free;			/* beginning of free chain (if non-0) */
+    struct colordesc *cd;
+#define	CDEND(cm)	(&(cm)->cd[(cm)->max + 1])
+#define	NINLINECDS	((size_t)10)
+    struct colordesc cdspace[NINLINECDS];
+    union tree tree[NBYTS];	/* tree top, plus fill blocks */
+};
+
+/* optimization magic to do fast chr->color mapping */
+#define	B0(c)	((c) & BYTMASK)
+#define	B1(c)	(((c)>>BYTBITS) & BYTMASK)
+#define	B2(c)	(((c)>>(2*BYTBITS)) & BYTMASK)
+#define	B3(c)	(((c)>>(3*BYTBITS)) & BYTMASK)
+#if NBYTS == 1
+#define	GETCOLOR(cm, c)	((cm)->tree->tcolor[B0(c)])
+#endif
+/* beware, for NBYTS>1, GETCOLOR() is unsafe -- 2nd arg used repeatedly */
+#if NBYTS == 2
+#define	GETCOLOR(cm, c)	((cm)->tree->tptr[B1(c)]->tcolor[B0(c)])
+#endif
+#if NBYTS == 4
+#define	GETCOLOR(cm, c)	((cm)->tree->tptr[B3(c)]->tptr[B2(c)]->tptr[B1(c)]->tcolor[B0(c)])
+#endif
+
+/*
+ * Interface definitions for locale-interface functions in locale.c.
+ */
+
+/* Representation of a set of characters. */
+struct cvec {
+    int nchrs;			/* number of chrs */
+    int chrspace;		/* number of chrs possible */
+    chr *chrs;			/* pointer to vector of chrs */
+    int nranges;		/* number of ranges (chr pairs) */
+    int rangespace;		/* number of chrs possible */
+    chr *ranges;		/* pointer to vector of chr pairs */
+};
+
+/*
+ * definitions for non-deterministic finite autmaton (NFA) internal
+ * representation
+ *
+ * Having a "from" pointer within each arc may seem redundant, but it saves a
+ * lot of hassle.
+ */
+
+struct state;
+
+struct arc {
+    int type;			/* 0 if free, else an NFA arc type code */
+    color co;
+    struct state *from;		/* where it's from (and contained within) */
+    struct state *to;		/* where it's to */
+    struct arc *outchain;	/* link in *from's outs chain or free chain */
+    struct arc *outchainRev;	/* back-link in *from's outs chain */
+#define	freechain outchain	/* we do not maintain "freechainRev" */
+    struct arc *inchain;	/* *to's ins chain */
+    struct arc *inchainRev;	/* back-link in *to's ins chain */
+    struct arc *colorchain;	/* color's arc chain */
+    struct arc *colorchainRev;	/* back-link in color's arc chain */
+};
+
+struct arcbatch {		/* for bulk allocation of arcs */
+    struct arcbatch *next;
+#define	ABSIZE	10
+    struct arc a[ABSIZE];
+};
+
+struct state {
+    int no;
+#define	FREESTATE	(-1)
+    char flag;			/* marks special states */
+    int nins;			/* number of inarcs */
+    struct arc *ins;		/* chain of inarcs */
+    int nouts;			/* number of outarcs */
+    struct arc *outs;		/* chain of outarcs */
+    struct arc *free;		/* chain of free arcs */
+    struct state *tmp;		/* temporary for traversal algorithms */
+    struct state *next;		/* chain for traversing all */
+    struct state *prev;		/* back chain */
+    struct arcbatch oas;	/* first arcbatch, avoid malloc in easy case */
+    int noas;			/* number of arcs used in first arcbatch */
+};
+
+struct nfa {
+    struct state *pre;		/* pre-initial state */
+    struct state *init;		/* initial state */
+    struct state *final;	/* final state */
+    struct state *post;		/* post-final state */
+    int nstates;		/* for numbering states */
+    struct state *states;	/* state-chain header */
+    struct state *slast;	/* tail of the chain */
+    struct state *free;		/* free list */
+    struct colormap *cm;	/* the color map */
+    color bos[2];		/* colors, if any, assigned to BOS and BOL */
+    color eos[2];		/* colors, if any, assigned to EOS and EOL */
+    struct vars *v;		/* simplifies compile error reporting */
+    struct nfa *parent;		/* parent NFA, if any */
+};
+
+/*
+ * definitions for compacted NFA
+ *
+ * The main space savings in a compacted NFA is from making the arcs as small
+ * as possible.  We store only the transition color and next-state number for
+ * each arc.  The list of out arcs for each state is an array beginning at
+ * cnfa.states[statenumber], and terminated by a dummy carc struct with
+ * co == COLORLESS.
+ *
+ * The non-dummy carc structs are of two types: plain arcs and LACON arcs.
+ * Plain arcs just store the transition color number as "co".  LACON arcs
+ * store the lookahead constraint number plus cnfa.ncolors as "co".  LACON
+ * arcs can be distinguished from plain by testing for co >= cnfa.ncolors.
+ */
+
+struct carc {
+    color co;			/* COLORLESS is list terminator */
+    int to;			/* next-state number */
+};
+
+struct cnfa {
+    int nstates;		/* number of states */
+    int ncolors;		/* number of colors */
+    int flags;
+#define	HASLACONS	01	/* uses lookahead constraints */
+    int pre;			/* setup state number */
+    int post;			/* teardown state number */
+    color bos[2];		/* colors, if any, assigned to BOS and BOL */
+    color eos[2];		/* colors, if any, assigned to EOS and EOL */
+    char *stflags;		/* vector of per-state flags bytes */
+#define CNFA_NOPROGRESS	01	/* flag bit for a no-progress state */
+    struct carc **states;	/* vector of pointers to outarc lists */
+    /* states[n] are pointers into a single malloc'd array of arcs */
+    struct carc *arcs;		/* the area for the lists */
+};
+#define	ZAPCNFA(cnfa)	((cnfa).nstates = 0)
+#define	NULLCNFA(cnfa)	((cnfa).nstates == 0)
+
+/*
+ * This symbol limits the transient heap space used by the regex compiler,
+ * and thereby also the maximum complexity of NFAs that we'll deal with.
+ * Currently we only count NFA states and arcs against this; the other
+ * transient data is generally not large enough to notice compared to those.
+ * Note that we do not charge anything for the final output data structures
+ * (the compacted NFA and the colormap).
+ */
+#ifndef REG_MAX_COMPILE_SPACE
+#define REG_MAX_COMPILE_SPACE  \
+	(100000 * sizeof(struct state) + 100000 * sizeof(struct arcbatch))
+#endif
+
+/*
+ * subexpression tree
+ *
+ * "op" is one of:
+ *	'='  plain regex without interesting substructure (implemented as DFA)
+ *	'b'  back-reference (has no substructure either)
+ *	'('  capture node: captures the match of its single child
+ *	'.'  concatenation: matches a match for left, then a match for right
+ *	'|'  alternation: matches a match for left or a match for right
+ *	'*'  iteration: matches some number of matches of its single child
+ *
+ * Note: the right child of an alternation must be another alternation or
+ * NULL; hence, an N-way branch requires N alternation nodes, not N-1 as you
+ * might expect.  This could stand to be changed.  Actually I'd rather see
+ * a single alternation node with N children, but that will take revising
+ * the representation of struct subre.
+ *
+ * Note: when a backref is directly quantified, we stick the min/max counts
+ * into the backref rather than plastering an iteration node on top.  This is
+ * for efficiency: there is no need to search for possible division points.
+ */
+
+struct subre {
+    char op;			/* see type codes above */
+    char flags;
+#define	LONGER	01		/* prefers longer match */
+#define	SHORTER	02		/* prefers shorter match */
+#define	MIXED	04		/* mixed preference below */
+#define	CAP	010		/* capturing parens below */
+#define	BACKR	020		/* back reference below */
+#define	INUSE	0100		/* in use in final tree */
+#define	NOPROP	03		/* bits which may not propagate up */
+#define	LMIX(f)	((f)<<2)	/* LONGER -> MIXED */
+#define	SMIX(f)	((f)<<1)	/* SHORTER -> MIXED */
+#define	UP(f)	(((f)&~NOPROP) | (LMIX(f) & SMIX(f) & MIXED))
+#define	MESSY(f)	((f)&(MIXED|CAP|BACKR))
+#define	PREF(f)	((f)&NOPROP)
+#define	PREF2(f1, f2)	((PREF(f1) != 0) ? PREF(f1) : PREF(f2))
+#define	COMBINE(f1, f2)	(UP((f1)|(f2)) | PREF2(f1, f2))
+    short id;			/* ID of subre (1..ntree-1) */
+    int subno;			/* subexpression number (for 'b' and '(') */
+    short min;			/* min repetitions for iteration or backref */
+    short max;			/* max repetitions for iteration or backref */
+    struct subre *left;		/* left child, if any (also freelist chain) */
+    struct subre *right;	/* right child, if any */
+    struct state *begin;	/* outarcs from here... */
+    struct state *end;		/* ...ending in inarcs here */
+    struct cnfa cnfa;		/* compacted NFA, if any */
+    struct subre *chain;	/* for bookkeeping and error cleanup */
+};
+
+/*
+ * table of function pointers for generic manipulation functions. A regex_t's
+ * re_fns points to one of these.
+ */
+
+struct fns {
+    void (*free) (regex_t *);
+};
+
+/*
+ * the insides of a regex_t, hidden behind a void *
+ */
+
+struct guts {
+    int magic;
+#define	GUTSMAGIC	0xfed9
+    int cflags;			/* copy of compile flags */
+    long info;			/* copy of re_info */
+    size_t nsub;		/* copy of re_nsub */
+    struct subre *tree;
+    struct cnfa search;		/* for fast preliminary search */
+    int ntree;			/* number of subre's, plus one */
+    struct colormap cmap;
+    int (*compare) (const chr *, const chr *, size_t);
+    struct subre *lacons;	/* lookahead-constraint vector */
+    int nlacons;		/* size of lacons */
+};
+
+/*
+ * Magic for allocating a variable workspace. This default version is
+ * stack-hungry.
+ */
+
+#ifndef AllocVars
+#define AllocVars(vPtr) \
+    struct vars var; \
+    register struct vars *vPtr = &var
+#endif
+#ifndef FreeVars
+#define FreeVars(vPtr) ((void) 0)
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tcl.decls b/generic/tcl.decls
new file mode 100644
index 0000000..b2b91a9
--- /dev/null
+++ b/generic/tcl.decls
@@ -0,0 +1,2402 @@
+# tcl.decls --
+#
+#	This file contains the declarations for all supported public
+#	functions that are exported by the Tcl library via the stubs table.
+#	This file is used to generate the tclDecls.h, tclPlatDecls.h
+#	and tclStubInit.c files.
+#
+# Copyright (c) 1998-1999 by Scriptics Corporation.
+# Copyright (c) 2001, 2002 by Kevin B. Kenny.  All rights reserved.
+# Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+#
+# See the file "license.terms" for information on usage and redistribution
+# of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+
+library tcl
+
+# Define the tcl interface with several sub interfaces:
+#     tclPlat	 - platform specific public
+#     tclInt	 - generic private
+#     tclPlatInt - platform specific private
+
+interface tcl
+hooks {tclPlat tclInt tclIntPlat}
+scspec EXTERN
+
+# Declare each of the functions in the public Tcl interface.  Note that
+# the an index should never be reused for a different function in order
+# to preserve backwards compatibility.
+
+declare 0 {
+    int Tcl_PkgProvideEx(Tcl_Interp *interp, const char *name,
+	    const char *version, const void *clientData)
+}
+declare 1 {
+    CONST84_RETURN char *Tcl_PkgRequireEx(Tcl_Interp *interp,
+	    const char *name, const char *version, int exact,
+	    void *clientDataPtr)
+}
+declare 2 {
+    TCL_NORETURN void Tcl_Panic(const char *format, ...)
+}
+declare 3 {
+    char *Tcl_Alloc(unsigned int size)
+}
+declare 4 {
+    void Tcl_Free(char *ptr)
+}
+declare 5 {
+    char *Tcl_Realloc(char *ptr, unsigned int size)
+}
+declare 6 {
+    char *Tcl_DbCkalloc(unsigned int size, const char *file, int line)
+}
+declare 7 {
+    void Tcl_DbCkfree(char *ptr, const char *file, int line)
+}
+declare 8 {
+    char *Tcl_DbCkrealloc(char *ptr, unsigned int size,
+	    const char *file, int line)
+}
+
+# Tcl_CreateFileHandler and Tcl_DeleteFileHandler are only available on unix,
+# but they are part of the old generic interface, so we include them here for
+# compatibility reasons.
+
+declare 9 unix {
+    void Tcl_CreateFileHandler(int fd, int mask, Tcl_FileProc *proc,
+	    ClientData clientData)
+}
+declare 10 unix {
+    void Tcl_DeleteFileHandler(int fd)
+}
+declare 11 {
+    void Tcl_SetTimer(const Tcl_Time *timePtr)
+}
+declare 12 {
+    void Tcl_Sleep(int ms)
+}
+declare 13 {
+    int Tcl_WaitForEvent(const Tcl_Time *timePtr)
+}
+declare 14 {
+    int Tcl_AppendAllObjTypes(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 15 {
+    void Tcl_AppendStringsToObj(Tcl_Obj *objPtr, ...)
+}
+declare 16 {
+    void Tcl_AppendToObj(Tcl_Obj *objPtr, const char *bytes, int length)
+}
+declare 17 {
+    Tcl_Obj *Tcl_ConcatObj(int objc, Tcl_Obj *const objv[])
+}
+declare 18 {
+    int Tcl_ConvertToType(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    const Tcl_ObjType *typePtr)
+}
+declare 19 {
+    void Tcl_DbDecrRefCount(Tcl_Obj *objPtr, const char *file, int line)
+}
+declare 20 {
+    void Tcl_DbIncrRefCount(Tcl_Obj *objPtr, const char *file, int line)
+}
+declare 21 {
+    int Tcl_DbIsShared(Tcl_Obj *objPtr, const char *file, int line)
+}
+declare 22 {
+    Tcl_Obj *Tcl_DbNewBooleanObj(int boolValue, const char *file, int line)
+}
+declare 23 {
+    Tcl_Obj *Tcl_DbNewByteArrayObj(const unsigned char *bytes, int length,
+	    const char *file, int line)
+}
+declare 24 {
+    Tcl_Obj *Tcl_DbNewDoubleObj(double doubleValue, const char *file,
+	    int line)
+}
+declare 25 {
+    Tcl_Obj *Tcl_DbNewListObj(int objc, Tcl_Obj *const *objv,
+	    const char *file, int line)
+}
+declare 26 {
+    Tcl_Obj *Tcl_DbNewLongObj(long longValue, const char *file, int line)
+}
+declare 27 {
+    Tcl_Obj *Tcl_DbNewObj(const char *file, int line)
+}
+declare 28 {
+    Tcl_Obj *Tcl_DbNewStringObj(const char *bytes, int length,
+	    const char *file, int line)
+}
+declare 29 {
+    Tcl_Obj *Tcl_DuplicateObj(Tcl_Obj *objPtr)
+}
+declare 30 {
+    void TclFreeObj(Tcl_Obj *objPtr)
+}
+declare 31 {
+    int Tcl_GetBoolean(Tcl_Interp *interp, const char *src, int *boolPtr)
+}
+declare 32 {
+    int Tcl_GetBooleanFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    int *boolPtr)
+}
+declare 33 {
+    unsigned char *Tcl_GetByteArrayFromObj(Tcl_Obj *objPtr, int *lengthPtr)
+}
+declare 34 {
+    int Tcl_GetDouble(Tcl_Interp *interp, const char *src, double *doublePtr)
+}
+declare 35 {
+    int Tcl_GetDoubleFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    double *doublePtr)
+}
+declare 36 {
+    int Tcl_GetIndexFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    CONST84 char *const *tablePtr, const char *msg, int flags, int *indexPtr)
+}
+declare 37 {
+    int Tcl_GetInt(Tcl_Interp *interp, const char *src, int *intPtr)
+}
+declare 38 {
+    int Tcl_GetIntFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr, int *intPtr)
+}
+declare 39 {
+    int Tcl_GetLongFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr, long *longPtr)
+}
+declare 40 {
+    CONST86 Tcl_ObjType *Tcl_GetObjType(const char *typeName)
+}
+declare 41 {
+    char *Tcl_GetStringFromObj(Tcl_Obj *objPtr, int *lengthPtr)
+}
+declare 42 {
+    void Tcl_InvalidateStringRep(Tcl_Obj *objPtr)
+}
+declare 43 {
+    int Tcl_ListObjAppendList(Tcl_Interp *interp, Tcl_Obj *listPtr,
+	    Tcl_Obj *elemListPtr)
+}
+declare 44 {
+    int Tcl_ListObjAppendElement(Tcl_Interp *interp, Tcl_Obj *listPtr,
+	    Tcl_Obj *objPtr)
+}
+declare 45 {
+    int Tcl_ListObjGetElements(Tcl_Interp *interp, Tcl_Obj *listPtr,
+	    int *objcPtr, Tcl_Obj ***objvPtr)
+}
+declare 46 {
+    int Tcl_ListObjIndex(Tcl_Interp *interp, Tcl_Obj *listPtr, int index,
+	    Tcl_Obj **objPtrPtr)
+}
+declare 47 {
+    int Tcl_ListObjLength(Tcl_Interp *interp, Tcl_Obj *listPtr,
+	    int *lengthPtr)
+}
+declare 48 {
+    int Tcl_ListObjReplace(Tcl_Interp *interp, Tcl_Obj *listPtr, int first,
+	    int count, int objc, Tcl_Obj *const objv[])
+}
+declare 49 {
+    Tcl_Obj *Tcl_NewBooleanObj(int boolValue)
+}
+declare 50 {
+    Tcl_Obj *Tcl_NewByteArrayObj(const unsigned char *bytes, int length)
+}
+declare 51 {
+    Tcl_Obj *Tcl_NewDoubleObj(double doubleValue)
+}
+declare 52 {
+    Tcl_Obj *Tcl_NewIntObj(int intValue)
+}
+declare 53 {
+    Tcl_Obj *Tcl_NewListObj(int objc, Tcl_Obj *const objv[])
+}
+declare 54 {
+    Tcl_Obj *Tcl_NewLongObj(long longValue)
+}
+declare 55 {
+    Tcl_Obj *Tcl_NewObj(void)
+}
+declare 56 {
+    Tcl_Obj *Tcl_NewStringObj(const char *bytes, int length)
+}
+declare 57 {
+    void Tcl_SetBooleanObj(Tcl_Obj *objPtr, int boolValue)
+}
+declare 58 {
+    unsigned char *Tcl_SetByteArrayLength(Tcl_Obj *objPtr, int length)
+}
+declare 59 {
+    void Tcl_SetByteArrayObj(Tcl_Obj *objPtr, const unsigned char *bytes,
+	    int length)
+}
+declare 60 {
+    void Tcl_SetDoubleObj(Tcl_Obj *objPtr, double doubleValue)
+}
+declare 61 {
+    void Tcl_SetIntObj(Tcl_Obj *objPtr, int intValue)
+}
+declare 62 {
+    void Tcl_SetListObj(Tcl_Obj *objPtr, int objc, Tcl_Obj *const objv[])
+}
+declare 63 {
+    void Tcl_SetLongObj(Tcl_Obj *objPtr, long longValue)
+}
+declare 64 {
+    void Tcl_SetObjLength(Tcl_Obj *objPtr, int length)
+}
+declare 65 {
+    void Tcl_SetStringObj(Tcl_Obj *objPtr, const char *bytes, int length)
+}
+declare 66 {
+    void Tcl_AddErrorInfo(Tcl_Interp *interp, const char *message)
+}
+declare 67 {
+    void Tcl_AddObjErrorInfo(Tcl_Interp *interp, const char *message,
+	    int length)
+}
+declare 68 {
+    void Tcl_AllowExceptions(Tcl_Interp *interp)
+}
+declare 69 {
+    void Tcl_AppendElement(Tcl_Interp *interp, const char *element)
+}
+declare 70 {
+    void Tcl_AppendResult(Tcl_Interp *interp, ...)
+}
+declare 71 {
+    Tcl_AsyncHandler Tcl_AsyncCreate(Tcl_AsyncProc *proc,
+	    ClientData clientData)
+}
+declare 72 {
+    void Tcl_AsyncDelete(Tcl_AsyncHandler async)
+}
+declare 73 {
+    int Tcl_AsyncInvoke(Tcl_Interp *interp, int code)
+}
+declare 74 {
+    void Tcl_AsyncMark(Tcl_AsyncHandler async)
+}
+declare 75 {
+    int Tcl_AsyncReady(void)
+}
+declare 76 {
+    void Tcl_BackgroundError(Tcl_Interp *interp)
+}
+declare 77 {
+    char Tcl_Backslash(const char *src, int *readPtr)
+}
+declare 78 {
+    int Tcl_BadChannelOption(Tcl_Interp *interp, const char *optionName,
+	    const char *optionList)
+}
+declare 79 {
+    void Tcl_CallWhenDeleted(Tcl_Interp *interp, Tcl_InterpDeleteProc *proc,
+	    ClientData clientData)
+}
+declare 80 {
+    void Tcl_CancelIdleCall(Tcl_IdleProc *idleProc, ClientData clientData)
+}
+declare 81 {
+    int Tcl_Close(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 82 {
+    int Tcl_CommandComplete(const char *cmd)
+}
+declare 83 {
+    char *Tcl_Concat(int argc, CONST84 char *const *argv)
+}
+declare 84 {
+    int Tcl_ConvertElement(const char *src, char *dst, int flags)
+}
+declare 85 {
+    int Tcl_ConvertCountedElement(const char *src, int length, char *dst,
+	    int flags)
+}
+declare 86 {
+    int Tcl_CreateAlias(Tcl_Interp *slave, const char *slaveCmd,
+	    Tcl_Interp *target, const char *targetCmd, int argc,
+	    CONST84 char *const *argv)
+}
+declare 87 {
+    int Tcl_CreateAliasObj(Tcl_Interp *slave, const char *slaveCmd,
+	    Tcl_Interp *target, const char *targetCmd, int objc,
+	    Tcl_Obj *const objv[])
+}
+declare 88 {
+    Tcl_Channel Tcl_CreateChannel(const Tcl_ChannelType *typePtr,
+	    const char *chanName, ClientData instanceData, int mask)
+}
+declare 89 {
+    void Tcl_CreateChannelHandler(Tcl_Channel chan, int mask,
+	    Tcl_ChannelProc *proc, ClientData clientData)
+}
+declare 90 {
+    void Tcl_CreateCloseHandler(Tcl_Channel chan, Tcl_CloseProc *proc,
+	    ClientData clientData)
+}
+declare 91 {
+    Tcl_Command Tcl_CreateCommand(Tcl_Interp *interp, const char *cmdName,
+	    Tcl_CmdProc *proc, ClientData clientData,
+	    Tcl_CmdDeleteProc *deleteProc)
+}
+declare 92 {
+    void Tcl_CreateEventSource(Tcl_EventSetupProc *setupProc,
+	    Tcl_EventCheckProc *checkProc, ClientData clientData)
+}
+declare 93 {
+    void Tcl_CreateExitHandler(Tcl_ExitProc *proc, ClientData clientData)
+}
+declare 94 {
+    Tcl_Interp *Tcl_CreateInterp(void)
+}
+declare 95 {
+    void Tcl_CreateMathFunc(Tcl_Interp *interp, const char *name,
+	    int numArgs, Tcl_ValueType *argTypes,
+	    Tcl_MathProc *proc, ClientData clientData)
+}
+declare 96 {
+    Tcl_Command Tcl_CreateObjCommand(Tcl_Interp *interp,
+	    const char *cmdName,
+	    Tcl_ObjCmdProc *proc, ClientData clientData,
+	    Tcl_CmdDeleteProc *deleteProc)
+}
+declare 97 {
+    Tcl_Interp *Tcl_CreateSlave(Tcl_Interp *interp, const char *slaveName,
+	    int isSafe)
+}
+declare 98 {
+    Tcl_TimerToken Tcl_CreateTimerHandler(int milliseconds,
+	    Tcl_TimerProc *proc, ClientData clientData)
+}
+declare 99 {
+    Tcl_Trace Tcl_CreateTrace(Tcl_Interp *interp, int level,
+	    Tcl_CmdTraceProc *proc, ClientData clientData)
+}
+declare 100 {
+    void Tcl_DeleteAssocData(Tcl_Interp *interp, const char *name)
+}
+declare 101 {
+    void Tcl_DeleteChannelHandler(Tcl_Channel chan, Tcl_ChannelProc *proc,
+	    ClientData clientData)
+}
+declare 102 {
+    void Tcl_DeleteCloseHandler(Tcl_Channel chan, Tcl_CloseProc *proc,
+	    ClientData clientData)
+}
+declare 103 {
+    int Tcl_DeleteCommand(Tcl_Interp *interp, const char *cmdName)
+}
+declare 104 {
+    int Tcl_DeleteCommandFromToken(Tcl_Interp *interp, Tcl_Command command)
+}
+declare 105 {
+    void Tcl_DeleteEvents(Tcl_EventDeleteProc *proc, ClientData clientData)
+}
+declare 106 {
+    void Tcl_DeleteEventSource(Tcl_EventSetupProc *setupProc,
+	    Tcl_EventCheckProc *checkProc, ClientData clientData)
+}
+declare 107 {
+    void Tcl_DeleteExitHandler(Tcl_ExitProc *proc, ClientData clientData)
+}
+declare 108 {
+    void Tcl_DeleteHashEntry(Tcl_HashEntry *entryPtr)
+}
+declare 109 {
+    void Tcl_DeleteHashTable(Tcl_HashTable *tablePtr)
+}
+declare 110 {
+    void Tcl_DeleteInterp(Tcl_Interp *interp)
+}
+declare 111 {
+    void Tcl_DetachPids(int numPids, Tcl_Pid *pidPtr)
+}
+declare 112 {
+    void Tcl_DeleteTimerHandler(Tcl_TimerToken token)
+}
+declare 113 {
+    void Tcl_DeleteTrace(Tcl_Interp *interp, Tcl_Trace trace)
+}
+declare 114 {
+    void Tcl_DontCallWhenDeleted(Tcl_Interp *interp,
+	    Tcl_InterpDeleteProc *proc, ClientData clientData)
+}
+declare 115 {
+    int Tcl_DoOneEvent(int flags)
+}
+declare 116 {
+    void Tcl_DoWhenIdle(Tcl_IdleProc *proc, ClientData clientData)
+}
+declare 117 {
+    char *Tcl_DStringAppend(Tcl_DString *dsPtr, const char *bytes, int length)
+}
+declare 118 {
+    char *Tcl_DStringAppendElement(Tcl_DString *dsPtr, const char *element)
+}
+declare 119 {
+    void Tcl_DStringEndSublist(Tcl_DString *dsPtr)
+}
+declare 120 {
+    void Tcl_DStringFree(Tcl_DString *dsPtr)
+}
+declare 121 {
+    void Tcl_DStringGetResult(Tcl_Interp *interp, Tcl_DString *dsPtr)
+}
+declare 122 {
+    void Tcl_DStringInit(Tcl_DString *dsPtr)
+}
+declare 123 {
+    void Tcl_DStringResult(Tcl_Interp *interp, Tcl_DString *dsPtr)
+}
+declare 124 {
+    void Tcl_DStringSetLength(Tcl_DString *dsPtr, int length)
+}
+declare 125 {
+    void Tcl_DStringStartSublist(Tcl_DString *dsPtr)
+}
+declare 126 {
+    int Tcl_Eof(Tcl_Channel chan)
+}
+declare 127 {
+    CONST84_RETURN char *Tcl_ErrnoId(void)
+}
+declare 128 {
+    CONST84_RETURN char *Tcl_ErrnoMsg(int err)
+}
+declare 129 {
+    int Tcl_Eval(Tcl_Interp *interp, const char *script)
+}
+# This is obsolete, use Tcl_FSEvalFile
+declare 130 {
+    int Tcl_EvalFile(Tcl_Interp *interp, const char *fileName)
+}
+declare 131 {
+    int Tcl_EvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 132 {
+    void Tcl_EventuallyFree(ClientData clientData, Tcl_FreeProc *freeProc)
+}
+declare 133 {
+    TCL_NORETURN void Tcl_Exit(int status)
+}
+declare 134 {
+    int Tcl_ExposeCommand(Tcl_Interp *interp, const char *hiddenCmdToken,
+	    const char *cmdName)
+}
+declare 135 {
+    int Tcl_ExprBoolean(Tcl_Interp *interp, const char *expr, int *ptr)
+}
+declare 136 {
+    int Tcl_ExprBooleanObj(Tcl_Interp *interp, Tcl_Obj *objPtr, int *ptr)
+}
+declare 137 {
+    int Tcl_ExprDouble(Tcl_Interp *interp, const char *expr, double *ptr)
+}
+declare 138 {
+    int Tcl_ExprDoubleObj(Tcl_Interp *interp, Tcl_Obj *objPtr, double *ptr)
+}
+declare 139 {
+    int Tcl_ExprLong(Tcl_Interp *interp, const char *expr, long *ptr)
+}
+declare 140 {
+    int Tcl_ExprLongObj(Tcl_Interp *interp, Tcl_Obj *objPtr, long *ptr)
+}
+declare 141 {
+    int Tcl_ExprObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    Tcl_Obj **resultPtrPtr)
+}
+declare 142 {
+    int Tcl_ExprString(Tcl_Interp *interp, const char *expr)
+}
+declare 143 {
+    void Tcl_Finalize(void)
+}
+declare 144 {
+    void Tcl_FindExecutable(const char *argv0)
+}
+declare 145 {
+    Tcl_HashEntry *Tcl_FirstHashEntry(Tcl_HashTable *tablePtr,
+	    Tcl_HashSearch *searchPtr)
+}
+declare 146 {
+    int Tcl_Flush(Tcl_Channel chan)
+}
+declare 147 {
+    void Tcl_FreeResult(Tcl_Interp *interp)
+}
+declare 148 {
+    int Tcl_GetAlias(Tcl_Interp *interp, const char *slaveCmd,
+	    Tcl_Interp **targetInterpPtr, CONST84 char **targetCmdPtr,
+	    int *argcPtr, CONST84 char ***argvPtr)
+}
+declare 149 {
+    int Tcl_GetAliasObj(Tcl_Interp *interp, const char *slaveCmd,
+	    Tcl_Interp **targetInterpPtr, CONST84 char **targetCmdPtr,
+	    int *objcPtr, Tcl_Obj ***objv)
+}
+declare 150 {
+    ClientData Tcl_GetAssocData(Tcl_Interp *interp, const char *name,
+	    Tcl_InterpDeleteProc **procPtr)
+}
+declare 151 {
+    Tcl_Channel Tcl_GetChannel(Tcl_Interp *interp, const char *chanName,
+	    int *modePtr)
+}
+declare 152 {
+    int Tcl_GetChannelBufferSize(Tcl_Channel chan)
+}
+declare 153 {
+    int Tcl_GetChannelHandle(Tcl_Channel chan, int direction,
+	    ClientData *handlePtr)
+}
+declare 154 {
+    ClientData Tcl_GetChannelInstanceData(Tcl_Channel chan)
+}
+declare 155 {
+    int Tcl_GetChannelMode(Tcl_Channel chan)
+}
+declare 156 {
+    CONST84_RETURN char *Tcl_GetChannelName(Tcl_Channel chan)
+}
+declare 157 {
+    int Tcl_GetChannelOption(Tcl_Interp *interp, Tcl_Channel chan,
+	    const char *optionName, Tcl_DString *dsPtr)
+}
+declare 158 {
+    CONST86 Tcl_ChannelType *Tcl_GetChannelType(Tcl_Channel chan)
+}
+declare 159 {
+    int Tcl_GetCommandInfo(Tcl_Interp *interp, const char *cmdName,
+	    Tcl_CmdInfo *infoPtr)
+}
+declare 160 {
+    CONST84_RETURN char *Tcl_GetCommandName(Tcl_Interp *interp,
+	    Tcl_Command command)
+}
+declare 161 {
+    int Tcl_GetErrno(void)
+}
+declare 162 {
+    CONST84_RETURN char *Tcl_GetHostName(void)
+}
+declare 163 {
+    int Tcl_GetInterpPath(Tcl_Interp *askInterp, Tcl_Interp *slaveInterp)
+}
+declare 164 {
+    Tcl_Interp *Tcl_GetMaster(Tcl_Interp *interp)
+}
+declare 165 {
+    const char *Tcl_GetNameOfExecutable(void)
+}
+declare 166 {
+    Tcl_Obj *Tcl_GetObjResult(Tcl_Interp *interp)
+}
+
+# Tcl_GetOpenFile is only available on unix, but it is a part of the old
+# generic interface, so we inlcude it here for compatibility reasons.
+
+declare 167 unix {
+    int Tcl_GetOpenFile(Tcl_Interp *interp, const char *chanID, int forWriting,
+	    int checkUsage, ClientData *filePtr)
+}
+# Obsolete.  Should now use Tcl_FSGetPathType which is objectified
+# and therefore usually faster.
+declare 168 {
+    Tcl_PathType Tcl_GetPathType(const char *path)
+}
+declare 169 {
+    int Tcl_Gets(Tcl_Channel chan, Tcl_DString *dsPtr)
+}
+declare 170 {
+    int Tcl_GetsObj(Tcl_Channel chan, Tcl_Obj *objPtr)
+}
+declare 171 {
+    int Tcl_GetServiceMode(void)
+}
+declare 172 {
+    Tcl_Interp *Tcl_GetSlave(Tcl_Interp *interp, const char *slaveName)
+}
+declare 173 {
+    Tcl_Channel Tcl_GetStdChannel(int type)
+}
+declare 174 {
+    CONST84_RETURN char *Tcl_GetStringResult(Tcl_Interp *interp)
+}
+declare 175 {
+    CONST84_RETURN char *Tcl_GetVar(Tcl_Interp *interp, const char *varName,
+	    int flags)
+}
+declare 176 {
+    CONST84_RETURN char *Tcl_GetVar2(Tcl_Interp *interp, const char *part1,
+	    const char *part2, int flags)
+}
+declare 177 {
+    int Tcl_GlobalEval(Tcl_Interp *interp, const char *command)
+}
+declare 178 {
+    int Tcl_GlobalEvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 179 {
+    int Tcl_HideCommand(Tcl_Interp *interp, const char *cmdName,
+	    const char *hiddenCmdToken)
+}
+declare 180 {
+    int Tcl_Init(Tcl_Interp *interp)
+}
+declare 181 {
+    void Tcl_InitHashTable(Tcl_HashTable *tablePtr, int keyType)
+}
+declare 182 {
+    int Tcl_InputBlocked(Tcl_Channel chan)
+}
+declare 183 {
+    int Tcl_InputBuffered(Tcl_Channel chan)
+}
+declare 184 {
+    int Tcl_InterpDeleted(Tcl_Interp *interp)
+}
+declare 185 {
+    int Tcl_IsSafe(Tcl_Interp *interp)
+}
+# Obsolete, use Tcl_FSJoinPath
+declare 186 {
+    char *Tcl_JoinPath(int argc, CONST84 char *const *argv,
+	    Tcl_DString *resultPtr)
+}
+declare 187 {
+    int Tcl_LinkVar(Tcl_Interp *interp, const char *varName, char *addr,
+	    int type)
+}
+
+# This slot is reserved for use by the plus patch:
+#  declare 188 {
+#	Tcl_MainLoop
+#  }
+
+declare 189 {
+    Tcl_Channel Tcl_MakeFileChannel(ClientData handle, int mode)
+}
+declare 190 {
+    int Tcl_MakeSafe(Tcl_Interp *interp)
+}
+declare 191 {
+    Tcl_Channel Tcl_MakeTcpClientChannel(ClientData tcpSocket)
+}
+declare 192 {
+    char *Tcl_Merge(int argc, CONST84 char *const *argv)
+}
+declare 193 {
+    Tcl_HashEntry *Tcl_NextHashEntry(Tcl_HashSearch *searchPtr)
+}
+declare 194 {
+    void Tcl_NotifyChannel(Tcl_Channel channel, int mask)
+}
+declare 195 {
+    Tcl_Obj *Tcl_ObjGetVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+	    Tcl_Obj *part2Ptr, int flags)
+}
+declare 196 {
+    Tcl_Obj *Tcl_ObjSetVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+	    Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr, int flags)
+}
+declare 197 {
+    Tcl_Channel Tcl_OpenCommandChannel(Tcl_Interp *interp, int argc,
+	    CONST84 char **argv, int flags)
+}
+# This is obsolete, use Tcl_FSOpenFileChannel
+declare 198 {
+    Tcl_Channel Tcl_OpenFileChannel(Tcl_Interp *interp, const char *fileName,
+	    const char *modeString, int permissions)
+}
+declare 199 {
+    Tcl_Channel Tcl_OpenTcpClient(Tcl_Interp *interp, int port,
+	    const char *address, const char *myaddr, int myport, int async)
+}
+declare 200 {
+    Tcl_Channel Tcl_OpenTcpServer(Tcl_Interp *interp, int port,
+	    const char *host, Tcl_TcpAcceptProc *acceptProc,
+	    ClientData callbackData)
+}
+declare 201 {
+    void Tcl_Preserve(ClientData data)
+}
+declare 202 {
+    void Tcl_PrintDouble(Tcl_Interp *interp, double value, char *dst)
+}
+declare 203 {
+    int Tcl_PutEnv(const char *assignment)
+}
+declare 204 {
+    CONST84_RETURN char *Tcl_PosixError(Tcl_Interp *interp)
+}
+declare 205 {
+    void Tcl_QueueEvent(Tcl_Event *evPtr, Tcl_QueuePosition position)
+}
+declare 206 {
+    int Tcl_Read(Tcl_Channel chan, char *bufPtr, int toRead)
+}
+declare 207 {
+    void Tcl_ReapDetachedProcs(void)
+}
+declare 208 {
+    int Tcl_RecordAndEval(Tcl_Interp *interp, const char *cmd, int flags)
+}
+declare 209 {
+    int Tcl_RecordAndEvalObj(Tcl_Interp *interp, Tcl_Obj *cmdPtr, int flags)
+}
+declare 210 {
+    void Tcl_RegisterChannel(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 211 {
+    void Tcl_RegisterObjType(const Tcl_ObjType *typePtr)
+}
+declare 212 {
+    Tcl_RegExp Tcl_RegExpCompile(Tcl_Interp *interp, const char *pattern)
+}
+declare 213 {
+    int Tcl_RegExpExec(Tcl_Interp *interp, Tcl_RegExp regexp,
+	    const char *text, const char *start)
+}
+declare 214 {
+    int Tcl_RegExpMatch(Tcl_Interp *interp, const char *text,
+	    const char *pattern)
+}
+declare 215 {
+    void Tcl_RegExpRange(Tcl_RegExp regexp, int index,
+	    CONST84 char **startPtr, CONST84 char **endPtr)
+}
+declare 216 {
+    void Tcl_Release(ClientData clientData)
+}
+declare 217 {
+    void Tcl_ResetResult(Tcl_Interp *interp)
+}
+declare 218 {
+    int Tcl_ScanElement(const char *src, int *flagPtr)
+}
+declare 219 {
+    int Tcl_ScanCountedElement(const char *src, int length, int *flagPtr)
+}
+# Obsolete
+declare 220 {
+    int Tcl_SeekOld(Tcl_Channel chan, int offset, int mode)
+}
+declare 221 {
+    int Tcl_ServiceAll(void)
+}
+declare 222 {
+    int Tcl_ServiceEvent(int flags)
+}
+declare 223 {
+    void Tcl_SetAssocData(Tcl_Interp *interp, const char *name,
+	    Tcl_InterpDeleteProc *proc, ClientData clientData)
+}
+declare 224 {
+    void Tcl_SetChannelBufferSize(Tcl_Channel chan, int sz)
+}
+declare 225 {
+    int Tcl_SetChannelOption(Tcl_Interp *interp, Tcl_Channel chan,
+	    const char *optionName, const char *newValue)
+}
+declare 226 {
+    int Tcl_SetCommandInfo(Tcl_Interp *interp, const char *cmdName,
+	    const Tcl_CmdInfo *infoPtr)
+}
+declare 227 {
+    void Tcl_SetErrno(int err)
+}
+declare 228 {
+    void Tcl_SetErrorCode(Tcl_Interp *interp, ...)
+}
+declare 229 {
+    void Tcl_SetMaxBlockTime(const Tcl_Time *timePtr)
+}
+declare 230 {
+    void Tcl_SetPanicProc(TCL_NORETURN1 Tcl_PanicProc *panicProc)
+}
+declare 231 {
+    int Tcl_SetRecursionLimit(Tcl_Interp *interp, int depth)
+}
+declare 232 {
+    void Tcl_SetResult(Tcl_Interp *interp, char *result,
+	    Tcl_FreeProc *freeProc)
+}
+declare 233 {
+    int Tcl_SetServiceMode(int mode)
+}
+declare 234 {
+    void Tcl_SetObjErrorCode(Tcl_Interp *interp, Tcl_Obj *errorObjPtr)
+}
+declare 235 {
+    void Tcl_SetObjResult(Tcl_Interp *interp, Tcl_Obj *resultObjPtr)
+}
+declare 236 {
+    void Tcl_SetStdChannel(Tcl_Channel channel, int type)
+}
+declare 237 {
+    CONST84_RETURN char *Tcl_SetVar(Tcl_Interp *interp, const char *varName,
+	    const char *newValue, int flags)
+}
+declare 238 {
+    CONST84_RETURN char *Tcl_SetVar2(Tcl_Interp *interp, const char *part1,
+	    const char *part2, const char *newValue, int flags)
+}
+declare 239 {
+    CONST84_RETURN char *Tcl_SignalId(int sig)
+}
+declare 240 {
+    CONST84_RETURN char *Tcl_SignalMsg(int sig)
+}
+declare 241 {
+    void Tcl_SourceRCFile(Tcl_Interp *interp)
+}
+declare 242 {
+    int Tcl_SplitList(Tcl_Interp *interp, const char *listStr, int *argcPtr,
+	    CONST84 char ***argvPtr)
+}
+# Obsolete, use Tcl_FSSplitPath
+declare 243 {
+    void Tcl_SplitPath(const char *path, int *argcPtr, CONST84 char ***argvPtr)
+}
+declare 244 {
+    void Tcl_StaticPackage(Tcl_Interp *interp, const char *pkgName,
+	    Tcl_PackageInitProc *initProc, Tcl_PackageInitProc *safeInitProc)
+}
+declare 245 {
+    int Tcl_StringMatch(const char *str, const char *pattern)
+}
+# Obsolete
+declare 246 {
+    int Tcl_TellOld(Tcl_Channel chan)
+}
+declare 247 {
+    int Tcl_TraceVar(Tcl_Interp *interp, const char *varName, int flags,
+	    Tcl_VarTraceProc *proc, ClientData clientData)
+}
+declare 248 {
+    int Tcl_TraceVar2(Tcl_Interp *interp, const char *part1, const char *part2,
+	    int flags, Tcl_VarTraceProc *proc, ClientData clientData)
+}
+declare 249 {
+    char *Tcl_TranslateFileName(Tcl_Interp *interp, const char *name,
+	    Tcl_DString *bufferPtr)
+}
+declare 250 {
+    int Tcl_Ungets(Tcl_Channel chan, const char *str, int len, int atHead)
+}
+declare 251 {
+    void Tcl_UnlinkVar(Tcl_Interp *interp, const char *varName)
+}
+declare 252 {
+    int Tcl_UnregisterChannel(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 253 {
+    int Tcl_UnsetVar(Tcl_Interp *interp, const char *varName, int flags)
+}
+declare 254 {
+    int Tcl_UnsetVar2(Tcl_Interp *interp, const char *part1, const char *part2,
+	    int flags)
+}
+declare 255 {
+    void Tcl_UntraceVar(Tcl_Interp *interp, const char *varName, int flags,
+	    Tcl_VarTraceProc *proc, ClientData clientData)
+}
+declare 256 {
+    void Tcl_UntraceVar2(Tcl_Interp *interp, const char *part1,
+	    const char *part2, int flags, Tcl_VarTraceProc *proc,
+	    ClientData clientData)
+}
+declare 257 {
+    void Tcl_UpdateLinkedVar(Tcl_Interp *interp, const char *varName)
+}
+declare 258 {
+    int Tcl_UpVar(Tcl_Interp *interp, const char *frameName,
+	    const char *varName, const char *localName, int flags)
+}
+declare 259 {
+    int Tcl_UpVar2(Tcl_Interp *interp, const char *frameName, const char *part1,
+	    const char *part2, const char *localName, int flags)
+}
+declare 260 {
+    int Tcl_VarEval(Tcl_Interp *interp, ...)
+}
+declare 261 {
+    ClientData Tcl_VarTraceInfo(Tcl_Interp *interp, const char *varName,
+	    int flags, Tcl_VarTraceProc *procPtr, ClientData prevClientData)
+}
+declare 262 {
+    ClientData Tcl_VarTraceInfo2(Tcl_Interp *interp, const char *part1,
+	    const char *part2, int flags, Tcl_VarTraceProc *procPtr,
+	    ClientData prevClientData)
+}
+declare 263 {
+    int Tcl_Write(Tcl_Channel chan, const char *s, int slen)
+}
+declare 264 {
+    void Tcl_WrongNumArgs(Tcl_Interp *interp, int objc,
+	    Tcl_Obj *const objv[], const char *message)
+}
+declare 265 {
+    int Tcl_DumpActiveMemory(const char *fileName)
+}
+declare 266 {
+    void Tcl_ValidateAllMemory(const char *file, int line)
+}
+declare 267 {
+    void Tcl_AppendResultVA(Tcl_Interp *interp, va_list argList)
+}
+declare 268 {
+    void Tcl_AppendStringsToObjVA(Tcl_Obj *objPtr, va_list argList)
+}
+declare 269 {
+    char *Tcl_HashStats(Tcl_HashTable *tablePtr)
+}
+declare 270 {
+    CONST84_RETURN char *Tcl_ParseVar(Tcl_Interp *interp, const char *start,
+	    CONST84 char **termPtr)
+}
+declare 271 {
+    CONST84_RETURN char *Tcl_PkgPresent(Tcl_Interp *interp, const char *name,
+	    const char *version, int exact)
+}
+declare 272 {
+    CONST84_RETURN char *Tcl_PkgPresentEx(Tcl_Interp *interp,
+	    const char *name, const char *version, int exact,
+	    void *clientDataPtr)
+}
+declare 273 {
+    int Tcl_PkgProvide(Tcl_Interp *interp, const char *name,
+	    const char *version)
+}
+# TIP #268: The internally used new Require function is in slot 573.
+declare 274 {
+    CONST84_RETURN char *Tcl_PkgRequire(Tcl_Interp *interp, const char *name,
+	    const char *version, int exact)
+}
+declare 275 {
+    void Tcl_SetErrorCodeVA(Tcl_Interp *interp, va_list argList)
+}
+declare 276 {
+    int  Tcl_VarEvalVA(Tcl_Interp *interp, va_list argList)
+}
+declare 277 {
+    Tcl_Pid Tcl_WaitPid(Tcl_Pid pid, int *statPtr, int options)
+}
+declare 278 {
+    TCL_NORETURN void Tcl_PanicVA(const char *format, va_list argList)
+}
+declare 279 {
+    void Tcl_GetVersion(int *major, int *minor, int *patchLevel, int *type)
+}
+declare 280 {
+    void Tcl_InitMemory(Tcl_Interp *interp)
+}
+
+# Andreas Kupries <a.kupries@westend.com>, 03/21/1999
+# "Trf-Patch for filtering channels"
+#
+# C-Level API for (un)stacking of channels. This allows the introduction
+# of filtering channels with relatively little changes to the core.
+# This patch was created in cooperation with Jan Nijtmans j.nijtmans@chello.nl
+# and is therefore part of his plus-patches too.
+#
+# It would have been possible to place the following definitions according
+# to the alphabetical order used elsewhere in this file, but I decided
+# against that to ease the maintenance of the patch across new tcl versions
+# (patch usually has no problems to integrate the patch file for the last
+# version into the new one).
+
+declare 281 {
+    Tcl_Channel Tcl_StackChannel(Tcl_Interp *interp,
+	    const Tcl_ChannelType *typePtr, ClientData instanceData,
+	    int mask, Tcl_Channel prevChan)
+}
+declare 282 {
+    int Tcl_UnstackChannel(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 283 {
+    Tcl_Channel Tcl_GetStackedChannel(Tcl_Channel chan)
+}
+
+# 284 was reserved, but added in 8.4a2
+declare 284 {
+    void Tcl_SetMainLoop(Tcl_MainLoopProc *proc)
+}
+
+# Reserved for future use (8.0.x vs. 8.1)
+#  declare 285 {
+#  }
+
+# Added in 8.1:
+
+declare 286 {
+    void Tcl_AppendObjToObj(Tcl_Obj *objPtr, Tcl_Obj *appendObjPtr)
+}
+declare 287 {
+    Tcl_Encoding Tcl_CreateEncoding(const Tcl_EncodingType *typePtr)
+}
+declare 288 {
+    void Tcl_CreateThreadExitHandler(Tcl_ExitProc *proc, ClientData clientData)
+}
+declare 289 {
+    void Tcl_DeleteThreadExitHandler(Tcl_ExitProc *proc, ClientData clientData)
+}
+declare 290 {
+    void Tcl_DiscardResult(Tcl_SavedResult *statePtr)
+}
+declare 291 {
+    int Tcl_EvalEx(Tcl_Interp *interp, const char *script, int numBytes,
+	    int flags)
+}
+declare 292 {
+    int Tcl_EvalObjv(Tcl_Interp *interp, int objc, Tcl_Obj *const objv[],
+	    int flags)
+}
+declare 293 {
+    int Tcl_EvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags)
+}
+declare 294 {
+    TCL_NORETURN void Tcl_ExitThread(int status)
+}
+declare 295 {
+    int Tcl_ExternalToUtf(Tcl_Interp *interp, Tcl_Encoding encoding,
+	    const char *src, int srcLen, int flags,
+	    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+	    int *srcReadPtr, int *dstWrotePtr, int *dstCharsPtr)
+}
+declare 296 {
+    char *Tcl_ExternalToUtfDString(Tcl_Encoding encoding,
+	    const char *src, int srcLen, Tcl_DString *dsPtr)
+}
+declare 297 {
+    void Tcl_FinalizeThread(void)
+}
+declare 298 {
+    void Tcl_FinalizeNotifier(ClientData clientData)
+}
+declare 299 {
+    void Tcl_FreeEncoding(Tcl_Encoding encoding)
+}
+declare 300 {
+    Tcl_ThreadId Tcl_GetCurrentThread(void)
+}
+declare 301 {
+    Tcl_Encoding Tcl_GetEncoding(Tcl_Interp *interp, const char *name)
+}
+declare 302 {
+    CONST84_RETURN char *Tcl_GetEncodingName(Tcl_Encoding encoding)
+}
+declare 303 {
+    void Tcl_GetEncodingNames(Tcl_Interp *interp)
+}
+declare 304 {
+    int Tcl_GetIndexFromObjStruct(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    const void *tablePtr, int offset, const char *msg, int flags,
+	    int *indexPtr)
+}
+declare 305 {
+    void *Tcl_GetThreadData(Tcl_ThreadDataKey *keyPtr, int size)
+}
+declare 306 {
+    Tcl_Obj *Tcl_GetVar2Ex(Tcl_Interp *interp, const char *part1,
+	    const char *part2, int flags)
+}
+declare 307 {
+    ClientData Tcl_InitNotifier(void)
+}
+declare 308 {
+    void Tcl_MutexLock(Tcl_Mutex *mutexPtr)
+}
+declare 309 {
+    void Tcl_MutexUnlock(Tcl_Mutex *mutexPtr)
+}
+declare 310 {
+    void Tcl_ConditionNotify(Tcl_Condition *condPtr)
+}
+declare 311 {
+    void Tcl_ConditionWait(Tcl_Condition *condPtr, Tcl_Mutex *mutexPtr,
+	    const Tcl_Time *timePtr)
+}
+declare 312 {
+    int Tcl_NumUtfChars(const char *src, int length)
+}
+declare 313 {
+    int Tcl_ReadChars(Tcl_Channel channel, Tcl_Obj *objPtr, int charsToRead,
+	    int appendFlag)
+}
+declare 314 {
+    void Tcl_RestoreResult(Tcl_Interp *interp, Tcl_SavedResult *statePtr)
+}
+declare 315 {
+    void Tcl_SaveResult(Tcl_Interp *interp, Tcl_SavedResult *statePtr)
+}
+declare 316 {
+    int Tcl_SetSystemEncoding(Tcl_Interp *interp, const char *name)
+}
+declare 317 {
+    Tcl_Obj *Tcl_SetVar2Ex(Tcl_Interp *interp, const char *part1,
+	    const char *part2, Tcl_Obj *newValuePtr, int flags)
+}
+declare 318 {
+    void Tcl_ThreadAlert(Tcl_ThreadId threadId)
+}
+declare 319 {
+    void Tcl_ThreadQueueEvent(Tcl_ThreadId threadId, Tcl_Event *evPtr,
+	    Tcl_QueuePosition position)
+}
+declare 320 {
+    Tcl_UniChar Tcl_UniCharAtIndex(const char *src, int index)
+}
+declare 321 {
+    Tcl_UniChar Tcl_UniCharToLower(int ch)
+}
+declare 322 {
+    Tcl_UniChar Tcl_UniCharToTitle(int ch)
+}
+declare 323 {
+    Tcl_UniChar Tcl_UniCharToUpper(int ch)
+}
+declare 324 {
+    int Tcl_UniCharToUtf(int ch, char *buf)
+}
+declare 325 {
+    CONST84_RETURN char *Tcl_UtfAtIndex(const char *src, int index)
+}
+declare 326 {
+    int Tcl_UtfCharComplete(const char *src, int length)
+}
+declare 327 {
+    int Tcl_UtfBackslash(const char *src, int *readPtr, char *dst)
+}
+declare 328 {
+    CONST84_RETURN char *Tcl_UtfFindFirst(const char *src, int ch)
+}
+declare 329 {
+    CONST84_RETURN char *Tcl_UtfFindLast(const char *src, int ch)
+}
+declare 330 {
+    CONST84_RETURN char *Tcl_UtfNext(const char *src)
+}
+declare 331 {
+    CONST84_RETURN char *Tcl_UtfPrev(const char *src, const char *start)
+}
+declare 332 {
+    int Tcl_UtfToExternal(Tcl_Interp *interp, Tcl_Encoding encoding,
+	    const char *src, int srcLen, int flags,
+	    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+	    int *srcReadPtr, int *dstWrotePtr, int *dstCharsPtr)
+}
+declare 333 {
+    char *Tcl_UtfToExternalDString(Tcl_Encoding encoding,
+	    const char *src, int srcLen, Tcl_DString *dsPtr)
+}
+declare 334 {
+    int Tcl_UtfToLower(char *src)
+}
+declare 335 {
+    int Tcl_UtfToTitle(char *src)
+}
+declare 336 {
+    int Tcl_UtfToUniChar(const char *src, Tcl_UniChar *chPtr)
+}
+declare 337 {
+    int Tcl_UtfToUpper(char *src)
+}
+declare 338 {
+    int Tcl_WriteChars(Tcl_Channel chan, const char *src, int srcLen)
+}
+declare 339 {
+    int Tcl_WriteObj(Tcl_Channel chan, Tcl_Obj *objPtr)
+}
+declare 340 {
+    char *Tcl_GetString(Tcl_Obj *objPtr)
+}
+declare 341 {
+    CONST84_RETURN char *Tcl_GetDefaultEncodingDir(void)
+}
+declare 342 {
+    void Tcl_SetDefaultEncodingDir(const char *path)
+}
+declare 343 {
+    void Tcl_AlertNotifier(ClientData clientData)
+}
+declare 344 {
+    void Tcl_ServiceModeHook(int mode)
+}
+declare 345 {
+    int Tcl_UniCharIsAlnum(int ch)
+}
+declare 346 {
+    int Tcl_UniCharIsAlpha(int ch)
+}
+declare 347 {
+    int Tcl_UniCharIsDigit(int ch)
+}
+declare 348 {
+    int Tcl_UniCharIsLower(int ch)
+}
+declare 349 {
+    int Tcl_UniCharIsSpace(int ch)
+}
+declare 350 {
+    int Tcl_UniCharIsUpper(int ch)
+}
+declare 351 {
+    int Tcl_UniCharIsWordChar(int ch)
+}
+declare 352 {
+    int Tcl_UniCharLen(const Tcl_UniChar *uniStr)
+}
+declare 353 {
+    int Tcl_UniCharNcmp(const Tcl_UniChar *ucs, const Tcl_UniChar *uct,
+	    unsigned long numChars)
+}
+declare 354 {
+    char *Tcl_UniCharToUtfDString(const Tcl_UniChar *uniStr,
+	    int uniLength, Tcl_DString *dsPtr)
+}
+declare 355 {
+    Tcl_UniChar *Tcl_UtfToUniCharDString(const char *src,
+	    int length, Tcl_DString *dsPtr)
+}
+declare 356 {
+    Tcl_RegExp Tcl_GetRegExpFromObj(Tcl_Interp *interp, Tcl_Obj *patObj,
+	    int flags)
+}
+declare 357 {
+    Tcl_Obj *Tcl_EvalTokens(Tcl_Interp *interp, Tcl_Token *tokenPtr,
+	    int count)
+}
+declare 358 {
+    void Tcl_FreeParse(Tcl_Parse *parsePtr)
+}
+declare 359 {
+    void Tcl_LogCommandInfo(Tcl_Interp *interp, const char *script,
+	    const char *command, int length)
+}
+declare 360 {
+    int Tcl_ParseBraces(Tcl_Interp *interp, const char *start, int numBytes,
+	    Tcl_Parse *parsePtr, int append, CONST84 char **termPtr)
+}
+declare 361 {
+    int Tcl_ParseCommand(Tcl_Interp *interp, const char *start, int numBytes,
+	    int nested, Tcl_Parse *parsePtr)
+}
+declare 362 {
+    int Tcl_ParseExpr(Tcl_Interp *interp, const char *start, int numBytes,
+	    Tcl_Parse *parsePtr)
+}
+declare 363 {
+    int Tcl_ParseQuotedString(Tcl_Interp *interp, const char *start,
+	    int numBytes, Tcl_Parse *parsePtr, int append,
+	    CONST84 char **termPtr)
+}
+declare 364 {
+    int Tcl_ParseVarName(Tcl_Interp *interp, const char *start, int numBytes,
+	    Tcl_Parse *parsePtr, int append)
+}
+# These 4 functions are obsolete, use Tcl_FSGetCwd, Tcl_FSChdir,
+# Tcl_FSAccess and Tcl_FSStat
+declare 365 {
+    char *Tcl_GetCwd(Tcl_Interp *interp, Tcl_DString *cwdPtr)
+}
+declare 366 {
+   int Tcl_Chdir(const char *dirName)
+}
+declare 367 {
+   int Tcl_Access(const char *path, int mode)
+}
+declare 368 {
+    int Tcl_Stat(const char *path, struct stat *bufPtr)
+}
+declare 369 {
+    int Tcl_UtfNcmp(const char *s1, const char *s2, unsigned long n)
+}
+declare 370 {
+    int Tcl_UtfNcasecmp(const char *s1, const char *s2, unsigned long n)
+}
+declare 371 {
+    int Tcl_StringCaseMatch(const char *str, const char *pattern, int nocase)
+}
+declare 372 {
+    int Tcl_UniCharIsControl(int ch)
+}
+declare 373 {
+    int Tcl_UniCharIsGraph(int ch)
+}
+declare 374 {
+    int Tcl_UniCharIsPrint(int ch)
+}
+declare 375 {
+    int Tcl_UniCharIsPunct(int ch)
+}
+declare 376 {
+    int Tcl_RegExpExecObj(Tcl_Interp *interp, Tcl_RegExp regexp,
+	    Tcl_Obj *textObj, int offset, int nmatches, int flags)
+}
+declare 377 {
+    void Tcl_RegExpGetInfo(Tcl_RegExp regexp, Tcl_RegExpInfo *infoPtr)
+}
+declare 378 {
+    Tcl_Obj *Tcl_NewUnicodeObj(const Tcl_UniChar *unicode, int numChars)
+}
+declare 379 {
+    void Tcl_SetUnicodeObj(Tcl_Obj *objPtr, const Tcl_UniChar *unicode,
+	    int numChars)
+}
+declare 380 {
+    int Tcl_GetCharLength(Tcl_Obj *objPtr)
+}
+declare 381 {
+    Tcl_UniChar Tcl_GetUniChar(Tcl_Obj *objPtr, int index)
+}
+declare 382 {
+    Tcl_UniChar *Tcl_GetUnicode(Tcl_Obj *objPtr)
+}
+declare 383 {
+    Tcl_Obj *Tcl_GetRange(Tcl_Obj *objPtr, int first, int last)
+}
+declare 384 {
+    void Tcl_AppendUnicodeToObj(Tcl_Obj *objPtr, const Tcl_UniChar *unicode,
+	    int length)
+}
+declare 385 {
+    int Tcl_RegExpMatchObj(Tcl_Interp *interp, Tcl_Obj *textObj,
+	    Tcl_Obj *patternObj)
+}
+declare 386 {
+    void Tcl_SetNotifier(Tcl_NotifierProcs *notifierProcPtr)
+}
+declare 387 {
+    Tcl_Mutex *Tcl_GetAllocMutex(void)
+}
+declare 388 {
+    int Tcl_GetChannelNames(Tcl_Interp *interp)
+}
+declare 389 {
+    int Tcl_GetChannelNamesEx(Tcl_Interp *interp, const char *pattern)
+}
+declare 390 {
+    int Tcl_ProcObjCmd(ClientData clientData, Tcl_Interp *interp,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 391 {
+    void Tcl_ConditionFinalize(Tcl_Condition *condPtr)
+}
+declare 392 {
+    void Tcl_MutexFinalize(Tcl_Mutex *mutex)
+}
+declare 393 {
+    int Tcl_CreateThread(Tcl_ThreadId *idPtr, Tcl_ThreadCreateProc *proc,
+	    ClientData clientData, int stackSize, int flags)
+}
+
+# Introduced in 8.3.2
+declare 394 {
+    int Tcl_ReadRaw(Tcl_Channel chan, char *dst, int bytesToRead)
+}
+declare 395 {
+    int Tcl_WriteRaw(Tcl_Channel chan, const char *src, int srcLen)
+}
+declare 396 {
+    Tcl_Channel Tcl_GetTopChannel(Tcl_Channel chan)
+}
+declare 397 {
+    int Tcl_ChannelBuffered(Tcl_Channel chan)
+}
+declare 398 {
+    CONST84_RETURN char *Tcl_ChannelName(const Tcl_ChannelType *chanTypePtr)
+}
+declare 399 {
+    Tcl_ChannelTypeVersion Tcl_ChannelVersion(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 400 {
+    Tcl_DriverBlockModeProc *Tcl_ChannelBlockModeProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 401 {
+    Tcl_DriverCloseProc *Tcl_ChannelCloseProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 402 {
+    Tcl_DriverClose2Proc *Tcl_ChannelClose2Proc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 403 {
+    Tcl_DriverInputProc *Tcl_ChannelInputProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 404 {
+    Tcl_DriverOutputProc *Tcl_ChannelOutputProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 405 {
+    Tcl_DriverSeekProc *Tcl_ChannelSeekProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 406 {
+    Tcl_DriverSetOptionProc *Tcl_ChannelSetOptionProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 407 {
+    Tcl_DriverGetOptionProc *Tcl_ChannelGetOptionProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 408 {
+    Tcl_DriverWatchProc *Tcl_ChannelWatchProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 409 {
+    Tcl_DriverGetHandleProc *Tcl_ChannelGetHandleProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 410 {
+    Tcl_DriverFlushProc *Tcl_ChannelFlushProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+declare 411 {
+    Tcl_DriverHandlerProc *Tcl_ChannelHandlerProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+
+# Introduced in 8.4a2
+declare 412 {
+    int Tcl_JoinThread(Tcl_ThreadId threadId, int *result)
+}
+declare 413 {
+    int Tcl_IsChannelShared(Tcl_Channel channel)
+}
+declare 414 {
+    int Tcl_IsChannelRegistered(Tcl_Interp *interp, Tcl_Channel channel)
+}
+declare 415 {
+    void Tcl_CutChannel(Tcl_Channel channel)
+}
+declare 416 {
+    void Tcl_SpliceChannel(Tcl_Channel channel)
+}
+declare 417 {
+    void Tcl_ClearChannelHandlers(Tcl_Channel channel)
+}
+declare 418 {
+    int Tcl_IsChannelExisting(const char *channelName)
+}
+declare 419 {
+    int Tcl_UniCharNcasecmp(const Tcl_UniChar *ucs, const Tcl_UniChar *uct,
+	    unsigned long numChars)
+}
+declare 420 {
+    int Tcl_UniCharCaseMatch(const Tcl_UniChar *uniStr,
+	    const Tcl_UniChar *uniPattern, int nocase)
+}
+declare 421 {
+    Tcl_HashEntry *Tcl_FindHashEntry(Tcl_HashTable *tablePtr, const void *key)
+}
+declare 422 {
+    Tcl_HashEntry *Tcl_CreateHashEntry(Tcl_HashTable *tablePtr,
+	    const void *key, int *newPtr)
+}
+declare 423 {
+    void Tcl_InitCustomHashTable(Tcl_HashTable *tablePtr, int keyType,
+	    const Tcl_HashKeyType *typePtr)
+}
+declare 424 {
+    void Tcl_InitObjHashTable(Tcl_HashTable *tablePtr)
+}
+declare 425 {
+    ClientData Tcl_CommandTraceInfo(Tcl_Interp *interp, const char *varName,
+	    int flags, Tcl_CommandTraceProc *procPtr,
+	    ClientData prevClientData)
+}
+declare 426 {
+    int Tcl_TraceCommand(Tcl_Interp *interp, const char *varName, int flags,
+	    Tcl_CommandTraceProc *proc, ClientData clientData)
+}
+declare 427 {
+    void Tcl_UntraceCommand(Tcl_Interp *interp, const char *varName,
+	    int flags, Tcl_CommandTraceProc *proc, ClientData clientData)
+}
+declare 428 {
+    char *Tcl_AttemptAlloc(unsigned int size)
+}
+declare 429 {
+    char *Tcl_AttemptDbCkalloc(unsigned int size, const char *file, int line)
+}
+declare 430 {
+    char *Tcl_AttemptRealloc(char *ptr, unsigned int size)
+}
+declare 431 {
+    char *Tcl_AttemptDbCkrealloc(char *ptr, unsigned int size,
+	    const char *file, int line)
+}
+declare 432 {
+    int Tcl_AttemptSetObjLength(Tcl_Obj *objPtr, int length)
+}
+
+# TIP#10 (thread-aware channels) akupries
+declare 433 {
+    Tcl_ThreadId Tcl_GetChannelThread(Tcl_Channel channel)
+}
+
+# introduced in 8.4a3
+declare 434 {
+    Tcl_UniChar *Tcl_GetUnicodeFromObj(Tcl_Obj *objPtr, int *lengthPtr)
+}
+
+# TIP#15 (math function introspection) dkf
+declare 435 {
+    int Tcl_GetMathFuncInfo(Tcl_Interp *interp, const char *name,
+	    int *numArgsPtr, Tcl_ValueType **argTypesPtr,
+	    Tcl_MathProc **procPtr, ClientData *clientDataPtr)
+}
+declare 436 {
+    Tcl_Obj *Tcl_ListMathFuncs(Tcl_Interp *interp, const char *pattern)
+}
+
+# TIP#36 (better access to 'subst') dkf
+declare 437 {
+    Tcl_Obj *Tcl_SubstObj(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags)
+}
+
+# TIP#17 (virtual filesystem layer) vdarley
+declare 438 {
+    int Tcl_DetachChannel(Tcl_Interp *interp, Tcl_Channel channel)
+}
+declare 439 {
+    int Tcl_IsStandardChannel(Tcl_Channel channel)
+}
+declare 440 {
+    int	Tcl_FSCopyFile(Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr)
+}
+declare 441 {
+    int	Tcl_FSCopyDirectory(Tcl_Obj *srcPathPtr,
+	    Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr)
+}
+declare 442 {
+    int	Tcl_FSCreateDirectory(Tcl_Obj *pathPtr)
+}
+declare 443 {
+    int	Tcl_FSDeleteFile(Tcl_Obj *pathPtr)
+}
+declare 444 {
+    int	Tcl_FSLoadFile(Tcl_Interp *interp, Tcl_Obj *pathPtr, const char *sym1,
+	    const char *sym2, Tcl_PackageInitProc **proc1Ptr,
+	    Tcl_PackageInitProc **proc2Ptr, Tcl_LoadHandle *handlePtr,
+	    Tcl_FSUnloadFileProc **unloadProcPtr)
+}
+declare 445 {
+    int	Tcl_FSMatchInDirectory(Tcl_Interp *interp, Tcl_Obj *result,
+	    Tcl_Obj *pathPtr, const char *pattern, Tcl_GlobTypeData *types)
+}
+declare 446 {
+    Tcl_Obj *Tcl_FSLink(Tcl_Obj *pathPtr, Tcl_Obj *toPtr, int linkAction)
+}
+declare 447 {
+    int Tcl_FSRemoveDirectory(Tcl_Obj *pathPtr,
+	    int recursive, Tcl_Obj **errorPtr)
+}
+declare 448 {
+    int	Tcl_FSRenameFile(Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr)
+}
+declare 449 {
+    int	Tcl_FSLstat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf)
+}
+declare 450 {
+    int Tcl_FSUtime(Tcl_Obj *pathPtr, struct utimbuf *tval)
+}
+declare 451 {
+    int Tcl_FSFileAttrsGet(Tcl_Interp *interp,
+	    int index, Tcl_Obj *pathPtr, Tcl_Obj **objPtrRef)
+}
+declare 452 {
+    int Tcl_FSFileAttrsSet(Tcl_Interp *interp,
+	    int index, Tcl_Obj *pathPtr, Tcl_Obj *objPtr)
+}
+declare 453 {
+    const char *CONST86 *Tcl_FSFileAttrStrings(Tcl_Obj *pathPtr,
+	    Tcl_Obj **objPtrRef)
+}
+declare 454 {
+    int Tcl_FSStat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf)
+}
+declare 455 {
+    int Tcl_FSAccess(Tcl_Obj *pathPtr, int mode)
+}
+declare 456 {
+    Tcl_Channel Tcl_FSOpenFileChannel(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+	    const char *modeString, int permissions)
+}
+declare 457 {
+    Tcl_Obj *Tcl_FSGetCwd(Tcl_Interp *interp)
+}
+declare 458 {
+    int Tcl_FSChdir(Tcl_Obj *pathPtr)
+}
+declare 459 {
+    int Tcl_FSConvertToPathType(Tcl_Interp *interp, Tcl_Obj *pathPtr)
+}
+declare 460 {
+    Tcl_Obj *Tcl_FSJoinPath(Tcl_Obj *listObj, int elements)
+}
+declare 461 {
+    Tcl_Obj *Tcl_FSSplitPath(Tcl_Obj *pathPtr, int *lenPtr)
+}
+declare 462 {
+    int Tcl_FSEqualPaths(Tcl_Obj *firstPtr, Tcl_Obj *secondPtr)
+}
+declare 463 {
+    Tcl_Obj *Tcl_FSGetNormalizedPath(Tcl_Interp *interp, Tcl_Obj *pathPtr)
+}
+declare 464 {
+    Tcl_Obj *Tcl_FSJoinToPath(Tcl_Obj *pathPtr, int objc,
+	    Tcl_Obj *const objv[])
+}
+declare 465 {
+    ClientData Tcl_FSGetInternalRep(Tcl_Obj *pathPtr,
+	    const Tcl_Filesystem *fsPtr)
+}
+declare 466 {
+    Tcl_Obj *Tcl_FSGetTranslatedPath(Tcl_Interp *interp, Tcl_Obj *pathPtr)
+}
+declare 467 {
+    int Tcl_FSEvalFile(Tcl_Interp *interp, Tcl_Obj *fileName)
+}
+declare 468 {
+    Tcl_Obj *Tcl_FSNewNativePath(const Tcl_Filesystem *fromFilesystem,
+	    ClientData clientData)
+}
+declare 469 {
+    const void *Tcl_FSGetNativePath(Tcl_Obj *pathPtr)
+}
+declare 470 {
+    Tcl_Obj *Tcl_FSFileSystemInfo(Tcl_Obj *pathPtr)
+}
+declare 471 {
+    Tcl_Obj *Tcl_FSPathSeparator(Tcl_Obj *pathPtr)
+}
+declare 472 {
+    Tcl_Obj *Tcl_FSListVolumes(void)
+}
+declare 473 {
+    int Tcl_FSRegister(ClientData clientData, const Tcl_Filesystem *fsPtr)
+}
+declare 474 {
+    int Tcl_FSUnregister(const Tcl_Filesystem *fsPtr)
+}
+declare 475 {
+    ClientData Tcl_FSData(const Tcl_Filesystem *fsPtr)
+}
+declare 476 {
+    const char *Tcl_FSGetTranslatedStringPath(Tcl_Interp *interp,
+	    Tcl_Obj *pathPtr)
+}
+declare 477 {
+    CONST86 Tcl_Filesystem *Tcl_FSGetFileSystemForPath(Tcl_Obj *pathPtr)
+}
+declare 478 {
+    Tcl_PathType Tcl_FSGetPathType(Tcl_Obj *pathPtr)
+}
+
+# TIP#49 (detection of output buffering) akupries
+declare 479 {
+    int Tcl_OutputBuffered(Tcl_Channel chan)
+}
+declare 480 {
+    void Tcl_FSMountsChanged(const Tcl_Filesystem *fsPtr)
+}
+
+# TIP#56 (evaluate a parsed script) msofer
+declare 481 {
+    int Tcl_EvalTokensStandard(Tcl_Interp *interp, Tcl_Token *tokenPtr,
+	    int count)
+}
+
+# TIP#73 (access to current time) kbk
+declare 482 {
+    void Tcl_GetTime(Tcl_Time *timeBuf)
+}
+
+# TIP#32 (object-enabled traces) kbk
+declare 483 {
+    Tcl_Trace Tcl_CreateObjTrace(Tcl_Interp *interp, int level, int flags,
+	    Tcl_CmdObjTraceProc *objProc, ClientData clientData,
+	    Tcl_CmdObjTraceDeleteProc *delProc)
+}
+declare 484 {
+    int Tcl_GetCommandInfoFromToken(Tcl_Command token, Tcl_CmdInfo *infoPtr)
+}
+declare 485 {
+    int Tcl_SetCommandInfoFromToken(Tcl_Command token,
+	    const Tcl_CmdInfo *infoPtr)
+}
+
+### New functions on 64-bit dev branch ###
+# TIP#72 (64-bit values) dkf
+declare 486 {
+    Tcl_Obj *Tcl_DbNewWideIntObj(Tcl_WideInt wideValue,
+	    const char *file, int line)
+}
+declare 487 {
+    int Tcl_GetWideIntFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    Tcl_WideInt *widePtr)
+}
+declare 488 {
+    Tcl_Obj *Tcl_NewWideIntObj(Tcl_WideInt wideValue)
+}
+declare 489 {
+    void Tcl_SetWideIntObj(Tcl_Obj *objPtr, Tcl_WideInt wideValue)
+}
+declare 490 {
+    Tcl_StatBuf *Tcl_AllocStatBuf(void)
+}
+declare 491 {
+    Tcl_WideInt Tcl_Seek(Tcl_Channel chan, Tcl_WideInt offset, int mode)
+}
+declare 492 {
+    Tcl_WideInt Tcl_Tell(Tcl_Channel chan)
+}
+
+# TIP#91 (back-compat enhancements for channels) dkf
+declare 493 {
+    Tcl_DriverWideSeekProc *Tcl_ChannelWideSeekProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+
+# ----- BASELINE -- FOR -- 8.4.0 ----- #
+
+# TIP#111 (dictionaries) dkf
+declare 494 {
+    int Tcl_DictObjPut(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+	    Tcl_Obj *keyPtr, Tcl_Obj *valuePtr)
+}
+declare 495 {
+    int Tcl_DictObjGet(Tcl_Interp *interp, Tcl_Obj *dictPtr, Tcl_Obj *keyPtr,
+	    Tcl_Obj **valuePtrPtr)
+}
+declare 496 {
+    int Tcl_DictObjRemove(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+	    Tcl_Obj *keyPtr)
+}
+declare 497 {
+    int Tcl_DictObjSize(Tcl_Interp *interp, Tcl_Obj *dictPtr, int *sizePtr)
+}
+declare 498 {
+    int Tcl_DictObjFirst(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+	    Tcl_DictSearch *searchPtr,
+	    Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr, int *donePtr)
+}
+declare 499 {
+    void Tcl_DictObjNext(Tcl_DictSearch *searchPtr,
+	    Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr, int *donePtr)
+}
+declare 500 {
+    void Tcl_DictObjDone(Tcl_DictSearch *searchPtr)
+}
+declare 501 {
+    int Tcl_DictObjPutKeyList(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+	    int keyc, Tcl_Obj *const *keyv, Tcl_Obj *valuePtr)
+}
+declare 502 {
+    int Tcl_DictObjRemoveKeyList(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+	    int keyc, Tcl_Obj *const *keyv)
+}
+declare 503 {
+    Tcl_Obj *Tcl_NewDictObj(void)
+}
+declare 504 {
+    Tcl_Obj *Tcl_DbNewDictObj(const char *file, int line)
+}
+
+# TIP#59 (configuration reporting) akupries
+declare 505 {
+    void Tcl_RegisterConfig(Tcl_Interp *interp, const char *pkgName,
+	    const Tcl_Config *configuration, const char *valEncoding)
+}
+
+# TIP #139 (partial exposure of namespace API - transferred from tclInt.decls)
+# dkf, API by Brent Welch?
+declare 506 {
+    Tcl_Namespace *Tcl_CreateNamespace(Tcl_Interp *interp, const char *name,
+	    ClientData clientData, Tcl_NamespaceDeleteProc *deleteProc)
+}
+declare 507 {
+    void Tcl_DeleteNamespace(Tcl_Namespace *nsPtr)
+}
+declare 508 {
+    int Tcl_AppendExportList(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    Tcl_Obj *objPtr)
+}
+declare 509 {
+    int Tcl_Export(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern, int resetListFirst)
+}
+declare 510 {
+    int Tcl_Import(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern, int allowOverwrite)
+}
+declare 511 {
+    int Tcl_ForgetImport(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern)
+}
+declare 512 {
+    Tcl_Namespace *Tcl_GetCurrentNamespace(Tcl_Interp *interp)
+}
+declare 513 {
+    Tcl_Namespace *Tcl_GetGlobalNamespace(Tcl_Interp *interp)
+}
+declare 514 {
+    Tcl_Namespace *Tcl_FindNamespace(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *contextNsPtr, int flags)
+}
+declare 515 {
+    Tcl_Command Tcl_FindCommand(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *contextNsPtr, int flags)
+}
+declare 516 {
+    Tcl_Command Tcl_GetCommandFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 517 {
+    void Tcl_GetCommandFullName(Tcl_Interp *interp, Tcl_Command command,
+	    Tcl_Obj *objPtr)
+}
+
+# TIP#137 (encoding-aware source command) dgp for Anton Kovalenko
+declare 518 {
+    int Tcl_FSEvalFileEx(Tcl_Interp *interp, Tcl_Obj *fileName,
+	    const char *encodingName)
+}
+
+# TIP#121 (exit handler) dkf for Joe Mistachkin
+declare 519 {
+    Tcl_ExitProc *Tcl_SetExitProc(TCL_NORETURN1 Tcl_ExitProc *proc)
+}
+
+# TIP#143 (resource limits) dkf
+declare 520 {
+    void Tcl_LimitAddHandler(Tcl_Interp *interp, int type,
+	    Tcl_LimitHandlerProc *handlerProc, ClientData clientData,
+	    Tcl_LimitHandlerDeleteProc *deleteProc)
+}
+declare 521 {
+    void Tcl_LimitRemoveHandler(Tcl_Interp *interp, int type,
+	    Tcl_LimitHandlerProc *handlerProc, ClientData clientData)
+}
+declare 522 {
+    int Tcl_LimitReady(Tcl_Interp *interp)
+}
+declare 523 {
+    int Tcl_LimitCheck(Tcl_Interp *interp)
+}
+declare 524 {
+    int Tcl_LimitExceeded(Tcl_Interp *interp)
+}
+declare 525 {
+    void Tcl_LimitSetCommands(Tcl_Interp *interp, int commandLimit)
+}
+declare 526 {
+    void Tcl_LimitSetTime(Tcl_Interp *interp, Tcl_Time *timeLimitPtr)
+}
+declare 527 {
+    void Tcl_LimitSetGranularity(Tcl_Interp *interp, int type, int granularity)
+}
+declare 528 {
+    int Tcl_LimitTypeEnabled(Tcl_Interp *interp, int type)
+}
+declare 529 {
+    int Tcl_LimitTypeExceeded(Tcl_Interp *interp, int type)
+}
+declare 530 {
+    void Tcl_LimitTypeSet(Tcl_Interp *interp, int type)
+}
+declare 531 {
+    void Tcl_LimitTypeReset(Tcl_Interp *interp, int type)
+}
+declare 532 {
+    int Tcl_LimitGetCommands(Tcl_Interp *interp)
+}
+declare 533 {
+    void Tcl_LimitGetTime(Tcl_Interp *interp, Tcl_Time *timeLimitPtr)
+}
+declare 534 {
+    int Tcl_LimitGetGranularity(Tcl_Interp *interp, int type)
+}
+
+# TIP#226 (interpreter result state management) dgp
+declare 535 {
+    Tcl_InterpState Tcl_SaveInterpState(Tcl_Interp *interp, int status)
+}
+declare 536 {
+    int Tcl_RestoreInterpState(Tcl_Interp *interp, Tcl_InterpState state)
+}
+declare 537 {
+    void Tcl_DiscardInterpState(Tcl_InterpState state)
+}
+
+# TIP#227 (return options interface) dgp
+declare 538 {
+    int Tcl_SetReturnOptions(Tcl_Interp *interp, Tcl_Obj *options)
+}
+declare 539 {
+    Tcl_Obj *Tcl_GetReturnOptions(Tcl_Interp *interp, int result)
+}
+
+# TIP#235 (ensembles) dkf
+declare 540 {
+    int Tcl_IsEnsemble(Tcl_Command token)
+}
+declare 541 {
+    Tcl_Command Tcl_CreateEnsemble(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *namespacePtr, int flags)
+}
+declare 542 {
+    Tcl_Command Tcl_FindEnsemble(Tcl_Interp *interp, Tcl_Obj *cmdNameObj,
+	    int flags)
+}
+declare 543 {
+    int Tcl_SetEnsembleSubcommandList(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj *subcmdList)
+}
+declare 544 {
+    int Tcl_SetEnsembleMappingDict(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj *mapDict)
+}
+declare 545 {
+    int Tcl_SetEnsembleUnknownHandler(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj *unknownList)
+}
+declare 546 {
+    int Tcl_SetEnsembleFlags(Tcl_Interp *interp, Tcl_Command token, int flags)
+}
+declare 547 {
+    int Tcl_GetEnsembleSubcommandList(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj **subcmdListPtr)
+}
+declare 548 {
+    int Tcl_GetEnsembleMappingDict(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj **mapDictPtr)
+}
+declare 549 {
+    int Tcl_GetEnsembleUnknownHandler(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj **unknownListPtr)
+}
+declare 550 {
+    int Tcl_GetEnsembleFlags(Tcl_Interp *interp, Tcl_Command token,
+	    int *flagsPtr)
+}
+declare 551 {
+    int Tcl_GetEnsembleNamespace(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Namespace **namespacePtrPtr)
+}
+
+# TIP#233 (virtualized time) akupries
+declare 552 {
+    void Tcl_SetTimeProc(Tcl_GetTimeProc *getProc,
+	    Tcl_ScaleTimeProc *scaleProc,
+	    ClientData clientData)
+}
+declare 553 {
+    void Tcl_QueryTimeProc(Tcl_GetTimeProc **getProc,
+	    Tcl_ScaleTimeProc **scaleProc,
+	    ClientData *clientData)
+}
+
+# TIP#218 (driver thread actions) davygrvy/akupries ChannelType ver 4
+declare 554 {
+    Tcl_DriverThreadActionProc *Tcl_ChannelThreadActionProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+
+# TIP#237 (arbitrary-precision integers) kbk
+declare 555 {
+    Tcl_Obj *Tcl_NewBignumObj(mp_int *value)
+}
+declare 556 {
+    Tcl_Obj *Tcl_DbNewBignumObj(mp_int *value, const char *file, int line)
+}
+declare 557 {
+    void Tcl_SetBignumObj(Tcl_Obj *obj, mp_int *value)
+}
+declare 558 {
+    int Tcl_GetBignumFromObj(Tcl_Interp *interp, Tcl_Obj *obj, mp_int *value)
+}
+declare 559 {
+    int Tcl_TakeBignumFromObj(Tcl_Interp *interp, Tcl_Obj *obj, mp_int *value)
+}
+
+# TIP #208 ('chan' command) jeffh
+declare 560 {
+    int Tcl_TruncateChannel(Tcl_Channel chan, Tcl_WideInt length)
+}
+declare 561 {
+    Tcl_DriverTruncateProc *Tcl_ChannelTruncateProc(
+	    const Tcl_ChannelType *chanTypePtr)
+}
+
+# TIP#219 (channel reflection api) akupries
+declare 562 {
+    void Tcl_SetChannelErrorInterp(Tcl_Interp *interp, Tcl_Obj *msg)
+}
+declare 563 {
+    void Tcl_GetChannelErrorInterp(Tcl_Interp *interp, Tcl_Obj **msg)
+}
+declare 564 {
+    void Tcl_SetChannelError(Tcl_Channel chan, Tcl_Obj *msg)
+}
+declare 565 {
+    void Tcl_GetChannelError(Tcl_Channel chan, Tcl_Obj **msg)
+}
+
+# TIP #237 (additional conversion functions for bignum support) kbk/dgp
+declare 566 {
+    int Tcl_InitBignumFromDouble(Tcl_Interp *interp, double initval,
+	    mp_int *toInit)
+}
+
+# TIP#181 (namespace unknown command) dgp for Neil Madden
+declare 567 {
+    Tcl_Obj *Tcl_GetNamespaceUnknownHandler(Tcl_Interp *interp,
+	    Tcl_Namespace *nsPtr)
+}
+declare 568 {
+    int Tcl_SetNamespaceUnknownHandler(Tcl_Interp *interp,
+	    Tcl_Namespace *nsPtr, Tcl_Obj *handlerPtr)
+}
+
+# TIP#258 (enhanced interface for encodings) dgp
+declare 569 {
+    int Tcl_GetEncodingFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    Tcl_Encoding *encodingPtr)
+}
+declare 570 {
+    Tcl_Obj *Tcl_GetEncodingSearchPath(void)
+}
+declare 571 {
+    int Tcl_SetEncodingSearchPath(Tcl_Obj *searchPath)
+}
+declare 572 {
+    const char *Tcl_GetEncodingNameFromEnvironment(Tcl_DString *bufPtr)
+}
+
+# TIP#268 (extended version numbers and requirements) akupries
+declare 573 {
+    int Tcl_PkgRequireProc(Tcl_Interp *interp, const char *name,
+	    int objc, Tcl_Obj *const objv[], void *clientDataPtr)
+}
+
+# TIP#270 (utility C routines for string formatting) dgp
+declare 574 {
+    void Tcl_AppendObjToErrorInfo(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 575 {
+    void Tcl_AppendLimitedToObj(Tcl_Obj *objPtr, const char *bytes, int length,
+	    int limit, const char *ellipsis)
+}
+declare 576 {
+    Tcl_Obj *Tcl_Format(Tcl_Interp *interp, const char *format, int objc,
+	    Tcl_Obj *const objv[])
+}
+declare 577 {
+    int Tcl_AppendFormatToObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    const char *format, int objc, Tcl_Obj *const objv[])
+}
+declare 578 {
+    Tcl_Obj *Tcl_ObjPrintf(const char *format, ...)
+}
+declare 579 {
+    void Tcl_AppendPrintfToObj(Tcl_Obj *objPtr, const char *format, ...)
+}
+
+# ----- BASELINE -- FOR -- 8.5.0 ----- #
+
+# TIP #285 (script cancellation support) jmistachkin
+declare 580 {
+    int Tcl_CancelEval(Tcl_Interp *interp, Tcl_Obj *resultObjPtr,
+	    ClientData clientData, int flags)
+}
+declare 581 {
+    int Tcl_Canceled(Tcl_Interp *interp, int flags)
+}
+
+# TIP#304 (chan pipe) aferrieux
+declare 582 {
+    int Tcl_CreatePipe(Tcl_Interp  *interp, Tcl_Channel *rchan,
+	    Tcl_Channel *wchan, int flags)
+}
+
+# TIP #322 (NRE public interface) msofer
+declare 583 {
+    Tcl_Command Tcl_NRCreateCommand(Tcl_Interp *interp,
+	    const char *cmdName, Tcl_ObjCmdProc *proc,
+	    Tcl_ObjCmdProc *nreProc, ClientData clientData,
+	    Tcl_CmdDeleteProc *deleteProc)
+}
+declare 584 {
+    int Tcl_NREvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags)
+}
+declare 585 {
+    int Tcl_NREvalObjv(Tcl_Interp *interp, int objc, Tcl_Obj *const objv[],
+	    int flags)
+}
+declare 586 {
+    int Tcl_NRCmdSwap(Tcl_Interp *interp, Tcl_Command cmd, int objc,
+	    Tcl_Obj *const objv[], int flags)
+}
+declare 587 {
+    void Tcl_NRAddCallback(Tcl_Interp *interp, Tcl_NRPostProc *postProcPtr,
+	    ClientData data0, ClientData data1, ClientData data2,
+	    ClientData data3)
+}
+# For use by NR extenders, to have a simple way to also provide a (required!)
+# classic objProc
+declare 588 {
+    int Tcl_NRCallObjProc(Tcl_Interp *interp, Tcl_ObjCmdProc *objProc,
+	    ClientData clientData, int objc, Tcl_Obj *const objv[])
+}
+
+# TIP#316 (Tcl_StatBuf reader functions) dkf
+declare 589 {
+    unsigned Tcl_GetFSDeviceFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 590 {
+    unsigned Tcl_GetFSInodeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 591 {
+    unsigned Tcl_GetModeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 592 {
+    int Tcl_GetLinkCountFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 593 {
+    int Tcl_GetUserIdFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 594 {
+    int Tcl_GetGroupIdFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 595 {
+    int Tcl_GetDeviceTypeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 596 {
+    Tcl_WideInt Tcl_GetAccessTimeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 597 {
+    Tcl_WideInt Tcl_GetModificationTimeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 598 {
+    Tcl_WideInt Tcl_GetChangeTimeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 599 {
+    Tcl_WideUInt Tcl_GetSizeFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 600 {
+    Tcl_WideUInt Tcl_GetBlocksFromStat(const Tcl_StatBuf *statPtr)
+}
+declare 601 {
+    unsigned Tcl_GetBlockSizeFromStat(const Tcl_StatBuf *statPtr)
+}
+
+# TIP#314 (ensembles with parameters) dkf for Lars Hellstr"om
+declare 602 {
+    int Tcl_SetEnsembleParameterList(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj *paramList)
+}
+declare 603 {
+    int Tcl_GetEnsembleParameterList(Tcl_Interp *interp, Tcl_Command token,
+	    Tcl_Obj **paramListPtr)
+}
+
+# TIP#265 (option parser) dkf for Sam Bromley
+declare 604 {
+    int Tcl_ParseArgsObjv(Tcl_Interp *interp, const Tcl_ArgvInfo *argTable,
+	    int *objcPtr, Tcl_Obj *const *objv, Tcl_Obj ***remObjv)
+}
+
+# TIP#336 (manipulate the error line) dgp
+declare 605 {
+    int Tcl_GetErrorLine(Tcl_Interp *interp)
+}
+declare 606 {
+    void Tcl_SetErrorLine(Tcl_Interp *interp, int lineNum)
+}
+
+# TIP#307 (move results between interpreters) dkf
+declare 607 {
+    void Tcl_TransferResult(Tcl_Interp *sourceInterp, int result,
+	    Tcl_Interp *targetInterp)
+}
+
+# TIP#335 (detect if interpreter in use) jmistachkin
+declare 608 {
+    int Tcl_InterpActive(Tcl_Interp *interp)
+}
+
+# TIP#337 (log exception for background processing) dgp
+declare 609 {
+    void Tcl_BackgroundException(Tcl_Interp *interp, int code)
+}
+
+# TIP#234 (zlib interface) dkf/Pascal Scheffers
+declare 610 {
+    int Tcl_ZlibDeflate(Tcl_Interp *interp, int format, Tcl_Obj *data,
+	    int level, Tcl_Obj *gzipHeaderDictObj)
+}
+declare 611 {
+    int Tcl_ZlibInflate(Tcl_Interp *interp, int format, Tcl_Obj *data,
+	    int buffersize, Tcl_Obj *gzipHeaderDictObj)
+}
+declare 612 {
+    unsigned int Tcl_ZlibCRC32(unsigned int crc, const unsigned char *buf,
+	    int len)
+}
+declare 613 {
+    unsigned int Tcl_ZlibAdler32(unsigned int adler, const unsigned char *buf,
+	    int len)
+}
+declare 614 {
+    int Tcl_ZlibStreamInit(Tcl_Interp *interp, int mode, int format,
+	    int level, Tcl_Obj *dictObj, Tcl_ZlibStream *zshandle)
+}
+declare 615 {
+    Tcl_Obj *Tcl_ZlibStreamGetCommandName(Tcl_ZlibStream zshandle)
+}
+declare 616 {
+    int Tcl_ZlibStreamEof(Tcl_ZlibStream zshandle)
+}
+declare 617 {
+    int Tcl_ZlibStreamChecksum(Tcl_ZlibStream zshandle)
+}
+declare 618 {
+    int Tcl_ZlibStreamPut(Tcl_ZlibStream zshandle, Tcl_Obj *data, int flush)
+}
+declare 619 {
+    int Tcl_ZlibStreamGet(Tcl_ZlibStream zshandle, Tcl_Obj *data, int count)
+}
+declare 620 {
+    int Tcl_ZlibStreamClose(Tcl_ZlibStream zshandle)
+}
+declare 621 {
+    int Tcl_ZlibStreamReset(Tcl_ZlibStream zshandle)
+}
+
+# TIP 338 (control over startup script) dgp
+declare 622 {
+    void Tcl_SetStartupScript(Tcl_Obj *path, const char *encoding)
+}
+declare 623 {
+    Tcl_Obj *Tcl_GetStartupScript(const char **encodingPtr)
+}
+
+# TIP#332 (half-close made public) aferrieux
+declare 624 {
+    int Tcl_CloseEx(Tcl_Interp *interp, Tcl_Channel chan, int flags)
+}
+
+# TIP #353 (NR-enabled expressions) dgp
+declare 625 {
+    int Tcl_NRExprObj(Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_Obj *resultPtr)
+}
+
+# TIP #356 (NR-enabled substitution) dgp
+declare 626 {
+    int Tcl_NRSubstObj(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags)
+}
+
+# TIP #357 (Export TclLoadFile and TclpFindSymbol) kbk
+declare 627 {
+    int Tcl_LoadFile(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+		     const char *const symv[], int flags, void *procPtrs,
+		     Tcl_LoadHandle *handlePtr)
+}
+declare 628 {
+    void *Tcl_FindSymbol(Tcl_Interp *interp, Tcl_LoadHandle handle,
+			 const char *symbol)
+}
+declare 629 {
+    int Tcl_FSUnloadFile(Tcl_Interp *interp, Tcl_LoadHandle handlePtr)
+}
+
+# TIP #400
+declare 630 {
+    void Tcl_ZlibStreamSetCompressionDictionary(Tcl_ZlibStream zhandle,
+	    Tcl_Obj *compressionDictionaryObj)
+}
+
+# ----- BASELINE -- FOR -- 8.6.0 ----- #
+
+# TIP #456
+declare 631 {
+    Tcl_Channel Tcl_OpenTcpServerEx(Tcl_Interp *interp, const char *service,
+	    const char *host, unsigned int flags, Tcl_TcpAcceptProc *acceptProc,
+	    ClientData callbackData)
+}
+
+# ----- BASELINE -- FOR -- 8.7.0 ----- #
+
+
+
+##############################################################################
+
+# Define the platform specific public Tcl interface. These functions are only
+# available on the designated platform.
+
+interface tclPlat
+
+################################
+# Unix specific functions
+#   (none)
+
+################################
+# Windows specific functions
+
+# Added in Tcl 8.1
+
+declare 0 win {
+    TCHAR *Tcl_WinUtfToTChar(const char *str, int len, Tcl_DString *dsPtr)
+}
+declare 1 win {
+    char *Tcl_WinTCharToUtf(const TCHAR *str, int len, Tcl_DString *dsPtr)
+}
+
+################################
+# Mac OS X specific functions
+
+declare 0 macosx {
+    int Tcl_MacOSXOpenBundleResources(Tcl_Interp *interp,
+	    const char *bundleName, int hasResourceFile,
+	    int maxPathLen, char *libraryPath)
+}
+declare 1 macosx {
+    int Tcl_MacOSXOpenVersionedBundleResources(Tcl_Interp *interp,
+	    const char *bundleName, const char *bundleVersion,
+	    int hasResourceFile, int maxPathLen, char *libraryPath)
+}
+
+##############################################################################
+
+# Public functions that are not accessible via the stubs table.
+
+export {
+    void Tcl_Main(int argc, char **argv, Tcl_AppInitProc *appInitProc)
+}
+export {
+    const char *Tcl_InitStubs(Tcl_Interp *interp, const char *version,
+	int exact)
+}
+export {
+    const char *TclTomMathInitializeStubs(Tcl_Interp* interp,
+	const char* version, int epoch, int revision)
+}
+export {
+    const char *Tcl_PkgInitStubsCheck(Tcl_Interp *interp, const char *version,
+	int exact)
+}
+export {
+    void Tcl_GetMemoryInfo(Tcl_DString *dsPtr)
+}
+
+# Local Variables:
+# mode: tcl
+# End:
diff --git a/generic/tcl.h b/generic/tcl.h
new file mode 100644
index 0000000..6fa26f9
--- /dev/null
+++ b/generic/tcl.h
@@ -0,0 +1,2661 @@
+/*
+ * tcl.h --
+ *
+ *	This header file describes the externally-visible facilities of the
+ *	Tcl interpreter.
+ *
+ * Copyright (c) 1987-1994 The Regents of the University of California.
+ * Copyright (c) 1993-1996 Lucent Technologies.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 by Scriptics Corporation.
+ * Copyright (c) 2002 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCL
+#define _TCL
+
+/*
+ * For C++ compilers, use extern "C"
+ */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * The following defines are used to indicate the various release levels.
+ */
+
+#define TCL_ALPHA_RELEASE	0
+#define TCL_BETA_RELEASE	1
+#define TCL_FINAL_RELEASE	2
+
+/*
+ * When version numbers change here, must also go into the following files and
+ * update the version numbers:
+ *
+ * library/init.tcl	(1 LOC patch)
+ * unix/configure.ac	(2 LOC Major, 2 LOC minor, 1 LOC patch)
+ * win/configure.ac	(as above)
+ * win/tcl.m4		(not patchlevel)
+ * README		(sections 0 and 2, with and without separator)
+ * macosx/Tcl.pbproj/project.pbxproj (not patchlevel) 1 LOC
+ * macosx/Tcl.pbproj/default.pbxuser (not patchlevel) 1 LOC
+ * macosx/Tcl.xcode/project.pbxproj (not patchlevel) 2 LOC
+ * macosx/Tcl.xcode/default.pbxuser (not patchlevel) 1 LOC
+ * macosx/Tcl-Common.xcconfig (not patchlevel) 1 LOC
+ * win/README		(not patchlevel) (sections 0 and 2)
+ * unix/tcl.spec	(1 LOC patch)
+ * tools/tcl.hpj.in	(not patchlevel, for windows installer)
+ */
+
+#define TCL_MAJOR_VERSION   8
+#define TCL_MINOR_VERSION   7
+#define TCL_RELEASE_LEVEL   TCL_ALPHA_RELEASE
+#define TCL_RELEASE_SERIAL  0
+
+#define TCL_VERSION	    "8.7"
+#define TCL_PATCH_LEVEL	    "8.7a0"
+
+#if !defined(TCL_NO_DEPRECATED) || defined(RC_INVOKED)
+/*
+ *----------------------------------------------------------------------------
+ * The following definitions set up the proper options for Windows compilers.
+ * We use this method because there is no autoconf equivalent.
+ */
+
+#ifdef _WIN32
+#   ifndef __WIN32__
+#	define __WIN32__
+#   endif
+#   ifndef WIN32
+#	define WIN32
+#   endif
+#endif
+
+/*
+ * Utility macros: STRINGIFY takes an argument and wraps it in "" (double
+ * quotation marks), JOIN joins two arguments.
+ */
+
+#ifndef STRINGIFY
+#  define STRINGIFY(x) STRINGIFY1(x)
+#  define STRINGIFY1(x) #x
+#endif
+#ifndef JOIN
+#  define JOIN(a,b) JOIN1(a,b)
+#  define JOIN1(a,b) a##b
+#endif
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * A special definition used to allow this header file to be included from
+ * windows resource files so that they can obtain version information.
+ * RC_INVOKED is defined by default by the windows RC tool.
+ *
+ * Resource compilers don't like all the C stuff, like typedefs and function
+ * declarations, that occur below, so block them out.
+ */
+
+#ifndef RC_INVOKED
+
+/*
+ * Special macro to define mutexes, that doesn't do anything if we are not
+ * using threads.
+ */
+
+#ifdef TCL_THREADS
+#define TCL_DECLARE_MUTEX(name) static Tcl_Mutex name;
+#else
+#define TCL_DECLARE_MUTEX(name)
+#endif
+
+/*
+ * Tcl's public routine Tcl_FSSeek() uses the values SEEK_SET, SEEK_CUR, and
+ * SEEK_END, all #define'd by stdio.h .
+ *
+ * Also, many extensions need stdio.h, and they've grown accustomed to tcl.h
+ * providing it for them rather than #include-ing it themselves as they
+ * should, so also for their sake, we keep the #include to be consistent with
+ * prior Tcl releases.
+ */
+
+#include <stdio.h>
+
+/*
+ *----------------------------------------------------------------------------
+ * Support for functions with a variable number of arguments.
+ *
+ * The following TCL_VARARGS* macros are to support old extensions
+ * written for older versions of Tcl where the macros permitted
+ * support for the varargs.h system as well as stdarg.h .
+ *
+ * New code should just directly be written to use stdarg.h conventions.
+ */
+
+#include <stdarg.h>
+#ifndef TCL_NO_DEPRECATED
+#    define TCL_VARARGS(type, name) (type name, ...)
+#    define TCL_VARARGS_DEF(type, name) (type name, ...)
+#    define TCL_VARARGS_START(type, name, list) (va_start(list, name), name)
+#endif /* !TCL_NO_DEPRECATED */
+#if defined(__GNUC__) && (__GNUC__ > 2)
+#   define TCL_FORMAT_PRINTF(a,b) __attribute__ ((__format__ (__printf__, a, b)))
+#   define TCL_NORETURN __attribute__ ((noreturn))
+#   define TCL_NOINLINE __attribute__ ((noinline))
+#   if defined(BUILD_tcl) || defined(BUILD_tk)
+#	define TCL_NORETURN1 __attribute__ ((noreturn))
+#   else
+#	define TCL_NORETURN1 /* nothing */
+#   endif
+#else
+#   define TCL_FORMAT_PRINTF(a,b)
+#   if defined(_MSC_VER) && (_MSC_VER >= 1310)
+#	define TCL_NORETURN _declspec(noreturn)
+#	define TCL_NOINLINE __declspec(noinline)
+#   else
+#	define TCL_NORETURN /* nothing */
+#	define TCL_NOINLINE /* nothing */
+#   endif
+#   define TCL_NORETURN1 /* nothing */
+#endif
+
+/*
+ * Allow a part of Tcl's API to be explicitly marked as deprecated.
+ *
+ * Used to make TIP 330/336 generate moans even if people use the
+ * compatibility macros. Change your code, guys! We won't support you forever.
+ */
+
+#if defined(__GNUC__) && ((__GNUC__ >= 4) || ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 1)))
+#   if (__GNUC__ > 4) || ((__GNUC__ == 4) && (__GNUC_MINOR__ >= 5))
+#	define TCL_DEPRECATED_API(msg)	__attribute__ ((__deprecated__ (msg)))
+#   else
+#	define TCL_DEPRECATED_API(msg)	__attribute__ ((__deprecated__))
+#   endif
+#else
+#   define TCL_DEPRECATED_API(msg)	/* nothing portable */
+#endif
+
+/*
+ *----------------------------------------------------------------------------
+ * Macros used to declare a function to be exported by a DLL. Used by Windows,
+ * maps to no-op declarations on non-Windows systems. The default build on
+ * windows is for a DLL, which causes the DLLIMPORT and DLLEXPORT macros to be
+ * nonempty. To build a static library, the macro STATIC_BUILD should be
+ * defined.
+ *
+ * Note: when building static but linking dynamically to MSVCRT we must still
+ *       correctly decorate the C library imported function.  Use CRTIMPORT
+ *       for this purpose.  _DLL is defined by the compiler when linking to
+ *       MSVCRT.
+ */
+
+#if (defined(_WIN32) && (defined(_MSC_VER) || (defined(__BORLANDC__) && (__BORLANDC__ >= 0x0550)) || defined(__LCC__) || defined(__WATCOMC__) || (defined(__GNUC__) && defined(__declspec))))
+#   define HAVE_DECLSPEC 1
+#   ifdef STATIC_BUILD
+#       define DLLIMPORT
+#       define DLLEXPORT
+#       ifdef _DLL
+#           define CRTIMPORT __declspec(dllimport)
+#       else
+#           define CRTIMPORT
+#       endif
+#   else
+#       define DLLIMPORT __declspec(dllimport)
+#       define DLLEXPORT __declspec(dllexport)
+#       define CRTIMPORT __declspec(dllimport)
+#   endif
+#else
+#   define DLLIMPORT
+#   if defined(__GNUC__) && __GNUC__ > 3
+#       define DLLEXPORT __attribute__ ((visibility("default")))
+#   else
+#       define DLLEXPORT
+#   endif
+#   define CRTIMPORT
+#endif
+
+/*
+ * These macros are used to control whether functions are being declared for
+ * import or export. If a function is being declared while it is being built
+ * to be included in a shared library, then it should have the DLLEXPORT
+ * storage class. If is being declared for use by a module that is going to
+ * link against the shared library, then it should have the DLLIMPORT storage
+ * class. If the symbol is beind declared for a static build or for use from a
+ * stub library, then the storage class should be empty.
+ *
+ * The convention is that a macro called BUILD_xxxx, where xxxx is the name of
+ * a library we are building, is set on the compile line for sources that are
+ * to be placed in the library. When this macro is set, the storage class will
+ * be set to DLLEXPORT. At the end of the header file, the storage class will
+ * be reset to DLLIMPORT.
+ */
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/*
+ * The following _ANSI_ARGS_ macro is to support old extensions
+ * written for older versions of Tcl where it permitted support
+ * for compilers written in the pre-prototype era of C.
+ *
+ * New code should use prototypes.
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#   undef _ANSI_ARGS_
+#   define _ANSI_ARGS_(x)	x
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * Definitions that allow this header file to be used either with or without
+ * ANSI C features.
+ */
+
+#ifndef INLINE
+#   define INLINE
+#endif
+
+#ifdef NO_CONST
+#   ifndef const
+#      define const
+#   endif
+#endif
+#ifndef CONST
+#   define CONST const
+#endif
+
+#ifdef USE_NON_CONST
+#   ifdef USE_COMPAT_CONST
+#      error define at most one of USE_NON_CONST and USE_COMPAT_CONST
+#   endif
+#   define CONST84
+#   define CONST84_RETURN
+#else
+#   ifdef USE_COMPAT_CONST
+#      define CONST84
+#      define CONST84_RETURN const
+#   else
+#      define CONST84 const
+#      define CONST84_RETURN const
+#   endif
+#endif
+
+#ifndef CONST86
+#      define CONST86 CONST84
+#endif
+
+/*
+ * Make sure EXTERN isn't defined elsewhere.
+ */
+
+#ifdef EXTERN
+#   undef EXTERN
+#endif /* EXTERN */
+
+#ifdef __cplusplus
+#   define EXTERN extern "C" TCL_STORAGE_CLASS
+#else
+#   define EXTERN extern TCL_STORAGE_CLASS
+#endif
+
+/*
+ *----------------------------------------------------------------------------
+ * The following code is copied from winnt.h. If we don't replicate it here,
+ * then <windows.h> can't be included after tcl.h, since tcl.h also defines
+ * VOID. This block is skipped under Cygwin and Mingw.
+ */
+
+#if defined(_WIN32) && !defined(HAVE_WINNT_IGNORE_VOID)
+#ifndef VOID
+#define VOID void
+typedef char CHAR;
+typedef short SHORT;
+typedef long LONG;
+#endif
+#endif /* _WIN32 && !HAVE_WINNT_IGNORE_VOID */
+
+/*
+ * Macro to use instead of "void" for arguments that must have type "void *"
+ * in ANSI C; maps them to type "char *" in non-ANSI systems.
+ */
+
+#ifndef __VXWORKS__
+#   ifndef NO_VOID
+#	define VOID void
+#   else
+#	define VOID char
+#   endif
+#endif
+
+/*
+ * Miscellaneous declarations.
+ */
+
+#ifndef _CLIENTDATA
+#   ifndef NO_VOID
+	typedef void *ClientData;
+#   else
+	typedef int *ClientData;
+#   endif
+#   define _CLIENTDATA
+#endif
+
+/*
+ * Darwin specific configure overrides (to support fat compiles, where
+ * configure runs only once for multiple architectures):
+ */
+
+#ifdef __APPLE__
+#   ifdef __LP64__
+#	undef TCL_WIDE_INT_TYPE
+#	define TCL_WIDE_INT_IS_LONG 1
+#	define TCL_CFG_DO64BIT 1
+#    else /* !__LP64__ */
+#	define TCL_WIDE_INT_TYPE long long
+#	undef TCL_WIDE_INT_IS_LONG
+#	undef TCL_CFG_DO64BIT
+#    endif /* __LP64__ */
+#    undef HAVE_STRUCT_STAT64
+#endif /* __APPLE__ */
+
+/*
+ * Define Tcl_WideInt to be a type that is (at least) 64-bits wide, and define
+ * Tcl_WideUInt to be the unsigned variant of that type (assuming that where
+ * we have one, we can have the other.)
+ *
+ * Also defines the following macros:
+ * TCL_WIDE_INT_IS_LONG - if wide ints are really longs (i.e. we're on a
+ *	LP64 system such as modern Solaris or Linux ... not including Win64)
+ * Tcl_WideAsLong - forgetful converter from wideInt to long.
+ * Tcl_LongAsWide - sign-extending converter from long to wideInt.
+ * Tcl_WideAsDouble - converter from wideInt to double.
+ * Tcl_DoubleAsWide - converter from double to wideInt.
+ *
+ * The following invariant should hold for any long value 'longVal':
+ *	longVal == Tcl_WideAsLong(Tcl_LongAsWide(longVal))
+ *
+ * Note on converting between Tcl_WideInt and strings. This implementation (in
+ * tclObj.c) depends on the function
+ * sprintf(...,"%" TCL_LL_MODIFIER "d",...).
+ */
+
+#if !defined(TCL_WIDE_INT_TYPE)&&!defined(TCL_WIDE_INT_IS_LONG)
+#   if defined(_WIN32)
+#      define TCL_WIDE_INT_TYPE __int64
+#      define TCL_LL_MODIFIER	"I64"
+#   elif defined(__GNUC__)
+#      define TCL_WIDE_INT_TYPE long long
+#      define TCL_LL_MODIFIER	"ll"
+#   else /* ! _WIN32 && ! __GNUC__ */
+/*
+ * Don't know what platform it is and configure hasn't discovered what is
+ * going on for us. Try to guess...
+ */
+#      include <limits.h>
+#      if (INT_MAX < LONG_MAX)
+#         define TCL_WIDE_INT_IS_LONG	1
+#      else
+#         define TCL_WIDE_INT_TYPE long long
+#      endif
+#   endif /* _WIN32 */
+#endif /* !TCL_WIDE_INT_TYPE & !TCL_WIDE_INT_IS_LONG */
+#ifdef TCL_WIDE_INT_IS_LONG
+#   undef TCL_WIDE_INT_TYPE
+#   define TCL_WIDE_INT_TYPE	long
+#endif /* TCL_WIDE_INT_IS_LONG */
+
+typedef TCL_WIDE_INT_TYPE		Tcl_WideInt;
+typedef unsigned TCL_WIDE_INT_TYPE	Tcl_WideUInt;
+
+#ifdef TCL_WIDE_INT_IS_LONG
+#   ifndef TCL_LL_MODIFIER
+#      define TCL_LL_MODIFIER		"l"
+#   endif /* !TCL_LL_MODIFIER */
+#else /* TCL_WIDE_INT_IS_LONG */
+/*
+ * The next short section of defines are only done when not running on Windows
+ * or some other strange platform.
+ */
+#   ifndef TCL_LL_MODIFIER
+#      define TCL_LL_MODIFIER		"ll"
+#   endif /* !TCL_LL_MODIFIER */
+#endif /* TCL_WIDE_INT_IS_LONG */
+
+#define Tcl_WideAsLong(val)	((long)((Tcl_WideInt)(val)))
+#define Tcl_LongAsWide(val)	((Tcl_WideInt)((long)(val)))
+#define Tcl_WideAsDouble(val)	((double)((Tcl_WideInt)(val)))
+#define Tcl_DoubleAsWide(val)	((Tcl_WideInt)((double)(val)))
+
+#if defined(_WIN32)
+#   ifdef __BORLANDC__
+	typedef struct stati64 Tcl_StatBuf;
+#   elif defined(_WIN64)
+	typedef struct __stat64 Tcl_StatBuf;
+#   elif (defined(_MSC_VER) && (_MSC_VER < 1400)) || defined(_USE_32BIT_TIME_T)
+	typedef struct _stati64	Tcl_StatBuf;
+#   else
+	typedef struct _stat32i64 Tcl_StatBuf;
+#   endif /* _MSC_VER < 1400 */
+#elif defined(__CYGWIN__)
+    typedef struct {
+	dev_t st_dev;
+	unsigned short st_ino;
+	unsigned short st_mode;
+	short st_nlink;
+	short st_uid;
+	short st_gid;
+	/* Here is a 2-byte gap */
+	dev_t st_rdev;
+	/* Here is a 4-byte gap */
+	long long st_size;
+	struct {long tv_sec;} st_atim;
+	struct {long tv_sec;} st_mtim;
+	struct {long tv_sec;} st_ctim;
+	/* Here is a 4-byte gap */
+    } Tcl_StatBuf;
+#elif defined(HAVE_STRUCT_STAT64) && !defined(__APPLE__)
+    typedef struct stat64 Tcl_StatBuf;
+#else
+    typedef struct stat Tcl_StatBuf;
+#endif
+
+/*
+ *----------------------------------------------------------------------------
+ * Data structures defined opaquely in this module. The definitions below just
+ * provide dummy types. A few fields are made visible in Tcl_Interp
+ * structures, namely those used for returning a string result from commands.
+ * Direct access to the result field is discouraged in Tcl 8.0. The
+ * interpreter result is either an object or a string, and the two values are
+ * kept consistent unless some C code sets interp->result directly.
+ * Programmers should use either the function Tcl_GetObjResult() or
+ * Tcl_GetStringResult() to read the interpreter's result. See the SetResult
+ * man page for details.
+ *
+ * Note: any change to the Tcl_Interp definition below must be mirrored in the
+ * "real" definition in tclInt.h.
+ *
+ * Note: Tcl_ObjCmdProc functions do not directly set result and freeProc.
+ * Instead, they set a Tcl_Obj member in the "real" structure that can be
+ * accessed with Tcl_GetObjResult() and Tcl_SetObjResult().
+ */
+
+typedef struct Tcl_Interp
+#ifndef TCL_NO_DEPRECATED
+{
+    /* TIP #330: Strongly discourage extensions from using the string
+     * result. */
+#ifdef USE_INTERP_RESULT
+    char *result TCL_DEPRECATED_API("use Tcl_GetStringResult/Tcl_SetResult");
+				/* If the last command returned a string
+				 * result, this points to it. */
+    void (*freeProc) (char *blockPtr)
+	    TCL_DEPRECATED_API("use Tcl_GetStringResult/Tcl_SetResult");
+				/* Zero means the string result is statically
+				 * allocated. TCL_DYNAMIC means it was
+				 * allocated with ckalloc and should be freed
+				 * with ckfree. Other values give the address
+				 * of function to invoke to free the result.
+				 * Tcl_Eval must free it before executing next
+				 * command. */
+#else
+    char *resultDontUse; /* Don't use in extensions! */
+    void (*freeProcDontUse) (char *); /* Don't use in extensions! */
+#endif
+#ifdef USE_INTERP_ERRORLINE
+    int errorLine TCL_DEPRECATED_API("use Tcl_GetErrorLine/Tcl_SetErrorLine");
+				/* When TCL_ERROR is returned, this gives the
+				 * line number within the command where the
+				 * error occurred (1 if first line). */
+#else
+    int errorLineDontUse; /* Don't use in extensions! */
+#endif
+}
+#endif /* !TCL_NO_DEPRECATED */
+Tcl_Interp;
+
+typedef struct Tcl_AsyncHandler_ *Tcl_AsyncHandler;
+typedef struct Tcl_Channel_ *Tcl_Channel;
+typedef struct Tcl_ChannelTypeVersion_ *Tcl_ChannelTypeVersion;
+typedef struct Tcl_Command_ *Tcl_Command;
+typedef struct Tcl_Condition_ *Tcl_Condition;
+typedef struct Tcl_Dict_ *Tcl_Dict;
+typedef struct Tcl_EncodingState_ *Tcl_EncodingState;
+typedef struct Tcl_Encoding_ *Tcl_Encoding;
+typedef struct Tcl_Event Tcl_Event;
+typedef struct Tcl_InterpState_ *Tcl_InterpState;
+typedef struct Tcl_LoadHandle_ *Tcl_LoadHandle;
+typedef struct Tcl_Mutex_ *Tcl_Mutex;
+typedef struct Tcl_Pid_ *Tcl_Pid;
+typedef struct Tcl_RegExp_ *Tcl_RegExp;
+typedef struct Tcl_ThreadDataKey_ *Tcl_ThreadDataKey;
+typedef struct Tcl_ThreadId_ *Tcl_ThreadId;
+typedef struct Tcl_TimerToken_ *Tcl_TimerToken;
+typedef struct Tcl_Trace_ *Tcl_Trace;
+typedef struct Tcl_Var_ *Tcl_Var;
+typedef struct Tcl_ZLibStream_ *Tcl_ZlibStream;
+
+/*
+ *----------------------------------------------------------------------------
+ * Definition of the interface to functions implementing threads. A function
+ * following this definition is given to each call of 'Tcl_CreateThread' and
+ * will be called as the main fuction of the new thread created by that call.
+ */
+
+#if defined _WIN32
+typedef unsigned (__stdcall Tcl_ThreadCreateProc) (ClientData clientData);
+#else
+typedef void (Tcl_ThreadCreateProc) (ClientData clientData);
+#endif
+
+/*
+ * Threading function return types used for abstracting away platform
+ * differences when writing a Tcl_ThreadCreateProc. See the NewThread function
+ * in generic/tclThreadTest.c for it's usage.
+ */
+
+#if defined _WIN32
+#   define Tcl_ThreadCreateType		unsigned __stdcall
+#   define TCL_THREAD_CREATE_RETURN	return 0
+#else
+#   define Tcl_ThreadCreateType		void
+#   define TCL_THREAD_CREATE_RETURN
+#endif
+
+/*
+ * Definition of values for default stacksize and the possible flags to be
+ * given to Tcl_CreateThread.
+ */
+
+#define TCL_THREAD_STACK_DEFAULT (0)    /* Use default size for stack. */
+#define TCL_THREAD_NOFLAGS	 (0000) /* Standard flags, default
+					 * behaviour. */
+#define TCL_THREAD_JOINABLE	 (0001) /* Mark the thread as joinable. */
+
+/*
+ * Flag values passed to Tcl_StringCaseMatch.
+ */
+
+#define TCL_MATCH_NOCASE	(1<<0)
+
+/*
+ * Flag values passed to Tcl_GetRegExpFromObj.
+ */
+
+#define	TCL_REG_BASIC		000000	/* BREs (convenience). */
+#define	TCL_REG_EXTENDED	000001	/* EREs. */
+#define	TCL_REG_ADVF		000002	/* Advanced features in EREs. */
+#define	TCL_REG_ADVANCED	000003	/* AREs (which are also EREs). */
+#define	TCL_REG_QUOTE		000004	/* No special characters, none. */
+#define	TCL_REG_NOCASE		000010	/* Ignore case. */
+#define	TCL_REG_NOSUB		000020	/* Don't care about subexpressions. */
+#define	TCL_REG_EXPANDED	000040	/* Expanded format, white space &
+					 * comments. */
+#define	TCL_REG_NLSTOP		000100  /* \n doesn't match . or [^ ] */
+#define	TCL_REG_NLANCH		000200  /* ^ matches after \n, $ before. */
+#define	TCL_REG_NEWLINE		000300  /* Newlines are line terminators. */
+#define	TCL_REG_CANMATCH	001000  /* Report details on partial/limited
+					 * matches. */
+
+/*
+ * Flags values passed to Tcl_RegExpExecObj.
+ */
+
+#define	TCL_REG_NOTBOL	0001	/* Beginning of string does not match ^.  */
+#define	TCL_REG_NOTEOL	0002	/* End of string does not match $. */
+
+/*
+ * Structures filled in by Tcl_RegExpInfo. Note that all offset values are
+ * relative to the start of the match string, not the beginning of the entire
+ * string.
+ */
+
+typedef struct Tcl_RegExpIndices {
+    long start;			/* Character offset of first character in
+				 * match. */
+    long end;			/* Character offset of first character after
+				 * the match. */
+} Tcl_RegExpIndices;
+
+typedef struct Tcl_RegExpInfo {
+    int nsubs;			/* Number of subexpressions in the compiled
+				 * expression. */
+    Tcl_RegExpIndices *matches;	/* Array of nsubs match offset pairs. */
+    long extendStart;		/* The offset at which a subsequent match
+				 * might begin. */
+    long reserved;		/* Reserved for later use. */
+} Tcl_RegExpInfo;
+
+/*
+ * Picky compilers complain if this typdef doesn't appear before the struct's
+ * reference in tclDecls.h.
+ */
+
+typedef Tcl_StatBuf *Tcl_Stat_;
+typedef struct stat *Tcl_OldStat_;
+
+/*
+ *----------------------------------------------------------------------------
+ * When a TCL command returns, the interpreter contains a result from the
+ * command. Programmers are strongly encouraged to use one of the functions
+ * Tcl_GetObjResult() or Tcl_GetStringResult() to read the interpreter's
+ * result. See the SetResult man page for details. Besides this result, the
+ * command function returns an integer code, which is one of the following:
+ *
+ * TCL_OK		Command completed normally; the interpreter's result
+ *			contains the command's result.
+ * TCL_ERROR		The command couldn't be completed successfully; the
+ *			interpreter's result describes what went wrong.
+ * TCL_RETURN		The command requests that the current function return;
+ *			the interpreter's result contains the function's
+ *			return value.
+ * TCL_BREAK		The command requests that the innermost loop be
+ *			exited; the interpreter's result is meaningless.
+ * TCL_CONTINUE		Go on to the next iteration of the current loop; the
+ *			interpreter's result is meaningless.
+ */
+
+#define TCL_OK			0
+#define TCL_ERROR		1
+#define TCL_RETURN		2
+#define TCL_BREAK		3
+#define TCL_CONTINUE		4
+
+#define TCL_RESULT_SIZE		200
+
+/*
+ *----------------------------------------------------------------------------
+ * Flags to control what substitutions are performed by Tcl_SubstObj():
+ */
+
+#define TCL_SUBST_COMMANDS	001
+#define TCL_SUBST_VARIABLES	002
+#define TCL_SUBST_BACKSLASHES	004
+#define TCL_SUBST_ALL		007
+
+/*
+ * Argument descriptors for math function callbacks in expressions:
+ */
+
+typedef enum {
+    TCL_INT, TCL_DOUBLE, TCL_EITHER, TCL_WIDE_INT
+} Tcl_ValueType;
+
+typedef struct Tcl_Value {
+    Tcl_ValueType type;		/* Indicates intValue or doubleValue is valid,
+				 * or both. */
+    long intValue;		/* Integer value. */
+    double doubleValue;		/* Double-precision floating value. */
+    Tcl_WideInt wideValue;	/* Wide (min. 64-bit) integer value. */
+} Tcl_Value;
+
+/*
+ * Forward declaration of Tcl_Obj to prevent an error when the forward
+ * reference to Tcl_Obj is encountered in the function types declared below.
+ */
+
+struct Tcl_Obj;
+
+/*
+ *----------------------------------------------------------------------------
+ * Function types defined by Tcl:
+ */
+
+typedef int (Tcl_AppInitProc) (Tcl_Interp *interp);
+typedef int (Tcl_AsyncProc) (ClientData clientData, Tcl_Interp *interp,
+	int code);
+typedef void (Tcl_ChannelProc) (ClientData clientData, int mask);
+typedef void (Tcl_CloseProc) (ClientData data);
+typedef void (Tcl_CmdDeleteProc) (ClientData clientData);
+typedef int (Tcl_CmdProc) (ClientData clientData, Tcl_Interp *interp,
+	int argc, CONST84 char *argv[]);
+typedef void (Tcl_CmdTraceProc) (ClientData clientData, Tcl_Interp *interp,
+	int level, char *command, Tcl_CmdProc *proc,
+	ClientData cmdClientData, int argc, CONST84 char *argv[]);
+typedef int (Tcl_CmdObjTraceProc) (ClientData clientData, Tcl_Interp *interp,
+	int level, const char *command, Tcl_Command commandInfo, int objc,
+	struct Tcl_Obj *const *objv);
+typedef void (Tcl_CmdObjTraceDeleteProc) (ClientData clientData);
+typedef void (Tcl_DupInternalRepProc) (struct Tcl_Obj *srcPtr,
+	struct Tcl_Obj *dupPtr);
+typedef int (Tcl_EncodingConvertProc) (ClientData clientData, const char *src,
+	int srcLen, int flags, Tcl_EncodingState *statePtr, char *dst,
+	int dstLen, int *srcReadPtr, int *dstWrotePtr, int *dstCharsPtr);
+typedef void (Tcl_EncodingFreeProc) (ClientData clientData);
+typedef int (Tcl_EventProc) (Tcl_Event *evPtr, int flags);
+typedef void (Tcl_EventCheckProc) (ClientData clientData, int flags);
+typedef int (Tcl_EventDeleteProc) (Tcl_Event *evPtr, ClientData clientData);
+typedef void (Tcl_EventSetupProc) (ClientData clientData, int flags);
+typedef void (Tcl_ExitProc) (ClientData clientData);
+typedef void (Tcl_FileProc) (ClientData clientData, int mask);
+typedef void (Tcl_FileFreeProc) (ClientData clientData);
+typedef void (Tcl_FreeInternalRepProc) (struct Tcl_Obj *objPtr);
+typedef void (Tcl_FreeProc) (char *blockPtr);
+typedef void (Tcl_IdleProc) (ClientData clientData);
+typedef void (Tcl_InterpDeleteProc) (ClientData clientData,
+	Tcl_Interp *interp);
+typedef int (Tcl_MathProc) (ClientData clientData, Tcl_Interp *interp,
+	Tcl_Value *args, Tcl_Value *resultPtr);
+typedef void (Tcl_NamespaceDeleteProc) (ClientData clientData);
+typedef int (Tcl_ObjCmdProc) (ClientData clientData, Tcl_Interp *interp,
+	int objc, struct Tcl_Obj *const *objv);
+typedef int (Tcl_PackageInitProc) (Tcl_Interp *interp);
+typedef int (Tcl_PackageUnloadProc) (Tcl_Interp *interp, int flags);
+typedef void (Tcl_PanicProc) (const char *format, ...);
+typedef void (Tcl_TcpAcceptProc) (ClientData callbackData, Tcl_Channel chan,
+	char *address, int port);
+typedef void (Tcl_TimerProc) (ClientData clientData);
+typedef int (Tcl_SetFromAnyProc) (Tcl_Interp *interp, struct Tcl_Obj *objPtr);
+typedef void (Tcl_UpdateStringProc) (struct Tcl_Obj *objPtr);
+typedef char * (Tcl_VarTraceProc) (ClientData clientData, Tcl_Interp *interp,
+	CONST84 char *part1, CONST84 char *part2, int flags);
+typedef void (Tcl_CommandTraceProc) (ClientData clientData, Tcl_Interp *interp,
+	const char *oldName, const char *newName, int flags);
+typedef void (Tcl_CreateFileHandlerProc) (int fd, int mask, Tcl_FileProc *proc,
+	ClientData clientData);
+typedef void (Tcl_DeleteFileHandlerProc) (int fd);
+typedef void (Tcl_AlertNotifierProc) (ClientData clientData);
+typedef void (Tcl_ServiceModeHookProc) (int mode);
+typedef ClientData (Tcl_InitNotifierProc) (void);
+typedef void (Tcl_FinalizeNotifierProc) (ClientData clientData);
+typedef void (Tcl_MainLoopProc) (void);
+
+/*
+ *----------------------------------------------------------------------------
+ * The following structure represents a type of object, which is a particular
+ * internal representation for an object plus a set of functions that provide
+ * standard operations on objects of that type.
+ */
+
+typedef struct Tcl_ObjType {
+    const char *name;		/* Name of the type, e.g. "int". */
+    Tcl_FreeInternalRepProc *freeIntRepProc;
+				/* Called to free any storage for the type's
+				 * internal rep. NULL if the internal rep does
+				 * not need freeing. */
+    Tcl_DupInternalRepProc *dupIntRepProc;
+				/* Called to create a new object as a copy of
+				 * an existing object. */
+    Tcl_UpdateStringProc *updateStringProc;
+				/* Called to update the string rep from the
+				 * type's internal representation. */
+    Tcl_SetFromAnyProc *setFromAnyProc;
+				/* Called to convert the object's internal rep
+				 * to this type. Frees the internal rep of the
+				 * old type. Returns TCL_ERROR on failure. */
+} Tcl_ObjType;
+
+/*
+ * One of the following structures exists for each object in the Tcl system.
+ * An object stores a value as either a string, some internal representation,
+ * or both.
+ */
+
+typedef struct Tcl_Obj {
+    int refCount;		/* When 0 the object will be freed. */
+    char *bytes;		/* This points to the first byte of the
+				 * object's string representation. The array
+				 * must be followed by a null byte (i.e., at
+				 * offset length) but may also contain
+				 * embedded null characters. The array's
+				 * storage is allocated by ckalloc. NULL means
+				 * the string rep is invalid and must be
+				 * regenerated from the internal rep.  Clients
+				 * should use Tcl_GetStringFromObj or
+				 * Tcl_GetString to get a pointer to the byte
+				 * array as a readonly value. */
+    int length;			/* The number of bytes at *bytes, not
+				 * including the terminating null. */
+    const Tcl_ObjType *typePtr;	/* Denotes the object's type. Always
+				 * corresponds to the type of the object's
+				 * internal rep. NULL indicates the object has
+				 * no internal rep (has no type). */
+    union {			/* The internal representation: */
+	long longValue;		/*   - an long integer value. */
+	double doubleValue;	/*   - a double-precision floating value. */
+	void *otherValuePtr;	/*   - another, type-specific value, not used
+				 *     internally any more. */
+	Tcl_WideInt wideValue;	/*   - a long long value. */
+	struct {		/*   - internal rep as two pointers.
+				 *     Many uses in Tcl, including a bignum's
+				 *     tightly packed fields, where the alloc,
+				 *     used and signum flags are packed into
+				 *     ptr2 with everything else hung off
+				 *     ptr1. */
+	    void *ptr1;
+	    void *ptr2;
+	} twoPtrValue;
+	struct {		/*   - internal rep as a pointer and a long,
+				 *     not used internally any more. */
+	    void *ptr;
+	    unsigned long value;
+	} ptrAndLongRep;
+    } internalRep;
+} Tcl_Obj;
+
+/*
+ * Macros to increment and decrement a Tcl_Obj's reference count, and to test
+ * whether an object is shared (i.e. has reference count > 1). Note: clients
+ * should use Tcl_DecrRefCount() when they are finished using an object, and
+ * should never call TclFreeObj() directly. TclFreeObj() is only defined and
+ * made public in tcl.h to support Tcl_DecrRefCount's macro definition.
+ */
+
+void		Tcl_IncrRefCount(Tcl_Obj *objPtr);
+void		Tcl_DecrRefCount(Tcl_Obj *objPtr);
+int		Tcl_IsShared(Tcl_Obj *objPtr);
+
+/*
+ *----------------------------------------------------------------------------
+ * The following structure contains the state needed by Tcl_SaveResult. No-one
+ * outside of Tcl should access any of these fields. This structure is
+ * typically allocated on the stack.
+ */
+
+typedef struct Tcl_SavedResult {
+    char *result;
+    Tcl_FreeProc *freeProc;
+    Tcl_Obj *objResultPtr;
+    char *appendResult;
+    int appendAvl;
+    int appendUsed;
+    char resultSpace[TCL_RESULT_SIZE+1];
+} Tcl_SavedResult;
+
+/*
+ *----------------------------------------------------------------------------
+ * The following definitions support Tcl's namespace facility. Note: the first
+ * five fields must match exactly the fields in a Namespace structure (see
+ * tclInt.h).
+ */
+
+typedef struct Tcl_Namespace {
+    char *name;			/* The namespace's name within its parent
+				 * namespace. This contains no ::'s. The name
+				 * of the global namespace is "" although "::"
+				 * is an synonym. */
+    char *fullName;		/* The namespace's fully qualified name. This
+				 * starts with ::. */
+    ClientData clientData;	/* Arbitrary value associated with this
+				 * namespace. */
+    Tcl_NamespaceDeleteProc *deleteProc;
+				/* Function invoked when deleting the
+				 * namespace to, e.g., free clientData. */
+    struct Tcl_Namespace *parentPtr;
+				/* Points to the namespace that contains this
+				 * one. NULL if this is the global
+				 * namespace. */
+} Tcl_Namespace;
+
+/*
+ *----------------------------------------------------------------------------
+ * The following structure represents a call frame, or activation record. A
+ * call frame defines a naming context for a procedure call: its local scope
+ * (for local variables) and its namespace scope (used for non-local
+ * variables; often the global :: namespace). A call frame can also define the
+ * naming context for a namespace eval or namespace inscope command: the
+ * namespace in which the command's code should execute. The Tcl_CallFrame
+ * structures exist only while procedures or namespace eval/inscope's are
+ * being executed, and provide a Tcl call stack.
+ *
+ * A call frame is initialized and pushed using Tcl_PushCallFrame and popped
+ * using Tcl_PopCallFrame. Storage for a Tcl_CallFrame must be provided by the
+ * Tcl_PushCallFrame caller, and callers typically allocate them on the C call
+ * stack for efficiency. For this reason, Tcl_CallFrame is defined as a
+ * structure and not as an opaque token. However, most Tcl_CallFrame fields
+ * are hidden since applications should not access them directly; others are
+ * declared as "dummyX".
+ *
+ * WARNING!! The structure definition must be kept consistent with the
+ * CallFrame structure in tclInt.h. If you change one, change the other.
+ */
+
+typedef struct Tcl_CallFrame {
+    Tcl_Namespace *nsPtr;
+    int dummy1;
+    int dummy2;
+    void *dummy3;
+    void *dummy4;
+    void *dummy5;
+    int dummy6;
+    void *dummy7;
+    void *dummy8;
+    int dummy9;
+    void *dummy10;
+    void *dummy11;
+    void *dummy12;
+    void *dummy13;
+} Tcl_CallFrame;
+
+/*
+ *----------------------------------------------------------------------------
+ * Information about commands that is returned by Tcl_GetCommandInfo and
+ * passed to Tcl_SetCommandInfo. objProc is an objc/objv object-based command
+ * function while proc is a traditional Tcl argc/argv string-based function.
+ * Tcl_CreateObjCommand and Tcl_CreateCommand ensure that both objProc and
+ * proc are non-NULL and can be called to execute the command. However, it may
+ * be faster to call one instead of the other. The member isNativeObjectProc
+ * is set to 1 if an object-based function was registered by
+ * Tcl_CreateObjCommand, and to 0 if a string-based function was registered by
+ * Tcl_CreateCommand. The other function is typically set to a compatibility
+ * wrapper that does string-to-object or object-to-string argument conversions
+ * then calls the other function.
+ */
+
+typedef struct Tcl_CmdInfo {
+    int isNativeObjectProc;	/* 1 if objProc was registered by a call to
+				 * Tcl_CreateObjCommand; 0 otherwise.
+				 * Tcl_SetCmdInfo does not modify this
+				 * field. */
+    Tcl_ObjCmdProc *objProc;	/* Command's object-based function. */
+    ClientData objClientData;	/* ClientData for object proc. */
+    Tcl_CmdProc *proc;		/* Command's string-based function. */
+    ClientData clientData;	/* ClientData for string proc. */
+    Tcl_CmdDeleteProc *deleteProc;
+				/* Function to call when command is
+				 * deleted. */
+    ClientData deleteData;	/* Value to pass to deleteProc (usually the
+				 * same as clientData). */
+    Tcl_Namespace *namespacePtr;/* Points to the namespace that contains this
+				 * command. Note that Tcl_SetCmdInfo will not
+				 * change a command's namespace; use
+				 * TclRenameCommand or Tcl_Eval (of 'rename')
+				 * to do that. */
+} Tcl_CmdInfo;
+
+/*
+ *----------------------------------------------------------------------------
+ * The structure defined below is used to hold dynamic strings. The only
+ * fields that clients should use are string and length, accessible via the
+ * macros Tcl_DStringValue and Tcl_DStringLength.
+ */
+
+#define TCL_DSTRING_STATIC_SIZE 200
+typedef struct Tcl_DString {
+    char *string;		/* Points to beginning of string: either
+				 * staticSpace below or a malloced array. */
+    int length;			/* Number of non-NULL characters in the
+				 * string. */
+    int spaceAvl;		/* Total number of bytes available for the
+				 * string and its terminating NULL char. */
+    char staticSpace[TCL_DSTRING_STATIC_SIZE];
+				/* Space to use in common case where string is
+				 * small. */
+} Tcl_DString;
+
+#define Tcl_DStringLength(dsPtr) ((dsPtr)->length)
+#define Tcl_DStringValue(dsPtr) ((dsPtr)->string)
+#ifndef TCL_NO_DEPRECATED
+#   define Tcl_DStringTrunc Tcl_DStringSetLength
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * Definitions for the maximum number of digits of precision that may be
+ * specified in the "tcl_precision" variable, and the number of bytes of
+ * buffer space required by Tcl_PrintDouble.
+ */
+
+#define TCL_MAX_PREC		17
+#define TCL_DOUBLE_SPACE	(TCL_MAX_PREC+10)
+
+/*
+ * Definition for a number of bytes of buffer space sufficient to hold the
+ * string representation of an integer in base 10 (assuming the existence of
+ * 64-bit integers).
+ */
+
+#define TCL_INTEGER_SPACE	24
+
+/*
+ * Flag values passed to Tcl_ConvertElement.
+ * TCL_DONT_USE_BRACES forces it not to enclose the element in braces, but to
+ *	use backslash quoting instead.
+ * TCL_DONT_QUOTE_HASH disables the default quoting of the '#' character. It
+ *	is safe to leave the hash unquoted when the element is not the first
+ *	element of a list, and this flag can be used by the caller to indicate
+ *	that condition.
+ */
+
+#define TCL_DONT_USE_BRACES	1
+#define TCL_DONT_QUOTE_HASH	8
+
+/*
+ * Flag that may be passed to Tcl_GetIndexFromObj to force it to disallow
+ * abbreviated strings.
+ */
+
+#define TCL_EXACT	1
+
+/*
+ *----------------------------------------------------------------------------
+ * Flag values passed to Tcl_RecordAndEval, Tcl_EvalObj, Tcl_EvalObjv.
+ * WARNING: these bit choices must not conflict with the bit choices for
+ * evalFlag bits in tclInt.h!
+ *
+ * Meanings:
+ *	TCL_NO_EVAL:		Just record this command
+ *	TCL_EVAL_GLOBAL:	Execute script in global namespace
+ *	TCL_EVAL_DIRECT:	Do not compile this script
+ *	TCL_EVAL_INVOKE:	Magical Tcl_EvalObjv mode for aliases/ensembles
+ *				o Run in iPtr->lookupNsPtr or global namespace
+ *				o Cut out of error traces
+ *				o Don't reset the flags controlling ensemble
+ *				  error message rewriting.
+ *	TCL_CANCEL_UNWIND:	Magical Tcl_CancelEval mode that causes the
+ *				stack for the script in progress to be
+ *				completely unwound.
+ *	TCL_EVAL_NOERR:	Do no exception reporting at all, just return
+ *				as the caller will report.
+ */
+
+#define TCL_NO_EVAL		0x010000
+#define TCL_EVAL_GLOBAL		0x020000
+#define TCL_EVAL_DIRECT		0x040000
+#define TCL_EVAL_INVOKE		0x080000
+#define TCL_CANCEL_UNWIND	0x100000
+#define TCL_EVAL_NOERR          0x200000
+
+/*
+ * Special freeProc values that may be passed to Tcl_SetResult (see the man
+ * page for details):
+ */
+
+#define TCL_VOLATILE		((Tcl_FreeProc *) 1)
+#define TCL_STATIC		((Tcl_FreeProc *) 0)
+#define TCL_DYNAMIC		((Tcl_FreeProc *) 3)
+
+/*
+ * Flag values passed to variable-related functions.
+ * WARNING: these bit choices must not conflict with the bit choice for
+ * TCL_CANCEL_UNWIND, above.
+ */
+
+#define TCL_GLOBAL_ONLY		 1
+#define TCL_NAMESPACE_ONLY	 2
+#define TCL_APPEND_VALUE	 4
+#define TCL_LIST_ELEMENT	 8
+#define TCL_TRACE_READS		 0x10
+#define TCL_TRACE_WRITES	 0x20
+#define TCL_TRACE_UNSETS	 0x40
+#define TCL_TRACE_DESTROYED	 0x80
+#define TCL_INTERP_DESTROYED	 0x100
+#define TCL_LEAVE_ERR_MSG	 0x200
+#define TCL_TRACE_ARRAY		 0x800
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+/* Required to support old variable/vdelete/vinfo traces. */
+#define TCL_TRACE_OLD_STYLE	 0x1000
+#endif
+/* Indicate the semantics of the result of a trace. */
+#define TCL_TRACE_RESULT_DYNAMIC 0x8000
+#define TCL_TRACE_RESULT_OBJECT  0x10000
+
+/*
+ * Flag values for ensemble commands.
+ */
+
+#define TCL_ENSEMBLE_PREFIX 0x02/* Flag value to say whether to allow
+				 * unambiguous prefixes of commands or to
+				 * require exact matches for command names. */
+
+/*
+ * Flag values passed to command-related functions.
+ */
+
+#define TCL_TRACE_RENAME	0x2000
+#define TCL_TRACE_DELETE	0x4000
+
+#define TCL_ALLOW_INLINE_COMPILATION 0x20000
+
+/*
+ * The TCL_PARSE_PART1 flag is deprecated and has no effect. The part1 is now
+ * always parsed whenever the part2 is NULL. (This is to avoid a common error
+ * when converting code to use the new object based APIs and forgetting to
+ * give the flag)
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#   define TCL_PARSE_PART1	0x400
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * Types for linked variables:
+ */
+
+#define TCL_LINK_INT		1
+#define TCL_LINK_DOUBLE		2
+#define TCL_LINK_BOOLEAN	3
+#define TCL_LINK_STRING		4
+#define TCL_LINK_WIDE_INT	5
+#define TCL_LINK_CHAR		6
+#define TCL_LINK_UCHAR		7
+#define TCL_LINK_SHORT		8
+#define TCL_LINK_USHORT		9
+#define TCL_LINK_UINT		10
+#if defined(TCL_WIDE_INT_IS_LONG) || defined(_WIN32) || defined(__CYGWIN__)
+#define TCL_LINK_LONG		((sizeof(long) != sizeof(int)) ? TCL_LINK_WIDE_INT : TCL_LINK_INT)
+#define TCL_LINK_ULONG		((sizeof(long) != sizeof(int)) ? TCL_LINK_WIDE_UINT : TCL_LINK_UINT)
+#else
+#define TCL_LINK_LONG		11
+#define TCL_LINK_ULONG		12
+#endif
+#define TCL_LINK_FLOAT		13
+#define TCL_LINK_WIDE_UINT	14
+#define TCL_LINK_READ_ONLY	0x80
+
+/*
+ *----------------------------------------------------------------------------
+ * Forward declarations of Tcl_HashTable and related types.
+ */
+
+#ifndef TCL_HASH_TYPE
+#  define TCL_HASH_TYPE unsigned
+#endif
+
+typedef struct Tcl_HashKeyType Tcl_HashKeyType;
+typedef struct Tcl_HashTable Tcl_HashTable;
+typedef struct Tcl_HashEntry Tcl_HashEntry;
+
+typedef TCL_HASH_TYPE (Tcl_HashKeyProc) (Tcl_HashTable *tablePtr, void *keyPtr);
+typedef int (Tcl_CompareHashKeysProc) (void *keyPtr, Tcl_HashEntry *hPtr);
+typedef Tcl_HashEntry * (Tcl_AllocHashEntryProc) (Tcl_HashTable *tablePtr,
+	void *keyPtr);
+typedef void (Tcl_FreeHashEntryProc) (Tcl_HashEntry *hPtr);
+
+/*
+ * Structure definition for an entry in a hash table. No-one outside Tcl
+ * should access any of these fields directly; use the macros defined below.
+ */
+
+struct Tcl_HashEntry {
+    Tcl_HashEntry *nextPtr;	/* Pointer to next entry in this hash bucket,
+				 * or NULL for end of chain. */
+    Tcl_HashTable *tablePtr;	/* Pointer to table containing entry. */
+    void *hash;			/* Hash value, stored as pointer to ensure
+				 * that the offsets of the fields in this
+				 * structure are not changed. */
+    ClientData clientData;	/* Application stores something here with
+				 * Tcl_SetHashValue. */
+    union {			/* Key has one of these forms: */
+	char *oneWordValue;	/* One-word value for key. */
+	Tcl_Obj *objPtr;	/* Tcl_Obj * key value. */
+	int words[1];		/* Multiple integer words for key. The actual
+				 * size will be as large as necessary for this
+				 * table's keys. */
+	char string[1];		/* String for key. The actual size will be as
+				 * large as needed to hold the key. */
+    } key;			/* MUST BE LAST FIELD IN RECORD!! */
+};
+
+/*
+ * Flags used in Tcl_HashKeyType.
+ *
+ * TCL_HASH_KEY_RANDOMIZE_HASH -
+ *				There are some things, pointers for example
+ *				which don't hash well because they do not use
+ *				the lower bits. If this flag is set then the
+ *				hash table will attempt to rectify this by
+ *				randomising the bits and then using the upper
+ *				N bits as the index into the table.
+ * TCL_HASH_KEY_SYSTEM_HASH -	If this flag is set then all memory internally
+ *                              allocated for the hash table that is not for an
+ *                              entry will use the system heap.
+ */
+
+#define TCL_HASH_KEY_RANDOMIZE_HASH 0x1
+#define TCL_HASH_KEY_SYSTEM_HASH    0x2
+
+/*
+ * Structure definition for the methods associated with a hash table key type.
+ */
+
+#define TCL_HASH_KEY_TYPE_VERSION 1
+struct Tcl_HashKeyType {
+    int version;		/* Version of the table. If this structure is
+				 * extended in future then the version can be
+				 * used to distinguish between different
+				 * structures. */
+    int flags;			/* Flags, see above for details. */
+    Tcl_HashKeyProc *hashKeyProc;
+				/* Calculates a hash value for the key. If
+				 * this is NULL then the pointer itself is
+				 * used as a hash value. */
+    Tcl_CompareHashKeysProc *compareKeysProc;
+				/* Compares two keys and returns zero if they
+				 * do not match, and non-zero if they do. If
+				 * this is NULL then the pointers are
+				 * compared. */
+    Tcl_AllocHashEntryProc *allocEntryProc;
+				/* Called to allocate memory for a new entry,
+				 * i.e. if the key is a string then this could
+				 * allocate a single block which contains
+				 * enough space for both the entry and the
+				 * string. Only the key field of the allocated
+				 * Tcl_HashEntry structure needs to be filled
+				 * in. If something else needs to be done to
+				 * the key, i.e. incrementing a reference
+				 * count then that should be done by this
+				 * function. If this is NULL then Tcl_Alloc is
+				 * used to allocate enough space for a
+				 * Tcl_HashEntry and the key pointer is
+				 * assigned to key.oneWordValue. */
+    Tcl_FreeHashEntryProc *freeEntryProc;
+				/* Called to free memory associated with an
+				 * entry. If something else needs to be done
+				 * to the key, i.e. decrementing a reference
+				 * count then that should be done by this
+				 * function. If this is NULL then Tcl_Free is
+				 * used to free the Tcl_HashEntry. */
+};
+
+/*
+ * Structure definition for a hash table.  Must be in tcl.h so clients can
+ * allocate space for these structures, but clients should never access any
+ * fields in this structure.
+ */
+
+#define TCL_SMALL_HASH_TABLE 4
+struct Tcl_HashTable {
+    Tcl_HashEntry **buckets;	/* Pointer to bucket array. Each element
+				 * points to first entry in bucket's hash
+				 * chain, or NULL. */
+    Tcl_HashEntry *staticBuckets[TCL_SMALL_HASH_TABLE];
+				/* Bucket array used for small tables (to
+				 * avoid mallocs and frees). */
+    int numBuckets;		/* Total number of buckets allocated at
+				 * **bucketPtr. */
+    int numEntries;		/* Total number of entries present in
+				 * table. */
+    int rebuildSize;		/* Enlarge table when numEntries gets to be
+				 * this large. */
+    int downShift;		/* Shift count used in hashing function.
+				 * Designed to use high-order bits of
+				 * randomized keys. */
+    int mask;			/* Mask value used in hashing function. */
+    int keyType;		/* Type of keys used in this table. It's
+				 * either TCL_CUSTOM_KEYS, TCL_STRING_KEYS,
+				 * TCL_ONE_WORD_KEYS, or an integer giving the
+				 * number of ints that is the size of the
+				 * key. */
+    Tcl_HashEntry *(*findProc) (Tcl_HashTable *tablePtr, const char *key);
+    Tcl_HashEntry *(*createProc) (Tcl_HashTable *tablePtr, const char *key,
+	    int *newPtr);
+    const Tcl_HashKeyType *typePtr;
+				/* Type of the keys used in the
+				 * Tcl_HashTable. */
+};
+
+/*
+ * Structure definition for information used to keep track of searches through
+ * hash tables:
+ */
+
+typedef struct Tcl_HashSearch {
+    Tcl_HashTable *tablePtr;	/* Table being searched. */
+    int nextIndex;		/* Index of next bucket to be enumerated after
+				 * present one. */
+    Tcl_HashEntry *nextEntryPtr;/* Next entry to be enumerated in the current
+				 * bucket. */
+} Tcl_HashSearch;
+
+/*
+ * Acceptable key types for hash tables:
+ *
+ * TCL_STRING_KEYS:		The keys are strings, they are copied into the
+ *				entry.
+ * TCL_ONE_WORD_KEYS:		The keys are pointers, the pointer is stored
+ *				in the entry.
+ * TCL_CUSTOM_TYPE_KEYS:	The keys are arbitrary types which are copied
+ *				into the entry.
+ * TCL_CUSTOM_PTR_KEYS:		The keys are pointers to arbitrary types, the
+ *				pointer is stored in the entry.
+ *
+ * While maintaining binary compatibility the above have to be distinct values
+ * as they are used to differentiate between old versions of the hash table
+ * which don't have a typePtr and new ones which do. Once binary compatibility
+ * is discarded in favour of making more wide spread changes TCL_STRING_KEYS
+ * can be the same as TCL_CUSTOM_TYPE_KEYS, and TCL_ONE_WORD_KEYS can be the
+ * same as TCL_CUSTOM_PTR_KEYS because they simply determine how the key is
+ * accessed from the entry and not the behaviour.
+ */
+
+#define TCL_STRING_KEYS		(0)
+#define TCL_ONE_WORD_KEYS	(1)
+#define TCL_CUSTOM_TYPE_KEYS	(-2)
+#define TCL_CUSTOM_PTR_KEYS	(-1)
+
+/*
+ * Structure definition for information used to keep track of searches through
+ * dictionaries. These fields should not be accessed by code outside
+ * tclDictObj.c
+ */
+
+typedef struct {
+    void *next;			/* Search position for underlying hash
+				 * table. */
+    int epoch;			/* Epoch marker for dictionary being searched,
+				 * or -1 if search has terminated. */
+    Tcl_Dict dictionaryPtr;	/* Reference to dictionary being searched. */
+} Tcl_DictSearch;
+
+/*
+ *----------------------------------------------------------------------------
+ * Flag values to pass to Tcl_DoOneEvent to disable searches for some kinds of
+ * events:
+ */
+
+#define TCL_DONT_WAIT		(1<<1)
+#define TCL_WINDOW_EVENTS	(1<<2)
+#define TCL_FILE_EVENTS		(1<<3)
+#define TCL_TIMER_EVENTS	(1<<4)
+#define TCL_IDLE_EVENTS		(1<<5)	/* WAS 0x10 ???? */
+#define TCL_ALL_EVENTS		(~TCL_DONT_WAIT)
+
+/*
+ * The following structure defines a generic event for the Tcl event system.
+ * These are the things that are queued in calls to Tcl_QueueEvent and
+ * serviced later by Tcl_DoOneEvent. There can be many different kinds of
+ * events with different fields, corresponding to window events, timer events,
+ * etc. The structure for a particular event consists of a Tcl_Event header
+ * followed by additional information specific to that event.
+ */
+
+struct Tcl_Event {
+    Tcl_EventProc *proc;	/* Function to call to service this event. */
+    struct Tcl_Event *nextPtr;	/* Next in list of pending events, or NULL. */
+};
+
+/*
+ * Positions to pass to Tcl_QueueEvent:
+ */
+
+typedef enum {
+    TCL_QUEUE_TAIL, TCL_QUEUE_HEAD, TCL_QUEUE_MARK
+} Tcl_QueuePosition;
+
+/*
+ * Values to pass to Tcl_SetServiceMode to specify the behavior of notifier
+ * event routines.
+ */
+
+#define TCL_SERVICE_NONE 0
+#define TCL_SERVICE_ALL 1
+
+/*
+ * The following structure keeps is used to hold a time value, either as an
+ * absolute time (the number of seconds from the epoch) or as an elapsed time.
+ * On Unix systems the epoch is Midnight Jan 1, 1970 GMT.
+ */
+
+typedef struct Tcl_Time {
+    long sec;			/* Seconds. */
+    long usec;			/* Microseconds. */
+} Tcl_Time;
+
+typedef void (Tcl_SetTimerProc) (CONST86 Tcl_Time *timePtr);
+typedef int (Tcl_WaitForEventProc) (CONST86 Tcl_Time *timePtr);
+
+/*
+ * TIP #233 (Virtualized Time)
+ */
+
+typedef void (Tcl_GetTimeProc)   (Tcl_Time *timebuf, ClientData clientData);
+typedef void (Tcl_ScaleTimeProc) (Tcl_Time *timebuf, ClientData clientData);
+
+/*
+ *----------------------------------------------------------------------------
+ * Bits to pass to Tcl_CreateFileHandler and Tcl_CreateChannelHandler to
+ * indicate what sorts of events are of interest:
+ */
+
+#define TCL_READABLE		(1<<1)
+#define TCL_WRITABLE		(1<<2)
+#define TCL_EXCEPTION		(1<<3)
+
+/*
+ * Flag values to pass to Tcl_OpenCommandChannel to indicate the disposition
+ * of the stdio handles. TCL_STDIN, TCL_STDOUT, TCL_STDERR, are also used in
+ * Tcl_GetStdChannel.
+ */
+
+#define TCL_STDIN		(1<<1)
+#define TCL_STDOUT		(1<<2)
+#define TCL_STDERR		(1<<3)
+#define TCL_ENFORCE_MODE	(1<<4)
+
+/*
+ * Bits passed to Tcl_DriverClose2Proc to indicate which side of a channel
+ * should be closed.
+ */
+
+#define TCL_CLOSE_READ		(1<<1)
+#define TCL_CLOSE_WRITE		(1<<2)
+
+/*
+ * Value to use as the closeProc for a channel that supports the close2Proc
+ * interface.
+ */
+
+#define TCL_CLOSE2PROC		((Tcl_DriverCloseProc *) 1)
+
+/*
+ * Channel version tag. This was introduced in 8.3.2/8.4.
+ */
+
+#define TCL_CHANNEL_VERSION_1	((Tcl_ChannelTypeVersion) 0x1)
+#define TCL_CHANNEL_VERSION_2	((Tcl_ChannelTypeVersion) 0x2)
+#define TCL_CHANNEL_VERSION_3	((Tcl_ChannelTypeVersion) 0x3)
+#define TCL_CHANNEL_VERSION_4	((Tcl_ChannelTypeVersion) 0x4)
+#define TCL_CHANNEL_VERSION_5	((Tcl_ChannelTypeVersion) 0x5)
+
+/*
+ * TIP #218: Channel Actions, Ids for Tcl_DriverThreadActionProc.
+ */
+
+#define TCL_CHANNEL_THREAD_INSERT (0)
+#define TCL_CHANNEL_THREAD_REMOVE (1)
+
+/*
+ * Typedefs for the various operations in a channel type:
+ */
+
+typedef int	(Tcl_DriverBlockModeProc) (ClientData instanceData, int mode);
+typedef int	(Tcl_DriverCloseProc) (ClientData instanceData,
+			Tcl_Interp *interp);
+typedef int	(Tcl_DriverClose2Proc) (ClientData instanceData,
+			Tcl_Interp *interp, int flags);
+typedef int	(Tcl_DriverInputProc) (ClientData instanceData, char *buf,
+			int toRead, int *errorCodePtr);
+typedef int	(Tcl_DriverOutputProc) (ClientData instanceData,
+			CONST84 char *buf, int toWrite, int *errorCodePtr);
+typedef int	(Tcl_DriverSeekProc) (ClientData instanceData, long offset,
+			int mode, int *errorCodePtr);
+typedef int	(Tcl_DriverSetOptionProc) (ClientData instanceData,
+			Tcl_Interp *interp, const char *optionName,
+			const char *value);
+typedef int	(Tcl_DriverGetOptionProc) (ClientData instanceData,
+			Tcl_Interp *interp, CONST84 char *optionName,
+			Tcl_DString *dsPtr);
+typedef void	(Tcl_DriverWatchProc) (ClientData instanceData, int mask);
+typedef int	(Tcl_DriverGetHandleProc) (ClientData instanceData,
+			int direction, ClientData *handlePtr);
+typedef int	(Tcl_DriverFlushProc) (ClientData instanceData);
+typedef int	(Tcl_DriverHandlerProc) (ClientData instanceData,
+			int interestMask);
+typedef Tcl_WideInt (Tcl_DriverWideSeekProc) (ClientData instanceData,
+			Tcl_WideInt offset, int mode, int *errorCodePtr);
+/*
+ * TIP #218, Channel Thread Actions
+ */
+typedef void	(Tcl_DriverThreadActionProc) (ClientData instanceData,
+			int action);
+/*
+ * TIP #208, File Truncation (etc.)
+ */
+typedef int	(Tcl_DriverTruncateProc) (ClientData instanceData,
+			Tcl_WideInt length);
+
+/*
+ * struct Tcl_ChannelType:
+ *
+ * One such structure exists for each type (kind) of channel. It collects
+ * together in one place all the functions that are part of the specific
+ * channel type.
+ *
+ * It is recommend that the Tcl_Channel* functions are used to access elements
+ * of this structure, instead of direct accessing.
+ */
+
+typedef struct Tcl_ChannelType {
+    const char *typeName;	/* The name of the channel type in Tcl
+				 * commands. This storage is owned by channel
+				 * type. */
+    Tcl_ChannelTypeVersion version;
+				/* Version of the channel type. */
+    Tcl_DriverCloseProc *closeProc;
+				/* Function to call to close the channel, or
+				 * TCL_CLOSE2PROC if the close2Proc should be
+				 * used instead. */
+    Tcl_DriverInputProc *inputProc;
+				/* Function to call for input on channel. */
+    Tcl_DriverOutputProc *outputProc;
+				/* Function to call for output on channel. */
+    Tcl_DriverSeekProc *seekProc;
+				/* Function to call to seek on the channel.
+				 * May be NULL. */
+    Tcl_DriverSetOptionProc *setOptionProc;
+				/* Set an option on a channel. */
+    Tcl_DriverGetOptionProc *getOptionProc;
+				/* Get an option from a channel. */
+    Tcl_DriverWatchProc *watchProc;
+				/* Set up the notifier to watch for events on
+				 * this channel. */
+    Tcl_DriverGetHandleProc *getHandleProc;
+				/* Get an OS handle from the channel or NULL
+				 * if not supported. */
+    Tcl_DriverClose2Proc *close2Proc;
+				/* Function to call to close the channel if
+				 * the device supports closing the read &
+				 * write sides independently. */
+    Tcl_DriverBlockModeProc *blockModeProc;
+				/* Set blocking mode for the raw channel. May
+				 * be NULL. */
+    /*
+     * Only valid in TCL_CHANNEL_VERSION_2 channels or later.
+     */
+    Tcl_DriverFlushProc *flushProc;
+				/* Function to call to flush a channel. May be
+				 * NULL. */
+    Tcl_DriverHandlerProc *handlerProc;
+				/* Function to call to handle a channel event.
+				 * This will be passed up the stacked channel
+				 * chain. */
+    /*
+     * Only valid in TCL_CHANNEL_VERSION_3 channels or later.
+     */
+    Tcl_DriverWideSeekProc *wideSeekProc;
+				/* Function to call to seek on the channel
+				 * which can handle 64-bit offsets. May be
+				 * NULL, and must be NULL if seekProc is
+				 * NULL. */
+    /*
+     * Only valid in TCL_CHANNEL_VERSION_4 channels or later.
+     * TIP #218, Channel Thread Actions.
+     */
+    Tcl_DriverThreadActionProc *threadActionProc;
+				/* Function to call to notify the driver of
+				 * thread specific activity for a channel. May
+				 * be NULL. */
+    /*
+     * Only valid in TCL_CHANNEL_VERSION_5 channels or later.
+     * TIP #208, File Truncation.
+     */
+    Tcl_DriverTruncateProc *truncateProc;
+				/* Function to call to truncate the underlying
+				 * file to a particular length. May be NULL if
+				 * the channel does not support truncation. */
+} Tcl_ChannelType;
+
+/*
+ * The following flags determine whether the blockModeProc above should set
+ * the channel into blocking or nonblocking mode. They are passed as arguments
+ * to the blockModeProc function in the above structure.
+ */
+
+#define TCL_MODE_BLOCKING	0	/* Put channel into blocking mode. */
+#define TCL_MODE_NONBLOCKING	1	/* Put channel into nonblocking
+					 * mode. */
+
+/*
+ *----------------------------------------------------------------------------
+ * Enum for different types of file paths.
+ */
+
+typedef enum Tcl_PathType {
+    TCL_PATH_ABSOLUTE,
+    TCL_PATH_RELATIVE,
+    TCL_PATH_VOLUME_RELATIVE
+} Tcl_PathType;
+
+/*
+ * The following structure is used to pass glob type data amongst the various
+ * glob routines and Tcl_FSMatchInDirectory.
+ */
+
+typedef struct Tcl_GlobTypeData {
+    int type;			/* Corresponds to bcdpfls as in 'find -t'. */
+    int perm;			/* Corresponds to file permissions. */
+    Tcl_Obj *macType;		/* Acceptable Mac type. */
+    Tcl_Obj *macCreator;	/* Acceptable Mac creator. */
+} Tcl_GlobTypeData;
+
+/*
+ * Type and permission definitions for glob command.
+ */
+
+#define TCL_GLOB_TYPE_BLOCK		(1<<0)
+#define TCL_GLOB_TYPE_CHAR		(1<<1)
+#define TCL_GLOB_TYPE_DIR		(1<<2)
+#define TCL_GLOB_TYPE_PIPE		(1<<3)
+#define TCL_GLOB_TYPE_FILE		(1<<4)
+#define TCL_GLOB_TYPE_LINK		(1<<5)
+#define TCL_GLOB_TYPE_SOCK		(1<<6)
+#define TCL_GLOB_TYPE_MOUNT		(1<<7)
+
+#define TCL_GLOB_PERM_RONLY		(1<<0)
+#define TCL_GLOB_PERM_HIDDEN		(1<<1)
+#define TCL_GLOB_PERM_R			(1<<2)
+#define TCL_GLOB_PERM_W			(1<<3)
+#define TCL_GLOB_PERM_X			(1<<4)
+
+/*
+ * Flags for the unload callback function.
+ */
+
+#define TCL_UNLOAD_DETACH_FROM_INTERPRETER	(1<<0)
+#define TCL_UNLOAD_DETACH_FROM_PROCESS		(1<<1)
+
+/*
+ * Typedefs for the various filesystem operations:
+ */
+
+typedef int (Tcl_FSStatProc) (Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+typedef int (Tcl_FSAccessProc) (Tcl_Obj *pathPtr, int mode);
+typedef Tcl_Channel (Tcl_FSOpenFileChannelProc) (Tcl_Interp *interp,
+	Tcl_Obj *pathPtr, int mode, int permissions);
+typedef int (Tcl_FSMatchInDirectoryProc) (Tcl_Interp *interp, Tcl_Obj *result,
+	Tcl_Obj *pathPtr, const char *pattern, Tcl_GlobTypeData *types);
+typedef Tcl_Obj * (Tcl_FSGetCwdProc) (Tcl_Interp *interp);
+typedef int (Tcl_FSChdirProc) (Tcl_Obj *pathPtr);
+typedef int (Tcl_FSLstatProc) (Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+typedef int (Tcl_FSCreateDirectoryProc) (Tcl_Obj *pathPtr);
+typedef int (Tcl_FSDeleteFileProc) (Tcl_Obj *pathPtr);
+typedef int (Tcl_FSCopyDirectoryProc) (Tcl_Obj *srcPathPtr,
+	Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr);
+typedef int (Tcl_FSCopyFileProc) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr);
+typedef int (Tcl_FSRemoveDirectoryProc) (Tcl_Obj *pathPtr, int recursive,
+	Tcl_Obj **errorPtr);
+typedef int (Tcl_FSRenameFileProc) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr);
+typedef void (Tcl_FSUnloadFileProc) (Tcl_LoadHandle loadHandle);
+typedef Tcl_Obj * (Tcl_FSListVolumesProc) (void);
+/* We have to declare the utime structure here. */
+struct utimbuf;
+typedef int (Tcl_FSUtimeProc) (Tcl_Obj *pathPtr, struct utimbuf *tval);
+typedef int (Tcl_FSNormalizePathProc) (Tcl_Interp *interp, Tcl_Obj *pathPtr,
+	int nextCheckpoint);
+typedef int (Tcl_FSFileAttrsGetProc) (Tcl_Interp *interp, int index,
+	Tcl_Obj *pathPtr, Tcl_Obj **objPtrRef);
+typedef const char *CONST86 * (Tcl_FSFileAttrStringsProc) (Tcl_Obj *pathPtr,
+	Tcl_Obj **objPtrRef);
+typedef int (Tcl_FSFileAttrsSetProc) (Tcl_Interp *interp, int index,
+	Tcl_Obj *pathPtr, Tcl_Obj *objPtr);
+typedef Tcl_Obj * (Tcl_FSLinkProc) (Tcl_Obj *pathPtr, Tcl_Obj *toPtr,
+	int linkType);
+typedef int (Tcl_FSLoadFileProc) (Tcl_Interp *interp, Tcl_Obj *pathPtr,
+	Tcl_LoadHandle *handlePtr, Tcl_FSUnloadFileProc **unloadProcPtr);
+typedef int (Tcl_FSPathInFilesystemProc) (Tcl_Obj *pathPtr,
+	ClientData *clientDataPtr);
+typedef Tcl_Obj * (Tcl_FSFilesystemPathTypeProc) (Tcl_Obj *pathPtr);
+typedef Tcl_Obj * (Tcl_FSFilesystemSeparatorProc) (Tcl_Obj *pathPtr);
+typedef void (Tcl_FSFreeInternalRepProc) (ClientData clientData);
+typedef ClientData (Tcl_FSDupInternalRepProc) (ClientData clientData);
+typedef Tcl_Obj * (Tcl_FSInternalToNormalizedProc) (ClientData clientData);
+typedef ClientData (Tcl_FSCreateInternalRepProc) (Tcl_Obj *pathPtr);
+
+typedef struct Tcl_FSVersion_ *Tcl_FSVersion;
+
+/*
+ *----------------------------------------------------------------------------
+ * Data structures related to hooking into the filesystem
+ */
+
+/*
+ * Filesystem version tag.  This was introduced in 8.4.
+ */
+
+#define TCL_FILESYSTEM_VERSION_1	((Tcl_FSVersion) 0x1)
+
+/*
+ * struct Tcl_Filesystem:
+ *
+ * One such structure exists for each type (kind) of filesystem. It collects
+ * together in one place all the functions that are part of the specific
+ * filesystem. Tcl always accesses the filesystem through one of these
+ * structures.
+ *
+ * Not all entries need be non-NULL; any which are NULL are simply ignored.
+ * However, a complete filesystem should provide all of these functions. The
+ * explanations in the structure show the importance of each function.
+ */
+
+typedef struct Tcl_Filesystem {
+    const char *typeName;	/* The name of the filesystem. */
+    int structureLength;	/* Length of this structure, so future binary
+				 * compatibility can be assured. */
+    Tcl_FSVersion version;	/* Version of the filesystem type. */
+    Tcl_FSPathInFilesystemProc *pathInFilesystemProc;
+				/* Function to check whether a path is in this
+				 * filesystem. This is the most important
+				 * filesystem function. */
+    Tcl_FSDupInternalRepProc *dupInternalRepProc;
+				/* Function to duplicate internal fs rep. May
+				 * be NULL (but then fs is less efficient). */
+    Tcl_FSFreeInternalRepProc *freeInternalRepProc;
+				/* Function to free internal fs rep. Must be
+				 * implemented if internal representations
+				 * need freeing, otherwise it can be NULL. */
+    Tcl_FSInternalToNormalizedProc *internalToNormalizedProc;
+				/* Function to convert internal representation
+				 * to a normalized path. Only required if the
+				 * fs creates pure path objects with no
+				 * string/path representation. */
+    Tcl_FSCreateInternalRepProc *createInternalRepProc;
+				/* Function to create a filesystem-specific
+				 * internal representation. May be NULL if
+				 * paths have no internal representation, or
+				 * if the Tcl_FSPathInFilesystemProc for this
+				 * filesystem always immediately creates an
+				 * internal representation for paths it
+				 * accepts. */
+    Tcl_FSNormalizePathProc *normalizePathProc;
+				/* Function to normalize a path.  Should be
+				 * implemented for all filesystems which can
+				 * have multiple string representations for
+				 * the same path object. */
+    Tcl_FSFilesystemPathTypeProc *filesystemPathTypeProc;
+				/* Function to determine the type of a path in
+				 * this filesystem. May be NULL. */
+    Tcl_FSFilesystemSeparatorProc *filesystemSeparatorProc;
+				/* Function to return the separator
+				 * character(s) for this filesystem. Must be
+				 * implemented. */
+    Tcl_FSStatProc *statProc;	/* Function to process a 'Tcl_FSStat()' call.
+				 * Must be implemented for any reasonable
+				 * filesystem. */
+    Tcl_FSAccessProc *accessProc;
+				/* Function to process a 'Tcl_FSAccess()'
+				 * call. Must be implemented for any
+				 * reasonable filesystem. */
+    Tcl_FSOpenFileChannelProc *openFileChannelProc;
+				/* Function to process a
+				 * 'Tcl_FSOpenFileChannel()' call. Must be
+				 * implemented for any reasonable
+				 * filesystem. */
+    Tcl_FSMatchInDirectoryProc *matchInDirectoryProc;
+				/* Function to process a
+				 * 'Tcl_FSMatchInDirectory()'.  If not
+				 * implemented, then glob and recursive copy
+				 * functionality will be lacking in the
+				 * filesystem. */
+    Tcl_FSUtimeProc *utimeProc;	/* Function to process a 'Tcl_FSUtime()' call.
+				 * Required to allow setting (not reading) of
+				 * times with 'file mtime', 'file atime' and
+				 * the open-r/open-w/fcopy implementation of
+				 * 'file copy'. */
+    Tcl_FSLinkProc *linkProc;	/* Function to process a 'Tcl_FSLink()' call.
+				 * Should be implemented only if the
+				 * filesystem supports links (reading or
+				 * creating). */
+    Tcl_FSListVolumesProc *listVolumesProc;
+				/* Function to list any filesystem volumes
+				 * added by this filesystem. Should be
+				 * implemented only if the filesystem adds
+				 * volumes at the head of the filesystem. */
+    Tcl_FSFileAttrStringsProc *fileAttrStringsProc;
+				/* Function to list all attributes strings
+				 * which are valid for this filesystem. If not
+				 * implemented the filesystem will not support
+				 * the 'file attributes' command. This allows
+				 * arbitrary additional information to be
+				 * attached to files in the filesystem. */
+    Tcl_FSFileAttrsGetProc *fileAttrsGetProc;
+				/* Function to process a
+				 * 'Tcl_FSFileAttrsGet()' call, used by 'file
+				 * attributes'. */
+    Tcl_FSFileAttrsSetProc *fileAttrsSetProc;
+				/* Function to process a
+				 * 'Tcl_FSFileAttrsSet()' call, used by 'file
+				 * attributes'.  */
+    Tcl_FSCreateDirectoryProc *createDirectoryProc;
+				/* Function to process a
+				 * 'Tcl_FSCreateDirectory()' call. Should be
+				 * implemented unless the FS is read-only. */
+    Tcl_FSRemoveDirectoryProc *removeDirectoryProc;
+				/* Function to process a
+				 * 'Tcl_FSRemoveDirectory()' call. Should be
+				 * implemented unless the FS is read-only. */
+    Tcl_FSDeleteFileProc *deleteFileProc;
+				/* Function to process a 'Tcl_FSDeleteFile()'
+				 * call. Should be implemented unless the FS
+				 * is read-only. */
+    Tcl_FSCopyFileProc *copyFileProc;
+				/* Function to process a 'Tcl_FSCopyFile()'
+				 * call. If not implemented Tcl will fall back
+				 * on open-r, open-w and fcopy as a copying
+				 * mechanism, for copying actions initiated in
+				 * Tcl (not C). */
+    Tcl_FSRenameFileProc *renameFileProc;
+				/* Function to process a 'Tcl_FSRenameFile()'
+				 * call. If not implemented, Tcl will fall
+				 * back on a copy and delete mechanism, for
+				 * rename actions initiated in Tcl (not C). */
+    Tcl_FSCopyDirectoryProc *copyDirectoryProc;
+				/* Function to process a
+				 * 'Tcl_FSCopyDirectory()' call. If not
+				 * implemented, Tcl will fall back on a
+				 * recursive create-dir, file copy mechanism,
+				 * for copying actions initiated in Tcl (not
+				 * C). */
+    Tcl_FSLstatProc *lstatProc;	/* Function to process a 'Tcl_FSLstat()' call.
+				 * If not implemented, Tcl will attempt to use
+				 * the 'statProc' defined above instead. */
+    Tcl_FSLoadFileProc *loadFileProc;
+				/* Function to process a 'Tcl_FSLoadFile()'
+				 * call. If not implemented, Tcl will fall
+				 * back on a copy to native-temp followed by a
+				 * Tcl_FSLoadFile on that temporary copy. */
+    Tcl_FSGetCwdProc *getCwdProc;
+				/* Function to process a 'Tcl_FSGetCwd()'
+				 * call. Most filesystems need not implement
+				 * this. It will usually only be called once,
+				 * if 'getcwd' is called before 'chdir'. May
+				 * be NULL. */
+    Tcl_FSChdirProc *chdirProc;	/* Function to process a 'Tcl_FSChdir()' call.
+				 * If filesystems do not implement this, it
+				 * will be emulated by a series of directory
+				 * access checks. Otherwise, virtual
+				 * filesystems which do implement it need only
+				 * respond with a positive return result if
+				 * the dirName is a valid directory in their
+				 * filesystem. They need not remember the
+				 * result, since that will be automatically
+				 * remembered for use by GetCwd. Real
+				 * filesystems should carry out the correct
+				 * action (i.e. call the correct system
+				 * 'chdir' api). If not implemented, then 'cd'
+				 * and 'pwd' will fail inside the
+				 * filesystem. */
+} Tcl_Filesystem;
+
+/*
+ * The following definitions are used as values for the 'linkAction' flag to
+ * Tcl_FSLink, or the linkProc of any filesystem. Any combination of flags can
+ * be given. For link creation, the linkProc should create a link which
+ * matches any of the types given.
+ *
+ * TCL_CREATE_SYMBOLIC_LINK -	Create a symbolic or soft link.
+ * TCL_CREATE_HARD_LINK -	Create a hard link.
+ */
+
+#define TCL_CREATE_SYMBOLIC_LINK	0x01
+#define TCL_CREATE_HARD_LINK		0x02
+
+/*
+ *----------------------------------------------------------------------------
+ * The following structure represents the Notifier functions that you can
+ * override with the Tcl_SetNotifier call.
+ */
+
+typedef struct Tcl_NotifierProcs {
+    Tcl_SetTimerProc *setTimerProc;
+    Tcl_WaitForEventProc *waitForEventProc;
+    Tcl_CreateFileHandlerProc *createFileHandlerProc;
+    Tcl_DeleteFileHandlerProc *deleteFileHandlerProc;
+    Tcl_InitNotifierProc *initNotifierProc;
+    Tcl_FinalizeNotifierProc *finalizeNotifierProc;
+    Tcl_AlertNotifierProc *alertNotifierProc;
+    Tcl_ServiceModeHookProc *serviceModeHookProc;
+} Tcl_NotifierProcs;
+
+/*
+ *----------------------------------------------------------------------------
+ * The following data structures and declarations are for the new Tcl parser.
+ *
+ * For each word of a command, and for each piece of a word such as a variable
+ * reference, one of the following structures is created to describe the
+ * token.
+ */
+
+typedef struct Tcl_Token {
+    int type;			/* Type of token, such as TCL_TOKEN_WORD; see
+				 * below for valid types. */
+    const char *start;		/* First character in token. */
+    int size;			/* Number of bytes in token. */
+    int numComponents;		/* If this token is composed of other tokens,
+				 * this field tells how many of them there are
+				 * (including components of components, etc.).
+				 * The component tokens immediately follow
+				 * this one. */
+} Tcl_Token;
+
+/*
+ * Type values defined for Tcl_Token structures. These values are defined as
+ * mask bits so that it's easy to check for collections of types.
+ *
+ * TCL_TOKEN_WORD -		The token describes one word of a command,
+ *				from the first non-blank character of the word
+ *				(which may be " or {) up to but not including
+ *				the space, semicolon, or bracket that
+ *				terminates the word. NumComponents counts the
+ *				total number of sub-tokens that make up the
+ *				word. This includes, for example, sub-tokens
+ *				of TCL_TOKEN_VARIABLE tokens.
+ * TCL_TOKEN_SIMPLE_WORD -	This token is just like TCL_TOKEN_WORD except
+ *				that the word is guaranteed to consist of a
+ *				single TCL_TOKEN_TEXT sub-token.
+ * TCL_TOKEN_TEXT -		The token describes a range of literal text
+ *				that is part of a word. NumComponents is
+ *				always 0.
+ * TCL_TOKEN_BS -		The token describes a backslash sequence that
+ *				must be collapsed. NumComponents is always 0.
+ * TCL_TOKEN_COMMAND -		The token describes a command whose result
+ *				must be substituted into the word. The token
+ *				includes the enclosing brackets. NumComponents
+ *				is always 0.
+ * TCL_TOKEN_VARIABLE -		The token describes a variable substitution,
+ *				including the dollar sign, variable name, and
+ *				array index (if there is one) up through the
+ *				right parentheses. NumComponents tells how
+ *				many additional tokens follow to represent the
+ *				variable name. The first token will be a
+ *				TCL_TOKEN_TEXT token that describes the
+ *				variable name. If the variable is an array
+ *				reference then there will be one or more
+ *				additional tokens, of type TCL_TOKEN_TEXT,
+ *				TCL_TOKEN_BS, TCL_TOKEN_COMMAND, and
+ *				TCL_TOKEN_VARIABLE, that describe the array
+ *				index; numComponents counts the total number
+ *				of nested tokens that make up the variable
+ *				reference, including sub-tokens of
+ *				TCL_TOKEN_VARIABLE tokens.
+ * TCL_TOKEN_SUB_EXPR -		The token describes one subexpression of an
+ *				expression, from the first non-blank character
+ *				of the subexpression up to but not including
+ *				the space, brace, or bracket that terminates
+ *				the subexpression. NumComponents counts the
+ *				total number of following subtokens that make
+ *				up the subexpression; this includes all
+ *				subtokens for any nested TCL_TOKEN_SUB_EXPR
+ *				tokens. For example, a numeric value used as a
+ *				primitive operand is described by a
+ *				TCL_TOKEN_SUB_EXPR token followed by a
+ *				TCL_TOKEN_TEXT token. A binary subexpression
+ *				is described by a TCL_TOKEN_SUB_EXPR token
+ *				followed by the TCL_TOKEN_OPERATOR token for
+ *				the operator, then TCL_TOKEN_SUB_EXPR tokens
+ *				for the left then the right operands.
+ * TCL_TOKEN_OPERATOR -		The token describes one expression operator.
+ *				An operator might be the name of a math
+ *				function such as "abs". A TCL_TOKEN_OPERATOR
+ *				token is always preceeded by one
+ *				TCL_TOKEN_SUB_EXPR token for the operator's
+ *				subexpression, and is followed by zero or more
+ *				TCL_TOKEN_SUB_EXPR tokens for the operator's
+ *				operands. NumComponents is always 0.
+ * TCL_TOKEN_EXPAND_WORD -	This token is just like TCL_TOKEN_WORD except
+ *				that it marks a word that began with the
+ *				literal character prefix "{*}". This word is
+ *				marked to be expanded - that is, broken into
+ *				words after substitution is complete.
+ */
+
+#define TCL_TOKEN_WORD		1
+#define TCL_TOKEN_SIMPLE_WORD	2
+#define TCL_TOKEN_TEXT		4
+#define TCL_TOKEN_BS		8
+#define TCL_TOKEN_COMMAND	16
+#define TCL_TOKEN_VARIABLE	32
+#define TCL_TOKEN_SUB_EXPR	64
+#define TCL_TOKEN_OPERATOR	128
+#define TCL_TOKEN_EXPAND_WORD	256
+
+/*
+ * Parsing error types. On any parsing error, one of these values will be
+ * stored in the error field of the Tcl_Parse structure defined below.
+ */
+
+#define TCL_PARSE_SUCCESS		0
+#define TCL_PARSE_QUOTE_EXTRA		1
+#define TCL_PARSE_BRACE_EXTRA		2
+#define TCL_PARSE_MISSING_BRACE		3
+#define TCL_PARSE_MISSING_BRACKET	4
+#define TCL_PARSE_MISSING_PAREN		5
+#define TCL_PARSE_MISSING_QUOTE		6
+#define TCL_PARSE_MISSING_VAR_BRACE	7
+#define TCL_PARSE_SYNTAX		8
+#define TCL_PARSE_BAD_NUMBER		9
+
+/*
+ * A structure of the following type is filled in by Tcl_ParseCommand. It
+ * describes a single command parsed from an input string.
+ */
+
+#define NUM_STATIC_TOKENS 20
+
+typedef struct Tcl_Parse {
+    const char *commentStart;	/* Pointer to # that begins the first of one
+				 * or more comments preceding the command. */
+    int commentSize;		/* Number of bytes in comments (up through
+				 * newline character that terminates the last
+				 * comment). If there were no comments, this
+				 * field is 0. */
+    const char *commandStart;	/* First character in first word of
+				 * command. */
+    int commandSize;		/* Number of bytes in command, including first
+				 * character of first word, up through the
+				 * terminating newline, close bracket, or
+				 * semicolon. */
+    int numWords;		/* Total number of words in command. May be
+				 * 0. */
+    Tcl_Token *tokenPtr;	/* Pointer to first token representing the
+				 * words of the command. Initially points to
+				 * staticTokens, but may change to point to
+				 * malloc-ed space if command exceeds space in
+				 * staticTokens. */
+    int numTokens;		/* Total number of tokens in command. */
+    int tokensAvailable;	/* Total number of tokens available at
+				 * *tokenPtr. */
+    int errorType;		/* One of the parsing error types defined
+				 * above. */
+
+    /*
+     * The fields below are intended only for the private use of the parser.
+     * They should not be used by functions that invoke Tcl_ParseCommand.
+     */
+
+    const char *string;		/* The original command string passed to
+				 * Tcl_ParseCommand. */
+    const char *end;		/* Points to the character just after the last
+				 * one in the command string. */
+    Tcl_Interp *interp;		/* Interpreter to use for error reporting, or
+				 * NULL. */
+    const char *term;		/* Points to character in string that
+				 * terminated most recent token. Filled in by
+				 * ParseTokens. If an error occurs, points to
+				 * beginning of region where the error
+				 * occurred (e.g. the open brace if the close
+				 * brace is missing). */
+    int incomplete;		/* This field is set to 1 by Tcl_ParseCommand
+				 * if the command appears to be incomplete.
+				 * This information is used by
+				 * Tcl_CommandComplete. */
+    Tcl_Token staticTokens[NUM_STATIC_TOKENS];
+				/* Initial space for tokens for command. This
+				 * space should be large enough to accommodate
+				 * most commands; dynamic space is allocated
+				 * for very large commands that don't fit
+				 * here. */
+} Tcl_Parse;
+
+/*
+ *----------------------------------------------------------------------------
+ * The following structure represents a user-defined encoding. It collects
+ * together all the functions that are used by the specific encoding.
+ */
+
+typedef struct Tcl_EncodingType {
+    const char *encodingName;	/* The name of the encoding, e.g. "euc-jp".
+				 * This name is the unique key for this
+				 * encoding type. */
+    Tcl_EncodingConvertProc *toUtfProc;
+				/* Function to convert from external encoding
+				 * into UTF-8. */
+    Tcl_EncodingConvertProc *fromUtfProc;
+				/* Function to convert from UTF-8 into
+				 * external encoding. */
+    Tcl_EncodingFreeProc *freeProc;
+				/* If non-NULL, function to call when this
+				 * encoding is deleted. */
+    ClientData clientData;	/* Arbitrary value associated with encoding
+				 * type. Passed to conversion functions. */
+    int nullSize;		/* Number of zero bytes that signify
+				 * end-of-string in this encoding. This number
+				 * is used to determine the source string
+				 * length when the srcLen argument is
+				 * negative. Must be 1 or 2. */
+} Tcl_EncodingType;
+
+/*
+ * The following definitions are used as values for the conversion control
+ * flags argument when converting text from one character set to another:
+ *
+ * TCL_ENCODING_START -		Signifies that the source buffer is the first
+ *				block in a (potentially multi-block) input
+ *				stream. Tells the conversion function to reset
+ *				to an initial state and perform any
+ *				initialization that needs to occur before the
+ *				first byte is converted. If the source buffer
+ *				contains the entire input stream to be
+ *				converted, this flag should be set.
+ * TCL_ENCODING_END -		Signifies that the source buffer is the last
+ *				block in a (potentially multi-block) input
+ *				stream. Tells the conversion routine to
+ *				perform any finalization that needs to occur
+ *				after the last byte is converted and then to
+ *				reset to an initial state. If the source
+ *				buffer contains the entire input stream to be
+ *				converted, this flag should be set.
+ * TCL_ENCODING_STOPONERROR -	If set, then the converter will return
+ *				immediately upon encountering an invalid byte
+ *				sequence or a source character that has no
+ *				mapping in the target encoding. If clear, then
+ *				the converter will skip the problem,
+ *				substituting one or more "close" characters in
+ *				the destination buffer and then continue to
+ *				convert the source.
+ * TCL_ENCODING_NO_TERMINATE - 	If set, Tcl_ExternalToUtf will not append a
+ *				terminating NUL byte.  Knowing that it will
+ *				not need space to do so, it will fill all
+ *				dstLen bytes with encoded UTF-8 content, as
+ *				other circumstances permit.  If clear, the
+ *				default behavior is to reserve a byte in
+ *				the dst space for NUL termination, and to
+ *				append the NUL byte.
+ * TCL_ENCODING_CHAR_LIMIT -	If set and dstCharsPtr is not NULL, then
+ *				Tcl_ExternalToUtf takes the initial value
+ *				of *dstCharsPtr is taken as a limit of the
+ *				maximum number of chars to produce in the
+ *				encoded UTF-8 content.  Otherwise, the
+ *				number of chars produced is controlled only
+ *				by other limiting factors.
+ */
+
+#define TCL_ENCODING_START		0x01
+#define TCL_ENCODING_END		0x02
+#define TCL_ENCODING_STOPONERROR	0x04
+#define TCL_ENCODING_NO_TERMINATE	0x08
+#define TCL_ENCODING_CHAR_LIMIT		0x10
+
+/*
+ * The following definitions are the error codes returned by the conversion
+ * routines:
+ *
+ * TCL_OK -			All characters were converted.
+ * TCL_CONVERT_NOSPACE -	The output buffer would not have been large
+ *				enough for all of the converted data; as many
+ *				characters as could fit were converted though.
+ * TCL_CONVERT_MULTIBYTE -	The last few bytes in the source string were
+ *				the beginning of a multibyte sequence, but
+ *				more bytes were needed to complete this
+ *				sequence. A subsequent call to the conversion
+ *				routine should pass the beginning of this
+ *				unconverted sequence plus additional bytes
+ *				from the source stream to properly convert the
+ *				formerly split-up multibyte sequence.
+ * TCL_CONVERT_SYNTAX -		The source stream contained an invalid
+ *				character sequence. This may occur if the
+ *				input stream has been damaged or if the input
+ *				encoding method was misidentified. This error
+ *				is reported only if TCL_ENCODING_STOPONERROR
+ *				was specified.
+ * TCL_CONVERT_UNKNOWN -	The source string contained a character that
+ *				could not be represented in the target
+ *				encoding. This error is reported only if
+ *				TCL_ENCODING_STOPONERROR was specified.
+ */
+
+#define TCL_CONVERT_MULTIBYTE	(-1)
+#define TCL_CONVERT_SYNTAX	(-2)
+#define TCL_CONVERT_UNKNOWN	(-3)
+#define TCL_CONVERT_NOSPACE	(-4)
+
+/*
+ * The maximum number of bytes that are necessary to represent a single
+ * Unicode character in UTF-8. The valid values should be 3, 4 or 6
+ * (or perhaps 1 if we want to support a non-unicode enabled core). If 3 or
+ * 4, then Tcl_UniChar must be 2-bytes in size (UCS-2) (the default). If 6,
+ * then Tcl_UniChar must be 4-bytes in size (UCS-4). At this time UCS-2 mode
+ * is the default and recommended mode. UCS-4 is experimental and not
+ * recommended. It works for the core, but most extensions expect UCS-2.
+ */
+
+#ifndef TCL_UTF_MAX
+#define TCL_UTF_MAX		3
+#endif
+
+/*
+ * This represents a Unicode character. Any changes to this should also be
+ * reflected in regcustom.h.
+ */
+
+#if TCL_UTF_MAX > 4
+    /*
+     * unsigned int isn't 100% accurate as it should be a strict 4-byte value
+     * (perhaps wchar_t). 64-bit systems may have troubles. The size of this
+     * value must be reflected correctly in regcustom.h and
+     * in tclEncoding.c.
+     * XXX: Tcl is currently UCS-2 and planning UTF-16 for the Unicode
+     * XXX: string rep that Tcl_UniChar represents.  Changing the size
+     * XXX: of Tcl_UniChar is /not/ supported.
+     */
+typedef unsigned int Tcl_UniChar;
+#else
+typedef unsigned short Tcl_UniChar;
+#endif
+
+/*
+ *----------------------------------------------------------------------------
+ * TIP #59: The following structure is used in calls 'Tcl_RegisterConfig' to
+ * provide the system with the embedded configuration data.
+ */
+
+typedef struct Tcl_Config {
+    const char *key;		/* Configuration key to register. ASCII
+				 * encoded, thus UTF-8. */
+    const char *value;		/* The value associated with the key. System
+				 * encoding. */
+} Tcl_Config;
+
+/*
+ *----------------------------------------------------------------------------
+ * Flags for TIP#143 limits, detailing which limits are active in an
+ * interpreter. Used for Tcl_{Add,Remove}LimitHandler type argument.
+ */
+
+#define TCL_LIMIT_COMMANDS	0x01
+#define TCL_LIMIT_TIME		0x02
+
+/*
+ * Structure containing information about a limit handler to be called when a
+ * command- or time-limit is exceeded by an interpreter.
+ */
+
+typedef void (Tcl_LimitHandlerProc) (ClientData clientData, Tcl_Interp *interp);
+typedef void (Tcl_LimitHandlerDeleteProc) (ClientData clientData);
+
+/*
+ *----------------------------------------------------------------------------
+ * Override definitions for libtommath.
+ */
+
+typedef struct mp_int mp_int;
+#define MP_INT_DECLARED
+typedef unsigned int mp_digit;
+#define MP_DIGIT_DECLARED
+
+/*
+ *----------------------------------------------------------------------------
+ * Definitions needed for Tcl_ParseArgvObj routines.
+ * Based on tkArgv.c.
+ * Modifications from the original are copyright (c) Sam Bromley 2006
+ */
+
+typedef struct {
+    int type;			/* Indicates the option type; see below. */
+    const char *keyStr;		/* The key string that flags the option in the
+				 * argv array. */
+    void *srcPtr;		/* Value to be used in setting dst; usage
+				 * depends on type.*/
+    void *dstPtr;		/* Address of value to be modified; usage
+				 * depends on type.*/
+    const char *helpStr;	/* Documentation message describing this
+				 * option. */
+    ClientData clientData;	/* Word to pass to function callbacks. */
+} Tcl_ArgvInfo;
+
+/*
+ * Legal values for the type field of a Tcl_ArgInfo: see the user
+ * documentation for details.
+ */
+
+#define TCL_ARGV_CONSTANT	15
+#define TCL_ARGV_INT		16
+#define TCL_ARGV_STRING		17
+#define TCL_ARGV_REST		18
+#define TCL_ARGV_FLOAT		19
+#define TCL_ARGV_FUNC		20
+#define TCL_ARGV_GENFUNC	21
+#define TCL_ARGV_HELP		22
+#define TCL_ARGV_END		23
+
+/*
+ * Types of callback functions for the TCL_ARGV_FUNC and TCL_ARGV_GENFUNC
+ * argument types:
+ */
+
+typedef int (Tcl_ArgvFuncProc)(ClientData clientData, Tcl_Obj *objPtr,
+	void *dstPtr);
+typedef int (Tcl_ArgvGenFuncProc)(ClientData clientData, Tcl_Interp *interp,
+	int objc, Tcl_Obj *const *objv, void *dstPtr);
+
+/*
+ * Shorthand for commonly used argTable entries.
+ */
+
+#define TCL_ARGV_AUTO_HELP \
+    {TCL_ARGV_HELP,	"-help",	NULL,	NULL, \
+	    "Print summary of command-line options and abort", NULL}
+#define TCL_ARGV_AUTO_REST \
+    {TCL_ARGV_REST,	"--",		NULL,	NULL, \
+	    "Marks the end of the options", NULL}
+#define TCL_ARGV_TABLE_END \
+    {TCL_ARGV_END, NULL, NULL, NULL, NULL, NULL}
+
+/*
+ *----------------------------------------------------------------------------
+ * Definitions needed for Tcl_Zlib routines. [TIP #234]
+ *
+ * Constants for the format flags describing what sort of data format is
+ * desired/expected for the Tcl_ZlibDeflate, Tcl_ZlibInflate and
+ * Tcl_ZlibStreamInit functions.
+ */
+
+#define TCL_ZLIB_FORMAT_RAW	1
+#define TCL_ZLIB_FORMAT_ZLIB	2
+#define TCL_ZLIB_FORMAT_GZIP	4
+#define TCL_ZLIB_FORMAT_AUTO	8
+
+/*
+ * Constants that describe whether the stream is to operate in compressing or
+ * decompressing mode.
+ */
+
+#define TCL_ZLIB_STREAM_DEFLATE	16
+#define TCL_ZLIB_STREAM_INFLATE	32
+
+/*
+ * Constants giving compression levels. Use of TCL_ZLIB_COMPRESS_DEFAULT is
+ * recommended.
+ */
+
+#define TCL_ZLIB_COMPRESS_NONE	0
+#define TCL_ZLIB_COMPRESS_FAST	1
+#define TCL_ZLIB_COMPRESS_BEST	9
+#define TCL_ZLIB_COMPRESS_DEFAULT (-1)
+
+/*
+ * Constants for types of flushing, used with Tcl_ZlibFlush.
+ */
+
+#define TCL_ZLIB_NO_FLUSH	0
+#define TCL_ZLIB_FLUSH		2
+#define TCL_ZLIB_FULLFLUSH	3
+#define TCL_ZLIB_FINALIZE	4
+
+/*
+ *----------------------------------------------------------------------------
+ * Definitions needed for the Tcl_LoadFile function. [TIP #416]
+ */
+
+#define TCL_LOAD_GLOBAL 1
+#define TCL_LOAD_LAZY 2
+
+/*
+ *----------------------------------------------------------------------------
+ * Definitions needed for the Tcl_OpenTcpServerEx function. [TIP #456]
+ */
+#define TCL_TCPSERVER_REUSEADDR (1<<0)
+#define TCL_TCPSERVER_REUSEPORT (1<<1)
+
+/*
+ *----------------------------------------------------------------------------
+ * Single public declaration for NRE.
+ */
+
+typedef int (Tcl_NRPostProc) (ClientData data[], Tcl_Interp *interp,
+				int result);
+
+/*
+ *----------------------------------------------------------------------------
+ * The following constant is used to test for older versions of Tcl in the
+ * stubs tables.
+ */
+
+#define TCL_STUB_MAGIC		((int) 0xFCA3BACF)
+
+/*
+ * The following function is required to be defined in all stubs aware
+ * extensions. The function is actually implemented in the stub library, not
+ * the main Tcl library, although there is a trivial implementation in the
+ * main library in case an extension is statically linked into an application.
+ */
+
+const char *		Tcl_InitStubs(Tcl_Interp *interp, const char *version,
+			    int exact, int magic);
+const char *		TclTomMathInitializeStubs(Tcl_Interp *interp,
+			    const char *version, int epoch, int revision);
+
+#ifdef USE_TCL_STUBS
+#define Tcl_InitStubs(interp, version, exact) \
+    (Tcl_InitStubs)(interp, version, \
+	    (exact)|(TCL_MAJOR_VERSION<<8)|(TCL_MINOR_VERSION<<16), \
+	    TCL_STUB_MAGIC)
+#else
+#define Tcl_InitStubs(interp, version, exact) \
+    Tcl_PkgInitStubsCheck(interp, version, \
+	    (exact)|(TCL_MAJOR_VERSION<<8)|(TCL_MINOR_VERSION<<16))
+#endif
+
+/*
+ * Public functions that are not accessible via the stubs table.
+ * Tcl_GetMemoryInfo is needed for AOLserver. [Bug 1868171]
+ */
+
+#define Tcl_Main(argc, argv, proc) Tcl_MainEx(argc, argv, proc, \
+	    ((Tcl_CreateInterp)()))
+EXTERN void		Tcl_MainEx(int argc, char **argv,
+			    Tcl_AppInitProc *appInitProc, Tcl_Interp *interp);
+EXTERN const char *	Tcl_PkgInitStubsCheck(Tcl_Interp *interp,
+			    const char *version, int exact);
+EXTERN void		Tcl_GetMemoryInfo(Tcl_DString *dsPtr);
+
+/*
+ *----------------------------------------------------------------------------
+ * Include the public function declarations that are accessible via the stubs
+ * table.
+ */
+
+#include "tclDecls.h"
+
+/*
+ * Include platform specific public function declarations that are accessible
+ * via the stubs table. Make all TclOO symbols MODULE_SCOPE (which only
+ * has effect on building it as a shared library). See ticket [3010352].
+ */
+
+#if defined(BUILD_tcl)
+#   undef TCLAPI
+#   define TCLAPI MODULE_SCOPE
+#endif
+
+#include "tclPlatDecls.h"
+
+/*
+ *----------------------------------------------------------------------------
+ * The following declarations either map ckalloc and ckfree to malloc and
+ * free, or they map them to functions with all sorts of debugging hooks
+ * defined in tclCkalloc.c.
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+#   define ckalloc(x) \
+    ((void *) Tcl_DbCkalloc((unsigned)(x), __FILE__, __LINE__))
+#   define ckfree(x) \
+    Tcl_DbCkfree((char *)(x), __FILE__, __LINE__)
+#   define ckrealloc(x,y) \
+    ((void *) Tcl_DbCkrealloc((char *)(x), (unsigned)(y), __FILE__, __LINE__))
+#   define attemptckalloc(x) \
+    ((void *) Tcl_AttemptDbCkalloc((unsigned)(x), __FILE__, __LINE__))
+#   define attemptckrealloc(x,y) \
+    ((void *) Tcl_AttemptDbCkrealloc((char *)(x), (unsigned)(y), __FILE__, __LINE__))
+
+#else /* !TCL_MEM_DEBUG */
+
+/*
+ * If we are not using the debugging allocator, we should call the Tcl_Alloc,
+ * et al. routines in order to guarantee that every module is using the same
+ * memory allocator both inside and outside of the Tcl library.
+ */
+
+#   define ckalloc(x) \
+    ((void *) Tcl_Alloc((unsigned)(x)))
+#   define ckfree(x) \
+    Tcl_Free((char *)(x))
+#   define ckrealloc(x,y) \
+    ((void *) Tcl_Realloc((char *)(x), (unsigned)(y)))
+#   define attemptckalloc(x) \
+    ((void *) Tcl_AttemptAlloc((unsigned)(x)))
+#   define attemptckrealloc(x,y) \
+    ((void *) Tcl_AttemptRealloc((char *)(x), (unsigned)(y)))
+#   undef  Tcl_InitMemory
+#   define Tcl_InitMemory(x)
+#   undef  Tcl_DumpActiveMemory
+#   define Tcl_DumpActiveMemory(x)
+#   undef  Tcl_ValidateAllMemory
+#   define Tcl_ValidateAllMemory(x,y)
+
+#endif /* !TCL_MEM_DEBUG */
+
+#ifdef TCL_MEM_DEBUG
+#   define Tcl_IncrRefCount(objPtr) \
+	Tcl_DbIncrRefCount(objPtr, __FILE__, __LINE__)
+#   define Tcl_DecrRefCount(objPtr) \
+	Tcl_DbDecrRefCount(objPtr, __FILE__, __LINE__)
+#   define Tcl_IsShared(objPtr) \
+	Tcl_DbIsShared(objPtr, __FILE__, __LINE__)
+#else
+#   define Tcl_IncrRefCount(objPtr) \
+	++(objPtr)->refCount
+    /*
+     * Use do/while0 idiom for optimum correctness without compiler warnings.
+     * http://c2.com/cgi/wiki?TrivialDoWhileLoop
+     */
+#   define Tcl_DecrRefCount(objPtr) \
+	do { \
+	    Tcl_Obj *_objPtr = (objPtr); \
+	    if ((_objPtr)->refCount-- <= 1) { \
+		TclFreeObj(_objPtr); \
+	    } \
+	} while(0)
+#   define Tcl_IsShared(objPtr) \
+	((objPtr)->refCount > 1)
+#endif
+
+/*
+ * Macros and definitions that help to debug the use of Tcl objects. When
+ * TCL_MEM_DEBUG is defined, the Tcl_New declarations are overridden to call
+ * debugging versions of the object creation functions.
+ */
+
+#ifdef TCL_MEM_DEBUG
+#  undef  Tcl_NewBignumObj
+#  define Tcl_NewBignumObj(val) \
+     Tcl_DbNewBignumObj(val, __FILE__, __LINE__)
+#  undef  Tcl_NewBooleanObj
+#  define Tcl_NewBooleanObj(val) \
+     Tcl_DbNewLongObj((val)!=0, __FILE__, __LINE__)
+#  undef  Tcl_NewByteArrayObj
+#  define Tcl_NewByteArrayObj(bytes, len) \
+     Tcl_DbNewByteArrayObj(bytes, len, __FILE__, __LINE__)
+#  undef  Tcl_NewDoubleObj
+#  define Tcl_NewDoubleObj(val) \
+     Tcl_DbNewDoubleObj(val, __FILE__, __LINE__)
+#  undef  Tcl_NewIntObj
+#  define Tcl_NewIntObj(val) \
+     Tcl_DbNewLongObj(val, __FILE__, __LINE__)
+#  undef  Tcl_NewListObj
+#  define Tcl_NewListObj(objc, objv) \
+     Tcl_DbNewListObj(objc, objv, __FILE__, __LINE__)
+#  undef  Tcl_NewLongObj
+#  define Tcl_NewLongObj(val) \
+     Tcl_DbNewLongObj(val, __FILE__, __LINE__)
+#  undef  Tcl_NewObj
+#  define Tcl_NewObj() \
+     Tcl_DbNewObj(__FILE__, __LINE__)
+#  undef  Tcl_NewStringObj
+#  define Tcl_NewStringObj(bytes, len) \
+     Tcl_DbNewStringObj(bytes, len, __FILE__, __LINE__)
+#  undef  Tcl_NewWideIntObj
+#  define Tcl_NewWideIntObj(val) \
+     Tcl_DbNewWideIntObj(val, __FILE__, __LINE__)
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------------
+ * Macros for clients to use to access fields of hash entries:
+ */
+
+#define Tcl_GetHashValue(h) ((h)->clientData)
+#define Tcl_SetHashValue(h, value) ((h)->clientData = (ClientData) (value))
+#define Tcl_GetHashKey(tablePtr, h) \
+	((void *) (((tablePtr)->keyType == TCL_ONE_WORD_KEYS || \
+		    (tablePtr)->keyType == TCL_CUSTOM_PTR_KEYS) \
+		   ? (h)->key.oneWordValue \
+		   : (h)->key.string))
+
+/*
+ * Macros to use for clients to use to invoke find and create functions for
+ * hash tables:
+ */
+
+#undef  Tcl_FindHashEntry
+#define Tcl_FindHashEntry(tablePtr, key) \
+	(*((tablePtr)->findProc))(tablePtr, (const char *)(key))
+#undef  Tcl_CreateHashEntry
+#define Tcl_CreateHashEntry(tablePtr, key, newPtr) \
+	(*((tablePtr)->createProc))(tablePtr, (const char *)(key), newPtr)
+
+/*
+ *----------------------------------------------------------------------------
+ * Macros that eliminate the overhead of the thread synchronization functions
+ * when compiling without thread support.
+ */
+
+#ifndef TCL_THREADS
+#undef  Tcl_MutexLock
+#define Tcl_MutexLock(mutexPtr)
+#undef  Tcl_MutexUnlock
+#define Tcl_MutexUnlock(mutexPtr)
+#undef  Tcl_MutexFinalize
+#define Tcl_MutexFinalize(mutexPtr)
+#undef  Tcl_ConditionNotify
+#define Tcl_ConditionNotify(condPtr)
+#undef  Tcl_ConditionWait
+#define Tcl_ConditionWait(condPtr, mutexPtr, timePtr)
+#undef  Tcl_ConditionFinalize
+#define Tcl_ConditionFinalize(condPtr)
+#endif /* TCL_THREADS */
+
+/*
+ *----------------------------------------------------------------------------
+ * Deprecated Tcl functions:
+ */
+
+#ifndef TCL_NO_DEPRECATED
+/*
+ * These function have been renamed. The old names are deprecated, but we
+ * define these macros for backwards compatibilty.
+ */
+
+#   define Tcl_Ckalloc		Tcl_Alloc
+#   define Tcl_Ckfree		Tcl_Free
+#   define Tcl_Ckrealloc	Tcl_Realloc
+#   define Tcl_Return		Tcl_SetResult
+#   define Tcl_TildeSubst	Tcl_TranslateFileName
+#if !defined(__APPLE__) /* On OSX, there is a conflict with "mach/mach.h" */
+#   define panic		Tcl_Panic
+#endif
+#   define panicVA		Tcl_PanicVA
+
+/*
+ *----------------------------------------------------------------------------
+ * Convenience declaration of Tcl_AppInit for backwards compatibility. This
+ * function is not *implemented* by the tcl library, so the storage class is
+ * neither DLLEXPORT nor DLLIMPORT.
+ */
+
+extern Tcl_AppInitProc Tcl_AppInit;
+
+#endif /* !TCL_NO_DEPRECATED */
+
+#endif /* RC_INVOKED */
+
+/*
+ * end block for C++
+ */
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _TCL */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclAlloc.c b/generic/tclAlloc.c
new file mode 100644
index 0000000..64df1a2
--- /dev/null
+++ b/generic/tclAlloc.c
@@ -0,0 +1,759 @@
+/*
+ * tclAlloc.c --
+ *
+ *	This is a very fast storage allocator. It allocates blocks of a small
+ *	number of different sizes, and keeps free lists of each size. Blocks
+ *	that don't exactly fit are passed up to the next larger size. Blocks
+ *	over a certain size are directly allocated from the system.
+ *
+ * Copyright (c) 1983 Regents of the University of California.
+ * Copyright (c) 1996-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * Portions contributed by Chris Kingsley, Jack Jansen and Ray Johnson.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/*
+ * Windows and Unix use an alternative allocator when building with threads
+ * that has significantly reduced lock contention.
+ */
+
+#include "tclInt.h"
+#if !defined(TCL_THREADS) || !defined(USE_THREAD_ALLOC)
+
+#if USE_TCLALLOC
+
+/*
+ * We should really make use of AC_CHECK_TYPE(caddr_t) here, but it can wait
+ * until Tcl uses config.h properly.
+ */
+
+#if defined(_MSC_VER) || defined(__MSVCRT__) || defined(__BORLANDC__)
+typedef size_t caddr_t;
+#endif
+
+/*
+ * The overhead on a block is at least 8 bytes. When free, this space contains
+ * a pointer to the next free block, and the bottom two bits must be zero.
+ * When in use, the first byte is set to MAGIC, and the second byte is the
+ * size index. The remaining bytes are for alignment. If range checking is
+ * enabled then a second word holds the size of the requested block, less 1,
+ * rounded up to a multiple of sizeof(RMAGIC). The order of elements is
+ * critical: ov.magic must overlay the low order bits of ov.next, and ov.magic
+ * can not be a valid ov.next bit pattern.
+ */
+
+union overhead {
+    union overhead *next;		/* when free */
+    unsigned char padding[TCL_ALLOCALIGN];	/* align struct to TCL_ALLOCALIGN bytes */
+    struct {
+	unsigned char magic0;		/* magic number */
+	unsigned char index;		/* bucket # */
+	unsigned char unused;		/* unused */
+	unsigned char magic1;		/* other magic number */
+#ifndef NDEBUG
+	unsigned short rmagic;		/* range magic number */
+	size_t size;		/* actual block size */
+	unsigned short unused2;		/* padding to 8-byte align */
+#endif
+    } ovu;
+#define overMagic0	ovu.magic0
+#define overMagic1	ovu.magic1
+#define bucketIndex	ovu.index
+#define rangeCheckMagic	ovu.rmagic
+#define realBlockSize	ovu.size
+};
+
+
+#define MAGIC		0xef	/* magic # on accounting info */
+#define RMAGIC		0x5555	/* magic # on range info */
+
+#ifndef NDEBUG
+#define	RSLOP		sizeof(unsigned short)
+#else
+#define	RSLOP		0
+#endif
+
+#define OVERHEAD (sizeof(union overhead) + RSLOP)
+
+/*
+ * Macro to make it easier to refer to the end-of-block guard magic.
+ */
+
+#define BLOCK_END(overPtr) \
+    (*(unsigned short *)((caddr_t)((overPtr) + 1) + (overPtr)->realBlockSize))
+
+/*
+ * nextf[i] is the pointer to the next free block of size 2^(i+3). The
+ * smallest allocatable block is MINBLOCK bytes. The overhead information
+ * precedes the data area returned to the user.
+ */
+
+#define MINBLOCK	((sizeof(union overhead) + (TCL_ALLOCALIGN-1)) & ~(TCL_ALLOCALIGN-1))
+#define NBUCKETS	(13 - (MINBLOCK >> 4))
+#define MAXMALLOC	(1<<(NBUCKETS+2))
+static union overhead *nextf[NBUCKETS];
+
+/*
+ * The following structure is used to keep track of all system memory
+ * currently owned by Tcl. When finalizing, all this memory will be returned
+ * to the system.
+ */
+
+struct block {
+    struct block *nextPtr;	/* Linked list. */
+    struct block *prevPtr;	/* Linked list for big blocks, ensures 8-byte
+				 * alignment for suballocated blocks. */
+};
+
+static struct block *blockList;	/* Tracks the suballocated blocks. */
+static struct block bigBlocks={	/* Big blocks aren't suballocated. */
+    &bigBlocks, &bigBlocks
+};
+
+/*
+ * The allocator is protected by a special mutex that must be explicitly
+ * initialized. Futhermore, because Tcl_Alloc may be used before anything else
+ * in Tcl, we make this module self-initializing after all with the allocInit
+ * variable.
+ */
+
+#ifdef TCL_THREADS
+static Tcl_Mutex *allocMutexPtr;
+#endif
+static int allocInit = 0;
+
+#ifdef MSTATS
+
+/*
+ * numMallocs[i] is the difference between the number of mallocs and frees for
+ * a given block size.
+ */
+
+static	size_t numMallocs[NBUCKETS+1];
+#endif
+
+#if !defined(NDEBUG)
+#define	ASSERT(p)	if (!(p)) Tcl_Panic(# p)
+#define RANGE_ASSERT(p) if (!(p)) Tcl_Panic(# p)
+#else
+#define	ASSERT(p)
+#define RANGE_ASSERT(p)
+#endif
+
+/*
+ * Prototypes for functions used only in this file.
+ */
+
+static void		MoreCore(size_t bucket);
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclInitAlloc --
+ *
+ *	Initialize the memory system.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Initialize the mutex used to serialize allocations.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+TclInitAlloc(void)
+{
+    if (!allocInit) {
+	allocInit = 1;
+#ifdef TCL_THREADS
+	allocMutexPtr = Tcl_GetAllocMutex();
+#endif
+    }
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclFinalizeAllocSubsystem --
+ *
+ *	Release all resources being used by this subsystem, including
+ *	aggressively freeing all memory allocated by TclpAlloc() that has not
+ *	yet been released with TclpFree().
+ *
+ *	After this function is called, all memory allocated with TclpAlloc()
+ *	should be considered unusable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	This subsystem is self-initializing, since memory can be allocated
+ *	before Tcl is formally initialized. After this call, this subsystem
+ *	has been reset to its initial state and is usable again.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+TclFinalizeAllocSubsystem(void)
+{
+    unsigned int i;
+    struct block *blockPtr, *nextPtr;
+
+    Tcl_MutexLock(allocMutexPtr);
+    for (blockPtr = blockList; blockPtr != NULL; blockPtr = nextPtr) {
+	nextPtr = blockPtr->nextPtr;
+	TclpSysFree(blockPtr);
+    }
+    blockList = NULL;
+
+    for (blockPtr = bigBlocks.nextPtr; blockPtr != &bigBlocks; ) {
+	nextPtr = blockPtr->nextPtr;
+	TclpSysFree(blockPtr);
+	blockPtr = nextPtr;
+    }
+    bigBlocks.nextPtr = &bigBlocks;
+    bigBlocks.prevPtr = &bigBlocks;
+
+    for (i=0 ; i<NBUCKETS ; i++) {
+	nextf[i] = NULL;
+#ifdef MSTATS
+	numMallocs[i] = 0;
+#endif
+    }
+#ifdef MSTATS
+    numMallocs[i] = 0;
+#endif
+    Tcl_MutexUnlock(allocMutexPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpAlloc --
+ *
+ *	Allocate more memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpAlloc(
+    unsigned int numBytes)	/* Number of bytes to allocate. */
+{
+    register union overhead *overPtr;
+    register size_t bucket;
+    register unsigned amount;
+    struct block *bigBlockPtr = NULL;
+
+    if (!allocInit) {
+	/*
+	 * We have to make the "self initializing" because Tcl_Alloc may be
+	 * used before any other part of Tcl. E.g., see main() for tclsh!
+	 */
+
+	TclInitAlloc();
+    }
+    Tcl_MutexLock(allocMutexPtr);
+
+    /*
+     * First the simple case: we simple allocate big blocks directly.
+     */
+
+    if (numBytes >= MAXMALLOC - OVERHEAD) {
+	if (numBytes <= UINT_MAX - OVERHEAD -sizeof(struct block)) {
+	    bigBlockPtr = (struct block *) TclpSysAlloc((unsigned)
+		    (sizeof(struct block) + OVERHEAD + numBytes), 0);
+	}
+	if (bigBlockPtr == NULL) {
+	    Tcl_MutexUnlock(allocMutexPtr);
+	    return NULL;
+	}
+	bigBlockPtr->nextPtr = bigBlocks.nextPtr;
+	bigBlocks.nextPtr = bigBlockPtr;
+	bigBlockPtr->prevPtr = &bigBlocks;
+	bigBlockPtr->nextPtr->prevPtr = bigBlockPtr;
+
+	overPtr = (union overhead *) (bigBlockPtr + 1);
+	overPtr->overMagic0 = overPtr->overMagic1 = MAGIC;
+	overPtr->bucketIndex = 0xff;
+#ifdef MSTATS
+	numMallocs[NBUCKETS]++;
+#endif
+
+#ifndef NDEBUG
+	/*
+	 * Record allocated size of block and bound space with magic numbers.
+	 */
+
+	overPtr->realBlockSize = (numBytes + RSLOP - 1) & ~(RSLOP - 1);
+	overPtr->rangeCheckMagic = RMAGIC;
+	BLOCK_END(overPtr) = RMAGIC;
+#endif
+
+	Tcl_MutexUnlock(allocMutexPtr);
+	return (void *)(overPtr+1);
+    }
+
+    /*
+     * Convert amount of memory requested into closest block size stored in
+     * hash buckets which satisfies request. Account for space used per block
+     * for accounting.
+     */
+
+    amount = MINBLOCK;		/* size of first bucket */
+    bucket = MINBLOCK >> 4;
+
+    while (numBytes + OVERHEAD > amount) {
+	amount <<= 1;
+	if (amount == 0) {
+	    Tcl_MutexUnlock(allocMutexPtr);
+	    return NULL;
+	}
+	bucket++;
+    }
+    ASSERT(bucket < NBUCKETS);
+
+    /*
+     * If nothing in hash bucket right now, request more memory from the
+     * system.
+     */
+
+    if ((overPtr = nextf[bucket]) == NULL) {
+	MoreCore(bucket);
+	if ((overPtr = nextf[bucket]) == NULL) {
+	    Tcl_MutexUnlock(allocMutexPtr);
+	    return NULL;
+	}
+    }
+
+    /*
+     * Remove from linked list
+     */
+
+    nextf[bucket] = overPtr->next;
+    overPtr->overMagic0 = overPtr->overMagic1 = MAGIC;
+    overPtr->bucketIndex = (unsigned char) bucket;
+
+#ifdef MSTATS
+    numMallocs[bucket]++;
+#endif
+
+#ifndef NDEBUG
+    /*
+     * Record allocated size of block and bound space with magic numbers.
+     */
+
+    overPtr->realBlockSize = (numBytes + RSLOP - 1) & ~(RSLOP - 1);
+    overPtr->rangeCheckMagic = RMAGIC;
+    BLOCK_END(overPtr) = RMAGIC;
+#endif
+
+    Tcl_MutexUnlock(allocMutexPtr);
+    return ((char *)(overPtr + 1));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MoreCore --
+ *
+ *	Allocate more memory to the indicated bucket.
+ *
+ *	Assumes Mutex is already held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Attempts to get more memory from the system.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MoreCore(
+    size_t bucket)	/* What bucket to allocate to. */
+{
+    register union overhead *overPtr;
+    register size_t size;	/* size of desired block */
+    size_t amount;		/* amount to allocate */
+    size_t numBlocks;		/* how many blocks we get */
+    struct block *blockPtr;
+
+    /*
+     * sbrk_size <= 0 only for big, FLUFFY, requests (about 2^30 bytes on a
+     * VAX, I think) or for a negative arg.
+     */
+
+    size = ((size_t)1) << (bucket + 3);
+    ASSERT(size > 0);
+
+    amount = MAXMALLOC;
+    numBlocks = amount / size;
+    ASSERT(numBlocks*size == amount);
+
+    blockPtr = (struct block *) TclpSysAlloc(
+	    (sizeof(struct block) + amount), 1);
+    /* no more room! */
+    if (blockPtr == NULL) {
+	return;
+    }
+    blockPtr->nextPtr = blockList;
+    blockList = blockPtr;
+
+    overPtr = (union overhead *) (blockPtr + 1);
+
+    /*
+     * Add new memory allocated to that on free list for this hash bucket.
+     */
+
+    nextf[bucket] = overPtr;
+    while (--numBlocks > 0) {
+	overPtr->next = (union overhead *)((caddr_t)overPtr + size);
+	overPtr = (union overhead *)((caddr_t)overPtr + size);
+    }
+    overPtr->next = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpFree --
+ *
+ *	Free memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclpFree(
+    char *oldPtr)		/* Pointer to memory to free. */
+{
+    register size_t size;
+    register union overhead *overPtr;
+    struct block *bigBlockPtr;
+
+    if (oldPtr == NULL) {
+	return;
+    }
+
+    Tcl_MutexLock(allocMutexPtr);
+    overPtr = (union overhead *)((caddr_t)oldPtr - sizeof(union overhead));
+
+    ASSERT(overPtr->overMagic0 == MAGIC);	/* make sure it was in use */
+    ASSERT(overPtr->overMagic1 == MAGIC);
+    if (overPtr->overMagic0 != MAGIC || overPtr->overMagic1 != MAGIC) {
+	Tcl_MutexUnlock(allocMutexPtr);
+	return;
+    }
+
+    RANGE_ASSERT(overPtr->rangeCheckMagic == RMAGIC);
+    RANGE_ASSERT(BLOCK_END(overPtr) == RMAGIC);
+    size = overPtr->bucketIndex;
+    if (size == 0xff) {
+#ifdef MSTATS
+	numMallocs[NBUCKETS]--;
+#endif
+
+	bigBlockPtr = (struct block *) overPtr - 1;
+	bigBlockPtr->prevPtr->nextPtr = bigBlockPtr->nextPtr;
+	bigBlockPtr->nextPtr->prevPtr = bigBlockPtr->prevPtr;
+	TclpSysFree(bigBlockPtr);
+
+	Tcl_MutexUnlock(allocMutexPtr);
+	return;
+    }
+    ASSERT(size < NBUCKETS);
+    overPtr->next = nextf[size];	/* also clobbers overMagic */
+    nextf[size] = overPtr;
+
+#ifdef MSTATS
+    numMallocs[size]--;
+#endif
+
+    Tcl_MutexUnlock(allocMutexPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpRealloc --
+ *
+ *	Reallocate memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpRealloc(
+    char *oldPtr,		/* Pointer to alloced block. */
+    unsigned int numBytes)	/* New size of memory. */
+{
+    int i;
+    union overhead *overPtr;
+    struct block *bigBlockPtr;
+    int expensive;
+    size_t maxSize;
+
+    if (oldPtr == NULL) {
+	return TclpAlloc(numBytes);
+    }
+
+    Tcl_MutexLock(allocMutexPtr);
+
+    overPtr = (union overhead *)((caddr_t)oldPtr - sizeof(union overhead));
+
+    ASSERT(overPtr->overMagic0 == MAGIC);	/* make sure it was in use */
+    ASSERT(overPtr->overMagic1 == MAGIC);
+    if (overPtr->overMagic0 != MAGIC || overPtr->overMagic1 != MAGIC) {
+	Tcl_MutexUnlock(allocMutexPtr);
+	return NULL;
+    }
+
+    RANGE_ASSERT(overPtr->rangeCheckMagic == RMAGIC);
+    RANGE_ASSERT(BLOCK_END(overPtr) == RMAGIC);
+    i = overPtr->bucketIndex;
+
+    /*
+     * If the block isn't in a bin, just realloc it.
+     */
+
+    if (i == 0xff) {
+	struct block *prevPtr, *nextPtr;
+	bigBlockPtr = (struct block *) overPtr - 1;
+	prevPtr = bigBlockPtr->prevPtr;
+	nextPtr = bigBlockPtr->nextPtr;
+	bigBlockPtr = (struct block *) TclpSysRealloc(bigBlockPtr,
+		sizeof(struct block) + OVERHEAD + numBytes);
+	if (bigBlockPtr == NULL) {
+	    Tcl_MutexUnlock(allocMutexPtr);
+	    return NULL;
+	}
+
+	if (prevPtr->nextPtr != bigBlockPtr) {
+	    /*
+	     * If the block has moved, splice the new block into the list
+	     * where the old block used to be.
+	     */
+
+	    prevPtr->nextPtr = bigBlockPtr;
+	    nextPtr->prevPtr = bigBlockPtr;
+	}
+
+	overPtr = (union overhead *) (bigBlockPtr + 1);
+
+#ifdef MSTATS
+	numMallocs[NBUCKETS]++;
+#endif
+
+#ifndef NDEBUG
+	/*
+	 * Record allocated size of block and update magic number bounds.
+	 */
+
+	overPtr->realBlockSize = (numBytes + RSLOP - 1) & ~(RSLOP - 1);
+	BLOCK_END(overPtr) = RMAGIC;
+#endif
+
+	Tcl_MutexUnlock(allocMutexPtr);
+	return (char *)(overPtr+1);
+    }
+    maxSize = 1 << (i+3);
+    expensive = 0;
+    if (numBytes+OVERHEAD > maxSize) {
+	expensive = 1;
+    } else if (i>0 && numBytes+OVERHEAD < maxSize/2) {
+	expensive = 1;
+    }
+
+    if (expensive) {
+	void *newPtr;
+
+	Tcl_MutexUnlock(allocMutexPtr);
+
+	newPtr = TclpAlloc(numBytes);
+	if (newPtr == NULL) {
+	    return NULL;
+	}
+	maxSize -= OVERHEAD;
+	if (maxSize < numBytes) {
+	    numBytes = maxSize;
+	}
+	memcpy(newPtr, oldPtr, (size_t) numBytes);
+	TclpFree(oldPtr);
+	return newPtr;
+    }
+
+    /*
+     * Ok, we don't have to copy, it fits as-is
+     */
+
+#ifndef NDEBUG
+    overPtr->realBlockSize = (numBytes + RSLOP - 1) & ~(RSLOP - 1);
+    BLOCK_END(overPtr) = RMAGIC;
+#endif
+
+    Tcl_MutexUnlock(allocMutexPtr);
+    return(oldPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * mstats --
+ *
+ *	Prints two lines of numbers, one showing the length of the free list
+ *	for each size category, the second showing the number of mallocs -
+ *	frees for each size category.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef MSTATS
+void
+mstats(
+    char *s)			/* Where to write info. */
+{
+    register unsigned int i, j;
+    register union overhead *overPtr;
+    size_t totalFree = 0, totalUsed = 0;
+
+    Tcl_MutexLock(allocMutexPtr);
+
+    fprintf(stderr, "Memory allocation statistics %s\nTclpFree:\t", s);
+    for (i = 0; i < NBUCKETS; i++) {
+	for (j=0, overPtr=nextf[i]; overPtr; overPtr=overPtr->next, j++) {
+	    fprintf(stderr, " %u", j);
+	}
+	totalFree += ((size_t)j) * (1 << (i + 3));
+    }
+
+    fprintf(stderr, "\nused:\t");
+    for (i = 0; i < NBUCKETS; i++) {
+	fprintf(stderr, " %" TCL_LL_MODIFIER "d", (Tcl_WideInt)numMallocs[i]);
+	totalUsed += numMallocs[i] * (1 << (i + 3));
+    }
+
+    fprintf(stderr, "\n\tTotal small in use: %" TCL_LL_MODIFIER "d, total free: %" TCL_LL_MODIFIER "d\n",
+	(Tcl_WideInt)totalUsed, (Tcl_WideInt)totalFree);
+    fprintf(stderr, "\n\tNumber of big (>%d) blocks in use: %" TCL_LL_MODIFIER "d\n",
+	    MAXMALLOC, (Tcl_WideInt)numMallocs[NBUCKETS]);
+
+    Tcl_MutexUnlock(allocMutexPtr);
+}
+#endif
+
+#else	/* !USE_TCLALLOC */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpAlloc --
+ *
+ *	Allocate more memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpAlloc(
+    unsigned int numBytes)	/* Number of bytes to allocate. */
+{
+    return (char *) malloc(numBytes);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpFree --
+ *
+ *	Free memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclpFree(
+    char *oldPtr)		/* Pointer to memory to free. */
+{
+    free(oldPtr);
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpRealloc --
+ *
+ *	Reallocate memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpRealloc(
+    char *oldPtr,		/* Pointer to alloced block. */
+    unsigned int numBytes)	/* New size of memory. */
+{
+    return (char *) realloc(oldPtr, numBytes);
+}
+
+#endif /* !USE_TCLALLOC */
+#endif /* !TCL_THREADS */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclAssembly.c b/generic/tclAssembly.c
new file mode 100644
index 0000000..4c5ae68
--- /dev/null
+++ b/generic/tclAssembly.c
@@ -0,0 +1,4345 @@
+/*
+ * tclAssembly.c --
+ *
+ *	Assembler for Tcl bytecodes.
+ *
+ * This file contains the procedures that convert Tcl Assembly Language (TAL)
+ * to a sequence of bytecode instructions for the Tcl execution engine.
+ *
+ * Copyright (c) 2010 by Ozgur Dogan Ugurlu.
+ * Copyright (c) 2010 by Kevin B. Kenny.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/*-
+ *- THINGS TO DO:
+ *- More instructions:
+ *-   done - alternate exit point (affects stack and exception range checking)
+ *-   break and continue - if exception ranges can be sorted out.
+ *-   foreach_start4, foreach_step4
+ *-   returnImm, returnStk
+ *-   expandStart, expandStkTop, invokeExpanded, expandDrop
+ *-   dictFirst, dictNext, dictDone
+ *-   dictUpdateStart, dictUpdateEnd
+ *-   jumpTable testing
+ *-   syntax (?)
+ *-   returnCodeBranch
+ *-   tclooNext, tclooNextClass
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include "tclOOInt.h"
+
+/*
+ * Structure that represents a range of instructions in the bytecode.
+ */
+
+typedef struct CodeRange {
+    int startOffset;		/* Start offset in the bytecode array */
+    int endOffset;		/* End offset in the bytecode array */
+} CodeRange;
+
+/*
+ * State identified for a basic block's catch context.
+ */
+
+typedef enum BasicBlockCatchState {
+    BBCS_UNKNOWN = 0,		/* Catch context has not yet been identified */
+    BBCS_NONE,			/* Block is outside of any catch */
+    BBCS_INCATCH,		/* Block is within a catch context */
+    BBCS_CAUGHT 		/* Block is within a catch context and
+				 * may be executed after an exception fires */
+} BasicBlockCatchState;
+
+/*
+ * Structure that defines a basic block - a linear sequence of bytecode
+ * instructions with no jumps in or out (including not changing the
+ * state of any exception range).
+ */
+
+typedef struct BasicBlock {
+    int originalStartOffset;	/* Instruction offset before JUMP1s were
+				 * substituted with JUMP4's */
+    int startOffset;		/* Instruction offset of the start of the
+				 * block */
+    int startLine;		/* Line number in the input script of the
+				 * instruction at the start of the block */
+    int jumpOffset;		/* Bytecode offset of the 'jump' instruction
+				 * that ends the block, or -1 if there is no
+				 * jump. */
+    int jumpLine;		/* Line number in the input script of the
+				 * 'jump' instruction that ends the block, or
+				 * -1 if there is no jump */
+    struct BasicBlock* prevPtr;	/* Immediate predecessor of this block */
+    struct BasicBlock* predecessor;
+				/* Predecessor of this block in the spanning
+				 * tree */
+    struct BasicBlock* successor1;
+				/* BasicBlock structure of the following
+				 * block: NULL at the end of the bytecode
+				 * sequence. */
+    Tcl_Obj* jumpTarget;	/* Jump target label if the jump target is
+				 * unresolved */
+    int initialStackDepth;	/* Absolute stack depth on entry */
+    int minStackDepth;		/* Low-water relative stack depth */
+    int maxStackDepth;		/* High-water relative stack depth */
+    int finalStackDepth;	/* Relative stack depth on exit */
+    enum BasicBlockCatchState catchState;
+				/* State of the block for 'catch' analysis */
+    int catchDepth;		/* Number of nested catches in which the basic
+				 * block appears */
+    struct BasicBlock* enclosingCatch;
+				/* BasicBlock structure of the last startCatch
+				 * executed on a path to this block, or NULL
+				 * if there is no enclosing catch */
+    int foreignExceptionBase;	/* Base index of foreign exceptions */
+    int foreignExceptionCount;	/* Count of foreign exceptions */
+    ExceptionRange* foreignExceptions;
+				/* ExceptionRange structures for exception
+				 * ranges belonging to embedded scripts and
+				 * expressions in this block */
+    JumptableInfo* jtPtr;	/* Jump table at the end of this basic block */
+    int flags;			/* Boolean flags */
+} BasicBlock;
+
+/*
+ * Flags that pertain to a basic block.
+ */
+
+enum BasicBlockFlags {
+    BB_VISITED = (1 << 0),	/* Block has been visited in the current
+				 * traversal */
+    BB_FALLTHRU = (1 << 1),	/* Control may pass from this block to a
+				 * successor */
+    BB_JUMP1 = (1 << 2),	/* Basic block ends with a 1-byte-offset jump
+				 * and may need expansion */
+    BB_JUMPTABLE = (1 << 3),	/* Basic block ends with a jump table */
+    BB_BEGINCATCH = (1 << 4),	/* Block ends with a 'beginCatch' instruction,
+				 * marking it as the start of a 'catch'
+				 * sequence. The 'jumpTarget' is the exception
+				 * exit from the catch block. */
+    BB_ENDCATCH = (1 << 5)	/* Block ends with an 'endCatch' instruction,
+				 * unwinding the catch from the exception
+				 * stack. */
+};
+
+/*
+ * Source instruction type recognized by the assembler.
+ */
+
+typedef enum TalInstType {
+    ASSEM_1BYTE,		/* Fixed arity, 1-byte instruction */
+    ASSEM_BEGIN_CATCH,		/* Begin catch: one 4-byte jump offset to be
+				 * converted to appropriate exception
+				 * ranges */
+    ASSEM_BOOL,			/* One Boolean operand */
+    ASSEM_BOOL_LVT4,		/* One Boolean, one 4-byte LVT ref. */
+    ASSEM_CLOCK_READ,		/* 1-byte unsigned-integer case number, in the
+				 * range 0-3 */
+    ASSEM_CONCAT1,		/* 1-byte unsigned-integer operand count, must
+				 * be strictly positive, consumes N, produces
+				 * 1 */
+    ASSEM_DICT_GET,		/* 'dict get' and related - consumes N+1
+				 * operands, produces 1, N > 0 */
+    ASSEM_DICT_SET,		/* specifies key count and LVT index, consumes
+				 * N+1 operands, produces 1, N > 0 */
+    ASSEM_DICT_UNSET,		/* specifies key count and LVT index, consumes
+				 * N operands, produces 1, N > 0 */
+    ASSEM_END_CATCH,		/* End catch. No args. Exception range popped
+				 * from stack and stack pointer restored. */
+    ASSEM_EVAL,			/* 'eval' - evaluate a constant script (by
+				 * compiling it in line with the assembly
+				 * code! I love Tcl!) */
+    ASSEM_INDEX,		/* 4 byte operand, integer or end-integer */
+    ASSEM_INVOKE,		/* 1- or 4-byte operand count, must be
+				 * strictly positive, consumes N, produces
+				 * 1. */
+    ASSEM_JUMP,			/* Jump instructions */
+    ASSEM_JUMP4,		/* Jump instructions forcing a 4-byte offset */
+    ASSEM_JUMPTABLE,		/* Jumptable (switch -exact) */
+    ASSEM_LABEL,		/* The assembly directive that defines a
+				 * label */
+    ASSEM_LINDEX_MULTI,		/* 4-byte operand count, must be strictly
+				 * positive, consumes N, produces 1 */
+    ASSEM_LIST,			/* 4-byte operand count, must be nonnegative,
+				 * consumses N, produces 1 */
+    ASSEM_LSET_FLAT,		/* 4-byte operand count, must be >= 3,
+				 * consumes N, produces 1 */
+    ASSEM_LVT,			/* One operand that references a local
+				 * variable */
+    ASSEM_LVT1,			/* One 1-byte operand that references a local
+				 * variable */
+    ASSEM_LVT1_SINT1,		/* One 1-byte operand that references a local
+				 * variable, one signed-integer 1-byte
+				 * operand */
+    ASSEM_LVT4,			/* One 4-byte operand that references a local
+				 * variable */
+    ASSEM_OVER,			/* OVER: 4-byte operand count, consumes N+1,
+				 * produces N+2 */
+    ASSEM_PUSH,			/* one literal operand */
+    ASSEM_REGEXP,		/* One Boolean operand, but weird mapping to
+				 * call flags */
+    ASSEM_REVERSE,		/* REVERSE: 4-byte operand count, consumes N,
+				 * produces N */
+    ASSEM_SINT1,		/* One 1-byte signed-integer operand
+				 * (INCR_STK_IMM) */
+    ASSEM_SINT4_LVT4		/* Signed 4-byte integer operand followed by
+				 * LVT entry.  Fixed arity */
+} TalInstType;
+
+/*
+ * Description of an instruction recognized by the assembler.
+ */
+
+typedef struct TalInstDesc {
+    const char *name;		/* Name of instruction. */
+    TalInstType instType;	/* The type of instruction */
+    int tclInstCode;		/* Instruction code. For instructions having
+				 * 1- and 4-byte variables, tclInstCode is
+				 * ((1byte)<<8) || (4byte) */
+    int operandsConsumed;	/* Number of operands consumed by the
+				 * operation, or INT_MIN if the operation is
+				 * variadic */
+    int operandsProduced;	/* Number of operands produced by the
+				 * operation. If negative, the operation has a
+				 * net stack effect of -1-operandsProduced */
+} TalInstDesc;
+
+/*
+ * Structure that holds the state of the assembler while generating code.
+ */
+
+typedef struct AssemblyEnv {
+    CompileEnv* envPtr;		/* Compilation environment being used for code
+				 * generation */
+    Tcl_Parse* parsePtr;	/* Parse of the current line of source */
+    Tcl_HashTable labelHash;	/* Hash table whose keys are labels and whose
+				 * values are 'label' objects storing the code
+				 * offsets of the labels. */
+    int cmdLine;		/* Current line number within the assembly
+				 * code */
+    int* clNext;		/* Invisible continuation line for
+				 * [info frame] */
+    BasicBlock* head_bb;	/* First basic block in the code */
+    BasicBlock* curr_bb;	/* Current basic block */
+    int maxDepth;		/* Maximum stack depth encountered */
+    int curCatchDepth;		/* Current depth of catches */
+    int maxCatchDepth;		/* Maximum depth of catches encountered */
+    int flags;			/* Compilation flags (TCL_EVAL_DIRECT) */
+} AssemblyEnv;
+
+/*
+ * Static functions defined in this file.
+ */
+
+static void		AddBasicBlockRangeToErrorInfo(AssemblyEnv*,
+			    BasicBlock*);
+static BasicBlock *	AllocBB(AssemblyEnv*);
+static int		AssembleOneLine(AssemblyEnv* envPtr);
+static void		BBAdjustStackDepth(BasicBlock* bbPtr, int consumed,
+			    int produced);
+static void		BBUpdateStackReqs(BasicBlock* bbPtr, int tblIdx,
+			    int count);
+static void		BBEmitInstInt1(AssemblyEnv* assemEnvPtr, int tblIdx,
+			    int opnd, int count);
+static void		BBEmitInstInt4(AssemblyEnv* assemEnvPtr, int tblIdx,
+			    int opnd, int count);
+static void		BBEmitInst1or4(AssemblyEnv* assemEnvPtr, int tblIdx,
+			    int param, int count);
+static void		BBEmitOpcode(AssemblyEnv* assemEnvPtr, int tblIdx,
+			    int count);
+static int		BuildExceptionRanges(AssemblyEnv* assemEnvPtr);
+static int		CalculateJumpRelocations(AssemblyEnv*, int*);
+static int		CheckForUnclosedCatches(AssemblyEnv*);
+static int		CheckForThrowInWrongContext(AssemblyEnv*);
+static int		CheckNonThrowingBlock(AssemblyEnv*, BasicBlock*);
+static int		BytecodeMightThrow(unsigned char);
+static int		CheckJumpTableLabels(AssemblyEnv*, BasicBlock*);
+static int		CheckNamespaceQualifiers(Tcl_Interp*, const char*,
+			    int);
+static int		CheckNonNegative(Tcl_Interp*, int);
+static int		CheckOneByte(Tcl_Interp*, int);
+static int		CheckSignedOneByte(Tcl_Interp*, int);
+static int		CheckStack(AssemblyEnv*);
+static int		CheckStrictlyPositive(Tcl_Interp*, int);
+static ByteCode *	CompileAssembleObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static void		CompileEmbeddedScript(AssemblyEnv*, Tcl_Token*,
+			    const TalInstDesc*);
+static int		DefineLabel(AssemblyEnv* envPtr, const char* label);
+static void		DeleteMirrorJumpTable(JumptableInfo* jtPtr);
+static void		DupAssembleCodeInternalRep(Tcl_Obj* src,
+			    Tcl_Obj* dest);
+static void		FillInJumpOffsets(AssemblyEnv*);
+static int		CreateMirrorJumpTable(AssemblyEnv* assemEnvPtr,
+			    Tcl_Obj* jumpTable);
+static int		FindLocalVar(AssemblyEnv* envPtr,
+			    Tcl_Token** tokenPtrPtr);
+static int		FinishAssembly(AssemblyEnv*);
+static void		FreeAssembleCodeInternalRep(Tcl_Obj *objPtr);
+static void		FreeAssemblyEnv(AssemblyEnv*);
+static int		GetBooleanOperand(AssemblyEnv*, Tcl_Token**, int*);
+static int		GetListIndexOperand(AssemblyEnv*, Tcl_Token**, int*);
+static int		GetIntegerOperand(AssemblyEnv*, Tcl_Token**, int*);
+static int		GetNextOperand(AssemblyEnv*, Tcl_Token**, Tcl_Obj**);
+static void		LookForFreshCatches(BasicBlock*, BasicBlock**);
+static void		MoveCodeForJumps(AssemblyEnv*, int);
+static void		MoveExceptionRangesToBasicBlock(AssemblyEnv*, int,
+			    int);
+static AssemblyEnv*	NewAssemblyEnv(CompileEnv*, int);
+static int		ProcessCatches(AssemblyEnv*);
+static int		ProcessCatchesInBasicBlock(AssemblyEnv*, BasicBlock*,
+			    BasicBlock*, enum BasicBlockCatchState, int);
+static void		ResetVisitedBasicBlocks(AssemblyEnv*);
+static void		ResolveJumpTableTargets(AssemblyEnv*, BasicBlock*);
+static void		ReportUndefinedLabel(AssemblyEnv*, BasicBlock*,
+			    Tcl_Obj*);
+static void		RestoreEmbeddedExceptionRanges(AssemblyEnv*);
+static int		StackCheckBasicBlock(AssemblyEnv*, BasicBlock *,
+			    BasicBlock *, int);
+static BasicBlock*	StartBasicBlock(AssemblyEnv*, int fallthrough,
+			    Tcl_Obj* jumpLabel);
+/* static int		AdvanceIp(const unsigned char *pc); */
+static int		StackCheckBasicBlock(AssemblyEnv*, BasicBlock *,
+			    BasicBlock *, int);
+static int		StackCheckExit(AssemblyEnv*);
+static void		StackFreshCatches(AssemblyEnv*, BasicBlock*, int,
+			    BasicBlock**, int*);
+static void		SyncStackDepth(AssemblyEnv*);
+static int		TclAssembleCode(CompileEnv* envPtr, const char* code,
+			    int codeLen, int flags);
+static void		UnstackExpiredCatches(CompileEnv*, BasicBlock*, int,
+			    BasicBlock**, int*);
+
+/*
+ * Tcl_ObjType that describes bytecode emitted by the assembler.
+ */
+
+static const Tcl_ObjType assembleCodeType = {
+    "assemblecode",
+    FreeAssembleCodeInternalRep, /* freeIntRepProc */
+    DupAssembleCodeInternalRep,	 /* dupIntRepProc */
+    NULL,			 /* updateStringProc */
+    NULL			 /* setFromAnyProc */
+};
+
+/*
+ * Source instructions recognized in the Tcl Assembly Language (TAL)
+ */
+
+static const TalInstDesc TalInstructionTable[] = {
+    /* PUSH must be first, see the code near the end of TclAssembleCode */
+    {"push",		ASSEM_PUSH,	(INST_PUSH1<<8
+					 | INST_PUSH4),		0,	1},
+
+    {"add",		ASSEM_1BYTE,	INST_ADD,		2,	1},
+    {"append",		ASSEM_LVT,	(INST_APPEND_SCALAR1<<8
+					 | INST_APPEND_SCALAR4),1,	1},
+    {"appendArray",	ASSEM_LVT,	(INST_APPEND_ARRAY1<<8
+					 | INST_APPEND_ARRAY4),	2,	1},
+    {"appendArrayStk",	ASSEM_1BYTE,	INST_APPEND_ARRAY_STK,	3,	1},
+    {"appendStk",	ASSEM_1BYTE,	INST_APPEND_STK,	2,	1},
+    {"arrayExistsImm",	ASSEM_LVT4,	INST_ARRAY_EXISTS_IMM,	0,	1},
+    {"arrayExistsStk",	ASSEM_1BYTE,	INST_ARRAY_EXISTS_STK,	1,	1},
+    {"arrayMakeImm",	ASSEM_LVT4,	INST_ARRAY_MAKE_IMM,	0,	0},
+    {"arrayMakeStk",	ASSEM_1BYTE,	INST_ARRAY_MAKE_STK,	1,	0},
+    {"beginCatch",	ASSEM_BEGIN_CATCH,
+					INST_BEGIN_CATCH4,	0,	0},
+    {"bitand",		ASSEM_1BYTE,	INST_BITAND,		2,	1},
+    {"bitnot",		ASSEM_1BYTE,	INST_BITNOT,		1,	1},
+    {"bitor",		ASSEM_1BYTE,	INST_BITOR,		2,	1},
+    {"bitxor",		ASSEM_1BYTE,	INST_BITXOR,		2,	1},
+    {"clockRead",	ASSEM_CLOCK_READ, INST_CLOCK_READ,	0,	1},
+    {"concat",		ASSEM_CONCAT1,	INST_STR_CONCAT1,	INT_MIN,1},
+    {"concatStk",	ASSEM_LIST,	INST_CONCAT_STK,	INT_MIN,1},
+    {"coroName",	ASSEM_1BYTE,	INST_COROUTINE_NAME,	0,	1},
+    {"currentNamespace",ASSEM_1BYTE,	INST_NS_CURRENT,	0,	1},
+    {"dictAppend",	ASSEM_LVT4,	INST_DICT_APPEND,	2,	1},
+    {"dictExists",	ASSEM_DICT_GET, INST_DICT_EXISTS,	INT_MIN,1},
+    {"dictExpand",	ASSEM_1BYTE,	INST_DICT_EXPAND,	3,	1},
+    {"dictGet",		ASSEM_DICT_GET, INST_DICT_GET,		INT_MIN,1},
+    {"dictIncrImm",	ASSEM_SINT4_LVT4,
+					INST_DICT_INCR_IMM,	1,	1},
+    {"dictLappend",	ASSEM_LVT4,	INST_DICT_LAPPEND,	2,	1},
+    {"dictRecombineStk",ASSEM_1BYTE,	INST_DICT_RECOMBINE_STK,3,	0},
+    {"dictRecombineImm",ASSEM_LVT4,	INST_DICT_RECOMBINE_IMM,2,	0},
+    {"dictSet",		ASSEM_DICT_SET, INST_DICT_SET,		INT_MIN,1},
+    {"dictUnset",	ASSEM_DICT_UNSET,
+					INST_DICT_UNSET,	INT_MIN,1},
+    {"div",		ASSEM_1BYTE,	INST_DIV,		2,	1},
+    {"dup",		ASSEM_1BYTE,	INST_DUP,		1,	2},
+    {"endCatch",	ASSEM_END_CATCH,INST_END_CATCH,		0,	0},
+    {"eq",		ASSEM_1BYTE,	INST_EQ,		2,	1},
+    {"eval",		ASSEM_EVAL,	INST_EVAL_STK,		1,	1},
+    {"evalStk",		ASSEM_1BYTE,	INST_EVAL_STK,		1,	1},
+    {"exist",		ASSEM_LVT4,	INST_EXIST_SCALAR,	0,	1},
+    {"existArray",	ASSEM_LVT4,	INST_EXIST_ARRAY,	1,	1},
+    {"existArrayStk",	ASSEM_1BYTE,	INST_EXIST_ARRAY_STK,	2,	1},
+    {"existStk",	ASSEM_1BYTE,	INST_EXIST_STK,		1,	1},
+    {"expon",		ASSEM_1BYTE,	INST_EXPON,		2,	1},
+    {"expr",		ASSEM_EVAL,	INST_EXPR_STK,		1,	1},
+    {"exprStk",		ASSEM_1BYTE,	INST_EXPR_STK,		1,	1},
+    {"ge",		ASSEM_1BYTE,	INST_GE,		2,	1},
+    {"gt",		ASSEM_1BYTE,	INST_GT,		2,	1},
+    {"incr",		ASSEM_LVT1,	INST_INCR_SCALAR1,	1,	1},
+    {"incrArray",	ASSEM_LVT1,	INST_INCR_ARRAY1,	2,	1},
+    {"incrArrayImm",	ASSEM_LVT1_SINT1,
+					INST_INCR_ARRAY1_IMM,	1,	1},
+    {"incrArrayStk",	ASSEM_1BYTE,	INST_INCR_ARRAY_STK,	3,	1},
+    {"incrArrayStkImm", ASSEM_SINT1,	INST_INCR_ARRAY_STK_IMM,2,	1},
+    {"incrImm",		ASSEM_LVT1_SINT1,
+					INST_INCR_SCALAR1_IMM,	0,	1},
+    {"incrStk",		ASSEM_1BYTE,	INST_INCR_STK,		2,	1},
+    {"incrStkImm",	ASSEM_SINT1,	INST_INCR_STK_IMM,	1,	1},
+    {"infoLevelArgs",	ASSEM_1BYTE,	INST_INFO_LEVEL_ARGS,	1,	1},
+    {"infoLevelNumber",	ASSEM_1BYTE,	INST_INFO_LEVEL_NUM,	0,	1},
+    {"invokeStk",	ASSEM_INVOKE,	(INST_INVOKE_STK1 << 8
+					 | INST_INVOKE_STK4),	INT_MIN,1},
+    {"jump",		ASSEM_JUMP,	INST_JUMP1,		0,	0},
+    {"jump4",		ASSEM_JUMP4,	INST_JUMP4,		0,	0},
+    {"jumpFalse",	ASSEM_JUMP,	INST_JUMP_FALSE1,	1,	0},
+    {"jumpFalse4",	ASSEM_JUMP4,	INST_JUMP_FALSE4,	1,	0},
+    {"jumpTable",	ASSEM_JUMPTABLE,INST_JUMP_TABLE,	1,	0},
+    {"jumpTrue",	ASSEM_JUMP,	INST_JUMP_TRUE1,	1,	0},
+    {"jumpTrue4",	ASSEM_JUMP4,	INST_JUMP_TRUE4,	1,	0},
+    {"label",		ASSEM_LABEL,	0,			0,	0},
+    {"land",		ASSEM_1BYTE,	INST_LAND,		2,	1},
+    {"lappend",		ASSEM_LVT,	(INST_LAPPEND_SCALAR1<<8
+					 | INST_LAPPEND_SCALAR4),
+								1,	1},
+    {"lappendArray",	ASSEM_LVT,	(INST_LAPPEND_ARRAY1<<8
+					 | INST_LAPPEND_ARRAY4),2,	1},
+    {"lappendArrayStk", ASSEM_1BYTE,	INST_LAPPEND_ARRAY_STK,	3,	1},
+    {"lappendList",	ASSEM_LVT4,	INST_LAPPEND_LIST,	1,	1},
+    {"lappendListArray",ASSEM_LVT4,	INST_LAPPEND_LIST_ARRAY,2,	1},
+    {"lappendListArrayStk", ASSEM_1BYTE,INST_LAPPEND_LIST_ARRAY_STK, 3,	1},
+    {"lappendListStk",	ASSEM_1BYTE,	INST_LAPPEND_LIST_STK,	2,	1},
+    {"lappendStk",	ASSEM_1BYTE,	INST_LAPPEND_STK,	2,	1},
+    {"le",		ASSEM_1BYTE,	INST_LE,		2,	1},
+    {"lindexMulti",	ASSEM_LINDEX_MULTI,
+					INST_LIST_INDEX_MULTI,	INT_MIN,1},
+    {"list",		ASSEM_LIST,	INST_LIST,		INT_MIN,1},
+    {"listConcat",	ASSEM_1BYTE,	INST_LIST_CONCAT,	2,	1},
+    {"listIn",		ASSEM_1BYTE,	INST_LIST_IN,		2,	1},
+    {"listIndex",	ASSEM_1BYTE,	INST_LIST_INDEX,	2,	1},
+    {"listIndexImm",	ASSEM_INDEX,	INST_LIST_INDEX_IMM,	1,	1},
+    {"listLength",	ASSEM_1BYTE,	INST_LIST_LENGTH,	1,	1},
+    {"listNotIn",	ASSEM_1BYTE,	INST_LIST_NOT_IN,	2,	1},
+    {"load",		ASSEM_LVT,	(INST_LOAD_SCALAR1 << 8
+					 | INST_LOAD_SCALAR4),	0,	1},
+    {"loadArray",	ASSEM_LVT,	(INST_LOAD_ARRAY1<<8
+					 | INST_LOAD_ARRAY4),	1,	1},
+    {"loadArrayStk",	ASSEM_1BYTE,	INST_LOAD_ARRAY_STK,	2,	1},
+    {"loadStk",		ASSEM_1BYTE,	INST_LOAD_STK,		1,	1},
+    {"lor",		ASSEM_1BYTE,	INST_LOR,		2,	1},
+    {"lsetFlat",	ASSEM_LSET_FLAT,INST_LSET_FLAT,		INT_MIN,1},
+    {"lsetList",	ASSEM_1BYTE,	INST_LSET_LIST,		3,	1},
+    {"lshift",		ASSEM_1BYTE,	INST_LSHIFT,		2,	1},
+    {"lt",		ASSEM_1BYTE,	INST_LT,		2,	1},
+    {"mod",		ASSEM_1BYTE,	INST_MOD,		2,	1},
+    {"mult",		ASSEM_1BYTE,	INST_MULT,		2,	1},
+    {"neq",		ASSEM_1BYTE,	INST_NEQ,		2,	1},
+    {"nop",		ASSEM_1BYTE,	INST_NOP,		0,	0},
+    {"not",		ASSEM_1BYTE,	INST_LNOT,		1,	1},
+    {"nsupvar",		ASSEM_LVT4,	INST_NSUPVAR,		2,	1},
+    {"numericType",	ASSEM_1BYTE,	INST_NUM_TYPE,		1,	1},
+    {"originCmd",	ASSEM_1BYTE,	INST_ORIGIN_COMMAND,	1,	1},
+    {"over",		ASSEM_OVER,	INST_OVER,		INT_MIN,-1-1},
+    {"pop",		ASSEM_1BYTE,	INST_POP,		1,	0},
+    {"pushReturnCode",	ASSEM_1BYTE,	INST_PUSH_RETURN_CODE,	0,	1},
+    {"pushReturnOpts",	ASSEM_1BYTE,	INST_PUSH_RETURN_OPTIONS,
+								0,	1},
+    {"pushResult",	ASSEM_1BYTE,	INST_PUSH_RESULT,	0,	1},
+    {"regexp",		ASSEM_REGEXP,	INST_REGEXP,		2,	1},
+    {"resolveCmd",	ASSEM_1BYTE,	INST_RESOLVE_COMMAND,	1,	1},
+    {"reverse",		ASSEM_REVERSE,	INST_REVERSE,		INT_MIN,-1-0},
+    {"rshift",		ASSEM_1BYTE,	INST_RSHIFT,		2,	1},
+    {"store",		ASSEM_LVT,	(INST_STORE_SCALAR1<<8
+					 | INST_STORE_SCALAR4),	1,	1},
+    {"storeArray",	ASSEM_LVT,	(INST_STORE_ARRAY1<<8
+					 | INST_STORE_ARRAY4),	2,	1},
+    {"storeArrayStk",	ASSEM_1BYTE,	INST_STORE_ARRAY_STK,	3,	1},
+    {"storeStk",	ASSEM_1BYTE,	INST_STORE_STK,		2,	1},
+    {"strcaseLower",	ASSEM_1BYTE,	INST_STR_LOWER,		1,	1},
+    {"strcaseTitle",	ASSEM_1BYTE,	INST_STR_TITLE,		1,	1},
+    {"strcaseUpper",	ASSEM_1BYTE,	INST_STR_UPPER,		1,	1},
+    {"strcmp",		ASSEM_1BYTE,	INST_STR_CMP,		2,	1},
+    {"strcat",		ASSEM_CONCAT1,	INST_STR_CONCAT1,	INT_MIN,1},
+    {"streq",		ASSEM_1BYTE,	INST_STR_EQ,		2,	1},
+    {"strfind",		ASSEM_1BYTE,	INST_STR_FIND,		2,	1},
+    {"strindex",	ASSEM_1BYTE,	INST_STR_INDEX,		2,	1},
+    {"strlen",		ASSEM_1BYTE,	INST_STR_LEN,		1,	1},
+    {"strmap",		ASSEM_1BYTE,	INST_STR_MAP,		3,	1},
+    {"strmatch",	ASSEM_BOOL,	INST_STR_MATCH,		2,	1},
+    {"strneq",		ASSEM_1BYTE,	INST_STR_NEQ,		2,	1},
+    {"strrange",	ASSEM_1BYTE,	INST_STR_RANGE,		3,	1},
+    {"strreplace",	ASSEM_1BYTE,	INST_STR_REPLACE,	4,	1},
+    {"strrfind",	ASSEM_1BYTE,	INST_STR_FIND_LAST,	2,	1},
+    {"strtrim",		ASSEM_1BYTE,	INST_STR_TRIM,		2,	1},
+    {"strtrimLeft",	ASSEM_1BYTE,	INST_STR_TRIM_LEFT,	2,	1},
+    {"strtrimRight",	ASSEM_1BYTE,	INST_STR_TRIM_RIGHT,	2,	1},
+    {"sub",		ASSEM_1BYTE,	INST_SUB,		2,	1},
+    {"tclooClass",	ASSEM_1BYTE,	INST_TCLOO_CLASS,	1,	1},
+    {"tclooIsObject",	ASSEM_1BYTE,	INST_TCLOO_IS_OBJECT,	1,	1},
+    {"tclooNamespace",	ASSEM_1BYTE,	INST_TCLOO_NS,		1,	1},
+    {"tclooSelf",	ASSEM_1BYTE,	INST_TCLOO_SELF,	0,	1},
+    {"tryCvtToBoolean",	ASSEM_1BYTE,	INST_TRY_CVT_TO_BOOLEAN,1,	2},
+    {"tryCvtToNumeric",	ASSEM_1BYTE,	INST_TRY_CVT_TO_NUMERIC,1,	1},
+    {"uminus",		ASSEM_1BYTE,	INST_UMINUS,		1,	1},
+    {"unset",		ASSEM_BOOL_LVT4,INST_UNSET_SCALAR,	0,	0},
+    {"unsetArray",	ASSEM_BOOL_LVT4,INST_UNSET_ARRAY,	1,	0},
+    {"unsetArrayStk",	ASSEM_BOOL,	INST_UNSET_ARRAY_STK,	2,	0},
+    {"unsetStk",	ASSEM_BOOL,	INST_UNSET_STK,		1,	0},
+    {"uplus",		ASSEM_1BYTE,	INST_UPLUS,		1,	1},
+    {"upvar",		ASSEM_LVT4,	INST_UPVAR,		2,	1},
+    {"variable",	ASSEM_LVT4,	INST_VARIABLE,		1,	0},
+    {"verifyDict",	ASSEM_1BYTE,	INST_DICT_VERIFY,	1,	0},
+    {"yield",		ASSEM_1BYTE,	INST_YIELD,		1,	1},
+    {NULL,		0,		0,			0,	0}
+};
+
+/*
+ * List of instructions that cannot throw an exception under any
+ * circumstances.  These instructions are the ones that are permissible after
+ * an exception is caught but before the corresponding exception range is
+ * popped from the stack.
+ * The instructions must be in ascending order by numeric operation code.
+ */
+
+static const unsigned char NonThrowingByteCodes[] = {
+    INST_PUSH1, INST_PUSH4, INST_POP, INST_DUP,			/* 1-4 */
+    INST_JUMP1, INST_JUMP4,					/* 34-35 */
+    INST_END_CATCH, INST_PUSH_RESULT, INST_PUSH_RETURN_CODE,	/* 70-72 */
+    INST_LIST,							/* 79 */
+    INST_OVER,							/* 95 */
+    INST_PUSH_RETURN_OPTIONS,					/* 108 */
+    INST_REVERSE,						/* 126 */
+    INST_NOP,							/* 132 */
+    INST_STR_MAP,						/* 143 */
+    INST_STR_FIND,						/* 144 */
+    INST_COROUTINE_NAME,					/* 149 */
+    INST_NS_CURRENT,						/* 151 */
+    INST_INFO_LEVEL_NUM,					/* 152 */
+    INST_RESOLVE_COMMAND,					/* 154 */
+    INST_STR_TRIM, INST_STR_TRIM_LEFT, INST_STR_TRIM_RIGHT,	/* 166-168 */
+    INST_CONCAT_STK,						/* 169 */
+    INST_STR_UPPER, INST_STR_LOWER, INST_STR_TITLE,		/* 170-172 */
+    INST_NUM_TYPE						/* 180 */
+};
+
+/*
+ * Helper macros.
+ */
+
+#if defined(TCL_DEBUG_ASSEMBLY) && defined(__GNUC__) && __GNUC__ > 2
+#define DEBUG_PRINT(...)	fprintf(stderr, ##__VA_ARGS__);fflush(stderr)
+#elif defined(__GNUC__) && __GNUC__ > 2
+#define DEBUG_PRINT(...)	/* nothing */
+#else
+#define DEBUG_PRINT		/* nothing */
+#endif
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BBAdjustStackDepth --
+ *
+ *	When an opcode is emitted, adjusts the stack information in the basic
+ *	block to reflect the number of operands produced and consumed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Updates minimum, maximum and final stack requirements in the basic
+ *	block.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+BBAdjustStackDepth(
+    BasicBlock *bbPtr,		/* Structure describing the basic block */
+    int consumed,		/* Count of operands consumed by the
+				 * operation */
+    int produced)		/* Count of operands produced by the
+				 * operation */
+{
+    int depth = bbPtr->finalStackDepth;
+
+    depth -= consumed;
+    if (depth < bbPtr->minStackDepth) {
+	bbPtr->minStackDepth = depth;
+    }
+    depth += produced;
+    if (depth > bbPtr->maxStackDepth) {
+	bbPtr->maxStackDepth = depth;
+    }
+    bbPtr->finalStackDepth = depth;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BBUpdateStackReqs --
+ *
+ *	Updates the stack requirements of a basic block, given the opcode
+ *	being emitted and an operand count.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Updates min, max and final stack requirements in the basic block.
+ *
+ * Notes:
+ *	This function must not be called for instructions such as REVERSE and
+ *	OVER that are variadic but do not consume all their operands. Instead,
+ *	BBAdjustStackDepth should be called directly.
+ *
+ *	count should be provided only for variadic operations. For operations
+ *	with known arity, count should be 0.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+BBUpdateStackReqs(
+    BasicBlock* bbPtr,		/* Structure describing the basic block */
+    int tblIdx,			/* Index in TalInstructionTable of the
+				 * operation being assembled */
+    int count)			/* Count of operands for variadic insts */
+{
+    int consumed = TalInstructionTable[tblIdx].operandsConsumed;
+    int produced = TalInstructionTable[tblIdx].operandsProduced;
+
+    if (consumed == INT_MIN) {
+	/*
+	 * The instruction is variadic; it consumes 'count' operands.
+	 */
+
+	consumed = count;
+    }
+    if (produced < 0) {
+	/*
+	 * The instruction leaves some of its variadic operands on the stack,
+	 * with net stack effect of '-1-produced'
+	 */
+
+	produced = consumed - produced - 1;
+    }
+    BBAdjustStackDepth(bbPtr, consumed, produced);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BBEmitOpcode, BBEmitInstInt1, BBEmitInstInt4 --
+ *
+ *	Emit the opcode part of an instruction, or the entirety of an
+ *	instruction with a 1- or 4-byte operand, and adjust stack
+ *	requirements.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores instruction and operand in the operand stream, and adjusts the
+ *	stack.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+BBEmitOpcode(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int tblIdx,			/* Table index in TalInstructionTable of op */
+    int count)			/* Operand count for variadic ops */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr = assemEnvPtr->curr_bb;
+				/* Current basic block */
+    int op = TalInstructionTable[tblIdx].tclInstCode & 0xff;
+
+    /*
+     * If this is the first instruction in a basic block, record its line
+     * number.
+     */
+
+    if (bbPtr->startOffset == envPtr->codeNext - envPtr->codeStart) {
+	bbPtr->startLine = assemEnvPtr->cmdLine;
+    }
+
+    TclEmitInt1(op, envPtr);
+    TclUpdateAtCmdStart(op, envPtr);
+    BBUpdateStackReqs(bbPtr, tblIdx, count);
+}
+
+static void
+BBEmitInstInt1(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int tblIdx,			/* Index in TalInstructionTable of op */
+    int opnd,			/* 1-byte operand */
+    int count)			/* Operand count for variadic ops */
+{
+    BBEmitOpcode(assemEnvPtr, tblIdx, count);
+    TclEmitInt1(opnd, assemEnvPtr->envPtr);
+}
+
+static void
+BBEmitInstInt4(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int tblIdx,			/* Index in TalInstructionTable of op */
+    int opnd,			/* 4-byte operand */
+    int count)			/* Operand count for variadic ops */
+{
+    BBEmitOpcode(assemEnvPtr, tblIdx, count);
+    TclEmitInt4(opnd, assemEnvPtr->envPtr);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BBEmitInst1or4 --
+ *
+ *	Emits a 1- or 4-byte operation according to the magnitude of the
+ *	operand.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+BBEmitInst1or4(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int tblIdx,			/* Index in TalInstructionTable of op */
+    int param,			/* Variable-length parameter */
+    int count)			/* Arity if variadic */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr = assemEnvPtr->curr_bb;
+				/* Current basic block */
+    int op = TalInstructionTable[tblIdx].tclInstCode;
+
+    if (param <= 0xff) {
+	op >>= 8;
+    } else {
+	op &= 0xff;
+    }
+    TclEmitInt1(op, envPtr);
+    if (param <= 0xff) {
+	TclEmitInt1(param, envPtr);
+    } else {
+	TclEmitInt4(param, envPtr);
+    }
+    TclUpdateAtCmdStart(op, envPtr);
+    BBUpdateStackReqs(bbPtr, tblIdx, count);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * Tcl_AssembleObjCmd, TclNRAssembleObjCmd --
+ *
+ *	Direct evaluation path for tcl::unsupported::assemble
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Assembles the code in objv[1], and executes it, so side effects
+ *	include whatever the code does.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+int
+Tcl_AssembleObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    /*
+     * Boilerplate - make sure that there is an NRE trampoline on the C stack
+     * because there needs to be one in place to execute bytecode.
+     */
+
+    return Tcl_NRCallObjProc(interp, TclNRAssembleObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRAssembleObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    ByteCode *codePtr;		/* Pointer to the bytecode to execute */
+    Tcl_Obj* backtrace;		/* Object where extra error information is
+				 * constructed. */
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "bytecodeList");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Assemble the source to bytecode.
+     */
+
+    codePtr = CompileAssembleObj(interp, objv[1]);
+
+    /*
+     * On failure, report error line.
+     */
+
+    if (codePtr == NULL) {
+	Tcl_AddErrorInfo(interp, "\n    (\"");
+	Tcl_AppendObjToErrorInfo(interp, objv[0]);
+	Tcl_AddErrorInfo(interp, "\" body, line ");
+	backtrace = Tcl_NewIntObj(Tcl_GetErrorLine(interp));
+	Tcl_AppendObjToErrorInfo(interp, backtrace);
+	Tcl_AddErrorInfo(interp, ")");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Use NRE to evaluate the bytecode from the trampoline.
+     */
+
+    return TclNRExecuteByteCode(interp, codePtr);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CompileAssembleObj --
+ *
+ *	Sets up and assembles Tcl bytecode for the direct-execution path in
+ *	the Tcl bytecode assembler.
+ *
+ * Results:
+ *	Returns a pointer to the assembled code. Returns NULL if the assembly
+ *	fails for any reason, with an appropriate error message in the
+ *	interpreter.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static ByteCode *
+CompileAssembleObj(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    Tcl_Obj *objPtr)		/* Source code to assemble */
+{
+    Interp *iPtr = (Interp *) interp;
+				/* Internals of the interpreter */
+    CompileEnv compEnv;		/* Compilation environment structure */
+    register ByteCode *codePtr = NULL;
+				/* Bytecode resulting from the assembly */
+    Namespace* namespacePtr;	/* Namespace in which variable and command
+				 * names in the bytecode resolve */
+    int status;			/* Status return from Tcl_AssembleCode */
+    const char* source;		/* String representation of the source code */
+    int sourceLen;		/* Length of the source code in bytes */
+
+
+    /*
+     * Get the expression ByteCode from the object. If it exists, make sure it
+     * is valid in the current context.
+     */
+
+    if (objPtr->typePtr == &assembleCodeType) {
+	namespacePtr = iPtr->varFramePtr->nsPtr;
+	codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	if (((Interp *) *codePtr->interpHandle == iPtr)
+		&& (codePtr->compileEpoch == iPtr->compileEpoch)
+		&& (codePtr->nsPtr == namespacePtr)
+		&& (codePtr->nsEpoch == namespacePtr->resolverEpoch)
+		&& (codePtr->localCachePtr
+			== iPtr->varFramePtr->localCachePtr)) {
+	    return codePtr;
+	}
+
+	/*
+	 * Not valid, so free it and regenerate.
+	 */
+
+	TclFreeIntRep(objPtr);
+    }
+
+    /*
+     * Set up the compilation environment, and assemble the code.
+     */
+
+    source = TclGetStringFromObj(objPtr, &sourceLen);
+    TclInitCompileEnv(interp, &compEnv, source, sourceLen, NULL, 0);
+    status = TclAssembleCode(&compEnv, source, sourceLen, TCL_EVAL_DIRECT);
+    if (status != TCL_OK) {
+	/*
+	 * Assembly failed. Clean up and report the error.
+	 */
+	TclFreeCompileEnv(&compEnv);
+	return NULL;
+    }
+
+    /*
+     * Add a "done" instruction as the last instruction and change the object
+     * into a ByteCode object. Ownership of the literal objects and aux data
+     * items is given to the ByteCode object.
+     */
+
+    TclEmitOpcode(INST_DONE, &compEnv);
+    codePtr = TclInitByteCodeObj(objPtr, &assembleCodeType, &compEnv);
+    TclFreeCompileEnv(&compEnv);
+
+    /*
+     * Record the local variable context to which the bytecode pertains
+     */
+
+    if (iPtr->varFramePtr->localCachePtr) {
+	codePtr->localCachePtr = iPtr->varFramePtr->localCachePtr;
+	codePtr->localCachePtr->refCount++;
+    }
+
+    /*
+     * Report on what the assembler did.
+     */
+
+#ifdef TCL_COMPILE_DEBUG
+    if (tclTraceCompile >= 2) {
+	TclPrintByteCodeObj(interp, objPtr);
+	fflush(stdout);
+    }
+#endif /* TCL_COMPILE_DEBUG */
+
+    return codePtr;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TclCompileAssembleCmd --
+ *
+ *	Compilation procedure for the '::tcl::unsupported::assemble' command.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Puts the result of assembling the code into the bytecode stream in
+ *	'compileEnv'.
+ *
+ * This procedure makes sure that the command has a single arg, which is
+ * constant. If that condition is met, the procedure calls TclAssembleCode to
+ * produce bytecode for the given assembly code, and returns any error
+ * resulting from the assembly.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+int
+TclCompileAssembleCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;	/* Token in the input script */
+
+    int numCommands = envPtr->numCommands;
+    int offset = envPtr->codeNext - envPtr->codeStart;
+    int depth = envPtr->currStackDepth;
+
+    /*
+     * Make sure that the command has a single arg that is a simple word.
+     */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Compile the code and convert any error from the compilation into
+     * bytecode reporting the error;
+     */
+
+    if (TCL_ERROR == TclAssembleCode(envPtr, tokenPtr[1].start,
+	    tokenPtr[1].size, TCL_EVAL_DIRECT)) {
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"%.*s\" body, line %d)",
+		parsePtr->tokenPtr->size, parsePtr->tokenPtr->start,
+		Tcl_GetErrorLine(interp)));
+	envPtr->numCommands = numCommands;
+	envPtr->codeNext = envPtr->codeStart + offset;
+	envPtr->currStackDepth = depth;
+	TclCompileSyntaxError(interp, envPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TclAssembleCode --
+ *
+ *	Take a list of instructions in a Tcl_Obj, and assemble them to Tcl
+ *	bytecodes
+ *
+ * Results:
+ *	Returns TCL_OK on success, TCL_ERROR on failure.  If 'flags' includes
+ *	TCL_EVAL_DIRECT, places an error message in the interpreter result.
+ *
+ * Side effects:
+ *	Adds byte codes to the compile environment, and updates the
+ *	environment's stack depth.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+TclAssembleCode(
+    CompileEnv *envPtr,		/* Compilation environment that is to receive
+				 * the generated bytecode */
+    const char* codePtr,	/* Assembly-language code to be processed */
+    int codeLen,		/* Length of the code */
+    int flags)			/* OR'ed combination of flags */
+{
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    /*
+     * Walk through the assembly script using the Tcl parser.  Each 'command'
+     * will be an instruction or assembly directive.
+     */
+
+    const char* instPtr = codePtr;
+				/* Where to start looking for a line of code */
+    const char* nextPtr;	/* Pointer to the end of the line of code */
+    int bytesLeft = codeLen;	/* Number of bytes of source code remaining to
+				 * be parsed */
+    int status;			/* Tcl status return */
+    AssemblyEnv* assemEnvPtr = NewAssemblyEnv(envPtr, flags);
+    Tcl_Parse* parsePtr = assemEnvPtr->parsePtr;
+
+    do {
+	/*
+	 * Parse out one command line from the assembly script.
+	 */
+
+	status = Tcl_ParseCommand(interp, instPtr, bytesLeft, 0, parsePtr);
+
+	/*
+	 * Report errors in the parse.
+	 */
+
+	if (status != TCL_OK) {
+	    if (flags & TCL_EVAL_DIRECT) {
+		Tcl_LogCommandInfo(interp, codePtr, parsePtr->commandStart,
+			parsePtr->term + 1 - parsePtr->commandStart);
+	    }
+	    FreeAssemblyEnv(assemEnvPtr);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Advance the pointers around any leading commentary.
+	 */
+
+	TclAdvanceLines(&assemEnvPtr->cmdLine, instPtr,
+		parsePtr->commandStart);
+	TclAdvanceContinuations(&assemEnvPtr->cmdLine, &assemEnvPtr->clNext,
+		parsePtr->commandStart - envPtr->source);
+
+	/*
+	 * Process the line of code.
+	 */
+
+	if (parsePtr->numWords > 0) {
+	    int instLen = parsePtr->commandSize;
+		    /* Length in bytes of the current command */
+
+	    if (parsePtr->term == parsePtr->commandStart + instLen - 1) {
+		--instLen;
+	    }
+
+	    /*
+	     * If tracing, show each line assembled as it happens.
+	     */
+
+#ifdef TCL_COMPILE_DEBUG
+	    if ((tclTraceCompile >= 2) && (envPtr->procPtr == NULL)) {
+		printf("  %4ld Assembling: ",
+			(long)(envPtr->codeNext - envPtr->codeStart));
+		TclPrintSource(stdout, parsePtr->commandStart,
+			TclMin(instLen, 55));
+		printf("\n");
+	    }
+#endif
+	    if (AssembleOneLine(assemEnvPtr) != TCL_OK) {
+		if (flags & TCL_EVAL_DIRECT) {
+		    Tcl_LogCommandInfo(interp, codePtr,
+			    parsePtr->commandStart, instLen);
+		}
+		Tcl_FreeParse(parsePtr);
+		FreeAssemblyEnv(assemEnvPtr);
+		return TCL_ERROR;
+	    }
+	}
+
+	/*
+	 * Advance to the next line of code.
+	 */
+
+	nextPtr = parsePtr->commandStart + parsePtr->commandSize;
+	bytesLeft -= (nextPtr - instPtr);
+	instPtr = nextPtr;
+	TclAdvanceLines(&assemEnvPtr->cmdLine, parsePtr->commandStart,
+		instPtr);
+	TclAdvanceContinuations(&assemEnvPtr->cmdLine, &assemEnvPtr->clNext,
+		instPtr - envPtr->source);
+	Tcl_FreeParse(parsePtr);
+    } while (bytesLeft > 0);
+
+    /*
+     * Done with parsing the code.
+     */
+
+    status = FinishAssembly(assemEnvPtr);
+    FreeAssemblyEnv(assemEnvPtr);
+    return status;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * NewAssemblyEnv --
+ *
+ *	Creates an environment for the assembler to run in.
+ *
+ * Results:
+ *	Allocates, initialises and returns an assembler environment
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static AssemblyEnv*
+NewAssemblyEnv(
+    CompileEnv* envPtr,		/* Compilation environment being used for code
+				 * generation*/
+    int flags)			/* Compilation flags (TCL_EVAL_DIRECT) */
+{
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    AssemblyEnv* assemEnvPtr = TclStackAlloc(interp, sizeof(AssemblyEnv));
+				/* Assembler environment under construction */
+    Tcl_Parse* parsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse));
+				/* Parse of one line of assembly code */
+
+    assemEnvPtr->envPtr = envPtr;
+    assemEnvPtr->parsePtr = parsePtr;
+    assemEnvPtr->cmdLine = 1;
+    assemEnvPtr->clNext = envPtr->clNext;
+
+    /*
+     * Make the hashtables that store symbol resolution.
+     */
+
+    Tcl_InitHashTable(&assemEnvPtr->labelHash, TCL_STRING_KEYS);
+
+    /*
+     * Start the first basic block.
+     */
+
+    assemEnvPtr->curr_bb = NULL;
+    assemEnvPtr->head_bb = AllocBB(assemEnvPtr);
+    assemEnvPtr->curr_bb = assemEnvPtr->head_bb;
+    assemEnvPtr->head_bb->startLine = 1;
+
+    /*
+     * Stash compilation flags.
+     */
+
+    assemEnvPtr->flags = flags;
+    return assemEnvPtr;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * FreeAssemblyEnv --
+ *
+ *	Cleans up the assembler environment when assembly is complete.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+FreeAssemblyEnv(
+    AssemblyEnv* assemEnvPtr)	/* Environment to free */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment being used for code
+				 * generation */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    BasicBlock* thisBB;		/* Pointer to a basic block being deleted */
+    BasicBlock* nextBB;		/* Pointer to a deleted basic block's
+				 * successor */
+
+    /*
+     * Free all the basic block structures.
+     */
+
+    for (thisBB = assemEnvPtr->head_bb; thisBB != NULL; thisBB = nextBB) {
+	if (thisBB->jumpTarget != NULL) {
+	    Tcl_DecrRefCount(thisBB->jumpTarget);
+	}
+	if (thisBB->foreignExceptions != NULL) {
+	    ckfree(thisBB->foreignExceptions);
+	}
+	nextBB = thisBB->successor1;
+	if (thisBB->jtPtr != NULL) {
+	    DeleteMirrorJumpTable(thisBB->jtPtr);
+	    thisBB->jtPtr = NULL;
+	}
+	ckfree(thisBB);
+    }
+
+    /*
+     * Dispose what's left.
+     */
+
+    Tcl_DeleteHashTable(&assemEnvPtr->labelHash);
+    TclStackFree(interp, assemEnvPtr->parsePtr);
+    TclStackFree(interp, assemEnvPtr);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * AssembleOneLine --
+ *
+ *	Assembles a single command from an assembly language source.
+ *
+ * Results:
+ *	Returns TCL_ERROR with an appropriate error message if the assembly
+ *	fails. Returns TCL_OK if the assembly succeeds. Updates the assembly
+ *	environment with the state of the assembly.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+AssembleOneLine(
+    AssemblyEnv* assemEnvPtr)	/* State of the assembly */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment being used for code
+				 * gen */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Parse* parsePtr = assemEnvPtr->parsePtr;
+				/* Parse of the line of code */
+    Tcl_Token* tokenPtr;	/* Current token within the line of code */
+    Tcl_Obj* instNameObj;	/* Name of the instruction */
+    int tblIdx;			/* Index in TalInstructionTable of the
+				 * instruction */
+    enum TalInstType instType;	/* Type of the instruction */
+    Tcl_Obj* operand1Obj = NULL;
+				/* First operand to the instruction */
+    const char* operand1;	/* String rep of the operand */
+    int operand1Len;		/* String length of the operand */
+    int opnd;			/* Integer representation of an operand */
+    int litIndex;		/* Literal pool index of a constant */
+    int localVar;		/* LVT index of a local variable */
+    int flags;			/* Flags for a basic block */
+    JumptableInfo* jtPtr;	/* Pointer to a jumptable */
+    int infoIndex;		/* Index of the jumptable in auxdata */
+    int status = TCL_ERROR;	/* Return value from this function */
+
+    /*
+     * Make sure that the instruction name is known at compile time.
+     */
+
+    tokenPtr = parsePtr->tokenPtr;
+    if (GetNextOperand(assemEnvPtr, &tokenPtr, &instNameObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Look up the instruction name.
+     */
+
+    if (Tcl_GetIndexFromObjStruct(interp, instNameObj,
+	    &TalInstructionTable[0].name, sizeof(TalInstDesc), "instruction",
+	    TCL_EXACT, &tblIdx) != TCL_OK) {
+	goto cleanup;
+    }
+
+    /*
+     * Vector on the type of instruction being processed.
+     */
+
+    instType = TalInstructionTable[tblIdx].instType;
+    switch (instType) {
+
+    case ASSEM_PUSH:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "value");
+	    goto cleanup;
+	}
+	if (GetNextOperand(assemEnvPtr, &tokenPtr, &operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+	operand1 = TclGetStringFromObj(operand1Obj, &operand1Len);
+	litIndex = TclRegisterLiteral(envPtr, operand1, operand1Len, 0);
+	BBEmitInst1or4(assemEnvPtr, tblIdx, litIndex, 0);
+	break;
+
+    case ASSEM_1BYTE:
+	if (parsePtr->numWords != 1) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "");
+	    goto cleanup;
+	}
+	BBEmitOpcode(assemEnvPtr, tblIdx, 0);
+	break;
+
+    case ASSEM_BEGIN_CATCH:
+	/*
+	 * Emit the BEGIN_CATCH instruction with the code offset of the
+	 * exception branch target instead of the exception range index. The
+	 * correct index will be generated and inserted later, when catches
+	 * are being resolved.
+	 */
+
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "label");
+	    goto cleanup;
+	}
+	if (GetNextOperand(assemEnvPtr, &tokenPtr, &operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+	assemEnvPtr->curr_bb->jumpLine = assemEnvPtr->cmdLine;
+	assemEnvPtr->curr_bb->jumpOffset = envPtr->codeNext-envPtr->codeStart;
+	BBEmitInstInt4(assemEnvPtr, tblIdx, 0, 0);
+	assemEnvPtr->curr_bb->flags |= BB_BEGINCATCH;
+	StartBasicBlock(assemEnvPtr, BB_FALLTHRU, operand1Obj);
+	break;
+
+    case ASSEM_BOOL:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "boolean");
+	    goto cleanup;
+	}
+	if (GetBooleanOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, opnd, 0);
+	break;
+
+    case ASSEM_BOOL_LVT4:
+	if (parsePtr->numWords != 3) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "boolean varName");
+	    goto cleanup;
+	}
+	if (GetBooleanOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, opnd, 0);
+	TclEmitInt4(localVar, envPtr);
+	break;
+
+    case ASSEM_CLOCK_READ:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "imm8");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	if (opnd < 0 || opnd > 3) {
+	    Tcl_SetObjResult(interp,
+			     Tcl_NewStringObj("operand must be [0..3]", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "OPERAND<0,>3", NULL);
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_CONCAT1:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "imm8");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckOneByte(interp, opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_DICT_GET:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd+1);
+	break;
+
+    case ASSEM_DICT_SET:
+	if (parsePtr->numWords != 3) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count varName");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd+1);
+	TclEmitInt4(localVar, envPtr);
+	break;
+
+    case ASSEM_DICT_UNSET:
+	if (parsePtr->numWords != 3) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count varName");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	TclEmitInt4(localVar, envPtr);
+	break;
+
+    case ASSEM_END_CATCH:
+	if (parsePtr->numWords != 1) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "");
+	    goto cleanup;
+	}
+	assemEnvPtr->curr_bb->flags |= BB_ENDCATCH;
+	BBEmitOpcode(assemEnvPtr, tblIdx, 0);
+	StartBasicBlock(assemEnvPtr, BB_FALLTHRU, NULL);
+	break;
+
+    case ASSEM_EVAL:
+	/* TODO - Refactor this stuff into a subroutine that takes the inst
+	 * code, the message ("script" or "expression") and an evaluator
+	 * callback that calls TclCompileScript or TclCompileExpr. */
+
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj,
+		    ((TalInstructionTable[tblIdx].tclInstCode
+		    == INST_EVAL_STK) ? "script" : "expression"));
+	    goto cleanup;
+	}
+	if (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	    CompileEmbeddedScript(assemEnvPtr, tokenPtr+1,
+		    TalInstructionTable+tblIdx);
+	} else if (GetNextOperand(assemEnvPtr, &tokenPtr,
+		&operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	} else {
+	    operand1 = TclGetStringFromObj(operand1Obj, &operand1Len);
+	    litIndex = TclRegisterLiteral(envPtr, operand1, operand1Len, 0);
+
+	    /*
+	     * Assumes that PUSH is the first slot!
+	     */
+
+	    BBEmitInst1or4(assemEnvPtr, 0, litIndex, 0);
+	    BBEmitOpcode(assemEnvPtr, tblIdx, 0);
+	}
+	break;
+
+    case ASSEM_INVOKE:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+
+	BBEmitInst1or4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_JUMP:
+    case ASSEM_JUMP4:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "label");
+	    goto cleanup;
+	}
+	if (GetNextOperand(assemEnvPtr, &tokenPtr, &operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+	assemEnvPtr->curr_bb->jumpOffset = envPtr->codeNext-envPtr->codeStart;
+	if (instType == ASSEM_JUMP) {
+	    flags = BB_JUMP1;
+	    BBEmitInstInt1(assemEnvPtr, tblIdx, 0, 0);
+	} else {
+	    flags = 0;
+	    BBEmitInstInt4(assemEnvPtr, tblIdx, 0, 0);
+	}
+
+	/*
+	 * Start a new basic block at the instruction following the jump.
+	 */
+
+	assemEnvPtr->curr_bb->jumpLine = assemEnvPtr->cmdLine;
+	if (TalInstructionTable[tblIdx].operandsConsumed != 0) {
+	    flags |= BB_FALLTHRU;
+	}
+	StartBasicBlock(assemEnvPtr, flags, operand1Obj);
+	break;
+
+    case ASSEM_JUMPTABLE:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "table");
+	    goto cleanup;
+	}
+	if (GetNextOperand(assemEnvPtr, &tokenPtr, &operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+
+	jtPtr = ckalloc(sizeof(JumptableInfo));
+
+	Tcl_InitHashTable(&jtPtr->hashTable, TCL_STRING_KEYS);
+	assemEnvPtr->curr_bb->jumpLine = assemEnvPtr->cmdLine;
+	assemEnvPtr->curr_bb->jumpOffset = envPtr->codeNext-envPtr->codeStart;
+	DEBUG_PRINT("bb %p jumpLine %d jumpOffset %d\n",
+		assemEnvPtr->curr_bb, assemEnvPtr->cmdLine,
+		envPtr->codeNext - envPtr->codeStart);
+
+	infoIndex = TclCreateAuxData(jtPtr, &tclJumptableInfoType, envPtr);
+	DEBUG_PRINT("auxdata index=%d\n", infoIndex);
+
+	BBEmitInstInt4(assemEnvPtr, tblIdx, infoIndex, 0);
+	if (CreateMirrorJumpTable(assemEnvPtr, operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+	StartBasicBlock(assemEnvPtr, BB_JUMPTABLE|BB_FALLTHRU, NULL);
+	break;
+
+    case ASSEM_LABEL:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "name");
+	    goto cleanup;
+	}
+	if (GetNextOperand(assemEnvPtr, &tokenPtr, &operand1Obj) != TCL_OK) {
+	    goto cleanup;
+	}
+
+	/*
+	 * Add the (label_name, address) pair to the hash table.
+	 */
+
+	if (DefineLabel(assemEnvPtr, TclGetString(operand1Obj)) != TCL_OK) {
+	    goto cleanup;
+	}
+	break;
+
+    case ASSEM_LINDEX_MULTI:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckStrictlyPositive(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_LIST:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckNonNegative(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_INDEX:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetListIndexOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_LSET_FLAT:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	if (opnd < 2) {
+	    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+		Tcl_SetObjResult(interp,
+			Tcl_NewStringObj("operand must be >=2", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ASSEM", "OPERAND>=2", NULL);
+	    }
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_LVT:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "varname");
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInst1or4(assemEnvPtr, tblIdx, localVar, 0);
+	break;
+
+    case ASSEM_LVT1:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "varname");
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0 || CheckOneByte(interp, localVar)) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, localVar, 0);
+	break;
+
+    case ASSEM_LVT1_SINT1:
+	if (parsePtr->numWords != 3) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "varName imm8");
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0 || CheckOneByte(interp, localVar)
+		|| GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckSignedOneByte(interp, opnd)) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, localVar, 0);
+	TclEmitInt1(opnd, envPtr);
+	break;
+
+    case ASSEM_LVT4:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "varname");
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, localVar, 0);
+	break;
+
+    case ASSEM_OVER:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckNonNegative(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd+1);
+	break;
+
+    case ASSEM_REGEXP:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "boolean");
+	    goto cleanup;
+	}
+	if (GetBooleanOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	{
+	    int flags = TCL_REG_ADVANCED | (opnd ? TCL_REG_NOCASE : 0);
+
+	    BBEmitInstInt1(assemEnvPtr, tblIdx, flags, 0);
+	}
+	break;
+
+    case ASSEM_REVERSE:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckNonNegative(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, opnd);
+	break;
+
+    case ASSEM_SINT1:
+	if (parsePtr->numWords != 2) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "imm8");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK
+		|| CheckSignedOneByte(interp, opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	BBEmitInstInt1(assemEnvPtr, tblIdx, opnd, 0);
+	break;
+
+    case ASSEM_SINT4_LVT4:
+	if (parsePtr->numWords != 3) {
+	    Tcl_WrongNumArgs(interp, 1, &instNameObj, "count varName");
+	    goto cleanup;
+	}
+	if (GetIntegerOperand(assemEnvPtr, &tokenPtr, &opnd) != TCL_OK) {
+	    goto cleanup;
+	}
+	localVar = FindLocalVar(assemEnvPtr, &tokenPtr);
+	if (localVar < 0) {
+	    goto cleanup;
+	}
+	BBEmitInstInt4(assemEnvPtr, tblIdx, opnd, 0);
+	TclEmitInt4(localVar, envPtr);
+	break;
+
+    default:
+	Tcl_Panic("Instruction \"%s\" could not be found, can't happen\n",
+		TclGetString(instNameObj));
+    }
+
+    status = TCL_OK;
+ cleanup:
+    Tcl_DecrRefCount(instNameObj);
+    if (operand1Obj) {
+	Tcl_DecrRefCount(operand1Obj);
+    }
+    return status;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CompileEmbeddedScript --
+ *
+ *	Compile an embedded 'eval' or 'expr' that appears in assembly code.
+ *
+ * This procedure is called when the 'eval' or 'expr' assembly directive is
+ * encountered, and the argument to the directive is a simple word that
+ * requires no substitution. The appropriate compiler (TclCompileScript or
+ * TclCompileExpr) is invoked recursively, and emits bytecode.
+ *
+ * Before the compiler is invoked, the compilation environment's stack
+ * consumption is reset to zero. Upon return from the compilation, the net
+ * stack effect of the compilation is in the compiler env, and this stack
+ * effect is posted to the assembler environment. The compile environment's
+ * stack consumption is then restored to what it was before (which is actually
+ * the state of the stack on entry to the block of assembly code).
+ *
+ * Any exception ranges pushed by the compilation are copied to the basic
+ * block and removed from the compiler environment. They will be rebuilt at
+ * the end of assembly, when the exception stack depth is actually known.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+CompileEmbeddedScript(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token* tokenPtr,	/* Tcl_Token containing the script */
+    const TalInstDesc* instPtr)	/* Instruction that determines whether
+				 * the script is 'expr' or 'eval' */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+
+    /*
+     * The expression or script is not only known at compile time, but
+     * actually a "simple word". It can be compiled inline by invoking the
+     * compiler recursively.
+     *
+     * Save away the stack depth and reset it before compiling the script.
+     * We'll record the stack usage of the script in the BasicBlock, and
+     * accumulate it together with the stack usage of the enclosing assembly
+     * code.
+     */
+
+    int savedStackDepth = envPtr->currStackDepth;
+    int savedMaxStackDepth = envPtr->maxStackDepth;
+    int savedCodeIndex = envPtr->codeNext - envPtr->codeStart;
+    int savedExceptArrayNext = envPtr->exceptArrayNext;
+
+    envPtr->currStackDepth = 0;
+    envPtr->maxStackDepth = 0;
+
+    StartBasicBlock(assemEnvPtr, BB_FALLTHRU, NULL);
+    switch(instPtr->tclInstCode) {
+    case INST_EVAL_STK:
+	TclCompileScript(interp, tokenPtr->start, tokenPtr->size, envPtr);
+	break;
+    case INST_EXPR_STK:
+	TclCompileExpr(interp, tokenPtr->start, tokenPtr->size, envPtr, 1);
+	break;
+    default:
+	Tcl_Panic("no ASSEM_EVAL case for %s (%d), can't happen",
+		instPtr->name, instPtr->tclInstCode);
+    }
+
+    /*
+     * Roll up the stack usage of the embedded block into the assembler
+     * environment.
+     */
+
+    SyncStackDepth(assemEnvPtr);
+    envPtr->currStackDepth = savedStackDepth;
+    envPtr->maxStackDepth = savedMaxStackDepth;
+
+    /*
+     * Save any exception ranges that were pushed by the compiler; they will
+     * need to be fixed up once the stack depth is known.
+     */
+
+    MoveExceptionRangesToBasicBlock(assemEnvPtr, savedCodeIndex,
+	    savedExceptArrayNext);
+
+    /*
+     * Flush the current basic block.
+     */
+
+    StartBasicBlock(assemEnvPtr, BB_FALLTHRU, NULL);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * SyncStackDepth --
+ *
+ *	Copies the stack depth from the compile environment to a basic block.
+ *
+ * Side effects:
+ *	Current and max stack depth in the current basic block are adjusted.
+ *
+ * This procedure is called on return from invoking the compiler for the
+ * 'eval' and 'expr' operations. It adjusts the stack depth of the current
+ * basic block to reflect the stack required by the just-compiled code.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+SyncStackDepth(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* curr_bb = assemEnvPtr->curr_bb;
+				/* Current basic block */
+    int maxStackDepth = curr_bb->finalStackDepth + envPtr->maxStackDepth;
+				/* Max stack depth in the basic block */
+
+    if (maxStackDepth > curr_bb->maxStackDepth) {
+	curr_bb->maxStackDepth = maxStackDepth;
+    }
+    curr_bb->finalStackDepth += envPtr->currStackDepth;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * MoveExceptionRangesToBasicBlock --
+ *
+ *	Removes exception ranges that were created by compiling an embedded
+ *	script from the CompileEnv, and stores them in the BasicBlock. They
+ *	will be reinstalled, at the correct stack depth, after control flow
+ *	analysis is complete on the assembly code.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+MoveExceptionRangesToBasicBlock(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int savedCodeIndex,		/* Start of the embedded code */
+    int savedExceptArrayNext)	/* Saved index of the end of the exception
+				 * range array */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* curr_bb = assemEnvPtr->curr_bb;
+				/* Current basic block */
+    int exceptionCount = envPtr->exceptArrayNext - savedExceptArrayNext;
+				/* Number of ranges that must be moved */
+    int i;
+
+    if (exceptionCount == 0) {
+	/* Nothing to do */
+	return;
+    }
+
+    /*
+     * Save the exception ranges in the basic block. They will be re-added at
+     * the conclusion of assembly; at this time, the INST_BEGIN_CATCH
+     * instructions in the block will be adjusted from whatever range indices
+     * they have [savedExceptArrayNext .. envPtr->exceptArrayNext) to the
+     * indices that the exceptions acquire. The saved exception ranges are
+     * converted to a relative nesting depth. The depth will be recomputed
+     * once flow analysis has determined the actual stack depth of the block.
+     */
+
+    DEBUG_PRINT("basic block %p has %d exceptions starting at %d\n",
+	    curr_bb, exceptionCount, savedExceptArrayNext);
+    curr_bb->foreignExceptionBase = savedExceptArrayNext;
+    curr_bb->foreignExceptionCount = exceptionCount;
+    curr_bb->foreignExceptions =
+	    ckalloc(exceptionCount * sizeof(ExceptionRange));
+    memcpy(curr_bb->foreignExceptions,
+	    envPtr->exceptArrayPtr + savedExceptArrayNext,
+	    exceptionCount * sizeof(ExceptionRange));
+    for (i = 0; i < exceptionCount; ++i) {
+	curr_bb->foreignExceptions[i].nestingLevel -= envPtr->exceptDepth;
+    }
+    envPtr->exceptArrayNext = savedExceptArrayNext;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CreateMirrorJumpTable --
+ *
+ *	Makes a jump table with comparison values and assembly code labels.
+ *
+ * Results:
+ *	Returns a standard Tcl status, with an error message in the
+ *	interpreter on error.
+ *
+ * Side effects:
+ *	Initializes the jump table pointer in the current basic block to a
+ *	JumptableInfo. The keys in the JumptableInfo are the comparison
+ *	strings. The values, instead of being jump displacements, are
+ *	Tcl_Obj's with the code labels.
+ */
+
+static int
+CreateMirrorJumpTable(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Obj* jumps)		/* List of alternating keywords and labels */
+{
+    int objc;			/* Number of elements in the 'jumps' list */
+    Tcl_Obj** objv;		/* Pointers to the elements in the list */
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    BasicBlock* bbPtr = assemEnvPtr->curr_bb;
+				/* Current basic block */
+    JumptableInfo* jtPtr;
+    Tcl_HashTable* jtHashPtr;	/* Hashtable in the JumptableInfo */
+    Tcl_HashEntry* hashEntry;	/* Entry for a key in the hashtable */
+    int isNew;			/* Flag==1 if the key is not yet in the
+				 * table. */
+    int i;
+
+    if (Tcl_ListObjGetElements(interp, jumps, &objc, &objv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (objc % 2 != 0) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "jump table must have an even number of list elements",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADJUMPTABLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Allocate the jumptable.
+     */
+
+    jtPtr = ckalloc(sizeof(JumptableInfo));
+    jtHashPtr = &jtPtr->hashTable;
+    Tcl_InitHashTable(jtHashPtr, TCL_STRING_KEYS);
+
+    /*
+     * Fill the keys and labels into the table.
+     */
+
+    DEBUG_PRINT("jump table {\n");
+    for (i = 0; i < objc; i+=2) {
+	DEBUG_PRINT("  %s -> %s\n", TclGetString(objv[i]),
+		TclGetString(objv[i+1]));
+	hashEntry = Tcl_CreateHashEntry(jtHashPtr, TclGetString(objv[i]),
+		&isNew);
+	if (!isNew) {
+	    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"duplicate entry in jump table for \"%s\"",
+			TclGetString(objv[i])));
+		Tcl_SetErrorCode(interp, "TCL", "ASSEM", "DUPJUMPTABLEENTRY");
+		DeleteMirrorJumpTable(jtPtr);
+		return TCL_ERROR;
+	    }
+	}
+	Tcl_SetHashValue(hashEntry, objv[i+1]);
+	Tcl_IncrRefCount(objv[i+1]);
+    }
+    DEBUG_PRINT("}\n");
+
+    /*
+     * Put the mirror jumptable in the basic block struct.
+     */
+
+    bbPtr->jtPtr = jtPtr;
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * DeleteMirrorJumpTable --
+ *
+ *	Cleans up a jump table when the basic block is deleted.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+DeleteMirrorJumpTable(
+    JumptableInfo* jtPtr)
+{
+    Tcl_HashTable* jtHashPtr = &jtPtr->hashTable;
+				/* Hash table pointer */
+    Tcl_HashSearch search;	/* Hash search control */
+    Tcl_HashEntry* entry;	/* Hash table entry containing a jump label */
+    Tcl_Obj* label;		/* Jump label from the hash table */
+
+    for (entry = Tcl_FirstHashEntry(jtHashPtr, &search);
+	    entry != NULL;
+	    entry = Tcl_NextHashEntry(&search)) {
+	label = Tcl_GetHashValue(entry);
+	Tcl_DecrRefCount(label);
+	Tcl_SetHashValue(entry, NULL);
+    }
+    Tcl_DeleteHashTable(jtHashPtr);
+    ckfree(jtPtr);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * GetNextOperand --
+ *
+ *	Retrieves the next operand in sequence from an assembly instruction,
+ *	and makes sure that its value is known at compile time.
+ *
+ * Results:
+ *	If successful, returns TCL_OK and leaves a Tcl_Obj with the operand
+ *	text in *operandObjPtr. In case of failure, returns TCL_ERROR and
+ *	leaves *operandObjPtr untouched.
+ *
+ * Side effects:
+ *	Advances *tokenPtrPtr around the token just processed.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+GetNextOperand(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token** tokenPtrPtr,	/* INPUT/OUTPUT: Pointer to the token holding
+				 * the operand */
+    Tcl_Obj** operandObjPtr)	/* OUTPUT: Tcl object holding the operand text
+				 * with \-substitutions done. */
+{
+    Tcl_Interp* interp = (Tcl_Interp*) assemEnvPtr->envPtr->iPtr;
+    Tcl_Obj* operandObj = Tcl_NewObj();
+
+    if (!TclWordKnownAtCompileTime(*tokenPtrPtr, operandObj)) {
+	Tcl_DecrRefCount(operandObj);
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "assembly code may not contain substitutions", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "NOSUBST", NULL);
+	}
+	return TCL_ERROR;
+    }
+    *tokenPtrPtr = TokenAfter(*tokenPtrPtr);
+    Tcl_IncrRefCount(operandObj);
+    *operandObjPtr = operandObj;
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * GetBooleanOperand --
+ *
+ *	Retrieves a Boolean operand from the input stream and advances
+ *	the token pointer.
+ *
+ * Results:
+ *	Returns a standard Tcl result (with an error message in the
+ *	interpreter on failure).
+ *
+ * Side effects:
+ *	Stores the Boolean value in (*result) and advances (*tokenPtrPtr)
+ *	to the next token.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+GetBooleanOperand(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token** tokenPtrPtr,	/* Current token from the parser */
+    int* result)		/* OUTPUT: Integer extracted from the token */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Token* tokenPtr = *tokenPtrPtr;
+				/* INOUT: Pointer to the next token in the
+				 * source code */
+    Tcl_Obj* intObj;		/* Integer from the source code */
+    int status;			/* Tcl status return */
+
+    /*
+     * Extract the next token as a string.
+     */
+
+    if (GetNextOperand(assemEnvPtr, tokenPtrPtr, &intObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert to an integer, advance to the next token and return.
+     */
+
+    status = Tcl_GetBooleanFromObj(interp, intObj, result);
+    Tcl_DecrRefCount(intObj);
+    *tokenPtrPtr = TokenAfter(tokenPtr);
+    return status;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * GetIntegerOperand --
+ *
+ *	Retrieves an integer operand from the input stream and advances the
+ *	token pointer.
+ *
+ * Results:
+ *	Returns a standard Tcl result (with an error message in the
+ *	interpreter on failure).
+ *
+ * Side effects:
+ *	Stores the integer value in (*result) and advances (*tokenPtrPtr) to
+ *	the next token.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+GetIntegerOperand(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token** tokenPtrPtr,	/* Current token from the parser */
+    int* result)		/* OUTPUT: Integer extracted from the token */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Token* tokenPtr = *tokenPtrPtr;
+				/* INOUT: Pointer to the next token in the
+				 * source code */
+    Tcl_Obj* intObj;		/* Integer from the source code */
+    int status;			/* Tcl status return */
+
+    /*
+     * Extract the next token as a string.
+     */
+
+    if (GetNextOperand(assemEnvPtr, tokenPtrPtr, &intObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert to an integer, advance to the next token and return.
+     */
+
+    status = Tcl_GetIntFromObj(interp, intObj, result);
+    Tcl_DecrRefCount(intObj);
+    *tokenPtrPtr = TokenAfter(tokenPtr);
+    return status;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * GetListIndexOperand --
+ *
+ *	Gets the value of an operand intended to serve as a list index.
+ *
+ * Results:
+ *	Returns a standard Tcl result: TCL_OK if the parse is successful and
+ *	TCL_ERROR (with an appropriate error message) if the parse fails.
+ *
+ * Side effects:
+ *	Stores the list index at '*index'. Values between -1 and 0x7fffffff
+ *	have their natural meaning; values between -2 and -0x80000000
+ *	represent 'end-2-N'.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+GetListIndexOperand(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token** tokenPtrPtr,	/* Current token from the parser */
+    int* result)		/* OUTPUT: Integer extracted from the token */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Token* tokenPtr = *tokenPtrPtr;
+				/* INOUT: Pointer to the next token in the
+				 * source code */
+    Tcl_Obj* intObj;		/* Integer from the source code */
+    int status;			/* Tcl status return */
+
+    /*
+     * Extract the next token as a string.
+     */
+
+    if (GetNextOperand(assemEnvPtr, tokenPtrPtr, &intObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert to an integer, advance to the next token and return.
+     */
+
+    status = TclGetIntForIndex(interp, intObj, -2, result);
+    Tcl_DecrRefCount(intObj);
+    *tokenPtrPtr = TokenAfter(tokenPtr);
+    return status;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * FindLocalVar --
+ *
+ *	Gets the name of a local variable from the input stream and advances
+ *	the token pointer.
+ *
+ * Results:
+ *	Returns the LVT index of the local variable.  Returns -1 if the
+ *	variable is non-local, not known at compile time, or cannot be
+ *	installed in the LVT (leaving an error message in the interpreter
+ *	result if necessary).
+ *
+ * Side effects:
+ *	Advances the token pointer.  May define a new LVT slot if the variable
+ *	has not yet been seen and the execution context allows for it.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+FindLocalVar(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    Tcl_Token** tokenPtrPtr)
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Token* tokenPtr = *tokenPtrPtr;
+				/* INOUT: Pointer to the next token in the
+				 * source code. */
+    Tcl_Obj* varNameObj;	/* Name of the variable */
+    const char* varNameStr;
+    int varNameLen;
+    int localVar;		/* Index of the variable in the LVT */
+
+    if (GetNextOperand(assemEnvPtr, tokenPtrPtr, &varNameObj) != TCL_OK) {
+	return -1;
+    }
+    varNameStr = TclGetStringFromObj(varNameObj, &varNameLen);
+    if (CheckNamespaceQualifiers(interp, varNameStr, varNameLen)) {
+	Tcl_DecrRefCount(varNameObj);
+	return -1;
+    }
+    localVar = TclFindCompiledLocal(varNameStr, varNameLen, 1, envPtr);
+    Tcl_DecrRefCount(varNameObj);
+    if (localVar == -1) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "cannot use this instruction to create a variable"
+		    " in a non-proc context", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "LVT", NULL);
+	}
+	return -1;
+    }
+    *tokenPtrPtr = TokenAfter(tokenPtr);
+    return localVar;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckNamespaceQualifiers --
+ *
+ *	Verify that a variable name has no namespace qualifiers before
+ *	attempting to install it in the LVT.
+ *
+ * Results:
+ *	On success, returns TCL_OK. On failure, returns TCL_ERROR and stores
+ *	an error message in the interpreter result.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckNamespaceQualifiers(
+    Tcl_Interp* interp,		/* Tcl interpreter for error reporting */
+    const char* name,		/* Variable name to check */
+    int nameLen)		/* Length of the variable */
+{
+    const char* p;
+
+    for (p = name; p+2 < name+nameLen;  p++) {
+	if ((*p == ':') && (p[1] == ':')) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "variable \"%s\" is not local", name));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "NONLOCAL", name, NULL);
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckOneByte --
+ *
+ *	Verify that a constant fits in a single byte in the instruction
+ *	stream.
+ *
+ * Results:
+ *	On success, returns TCL_OK. On failure, returns TCL_ERROR and stores
+ *	an error message in the interpreter result.
+ *
+ * This code is here primarily to verify that instructions like INCR_SCALAR1
+ * are possible on a given local variable. The fact that there is no
+ * INCR_SCALAR4 is puzzling.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckOneByte(
+    Tcl_Interp* interp,		/* Tcl interpreter for error reporting */
+    int value)			/* Value to check */
+{
+    Tcl_Obj* result;		/* Error message */
+
+    if (value < 0 || value > 0xff) {
+	result = Tcl_NewStringObj("operand does not fit in one byte", -1);
+	Tcl_SetObjResult(interp, result);
+	Tcl_SetErrorCode(interp, "TCL", "ASSEM", "1BYTE", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckSignedOneByte --
+ *
+ *	Verify that a constant fits in a single signed byte in the instruction
+ *	stream.
+ *
+ * Results:
+ *	On success, returns TCL_OK. On failure, returns TCL_ERROR and stores
+ *	an error message in the interpreter result.
+ *
+ * This code is here primarily to verify that instructions like INCR_SCALAR1
+ * are possible on a given local variable. The fact that there is no
+ * INCR_SCALAR4 is puzzling.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckSignedOneByte(
+    Tcl_Interp* interp,		/* Tcl interpreter for error reporting */
+    int value)			/* Value to check */
+{
+    Tcl_Obj* result;		/* Error message */
+
+    if (value > 0x7f || value < -0x80) {
+	result = Tcl_NewStringObj("operand does not fit in one byte", -1);
+	Tcl_SetObjResult(interp, result);
+	Tcl_SetErrorCode(interp, "TCL", "ASSEM", "1BYTE", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckNonNegative --
+ *
+ *	Verify that a constant is nonnegative
+ *
+ * Results:
+ *	On success, returns TCL_OK. On failure, returns TCL_ERROR and stores
+ *	an error message in the interpreter result.
+ *
+ * This code is here primarily to verify that instructions like INCR_INVOKE
+ * are consuming a positive number of operands
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckNonNegative(
+    Tcl_Interp* interp,		/* Tcl interpreter for error reporting */
+    int value)			/* Value to check */
+{
+    Tcl_Obj* result;		/* Error message */
+
+    if (value < 0) {
+	result = Tcl_NewStringObj("operand must be nonnegative", -1);
+	Tcl_SetObjResult(interp, result);
+	Tcl_SetErrorCode(interp, "TCL", "ASSEM", "NONNEGATIVE", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckStrictlyPositive --
+ *
+ *	Verify that a constant is positive
+ *
+ * Results:
+ *	On success, returns TCL_OK. On failure, returns TCL_ERROR and
+ *	stores an error message in the interpreter result.
+ *
+ * This code is here primarily to verify that instructions like INCR_INVOKE
+ * are consuming a positive number of operands
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckStrictlyPositive(
+    Tcl_Interp* interp,		/* Tcl interpreter for error reporting */
+    int value)			/* Value to check */
+{
+    Tcl_Obj* result;		/* Error message */
+
+    if (value <= 0) {
+	result = Tcl_NewStringObj("operand must be positive", -1);
+	Tcl_SetObjResult(interp, result);
+	Tcl_SetErrorCode(interp, "TCL", "ASSEM", "POSITIVE", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * DefineLabel --
+ *
+ *	Defines a label appearing in the assembly sequence.
+ *
+ * Results:
+ *	Returns a standard Tcl result. Returns TCL_OK and an empty result if
+ *	the definition succeeds; returns TCL_ERROR and an appropriate message
+ *	if a duplicate definition is found.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+DefineLabel(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    const char* labelName)	/* Label being defined */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_HashEntry* entry;	/* Label's entry in the symbol table */
+    int isNew;			/* Flag == 1 iff the label was previously
+				 * undefined */
+
+    /* TODO - This can now be simplified! */
+
+    StartBasicBlock(assemEnvPtr, BB_FALLTHRU, NULL);
+
+    /*
+     * Look up the newly-defined label in the symbol table.
+     */
+
+    entry = Tcl_CreateHashEntry(&assemEnvPtr->labelHash, labelName, &isNew);
+    if (!isNew) {
+	/*
+	 * This is a duplicate label.
+	 */
+
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "duplicate definition of label \"%s\"", labelName));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "DUPLABEL", labelName,
+		    NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * This is the first appearance of the label in the code.
+     */
+
+    Tcl_SetHashValue(entry, assemEnvPtr->curr_bb);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * StartBasicBlock --
+ *
+ *	Starts a new basic block when a label or jump is encountered.
+ *
+ * Results:
+ *	Returns a pointer to the BasicBlock structure of the new
+ *	basic block.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static BasicBlock*
+StartBasicBlock(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int flags,			/* Flags to apply to the basic block being
+				 * closed, if there is one. */
+    Tcl_Obj* jumpLabel)		/* Label of the location that the block jumps
+				 * to, or NULL if the block does not jump */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* newBB;		/* BasicBlock structure for the new block */
+    BasicBlock* currBB = assemEnvPtr->curr_bb;
+
+    /*
+     * Coalesce zero-length blocks.
+     */
+
+    if (currBB->startOffset == envPtr->codeNext - envPtr->codeStart) {
+	currBB->startLine = assemEnvPtr->cmdLine;
+	return currBB;
+    }
+
+    /*
+     * Make the new basic block.
+     */
+
+    newBB = AllocBB(assemEnvPtr);
+
+    /*
+     * Record the jump target if there is one.
+     */
+
+    currBB->jumpTarget = jumpLabel;
+    if (jumpLabel != NULL) {
+	Tcl_IncrRefCount(currBB->jumpTarget);
+    }
+
+    /*
+     * Record the fallthrough if there is one.
+     */
+
+    currBB->flags |= flags;
+
+    /*
+     * Record the successor block.
+     */
+
+    currBB->successor1 = newBB;
+    assemEnvPtr->curr_bb = newBB;
+    return newBB;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * AllocBB --
+ *
+ *	Allocates a new basic block
+ *
+ * Results:
+ *	Returns a pointer to the newly allocated block, which is initialized
+ *	to contain no code and begin at the current instruction pointer.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static BasicBlock *
+AllocBB(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+    BasicBlock *bb = ckalloc(sizeof(BasicBlock));
+
+    bb->originalStartOffset =
+	    bb->startOffset = envPtr->codeNext - envPtr->codeStart;
+    bb->startLine = assemEnvPtr->cmdLine + 1;
+    bb->jumpOffset = -1;
+    bb->jumpLine = -1;
+    bb->prevPtr = assemEnvPtr->curr_bb;
+    bb->predecessor = NULL;
+    bb->successor1 = NULL;
+    bb->jumpTarget = NULL;
+    bb->initialStackDepth = 0;
+    bb->minStackDepth = 0;
+    bb->maxStackDepth = 0;
+    bb->finalStackDepth = 0;
+    bb->catchDepth = 0;
+    bb->enclosingCatch = NULL;
+    bb->foreignExceptionBase = -1;
+    bb->foreignExceptionCount = 0;
+    bb->foreignExceptions = NULL;
+    bb->jtPtr = NULL;
+    bb->flags = 0;
+
+    return bb;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * FinishAssembly --
+ *
+ *	Postprocessing after all bytecode has been generated for a block of
+ *	assembly code.
+ *
+ * Results:
+ *	Returns a standard Tcl result, with an error message left in the
+ *	interpreter if appropriate.
+ *
+ * Side effects:
+ *	The program is checked to see if any undefined labels remain.  The
+ *	initial stack depth of all the basic blocks in the flow graph is
+ *	calculated and saved.  The stack balance on exit is computed, checked
+ *	and saved.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+FinishAssembly(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    int mustMove;		/* Amount by which the code needs to be grown
+				 * because of expanding jumps */
+
+    /*
+     * Resolve the targets of all jumps and determine whether code needs to be
+     * moved around.
+     */
+
+    if (CalculateJumpRelocations(assemEnvPtr, &mustMove)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Move the code if necessary.
+     */
+
+    if (mustMove) {
+	MoveCodeForJumps(assemEnvPtr, mustMove);
+    }
+
+    /*
+     * Resolve jump target labels to bytecode offsets.
+     */
+
+    FillInJumpOffsets(assemEnvPtr);
+
+    /*
+     * Label each basic block with its catch context. Quit on inconsistency.
+     */
+
+    if (ProcessCatches(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure that no block accessible from a catch's error exit that hasn't
+     * popped the exception stack can throw an exception.
+     */
+
+    if (CheckForThrowInWrongContext(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Compute stack balance throughout the program.
+     */
+
+    if (CheckStack(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * TODO - Check for unreachable code. Or maybe not; unreachable code is
+     * Mostly Harmless.
+     */
+
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CalculateJumpRelocations --
+ *
+ *	Calculate any movement that has to be done in the assembly code to
+ *	expand JUMP1 instructions to JUMP4 (because they jump more than a
+ *	1-byte range).
+ *
+ * Results:
+ *	Returns a standard Tcl result, with an appropriate error message if
+ *	anything fails.
+ *
+ * Side effects:
+ *	Sets the 'startOffset' pointer in every basic block to the new origin
+ *	of the block, and turns off JUMP1 flags on instructions that must be
+ *	expanded (and adjusts them to the corresponding JUMP4's).  Does *not*
+ *	store the jump offsets at this point.
+ *
+ *	Sets *mustMove to 1 if and only if at least one instruction changed
+ *	size so the code must be moved.
+ *
+ *	As a side effect, also checks for undefined labels and reports them.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CalculateJumpRelocations(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    int* mustMove)		/* OUTPUT: Number of bytes that have been
+				 * added to the code */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr;		/* Pointer to a basic block being checked */
+    Tcl_HashEntry* entry;	/* Exit label's entry in the symbol table */
+    BasicBlock* jumpTarget;	/* Basic block where the jump goes */
+    int motion;			/* Amount by which the code has expanded */
+    int offset;			/* Offset in the bytecode from a jump
+				 * instruction to its target */
+    unsigned opcode;		/* Opcode in the bytecode being adjusted */
+
+    /*
+     * Iterate through basic blocks as long as a change results in code
+     * expansion.
+     */
+
+    *mustMove = 0;
+    do {
+	motion = 0;
+	for (bbPtr = assemEnvPtr->head_bb;
+		bbPtr != NULL;
+		bbPtr = bbPtr->successor1) {
+	    /*
+	     * Advance the basic block start offset by however many bytes we
+	     * have inserted in the code up to this point
+	     */
+
+	    bbPtr->startOffset += motion;
+
+	    /*
+	     * If the basic block references a label (and hence performs a
+	     * jump), find the location of the label. Report an error if the
+	     * label is missing.
+	     */
+
+	    if (bbPtr->jumpTarget != NULL) {
+		entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+			TclGetString(bbPtr->jumpTarget));
+		if (entry == NULL) {
+		    ReportUndefinedLabel(assemEnvPtr, bbPtr,
+			    bbPtr->jumpTarget);
+		    return TCL_ERROR;
+		}
+
+		/*
+		 * If the instruction is a JUMP1, turn it into a JUMP4 if its
+		 * target is out of range.
+		 */
+
+		jumpTarget = Tcl_GetHashValue(entry);
+		if (bbPtr->flags & BB_JUMP1) {
+		    offset = jumpTarget->startOffset
+			    - (bbPtr->jumpOffset + motion);
+		    if (offset < -0x80 || offset > 0x7f) {
+			opcode = TclGetUInt1AtPtr(envPtr->codeStart
+				+ bbPtr->jumpOffset);
+			++opcode;
+			TclStoreInt1AtPtr(opcode,
+				envPtr->codeStart + bbPtr->jumpOffset);
+			motion += 3;
+			bbPtr->flags &= ~BB_JUMP1;
+		    }
+		}
+	    }
+
+	    /*
+	     * If the basic block references a jump table, that doesn't affect
+	     * the code locations, but resolve the labels now, and store basic
+	     * block pointers in the jumptable hash.
+	     */
+
+	    if (bbPtr->flags & BB_JUMPTABLE) {
+		if (CheckJumpTableLabels(assemEnvPtr, bbPtr) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+	    }
+	}
+	*mustMove += motion;
+    } while (motion != 0);
+
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckJumpTableLabels --
+ *
+ *	Make sure that all the labels in a jump table are defined.
+ *
+ * Results:
+ *	Returns TCL_OK if they are, TCL_ERROR if they aren't.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckJumpTableLabels(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr)		/* Basic block that ends in a jump table */
+{
+    Tcl_HashTable* symHash = &bbPtr->jtPtr->hashTable;
+				/* Hash table with the symbols */
+    Tcl_HashSearch search;	/* Hash table iterator */
+    Tcl_HashEntry* symEntryPtr;	/* Hash entry for the symbols */
+    Tcl_Obj* symbolObj;		/* Jump target */
+    Tcl_HashEntry* valEntryPtr;	/* Hash entry for the resolutions */
+
+    /*
+     * Look up every jump target in the jump hash.
+     */
+
+    DEBUG_PRINT("check jump table labels %p {\n", bbPtr);
+    for (symEntryPtr = Tcl_FirstHashEntry(symHash, &search);
+	    symEntryPtr != NULL;
+	    symEntryPtr = Tcl_NextHashEntry(&search)) {
+	symbolObj = Tcl_GetHashValue(symEntryPtr);
+	valEntryPtr = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		TclGetString(symbolObj));
+	DEBUG_PRINT("  %s -> %s (%d)\n",
+		(char*) Tcl_GetHashKey(symHash, symEntryPtr),
+		TclGetString(symbolObj), (valEntryPtr != NULL));
+	if (valEntryPtr == NULL) {
+	    ReportUndefinedLabel(assemEnvPtr, bbPtr, symbolObj);
+	    return TCL_ERROR;
+	}
+    }
+    DEBUG_PRINT("}\n");
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ReportUndefinedLabel --
+ *
+ *	Report that a basic block refers to an undefined jump label
+ *
+ * Side effects:
+ *	Stores an error message, error code, and line number information in
+ *	the assembler's Tcl interpreter.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+ReportUndefinedLabel(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr,		/* Basic block that contains the undefined
+				 * label */
+    Tcl_Obj* jumpTarget)	/* Label of a jump target */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+
+    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"undefined label \"%s\"", TclGetString(jumpTarget)));
+	Tcl_SetErrorCode(interp, "TCL", "ASSEM", "NOLABEL",
+		TclGetString(jumpTarget), NULL);
+	Tcl_SetErrorLine(interp, bbPtr->jumpLine);
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * MoveCodeForJumps --
+ *
+ *	Move bytecodes in memory to accommodate JUMP1 instructions that have
+ *	expanded to become JUMP4's.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+MoveCodeForJumps(
+    AssemblyEnv* assemEnvPtr,	/* Assembler environment */
+    int mustMove)		/* Number of bytes of added code */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr;		/* Pointer to a basic block being checked */
+    int topOffset;		/* Bytecode offset of the following basic
+				 * block before code motion */
+
+    /*
+     * Make sure that there is enough space in the bytecode array to
+     * accommodate the expanded code.
+     */
+
+    while (envPtr->codeEnd < envPtr->codeNext + mustMove) {
+	TclExpandCodeArray(envPtr);
+    }
+
+    /*
+     * Iterate through the bytecodes in reverse order, and move them upward to
+     * their new homes.
+     */
+
+    topOffset = envPtr->codeNext - envPtr->codeStart;
+    for (bbPtr = assemEnvPtr->curr_bb; bbPtr != NULL; bbPtr = bbPtr->prevPtr) {
+	DEBUG_PRINT("move code from %d to %d\n",
+		bbPtr->originalStartOffset, bbPtr->startOffset);
+	memmove(envPtr->codeStart + bbPtr->startOffset,
+		envPtr->codeStart + bbPtr->originalStartOffset,
+		topOffset - bbPtr->originalStartOffset);
+	topOffset = bbPtr->originalStartOffset;
+	bbPtr->jumpOffset += (bbPtr->startOffset - bbPtr->originalStartOffset);
+    }
+    envPtr->codeNext += mustMove;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * FillInJumpOffsets --
+ *
+ *	Fill in the final offsets of all jump instructions once bytecode
+ *	locations have been completely determined.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+FillInJumpOffsets(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr;		/* Pointer to a basic block being checked */
+    Tcl_HashEntry* entry;	/* Hashtable entry for a jump target label */
+    BasicBlock* jumpTarget;	/* Basic block where a jump goes */
+    int fromOffset;		/* Bytecode location of a jump instruction */
+    int targetOffset;		/* Bytecode location of a jump instruction's
+				 * target */
+
+    for (bbPtr = assemEnvPtr->head_bb;
+	    bbPtr != NULL;
+	    bbPtr = bbPtr->successor1) {
+	if (bbPtr->jumpTarget != NULL) {
+	    entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		    TclGetString(bbPtr->jumpTarget));
+	    jumpTarget = Tcl_GetHashValue(entry);
+	    fromOffset = bbPtr->jumpOffset;
+	    targetOffset = jumpTarget->startOffset;
+	    if (bbPtr->flags & BB_JUMP1) {
+		TclStoreInt1AtPtr(targetOffset - fromOffset,
+			envPtr->codeStart + fromOffset + 1);
+	    } else {
+		TclStoreInt4AtPtr(targetOffset - fromOffset,
+			envPtr->codeStart + fromOffset + 1);
+	    }
+	}
+	if (bbPtr->flags & BB_JUMPTABLE) {
+	    ResolveJumpTableTargets(assemEnvPtr, bbPtr);
+	}
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ResolveJumpTableTargets --
+ *
+ *	Puts bytecode addresses for the targets of a jumptable into the
+ *	table
+ *
+ * Results:
+ *	Returns TCL_OK if they are, TCL_ERROR if they aren't.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+ResolveJumpTableTargets(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr)		/* Basic block that ends in a jump table */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_HashTable* symHash = &bbPtr->jtPtr->hashTable;
+				/* Hash table with the symbols */
+    Tcl_HashSearch search;	/* Hash table iterator */
+    Tcl_HashEntry* symEntryPtr;	/* Hash entry for the symbols */
+    Tcl_Obj* symbolObj;		/* Jump target */
+    Tcl_HashEntry* valEntryPtr;	/* Hash entry for the resolutions */
+    int auxDataIndex;		/* Index of the auxdata */
+    JumptableInfo* realJumpTablePtr;
+				/* Jump table in the actual code */
+    Tcl_HashTable* realJumpHashPtr;
+				/* Jump table hash in the actual code */
+    Tcl_HashEntry* realJumpEntryPtr;
+				/* Entry in the jump table hash in
+				 * the actual code */
+    BasicBlock* jumpTargetBBPtr;
+				/* Basic block that the jump proceeds to */
+    int junk;
+
+    auxDataIndex = TclGetInt4AtPtr(envPtr->codeStart + bbPtr->jumpOffset + 1);
+    DEBUG_PRINT("bbPtr = %p jumpOffset = %d auxDataIndex = %d\n",
+	    bbPtr, bbPtr->jumpOffset, auxDataIndex);
+    realJumpTablePtr = TclFetchAuxData(envPtr, auxDataIndex);
+    realJumpHashPtr = &realJumpTablePtr->hashTable;
+
+    /*
+     * Look up every jump target in the jump hash.
+     */
+
+    DEBUG_PRINT("resolve jump table {\n");
+    for (symEntryPtr = Tcl_FirstHashEntry(symHash, &search);
+	    symEntryPtr != NULL;
+	    symEntryPtr = Tcl_NextHashEntry(&search)) {
+	symbolObj = Tcl_GetHashValue(symEntryPtr);
+	DEBUG_PRINT("     symbol %s\n", TclGetString(symbolObj));
+
+	valEntryPtr = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		TclGetString(symbolObj));
+	jumpTargetBBPtr = Tcl_GetHashValue(valEntryPtr);
+
+	realJumpEntryPtr = Tcl_CreateHashEntry(realJumpHashPtr,
+		Tcl_GetHashKey(symHash, symEntryPtr), &junk);
+	DEBUG_PRINT("  %s -> %s -> bb %p (pc %d)    hash entry %p\n",
+		(char*) Tcl_GetHashKey(symHash, symEntryPtr),
+		TclGetString(symbolObj), jumpTargetBBPtr,
+		jumpTargetBBPtr->startOffset, realJumpEntryPtr);
+
+	Tcl_SetHashValue(realJumpEntryPtr,
+		INT2PTR(jumpTargetBBPtr->startOffset - bbPtr->jumpOffset));
+    }
+    DEBUG_PRINT("}\n");
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckForThrowInWrongContext --
+ *
+ *	Verify that no beginCatch/endCatch sequence can throw an exception
+ *	after an original exception is caught and before its exception context
+ *	is removed from the stack.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Stores an appropriate error message in the interpreter as needed.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckForThrowInWrongContext(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    BasicBlock* blockPtr;	/* Current basic block */
+
+    /*
+     * Walk through the basic blocks in turn, checking all the ones that have
+     * caught an exception and not disposed of it properly.
+     */
+
+    for (blockPtr = assemEnvPtr->head_bb;
+	    blockPtr != NULL;
+	    blockPtr = blockPtr->successor1) {
+	if (blockPtr->catchState == BBCS_CAUGHT) {
+	    /*
+	     * Walk through the instructions in the basic block.
+	     */
+
+	    if (CheckNonThrowingBlock(assemEnvPtr, blockPtr) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckNonThrowingBlock --
+ *
+ *	Check that a basic block cannot throw an exception.
+ *
+ * Results:
+ *	Returns TCL_ERROR if the block cannot be proven to be nonthrowing.
+ *
+ * Side effects:
+ *	Stashes an error message in the interpreter result.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckNonThrowingBlock(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* blockPtr)	/* Basic block where exceptions are not
+				 * allowed */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    BasicBlock* nextPtr;	/* Pointer to the succeeding basic block */
+    int offset;			/* Bytecode offset of the current
+				 * instruction */
+    int bound;			/* Bytecode offset following the last
+				 * instruction of the block. */
+    unsigned char opcode;	/* Current bytecode instruction */
+
+    /*
+     * Determine where in the code array the basic block ends.
+     */
+
+    nextPtr = blockPtr->successor1;
+    if (nextPtr == NULL) {
+	bound = envPtr->codeNext - envPtr->codeStart;
+    } else {
+	bound = nextPtr->startOffset;
+    }
+
+    /*
+     * Walk through the instructions of the block.
+     */
+
+    offset = blockPtr->startOffset;
+    while (offset < bound) {
+	/*
+	 * Determine whether an instruction is nonthrowing.
+	 */
+
+	opcode = (envPtr->codeStart)[offset];
+	if (BytecodeMightThrow(opcode)) {
+	    /*
+	     * Report an error for a throw in the wrong context.
+	     */
+
+	    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"\"%s\" instruction may not appear in "
+			"a context where an exception has been "
+			"caught and not disposed of.",
+			tclInstructionTable[opcode].name));
+		Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADTHROW", NULL);
+		AddBasicBlockRangeToErrorInfo(assemEnvPtr, blockPtr);
+	    }
+	    return TCL_ERROR;
+	}
+	offset += tclInstructionTable[opcode].numBytes;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BytecodeMightThrow --
+ *
+ *	Tests if a given bytecode instruction might throw an exception.
+ *
+ * Results:
+ *	Returns 1 if the bytecode might throw an exception, 0 if the
+ *	instruction is known never to throw.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+BytecodeMightThrow(
+    unsigned char opcode)
+{
+    /*
+     * Binary search on the non-throwing bytecode list.
+     */
+
+    int min = 0;
+    int max = sizeof(NonThrowingByteCodes) - 1;
+    int mid;
+    unsigned char c;
+
+    while (max >= min) {
+	mid = (min + max) / 2;
+	c = NonThrowingByteCodes[mid];
+	if (opcode < c) {
+	    max = mid-1;
+	} else if (opcode > c) {
+	    min = mid+1;
+	} else {
+	    /*
+	     * Opcode is nonthrowing.
+	     */
+
+	    return 0;
+	}
+    }
+
+    return 1;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckStack --
+ *
+ *	Audit stack usage in a block of assembly code.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Updates stack depth on entry for all basic blocks in the flowgraph.
+ *	Calculates the max stack depth used in the program, and updates the
+ *	compilation environment to reflect it.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckStack(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    int maxDepth;		/* Maximum stack depth overall */
+
+    /*
+     * Checking the head block will check all the other blocks recursively.
+     */
+
+    assemEnvPtr->maxDepth = 0;
+    if (StackCheckBasicBlock(assemEnvPtr, assemEnvPtr->head_bb, NULL,
+	    0) == TCL_ERROR) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Post the max stack depth back to the compilation environment.
+     */
+
+    maxDepth = assemEnvPtr->maxDepth + envPtr->currStackDepth;
+    if (maxDepth > envPtr->maxStackDepth) {
+	envPtr->maxStackDepth = maxDepth;
+    }
+
+    /*
+     * If the exit is reachable, make sure that the program exits with 1
+     * operand on the stack.
+     */
+
+    if (StackCheckExit(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Reset the visited state on all basic blocks.
+     */
+
+    ResetVisitedBasicBlocks(assemEnvPtr);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * StackCheckBasicBlock --
+ *
+ *	Checks stack consumption for a basic block (and recursively for its
+ *	successors).
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Updates initial stack depth for the basic block and its successors.
+ *	(Final and maximum stack depth are relative to initial, and are not
+ *	touched).
+ *
+ * This procedure eventually checks, for the entire flow graph, whether stack
+ * balance is consistent.  It is an error for a given basic block to be
+ * reachable along multiple flow paths with different stack depths.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+StackCheckBasicBlock(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* blockPtr,	/* Pointer to the basic block being checked */
+    BasicBlock* predecessor,	/* Pointer to the block that passed control to
+				 * this one. */
+    int initialStackDepth)	/* Stack depth on entry to the block */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    BasicBlock* jumpTarget;	/* Basic block where a jump goes */
+    int stackDepth;		/* Current stack depth */
+    int maxDepth;		/* Maximum stack depth so far */
+    int result;			/* Tcl status return */
+    Tcl_HashSearch jtSearch;	/* Search structure for the jump table */
+    Tcl_HashEntry* jtEntry;	/* Hash entry in the jump table */
+    Tcl_Obj* targetLabel;	/* Target label from the jump table */
+    Tcl_HashEntry* entry;	/* Hash entry in the label table */
+
+    if (blockPtr->flags & BB_VISITED) {
+	/*
+	 * If the block is already visited, check stack depth for consistency
+	 * among the paths that reach it.
+	 */
+
+	if (blockPtr->initialStackDepth == initialStackDepth) {
+	    return TCL_OK;
+	}
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "inconsistent stack depths on two execution paths", -1));
+
+	    /*
+	     * TODO - add execution trace of both paths
+	     */
+
+	    Tcl_SetErrorLine(interp, blockPtr->startLine);
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADSTACK", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the block is not already visited, set the 'predecessor' link to
+     * indicate how control got to it. Set the initial stack depth to the
+     * current stack depth in the flow of control.
+     */
+
+    blockPtr->flags |= BB_VISITED;
+    blockPtr->predecessor = predecessor;
+    blockPtr->initialStackDepth = initialStackDepth;
+
+    /*
+     * Calculate minimum stack depth, and flag an error if the block
+     * underflows the stack.
+     */
+
+    if (initialStackDepth + blockPtr->minStackDepth < 0) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("stack underflow", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADSTACK", NULL);
+	    AddBasicBlockRangeToErrorInfo(assemEnvPtr, blockPtr);
+	    Tcl_SetErrorLine(interp, blockPtr->startLine);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure that the block doesn't try to pop below the stack level of an
+     * enclosing catch.
+     */
+
+    if (blockPtr->enclosingCatch != 0 &&
+	    initialStackDepth + blockPtr->minStackDepth
+	    < (blockPtr->enclosingCatch->initialStackDepth
+		+ blockPtr->enclosingCatch->finalStackDepth)) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "code pops stack below level of enclosing catch", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADSTACKINCATCH", -1);
+	    AddBasicBlockRangeToErrorInfo(assemEnvPtr, blockPtr);
+	    Tcl_SetErrorLine(interp, blockPtr->startLine);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Update maximum stgack depth.
+     */
+
+    maxDepth = initialStackDepth + blockPtr->maxStackDepth;
+    if (maxDepth > assemEnvPtr->maxDepth) {
+	assemEnvPtr->maxDepth = maxDepth;
+    }
+
+    /*
+     * Calculate stack depth on exit from the block, and invoke this procedure
+     * recursively to check successor blocks.
+     */
+
+    stackDepth = initialStackDepth + blockPtr->finalStackDepth;
+    result = TCL_OK;
+    if (blockPtr->flags & BB_FALLTHRU) {
+	result = StackCheckBasicBlock(assemEnvPtr, blockPtr->successor1,
+		blockPtr, stackDepth);
+    }
+
+    if (result == TCL_OK && blockPtr->jumpTarget != NULL) {
+	entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		TclGetString(blockPtr->jumpTarget));
+	jumpTarget = Tcl_GetHashValue(entry);
+	result = StackCheckBasicBlock(assemEnvPtr, jumpTarget, blockPtr,
+		stackDepth);
+    }
+
+    /*
+     * All blocks referenced in a jump table are successors.
+     */
+
+    if (blockPtr->flags & BB_JUMPTABLE) {
+	for (jtEntry = Tcl_FirstHashEntry(&blockPtr->jtPtr->hashTable,
+		    &jtSearch);
+		result == TCL_OK && jtEntry != NULL;
+		jtEntry = Tcl_NextHashEntry(&jtSearch)) {
+	    targetLabel = Tcl_GetHashValue(jtEntry);
+	    entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		    TclGetString(targetLabel));
+	    jumpTarget = Tcl_GetHashValue(entry);
+	    result = StackCheckBasicBlock(assemEnvPtr, jumpTarget,
+		    blockPtr, stackDepth);
+	}
+    }
+
+    return result;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * StackCheckExit --
+ *
+ *	Makes sure that the net stack effect of an entire assembly language
+ *	script is to push 1 result.
+ *
+ * Results:
+ *	Returns a standard Tcl result, with an error message in the
+ *	interpreter result if the stack is wrong.
+ *
+ * Side effects:
+ *	If the assembly code had a net stack effect of zero, emits code to the
+ *	concluding block to push a null result. In any case, updates the stack
+ *	depth in the compile environment to reflect the net effect of the
+ *	assembly code.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+StackCheckExit(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    int depth;			/* Net stack effect */
+    int litIndex;		/* Index in the literal pool of the empty
+				 * string */
+    BasicBlock* curr_bb = assemEnvPtr->curr_bb;
+				/* Final basic block in the assembly */
+
+    /*
+     * Don't perform these checks if execution doesn't reach the exit (either
+     * because of an infinite loop or because the only return is from the
+     * middle.
+     */
+
+    if (curr_bb->flags & BB_VISITED) {
+	/*
+	 * Exit with no operands; push an empty one.
+	 */
+
+	depth = curr_bb->finalStackDepth + curr_bb->initialStackDepth;
+	if (depth == 0) {
+	    /*
+	     * Emit a 'push' of the empty literal.
+	     */
+
+	    litIndex = TclRegisterLiteral(envPtr, "", 0, 0);
+
+	    /*
+	     * Assumes that 'push' is at slot 0 in TalInstructionTable.
+	     */
+
+	    BBEmitInst1or4(assemEnvPtr, 0, litIndex, 0);
+	    ++depth;
+	}
+
+	/*
+	 * Exit with unbalanced stack.
+	 */
+
+	if (depth != 1) {
+	    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"stack is unbalanced on exit from the code (depth=%d)",
+			depth));
+		Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADSTACK", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Record stack usage.
+	 */
+
+	envPtr->currStackDepth += depth;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ProcessCatches --
+ *
+ *	First pass of 'catch' processing.
+ *
+ * Results:
+ *	Returns a standard Tcl result, with an appropriate error message if
+ *	the result is TCL_ERROR.
+ *
+ * Side effects:
+ *	Labels all basic blocks with their enclosing catches.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+ProcessCatches(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    BasicBlock* blockPtr;	/* Pointer to a basic block */
+
+    /*
+     * Clear the catch state of all basic blocks.
+     */
+
+    for (blockPtr = assemEnvPtr->head_bb;
+	    blockPtr != NULL;
+	    blockPtr = blockPtr->successor1) {
+	blockPtr->catchState = BBCS_UNKNOWN;
+	blockPtr->enclosingCatch = NULL;
+    }
+
+    /*
+     * Start the check recursively from the first basic block, which is
+     * outside any exception context
+     */
+
+    if (ProcessCatchesInBasicBlock(assemEnvPtr, assemEnvPtr->head_bb,
+	    NULL, BBCS_NONE, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check for unclosed catch on exit.
+     */
+
+    if (CheckForUnclosedCatches(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now there's enough information to build the exception ranges.
+     */
+
+    if (BuildExceptionRanges(assemEnvPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Finally, restore any exception ranges from embedded scripts.
+     */
+
+    RestoreEmbeddedExceptionRanges(assemEnvPtr);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ProcessCatchesInBasicBlock --
+ *
+ *	First-pass catch processing for one basic block.
+ *
+ * Results:
+ *	Returns a standard Tcl result, with error message in the interpreter
+ *	result if an error occurs.
+ *
+ * This procedure checks consistency of the exception context through the
+ * assembler program, and records the enclosing 'catch' for every basic block.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+ProcessCatchesInBasicBlock(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr,		/* Basic block being processed */
+    BasicBlock* enclosing,	/* Start basic block of the enclosing catch */
+    enum BasicBlockCatchState state,
+				/* BBCS_NONE, BBCS_INCATCH, or BBCS_CAUGHT */
+    int catchDepth)		/* Depth of nesting of catches */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    int result;			/* Return value from this procedure */
+    BasicBlock* fallThruEnclosing;
+				/* Enclosing catch if execution falls thru */
+    enum BasicBlockCatchState fallThruState;
+				/* Catch state of the successor block */
+    BasicBlock* jumpEnclosing;	/* Enclosing catch if execution goes to jump
+				 * target */
+    enum BasicBlockCatchState jumpState;
+				/* Catch state of the jump target */
+    int changed = 0;		/* Flag == 1 iff successor blocks need to be
+				 * checked because the state of this block has
+				 * changed. */
+    BasicBlock* jumpTarget;	/* Basic block where a jump goes */
+    Tcl_HashSearch jtSearch;	/* Hash search control for a jumptable */
+    Tcl_HashEntry* jtEntry;	/* Entry in a jumptable */
+    Tcl_Obj* targetLabel;	/* Target label from a jumptable */
+    Tcl_HashEntry* entry;	/* Entry from the label table */
+
+    /*
+     * Update the state of the current block, checking for consistency.  Set
+     * 'changed' to 1 if the state changes and successor blocks need to be
+     * rechecked.
+     */
+
+    if (bbPtr->catchState == BBCS_UNKNOWN) {
+	bbPtr->enclosingCatch = enclosing;
+    } else if (bbPtr->enclosingCatch != enclosing) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "execution reaches an instruction in inconsistent "
+		    "exception contexts", -1));
+	    Tcl_SetErrorLine(interp, bbPtr->startLine);
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADCATCH", NULL);
+	}
+	return TCL_ERROR;
+    }
+    if (state > bbPtr->catchState) {
+	bbPtr->catchState = state;
+	changed = 1;
+    }
+
+    /*
+     * If this block has been visited before, and its state hasn't changed,
+     * we're done with it for now.
+     */
+
+    if (!changed) {
+	return TCL_OK;
+    }
+    bbPtr->catchDepth = catchDepth;
+
+    /*
+     * Determine enclosing catch and 'caught' state for the fallthrough and
+     * the jump target. Default for both is the state of the current block.
+     */
+
+    fallThruEnclosing = enclosing;
+    fallThruState = state;
+    jumpEnclosing = enclosing;
+    jumpState = state;
+
+    /*
+     * TODO: Make sure that the test cases include validating that a natural
+     * loop can't include 'beginCatch' or 'endCatch'
+     */
+
+    if (bbPtr->flags & BB_BEGINCATCH) {
+	/*
+	 * If the block begins a catch, the state for the successor is 'in
+	 * catch'. The jump target is the exception exit, and the state of the
+	 * jump target is 'caught.'
+	 */
+
+	fallThruEnclosing = bbPtr;
+	fallThruState = BBCS_INCATCH;
+	jumpEnclosing = bbPtr;
+	jumpState = BBCS_CAUGHT;
+	++catchDepth;
+    }
+
+    if (bbPtr->flags & BB_ENDCATCH) {
+	/*
+	 * If the block ends a catch, the state for the successor is whatever
+	 * the state was on entry to the catch.
+	 */
+
+	if (enclosing == NULL) {
+	    if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"endCatch without a corresponding beginCatch", -1));
+		Tcl_SetErrorLine(interp, bbPtr->startLine);
+		Tcl_SetErrorCode(interp, "TCL", "ASSEM", "BADENDCATCH", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	fallThruEnclosing = enclosing->enclosingCatch;
+	fallThruState = enclosing->catchState;
+	--catchDepth;
+    }
+
+    /*
+     * Visit any successor blocks with the appropriate exception context
+     */
+
+    result = TCL_OK;
+    if (bbPtr->flags & BB_FALLTHRU) {
+	result = ProcessCatchesInBasicBlock(assemEnvPtr, bbPtr->successor1,
+		fallThruEnclosing, fallThruState, catchDepth);
+    }
+    if (result == TCL_OK && bbPtr->jumpTarget != NULL) {
+	entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		TclGetString(bbPtr->jumpTarget));
+	jumpTarget = Tcl_GetHashValue(entry);
+	result = ProcessCatchesInBasicBlock(assemEnvPtr, jumpTarget,
+		jumpEnclosing, jumpState, catchDepth);
+    }
+
+    /*
+     * All blocks referenced in a jump table are successors.
+     */
+
+    if (bbPtr->flags & BB_JUMPTABLE) {
+	for (jtEntry = Tcl_FirstHashEntry(&bbPtr->jtPtr->hashTable,&jtSearch);
+		result == TCL_OK && jtEntry != NULL;
+		jtEntry = Tcl_NextHashEntry(&jtSearch)) {
+	    targetLabel = Tcl_GetHashValue(jtEntry);
+	    entry = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		    TclGetString(targetLabel));
+	    jumpTarget = Tcl_GetHashValue(entry);
+	    result = ProcessCatchesInBasicBlock(assemEnvPtr, jumpTarget,
+		    jumpEnclosing, jumpState, catchDepth);
+	}
+    }
+
+    return result;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * CheckForUnclosedCatches --
+ *
+ *	Checks that a sequence of assembly code has no unclosed catches on
+ *	exit.
+ *
+ * Results:
+ *	Returns a standard Tcl result, with an error message for unclosed
+ *	catches.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+CheckForUnclosedCatches(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+
+    if (assemEnvPtr->curr_bb->catchState >= BBCS_INCATCH) {
+	if (assemEnvPtr->flags & TCL_EVAL_DIRECT) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "catch still active on exit from assembly code", -1));
+	    Tcl_SetErrorLine(interp,
+		    assemEnvPtr->curr_bb->enclosingCatch->startLine);
+	    Tcl_SetErrorCode(interp, "TCL", "ASSEM", "UNCLOSEDCATCH", NULL);
+	}
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * BuildExceptionRanges --
+ *
+ *	Walks through the assembly code and builds exception ranges for the
+ *	catches embedded therein.
+ *
+ * Results:
+ *	Returns a standard Tcl result with an error message in the interpreter
+ *	if anything is unsuccessful.
+ *
+ * Side effects:
+ *	Each contiguous block of code with a given catch exit is assigned an
+ *	exception range at the appropriate level.
+ *	Exception ranges in embedded blocks have their levels corrected and
+ *	collated into the table.
+ *	Blocks that end with 'beginCatch' are associated with the innermost
+ *	exception range of the following block.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+BuildExceptionRanges(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr;		/* Current basic block */
+    BasicBlock* prevPtr = NULL;	/* Previous basic block */
+    int catchDepth = 0;		/* Current catch depth */
+    int maxCatchDepth = 0;	/* Maximum catch depth in the program */
+    BasicBlock** catches;	/* Stack of catches in progress */
+    int* catchIndices;		/* Indices of the exception ranges of catches
+				 * in progress */
+    int i;
+
+    /*
+     * Determine the max catch depth for the entire assembly script
+     * (excluding embedded eval's and expr's, which will be handled later).
+     */
+
+    for (bbPtr=assemEnvPtr->head_bb; bbPtr != NULL; bbPtr=bbPtr->successor1) {
+	if (bbPtr->catchDepth > maxCatchDepth) {
+	    maxCatchDepth = bbPtr->catchDepth;
+	}
+    }
+
+    /*
+     * Allocate memory for a stack of active catches.
+     */
+
+    catches = ckalloc(maxCatchDepth * sizeof(BasicBlock*));
+    catchIndices = ckalloc(maxCatchDepth * sizeof(int));
+    for (i = 0; i < maxCatchDepth; ++i) {
+	catches[i] = NULL;
+	catchIndices[i] = -1;
+    }
+
+    /*
+     * Walk through the basic blocks and manage exception ranges.
+     */
+
+    for (bbPtr=assemEnvPtr->head_bb; bbPtr != NULL; bbPtr=bbPtr->successor1) {
+	UnstackExpiredCatches(envPtr, bbPtr, catchDepth, catches,
+		catchIndices);
+	LookForFreshCatches(bbPtr, catches);
+	StackFreshCatches(assemEnvPtr, bbPtr, catchDepth, catches,
+		catchIndices);
+
+	/*
+	 * If the last block was a 'begin catch', fill in the exception range.
+	 */
+
+	catchDepth = bbPtr->catchDepth;
+	if (prevPtr != NULL && (prevPtr->flags & BB_BEGINCATCH)) {
+	    TclStoreInt4AtPtr(catchIndices[catchDepth-1],
+		    envPtr->codeStart + bbPtr->startOffset - 4);
+	}
+
+	prevPtr = bbPtr;
+    }
+
+    /* Make sure that all catches are closed */
+
+    if (catchDepth != 0) {
+	Tcl_Panic("unclosed catch at end of code in "
+		"tclAssembly.c:BuildExceptionRanges, can't happen");
+    }
+
+    /* Free temp storage */
+
+    ckfree(catchIndices);
+    ckfree(catches);
+
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * UnstackExpiredCatches --
+ *
+ *	Unstacks and closes the exception ranges for any catch contexts that
+ *	were active in the previous basic block but are inactive in the
+ *	current one.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+UnstackExpiredCatches(
+    CompileEnv* envPtr,		/* Compilation environment */
+    BasicBlock* bbPtr,		/* Basic block being processed */
+    int catchDepth,		/* Depth of nesting of catches prior to entry
+				 * to this block */
+    BasicBlock** catches,	/* Array of catch contexts */
+    int* catchIndices)		/* Indices of the exception ranges
+				 * corresponding to the catch contexts */
+{
+    ExceptionRange* range;	/* Exception range for a specific catch */
+    BasicBlock* catch;		/* Catch block being examined */
+    BasicBlockCatchState catchState;
+				/* State of the code relative to the catch
+				 * block being examined ("in catch" or
+				 * "caught"). */
+
+    /*
+     * Unstack any catches that are deeper than the nesting level of the basic
+     * block being entered.
+     */
+
+    while (catchDepth > bbPtr->catchDepth) {
+	--catchDepth;
+	if (catches[catchDepth] != NULL) {
+	    range = envPtr->exceptArrayPtr + catchIndices[catchDepth];
+	    range->numCodeBytes = bbPtr->startOffset - range->codeOffset;
+	    catches[catchDepth] = NULL;
+	    catchIndices[catchDepth] = -1;
+	}
+    }
+
+    /*
+     * Unstack any catches that don't match the basic block being entered,
+     * either because they are no longer part of the context, or because the
+     * context has changed from INCATCH to CAUGHT.
+     */
+
+    catchState = bbPtr->catchState;
+    catch = bbPtr->enclosingCatch;
+    while (catchDepth > 0) {
+	--catchDepth;
+	if (catches[catchDepth] != NULL) {
+	    if (catches[catchDepth] != catch || catchState >= BBCS_CAUGHT) {
+		range = envPtr->exceptArrayPtr + catchIndices[catchDepth];
+		range->numCodeBytes = bbPtr->startOffset - range->codeOffset;
+		catches[catchDepth] = NULL;
+		catchIndices[catchDepth] = -1;
+	    }
+	    catchState = catch->catchState;
+	    catch = catch->enclosingCatch;
+	}
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * LookForFreshCatches --
+ *
+ *	Determines whether a basic block being entered needs any exception
+ *	ranges that are not already stacked.
+ *
+ * Does not create the ranges: this procedure iterates from the innermost
+ * catch outward, but exception ranges must be created from the outermost
+ * catch inward.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+LookForFreshCatches(
+    BasicBlock* bbPtr,		/* Basic block being entered */
+    BasicBlock** catches)	/* Array of catch contexts that are already
+				 * entered */
+{
+    BasicBlockCatchState catchState;
+				/* State ("in catch" or "caught") of the
+				 * current catch. */
+    BasicBlock* catch;		/* Current enclosing catch */
+    int catchDepth;		/* Nesting depth of the current catch */
+
+    catchState = bbPtr->catchState;
+    catch = bbPtr->enclosingCatch;
+    catchDepth = bbPtr->catchDepth;
+    while (catchDepth > 0) {
+	--catchDepth;
+	if (catches[catchDepth] != catch && catchState < BBCS_CAUGHT) {
+	    catches[catchDepth] = catch;
+	}
+	catchState = catch->catchState;
+	catch = catch->enclosingCatch;
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * StackFreshCatches --
+ *
+ *	Make ExceptionRange records for any catches that are in the basic
+ *	block being entered and were not in the previous basic block.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+StackFreshCatches(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr,		/* Basic block being processed */
+    int catchDepth,		/* Depth of nesting of catches prior to entry
+				 * to this block */
+    BasicBlock** catches,	/* Array of catch contexts */
+    int* catchIndices)		/* Indices of the exception ranges
+				 * corresponding to the catch contexts */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    ExceptionRange* range;	/* Exception range for a specific catch */
+    BasicBlock* catch;		/* Catch block being examined */
+    BasicBlock* errorExit;	/* Error exit from the catch block */
+    Tcl_HashEntry* entryPtr;
+
+    catchDepth = 0;
+
+    /*
+     * Iterate through the enclosing catch blocks from the outside in,
+     * looking for ones that don't have exception ranges (and are uncaught)
+     */
+
+    for (catchDepth = 0; catchDepth < bbPtr->catchDepth; ++catchDepth) {
+	if (catchIndices[catchDepth] == -1 && catches[catchDepth] != NULL) {
+	    /*
+	     * Create an exception range for a block that needs one.
+	     */
+
+	    catch = catches[catchDepth];
+	    catchIndices[catchDepth] =
+		    TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+	    range = envPtr->exceptArrayPtr + catchIndices[catchDepth];
+	    range->nestingLevel = envPtr->exceptDepth + catchDepth;
+	    envPtr->maxExceptDepth =
+		    TclMax(range->nestingLevel + 1, envPtr->maxExceptDepth);
+	    range->codeOffset = bbPtr->startOffset;
+
+	    entryPtr = Tcl_FindHashEntry(&assemEnvPtr->labelHash,
+		    TclGetString(catch->jumpTarget));
+	    if (entryPtr == NULL) {
+		Tcl_Panic("undefined label in tclAssembly.c:"
+			"BuildExceptionRanges, can't happen");
+	    }
+
+	    errorExit = Tcl_GetHashValue(entryPtr);
+	    range->catchOffset = errorExit->startOffset;
+	}
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * RestoreEmbeddedExceptionRanges --
+ *
+ *	Processes an assembly script, replacing any exception ranges that
+ *	were present in embedded code.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+RestoreEmbeddedExceptionRanges(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    BasicBlock* bbPtr;		/* Current basic block */
+    int rangeBase;		/* Base of the foreign exception ranges when
+				 * they are reinstalled */
+    int rangeIndex;		/* Index of the current foreign exception
+				 * range as reinstalled */
+    ExceptionRange* range;	/* Current foreign exception range */
+    unsigned char opcode;	/* Current instruction's opcode */
+    int catchIndex;		/* Index of the exception range to which the
+				 * current instruction refers */
+    int i;
+
+    /*
+     * Walk the basic blocks looking for exceptions in embedded scripts.
+     */
+
+    for (bbPtr = assemEnvPtr->head_bb;
+	    bbPtr != NULL;
+	    bbPtr = bbPtr->successor1) {
+	if (bbPtr->foreignExceptionCount != 0) {
+	    /*
+	     * Reinstall the embedded exceptions and track their nesting level
+	     */
+
+	    rangeBase = envPtr->exceptArrayNext;
+	    for (i = 0; i < bbPtr->foreignExceptionCount; ++i) {
+		range = bbPtr->foreignExceptions + i;
+		rangeIndex = TclCreateExceptRange(range->type, envPtr);
+		range->nestingLevel += envPtr->exceptDepth + bbPtr->catchDepth;
+		memcpy(envPtr->exceptArrayPtr + rangeIndex, range,
+			sizeof(ExceptionRange));
+		if (range->nestingLevel >= envPtr->maxExceptDepth) {
+		    envPtr->maxExceptDepth = range->nestingLevel + 1;
+		}
+	    }
+
+	    /*
+	     * Walk through the bytecode of the basic block, and relocate
+	     * INST_BEGIN_CATCH4 instructions to the new locations
+	     */
+
+	    i = bbPtr->startOffset;
+	    while (i < bbPtr->successor1->startOffset) {
+		opcode = envPtr->codeStart[i];
+		if (opcode == INST_BEGIN_CATCH4) {
+		    catchIndex = TclGetUInt4AtPtr(envPtr->codeStart + i + 1);
+		    if (catchIndex >= bbPtr->foreignExceptionBase
+			    && catchIndex < (bbPtr->foreignExceptionBase +
+			    bbPtr->foreignExceptionCount)) {
+			catchIndex -= bbPtr->foreignExceptionBase;
+			catchIndex += rangeBase;
+			TclStoreInt4AtPtr(catchIndex, envPtr->codeStart+i+1);
+		    }
+		}
+		i += tclInstructionTable[opcode].numBytes;
+	    }
+	}
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ResetVisitedBasicBlocks --
+ *
+ *	Turns off the 'visited' flag in all basic blocks at the conclusion
+ *	of a pass.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+ResetVisitedBasicBlocks(
+    AssemblyEnv* assemEnvPtr)	/* Assembly environment */
+{
+    BasicBlock* block;
+
+    for (block = assemEnvPtr->head_bb; block != NULL;
+	    block = block->successor1) {
+	block->flags &= ~BB_VISITED;
+    }
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * AddBasicBlockRangeToErrorInfo --
+ *
+ *	Updates the error info of the Tcl interpreter to show a given basic
+ *	block in the code.
+ *
+ * This procedure is used to label the callstack with source location
+ * information when reporting an error in stack checking.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+AddBasicBlockRangeToErrorInfo(
+    AssemblyEnv* assemEnvPtr,	/* Assembly environment */
+    BasicBlock* bbPtr)		/* Basic block in which the error is found */
+{
+    CompileEnv* envPtr = assemEnvPtr->envPtr;
+				/* Compilation environment */
+    Tcl_Interp* interp = (Tcl_Interp*) envPtr->iPtr;
+				/* Tcl interpreter */
+    Tcl_Obj* lineNo;		/* Line number in the source */
+
+    Tcl_AddErrorInfo(interp, "\n    in assembly code between lines ");
+    lineNo = Tcl_NewIntObj(bbPtr->startLine);
+    Tcl_IncrRefCount(lineNo);
+    Tcl_AppendObjToErrorInfo(interp, lineNo);
+    Tcl_AddErrorInfo(interp, " and ");
+    if (bbPtr->successor1 != NULL) {
+	TclSetLongObj(lineNo, bbPtr->successor1->startLine);
+	Tcl_AppendObjToErrorInfo(interp, lineNo);
+    } else {
+	Tcl_AddErrorInfo(interp, "end of assembly code");
+    }
+    Tcl_DecrRefCount(lineNo);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * DupAssembleCodeInternalRep --
+ *
+ *	Part of the Tcl object type implementation for Tcl assembly language
+ *	bytecode. We do not copy the bytecode intrep. Instead, we return
+ *	without setting copyPtr->typePtr, so the copy is a plain string copy
+ *	of the assembly source, and if it is to be used as a compiled
+ *	expression, it will need to be reprocessed.
+ *
+ *	This makes sense, because with Tcl's copy-on-write practices, the
+ *	usual (only?) time Tcl_DuplicateObj() will be called is when the copy
+ *	is about to be modified, which would invalidate any copied bytecode
+ *	anyway. The only reason it might make sense to copy the bytecode is if
+ *	we had some modifying routines that operated directly on the intrep,
+ *	as we do for lists and dicts.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+DupAssembleCodeInternalRep(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *copyPtr)
+{
+    return;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * FreeAssembleCodeInternalRep --
+ *
+ *	Part of the Tcl object type implementation for Tcl expression
+ *	bytecode. Frees the storage allocated to hold the internal rep, unless
+ *	ref counts indicate bytecode execution is still in progress.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May free allocated memory. Leaves objPtr untyped.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static void
+FreeAssembleCodeInternalRep(
+    Tcl_Obj *objPtr)
+{
+    ByteCode *codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclReleaseByteCode(codePtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclAsync.c b/generic/tclAsync.c
new file mode 100644
index 0000000..14804e4
--- /dev/null
+++ b/generic/tclAsync.c
@@ -0,0 +1,355 @@
+/*
+ * tclAsync.c --
+ *
+ *	This file provides low-level support needed to invoke signal handlers
+ *	in a safe way. The code here doesn't actually handle signals, though.
+ *	This code is based on proposals made by Mark Diekhans and Don Libes.
+ *
+ * Copyright (c) 1993 The Regents of the University of California.
+ * Copyright (c) 1994 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/* Forward declaration */
+struct ThreadSpecificData;
+
+/*
+ * One of the following structures exists for each asynchronous handler:
+ */
+
+typedef struct AsyncHandler {
+    int ready;			/* Non-zero means this handler should be
+				 * invoked in the next call to
+				 * Tcl_AsyncInvoke. */
+    struct AsyncHandler *nextPtr;
+				/* Next in list of all handlers for the
+				 * process. */
+    Tcl_AsyncProc *proc;	/* Procedure to call when handler is
+				 * invoked. */
+    ClientData clientData;	/* Value to pass to handler when it is
+				 * invoked. */
+    struct ThreadSpecificData *originTsd;
+				/* Used in Tcl_AsyncMark to modify thread-
+				 * specific data from outside the thread it is
+				 * associated to. */
+    Tcl_ThreadId originThrdId;	/* Origin thread where this token was created
+				 * and where it will be yielded. */
+} AsyncHandler;
+
+typedef struct ThreadSpecificData {
+    /*
+     * The variables below maintain a list of all existing handlers specific
+     * to the calling thread.
+     */
+    AsyncHandler *firstHandler;	/* First handler defined for process, or NULL
+				 * if none. */
+    AsyncHandler *lastHandler;	/* Last handler or NULL. */
+    int asyncReady;		/* This is set to 1 whenever a handler becomes
+				 * ready and it is cleared to zero whenever
+				 * Tcl_AsyncInvoke is called. It can be
+				 * checked elsewhere in the application by
+				 * calling Tcl_AsyncReady to see if
+				 * Tcl_AsyncInvoke should be invoked. */
+    int asyncActive;		/* Indicates whether Tcl_AsyncInvoke is
+				 * currently working. If so then we won't set
+				 * asyncReady again until Tcl_AsyncInvoke
+				 * returns. */
+    Tcl_Mutex asyncMutex;	/* Thread-specific AsyncHandler linked-list
+				 * lock */
+} ThreadSpecificData;
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeAsync --
+ *
+ *	Finalizes the mutex in the thread local data structure for the async
+ *	subsystem.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Forgets knowledge of the mutex should it have been created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeAsync(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->asyncMutex != NULL) {
+	Tcl_MutexFinalize(&tsdPtr->asyncMutex);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AsyncCreate --
+ *
+ *	This procedure creates the data structures for an asynchronous
+ *	handler, so that no memory has to be allocated when the handler is
+ *	activated.
+ *
+ * Results:
+ *	The return value is a token for the handler, which can be used to
+ *	activate it later on.
+ *
+ * Side effects:
+ *	Information about the handler is recorded.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_AsyncHandler
+Tcl_AsyncCreate(
+    Tcl_AsyncProc *proc,	/* Procedure to call when handler is
+				 * invoked. */
+    ClientData clientData)	/* Argument to pass to handler. */
+{
+    AsyncHandler *asyncPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    asyncPtr = ckalloc(sizeof(AsyncHandler));
+    asyncPtr->ready = 0;
+    asyncPtr->nextPtr = NULL;
+    asyncPtr->proc = proc;
+    asyncPtr->clientData = clientData;
+    asyncPtr->originTsd = tsdPtr;
+    asyncPtr->originThrdId = Tcl_GetCurrentThread();
+
+    Tcl_MutexLock(&tsdPtr->asyncMutex);
+    if (tsdPtr->firstHandler == NULL) {
+	tsdPtr->firstHandler = asyncPtr;
+    } else {
+	tsdPtr->lastHandler->nextPtr = asyncPtr;
+    }
+    tsdPtr->lastHandler = asyncPtr;
+    Tcl_MutexUnlock(&tsdPtr->asyncMutex);
+    return (Tcl_AsyncHandler) asyncPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AsyncMark --
+ *
+ *	This procedure is called to request that an asynchronous handler be
+ *	invoked as soon as possible. It's typically called from an interrupt
+ *	handler, where it isn't safe to do anything that depends on or
+ *	modifies application state.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The handler gets marked for invocation later.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AsyncMark(
+    Tcl_AsyncHandler async)		/* Token for handler. */
+{
+    AsyncHandler *token = (AsyncHandler *) async;
+
+    Tcl_MutexLock(&token->originTsd->asyncMutex);
+    token->ready = 1;
+    if (!token->originTsd->asyncActive) {
+	token->originTsd->asyncReady = 1;
+	Tcl_ThreadAlert(token->originThrdId);
+    }
+    Tcl_MutexUnlock(&token->originTsd->asyncMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AsyncInvoke --
+ *
+ *	This procedure is called at a "safe" time at background level to
+ *	invoke any active asynchronous handlers.
+ *
+ * Results:
+ *	The return value is a normal Tcl result, which is intended to replace
+ *	the code argument as the current completion code for interp.
+ *
+ * Side effects:
+ *	Depends on the handlers that are active.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AsyncInvoke(
+    Tcl_Interp *interp,		/* If invoked from Tcl_Eval just after
+				 * completing a command, points to
+				 * interpreter. Otherwise it is NULL. */
+    int code)			/* If interp is non-NULL, this gives
+				 * completion code from command that just
+				 * completed. */
+{
+    AsyncHandler *asyncPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    Tcl_MutexLock(&tsdPtr->asyncMutex);
+
+    if (tsdPtr->asyncReady == 0) {
+	Tcl_MutexUnlock(&tsdPtr->asyncMutex);
+	return code;
+    }
+    tsdPtr->asyncReady = 0;
+    tsdPtr->asyncActive = 1;
+    if (interp == NULL) {
+	code = 0;
+    }
+
+    /*
+     * Make one or more passes over the list of handlers, invoking at most one
+     * handler in each pass. After invoking a handler, go back to the start of
+     * the list again so that (a) if a new higher-priority handler gets marked
+     * while executing a lower priority handler, we execute the higher-
+     * priority handler next, and (b) if a handler gets deleted during the
+     * execution of a handler, then the list structure may change so it isn't
+     * safe to continue down the list anyway.
+     */
+
+    while (1) {
+	for (asyncPtr = tsdPtr->firstHandler; asyncPtr != NULL;
+		asyncPtr = asyncPtr->nextPtr) {
+	    if (asyncPtr->ready) {
+		break;
+	    }
+	}
+	if (asyncPtr == NULL) {
+	    break;
+	}
+	asyncPtr->ready = 0;
+	Tcl_MutexUnlock(&tsdPtr->asyncMutex);
+	code = asyncPtr->proc(asyncPtr->clientData, interp, code);
+	Tcl_MutexLock(&tsdPtr->asyncMutex);
+    }
+    tsdPtr->asyncActive = 0;
+    Tcl_MutexUnlock(&tsdPtr->asyncMutex);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AsyncDelete --
+ *
+ *	Frees up all the state for an asynchronous handler. The handler should
+ *	never be used again.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The state associated with the handler is deleted.
+ *
+ *	Failure to locate the handler in current thread private list
+ *	of async handlers will result in panic; exception: the list
+ *	is already empty (potential trouble?).
+ *	Consequently, threads should create and delete handlers
+ *	themselves.  I.e. a handler created by one should not be
+ *	deleted by some other thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AsyncDelete(
+    Tcl_AsyncHandler async)		/* Token for handler to delete. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    AsyncHandler *asyncPtr = (AsyncHandler *) async;
+    AsyncHandler *prevPtr, *thisPtr;
+
+    /*
+     * Assure early handling of the constraint
+     */
+
+    if (asyncPtr->originThrdId != Tcl_GetCurrentThread()) {
+	Tcl_Panic("Tcl_AsyncDelete: async handler deleted by the wrong thread");
+    }
+
+    /*
+     * If we come to this point when TSD's for the current
+     * thread have already been garbage-collected, we are
+     * in the _serious_ trouble. OTOH, we tolerate calling
+     * with already cleaned-up handler list (should we?).
+     */
+
+    Tcl_MutexLock(&tsdPtr->asyncMutex);
+    if (tsdPtr->firstHandler != NULL) {
+	prevPtr = thisPtr = tsdPtr->firstHandler;
+	while (thisPtr != NULL && thisPtr != asyncPtr) {
+	    prevPtr = thisPtr;
+	    thisPtr = thisPtr->nextPtr;
+	}
+	if (thisPtr == NULL) {
+	    Tcl_Panic("Tcl_AsyncDelete: cannot find async handler");
+	}
+	if (asyncPtr == tsdPtr->firstHandler) {
+	    tsdPtr->firstHandler = asyncPtr->nextPtr;
+	} else {
+	    prevPtr->nextPtr = asyncPtr->nextPtr;
+	}
+	if (asyncPtr == tsdPtr->lastHandler) {
+	    tsdPtr->lastHandler = prevPtr;
+	}
+    }
+    Tcl_MutexUnlock(&tsdPtr->asyncMutex);
+    ckfree(asyncPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AsyncReady --
+ *
+ *	This procedure can be used to tell whether Tcl_AsyncInvoke needs to be
+ *	called. This procedure is the external interface for checking the
+ *	thread-specific asyncReady variable.
+ *
+ * Results:
+ *	The return value is 1 whenever a handler is ready and is 0 when no
+ *	handlers are ready.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AsyncReady(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    return tsdPtr->asyncReady;
+}
+
+int *
+TclGetAsyncReadyPtr(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    return &(tsdPtr->asyncReady);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclBasic.c b/generic/tclBasic.c
new file mode 100644
index 0000000..9d1e98d
--- /dev/null
+++ b/generic/tclBasic.c
@@ -0,0 +1,9108 @@
+/*
+ * tclBasic.c --
+ *
+ *	Contains the basic facilities for TCL command interpretation,
+ *	including interpreter creation and deletion, command creation and
+ *	deletion, and command/script execution.
+ *
+ * Copyright (c) 1987-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 2001, 2002 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ * Copyright (c) 2006-2008 by Joe Mistachkin.  All rights reserved.
+ * Copyright (c) 2008 Miguel Sofer <msofer@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclOOInt.h"
+#include "tclCompile.h"
+#include "tommath.h"
+#include <math.h>
+#include <assert.h>
+
+#define INTERP_STACK_INITIAL_SIZE 2000
+#define CORO_STACK_INITIAL_SIZE    200
+
+/*
+ * Determine whether we're using IEEE floating point
+ */
+
+#if (FLT_RADIX == 2) && (DBL_MANT_DIG == 53) && (DBL_MAX_EXP == 1024)
+#   define IEEE_FLOATING_POINT
+/* Largest odd integer that can be represented exactly in a double */
+#   define MAX_EXACT 9007199254740991.0
+#endif
+
+/*
+ * The following structure defines the client data for a math function
+ * registered with Tcl_CreateMathFunc
+ */
+
+typedef struct OldMathFuncData {
+    Tcl_MathProc *proc;		/* Handler function */
+    int numArgs;		/* Number of args expected */
+    Tcl_ValueType *argTypes;	/* Types of the args */
+    ClientData clientData;	/* Client data for the handler function */
+} OldMathFuncData;
+
+/*
+ * This is the script cancellation struct and hash table. The hash table is
+ * used to keep track of the information necessary to process script
+ * cancellation requests, including the original interp, asynchronous handler
+ * tokens (created by Tcl_AsyncCreate), and the clientData and flags arguments
+ * passed to Tcl_CancelEval on a per-interp basis. The cancelLock mutex is
+ * used for protecting calls to Tcl_CancelEval as well as protecting access to
+ * the hash table below.
+ */
+
+typedef struct {
+    Tcl_Interp *interp;		/* Interp this struct belongs to. */
+    Tcl_AsyncHandler async;	/* Async handler token for script
+				 * cancellation. */
+    char *result;		/* The script cancellation result or NULL for
+				 * a default result. */
+    int length;			/* Length of the above error message. */
+    ClientData clientData;	/* Ignored */
+    int flags;			/* Additional flags */
+} CancelInfo;
+static Tcl_HashTable cancelTable;
+static int cancelTableInitialized = 0;	/* 0 means not yet initialized. */
+TCL_DECLARE_MUTEX(cancelLock)
+
+/*
+ * Declarations for managing contexts for non-recursive coroutines. Contexts
+ * are used to save the evaluation state between NR calls to each coro.
+ */
+
+#define SAVE_CONTEXT(context)				\
+    (context).framePtr = iPtr->framePtr;		\
+    (context).varFramePtr = iPtr->varFramePtr;		\
+    (context).cmdFramePtr = iPtr->cmdFramePtr;		\
+    (context).lineLABCPtr = iPtr->lineLABCPtr
+
+#define RESTORE_CONTEXT(context)			\
+    iPtr->framePtr = (context).framePtr;		\
+    iPtr->varFramePtr = (context).varFramePtr;		\
+    iPtr->cmdFramePtr = (context).cmdFramePtr;		\
+    iPtr->lineLABCPtr = (context).lineLABCPtr
+
+/*
+ * Static functions in this file:
+ */
+
+static char *		CallCommandTraces(Interp *iPtr, Command *cmdPtr,
+			    const char *oldName, const char *newName,
+			    int flags);
+static int		CancelEvalProc(ClientData clientData,
+			    Tcl_Interp *interp, int code);
+static int		CheckDoubleResult(Tcl_Interp *interp, double dResult);
+static void		DeleteCoroutine(ClientData clientData);
+static void		DeleteInterpProc(Tcl_Interp *interp);
+static void		DeleteOpCmdClientData(ClientData clientData);
+#ifdef USE_DTRACE
+static Tcl_ObjCmdProc	DTraceObjCmd;
+static Tcl_NRPostProc	DTraceCmdReturn;
+#else
+#   define DTraceCmdReturn	NULL
+#endif /* USE_DTRACE */
+static Tcl_ObjCmdProc	ExprAbsFunc;
+static Tcl_ObjCmdProc	ExprBinaryFunc;
+static Tcl_ObjCmdProc	ExprBoolFunc;
+static Tcl_ObjCmdProc	ExprCeilFunc;
+static Tcl_ObjCmdProc	ExprDoubleFunc;
+static Tcl_ObjCmdProc	ExprEntierFunc;
+static Tcl_ObjCmdProc	ExprFloorFunc;
+static Tcl_ObjCmdProc	ExprIntFunc;
+static Tcl_ObjCmdProc	ExprIsqrtFunc;
+static Tcl_ObjCmdProc	ExprRandFunc;
+static Tcl_ObjCmdProc	ExprRoundFunc;
+static Tcl_ObjCmdProc	ExprSqrtFunc;
+static Tcl_ObjCmdProc	ExprSrandFunc;
+static Tcl_ObjCmdProc	ExprUnaryFunc;
+static Tcl_ObjCmdProc	ExprWideFunc;
+static void		MathFuncWrongNumArgs(Tcl_Interp *interp, int expected,
+			    int actual, Tcl_Obj *const *objv);
+static Tcl_NRPostProc	NRCoroutineCallerCallback;
+static Tcl_NRPostProc	NRCoroutineExitCallback;
+static Tcl_NRPostProc	NRCommand;
+
+static Tcl_ObjCmdProc	OldMathFuncProc;
+static void		OldMathFuncDeleteProc(ClientData clientData);
+static void		ProcessUnexpectedResult(Tcl_Interp *interp,
+			    int returnCode);
+static int		RewindCoroutine(CoroutineData *corPtr, int result);
+static void		TEOV_SwitchVarFrame(Tcl_Interp *interp);
+static void		TEOV_PushExceptionHandlers(Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[], int flags);
+static inline Command *	TEOV_LookupCmdFromObj(Tcl_Interp *interp,
+			    Tcl_Obj *namePtr, Namespace *lookupNsPtr);
+static int		TEOV_NotFound(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[], Namespace *lookupNsPtr);
+static int		TEOV_RunEnterTraces(Tcl_Interp *interp,
+			    Command **cmdPtrPtr, Tcl_Obj *commandPtr, int objc,
+			    Tcl_Obj *const objv[]);
+static Tcl_NRPostProc	RewindCoroutineCallback;
+static Tcl_NRPostProc	TEOEx_ByteCodeCallback;
+static Tcl_NRPostProc	TEOEx_ListCallback;
+static Tcl_NRPostProc	TEOV_Error;
+static Tcl_NRPostProc	TEOV_Exception;
+static Tcl_NRPostProc	TEOV_NotFoundCallback;
+static Tcl_NRPostProc	TEOV_RestoreVarFrame;
+static Tcl_NRPostProc	TEOV_RunLeaveTraces;
+static Tcl_NRPostProc	EvalObjvCore;
+static Tcl_NRPostProc	Dispatch;
+
+static Tcl_ObjCmdProc NRCoroInjectObjCmd;
+static Tcl_NRPostProc NRPostInvoke;
+
+MODULE_SCOPE const TclStubs tclStubs;
+
+/*
+ * Magical counts for the number of arguments accepted by a coroutine command
+ * after particular kinds of [yield].
+ */
+
+#define CORO_ACTIVATE_YIELD    PTR2INT(NULL)
+#define CORO_ACTIVATE_YIELDM   PTR2INT(NULL)+1
+
+#define COROUTINE_ARGUMENTS_SINGLE_OPTIONAL     (-1)
+#define COROUTINE_ARGUMENTS_ARBITRARY           (-2)
+
+/*
+ * The following structure define the commands in the Tcl core.
+ */
+
+typedef struct {
+    const char *name;		/* Name of object-based command. */
+    Tcl_ObjCmdProc *objProc;	/* Object-based function for command. */
+    CompileProc *compileProc;	/* Function called to compile command. */
+    Tcl_ObjCmdProc *nreProc;	/* NR-based function for command */
+    int flags;			/* Various flag bits, as defined below. */
+} CmdInfo;
+
+#define CMD_IS_SAFE         1   /* Whether this command is part of the set of
+                                 * commands present by default in a safe
+                                 * interpreter. */
+/* CMD_COMPILES_EXPANDED - Whether the compiler for this command can handle
+ * expansion for itself rather than needing the generic layer to take care of
+ * it for it. Defined in tclInt.h. */
+
+/*
+ * The built-in commands, and the functions that implement them:
+ */
+
+static const CmdInfo builtInCmds[] = {
+    /*
+     * Commands in the generic core.
+     */
+
+    {"append",		Tcl_AppendObjCmd,	TclCompileAppendCmd,	NULL,	CMD_IS_SAFE},
+    {"apply",		Tcl_ApplyObjCmd,	NULL,			TclNRApplyObjCmd,	CMD_IS_SAFE},
+    {"break",		Tcl_BreakObjCmd,	TclCompileBreakCmd,	NULL,	CMD_IS_SAFE},
+#ifndef TCL_NO_DEPRECATED
+    {"case",		Tcl_CaseObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+#endif
+    {"catch",		Tcl_CatchObjCmd,	TclCompileCatchCmd,	TclNRCatchObjCmd,	CMD_IS_SAFE},
+    {"concat",		Tcl_ConcatObjCmd,	TclCompileConcatCmd,	NULL,	CMD_IS_SAFE},
+    {"continue",	Tcl_ContinueObjCmd,	TclCompileContinueCmd,	NULL,	CMD_IS_SAFE},
+    {"coroutine",	NULL,			NULL,			TclNRCoroutineObjCmd,	CMD_IS_SAFE},
+    {"error",		Tcl_ErrorObjCmd,	TclCompileErrorCmd,	NULL,	CMD_IS_SAFE},
+    {"eval",		Tcl_EvalObjCmd,		NULL,			TclNREvalObjCmd,	CMD_IS_SAFE},
+    {"expr",		Tcl_ExprObjCmd,		TclCompileExprCmd,	TclNRExprObjCmd,	CMD_IS_SAFE},
+    {"for",		Tcl_ForObjCmd,		TclCompileForCmd,	TclNRForObjCmd,	CMD_IS_SAFE},
+    {"foreach",		Tcl_ForeachObjCmd,	TclCompileForeachCmd,	TclNRForeachCmd,	CMD_IS_SAFE},
+    {"format",		Tcl_FormatObjCmd,	TclCompileFormatCmd,	NULL,	CMD_IS_SAFE},
+    {"global",		Tcl_GlobalObjCmd,	TclCompileGlobalCmd,	NULL,	CMD_IS_SAFE},
+    {"if",		Tcl_IfObjCmd,		TclCompileIfCmd,	TclNRIfObjCmd,	CMD_IS_SAFE},
+    {"incr",		Tcl_IncrObjCmd,		TclCompileIncrCmd,	NULL,	CMD_IS_SAFE},
+    {"join",		Tcl_JoinObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"lappend",		Tcl_LappendObjCmd,	TclCompileLappendCmd,	NULL,	CMD_IS_SAFE},
+    {"lassign",		Tcl_LassignObjCmd,	TclCompileLassignCmd,	NULL,	CMD_IS_SAFE},
+    {"lindex",		Tcl_LindexObjCmd,	TclCompileLindexCmd,	NULL,	CMD_IS_SAFE},
+    {"linsert",		Tcl_LinsertObjCmd,	TclCompileLinsertCmd,	NULL,	CMD_IS_SAFE},
+    {"list",		Tcl_ListObjCmd,		TclCompileListCmd,	NULL,	CMD_IS_SAFE|CMD_COMPILES_EXPANDED},
+    {"llength",		Tcl_LlengthObjCmd,	TclCompileLlengthCmd,	NULL,	CMD_IS_SAFE},
+    {"lmap",		Tcl_LmapObjCmd,		TclCompileLmapCmd,	TclNRLmapCmd,	CMD_IS_SAFE},
+    {"lrange",		Tcl_LrangeObjCmd,	TclCompileLrangeCmd,	NULL,	CMD_IS_SAFE},
+    {"lrepeat",		Tcl_LrepeatObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"lreplace",	Tcl_LreplaceObjCmd,	TclCompileLreplaceCmd,	NULL,	CMD_IS_SAFE},
+    {"lreverse",	Tcl_LreverseObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"lsearch",		Tcl_LsearchObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"lset",		Tcl_LsetObjCmd,		TclCompileLsetCmd,	NULL,	CMD_IS_SAFE},
+    {"lsort",		Tcl_LsortObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"package",		Tcl_PackageObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"proc",		Tcl_ProcObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"regexp",		Tcl_RegexpObjCmd,	TclCompileRegexpCmd,	NULL,	CMD_IS_SAFE},
+    {"regsub",		Tcl_RegsubObjCmd,	TclCompileRegsubCmd,	NULL,	CMD_IS_SAFE},
+    {"rename",		Tcl_RenameObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"return",		Tcl_ReturnObjCmd,	TclCompileReturnCmd,	NULL,	CMD_IS_SAFE},
+    {"scan",		Tcl_ScanObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"set",		Tcl_SetObjCmd,		TclCompileSetCmd,	NULL,	CMD_IS_SAFE},
+    {"split",		Tcl_SplitObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"subst",		Tcl_SubstObjCmd,	TclCompileSubstCmd,	TclNRSubstObjCmd,	CMD_IS_SAFE},
+    {"switch",		Tcl_SwitchObjCmd,	TclCompileSwitchCmd,	TclNRSwitchObjCmd, CMD_IS_SAFE},
+    {"tailcall",	NULL,			TclCompileTailcallCmd,	TclNRTailcallObjCmd,	CMD_IS_SAFE},
+    {"throw",		Tcl_ThrowObjCmd,	TclCompileThrowCmd,	NULL,	CMD_IS_SAFE},
+    {"trace",		Tcl_TraceObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"try",		Tcl_TryObjCmd,		TclCompileTryCmd,	TclNRTryObjCmd,	CMD_IS_SAFE},
+    {"unset",		Tcl_UnsetObjCmd,	TclCompileUnsetCmd,	NULL,	CMD_IS_SAFE},
+    {"uplevel",		Tcl_UplevelObjCmd,	NULL,			TclNRUplevelObjCmd,	CMD_IS_SAFE},
+    {"upvar",		Tcl_UpvarObjCmd,	TclCompileUpvarCmd,	NULL,	CMD_IS_SAFE},
+    {"variable",	Tcl_VariableObjCmd,	TclCompileVariableCmd,	NULL,	CMD_IS_SAFE},
+    {"while",		Tcl_WhileObjCmd,	TclCompileWhileCmd,	TclNRWhileObjCmd,	CMD_IS_SAFE},
+    {"yield",		NULL,			TclCompileYieldCmd,	TclNRYieldObjCmd,	CMD_IS_SAFE},
+    {"yieldto",		NULL,			TclCompileYieldToCmd,	TclNRYieldToObjCmd,	CMD_IS_SAFE},
+
+    /*
+     * Commands in the OS-interface. Note that many of these are unsafe.
+     */
+
+    {"after",		Tcl_AfterObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"cd",		Tcl_CdObjCmd,		NULL,			NULL,	0},
+    {"close",		Tcl_CloseObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"eof",		Tcl_EofObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"exec",		Tcl_ExecObjCmd,		NULL,			NULL,	0},
+    {"exit",		Tcl_ExitObjCmd,		NULL,			NULL,	0},
+    {"fblocked",	Tcl_FblockedObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"fconfigure",	Tcl_FconfigureObjCmd,	NULL,			NULL,	0},
+    {"fcopy",		Tcl_FcopyObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"fileevent",	Tcl_FileEventObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"flush",		Tcl_FlushObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"gets",		Tcl_GetsObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"glob",		Tcl_GlobObjCmd,		NULL,			NULL,	0},
+    {"load",		Tcl_LoadObjCmd,		NULL,			NULL,	0},
+    {"open",		Tcl_OpenObjCmd,		NULL,			NULL,	0},
+    {"pid",		Tcl_PidObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"puts",		Tcl_PutsObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"pwd",		Tcl_PwdObjCmd,		NULL,			NULL,	0},
+    {"read",		Tcl_ReadObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"seek",		Tcl_SeekObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"socket",		Tcl_SocketObjCmd,	NULL,			NULL,	0},
+    {"source",		Tcl_SourceObjCmd,	NULL,			TclNRSourceObjCmd,	0},
+    {"tell",		Tcl_TellObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"time",		Tcl_TimeObjCmd,		NULL,			NULL,	CMD_IS_SAFE},
+    {"unload",		Tcl_UnloadObjCmd,	NULL,			NULL,	0},
+    {"update",		Tcl_UpdateObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {"vwait",		Tcl_VwaitObjCmd,	NULL,			NULL,	CMD_IS_SAFE},
+    {NULL,		NULL,			NULL,			NULL,	0}
+};
+
+/*
+ * Math functions. All are safe.
+ */
+
+typedef struct {
+    const char *name;		/* Name of the function. The full name is
+				 * "::tcl::mathfunc::<name>". */
+    Tcl_ObjCmdProc *objCmdProc;	/* Function that evaluates the function */
+    ClientData clientData;	/* Client data for the function */
+} BuiltinFuncDef;
+static const BuiltinFuncDef BuiltinFuncTable[] = {
+    { "abs",	ExprAbsFunc,	NULL			},
+    { "acos",	ExprUnaryFunc,	(ClientData) acos	},
+    { "asin",	ExprUnaryFunc,	(ClientData) asin	},
+    { "atan",	ExprUnaryFunc,	(ClientData) atan	},
+    { "atan2",	ExprBinaryFunc,	(ClientData) atan2	},
+    { "bool",	ExprBoolFunc,	NULL			},
+    { "ceil",	ExprCeilFunc,	NULL			},
+    { "cos",	ExprUnaryFunc,	(ClientData) cos	},
+    { "cosh",	ExprUnaryFunc,	(ClientData) cosh	},
+    { "double",	ExprDoubleFunc,	NULL			},
+    { "entier",	ExprEntierFunc,	NULL			},
+    { "exp",	ExprUnaryFunc,	(ClientData) exp	},
+    { "floor",	ExprFloorFunc,	NULL			},
+    { "fmod",	ExprBinaryFunc,	(ClientData) fmod	},
+    { "hypot",	ExprBinaryFunc,	(ClientData) hypot	},
+    { "int",	ExprIntFunc,	NULL			},
+    { "isqrt",	ExprIsqrtFunc,	NULL			},
+    { "log",	ExprUnaryFunc,	(ClientData) log	},
+    { "log10",	ExprUnaryFunc,	(ClientData) log10	},
+    { "pow",	ExprBinaryFunc,	(ClientData) pow	},
+    { "rand",	ExprRandFunc,	NULL			},
+    { "round",	ExprRoundFunc,	NULL			},
+    { "sin",	ExprUnaryFunc,	(ClientData) sin	},
+    { "sinh",	ExprUnaryFunc,	(ClientData) sinh	},
+    { "sqrt",	ExprSqrtFunc,	NULL			},
+    { "srand",	ExprSrandFunc,	NULL			},
+    { "tan",	ExprUnaryFunc,	(ClientData) tan	},
+    { "tanh",	ExprUnaryFunc,	(ClientData) tanh	},
+    { "wide",	ExprWideFunc,	NULL			},
+    { NULL, NULL, NULL }
+};
+
+/*
+ * TIP#174's math operators. All are safe.
+ */
+
+typedef struct {
+    const char *name;		/* Name of object-based command. */
+    Tcl_ObjCmdProc *objProc;	/* Object-based function for command. */
+    CompileProc *compileProc;	/* Function called to compile command. */
+    union {
+	int numArgs;
+	int identity;
+    } i;
+    const char *expected;	/* For error message, what argument(s)
+				 * were expected. */
+} OpCmdInfo;
+static const OpCmdInfo mathOpCmds[] = {
+    { "~",	TclSingleOpCmd,		TclCompileInvertOpCmd,
+		/* numArgs */ {1},	"integer"},
+    { "!",	TclSingleOpCmd,		TclCompileNotOpCmd,
+		/* numArgs */ {1},	"boolean"},
+    { "+",	TclVariadicOpCmd,	TclCompileAddOpCmd,
+		/* identity */ {0},	NULL},
+    { "*",	TclVariadicOpCmd,	TclCompileMulOpCmd,
+		/* identity */ {1},	NULL},
+    { "&",	TclVariadicOpCmd,	TclCompileAndOpCmd,
+		/* identity */ {-1},	NULL},
+    { "|",	TclVariadicOpCmd,	TclCompileOrOpCmd,
+		/* identity */ {0},	NULL},
+    { "^",	TclVariadicOpCmd,	TclCompileXorOpCmd,
+		/* identity */ {0},	NULL},
+    { "**",	TclVariadicOpCmd,	TclCompilePowOpCmd,
+		/* identity */ {1},	NULL},
+    { "<<",	TclSingleOpCmd,		TclCompileLshiftOpCmd,
+		/* numArgs */ {2},	"integer shift"},
+    { ">>",	TclSingleOpCmd,		TclCompileRshiftOpCmd,
+		/* numArgs */ {2},	"integer shift"},
+    { "%",	TclSingleOpCmd,		TclCompileModOpCmd,
+		/* numArgs */ {2},	"integer integer"},
+    { "!=",	TclSingleOpCmd,		TclCompileNeqOpCmd,
+		/* numArgs */ {2},	"value value"},
+    { "ne",	TclSingleOpCmd,		TclCompileStrneqOpCmd,
+		/* numArgs */ {2},	"value value"},
+    { "in",	TclSingleOpCmd,		TclCompileInOpCmd,
+		/* numArgs */ {2},	"value list"},
+    { "ni",	TclSingleOpCmd,		TclCompileNiOpCmd,
+		/* numArgs */ {2},	"value list"},
+    { "-",	TclNoIdentOpCmd,	TclCompileMinusOpCmd,
+		/* unused */ {0},	"value ?value ...?"},
+    { "/",	TclNoIdentOpCmd,	TclCompileDivOpCmd,
+		/* unused */ {0},	"value ?value ...?"},
+    { "<",	TclSortingOpCmd,	TclCompileLessOpCmd,
+		/* unused */ {0},	NULL},
+    { "<=",	TclSortingOpCmd,	TclCompileLeqOpCmd,
+		/* unused */ {0},	NULL},
+    { ">",	TclSortingOpCmd,	TclCompileGreaterOpCmd,
+		/* unused */ {0},	NULL},
+    { ">=",	TclSortingOpCmd,	TclCompileGeqOpCmd,
+		/* unused */ {0},	NULL},
+    { "==",	TclSortingOpCmd,	TclCompileEqOpCmd,
+		/* unused */ {0},	NULL},
+    { "eq",	TclSortingOpCmd,	TclCompileStreqOpCmd,
+		/* unused */ {0},	NULL},
+    { NULL,	NULL,			NULL,
+		{0},			NULL}
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeEvaluation --
+ *
+ *	Finalizes the script cancellation hash table.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeEvaluation(void)
+{
+    Tcl_MutexLock(&cancelLock);
+    if (cancelTableInitialized == 1) {
+	Tcl_DeleteHashTable(&cancelTable);
+	cancelTableInitialized = 0;
+    }
+    Tcl_MutexUnlock(&cancelLock);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateInterp --
+ *
+ *	Create a new TCL command interpreter.
+ *
+ * Results:
+ *	The return value is a token for the interpreter, which may be used in
+ *	calls to functions like Tcl_CreateCmd, Tcl_Eval, or Tcl_DeleteInterp.
+ *
+ * Side effects:
+ *	The command interpreter is initialized with the built-in commands and
+ *	with the variables documented in tclvars(n).
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Interp *
+Tcl_CreateInterp(void)
+{
+    Interp *iPtr;
+    Tcl_Interp *interp;
+    Command *cmdPtr;
+    const BuiltinFuncDef *builtinFuncPtr;
+    const OpCmdInfo *opcmdInfoPtr;
+    const CmdInfo *cmdInfoPtr;
+    Tcl_Namespace *mathfuncNSPtr, *mathopNSPtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+    CancelInfo *cancelInfo;
+    union {
+	char c[sizeof(short)];
+	short s;
+    } order;
+#ifdef TCL_COMPILE_STATS
+    ByteCodeStats *statsPtr;
+#endif /* TCL_COMPILE_STATS */
+    char mathFuncName[32];
+    CallFrame *framePtr;
+
+    TclInitSubsystems();
+
+    /*
+     * Panic if someone updated the CallFrame structure without also updating
+     * the Tcl_CallFrame structure (or vice versa).
+     */
+
+    if (sizeof(Tcl_CallFrame) < sizeof(CallFrame)) {
+	/*NOTREACHED*/
+	Tcl_Panic("Tcl_CallFrame must not be smaller than CallFrame");
+    }
+
+#if defined(_WIN32) && !defined(_WIN64)
+    if (sizeof(time_t) != 4) {
+	/*NOTREACHED*/
+	Tcl_Panic("<time.h> is not compatible with MSVC");
+    }
+    if ((TclOffset(Tcl_StatBuf,st_atime) != 32)
+	    || (TclOffset(Tcl_StatBuf,st_ctime) != 40)) {
+	/*NOTREACHED*/
+	Tcl_Panic("<sys/stat.h> is not compatible with MSVC");
+    }
+#endif
+
+    if (cancelTableInitialized == 0) {
+	Tcl_MutexLock(&cancelLock);
+	if (cancelTableInitialized == 0) {
+	    Tcl_InitHashTable(&cancelTable, TCL_ONE_WORD_KEYS);
+	    cancelTableInitialized = 1;
+	}
+	Tcl_MutexUnlock(&cancelLock);
+    }
+
+    /*
+     * Initialize support for namespaces and create the global namespace
+     * (whose name is ""; an alias is "::"). This also initializes the Tcl
+     * object type table and other object management code.
+     */
+
+    iPtr = ckalloc(sizeof(Interp));
+    interp = (Tcl_Interp *) iPtr;
+
+#ifdef TCL_NO_DEPRECATED
+    iPtr->result = &tclEmptyString;
+#else
+    iPtr->result = iPtr->resultSpace;
+#endif
+    iPtr->freeProc = NULL;
+    iPtr->errorLine = 0;
+    iPtr->objResultPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(iPtr->objResultPtr);
+    iPtr->handle = TclHandleCreate(iPtr);
+    iPtr->globalNsPtr = NULL;
+    iPtr->hiddenCmdTablePtr = NULL;
+    iPtr->interpInfo = NULL;
+
+    TCL_CT_ASSERT(sizeof(iPtr->extra) <= sizeof(Tcl_HashTable));
+    iPtr->extra.optimizer = TclOptimizeBytecode;
+
+    iPtr->numLevels = 0;
+    iPtr->maxNestingDepth = MAX_NESTING_DEPTH;
+    iPtr->framePtr = NULL;	/* Initialise as soon as :: is available */
+    iPtr->varFramePtr = NULL;	/* Initialise as soon as :: is available */
+
+    /*
+     * TIP #280 - Initialize the arrays used to extend the ByteCode and Proc
+     * structures.
+     */
+
+    iPtr->cmdFramePtr = NULL;
+    iPtr->linePBodyPtr = ckalloc(sizeof(Tcl_HashTable));
+    iPtr->lineBCPtr = ckalloc(sizeof(Tcl_HashTable));
+    iPtr->lineLAPtr = ckalloc(sizeof(Tcl_HashTable));
+    iPtr->lineLABCPtr = ckalloc(sizeof(Tcl_HashTable));
+    Tcl_InitHashTable(iPtr->linePBodyPtr, TCL_ONE_WORD_KEYS);
+    Tcl_InitHashTable(iPtr->lineBCPtr, TCL_ONE_WORD_KEYS);
+    Tcl_InitHashTable(iPtr->lineLAPtr, TCL_ONE_WORD_KEYS);
+    Tcl_InitHashTable(iPtr->lineLABCPtr, TCL_ONE_WORD_KEYS);
+    iPtr->scriptCLLocPtr = NULL;
+
+    iPtr->activeVarTracePtr = NULL;
+
+    iPtr->returnOpts = NULL;
+    iPtr->errorInfo = NULL;
+    TclNewLiteralStringObj(iPtr->eiVar, "::errorInfo");
+    Tcl_IncrRefCount(iPtr->eiVar);
+    iPtr->errorStack = Tcl_NewListObj(0, NULL);
+    Tcl_IncrRefCount(iPtr->errorStack);
+    iPtr->resetErrorStack = 1;
+    TclNewLiteralStringObj(iPtr->upLiteral,"UP");
+    Tcl_IncrRefCount(iPtr->upLiteral);
+    TclNewLiteralStringObj(iPtr->callLiteral,"CALL");
+    Tcl_IncrRefCount(iPtr->callLiteral);
+    TclNewLiteralStringObj(iPtr->innerLiteral,"INNER");
+    Tcl_IncrRefCount(iPtr->innerLiteral);
+    iPtr->innerContext = Tcl_NewListObj(0, NULL);
+    Tcl_IncrRefCount(iPtr->innerContext);
+    iPtr->errorCode = NULL;
+    TclNewLiteralStringObj(iPtr->ecVar, "::errorCode");
+    Tcl_IncrRefCount(iPtr->ecVar);
+    iPtr->returnLevel = 1;
+    iPtr->returnCode = TCL_OK;
+
+    iPtr->rootFramePtr = NULL;	/* Initialise as soon as :: is available */
+    iPtr->lookupNsPtr = NULL;
+
+#ifndef TCL_NO_DEPRECATED
+    iPtr->appendResult = NULL;
+    iPtr->appendAvl = 0;
+    iPtr->appendUsed = 0;
+#endif
+
+    Tcl_InitHashTable(&iPtr->packageTable, TCL_STRING_KEYS);
+    iPtr->packageUnknown = NULL;
+
+    /* TIP #268 */
+#if (TCL_RELEASE_LEVEL == TCL_FINAL_RELEASE)
+    if (getenv("TCL_PKG_PREFER_LATEST") == NULL) {
+	iPtr->packagePrefer = PKG_PREFER_STABLE;
+    } else
+#endif
+	iPtr->packagePrefer = PKG_PREFER_LATEST;
+
+    iPtr->cmdCount = 0;
+    TclInitLiteralTable(&iPtr->literalTable);
+    iPtr->compileEpoch = 1;
+    iPtr->compiledProcPtr = NULL;
+    iPtr->resolverPtr = NULL;
+    iPtr->evalFlags = 0;
+    iPtr->scriptFile = NULL;
+    iPtr->flags = 0;
+    iPtr->tracePtr = NULL;
+    iPtr->tracesForbiddingInline = 0;
+    iPtr->activeCmdTracePtr = NULL;
+    iPtr->activeInterpTracePtr = NULL;
+    iPtr->assocData = NULL;
+    iPtr->execEnvPtr = NULL;	/* Set after namespaces initialized. */
+    iPtr->emptyObjPtr = Tcl_NewObj();
+				/* Another empty object. */
+    Tcl_IncrRefCount(iPtr->emptyObjPtr);
+#ifndef TCL_NO_DEPRECATED
+    iPtr->resultSpace[0] = 0;
+#endif
+    iPtr->threadId = Tcl_GetCurrentThread();
+
+    /* TIP #378 */
+#ifdef TCL_INTERP_DEBUG_FRAME
+    iPtr->flags |= INTERP_DEBUG_FRAME;
+#else
+    if (getenv("TCL_INTERP_DEBUG_FRAME") != NULL) {
+        iPtr->flags |= INTERP_DEBUG_FRAME;
+    }
+#endif
+
+    /*
+     * Initialise the tables for variable traces and searches *before*
+     * creating the global ns - so that the trace on errorInfo can be
+     * recorded.
+     */
+
+    Tcl_InitHashTable(&iPtr->varTraces, TCL_ONE_WORD_KEYS);
+    Tcl_InitHashTable(&iPtr->varSearches, TCL_ONE_WORD_KEYS);
+
+    iPtr->globalNsPtr = NULL;	/* Force creation of global ns below. */
+    iPtr->globalNsPtr = (Namespace *) Tcl_CreateNamespace(interp, "",
+	    NULL, NULL);
+    if (iPtr->globalNsPtr == NULL) {
+	Tcl_Panic("Tcl_CreateInterp: can't create global namespace");
+    }
+
+    /*
+     * Initialise the rootCallframe. It cannot be allocated on the stack, as
+     * it has to be in place before TclCreateExecEnv tries to use a variable.
+     */
+
+    /* This is needed to satisfy GCC 3.3's strict aliasing rules */
+    framePtr = ckalloc(sizeof(CallFrame));
+    (void) Tcl_PushCallFrame(interp, (Tcl_CallFrame *) framePtr,
+	    (Tcl_Namespace *) iPtr->globalNsPtr, /*isProcCallFrame*/ 0);
+    framePtr->objc = 0;
+
+    iPtr->framePtr = framePtr;
+    iPtr->varFramePtr = framePtr;
+    iPtr->rootFramePtr = framePtr;
+
+    /*
+     * Initialize support for code compilation and execution. We call
+     * TclCreateExecEnv after initializing namespaces since it tries to
+     * reference a Tcl variable (it links to the Tcl "tcl_traceExec"
+     * variable).
+     */
+
+    iPtr->execEnvPtr = TclCreateExecEnv(interp, INTERP_STACK_INITIAL_SIZE);
+
+    /*
+     * TIP #219, Tcl Channel Reflection API support.
+     */
+
+    iPtr->chanMsg = NULL;
+
+    /*
+     * TIP #285, Script cancellation support.
+     */
+
+    iPtr->asyncCancelMsg = Tcl_NewObj();
+
+    cancelInfo = ckalloc(sizeof(CancelInfo));
+    cancelInfo->interp = interp;
+
+    iPtr->asyncCancel = Tcl_AsyncCreate(CancelEvalProc, cancelInfo);
+    cancelInfo->async = iPtr->asyncCancel;
+    cancelInfo->result = NULL;
+    cancelInfo->length = 0;
+
+    Tcl_MutexLock(&cancelLock);
+    hPtr = Tcl_CreateHashEntry(&cancelTable, iPtr, &isNew);
+    Tcl_SetHashValue(hPtr, cancelInfo);
+    Tcl_MutexUnlock(&cancelLock);
+
+    /*
+     * Initialize the compilation and execution statistics kept for this
+     * interpreter.
+     */
+
+#ifdef TCL_COMPILE_STATS
+    statsPtr = &iPtr->stats;
+    statsPtr->numExecutions = 0;
+    statsPtr->numCompilations = 0;
+    statsPtr->numByteCodesFreed = 0;
+    memset(statsPtr->instructionCount, 0,
+	    sizeof(statsPtr->instructionCount));
+
+    statsPtr->totalSrcBytes = 0.0;
+    statsPtr->totalByteCodeBytes = 0.0;
+    statsPtr->currentSrcBytes = 0.0;
+    statsPtr->currentByteCodeBytes = 0.0;
+    memset(statsPtr->srcCount, 0, sizeof(statsPtr->srcCount));
+    memset(statsPtr->byteCodeCount, 0, sizeof(statsPtr->byteCodeCount));
+    memset(statsPtr->lifetimeCount, 0, sizeof(statsPtr->lifetimeCount));
+
+    statsPtr->currentInstBytes = 0.0;
+    statsPtr->currentLitBytes = 0.0;
+    statsPtr->currentExceptBytes = 0.0;
+    statsPtr->currentAuxBytes = 0.0;
+    statsPtr->currentCmdMapBytes = 0.0;
+
+    statsPtr->numLiteralsCreated = 0;
+    statsPtr->totalLitStringBytes = 0.0;
+    statsPtr->currentLitStringBytes = 0.0;
+    memset(statsPtr->literalCount, 0, sizeof(statsPtr->literalCount));
+#endif /* TCL_COMPILE_STATS */
+
+    /*
+     * Initialise the stub table pointer.
+     */
+
+    iPtr->stubTable = &tclStubs;
+
+    /*
+     * Initialize the ensemble error message rewriting support.
+     */
+
+    TclResetRewriteEnsemble(interp, 1);
+
+    /*
+     * TIP#143: Initialise the resource limit support.
+     */
+
+    TclInitLimitSupport(interp);
+
+    /*
+     * Initialise the thread-specific data ekeko. Note that the thread's alloc
+     * cache was already initialised by the call to alloc the interp struct.
+     */
+
+#if defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+    iPtr->allocCache = TclpGetAllocCache();
+#else
+    iPtr->allocCache = NULL;
+#endif
+    iPtr->pendingObjDataPtr = NULL;
+    iPtr->asyncReadyPtr = TclGetAsyncReadyPtr();
+    iPtr->deferredCallbacks = NULL;
+
+    /*
+     * Create the core commands. Do it here, rather than calling
+     * Tcl_CreateCommand, because it's faster (there's no need to check for a
+     * pre-existing command by the same name). If a command has a Tcl_CmdProc
+     * but no Tcl_ObjCmdProc, set the Tcl_ObjCmdProc to
+     * TclInvokeStringCommand. This is an object-based wrapper function that
+     * extracts strings, calls the string function, and creates an object for
+     * the result. Similarly, if a command has a Tcl_ObjCmdProc but no
+     * Tcl_CmdProc, set the Tcl_CmdProc to TclInvokeObjectCommand.
+     */
+
+    for (cmdInfoPtr = builtInCmds; cmdInfoPtr->name != NULL; cmdInfoPtr++) {
+	if ((cmdInfoPtr->objProc == NULL)
+		&& (cmdInfoPtr->compileProc == NULL)
+		&& (cmdInfoPtr->nreProc == NULL)) {
+	    Tcl_Panic("builtin command with NULL object command proc and a NULL compile proc");
+	}
+
+	hPtr = Tcl_CreateHashEntry(&iPtr->globalNsPtr->cmdTable,
+		cmdInfoPtr->name, &isNew);
+	if (isNew) {
+	    cmdPtr = ckalloc(sizeof(Command));
+	    cmdPtr->hPtr = hPtr;
+	    cmdPtr->nsPtr = iPtr->globalNsPtr;
+	    cmdPtr->refCount = 1;
+	    cmdPtr->cmdEpoch = 0;
+	    cmdPtr->compileProc = cmdInfoPtr->compileProc;
+	    cmdPtr->proc = TclInvokeObjectCommand;
+	    cmdPtr->clientData = cmdPtr;
+	    cmdPtr->objProc = cmdInfoPtr->objProc;
+	    cmdPtr->objClientData = NULL;
+	    cmdPtr->deleteProc = NULL;
+	    cmdPtr->deleteData = NULL;
+	    cmdPtr->flags = 0;
+            if (cmdInfoPtr->flags & CMD_COMPILES_EXPANDED) {
+                cmdPtr->flags |= CMD_COMPILES_EXPANDED;
+            }
+	    cmdPtr->importRefPtr = NULL;
+	    cmdPtr->tracePtr = NULL;
+	    cmdPtr->nreProc = cmdInfoPtr->nreProc;
+	    Tcl_SetHashValue(hPtr, cmdPtr);
+	}
+    }
+
+    /*
+     * Create the "array", "binary", "chan", "clock", "dict", "encoding",
+     * "file", "info", "namespace" and "string" ensembles. Note that all these
+     * commands (and their subcommands that are not present in the global
+     * namespace) are wholly safe *except* for "clock", "encoding" and "file".
+     */
+
+    TclInitArrayCmd(interp);
+    TclInitBinaryCmd(interp);
+    TclInitChanCmd(interp);
+    TclInitDictCmd(interp);
+    TclInitEncodingCmd(interp);
+    TclInitFileCmd(interp);
+    TclInitInfoCmd(interp);
+    TclInitNamespaceCmd(interp);
+    TclInitStringCmd(interp);
+    TclInitPrefixCmd(interp);
+
+    /*
+     * Register "clock" subcommands. These *do* go through
+     * Tcl_CreateObjCommand, since they aren't in the global namespace and
+     * involve ensembles.
+     */
+
+    TclClockInit(interp);
+
+    /*
+     * Register the built-in functions. This is empty now that they are
+     * implemented as commands in the ::tcl::mathfunc namespace.
+     */
+
+    /*
+     * Register the default [interp bgerror] handler.
+     */
+
+    Tcl_CreateObjCommand(interp, "::tcl::Bgerror",
+	    TclDefaultBgErrorHandlerObjCmd, NULL, NULL);
+
+    /*
+     * Create unsupported commands for debugging bytecode and objects.
+     */
+
+    Tcl_CreateObjCommand(interp, "::tcl::unsupported::disassemble",
+	    Tcl_DisassembleObjCmd, INT2PTR(0), NULL);
+    Tcl_CreateObjCommand(interp, "::tcl::unsupported::getbytecode",
+	    Tcl_DisassembleObjCmd, INT2PTR(1), NULL);
+    Tcl_CreateObjCommand(interp, "::tcl::unsupported::representation",
+	    Tcl_RepresentationCmd, NULL, NULL);
+
+    /* Adding the bytecode assembler command */
+    cmdPtr = (Command *) Tcl_NRCreateCommand(interp,
+            "::tcl::unsupported::assemble", Tcl_AssembleObjCmd,
+            TclNRAssembleObjCmd, NULL, NULL);
+    cmdPtr->compileProc = &TclCompileAssembleCmd;
+
+    Tcl_NRCreateCommand(interp, "::tcl::unsupported::inject", NULL,
+	    NRCoroInjectObjCmd, NULL, NULL);
+
+#ifdef USE_DTRACE
+    /*
+     * Register the tcl::dtrace command.
+     */
+
+    Tcl_CreateObjCommand(interp, "::tcl::dtrace", DTraceObjCmd, NULL, NULL);
+#endif /* USE_DTRACE */
+
+    /*
+     * Register the builtin math functions.
+     */
+
+    mathfuncNSPtr = Tcl_CreateNamespace(interp, "::tcl::mathfunc", NULL,NULL);
+    if (mathfuncNSPtr == NULL) {
+	Tcl_Panic("Can't create math function namespace");
+    }
+#define MATH_FUNC_PREFIX_LEN 17 /* == strlen("::tcl::mathfunc::") */
+    memcpy(mathFuncName, "::tcl::mathfunc::", MATH_FUNC_PREFIX_LEN);
+    for (builtinFuncPtr = BuiltinFuncTable; builtinFuncPtr->name != NULL;
+	    builtinFuncPtr++) {
+	strcpy(mathFuncName+MATH_FUNC_PREFIX_LEN, builtinFuncPtr->name);
+	Tcl_CreateObjCommand(interp, mathFuncName,
+		builtinFuncPtr->objCmdProc, builtinFuncPtr->clientData, NULL);
+	Tcl_Export(interp, mathfuncNSPtr, builtinFuncPtr->name, 0);
+    }
+
+    /*
+     * Register the mathematical "operator" commands. [TIP #174]
+     */
+
+    mathopNSPtr = Tcl_CreateNamespace(interp, "::tcl::mathop", NULL, NULL);
+    if (mathopNSPtr == NULL) {
+	Tcl_Panic("can't create math operator namespace");
+    }
+    Tcl_Export(interp, mathopNSPtr, "*", 1);
+#define MATH_OP_PREFIX_LEN 15 /* == strlen("::tcl::mathop::") */
+    memcpy(mathFuncName, "::tcl::mathop::", MATH_OP_PREFIX_LEN);
+    for (opcmdInfoPtr=mathOpCmds ; opcmdInfoPtr->name!=NULL ; opcmdInfoPtr++){
+	TclOpCmdClientData *occdPtr = ckalloc(sizeof(TclOpCmdClientData));
+
+	occdPtr->op = opcmdInfoPtr->name;
+	occdPtr->i.numArgs = opcmdInfoPtr->i.numArgs;
+	occdPtr->expected = opcmdInfoPtr->expected;
+	strcpy(mathFuncName + MATH_OP_PREFIX_LEN, opcmdInfoPtr->name);
+	cmdPtr = (Command *) Tcl_CreateObjCommand(interp, mathFuncName,
+		opcmdInfoPtr->objProc, occdPtr, DeleteOpCmdClientData);
+	if (cmdPtr == NULL) {
+	    Tcl_Panic("failed to create math operator %s",
+		    opcmdInfoPtr->name);
+	} else if (opcmdInfoPtr->compileProc != NULL) {
+	    cmdPtr->compileProc = opcmdInfoPtr->compileProc;
+	}
+    }
+
+    /*
+     * Do Multiple/Safe Interps Tcl init stuff
+     */
+
+    TclInterpInit(interp);
+    TclSetupEnv(interp);
+
+    /*
+     * TIP #59: Make embedded configuration information available.
+     */
+
+    TclInitEmbeddedConfigurationInformation(interp);
+
+    /*
+     * TIP #440: Declare the name of the script engine to be "Tcl".
+     */
+
+    Tcl_SetVar2(interp, "tcl_platform", "engine", "Tcl",
+	    TCL_GLOBAL_ONLY);
+
+    /*
+     * Compute the byte order of this machine.
+     */
+
+    order.s = 1;
+    Tcl_SetVar2(interp, "tcl_platform", "byteOrder",
+	    ((order.c[0] == 1) ? "littleEndian" : "bigEndian"),
+	    TCL_GLOBAL_ONLY);
+
+    Tcl_SetVar2Ex(interp, "tcl_platform", "wordSize",
+	    Tcl_NewLongObj((long) sizeof(long)), TCL_GLOBAL_ONLY);
+
+    /* TIP #291 */
+    Tcl_SetVar2Ex(interp, "tcl_platform", "pointerSize",
+	    Tcl_NewLongObj((long) sizeof(void *)), TCL_GLOBAL_ONLY);
+
+    /*
+     * Set up other variables such as tcl_version and tcl_library
+     */
+
+    Tcl_SetVar2(interp, "tcl_patchLevel", NULL, TCL_PATCH_LEVEL, TCL_GLOBAL_ONLY);
+    Tcl_SetVar2(interp, "tcl_version", NULL, TCL_VERSION, TCL_GLOBAL_ONLY);
+    Tcl_TraceVar2(interp, "tcl_precision", NULL,
+	    TCL_GLOBAL_ONLY|TCL_TRACE_READS|TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    TclPrecTraceProc, NULL);
+    TclpSetVariables(interp);
+
+#ifdef TCL_THREADS
+    /*
+     * The existence of the "threaded" element of the tcl_platform array
+     * indicates that this particular Tcl shell has been compiled with threads
+     * turned on. Using "info exists tcl_platform(threaded)" a Tcl script can
+     * introspect on the interpreter level of thread safety.
+     */
+
+    Tcl_SetVar2(interp, "tcl_platform", "threaded", "1", TCL_GLOBAL_ONLY);
+#endif
+
+    /*
+     * Register Tcl's version number.
+     * TIP #268: Full patchlevel instead of just major.minor
+     */
+
+    Tcl_PkgProvideEx(interp, "Tcl", TCL_PATCH_LEVEL, &tclStubs);
+
+    if (TclTommath_Init(interp) != TCL_OK) {
+	Tcl_Panic("%s", TclGetString(Tcl_GetObjResult(interp)));
+    }
+
+    if (TclOOInit(interp) != TCL_OK) {
+	Tcl_Panic("%s", TclGetString(Tcl_GetObjResult(interp)));
+    }
+
+    /*
+     * Only build in zlib support if we've successfully detected a library to
+     * compile and link against.
+     */
+
+#ifdef HAVE_ZLIB
+    if (TclZlibInit(interp) != TCL_OK) {
+	Tcl_Panic("%s", TclGetString(Tcl_GetObjResult(interp)));
+    }
+#endif
+
+    TOP_CB(iPtr) = NULL;
+    return interp;
+}
+
+static void
+DeleteOpCmdClientData(
+    ClientData clientData)
+{
+    TclOpCmdClientData *occdPtr = clientData;
+
+    ckfree(occdPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclHideUnsafeCommands --
+ *
+ *	Hides base commands that are not marked as safe from this interpreter.
+ *
+ * Results:
+ *	TCL_OK if it succeeds, TCL_ERROR else.
+ *
+ * Side effects:
+ *	Hides functionality in an interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclHideUnsafeCommands(
+    Tcl_Interp *interp)		/* Hide commands in this interpreter. */
+{
+    register const CmdInfo *cmdInfoPtr;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+    for (cmdInfoPtr = builtInCmds; cmdInfoPtr->name != NULL; cmdInfoPtr++) {
+	if (!(cmdInfoPtr->flags & CMD_IS_SAFE)) {
+	    Tcl_HideCommand(interp, cmdInfoPtr->name, cmdInfoPtr->name);
+	}
+    }
+    TclMakeEncodingCommandSafe(interp); /* Ugh! */
+    TclMakeFileCommandSafe(interp);     /* Ugh! */
+    return TCL_OK;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_CallWhenDeleted --
+ *
+ *	Arrange for a function to be called before a given interpreter is
+ *	deleted. The function is called as soon as Tcl_DeleteInterp is called;
+ *	if Tcl_CallWhenDeleted is called on an interpreter that has already
+ *	been deleted, the function will be called when the last Tcl_Release is
+ *	done on the interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	When Tcl_DeleteInterp is invoked to delete interp, proc will be
+ *	invoked. See the manual entry for details.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+Tcl_CallWhenDeleted(
+    Tcl_Interp *interp,		/* Interpreter to watch. */
+    Tcl_InterpDeleteProc *proc,	/* Function to call when interpreter is about
+				 * to be deleted. */
+    ClientData clientData)	/* One-word value to pass to proc. */
+{
+    Interp *iPtr = (Interp *) interp;
+    static Tcl_ThreadDataKey assocDataCounterKey;
+    int *assocDataCounterPtr =
+	    Tcl_GetThreadData(&assocDataCounterKey, sizeof(int));
+    int isNew;
+    char buffer[32 + TCL_INTEGER_SPACE];
+    AssocData *dPtr = ckalloc(sizeof(AssocData));
+    Tcl_HashEntry *hPtr;
+
+    sprintf(buffer, "Assoc Data Key #%d", *assocDataCounterPtr);
+    (*assocDataCounterPtr)++;
+
+    if (iPtr->assocData == NULL) {
+	iPtr->assocData = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(iPtr->assocData, TCL_STRING_KEYS);
+    }
+    hPtr = Tcl_CreateHashEntry(iPtr->assocData, buffer, &isNew);
+    dPtr->proc = proc;
+    dPtr->clientData = clientData;
+    Tcl_SetHashValue(hPtr, dPtr);
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_DontCallWhenDeleted --
+ *
+ *	Cancel the arrangement for a function to be called when a given
+ *	interpreter is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If proc and clientData were previously registered as a callback via
+ *	Tcl_CallWhenDeleted, they are unregistered. If they weren't previously
+ *	registered then nothing happens.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+Tcl_DontCallWhenDeleted(
+    Tcl_Interp *interp,		/* Interpreter to watch. */
+    Tcl_InterpDeleteProc *proc,	/* Function to call when interpreter is about
+				 * to be deleted. */
+    ClientData clientData)	/* One-word value to pass to proc. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashTable *hTablePtr;
+    Tcl_HashSearch hSearch;
+    Tcl_HashEntry *hPtr;
+    AssocData *dPtr;
+
+    hTablePtr = iPtr->assocData;
+    if (hTablePtr == NULL) {
+	return;
+    }
+    for (hPtr = Tcl_FirstHashEntry(hTablePtr, &hSearch); hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&hSearch)) {
+	dPtr = Tcl_GetHashValue(hPtr);
+	if ((dPtr->proc == proc) && (dPtr->clientData == clientData)) {
+	    ckfree(dPtr);
+	    Tcl_DeleteHashEntry(hPtr);
+	    return;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetAssocData --
+ *
+ *	Creates a named association between user-specified data, a delete
+ *	function and this interpreter. If the association already exists the
+ *	data is overwritten with the new data. The delete function will be
+ *	invoked when the interpreter is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Sets the associated data, creates the association if needed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetAssocData(
+    Tcl_Interp *interp,		/* Interpreter to associate with. */
+    const char *name,		/* Name for association. */
+    Tcl_InterpDeleteProc *proc,	/* Proc to call when interpreter is about to
+				 * be deleted. */
+    ClientData clientData)	/* One-word value to pass to proc. */
+{
+    Interp *iPtr = (Interp *) interp;
+    AssocData *dPtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (iPtr->assocData == NULL) {
+	iPtr->assocData = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(iPtr->assocData, TCL_STRING_KEYS);
+    }
+    hPtr = Tcl_CreateHashEntry(iPtr->assocData, name, &isNew);
+    if (isNew == 0) {
+	dPtr = Tcl_GetHashValue(hPtr);
+    } else {
+	dPtr = ckalloc(sizeof(AssocData));
+    }
+    dPtr->proc = proc;
+    dPtr->clientData = clientData;
+
+    Tcl_SetHashValue(hPtr, dPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteAssocData --
+ *
+ *	Deletes a named association of user-specified data with the specified
+ *	interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the association.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteAssocData(
+    Tcl_Interp *interp,		/* Interpreter to associate with. */
+    const char *name)		/* Name of association. */
+{
+    Interp *iPtr = (Interp *) interp;
+    AssocData *dPtr;
+    Tcl_HashEntry *hPtr;
+
+    if (iPtr->assocData == NULL) {
+	return;
+    }
+    hPtr = Tcl_FindHashEntry(iPtr->assocData, name);
+    if (hPtr == NULL) {
+	return;
+    }
+    dPtr = Tcl_GetHashValue(hPtr);
+    if (dPtr->proc != NULL) {
+	dPtr->proc(dPtr->clientData, interp);
+    }
+    ckfree(dPtr);
+    Tcl_DeleteHashEntry(hPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetAssocData --
+ *
+ *	Returns the client data associated with this name in the specified
+ *	interpreter.
+ *
+ * Results:
+ *	The client data in the AssocData record denoted by the named
+ *	association, or NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_GetAssocData(
+    Tcl_Interp *interp,		/* Interpreter associated with. */
+    const char *name,		/* Name of association. */
+    Tcl_InterpDeleteProc **procPtr)
+				/* Pointer to place to store address of
+				 * current deletion callback. */
+{
+    Interp *iPtr = (Interp *) interp;
+    AssocData *dPtr;
+    Tcl_HashEntry *hPtr;
+
+    if (iPtr->assocData == NULL) {
+	return NULL;
+    }
+    hPtr = Tcl_FindHashEntry(iPtr->assocData, name);
+    if (hPtr == NULL) {
+	return NULL;
+    }
+    dPtr = Tcl_GetHashValue(hPtr);
+    if (procPtr != NULL) {
+	*procPtr = dPtr->proc;
+    }
+    return dPtr->clientData;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InterpDeleted --
+ *
+ *	Returns nonzero if the interpreter has been deleted with a call to
+ *	Tcl_DeleteInterp.
+ *
+ * Results:
+ *	Nonzero if the interpreter is deleted, zero otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_InterpDeleted(
+    Tcl_Interp *interp)
+{
+    return (((Interp *) interp)->flags & DELETED) ? 1 : 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteInterp --
+ *
+ *	Ensures that the interpreter will be deleted eventually. If there are
+ *	no Tcl_Preserve calls in effect for this interpreter, it is deleted
+ *	immediately, otherwise the interpreter is deleted when the last
+ *	Tcl_Preserve is matched by a call to Tcl_Release. In either case, the
+ *	function runs the currently registered deletion callbacks.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interpreter is marked as deleted. The caller may still use it
+ *	safely if there are calls to Tcl_Preserve in effect for the
+ *	interpreter, but further calls to Tcl_Eval etc in this interpreter
+ *	will fail.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteInterp(
+    Tcl_Interp *interp)		/* Token for command interpreter (returned by
+				 * a previous call to Tcl_CreateInterp). */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    /*
+     * If the interpreter has already been marked deleted, just punt.
+     */
+
+    if (iPtr->flags & DELETED) {
+	return;
+    }
+
+    /*
+     * Mark the interpreter as deleted. No further evals will be allowed.
+     * Increase the compileEpoch as a signal to compiled bytecodes.
+     */
+
+    iPtr->flags |= DELETED;
+    iPtr->compileEpoch++;
+
+    /*
+     * Ensure that the interpreter is eventually deleted.
+     */
+
+    Tcl_EventuallyFree(interp, (Tcl_FreeProc *) DeleteInterpProc);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteInterpProc --
+ *
+ *	Helper function to delete an interpreter. This function is called when
+ *	the last call to Tcl_Preserve on this interpreter is matched by a call
+ *	to Tcl_Release. The function cleans up all resources used in the
+ *	interpreter and calls all currently registered interpreter deletion
+ *	callbacks.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Whatever the interpreter deletion callbacks do. Frees resources used
+ *	by the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteInterpProc(
+    Tcl_Interp *interp)		/* Interpreter to delete. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+    Tcl_HashTable *hTablePtr;
+    ResolverScheme *resPtr, *nextResPtr;
+    int i;
+
+    /*
+     * Punt if there is an error in the Tcl_Release/Tcl_Preserve matchup,
+	 * unless we are exiting.
+     */
+
+    if ((iPtr->numLevels > 0) && !TclInExit()) {
+	Tcl_Panic("DeleteInterpProc called with active evals");
+    }
+
+    /*
+     * The interpreter should already be marked deleted; otherwise how did we
+     * get here?
+     */
+
+    if (!(iPtr->flags & DELETED)) {
+	Tcl_Panic("DeleteInterpProc called on interpreter not marked deleted");
+    }
+
+    /*
+     * TIP #219, Tcl Channel Reflection API. Discard a leftover state.
+     */
+
+    if (iPtr->chanMsg != NULL) {
+	Tcl_DecrRefCount(iPtr->chanMsg);
+	iPtr->chanMsg = NULL;
+    }
+
+    /*
+     * TIP #285, Script cancellation support. Delete this interp from the
+     * global hash table of CancelInfo structs.
+     */
+
+    Tcl_MutexLock(&cancelLock);
+    hPtr = Tcl_FindHashEntry(&cancelTable, (char *) iPtr);
+    if (hPtr != NULL) {
+	CancelInfo *cancelInfo = Tcl_GetHashValue(hPtr);
+
+	if (cancelInfo != NULL) {
+	    if (cancelInfo->result != NULL) {
+		ckfree(cancelInfo->result);
+	    }
+	    ckfree(cancelInfo);
+	}
+
+	Tcl_DeleteHashEntry(hPtr);
+    }
+
+    if (iPtr->asyncCancel != NULL) {
+	Tcl_AsyncDelete(iPtr->asyncCancel);
+	iPtr->asyncCancel = NULL;
+    }
+
+    if (iPtr->asyncCancelMsg != NULL) {
+	Tcl_DecrRefCount(iPtr->asyncCancelMsg);
+	iPtr->asyncCancelMsg = NULL;
+    }
+    Tcl_MutexUnlock(&cancelLock);
+
+    /*
+     * Shut down all limit handler callback scripts that call back into this
+     * interpreter. Then eliminate all limit handlers for this interpreter.
+     */
+
+    TclRemoveScriptLimitCallbacks(interp);
+    TclLimitRemoveAllHandlers(interp);
+
+    /*
+     * Dismantle the namespace here, before we clear the assocData. If any
+     * background errors occur here, they will be deleted below.
+     *
+     * Dismantle the namespace after freeing the iPtr->handle so that each
+     * bytecode releases its literals without caring to update the literal
+     * table, as it will be freed later in this function without further use.
+     */
+
+    TclHandleFree(iPtr->handle);
+    TclTeardownNamespace(iPtr->globalNsPtr);
+
+    /*
+     * Delete all the hidden commands.
+     */
+
+    hTablePtr = iPtr->hiddenCmdTablePtr;
+    if (hTablePtr != NULL) {
+	/*
+	 * Non-pernicious deletion. The deletion callbacks will not be allowed
+	 * to create any new hidden or non-hidden commands.
+	 * Tcl_DeleteCommandFromToken will remove the entry from the
+	 * hiddenCmdTablePtr.
+	 */
+
+	hPtr = Tcl_FirstHashEntry(hTablePtr, &search);
+	for (; hPtr != NULL; hPtr = Tcl_NextHashEntry(&search)) {
+	    Tcl_DeleteCommandFromToken(interp, Tcl_GetHashValue(hPtr));
+	}
+	Tcl_DeleteHashTable(hTablePtr);
+	ckfree(hTablePtr);
+    }
+
+    /*
+     * Invoke deletion callbacks; note that a callback can create new
+     * callbacks, so we iterate.
+     */
+
+    while (iPtr->assocData != NULL) {
+	AssocData *dPtr;
+
+	hTablePtr = iPtr->assocData;
+	iPtr->assocData = NULL;
+	for (hPtr = Tcl_FirstHashEntry(hTablePtr, &search);
+		hPtr != NULL;
+		hPtr = Tcl_FirstHashEntry(hTablePtr, &search)) {
+	    dPtr = Tcl_GetHashValue(hPtr);
+	    Tcl_DeleteHashEntry(hPtr);
+	    if (dPtr->proc != NULL) {
+		dPtr->proc(dPtr->clientData, interp);
+	    }
+	    ckfree(dPtr);
+	}
+	Tcl_DeleteHashTable(hTablePtr);
+	ckfree(hTablePtr);
+    }
+
+    /*
+     * Pop the root frame pointer and finish deleting the global
+     * namespace. The order is important [Bug 1658572].
+     */
+
+    if ((iPtr->framePtr != iPtr->rootFramePtr) && !TclInExit()) {
+	Tcl_Panic("DeleteInterpProc: popping rootCallFrame with other frames on top");
+    }
+    Tcl_PopCallFrame(interp);
+    ckfree(iPtr->rootFramePtr);
+    iPtr->rootFramePtr = NULL;
+    Tcl_DeleteNamespace((Tcl_Namespace *) iPtr->globalNsPtr);
+
+    /*
+     * Free up the result *after* deleting variables, since variable deletion
+     * could have transferred ownership of the result string to Tcl.
+     */
+
+    Tcl_FreeResult(interp);
+    iPtr->result = NULL;
+    Tcl_DecrRefCount(iPtr->objResultPtr);
+    iPtr->objResultPtr = NULL;
+    Tcl_DecrRefCount(iPtr->ecVar);
+    if (iPtr->errorCode) {
+	Tcl_DecrRefCount(iPtr->errorCode);
+	iPtr->errorCode = NULL;
+    }
+    Tcl_DecrRefCount(iPtr->eiVar);
+    if (iPtr->errorInfo) {
+	Tcl_DecrRefCount(iPtr->errorInfo);
+	iPtr->errorInfo = NULL;
+    }
+    Tcl_DecrRefCount(iPtr->errorStack);
+    iPtr->errorStack = NULL;
+    Tcl_DecrRefCount(iPtr->upLiteral);
+    Tcl_DecrRefCount(iPtr->callLiteral);
+    Tcl_DecrRefCount(iPtr->innerLiteral);
+    Tcl_DecrRefCount(iPtr->innerContext);
+    if (iPtr->returnOpts) {
+	Tcl_DecrRefCount(iPtr->returnOpts);
+    }
+#ifndef TCL_NO_DEPRECATED
+    if (iPtr->appendResult != NULL) {
+	ckfree(iPtr->appendResult);
+	iPtr->appendResult = NULL;
+    }
+#endif
+    TclFreePackageInfo(iPtr);
+    while (iPtr->tracePtr != NULL) {
+	Tcl_DeleteTrace((Tcl_Interp *) iPtr, (Tcl_Trace) iPtr->tracePtr);
+    }
+    if (iPtr->execEnvPtr != NULL) {
+	TclDeleteExecEnv(iPtr->execEnvPtr);
+    }
+    if (iPtr->scriptFile) {
+	Tcl_DecrRefCount(iPtr->scriptFile);
+	iPtr->scriptFile = NULL;
+    }
+    Tcl_DecrRefCount(iPtr->emptyObjPtr);
+    iPtr->emptyObjPtr = NULL;
+
+    resPtr = iPtr->resolverPtr;
+    while (resPtr) {
+	nextResPtr = resPtr->nextPtr;
+	ckfree(resPtr->name);
+	ckfree(resPtr);
+	resPtr = nextResPtr;
+    }
+
+    /*
+     * Free up literal objects created for scripts compiled by the
+     * interpreter.
+     */
+
+    TclDeleteLiteralTable(interp, &iPtr->literalTable);
+
+    /*
+     * TIP #280 - Release the arrays for ByteCode/Proc extension, and
+     * contents.
+     */
+
+    for (hPtr = Tcl_FirstHashEntry(iPtr->linePBodyPtr, &search);
+	    hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	CmdFrame *cfPtr = Tcl_GetHashValue(hPtr);
+	Proc *procPtr = (Proc *) Tcl_GetHashKey(iPtr->linePBodyPtr, hPtr);
+
+	procPtr->iPtr = NULL;
+	if (cfPtr) {
+	    if (cfPtr->type == TCL_LOCATION_SOURCE) {
+		Tcl_DecrRefCount(cfPtr->data.eval.path);
+	    }
+	    ckfree(cfPtr->line);
+	    ckfree(cfPtr);
+	}
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    Tcl_DeleteHashTable(iPtr->linePBodyPtr);
+    ckfree(iPtr->linePBodyPtr);
+    iPtr->linePBodyPtr = NULL;
+
+    /*
+     * See also tclCompile.c, TclCleanupByteCode
+     */
+
+    for (hPtr = Tcl_FirstHashEntry(iPtr->lineBCPtr, &search);
+	    hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	ExtCmdLoc *eclPtr = Tcl_GetHashValue(hPtr);
+
+	if (eclPtr->type == TCL_LOCATION_SOURCE) {
+	    Tcl_DecrRefCount(eclPtr->path);
+	}
+	for (i=0; i< eclPtr->nuloc; i++) {
+	    ckfree(eclPtr->loc[i].line);
+	}
+
+	if (eclPtr->loc != NULL) {
+	    ckfree(eclPtr->loc);
+	}
+
+	ckfree(eclPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    Tcl_DeleteHashTable(iPtr->lineBCPtr);
+    ckfree(iPtr->lineBCPtr);
+    iPtr->lineBCPtr = NULL;
+
+    /*
+     * Location stack for uplevel/eval/... scripts which were passed through
+     * proc arguments. Actually we track all arguments as we do not and cannot
+     * know which arguments will be used as scripts and which will not.
+     */
+
+    if (iPtr->lineLAPtr->numEntries && !TclInExit()) {
+	/*
+	 * When the interp goes away we have nothing on the stack, so there
+	 * are no arguments, so this table has to be empty.
+	 */
+
+	Tcl_Panic("Argument location tracking table not empty");
+    }
+
+    Tcl_DeleteHashTable(iPtr->lineLAPtr);
+    ckfree(iPtr->lineLAPtr);
+    iPtr->lineLAPtr = NULL;
+
+    if (iPtr->lineLABCPtr->numEntries && !TclInExit()) {
+	/*
+	 * When the interp goes away we have nothing on the stack, so there
+	 * are no arguments, so this table has to be empty.
+	 */
+
+	Tcl_Panic("Argument location tracking table not empty");
+    }
+
+    Tcl_DeleteHashTable(iPtr->lineLABCPtr);
+    ckfree(iPtr->lineLABCPtr);
+    iPtr->lineLABCPtr = NULL;
+
+    /*
+     * Squelch the tables of traces on variables and searches over arrays in
+     * the in the interpreter.
+     */
+
+    Tcl_DeleteHashTable(&iPtr->varTraces);
+    Tcl_DeleteHashTable(&iPtr->varSearches);
+
+    ckfree(iPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_HideCommand --
+ *
+ *	Makes a command hidden so that it cannot be invoked from within an
+ *	interpreter, only from within an ancestor.
+ *
+ * Results:
+ *	A standard Tcl result; also leaves a message in the interp's result if
+ *	an error occurs.
+ *
+ * Side effects:
+ *	Removes a command from the command table and create an entry into the
+ *	hidden command table under the specified token name.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_HideCommand(
+    Tcl_Interp *interp,		/* Interpreter in which to hide command. */
+    const char *cmdName,	/* Name of command to hide. */
+    const char *hiddenCmdToken)	/* Token name of the to-be-hidden command. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Command cmd;
+    Command *cmdPtr;
+    Tcl_HashTable *hiddenCmdTablePtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (iPtr->flags & DELETED) {
+	/*
+	 * The interpreter is being deleted. Do not create any new structures,
+	 * because it is not safe to modify the interpreter.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Disallow hiding of commands that are currently in a namespace or
+     * renaming (as part of hiding) into a namespace (because the current
+     * implementation with a single global table and the needed uniqueness of
+     * names cause problems with namespaces).
+     *
+     * We don't need to check for "::" in cmdName because the real check is on
+     * the nsPtr below.
+     *
+     * hiddenCmdToken is just a string which is not interpreted in any way. It
+     * may contain :: but the string is not interpreted as a namespace
+     * qualifier command name. Thus, hiding foo::bar to foo::bar and then
+     * trying to expose or invoke ::foo::bar will NOT work; but if the
+     * application always uses the same strings it will get consistent
+     * behaviour.
+     *
+     * But as we currently limit ourselves to the global namespace only for
+     * the source, in order to avoid potential confusion, lets prevent "::" in
+     * the token too. - dl
+     */
+
+    if (strstr(hiddenCmdToken, "::") != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"cannot use namespace qualifiers in hidden command"
+		" token (rename)", -1));
+        Tcl_SetErrorCode(interp, "TCL", "VALUE", "HIDDENTOKEN", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the command to hide. An error is returned if cmdName can't be
+     * found. Look up the command only from the global namespace. Full path of
+     * the command must be given if using namespaces.
+     */
+
+    cmd = Tcl_FindCommand(interp, cmdName, NULL,
+	    /*flags*/ TCL_LEAVE_ERR_MSG | TCL_GLOBAL_ONLY);
+    if (cmd == (Tcl_Command) NULL) {
+	return TCL_ERROR;
+    }
+    cmdPtr = (Command *) cmd;
+
+    /*
+     * Check that the command is really in global namespace
+     */
+
+    if (cmdPtr->nsPtr != iPtr->globalNsPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "can only hide global namespace commands (use rename then hide)",
+                -1));
+        Tcl_SetErrorCode(interp, "TCL", "HIDE", "NON_GLOBAL", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Initialize the hidden command table if necessary.
+     */
+
+    hiddenCmdTablePtr = iPtr->hiddenCmdTablePtr;
+    if (hiddenCmdTablePtr == NULL) {
+	hiddenCmdTablePtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(hiddenCmdTablePtr, TCL_STRING_KEYS);
+	iPtr->hiddenCmdTablePtr = hiddenCmdTablePtr;
+    }
+
+    /*
+     * It is an error to move an exposed command to a hidden command with
+     * hiddenCmdToken if a hidden command with the name hiddenCmdToken already
+     * exists.
+     */
+
+    hPtr = Tcl_CreateHashEntry(hiddenCmdTablePtr, hiddenCmdToken, &isNew);
+    if (!isNew) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "hidden command named \"%s\" already exists",
+                hiddenCmdToken));
+        Tcl_SetErrorCode(interp, "TCL", "HIDE", "ALREADY_HIDDEN", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * NB: This code is currently 'like' a rename to a specialy set apart name
+     * table. Changes here and in TclRenameCommand must be kept in synch until
+     * the common parts are actually factorized out.
+     */
+
+    /*
+     * Remove the hash entry for the command from the interpreter command
+     * table. This is like deleting the command, so bump its command epoch;
+     * this invalidates any cached references that point to the command.
+     */
+
+    if (cmdPtr->hPtr != NULL) {
+	Tcl_DeleteHashEntry(cmdPtr->hPtr);
+	cmdPtr->hPtr = NULL;
+	cmdPtr->cmdEpoch++;
+    }
+
+    /*
+     * The list of command exported from the namespace might have changed.
+     * However, we do not need to recompute this just yet; next time we need
+     * the info will be soon enough.
+     */
+
+    TclInvalidateNsCmdLookup(cmdPtr->nsPtr);
+
+    /*
+     * Now link the hash table entry with the command structure. We ensured
+     * above that the nsPtr was right.
+     */
+
+    cmdPtr->hPtr = hPtr;
+    Tcl_SetHashValue(hPtr, cmdPtr);
+
+    /*
+     * If the command being hidden has a compile function, increment the
+     * interpreter's compileEpoch to invalidate its compiled code. This makes
+     * sure that we don't later try to execute old code compiled with
+     * command-specific (i.e., inline) bytecodes for the now-hidden command.
+     * This field is checked in Tcl_EvalObj and ObjInterpProc, and code whose
+     * compilation epoch doesn't match is recompiled.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	iPtr->compileEpoch++;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ExposeCommand --
+ *
+ *	Makes a previously hidden command callable from inside the interpreter
+ *	instead of only by its ancestors.
+ *
+ * Results:
+ *	A standard Tcl result. If an error occurs, a message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	Moves commands from one hash table to another.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ExposeCommand(
+    Tcl_Interp *interp,		/* Interpreter in which to make command
+				 * callable. */
+    const char *hiddenCmdToken,	/* Name of hidden command. */
+    const char *cmdName)	/* Name of to-be-exposed command. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Command *cmdPtr;
+    Namespace *nsPtr;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashTable *hiddenCmdTablePtr;
+    int isNew;
+
+    if (iPtr->flags & DELETED) {
+	/*
+	 * The interpreter is being deleted. Do not create any new structures,
+	 * because it is not safe to modify the interpreter.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check that we have a regular name for the command (that the user is not
+     * trying to do an expose and a rename (to another namespace) at the same
+     * time).
+     */
+
+    if (strstr(cmdName, "::") != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "cannot expose to a namespace (use expose to toplevel, then rename)",
+                -1));
+        Tcl_SetErrorCode(interp, "TCL", "EXPOSE", "NON_GLOBAL", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the command from the hidden command table:
+     */
+
+    hPtr = NULL;
+    hiddenCmdTablePtr = iPtr->hiddenCmdTablePtr;
+    if (hiddenCmdTablePtr != NULL) {
+	hPtr = Tcl_FindHashEntry(hiddenCmdTablePtr, hiddenCmdToken);
+    }
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "unknown hidden command \"%s\"", hiddenCmdToken));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "HIDDENTOKEN",
+                hiddenCmdToken, NULL);
+	return TCL_ERROR;
+    }
+    cmdPtr = Tcl_GetHashValue(hPtr);
+
+    /*
+     * Check that we have a true global namespace command (enforced by
+     * Tcl_HideCommand but let's double check. (If it was not, we would not
+     * really know how to handle it).
+     */
+
+    if (cmdPtr->nsPtr != iPtr->globalNsPtr) {
+	/*
+	 * This case is theoritically impossible, we might rather Tcl_Panic
+	 * than 'nicely' erroring out ?
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"trying to expose a non-global command namespace command",
+		-1));
+	return TCL_ERROR;
+    }
+
+    /*
+     * This is the global table.
+     */
+
+    nsPtr = cmdPtr->nsPtr;
+
+    /*
+     * It is an error to overwrite an existing exposed command as a result of
+     * exposing a previously hidden command.
+     */
+
+    hPtr = Tcl_CreateHashEntry(&nsPtr->cmdTable, cmdName, &isNew);
+    if (!isNew) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "exposed command \"%s\" already exists", cmdName));
+        Tcl_SetErrorCode(interp, "TCL", "EXPOSE", "COMMAND_EXISTS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Command resolvers (per-interp, per-namespace) might have resolved to a
+     * command for the given namespace scope with this command not being
+     * registered with the namespace's command table. During BC compilation,
+     * the so-resolved command turns into a CmdName literal. Without
+     * invalidating a possible CmdName literal here explicitly, such literals
+     * keep being reused while pointing to overhauled commands.
+     */
+
+    TclInvalidateCmdLiteral(interp, cmdName, nsPtr);
+
+    /*
+     * The list of command exported from the namespace might have changed.
+     * However, we do not need to recompute this just yet; next time we need
+     * the info will be soon enough.
+     */
+
+    TclInvalidateNsCmdLookup(nsPtr);
+
+    /*
+     * Remove the hash entry for the command from the interpreter hidden
+     * command table.
+     */
+
+    if (cmdPtr->hPtr != NULL) {
+	Tcl_DeleteHashEntry(cmdPtr->hPtr);
+	cmdPtr->hPtr = NULL;
+    }
+
+    /*
+     * Now link the hash table entry with the command structure. This is like
+     * creating a new command, so deal with any shadowing of commands in the
+     * global namespace.
+     */
+
+    cmdPtr->hPtr = hPtr;
+
+    Tcl_SetHashValue(hPtr, cmdPtr);
+
+    /*
+     * Not needed as we are only in the global namespace (but would be needed
+     * again if we supported namespace command hiding)
+     *
+     * TclResetShadowedCmdRefs(interp, cmdPtr);
+     */
+
+    /*
+     * If the command being exposed has a compile function, increment
+     * interpreter's compileEpoch to invalidate its compiled code. This makes
+     * sure that we don't later try to execute old code compiled assuming the
+     * command is hidden. This field is checked in Tcl_EvalObj and
+     * ObjInterpProc, and code whose compilation epoch doesn't match is
+     * recompiled.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	iPtr->compileEpoch++;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateCommand --
+ *
+ *	Define a new command in a command table.
+ *
+ * Results:
+ *	The return value is a token for the command, which can be used in
+ *	future calls to Tcl_GetCommandName.
+ *
+ * Side effects:
+ *	If a command named cmdName already exists for interp, it is deleted.
+ *	In the future, when cmdName is seen as the name of a command by
+ *	Tcl_Eval, proc will be called. To support the bytecode interpreter,
+ *	the command is created with a wrapper Tcl_ObjCmdProc
+ *	(TclInvokeStringCommand) that eventially calls proc. When the command
+ *	is deleted from the table, deleteProc will be called. See the manual
+ *	entry for details on the calling sequence.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_CreateCommand(
+    Tcl_Interp *interp,		/* Token for command interpreter returned by a
+				 * previous call to Tcl_CreateInterp. */
+    const char *cmdName,	/* Name of command. If it contains namespace
+				 * qualifiers, the new command is put in the
+				 * specified namespace; otherwise it is put in
+				 * the global namespace. */
+    Tcl_CmdProc *proc,		/* Function to associate with cmdName. */
+    ClientData clientData,	/* Arbitrary value passed to string proc. */
+    Tcl_CmdDeleteProc *deleteProc)
+				/* If not NULL, gives a function to call when
+				 * this command is deleted. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ImportRef *oldRefPtr = NULL;
+    Namespace *nsPtr, *dummy1, *dummy2;
+    Command *cmdPtr, *refCmdPtr;
+    Tcl_HashEntry *hPtr;
+    const char *tail;
+    int isNew;
+    ImportedCmdData *dataPtr;
+
+    if (iPtr->flags & DELETED) {
+	/*
+	 * The interpreter is being deleted. Don't create any new commands;
+	 * it's not safe to muck with the interpreter anymore.
+	 */
+
+	return (Tcl_Command) NULL;
+    }
+
+    /*
+     * Determine where the command should reside. If its name contains
+     * namespace qualifiers, we put it in the specified namespace; otherwise,
+     * we always put it in the global namespace.
+     */
+
+    if (strstr(cmdName, "::") != NULL) {
+	TclGetNamespaceForQualName(interp, cmdName, NULL,
+		TCL_CREATE_NS_IF_UNKNOWN, &nsPtr, &dummy1, &dummy2, &tail);
+	if ((nsPtr == NULL) || (tail == NULL)) {
+	    return (Tcl_Command) NULL;
+	}
+    } else {
+	nsPtr = iPtr->globalNsPtr;
+	tail = cmdName;
+    }
+
+    hPtr = Tcl_CreateHashEntry(&nsPtr->cmdTable, tail, &isNew);
+    if (!isNew) {
+	/*
+	 * Command already exists. Delete the old one. Be careful to preserve
+	 * any existing import links so we can restore them down below. That
+	 * way, you can redefine a command and its import status will remain
+	 * intact.
+	 */
+
+	cmdPtr = Tcl_GetHashValue(hPtr);
+	cmdPtr->refCount++;
+	if (cmdPtr->importRefPtr) {
+	    cmdPtr->flags |= CMD_REDEF_IN_PROGRESS;
+	}
+
+	Tcl_DeleteCommandFromToken(interp, (Tcl_Command) cmdPtr);
+
+	if (cmdPtr->flags & CMD_REDEF_IN_PROGRESS) {
+	    oldRefPtr = cmdPtr->importRefPtr;
+	    cmdPtr->importRefPtr = NULL;
+	}
+	TclCleanupCommandMacro(cmdPtr);
+
+	hPtr = Tcl_CreateHashEntry(&nsPtr->cmdTable, tail, &isNew);
+	if (!isNew) {
+	    /*
+	     * If the deletion callback recreated the command, just throw away
+	     * the new command (if we try to delete it again, we could get
+	     * stuck in an infinite loop).
+	     */
+
+	    ckfree(Tcl_GetHashValue(hPtr));
+	}
+    } else {
+	/*
+	 * Command resolvers (per-interp, per-namespace) might have resolved
+	 * to a command for the given namespace scope with this command not
+	 * being registered with the namespace's command table. During BC
+	 * compilation, the so-resolved command turns into a CmdName literal.
+	 * Without invalidating a possible CmdName literal here explicitly,
+	 * such literals keep being reused while pointing to overhauled
+	 * commands.
+	 */
+
+	TclInvalidateCmdLiteral(interp, tail, nsPtr);
+
+	/*
+	 * The list of command exported from the namespace might have changed.
+	 * However, we do not need to recompute this just yet; next time we
+	 * need the info will be soon enough.
+	 */
+
+	TclInvalidateNsCmdLookup(nsPtr);
+	TclInvalidateNsPath(nsPtr);
+    }
+    cmdPtr = ckalloc(sizeof(Command));
+    Tcl_SetHashValue(hPtr, cmdPtr);
+    cmdPtr->hPtr = hPtr;
+    cmdPtr->nsPtr = nsPtr;
+    cmdPtr->refCount = 1;
+    cmdPtr->cmdEpoch = 0;
+    cmdPtr->compileProc = NULL;
+    cmdPtr->objProc = TclInvokeStringCommand;
+    cmdPtr->objClientData = cmdPtr;
+    cmdPtr->proc = proc;
+    cmdPtr->clientData = clientData;
+    cmdPtr->deleteProc = deleteProc;
+    cmdPtr->deleteData = clientData;
+    cmdPtr->flags = 0;
+    cmdPtr->importRefPtr = NULL;
+    cmdPtr->tracePtr = NULL;
+    cmdPtr->nreProc = NULL;
+
+    /*
+     * Plug in any existing import references found above. Be sure to update
+     * all of these references to point to the new command.
+     */
+
+    if (oldRefPtr != NULL) {
+	cmdPtr->importRefPtr = oldRefPtr;
+	while (oldRefPtr != NULL) {
+	    refCmdPtr = oldRefPtr->importedCmdPtr;
+	    dataPtr = refCmdPtr->objClientData;
+	    dataPtr->realCmdPtr = cmdPtr;
+	    oldRefPtr = oldRefPtr->nextPtr;
+	}
+    }
+
+    /*
+     * We just created a command, so in its namespace and all of its parent
+     * namespaces, it may shadow global commands with the same name. If any
+     * shadowed commands are found, invalidate all cached command references
+     * in the affected namespaces.
+     */
+
+    TclResetShadowedCmdRefs(interp, cmdPtr);
+    return (Tcl_Command) cmdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateObjCommand --
+ *
+ *	Define a new object-based command in a command table.
+ *
+ * Results:
+ *	The return value is a token for the command, which can be used in
+ *	future calls to Tcl_GetCommandName.
+ *
+ * Side effects:
+ *	If a command named "cmdName" already exists for interp, it is
+ *	first deleted.  Then the new command is created from the arguments.
+ *	[***] (See below for exception).
+ *
+ *	In the future, during bytecode evaluation when "cmdName" is seen as
+ *	the name of a command by Tcl_EvalObj or Tcl_Eval, the object-based
+ *	Tcl_ObjCmdProc proc will be called. When the command is deleted from
+ *	the table, deleteProc will be called. See the manual entry for details
+ *	on the calling sequence.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_CreateObjCommand(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * previous call to Tcl_CreateInterp). */
+    const char *cmdName,	/* Name of command. If it contains namespace
+				 * qualifiers, the new command is put in the
+				 * specified namespace; otherwise it is put in
+				 * the global namespace. */
+    Tcl_ObjCmdProc *proc,	/* Object-based function to associate with
+				 * name. */
+    ClientData clientData,	/* Arbitrary value to pass to object
+				 * function. */
+    Tcl_CmdDeleteProc *deleteProc)
+				/* If not NULL, gives a function to call when
+				 * this command is deleted. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ImportRef *oldRefPtr = NULL;
+    Namespace *nsPtr, *dummy1, *dummy2;
+    Command *cmdPtr, *refCmdPtr;
+    Tcl_HashEntry *hPtr;
+    const char *tail;
+    int isNew;
+    ImportedCmdData *dataPtr;
+
+    if (iPtr->flags & DELETED) {
+	/*
+	 * The interpreter is being deleted. Don't create any new commands;
+	 * it's not safe to muck with the interpreter anymore.
+	 */
+
+	return (Tcl_Command) NULL;
+    }
+
+    /*
+     * Determine where the command should reside. If its name contains
+     * namespace qualifiers, we put it in the specified namespace; otherwise,
+     * we always put it in the global namespace.
+     */
+
+    if (strstr(cmdName, "::") != NULL) {
+	TclGetNamespaceForQualName(interp, cmdName, NULL,
+		TCL_CREATE_NS_IF_UNKNOWN, &nsPtr, &dummy1, &dummy2, &tail);
+	if ((nsPtr == NULL) || (tail == NULL)) {
+	    return (Tcl_Command) NULL;
+	}
+    } else {
+	nsPtr = iPtr->globalNsPtr;
+	tail = cmdName;
+    }
+
+    hPtr = Tcl_CreateHashEntry(&nsPtr->cmdTable, tail, &isNew);
+    TclInvalidateNsPath(nsPtr);
+    if (!isNew) {
+	cmdPtr = Tcl_GetHashValue(hPtr);
+
+	/* Command already exists. */
+
+	/*
+	 * [***] This is wrong.  See Tcl Bug a16752c252.
+	 * However, this buggy behavior is kept under particular
+	 * circumstances to accommodate deployed binaries of the
+	 * "tclcompiler" program. http://sourceforge.net/projects/tclpro/
+	 * that crash if the bug is fixed.
+	 */
+
+	if (cmdPtr->objProc == TclInvokeStringCommand
+		&& cmdPtr->clientData == clientData
+		&& cmdPtr->deleteData == clientData
+		&& cmdPtr->deleteProc == deleteProc) {
+	    cmdPtr->objProc = proc;
+	    cmdPtr->objClientData = clientData;
+	    return (Tcl_Command) cmdPtr;
+	}
+
+	/*
+	 * Otherwise, we delete the old command. Be careful to preserve any
+	 * existing import links so we can restore them down below. That way,
+	 * you can redefine a command and its import status will remain
+	 * intact.
+	 */
+
+	cmdPtr->refCount++;
+	if (cmdPtr->importRefPtr) {
+	    cmdPtr->flags |= CMD_REDEF_IN_PROGRESS;
+	}
+
+	Tcl_DeleteCommandFromToken(interp, (Tcl_Command) cmdPtr);
+
+	if (cmdPtr->flags & CMD_REDEF_IN_PROGRESS) {
+	    oldRefPtr = cmdPtr->importRefPtr;
+	    cmdPtr->importRefPtr = NULL;
+	}
+	TclCleanupCommandMacro(cmdPtr);
+
+	hPtr = Tcl_CreateHashEntry(&nsPtr->cmdTable, tail, &isNew);
+	if (!isNew) {
+	    /*
+	     * If the deletion callback recreated the command, just throw away
+	     * the new command (if we try to delete it again, we could get
+	     * stuck in an infinite loop).
+	     */
+
+	    ckfree(Tcl_GetHashValue(hPtr));
+	}
+    } else {
+	/*
+	 * Command resolvers (per-interp, per-namespace) might have resolved
+	 * to a command for the given namespace scope with this command not
+	 * being registered with the namespace's command table. During BC
+	 * compilation, the so-resolved command turns into a CmdName literal.
+	 * Without invalidating a possible CmdName literal here explicitly,
+	 * such literals keep being reused while pointing to overhauled
+	 * commands.
+	 */
+
+	TclInvalidateCmdLiteral(interp, tail, nsPtr);
+
+	/*
+	 * The list of command exported from the namespace might have changed.
+	 * However, we do not need to recompute this just yet; next time we
+	 * need the info will be soon enough.
+	 */
+
+	TclInvalidateNsCmdLookup(nsPtr);
+    }
+    cmdPtr = ckalloc(sizeof(Command));
+    Tcl_SetHashValue(hPtr, cmdPtr);
+    cmdPtr->hPtr = hPtr;
+    cmdPtr->nsPtr = nsPtr;
+    cmdPtr->refCount = 1;
+    cmdPtr->cmdEpoch = 0;
+    cmdPtr->compileProc = NULL;
+    cmdPtr->objProc = proc;
+    cmdPtr->objClientData = clientData;
+    cmdPtr->proc = TclInvokeObjectCommand;
+    cmdPtr->clientData = cmdPtr;
+    cmdPtr->deleteProc = deleteProc;
+    cmdPtr->deleteData = clientData;
+    cmdPtr->flags = 0;
+    cmdPtr->importRefPtr = NULL;
+    cmdPtr->tracePtr = NULL;
+    cmdPtr->nreProc = NULL;
+
+    /*
+     * Plug in any existing import references found above. Be sure to update
+     * all of these references to point to the new command.
+     */
+
+    if (oldRefPtr != NULL) {
+	cmdPtr->importRefPtr = oldRefPtr;
+	while (oldRefPtr != NULL) {
+	    refCmdPtr = oldRefPtr->importedCmdPtr;
+	    dataPtr = refCmdPtr->objClientData;
+	    dataPtr->realCmdPtr = cmdPtr;
+	    oldRefPtr = oldRefPtr->nextPtr;
+	}
+    }
+
+    /*
+     * We just created a command, so in its namespace and all of its parent
+     * namespaces, it may shadow global commands with the same name. If any
+     * shadowed commands are found, invalidate all cached command references
+     * in the affected namespaces.
+     */
+
+    TclResetShadowedCmdRefs(interp, cmdPtr);
+    return (Tcl_Command) cmdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInvokeStringCommand --
+ *
+ *	"Wrapper" Tcl_ObjCmdProc used to call an existing string-based
+ *	Tcl_CmdProc if no object-based function exists for a command. A
+ *	pointer to this function is stored as the Tcl_ObjCmdProc in a Command
+ *	structure. It simply turns around and calls the string Tcl_CmdProc in
+ *	the Command structure.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	Besides those side effects of the called Tcl_CmdProc,
+ *	TclInvokeStringCommand allocates and frees storage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInvokeStringCommand(
+    ClientData clientData,	/* Points to command's Command structure. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    register int objc,		/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Command *cmdPtr = clientData;
+    int i, result;
+    const char **argv =
+	    TclStackAlloc(interp, (unsigned)(objc + 1) * sizeof(char *));
+
+    for (i = 0; i < objc; i++) {
+	argv[i] = TclGetString(objv[i]);
+    }
+    argv[objc] = 0;
+
+    /*
+     * Invoke the command's string-based Tcl_CmdProc.
+     */
+
+    result = cmdPtr->proc(cmdPtr->clientData, interp, objc, argv);
+
+    TclStackFree(interp, (void *) argv);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInvokeObjectCommand --
+ *
+ *	"Wrapper" Tcl_CmdProc used to call an existing object-based
+ *	Tcl_ObjCmdProc if no string-based function exists for a command. A
+ *	pointer to this function is stored as the Tcl_CmdProc in a Command
+ *	structure. It simply turns around and calls the object Tcl_ObjCmdProc
+ *	in the Command structure.
+ *
+ * Results:
+ *	A standard Tcl string result value.
+ *
+ * Side effects:
+ *	Besides those side effects of the called Tcl_ObjCmdProc,
+ *	TclInvokeObjectCommand allocates and frees storage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInvokeObjectCommand(
+    ClientData clientData,	/* Points to command's Command structure. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    register const char **argv)	/* Argument strings. */
+{
+    Command *cmdPtr = clientData;
+    Tcl_Obj *objPtr;
+    int i, length, result;
+    Tcl_Obj **objv =
+	    TclStackAlloc(interp, (unsigned)(argc * sizeof(Tcl_Obj *)));
+
+    for (i = 0; i < argc; i++) {
+	length = strlen(argv[i]);
+	TclNewStringObj(objPtr, argv[i], length);
+	Tcl_IncrRefCount(objPtr);
+	objv[i] = objPtr;
+    }
+
+    /*
+     * Invoke the command's object-based Tcl_ObjCmdProc.
+     */
+
+    if (cmdPtr->objProc != NULL) {
+	result = cmdPtr->objProc(cmdPtr->objClientData, interp, argc, objv);
+    } else {
+	result = Tcl_NRCallObjProc(interp, cmdPtr->nreProc,
+		cmdPtr->objClientData, argc, objv);
+    }
+
+    /*
+     * Move the interpreter's object result to the string result, then reset
+     * the object result.
+     */
+
+    (void) Tcl_GetStringResult(interp);
+
+    /*
+     * Decrement the ref counts for the argument objects created above, then
+     * free the objv array if malloc'ed storage was used.
+     */
+
+    for (i = 0; i < argc; i++) {
+	objPtr = objv[i];
+	Tcl_DecrRefCount(objPtr);
+    }
+    TclStackFree(interp, objv);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRenameCommand --
+ *
+ *	Called to give an existing Tcl command a different name. Both the old
+ *	command name and the new command name can have "::" namespace
+ *	qualifiers. If the new command has a different namespace context, the
+ *	command will be moved to that namespace and will execute in the
+ *	context of that new namespace.
+ *
+ *	If the new command name is NULL or the null string, the command is
+ *	deleted.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	If anything goes wrong, an error message is returned in the
+ *	interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclRenameCommand(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    const char *oldName,	/* Existing command name. */
+    const char *newName)	/* New command name. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *newTail;
+    Namespace *cmdNsPtr, *newNsPtr, *dummy1, *dummy2;
+    Tcl_Command cmd;
+    Command *cmdPtr;
+    Tcl_HashEntry *hPtr, *oldHPtr;
+    int isNew, result;
+    Tcl_Obj *oldFullName;
+    Tcl_DString newFullName;
+
+    /*
+     * Find the existing command. An error is returned if cmdName can't be
+     * found.
+     */
+
+    cmd = Tcl_FindCommand(interp, oldName, NULL, /*flags*/ 0);
+    cmdPtr = (Command *) cmd;
+    if (cmdPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't %s \"%s\": command doesn't exist",
+		((newName == NULL)||(*newName == '\0'))? "delete":"rename",
+		oldName));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND", oldName, NULL);
+	return TCL_ERROR;
+    }
+    cmdNsPtr = cmdPtr->nsPtr;
+    oldFullName = Tcl_NewObj();
+    Tcl_IncrRefCount(oldFullName);
+    Tcl_GetCommandFullName(interp, cmd, oldFullName);
+
+    /*
+     * If the new command name is NULL or empty, delete the command. Do this
+     * with Tcl_DeleteCommandFromToken, since we already have the command.
+     */
+
+    if ((newName == NULL) || (*newName == '\0')) {
+	Tcl_DeleteCommandFromToken(interp, cmd);
+	result = TCL_OK;
+	goto done;
+    }
+
+    /*
+     * Make sure that the destination command does not already exist. The
+     * rename operation is like creating a command, so we should automatically
+     * create the containing namespaces just like Tcl_CreateCommand would.
+     */
+
+    TclGetNamespaceForQualName(interp, newName, NULL,
+	    TCL_CREATE_NS_IF_UNKNOWN, &newNsPtr, &dummy1, &dummy2, &newTail);
+
+    if ((newNsPtr == NULL) || (newTail == NULL)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't rename to \"%s\": bad command name", newName));
+        Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", NULL);
+	result = TCL_ERROR;
+	goto done;
+    }
+    if (Tcl_FindHashEntry(&newNsPtr->cmdTable, newTail) != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't rename to \"%s\": command already exists", newName));
+        Tcl_SetErrorCode(interp, "TCL", "OPERATION", "RENAME",
+                "TARGET_EXISTS", NULL);
+	result = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Warning: any changes done in the code here are likely to be needed in
+     * Tcl_HideCommand code too (until the common parts are extracted out).
+     * - dl
+     */
+
+    /*
+     * Put the command in the new namespace so we can check for an alias loop.
+     * Since we are adding a new command to a namespace, we must handle any
+     * shadowing of the global commands that this might create.
+     */
+
+    oldHPtr = cmdPtr->hPtr;
+    hPtr = Tcl_CreateHashEntry(&newNsPtr->cmdTable, newTail, &isNew);
+    Tcl_SetHashValue(hPtr, cmdPtr);
+    cmdPtr->hPtr = hPtr;
+    cmdPtr->nsPtr = newNsPtr;
+    TclResetShadowedCmdRefs(interp, cmdPtr);
+
+    /*
+     * Now check for an alias loop. If we detect one, put everything back the
+     * way it was and report the error.
+     */
+
+    result = TclPreventAliasLoop(interp, interp, (Tcl_Command) cmdPtr);
+    if (result != TCL_OK) {
+	Tcl_DeleteHashEntry(cmdPtr->hPtr);
+	cmdPtr->hPtr = oldHPtr;
+	cmdPtr->nsPtr = cmdNsPtr;
+	goto done;
+    }
+
+    /*
+     * The list of command exported from the namespace might have changed.
+     * However, we do not need to recompute this just yet; next time we need
+     * the info will be soon enough. These might refer to the same variable,
+     * but that's no big deal.
+     */
+
+    TclInvalidateNsCmdLookup(cmdNsPtr);
+    TclInvalidateNsCmdLookup(cmdPtr->nsPtr);
+
+    /*
+     * Command resolvers (per-interp, per-namespace) might have resolved to a
+     * command for the given namespace scope with this command not being
+     * registered with the namespace's command table. During BC compilation,
+     * the so-resolved command turns into a CmdName literal. Without
+     * invalidating a possible CmdName literal here explicitly, such literals
+     * keep being reused while pointing to overhauled commands.
+     */
+
+    TclInvalidateCmdLiteral(interp, newTail, cmdPtr->nsPtr);
+
+    /*
+     * Script for rename traces can delete the command "oldName". Therefore
+     * increment the reference count for cmdPtr so that it's Command structure
+     * is freed only towards the end of this function by calling
+     * TclCleanupCommand.
+     *
+     * The trace function needs to get a fully qualified name for old and new
+     * commands [Tcl bug #651271], or else there's no way for the trace
+     * function to get the namespace from which the old command is being
+     * renamed!
+     */
+
+    Tcl_DStringInit(&newFullName);
+    Tcl_DStringAppend(&newFullName, newNsPtr->fullName, -1);
+    if (newNsPtr != iPtr->globalNsPtr) {
+	TclDStringAppendLiteral(&newFullName, "::");
+    }
+    Tcl_DStringAppend(&newFullName, newTail, -1);
+    cmdPtr->refCount++;
+    CallCommandTraces(iPtr, cmdPtr, TclGetString(oldFullName),
+	    Tcl_DStringValue(&newFullName), TCL_TRACE_RENAME);
+    Tcl_DStringFree(&newFullName);
+
+    /*
+     * The new command name is okay, so remove the command from its current
+     * namespace. This is like deleting the command, so bump the cmdEpoch to
+     * invalidate any cached references to the command.
+     */
+
+    Tcl_DeleteHashEntry(oldHPtr);
+    cmdPtr->cmdEpoch++;
+
+    /*
+     * If the command being renamed has a compile function, increment the
+     * interpreter's compileEpoch to invalidate its compiled code. This makes
+     * sure that we don't later try to execute old code compiled for the
+     * now-renamed command.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	iPtr->compileEpoch++;
+    }
+
+    /*
+     * Now free the Command structure, if the "oldName" command has been
+     * deleted by invocation of rename traces.
+     */
+
+    TclCleanupCommandMacro(cmdPtr);
+    result = TCL_OK;
+
+  done:
+    TclDecrRefCount(oldFullName);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetCommandInfo --
+ *
+ *	Modifies various information about a Tcl command. Note that this
+ *	function will not change a command's namespace; use TclRenameCommand
+ *	to do that. Also, the isNativeObjectProc member of *infoPtr is
+ *	ignored.
+ *
+ * Results:
+ *	If cmdName exists in interp, then the information at *infoPtr is
+ *	stored with the command in place of the current information and 1 is
+ *	returned. If the command doesn't exist then 0 is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetCommandInfo(
+    Tcl_Interp *interp,		/* Interpreter in which to look for
+				 * command. */
+    const char *cmdName,	/* Name of desired command. */
+    const Tcl_CmdInfo *infoPtr)	/* Where to find information to store in the
+				 * command. */
+{
+    Tcl_Command cmd;
+
+    cmd = Tcl_FindCommand(interp, cmdName, NULL, /*flags*/ 0);
+    return Tcl_SetCommandInfoFromToken(cmd, infoPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetCommandInfoFromToken --
+ *
+ *	Modifies various information about a Tcl command. Note that this
+ *	function will not change a command's namespace; use TclRenameCommand
+ *	to do that. Also, the isNativeObjectProc member of *infoPtr is
+ *	ignored.
+ *
+ * Results:
+ *	If cmdName exists in interp, then the information at *infoPtr is
+ *	stored with the command in place of the current information and 1 is
+ *	returned. If the command doesn't exist then 0 is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetCommandInfoFromToken(
+    Tcl_Command cmd,
+    const Tcl_CmdInfo *infoPtr)
+{
+    Command *cmdPtr;		/* Internal representation of the command */
+
+    if (cmd == NULL) {
+	return 0;
+    }
+
+    /*
+     * The isNativeObjectProc and nsPtr members of *infoPtr are ignored.
+     */
+
+    cmdPtr = (Command *) cmd;
+    cmdPtr->proc = infoPtr->proc;
+    cmdPtr->clientData = infoPtr->clientData;
+    if (infoPtr->objProc == NULL) {
+	cmdPtr->objProc = TclInvokeStringCommand;
+	cmdPtr->objClientData = cmdPtr;
+	cmdPtr->nreProc = NULL;
+    } else {
+	if (infoPtr->objProc != cmdPtr->objProc) {
+	    cmdPtr->nreProc = NULL;
+	    cmdPtr->objProc = infoPtr->objProc;
+	}
+	cmdPtr->objClientData = infoPtr->objClientData;
+    }
+    cmdPtr->deleteProc = infoPtr->deleteProc;
+    cmdPtr->deleteData = infoPtr->deleteData;
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCommandInfo --
+ *
+ *	Returns various information about a Tcl command.
+ *
+ * Results:
+ *	If cmdName exists in interp, then *infoPtr is modified to hold
+ *	information about cmdName and 1 is returned. If the command doesn't
+ *	exist then 0 is returned and *infoPtr isn't modified.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetCommandInfo(
+    Tcl_Interp *interp,		/* Interpreter in which to look for
+				 * command. */
+    const char *cmdName,	/* Name of desired command. */
+    Tcl_CmdInfo *infoPtr)	/* Where to store information about
+				 * command. */
+{
+    Tcl_Command cmd;
+
+    cmd = Tcl_FindCommand(interp, cmdName, NULL, /*flags*/ 0);
+    return Tcl_GetCommandInfoFromToken(cmd, infoPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCommandInfoFromToken --
+ *
+ *	Returns various information about a Tcl command.
+ *
+ * Results:
+ *	Copies information from the command identified by 'cmd' into a
+ *	caller-supplied structure and returns 1. If the 'cmd' is NULL, leaves
+ *	the structure untouched and returns 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetCommandInfoFromToken(
+    Tcl_Command cmd,
+    Tcl_CmdInfo *infoPtr)
+{
+    Command *cmdPtr;		/* Internal representation of the command */
+
+    if (cmd == NULL) {
+	return 0;
+    }
+
+    /*
+     * Set isNativeObjectProc 1 if objProc was registered by a call to
+     * Tcl_CreateObjCommand. Otherwise set it to 0.
+     */
+
+    cmdPtr = (Command *) cmd;
+    infoPtr->isNativeObjectProc =
+	    (cmdPtr->objProc != TclInvokeStringCommand);
+    infoPtr->objProc = cmdPtr->objProc;
+    infoPtr->objClientData = cmdPtr->objClientData;
+    infoPtr->proc = cmdPtr->proc;
+    infoPtr->clientData = cmdPtr->clientData;
+    infoPtr->deleteProc = cmdPtr->deleteProc;
+    infoPtr->deleteData = cmdPtr->deleteData;
+    infoPtr->namespacePtr = (Tcl_Namespace *) cmdPtr->nsPtr;
+
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCommandName --
+ *
+ *	Given a token returned by Tcl_CreateCommand, this function returns the
+ *	current name of the command (which may have changed due to renaming).
+ *
+ * Results:
+ *	The return value is the name of the given command.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetCommandName(
+    Tcl_Interp *interp,		/* Interpreter containing the command. */
+    Tcl_Command command)	/* Token for command returned by a previous
+				 * call to Tcl_CreateCommand. The command must
+				 * not have been deleted. */
+{
+    Command *cmdPtr = (Command *) command;
+
+    if ((cmdPtr == NULL) || (cmdPtr->hPtr == NULL)) {
+	/*
+	 * This should only happen if command was "created" after the
+	 * interpreter began to be deleted, so there isn't really any command.
+	 * Just return an empty string.
+	 */
+
+	return "";
+    }
+
+    return Tcl_GetHashKey(cmdPtr->hPtr->tablePtr, cmdPtr->hPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCommandFullName --
+ *
+ *	Given a token returned by, e.g., Tcl_CreateCommand or Tcl_FindCommand,
+ *	this function appends to an object the command's full name, qualified
+ *	by a sequence of parent namespace names. The command's fully-qualified
+ *	name may have changed due to renaming.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The command's fully-qualified name is appended to the string
+ *	representation of objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetCommandFullName(
+    Tcl_Interp *interp,		/* Interpreter containing the command. */
+    Tcl_Command command,	/* Token for command returned by a previous
+				 * call to Tcl_CreateCommand. The command must
+				 * not have been deleted. */
+    Tcl_Obj *objPtr)		/* Points to the object onto which the
+				 * command's full name is appended. */
+
+{
+    Interp *iPtr = (Interp *) interp;
+    register Command *cmdPtr = (Command *) command;
+    char *name;
+
+    /*
+     * Add the full name of the containing namespace, followed by the "::"
+     * separator, and the command name.
+     */
+
+    if (cmdPtr != NULL) {
+	if (cmdPtr->nsPtr != NULL) {
+	    Tcl_AppendToObj(objPtr, cmdPtr->nsPtr->fullName, -1);
+	    if (cmdPtr->nsPtr != iPtr->globalNsPtr) {
+		Tcl_AppendToObj(objPtr, "::", 2);
+	    }
+	}
+	if (cmdPtr->hPtr != NULL) {
+	    name = Tcl_GetHashKey(cmdPtr->hPtr->tablePtr, cmdPtr->hPtr);
+	    Tcl_AppendToObj(objPtr, name, -1);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteCommand --
+ *
+ *	Remove the given command from the given interpreter.
+ *
+ * Results:
+ *	0 is returned if the command was deleted successfully. -1 is returned
+ *	if there didn't exist a command by that name.
+ *
+ * Side effects:
+ *	cmdName will no longer be recognized as a valid command for interp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DeleteCommand(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * a previous Tcl_CreateInterp call). */
+    const char *cmdName)	/* Name of command to remove. */
+{
+    Tcl_Command cmd;
+
+    /*
+     * Find the desired command and delete it.
+     */
+
+    cmd = Tcl_FindCommand(interp, cmdName, NULL, /*flags*/ 0);
+    if (cmd == NULL) {
+	return -1;
+    }
+    return Tcl_DeleteCommandFromToken(interp, cmd);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteCommandFromToken --
+ *
+ *	Removes the given command from the given interpreter. This function
+ *	resembles Tcl_DeleteCommand, but takes a Tcl_Command token instead of
+ *	a command name for efficiency.
+ *
+ * Results:
+ *	0 is returned if the command was deleted successfully. -1 is returned
+ *	if there didn't exist a command by that name.
+ *
+ * Side effects:
+ *	The command specified by "cmd" will no longer be recognized as a valid
+ *	command for "interp".
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DeleteCommandFromToken(
+    Tcl_Interp *interp,		/* Token for command interpreter returned by a
+				 * previous call to Tcl_CreateInterp. */
+    Tcl_Command cmd)		/* Token for command to delete. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Command *cmdPtr = (Command *) cmd;
+    ImportRef *refPtr, *nextRefPtr;
+    Tcl_Command importCmd;
+
+    /*
+     * The code here is tricky. We can't delete the hash table entry before
+     * invoking the deletion callback because there are cases where the
+     * deletion callback needs to invoke the command (e.g. object systems such
+     * as OTcl). However, this means that the callback could try to delete or
+     * rename the command. The deleted flag allows us to detect these cases
+     * and skip nested deletes.
+     */
+
+    if (cmdPtr->flags & CMD_IS_DELETED) {
+	/*
+	 * Another deletion is already in progress. Remove the hash table
+	 * entry now, but don't invoke a callback or free the command
+	 * structure. Take care to only remove the hash entry if it has not
+	 * already been removed; otherwise if we manage to hit this function
+	 * three times, everything goes up in smoke. [Bug 1220058]
+	 */
+
+	if (cmdPtr->hPtr != NULL) {
+	    Tcl_DeleteHashEntry(cmdPtr->hPtr);
+	    cmdPtr->hPtr = NULL;
+	}
+
+	/*
+	 * Bump the command epoch counter. This will invalidate all cached
+	 * references that point to this command.
+	 */
+
+	cmdPtr->cmdEpoch++;
+
+	return 0;
+    }
+
+    /*
+     * We must delete this command, even though both traces and delete procs
+     * may try to avoid this (renaming the command etc). Also traces and
+     * delete procs may try to delete the command themsevles. This flag
+     * declares that a delete is in progress and that recursive deletes should
+     * be ignored.
+     */
+
+    cmdPtr->flags |= CMD_IS_DELETED;
+
+    /*
+     * Call trace functions for the command being deleted. Then delete its
+     * traces.
+     */
+
+    if (cmdPtr->tracePtr != NULL) {
+	CommandTrace *tracePtr;
+	CallCommandTraces(iPtr,cmdPtr,NULL,NULL,TCL_TRACE_DELETE);
+
+	/*
+	 * Now delete these traces.
+	 */
+
+	tracePtr = cmdPtr->tracePtr;
+	while (tracePtr != NULL) {
+	    CommandTrace *nextPtr = tracePtr->nextPtr;
+
+	    if (tracePtr->refCount-- <= 1) {
+		ckfree(tracePtr);
+	    }
+	    tracePtr = nextPtr;
+	}
+	cmdPtr->tracePtr = NULL;
+    }
+
+    /*
+     * The list of command exported from the namespace might have changed.
+     * However, we do not need to recompute this just yet; next time we need
+     * the info will be soon enough.
+     */
+
+    TclInvalidateNsCmdLookup(cmdPtr->nsPtr);
+
+    /*
+     * If the command being deleted has a compile function, increment the
+     * interpreter's compileEpoch to invalidate its compiled code. This makes
+     * sure that we don't later try to execute old code compiled with
+     * command-specific (i.e., inline) bytecodes for the now-deleted command.
+     * This field is checked in Tcl_EvalObj and ObjInterpProc, and code whose
+     * compilation epoch doesn't match is recompiled.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	iPtr->compileEpoch++;
+    }
+
+    if (cmdPtr->deleteProc != NULL) {
+	/*
+	 * Delete the command's client data. If this was an imported command
+	 * created when a command was imported into a namespace, this client
+	 * data will be a pointer to a ImportedCmdData structure describing
+	 * the "real" command that this imported command refers to.
+	 *
+	 * If you are getting a crash during the call to deleteProc and
+	 * cmdPtr->deleteProc is a pointer to the function free(), the most
+	 * likely cause is that your extension allocated memory for the
+	 * clientData argument to Tcl_CreateObjCommand with the ckalloc()
+	 * macro and you are now trying to deallocate this memory with free()
+	 * instead of ckfree(). You should pass a pointer to your own method
+	 * that calls ckfree().
+	 */
+
+	cmdPtr->deleteProc(cmdPtr->deleteData);
+    }
+
+    /*
+     * If this command was imported into other namespaces, then imported
+     * commands were created that refer back to this command. Delete these
+     * imported commands now.
+     */
+    if (!(cmdPtr->flags & CMD_REDEF_IN_PROGRESS)) {
+	for (refPtr = cmdPtr->importRefPtr; refPtr != NULL;
+		refPtr = nextRefPtr) {
+	    nextRefPtr = refPtr->nextPtr;
+	    importCmd = (Tcl_Command) refPtr->importedCmdPtr;
+	    Tcl_DeleteCommandFromToken(interp, importCmd);
+	}
+    }
+
+    /*
+     * Don't use hPtr to delete the hash entry here, because it's possible
+     * that the deletion callback renamed the command. Instead, use
+     * cmdPtr->hptr, and make sure that no-one else has already deleted the
+     * hash entry.
+     */
+
+    if (cmdPtr->hPtr != NULL) {
+	Tcl_DeleteHashEntry(cmdPtr->hPtr);
+	cmdPtr->hPtr = NULL;
+
+	/*
+	 * Bump the command epoch counter. This will invalidate all cached
+	 * references that point to this command.
+	 */
+
+	cmdPtr->cmdEpoch++;
+    }
+
+    /*
+     * A number of tests for particular kinds of commands are done by checking
+     * whether the objProc field holds a known value. Set the field to NULL so
+     * that such tests won't have false positives when applied to deleted
+     * commands.
+     */
+
+    cmdPtr->objProc = NULL;
+
+    /*
+     * Now free the Command structure, unless there is another reference to it
+     * from a CmdName Tcl object in some ByteCode code sequence. In that case,
+     * delay the cleanup until all references are either discarded (when a
+     * ByteCode is freed) or replaced by a new reference (when a cached
+     * CmdName Command reference is found to be invalid and
+     * TclNRExecuteByteCode looks up the command in the command hashtable).
+     */
+
+    TclCleanupCommandMacro(cmdPtr);
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CallCommandTraces --
+ *
+ *	Abstraction of the code to call traces on a command.
+ *
+ * Results:
+ *	Currently always NULL.
+ *
+ * Side effects:
+ *	Anything; this may recursively evaluate scripts and code exists to do
+ *	just that.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+CallCommandTraces(
+    Interp *iPtr,		/* Interpreter containing command. */
+    Command *cmdPtr,		/* Command whose traces are to be invoked. */
+    const char *oldName,	/* Command's old name, or NULL if we must get
+				 * the name from cmdPtr */
+    const char *newName,	/* Command's new name, or NULL if the command
+				 * is not being renamed */
+    int flags)			/* Flags indicating the type of traces to
+				 * trigger, either TCL_TRACE_DELETE or
+				 * TCL_TRACE_RENAME. */
+{
+    register CommandTrace *tracePtr;
+    ActiveCommandTrace active;
+    char *result;
+    Tcl_Obj *oldNamePtr = NULL;
+    Tcl_InterpState state = NULL;
+
+    if (cmdPtr->flags & CMD_TRACE_ACTIVE) {
+	/*
+	 * While a rename trace is active, we will not process any more rename
+	 * traces; while a delete trace is active we will never reach here -
+	 * because Tcl_DeleteCommandFromToken checks for the condition
+	 * (cmdPtr->flags & CMD_IS_DELETED) and returns immediately when a
+	 * command deletion is in progress. For all other traces, delete
+	 * traces will not be invoked but a call to TraceCommandProc will
+	 * ensure that tracePtr->clientData is freed whenever the command
+	 * "oldName" is deleted.
+	 */
+
+	if (cmdPtr->flags & TCL_TRACE_RENAME) {
+	    flags &= ~TCL_TRACE_RENAME;
+	}
+	if (flags == 0) {
+	    return NULL;
+	}
+    }
+    cmdPtr->flags |= CMD_TRACE_ACTIVE;
+    cmdPtr->refCount++;
+
+    result = NULL;
+    active.nextPtr = iPtr->activeCmdTracePtr;
+    active.reverseScan = 0;
+    iPtr->activeCmdTracePtr = &active;
+
+    if (flags & TCL_TRACE_DELETE) {
+	flags |= TCL_TRACE_DESTROYED;
+    }
+    active.cmdPtr = cmdPtr;
+
+    Tcl_Preserve(iPtr);
+
+    for (tracePtr = cmdPtr->tracePtr; tracePtr != NULL;
+	    tracePtr = active.nextTracePtr) {
+	active.nextTracePtr = tracePtr->nextPtr;
+	if (!(tracePtr->flags & flags)) {
+	    continue;
+	}
+	cmdPtr->flags |= tracePtr->flags;
+	if (oldName == NULL) {
+	    TclNewObj(oldNamePtr);
+	    Tcl_IncrRefCount(oldNamePtr);
+	    Tcl_GetCommandFullName((Tcl_Interp *) iPtr,
+		    (Tcl_Command) cmdPtr, oldNamePtr);
+	    oldName = TclGetString(oldNamePtr);
+	}
+	tracePtr->refCount++;
+	if (state == NULL) {
+	    state = Tcl_SaveInterpState((Tcl_Interp *) iPtr, TCL_OK);
+	}
+	tracePtr->traceProc(tracePtr->clientData, (Tcl_Interp *) iPtr,
+		oldName, newName, flags);
+	cmdPtr->flags &= ~tracePtr->flags;
+	if (tracePtr->refCount-- <= 1) {
+	    ckfree(tracePtr);
+	}
+    }
+
+    if (state) {
+	Tcl_RestoreInterpState((Tcl_Interp *) iPtr, state);
+    }
+
+    /*
+     * If a new object was created to hold the full oldName, free it now.
+     */
+
+    if (oldNamePtr != NULL) {
+	TclDecrRefCount(oldNamePtr);
+    }
+
+    /*
+     * Restore the variable's flags, remove the record of our active traces,
+     * and then return.
+     */
+
+    cmdPtr->flags &= ~CMD_TRACE_ACTIVE;
+    cmdPtr->refCount--;
+    iPtr->activeCmdTracePtr = active.nextPtr;
+    Tcl_Release(iPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CancelEvalProc --
+ *
+ *	Marks this interpreter as being canceled. This causes current
+ *	executions to be unwound as the interpreter enters a state where it
+ *	refuses to execute more commands or handle [catch] or [try], yet the
+ *	interpreter is still able to execute further commands after the
+ *	cancelation is cleared (unlike if it is deleted).
+ *
+ * Results:
+ *	The value given for the code argument.
+ *
+ * Side effects:
+ *	Transfers a message from the cancelation message to the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CancelEvalProc(
+    ClientData clientData,	/* Interp to cancel the script in progress. */
+    Tcl_Interp *interp,		/* Ignored */
+    int code)			/* Current return code from command. */
+{
+    CancelInfo *cancelInfo = clientData;
+    Interp *iPtr;
+
+    if (cancelInfo != NULL) {
+	Tcl_MutexLock(&cancelLock);
+	iPtr = (Interp *) cancelInfo->interp;
+
+	if (iPtr != NULL) {
+	    /*
+	     * Setting the CANCELED flag will cause the script in progress to
+	     * be canceled as soon as possible. The core honors this flag at
+	     * all the necessary places to ensure script cancellation is
+	     * responsive. Extensions can check for this flag by calling
+	     * Tcl_Canceled and checking if TCL_ERROR is returned or they can
+	     * choose to ignore the script cancellation flag and the
+	     * associated functionality altogether. Currently, the only other
+	     * flag we care about here is the TCL_CANCEL_UNWIND flag (from
+	     * Tcl_CancelEval). We do not want to simply combine all the flags
+	     * from original Tcl_CancelEval call with the interp flags here
+	     * just in case the caller passed flags that might cause behaviour
+	     * unrelated to script cancellation.
+	     */
+
+	    TclSetCancelFlags(iPtr, cancelInfo->flags | CANCELED);
+
+	    /*
+	     * Now, we must set the script cancellation flags on all the slave
+	     * interpreters belonging to this one.
+	     */
+
+	    TclSetSlaveCancelFlags((Tcl_Interp *) iPtr,
+		    cancelInfo->flags | CANCELED, 0);
+
+	    /*
+	     * Create the result object now so that Tcl_Canceled can avoid
+	     * locking the cancelLock mutex.
+	     */
+
+	    if (cancelInfo->result != NULL) {
+		Tcl_SetStringObj(iPtr->asyncCancelMsg, cancelInfo->result,
+			cancelInfo->length);
+	    } else {
+		Tcl_SetObjLength(iPtr->asyncCancelMsg, 0);
+	    }
+	}
+	Tcl_MutexUnlock(&cancelLock);
+    }
+
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCleanupCommand --
+ *
+ *	This function frees up a Command structure unless it is still
+ *	referenced from an interpreter's command hashtable or from a CmdName
+ *	Tcl object representing the name of a command in a ByteCode
+ *	instruction sequence.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory gets freed unless a reference to the Command structure still
+ *	exists. In that case the cleanup is delayed until the command is
+ *	deleted or when the last ByteCode referring to it is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCleanupCommand(
+    register Command *cmdPtr)	/* Points to the Command structure to
+				 * be freed. */
+{
+    if (cmdPtr->refCount-- <= 1) {
+	ckfree(cmdPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateMathFunc --
+ *
+ *	Creates a new math function for expressions in a given interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The Tcl function defined by "name" is created or redefined. If the
+ *	function already exists then its definition is replaced; this includes
+ *	the builtin functions. Redefining a builtin function forces all
+ *	existing code to be invalidated since that code may be compiled using
+ *	an instruction specific to the replaced function. In addition,
+ *	redefioning a non-builtin function will force existing code to be
+ *	invalidated if the number of arguments has changed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateMathFunc(
+    Tcl_Interp *interp,		/* Interpreter in which function is to be
+				 * available. */
+    const char *name,		/* Name of function (e.g. "sin"). */
+    int numArgs,		/* Nnumber of arguments required by
+				 * function. */
+    Tcl_ValueType *argTypes,	/* Array of types acceptable for each
+				 * argument. */
+    Tcl_MathProc *proc,		/* C function that implements the math
+				 * function. */
+    ClientData clientData)	/* Additional value to pass to the
+				 * function. */
+{
+    Tcl_DString bigName;
+    OldMathFuncData *data = ckalloc(sizeof(OldMathFuncData));
+
+    data->proc = proc;
+    data->numArgs = numArgs;
+    data->argTypes = ckalloc(numArgs * sizeof(Tcl_ValueType));
+    memcpy(data->argTypes, argTypes, numArgs * sizeof(Tcl_ValueType));
+    data->clientData = clientData;
+
+    Tcl_DStringInit(&bigName);
+    TclDStringAppendLiteral(&bigName, "::tcl::mathfunc::");
+    Tcl_DStringAppend(&bigName, name, -1);
+
+    Tcl_CreateObjCommand(interp, Tcl_DStringValue(&bigName),
+	    OldMathFuncProc, data, OldMathFuncDeleteProc);
+    Tcl_DStringFree(&bigName);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * OldMathFuncProc --
+ *
+ *	Dispatch to a math function created with Tcl_CreateMathFunc
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Whatever the math function does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+OldMathFuncProc(
+    ClientData clientData,	/* Ponter to OldMathFuncData describing the
+				 * function being called */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Actual parameter count */
+    Tcl_Obj *const *objv)	/* Parameter vector */
+{
+    Tcl_Obj *valuePtr;
+    OldMathFuncData *dataPtr = clientData;
+    Tcl_Value funcResult, *args;
+    int result;
+    int j, k;
+    double d;
+
+    /*
+     * Check argument count.
+     */
+
+    if (objc != dataPtr->numArgs + 1) {
+	MathFuncWrongNumArgs(interp, dataPtr->numArgs+1, objc, objv);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert arguments from Tcl_Obj's to Tcl_Value's.
+     */
+
+    args = ckalloc(dataPtr->numArgs * sizeof(Tcl_Value));
+    for (j = 1, k = 0; j < objc; ++j, ++k) {
+	/* TODO: Convert to TclGetNumberFromObj? */
+	valuePtr = objv[j];
+	result = Tcl_GetDoubleFromObj(NULL, valuePtr, &d);
+#ifdef ACCEPT_NAN
+	if ((result != TCL_OK) && (valuePtr->typePtr == &tclDoubleType)) {
+	    d = valuePtr->internalRep.doubleValue;
+	    result = TCL_OK;
+	}
+#endif
+	if (result != TCL_OK) {
+	    /*
+	     * We have a non-numeric argument.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "argument to math function didn't have numeric value",
+		    -1));
+	    TclCheckBadOctal(interp, TclGetString(valuePtr));
+	    ckfree(args);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Copy the object's numeric value to the argument record, converting
+	 * it if necessary.
+	 *
+	 * NOTE: no bignum support; use the new mathfunc interface for that.
+	 */
+
+	args[k].type = dataPtr->argTypes[k];
+	switch (args[k].type) {
+	case TCL_EITHER:
+	    if (Tcl_GetLongFromObj(NULL, valuePtr, &args[k].intValue)
+		    == TCL_OK) {
+		args[k].type = TCL_INT;
+		break;
+	    }
+	    if (TclGetWideIntFromObj(interp, valuePtr, &args[k].wideValue)
+		    == TCL_OK) {
+		args[k].type = TCL_WIDE_INT;
+		break;
+	    }
+	    args[k].type = TCL_DOUBLE;
+	    /* FALLTHROUGH */
+
+	case TCL_DOUBLE:
+	    args[k].doubleValue = d;
+	    break;
+	case TCL_INT:
+	    if (ExprIntFunc(NULL, interp, 2, &objv[j-1]) != TCL_OK) {
+		ckfree(args);
+		return TCL_ERROR;
+	    }
+	    valuePtr = Tcl_GetObjResult(interp);
+	    Tcl_GetLongFromObj(NULL, valuePtr, &args[k].intValue);
+	    Tcl_ResetResult(interp);
+	    break;
+	case TCL_WIDE_INT:
+	    if (ExprWideFunc(NULL, interp, 2, &objv[j-1]) != TCL_OK) {
+		ckfree(args);
+		return TCL_ERROR;
+	    }
+	    valuePtr = Tcl_GetObjResult(interp);
+	    TclGetWideIntFromObj(NULL, valuePtr, &args[k].wideValue);
+	    Tcl_ResetResult(interp);
+	    break;
+	}
+    }
+
+    /*
+     * Call the function.
+     */
+
+    errno = 0;
+    result = dataPtr->proc(dataPtr->clientData, interp, args, &funcResult);
+    ckfree(args);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    /*
+     * Return the result of the call.
+     */
+
+    if (funcResult.type == TCL_INT) {
+	TclNewLongObj(valuePtr, funcResult.intValue);
+    } else if (funcResult.type == TCL_WIDE_INT) {
+	valuePtr = Tcl_NewWideIntObj(funcResult.wideValue);
+    } else {
+	return CheckDoubleResult(interp, funcResult.doubleValue);
+    }
+    Tcl_SetObjResult(interp, valuePtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * OldMathFuncDeleteProc --
+ *
+ *	Cleans up after deleting a math function registered with
+ *	Tcl_CreateMathFunc
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees allocated memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+OldMathFuncDeleteProc(
+    ClientData clientData)
+{
+    OldMathFuncData *dataPtr = clientData;
+
+    ckfree(dataPtr->argTypes);
+    ckfree(dataPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetMathFuncInfo --
+ *
+ *	Discovers how a particular math function was created in a given
+ *	interpreter.
+ *
+ * Results:
+ *	TCL_OK if it succeeds, TCL_ERROR else (leaving an error message in the
+ *	interpreter result if that happens.)
+ *
+ * Side effects:
+ *	If this function succeeds, the variables pointed to by the numArgsPtr
+ *	and argTypePtr arguments will be updated to detail the arguments
+ *	allowed by the function. The variable pointed to by the procPtr
+ *	argument will be set to NULL if the function is a builtin function,
+ *	and will be set to the address of the C function used to implement the
+ *	math function otherwise (in which case the variable pointed to by the
+ *	clientDataPtr argument will also be updated.)
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetMathFuncInfo(
+    Tcl_Interp *interp,
+    const char *name,
+    int *numArgsPtr,
+    Tcl_ValueType **argTypesPtr,
+    Tcl_MathProc **procPtr,
+    ClientData *clientDataPtr)
+{
+    Tcl_Obj *cmdNameObj;
+    Command *cmdPtr;
+
+    /*
+     * Get the command that implements the math function.
+     */
+
+    TclNewLiteralStringObj(cmdNameObj, "tcl::mathfunc::");
+    Tcl_AppendToObj(cmdNameObj, name, -1);
+    Tcl_IncrRefCount(cmdNameObj);
+    cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, cmdNameObj);
+    Tcl_DecrRefCount(cmdNameObj);
+
+    /*
+     * Report unknown functions.
+     */
+
+    if (cmdPtr == NULL) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "unknown math function \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "MATHFUNC", name, NULL);
+	*numArgsPtr = -1;
+	*argTypesPtr = NULL;
+	*procPtr = NULL;
+	*clientDataPtr = NULL;
+	return TCL_ERROR;
+    }
+
+    /*
+     * Retrieve function info for user defined functions; return dummy
+     * information for builtins.
+     */
+
+    if (cmdPtr->objProc == &OldMathFuncProc) {
+	OldMathFuncData *dataPtr = cmdPtr->clientData;
+
+	*procPtr = dataPtr->proc;
+	*numArgsPtr = dataPtr->numArgs;
+	*argTypesPtr = dataPtr->argTypes;
+	*clientDataPtr = dataPtr->clientData;
+    } else {
+	*procPtr = NULL;
+	*numArgsPtr = -1;
+	*argTypesPtr = NULL;
+	*procPtr = NULL;
+	*clientDataPtr = NULL;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListMathFuncs --
+ *
+ *	Produces a list of all the math functions defined in a given
+ *	interpreter.
+ *
+ * Results:
+ *	A pointer to a Tcl_Obj structure with a reference count of zero, or
+ *	NULL in the case of an error (in which case a suitable error message
+ *	will be left in the interpreter result.)
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ListMathFuncs(
+    Tcl_Interp *interp,
+    const char *pattern)
+{
+    Tcl_Obj *script = Tcl_NewStringObj("::info functions ", -1);
+    Tcl_Obj *result;
+    Tcl_InterpState state;
+
+    if (pattern) {
+	Tcl_Obj *patternObj = Tcl_NewStringObj(pattern, -1);
+	Tcl_Obj *arg = Tcl_NewListObj(1, &patternObj);
+
+	Tcl_AppendObjToObj(script, arg);
+	Tcl_DecrRefCount(arg);	/* Should tear down patternObj too */
+    }
+
+    state = Tcl_SaveInterpState(interp, TCL_OK);
+    Tcl_IncrRefCount(script);
+    if (TCL_OK == Tcl_EvalObjEx(interp, script, 0)) {
+	result = Tcl_DuplicateObj(Tcl_GetObjResult(interp));
+    } else {
+	result = Tcl_NewObj();
+    }
+    Tcl_DecrRefCount(script);
+    Tcl_RestoreInterpState(interp, state);
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInterpReady --
+ *
+ *	Check if an interpreter is ready to eval commands or scripts, i.e., if
+ *	it was not deleted and if the nesting level is not too high.
+ *
+ * Results:
+ *	The return value is TCL_OK if it the interpreter is ready, TCL_ERROR
+ *	otherwise.
+ *
+ * Side effects:
+ *	The interpreters object and string results are cleared.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInterpReady(
+    Tcl_Interp *interp)
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    /*
+     * Reset both the interpreter's string and object results and clear out
+     * any previous error information.
+     */
+
+    Tcl_ResetResult(interp);
+
+    /*
+     * If the interpreter has been deleted, return an error.
+     */
+
+    if (iPtr->flags & DELETED) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to call eval in deleted interpreter", -1));
+	Tcl_SetErrorCode(interp, "TCL", "IDELETE",
+		"attempt to call eval in deleted interpreter", NULL);
+	return TCL_ERROR;
+    }
+
+    if (iPtr->execEnvPtr->rewind) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure the script being evaluated (if any) has not been canceled.
+     */
+
+    if (TclCanceled(iPtr) &&
+	    (TCL_OK != Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG))) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check depth of nested calls to Tcl_Eval: if this gets too large, it's
+     * probably because of an infinite loop somewhere.
+     */
+
+    if (((iPtr->numLevels) <= iPtr->maxNestingDepth)) {
+	return TCL_OK;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    "too many nested evaluations (infinite loop?)", -1));
+    Tcl_SetErrorCode(interp, "TCL", "LIMIT", "STACK", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclResetCancellation --
+ *
+ *	Reset the script cancellation flags if the nesting level
+ *	(iPtr->numLevels) for the interp is zero or argument force is
+ *	non-zero.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	The script cancellation flags for the interp may be reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclResetCancellation(
+    Tcl_Interp *interp,
+    int force)
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    if (iPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (force || (iPtr->numLevels == 0)) {
+	TclUnsetCancelFlags(iPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Canceled --
+ *
+ *	Check if the script in progress has been canceled, i.e.,
+ *	Tcl_CancelEval was called for this interpreter or any of its master
+ *	interpreters.
+ *
+ * Results:
+ *	The return value is TCL_OK if the script evaluation has not been
+ *	canceled, TCL_ERROR otherwise.
+ *
+ *	If "flags" contains TCL_LEAVE_ERR_MSG, an error message is returned in
+ *	the interpreter's result object. Otherwise, the interpreter's result
+ *	object is left unchanged. If "flags" contains TCL_CANCEL_UNWIND,
+ *	TCL_ERROR will only be returned if the script evaluation is being
+ *	completely unwound.
+ *
+ * Side effects:
+ *	The CANCELED flag for the interp will be reset if it is set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Canceled(
+    Tcl_Interp *interp,
+    int flags)
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    /*
+     * Has the current script in progress for this interpreter been canceled
+     * or is the stack being unwound due to the previous script cancellation?
+     */
+
+    if (!TclCanceled(iPtr)) {
+        return TCL_OK;
+    }
+
+    /*
+     * The CANCELED flag is a one-shot flag that is reset immediately upon
+     * being detected; however, if the TCL_CANCEL_UNWIND flag is set we will
+     * continue to report that the script in progress has been canceled
+     * thereby allowing the evaluation stack for the interp to be fully
+     * unwound.
+     */
+
+    iPtr->flags &= ~CANCELED;
+
+    /*
+     * The CANCELED flag was detected and reset; however, if the caller
+     * specified the TCL_CANCEL_UNWIND flag, we only return TCL_ERROR
+     * (indicating that the script in progress has been canceled) if the
+     * evaluation stack for the interp is being fully unwound.
+     */
+
+    if ((flags & TCL_CANCEL_UNWIND) && !(iPtr->flags & TCL_CANCEL_UNWIND)) {
+        return TCL_OK;
+    }
+
+    /*
+     * If the TCL_LEAVE_ERR_MSG flags bit is set, place an error in the
+     * interp's result; otherwise, we leave it alone.
+     */
+
+    if (flags & TCL_LEAVE_ERR_MSG) {
+        const char *id, *message = NULL;
+        int length;
+
+        /*
+         * Setup errorCode variables so that we can differentiate between
+         * being canceled and unwound.
+         */
+
+        if (iPtr->asyncCancelMsg != NULL) {
+            message = TclGetStringFromObj(iPtr->asyncCancelMsg, &length);
+        } else {
+            length = 0;
+        }
+
+        if (iPtr->flags & TCL_CANCEL_UNWIND) {
+            id = "IUNWIND";
+            if (length == 0) {
+                message = "eval unwound";
+            }
+        } else {
+            id = "ICANCEL";
+            if (length == 0) {
+                message = "eval canceled";
+            }
+        }
+
+        Tcl_SetObjResult(interp, Tcl_NewStringObj(message, -1));
+        Tcl_SetErrorCode(interp, "TCL", "CANCEL", id, message, NULL);
+    }
+
+    /*
+     * Return TCL_ERROR to the caller (not necessarily just the Tcl core
+     * itself) that indicates further processing of the script or command in
+     * progress should halt gracefully and as soon as possible.
+     */
+
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CancelEval --
+ *
+ *	This function schedules the cancellation of the current script in the
+ *	given interpreter.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR. Since the interp may belong to a different thread, no error
+ *	message can be left in the interp's result.
+ *
+ * Side effects:
+ *	The script in progress in the specified interpreter will be canceled
+ *	with TCL_ERROR after asynchronous handlers are invoked at the next
+ *	Tcl_Canceled check.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_CancelEval(
+    Tcl_Interp *interp,		/* Interpreter in which to cancel the
+				 * script. */
+    Tcl_Obj *resultObjPtr,	/* The script cancellation error message or
+				 * NULL for a default error message. */
+    ClientData clientData,	/* Passed to CancelEvalProc. */
+    int flags)			/* Collection of OR-ed bits that control
+				 * the cancellation of the script. Only
+				 * TCL_CANCEL_UNWIND is currently
+				 * supported. */
+{
+    Tcl_HashEntry *hPtr;
+    CancelInfo *cancelInfo;
+    int code = TCL_ERROR;
+    const char *result;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+
+    Tcl_MutexLock(&cancelLock);
+    if (cancelTableInitialized != 1) {
+	/*
+	 * No CancelInfo hash table (Tcl_CreateInterp has never been called?)
+	 */
+
+	goto done;
+    }
+    hPtr = Tcl_FindHashEntry(&cancelTable, (char *) interp);
+    if (hPtr == NULL) {
+	/*
+	 * No CancelInfo record for this interpreter.
+	 */
+
+	goto done;
+    }
+    cancelInfo = Tcl_GetHashValue(hPtr);
+
+    /*
+     * Populate information needed by the interpreter thread to fulfill the
+     * cancellation request. Currently, clientData is ignored. If the
+     * TCL_CANCEL_UNWIND flags bit is set, the script in progress is not
+     * allowed to catch the script cancellation because the evaluation stack
+     * for the interp is completely unwound.
+     */
+
+    if (resultObjPtr != NULL) {
+	result = TclGetStringFromObj(resultObjPtr, &cancelInfo->length);
+	cancelInfo->result = ckrealloc(cancelInfo->result,cancelInfo->length);
+	memcpy(cancelInfo->result, result, (size_t) cancelInfo->length);
+	TclDecrRefCount(resultObjPtr);	/* Discard their result object. */
+    } else {
+	cancelInfo->result = NULL;
+	cancelInfo->length = 0;
+    }
+    cancelInfo->clientData = clientData;
+    cancelInfo->flags = flags;
+    Tcl_AsyncMark(cancelInfo->async);
+    code = TCL_OK;
+
+  done:
+    Tcl_MutexUnlock(&cancelLock);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InterpActive --
+ *
+ *	Returns non-zero if the specified interpreter is in use, i.e. if there
+ *	is an evaluation currently active in the interpreter.
+ *
+ * Results:
+ *	See above.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_InterpActive(
+    Tcl_Interp *interp)
+{
+    return ((Interp *) interp)->numLevels > 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalObjv --
+ *
+ *	This function evaluates a Tcl command that has already been parsed
+ *	into words, with one Tcl_Obj holding each word.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR. A result or error message is left in interp's result.
+ *
+ * Side effects:
+ *	Always pushes a callback. Other side effects depend on the command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_EvalObjv(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate the
+				 * command. Also used for error reporting. */
+    int objc,			/* Number of words in command. */
+    Tcl_Obj *const objv[],	/* An array of pointers to objects that are
+				 * the words that make up the command. */
+    int flags)			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Only
+				 * TCL_EVAL_GLOBAL, TCL_EVAL_INVOKE and
+				 * TCL_EVAL_NOERR are currently supported. */
+{
+    int result;
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    result = TclNREvalObjv(interp, objc, objv, flags, NULL);
+    return TclNRRunCallbacks(interp, result, rootPtr);
+}
+
+int
+TclNREvalObjv(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate the
+				 * command. Also used for error reporting. */
+    int objc,			/* Number of words in command. */
+    Tcl_Obj *const objv[],	/* An array of pointers to objects that are
+				 * the words that make up the command. */
+    int flags,			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Only
+				 * TCL_EVAL_GLOBAL, TCL_EVAL_INVOKE and
+				 * TCL_EVAL_NOERR are currently supported. */
+    Command *cmdPtr)		/* NULL if the Command is to be looked up
+				 * here, otherwise the pointer to the
+				 * requested Command struct to be invoked. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    /*
+     * data[1] stores a marker for use by tailcalls; it will be set to 1 by
+     * command redirectors (imports, alias, ensembles) so that tailcall skips
+     * this callback (that marks the end of the target command) and goes back
+     * to the end of the source command.
+     */
+
+    if (iPtr->deferredCallbacks) {
+        iPtr->deferredCallbacks = NULL;
+    } else {
+	TclNRAddCallback(interp, NRCommand, NULL, NULL, NULL, NULL);
+    }
+
+    iPtr->numLevels++;
+    TclNRAddCallback(interp, EvalObjvCore, cmdPtr, INT2PTR(flags),
+	    INT2PTR(objc), objv);
+    return TCL_OK;
+}
+
+static int
+EvalObjvCore(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Command *cmdPtr = NULL, *preCmdPtr = data[0];
+    int flags = PTR2INT(data[1]);
+    int objc = PTR2INT(data[2]);
+    Tcl_Obj **objv = data[3];
+    Interp *iPtr = (Interp *) interp;
+    Namespace *lookupNsPtr = NULL;
+    int enterTracesDone = 0;
+
+    /*
+     * Push records for task to be done on return, in INVERSE order. First, if
+     * needed, the exception handlers (as they should happen last).
+     */
+
+    if (!(flags & TCL_EVAL_NOERR)) {
+	TEOV_PushExceptionHandlers(interp, objc, objv, flags);
+    }
+
+    if (TCL_OK != TclInterpReady(interp)) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 0) {
+	return TCL_OK;
+    }
+
+    if (TclLimitExceeded(iPtr->limit)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Configure evaluation context to match the requested flags.
+     */
+
+    if (iPtr->lookupNsPtr) {
+
+	/*
+	 * Capture the namespace we should do command name resolution in, as
+	 * instructed by our caller sneaking it in to us in a private interp
+	 * field.  Clear that field right away so we cannot possibly have its
+	 * use leak where it should not.  The sneaky message pass is done.
+	 *
+	 * Use of this mechanism overrides the TCL_EVAL_GLOBAL flag.
+	 * TODO: Is that a bug?
+	 */
+
+	lookupNsPtr = iPtr->lookupNsPtr;
+	iPtr->lookupNsPtr = NULL;
+    } else if (flags & TCL_EVAL_INVOKE) {
+	lookupNsPtr = iPtr->globalNsPtr;
+    } else {
+
+	/*
+	 * TCL_EVAL_INVOKE was not set: clear rewrite rules
+	 */
+
+	TclResetRewriteEnsemble(interp, 1);
+
+	if (flags & TCL_EVAL_GLOBAL) {
+	    TEOV_SwitchVarFrame(interp);
+	    lookupNsPtr = iPtr->globalNsPtr;
+	}
+    }
+
+    /*
+     * Lookup the Command to dispatch.
+     */
+
+    reresolve:
+    assert(cmdPtr == NULL);
+    if (preCmdPtr) {
+	/* Caller gave it to us */
+	if (!(preCmdPtr->flags & CMD_IS_DELETED)) {
+	    /* So long as it exists, use it. */
+	    cmdPtr = preCmdPtr;
+	} else if (flags & TCL_EVAL_NORESOLVE) {
+	    /*
+	     * When it's been deleted, and we're told not to attempt
+	     * resolving it ourselves, all we can do is raise an error.
+	     */
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "attempt to invoke a deleted command"));
+	    Tcl_SetErrorCode(interp, "TCL", "EVAL", "DELETEDCOMMAND", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    if (cmdPtr == NULL) {
+	cmdPtr = TEOV_LookupCmdFromObj(interp, objv[0], lookupNsPtr);
+	if (!cmdPtr) {
+	    return TEOV_NotFound(interp, objc, objv, lookupNsPtr);
+	}
+    }
+
+    if (enterTracesDone || iPtr->tracePtr
+	    || (cmdPtr->flags & CMD_HAS_EXEC_TRACES)) {
+
+	Tcl_Obj *commandPtr = TclGetSourceFromFrame(
+		flags & TCL_EVAL_SOURCE_IN_FRAME ?  iPtr->cmdFramePtr : NULL,
+		objc, objv);
+	Tcl_IncrRefCount(commandPtr);
+
+	if (!enterTracesDone) {
+
+	    int code = TEOV_RunEnterTraces(interp, &cmdPtr, commandPtr,
+		    objc, objv);
+
+	    /*
+	     * Send any exception from enter traces back as an exception
+	     * raised by the traced command.
+	     * TODO: Is this a bug?  Letting an execution trace BREAK or
+	     * CONTINUE or RETURN in the place of the traced command?
+	     * Would either converting all exceptions to TCL_ERROR, or
+	     * just swallowing them be better?  (Swallowing them has the
+	     * problem of permanently hiding program errors.)
+	     */
+
+	    if (code != TCL_OK) {
+		Tcl_DecrRefCount(commandPtr);
+		return code;
+	    }
+
+	    /*
+	     * If the enter traces made the resolved cmdPtr unusable, go
+	     * back and resolve again, but next time don't run enter
+	     * traces again.
+	     */
+
+	    if (cmdPtr == NULL) {
+		enterTracesDone = 1;
+		Tcl_DecrRefCount(commandPtr);
+		goto reresolve;
+	    }
+	}
+
+	/*
+	 * Schedule leave traces.  Raise the refCount on the resolved
+	 * cmdPtr, so that when it passes to the leave traces we know
+	 * it's still valid.
+	 */
+
+	cmdPtr->refCount++;
+	TclNRAddCallback(interp, TEOV_RunLeaveTraces, INT2PTR(objc),
+		    commandPtr, cmdPtr, objv);
+    }
+
+    TclNRAddCallback(interp, Dispatch,
+	    cmdPtr->nreProc ? cmdPtr->nreProc : cmdPtr->objProc,
+	    cmdPtr->objClientData, INT2PTR(objc), objv);
+    return TCL_OK;
+}
+
+static int
+Dispatch(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_ObjCmdProc *objProc = data[0];
+    ClientData clientData = data[1];
+    int objc = PTR2INT(data[2]);
+    Tcl_Obj **objv = data[3];
+    Interp *iPtr = (Interp *) interp;
+
+#ifdef USE_DTRACE
+    if (TCL_DTRACE_CMD_ARGS_ENABLED()) {
+	const char *a[10];
+	int i = 0;
+
+	while (i < 10) {
+	    a[i] = i < objc ? TclGetString(objv[i]) : NULL; i++;
+	}
+	TCL_DTRACE_CMD_ARGS(a[0], a[1], a[2], a[3], a[4], a[5], a[6], a[7],
+		a[8], a[9]);
+    }
+    if (TCL_DTRACE_CMD_INFO_ENABLED() && iPtr->cmdFramePtr) {
+	Tcl_Obj *info = TclInfoFrame(interp, iPtr->cmdFramePtr);
+	const char *a[6]; int i[2];
+
+	TclDTraceInfo(info, a, i);
+	TCL_DTRACE_CMD_INFO(a[0], a[1], a[2], a[3], i[0], i[1], a[4], a[5]);
+	TclDecrRefCount(info);
+    }
+    if ((TCL_DTRACE_CMD_RETURN_ENABLED() || TCL_DTRACE_CMD_RESULT_ENABLED())
+	    && objc) {
+	TclNRAddCallback(interp, DTraceCmdReturn, objv[0], NULL, NULL, NULL);
+    }
+    if (TCL_DTRACE_CMD_ENTRY_ENABLED() && objc) {
+	TCL_DTRACE_CMD_ENTRY(TclGetString(objv[0]), objc - 1,
+		(Tcl_Obj **)(objv + 1));
+    }
+#endif /* USE_DTRACE */
+
+    iPtr->cmdCount++;
+    return objProc(clientData, interp, objc, objv);
+}
+
+int
+TclNRRunCallbacks(
+    Tcl_Interp *interp,
+    int result,
+    struct NRE_callback *rootPtr)
+				/* All callbacks down to rootPtr not inclusive
+				 * are to be run. */
+{
+    Interp *iPtr = (Interp *) interp;
+    NRE_callback *callbackPtr;
+    Tcl_NRPostProc *procPtr;
+
+    /*
+     * If the interpreter has a non-empty string result, the result object is
+     * either empty or stale because some function set interp->result
+     * directly. If so, move the string result to the result object, then
+     * reset the string result.
+     *
+     * This only needs to be done for the first item in the list: all other
+     * are for NR function calls, and those are Tcl_Obj based.
+     */
+
+    if (*(iPtr->result) != 0) {
+	(void) Tcl_GetObjResult(interp);
+    }
+
+    while (TOP_CB(interp) != rootPtr) {
+	callbackPtr = TOP_CB(interp);
+	procPtr = callbackPtr->procPtr;
+	TOP_CB(interp) = callbackPtr->nextPtr;
+	result = procPtr(callbackPtr->data, interp, result);
+	TCLNR_FREE(interp, callbackPtr);
+    }
+    return result;
+}
+
+static int
+NRCommand(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->numLevels--;
+
+     /*
+      * If there is a tailcall, schedule it next
+      */
+
+    if (data[1] && (data[1] != INT2PTR(1))) {
+        TclNRAddCallback(interp, TclNRTailcallEval, data[1], NULL, NULL, NULL);
+    }
+
+    /* OPT ??
+     * Do not interrupt a series of cleanups with async or limit checks:
+     * just check at the end?
+     */
+
+    if (TclAsyncReady(iPtr)) {
+	result = Tcl_AsyncInvoke(interp, result);
+    }
+    if ((result == TCL_OK) && TclCanceled(iPtr)) {
+	result = Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG);
+    }
+    if (result == TCL_OK && TclLimitReady(iPtr->limit)) {
+	result = Tcl_LimitCheck(interp);
+    }
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TEOV_Exception	 -
+ * TEOV_LookupCmdFromObj -
+ * TEOV_RunEnterTraces	 -
+ * TEOV_RunLeaveTraces	 -
+ * TEOV_NotFound	 -
+ *
+ *	These are helper functions for Tcl_EvalObjv.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TEOV_PushExceptionHandlers(
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[],
+    int flags)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    /*
+     * If any error processing is necessary, push the appropriate records.
+     * Note that we have to push them in the inverse order: first the one that
+     * has to run last.
+     */
+
+    if (!(flags & TCL_EVAL_INVOKE)) {
+	/*
+	 * Error messages
+	 */
+
+	TclNRAddCallback(interp, TEOV_Error, INT2PTR(objc),
+		(ClientData) objv, NULL, NULL);
+    }
+
+    if (iPtr->numLevels == 1) {
+	/*
+	 * No CONTINUE or BREAK at level 0, manage RETURN
+	 */
+
+	TclNRAddCallback(interp, TEOV_Exception, INT2PTR(iPtr->evalFlags),
+		NULL, NULL, NULL);
+    }
+}
+
+static void
+TEOV_SwitchVarFrame(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    /*
+     * Change the varFrame to be the rootVarFrame, and push a record to
+     * restore things at the end.
+     */
+
+    TclNRAddCallback(interp, TEOV_RestoreVarFrame, iPtr->varFramePtr, NULL,
+	    NULL, NULL);
+    iPtr->varFramePtr = iPtr->rootFramePtr;
+}
+
+static int
+TEOV_RestoreVarFrame(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ((Interp *) interp)->varFramePtr = data[0];
+    return result;
+}
+
+static int
+TEOV_Exception(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    int allowExceptions = (PTR2INT(data[0]) & TCL_ALLOW_EXCEPTIONS);
+
+    if (result != TCL_OK) {
+	if (result == TCL_RETURN) {
+	    result = TclUpdateReturnInfo(iPtr);
+	}
+	if ((result != TCL_ERROR) && !allowExceptions) {
+	    ProcessUnexpectedResult(interp, result);
+	    result = TCL_ERROR;
+	}
+    }
+
+    /*
+     * We are returning to level 0, so should process TclResetCancellation. As
+     * numLevels has not *yet* been decreased, do not call it: do the thing
+     * here directly.
+     */
+
+    TclUnsetCancelFlags(iPtr);
+    return result;
+}
+
+static int
+TEOV_Error(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *listPtr;
+    const char *cmdString;
+    int cmdLen;
+    int objc = PTR2INT(data[0]);
+    Tcl_Obj **objv = data[1];
+
+    if ((result == TCL_ERROR) && !(iPtr->flags & ERR_ALREADY_LOGGED)){
+	/*
+	 * If there was an error, a command string will be needed for the
+	 * error log: get it out of the itemPtr. The details depend on the
+	 * type.
+	 */
+
+	listPtr = Tcl_NewListObj(objc, objv);
+	cmdString = TclGetStringFromObj(listPtr, &cmdLen);
+	Tcl_LogCommandInfo(interp, cmdString, cmdString, cmdLen);
+	Tcl_DecrRefCount(listPtr);
+    }
+    iPtr->flags &= ~ERR_ALREADY_LOGGED;
+    return result;
+}
+
+static int
+TEOV_NotFound(
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[],
+    Namespace *lookupNsPtr)
+{
+    Command * cmdPtr;
+    Interp *iPtr = (Interp *) interp;
+    int i, newObjc, handlerObjc;
+    Tcl_Obj **newObjv, **handlerObjv;
+    CallFrame *varFramePtr = iPtr->varFramePtr;
+    Namespace *currNsPtr = NULL;/* Used to check for and invoke any registered
+				 * unknown command handler for the current
+				 * namespace (TIP 181). */
+    Namespace *savedNsPtr = NULL;
+
+    currNsPtr = varFramePtr->nsPtr;
+    if ((currNsPtr == NULL) || (currNsPtr->unknownHandlerPtr == NULL)) {
+	currNsPtr = iPtr->globalNsPtr;
+	if (currNsPtr == NULL) {
+	    Tcl_Panic("Tcl_EvalObjv: NULL global namespace pointer");
+	}
+    }
+
+    /*
+     * Check to see if the resolution namespace has lost its unknown handler.
+     * If so, reset it to "::unknown".
+     */
+
+    if (currNsPtr->unknownHandlerPtr == NULL) {
+	TclNewLiteralStringObj(currNsPtr->unknownHandlerPtr, "::unknown");
+	Tcl_IncrRefCount(currNsPtr->unknownHandlerPtr);
+    }
+
+    /*
+     * Get the list of words for the unknown handler and allocate enough space
+     * to hold both the handler prefix and all words of the command invokation
+     * itself.
+     */
+
+    Tcl_ListObjGetElements(NULL, currNsPtr->unknownHandlerPtr,
+	    &handlerObjc, &handlerObjv);
+    newObjc = objc + handlerObjc;
+    newObjv = TclStackAlloc(interp, (int) sizeof(Tcl_Obj *) * newObjc);
+
+    /*
+     * Copy command prefix from unknown handler and add on the real command's
+     * full argument list. Note that we only use memcpy() once because we have
+     * to increment the reference count of all the handler arguments anyway.
+     */
+
+    for (i = 0; i < handlerObjc; ++i) {
+	newObjv[i] = handlerObjv[i];
+	Tcl_IncrRefCount(newObjv[i]);
+    }
+    memcpy(newObjv+handlerObjc, objv, sizeof(Tcl_Obj *) * (unsigned)objc);
+
+    /*
+     * Look up and invoke the handler (by recursive call to this function). If
+     * there is no handler at all, instead of doing the recursive call we just
+     * generate a generic error message; it would be an infinite-recursion
+     * nightmare otherwise.
+     *
+     * In this case we worry a bit less about recursion for now, and call the
+     * "blocking" interface.
+     */
+
+    cmdPtr = TEOV_LookupCmdFromObj(interp, newObjv[0], lookupNsPtr);
+    if (cmdPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "invalid command name \"%s\"", TclGetString(objv[0])));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND",
+                TclGetString(objv[0]), NULL);
+
+	/*
+	 * Release any resources we locked and allocated during the handler
+	 * call.
+	 */
+
+	for (i = 0; i < handlerObjc; ++i) {
+	    Tcl_DecrRefCount(newObjv[i]);
+	}
+	TclStackFree(interp, newObjv);
+	return TCL_ERROR;
+    }
+
+    if (lookupNsPtr) {
+	savedNsPtr = varFramePtr->nsPtr;
+	varFramePtr->nsPtr = lookupNsPtr;
+    }
+    TclSkipTailcall(interp);
+    TclNRAddCallback(interp, TEOV_NotFoundCallback, INT2PTR(handlerObjc),
+	    newObjv, savedNsPtr, NULL);
+    return TclNREvalObjv(interp, newObjc, newObjv, TCL_EVAL_NOERR, NULL);
+}
+
+static int
+TEOV_NotFoundCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    int objc = PTR2INT(data[0]);
+    Tcl_Obj **objv = data[1];
+    Namespace *savedNsPtr = data[2];
+
+    int i;
+
+    if (savedNsPtr) {
+	iPtr->varFramePtr->nsPtr = savedNsPtr;
+    }
+
+    /*
+     * Release any resources we locked and allocated during the handler call.
+     */
+
+    for (i = 0; i < objc; ++i) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    TclStackFree(interp, objv);
+
+    return result;
+}
+
+static int
+TEOV_RunEnterTraces(
+    Tcl_Interp *interp,
+    Command **cmdPtrPtr,
+    Tcl_Obj *commandPtr,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Command *cmdPtr = *cmdPtrPtr;
+    size_t newEpoch, cmdEpoch = cmdPtr->cmdEpoch;
+    int length, traceCode = TCL_OK;
+    const char *command = TclGetStringFromObj(commandPtr, &length);
+
+    /*
+     * Call trace functions.
+     * Execute any command or execution traces. Note that we bump up the
+     * command's reference count for the duration of the calling of the
+     * traces so that the structure doesn't go away underneath our feet.
+     */
+
+    cmdPtr->refCount++;
+    if (iPtr->tracePtr) {
+	traceCode = TclCheckInterpTraces(interp, command, length,
+		cmdPtr, TCL_OK, TCL_TRACE_ENTER_EXEC, objc, objv);
+    }
+    if ((cmdPtr->flags & CMD_HAS_EXEC_TRACES) && (traceCode == TCL_OK)) {
+	traceCode = TclCheckExecutionTraces(interp, command, length,
+		cmdPtr, TCL_OK, TCL_TRACE_ENTER_EXEC, objc, objv);
+    }
+    newEpoch = cmdPtr->cmdEpoch;
+    TclCleanupCommandMacro(cmdPtr);
+
+    if (traceCode != TCL_OK) {
+	if (traceCode == TCL_ERROR) {
+	    Tcl_Obj *info;
+
+	    TclNewLiteralStringObj(info, "\n    (enter trace on \"");
+	    Tcl_AppendLimitedToObj(info, command, length, 55, "...");
+	    Tcl_AppendToObj(info, "\")", 2);
+	    Tcl_AppendObjToErrorInfo(interp, info);
+	    iPtr->flags |= ERR_ALREADY_LOGGED;
+	}
+	return traceCode;
+    }
+    if (cmdEpoch != newEpoch) {
+	*cmdPtrPtr = NULL;
+    }
+    return TCL_OK;
+}
+
+static int
+TEOV_RunLeaveTraces(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    int traceCode = TCL_OK;
+    int objc = PTR2INT(data[0]);
+    Tcl_Obj *commandPtr = data[1];
+    Command *cmdPtr = data[2];
+    Tcl_Obj **objv = data[3];
+    int length;
+    const char *command = TclGetStringFromObj(commandPtr, &length);
+
+    if (!(cmdPtr->flags & CMD_IS_DELETED)) {
+	if (cmdPtr->flags & CMD_HAS_EXEC_TRACES){
+	    traceCode = TclCheckExecutionTraces(interp, command, length,
+		    cmdPtr, result, TCL_TRACE_LEAVE_EXEC, objc, objv);
+	}
+	if (iPtr->tracePtr != NULL && traceCode == TCL_OK) {
+	    traceCode = TclCheckInterpTraces(interp, command, length,
+		    cmdPtr, result, TCL_TRACE_LEAVE_EXEC, objc, objv);
+	}
+    }
+
+    /*
+     * As cmdPtr is set, TclNRRunCallbacks is about to reduce the numlevels.
+     * Prevent that by resetting the cmdPtr field and dealing right here with
+     * cmdPtr->refCount.
+     */
+
+    TclCleanupCommandMacro(cmdPtr);
+
+    if (traceCode != TCL_OK) {
+	if (traceCode == TCL_ERROR) {
+	    Tcl_Obj *info;
+
+	    TclNewLiteralStringObj(info, "\n    (leave trace on \"");
+	    Tcl_AppendLimitedToObj(info, command, length, 55, "...");
+	    Tcl_AppendToObj(info, "\")", 2);
+	    Tcl_AppendObjToErrorInfo(interp, info);
+	    iPtr->flags |= ERR_ALREADY_LOGGED;
+	}
+	result = traceCode;
+    }
+    Tcl_DecrRefCount(commandPtr);
+    return result;
+}
+
+static inline Command *
+TEOV_LookupCmdFromObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *namePtr,
+    Namespace *lookupNsPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    Command *cmdPtr;
+    Namespace *savedNsPtr = iPtr->varFramePtr->nsPtr;
+
+    if (lookupNsPtr) {
+	iPtr->varFramePtr->nsPtr = lookupNsPtr;
+    }
+    cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, namePtr);
+    iPtr->varFramePtr->nsPtr = savedNsPtr;
+    return cmdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalTokensStandard --
+ *
+ *	Given an array of tokens parsed from a Tcl command (e.g., the tokens
+ *	that make up a word or the index for an array variable) this function
+ *	evaluates the tokens and concatenates their values to form a single
+ *	result value.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR. A result or error message is left in interp's result.
+ *
+ * Side effects:
+ *	Depends on the array of tokens being evaled.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_EvalTokensStandard(
+    Tcl_Interp *interp,		/* Interpreter in which to lookup variables,
+				 * execute nested commands, and report
+				 * errors. */
+    Tcl_Token *tokenPtr,	/* Pointer to first in an array of tokens to
+				 * evaluate and concatenate. */
+    int count)			/* Number of tokens to consider at tokenPtr.
+				 * Must be at least 1. */
+{
+    return TclSubstTokens(interp, tokenPtr, count, /* numLeftPtr */ NULL, 1,
+	    NULL, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalTokens --
+ *
+ *	Given an array of tokens parsed from a Tcl command (e.g., the tokens
+ *	that make up a word or the index for an array variable) this function
+ *	evaluates the tokens and concatenates their values to form a single
+ *	result value.
+ *
+ * Results:
+ *	The return value is a pointer to a newly allocated Tcl_Obj containing
+ *	the value of the array of tokens. The reference count of the returned
+ *	object has been incremented. If an error occurs in evaluating the
+ *	tokens then a NULL value is returned and an error message is left in
+ *	interp's result.
+ *
+ * Side effects:
+ *	A new object is allocated to hold the result.
+ *
+ *----------------------------------------------------------------------
+ *
+ * This uses a non-standard return convention; its use is now deprecated. It
+ * is a wrapper for the new function Tcl_EvalTokensStandard, and is not used
+ * in the core any longer. It is only kept for backward compatibility.
+ */
+
+Tcl_Obj *
+Tcl_EvalTokens(
+    Tcl_Interp *interp,		/* Interpreter in which to lookup variables,
+				 * execute nested commands, and report
+				 * errors. */
+    Tcl_Token *tokenPtr,	/* Pointer to first in an array of tokens to
+				 * evaluate and concatenate. */
+    int count)			/* Number of tokens to consider at tokenPtr.
+				 * Must be at least 1. */
+{
+    Tcl_Obj *resPtr;
+
+    if (Tcl_EvalTokensStandard(interp, tokenPtr, count) != TCL_OK) {
+	return NULL;
+    }
+    resPtr = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(resPtr);
+    Tcl_ResetResult(interp);
+    return resPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalEx, TclEvalEx --
+ *
+ *	This function evaluates a Tcl script without using the compiler or
+ *	byte-code interpreter. It just parses the script, creates values for
+ *	each word of each command, then calls EvalObjv to execute each
+ *	command.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR. A result or error message is left in interp's result.
+ *
+ * Side effects:
+ *	Depends on the script.
+ *
+ * TIP #280 : Keep public API, internally extended API.
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_EvalEx(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate the
+				 * script. Also used for error reporting. */
+    const char *script,		/* First character of script to evaluate. */
+    int numBytes,		/* Number of bytes in script. If < 0, the
+				 * script consists of all bytes up to the
+				 * first null character. */
+    int flags)			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Only
+				 * TCL_EVAL_GLOBAL is currently supported. */
+{
+    return TclEvalEx(interp, script, numBytes, flags, 1, NULL, script);
+}
+
+int
+TclEvalEx(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate the
+				 * script. Also used for error reporting. */
+    const char *script,		/* First character of script to evaluate. */
+    int numBytes,		/* Number of bytes in script. If < 0, the
+				 * script consists of all bytes up to the
+				 * first NUL character. */
+    int flags,			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Only
+				 * TCL_EVAL_GLOBAL is currently supported. */
+    int line,			/* The line the script starts on. */
+    int *clNextOuter,		/* Information about an outer context for */
+    const char *outerScript)	/* continuation line data. This is set only in
+				 * TclSubstTokens(), to properly handle
+				 * [...]-nested commands. The 'outerScript'
+				 * refers to the most-outer script containing
+				 * the embedded command, which is refered to
+				 * by 'script'. The 'clNextOuter' refers to
+				 * the current entry in the table of
+				 * continuation lines in this "master script",
+				 * and the character offsets are relative to
+				 * the 'outerScript' as well.
+				 *
+				 * If outerScript == script, then this call is
+				 * for the outer-most script/command. See
+				 * Tcl_EvalEx() and TclEvalObjEx() for places
+				 * generating arguments for which this is
+				 * true. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *p, *next;
+    const unsigned int minObjs = 20;
+    Tcl_Obj **objv, **objvSpace;
+    int *expand, *lines, *lineSpace;
+    Tcl_Token *tokenPtr;
+    int commandLength, bytesLeft, expandRequested, code = TCL_OK;
+    CallFrame *savedVarFramePtr;/* Saves old copy of iPtr->varFramePtr in case
+				 * TCL_EVAL_GLOBAL was set. */
+    int allowExceptions = (iPtr->evalFlags & TCL_ALLOW_EXCEPTIONS);
+    int gotParse = 0;
+    unsigned int i, objectsUsed = 0;
+				/* These variables keep track of how much
+				 * state has been allocated while evaluating
+				 * the script, so that it can be freed
+				 * properly if an error occurs. */
+    Tcl_Parse *parsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse));
+    CmdFrame *eeFramePtr = TclStackAlloc(interp, sizeof(CmdFrame));
+    Tcl_Obj **stackObjArray =
+	    TclStackAlloc(interp, minObjs * sizeof(Tcl_Obj *));
+    int *expandStack = TclStackAlloc(interp, minObjs * sizeof(int));
+    int *linesStack = TclStackAlloc(interp, minObjs * sizeof(int));
+				/* TIP #280 Structures for tracking of command
+				 * locations. */
+    int *clNext = NULL;		/* Pointer for the tracking of invisible
+				 * continuation lines. Initialized only if the
+				 * caller gave us a table of locations to
+				 * track, via scriptCLLocPtr. It always refers
+				 * to the table entry holding the location of
+				 * the next invisible continuation line to
+				 * look for, while parsing the script. */
+
+    if (iPtr->scriptCLLocPtr) {
+	if (clNextOuter) {
+	    clNext = clNextOuter;
+	} else {
+	    clNext = &iPtr->scriptCLLocPtr->loc[0];
+	}
+    }
+
+    if (numBytes < 0) {
+	numBytes = strlen(script);
+    }
+    Tcl_ResetResult(interp);
+
+    savedVarFramePtr = iPtr->varFramePtr;
+    if (flags & TCL_EVAL_GLOBAL) {
+	iPtr->varFramePtr = iPtr->rootFramePtr;
+    }
+
+    /*
+     * Each iteration through the following loop parses the next command from
+     * the script and then executes it.
+     */
+
+    objv = objvSpace = stackObjArray;
+    lines = lineSpace = linesStack;
+    expand = expandStack;
+    p = script;
+    bytesLeft = numBytes;
+
+    /*
+     * TIP #280 Initialize tracking. Do not push on the frame stack yet.
+     *
+     * We open a new context, either for a sourced script, or 'eval'.
+     * For sourced files we always have a path object, even if nothing was
+     * specified in the interp itself. That makes code using it simpler as
+     * NULL checks can be left out. Sourced file without path in the
+     * 'scriptFile' is possible during Tcl initialization.
+     */
+
+    eeFramePtr->level = iPtr->cmdFramePtr ? iPtr->cmdFramePtr->level + 1 : 1;
+    eeFramePtr->framePtr = iPtr->framePtr;
+    eeFramePtr->nextPtr = iPtr->cmdFramePtr;
+    eeFramePtr->nline = 0;
+    eeFramePtr->line = NULL;
+    eeFramePtr->cmdObj = NULL;
+
+    iPtr->cmdFramePtr = eeFramePtr;
+    if (iPtr->evalFlags & TCL_EVAL_FILE) {
+	/*
+	 * Set up for a sourced file.
+	 */
+
+	eeFramePtr->type = TCL_LOCATION_SOURCE;
+
+	if (iPtr->scriptFile) {
+	    /*
+	     * Normalization here, to have the correct pwd. Should have
+	     * negligible impact on performance, as the norm should have been
+	     * done already by the 'source' invoking us, and it caches the
+	     * result.
+	     */
+
+	    Tcl_Obj *norm = Tcl_FSGetNormalizedPath(interp, iPtr->scriptFile);
+
+	    if (norm == NULL) {
+		/*
+		 * Error message in the interp result.
+		 */
+
+		code = TCL_ERROR;
+		goto error;
+	    }
+	    eeFramePtr->data.eval.path = norm;
+	} else {
+	    TclNewLiteralStringObj(eeFramePtr->data.eval.path, "");
+	}
+	Tcl_IncrRefCount(eeFramePtr->data.eval.path);
+    } else {
+	/*
+	 * Set up for plain eval.
+	 */
+
+	eeFramePtr->type = TCL_LOCATION_EVAL;
+	eeFramePtr->data.eval.path = NULL;
+    }
+
+    iPtr->evalFlags = 0;
+    do {
+	if (Tcl_ParseCommand(interp, p, bytesLeft, 0, parsePtr) != TCL_OK) {
+	    code = TCL_ERROR;
+	    Tcl_LogCommandInfo(interp, script, parsePtr->commandStart,
+		    parsePtr->term + 1 - parsePtr->commandStart);
+	    goto posterror;
+	}
+
+	/*
+	 * TIP #280 Track lines. The parser may have skipped text till it
+	 * found the command we are now at. We have to count the lines in this
+	 * block, and do not forget invisible continuation lines.
+	 */
+
+	TclAdvanceLines(&line, p, parsePtr->commandStart);
+	TclAdvanceContinuations(&line, &clNext,
+		parsePtr->commandStart - outerScript);
+
+	gotParse = 1;
+	if (parsePtr->numWords > 0) {
+	    /*
+	     * TIP #280. Track lines within the words of the current
+	     * command. We use a separate pointer into the table of
+	     * continuation line locations to not lose our position for the
+	     * per-command parsing.
+	     */
+
+	    int wordLine = line;
+	    const char *wordStart = parsePtr->commandStart;
+	    int *wordCLNext = clNext;
+	    unsigned int objectsNeeded = 0;
+	    unsigned int numWords = parsePtr->numWords;
+
+	    /*
+	     * Generate an array of objects for the words of the command.
+	     */
+
+	    if (numWords > minObjs) {
+		expand =    ckalloc(numWords * sizeof(int));
+		objvSpace = ckalloc(numWords * sizeof(Tcl_Obj *));
+		lineSpace = ckalloc(numWords * sizeof(int));
+	    }
+	    expandRequested = 0;
+	    objv = objvSpace;
+	    lines = lineSpace;
+
+	    iPtr->cmdFramePtr = eeFramePtr->nextPtr;
+	    for (objectsUsed = 0, tokenPtr = parsePtr->tokenPtr;
+		    objectsUsed < numWords;
+		    objectsUsed++, tokenPtr += tokenPtr->numComponents+1) {
+		/*
+		 * TIP #280. Track lines to current word. Save the information
+		 * on a per-word basis, signaling dynamic words as needed.
+		 * Make the information available to the recursively called
+		 * evaluator as well, including the type of context (source
+		 * vs. eval).
+		 */
+
+		TclAdvanceLines(&wordLine, wordStart, tokenPtr->start);
+		TclAdvanceContinuations(&wordLine, &wordCLNext,
+			tokenPtr->start - outerScript);
+		wordStart = tokenPtr->start;
+
+		lines[objectsUsed] = TclWordKnownAtCompileTime(tokenPtr, NULL)
+			? wordLine : -1;
+
+		if (eeFramePtr->type == TCL_LOCATION_SOURCE) {
+		    iPtr->evalFlags |= TCL_EVAL_FILE;
+		}
+
+		code = TclSubstTokens(interp, tokenPtr+1,
+			tokenPtr->numComponents, NULL, wordLine,
+			wordCLNext, outerScript);
+
+		iPtr->evalFlags = 0;
+
+		if (code != TCL_OK) {
+		    break;
+		}
+		objv[objectsUsed] = Tcl_GetObjResult(interp);
+		Tcl_IncrRefCount(objv[objectsUsed]);
+		if (tokenPtr->type == TCL_TOKEN_EXPAND_WORD) {
+		    int numElements;
+
+		    code = TclListObjLength(interp, objv[objectsUsed],
+			    &numElements);
+		    if (code == TCL_ERROR) {
+			/*
+			 * Attempt to expand a non-list.
+			 */
+
+			Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+				"\n    (expanding word %d)", objectsUsed));
+			Tcl_DecrRefCount(objv[objectsUsed]);
+			break;
+		    }
+		    expandRequested = 1;
+		    expand[objectsUsed] = 1;
+
+		    objectsNeeded += (numElements ? numElements : 1);
+		} else {
+		    expand[objectsUsed] = 0;
+		    objectsNeeded++;
+		}
+
+		if (wordCLNext) {
+		    TclContinuationsEnterDerived(objv[objectsUsed],
+			    wordStart - outerScript, wordCLNext);
+		}
+	    } /* for loop */
+	    iPtr->cmdFramePtr = eeFramePtr;
+	    if (code != TCL_OK) {
+		goto error;
+	    }
+	    if (expandRequested) {
+		/*
+		 * Some word expansion was requested. Check for objv resize.
+		 */
+
+		Tcl_Obj **copy = objvSpace;
+		int *lcopy = lineSpace;
+		int wordIdx = numWords;
+		int objIdx = objectsNeeded - 1;
+
+		if ((numWords > minObjs) || (objectsNeeded > minObjs)) {
+		    objv = objvSpace =
+			    ckalloc(objectsNeeded * sizeof(Tcl_Obj *));
+		    lines = lineSpace = ckalloc(objectsNeeded * sizeof(int));
+		}
+
+		objectsUsed = 0;
+		while (wordIdx--) {
+		    if (expand[wordIdx]) {
+			int numElements;
+			Tcl_Obj **elements, *temp = copy[wordIdx];
+
+			Tcl_ListObjGetElements(NULL, temp, &numElements,
+				&elements);
+			objectsUsed += numElements;
+			while (numElements--) {
+			    lines[objIdx] = -1;
+			    objv[objIdx--] = elements[numElements];
+			    Tcl_IncrRefCount(elements[numElements]);
+			}
+			Tcl_DecrRefCount(temp);
+		    } else {
+			lines[objIdx] = lcopy[wordIdx];
+			objv[objIdx--] = copy[wordIdx];
+			objectsUsed++;
+		    }
+		}
+		objv += objIdx+1;
+
+		if (copy != stackObjArray) {
+		    ckfree(copy);
+		}
+		if (lcopy != linesStack) {
+		    ckfree(lcopy);
+		}
+	    }
+
+	    /*
+	     * Execute the command and free the objects for its words.
+	     *
+	     * TIP #280: Remember the command itself for 'info frame'. We
+	     * shorten the visible command by one char to exclude the
+	     * termination character, if necessary. Here is where we put our
+	     * frame on the stack of frames too. _After_ the nested commands
+	     * have been executed.
+	     */
+
+	    eeFramePtr->cmd = parsePtr->commandStart;
+	    eeFramePtr->len = parsePtr->commandSize;
+
+	    if (parsePtr->term ==
+		    parsePtr->commandStart + parsePtr->commandSize - 1) {
+		eeFramePtr->len--;
+	    }
+
+	    eeFramePtr->nline = objectsUsed;
+	    eeFramePtr->line = lines;
+
+	    TclArgumentEnter(interp, objv, objectsUsed, eeFramePtr);
+	    code = Tcl_EvalObjv(interp, objectsUsed, objv,
+		    TCL_EVAL_NOERR | TCL_EVAL_SOURCE_IN_FRAME);
+	    TclArgumentRelease(interp, objv, objectsUsed);
+
+	    eeFramePtr->line = NULL;
+	    eeFramePtr->nline = 0;
+	    if (eeFramePtr->cmdObj) {
+		Tcl_DecrRefCount(eeFramePtr->cmdObj);
+		eeFramePtr->cmdObj = NULL;
+	    }
+
+	    if (code != TCL_OK) {
+		goto error;
+	    }
+	    for (i = 0; i < objectsUsed; i++) {
+		Tcl_DecrRefCount(objv[i]);
+	    }
+	    objectsUsed = 0;
+	    if (objvSpace != stackObjArray) {
+		ckfree(objvSpace);
+		objvSpace = stackObjArray;
+		ckfree(lineSpace);
+		lineSpace = linesStack;
+	    }
+
+	    /*
+	     * Free expand separately since objvSpace could have been
+	     * reallocated above.
+	     */
+
+	    if (expand != expandStack) {
+		ckfree(expand);
+		expand = expandStack;
+	    }
+	}
+
+	/*
+	 * Advance to the next command in the script.
+	 *
+	 * TIP #280 Track Lines. Now we track how many lines were in the
+	 * executed command.
+	 */
+
+	next = parsePtr->commandStart + parsePtr->commandSize;
+	bytesLeft -= next - p;
+	p = next;
+	TclAdvanceLines(&line, parsePtr->commandStart, p);
+	Tcl_FreeParse(parsePtr);
+	gotParse = 0;
+    } while (bytesLeft > 0);
+    iPtr->varFramePtr = savedVarFramePtr;
+    code = TCL_OK;
+    goto cleanup_return;
+
+  error:
+    /*
+     * Generate and log various pieces of error information.
+     */
+
+    if (iPtr->numLevels == 0) {
+	if (code == TCL_RETURN) {
+	    code = TclUpdateReturnInfo(iPtr);
+	}
+	if ((code != TCL_OK) && (code != TCL_ERROR) && !allowExceptions) {
+	    ProcessUnexpectedResult(interp, code);
+	    code = TCL_ERROR;
+	}
+    }
+    if ((code == TCL_ERROR) && !(iPtr->flags & ERR_ALREADY_LOGGED)) {
+	commandLength = parsePtr->commandSize;
+	if (parsePtr->term == parsePtr->commandStart + commandLength - 1) {
+	    /*
+	     * The terminator character (such as ; or ]) of the command where
+	     * the error occurred is the last character in the parsed command.
+	     * Reduce the length by one so that the error message doesn't
+	     * include the terminator character.
+	     */
+
+	    commandLength -= 1;
+	}
+	Tcl_LogCommandInfo(interp, script, parsePtr->commandStart,
+		commandLength);
+    }
+ posterror:
+    iPtr->flags &= ~ERR_ALREADY_LOGGED;
+
+    /*
+     * Then free resources that had been allocated to the command.
+     */
+
+    for (i = 0; i < objectsUsed; i++) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    if (gotParse) {
+	Tcl_FreeParse(parsePtr);
+    }
+    if (objvSpace != stackObjArray) {
+	ckfree(objvSpace);
+	ckfree(lineSpace);
+    }
+    if (expand != expandStack) {
+	ckfree(expand);
+    }
+    iPtr->varFramePtr = savedVarFramePtr;
+
+ cleanup_return:
+    /*
+     * TIP #280. Release the local CmdFrame, and its contents.
+     */
+
+    iPtr->cmdFramePtr = iPtr->cmdFramePtr->nextPtr;
+    if (eeFramePtr->type == TCL_LOCATION_SOURCE) {
+	Tcl_DecrRefCount(eeFramePtr->data.eval.path);
+    }
+    TclStackFree(interp, linesStack);
+    TclStackFree(interp, expandStack);
+    TclStackFree(interp, stackObjArray);
+    TclStackFree(interp, eeFramePtr);
+    TclStackFree(interp, parsePtr);
+
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclAdvanceLines --
+ *
+ *	This function is a helper which counts the number of lines in a block
+ *	of text and advances an external counter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The specified counter is advanced per the number of lines found.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclAdvanceLines(
+    int *line,
+    const char *start,
+    const char *end)
+{
+    register const char *p;
+
+    for (p = start; p < end; p++) {
+	if (*p == '\n') {
+	    (*line)++;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclAdvanceContinuations --
+ *
+ *	This procedure is a helper which counts the number of continuation
+ *	lines (CL) in a block of text using a table of CL locations and
+ *	advances an external counter, and the pointer into the table.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The specified counter is advanced per the number of continuation lines
+ *	found.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclAdvanceContinuations(
+    int *line,
+    int **clNextPtrPtr,
+    int loc)
+{
+    /*
+     * Track the invisible continuation lines embedded in a script, if any.
+     * Here they are just spaces (already). They were removed by
+     * TclSubstTokens via TclParseBackslash.
+     *
+     * *clNextPtrPtr         <=> We have continuation lines to track.
+     * **clNextPtrPtr >= 0   <=> We are not beyond the last possible location.
+     * loc >= **clNextPtrPtr <=> We stepped beyond the current cont. line.
+     */
+
+    while (*clNextPtrPtr && (**clNextPtrPtr >= 0)
+	    && (loc >= **clNextPtrPtr)) {
+	/*
+	 * We just stepped over an invisible continuation line. Adjust the
+	 * line counter and step to the table entry holding the location of
+	 * the next continuation line to track.
+	 */
+
+	(*line)++;
+	(*clNextPtrPtr)++;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ * Note: The whole data structure access for argument location tracking is
+ * hidden behind these three functions. The only parts open are the lineLAPtr
+ * field in the Interp structure. The CFWord definition is internal to here.
+ * Should make it easier to redo the data structures if we find something more
+ * space/time efficient.
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArgumentEnter --
+ *
+ *	This procedure is a helper for the TIP #280 uplevel extension. It
+ *	enters location references for the arguments of a command to be
+ *	invoked. Only the first entry has the actual data, further entries
+ *	simply count the usage up.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May allocate memory.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclArgumentEnter(
+    Tcl_Interp *interp,
+    Tcl_Obj **objv,
+    int objc,
+    CmdFrame *cfPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    int new, i;
+    Tcl_HashEntry *hPtr;
+    CFWord *cfwPtr;
+
+    for (i = 1; i < objc; i++) {
+	/*
+	 * Ignore argument words without line information (= dynamic). If they
+	 * are variables they may have location information associated with
+	 * that, either through globally recorded 'set' invokations, or
+	 * literals in bytecode. Eitehr way there is no need to record
+	 * something here.
+	 */
+
+	if (cfPtr->line[i] < 0) {
+	    continue;
+	}
+	hPtr = Tcl_CreateHashEntry(iPtr->lineLAPtr, objv[i], &new);
+	if (new) {
+	    /*
+	     * The word is not on the stack yet, remember the current location
+	     * and initialize references.
+	     */
+
+	    cfwPtr = ckalloc(sizeof(CFWord));
+	    cfwPtr->framePtr = cfPtr;
+	    cfwPtr->word = i;
+	    cfwPtr->refCount = 1;
+	    Tcl_SetHashValue(hPtr, cfwPtr);
+	} else {
+	    /*
+	     * The word is already on the stack, its current location is not
+	     * relevant. Just remember the reference to prevent early removal.
+	     */
+
+	    cfwPtr = Tcl_GetHashValue(hPtr);
+	    cfwPtr->refCount++;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArgumentRelease --
+ *
+ *	This procedure is a helper for the TIP #280 uplevel extension. It
+ *	removes the location references for the arguments of a command just
+ *	done. Usage is counted down, the data is removed only when no user is
+ *	left over.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May release memory.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclArgumentRelease(
+    Tcl_Interp *interp,
+    Tcl_Obj **objv,
+    int objc)
+{
+    Interp *iPtr = (Interp *) interp;
+    int i;
+
+    for (i = 1; i < objc; i++) {
+	CFWord *cfwPtr;
+	Tcl_HashEntry *hPtr =
+		Tcl_FindHashEntry(iPtr->lineLAPtr, (char *) objv[i]);
+
+	if (!hPtr) {
+	    continue;
+	}
+	cfwPtr = Tcl_GetHashValue(hPtr);
+
+	if (cfwPtr->refCount-- > 1) {
+	    continue;
+	}
+
+	ckfree(cfwPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArgumentBCEnter --
+ *
+ *	This procedure is a helper for the TIP #280 uplevel extension. It
+ *	enters location references for the literal arguments of commands in
+ *	bytecode about to be invoked. Only the first entry has the actual
+ *	data, further entries simply count the usage up.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May allocate memory.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclArgumentBCEnter(
+    Tcl_Interp *interp,
+    Tcl_Obj *objv[],
+    int objc,
+    void *codePtr,
+    CmdFrame *cfPtr,
+    int cmd,
+    int pc)
+{
+    ExtCmdLoc *eclPtr;
+    int word;
+    ECL *ePtr;
+    CFWordBC *lastPtr = NULL;
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hePtr =
+	    Tcl_FindHashEntry(iPtr->lineBCPtr, (char *) codePtr);
+
+    if (!hePtr) {
+	return;
+    }
+    eclPtr = Tcl_GetHashValue(hePtr);
+    ePtr = &eclPtr->loc[cmd];
+
+    /*
+     * ePtr->nline is the number of words originally parsed.
+     *
+     * objc is the number of elements getting invoked.
+     *
+     * If they are not the same, we arrived here by compiling an
+     * ensemble dispatch.  Ensemble subcommands that lead to script
+     * evaluation are not supposed to get compiled, because a command
+     * such as [info level] in the script can expose some of the dispatch
+     * shenanigans.  This means that we don't have to tend to the
+     * housekeeping, and can escape now.
+     */
+
+    if (ePtr->nline != objc) {
+        return;
+    }
+
+    /*
+     * Having disposed of the ensemble cases, we can state...
+     * A few truths ...
+     * (1) ePtr->nline == objc
+     * (2) (ePtr->line[word] < 0) => !literal, for all words
+     * (3) (word == 0) => !literal
+     *
+     * Item (2) is why we can use objv to get the literals, and do not
+     * have to save them at compile time.
+     */
+
+    for (word = 1; word < objc; word++) {
+	if (ePtr->line[word] >= 0) {
+	    int isnew;
+	    Tcl_HashEntry *hPtr = Tcl_CreateHashEntry(iPtr->lineLABCPtr,
+		objv[word], &isnew);
+	    CFWordBC *cfwPtr = ckalloc(sizeof(CFWordBC));
+
+	    cfwPtr->framePtr = cfPtr;
+	    cfwPtr->obj = objv[word];
+	    cfwPtr->pc = pc;
+	    cfwPtr->word = word;
+	    cfwPtr->nextPtr = lastPtr;
+	    lastPtr = cfwPtr;
+
+	    if (isnew) {
+		/*
+		 * The word is not on the stack yet, remember the current
+		 * location and initialize references.
+		 */
+
+		cfwPtr->prevPtr = NULL;
+	    } else {
+		/*
+		 * The object is already on the stack, however it may have
+		 * a different location now (literal sharing may map
+		 * multiple location to a single Tcl_Obj*. Save the old
+		 * information in the new structure.
+		 */
+
+		cfwPtr->prevPtr = Tcl_GetHashValue(hPtr);
+	    }
+
+	    Tcl_SetHashValue(hPtr, cfwPtr);
+	}
+    } /* for */
+
+    cfPtr->litarg = lastPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArgumentBCRelease --
+ *
+ *	This procedure is a helper for the TIP #280 uplevel extension. It
+ *	removes the location references for the literal arguments of commands
+ *	in bytecode just done. Usage is counted down, the data is removed only
+ *	when no user is left over.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May release memory.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclArgumentBCRelease(
+    Tcl_Interp *interp,
+    CmdFrame *cfPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    CFWordBC *cfwPtr = (CFWordBC *) cfPtr->litarg;
+
+    while (cfwPtr) {
+	CFWordBC *nextPtr = cfwPtr->nextPtr;
+	Tcl_HashEntry *hPtr =
+		Tcl_FindHashEntry(iPtr->lineLABCPtr, (char *) cfwPtr->obj);
+	CFWordBC *xPtr = Tcl_GetHashValue(hPtr);
+
+	if (xPtr != cfwPtr) {
+	    Tcl_Panic("TclArgumentBC Enter/Release Mismatch");
+	}
+
+	if (cfwPtr->prevPtr) {
+	    Tcl_SetHashValue(hPtr, cfwPtr->prevPtr);
+	} else {
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+
+	ckfree(cfwPtr);
+	cfwPtr = nextPtr;
+    }
+
+    cfPtr->litarg = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArgumentGet --
+ *
+ *	This procedure is a helper for the TIP #280 uplevel extension. It
+ *	finds the location references for a Tcl_Obj, if any.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Writes found location information into the result arguments.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclArgumentGet(
+    Tcl_Interp *interp,
+    Tcl_Obj *obj,
+    CmdFrame **cfPtrPtr,
+    int *wordPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hPtr;
+    CmdFrame *framePtr;
+
+    /*
+     * An object which either has no string rep or else is a canonical list is
+     * guaranteed to have been generated dynamically: bail out, this cannot
+     * have a usable absolute location. _Do not touch_ the information the set
+     * up by the caller. It knows better than us.
+     */
+
+    if ((obj->bytes == NULL) || TclListObjIsCanonical(obj)) {
+	return;
+    }
+
+    /*
+     * First look for location information recorded in the argument
+     * stack. That is nearest.
+     */
+
+    hPtr = Tcl_FindHashEntry(iPtr->lineLAPtr, (char *) obj);
+    if (hPtr) {
+	CFWord *cfwPtr = Tcl_GetHashValue(hPtr);
+
+	*wordPtr = cfwPtr->word;
+	*cfPtrPtr = cfwPtr->framePtr;
+	return;
+    }
+
+    /*
+     * Check if the Tcl_Obj has location information as a bytecode literal, in
+     * that stack.
+     */
+
+    hPtr = Tcl_FindHashEntry(iPtr->lineLABCPtr, (char *) obj);
+    if (hPtr) {
+	CFWordBC *cfwPtr = Tcl_GetHashValue(hPtr);
+
+	framePtr = cfwPtr->framePtr;
+	framePtr->data.tebc.pc = (char *) (((ByteCode *)
+		framePtr->data.tebc.codePtr)->codeStart + cfwPtr->pc);
+	*cfPtrPtr = cfwPtr->framePtr;
+	*wordPtr = cfwPtr->word;
+	return;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Eval --
+ *
+ *	Execute a Tcl command in a string. This function executes the script
+ *	directly, rather than compiling it to bytecodes. Before the arrival of
+ *	the bytecode compiler in Tcl 8.0 Tcl_Eval was the main function used
+ *	for executing Tcl commands, but nowadays it isn't used much.
+ *
+ * Results:
+ *	The return value is one of the return codes defined in tcl.h (such as
+ *	TCL_OK), and interp's result contains a value to supplement the return
+ *	code. The value of the result will persist only until the next call to
+ *	Tcl_Eval or Tcl_EvalObj: you must copy it or lose it!
+ *
+ * Side effects:
+ *	Can be almost arbitrary, depending on the commands in the script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_Eval
+int
+Tcl_Eval(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * previous call to Tcl_CreateInterp). */
+    const char *script)		/* Pointer to TCL command to execute. */
+{
+    int code = Tcl_EvalEx(interp, script, -1, 0);
+
+    /*
+     * For backwards compatibility with old C code that predates the object
+     * system in Tcl 8.0, we have to mirror the object result back into the
+     * string result (some callers may expect it there).
+     */
+
+    (void) Tcl_GetStringResult(interp);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalObj, Tcl_GlobalEvalObj --
+ *
+ *	These functions are deprecated but we keep them around for backwards
+ *	compatibility reasons.
+ *
+ * Results:
+ *	See the functions they call.
+ *
+ * Side effects:
+ *	See the functions they call.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_EvalObj
+int
+Tcl_EvalObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)
+{
+    return Tcl_EvalObjEx(interp, objPtr, 0);
+}
+#undef Tcl_GlobalEvalObj
+int
+Tcl_GlobalEvalObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)
+{
+    return Tcl_EvalObjEx(interp, objPtr, TCL_EVAL_GLOBAL);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalObjEx, TclEvalObjEx --
+ *
+ *	Execute Tcl commands stored in a Tcl object. These commands are
+ *	compiled into bytecodes if necessary, unless TCL_EVAL_DIRECT is
+ *	specified.
+ *
+ *	If the flag TCL_EVAL_DIRECT is passed in, the value of invoker
+ *	must be NULL.  Support for non-NULL invokers in that mode has
+ *	been removed since it was unused and untested.  Failure to
+ *	follow this limitation will lead to an assertion panic.
+ *
+ * Results:
+ *	The return value is one of the return codes defined in tcl.h (such as
+ *	TCL_OK), and the interpreter's result contains a value to supplement
+ *	the return code.
+ *
+ * Side effects:
+ *	The object is converted, if necessary, to a ByteCode object that holds
+ *	the bytecode instructions for the commands. Executing the commands
+ *	will almost certainly have side effects that depend on those commands.
+ *
+ * TIP #280 : Keep public API, internally extended API.
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_EvalObjEx(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * a previous call to Tcl_CreateInterp). */
+    register Tcl_Obj *objPtr,	/* Pointer to object containing commands to
+				 * execute. */
+    int flags)			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Supported values
+				 * are TCL_EVAL_GLOBAL and TCL_EVAL_DIRECT. */
+{
+    return TclEvalObjEx(interp, objPtr, flags, NULL, 0);
+}
+
+int
+TclEvalObjEx(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * a previous call to Tcl_CreateInterp). */
+    register Tcl_Obj *objPtr,	/* Pointer to object containing commands to
+				 * execute. */
+    int flags,			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Supported values
+				 * are TCL_EVAL_GLOBAL and TCL_EVAL_DIRECT. */
+    const CmdFrame *invoker,	/* Frame of the command doing the eval. */
+    int word)			/* Index of the word which is in objPtr. */
+{
+    int result = TCL_OK;
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    result = TclNREvalObjEx(interp, objPtr, flags, invoker, word);
+    return TclNRRunCallbacks(interp, result, rootPtr);
+}
+
+int
+TclNREvalObjEx(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * a previous call to Tcl_CreateInterp). */
+    register Tcl_Obj *objPtr,	/* Pointer to object containing commands to
+				 * execute. */
+    int flags,			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Supported values
+				 * are TCL_EVAL_GLOBAL and TCL_EVAL_DIRECT. */
+    const CmdFrame *invoker,	/* Frame of the command doing the eval. */
+    int word)			/* Index of the word which is in objPtr. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int result;
+
+    /*
+     * This function consists of three independent blocks for: direct
+     * evaluation of canonical lists, compilation and bytecode execution and
+     * finally direct evaluation. Precisely one of these blocks will be run.
+     */
+
+    if (TclListObjIsCanonical(objPtr)) {
+	CmdFrame *eoFramePtr = NULL;
+	int objc;
+	Tcl_Obj *listPtr, **objv;
+
+	/*
+	 * Canonical List Optimization:  In this case, we
+	 * can safely use Tcl_EvalObjv instead and get an appreciable
+	 * improvement in execution speed. This is because it allows us to
+	 * avoid a setFromAny step that would just pack everything into a
+	 * string and back out again.
+	 *
+	 * This also preserves any associations between list elements and
+	 * location information for such elements.
+	 */
+
+	/*
+	 * Shimmer protection! Always pass an unshared obj. The caller could
+	 * incr the refCount of objPtr AFTER calling us! To be completely safe
+	 * we always make a copy. The callback takes care od the refCounts for
+	 * both listPtr and objPtr.
+	 *
+	 * TODO: Create a test to demo this need, or eliminate it.
+	 * FIXME OPT: preserve just the internal rep?
+	 */
+
+	Tcl_IncrRefCount(objPtr);
+	listPtr = TclListObjCopy(interp, objPtr);
+	Tcl_IncrRefCount(listPtr);
+
+	if (word != INT_MIN) {
+	    /*
+	     * TIP #280 Structures for tracking lines. As we know that this is
+	     * dynamic execution we ignore the invoker, even if known.
+	     *
+	     * TIP #280. We do _not_ compute all the line numbers for the
+	     * words in the command. For the eval of a pure list the most
+	     * sensible choice is to put all words on line 1. Given that we
+	     * neither need memory for them nor compute anything. 'line' is
+	     * left NULL. The two places using this information (TclInfoFrame,
+	     * and TclInitCompileEnv), are special-cased to use the proper
+	     * line number directly instead of accessing the 'line' array.
+	     *
+	     * Note that we use (word==INTMIN) to signal that no command frame
+	     * should be pushed, as needed by alias and ensemble redirections.
+	     */
+
+	    eoFramePtr = TclStackAlloc(interp, sizeof(CmdFrame));
+	    eoFramePtr->nline = 0;
+	    eoFramePtr->line = NULL;
+
+	    eoFramePtr->type = TCL_LOCATION_EVAL;
+	    eoFramePtr->level = (iPtr->cmdFramePtr == NULL?
+		    1 : iPtr->cmdFramePtr->level + 1);
+	    eoFramePtr->framePtr = iPtr->framePtr;
+	    eoFramePtr->nextPtr = iPtr->cmdFramePtr;
+
+	    eoFramePtr->cmdObj = objPtr;
+	    eoFramePtr->cmd = NULL;
+	    eoFramePtr->len = 0;
+	    eoFramePtr->data.eval.path = NULL;
+
+	    iPtr->cmdFramePtr = eoFramePtr;
+
+	    flags |= TCL_EVAL_SOURCE_IN_FRAME;
+	}
+
+	TclMarkTailcall(interp);
+        TclNRAddCallback(interp, TEOEx_ListCallback, listPtr, eoFramePtr,
+		objPtr, NULL);
+
+	TclListObjGetElements(NULL, listPtr, &objc, &objv);
+	return TclNREvalObjv(interp, objc, objv, flags, NULL);
+    }
+
+    if (!(flags & TCL_EVAL_DIRECT)) {
+	/*
+	 * Let the compiler/engine subsystem do the evaluation.
+	 *
+	 * TIP #280 The invoker provides us with the context for the script.
+	 * We transfer this to the byte code compiler.
+	 */
+
+	int allowExceptions = (iPtr->evalFlags & TCL_ALLOW_EXCEPTIONS);
+	ByteCode *codePtr;
+	CallFrame *savedVarFramePtr = NULL;	/* Saves old copy of
+						 * iPtr->varFramePtr in case
+						 * TCL_EVAL_GLOBAL was set. */
+
+        if (TclInterpReady(interp) != TCL_OK) {
+            return TCL_ERROR;
+        }
+	if (flags & TCL_EVAL_GLOBAL) {
+	    savedVarFramePtr = iPtr->varFramePtr;
+	    iPtr->varFramePtr = iPtr->rootFramePtr;
+	}
+	Tcl_IncrRefCount(objPtr);
+	codePtr = TclCompileObj(interp, objPtr, invoker, word);
+
+	TclNRAddCallback(interp, TEOEx_ByteCodeCallback, savedVarFramePtr,
+		objPtr, INT2PTR(allowExceptions), NULL);
+        return TclNRExecuteByteCode(interp, codePtr);
+    }
+
+    {
+	/*
+	 * We're not supposed to use the compiler or byte-code
+	 * interpreter. Let Tcl_EvalEx evaluate the command directly (and
+	 * probably more slowly).
+	 */
+
+	const char *script;
+	int numSrcBytes;
+
+	/*
+	 * Now we check if we have data about invisible continuation lines for
+	 * the script, and make it available to the direct script parser and
+	 * evaluator we are about to call, if so.
+	 *
+	 * It may be possible that the script Tcl_Obj* can be free'd while the
+	 * evaluator is using it, leading to the release of the associated
+	 * ContLineLoc structure as well. To ensure that the latter doesn't
+	 * happen we set a lock on it. We release this lock later in this
+	 * function, after the evaluator is done. The relevant "lineCLPtr"
+	 * hashtable is managed in the file "tclObj.c".
+	 *
+	 * Another important action is to save (and later restore) the
+	 * continuation line information of the caller, in case we are
+	 * executing nested commands in the eval/direct path.
+	 */
+
+	ContLineLoc *saveCLLocPtr = iPtr->scriptCLLocPtr;
+
+	assert(invoker == NULL);
+
+	iPtr->scriptCLLocPtr = TclContinuationsGet(objPtr);
+
+	Tcl_IncrRefCount(objPtr);
+
+	script = TclGetStringFromObj(objPtr, &numSrcBytes);
+	result = Tcl_EvalEx(interp, script, numSrcBytes, flags);
+
+	TclDecrRefCount(objPtr);
+
+	iPtr->scriptCLLocPtr = saveCLLocPtr;
+	return result;
+    }
+}
+
+static int
+TEOEx_ByteCodeCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *savedVarFramePtr = data[0];
+    Tcl_Obj *objPtr = data[1];
+    int allowExceptions = PTR2INT(data[2]);
+
+    if (iPtr->numLevels == 0) {
+	if (result == TCL_RETURN) {
+	    result = TclUpdateReturnInfo(iPtr);
+	}
+	if ((result != TCL_OK) && (result != TCL_ERROR) && !allowExceptions) {
+	    const char *script;
+	    int numSrcBytes;
+
+	    ProcessUnexpectedResult(interp, result);
+	    result = TCL_ERROR;
+	    script = TclGetStringFromObj(objPtr, &numSrcBytes);
+	    Tcl_LogCommandInfo(interp, script, script, numSrcBytes);
+	}
+
+	/*
+	 * We are returning to level 0, so should call TclResetCancellation.
+	 * Let us just unset the flags inline.
+	 */
+
+	TclUnsetCancelFlags(iPtr);
+    }
+    iPtr->evalFlags = 0;
+
+    /*
+     * Restore the callFrame if this was a TCL_EVAL_GLOBAL.
+     */
+
+    if (savedVarFramePtr) {
+	iPtr->varFramePtr = savedVarFramePtr;
+    }
+
+    TclDecrRefCount(objPtr);
+    return result;
+}
+
+static int
+TEOEx_ListCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *listPtr = data[0];
+    CmdFrame *eoFramePtr = data[1];
+    Tcl_Obj *objPtr = data[2];
+
+    /*
+     * Remove the cmdFrame
+     */
+
+    if (eoFramePtr) {
+	iPtr->cmdFramePtr = eoFramePtr->nextPtr;
+	TclStackFree(interp, eoFramePtr);
+    }
+    TclDecrRefCount(objPtr);
+    TclDecrRefCount(listPtr);
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ProcessUnexpectedResult --
+ *
+ *	Function called by Tcl_EvalObj to set the interpreter's result value
+ *	to an appropriate error message when the code it evaluates returns an
+ *	unexpected result code (not TCL_OK and not TCL_ERROR) to the topmost
+ *	evaluation level.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interpreter result is set to an error message appropriate to the
+ *	result code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ProcessUnexpectedResult(
+    Tcl_Interp *interp,		/* The interpreter in which the unexpected
+				 * result code was returned. */
+    int returnCode)		/* The unexpected result code. */
+{
+    char buf[TCL_INTEGER_SPACE];
+
+    Tcl_ResetResult(interp);
+    if (returnCode == TCL_BREAK) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"invoked \"break\" outside of a loop", -1));
+    } else if (returnCode == TCL_CONTINUE) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"invoked \"continue\" outside of a loop", -1));
+    } else {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"command returned bad code: %d", returnCode));
+    }
+    sprintf(buf, "%d", returnCode);
+    Tcl_SetErrorCode(interp, "TCL", "UNEXPECTED_RESULT_CODE", buf, NULL);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_ExprLong, Tcl_ExprDouble, Tcl_ExprBoolean --
+ *
+ *	Functions to evaluate an expression and return its value in a
+ *	particular form.
+ *
+ * Results:
+ *	Each of the functions below returns a standard Tcl result. If an error
+ *	occurs then an error message is left in the interp's result. Otherwise
+ *	the value of the expression, in the appropriate form, is stored at
+ *	*ptr. If the expression had a result that was incompatible with the
+ *	desired form then an error is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_ExprLong(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    const char *exprstring,	/* Expression to evaluate. */
+    long *ptr)			/* Where to store result. */
+{
+    register Tcl_Obj *exprPtr;
+    int result = TCL_OK;
+    if (*exprstring == '\0') {
+	/*
+	 * Legacy compatibility - return 0 for the zero-length string.
+	 */
+
+	*ptr = 0;
+    } else {
+	exprPtr = Tcl_NewStringObj(exprstring, -1);
+	Tcl_IncrRefCount(exprPtr);
+	result = Tcl_ExprLongObj(interp, exprPtr, ptr);
+	Tcl_DecrRefCount(exprPtr);
+	if (result != TCL_OK) {
+	    (void) Tcl_GetStringResult(interp);
+	}
+    }
+    return result;
+}
+
+int
+Tcl_ExprDouble(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    const char *exprstring,	/* Expression to evaluate. */
+    double *ptr)		/* Where to store result. */
+{
+    register Tcl_Obj *exprPtr;
+    int result = TCL_OK;
+
+    if (*exprstring == '\0') {
+	/*
+	 * Legacy compatibility - return 0 for the zero-length string.
+	 */
+
+	*ptr = 0.0;
+    } else {
+	exprPtr = Tcl_NewStringObj(exprstring, -1);
+	Tcl_IncrRefCount(exprPtr);
+	result = Tcl_ExprDoubleObj(interp, exprPtr, ptr);
+	Tcl_DecrRefCount(exprPtr);
+				/* Discard the expression object. */
+	if (result != TCL_OK) {
+	    (void) Tcl_GetStringResult(interp);
+	}
+    }
+    return result;
+}
+
+int
+Tcl_ExprBoolean(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    const char *exprstring,	/* Expression to evaluate. */
+    int *ptr)			/* Where to store 0/1 result. */
+{
+    if (*exprstring == '\0') {
+	/*
+	 * An empty string. Just set the result boolean to 0 (false).
+	 */
+
+	*ptr = 0;
+	return TCL_OK;
+    } else {
+	int result;
+	Tcl_Obj *exprPtr = Tcl_NewStringObj(exprstring, -1);
+
+	Tcl_IncrRefCount(exprPtr);
+	result = Tcl_ExprBooleanObj(interp, exprPtr, ptr);
+	Tcl_DecrRefCount(exprPtr);
+	if (result != TCL_OK) {
+	    /*
+	     * Move the interpreter's object result to the string result, then
+	     * reset the object result.
+	     */
+
+	    (void) Tcl_GetStringResult(interp);
+	}
+	return result;
+    }
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_ExprLongObj, Tcl_ExprDoubleObj, Tcl_ExprBooleanObj --
+ *
+ *	Functions to evaluate an expression in an object and return its value
+ *	in a particular form.
+ *
+ * Results:
+ *	Each of the functions below returns a standard Tcl result object. If
+ *	an error occurs then an error message is left in the interpreter's
+ *	result. Otherwise the value of the expression, in the appropriate
+ *	form, is stored at *ptr. If the expression had a result that was
+ *	incompatible with the desired form then an error is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *--------------------------------------------------------------
+ */
+
+int
+Tcl_ExprLongObj(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    register Tcl_Obj *objPtr,	/* Expression to evaluate. */
+    long *ptr)			/* Where to store long result. */
+{
+    Tcl_Obj *resultPtr;
+    int result, type;
+    double d;
+    ClientData internalPtr;
+
+    result = Tcl_ExprObj(interp, objPtr, &resultPtr);
+    if (result != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (TclGetNumberFromObj(interp, resultPtr, &internalPtr, &type)!=TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch (type) {
+    case TCL_NUMBER_DOUBLE: {
+	mp_int big;
+
+	d = *((const double *) internalPtr);
+	Tcl_DecrRefCount(resultPtr);
+	if (Tcl_InitBignumFromDouble(interp, d, &big) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	resultPtr = Tcl_NewBignumObj(&big);
+	/* FALLTHROUGH */
+    }
+    case TCL_NUMBER_LONG:
+    case TCL_NUMBER_WIDE:
+    case TCL_NUMBER_BIG:
+	result = TclGetLongFromObj(interp, resultPtr, ptr);
+	break;
+
+    case TCL_NUMBER_NAN:
+	Tcl_GetDoubleFromObj(interp, resultPtr, &d);
+	result = TCL_ERROR;
+    }
+
+    Tcl_DecrRefCount(resultPtr);/* Discard the result object. */
+    return result;
+}
+
+int
+Tcl_ExprDoubleObj(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    register Tcl_Obj *objPtr,	/* Expression to evaluate. */
+    double *ptr)		/* Where to store double result. */
+{
+    Tcl_Obj *resultPtr;
+    int result, type;
+    ClientData internalPtr;
+
+    result = Tcl_ExprObj(interp, objPtr, &resultPtr);
+    if (result != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    result = TclGetNumberFromObj(interp, resultPtr, &internalPtr, &type);
+    if (result == TCL_OK) {
+	switch (type) {
+	case TCL_NUMBER_NAN:
+#ifndef ACCEPT_NAN
+	    result = Tcl_GetDoubleFromObj(interp, resultPtr, ptr);
+	    break;
+#endif
+	case TCL_NUMBER_DOUBLE:
+	    *ptr = *((const double *) internalPtr);
+	    result = TCL_OK;
+	    break;
+	default:
+	    result = Tcl_GetDoubleFromObj(interp, resultPtr, ptr);
+	}
+    }
+    Tcl_DecrRefCount(resultPtr);/* Discard the result object. */
+    return result;
+}
+
+int
+Tcl_ExprBooleanObj(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    register Tcl_Obj *objPtr,	/* Expression to evaluate. */
+    int *ptr)			/* Where to store 0/1 result. */
+{
+    Tcl_Obj *resultPtr;
+    int result;
+
+    result = Tcl_ExprObj(interp, objPtr, &resultPtr);
+    if (result == TCL_OK) {
+	result = Tcl_GetBooleanFromObj(interp, resultPtr, ptr);
+	Tcl_DecrRefCount(resultPtr);
+				/* Discard the result object. */
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjInvokeNamespace --
+ *
+ *	Object version: Invokes a Tcl command, given an objv/objc, from either
+ *	the exposed or hidden set of commands in the given interpreter.
+ *
+ *	NOTE: The command is invoked in the global stack frame of the
+ *	interpreter or namespace, thus it cannot see any current state on the
+ *	stack of that interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Whatever the command does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjInvokeNamespace(
+    Tcl_Interp *interp,		/* Interpreter in which command is to be
+				 * invoked. */
+    int objc,			/* Count of arguments. */
+    Tcl_Obj *const objv[],	/* Argument objects; objv[0] points to the
+				 * name of the command to invoke. */
+    Tcl_Namespace *nsPtr,	/* The namespace to use. */
+    int flags)			/* Combination of flags controlling the call:
+				 * TCL_INVOKE_HIDDEN, TCL_INVOKE_NO_UNKNOWN,
+				 * or TCL_INVOKE_NO_TRACEBACK. */
+{
+    int result;
+    Tcl_CallFrame *framePtr;
+
+    /*
+     * Make the specified namespace the current namespace and invoke the
+     * command.
+     */
+
+    (void) TclPushStackFrame(interp, &framePtr, nsPtr, /*isProcFrame*/0);
+    result = TclObjInvoke(interp, objc, objv, flags);
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjInvoke --
+ *
+ *	Invokes a Tcl command, given an objv/objc, from either the exposed or
+ *	the hidden sets of commands in the given interpreter.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Whatever the command does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjInvoke(
+    Tcl_Interp *interp,		/* Interpreter in which command is to be
+				 * invoked. */
+    int objc,			/* Count of arguments. */
+    Tcl_Obj *const objv[],	/* Argument objects; objv[0] points to the
+				 * name of the command to invoke. */
+    int flags)			/* Combination of flags controlling the call:
+				 * TCL_INVOKE_HIDDEN, TCL_INVOKE_NO_UNKNOWN,
+				 * or TCL_INVOKE_NO_TRACEBACK. */
+{
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+    if ((objc < 1) || (objv == NULL)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "illegal argument vector", -1));
+	return TCL_ERROR;
+    }
+    if ((flags & TCL_INVOKE_HIDDEN) == 0) {
+	Tcl_Panic("TclObjInvoke: called without TCL_INVOKE_HIDDEN");
+    }
+    return Tcl_NRCallObjProc(interp, TclNRInvoke, NULL, objc, objv);
+}
+
+int
+TclNRInvoke(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    register Interp *iPtr = (Interp *) interp;
+    Tcl_HashTable *hTblPtr;	/* Table of hidden commands. */
+    const char *cmdName;	/* Name of the command from objv[0]. */
+    Tcl_HashEntry *hPtr = NULL;
+    Command *cmdPtr;
+
+    cmdName = TclGetString(objv[0]);
+    hTblPtr = iPtr->hiddenCmdTablePtr;
+    if (hTblPtr != NULL) {
+	hPtr = Tcl_FindHashEntry(hTblPtr, cmdName);
+    }
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "invalid hidden command name \"%s\"", cmdName));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "HIDDENTOKEN", cmdName,
+                NULL);
+	return TCL_ERROR;
+    }
+    cmdPtr = Tcl_GetHashValue(hPtr);
+
+    /* Avoid the exception-handling brain damage when numLevels == 0 . */
+    iPtr->numLevels++;
+    Tcl_NRAddCallback(interp, NRPostInvoke, NULL, NULL, NULL, NULL);
+
+    /*
+     * Normal command resolution of objv[0] isn't going to find cmdPtr.
+     * That's the whole point of **hidden** commands.  So tell the
+     * Eval core machinery not to even try (and risk finding something wrong).
+     */
+
+    return TclNREvalObjv(interp, objc, objv, TCL_EVAL_NORESOLVE, cmdPtr);
+}
+
+static int
+NRPostInvoke(
+    ClientData clientData[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *)interp;
+    iPtr->numLevels--;
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_ExprString --
+ *
+ *	Evaluate an expression in a string and return its value in string
+ *	form.
+ *
+ * Results:
+ *	A standard Tcl result. If the result is TCL_OK, then the interp's
+ *	result is set to the string value of the expression. If the result is
+ *	TCL_ERROR, then the interp's result contains an error message.
+ *
+ * Side effects:
+ *	A Tcl object is allocated to hold a copy of the expression string.
+ *	This expression object is passed to Tcl_ExprObj and then deallocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_ExprString(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    const char *expr)		/* Expression to evaluate. */
+{
+    int code = TCL_OK;
+
+    if (expr[0] == '\0') {
+	/*
+	 * An empty string. Just set the interpreter's result to 0.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(0));
+    } else {
+	Tcl_Obj *resultPtr, *exprObj = Tcl_NewStringObj(expr, -1);
+
+	Tcl_IncrRefCount(exprObj);
+	code = Tcl_ExprObj(interp, exprObj, &resultPtr);
+	Tcl_DecrRefCount(exprObj);
+	if (code == TCL_OK) {
+	    Tcl_SetObjResult(interp, resultPtr);
+	    Tcl_DecrRefCount(resultPtr);
+	}
+    }
+
+    /*
+     * Force the string rep of the interp result.
+     */
+
+    (void) Tcl_GetStringResult(interp);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendObjToErrorInfo --
+ *
+ *	Add a Tcl_Obj value to the errorInfo field that describes the current
+ *	error.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The value of the Tcl_obj is appended to the errorInfo field. If we are
+ *	just starting to log an error, errorInfo is initialized from the error
+ *	message in the interpreter's result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_AddObjErrorInfo
+void
+Tcl_AppendObjToErrorInfo(
+    Tcl_Interp *interp,		/* Interpreter to which error information
+				 * pertains. */
+    Tcl_Obj *objPtr)		/* Message to record. */
+{
+    const char *message = TclGetString(objPtr);
+
+    Tcl_IncrRefCount(objPtr);
+    Tcl_AddObjErrorInfo(interp, message, objPtr->length);
+    Tcl_DecrRefCount(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AddErrorInfo --
+ *
+ *	Add information to the errorInfo field that describes the current
+ *	error.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The contents of message are appended to the errorInfo field. If we are
+ *	just starting to log an error, errorInfo is initialized from the error
+ *	message in the interpreter's result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_AddErrorInfo
+void
+Tcl_AddErrorInfo(
+    Tcl_Interp *interp,		/* Interpreter to which error information
+				 * pertains. */
+    const char *message)	/* Message to record. */
+{
+    Tcl_AddObjErrorInfo(interp, message, -1);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AddObjErrorInfo --
+ *
+ *	Add information to the errorInfo field that describes the current
+ *	error. This routine differs from Tcl_AddErrorInfo by taking a byte
+ *	pointer and length.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	"length" bytes from "message" are appended to the errorInfo field. If
+ *	"length" is negative, use bytes up to the first NULL byte. If we are
+ *	just starting to log an error, errorInfo is initialized from the error
+ *	message in the interpreter's result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AddObjErrorInfo(
+    Tcl_Interp *interp,		/* Interpreter to which error information
+				 * pertains. */
+    const char *message,	/* Points to the first byte of an array of
+				 * bytes of the message. */
+    int length)			/* The number of bytes in the message. If < 0,
+				 * then append all bytes up to a NULL byte. */
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    /*
+     * If we are just starting to log an error, errorInfo is initialized from
+     * the error message in the interpreter's result.
+     */
+
+    iPtr->flags |= ERR_LEGACY_COPY;
+    if (iPtr->errorInfo == NULL) {
+	if (iPtr->result[0] != 0) {
+	    /*
+	     * The interp's string result is set, apparently by some extension
+	     * making a deprecated direct write to it. That extension may
+	     * expect interp->result to continue to be set, so we'll take
+	     * special pains to avoid clearing it, until we drop support for
+	     * interp->result completely.
+	     */
+
+	    iPtr->errorInfo = Tcl_NewStringObj(iPtr->result, -1);
+	} else {
+	    iPtr->errorInfo = iPtr->objResultPtr;
+	}
+	Tcl_IncrRefCount(iPtr->errorInfo);
+	if (!iPtr->errorCode) {
+	    Tcl_SetErrorCode(interp, "NONE", NULL);
+	}
+    }
+
+    /*
+     * Now append "message" to the end of errorInfo.
+     */
+
+    if (length != 0) {
+	if (Tcl_IsShared(iPtr->errorInfo)) {
+	    Tcl_DecrRefCount(iPtr->errorInfo);
+	    iPtr->errorInfo = Tcl_DuplicateObj(iPtr->errorInfo);
+	    Tcl_IncrRefCount(iPtr->errorInfo);
+	}
+	Tcl_AppendToObj(iPtr->errorInfo, message, length);
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_VarEvalVA --
+ *
+ *	Given a variable number of string arguments, concatenate them all
+ *	together and execute the result as a Tcl command.
+ *
+ * Results:
+ *	A standard Tcl return result. An error message or other result may be
+ *	left in the interp's result.
+ *
+ * Side effects:
+ *	Depends on what was done by the command.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_VarEvalVA(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate command */
+    va_list argList)		/* Variable argument list. */
+{
+    Tcl_DString buf;
+    char *string;
+    int result;
+
+    /*
+     * Copy the strings one after the other into a single larger string. Use
+     * stack-allocated space for small commands, but if the command gets too
+     * large than call ckalloc to create the space.
+     */
+
+    Tcl_DStringInit(&buf);
+    while (1) {
+	string = va_arg(argList, char *);
+	if (string == NULL) {
+	    break;
+	}
+	Tcl_DStringAppend(&buf, string, -1);
+    }
+
+    result = Tcl_EvalEx(interp, Tcl_DStringValue(&buf), -1, 0);
+    Tcl_DStringFree(&buf);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_VarEval --
+ *
+ *	Given a variable number of string arguments, concatenate them all
+ *	together and execute the result as a Tcl command.
+ *
+ * Results:
+ *	A standard Tcl return result. An error message or other result may be
+ *	left in interp->result.
+ *
+ * Side effects:
+ *	Depends on what was done by the command.
+ *
+ *----------------------------------------------------------------------
+ */
+	/* ARGSUSED */
+int
+Tcl_VarEval(
+    Tcl_Interp *interp,
+    ...)
+{
+    va_list argList;
+    int result;
+
+    va_start(argList, interp);
+    result = Tcl_VarEvalVA(interp, argList);
+    va_end(argList);
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GlobalEval --
+ *
+ *	Evaluate a command at global level in an interpreter.
+ *
+ * Results:
+ *	A standard Tcl result is returned, and the interp's result is modified
+ *	accordingly.
+ *
+ * Side effects:
+ *	The command string is executed in interp, and the execution is carried
+ *	out in the variable context of global level (no functions active),
+ *	just as if an "uplevel #0" command were being executed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_GlobalEval
+int
+Tcl_GlobalEval(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate
+				 * command. */
+    const char *command)	/* Command to evaluate. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    int result;
+    CallFrame *savedVarFramePtr;
+
+    savedVarFramePtr = iPtr->varFramePtr;
+    iPtr->varFramePtr = iPtr->rootFramePtr;
+    result = Tcl_EvalEx(interp, command, -1, 0);
+    iPtr->varFramePtr = savedVarFramePtr;
+    return result;
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetRecursionLimit --
+ *
+ *	Set the maximum number of recursive calls that may be active for an
+ *	interpreter at once.
+ *
+ * Results:
+ *	The return value is the old limit on nesting for interp.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetRecursionLimit(
+    Tcl_Interp *interp,		/* Interpreter whose nesting limit is to be
+				 * set. */
+    int depth)			/* New value for maximimum depth. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int old;
+
+    old = iPtr->maxNestingDepth;
+    if (depth > 0) {
+	iPtr->maxNestingDepth = depth;
+    }
+    return old;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AllowExceptions --
+ *
+ *	Sets a flag in an interpreter so that exceptions can occur in the next
+ *	call to Tcl_Eval without them being turned into errors.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The TCL_ALLOW_EXCEPTIONS flag gets set in the interpreter's evalFlags
+ *	structure. See the reference documentation for more details.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AllowExceptions(
+    Tcl_Interp *interp)		/* Interpreter in which to set flag. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->evalFlags |= TCL_ALLOW_EXCEPTIONS;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetVersion --
+ *
+ *	Get the Tcl major, minor, and patchlevel version numbers and the
+ *	release type. A patch is a release type TCL_FINAL_RELEASE with a
+ *	patchLevel > 0.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetVersion(
+    int *majorV,
+    int *minorV,
+    int *patchLevelV,
+    int *type)
+{
+    if (majorV != NULL) {
+	*majorV = TCL_MAJOR_VERSION;
+    }
+    if (minorV != NULL) {
+	*minorV = TCL_MINOR_VERSION;
+    }
+    if (patchLevelV != NULL) {
+	*patchLevelV = TCL_RELEASE_SERIAL;
+    }
+    if (type != NULL) {
+	*type = TCL_RELEASE_LEVEL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Math Functions --
+ *
+ *	This page contains the functions that implement all of the built-in
+ *	math functions for expressions.
+ *
+ * Results:
+ *	Each function returns TCL_OK if it succeeds and pushes an Tcl object
+ *	holding the result. If it fails it returns TCL_ERROR and leaves an
+ *	error message in the interpreter's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ExprCeilFunc(
+    ClientData clientData,	/* Ignored */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter list. */
+{
+    int code;
+    double d;
+    mp_int big;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[1], &d);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[1]->typePtr == &tclDoubleType)) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetBignumFromObj(NULL, objv[1], &big) == TCL_OK) {
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(TclCeil(&big)));
+	mp_clear(&big);
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(ceil(d)));
+    }
+    return TCL_OK;
+}
+
+static int
+ExprFloorFunc(
+    ClientData clientData,	/* Ignored */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter list. */
+{
+    int code;
+    double d;
+    mp_int big;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[1], &d);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[1]->typePtr == &tclDoubleType)) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetBignumFromObj(NULL, objv[1], &big) == TCL_OK) {
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(TclFloor(&big)));
+	mp_clear(&big);
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(floor(d)));
+    }
+    return TCL_OK;
+}
+
+static int
+ExprIsqrtFunc(
+    ClientData clientData,	/* Ignored */
+    Tcl_Interp *interp,		/* The interpreter in which to execute. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter list. */
+{
+    ClientData ptr;
+    int type;
+    double d;
+    Tcl_WideInt w;
+    mp_int big;
+    int exact = 0;		/* Flag ==1 if the argument can be represented
+				 * in a double as an exact integer. */
+
+    /*
+     * Check syntax.
+     */
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure that the arg is a number.
+     */
+
+    if (TclGetNumberFromObj(interp, objv[1], &ptr, &type) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch (type) {
+    case TCL_NUMBER_NAN:
+	Tcl_GetDoubleFromObj(interp, objv[1], &d);
+	return TCL_ERROR;
+    case TCL_NUMBER_DOUBLE:
+	d = *((const double *) ptr);
+	if (d < 0) {
+	    goto negarg;
+	}
+#ifdef IEEE_FLOATING_POINT
+	if (d <= MAX_EXACT) {
+	    exact = 1;
+	}
+#endif
+	if (!exact) {
+	    if (Tcl_InitBignumFromDouble(interp, d, &big) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	break;
+    case TCL_NUMBER_BIG:
+	if (Tcl_GetBignumFromObj(interp, objv[1], &big) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (SIGN(&big) == MP_NEG) {
+	    mp_clear(&big);
+	    goto negarg;
+	}
+	break;
+    default:
+	if (TclGetWideIntFromObj(interp, objv[1], &w) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (w < 0) {
+	    goto negarg;
+	}
+	d = (double) w;
+#ifdef IEEE_FLOATING_POINT
+	if (d < MAX_EXACT) {
+	    exact = 1;
+	}
+#endif
+	if (!exact) {
+	    Tcl_GetBignumFromObj(interp, objv[1], &big);
+	}
+	break;
+    }
+
+    if (exact) {
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt) sqrt(d)));
+    } else {
+	mp_int root;
+
+	mp_init(&root);
+	mp_sqrt(&big, &root);
+	mp_clear(&big);
+	Tcl_SetObjResult(interp, Tcl_NewBignumObj(&root));
+    }
+    return TCL_OK;
+
+  negarg:
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+            "square root of negative argument", -1));
+    Tcl_SetErrorCode(interp, "ARITH", "DOMAIN",
+	    "domain error: argument not in valid range", NULL);
+    return TCL_ERROR;
+}
+
+static int
+ExprSqrtFunc(
+    ClientData clientData,	/* Ignored */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter list. */
+{
+    int code;
+    double d;
+    mp_int big;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[1], &d);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[1]->typePtr == &tclDoubleType)) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if ((d >= 0.0) && TclIsInfinite(d)
+	    && (Tcl_GetBignumFromObj(NULL, objv[1], &big) == TCL_OK)) {
+	mp_int root;
+
+	mp_init(&root);
+	mp_sqrt(&big, &root);
+	mp_clear(&big);
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(TclBignumToDouble(&root)));
+	mp_clear(&root);
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(sqrt(d)));
+    }
+    return TCL_OK;
+}
+
+static int
+ExprUnaryFunc(
+    ClientData clientData,	/* Contains the address of a function that
+				 * takes one double argument and returns a
+				 * double result. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count */
+    Tcl_Obj *const *objv)	/* Actual parameter list */
+{
+    int code;
+    double d;
+    double (*func)(double) = (double (*)(double)) clientData;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[1], &d);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[1]->typePtr == &tclDoubleType)) {
+	d = objv[1]->internalRep.doubleValue;
+	Tcl_ResetResult(interp);
+	code = TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+    errno = 0;
+    return CheckDoubleResult(interp, func(d));
+}
+
+static int
+CheckDoubleResult(
+    Tcl_Interp *interp,
+    double dResult)
+{
+#ifndef ACCEPT_NAN
+    if (TclIsNaN(dResult)) {
+	TclExprFloatError(interp, dResult);
+	return TCL_ERROR;
+    }
+#endif
+    if ((errno == ERANGE) && ((dResult == 0.0) || TclIsInfinite(dResult))) {
+	/*
+	 * When ERANGE signals under/overflow, just accept 0.0 or +/-Inf
+	 */
+    } else if (errno != 0) {
+	/*
+	 * Report other errno values as errors.
+	 */
+
+	TclExprFloatError(interp, dResult);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewDoubleObj(dResult));
+    return TCL_OK;
+}
+
+static int
+ExprBinaryFunc(
+    ClientData clientData,	/* Contains the address of a function that
+				 * takes two double arguments and returns a
+				 * double result. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Parameter vector. */
+{
+    int code;
+    double d1, d2;
+    double (*func)(double, double) = (double (*)(double, double)) clientData;
+
+    if (objc != 3) {
+	MathFuncWrongNumArgs(interp, 3, objc, objv);
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[1], &d1);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[1]->typePtr == &tclDoubleType)) {
+	d1 = objv[1]->internalRep.doubleValue;
+	Tcl_ResetResult(interp);
+	code = TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+    code = Tcl_GetDoubleFromObj(interp, objv[2], &d2);
+#ifdef ACCEPT_NAN
+    if ((code != TCL_OK) && (objv[2]->typePtr == &tclDoubleType)) {
+	d2 = objv[2]->internalRep.doubleValue;
+	Tcl_ResetResult(interp);
+	code = TCL_OK;
+    }
+#endif
+    if (code != TCL_OK) {
+	return TCL_ERROR;
+    }
+    errno = 0;
+    return CheckDoubleResult(interp, func(d1, d2));
+}
+
+static int
+ExprAbsFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Parameter vector. */
+{
+    ClientData ptr;
+    int type;
+    mp_int big;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+
+    if (TclGetNumberFromObj(interp, objv[1], &ptr, &type) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (type == TCL_NUMBER_LONG) {
+	long l = *((const long *) ptr);
+
+	if (l > (long)0) {
+	    goto unChanged;
+	} else if (l == (long)0) {
+	    const char *string = objv[1]->bytes;
+	    if (string) {
+		while (*string != '0') {
+		    if (*string == '-') {
+			Tcl_SetObjResult(interp, Tcl_NewLongObj(0));
+			return TCL_OK;
+		    }
+		    string++;
+		}
+	    }
+	    goto unChanged;
+	} else if (l == LONG_MIN) {
+	    TclBNInitBignumFromLong(&big, l);
+	    goto tooLarge;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewLongObj(-l));
+	return TCL_OK;
+    }
+
+    if (type == TCL_NUMBER_DOUBLE) {
+	double d = *((const double *) ptr);
+	static const double poszero = 0.0;
+
+	/*
+	 * We need to distinguish here between positive 0.0 and negative -0.0.
+	 * [Bug 2954959]
+	 */
+
+	if (d == -0.0) {
+	    if (!memcmp(&d, &poszero, sizeof(double))) {
+		goto unChanged;
+	    }
+	} else if (d > -0.0) {
+	    goto unChanged;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewDoubleObj(-d));
+	return TCL_OK;
+    }
+
+#ifndef TCL_WIDE_INT_IS_LONG
+    if (type == TCL_NUMBER_WIDE) {
+	Tcl_WideInt w = *((const Tcl_WideInt *) ptr);
+
+	if (w >= (Tcl_WideInt)0) {
+	    goto unChanged;
+	}
+	if (w == LLONG_MIN) {
+	    TclBNInitBignumFromWideInt(&big, w);
+	    goto tooLarge;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj(-w));
+	return TCL_OK;
+    }
+#endif
+
+    if (type == TCL_NUMBER_BIG) {
+	if (mp_cmp_d((const mp_int *) ptr, 0) == MP_LT) {
+	    Tcl_GetBignumFromObj(NULL, objv[1], &big);
+	tooLarge:
+	    mp_neg(&big, &big);
+	    Tcl_SetObjResult(interp, Tcl_NewBignumObj(&big));
+	} else {
+	unChanged:
+	    Tcl_SetObjResult(interp, objv[1]);
+	}
+	return TCL_OK;
+    }
+
+    if (type == TCL_NUMBER_NAN) {
+#ifdef ACCEPT_NAN
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+#else
+	double d;
+
+	Tcl_GetDoubleFromObj(interp, objv[1], &d);
+	return TCL_ERROR;
+#endif
+    }
+    return TCL_OK;
+}
+
+static int
+ExprBoolFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    int value;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetBooleanFromObj(interp, objv[1], &value) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(value));
+    return TCL_OK;
+}
+
+static int
+ExprDoubleFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    double dResult;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetDoubleFromObj(interp, objv[1], &dResult) != TCL_OK) {
+#ifdef ACCEPT_NAN
+	if (objv[1]->typePtr == &tclDoubleType) {
+	    Tcl_SetObjResult(interp, objv[1]);
+	    return TCL_OK;
+	}
+#endif
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewDoubleObj(dResult));
+    return TCL_OK;
+}
+
+static int
+ExprEntierFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    double d;
+    int type;
+    ClientData ptr;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+    if (TclGetNumberFromObj(interp, objv[1], &ptr, &type) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (type == TCL_NUMBER_DOUBLE) {
+	d = *((const double *) ptr);
+	if ((d >= (double)LONG_MAX) || (d <= (double)LONG_MIN)) {
+	    mp_int big;
+
+	    if (Tcl_InitBignumFromDouble(interp, d, &big) != TCL_OK) {
+		/* Infinity */
+		return TCL_ERROR;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewBignumObj(&big));
+	    return TCL_OK;
+	} else {
+	    long result = (long) d;
+
+	    Tcl_SetObjResult(interp, Tcl_NewLongObj(result));
+	    return TCL_OK;
+	}
+    }
+
+    if (type != TCL_NUMBER_NAN) {
+	/*
+	 * All integers are already of integer type.
+	 */
+
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    /*
+     * Get the error message for NaN.
+     */
+
+    Tcl_GetDoubleFromObj(interp, objv[1], &d);
+    return TCL_ERROR;
+}
+
+static int
+ExprIntFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    long iResult;
+    Tcl_Obj *objPtr;
+    if (ExprEntierFunc(NULL, interp, objc, objv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    objPtr = Tcl_GetObjResult(interp);
+    if (TclGetLongFromObj(NULL, objPtr, &iResult) != TCL_OK) {
+	/*
+	 * Truncate the bignum; keep only bits in long range.
+	 */
+
+	mp_int big;
+
+	Tcl_GetBignumFromObj(NULL, objPtr, &big);
+	mp_mod_2d(&big, (int) CHAR_BIT * sizeof(long), &big);
+	objPtr = Tcl_NewBignumObj(&big);
+	Tcl_IncrRefCount(objPtr);
+	TclGetLongFromObj(NULL, objPtr, &iResult);
+	Tcl_DecrRefCount(objPtr);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewLongObj(iResult));
+    return TCL_OK;
+}
+
+static int
+ExprWideFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    Tcl_WideInt wResult;
+    Tcl_Obj *objPtr;
+
+    if (ExprEntierFunc(NULL, interp, objc, objv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    objPtr = Tcl_GetObjResult(interp);
+    if (TclGetWideIntFromObj(NULL, objPtr, &wResult) != TCL_OK) {
+	/*
+	 * Truncate the bignum; keep only bits in wide int range.
+	 */
+
+	mp_int big;
+
+	Tcl_GetBignumFromObj(NULL, objPtr, &big);
+	mp_mod_2d(&big, (int) CHAR_BIT * sizeof(Tcl_WideInt), &big);
+	objPtr = Tcl_NewBignumObj(&big);
+	Tcl_IncrRefCount(objPtr);
+	TclGetWideIntFromObj(NULL, objPtr, &wResult);
+	Tcl_DecrRefCount(objPtr);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj(wResult));
+    return TCL_OK;
+}
+
+static int
+ExprRandFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    Interp *iPtr = (Interp *) interp;
+    double dResult;
+    long tmp;			/* Algorithm assumes at least 32 bits. Only
+				 * long guarantees that. See below. */
+    Tcl_Obj *oResult;
+
+    if (objc != 1) {
+	MathFuncWrongNumArgs(interp, 1, objc, objv);
+	return TCL_ERROR;
+    }
+
+    if (!(iPtr->flags & RAND_SEED_INITIALIZED)) {
+	iPtr->flags |= RAND_SEED_INITIALIZED;
+
+	/*
+	 * Take into consideration the thread this interp is running in order
+	 * to insure different seeds in different threads (bug #416643)
+	 */
+
+	iPtr->randSeed = TclpGetClicks() + (PTR2INT(Tcl_GetCurrentThread())<<12);
+
+	/*
+	 * Make sure 1 <= randSeed <= (2^31) - 2. See below.
+	 */
+
+	iPtr->randSeed &= (unsigned long) 0x7fffffff;
+	if ((iPtr->randSeed == 0) || (iPtr->randSeed == 0x7fffffff)) {
+	    iPtr->randSeed ^= 123459876;
+	}
+    }
+
+    /*
+     * Generate the random number using the linear congruential generator
+     * defined by the following recurrence:
+     *		seed = ( IA * seed ) mod IM
+     * where IA is 16807 and IM is (2^31) - 1. The recurrence maps a seed in
+     * the range [1, IM - 1] to a new seed in that same range. The recurrence
+     * maps IM to 0, and maps 0 back to 0, so those two values must not be
+     * allowed as initial values of seed.
+     *
+     * In order to avoid potential problems with integer overflow, the
+     * recurrence is implemented in terms of additional constants IQ and IR
+     * such that
+     *		IM = IA*IQ + IR
+     * None of the operations in the implementation overflows a 32-bit signed
+     * integer, and the C type long is guaranteed to be at least 32 bits wide.
+     *
+     * For more details on how this algorithm works, refer to the following
+     * papers:
+     *
+     *	S.K. Park & K.W. Miller, "Random number generators: good ones are hard
+     *	to find," Comm ACM 31(10):1192-1201, Oct 1988
+     *
+     *	W.H. Press & S.A. Teukolsky, "Portable random number generators,"
+     *	Computers in Physics 6(5):522-524, Sep/Oct 1992.
+     */
+
+#define RAND_IA		16807
+#define RAND_IM		2147483647
+#define RAND_IQ		127773
+#define RAND_IR		2836
+#define RAND_MASK	123459876
+
+    tmp = iPtr->randSeed/RAND_IQ;
+    iPtr->randSeed = RAND_IA*(iPtr->randSeed - tmp*RAND_IQ) - RAND_IR*tmp;
+    if (iPtr->randSeed < 0) {
+	iPtr->randSeed += RAND_IM;
+    }
+
+    /*
+     * Since the recurrence keeps seed values in the range [1, RAND_IM - 1],
+     * dividing by RAND_IM yields a double in the range (0, 1).
+     */
+
+    dResult = iPtr->randSeed * (1.0/RAND_IM);
+
+    /*
+     * Push a Tcl object with the result.
+     */
+
+    TclNewDoubleObj(oResult, dResult);
+    Tcl_SetObjResult(interp, oResult);
+    return TCL_OK;
+}
+
+static int
+ExprRoundFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Parameter vector. */
+{
+    double d;
+    ClientData ptr;
+    int type;
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+
+    if (TclGetNumberFromObj(interp, objv[1], &ptr, &type) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (type == TCL_NUMBER_DOUBLE) {
+	double fractPart, intPart;
+	long max = LONG_MAX, min = LONG_MIN;
+
+	fractPart = modf(*((const double *) ptr), &intPart);
+	if (fractPart <= -0.5) {
+	    min++;
+	} else if (fractPart >= 0.5) {
+	    max--;
+	}
+	if ((intPart >= (double)max) || (intPart <= (double)min)) {
+	    mp_int big;
+
+	    if (Tcl_InitBignumFromDouble(interp, intPart, &big) != TCL_OK) {
+		/* Infinity */
+		return TCL_ERROR;
+	    }
+	    if (fractPart <= -0.5) {
+		mp_sub_d(&big, 1, &big);
+	    } else if (fractPart >= 0.5) {
+		mp_add_d(&big, 1, &big);
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewBignumObj(&big));
+	    return TCL_OK;
+	} else {
+	    long result = (long)intPart;
+
+	    if (fractPart <= -0.5) {
+		result--;
+	    } else if (fractPart >= 0.5) {
+		result++;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewLongObj(result));
+	    return TCL_OK;
+	}
+    }
+
+    if (type != TCL_NUMBER_NAN) {
+	/*
+	 * All integers are already rounded
+	 */
+
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    /*
+     * Get the error message for NaN.
+     */
+
+    Tcl_GetDoubleFromObj(interp, objv[1], &d);
+    return TCL_ERROR;
+}
+
+static int
+ExprSrandFunc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* The interpreter in which to execute the
+				 * function. */
+    int objc,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Parameter vector. */
+{
+    Interp *iPtr = (Interp *) interp;
+    long i = 0;			/* Initialized to avoid compiler warning. */
+
+    /*
+     * Convert argument and use it to reset the seed.
+     */
+
+    if (objc != 2) {
+	MathFuncWrongNumArgs(interp, 2, objc, objv);
+	return TCL_ERROR;
+    }
+
+    if (TclGetLongFromObj(NULL, objv[1], &i) != TCL_OK) {
+	Tcl_Obj *objPtr;
+	mp_int big;
+
+	if (Tcl_GetBignumFromObj(interp, objv[1], &big) != TCL_OK) {
+	    /* TODO: more ::errorInfo here? or in caller? */
+	    return TCL_ERROR;
+	}
+
+	mp_mod_2d(&big, (int) CHAR_BIT * sizeof(long), &big);
+	objPtr = Tcl_NewBignumObj(&big);
+	Tcl_IncrRefCount(objPtr);
+	TclGetLongFromObj(NULL, objPtr, &i);
+	Tcl_DecrRefCount(objPtr);
+    }
+
+    /*
+     * Reset the seed. Make sure 1 <= randSeed <= 2^31 - 2. See comments in
+     * ExprRandFunc for more details.
+     */
+
+    iPtr->flags |= RAND_SEED_INITIALIZED;
+    iPtr->randSeed = i;
+    iPtr->randSeed &= (unsigned long) 0x7fffffff;
+    if ((iPtr->randSeed == 0) || (iPtr->randSeed == 0x7fffffff)) {
+	iPtr->randSeed ^= 123459876;
+    }
+
+    /*
+     * To avoid duplicating the random number generation code we simply clean
+     * up our state and call the real random number function. That function
+     * will always succeed.
+     */
+
+    return ExprRandFunc(clientData, interp, 1, objv);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MathFuncWrongNumArgs --
+ *
+ *	Generate an error message when a math function presents the wrong
+ *	number of arguments.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	An error message is stored in the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MathFuncWrongNumArgs(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int expected,		/* Formal parameter count. */
+    int found,			/* Actual parameter count. */
+    Tcl_Obj *const *objv)	/* Actual parameter vector. */
+{
+    const char *name = TclGetString(objv[0]);
+    const char *tail = name + strlen(name);
+
+    while (tail > name+1) {
+	tail--;
+	if (*tail == ':' && tail[-1] == ':') {
+	    name = tail+1;
+	    break;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "too %s arguments for math function \"%s\"",
+	    (found < expected ? "few" : "many"), name));
+    Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+}
+
+#ifdef USE_DTRACE
+/*
+ *----------------------------------------------------------------------
+ *
+ * DTraceObjCmd --
+ *
+ *	This function is invoked to process the "::tcl::dtrace" Tcl command.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	The 'tcl-probe' DTrace probe is triggered (if it is enabled).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DTraceObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (TCL_DTRACE_TCL_PROBE_ENABLED()) {
+	char *a[10];
+	int i = 0;
+
+	while (i++ < 10) {
+	    a[i-1] = i < objc ? TclGetString(objv[i]) : NULL;
+	}
+	TCL_DTRACE_TCL_PROBE(a[0], a[1], a[2], a[3], a[4], a[5], a[6], a[7],
+		a[8], a[9]);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDTraceInfo --
+ *
+ *	Extract information from a TIP280 dict for use by DTrace probes.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDTraceInfo(
+    Tcl_Obj *info,
+    const char **args,
+    int *argsi)
+{
+    static Tcl_Obj *keys[10] = { NULL };
+    Tcl_Obj **k = keys, *val;
+    int i = 0;
+
+    if (!*k) {
+#define kini(s) TclNewLiteralStringObj(keys[i], s); i++
+	kini("cmd");	kini("type");	kini("proc");	kini("file");
+	kini("method");	kini("class");	kini("lambda");	kini("object");
+	kini("line");	kini("level");
+#undef kini
+    }
+    for (i = 0; i < 6; i++) {
+	Tcl_DictObjGet(NULL, info, *k++, &val);
+	args[i] = val ? TclGetString(val) : NULL;
+    }
+    /* no "proc" -> use "lambda" */
+    if (!args[2]) {
+	Tcl_DictObjGet(NULL, info, *k, &val);
+	args[2] = val ? TclGetString(val) : NULL;
+    }
+    k++;
+    /* no "class" -> use "object" */
+    if (!args[5]) {
+	Tcl_DictObjGet(NULL, info, *k, &val);
+	args[5] = val ? TclGetString(val) : NULL;
+    }
+    k++;
+    for (i = 0; i < 2; i++) {
+	Tcl_DictObjGet(NULL, info, *k++, &val);
+	if (val) {
+	    TclGetIntFromObj(NULL, val, &argsi[i]);
+	} else {
+	    argsi[i] = 0;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DTraceCmdReturn --
+ *
+ *	NR callback for DTrace command return probes.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DTraceCmdReturn(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    char *cmdName = TclGetString((Tcl_Obj *) data[0]);
+
+    if (TCL_DTRACE_CMD_RETURN_ENABLED()) {
+	TCL_DTRACE_CMD_RETURN(cmdName, result);
+    }
+    if (TCL_DTRACE_CMD_RESULT_ENABLED()) {
+	Tcl_Obj *r = Tcl_GetObjResult(interp);
+
+	TCL_DTRACE_CMD_RESULT(cmdName, result, TclGetString(r), r);
+    }
+    return result;
+}
+
+TCL_DTRACE_DEBUG_LOG()
+
+#endif /* USE_DTRACE */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NRCallObjProc --
+ *
+ *	This function calls an objProc directly while managing things properly
+ *	if it happens to be an NR objProc. It is meant to be used by extenders
+ *	that provide an NR implementation of a command, as this function
+ *	permits a trivial coding of the non-NR objProc.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR. A result or error message is left in interp's result.
+ *
+ * Side effects:
+ *	Depends on the objProc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_NRCallObjProc(
+    Tcl_Interp *interp,
+    Tcl_ObjCmdProc *objProc,
+    ClientData clientData,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    TclNRAddCallback(interp, Dispatch, objProc, clientData,
+	    INT2PTR(objc), objv);
+    return TclNRRunCallbacks(interp, TCL_OK, rootPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NRCreateCommand --
+ *
+ *	Define a new NRE-enabled object-based command in a command table.
+ *
+ * Results:
+ *	The return value is a token for the command, which can be used in
+ *	future calls to Tcl_GetCommandName.
+ *
+ * Side effects:
+ *	If no command named "cmdName" already exists for interp, one is
+ *	created. Otherwise, if a command does exist, then if the object-based
+ *	Tcl_ObjCmdProc is TclInvokeStringCommand, we assume Tcl_CreateCommand
+ *	was called previously for the same command and just set its
+ *	Tcl_ObjCmdProc to the argument "proc"; otherwise, we delete the old
+ *	command.
+ *
+ *	In the future, during bytecode evaluation when "cmdName" is seen as
+ *	the name of a command by Tcl_EvalObj or Tcl_Eval, the object-based
+ *	Tcl_ObjCmdProc proc will be called. When the command is deleted from
+ *	the table, deleteProc will be called. See the manual entry for details
+ *	on the calling sequence.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_NRCreateCommand(
+    Tcl_Interp *interp,		/* Token for command interpreter (returned by
+				 * previous call to Tcl_CreateInterp). */
+    const char *cmdName,	/* Name of command. If it contains namespace
+				 * qualifiers, the new command is put in the
+				 * specified namespace; otherwise it is put in
+				 * the global namespace. */
+    Tcl_ObjCmdProc *proc,	/* Object-based function to associate with
+				 * name, provides direct access for direct
+				 * calls. */
+    Tcl_ObjCmdProc *nreProc,	/* Object-based function to associate with
+				 * name, provides NR implementation */
+    ClientData clientData,	/* Arbitrary value to pass to object
+				 * function. */
+    Tcl_CmdDeleteProc *deleteProc)
+				/* If not NULL, gives a function to call when
+				 * this command is deleted. */
+{
+    Command *cmdPtr = (Command *)
+	    Tcl_CreateObjCommand(interp,cmdName,proc,clientData,deleteProc);
+
+    cmdPtr->nreProc = nreProc;
+    return (Tcl_Command) cmdPtr;
+}
+
+/****************************************************************************
+ * Stuff for the public api
+ ****************************************************************************/
+
+int
+Tcl_NREvalObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    int flags)
+{
+    return TclNREvalObjEx(interp, objPtr, flags, NULL, INT_MIN);
+}
+
+int
+Tcl_NREvalObjv(
+    Tcl_Interp *interp,		/* Interpreter in which to evaluate the
+				 * command. Also used for error reporting. */
+    int objc,			/* Number of words in command. */
+    Tcl_Obj *const objv[],	/* An array of pointers to objects that are
+				 * the words that make up the command. */
+    int flags)			/* Collection of OR-ed bits that control the
+				 * evaluation of the script. Only
+				 * TCL_EVAL_GLOBAL, TCL_EVAL_INVOKE and
+				 * TCL_EVAL_NOERR are currently supported. */
+{
+    return TclNREvalObjv(interp, objc, objv, flags, NULL);
+}
+
+int
+Tcl_NRCmdSwap(
+    Tcl_Interp *interp,
+    Tcl_Command cmd,
+    int objc,
+    Tcl_Obj *const objv[],
+    int flags)
+{
+    return TclNREvalObjv(interp, objc, objv, flags|TCL_EVAL_NOERR,
+	    (Command *) cmd);
+}
+
+/*****************************************************************************
+ * Tailcall related code
+ *****************************************************************************
+ *
+ * The steps of the tailcall dance are as follows:
+ *
+ *   1. when [tailcall] is invoked, it stores the corresponding callback in
+ *      the current CallFrame and returns TCL_RETURN
+ *   2. when the CallFrame is popped, it calls TclSetTailcall to store the
+ *      callback in the proper NRCommand callback - the spot where the command
+ *      that pushed the CallFrame is completely cleaned up
+ *   3. when the NRCommand callback runs, it schedules the tailcall callback
+ *      to run immediately after it returns
+ *
+ *   One delicate point is to properly define the NRCommand where the tailcall
+ *   will execute. There are functions whose purpose is to help define the
+ *   precise spot:
+ *     TclMarkTailcall: if the NEXT command to be pushed tailcalls, execution
+ *         should continue right here
+ *     TclSkipTailcall:  if the NEXT command to be pushed tailcalls, execution
+ *         should continue after the CURRENT command is fully returned ("skip
+ *         the next command: we are redirecting to it, tailcalls should run
+ *         after WE return")
+ *     TclPushTailcallPoint: the search for a tailcalling spot cannot traverse
+ *         this point. This is special for OO, as some of the oo constructs
+ *         that behave like commands may not push an NRCommand callback.
+ */
+
+void
+TclMarkTailcall(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (iPtr->deferredCallbacks == NULL) {
+	TclNRAddCallback(interp, NRCommand, NULL, NULL,
+                NULL, NULL);
+        iPtr->deferredCallbacks = TOP_CB(interp);
+    }
+}
+
+void
+TclSkipTailcall(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    TclMarkTailcall(interp);
+    iPtr->deferredCallbacks->data[1] = INT2PTR(1);
+}
+
+void
+TclPushTailcallPoint(
+    Tcl_Interp *interp)
+{
+    TclNRAddCallback(interp, NRCommand, NULL, NULL, NULL, NULL);
+    ((Interp *) interp)->numLevels++;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetTailcall --
+ *
+ *	Splice a tailcall command in the proper spot of the NRE callback
+ *	stack, so that it runs at the right time.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetTailcall(
+    Tcl_Interp *interp,
+    Tcl_Obj *listPtr)
+{
+    /*
+     * Find the splicing spot: right before the NRCommand of the thing
+     * being tailcalled. Note that we skip NRCommands marked by a 1 in data[1]
+     * (used by command redirectors).
+     */
+
+    NRE_callback *runPtr;
+
+    for (runPtr = TOP_CB(interp); runPtr; runPtr = runPtr->nextPtr) {
+        if (((runPtr->procPtr) == NRCommand) && !runPtr->data[1]) {
+            break;
+        }
+    }
+    if (!runPtr) {
+        Tcl_Panic("tailcall cannot find the right splicing spot: should not happen!");
+    }
+    runPtr->data[1] = listPtr;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRTailcallObjCmd --
+ *
+ *	Prepare the tailcall as a list and store it in the current
+ *	varFrame. When the frame is later popped the tailcall will be spliced
+ *	at the proper place.
+ *
+ * Results:
+ *	The first NRCommand callback that is not marked to be skipped is
+ *	updated so that its data[1] field contains the tailcall list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNRTailcallObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (objc < 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?command? ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    if (!(iPtr->varFramePtr->isProcCallFrame & 1)) {
+        Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "tailcall can only be called from a proc, lambda or method", -1));
+        Tcl_SetErrorCode(interp, "TCL", "TAILCALL", "ILLEGAL", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Invocation without args just clears a scheduled tailcall; invocation
+     * with an argument replaces any previously scheduled tailcall.
+     */
+
+    if (iPtr->varFramePtr->tailcallPtr) {
+        Tcl_DecrRefCount(iPtr->varFramePtr->tailcallPtr);
+        iPtr->varFramePtr->tailcallPtr = NULL;
+    }
+
+    /*
+     * Create the callback to actually evaluate the tailcalled
+     * command, then set it in the varFrame so that PopCallFrame can use it
+     * at the proper time.
+     */
+
+    if (objc > 1) {
+        Tcl_Obj *listPtr, *nsObjPtr;
+        Tcl_Namespace *nsPtr = (Tcl_Namespace *) iPtr->varFramePtr->nsPtr;
+        Tcl_Namespace *ns1Ptr;
+
+        /* The tailcall data is in a Tcl list: the first element is the
+         * namespace, the rest the command to be tailcalled. */
+
+        listPtr = Tcl_NewListObj(objc, objv);
+
+        nsObjPtr = Tcl_NewStringObj(nsPtr->fullName, -1);
+        if ((TCL_OK != TclGetNamespaceFromObj(interp, nsObjPtr, &ns1Ptr))
+                || (nsPtr != ns1Ptr)) {
+            Tcl_Panic("Tailcall failed to find the proper namespace");
+        }
+ 	TclListObjSetElement(interp, listPtr, 0, nsObjPtr);
+
+        iPtr->varFramePtr->tailcallPtr = listPtr;
+    }
+    return TCL_RETURN;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRTailcallEval --
+ *
+ *	This NREcallback actually causes the tailcall to be evaluated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNRTailcallEval(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *listPtr = data[0], *nsObjPtr;
+    Tcl_Namespace *nsPtr;
+    int objc;
+    Tcl_Obj **objv;
+
+    Tcl_ListObjGetElements(interp, listPtr, &objc, &objv);
+    nsObjPtr = objv[0];
+
+    if (result == TCL_OK) {
+	result = TclGetNamespaceFromObj(interp, nsObjPtr, &nsPtr);
+    }
+
+    if (result != TCL_OK) {
+        /*
+         * Tailcall execution was preempted, eg by an intervening catch or by
+         * a now-gone namespace: cleanup and return.
+         */
+
+	Tcl_DecrRefCount(listPtr);
+        return result;
+    }
+
+    /*
+     * Perform the tailcall
+     */
+
+    TclMarkTailcall(interp);
+    TclNRAddCallback(interp, TclNRReleaseValues, listPtr, NULL, NULL,NULL);
+    iPtr->lookupNsPtr = (Namespace *) nsPtr;
+    return TclNREvalObjv(interp, objc-1, objv+1, 0, NULL);
+}
+
+int
+TclNRReleaseValues(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    int i = 0;
+    while (i < 4) {
+	if (data[i]) {
+	    Tcl_DecrRefCount((Tcl_Obj *) data[i]);
+	} else {
+	    break;
+	}
+	i++;
+    }
+    return result;
+}
+
+
+void
+Tcl_NRAddCallback(
+    Tcl_Interp *interp,
+    Tcl_NRPostProc *postProcPtr,
+    ClientData data0,
+    ClientData data1,
+    ClientData data2,
+    ClientData data3)
+{
+    if (!(postProcPtr)) {
+	Tcl_Panic("Adding a callback without an objProc?!");
+    }
+    TclNRAddCallback(interp, postProcPtr, data0, data1, data2, data3);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRCoroutineObjCmd -- (and friends)
+ *
+ *	This object-based function is invoked to process the "coroutine" Tcl
+ *	command. It is heavily based on "apply".
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	A new procedure gets created.
+ *
+ * ** FIRST EXPERIMENTAL IMPLEMENTATION **
+ *
+ * It is fairly amateurish and not up to our standards - mainly in terms of
+ * error messages and [info] interaction. Just to test the infrastructure in
+ * teov and tebc.
+ *----------------------------------------------------------------------
+ */
+
+#define iPtr ((Interp *) interp)
+
+int
+TclNRYieldObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    CoroutineData *corPtr = iPtr->execEnvPtr->corPtr;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?returnValue?");
+	return TCL_ERROR;
+    }
+
+    if (!corPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "yield can only be called in a coroutine", -1));
+	Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "ILLEGAL_YIELD", NULL);
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	Tcl_SetObjResult(interp, objv[1]);
+    }
+
+    NRE_ASSERT(!COR_IS_SUSPENDED(corPtr));
+    TclNRAddCallback(interp, TclNRCoroutineActivateCallback, corPtr,
+            clientData, NULL, NULL);
+    return TCL_OK;
+}
+
+int
+TclNRYieldToObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    CoroutineData *corPtr = iPtr->execEnvPtr->corPtr;
+    Tcl_Obj *listPtr, *nsObjPtr;
+    Tcl_Namespace *nsPtr = TclGetCurrentNamespace(interp);
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "command ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    if (!corPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "yieldto can only be called in a coroutine", -1));
+	Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "ILLEGAL_YIELD", NULL);
+	return TCL_ERROR;
+    }
+
+    if (((Namespace *) nsPtr)->flags & NS_DYING) {
+        Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"yieldto called in deleted namespace", -1));
+        Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "YIELDTO_IN_DELETED",
+		NULL);
+        return TCL_ERROR;
+    }
+
+    /*
+     * Add the tailcall in the caller env, then just yield.
+     *
+     * This is essentially code from TclNRTailcallObjCmd
+     */
+
+    listPtr = Tcl_NewListObj(objc, objv);
+    nsObjPtr = Tcl_NewStringObj(nsPtr->fullName, -1);
+    TclListObjSetElement(interp, listPtr, 0, nsObjPtr);
+
+    /*
+     * Add the callback in the caller's env, then instruct TEBC to yield.
+     */
+
+    iPtr->execEnvPtr = corPtr->callerEEPtr;
+    TclSetTailcall(interp, listPtr);
+    iPtr->execEnvPtr = corPtr->eePtr;
+
+    return TclNRYieldObjCmd(INT2PTR(CORO_ACTIVATE_YIELDM), interp, 1, objv);
+}
+
+static int
+RewindCoroutineCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    return Tcl_RestoreInterpState(interp, data[0]);
+}
+
+static int
+RewindCoroutine(
+    CoroutineData *corPtr,
+    int result)
+{
+    Tcl_Interp *interp = corPtr->eePtr->interp;
+    Tcl_InterpState state = Tcl_SaveInterpState(interp, result);
+
+    NRE_ASSERT(COR_IS_SUSPENDED(corPtr));
+    NRE_ASSERT(corPtr->eePtr != NULL);
+    NRE_ASSERT(corPtr->eePtr != iPtr->execEnvPtr);
+
+    corPtr->eePtr->rewind = 1;
+    TclNRAddCallback(interp, RewindCoroutineCallback, state,
+	    NULL, NULL, NULL);
+    return TclNRInterpCoroutine(corPtr, interp, 0, NULL);
+}
+
+static void
+DeleteCoroutine(
+    ClientData clientData)
+{
+    CoroutineData *corPtr = clientData;
+    Tcl_Interp *interp = corPtr->eePtr->interp;
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    if (COR_IS_SUSPENDED(corPtr)) {
+	TclNRRunCallbacks(interp, RewindCoroutine(corPtr,TCL_OK), rootPtr);
+    }
+}
+
+static int
+NRCoroutineCallerCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CoroutineData *corPtr = data[0];
+    Command *cmdPtr = corPtr->cmdPtr;
+
+    /*
+     * This is the last callback in the caller execEnv, right before switching
+     * to the coroutine's
+     */
+
+    NRE_ASSERT(iPtr->execEnvPtr == corPtr->callerEEPtr);
+
+    if (!corPtr->eePtr) {
+	/*
+	 * The execEnv was wound down but not deleted for our sake. We finish
+	 * the job here. The caller context has already been restored.
+	 */
+
+	NRE_ASSERT(iPtr->varFramePtr == corPtr->caller.varFramePtr);
+	NRE_ASSERT(iPtr->framePtr == corPtr->caller.framePtr);
+	NRE_ASSERT(iPtr->cmdFramePtr == corPtr->caller.cmdFramePtr);
+	ckfree(corPtr);
+	return result;
+    }
+
+    NRE_ASSERT(COR_IS_SUSPENDED(corPtr));
+    SAVE_CONTEXT(corPtr->running);
+    RESTORE_CONTEXT(corPtr->caller);
+
+    if (cmdPtr->flags & CMD_IS_DELETED) {
+	/*
+	 * The command was deleted while it was running: wind down the
+	 * execEnv, this will do the complete cleanup. RewindCoroutine will
+	 * restore both the caller's context and interp state.
+	 */
+
+	return RewindCoroutine(corPtr, result);
+    }
+
+    return result;
+}
+
+static int
+NRCoroutineExitCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CoroutineData *corPtr = data[0];
+    Command *cmdPtr = corPtr->cmdPtr;
+
+    /*
+     * This runs at the bottom of the Coroutine's execEnv: it will be executed
+     * when the coroutine returns or is wound down, but not when it yields. It
+     * deletes the coroutine and restores the caller's environment.
+     */
+
+    NRE_ASSERT(interp == corPtr->eePtr->interp);
+    NRE_ASSERT(TOP_CB(interp) == NULL);
+    NRE_ASSERT(iPtr->execEnvPtr == corPtr->eePtr);
+    NRE_ASSERT(!COR_IS_SUSPENDED(corPtr));
+    NRE_ASSERT((corPtr->callerEEPtr->callbackPtr->procPtr == NRCoroutineCallerCallback));
+
+    cmdPtr->deleteProc = NULL;
+    Tcl_DeleteCommandFromToken(interp, (Tcl_Command) cmdPtr);
+    TclCleanupCommandMacro(cmdPtr);
+
+    corPtr->eePtr->corPtr = NULL;
+    TclDeleteExecEnv(corPtr->eePtr);
+    corPtr->eePtr = NULL;
+
+    corPtr->stackLevel = NULL;
+
+    /*
+     * #280.
+     * Drop the coroutine-owned copy of the lineLABCPtr hashtable for literal
+     * command arguments in bytecode.
+     */
+
+    Tcl_DeleteHashTable(corPtr->lineLABCPtr);
+    ckfree(corPtr->lineLABCPtr);
+    corPtr->lineLABCPtr = NULL;
+
+    RESTORE_CONTEXT(corPtr->caller);
+    iPtr->execEnvPtr = corPtr->callerEEPtr;
+    iPtr->numLevels++;
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRCoroutineActivateCallback --
+ *
+ *      This is the workhorse for coroutines: it implements both yield and
+ *      resume.
+ *
+ *      It is important that both be implemented in the same callback: the
+ *      detection of the impossibility to suspend due to a busy C-stack relies
+ *      on the precise position of a local variable in the stack. We do not
+ *      want the compiler to play tricks on us, either by moving things around
+ *      or inlining.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNRCoroutineActivateCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CoroutineData *corPtr = data[0];
+    int type = PTR2INT(data[1]);
+    int numLevels, unused;
+    int *stackLevel = &unused;
+
+    if (!corPtr->stackLevel) {
+        /*
+         * -- Coroutine is suspended --
+         * Push the callback to restore the caller's context on yield or
+         * return.
+         */
+
+        TclNRAddCallback(interp, NRCoroutineCallerCallback, corPtr,
+                NULL, NULL, NULL);
+
+        /*
+         * Record the stackLevel at which the resume is happening, then swap
+         * the interp's environment to make it suitable to run this coroutine.
+         */
+
+        corPtr->stackLevel = stackLevel;
+        numLevels = corPtr->auxNumLevels;
+        corPtr->auxNumLevels = iPtr->numLevels;
+
+        SAVE_CONTEXT(corPtr->caller);
+        corPtr->callerEEPtr = iPtr->execEnvPtr;
+        RESTORE_CONTEXT(corPtr->running);
+        iPtr->execEnvPtr = corPtr->eePtr;
+        iPtr->numLevels += numLevels;
+    } else {
+        /*
+         * Coroutine is active: yield
+         */
+
+        if (corPtr->stackLevel != stackLevel) {
+            Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "cannot yield: C stack busy", -1));
+            Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "CANT_YIELD",
+                    NULL);
+            return TCL_ERROR;
+        }
+
+        if (type == CORO_ACTIVATE_YIELD) {
+            corPtr->nargs = COROUTINE_ARGUMENTS_SINGLE_OPTIONAL;
+        } else if (type == CORO_ACTIVATE_YIELDM) {
+            corPtr->nargs = COROUTINE_ARGUMENTS_ARBITRARY;
+        } else {
+            Tcl_Panic("Yield received an option which is not implemented");
+        }
+
+        corPtr->stackLevel = NULL;
+
+        numLevels = iPtr->numLevels;
+        iPtr->numLevels = corPtr->auxNumLevels;
+        corPtr->auxNumLevels = numLevels - corPtr->auxNumLevels;
+
+        iPtr->execEnvPtr = corPtr->callerEEPtr;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNREvalList --
+ *
+ *      Callback to invoke command as list, used in order to delayed
+ *	processing of canonical list command in sane environment.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TclNREvalList(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    int objc;
+    Tcl_Obj **objv;
+    Tcl_Obj *listPtr = data[0];
+
+    Tcl_IncrRefCount(listPtr);
+
+    TclMarkTailcall(interp);
+    TclNRAddCallback(interp, TclNRReleaseValues, listPtr, NULL, NULL,NULL);
+    TclListObjGetElements(NULL, listPtr, &objc, &objv);
+    return TclNREvalObjv(interp, objc, objv, 0, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NRCoroInjectObjCmd --
+ *
+ *      Implementation of [::tcl::unsupported::inject] command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NRCoroInjectObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Command *cmdPtr;
+    CoroutineData *corPtr;
+    ExecEnv *savedEEPtr = iPtr->execEnvPtr;
+
+    /*
+     * Usage more or less like tailcall:
+     *   inject coroName cmd ?arg1 arg2 ...?
+     */
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "coroName cmd ?arg1 arg2 ...?");
+	return TCL_ERROR;
+    }
+
+    cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, objv[1]);
+    if ((!cmdPtr) || (cmdPtr->nreProc != TclNRInterpCoroutine)) {
+        Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "can only inject a command into a coroutine", -1));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COROUTINE",
+                TclGetString(objv[1]), NULL);
+        return TCL_ERROR;
+    }
+
+    corPtr = cmdPtr->objClientData;
+    if (!COR_IS_SUSPENDED(corPtr)) {
+        Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "can only inject a command into a suspended coroutine", -1));
+        Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "ACTIVE", NULL);
+        return TCL_ERROR;
+    }
+
+    /*
+     * Add the callback to the coro's execEnv, so that it is the first thing
+     * to happen when the coro is resumed.
+     */
+
+    iPtr->execEnvPtr = corPtr->eePtr;
+    TclNRAddCallback(interp, TclNREvalList, Tcl_NewListObj(objc-2, objv+2),
+	NULL, NULL, NULL);
+    iPtr->execEnvPtr = savedEEPtr;
+
+    return TCL_OK;
+}
+
+int
+TclNRInterpCoroutine(
+    ClientData clientData,
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    CoroutineData *corPtr = clientData;
+
+    if (!COR_IS_SUSPENDED(corPtr)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "coroutine \"%s\" is already running",
+                TclGetString(objv[0])));
+	Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "BUSY", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse all the arguments to work out what to feed as the result of the
+     * [yield]. TRICKY POINT: objc==0 happens here! It occurs when a coroutine
+     * is deleted!
+     */
+
+    switch (corPtr->nargs) {
+    case COROUTINE_ARGUMENTS_SINGLE_OPTIONAL:
+        if (objc == 2) {
+            Tcl_SetObjResult(interp, objv[1]);
+        } else if (objc > 2) {
+            Tcl_WrongNumArgs(interp, 1, objv, "?arg?");
+            return TCL_ERROR;
+        }
+        break;
+    default:
+        if (corPtr->nargs != objc-1) {
+            Tcl_SetObjResult(interp,
+                    Tcl_NewStringObj("wrong coro nargs; how did we get here? "
+                    "not implemented!", -1));
+            Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+            return TCL_ERROR;
+        }
+        /* fallthrough */
+    case COROUTINE_ARGUMENTS_ARBITRARY:
+        if (objc > 1) {
+            Tcl_SetObjResult(interp, Tcl_NewListObj(objc-1, objv+1));
+        }
+        break;
+    }
+
+    TclNRAddCallback(interp, TclNRCoroutineActivateCallback, corPtr,
+            NULL, NULL, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRCoroutineObjCmd --
+ *
+ *      Implementation of [coroutine] command; see documentation for
+ *      description of what this does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNRCoroutineObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Command *cmdPtr;
+    CoroutineData *corPtr;
+    const char *fullName, *procName;
+    Namespace *nsPtr, *altNsPtr, *cxtNsPtr;
+    Tcl_DString ds;
+    Namespace *lookupNsPtr = iPtr->varFramePtr->nsPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name cmd ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * FIXME: this is copy/pasted from Tcl_ProcObjCommand. Should have
+     * something in tclUtil.c to find the FQ name.
+     */
+
+    fullName = TclGetString(objv[1]);
+    TclGetNamespaceForQualName(interp, fullName, NULL, 0,
+	    &nsPtr, &altNsPtr, &cxtNsPtr, &procName);
+
+    if (nsPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't create procedure \"%s\": unknown namespace",
+                fullName));
+        Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE", NULL);
+	return TCL_ERROR;
+    }
+    if (procName == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't create procedure \"%s\": bad procedure name",
+                fullName));
+        Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", fullName, NULL);
+	return TCL_ERROR;
+    }
+    if ((nsPtr != iPtr->globalNsPtr)
+	    && (procName != NULL) && (procName[0] == ':')) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't create procedure \"%s\" in non-global namespace with"
+                " name starting with \":\"", procName));
+        Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", procName, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * We ARE creating the coroutine command: allocate the corresponding
+     * struct and create the corresponding command.
+     */
+
+    corPtr = ckalloc(sizeof(CoroutineData));
+
+    Tcl_DStringInit(&ds);
+    if (nsPtr != iPtr->globalNsPtr) {
+	Tcl_DStringAppend(&ds, nsPtr->fullName, -1);
+	TclDStringAppendLiteral(&ds, "::");
+    }
+    Tcl_DStringAppend(&ds, procName, -1);
+
+    cmdPtr = (Command *) Tcl_NRCreateCommand(interp, Tcl_DStringValue(&ds),
+	    /*objProc*/ NULL, TclNRInterpCoroutine, corPtr, DeleteCoroutine);
+    Tcl_DStringFree(&ds);
+
+    corPtr->cmdPtr = cmdPtr;
+    cmdPtr->refCount++;
+
+    /*
+     * #280.
+     * Provide the new coroutine with its own copy of the lineLABCPtr
+     * hashtable for literal command arguments in bytecode. Note that that
+     * CFWordBC chains are not duplicated, only the entrypoints to them. This
+     * means that in the presence of coroutines each chain is potentially a
+     * tree. Like the chain -> tree conversion of the CmdFrame stack.
+     */
+
+    {
+	Tcl_HashSearch hSearch;
+	Tcl_HashEntry *hePtr;
+
+	corPtr->lineLABCPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(corPtr->lineLABCPtr, TCL_ONE_WORD_KEYS);
+
+	for (hePtr = Tcl_FirstHashEntry(iPtr->lineLABCPtr,&hSearch);
+		hePtr; hePtr = Tcl_NextHashEntry(&hSearch)) {
+	    int isNew;
+	    Tcl_HashEntry *newPtr =
+		    Tcl_CreateHashEntry(corPtr->lineLABCPtr,
+		    Tcl_GetHashKey(iPtr->lineLABCPtr, hePtr),
+		    &isNew);
+
+	    Tcl_SetHashValue(newPtr, Tcl_GetHashValue(hePtr));
+	}
+    }
+
+    /*
+     * Create the base context.
+     */
+
+    corPtr->running.framePtr = iPtr->rootFramePtr;
+    corPtr->running.varFramePtr = iPtr->rootFramePtr;
+    corPtr->running.cmdFramePtr = NULL;
+    corPtr->running.lineLABCPtr = corPtr->lineLABCPtr;
+    corPtr->stackLevel = NULL;
+    corPtr->auxNumLevels = 0;
+
+    /*
+     * Create the coro's execEnv, switch to it to push the exit and coro
+     * command callbacks, then switch back.
+     */
+
+    corPtr->eePtr = TclCreateExecEnv(interp, CORO_STACK_INITIAL_SIZE);
+    corPtr->callerEEPtr = iPtr->execEnvPtr;
+    corPtr->eePtr->corPtr = corPtr;
+
+    SAVE_CONTEXT(corPtr->caller);
+    corPtr->callerEEPtr = iPtr->execEnvPtr;
+    RESTORE_CONTEXT(corPtr->running);
+    iPtr->execEnvPtr = corPtr->eePtr;
+
+    TclNRAddCallback(interp, NRCoroutineExitCallback, corPtr,
+	    NULL, NULL, NULL);
+
+    /* insure that the command is looked up in the correct namespace */
+    iPtr->lookupNsPtr = lookupNsPtr;
+    Tcl_NREvalObj(interp, Tcl_NewListObj(objc-2, objv+2), 0);
+    iPtr->numLevels--;
+
+    SAVE_CONTEXT(corPtr->running);
+    RESTORE_CONTEXT(corPtr->caller);
+    iPtr->execEnvPtr = corPtr->callerEEPtr;
+
+    /*
+     * Now just resume the coroutine.
+     */
+
+    TclNRAddCallback(interp, TclNRCoroutineActivateCallback, corPtr,
+            NULL, NULL, NULL);
+    return TCL_OK;
+}
+
+/*
+ * This is used in the [info] ensemble
+ */
+
+int
+TclInfoCoroutineCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    CoroutineData *corPtr = iPtr->execEnvPtr->corPtr;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    if (corPtr && !(corPtr->cmdPtr->flags & CMD_IS_DELETED)) {
+	Tcl_Obj *namePtr;
+
+	TclNewObj(namePtr);
+	Tcl_GetCommandFullName(interp, (Tcl_Command) corPtr->cmdPtr, namePtr);
+	Tcl_SetObjResult(interp, namePtr);
+    }
+    return TCL_OK;
+}
+
+#undef iPtr
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclBinary.c b/generic/tclBinary.c
new file mode 100644
index 0000000..a693894
--- /dev/null
+++ b/generic/tclBinary.c
@@ -0,0 +1,3086 @@
+/*
+ * tclBinary.c --
+ *
+ *	This file contains the implementation of the "binary" Tcl built-in
+ *	command and the Tcl binary data object.
+ *
+ * Copyright (c) 1997 by Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+
+#include <math.h>
+
+/*
+ * The following constants are used by GetFormatSpec to indicate various
+ * special conditions in the parsing of a format specifier.
+ */
+
+#define BINARY_ALL -1		/* Use all elements in the argument. */
+#define BINARY_NOCOUNT -2	/* No count was specified in format. */
+
+/*
+ * The following flags may be ORed together and returned by GetFormatSpec
+ */
+
+#define BINARY_SIGNED 0		/* Field to be read as signed data */
+#define BINARY_UNSIGNED 1	/* Field to be read as unsigned data */
+
+/*
+ * The following defines the maximum number of different (integer) numbers
+ * placed in the object cache by 'binary scan' before it bails out and
+ * switches back to Plan A (creating a new object for each value.)
+ * Theoretically, it would be possible to keep the cache about for the values
+ * that are already in it, but that makes the code slower in practise when
+ * overflow happens, and makes little odds the rest of the time (as measured
+ * on my machine.) It is also slower (on the sample I tried at least) to grow
+ * the cache to hold all items we might want to put in it; presumably the
+ * extra cost of managing the memory for the enlarged table outweighs the
+ * benefit from allocating fewer objects. This is probably because as the
+ * number of objects increases, the likelihood of reuse of any particular one
+ * drops, and there is very little gain from larger maximum cache sizes (the
+ * value below is chosen to allow caching to work in full with conversion of
+ * bytes.) - DKF
+ */
+
+#define BINARY_SCAN_MAX_CACHE	260
+
+/*
+ * Prototypes for local procedures defined in this file:
+ */
+
+static void		DupByteArrayInternalRep(Tcl_Obj *srcPtr,
+			    Tcl_Obj *copyPtr);
+static int		FormatNumber(Tcl_Interp *interp, int type,
+			    Tcl_Obj *src, unsigned char **cursorPtr);
+static void		FreeByteArrayInternalRep(Tcl_Obj *objPtr);
+static int		GetFormatSpec(const char **formatPtr, char *cmdPtr,
+			    int *countPtr, int *flagsPtr);
+static Tcl_Obj *	ScanNumber(unsigned char *buffer, int type,
+			    int flags, Tcl_HashTable **numberCachePtr);
+static int		SetByteArrayFromAny(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static void		UpdateStringOfByteArray(Tcl_Obj *listPtr);
+static void		DeleteScanNumberCache(Tcl_HashTable *numberCachePtr);
+static int		NeedReversing(int format);
+static void		CopyNumber(const void *from, void *to,
+			    unsigned length, int type);
+/* Binary ensemble commands */
+static int		BinaryFormatCmd(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		BinaryScanCmd(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+/* Binary encoding sub-ensemble commands */
+static int		BinaryEncodeHex(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		BinaryDecodeHex(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		BinaryEncode64(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		BinaryDecode64(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		BinaryEncodeUu(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		BinaryDecodeUu(ClientData clientData,
+			    Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+
+/*
+ * The following tables are used by the binary encoders
+ */
+
+static const char HexDigits[16] = {
+    '0', '1', '2', '3', '4', '5', '6', '7',
+    '8', '9', 'a', 'b', 'c', 'd', 'e', 'f'
+};
+
+static const char UueDigits[65] = {
+    '`', '!', '"', '#', '$', '%', '&', '\'',
+    '(', ')', '*', '+', ',', '-', '.', '/',
+    '0', '1', '2', '3', '4', '5', '6', '7',
+    '8', '9', ':', ';', '<', '=', '>', '?',
+    '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
+    'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
+    'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
+    'X', 'Y', 'Z', '[', '\\',']', '^', '_',
+    '`'
+};
+
+static const char B64Digits[65] = {
+    'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
+    'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
+    'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
+    'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f',
+    'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n',
+    'o', 'p', 'q', 'r', 's', 't', 'u', 'v',
+    'w', 'x', 'y', 'z', '0', '1', '2', '3',
+    '4', '5', '6', '7', '8', '9', '+', '/',
+    '='
+};
+
+/*
+ * How to construct the ensembles.
+ */
+
+static const EnsembleImplMap binaryMap[] = {
+    { "format", BinaryFormatCmd, TclCompileBasicMin1ArgCmd, NULL, NULL, 0 },
+    { "scan",   BinaryScanCmd, TclCompileBasicMin2ArgCmd, NULL, NULL, 0 },
+    { "encode", NULL, NULL, NULL, NULL, 0 },
+    { "decode", NULL, NULL, NULL, NULL, 0 },
+    { NULL, NULL, NULL, NULL, NULL, 0 }
+};
+static const EnsembleImplMap encodeMap[] = {
+    { "hex",      BinaryEncodeHex, TclCompileBasic1ArgCmd, NULL, NULL, 0 },
+    { "uuencode", BinaryEncodeUu,  NULL, NULL, NULL, 0 },
+    { "base64",   BinaryEncode64,  NULL, NULL, NULL, 0 },
+    { NULL, NULL, NULL, NULL, NULL, 0 }
+};
+static const EnsembleImplMap decodeMap[] = {
+    { "hex",      BinaryDecodeHex, TclCompileBasic1Or2ArgCmd, NULL, NULL, 0 },
+    { "uuencode", BinaryDecodeUu,  TclCompileBasic1Or2ArgCmd, NULL, NULL, 0 },
+    { "base64",   BinaryDecode64,  TclCompileBasic1Or2ArgCmd, NULL, NULL, 0 },
+    { NULL, NULL, NULL, NULL, NULL, 0 }
+};
+
+/*
+ * The following object types represent an array of bytes. The intent is
+ * to allow arbitrary binary data to pass through Tcl as a Tcl value
+ * without loss or damage. Such values are useful for things like
+ * encoded strings or Tk images to name just two.
+ *
+ * It's strange to have two Tcl_ObjTypes in place for this task when
+ * one would do, so a bit of detail and history how we got to this point
+ * and where we might go from here.
+ *
+ * A bytearray is an ordered sequence of bytes. Each byte is an integer
+ * value in the range [0-255].  To be a Tcl value type, we need a way to
+ * encode each value in the value set as a Tcl string.  The simplest
+ * encoding is to represent each byte value as the same codepoint value.
+ * A bytearray of N bytes is encoded into a Tcl string of N characters
+ * where the codepoint of each character is the value of corresponding byte.
+ * This approach creates a one-to-one map between all bytearray values
+ * and a subset of Tcl string values.
+ *
+ * When converting a Tcl string value to the bytearray internal rep, the
+ * question arises what to do with strings outside that subset?  That is,
+ * those Tcl strings containing at least one codepoint greater than 255?
+ * The obviously correct answer is to raise an error!  That string value
+ * does not represent any valid bytearray value. Full Stop.  The
+ * setFromAnyProc signature has a completion code return value for just
+ * this reason, to reject invalid inputs.
+ *
+ * Unfortunately this was not the path taken by the authors of the
+ * original tclByteArrayType.  They chose to accept all Tcl string values
+ * as acceptable string encodings of the bytearray values that result
+ * from masking away the high bits of any codepoint value at all. This
+ * meant that every bytearray value had multiple accepted string
+ * representations.
+ *
+ * The implications of this choice are truly ugly.  When a Tcl value has
+ * a string representation, we are required to accept that as the true
+ * value.  Bytearray values that possess a string representation cannot
+ * be processed as bytearrays because we cannot know which true value
+ * that bytearray represents.  The consequence is that we drag around
+ * an internal rep that we cannot make any use of.  This painful price
+ * is extracted at any point after a string rep happens to be generated
+ * for the value.  This happens even when the troublesome codepoints
+ * outside the byte range never show up.  This happens rather routinely
+ * in normal Tcl operations unless we burden the script writer with the
+ * cognitive burden of avoiding it.  The price is also paid by callers
+ * of the C interface.  The routine
+ *
+ *	unsigned char *Tcl_GetByteArrayFromObj(objPtr, lenPtr)
+ *
+ * has a guarantee to always return a non-NULL value, but that value
+ * points to a byte sequence that cannot be used by the caller to
+ * process the Tcl value absent some sideband testing that objPtr
+ * is "pure".  Tcl offers no public interface to perform this test,
+ * so callers either break encapsulation or are unavoidably buggy.  Tcl
+ * has defined a public interface that cannot be used correctly. The
+ * Tcl source code itself suffers the same problem, and has been buggy,
+ * but progressively less so as more and more portions of the code have
+ * been retrofitted with the required "purity testing".  The set of values
+ * able to pass the purity test can be increased via the introduction of
+ * a "canonical" flag marker, but the only way the broken interface itself
+ * can be discarded is to start over and define the Tcl_ObjType properly.
+ * Bytearrays should simply be usable as bytearrays without a kabuki
+ * dance of testing.
+ *
+ * The Tcl_ObjType "properByteArrayType" is (nearly) a correct
+ * implementation of bytearrays.  Any Tcl value with the type
+ * properByteArrayType can have its bytearray value fetched and
+ * used with confidence that acting on that value is equivalent to
+ * acting on the true Tcl string value.  This still implies a side
+ * testing burden -- past mistakes will not let us avoid that
+ * immediately, but it is at least a conventional test of type, and
+ * can be implemented entirely by examining the objPtr fields, with
+ * no need to query the intrep, as a canonical flag would require.
+ *
+ * Until Tcl_GetByteArrayFromObj() and Tcl_SetByteArrayLength() can
+ * be revised to admit the possibility of returning NULL when the true
+ * value is not a valid bytearray, we need a mechanism to retain
+ * compatibility with the deployed callers of the broken interface.
+ * That's what the retained "tclByteArrayType" provides.  In those
+ * unusual circumstances where we convert an invalid bytearray value
+ * to a bytearray type, it is to this legacy type.  Essentially any
+ * time this legacy type gets used, it's a signal of a bug being ignored.
+ * A TIP should be drafted to remove this connection to the broken past
+ * so that Tcl 9 will no longer have any trace of it.  Prescribing a
+ * migration path will be the key element of that work.  The internal
+ * changes now in place are the limit of what can be done short of
+ * interface repair.  They provide a great expansion of the histories
+ * over which bytearray values can be useful in the meanwhile.
+ */
+
+static const Tcl_ObjType properByteArrayType = {
+    "bytearray",
+    FreeByteArrayInternalRep,
+    DupByteArrayInternalRep,
+    UpdateStringOfByteArray,
+    NULL
+};
+
+const Tcl_ObjType tclByteArrayType = {
+    "bytearray",
+    FreeByteArrayInternalRep,
+    DupByteArrayInternalRep,
+    NULL,
+    SetByteArrayFromAny
+};
+
+/*
+ * The following structure is the internal rep for a ByteArray object. Keeps
+ * track of how much memory has been used and how much has been allocated for
+ * the byte array to enable growing and shrinking of the ByteArray object with
+ * fewer mallocs.
+ */
+
+typedef struct ByteArray {
+    int used;			/* The number of bytes used in the byte
+				 * array. */
+    int allocated;		/* The amount of space actually allocated
+				 * minus 1 byte. */
+    unsigned char bytes[1];	/* The array of bytes. The actual size of this
+				 * field depends on the 'allocated' field
+				 * above. */
+} ByteArray;
+
+#define BYTEARRAY_SIZE(len) \
+		((unsigned) (TclOffset(ByteArray, bytes) + (len)))
+#define GET_BYTEARRAY(objPtr) \
+		((ByteArray *) (objPtr)->internalRep.twoPtrValue.ptr1)
+#define SET_BYTEARRAY(objPtr, baPtr) \
+		(objPtr)->internalRep.twoPtrValue.ptr1 = (void *) (baPtr)
+
+int
+TclIsPureByteArray(
+    Tcl_Obj * objPtr)
+{
+    return (objPtr->typePtr == &properByteArrayType);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewByteArrayObj --
+ *
+ *	This procedure is creates a new ByteArray object and initializes it
+ *	from the given array of bytes.
+ *
+ * Results:
+ *	The newly create object is returned. This object will have no initial
+ *	string representation. The returned object has a ref count of 0.
+ *
+ * Side effects:
+ *	Memory allocated for new object and copy of byte array argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_NewByteArrayObj
+
+Tcl_Obj *
+Tcl_NewByteArrayObj(
+    const unsigned char *bytes,	/* The array of bytes used to initialize the
+				 * new object. */
+    int length)			/* Length of the array of bytes, which must be
+				 * >= 0. */
+{
+#ifdef TCL_MEM_DEBUG
+    return Tcl_DbNewByteArrayObj(bytes, length, "unknown", 0);
+#else /* if not TCL_MEM_DEBUG */
+    Tcl_Obj *objPtr;
+
+    TclNewObj(objPtr);
+    Tcl_SetByteArrayObj(objPtr, bytes, length);
+    return objPtr;
+#endif /* TCL_MEM_DEBUG */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewByteArrayObj --
+ *
+ *	This procedure is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It is the same as the Tcl_NewByteArrayObj
+ *	above except that it calls Tcl_DbCkalloc directly with the file name
+ *	and line number from its caller. This simplifies debugging since then
+ *	the [memory active] command will report the correct file name and line
+ *	number when reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this procedure just returns the
+ *	result of calling Tcl_NewByteArrayObj.
+ *
+ * Results:
+ *	The newly create object is returned. This object will have no initial
+ *	string representation. The returned object has a ref count of 0.
+ *
+ * Side effects:
+ *	Memory allocated for new object and copy of byte array argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_DbNewByteArrayObj(
+    const unsigned char *bytes,	/* The array of bytes used to initialize the
+				 * new object. */
+    int length,			/* Length of the array of bytes, which must be
+				 * >= 0. */
+    const char *file,		/* The name of the source file calling this
+				 * procedure; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+#ifdef TCL_MEM_DEBUG
+    Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    Tcl_SetByteArrayObj(objPtr, bytes, length);
+    return objPtr;
+#else /* if not TCL_MEM_DEBUG */
+    return Tcl_NewByteArrayObj(bytes, length);
+#endif /* TCL_MEM_DEBUG */
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_SetByteArrayObj --
+ *
+ *	Modify an object to be a ByteArray object and to have the specified
+ *	array of bytes as its value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep and internal rep is freed. Memory
+ *	allocated for copy of byte array argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetByteArrayObj(
+    Tcl_Obj *objPtr,		/* Object to initialize as a ByteArray. */
+    const unsigned char *bytes,	/* The array of bytes to use as the new
+				   value. May be NULL even if length > 0. */
+    int length)			/* Length of the array of bytes, which must
+				   be >= 0. */
+{
+    ByteArray *byteArrayPtr;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetByteArrayObj");
+    }
+    TclFreeIntRep(objPtr);
+    TclInvalidateStringRep(objPtr);
+
+    if (length < 0) {
+	length = 0;
+    }
+    byteArrayPtr = ckalloc(BYTEARRAY_SIZE(length));
+    byteArrayPtr->used = length;
+    byteArrayPtr->allocated = length;
+
+    if ((bytes != NULL) && (length > 0)) {
+	memcpy(byteArrayPtr->bytes, bytes, (size_t) length);
+    }
+    objPtr->typePtr = &properByteArrayType;
+    SET_BYTEARRAY(objPtr, byteArrayPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetByteArrayFromObj --
+ *
+ *	Attempt to get the array of bytes from the Tcl object. If the object
+ *	is not already a ByteArray object, an attempt will be made to convert
+ *	it to one.
+ *
+ * Results:
+ *	Pointer to array of bytes representing the ByteArray object.
+ *
+ * Side effects:
+ *	Frees old internal rep. Allocates memory for new internal rep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+unsigned char *
+Tcl_GetByteArrayFromObj(
+    Tcl_Obj *objPtr,		/* The ByteArray object. */
+    int *lengthPtr)		/* If non-NULL, filled with length of the
+				 * array of bytes in the ByteArray object. */
+{
+    ByteArray *baPtr;
+
+    if ((objPtr->typePtr != &properByteArrayType)
+	    && (objPtr->typePtr != &tclByteArrayType)) {
+	SetByteArrayFromAny(NULL, objPtr);
+    }
+    baPtr = GET_BYTEARRAY(objPtr);
+
+    if (lengthPtr != NULL) {
+	*lengthPtr = baPtr->used;
+    }
+    return (unsigned char *) baPtr->bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetByteArrayLength --
+ *
+ *	This procedure changes the length of the byte array for this object.
+ *	Once the caller has set the length of the array, it is acceptable to
+ *	directly modify the bytes in the array up until Tcl_GetStringFromObj()
+ *	has been called on this object.
+ *
+ * Results:
+ *	The new byte array of the specified length.
+ *
+ * Side effects:
+ *	Allocates enough memory for an array of bytes of the requested size.
+ *	When growing the array, the old array is copied to the new array; new
+ *	bytes are undefined. When shrinking, the old array is truncated to the
+ *	specified length.
+ *
+ *----------------------------------------------------------------------
+ */
+
+unsigned char *
+Tcl_SetByteArrayLength(
+    Tcl_Obj *objPtr,		/* The ByteArray object. */
+    int length)			/* New length for internal byte array. */
+{
+    ByteArray *byteArrayPtr;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetByteArrayLength");
+    }
+    if ((objPtr->typePtr != &properByteArrayType)
+	    && (objPtr->typePtr != &tclByteArrayType)) {
+	SetByteArrayFromAny(NULL, objPtr);
+    }
+
+    byteArrayPtr = GET_BYTEARRAY(objPtr);
+    if (length > byteArrayPtr->allocated) {
+	byteArrayPtr = ckrealloc(byteArrayPtr, BYTEARRAY_SIZE(length));
+	byteArrayPtr->allocated = length;
+	SET_BYTEARRAY(objPtr, byteArrayPtr);
+    }
+    TclInvalidateStringRep(objPtr);
+    byteArrayPtr->used = length;
+    return byteArrayPtr->bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetByteArrayFromAny --
+ *
+ *	Generate the ByteArray internal rep from the string rep.
+ *
+ * Results:
+ *	The return value is always TCL_OK.
+ *
+ * Side effects:
+ *	A ByteArray object is stored as the internal rep of objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetByteArrayFromAny(
+    Tcl_Interp *interp,		/* Not used. */
+    Tcl_Obj *objPtr)		/* The object to convert to type ByteArray. */
+{
+    size_t length;
+    int improper = 0;
+    const char *src, *srcEnd;
+    unsigned char *dst;
+    ByteArray *byteArrayPtr;
+    Tcl_UniChar ch = 0;
+
+    if (objPtr->typePtr == &properByteArrayType) {
+	return TCL_OK;
+    }
+    if (objPtr->typePtr == &tclByteArrayType) {
+	return TCL_OK;
+    }
+
+    src = TclGetString(objPtr);
+    length = objPtr->length;
+    srcEnd = src + length;
+
+    byteArrayPtr = ckalloc(BYTEARRAY_SIZE(length));
+    for (dst = byteArrayPtr->bytes; src < srcEnd; ) {
+	src += TclUtfToUniChar(src, &ch);
+	improper = improper || (ch > 255);
+	*dst++ = UCHAR(ch);
+    }
+
+    byteArrayPtr->used = dst - byteArrayPtr->bytes;
+    byteArrayPtr->allocated = length;
+
+    TclFreeIntRep(objPtr);
+    objPtr->typePtr = improper ? &tclByteArrayType : &properByteArrayType;
+    SET_BYTEARRAY(objPtr, byteArrayPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeByteArrayInternalRep --
+ *
+ *	Deallocate the storage associated with a ByteArray data object's
+ *	internal representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeByteArrayInternalRep(
+    Tcl_Obj *objPtr)		/* Object with internal rep to free. */
+{
+    ckfree(GET_BYTEARRAY(objPtr));
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupByteArrayInternalRep --
+ *
+ *	Initialize the internal representation of a ByteArray Tcl_Obj to a
+ *	copy of the internal representation of an existing ByteArray object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupByteArrayInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    Tcl_Obj *copyPtr)		/* Object with internal rep to set. */
+{
+    int length;
+    ByteArray *srcArrayPtr, *copyArrayPtr;
+
+    srcArrayPtr = GET_BYTEARRAY(srcPtr);
+    length = srcArrayPtr->used;
+
+    copyArrayPtr = ckalloc(BYTEARRAY_SIZE(length));
+    copyArrayPtr->used = length;
+    copyArrayPtr->allocated = length;
+    memcpy(copyArrayPtr->bytes, srcArrayPtr->bytes, (size_t) length);
+    SET_BYTEARRAY(copyPtr, copyArrayPtr);
+
+    copyPtr->typePtr = srcPtr->typePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfByteArray --
+ *
+ *	Update the string representation for a ByteArray data object. Note:
+ *	This procedure does not invalidate an existing old string rep so
+ *	storage will be lost if this has not already been done.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	ByteArray-to-string conversion.
+ *
+ *	The object becomes a string object -- the internal rep is discarded
+ *	and the typePtr becomes NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfByteArray(
+    Tcl_Obj *objPtr)		/* ByteArray object whose string rep to
+				 * update. */
+{
+    int i, length, size;
+    unsigned char *src;
+    char *dst;
+    ByteArray *byteArrayPtr;
+
+    byteArrayPtr = GET_BYTEARRAY(objPtr);
+    src = byteArrayPtr->bytes;
+    length = byteArrayPtr->used;
+
+    /*
+     * How much space will string rep need?
+     */
+
+    size = length;
+    for (i = 0; i < length && size >= 0; i++) {
+	if ((src[i] == 0) || (src[i] > 127)) {
+	    size++;
+	}
+    }
+    if (size < 0) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+
+    dst = ckalloc(size + 1);
+    objPtr->bytes = dst;
+    objPtr->length = size;
+
+    if (size == length) {
+	memcpy(dst, src, (size_t) size);
+	dst[size] = '\0';
+    } else {
+	for (i = 0; i < length; i++) {
+	    dst += Tcl_UniCharToUtf(src[i], dst);
+	}
+	*dst = '\0';
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclAppendBytesToByteArray --
+ *
+ *	This function appends an array of bytes to a byte array object. Note
+ *	that the object *must* be unshared, and the array of bytes *must not*
+ *	refer to the object being appended to.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates enough memory for an array of bytes of the requested total
+ *	size, or possibly larger. [Bug 2992970]
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclAppendBytesToByteArray(
+    Tcl_Obj *objPtr,
+    const unsigned char *bytes,
+    int len)
+{
+    ByteArray *byteArrayPtr;
+    int needed;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object","TclAppendBytesToByteArray");
+    }
+    if (len < 0) {
+	Tcl_Panic("%s must be called with definite number of bytes to append",
+		"TclAppendBytesToByteArray");
+    }
+    if (len == 0) {
+	/* Append zero bytes is a no-op. */
+	return;
+    }
+    if ((objPtr->typePtr != &properByteArrayType)
+	    && (objPtr->typePtr != &tclByteArrayType)) {
+	SetByteArrayFromAny(NULL, objPtr);
+    }
+    byteArrayPtr = GET_BYTEARRAY(objPtr);
+
+    if (len > INT_MAX - byteArrayPtr->used) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+
+    needed = byteArrayPtr->used + len;
+    /*
+     * If we need to, resize the allocated space in the byte array.
+     */
+
+    if (needed > byteArrayPtr->allocated) {
+	ByteArray *ptr = NULL;
+	int attempt;
+
+	if (needed <= INT_MAX/2) {
+	    /* Try to allocate double the total space that is needed. */
+	    attempt = 2 * needed;
+	    ptr = attemptckrealloc(byteArrayPtr, BYTEARRAY_SIZE(attempt));
+	}
+	if (ptr == NULL) {
+	    /* Try to allocate double the increment that is needed (plus). */
+	    unsigned int limit = INT_MAX - needed;
+	    unsigned int extra = len + TCL_MIN_GROWTH;
+	    int growth = (int) ((extra > limit) ? limit : extra);
+
+	    attempt = needed + growth;
+	    ptr = attemptckrealloc(byteArrayPtr, BYTEARRAY_SIZE(attempt));
+	}
+	if (ptr == NULL) {
+	    /* Last chance: Try to allocate exactly what is needed. */
+	    attempt = needed;
+	    ptr = ckrealloc(byteArrayPtr, BYTEARRAY_SIZE(attempt));
+	}
+	byteArrayPtr = ptr;
+	byteArrayPtr->allocated = attempt;
+	SET_BYTEARRAY(objPtr, byteArrayPtr);
+    }
+
+    if (bytes) {
+	memcpy(byteArrayPtr->bytes + byteArrayPtr->used, bytes, len);
+    }
+    byteArrayPtr->used += len;
+    TclInvalidateStringRep(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitBinaryCmd --
+ *
+ *	This function is called to create the "binary" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A command token for the new command.
+ *
+ * Side effects:
+ *	Creates a new binary command as a mapped ensemble.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitBinaryCmd(
+    Tcl_Interp *interp)
+{
+    Tcl_Command binaryEnsemble;
+
+    binaryEnsemble = TclMakeEnsemble(interp, "binary", binaryMap);
+    TclMakeEnsemble(interp, "binary encode", encodeMap);
+    TclMakeEnsemble(interp, "binary decode", decodeMap);
+    return binaryEnsemble;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryFormatCmd --
+ *
+ *	This procedure implements the "binary format" Tcl command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryFormatCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int arg;			/* Index of next argument to consume. */
+    int value = 0;		/* Current integer value to be packed.
+				 * Initialized to avoid compiler warning. */
+    char cmd;			/* Current format character. */
+    int count;			/* Count associated with current format
+				 * character. */
+    int flags;			/* Format field flags */
+    const char *format;	/* Pointer to current position in format
+				 * string. */
+    Tcl_Obj *resultPtr = NULL;	/* Object holding result buffer. */
+    unsigned char *buffer;	/* Start of result buffer. */
+    unsigned char *cursor;	/* Current position within result buffer. */
+    unsigned char *maxPos;	/* Greatest position within result buffer that
+				 * cursor has visited.*/
+    const char *errorString;
+    const char *errorValue, *str;
+    int offset, size, length;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "formatString ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * To avoid copying the data, we format the string in two passes. The
+     * first pass computes the size of the output buffer. The second pass
+     * places the formatted data into the buffer.
+     */
+
+    format = TclGetString(objv[1]);
+    arg = 2;
+    offset = 0;
+    length = 0;
+    while (*format != '\0') {
+	str = format;
+	flags = 0;
+	if (!GetFormatSpec(&format, &cmd, &count, &flags)) {
+	    break;
+	}
+	switch (cmd) {
+	case 'a':
+	case 'A':
+	case 'b':
+	case 'B':
+	case 'h':
+	case 'H':
+	    /*
+	     * For string-type specifiers, the count corresponds to the number
+	     * of bytes in a single argument.
+	     */
+
+	    if (arg >= objc) {
+		goto badIndex;
+	    }
+	    if (count == BINARY_ALL) {
+		Tcl_GetByteArrayFromObj(objv[arg], &count);
+	    } else if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    arg++;
+	    if (cmd == 'a' || cmd == 'A') {
+		offset += count;
+	    } else if (cmd == 'b' || cmd == 'B') {
+		offset += (count + 7) / 8;
+	    } else {
+		offset += (count + 1) / 2;
+	    }
+	    break;
+	case 'c':
+	    size = 1;
+	    goto doNumbers;
+	case 't':
+	case 's':
+	case 'S':
+	    size = 2;
+	    goto doNumbers;
+	case 'n':
+	case 'i':
+	case 'I':
+	    size = 4;
+	    goto doNumbers;
+	case 'm':
+	case 'w':
+	case 'W':
+	    size = 8;
+	    goto doNumbers;
+	case 'r':
+	case 'R':
+	case 'f':
+	    size = sizeof(float);
+	    goto doNumbers;
+	case 'q':
+	case 'Q':
+	case 'd':
+	    size = sizeof(double);
+
+	doNumbers:
+	    if (arg >= objc) {
+		goto badIndex;
+	    }
+
+	    /*
+	     * For number-type specifiers, the count corresponds to the number
+	     * of elements in the list stored in a single argument. If no
+	     * count is specified, then the argument is taken as a single
+	     * non-list value.
+	     */
+
+	    if (count == BINARY_NOCOUNT) {
+		arg++;
+		count = 1;
+	    } else {
+		int listc;
+		Tcl_Obj **listv;
+
+		/*
+		 * The macro evals its args more than once: avoid arg++
+		 */
+
+		if (TclListObjGetElements(interp, objv[arg], &listc,
+			&listv) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		arg++;
+
+		if (count == BINARY_ALL) {
+		    count = listc;
+		} else if (count > listc) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "number of elements in list does not match count",
+			    -1));
+		    return TCL_ERROR;
+		}
+	    }
+	    offset += count*size;
+	    break;
+
+	case 'x':
+	    if (count == BINARY_ALL) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"cannot use \"*\" in format string with \"x\"", -1));
+		return TCL_ERROR;
+	    } else if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    offset += count;
+	    break;
+	case 'X':
+	    if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    if ((count > offset) || (count == BINARY_ALL)) {
+		count = offset;
+	    }
+	    if (offset > length) {
+		length = offset;
+	    }
+	    offset -= count;
+	    break;
+	case '@':
+	    if (offset > length) {
+		length = offset;
+	    }
+	    if (count == BINARY_ALL) {
+		offset = length;
+	    } else if (count == BINARY_NOCOUNT) {
+		goto badCount;
+	    } else {
+		offset = count;
+	    }
+	    break;
+	default:
+	    errorString = str;
+	    goto badField;
+	}
+    }
+    if (offset > length) {
+	length = offset;
+    }
+    if (length == 0) {
+	return TCL_OK;
+    }
+
+    /*
+     * Prepare the result object by preallocating the caclulated number of
+     * bytes and filling with nulls.
+     */
+
+    resultPtr = Tcl_NewObj();
+    buffer = Tcl_SetByteArrayLength(resultPtr, length);
+    memset(buffer, 0, (size_t) length);
+
+    /*
+     * Pack the data into the result object. Note that we can skip the
+     * error checking during this pass, since we have already parsed the
+     * string once.
+     */
+
+    arg = 2;
+    format = TclGetString(objv[1]);
+    cursor = buffer;
+    maxPos = cursor;
+    while (*format != 0) {
+	flags = 0;
+	if (!GetFormatSpec(&format, &cmd, &count, &flags)) {
+	    break;
+	}
+	if ((count == 0) && (cmd != '@')) {
+	    if (cmd != 'x') {
+		arg++;
+	    }
+	    continue;
+	}
+	switch (cmd) {
+	case 'a':
+	case 'A': {
+	    char pad = (char) (cmd == 'a' ? '\0' : ' ');
+	    unsigned char *bytes;
+
+	    bytes = Tcl_GetByteArrayFromObj(objv[arg++], &length);
+
+	    if (count == BINARY_ALL) {
+		count = length;
+	    } else if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    if (length >= count) {
+		memcpy(cursor, bytes, (size_t) count);
+	    } else {
+		memcpy(cursor, bytes, (size_t) length);
+		memset(cursor + length, pad, (size_t) (count - length));
+	    }
+	    cursor += count;
+	    break;
+	}
+	case 'b':
+	case 'B': {
+	    unsigned char *last;
+
+	    str = TclGetStringFromObj(objv[arg], &length);
+	    arg++;
+	    if (count == BINARY_ALL) {
+		count = length;
+	    } else if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    last = cursor + ((count + 7) / 8);
+	    if (count > length) {
+		count = length;
+	    }
+	    value = 0;
+	    errorString = "binary";
+	    if (cmd == 'B') {
+		for (offset = 0; offset < count; offset++) {
+		    value <<= 1;
+		    if (str[offset] == '1') {
+			value |= 1;
+		    } else if (str[offset] != '0') {
+			errorValue = str;
+			Tcl_DecrRefCount(resultPtr);
+			goto badValue;
+		    }
+		    if (((offset + 1) % 8) == 0) {
+			*cursor++ = UCHAR(value);
+			value = 0;
+		    }
+		}
+	    } else {
+		for (offset = 0; offset < count; offset++) {
+		    value >>= 1;
+		    if (str[offset] == '1') {
+			value |= 128;
+		    } else if (str[offset] != '0') {
+			errorValue = str;
+			Tcl_DecrRefCount(resultPtr);
+			goto badValue;
+		    }
+		    if (!((offset + 1) % 8)) {
+			*cursor++ = UCHAR(value);
+			value = 0;
+		    }
+		}
+	    }
+	    if ((offset % 8) != 0) {
+		if (cmd == 'B') {
+		    value <<= 8 - (offset % 8);
+		} else {
+		    value >>= 8 - (offset % 8);
+		}
+		*cursor++ = UCHAR(value);
+	    }
+	    while (cursor < last) {
+		*cursor++ = '\0';
+	    }
+	    break;
+	}
+	case 'h':
+	case 'H': {
+	    unsigned char *last;
+	    int c;
+
+	    str = TclGetStringFromObj(objv[arg], &length);
+	    arg++;
+	    if (count == BINARY_ALL) {
+		count = length;
+	    } else if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    last = cursor + ((count + 1) / 2);
+	    if (count > length) {
+		count = length;
+	    }
+	    value = 0;
+	    errorString = "hexadecimal";
+	    if (cmd == 'H') {
+		for (offset = 0; offset < count; offset++) {
+		    value <<= 4;
+		    if (!isxdigit(UCHAR(str[offset]))) {     /* INTL: digit */
+			errorValue = str;
+			Tcl_DecrRefCount(resultPtr);
+			goto badValue;
+		    }
+		    c = str[offset] - '0';
+		    if (c > 9) {
+			c += ('0' - 'A') + 10;
+		    }
+		    if (c > 16) {
+			c += ('A' - 'a');
+		    }
+		    value |= (c & 0xf);
+		    if (offset % 2) {
+			*cursor++ = (char) value;
+			value = 0;
+		    }
+		}
+	    } else {
+		for (offset = 0; offset < count; offset++) {
+		    value >>= 4;
+
+		    if (!isxdigit(UCHAR(str[offset]))) {     /* INTL: digit */
+			errorValue = str;
+			Tcl_DecrRefCount(resultPtr);
+			goto badValue;
+		    }
+		    c = str[offset] - '0';
+		    if (c > 9) {
+			c += ('0' - 'A') + 10;
+		    }
+		    if (c > 16) {
+			c += ('A' - 'a');
+		    }
+		    value |= ((c << 4) & 0xf0);
+		    if (offset % 2) {
+			*cursor++ = UCHAR(value & 0xff);
+			value = 0;
+		    }
+		}
+	    }
+	    if (offset % 2) {
+		if (cmd == 'H') {
+		    value <<= 4;
+		} else {
+		    value >>= 4;
+		}
+		*cursor++ = UCHAR(value);
+	    }
+
+	    while (cursor < last) {
+		*cursor++ = '\0';
+	    }
+	    break;
+	}
+	case 'c':
+	case 't':
+	case 's':
+	case 'S':
+	case 'n':
+	case 'i':
+	case 'I':
+	case 'm':
+	case 'w':
+	case 'W':
+	case 'r':
+	case 'R':
+	case 'd':
+	case 'q':
+	case 'Q':
+	case 'f': {
+	    int listc, i;
+	    Tcl_Obj **listv;
+
+	    if (count == BINARY_NOCOUNT) {
+		/*
+		 * Note that we are casting away the const-ness of objv, but
+		 * this is safe since we aren't going to modify the array.
+		 */
+
+		listv = (Tcl_Obj **) (objv + arg);
+		listc = 1;
+		count = 1;
+	    } else {
+		TclListObjGetElements(interp, objv[arg], &listc, &listv);
+		if (count == BINARY_ALL) {
+		    count = listc;
+		}
+	    }
+	    arg++;
+	    for (i = 0; i < count; i++) {
+		if (FormatNumber(interp, cmd, listv[i], &cursor)!=TCL_OK) {
+		    Tcl_DecrRefCount(resultPtr);
+		    return TCL_ERROR;
+		}
+	    }
+	    break;
+	}
+	case 'x':
+	    if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    memset(cursor, 0, (size_t) count);
+	    cursor += count;
+	    break;
+	case 'X':
+	    if (cursor > maxPos) {
+		maxPos = cursor;
+	    }
+	    if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    if ((count == BINARY_ALL) || (count > (cursor - buffer))) {
+		cursor = buffer;
+	    } else {
+		cursor -= count;
+	    }
+	    break;
+	case '@':
+	    if (cursor > maxPos) {
+		maxPos = cursor;
+	    }
+	    if (count == BINARY_ALL) {
+		cursor = maxPos;
+	    } else {
+		cursor = buffer + count;
+	    }
+	    break;
+	}
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+
+ badValue:
+    Tcl_ResetResult(interp);
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "expected %s string but got \"%s\" instead",
+	    errorString, errorValue));
+    return TCL_ERROR;
+
+ badCount:
+    errorString = "missing count for \"@\" field specifier";
+    goto error;
+
+ badIndex:
+    errorString = "not enough arguments for all format specifiers";
+    goto error;
+
+ badField:
+    {
+	Tcl_UniChar ch = 0;
+	char buf[TCL_UTF_MAX + 1];
+
+	TclUtfToUniChar(errorString, &ch);
+	buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad field specifier \"%s\"", buf));
+	return TCL_ERROR;
+    }
+
+ error:
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(errorString, -1));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryScanCmd --
+ *
+ *	This procedure implements the "binary scan" Tcl command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+BinaryScanCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int arg;			/* Index of next argument to consume. */
+    int value = 0;		/* Current integer value to be packed.
+				 * Initialized to avoid compiler warning. */
+    char cmd;			/* Current format character. */
+    int count;			/* Count associated with current format
+				 * character. */
+    int flags;			/* Format field flags */
+    const char *format;	/* Pointer to current position in format
+				 * string. */
+    Tcl_Obj *resultPtr = NULL;	/* Object holding result buffer. */
+    unsigned char *buffer;	/* Start of result buffer. */
+    const char *errorString;
+    const char *str;
+    int offset, size, length;
+
+    int i;
+    Tcl_Obj *valuePtr, *elementPtr;
+    Tcl_HashTable numberCacheHash;
+    Tcl_HashTable *numberCachePtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"value formatString ?varName ...?");
+	return TCL_ERROR;
+    }
+    numberCachePtr = &numberCacheHash;
+    Tcl_InitHashTable(numberCachePtr, TCL_ONE_WORD_KEYS);
+    buffer = Tcl_GetByteArrayFromObj(objv[1], &length);
+    format = TclGetString(objv[2]);
+    arg = 3;
+    offset = 0;
+    while (*format != '\0') {
+	str = format;
+	flags = 0;
+	if (!GetFormatSpec(&format, &cmd, &count, &flags)) {
+	    goto done;
+	}
+	switch (cmd) {
+	case 'a':
+	case 'A': {
+	    unsigned char *src;
+
+	    if (arg >= objc) {
+		DeleteScanNumberCache(numberCachePtr);
+		goto badIndex;
+	    }
+	    if (count == BINARY_ALL) {
+		count = length - offset;
+	    } else {
+		if (count == BINARY_NOCOUNT) {
+		    count = 1;
+		}
+		if (count > (length - offset)) {
+		    goto done;
+		}
+	    }
+
+	    src = buffer + offset;
+	    size = count;
+
+	    /*
+	     * Trim trailing nulls and spaces, if necessary.
+	     */
+
+	    if (cmd == 'A') {
+		while (size > 0) {
+		    if (src[size-1] != '\0' && src[size-1] != ' ') {
+			break;
+		    }
+		    size--;
+		}
+	    }
+
+	    /*
+	     * Have to do this #ifdef-fery because (as part of defining
+	     * Tcl_NewByteArrayObj) we removed the #def that hides this stuff
+	     * normally. If this code ever gets copied to another file, it
+	     * should be changed back to the simpler version.
+	     */
+
+#ifdef TCL_MEM_DEBUG
+	    valuePtr = Tcl_DbNewByteArrayObj(src, size, __FILE__, __LINE__);
+#else
+	    valuePtr = Tcl_NewByteArrayObj(src, size);
+#endif /* TCL_MEM_DEBUG */
+
+	    resultPtr = Tcl_ObjSetVar2(interp, objv[arg], NULL, valuePtr,
+		    TCL_LEAVE_ERR_MSG);
+	    arg++;
+	    if (resultPtr == NULL) {
+		DeleteScanNumberCache(numberCachePtr);
+		return TCL_ERROR;
+	    }
+	    offset += count;
+	    break;
+	}
+	case 'b':
+	case 'B': {
+	    unsigned char *src;
+	    char *dest;
+
+	    if (arg >= objc) {
+		DeleteScanNumberCache(numberCachePtr);
+		goto badIndex;
+	    }
+	    if (count == BINARY_ALL) {
+		count = (length - offset) * 8;
+	    } else {
+		if (count == BINARY_NOCOUNT) {
+		    count = 1;
+		}
+		if (count > (length - offset) * 8) {
+		    goto done;
+		}
+	    }
+	    src = buffer + offset;
+	    valuePtr = Tcl_NewObj();
+	    Tcl_SetObjLength(valuePtr, count);
+	    dest = TclGetString(valuePtr);
+
+	    if (cmd == 'b') {
+		for (i = 0; i < count; i++) {
+		    if (i % 8) {
+			value >>= 1;
+		    } else {
+			value = *src++;
+		    }
+		    *dest++ = (char) ((value & 1) ? '1' : '0');
+		}
+	    } else {
+		for (i = 0; i < count; i++) {
+		    if (i % 8) {
+			value <<= 1;
+		    } else {
+			value = *src++;
+		    }
+		    *dest++ = (char) ((value & 0x80) ? '1' : '0');
+		}
+	    }
+
+	    resultPtr = Tcl_ObjSetVar2(interp, objv[arg], NULL, valuePtr,
+		    TCL_LEAVE_ERR_MSG);
+	    arg++;
+	    if (resultPtr == NULL) {
+		DeleteScanNumberCache(numberCachePtr);
+		return TCL_ERROR;
+	    }
+	    offset += (count + 7) / 8;
+	    break;
+	}
+	case 'h':
+	case 'H': {
+	    char *dest;
+	    unsigned char *src;
+	    static const char hexdigit[] = "0123456789abcdef";
+
+	    if (arg >= objc) {
+		DeleteScanNumberCache(numberCachePtr);
+		goto badIndex;
+	    }
+	    if (count == BINARY_ALL) {
+		count = (length - offset)*2;
+	    } else {
+		if (count == BINARY_NOCOUNT) {
+		    count = 1;
+		}
+		if (count > (length - offset)*2) {
+		    goto done;
+		}
+	    }
+	    src = buffer + offset;
+	    valuePtr = Tcl_NewObj();
+	    Tcl_SetObjLength(valuePtr, count);
+	    dest = TclGetString(valuePtr);
+
+	    if (cmd == 'h') {
+		for (i = 0; i < count; i++) {
+		    if (i % 2) {
+			value >>= 4;
+		    } else {
+			value = *src++;
+		    }
+		    *dest++ = hexdigit[value & 0xf];
+		}
+	    } else {
+		for (i = 0; i < count; i++) {
+		    if (i % 2) {
+			value <<= 4;
+		    } else {
+			value = *src++;
+		    }
+		    *dest++ = hexdigit[(value >> 4) & 0xf];
+		}
+	    }
+
+	    resultPtr = Tcl_ObjSetVar2(interp, objv[arg], NULL, valuePtr,
+		    TCL_LEAVE_ERR_MSG);
+	    arg++;
+	    if (resultPtr == NULL) {
+		DeleteScanNumberCache(numberCachePtr);
+		return TCL_ERROR;
+	    }
+	    offset += (count + 1) / 2;
+	    break;
+	}
+	case 'c':
+	    size = 1;
+	    goto scanNumber;
+	case 't':
+	case 's':
+	case 'S':
+	    size = 2;
+	    goto scanNumber;
+	case 'n':
+	case 'i':
+	case 'I':
+	    size = 4;
+	    goto scanNumber;
+	case 'm':
+	case 'w':
+	case 'W':
+	    size = 8;
+	    goto scanNumber;
+	case 'r':
+	case 'R':
+	case 'f':
+	    size = sizeof(float);
+	    goto scanNumber;
+	case 'q':
+	case 'Q':
+	case 'd': {
+	    unsigned char *src;
+
+	    size = sizeof(double);
+	    /* fall through */
+
+	scanNumber:
+	    if (arg >= objc) {
+		DeleteScanNumberCache(numberCachePtr);
+		goto badIndex;
+	    }
+	    if (count == BINARY_NOCOUNT) {
+		if ((length - offset) < size) {
+		    goto done;
+		}
+		valuePtr = ScanNumber(buffer+offset, cmd, flags,
+			&numberCachePtr);
+		offset += size;
+	    } else {
+		if (count == BINARY_ALL) {
+		    count = (length - offset) / size;
+		}
+		if ((length - offset) < (count * size)) {
+		    goto done;
+		}
+		valuePtr = Tcl_NewObj();
+		src = buffer + offset;
+		for (i = 0; i < count; i++) {
+		    elementPtr = ScanNumber(src, cmd, flags, &numberCachePtr);
+		    src += size;
+		    Tcl_ListObjAppendElement(NULL, valuePtr, elementPtr);
+		}
+		offset += count * size;
+	    }
+
+	    resultPtr = Tcl_ObjSetVar2(interp, objv[arg], NULL, valuePtr,
+		    TCL_LEAVE_ERR_MSG);
+	    arg++;
+	    if (resultPtr == NULL) {
+		DeleteScanNumberCache(numberCachePtr);
+		return TCL_ERROR;
+	    }
+	    break;
+	}
+	case 'x':
+	    if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    if ((count == BINARY_ALL) || (count > (length - offset))) {
+		offset = length;
+	    } else {
+		offset += count;
+	    }
+	    break;
+	case 'X':
+	    if (count == BINARY_NOCOUNT) {
+		count = 1;
+	    }
+	    if ((count == BINARY_ALL) || (count > offset)) {
+		offset = 0;
+	    } else {
+		offset -= count;
+	    }
+	    break;
+	case '@':
+	    if (count == BINARY_NOCOUNT) {
+		DeleteScanNumberCache(numberCachePtr);
+		goto badCount;
+	    }
+	    if ((count == BINARY_ALL) || (count > length)) {
+		offset = length;
+	    } else {
+		offset = count;
+	    }
+	    break;
+	default:
+	    DeleteScanNumberCache(numberCachePtr);
+	    errorString = str;
+	    goto badField;
+	}
+    }
+
+    /*
+     * Set the result to the last position of the cursor.
+     */
+
+ done:
+    Tcl_SetObjResult(interp, Tcl_NewLongObj(arg - 3));
+    DeleteScanNumberCache(numberCachePtr);
+
+    return TCL_OK;
+
+ badCount:
+    errorString = "missing count for \"@\" field specifier";
+    goto error;
+
+ badIndex:
+    errorString = "not enough arguments for all format specifiers";
+    goto error;
+
+ badField:
+    {
+	Tcl_UniChar ch = 0;
+	char buf[TCL_UTF_MAX + 1];
+
+	TclUtfToUniChar(errorString, &ch);
+	buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad field specifier \"%s\"", buf));
+	return TCL_ERROR;
+    }
+
+ error:
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(errorString, -1));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetFormatSpec --
+ *
+ *	This function parses the format strings used in the binary format and
+ *	scan commands.
+ *
+ * Results:
+ *	Moves the formatPtr to the start of the next command. Returns the
+ *	current command character and count in cmdPtr and countPtr. The count
+ *	is set to BINARY_ALL if the count character was '*' or BINARY_NOCOUNT
+ *	if no count was specified. Returns 1 on success, or 0 if the string
+ *	did not have a format specifier.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetFormatSpec(
+    const char **formatPtr,	/* Pointer to format string. */
+    char *cmdPtr,		/* Pointer to location of command char. */
+    int *countPtr,		/* Pointer to repeat count value. */
+    int *flagsPtr)		/* Pointer to field flags */
+{
+    /*
+     * Skip any leading blanks.
+     */
+
+    while (**formatPtr == ' ') {
+	(*formatPtr)++;
+    }
+
+    /*
+     * The string was empty, except for whitespace, so fail.
+     */
+
+    if (!(**formatPtr)) {
+	return 0;
+    }
+
+    /*
+     * Extract the command character and any trailing digits or '*'.
+     */
+
+    *cmdPtr = **formatPtr;
+    (*formatPtr)++;
+    if (**formatPtr == 'u') {
+	(*formatPtr)++;
+	*flagsPtr |= BINARY_UNSIGNED;
+    }
+    if (**formatPtr == '*') {
+	(*formatPtr)++;
+	*countPtr = BINARY_ALL;
+    } else if (isdigit(UCHAR(**formatPtr))) { /* INTL: digit */
+	unsigned long int count;
+
+	errno = 0;
+	count = strtoul(*formatPtr, (char **) formatPtr, 10);
+	if (errno || (count > (unsigned long) INT_MAX)) {
+	    *countPtr = INT_MAX;
+	} else {
+	    *countPtr = (int) count;
+	}
+    } else {
+	*countPtr = BINARY_NOCOUNT;
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NeedReversing --
+ *
+ *	This routine determines, if bytes of a number need to be re-ordered,
+ *	and returns a numeric code indicating the re-ordering to be done.
+ *	This depends on the endiannes of the machine and the desired format.
+ *	It is in effect a table (whose contents depend on the endianness of
+ *	the system) describing whether a value needs reversing or not. Anyone
+ *	porting the code to a big-endian platform should take care to make
+ *	sure that they define WORDS_BIGENDIAN though this is already done by
+ *	configure for the Unix build; little-endian platforms (including
+ *	Windows) don't need to do anything.
+ *
+ * Results:
+ *	0	No re-ordering needed.
+ *	1	Reverse the bytes:	01234567 <-> 76543210 (little to big)
+ *	2	Apply this re-ordering: 01234567 <-> 45670123 (Nokia to little)
+ *	3	Apply this re-ordering: 01234567 <-> 32107654 (Nokia to big)
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NeedReversing(
+    int format)
+{
+    switch (format) {
+	/* native floats and doubles: never reverse */
+    case 'd':
+    case 'f':
+	/* big endian ints: never reverse */
+    case 'I':
+    case 'S':
+    case 'W':
+#ifdef WORDS_BIGENDIAN
+	/* native ints: reverse if we're little-endian */
+    case 'n':
+    case 't':
+    case 'm':
+	/* f: reverse if we're little-endian */
+    case 'Q':
+    case 'R':
+#else /* !WORDS_BIGENDIAN */
+	/* small endian floats: reverse if we're big-endian */
+    case 'r':
+#endif /* WORDS_BIGENDIAN */
+	return 0;
+
+#ifdef WORDS_BIGENDIAN
+	/* small endian floats: reverse if we're big-endian */
+    case 'q':
+    case 'r':
+#else /* !WORDS_BIGENDIAN */
+	/* native ints: reverse if we're little-endian */
+    case 'n':
+    case 't':
+    case 'm':
+	/* f: reverse if we're little-endian */
+    case 'R':
+#endif /* WORDS_BIGENDIAN */
+	/* small endian ints: always reverse */
+    case 'i':
+    case 's':
+    case 'w':
+	return 1;
+
+#ifndef WORDS_BIGENDIAN
+    /*
+     * The Q and q formats need special handling to account for the unusual
+     * byte ordering of 8-byte floats on Nokia 770 systems, which claim to be
+     * little-endian, but also reverse word order.
+     */
+
+    case 'Q':
+	if (TclNokia770Doubles()) {
+	    return 3;
+	}
+	return 1;
+    case 'q':
+	if (TclNokia770Doubles()) {
+	    return 2;
+	}
+	return 0;
+#endif
+    }
+
+    Tcl_Panic("unexpected fallthrough");
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CopyNumber --
+ *
+ *	This routine is called by FormatNumber and ScanNumber to copy a
+ *	floating-point number. If required, bytes are reversed while copying.
+ *	The behaviour is only fully defined when used with IEEE float and
+ *	double values (guaranteed to be 4 and 8 bytes long, respectively.)
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Copies length bytes
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CopyNumber(
+    const void *from,		/* source */
+    void *to,			/* destination */
+    unsigned length,		/* Number of bytes to copy */
+    int type)			/* What type of thing are we copying? */
+{
+    switch (NeedReversing(type)) {
+    case 0:
+	memcpy(to, from, length);
+	break;
+    case 1: {
+	const unsigned char *fromPtr = from;
+	unsigned char *toPtr = to;
+
+	switch (length) {
+	case 4:
+	    toPtr[0] = fromPtr[3];
+	    toPtr[1] = fromPtr[2];
+	    toPtr[2] = fromPtr[1];
+	    toPtr[3] = fromPtr[0];
+	    break;
+	case 8:
+	    toPtr[0] = fromPtr[7];
+	    toPtr[1] = fromPtr[6];
+	    toPtr[2] = fromPtr[5];
+	    toPtr[3] = fromPtr[4];
+	    toPtr[4] = fromPtr[3];
+	    toPtr[5] = fromPtr[2];
+	    toPtr[6] = fromPtr[1];
+	    toPtr[7] = fromPtr[0];
+	    break;
+	}
+	break;
+    }
+    case 2: {
+	const unsigned char *fromPtr = from;
+	unsigned char *toPtr = to;
+
+	toPtr[0] = fromPtr[4];
+	toPtr[1] = fromPtr[5];
+	toPtr[2] = fromPtr[6];
+	toPtr[3] = fromPtr[7];
+	toPtr[4] = fromPtr[0];
+	toPtr[5] = fromPtr[1];
+	toPtr[6] = fromPtr[2];
+	toPtr[7] = fromPtr[3];
+	break;
+    }
+    case 3: {
+	const unsigned char *fromPtr = from;
+	unsigned char *toPtr = to;
+
+	toPtr[0] = fromPtr[3];
+	toPtr[1] = fromPtr[2];
+	toPtr[2] = fromPtr[1];
+	toPtr[3] = fromPtr[0];
+	toPtr[4] = fromPtr[7];
+	toPtr[5] = fromPtr[6];
+	toPtr[6] = fromPtr[5];
+	toPtr[7] = fromPtr[4];
+	break;
+    }
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FormatNumber --
+ *
+ *	This routine is called by Tcl_BinaryObjCmd to format a number into a
+ *	location pointed at by cursor.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Moves the cursor to the next location to be written into.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FormatNumber(
+    Tcl_Interp *interp,		/* Current interpreter, used to report
+				 * errors. */
+    int type,			/* Type of number to format. */
+    Tcl_Obj *src,		/* Number to format. */
+    unsigned char **cursorPtr)	/* Pointer to index into destination buffer. */
+{
+    long value;
+    double dvalue;
+    Tcl_WideInt wvalue;
+    float fvalue;
+
+    switch (type) {
+    case 'd':
+    case 'q':
+    case 'Q':
+	/*
+	 * Double-precision floating point values. Tcl_GetDoubleFromObj
+	 * returns TCL_ERROR for NaN, but we can check by comparing the
+	 * object's type pointer.
+	 */
+
+	if (Tcl_GetDoubleFromObj(interp, src, &dvalue) != TCL_OK) {
+	    if (src->typePtr != &tclDoubleType) {
+		return TCL_ERROR;
+	    }
+	    dvalue = src->internalRep.doubleValue;
+	}
+	CopyNumber(&dvalue, *cursorPtr, sizeof(double), type);
+	*cursorPtr += sizeof(double);
+	return TCL_OK;
+
+    case 'f':
+    case 'r':
+    case 'R':
+	/*
+	 * Single-precision floating point values. Tcl_GetDoubleFromObj
+	 * returns TCL_ERROR for NaN, but we can check by comparing the
+	 * object's type pointer.
+	 */
+
+	if (Tcl_GetDoubleFromObj(interp, src, &dvalue) != TCL_OK) {
+	    if (src->typePtr != &tclDoubleType) {
+		return TCL_ERROR;
+	    }
+	    dvalue = src->internalRep.doubleValue;
+	}
+
+	/*
+	 * Because some compilers will generate floating point exceptions on
+	 * an overflow cast (e.g. Borland), we restrict the values to the
+	 * valid range for float.
+	 */
+
+	if (fabs(dvalue) > (double)FLT_MAX) {
+	    fvalue = (dvalue >= 0.0) ? FLT_MAX : -FLT_MAX;
+	} else {
+	    fvalue = (float) dvalue;
+	}
+	CopyNumber(&fvalue, *cursorPtr, sizeof(float), type);
+	*cursorPtr += sizeof(float);
+	return TCL_OK;
+
+	/*
+	 * 64-bit integer values.
+	 */
+    case 'w':
+    case 'W':
+    case 'm':
+	if (Tcl_GetWideIntFromObj(interp, src, &wvalue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (NeedReversing(type)) {
+	    *(*cursorPtr)++ = UCHAR(wvalue);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 8);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 16);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 24);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 32);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 40);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 48);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 56);
+	} else {
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 56);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 48);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 40);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 32);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 24);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 16);
+	    *(*cursorPtr)++ = UCHAR(wvalue >> 8);
+	    *(*cursorPtr)++ = UCHAR(wvalue);
+	}
+	return TCL_OK;
+
+	/*
+	 * 32-bit integer values.
+	 */
+    case 'i':
+    case 'I':
+    case 'n':
+	if (TclGetLongFromObj(interp, src, &value) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (NeedReversing(type)) {
+	    *(*cursorPtr)++ = UCHAR(value);
+	    *(*cursorPtr)++ = UCHAR(value >> 8);
+	    *(*cursorPtr)++ = UCHAR(value >> 16);
+	    *(*cursorPtr)++ = UCHAR(value >> 24);
+	} else {
+	    *(*cursorPtr)++ = UCHAR(value >> 24);
+	    *(*cursorPtr)++ = UCHAR(value >> 16);
+	    *(*cursorPtr)++ = UCHAR(value >> 8);
+	    *(*cursorPtr)++ = UCHAR(value);
+	}
+	return TCL_OK;
+
+	/*
+	 * 16-bit integer values.
+	 */
+    case 's':
+    case 'S':
+    case 't':
+	if (TclGetLongFromObj(interp, src, &value) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (NeedReversing(type)) {
+	    *(*cursorPtr)++ = UCHAR(value);
+	    *(*cursorPtr)++ = UCHAR(value >> 8);
+	} else {
+	    *(*cursorPtr)++ = UCHAR(value >> 8);
+	    *(*cursorPtr)++ = UCHAR(value);
+	}
+	return TCL_OK;
+
+	/*
+	 * 8-bit integer values.
+	 */
+    case 'c':
+	if (TclGetLongFromObj(interp, src, &value) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	*(*cursorPtr)++ = UCHAR(value);
+	return TCL_OK;
+
+    default:
+	Tcl_Panic("unexpected fallthrough");
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ScanNumber --
+ *
+ *	This routine is called by Tcl_BinaryObjCmd to scan a number out of a
+ *	buffer.
+ *
+ * Results:
+ *	Returns a newly created object containing the scanned number. This
+ *	object has a ref count of zero.
+ *
+ * Side effects:
+ *	Might reuse an object in the number cache, place a new object in the
+ *	cache, or delete the cache and set the reference to it (itself passed
+ *	in by reference) to NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+ScanNumber(
+    unsigned char *buffer,	/* Buffer to scan number from. */
+    int type,			/* Format character from "binary scan" */
+    int flags,			/* Format field flags */
+    Tcl_HashTable **numberCachePtrPtr)
+				/* Place to look for cache of scanned
+				 * value objects, or NULL if too many
+				 * different numbers have been scanned. */
+{
+    long value;
+    float fvalue;
+    double dvalue;
+    Tcl_WideUInt uwvalue;
+
+    /*
+     * We cannot rely on the compiler to properly sign extend integer values
+     * when we cast from smaller values to larger values because we don't know
+     * the exact size of the integer types. So, we have to handle sign
+     * extension explicitly by checking the high bit and padding with 1's as
+     * needed. This practice is disabled if the BINARY_UNSIGNED flag is set.
+     */
+
+    switch (type) {
+    case 'c':
+	/*
+	 * Characters need special handling. We want to produce a signed
+	 * result, but on some platforms (such as AIX) chars are unsigned. To
+	 * deal with this, check for a value that should be negative but
+	 * isn't.
+	 */
+
+	value = buffer[0];
+	if (!(flags & BINARY_UNSIGNED)) {
+	    if (value & 0x80) {
+		value |= -0x100;
+	    }
+	}
+	goto returnNumericObject;
+
+	/*
+	 * 16-bit numeric values. We need the sign extension trick (see above)
+	 * here as well.
+	 */
+
+    case 's':
+    case 'S':
+    case 't':
+	if (NeedReversing(type)) {
+	    value = (long) (buffer[0] + (buffer[1] << 8));
+	} else {
+	    value = (long) (buffer[1] + (buffer[0] << 8));
+	}
+	if (!(flags & BINARY_UNSIGNED)) {
+	    if (value & 0x8000) {
+		value |= -0x10000;
+	    }
+	}
+	goto returnNumericObject;
+
+	/*
+	 * 32-bit numeric values.
+	 */
+
+    case 'i':
+    case 'I':
+    case 'n':
+	if (NeedReversing(type)) {
+	    value = (long) (buffer[0]
+		    + (buffer[1] << 8)
+		    + (buffer[2] << 16)
+		    + (((long)buffer[3]) << 24));
+	} else {
+	    value = (long) (buffer[3]
+		    + (buffer[2] << 8)
+		    + (buffer[1] << 16)
+		    + (((long) buffer[0]) << 24));
+	}
+
+	/*
+	 * Check to see if the value was sign extended properly on systems
+	 * where an int is more than 32-bits.
+	 * We avoid caching unsigned integers as we cannot distinguish between
+	 * 32bit signed and unsigned in the hash (short and char are ok).
+	 */
+
+	if (flags & BINARY_UNSIGNED) {
+	    return Tcl_NewWideIntObj((Tcl_WideInt)(unsigned long)value);
+	}
+	if ((value & (((unsigned) 1)<<31)) && (value > 0)) {
+	    value -= (((unsigned) 1)<<31);
+	    value -= (((unsigned) 1)<<31);
+	}
+
+    returnNumericObject:
+	if (*numberCachePtrPtr == NULL) {
+	    return Tcl_NewLongObj(value);
+	} else {
+	    register Tcl_HashTable *tablePtr = *numberCachePtrPtr;
+	    register Tcl_HashEntry *hPtr;
+	    int isNew;
+
+	    hPtr = Tcl_CreateHashEntry(tablePtr, INT2PTR(value), &isNew);
+	    if (!isNew) {
+		return Tcl_GetHashValue(hPtr);
+	    }
+	    if (tablePtr->numEntries <= BINARY_SCAN_MAX_CACHE) {
+		register Tcl_Obj *objPtr = Tcl_NewLongObj(value);
+
+		Tcl_IncrRefCount(objPtr);
+		Tcl_SetHashValue(hPtr, objPtr);
+		return objPtr;
+	    }
+
+	    /*
+	     * We've overflowed the cache! Someone's parsing a LOT of varied
+	     * binary data in a single call! Bail out by switching back to the
+	     * old behaviour for the rest of the scan.
+	     *
+	     * Note that anyone just using the 'c' conversion (for bytes)
+	     * cannot trigger this.
+	     */
+
+	    DeleteScanNumberCache(tablePtr);
+	    *numberCachePtrPtr = NULL;
+	    return Tcl_NewLongObj(value);
+	}
+
+	/*
+	 * Do not cache wide (64-bit) values; they are already too large to
+	 * use as keys.
+	 */
+
+    case 'w':
+    case 'W':
+    case 'm':
+	if (NeedReversing(type)) {
+	    uwvalue = ((Tcl_WideUInt) buffer[0])
+		    | (((Tcl_WideUInt) buffer[1]) << 8)
+		    | (((Tcl_WideUInt) buffer[2]) << 16)
+		    | (((Tcl_WideUInt) buffer[3]) << 24)
+		    | (((Tcl_WideUInt) buffer[4]) << 32)
+		    | (((Tcl_WideUInt) buffer[5]) << 40)
+		    | (((Tcl_WideUInt) buffer[6]) << 48)
+		    | (((Tcl_WideUInt) buffer[7]) << 56);
+	} else {
+	    uwvalue = ((Tcl_WideUInt) buffer[7])
+		    | (((Tcl_WideUInt) buffer[6]) << 8)
+		    | (((Tcl_WideUInt) buffer[5]) << 16)
+		    | (((Tcl_WideUInt) buffer[4]) << 24)
+		    | (((Tcl_WideUInt) buffer[3]) << 32)
+		    | (((Tcl_WideUInt) buffer[2]) << 40)
+		    | (((Tcl_WideUInt) buffer[1]) << 48)
+		    | (((Tcl_WideUInt) buffer[0]) << 56);
+	}
+	if (flags & BINARY_UNSIGNED) {
+	    Tcl_Obj *bigObj = NULL;
+	    mp_int big;
+
+	    TclBNInitBignumFromWideUInt(&big, uwvalue);
+	    bigObj = Tcl_NewBignumObj(&big);
+	    return bigObj;
+	}
+	return Tcl_NewWideIntObj((Tcl_WideInt) uwvalue);
+
+	/*
+	 * Do not cache double values; they are already too large to use as
+	 * keys and the values stored are utterly incompatible with the
+	 * integer part of the cache.
+	 */
+
+	/*
+	 * 32-bit IEEE single-precision floating point.
+	 */
+
+    case 'f':
+    case 'R':
+    case 'r':
+	CopyNumber(buffer, &fvalue, sizeof(float), type);
+	return Tcl_NewDoubleObj(fvalue);
+
+	/*
+	 * 64-bit IEEE double-precision floating point.
+	 */
+
+    case 'd':
+    case 'Q':
+    case 'q':
+	CopyNumber(buffer, &dvalue, sizeof(double), type);
+	return Tcl_NewDoubleObj(dvalue);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteScanNumberCache --
+ *
+ *	Deletes the hash table acting as a scan number cache.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Decrements the reference counts of the objects in the cache.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteScanNumberCache(
+    Tcl_HashTable *numberCachePtr)
+				/* Pointer to the hash table, or NULL (when
+				 * the cache has already been deleted due to
+				 * overflow.) */
+{
+    Tcl_HashEntry *hEntry;
+    Tcl_HashSearch search;
+
+    if (numberCachePtr == NULL) {
+	return;
+    }
+
+    hEntry = Tcl_FirstHashEntry(numberCachePtr, &search);
+    while (hEntry != NULL) {
+	register Tcl_Obj *value = Tcl_GetHashValue(hEntry);
+
+	if (value != NULL) {
+	    Tcl_DecrRefCount(value);
+	}
+	hEntry = Tcl_NextHashEntry(&search);
+    }
+    Tcl_DeleteHashTable(numberCachePtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * NOTES --
+ *
+ *	Some measurements show that it is faster to use a table to to perform
+ *	uuencode and base64 value encoding than to calculate the output (at
+ *	least on intel P4 arch).
+ *
+ *	Conversely using a lookup table for the decoding is slower than just
+ *	calculating the values. We therefore use the fastest of each method.
+ *
+ *	Presumably this has to do with the size of the tables. The base64
+ *	decode table is 255 bytes while the encode table is only 65 bytes. The
+ *	choice likely depends on CPU memory cache sizes.
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryEncodeHex --
+ *
+ *	Implement the [binary encode hex] binary encoding. clientData must be
+ *	a table to convert values to hexadecimal digits.
+ *
+ * Results:
+ *	Interp result set to an encoded byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryEncodeHex(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj = NULL;
+    unsigned char *data = NULL;
+    unsigned char *cursor = NULL;
+    int offset = 0, count = 0;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "data");
+	return TCL_ERROR;
+    }
+
+    TclNewObj(resultObj);
+    data = Tcl_GetByteArrayFromObj(objv[1], &count);
+    cursor = Tcl_SetByteArrayLength(resultObj, count * 2);
+    for (offset = 0; offset < count; ++offset) {
+	*cursor++ = HexDigits[((data[offset] >> 4) & 0x0f)];
+	*cursor++ = HexDigits[(data[offset] & 0x0f)];
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryDecodeHex --
+ *
+ *	Implement the [binary decode hex] binary encoding.
+ *
+ * Results:
+ *	Interp result set to an decoded byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryDecodeHex(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj = NULL;
+    unsigned char *data, *datastart, *dataend;
+    unsigned char *begin, *cursor, c;
+    int i, index, value, size, count = 0, cut = 0, strict = 0;
+    enum {OPT_STRICT };
+    static const char *const optStrings[] = { "-strict", NULL };
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?options? data");
+	return TCL_ERROR;
+    }
+    for (i = 1; i < objc-1; ++i) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], optStrings, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case OPT_STRICT:
+	    strict = 1;
+	    break;
+	}
+    }
+
+    TclNewObj(resultObj);
+    datastart = data = (unsigned char *)
+	    TclGetStringFromObj(objv[objc-1], &count);
+    dataend = data + count;
+    size = (count + 1) / 2;
+    begin = cursor = Tcl_SetByteArrayLength(resultObj, size);
+    while (data < dataend) {
+	value = 0;
+	for (i=0 ; i<2 ; i++) {
+	    if (data >= dataend) {
+		value <<= 4;
+		break;
+	    }
+
+	    c = *data++;
+	    if (!isxdigit((int) c)) {
+		if (strict || !isspace(c)) {
+		    goto badChar;
+		}
+		i--;
+		continue;
+	    }
+
+	    value <<= 4;
+	    c -= '0';
+	    if (c > 9) {
+		c += ('0' - 'A') + 10;
+	    }
+	    if (c > 16) {
+		c += ('A' - 'a');
+	    }
+	    value |= (c & 0xf);
+	}
+	if (i < 2) {
+	    cut++;
+	}
+	*cursor++ = UCHAR(value);
+	value = 0;
+    }
+    if (cut > size) {
+	cut = size;
+    }
+    Tcl_SetByteArrayLength(resultObj, cursor - begin - cut);
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+
+  badChar:
+    TclDecrRefCount(resultObj);
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "invalid hexadecimal digit \"%c\" at position %d",
+	    c, (int) (data - datastart - 1)));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryEncode64 --
+ *
+ *	This implements a generic 6 bit binary encoding. Input is broken into
+ *	6 bit chunks and a lookup table passed in via clientData is used to
+ *	turn these values into output characters. This is used to implement
+ *	base64 binary encodings.
+ *
+ * Results:
+ *	Interp result set to an encoded byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+#define OUTPUT(c) \
+    do {						\
+	*cursor++ = (c);				\
+	outindex++;					\
+	if (maxlen > 0 && cursor != limit) {		\
+	    if (outindex == maxlen) {			\
+		memcpy(cursor, wrapchar, wrapcharlen);	\
+		cursor += wrapcharlen;			\
+		outindex = 0;				\
+	    }						\
+	}						\
+	if (cursor > limit) {				\
+	    Tcl_Panic("limit hit");			\
+	}						\
+    } while (0)
+
+static int
+BinaryEncode64(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj;
+    unsigned char *data, *cursor, *limit;
+    int maxlen = 0;
+    const char *wrapchar = "\n";
+    int wrapcharlen = 1;
+    int offset, i, index, size, outindex = 0, count = 0;
+    enum {OPT_MAXLEN, OPT_WRAPCHAR };
+    static const char *const optStrings[] = { "-maxlen", "-wrapchar", NULL };
+
+    if (objc < 2 || objc%2 != 0) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-maxlen len? ?-wrapchar char? data");
+	return TCL_ERROR;
+    }
+    for (i = 1; i < objc-1; i += 2) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], optStrings, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case OPT_MAXLEN:
+	    if (Tcl_GetIntFromObj(interp, objv[i+1], &maxlen) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (maxlen < 0) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"line length out of range", -1));
+		Tcl_SetErrorCode(interp, "TCL", "BINARY", "ENCODE",
+			"LINE_LENGTH", NULL);
+		return TCL_ERROR;
+	    }
+	    break;
+	case OPT_WRAPCHAR:
+	    wrapchar = TclGetStringFromObj(objv[i+1], &wrapcharlen);
+	    if (wrapcharlen == 0) {
+		maxlen = 0;
+	    }
+	    break;
+	}
+    }
+
+    resultObj = Tcl_NewObj();
+    data = Tcl_GetByteArrayFromObj(objv[objc-1], &count);
+    if (count > 0) {
+	size = (((count * 4) / 3) + 3) & ~3; /* ensure 4 byte chunks */
+	if (maxlen > 0 && size > maxlen) {
+	    int adjusted = size + (wrapcharlen * (size / maxlen));
+
+	    if (size % maxlen == 0) {
+		adjusted -= wrapcharlen;
+	    }
+	    size = adjusted;
+	}
+	cursor = Tcl_SetByteArrayLength(resultObj, size);
+	limit = cursor + size;
+	for (offset = 0; offset < count; offset+=3) {
+	    unsigned char d[3] = {0, 0, 0};
+
+	    for (i = 0; i < 3 && offset+i < count; ++i) {
+		d[i] = data[offset + i];
+	    }
+	    OUTPUT(B64Digits[d[0] >> 2]);
+	    OUTPUT(B64Digits[((d[0] & 0x03) << 4) | (d[1] >> 4)]);
+	    if (offset+1 < count) {
+		OUTPUT(B64Digits[((d[1] & 0x0f) << 2) | (d[2] >> 6)]);
+	    } else {
+		OUTPUT(B64Digits[64]);
+	    }
+	    if (offset+2 < count) {
+		OUTPUT(B64Digits[d[2] & 0x3f]);
+	    } else {
+		OUTPUT(B64Digits[64]);
+	    }
+	}
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+#undef OUTPUT
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryEncodeUu --
+ *
+ *	This implements the uuencode binary encoding. Input is broken into 6
+ *	bit chunks and a lookup table is used to turn these values into output
+ *	characters. This differs from the generic code above in that line
+ *	lengths are also encoded.
+ *
+ * Results:
+ *	Interp result set to an encoded byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryEncodeUu(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj;
+    unsigned char *data, *start, *cursor;
+    int offset, count, rawLength, n, i, j, bits, index;
+    int lineLength = 61;
+    const unsigned char SingleNewline[] = { (unsigned char) '\n' };
+    const unsigned char *wrapchar = SingleNewline;
+    int wrapcharlen = sizeof(SingleNewline);
+    enum { OPT_MAXLEN, OPT_WRAPCHAR };
+    static const char *const optStrings[] = { "-maxlen", "-wrapchar", NULL };
+
+    if (objc < 2 || objc%2 != 0) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-maxlen len? ?-wrapchar char? data");
+	return TCL_ERROR;
+    }
+    for (i = 1; i < objc-1; i += 2) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], optStrings, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case OPT_MAXLEN:
+	    if (Tcl_GetIntFromObj(interp, objv[i+1], &lineLength) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (lineLength < 3 || lineLength > 85) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"line length out of range", -1));
+		Tcl_SetErrorCode(interp, "TCL", "BINARY", "ENCODE",
+			"LINE_LENGTH", NULL);
+		return TCL_ERROR;
+	    }
+	    break;
+	case OPT_WRAPCHAR:
+	    wrapchar = Tcl_GetByteArrayFromObj(objv[i+1], &wrapcharlen);
+	    break;
+	}
+    }
+
+    /*
+     * Allocate the buffer. This is a little bit too long, but is "good
+     * enough".
+     */
+
+    resultObj = Tcl_NewObj();
+    offset = 0;
+    data = Tcl_GetByteArrayFromObj(objv[objc-1], &count);
+    rawLength = (lineLength - 1) * 3 / 4;
+    start = cursor = Tcl_SetByteArrayLength(resultObj,
+	    (lineLength + wrapcharlen) *
+	    ((count + (rawLength - 1)) / rawLength));
+    n = bits = 0;
+
+    /*
+     * Encode the data. Each output line first has the length of raw data
+     * encoded by the output line described in it by one encoded byte, then
+     * the encoded data follows (encoding each 6 bits as one character).
+     * Encoded lines are always terminated by a newline.
+     */
+
+    while (offset < count) {
+	int lineLen = count - offset;
+
+	if (lineLen > rawLength) {
+	    lineLen = rawLength;
+	}
+	*cursor++ = UueDigits[lineLen];
+	for (i=0 ; i<lineLen ; i++) {
+	    n <<= 8;
+	    n |= data[offset++];
+	    for (bits += 8; bits > 6 ; bits -= 6) {
+		*cursor++ = UueDigits[(n >> (bits-6)) & 0x3f];
+	    }
+	}
+	if (bits > 0) {
+	    n <<= 8;
+	    *cursor++ = UueDigits[(n >> (bits + 2)) & 0x3f];
+	    bits = 0;
+	}
+	for (j=0 ; j<wrapcharlen ; ++j) {
+	    *cursor++ = wrapchar[j];
+	}
+    }
+
+    /*
+     * Fix the length of the output bytearray.
+     */
+
+    Tcl_SetByteArrayLength(resultObj, cursor-start);
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryDecodeUu --
+ *
+ *	Decode a uuencoded string.
+ *
+ * Results:
+ *	Interp result set to an byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryDecodeUu(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj = NULL;
+    unsigned char *data, *datastart, *dataend;
+    unsigned char *begin, *cursor;
+    int i, index, size, count = 0, strict = 0, lineLen;
+    unsigned char c;
+    enum {OPT_STRICT };
+    static const char *const optStrings[] = { "-strict", NULL };
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?options? data");
+	return TCL_ERROR;
+    }
+    for (i = 1; i < objc-1; ++i) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], optStrings, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case OPT_STRICT:
+	    strict = 1;
+	    break;
+	}
+    }
+
+    TclNewObj(resultObj);
+    datastart = data = (unsigned char *)
+	    TclGetStringFromObj(objv[objc-1], &count);
+    dataend = data + count;
+    size = ((count + 3) & ~3) * 3 / 4;
+    begin = cursor = Tcl_SetByteArrayLength(resultObj, size);
+    lineLen = -1;
+
+    /*
+     * The decoding loop. First, we get the length of line (strictly, the
+     * number of data bytes we expect to generate from the line) we're
+     * processing this time round if it is not already known (i.e., when the
+     * lineLen variable is set to the magic value, -1).
+     */
+
+    while (data < dataend) {
+	char d[4] = {0, 0, 0, 0};
+
+	if (lineLen < 0) {
+	    c = *data++;
+	    if (c < 32 || c > 96) {
+		if (strict || !isspace(c)) {
+		    goto badUu;
+		}
+		i--;
+		continue;
+	    }
+	    lineLen = (c - 32) & 0x3f;
+	}
+
+	/*
+	 * Now we read a four-character grouping.
+	 */
+
+	for (i=0 ; i<4 ; i++) {
+	    if (data < dataend) {
+		d[i] = c = *data++;
+		if (c < 32 || c > 96) {
+		    if (strict) {
+			if (!isspace(c)) {
+			    goto badUu;
+			} else if (c == '\n') {
+			    goto shortUu;
+			}
+		    }
+		    i--;
+		    continue;
+		}
+	    }
+	}
+
+	/*
+	 * Translate that grouping into (up to) three binary bytes output.
+	 */
+
+	if (lineLen > 0) {
+	    *cursor++ = (((d[0] - 0x20) & 0x3f) << 2)
+		    | (((d[1] - 0x20) & 0x3f) >> 4);
+	    if (--lineLen > 0) {
+		*cursor++ = (((d[1] - 0x20) & 0x3f) << 4)
+			| (((d[2] - 0x20) & 0x3f) >> 2);
+		if (--lineLen > 0) {
+		    *cursor++ = (((d[2] - 0x20) & 0x3f) << 6)
+			    | (((d[3] - 0x20) & 0x3f));
+		    lineLen--;
+		}
+	    }
+	}
+
+	/*
+	 * If we've reached the end of the line, skip until we process a
+	 * newline.
+	 */
+
+	if (lineLen == 0 && data < dataend) {
+	    lineLen = -1;
+	    do {
+		c = *data++;
+		if (c == '\n') {
+		    break;
+		} else if (c >= 32 && c <= 96) {
+		    data--;
+		    break;
+		} else if (strict || !isspace(c)) {
+		    goto badUu;
+		}
+	    } while (data < dataend);
+	}
+    }
+
+    /*
+     * Sanity check, clean up and finish.
+     */
+
+    if (lineLen > 0 && strict) {
+	goto shortUu;
+    }
+    Tcl_SetByteArrayLength(resultObj, cursor - begin);
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+
+  shortUu:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf("short uuencode data"));
+    Tcl_SetErrorCode(interp, "TCL", "BINARY", "DECODE", "SHORT", NULL);
+    TclDecrRefCount(resultObj);
+    return TCL_ERROR;
+
+  badUu:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "invalid uuencode character \"%c\" at position %d",
+	    c, (int) (data - datastart - 1)));
+    Tcl_SetErrorCode(interp, "TCL", "BINARY", "DECODE", "INVALID", NULL);
+    TclDecrRefCount(resultObj);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BinaryDecode64 --
+ *
+ *	Decode a base64 encoded string.
+ *
+ * Results:
+ *	Interp result set to an byte array object
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BinaryDecode64(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *resultObj = NULL;
+    unsigned char *data, *datastart, *dataend, c = '\0';
+    unsigned char *begin = NULL;
+    unsigned char *cursor = NULL;
+    int strict = 0;
+    int i, index, size, cut = 0, count = 0;
+    enum { OPT_STRICT };
+    static const char *const optStrings[] = { "-strict", NULL };
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?options? data");
+	return TCL_ERROR;
+    }
+    for (i = 1; i < objc-1; ++i) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], optStrings, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case OPT_STRICT:
+	    strict = 1;
+	    break;
+	}
+    }
+
+    TclNewObj(resultObj);
+    datastart = data = (unsigned char *)
+	    TclGetStringFromObj(objv[objc-1], &count);
+    dataend = data + count;
+    size = ((count + 3) & ~3) * 3 / 4;
+    begin = cursor = Tcl_SetByteArrayLength(resultObj, size);
+    while (data < dataend) {
+	unsigned long value = 0;
+
+	/*
+	 * Decode the current block. Each base64 block consists of four input
+	 * characters A-Z, a-z, 0-9, +, or /. Each character supplies six bits
+	 * of output data, so each block's output is 24 bits (three bytes) in
+	 * length. The final block can be shorter by one or two bytes, denoted
+	 * by the input ending with one or two ='s, respectively.
+	 */
+
+	for (i = 0; i < 4; i++) {
+	    /*
+	     * Get the next input character. At end of input, pad with at most
+	     * two ='s. If more than two ='s would be needed, instead discard
+	     * the block read thus far.
+	     */
+
+	    if (data < dataend) {
+		c = *data++;
+	    } else if (i > 1) {
+		c = '=';
+	    } else {
+		cut += 3;
+		break;
+	    }
+
+	    /*
+	     * Load the character into the block value. Handle ='s specially
+	     * because they're only valid as the last character or two of the
+	     * final block of input. Unless strict mode is enabled, skip any
+	     * input whitespace characters.
+	     */
+
+	    if (cut) {
+		if (c == '=' && i > 1) {
+		     value <<= 6;
+		     cut++;
+		} else if (!strict && isspace(c)) {
+		     i--;
+		} else {
+		    goto bad64;
+		}
+	    } else if (c >= 'A' && c <= 'Z') {
+		value = (value << 6) | ((c - 'A') & 0x3f);
+	    } else if (c >= 'a' && c <= 'z') {
+		value = (value << 6) | ((c - 'a' + 26) & 0x3f);
+	    } else if (c >= '0' && c <= '9') {
+		value = (value << 6) | ((c - '0' + 52) & 0x3f);
+	    } else if (c == '+') {
+		value = (value << 6) | 0x3e;
+	    } else if (c == '/') {
+		value = (value << 6) | 0x3f;
+	    } else if (c == '=') {
+		value <<= 6;
+		cut++;
+	    } else if (strict || !isspace(c)) {
+		goto bad64;
+	    } else {
+		i--;
+	    }
+	}
+	*cursor++ = UCHAR((value >> 16) & 0xff);
+	*cursor++ = UCHAR((value >> 8) & 0xff);
+	*cursor++ = UCHAR(value & 0xff);
+
+	/*
+	 * Since = is only valid within the final block, if it was encountered
+	 * but there are still more input characters, confirm that strict mode
+	 * is off and all subsequent characters are whitespace.
+	 */
+
+	if (cut && data < dataend) {
+	    if (strict) {
+		goto bad64;
+	    }
+	    for (; data < dataend; data++) {
+		if (!isspace(*data)) {
+		    goto bad64;
+		}
+	    }
+	}
+    }
+    Tcl_SetByteArrayLength(resultObj, cursor - begin - cut);
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+
+  bad64:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "invalid base64 character \"%c\" at position %d",
+	    (char) c, (int) (data - datastart - 1)));
+    TclDecrRefCount(resultObj);
+    return TCL_ERROR;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
+
diff --git a/generic/tclCkalloc.c b/generic/tclCkalloc.c
new file mode 100644
index 0000000..123d872
--- /dev/null
+++ b/generic/tclCkalloc.c
@@ -0,0 +1,1330 @@
+/*
+ * tclCkalloc.c --
+ *
+ *    Interface to malloc and free that provides support for debugging
+ *    problems involving overwritten, double freeing memory and loss of
+ *    memory.
+ *
+ * Copyright (c) 1991-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ *
+ * This code contributed by Karl Lehenbauer and Mark Diekhans
+ */
+
+#include "tclInt.h"
+
+#define FALSE	0
+#define TRUE	1
+
+#undef Tcl_Alloc
+#undef Tcl_Free
+#undef Tcl_Realloc
+#undef Tcl_AttemptAlloc
+#undef Tcl_AttemptRealloc
+
+#ifdef TCL_MEM_DEBUG
+
+/*
+ * One of the following structures is allocated each time the
+ * "memory tag" command is invoked, to hold the current tag.
+ */
+
+typedef struct MemTag {
+    size_t refCount;		/* Number of mem_headers referencing this
+				 * tag. */
+    char string[1];		/* Actual size of string will be as large as
+				 * needed for actual tag. This must be the
+				 * last field in the structure. */
+} MemTag;
+
+#define TAG_SIZE(bytesInString) ((TclOffset(MemTag, string) + 1) + bytesInString)
+
+static MemTag *curTagPtr = NULL;/* Tag to use in all future mem_headers (set
+				 * by "memory tag" command). */
+
+/*
+ * One of the following structures is allocated just before each dynamically
+ * allocated chunk of memory, both to record information about the chunk and
+ * to help detect chunk under-runs.
+ */
+
+#define LOW_GUARD_SIZE (8 + (32 - (sizeof(size_t) + sizeof(int)))%8)
+struct mem_header {
+    struct mem_header *flink;
+    struct mem_header *blink;
+    MemTag *tagPtr;		/* Tag from "memory tag" command; may be
+				 * NULL. */
+    const char *file;
+    size_t length;
+    int line;
+    unsigned char low_guard[LOW_GUARD_SIZE];
+				/* Aligns body on 8-byte boundary, plus
+				 * provides at least 8 additional guard bytes
+				 * to detect underruns. */
+    char body[1];		/* First byte of client's space. Actual size
+				 * of this field will be larger than one. */
+};
+
+static struct mem_header *allocHead = NULL;  /* List of allocated structures */
+
+#define GUARD_VALUE  0141
+
+/*
+ * The following macro determines the amount of guard space *above* each chunk
+ * of memory.
+ */
+
+#define HIGH_GUARD_SIZE 8
+
+/*
+ * The following macro computes the offset of the "body" field within
+ * mem_header. It is used to get back to the header pointer from the body
+ * pointer that's used by clients.
+ */
+
+#define BODY_OFFSET \
+	((size_t) (&((struct mem_header *) 0)->body))
+
+static unsigned int total_mallocs = 0;
+static unsigned int total_frees = 0;
+static size_t current_bytes_malloced = 0;
+static size_t maximum_bytes_malloced = 0;
+static unsigned int current_malloc_packets = 0;
+static unsigned int  maximum_malloc_packets = 0;
+static unsigned int break_on_malloc = 0;
+static unsigned int trace_on_at_malloc = 0;
+static int alloc_tracing = FALSE;
+static int init_malloced_bodies = TRUE;
+#ifdef MEM_VALIDATE
+static int validate_memory = TRUE;
+#else
+static int validate_memory = FALSE;
+#endif
+
+/*
+ * The following variable indicates to TclFinalizeMemorySubsystem() that it
+ * should dump out the state of memory before exiting. If the value is
+ * non-NULL, it gives the name of the file in which to dump memory usage
+ * information.
+ */
+
+char *tclMemDumpFileName = NULL;
+
+static char *onExitMemDumpFileName = NULL;
+static char dumpFile[100];	/* Records where to dump memory allocation
+				 * information. */
+
+/*
+ * Mutex to serialize allocations. This is a low-level mutex that must be
+ * explicitly initialized. This is necessary because the self initializing
+ * mutexes use ckalloc...
+ */
+
+static Tcl_Mutex *ckallocMutexPtr;
+static int ckallocInit = 0;
+
+/*
+ * Prototypes for procedures defined in this file:
+ */
+
+static int		CheckmemCmd(ClientData clientData, Tcl_Interp *interp,
+			    int argc, const char *argv[]);
+static int		MemoryCmd(ClientData clientData, Tcl_Interp *interp,
+			    int argc, const char *argv[]);
+static void		ValidateMemory(struct mem_header *memHeaderP,
+			    const char *file, int line, int nukeGuards);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitDbCkalloc --
+ *
+ *	Initialize the locks used by the allocator. This is only appropriate
+ *	to call in a single threaded environment, such as during
+ *	TclInitSubsystems.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitDbCkalloc(void)
+{
+    if (!ckallocInit) {
+	ckallocInit = 1;
+	ckallocMutexPtr = Tcl_GetAllocMutex();
+#ifndef TCL_THREADS
+	/* Silence compiler warning */
+	(void)ckallocMutexPtr;
+#endif
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDumpMemoryInfo --
+ *
+ *	Display the global memory management statistics.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclDumpMemoryInfo(
+    ClientData clientData,
+    int flags)
+{
+    char buf[1024];
+
+    if (clientData == NULL) {
+        return 0;
+    }
+    sprintf(buf,
+	    "total mallocs             %10u\n"
+	    "total frees               %10u\n"
+	    "current packets allocated %10u\n"
+	    "current bytes allocated   %10" TCL_LL_MODIFIER "u\n"
+	    "maximum packets allocated %10u\n"
+	    "maximum bytes allocated   %10" TCL_LL_MODIFIER "u\n",
+	    total_mallocs,
+	    total_frees,
+	    current_malloc_packets,
+	    (Tcl_WideInt)current_bytes_malloced,
+	    maximum_malloc_packets,
+	    (Tcl_WideInt)maximum_bytes_malloced);
+    if (flags == 0) {
+	fprintf((FILE *)clientData, "%s", buf);
+    } else {
+	/* Assume objPtr to append to */
+	Tcl_AppendToObj((Tcl_Obj *) clientData, buf, -1);
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ValidateMemory --
+ *
+ *	Validate memory guard zones for a particular chunk of allocated
+ *	memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Prints validation information about the allocated memory to stderr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ValidateMemory(
+    struct mem_header *memHeaderP,
+				/* Memory chunk to validate */
+    const char *file,		/* File containing the call to
+				 * Tcl_ValidateAllMemory */
+    int line,			/* Line number of call to
+				 * Tcl_ValidateAllMemory */
+    int nukeGuards)		/* If non-zero, indicates that the memory
+				 * guards are to be reset to 0 after they have
+				 * been printed */
+{
+    unsigned char *hiPtr;
+    size_t idx;
+    int guard_failed = FALSE;
+    int byte;
+
+    for (idx = 0; idx < LOW_GUARD_SIZE; idx++) {
+	byte = *(memHeaderP->low_guard + idx);
+	if (byte != GUARD_VALUE) {
+	    guard_failed = TRUE;
+	    fflush(stdout);
+	    byte &= 0xff;
+	    fprintf(stderr, "low guard byte %d is 0x%x  \t%c\n", (int)idx, byte,
+		    (isprint(UCHAR(byte)) ? byte : ' ')); /* INTL: bytes */
+	}
+    }
+    if (guard_failed) {
+	TclDumpMemoryInfo((ClientData) stderr, 0);
+	fprintf(stderr, "low guard failed at %p, %s %d\n",
+		memHeaderP->body, file, line);
+	fflush(stderr);			/* In case name pointer is bad. */
+	fprintf(stderr, "%" TCL_LL_MODIFIER "d bytes allocated at (%s %d)\n", (Tcl_WideInt) memHeaderP->length,
+		memHeaderP->file, memHeaderP->line);
+	Tcl_Panic("Memory validation failure");
+    }
+
+    hiPtr = (unsigned char *)memHeaderP->body + memHeaderP->length;
+    for (idx = 0; idx < HIGH_GUARD_SIZE; idx++) {
+	byte = *(hiPtr + idx);
+	if (byte != GUARD_VALUE) {
+	    guard_failed = TRUE;
+	    fflush(stdout);
+	    byte &= 0xff;
+	    fprintf(stderr, "hi guard byte %d is 0x%x  \t%c\n", (int)idx, byte,
+		    (isprint(UCHAR(byte)) ? byte : ' ')); /* INTL: bytes */
+	}
+    }
+
+    if (guard_failed) {
+	TclDumpMemoryInfo((ClientData) stderr, 0);
+	fprintf(stderr, "high guard failed at %p, %s %d\n",
+		memHeaderP->body, file, line);
+	fflush(stderr);			/* In case name pointer is bad. */
+	fprintf(stderr, "%" TCL_LL_MODIFIER "d bytes allocated at (%s %d)\n",
+		(Tcl_WideInt)memHeaderP->length, memHeaderP->file,
+		memHeaderP->line);
+	Tcl_Panic("Memory validation failure");
+    }
+
+    if (nukeGuards) {
+	memset(memHeaderP->low_guard, 0, LOW_GUARD_SIZE);
+	memset(hiPtr, 0, HIGH_GUARD_SIZE);
+    }
+
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ValidateAllMemory --
+ *
+ *	Validate memory guard regions for all allocated memory.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Displays memory validation information to stderr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ValidateAllMemory(
+    const char *file,		/* File from which Tcl_ValidateAllMemory was
+				 * called. */
+    int line)			/* Line number of call to
+				 * Tcl_ValidateAllMemory */
+{
+    struct mem_header *memScanP;
+
+    if (!ckallocInit) {
+	TclInitDbCkalloc();
+    }
+    Tcl_MutexLock(ckallocMutexPtr);
+    for (memScanP = allocHead; memScanP != NULL; memScanP = memScanP->flink) {
+	ValidateMemory(memScanP, file, line, FALSE);
+    }
+    Tcl_MutexUnlock(ckallocMutexPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DumpActiveMemory --
+ *
+ *	Displays all allocated memory to a file; if no filename is given,
+ *	information will be written to stderr.
+ *
+ * Results:
+ *	Return TCL_ERROR if an error accessing the file occurs, `errno' will
+ *	have the file error number left in it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DumpActiveMemory(
+    const char *fileName)	/* Name of the file to write info to */
+{
+    FILE *fileP;
+    struct mem_header *memScanP;
+    char *address;
+
+    if (fileName == NULL) {
+	fileP = stderr;
+    } else {
+	fileP = fopen(fileName, "w");
+	if (fileP == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_MutexLock(ckallocMutexPtr);
+    for (memScanP = allocHead; memScanP != NULL; memScanP = memScanP->flink) {
+	address = &memScanP->body[0];
+	fprintf(fileP, "%p - %p  %" TCL_LL_MODIFIER "d @ %s %d %s",
+		address, address + memScanP->length - 1,
+		(Tcl_WideInt)memScanP->length, memScanP->file, memScanP->line,
+		(memScanP->tagPtr == NULL) ? "" : memScanP->tagPtr->string);
+	(void) fputc('\n', fileP);
+    }
+    Tcl_MutexUnlock(ckallocMutexPtr);
+
+    if (fileP != stderr) {
+	fclose(fileP);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbCkalloc - debugging ckalloc
+ *
+ *	Allocate the requested amount of space plus some extra for guard bands
+ *	at both ends of the request, plus a size, panicing if there isn't
+ *	enough space, then write in the guard bands and return the address of
+ *	the space in the middle that the user asked for.
+ *
+ *	The second and third arguments are file and line, these contain the
+ *	filename and line number corresponding to the caller. These are sent
+ *	by the ckalloc macro; it uses the preprocessor autodefines __FILE__
+ *	and __LINE__.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_DbCkalloc(
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    struct mem_header *result = NULL;
+
+    if (validate_memory) {
+	Tcl_ValidateAllMemory(file, line);
+    }
+
+    /* Don't let size argument to TclpAlloc overflow */
+    if (size <= UINT_MAX - HIGH_GUARD_SIZE -sizeof(struct mem_header)) {
+	result = (struct mem_header *) TclpAlloc((unsigned)size +
+		sizeof(struct mem_header) + HIGH_GUARD_SIZE);
+    }
+    if (result == NULL) {
+	fflush(stdout);
+	TclDumpMemoryInfo((ClientData) stderr, 0);
+	Tcl_Panic("unable to alloc %u bytes, %s line %d", size, file, line);
+    }
+
+    /*
+     * Fill in guard zones and size. Also initialize the contents of the block
+     * with bogus bytes to detect uses of initialized data. Link into
+     * allocated list.
+     */
+
+    if (init_malloced_bodies) {
+	memset(result, GUARD_VALUE,
+		size + sizeof(struct mem_header) + HIGH_GUARD_SIZE);
+    } else {
+	memset(result->low_guard, GUARD_VALUE, LOW_GUARD_SIZE);
+	memset(result->body + size, GUARD_VALUE, HIGH_GUARD_SIZE);
+    }
+    if (!ckallocInit) {
+	TclInitDbCkalloc();
+    }
+    Tcl_MutexLock(ckallocMutexPtr);
+    result->length = size;
+    result->tagPtr = curTagPtr;
+    if (curTagPtr != NULL) {
+	curTagPtr->refCount++;
+    }
+    result->file = file;
+    result->line = line;
+    result->flink = allocHead;
+    result->blink = NULL;
+
+    if (allocHead != NULL) {
+	allocHead->blink = result;
+    }
+    allocHead = result;
+
+    total_mallocs++;
+    if (trace_on_at_malloc && (total_mallocs >= trace_on_at_malloc)) {
+	(void) fflush(stdout);
+	fprintf(stderr, "reached malloc trace enable point (%u)\n",
+		total_mallocs);
+	fflush(stderr);
+	alloc_tracing = TRUE;
+	trace_on_at_malloc = 0;
+    }
+
+    if (alloc_tracing) {
+	fprintf(stderr,"ckalloc %p %u %s %d\n",
+		result->body, size, file, line);
+    }
+
+    if (break_on_malloc && (total_mallocs >= break_on_malloc)) {
+	break_on_malloc = 0;
+	(void) fflush(stdout);
+	Tcl_Panic("reached malloc break limit (%u)", total_mallocs);
+    }
+
+    current_malloc_packets++;
+    if (current_malloc_packets > maximum_malloc_packets) {
+	maximum_malloc_packets = current_malloc_packets;
+    }
+    current_bytes_malloced += size;
+    if (current_bytes_malloced > maximum_bytes_malloced) {
+	maximum_bytes_malloced = current_bytes_malloced;
+    }
+
+    Tcl_MutexUnlock(ckallocMutexPtr);
+
+    return result->body;
+}
+
+char *
+Tcl_AttemptDbCkalloc(
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    struct mem_header *result = NULL;
+
+    if (validate_memory) {
+	Tcl_ValidateAllMemory(file, line);
+    }
+
+    /* Don't let size argument to TclpAlloc overflow */
+    if (size <= UINT_MAX - HIGH_GUARD_SIZE - sizeof(struct mem_header)) {
+	result = (struct mem_header *) TclpAlloc((unsigned)size +
+		sizeof(struct mem_header) + HIGH_GUARD_SIZE);
+    }
+    if (result == NULL) {
+	fflush(stdout);
+	TclDumpMemoryInfo((ClientData) stderr, 0);
+	return NULL;
+    }
+
+    /*
+     * Fill in guard zones and size. Also initialize the contents of the block
+     * with bogus bytes to detect uses of initialized data. Link into
+     * allocated list.
+     */
+    if (init_malloced_bodies) {
+	memset(result, GUARD_VALUE,
+		size + sizeof(struct mem_header) + HIGH_GUARD_SIZE);
+    } else {
+	memset(result->low_guard, GUARD_VALUE, LOW_GUARD_SIZE);
+	memset(result->body + size, GUARD_VALUE, HIGH_GUARD_SIZE);
+    }
+    if (!ckallocInit) {
+	TclInitDbCkalloc();
+    }
+    Tcl_MutexLock(ckallocMutexPtr);
+    result->length = size;
+    result->tagPtr = curTagPtr;
+    if (curTagPtr != NULL) {
+	curTagPtr->refCount++;
+    }
+    result->file = file;
+    result->line = line;
+    result->flink = allocHead;
+    result->blink = NULL;
+
+    if (allocHead != NULL) {
+	allocHead->blink = result;
+    }
+    allocHead = result;
+
+    total_mallocs++;
+    if (trace_on_at_malloc && (total_mallocs >= trace_on_at_malloc)) {
+	(void) fflush(stdout);
+	fprintf(stderr, "reached malloc trace enable point (%d)\n",
+		total_mallocs);
+	fflush(stderr);
+	alloc_tracing = TRUE;
+	trace_on_at_malloc = 0;
+    }
+
+    if (alloc_tracing) {
+	fprintf(stderr,"ckalloc %p %u %s %d\n",
+		result->body, size, file, line);
+    }
+
+    if (break_on_malloc && (total_mallocs >= break_on_malloc)) {
+	break_on_malloc = 0;
+	(void) fflush(stdout);
+	Tcl_Panic("reached malloc break limit (%d)", total_mallocs);
+    }
+
+    current_malloc_packets++;
+    if (current_malloc_packets > maximum_malloc_packets) {
+	maximum_malloc_packets = current_malloc_packets;
+    }
+    current_bytes_malloced += size;
+    if (current_bytes_malloced > maximum_bytes_malloced) {
+	maximum_bytes_malloced = current_bytes_malloced;
+    }
+
+    Tcl_MutexUnlock(ckallocMutexPtr);
+
+    return result->body;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbCkfree - debugging ckfree
+ *
+ *	Verify that the low and high guards are intact, and if so then free
+ *	the buffer else Tcl_Panic.
+ *
+ *	The guards are erased after being checked to catch duplicate frees.
+ *
+ *	The second and third arguments are file and line, these contain the
+ *	filename and line number corresponding to the caller. These are sent
+ *	by the ckfree macro; it uses the preprocessor autodefines __FILE__ and
+ *	__LINE__.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DbCkfree(
+    char *ptr,
+    const char *file,
+    int line)
+{
+    struct mem_header *memp;
+
+    if (ptr == NULL) {
+	return;
+    }
+
+    /*
+     * The following cast is *very* tricky. Must convert the pointer to an
+     * integer before doing arithmetic on it, because otherwise the arithmetic
+     * will be done differently (and incorrectly) on word-addressed machines
+     * such as Crays (will subtract only bytes, even though BODY_OFFSET is in
+     * words on these machines).
+     */
+
+    memp = (struct mem_header *) (((size_t) ptr) - BODY_OFFSET);
+
+    if (alloc_tracing) {
+	fprintf(stderr, "ckfree %p %" TCL_LL_MODIFIER "d %s %d\n",
+		memp->body, (Tcl_WideInt) memp->length, file, line);
+    }
+
+    if (validate_memory) {
+	Tcl_ValidateAllMemory(file, line);
+    }
+
+    Tcl_MutexLock(ckallocMutexPtr);
+    ValidateMemory(memp, file, line, TRUE);
+    if (init_malloced_bodies) {
+	memset(ptr, GUARD_VALUE, memp->length);
+    }
+
+    total_frees++;
+    current_malloc_packets--;
+    current_bytes_malloced -= memp->length;
+
+    if (memp->tagPtr != NULL) {
+	if ((memp->tagPtr->refCount-- <= 1) && (curTagPtr != memp->tagPtr)) {
+	    TclpFree((char *) memp->tagPtr);
+	}
+    }
+
+    /*
+     * Delink from allocated list
+     */
+
+    if (memp->flink != NULL) {
+	memp->flink->blink = memp->blink;
+    }
+    if (memp->blink != NULL) {
+	memp->blink->flink = memp->flink;
+    }
+    if (allocHead == memp) {
+	allocHead = memp->flink;
+    }
+    TclpFree((char *) memp);
+    Tcl_MutexUnlock(ckallocMutexPtr);
+}
+
+/*
+ *--------------------------------------------------------------------
+ *
+ * Tcl_DbCkrealloc - debugging ckrealloc
+ *
+ *	Reallocate a chunk of memory by allocating a new one of the right
+ *	size, copying the old data to the new location, and then freeing the
+ *	old memory space, using all the memory checking features of this
+ *	package.
+ *
+ *--------------------------------------------------------------------
+ */
+
+char *
+Tcl_DbCkrealloc(
+    char *ptr,
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *newPtr;
+    size_t copySize;
+    struct mem_header *memp;
+
+    if (ptr == NULL) {
+	return Tcl_DbCkalloc(size, file, line);
+    }
+
+    /*
+     * See comment from Tcl_DbCkfree before you change the following line.
+     */
+
+    memp = (struct mem_header *) (((size_t) ptr) - BODY_OFFSET);
+
+    copySize = size;
+    if (copySize > memp->length) {
+	copySize = memp->length;
+    }
+    newPtr = Tcl_DbCkalloc(size, file, line);
+    memcpy(newPtr, ptr, (size_t) copySize);
+    Tcl_DbCkfree(ptr, file, line);
+    return newPtr;
+}
+
+char *
+Tcl_AttemptDbCkrealloc(
+    char *ptr,
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *newPtr;
+    size_t copySize;
+    struct mem_header *memp;
+
+    if (ptr == NULL) {
+	return Tcl_AttemptDbCkalloc(size, file, line);
+    }
+
+    /*
+     * See comment from Tcl_DbCkfree before you change the following line.
+     */
+
+    memp = (struct mem_header *) (((size_t) ptr) - BODY_OFFSET);
+
+    copySize = size;
+    if (copySize > memp->length) {
+	copySize = memp->length;
+    }
+    newPtr = Tcl_AttemptDbCkalloc(size, file, line);
+    if (newPtr == NULL) {
+	return NULL;
+    }
+    memcpy(newPtr, ptr, (size_t) copySize);
+    Tcl_DbCkfree(ptr, file, line);
+    return newPtr;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Alloc, et al. --
+ *
+ *	These functions are defined in terms of the debugging versions when
+ *	TCL_MEM_DEBUG is set.
+ *
+ * Results:
+ *	Same as the debug versions.
+ *
+ * Side effects:
+ *	Same as the debug versions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_Alloc(
+    unsigned int size)
+{
+    return Tcl_DbCkalloc(size, "unknown", 0);
+}
+
+char *
+Tcl_AttemptAlloc(
+    unsigned int size)
+{
+    return Tcl_AttemptDbCkalloc(size, "unknown", 0);
+}
+
+void
+Tcl_Free(
+    char *ptr)
+{
+    Tcl_DbCkfree(ptr, "unknown", 0);
+}
+
+char *
+Tcl_Realloc(
+    char *ptr,
+    unsigned int size)
+{
+    return Tcl_DbCkrealloc(ptr, size, "unknown", 0);
+}
+char *
+Tcl_AttemptRealloc(
+    char *ptr,
+    unsigned int size)
+{
+    return Tcl_AttemptDbCkrealloc(ptr, size, "unknown", 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MemoryCmd --
+ *
+ *	Implements the Tcl "memory" command, which provides Tcl-level control
+ *	of Tcl memory debugging information.
+ *		memory active $file
+ *		memory break_on_malloc $count
+ *		memory info
+ *		memory init on|off
+ *		memory onexit $file
+ *		memory tag $string
+ *		memory trace on|off
+ *		memory trace_on_at_malloc $count
+ *		memory validate on|off
+ *
+ * Results:
+ *	Standard TCL results.
+ *
+ *----------------------------------------------------------------------
+ */
+	/* ARGSUSED */
+static int
+MemoryCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int argc,
+    const char *argv[])
+{
+    const char *fileName;
+    FILE *fileP;
+    Tcl_DString buffer;
+    int result;
+    size_t len;
+
+    if (argc < 2) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "wrong # args: should be \"%s option [args..]\"", argv[0]));
+	return TCL_ERROR;
+    }
+
+    if (strcmp(argv[1], "active") == 0 || strcmp(argv[1], "display") == 0) {
+	if (argc != 3) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "wrong # args: should be \"%s %s file\"",
+                    argv[0], argv[1]));
+	    return TCL_ERROR;
+	}
+	fileName = Tcl_TranslateFileName(interp, argv[2], &buffer);
+	if (fileName == NULL) {
+	    return TCL_ERROR;
+	}
+	result = Tcl_DumpActiveMemory(fileName);
+	Tcl_DStringFree(&buffer);
+	if (result != TCL_OK) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf("error accessing %s: %s",
+                    argv[2], Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"break_on_malloc") == 0) {
+	int value;
+	if (argc != 3) {
+	    goto argError;
+	}
+	if (Tcl_GetInt(interp, argv[2], &value) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	break_on_malloc = (unsigned int) value;
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"info") == 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%-25s %10u\n%-25s %10u\n%-25s %10u\n%-25s %10" TCL_LL_MODIFIER"d\n%-25s %10u\n%-25s %10" TCL_LL_MODIFIER "d\n",
+		"total mallocs", total_mallocs, "total frees", total_frees,
+		"current packets allocated", current_malloc_packets,
+		"current bytes allocated", (Tcl_WideInt)current_bytes_malloced,
+		"maximum packets allocated", maximum_malloc_packets,
+		"maximum bytes allocated", (Tcl_WideInt)maximum_bytes_malloced));
+	return TCL_OK;
+    }
+    if (strcmp(argv[1], "init") == 0) {
+	if (argc != 3) {
+	    goto bad_suboption;
+	}
+	init_malloced_bodies = (strcmp(argv[2],"on") == 0);
+	return TCL_OK;
+    }
+    if (strcmp(argv[1], "objs") == 0) {
+	if (argc != 3) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "wrong # args: should be \"%s objs file\"", argv[0]));
+	    return TCL_ERROR;
+	}
+	fileName = Tcl_TranslateFileName(interp, argv[2], &buffer);
+	if (fileName == NULL) {
+	    return TCL_ERROR;
+	}
+	fileP = fopen(fileName, "w");
+	if (fileP == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "cannot open output file: %s",
+                    Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+	TclDbDumpActiveObjects(fileP);
+	fclose(fileP);
+	Tcl_DStringFree(&buffer);
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"onexit") == 0) {
+	if (argc != 3) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "wrong # args: should be \"%s onexit file\"", argv[0]));
+	    return TCL_ERROR;
+	}
+	fileName = Tcl_TranslateFileName(interp, argv[2], &buffer);
+	if (fileName == NULL) {
+	    return TCL_ERROR;
+	}
+	onExitMemDumpFileName = dumpFile;
+	strcpy(onExitMemDumpFileName,fileName);
+	Tcl_DStringFree(&buffer);
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"tag") == 0) {
+	if (argc != 3) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "wrong # args: should be \"%s tag string\"", argv[0]));
+	    return TCL_ERROR;
+	}
+	if ((curTagPtr != NULL) && (curTagPtr->refCount == 0)) {
+	    TclpFree((char *) curTagPtr);
+	}
+	len = strlen(argv[2]);
+	curTagPtr = (MemTag *) TclpAlloc(TAG_SIZE(len));
+	curTagPtr->refCount = 0;
+	memcpy(curTagPtr->string, argv[2], len + 1);
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"trace") == 0) {
+	if (argc != 3) {
+	    goto bad_suboption;
+	}
+	alloc_tracing = (strcmp(argv[2],"on") == 0);
+	return TCL_OK;
+    }
+
+    if (strcmp(argv[1],"trace_on_at_malloc") == 0) {
+	int value;
+	if (argc != 3) {
+	    goto argError;
+	}
+	if (Tcl_GetInt(interp, argv[2], &value) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	trace_on_at_malloc = value;
+	return TCL_OK;
+    }
+    if (strcmp(argv[1],"validate") == 0) {
+	if (argc != 3) {
+	    goto bad_suboption;
+	}
+	validate_memory = (strcmp(argv[2],"on") == 0);
+	return TCL_OK;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+            "bad option \"%s\": should be active, break_on_malloc, info, "
+            "init, objs, onexit, tag, trace, trace_on_at_malloc, or validate",
+            argv[1]));
+    return TCL_ERROR;
+
+  argError:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+            "wrong # args: should be \"%s %s count\"", argv[0], argv[1]));
+    return TCL_ERROR;
+
+  bad_suboption:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+            "wrong # args: should be \"%s %s on|off\"", argv[0], argv[1]));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckmemCmd --
+ *
+ *	This is the command procedure for the "checkmem" command, which causes
+ *	the application to exit after printing information about memory usage
+ *	to the file passed to this command as its first argument.
+ *
+ * Results:
+ *	Returns a standard Tcl completion code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckmemCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter for evaluation. */
+    int argc,			/* Number of arguments. */
+    const char *argv[])		/* String values of arguments. */
+{
+    if (argc != 2) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "wrong # args: should be \"%s fileName\"", argv[0]));
+	return TCL_ERROR;
+    }
+    tclMemDumpFileName = dumpFile;
+    strcpy(tclMemDumpFileName, argv[1]);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitMemory --
+ *
+ *	Create the "memory" and "checkmem" commands in the given interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	New commands are added to the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_InitMemory(
+    Tcl_Interp *interp)		/* Interpreter in which commands should be
+				 * added */
+{
+    TclInitDbCkalloc();
+    Tcl_CreateCommand(interp, "memory", MemoryCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "checkmem", CheckmemCmd, NULL, NULL);
+}
+
+
+#else	/* TCL_MEM_DEBUG */
+
+/* This is the !TCL_MEM_DEBUG case */
+
+#undef Tcl_InitMemory
+#undef Tcl_DumpActiveMemory
+#undef Tcl_ValidateAllMemory
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Alloc --
+ *
+ *	Interface to TclpAlloc when TCL_MEM_DEBUG is disabled. It does check
+ *	that memory was actually allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_Alloc(
+    unsigned int size)
+{
+    char *result;
+
+    result = TclpAlloc(size);
+
+    /*
+     * Most systems will not alloc(0), instead bumping it to one so that NULL
+     * isn't returned. Some systems (AIX, Tru64) will alloc(0) by returning
+     * NULL, so we have to check that the NULL we get is not in response to
+     * alloc(0).
+     *
+     * The ANSI spec actually says that systems either return NULL *or* a
+     * special pointer on failure, but we only check for NULL
+     */
+
+    if ((result == NULL) && size) {
+	Tcl_Panic("unable to alloc %u bytes", size);
+    }
+    return result;
+}
+
+char *
+Tcl_DbCkalloc(
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *result;
+
+    result = (char *) TclpAlloc(size);
+
+    if ((result == NULL) && size) {
+	fflush(stdout);
+	Tcl_Panic("unable to alloc %u bytes, %s line %d", size, file, line);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AttemptAlloc --
+ *
+ *	Interface to TclpAlloc when TCL_MEM_DEBUG is disabled. It does not
+ *	check that memory was actually allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_AttemptAlloc(
+    unsigned int size)
+{
+    char *result;
+
+    result = TclpAlloc(size);
+    return result;
+}
+
+char *
+Tcl_AttemptDbCkalloc(
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *result;
+
+    result = (char *) TclpAlloc(size);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Realloc --
+ *
+ *	Interface to TclpRealloc when TCL_MEM_DEBUG is disabled. It does check
+ *	that memory was actually allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_Realloc(
+    char *ptr,
+    unsigned int size)
+{
+    char *result;
+
+    result = TclpRealloc(ptr, size);
+
+    if ((result == NULL) && size) {
+	Tcl_Panic("unable to realloc %u bytes", size);
+    }
+    return result;
+}
+
+char *
+Tcl_DbCkrealloc(
+    char *ptr,
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *result;
+
+    result = (char *) TclpRealloc(ptr, size);
+
+    if ((result == NULL) && size) {
+	fflush(stdout);
+	Tcl_Panic("unable to realloc %u bytes, %s line %d", size, file, line);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AttemptRealloc --
+ *
+ *	Interface to TclpRealloc when TCL_MEM_DEBUG is disabled. It does not
+ *	check that memory was actually allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_AttemptRealloc(
+    char *ptr,
+    unsigned int size)
+{
+    char *result;
+
+    result = TclpRealloc(ptr, size);
+    return result;
+}
+
+char *
+Tcl_AttemptDbCkrealloc(
+    char *ptr,
+    unsigned int size,
+    const char *file,
+    int line)
+{
+    char *result;
+
+    result = (char *) TclpRealloc(ptr, size);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Free --
+ *
+ *	Interface to TclpFree when TCL_MEM_DEBUG is disabled. Done here rather
+ *	in the macro to keep some modules from being compiled with
+ *	TCL_MEM_DEBUG enabled and some with it disabled.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_Free(
+    char *ptr)
+{
+    TclpFree(ptr);
+}
+
+void
+Tcl_DbCkfree(
+    char *ptr,
+    const char *file,
+    int line)
+{
+    TclpFree(ptr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitMemory --
+ *
+ *	Dummy initialization for memory command, which is only available if
+ *	TCL_MEM_DEBUG is on.
+ *
+ *----------------------------------------------------------------------
+ */
+	/* ARGSUSED */
+void
+Tcl_InitMemory(
+    Tcl_Interp *interp)
+{
+}
+
+int
+Tcl_DumpActiveMemory(
+    const char *fileName)
+{
+    return TCL_OK;
+}
+
+void
+Tcl_ValidateAllMemory(
+    const char *file,
+    int line)
+{
+}
+
+int
+TclDumpMemoryInfo(
+    ClientData clientData,
+    int flags)
+{
+    return 1;
+}
+
+#endif	/* TCL_MEM_DEBUG */
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFinalizeMemorySubsystem --
+ *
+ *	This procedure is called to finalize all the structures that are used
+ *	by the memory allocator on a per-process basis.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	This subsystem is self-initializing, since memory can be allocated
+ *	before Tcl is formally initialized. After this call, this subsystem
+ *	has been reset to its initial state and is usable again.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclFinalizeMemorySubsystem(void)
+{
+#ifdef TCL_MEM_DEBUG
+    if (tclMemDumpFileName != NULL) {
+	Tcl_DumpActiveMemory(tclMemDumpFileName);
+    } else if (onExitMemDumpFileName != NULL) {
+	Tcl_DumpActiveMemory(onExitMemDumpFileName);
+    }
+
+    Tcl_MutexLock(ckallocMutexPtr);
+
+    if (curTagPtr != NULL) {
+	TclpFree((char *) curTagPtr);
+	curTagPtr = NULL;
+    }
+    allocHead = NULL;
+
+    Tcl_MutexUnlock(ckallocMutexPtr);
+#endif
+
+#if USE_TCLALLOC
+    TclFinalizeAllocSubsystem();
+#endif
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclClock.c b/generic/tclClock.c
new file mode 100644
index 0000000..bbfc83b
--- /dev/null
+++ b/generic/tclClock.c
@@ -0,0 +1,2090 @@
+/*
+ * tclClock.c --
+ *
+ *	Contains the time and date related commands. This code is derived from
+ *	the time and date facilities of TclX, by Mark Diekhans and Karl
+ *	Lehenbauer.
+ *
+ * Copyright 1991-1995 Karl Lehenbauer and Mark Diekhans.
+ * Copyright (c) 1995 Sun Microsystems, Inc.
+ * Copyright (c) 2004 by Kevin B. Kenny. All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Windows has mktime. The configurators do not check.
+ */
+
+#ifdef _WIN32
+#define HAVE_MKTIME 1
+#endif
+
+/*
+ * Constants
+ */
+
+#define JULIAN_DAY_POSIX_EPOCH		2440588
+#define SECONDS_PER_DAY			86400
+#define JULIAN_SEC_POSIX_EPOCH	      (((Tcl_WideInt) JULIAN_DAY_POSIX_EPOCH) \
+					* SECONDS_PER_DAY)
+#define FOUR_CENTURIES			146097	/* days */
+#define JDAY_1_JAN_1_CE_JULIAN		1721424
+#define JDAY_1_JAN_1_CE_GREGORIAN	1721426
+#define ONE_CENTURY_GREGORIAN		36524	/* days */
+#define FOUR_YEARS			1461	/* days */
+#define ONE_YEAR			365	/* days */
+
+/*
+ * Table of the days in each month, leap and common years
+ */
+
+static const int hath[2][12] = {
+    {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31},
+    {31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}
+};
+static const int daysInPriorMonths[2][13] = {
+    {0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365},
+    {0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366}
+};
+
+/*
+ * Enumeration of the string literals used in [clock]
+ */
+
+typedef enum ClockLiteral {
+    LIT__NIL,
+    LIT__DEFAULT_FORMAT,
+    LIT_BCE,		LIT_C,
+    LIT_CANNOT_USE_GMT_AND_TIMEZONE,
+    LIT_CE,
+    LIT_DAYOFMONTH,	LIT_DAYOFWEEK,		LIT_DAYOFYEAR,
+    LIT_ERA,		LIT_GMT,		LIT_GREGORIAN,
+    LIT_INTEGER_VALUE_TOO_LARGE,
+    LIT_ISO8601WEEK,	LIT_ISO8601YEAR,
+    LIT_JULIANDAY,	LIT_LOCALSECONDS,
+    LIT_MONTH,
+    LIT_SECONDS,	LIT_TZNAME,		LIT_TZOFFSET,
+    LIT_YEAR,
+    LIT__END
+} ClockLiteral;
+static const char *const literals[] = {
+    "",
+    "%a %b %d %H:%M:%S %Z %Y",
+    "BCE",		"C",
+    "cannot use -gmt and -timezone in same call",
+    "CE",
+    "dayOfMonth",	"dayOfWeek",		"dayOfYear",
+    "era",		":GMT",			"gregorian",
+    "integer value too large to represent",
+    "iso8601Week",	"iso8601Year",
+    "julianDay",	"localSeconds",
+    "month",
+    "seconds",		"tzName",		"tzOffset",
+    "year"
+};
+
+/*
+ * Structure containing the client data for [clock]
+ */
+
+typedef struct {
+    size_t refCount;		/* Number of live references. */
+    Tcl_Obj **literals;		/* Pool of object literals. */
+} ClockClientData;
+
+/*
+ * Structure containing the fields used in [clock format] and [clock scan]
+ */
+
+typedef struct TclDateFields {
+    Tcl_WideInt seconds;	/* Time expressed in seconds from the Posix
+				 * epoch */
+    Tcl_WideInt localSeconds;	/* Local time expressed in nominal seconds
+				 * from the Posix epoch */
+    int tzOffset;		/* Time zone offset in seconds east of
+				 * Greenwich */
+    Tcl_Obj *tzName;		/* Time zone name */
+    int julianDay;		/* Julian Day Number in local time zone */
+    enum {BCE=1, CE=0} era;	/* Era */
+    int gregorian;		/* Flag == 1 if the date is Gregorian */
+    int year;			/* Year of the era */
+    int dayOfYear;		/* Day of the year (1 January == 1) */
+    int month;			/* Month number */
+    int dayOfMonth;		/* Day of the month */
+    int iso8601Year;		/* ISO8601 week-based year */
+    int iso8601Week;		/* ISO8601 week number */
+    int dayOfWeek;		/* Day of the week */
+} TclDateFields;
+static const char *const eras[] = { "CE", "BCE", NULL };
+
+/*
+ * Thread specific data block holding a 'struct tm' for the 'gmtime' and
+ * 'localtime' library calls.
+ */
+
+static Tcl_ThreadDataKey tmKey;
+
+/*
+ * Mutex protecting 'gmtime', 'localtime' and 'mktime' calls and the statics
+ * in the date parsing code.
+ */
+
+TCL_DECLARE_MUTEX(clockMutex)
+
+/*
+ * Function prototypes for local procedures in this file:
+ */
+
+static int		ConvertUTCToLocal(Tcl_Interp *,
+			    TclDateFields *, Tcl_Obj *, int);
+static int		ConvertUTCToLocalUsingTable(Tcl_Interp *,
+			    TclDateFields *, int, Tcl_Obj *const[]);
+static int		ConvertUTCToLocalUsingC(Tcl_Interp *,
+			    TclDateFields *, int);
+static int		ConvertLocalToUTC(Tcl_Interp *,
+			    TclDateFields *, Tcl_Obj *, int);
+static int		ConvertLocalToUTCUsingTable(Tcl_Interp *,
+			    TclDateFields *, int, Tcl_Obj *const[]);
+static int		ConvertLocalToUTCUsingC(Tcl_Interp *,
+			    TclDateFields *, int);
+static Tcl_Obj *	LookupLastTransition(Tcl_Interp *, Tcl_WideInt,
+			    int, Tcl_Obj *const *);
+static void		GetYearWeekDay(TclDateFields *, int);
+static void		GetGregorianEraYearDay(TclDateFields *, int);
+static void		GetMonthDay(TclDateFields *);
+static void		GetJulianDayFromEraYearWeekDay(TclDateFields *, int);
+static void		GetJulianDayFromEraYearMonthDay(TclDateFields *, int);
+static int		IsGregorianLeapYear(TclDateFields *);
+static int		WeekdayOnOrBefore(int, int);
+static int		ClockClicksObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockConvertlocaltoutcObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockGetdatefieldsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockGetjuliandayfromerayearmonthdayObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockGetjuliandayfromerayearweekdayObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockGetenvObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockMicrosecondsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockMillisecondsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockParseformatargsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		ClockSecondsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static struct tm *	ThreadSafeLocalTime(const time_t *);
+static void		TzsetIfNecessary(void);
+static void		ClockDeleteCmdProc(ClientData);
+
+/*
+ * Structure containing description of "native" clock commands to create.
+ */
+
+struct ClockCommand {
+    const char *name;		/* The tail of the command name. The full name
+				 * is "::tcl::clock::<name>". When NULL marks
+				 * the end of the table. */
+    Tcl_ObjCmdProc *objCmdProc;	/* Function that implements the command. This
+				 * will always have the ClockClientData sent
+				 * to it, but may well ignore this data. */
+};
+
+static const struct ClockCommand clockCommands[] = {
+    { "getenv",			ClockGetenvObjCmd },
+    { "Oldscan",		TclClockOldscanObjCmd },
+    { "ConvertLocalToUTC",	ClockConvertlocaltoutcObjCmd },
+    { "GetDateFields",		ClockGetdatefieldsObjCmd },
+    { "GetJulianDayFromEraYearMonthDay",
+		ClockGetjuliandayfromerayearmonthdayObjCmd },
+    { "GetJulianDayFromEraYearWeekDay",
+		ClockGetjuliandayfromerayearweekdayObjCmd },
+    { "ParseFormatArgs",	ClockParseformatargsObjCmd },
+    { NULL, NULL }
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclClockInit --
+ *
+ *	Registers the 'clock' subcommands with the Tcl interpreter and
+ *	initializes its client data (which consists mostly of constant
+ *	Tcl_Obj's that it is too much trouble to keep recreating).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Installs the commands and creates the client data
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclClockInit(
+    Tcl_Interp *interp)		/* Tcl interpreter */
+{
+    const struct ClockCommand *clockCmdPtr;
+    char cmdName[50];		/* Buffer large enough to hold the string
+				 *::tcl::clock::GetJulianDayFromEraYearMonthDay
+				 * plus a terminating NUL. */
+    ClockClientData *data;
+    int i;
+
+    /* Structure of the 'clock' ensemble */
+
+    static const EnsembleImplMap clockImplMap[] = {
+	{"add",          NULL,                    TclCompileBasicMin1ArgCmd, NULL, NULL,       0},
+	{"clicks",       ClockClicksObjCmd,       TclCompileClockClicksCmd,  NULL, NULL,       0},
+	{"format",       NULL,                    TclCompileBasicMin1ArgCmd, NULL, NULL,       0},
+	{"microseconds", ClockMicrosecondsObjCmd, TclCompileClockReadingCmd, NULL, INT2PTR(1), 0},
+	{"milliseconds", ClockMillisecondsObjCmd, TclCompileClockReadingCmd, NULL, INT2PTR(2), 0},
+	{"scan",         NULL,                    TclCompileBasicMin1ArgCmd, NULL, NULL      , 0},
+	{"seconds",      ClockSecondsObjCmd,      TclCompileClockReadingCmd, NULL, INT2PTR(3), 0},
+	{NULL,           NULL,                    NULL,                      NULL, NULL,       0}
+    };
+
+    /*
+     * Safe interps get [::clock] as alias to a master, so do not need their
+     * own copies of the support routines.
+     */
+
+    if (Tcl_IsSafe(interp)) {
+	return;
+    }
+
+    /*
+     * Create the client data, which is a refcounted literal pool.
+     */
+
+    data = ckalloc(sizeof(ClockClientData));
+    data->refCount = 0;
+    data->literals = ckalloc(LIT__END * sizeof(Tcl_Obj*));
+    for (i = 0; i < LIT__END; ++i) {
+	data->literals[i] = Tcl_NewStringObj(literals[i], -1);
+	Tcl_IncrRefCount(data->literals[i]);
+    }
+
+    /*
+     * Install the commands.
+     * TODO - Let Tcl_MakeEnsemble do this?
+     */
+
+#define TCL_CLOCK_PREFIX_LEN 14 /* == strlen("::tcl::clock::") */
+    memcpy(cmdName, "::tcl::clock::", TCL_CLOCK_PREFIX_LEN);
+    for (clockCmdPtr=clockCommands ; clockCmdPtr->name!=NULL ; clockCmdPtr++) {
+	strcpy(cmdName + TCL_CLOCK_PREFIX_LEN, clockCmdPtr->name);
+	data->refCount++;
+	Tcl_CreateObjCommand(interp, cmdName, clockCmdPtr->objCmdProc, data,
+		ClockDeleteCmdProc);
+    }
+
+    /* Make the clock ensemble */
+
+    TclMakeEnsemble(interp, "clock", clockImplMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockConvertlocaltoutcObjCmd --
+ *
+ *	Tcl command that converts a UTC time to a local time by whatever means
+ *	is available.
+ *
+ * Usage:
+ *	::tcl::clock::ConvertUTCToLocal dictionary tzdata changeover
+ *
+ * Parameters:
+ *	dict - Dictionary containing a 'localSeconds' entry.
+ *	tzdata - Time zone data
+ *	changeover - Julian Day of the adoption of the Gregorian calendar.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	On success, sets the interpreter result to the given dictionary
+ *	augmented with a 'seconds' field giving the UTC time. On failure,
+ *	leaves an error message in the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ClockConvertlocaltoutcObjCmd(
+    ClientData clientData,	/* Client data */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter vector */
+{
+    ClockClientData *data = clientData;
+    Tcl_Obj *const *literals = data->literals;
+    Tcl_Obj *secondsObj;
+    Tcl_Obj *dict;
+    int changeover;
+    TclDateFields fields;
+    int created = 0;
+    int status;
+
+    /*
+     * Check params and convert time.
+     */
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dict tzdata changeover");
+	return TCL_ERROR;
+    }
+    dict = objv[1];
+    if (Tcl_DictObjGet(interp, dict, literals[LIT_LOCALSECONDS],
+	    &secondsObj)!= TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (secondsObj == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("key \"localseconds\" not "
+		"found in dictionary", -1));
+	return TCL_ERROR;
+    }
+    if ((TclGetWideIntFromObj(interp, secondsObj,
+	    &fields.localSeconds) != TCL_OK)
+	|| (TclGetIntFromObj(interp, objv[3], &changeover) != TCL_OK)
+	|| ConvertLocalToUTC(interp, &fields, objv[2], changeover)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Copy-on-write; set the 'seconds' field in the dictionary and place the
+     * modified dictionary in the interpreter result.
+     */
+
+    if (Tcl_IsShared(dict)) {
+	dict = Tcl_DuplicateObj(dict);
+	created = 1;
+	Tcl_IncrRefCount(dict);
+    }
+    status = Tcl_DictObjPut(interp, dict, literals[LIT_SECONDS],
+	    Tcl_NewWideIntObj(fields.seconds));
+    if (status == TCL_OK) {
+	Tcl_SetObjResult(interp, dict);
+    }
+    if (created) {
+	Tcl_DecrRefCount(dict);
+    }
+    return status;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockGetdatefieldsObjCmd --
+ *
+ *	Tcl command that determines the values that [clock format] will use in
+ *	formatting a date, and populates a dictionary with them.
+ *
+ * Usage:
+ *	::tcl::clock::GetDateFields seconds tzdata changeover
+ *
+ * Parameters:
+ *	seconds - Time expressed in seconds from the Posix epoch.
+ *	tzdata - Time zone data of the time zone in which time is to be
+ *		 expressed.
+ *	changeover - Julian Day Number at which the current locale adopted
+ *		     the Gregorian calendar
+ *
+ * Results:
+ *	Returns a dictonary populated with the fields:
+ *		seconds - Seconds from the Posix epoch
+ *		localSeconds - Nominal seconds from the Posix epoch in the
+ *			       local time zone.
+ *		tzOffset - Time zone offset in seconds east of Greenwich
+ *		tzName - Time zone name
+ *		julianDay - Julian Day Number in the local time zone
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockGetdatefieldsObjCmd(
+    ClientData clientData,	/* Opaque pointer to literal pool, etc. */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter vector */
+{
+    TclDateFields fields;
+    Tcl_Obj *dict;
+    ClockClientData *data = clientData;
+    Tcl_Obj *const *literals = data->literals;
+    int changeover;
+
+    /*
+     * Check params.
+     */
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "seconds tzdata changeover");
+	return TCL_ERROR;
+    }
+    if (TclGetWideIntFromObj(interp, objv[1], &fields.seconds) != TCL_OK
+	    || TclGetIntFromObj(interp, objv[3], &changeover) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * fields.seconds could be an unsigned number that overflowed. Make sure
+     * that it isn't.
+     */
+
+    if (objv[1]->typePtr == &tclBignumType) {
+	Tcl_SetObjResult(interp, literals[LIT_INTEGER_VALUE_TOO_LARGE]);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert UTC time to local.
+     */
+
+    if (ConvertUTCToLocal(interp, &fields, objv[2], changeover) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Extract Julian day.
+     */
+
+    fields.julianDay = (int) ((fields.localSeconds + JULIAN_SEC_POSIX_EPOCH)
+	    / SECONDS_PER_DAY);
+
+    /*
+     * Convert to Julian or Gregorian calendar.
+     */
+
+    GetGregorianEraYearDay(&fields, changeover);
+    GetMonthDay(&fields);
+    GetYearWeekDay(&fields, changeover);
+
+    dict = Tcl_NewDictObj();
+    Tcl_DictObjPut(NULL, dict, literals[LIT_LOCALSECONDS],
+	    Tcl_NewWideIntObj(fields.localSeconds));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_SECONDS],
+	    Tcl_NewWideIntObj(fields.seconds));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_TZNAME], fields.tzName);
+    Tcl_DecrRefCount(fields.tzName);
+    Tcl_DictObjPut(NULL, dict, literals[LIT_TZOFFSET],
+	    Tcl_NewIntObj(fields.tzOffset));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_JULIANDAY],
+	    Tcl_NewIntObj(fields.julianDay));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_GREGORIAN],
+	    Tcl_NewIntObj(fields.gregorian));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_ERA],
+	    literals[fields.era ? LIT_BCE : LIT_CE]);
+    Tcl_DictObjPut(NULL, dict, literals[LIT_YEAR],
+	    Tcl_NewIntObj(fields.year));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_DAYOFYEAR],
+	    Tcl_NewIntObj(fields.dayOfYear));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_MONTH],
+	    Tcl_NewIntObj(fields.month));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_DAYOFMONTH],
+	    Tcl_NewIntObj(fields.dayOfMonth));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_ISO8601YEAR],
+	    Tcl_NewIntObj(fields.iso8601Year));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_ISO8601WEEK],
+	    Tcl_NewIntObj(fields.iso8601Week));
+    Tcl_DictObjPut(NULL, dict, literals[LIT_DAYOFWEEK],
+	    Tcl_NewIntObj(fields.dayOfWeek));
+    Tcl_SetObjResult(interp, dict);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockGetjuliandayfromerayearmonthdayObjCmd --
+ *
+ *	Tcl command that converts a time from era-year-month-day to a Julian
+ *	Day Number.
+ *
+ * Parameters:
+ *	dict - Dictionary that contains 'era', 'year', 'month' and
+ *	       'dayOfMonth' keys.
+ *	changeover - Julian Day of changeover to the Gregorian calendar
+ *
+ * Results:
+ *	Result is either TCL_OK, with the interpreter result being the
+ *	dictionary augmented with a 'julianDay' key, or TCL_ERROR,
+ *	with the result being an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FetchEraField(
+    Tcl_Interp *interp,
+    Tcl_Obj *dict,
+    Tcl_Obj *key,
+    int *storePtr)
+{
+    Tcl_Obj *value = NULL;
+
+    if (Tcl_DictObjGet(interp, dict, key, &value) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (value == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"expected key(s) not found in dictionary", -1));
+	return TCL_ERROR;
+    }
+    return Tcl_GetIndexFromObj(interp, value, eras, "era", TCL_EXACT, storePtr);
+}
+
+static int
+FetchIntField(
+    Tcl_Interp *interp,
+    Tcl_Obj *dict,
+    Tcl_Obj *key,
+    int *storePtr)
+{
+    Tcl_Obj *value = NULL;
+
+    if (Tcl_DictObjGet(interp, dict, key, &value) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (value == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"expected key(s) not found in dictionary", -1));
+	return TCL_ERROR;
+    }
+    return TclGetIntFromObj(interp, value, storePtr);
+}
+
+static int
+ClockGetjuliandayfromerayearmonthdayObjCmd(
+    ClientData clientData,	/* Opaque pointer to literal pool, etc. */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter vector */
+{
+    TclDateFields fields;
+    Tcl_Obj *dict;
+    ClockClientData *data = clientData;
+    Tcl_Obj *const *literals = data->literals;
+    int changeover;
+    int copied = 0;
+    int status;
+    int era = 0;
+
+    /*
+     * Check params.
+     */
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dict changeover");
+	return TCL_ERROR;
+    }
+    dict = objv[1];
+    if (FetchEraField(interp, dict, literals[LIT_ERA], &era) != TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_YEAR], &fields.year)
+		!= TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_MONTH], &fields.month)
+		!= TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_DAYOFMONTH],
+		&fields.dayOfMonth) != TCL_OK
+	    || TclGetIntFromObj(interp, objv[2], &changeover) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    fields.era = era;
+
+    /*
+     * Get Julian day.
+     */
+
+    GetJulianDayFromEraYearMonthDay(&fields, changeover);
+
+    /*
+     * Store Julian day in the dictionary - copy on write.
+     */
+
+    if (Tcl_IsShared(dict)) {
+	dict = Tcl_DuplicateObj(dict);
+	Tcl_IncrRefCount(dict);
+	copied = 1;
+    }
+    status = Tcl_DictObjPut(interp, dict, literals[LIT_JULIANDAY],
+	    Tcl_NewIntObj(fields.julianDay));
+    if (status == TCL_OK) {
+	Tcl_SetObjResult(interp, dict);
+    }
+    if (copied) {
+	Tcl_DecrRefCount(dict);
+    }
+    return status;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockGetjuliandayfromerayearweekdayObjCmd --
+ *
+ *	Tcl command that converts a time from the ISO calendar to a Julian Day
+ *	Number.
+ *
+ * Parameters:
+ *	dict - Dictionary that contains 'era', 'iso8601Year', 'iso8601Week'
+ *	       and 'dayOfWeek' keys.
+ *	changeover - Julian Day of changeover to the Gregorian calendar
+ *
+ * Results:
+ *	Result is either TCL_OK, with the interpreter result being the
+ *	dictionary augmented with a 'julianDay' key, or TCL_ERROR, with the
+ *	result being an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ClockGetjuliandayfromerayearweekdayObjCmd(
+    ClientData clientData,	/* Opaque pointer to literal pool, etc. */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter vector */
+{
+    TclDateFields fields;
+    Tcl_Obj *dict;
+    ClockClientData *data = clientData;
+    Tcl_Obj *const *literals = data->literals;
+    int changeover;
+    int copied = 0;
+    int status;
+    int era = 0;
+
+    /*
+     * Check params.
+     */
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dict changeover");
+	return TCL_ERROR;
+    }
+    dict = objv[1];
+    if (FetchEraField(interp, dict, literals[LIT_ERA], &era) != TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_ISO8601YEAR],
+		&fields.iso8601Year) != TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_ISO8601WEEK],
+		&fields.iso8601Week) != TCL_OK
+	    || FetchIntField(interp, dict, literals[LIT_DAYOFWEEK],
+		&fields.dayOfWeek) != TCL_OK
+	    || TclGetIntFromObj(interp, objv[2], &changeover) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    fields.era = era;
+
+    /*
+     * Get Julian day.
+     */
+
+    GetJulianDayFromEraYearWeekDay(&fields, changeover);
+
+    /*
+     * Store Julian day in the dictionary - copy on write.
+     */
+
+    if (Tcl_IsShared(dict)) {
+	dict = Tcl_DuplicateObj(dict);
+	Tcl_IncrRefCount(dict);
+	copied = 1;
+    }
+    status = Tcl_DictObjPut(interp, dict, literals[LIT_JULIANDAY],
+	    Tcl_NewIntObj(fields.julianDay));
+    if (status == TCL_OK) {
+	Tcl_SetObjResult(interp, dict);
+    }
+    if (copied) {
+	Tcl_DecrRefCount(dict);
+    }
+    return status;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertLocalToUTC --
+ *
+ *	Converts a time (in a TclDateFields structure) from the local wall
+ *	clock to UTC.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Populates the 'seconds' field if successful; stores an error message
+ *	in the interpreter result on failure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertLocalToUTC(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Fields of the time */
+    Tcl_Obj *tzdata,		/* Time zone data */
+    int changeover)		/* Julian Day of the Gregorian transition */
+{
+    int rowc;			/* Number of rows in tzdata */
+    Tcl_Obj **rowv;		/* Pointers to the rows */
+
+    /*
+     * Unpack the tz data.
+     */
+
+    if (TclListObjGetElements(interp, tzdata, &rowc, &rowv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Special case: If the time zone is :localtime, the tzdata will be empty.
+     * Use 'mktime' to convert the time to local
+     */
+
+    if (rowc == 0) {
+	return ConvertLocalToUTCUsingC(interp, fields, changeover);
+    } else {
+	return ConvertLocalToUTCUsingTable(interp, fields, rowc, rowv);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertLocalToUTCUsingTable --
+ *
+ *	Converts a time (in a TclDateFields structure) from local time in a
+ *	given time zone to UTC.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Stores an error message in the interpreter if an error occurs; if
+ *	successful, stores the 'seconds' field in 'fields.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertLocalToUTCUsingTable(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Time to convert, with 'seconds' filled in */
+    int rowc,			/* Number of points at which time changes */
+    Tcl_Obj *const rowv[])	/* Points at which time changes */
+{
+    Tcl_Obj *row;
+    int cellc;
+    Tcl_Obj **cellv;
+    int have[8];
+    int nHave = 0;
+    int i;
+    int found;
+
+    /*
+     * Perform an initial lookup assuming that local == UTC, and locate the
+     * last time conversion prior to that time. Get the offset from that row,
+     * and look up again. Continue until we find an offset that we found
+     * before. This definition, rather than "the same offset" ensures that we
+     * don't enter an endless loop, as would otherwise happen when trying to
+     * convert a non-existent time such as 02:30 during the US Spring Daylight
+     * Saving Time transition.
+     */
+
+    found = 0;
+    fields->tzOffset = 0;
+    fields->seconds = fields->localSeconds;
+    while (!found) {
+	row = LookupLastTransition(interp, fields->seconds, rowc, rowv);
+	if ((row == NULL)
+		|| TclListObjGetElements(interp, row, &cellc,
+		    &cellv) != TCL_OK
+		|| TclGetIntFromObj(interp, cellv[1],
+		    &fields->tzOffset) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	found = 0;
+	for (i = 0; !found && i < nHave; ++i) {
+	    if (have[i] == fields->tzOffset) {
+		found = 1;
+		break;
+	    }
+	}
+	if (!found) {
+	    if (nHave == 8) {
+		Tcl_Panic("loop in ConvertLocalToUTCUsingTable");
+	    }
+	    have[nHave++] = fields->tzOffset;
+	}
+	fields->seconds = fields->localSeconds - fields->tzOffset;
+    }
+    fields->tzOffset = have[i];
+    fields->seconds = fields->localSeconds - fields->tzOffset;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertLocalToUTCUsingC --
+ *
+ *	Converts a time from local wall clock to UTC when the local time zone
+ *	cannot be determined. Uses 'mktime' to do the job.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Stores an error message in the interpreter if an error occurs; if
+ *	successful, stores the 'seconds' field in 'fields.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertLocalToUTCUsingC(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Time to convert, with 'seconds' filled in */
+    int changeover)		/* Julian Day of the Gregorian transition */
+{
+    struct tm timeVal;
+    int localErrno;
+    int secondOfDay;
+    Tcl_WideInt jsec;
+
+    /*
+     * Convert the given time to a date.
+     */
+
+    jsec = fields->localSeconds + JULIAN_SEC_POSIX_EPOCH;
+    fields->julianDay = (int) (jsec / SECONDS_PER_DAY);
+    secondOfDay = (int)(jsec % SECONDS_PER_DAY);
+    if (secondOfDay < 0) {
+	secondOfDay += SECONDS_PER_DAY;
+	fields->julianDay--;
+    }
+    GetGregorianEraYearDay(fields, changeover);
+    GetMonthDay(fields);
+
+    /*
+     * Convert the date/time to a 'struct tm'.
+     */
+
+    timeVal.tm_year = fields->year - 1900;
+    timeVal.tm_mon = fields->month - 1;
+    timeVal.tm_mday = fields->dayOfMonth;
+    timeVal.tm_hour = (secondOfDay / 3600) % 24;
+    timeVal.tm_min = (secondOfDay / 60) % 60;
+    timeVal.tm_sec = secondOfDay % 60;
+    timeVal.tm_isdst = -1;
+    timeVal.tm_wday = -1;
+    timeVal.tm_yday = -1;
+
+    /*
+     * Get local time. It is rumored that mktime is not thread safe on some
+     * platforms, so seize a mutex before attempting this.
+     */
+
+    TzsetIfNecessary();
+    Tcl_MutexLock(&clockMutex);
+    errno = 0;
+    fields->seconds = (Tcl_WideInt) mktime(&timeVal);
+    localErrno = errno;
+    Tcl_MutexUnlock(&clockMutex);
+
+    /*
+     * If conversion fails, report an error.
+     */
+
+    if (localErrno != 0
+	    || (fields->seconds == -1 && timeVal.tm_yday == -1)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"time value too large/small to represent", -1));
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertUTCToLocal --
+ *
+ *	Converts a time (in a TclDateFields structure) from UTC to local time.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Populates the 'tzName' and 'tzOffset' fields.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertUTCToLocal(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Fields of the time */
+    Tcl_Obj *tzdata,		/* Time zone data */
+    int changeover)		/* Julian Day of the Gregorian transition */
+{
+    int rowc;			/* Number of rows in tzdata */
+    Tcl_Obj **rowv;		/* Pointers to the rows */
+
+    /*
+     * Unpack the tz data.
+     */
+
+    if (TclListObjGetElements(interp, tzdata, &rowc, &rowv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Special case: If the time zone is :localtime, the tzdata will be empty.
+     * Use 'localtime' to convert the time to local
+     */
+
+    if (rowc == 0) {
+	return ConvertUTCToLocalUsingC(interp, fields, changeover);
+    } else {
+	return ConvertUTCToLocalUsingTable(interp, fields, rowc, rowv);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertUTCToLocalUsingTable --
+ *
+ *	Converts UTC to local time, given a table of transition points
+ *
+ * Results:
+ *	Returns a standard Tcl result
+ *
+ * Side effects:
+ *	On success, fills fields->tzName, fields->tzOffset and
+ *	fields->localSeconds. On failure, places an error message in the
+ *	interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertUTCToLocalUsingTable(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Fields of the date */
+    int rowc,			/* Number of rows in the conversion table
+				 * (>= 1) */
+    Tcl_Obj *const rowv[])	/* Rows of the conversion table */
+{
+    Tcl_Obj *row;		/* Row containing the current information */
+    int cellc;			/* Count of cells in the row (must be 4) */
+    Tcl_Obj **cellv;		/* Pointers to the cells */
+
+    /*
+     * Look up the nearest transition time.
+     */
+
+    row = LookupLastTransition(interp, fields->seconds, rowc, rowv);
+    if (row == NULL ||
+	    TclListObjGetElements(interp, row, &cellc, &cellv) != TCL_OK ||
+	    TclGetIntFromObj(interp, cellv[1], &fields->tzOffset) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert the time.
+     */
+
+    fields->tzName = cellv[3];
+    Tcl_IncrRefCount(fields->tzName);
+    fields->localSeconds = fields->seconds + fields->tzOffset;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertUTCToLocalUsingC --
+ *
+ *	Converts UTC to localtime in cases where the local time zone is not
+ *	determinable, using the C 'localtime' function to do it.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	On success, fills fields->tzName, fields->tzOffset and
+ *	fields->localSeconds. On failure, places an error message in the
+ *	interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ConvertUTCToLocalUsingC(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    TclDateFields *fields,	/* Time to convert, with 'seconds' filled in */
+    int changeover)		/* Julian Day of the Gregorian transition */
+{
+    time_t tock;
+    struct tm *timeVal;		/* Time after conversion */
+    int diff;			/* Time zone diff local-Greenwich */
+    char buffer[8];		/* Buffer for time zone name */
+
+    /*
+     * Use 'localtime' to determine local year, month, day, time of day.
+     */
+
+    tock = (time_t) fields->seconds;
+    if ((Tcl_WideInt) tock != fields->seconds) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"number too large to represent as a Posix time", -1));
+	Tcl_SetErrorCode(interp, "CLOCK", "argTooLarge", NULL);
+	return TCL_ERROR;
+    }
+    TzsetIfNecessary();
+    timeVal = ThreadSafeLocalTime(&tock);
+    if (timeVal == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"localtime failed (clock value may be too "
+		"large/small to represent)", -1));
+	Tcl_SetErrorCode(interp, "CLOCK", "localtimeFailed", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Fill in the date in 'fields' and use it to derive Julian Day.
+     */
+
+    fields->era = CE;
+    fields->year = timeVal->tm_year + 1900;
+    fields->month = timeVal->tm_mon + 1;
+    fields->dayOfMonth = timeVal->tm_mday;
+    GetJulianDayFromEraYearMonthDay(fields, changeover);
+
+    /*
+     * Convert that value to seconds.
+     */
+
+    fields->localSeconds = (((fields->julianDay * (Tcl_WideInt) 24
+	    + timeVal->tm_hour) * 60 + timeVal->tm_min) * 60
+	    + timeVal->tm_sec) - JULIAN_SEC_POSIX_EPOCH;
+
+    /*
+     * Determine a time zone offset and name; just use +hhmm for the name.
+     */
+
+    diff = (int) (fields->localSeconds - fields->seconds);
+    fields->tzOffset = diff;
+    if (diff < 0) {
+	*buffer = '-';
+	diff = -diff;
+    } else {
+	*buffer = '+';
+    }
+    sprintf(buffer+1, "%02d", diff / 3600);
+    diff %= 3600;
+    sprintf(buffer+3, "%02d", diff / 60);
+    diff %= 60;
+    if (diff > 0) {
+	sprintf(buffer+5, "%02d", diff);
+    }
+    fields->tzName = Tcl_NewStringObj(buffer, -1);
+    Tcl_IncrRefCount(fields->tzName);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * LookupLastTransition --
+ *
+ *	Given a UTC time and a tzdata array, looks up the last transition on
+ *	or before the given time.
+ *
+ * Results:
+ *	Returns a pointer to the row, or NULL if an error occurs.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+LookupLastTransition(
+    Tcl_Interp *interp,		/* Interpreter for error messages */
+    Tcl_WideInt tick,		/* Time from the epoch */
+    int rowc,			/* Number of rows of tzdata */
+    Tcl_Obj *const *rowv)	/* Rows in tzdata */
+{
+    int l;
+    int u;
+    Tcl_Obj *compObj;
+    Tcl_WideInt compVal;
+
+    /*
+     * Examine the first row to make sure we're in bounds.
+     */
+
+    if (Tcl_ListObjIndex(interp, rowv[0], 0, &compObj) != TCL_OK
+	    || TclGetWideIntFromObj(interp, compObj, &compVal) != TCL_OK) {
+	return NULL;
+    }
+
+    /*
+     * Bizarre case - first row doesn't begin at MIN_WIDE_INT. Return it
+     * anyway.
+     */
+
+    if (tick < compVal) {
+	return rowv[0];
+    }
+
+    /*
+     * Binary-search to find the transition.
+     */
+
+    l = 0;
+    u = rowc-1;
+    while (l < u) {
+	int m = (l + u + 1) / 2;
+
+	if (Tcl_ListObjIndex(interp, rowv[m], 0, &compObj) != TCL_OK ||
+		TclGetWideIntFromObj(interp, compObj, &compVal) != TCL_OK) {
+	    return NULL;
+	}
+	if (tick >= compVal) {
+	    l = m;
+	} else {
+	    u = m-1;
+	}
+    }
+    return rowv[l];
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetYearWeekDay --
+ *
+ *	Given a date with Julian Calendar Day, compute the year, week, and day
+ *	in the ISO8601 calendar.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores 'iso8601Year', 'iso8601Week' and 'dayOfWeek' in the date
+ *	fields.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetYearWeekDay(
+    TclDateFields *fields,	/* Date to convert, must have 'julianDay' */
+    int changeover)		/* Julian Day Number of the Gregorian
+				 * transition */
+{
+    TclDateFields temp;
+    int dayOfFiscalYear;
+
+    /*
+     * Find the given date, minus three days, plus one year. That date's
+     * iso8601 year is an upper bound on the ISO8601 year of the given date.
+     */
+
+    temp.julianDay = fields->julianDay - 3;
+    GetGregorianEraYearDay(&temp, changeover);
+    if (temp.era == BCE) {
+	temp.iso8601Year = temp.year - 1;
+    } else {
+	temp.iso8601Year = temp.year + 1;
+    }
+    temp.iso8601Week = 1;
+    temp.dayOfWeek = 1;
+    GetJulianDayFromEraYearWeekDay(&temp, changeover);
+
+    /*
+     * temp.julianDay is now the start of an ISO8601 year, either the one
+     * corresponding to the given date, or the one after. If we guessed high,
+     * move one year earlier
+     */
+
+    if (fields->julianDay < temp.julianDay) {
+	if (temp.era == BCE) {
+	    temp.iso8601Year += 1;
+	} else {
+	    temp.iso8601Year -= 1;
+	}
+	GetJulianDayFromEraYearWeekDay(&temp, changeover);
+    }
+
+    fields->iso8601Year = temp.iso8601Year;
+    dayOfFiscalYear = fields->julianDay - temp.julianDay;
+    fields->iso8601Week = (dayOfFiscalYear / 7) + 1;
+    fields->dayOfWeek = (dayOfFiscalYear + 1) % 7;
+    if (fields->dayOfWeek < 1) {
+	fields->dayOfWeek += 7;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetGregorianEraYearDay --
+ *
+ *	Given a Julian Day Number, extracts the year and day of the year and
+ *	puts them into TclDateFields, along with the era (BCE or CE) and a
+ *	flag indicating whether the date is Gregorian or Julian.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores 'era', 'gregorian', 'year', and 'dayOfYear'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetGregorianEraYearDay(
+    TclDateFields *fields,	/* Date fields containing 'julianDay' */
+    int changeover)		/* Gregorian transition date */
+{
+    int jday = fields->julianDay;
+    int day;
+    int year;
+    int n;
+
+    if (jday >= changeover) {
+	/*
+	 * Gregorian calendar.
+	 */
+
+	fields->gregorian = 1;
+	year = 1;
+
+	/*
+	 * n = Number of 400-year cycles since 1 January, 1 CE in the
+	 * proleptic Gregorian calendar. day = remaining days.
+	 */
+
+	day = jday - JDAY_1_JAN_1_CE_GREGORIAN;
+	n = day / FOUR_CENTURIES;
+	day %= FOUR_CENTURIES;
+	if (day < 0) {
+	    day += FOUR_CENTURIES;
+	    n--;
+	}
+	year += 400 * n;
+
+	/*
+	 * n = number of centuries since the start of (year);
+	 * day = remaining days
+	 */
+
+	n = day / ONE_CENTURY_GREGORIAN;
+	day %= ONE_CENTURY_GREGORIAN;
+	if (n > 3) {
+	    /*
+	     * 31 December in the last year of a 400-year cycle.
+	     */
+
+	    n = 3;
+	    day += ONE_CENTURY_GREGORIAN;
+	}
+	year += 100 * n;
+    } else {
+	/*
+	 * Julian calendar.
+	 */
+
+	fields->gregorian = 0;
+	year = 1;
+	day = jday - JDAY_1_JAN_1_CE_JULIAN;
+    }
+
+    /*
+     * n = number of 4-year cycles; days = remaining days.
+     */
+
+    n = day / FOUR_YEARS;
+    day %= FOUR_YEARS;
+    if (day < 0) {
+	day += FOUR_YEARS;
+	n--;
+    }
+    year += 4 * n;
+
+    /*
+     * n = number of years; days = remaining days.
+     */
+
+    n = day / ONE_YEAR;
+    day %= ONE_YEAR;
+    if (n > 3) {
+	/*
+	 * 31 December of a leap year.
+	 */
+
+	n = 3;
+	day += 365;
+    }
+    year += n;
+
+    /*
+     * store era/year/day back into fields.
+     */
+
+    if (year <= 0) {
+	fields->era = BCE;
+	fields->year = 1 - year;
+    } else {
+	fields->era = CE;
+	fields->year = year;
+    }
+    fields->dayOfYear = day + 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetMonthDay --
+ *
+ *	Given a date as year and day-of-year, find month and day.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores 'month' and 'dayOfMonth' in the 'fields' structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetMonthDay(
+    TclDateFields *fields)	/* Date to convert */
+{
+    int day = fields->dayOfYear;
+    int month;
+    const int *h = hath[IsGregorianLeapYear(fields)];
+
+    for (month = 0; month < 12 && day > h[month]; ++month) {
+	day -= h[month];
+    }
+    fields->month = month+1;
+    fields->dayOfMonth = day;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetJulianDayFromEraYearWeekDay --
+ *
+ *	Given a TclDateFields structure containing era, ISO8601 year, ISO8601
+ *	week, and day of week, computes the Julian Day Number.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores 'julianDay' in the fields.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetJulianDayFromEraYearWeekDay(
+    TclDateFields *fields,	/* Date to convert */
+    int changeover)		/* Julian Day Number of the Gregorian
+				 * transition */
+{
+    int firstMonday;		/* Julian day number of week 1, day 1 in the
+				 * given year */
+    TclDateFields firstWeek;
+
+    /*
+     * Find January 4 in the ISO8601 year, which will always be in week 1.
+     */
+
+    firstWeek.era = fields->era;
+    firstWeek.year = fields->iso8601Year;
+    firstWeek.month = 1;
+    firstWeek.dayOfMonth = 4;
+    GetJulianDayFromEraYearMonthDay(&firstWeek, changeover);
+
+    /*
+     * Find Monday of week 1.
+     */
+
+    firstMonday = WeekdayOnOrBefore(1, firstWeek.julianDay);
+
+    /*
+     * Advance to the given week and day.
+     */
+
+    fields->julianDay = firstMonday + 7 * (fields->iso8601Week - 1)
+	    + fields->dayOfWeek - 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetJulianDayFromEraYearMonthDay --
+ *
+ *	Given era, year, month, and dayOfMonth (in TclDateFields), and the
+ *	Gregorian transition date, computes the Julian Day Number.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores day number in 'julianDay'
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetJulianDayFromEraYearMonthDay(
+    TclDateFields *fields,	/* Date to convert */
+    int changeover)		/* Gregorian transition date as a Julian Day */
+{
+    int year, ym1, month, mm1, q, r, ym1o4, ym1o100, ym1o400;
+
+    if (fields->era == BCE) {
+	year = 1 - fields->year;
+    } else {
+	year = fields->year;
+    }
+
+    /*
+     * Reduce month modulo 12.
+     */
+
+    month = fields->month;
+    mm1 = month - 1;
+    q = mm1 / 12;
+    r = (mm1 % 12);
+    if (r < 0) {
+	r += 12;
+	q -= 1;
+    }
+    year += q;
+    month = r + 1;
+    ym1 = year - 1;
+
+    /*
+     * Adjust the year after reducing the month.
+     */
+
+    fields->gregorian = 1;
+    if (year < 1) {
+	fields->era = BCE;
+	fields->year = 1-year;
+    } else {
+	fields->era = CE;
+	fields->year = year;
+    }
+
+    /*
+     * Try an initial conversion in the Gregorian calendar.
+     */
+
+#if 0 /* BUG http://core.tcl.tk/tcl/tktview?name=da340d4f32 */
+    ym1o4 = ym1 / 4;
+#else
+    /*
+     * Have to make sure quotient is truncated towards 0 when negative.
+     * See above bug for details. The casts are necessary.
+     */
+    if (ym1 >= 0)
+	ym1o4 = ym1 / 4;
+    else {
+	ym1o4 = - (int) (((unsigned int) -ym1) / 4);
+    }
+#endif
+    if (ym1 % 4 < 0) {
+	ym1o4--;
+    }
+    ym1o100 = ym1 / 100;
+    if (ym1 % 100 < 0) {
+	ym1o100--;
+    }
+    ym1o400 = ym1 / 400;
+    if (ym1 % 400 < 0) {
+	ym1o400--;
+    }
+    fields->julianDay = JDAY_1_JAN_1_CE_GREGORIAN - 1
+	    + fields->dayOfMonth
+	    + daysInPriorMonths[IsGregorianLeapYear(fields)][month - 1]
+	    + (ONE_YEAR * ym1)
+	    + ym1o4
+	    - ym1o100
+	    + ym1o400;
+
+    /*
+     * If the resulting date is before the Gregorian changeover, convert in
+     * the Julian calendar instead.
+     */
+
+    if (fields->julianDay < changeover) {
+	fields->gregorian = 0;
+	fields->julianDay = JDAY_1_JAN_1_CE_JULIAN - 1
+		+ fields->dayOfMonth
+		+ daysInPriorMonths[year%4 == 0][month - 1]
+		+ (365 * ym1)
+		+ ym1o4;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IsGregorianLeapYear --
+ *
+ *	Tests whether a given year is a leap year, in either Julian or
+ *	Gregorian calendar.
+ *
+ * Results:
+ *	Returns 1 for a leap year, 0 otherwise.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+IsGregorianLeapYear(
+    TclDateFields *fields)	/* Date to test */
+{
+    int year = fields->year;
+
+    if (fields->era == BCE) {
+	year = 1 - year;
+    }
+    if (year%4 != 0) {
+	return 0;
+    } else if (!(fields->gregorian)) {
+	return 1;
+    } else if (year%400 == 0) {
+	return 1;
+    } else if (year%100 == 0) {
+	return 0;
+    } else {
+	return 1;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * WeekdayOnOrBefore --
+ *
+ *	Finds the Julian Day Number of a given day of the week that falls on
+ *	or before a given date, expressed as Julian Day Number.
+ *
+ * Results:
+ *	Returns the Julian Day Number
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+WeekdayOnOrBefore(
+    int dayOfWeek,		/* Day of week; Sunday == 0 or 7 */
+    int julianDay)		/* Reference date */
+{
+    int k = (dayOfWeek + 6) % 7;
+    if (k < 0) {
+	k += 7;
+    }
+    return julianDay - ((julianDay - k) % 7);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockGetenvObjCmd --
+ *
+ *	Tcl command that reads an environment variable from the system
+ *
+ * Usage:
+ *	::tcl::clock::getEnv NAME
+ *
+ * Parameters:
+ *	NAME - Name of the environment variable desired
+ *
+ * Results:
+ *	Returns a standard Tcl result. Returns an error if the variable does
+ *	not exist, with a message left in the interpreter. Returns TCL_OK and
+ *	the value of the variable if the variable does exist,
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockGetenvObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    const char *varName;
+    const char *varValue;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    varName = TclGetString(objv[1]);
+    varValue = getenv(varName);
+    if (varValue == NULL) {
+	varValue = "";
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(varValue, -1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ThreadSafeLocalTime --
+ *
+ *	Wrapper around the 'localtime' library function to make it thread
+ *	safe.
+ *
+ * Results:
+ *	Returns a pointer to a 'struct tm' in thread-specific data.
+ *
+ * Side effects:
+ *	Invokes localtime or localtime_r as appropriate.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static struct tm *
+ThreadSafeLocalTime(
+    const time_t *timePtr)	/* Pointer to the number of seconds since the
+				 * local system's epoch */
+{
+    /*
+     * Get a thread-local buffer to hold the returned time.
+     */
+
+    struct tm *tmPtr = Tcl_GetThreadData(&tmKey, sizeof(struct tm));
+#ifdef HAVE_LOCALTIME_R
+    localtime_r(timePtr, tmPtr);
+#else
+    struct tm *sysTmPtr;
+
+    Tcl_MutexLock(&clockMutex);
+    sysTmPtr = localtime(timePtr);
+    if (sysTmPtr == NULL) {
+	Tcl_MutexUnlock(&clockMutex);
+	return NULL;
+    }
+    memcpy(tmPtr, localtime(timePtr), sizeof(struct tm));
+    Tcl_MutexUnlock(&clockMutex);
+#endif
+    return tmPtr;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * ClockClicksObjCmd --
+ *
+ *	Returns a high-resolution counter.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ * This function implements the 'clock clicks' Tcl command. Refer to the user
+ * documentation for details on what it does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockClicksObjCmd(
+    ClientData clientData,	/* Client data is unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter values */
+{
+    static const char *const clicksSwitches[] = {
+	"-milliseconds", "-microseconds", NULL
+    };
+    enum ClicksSwitch {
+	CLICKS_MILLIS, CLICKS_MICROS, CLICKS_NATIVE
+    };
+    int index = CLICKS_NATIVE;
+    Tcl_Time now;
+    Tcl_WideInt clicks = 0;
+
+    switch (objc) {
+    case 1:
+	break;
+    case 2:
+	if (Tcl_GetIndexFromObj(interp, objv[1], clicksSwitches, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	break;
+    default:
+	Tcl_WrongNumArgs(interp, 1, objv, "?-switch?");
+	return TCL_ERROR;
+    }
+
+    switch (index) {
+    case CLICKS_MILLIS:
+	Tcl_GetTime(&now);
+	clicks = (Tcl_WideInt) now.sec * 1000 + now.usec / 1000;
+	break;
+    case CLICKS_NATIVE:
+#ifdef TCL_WIDE_CLICKS
+	clicks = TclpGetWideClicks();
+#else
+	clicks = (Tcl_WideInt) TclpGetClicks();
+#endif
+	break;
+    case CLICKS_MICROS:
+	Tcl_GetTime(&now);
+	clicks = ((Tcl_WideInt) now.sec * 1000000) + now.usec;
+	break;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj(clicks));
+    return TCL_OK;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * ClockMillisecondsObjCmd -
+ *
+ *	Returns a count of milliseconds since the epoch.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ * This function implements the 'clock milliseconds' Tcl command. Refer to the
+ * user documentation for details on what it does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockMillisecondsObjCmd(
+    ClientData clientData,	/* Client data is unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter values */
+{
+    Tcl_Time now;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_GetTime(&now);
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt)
+	    now.sec * 1000 + now.usec / 1000));
+    return TCL_OK;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * ClockMicrosecondsObjCmd -
+ *
+ *	Returns a count of microseconds since the epoch.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ * This function implements the 'clock microseconds' Tcl command. Refer to the
+ * user documentation for details on what it does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockMicrosecondsObjCmd(
+    ClientData clientData,	/* Client data is unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter values */
+{
+    Tcl_Time now;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_GetTime(&now);
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj(
+	    ((Tcl_WideInt) now.sec * 1000000) + now.usec));
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * ClockParseformatargsObjCmd --
+ *
+ *	Parses the arguments for [clock format].
+ *
+ * Results:
+ *	Returns a standard Tcl result, whose value is a four-element list
+ *	comprising the time format, the locale, and the timezone.
+ *
+ * This function exists because the loop that parses the [clock format]
+ * options is a known performance "hot spot", and is implemented in an effort
+ * to speed that particular code up.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+ClockParseformatargsObjCmd(
+    ClientData clientData,	/* Client data containing literal pool */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const objv[])	/* Parameter vector */
+{
+    ClockClientData *dataPtr = clientData;
+    Tcl_Obj **litPtr = dataPtr->literals;
+    Tcl_Obj *results[3];	/* Format, locale and timezone */
+#define formatObj results[0]
+#define localeObj results[1]
+#define timezoneObj results[2]
+    int gmtFlag = 0;
+    static const char *const options[] = { /* Command line options expected */
+	"-format",	"-gmt",		"-locale",
+	"-timezone",	NULL };
+    enum optionInd {
+	CLOCK_FORMAT_FORMAT,	CLOCK_FORMAT_GMT,	CLOCK_FORMAT_LOCALE,
+	CLOCK_FORMAT_TIMEZONE
+    };
+    int optionIndex;		/* Index of an option. */
+    int saw = 0;		/* Flag == 1 if option was seen already. */
+    Tcl_WideInt clockVal;	/* Clock value - just used to parse. */
+    int i;
+
+    /*
+     * Args consist of a time followed by keyword-value pairs.
+     */
+
+    if (objc < 2 || (objc % 2) != 0) {
+	Tcl_WrongNumArgs(interp, 0, objv,
+		"clock format clockval ?-format string? "
+		"?-gmt boolean? ?-locale LOCALE? ?-timezone ZONE?");
+	Tcl_SetErrorCode(interp, "CLOCK", "wrongNumArgs", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Extract values for the keywords.
+     */
+
+    formatObj = litPtr[LIT__DEFAULT_FORMAT];
+    localeObj = litPtr[LIT_C];
+    timezoneObj = litPtr[LIT__NIL];
+    for (i = 2; i < objc; i+=2) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		&optionIndex) != TCL_OK) {
+	    Tcl_SetErrorCode(interp, "CLOCK", "badOption",
+		    Tcl_GetString(objv[i]), NULL);
+	    return TCL_ERROR;
+	}
+	switch (optionIndex) {
+	case CLOCK_FORMAT_FORMAT:
+	    formatObj = objv[i+1];
+	    break;
+	case CLOCK_FORMAT_GMT:
+	    if (Tcl_GetBooleanFromObj(interp, objv[i+1], &gmtFlag) != TCL_OK){
+		return TCL_ERROR;
+	    }
+	    break;
+	case CLOCK_FORMAT_LOCALE:
+	    localeObj = objv[i+1];
+	    break;
+	case CLOCK_FORMAT_TIMEZONE:
+	    timezoneObj = objv[i+1];
+	    break;
+	}
+	saw |= 1 << optionIndex;
+    }
+
+    /*
+     * Check options.
+     */
+
+    if (TclGetWideIntFromObj(interp, objv[1], &clockVal) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if ((saw & (1 << CLOCK_FORMAT_GMT))
+	    && (saw & (1 << CLOCK_FORMAT_TIMEZONE))) {
+	Tcl_SetObjResult(interp, litPtr[LIT_CANNOT_USE_GMT_AND_TIMEZONE]);
+	Tcl_SetErrorCode(interp, "CLOCK", "gmtWithTimezone", NULL);
+	return TCL_ERROR;
+    }
+    if (gmtFlag) {
+	timezoneObj = litPtr[LIT_GMT];
+    }
+
+    /*
+     * Return options as a list.
+     */
+
+    Tcl_SetObjResult(interp, Tcl_NewListObj(3, results));
+    return TCL_OK;
+
+#undef timezoneObj
+#undef localeObj
+#undef formatObj
+}
+
+/*----------------------------------------------------------------------
+ *
+ * ClockSecondsObjCmd -
+ *
+ *	Returns a count of microseconds since the epoch.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ * This function implements the 'clock seconds' Tcl command. Refer to the user
+ * documentation for details on what it does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+ClockSecondsObjCmd(
+    ClientData clientData,	/* Client data is unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const *objv)	/* Parameter values */
+{
+    Tcl_Time now;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_GetTime(&now);
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt) now.sec));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TzsetIfNecessary --
+ *
+ *	Calls the tzset() library function if the contents of the TZ
+ *	environment variable has changed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Calls tzset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TzsetIfNecessary(void)
+{
+    static char* tzWas = INT2PTR(-1);	/* Previous value of TZ, protected by
+				 * clockMutex. */
+    const char *tzIsNow;	/* Current value of TZ */
+
+    Tcl_MutexLock(&clockMutex);
+    tzIsNow = getenv("TZ");
+    if (tzIsNow != NULL && (tzWas == NULL || tzWas == INT2PTR(-1)
+	    || strcmp(tzIsNow, tzWas) != 0)) {
+	tzset();
+	if (tzWas != NULL && tzWas != INT2PTR(-1)) {
+	    ckfree(tzWas);
+	}
+	tzWas = ckalloc(strlen(tzIsNow) + 1);
+	strcpy(tzWas, tzIsNow);
+    } else if (tzIsNow == NULL && tzWas != NULL) {
+	tzset();
+	if (tzWas != INT2PTR(-1)) ckfree(tzWas);
+	tzWas = NULL;
+    }
+    Tcl_MutexUnlock(&clockMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClockDeleteCmdProc --
+ *
+ *	Remove a reference to the clock client data, and clean up memory
+ *	when it's all gone.
+ *
+ * Results:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ClockDeleteCmdProc(
+    ClientData clientData)	/* Opaque pointer to the client data */
+{
+    ClockClientData *data = clientData;
+    int i;
+
+    if (data->refCount-- <= 1) {
+	for (i = 0; i < LIT__END; ++i) {
+	    Tcl_DecrRefCount(data->literals[i]);
+	}
+	ckfree(data->literals);
+	ckfree(data);
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCmdAH.c b/generic/tclCmdAH.c
new file mode 100644
index 0000000..807a1ac
--- /dev/null
+++ b/generic/tclCmdAH.c
@@ -0,0 +1,3232 @@
+/*
+ * tclCmdAH.c --
+ *
+ *	This file contains the top-level command routines for most of the Tcl
+ *	built-in commands whose names begin with the letters A to H.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#ifdef _WIN32
+#   include "tclWinInt.h"
+#endif
+#include <locale.h>
+
+/*
+ * The state structure used by [foreach]. Note that the actual structure has
+ * all its working arrays appended afterwards so they can be allocated and
+ * freed in a single step.
+ */
+
+struct ForeachState {
+    Tcl_Obj *bodyPtr;		/* The script body of the command. */
+    int bodyIdx;		/* The argument index of the body. */
+    int j, maxj;		/* Number of loop iterations. */
+    int numLists;		/* Count of value lists. */
+    int *index;			/* Array of value list indices. */
+    int *varcList;		/* # loop variables per list. */
+    Tcl_Obj ***varvList;	/* Array of var name lists. */
+    Tcl_Obj **vCopyList;	/* Copies of var name list arguments. */
+    int *argcList;		/* Array of value list sizes. */
+    Tcl_Obj ***argvList;	/* Array of value lists. */
+    Tcl_Obj **aCopyList;	/* Copies of value list arguments. */
+    Tcl_Obj *resultList;	/* List of result values from the loop body,
+				 * or NULL if we're not collecting them
+				 * ([lmap] vs [foreach]). */
+};
+
+/*
+ * Prototypes for local procedures defined in this file:
+ */
+
+static int		CheckAccess(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    int mode);
+static int		BadEncodingSubcommand(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		EncodingConvertfromObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		EncodingConverttoObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		EncodingDirsObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		EncodingNamesObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		EncodingSystemObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static inline int	ForeachAssignments(Tcl_Interp *interp,
+			    struct ForeachState *statePtr);
+static inline void	ForeachCleanup(Tcl_Interp *interp,
+			    struct ForeachState *statePtr);
+static int		GetStatBuf(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    Tcl_FSStatProc *statProc, Tcl_StatBuf *statPtr);
+static const char *	GetTypeFromMode(int mode);
+static int		StoreStatData(Tcl_Interp *interp, Tcl_Obj *varName,
+			    Tcl_StatBuf *statPtr);
+static inline int	EachloopCmd(Tcl_Interp *interp, int collect,
+			    int objc, Tcl_Obj *const objv[]);
+static Tcl_NRPostProc	CatchObjCmdCallback;
+static Tcl_NRPostProc	ExprCallback;
+static Tcl_NRPostProc	ForSetupCallback;
+static Tcl_NRPostProc	ForCondCallback;
+static Tcl_NRPostProc	ForNextCallback;
+static Tcl_NRPostProc	ForPostNextCallback;
+static Tcl_NRPostProc	ForeachLoopStep;
+static Tcl_NRPostProc	EvalCmdErrMsg;
+
+static Tcl_ObjCmdProc	BadFileSubcommand;
+static Tcl_ObjCmdProc FileAttrAccessTimeCmd;
+static Tcl_ObjCmdProc FileAttrIsDirectoryCmd;
+static Tcl_ObjCmdProc FileAttrIsExecutableCmd;
+static Tcl_ObjCmdProc FileAttrIsExistingCmd;
+static Tcl_ObjCmdProc FileAttrIsFileCmd;
+static Tcl_ObjCmdProc FileAttrIsOwnedCmd;
+static Tcl_ObjCmdProc FileAttrIsReadableCmd;
+static Tcl_ObjCmdProc FileAttrIsWritableCmd;
+static Tcl_ObjCmdProc FileAttrLinkStatCmd;
+static Tcl_ObjCmdProc FileAttrModifyTimeCmd;
+static Tcl_ObjCmdProc FileAttrSizeCmd;
+static Tcl_ObjCmdProc FileAttrStatCmd;
+static Tcl_ObjCmdProc FileAttrTypeCmd;
+static Tcl_ObjCmdProc FilesystemSeparatorCmd;
+static Tcl_ObjCmdProc FilesystemVolumesCmd;
+static Tcl_ObjCmdProc PathDirNameCmd;
+static Tcl_ObjCmdProc PathExtensionCmd;
+static Tcl_ObjCmdProc PathFilesystemCmd;
+static Tcl_ObjCmdProc PathJoinCmd;
+static Tcl_ObjCmdProc PathNativeNameCmd;
+static Tcl_ObjCmdProc PathNormalizeCmd;
+static Tcl_ObjCmdProc PathRootNameCmd;
+static Tcl_ObjCmdProc PathSplitCmd;
+static Tcl_ObjCmdProc PathTailCmd;
+static Tcl_ObjCmdProc PathTypeCmd;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_BreakObjCmd --
+ *
+ *	This procedure is invoked to process the "break" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "break" or the name to
+ *	which "break" was renamed: e.g., "set z break; $z"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_BreakObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    return TCL_BREAK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CaseObjCmd --
+ *
+ *	This procedure is invoked to process the "case" Tcl command. See the
+ *	user documentation for details on what it does. THIS COMMAND IS
+ *	OBSOLETE AND DEPRECATED. SLATED FOR REMOVAL IN TCL 9.0.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+#ifndef TCL_NO_DEPRECATED
+	/* ARGSUSED */
+int
+Tcl_CaseObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register int i;
+    int body, result, caseObjc;
+    const char *stringPtr, *arg;
+    Tcl_Obj *const *caseObjv;
+    Tcl_Obj *armPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"string ?in? ?pattern body ...? ?default body?");
+	return TCL_ERROR;
+    }
+
+    stringPtr = TclGetString(objv[1]);
+    body = -1;
+
+    arg = TclGetString(objv[2]);
+    if (strcmp(arg, "in") == 0) {
+	i = 3;
+    } else {
+	i = 2;
+    }
+    caseObjc = objc - i;
+    caseObjv = objv + i;
+
+    /*
+     * If all of the pattern/command pairs are lumped into a single argument,
+     * split them out again.
+     */
+
+    if (caseObjc == 1) {
+	Tcl_Obj **newObjv;
+
+	TclListObjGetElements(interp, caseObjv[0], &caseObjc, &newObjv);
+	caseObjv = newObjv;
+    }
+
+    for (i = 0;  i < caseObjc;  i += 2) {
+	int patObjc, j;
+	const char **patObjv;
+	const char *pat, *p;
+
+	if (i == caseObjc-1) {
+	    Tcl_ResetResult(interp);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "extra case pattern with no body", -1));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Check for special case of single pattern (no list) with no
+	 * backslash sequences.
+	 */
+
+	pat = TclGetString(caseObjv[i]);
+	for (p = pat; *p != '\0'; p++) {
+	    if (TclIsSpaceProc(*p) || (*p == '\\')) {
+		break;
+	    }
+	}
+	if (*p == '\0') {
+	    if ((*pat == 'd') && (strcmp(pat, "default") == 0)) {
+		body = i + 1;
+	    }
+	    if (Tcl_StringMatch(stringPtr, pat)) {
+		body = i + 1;
+		goto match;
+	    }
+	    continue;
+	}
+
+	/*
+	 * Break up pattern lists, then check each of the patterns in the
+	 * list.
+	 */
+
+	result = Tcl_SplitList(interp, pat, &patObjc, &patObjv);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	for (j = 0; j < patObjc; j++) {
+	    if (Tcl_StringMatch(stringPtr, patObjv[j])) {
+		body = i + 1;
+		break;
+	    }
+	}
+	ckfree(patObjv);
+	if (j < patObjc) {
+	    break;
+	}
+    }
+
+  match:
+    if (body != -1) {
+	armPtr = caseObjv[body - 1];
+	result = Tcl_EvalObjEx(interp, caseObjv[body], 0);
+	if (result == TCL_ERROR) {
+	    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		    "\n    (\"%.50s\" arm line %d)",
+		    TclGetString(armPtr), Tcl_GetErrorLine(interp)));
+	}
+	return result;
+    }
+
+    /*
+     * Nothing matched: return nothing.
+     */
+
+    return TCL_OK;
+}
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CatchObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "catch" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_CatchObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRCatchObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRCatchObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *varNamePtr = NULL;
+    Tcl_Obj *optionVarNamePtr = NULL;
+    Interp *iPtr = (Interp *) interp;
+
+    if ((objc < 2) || (objc > 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"script ?resultVarName? ?optionVarName?");
+	return TCL_ERROR;
+    }
+
+    if (objc >= 3) {
+	varNamePtr = objv[2];
+    }
+    if (objc == 4) {
+	optionVarNamePtr = objv[3];
+    }
+
+    TclNRAddCallback(interp, CatchObjCmdCallback, INT2PTR(objc),
+	    varNamePtr, optionVarNamePtr, NULL);
+
+    /*
+     * TIP #280. Make invoking context available to caught script.
+     */
+
+    return TclNREvalObjEx(interp, objv[1], 0, iPtr->cmdFramePtr, 1);
+}
+
+static int
+CatchObjCmdCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    int objc = PTR2INT(data[0]);
+    Tcl_Obj *varNamePtr = data[1];
+    Tcl_Obj *optionVarNamePtr = data[2];
+    int rewind = iPtr->execEnvPtr->rewind;
+
+    /*
+     * We disable catch in interpreters where the limit has been exceeded.
+     */
+
+    if (rewind || Tcl_LimitExceeded(interp)) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"catch\" body line %d)", Tcl_GetErrorLine(interp)));
+	return TCL_ERROR;
+    }
+
+    if (objc >= 3) {
+	if (NULL == Tcl_ObjSetVar2(interp, varNamePtr, NULL,
+		Tcl_GetObjResult(interp), TCL_LEAVE_ERR_MSG)) {
+	    return TCL_ERROR;
+	}
+    }
+    if (objc == 4) {
+	Tcl_Obj *options = Tcl_GetReturnOptions(interp, result);
+
+	if (NULL == Tcl_ObjSetVar2(interp, optionVarNamePtr, NULL,
+		options, TCL_LEAVE_ERR_MSG)) {
+	    /* Do not decrRefCount 'options', it was already done by
+	     * Tcl_ObjSetVar2 */
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_ResetResult(interp);
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(result));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CdObjCmd --
+ *
+ *	This procedure is invoked to process the "cd" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_CdObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *dir;
+    int result;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?dirName?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	dir = objv[1];
+    } else {
+	TclNewLiteralStringObj(dir, "~");
+	Tcl_IncrRefCount(dir);
+    }
+    if (Tcl_FSConvertToPathType(interp, dir) != TCL_OK) {
+	result = TCL_ERROR;
+    } else {
+	result = Tcl_FSChdir(dir);
+	if (result != TCL_OK) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "couldn't change working directory to \"%s\": %s",
+		    TclGetString(dir), Tcl_PosixError(interp)));
+	    result = TCL_ERROR;
+	}
+    }
+    if (objc != 2) {
+	Tcl_DecrRefCount(dir);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConcatObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "concat" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ConcatObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc >= 2) {
+	Tcl_SetObjResult(interp, Tcl_ConcatObj(objc-1, objv+1));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ContinueObjCmd --
+ *
+ *	This procedure is invoked to process the "continue" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "continue" or the name to
+ *	which "continue" was renamed: e.g., "set z continue; $z"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ContinueObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    return TCL_CONTINUE;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EncodingObjCmd --
+ *
+ *	This command manipulates encodings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_EncodingObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int index;
+
+    static const char *const optionStrings[] = {
+	"convertfrom", "convertto", "dirs", "names", "system",
+	NULL
+    };
+    enum options {
+	ENC_CONVERTFROM, ENC_CONVERTTO, ENC_DIRS, ENC_NAMES, ENC_SYSTEM
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], optionStrings, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum options) index) {
+    case ENC_CONVERTTO:
+	return EncodingConverttoObjCmd(dummy, interp, objc, objv);
+    case ENC_CONVERTFROM:
+	return EncodingConvertfromObjCmd(dummy, interp, objc, objv);
+    case ENC_DIRS:
+	return EncodingDirsObjCmd(dummy, interp, objc, objv);
+    case ENC_NAMES:
+	return EncodingNamesObjCmd(dummy, interp, objc, objv);
+    case ENC_SYSTEM:
+	return EncodingSystemObjCmd(dummy, interp, objc, objv);
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TclInitEncodingCmd --
+ *
+ *	This function creates the 'encoding' ensemble.
+ *
+ * Results:
+ *	Returns the Tcl_Command so created.
+ *
+ * Side effects:
+ *	The ensemble is initialized.
+ *
+ * This command is hidden in a safe interpreter.
+ */
+
+Tcl_Command
+TclInitEncodingCmd(
+    Tcl_Interp* interp)		/* Tcl interpreter */
+{
+    static const EnsembleImplMap encodingImplMap[] = {
+	{"convertfrom", EncodingConvertfromObjCmd, TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"convertto",   EncodingConverttoObjCmd,   TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"dirs",        EncodingDirsObjCmd,        TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{"names",       EncodingNamesObjCmd,       TclCompileBasic0ArgCmd,    NULL, NULL, 0},
+	{"system",      EncodingSystemObjCmd,      TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{NULL,          NULL,                      NULL,                      NULL, NULL, 0}
+    };
+
+    return TclMakeEnsemble(interp, "encoding", encodingImplMap);
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TclMakeEncodingCommandSafe --
+ *
+ *	This function hides the unsafe 'dirs' and 'system' subcommands of
+ *	the "encoding" Tcl command ensemble. It must be called only from
+ *	TclHideUnsafeCommands.
+ *
+ * Results:
+ *	A standard Tcl result
+ *
+ * Side effects:
+ *	Adds commands to the table of hidden commands.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+int
+TclMakeEncodingCommandSafe(
+    Tcl_Interp* interp)		/* Tcl interpreter */
+{
+    static const struct {
+	const char *cmdName;
+	int unsafe;
+    } unsafeInfo[] = {
+	{"convertfrom", 0},
+	{"convertto",   0},
+	{"dirs",        1},
+	{"names",       0},
+	{"system",      0},
+	{NULL,          0}
+    };
+
+    int i;
+    Tcl_DString oldBuf, newBuf;
+
+    Tcl_DStringInit(&oldBuf);
+    TclDStringAppendLiteral(&oldBuf, "::tcl::encoding::");
+    Tcl_DStringInit(&newBuf);
+    TclDStringAppendLiteral(&newBuf, "tcl:encoding:");
+    for (i=0 ; unsafeInfo[i].cmdName != NULL ; i++) {
+	if (unsafeInfo[i].unsafe) {
+	    const char *oldName, *newName;
+
+	    Tcl_DStringSetLength(&oldBuf, 17);
+	    oldName = Tcl_DStringAppend(&oldBuf, unsafeInfo[i].cmdName, -1);
+	    Tcl_DStringSetLength(&newBuf, 13);
+	    newName = Tcl_DStringAppend(&newBuf, unsafeInfo[i].cmdName, -1);
+	    if (TclRenameCommand(interp, oldName, "___tmp") != TCL_OK
+		    || Tcl_HideCommand(interp, "___tmp", newName) != TCL_OK) {
+		Tcl_Panic("problem making 'encoding %s' safe: %s",
+			unsafeInfo[i].cmdName,
+			Tcl_GetString(Tcl_GetObjResult(interp)));
+	    }
+	    Tcl_CreateObjCommand(interp, oldName, BadEncodingSubcommand,
+		    (ClientData) unsafeInfo[i].cmdName, NULL);
+	}
+    }
+    Tcl_DStringFree(&oldBuf);
+    Tcl_DStringFree(&newBuf);
+
+    /*
+     * Ugh. The [encoding] command is now actually safe, but it is assumed by
+     * scripts that it is not, which messes up security policies.
+     */
+
+    if (Tcl_HideCommand(interp, "encoding", "encoding") != TCL_OK) {
+	Tcl_Panic("problem making 'encoding' safe: %s",
+		Tcl_GetString(Tcl_GetObjResult(interp)));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BadEncodingSubcommand --
+ *
+ *	Command used to act as a backstop implementation when subcommands of
+ *	"encoding" are unsafe (the real implementations of the subcommands are
+ *	hidden). The clientData is always the full official subcommand name.
+ *
+ * Results:
+ *	A standard Tcl result (always a TCL_ERROR).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BadEncodingSubcommand(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    const char *subcommandName = (const char *) clientData;
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "not allowed to invoke subcommand %s of encoding", subcommandName));
+    Tcl_SetErrorCode(interp, "TCL", "SAFE", "SUBCOMMAND", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EncodingConvertfromObjCmd --
+ *
+ *	This command converts a byte array in an external encoding into a
+ *	Tcl string
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+EncodingConvertfromObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *data;		/* Byte array to convert */
+    Tcl_DString ds;		/* Buffer to hold the string */
+    Tcl_Encoding encoding;	/* Encoding to use */
+    int length;			/* Length of the byte array being converted */
+    const char *bytesPtr;	/* Pointer to the first byte of the array */
+
+    if (objc == 2) {
+	encoding = Tcl_GetEncoding(interp, NULL);
+	data = objv[1];
+    } else if (objc == 3) {
+	if (Tcl_GetEncodingFromObj(interp, objv[1], &encoding) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	data = objv[2];
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?encoding? data");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert the string into a byte array in 'ds'
+     */
+    bytesPtr = (char *) Tcl_GetByteArrayFromObj(data, &length);
+    Tcl_ExternalToUtfDString(encoding, bytesPtr, length, &ds);
+
+    /*
+     * Note that we cannot use Tcl_DStringResult here because it will
+     * truncate the string at the first null byte.
+     */
+
+    Tcl_SetObjResult(interp, TclDStringToObj(&ds));
+
+    /*
+     * We're done with the encoding
+     */
+
+    Tcl_FreeEncoding(encoding);
+    return TCL_OK;
+
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EncodingConverttoObjCmd --
+ *
+ *	This command converts a Tcl string into a byte array that
+ *	encodes the string according to some encoding.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+EncodingConverttoObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *data;		/* String to convert */
+    Tcl_DString ds;		/* Buffer to hold the byte array */
+    Tcl_Encoding encoding;	/* Encoding to use */
+    int length;			/* Length of the string being converted */
+    const char *stringPtr;	/* Pointer to the first byte of the string */
+
+    /* TODO - ADJUST OBJ INDICES WHEN ENSEMBLIFYING THIS */
+
+    if (objc == 2) {
+	encoding = Tcl_GetEncoding(interp, NULL);
+	data = objv[1];
+    } else if (objc == 3) {
+	if (Tcl_GetEncodingFromObj(interp, objv[1], &encoding) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	data = objv[2];
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?encoding? data");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert the string to a byte array in 'ds'
+     */
+
+    stringPtr = TclGetStringFromObj(data, &length);
+    Tcl_UtfToExternalDString(encoding, stringPtr, length, &ds);
+    Tcl_SetObjResult(interp,
+		     Tcl_NewByteArrayObj((unsigned char*) Tcl_DStringValue(&ds),
+					 Tcl_DStringLength(&ds)));
+    Tcl_DStringFree(&ds);
+
+    /*
+     * We're done with the encoding
+     */
+
+    Tcl_FreeEncoding(encoding);
+    return TCL_OK;
+
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EncodingDirsObjCmd --
+ *
+ *	This command manipulates the encoding search path.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Can set the encoding search path.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+EncodingDirsObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *dirListObj;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?dirList?");
+	return TCL_ERROR;
+    }
+    if (objc == 1) {
+	Tcl_SetObjResult(interp, Tcl_GetEncodingSearchPath());
+	return TCL_OK;
+    }
+
+    dirListObj = objv[1];
+    if (Tcl_SetEncodingSearchPath(dirListObj) == TCL_ERROR) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"expected directory list but got \"%s\"",
+		TclGetString(dirListObj)));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "ENCODING", "BADPATH",
+		NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, dirListObj);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * EncodingNamesObjCmd --
+ *
+ *	This command returns a list of the available encoding names
+ *
+ * Results:
+ *	Returns a standard Tcl result
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+int
+EncodingNamesObjCmd(ClientData dummy,       /* Unused */
+		    Tcl_Interp* interp,	    /* Tcl interpreter */
+		    int objc,		    /* Number of command line args */
+		    Tcl_Obj* const objv[])  /* Vector of command line args */
+{
+    if (objc > 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_GetEncodingNames(interp);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * EncodingSystemObjCmd --
+ *
+ *	This command retrieves or changes the system encoding
+ *
+ * Results:
+ *	Returns a standard Tcl result
+ *
+ * Side effects:
+ *	May change the system encoding.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+int
+EncodingSystemObjCmd(ClientData dummy,      /* Unused */
+		     Tcl_Interp* interp,    /* Tcl interpreter */
+		     int objc,		    /* Number of command line args */
+		     Tcl_Obj* const objv[]) /* Vector of command line args */
+{
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?encoding?");
+	return TCL_ERROR;
+    }
+    if (objc == 1) {
+	Tcl_SetObjResult(interp,
+			 Tcl_NewStringObj(Tcl_GetEncodingName(NULL), -1));
+    } else {
+	return Tcl_SetSystemEncoding(interp, TclGetString(objv[1]));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ErrorObjCmd --
+ *
+ *	This procedure is invoked to process the "error" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ErrorObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *options, *optName;
+
+    if ((objc < 2) || (objc > 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "message ?errorInfo? ?errorCode?");
+	return TCL_ERROR;
+    }
+
+    TclNewLiteralStringObj(options, "-code error -level 0");
+
+    if (objc >= 3) {		/* Process the optional info argument */
+	TclNewLiteralStringObj(optName, "-errorinfo");
+	Tcl_ListObjAppendElement(NULL, options, optName);
+	Tcl_ListObjAppendElement(NULL, options, objv[2]);
+    }
+
+    if (objc >= 4) {		/* Process the optional code argument */
+	TclNewLiteralStringObj(optName, "-errorcode");
+	Tcl_ListObjAppendElement(NULL, options, optName);
+	Tcl_ListObjAppendElement(NULL, options, objv[3]);
+    }
+
+    Tcl_SetObjResult(interp, objv[1]);
+    return Tcl_SetReturnOptions(interp, options);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EvalObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "eval" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+EvalCmdErrMsg(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    if (result == TCL_ERROR) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"eval\" body line %d)", Tcl_GetErrorLine(interp)));
+    }
+    return result;
+}
+
+int
+Tcl_EvalObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNREvalObjCmd, dummy, objc, objv);
+}
+
+int
+TclNREvalObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Tcl_Obj *objPtr;
+    Interp *iPtr = (Interp *) interp;
+    CmdFrame *invoker = NULL;
+    int word = 0;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	/*
+	 * TIP #280. Make argument location available to eval'd script.
+	 */
+
+	invoker = iPtr->cmdFramePtr;
+	word = 1;
+	objPtr = objv[1];
+	TclArgumentGet(interp, objPtr, &invoker, &word);
+    } else {
+	/*
+	 * More than one argument: concatenate them together with spaces
+	 * between, then evaluate the result. Tcl_EvalObjEx will delete the
+	 * object when it decrements its refcount after eval'ing it.
+	 *
+	 * TIP #280. Make invoking context available to eval'd script, done
+	 * with the default values.
+	 */
+
+	objPtr = Tcl_ConcatObj(objc-1, objv+1);
+    }
+    TclNRAddCallback(interp, EvalCmdErrMsg, NULL, NULL, NULL, NULL);
+    return TclNREvalObjEx(interp, objPtr, 0, invoker, word);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ExitObjCmd --
+ *
+ *	This procedure is invoked to process the "exit" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ExitObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int value;
+
+    if ((objc != 1) && (objc != 2)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?returnCode?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 1) {
+	value = 0;
+    } else if (Tcl_GetIntFromObj(interp, objv[1], &value) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    Tcl_Exit(value);
+    /*NOTREACHED*/
+    return TCL_OK;		/* Better not ever reach this! */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ExprObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "expr" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is called in two
+ *	circumstances: 1) to execute expr commands that are too complicated or
+ *	too unsafe to try compiling directly into an inline sequence of
+ *	instructions, and 2) to execute commands where the command name is
+ *	computed at runtime and is "expr" or the name to which "expr" was
+ *	renamed (e.g., "set z expr; $z 2+3")
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ExprObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRExprObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRExprObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *resultPtr, *objPtr;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    TclNewObj(resultPtr);
+    Tcl_IncrRefCount(resultPtr);
+    if (objc == 2) {
+	objPtr = objv[1];
+	TclNRAddCallback(interp, ExprCallback, resultPtr, NULL, NULL, NULL);
+    } else {
+	objPtr = Tcl_ConcatObj(objc-1, objv+1);
+	TclNRAddCallback(interp, ExprCallback, resultPtr, objPtr, NULL, NULL);
+    }
+
+    return Tcl_NRExprObj(interp, objPtr, resultPtr);
+}
+
+static int
+ExprCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj *resultPtr = data[0];
+    Tcl_Obj *objPtr = data[1];
+
+    if (objPtr != NULL) {
+	Tcl_DecrRefCount(objPtr);
+    }
+
+    if (result == TCL_OK) {
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+    Tcl_DecrRefCount(resultPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitFileCmd --
+ *
+ *	This function builds the "file" Tcl command ensemble. See the user
+ *	documentation for details on what that ensemble does.
+ *
+ *	PLEASE NOTE THAT THIS FAILS WITH FILENAMES AND PATHS WITH EMBEDDED
+ *	NULLS. With the object-based Tcl_FS APIs, the above NOTE may no longer
+ *	be true. In any case this assertion should be tested.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitFileCmd(
+    Tcl_Interp *interp)
+{
+    /*
+     * Note that most subcommands are unsafe because either they manipulate
+     * the native filesystem or because they reveal information about the
+     * native filesystem.
+     */
+
+    static const EnsembleImplMap initMap[] = {
+	{"atime",	FileAttrAccessTimeCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"attributes",	TclFileAttrsCmd,	NULL, NULL, NULL, 0},
+	{"channels",	TclChannelNamesCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{"copy",	TclFileCopyCmd,		NULL, NULL, NULL, 0},
+	{"delete",	TclFileDeleteCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+	{"dirname",	PathDirNameCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"executable",	FileAttrIsExecutableCmd, TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"exists",	FileAttrIsExistingCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"extension",	PathExtensionCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"isdirectory",	FileAttrIsDirectoryCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"isfile",	FileAttrIsFileCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"join",	PathJoinCmd,		TclCompileBasicMin1ArgCmd, NULL, NULL, 0},
+	{"link",	TclFileLinkCmd,		TclCompileBasic1To3ArgCmd, NULL, NULL, 0},
+	{"lstat",	FileAttrLinkStatCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"mtime",	FileAttrModifyTimeCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"mkdir",	TclFileMakeDirsCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+	{"nativename",	PathNativeNameCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"normalize",	PathNormalizeCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"owned",	FileAttrIsOwnedCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"pathtype",	PathTypeCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"readable",	FileAttrIsReadableCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"readlink",	TclFileReadLinkCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"rename",	TclFileRenameCmd,	NULL, NULL, NULL, 0},
+	{"rootname",	PathRootNameCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"separator",	FilesystemSeparatorCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{"size",	FileAttrSizeCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"split",	PathSplitCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"stat",	FileAttrStatCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"system",	PathFilesystemCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{"tail",	PathTailCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"tempfile",	TclFileTemporaryCmd,	TclCompileBasic0To2ArgCmd, NULL, NULL, 0},
+	{"type",	FileAttrTypeCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"volumes",	FilesystemVolumesCmd,	TclCompileBasic0ArgCmd, NULL, NULL, 0},
+	{"writable",	FileAttrIsWritableCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{NULL, NULL, NULL, NULL, NULL, 0}
+    };
+    return TclMakeEnsemble(interp, "file", initMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclMakeFileCommandSafe --
+ *
+ *	This function hides the unsafe subcommands of the "file" Tcl command
+ *	ensemble. It must only be called from TclHideUnsafeCommands.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Adds commands to the table of hidden commands.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclMakeFileCommandSafe(
+    Tcl_Interp *interp)
+{
+    static const struct {
+	const char *cmdName;
+	int unsafe;
+    } unsafeInfo[] = {
+	{"atime",	 1},
+	{"attributes",	 1},
+	{"channels",	 0},
+	{"copy",	 1},
+	{"delete",	 1},
+	{"dirname",	 1},
+	{"executable",	 1},
+	{"exists",	 1},
+	{"extension",	 1},
+	{"isdirectory",	 1},
+	{"isfile",	 1},
+	{"join",	 0},
+	{"link",	 1},
+	{"lstat",	 1},
+	{"mtime",	 1},
+	{"mkdir",	 1},
+	{"nativename",	 1},
+	{"normalize",	 1},
+	{"owned",	 1},
+	{"pathtype",	 0},
+	{"readable",	 1},
+	{"readlink",	 1},
+	{"rename",	 1},
+	{"rootname",	 1},
+	{"separator",	 0},
+	{"size",	 1},
+	{"split",	 0},
+	{"stat",	 1},
+	{"system",	 0},
+	{"tail",	 1},
+	{"tempfile",	 1},
+	{"type",	 1},
+	{"volumes",	 1},
+	{"writable",	 1},
+	{NULL, 0}
+    };
+    int i;
+    Tcl_DString oldBuf, newBuf;
+
+    Tcl_DStringInit(&oldBuf);
+    TclDStringAppendLiteral(&oldBuf, "::tcl::file::");
+    Tcl_DStringInit(&newBuf);
+    TclDStringAppendLiteral(&newBuf, "tcl:file:");
+    for (i=0 ; unsafeInfo[i].cmdName != NULL ; i++) {
+	if (unsafeInfo[i].unsafe) {
+	    const char *oldName, *newName;
+
+	    Tcl_DStringSetLength(&oldBuf, 13);
+	    oldName = Tcl_DStringAppend(&oldBuf, unsafeInfo[i].cmdName, -1);
+	    Tcl_DStringSetLength(&newBuf, 9);
+	    newName = Tcl_DStringAppend(&newBuf, unsafeInfo[i].cmdName, -1);
+	    if (TclRenameCommand(interp, oldName, "___tmp") != TCL_OK
+		    || Tcl_HideCommand(interp, "___tmp", newName) != TCL_OK) {
+		Tcl_Panic("problem making 'file %s' safe: %s",
+			unsafeInfo[i].cmdName,
+			Tcl_GetString(Tcl_GetObjResult(interp)));
+	    }
+	    Tcl_CreateObjCommand(interp, oldName, BadFileSubcommand,
+		    (ClientData) unsafeInfo[i].cmdName, NULL);
+	}
+    }
+    Tcl_DStringFree(&oldBuf);
+    Tcl_DStringFree(&newBuf);
+
+    /*
+     * Ugh. The [file] command is now actually safe, but it is assumed by
+     * scripts that it is not, which messes up security policies. [Bug
+     * 3211758]
+     */
+
+    if (Tcl_HideCommand(interp, "file", "file") != TCL_OK) {
+	Tcl_Panic("problem making 'file' safe: %s",
+		Tcl_GetString(Tcl_GetObjResult(interp)));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BadFileSubcommand --
+ *
+ *	Command used to act as a backstop implementation when subcommands of
+ *	"file" are unsafe (the real implementations of the subcommands are
+ *	hidden). The clientData is always the full official subcommand name.
+ *
+ * Results:
+ *	A standard Tcl result (always a TCL_ERROR).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+BadFileSubcommand(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    const char *subcommandName = (const char *) clientData;
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "not allowed to invoke subcommand %s of file", subcommandName));
+    Tcl_SetErrorCode(interp, "TCL", "SAFE", "SUBCOMMAND", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrAccessTimeCmd --
+ *
+ *	This function is invoked to process the "file atime" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May update the access time on the file, if requested by the user.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrAccessTimeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+    struct utimbuf tval;
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?time?");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+#if defined(_WIN32)
+    /* We use a value of 0 to indicate the access time not available */
+    if (buf.st_atime == 0) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                             "could not get access time for file \"%s\"",
+                             TclGetString(objv[1])));
+        return TCL_ERROR;
+    }
+#endif
+
+    if (objc == 3) {
+	/*
+	 * Need separate variable for reading longs from an object on 64-bit
+	 * platforms. [Bug 698146]
+	 */
+
+	long newTime;
+
+	if (TclGetLongFromObj(interp, objv[2], &newTime) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	tval.actime = newTime;
+	tval.modtime = buf.st_mtime;
+
+	if (Tcl_FSUtime(objv[1], &tval) != 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "could not set access time for file \"%s\": %s",
+		    TclGetString(objv[1]), Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Do another stat to ensure that the we return the new recognized
+	 * atime - hopefully the same as the one we sent in. However, fs's
+	 * like FAT don't even know what atime is.
+	 */
+
+	if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewLongObj((long) buf.st_atime));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrModifyTimeCmd --
+ *
+ *	This function is invoked to process the "file mtime" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May update the modification time on the file, if requested by the
+ *	user.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrModifyTimeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+    struct utimbuf tval;
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?time?");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+#if defined(_WIN32)
+    /* We use a value of 0 to indicate the modification time not available */
+    if (buf.st_mtime == 0) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                             "could not get modification time for file \"%s\"",
+                             TclGetString(objv[1])));
+        return TCL_ERROR;
+    }
+#endif
+    if (objc == 3) {
+	/*
+	 * Need separate variable for reading longs from an object on 64-bit
+	 * platforms. [Bug 698146]
+	 */
+
+	long newTime;
+
+	if (TclGetLongFromObj(interp, objv[2], &newTime) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	tval.actime = buf.st_atime;
+	tval.modtime = newTime;
+
+	if (Tcl_FSUtime(objv[1], &tval) != 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "could not set modification time for file \"%s\": %s",
+		    TclGetString(objv[1]), Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Do another stat to ensure that the we return the new recognized
+	 * mtime - hopefully the same as the one we sent in.
+	 */
+
+	if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewLongObj((long) buf.st_mtime));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrLinkStatCmd --
+ *
+ *	This function is invoked to process the "file lstat" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Writes to an array named by the user.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrLinkStatCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name varName");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSLstat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return StoreStatData(interp, objv[2], &buf);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrStatCmd --
+ *
+ *	This function is invoked to process the "file stat" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Writes to an array named by the user.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrStatCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name varName");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return StoreStatData(interp, objv[2], &buf);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrTypeCmd --
+ *
+ *	This function is invoked to process the "file type" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrTypeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSLstat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    GetTypeFromMode((unsigned short) buf.st_mode), -1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrSizeCmd --
+ *
+ *	This function is invoked to process the "file size" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrSizeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(interp, objv[1], Tcl_FSStat, &buf) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt) buf.st_size));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsDirectoryCmd --
+ *
+ *	This function is invoked to process the "file isdirectory" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsDirectoryCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+    int value = 0;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(NULL, objv[1], Tcl_FSStat, &buf) == TCL_OK) {
+	value = S_ISDIR(buf.st_mode);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(value));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsExecutableCmd --
+ *
+ *	This function is invoked to process the "file executable" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsExecutableCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    return CheckAccess(interp, objv[1], X_OK);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsExistingCmd --
+ *
+ *	This function is invoked to process the "file exists" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsExistingCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    return CheckAccess(interp, objv[1], F_OK);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsFileCmd --
+ *
+ *	This function is invoked to process the "file isfile" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsFileCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_StatBuf buf;
+    int value = 0;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    if (GetStatBuf(NULL, objv[1], Tcl_FSStat, &buf) == TCL_OK) {
+	value = S_ISREG(buf.st_mode);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(value));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsOwnedCmd --
+ *
+ *	This function is invoked to process the "file owned" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsOwnedCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+#ifdef __CYGWIN__
+#define geteuid() (short)(geteuid)()
+#endif
+#if !defined(_WIN32)
+    Tcl_StatBuf buf;
+#endif
+    int value = 0;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+#if defined(_WIN32)
+    value = TclWinFileOwned(objv[1]);
+#else
+    if (GetStatBuf(NULL, objv[1], Tcl_FSStat, &buf) == TCL_OK) {
+	value = (geteuid() == buf.st_uid);
+    }
+#endif
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(value));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsReadableCmd --
+ *
+ *	This function is invoked to process the "file readable" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsReadableCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    return CheckAccess(interp, objv[1], R_OK);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileAttrIsWritableCmd --
+ *
+ *	This function is invoked to process the "file writable" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FileAttrIsWritableCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    return CheckAccess(interp, objv[1], W_OK);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathDirNameCmd --
+ *
+ *	This function is invoked to process the "file dirname" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathDirNameCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *dirPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    dirPtr = TclPathPart(interp, objv[1], TCL_PATH_DIRNAME);
+    if (dirPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, dirPtr);
+    Tcl_DecrRefCount(dirPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathExtensionCmd --
+ *
+ *	This function is invoked to process the "file extension" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathExtensionCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *dirPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    dirPtr = TclPathPart(interp, objv[1], TCL_PATH_EXTENSION);
+    if (dirPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, dirPtr);
+    Tcl_DecrRefCount(dirPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathRootNameCmd --
+ *
+ *	This function is invoked to process the "file root" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathRootNameCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *dirPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    dirPtr = TclPathPart(interp, objv[1], TCL_PATH_ROOT);
+    if (dirPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, dirPtr);
+    Tcl_DecrRefCount(dirPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathTailCmd --
+ *
+ *	This function is invoked to process the "file tail" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathTailCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *dirPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    dirPtr = TclPathPart(interp, objv[1], TCL_PATH_TAIL);
+    if (dirPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, dirPtr);
+    Tcl_DecrRefCount(dirPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathFilesystemCmd --
+ *
+ *	This function is invoked to process the "file system" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathFilesystemCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *fsInfo;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    fsInfo = Tcl_FSFileSystemInfo(objv[1]);
+    if (fsInfo == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("unrecognised path", -1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "FILESYSTEM",
+		Tcl_GetString(objv[1]), NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, fsInfo);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathJoinCmd --
+ *
+ *	This function is invoked to process the "file join" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathJoinCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?name ...?");
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, TclJoinPath(objc - 1, objv + 1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathNativeNameCmd --
+ *
+ *	This function is invoked to process the "file nativename" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathNativeNameCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_DString ds;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    if (Tcl_TranslateFileName(interp, TclGetString(objv[1]), &ds) == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, TclDStringToObj(&ds));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathNormalizeCmd --
+ *
+ *	This function is invoked to process the "file normalize" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathNormalizeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *fileName;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    fileName = Tcl_FSGetNormalizedPath(interp, objv[1]);
+    if (fileName == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, fileName);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathSplitCmd --
+ *
+ *	This function is invoked to process the "file split" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathSplitCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *res;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    res = Tcl_FSSplitPath(objv[1], NULL);
+    if (res == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"could not read \"%s\": no such file or directory",
+		TclGetString(objv[1])));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PATHSPLIT", "NONESUCH",
+		NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, res);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PathTypeCmd --
+ *
+ *	This function is invoked to process the "file pathtype" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PathTypeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *typeName;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    switch (Tcl_FSGetPathType(objv[1])) {
+    case TCL_PATH_ABSOLUTE:
+	TclNewLiteralStringObj(typeName, "absolute");
+	break;
+    case TCL_PATH_RELATIVE:
+	TclNewLiteralStringObj(typeName, "relative");
+	break;
+    case TCL_PATH_VOLUME_RELATIVE:
+	TclNewLiteralStringObj(typeName, "volumerelative");
+	break;
+    default:
+	/* Should be unreachable */
+	return TCL_OK;
+    }
+    Tcl_SetObjResult(interp, typeName);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FilesystemSeparatorCmd --
+ *
+ *	This function is invoked to process the "file separator" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FilesystemSeparatorCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc < 1 || objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?name?");
+	return TCL_ERROR;
+    }
+    if (objc == 1) {
+	const char *separator = NULL; /* lint */
+
+	switch (tclPlatform) {
+	case TCL_PLATFORM_UNIX:
+	    separator = "/";
+	    break;
+	case TCL_PLATFORM_WINDOWS:
+	    separator = "\\";
+	    break;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(separator, 1));
+    } else {
+	Tcl_Obj *separatorObj = Tcl_FSPathSeparator(objv[1]);
+
+	if (separatorObj == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unrecognised path", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "FILESYSTEM",
+		    Tcl_GetString(objv[1]), NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, separatorObj);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FilesystemVolumesCmd --
+ *
+ *	This function is invoked to process the "file volumes" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FilesystemVolumesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_FSListVolumes());
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * CheckAccess --
+ *
+ *	Utility procedure used by Tcl_FileObjCmd() to query file attributes
+ *	available through the access() system call.
+ *
+ * Results:
+ *	Always returns TCL_OK. Sets interp's result to boolean true or false
+ *	depending on whether the file has the specified attribute.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+CheckAccess(
+    Tcl_Interp *interp,		/* Interp for status return. Must not be
+				 * NULL. */
+    Tcl_Obj *pathPtr,		/* Name of file to check. */
+    int mode)			/* Attribute to check; passed as argument to
+				 * access(). */
+{
+    int value;
+
+    if (Tcl_FSConvertToPathType(interp, pathPtr) != TCL_OK) {
+	value = 0;
+    } else {
+	value = (Tcl_FSAccess(pathPtr, mode) == 0);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(value));
+
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * GetStatBuf --
+ *
+ *	Utility procedure used by Tcl_FileObjCmd() to query file attributes
+ *	available through the stat() or lstat() system call.
+ *
+ * Results:
+ *	The return value is TCL_OK if the specified file exists and can be
+ *	stat'ed, TCL_ERROR otherwise. If TCL_ERROR is returned, an error
+ *	message is left in interp's result. If TCL_OK is returned, *statPtr is
+ *	filled with information about the specified file.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+GetStatBuf(
+    Tcl_Interp *interp,		/* Interp for error return. May be NULL. */
+    Tcl_Obj *pathPtr,		/* Path name to examine. */
+    Tcl_FSStatProc *statProc,	/* Either stat() or lstat() depending on
+				 * desired behavior. */
+    Tcl_StatBuf *statPtr)	/* Filled with info about file obtained by
+				 * calling (*statProc)(). */
+{
+    int status;
+
+    if (Tcl_FSConvertToPathType(interp, pathPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    status = statProc(pathPtr, statPtr);
+
+    if (status < 0) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "could not read \"%s\": %s",
+		    TclGetString(pathPtr), Tcl_PosixError(interp)));
+	}
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StoreStatData --
+ *
+ *	This is a utility procedure that breaks out the fields of a "stat"
+ *	structure and stores them in textual form into the elements of an
+ *	associative array.
+ *
+ * Results:
+ *	Returns a standard Tcl return value. If an error occurs then a message
+ *	is left in interp's result.
+ *
+ * Side effects:
+ *	Elements of the associative array given by "varName" are modified.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StoreStatData(
+    Tcl_Interp *interp,		/* Interpreter for error reports. */
+    Tcl_Obj *varName,		/* Name of associative array variable in which
+				 * to store stat results. */
+    Tcl_StatBuf *statPtr)	/* Pointer to buffer containing stat data to
+				 * store in varName. */
+{
+    Tcl_Obj *field, *value;
+    register unsigned short mode;
+
+    /*
+     * Assume Tcl_ObjSetVar2() does not keep a copy of the field name!
+     *
+     * Might be a better idea to call Tcl_SetVar2Ex() instead, except we want
+     * to have an object (i.e. possibly cached) array variable name but a
+     * string element name, so no API exists. Messy.
+     */
+
+#define STORE_ARY(fieldName, object) \
+    TclNewLiteralStringObj(field, fieldName);				\
+    Tcl_IncrRefCount(field);						\
+    value = (object);							\
+    if (Tcl_ObjSetVar2(interp,varName,field,value,TCL_LEAVE_ERR_MSG)==NULL) { \
+	TclDecrRefCount(field);						\
+	return TCL_ERROR;						\
+    }									\
+    TclDecrRefCount(field);
+
+    /*
+     * Watch out porters; the inode is meant to be an *unsigned* value, so the
+     * cast might fail when there isn't a real arithmetic 'long long' type...
+     */
+
+    STORE_ARY("dev",	Tcl_NewLongObj((long)statPtr->st_dev));
+    STORE_ARY("ino",	Tcl_NewWideIntObj((Tcl_WideInt)statPtr->st_ino));
+    STORE_ARY("nlink",	Tcl_NewLongObj((long)statPtr->st_nlink));
+    STORE_ARY("uid",	Tcl_NewLongObj((long)statPtr->st_uid));
+    STORE_ARY("gid",	Tcl_NewLongObj((long)statPtr->st_gid));
+    STORE_ARY("size",	Tcl_NewWideIntObj((Tcl_WideInt)statPtr->st_size));
+#ifdef HAVE_STRUCT_STAT_ST_BLOCKS
+    STORE_ARY("blocks",	Tcl_NewWideIntObj((Tcl_WideInt)statPtr->st_blocks));
+#endif
+#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
+    STORE_ARY("blksize", Tcl_NewLongObj((long)statPtr->st_blksize));
+#endif
+    STORE_ARY("atime",	Tcl_NewLongObj((long)statPtr->st_atime));
+    STORE_ARY("mtime",	Tcl_NewLongObj((long)statPtr->st_mtime));
+    STORE_ARY("ctime",	Tcl_NewLongObj((long)statPtr->st_ctime));
+    mode = (unsigned short) statPtr->st_mode;
+    STORE_ARY("mode",	Tcl_NewIntObj(mode));
+    STORE_ARY("type",	Tcl_NewStringObj(GetTypeFromMode(mode), -1));
+#undef STORE_ARY
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetTypeFromMode --
+ *
+ *	Given a mode word, returns a string identifying the type of a file.
+ *
+ * Results:
+ *	A static text string giving the file type from mode.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *
+GetTypeFromMode(
+    int mode)
+{
+    if (S_ISREG(mode)) {
+	return "file";
+    } else if (S_ISDIR(mode)) {
+	return "directory";
+    } else if (S_ISCHR(mode)) {
+	return "characterSpecial";
+    } else if (S_ISBLK(mode)) {
+	return "blockSpecial";
+    } else if (S_ISFIFO(mode)) {
+	return "fifo";
+#ifdef S_ISLNK
+    } else if (S_ISLNK(mode)) {
+	return "link";
+#endif
+#ifdef S_ISSOCK
+    } else if (S_ISSOCK(mode)) {
+	return "socket";
+#endif
+    }
+    return "unknown";
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ForObjCmd --
+ *
+ *	This procedure is invoked to process the "for" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "for" or the name to which
+ *	"for" was renamed: e.g.,
+ *	"set z for; $z {set i 0} {$i<100} {incr i} {puts $i}"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ * Notes:
+ *	This command is split into a lot of pieces so that it can avoid doing
+ *	reentrant TEBC calls. This makes things rather hard to follow, but
+ *	here's the plan:
+ *
+ *	NR:	---------------_\
+ *	Direct:	Tcl_ForObjCmd -> TclNRForObjCmd
+ *					|
+ *				ForSetupCallback
+ *					|
+ *	[while] ------------> TclNRForIterCallback <---------.
+ *					|		     |
+ *				 ForCondCallback	     |
+ *					|		     |
+ *				 ForNextCallback ------------|
+ *					|		     |
+ *			       ForPostNextCallback	     |
+ *					|____________________|
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ForObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRForObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRForObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ForIterData *iterPtr;
+
+    if (objc != 5) {
+	Tcl_WrongNumArgs(interp, 1, objv, "start test next command");
+	return TCL_ERROR;
+    }
+
+    TclSmallAllocEx(interp, sizeof(ForIterData), iterPtr);
+    iterPtr->cond = objv[2];
+    iterPtr->body = objv[4];
+    iterPtr->next = objv[3];
+    iterPtr->msg  = "\n    (\"for\" body line %d)";
+    iterPtr->word = 4;
+
+    TclNRAddCallback(interp, ForSetupCallback, iterPtr, NULL, NULL, NULL);
+
+    /*
+     * TIP #280. Make invoking context available to initial script.
+     */
+
+    return TclNREvalObjEx(interp, objv[1], 0, iPtr->cmdFramePtr, 1);
+}
+
+static int
+ForSetupCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ForIterData *iterPtr = data[0];
+
+    if (result != TCL_OK) {
+	if (result == TCL_ERROR) {
+	    Tcl_AddErrorInfo(interp, "\n    (\"for\" initial command)");
+	}
+	TclSmallFreeEx(interp, iterPtr);
+	return result;
+    }
+    TclNRAddCallback(interp, TclNRForIterCallback, iterPtr, NULL, NULL, NULL);
+    return TCL_OK;
+}
+
+int
+TclNRForIterCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ForIterData *iterPtr = data[0];
+    Tcl_Obj *boolObj;
+
+    switch (result) {
+    case TCL_OK:
+    case TCL_CONTINUE:
+	/*
+	 * We need to reset the result before evaluating the expression.
+	 * Otherwise, any error message will be appended to the result of the
+	 * last evaluation.
+	 */
+
+	Tcl_ResetResult(interp);
+	TclNewObj(boolObj);
+	TclNRAddCallback(interp, ForCondCallback, iterPtr, boolObj, NULL,
+		NULL);
+	return Tcl_NRExprObj(interp, iterPtr->cond, boolObj);
+    case TCL_BREAK:
+	result = TCL_OK;
+	Tcl_ResetResult(interp);
+	break;
+    case TCL_ERROR:
+	Tcl_AppendObjToErrorInfo(interp,
+		Tcl_ObjPrintf(iterPtr->msg, Tcl_GetErrorLine(interp)));
+    }
+    TclSmallFreeEx(interp, iterPtr);
+    return result;
+}
+
+static int
+ForCondCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    ForIterData *iterPtr = data[0];
+    Tcl_Obj *boolObj = data[1];
+    int value;
+
+    if (result != TCL_OK) {
+	Tcl_DecrRefCount(boolObj);
+	TclSmallFreeEx(interp, iterPtr);
+	return result;
+    } else if (Tcl_GetBooleanFromObj(interp, boolObj, &value) != TCL_OK) {
+	Tcl_DecrRefCount(boolObj);
+	TclSmallFreeEx(interp, iterPtr);
+	return TCL_ERROR;
+    }
+    Tcl_DecrRefCount(boolObj);
+
+    if (value) {
+	/* TIP #280. */
+	if (iterPtr->next) {
+	    TclNRAddCallback(interp, ForNextCallback, iterPtr, NULL, NULL,
+		    NULL);
+	} else {
+	    TclNRAddCallback(interp, TclNRForIterCallback, iterPtr, NULL,
+		    NULL, NULL);
+	}
+	return TclNREvalObjEx(interp, iterPtr->body, 0, iPtr->cmdFramePtr,
+		iterPtr->word);
+    }
+    TclSmallFreeEx(interp, iterPtr);
+    return result;
+}
+
+static int
+ForNextCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    ForIterData *iterPtr = data[0];
+    Tcl_Obj *next = iterPtr->next;
+
+    if ((result == TCL_OK) || (result == TCL_CONTINUE)) {
+	TclNRAddCallback(interp, ForPostNextCallback, iterPtr, NULL, NULL,
+		NULL);
+
+	/*
+	 * TIP #280. Make invoking context available to next script.
+	 */
+
+	return TclNREvalObjEx(interp, next, 0, iPtr->cmdFramePtr, 3);
+    }
+
+    TclNRAddCallback(interp, TclNRForIterCallback, iterPtr, NULL, NULL, NULL);
+    return result;
+}
+
+static int
+ForPostNextCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ForIterData *iterPtr = data[0];
+
+    if ((result != TCL_BREAK) && (result != TCL_OK)) {
+	if (result == TCL_ERROR) {
+	    Tcl_AddErrorInfo(interp, "\n    (\"for\" loop-end command)");
+	    TclSmallFreeEx(interp, iterPtr);
+	}
+	return result;
+    }
+    TclNRAddCallback(interp, TclNRForIterCallback, iterPtr, NULL, NULL, NULL);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ForeachObjCmd, TclNRForeachCmd, EachloopCmd --
+ *
+ *	This object-based procedure is invoked to process the "foreach" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ForeachObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRForeachCmd, dummy, objc, objv);
+}
+
+int
+TclNRForeachCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    return EachloopCmd(interp, TCL_EACH_KEEP_NONE, objc, objv);
+}
+
+int
+Tcl_LmapObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRLmapCmd, dummy, objc, objv);
+}
+
+int
+TclNRLmapCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    return EachloopCmd(interp, TCL_EACH_COLLECT, objc, objv);
+}
+
+static inline int
+EachloopCmd(
+    Tcl_Interp *interp,		/* Our context for variables and script
+				 * evaluation. */
+    int collect,		/* Select collecting or accumulating mode
+				 * (TCL_EACH_*) */
+    int objc,			/* The arguments being passed in... */
+    Tcl_Obj *const objv[])
+{
+    int numLists = (objc-2) / 2;
+    register struct ForeachState *statePtr;
+    int i, j, result;
+
+    if (objc < 4 || (objc%2 != 0)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"varList list ?varList list ...? command");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Manage numList parallel value lists.
+     * statePtr->argvList[i] is a value list counted by statePtr->argcList[i];
+     * statePtr->varvList[i] is the list of variables associated with the
+     *		value list;
+     * statePtr->varcList[i] is the number of variables associated with the
+     *		value list;
+     * statePtr->index[i] is the current pointer into the value list
+     *		statePtr->argvList[i].
+     *
+     * The setting up of all of these pointers is moderately messy, but allows
+     * the rest of this code to be simple and for us to use a single memory
+     * allocation for better performance.
+     */
+
+    statePtr = TclStackAlloc(interp,
+	    sizeof(struct ForeachState) + 3 * numLists * sizeof(int)
+	    + 2 * numLists * (sizeof(Tcl_Obj **) + sizeof(Tcl_Obj *)));
+    memset(statePtr, 0,
+	    sizeof(struct ForeachState) + 3 * numLists * sizeof(int)
+	    + 2 * numLists * (sizeof(Tcl_Obj **) + sizeof(Tcl_Obj *)));
+    statePtr->varvList = (Tcl_Obj ***) (statePtr + 1);
+    statePtr->argvList = statePtr->varvList + numLists;
+    statePtr->vCopyList = (Tcl_Obj **) (statePtr->argvList + numLists);
+    statePtr->aCopyList = statePtr->vCopyList + numLists;
+    statePtr->index = (int *) (statePtr->aCopyList + numLists);
+    statePtr->varcList = statePtr->index + numLists;
+    statePtr->argcList = statePtr->varcList + numLists;
+
+    statePtr->numLists = numLists;
+    statePtr->bodyPtr = objv[objc - 1];
+    statePtr->bodyIdx = objc - 1;
+
+    if (collect == TCL_EACH_COLLECT) {
+	statePtr->resultList = Tcl_NewListObj(0, NULL);
+    } else {
+	statePtr->resultList = NULL;
+    }
+
+    /*
+     * Break up the value lists and variable lists into elements.
+     */
+
+    for (i=0 ; i<numLists ; i++) {
+	statePtr->vCopyList[i] = TclListObjCopy(interp, objv[1+i*2]);
+	if (statePtr->vCopyList[i] == NULL) {
+	    result = TCL_ERROR;
+	    goto done;
+	}
+	TclListObjGetElements(NULL, statePtr->vCopyList[i],
+		&statePtr->varcList[i], &statePtr->varvList[i]);
+	if (statePtr->varcList[i] < 1) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "%s varlist is empty",
+		    (statePtr->resultList != NULL ? "lmap" : "foreach")));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION",
+		    (statePtr->resultList != NULL ? "LMAP" : "FOREACH"),
+		    "NEEDVARS", NULL);
+	    result = TCL_ERROR;
+	    goto done;
+	}
+
+	statePtr->aCopyList[i] = TclListObjCopy(interp, objv[2+i*2]);
+	if (statePtr->aCopyList[i] == NULL) {
+	    result = TCL_ERROR;
+	    goto done;
+	}
+	TclListObjGetElements(NULL, statePtr->aCopyList[i],
+		&statePtr->argcList[i], &statePtr->argvList[i]);
+
+	j = statePtr->argcList[i] / statePtr->varcList[i];
+	if ((statePtr->argcList[i] % statePtr->varcList[i]) != 0) {
+	    j++;
+	}
+	if (j > statePtr->maxj) {
+	    statePtr->maxj = j;
+	}
+    }
+
+    /*
+     * If there is any work to do, assign the variables and set things going
+     * non-recursively.
+     */
+
+    if (statePtr->maxj > 0) {
+	result = ForeachAssignments(interp, statePtr);
+	if (result == TCL_ERROR) {
+	    goto done;
+	}
+
+	TclNRAddCallback(interp, ForeachLoopStep, statePtr, NULL, NULL, NULL);
+	return TclNREvalObjEx(interp, objv[objc-1], 0,
+		((Interp *) interp)->cmdFramePtr, objc-1);
+    }
+
+    /*
+     * This cleanup stage is only used when an error occurs during setup or if
+     * there is no work to do.
+     */
+
+    result = TCL_OK;
+  done:
+    ForeachCleanup(interp, statePtr);
+    return result;
+}
+
+/*
+ * Post-body processing handler.
+ */
+
+static int
+ForeachLoopStep(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    register struct ForeachState *statePtr = data[0];
+
+    /*
+     * Process the result code from this run of the [foreach] body. Note that
+     * this switch uses fallthroughs in several places. Maintainer aware!
+     */
+
+    switch (result) {
+    case TCL_CONTINUE:
+	result = TCL_OK;
+	break;
+    case TCL_OK:
+	if (statePtr->resultList != NULL) {
+	    Tcl_ListObjAppendElement(interp, statePtr->resultList,
+		    Tcl_GetObjResult(interp));
+	}
+	break;
+    case TCL_BREAK:
+	result = TCL_OK;
+	goto finish;
+    case TCL_ERROR:
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"%s\" body line %d)",
+		(statePtr->resultList != NULL ? "lmap" : "foreach"),
+		Tcl_GetErrorLine(interp)));
+    default:
+	goto done;
+    }
+
+    /*
+     * Test if there is work still to be done. If so, do the next round of
+     * variable assignments, reschedule ourselves and run the body again.
+     */
+
+    if (statePtr->maxj > ++statePtr->j) {
+	result = ForeachAssignments(interp, statePtr);
+	if (result == TCL_ERROR) {
+	    goto done;
+	}
+
+	TclNRAddCallback(interp, ForeachLoopStep, statePtr, NULL, NULL, NULL);
+	return TclNREvalObjEx(interp, statePtr->bodyPtr, 0,
+		((Interp *) interp)->cmdFramePtr, statePtr->bodyIdx);
+    }
+
+    /*
+     * We're done. Tidy up our work space and finish off.
+     */
+
+  finish:
+    if (statePtr->resultList == NULL) {
+	Tcl_ResetResult(interp);
+    } else {
+	Tcl_SetObjResult(interp, statePtr->resultList);
+	statePtr->resultList = NULL;	/* Don't clean it up */
+    }
+
+  done:
+    ForeachCleanup(interp, statePtr);
+    return result;
+}
+
+/*
+ * Factored out code to do the assignments in [foreach].
+ */
+
+static inline int
+ForeachAssignments(
+    Tcl_Interp *interp,
+    struct ForeachState *statePtr)
+{
+    int i, v, k;
+    Tcl_Obj *valuePtr, *varValuePtr;
+
+    for (i=0 ; i<statePtr->numLists ; i++) {
+	for (v=0 ; v<statePtr->varcList[i] ; v++) {
+	    k = statePtr->index[i]++;
+
+	    if (k < statePtr->argcList[i]) {
+		valuePtr = statePtr->argvList[i][k];
+	    } else {
+		TclNewObj(valuePtr);	/* Empty string */
+	    }
+
+	    varValuePtr = Tcl_ObjSetVar2(interp, statePtr->varvList[i][v],
+		    NULL, valuePtr, TCL_LEAVE_ERR_MSG);
+
+	    if (varValuePtr == NULL) {
+		Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			"\n    (setting %s loop variable \"%s\")",
+			(statePtr->resultList != NULL ? "lmap" : "foreach"),
+			TclGetString(statePtr->varvList[i][v])));
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ * Factored out code for cleaning up the state of the foreach.
+ */
+
+static inline void
+ForeachCleanup(
+    Tcl_Interp *interp,
+    struct ForeachState *statePtr)
+{
+    int i;
+
+    for (i=0 ; i<statePtr->numLists ; i++) {
+	if (statePtr->vCopyList[i]) {
+	    TclDecrRefCount(statePtr->vCopyList[i]);
+	}
+	if (statePtr->aCopyList[i]) {
+	    TclDecrRefCount(statePtr->aCopyList[i]);
+	}
+    }
+    if (statePtr->resultList != NULL) {
+	TclDecrRefCount(statePtr->resultList);
+    }
+    TclStackFree(interp, statePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FormatObjCmd --
+ *
+ *	This procedure is invoked to process the "format" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_FormatObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *resultPtr;		/* Where result is stored finally. */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "formatString ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    resultPtr = Tcl_Format(interp, TclGetString(objv[1]), objc-2, objv+2);
+    if (resultPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCmdIL.c b/generic/tclCmdIL.c
new file mode 100644
index 0000000..47076ec
--- /dev/null
+++ b/generic/tclCmdIL.c
@@ -0,0 +1,4565 @@
+/*
+ * tclCmdIL.c --
+ *
+ *	This file contains the top-level command routines for most of the Tcl
+ *	built-in commands whose names begin with the letters I through L. It
+ *	contains only commands in the generic core (i.e., those that don't
+ *	depend much upon UNIX facilities).
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1993-1997 Lucent Technologies.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ * Copyright (c) 2005 Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclRegexp.h"
+
+/*
+ * During execution of the "lsort" command, structures of the following type
+ * are used to arrange the objects being sorted into a collection of linked
+ * lists.
+ */
+
+typedef struct SortElement {
+    union {			/* The value that we sorting by. */
+	const char *strValuePtr;
+	Tcl_WideInt wideValue;
+	double doubleValue;
+	Tcl_Obj *objValuePtr;
+    } collationKey;
+    union {			/* Object being sorted, or its index. */
+	Tcl_Obj *objPtr;
+	int index;
+    } payload;
+    struct SortElement *nextPtr;/* Next element in the list, or NULL for end
+				 * of list. */
+} SortElement;
+
+/*
+ * These function pointer types are used with the "lsearch" and "lsort"
+ * commands to facilitate the "-nocase" option.
+ */
+
+typedef int (*SortStrCmpFn_t) (const char *, const char *);
+typedef int (*SortMemCmpFn_t) (const void *, const void *, size_t);
+
+/*
+ * The "lsort" command needs to pass certain information down to the function
+ * that compares two list elements, and the comparison function needs to pass
+ * success or failure information back up to the top-level "lsort" command.
+ * The following structure is used to pass this information.
+ */
+
+typedef struct SortInfo {
+    int isIncreasing;		/* Nonzero means sort in increasing order. */
+    int sortMode;		/* The sort mode. One of SORTMODE_* values
+				 * defined below. */
+    Tcl_Obj *compareCmdPtr;	/* The Tcl comparison command when sortMode is
+				 * SORTMODE_COMMAND. Pre-initialized to hold
+				 * base of command. */
+    int *indexv;		/* If the -index option was specified, this
+				 * holds the indexes contained in the list
+				 * supplied as an argument to that option.
+				 * NULL if no indexes supplied, and points to
+				 * singleIndex field when only one
+				 * supplied. */
+    int indexc;			/* Number of indexes in indexv array. */
+    int singleIndex;		/* Static space for common index case. */
+    int unique;
+    int numElements;
+    Tcl_Interp *interp;		/* The interpreter in which the sort is being
+				 * done. */
+    int resultCode;		/* Completion code for the lsort command. If
+				 * an error occurs during the sort this is
+				 * changed from TCL_OK to TCL_ERROR. */
+} SortInfo;
+
+/*
+ * The "sortMode" field of the SortInfo structure can take on any of the
+ * following values.
+ */
+
+#define SORTMODE_ASCII		0
+#define SORTMODE_INTEGER	1
+#define SORTMODE_REAL		2
+#define SORTMODE_COMMAND	3
+#define SORTMODE_DICTIONARY	4
+#define SORTMODE_ASCII_NC	8
+
+/*
+ * Magic values for the index field of the SortInfo structure. Note that the
+ * index "end-1" will be translated to SORTIDX_END-1, etc.
+ */
+
+#define SORTIDX_NONE	-1	/* Not indexed; use whole value. */
+#define SORTIDX_END	-2	/* Indexed from end. */
+
+/*
+ * Forward declarations for procedures defined in this file:
+ */
+
+static int		DictionaryCompare(const char *left, const char *right);
+static Tcl_NRPostProc	IfConditionCallback;
+static int		InfoArgsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoBodyCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoCmdCountCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoCommandsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoCompleteCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoDefaultCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+/* TIP #348 - New 'info' subcommand 'errorstack' */
+static int		InfoErrorStackCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+/* TIP #280 - New 'info' subcommand 'frame' */
+static int		InfoFrameCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoFunctionsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoHostnameCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoLevelCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoLibraryCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoLoadedCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoNameOfExecutableCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		InfoPatchLevelCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoProcsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoScriptCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoSharedlibCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		InfoTclVersionCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static SortElement *	MergeLists(SortElement *leftPtr, SortElement *rightPtr,
+			    SortInfo *infoPtr);
+static int		SortCompare(SortElement *firstPtr, SortElement *second,
+			    SortInfo *infoPtr);
+static Tcl_Obj *	SelectObjFromSublist(Tcl_Obj *firstPtr,
+			    SortInfo *infoPtr);
+
+/*
+ * Array of values describing how to implement each standard subcommand of the
+ * "info" command.
+ */
+
+static const EnsembleImplMap defaultInfoMap[] = {
+    {"args",		   InfoArgsCmd,		    TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"body",		   InfoBodyCmd,		    TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"cmdcount",	   InfoCmdCountCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"commands",	   InfoCommandsCmd,	    TclCompileInfoCommandsCmd, NULL, NULL, 0},
+    {"complete",	   InfoCompleteCmd,	    TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"coroutine",	   TclInfoCoroutineCmd,     TclCompileInfoCoroutineCmd, NULL, NULL, 0},
+    {"default",		   InfoDefaultCmd,	    TclCompileBasic3ArgCmd, NULL, NULL, 0},
+    {"errorstack",	   InfoErrorStackCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"exists",		   TclInfoExistsCmd,	    TclCompileInfoExistsCmd, NULL, NULL, 0},
+    {"frame",		   InfoFrameCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"functions",	   InfoFunctionsCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"globals",		   TclInfoGlobalsCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"hostname",	   InfoHostnameCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"level",		   InfoLevelCmd,	    TclCompileInfoLevelCmd, NULL, NULL, 0},
+    {"library",		   InfoLibraryCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"loaded",		   InfoLoadedCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"locals",		   TclInfoLocalsCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"nameofexecutable",   InfoNameOfExecutableCmd, TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"patchlevel",	   InfoPatchLevelCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"procs",		   InfoProcsCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"script",		   InfoScriptCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"sharedlibextension", InfoSharedlibCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"tclversion",	   InfoTclVersionCmd,	    TclCompileBasic0ArgCmd, NULL, NULL, 0},
+    {"vars",		   TclInfoVarsCmd,	    TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {NULL, NULL, NULL, NULL, NULL, 0}
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IfObjCmd --
+ *
+ *	This procedure is invoked to process the "if" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "if" or the name to which
+ *	"if" was renamed: e.g., "set z if; $z 1 {puts foo}"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IfObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRIfObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRIfObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *boolObj;
+
+    if (objc <= 1) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"wrong # args: no expression after \"%s\" argument",
+		TclGetString(objv[0])));
+	Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * At this point, objv[1] refers to the main expression to test. The
+     * arguments after the expression must be "then" (optional) and a script
+     * to execute if the expression is true.
+     */
+
+    TclNewObj(boolObj);
+    Tcl_NRAddCallback(interp, IfConditionCallback, INT2PTR(objc),
+	    (ClientData) objv, INT2PTR(1), boolObj);
+    return Tcl_NRExprObj(interp, objv[1], boolObj);
+}
+
+static int
+IfConditionCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    int objc = PTR2INT(data[0]);
+    Tcl_Obj *const *objv = data[1];
+    int i = PTR2INT(data[2]);
+    Tcl_Obj *boolObj = data[3];
+    int value, thenScriptIndex = 0;
+    const char *clause;
+
+    if (result != TCL_OK) {
+	TclDecrRefCount(boolObj);
+	return result;
+    }
+    if (Tcl_GetBooleanFromObj(interp, boolObj, &value) != TCL_OK) {
+	TclDecrRefCount(boolObj);
+	return TCL_ERROR;
+    }
+    TclDecrRefCount(boolObj);
+
+    while (1) {
+	i++;
+	if (i >= objc) {
+	    goto missingScript;
+	}
+	clause = TclGetString(objv[i]);
+	if ((i < objc) && (strcmp(clause, "then") == 0)) {
+	    i++;
+	}
+	if (i >= objc) {
+	    goto missingScript;
+	}
+	if (value) {
+	    thenScriptIndex = i;
+	    value = 0;
+	}
+
+	/*
+	 * The expression evaluated to false. Skip the command, then see if
+	 * there is an "else" or "elseif" clause.
+	 */
+
+	i++;
+	if (i >= objc) {
+	    if (thenScriptIndex) {
+		/*
+		 * TIP #280. Make invoking context available to branch.
+		 */
+
+		return TclNREvalObjEx(interp, objv[thenScriptIndex], 0,
+			iPtr->cmdFramePtr, thenScriptIndex);
+	    }
+	    return TCL_OK;
+	}
+	clause = TclGetString(objv[i]);
+	if ((clause[0] != 'e') || (strcmp(clause, "elseif") != 0)) {
+	    break;
+	}
+	i++;
+
+	/*
+	 * At this point in the loop, objv and objc refer to an expression to
+	 * test, either for the main expression or an expression following an
+	 * "elseif". The arguments after the expression must be "then"
+	 * (optional) and a script to execute if the expression is true.
+	 */
+
+	if (i >= objc) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "wrong # args: no expression after \"%s\" argument",
+		    clause));
+	    Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+	    return TCL_ERROR;
+	}
+	if (!thenScriptIndex) {
+	    TclNewObj(boolObj);
+	    Tcl_NRAddCallback(interp, IfConditionCallback, data[0], data[1],
+		    INT2PTR(i), boolObj);
+	    return Tcl_NRExprObj(interp, objv[i], boolObj);
+	}
+    }
+
+    /*
+     * Couldn't find a "then" or "elseif" clause to execute. Check now for an
+     * "else" clause. We know that there's at least one more argument when we
+     * get here.
+     */
+
+    if (strcmp(clause, "else") == 0) {
+	i++;
+	if (i >= objc) {
+	    goto missingScript;
+	}
+    }
+    if (i < objc - 1) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"wrong # args: extra words after \"else\" clause in \"if\" command",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+	return TCL_ERROR;
+    }
+    if (thenScriptIndex) {
+	/*
+	 * TIP #280. Make invoking context available to branch/else.
+	 */
+
+	return TclNREvalObjEx(interp, objv[thenScriptIndex], 0,
+		iPtr->cmdFramePtr, thenScriptIndex);
+    }
+    return TclNREvalObjEx(interp, objv[i], 0, iPtr->cmdFramePtr, i);
+
+  missingScript:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "wrong # args: no script following \"%s\" argument",
+	    TclGetString(objv[i-1])));
+    Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IncrObjCmd --
+ *
+ *	This procedure is invoked to process the "incr" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "incr" or the name to
+ *	which "incr" was renamed: e.g., "set z incr; $z i -1"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IncrObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *newValuePtr, *incrPtr;
+
+    if ((objc != 2) && (objc != 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName ?increment?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	incrPtr = objv[2];
+    } else {
+	incrPtr = Tcl_NewIntObj(1);
+    }
+    Tcl_IncrRefCount(incrPtr);
+    newValuePtr = TclIncrObjVar2(interp, objv[1], NULL,
+	    incrPtr, TCL_LEAVE_ERR_MSG);
+    Tcl_DecrRefCount(incrPtr);
+
+    if (newValuePtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set the interpreter's object result to refer to the variable's new
+     * value object.
+     */
+
+    Tcl_SetObjResult(interp, newValuePtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitInfoCmd --
+ *
+ *	This function is called to create the "info" Tcl command. See the user
+ *	documentation for details on what it does.
+ *
+ * Results:
+ *	Handle for the info command, or NULL on failure.
+ *
+ * Side effects:
+ *	none
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitInfoCmd(
+    Tcl_Interp *interp)		/* Current interpreter. */
+{
+    return TclMakeEnsemble(interp, "info", defaultInfoMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoArgsCmd --
+ *
+ *	Called to implement the "info args" command that returns the argument
+ *	list for a procedure. Handles the following syntax:
+ *
+ *	    info args procName
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoArgsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    const char *name;
+    Proc *procPtr;
+    CompiledLocal *localPtr;
+    Tcl_Obj *listObjPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "procname");
+	return TCL_ERROR;
+    }
+
+    name = TclGetString(objv[1]);
+    procPtr = TclFindProc(iPtr, name);
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" isn't a procedure", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PROCEDURE", name, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Build a return list containing the arguments.
+     */
+
+    listObjPtr = Tcl_NewListObj(0, NULL);
+    for (localPtr = procPtr->firstLocalPtr;  localPtr != NULL;
+	    localPtr = localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)) {
+	    Tcl_ListObjAppendElement(interp, listObjPtr,
+		    Tcl_NewStringObj(localPtr->name, -1));
+	}
+    }
+    Tcl_SetObjResult(interp, listObjPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoBodyCmd --
+ *
+ *	Called to implement the "info body" command that returns the body for
+ *	a procedure. Handles the following syntax:
+ *
+ *	    info body procName
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoBodyCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    const char *name;
+    Proc *procPtr;
+    Tcl_Obj *bodyPtr, *resultPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "procname");
+	return TCL_ERROR;
+    }
+
+    name = TclGetString(objv[1]);
+    procPtr = TclFindProc(iPtr, name);
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" isn't a procedure", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PROCEDURE", name, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Here we used to return procPtr->bodyPtr, except when the body was
+     * bytecompiled - in that case, the return was a copy of the body's string
+     * rep. In order to better isolate the implementation details of the
+     * compiler/engine subsystem, we now always return a copy of the string
+     * rep. It is important to return a copy so that later manipulations of
+     * the object do not invalidate the internal rep.
+     */
+
+    bodyPtr = procPtr->bodyPtr;
+    if (bodyPtr->bytes == NULL) {
+	/*
+	 * The string rep might not be valid if the procedure has never been
+	 * run before. [Bug #545644]
+	 */
+
+	TclGetString(bodyPtr);
+    }
+    resultPtr = Tcl_NewStringObj(bodyPtr->bytes, bodyPtr->length);
+
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoCmdCountCmd --
+ *
+ *	Called to implement the "info cmdcount" command that returns the
+ *	number of commands that have been executed. Handles the following
+ *	syntax:
+ *
+ *	    info cmdcount
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoCmdCountCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(iPtr->cmdCount));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoCommandsCmd --
+ *
+ *	Called to implement the "info commands" command that returns the list
+ *	of commands in the interpreter that match an optional pattern. The
+ *	pattern, if any, consists of an optional sequence of namespace names
+ *	separated by "::" qualifiers, which is followed by a glob-style
+ *	pattern that restricts which commands are returned. Handles the
+ *	following syntax:
+ *
+ *	    info commands ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoCommandsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *cmdName, *pattern;
+    const char *simplePattern;
+    register Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch search;
+    Namespace *nsPtr;
+    Namespace *globalNsPtr = (Namespace *) Tcl_GetGlobalNamespace(interp);
+    Namespace *currNsPtr = (Namespace *) Tcl_GetCurrentNamespace(interp);
+    Tcl_Obj *listPtr, *elemObjPtr;
+    int specificNsInPattern = 0;/* Init. to avoid compiler warning. */
+    Tcl_Command cmd;
+    int i;
+
+    /*
+     * Get the pattern and find the "effective namespace" in which to list
+     * commands.
+     */
+
+    if (objc == 1) {
+	simplePattern = NULL;
+	nsPtr = currNsPtr;
+	specificNsInPattern = 0;
+    } else if (objc == 2) {
+	/*
+	 * From the pattern, get the effective namespace and the simple
+	 * pattern (no namespace qualifiers or ::'s) at the end. If an error
+	 * was found while parsing the pattern, return it. Otherwise, if the
+	 * namespace wasn't found, just leave nsPtr NULL: we will return an
+	 * empty list since no commands there can be found.
+	 */
+
+	Namespace *dummy1NsPtr, *dummy2NsPtr;
+
+	pattern = TclGetString(objv[1]);
+	TclGetNamespaceForQualName(interp, pattern, NULL, 0, &nsPtr,
+		&dummy1NsPtr, &dummy2NsPtr, &simplePattern);
+
+	if (nsPtr != NULL) {	/* We successfully found the pattern's ns. */
+	    specificNsInPattern = (strcmp(simplePattern, pattern) != 0);
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Exit as quickly as possible if we couldn't find the namespace.
+     */
+
+    if (nsPtr == NULL) {
+	return TCL_OK;
+    }
+
+    /*
+     * Scan through the effective namespace's command table and create a list
+     * with all commands that match the pattern. If a specific namespace was
+     * requested in the pattern, qualify the command names with the namespace
+     * name.
+     */
+
+    listPtr = Tcl_NewListObj(0, NULL);
+
+    if (simplePattern != NULL && TclMatchIsTrivial(simplePattern)) {
+	/*
+	 * Special case for when the pattern doesn't include any of glob's
+	 * special characters. This lets us avoid scans of any hash tables.
+	 */
+
+	entryPtr = Tcl_FindHashEntry(&nsPtr->cmdTable, simplePattern);
+	if (entryPtr != NULL) {
+	    if (specificNsInPattern) {
+		cmd = Tcl_GetHashValue(entryPtr);
+		elemObjPtr = Tcl_NewObj();
+		Tcl_GetCommandFullName(interp, cmd, elemObjPtr);
+	    } else {
+		cmdName = Tcl_GetHashKey(&nsPtr->cmdTable, entryPtr);
+		elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+	    }
+	    Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+	    Tcl_SetObjResult(interp, listPtr);
+	    return TCL_OK;
+	}
+	if ((nsPtr != globalNsPtr) && !specificNsInPattern) {
+	    Tcl_HashTable *tablePtr = NULL;	/* Quell warning. */
+
+	    for (i=0 ; i<nsPtr->commandPathLength ; i++) {
+		Namespace *pathNsPtr = nsPtr->commandPathArray[i].nsPtr;
+
+		if (pathNsPtr == NULL) {
+		    continue;
+		}
+		tablePtr = &pathNsPtr->cmdTable;
+		entryPtr = Tcl_FindHashEntry(tablePtr, simplePattern);
+		if (entryPtr != NULL) {
+		    break;
+		}
+	    }
+	    if (entryPtr == NULL) {
+		tablePtr = &globalNsPtr->cmdTable;
+		entryPtr = Tcl_FindHashEntry(tablePtr, simplePattern);
+	    }
+	    if (entryPtr != NULL) {
+		cmdName = Tcl_GetHashKey(tablePtr, entryPtr);
+		Tcl_ListObjAppendElement(interp, listPtr,
+			Tcl_NewStringObj(cmdName, -1));
+		Tcl_SetObjResult(interp, listPtr);
+		return TCL_OK;
+	    }
+	}
+    } else if (nsPtr->commandPathLength == 0 || specificNsInPattern) {
+	/*
+	 * The pattern is non-trivial, but either there is no explicit path or
+	 * there is an explicit namespace in the pattern. In both cases, the
+	 * old matching scheme is perfect.
+	 */
+
+	entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+	while (entryPtr != NULL) {
+	    cmdName = Tcl_GetHashKey(&nsPtr->cmdTable, entryPtr);
+	    if ((simplePattern == NULL)
+		    || Tcl_StringMatch(cmdName, simplePattern)) {
+		if (specificNsInPattern) {
+		    cmd = Tcl_GetHashValue(entryPtr);
+		    elemObjPtr = Tcl_NewObj();
+		    Tcl_GetCommandFullName(interp, cmd, elemObjPtr);
+		} else {
+		    elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+		}
+		Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+	    }
+	    entryPtr = Tcl_NextHashEntry(&search);
+	}
+
+	/*
+	 * If the effective namespace isn't the global :: namespace, and a
+	 * specific namespace wasn't requested in the pattern, then add in all
+	 * global :: commands that match the simple pattern. Of course, we add
+	 * in only those commands that aren't hidden by a command in the
+	 * effective namespace.
+	 */
+
+	if ((nsPtr != globalNsPtr) && !specificNsInPattern) {
+	    entryPtr = Tcl_FirstHashEntry(&globalNsPtr->cmdTable, &search);
+	    while (entryPtr != NULL) {
+		cmdName = Tcl_GetHashKey(&globalNsPtr->cmdTable, entryPtr);
+		if ((simplePattern == NULL)
+			|| Tcl_StringMatch(cmdName, simplePattern)) {
+		    if (Tcl_FindHashEntry(&nsPtr->cmdTable,cmdName) == NULL) {
+			Tcl_ListObjAppendElement(interp, listPtr,
+				Tcl_NewStringObj(cmdName, -1));
+		    }
+		}
+		entryPtr = Tcl_NextHashEntry(&search);
+	    }
+	}
+    } else {
+	/*
+	 * The pattern is non-trivial (can match more than one command name),
+	 * there is an explicit path, and there is no explicit namespace in
+	 * the pattern. This means that we have to traverse the path to
+	 * discover all the commands defined.
+	 */
+
+	Tcl_HashTable addedCommandsTable;
+	int isNew;
+	int foundGlobal = (nsPtr == globalNsPtr);
+
+	/*
+	 * We keep a hash of the objects already added to the result list.
+	 */
+
+	Tcl_InitObjHashTable(&addedCommandsTable);
+
+	entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+	while (entryPtr != NULL) {
+	    cmdName = Tcl_GetHashKey(&nsPtr->cmdTable, entryPtr);
+	    if ((simplePattern == NULL)
+		    || Tcl_StringMatch(cmdName, simplePattern)) {
+		elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+		Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		(void) Tcl_CreateHashEntry(&addedCommandsTable,
+			elemObjPtr, &isNew);
+	    }
+	    entryPtr = Tcl_NextHashEntry(&search);
+	}
+
+	/*
+	 * Search the path next.
+	 */
+
+	for (i=0 ; i<nsPtr->commandPathLength ; i++) {
+	    Namespace *pathNsPtr = nsPtr->commandPathArray[i].nsPtr;
+
+	    if (pathNsPtr == NULL) {
+		continue;
+	    }
+	    if (pathNsPtr == globalNsPtr) {
+		foundGlobal = 1;
+	    }
+	    entryPtr = Tcl_FirstHashEntry(&pathNsPtr->cmdTable, &search);
+	    while (entryPtr != NULL) {
+		cmdName = Tcl_GetHashKey(&pathNsPtr->cmdTable, entryPtr);
+		if ((simplePattern == NULL)
+			|| Tcl_StringMatch(cmdName, simplePattern)) {
+		    elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+		    (void) Tcl_CreateHashEntry(&addedCommandsTable,
+			    elemObjPtr, &isNew);
+		    if (isNew) {
+			Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		    } else {
+			TclDecrRefCount(elemObjPtr);
+		    }
+		}
+		entryPtr = Tcl_NextHashEntry(&search);
+	    }
+	}
+
+	/*
+	 * If the effective namespace isn't the global :: namespace, and a
+	 * specific namespace wasn't requested in the pattern, then add in all
+	 * global :: commands that match the simple pattern. Of course, we add
+	 * in only those commands that aren't hidden by a command in the
+	 * effective namespace.
+	 */
+
+	if (!foundGlobal) {
+	    entryPtr = Tcl_FirstHashEntry(&globalNsPtr->cmdTable, &search);
+	    while (entryPtr != NULL) {
+		cmdName = Tcl_GetHashKey(&globalNsPtr->cmdTable, entryPtr);
+		if ((simplePattern == NULL)
+			|| Tcl_StringMatch(cmdName, simplePattern)) {
+		    elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+		    if (Tcl_FindHashEntry(&addedCommandsTable,
+			    (char *) elemObjPtr) == NULL) {
+			Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		    } else {
+			TclDecrRefCount(elemObjPtr);
+		    }
+		}
+		entryPtr = Tcl_NextHashEntry(&search);
+	    }
+	}
+
+	Tcl_DeleteHashTable(&addedCommandsTable);
+    }
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoCompleteCmd --
+ *
+ *	Called to implement the "info complete" command that determines
+ *	whether a string is a complete Tcl command. Handles the following
+ *	syntax:
+ *
+ *	    info complete command
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoCompleteCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "command");
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(
+	    TclObjCommandComplete(objv[1])));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoDefaultCmd --
+ *
+ *	Called to implement the "info default" command that returns the
+ *	default value for a procedure argument. Handles the following syntax:
+ *
+ *	    info default procName arg varName
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoDefaultCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *procName, *argName;
+    Proc *procPtr;
+    CompiledLocal *localPtr;
+    Tcl_Obj *valueObjPtr;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "procname arg varname");
+	return TCL_ERROR;
+    }
+
+    procName = TclGetString(objv[1]);
+    argName = TclGetString(objv[2]);
+
+    procPtr = TclFindProc(iPtr, procName);
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" isn't a procedure", procName));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PROCEDURE", procName,
+		NULL);
+	return TCL_ERROR;
+    }
+
+    for (localPtr = procPtr->firstLocalPtr;  localPtr != NULL;
+	    localPtr = localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)
+		&& (strcmp(argName, localPtr->name) == 0)) {
+	    if (localPtr->defValuePtr != NULL) {
+		valueObjPtr = Tcl_ObjSetVar2(interp, objv[3], NULL,
+			localPtr->defValuePtr, TCL_LEAVE_ERR_MSG);
+		if (valueObjPtr == NULL) {
+		    return TCL_ERROR;
+		}
+		Tcl_SetObjResult(interp, Tcl_NewIntObj(1));
+	    } else {
+		Tcl_Obj *nullObjPtr = Tcl_NewObj();
+
+		valueObjPtr = Tcl_ObjSetVar2(interp, objv[3], NULL,
+			nullObjPtr, TCL_LEAVE_ERR_MSG);
+		if (valueObjPtr == NULL) {
+		    return TCL_ERROR;
+		}
+		Tcl_SetObjResult(interp, Tcl_NewIntObj(0));
+	    }
+	    return TCL_OK;
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "procedure \"%s\" doesn't have an argument \"%s\"",
+	    procName, argName));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ARGUMENT", argName, NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoErrorStackCmd --
+ *
+ *	Called to implement the "info errorstack" command that returns information
+ *	about the last error's call stack. Handles the following syntax:
+ *
+ *	    info errorstack ?interp?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoErrorStackCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Interp *target;
+    Interp *iPtr;
+
+    if ((objc != 1) && (objc != 2)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?interp?");
+	return TCL_ERROR;
+    }
+
+    target = interp;
+    if (objc == 2) {
+	target = Tcl_GetSlave(interp, Tcl_GetString(objv[1]));
+	if (target == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    iPtr = (Interp *) target;
+    Tcl_SetObjResult(interp, iPtr->errorStack);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInfoExistsCmd --
+ *
+ *	Called to implement the "info exists" command that determines whether
+ *	a variable exists. Handles the following syntax:
+ *
+ *	    info exists varName
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInfoExistsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *varName;
+    Var *varPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName");
+	return TCL_ERROR;
+    }
+
+    varName = TclGetString(objv[1]);
+    varPtr = TclVarTraceExists(interp, varName);
+
+    Tcl_SetObjResult(interp,
+	    Tcl_NewBooleanObj(varPtr && varPtr->value.objPtr));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoFrameCmd --
+ *	TIP #280
+ *
+ *	Called to implement the "info frame" command that returns the location
+ *	of either the currently executing command, or its caller. Handles the
+ *	following syntax:
+ *
+ *		info frame ?number?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoFrameCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int level, code = TCL_OK;
+    CmdFrame *framePtr, **cmdFramePtrPtr = &iPtr->cmdFramePtr;
+    CoroutineData *corPtr = iPtr->execEnvPtr->corPtr;
+    int topLevel = 0;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?number?");
+	return TCL_ERROR;
+    }
+
+    while (corPtr) {
+	while (*cmdFramePtrPtr) {
+	    topLevel++;
+	    cmdFramePtrPtr = &((*cmdFramePtrPtr)->nextPtr);
+	}
+	if (corPtr->caller.cmdFramePtr) {
+	    *cmdFramePtrPtr = corPtr->caller.cmdFramePtr;
+	}
+	corPtr = corPtr->callerEEPtr->corPtr;
+    }
+    topLevel += (*cmdFramePtrPtr)->level;
+
+    if (topLevel != iPtr->cmdFramePtr->level) {
+	framePtr = iPtr->cmdFramePtr;
+	while (framePtr) {
+	    framePtr->level = topLevel--;
+	    framePtr = framePtr->nextPtr;
+	}
+	if (topLevel) {
+	    Tcl_Panic("Broken frame level calculation");
+	}
+	topLevel = iPtr->cmdFramePtr->level;
+    }
+
+    if (objc == 1) {
+	/*
+	 * Just "info frame".
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(topLevel));
+	goto done;
+    }
+
+    /*
+     * We've got "info frame level" and must parse the level first.
+     */
+
+    if (TclGetIntFromObj(interp, objv[1], &level) != TCL_OK) {
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    if ((level > topLevel) || (level <= - topLevel)) {
+    levelError:
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad level \"%s\"", TclGetString(objv[1])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "LEVEL",
+		TclGetString(objv[1]), NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Let us convert to relative so that we know how many levels to go back
+     */
+
+    if (level > 0) {
+	level -= topLevel;
+    }
+
+    framePtr = iPtr->cmdFramePtr;
+    while (++level <= 0) {
+	framePtr = framePtr->nextPtr;
+	if (!framePtr) {
+	    goto levelError;
+	}
+    }
+
+    Tcl_SetObjResult(interp, TclInfoFrame(interp, framePtr));
+
+  done:
+    cmdFramePtrPtr = &iPtr->cmdFramePtr;
+    corPtr = iPtr->execEnvPtr->corPtr;
+    while (corPtr) {
+	CmdFrame *endPtr = corPtr->caller.cmdFramePtr;
+
+	if (endPtr) {
+	    if (*cmdFramePtrPtr == endPtr) {
+		*cmdFramePtrPtr = NULL;
+	    } else {
+		CmdFrame *runPtr = *cmdFramePtrPtr;
+
+		while (runPtr->nextPtr != endPtr) {
+		    runPtr->level -= endPtr->level;
+		    runPtr = runPtr->nextPtr;
+		}
+		runPtr->level = 1;
+		runPtr->nextPtr = NULL;
+	    }
+	    cmdFramePtrPtr = &corPtr->caller.cmdFramePtr;
+	}
+	corPtr = corPtr->callerEEPtr->corPtr;
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInfoFrame --
+ *
+ *	Core of InfoFrameCmd, returns TIP280 dict for a given frame.
+ *
+ * Results:
+ *	Returns TIP280 dict.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclInfoFrame(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    CmdFrame *framePtr)		/* Frame to get info for. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *tmpObj;
+    Tcl_Obj *lv[20];		/* Keep uptodate when more keys are added to
+				 * the dict. */
+    int lc = 0;
+    /*
+     * This array is indexed by the TCL_LOCATION_... values, except
+     * for _LAST.
+     */
+    static const char *const typeString[TCL_LOCATION_LAST] = {
+	"eval", "eval", "eval", "precompiled", "source", "proc"
+    };
+    Proc *procPtr = framePtr->framePtr ? framePtr->framePtr->procPtr : NULL;
+    int needsFree = -1;
+
+    /*
+     * Pull the information and construct the dictionary to return, as list.
+     * Regarding use of the CmdFrame fields see tclInt.h, and its definition.
+     */
+
+#define ADD_PAIR(name, value) \
+	TclNewLiteralStringObj(tmpObj, name); \
+	lv[lc++] = tmpObj; \
+	lv[lc++] = (value)
+
+    switch (framePtr->type) {
+    case TCL_LOCATION_EVAL:
+	/*
+	 * Evaluation, dynamic script. Type, line, cmd, the latter through
+	 * str.
+	 */
+
+	ADD_PAIR("type", Tcl_NewStringObj(typeString[framePtr->type], -1));
+	if (framePtr->line) {
+	    ADD_PAIR("line", Tcl_NewIntObj(framePtr->line[0]));
+	} else {
+	    ADD_PAIR("line", Tcl_NewIntObj(1));
+	}
+	ADD_PAIR("cmd", TclGetSourceFromFrame(framePtr, 0, NULL));
+	break;
+
+    case TCL_LOCATION_PREBC:
+	/*
+	 * Precompiled. Result contains the type as signal, nothing else.
+	 */
+
+	ADD_PAIR("type", Tcl_NewStringObj(typeString[framePtr->type], -1));
+	break;
+
+    case TCL_LOCATION_BC: {
+	/*
+	 * Execution of bytecode. Talk to the BC engine to fill out the frame.
+	 */
+
+	CmdFrame *fPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+
+	*fPtr = *framePtr;
+
+	/*
+	 * Note:
+	 * Type BC => f.data.eval.path	  is not used.
+	 *	      f.data.tebc.codePtr is used instead.
+	 */
+
+	TclGetSrcInfoForPc(fPtr);
+
+	/*
+	 * Now filled: cmd.str.(cmd,len), line
+	 * Possibly modified: type, path!
+	 */
+
+	ADD_PAIR("type", Tcl_NewStringObj(typeString[fPtr->type], -1));
+	if (fPtr->line) {
+	    ADD_PAIR("line", Tcl_NewIntObj(fPtr->line[0]));
+	}
+
+	if (fPtr->type == TCL_LOCATION_SOURCE) {
+	    ADD_PAIR("file", fPtr->data.eval.path);
+
+	    /*
+	     * Death of reference by TclGetSrcInfoForPc.
+	     */
+
+	    Tcl_DecrRefCount(fPtr->data.eval.path);
+	}
+
+	ADD_PAIR("cmd", TclGetSourceFromFrame(fPtr, 0, NULL));
+	if (fPtr->cmdObj && framePtr->cmdObj == NULL) {
+	    needsFree = lc - 1;
+	}
+	TclStackFree(interp, fPtr);
+	break;
+    }
+
+    case TCL_LOCATION_SOURCE:
+	/*
+	 * Evaluation of a script file.
+	 */
+
+	ADD_PAIR("type", Tcl_NewStringObj(typeString[framePtr->type], -1));
+	ADD_PAIR("line", Tcl_NewIntObj(framePtr->line[0]));
+	ADD_PAIR("file", framePtr->data.eval.path);
+
+	/*
+	 * Refcount framePtr->data.eval.path goes up when lv is converted into
+	 * the result list object.
+	 */
+
+	ADD_PAIR("cmd", TclGetSourceFromFrame(framePtr, 0, NULL));
+	break;
+
+    case TCL_LOCATION_PROC:
+	Tcl_Panic("TCL_LOCATION_PROC found in standard frame");
+	break;
+    }
+
+    /*
+     * 'proc'. Common to all frame types. Conditional on having an associated
+     * Procedure CallFrame.
+     */
+
+    if (procPtr != NULL) {
+	Tcl_HashEntry *namePtr = procPtr->cmdPtr->hPtr;
+
+	if (namePtr) {
+	    Tcl_Obj *procNameObj;
+
+	    /*
+	     * This is a regular command.
+	     */
+
+	    TclNewObj(procNameObj);
+	    Tcl_GetCommandFullName(interp, (Tcl_Command) procPtr->cmdPtr,
+		    procNameObj);
+	    ADD_PAIR("proc", procNameObj);
+	} else if (procPtr->cmdPtr->clientData) {
+	    ExtraFrameInfo *efiPtr = procPtr->cmdPtr->clientData;
+	    int i;
+
+	    /*
+	     * This is a non-standard command. Luckily, it's told us how to
+	     * render extra information about its frame.
+	     */
+
+	    for (i=0 ; i<efiPtr->length ; i++) {
+		lv[lc++] = Tcl_NewStringObj(efiPtr->fields[i].name, -1);
+		if (efiPtr->fields[i].proc) {
+		    lv[lc++] =
+			efiPtr->fields[i].proc(efiPtr->fields[i].clientData);
+		} else {
+		    lv[lc++] = efiPtr->fields[i].clientData;
+		}
+	    }
+	}
+    }
+
+    /*
+     * 'level'. Common to all frame types. Conditional on having an associated
+     * _visible_ CallFrame.
+     */
+
+    if ((framePtr->framePtr != NULL) && (iPtr->varFramePtr != NULL)) {
+	CallFrame *current = framePtr->framePtr;
+	CallFrame *top = iPtr->varFramePtr;
+	CallFrame *idx;
+
+	for (idx=top ; idx!=NULL ; idx=idx->callerVarPtr) {
+	    if (idx == current) {
+		int c = framePtr->framePtr->level;
+		int t = iPtr->varFramePtr->level;
+
+		ADD_PAIR("level", Tcl_NewIntObj(t - c));
+		break;
+	    }
+	}
+    }
+
+    tmpObj = Tcl_NewListObj(lc, lv);
+    if (needsFree >= 0) {
+	Tcl_DecrRefCount(lv[needsFree]);
+    }
+    return tmpObj;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoFunctionsCmd --
+ *
+ *	Called to implement the "info functions" command that returns the list
+ *	of math functions matching an optional pattern. Handles the following
+ *	syntax:
+ *
+ *	    info functions ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoFunctionsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *script;
+    int code;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    script = Tcl_NewStringObj(
+"	    ::apply [::list {{pattern *}} {\n"
+"		::set cmds {}\n"
+"		::foreach cmd [::info commands ::tcl::mathfunc::$pattern] {\n"
+"		    ::lappend cmds [::namespace tail $cmd]\n"
+"		}\n"
+"		::foreach cmd [::info commands tcl::mathfunc::$pattern] {\n"
+"		    ::set cmd [::namespace tail $cmd]\n"
+"		    ::if {$cmd ni $cmds} {\n"
+"			::lappend cmds $cmd\n"
+"		    }\n"
+"		}\n"
+"		::return $cmds\n"
+"	    } [::namespace current]] ", -1);
+
+    if (objc == 2) {
+	Tcl_Obj *arg = Tcl_NewListObj(1, &(objv[1]));
+
+	Tcl_AppendObjToObj(script, arg);
+	Tcl_DecrRefCount(arg);
+    }
+
+    Tcl_IncrRefCount(script);
+    code = Tcl_EvalObjEx(interp, script, 0);
+
+    Tcl_DecrRefCount(script);
+
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoHostnameCmd --
+ *
+ *	Called to implement the "info hostname" command that returns the host
+ *	name. Handles the following syntax:
+ *
+ *	    info hostname
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoHostnameCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *name;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    name = Tcl_GetHostName();
+    if (name) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(name, -1));
+	return TCL_OK;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    "unable to determine name of host", -1));
+    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "HOSTNAME", "UNKNOWN", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoLevelCmd --
+ *
+ *	Called to implement the "info level" command that returns information
+ *	about the call stack. Handles the following syntax:
+ *
+ *	    info level ?number?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoLevelCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (objc == 1) {		/* Just "info level" */
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(iPtr->varFramePtr->level));
+	return TCL_OK;
+    }
+
+    if (objc == 2) {
+	int level;
+	CallFrame *framePtr, *rootFramePtr = iPtr->rootFramePtr;
+
+	if (TclGetIntFromObj(interp, objv[1], &level) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (level <= 0) {
+	    if (iPtr->varFramePtr == rootFramePtr) {
+		goto levelError;
+	    }
+	    level += iPtr->varFramePtr->level;
+	}
+	for (framePtr=iPtr->varFramePtr ; framePtr!=rootFramePtr;
+		framePtr=framePtr->callerVarPtr) {
+	    if (framePtr->level == level) {
+		break;
+	    }
+	}
+	if (framePtr == rootFramePtr) {
+	    goto levelError;
+	}
+
+	Tcl_SetObjResult(interp,
+		Tcl_NewListObj(framePtr->objc, framePtr->objv));
+	return TCL_OK;
+    }
+
+    Tcl_WrongNumArgs(interp, 1, objv, "?number?");
+    return TCL_ERROR;
+
+  levelError:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "bad level \"%s\"", TclGetString(objv[1])));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "LEVEL",
+	    TclGetString(objv[1]), NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoLibraryCmd --
+ *
+ *	Called to implement the "info library" command that returns the
+ *	library directory for the Tcl installation. Handles the following
+ *	syntax:
+ *
+ *	    info library
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoLibraryCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *libDirName;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    libDirName = Tcl_GetVar2(interp, "tcl_library", NULL, TCL_GLOBAL_ONLY);
+    if (libDirName != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(libDirName, -1));
+	return TCL_OK;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    "no library has been specified for Tcl", -1));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARIABLE", "tcl_library",NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoLoadedCmd --
+ *
+ *	Called to implement the "info loaded" command that returns the
+ *	packages that have been loaded into an interpreter. Handles the
+ *	following syntax:
+ *
+ *	    info loaded ?interp?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoLoadedCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *interpName, *packageName;
+
+    if (objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?interp? ?packageName?");
+	return TCL_ERROR;
+    }
+
+    if (objc < 2) {		/* Get loaded pkgs in all interpreters. */
+	interpName = NULL;
+    } else {			/* Get pkgs just in specified interp. */
+	interpName = TclGetString(objv[1]);
+    }
+    if (objc < 3) {		/* Get loaded files in all packages. */
+	packageName = NULL;
+    } else {			/* Get pkgs just in specified interp. */
+	packageName = TclGetString(objv[2]);
+    }
+    return TclGetLoadedPackagesEx(interp, interpName, packageName);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoNameOfExecutableCmd --
+ *
+ *	Called to implement the "info nameofexecutable" command that returns
+ *	the name of the binary file running this application. Handles the
+ *	following syntax:
+ *
+ *	    info nameofexecutable
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoNameOfExecutableCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, TclGetObjNameOfExecutable());
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoPatchLevelCmd --
+ *
+ *	Called to implement the "info patchlevel" command that returns the
+ *	default value for an argument to a procedure. Handles the following
+ *	syntax:
+ *
+ *	    info patchlevel
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoPatchLevelCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *patchlevel;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    patchlevel = Tcl_GetVar2(interp, "tcl_patchLevel", NULL,
+	    (TCL_GLOBAL_ONLY | TCL_LEAVE_ERR_MSG));
+    if (patchlevel != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(patchlevel, -1));
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoProcsCmd --
+ *
+ *	Called to implement the "info procs" command that returns the list of
+ *	procedures in the interpreter that match an optional pattern. The
+ *	pattern, if any, consists of an optional sequence of namespace names
+ *	separated by "::" qualifiers, which is followed by a glob-style
+ *	pattern that restricts which commands are returned. Handles the
+ *	following syntax:
+ *
+ *	    info procs ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoProcsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *cmdName, *pattern;
+    const char *simplePattern;
+    Namespace *nsPtr;
+#ifdef INFO_PROCS_SEARCH_GLOBAL_NS
+    Namespace *globalNsPtr = (Namespace *) Tcl_GetGlobalNamespace(interp);
+#endif
+    Namespace *currNsPtr = (Namespace *) Tcl_GetCurrentNamespace(interp);
+    Tcl_Obj *listPtr, *elemObjPtr;
+    int specificNsInPattern = 0;/* Init. to avoid compiler warning. */
+    register Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch search;
+    Command *cmdPtr, *realCmdPtr;
+
+    /*
+     * Get the pattern and find the "effective namespace" in which to list
+     * procs.
+     */
+
+    if (objc == 1) {
+	simplePattern = NULL;
+	nsPtr = currNsPtr;
+	specificNsInPattern = 0;
+    } else if (objc == 2) {
+	/*
+	 * From the pattern, get the effective namespace and the simple
+	 * pattern (no namespace qualifiers or ::'s) at the end. If an error
+	 * was found while parsing the pattern, return it. Otherwise, if the
+	 * namespace wasn't found, just leave nsPtr NULL: we will return an
+	 * empty list since no commands there can be found.
+	 */
+
+	Namespace *dummy1NsPtr, *dummy2NsPtr;
+
+	pattern = TclGetString(objv[1]);
+	TclGetNamespaceForQualName(interp, pattern, NULL, /*flags*/ 0, &nsPtr,
+		&dummy1NsPtr, &dummy2NsPtr, &simplePattern);
+
+	if (nsPtr != NULL) {	/* We successfully found the pattern's ns. */
+	    specificNsInPattern = (strcmp(simplePattern, pattern) != 0);
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    if (nsPtr == NULL) {
+	return TCL_OK;
+    }
+
+    /*
+     * Scan through the effective namespace's command table and create a list
+     * with all procs that match the pattern. If a specific namespace was
+     * requested in the pattern, qualify the command names with the namespace
+     * name.
+     */
+
+    listPtr = Tcl_NewListObj(0, NULL);
+#ifndef INFO_PROCS_SEARCH_GLOBAL_NS
+    if (simplePattern != NULL && TclMatchIsTrivial(simplePattern)) {
+	entryPtr = Tcl_FindHashEntry(&nsPtr->cmdTable, simplePattern);
+	if (entryPtr != NULL) {
+	    cmdPtr = Tcl_GetHashValue(entryPtr);
+
+	    if (!TclIsProc(cmdPtr)) {
+		realCmdPtr = (Command *)
+			TclGetOriginalCommand((Tcl_Command) cmdPtr);
+		if (realCmdPtr != NULL && TclIsProc(realCmdPtr)) {
+		    goto simpleProcOK;
+		}
+	    } else {
+	    simpleProcOK:
+		if (specificNsInPattern) {
+		    elemObjPtr = Tcl_NewObj();
+		    Tcl_GetCommandFullName(interp, (Tcl_Command) cmdPtr,
+			    elemObjPtr);
+		} else {
+		    elemObjPtr = Tcl_NewStringObj(simplePattern, -1);
+		}
+		Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+	    }
+	}
+    } else
+#endif /* !INFO_PROCS_SEARCH_GLOBAL_NS */
+    {
+	entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+	while (entryPtr != NULL) {
+	    cmdName = Tcl_GetHashKey(&nsPtr->cmdTable, entryPtr);
+	    if ((simplePattern == NULL)
+		    || Tcl_StringMatch(cmdName, simplePattern)) {
+		cmdPtr = Tcl_GetHashValue(entryPtr);
+
+		if (!TclIsProc(cmdPtr)) {
+		    realCmdPtr = (Command *)
+			    TclGetOriginalCommand((Tcl_Command) cmdPtr);
+		    if (realCmdPtr != NULL && TclIsProc(realCmdPtr)) {
+			goto procOK;
+		    }
+		} else {
+		procOK:
+		    if (specificNsInPattern) {
+			elemObjPtr = Tcl_NewObj();
+			Tcl_GetCommandFullName(interp, (Tcl_Command) cmdPtr,
+				elemObjPtr);
+		    } else {
+			elemObjPtr = Tcl_NewStringObj(cmdName, -1);
+		    }
+		    Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		}
+	    }
+	    entryPtr = Tcl_NextHashEntry(&search);
+	}
+
+	/*
+	 * If the effective namespace isn't the global :: namespace, and a
+	 * specific namespace wasn't requested in the pattern, then add in all
+	 * global :: procs that match the simple pattern. Of course, we add in
+	 * only those procs that aren't hidden by a proc in the effective
+	 * namespace.
+	 */
+
+#ifdef INFO_PROCS_SEARCH_GLOBAL_NS
+	/*
+	 * If "info procs" worked like "info commands", returning the commands
+	 * also seen in the global namespace, then you would include this
+	 * code. As this could break backwards compatibilty with 8.0-8.2, we
+	 * decided not to "fix" it in 8.3, leaving the behavior slightly
+	 * different.
+	 */
+
+	if ((nsPtr != globalNsPtr) && !specificNsInPattern) {
+	    entryPtr = Tcl_FirstHashEntry(&globalNsPtr->cmdTable, &search);
+	    while (entryPtr != NULL) {
+		cmdName = Tcl_GetHashKey(&globalNsPtr->cmdTable, entryPtr);
+		if ((simplePattern == NULL)
+			|| Tcl_StringMatch(cmdName, simplePattern)) {
+		    if (Tcl_FindHashEntry(&nsPtr->cmdTable,cmdName) == NULL) {
+			cmdPtr = Tcl_GetHashValue(entryPtr);
+			realCmdPtr = (Command *) TclGetOriginalCommand(
+				(Tcl_Command) cmdPtr);
+
+			if (TclIsProc(cmdPtr) || ((realCmdPtr != NULL)
+				&& TclIsProc(realCmdPtr))) {
+			    Tcl_ListObjAppendElement(interp, listPtr,
+				    Tcl_NewStringObj(cmdName, -1));
+			}
+		    }
+		}
+		entryPtr = Tcl_NextHashEntry(&search);
+	    }
+	}
+#endif
+    }
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoScriptCmd --
+ *
+ *	Called to implement the "info script" command that returns the script
+ *	file that is currently being evaluated. Handles the following syntax:
+ *
+ *	    info script ?newName?
+ *
+ *	If newName is specified, it will set that as the internal name.
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message. It may change the internal
+ *	script filename.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoScriptCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    if ((objc != 1) && (objc != 2)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?filename?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	if (iPtr->scriptFile != NULL) {
+	    Tcl_DecrRefCount(iPtr->scriptFile);
+	}
+	iPtr->scriptFile = objv[1];
+	Tcl_IncrRefCount(iPtr->scriptFile);
+    }
+    if (iPtr->scriptFile != NULL) {
+	Tcl_SetObjResult(interp, iPtr->scriptFile);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoSharedlibCmd --
+ *
+ *	Called to implement the "info sharedlibextension" command that returns
+ *	the file extension used for shared libraries. Handles the following
+ *	syntax:
+ *
+ *	    info sharedlibextension
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoSharedlibCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+#ifdef TCL_SHLIB_EXT
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(TCL_SHLIB_EXT, -1));
+#endif
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoTclVersionCmd --
+ *
+ *	Called to implement the "info tclversion" command that returns the
+ *	version number for this Tcl library. Handles the following syntax:
+ *
+ *	    info tclversion
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InfoTclVersionCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *version;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    version = Tcl_GetVar2Ex(interp, "tcl_version", NULL,
+	    (TCL_GLOBAL_ONLY | TCL_LEAVE_ERR_MSG));
+    if (version != NULL) {
+	Tcl_SetObjResult(interp, version);
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_JoinObjCmd --
+ *
+ *	This procedure is invoked to process the "join" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_JoinObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    int listLen;
+    Tcl_Obj *resObjPtr = NULL, *joinObjPtr, **elemPtrs;
+
+    if ((objc < 2) || (objc > 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list ?joinString?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure the list argument is a list object and get its length and a
+     * pointer to its array of element pointers.
+     */
+
+    if (TclListObjGetElements(interp, objv[1], &listLen,
+	    &elemPtrs) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (listLen == 0) {
+	/* No elements to join; default empty result is correct. */
+	return TCL_OK;
+    }
+    if (listLen == 1) {
+	/* One element; return it */
+	Tcl_SetObjResult(interp, elemPtrs[0]);
+	return TCL_OK;
+    }
+
+    joinObjPtr = (objc == 2) ? Tcl_NewStringObj(" ", 1) : objv[2];
+    Tcl_IncrRefCount(joinObjPtr);
+
+    if (Tcl_GetCharLength(joinObjPtr) == 0) {
+	TclStringCatObjv(interp, /* inPlace */ 0, listLen, elemPtrs,
+		&resObjPtr);
+    } else {
+	int i;
+
+	resObjPtr = Tcl_NewObj();
+	for (i = 0;  i < listLen;  i++) {
+	    if (i > 0) {
+
+		/*
+		 * NOTE: This code is relying on Tcl_AppendObjToObj() **NOT**
+		 * to shimmer joinObjPtr.  If it did, then the case where
+		 * objv[1] and objv[2] are the same value would not be safe.
+		 * Accessing elemPtrs would crash.
+		 */
+
+		Tcl_AppendObjToObj(resObjPtr, joinObjPtr);
+	    }
+	    Tcl_AppendObjToObj(resObjPtr, elemPtrs[i]);
+	}
+    }
+    Tcl_DecrRefCount(joinObjPtr);
+    if (resObjPtr) {
+	Tcl_SetObjResult(interp, resObjPtr);
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LassignObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "lassign" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LassignObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *listCopyPtr;
+    Tcl_Obj **listObjv;		/* The contents of the list. */
+    int listObjc;		/* The length of the list. */
+    int code = TCL_OK;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list ?varName ...?");
+	return TCL_ERROR;
+    }
+
+    listCopyPtr = TclListObjCopy(interp, objv[1]);
+    if (listCopyPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    TclListObjGetElements(NULL, listCopyPtr, &listObjc, &listObjv);
+
+    objc -= 2;
+    objv += 2;
+    while (code == TCL_OK && objc > 0 && listObjc > 0) {
+	if (Tcl_ObjSetVar2(interp, *objv++, NULL, *listObjv++,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    code = TCL_ERROR;
+	}
+	objc--;
+	listObjc--;
+    }
+
+    if (code == TCL_OK && objc > 0) {
+	Tcl_Obj *emptyObj;
+
+	TclNewObj(emptyObj);
+	Tcl_IncrRefCount(emptyObj);
+	while (code == TCL_OK && objc-- > 0) {
+	    if (Tcl_ObjSetVar2(interp, *objv++, NULL, emptyObj,
+		    TCL_LEAVE_ERR_MSG) == NULL) {
+		code = TCL_ERROR;
+	    }
+	}
+	Tcl_DecrRefCount(emptyObj);
+    }
+
+    if (code == TCL_OK && listObjc > 0) {
+	Tcl_SetObjResult(interp, Tcl_NewListObj(listObjc, listObjv));
+    }
+
+    Tcl_DecrRefCount(listCopyPtr);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LindexObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "lindex" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LindexObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+
+    Tcl_Obj *elemPtr;		/* Pointer to the element being extracted. */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list ?index ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If objc==3, then objv[2] may be either a single index or a list of
+     * indices: go to TclLindexList to determine which. If objc>=4, or
+     * objc==2, then objv[2 .. objc-2] are all single indices and processed as
+     * such in TclLindexFlat.
+     */
+
+    if (objc == 3) {
+	elemPtr = TclLindexList(interp, objv[1], objv[2]);
+    } else {
+	elemPtr = TclLindexFlat(interp, objv[1], objc-2, objv+2);
+    }
+
+    /*
+     * Set the interpreter's object result to the last element extracted.
+     */
+
+    if (elemPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, elemPtr);
+    Tcl_DecrRefCount(elemPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LinsertObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "linsert" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A new Tcl list object formed by inserting zero or more elements into a
+ *	list.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LinsertObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    register int objc,		/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *listPtr;
+    int index, len, result;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list index ?element ...?");
+	return TCL_ERROR;
+    }
+
+    result = TclListObjLength(interp, objv[1], &len);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    /*
+     * Get the index. "end" is interpreted to be the index after the last
+     * element, such that using it will cause any inserted elements to be
+     * appended to the list.
+     */
+
+    result = TclGetIntForIndexM(interp, objv[2], /*end*/ len, &index);
+    if (result != TCL_OK) {
+	return result;
+    }
+    if (index > len) {
+	index = len;
+    }
+
+    /*
+     * If the list object is unshared we can modify it directly. Otherwise we
+     * create a copy to modify: this is "copy on write".
+     */
+
+    listPtr = objv[1];
+    if (Tcl_IsShared(listPtr)) {
+	listPtr = TclListObjCopy(NULL, listPtr);
+    }
+
+    if ((objc == 4) && (index == len)) {
+	/*
+	 * Special case: insert one element at the end of the list.
+	 */
+
+	Tcl_ListObjAppendElement(NULL, listPtr, objv[3]);
+    } else {
+	if (TCL_OK != Tcl_ListObjReplace(interp, listPtr, index, 0,
+		(objc-3), &(objv[3]))) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Set the interpreter's object result.
+     */
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjCmd --
+ *
+ *	This procedure is invoked to process the "list" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    register int objc,		/* Number of arguments. */
+    register Tcl_Obj *const objv[])
+				/* The argument objects. */
+{
+    /*
+     * If there are no list elements, the result is an empty object.
+     * Otherwise set the interpreter's result object to be a list object.
+     */
+
+    if (objc > 1) {
+	Tcl_SetObjResult(interp, Tcl_NewListObj(objc-1, &objv[1]));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LlengthObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "llength" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LlengthObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    register Tcl_Obj *const objv[])
+				/* Argument objects. */
+{
+    int listLen, result;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list");
+	return TCL_ERROR;
+    }
+
+    result = TclListObjLength(interp, objv[1], &listLen);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    /*
+     * Set the interpreter's object result to an integer object holding the
+     * length.
+     */
+
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(listLen));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LrangeObjCmd --
+ *
+ *	This procedure is invoked to process the "lrange" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LrangeObjCmd(
+    ClientData notUsed,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    register Tcl_Obj *const objv[])
+				/* Argument objects. */
+{
+    Tcl_Obj **elemPtrs;
+    int listLen, first, last, result;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list first last");
+	return TCL_ERROR;
+    }
+
+    result = TclListObjLength(interp, objv[1], &listLen);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    result = TclGetIntForIndexM(interp, objv[2], /*endValue*/ listLen - 1,
+	    &first);
+    if (result != TCL_OK) {
+	return result;
+    }
+    if (first < 0) {
+	first = 0;
+    }
+
+    result = TclGetIntForIndexM(interp, objv[3], /*endValue*/ listLen - 1,
+	    &last);
+    if (result != TCL_OK) {
+	return result;
+    }
+    if (last >= listLen) {
+	last = listLen - 1;
+    }
+
+    if (first > last) {
+	/*
+	 * Returning an empty list is easy.
+	 */
+
+	return TCL_OK;
+    }
+
+    result = TclListObjGetElements(interp, objv[1], &listLen, &elemPtrs);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    if (Tcl_IsShared(objv[1]) ||
+	    ((ListRepPtr(objv[1])->refCount > 1))) {
+	Tcl_SetObjResult(interp, Tcl_NewListObj(last - first + 1,
+		&elemPtrs[first]));
+    } else {
+	/*
+	 * In-place is possible.
+	 */
+
+	if (last < (listLen - 1)) {
+	    Tcl_ListObjReplace(interp, objv[1], last + 1, listLen - 1 - last,
+		    0, NULL);
+	}
+
+	/*
+	 * This one is not conditioned on (first > 0) in order to preserve the
+	 * string-canonizing effect of [lrange 0 end].
+	 */
+
+	Tcl_ListObjReplace(interp, objv[1], 0, first, 0, NULL);
+	Tcl_SetObjResult(interp, objv[1]);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LrepeatObjCmd --
+ *
+ *	This procedure is invoked to process the "lrepeat" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LrepeatObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    register int objc,		/* Number of arguments. */
+    register Tcl_Obj *const objv[])
+				/* The argument objects. */
+{
+    int elementCount, i, totalElems;
+    Tcl_Obj *listPtr, **dataArray = NULL;
+
+    /*
+     * Check arguments for legality:
+     *		lrepeat count ?value ...?
+     */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "count ?value ...?");
+	return TCL_ERROR;
+    }
+    if (TCL_OK != TclGetIntFromObj(interp, objv[1], &elementCount)) {
+	return TCL_ERROR;
+    }
+    if (elementCount < 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad count \"%d\": must be integer >= 0", elementCount));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LREPEAT", "NEGARG",
+		NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Skip forward to the interesting arguments now we've finished parsing.
+     */
+
+    objc -= 2;
+    objv += 2;
+
+    /* Final sanity check. Do not exceed limits on max list length. */
+
+    if (elementCount && objc > LIST_MAX/elementCount) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"max length of a Tcl list (%d elements) exceeded", LIST_MAX));
+	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	return TCL_ERROR;
+    }
+    totalElems = objc * elementCount;
+
+    /*
+     * Get an empty list object that is allocated large enough to hold each
+     * init value elementCount times.
+     */
+
+    listPtr = Tcl_NewListObj(totalElems, NULL);
+    if (totalElems) {
+	List *listRepPtr = ListRepPtr(listPtr);
+
+	listRepPtr->elemCount = elementCount*objc;
+	dataArray = &listRepPtr->elements;
+    }
+
+    /*
+     * Set the elements. Note that we handle the common degenerate case of a
+     * single value being repeated separately to permit the compiler as much
+     * room as possible to optimize a loop that might be run a very large
+     * number of times.
+     */
+
+    CLANG_ASSERT(dataArray || totalElems == 0 );
+    if (objc == 1) {
+	register Tcl_Obj *tmpPtr = objv[0];
+
+	tmpPtr->refCount += elementCount;
+	for (i=0 ; i<elementCount ; i++) {
+	    dataArray[i] = tmpPtr;
+	}
+    } else {
+	int j, k = 0;
+
+	for (i=0 ; i<elementCount ; i++) {
+	    for (j=0 ; j<objc ; j++) {
+		Tcl_IncrRefCount(objv[j]);
+		dataArray[k++] = objv[j];
+	    }
+	}
+    }
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LreplaceObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "lreplace" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A new Tcl list object formed by replacing zero or more elements of a
+ *	list.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LreplaceObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Tcl_Obj *listPtr;
+    int first, last, listLen, numToDelete, result;
+
+    if (objc < 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"list first last ?element ...?");
+	return TCL_ERROR;
+    }
+
+    result = TclListObjLength(interp, objv[1], &listLen);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    /*
+     * Get the first and last indexes. "end" is interpreted to be the index
+     * for the last element, such that using it will cause that element to be
+     * included for deletion.
+     */
+
+    result = TclGetIntForIndexM(interp, objv[2], /*end*/ listLen-1, &first);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    result = TclGetIntForIndexM(interp, objv[3], /*end*/ listLen-1, &last);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    if (first < 0) {
+	first = 0;
+    }
+
+    /*
+     * Complain if the user asked for a start element that is greater than the
+     * list length. This won't ever trigger for the "end-*" case as that will
+     * be properly constrained by TclGetIntForIndex because we use listLen-1
+     * (to allow for replacing the last elem).
+     */
+
+    if ((first > listLen) && (listLen > 0)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"list doesn't contain element %s", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LREPLACE", "BADIDX",
+		NULL);
+	return TCL_ERROR;
+    }
+    if (last >= listLen) {
+	last = listLen - 1;
+    }
+    if (first <= last) {
+	numToDelete = last - first + 1;
+    } else {
+	numToDelete = 0;
+    }
+
+    /*
+     * If the list object is unshared we can modify it directly, otherwise we
+     * create a copy to modify: this is "copy on write".
+     */
+
+    listPtr = objv[1];
+    if (Tcl_IsShared(listPtr)) {
+	listPtr = TclListObjCopy(NULL, listPtr);
+    }
+
+    /*
+     * Note that we call Tcl_ListObjReplace even when numToDelete == 0 and
+     * objc == 4. In this case, the list value of listPtr is not changed (no
+     * elements are removed or added), but by making the call we are assured
+     * we end up with a list in canonical form. Resist any temptation to
+     * optimize this case away.
+     */
+
+    if (TCL_OK != Tcl_ListObjReplace(interp, listPtr, first, numToDelete,
+	    objc-4, objv+4)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set the interpreter's object result.
+     */
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LreverseObjCmd --
+ *
+ *	This procedure is invoked to process the "lreverse" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LreverseObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument values. */
+{
+    Tcl_Obj **elemv;
+    int elemc, i, j;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "list");
+	return TCL_ERROR;
+    }
+    if (TclListObjGetElements(interp, objv[1], &elemc, &elemv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the list is empty, just return it. [Bug 1876793]
+     */
+
+    if (!elemc) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    if (Tcl_IsShared(objv[1])
+	    || (ListRepPtr(objv[1])->refCount > 1)) {	/* Bug 1675044 */
+	Tcl_Obj *resultObj, **dataArray;
+	List *listRepPtr;
+
+	resultObj = Tcl_NewListObj(elemc, NULL);
+	listRepPtr = ListRepPtr(resultObj);
+	listRepPtr->elemCount = elemc;
+	dataArray = &listRepPtr->elements;
+
+	for (i=0,j=elemc-1 ; i<elemc ; i++,j--) {
+	    dataArray[j] = elemv[i];
+	    Tcl_IncrRefCount(elemv[i]);
+	}
+
+	Tcl_SetObjResult(interp, resultObj);
+    } else {
+
+	/*
+	 * Not shared, so swap "in place". This relies on Tcl_LOGE above
+	 * returning a pointer to the live array of Tcl_Obj values.
+	 */
+
+	for (i=0,j=elemc-1 ; i<j ; i++,j--) {
+	    Tcl_Obj *tmp = elemv[i];
+
+	    elemv[i] = elemv[j];
+	    elemv[j] = tmp;
+	}
+	TclInvalidateStringRep(objv[1]);
+	Tcl_SetObjResult(interp, objv[1]);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LsearchObjCmd --
+ *
+ *	This procedure is invoked to process the "lsearch" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LsearchObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument values. */
+{
+    const char *bytes, *patternBytes;
+    int i, match, index, result, listc, length, elemLen, bisect;
+    int dataType, isIncreasing, lower, upper, offset;
+    Tcl_WideInt patWide, objWide;
+    int allMatches, inlineReturn, negatedMatch, returnSubindices, noCase;
+    double patDouble, objDouble;
+    SortInfo sortInfo;
+    Tcl_Obj *patObj, **listv, *listPtr, *startPtr, *itemPtr;
+    SortStrCmpFn_t strCmpFn = strcmp;
+    Tcl_RegExp regexp = NULL;
+    static const char *const options[] = {
+	"-all",	    "-ascii",   "-bisect", "-decreasing", "-dictionary",
+	"-exact",   "-glob",    "-increasing", "-index",
+	"-inline",  "-integer", "-nocase",     "-not",
+	"-real",    "-regexp",  "-sorted",     "-start",
+	"-subindices", NULL
+    };
+    enum options {
+	LSEARCH_ALL, LSEARCH_ASCII, LSEARCH_BISECT, LSEARCH_DECREASING,
+	LSEARCH_DICTIONARY, LSEARCH_EXACT, LSEARCH_GLOB, LSEARCH_INCREASING,
+	LSEARCH_INDEX, LSEARCH_INLINE, LSEARCH_INTEGER, LSEARCH_NOCASE,
+	LSEARCH_NOT, LSEARCH_REAL, LSEARCH_REGEXP, LSEARCH_SORTED,
+	LSEARCH_START, LSEARCH_SUBINDICES
+    };
+    enum datatypes {
+	ASCII, DICTIONARY, INTEGER, REAL
+    };
+    enum modes {
+	EXACT, GLOB, REGEXP, SORTED
+    };
+    enum modes mode;
+
+    mode = GLOB;
+    dataType = ASCII;
+    isIncreasing = 1;
+    allMatches = 0;
+    inlineReturn = 0;
+    returnSubindices = 0;
+    negatedMatch = 0;
+    bisect = 0;
+    listPtr = NULL;
+    startPtr = NULL;
+    offset = 0;
+    noCase = 0;
+    sortInfo.compareCmdPtr = NULL;
+    sortInfo.isIncreasing = 1;
+    sortInfo.sortMode = 0;
+    sortInfo.interp = interp;
+    sortInfo.resultCode = TCL_OK;
+    sortInfo.indexv = NULL;
+    sortInfo.indexc = 0;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-option value ...? list pattern");
+	return TCL_ERROR;
+    }
+
+    for (i = 1; i < objc-2; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0, &index)
+		!= TCL_OK) {
+	    if (startPtr != NULL) {
+		Tcl_DecrRefCount(startPtr);
+	    }
+	    result = TCL_ERROR;
+	    goto done;
+	}
+	switch ((enum options) index) {
+	case LSEARCH_ALL:		/* -all */
+	    allMatches = 1;
+	    break;
+	case LSEARCH_ASCII:		/* -ascii */
+	    dataType = ASCII;
+	    break;
+	case LSEARCH_BISECT:		/* -bisect */
+	    mode = SORTED;
+	    bisect = 1;
+	    break;
+	case LSEARCH_DECREASING:	/* -decreasing */
+	    isIncreasing = 0;
+	    sortInfo.isIncreasing = 0;
+	    break;
+	case LSEARCH_DICTIONARY:	/* -dictionary */
+	    dataType = DICTIONARY;
+	    break;
+	case LSEARCH_EXACT:		/* -increasing */
+	    mode = EXACT;
+	    break;
+	case LSEARCH_GLOB:		/* -glob */
+	    mode = GLOB;
+	    break;
+	case LSEARCH_INCREASING:	/* -increasing */
+	    isIncreasing = 1;
+	    sortInfo.isIncreasing = 1;
+	    break;
+	case LSEARCH_INLINE:		/* -inline */
+	    inlineReturn = 1;
+	    break;
+	case LSEARCH_INTEGER:		/* -integer */
+	    dataType = INTEGER;
+	    break;
+	case LSEARCH_NOCASE:		/* -nocase */
+	    strCmpFn = TclUtfCasecmp;
+	    noCase = 1;
+	    break;
+	case LSEARCH_NOT:		/* -not */
+	    negatedMatch = 1;
+	    break;
+	case LSEARCH_REAL:		/* -real */
+	    dataType = REAL;
+	    break;
+	case LSEARCH_REGEXP:		/* -regexp */
+	    mode = REGEXP;
+	    break;
+	case LSEARCH_SORTED:		/* -sorted */
+	    mode = SORTED;
+	    break;
+	case LSEARCH_SUBINDICES:	/* -subindices */
+	    returnSubindices = 1;
+	    break;
+	case LSEARCH_START:		/* -start */
+	    /*
+	     * If there was a previous -start option, release its saved index
+	     * because it will either be replaced or there will be an error.
+	     */
+
+	    if (startPtr != NULL) {
+		Tcl_DecrRefCount(startPtr);
+	    }
+	    if (i > objc-4) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing starting index", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		result = TCL_ERROR;
+		goto done;
+	    }
+	    i++;
+	    if (objv[i] == objv[objc - 2]) {
+		/*
+		 * Take copy to prevent shimmering problems. Note that it does
+		 * not matter if the index obj is also a component of the list
+		 * being searched. We only need to copy where the list and the
+		 * index are one-and-the-same.
+		 */
+
+		startPtr = Tcl_DuplicateObj(objv[i]);
+	    } else {
+		startPtr = objv[i];
+		Tcl_IncrRefCount(startPtr);
+	    }
+	    break;
+	case LSEARCH_INDEX: {		/* -index */
+	    Tcl_Obj **indices;
+	    int j;
+
+	    if (sortInfo.indexc > 1) {
+		TclStackFree(interp, sortInfo.indexv);
+	    }
+	    if (i > objc-4) {
+		if (startPtr != NULL) {
+		    Tcl_DecrRefCount(startPtr);
+		}
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-index\" option must be followed by list index",
+			-1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		return TCL_ERROR;
+	    }
+
+	    /*
+	     * Store the extracted indices for processing by sublist
+	     * extraction. Note that we don't do this using objects because
+	     * that has shimmering problems.
+	     */
+
+	    i++;
+	    if (TclListObjGetElements(interp, objv[i],
+		    &sortInfo.indexc, &indices) != TCL_OK) {
+		if (startPtr != NULL) {
+		    Tcl_DecrRefCount(startPtr);
+		}
+		return TCL_ERROR;
+	    }
+	    switch (sortInfo.indexc) {
+	    case 0:
+		sortInfo.indexv = NULL;
+		break;
+	    case 1:
+		sortInfo.indexv = &sortInfo.singleIndex;
+		break;
+	    default:
+		sortInfo.indexv =
+			TclStackAlloc(interp, sizeof(int) * sortInfo.indexc);
+	    }
+
+	    /*
+	     * Fill the array by parsing each index. We don't know whether
+	     * their scale is sensible yet, but we at least perform the
+	     * syntactic check here.
+	     */
+
+	    for (j=0 ; j<sortInfo.indexc ; j++) {
+		if (TclGetIntForIndexM(interp, indices[j], SORTIDX_END,
+			&sortInfo.indexv[j]) != TCL_OK) {
+		    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			    "\n    (-index option item number %d)", j));
+		    result = TCL_ERROR;
+		    goto done;
+		}
+	    }
+	    break;
+	}
+	}
+    }
+
+    /*
+     * Subindices only make sense if asked for with -index option set.
+     */
+
+    if (returnSubindices && sortInfo.indexc==0) {
+	if (startPtr != NULL) {
+	    Tcl_DecrRefCount(startPtr);
+	}
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"-subindices cannot be used without -index option", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSEARCH",
+		"BAD_OPTION_MIX", NULL);
+	return TCL_ERROR;
+    }
+
+    if (bisect && (allMatches || negatedMatch)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"-bisect is not compatible with -all or -not", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSEARCH",
+		"BAD_OPTION_MIX", NULL);
+	return TCL_ERROR;
+    }
+
+    if (mode == REGEXP) {
+	/*
+	 * We can shimmer regexp/list if listv[i] == pattern, so get the
+	 * regexp rep before the list rep. First time round, omit the interp
+	 * and hope that the compilation will succeed. If it fails, we'll
+	 * recompile in "expensive" mode with a place to put error messages.
+	 */
+
+	regexp = Tcl_GetRegExpFromObj(NULL, objv[objc - 1],
+		TCL_REG_ADVANCED | TCL_REG_NOSUB |
+		(noCase ? TCL_REG_NOCASE : 0));
+	if (regexp == NULL) {
+	    /*
+	     * Failed to compile the RE. Try again without the TCL_REG_NOSUB
+	     * flag in case the RE had sub-expressions in it [Bug 1366683]. If
+	     * this fails, an error message will be left in the interpreter.
+	     */
+
+	    regexp = Tcl_GetRegExpFromObj(interp, objv[objc - 1],
+		    TCL_REG_ADVANCED | (noCase ? TCL_REG_NOCASE : 0));
+	}
+
+	if (regexp == NULL) {
+	    if (startPtr != NULL) {
+		Tcl_DecrRefCount(startPtr);
+	    }
+	    result = TCL_ERROR;
+	    goto done;
+	}
+    }
+
+    /*
+     * Make sure the list argument is a list object and get its length and a
+     * pointer to its array of element pointers.
+     */
+
+    result = TclListObjGetElements(interp, objv[objc - 2], &listc, &listv);
+    if (result != TCL_OK) {
+	if (startPtr != NULL) {
+	    Tcl_DecrRefCount(startPtr);
+	}
+	goto done;
+    }
+
+    /*
+     * Get the user-specified start offset.
+     */
+
+    if (startPtr) {
+	result = TclGetIntForIndexM(interp, startPtr, listc-1, &offset);
+	Tcl_DecrRefCount(startPtr);
+	if (result != TCL_OK) {
+	    goto done;
+	}
+	if (offset < 0) {
+	    offset = 0;
+	}
+
+	/*
+	 * If the search started past the end of the list, we just return a
+	 * "did not match anything at all" result straight away. [Bug 1374778]
+	 */
+
+	if (offset > listc-1) {
+	    if (sortInfo.indexc > 1) {
+		TclStackFree(interp, sortInfo.indexv);
+	    }
+	    if (allMatches || inlineReturn) {
+		Tcl_ResetResult(interp);
+	    } else {
+		Tcl_SetObjResult(interp, Tcl_NewIntObj(-1));
+	    }
+	    return TCL_OK;
+	}
+    }
+
+    patObj = objv[objc - 1];
+    patternBytes = NULL;
+    if (mode == EXACT || mode == SORTED) {
+	switch ((enum datatypes) dataType) {
+	case ASCII:
+	case DICTIONARY:
+	    patternBytes = TclGetStringFromObj(patObj, &length);
+	    break;
+	case INTEGER:
+	    result = TclGetWideIntFromObj(interp, patObj, &patWide);
+	    if (result != TCL_OK) {
+		goto done;
+	    }
+
+	    /*
+	     * List representation might have been shimmered; restore it. [Bug
+	     * 1844789]
+	     */
+
+	    TclListObjGetElements(NULL, objv[objc - 2], &listc, &listv);
+	    break;
+	case REAL:
+	    result = Tcl_GetDoubleFromObj(interp, patObj, &patDouble);
+	    if (result != TCL_OK) {
+		goto done;
+	    }
+
+	    /*
+	     * List representation might have been shimmered; restore it. [Bug
+	     * 1844789]
+	     */
+
+	    TclListObjGetElements(NULL, objv[objc - 2], &listc, &listv);
+	    break;
+	}
+    } else {
+	patternBytes = TclGetStringFromObj(patObj, &length);
+    }
+
+    /*
+     * Set default index value to -1, indicating failure; if we find the item
+     * in the course of our search, index will be set to the correct value.
+     */
+
+    index = -1;
+    match = 0;
+
+    if (mode == SORTED && !allMatches && !negatedMatch) {
+	/*
+	 * If the data is sorted, we can do a more intelligent search. Note
+	 * that there is no point in being smart when -all was specified; in
+	 * that case, we have to look at all items anyway, and there is no
+	 * sense in doing this when the match sense is inverted.
+	 */
+
+	lower = offset - 1;
+	upper = listc;
+	while (lower + 1 != upper && sortInfo.resultCode == TCL_OK) {
+	    i = (lower + upper)/2;
+	    if (sortInfo.indexc != 0) {
+		itemPtr = SelectObjFromSublist(listv[i], &sortInfo);
+		if (sortInfo.resultCode != TCL_OK) {
+		    result = sortInfo.resultCode;
+		    goto done;
+		}
+	    } else {
+		itemPtr = listv[i];
+	    }
+	    switch ((enum datatypes) dataType) {
+	    case ASCII:
+		bytes = TclGetString(itemPtr);
+		match = strCmpFn(patternBytes, bytes);
+		break;
+	    case DICTIONARY:
+		bytes = TclGetString(itemPtr);
+		match = DictionaryCompare(patternBytes, bytes);
+		break;
+	    case INTEGER:
+		result = TclGetWideIntFromObj(interp, itemPtr, &objWide);
+		if (result != TCL_OK) {
+		    goto done;
+		}
+		if (patWide == objWide) {
+		    match = 0;
+		} else if (patWide < objWide) {
+		    match = -1;
+		} else {
+		    match = 1;
+		}
+		break;
+	    case REAL:
+		result = Tcl_GetDoubleFromObj(interp, itemPtr, &objDouble);
+		if (result != TCL_OK) {
+		    goto done;
+		}
+		if (patDouble == objDouble) {
+		    match = 0;
+		} else if (patDouble < objDouble) {
+		    match = -1;
+		} else {
+		    match = 1;
+		}
+		break;
+	    }
+	    if (match == 0) {
+		/*
+		 * Normally, binary search is written to stop when it finds a
+		 * match. If there are duplicates of an element in the list,
+		 * our first match might not be the first occurance.
+		 * Consider: 0 0 0 1 1 1 2 2 2
+		 *
+		 * To maintain consistancy with standard lsearch semantics, we
+		 * must find the leftmost occurance of the pattern in the
+		 * list. Thus we don't just stop searching here. This
+		 * variation means that a search always makes log n
+		 * comparisons (normal binary search might "get lucky" with an
+		 * early comparison).
+		 *
+		 * In bisect mode though, we want the last of equals.
+		 */
+
+		index = i;
+		if (bisect) {
+		    lower = i;
+		} else {
+		    upper = i;
+		}
+	    } else if (match > 0) {
+		if (isIncreasing) {
+		    lower = i;
+		} else {
+		    upper = i;
+		}
+	    } else {
+		if (isIncreasing) {
+		    upper = i;
+		} else {
+		    lower = i;
+		}
+	    }
+	}
+	if (bisect && index < 0) {
+	    index = lower;
+	}
+    } else {
+	/*
+	 * We need to do a linear search, because (at least one) of:
+	 *   - our matcher can only tell equal vs. not equal
+	 *   - our matching sense is negated
+	 *   - we're building a list of all matched items
+	 */
+
+	if (allMatches) {
+	    listPtr = Tcl_NewListObj(0, NULL);
+	}
+	for (i = offset; i < listc; i++) {
+	    match = 0;
+	    if (sortInfo.indexc != 0) {
+		itemPtr = SelectObjFromSublist(listv[i], &sortInfo);
+		if (sortInfo.resultCode != TCL_OK) {
+		    if (listPtr != NULL) {
+			Tcl_DecrRefCount(listPtr);
+		    }
+		    result = sortInfo.resultCode;
+		    goto done;
+		}
+	    } else {
+		itemPtr = listv[i];
+	    }
+
+	    switch (mode) {
+	    case SORTED:
+	    case EXACT:
+		switch ((enum datatypes) dataType) {
+		case ASCII:
+		    bytes = TclGetStringFromObj(itemPtr, &elemLen);
+		    if (length == elemLen) {
+			/*
+			 * This split allows for more optimal compilation of
+			 * memcmp/strcasecmp.
+			 */
+
+			if (noCase) {
+			    match = (TclUtfCasecmp(bytes, patternBytes) == 0);
+			} else {
+			    match = (memcmp(bytes, patternBytes,
+				    (size_t) length) == 0);
+			}
+		    }
+		    break;
+
+		case DICTIONARY:
+		    bytes = TclGetString(itemPtr);
+		    match = (DictionaryCompare(bytes, patternBytes) == 0);
+		    break;
+
+		case INTEGER:
+		    result = TclGetWideIntFromObj(interp, itemPtr, &objWide);
+		    if (result != TCL_OK) {
+			if (listPtr != NULL) {
+			    Tcl_DecrRefCount(listPtr);
+			}
+			goto done;
+		    }
+		    match = (objWide == patWide);
+		    break;
+
+		case REAL:
+		    result = Tcl_GetDoubleFromObj(interp,itemPtr, &objDouble);
+		    if (result != TCL_OK) {
+			if (listPtr) {
+			    Tcl_DecrRefCount(listPtr);
+			}
+			goto done;
+		    }
+		    match = (objDouble == patDouble);
+		    break;
+		}
+		break;
+
+	    case GLOB:
+		match = Tcl_StringCaseMatch(TclGetString(itemPtr),
+			patternBytes, noCase);
+		break;
+
+	    case REGEXP:
+		match = Tcl_RegExpExecObj(interp, regexp, itemPtr, 0, 0, 0);
+		if (match < 0) {
+		    Tcl_DecrRefCount(patObj);
+		    if (listPtr != NULL) {
+			Tcl_DecrRefCount(listPtr);
+		    }
+		    result = TCL_ERROR;
+		    goto done;
+		}
+		break;
+	    }
+
+	    /*
+	     * Invert match condition for -not.
+	     */
+
+	    if (negatedMatch) {
+		match = !match;
+	    }
+	    if (!match) {
+		continue;
+	    }
+	    if (!allMatches) {
+		index = i;
+		break;
+	    } else if (inlineReturn) {
+		/*
+		 * Note that these appends are not expected to fail.
+		 */
+
+		if (returnSubindices && (sortInfo.indexc != 0)) {
+		    itemPtr = SelectObjFromSublist(listv[i], &sortInfo);
+		} else {
+		    itemPtr = listv[i];
+		}
+		Tcl_ListObjAppendElement(interp, listPtr, itemPtr);
+	    } else if (returnSubindices) {
+		int j;
+
+		itemPtr = Tcl_NewIntObj(i);
+		for (j=0 ; j<sortInfo.indexc ; j++) {
+		    Tcl_ListObjAppendElement(interp, itemPtr,
+			    Tcl_NewIntObj(sortInfo.indexv[j]));
+		}
+		Tcl_ListObjAppendElement(interp, listPtr, itemPtr);
+	    } else {
+		Tcl_ListObjAppendElement(interp, listPtr, Tcl_NewIntObj(i));
+	    }
+	}
+    }
+
+    /*
+     * Return everything or a single value.
+     */
+
+    if (allMatches) {
+	Tcl_SetObjResult(interp, listPtr);
+    } else if (!inlineReturn) {
+	if (returnSubindices) {
+	    int j;
+
+	    itemPtr = Tcl_NewIntObj(index);
+	    for (j=0 ; j<sortInfo.indexc ; j++) {
+		Tcl_ListObjAppendElement(interp, itemPtr,
+			Tcl_NewIntObj(sortInfo.indexv[j]));
+	    }
+	    Tcl_SetObjResult(interp, itemPtr);
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(index));
+	}
+    } else if (index < 0) {
+	/*
+	 * Is this superfluous? The result should be a blank object by
+	 * default...
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewObj());
+    } else {
+	Tcl_SetObjResult(interp, listv[index]);
+    }
+    result = TCL_OK;
+
+    /*
+     * Cleanup the index list array.
+     */
+
+  done:
+    if (sortInfo.indexc > 1) {
+	TclStackFree(interp, sortInfo.indexv);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LsetObjCmd --
+ *
+ *	This procedure is invoked to process the "lset" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LsetObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument values. */
+{
+    Tcl_Obj *listPtr;		/* Pointer to the list being altered. */
+    Tcl_Obj *finalValuePtr;	/* Value finally assigned to the variable. */
+
+    /*
+     * Check parameter count.
+     */
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"listVar ?index? ?index ...? value");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Look up the list variable's value.
+     */
+
+    listPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, TCL_LEAVE_ERR_MSG);
+    if (listPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Substitute the value in the value. Return either the value or else an
+     * unshared copy of it.
+     */
+
+    if (objc == 4) {
+	finalValuePtr = TclLsetList(interp, listPtr, objv[2], objv[3]);
+    } else {
+	finalValuePtr = TclLsetFlat(interp, listPtr, objc-3, objv+2,
+		objv[objc-1]);
+    }
+
+    /*
+     * If substitution has failed, bail out.
+     */
+
+    if (finalValuePtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Finally, update the variable so that traces fire.
+     */
+
+    listPtr = Tcl_ObjSetVar2(interp, objv[1], NULL, finalValuePtr,
+	    TCL_LEAVE_ERR_MSG);
+    Tcl_DecrRefCount(finalValuePtr);
+    if (listPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Return the new value of the variable as the interpreter result.
+     */
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LsortObjCmd --
+ *
+ *	This procedure is invoked to process the "lsort" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LsortObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument values. */
+{
+    int i, j, index, indices, length, nocase = 0, indexc;
+    int sortMode = SORTMODE_ASCII;
+    int group, groupSize, groupOffset, idx, allocatedIndexVector = 0;
+    Tcl_Obj *resultPtr, *cmdPtr, **listObjPtrs, *listObj, *indexPtr;
+    SortElement *elementArray = NULL, *elementPtr;
+    SortInfo sortInfo;		/* Information about this sort that needs to
+				 * be passed to the comparison function. */
+#   define NUM_LISTS 30
+    SortElement *subList[NUM_LISTS+1];
+				/* This array holds pointers to temporary
+				 * lists built during the merge sort. Element
+				 * i of the array holds a list of length
+				 * 2**i. */
+    static const char *const switches[] = {
+	"-ascii", "-command", "-decreasing", "-dictionary", "-increasing",
+	"-index", "-indices", "-integer", "-nocase", "-real", "-stride",
+	"-unique", NULL
+    };
+    enum Lsort_Switches {
+	LSORT_ASCII, LSORT_COMMAND, LSORT_DECREASING, LSORT_DICTIONARY,
+	LSORT_INCREASING, LSORT_INDEX, LSORT_INDICES, LSORT_INTEGER,
+	LSORT_NOCASE, LSORT_REAL, LSORT_STRIDE, LSORT_UNIQUE
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-option value ...? list");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse arguments to set up the mode for the sort.
+     */
+
+    sortInfo.isIncreasing = 1;
+    sortInfo.sortMode = SORTMODE_ASCII;
+    sortInfo.indexv = NULL;
+    sortInfo.indexc = 0;
+    sortInfo.unique = 0;
+    sortInfo.interp = interp;
+    sortInfo.resultCode = TCL_OK;
+    cmdPtr = NULL;
+    indices = 0;
+    group = 0;
+    groupSize = 1;
+    groupOffset = 0;
+    indexPtr = NULL;
+    for (i = 1; i < objc-1; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], switches, "option", 0,
+		&index) != TCL_OK) {
+	    sortInfo.resultCode = TCL_ERROR;
+	    goto done;
+	}
+	switch ((enum Lsort_Switches) index) {
+	case LSORT_ASCII:
+	    sortInfo.sortMode = SORTMODE_ASCII;
+	    break;
+	case LSORT_COMMAND:
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-command\" option must be followed "
+			"by comparison command", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    sortInfo.sortMode = SORTMODE_COMMAND;
+	    cmdPtr = objv[i+1];
+	    i++;
+	    break;
+	case LSORT_DECREASING:
+	    sortInfo.isIncreasing = 0;
+	    break;
+	case LSORT_DICTIONARY:
+	    sortInfo.sortMode = SORTMODE_DICTIONARY;
+	    break;
+	case LSORT_INCREASING:
+	    sortInfo.isIncreasing = 1;
+	    break;
+	case LSORT_INDEX: {
+	    int indexc, dummy;
+	    Tcl_Obj **indexv;
+
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-index\" option must be followed by list index",
+			-1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    if (TclListObjGetElements(interp, objv[i+1], &indexc,
+		    &indexv) != TCL_OK) {
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+
+	    /*
+	     * Check each of the indices for syntactic correctness. Note that
+	     * we do not store the converted values here because we do not
+	     * know if this is the only -index option yet and so we can't
+	     * allocate any space; that happens after the scan through all the
+	     * options is done.
+	     */
+
+	    for (j=0 ; j<indexc ; j++) {
+		if (TclGetIntForIndexM(interp, indexv[j], SORTIDX_END,
+			&dummy) != TCL_OK) {
+		    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			    "\n    (-index option item number %d)", j));
+		    sortInfo.resultCode = TCL_ERROR;
+		    goto done;
+		}
+	    }
+	    indexPtr = objv[i+1];
+	    i++;
+	    break;
+	}
+	case LSORT_INTEGER:
+	    sortInfo.sortMode = SORTMODE_INTEGER;
+	    break;
+	case LSORT_NOCASE:
+	    nocase = 1;
+	    break;
+	case LSORT_REAL:
+	    sortInfo.sortMode = SORTMODE_REAL;
+	    break;
+	case LSORT_UNIQUE:
+	    sortInfo.unique = 1;
+	    break;
+	case LSORT_INDICES:
+	    indices = 1;
+	    break;
+	case LSORT_STRIDE:
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-stride\" option must be "
+			"followed by stride length", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    if (Tcl_GetIntFromObj(interp, objv[i+1], &groupSize) != TCL_OK) {
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    if (groupSize < 2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"stride length must be at least 2", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSORT",
+			"BADSTRIDE", NULL);
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    group = 1;
+	    i++;
+	    break;
+	}
+    }
+    if (nocase && (sortInfo.sortMode == SORTMODE_ASCII)) {
+	sortInfo.sortMode = SORTMODE_ASCII_NC;
+    }
+
+    /*
+     * Now extract the -index list for real, if present. No failures are
+     * expected here; the values are all of the right type or convertible to
+     * it.
+     */
+
+    if (indexPtr) {
+	Tcl_Obj **indexv;
+
+	TclListObjGetElements(interp, indexPtr, &sortInfo.indexc, &indexv);
+	switch (sortInfo.indexc) {
+	case 0:
+	    sortInfo.indexv = NULL;
+	    break;
+	case 1:
+	    sortInfo.indexv = &sortInfo.singleIndex;
+	    break;
+	default:
+	    sortInfo.indexv =
+		    TclStackAlloc(interp, sizeof(int) * sortInfo.indexc);
+	    allocatedIndexVector = 1;	/* Cannot use indexc field, as it
+					 * might be decreased by 1 later. */
+	}
+	for (j=0 ; j<sortInfo.indexc ; j++) {
+	    TclGetIntForIndexM(interp, indexv[j], SORTIDX_END,
+		    &sortInfo.indexv[j]);
+	}
+    }
+
+    listObj = objv[objc-1];
+
+    if (sortInfo.sortMode == SORTMODE_COMMAND) {
+	Tcl_Obj *newCommandPtr, *newObjPtr;
+
+	/*
+	 * When sorting using a command, we are reentrant and therefore might
+	 * have the representation of the list being sorted shimmered out from
+	 * underneath our feet. Take a copy (cheap) to prevent this. [Bug
+	 * 1675116]
+	 */
+
+	listObj = TclListObjCopy(interp, listObj);
+	if (listObj == NULL) {
+	    sortInfo.resultCode = TCL_ERROR;
+	    goto done;
+	}
+
+	/*
+	 * The existing command is a list. We want to flatten it, append two
+	 * dummy arguments on the end, and replace these arguments later.
+	 */
+
+	newCommandPtr = Tcl_DuplicateObj(cmdPtr);
+	TclNewObj(newObjPtr);
+	Tcl_IncrRefCount(newCommandPtr);
+	if (Tcl_ListObjAppendElement(interp, newCommandPtr, newObjPtr)
+		!= TCL_OK) {
+	    TclDecrRefCount(newCommandPtr);
+	    TclDecrRefCount(listObj);
+	    Tcl_IncrRefCount(newObjPtr);
+	    TclDecrRefCount(newObjPtr);
+	    sortInfo.resultCode = TCL_ERROR;
+	    goto done;
+	}
+	Tcl_ListObjAppendElement(interp, newCommandPtr, Tcl_NewObj());
+	sortInfo.compareCmdPtr = newCommandPtr;
+    }
+
+    sortInfo.resultCode = TclListObjGetElements(interp, listObj,
+	    &length, &listObjPtrs);
+    if (sortInfo.resultCode != TCL_OK || length <= 0) {
+	goto done;
+    }
+
+    /*
+     * Check for sanity when grouping elements of the overall list together
+     * because of the -stride option. [TIP #326]
+     */
+
+    if (group) {
+	if (length % groupSize) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "list size must be a multiple of the stride length",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSORT", "BADSTRIDE",
+		    NULL);
+	    sortInfo.resultCode = TCL_ERROR;
+	    goto done;
+	}
+	length = length / groupSize;
+	if (sortInfo.indexc > 0) {
+	    /*
+	     * Use the first value in the list supplied to -index as the
+	     * offset of the element within each group by which to sort.
+	     */
+
+	    groupOffset = sortInfo.indexv[0];
+	    if (groupOffset <= SORTIDX_END) {
+		groupOffset = (groupOffset - SORTIDX_END) + groupSize - 1;
+	    }
+	    if (groupOffset < 0 || groupOffset >= groupSize) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"when used with \"-stride\", the leading \"-index\""
+			" value must be within the group", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSORT",
+			"BADINDEX", NULL);
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    if (sortInfo.indexc == 1) {
+		sortInfo.indexc = 0;
+		sortInfo.indexv = NULL;
+	    } else {
+		sortInfo.indexc--;
+
+		/*
+		 * Do not shrink the actual memory block used; that doesn't
+		 * work with TclStackAlloc-allocated memory. [Bug 2918962]
+		 */
+
+		for (i = 0; i < sortInfo.indexc; i++) {
+		    sortInfo.indexv[i] = sortInfo.indexv[i+1];
+		}
+	    }
+	}
+    }
+
+    sortInfo.numElements = length;
+
+    indexc = sortInfo.indexc;
+    sortMode = sortInfo.sortMode;
+    if ((sortMode == SORTMODE_ASCII_NC)
+	    || (sortMode == SORTMODE_DICTIONARY)) {
+	/*
+	 * For this function's purpose all string-based modes are equivalent
+	 */
+
+	sortMode = SORTMODE_ASCII;
+    }
+
+    /*
+     * Initialize the sublists. After the following loop, subList[i] will
+     * contain a sorted sublist of length 2**i. Use one extra subList at the
+     * end, always at NULL, to indicate the end of the lists.
+     */
+
+    for (j=0 ; j<=NUM_LISTS ; j++) {
+	subList[j] = NULL;
+    }
+
+    /*
+     * The following loop creates a SortElement for each list element and
+     * begins sorting it into the sublists as it appears.
+     */
+
+    elementArray = ckalloc(length * sizeof(SortElement));
+
+    for (i=0; i < length; i++){
+	idx = groupSize * i + groupOffset;
+	if (indexc) {
+	    /*
+	     * If this is an indexed sort, retrieve the corresponding element
+	     */
+	    indexPtr = SelectObjFromSublist(listObjPtrs[idx], &sortInfo);
+	    if (sortInfo.resultCode != TCL_OK) {
+		goto done;
+	    }
+	} else {
+	    indexPtr = listObjPtrs[idx];
+	}
+
+	/*
+	 * Determine the "value" of this object for sorting purposes
+	 */
+
+	if (sortMode == SORTMODE_ASCII) {
+	    elementArray[i].collationKey.strValuePtr = TclGetString(indexPtr);
+	} else if (sortMode == SORTMODE_INTEGER) {
+	    Tcl_WideInt a;
+
+	    if (TclGetWideIntFromObj(sortInfo.interp, indexPtr, &a) != TCL_OK) {
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    elementArray[i].collationKey.wideValue = a;
+	} else if (sortMode == SORTMODE_REAL) {
+	    double a;
+
+	    if (Tcl_GetDoubleFromObj(sortInfo.interp, indexPtr,
+		    &a) != TCL_OK) {
+		sortInfo.resultCode = TCL_ERROR;
+		goto done;
+	    }
+	    elementArray[i].collationKey.doubleValue = a;
+	} else {
+	    elementArray[i].collationKey.objValuePtr = indexPtr;
+	}
+
+	/*
+	 * Determine the representation of this element in the result: either
+	 * the objPtr itself, or its index in the original list.
+	 */
+
+	if (indices || group) {
+	    elementArray[i].payload.index = idx;
+	} else {
+	    elementArray[i].payload.objPtr = listObjPtrs[idx];
+	}
+
+	/*
+	 * Merge this element in the pre-existing sublists (and merge together
+	 * sublists when we have two of the same size).
+	 */
+
+	elementArray[i].nextPtr = NULL;
+	elementPtr = &elementArray[i];
+	for (j=0 ; subList[j] ; j++) {
+	    elementPtr = MergeLists(subList[j], elementPtr, &sortInfo);
+	    subList[j] = NULL;
+	}
+	if (j >= NUM_LISTS) {
+	    j = NUM_LISTS-1;
+	}
+	subList[j] = elementPtr;
+    }
+
+    /*
+     * Merge all sublists
+     */
+
+    elementPtr = subList[0];
+    for (j=1 ; j<NUM_LISTS ; j++) {
+	elementPtr = MergeLists(subList[j], elementPtr, &sortInfo);
+    }
+
+    /*
+     * Now store the sorted elements in the result list.
+     */
+
+    if (sortInfo.resultCode == TCL_OK) {
+	List *listRepPtr;
+	Tcl_Obj **newArray, *objPtr;
+
+	resultPtr = Tcl_NewListObj(sortInfo.numElements * groupSize, NULL);
+	listRepPtr = ListRepPtr(resultPtr);
+	newArray = &listRepPtr->elements;
+	if (group) {
+	    for (i=0; elementPtr!=NULL ; elementPtr=elementPtr->nextPtr) {
+		idx = elementPtr->payload.index;
+		for (j = 0; j < groupSize; j++) {
+		    if (indices) {
+			objPtr = Tcl_NewIntObj(idx + j - groupOffset);
+			newArray[i++] = objPtr;
+			Tcl_IncrRefCount(objPtr);
+		    } else {
+			objPtr = listObjPtrs[idx + j - groupOffset];
+			newArray[i++] = objPtr;
+			Tcl_IncrRefCount(objPtr);
+		    }
+		}
+	    }
+	} else if (indices) {
+	    for (i=0; elementPtr != NULL ; elementPtr = elementPtr->nextPtr) {
+		objPtr = Tcl_NewIntObj(elementPtr->payload.index);
+		newArray[i++] = objPtr;
+		Tcl_IncrRefCount(objPtr);
+	    }
+	} else {
+	    for (i=0; elementPtr != NULL ; elementPtr = elementPtr->nextPtr) {
+		objPtr = elementPtr->payload.objPtr;
+		newArray[i++] = objPtr;
+		Tcl_IncrRefCount(objPtr);
+	    }
+	}
+	listRepPtr->elemCount = i;
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+
+  done:
+    if (sortMode == SORTMODE_COMMAND) {
+	TclDecrRefCount(sortInfo.compareCmdPtr);
+	TclDecrRefCount(listObj);
+	sortInfo.compareCmdPtr = NULL;
+    }
+    if (allocatedIndexVector) {
+	TclStackFree(interp, sortInfo.indexv);
+    }
+    if (elementArray) {
+	ckfree(elementArray);
+    }
+    return sortInfo.resultCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MergeLists -
+ *
+ *	This procedure combines two sorted lists of SortElement structures
+ *	into a single sorted list.
+ *
+ * Results:
+ *	The unified list of SortElement structures.
+ *
+ * Side effects:
+ *	If infoPtr->unique is set then infoPtr->numElements may be updated.
+ *	Possibly others, if a user-defined comparison command does something
+ *	weird.
+ *
+ * Note:
+ *	If infoPtr->unique is set, the merge assumes that there are no
+ *	"repeated" elements in each of the left and right lists. In that case,
+ *	if any element of the left list is equivalent to one in the right list
+ *	it is omitted from the merged list.
+ *
+ *	This simplified mechanism works because of the special way our
+ *	MergeSort creates the sublists to be merged and will fail to eliminate
+ *	all repeats in the general case where they are already present in
+ *	either the left or right list. A general code would need to skip
+ *	adjacent initial repeats in the left and right lists before comparing
+ *	their initial elements, at each step.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static SortElement *
+MergeLists(
+    SortElement *leftPtr,	/* First list to be merged; may be NULL. */
+    SortElement *rightPtr,	/* Second list to be merged; may be NULL. */
+    SortInfo *infoPtr)		/* Information needed by the comparison
+				 * operator. */
+{
+    SortElement *headPtr, *tailPtr;
+    int cmp;
+
+    if (leftPtr == NULL) {
+	return rightPtr;
+    }
+    if (rightPtr == NULL) {
+	return leftPtr;
+    }
+    cmp = SortCompare(leftPtr, rightPtr, infoPtr);
+    if (cmp > 0 || (cmp == 0 && infoPtr->unique)) {
+	if (cmp == 0) {
+	    infoPtr->numElements--;
+	    leftPtr = leftPtr->nextPtr;
+	}
+	tailPtr = rightPtr;
+	rightPtr = rightPtr->nextPtr;
+    } else {
+	tailPtr = leftPtr;
+	leftPtr = leftPtr->nextPtr;
+    }
+    headPtr = tailPtr;
+    if (!infoPtr->unique) {
+	while ((leftPtr != NULL) && (rightPtr != NULL)) {
+	    cmp = SortCompare(leftPtr, rightPtr, infoPtr);
+	    if (cmp > 0) {
+		tailPtr->nextPtr = rightPtr;
+		tailPtr = rightPtr;
+		rightPtr = rightPtr->nextPtr;
+	    } else {
+		tailPtr->nextPtr = leftPtr;
+		tailPtr = leftPtr;
+		leftPtr = leftPtr->nextPtr;
+	    }
+	}
+    } else {
+	while ((leftPtr != NULL) && (rightPtr != NULL)) {
+	    cmp = SortCompare(leftPtr, rightPtr, infoPtr);
+	    if (cmp >= 0) {
+		if (cmp == 0) {
+		    infoPtr->numElements--;
+		    leftPtr = leftPtr->nextPtr;
+		}
+		tailPtr->nextPtr = rightPtr;
+		tailPtr = rightPtr;
+		rightPtr = rightPtr->nextPtr;
+	    } else {
+		tailPtr->nextPtr = leftPtr;
+		tailPtr = leftPtr;
+		leftPtr = leftPtr->nextPtr;
+	    }
+	}
+    }
+    if (leftPtr != NULL) {
+	tailPtr->nextPtr = leftPtr;
+    } else {
+	tailPtr->nextPtr = rightPtr;
+    }
+    return headPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SortCompare --
+ *
+ *	This procedure is invoked by MergeLists to determine the proper
+ *	ordering between two elements.
+ *
+ * Results:
+ *	A negative results means the the first element comes before the
+ *	second, and a positive results means that the second element should
+ *	come first. A result of zero means the two elements are equal and it
+ *	doesn't matter which comes first.
+ *
+ * Side effects:
+ *	None, unless a user-defined comparison command does something weird.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SortCompare(
+    SortElement *elemPtr1, SortElement *elemPtr2,
+				/* Values to be compared. */
+    SortInfo *infoPtr)		/* Information passed from the top-level
+				 * "lsort" command. */
+{
+    int order = 0;
+
+    if (infoPtr->sortMode == SORTMODE_ASCII) {
+	order = strcmp(elemPtr1->collationKey.strValuePtr,
+		elemPtr2->collationKey.strValuePtr);
+    } else if (infoPtr->sortMode == SORTMODE_ASCII_NC) {
+	order = TclUtfCasecmp(elemPtr1->collationKey.strValuePtr,
+		elemPtr2->collationKey.strValuePtr);
+    } else if (infoPtr->sortMode == SORTMODE_DICTIONARY) {
+	order = DictionaryCompare(elemPtr1->collationKey.strValuePtr,
+		elemPtr2->collationKey.strValuePtr);
+    } else if (infoPtr->sortMode == SORTMODE_INTEGER) {
+	Tcl_WideInt a, b;
+
+	a = elemPtr1->collationKey.wideValue;
+	b = elemPtr2->collationKey.wideValue;
+	order = ((a >= b) - (a <= b));
+    } else if (infoPtr->sortMode == SORTMODE_REAL) {
+	double a, b;
+
+	a = elemPtr1->collationKey.doubleValue;
+	b = elemPtr2->collationKey.doubleValue;
+	order = ((a >= b) - (a <= b));
+    } else {
+	Tcl_Obj **objv, *paramObjv[2];
+	int objc;
+	Tcl_Obj *objPtr1, *objPtr2;
+
+	if (infoPtr->resultCode != TCL_OK) {
+	    /*
+	     * Once an error has occurred, skip any future comparisons so as
+	     * to preserve the error message in sortInterp->result.
+	     */
+
+	    return 0;
+	}
+
+
+	objPtr1 = elemPtr1->collationKey.objValuePtr;
+	objPtr2 = elemPtr2->collationKey.objValuePtr;
+
+	paramObjv[0] = objPtr1;
+	paramObjv[1] = objPtr2;
+
+	/*
+	 * We made space in the command list for the two things to compare.
+	 * Replace them and evaluate the result.
+	 */
+
+	TclListObjLength(infoPtr->interp, infoPtr->compareCmdPtr, &objc);
+	Tcl_ListObjReplace(infoPtr->interp, infoPtr->compareCmdPtr, objc - 2,
+		2, 2, paramObjv);
+	TclListObjGetElements(infoPtr->interp, infoPtr->compareCmdPtr,
+		&objc, &objv);
+
+	infoPtr->resultCode = Tcl_EvalObjv(infoPtr->interp, objc, objv, 0);
+
+	if (infoPtr->resultCode != TCL_OK) {
+	    Tcl_AddErrorInfo(infoPtr->interp, "\n    (-compare command)");
+	    return 0;
+	}
+
+	/*
+	 * Parse the result of the command.
+	 */
+
+	if (TclGetIntFromObj(infoPtr->interp,
+		Tcl_GetObjResult(infoPtr->interp), &order) != TCL_OK) {
+	    Tcl_SetObjResult(infoPtr->interp, Tcl_NewStringObj(
+		    "-compare command returned non-integer result", -1));
+	    Tcl_SetErrorCode(infoPtr->interp, "TCL", "OPERATION", "LSORT",
+		    "COMPARISONFAILED", NULL);
+	    infoPtr->resultCode = TCL_ERROR;
+	    return 0;
+	}
+    }
+    if (!infoPtr->isIncreasing) {
+	order = -order;
+    }
+    return order;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictionaryCompare
+ *
+ *	This function compares two strings as if they were being used in an
+ *	index or card catalog. The case of alphabetic characters is ignored,
+ *	except to break ties. Thus "B" comes before "b" but after "a". Also,
+ *	integers embedded in the strings compare in numerical order. In other
+ *	words, "x10y" comes after "x9y", not * before it as it would when
+ *	using strcmp().
+ *
+ * Results:
+ *	A negative result means that the first element comes before the
+ *	second, and a positive result means that the second element should
+ *	come first. A result of zero means the two elements are equal and it
+ *	doesn't matter which comes first.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictionaryCompare(
+    const char *left, const char *right)	/* The strings to compare. */
+{
+    Tcl_UniChar uniLeft = 0, uniRight = 0, uniLeftLower, uniRightLower;
+    int diff, zeros;
+    int secondaryDiff = 0;
+
+    while (1) {
+	if (isdigit(UCHAR(*right))		/* INTL: digit */
+		&& isdigit(UCHAR(*left))) {	/* INTL: digit */
+	    /*
+	     * There are decimal numbers embedded in the two strings. Compare
+	     * them as numbers, rather than strings. If one number has more
+	     * leading zeros than the other, the number with more leading
+	     * zeros sorts later, but only as a secondary choice.
+	     */
+
+	    zeros = 0;
+	    while ((*right == '0') && isdigit(UCHAR(right[1]))) {
+		right++;
+		zeros--;
+	    }
+	    while ((*left == '0') && isdigit(UCHAR(left[1]))) {
+		left++;
+		zeros++;
+	    }
+	    if (secondaryDiff == 0) {
+		secondaryDiff = zeros;
+	    }
+
+	    /*
+	     * The code below compares the numbers in the two strings without
+	     * ever converting them to integers. It does this by first
+	     * comparing the lengths of the numbers and then comparing the
+	     * digit values.
+	     */
+
+	    diff = 0;
+	    while (1) {
+		if (diff == 0) {
+		    diff = UCHAR(*left) - UCHAR(*right);
+		}
+		right++;
+		left++;
+		if (!isdigit(UCHAR(*right))) {		/* INTL: digit */
+		    if (isdigit(UCHAR(*left))) {	/* INTL: digit */
+			return 1;
+		    } else {
+			/*
+			 * The two numbers have the same length. See if their
+			 * values are different.
+			 */
+
+			if (diff != 0) {
+			    return diff;
+			}
+			break;
+		    }
+		} else if (!isdigit(UCHAR(*left))) {	/* INTL: digit */
+		    return -1;
+		}
+	    }
+	    continue;
+	}
+
+	/*
+	 * Convert character to Unicode for comparison purposes. If either
+	 * string is at the terminating null, do a byte-wise comparison and
+	 * bail out immediately.
+	 */
+
+	if ((*left != '\0') && (*right != '\0')) {
+	    left += TclUtfToUniChar(left, &uniLeft);
+	    right += TclUtfToUniChar(right, &uniRight);
+
+	    /*
+	     * Convert both chars to lower for the comparison, because
+	     * dictionary sorts are case insensitve. Covert to lower, not
+	     * upper, so chars between Z and a will sort before A (where most
+	     * other interesting punctuations occur).
+	     */
+
+	    uniLeftLower = Tcl_UniCharToLower(uniLeft);
+	    uniRightLower = Tcl_UniCharToLower(uniRight);
+	} else {
+	    diff = UCHAR(*left) - UCHAR(*right);
+	    break;
+	}
+
+	diff = uniLeftLower - uniRightLower;
+	if (diff) {
+	    return diff;
+	}
+	if (secondaryDiff == 0) {
+	    if (Tcl_UniCharIsUpper(uniLeft) && Tcl_UniCharIsLower(uniRight)) {
+		secondaryDiff = -1;
+	    } else if (Tcl_UniCharIsUpper(uniRight)
+		    && Tcl_UniCharIsLower(uniLeft)) {
+		secondaryDiff = 1;
+	    }
+	}
+    }
+    if (diff == 0) {
+	diff = secondaryDiff;
+    }
+    return diff;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SelectObjFromSublist --
+ *
+ *	This procedure is invoked from lsearch and SortCompare. It is used for
+ *	implementing the -index option, for the lsort and lsearch commands.
+ *
+ * Results:
+ *	Returns NULL if a failure occurs, and sets the result in the infoPtr.
+ *	Otherwise returns the Tcl_Obj* to the item.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Note:
+ *	No reference counting is done, as the result is only used internally
+ *	and never passed directly to user code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+SelectObjFromSublist(
+    Tcl_Obj *objPtr,		/* Obj to select sublist from. */
+    SortInfo *infoPtr)		/* Information passed from the top-level
+				 * "lsearch" or "lsort" command. */
+{
+    int i;
+
+    /*
+     * Quick check for case when no "-index" option is there.
+     */
+
+    if (infoPtr->indexc == 0) {
+	return objPtr;
+    }
+
+    /*
+     * Iterate over the indices, traversing through the nested sublists as we
+     * go.
+     */
+
+    for (i=0 ; i<infoPtr->indexc ; i++) {
+	int listLen, index;
+	Tcl_Obj *currentObj;
+
+	if (TclListObjLength(infoPtr->interp, objPtr, &listLen) != TCL_OK) {
+	    infoPtr->resultCode = TCL_ERROR;
+	    return NULL;
+	}
+	index = infoPtr->indexv[i];
+
+	/*
+	 * Adjust for end-based indexing.
+	 */
+
+	if (index < SORTIDX_NONE) {
+	    index += listLen + 1;
+	}
+
+	if (Tcl_ListObjIndex(infoPtr->interp, objPtr, index,
+		&currentObj) != TCL_OK) {
+	    infoPtr->resultCode = TCL_ERROR;
+	    return NULL;
+	}
+	if (currentObj == NULL) {
+	    Tcl_SetObjResult(infoPtr->interp, Tcl_ObjPrintf(
+		    "element %d missing from sublist \"%s\"",
+		    index, TclGetString(objPtr)));
+	    Tcl_SetErrorCode(infoPtr->interp, "TCL", "OPERATION", "LSORT",
+		    "INDEXFAILED", NULL);
+	    infoPtr->resultCode = TCL_ERROR;
+	    return NULL;
+	}
+	objPtr = currentObj;
+    }
+    return objPtr;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * End:
+ */
diff --git a/generic/tclCmdMZ.c b/generic/tclCmdMZ.c
new file mode 100644
index 0000000..2195aa1
--- /dev/null
+++ b/generic/tclCmdMZ.c
@@ -0,0 +1,4855 @@
+/*
+ * tclCmdMZ.c --
+ *
+ *	This file contains the top-level command routines for most of the Tcl
+ *	built-in commands whose names begin with the letters M to Z. It
+ *	contains only commands in the generic core (i.e. those that don't
+ *	depend much upon UNIX facilities).
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 Scriptics Corporation.
+ * Copyright (c) 2002 ActiveState Corporation.
+ * Copyright (c) 2003-2009 Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclRegexp.h"
+#include "tclStringTrim.h"
+
+static inline Tcl_Obj *	During(Tcl_Interp *interp, int resultCode,
+			    Tcl_Obj *oldOptions, Tcl_Obj *errorInfo);
+static Tcl_NRPostProc	SwitchPostProc;
+static Tcl_NRPostProc	TryPostBody;
+static Tcl_NRPostProc	TryPostFinal;
+static Tcl_NRPostProc	TryPostHandler;
+static int		UniCharIsAscii(int character);
+static int		UniCharIsHexDigit(int character);
+
+/*
+ * Default set of characters to trim in [string trim] and friends. This is a
+ * UTF-8 literal string containing all Unicode space characters [TIP #413]
+ */
+
+const char tclDefaultTrimSet[] =
+	"\x09\x0a\x0b\x0c\x0d " /* ASCII */
+	"\xc0\x80" /*     nul (U+0000) */
+	"\xc2\x85" /*     next line (U+0085) */
+	"\xc2\xa0" /*     non-breaking space (U+00a0) */
+	"\xe1\x9a\x80" /* ogham space mark (U+1680) */
+	"\xe1\xa0\x8e" /* mongolian vowel separator (U+180e) */
+	"\xe2\x80\x80" /* en quad (U+2000) */
+	"\xe2\x80\x81" /* em quad (U+2001) */
+	"\xe2\x80\x82" /* en space (U+2002) */
+	"\xe2\x80\x83" /* em space (U+2003) */
+	"\xe2\x80\x84" /* three-per-em space (U+2004) */
+	"\xe2\x80\x85" /* four-per-em space (U+2005) */
+	"\xe2\x80\x86" /* six-per-em space (U+2006) */
+	"\xe2\x80\x87" /* figure space (U+2007) */
+	"\xe2\x80\x88" /* punctuation space (U+2008) */
+	"\xe2\x80\x89" /* thin space (U+2009) */
+	"\xe2\x80\x8a" /* hair space (U+200a) */
+	"\xe2\x80\x8b" /* zero width space (U+200b) */
+	"\xe2\x80\xa8" /* line separator (U+2028) */
+	"\xe2\x80\xa9" /* paragraph separator (U+2029) */
+	"\xe2\x80\xaf" /* narrow no-break space (U+202f) */
+	"\xe2\x81\x9f" /* medium mathematical space (U+205f) */
+	"\xe2\x81\xa0" /* word joiner (U+2060) */
+	"\xe3\x80\x80" /* ideographic space (U+3000) */
+	"\xef\xbb\xbf" /* zero width no-break space (U+feff) */
+;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PwdObjCmd --
+ *
+ *	This procedure is invoked to process the "pwd" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_PwdObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *retVal;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    retVal = Tcl_FSGetCwd(interp);
+    if (retVal == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, retVal);
+    Tcl_DecrRefCount(retVal);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegexpObjCmd --
+ *
+ *	This procedure is invoked to process the "regexp" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegexpObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int i, indices, match, about, offset, all, doinline, numMatchesSaved;
+    int cflags, eflags, stringLength, matchLength;
+    Tcl_RegExp regExpr;
+    Tcl_Obj *objPtr, *startIndex = NULL, *resultPtr = NULL;
+    Tcl_RegExpInfo info;
+    static const char *const options[] = {
+	"-all",		"-about",	"-indices",	"-inline",
+	"-expanded",	"-line",	"-linestop",	"-lineanchor",
+	"-nocase",	"-start",	"--",		NULL
+    };
+    enum options {
+	REGEXP_ALL,	REGEXP_ABOUT,	REGEXP_INDICES,	REGEXP_INLINE,
+	REGEXP_EXPANDED,REGEXP_LINE,	REGEXP_LINESTOP,REGEXP_LINEANCHOR,
+	REGEXP_NOCASE,	REGEXP_START,	REGEXP_LAST
+    };
+
+    indices = 0;
+    about = 0;
+    cflags = TCL_REG_ADVANCED;
+    offset = 0;
+    all = 0;
+    doinline = 0;
+
+    for (i = 1; i < objc; i++) {
+	const char *name;
+	int index;
+
+	name = TclGetString(objv[i]);
+	if (name[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", TCL_EXACT,
+		&index) != TCL_OK) {
+	    goto optionError;
+	}
+	switch ((enum options) index) {
+	case REGEXP_ALL:
+	    all = 1;
+	    break;
+	case REGEXP_INDICES:
+	    indices = 1;
+	    break;
+	case REGEXP_INLINE:
+	    doinline = 1;
+	    break;
+	case REGEXP_NOCASE:
+	    cflags |= TCL_REG_NOCASE;
+	    break;
+	case REGEXP_ABOUT:
+	    about = 1;
+	    break;
+	case REGEXP_EXPANDED:
+	    cflags |= TCL_REG_EXPANDED;
+	    break;
+	case REGEXP_LINE:
+	    cflags |= TCL_REG_NEWLINE;
+	    break;
+	case REGEXP_LINESTOP:
+	    cflags |= TCL_REG_NLSTOP;
+	    break;
+	case REGEXP_LINEANCHOR:
+	    cflags |= TCL_REG_NLANCH;
+	    break;
+	case REGEXP_START: {
+	    int temp;
+	    if (++i >= objc) {
+		goto endOfForLoop;
+	    }
+	    if (TclGetIntForIndexM(interp, objv[i], 0, &temp) != TCL_OK) {
+		goto optionError;
+	    }
+	    if (startIndex) {
+		Tcl_DecrRefCount(startIndex);
+	    }
+	    startIndex = objv[i];
+	    Tcl_IncrRefCount(startIndex);
+	    break;
+	}
+	case REGEXP_LAST:
+	    i++;
+	    goto endOfForLoop;
+	}
+    }
+
+  endOfForLoop:
+    if ((objc - i) < (2 - about)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-option ...? exp string ?matchVar? ?subMatchVar ...?");
+	goto optionError;
+    }
+    objc -= i;
+    objv += i;
+
+    /*
+     * Check if the user requested -inline, but specified match variables; a
+     * no-no.
+     */
+
+    if (doinline && ((objc - 2) != 0)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"regexp match variables not allowed when using -inline", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "REGEXP",
+		"MIX_VAR_INLINE", NULL);
+	goto optionError;
+    }
+
+    /*
+     * Handle the odd about case separately.
+     */
+
+    if (about) {
+	regExpr = Tcl_GetRegExpFromObj(interp, objv[0], cflags);
+	if ((regExpr == NULL) || (TclRegAbout(interp, regExpr) < 0)) {
+	optionError:
+	    if (startIndex) {
+		Tcl_DecrRefCount(startIndex);
+	    }
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * Get the length of the string that we are matching against so we can do
+     * the termination test for -all matches. Do this before getting the
+     * regexp to avoid shimmering problems.
+     */
+
+    objPtr = objv[1];
+    stringLength = Tcl_GetCharLength(objPtr);
+
+    if (startIndex) {
+	TclGetIntForIndexM(NULL, startIndex, stringLength, &offset);
+	Tcl_DecrRefCount(startIndex);
+	if (offset < 0) {
+	    offset = 0;
+	}
+    }
+
+    regExpr = Tcl_GetRegExpFromObj(interp, objv[0], cflags);
+    if (regExpr == NULL) {
+	return TCL_ERROR;
+    }
+
+    objc -= 2;
+    objv += 2;
+
+    if (doinline) {
+	/*
+	 * Save all the subexpressions, as we will return them as a list
+	 */
+
+	numMatchesSaved = -1;
+    } else {
+	/*
+	 * Save only enough subexpressions for matches we want to keep, expect
+	 * in the case of -all, where we need to keep at least one to know
+	 * where to move the offset.
+	 */
+
+	numMatchesSaved = (objc == 0) ? all : objc;
+    }
+
+    /*
+     * The following loop is to handle multiple matches within the same source
+     * string; each iteration handles one match. If "-all" hasn't been
+     * specified then the loop body only gets executed once. We terminate the
+     * loop when the starting offset is past the end of the string.
+     */
+
+    while (1) {
+	/*
+	 * Pass either 0 or TCL_REG_NOTBOL in the eflags. Passing
+	 * TCL_REG_NOTBOL indicates that the character at offset should not be
+	 * considered the start of the line. If for example the pattern {^} is
+	 * passed and -start is positive, then the pattern will not match the
+	 * start of the string unless the previous character is a newline.
+	 */
+
+	if (offset == 0) {
+	    eflags = 0;
+	} else if (offset > stringLength) {
+	    eflags = TCL_REG_NOTBOL;
+	} else if (Tcl_GetUniChar(objPtr, offset-1) == (Tcl_UniChar)'\n') {
+	    eflags = 0;
+	} else {
+	    eflags = TCL_REG_NOTBOL;
+	}
+
+	match = Tcl_RegExpExecObj(interp, regExpr, objPtr, offset,
+		numMatchesSaved, eflags);
+	if (match < 0) {
+	    return TCL_ERROR;
+	}
+
+	if (match == 0) {
+	    /*
+	     * We want to set the value of the interpreter result only when
+	     * this is the first time through the loop.
+	     */
+
+	    if (all <= 1) {
+		/*
+		 * If inlining, the interpreter's object result remains an
+		 * empty list, otherwise set it to an integer object w/ value
+		 * 0.
+		 */
+
+		if (!doinline) {
+		    Tcl_SetObjResult(interp, Tcl_NewIntObj(0));
+		}
+		return TCL_OK;
+	    }
+	    break;
+	}
+
+	/*
+	 * If additional variable names have been specified, return index
+	 * information in those variables.
+	 */
+
+	Tcl_RegExpGetInfo(regExpr, &info);
+	if (doinline) {
+	    /*
+	     * It's the number of substitutions, plus one for the matchVar at
+	     * index 0
+	     */
+
+	    objc = info.nsubs + 1;
+	    if (all <= 1) {
+		resultPtr = Tcl_NewObj();
+	    }
+	}
+	for (i = 0; i < objc; i++) {
+	    Tcl_Obj *newPtr;
+
+	    if (indices) {
+		int start, end;
+		Tcl_Obj *objs[2];
+
+		/*
+		 * Only adjust the match area if there was a match for that
+		 * area. (Scriptics Bug 4391/SF Bug #219232)
+		 */
+
+		if (i <= info.nsubs && info.matches[i].start >= 0) {
+		    start = offset + info.matches[i].start;
+		    end = offset + info.matches[i].end;
+
+		    /*
+		     * Adjust index so it refers to the last character in the
+		     * match instead of the first character after the match.
+		     */
+
+		    if (end >= offset) {
+			end--;
+		    }
+		} else {
+		    start = -1;
+		    end = -1;
+		}
+
+		objs[0] = Tcl_NewLongObj(start);
+		objs[1] = Tcl_NewLongObj(end);
+
+		newPtr = Tcl_NewListObj(2, objs);
+	    } else {
+		if (i <= info.nsubs) {
+		    newPtr = Tcl_GetRange(objPtr,
+			    offset + info.matches[i].start,
+			    offset + info.matches[i].end - 1);
+		} else {
+		    newPtr = Tcl_NewObj();
+		}
+	    }
+	    if (doinline) {
+		if (Tcl_ListObjAppendElement(interp, resultPtr, newPtr)
+			!= TCL_OK) {
+		    Tcl_DecrRefCount(newPtr);
+		    Tcl_DecrRefCount(resultPtr);
+		    return TCL_ERROR;
+		}
+	    } else {
+		if (Tcl_ObjSetVar2(interp, objv[i], NULL, newPtr,
+			TCL_LEAVE_ERR_MSG) == NULL) {
+		    return TCL_ERROR;
+		}
+	    }
+	}
+
+	if (all == 0) {
+	    break;
+	}
+
+	/*
+	 * Adjust the offset to the character just after the last one in the
+	 * matchVar and increment all to count how many times we are making a
+	 * match. We always increment the offset by at least one to prevent
+	 * endless looping (as in the case: regexp -all {a*} a). Otherwise,
+	 * when we match the NULL string at the end of the input string, we
+	 * will loop indefinately (because the length of the match is 0, so
+	 * offset never changes).
+	 */
+
+	matchLength = (info.matches[0].end - info.matches[0].start);
+
+	offset += info.matches[0].end;
+
+	/*
+	 * A match of length zero could happen for {^} {$} or {.*} and in
+	 * these cases we always want to bump the index up one.
+	 */
+
+	if (matchLength == 0) {
+	    offset++;
+	}
+	all++;
+	if (offset >= stringLength) {
+	    break;
+	}
+    }
+
+    /*
+     * Set the interpreter's object result to an integer object with value 1
+     * if -all wasn't specified, otherwise it's all-1 (the number of times
+     * through the while - 1).
+     */
+
+    if (doinline) {
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(all ? all-1 : 1));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegsubObjCmd --
+ *
+ *	This procedure is invoked to process the "regsub" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegsubObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int idx, result, cflags, all, wlen, wsublen, numMatches, offset;
+    int start, end, subStart, subEnd, match, command, numParts;
+    Tcl_RegExp regExpr;
+    Tcl_RegExpInfo info;
+    Tcl_Obj *resultPtr, *subPtr, *objPtr, *startIndex = NULL;
+    Tcl_UniChar ch, *wsrc, *wfirstChar, *wstring, *wsubspec, *wend;
+
+    static const char *const options[] = {
+	"-all",		"-command",	"-expanded",	"-line",
+	"-linestop",	"-lineanchor",	"-nocase",	"-start",
+	"--",		NULL
+    };
+    enum options {
+	REGSUB_ALL,	 REGSUB_COMMAND,    REGSUB_EXPANDED, REGSUB_LINE,
+	REGSUB_LINESTOP, REGSUB_LINEANCHOR, REGSUB_NOCASE,   REGSUB_START,
+	REGSUB_LAST
+    };
+
+    cflags = TCL_REG_ADVANCED;
+    all = 0;
+    offset = 0;
+    command = 0;
+    resultPtr = NULL;
+
+    for (idx = 1; idx < objc; idx++) {
+	const char *name;
+	int index;
+
+	name = TclGetString(objv[idx]);
+	if (name[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[idx], options, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    goto optionError;
+	}
+	switch ((enum options) index) {
+	case REGSUB_ALL:
+	    all = 1;
+	    break;
+	case REGSUB_NOCASE:
+	    cflags |= TCL_REG_NOCASE;
+	    break;
+	case REGSUB_COMMAND:
+	    command = 1;
+	    break;
+	case REGSUB_EXPANDED:
+	    cflags |= TCL_REG_EXPANDED;
+	    break;
+	case REGSUB_LINE:
+	    cflags |= TCL_REG_NEWLINE;
+	    break;
+	case REGSUB_LINESTOP:
+	    cflags |= TCL_REG_NLSTOP;
+	    break;
+	case REGSUB_LINEANCHOR:
+	    cflags |= TCL_REG_NLANCH;
+	    break;
+	case REGSUB_START: {
+	    int temp;
+	    if (++idx >= objc) {
+		goto endOfForLoop;
+	    }
+	    if (TclGetIntForIndexM(interp, objv[idx], 0, &temp) != TCL_OK) {
+		goto optionError;
+	    }
+	    if (startIndex) {
+		Tcl_DecrRefCount(startIndex);
+	    }
+	    startIndex = objv[idx];
+	    Tcl_IncrRefCount(startIndex);
+	    break;
+	}
+	case REGSUB_LAST:
+	    idx++;
+	    goto endOfForLoop;
+	}
+    }
+
+  endOfForLoop:
+    if (objc-idx < 3 || objc-idx > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-option ...? exp string subSpec ?varName?");
+    optionError:
+	if (startIndex) {
+	    Tcl_DecrRefCount(startIndex);
+	}
+	return TCL_ERROR;
+    }
+
+    objc -= idx;
+    objv += idx;
+
+    if (startIndex) {
+	int stringLength = Tcl_GetCharLength(objv[1]);
+
+	TclGetIntForIndexM(NULL, startIndex, stringLength, &offset);
+	Tcl_DecrRefCount(startIndex);
+	if (offset < 0) {
+	    offset = 0;
+	}
+    }
+
+    if (all && (offset == 0) && (command == 0)
+	    && (strpbrk(TclGetString(objv[2]), "&\\") == NULL)
+	    && (strpbrk(TclGetString(objv[0]), "*+?{}()[].\\|^$") == NULL)) {
+	/*
+	 * This is a simple one pair string map situation. We make use of a
+	 * slightly modified version of the one pair STR_MAP code.
+	 */
+
+	int slen, nocase;
+	int (*strCmpFn)(const Tcl_UniChar*,const Tcl_UniChar*,unsigned long);
+	Tcl_UniChar *p, wsrclc;
+
+	numMatches = 0;
+	nocase = (cflags & TCL_REG_NOCASE);
+	strCmpFn = nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp;
+
+	wsrc = Tcl_GetUnicodeFromObj(objv[0], &slen);
+	wstring = Tcl_GetUnicodeFromObj(objv[1], &wlen);
+	wsubspec = Tcl_GetUnicodeFromObj(objv[2], &wsublen);
+	wend = wstring + wlen - (slen ? slen - 1 : 0);
+	result = TCL_OK;
+
+	if (slen == 0) {
+	    /*
+	     * regsub behavior for "" matches between each character. 'string
+	     * map' skips the "" case.
+	     */
+
+	    if (wstring < wend) {
+		resultPtr = Tcl_NewUnicodeObj(wstring, 0);
+		Tcl_IncrRefCount(resultPtr);
+		for (; wstring < wend; wstring++) {
+		    Tcl_AppendUnicodeToObj(resultPtr, wsubspec, wsublen);
+		    Tcl_AppendUnicodeToObj(resultPtr, wstring, 1);
+		    numMatches++;
+		}
+		wlen = 0;
+	    }
+	} else {
+	    wsrclc = Tcl_UniCharToLower(*wsrc);
+	    for (p = wfirstChar = wstring; wstring < wend; wstring++) {
+		if ((*wstring == *wsrc ||
+			(nocase && Tcl_UniCharToLower(*wstring)==wsrclc)) &&
+			(slen==1 || (strCmpFn(wstring, wsrc,
+				(unsigned long) slen) == 0))) {
+		    if (numMatches == 0) {
+			resultPtr = Tcl_NewUnicodeObj(wstring, 0);
+			Tcl_IncrRefCount(resultPtr);
+		    }
+		    if (p != wstring) {
+			Tcl_AppendUnicodeToObj(resultPtr, p, wstring - p);
+			p = wstring + slen;
+		    } else {
+			p += slen;
+		    }
+		    wstring = p - 1;
+
+		    Tcl_AppendUnicodeToObj(resultPtr, wsubspec, wsublen);
+		    numMatches++;
+		}
+	    }
+	    if (numMatches) {
+		wlen    = wfirstChar + wlen - p;
+		wstring = p;
+	    }
+	}
+	objPtr = NULL;
+	subPtr = NULL;
+	goto regsubDone;
+    }
+
+    regExpr = Tcl_GetRegExpFromObj(interp, objv[0], cflags);
+    if (regExpr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (command) {
+	/*
+	 * In command-prefix mode, we require that the third non-option
+	 * argument be a list, so we enforce that here. Afterwards, we fetch
+	 * the RE compilation again in case objv[0] and objv[2] are the same
+	 * object. (If they aren't, that's cheap to do.)
+	 */
+
+	if (Tcl_ListObjLength(interp, objv[2], &numParts) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (numParts < 1) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command prefix must be a list of at least one element",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "REGSUB",
+		    "CMDEMPTY", NULL);
+	    return TCL_ERROR;
+	}
+	regExpr = Tcl_GetRegExpFromObj(interp, objv[0], cflags);
+    }
+
+    /*
+     * Make sure to avoid problems where the objects are shared. This can
+     * cause RegExpObj <> UnicodeObj shimmering that causes data corruption.
+     * [Bug #461322]
+     */
+
+    if (objv[1] == objv[0]) {
+	objPtr = Tcl_DuplicateObj(objv[1]);
+    } else {
+	objPtr = objv[1];
+    }
+    wstring = Tcl_GetUnicodeFromObj(objPtr, &wlen);
+    if (objv[2] == objv[0]) {
+	subPtr = Tcl_DuplicateObj(objv[2]);
+    } else {
+	subPtr = objv[2];
+    }
+    if (!command) {
+	wsubspec = Tcl_GetUnicodeFromObj(subPtr, &wsublen);
+    }
+
+    result = TCL_OK;
+
+    /*
+     * The following loop is to handle multiple matches within the same source
+     * string; each iteration handles one match and its corresponding
+     * substitution. If "-all" hasn't been specified then the loop body only
+     * gets executed once. We must use 'offset <= wlen' in particular for the
+     * case where the regexp pattern can match the empty string - this is
+     * useful when doing, say, 'regsub -- ^ $str ...' when $str might be
+     * empty.
+     */
+
+    numMatches = 0;
+    for ( ; offset <= wlen; ) {
+
+	/*
+	 * The flags argument is set if string is part of a larger string, so
+	 * that "^" won't match.
+	 */
+
+	match = Tcl_RegExpExecObj(interp, regExpr, objPtr, offset,
+		10 /* matches */, ((offset > 0 &&
+		(wstring[offset-1] != (Tcl_UniChar)'\n'))
+		? TCL_REG_NOTBOL : 0));
+
+	if (match < 0) {
+	    result = TCL_ERROR;
+	    goto done;
+	}
+	if (match == 0) {
+	    break;
+	}
+	if (numMatches == 0) {
+	    resultPtr = Tcl_NewUnicodeObj(wstring, 0);
+	    Tcl_IncrRefCount(resultPtr);
+	    if (offset > 0) {
+		/*
+		 * Copy the initial portion of the string in if an offset was
+		 * specified.
+		 */
+
+		Tcl_AppendUnicodeToObj(resultPtr, wstring, offset);
+	    }
+	}
+	numMatches++;
+
+	/*
+	 * Copy the portion of the source string before the match to the
+	 * result variable.
+	 */
+
+	Tcl_RegExpGetInfo(regExpr, &info);
+	start = info.matches[0].start;
+	end = info.matches[0].end;
+	Tcl_AppendUnicodeToObj(resultPtr, wstring + offset, start);
+
+	/*
+	 * In command-prefix mode, the substitutions are added as quoted
+	 * arguments to the subSpec to form a command, that is then executed
+	 * and the result used as the string to substitute in. Actually,
+	 * everything is passed through Tcl_EvalObjv, as that's much faster.
+	 */
+
+	if (command) {
+	    Tcl_Obj **args = NULL, **parts;
+	    int numArgs;
+
+	    Tcl_ListObjGetElements(interp, subPtr, &numParts, &parts);
+	    numArgs = numParts + info.nsubs + 1;
+	    args = ckalloc(sizeof(Tcl_Obj*) * numArgs);
+	    memcpy(args, parts, sizeof(Tcl_Obj*) * numParts);
+
+	    for (idx = 0 ; idx <= info.nsubs ; idx++) {
+		subStart = info.matches[idx].start;
+		subEnd = info.matches[idx].end;
+		if ((subStart >= 0) && (subEnd >= 0)) {
+		    args[idx + numParts] = Tcl_NewUnicodeObj(
+			    wstring + offset + subStart, subEnd - subStart);
+		} else {
+		    args[idx + numParts] = Tcl_NewObj();
+		}
+		Tcl_IncrRefCount(args[idx + numParts]);
+	    }
+
+	    /*
+	     * At this point, we're locally holding the references to the
+	     * argument words we added for this time round the loop, and the
+	     * subPtr is holding the references to the words that the user
+	     * supplied directly. None are zero-refcount, which is important
+	     * because Tcl_EvalObjv is "hairy monster" in terms of refcount
+	     * handling, being able to optionally add references to any of its
+	     * argument words. We'll drop the local refs immediately
+	     * afterwards; subPtr is handled in the main exit stanza.
+	     */
+
+	    result = Tcl_EvalObjv(interp, numArgs, args, 0);
+	    for (idx = 0 ; idx <= info.nsubs ; idx++) {
+		TclDecrRefCount(args[idx + numParts]);
+	    }
+	    ckfree(args);
+	    if (result != TCL_OK) {
+		if (result == TCL_ERROR) {
+		    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			    "\n    (%s substitution computation script)",
+			    options[REGSUB_COMMAND]));
+		}
+		goto done;
+	    }
+
+	    Tcl_AppendObjToObj(resultPtr, Tcl_GetObjResult(interp));
+	    Tcl_ResetResult(interp);
+
+	    /*
+	     * Refetch the unicode, in case the representation was smashed by
+	     * the user code.
+	     */
+
+	    wstring = Tcl_GetUnicodeFromObj(objPtr, &wlen);
+
+	    offset += end;
+	    if (end == 0 || start == end) {
+		/*
+		 * Always consume at least one character of the input string
+		 * in order to prevent infinite loops, even when we
+		 * technically matched the empty string; we must not match
+		 * again at the same spot.
+		 */
+
+		if (offset < wlen) {
+		    Tcl_AppendUnicodeToObj(resultPtr, wstring + offset, 1);
+		}
+		offset++;
+	    }
+	    if (all) {
+		continue;
+	    } else {
+		break;
+	    }
+	}
+
+	/*
+	 * Append the subSpec argument to the variable, making appropriate
+	 * substitutions. This code is a bit hairy because of the backslash
+	 * conventions and because the code saves up ranges of characters in
+	 * subSpec to reduce the number of calls to Tcl_SetVar.
+	 */
+
+	wsrc = wfirstChar = wsubspec;
+	wend = wsubspec + wsublen;
+	for (ch = *wsrc; wsrc != wend; wsrc++, ch = *wsrc) {
+	    if (ch == '&') {
+		idx = 0;
+	    } else if (ch == '\\') {
+		ch = wsrc[1];
+		if ((ch >= '0') && (ch <= '9')) {
+		    idx = ch - '0';
+		} else if ((ch == '\\') || (ch == '&')) {
+		    *wsrc = ch;
+		    Tcl_AppendUnicodeToObj(resultPtr, wfirstChar,
+			    wsrc - wfirstChar + 1);
+		    *wsrc = '\\';
+		    wfirstChar = wsrc + 2;
+		    wsrc++;
+		    continue;
+		} else {
+		    continue;
+		}
+	    } else {
+		continue;
+	    }
+
+	    if (wfirstChar != wsrc) {
+		Tcl_AppendUnicodeToObj(resultPtr, wfirstChar,
+			wsrc - wfirstChar);
+	    }
+
+	    if (idx <= info.nsubs) {
+		subStart = info.matches[idx].start;
+		subEnd = info.matches[idx].end;
+		if ((subStart >= 0) && (subEnd >= 0)) {
+		    Tcl_AppendUnicodeToObj(resultPtr,
+			    wstring + offset + subStart, subEnd - subStart);
+		}
+	    }
+
+	    if (*wsrc == '\\') {
+		wsrc++;
+	    }
+	    wfirstChar = wsrc + 1;
+	}
+
+	if (wfirstChar != wsrc) {
+	    Tcl_AppendUnicodeToObj(resultPtr, wfirstChar, wsrc - wfirstChar);
+	}
+
+	if (end == 0) {
+	    /*
+	     * Always consume at least one character of the input string in
+	     * order to prevent infinite loops.
+	     */
+
+	    if (offset < wlen) {
+		Tcl_AppendUnicodeToObj(resultPtr, wstring + offset, 1);
+	    }
+	    offset++;
+	} else {
+	    offset += end;
+	    if (start == end) {
+		/*
+		 * We matched an empty string, which means we must go forward
+		 * one more step so we don't match again at the same spot.
+		 */
+
+		if (offset < wlen) {
+		    Tcl_AppendUnicodeToObj(resultPtr, wstring + offset, 1);
+		}
+		offset++;
+	    }
+	}
+	if (!all) {
+	    break;
+	}
+    }
+
+    /*
+     * Copy the portion of the source string after the last match to the
+     * result variable.
+     */
+
+  regsubDone:
+    if (numMatches == 0) {
+	/*
+	 * On zero matches, just ignore the offset, since it shouldn't matter
+	 * to us in this case, and the user may have skewed it.
+	 */
+
+	resultPtr = objv[1];
+	Tcl_IncrRefCount(resultPtr);
+    } else if (offset < wlen) {
+	Tcl_AppendUnicodeToObj(resultPtr, wstring + offset, wlen - offset);
+    }
+    if (objc == 4) {
+	if (Tcl_ObjSetVar2(interp, objv[3], NULL, resultPtr,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    result = TCL_ERROR;
+	} else {
+	    /*
+	     * Set the interpreter's object result to an integer object
+	     * holding the number of matches.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(numMatches));
+	}
+    } else {
+	/*
+	 * No varname supplied, so just return the modified string.
+	 */
+
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+
+  done:
+    if (objPtr && (objv[1] == objv[0])) {
+	Tcl_DecrRefCount(objPtr);
+    }
+    if (subPtr && (objv[2] == objv[0])) {
+	Tcl_DecrRefCount(subPtr);
+    }
+    if (resultPtr) {
+	Tcl_DecrRefCount(resultPtr);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RenameObjCmd --
+ *
+ *	This procedure is invoked to process the "rename" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RenameObjCmd(
+    ClientData dummy,		/* Arbitrary value passed to the command. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *oldName, *newName;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "oldName newName");
+	return TCL_ERROR;
+    }
+
+    oldName = TclGetString(objv[1]);
+    newName = TclGetString(objv[2]);
+    return TclRenameCommand(interp, oldName, newName);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ReturnObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "return" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ReturnObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int code, level;
+    Tcl_Obj *returnOpts;
+
+    /*
+     * General syntax: [return ?-option value ...? ?result?]
+     * An even number of words means an explicit result argument is present.
+     */
+
+    int explicitResult = (0 == (objc % 2));
+    int numOptionWords = objc - 1 - explicitResult;
+
+    if (TCL_ERROR == TclMergeReturnOptions(interp, numOptionWords, objv+1,
+	    &returnOpts, &code, &level)) {
+	return TCL_ERROR;
+    }
+
+    code = TclProcessReturn(interp, code, level, returnOpts);
+    if (explicitResult) {
+	Tcl_SetObjResult(interp, objv[objc-1]);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SourceObjCmd --
+ *
+ *	This procedure is invoked to process the "source" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SourceObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRSourceObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRSourceObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *encodingName = NULL;
+    Tcl_Obj *fileName;
+    int result;
+    void **pkgFiles = NULL;
+    void *names = NULL;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-encoding name? fileName");
+	return TCL_ERROR;
+    }
+
+    fileName = objv[objc-1];
+
+    if (objc == 4) {
+	static const char *const options[] = {
+	    "-encoding", NULL
+	};
+	int index;
+
+	if (TCL_ERROR == Tcl_GetIndexFromObj(interp, objv[1], options,
+		"option", TCL_EXACT, &index)) {
+	    return TCL_ERROR;
+	}
+	encodingName = TclGetString(objv[2]);
+    } else if (objc == 3) {
+	/* Handle undocumented -nopkg option. This should only be
+	 * used by the internal ::tcl::Pkg::source utility function. */
+	static const char *const nopkgoptions[] = {
+	    "-nopkg", NULL
+	};
+	int index;
+
+	if (TCL_ERROR == Tcl_GetIndexFromObj(interp, objv[1], nopkgoptions,
+		"option", TCL_EXACT, &index)) {
+	    return TCL_ERROR;
+	}
+	pkgFiles = Tcl_GetAssocData(interp, "tclPkgFiles", NULL);
+	/* Make sure that during the following TclNREvalFile no filenames
+	 * are recorded for inclusion in the "package files" command */
+	names = *pkgFiles;
+	*pkgFiles = NULL;
+    }
+    result = TclNREvalFile(interp, fileName, encodingName);
+    if (pkgFiles) {
+	/* restore "tclPkgFiles" assocdata to how it was. */
+	*pkgFiles = names;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SplitObjCmd --
+ *
+ *	This procedure is invoked to process the "split" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SplitObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_UniChar ch = 0;
+    int len;
+    const char *splitChars;
+    const char *stringPtr;
+    const char *end;
+    int splitCharLen, stringLen;
+    Tcl_Obj *listPtr, *objPtr;
+
+    if (objc == 2) {
+	splitChars = " \n\t\r";
+	splitCharLen = 4;
+    } else if (objc == 3) {
+	splitChars = TclGetStringFromObj(objv[2], &splitCharLen);
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?splitChars?");
+	return TCL_ERROR;
+    }
+
+    stringPtr = TclGetStringFromObj(objv[1], &stringLen);
+    end = stringPtr + stringLen;
+    listPtr = Tcl_NewObj();
+
+    if (stringLen == 0) {
+	/*
+	 * Do nothing.
+	 */
+    } else if (splitCharLen == 0) {
+	Tcl_HashTable charReuseTable;
+	Tcl_HashEntry *hPtr;
+	int isNew;
+
+	/*
+	 * Handle the special case of splitting on every character.
+	 *
+	 * Uses a hash table to ensure that each kind of character has only
+	 * one Tcl_Obj instance (multiply-referenced) in the final list. This
+	 * is a *major* win when splitting on a long string (especially in the
+	 * megabyte range!) - DKF
+	 */
+
+	Tcl_InitHashTable(&charReuseTable, TCL_ONE_WORD_KEYS);
+
+	for ( ; stringPtr < end; stringPtr += len) {
+	    len = TclUtfToUniChar(stringPtr, &ch);
+
+	    /*
+	     * Assume Tcl_UniChar is an integral type...
+	     */
+
+	    hPtr = Tcl_CreateHashEntry(&charReuseTable, INT2PTR((int) ch),
+		    &isNew);
+	    if (isNew) {
+		TclNewStringObj(objPtr, stringPtr, len);
+
+		/*
+		 * Don't need to fiddle with refcount...
+		 */
+
+		Tcl_SetHashValue(hPtr, objPtr);
+	    } else {
+		objPtr = Tcl_GetHashValue(hPtr);
+	    }
+	    Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
+	}
+	Tcl_DeleteHashTable(&charReuseTable);
+
+    } else if (splitCharLen == 1) {
+	char *p;
+
+	/*
+	 * Handle the special case of splitting on a single character. This is
+	 * only true for the one-char ASCII case, as one unicode char is > 1
+	 * byte in length.
+	 */
+
+	while (*stringPtr && (p=strchr(stringPtr,(int)*splitChars)) != NULL) {
+	    objPtr = Tcl_NewStringObj(stringPtr, p - stringPtr);
+	    Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
+	    stringPtr = p + 1;
+	}
+	TclNewStringObj(objPtr, stringPtr, end - stringPtr);
+	Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
+    } else {
+	const char *element, *p, *splitEnd;
+	int splitLen;
+	Tcl_UniChar splitChar = 0;
+
+	/*
+	 * Normal case: split on any of a given set of characters. Discard
+	 * instances of the split characters.
+	 */
+
+	splitEnd = splitChars + splitCharLen;
+
+	for (element = stringPtr; stringPtr < end; stringPtr += len) {
+	    len = TclUtfToUniChar(stringPtr, &ch);
+	    for (p = splitChars; p < splitEnd; p += splitLen) {
+		splitLen = TclUtfToUniChar(p, &splitChar);
+		if (ch == splitChar) {
+		    TclNewStringObj(objPtr, element, stringPtr - element);
+		    Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
+		    element = stringPtr + len;
+		    break;
+		}
+	    }
+	}
+
+	TclNewStringObj(objPtr, element, stringPtr - element);
+	Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
+    }
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringFirstCmd --
+ *
+ *	This procedure is invoked to process the "string first" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringFirstCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int start = 0;
+
+    if (objc < 3 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"needleString haystackString ?startIndex?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 4) {
+	int size = Tcl_GetCharLength(objv[2]);
+
+	if (TCL_OK != TclGetIntForIndexM(interp, objv[3], size - 1, &start)) {
+	    return TCL_ERROR;
+	}
+
+	if (start < 0) {
+	    start = 0;
+	}
+	if (start >= size) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(-1));
+	    return TCL_OK;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(TclStringFind(objv[1],
+	    objv[2], start)));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringLastCmd --
+ *
+ *	This procedure is invoked to process the "string last" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringLastCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int last = INT_MAX - 1;
+
+    if (objc < 3 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"needleString haystackString ?lastIndex?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 4) {
+	int size = Tcl_GetCharLength(objv[2]);
+
+	if (TCL_OK != TclGetIntForIndexM(interp, objv[3], size - 1, &last)) {
+	    return TCL_ERROR;
+	}
+
+	if (last < 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(-1));
+	    return TCL_OK;
+	}
+	if (last >= size) {
+	    last = size - 1;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(TclStringLast(objv[1],
+	    objv[2], last)));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringIndexCmd --
+ *
+ *	This procedure is invoked to process the "string index" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringIndexCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length, index;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string charIndex");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the char length to calulate what 'end' means.
+     */
+
+    length = Tcl_GetCharLength(objv[1]);
+    if (TclGetIntForIndexM(interp, objv[2], length-1, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if ((index >= 0) && (index < length)) {
+	Tcl_UniChar ch = Tcl_GetUniChar(objv[1], index);
+
+	/*
+	 * If we have a ByteArray object, we're careful to generate a new
+	 * bytearray for a result.
+	 */
+
+	if (TclIsPureByteArray(objv[1])) {
+	    unsigned char uch = (unsigned char) ch;
+
+	    Tcl_SetObjResult(interp, Tcl_NewByteArrayObj(&uch, 1));
+	} else {
+	    char buf[TCL_UTF_MAX];
+
+	    length = Tcl_UniCharToUtf(ch, buf);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(buf, length));
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringIsCmd --
+ *
+ *	This procedure is invoked to process the "string is" Tcl command. See
+ *	the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringIsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *string1, *end, *stop;
+    Tcl_UniChar ch = 0;
+    int (*chcomp)(int) = NULL;	/* The UniChar comparison function. */
+    int i, failat = 0, result = 1, strict = 0, index, length1, length2;
+    Tcl_Obj *objPtr, *failVarObj = NULL;
+    Tcl_WideInt w;
+
+    static const char *const isClasses[] = {
+	"alnum",	"alpha",	"ascii",	"control",
+	"boolean",	"digit",	"double",	"entier",
+	"false",	"graph",	"integer",	"list",
+	"lower",	"print",	"punct",	"space",
+	"true",		"upper",	"wideinteger",	"wordchar",
+	"xdigit",	NULL
+    };
+    enum isClasses {
+	STR_IS_ALNUM,	STR_IS_ALPHA,	STR_IS_ASCII,	STR_IS_CONTROL,
+	STR_IS_BOOL,	STR_IS_DIGIT,	STR_IS_DOUBLE,	STR_IS_ENTIER,
+	STR_IS_FALSE,	STR_IS_GRAPH,	STR_IS_INT,	STR_IS_LIST,
+	STR_IS_LOWER,	STR_IS_PRINT,	STR_IS_PUNCT,	STR_IS_SPACE,
+	STR_IS_TRUE,	STR_IS_UPPER,	STR_IS_WIDE,	STR_IS_WORD,
+	STR_IS_XDIGIT
+    };
+    static const char *const isOptions[] = {
+	"-strict", "-failindex", NULL
+    };
+    enum isOptions {
+	OPT_STRICT, OPT_FAILIDX
+    };
+
+    if (objc < 3 || objc > 6) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"class ?-strict? ?-failindex var? str");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], isClasses, "class", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc != 3) {
+	for (i = 2; i < objc-1; i++) {
+	    int idx2;
+
+	    if (Tcl_GetIndexFromObj(interp, objv[i], isOptions, "option", 0,
+		    &idx2) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum isOptions) idx2) {
+	    case OPT_STRICT:
+		strict = 1;
+		break;
+	    case OPT_FAILIDX:
+		if (i+1 >= objc-1) {
+		    Tcl_WrongNumArgs(interp, 2, objv,
+			    "?-strict? ?-failindex var? str");
+		    return TCL_ERROR;
+		}
+		failVarObj = objv[++i];
+		break;
+	    }
+	}
+    }
+
+    /*
+     * We get the objPtr so that we can short-cut for some classes by checking
+     * the object type (int and double), but we need the string otherwise,
+     * because we don't want any conversion of type occuring (as, for example,
+     * Tcl_Get*FromObj would do).
+     */
+
+    objPtr = objv[objc-1];
+
+    /*
+     * When entering here, result == 1 and failat == 0.
+     */
+
+    switch ((enum isClasses) index) {
+    case STR_IS_ALNUM:
+	chcomp = Tcl_UniCharIsAlnum;
+	break;
+    case STR_IS_ALPHA:
+	chcomp = Tcl_UniCharIsAlpha;
+	break;
+    case STR_IS_ASCII:
+	chcomp = UniCharIsAscii;
+	break;
+    case STR_IS_BOOL:
+    case STR_IS_TRUE:
+    case STR_IS_FALSE:
+	if ((objPtr->typePtr != &tclBooleanType)
+		&& (TCL_OK != TclSetBooleanFromAny(NULL, objPtr))) {
+	    if (strict) {
+		result = 0;
+	    } else {
+		string1 = TclGetStringFromObj(objPtr, &length1);
+		result = length1 == 0;
+	    }
+	} else if (((index == STR_IS_TRUE) &&
+		objPtr->internalRep.longValue == 0)
+	    || ((index == STR_IS_FALSE) &&
+		objPtr->internalRep.longValue != 0)) {
+	    result = 0;
+	}
+	break;
+    case STR_IS_CONTROL:
+	chcomp = Tcl_UniCharIsControl;
+	break;
+    case STR_IS_DIGIT:
+	chcomp = Tcl_UniCharIsDigit;
+	break;
+    case STR_IS_DOUBLE: {
+	/* TODO */
+	if ((objPtr->typePtr == &tclDoubleType) ||
+		(objPtr->typePtr == &tclIntType) ||
+#ifndef TCL_WIDE_INT_IS_LONG
+		(objPtr->typePtr == &tclWideIntType) ||
+#endif
+		(objPtr->typePtr == &tclBignumType)) {
+	    break;
+	}
+	string1 = TclGetStringFromObj(objPtr, &length1);
+	if (length1 == 0) {
+	    if (strict) {
+		result = 0;
+	    }
+	    goto str_is_done;
+	}
+	end = string1 + length1;
+	if (TclParseNumber(NULL, objPtr, NULL, NULL, -1,
+		(const char **) &stop, 0) != TCL_OK) {
+	    result = 0;
+	    failat = 0;
+	} else {
+	    failat = stop - string1;
+	    if (stop < end) {
+		result = 0;
+		TclFreeIntRep(objPtr);
+	    }
+	}
+	break;
+    }
+    case STR_IS_GRAPH:
+	chcomp = Tcl_UniCharIsGraph;
+	break;
+    case STR_IS_INT:
+	if (TCL_OK == TclGetIntFromObj(NULL, objPtr, &i)) {
+	    break;
+	}
+	goto failedIntParse;
+    case STR_IS_ENTIER:
+	if ((objPtr->typePtr == &tclIntType) ||
+#ifndef TCL_WIDE_INT_IS_LONG
+		(objPtr->typePtr == &tclWideIntType) ||
+#endif
+		(objPtr->typePtr == &tclBignumType)) {
+	    break;
+	}
+	string1 = TclGetStringFromObj(objPtr, &length1);
+	if (length1 == 0) {
+	    if (strict) {
+		result = 0;
+	    }
+	    goto str_is_done;
+	}
+	end = string1 + length1;
+	if (TclParseNumber(NULL, objPtr, NULL, NULL, -1,
+		(const char **) &stop, TCL_PARSE_INTEGER_ONLY) == TCL_OK) {
+	    if (stop == end) {
+		/*
+		 * Entire string parses as an integer.
+		 */
+
+		break;
+	    } else {
+		/*
+		 * Some prefix parsed as an integer, but not the whole string,
+		 * so return failure index as the point where parsing stopped.
+		 * Clear out the internal rep, since keeping it would leave
+		 * *objPtr in an inconsistent state.
+		 */
+
+		result = 0;
+		failat = stop - string1;
+		TclFreeIntRep(objPtr);
+	    }
+	} else {
+	    /*
+	     * No prefix is a valid integer. Fail at beginning.
+	     */
+
+	    result = 0;
+	    failat = 0;
+	}
+	break;
+    case STR_IS_WIDE:
+	if (TCL_OK == TclGetWideIntFromObj(NULL, objPtr, &w)) {
+	    break;
+	}
+
+    failedIntParse:
+	string1 = TclGetStringFromObj(objPtr, &length1);
+	if (length1 == 0) {
+	    if (strict) {
+		result = 0;
+	    }
+	    goto str_is_done;
+	}
+	result = 0;
+	if (failVarObj == NULL) {
+	    /*
+	     * Don't bother computing the failure point if we're not going to
+	     * return it.
+	     */
+
+	    break;
+	}
+	end = string1 + length1;
+	if (TclParseNumber(NULL, objPtr, NULL, NULL, -1,
+		(const char **) &stop, TCL_PARSE_INTEGER_ONLY) == TCL_OK) {
+	    if (stop == end) {
+		/*
+		 * Entire string parses as an integer, but rejected by
+		 * Tcl_Get(Wide)IntFromObj() so we must have overflowed the
+		 * target type, and our convention is to return failure at
+		 * index -1 in that situation.
+		 */
+
+		failat = -1;
+	    } else {
+		/*
+		 * Some prefix parsed as an integer, but not the whole string,
+		 * so return failure index as the point where parsing stopped.
+		 * Clear out the internal rep, since keeping it would leave
+		 * *objPtr in an inconsistent state.
+		 */
+
+		failat = stop - string1;
+		TclFreeIntRep(objPtr);
+	    }
+	} else {
+	    /*
+	     * No prefix is a valid integer. Fail at beginning.
+	     */
+
+	    failat = 0;
+	}
+	break;
+    case STR_IS_LIST:
+	/*
+	 * We ignore the strictness here, since empty strings are always
+	 * well-formed lists.
+	 */
+
+	if (TCL_OK == TclListObjLength(NULL, objPtr, &length2)) {
+	    break;
+	}
+
+	if (failVarObj != NULL) {
+	    /*
+	     * Need to figure out where the list parsing failed, which is
+	     * fairly expensive. This is adapted from the core of
+	     * SetListFromAny().
+	     */
+
+	    const char *elemStart, *nextElem;
+	    int lenRemain, elemSize;
+	    register const char *p;
+
+	    string1 = TclGetStringFromObj(objPtr, &length1);
+	    end = string1 + length1;
+	    failat = -1;
+	    for (p=string1, lenRemain=length1; lenRemain > 0;
+		    p=nextElem, lenRemain=end-nextElem) {
+		if (TCL_ERROR == TclFindElement(NULL, p, lenRemain,
+			&elemStart, &nextElem, &elemSize, NULL)) {
+		    Tcl_Obj *tmpStr;
+
+		    /*
+		     * This is the simplest way of getting the number of
+		     * characters parsed. Note that this is not the same as
+		     * the number of bytes when parsing strings with non-ASCII
+		     * characters in them.
+		     *
+		     * Skip leading spaces first. This is only really an issue
+		     * if it is the first "element" that has the failure.
+		     */
+
+		    while (TclIsSpaceProc(*p)) {
+			p++;
+		    }
+		    TclNewStringObj(tmpStr, string1, p-string1);
+		    failat = Tcl_GetCharLength(tmpStr);
+		    TclDecrRefCount(tmpStr);
+		    break;
+		}
+	    }
+	}
+	result = 0;
+	break;
+    case STR_IS_LOWER:
+	chcomp = Tcl_UniCharIsLower;
+	break;
+    case STR_IS_PRINT:
+	chcomp = Tcl_UniCharIsPrint;
+	break;
+    case STR_IS_PUNCT:
+	chcomp = Tcl_UniCharIsPunct;
+	break;
+    case STR_IS_SPACE:
+	chcomp = Tcl_UniCharIsSpace;
+	break;
+    case STR_IS_UPPER:
+	chcomp = Tcl_UniCharIsUpper;
+	break;
+    case STR_IS_WORD:
+	chcomp = Tcl_UniCharIsWordChar;
+	break;
+    case STR_IS_XDIGIT:
+	chcomp = UniCharIsHexDigit;
+	break;
+    }
+
+    if (chcomp != NULL) {
+	string1 = TclGetStringFromObj(objPtr, &length1);
+	if (length1 == 0) {
+	    if (strict) {
+		result = 0;
+	    }
+	    goto str_is_done;
+	}
+	end = string1 + length1;
+	for (; string1 < end; string1 += length2, failat++) {
+	    length2 = TclUtfToUniChar(string1, &ch);
+	    if (!chcomp(ch)) {
+		result = 0;
+		break;
+	    }
+	}
+    }
+
+    /*
+     * Only set the failVarObj when we will return 0 and we have indicated a
+     * valid fail index (>= 0).
+     */
+
+ str_is_done:
+    if ((result == 0) && (failVarObj != NULL) &&
+	Tcl_ObjSetVar2(interp, failVarObj, NULL, Tcl_NewIntObj(failat),
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(result));
+    return TCL_OK;
+}
+
+static int
+UniCharIsAscii(
+    int character)
+{
+    return (character >= 0) && (character < 0x80);
+}
+
+static int
+UniCharIsHexDigit(
+    int character)
+{
+    return (character >= 0) && (character < 0x80) && isxdigit(character);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringMapCmd --
+ *
+ *	This procedure is invoked to process the "string map" Tcl command. See
+ *	the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringMapCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length1, length2, mapElemc, index;
+    int nocase = 0, mapWithDict = 0, copySource = 0;
+    Tcl_Obj **mapElemv, *sourceObj, *resultPtr;
+    Tcl_UniChar *ustring1, *ustring2, *p, *end;
+    int (*strCmpFn)(const Tcl_UniChar*, const Tcl_UniChar*, unsigned long);
+
+    if (objc < 3 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-nocase? charMap string");
+	return TCL_ERROR;
+    }
+
+    if (objc == 4) {
+	const char *string = TclGetStringFromObj(objv[1], &length2);
+
+	if ((length2 > 1) &&
+		strncmp(string, "-nocase", (size_t) length2) == 0) {
+	    nocase = 1;
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\": must be -nocase", string));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
+		    string, NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * This test is tricky, but has to be that way or you get other strange
+     * inconsistencies (see test string-10.20.1 for illustration why!)
+     */
+
+    if (objv[objc-2]->typePtr == &tclDictType && objv[objc-2]->bytes == NULL){
+	int i, done;
+	Tcl_DictSearch search;
+
+	/*
+	 * We know the type exactly, so all dict operations will succeed for
+	 * sure. This shortens this code quite a bit.
+	 */
+
+	Tcl_DictObjSize(interp, objv[objc-2], &mapElemc);
+	if (mapElemc == 0) {
+	    /*
+	     * Empty charMap, just return whatever string was given.
+	     */
+
+	    Tcl_SetObjResult(interp, objv[objc-1]);
+	    return TCL_OK;
+	}
+
+	mapElemc *= 2;
+	mapWithDict = 1;
+
+	/*
+	 * Copy the dictionary out into an array; that's the easiest way to
+	 * adapt this code...
+	 */
+
+	mapElemv = TclStackAlloc(interp, sizeof(Tcl_Obj *) * mapElemc);
+	Tcl_DictObjFirst(interp, objv[objc-2], &search, mapElemv+0,
+		mapElemv+1, &done);
+	for (i=2 ; i<mapElemc ; i+=2) {
+	    Tcl_DictObjNext(&search, mapElemv+i, mapElemv+i+1, &done);
+	}
+	Tcl_DictObjDone(&search);
+    } else {
+	if (TclListObjGetElements(interp, objv[objc-2], &mapElemc,
+		&mapElemv) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (mapElemc == 0) {
+	    /*
+	     * empty charMap, just return whatever string was given.
+	     */
+
+	    Tcl_SetObjResult(interp, objv[objc-1]);
+	    return TCL_OK;
+	} else if (mapElemc & 1) {
+	    /*
+	     * The charMap must be an even number of key/value items.
+	     */
+
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("char map list unbalanced", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "MAP",
+		    "UNBALANCED", NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Take a copy of the source string object if it is the same as the map
+     * string to cut out nasty sharing crashes. [Bug 1018562]
+     */
+
+    if (objv[objc-2] == objv[objc-1]) {
+	sourceObj = Tcl_DuplicateObj(objv[objc-1]);
+	copySource = 1;
+    } else {
+	sourceObj = objv[objc-1];
+    }
+    ustring1 = Tcl_GetUnicodeFromObj(sourceObj, &length1);
+    if (length1 == 0) {
+	/*
+	 * Empty input string, just stop now.
+	 */
+
+	goto done;
+    }
+    end = ustring1 + length1;
+
+    strCmpFn = (nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp);
+
+    /*
+     * Force result to be Unicode
+     */
+
+    resultPtr = Tcl_NewUnicodeObj(ustring1, 0);
+
+    if (mapElemc == 2) {
+	/*
+	 * Special case for one map pair which avoids the extra for loop and
+	 * extra calls to get Unicode data. The algorithm is otherwise
+	 * identical to the multi-pair case. This will be >30% faster on
+	 * larger strings.
+	 */
+
+	int mapLen;
+	Tcl_UniChar *mapString, u2lc;
+
+	ustring2 = Tcl_GetUnicodeFromObj(mapElemv[0], &length2);
+	p = ustring1;
+	if ((length2 > length1) || (length2 == 0)) {
+	    /*
+	     * Match string is either longer than input or empty.
+	     */
+
+	    ustring1 = end;
+	} else {
+	    mapString = Tcl_GetUnicodeFromObj(mapElemv[1], &mapLen);
+	    u2lc = (nocase ? Tcl_UniCharToLower(*ustring2) : 0);
+	    for (; ustring1 < end; ustring1++) {
+		if (((*ustring1 == *ustring2) ||
+			(nocase&&Tcl_UniCharToLower(*ustring1)==u2lc)) &&
+			(length2==1 || strCmpFn(ustring1, ustring2,
+				(unsigned long) length2) == 0)) {
+		    if (p != ustring1) {
+			Tcl_AppendUnicodeToObj(resultPtr, p, ustring1-p);
+			p = ustring1 + length2;
+		    } else {
+			p += length2;
+		    }
+		    ustring1 = p - 1;
+
+		    Tcl_AppendUnicodeToObj(resultPtr, mapString, mapLen);
+		}
+	    }
+	}
+    } else {
+	Tcl_UniChar **mapStrings, *u2lc = NULL;
+	int *mapLens;
+
+	/*
+	 * Precompute pointers to the unicode string and length. This saves us
+	 * repeated function calls later, significantly speeding up the
+	 * algorithm. We only need the lowercase first char in the nocase
+	 * case.
+	 */
+
+	mapStrings = TclStackAlloc(interp, mapElemc*2*sizeof(Tcl_UniChar *));
+	mapLens = TclStackAlloc(interp, mapElemc * 2 * sizeof(int));
+	if (nocase) {
+	    u2lc = TclStackAlloc(interp, mapElemc * sizeof(Tcl_UniChar));
+	}
+	for (index = 0; index < mapElemc; index++) {
+	    mapStrings[index] = Tcl_GetUnicodeFromObj(mapElemv[index],
+		    mapLens+index);
+	    if (nocase && ((index % 2) == 0)) {
+		u2lc[index/2] = Tcl_UniCharToLower(*mapStrings[index]);
+	    }
+	}
+	for (p = ustring1; ustring1 < end; ustring1++) {
+	    for (index = 0; index < mapElemc; index += 2) {
+		/*
+		 * Get the key string to match on.
+		 */
+
+		ustring2 = mapStrings[index];
+		length2 = mapLens[index];
+		if ((length2 > 0) && ((*ustring1 == *ustring2) || (nocase &&
+			(Tcl_UniCharToLower(*ustring1) == u2lc[index/2]))) &&
+			/* Restrict max compare length. */
+			(end-ustring1 >= length2) && ((length2 == 1) ||
+			!strCmpFn(ustring2, ustring1, (unsigned) length2))) {
+		    if (p != ustring1) {
+			/*
+			 * Put the skipped chars onto the result first.
+			 */
+
+			Tcl_AppendUnicodeToObj(resultPtr, p, ustring1-p);
+			p = ustring1 + length2;
+		    } else {
+			p += length2;
+		    }
+
+		    /*
+		     * Adjust len to be full length of matched string.
+		     */
+
+		    ustring1 = p - 1;
+
+		    /*
+		     * Append the map value to the unicode string.
+		     */
+
+		    Tcl_AppendUnicodeToObj(resultPtr,
+			    mapStrings[index+1], mapLens[index+1]);
+		    break;
+		}
+	    }
+	}
+	if (nocase) {
+	    TclStackFree(interp, u2lc);
+	}
+	TclStackFree(interp, mapLens);
+	TclStackFree(interp, mapStrings);
+    }
+    if (p != ustring1) {
+	/*
+	 * Put the rest of the unmapped chars onto result.
+	 */
+
+	Tcl_AppendUnicodeToObj(resultPtr, p, ustring1 - p);
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+  done:
+    if (mapWithDict) {
+	TclStackFree(interp, mapElemv);
+    }
+    if (copySource) {
+	Tcl_DecrRefCount(sourceObj);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringMatchCmd --
+ *
+ *	This procedure is invoked to process the "string match" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringMatchCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int nocase = 0;
+
+    if (objc < 3 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-nocase? pattern string");
+	return TCL_ERROR;
+    }
+
+    if (objc == 4) {
+	int length;
+	const char *string = TclGetStringFromObj(objv[1], &length);
+
+	if ((length > 1) &&
+	    strncmp(string, "-nocase", (size_t) length) == 0) {
+	    nocase = TCL_MATCH_NOCASE;
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\": must be -nocase", string));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
+		    string, NULL);
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(
+		TclStringMatchObj(objv[objc-1], objv[objc-2], nocase)));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringRangeCmd --
+ *
+ *	This procedure is invoked to process the "string range" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringRangeCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length, first, last;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string first last");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the length in actual characters; Then reduce it by one because
+     * 'end' refers to the last character, not one past it.
+     */
+
+    length = Tcl_GetCharLength(objv[1]) - 1;
+
+    if (TclGetIntForIndexM(interp, objv[2], length, &first) != TCL_OK ||
+	    TclGetIntForIndexM(interp, objv[3], length, &last) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (first < 0) {
+	first = 0;
+    }
+    if (last >= length) {
+	last = length;
+    }
+    if (last >= first) {
+	Tcl_SetObjResult(interp, Tcl_GetRange(objv[1], first, last));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringReptCmd --
+ *
+ *	This procedure is invoked to process the "string repeat" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringReptCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int count;
+    Tcl_Obj *resultPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string count");
+	return TCL_ERROR;
+    }
+
+    if (TclGetIntFromObj(interp, objv[2], &count) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check for cases that allow us to skip copying stuff.
+     */
+
+    if (count == 1) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    } else if (count < 1) {
+	return TCL_OK;
+    }
+
+    if (TCL_OK != TclStringRepeat(interp, objv[1], count, &resultPtr)) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringRplcCmd --
+ *
+ *	This procedure is invoked to process the "string replace" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringRplcCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_UniChar *ustring;
+    int first, last, length;
+
+    if (objc < 4 || objc > 5) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string first last ?string?");
+	return TCL_ERROR;
+    }
+
+    ustring = Tcl_GetUnicodeFromObj(objv[1], &length);
+    length--;
+
+    if (TclGetIntForIndexM(interp, objv[2], length, &first) != TCL_OK ||
+	    TclGetIntForIndexM(interp, objv[3], length, &last) != TCL_OK){
+	return TCL_ERROR;
+    }
+
+    if ((last < first) || (last < 0) || (first > length)) {
+	Tcl_SetObjResult(interp, objv[1]);
+    } else {
+	Tcl_Obj *resultPtr;
+
+	ustring = Tcl_GetUnicodeFromObj(objv[1], &length);
+	length--;
+
+	if (first < 0) {
+	    first = 0;
+	}
+
+	resultPtr = Tcl_NewUnicodeObj(ustring, first);
+	if (objc == 5) {
+	    Tcl_AppendObjToObj(resultPtr, objv[4]);
+	}
+	if (last < length) {
+	    Tcl_AppendUnicodeToObj(resultPtr, ustring + last + 1,
+		    length - last);
+	}
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringRevCmd --
+ *
+ *	This procedure is invoked to process the "string reverse" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringRevCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string");
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, TclStringObjReverse(objv[1]));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringStartCmd --
+ *
+ *	This procedure is invoked to process the "string wordstart" Tcl
+ *	command. See the user documentation for details on what it does. Note
+ *	that this command only functions correctly on properly formed Tcl UTF
+ *	strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringStartCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_UniChar ch = 0;
+    const char *p, *string;
+    int cur, index, length, numChars;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string index");
+	return TCL_ERROR;
+    }
+
+    string = TclGetStringFromObj(objv[1], &length);
+    numChars = Tcl_NumUtfChars(string, length);
+    if (TclGetIntForIndexM(interp, objv[2], numChars-1, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    string = TclGetStringFromObj(objv[1], &length);
+    if (index >= numChars) {
+	index = numChars - 1;
+    }
+    cur = 0;
+    if (index > 0) {
+	p = Tcl_UtfAtIndex(string, index);
+	for (cur = index; cur >= 0; cur--) {
+	    TclUtfToUniChar(p, &ch);
+	    if (!Tcl_UniCharIsWordChar(ch)) {
+		break;
+	    }
+	    p = Tcl_UtfPrev(p, string);
+	}
+	if (cur != index) {
+	    cur += 1;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(cur));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringEndCmd --
+ *
+ *	This procedure is invoked to process the "string wordend" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringEndCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_UniChar ch = 0;
+    const char *p, *end, *string;
+    int cur, index, length, numChars;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string index");
+	return TCL_ERROR;
+    }
+
+    string = TclGetStringFromObj(objv[1], &length);
+    numChars = Tcl_NumUtfChars(string, length);
+    if (TclGetIntForIndexM(interp, objv[2], numChars-1, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    string = TclGetStringFromObj(objv[1], &length);
+    if (index < 0) {
+	index = 0;
+    }
+    if (index < numChars) {
+	p = Tcl_UtfAtIndex(string, index);
+	end = string+length;
+	for (cur = index; p < end; cur++) {
+	    p += TclUtfToUniChar(p, &ch);
+	    if (!Tcl_UniCharIsWordChar(ch)) {
+		break;
+	    }
+	}
+	if (cur == index) {
+	    cur++;
+	}
+    } else {
+	cur = numChars;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(cur));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringEqualCmd --
+ *
+ *	This procedure is invoked to process the "string equal" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringEqualCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    /*
+     * Remember to keep code here in some sync with the byte-compiled versions
+     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
+     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
+     */
+
+    const char *string1, *string2;
+    int length1, length2, i, match, length, nocase = 0, reqlength = -1;
+    typedef int (*strCmpFn_t)(const char *, const char *, unsigned int);
+    strCmpFn_t strCmpFn;
+
+    if (objc < 3 || objc > 6) {
+    str_cmp_args:
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-nocase? ?-length int? string1 string2");
+	return TCL_ERROR;
+    }
+
+    for (i = 1; i < objc-2; i++) {
+	string2 = TclGetStringFromObj(objv[i], &length2);
+	if ((length2 > 1) && !strncmp(string2, "-nocase", (size_t)length2)) {
+	    nocase = 1;
+	} else if ((length2 > 1)
+		&& !strncmp(string2, "-length", (size_t)length2)) {
+	    if (i+1 >= objc-2) {
+		goto str_cmp_args;
+	    }
+	    i++;
+	    if (TclGetIntFromObj(interp, objv[i], &reqlength) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\": must be -nocase or -length",
+		    string2));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
+		    string2, NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * From now on, we only access the two objects at the end of the argument
+     * array.
+     */
+
+    objv += objc-2;
+
+    if ((reqlength == 0) || (objv[0] == objv[1])) {
+	/*
+	 * Always match at 0 chars of if it is the same obj.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(1));
+	return TCL_OK;
+    }
+
+    if (!nocase && TclIsPureByteArray(objv[0]) &&
+	    TclIsPureByteArray(objv[1])) {
+	/*
+	 * Use binary versions of comparisons since that won't cause undue
+	 * type conversions and it is much faster. Only do this if we're
+	 * case-sensitive (which is all that really makes sense with byte
+	 * arrays anyway, and we have no memcasecmp() for some reason... :^)
+	 */
+
+	string1 = (char *) Tcl_GetByteArrayFromObj(objv[0], &length1);
+	string2 = (char *) Tcl_GetByteArrayFromObj(objv[1], &length2);
+	strCmpFn = (strCmpFn_t) memcmp;
+    } else if ((objv[0]->typePtr == &tclStringType)
+	    && (objv[1]->typePtr == &tclStringType)) {
+	/*
+	 * Do a unicode-specific comparison if both of the args are of String
+	 * type. In benchmark testing this proved the most efficient check
+	 * between the unicode and string comparison operations.
+	 */
+
+	string1 = (char *) Tcl_GetUnicodeFromObj(objv[0], &length1);
+	string2 = (char *) Tcl_GetUnicodeFromObj(objv[1], &length2);
+	strCmpFn = (strCmpFn_t)
+		(nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp);
+    } else {
+	/*
+	 * As a catch-all we will work with UTF-8. We cannot use memcmp() as
+	 * that is unsafe with any string containing NUL (\xC0\x80 in Tcl's
+	 * utf rep). We can use the more efficient TclpUtfNcmp2 if we are
+	 * case-sensitive and no specific length was requested.
+	 */
+
+	string1 = (char *) TclGetStringFromObj(objv[0], &length1);
+	string2 = (char *) TclGetStringFromObj(objv[1], &length2);
+	if ((reqlength < 0) && !nocase) {
+	    strCmpFn = (strCmpFn_t) TclpUtfNcmp2;
+	} else {
+	    length1 = Tcl_NumUtfChars(string1, length1);
+	    length2 = Tcl_NumUtfChars(string2, length2);
+	    strCmpFn = (strCmpFn_t) (nocase ? Tcl_UtfNcasecmp : Tcl_UtfNcmp);
+	}
+    }
+
+    if ((reqlength < 0) && (length1 != length2)) {
+	match = 1;		/* This will be reversed below. */
+    } else {
+	length = (length1 < length2) ? length1 : length2;
+	if (reqlength > 0 && reqlength < length) {
+	    length = reqlength;
+	} else if (reqlength < 0) {
+	    /*
+	     * The requested length is negative, so we ignore it by setting it
+	     * to length + 1 so we correct the match var.
+	     */
+
+	    reqlength = length + 1;
+	}
+
+	match = strCmpFn(string1, string2, (unsigned) length);
+	if ((match == 0) && (reqlength > length)) {
+	    match = length1 - length2;
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(match ? 0 : 1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringCmpCmd --
+ *
+ *	This procedure is invoked to process the "string compare" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringCmpCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    /*
+     * Remember to keep code here in some sync with the byte-compiled versions
+     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
+     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
+     */
+
+    const char *string1, *string2;
+    int length1, length2, i, match, length, nocase = 0, reqlength = -1;
+    typedef int (*strCmpFn_t)(const char *, const char *, unsigned int);
+    strCmpFn_t strCmpFn;
+
+    if (objc < 3 || objc > 6) {
+    str_cmp_args:
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-nocase? ?-length int? string1 string2");
+	return TCL_ERROR;
+    }
+
+    for (i = 1; i < objc-2; i++) {
+	string2 = TclGetStringFromObj(objv[i], &length2);
+	if ((length2 > 1) && !strncmp(string2, "-nocase", (size_t)length2)) {
+	    nocase = 1;
+	} else if ((length2 > 1)
+		&& !strncmp(string2, "-length", (size_t)length2)) {
+	    if (i+1 >= objc-2) {
+		goto str_cmp_args;
+	    }
+	    i++;
+	    if (TclGetIntFromObj(interp, objv[i], &reqlength) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\": must be -nocase or -length",
+		    string2));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "option",
+		    string2, NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * From now on, we only access the two objects at the end of the argument
+     * array.
+     */
+
+    objv += objc-2;
+
+    if ((reqlength == 0) || (objv[0] == objv[1])) {
+	/*
+	 * Always match at 0 chars of if it is the same obj.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(0));
+	return TCL_OK;
+    }
+
+    if (!nocase && TclIsPureByteArray(objv[0]) &&
+	    TclIsPureByteArray(objv[1])) {
+	/*
+	 * Use binary versions of comparisons since that won't cause undue
+	 * type conversions and it is much faster. Only do this if we're
+	 * case-sensitive (which is all that really makes sense with byte
+	 * arrays anyway, and we have no memcasecmp() for some reason... :^)
+	 */
+
+	string1 = (char *) Tcl_GetByteArrayFromObj(objv[0], &length1);
+	string2 = (char *) Tcl_GetByteArrayFromObj(objv[1], &length2);
+	strCmpFn = (strCmpFn_t) memcmp;
+    } else if ((objv[0]->typePtr == &tclStringType)
+	    && (objv[1]->typePtr == &tclStringType)) {
+	/*
+	 * Do a unicode-specific comparison if both of the args are of String
+	 * type. In benchmark testing this proved the most efficient check
+	 * between the unicode and string comparison operations.
+	 */
+
+	string1 = (char *) Tcl_GetUnicodeFromObj(objv[0], &length1);
+	string2 = (char *) Tcl_GetUnicodeFromObj(objv[1], &length2);
+	strCmpFn = (strCmpFn_t)
+		(nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp);
+    } else {
+	/*
+	 * As a catch-all we will work with UTF-8. We cannot use memcmp() as
+	 * that is unsafe with any string containing NUL (\xC0\x80 in Tcl's
+	 * utf rep). We can use the more efficient TclpUtfNcmp2 if we are
+	 * case-sensitive and no specific length was requested.
+	 */
+
+	string1 = (char *) TclGetStringFromObj(objv[0], &length1);
+	string2 = (char *) TclGetStringFromObj(objv[1], &length2);
+	if ((reqlength < 0) && !nocase) {
+	    strCmpFn = (strCmpFn_t) TclpUtfNcmp2;
+	} else {
+	    length1 = Tcl_NumUtfChars(string1, length1);
+	    length2 = Tcl_NumUtfChars(string2, length2);
+	    strCmpFn = (strCmpFn_t) (nocase ? Tcl_UtfNcasecmp : Tcl_UtfNcmp);
+	}
+    }
+
+    length = (length1 < length2) ? length1 : length2;
+    if (reqlength > 0 && reqlength < length) {
+	length = reqlength;
+    } else if (reqlength < 0) {
+	/*
+	 * The requested length is negative, so we ignore it by setting it to
+	 * length + 1 so we correct the match var.
+	 */
+
+	reqlength = length + 1;
+    }
+
+    match = strCmpFn(string1, string2, (unsigned) length);
+    if ((match == 0) && (reqlength > length)) {
+	match = length1 - length2;
+    }
+
+    Tcl_SetObjResult(interp,
+	    Tcl_NewIntObj((match > 0) ? 1 : (match < 0) ? -1 : 0));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringCatCmd --
+ *
+ *	This procedure is invoked to process the "string cat" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringCatCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int code;
+    Tcl_Obj *objResultPtr;
+
+    if (objc < 2) {
+	/*
+	 * If there are no args, the result is an empty object.
+	 * Just leave the preset empty interp result.
+	 */
+	return TCL_OK;
+    }
+    if (objc == 2) {
+	/*
+	 * Other trivial case, single arg, just return it.
+	 */
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    code = TclStringCatObjv(interp, /* inPlace */ 1, objc-1, objv+1,
+	    &objResultPtr);
+
+    if (code == TCL_OK) {
+	Tcl_SetObjResult(interp, objResultPtr);
+	return TCL_OK;
+    }
+
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringBytesCmd --
+ *
+ *	This procedure is invoked to process the "string bytelength" Tcl
+ *	command. See the user documentation for details on what it does. Note
+ *	that this command only functions correctly on properly formed Tcl UTF
+ *	strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringBytesCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string");
+	return TCL_ERROR;
+    }
+
+    (void) TclGetStringFromObj(objv[1], &length);
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(length));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringLenCmd --
+ *
+ *	This procedure is invoked to process the "string length" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringLenCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string");
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(Tcl_GetCharLength(objv[1])));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringLowerCmd --
+ *
+ *	This procedure is invoked to process the "string tolower" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringLowerCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length1, length2;
+    const char *string1;
+    char *string2;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?first? ?last?");
+	return TCL_ERROR;
+    }
+
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    if (objc == 2) {
+	Tcl_Obj *resultPtr = Tcl_NewStringObj(string1, length1);
+
+	length1 = Tcl_UtfToLower(TclGetString(resultPtr));
+	Tcl_SetObjLength(resultPtr, length1);
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	int first, last;
+	const char *start, *end;
+	Tcl_Obj *resultPtr;
+
+	length1 = Tcl_NumUtfChars(string1, length1) - 1;
+	if (TclGetIntForIndexM(interp,objv[2],length1, &first) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (first < 0) {
+	    first = 0;
+	}
+	last = first;
+
+	if ((objc == 4) && (TclGetIntForIndexM(interp, objv[3], length1,
+		&last) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+
+	if (last >= length1) {
+	    last = length1;
+	}
+	if (last < first) {
+	    Tcl_SetObjResult(interp, objv[1]);
+	    return TCL_OK;
+	}
+
+	string1 = TclGetStringFromObj(objv[1], &length1);
+	start = Tcl_UtfAtIndex(string1, first);
+	end = Tcl_UtfAtIndex(start, last - first + 1);
+	resultPtr = Tcl_NewStringObj(string1, end - string1);
+	string2 = TclGetString(resultPtr) + (start - string1);
+
+	length2 = Tcl_UtfToLower(string2);
+	Tcl_SetObjLength(resultPtr, length2 + (start - string1));
+
+	Tcl_AppendToObj(resultPtr, end, -1);
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringUpperCmd --
+ *
+ *	This procedure is invoked to process the "string toupper" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringUpperCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length1, length2;
+    const char *string1;
+    char *string2;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?first? ?last?");
+	return TCL_ERROR;
+    }
+
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    if (objc == 2) {
+	Tcl_Obj *resultPtr = Tcl_NewStringObj(string1, length1);
+
+	length1 = Tcl_UtfToUpper(TclGetString(resultPtr));
+	Tcl_SetObjLength(resultPtr, length1);
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	int first, last;
+	const char *start, *end;
+	Tcl_Obj *resultPtr;
+
+	length1 = Tcl_NumUtfChars(string1, length1) - 1;
+	if (TclGetIntForIndexM(interp,objv[2],length1, &first) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (first < 0) {
+	    first = 0;
+	}
+	last = first;
+
+	if ((objc == 4) && (TclGetIntForIndexM(interp, objv[3], length1,
+		&last) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+
+	if (last >= length1) {
+	    last = length1;
+	}
+	if (last < first) {
+	    Tcl_SetObjResult(interp, objv[1]);
+	    return TCL_OK;
+	}
+
+	string1 = TclGetStringFromObj(objv[1], &length1);
+	start = Tcl_UtfAtIndex(string1, first);
+	end = Tcl_UtfAtIndex(start, last - first + 1);
+	resultPtr = Tcl_NewStringObj(string1, end - string1);
+	string2 = TclGetString(resultPtr) + (start - string1);
+
+	length2 = Tcl_UtfToUpper(string2);
+	Tcl_SetObjLength(resultPtr, length2 + (start - string1));
+
+	Tcl_AppendToObj(resultPtr, end, -1);
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTitleCmd --
+ *
+ *	This procedure is invoked to process the "string totitle" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringTitleCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length1, length2;
+    const char *string1;
+    char *string2;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?first? ?last?");
+	return TCL_ERROR;
+    }
+
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    if (objc == 2) {
+	Tcl_Obj *resultPtr = Tcl_NewStringObj(string1, length1);
+
+	length1 = Tcl_UtfToTitle(TclGetString(resultPtr));
+	Tcl_SetObjLength(resultPtr, length1);
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	int first, last;
+	const char *start, *end;
+	Tcl_Obj *resultPtr;
+
+	length1 = Tcl_NumUtfChars(string1, length1) - 1;
+	if (TclGetIntForIndexM(interp,objv[2],length1, &first) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (first < 0) {
+	    first = 0;
+	}
+	last = first;
+
+	if ((objc == 4) && (TclGetIntForIndexM(interp, objv[3], length1,
+		&last) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+
+	if (last >= length1) {
+	    last = length1;
+	}
+	if (last < first) {
+	    Tcl_SetObjResult(interp, objv[1]);
+	    return TCL_OK;
+	}
+
+	string1 = TclGetStringFromObj(objv[1], &length1);
+	start = Tcl_UtfAtIndex(string1, first);
+	end = Tcl_UtfAtIndex(start, last - first + 1);
+	resultPtr = Tcl_NewStringObj(string1, end - string1);
+	string2 = TclGetString(resultPtr) + (start - string1);
+
+	length2 = Tcl_UtfToTitle(string2);
+	Tcl_SetObjLength(resultPtr, length2 + (start - string1));
+
+	Tcl_AppendToObj(resultPtr, end, -1);
+	Tcl_SetObjResult(interp, resultPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTrimCmd --
+ *
+ *	This procedure is invoked to process the "string trim" Tcl command.
+ *	See the user documentation for details on what it does. Note that this
+ *	command only functions correctly on properly formed Tcl UTF strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringTrimCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *string1, *string2;
+    int triml, trimr, length1, length2;
+
+    if (objc == 3) {
+	string2 = TclGetStringFromObj(objv[2], &length2);
+    } else if (objc == 2) {
+	string2 = tclDefaultTrimSet;
+	length2 = strlen(tclDefaultTrimSet);
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?chars?");
+	return TCL_ERROR;
+    }
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    triml = TclTrimLeft(string1, length1, string2, length2);
+    trimr = TclTrimRight(string1 + triml, length1 - triml, string2, length2);
+
+    Tcl_SetObjResult(interp,
+	    Tcl_NewStringObj(string1 + triml, length1 - triml - trimr));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTrimLCmd --
+ *
+ *	This procedure is invoked to process the "string trimleft" Tcl
+ *	command. See the user documentation for details on what it does. Note
+ *	that this command only functions correctly on properly formed Tcl UTF
+ *	strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringTrimLCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *string1, *string2;
+    int trim, length1, length2;
+
+    if (objc == 3) {
+	string2 = TclGetStringFromObj(objv[2], &length2);
+    } else if (objc == 2) {
+	string2 = tclDefaultTrimSet;
+	length2 = strlen(tclDefaultTrimSet);
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?chars?");
+	return TCL_ERROR;
+    }
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    trim = TclTrimLeft(string1, length1, string2, length2);
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(string1+trim, length1-trim));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTrimRCmd --
+ *
+ *	This procedure is invoked to process the "string trimright" Tcl
+ *	command. See the user documentation for details on what it does. Note
+ *	that this command only functions correctly on properly formed Tcl UTF
+ *	strings.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringTrimRCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *string1, *string2;
+    int trim, length1, length2;
+
+    if (objc == 3) {
+	string2 = TclGetStringFromObj(objv[2], &length2);
+    } else if (objc == 2) {
+	string2 = tclDefaultTrimSet;
+	length2 = strlen(tclDefaultTrimSet);
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "string ?chars?");
+	return TCL_ERROR;
+    }
+    string1 = TclGetStringFromObj(objv[1], &length1);
+
+    trim = TclTrimRight(string1, length1, string2, length2);
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(string1, length1-trim));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitStringCmd --
+ *
+ *	This procedure creates the "string" Tcl command. See the user
+ *	documentation for details on what it does. Note that this command only
+ *	functions correctly on properly formed Tcl UTF strings.
+ *
+ *	Also note that the primary methods here (equal, compare, match, ...)
+ *	have bytecode equivalents. You will find the code for those in
+ *	tclExecute.c. The code here will only be used in the non-bc case (like
+ *	in an 'eval').
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitStringCmd(
+    Tcl_Interp *interp)		/* Current interpreter. */
+{
+    static const EnsembleImplMap stringImplMap[] = {
+	{"bytelength",	StringBytesCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"cat",		StringCatCmd,	TclCompileStringCatCmd, NULL, NULL, 0},
+	{"compare",	StringCmpCmd,	TclCompileStringCmpCmd, NULL, NULL, 0},
+	{"equal",	StringEqualCmd,	TclCompileStringEqualCmd, NULL, NULL, 0},
+	{"first",	StringFirstCmd,	TclCompileStringFirstCmd, NULL, NULL, 0},
+	{"index",	StringIndexCmd,	TclCompileStringIndexCmd, NULL, NULL, 0},
+	{"is",		StringIsCmd,	TclCompileStringIsCmd, NULL, NULL, 0},
+	{"last",	StringLastCmd,	TclCompileStringLastCmd, NULL, NULL, 0},
+	{"length",	StringLenCmd,	TclCompileStringLenCmd, NULL, NULL, 0},
+	{"map",		StringMapCmd,	TclCompileStringMapCmd, NULL, NULL, 0},
+	{"match",	StringMatchCmd,	TclCompileStringMatchCmd, NULL, NULL, 0},
+	{"range",	StringRangeCmd,	TclCompileStringRangeCmd, NULL, NULL, 0},
+	{"repeat",	StringReptCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"replace",	StringRplcCmd,	TclCompileStringReplaceCmd, NULL, NULL, 0},
+	{"reverse",	StringRevCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"tolower",	StringLowerCmd,	TclCompileStringToLowerCmd, NULL, NULL, 0},
+	{"toupper",	StringUpperCmd,	TclCompileStringToUpperCmd, NULL, NULL, 0},
+	{"totitle",	StringTitleCmd,	TclCompileStringToTitleCmd, NULL, NULL, 0},
+	{"trim",	StringTrimCmd,	TclCompileStringTrimCmd, NULL, NULL, 0},
+	{"trimleft",	StringTrimLCmd,	TclCompileStringTrimLCmd, NULL, NULL, 0},
+	{"trimright",	StringTrimRCmd,	TclCompileStringTrimRCmd, NULL, NULL, 0},
+	{"wordend",	StringEndCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"wordstart",	StringStartCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{NULL, NULL, NULL, NULL, NULL, 0}
+    };
+
+    return TclMakeEnsemble(interp, "string", stringImplMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SubstObjCmd --
+ *
+ *	This procedure is invoked to process the "subst" Tcl command. See the
+ *	user documentation for details on what it does. This command relies on
+ *	Tcl_SubstObj() for its implementation.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSubstOptions(
+    Tcl_Interp *interp,
+    int numOpts,
+    Tcl_Obj *const opts[],
+    int *flagPtr)
+{
+    static const char *const substOptions[] = {
+	"-nobackslashes", "-nocommands", "-novariables", NULL
+    };
+    enum {
+	SUBST_NOBACKSLASHES, SUBST_NOCOMMANDS, SUBST_NOVARS
+    };
+    int i, flags = TCL_SUBST_ALL;
+
+    for (i = 0; i < numOpts; i++) {
+	int optionIndex;
+
+	if (Tcl_GetIndexFromObj(interp, opts[i], substOptions, "option", 0,
+		&optionIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (optionIndex) {
+	case SUBST_NOBACKSLASHES:
+	    flags &= ~TCL_SUBST_BACKSLASHES;
+	    break;
+	case SUBST_NOCOMMANDS:
+	    flags &= ~TCL_SUBST_COMMANDS;
+	    break;
+	case SUBST_NOVARS:
+	    flags &= ~TCL_SUBST_VARIABLES;
+	    break;
+	default:
+	    Tcl_Panic("Tcl_SubstObjCmd: bad option index to SubstOptions");
+	}
+    }
+    *flagPtr = flags;
+    return TCL_OK;
+}
+
+int
+Tcl_SubstObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRSubstObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRSubstObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int flags;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-nobackslashes? ?-nocommands? ?-novariables? string");
+	return TCL_ERROR;
+    }
+
+    if (TclSubstOptions(interp, objc-2, objv+1, &flags) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return Tcl_NRSubstObj(interp, objv[objc-1], flags);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SwitchObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "switch" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SwitchObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRSwitchObjCmd, dummy, objc, objv);
+}
+int
+TclNRSwitchObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int i,j, index, mode, foundmode, splitObjs, numMatchesSaved;
+    int noCase, patternLength;
+    const char *pattern;
+    Tcl_Obj *stringObj, *indexVarObj, *matchVarObj;
+    Tcl_Obj *const *savedObjv = objv;
+    Tcl_RegExp regExpr = NULL;
+    Interp *iPtr = (Interp *) interp;
+    int pc = 0;
+    int bidx = 0;		/* Index of body argument. */
+    Tcl_Obj *blist = NULL;	/* List obj which is the body */
+    CmdFrame *ctxPtr;		/* Copy of the topmost cmdframe, to allow us
+				 * to mess with the line information */
+
+    /*
+     * If you add options that make -e and -g not unique prefixes of -exact or
+     * -glob, you *must* fix TclCompileSwitchCmd's option parser as well.
+     */
+
+    static const char *const options[] = {
+	"-exact", "-glob", "-indexvar", "-matchvar", "-nocase", "-regexp",
+	"--", NULL
+    };
+    enum options {
+	OPT_EXACT, OPT_GLOB, OPT_INDEXV, OPT_MATCHV, OPT_NOCASE, OPT_REGEXP,
+	OPT_LAST
+    };
+    typedef int (*strCmpFn_t)(const char *, const char *);
+    strCmpFn_t strCmpFn = strcmp;
+
+    mode = OPT_EXACT;
+    foundmode = 0;
+    indexVarObj = NULL;
+    matchVarObj = NULL;
+    numMatchesSaved = 0;
+    noCase = 0;
+    for (i = 1; i < objc-2; i++) {
+	if (TclGetString(objv[i])[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum options) index) {
+	    /*
+	     * General options.
+	     */
+
+	case OPT_LAST:
+	    i++;
+	    goto finishedOptions;
+	case OPT_NOCASE:
+	    strCmpFn = TclUtfCasecmp;
+	    noCase = 1;
+	    break;
+
+	    /*
+	     * Handle the different switch mode options.
+	     */
+
+	default:
+	    if (foundmode) {
+		/*
+		 * Mode already set via -exact, -glob, or -regexp.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"bad option \"%s\": %s option already found",
+			TclGetString(objv[i]), options[mode]));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+			"DOUBLEOPT", NULL);
+		return TCL_ERROR;
+	    }
+	    foundmode = 1;
+	    mode = index;
+	    break;
+
+	    /*
+	     * Check for TIP#75 options specifying the variables to write
+	     * regexp information into.
+	     */
+
+	case OPT_INDEXV:
+	    i++;
+	    if (i >= objc-2) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"missing variable name argument to %s option",
+			"-indexvar"));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+			"NOVAR", NULL);
+		return TCL_ERROR;
+	    }
+	    indexVarObj = objv[i];
+	    numMatchesSaved = -1;
+	    break;
+	case OPT_MATCHV:
+	    i++;
+	    if (i >= objc-2) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"missing variable name argument to %s option",
+			"-matchvar"));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+			"NOVAR", NULL);
+		return TCL_ERROR;
+	    }
+	    matchVarObj = objv[i];
+	    numMatchesSaved = -1;
+	    break;
+	}
+    }
+
+  finishedOptions:
+    if (objc - i < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-option ...? string ?pattern body ...? ?default body?");
+	return TCL_ERROR;
+    }
+    if (indexVarObj != NULL && mode != OPT_REGEXP) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s option requires -regexp option", "-indexvar"));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+		"MODERESTRICTION", NULL);
+	return TCL_ERROR;
+    }
+    if (matchVarObj != NULL && mode != OPT_REGEXP) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s option requires -regexp option", "-matchvar"));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+		"MODERESTRICTION", NULL);
+	return TCL_ERROR;
+    }
+
+    stringObj = objv[i];
+    objc -= i + 1;
+    objv += i + 1;
+    bidx = i + 1;		/* First after the match string. */
+
+    /*
+     * If all of the pattern/command pairs are lumped into a single argument,
+     * split them out again.
+     *
+     * TIP #280: Determine the lines the words in the list start at, based on
+     * the same data for the list word itself. The cmdFramePtr line
+     * information is manipulated directly.
+     */
+
+    splitObjs = 0;
+    if (objc == 1) {
+	Tcl_Obj **listv;
+
+	blist = objv[0];
+	if (TclListObjGetElements(interp, objv[0], &objc, &listv) != TCL_OK){
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Ensure that the list is non-empty.
+	 */
+
+	if (objc < 1) {
+	    Tcl_WrongNumArgs(interp, 1, savedObjv,
+		    "?-option ...? string {?pattern body ...? ?default body?}");
+	    return TCL_ERROR;
+	}
+	objv = listv;
+	splitObjs = 1;
+    }
+
+    /*
+     * Complain if there is an odd number of words in the list of patterns and
+     * bodies.
+     */
+
+    if (objc % 2) {
+	Tcl_ResetResult(interp);
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"extra switch pattern with no body", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH", "BADARM",
+		NULL);
+
+	/*
+	 * Check if this can be due to a badly placed comment in the switch
+	 * block.
+	 *
+	 * The following is an heuristic to detect the infamous "comment in
+	 * switch" error: just check if a pattern begins with '#'.
+	 */
+
+	if (splitObjs) {
+	    for (i=0 ; i<objc ; i+=2) {
+		if (TclGetString(objv[i])[0] == '#') {
+		    Tcl_AppendToObj(Tcl_GetObjResult(interp),
+			    ", this may be due to a comment incorrectly"
+			    " placed outside of a switch body - see the"
+			    " \"switch\" documentation", -1);
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH",
+			    "BADARM", "COMMENT?", NULL);
+		    break;
+		}
+	    }
+	}
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Complain if the last body is a continuation. Note that this check
+     * assumes that the list is non-empty!
+     */
+
+    if (strcmp(TclGetString(objv[objc-1]), "-") == 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"no body specified for pattern \"%s\"",
+		TclGetString(objv[objc-2])));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "SWITCH", "BADARM",
+		"FALLTHROUGH", NULL);
+	return TCL_ERROR;
+    }
+
+    for (i = 0; i < objc; i += 2) {
+	/*
+	 * See if the pattern matches the string.
+	 */
+
+	pattern = TclGetStringFromObj(objv[i], &patternLength);
+
+	if ((i == objc - 2) && (*pattern == 'd')
+		&& (strcmp(pattern, "default") == 0)) {
+	    Tcl_Obj *emptyObj = NULL;
+
+	    /*
+	     * If either indexVarObj or matchVarObj are non-NULL, we're in
+	     * REGEXP mode but have reached the default clause anyway. TIP#75
+	     * specifies that we set the variables to empty lists (== empty
+	     * objects) in that case.
+	     */
+
+	    if (indexVarObj != NULL) {
+		TclNewObj(emptyObj);
+		if (Tcl_ObjSetVar2(interp, indexVarObj, NULL, emptyObj,
+			TCL_LEAVE_ERR_MSG) == NULL) {
+		    return TCL_ERROR;
+		}
+	    }
+	    if (matchVarObj != NULL) {
+		if (emptyObj == NULL) {
+		    TclNewObj(emptyObj);
+		}
+		if (Tcl_ObjSetVar2(interp, matchVarObj, NULL, emptyObj,
+			TCL_LEAVE_ERR_MSG) == NULL) {
+		    return TCL_ERROR;
+		}
+	    }
+	    goto matchFound;
+	}
+
+	switch (mode) {
+	case OPT_EXACT:
+	    if (strCmpFn(TclGetString(stringObj), pattern) == 0) {
+		goto matchFound;
+	    }
+	    break;
+	case OPT_GLOB:
+	    if (Tcl_StringCaseMatch(TclGetString(stringObj),pattern,noCase)) {
+		goto matchFound;
+	    }
+	    break;
+	case OPT_REGEXP:
+	    regExpr = Tcl_GetRegExpFromObj(interp, objv[i],
+		    TCL_REG_ADVANCED | (noCase ? TCL_REG_NOCASE : 0));
+	    if (regExpr == NULL) {
+		return TCL_ERROR;
+	    } else {
+		int matched = Tcl_RegExpExecObj(interp, regExpr, stringObj, 0,
+			numMatchesSaved, 0);
+
+		if (matched < 0) {
+		    return TCL_ERROR;
+		} else if (matched) {
+		    goto matchFoundRegexp;
+		}
+	    }
+	    break;
+	}
+    }
+    return TCL_OK;
+
+  matchFoundRegexp:
+    /*
+     * We are operating in REGEXP mode and we need to store information about
+     * what we matched in some user-nominated arrays. So build the lists of
+     * values and indices to write here. [TIP#75]
+     */
+
+    if (numMatchesSaved) {
+	Tcl_RegExpInfo info;
+	Tcl_Obj *matchesObj, *indicesObj = NULL;
+
+	Tcl_RegExpGetInfo(regExpr, &info);
+	if (matchVarObj != NULL) {
+	    TclNewObj(matchesObj);
+	} else {
+	    matchesObj = NULL;
+	}
+	if (indexVarObj != NULL) {
+	    TclNewObj(indicesObj);
+	}
+
+	for (j=0 ; j<=info.nsubs ; j++) {
+	    if (indexVarObj != NULL) {
+		Tcl_Obj *rangeObjAry[2];
+
+		if (info.matches[j].end > 0) {
+		    rangeObjAry[0] = Tcl_NewLongObj(info.matches[j].start);
+		    rangeObjAry[1] = Tcl_NewLongObj(info.matches[j].end-1);
+		} else {
+		    rangeObjAry[0] = rangeObjAry[1] = Tcl_NewIntObj(-1);
+		}
+
+		/*
+		 * Never fails; the object is always clean at this point.
+		 */
+
+		Tcl_ListObjAppendElement(NULL, indicesObj,
+			Tcl_NewListObj(2, rangeObjAry));
+	    }
+
+	    if (matchVarObj != NULL) {
+		Tcl_Obj *substringObj;
+
+		substringObj = Tcl_GetRange(stringObj,
+			info.matches[j].start, info.matches[j].end-1);
+
+		/*
+		 * Never fails; the object is always clean at this point.
+		 */
+
+		Tcl_ListObjAppendElement(NULL, matchesObj, substringObj);
+	    }
+	}
+
+	if (indexVarObj != NULL) {
+	    if (Tcl_ObjSetVar2(interp, indexVarObj, NULL, indicesObj,
+		    TCL_LEAVE_ERR_MSG) == NULL) {
+		/*
+		 * Careful! Check to see if we have allocated the list of
+		 * matched strings; if so (but there was an error assigning
+		 * the indices list) we have a potential memory leak because
+		 * the match list has not been written to a variable. Except
+		 * that we'll clean that up right now.
+		 */
+
+		if (matchesObj != NULL) {
+		    Tcl_DecrRefCount(matchesObj);
+		}
+		return TCL_ERROR;
+	    }
+	}
+	if (matchVarObj != NULL) {
+	    if (Tcl_ObjSetVar2(interp, matchVarObj, NULL, matchesObj,
+		    TCL_LEAVE_ERR_MSG) == NULL) {
+		/*
+		 * Unlike above, if indicesObj is non-NULL at this point, it
+		 * will have been written to a variable already and will hence
+		 * not be leaked.
+		 */
+
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    /*
+     * We've got a match. Find a body to execute, skipping bodies that are
+     * "-".
+     */
+
+  matchFound:
+    ctxPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+    *ctxPtr = *iPtr->cmdFramePtr;
+
+    if (splitObjs) {
+	/*
+	 * We have to perform the GetSrc and other type dependent handling of
+	 * the frame here because we are munging with the line numbers,
+	 * something the other commands like if, etc. are not doing. Them are
+	 * fine with simply passing the CmdFrame through and having the
+	 * special handling done in 'info frame', or the bc compiler
+	 */
+
+	if (ctxPtr->type == TCL_LOCATION_BC) {
+	    /*
+	     * Type BC => ctxPtr->data.eval.path    is not used.
+	     *		  ctxPtr->data.tebc.codePtr is used instead.
+	     */
+
+	    TclGetSrcInfoForPc(ctxPtr);
+	    pc = 1;
+
+	    /*
+	     * The line information in the cmdFrame is now a copy we do not
+	     * own.
+	     */
+	}
+
+	if (ctxPtr->type == TCL_LOCATION_SOURCE && ctxPtr->line[bidx] >= 0) {
+	    int bline = ctxPtr->line[bidx];
+
+	    ctxPtr->line = ckalloc(objc * sizeof(int));
+	    ctxPtr->nline = objc;
+	    TclListLines(blist, bline, objc, ctxPtr->line, objv);
+	} else {
+	    /*
+	     * This is either a dynamic code word, when all elements are
+	     * relative to themselves, or something else less expected and
+	     * where we have no information. The result is the same in both
+	     * cases; tell the code to come that it doesn't know where it is,
+	     * which triggers reversion to the old behavior.
+	     */
+
+	    int k;
+
+	    ctxPtr->line = ckalloc(objc * sizeof(int));
+	    ctxPtr->nline = objc;
+	    for (k=0; k < objc; k++) {
+		ctxPtr->line[k] = -1;
+	    }
+	}
+    }
+
+    for (j = i + 1; ; j += 2) {
+	if (j >= objc) {
+	    /*
+	     * This shouldn't happen since we've checked that the last body is
+	     * not a continuation...
+	     */
+
+	    Tcl_Panic("fall-out when searching for body to match pattern");
+	}
+	if (strcmp(TclGetString(objv[j]), "-") != 0) {
+	    break;
+	}
+    }
+
+    /*
+     * TIP #280: Make invoking context available to switch branch.
+     */
+
+    Tcl_NRAddCallback(interp, SwitchPostProc, INT2PTR(splitObjs), ctxPtr,
+	    INT2PTR(pc), (ClientData) pattern);
+    return TclNREvalObjEx(interp, objv[j], 0, ctxPtr, splitObjs ? j : bidx+j);
+}
+
+static int
+SwitchPostProc(
+    ClientData data[],		/* Data passed from Tcl_NRAddCallback above */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int result)			/* Result to return*/
+{
+    /* Unpack the preserved data */
+
+    int splitObjs = PTR2INT(data[0]);
+    CmdFrame *ctxPtr = data[1];
+    int pc = PTR2INT(data[2]);
+    const char *pattern = data[3];
+    int patternLength = strlen(pattern);
+
+    /*
+     * Clean up TIP 280 context information
+     */
+
+    if (splitObjs) {
+	ckfree(ctxPtr->line);
+	if (pc && (ctxPtr->type == TCL_LOCATION_SOURCE)) {
+	    /*
+	     * Death of SrcInfo reference.
+	     */
+
+	    Tcl_DecrRefCount(ctxPtr->data.eval.path);
+	}
+    }
+
+    /*
+     * Generate an error message if necessary.
+     */
+
+    if (result == TCL_ERROR) {
+	int limit = 50;
+	int overflow = (patternLength > limit);
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"%.*s%s\" arm line %d)",
+		(overflow ? limit : patternLength), pattern,
+		(overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+    }
+    TclStackFree(interp, ctxPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ThrowObjCmd --
+ *
+ *	This procedure is invoked to process the "throw" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ThrowObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *options;
+    int len;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "type message");
+	return TCL_ERROR;
+    }
+
+    /*
+     * The type must be a list of at least length 1.
+     */
+
+    if (Tcl_ListObjLength(interp, objv[1], &len) != TCL_OK) {
+	return TCL_ERROR;
+    } else if (len < 1) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"type must be non-empty list", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "THROW", "BADEXCEPTION",
+		NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now prepare the result options dictionary. We use the list API as it is
+     * slightly more convenient.
+     */
+
+    TclNewLiteralStringObj(options, "-code error -level 0 -errorcode");
+    Tcl_ListObjAppendElement(NULL, options, objv[1]);
+
+    /*
+     * We're ready to go. Fire things into the low-level result machinery.
+     */
+
+    Tcl_SetObjResult(interp, objv[2]);
+    return Tcl_SetReturnOptions(interp, options);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TimeObjCmd --
+ *
+ *	This object-based procedure is invoked to process the "time" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_TimeObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Tcl_Obj *objPtr;
+    Tcl_Obj *objs[4];
+    register int i, result;
+    int count;
+    double totalMicroSec;
+#ifndef TCL_WIDE_CLICKS
+    Tcl_Time start, stop;
+#else
+    Tcl_WideInt start, stop;
+#endif
+
+    if (objc == 2) {
+	count = 1;
+    } else if (objc == 3) {
+	result = TclGetIntFromObj(interp, objv[2], &count);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "command ?count?");
+	return TCL_ERROR;
+    }
+
+    objPtr = objv[1];
+    i = count;
+#ifndef TCL_WIDE_CLICKS
+    Tcl_GetTime(&start);
+#else
+    start = TclpGetWideClicks();
+#endif
+    while (i-- > 0) {
+	result = Tcl_EvalObjEx(interp, objPtr, 0);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+#ifndef TCL_WIDE_CLICKS
+    Tcl_GetTime(&stop);
+    totalMicroSec = ((double) (stop.sec - start.sec)) * 1.0e6
+	    + (stop.usec - start.usec);
+#else
+    stop = TclpGetWideClicks();
+    totalMicroSec = ((double) TclpWideClicksToNanoseconds(stop - start))/1.0e3;
+#endif
+
+    if (count <= 1) {
+	/*
+	 * Use int obj since we know time is not fractional. [Bug 1202178]
+	 */
+
+	objs[0] = Tcl_NewIntObj((count <= 0) ? 0 : (int) totalMicroSec);
+    } else {
+	objs[0] = Tcl_NewDoubleObj(totalMicroSec/count);
+    }
+
+    /*
+     * Construct the result as a list because many programs have always parsed
+     * as such (extracting the first element, typically).
+     */
+
+    TclNewLiteralStringObj(objs[1], "microseconds");
+    TclNewLiteralStringObj(objs[2], "per");
+    TclNewLiteralStringObj(objs[3], "iteration");
+    Tcl_SetObjResult(interp, Tcl_NewListObj(4, objs));
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TryObjCmd, TclNRTryObjCmd --
+ *
+ *	This procedure is invoked to process the "try" Tcl command. See the
+ *	user documentation (or TIP #329) for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_TryObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRTryObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRTryObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *bodyObj, *handlersObj, *finallyObj = NULL;
+    int i, bodyShared, haveHandlers, dummy, code;
+    static const char *const handlerNames[] = {
+	"finally", "on", "trap", NULL
+    };
+    enum Handlers {
+	TryFinally, TryOn, TryTrap
+    };
+
+    /*
+     * Parse the arguments. The handlers are passed to subsequent callbacks as
+     * a Tcl_Obj list of the 5-tuples like (type, returnCode, errorCodePrefix,
+     * bindVariables, script), and the finally script is just passed as it is.
+     */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"body ?handler ...? ?finally script?");
+	return TCL_ERROR;
+    }
+    bodyObj = objv[1];
+    handlersObj = Tcl_NewObj();
+    bodyShared = 0;
+    haveHandlers = 0;
+    for (i=2 ; i<objc ; i++) {
+	int type;
+	Tcl_Obj *info[5];
+
+	if (Tcl_GetIndexFromObj(interp, objv[i], handlerNames, "handler type",
+		0, &type) != TCL_OK) {
+	    Tcl_DecrRefCount(handlersObj);
+	    return TCL_ERROR;
+	}
+	switch ((enum Handlers) type) {
+	case TryFinally:	/* finally script */
+	    if (i < objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"finally clause must be last", -1));
+		Tcl_DecrRefCount(handlersObj);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "FINALLY",
+			"NONTERMINAL", NULL);
+		return TCL_ERROR;
+	    } else if (i == objc-1) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"wrong # args to finally clause: must be"
+			" \"... finally script\"", -1));
+		Tcl_DecrRefCount(handlersObj);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "FINALLY",
+			"ARGUMENT", NULL);
+		return TCL_ERROR;
+	    }
+	    finallyObj = objv[++i];
+	    break;
+
+	case TryOn:		/* on code variableList script */
+	    if (i > objc-4) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"wrong # args to on clause: must be \"... on code"
+			" variableList script\"", -1));
+		Tcl_DecrRefCount(handlersObj);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "ON",
+			"ARGUMENT", NULL);
+		return TCL_ERROR;
+	    }
+	    if (TclGetCompletionCodeFromObj(interp, objv[i+1],
+		    &code) != TCL_OK) {
+		Tcl_DecrRefCount(handlersObj);
+		return TCL_ERROR;
+	    }
+	    info[2] = NULL;
+	    goto commonHandler;
+
+	case TryTrap:		/* trap pattern variableList script */
+	    if (i > objc-4) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"wrong # args to trap clause: "
+			"must be \"... trap pattern variableList script\"",
+			-1));
+		Tcl_DecrRefCount(handlersObj);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "TRAP",
+			"ARGUMENT", NULL);
+		return TCL_ERROR;
+	    }
+	    code = 1;
+	    if (Tcl_ListObjLength(NULL, objv[i+1], &dummy) != TCL_OK) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"bad prefix '%s': must be a list",
+			Tcl_GetString(objv[i+1])));
+		Tcl_DecrRefCount(handlersObj);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "TRAP",
+			"EXNFORMAT", NULL);
+		return TCL_ERROR;
+	    }
+	    info[2] = objv[i+1];
+
+	commonHandler:
+	    if (Tcl_ListObjLength(interp, objv[i+2], &dummy) != TCL_OK) {
+		Tcl_DecrRefCount(handlersObj);
+		return TCL_ERROR;
+	    }
+
+	    info[0] = objv[i];			/* type */
+	    TclNewLongObj(info[1], code);	/* returnCode */
+	    if (info[2] == NULL) {		/* errorCodePrefix */
+		TclNewObj(info[2]);
+	    }
+	    info[3] = objv[i+2];		/* bindVariables */
+	    info[4] = objv[i+3];		/* script */
+
+	    bodyShared = !strcmp(TclGetString(objv[i+3]), "-");
+	    Tcl_ListObjAppendElement(NULL, handlersObj,
+		    Tcl_NewListObj(5, info));
+	    haveHandlers = 1;
+	    i += 3;
+	    break;
+	}
+    }
+    if (bodyShared) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"last non-finally clause must not have a body of \"-\"", -1));
+	Tcl_DecrRefCount(handlersObj);
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRY", "BADFALLTHROUGH",
+		NULL);
+	return TCL_ERROR;
+    }
+    if (!haveHandlers) {
+	Tcl_DecrRefCount(handlersObj);
+	handlersObj = NULL;
+    }
+
+    /*
+     * Execute the body.
+     */
+
+    Tcl_NRAddCallback(interp, TryPostBody, handlersObj, finallyObj,
+	    (ClientData)objv, INT2PTR(objc));
+    return TclNREvalObjEx(interp, bodyObj, 0,
+	    ((Interp *) interp)->cmdFramePtr, 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * During --
+ *
+ *	This helper function patches together the updates to the interpreter's
+ *	return options that are needed when things fail during the processing
+ *	of a handler or finally script for the [try] command.
+ *
+ * Returns:
+ *	The new option dictionary.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline Tcl_Obj *
+During(
+    Tcl_Interp *interp,
+    int resultCode,		/* The result code from the just-evaluated
+				 * script. */
+    Tcl_Obj *oldOptions,	/* The old option dictionary. */
+    Tcl_Obj *errorInfo)		/* An object to append to the errorinfo and
+				 * release, or NULL if nothing is to be added.
+				 * Designed to be used with Tcl_ObjPrintf. */
+{
+    Tcl_Obj *during, *options;
+
+    if (errorInfo != NULL) {
+	Tcl_AppendObjToErrorInfo(interp, errorInfo);
+    }
+    options = Tcl_GetReturnOptions(interp, resultCode);
+    TclNewLiteralStringObj(during, "-during");
+    Tcl_IncrRefCount(during);
+    Tcl_DictObjPut(interp, options, during, oldOptions);
+    Tcl_DecrRefCount(during);
+    Tcl_IncrRefCount(options);
+    Tcl_DecrRefCount(oldOptions);
+    return options;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TryPostBody --
+ *
+ *	Callback to handle the outcome of the execution of the body of a 'try'
+ *	command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TryPostBody(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj *resultObj, *options, *handlersObj, *finallyObj, *cmdObj, **objv;
+    int i, dummy, code, objc;
+    int numHandlers = 0;
+
+    handlersObj = data[0];
+    finallyObj = data[1];
+    objv = data[2];
+    objc = PTR2INT(data[3]);
+
+    cmdObj = objv[0];
+
+    /*
+     * Check for limits/rewinding, which override normal trapping behaviour.
+     */
+
+    if (((Interp*) interp)->execEnvPtr->rewind || Tcl_LimitExceeded(interp)) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"%s\" body line %d)", TclGetString(cmdObj),
+		Tcl_GetErrorLine(interp)));
+	if (handlersObj != NULL) {
+	    Tcl_DecrRefCount(handlersObj);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Basic processing of the outcome of the script, including adding of
+     * errorinfo trace.
+     */
+
+    if (result == TCL_ERROR) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"%s\" body line %d)", TclGetString(cmdObj),
+		Tcl_GetErrorLine(interp)));
+    }
+    resultObj = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(resultObj);
+    options = Tcl_GetReturnOptions(interp, result);
+    Tcl_IncrRefCount(options);
+    Tcl_ResetResult(interp);
+
+    /*
+     * Handle the results.
+     */
+
+    if (handlersObj != NULL) {
+	int found = 0;
+	Tcl_Obj **handlers, **info;
+
+	Tcl_ListObjGetElements(NULL, handlersObj, &numHandlers, &handlers);
+	for (i=0 ; i<numHandlers ; i++) {
+	    Tcl_Obj *handlerBodyObj;
+
+	    Tcl_ListObjGetElements(NULL, handlers[i], &dummy, &info);
+	    if (!found) {
+		Tcl_GetIntFromObj(NULL, info[1], &code);
+		if (code != result) {
+		    continue;
+		}
+
+		/*
+		 * When processing an error, we must also perform list-prefix
+		 * matching of the errorcode list. However, if this was an
+		 * 'on' handler, the list that we are matching against will be
+		 * empty.
+		 */
+
+		if (code == TCL_ERROR) {
+		    Tcl_Obj *errorCodeName, *errcode, **bits1, **bits2;
+		    int len1, len2, j;
+
+		    TclNewLiteralStringObj(errorCodeName, "-errorcode");
+		    Tcl_DictObjGet(NULL, options, errorCodeName, &errcode);
+		    Tcl_DecrRefCount(errorCodeName);
+		    Tcl_ListObjGetElements(NULL, info[2], &len1, &bits1);
+		    if (Tcl_ListObjGetElements(NULL, errcode, &len2,
+			    &bits2) != TCL_OK) {
+			continue;
+		    }
+		    if (len2 < len1) {
+			continue;
+		    }
+		    for (j=0 ; j<len1 ; j++) {
+			if (strcmp(TclGetString(bits1[j]),
+				TclGetString(bits2[j])) != 0) {
+			    /*
+			     * Really want 'continue outerloop;', but C does
+			     * not give us that.
+			     */
+
+			    goto didNotMatch;
+			}
+		    }
+		}
+
+		found = 1;
+	    }
+
+	    /*
+	     * Now we need to scan forward over "-" bodies. Note that we've
+	     * already checked that the last body is not a "-", so this search
+	     * will terminate successfully.
+	     */
+
+	    if (!strcmp(TclGetString(info[4]), "-")) {
+		continue;
+	    }
+
+	    /*
+	     * Bind the variables. We already know this is a list of variable
+	     * names, but it might be empty.
+	     */
+
+	    Tcl_ResetResult(interp);
+	    result = TCL_ERROR;
+	    Tcl_ListObjLength(NULL, info[3], &dummy);
+	    if (dummy > 0) {
+		Tcl_Obj *varName;
+
+		Tcl_ListObjIndex(NULL, info[3], 0, &varName);
+		if (Tcl_ObjSetVar2(interp, varName, NULL, resultObj,
+			TCL_LEAVE_ERR_MSG) == NULL) {
+		    Tcl_DecrRefCount(resultObj);
+		    goto handlerFailed;
+		}
+		Tcl_DecrRefCount(resultObj);
+		if (dummy > 1) {
+		    Tcl_ListObjIndex(NULL, info[3], 1, &varName);
+		    if (Tcl_ObjSetVar2(interp, varName, NULL, options,
+			    TCL_LEAVE_ERR_MSG) == NULL) {
+			goto handlerFailed;
+		    }
+		}
+	    } else {
+		/*
+		 * Dispose of the result to prevent a memleak. [Bug 2910044]
+		 */
+
+		Tcl_DecrRefCount(resultObj);
+	    }
+
+	    /*
+	     * Evaluate the handler body and process the outcome. Note that we
+	     * need to keep the kind of handler for debugging purposes, and in
+	     * any case anything we want from info[] must be extracted right
+	     * now because the info[] array is about to become invalid. There
+	     * is very little refcount handling here however, since we know
+	     * that the objects that we still want to refer to now were input
+	     * arguments to [try] and so are still on the Tcl value stack.
+	     */
+
+	    handlerBodyObj = info[4];
+	    Tcl_NRAddCallback(interp, TryPostHandler, objv, options, info[0],
+		    INT2PTR((finallyObj == NULL) ? 0 : objc - 1));
+	    Tcl_DecrRefCount(handlersObj);
+	    return TclNREvalObjEx(interp, handlerBodyObj, 0,
+		    ((Interp *) interp)->cmdFramePtr, 4*i + 5);
+
+	handlerFailed:
+	    resultObj = Tcl_GetObjResult(interp);
+	    Tcl_IncrRefCount(resultObj);
+	    options = During(interp, result, options, NULL);
+	    break;
+
+	didNotMatch:
+	    continue;
+	}
+
+	/*
+	 * No handler matched; get rid of the list of handlers.
+	 */
+
+	Tcl_DecrRefCount(handlersObj);
+    }
+
+    /*
+     * Process the finally clause.
+     */
+
+    if (finallyObj != NULL) {
+	Tcl_NRAddCallback(interp, TryPostFinal, resultObj, options, cmdObj,
+		NULL);
+	return TclNREvalObjEx(interp, finallyObj, 0,
+		((Interp *) interp)->cmdFramePtr, objc - 1);
+    }
+
+    /*
+     * Install the correct result/options into the interpreter and clean up
+     * any temporary storage.
+     */
+
+    result = Tcl_SetReturnOptions(interp, options);
+    Tcl_DecrRefCount(options);
+    Tcl_SetObjResult(interp, resultObj);
+    Tcl_DecrRefCount(resultObj);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TryPostHandler --
+ *
+ *	Callback to handle the outcome of the execution of a handler of a
+ *	'try' command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TryPostHandler(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj *resultObj, *cmdObj, *options, *handlerKindObj, **objv;
+    Tcl_Obj *finallyObj;
+    int finally;
+
+    objv = data[0];
+    options = data[1];
+    handlerKindObj = data[2];
+    finally = PTR2INT(data[3]);
+
+    cmdObj = objv[0];
+    finallyObj = finally ? objv[finally] : 0;
+
+    /*
+     * Check for limits/rewinding, which override normal trapping behaviour.
+     */
+
+    if (((Interp*) interp)->execEnvPtr->rewind || Tcl_LimitExceeded(interp)) {
+	options = During(interp, result, options, Tcl_ObjPrintf(
+		"\n    (\"%s ... %s\" handler line %d)",
+		TclGetString(cmdObj), TclGetString(handlerKindObj),
+		Tcl_GetErrorLine(interp)));
+	Tcl_DecrRefCount(options);
+	return TCL_ERROR;
+    }
+
+    /*
+     * The handler result completely substitutes for the result of the body.
+     */
+
+    resultObj = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(resultObj);
+    if (result == TCL_ERROR) {
+	options = During(interp, result, options, Tcl_ObjPrintf(
+		"\n    (\"%s ... %s\" handler line %d)",
+		TclGetString(cmdObj), TclGetString(handlerKindObj),
+		Tcl_GetErrorLine(interp)));
+    } else {
+	Tcl_DecrRefCount(options);
+	options = Tcl_GetReturnOptions(interp, result);
+	Tcl_IncrRefCount(options);
+    }
+
+    /*
+     * Process the finally clause if it is present.
+     */
+
+    if (finallyObj != NULL) {
+	Interp *iPtr = (Interp *) interp;
+
+	Tcl_NRAddCallback(interp, TryPostFinal, resultObj, options, cmdObj,
+		NULL);
+
+	/* The 'finally' script is always the last argument word. */
+	return TclNREvalObjEx(interp, finallyObj, 0, iPtr->cmdFramePtr,
+		finally);
+    }
+
+    /*
+     * Install the correct result/options into the interpreter and clean up
+     * any temporary storage.
+     */
+
+    result = Tcl_SetReturnOptions(interp, options);
+    Tcl_DecrRefCount(options);
+    Tcl_SetObjResult(interp, resultObj);
+    Tcl_DecrRefCount(resultObj);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TryPostFinal --
+ *
+ *	Callback to handle the outcome of the execution of the finally script
+ *	of a 'try' command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TryPostFinal(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj *resultObj, *options, *cmdObj;
+
+    resultObj = data[0];
+    options = data[1];
+    cmdObj = data[2];
+
+    /*
+     * If the result wasn't OK, we need to adjust the result options.
+     */
+
+    if (result != TCL_OK) {
+	Tcl_DecrRefCount(resultObj);
+	resultObj = NULL;
+	if (result == TCL_ERROR) {
+	    options = During(interp, result, options, Tcl_ObjPrintf(
+		    "\n    (\"%s ... finally\" body line %d)",
+		    TclGetString(cmdObj), Tcl_GetErrorLine(interp)));
+	} else {
+	    Tcl_Obj *origOptions = options;
+
+	    options = Tcl_GetReturnOptions(interp, result);
+	    Tcl_IncrRefCount(options);
+	    Tcl_DecrRefCount(origOptions);
+	}
+    }
+
+    /*
+     * Install the correct result/options into the interpreter and clean up
+     * any temporary storage.
+     */
+
+    result = Tcl_SetReturnOptions(interp, options);
+    Tcl_DecrRefCount(options);
+    if (resultObj != NULL) {
+	Tcl_SetObjResult(interp, resultObj);
+	Tcl_DecrRefCount(resultObj);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_WhileObjCmd --
+ *
+ *	This procedure is invoked to process the "while" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	With the bytecode compiler, this procedure is only called when a
+ *	command name is computed at runtime, and is "while" or the name to
+ *	which "while" was renamed: e.g., "set z while; $z {$i<100} {}"
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_WhileObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRWhileObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRWhileObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    ForIterData *iterPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "test command");
+	return TCL_ERROR;
+    }
+
+    /*
+     * We reuse [for]'s callback, passing a NULL for the 'next' script.
+     */
+
+    TclSmallAllocEx(interp, sizeof(ForIterData), iterPtr);
+    iterPtr->cond = objv[1];
+    iterPtr->body = objv[2];
+    iterPtr->next = NULL;
+    iterPtr->msg  = "\n    (\"while\" body line %d)";
+    iterPtr->word = 2;
+
+    TclNRAddCallback(interp, TclNRForIterCallback, iterPtr, NULL,
+	    NULL, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclListLines --
+ *
+ *	???
+ *
+ * Results:
+ *	Filled in array of line numbers?
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclListLines(
+    Tcl_Obj *listObj,		/* Pointer to obj holding a string with list
+				 * structure. Assumed to be valid. Assumed to
+				 * contain n elements. */
+    int line,			/* Line the list as a whole starts on. */
+    int n,			/* #elements in lines */
+    int *lines,			/* Array of line numbers, to fill. */
+    Tcl_Obj *const *elems)      /* The list elems as Tcl_Obj*, in need of
+				 * derived continuation data */
+{
+    const char *listStr = Tcl_GetString(listObj);
+    const char *listHead = listStr;
+    int i, length = strlen(listStr);
+    const char *element = NULL, *next = NULL;
+    ContLineLoc *clLocPtr = TclContinuationsGet(listObj);
+    int *clNext = (clLocPtr ? &clLocPtr->loc[0] : NULL);
+
+    for (i = 0; i < n; i++) {
+	TclFindElement(NULL, listStr, length, &element, &next, NULL, NULL);
+
+	TclAdvanceLines(&line, listStr, element);
+				/* Leading whitespace */
+	TclAdvanceContinuations(&line, &clNext, element - listHead);
+	if (elems && clNext) {
+	    TclContinuationsEnterDerived(elems[i], element-listHead, clNext);
+	}
+	lines[i] = line;
+	length -= (next - listStr);
+	TclAdvanceLines(&line, element, next);
+				/* Element */
+	listStr = next;
+
+	if (*element == 0) {
+	    /* ASSERT i == n */
+	    break;
+	}
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCompCmds.c b/generic/tclCompCmds.c
new file mode 100644
index 0000000..b9bc228
--- /dev/null
+++ b/generic/tclCompCmds.c
@@ -0,0 +1,3599 @@
+/*
+ * tclCompCmds.c --
+ *
+ *	This file contains compilation procedures that compile various Tcl
+ *	commands into a sequence of instructions ("bytecodes").
+ *
+ * Copyright (c) 1997-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2002 ActiveState Corporation.
+ * Copyright (c) 2004-2013 by Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include <assert.h>
+
+/*
+ * Prototypes for procedures defined later in this file:
+ */
+
+static ClientData	DupDictUpdateInfo(ClientData clientData);
+static void		FreeDictUpdateInfo(ClientData clientData);
+static void		PrintDictUpdateInfo(ClientData clientData,
+			    Tcl_Obj *appendObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static void		DisassembleDictUpdateInfo(ClientData clientData,
+			    Tcl_Obj *dictObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static ClientData	DupForeachInfo(ClientData clientData);
+static void		FreeForeachInfo(ClientData clientData);
+static void		PrintForeachInfo(ClientData clientData,
+			    Tcl_Obj *appendObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static void		DisassembleForeachInfo(ClientData clientData,
+			    Tcl_Obj *dictObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static void		PrintNewForeachInfo(ClientData clientData,
+			    Tcl_Obj *appendObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static void		DisassembleNewForeachInfo(ClientData clientData,
+			    Tcl_Obj *dictObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static int		CompileEachloopCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    CompileEnv *envPtr, int collect);
+static int		CompileDictEachCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr, int collect);
+
+/*
+ * The structures below define the AuxData types defined in this file.
+ */
+
+static const AuxDataType foreachInfoType = {
+    "ForeachInfo",		/* name */
+    DupForeachInfo,		/* dupProc */
+    FreeForeachInfo,		/* freeProc */
+    PrintForeachInfo,		/* printProc */
+    DisassembleForeachInfo	/* disassembleProc */
+};
+
+static const AuxDataType newForeachInfoType = {
+    "NewForeachInfo",		/* name */
+    DupForeachInfo,		/* dupProc */
+    FreeForeachInfo,		/* freeProc */
+    PrintNewForeachInfo,	/* printProc */
+    DisassembleNewForeachInfo	/* disassembleProc */
+};
+
+static const AuxDataType dictUpdateInfoType = {
+    "DictUpdateInfo",		/* name */
+    DupDictUpdateInfo,		/* dupProc */
+    FreeDictUpdateInfo,		/* freeProc */
+    PrintDictUpdateInfo,	/* printProc */
+    DisassembleDictUpdateInfo	/* disassembleProc */
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetAuxDataType --
+ *
+ *	This procedure looks up an Auxdata type by name.
+ *
+ * Results:
+ *	If an AuxData type with name matching "typeName" is found, a pointer
+ *	to its AuxDataType structure is returned; otherwise, NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const AuxDataType *
+TclGetAuxDataType(
+    const char *typeName)	/* Name of AuxData type to look up. */
+{
+    if (!strcmp(typeName, foreachInfoType.name)) {
+	return &foreachInfoType;
+    } else if (!strcmp(typeName, newForeachInfoType.name)) {
+	return &newForeachInfoType;
+    } else if (!strcmp(typeName, dictUpdateInfoType.name)) {
+	return &dictUpdateInfoType;
+    } else if (!strcmp(typeName, tclJumptableInfoType.name)) {
+	return &tclJumptableInfoType;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileAppendCmd --
+ *
+ *	Procedure called to compile the "append" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "append" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileAppendCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr, *valueTokenPtr;
+    int isScalar, localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* TODO: Consider support for compiling expanded args. */
+    numWords = parsePtr->numWords;
+    if (numWords == 1) {
+	return TCL_ERROR;
+    } else if (numWords == 2) {
+	/*
+	 * append varName == set varName
+	 */
+
+	return TclCompileSetCmd(interp, parsePtr, cmdPtr, envPtr);
+    } else if (numWords > 3) {
+	/*
+	 * APPEND instructions currently only handle one value, but we can
+	 * handle some multi-value cases by stringing them together.
+	 */
+
+	goto appendMultiple;
+    }
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we need
+     * to emit code to compute and push the name at runtime. We use a frame
+     * slot (entry in the array of local vars) if we are compiling a procedure
+     * body and if the name is simple text that does not include namespace
+     * qualifiers.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+	    &localIndex, &isScalar, 1);
+
+    /*
+     * We are doing an assignment, otherwise TclCompileSetCmd was called, so
+     * push the new value. This will need to be extended to push a value for
+     * each argument.
+     */
+
+	valueTokenPtr = TokenAfter(varTokenPtr);
+	CompileWord(envPtr, valueTokenPtr, interp, 2);
+
+    /*
+     * Emit instructions to set/get the variable.
+     */
+
+	if (isScalar) {
+	    if (localIndex < 0) {
+		TclEmitOpcode(INST_APPEND_STK, envPtr);
+	    } else {
+		Emit14Inst(INST_APPEND_SCALAR, localIndex, envPtr);
+	    }
+	} else {
+	    if (localIndex < 0) {
+		TclEmitOpcode(INST_APPEND_ARRAY_STK, envPtr);
+	    } else {
+		Emit14Inst(INST_APPEND_ARRAY, localIndex, envPtr);
+	    }
+	}
+
+    return TCL_OK;
+
+  appendMultiple:
+    /*
+     * Can only handle the case where we are appending to a local scalar when
+     * there are multiple values to append.  Fortunately, this is common.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    localIndex = LocalScalarFromToken(varTokenPtr, envPtr);
+    if (localIndex < 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Definitely appending to a local scalar; generate the words and append
+     * them.
+     */
+
+    valueTokenPtr = TokenAfter(varTokenPtr);
+    for (i = 2 ; i < numWords ; i++) {
+	CompileWord(envPtr, valueTokenPtr, interp, i);
+	valueTokenPtr = TokenAfter(valueTokenPtr);
+    }
+    TclEmitInstInt4(	  INST_REVERSE, numWords-2,		envPtr);
+    for (i = 2 ; i < numWords ;) {
+	Emit14Inst(	  INST_APPEND_SCALAR, localIndex,	envPtr);
+	if (++i < numWords) {
+	    TclEmitOpcode(INST_POP,				envPtr);
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileArray*Cmd --
+ *
+ *	Functions called to compile "array" sucommands.
+ *
+ * Results:
+ *	All return TCL_OK for a successful compile, and TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "array" subcommand at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileArrayExistsCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    int isScalar, localIndex;
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    PushVarNameWord(interp, tokenPtr, envPtr, TCL_NO_ELEMENT,
+	    &localIndex, &isScalar, 1);
+    if (!isScalar) {
+	return TCL_ERROR;
+    }
+
+    if (localIndex >= 0) {
+	TclEmitInstInt4(INST_ARRAY_EXISTS_IMM, localIndex,	envPtr);
+    } else {
+	TclEmitOpcode(	INST_ARRAY_EXISTS_STK,			envPtr);
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileArraySetCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *varTokenPtr, *dataTokenPtr;
+    int isScalar, localIndex, code = TCL_OK;
+    int isDataLiteral, isDataValid, isDataEven, len;
+    int keyVar, valVar, infoIndex;
+    int fwd, offsetBack, offsetFwd;
+    Tcl_Obj *literalObj;
+    ForeachInfo *infoPtr;
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dataTokenPtr = TokenAfter(varTokenPtr);
+    literalObj = Tcl_NewObj();
+    isDataLiteral = TclWordKnownAtCompileTime(dataTokenPtr, literalObj);
+    isDataValid = (isDataLiteral
+	    && Tcl_ListObjLength(NULL, literalObj, &len) == TCL_OK);
+    isDataEven = (isDataValid && (len & 1) == 0);
+
+    /*
+     * Special case: literal odd-length argument is always an error.
+     */
+
+    if (isDataValid && !isDataEven) {
+	PushStringLiteral(envPtr, "list must have an even number of elements");
+	PushStringLiteral(envPtr, "-errorcode {TCL ARGUMENT FORMAT}");
+	TclEmitInstInt4(INST_RETURN_IMM, TCL_ERROR,		envPtr);
+	TclEmitInt4(		0,				envPtr);
+	goto done;
+    }
+
+    /*
+     * Except for the special "ensure array" case below, when we're not in
+     * a proc, we cannot do a better compile than generic.
+     */
+
+    if ((varTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) ||
+	    (envPtr->procPtr == NULL && !(isDataEven && len == 0))) {
+	code = TclCompileBasic2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+	goto done;
+    }
+
+    PushVarNameWord(interp, varTokenPtr, envPtr, TCL_NO_ELEMENT,
+	    &localIndex, &isScalar, 1);
+    if (!isScalar) {
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Special case: literal empty value argument is just an "ensure array"
+     * operation.
+     */
+
+    if (isDataEven && len == 0) {
+	if (localIndex >= 0) {
+	    TclEmitInstInt4(INST_ARRAY_EXISTS_IMM, localIndex,	envPtr);
+	    TclEmitInstInt1(INST_JUMP_TRUE1, 7,			envPtr);
+	    TclEmitInstInt4(INST_ARRAY_MAKE_IMM, localIndex,	envPtr);
+	} else {
+	    TclEmitOpcode(  INST_DUP,				envPtr);
+	    TclEmitOpcode(  INST_ARRAY_EXISTS_STK,		envPtr);
+	    TclEmitInstInt1(INST_JUMP_TRUE1, 5,			envPtr);
+	    TclEmitOpcode(  INST_ARRAY_MAKE_STK,		envPtr);
+	    TclEmitInstInt1(INST_JUMP1, 3,			envPtr);
+	    /* Each branch decrements stack depth, but we only take one. */
+	    TclAdjustStackDepth(1, envPtr);
+	    TclEmitOpcode(  INST_POP,				envPtr);
+	}
+	PushStringLiteral(envPtr, "");
+	goto done;
+    }
+
+    if (localIndex < 0) {
+	/*
+	 * a non-local variable: upvar from a local one! This consumes the
+	 * variable name that was left at stacktop.
+	 */
+
+	localIndex = TclFindCompiledLocal(varTokenPtr->start,
+		varTokenPtr->size, 1, envPtr);
+	PushStringLiteral(envPtr, "0");
+	TclEmitInstInt4(INST_REVERSE, 2,        		envPtr);
+	TclEmitInstInt4(INST_UPVAR, localIndex, 		envPtr);
+	TclEmitOpcode(INST_POP,          			envPtr);
+    }
+
+    /*
+     * Prepare for the internal foreach.
+     */
+
+    keyVar = AnonymousLocal(envPtr);
+    valVar = AnonymousLocal(envPtr);
+
+    infoPtr = ckalloc(sizeof(ForeachInfo));
+    infoPtr->numLists = 1;
+    infoPtr->varLists[0] = ckalloc(sizeof(ForeachVarList) + sizeof(int));
+    infoPtr->varLists[0]->numVars = 2;
+    infoPtr->varLists[0]->varIndexes[0] = keyVar;
+    infoPtr->varLists[0]->varIndexes[1] = valVar;
+    infoIndex = TclCreateAuxData(infoPtr, &newForeachInfoType, envPtr);
+
+    /*
+     * Start issuing instructions to write to the array.
+     */
+
+    CompileWord(envPtr, dataTokenPtr, interp, 2);
+    if (!isDataLiteral || !isDataValid) {
+	/*
+	 * Only need this safety check if we're handling a non-literal or list
+	 * containing an invalid literal; with valid list literals, we've
+	 * already checked (worth it because literals are a very common
+	 * use-case with [array set]).
+	 */
+
+	TclEmitOpcode(	INST_DUP,				envPtr);
+	TclEmitOpcode(	INST_LIST_LENGTH,			envPtr);
+	PushStringLiteral(envPtr, "1");
+	TclEmitOpcode(	INST_BITAND,				envPtr);
+	offsetFwd = CurrentOffset(envPtr);
+	TclEmitInstInt1(INST_JUMP_FALSE1, 0,			envPtr);
+	PushStringLiteral(envPtr, "list must have an even number of elements");
+	PushStringLiteral(envPtr, "-errorcode {TCL ARGUMENT FORMAT}");
+	TclEmitInstInt4(INST_RETURN_IMM, TCL_ERROR,		envPtr);
+	TclEmitInt4(		0,				envPtr);
+	TclAdjustStackDepth(-1, envPtr);
+	fwd = CurrentOffset(envPtr) - offsetFwd;
+	TclStoreInt1AtPtr(fwd, envPtr->codeStart+offsetFwd+1);
+    }
+
+    TclEmitInstInt4(INST_ARRAY_EXISTS_IMM, localIndex,	envPtr);
+    TclEmitInstInt1(INST_JUMP_TRUE1, 7,			envPtr);
+    TclEmitInstInt4(INST_ARRAY_MAKE_IMM, localIndex,	envPtr);
+    TclEmitInstInt4(INST_FOREACH_START, infoIndex,	envPtr);
+    offsetBack = CurrentOffset(envPtr);
+    Emit14Inst(	INST_LOAD_SCALAR, keyVar,		envPtr);
+    Emit14Inst(	INST_LOAD_SCALAR, valVar,		envPtr);
+    Emit14Inst(	INST_STORE_ARRAY, localIndex,		envPtr);
+    TclEmitOpcode(	INST_POP,			envPtr);
+    infoPtr->loopCtTemp = offsetBack - CurrentOffset(envPtr); /*misuse */
+    TclEmitOpcode( INST_FOREACH_STEP,			envPtr);
+    TclEmitOpcode( INST_FOREACH_END,			envPtr);
+    TclAdjustStackDepth(-3, envPtr);
+    PushStringLiteral(envPtr,	"");
+
+    done:
+    Tcl_DecrRefCount(literalObj);
+    return code;
+}
+
+int
+TclCompileArrayUnsetCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    int isScalar, localIndex;
+
+    if (parsePtr->numWords != 2) {
+	return TclCompileBasic2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    PushVarNameWord(interp, tokenPtr, envPtr, TCL_NO_ELEMENT,
+	    &localIndex, &isScalar, 1);
+    if (!isScalar) {
+	return TCL_ERROR;
+    }
+
+    if (localIndex >= 0) {
+	TclEmitInstInt4(INST_ARRAY_EXISTS_IMM, localIndex,	envPtr);
+	TclEmitInstInt1(INST_JUMP_FALSE1, 8,			envPtr);
+	TclEmitInstInt1(INST_UNSET_SCALAR, 1,			envPtr);
+	TclEmitInt4(		localIndex,			envPtr);
+    } else {
+	TclEmitOpcode(	INST_DUP,				envPtr);
+	TclEmitOpcode(	INST_ARRAY_EXISTS_STK,			envPtr);
+	TclEmitInstInt1(INST_JUMP_FALSE1, 6,			envPtr);
+	TclEmitInstInt1(INST_UNSET_STK, 1,			envPtr);
+	TclEmitInstInt1(INST_JUMP1, 3,				envPtr);
+	/* Each branch decrements stack depth, but we only take one. */
+	TclAdjustStackDepth(1, envPtr);
+	TclEmitOpcode(	INST_POP,				envPtr);
+    }
+    PushStringLiteral(envPtr,	"");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileBreakCmd --
+ *
+ *	Procedure called to compile the "break" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "break" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileBreakCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    ExceptionRange *rangePtr;
+    ExceptionAux *auxPtr;
+
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the innermost exception range that contains this command.
+     */
+
+    rangePtr = TclGetInnermostExceptionRange(envPtr, TCL_BREAK, &auxPtr);
+    if (rangePtr && rangePtr->type == LOOP_EXCEPTION_RANGE) {
+	/*
+	 * Found the target! No need for a nasty INST_BREAK here.
+	 */
+
+	TclCleanupStackForBreakContinue(envPtr, auxPtr);
+	TclAddLoopBreakFixup(envPtr, auxPtr);
+    } else {
+	/*
+	 * Emit a real break.
+	 */
+
+	TclEmitOpcode(INST_BREAK, envPtr);
+    }
+    TclAdjustStackDepth(1, envPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileCatchCmd --
+ *
+ *	Procedure called to compile the "catch" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "catch" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileCatchCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    JumpFixup jumpFixup;
+    Tcl_Token *cmdTokenPtr, *resultNameTokenPtr, *optsNameTokenPtr;
+    int resultIndex, optsIndex, range, dropScript = 0;
+    DefineLineInformation;	/* TIP #280 */
+    int depth = TclGetStackDepth(envPtr);
+
+    /*
+     * If syntax does not match what we expect for [catch], do not compile.
+     * Let runtime checks determine if syntax has changed.
+     */
+
+    if ((parsePtr->numWords < 2) || (parsePtr->numWords > 4)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * If variables were specified and the catch command is at global level
+     * (not in a procedure), don't compile it inline: the payoff is too small.
+     */
+
+    if ((parsePtr->numWords >= 3) && !EnvHasLVT(envPtr)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure the variable names, if any, have no substitutions and just
+     * refer to local scalars.
+     */
+
+    resultIndex = optsIndex = -1;
+    cmdTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (parsePtr->numWords >= 3) {
+	resultNameTokenPtr = TokenAfter(cmdTokenPtr);
+	/* DGP */
+	resultIndex = LocalScalarFromToken(resultNameTokenPtr, envPtr);
+	if (resultIndex < 0) {
+	    return TCL_ERROR;
+	}
+
+	/* DKF */
+	if (parsePtr->numWords == 4) {
+	    optsNameTokenPtr = TokenAfter(resultNameTokenPtr);
+	    optsIndex = LocalScalarFromToken(optsNameTokenPtr, envPtr);
+	    if (optsIndex < 0) {
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    /*
+     * We will compile the catch command. Declare the exception range that it
+     * uses.
+     *
+     * If the body is a simple word, compile a BEGIN_CATCH instruction,
+     * followed by the instructions to eval the body.
+     * Otherwise, compile instructions to substitute the body text before
+     * starting the catch, then BEGIN_CATCH, and then EVAL_STK to evaluate the
+     * substituted body.
+     * Care has to be taken to make sure that substitution happens outside the
+     * catch range so that errors in the substitution are not caught.
+     * [Bug 219184]
+     * The reason for duplicating the script is that EVAL_STK would otherwise
+     * begin by undeflowing the stack below the mark set by BEGIN_CATCH4.
+     */
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    if (cmdTokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	TclEmitInstInt4(	INST_BEGIN_CATCH4, range,	envPtr);
+	ExceptionRangeStarts(envPtr, range);
+	BODY(cmdTokenPtr, 1);
+    } else {
+	SetLineInformation(1);
+	CompileTokens(envPtr, cmdTokenPtr, interp);
+	TclEmitInstInt4(	INST_BEGIN_CATCH4, range,	envPtr);
+	ExceptionRangeStarts(envPtr, range);
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitInvoke(envPtr,	INST_EVAL_STK);
+	/* drop the script */
+	dropScript = 1;
+	TclEmitInstInt4(	INST_REVERSE, 2,		envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+    ExceptionRangeEnds(envPtr, range);
+
+
+    /*
+     * Emit the "no errors" epilogue: push "0" (TCL_OK) as the catch result,
+     * and jump around the "error case" code.
+     */
+
+    TclCheckStackDepth(depth+1, envPtr);
+    PushStringLiteral(envPtr, "0");
+    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &jumpFixup);
+
+    /*
+     * Emit the "error case" epilogue. Push the interpreter result and the
+     * return code.
+     */
+
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    TclSetStackDepth(depth + dropScript, envPtr);
+
+    if (dropScript) {
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+
+
+    /* Stack at this point is empty */
+    TclEmitOpcode(		INST_PUSH_RESULT,		envPtr);
+    TclEmitOpcode(		INST_PUSH_RETURN_CODE,		envPtr);
+
+    /* Stack at this point on both branches: result returnCode */
+
+    if (TclFixupForwardJumpToHere(envPtr, &jumpFixup, 127)) {
+	Tcl_Panic("TclCompileCatchCmd: bad jump distance %d",
+		(int)(CurrentOffset(envPtr) - jumpFixup.codeOffset));
+    }
+
+    /*
+     * Push the return options if the caller wants them. This needs to happen
+     * before INST_END_CATCH
+     */
+
+    if (optsIndex != -1) {
+	TclEmitOpcode(		INST_PUSH_RETURN_OPTIONS,	envPtr);
+    }
+
+    /*
+     * End the catch
+     */
+
+    TclEmitOpcode(		INST_END_CATCH,			envPtr);
+
+    /*
+     * Save the result and return options if the caller wants them. This needs
+     * to happen after INST_END_CATCH (compile-3.6/7).
+     */
+
+    if (optsIndex != -1) {
+	Emit14Inst(		INST_STORE_SCALAR, optsIndex,	envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+
+    /*
+     * At this point, the top of the stack is inconveniently ordered:
+     *		result returnCode
+     * Reverse the stack to store the result.
+     */
+
+    TclEmitInstInt4(	INST_REVERSE, 2,		envPtr);
+    if (resultIndex != -1) {
+	Emit14Inst(	INST_STORE_SCALAR, resultIndex,	envPtr);
+    }
+    TclEmitOpcode(	INST_POP,			envPtr);
+
+    TclCheckStackDepth(depth+1, envPtr);
+    return TCL_OK;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * TclCompileClockClicksCmd --
+ *
+ *	Procedure called to compile the "tcl::clock::clicks" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to run time.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "clock clicks"
+ *	command at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileClockClicksCmd(
+    Tcl_Interp* interp,		/* Tcl interpreter */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token* tokenPtr;
+
+    switch (parsePtr->numWords) {
+    case 1:
+	/*
+	 * No args
+	 */
+	TclEmitInstInt1(INST_CLOCK_READ, 0, envPtr);
+	break;
+    case 2:
+	/*
+	 * -milliseconds or -microseconds
+	 */
+	tokenPtr = TokenAfter(parsePtr->tokenPtr);
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD
+	    || tokenPtr[1].size < 4
+	    || tokenPtr[1].size > 13) {
+	    return TCL_ERROR;
+	} else if (!strncmp(tokenPtr[1].start, "-microseconds",
+			    tokenPtr[1].size)) {
+	    TclEmitInstInt1(INST_CLOCK_READ, 1, envPtr);
+	    break;
+	} else if (!strncmp(tokenPtr[1].start, "-milliseconds",
+			    tokenPtr[1].size)) {
+	    TclEmitInstInt1(INST_CLOCK_READ, 2, envPtr);
+	    break;
+	} else {
+	    return TCL_ERROR;
+	}
+    default:
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+
+/*----------------------------------------------------------------------
+ *
+ * TclCompileClockReadingCmd --
+ *
+ *	Procedure called to compile the "tcl::clock::microseconds",
+ *	"tcl::clock::milliseconds" and "tcl::clock::seconds" commands.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to run time.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "clock clicks"
+ *	command at runtime.
+ *
+ * Client data is 1 for microseconds, 2 for milliseconds, 3 for seconds.
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileClockReadingCmd(
+    Tcl_Interp* interp,		/* Tcl interpreter */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    TclEmitInstInt1(INST_CLOCK_READ, PTR2INT(cmdPtr->objClientData), envPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileConcatCmd --
+ *
+ *	Procedure called to compile the "concat" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "concat" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileConcatCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Obj *objPtr, *listObj;
+    Tcl_Token *tokenPtr;
+    int i;
+
+    /* TODO: Consider compiling expansion case. */
+    if (parsePtr->numWords == 1) {
+	/*
+	 * [concat] without arguments just pushes an empty object.
+	 */
+
+	PushStringLiteral(envPtr, "");
+	return TCL_OK;
+    }
+
+    /*
+     * Test if all arguments are compile-time known. If they are, we can
+     * implement with a simple push.
+     */
+
+    listObj = Tcl_NewObj();
+    for (i = 1, tokenPtr = parsePtr->tokenPtr; i < parsePtr->numWords; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	objPtr = Tcl_NewObj();
+	if (!TclWordKnownAtCompileTime(tokenPtr, objPtr)) {
+	    Tcl_DecrRefCount(objPtr);
+	    Tcl_DecrRefCount(listObj);
+	    listObj = NULL;
+	    break;
+	}
+	(void) Tcl_ListObjAppendElement(NULL, listObj, objPtr);
+    }
+    if (listObj != NULL) {
+	Tcl_Obj **objs;
+	const char *bytes;
+	int len;
+
+	Tcl_ListObjGetElements(NULL, listObj, &len, &objs);
+	objPtr = Tcl_ConcatObj(len, objs);
+	Tcl_DecrRefCount(listObj);
+	bytes = TclGetStringFromObj(objPtr, &len);
+	PushLiteral(envPtr, bytes, len);
+	Tcl_DecrRefCount(objPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * General case: runtime concat.
+     */
+
+    for (i = 1, tokenPtr = parsePtr->tokenPtr; i < parsePtr->numWords; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i);
+    }
+
+    TclEmitInstInt4(	INST_CONCAT_STK, i-1,		envPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileContinueCmd --
+ *
+ *	Procedure called to compile the "continue" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "continue" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileContinueCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    ExceptionRange *rangePtr;
+    ExceptionAux *auxPtr;
+
+    /*
+     * There should be no argument after the "continue".
+     */
+
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * See if we can find a valid continueOffset (i.e., not -1) in the
+     * innermost containing exception range.
+     */
+
+    rangePtr = TclGetInnermostExceptionRange(envPtr, TCL_CONTINUE, &auxPtr);
+    if (rangePtr && rangePtr->type == LOOP_EXCEPTION_RANGE) {
+	/*
+	 * Found the target! No need for a nasty INST_CONTINUE here.
+	 */
+
+	TclCleanupStackForBreakContinue(envPtr, auxPtr);
+	TclAddLoopContinueFixup(envPtr, auxPtr);
+    } else {
+	/*
+	 * Emit a real continue.
+	 */
+
+	TclEmitOpcode(INST_CONTINUE, envPtr);
+    }
+    TclAdjustStackDepth(1, envPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileDict*Cmd --
+ *
+ *	Functions called to compile "dict" sucommands.
+ *
+ * Results:
+ *	All return TCL_OK for a successful compile, and TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "dict" subcommand at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileDictSetCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int i, dictVarIndex;
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *varTokenPtr;
+
+    /*
+     * There must be at least one argument after the command.
+     */
+
+    if (parsePtr->numWords < 4) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * The dictionary variable must be a local scalar that is knowable at
+     * compile time; anything else exceeds the complexity of the opcode. So
+     * discover what the index is.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictVarIndex = LocalScalarFromToken(varTokenPtr, envPtr);
+    if (dictVarIndex < 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Remaining words (key path and value to set) can be handled normally.
+     */
+
+    tokenPtr = TokenAfter(varTokenPtr);
+    for (i=2 ; i< parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    /*
+     * Now emit the instruction to do the dict manipulation.
+     */
+
+    TclEmitInstInt4( INST_DICT_SET, parsePtr->numWords-3,	envPtr);
+    TclEmitInt4(     dictVarIndex,			envPtr);
+    TclAdjustStackDepth(-1, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictIncrCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *varTokenPtr, *keyTokenPtr;
+    int dictVarIndex, incrAmount;
+
+    /*
+     * There must be at least two arguments after the command.
+     */
+
+    if (parsePtr->numWords < 3 || parsePtr->numWords > 4) {
+	return TCL_ERROR;
+    }
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    keyTokenPtr = TokenAfter(varTokenPtr);
+
+    /*
+     * Parse the increment amount, if present.
+     */
+
+    if (parsePtr->numWords == 4) {
+	const char *word;
+	int numBytes, code;
+	Tcl_Token *incrTokenPtr;
+	Tcl_Obj *intObj;
+
+	incrTokenPtr = TokenAfter(keyTokenPtr);
+	if (incrTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    return TclCompileBasic2Or3ArgCmd(interp, parsePtr,cmdPtr, envPtr);
+	}
+	word = incrTokenPtr[1].start;
+	numBytes = incrTokenPtr[1].size;
+
+	intObj = Tcl_NewStringObj(word, numBytes);
+	Tcl_IncrRefCount(intObj);
+	code = TclGetIntFromObj(NULL, intObj, &incrAmount);
+	TclDecrRefCount(intObj);
+	if (code != TCL_OK) {
+	    return TclCompileBasic2Or3ArgCmd(interp, parsePtr,cmdPtr, envPtr);
+	}
+    } else {
+	incrAmount = 1;
+    }
+
+    /*
+     * The dictionary variable must be a local scalar that is knowable at
+     * compile time; anything else exceeds the complexity of the opcode. So
+     * discover what the index is.
+     */
+
+    dictVarIndex = LocalScalarFromToken(varTokenPtr, envPtr);
+    if (dictVarIndex < 0) {
+	return TclCompileBasic2Or3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Emit the key and the code to actually do the increment.
+     */
+
+    CompileWord(envPtr, keyTokenPtr, interp, 2);
+    TclEmitInstInt4( INST_DICT_INCR_IMM, incrAmount,	envPtr);
+    TclEmitInt4(     dictVarIndex,			envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictGetCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * There must be at least two arguments after the command (the single-arg
+     * case is legal, but too special and magic for us to deal with here).
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Only compile this because we need INST_DICT_GET anyway.
+     */
+
+    for (i=1 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    TclEmitInstInt4(INST_DICT_GET, parsePtr->numWords-2, envPtr);
+    TclAdjustStackDepth(-1, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictExistsCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * There must be at least two arguments after the command (the single-arg
+     * case is legal, but too special and magic for us to deal with here).
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Now we do the code generation.
+     */
+
+    for (i=1 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    TclEmitInstInt4(INST_DICT_EXISTS, parsePtr->numWords-2, envPtr);
+    TclAdjustStackDepth(-1, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictUnsetCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int i, dictVarIndex;
+
+    /*
+     * There must be at least one argument after the variable name for us to
+     * compile to bytecode.
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * The dictionary variable must be a local scalar that is knowable at
+     * compile time; anything else exceeds the complexity of the opcode. So
+     * discover what the index is.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictVarIndex = LocalScalarFromToken(tokenPtr, envPtr);
+    if (dictVarIndex < 0) {
+	return TclCompileBasicMin2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Remaining words (the key path) can be handled normally.
+     */
+
+    for (i=2 ; i<parsePtr->numWords ; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i);
+    }
+
+    /*
+     * Now emit the instruction to do the dict manipulation.
+     */
+
+    TclEmitInstInt4( INST_DICT_UNSET, parsePtr->numWords-2,	envPtr);
+    TclEmitInt4(	dictVarIndex,				envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictCreateCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    int worker;			/* Temp var for building the value in. */
+    Tcl_Token *tokenPtr;
+    Tcl_Obj *keyObj, *valueObj, *dictObj;
+    const char *bytes;
+    int i, len;
+
+    if ((parsePtr->numWords & 1) == 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * See if we can build the value at compile time...
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictObj = Tcl_NewObj();
+    Tcl_IncrRefCount(dictObj);
+    for (i=1 ; i<parsePtr->numWords ; i+=2) {
+	keyObj = Tcl_NewObj();
+	Tcl_IncrRefCount(keyObj);
+	if (!TclWordKnownAtCompileTime(tokenPtr, keyObj)) {
+	    Tcl_DecrRefCount(keyObj);
+	    Tcl_DecrRefCount(dictObj);
+	    goto nonConstant;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+	valueObj = Tcl_NewObj();
+	Tcl_IncrRefCount(valueObj);
+	if (!TclWordKnownAtCompileTime(tokenPtr, valueObj)) {
+	    Tcl_DecrRefCount(keyObj);
+	    Tcl_DecrRefCount(valueObj);
+	    Tcl_DecrRefCount(dictObj);
+	    goto nonConstant;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+	Tcl_DictObjPut(NULL, dictObj, keyObj, valueObj);
+	Tcl_DecrRefCount(keyObj);
+	Tcl_DecrRefCount(valueObj);
+    }
+
+    /*
+     * We did! Excellent. The "verifyDict" is to do type forcing.
+     */
+
+    bytes = TclGetStringFromObj(dictObj, &len);
+    PushLiteral(envPtr, bytes, len);
+    TclEmitOpcode(		INST_DUP,			envPtr);
+    TclEmitOpcode(		INST_DICT_VERIFY,		envPtr);
+    Tcl_DecrRefCount(dictObj);
+    return TCL_OK;
+
+    /*
+     * Otherwise, we've got to issue runtime code to do the building, which we
+     * do by [dict set]ting into an unnamed local variable. This requires that
+     * we are in a context with an LVT.
+     */
+
+  nonConstant:
+    worker = AnonymousLocal(envPtr);
+    if (worker < 0) {
+	return TclCompileBasicMin0ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    PushStringLiteral(envPtr,		"");
+    Emit14Inst(			INST_STORE_SCALAR, worker,	envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    for (i=1 ; i<parsePtr->numWords ; i+=2) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i+1);
+	tokenPtr = TokenAfter(tokenPtr);
+	TclEmitInstInt4(	INST_DICT_SET, 1,		envPtr);
+	TclEmitInt4(			worker,			envPtr);
+	TclAdjustStackDepth(-1, envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+    Emit14Inst(			INST_LOAD_SCALAR, worker,	envPtr);
+    TclEmitInstInt1(		INST_UNSET_SCALAR, 0,		envPtr);
+    TclEmitInt4(			worker,			envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictMergeCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    int i, workerIndex, infoIndex, outLoop;
+
+    /*
+     * Deal with some special edge cases. Note that in the case with one
+     * argument, the only thing to do is to verify the dict-ness.
+     */
+
+    /* TODO: Consider support for compiling expanded args. (less likely) */
+    if (parsePtr->numWords < 2) {
+	PushStringLiteral(envPtr, "");
+	return TCL_OK;
+    } else if (parsePtr->numWords == 2) {
+	tokenPtr = TokenAfter(parsePtr->tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 1);
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitOpcode(		INST_DICT_VERIFY,		envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * There's real merging work to do.
+     *
+     * Allocate some working space. This means we'll only ever compile this
+     * command when there's an LVT present.
+     */
+
+    workerIndex = AnonymousLocal(envPtr);
+    if (workerIndex < 0) {
+	return TclCompileBasicMin2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+    infoIndex = AnonymousLocal(envPtr);
+
+    /*
+     * Get the first dictionary and verify that it is so.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    TclEmitOpcode(		INST_DUP,			envPtr);
+    TclEmitOpcode(		INST_DICT_VERIFY,		envPtr);
+    Emit14Inst(			INST_STORE_SCALAR, workerIndex,	envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+
+    /*
+     * For each of the remaining dictionaries...
+     */
+
+    outLoop = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    TclEmitInstInt4(		INST_BEGIN_CATCH4, outLoop,	envPtr);
+    ExceptionRangeStarts(envPtr, outLoop);
+    for (i=2 ; i<parsePtr->numWords ; i++) {
+	/*
+	 * Get the dictionary, and merge its pairs into the first dict (using
+	 * a small loop).
+	 */
+
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i);
+	TclEmitInstInt4(	INST_DICT_FIRST, infoIndex,	envPtr);
+	TclEmitInstInt1(	INST_JUMP_TRUE1, 24,		envPtr);
+	TclEmitInstInt4(	INST_REVERSE, 2,		envPtr);
+	TclEmitInstInt4(	INST_DICT_SET, 1,		envPtr);
+	TclEmitInt4(			workerIndex,		envPtr);
+	TclAdjustStackDepth(-1, envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+	TclEmitInstInt4(	INST_DICT_NEXT, infoIndex,	envPtr);
+	TclEmitInstInt1(	INST_JUMP_FALSE1, -20,		envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+	TclEmitInstInt1(	INST_UNSET_SCALAR, 0,		envPtr);
+	TclEmitInt4(			infoIndex,		envPtr);
+    }
+    ExceptionRangeEnds(envPtr, outLoop);
+    TclEmitOpcode(		INST_END_CATCH,			envPtr);
+
+    /*
+     * Clean up any state left over.
+     */
+
+    Emit14Inst(			INST_LOAD_SCALAR, workerIndex,	envPtr);
+    TclEmitInstInt1(		INST_UNSET_SCALAR, 0,		envPtr);
+    TclEmitInt4(			workerIndex,		envPtr);
+    TclEmitInstInt1(		INST_JUMP1, 18,			envPtr);
+
+    /*
+     * If an exception happens when starting to iterate over the second (and
+     * subsequent) dicts. This is strictly not necessary, but it is nice.
+     */
+
+    TclAdjustStackDepth(-1, envPtr);
+    ExceptionRangeTarget(envPtr, outLoop, catchOffset);
+    TclEmitOpcode(		INST_PUSH_RETURN_OPTIONS,	envPtr);
+    TclEmitOpcode(		INST_PUSH_RESULT,		envPtr);
+    TclEmitOpcode(		INST_END_CATCH,			envPtr);
+    TclEmitInstInt1(		INST_UNSET_SCALAR, 0,		envPtr);
+    TclEmitInt4(			workerIndex,		envPtr);
+    TclEmitInstInt1(		INST_UNSET_SCALAR, 0,		envPtr);
+    TclEmitInt4(			infoIndex,		envPtr);
+    TclEmitOpcode(		INST_RETURN_STK,		envPtr);
+
+    return TCL_OK;
+}
+
+int
+TclCompileDictForCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    return CompileDictEachCmd(interp, parsePtr, cmdPtr, envPtr,
+	    TCL_EACH_KEEP_NONE);
+}
+
+int
+TclCompileDictMapCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    return CompileDictEachCmd(interp, parsePtr, cmdPtr, envPtr,
+	    TCL_EACH_COLLECT);
+}
+
+int
+CompileDictEachCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int collect)		/* Flag == TCL_EACH_COLLECT to collect and
+				 * construct a new dictionary with the loop
+				 * body result. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *varsTokenPtr, *dictTokenPtr, *bodyTokenPtr;
+    int keyVarIndex, valueVarIndex, nameChars, loopRange, catchRange;
+    int infoIndex, jumpDisplacement, bodyTargetOffset, emptyTargetOffset;
+    int numVars, endTargetOffset;
+    int collectVar = -1;	/* Index of temp var holding the result
+				 * dict. */
+    const char **argv;
+    Tcl_DString buffer;
+
+    /*
+     * There must be three arguments after the command.
+     */
+
+    if (parsePtr->numWords != 4) {
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    varsTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictTokenPtr = TokenAfter(varsTokenPtr);
+    bodyTokenPtr = TokenAfter(dictTokenPtr);
+    if (varsTokenPtr->type != TCL_TOKEN_SIMPLE_WORD ||
+	    bodyTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Create temporary variable to capture return values from loop body when
+     * we're collecting results.
+     */
+
+    if (collect == TCL_EACH_COLLECT) {
+	collectVar = AnonymousLocal(envPtr);
+	if (collectVar < 0) {
+	    return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+	}
+    }
+
+    /*
+     * Check we've got a pair of variables and that they are local variables.
+     * Then extract their indices in the LVT.
+     */
+
+    Tcl_DStringInit(&buffer);
+    TclDStringAppendToken(&buffer, &varsTokenPtr[1]);
+    if (Tcl_SplitList(NULL, Tcl_DStringValue(&buffer), &numVars,
+	    &argv) != TCL_OK) {
+	Tcl_DStringFree(&buffer);
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+    Tcl_DStringFree(&buffer);
+    if (numVars != 2) {
+	ckfree(argv);
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    nameChars = strlen(argv[0]);
+    keyVarIndex = LocalScalar(argv[0], nameChars, envPtr);
+    nameChars = strlen(argv[1]);
+    valueVarIndex = LocalScalar(argv[1], nameChars, envPtr);
+    ckfree(argv);
+
+    if ((keyVarIndex < 0) || (valueVarIndex < 0)) {
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Allocate a temporary variable to store the iterator reference. The
+     * variable will contain a Tcl_DictSearch reference which will be
+     * allocated by INST_DICT_FIRST and disposed when the variable is unset
+     * (at which point it should also have been finished with).
+     */
+
+    infoIndex = AnonymousLocal(envPtr);
+    if (infoIndex < 0) {
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Preparation complete; issue instructions. Note that this code issues
+     * fixed-sized jumps. That simplifies things a lot!
+     *
+     * First up, initialize the accumulator dictionary if needed.
+     */
+
+    if (collect == TCL_EACH_COLLECT) {
+	PushStringLiteral(envPtr, "");
+	Emit14Inst(	INST_STORE_SCALAR, collectVar,		envPtr);
+	TclEmitOpcode(	INST_POP,				envPtr);
+    }
+
+    /*
+     * Get the dictionary and start the iteration. No catching of errors at
+     * this point.
+     */
+
+    CompileWord(envPtr, dictTokenPtr, interp, 2);
+
+    /*
+     * Now we catch errors from here on so that we can finalize the search
+     * started by Tcl_DictObjFirst above.
+     */
+
+    catchRange = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    TclEmitInstInt4(	INST_BEGIN_CATCH4, catchRange,		envPtr);
+    ExceptionRangeStarts(envPtr, catchRange);
+
+    TclEmitInstInt4(	INST_DICT_FIRST, infoIndex,		envPtr);
+    emptyTargetOffset = CurrentOffset(envPtr);
+    TclEmitInstInt4(	INST_JUMP_TRUE4, 0,			envPtr);
+
+    /*
+     * Inside the iteration, write the loop variables.
+     */
+
+    bodyTargetOffset = CurrentOffset(envPtr);
+    Emit14Inst(		INST_STORE_SCALAR, keyVarIndex,		envPtr);
+    TclEmitOpcode(	INST_POP,				envPtr);
+    Emit14Inst(		INST_STORE_SCALAR, valueVarIndex,	envPtr);
+    TclEmitOpcode(	INST_POP,				envPtr);
+
+    /*
+     * Set up the loop exception targets.
+     */
+
+    loopRange = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+    ExceptionRangeStarts(envPtr, loopRange);
+
+    /*
+     * Compile the loop body itself. It should be stack-neutral.
+     */
+
+    BODY(bodyTokenPtr, 3);
+    if (collect == TCL_EACH_COLLECT) {
+	Emit14Inst(	INST_LOAD_SCALAR, keyVarIndex,		envPtr);
+	TclEmitInstInt4(INST_OVER, 1,				envPtr);
+	TclEmitInstInt4(INST_DICT_SET, 1,			envPtr);
+	TclEmitInt4(		collectVar,			envPtr);
+	TclAdjustStackDepth(-1, envPtr);
+	TclEmitOpcode(	INST_POP,				envPtr);
+    }
+    TclEmitOpcode(	INST_POP,				envPtr);
+
+    /*
+     * Both exception target ranges (error and loop) end here.
+     */
+
+    ExceptionRangeEnds(envPtr, loopRange);
+    ExceptionRangeEnds(envPtr, catchRange);
+
+    /*
+     * Continue (or just normally process) by getting the next pair of items
+     * from the dictionary and jumping back to the code to write them into
+     * variables if there is another pair.
+     */
+
+    ExceptionRangeTarget(envPtr, loopRange, continueOffset);
+    TclEmitInstInt4(	INST_DICT_NEXT, infoIndex,		envPtr);
+    jumpDisplacement = bodyTargetOffset - CurrentOffset(envPtr);
+    TclEmitInstInt4(	INST_JUMP_FALSE4, jumpDisplacement,	envPtr);
+    endTargetOffset = CurrentOffset(envPtr);
+    TclEmitInstInt1(	INST_JUMP1, 0,				envPtr);
+
+    /*
+     * Error handler "finally" clause, which force-terminates the iteration
+     * and rethrows the error.
+     */
+
+    TclAdjustStackDepth(-1, envPtr);
+    ExceptionRangeTarget(envPtr, catchRange, catchOffset);
+    TclEmitOpcode(	INST_PUSH_RETURN_OPTIONS,		envPtr);
+    TclEmitOpcode(	INST_PUSH_RESULT,			envPtr);
+    TclEmitOpcode(	INST_END_CATCH,				envPtr);
+    TclEmitInstInt1(	INST_UNSET_SCALAR, 0,			envPtr);
+    TclEmitInt4(		infoIndex,			envPtr);
+    if (collect == TCL_EACH_COLLECT) {
+	TclEmitInstInt1(INST_UNSET_SCALAR, 0,			envPtr);
+	TclEmitInt4(		collectVar,			envPtr);
+    }
+    TclEmitOpcode(	INST_RETURN_STK,			envPtr);
+
+    /*
+     * Otherwise we're done (the jump after the DICT_FIRST points here) and we
+     * need to pop the bogus key/value pair (pushed to keep stack calculations
+     * easy!) Note that we skip the END_CATCH. [Bug 1382528]
+     */
+
+    jumpDisplacement = CurrentOffset(envPtr) - emptyTargetOffset;
+    TclUpdateInstInt4AtPc(INST_JUMP_TRUE4, jumpDisplacement,
+	    envPtr->codeStart + emptyTargetOffset);
+    jumpDisplacement = CurrentOffset(envPtr) - endTargetOffset;
+    TclUpdateInstInt1AtPc(INST_JUMP1, jumpDisplacement,
+	    envPtr->codeStart + endTargetOffset);
+    TclEmitOpcode(	INST_POP,				envPtr);
+    TclEmitOpcode(	INST_POP,				envPtr);
+    ExceptionRangeTarget(envPtr, loopRange, breakOffset);
+    TclFinalizeLoopExceptionRange(envPtr, loopRange);
+    TclEmitOpcode(	INST_END_CATCH,				envPtr);
+
+    /*
+     * Final stage of the command (normal case) is that we push an empty
+     * object (or push the accumulator as the result object). This is done
+     * last to promote peephole optimization when it's dropped immediately.
+     */
+
+    TclEmitInstInt1(	INST_UNSET_SCALAR, 0,			envPtr);
+    TclEmitInt4(		infoIndex,			envPtr);
+    if (collect == TCL_EACH_COLLECT) {
+	Emit14Inst(	INST_LOAD_SCALAR, collectVar,		envPtr);
+	TclEmitInstInt1(INST_UNSET_SCALAR, 0,			envPtr);
+	TclEmitInt4(		collectVar,			envPtr);
+    } else {
+	PushStringLiteral(envPtr, "");
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileDictUpdateCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    int i, dictIndex, numVars, range, infoIndex;
+    Tcl_Token **keyTokenPtrs, *dictVarTokenPtr, *bodyTokenPtr, *tokenPtr;
+    DictUpdateInfo *duiPtr;
+    JumpFixup jumpFixup;
+
+    /*
+     * There must be at least one argument after the command.
+     */
+
+    if (parsePtr->numWords < 5) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the command. Expect the following:
+     *   dict update <lit(eral)> <any> <lit> ?<any> <lit> ...? <lit>
+     */
+
+    if ((parsePtr->numWords - 1) & 1) {
+	return TCL_ERROR;
+    }
+    numVars = (parsePtr->numWords - 3) / 2;
+
+    /*
+     * The dictionary variable must be a local scalar that is knowable at
+     * compile time; anything else exceeds the complexity of the opcode. So
+     * discover what the index is.
+     */
+
+    dictVarTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictIndex = LocalScalarFromToken(dictVarTokenPtr, envPtr);
+    if (dictIndex < 0) {
+	goto issueFallback;
+    }
+
+    /*
+     * Assemble the instruction metadata. This is complex enough that it is
+     * represented as auxData; it holds an ordered list of variable indices
+     * that are to be used.
+     */
+
+    duiPtr = ckalloc(sizeof(DictUpdateInfo) + sizeof(int) * (numVars - 1));
+    duiPtr->length = numVars;
+    keyTokenPtrs = TclStackAlloc(interp, sizeof(Tcl_Token *) * numVars);
+    tokenPtr = TokenAfter(dictVarTokenPtr);
+
+    for (i=0 ; i<numVars ; i++) {
+	/*
+	 * Put keys to one side for later compilation to bytecode.
+	 */
+
+	keyTokenPtrs[i] = tokenPtr;
+	tokenPtr = TokenAfter(tokenPtr);
+
+	/*
+	 * Stash the index in the auxiliary data (if it is indeed a local
+	 * scalar that is resolvable at compile-time).
+	 */
+
+	duiPtr->varIndices[i] = LocalScalarFromToken(tokenPtr, envPtr);
+	if (duiPtr->varIndices[i] < 0) {
+	    goto failedUpdateInfoAssembly;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	goto failedUpdateInfoAssembly;
+    }
+    bodyTokenPtr = tokenPtr;
+
+    /*
+     * The list of variables to bind is stored in auxiliary data so that it
+     * can't be snagged by literal sharing and forced to shimmer dangerously.
+     */
+
+    infoIndex = TclCreateAuxData(duiPtr, &dictUpdateInfoType, envPtr);
+
+    for (i=0 ; i<numVars ; i++) {
+	CompileWord(envPtr, keyTokenPtrs[i], interp, 2*i+2);
+    }
+    TclEmitInstInt4(	INST_LIST, numVars,			envPtr);
+    TclEmitInstInt4(	INST_DICT_UPDATE_START, dictIndex,	envPtr);
+    TclEmitInt4(		infoIndex,			envPtr);
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    TclEmitInstInt4(	INST_BEGIN_CATCH4, range,		envPtr);
+
+    ExceptionRangeStarts(envPtr, range);
+    BODY(bodyTokenPtr, parsePtr->numWords - 1);
+    ExceptionRangeEnds(envPtr, range);
+
+    /*
+     * Normal termination code: the stack has the key list below the result of
+     * the body evaluation: swap them and finish the update code.
+     */
+
+    TclEmitOpcode(	INST_END_CATCH,				envPtr);
+    TclEmitInstInt4(	INST_REVERSE, 2,			envPtr);
+    TclEmitInstInt4(	INST_DICT_UPDATE_END, dictIndex,	envPtr);
+    TclEmitInt4(		infoIndex,			envPtr);
+
+    /*
+     * Jump around the exceptional termination code.
+     */
+
+    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &jumpFixup);
+
+    /*
+     * Termination code for non-ok returns: stash the result and return
+     * options in the stack, bring up the key list, finish the update code,
+     * and finally return with the catched return data
+     */
+
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    TclEmitOpcode(	INST_PUSH_RESULT,			envPtr);
+    TclEmitOpcode(	INST_PUSH_RETURN_OPTIONS,		envPtr);
+    TclEmitOpcode(	INST_END_CATCH,				envPtr);
+    TclEmitInstInt4(	INST_REVERSE, 3,			envPtr);
+
+    TclEmitInstInt4(	INST_DICT_UPDATE_END, dictIndex,	envPtr);
+    TclEmitInt4(		infoIndex,			envPtr);
+    TclEmitInvoke(envPtr,INST_RETURN_STK);
+
+    if (TclFixupForwardJumpToHere(envPtr, &jumpFixup, 127)) {
+	Tcl_Panic("TclCompileDictCmd(update): bad jump distance %d",
+		(int) (CurrentOffset(envPtr) - jumpFixup.codeOffset));
+    }
+    TclStackFree(interp, keyTokenPtrs);
+    return TCL_OK;
+
+    /*
+     * Clean up after a failure to create the DictUpdateInfo structure.
+     */
+
+  failedUpdateInfoAssembly:
+    ckfree(duiPtr);
+    TclStackFree(interp, keyTokenPtrs);
+  issueFallback:
+    return TclCompileBasicMin2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileDictAppendCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    int i, dictVarIndex;
+
+    /*
+     * There must be at least two argument after the command. And we impose an
+     * (arbirary) safe limit; anyone exceeding it should stop worrying about
+     * speed quite so much. ;-)
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords<4 || parsePtr->numWords>100) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the index of the local variable that we will be working with.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    dictVarIndex = LocalScalarFromToken(tokenPtr, envPtr);
+    if (dictVarIndex < 0) {
+	return TclCompileBasicMin2ArgCmd(interp, parsePtr,cmdPtr, envPtr);
+    }
+
+    /*
+     * Produce the string to concatenate onto the dictionary entry.
+     */
+
+    tokenPtr = TokenAfter(tokenPtr);
+    for (i=2 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    if (parsePtr->numWords > 4) {
+	TclEmitInstInt1(INST_STR_CONCAT1, parsePtr->numWords-3, envPtr);
+    }
+
+    /*
+     * Do the concatenation.
+     */
+
+    TclEmitInstInt4(INST_DICT_APPEND, dictVarIndex, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictLappendCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *varTokenPtr, *keyTokenPtr, *valueTokenPtr;
+    int dictVarIndex;
+
+    /*
+     * There must be three arguments after the command.
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    /* Probably not.  Why is INST_DICT_LAPPEND limited to one value? */
+    if (parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the arguments.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    keyTokenPtr = TokenAfter(varTokenPtr);
+    valueTokenPtr = TokenAfter(keyTokenPtr);
+    dictVarIndex = LocalScalarFromToken(varTokenPtr, envPtr);
+    if (dictVarIndex < 0) {
+	return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Issue the implementation.
+     */
+
+    CompileWord(envPtr, keyTokenPtr, interp, 2);
+    CompileWord(envPtr, valueTokenPtr, interp, 3);
+    TclEmitInstInt4(	INST_DICT_LAPPEND, dictVarIndex,	envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileDictWithCmd(
+    Tcl_Interp *interp,		/* Used for looking up stuff. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    int i, range, varNameTmp = -1, pathTmp = -1, keysTmp, gotPath;
+    int dictVar, bodyIsEmpty = 1;
+    Tcl_Token *varTokenPtr, *tokenPtr;
+    JumpFixup jumpFixup;
+    const char *ptr, *end;
+
+    /*
+     * There must be at least one argument after the command.
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the command (trivially). Expect the following:
+     *   dict with <any (varName)> ?<any> ...? <literal>
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    tokenPtr = TokenAfter(varTokenPtr);
+    for (i=3 ; i<parsePtr->numWords ; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	return TclCompileBasicMin2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Test if the last word is an empty script; if so, we can compile it in
+     * all cases, but if it is non-empty we need local variable table entries
+     * to hold the temporary variables (used to keep stack usage simple).
+     */
+
+    for (ptr=tokenPtr[1].start,end=ptr+tokenPtr[1].size ; ptr!=end ; ptr++) {
+	if (*ptr!=' ' && *ptr!='\t' && *ptr!='\n' && *ptr!='\r') {
+	    if (envPtr->procPtr == NULL) {
+		return TclCompileBasicMin2ArgCmd(interp, parsePtr, cmdPtr,
+			envPtr);
+	    }
+	    bodyIsEmpty = 0;
+	    break;
+	}
+    }
+
+    /*
+     * Determine if we're manipulating a dict in a simple local variable.
+     */
+
+    gotPath = (parsePtr->numWords > 3);
+    dictVar = LocalScalarFromToken(varTokenPtr, envPtr);
+
+    /*
+     * Special case: an empty body means we definitely have no need to issue
+     * try-finally style code or to allocate local variable table entries for
+     * storing temporaries. Still need to do both INST_DICT_EXPAND and
+     * INST_DICT_RECOMBINE_* though, because we can't determine if we're free
+     * of traces.
+     */
+
+    if (bodyIsEmpty) {
+	if (dictVar >= 0) {
+	    if (gotPath) {
+		/*
+		 * Case: Path into dict in LVT with empty body.
+		 */
+
+		tokenPtr = TokenAfter(varTokenPtr);
+		for (i=2 ; i<parsePtr->numWords-1 ; i++) {
+		    CompileWord(envPtr, tokenPtr, interp, i);
+		    tokenPtr = TokenAfter(tokenPtr);
+		}
+		TclEmitInstInt4(INST_LIST, parsePtr->numWords-3,envPtr);
+		Emit14Inst(	INST_LOAD_SCALAR, dictVar,	envPtr);
+		TclEmitInstInt4(INST_OVER, 1,			envPtr);
+		TclEmitOpcode(	INST_DICT_EXPAND,		envPtr);
+		TclEmitInstInt4(INST_DICT_RECOMBINE_IMM, dictVar, envPtr);
+	    } else {
+		/*
+		 * Case: Direct dict in LVT with empty body.
+		 */
+
+		PushStringLiteral(envPtr, "");
+		Emit14Inst(	INST_LOAD_SCALAR, dictVar,	envPtr);
+		PushStringLiteral(envPtr, "");
+		TclEmitOpcode(	INST_DICT_EXPAND,		envPtr);
+		TclEmitInstInt4(INST_DICT_RECOMBINE_IMM, dictVar, envPtr);
+	    }
+	} else {
+	    if (gotPath) {
+		/*
+		 * Case: Path into dict in non-simple var with empty body.
+		 */
+
+		tokenPtr = varTokenPtr;
+		for (i=1 ; i<parsePtr->numWords-1 ; i++) {
+		    CompileWord(envPtr, tokenPtr, interp, i);
+		    tokenPtr = TokenAfter(tokenPtr);
+		}
+		TclEmitInstInt4(INST_LIST, parsePtr->numWords-3,envPtr);
+		TclEmitInstInt4(INST_OVER, 1,			envPtr);
+		TclEmitOpcode(	INST_LOAD_STK,			envPtr);
+		TclEmitInstInt4(INST_OVER, 1,			envPtr);
+		TclEmitOpcode(	INST_DICT_EXPAND,		envPtr);
+		TclEmitOpcode(	INST_DICT_RECOMBINE_STK,	envPtr);
+	    } else {
+		/*
+		 * Case: Direct dict in non-simple var with empty body.
+		 */
+
+		CompileWord(envPtr, varTokenPtr, interp, 1);
+		TclEmitOpcode(	INST_DUP,			envPtr);
+		TclEmitOpcode(	INST_LOAD_STK,			envPtr);
+		PushStringLiteral(envPtr, "");
+		TclEmitOpcode(	INST_DICT_EXPAND,		envPtr);
+		PushStringLiteral(envPtr, "");
+		TclEmitInstInt4(INST_REVERSE, 2,		envPtr);
+		TclEmitOpcode(	INST_DICT_RECOMBINE_STK,	envPtr);
+	    }
+	}
+	PushStringLiteral(envPtr, "");
+	return TCL_OK;
+    }
+
+    /*
+     * OK, we have a non-trivial body. This means that the focus is on
+     * generating a try-finally structure where the INST_DICT_RECOMBINE_* goes
+     * in the 'finally' clause.
+     *
+     * Start by allocating local (unnamed, untraced) working variables.
+     */
+
+    if (dictVar == -1) {
+	varNameTmp = AnonymousLocal(envPtr);
+    }
+    if (gotPath) {
+	pathTmp = AnonymousLocal(envPtr);
+    }
+    keysTmp = AnonymousLocal(envPtr);
+
+    /*
+     * Issue instructions. First, the part to expand the dictionary.
+     */
+
+    if (dictVar == -1) {
+	CompileWord(envPtr, varTokenPtr, interp, 1);
+	Emit14Inst(		INST_STORE_SCALAR, varNameTmp,	envPtr);
+    }
+    tokenPtr = TokenAfter(varTokenPtr);
+    if (gotPath) {
+	for (i=2 ; i<parsePtr->numWords-1 ; i++) {
+	    CompileWord(envPtr, tokenPtr, interp, i);
+	    tokenPtr = TokenAfter(tokenPtr);
+	}
+	TclEmitInstInt4(	INST_LIST, parsePtr->numWords-3,envPtr);
+	Emit14Inst(		INST_STORE_SCALAR, pathTmp,	envPtr);
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+    if (dictVar == -1) {
+	TclEmitOpcode(		INST_LOAD_STK,			envPtr);
+    } else {
+	Emit14Inst(		INST_LOAD_SCALAR, dictVar,	envPtr);
+    }
+    if (gotPath) {
+	Emit14Inst(		INST_LOAD_SCALAR, pathTmp,	envPtr);
+    } else {
+	PushStringLiteral(envPtr, "");
+    }
+    TclEmitOpcode(		INST_DICT_EXPAND,		envPtr);
+    Emit14Inst(			INST_STORE_SCALAR, keysTmp,	envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+
+    /*
+     * Now the body of the [dict with].
+     */
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    TclEmitInstInt4(		INST_BEGIN_CATCH4, range,	envPtr);
+
+    ExceptionRangeStarts(envPtr, range);
+    BODY(tokenPtr, parsePtr->numWords - 1);
+    ExceptionRangeEnds(envPtr, range);
+
+    /*
+     * Now fold the results back into the dictionary in the OK case.
+     */
+
+    TclEmitOpcode(		INST_END_CATCH,			envPtr);
+    if (dictVar == -1) {
+	Emit14Inst(		INST_LOAD_SCALAR, varNameTmp,	envPtr);
+    }
+    if (gotPath) {
+	Emit14Inst(		INST_LOAD_SCALAR, pathTmp,	envPtr);
+    } else {
+	PushStringLiteral(envPtr, "");
+    }
+    Emit14Inst(			INST_LOAD_SCALAR, keysTmp,	envPtr);
+    if (dictVar == -1) {
+	TclEmitOpcode(		INST_DICT_RECOMBINE_STK,	envPtr);
+    } else {
+	TclEmitInstInt4(	INST_DICT_RECOMBINE_IMM, dictVar, envPtr);
+    }
+    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &jumpFixup);
+
+    /*
+     * Now fold the results back into the dictionary in the exception case.
+     */
+
+    TclAdjustStackDepth(-1, envPtr);
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    TclEmitOpcode(		INST_PUSH_RETURN_OPTIONS,	envPtr);
+    TclEmitOpcode(		INST_PUSH_RESULT,		envPtr);
+    TclEmitOpcode(		INST_END_CATCH,			envPtr);
+    if (dictVar == -1) {
+	Emit14Inst(		INST_LOAD_SCALAR, varNameTmp,	envPtr);
+    }
+    if (parsePtr->numWords > 3) {
+	Emit14Inst(		INST_LOAD_SCALAR, pathTmp,	envPtr);
+    } else {
+	PushStringLiteral(envPtr, "");
+    }
+    Emit14Inst(			INST_LOAD_SCALAR, keysTmp,	envPtr);
+    if (dictVar == -1) {
+	TclEmitOpcode(		INST_DICT_RECOMBINE_STK,	envPtr);
+    } else {
+	TclEmitInstInt4(	INST_DICT_RECOMBINE_IMM, dictVar, envPtr);
+    }
+    TclEmitInvoke(envPtr,	INST_RETURN_STK);
+
+    /*
+     * Prepare for the start of the next command.
+     */
+
+    if (TclFixupForwardJumpToHere(envPtr, &jumpFixup, 127)) {
+	Tcl_Panic("TclCompileDictCmd(update): bad jump distance %d",
+		(int) (CurrentOffset(envPtr) - jumpFixup.codeOffset));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupDictUpdateInfo, FreeDictUpdateInfo --
+ *
+ *	Functions to duplicate, release and print the aux data created for use
+ *	with the INST_DICT_UPDATE_START and INST_DICT_UPDATE_END instructions.
+ *
+ * Results:
+ *	DupDictUpdateInfo: a copy of the auxiliary data
+ *	FreeDictUpdateInfo: none
+ *	PrintDictUpdateInfo: none
+ *	DisassembleDictUpdateInfo: none
+ *
+ * Side effects:
+ *	DupDictUpdateInfo: allocates memory
+ *	FreeDictUpdateInfo: releases memory
+ *	PrintDictUpdateInfo: none
+ *	DisassembleDictUpdateInfo: none
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ClientData
+DupDictUpdateInfo(
+    ClientData clientData)
+{
+    DictUpdateInfo *dui1Ptr, *dui2Ptr;
+    unsigned len;
+
+    dui1Ptr = clientData;
+    len = sizeof(DictUpdateInfo) + sizeof(int) * (dui1Ptr->length - 1);
+    dui2Ptr = ckalloc(len);
+    memcpy(dui2Ptr, dui1Ptr, len);
+    return dui2Ptr;
+}
+
+static void
+FreeDictUpdateInfo(
+    ClientData clientData)
+{
+    ckfree(clientData);
+}
+
+static void
+PrintDictUpdateInfo(
+    ClientData clientData,
+    Tcl_Obj *appendObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    DictUpdateInfo *duiPtr = clientData;
+    int i;
+
+    for (i=0 ; i<duiPtr->length ; i++) {
+	if (i) {
+	    Tcl_AppendToObj(appendObj, ", ", -1);
+	}
+	Tcl_AppendPrintfToObj(appendObj, "%%v%u", duiPtr->varIndices[i]);
+    }
+}
+
+static void
+DisassembleDictUpdateInfo(
+    ClientData clientData,
+    Tcl_Obj *dictObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    DictUpdateInfo *duiPtr = clientData;
+    int i;
+    Tcl_Obj *variables = Tcl_NewObj();
+
+    for (i=0 ; i<duiPtr->length ; i++) {
+	Tcl_ListObjAppendElement(NULL, variables,
+		Tcl_NewIntObj(duiPtr->varIndices[i]));
+    }
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("variables", -1),
+	    variables);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileErrorCmd --
+ *
+ *	Procedure called to compile the "error" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "error" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileErrorCmd(
+    Tcl_Interp *interp,		/* Used for context. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * General syntax: [error message ?errorInfo? ?errorCode?]
+     */
+
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords < 2 || parsePtr->numWords > 4) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Handle the message.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+
+    /*
+     * Construct the options. Note that -code and -level are not here.
+     */
+
+    if (parsePtr->numWords == 2) {
+	PushStringLiteral(envPtr, "");
+    } else {
+	PushStringLiteral(envPtr, "-errorinfo");
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 2);
+	if (parsePtr->numWords == 3) {
+	    TclEmitInstInt4(	INST_LIST, 2,			envPtr);
+	} else {
+	    PushStringLiteral(envPtr, "-errorcode");
+	    tokenPtr = TokenAfter(tokenPtr);
+	    CompileWord(envPtr, tokenPtr, interp, 3);
+	    TclEmitInstInt4(	INST_LIST, 4,			envPtr);
+	}
+    }
+
+    /*
+     * Issue the error via 'returnImm error 0'.
+     */
+
+    TclEmitInstInt4(		INST_RETURN_IMM, TCL_ERROR,	envPtr);
+    TclEmitInt4(			0,			envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileExprCmd --
+ *
+ *	Procedure called to compile the "expr" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "expr" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileExprCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *firstWordPtr;
+
+    if (parsePtr->numWords == 1) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * TIP #280: Use the per-word line information of the current command.
+     */
+
+    envPtr->line = envPtr->extCmdMapPtr->loc[
+	    envPtr->extCmdMapPtr->nuloc-1].line[1];
+
+    firstWordPtr = TokenAfter(parsePtr->tokenPtr);
+    TclCompileExprWords(interp, firstWordPtr, parsePtr->numWords-1, envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileForCmd --
+ *
+ *	Procedure called to compile the "for" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "for" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileForCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *startTokenPtr, *testTokenPtr, *nextTokenPtr, *bodyTokenPtr;
+    JumpFixup jumpEvalCondFixup;
+    int bodyCodeOffset, nextCodeOffset, jumpDist;
+    int bodyRange, nextRange;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 5) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the test expression requires substitutions, don't compile the for
+     * command inline. E.g., the expression might cause the loop to never
+     * execute or execute forever, as in "for {} "$x > 5" {incr x} {}".
+     */
+
+    startTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    testTokenPtr = TokenAfter(startTokenPtr);
+    if (testTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Bail out also if the body or the next expression require substitutions
+     * in order to insure correct behaviour [Bug 219166]
+     */
+
+    nextTokenPtr = TokenAfter(testTokenPtr);
+    bodyTokenPtr = TokenAfter(nextTokenPtr);
+    if ((nextTokenPtr->type != TCL_TOKEN_SIMPLE_WORD)
+	    || (bodyTokenPtr->type != TCL_TOKEN_SIMPLE_WORD)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Inline compile the initial command.
+     */
+
+    BODY(startTokenPtr, 1);
+    TclEmitOpcode(INST_POP, envPtr);
+
+    /*
+     * Jump to the evaluation of the condition. This code uses the "loop
+     * rotation" optimisation (which eliminates one branch from the loop).
+     * "for start cond next body" produces then:
+     *       start
+     *       goto A
+     *    B: body                : bodyCodeOffset
+     *       next                : nextCodeOffset, continueOffset
+     *    A: cond -> result      : testCodeOffset
+     *       if (result) goto B
+     */
+
+    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &jumpEvalCondFixup);
+
+    /*
+     * Compile the loop body.
+     */
+
+    bodyRange = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+    bodyCodeOffset = ExceptionRangeStarts(envPtr, bodyRange);
+    BODY(bodyTokenPtr, 4);
+    ExceptionRangeEnds(envPtr, bodyRange);
+    TclEmitOpcode(INST_POP, envPtr);
+
+    /*
+     * Compile the "next" subcommand. Note that this exception range will not
+     * have a continueOffset (other than -1) connected to it; it won't trap
+     * TCL_CONTINUE but rather just TCL_BREAK.
+     */
+
+    nextRange = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+    envPtr->exceptAuxArrayPtr[nextRange].supportsContinue = 0;
+    nextCodeOffset = ExceptionRangeStarts(envPtr, nextRange);
+    BODY(nextTokenPtr, 3);
+    ExceptionRangeEnds(envPtr, nextRange);
+    TclEmitOpcode(INST_POP, envPtr);
+
+    /*
+     * Compile the test expression then emit the conditional jump that
+     * terminates the for.
+     */
+
+    if (TclFixupForwardJumpToHere(envPtr, &jumpEvalCondFixup, 127)) {
+	bodyCodeOffset += 3;
+	nextCodeOffset += 3;
+    }
+
+    SetLineInformation(2);
+    TclCompileExprWords(interp, testTokenPtr, 1, envPtr);
+
+    jumpDist = CurrentOffset(envPtr) - bodyCodeOffset;
+    if (jumpDist > 127) {
+	TclEmitInstInt4(INST_JUMP_TRUE4, -jumpDist, envPtr);
+    } else {
+	TclEmitInstInt1(INST_JUMP_TRUE1, -jumpDist, envPtr);
+    }
+
+    /*
+     * Fix the starting points of the exception ranges (may have moved due to
+     * jump type modification) and set where the exceptions target.
+     */
+
+    envPtr->exceptArrayPtr[bodyRange].codeOffset = bodyCodeOffset;
+    envPtr->exceptArrayPtr[bodyRange].continueOffset = nextCodeOffset;
+
+    envPtr->exceptArrayPtr[nextRange].codeOffset = nextCodeOffset;
+
+    ExceptionRangeTarget(envPtr, bodyRange, breakOffset);
+    ExceptionRangeTarget(envPtr, nextRange, breakOffset);
+    TclFinalizeLoopExceptionRange(envPtr, bodyRange);
+    TclFinalizeLoopExceptionRange(envPtr, nextRange);
+
+    /*
+     * The for command's result is an empty string.
+     */
+
+    PushStringLiteral(envPtr, "");
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileForeachCmd --
+ *
+ *	Procedure called to compile the "foreach" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "foreach" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileForeachCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    return CompileEachloopCmd(interp, parsePtr, cmdPtr, envPtr,
+	    TCL_EACH_KEEP_NONE);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLmapCmd --
+ *
+ *	Procedure called to compile the "lmap" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "lmap" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLmapCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    return CompileEachloopCmd(interp, parsePtr, cmdPtr, envPtr,
+	    TCL_EACH_COLLECT);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileEachloopCmd --
+ *
+ *	Procedure called to compile the "foreach" and "lmap" commands.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "foreach" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompileEachloopCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int collect)		/* Select collecting or accumulating mode
+				 * (TCL_EACH_*) */
+{
+    Proc *procPtr = envPtr->procPtr;
+    ForeachInfo *infoPtr=NULL;	/* Points to the structure describing this
+				 * foreach command. Stored in a AuxData
+				 * record in the ByteCode. */
+
+    Tcl_Token *tokenPtr, *bodyTokenPtr;
+    int jumpBackOffset, infoIndex, range;
+    int numWords, numLists, i, j, code = TCL_OK;
+    Tcl_Obj *varListObj = NULL;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * If the foreach command isn't in a procedure, don't compile it inline:
+     * the payoff is too small.
+     */
+
+    if (procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    numWords = parsePtr->numWords;
+    if ((numWords < 4) || (numWords%2 != 0)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Bail out if the body requires substitutions in order to insure correct
+     * behaviour. [Bug 219166]
+     */
+
+    for (i = 0, tokenPtr = parsePtr->tokenPtr; i < numWords-1; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    bodyTokenPtr = tokenPtr;
+    if (bodyTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create and initialize the ForeachInfo and ForeachVarList data
+     * structures describing this command. Then create a AuxData record
+     * pointing to the ForeachInfo structure.
+     */
+
+    numLists = (numWords - 2)/2;
+    infoPtr = ckalloc(sizeof(ForeachInfo)
+	    + (numLists - 1) * sizeof(ForeachVarList *));
+    infoPtr->numLists = 0;	/* Count this up as we go */
+
+    /*
+     * Parse each var list into sequence of var names.  Don't
+     * compile the foreach inline if any var name needs substitutions or isn't
+     * a scalar, or if any var list needs substitutions.
+     */
+
+    varListObj = Tcl_NewObj();
+    for (i = 0, tokenPtr = parsePtr->tokenPtr;
+	    i < numWords-1;
+	    i++, tokenPtr = TokenAfter(tokenPtr)) {
+	ForeachVarList *varListPtr;
+	int numVars;
+
+	if (i%2 != 1) {
+	    continue;
+	}
+
+	/*
+	 * If the variable list is empty, we can enter an infinite loop when
+	 * the interpreted version would not.  Take care to ensure this does
+	 * not happen.  [Bug 1671138]
+	 */
+
+	if (!TclWordKnownAtCompileTime(tokenPtr, varListObj) ||
+		TCL_OK != Tcl_ListObjLength(NULL, varListObj, &numVars) ||
+		numVars == 0) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+
+	varListPtr = ckalloc(sizeof(ForeachVarList)
+		+ (numVars - 1) * sizeof(int));
+	varListPtr->numVars = numVars;
+	infoPtr->varLists[i/2] = varListPtr;
+	infoPtr->numLists++;
+
+	for (j = 0;  j < numVars;  j++) {
+	    Tcl_Obj *varNameObj;
+	    const char *bytes;
+	    int numBytes, varIndex;
+
+	    Tcl_ListObjIndex(NULL, varListObj, j, &varNameObj);
+	    bytes = TclGetStringFromObj(varNameObj, &numBytes);
+	    varIndex = LocalScalar(bytes, numBytes, envPtr);
+	    if (varIndex < 0) {
+		code = TCL_ERROR;
+		goto done;
+	    }
+	    varListPtr->varIndexes[j] = varIndex;
+	}
+	Tcl_SetObjLength(varListObj, 0);
+    }
+
+    /*
+     * We will compile the foreach command.
+     */
+
+    infoIndex = TclCreateAuxData(infoPtr, &newForeachInfoType, envPtr);
+
+    /*
+     * Create the collecting object, unshared.
+     */
+
+    if (collect == TCL_EACH_COLLECT) {
+	TclEmitInstInt4(INST_LIST, 0, envPtr);
+    }
+
+    /*
+     * Evaluate each value list and leave it on stack.
+     */
+
+    for (i = 0, tokenPtr = parsePtr->tokenPtr;
+	    i < numWords-1;
+	    i++, tokenPtr = TokenAfter(tokenPtr)) {
+	if ((i%2 == 0) && (i > 0)) {
+	    CompileWord(envPtr, tokenPtr, interp, i);
+	}
+    }
+
+    TclEmitInstInt4(INST_FOREACH_START, infoIndex, envPtr);
+
+    /*
+     * Inline compile the loop body.
+     */
+
+    range = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+
+    ExceptionRangeStarts(envPtr, range);
+    BODY(bodyTokenPtr, numWords - 1);
+    ExceptionRangeEnds(envPtr, range);
+
+    if (collect == TCL_EACH_COLLECT) {
+	TclEmitOpcode(INST_LMAP_COLLECT, envPtr);
+    } else {
+	TclEmitOpcode(		INST_POP,			envPtr);
+    }
+
+    /*
+     * Bottom of loop code: assign each loop variable and check whether
+     * to terminate the loop. Set the loop's break target.
+     */
+
+    ExceptionRangeTarget(envPtr, range, continueOffset);
+    TclEmitOpcode(INST_FOREACH_STEP, envPtr);
+    ExceptionRangeTarget(envPtr, range, breakOffset);
+    TclFinalizeLoopExceptionRange(envPtr, range);
+    TclEmitOpcode(INST_FOREACH_END, envPtr);
+    TclAdjustStackDepth(-(numLists+2), envPtr);
+
+    /*
+     * Set the jumpback distance from INST_FOREACH_STEP to the start of the
+     * body's code. Misuse loopCtTemp for storing the jump size.
+     */
+
+    jumpBackOffset = envPtr->exceptArrayPtr[range].continueOffset -
+	    envPtr->exceptArrayPtr[range].codeOffset;
+    infoPtr->loopCtTemp = -jumpBackOffset;
+
+    /*
+     * The command's result is an empty string if not collecting. If
+     * collecting, it is automatically left on stack after FOREACH_END.
+     */
+
+    if (collect != TCL_EACH_COLLECT) {
+	PushStringLiteral(envPtr, "");
+    }
+
+    done:
+    if (code == TCL_ERROR) {
+	FreeForeachInfo(infoPtr);
+    }
+    Tcl_DecrRefCount(varListObj);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupForeachInfo --
+ *
+ *	This procedure duplicates a ForeachInfo structure created as auxiliary
+ *	data during the compilation of a foreach command.
+ *
+ * Results:
+ *	A pointer to a newly allocated copy of the existing ForeachInfo
+ *	structure is returned.
+ *
+ * Side effects:
+ *	Storage for the copied ForeachInfo record is allocated. If the
+ *	original ForeachInfo structure pointed to any ForeachVarList records,
+ *	these structures are also copied and pointers to them are stored in
+ *	the new ForeachInfo record.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ClientData
+DupForeachInfo(
+    ClientData clientData)	/* The foreach command's compilation auxiliary
+				 * data to duplicate. */
+{
+    register ForeachInfo *srcPtr = clientData;
+    ForeachInfo *dupPtr;
+    register ForeachVarList *srcListPtr, *dupListPtr;
+    int numVars, i, j, numLists = srcPtr->numLists;
+
+    dupPtr = ckalloc(sizeof(ForeachInfo)
+	    + numLists * sizeof(ForeachVarList *));
+    dupPtr->numLists = numLists;
+    dupPtr->firstValueTemp = srcPtr->firstValueTemp;
+    dupPtr->loopCtTemp = srcPtr->loopCtTemp;
+
+    for (i = 0;  i < numLists;  i++) {
+	srcListPtr = srcPtr->varLists[i];
+	numVars = srcListPtr->numVars;
+	dupListPtr = ckalloc(sizeof(ForeachVarList)
+		+ numVars * sizeof(int));
+	dupListPtr->numVars = numVars;
+	for (j = 0;  j < numVars;  j++) {
+	    dupListPtr->varIndexes[j] =	srcListPtr->varIndexes[j];
+	}
+	dupPtr->varLists[i] = dupListPtr;
+    }
+    return dupPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeForeachInfo --
+ *
+ *	Procedure to free a ForeachInfo structure created as auxiliary data
+ *	during the compilation of a foreach command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Storage for the ForeachInfo structure pointed to by the ClientData
+ *	argument is freed as is any ForeachVarList record pointed to by the
+ *	ForeachInfo structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeForeachInfo(
+    ClientData clientData)	/* The foreach command's compilation auxiliary
+				 * data to free. */
+{
+    register ForeachInfo *infoPtr = clientData;
+    register ForeachVarList *listPtr;
+    int numLists = infoPtr->numLists;
+    register int i;
+
+    for (i = 0;  i < numLists;  i++) {
+	listPtr = infoPtr->varLists[i];
+	ckfree(listPtr);
+    }
+    ckfree(infoPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PrintForeachInfo, DisassembleForeachInfo --
+ *
+ *	Functions to write a human-readable or script-readablerepresentation
+ *	of a ForeachInfo structure to a Tcl_Obj for debugging.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PrintForeachInfo(
+    ClientData clientData,
+    Tcl_Obj *appendObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register ForeachInfo *infoPtr = clientData;
+    register ForeachVarList *varsPtr;
+    int i, j;
+
+    Tcl_AppendToObj(appendObj, "data=[", -1);
+
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	if (i) {
+	    Tcl_AppendToObj(appendObj, ", ", -1);
+	}
+	Tcl_AppendPrintfToObj(appendObj, "%%v%u",
+		(unsigned) (infoPtr->firstValueTemp + i));
+    }
+    Tcl_AppendPrintfToObj(appendObj, "], loop=%%v%u",
+	    (unsigned) infoPtr->loopCtTemp);
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	if (i) {
+	    Tcl_AppendToObj(appendObj, ",", -1);
+	}
+	Tcl_AppendPrintfToObj(appendObj, "\n\t\t it%%v%u\t[",
+		(unsigned) (infoPtr->firstValueTemp + i));
+	varsPtr = infoPtr->varLists[i];
+	for (j=0 ; j<varsPtr->numVars ; j++) {
+	    if (j) {
+		Tcl_AppendToObj(appendObj, ", ", -1);
+	    }
+	    Tcl_AppendPrintfToObj(appendObj, "%%v%u",
+		    (unsigned) varsPtr->varIndexes[j]);
+	}
+	Tcl_AppendToObj(appendObj, "]", -1);
+    }
+}
+
+static void
+PrintNewForeachInfo(
+    ClientData clientData,
+    Tcl_Obj *appendObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register ForeachInfo *infoPtr = clientData;
+    register ForeachVarList *varsPtr;
+    int i, j;
+
+    Tcl_AppendPrintfToObj(appendObj, "jumpOffset=%+d, vars=",
+	    infoPtr->loopCtTemp);
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	if (i) {
+	    Tcl_AppendToObj(appendObj, ",", -1);
+	}
+	Tcl_AppendToObj(appendObj, "[", -1);
+	varsPtr = infoPtr->varLists[i];
+	for (j=0 ; j<varsPtr->numVars ; j++) {
+	    if (j) {
+		Tcl_AppendToObj(appendObj, ",", -1);
+	    }
+	    Tcl_AppendPrintfToObj(appendObj, "%%v%u",
+		    (unsigned) varsPtr->varIndexes[j]);
+	}
+	Tcl_AppendToObj(appendObj, "]", -1);
+    }
+}
+
+static void
+DisassembleForeachInfo(
+    ClientData clientData,
+    Tcl_Obj *dictObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register ForeachInfo *infoPtr = clientData;
+    register ForeachVarList *varsPtr;
+    int i, j;
+    Tcl_Obj *objPtr, *innerPtr;
+
+    /*
+     * Data stores.
+     */
+
+    objPtr = Tcl_NewObj();
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewIntObj(infoPtr->firstValueTemp + i));
+    }
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("data", -1), objPtr);
+
+    /*
+     * Loop counter.
+     */
+
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("loop", -1),
+	   Tcl_NewIntObj(infoPtr->loopCtTemp));
+
+    /*
+     * Assignment targets.
+     */
+
+    objPtr = Tcl_NewObj();
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	innerPtr = Tcl_NewObj();
+	varsPtr = infoPtr->varLists[i];
+	for (j=0 ; j<varsPtr->numVars ; j++) {
+	    Tcl_ListObjAppendElement(NULL, innerPtr,
+		    Tcl_NewIntObj(varsPtr->varIndexes[j]));
+	}
+	Tcl_ListObjAppendElement(NULL, objPtr, innerPtr);
+    }
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("assign", -1), objPtr);
+}
+
+static void
+DisassembleNewForeachInfo(
+    ClientData clientData,
+    Tcl_Obj *dictObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register ForeachInfo *infoPtr = clientData;
+    register ForeachVarList *varsPtr;
+    int i, j;
+    Tcl_Obj *objPtr, *innerPtr;
+
+    /*
+     * Jump offset.
+     */
+
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("jumpOffset", -1),
+	   Tcl_NewIntObj(infoPtr->loopCtTemp));
+
+    /*
+     * Assignment targets.
+     */
+
+    objPtr = Tcl_NewObj();
+    for (i=0 ; i<infoPtr->numLists ; i++) {
+	innerPtr = Tcl_NewObj();
+	varsPtr = infoPtr->varLists[i];
+	for (j=0 ; j<varsPtr->numVars ; j++) {
+	    Tcl_ListObjAppendElement(NULL, innerPtr,
+		    Tcl_NewIntObj(varsPtr->varIndexes[j]));
+	}
+	Tcl_ListObjAppendElement(NULL, objPtr, innerPtr);
+    }
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("assign", -1), objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileFormatCmd --
+ *
+ *	Procedure called to compile the "format" command. Handles cases that
+ *	can be done as constants or simple string concatenation only.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "format" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileFormatCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    Tcl_Obj **objv, *formatObj, *tmpObj;
+    char *bytes, *start;
+    int i, j, len;
+
+    /*
+     * Don't handle any guaranteed-error cases.
+     */
+
+    if (parsePtr->numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check if the argument words are all compile-time-known literals; that's
+     * a case we can handle by compiling to a constant.
+     */
+
+    formatObj = Tcl_NewObj();
+    Tcl_IncrRefCount(formatObj);
+    tokenPtr = TokenAfter(tokenPtr);
+    if (!TclWordKnownAtCompileTime(tokenPtr, formatObj)) {
+	Tcl_DecrRefCount(formatObj);
+	return TCL_ERROR;
+    }
+
+    objv = ckalloc((parsePtr->numWords-2) * sizeof(Tcl_Obj *));
+    for (i=0 ; i+2 < parsePtr->numWords ; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	objv[i] = Tcl_NewObj();
+	Tcl_IncrRefCount(objv[i]);
+	if (!TclWordKnownAtCompileTime(tokenPtr, objv[i])) {
+	    goto checkForStringConcatCase;
+	}
+    }
+
+    /*
+     * Everything is a literal, so the result is constant too (or an error if
+     * the format is broken). Do the format now.
+     */
+
+    tmpObj = Tcl_Format(interp, Tcl_GetString(formatObj),
+	    parsePtr->numWords-2, objv);
+    for (; --i>=0 ;) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    ckfree(objv);
+    Tcl_DecrRefCount(formatObj);
+    if (tmpObj == NULL) {
+	TclCompileSyntaxError(interp, envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Not an error, always a constant result, so just push the result as a
+     * literal. Job done.
+     */
+
+    bytes = TclGetStringFromObj(tmpObj, &len);
+    PushLiteral(envPtr, bytes, len);
+    Tcl_DecrRefCount(tmpObj);
+    return TCL_OK;
+
+  checkForStringConcatCase:
+    /*
+     * See if we can generate a sequence of things to concatenate. This
+     * requires that all the % sequences be %s or %%, as everything else is
+     * sufficiently complex that we don't bother.
+     *
+     * First, get the state of the system relatively sensible (cleaning up
+     * after our attempt to spot a literal).
+     */
+
+    for (; i>=0 ; i--) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    ckfree(objv);
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    tokenPtr = TokenAfter(tokenPtr);
+    i = 0;
+
+    /*
+     * Now scan through and check for non-%s and non-%% substitutions.
+     */
+
+    for (bytes = Tcl_GetString(formatObj) ; *bytes ; bytes++) {
+	if (*bytes == '%') {
+	    bytes++;
+	    if (*bytes == 's') {
+		i++;
+		continue;
+	    } else if (*bytes == '%') {
+		continue;
+	    }
+	    Tcl_DecrRefCount(formatObj);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Check if the number of things to concatenate will fit in a byte.
+     */
+
+    if (i+2 != parsePtr->numWords || i > 125) {
+	Tcl_DecrRefCount(formatObj);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Generate the pushes of the things to concatenate, a sequence of
+     * literals and compiled tokens (of which at least one is non-literal or
+     * we'd have the case in the first half of this function) which we will
+     * concatenate.
+     */
+
+    i = 0;			/* The count of things to concat. */
+    j = 2;			/* The index into the argument tokens, for
+				 * TIP#280 handling. */
+    start = Tcl_GetString(formatObj);
+				/* The start of the currently-scanned literal
+				 * in the format string. */
+    tmpObj = Tcl_NewObj();	/* The buffer used to accumulate the literal
+				 * being built. */
+    for (bytes = start ; *bytes ; bytes++) {
+	if (*bytes == '%') {
+	    Tcl_AppendToObj(tmpObj, start, bytes - start);
+	    if (*++bytes == '%') {
+		Tcl_AppendToObj(tmpObj, "%", 1);
+	    } else {
+		char *b = TclGetStringFromObj(tmpObj, &len);
+
+		/*
+		 * If there is a non-empty literal from the format string,
+		 * push it and reset.
+		 */
+
+		if (len > 0) {
+		    PushLiteral(envPtr, b, len);
+		    Tcl_DecrRefCount(tmpObj);
+		    tmpObj = Tcl_NewObj();
+		    i++;
+		}
+
+		/*
+		 * Push the code to produce the string that would be
+		 * substituted with %s, except we'll be concatenating
+		 * directly.
+		 */
+
+		CompileWord(envPtr, tokenPtr, interp, j);
+		tokenPtr = TokenAfter(tokenPtr);
+		j++;
+		i++;
+	    }
+	    start = bytes + 1;
+	}
+    }
+
+    /*
+     * Handle the case of a trailing literal.
+     */
+
+    Tcl_AppendToObj(tmpObj, start, bytes - start);
+    bytes = TclGetStringFromObj(tmpObj, &len);
+    if (len > 0) {
+	PushLiteral(envPtr, bytes, len);
+	i++;
+    }
+    Tcl_DecrRefCount(tmpObj);
+    Tcl_DecrRefCount(formatObj);
+
+    if (i > 1) {
+	/*
+	 * Do the concatenation, which produces the result.
+	 */
+
+	TclEmitInstInt1(INST_STR_CONCAT1, i, envPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLocalScalarFromToken --
+ *
+ *	Get the index into the table of compiled locals that corresponds
+ *	to a local scalar variable name.
+ *
+ * Results:
+ * 	Returns the non-negative integer index value into the table of
+ * 	compiled locals corresponding to a local scalar variable name.
+ * 	If the arguments passed in do not identify a local scalar variable
+ * 	then return -1.
+ *
+ * Side effects:
+ *	May add an entery into the table of compiled locals.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclLocalScalarFromToken(
+    Tcl_Token *tokenPtr,
+    CompileEnv *envPtr)
+{
+    int isScalar, index;
+
+    TclPushVarName(NULL, tokenPtr, envPtr, TCL_NO_ELEMENT, &index, &isScalar);
+    if (!isScalar) {
+	index = -1;
+    }
+    return index;
+}
+
+int
+TclLocalScalar(
+    const char *bytes,
+    int numBytes,
+    CompileEnv *envPtr)
+{
+    Tcl_Token token[2] =        {{TCL_TOKEN_SIMPLE_WORD, NULL, 0, 1},
+                                 {TCL_TOKEN_TEXT, NULL, 0, 0}};
+
+    token[1].start = bytes;
+    token[1].size = numBytes;
+    return TclLocalScalarFromToken(token, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPushVarName --
+ *
+ *	Procedure used in the compiling where pushing a variable name is
+ *	necessary (append, lappend, set).
+ *
+ * Results:
+ *	The values written to *localIndexPtr and *isScalarPtr signal to
+ *	the caller what the instructions emitted by this routine will do:
+ *
+ *	*isScalarPtr	(*localIndexPtr < 0)
+ *	1		1	Push the varname on the stack. (Stack +1)
+ *	1		0	*localIndexPtr is the index of the compiled
+ *				local for this varname.  No instructions
+ *				emitted.	(Stack +0)
+ *	0		1	Push part1 and part2 names of array element
+ *				on the stack.	(Stack +2)
+ *	0		0	*localIndexPtr is the index of the compiled
+ *				local for this array.  Element name is pushed
+ *				on the stack.	(Stack +1)
+ *
+ * Side effects:
+ *	Instructions are added to envPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclPushVarName(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Token *varTokenPtr,	/* Points to a variable token. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int flags,			/* TCL_NO_LARGE_INDEX | TCL_NO_ELEMENT. */
+    int *localIndexPtr,		/* Must not be NULL. */
+    int *isScalarPtr)		/* Must not be NULL. */
+{
+    register const char *p;
+    const char *name, *elName;
+    register int i, n;
+    Tcl_Token *elemTokenPtr = NULL;
+    int nameChars, elNameChars, simpleVarName, localIndex;
+    int elemTokenCount = 0, allocedTokens = 0, removedParen = 0;
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we need
+     * to emit code to compute and push the name at runtime. We use a frame
+     * slot (entry in the array of local vars) if we are compiling a procedure
+     * body and if the name is simple text that does not include namespace
+     * qualifiers.
+     */
+
+    simpleVarName = 0;
+    name = elName = NULL;
+    nameChars = elNameChars = 0;
+    localIndex = -1;
+
+    if (varTokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	/*
+	 * A simple variable name. Divide it up into "name" and "elName"
+	 * strings. If it is not a local variable, look it up at runtime.
+	 */
+
+	simpleVarName = 1;
+
+	name = varTokenPtr[1].start;
+	nameChars = varTokenPtr[1].size;
+	if (name[nameChars-1] == ')') {
+	    /*
+	     * last char is ')' => potential array reference.
+	     */
+
+	    for (i=0,p=name ; i<nameChars ; i++,p++) {
+		if (*p == '(') {
+		    elName = p + 1;
+		    elNameChars = nameChars - i - 2;
+		    nameChars = i;
+		    break;
+		}
+	    }
+
+	    if (!(flags & TCL_NO_ELEMENT) && (elName != NULL) && elNameChars) {
+		/*
+		 * An array element, the element name is a simple string:
+		 * assemble the corresponding token.
+		 */
+
+		elemTokenPtr = TclStackAlloc(interp, sizeof(Tcl_Token));
+		allocedTokens = 1;
+		elemTokenPtr->type = TCL_TOKEN_TEXT;
+		elemTokenPtr->start = elName;
+		elemTokenPtr->size = elNameChars;
+		elemTokenPtr->numComponents = 0;
+		elemTokenCount = 1;
+	    }
+	}
+    } else if (interp && ((n = varTokenPtr->numComponents) > 1)
+	    && (varTokenPtr[1].type == TCL_TOKEN_TEXT)
+	    && (varTokenPtr[n].type == TCL_TOKEN_TEXT)
+	    && (varTokenPtr[n].start[varTokenPtr[n].size - 1] == ')')) {
+	/*
+	 * Check for parentheses inside first token.
+	 */
+
+	simpleVarName = 0;
+	for (i = 0, p = varTokenPtr[1].start;
+		i < varTokenPtr[1].size; i++, p++) {
+	    if (*p == '(') {
+		simpleVarName = 1;
+		break;
+	    }
+	}
+	if (simpleVarName) {
+	    int remainingChars;
+
+	    /*
+	     * Check the last token: if it is just ')', do not count it.
+	     * Otherwise, remove the ')' and flag so that it is restored at
+	     * the end.
+	     */
+
+	    if (varTokenPtr[n].size == 1) {
+		n--;
+	    } else {
+		varTokenPtr[n].size--;
+		removedParen = n;
+	    }
+
+	    name = varTokenPtr[1].start;
+	    nameChars = p - varTokenPtr[1].start;
+	    elName = p + 1;
+	    remainingChars = (varTokenPtr[2].start - p) - 1;
+	    elNameChars = (varTokenPtr[n].start-p) + varTokenPtr[n].size - 1;
+
+	    if (!(flags & TCL_NO_ELEMENT)) {
+	      if (remainingChars) {
+		/*
+		 * Make a first token with the extra characters in the first
+		 * token.
+		 */
+
+		elemTokenPtr = TclStackAlloc(interp, n * sizeof(Tcl_Token));
+		allocedTokens = 1;
+		elemTokenPtr->type = TCL_TOKEN_TEXT;
+		elemTokenPtr->start = elName;
+		elemTokenPtr->size = remainingChars;
+		elemTokenPtr->numComponents = 0;
+		elemTokenCount = n;
+
+		/*
+		 * Copy the remaining tokens.
+		 */
+
+		memcpy(elemTokenPtr+1, varTokenPtr+2,
+			(n-1) * sizeof(Tcl_Token));
+	      } else {
+		/*
+		 * Use the already available tokens.
+		 */
+
+		elemTokenPtr = &varTokenPtr[2];
+		elemTokenCount = n - 1;
+	      }
+	    }
+	}
+    }
+
+    if (simpleVarName) {
+	/*
+	 * See whether name has any namespace separators (::'s).
+	 */
+
+	int hasNsQualifiers = 0;
+
+	for (i = 0, p = name;  i < nameChars;  i++, p++) {
+	    if ((*p == ':') && ((i+1) < nameChars) && (*(p+1) == ':')) {
+		hasNsQualifiers = 1;
+		break;
+	    }
+	}
+
+	/*
+	 * Look up the var name's index in the array of local vars in the proc
+	 * frame. If retrieving the var's value and it doesn't already exist,
+	 * push its name and look it up at runtime.
+	 */
+
+	if (!hasNsQualifiers) {
+	    localIndex = TclFindCompiledLocal(name, nameChars, 1, envPtr);
+	    if ((flags & TCL_NO_LARGE_INDEX) && (localIndex > 255)) {
+		/*
+		 * We'll push the name.
+		 */
+
+		localIndex = -1;
+	    }
+	}
+	if (interp && localIndex < 0) {
+	    PushLiteral(envPtr, name, nameChars);
+	}
+
+	/*
+	 * Compile the element script, if any, and only if not inhibited. [Bug
+	 * 3600328]
+	 */
+
+	if (elName != NULL && !(flags & TCL_NO_ELEMENT)) {
+	    if (elNameChars) {
+		TclCompileTokens(interp, elemTokenPtr, elemTokenCount,
+			envPtr);
+	    } else {
+		PushStringLiteral(envPtr, "");
+	    }
+	}
+    } else if (interp) {
+	/*
+	 * The var name isn't simple: compile and push it.
+	 */
+
+	CompileTokens(envPtr, varTokenPtr, interp);
+    }
+
+    if (removedParen) {
+	varTokenPtr[removedParen].size++;
+    }
+    if (allocedTokens) {
+	TclStackFree(interp, elemTokenPtr);
+    }
+    *localIndexPtr = localIndex;
+    *isScalarPtr = (elName == NULL);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCompCmdsGR.c b/generic/tclCompCmdsGR.c
new file mode 100644
index 0000000..ff5495c
--- /dev/null
+++ b/generic/tclCompCmdsGR.c
@@ -0,0 +1,3188 @@
+/*
+ * tclCompCmdsGR.c --
+ *
+ *	This file contains compilation procedures that compile various Tcl
+ *	commands (beginning with the letters 'g' through 'r') into a sequence
+ *	of instructions ("bytecodes").
+ *
+ * Copyright (c) 1997-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2002 ActiveState Corporation.
+ * Copyright (c) 2004-2013 by Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include <assert.h>
+
+/*
+ * Prototypes for procedures defined later in this file:
+ */
+
+static void		CompileReturnInternal(CompileEnv *envPtr,
+			    unsigned char op, int code, int level,
+			    Tcl_Obj *returnOpts);
+static int		IndexTailVarIfKnown(Tcl_Interp *interp,
+			    Tcl_Token *varTokenPtr, CompileEnv *envPtr);
+
+#define INDEX_END	(-2)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetIndexFromToken --
+ *
+ *	Parse a token and get the encoded version of the index (as understood
+ *	by TEBC), assuming it is at all knowable at compile time. Only handles
+ *	indices that are integers or 'end' or 'end-integer'.
+ *
+ * Returns:
+ *	TCL_OK if parsing succeeded, and TCL_ERROR if it failed.
+ *
+ * Side effects:
+ *	Sets *index to the index value if successful.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+GetIndexFromToken(
+    Tcl_Token *tokenPtr,
+    int *index)
+{
+    Tcl_Obj *tmpObj = Tcl_NewObj();
+    int result, idx;
+
+    if (!TclWordKnownAtCompileTime(tokenPtr, tmpObj)) {
+	Tcl_DecrRefCount(tmpObj);
+	return TCL_ERROR;
+    }
+
+    result = TclGetIntFromObj(NULL, tmpObj, &idx);
+    if (result == TCL_OK) {
+	if (idx < 0) {
+	    result = TCL_ERROR;
+	}
+    } else {
+	result = TclGetIntForIndexM(NULL, tmpObj, INDEX_END, &idx);
+	if (result == TCL_OK && idx > INDEX_END) {
+	    result = TCL_ERROR;
+	}
+    }
+    Tcl_DecrRefCount(tmpObj);
+
+    if (result == TCL_OK) {
+	*index = idx;
+    }
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileGlobalCmd --
+ *
+ *	Procedure called to compile the "global" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "global" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileGlobalCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr;
+    int localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* TODO: Consider support for compiling expanded args. */
+    numWords = parsePtr->numWords;
+    if (numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * 'global' has no effect outside of proc bodies; handle that at runtime
+     */
+
+    if (envPtr->procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the namespace
+     */
+
+    PushStringLiteral(envPtr, "::");
+
+    /*
+     * Loop over the variables.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    for (i=1; i<numWords; varTokenPtr = TokenAfter(varTokenPtr),i++) {
+	localIndex = IndexTailVarIfKnown(interp, varTokenPtr, envPtr);
+
+	if (localIndex < 0) {
+	    return TCL_ERROR;
+	}
+
+	/* TODO: Consider what value can pass throug the
+	 * IndexTailVarIfKnown() screen.  Full CompileWord()
+	 * likely does not apply here.  Push known value instead. */
+	CompileWord(envPtr, varTokenPtr, interp, i);
+	TclEmitInstInt4(	INST_NSUPVAR, localIndex,	envPtr);
+    }
+
+    /*
+     * Pop the namespace, and set the result to empty
+     */
+
+    TclEmitOpcode(		INST_POP,			envPtr);
+    PushStringLiteral(envPtr, "");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileIfCmd --
+ *
+ *	Procedure called to compile the "if" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "if" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileIfCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    JumpFixupArray jumpFalseFixupArray;
+				/* Used to fix the ifFalse jump after each
+				 * test when its target PC is determined. */
+    JumpFixupArray jumpEndFixupArray;
+				/* Used to fix the jump after each "then" body
+				 * to the end of the "if" when that PC is
+				 * determined. */
+    Tcl_Token *tokenPtr, *testTokenPtr;
+    int jumpIndex = 0;		/* Avoid compiler warning. */
+    int jumpFalseDist, numWords, wordIdx, numBytes, j, code;
+    const char *word;
+    int realCond = 1;		/* Set to 0 for static conditions:
+				 * "if 0 {..}" */
+    int boolVal;		/* Value of static condition. */
+    int compileScripts = 1;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * Only compile the "if" command if all arguments are simple words, in
+     * order to insure correct substitution [Bug 219166]
+     */
+
+    tokenPtr = parsePtr->tokenPtr;
+    wordIdx = 0;
+    numWords = parsePtr->numWords;
+
+    for (wordIdx = 0; wordIdx < numWords; wordIdx++) {
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    return TCL_ERROR;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    TclInitJumpFixupArray(&jumpFalseFixupArray);
+    TclInitJumpFixupArray(&jumpEndFixupArray);
+    code = TCL_OK;
+
+    /*
+     * Each iteration of this loop compiles one "if expr ?then? body" or
+     * "elseif expr ?then? body" clause.
+     */
+
+    tokenPtr = parsePtr->tokenPtr;
+    wordIdx = 0;
+    while (wordIdx < numWords) {
+	/*
+	 * Stop looping if the token isn't "if" or "elseif".
+	 */
+
+	word = tokenPtr[1].start;
+	numBytes = tokenPtr[1].size;
+	if ((tokenPtr == parsePtr->tokenPtr)
+		|| ((numBytes == 6) && (strncmp(word, "elseif", 6) == 0))) {
+	    tokenPtr = TokenAfter(tokenPtr);
+	    wordIdx++;
+	} else {
+	    break;
+	}
+	if (wordIdx >= numWords) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+
+	/*
+	 * Compile the test expression then emit the conditional jump around
+	 * the "then" part.
+	 */
+
+	testTokenPtr = tokenPtr;
+
+	if (realCond) {
+	    /*
+	     * Find out if the condition is a constant.
+	     */
+
+	    Tcl_Obj *boolObj = Tcl_NewStringObj(testTokenPtr[1].start,
+		    testTokenPtr[1].size);
+
+	    Tcl_IncrRefCount(boolObj);
+	    code = Tcl_GetBooleanFromObj(NULL, boolObj, &boolVal);
+	    TclDecrRefCount(boolObj);
+	    if (code == TCL_OK) {
+		/*
+		 * A static condition.
+		 */
+
+		realCond = 0;
+		if (!boolVal) {
+		    compileScripts = 0;
+		}
+	    } else {
+		SetLineInformation(wordIdx);
+		Tcl_ResetResult(interp);
+		TclCompileExprWords(interp, testTokenPtr, 1, envPtr);
+		if (jumpFalseFixupArray.next >= jumpFalseFixupArray.end) {
+		    TclExpandJumpFixupArray(&jumpFalseFixupArray);
+		}
+		jumpIndex = jumpFalseFixupArray.next;
+		jumpFalseFixupArray.next++;
+		TclEmitForwardJump(envPtr, TCL_FALSE_JUMP,
+			jumpFalseFixupArray.fixup+jumpIndex);
+	    }
+	    code = TCL_OK;
+	}
+
+	/*
+	 * Skip over the optional "then" before the then clause.
+	 */
+
+	tokenPtr = TokenAfter(testTokenPtr);
+	wordIdx++;
+	if (wordIdx >= numWords) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	if (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	    word = tokenPtr[1].start;
+	    numBytes = tokenPtr[1].size;
+	    if ((numBytes == 4) && (strncmp(word, "then", 4) == 0)) {
+		tokenPtr = TokenAfter(tokenPtr);
+		wordIdx++;
+		if (wordIdx >= numWords) {
+		    code = TCL_ERROR;
+		    goto done;
+		}
+	    }
+	}
+
+	/*
+	 * Compile the "then" command body.
+	 */
+
+	if (compileScripts) {
+	    BODY(tokenPtr, wordIdx);
+	}
+
+	if (realCond) {
+	    /*
+	     * Jump to the end of the "if" command. Both jumpFalseFixupArray
+	     * and jumpEndFixupArray are indexed by "jumpIndex".
+	     */
+
+	    if (jumpEndFixupArray.next >= jumpEndFixupArray.end) {
+		TclExpandJumpFixupArray(&jumpEndFixupArray);
+	    }
+	    jumpEndFixupArray.next++;
+	    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP,
+		    jumpEndFixupArray.fixup+jumpIndex);
+
+	    /*
+	     * Fix the target of the jumpFalse after the test. Generate a 4
+	     * byte jump if the distance is > 120 bytes. This is conservative,
+	     * and ensures that we won't have to replace this jump if we later
+	     * also need to replace the proceeding jump to the end of the "if"
+	     * with a 4 byte jump.
+	     */
+
+	    TclAdjustStackDepth(-1, envPtr);
+	    if (TclFixupForwardJumpToHere(envPtr,
+		    jumpFalseFixupArray.fixup+jumpIndex, 120)) {
+		/*
+		 * Adjust the code offset for the proceeding jump to the end
+		 * of the "if" command.
+		 */
+
+		jumpEndFixupArray.fixup[jumpIndex].codeOffset += 3;
+	    }
+	} else if (boolVal) {
+	    /*
+	     * We were processing an "if 1 {...}"; stop compiling scripts.
+	     */
+
+	    compileScripts = 0;
+	} else {
+	    /*
+	     * We were processing an "if 0 {...}"; reset so that the rest
+	     * (elseif, else) is compiled correctly.
+	     */
+
+	    realCond = 1;
+	    compileScripts = 1;
+	}
+
+	tokenPtr = TokenAfter(tokenPtr);
+	wordIdx++;
+    }
+
+    /*
+     * Check for the optional else clause. Do not compile anything if this was
+     * an "if 1 {...}" case.
+     */
+
+    if ((wordIdx < numWords) && (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD)) {
+	/*
+	 * There is an else clause. Skip over the optional "else" word.
+	 */
+
+	word = tokenPtr[1].start;
+	numBytes = tokenPtr[1].size;
+	if ((numBytes == 4) && (strncmp(word, "else", 4) == 0)) {
+	    tokenPtr = TokenAfter(tokenPtr);
+	    wordIdx++;
+	    if (wordIdx >= numWords) {
+		code = TCL_ERROR;
+		goto done;
+	    }
+	}
+
+	if (compileScripts) {
+	    /*
+	     * Compile the else command body.
+	     */
+
+	    BODY(tokenPtr, wordIdx);
+	}
+
+	/*
+	 * Make sure there are no words after the else clause.
+	 */
+
+	wordIdx++;
+	if (wordIdx < numWords) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+    } else {
+	/*
+	 * No else clause: the "if" command's result is an empty string.
+	 */
+
+	if (compileScripts) {
+	    PushStringLiteral(envPtr, "");
+	}
+    }
+
+    /*
+     * Fix the unconditional jumps to the end of the "if" command.
+     */
+
+    for (j = jumpEndFixupArray.next;  j > 0;  j--) {
+	jumpIndex = (j - 1);	/* i.e. process the closest jump first. */
+	if (TclFixupForwardJumpToHere(envPtr,
+		jumpEndFixupArray.fixup+jumpIndex, 127)) {
+	    /*
+	     * Adjust the immediately preceeding "ifFalse" jump. We moved it's
+	     * target (just after this jump) down three bytes.
+	     */
+
+	    unsigned char *ifFalsePc = envPtr->codeStart
+		    + jumpFalseFixupArray.fixup[jumpIndex].codeOffset;
+	    unsigned char opCode = *ifFalsePc;
+
+	    if (opCode == INST_JUMP_FALSE1) {
+		jumpFalseDist = TclGetInt1AtPtr(ifFalsePc + 1);
+		jumpFalseDist += 3;
+		TclStoreInt1AtPtr(jumpFalseDist, (ifFalsePc + 1));
+	    } else if (opCode == INST_JUMP_FALSE4) {
+		jumpFalseDist = TclGetInt4AtPtr(ifFalsePc + 1);
+		jumpFalseDist += 3;
+		TclStoreInt4AtPtr(jumpFalseDist, (ifFalsePc + 1));
+	    } else {
+		Tcl_Panic("TclCompileIfCmd: unexpected opcode \"%d\" updating ifFalse jump", (int) opCode);
+	    }
+	}
+    }
+
+    /*
+     * Free the jumpFixupArray array if malloc'ed storage was used.
+     */
+
+  done:
+    TclFreeJumpFixupArray(&jumpFalseFixupArray);
+    TclFreeJumpFixupArray(&jumpEndFixupArray);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileIncrCmd --
+ *
+ *	Procedure called to compile the "incr" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "incr" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileIncrCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr, *incrTokenPtr;
+    int isScalar, localIndex, haveImmValue, immValue;
+    DefineLineInformation;	/* TIP #280 */
+
+    if ((parsePtr->numWords != 2) && (parsePtr->numWords != 3)) {
+	return TCL_ERROR;
+    }
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    PushVarNameWord(interp, varTokenPtr, envPtr, TCL_NO_LARGE_INDEX,
+	    &localIndex, &isScalar, 1);
+
+    /*
+     * If an increment is given, push it, but see first if it's a small
+     * integer.
+     */
+
+    haveImmValue = 0;
+    immValue = 1;
+    if (parsePtr->numWords == 3) {
+	incrTokenPtr = TokenAfter(varTokenPtr);
+	if (incrTokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	    const char *word = incrTokenPtr[1].start;
+	    int numBytes = incrTokenPtr[1].size;
+	    int code;
+	    Tcl_Obj *intObj = Tcl_NewStringObj(word, numBytes);
+
+	    Tcl_IncrRefCount(intObj);
+	    code = TclGetIntFromObj(NULL, intObj, &immValue);
+	    TclDecrRefCount(intObj);
+	    if ((code == TCL_OK) && (-127 <= immValue) && (immValue <= 127)) {
+		haveImmValue = 1;
+	    }
+	    if (!haveImmValue) {
+		PushLiteral(envPtr, word, numBytes);
+	    }
+	} else {
+	    SetLineInformation(2);
+	    CompileTokens(envPtr, incrTokenPtr, interp);
+	}
+    } else {			/* No incr amount given so use 1. */
+	haveImmValue = 1;
+    }
+
+    /*
+     * Emit the instruction to increment the variable.
+     */
+
+    if (isScalar) {	/* Simple scalar variable. */
+	if (localIndex >= 0) {
+	    if (haveImmValue) {
+		TclEmitInstInt1(INST_INCR_SCALAR1_IMM, localIndex, envPtr);
+		TclEmitInt1(immValue, envPtr);
+	    } else {
+		TclEmitInstInt1(INST_INCR_SCALAR1, localIndex,	envPtr);
+	    }
+	} else {
+	    if (haveImmValue) {
+		TclEmitInstInt1(INST_INCR_STK_IMM, immValue, envPtr);
+	    } else {
+		TclEmitOpcode(	INST_INCR_STK,		envPtr);
+	    }
+	}
+    } else {			/* Simple array variable. */
+	if (localIndex >= 0) {
+	    if (haveImmValue) {
+		TclEmitInstInt1(INST_INCR_ARRAY1_IMM, localIndex, envPtr);
+		TclEmitInt1(immValue, envPtr);
+	    } else {
+		TclEmitInstInt1(INST_INCR_ARRAY1, localIndex,	envPtr);
+	    }
+	} else {
+	    if (haveImmValue) {
+		TclEmitInstInt1(INST_INCR_ARRAY_STK_IMM, immValue, envPtr);
+	    } else {
+		TclEmitOpcode(	INST_INCR_ARRAY_STK,		envPtr);
+	    }
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileInfo*Cmd --
+ *
+ *	Procedures called to compile "info" subcommands.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "info" subcommand at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileInfoCommandsCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    Tcl_Obj *objPtr;
+    char *bytes;
+
+    /*
+     * We require one compile-time known argument for the case we can compile.
+     */
+
+    if (parsePtr->numWords == 1) {
+	return TclCompileBasic0ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    } else if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    objPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(objPtr);
+    if (!TclWordKnownAtCompileTime(tokenPtr, objPtr)) {
+	goto notCompilable;
+    }
+    bytes = Tcl_GetString(objPtr);
+
+    /*
+     * We require that the argument start with "::" and not have any of "*\[?"
+     * in it. (Theoretically, we should look in only the final component, but
+     * the difference is so slight given current naming practices.)
+     */
+
+    if (bytes[0] != ':' || bytes[1] != ':' || !TclMatchIsTrivial(bytes)) {
+	goto notCompilable;
+    }
+    Tcl_DecrRefCount(objPtr);
+
+    /*
+     * Confirmed as a literal that will not frighten the horses. Compile. Note
+     * that the result needs to be list-ified.
+     */
+
+    /* TODO: Just push the known value */
+    CompileWord(envPtr, tokenPtr,		interp, 1);
+    TclEmitOpcode(	INST_RESOLVE_COMMAND,	envPtr);
+    TclEmitOpcode(	INST_DUP,		envPtr);
+    TclEmitOpcode(	INST_STR_LEN,		envPtr);
+    TclEmitInstInt1(	INST_JUMP_FALSE1, 7,	envPtr);
+    TclEmitInstInt4(	INST_LIST, 1,		envPtr);
+    return TCL_OK;
+
+  notCompilable:
+    Tcl_DecrRefCount(objPtr);
+    return TclCompileBasic1ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileInfoCoroutineCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Only compile [info coroutine] without arguments.
+     */
+
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Not much to do; we compile to a single instruction...
+     */
+
+    TclEmitOpcode(		INST_COROUTINE_NAME,		envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileInfoExistsCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int isScalar, localIndex;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we need
+     * to emit code to compute and push the name at runtime. We use a frame
+     * slot (entry in the array of local vars) if we are compiling a procedure
+     * body and if the name is simple text that does not include namespace
+     * qualifiers.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    PushVarNameWord(interp, tokenPtr, envPtr, 0, &localIndex, &isScalar, 1);
+
+    /*
+     * Emit instruction to check the variable for existence.
+     */
+
+    if (isScalar) {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_EXIST_STK,			envPtr);
+	} else {
+	    TclEmitInstInt4(	INST_EXIST_SCALAR, localIndex,	envPtr);
+	}
+    } else {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_EXIST_ARRAY_STK,		envPtr);
+	} else {
+	    TclEmitInstInt4(	INST_EXIST_ARRAY, localIndex,	envPtr);
+	}
+    }
+
+    return TCL_OK;
+}
+
+int
+TclCompileInfoLevelCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Only compile [info level] without arguments or with a single argument.
+     */
+
+    if (parsePtr->numWords == 1) {
+	/*
+	 * Not much to do; we compile to a single instruction...
+	 */
+
+	TclEmitOpcode(		INST_INFO_LEVEL_NUM,		envPtr);
+    } else if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    } else {
+	DefineLineInformation;	/* TIP #280 */
+
+	/*
+	 * Compile the argument, then add the instruction to convert it into a
+	 * list of arguments.
+	 */
+
+	CompileWord(envPtr, TokenAfter(parsePtr->tokenPtr), interp, 1);
+	TclEmitOpcode(		INST_INFO_LEVEL_ARGS,		envPtr);
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileInfoObjectClassCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    CompileWord(envPtr,		tokenPtr,		interp, 1);
+    TclEmitOpcode(		INST_TCLOO_CLASS,	envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileInfoObjectIsACmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * We only handle [info object isa object <somevalue>]. The first three
+     * words are compressed to a single token by the ensemble compilation
+     * engine.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || tokenPtr[1].size < 1
+	    || strncmp(tokenPtr[1].start, "object", tokenPtr[1].size)) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(tokenPtr);
+
+    /*
+     * Issue the code.
+     */
+
+    CompileWord(envPtr,		tokenPtr,		interp, 2);
+    TclEmitOpcode(		INST_TCLOO_IS_OBJECT,	envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileInfoObjectNamespaceCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    CompileWord(envPtr,		tokenPtr,		interp, 1);
+    TclEmitOpcode(		INST_TCLOO_NS,		envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLappendCmd --
+ *
+ *	Procedure called to compile the "lappend" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "lappend" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLappendCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr, *valueTokenPtr;
+    int isScalar, localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* TODO: Consider support for compiling expanded args. */
+    numWords = parsePtr->numWords;
+    if (numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    if (numWords != 3 || envPtr->procPtr == NULL) {
+	goto lappendMultiple;
+    }
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we
+     * need to emit code to compute and push the name at runtime. We use a
+     * frame slot (entry in the array of local vars) if we are compiling a
+     * procedure body and if the name is simple text that does not include
+     * namespace qualifiers.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+	    &localIndex, &isScalar, 1);
+
+    /*
+     * If we are doing an assignment, push the new value. In the no values
+     * case, create an empty object.
+     */
+
+    if (numWords > 2) {
+	Tcl_Token *valueTokenPtr = TokenAfter(varTokenPtr);
+
+	CompileWord(envPtr, valueTokenPtr, interp, 2);
+    }
+
+    /*
+     * Emit instructions to set/get the variable.
+     */
+
+    /*
+     * The *_STK opcodes should be refactored to make better use of existing
+     * LOAD/STORE instructions.
+     */
+
+    if (isScalar) {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_LAPPEND_STK,		envPtr);
+	} else {
+	    Emit14Inst(		INST_LAPPEND_SCALAR, localIndex, envPtr);
+	}
+    } else {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_LAPPEND_ARRAY_STK,		envPtr);
+	} else {
+	    Emit14Inst(		INST_LAPPEND_ARRAY, localIndex,	envPtr);
+	}
+    }
+
+    return TCL_OK;
+
+  lappendMultiple:
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+	    &localIndex, &isScalar, 1);
+    valueTokenPtr = TokenAfter(varTokenPtr);
+    for (i = 2 ; i < numWords ; i++) {
+	CompileWord(envPtr, valueTokenPtr, interp, i);
+	valueTokenPtr = TokenAfter(valueTokenPtr);
+    }
+    TclEmitInstInt4(	    INST_LIST, numWords-2,		envPtr);
+    if (isScalar) {
+	if (localIndex < 0) {
+	    TclEmitOpcode(  INST_LAPPEND_LIST_STK,		envPtr);
+	} else {
+	    TclEmitInstInt4(INST_LAPPEND_LIST, localIndex,	envPtr);
+	}
+    } else {
+	if (localIndex < 0) {
+	    TclEmitOpcode(  INST_LAPPEND_LIST_ARRAY_STK,	envPtr);
+	} else {
+	    TclEmitInstInt4(INST_LAPPEND_LIST_ARRAY, localIndex,envPtr);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLassignCmd --
+ *
+ *	Procedure called to compile the "lassign" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "lassign" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLassignCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int isScalar, localIndex, numWords, idx;
+    DefineLineInformation;	/* TIP #280 */
+
+    numWords = parsePtr->numWords;
+
+    /*
+     * Check for command syntax error, but we'll punt that to runtime.
+     */
+
+    if (numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Generate code to push list being taken apart by [lassign].
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+
+    /*
+     * Generate code to assign values from the list to variables.
+     */
+
+    for (idx=0 ; idx<numWords-2 ; idx++) {
+	tokenPtr = TokenAfter(tokenPtr);
+
+	/*
+	 * Generate the next variable name.
+	 */
+
+	PushVarNameWord(interp, tokenPtr, envPtr, 0, &localIndex,
+		&isScalar, idx+2);
+
+	/*
+	 * Emit instructions to get the idx'th item out of the list value on
+	 * the stack and assign it to the variable.
+	 */
+
+	if (isScalar) {
+	    if (localIndex >= 0) {
+		TclEmitOpcode(	INST_DUP,			envPtr);
+		TclEmitInstInt4(INST_LIST_INDEX_IMM, idx,	envPtr);
+		Emit14Inst(	INST_STORE_SCALAR, localIndex,	envPtr);
+		TclEmitOpcode(	INST_POP,			envPtr);
+	    } else {
+		TclEmitInstInt4(INST_OVER, 1,			envPtr);
+		TclEmitInstInt4(INST_LIST_INDEX_IMM, idx,	envPtr);
+		TclEmitOpcode(	INST_STORE_STK,			envPtr);
+		TclEmitOpcode(	INST_POP,			envPtr);
+	    }
+	} else {
+	    if (localIndex >= 0) {
+		TclEmitInstInt4(INST_OVER, 1,			envPtr);
+		TclEmitInstInt4(INST_LIST_INDEX_IMM, idx,	envPtr);
+		Emit14Inst(	INST_STORE_ARRAY, localIndex,	envPtr);
+		TclEmitOpcode(	INST_POP,			envPtr);
+	    } else {
+		TclEmitInstInt4(INST_OVER, 2,			envPtr);
+		TclEmitInstInt4(INST_LIST_INDEX_IMM, idx,	envPtr);
+		TclEmitOpcode(	INST_STORE_ARRAY_STK,		envPtr);
+		TclEmitOpcode(	INST_POP,			envPtr);
+	    }
+	}
+    }
+
+    /*
+     * Generate code to leave the rest of the list on the stack.
+     */
+
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx,	envPtr);
+    TclEmitInt4(			INDEX_END,		envPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLindexCmd --
+ *
+ *	Procedure called to compile the "lindex" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "lindex" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLindexCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *idxTokenPtr, *valTokenPtr;
+    int i, idx, numWords = parsePtr->numWords;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * Quit if too few args.
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (numWords <= 1) {
+	return TCL_ERROR;
+    }
+
+    valTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (numWords != 3) {
+	goto emitComplexLindex;
+    }
+
+    idxTokenPtr = TokenAfter(valTokenPtr);
+    if (GetIndexFromToken(idxTokenPtr, &idx) == TCL_OK) {
+	/*
+	 * All checks have been completed, and we have exactly one of these
+	 * constructs:
+	 *	 lindex <arbitraryValue> <posInt>
+	 *	 lindex <arbitraryValue> end-<posInt>
+	 * This is best compiled as a push of the arbitrary value followed by
+	 * an "immediate lindex" which is the most efficient variety.
+	 */
+
+	CompileWord(envPtr, valTokenPtr, interp, 1);
+	TclEmitInstInt4(	INST_LIST_INDEX_IMM, idx,	envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * If the value was not known at compile time, the conversion failed or
+     * the value was negative, we just keep on going with the more complex
+     * compilation.
+     */
+
+    /*
+     * Push the operands onto the stack.
+     */
+
+  emitComplexLindex:
+    for (i=1 ; i<numWords ; i++) {
+	CompileWord(envPtr, valTokenPtr, interp, i);
+	valTokenPtr = TokenAfter(valTokenPtr);
+    }
+
+    /*
+     * Emit INST_LIST_INDEX if objc==3, or INST_LIST_INDEX_MULTI if there are
+     * multiple index args.
+     */
+
+    if (numWords == 3) {
+	TclEmitOpcode(		INST_LIST_INDEX,		envPtr);
+    } else {
+	TclEmitInstInt4(	INST_LIST_INDEX_MULTI, numWords-1, envPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileListCmd --
+ *
+ *	Procedure called to compile the "list" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "list" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileListCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *valueTokenPtr;
+    int i, numWords, concat, build;
+    Tcl_Obj *listObj, *objPtr;
+
+    if (parsePtr->numWords == 1) {
+	/*
+	 * [list] without arguments just pushes an empty object.
+	 */
+
+	PushStringLiteral(envPtr, "");
+	return TCL_OK;
+    }
+
+    /*
+     * Test if all arguments are compile-time known. If they are, we can
+     * implement with a simple push.
+     */
+
+    numWords = parsePtr->numWords;
+    valueTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    listObj = Tcl_NewObj();
+    for (i = 1; i < numWords && listObj != NULL; i++) {
+	objPtr = Tcl_NewObj();
+	if (TclWordKnownAtCompileTime(valueTokenPtr, objPtr)) {
+	    (void) Tcl_ListObjAppendElement(NULL, listObj, objPtr);
+	} else {
+	    Tcl_DecrRefCount(objPtr);
+	    Tcl_DecrRefCount(listObj);
+	    listObj = NULL;
+	}
+	valueTokenPtr = TokenAfter(valueTokenPtr);
+    }
+    if (listObj != NULL) {
+	TclEmitPush(TclAddLiteralObj(envPtr, listObj, NULL), envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Push the all values onto the stack.
+     */
+
+    numWords = parsePtr->numWords;
+    valueTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    concat = build = 0;
+    for (i = 1; i < numWords; i++) {
+	if (valueTokenPtr->type == TCL_TOKEN_EXPAND_WORD && build > 0) {
+	    TclEmitInstInt4(	INST_LIST, build,	envPtr);
+	    if (concat) {
+		TclEmitOpcode(	INST_LIST_CONCAT,	envPtr);
+	    }
+	    build = 0;
+	    concat = 1;
+	}
+	CompileWord(envPtr, valueTokenPtr, interp, i);
+	if (valueTokenPtr->type == TCL_TOKEN_EXPAND_WORD) {
+	    if (concat) {
+		TclEmitOpcode(	INST_LIST_CONCAT,	envPtr);
+	    } else {
+		concat = 1;
+	    }
+	} else {
+	    build++;
+	}
+	valueTokenPtr = TokenAfter(valueTokenPtr);
+    }
+    if (build > 0) {
+	TclEmitInstInt4(	INST_LIST, build,	envPtr);
+	if (concat) {
+	    TclEmitOpcode(	INST_LIST_CONCAT,	envPtr);
+	}
+    }
+
+    /*
+     * If there was just one expanded word, we must ensure that it is a list
+     * at this point. We use an [lrange ... 0 end] for this (instead of
+     * [llength], as with literals) as we must drop any string representation
+     * that might be hanging around.
+     */
+
+    if (concat && numWords == 2) {
+	TclEmitInstInt4(	INST_LIST_RANGE_IMM, 0,	envPtr);
+	TclEmitInt4(			INDEX_END,	envPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLlengthCmd --
+ *
+ *	Procedure called to compile the "llength" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "llength" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLlengthCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    CompileWord(envPtr, varTokenPtr, interp, 1);
+    TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLrangeCmd --
+ *
+ *	How to compile the "lrange" command. We only bother because we needed
+ *	the opcode anyway for "lassign".
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLrangeCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for context. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *listTokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int idx1, idx2;
+
+    if (parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+    listTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Parse the indices. Will only compile if both are constants and not an
+     * _integer_ less than zero (since we reserve negative indices here for
+     * end-relative indexing) or an end-based index greater than 'end' itself.
+     */
+
+    tokenPtr = TokenAfter(listTokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx1) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(tokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx2) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Issue instructions. It's not safe to skip doing the LIST_RANGE, as
+     * we've not proved that the 'list' argument is really a list. Not that it
+     * is worth trying to do that given current knowledge.
+     */
+
+    CompileWord(envPtr, listTokenPtr, interp, 1);
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx1,	envPtr);
+    TclEmitInt4(		idx2,				envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLinsertCmd --
+ *
+ *	How to compile the "linsert" command. We only bother with the case
+ *	where the index is constant.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLinsertCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for context. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *listTokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int idx, i;
+
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+    listTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Parse the index. Will only compile if it is constant and not an
+     * _integer_ less than zero (since we reserve negative indices here for
+     * end-relative indexing) or an end-based index greater than 'end' itself.
+     */
+
+    tokenPtr = TokenAfter(listTokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * There are four main cases. If there are no values to insert, this is
+     * just a confirm-listiness check. If the index is '0', this is a prepend.
+     * If the index is 'end' (== INDEX_END), this is an append. Otherwise,
+     * this is a splice (== split, insert values as list, concat-3).
+     */
+
+    CompileWord(envPtr, listTokenPtr, interp, 1);
+    if (parsePtr->numWords == 3) {
+	TclEmitInstInt4(	INST_LIST_RANGE_IMM, 0,		envPtr);
+	TclEmitInt4(			INDEX_END,		envPtr);
+	return TCL_OK;
+    }
+
+    for (i=3 ; i<parsePtr->numWords ; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i);
+    }
+    TclEmitInstInt4(		INST_LIST, i-3,			envPtr);
+
+    if (idx == 0 /*start*/) {
+	TclEmitInstInt4(	INST_REVERSE, 2,		envPtr);
+	TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    } else if (idx == INDEX_END /*end*/) {
+	TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    } else {
+	if (idx < 0) {
+	    idx++;
+	}
+	TclEmitInstInt4(	INST_OVER, 1,			envPtr);
+	TclEmitInstInt4(	INST_LIST_RANGE_IMM, 0,		envPtr);
+	TclEmitInt4(			idx-1,			envPtr);
+	TclEmitInstInt4(	INST_REVERSE, 3,		envPtr);
+	TclEmitInstInt4(	INST_LIST_RANGE_IMM, idx,	envPtr);
+	TclEmitInt4(			INDEX_END,		envPtr);
+	TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+	TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLreplaceCmd --
+ *
+ *	How to compile the "lreplace" command. We only bother with the case
+ *	where the indices are constant.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLreplaceCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for context. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *listTokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Obj *tmpObj;
+    int idx1, idx2, i, offset, offset2;
+
+    if (parsePtr->numWords < 4) {
+	return TCL_ERROR;
+    }
+    listTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Parse the indices. Will only compile if both are constants and not an
+     * _integer_ less than zero (since we reserve negative indices here for
+     * end-relative indexing) or an end-based index greater than 'end' itself.
+     */
+
+    tokenPtr = TokenAfter(listTokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx1) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(tokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx2) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * idx1, idx2 are now in canonical form:
+     *
+     *  - integer:	[0,len+1]
+     *  - end index:    INDEX_END
+     *  - -ive offset:  INDEX_END-[len-1,0]
+     *  - +ive offset:  INDEX_END+1
+     */
+
+    /*
+     * Compilation fails when one index is end-based but the other isn't.
+     * Fixing this will require more bytecodes, but this is a workaround for
+     * now. [Bug 47ac84309b]
+     */
+
+    if ((idx1 <= INDEX_END) != (idx2 <= INDEX_END)) {
+	return TCL_ERROR;
+    }
+
+    if (idx2 != INDEX_END && idx2 >= 0 && idx2 < idx1) {
+	idx2 = idx1 - 1;
+    }
+
+    /*
+     * Work out what this [lreplace] is actually doing.
+     */
+
+    tmpObj = NULL;
+    CompileWord(envPtr, listTokenPtr, interp, 1);
+    if (parsePtr->numWords == 4) {
+	if (idx1 == 0) {
+	    if (idx2 == INDEX_END) {
+		goto dropAll;
+	    }
+	    idx1 = idx2 + 1;
+	    idx2 = INDEX_END;
+	    goto dropEnd;
+	} else if (idx2 == INDEX_END) {
+	    idx2 = idx1 - 1;
+	    idx1 = 0;
+	    goto dropEnd;
+	} else {
+	    if (idx2 < idx1) {
+		idx2 = idx1 - 1;
+	    }
+	    if (idx1 > 0) {
+		tmpObj = Tcl_NewIntObj(idx1);
+		Tcl_IncrRefCount(tmpObj);
+	    }
+	    goto dropRange;
+	}
+    }
+
+    tokenPtr = TokenAfter(tokenPtr);
+    for (i=4 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    TclEmitInstInt4(		INST_LIST, i - 4,		envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 2,		envPtr);
+    if (idx1 == 0) {
+	if (idx2 == INDEX_END) {
+	    goto replaceAll;
+	}
+	idx1 = idx2 + 1;
+	idx2 = INDEX_END;
+	goto replaceHead;
+    } else if (idx2 == INDEX_END) {
+	idx2 = idx1 - 1;
+	idx1 = 0;
+	goto replaceTail;
+    } else {
+	if (idx2 < idx1) {
+	    idx2 = idx1 - 1;
+	}
+	if (idx1 > 0) {
+	    tmpObj = Tcl_NewIntObj(idx1);
+	    Tcl_IncrRefCount(tmpObj);
+	}
+	goto replaceRange;
+    }
+
+    /*
+     * Issue instructions to perform the operations relating to configurations
+     * that just drop. The only argument pushed on the stack is the list to
+     * operate on.
+     */
+
+  dropAll:			/* This just ensures the arg is a list. */
+    TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+    PushStringLiteral(envPtr,	"");
+    goto done;
+
+  dropEnd:
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx1,	envPtr);
+    TclEmitInt4(			idx2,			envPtr);
+    goto done;
+
+  dropRange:
+    if (tmpObj != NULL) {
+	/*
+	 * Emit bytecode to check the list length.
+	 */
+
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+	TclEmitPush(TclAddLiteralObj(envPtr, tmpObj, NULL),	envPtr);
+	TclEmitOpcode(		INST_GE,			envPtr);
+	offset = CurrentOffset(envPtr);
+	TclEmitInstInt1(	INST_JUMP_TRUE1, 0,		envPtr);
+
+	/*
+	 * Emit an error if we've been given an empty list.
+	 */
+
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+	offset2 = CurrentOffset(envPtr);
+	TclEmitInstInt1(	INST_JUMP_FALSE1, 0,		envPtr);
+	TclEmitPush(TclAddLiteralObj(envPtr, Tcl_ObjPrintf(
+		"list doesn't contain element %d", idx1), NULL), envPtr);
+	CompileReturnInternal(envPtr, INST_RETURN_IMM, TCL_ERROR, 0,
+		Tcl_ObjPrintf("-errorcode {TCL OPERATION LREPLACE BADIDX}"));
+	TclStoreInt1AtPtr(CurrentOffset(envPtr) - offset,
+		envPtr->codeStart + offset + 1);
+	TclStoreInt1AtPtr(CurrentOffset(envPtr) - offset2,
+		envPtr->codeStart + offset2 + 1);
+	TclAdjustStackDepth(-1, envPtr);
+    }
+    TclEmitOpcode(		INST_DUP,			envPtr);
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, 0,		envPtr);
+    TclEmitInt4(			idx1 - 1,		envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 2,		envPtr);
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx2 + 1,	envPtr);
+    TclEmitInt4(			INDEX_END,		envPtr);
+    TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    goto done;
+
+    /*
+     * Issue instructions to perform the operations relating to configurations
+     * that do real replacement. All arguments are pushed and assembled into a
+     * pair: the list of values to replace with, and the list to do the
+     * surgery on.
+     */
+
+  replaceAll:
+    TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+    goto done;
+
+  replaceHead:
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx1,	envPtr);
+    TclEmitInt4(			idx2,			envPtr);
+    TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    goto done;
+
+  replaceTail:
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx1,	envPtr);
+    TclEmitInt4(			idx2,			envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 2,		envPtr);
+    TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    goto done;
+
+  replaceRange:
+    if (tmpObj != NULL) {
+	/*
+	 * Emit bytecode to check the list length.
+	 */
+
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+
+	/*
+	 * Check the list length vs idx1.
+	 */
+
+	TclEmitPush(TclAddLiteralObj(envPtr, tmpObj, NULL),	envPtr);
+	TclEmitOpcode(		INST_GE,			envPtr);
+	offset = CurrentOffset(envPtr);
+	TclEmitInstInt1(	INST_JUMP_TRUE1, 0,		envPtr);
+
+	/*
+	 * Emit an error if we've been given an empty list.
+	 */
+
+	TclEmitOpcode(		INST_DUP,			envPtr);
+	TclEmitOpcode(		INST_LIST_LENGTH,		envPtr);
+	offset2 = CurrentOffset(envPtr);
+	TclEmitInstInt1(	INST_JUMP_FALSE1, 0,		envPtr);
+	TclEmitPush(TclAddLiteralObj(envPtr, Tcl_ObjPrintf(
+		"list doesn't contain element %d", idx1), NULL), envPtr);
+	CompileReturnInternal(envPtr, INST_RETURN_IMM, TCL_ERROR, 0,
+		Tcl_ObjPrintf("-errorcode {TCL OPERATION LREPLACE BADIDX}"));
+	TclStoreInt1AtPtr(CurrentOffset(envPtr) - offset,
+		envPtr->codeStart + offset + 1);
+	TclStoreInt1AtPtr(CurrentOffset(envPtr) - offset2,
+		envPtr->codeStart + offset2 + 1);
+	TclAdjustStackDepth(-1, envPtr);
+    }
+    TclEmitOpcode(		INST_DUP,			envPtr);
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, 0,		envPtr);
+    TclEmitInt4(			idx1 - 1,		envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 2,		envPtr);
+    TclEmitInstInt4(		INST_LIST_RANGE_IMM, idx2 + 1,	envPtr);
+    TclEmitInt4(			INDEX_END,		envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 3,		envPtr);
+    TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    TclEmitInstInt4(		INST_REVERSE, 2,		envPtr);
+    TclEmitOpcode(		INST_LIST_CONCAT,		envPtr);
+    goto done;
+
+    /*
+     * Clean up the allocated memory.
+     */
+
+  done:
+    if (tmpObj != NULL) {
+	Tcl_DecrRefCount(tmpObj);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileLsetCmd --
+ *
+ *	Procedure called to compile the "lset" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "lset" command at
+ *	runtime.
+ *
+ * The general template for execution of the "lset" command is:
+ *	(1) Instructions to push the variable name, unless the variable is
+ *	    local to the stack frame.
+ *	(2) If the variable is an array element, instructions to push the
+ *	    array element name.
+ *	(3) Instructions to push each of zero or more "index" arguments to the
+ *	    stack, followed with the "newValue" element.
+ *	(4) Instructions to duplicate the variable name and/or array element
+ *	    name onto the top of the stack, if either was pushed at steps (1)
+ *	    and (2).
+ *	(5) The appropriate INST_LOAD_* instruction to place the original
+ *	    value of the list variable at top of stack.
+ *	(6) At this point, the stack contains:
+ *		varName? arrayElementName? index1 index2 ... newValue oldList
+ *	    The compiler emits one of INST_LSET_FLAT or INST_LSET_LIST
+ *	    according as whether there is exactly one index element (LIST) or
+ *	    either zero or else two or more (FLAT). This instruction removes
+ *	    everything from the stack except for the two names and pushes the
+ *	    new value of the variable.
+ *	(7) Finally, INST_STORE_* stores the new value in the variable and
+ *	    cleans up the stack.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileLsetCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    int tempDepth;		/* Depth used for emitting one part of the
+				 * code burst. */
+    Tcl_Token *varTokenPtr;	/* Pointer to the Tcl_Token representing the
+				 * parse of the variable name. */
+    int localIndex;		/* Index of var in local var table. */
+    int isScalar;		/* Flag == 1 if scalar, 0 if array. */
+    int i;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * Check argument count.
+     */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	/*
+	 * Fail at run time, not in compilation.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we need
+     * to emit code to compute and push the name at runtime. We use a frame
+     * slot (entry in the array of local vars) if we are compiling a procedure
+     * body and if the name is simple text that does not include namespace
+     * qualifiers.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+	    &localIndex, &isScalar, 1);
+
+    /*
+     * Push the "index" args and the new element value.
+     */
+
+    for (i=2 ; i<parsePtr->numWords ; ++i) {
+	varTokenPtr = TokenAfter(varTokenPtr);
+	CompileWord(envPtr, varTokenPtr, interp, i);
+    }
+
+    /*
+     * Duplicate the variable name if it's been pushed.
+     */
+
+    if (localIndex < 0) {
+	if (isScalar) {
+	    tempDepth = parsePtr->numWords - 2;
+	} else {
+	    tempDepth = parsePtr->numWords - 1;
+	}
+	TclEmitInstInt4(	INST_OVER, tempDepth,		envPtr);
+    }
+
+    /*
+     * Duplicate an array index if one's been pushed.
+     */
+
+    if (!isScalar) {
+	if (localIndex < 0) {
+	    tempDepth = parsePtr->numWords - 1;
+	} else {
+	    tempDepth = parsePtr->numWords - 2;
+	}
+	TclEmitInstInt4(	INST_OVER, tempDepth,		envPtr);
+    }
+
+    /*
+     * Emit code to load the variable's value.
+     */
+
+    if (isScalar) {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_LOAD_STK,			envPtr);
+	} else {
+	    Emit14Inst(		INST_LOAD_SCALAR, localIndex,	envPtr);
+	}
+    } else {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_LOAD_ARRAY_STK,		envPtr);
+	} else {
+	    Emit14Inst(		INST_LOAD_ARRAY, localIndex,	envPtr);
+	}
+    }
+
+    /*
+     * Emit the correct variety of 'lset' instruction.
+     */
+
+    if (parsePtr->numWords == 4) {
+	TclEmitOpcode(		INST_LSET_LIST,			envPtr);
+    } else {
+	TclEmitInstInt4(	INST_LSET_FLAT, parsePtr->numWords-1, envPtr);
+    }
+
+    /*
+     * Emit code to put the value back in the variable.
+     */
+
+    if (isScalar) {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_STORE_STK,			envPtr);
+	} else {
+	    Emit14Inst(		INST_STORE_SCALAR, localIndex,	envPtr);
+	}
+    } else {
+	if (localIndex < 0) {
+	    TclEmitOpcode(	INST_STORE_ARRAY_STK,		envPtr);
+	} else {
+	    Emit14Inst(		INST_STORE_ARRAY, localIndex,	envPtr);
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileNamespace*Cmd --
+ *
+ *	Procedures called to compile the "namespace" command; currently, only
+ *	the subcommands "namespace current" and "namespace upvar" are compiled
+ *	to bytecodes, and the latter only inside a procedure(-like) context.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "namespace upvar"
+ *	command at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileNamespaceCurrentCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Only compile [namespace current] without arguments.
+     */
+
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Not much to do; we compile to a single instruction...
+     */
+
+    TclEmitOpcode(		INST_NS_CURRENT,		envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceCodeCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * The specification of [namespace code] is rather shocking, in that it is
+     * supposed to check if the argument is itself the result of [namespace
+     * code] and not apply itself in that case. Which is excessively cautious,
+     * but what the test suite checks for.
+     */
+
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || (tokenPtr[1].size > 20
+	    && strncmp(tokenPtr[1].start, "::namespace inscope ", 20) == 0)) {
+	/*
+	 * Technically, we could just pass a literal '::namespace inscope '
+	 * term through, but that's something which really shouldn't be
+	 * occurring as something that the user writes so we'll just punt it.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now we can compile using the same strategy as [namespace code]'s normal
+     * implementation does internally. Note that we can't bind the namespace
+     * name directly here, because TclOO plays complex games with namespaces;
+     * the value needs to be determined at runtime for safety.
+     */
+
+    PushStringLiteral(envPtr,		"::namespace");
+    PushStringLiteral(envPtr,		"inscope");
+    TclEmitOpcode(		INST_NS_CURRENT,	envPtr);
+    CompileWord(envPtr,		tokenPtr,		interp, 1);
+    TclEmitInstInt4(		INST_LIST, 4,		envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceOriginCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    CompileWord(envPtr,	tokenPtr,			interp, 1);
+    TclEmitOpcode(	INST_ORIGIN_COMMAND,		envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceQualifiersCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    DefineLineInformation;	/* TIP #280 */
+    int off;
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    PushStringLiteral(envPtr, "0");
+    PushStringLiteral(envPtr, "::");
+    TclEmitInstInt4(	INST_OVER, 2,			envPtr);
+    TclEmitOpcode(	INST_STR_FIND_LAST,		envPtr);
+    off = CurrentOffset(envPtr);
+    PushStringLiteral(envPtr, "1");
+    TclEmitOpcode(	INST_SUB,			envPtr);
+    TclEmitInstInt4(	INST_OVER, 2,			envPtr);
+    TclEmitInstInt4(	INST_OVER, 1,			envPtr);
+    TclEmitOpcode(	INST_STR_INDEX,			envPtr);
+    PushStringLiteral(envPtr, ":");
+    TclEmitOpcode(	INST_STR_EQ,			envPtr);
+    off = off - CurrentOffset(envPtr);
+    TclEmitInstInt1(	INST_JUMP_TRUE1, off,		envPtr);
+    TclEmitOpcode(	INST_STR_RANGE,			envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceTailCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    DefineLineInformation;	/* TIP #280 */
+    JumpFixup jumpFixup;
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Take care; only add 2 to found index if the string was actually found.
+     */
+
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    PushStringLiteral(envPtr, "::");
+    TclEmitInstInt4(	INST_OVER, 1,			envPtr);
+    TclEmitOpcode(	INST_STR_FIND_LAST,		envPtr);
+    TclEmitOpcode(	INST_DUP,			envPtr);
+    PushStringLiteral(envPtr, "0");
+    TclEmitOpcode(	INST_GE,			envPtr);
+    TclEmitForwardJump(envPtr, TCL_FALSE_JUMP, &jumpFixup);
+    PushStringLiteral(envPtr, "2");
+    TclEmitOpcode(	INST_ADD,			envPtr);
+    TclFixupForwardJumpToHere(envPtr, &jumpFixup, 127);
+    PushStringLiteral(envPtr, "end");
+    TclEmitOpcode(	INST_STR_RANGE,			envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceUpvarCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *otherTokenPtr, *localTokenPtr;
+    int localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (envPtr->procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Only compile [namespace upvar ...]: needs an even number of args, >=4
+     */
+
+    numWords = parsePtr->numWords;
+    if ((numWords % 2) || (numWords < 4)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the namespace
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+
+    /*
+     * Loop over the (otherVar, thisVar) pairs. If any of the thisVar is not a
+     * local variable, return an error so that the non-compiled command will
+     * be called at runtime.
+     */
+
+    localTokenPtr = tokenPtr;
+    for (i=2; i<numWords; i+=2) {
+	otherTokenPtr = TokenAfter(localTokenPtr);
+	localTokenPtr = TokenAfter(otherTokenPtr);
+
+	CompileWord(envPtr, otherTokenPtr, interp, i);
+	localIndex = LocalScalarFromToken(localTokenPtr, envPtr);
+	if (localIndex < 0) {
+	    return TCL_ERROR;
+	}
+	TclEmitInstInt4(	INST_NSUPVAR, localIndex,	envPtr);
+    }
+
+    /*
+     * Pop the namespace, and set the result to empty
+     */
+
+    TclEmitOpcode(		INST_POP,			envPtr);
+    PushStringLiteral(envPtr, "");
+    return TCL_OK;
+}
+
+int
+TclCompileNamespaceWhichCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr, *opt;
+    int idx;
+
+    if (parsePtr->numWords < 2 || parsePtr->numWords > 3) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    idx = 1;
+
+    /*
+     * If there's an option, check that it's "-command". We don't handle
+     * "-variable" (currently) and anything else is an error.
+     */
+
+    if (parsePtr->numWords == 3) {
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    return TCL_ERROR;
+	}
+	opt = tokenPtr + 1;
+	if (opt->size < 2 || opt->size > 8
+		|| strncmp(opt->start, "-command", opt->size) != 0) {
+	    return TCL_ERROR;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+	idx++;
+    }
+
+    /*
+     * Issue the bytecode.
+     */
+
+    CompileWord(envPtr,		tokenPtr,		interp, idx);
+    TclEmitOpcode(		INST_RESOLVE_COMMAND,	envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileRegexpCmd --
+ *
+ *	Procedure called to compile the "regexp" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "regexp" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileRegexpCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *varTokenPtr;	/* Pointer to the Tcl_Token representing the
+				 * parse of the RE or string. */
+    int i, len, nocase, exact, sawLast, simple;
+    const char *str;
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * We are only interested in compiling simple regexp cases. Currently
+     * supported compile cases are:
+     *   regexp ?-nocase? ?--? staticString $var
+     *   regexp ?-nocase? ?--? {^staticString$} $var
+     */
+
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    simple = 0;
+    nocase = 0;
+    sawLast = 0;
+    varTokenPtr = parsePtr->tokenPtr;
+
+    /*
+     * We only look for -nocase and -- as options. Everything else gets pushed
+     * to runtime execution. This is different than regexp's runtime option
+     * handling, but satisfies our stricter needs.
+     */
+
+    for (i = 1; i < parsePtr->numWords - 2; i++) {
+	varTokenPtr = TokenAfter(varTokenPtr);
+	if (varTokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    /*
+	     * Not a simple string, so punt to runtime.
+	     */
+
+	    return TCL_ERROR;
+	}
+	str = varTokenPtr[1].start;
+	len = varTokenPtr[1].size;
+	if ((len == 2) && (str[0] == '-') && (str[1] == '-')) {
+	    sawLast++;
+	    i++;
+	    break;
+	} else if ((len > 1) && (strncmp(str,"-nocase",(unsigned)len) == 0)) {
+	    nocase = 1;
+	} else {
+	    /*
+	     * Not an option we recognize.
+	     */
+
+	    return TCL_ERROR;
+	}
+    }
+
+    if ((parsePtr->numWords - i) != 2) {
+	/*
+	 * We don't support capturing to variables.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the regexp string. If it is not a simple string or can't be
+     * converted to a glob pattern, push the word for the INST_REGEXP.
+     * Keep changes here in sync with TclCompileSwitchCmd Switch_Regexp.
+     */
+
+    varTokenPtr = TokenAfter(varTokenPtr);
+
+    if (varTokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	Tcl_DString ds;
+
+	str = varTokenPtr[1].start;
+	len = varTokenPtr[1].size;
+
+	/*
+	 * If it has a '-', it could be an incorrectly formed regexp command.
+	 */
+
+	if ((*str == '-') && !sawLast) {
+	    return TCL_ERROR;
+	}
+
+	if (len == 0) {
+	    /*
+	     * The semantics of regexp are always match on re == "".
+	     */
+
+	    PushStringLiteral(envPtr, "1");
+	    return TCL_OK;
+	}
+
+	/*
+	 * Attempt to convert pattern to glob.  If successful, push the
+	 * converted pattern as a literal.
+	 */
+
+	if (TclReToGlob(NULL, varTokenPtr[1].start, len, &ds, &exact, NULL)
+		== TCL_OK) {
+	    simple = 1;
+	    PushLiteral(envPtr, Tcl_DStringValue(&ds),Tcl_DStringLength(&ds));
+	    Tcl_DStringFree(&ds);
+	}
+    }
+
+    if (!simple) {
+	CompileWord(envPtr, varTokenPtr, interp, parsePtr->numWords-2);
+    }
+
+    /*
+     * Push the string arg.
+     */
+
+    varTokenPtr = TokenAfter(varTokenPtr);
+    CompileWord(envPtr, varTokenPtr, interp, parsePtr->numWords-1);
+
+    if (simple) {
+	if (exact && !nocase) {
+	    TclEmitOpcode(	INST_STR_EQ,			envPtr);
+	} else {
+	    TclEmitInstInt1(	INST_STR_MATCH, nocase,		envPtr);
+	}
+    } else {
+	/*
+	 * Pass correct RE compile flags.  We use only Int1 (8-bit), but
+	 * that handles all the flags we want to pass.
+	 * Don't use TCL_REG_NOSUB as we may have backrefs.
+	 */
+
+	int cflags = TCL_REG_ADVANCED | (nocase ? TCL_REG_NOCASE : 0);
+
+	TclEmitInstInt1(	INST_REGEXP, cflags,		envPtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileRegsubCmd --
+ *
+ *	Procedure called to compile the "regsub" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "regsub" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileRegsubCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    /*
+     * We only compile the case with [regsub -all] where the pattern is both
+     * known at compile time and simple (i.e., no RE metacharacters). That is,
+     * the pattern must be translatable into a glob like "*foo*" with no other
+     * glob metacharacters inside it; there must be some "foo" in there too.
+     * The substitution string must also be known at compile time and free of
+     * metacharacters ("\digit" and "&"). Finally, there must not be a
+     * variable mentioned in the [regsub] to write the result back to (because
+     * we can't get the count of substitutions that would be the result in
+     * that case). The key is that these are the conditions under which a
+     * [string map] could be used instead, in particular a [string map] of the
+     * form we can compile to bytecode.
+     *
+     * In short, we look for:
+     *
+     *   regsub -all [--] simpleRE string simpleReplacement
+     *
+     * The only optional part is the "--", and no other options are handled.
+     */
+
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr, *stringTokenPtr;
+    Tcl_Obj *patternObj = NULL, *replacementObj = NULL;
+    Tcl_DString pattern;
+    const char *bytes;
+    int len, exact, quantified, result = TCL_ERROR;
+
+    if (parsePtr->numWords < 5 || parsePtr->numWords > 6) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the "-all", which must be the first argument (other options not
+     * supported, non-"-all" substitution we can't compile).
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || tokenPtr[1].size != 4
+	    || strncmp(tokenPtr[1].start, "-all", 4)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the pattern into patternObj, checking for "--" in the process.
+     */
+
+    Tcl_DStringInit(&pattern);
+    tokenPtr = TokenAfter(tokenPtr);
+    patternObj = Tcl_NewObj();
+    if (!TclWordKnownAtCompileTime(tokenPtr, patternObj)) {
+	goto done;
+    }
+    if (Tcl_GetString(patternObj)[0] == '-') {
+	if (strcmp(Tcl_GetString(patternObj), "--") != 0
+		|| parsePtr->numWords == 5) {
+	    goto done;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+	Tcl_DecrRefCount(patternObj);
+	patternObj = Tcl_NewObj();
+	if (!TclWordKnownAtCompileTime(tokenPtr, patternObj)) {
+	    goto done;
+	}
+    } else if (parsePtr->numWords == 6) {
+	goto done;
+    }
+
+    /*
+     * Identify the code which produces the string to apply the substitution
+     * to (stringTokenPtr), and the replacement string (into replacementObj).
+     */
+
+    stringTokenPtr = TokenAfter(tokenPtr);
+    tokenPtr = TokenAfter(stringTokenPtr);
+    replacementObj = Tcl_NewObj();
+    if (!TclWordKnownAtCompileTime(tokenPtr, replacementObj)) {
+	goto done;
+    }
+
+    /*
+     * Next, higher-level checks. Is the RE a very simple glob? Is the
+     * replacement "simple"?
+     */
+
+    bytes = TclGetStringFromObj(patternObj, &len);
+    if (TclReToGlob(NULL, bytes, len, &pattern, &exact, &quantified)
+	    != TCL_OK || exact || quantified) {
+	goto done;
+    }
+    bytes = Tcl_DStringValue(&pattern);
+    if (*bytes++ != '*') {
+	goto done;
+    }
+    while (1) {
+	switch (*bytes) {
+	case '*':
+	    if (bytes[1] == '\0') {
+		/*
+		 * OK, we've proved there are no metacharacters except for the
+		 * '*' at each end.
+		 */
+
+		len = Tcl_DStringLength(&pattern) - 2;
+		if (len > 0) {
+		    goto isSimpleGlob;
+		}
+
+		/*
+		 * The pattern is "**"! I believe that should be impossible,
+		 * but we definitely can't handle that at all.
+		 */
+	    }
+	case '\0': case '?': case '[': case '\\':
+	    goto done;
+	}
+	bytes++;
+    }
+  isSimpleGlob:
+    for (bytes = Tcl_GetString(replacementObj); *bytes; bytes++) {
+	switch (*bytes) {
+	case '\\': case '&':
+	    goto done;
+	}
+    }
+
+    /*
+     * Proved the simplicity constraints! Time to issue the code.
+     */
+
+    result = TCL_OK;
+    bytes = Tcl_DStringValue(&pattern) + 1;
+    PushLiteral(envPtr,	bytes, len);
+    bytes = TclGetStringFromObj(replacementObj, &len);
+    PushLiteral(envPtr,	bytes, len);
+    CompileWord(envPtr,	stringTokenPtr, interp, parsePtr->numWords-2);
+    TclEmitOpcode(	INST_STR_MAP,	envPtr);
+
+  done:
+    Tcl_DStringFree(&pattern);
+    if (patternObj) {
+	Tcl_DecrRefCount(patternObj);
+    }
+    if (replacementObj) {
+	Tcl_DecrRefCount(replacementObj);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileReturnCmd --
+ *
+ *	Procedure called to compile the "return" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "return" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileReturnCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * General syntax: [return ?-option value ...? ?result?]
+     * An even number of words means an explicit result argument is present.
+     */
+    int level, code, objc, size, status = TCL_OK;
+    int numWords = parsePtr->numWords;
+    int explicitResult = (0 == (numWords % 2));
+    int numOptionWords = numWords - 1 - explicitResult;
+    Tcl_Obj *returnOpts, **objv;
+    Tcl_Token *wordTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    DefineLineInformation;	/* TIP #280 */
+
+    /*
+     * Check for special case which can always be compiled:
+     *	    return -options <opts> <msg>
+     * Unlike the normal [return] compilation, this version does everything at
+     * runtime so it can handle arbitrary words and not just literals. Note
+     * that if INST_RETURN_STK wasn't already needed for something else
+     * ('finally' clause processing) this piece of code would not be present.
+     */
+
+    if ((numWords == 4) && (wordTokenPtr->type == TCL_TOKEN_SIMPLE_WORD)
+	    && (wordTokenPtr[1].size == 8)
+	    && (strncmp(wordTokenPtr[1].start, "-options", 8) == 0)) {
+	Tcl_Token *optsTokenPtr = TokenAfter(wordTokenPtr);
+	Tcl_Token *msgTokenPtr = TokenAfter(optsTokenPtr);
+
+	CompileWord(envPtr, optsTokenPtr, interp, 2);
+	CompileWord(envPtr, msgTokenPtr,  interp, 3);
+	TclEmitInvoke(envPtr, INST_RETURN_STK);
+	return TCL_OK;
+    }
+
+    /*
+     * Allocate some working space.
+     */
+
+    objv = TclStackAlloc(interp, numOptionWords * sizeof(Tcl_Obj *));
+
+    /*
+     * Scan through the return options. If any are unknown at compile time,
+     * there is no value in bytecompiling. Save the option values known in an
+     * objv array for merging into a return options dictionary.
+     *
+     * TODO: There is potential for improvement if all option keys are known
+     * at compile time and all option values relating to '-code' and '-level'
+     * are known at compile time.
+     */
+
+    for (objc = 0; objc < numOptionWords; objc++) {
+	objv[objc] = Tcl_NewObj();
+	Tcl_IncrRefCount(objv[objc]);
+	if (!TclWordKnownAtCompileTime(wordTokenPtr, objv[objc])) {
+	    /*
+	     * Non-literal, so punt to run-time assembly of the dictionary.
+	     */
+
+	    for (; objc>=0 ; objc--) {
+		TclDecrRefCount(objv[objc]);
+	    }
+	    TclStackFree(interp, objv);
+	    goto issueRuntimeReturn;
+	}
+	wordTokenPtr = TokenAfter(wordTokenPtr);
+    }
+    status = TclMergeReturnOptions(interp, objc, objv,
+	    &returnOpts, &code, &level);
+    while (--objc >= 0) {
+	TclDecrRefCount(objv[objc]);
+    }
+    TclStackFree(interp, objv);
+    if (TCL_ERROR == status) {
+	/*
+	 * Something was bogus in the return options. Clear the error message,
+	 * and report back to the compiler that this must be interpreted at
+	 * runtime.
+	 */
+
+	Tcl_ResetResult(interp);
+	return TCL_ERROR;
+    }
+
+    /*
+     * All options are known at compile time, so we're going to bytecompile.
+     * Emit instructions to push the result on the stack.
+     */
+
+    if (explicitResult) {
+	 CompileWord(envPtr, wordTokenPtr, interp, numWords-1);
+    } else {
+	/*
+	 * No explict result argument, so default result is empty string.
+	 */
+
+	PushStringLiteral(envPtr, "");
+    }
+
+    /*
+     * Check for optimization: When [return] is in a proc, and there's no
+     * enclosing [catch], and there are no return options, then the INST_DONE
+     * instruction is equivalent, and may be more efficient.
+     */
+
+    if (numOptionWords == 0 && envPtr->procPtr != NULL) {
+	/*
+	 * We have default return options and we're in a proc ...
+	 */
+
+	int index = envPtr->exceptArrayNext - 1;
+	int enclosingCatch = 0;
+
+	while (index >= 0) {
+	    ExceptionRange range = envPtr->exceptArrayPtr[index];
+
+	    if ((range.type == CATCH_EXCEPTION_RANGE)
+		    && (range.catchOffset == -1)) {
+		enclosingCatch = 1;
+		break;
+	    }
+	    index--;
+	}
+	if (!enclosingCatch) {
+	    /*
+	     * ... and there is no enclosing catch. Issue the maximally
+	     * efficient exit instruction.
+	     */
+
+	    Tcl_DecrRefCount(returnOpts);
+	    TclEmitOpcode(INST_DONE, envPtr);
+	    TclAdjustStackDepth(1, envPtr);
+	    return TCL_OK;
+	}
+    }
+
+    /* Optimize [return -level 0 $x]. */
+    Tcl_DictObjSize(NULL, returnOpts, &size);
+    if (size == 0 && level == 0 && code == TCL_OK) {
+	Tcl_DecrRefCount(returnOpts);
+	return TCL_OK;
+    }
+
+    /*
+     * Could not use the optimization, so we push the return options dict, and
+     * emit the INST_RETURN_IMM instruction with code and level as operands.
+     */
+
+    CompileReturnInternal(envPtr, INST_RETURN_IMM, code, level, returnOpts);
+    return TCL_OK;
+
+  issueRuntimeReturn:
+    /*
+     * Assemble the option dictionary (as a list as that's good enough).
+     */
+
+    wordTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    for (objc=1 ; objc<=numOptionWords ; objc++) {
+	CompileWord(envPtr, wordTokenPtr, interp, objc);
+	wordTokenPtr = TokenAfter(wordTokenPtr);
+    }
+    TclEmitInstInt4(INST_LIST, numOptionWords, envPtr);
+
+    /*
+     * Push the result.
+     */
+
+    if (explicitResult) {
+	CompileWord(envPtr, wordTokenPtr, interp, numWords-1);
+    } else {
+	PushStringLiteral(envPtr, "");
+    }
+
+    /*
+     * Issue the RETURN itself.
+     */
+
+    TclEmitInvoke(envPtr, INST_RETURN_STK);
+    return TCL_OK;
+}
+
+static void
+CompileReturnInternal(
+    CompileEnv *envPtr,
+    unsigned char op,
+    int code,
+    int level,
+    Tcl_Obj *returnOpts)
+{
+    if (level == 0 && (code == TCL_BREAK || code == TCL_CONTINUE)) {
+	ExceptionRange *rangePtr;
+	ExceptionAux *exceptAux;
+
+	rangePtr = TclGetInnermostExceptionRange(envPtr, code, &exceptAux);
+	if (rangePtr && rangePtr->type == LOOP_EXCEPTION_RANGE) {
+	    TclCleanupStackForBreakContinue(envPtr, exceptAux);
+	    if (code == TCL_BREAK) {
+		TclAddLoopBreakFixup(envPtr, exceptAux);
+	    } else {
+		TclAddLoopContinueFixup(envPtr, exceptAux);
+	    }
+	    Tcl_DecrRefCount(returnOpts);
+	    return;
+	}
+    }
+
+    TclEmitPush(TclAddLiteralObj(envPtr, returnOpts, NULL), envPtr);
+    TclEmitInstInt4(op, code, envPtr);
+    TclEmitInt4(level, envPtr);
+}
+
+void
+TclCompileSyntaxError(
+    Tcl_Interp *interp,
+    CompileEnv *envPtr)
+{
+    Tcl_Obj *msg = Tcl_GetObjResult(interp);
+    int numBytes;
+    const char *bytes = TclGetStringFromObj(msg, &numBytes);
+
+    TclErrorStackResetIf(interp, bytes, numBytes);
+    TclEmitPush(TclRegisterLiteral(envPtr, bytes, numBytes, 0), envPtr);
+    CompileReturnInternal(envPtr, INST_SYNTAX, TCL_ERROR, 0,
+	    TclNoErrorStack(interp, Tcl_GetReturnOptions(interp, TCL_ERROR)));
+    Tcl_ResetResult(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileUpvarCmd --
+ *
+ *	Procedure called to compile the "upvar" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "upvar" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileUpvarCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *otherTokenPtr, *localTokenPtr;
+    int localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Obj *objPtr;
+
+    if (envPtr->procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    numWords = parsePtr->numWords;
+    if (numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the frame index if it is known at compile time
+     */
+
+    objPtr = Tcl_NewObj();
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (TclWordKnownAtCompileTime(tokenPtr, objPtr)) {
+	CallFrame *framePtr;
+	const Tcl_ObjType *newTypePtr, *typePtr = objPtr->typePtr;
+
+	/*
+	 * Attempt to convert to a level reference. Note that TclObjGetFrame
+	 * only changes the obj type when a conversion was successful.
+	 */
+
+	TclObjGetFrame(interp, objPtr, &framePtr);
+	newTypePtr = objPtr->typePtr;
+	Tcl_DecrRefCount(objPtr);
+
+	if (newTypePtr != typePtr) {
+	    if (numWords%2) {
+		return TCL_ERROR;
+	    }
+	    /* TODO: Push the known value instead? */
+	    CompileWord(envPtr, tokenPtr, interp, 1);
+	    otherTokenPtr = TokenAfter(tokenPtr);
+	    i = 2;
+	} else {
+	    if (!(numWords%2)) {
+		return TCL_ERROR;
+	    }
+	    PushStringLiteral(envPtr, "1");
+	    otherTokenPtr = tokenPtr;
+	    i = 1;
+	}
+    } else {
+	Tcl_DecrRefCount(objPtr);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Loop over the (otherVar, thisVar) pairs. If any of the thisVar is not a
+     * local variable, return an error so that the non-compiled command will
+     * be called at runtime.
+     */
+
+    for (; i<numWords; i+=2, otherTokenPtr = TokenAfter(localTokenPtr)) {
+	localTokenPtr = TokenAfter(otherTokenPtr);
+
+	CompileWord(envPtr, otherTokenPtr, interp, i);
+	localIndex = LocalScalarFromToken(localTokenPtr, envPtr);
+	if (localIndex < 0) {
+	    return TCL_ERROR;
+	}
+	TclEmitInstInt4(	INST_UPVAR, localIndex,		envPtr);
+    }
+
+    /*
+     * Pop the frame index, and set the result to empty
+     */
+
+    TclEmitOpcode(		INST_POP,			envPtr);
+    PushStringLiteral(envPtr, "");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileVariableCmd --
+ *
+ *	Procedure called to compile the "variable" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "variable" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileVariableCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr, *valueTokenPtr;
+    int localIndex, numWords, i;
+    DefineLineInformation;	/* TIP #280 */
+
+    numWords = parsePtr->numWords;
+    if (numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Bail out if not compiling a proc body
+     */
+
+    if (envPtr->procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Loop over the (var, value) pairs.
+     */
+
+    valueTokenPtr = parsePtr->tokenPtr;
+    for (i=1; i<numWords; i+=2) {
+	varTokenPtr = TokenAfter(valueTokenPtr);
+	valueTokenPtr = TokenAfter(varTokenPtr);
+
+	localIndex = IndexTailVarIfKnown(interp, varTokenPtr, envPtr);
+
+	if (localIndex < 0) {
+	    return TCL_ERROR;
+	}
+
+	/* TODO: Consider what value can pass throug the
+	 * IndexTailVarIfKnown() screen.  Full CompileWord()
+	 * likely does not apply here.  Push known value instead. */
+	CompileWord(envPtr, varTokenPtr, interp, i);
+	TclEmitInstInt4(	INST_VARIABLE, localIndex,	envPtr);
+
+	if (i+1 < numWords) {
+	    /*
+	     * A value has been given: set the variable, pop the value
+	     */
+
+	    CompileWord(envPtr, valueTokenPtr, interp, i+1);
+	    Emit14Inst(		INST_STORE_SCALAR, localIndex,	envPtr);
+	    TclEmitOpcode(	INST_POP,			envPtr);
+	}
+    }
+
+    /*
+     * Set the result to empty
+     */
+
+    PushStringLiteral(envPtr, "");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IndexTailVarIfKnown --
+ *
+ *	Procedure used in compiling [global] and [variable] commands. It
+ *	inspects the variable name described by varTokenPtr and, if the tail
+ *	is known at compile time, defines a corresponding local variable.
+ *
+ * Results:
+ *	Returns the variable's index in the table of compiled locals if the
+ *	tail is known at compile time, or -1 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+IndexTailVarIfKnown(
+    Tcl_Interp *interp,
+    Tcl_Token *varTokenPtr,	/* Token representing the variable name */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Obj *tailPtr;
+    const char *tailName, *p;
+    int len, n = varTokenPtr->numComponents;
+    Tcl_Token *lastTokenPtr;
+    int full, localIndex;
+
+    /*
+     * Determine if the tail is (a) known at compile time, and (b) not an
+     * array element. Should any of these fail, return an error so that the
+     * non-compiled command will be called at runtime.
+     *
+     * In order for the tail to be known at compile time, the last token in
+     * the word has to be constant and contain "::" if it is not the only one.
+     */
+
+    if (!EnvHasLVT(envPtr)) {
+	return -1;
+    }
+
+    TclNewObj(tailPtr);
+    if (TclWordKnownAtCompileTime(varTokenPtr, tailPtr)) {
+	full = 1;
+	lastTokenPtr = varTokenPtr;
+    } else {
+	full = 0;
+	lastTokenPtr = varTokenPtr + n;
+
+	if (lastTokenPtr->type != TCL_TOKEN_TEXT) {
+	    Tcl_DecrRefCount(tailPtr);
+	    return -1;
+	}
+	Tcl_SetStringObj(tailPtr, lastTokenPtr->start, lastTokenPtr->size);
+    }
+
+    tailName = TclGetStringFromObj(tailPtr, &len);
+
+    if (len) {
+	if (*(tailName+len-1) == ')') {
+	    /*
+	     * Possible array: bail out
+	     */
+
+	    Tcl_DecrRefCount(tailPtr);
+	    return -1;
+	}
+
+	/*
+	 * Get the tail: immediately after the last '::'
+	 */
+
+	for (p = tailName + len -1; p > tailName; p--) {
+	    if ((*p == ':') && (*(p-1) == ':')) {
+		p++;
+		break;
+	    }
+	}
+	if (!full && (p == tailName)) {
+	    /*
+	     * No :: in the last component.
+	     */
+
+	    Tcl_DecrRefCount(tailPtr);
+	    return -1;
+	}
+	len -= p - tailName;
+	tailName = p;
+    }
+
+    localIndex = TclFindCompiledLocal(tailName, len, 1, envPtr);
+    Tcl_DecrRefCount(tailPtr);
+    return localIndex;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclCompileObjectNextCmd, TclCompileObjectSelfCmd --
+ *
+ *	Compilations of the TclOO utility commands [next] and [self].
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclCompileObjectNextCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    int i;
+
+    if (parsePtr->numWords > 255) {
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    TclEmitInstInt1(	INST_TCLOO_NEXT, i,		envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileObjectNextToCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    int i;
+
+    if (parsePtr->numWords < 2 || parsePtr->numWords > 255) {
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    TclEmitInstInt1(	INST_TCLOO_NEXT_CLASS, i,	envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileObjectSelfCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * We only handle [self] and [self object] (which is the same operation).
+     * These are the only very common operations on [self] for which
+     * bytecoding is at all reasonable.
+     */
+
+    if (parsePtr->numWords == 1) {
+	goto compileSelfObject;
+    } else if (parsePtr->numWords == 2) {
+	Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr), *subcmd;
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || tokenPtr[1].size==0) {
+	    return TCL_ERROR;
+	}
+
+	subcmd = tokenPtr + 1;
+	if (strncmp(subcmd->start, "object", subcmd->size) == 0) {
+	    goto compileSelfObject;
+	} else if (strncmp(subcmd->start, "namespace", subcmd->size) == 0) {
+	    goto compileSelfNamespace;
+	}
+    }
+
+    /*
+     * Can't compile; handle with runtime call.
+     */
+
+    return TCL_ERROR;
+
+  compileSelfObject:
+
+    /*
+     * This delegates the entire problem to a single opcode.
+     */
+
+    TclEmitOpcode(		INST_TCLOO_SELF,		envPtr);
+    return TCL_OK;
+
+  compileSelfNamespace:
+
+    /*
+     * This is formally only correct with TclOO methods as they are currently
+     * implemented; it assumes that the current namespace is invariably when a
+     * TclOO context is present is the object's namespace, and that's
+     * technically only something that's a matter of current policy. But it
+     * avoids creating another opcode, so that's all good!
+     */
+
+    TclEmitOpcode(		INST_TCLOO_SELF,		envPtr);
+    TclEmitOpcode(		INST_POP,			envPtr);
+    TclEmitOpcode(		INST_NS_CURRENT,		envPtr);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCompCmdsSZ.c b/generic/tclCompCmdsSZ.c
new file mode 100644
index 0000000..25d10d6
--- /dev/null
+++ b/generic/tclCompCmdsSZ.c
@@ -0,0 +1,4485 @@
+/*
+ * tclCompCmdsSZ.c --
+ *
+ *	This file contains compilation procedures that compile various Tcl
+ *	commands (beginning with the letters 's' through 'z', except for
+ *	[upvar] and [variable]) into a sequence of instructions ("bytecodes").
+ *	Also includes the operator command compilers.
+ *
+ * Copyright (c) 1997-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2002 ActiveState Corporation.
+ * Copyright (c) 2004-2010 by Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include "tclStringTrim.h"
+
+/*
+ * Prototypes for procedures defined later in this file:
+ */
+
+static ClientData	DupJumptableInfo(ClientData clientData);
+static void		FreeJumptableInfo(ClientData clientData);
+static void		PrintJumptableInfo(ClientData clientData,
+			    Tcl_Obj *appendObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static void		DisassembleJumptableInfo(ClientData clientData,
+			    Tcl_Obj *dictObj, ByteCode *codePtr,
+			    unsigned int pcOffset);
+static int		CompileAssociativeBinaryOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, const char *identity,
+			    int instruction, CompileEnv *envPtr);
+static int		CompileComparisonOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, int instruction,
+			    CompileEnv *envPtr);
+static int		CompileStrictlyBinaryOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, int instruction,
+			    CompileEnv *envPtr);
+static int		CompileUnaryOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, int instruction,
+			    CompileEnv *envPtr);
+static void		IssueSwitchChainedTests(Tcl_Interp *interp,
+			    CompileEnv *envPtr, int mode, int noCase,
+			    int valueIndex, int numWords,
+			    Tcl_Token **bodyToken, int *bodyLines,
+			    int **bodyNext);
+static void		IssueSwitchJumpTable(Tcl_Interp *interp,
+			    CompileEnv *envPtr, int valueIndex,
+			    int numWords, Tcl_Token **bodyToken,
+			    int *bodyLines, int **bodyContLines);
+static int		IssueTryClausesInstructions(Tcl_Interp *interp,
+			    CompileEnv *envPtr, Tcl_Token *bodyToken,
+			    int numHandlers, int *matchCodes,
+			    Tcl_Obj **matchClauses, int *resultVarIndices,
+			    int *optionVarIndices, Tcl_Token **handlerTokens);
+static int		IssueTryClausesFinallyInstructions(Tcl_Interp *interp,
+			    CompileEnv *envPtr, Tcl_Token *bodyToken,
+			    int numHandlers, int *matchCodes,
+			    Tcl_Obj **matchClauses, int *resultVarIndices,
+			    int *optionVarIndices, Tcl_Token **handlerTokens,
+			    Tcl_Token *finallyToken);
+static int		IssueTryFinallyInstructions(Tcl_Interp *interp,
+			    CompileEnv *envPtr, Tcl_Token *bodyToken,
+			    Tcl_Token *finallyToken);
+
+/*
+ * The structures below define the AuxData types defined in this file.
+ */
+
+const AuxDataType tclJumptableInfoType = {
+    "JumptableInfo",		/* name */
+    DupJumptableInfo,		/* dupProc */
+    FreeJumptableInfo,		/* freeProc */
+    PrintJumptableInfo,		/* printProc */
+    DisassembleJumptableInfo	/* disassembleProc */
+};
+
+/*
+ * Shorthand macros for instruction issuing.
+ */
+
+#define OP(name)	TclEmitOpcode(INST_##name, envPtr)
+#define OP1(name,val)	TclEmitInstInt1(INST_##name,(val),envPtr)
+#define OP4(name,val)	TclEmitInstInt4(INST_##name,(val),envPtr)
+#define OP14(name,val1,val2) \
+    TclEmitInstInt1(INST_##name,(val1),envPtr);TclEmitInt4((val2),envPtr)
+#define OP44(name,val1,val2) \
+    TclEmitInstInt4(INST_##name,(val1),envPtr);TclEmitInt4((val2),envPtr)
+#define PUSH(str) \
+    PushStringLiteral(envPtr, str)
+#define JUMP4(name,var) \
+    (var) = CurrentOffset(envPtr);TclEmitInstInt4(INST_##name##4,0,envPtr)
+#define FIXJUMP4(var) \
+    TclStoreInt4AtPtr(CurrentOffset(envPtr)-(var),envPtr->codeStart+(var)+1)
+#define JUMP1(name,var) \
+    (var) = CurrentOffset(envPtr);TclEmitInstInt1(INST_##name##1,0,envPtr)
+#define FIXJUMP1(var) \
+    TclStoreInt1AtPtr(CurrentOffset(envPtr)-(var),envPtr->codeStart+(var)+1)
+#define LOAD(idx) \
+    if ((idx)<256) {OP1(LOAD_SCALAR1,(idx));} else {OP4(LOAD_SCALAR4,(idx));}
+#define STORE(idx) \
+    if ((idx)<256) {OP1(STORE_SCALAR1,(idx));} else {OP4(STORE_SCALAR4,(idx));}
+#define INVOKE(name) \
+    TclEmitInvoke(envPtr,INST_##name)
+
+#define INDEX_END	(-2)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetIndexFromToken --
+ *
+ *	Parse a token and get the encoded version of the index (as understood
+ *	by TEBC), assuming it is at all knowable at compile time. Only handles
+ *	indices that are integers or 'end' or 'end-integer'.
+ *
+ * Returns:
+ *	TCL_OK if parsing succeeded, and TCL_ERROR if it failed.
+ *
+ * Side effects:
+ *	Sets *index to the index value if successful.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+GetIndexFromToken(
+    Tcl_Token *tokenPtr,
+    int *index)
+{
+    Tcl_Obj *tmpObj = Tcl_NewObj();
+    int result, idx;
+
+    if (!TclWordKnownAtCompileTime(tokenPtr, tmpObj)) {
+	Tcl_DecrRefCount(tmpObj);
+	return TCL_ERROR;
+    }
+
+    result = TclGetIntFromObj(NULL, tmpObj, &idx);
+    if (result == TCL_OK) {
+	if (idx < 0) {
+	    result = TCL_ERROR;
+	}
+    } else {
+	result = TclGetIntForIndexM(NULL, tmpObj, INDEX_END, &idx);
+	if (result == TCL_OK && idx > INDEX_END) {
+	    result = TCL_ERROR;
+	}
+    }
+    Tcl_DecrRefCount(tmpObj);
+
+    if (result == TCL_OK) {
+	*index = idx;
+    }
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileSetCmd --
+ *
+ *	Procedure called to compile the "set" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "set" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileSetCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr, *valueTokenPtr;
+    int isAssignment, isScalar, localIndex, numWords;
+    DefineLineInformation;	/* TIP #280 */
+
+    numWords = parsePtr->numWords;
+    if ((numWords != 2) && (numWords != 3)) {
+	return TCL_ERROR;
+    }
+    isAssignment = (numWords == 3);
+
+    /*
+     * Decide if we can use a frame slot for the var/array name or if we need
+     * to emit code to compute and push the name at runtime. We use a frame
+     * slot (entry in the array of local vars) if we are compiling a procedure
+     * body and if the name is simple text that does not include namespace
+     * qualifiers.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+	    &localIndex, &isScalar, 1);
+
+    /*
+     * If we are doing an assignment, push the new value.
+     */
+
+    if (isAssignment) {
+	valueTokenPtr = TokenAfter(varTokenPtr);
+	CompileWord(envPtr, valueTokenPtr, interp, 2);
+    }
+
+    /*
+     * Emit instructions to set/get the variable.
+     */
+
+	if (isScalar) {
+	    if (localIndex < 0) {
+		TclEmitOpcode((isAssignment?
+			INST_STORE_STK : INST_LOAD_STK), envPtr);
+	    } else if (localIndex <= 255) {
+		TclEmitInstInt1((isAssignment?
+			INST_STORE_SCALAR1 : INST_LOAD_SCALAR1),
+			localIndex, envPtr);
+	    } else {
+		TclEmitInstInt4((isAssignment?
+			INST_STORE_SCALAR4 : INST_LOAD_SCALAR4),
+			localIndex, envPtr);
+	    }
+	} else {
+	    if (localIndex < 0) {
+		TclEmitOpcode((isAssignment?
+			INST_STORE_ARRAY_STK : INST_LOAD_ARRAY_STK), envPtr);
+	    } else if (localIndex <= 255) {
+		TclEmitInstInt1((isAssignment?
+			INST_STORE_ARRAY1 : INST_LOAD_ARRAY1),
+			localIndex, envPtr);
+	    } else {
+		TclEmitInstInt4((isAssignment?
+			INST_STORE_ARRAY4 : INST_LOAD_ARRAY4),
+			localIndex, envPtr);
+	    }
+	}
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileString*Cmd --
+ *
+ *	Procedures called to compile various subcommands of the "string"
+ *	command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "string" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileStringCatCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    int i, numWords = parsePtr->numWords, numArgs;
+    Tcl_Token *wordTokenPtr;
+    Tcl_Obj *obj, *folded;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* Trivial case, no arg */
+
+    if (numWords<2) {
+	PushStringLiteral(envPtr, "");
+	return TCL_OK;
+    }
+
+    /* General case: issue CONCAT1's (by chunks of 254 if needed), folding
+       contiguous constants along the way */
+
+    numArgs = 0;
+    folded = NULL;
+    wordTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    for (i = 1; i < numWords; i++) {
+	obj = Tcl_NewObj();
+	if (TclWordKnownAtCompileTime(wordTokenPtr, obj)) {
+	    if (folded) {
+		Tcl_AppendObjToObj(folded, obj);
+		Tcl_DecrRefCount(obj);
+	    } else {
+		folded = obj;
+	    }
+	} else {
+	    Tcl_DecrRefCount(obj);
+	    if (folded) {
+		int len;
+		const char *bytes = TclGetStringFromObj(folded, &len);
+
+		PushLiteral(envPtr, bytes, len);
+		Tcl_DecrRefCount(folded);
+		folded = NULL;
+		numArgs ++;
+	    }
+	    CompileWord(envPtr, wordTokenPtr, interp, i);
+	    numArgs ++;
+	    if (numArgs >= 254) { /* 254 to take care of the possible +1 of "folded" above */
+		TclEmitInstInt1(INST_STR_CONCAT1, numArgs, envPtr);
+		numArgs = 1;	/* concat pushes 1 obj, the result */
+	    }
+	}
+	wordTokenPtr = TokenAfter(wordTokenPtr);
+    }
+    if (folded) {
+	int len;
+	const char *bytes = TclGetStringFromObj(folded, &len);
+
+	PushLiteral(envPtr, bytes, len);
+	Tcl_DecrRefCount(folded);
+	folded = NULL;
+	numArgs ++;
+    }
+    if (numArgs > 1) {
+	TclEmitInstInt1(INST_STR_CONCAT1, numArgs, envPtr);
+    }
+
+    return TCL_OK;
+}
+
+int
+TclCompileStringCmpCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    /*
+     * We don't support any flags; the bytecode isn't that sophisticated.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the two operands onto the stack and then the test.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    tokenPtr = TokenAfter(tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 2);
+    TclEmitOpcode(INST_STR_CMP, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileStringEqualCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    /*
+     * We don't support any flags; the bytecode isn't that sophisticated.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the two operands onto the stack and then the test.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    tokenPtr = TokenAfter(tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 2);
+    TclEmitOpcode(INST_STR_EQ, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileStringFirstCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    /*
+     * We don't support any flags; the bytecode isn't that sophisticated.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the two operands onto the stack and then the test.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    tokenPtr = TokenAfter(tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 2);
+    OP(STR_FIND);
+    return TCL_OK;
+}
+
+int
+TclCompileStringLastCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    /*
+     * We don't support any flags; the bytecode isn't that sophisticated.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the two operands onto the stack and then the test.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    tokenPtr = TokenAfter(tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 2);
+    OP(STR_FIND_LAST);
+    return TCL_OK;
+}
+
+int
+TclCompileStringIndexCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Push the two operands onto the stack and then the index operation.
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    tokenPtr = TokenAfter(tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 2);
+    TclEmitOpcode(INST_STR_INDEX, envPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileStringIsCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    static const char *const isClasses[] = {
+	"alnum",	"alpha",	"ascii",	"control",
+	"boolean",	"digit",	"double",	"entier",
+	"false",	"graph",	"integer",	"list",
+	"lower",	"print",	"punct",	"space",
+	"true",		"upper",	"wideinteger",	"wordchar",
+	"xdigit",	NULL
+    };
+    enum isClasses {
+	STR_IS_ALNUM,	STR_IS_ALPHA,	STR_IS_ASCII,	STR_IS_CONTROL,
+	STR_IS_BOOL,	STR_IS_DIGIT,	STR_IS_DOUBLE,	STR_IS_ENTIER,
+	STR_IS_FALSE,	STR_IS_GRAPH,	STR_IS_INT,	STR_IS_LIST,
+	STR_IS_LOWER,	STR_IS_PRINT,	STR_IS_PUNCT,	STR_IS_SPACE,
+	STR_IS_TRUE,	STR_IS_UPPER,	STR_IS_WIDE,	STR_IS_WORD,
+	STR_IS_XDIGIT
+    };
+    int t, range, allowEmpty = 0, end;
+    InstStringClassType strClassType;
+    Tcl_Obj *isClass;
+
+    if (parsePtr->numWords < 3 || parsePtr->numWords > 6) {
+	return TCL_ERROR;
+    }
+    isClass = Tcl_NewObj();
+    if (!TclWordKnownAtCompileTime(tokenPtr, isClass)) {
+	Tcl_DecrRefCount(isClass);
+	return TCL_ERROR;
+    } else if (Tcl_GetIndexFromObj(interp, isClass, isClasses, "class", 0,
+	    &t) != TCL_OK) {
+	Tcl_DecrRefCount(isClass);
+	TclCompileSyntaxError(interp, envPtr);
+	return TCL_OK;
+    }
+    Tcl_DecrRefCount(isClass);
+
+#define GotLiteral(tokenPtr, word) \
+    ((tokenPtr)->type == TCL_TOKEN_SIMPLE_WORD &&			\
+     (tokenPtr)[1].size > 1 &&						\
+     (tokenPtr)[1].start[0] == word[0] &&				\
+     strncmp((tokenPtr)[1].start, (word), (tokenPtr)[1].size) == 0)
+
+    /*
+     * Cannot handle the -failindex option at all, and that's the only legal
+     * way to have more than 4 arguments.
+     */
+
+    if (parsePtr->numWords != 3 && parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(tokenPtr);
+    if (parsePtr->numWords == 3) {
+	allowEmpty = 1;
+    } else {
+	if (!GotLiteral(tokenPtr, "-strict")) {
+	    return TCL_ERROR;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+#undef GotLiteral
+
+    /*
+     * Compile the code. There are several main classes of check here.
+     *	1. Character classes
+     *	2. Booleans
+     *	3. Integers
+     *	4. Floats
+     *	5. Lists
+     */
+
+    CompileWord(envPtr, tokenPtr, interp, parsePtr->numWords-1);
+
+    switch ((enum isClasses) t) {
+    case STR_IS_ALNUM:
+	strClassType = STR_CLASS_ALNUM;
+	goto compileStrClass;
+    case STR_IS_ALPHA:
+	strClassType = STR_CLASS_ALPHA;
+	goto compileStrClass;
+    case STR_IS_ASCII:
+	strClassType = STR_CLASS_ASCII;
+	goto compileStrClass;
+    case STR_IS_CONTROL:
+	strClassType = STR_CLASS_CONTROL;
+	goto compileStrClass;
+    case STR_IS_DIGIT:
+	strClassType = STR_CLASS_DIGIT;
+	goto compileStrClass;
+    case STR_IS_GRAPH:
+	strClassType = STR_CLASS_GRAPH;
+	goto compileStrClass;
+    case STR_IS_LOWER:
+	strClassType = STR_CLASS_LOWER;
+	goto compileStrClass;
+    case STR_IS_PRINT:
+	strClassType = STR_CLASS_PRINT;
+	goto compileStrClass;
+    case STR_IS_PUNCT:
+	strClassType = STR_CLASS_PUNCT;
+	goto compileStrClass;
+    case STR_IS_SPACE:
+	strClassType = STR_CLASS_SPACE;
+	goto compileStrClass;
+    case STR_IS_UPPER:
+	strClassType = STR_CLASS_UPPER;
+	goto compileStrClass;
+    case STR_IS_WORD:
+	strClassType = STR_CLASS_WORD;
+	goto compileStrClass;
+    case STR_IS_XDIGIT:
+	strClassType = STR_CLASS_XDIGIT;
+    compileStrClass:
+	if (allowEmpty) {
+	    OP1(	STR_CLASS, strClassType);
+	} else {
+	    int over, over2;
+
+	    OP(		DUP);
+	    OP1(	STR_CLASS, strClassType);
+	    JUMP1(	JUMP_TRUE, over);
+	    OP(		POP);
+	    PUSH(	"0");
+	    JUMP1(	JUMP, over2);
+	    FIXJUMP1(over);
+	    PUSH(	"");
+	    OP(		STR_NEQ);
+	    FIXJUMP1(over2);
+	}
+	return TCL_OK;
+
+    case STR_IS_BOOL:
+    case STR_IS_FALSE:
+    case STR_IS_TRUE:
+	OP(		TRY_CVT_TO_BOOLEAN);
+	switch (t) {
+	    int over, over2;
+
+	case STR_IS_BOOL:
+	    if (allowEmpty) {
+		JUMP1(	JUMP_TRUE, over);
+		PUSH(	"");
+		OP(	STR_EQ);
+		JUMP1(	JUMP, over2);
+		FIXJUMP1(over);
+		OP(	POP);
+		PUSH(	"1");
+		FIXJUMP1(over2);
+	    } else {
+		OP4(	REVERSE, 2);
+		OP(	POP);
+	    }
+	    return TCL_OK;
+	case STR_IS_TRUE:
+	    JUMP1(	JUMP_TRUE, over);
+	    if (allowEmpty) {
+		PUSH(	"");
+		OP(	STR_EQ);
+	    } else {
+		OP(	POP);
+		PUSH(	"0");
+	    }
+	    FIXJUMP1(	over);
+	    OP(		LNOT);
+	    OP(		LNOT);
+	    return TCL_OK;
+	case STR_IS_FALSE:
+	    JUMP1(	JUMP_TRUE, over);
+	    if (allowEmpty) {
+		PUSH(	"");
+		OP(	STR_NEQ);
+	    } else {
+		OP(	POP);
+		PUSH(	"1");
+	    }
+	    FIXJUMP1(	over);
+	    OP(		LNOT);
+	    return TCL_OK;
+	}
+
+    case STR_IS_DOUBLE: {
+	int satisfied, isEmpty;
+
+	if (allowEmpty) {
+	    OP(		DUP);
+	    PUSH(	"");
+	    OP(		STR_EQ);
+	    JUMP1(	JUMP_TRUE, isEmpty);
+	    OP(		NUM_TYPE);
+	    JUMP1(	JUMP_TRUE, satisfied);
+	    PUSH(	"0");
+	    JUMP1(	JUMP, end);
+	    FIXJUMP1(	isEmpty);
+	    OP(		POP);
+	    FIXJUMP1(	satisfied);
+	} else {
+	    OP(		NUM_TYPE);
+	    JUMP1(	JUMP_TRUE, satisfied);
+	    PUSH(	"0");
+	    JUMP1(	JUMP, end);
+	    TclAdjustStackDepth(-1, envPtr);
+	    FIXJUMP1(	satisfied);
+	}
+	PUSH(		"1");
+	FIXJUMP1(	end);
+	return TCL_OK;
+    }
+
+    case STR_IS_INT:
+    case STR_IS_WIDE:
+    case STR_IS_ENTIER:
+	if (allowEmpty) {
+	    int testNumType;
+
+	    OP(		DUP);
+	    OP(		NUM_TYPE);
+	    OP(		DUP);
+	    JUMP1(	JUMP_TRUE, testNumType);
+	    OP(		POP);
+	    PUSH(	"");
+	    OP(		STR_EQ);
+	    JUMP1(	JUMP, end);
+	    TclAdjustStackDepth(1, envPtr);
+	    FIXJUMP1(	testNumType);
+	    OP4(	REVERSE, 2);
+	    OP(		POP);
+	} else {
+	    OP(		NUM_TYPE);
+	    OP(		DUP);
+	    JUMP1(	JUMP_FALSE, end);
+	}
+
+	switch (t) {
+	case STR_IS_INT:
+	    PUSH(	"1");
+	    OP(		EQ);
+	    break;
+	case STR_IS_WIDE:
+	    PUSH(	"2");
+	    OP(		LE);
+	    break;
+	case STR_IS_ENTIER:
+	    PUSH(	"3");
+	    OP(		LE);
+	    break;
+	}
+	FIXJUMP1(	end);
+	return TCL_OK;
+
+    case STR_IS_LIST:
+	range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+	OP4(		BEGIN_CATCH4, range);
+	ExceptionRangeStarts(envPtr, range);
+	OP(		DUP);
+	OP(		LIST_LENGTH);
+	OP(		POP);
+	ExceptionRangeEnds(envPtr, range);
+	ExceptionRangeTarget(envPtr, range, catchOffset);
+	OP(		POP);
+	OP(		PUSH_RETURN_CODE);
+	OP(		END_CATCH);
+	OP(		LNOT);
+	return TCL_OK;
+    }
+
+    return TclCompileBasicMin0ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileStringMatchCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    int i, length, exactMatch = 0, nocase = 0;
+    const char *str;
+
+    if (parsePtr->numWords < 3 || parsePtr->numWords > 4) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+    /*
+     * Check if we have a -nocase flag.
+     */
+
+    if (parsePtr->numWords == 4) {
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+	}
+	str = tokenPtr[1].start;
+	length = tokenPtr[1].size;
+	if ((length <= 1) || strncmp(str, "-nocase", (size_t) length)) {
+	    /*
+	     * Fail at run time, not in compilation.
+	     */
+
+	    return TclCompileBasic3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+	}
+	nocase = 1;
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    /*
+     * Push the strings to match against each other.
+     */
+
+    for (i = 0; i < 2; i++) {
+	if (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	    str = tokenPtr[1].start;
+	    length = tokenPtr[1].size;
+	    if (!nocase && (i == 0)) {
+		/*
+		 * Trivial matches can be done by 'string equal'. If -nocase
+		 * was specified, we can't do this because INST_STR_EQ has no
+		 * support for nocase.
+		 */
+
+		Tcl_Obj *copy = Tcl_NewStringObj(str, length);
+
+		Tcl_IncrRefCount(copy);
+		exactMatch = TclMatchIsTrivial(TclGetString(copy));
+		TclDecrRefCount(copy);
+	    }
+	    PushLiteral(envPtr, str, length);
+	} else {
+	    SetLineInformation(i+1+nocase);
+	    CompileTokens(envPtr, tokenPtr, interp);
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    /*
+     * Push the matcher.
+     */
+
+    if (exactMatch) {
+	TclEmitOpcode(INST_STR_EQ, envPtr);
+    } else {
+	TclEmitInstInt1(INST_STR_MATCH, nocase, envPtr);
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileStringLenCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+    Tcl_Obj *objPtr;
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    TclNewObj(objPtr);
+    if (TclWordKnownAtCompileTime(tokenPtr, objPtr)) {
+	/*
+	 * Here someone is asking for the length of a static string (or
+	 * something with backslashes). Just push the actual character (not
+	 * byte) length.
+	 */
+
+	char buf[TCL_INTEGER_SPACE];
+	int len = Tcl_GetCharLength(objPtr);
+
+	len = sprintf(buf, "%d", len);
+	PushLiteral(envPtr, buf, len);
+    } else {
+	SetLineInformation(1);
+	CompileTokens(envPtr, tokenPtr, interp);
+	TclEmitOpcode(INST_STR_LEN, envPtr);
+    }
+    TclDecrRefCount(objPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileStringMapCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *mapTokenPtr, *stringTokenPtr;
+    Tcl_Obj *mapObj, **objv;
+    char *bytes;
+    int len;
+
+    /*
+     * We only handle the case:
+     *
+     *    string map {foo bar} $thing
+     *
+     * That is, a literal two-element list (doesn't need to be brace-quoted,
+     * but does need to be compile-time knowable) and any old argument (the
+     * thing to map).
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+    mapTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    stringTokenPtr = TokenAfter(mapTokenPtr);
+    mapObj = Tcl_NewObj();
+    Tcl_IncrRefCount(mapObj);
+    if (!TclWordKnownAtCompileTime(mapTokenPtr, mapObj)) {
+	Tcl_DecrRefCount(mapObj);
+	return TclCompileBasic2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    } else if (Tcl_ListObjGetElements(NULL, mapObj, &len, &objv) != TCL_OK) {
+	Tcl_DecrRefCount(mapObj);
+	return TclCompileBasic2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    } else if (len != 2) {
+	Tcl_DecrRefCount(mapObj);
+	return TclCompileBasic2ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    /*
+     * Now issue the opcodes. Note that in the case that we know that the
+     * first word is an empty word, we don't issue the map at all. That is the
+     * correct semantics for mapping.
+     */
+
+    bytes = TclGetStringFromObj(objv[0], &len);
+    if (len == 0) {
+	CompileWord(envPtr, stringTokenPtr, interp, 2);
+    } else {
+	PushLiteral(envPtr, bytes, len);
+	bytes = TclGetStringFromObj(objv[1], &len);
+	PushLiteral(envPtr, bytes, len);
+	CompileWord(envPtr, stringTokenPtr, interp, 2);
+	OP(STR_MAP);
+    }
+    Tcl_DecrRefCount(mapObj);
+    return TCL_OK;
+}
+
+int
+TclCompileStringRangeCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *stringTokenPtr, *fromTokenPtr, *toTokenPtr;
+    int idx1, idx2;
+
+    if (parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+    stringTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    fromTokenPtr = TokenAfter(stringTokenPtr);
+    toTokenPtr = TokenAfter(fromTokenPtr);
+
+    /*
+     * Parse the two indices.
+     */
+
+    if (GetIndexFromToken(fromTokenPtr, &idx1) != TCL_OK) {
+	goto nonConstantIndices;
+    }
+    if (GetIndexFromToken(toTokenPtr, &idx2) != TCL_OK) {
+	goto nonConstantIndices;
+    }
+
+    /*
+     * Push the operand onto the stack and then the substring operation.
+     */
+
+    CompileWord(envPtr, stringTokenPtr,			interp, 1);
+    OP44(		STR_RANGE_IMM, idx1, idx2);
+    return TCL_OK;
+
+    /*
+     * Push the operands onto the stack and then the substring operation.
+     */
+
+  nonConstantIndices:
+    CompileWord(envPtr, stringTokenPtr,			interp, 1);
+    CompileWord(envPtr, fromTokenPtr,			interp, 2);
+    CompileWord(envPtr, toTokenPtr,			interp, 3);
+    OP(			STR_RANGE);
+    return TCL_OK;
+}
+
+int
+TclCompileStringReplaceCmd(
+    Tcl_Interp *interp,		/* Tcl interpreter for context. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the
+				 * command. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *tokenPtr, *valueTokenPtr, *replacementTokenPtr = NULL;
+    DefineLineInformation;	/* TIP #280 */
+    int idx1, idx2;
+
+    if (parsePtr->numWords < 4 || parsePtr->numWords > 5) {
+	return TCL_ERROR;
+    }
+    valueTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    if (parsePtr->numWords == 5) {
+	tokenPtr = TokenAfter(valueTokenPtr);
+	tokenPtr = TokenAfter(tokenPtr);
+	replacementTokenPtr = TokenAfter(tokenPtr);
+    }
+
+    /*
+     * Parse the indices. Will only compile special cases if both are
+     * constants and not an _integer_ less than zero (since we reserve
+     * negative indices here for end-relative indexing) or an end-based index
+     * greater than 'end' itself.
+     */
+
+    tokenPtr = TokenAfter(valueTokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx1) != TCL_OK) {
+	goto genericReplace;
+    }
+
+    tokenPtr = TokenAfter(tokenPtr);
+    if (GetIndexFromToken(tokenPtr, &idx2) != TCL_OK) {
+	goto genericReplace;
+    }
+
+    /*
+     * We handle these replacements specially: first character (where
+     * idx1=idx2=0) and last character (where idx1=idx2=INDEX_END). Anything
+     * else and the semantics get rather screwy.
+     */
+
+    if (idx1 == 0 && idx2 == 0) {
+	int notEq, end;
+
+	/*
+	 * Just working with the first character.
+	 */
+
+	CompileWord(envPtr, valueTokenPtr, interp, 1);
+	if (replacementTokenPtr == NULL) {
+	    /* Drop first */
+	    OP44(	STR_RANGE_IMM, 1, INDEX_END);
+	    return TCL_OK;
+	}
+	/* Replace first */
+	CompileWord(envPtr, replacementTokenPtr, interp, 4);
+	OP4(		OVER, 1);
+	PUSH(		"");
+	OP(		STR_EQ);
+	JUMP1(		JUMP_FALSE, notEq);
+	OP(		POP);
+	JUMP1(		JUMP, end);
+	FIXJUMP1(notEq);
+	TclAdjustStackDepth(1, envPtr);
+	OP4(		REVERSE, 2);
+	OP44(		STR_RANGE_IMM, 1, INDEX_END);
+	OP1(		STR_CONCAT1, 2);
+	FIXJUMP1(end);
+	return TCL_OK;
+
+    } else if (idx1 == INDEX_END && idx2 == INDEX_END) {
+	int notEq, end;
+
+	/*
+	 * Just working with the last character.
+	 */
+
+	CompileWord(envPtr, valueTokenPtr, interp, 1);
+	if (replacementTokenPtr == NULL) {
+	    /* Drop last */
+	    OP44(	STR_RANGE_IMM, 0, INDEX_END-1);
+	    return TCL_OK;
+	}
+	/* Replace last */
+	CompileWord(envPtr, replacementTokenPtr, interp, 4);
+	OP4(		OVER, 1);
+	PUSH(		"");
+	OP(		STR_EQ);
+	JUMP1(		JUMP_FALSE, notEq);
+	OP(		POP);
+	JUMP1(		JUMP, end);
+	FIXJUMP1(notEq);
+	TclAdjustStackDepth(1, envPtr);
+	OP4(		REVERSE, 2);
+	OP44(		STR_RANGE_IMM, 0, INDEX_END-1);
+	OP4(		REVERSE, 2);
+	OP1(		STR_CONCAT1, 2);
+	FIXJUMP1(end);
+	return TCL_OK;
+
+    } else {
+	/*
+	 * Need to process indices at runtime. This could be because the
+	 * indices are not constants, or because we need to resolve them to
+	 * absolute indices to work out if a replacement is going to happen.
+	 * In any case, to runtime it is.
+	 */
+
+    genericReplace:
+	CompileWord(envPtr, valueTokenPtr, interp, 1);
+	tokenPtr = TokenAfter(valueTokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 2);
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 3);
+	if (replacementTokenPtr != NULL) {
+	    CompileWord(envPtr, replacementTokenPtr, interp, 4);
+	} else {
+	    PUSH(	"");
+	}
+	OP(		STR_REPLACE);
+	return TCL_OK;
+    }
+}
+
+int
+TclCompileStringTrimLCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2 && parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    if (parsePtr->numWords == 3) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr,			interp, 2);
+    } else {
+	PushLiteral(envPtr, tclDefaultTrimSet, strlen(tclDefaultTrimSet));
+    }
+    OP(			STR_TRIM_LEFT);
+    return TCL_OK;
+}
+
+int
+TclCompileStringTrimRCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2 && parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    if (parsePtr->numWords == 3) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr,			interp, 2);
+    } else {
+	PushLiteral(envPtr, tclDefaultTrimSet, strlen(tclDefaultTrimSet));
+    }
+    OP(			STR_TRIM_RIGHT);
+    return TCL_OK;
+}
+
+int
+TclCompileStringTrimCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2 && parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    if (parsePtr->numWords == 3) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr,			interp, 2);
+    } else {
+	PushLiteral(envPtr, tclDefaultTrimSet, strlen(tclDefaultTrimSet));
+    }
+    OP(			STR_TRIM);
+    return TCL_OK;
+}
+
+int
+TclCompileStringToUpperCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2) {
+	return TclCompileBasic1To3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    OP(			STR_UPPER);
+    return TCL_OK;
+}
+
+int
+TclCompileStringToLowerCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2) {
+	return TclCompileBasic1To3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    OP(			STR_LOWER);
+    return TCL_OK;
+}
+
+int
+TclCompileStringToTitleCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr;
+
+    if (parsePtr->numWords != 2) {
+	return TclCompileBasic1To3ArgCmd(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr,			interp, 1);
+    OP(			STR_TITLE);
+    return TCL_OK;
+}
+
+/*
+ * Support definitions for the [string is] compilation.
+ */
+
+static int
+UniCharIsAscii(
+    int character)
+{
+    return (character >= 0) && (character < 0x80);
+}
+
+static int
+UniCharIsHexDigit(
+    int character)
+{
+    return (character >= 0) && (character < 0x80) && isxdigit(character);
+}
+
+StringClassDesc const tclStringClassTable[] = {
+    {"alnum",	Tcl_UniCharIsAlnum},
+    {"alpha",	Tcl_UniCharIsAlpha},
+    {"ascii",	UniCharIsAscii},
+    {"control", Tcl_UniCharIsControl},
+    {"digit",	Tcl_UniCharIsDigit},
+    {"graph",	Tcl_UniCharIsGraph},
+    {"lower",	Tcl_UniCharIsLower},
+    {"print",	Tcl_UniCharIsPrint},
+    {"punct",	Tcl_UniCharIsPunct},
+    {"space",	Tcl_UniCharIsSpace},
+    {"upper",	Tcl_UniCharIsUpper},
+    {"word",	Tcl_UniCharIsWordChar},
+    {"xdigit",	UniCharIsHexDigit},
+    {NULL,	NULL}
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileSubstCmd --
+ *
+ *	Procedure called to compile the "subst" command.
+ *
+ * Results:
+ *	Returns TCL_OK for successful compile, or TCL_ERROR to defer
+ *	evaluation to runtime (either when it is too complex to get the
+ *	semantics right, or when we know for sure that it is an error but need
+ *	the error to happen at the right time).
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "subst" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileSubstCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    int numArgs = parsePtr->numWords - 1;
+    int numOpts = numArgs - 1;
+    int objc, flags = TCL_SUBST_ALL;
+    Tcl_Obj **objv/*, *toSubst = NULL*/;
+    Tcl_Token *wordTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    int code = TCL_ERROR;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (numArgs == 0) {
+	return TCL_ERROR;
+    }
+
+    objv = TclStackAlloc(interp, /*numArgs*/ numOpts * sizeof(Tcl_Obj *));
+
+    for (objc = 0; objc < /*numArgs*/ numOpts; objc++) {
+	objv[objc] = Tcl_NewObj();
+	Tcl_IncrRefCount(objv[objc]);
+	if (!TclWordKnownAtCompileTime(wordTokenPtr, objv[objc])) {
+	    objc++;
+	    goto cleanup;
+	}
+	wordTokenPtr = TokenAfter(wordTokenPtr);
+    }
+
+/*
+    if (TclSubstOptions(NULL, numOpts, objv, &flags) == TCL_OK) {
+	toSubst = objv[numOpts];
+	Tcl_IncrRefCount(toSubst);
+    }
+*/
+
+    /* TODO: Figure out expansion to cover WordKnownAtCompileTime
+     *	The difficulty is that WKACT makes a copy, and if TclSubstParse
+     *	below parses the copy of the original source string, some deep
+     *	parts of the compile machinery get upset.  They want all pointers
+     *	stored in Tcl_Tokens to point back to the same original string.
+     */
+    if (wordTokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	code = TclSubstOptions(NULL, numOpts, objv, &flags);
+    }
+
+  cleanup:
+    while (--objc >= 0) {
+	TclDecrRefCount(objv[objc]);
+    }
+    TclStackFree(interp, objv);
+    if (/*toSubst == NULL*/ code != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    SetLineInformation(numArgs);
+    TclSubstCompile(interp, wordTokenPtr[1].start, wordTokenPtr[1].size,
+	    flags, mapPtr->loc[eclIndex].line[numArgs], envPtr);
+
+/*    TclDecrRefCount(toSubst);*/
+    return TCL_OK;
+}
+
+void
+TclSubstCompile(
+    Tcl_Interp *interp,
+    const char *bytes,
+    int numBytes,
+    int flags,
+    int line,
+    CompileEnv *envPtr)
+{
+    Tcl_Token *endTokenPtr, *tokenPtr;
+    int breakOffset = 0, count = 0, bline = line;
+    Tcl_Parse parse;
+    Tcl_InterpState state = NULL;
+
+    TclSubstParse(interp, bytes, numBytes, flags, &parse, &state);
+    if (state != NULL) {
+	Tcl_ResetResult(interp);
+    }
+
+    /*
+     * Tricky point! If the first token does not result in a *guaranteed* push
+     * of a Tcl_Obj on the stack, we must push an empty object. Otherwise it
+     * is possible to get to an INST_STR_CONCAT1 or INST_DONE without enough
+     * values on the stack, resulting in a crash. Thanks to Joe Mistachkin for
+     * identifying a script that could trigger this case.
+     */
+
+    tokenPtr = parse.tokenPtr;
+    if (tokenPtr->type != TCL_TOKEN_TEXT && tokenPtr->type != TCL_TOKEN_BS) {
+	PUSH("");
+	count++;
+    }
+
+    for (endTokenPtr = tokenPtr + parse.numTokens;
+	    tokenPtr < endTokenPtr; tokenPtr = TokenAfter(tokenPtr)) {
+	int length, literal, catchRange, breakJump;
+	char buf[TCL_UTF_MAX];
+	JumpFixup startFixup, okFixup, returnFixup, breakFixup;
+	JumpFixup continueFixup, otherFixup, endFixup;
+
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_TEXT:
+	    literal = TclRegisterLiteral(envPtr,
+		    tokenPtr->start, tokenPtr->size, 0);
+	    TclEmitPush(literal, envPtr);
+	    TclAdvanceLines(&bline, tokenPtr->start,
+		    tokenPtr->start + tokenPtr->size);
+	    count++;
+	    continue;
+	case TCL_TOKEN_BS:
+	    length = TclParseBackslash(tokenPtr->start, tokenPtr->size,
+		    NULL, buf);
+	    literal = TclRegisterLiteral(envPtr, buf, length, 0);
+	    TclEmitPush(literal, envPtr);
+	    count++;
+	    continue;
+	case TCL_TOKEN_VARIABLE:
+	    /*
+	     * Check for simple variable access; see if we can only generate
+	     * TCL_OK or TCL_ERROR from the substituted variable read; if so,
+	     * there is no need to generate elaborate exception-management
+	     * code. Note that the first component of TCL_TOKEN_VARIABLE is
+	     * always TCL_TOKEN_TEXT...
+	     */
+
+	    if (tokenPtr->numComponents > 1) {
+		int i, foundCommand = 0;
+
+		for (i=2 ; i<=tokenPtr->numComponents ; i++) {
+		    if (tokenPtr[i].type == TCL_TOKEN_COMMAND) {
+			foundCommand = 1;
+			break;
+		    }
+		}
+		if (foundCommand) {
+		    break;
+		}
+	    }
+
+	    envPtr->line = bline;
+	    TclCompileVarSubst(interp, tokenPtr, envPtr);
+	    bline = envPtr->line;
+	    count++;
+	    continue;
+	}
+
+	while (count > 255) {
+	    OP1(		STR_CONCAT1, 255);
+	    count -= 254;
+	}
+	if (count > 1) {
+	    OP1(		STR_CONCAT1, count);
+	    count = 1;
+	}
+
+	if (breakOffset == 0) {
+	    /* Jump to the start (jump over the jump to end) */
+	    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &startFixup);
+
+	    /* Jump to the end (all BREAKs land here) */
+	    breakOffset = CurrentOffset(envPtr);
+	    TclEmitInstInt4(INST_JUMP4, 0, envPtr);
+
+	    /* Start */
+	    if (TclFixupForwardJumpToHere(envPtr, &startFixup, 127)) {
+		Tcl_Panic("TclCompileSubstCmd: bad start jump distance %d",
+			(int) (CurrentOffset(envPtr) - startFixup.codeOffset));
+	    }
+	}
+
+	envPtr->line = bline;
+	catchRange = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+	OP4(	BEGIN_CATCH4, catchRange);
+	ExceptionRangeStarts(envPtr, catchRange);
+
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_COMMAND:
+	    TclCompileScript(interp, tokenPtr->start+1, tokenPtr->size-2,
+		    envPtr);
+	    count++;
+	    break;
+	case TCL_TOKEN_VARIABLE:
+	    TclCompileVarSubst(interp, tokenPtr, envPtr);
+	    count++;
+	    break;
+	default:
+	    Tcl_Panic("unexpected token type in TclCompileSubstCmd: %d",
+		    tokenPtr->type);
+	}
+
+	ExceptionRangeEnds(envPtr, catchRange);
+
+	/* Substitution produced TCL_OK */
+	OP(	END_CATCH);
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &okFixup);
+	TclAdjustStackDepth(-1, envPtr);
+
+	/* Exceptional return codes processed here */
+	ExceptionRangeTarget(envPtr, catchRange, catchOffset);
+	OP(	PUSH_RETURN_OPTIONS);
+	OP(	PUSH_RESULT);
+	OP(	PUSH_RETURN_CODE);
+	OP(	END_CATCH);
+	OP(	RETURN_CODE_BRANCH);
+
+	/* ERROR -> reraise it; NB: can't require BREAK/CONTINUE handling */
+	OP(	RETURN_STK);
+	OP(	NOP);
+
+	/* RETURN */
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &returnFixup);
+
+	/* BREAK */
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &breakFixup);
+
+	/* CONTINUE */
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &continueFixup);
+
+	/* OTHER */
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &otherFixup);
+
+	TclAdjustStackDepth(1, envPtr);
+	/* BREAK destination */
+	if (TclFixupForwardJumpToHere(envPtr, &breakFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad break jump distance %d",
+		    (int) (CurrentOffset(envPtr) - breakFixup.codeOffset));
+	}
+	OP(	POP);
+	OP(	POP);
+
+	breakJump = CurrentOffset(envPtr) - breakOffset;
+	if (breakJump > 127) {
+	    OP4(JUMP4, -breakJump);
+	} else {
+	    OP1(JUMP1, -breakJump);
+	}
+
+	TclAdjustStackDepth(2, envPtr);
+	/* CONTINUE destination */
+	if (TclFixupForwardJumpToHere(envPtr, &continueFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad continue jump distance %d",
+		    (int) (CurrentOffset(envPtr) - continueFixup.codeOffset));
+	}
+	OP(	POP);
+	OP(	POP);
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &endFixup);
+
+	TclAdjustStackDepth(2, envPtr);
+	/* RETURN + other destination */
+	if (TclFixupForwardJumpToHere(envPtr, &returnFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad return jump distance %d",
+		    (int) (CurrentOffset(envPtr) - returnFixup.codeOffset));
+	}
+	if (TclFixupForwardJumpToHere(envPtr, &otherFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad other jump distance %d",
+		    (int) (CurrentOffset(envPtr) - otherFixup.codeOffset));
+	}
+
+	/*
+	 * Pull the result to top of stack, discard options dict.
+	 */
+
+	OP4(	REVERSE, 2);
+	OP(	POP);
+
+	/* OK destination */
+	if (TclFixupForwardJumpToHere(envPtr, &okFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad ok jump distance %d",
+		    (int) (CurrentOffset(envPtr) - okFixup.codeOffset));
+	}
+	if (count > 1) {
+	    OP1(STR_CONCAT1, count);
+	    count = 1;
+	}
+
+	/* CONTINUE jump to here */
+	if (TclFixupForwardJumpToHere(envPtr, &endFixup, 127)) {
+	    Tcl_Panic("TclCompileSubstCmd: bad end jump distance %d",
+		    (int) (CurrentOffset(envPtr) - endFixup.codeOffset));
+	}
+	bline = envPtr->line;
+    }
+
+    while (count > 255) {
+	OP1(	STR_CONCAT1, 255);
+	count -= 254;
+    }
+    if (count > 1) {
+	OP1(	STR_CONCAT1, count);
+    }
+
+    Tcl_FreeParse(&parse);
+
+    if (state != NULL) {
+	Tcl_RestoreInterpState(interp, state);
+	TclCompileSyntaxError(interp, envPtr);
+	TclAdjustStackDepth(-1, envPtr);
+    }
+
+    /* Final target of the multi-jump from all BREAKs */
+    if (breakOffset > 0) {
+	TclUpdateInstInt4AtPc(INST_JUMP4, CurrentOffset(envPtr) - breakOffset,
+		envPtr->codeStart + breakOffset);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileSwitchCmd --
+ *
+ *	Procedure called to compile the "switch" command.
+ *
+ * Results:
+ *	Returns TCL_OK for successful compile, or TCL_ERROR to defer
+ *	evaluation to runtime (either when it is too complex to get the
+ *	semantics right, or when we know for sure that it is an error but need
+ *	the error to happen at the right time).
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "switch" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileSwitchCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;	/* Pointer to tokens in command. */
+    int numWords;		/* Number of words in command. */
+
+    Tcl_Token *valueTokenPtr;	/* Token for the value to switch on. */
+    enum {Switch_Exact, Switch_Glob, Switch_Regexp} mode;
+				/* What kind of switch are we doing? */
+
+    Tcl_Token *bodyTokenArray;	/* Array of real pattern list items. */
+    Tcl_Token **bodyToken;	/* Array of pointers to pattern list items. */
+    int *bodyLines;		/* Array of line numbers for body list
+				 * items. */
+    int **bodyContLines;	/* Array of continuation line info. */
+    int noCase;			/* Has the -nocase flag been given? */
+    int foundMode = 0;		/* Have we seen a mode flag yet? */
+    int i, valueIndex;
+    int result = TCL_ERROR;
+    DefineLineInformation;	/* TIP #280 */
+    int *clNext = envPtr->clNext;
+
+    /*
+     * Only handle the following versions:
+     *   switch         ?--? word {pattern body ...}
+     *   switch -exact  ?--? word {pattern body ...}
+     *   switch -glob   ?--? word {pattern body ...}
+     *   switch -regexp ?--? word {pattern body ...}
+     *   switch         --   word simpleWordPattern simpleWordBody ...
+     *   switch -exact  --   word simpleWordPattern simpleWordBody ...
+     *   switch -glob   --   word simpleWordPattern simpleWordBody ...
+     *   switch -regexp --   word simpleWordPattern simpleWordBody ...
+     * When the mode is -glob, can also handle a -nocase flag.
+     *
+     * First off, we don't care how the command's word was generated; we're
+     * compiling it anyway! So skip it...
+     */
+
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    valueIndex = 1;
+    numWords = parsePtr->numWords-1;
+
+    /*
+     * Check for options.
+     */
+
+    noCase = 0;
+    mode = Switch_Exact;
+    if (numWords == 2) {
+	/*
+	 * There's just the switch value and the bodies list. In that case, we
+	 * can skip all option parsing and move on to consider switch values
+	 * and the body list.
+	 */
+
+	goto finishedOptionParse;
+    }
+
+    /*
+     * There must be at least one option, --, because without that there is no
+     * way to statically avoid the problems you get from strings-to-be-matched
+     * that start with a - (the interpreted code falls apart if it encounters
+     * them, so we punt if we *might* encounter them as that is the easiest
+     * way of emulating the behaviour).
+     */
+
+    for (; numWords>=3 ; tokenPtr=TokenAfter(tokenPtr),numWords--) {
+	register unsigned size = tokenPtr[1].size;
+	register const char *chrs = tokenPtr[1].start;
+
+	/*
+	 * We only process literal options, and we assume that -e, -g and -n
+	 * are unique prefixes of -exact, -glob and -nocase respectively (true
+	 * at time of writing). Note that -exact and -glob may only be given
+	 * at most once or we bail out (error case).
+	 */
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || size < 2) {
+	    return TCL_ERROR;
+	}
+
+	if ((size <= 6) && !memcmp(chrs, "-exact", size)) {
+	    if (foundMode) {
+		return TCL_ERROR;
+	    }
+	    mode = Switch_Exact;
+	    foundMode = 1;
+	    valueIndex++;
+	    continue;
+	} else if ((size <= 5) && !memcmp(chrs, "-glob", size)) {
+	    if (foundMode) {
+		return TCL_ERROR;
+	    }
+	    mode = Switch_Glob;
+	    foundMode = 1;
+	    valueIndex++;
+	    continue;
+	} else if ((size <= 7) && !memcmp(chrs, "-regexp", size)) {
+	    if (foundMode) {
+		return TCL_ERROR;
+	    }
+	    mode = Switch_Regexp;
+	    foundMode = 1;
+	    valueIndex++;
+	    continue;
+	} else if ((size <= 7) && !memcmp(chrs, "-nocase", size)) {
+	    noCase = 1;
+	    valueIndex++;
+	    continue;
+	} else if ((size == 2) && !memcmp(chrs, "--", 2)) {
+	    valueIndex++;
+	    break;
+	}
+
+	/*
+	 * The switch command has many flags we cannot compile at all (e.g.
+	 * all the RE-related ones) which we must have encountered. Either
+	 * that or we have run off the end. The action here is the same: punt
+	 * to interpreted version.
+	 */
+
+	return TCL_ERROR;
+    }
+    if (numWords < 3) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(tokenPtr);
+    numWords--;
+    if (noCase && (mode == Switch_Exact)) {
+	/*
+	 * Can't compile this case; no opcode for case-insensitive equality!
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * The value to test against is going to always get pushed on the stack.
+     * But not yet; we need to verify that the rest of the command is
+     * compilable too.
+     */
+
+  finishedOptionParse:
+    valueTokenPtr = tokenPtr;
+    /* For valueIndex, see previous loop. */
+    tokenPtr = TokenAfter(tokenPtr);
+    numWords--;
+
+    /*
+     * Build an array of tokens for the matcher terms and script bodies. Note
+     * that in the case of the quoted bodies, this is tricky as we cannot use
+     * copies of the string from the input token for the generated tokens (it
+     * causes a crash during exception handling). When multiple tokens are
+     * available at this point, this is pretty easy.
+     */
+
+    if (numWords == 1) {
+	const char *bytes;
+	int maxLen, numBytes;
+	int bline;		/* TIP #280: line of the pattern/action list,
+				 * and start of list for when tracking the
+				 * location. This list comes immediately after
+				 * the value we switch on. */
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    return TCL_ERROR;
+	}
+	bytes = tokenPtr[1].start;
+	numBytes = tokenPtr[1].size;
+
+	/* Allocate enough space to work in. */
+	maxLen = TclMaxListLength(bytes, numBytes, NULL);
+	if (maxLen < 2)  {
+	    return TCL_ERROR;
+	}
+	bodyTokenArray = ckalloc(sizeof(Tcl_Token) * maxLen);
+	bodyToken = ckalloc(sizeof(Tcl_Token *) * maxLen);
+	bodyLines = ckalloc(sizeof(int) * maxLen);
+	bodyContLines = ckalloc(sizeof(int*) * maxLen);
+
+	bline = mapPtr->loc[eclIndex].line[valueIndex+1];
+	numWords = 0;
+
+	while (numBytes > 0) {
+	    const char *prevBytes = bytes;
+	    int literal;
+
+	    if (TCL_OK != TclFindElement(NULL, bytes, numBytes,
+		    &(bodyTokenArray[numWords].start), &bytes,
+		    &(bodyTokenArray[numWords].size), &literal) || !literal) {
+		goto abort;
+	    }
+
+	    bodyTokenArray[numWords].type = TCL_TOKEN_TEXT;
+	    bodyTokenArray[numWords].numComponents = 0;
+	    bodyToken[numWords] = bodyTokenArray + numWords;
+
+	    /*
+	     * TIP #280: Now determine the line the list element starts on
+	     * (there is no need to do it earlier, due to the possibility of
+	     * aborting, see above).
+	     */
+
+	    TclAdvanceLines(&bline, prevBytes, bodyTokenArray[numWords].start);
+	    TclAdvanceContinuations(&bline, &clNext,
+		    bodyTokenArray[numWords].start - envPtr->source);
+	    bodyLines[numWords] = bline;
+	    bodyContLines[numWords] = clNext;
+	    TclAdvanceLines(&bline, bodyTokenArray[numWords].start, bytes);
+	    TclAdvanceContinuations(&bline, &clNext, bytes - envPtr->source);
+
+	    numBytes -= (bytes - prevBytes);
+	    numWords++;
+	}
+	if (numWords % 2) {
+	abort:
+	    ckfree(bodyToken);
+	    ckfree(bodyTokenArray);
+	    ckfree(bodyLines);
+	    ckfree(bodyContLines);
+	    return TCL_ERROR;
+	}
+    } else if (numWords % 2 || numWords == 0) {
+	/*
+	 * Odd number of words (>1) available, or no words at all available.
+	 * Both are error cases, so punt and let the interpreted-version
+	 * generate the error message. Note that the second case probably
+	 * should get caught earlier, but it's easy to check here again anyway
+	 * because it'd cause a nasty crash otherwise.
+	 */
+
+	return TCL_ERROR;
+    } else {
+	/*
+	 * Multi-word definition of patterns & actions.
+	 */
+
+	bodyToken = ckalloc(sizeof(Tcl_Token *) * numWords);
+	bodyLines = ckalloc(sizeof(int) * numWords);
+	bodyContLines = ckalloc(sizeof(int*) * numWords);
+	bodyTokenArray = NULL;
+	for (i=0 ; i<numWords ; i++) {
+	    /*
+	     * We only handle the very simplest case. Anything more complex is
+	     * a good reason to go to the interpreted case anyway due to
+	     * traces, etc.
+	     */
+
+	    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+		goto freeTemporaries;
+	    }
+	    bodyToken[i] = tokenPtr+1;
+
+	    /*
+	     * TIP #280: Copy line information from regular cmd info.
+	     */
+
+	    bodyLines[i] = mapPtr->loc[eclIndex].line[valueIndex+1+i];
+	    bodyContLines[i] = mapPtr->loc[eclIndex].next[valueIndex+1+i];
+	    tokenPtr = TokenAfter(tokenPtr);
+	}
+    }
+
+    /*
+     * Fall back to interpreted if the last body is a continuation (it's
+     * illegal, but this makes the error happen at the right time).
+     */
+
+    if (bodyToken[numWords-1]->size == 1 &&
+	    bodyToken[numWords-1]->start[0] == '-') {
+	goto freeTemporaries;
+    }
+
+    /*
+     * Now we commit to generating code; the parsing stage per se is done.
+     * Check if we can generate a jump table, since if so that's faster than
+     * doing an explicit compare with each body. Note that we're definitely
+     * over-conservative with determining whether we can do the jump table,
+     * but it handles the most common case well enough.
+     */
+
+    /* Both methods push the value to match against onto the stack. */
+    CompileWord(envPtr, valueTokenPtr, interp, valueIndex);
+
+    if (mode == Switch_Exact) {
+	IssueSwitchJumpTable(interp, envPtr, valueIndex, numWords, bodyToken,
+		bodyLines, bodyContLines);
+    } else {
+	IssueSwitchChainedTests(interp, envPtr, mode, noCase, valueIndex,
+		numWords, bodyToken, bodyLines, bodyContLines);
+    }
+    result = TCL_OK;
+
+    /*
+     * Clean up all our temporary space and return.
+     */
+
+  freeTemporaries:
+    ckfree(bodyToken);
+    ckfree(bodyLines);
+    ckfree(bodyContLines);
+    if (bodyTokenArray != NULL) {
+	ckfree(bodyTokenArray);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IssueSwitchChainedTests --
+ *
+ *	Generate instructions for a [switch] command that is to be compiled
+ *	into a sequence of tests. This is the generic handle-everything mode
+ *	that inherently has performance that is (on average) linear in the
+ *	number of tests. It is the only mode that can handle -glob and -regexp
+ *	matches, or anything that is case-insensitive. It does not handle the
+ *	wild-and-wooly end of regexp matching (i.e., capture of match results)
+ *	so that's when we spill to the interpreted version.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+IssueSwitchChainedTests(
+    Tcl_Interp *interp,		/* Context for compiling script bodies. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int mode,			/* Exact, Glob or Regexp */
+    int noCase,			/* Case-insensitivity flag. */
+    int valueIndex,		/* The value to match against. */
+    int numBodyTokens,		/* Number of tokens describing things the
+				 * switch can match against and bodies to
+				 * execute when the match succeeds. */
+    Tcl_Token **bodyToken,	/* Array of pointers to pattern list items. */
+    int *bodyLines,		/* Array of line numbers for body list
+				 * items. */
+    int **bodyContLines)	/* Array of continuation line info. */
+{
+    enum {Switch_Exact, Switch_Glob, Switch_Regexp};
+    int foundDefault;		/* Flag to indicate whether a "default" clause
+				 * is present. */
+    JumpFixup *fixupArray;	/* Array of forward-jump fixup records. */
+    unsigned int *fixupTargetArray; /* Array of places for fixups to point at. */
+    int fixupCount;		/* Number of places to fix up. */
+    int contFixIndex;		/* Where the first of the jumps due to a group
+				 * of continuation bodies starts, or -1 if
+				 * there aren't any. */
+    int contFixCount;		/* Number of continuation bodies pointing to
+				 * the current (or next) real body. */
+    int nextArmFixupIndex;
+    int simple, exact;		/* For extracting the type of regexp. */
+    int i;
+
+    /*
+     * Generate a test for each arm.
+     */
+
+    contFixIndex = -1;
+    contFixCount = 0;
+    fixupArray = TclStackAlloc(interp, sizeof(JumpFixup) * numBodyTokens);
+    fixupTargetArray = TclStackAlloc(interp, sizeof(int) * numBodyTokens);
+    memset(fixupTargetArray, 0, numBodyTokens * sizeof(int));
+    fixupCount = 0;
+    foundDefault = 0;
+    for (i=0 ; i<numBodyTokens ; i+=2) {
+	nextArmFixupIndex = -1;
+	if (i!=numBodyTokens-2 || bodyToken[numBodyTokens-2]->size != 7 ||
+		memcmp(bodyToken[numBodyTokens-2]->start, "default", 7)) {
+	    /*
+	     * Generate the test for the arm.
+	     */
+
+	    switch (mode) {
+	    case Switch_Exact:
+		OP(	DUP);
+		TclCompileTokens(interp, bodyToken[i], 1,	envPtr);
+		OP(	STR_EQ);
+		break;
+	    case Switch_Glob:
+		TclCompileTokens(interp, bodyToken[i], 1,	envPtr);
+		OP4(	OVER, 1);
+		OP1(	STR_MATCH, noCase);
+		break;
+	    case Switch_Regexp:
+		simple = exact = 0;
+
+		/*
+		 * Keep in sync with TclCompileRegexpCmd.
+		 */
+
+		if (bodyToken[i]->type == TCL_TOKEN_TEXT) {
+		    Tcl_DString ds;
+
+		    if (bodyToken[i]->size == 0) {
+			/*
+			 * The semantics of regexps are that they always match
+			 * when the RE == "".
+			 */
+
+			PUSH("1");
+			break;
+		    }
+
+		    /*
+		     * Attempt to convert pattern to glob. If successful, push
+		     * the converted pattern.
+		     */
+
+		    if (TclReToGlob(NULL, bodyToken[i]->start,
+			    bodyToken[i]->size, &ds, &exact, NULL) == TCL_OK){
+			simple = 1;
+			PushLiteral(envPtr, Tcl_DStringValue(&ds),
+				Tcl_DStringLength(&ds));
+			Tcl_DStringFree(&ds);
+		    }
+		}
+		if (!simple) {
+		    TclCompileTokens(interp, bodyToken[i], 1, envPtr);
+		}
+
+		OP4(	OVER, 1);
+		if (!simple) {
+		    /*
+		     * Pass correct RE compile flags. We use only Int1
+		     * (8-bit), but that handles all the flags we want to
+		     * pass. Don't use TCL_REG_NOSUB as we may have backrefs
+		     * or capture vars.
+		     */
+
+		    int cflags = TCL_REG_ADVANCED
+			    | (noCase ? TCL_REG_NOCASE : 0);
+
+		    OP1(REGEXP, cflags);
+		} else if (exact && !noCase) {
+		    OP(	STR_EQ);
+		} else {
+		    OP1(STR_MATCH, noCase);
+		}
+		break;
+	    default:
+		Tcl_Panic("unknown switch mode: %d", mode);
+	    }
+
+	    /*
+	     * In a fall-through case, we will jump on _true_ to the place
+	     * where the body starts (generated later, with guarantee of this
+	     * ensured earlier; the final body is never a fall-through).
+	     */
+
+	    if (bodyToken[i+1]->size==1 && bodyToken[i+1]->start[0]=='-') {
+		if (contFixIndex == -1) {
+		    contFixIndex = fixupCount;
+		    contFixCount = 0;
+		}
+		TclEmitForwardJump(envPtr, TCL_TRUE_JUMP,
+			&fixupArray[contFixIndex+contFixCount]);
+		fixupCount++;
+		contFixCount++;
+		continue;
+	    }
+
+	    TclEmitForwardJump(envPtr, TCL_FALSE_JUMP,
+		    &fixupArray[fixupCount]);
+	    nextArmFixupIndex = fixupCount;
+	    fixupCount++;
+	} else {
+	    /*
+	     * Got a default clause; set a flag to inhibit the generation of
+	     * the jump after the body and the cleanup of the intermediate
+	     * value that we are switching against.
+	     *
+	     * Note that default clauses (which are always terminal clauses)
+	     * cannot be fall-through clauses as well, since the last clause
+	     * is never a fall-through clause (which we have already
+	     * verified).
+	     */
+
+	    foundDefault = 1;
+	}
+
+	/*
+	 * Generate the body for the arm. This is guaranteed not to be a
+	 * fall-through case, but it might have preceding fall-through cases,
+	 * so we must process those first.
+	 */
+
+	if (contFixIndex != -1) {
+	    int j;
+
+	    for (j=0 ; j<contFixCount ; j++) {
+		fixupTargetArray[contFixIndex+j] = CurrentOffset(envPtr);
+	    }
+	    contFixIndex = -1;
+	}
+
+	/*
+	 * Now do the actual compilation. Note that we do not use BODY()
+	 * because we may have synthesized the tokens in a non-standard
+	 * pattern.
+	 */
+
+	OP(	POP);
+	envPtr->line = bodyLines[i+1];		/* TIP #280 */
+	envPtr->clNext = bodyContLines[i+1];	/* TIP #280 */
+	TclCompileCmdWord(interp, bodyToken[i+1], 1, envPtr);
+
+	if (!foundDefault) {
+	    TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP,
+		    &fixupArray[fixupCount]);
+	    fixupCount++;
+	    fixupTargetArray[nextArmFixupIndex] = CurrentOffset(envPtr);
+	}
+    }
+
+    /*
+     * Discard the value we are matching against unless we've had a default
+     * clause (in which case it will already be gone due to the code at the
+     * start of processing an arm, guaranteed) and make the result of the
+     * command an empty string.
+     */
+
+    if (!foundDefault) {
+	OP(	POP);
+	PUSH("");
+    }
+
+    /*
+     * Do jump fixups for arms that were executed. First, fill in the jumps of
+     * all jumps that don't point elsewhere to point to here.
+     */
+
+    for (i=0 ; i<fixupCount ; i++) {
+	if (fixupTargetArray[i] == 0) {
+	    fixupTargetArray[i] = envPtr->codeNext-envPtr->codeStart;
+	}
+    }
+
+    /*
+     * Now scan backwards over all the jumps (all of which are forward jumps)
+     * doing each one. When we do one and there is a size changes, we must
+     * scan back over all the previous ones and see if they need adjusting
+     * before proceeding with further jump fixups (the interleaved nature of
+     * all the jumps makes this impossible to do without nested loops).
+     */
+
+    for (i=fixupCount-1 ; i>=0 ; i--) {
+	if (TclFixupForwardJump(envPtr, &fixupArray[i],
+		fixupTargetArray[i] - fixupArray[i].codeOffset, 127)) {
+	    int j;
+
+	    for (j=i-1 ; j>=0 ; j--) {
+		if (fixupTargetArray[j] > fixupArray[i].codeOffset) {
+		    fixupTargetArray[j] += 3;
+		}
+	    }
+	}
+    }
+    TclStackFree(interp, fixupTargetArray);
+    TclStackFree(interp, fixupArray);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IssueSwitchJumpTable --
+ *
+ *	Generate instructions for a [switch] command that is to be compiled
+ *	into a jump table. This only handles the case where case-sensitive,
+ *	exact matching is used, but this is actually the most common case in
+ *	real code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+IssueSwitchJumpTable(
+    Tcl_Interp *interp,		/* Context for compiling script bodies. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int valueIndex,		/* The value to match against. */
+    int numBodyTokens,		/* Number of tokens describing things the
+				 * switch can match against and bodies to
+				 * execute when the match succeeds. */
+    Tcl_Token **bodyToken,	/* Array of pointers to pattern list items. */
+    int *bodyLines,		/* Array of line numbers for body list
+				 * items. */
+    int **bodyContLines)	/* Array of continuation line info. */
+{
+    JumptableInfo *jtPtr;
+    int infoIndex, isNew, *finalFixups, numRealBodies = 0, jumpLocation;
+    int mustGenerate, foundDefault, jumpToDefault, i;
+    Tcl_DString buffer;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * Compile the switch by using a jump table, which is basically a
+     * hashtable that maps from literal values to match against to the offset
+     * (relative to the INST_JUMP_TABLE instruction) to jump to. The jump
+     * table itself is independent of any invokation of the bytecode, and as
+     * such is stored in an auxData block.
+     *
+     * Start by allocating the jump table itself, plus some workspace.
+     */
+
+    jtPtr = ckalloc(sizeof(JumptableInfo));
+    Tcl_InitHashTable(&jtPtr->hashTable, TCL_STRING_KEYS);
+    infoIndex = TclCreateAuxData(jtPtr, &tclJumptableInfoType, envPtr);
+    finalFixups = TclStackAlloc(interp, sizeof(int) * (numBodyTokens/2));
+    foundDefault = 0;
+    mustGenerate = 1;
+
+    /*
+     * Next, issue the instruction to do the jump, together with what we want
+     * to do if things do not work out (jump to either the default clause or
+     * the "default" default, which just sets the result to empty). Note that
+     * we will come back and rewrite the jump's offset parameter when we know
+     * what it should be, and that all jumps we issue are of the wide kind
+     * because that makes the code much easier to debug!
+     */
+
+    jumpLocation = CurrentOffset(envPtr);
+    OP4(	JUMP_TABLE, infoIndex);
+    jumpToDefault = CurrentOffset(envPtr);
+    OP4(	JUMP4, 0);
+
+    for (i=0 ; i<numBodyTokens ; i+=2) {
+	/*
+	 * For each arm, we must first work out what to do with the match
+	 * term.
+	 */
+
+	if (i!=numBodyTokens-2 || bodyToken[numBodyTokens-2]->size != 7 ||
+		memcmp(bodyToken[numBodyTokens-2]->start, "default", 7)) {
+	    /*
+	     * This is not a default clause, so insert the current location as
+	     * a target in the jump table (assuming it isn't already there,
+	     * which would indicate that this clause is probably masked by an
+	     * earlier one). Note that we use a Tcl_DString here simply
+	     * because the hash API does not let us specify the string length.
+	     */
+
+	    Tcl_DStringInit(&buffer);
+	    TclDStringAppendToken(&buffer, bodyToken[i]);
+	    hPtr = Tcl_CreateHashEntry(&jtPtr->hashTable,
+		    Tcl_DStringValue(&buffer), &isNew);
+	    if (isNew) {
+		/*
+		 * First time we've encountered this match clause, so it must
+		 * point to here.
+		 */
+
+		Tcl_SetHashValue(hPtr, CurrentOffset(envPtr) - jumpLocation);
+	    }
+	    Tcl_DStringFree(&buffer);
+	} else {
+	    /*
+	     * This is a default clause, so patch up the fallthrough from the
+	     * INST_JUMP_TABLE instruction to here.
+	     */
+
+	    foundDefault = 1;
+	    isNew = 1;
+	    TclStoreInt4AtPtr(CurrentOffset(envPtr)-jumpToDefault,
+		    envPtr->codeStart+jumpToDefault+1);
+	}
+
+	/*
+	 * Now, for each arm we must deal with the body of the clause.
+	 *
+	 * If this is a continuation body (never true of a final clause,
+	 * whether default or not) we're done because the next jump target
+	 * will also point here, so we advance to the next clause.
+	 */
+
+	if (bodyToken[i+1]->size == 1 && bodyToken[i+1]->start[0] == '-') {
+	    mustGenerate = 1;
+	    continue;
+	}
+
+	/*
+	 * Also skip this arm if its only match clause is masked. (We could
+	 * probably be more aggressive about this, but that would be much more
+	 * difficult to get right.)
+	 */
+
+	if (!isNew && !mustGenerate) {
+	    continue;
+	}
+	mustGenerate = 0;
+
+	/*
+	 * Compile the body of the arm.
+	 */
+
+	envPtr->line = bodyLines[i+1];		/* TIP #280 */
+	envPtr->clNext = bodyContLines[i+1];	/* TIP #280 */
+	TclCompileCmdWord(interp, bodyToken[i+1], 1, envPtr);
+
+	/*
+	 * Compile a jump in to the end of the command if this body is
+	 * anything other than a user-supplied default arm (to either skip
+	 * over the remaining bodies or the code that generates an empty
+	 * result).
+	 */
+
+	if (i+2 < numBodyTokens || !foundDefault) {
+	    finalFixups[numRealBodies++] = CurrentOffset(envPtr);
+
+	    /*
+	     * Easier by far to issue this jump as a fixed-width jump, since
+	     * otherwise we'd need to do a lot more (and more awkward)
+	     * rewriting when we fixed this all up.
+	     */
+
+	    OP4(	JUMP4, 0);
+	    TclAdjustStackDepth(-1, envPtr);
+	}
+    }
+
+    /*
+     * We're at the end. If we've not already done so through the processing
+     * of a user-supplied default clause, add in a "default" default clause
+     * now.
+     */
+
+    if (!foundDefault) {
+	TclStoreInt4AtPtr(CurrentOffset(envPtr)-jumpToDefault,
+		envPtr->codeStart+jumpToDefault+1);
+	PUSH("");
+    }
+
+    /*
+     * No more instructions to be issued; everything that needs to jump to the
+     * end of the command is fixed up at this point.
+     */
+
+    for (i=0 ; i<numRealBodies ; i++) {
+	TclStoreInt4AtPtr(CurrentOffset(envPtr)-finalFixups[i],
+		envPtr->codeStart+finalFixups[i]+1);
+    }
+
+    /*
+     * Clean up all our temporary space and return.
+     */
+
+    TclStackFree(interp, finalFixups);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupJumptableInfo, FreeJumptableInfo --
+ *
+ *	Functions to duplicate, release and print a jump-table created for use
+ *	with the INST_JUMP_TABLE instruction.
+ *
+ * Results:
+ *	DupJumptableInfo: a copy of the jump-table
+ *	FreeJumptableInfo: none
+ *	PrintJumptableInfo: none
+ *	DisassembleJumptableInfo: none
+ *
+ * Side effects:
+ *	DupJumptableInfo: allocates memory
+ *	FreeJumptableInfo: releases memory
+ *	PrintJumptableInfo: none
+ *	DisassembleJumptableInfo: none
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ClientData
+DupJumptableInfo(
+    ClientData clientData)
+{
+    JumptableInfo *jtPtr = clientData;
+    JumptableInfo *newJtPtr = ckalloc(sizeof(JumptableInfo));
+    Tcl_HashEntry *hPtr, *newHPtr;
+    Tcl_HashSearch search;
+    int isNew;
+
+    Tcl_InitHashTable(&newJtPtr->hashTable, TCL_STRING_KEYS);
+    hPtr = Tcl_FirstHashEntry(&jtPtr->hashTable, &search);
+    while (hPtr != NULL) {
+	newHPtr = Tcl_CreateHashEntry(&newJtPtr->hashTable,
+		Tcl_GetHashKey(&jtPtr->hashTable, hPtr), &isNew);
+	Tcl_SetHashValue(newHPtr, Tcl_GetHashValue(hPtr));
+    }
+    return newJtPtr;
+}
+
+static void
+FreeJumptableInfo(
+    ClientData clientData)
+{
+    JumptableInfo *jtPtr = clientData;
+
+    Tcl_DeleteHashTable(&jtPtr->hashTable);
+    ckfree(jtPtr);
+}
+
+static void
+PrintJumptableInfo(
+    ClientData clientData,
+    Tcl_Obj *appendObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register JumptableInfo *jtPtr = clientData;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+    const char *keyPtr;
+    int offset, i = 0;
+
+    hPtr = Tcl_FirstHashEntry(&jtPtr->hashTable, &search);
+    for (; hPtr ; hPtr = Tcl_NextHashEntry(&search)) {
+	keyPtr = Tcl_GetHashKey(&jtPtr->hashTable, hPtr);
+	offset = PTR2INT(Tcl_GetHashValue(hPtr));
+
+	if (i++) {
+	    Tcl_AppendToObj(appendObj, ", ", -1);
+	    if (i%4==0) {
+		Tcl_AppendToObj(appendObj, "\n\t\t", -1);
+	    }
+	}
+	Tcl_AppendPrintfToObj(appendObj, "\"%s\"->pc %d",
+		keyPtr, pcOffset + offset);
+    }
+}
+
+static void
+DisassembleJumptableInfo(
+    ClientData clientData,
+    Tcl_Obj *dictObj,
+    ByteCode *codePtr,
+    unsigned int pcOffset)
+{
+    register JumptableInfo *jtPtr = clientData;
+    Tcl_Obj *mapping = Tcl_NewObj();
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+    const char *keyPtr;
+    int offset;
+
+    hPtr = Tcl_FirstHashEntry(&jtPtr->hashTable, &search);
+    for (; hPtr ; hPtr = Tcl_NextHashEntry(&search)) {
+	keyPtr = Tcl_GetHashKey(&jtPtr->hashTable, hPtr);
+	offset = PTR2INT(Tcl_GetHashValue(hPtr));
+	Tcl_DictObjPut(NULL, mapping, Tcl_NewStringObj(keyPtr, -1),
+		Tcl_NewIntObj(offset));
+    }
+    Tcl_DictObjPut(NULL, dictObj, Tcl_NewStringObj("mapping", -1), mapping);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileTailcallCmd --
+ *
+ *	Procedure called to compile the "tailcall" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "tailcall" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileTailcallCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    int i;
+
+    if (parsePtr->numWords < 2 || parsePtr->numWords > 256
+	    || envPtr->procPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /* make room for the nsObjPtr */
+    /* TODO: Doesn't this have to be a known value? */
+    CompileWord(envPtr, tokenPtr, interp, 0);
+    for (i=1 ; i<parsePtr->numWords ; i++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, i);
+    }
+    TclEmitInstInt1(	INST_TAILCALL, parsePtr->numWords,	envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileThrowCmd --
+ *
+ *	Procedure called to compile the "throw" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "throw" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileThrowCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    int numWords = parsePtr->numWords;
+    Tcl_Token *codeToken, *msgToken;
+    Tcl_Obj *objPtr;
+    int codeKnown, codeIsList, codeIsValid, len;
+
+    if (numWords != 3) {
+	return TCL_ERROR;
+    }
+    codeToken = TokenAfter(parsePtr->tokenPtr);
+    msgToken = TokenAfter(codeToken);
+
+    TclNewObj(objPtr);
+    Tcl_IncrRefCount(objPtr);
+
+    codeKnown = TclWordKnownAtCompileTime(codeToken, objPtr);
+
+    /*
+     * First we must emit the code to substitute the arguments.  This
+     * must come first in case substitution raises errors.
+     */
+    if (!codeKnown) {
+	CompileWord(envPtr, codeToken, interp, 1);
+	PUSH(			"-errorcode");
+    }
+    CompileWord(envPtr, msgToken, interp, 2);
+
+    codeIsList = codeKnown && (TCL_OK ==
+	    Tcl_ListObjLength(interp, objPtr, &len));
+    codeIsValid = codeIsList && (len != 0);
+
+    if (codeIsValid) {
+	Tcl_Obj *errPtr, *dictPtr;
+
+	TclNewLiteralStringObj(errPtr, "-errorcode");
+	TclNewObj(dictPtr);
+	Tcl_DictObjPut(NULL, dictPtr, errPtr, objPtr);
+	TclEmitPush(TclAddLiteralObj(envPtr, dictPtr, NULL), envPtr);
+    }
+    TclDecrRefCount(objPtr);
+
+    /*
+     * Simpler bytecodes when we detect invalid arguments at compile time.
+     */
+    if (codeKnown && !codeIsValid) {
+	OP(			POP);
+	if (codeIsList) {
+	    /* Must be an empty list */
+	    goto issueErrorForEmptyCode;
+	}
+	TclCompileSyntaxError(interp, envPtr);
+	return TCL_OK;
+    }
+
+    if (!codeKnown) {
+	/*
+	 * Argument validity checking has to be done by bytecode at
+	 * run time.
+	 */
+	OP4(			REVERSE, 3);
+	OP(			DUP);
+	OP(			LIST_LENGTH);
+	OP1(			JUMP_FALSE1, 16);
+	OP4(			LIST, 2);
+	OP44(			RETURN_IMM, TCL_ERROR, 0);
+	TclAdjustStackDepth(2, envPtr);
+	OP(			POP);
+	OP(			POP);
+	OP(			POP);
+    issueErrorForEmptyCode:
+	PUSH(			"type must be non-empty list");
+	PUSH(			"-errorcode {TCL OPERATION THROW BADEXCEPTION}");
+    }
+    OP44(			RETURN_IMM, TCL_ERROR, 0);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileTryCmd --
+ *
+ *	Procedure called to compile the "try" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "try" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileTryCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    int numWords = parsePtr->numWords, numHandlers, result = TCL_ERROR;
+    Tcl_Token *bodyToken, *finallyToken, *tokenPtr;
+    Tcl_Token **handlerTokens = NULL;
+    Tcl_Obj **matchClauses = NULL;
+    int *matchCodes=NULL, *resultVarIndices=NULL, *optionVarIndices=NULL;
+    int i;
+
+    if (numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    bodyToken = TokenAfter(parsePtr->tokenPtr);
+
+    if (numWords == 2) {
+	/*
+	 * No handlers or finally; do nothing beyond evaluating the body.
+	 */
+
+	DefineLineInformation;	/* TIP #280 */
+	BODY(bodyToken, 1);
+	return TCL_OK;
+    }
+
+    numWords -= 2;
+    tokenPtr = TokenAfter(bodyToken);
+
+    /*
+     * Extract information about what handlers there are.
+     */
+
+    numHandlers = numWords >> 2;
+    numWords -= numHandlers * 4;
+    if (numHandlers > 0) {
+	handlerTokens = TclStackAlloc(interp, sizeof(Tcl_Token*)*numHandlers);
+	matchClauses = TclStackAlloc(interp, sizeof(Tcl_Obj *) * numHandlers);
+	memset(matchClauses, 0, sizeof(Tcl_Obj *) * numHandlers);
+	matchCodes = TclStackAlloc(interp, sizeof(int) * numHandlers);
+	resultVarIndices = TclStackAlloc(interp, sizeof(int) * numHandlers);
+	optionVarIndices = TclStackAlloc(interp, sizeof(int) * numHandlers);
+
+	for (i=0 ; i<numHandlers ; i++) {
+	    Tcl_Obj *tmpObj, **objv;
+	    int objc;
+
+	    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+		goto failedToCompile;
+	    }
+	    if (tokenPtr[1].size == 4
+		    && !strncmp(tokenPtr[1].start, "trap", 4)) {
+		/*
+		 * Parse the list of errorCode words to match against.
+		 */
+
+		matchCodes[i] = TCL_ERROR;
+		tokenPtr = TokenAfter(tokenPtr);
+		TclNewObj(tmpObj);
+		Tcl_IncrRefCount(tmpObj);
+		if (!TclWordKnownAtCompileTime(tokenPtr, tmpObj)
+			|| Tcl_ListObjLength(NULL, tmpObj, &objc) != TCL_OK
+			|| (objc == 0)) {
+		    TclDecrRefCount(tmpObj);
+		    goto failedToCompile;
+		}
+		Tcl_ListObjReplace(NULL, tmpObj, 0, 0, 0, NULL);
+		matchClauses[i] = tmpObj;
+	    } else if (tokenPtr[1].size == 2
+		    && !strncmp(tokenPtr[1].start, "on", 2)) {
+		int code;
+
+		/*
+		 * Parse the result code to look for.
+		 */
+
+		tokenPtr = TokenAfter(tokenPtr);
+		TclNewObj(tmpObj);
+		Tcl_IncrRefCount(tmpObj);
+		if (!TclWordKnownAtCompileTime(tokenPtr, tmpObj)) {
+		    TclDecrRefCount(tmpObj);
+		    goto failedToCompile;
+		}
+		if (TCL_ERROR == TclGetCompletionCodeFromObj(NULL, tmpObj, &code)) {
+		    TclDecrRefCount(tmpObj);
+		    goto failedToCompile;
+		}
+		matchCodes[i] = code;
+		TclDecrRefCount(tmpObj);
+	    } else {
+		goto failedToCompile;
+	    }
+
+	    /*
+	     * Parse the variable binding.
+	     */
+
+	    tokenPtr = TokenAfter(tokenPtr);
+	    TclNewObj(tmpObj);
+	    Tcl_IncrRefCount(tmpObj);
+	    if (!TclWordKnownAtCompileTime(tokenPtr, tmpObj)) {
+		TclDecrRefCount(tmpObj);
+		goto failedToCompile;
+	    }
+	    if (Tcl_ListObjGetElements(NULL, tmpObj, &objc, &objv) != TCL_OK
+		    || (objc > 2)) {
+		TclDecrRefCount(tmpObj);
+		goto failedToCompile;
+	    }
+	    if (objc > 0) {
+		int len;
+		const char *varname = TclGetStringFromObj(objv[0], &len);
+
+		resultVarIndices[i] = LocalScalar(varname, len, envPtr);
+		if (resultVarIndices[i] < 0) {
+		    TclDecrRefCount(tmpObj);
+		    goto failedToCompile;
+		}
+	    } else {
+		resultVarIndices[i] = -1;
+	    }
+	    if (objc == 2) {
+		int len;
+		const char *varname = TclGetStringFromObj(objv[1], &len);
+
+		optionVarIndices[i] = LocalScalar(varname, len, envPtr);
+		if (optionVarIndices[i] < 0) {
+		    TclDecrRefCount(tmpObj);
+		    goto failedToCompile;
+		}
+	    } else {
+		optionVarIndices[i] = -1;
+	    }
+	    TclDecrRefCount(tmpObj);
+
+	    /*
+	     * Extract the body for this handler.
+	     */
+
+	    tokenPtr = TokenAfter(tokenPtr);
+	    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+		goto failedToCompile;
+	    }
+	    if (tokenPtr[1].size == 1 && tokenPtr[1].start[0] == '-') {
+		handlerTokens[i] = NULL;
+	    } else {
+		handlerTokens[i] = tokenPtr;
+	    }
+
+	    tokenPtr = TokenAfter(tokenPtr);
+	}
+
+	if (handlerTokens[numHandlers-1] == NULL) {
+	    goto failedToCompile;
+	}
+    }
+
+    /*
+     * Parse the finally clause
+     */
+
+    if (numWords == 0) {
+	finallyToken = NULL;
+    } else if (numWords == 2) {
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD || tokenPtr[1].size != 7
+		|| strncmp(tokenPtr[1].start, "finally", 7)) {
+	    goto failedToCompile;
+	}
+	finallyToken = TokenAfter(tokenPtr);
+    } else {
+	goto failedToCompile;
+    }
+
+    /*
+     * Issue the bytecode.
+     */
+
+    if (!finallyToken) {
+	result = IssueTryClausesInstructions(interp, envPtr, bodyToken,
+		numHandlers, matchCodes, matchClauses, resultVarIndices,
+		optionVarIndices, handlerTokens);
+    } else if (numHandlers == 0) {
+	result = IssueTryFinallyInstructions(interp, envPtr, bodyToken,
+		finallyToken);
+    } else {
+	result = IssueTryClausesFinallyInstructions(interp, envPtr, bodyToken,
+		numHandlers, matchCodes, matchClauses, resultVarIndices,
+		optionVarIndices, handlerTokens, finallyToken);
+    }
+
+    /*
+     * Delete any temporary state and finish off.
+     */
+
+  failedToCompile:
+    if (numHandlers > 0) {
+	for (i=0 ; i<numHandlers ; i++) {
+	    if (matchClauses[i]) {
+		TclDecrRefCount(matchClauses[i]);
+	    }
+	}
+	TclStackFree(interp, optionVarIndices);
+	TclStackFree(interp, resultVarIndices);
+	TclStackFree(interp, matchCodes);
+	TclStackFree(interp, matchClauses);
+	TclStackFree(interp, handlerTokens);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IssueTryClausesInstructions, IssueTryClausesFinallyInstructions,
+ * IssueTryFinallyInstructions --
+ *
+ *	The code generators for [try]. Split from the parsing engine for
+ *	reasons of developer sanity, and also split between no-finally,
+ *	just-finally and with-finally cases because so many of the details of
+ *	generation vary between the three.
+ *
+ *	The macros below make the instruction issuing easier to follow.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+IssueTryClausesInstructions(
+    Tcl_Interp *interp,
+    CompileEnv *envPtr,
+    Tcl_Token *bodyToken,
+    int numHandlers,
+    int *matchCodes,
+    Tcl_Obj **matchClauses,
+    int *resultVars,
+    int *optionVars,
+    Tcl_Token **handlerTokens)
+{
+    DefineLineInformation;	/* TIP #280 */
+    int range, resultVar, optionsVar;
+    int i, j, len, forwardsNeedFixing = 0, trapZero = 0, afterBody = 0;
+    int *addrsToFix, *forwardsToFix, notCodeJumpSource, notECJumpSource;
+    int *noError;
+    char buf[TCL_INTEGER_SPACE];
+
+    resultVar = AnonymousLocal(envPtr);
+    optionsVar = AnonymousLocal(envPtr);
+    if (resultVar < 0 || optionsVar < 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check if we're supposed to trap a normal TCL_OK completion of the body.
+     * If not, we can handle that case much more efficiently.
+     */
+
+    for (i=0 ; i<numHandlers ; i++) {
+	if (matchCodes[i] == 0) {
+	    trapZero = 1;
+	    break;
+	}
+    }
+
+    /*
+     * Compile the body, trapping any error in it so that we can trap on it
+     * and/or run a finally clause. Note that there must be at least one
+     * on/trap clause; when none is present, this whole function is not called
+     * (and it's never called when there's a finally clause).
+     */
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    OP4(				BEGIN_CATCH4, range);
+    ExceptionRangeStarts(envPtr, range);
+    BODY(				bodyToken, 1);
+    ExceptionRangeEnds(envPtr, range);
+    if (!trapZero) {
+	OP(				END_CATCH);
+	JUMP4(				JUMP, afterBody);
+	TclAdjustStackDepth(-1, envPtr);
+    } else {
+	PUSH(				"0");
+	OP4(				REVERSE, 2);
+	OP1(				JUMP1, 4);
+	TclAdjustStackDepth(-2, envPtr);
+    }
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    OP(					PUSH_RETURN_CODE);
+    OP(					PUSH_RESULT);
+    OP(					PUSH_RETURN_OPTIONS);
+    OP(					END_CATCH);
+    STORE(				optionsVar);
+    OP(					POP);
+    STORE(				resultVar);
+    OP(					POP);
+
+    /*
+     * Now we handle all the registered 'on' and 'trap' handlers in order.
+     * For us to be here, there must be at least one handler.
+     *
+     * Slight overallocation, but reduces size of this function.
+     */
+
+    addrsToFix = TclStackAlloc(interp, sizeof(int)*numHandlers);
+    forwardsToFix = TclStackAlloc(interp, sizeof(int)*numHandlers);
+    noError = TclStackAlloc(interp, sizeof(int)*numHandlers);
+
+    for (i=0 ; i<numHandlers ; i++) {
+	noError[i] = -1;
+	sprintf(buf, "%d", matchCodes[i]);
+	OP(				DUP);
+	PushLiteral(envPtr, buf, strlen(buf));
+	OP(				EQ);
+	JUMP4(				JUMP_FALSE, notCodeJumpSource);
+	if (matchClauses[i]) {
+	    const char *p;
+	    Tcl_ListObjLength(NULL, matchClauses[i], &len);
+
+	    /*
+	     * Match the errorcode according to try/trap rules.
+	     */
+
+	    LOAD(			optionsVar);
+	    PUSH(			"-errorcode");
+	    OP4(			DICT_GET, 1);
+	    TclAdjustStackDepth(-1, envPtr);
+	    OP44(			LIST_RANGE_IMM, 0, len-1);
+	    p = TclGetStringFromObj(matchClauses[i], &len);
+	    PushLiteral(envPtr, p, len);
+	    OP(				STR_EQ);
+	    JUMP4(			JUMP_FALSE, notECJumpSource);
+	} else {
+	    notECJumpSource = -1; /* LINT */
+	}
+	OP(				POP);
+
+	/*
+	 * There is no finally clause, so we can avoid wrapping a catch
+	 * context around the handler. That simplifies what instructions need
+	 * to be issued a lot since we can let errors just fall through.
+	 */
+
+	if (resultVars[i] >= 0) {
+	    LOAD(			resultVar);
+	    STORE(			resultVars[i]);
+	    OP(				POP);
+	    if (optionVars[i] >= 0) {
+		LOAD(			optionsVar);
+		STORE(			optionVars[i]);
+		OP(			POP);
+	    }
+	}
+	if (!handlerTokens[i]) {
+	    forwardsNeedFixing = 1;
+	    JUMP4(			JUMP, forwardsToFix[i]);
+	    TclAdjustStackDepth(1, envPtr);
+	} else {
+	    int dontChangeOptions;
+
+	    forwardsToFix[i] = -1;
+	    if (forwardsNeedFixing) {
+		forwardsNeedFixing = 0;
+		for (j=0 ; j<i ; j++) {
+		    if (forwardsToFix[j] == -1) {
+			continue;
+		    }
+		    FIXJUMP4(forwardsToFix[j]);
+		    forwardsToFix[j] = -1;
+		}
+	    }
+	    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+	    OP4(			BEGIN_CATCH4, range);
+	    ExceptionRangeStarts(envPtr, range);
+	    BODY(			handlerTokens[i], 5+i*4);
+	    ExceptionRangeEnds(envPtr, range);
+	    OP(				END_CATCH);
+	    JUMP4(			JUMP, noError[i]);
+	    ExceptionRangeTarget(envPtr, range, catchOffset);
+	    TclAdjustStackDepth(-1, envPtr);
+	    OP(				PUSH_RESULT);
+	    OP(				PUSH_RETURN_OPTIONS);
+	    OP(				PUSH_RETURN_CODE);
+	    OP(				END_CATCH);
+	    PUSH(			"1");
+	    OP(				EQ);
+	    JUMP1(			JUMP_FALSE, dontChangeOptions);
+	    LOAD(			optionsVar);
+	    OP4(			REVERSE, 2);
+	    STORE(			optionsVar);
+	    OP(				POP);
+	    PUSH(			"-during");
+	    OP4(			REVERSE, 2);
+	    OP44(			DICT_SET, 1, optionsVar);
+	    TclAdjustStackDepth(-1, envPtr);
+	    FIXJUMP1(		dontChangeOptions);
+	    OP4(			REVERSE, 2);
+	    INVOKE(			RETURN_STK);
+	}
+
+	JUMP4(				JUMP, addrsToFix[i]);
+	if (matchClauses[i]) {
+	    FIXJUMP4(	notECJumpSource);
+	}
+	FIXJUMP4(	notCodeJumpSource);
+    }
+
+    /*
+     * Drop the result code since it didn't match any clause, and reissue the
+     * exception. Note also that INST_RETURN_STK can proceed to the next
+     * instruction.
+     */
+
+    OP(					POP);
+    LOAD(				optionsVar);
+    LOAD(				resultVar);
+    INVOKE(				RETURN_STK);
+
+    /*
+     * Fix all the jumps from taken clauses to here (which is the end of the
+     * [try]).
+     */
+
+    if (!trapZero) {
+	FIXJUMP4(afterBody);
+    }
+    for (i=0 ; i<numHandlers ; i++) {
+	FIXJUMP4(addrsToFix[i]);
+	if (noError[i] != -1) {
+	    FIXJUMP4(noError[i]);
+	}
+    }
+    TclStackFree(interp, noError);
+    TclStackFree(interp, forwardsToFix);
+    TclStackFree(interp, addrsToFix);
+    return TCL_OK;
+}
+
+static int
+IssueTryClausesFinallyInstructions(
+    Tcl_Interp *interp,
+    CompileEnv *envPtr,
+    Tcl_Token *bodyToken,
+    int numHandlers,
+    int *matchCodes,
+    Tcl_Obj **matchClauses,
+    int *resultVars,
+    int *optionVars,
+    Tcl_Token **handlerTokens,
+    Tcl_Token *finallyToken)	/* Not NULL */
+{
+    DefineLineInformation;	/* TIP #280 */
+    int range, resultVar, optionsVar, i, j, len, forwardsNeedFixing = 0;
+    int trapZero = 0, afterBody = 0, finalOK, finalError, noFinalError;
+    int *addrsToFix, *forwardsToFix, notCodeJumpSource, notECJumpSource;
+    char buf[TCL_INTEGER_SPACE];
+
+    resultVar = AnonymousLocal(envPtr);
+    optionsVar = AnonymousLocal(envPtr);
+    if (resultVar < 0 || optionsVar < 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check if we're supposed to trap a normal TCL_OK completion of the body.
+     * If not, we can handle that case much more efficiently.
+     */
+
+    for (i=0 ; i<numHandlers ; i++) {
+	if (matchCodes[i] == 0) {
+	    trapZero = 1;
+	    break;
+	}
+    }
+
+    /*
+     * Compile the body, trapping any error in it so that we can trap on it
+     * (if any trap matches) and run a finally clause.
+     */
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    OP4(				BEGIN_CATCH4, range);
+    ExceptionRangeStarts(envPtr, range);
+    BODY(				bodyToken, 1);
+    ExceptionRangeEnds(envPtr, range);
+    if (!trapZero) {
+	OP(				END_CATCH);
+	STORE(				resultVar);
+	OP(				POP);
+	PUSH(				"-level 0 -code 0");
+	STORE(				optionsVar);
+	OP(				POP);
+	JUMP4(				JUMP, afterBody);
+    } else {
+	PUSH(				"0");
+	OP4(				REVERSE, 2);
+	OP1(				JUMP1, 4);
+	TclAdjustStackDepth(-2, envPtr);
+    }
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    OP(					PUSH_RETURN_CODE);
+    OP(					PUSH_RESULT);
+    OP(					PUSH_RETURN_OPTIONS);
+    OP(					END_CATCH);
+    STORE(				optionsVar);
+    OP(					POP);
+    STORE(				resultVar);
+    OP(					POP);
+
+    /*
+     * Now we handle all the registered 'on' and 'trap' handlers in order.
+     *
+     * Slight overallocation, but reduces size of this function.
+     */
+
+    addrsToFix = TclStackAlloc(interp, sizeof(int)*numHandlers);
+    forwardsToFix = TclStackAlloc(interp, sizeof(int)*numHandlers);
+
+    for (i=0 ; i<numHandlers ; i++) {
+	int noTrapError, trapError;
+	const char *p;
+
+	sprintf(buf, "%d", matchCodes[i]);
+	OP(				DUP);
+	PushLiteral(envPtr, buf, strlen(buf));
+	OP(				EQ);
+	JUMP4(				JUMP_FALSE, notCodeJumpSource);
+	if (matchClauses[i]) {
+	    Tcl_ListObjLength(NULL, matchClauses[i], &len);
+
+	    /*
+	     * Match the errorcode according to try/trap rules.
+	     */
+
+	    LOAD(			optionsVar);
+	    PUSH(			"-errorcode");
+	    OP4(			DICT_GET, 1);
+	    TclAdjustStackDepth(-1, envPtr);
+	    OP44(			LIST_RANGE_IMM, 0, len-1);
+	    p = TclGetStringFromObj(matchClauses[i], &len);
+	    PushLiteral(envPtr, p, len);
+	    OP(				STR_EQ);
+	    JUMP4(			JUMP_FALSE, notECJumpSource);
+	} else {
+	    notECJumpSource = -1; /* LINT */
+	}
+	OP(				POP);
+
+	/*
+	 * There is a finally clause, so we need a fairly complex sequence of
+	 * instructions to deal with an on/trap handler because we must call
+	 * the finally handler *and* we need to substitute the result from a
+	 * failed trap for the result from the main script.
+	 */
+
+	if (resultVars[i] >= 0 || handlerTokens[i]) {
+	    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+	    OP4(			BEGIN_CATCH4, range);
+	    ExceptionRangeStarts(envPtr, range);
+	}
+	if (resultVars[i] >= 0) {
+	    LOAD(			resultVar);
+	    STORE(			resultVars[i]);
+	    OP(				POP);
+	    if (optionVars[i] >= 0) {
+		LOAD(			optionsVar);
+		STORE(			optionVars[i]);
+		OP(			POP);
+	    }
+
+	    if (!handlerTokens[i]) {
+		/*
+		 * No handler. Will not be the last handler (that is a
+		 * condition that is checked by the caller). Chain to the next
+		 * one.
+		 */
+
+		ExceptionRangeEnds(envPtr, range);
+		OP(			END_CATCH);
+		forwardsNeedFixing = 1;
+		JUMP4(			JUMP, forwardsToFix[i]);
+		goto finishTrapCatchHandling;
+	    }
+	} else if (!handlerTokens[i]) {
+	    /*
+	     * No handler. Will not be the last handler (that condition is
+	     * checked by the caller). Chain to the next one.
+	     */
+
+	    forwardsNeedFixing = 1;
+	    JUMP4(			JUMP, forwardsToFix[i]);
+	    goto endOfThisArm;
+	}
+
+	/*
+	 * Got a handler. Make sure that any pending patch-up actions from
+	 * previous unprocessed handlers are dealt with now that we know where
+	 * they are to jump to.
+	 */
+
+	if (forwardsNeedFixing) {
+	    forwardsNeedFixing = 0;
+	    OP1(			JUMP1, 7);
+	    for (j=0 ; j<i ; j++) {
+		if (forwardsToFix[j] == -1) {
+		    continue;
+		}
+		FIXJUMP4(	forwardsToFix[j]);
+		forwardsToFix[j] = -1;
+	    }
+	    OP4(			BEGIN_CATCH4, range);
+	}
+	BODY(				handlerTokens[i], 5+i*4);
+	ExceptionRangeEnds(envPtr, range);
+	PUSH(				"0");
+	OP(				PUSH_RETURN_OPTIONS);
+	OP4(				REVERSE, 3);
+	OP1(				JUMP1, 5);
+	TclAdjustStackDepth(-3, envPtr);
+	forwardsToFix[i] = -1;
+
+	/*
+	 * Error in handler or setting of variables; replace the stored
+	 * exception with the new one. Note that we only push this if we have
+	 * either a body or some variable setting here. Otherwise this code is
+	 * unreachable.
+	 */
+
+    finishTrapCatchHandling:
+	ExceptionRangeTarget(envPtr, range, catchOffset);
+	OP(				PUSH_RETURN_OPTIONS);
+	OP(				PUSH_RETURN_CODE);
+	OP(				PUSH_RESULT);
+	OP(				END_CATCH);
+	STORE(				resultVar);
+	OP(				POP);
+	PUSH(				"1");
+	OP(				EQ);
+	JUMP1(				JUMP_FALSE, noTrapError);
+	LOAD(				optionsVar);
+	PUSH(				"-during");
+	OP4(				REVERSE, 3);
+	STORE(				optionsVar);
+	OP(				POP);
+	OP44(				DICT_SET, 1, optionsVar);
+	TclAdjustStackDepth(-1, envPtr);
+	JUMP1(				JUMP, trapError);
+	FIXJUMP1(		noTrapError);
+	STORE(				optionsVar);
+	FIXJUMP1(		trapError);
+	/* Skip POP at end; can clean up with subsequent POP */
+	if (i+1 < numHandlers) {
+	    OP(				POP);
+	}
+
+    endOfThisArm:
+	if (i+1 < numHandlers) {
+	    JUMP4(			JUMP, addrsToFix[i]);
+	    TclAdjustStackDepth(1, envPtr);
+	}
+	if (matchClauses[i]) {
+	    FIXJUMP4(		notECJumpSource);
+	}
+	FIXJUMP4(		notCodeJumpSource);
+    }
+
+    /*
+     * Drop the result code, and fix all the jumps from taken clauses - which
+     * drop the result code as their first action - to point straight after
+     * (i.e., to the start of the finally clause).
+     */
+
+    OP(					POP);
+    for (i=0 ; i<numHandlers-1 ; i++) {
+	FIXJUMP4(		addrsToFix[i]);
+    }
+    TclStackFree(interp, forwardsToFix);
+    TclStackFree(interp, addrsToFix);
+
+    /*
+     * Process the finally clause (at last!) Note that we do not wrap this in
+     * error handlers because we would just rethrow immediately anyway. Then
+     * (on normal success) we reissue the exception. Note also that
+     * INST_RETURN_STK can proceed to the next instruction; that'll be the
+     * next command (or some inter-command manipulation).
+     */
+
+    if (!trapZero) {
+	FIXJUMP4(		afterBody);
+    }
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    OP4(				BEGIN_CATCH4, range);
+    ExceptionRangeStarts(envPtr, range);
+    BODY(				finallyToken, 3 + 4*numHandlers);
+    ExceptionRangeEnds(envPtr, range);
+    OP(					END_CATCH);
+    OP(					POP);
+    JUMP1(				JUMP, finalOK);
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    OP(					PUSH_RESULT);
+    OP(					PUSH_RETURN_OPTIONS);
+    OP(					PUSH_RETURN_CODE);
+    OP(					END_CATCH);
+    PUSH(				"1");
+    OP(					EQ);
+    JUMP1(				JUMP_FALSE, noFinalError);
+    LOAD(				optionsVar);
+    PUSH(				"-during");
+    OP4(				REVERSE, 3);
+    STORE(				optionsVar);
+    OP(					POP);
+    OP44(				DICT_SET, 1, optionsVar);
+    TclAdjustStackDepth(-1, envPtr);
+    OP(					POP);
+    JUMP1(				JUMP, finalError);
+    TclAdjustStackDepth(1, envPtr);
+    FIXJUMP1(			noFinalError);
+    STORE(				optionsVar);
+    OP(					POP);
+    FIXJUMP1(			finalError);
+    STORE(				resultVar);
+    OP(					POP);
+    FIXJUMP1(			finalOK);
+    LOAD(				optionsVar);
+    LOAD(				resultVar);
+    INVOKE(				RETURN_STK);
+
+    return TCL_OK;
+}
+
+static int
+IssueTryFinallyInstructions(
+    Tcl_Interp *interp,
+    CompileEnv *envPtr,
+    Tcl_Token *bodyToken,
+    Tcl_Token *finallyToken)
+{
+    DefineLineInformation;	/* TIP #280 */
+    int range, jumpOK, jumpSplice;
+
+    /*
+     * Note that this one is simple enough that we can issue it without
+     * needing a local variable table, making it a universal compilation.
+     */
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    OP4(				BEGIN_CATCH4, range);
+    ExceptionRangeStarts(envPtr, range);
+    BODY(				bodyToken, 1);
+    ExceptionRangeEnds(envPtr, range);
+    OP1(				JUMP1, 3);
+    TclAdjustStackDepth(-1, envPtr);
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    OP(					PUSH_RESULT);
+    OP(					PUSH_RETURN_OPTIONS);
+    OP(					END_CATCH);
+
+    range = TclCreateExceptRange(CATCH_EXCEPTION_RANGE, envPtr);
+    OP4(				BEGIN_CATCH4, range);
+    ExceptionRangeStarts(envPtr, range);
+    BODY(				finallyToken, 3);
+    ExceptionRangeEnds(envPtr, range);
+    OP(					END_CATCH);
+    OP(					POP);
+    JUMP1(				JUMP, jumpOK);
+    ExceptionRangeTarget(envPtr, range, catchOffset);
+    OP(					PUSH_RESULT);
+    OP(					PUSH_RETURN_OPTIONS);
+    OP(					PUSH_RETURN_CODE);
+    OP(					END_CATCH);
+    PUSH(				"1");
+    OP(					EQ);
+    JUMP1(				JUMP_FALSE, jumpSplice);
+    PUSH(				"-during");
+    OP4(				OVER, 3);
+    OP4(				LIST, 2);
+    OP(					LIST_CONCAT);
+    FIXJUMP1(		jumpSplice);
+    OP4(				REVERSE, 4);
+    OP(					POP);
+    OP(					POP);
+    OP1(				JUMP1, 7);
+    FIXJUMP1(		jumpOK);
+    OP4(				REVERSE, 2);
+    INVOKE(				RETURN_STK);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileUnsetCmd --
+ *
+ *	Procedure called to compile the "unset" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "unset" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileUnsetCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *varTokenPtr;
+    int isScalar, localIndex, flags = 1, i, varCount = 0, haveFlags = 0;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* TODO: Consider support for compiling expanded args. */
+
+    /*
+     * Verify that all words - except the first non-option one - are known at
+     * compile time so that we can handle them without needing to do a nasty
+     * push/rotate. [Bug 3970f54c4e]
+     */
+
+    for (i=1,varTokenPtr=parsePtr->tokenPtr ; i<parsePtr->numWords ; i++) {
+	Tcl_Obj *leadingWord = Tcl_NewObj();
+
+	varTokenPtr = TokenAfter(varTokenPtr);
+	if (!TclWordKnownAtCompileTime(varTokenPtr, leadingWord)) {
+	    TclDecrRefCount(leadingWord);
+
+	    /*
+	     * We can tolerate non-trivial substitutions in the first variable
+	     * to be unset. If a '--' or '-nocomplain' was present, anything
+	     * goes in that one place! (All subsequent variable names must be
+	     * constants since we don't want to have to push them all first.)
+	     */
+
+	    if (varCount == 0) {
+		if (haveFlags) {
+		    continue;
+		}
+
+		/*
+		 * In fact, we're OK as long as we're the first argument *and*
+		 * we provably don't start with a '-'. If that is true, then
+		 * even if everything else is varying, we still can't be a
+		 * flag. Otherwise we'll spill to runtime to place a limit on
+		 * the trickiness.
+		 */
+
+		if (varTokenPtr->type == TCL_TOKEN_WORD
+			&& varTokenPtr[1].type == TCL_TOKEN_TEXT
+			&& varTokenPtr[1].size > 0
+			&& varTokenPtr[1].start[0] != '-') {
+		    continue;
+		}
+	    }
+	    return TCL_ERROR;
+	}
+	if (varCount == 0) {
+	    const char *bytes;
+	    int len;
+
+	    bytes = TclGetStringFromObj(leadingWord, &len);
+	    if (i == 1 && len == 11 && !strncmp("-nocomplain", bytes, 11)) {
+		flags = 0;
+		haveFlags++;
+	    } else if (i == (2 - flags) && len == 2 && !strncmp("--", bytes, 2)) {
+		haveFlags++;
+	    } else {
+		varCount++;
+	    }
+	} else {
+	    varCount++;
+	}
+	TclDecrRefCount(leadingWord);
+    }
+
+    /*
+     * Issue instructions to unset each of the named variables.
+     */
+
+    varTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    for (i=0; i<haveFlags;i++) {
+	varTokenPtr = TokenAfter(varTokenPtr);
+    }
+    for (i=1+haveFlags ; i<parsePtr->numWords ; i++) {
+	/*
+	 * Decide if we can use a frame slot for the var/array name or if we
+	 * need to emit code to compute and push the name at runtime. We use a
+	 * frame slot (entry in the array of local vars) if we are compiling a
+	 * procedure body and if the name is simple text that does not include
+	 * namespace qualifiers.
+	 */
+
+	PushVarNameWord(interp, varTokenPtr, envPtr, 0,
+		&localIndex, &isScalar, i);
+
+	/*
+	 * Emit instructions to unset the variable.
+	 */
+
+	if (isScalar) {
+	    if (localIndex < 0) {
+		OP1(	UNSET_STK, flags);
+	    } else {
+		OP14(	UNSET_SCALAR, flags, localIndex);
+	    }
+	} else {
+	    if (localIndex < 0) {
+		OP1(	UNSET_ARRAY_STK, flags);
+	    } else {
+		OP14(	UNSET_ARRAY, flags, localIndex);
+	    }
+	}
+
+	varTokenPtr = TokenAfter(varTokenPtr);
+    }
+    PUSH("");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileWhileCmd --
+ *
+ *	Procedure called to compile the "while" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "while" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileWhileCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *testTokenPtr, *bodyTokenPtr;
+    JumpFixup jumpEvalCondFixup;
+    int testCodeOffset, bodyCodeOffset, jumpDist, range, code, boolVal;
+    int loopMayEnd = 1;		/* This is set to 0 if it is recognized as an
+				 * infinite loop. */
+    Tcl_Obj *boolObj;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the test expression requires substitutions, don't compile the while
+     * command inline. E.g., the expression might cause the loop to never
+     * execute or execute forever, as in "while "$x < 5" {}".
+     *
+     * Bail out also if the body expression requires substitutions in order to
+     * insure correct behaviour [Bug 219166]
+     */
+
+    testTokenPtr = TokenAfter(parsePtr->tokenPtr);
+    bodyTokenPtr = TokenAfter(testTokenPtr);
+
+    if ((testTokenPtr->type != TCL_TOKEN_SIMPLE_WORD)
+	    || (bodyTokenPtr->type != TCL_TOKEN_SIMPLE_WORD)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find out if the condition is a constant.
+     */
+
+    boolObj = Tcl_NewStringObj(testTokenPtr[1].start, testTokenPtr[1].size);
+    Tcl_IncrRefCount(boolObj);
+    code = Tcl_GetBooleanFromObj(NULL, boolObj, &boolVal);
+    TclDecrRefCount(boolObj);
+    if (code == TCL_OK) {
+	if (boolVal) {
+	    /*
+	     * It is an infinite loop; flag it so that we generate a more
+	     * efficient body.
+	     */
+
+	    loopMayEnd = 0;
+	} else {
+	    /*
+	     * This is an empty loop: "while 0 {...}" or such. Compile no
+	     * bytecodes.
+	     */
+
+	    goto pushResult;
+	}
+    }
+
+    /*
+     * Create a ExceptionRange record for the loop body. This is used to
+     * implement break and continue.
+     */
+
+    range = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+
+    /*
+     * Jump to the evaluation of the condition. This code uses the "loop
+     * rotation" optimisation (which eliminates one branch from the loop).
+     * "while cond body" produces then:
+     *       goto A
+     *    B: body                : bodyCodeOffset
+     *    A: cond -> result      : testCodeOffset, continueOffset
+     *       if (result) goto B
+     *
+     * The infinite loop "while 1 body" produces:
+     *    B: body                : all three offsets here
+     *       goto B
+     */
+
+    if (loopMayEnd) {
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP,
+		&jumpEvalCondFixup);
+	testCodeOffset = 0;	/* Avoid compiler warning. */
+    } else {
+	/*
+	 * Make sure that the first command in the body is preceded by an
+	 * INST_START_CMD, and hence counted properly. [Bug 1752146]
+	 */
+
+	envPtr->atCmdStart &= ~1;
+	testCodeOffset = CurrentOffset(envPtr);
+    }
+
+    /*
+     * Compile the loop body.
+     */
+
+    bodyCodeOffset = ExceptionRangeStarts(envPtr, range);
+    if (!loopMayEnd) {
+	envPtr->exceptArrayPtr[range].continueOffset = testCodeOffset;
+	envPtr->exceptArrayPtr[range].codeOffset = bodyCodeOffset;
+    }
+    BODY(bodyTokenPtr, 2);
+    ExceptionRangeEnds(envPtr, range);
+    OP(		POP);
+
+    /*
+     * Compile the test expression then emit the conditional jump that
+     * terminates the while. We already know it's a simple word.
+     */
+
+    if (loopMayEnd) {
+	testCodeOffset = CurrentOffset(envPtr);
+	jumpDist = testCodeOffset - jumpEvalCondFixup.codeOffset;
+	if (TclFixupForwardJump(envPtr, &jumpEvalCondFixup, jumpDist, 127)) {
+	    bodyCodeOffset += 3;
+	    testCodeOffset += 3;
+	}
+	SetLineInformation(1);
+	TclCompileExprWords(interp, testTokenPtr, 1, envPtr);
+
+	jumpDist = CurrentOffset(envPtr) - bodyCodeOffset;
+	if (jumpDist > 127) {
+	    TclEmitInstInt4(INST_JUMP_TRUE4, -jumpDist, envPtr);
+	} else {
+	    TclEmitInstInt1(INST_JUMP_TRUE1, -jumpDist, envPtr);
+	}
+    } else {
+	jumpDist = CurrentOffset(envPtr) - bodyCodeOffset;
+	if (jumpDist > 127) {
+	    TclEmitInstInt4(INST_JUMP4, -jumpDist, envPtr);
+	} else {
+	    TclEmitInstInt1(INST_JUMP1, -jumpDist, envPtr);
+	}
+    }
+
+    /*
+     * Set the loop's body, continue and break offsets.
+     */
+
+    envPtr->exceptArrayPtr[range].continueOffset = testCodeOffset;
+    envPtr->exceptArrayPtr[range].codeOffset = bodyCodeOffset;
+    ExceptionRangeTarget(envPtr, range, breakOffset);
+    TclFinalizeLoopExceptionRange(envPtr, range);
+
+    /*
+     * The while command's result is an empty string.
+     */
+
+  pushResult:
+    PUSH("");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileYieldCmd --
+ *
+ *	Procedure called to compile the "yield" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "yield" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileYieldCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    if (parsePtr->numWords < 1 || parsePtr->numWords > 2) {
+	return TCL_ERROR;
+    }
+
+    if (parsePtr->numWords == 1) {
+	PUSH("");
+    } else {
+	DefineLineInformation;	/* TIP #280 */
+	Tcl_Token *valueTokenPtr = TokenAfter(parsePtr->tokenPtr);
+
+	CompileWord(envPtr, valueTokenPtr, interp, 1);
+    }
+    OP(		YIELD);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileYieldToCmd --
+ *
+ *	Procedure called to compile the "yieldto" command.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the "yieldto" command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileYieldToCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    DefineLineInformation;	/* TIP #280 */
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    int i;
+
+    if (parsePtr->numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    OP(		NS_CURRENT);
+    for (i = 1 ; i < parsePtr->numWords ; i++) {
+	CompileWord(envPtr, tokenPtr, interp, i);
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    OP4(	LIST, i);
+    OP(		YIELD_TO_INVOKE);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileUnaryOpCmd --
+ *
+ *	Utility routine to compile the unary operator commands.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the compiled command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompileUnaryOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    int instruction,
+    CompileEnv *envPtr)
+{
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+    tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    CompileWord(envPtr, tokenPtr, interp, 1);
+    TclEmitOpcode(instruction, envPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileAssociativeBinaryOpCmd --
+ *
+ *	Utility routine to compile the binary operator commands that accept an
+ *	arbitrary number of arguments, and that are associative operations.
+ *	Because of the associativity, we may combine operations from right to
+ *	left, saving us any effort of re-ordering the arguments on the stack
+ *	after substitutions are completed.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the compiled command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompileAssociativeBinaryOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    const char *identity,
+    int instruction,
+    CompileEnv *envPtr)
+{
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int words;
+
+    /* TODO: Consider support for compiling expanded args. */
+    for (words=1 ; words<parsePtr->numWords ; words++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, words);
+    }
+    if (parsePtr->numWords <= 2) {
+	PushLiteral(envPtr, identity, -1);
+	words++;
+    }
+    if (words > 3) {
+	/*
+	 * Reverse order of arguments to get precise agreement with [expr] in
+	 * calcuations, including roundoff errors.
+	 */
+
+	OP4(	REVERSE, words-1);
+    }
+    while (--words > 1) {
+	TclEmitOpcode(instruction, envPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileStrictlyBinaryOpCmd --
+ *
+ *	Utility routine to compile the binary operator commands, that strictly
+ *	accept exactly two arguments.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the compiled command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompileStrictlyBinaryOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    int instruction,
+    CompileEnv *envPtr)
+{
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr,
+	    NULL, instruction, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileComparisonOpCmd --
+ *
+ *	Utility routine to compile the n-ary comparison operator commands.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the compiled command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompileComparisonOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    int instruction,
+    CompileEnv *envPtr)
+{
+    Tcl_Token *tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords < 3) {
+	PUSH("1");
+    } else if (parsePtr->numWords == 3) {
+	tokenPtr = TokenAfter(parsePtr->tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 1);
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 2);
+	TclEmitOpcode(instruction, envPtr);
+    } else if (envPtr->procPtr == NULL) {
+	/*
+	 * No local variable space!
+	 */
+
+	return TCL_ERROR;
+    } else {
+	int tmpIndex = AnonymousLocal(envPtr);
+	int words;
+
+	tokenPtr = TokenAfter(parsePtr->tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 1);
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, 2);
+	STORE(tmpIndex);
+	TclEmitOpcode(instruction, envPtr);
+	for (words=3 ; words<parsePtr->numWords ;) {
+	    LOAD(tmpIndex);
+	    tokenPtr = TokenAfter(tokenPtr);
+	    CompileWord(envPtr, tokenPtr, interp, words);
+	    if (++words < parsePtr->numWords) {
+		STORE(tmpIndex);
+	    }
+	    TclEmitOpcode(instruction, envPtr);
+	}
+	for (; words>3 ; words--) {
+	    OP(	BITAND);
+	}
+
+	/*
+	 * Drop the value from the temp variable; retaining that reference
+	 * might be expensive elsewhere.
+	 */
+
+	OP14(	UNSET_SCALAR, 0, tmpIndex);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompile*OpCmd --
+ *
+ *	Procedures called to compile the corresponding "::tcl::mathop::*"
+ *	commands. These are all wrappers around the utility operator command
+ *	compiler functions, except for the compilers for subtraction and
+ *	division, which are special.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the compiled command at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileInvertOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileUnaryOpCmd(interp, parsePtr, INST_BITNOT, envPtr);
+}
+
+int
+TclCompileNotOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileUnaryOpCmd(interp, parsePtr, INST_LNOT, envPtr);
+}
+
+int
+TclCompileAddOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr, "0", INST_ADD,
+	    envPtr);
+}
+
+int
+TclCompileMulOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr, "1", INST_MULT,
+	    envPtr);
+}
+
+int
+TclCompileAndOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr, "-1", INST_BITAND,
+	    envPtr);
+}
+
+int
+TclCompileOrOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr, "0", INST_BITOR,
+	    envPtr);
+}
+
+int
+TclCompileXorOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileAssociativeBinaryOpCmd(interp, parsePtr, "0", INST_BITXOR,
+	    envPtr);
+}
+
+int
+TclCompilePowOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    /*
+     * This one has its own implementation because the ** operator is the only
+     * one with right associativity.
+     */
+
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int words;
+
+    for (words=1 ; words<parsePtr->numWords ; words++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, words);
+    }
+    if (parsePtr->numWords <= 2) {
+	PUSH("1");
+	words++;
+    }
+    while (--words > 1) {
+	TclEmitOpcode(INST_EXPON, envPtr);
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileLshiftOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_LSHIFT, envPtr);
+}
+
+int
+TclCompileRshiftOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_RSHIFT, envPtr);
+}
+
+int
+TclCompileModOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_MOD, envPtr);
+}
+
+int
+TclCompileNeqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_NEQ, envPtr);
+}
+
+int
+TclCompileStrneqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_STR_NEQ, envPtr);
+}
+
+int
+TclCompileInOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_LIST_IN, envPtr);
+}
+
+int
+TclCompileNiOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileStrictlyBinaryOpCmd(interp, parsePtr, INST_LIST_NOT_IN,
+	    envPtr);
+}
+
+int
+TclCompileLessOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_LT, envPtr);
+}
+
+int
+TclCompileLeqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_LE, envPtr);
+}
+
+int
+TclCompileGreaterOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_GT, envPtr);
+}
+
+int
+TclCompileGeqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_GE, envPtr);
+}
+
+int
+TclCompileEqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_EQ, envPtr);
+}
+
+int
+TclCompileStreqOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    return CompileComparisonOpCmd(interp, parsePtr, INST_STR_EQ, envPtr);
+}
+
+int
+TclCompileMinusOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int words;
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords == 1) {
+	/*
+	 * Fallback to direct eval to report syntax error.
+	 */
+
+	return TCL_ERROR;
+    }
+    for (words=1 ; words<parsePtr->numWords ; words++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, words);
+    }
+    if (words == 2) {
+	TclEmitOpcode(INST_UMINUS, envPtr);
+	return TCL_OK;
+    }
+    if (words == 3) {
+	TclEmitOpcode(INST_SUB, envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Reverse order of arguments to get precise agreement with [expr] in
+     * calcuations, including roundoff errors.
+     */
+
+    TclEmitInstInt4(INST_REVERSE, words-1, envPtr);
+    while (--words > 1) {
+	TclEmitInstInt4(INST_REVERSE, 2, envPtr);
+	TclEmitOpcode(INST_SUB, envPtr);
+    }
+    return TCL_OK;
+}
+
+int
+TclCompileDivOpCmd(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)
+{
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    DefineLineInformation;	/* TIP #280 */
+    int words;
+
+    /* TODO: Consider support for compiling expanded args. */
+    if (parsePtr->numWords == 1) {
+	/*
+	 * Fallback to direct eval to report syntax error.
+	 */
+
+	return TCL_ERROR;
+    }
+    if (parsePtr->numWords == 2) {
+	PUSH("1.0");
+    }
+    for (words=1 ; words<parsePtr->numWords ; words++) {
+	tokenPtr = TokenAfter(tokenPtr);
+	CompileWord(envPtr, tokenPtr, interp, words);
+    }
+    if (words <= 3) {
+	TclEmitOpcode(INST_DIV, envPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Reverse order of arguments to get precise agreement with [expr] in
+     * calcuations, including roundoff errors.
+     */
+
+    TclEmitInstInt4(INST_REVERSE, words-1, envPtr);
+    while (--words > 1) {
+	TclEmitInstInt4(INST_REVERSE, 2, envPtr);
+	TclEmitOpcode(INST_DIV, envPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCompExpr.c b/generic/tclCompExpr.c
new file mode 100644
index 0000000..b854b0f
--- /dev/null
+++ b/generic/tclCompExpr.c
@@ -0,0 +1,2803 @@
+/*
+ * tclCompExpr.c --
+ *
+ *	This file contains the code to parse and compile Tcl expressions and
+ *	implementations of the Tcl commands corresponding to expression
+ *	operators, such as the command ::tcl::mathop::+ .
+ *
+ * Contributions from Don Porter, NIST, 2006-2007. (not subject to US copyright)
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"		/* CompileEnv */
+
+/*
+ * Expression parsing takes place in the routine ParseExpr(). It takes a
+ * string as input, parses that string, and generates a representation of the
+ * expression in the form of a tree of operators, a list of literals, a list
+ * of function names, and an array of Tcl_Token's within a Tcl_Parse struct.
+ * The tree is composed of OpNodes.
+ */
+
+typedef struct OpNode {
+    int left;			/* "Pointer" to the left operand. */
+    int right;			/* "Pointer" to the right operand. */
+    union {
+	int parent;		/* "Pointer" to the parent operand. */
+	int prev;		/* "Pointer" joining incomplete tree stack */
+    } p;
+    unsigned char lexeme;	/* Code that identifies the operator. */
+    unsigned char precedence;	/* Precedence of the operator */
+    unsigned char mark;		/* Mark used to control traversal. */
+    unsigned char constant;	/* Flag marking constant subexpressions. */
+} OpNode;
+
+/*
+ * The storage for the tree is dynamically allocated array of OpNodes. The
+ * array is grown as parsing needs dictate according to a scheme similar to
+ * Tcl's string growth algorithm, so that the resizing costs are O(N) and so
+ * that we use at least half the memory allocated as expressions get large.
+ *
+ * Each OpNode in the tree represents an operator in the expression, either
+ * unary or binary. When parsing is completed successfully, a binary operator
+ * OpNode will have its left and right fields filled with "pointers" to its
+ * left and right operands. A unary operator OpNode will have its right field
+ * filled with a pointer to its single operand. When an operand is a
+ * subexpression the "pointer" takes the form of the index -- a non-negative
+ * integer -- into the OpNode storage array where the root of that
+ * subexpression parse tree is found.
+ *
+ * Non-operator elements of the expression do not get stored in the OpNode
+ * tree. They are stored in the other structures according to their type.
+ * Literal values get appended to the literal list. Elements that denote forms
+ * of quoting or substitution known to the Tcl parser get stored as
+ * Tcl_Tokens. These non-operator elements of the expression are the leaves of
+ * the completed parse tree. When an operand of an OpNode is one of these leaf
+ * elements, the following negative integer codes are used to indicate which
+ * kind of elements it is.
+ */
+
+enum OperandTypes {
+    OT_LITERAL = -3,	/* Operand is a literal in the literal list */
+    OT_TOKENS = -2,	/* Operand is sequence of Tcl_Tokens */
+    OT_EMPTY = -1	/* "Operand" is an empty string. This is a special
+			 * case used only to represent the EMPTY lexeme. See
+			 * below. */
+};
+
+/*
+ * Readable macros to test whether a "pointer" value points to an operator.
+ * They operate on the "non-negative integer -> operator; negative integer ->
+ * a non-operator OperandType" distinction.
+ */
+
+#define IsOperator(l)	((l) >= 0)
+#define NotOperator(l)	((l) < 0)
+
+/*
+ * Note that it is sufficient to store in the tree just the type of leaf
+ * operand, without any explicit pointer to which leaf. This is true because
+ * the traversals of the completed tree we perform are known to visit the
+ * leaves in the same order as the original parse.
+ *
+ * In a completed parse tree, those OpNodes that are themselves (roots of
+ * subexpression trees that are) operands of some operator store in their
+ * p.parent field a "pointer" to the OpNode of that operator. The p.parent
+ * field permits a traversal of the tree within a non-recursive routine
+ * (ConvertTreeToTokens() and CompileExprTree()). This means that even
+ * expression trees of great depth pose no risk of blowing the C stack.
+ *
+ * While the parse tree is being constructed, the same memory space is used to
+ * hold the p.prev field which chains together a stack of incomplete trees
+ * awaiting their right operands.
+ *
+ * The lexeme field is filled in with the lexeme of the operator that is
+ * returned by the ParseLexeme() routine. Only lexemes for unary and binary
+ * operators get stored in an OpNode. Other lexmes get different treatement.
+ *
+ * The precedence field provides a place to store the precedence of the
+ * operator, so it need not be looked up again and again.
+ *
+ * The mark field is use to control the traversal of the tree, so that it can
+ * be done non-recursively. The mark values are:
+ */
+
+enum Marks {
+    MARK_LEFT,		/* Next step of traversal is to visit left subtree */
+    MARK_RIGHT,		/* Next step of traversal is to visit right subtree */
+    MARK_PARENT		/* Next step of traversal is to return to parent */
+};
+
+/*
+ * The constant field is a boolean flag marking which subexpressions are
+ * completely known at compile time, and are eligible for computing then
+ * rather than waiting until run time.
+ */
+
+/*
+ * Each lexeme belongs to one of four categories, which determine its place in
+ * the parse tree. We use the two high bits of the (unsigned char) value to
+ * store a NODE_TYPE code.
+ */
+
+#define NODE_TYPE	0xC0
+
+/*
+ * The four category values are LEAF, UNARY, and BINARY, explained below, and
+ * "uncategorized", which is used either temporarily, until context determines
+ * which of the other three categories is correct, or for lexemes like
+ * INVALID, which aren't really lexemes at all, but indicators of a parsing
+ * error. Note that the codes must be distinct to distinguish categories, but
+ * need not take the form of a bit array.
+ */
+
+#define BINARY		0x40	/* This lexeme is a binary operator. An OpNode
+				 * representing it should go into the parse
+				 * tree, and two operands should be parsed for
+				 * it in the expression. */
+#define UNARY		0x80	/* This lexeme is a unary operator. An OpNode
+				 * representing it should go into the parse
+				 * tree, and one operand should be parsed for
+				 * it in the expression. */
+#define LEAF		0xC0	/* This lexeme is a leaf operand in the parse
+				 * tree. No OpNode will be placed in the tree
+				 * for it. Either a literal value will be
+				 * appended to the list of literals in this
+				 * expression, or appropriate Tcl_Tokens will
+				 * be appended in a Tcl_Parse struct to
+				 * represent those leaves that require some
+				 * form of substitution. */
+
+/* Uncategorized lexemes */
+
+#define PLUS		1	/* Ambiguous. Resolves to UNARY_PLUS or
+				 * BINARY_PLUS according to context. */
+#define MINUS		2	/* Ambiguous. Resolves to UNARY_MINUS or
+				 * BINARY_MINUS according to context. */
+#define BAREWORD	3	/* Ambigous. Resolves to BOOLEAN or to
+				 * FUNCTION or a parse error according to
+				 * context and value. */
+#define INCOMPLETE	4	/* A parse error. Used only when the single
+				 * "=" is encountered.  */
+#define INVALID		5	/* A parse error. Used when any punctuation
+				 * appears that's not a supported operator. */
+
+/* Leaf lexemes */
+
+#define NUMBER		(LEAF | 1)
+				/* For literal numbers */
+#define SCRIPT		(LEAF | 2)
+				/* Script substitution; [foo] */
+#define BOOLEAN		(LEAF | BAREWORD)
+				/* For literal booleans */
+#define BRACED		(LEAF | 4)
+				/* Braced string; {foo bar} */
+#define VARIABLE	(LEAF | 5)
+				/* Variable substitution; $x */
+#define QUOTED		(LEAF | 6)
+				/* Quoted string; "foo $bar [soom]" */
+#define EMPTY		(LEAF | 7)
+				/* Used only for an empty argument list to a
+				 * function. Represents the empty string
+				 * within parens in the expression: rand() */
+
+/* Unary operator lexemes */
+
+#define UNARY_PLUS	(UNARY | PLUS)
+#define UNARY_MINUS	(UNARY | MINUS)
+#define FUNCTION	(UNARY | BAREWORD)
+				/* This is a bit of "creative interpretation"
+				 * on the part of the parser. A function call
+				 * is parsed into the parse tree according to
+				 * the perspective that the function name is a
+				 * unary operator and its argument list,
+				 * enclosed in parens, is its operand. The
+				 * additional requirements not implied
+				 * generally by treatment as a unary operator
+				 * -- for example, the requirement that the
+				 * operand be enclosed in parens -- are hard
+				 * coded in the relevant portions of
+				 * ParseExpr(). We trade off the need to
+				 * include such exceptional handling in the
+				 * code against the need we would otherwise
+				 * have for more lexeme categories. */
+#define START		(UNARY | 4)
+				/* This lexeme isn't parsed from the
+				 * expression text at all. It represents the
+				 * start of the expression and sits at the
+				 * root of the parse tree where it serves as
+				 * the start/end point of traversals. */
+#define OPEN_PAREN	(UNARY | 5)
+				/* Another bit of creative interpretation,
+				 * where we treat "(" as a unary operator with
+				 * the sub-expression between it and its
+				 * matching ")" as its operand. See
+				 * CLOSE_PAREN below. */
+#define NOT		(UNARY | 6)
+#define BIT_NOT		(UNARY | 7)
+
+/* Binary operator lexemes */
+
+#define BINARY_PLUS	(BINARY |  PLUS)
+#define BINARY_MINUS	(BINARY |  MINUS)
+#define COMMA		(BINARY |  3)
+				/* The "," operator is a low precedence binary
+				 * operator that separates the arguments in a
+				 * function call. The additional constraint
+				 * that this operator can only legally appear
+				 * at the right places within a function call
+				 * argument list are hard coded within
+				 * ParseExpr().  */
+#define MULT		(BINARY |  4)
+#define DIVIDE		(BINARY |  5)
+#define MOD		(BINARY |  6)
+#define LESS		(BINARY |  7)
+#define GREATER		(BINARY |  8)
+#define BIT_AND		(BINARY |  9)
+#define BIT_XOR		(BINARY | 10)
+#define BIT_OR		(BINARY | 11)
+#define QUESTION	(BINARY | 12)
+				/* These two lexemes make up the */
+#define COLON		(BINARY | 13)
+				/* ternary conditional operator, $x ? $y : $z.
+				 * We treat them as two binary operators to
+				 * avoid another lexeme category, and code the
+				 * additional constraints directly in
+				 * ParseExpr(). For instance, the right
+				 * operand of a "?" operator must be a ":"
+				 * operator. */
+#define LEFT_SHIFT	(BINARY | 14)
+#define RIGHT_SHIFT	(BINARY | 15)
+#define LEQ		(BINARY | 16)
+#define GEQ		(BINARY | 17)
+#define EQUAL		(BINARY | 18)
+#define NEQ		(BINARY | 19)
+#define AND		(BINARY | 20)
+#define OR		(BINARY | 21)
+#define STREQ		(BINARY | 22)
+#define STRNEQ		(BINARY | 23)
+#define EXPON		(BINARY | 24)
+				/* Unlike the other binary operators, EXPON is
+				 * right associative and this distinction is
+				 * coded directly in ParseExpr(). */
+#define IN_LIST		(BINARY | 25)
+#define NOT_IN_LIST	(BINARY | 26)
+#define CLOSE_PAREN	(BINARY | 27)
+				/* By categorizing the CLOSE_PAREN lexeme as a
+				 * BINARY operator, the normal parsing rules
+				 * for binary operators assure that a close
+				 * paren will not directly follow another
+				 * operator, and the machinery already in
+				 * place to connect operands to operators
+				 * according to precedence performs most of
+				 * the work of matching open and close parens
+				 * for us. In the end though, a close paren is
+				 * not really a binary operator, and some
+				 * special coding in ParseExpr() make sure we
+				 * never put an actual CLOSE_PAREN node in the
+				 * parse tree. The sub-expression between
+				 * parens becomes the single argument of the
+				 * matching OPEN_PAREN unary operator. */
+#define END		(BINARY | 28)
+				/* This lexeme represents the end of the
+				 * string being parsed. Treating it as a
+				 * binary operator follows the same logic as
+				 * the CLOSE_PAREN lexeme and END pairs with
+				 * START, in the same way that CLOSE_PAREN
+				 * pairs with OPEN_PAREN. */
+
+/*
+ * When ParseExpr() builds the parse tree it must choose which operands to
+ * connect to which operators.  This is done according to operator precedence.
+ * The greater an operator's precedence the greater claim it has to link to an
+ * available operand.  The Precedence enumeration lists the precedence values
+ * used by Tcl expression operators, from lowest to highest claim.  Each
+ * precedence level is commented with the operators that hold that precedence.
+ */
+
+enum Precedence {
+    PREC_END = 1,	/* END */
+    PREC_START,		/* START */
+    PREC_CLOSE_PAREN,	/* ")" */
+    PREC_OPEN_PAREN,	/* "(" */
+    PREC_COMMA,		/* "," */
+    PREC_CONDITIONAL,	/* "?", ":" */
+    PREC_OR,		/* "||" */
+    PREC_AND,		/* "&&" */
+    PREC_BIT_OR,	/* "|" */
+    PREC_BIT_XOR,	/* "^" */
+    PREC_BIT_AND,	/* "&" */
+    PREC_EQUAL,		/* "==", "!=", "eq", "ne", "in", "ni" */
+    PREC_COMPARE,	/* "<", ">", "<=", ">=" */
+    PREC_SHIFT,		/* "<<", ">>" */
+    PREC_ADD,		/* "+", "-" */
+    PREC_MULT,		/* "*", "/", "%" */
+    PREC_EXPON,		/* "**" */
+    PREC_UNARY		/* "+", "-", FUNCTION, "!", "~" */
+};
+
+/*
+ * Here the same information contained in the comments above is stored in
+ * inverted form, so that given a lexeme, one can quickly look up its
+ * precedence value.
+ */
+
+static const unsigned char prec[] = {
+    /* Non-operator lexemes */
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,
+    /* Binary operator lexemes */
+    PREC_ADD,		/* BINARY_PLUS */
+    PREC_ADD,		/* BINARY_MINUS */
+    PREC_COMMA,		/* COMMA */
+    PREC_MULT,		/* MULT */
+    PREC_MULT,		/* DIVIDE */
+    PREC_MULT,		/* MOD */
+    PREC_COMPARE,	/* LESS */
+    PREC_COMPARE,	/* GREATER */
+    PREC_BIT_AND,	/* BIT_AND */
+    PREC_BIT_XOR,	/* BIT_XOR */
+    PREC_BIT_OR,	/* BIT_OR */
+    PREC_CONDITIONAL,	/* QUESTION */
+    PREC_CONDITIONAL,	/* COLON */
+    PREC_SHIFT,		/* LEFT_SHIFT */
+    PREC_SHIFT,		/* RIGHT_SHIFT */
+    PREC_COMPARE,	/* LEQ */
+    PREC_COMPARE,	/* GEQ */
+    PREC_EQUAL,		/* EQUAL */
+    PREC_EQUAL,		/* NEQ */
+    PREC_AND,		/* AND */
+    PREC_OR,		/* OR */
+    PREC_EQUAL,		/* STREQ */
+    PREC_EQUAL,		/* STRNEQ */
+    PREC_EXPON,		/* EXPON */
+    PREC_EQUAL,		/* IN_LIST */
+    PREC_EQUAL,		/* NOT_IN_LIST */
+    PREC_CLOSE_PAREN,	/* CLOSE_PAREN */
+    PREC_END,		/* END */
+    /* Expansion room for more binary operators */
+    0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,
+    /* Unary operator lexemes */
+    PREC_UNARY,		/* UNARY_PLUS */
+    PREC_UNARY,		/* UNARY_MINUS */
+    PREC_UNARY,		/* FUNCTION */
+    PREC_START,		/* START */
+    PREC_OPEN_PAREN,	/* OPEN_PAREN */
+    PREC_UNARY,		/* NOT*/
+    PREC_UNARY,		/* BIT_NOT*/
+};
+
+/*
+ * A table mapping lexemes to bytecode instructions, used by CompileExprTree().
+ */
+
+static const unsigned char instruction[] = {
+    /* Non-operator lexemes */
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,
+    /* Binary operator lexemes */
+    INST_ADD,		/* BINARY_PLUS */
+    INST_SUB,		/* BINARY_MINUS */
+    0,			/* COMMA */
+    INST_MULT,		/* MULT */
+    INST_DIV,		/* DIVIDE */
+    INST_MOD,		/* MOD */
+    INST_LT,		/* LESS */
+    INST_GT,		/* GREATER */
+    INST_BITAND,	/* BIT_AND */
+    INST_BITXOR,	/* BIT_XOR */
+    INST_BITOR,		/* BIT_OR */
+    0,			/* QUESTION */
+    0,			/* COLON */
+    INST_LSHIFT,	/* LEFT_SHIFT */
+    INST_RSHIFT,	/* RIGHT_SHIFT */
+    INST_LE,		/* LEQ */
+    INST_GE,		/* GEQ */
+    INST_EQ,		/* EQUAL */
+    INST_NEQ,		/* NEQ */
+    0,			/* AND */
+    0,			/* OR */
+    INST_STR_EQ,	/* STREQ */
+    INST_STR_NEQ,	/* STRNEQ */
+    INST_EXPON,		/* EXPON */
+    INST_LIST_IN,	/* IN_LIST */
+    INST_LIST_NOT_IN,	/* NOT_IN_LIST */
+    0,			/* CLOSE_PAREN */
+    0,			/* END */
+    /* Expansion room for more binary operators */
+    0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,
+    /* Unary operator lexemes */
+    INST_UPLUS,		/* UNARY_PLUS */
+    INST_UMINUS,	/* UNARY_MINUS */
+    0,			/* FUNCTION */
+    0,			/* START */
+    0,			/* OPEN_PAREN */
+    INST_LNOT,		/* NOT*/
+    INST_BITNOT,	/* BIT_NOT*/
+};
+
+/*
+ * A table mapping a byte value to the corresponding lexeme for use by
+ * ParseLexeme().
+ */
+
+static const unsigned char Lexeme[] = {
+	INVALID		/* NUL */,	INVALID		/* SOH */,
+	INVALID		/* STX */,	INVALID		/* ETX */,
+	INVALID		/* EOT */,	INVALID		/* ENQ */,
+	INVALID		/* ACK */,	INVALID		/* BEL */,
+	INVALID		/* BS */,	INVALID		/* HT */,
+	INVALID		/* LF */,	INVALID		/* VT */,
+	INVALID		/* FF */,	INVALID		/* CR */,
+	INVALID		/* SO */,	INVALID		/* SI */,
+	INVALID		/* DLE */,	INVALID		/* DC1 */,
+	INVALID		/* DC2 */,	INVALID		/* DC3 */,
+	INVALID		/* DC4 */,	INVALID		/* NAK */,
+	INVALID		/* SYN */,	INVALID		/* ETB */,
+	INVALID		/* CAN */,	INVALID		/* EM */,
+	INVALID		/* SUB */,	INVALID		/* ESC */,
+	INVALID		/* FS */,	INVALID		/* GS */,
+	INVALID		/* RS */,	INVALID		/* US */,
+	INVALID		/* SPACE */,	0		/* ! or != */,
+	QUOTED		/* " */,	INVALID		/* # */,
+	VARIABLE	/* $ */,	MOD		/* % */,
+	0		/* & or && */,	INVALID		/* ' */,
+	OPEN_PAREN	/* ( */,	CLOSE_PAREN	/* ) */,
+	0		/* * or ** */,	PLUS		/* + */,
+	COMMA		/* , */,	MINUS		/* - */,
+	0		/* . */,	DIVIDE		/* / */,
+	0, 0, 0, 0, 0, 0, 0, 0, 0, 0,			/* 0-9 */
+	COLON		/* : */,	INVALID		/* ; */,
+	0		/* < or << or <= */,
+	0		/* == or INVALID */,
+	0		/* > or >> or >= */,
+	QUESTION	/* ? */,	INVALID		/* @ */,
+	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,		/* A-M */
+	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,		/* N-Z */
+	SCRIPT		/* [ */,	INVALID		/* \ */,
+	INVALID		/* ] */,	BIT_XOR		/* ^ */,
+	INVALID		/* _ */,	INVALID		/* ` */,
+	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,		/* a-m */
+	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,		/* n-z */
+	BRACED		/* { */,	0		/* | or || */,
+	INVALID		/* } */,	BIT_NOT		/* ~ */,
+	INVALID		/* DEL */
+};
+
+/*
+ * The JumpList struct is used to create a stack of data needed for the
+ * TclEmitForwardJump() and TclFixupForwardJump() calls that are performed
+ * when compiling the short-circuiting operators QUESTION/COLON, AND, and OR.
+ * Keeping a stack permits the CompileExprTree() routine to be non-recursive.
+ */
+
+typedef struct JumpList {
+    JumpFixup jump;		/* Pass this argument to matching calls of
+				 * TclEmitForwardJump() and
+				 * TclFixupForwardJump(). */
+    struct JumpList *next;	/* Point to next item on the stack */
+} JumpList;
+
+/*
+ * Declarations for local functions to this file:
+ */
+
+static void		CompileExprTree(Tcl_Interp *interp, OpNode *nodes,
+			    int index, Tcl_Obj *const **litObjvPtr,
+			    Tcl_Obj *const *funcObjv, Tcl_Token *tokenPtr,
+			    CompileEnv *envPtr, int optimize);
+static void		ConvertTreeToTokens(const char *start, int numBytes,
+			    OpNode *nodes, Tcl_Token *tokenPtr,
+			    Tcl_Parse *parsePtr);
+static int		ExecConstantExprTree(Tcl_Interp *interp, OpNode *nodes,
+			    int index, Tcl_Obj * const **litObjvPtr);
+static int		ParseExpr(Tcl_Interp *interp, const char *start,
+			    int numBytes, OpNode **opTreePtr,
+			    Tcl_Obj *litList, Tcl_Obj *funcList,
+			    Tcl_Parse *parsePtr, int parseOnly);
+static int		ParseLexeme(const char *start, int numBytes,
+			    unsigned char *lexemePtr, Tcl_Obj **literalPtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseExpr --
+ *
+ *	Given a string, the numBytes bytes starting at start, this function
+ *	parses it as a Tcl expression and constructs a tree representing the
+ *	structure of the expression. The caller must pass in empty lists as
+ *	the funcList and litList arguments. The elements of the parsed
+ *	expression are returned to the caller as that tree, a list of literal
+ *	values, a list of function names, and in Tcl_Tokens added to a
+ *	Tcl_Parse struct passed in by the caller.
+ *
+ * Results:
+ *	If the string is successfully parsed as a valid Tcl expression, TCL_OK
+ *	is returned, and data about the expression structure is written to the
+ *	last four arguments. If the string cannot be parsed as a valid Tcl
+ *	expression, TCL_ERROR is returned, and if interp is non-NULL, an error
+ *	message is written to interp.
+ *
+ * Side effects:
+ *	Memory will be allocated. If TCL_OK is returned, the caller must clean
+ *	up the returned data structures. The (OpNode *) value written to
+ *	opTreePtr should be passed to ckfree() and the parsePtr argument
+ *	should be passed to Tcl_FreeParse(). The elements appended to the
+ *	litList and funcList will automatically be freed whenever the refcount
+ *	on those lists indicates they can be freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseExpr(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    const char *start,		/* Start of source string to parse. */
+    int numBytes,		/* Number of bytes in string. */
+    OpNode **opTreePtr,		/* Points to space where a pointer to the
+				 * allocated OpNode tree should go. */
+    Tcl_Obj *litList,		/* List to append literals to. */
+    Tcl_Obj *funcList,		/* List to append function names to. */
+    Tcl_Parse *parsePtr,	/* Structure to fill with tokens representing
+				 * those operands that require run time
+				 * substitutions. */
+    int parseOnly)		/* A boolean indicating whether the caller's
+				 * aim is just a parse, or whether it will go
+				 * on to compile the expression. Different
+				 * optimizations are appropriate for the two
+				 * scenarios. */
+{
+    OpNode *nodes = NULL;	/* Pointer to the OpNode storage array where
+				 * we build the parse tree. */
+    unsigned int nodesAvailable = 64; /* Initial size of the storage array. This
+				 * value establishes a minimum tree memory
+				 * cost of only about 1 kibyte, and is large
+				 * enough for most expressions to parse with
+				 * no need for array growth and
+				 * reallocation. */
+    unsigned int nodesUsed = 0;	/* Number of OpNodes filled. */
+    int scanned = 0;		/* Capture number of byte scanned by parsing
+				 * routines. */
+    int lastParsed;		/* Stores info about what the lexeme parsed
+				 * the previous pass through the parsing loop
+				 * was. If it was an operator, lastParsed is
+				 * the index of the OpNode for that operator.
+				 * If it was not an operator, lastParsed holds
+				 * an OperandTypes value encoding what we need
+				 * to know about it. */
+    int incomplete;		/* Index of the most recent incomplete tree in
+				 * the OpNode array. Heads a stack of
+				 * incomplete trees linked by p.prev. */
+    int complete = OT_EMPTY;	/* "Index" of the complete tree (that is, a
+				 * complete subexpression) determined at the
+				 * moment. OT_EMPTY is a nonsense value used
+				 * only to silence compiler warnings. During a
+				 * parse, complete will always hold an index
+				 * or an OperandTypes value pointing to an
+				 * actual leaf at the time the complete tree
+				 * is needed. */
+
+    /*
+     * These variables control generation of the error message.
+     */
+
+    Tcl_Obj *msg = NULL;	/* The error message. */
+    Tcl_Obj *post = NULL;	/* In a few cases, an additional postscript
+				 * for the error message, supplying more
+				 * information after the error msg and
+				 * location have been reported. */
+    const char *errCode = NULL;	/* The detail word of the errorCode list, or
+				 * NULL to indicate that no changes to the
+				 * errorCode are to be done. */
+    const char *subErrCode = NULL;
+				/* Extra information for use in generating the
+				 * errorCode. */
+    const char *mark = "_@_";	/* In the portion of the complete error
+				 * message where the error location is
+				 * reported, this "mark" substring is inserted
+				 * into the string being parsed to aid in
+				 * pinpointing the location of the syntax
+				 * error in the expression. */
+    int insertMark = 0;		/* A boolean controlling whether the "mark"
+				 * should be inserted. */
+    const int limit = 25;	/* Portions of the error message are
+				 * constructed out of substrings of the
+				 * original expression. In order to keep the
+				 * error message readable, we impose this
+				 * limit on the substring size we extract. */
+
+    TclParseInit(interp, start, numBytes, parsePtr);
+
+    nodes = attemptckalloc(nodesAvailable * sizeof(OpNode));
+    if (nodes == NULL) {
+	TclNewLiteralStringObj(msg, "not enough memory to parse expression");
+	errCode = "NOMEM";
+	goto error;
+    }
+
+    /*
+     * Initialize the parse tree with the special "START" node.
+     */
+
+    nodes->lexeme = START;
+    nodes->precedence = prec[START];
+    nodes->mark = MARK_RIGHT;
+    nodes->constant = 1;
+    incomplete = lastParsed = nodesUsed;
+    nodesUsed++;
+
+    /*
+     * Main parsing loop parses one lexeme per iteration. We exit the loop
+     * only when there's a syntax error with a "goto error" which takes us to
+     * the error handling code following the loop, or when we've successfully
+     * completed the parse and we return to the caller.
+     */
+
+    while (1) {
+	OpNode *nodePtr;	/* Points to the OpNode we may fill this pass
+				 * through the loop. */
+	unsigned char lexeme;	/* The lexeme we parse this iteration. */
+	Tcl_Obj *literal;	/* Filled by the ParseLexeme() call when a
+				 * literal is parsed that has a Tcl_Obj rep
+				 * worth preserving. */
+
+	/*
+	 * Each pass through this loop adds up to one more OpNode. Allocate
+	 * space for one if required.
+	 */
+
+	if (nodesUsed >= nodesAvailable) {
+	    unsigned int size = nodesUsed * 2;
+	    OpNode *newPtr = NULL;
+
+	    do {
+	      if (size <= UINT_MAX/sizeof(OpNode)) {
+		newPtr = attemptckrealloc(nodes, size * sizeof(OpNode));
+	      }
+	    } while ((newPtr == NULL)
+		    && ((size -= (size - nodesUsed) / 2) > nodesUsed));
+	    if (newPtr == NULL) {
+		TclNewLiteralStringObj(msg,
+			"not enough memory to parse expression");
+		errCode = "NOMEM";
+		goto error;
+	    }
+	    nodesAvailable = size;
+	    nodes = newPtr;
+	}
+	nodePtr = nodes + nodesUsed;
+
+	/*
+	 * Skip white space between lexemes.
+	 */
+
+	scanned = TclParseAllWhiteSpace(start, numBytes);
+	start += scanned;
+	numBytes -= scanned;
+
+	scanned = ParseLexeme(start, numBytes, &lexeme, &literal);
+
+	/*
+	 * Use context to categorize the lexemes that are ambiguous.
+	 */
+
+	if ((NODE_TYPE & lexeme) == 0) {
+	    int b;
+
+	    switch (lexeme) {
+	    case INVALID:
+		msg = Tcl_ObjPrintf("invalid character \"%.*s\"",
+			scanned, start);
+		errCode = "BADCHAR";
+		goto error;
+	    case INCOMPLETE:
+		msg = Tcl_ObjPrintf("incomplete operator \"%.*s\"",
+			scanned, start);
+		errCode = "PARTOP";
+		goto error;
+	    case BAREWORD:
+
+		/*
+		 * Most barewords in an expression are a syntax error. The
+		 * exceptions are that when a bareword is followed by an open
+		 * paren, it might be a function call, and when the bareword
+		 * is a legal literal boolean value, we accept that as well.
+		 */
+
+		if (start[scanned+TclParseAllWhiteSpace(
+			start+scanned, numBytes-scanned)] == '(') {
+		    lexeme = FUNCTION;
+
+		    /*
+		     * When we compile the expression we'll need the function
+		     * name, and there's no place in the parse tree to store
+		     * it, so we keep a separate list of all the function
+		     * names we've parsed in the order we found them.
+		     */
+
+		    Tcl_ListObjAppendElement(NULL, funcList, literal);
+		} else if (Tcl_GetBooleanFromObj(NULL,literal,&b) == TCL_OK) {
+		    lexeme = BOOLEAN;
+		} else {
+		    Tcl_DecrRefCount(literal);
+		    msg = Tcl_ObjPrintf("invalid bareword \"%.*s%s\"",
+			    (scanned < limit) ? scanned : limit - 3, start,
+			    (scanned < limit) ? "" : "...");
+		    post = Tcl_ObjPrintf(
+			    "should be \"$%.*s%s\" or \"{%.*s%s}\"",
+			    (scanned < limit) ? scanned : limit - 3,
+			    start, (scanned < limit) ? "" : "...",
+			    (scanned < limit) ? scanned : limit - 3,
+			    start, (scanned < limit) ? "" : "...");
+		    Tcl_AppendPrintfToObj(post, " or \"%.*s%s(...)\" or ...",
+			    (scanned < limit) ? scanned : limit - 3,
+			    start, (scanned < limit) ? "" : "...");
+		    errCode = "BAREWORD";
+		    if (start[0] == '0') {
+			const char *stop;
+			TclParseNumber(NULL, NULL, NULL, start, scanned,
+				&stop, TCL_PARSE_NO_WHITESPACE);
+
+			if (isdigit(UCHAR(*stop)) || (stop == start + 1)) {
+			    switch (start[1]) {
+			    case 'b':
+				Tcl_AppendToObj(post,
+					" (invalid binary number?)", -1);
+				parsePtr->errorType = TCL_PARSE_BAD_NUMBER;
+				errCode = "BADNUMBER";
+				subErrCode = "BINARY";
+				break;
+			    case 'o':
+				Tcl_AppendToObj(post,
+					" (invalid octal number?)", -1);
+				parsePtr->errorType = TCL_PARSE_BAD_NUMBER;
+				errCode = "BADNUMBER";
+				subErrCode = "OCTAL";
+				break;
+			    default:
+				if (isdigit(UCHAR(start[1]))) {
+				    Tcl_AppendToObj(post,
+					    " (invalid octal number?)", -1);
+				    parsePtr->errorType = TCL_PARSE_BAD_NUMBER;
+				    errCode = "BADNUMBER";
+				    subErrCode = "OCTAL";
+				}
+				break;
+			    }
+			}
+		    }
+		    goto error;
+		}
+		break;
+	    case PLUS:
+	    case MINUS:
+		if (IsOperator(lastParsed)) {
+		    /*
+		     * A "+" or "-" coming just after another operator must be
+		     * interpreted as a unary operator.
+		     */
+
+		    lexeme |= UNARY;
+		} else {
+		    lexeme |= BINARY;
+		}
+	    }
+	}	/* Uncategorized lexemes */
+
+	/*
+	 * Handle lexeme based on its category.
+	 */
+
+	switch (NODE_TYPE & lexeme) {
+	case LEAF: {
+	    /*
+	     * Each LEAF results in either a literal getting appended to the
+	     * litList, or a sequence of Tcl_Tokens representing a Tcl word
+	     * getting appended to the parsePtr->tokens. No OpNode is filled
+	     * for this lexeme.
+	     */
+
+	    Tcl_Token *tokenPtr;
+	    const char *end = start;
+	    int wordIndex;
+	    int code = TCL_OK;
+
+	    /*
+	     * A leaf operand appearing just after something that's not an
+	     * operator is a syntax error.
+	     */
+
+	    if (NotOperator(lastParsed)) {
+		msg = Tcl_ObjPrintf("missing operator at %s", mark);
+		errCode = "MISSING";
+		scanned = 0;
+		insertMark = 1;
+
+		/*
+		 * Free any literal to avoid a memleak.
+		 */
+
+		if ((lexeme == NUMBER) || (lexeme == BOOLEAN)) {
+		    Tcl_DecrRefCount(literal);
+		}
+		goto error;
+	    }
+
+	    switch (lexeme) {
+	    case NUMBER:
+	    case BOOLEAN:
+		/*
+		 * TODO: Consider using a dict or hash to collapse all
+		 * duplicate literals into a single representative value.
+		 * (Like what is done with [split $s {}]).
+		 * Pro:	~75% memory saving on expressions like
+		 *	{1+1+1+1+1+.....+1} (Convert "pointer + Tcl_Obj" cost
+		 *	to "pointer" cost only)
+		 * Con:	Cost of the dict store/retrieve on every literal in
+		 *	every expression when expressions like the above tend
+		 *	to be uncommon.
+		 *	The memory savings is temporary; Compiling to bytecode
+		 *	will collapse things as literals are registered
+		 *	anyway, so the savings applies only to the time
+		 *	between parsing and compiling. Possibly important due
+		 *	to high-water mark nature of memory allocation.
+		 */
+
+		Tcl_ListObjAppendElement(NULL, litList, literal);
+		complete = lastParsed = OT_LITERAL;
+		start += scanned;
+		numBytes -= scanned;
+		continue;
+
+	    default:
+		break;
+	    }
+
+	    /*
+	     * Remaining LEAF cases may involve filling Tcl_Tokens, so make
+	     * room for at least 2 more tokens.
+	     */
+
+	    TclGrowParseTokenArray(parsePtr, 2);
+	    wordIndex = parsePtr->numTokens;
+	    tokenPtr = parsePtr->tokenPtr + wordIndex;
+	    tokenPtr->type = TCL_TOKEN_WORD;
+	    tokenPtr->start = start;
+	    parsePtr->numTokens++;
+
+	    switch (lexeme) {
+	    case QUOTED:
+		code = Tcl_ParseQuotedString(NULL, start, numBytes,
+			parsePtr, 1, &end);
+		scanned = end - start;
+		break;
+
+	    case BRACED:
+		code = Tcl_ParseBraces(NULL, start, numBytes,
+			parsePtr, 1, &end);
+		scanned = end - start;
+		break;
+
+	    case VARIABLE:
+		code = Tcl_ParseVarName(NULL, start, numBytes, parsePtr, 1);
+
+		/*
+		 * Handle the quirk that Tcl_ParseVarName reports a successful
+		 * parse even when it gets only a "$" with no variable name.
+		 */
+
+		tokenPtr = parsePtr->tokenPtr + wordIndex + 1;
+		if (code == TCL_OK && tokenPtr->type != TCL_TOKEN_VARIABLE) {
+		    TclNewLiteralStringObj(msg, "invalid character \"$\"");
+		    errCode = "BADCHAR";
+		    goto error;
+		}
+		scanned = tokenPtr->size;
+		break;
+
+	    case SCRIPT: {
+		Tcl_Parse *nestedPtr =
+			TclStackAlloc(interp, sizeof(Tcl_Parse));
+
+		tokenPtr = parsePtr->tokenPtr + parsePtr->numTokens;
+		tokenPtr->type = TCL_TOKEN_COMMAND;
+		tokenPtr->start = start;
+		tokenPtr->numComponents = 0;
+
+		end = start + numBytes;
+		start++;
+		while (1) {
+		    code = Tcl_ParseCommand(interp, start, end - start, 1,
+			    nestedPtr);
+		    if (code != TCL_OK) {
+			parsePtr->term = nestedPtr->term;
+			parsePtr->errorType = nestedPtr->errorType;
+			parsePtr->incomplete = nestedPtr->incomplete;
+			break;
+		    }
+		    start = nestedPtr->commandStart + nestedPtr->commandSize;
+		    Tcl_FreeParse(nestedPtr);
+		    if ((nestedPtr->term < end) && (nestedPtr->term[0] == ']')
+			    && !nestedPtr->incomplete) {
+			break;
+		    }
+
+		    if (start == end) {
+			TclNewLiteralStringObj(msg, "missing close-bracket");
+			parsePtr->term = tokenPtr->start;
+			parsePtr->errorType = TCL_PARSE_MISSING_BRACKET;
+			parsePtr->incomplete = 1;
+			code = TCL_ERROR;
+			errCode = "UNBALANCED";
+			break;
+		    }
+		}
+		TclStackFree(interp, nestedPtr);
+		end = start;
+		start = tokenPtr->start;
+		scanned = end - start;
+		tokenPtr->size = scanned;
+		parsePtr->numTokens++;
+		break;
+	    }			/* SCRIPT case */
+	    }
+	    if (code != TCL_OK) {
+		/*
+		 * Here we handle all the syntax errors generated by the
+		 * Tcl_Token generating parsing routines called in the switch
+		 * just above. If the value of parsePtr->incomplete is 1, then
+		 * the error was an unbalanced '[', '(', '{', or '"' and
+		 * parsePtr->term is pointing to that unbalanced character. If
+		 * the value of parsePtr->incomplete is 0, then the error is
+		 * one of lacking whitespace following a quoted word, for
+		 * example: expr {[an error {foo}bar]}, and parsePtr->term
+		 * points to where the whitespace is missing. We reset our
+		 * values of start and scanned so that when our error message
+		 * is constructed, the location of the syntax error is sure to
+		 * appear in it, even if the quoted expression is truncated.
+		 */
+
+		start = parsePtr->term;
+		scanned = parsePtr->incomplete;
+		if (parsePtr->incomplete) {
+		    errCode = "UNBALANCED";
+		}
+		goto error;
+	    }
+
+	    tokenPtr = parsePtr->tokenPtr + wordIndex;
+	    tokenPtr->size = scanned;
+	    tokenPtr->numComponents = parsePtr->numTokens - wordIndex - 1;
+	    if (!parseOnly && ((lexeme == QUOTED) || (lexeme == BRACED))) {
+		/*
+		 * When this expression is destined to be compiled, and a
+		 * braced or quoted word within an expression is known at
+		 * compile time (no runtime substitutions in it), we can store
+		 * it as a literal rather than in its tokenized form. This is
+		 * an advantage since the compiled bytecode is going to need
+		 * the argument in Tcl_Obj form eventually, so it's just as
+		 * well to get there now. Another advantage is that with this
+		 * conversion, larger constant expressions might be grown and
+		 * optimized.
+		 *
+		 * On the contrary, if the end goal of this parse is to fill a
+		 * Tcl_Parse for a caller of Tcl_ParseExpr(), then it's
+		 * wasteful to convert to a literal only to convert back again
+		 * later.
+		 */
+
+		literal = Tcl_NewObj();
+		if (TclWordKnownAtCompileTime(tokenPtr, literal)) {
+		    Tcl_ListObjAppendElement(NULL, litList, literal);
+		    complete = lastParsed = OT_LITERAL;
+		    parsePtr->numTokens = wordIndex;
+		    break;
+		}
+		Tcl_DecrRefCount(literal);
+	    }
+	    complete = lastParsed = OT_TOKENS;
+	    break;
+	} /* case LEAF */
+
+	case UNARY:
+
+	    /*
+	     * A unary operator appearing just after something that's not an
+	     * operator is a syntax error -- something trying to be the left
+	     * operand of an operator that doesn't take one.
+	     */
+
+	    if (NotOperator(lastParsed)) {
+		msg = Tcl_ObjPrintf("missing operator at %s", mark);
+		scanned = 0;
+		insertMark = 1;
+		errCode = "MISSING";
+		goto error;
+	    }
+
+	    /*
+	     * Create an OpNode for the unary operator.
+	     */
+
+	    nodePtr->lexeme = lexeme;
+	    nodePtr->precedence = prec[lexeme];
+	    nodePtr->mark = MARK_RIGHT;
+
+	    /*
+	     * A FUNCTION cannot be a constant expression, because Tcl allows
+	     * functions to return variable results with the same arguments;
+	     * for example, rand(). Other unary operators can root a constant
+	     * expression, so long as the argument is a constant expression.
+	     */
+
+	    nodePtr->constant = (lexeme != FUNCTION);
+
+	    /*
+	     * This unary operator is a new incomplete tree, so push it onto
+	     * our stack of incomplete trees. Also remember it as the last
+	     * lexeme we parsed.
+	     */
+
+	    nodePtr->p.prev = incomplete;
+	    incomplete = lastParsed = nodesUsed;
+	    nodesUsed++;
+	    break;
+
+	case BINARY: {
+	    OpNode *incompletePtr;
+	    unsigned char precedence = prec[lexeme];
+
+	    /*
+	     * A binary operator appearing just after another operator is a
+	     * syntax error -- one of the two operators is missing an operand.
+	     */
+
+	    if (IsOperator(lastParsed)) {
+		if ((lexeme == CLOSE_PAREN)
+			&& (nodePtr[-1].lexeme == OPEN_PAREN)) {
+		    if (nodePtr[-2].lexeme == FUNCTION) {
+			/*
+			 * Normally, "()" is a syntax error, but as a special
+			 * case accept it as an argument list for a function.
+			 * Treat this as a special LEAF lexeme, and restart
+			 * the parsing loop with zero characters scanned. We
+			 * will parse the ")" again the next time through, but
+			 * with the OT_EMPTY leaf as the subexpression between
+			 * the parens.
+			 */
+
+			scanned = 0;
+			complete = lastParsed = OT_EMPTY;
+			break;
+		    }
+		    msg = Tcl_ObjPrintf("empty subexpression at %s", mark);
+		    scanned = 0;
+		    insertMark = 1;
+		    errCode = "EMPTY";
+		    goto error;
+		}
+
+		if (nodePtr[-1].precedence > precedence) {
+		    if (nodePtr[-1].lexeme == OPEN_PAREN) {
+			TclNewLiteralStringObj(msg, "unbalanced open paren");
+			parsePtr->errorType = TCL_PARSE_MISSING_PAREN;
+			errCode = "UNBALANCED";
+		    } else if (nodePtr[-1].lexeme == COMMA) {
+			msg = Tcl_ObjPrintf(
+				"missing function argument at %s", mark);
+			scanned = 0;
+			insertMark = 1;
+			errCode = "MISSING";
+		    } else if (nodePtr[-1].lexeme == START) {
+			TclNewLiteralStringObj(msg, "empty expression");
+			errCode = "EMPTY";
+		    }
+		} else if (lexeme == CLOSE_PAREN) {
+		    TclNewLiteralStringObj(msg, "unbalanced close paren");
+		    errCode = "UNBALANCED";
+		} else if ((lexeme == COMMA)
+			&& (nodePtr[-1].lexeme == OPEN_PAREN)
+			&& (nodePtr[-2].lexeme == FUNCTION)) {
+		    msg = Tcl_ObjPrintf("missing function argument at %s",
+			    mark);
+		    scanned = 0;
+		    insertMark = 1;
+		    errCode = "UNBALANCED";
+		}
+		if (msg == NULL) {
+		    msg = Tcl_ObjPrintf("missing operand at %s", mark);
+		    scanned = 0;
+		    insertMark = 1;
+		    errCode = "MISSING";
+		}
+		goto error;
+	    }
+
+	    /*
+	     * Here is where the tree comes together. At this point, we have a
+	     * stack of incomplete trees corresponding to substrings that are
+	     * incomplete expressions, followed by a complete tree
+	     * corresponding to a substring that is itself a complete
+	     * expression, followed by the binary operator we have just
+	     * parsed. The incomplete trees can each be completed by adding a
+	     * right operand.
+	     *
+	     * To illustrate with an example, when we parse the expression
+	     * "1+2*3-4" and we reach this point having just parsed the "-"
+	     * operator, we have these incomplete trees: START, "1+", and
+	     * "2*". Next we have the complete subexpression "3". Last is the
+	     * "-" we've just parsed.
+	     *
+	     * The next step is to join our complete tree to an operator. The
+	     * choice is governed by the precedence and associativity of the
+	     * competing operators. If we connect it as the right operand of
+	     * our most recent incomplete tree, we get a new complete tree,
+	     * and we can repeat the process. The while loop following repeats
+	     * this until precedence indicates it is time to join the complete
+	     * tree as the left operand of the just parsed binary operator.
+	     *
+	     * Continuing the example, the first pass through the loop will
+	     * join "3" to "2*"; the next pass will join "2*3" to "1+". Then
+	     * we'll exit the loop and join "1+2*3" to "-". When we return to
+	     * parse another lexeme, our stack of incomplete trees is START
+	     * and "1+2*3-".
+	     */
+
+	    while (1) {
+		incompletePtr = nodes + incomplete;
+
+		if (incompletePtr->precedence < precedence) {
+		    break;
+		}
+
+		if (incompletePtr->precedence == precedence) {
+		    /*
+		     * Right association rules for exponentiation.
+		     */
+
+		    if (lexeme == EXPON) {
+			break;
+		    }
+
+		    /*
+		     * Special association rules for the conditional
+		     * operators. The "?" and ":" operators have equal
+		     * precedence, but must be linked up in sensible pairs.
+		     */
+
+		    if ((incompletePtr->lexeme == QUESTION)
+			    && (NotOperator(complete)
+			    || (nodes[complete].lexeme != COLON))) {
+			break;
+		    }
+		    if ((incompletePtr->lexeme == COLON)
+			    && (lexeme == QUESTION)) {
+			break;
+		    }
+		}
+
+		/*
+		 * Some special syntax checks...
+		 */
+
+		/* Parens must balance */
+		if ((incompletePtr->lexeme == OPEN_PAREN)
+			&& (lexeme != CLOSE_PAREN)) {
+		    TclNewLiteralStringObj(msg, "unbalanced open paren");
+		    parsePtr->errorType = TCL_PARSE_MISSING_PAREN;
+		    errCode = "UNBALANCED";
+		    goto error;
+		}
+
+		/* Right operand of "?" must be ":" */
+		if ((incompletePtr->lexeme == QUESTION)
+			&& (NotOperator(complete)
+			|| (nodes[complete].lexeme != COLON))) {
+		    msg = Tcl_ObjPrintf("missing operator \":\" at %s", mark);
+		    scanned = 0;
+		    insertMark = 1;
+		    errCode = "MISSING";
+		    goto error;
+		}
+
+		/* Operator ":" may only be right operand of "?" */
+		if (IsOperator(complete)
+			&& (nodes[complete].lexeme == COLON)
+			&& (incompletePtr->lexeme != QUESTION)) {
+		    TclNewLiteralStringObj(msg,
+			    "unexpected operator \":\" "
+			    "without preceding \"?\"");
+		    errCode = "SURPRISE";
+		    goto error;
+		}
+
+		/*
+		 * Attach complete tree as right operand of most recent
+		 * incomplete tree.
+		 */
+
+		incompletePtr->right = complete;
+		if (IsOperator(complete)) {
+		    nodes[complete].p.parent = incomplete;
+		    incompletePtr->constant = incompletePtr->constant
+			    && nodes[complete].constant;
+		} else {
+		    incompletePtr->constant = incompletePtr->constant
+			    && (complete == OT_LITERAL);
+		}
+
+		/*
+		 * The QUESTION/COLON and FUNCTION/OPEN_PAREN combinations
+		 * each make up a single operator. Force them to agree whether
+		 * they have a constant expression.
+		 */
+
+		if ((incompletePtr->lexeme == QUESTION)
+			|| (incompletePtr->lexeme == FUNCTION)) {
+		    nodes[complete].constant = incompletePtr->constant;
+		}
+
+		if (incompletePtr->lexeme == START) {
+		    /*
+		     * Completing the START tree indicates we're done.
+		     * Transfer the parse tree to the caller and return.
+		     */
+
+		    *opTreePtr = nodes;
+		    return TCL_OK;
+		}
+
+		/*
+		 * With a right operand attached, last incomplete tree has
+		 * become the complete tree. Pop it from the incomplete tree
+		 * stack.
+		 */
+
+		complete = incomplete;
+		incomplete = incompletePtr->p.prev;
+
+		/* CLOSE_PAREN can only close one OPEN_PAREN. */
+		if (incompletePtr->lexeme == OPEN_PAREN) {
+		    break;
+		}
+	    }
+
+	    /*
+	     * More syntax checks...
+	     */
+
+	    /* Parens must balance. */
+	    if (lexeme == CLOSE_PAREN) {
+		if (incompletePtr->lexeme != OPEN_PAREN) {
+		    TclNewLiteralStringObj(msg, "unbalanced close paren");
+		    errCode = "UNBALANCED";
+		    goto error;
+		}
+	    }
+
+	    /* Commas must appear only in function argument lists. */
+	    if (lexeme == COMMA) {
+		if  ((incompletePtr->lexeme != OPEN_PAREN)
+			|| (incompletePtr[-1].lexeme != FUNCTION)) {
+		    TclNewLiteralStringObj(msg,
+			    "unexpected \",\" outside function argument list");
+		    errCode = "SURPRISE";
+		    goto error;
+		}
+	    }
+
+	    /* Operator ":" may only be right operand of "?" */
+	    if (IsOperator(complete) && (nodes[complete].lexeme == COLON)) {
+		TclNewLiteralStringObj(msg,
+			"unexpected operator \":\" without preceding \"?\"");
+		errCode = "SURPRISE";
+		goto error;
+	    }
+
+	    /*
+	     * Create no node for a CLOSE_PAREN lexeme.
+	     */
+
+	    if (lexeme == CLOSE_PAREN) {
+		break;
+	    }
+
+	    /*
+	     * Link complete tree as left operand of new node.
+	     */
+
+	    nodePtr->lexeme = lexeme;
+	    nodePtr->precedence = precedence;
+	    nodePtr->mark = MARK_LEFT;
+	    nodePtr->left = complete;
+
+	    /*
+	     * The COMMA operator cannot be optimized, since the function
+	     * needs all of its arguments, and optimization would reduce the
+	     * number. Other binary operators root constant expressions when
+	     * both arguments are constant expressions.
+	     */
+
+	    nodePtr->constant = (lexeme != COMMA);
+
+	    if (IsOperator(complete)) {
+		nodes[complete].p.parent = nodesUsed;
+		nodePtr->constant = nodePtr->constant
+			&& nodes[complete].constant;
+	    } else {
+		nodePtr->constant = nodePtr->constant
+			&& (complete == OT_LITERAL);
+	    }
+
+	    /*
+	     * With a left operand attached and a right operand missing, the
+	     * just-parsed binary operator is root of a new incomplete tree.
+	     * Push it onto the stack of incomplete trees.
+	     */
+
+	    nodePtr->p.prev = incomplete;
+	    incomplete = lastParsed = nodesUsed;
+	    nodesUsed++;
+	    break;
+	}	/* case BINARY */
+	}	/* lexeme handler */
+
+	/* Advance past the just-parsed lexeme */
+	start += scanned;
+	numBytes -= scanned;
+    }	/* main parsing loop */
+
+    /*
+     * We only get here if there's been an error. Any errors that didn't get a
+     * suitable parsePtr->errorType, get recorded as syntax errors.
+     */
+
+  error:
+    if (parsePtr->errorType == TCL_PARSE_SUCCESS) {
+	parsePtr->errorType = TCL_PARSE_SYNTAX;
+    }
+
+    /*
+     * Free any partial parse tree we've built.
+     */
+
+    if (nodes != NULL) {
+	ckfree(nodes);
+    }
+
+    if (interp == NULL) {
+	/*
+	 * Nowhere to report an error message, so just free it.
+	 */
+
+	if (msg) {
+	    Tcl_DecrRefCount(msg);
+	}
+    } else {
+	/*
+	 * Construct the complete error message. Start with the simple error
+	 * message, pulled from the interp result if necessary...
+	 */
+
+	if (msg == NULL) {
+	    msg = Tcl_GetObjResult(interp);
+	}
+
+	/*
+	 * Add a detailed quote from the bad expression, displaying and
+	 * sometimes marking the precise location of the syntax error.
+	 */
+
+	Tcl_AppendPrintfToObj(msg, "\nin expression \"%s%.*s%.*s%s%s%.*s%s\"",
+		((start - limit) < parsePtr->string) ? "" : "...",
+		((start - limit) < parsePtr->string)
+			? (int) (start - parsePtr->string) : limit - 3,
+		((start - limit) < parsePtr->string)
+			? parsePtr->string : start - limit + 3,
+		(scanned < limit) ? scanned : limit - 3, start,
+		(scanned < limit) ? "" : "...", insertMark ? mark : "",
+		(start + scanned + limit > parsePtr->end)
+			? (int) (parsePtr->end - start) - scanned : limit-3,
+		start + scanned,
+		(start + scanned + limit > parsePtr->end) ? "" : "...");
+
+	/*
+	 * Next, append any postscript message.
+	 */
+
+	if (post != NULL) {
+	    Tcl_AppendToObj(msg, ";\n", -1);
+	    Tcl_AppendObjToObj(msg, post);
+	    Tcl_DecrRefCount(post);
+	}
+	Tcl_SetObjResult(interp, msg);
+
+	/*
+	 * Finally, place context information in the errorInfo.
+	 */
+
+	numBytes = parsePtr->end - parsePtr->string;
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (parsing expression \"%.*s%s\")",
+		(numBytes < limit) ? numBytes : limit - 3,
+		parsePtr->string, (numBytes < limit) ? "" : "..."));
+	if (errCode) {
+	    Tcl_SetErrorCode(interp, "TCL", "PARSE", "EXPR", errCode,
+		    subErrCode, NULL);
+	}
+    }
+
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertTreeToTokens --
+ *
+ *	Given a string, the numBytes bytes starting at start, and an OpNode
+ *	tree and Tcl_Token array created by passing that same string to
+ *	ParseExpr(), this function writes into *parsePtr the sequence of
+ *	Tcl_Tokens needed so to satisfy the historical interface provided by
+ *	Tcl_ParseExpr(). Note that this routine exists only for the sake of
+ *	the public Tcl_ParseExpr() routine. It is not used by Tcl itself at
+ *	all.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The Tcl_Parse *parsePtr is filled with Tcl_Tokens representing the
+ *	parsed expression.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ConvertTreeToTokens(
+    const char *start,
+    int numBytes,
+    OpNode *nodes,
+    Tcl_Token *tokenPtr,
+    Tcl_Parse *parsePtr)
+{
+    int subExprTokenIdx = 0;
+    OpNode *nodePtr = nodes;
+    int next = nodePtr->right;
+
+    while (1) {
+	Tcl_Token *subExprTokenPtr;
+	int scanned, parentIdx;
+	unsigned char lexeme;
+
+	/*
+	 * Advance the mark so the next exit from this node won't retrace
+	 * steps over ground already covered.
+	 */
+
+	nodePtr->mark++;
+
+	/*
+	 * Handle next child node or leaf.
+	 */
+
+	switch (next) {
+	case OT_EMPTY:
+
+	    /* No tokens and no characters for the OT_EMPTY leaf. */
+	    break;
+
+	case OT_LITERAL:
+
+	    /*
+	     * Skip any white space that comes before the literal.
+	     */
+
+	    scanned = TclParseAllWhiteSpace(start, numBytes);
+	    start += scanned;
+	    numBytes -= scanned;
+
+	    /*
+	     * Reparse the literal to get pointers into source string.
+	     */
+
+	    scanned = ParseLexeme(start, numBytes, &lexeme, NULL);
+
+	    TclGrowParseTokenArray(parsePtr, 2);
+	    subExprTokenPtr = parsePtr->tokenPtr + parsePtr->numTokens;
+	    subExprTokenPtr->type = TCL_TOKEN_SUB_EXPR;
+	    subExprTokenPtr->start = start;
+	    subExprTokenPtr->size = scanned;
+	    subExprTokenPtr->numComponents = 1;
+	    subExprTokenPtr[1].type = TCL_TOKEN_TEXT;
+	    subExprTokenPtr[1].start = start;
+	    subExprTokenPtr[1].size = scanned;
+	    subExprTokenPtr[1].numComponents = 0;
+
+	    parsePtr->numTokens += 2;
+	    start += scanned;
+	    numBytes -= scanned;
+	    break;
+
+	case OT_TOKENS: {
+	    /*
+	     * tokenPtr points to a token sequence that came from parsing a
+	     * Tcl word. A Tcl word is made up of a sequence of one or more
+	     * elements. When the word is only a single element, it's been the
+	     * historical practice to replace the TCL_TOKEN_WORD token
+	     * directly with a TCL_TOKEN_SUB_EXPR token. However, when the
+	     * word has multiple elements, a TCL_TOKEN_WORD token is kept as a
+	     * grouping device so that TCL_TOKEN_SUB_EXPR always has only one
+	     * element. Wise or not, these are the rules the Tcl expr parser
+	     * has followed, and for the sake of those few callers of
+	     * Tcl_ParseExpr() we do not change them now. Internally, we can
+	     * do better.
+	     */
+
+	    int toCopy = tokenPtr->numComponents + 1;
+
+	    if (tokenPtr->numComponents == tokenPtr[1].numComponents + 1) {
+		/*
+		 * Single element word. Copy tokens and convert the leading
+		 * token to TCL_TOKEN_SUB_EXPR.
+		 */
+
+		TclGrowParseTokenArray(parsePtr, toCopy);
+		subExprTokenPtr = parsePtr->tokenPtr + parsePtr->numTokens;
+		memcpy(subExprTokenPtr, tokenPtr,
+			(size_t) toCopy * sizeof(Tcl_Token));
+		subExprTokenPtr->type = TCL_TOKEN_SUB_EXPR;
+		parsePtr->numTokens += toCopy;
+	    } else {
+		/*
+		 * Multiple element word. Create a TCL_TOKEN_SUB_EXPR token to
+		 * lead, with fields initialized from the leading token, then
+		 * copy entire set of word tokens.
+		 */
+
+		TclGrowParseTokenArray(parsePtr, toCopy+1);
+		subExprTokenPtr = parsePtr->tokenPtr + parsePtr->numTokens;
+		*subExprTokenPtr = *tokenPtr;
+		subExprTokenPtr->type = TCL_TOKEN_SUB_EXPR;
+		subExprTokenPtr->numComponents++;
+		subExprTokenPtr++;
+		memcpy(subExprTokenPtr, tokenPtr,
+			(size_t) toCopy * sizeof(Tcl_Token));
+		parsePtr->numTokens += toCopy + 1;
+	    }
+
+	    scanned = tokenPtr->start + tokenPtr->size - start;
+	    start += scanned;
+	    numBytes -= scanned;
+	    tokenPtr += toCopy;
+	    break;
+	}
+
+	default:
+
+	    /*
+	     * Advance to the child node, which is an operator.
+	     */
+
+	    nodePtr = nodes + next;
+
+	    /*
+	     * Skip any white space that comes before the subexpression.
+	     */
+
+	    scanned = TclParseAllWhiteSpace(start, numBytes);
+	    start += scanned;
+	    numBytes -= scanned;
+
+	    /*
+	     * Generate tokens for the operator / subexpression...
+	     */
+
+	    switch (nodePtr->lexeme) {
+	    case OPEN_PAREN:
+	    case COMMA:
+	    case COLON:
+
+		/*
+		 * Historical practice has been to have no Tcl_Tokens for
+		 * these operators.
+		 */
+
+		break;
+
+	    default: {
+
+		/*
+		 * Remember the index of the last subexpression we were
+		 * working on -- that of our parent. We'll stack it later.
+		 */
+
+		parentIdx = subExprTokenIdx;
+
+		/*
+		 * Verify space for the two leading Tcl_Tokens representing
+		 * the subexpression rooted by this operator. The first
+		 * Tcl_Token will be of type TCL_TOKEN_SUB_EXPR; the second of
+		 * type TCL_TOKEN_OPERATOR.
+		 */
+
+		TclGrowParseTokenArray(parsePtr, 2);
+		subExprTokenIdx = parsePtr->numTokens;
+		subExprTokenPtr = parsePtr->tokenPtr + subExprTokenIdx;
+		parsePtr->numTokens += 2;
+		subExprTokenPtr->type = TCL_TOKEN_SUB_EXPR;
+		subExprTokenPtr[1].type = TCL_TOKEN_OPERATOR;
+
+		/*
+		 * Our current position scanning the string is the starting
+		 * point for this subexpression.
+		 */
+
+		subExprTokenPtr->start = start;
+
+		/*
+		 * Eventually, we know that the numComponents field of the
+		 * Tcl_Token of type TCL_TOKEN_OPERATOR will be 0. This means
+		 * we can make other use of this field for now to track the
+		 * stack of subexpressions we have pending.
+		 */
+
+		subExprTokenPtr[1].numComponents = parentIdx;
+		break;
+	    }
+	    }
+	    break;
+	}
+
+	/* Determine which way to exit the node on this pass. */
+    router:
+	switch (nodePtr->mark) {
+	case MARK_LEFT:
+	    next = nodePtr->left;
+	    break;
+
+	case MARK_RIGHT:
+	    next = nodePtr->right;
+
+	    /*
+	     * Skip any white space that comes before the operator.
+	     */
+
+	    scanned = TclParseAllWhiteSpace(start, numBytes);
+	    start += scanned;
+	    numBytes -= scanned;
+
+	    /*
+	     * Here we scan from the string the operator corresponding to
+	     * nodePtr->lexeme.
+	     */
+
+	    scanned = ParseLexeme(start, numBytes, &lexeme, NULL);
+
+	    switch(nodePtr->lexeme) {
+	    case OPEN_PAREN:
+	    case COMMA:
+	    case COLON:
+
+		/*
+		 * No tokens for these lexemes -> nothing to do.
+		 */
+
+		break;
+
+	    default:
+
+		/*
+		 * Record in the TCL_TOKEN_OPERATOR token the pointers into
+		 * the string marking where the operator is.
+		 */
+
+		subExprTokenPtr = parsePtr->tokenPtr + subExprTokenIdx;
+		subExprTokenPtr[1].start = start;
+		subExprTokenPtr[1].size = scanned;
+		break;
+	    }
+
+	    start += scanned;
+	    numBytes -= scanned;
+	    break;
+
+	case MARK_PARENT:
+	    switch (nodePtr->lexeme) {
+	    case START:
+
+		/* When we get back to the START node, we're done. */
+		return;
+
+	    case COMMA:
+	    case COLON:
+
+		/* No tokens for these lexemes -> nothing to do. */
+		break;
+
+	    case OPEN_PAREN:
+
+		/*
+		 * Skip past matching close paren.
+		 */
+
+		scanned = TclParseAllWhiteSpace(start, numBytes);
+		start += scanned;
+		numBytes -= scanned;
+		scanned = ParseLexeme(start, numBytes, &lexeme, NULL);
+		start += scanned;
+		numBytes -= scanned;
+		break;
+
+	    default:
+
+		/*
+		 * Before we leave this node/operator/subexpression for the
+		 * last time, finish up its tokens....
+		 *
+		 * Our current position scanning the string is where the
+		 * substring for the subexpression ends.
+		 */
+
+		subExprTokenPtr = parsePtr->tokenPtr + subExprTokenIdx;
+		subExprTokenPtr->size = start - subExprTokenPtr->start;
+
+		/*
+		 * All the Tcl_Tokens allocated and filled belong to
+		 * this subexpresion. The first token is the leading
+		 * TCL_TOKEN_SUB_EXPR token, and all the rest (one fewer)
+		 * are its components.
+		 */
+
+		subExprTokenPtr->numComponents =
+			(parsePtr->numTokens - subExprTokenIdx) - 1;
+
+		/*
+		 * Finally, as we return up the tree to our parent, pop the
+		 * parent subexpression off our subexpression stack, and
+		 * fill in the zero numComponents for the operator Tcl_Token.
+		 */
+
+		parentIdx = subExprTokenPtr[1].numComponents;
+		subExprTokenPtr[1].numComponents = 0;
+		subExprTokenIdx = parentIdx;
+		break;
+	    }
+
+	    /*
+	     * Since we're returning to parent, skip child handling code.
+	     */
+
+	    nodePtr = nodes + nodePtr->p.parent;
+	    goto router;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseExpr --
+ *
+ *	Given a string, the numBytes bytes starting at start, this function
+ *	parses it as a Tcl expression and stores information about the
+ *	structure of the expression in the Tcl_Parse struct indicated by the
+ *	caller.
+ *
+ * Results:
+ *	If the string is successfully parsed as a valid Tcl expression, TCL_OK
+ *	is returned, and data about the expression structure is written to
+ *	*parsePtr. If the string cannot be parsed as a valid Tcl expression,
+ *	TCL_ERROR is returned, and if interp is non-NULL, an error message is
+ *	written to interp.
+ *
+ * Side effects:
+ *	If there is insufficient space in parsePtr to hold all the information
+ *	about the expression, then additional space is malloc-ed. If the
+ *	function returns TCL_OK then the caller must eventually invoke
+ *	Tcl_FreeParse to release any additional space that was allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseExpr(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    const char *start,		/* Start of source string to parse. */
+    int numBytes,		/* Number of bytes in string. If < 0, the
+				 * string consists of all bytes up to the
+				 * first null character. */
+    Tcl_Parse *parsePtr)	/* Structure to fill with information about
+				 * the parsed expression; any previous
+				 * information in the structure is ignored. */
+{
+    int code;
+    OpNode *opTree = NULL;	/* Will point to the tree of operators. */
+    Tcl_Obj *litList = Tcl_NewObj();	/* List to hold the literals. */
+    Tcl_Obj *funcList = Tcl_NewObj();	/* List to hold the functon names. */
+    Tcl_Parse *exprParsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse));
+				/* Holds the Tcl_Tokens of substitutions. */
+
+    if (numBytes < 0) {
+	numBytes = (start ? strlen(start) : 0);
+    }
+
+    code = ParseExpr(interp, start, numBytes, &opTree, litList, funcList,
+	    exprParsePtr, 1 /* parseOnly */);
+    Tcl_DecrRefCount(funcList);
+    Tcl_DecrRefCount(litList);
+
+    TclParseInit(interp, start, numBytes, parsePtr);
+    if (code == TCL_OK) {
+	ConvertTreeToTokens(start, numBytes,
+		opTree, exprParsePtr->tokenPtr, parsePtr);
+    } else {
+	parsePtr->term = exprParsePtr->term;
+	parsePtr->errorType = exprParsePtr->errorType;
+    }
+
+    Tcl_FreeParse(exprParsePtr);
+    TclStackFree(interp, exprParsePtr);
+    ckfree(opTree);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseLexeme --
+ *
+ *	Parse a single lexeme from the start of a string, scanning no more
+ *	than numBytes bytes.
+ *
+ * Results:
+ *	Returns the number of bytes scanned to produce the lexeme.
+ *
+ * Side effects:
+ *	Code identifying lexeme parsed is writen to *lexemePtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseLexeme(
+    const char *start,		/* Start of lexeme to parse. */
+    int numBytes,		/* Number of bytes in string. */
+    unsigned char *lexemePtr,	/* Write code of parsed lexeme to this
+				 * storage. */
+    Tcl_Obj **literalPtr)	/* Write corresponding literal value to this
+				   storage, if non-NULL. */
+{
+    const char *end;
+    int scanned;
+    Tcl_UniChar ch = 0;
+    Tcl_Obj *literal = NULL;
+    unsigned char byte;
+
+    if (numBytes == 0) {
+	*lexemePtr = END;
+	return 0;
+    }
+    byte = UCHAR(*start);
+    if (byte < sizeof(Lexeme) && Lexeme[byte] != 0) {
+	*lexemePtr = Lexeme[byte];
+	return 1;
+    }
+    switch (byte) {
+    case '*':
+	if ((numBytes > 1) && (start[1] == '*')) {
+	    *lexemePtr = EXPON;
+	    return 2;
+	}
+	*lexemePtr = MULT;
+	return 1;
+
+    case '=':
+	if ((numBytes > 1) && (start[1] == '=')) {
+	    *lexemePtr = EQUAL;
+	    return 2;
+	}
+	*lexemePtr = INCOMPLETE;
+	return 1;
+
+    case '!':
+	if ((numBytes > 1) && (start[1] == '=')) {
+	    *lexemePtr = NEQ;
+	    return 2;
+	}
+	*lexemePtr = NOT;
+	return 1;
+
+    case '&':
+	if ((numBytes > 1) && (start[1] == '&')) {
+	    *lexemePtr = AND;
+	    return 2;
+	}
+	*lexemePtr = BIT_AND;
+	return 1;
+
+    case '|':
+	if ((numBytes > 1) && (start[1] == '|')) {
+	    *lexemePtr = OR;
+	    return 2;
+	}
+	*lexemePtr = BIT_OR;
+	return 1;
+
+    case '<':
+	if (numBytes > 1) {
+	    switch (start[1]) {
+	    case '<':
+		*lexemePtr = LEFT_SHIFT;
+		return 2;
+	    case '=':
+		*lexemePtr = LEQ;
+		return 2;
+	    }
+	}
+	*lexemePtr = LESS;
+	return 1;
+
+    case '>':
+	if (numBytes > 1) {
+	    switch (start[1]) {
+	    case '>':
+		*lexemePtr = RIGHT_SHIFT;
+		return 2;
+	    case '=':
+		*lexemePtr = GEQ;
+		return 2;
+	    }
+	}
+	*lexemePtr = GREATER;
+	return 1;
+
+    case 'i':
+	if ((numBytes > 1) && (start[1] == 'n')
+		&& ((numBytes == 2) || start[2] & 0x80 || !isalpha(UCHAR(start[2])))) {
+	    /*
+	     * Must make this check so we can tell the difference between the
+	     * "in" operator and the "int" function name and the "infinity"
+	     * numeric value.
+	     */
+
+	    *lexemePtr = IN_LIST;
+	    return 2;
+	}
+	break;
+
+    case 'e':
+	if ((numBytes > 1) && (start[1] == 'q')
+		&& ((numBytes == 2) || start[2] & 0x80 || !isalpha(UCHAR(start[2])))) {
+	    *lexemePtr = STREQ;
+	    return 2;
+	}
+	break;
+
+    case 'n':
+	if ((numBytes > 1)
+		&& ((numBytes == 2) || start[2] & 0x80 || !isalpha(UCHAR(start[2])))) {
+	    switch (start[1]) {
+	    case 'e':
+		*lexemePtr = STRNEQ;
+		return 2;
+	    case 'i':
+		*lexemePtr = NOT_IN_LIST;
+		return 2;
+	    }
+	}
+    }
+
+    literal = Tcl_NewObj();
+    if (TclParseNumber(NULL, literal, NULL, start, numBytes, &end,
+	    TCL_PARSE_NO_WHITESPACE) == TCL_OK) {
+	if (end < start + numBytes && !TclIsBareword(*end)) {
+
+	number:
+	    TclInitStringRep(literal, start, end-start);
+	    *lexemePtr = NUMBER;
+	    if (literalPtr) {
+		*literalPtr = literal;
+	    } else {
+		Tcl_DecrRefCount(literal);
+	    }
+	    return (end-start);
+	} else {
+	    unsigned char lexeme;
+
+	    /*
+	     * We have a number followed directly by bareword characters
+	     * (alpha, digit, underscore).  Is this a number followed by
+	     * bareword syntax error?  Or should we join into one bareword?
+	     * Example: Inf + luence + () becomes a valid function call.
+	     * [Bug 3401704]
+	     */
+	    if (literal->typePtr == &tclDoubleType) {
+		const char *p = start;
+
+		while (p < end) {
+		    if (!TclIsBareword(*p++)) {
+			/*
+			 * The number has non-bareword characters, so we
+			 * must treat it as a number.
+			 */
+			goto number;
+		    }
+		}
+	    }
+	    ParseLexeme(end, numBytes-(end-start), &lexeme, NULL);
+	    if ((NODE_TYPE & lexeme) == BINARY) {
+		/*
+		 * The bareword characters following the number take the
+		 * form of an operator (eq, ne, in, ni, ...) so we treat
+		 * as number + operator.
+		 */
+		goto number;
+	    }
+
+	    /*
+	     * Otherwise, fall through and parse the whole as a bareword.
+	     */
+	}
+    }
+
+    /*
+     * We reject leading underscores in bareword.  No sensible reason why.
+     * Might be inspired by reserved identifier rules in C, which of course
+     * have no direct relevance here.
+     */
+
+    if (!TclIsBareword(*start) || *start == '_') {
+	if (Tcl_UtfCharComplete(start, numBytes)) {
+	    scanned = TclUtfToUniChar(start, &ch);
+	} else {
+	    char utfBytes[TCL_UTF_MAX];
+
+	    memcpy(utfBytes, start, (size_t) numBytes);
+	    utfBytes[numBytes] = '\0';
+	    scanned = TclUtfToUniChar(utfBytes, &ch);
+	}
+	*lexemePtr = INVALID;
+	Tcl_DecrRefCount(literal);
+	return scanned;
+    }
+    end = start;
+    while (numBytes && TclIsBareword(*end)) {
+	end += 1;
+	numBytes -= 1;
+    }
+    *lexemePtr = BAREWORD;
+    if (literalPtr) {
+	Tcl_SetStringObj(literal, start, (int) (end-start));
+	*literalPtr = literal;
+    } else {
+	Tcl_DecrRefCount(literal);
+    }
+    return (end-start);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileExpr --
+ *
+ *	This procedure compiles a string containing a Tcl expression into Tcl
+ *	bytecodes.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Adds instructions to envPtr to evaluate the expression at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCompileExpr(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    const char *script,		/* The source script to compile. */
+    int numBytes,		/* Number of bytes in script. */
+    CompileEnv *envPtr,		/* Holds resulting instructions. */
+    int optimize)		/* 0 for one-off expressions. */
+{
+    OpNode *opTree = NULL;	/* Will point to the tree of operators */
+    Tcl_Obj *litList = Tcl_NewObj();	/* List to hold the literals */
+    Tcl_Obj *funcList = Tcl_NewObj();	/* List to hold the functon names*/
+    Tcl_Parse *parsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse));
+				/* Holds the Tcl_Tokens of substitutions */
+
+    int code = ParseExpr(interp, script, numBytes, &opTree, litList,
+	    funcList, parsePtr, 0 /* parseOnly */);
+
+    if (code == TCL_OK) {
+	/*
+	 * Valid parse; compile the tree.
+	 */
+
+	int objc;
+	Tcl_Obj *const *litObjv;
+	Tcl_Obj **funcObjv;
+
+	/* TIP #280 : Track Lines within the expression */
+	TclAdvanceLines(&envPtr->line, script,
+		script + TclParseAllWhiteSpace(script, numBytes));
+
+	TclListObjGetElements(NULL, litList, &objc, (Tcl_Obj ***)&litObjv);
+	TclListObjGetElements(NULL, funcList, &objc, &funcObjv);
+	CompileExprTree(interp, opTree, 0, &litObjv, funcObjv,
+		parsePtr->tokenPtr, envPtr, optimize);
+    } else {
+	TclCompileSyntaxError(interp, envPtr);
+    }
+
+    Tcl_FreeParse(parsePtr);
+    TclStackFree(interp, parsePtr);
+    Tcl_DecrRefCount(funcList);
+    Tcl_DecrRefCount(litList);
+    ckfree(opTree);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExecConstantExprTree --
+ *	Compiles and executes bytecode for the subexpression tree at index
+ *	in the nodes array.  This subexpression must be constant, made up
+ *	of only constant operators (not functions) and literals.
+ *
+ * Results:
+ *	A standard Tcl return code and result left in interp.
+ *
+ * Side effects:
+ *	Consumes subtree of nodes rooted at index.  Advances the pointer
+ *	*litObjvPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ExecConstantExprTree(
+    Tcl_Interp *interp,
+    OpNode *nodes,
+    int index,
+    Tcl_Obj *const **litObjvPtr)
+{
+    CompileEnv *envPtr;
+    ByteCode *byteCodePtr;
+    int code;
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    /*
+     * Note we are compiling an expression with literal arguments. This means
+     * there can be no [info frame] calls when we execute the resulting
+     * bytecode, so there's no need to tend to TIP 280 issues.
+     */
+
+    envPtr = TclStackAlloc(interp, sizeof(CompileEnv));
+    TclInitCompileEnv(interp, envPtr, NULL, 0, NULL, 0);
+    CompileExprTree(interp, nodes, index, litObjvPtr, NULL, NULL, envPtr,
+	    0 /* optimize */);
+    TclEmitOpcode(INST_DONE, envPtr);
+    byteCodePtr = TclInitByteCode(envPtr);
+    TclFreeCompileEnv(envPtr);
+    TclStackFree(interp, envPtr);
+    TclNRExecuteByteCode(interp, byteCodePtr);
+    code = TclNRRunCallbacks(interp, TCL_OK, rootPtr);
+    TclReleaseByteCode(byteCodePtr);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileExprTree --
+ *
+ *	Compiles and writes to envPtr instructions for the subexpression tree
+ *	at index in the nodes array. (*litObjvPtr) must point to the proper
+ *	location in a corresponding literals list. Likewise, when non-NULL,
+ *	funcObjv and tokenPtr must point into matching arrays of function
+ *	names and Tcl_Token's derived from earlier call to ParseExpr(). When
+ *	optimize is true, any constant subexpressions will be precomputed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Adds instructions to envPtr to evaluate the expression at runtime.
+ *	Consumes subtree of nodes rooted at index. Advances the pointer
+ *	*litObjvPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CompileExprTree(
+    Tcl_Interp *interp,
+    OpNode *nodes,
+    int index,
+    Tcl_Obj *const **litObjvPtr,
+    Tcl_Obj *const *funcObjv,
+    Tcl_Token *tokenPtr,
+    CompileEnv *envPtr,
+    int optimize)
+{
+    OpNode *nodePtr = nodes + index;
+    OpNode *rootPtr = nodePtr;
+    int numWords = 0;
+    JumpList *jumpPtr = NULL;
+    int convert = 1;
+
+    while (1) {
+	int next;
+	JumpList *freePtr, *newJump;
+
+	if (nodePtr->mark == MARK_LEFT) {
+	    next = nodePtr->left;
+
+	    if (nodePtr->lexeme == QUESTION) {
+		convert = 1;
+	    }
+	} else if (nodePtr->mark == MARK_RIGHT) {
+	    next = nodePtr->right;
+
+	    switch (nodePtr->lexeme) {
+	    case FUNCTION: {
+		Tcl_DString cmdName;
+		const char *p;
+		int length;
+
+		Tcl_DStringInit(&cmdName);
+		TclDStringAppendLiteral(&cmdName, "tcl::mathfunc::");
+		p = TclGetStringFromObj(*funcObjv, &length);
+		funcObjv++;
+		Tcl_DStringAppend(&cmdName, p, length);
+		TclEmitPush(TclRegisterLiteral(envPtr,
+			Tcl_DStringValue(&cmdName),
+			Tcl_DStringLength(&cmdName), LITERAL_CMD_NAME), envPtr);
+		Tcl_DStringFree(&cmdName);
+
+		/*
+		 * Start a count of the number of words in this function
+		 * command invocation. In case there's already a count in
+		 * progress (nested functions), save it in our unused "left"
+		 * field for restoring later.
+		 */
+
+		nodePtr->left = numWords;
+		numWords = 2;	/* Command plus one argument */
+		break;
+	    }
+	    case QUESTION:
+		newJump = TclStackAlloc(interp, sizeof(JumpList));
+		newJump->next = jumpPtr;
+		jumpPtr = newJump;
+		TclEmitForwardJump(envPtr, TCL_FALSE_JUMP, &jumpPtr->jump);
+		break;
+	    case COLON:
+		newJump = TclStackAlloc(interp, sizeof(JumpList));
+		newJump->next = jumpPtr;
+		jumpPtr = newJump;
+		TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP,
+			&jumpPtr->jump);
+		TclAdjustStackDepth(-1, envPtr);
+		if (convert) {
+		    jumpPtr->jump.jumpType = TCL_TRUE_JUMP;
+		}
+		convert = 1;
+		break;
+	    case AND:
+	    case OR:
+		newJump = TclStackAlloc(interp, sizeof(JumpList));
+		newJump->next = jumpPtr;
+		jumpPtr = newJump;
+		TclEmitForwardJump(envPtr, (nodePtr->lexeme == AND)
+			?  TCL_FALSE_JUMP : TCL_TRUE_JUMP, &jumpPtr->jump);
+		break;
+	    }
+	} else {
+	    int pc1, pc2, target;
+
+	    switch (nodePtr->lexeme) {
+	    case START:
+	    case QUESTION:
+		if (convert && (nodePtr == rootPtr)) {
+		    TclEmitOpcode(INST_TRY_CVT_TO_NUMERIC, envPtr);
+		}
+		break;
+	    case OPEN_PAREN:
+
+		/* do nothing */
+		break;
+	    case FUNCTION:
+		/*
+		 * Use the numWords count we've kept to invoke the function
+		 * command with the correct number of arguments.
+		 */
+
+		if (numWords < 255) {
+		    TclEmitInvoke(envPtr, INST_INVOKE_STK1, numWords);
+		} else {
+		    TclEmitInvoke(envPtr, INST_INVOKE_STK4, numWords);
+		}
+
+		/*
+		 * Restore any saved numWords value.
+		 */
+
+		numWords = nodePtr->left;
+		convert = 1;
+		break;
+	    case COMMA:
+		/*
+		 * Each comma implies another function argument.
+		 */
+
+		numWords++;
+		break;
+	    case COLON:
+		CLANG_ASSERT(jumpPtr);
+		if (jumpPtr->jump.jumpType == TCL_TRUE_JUMP) {
+		    jumpPtr->jump.jumpType = TCL_UNCONDITIONAL_JUMP;
+		    convert = 1;
+		}
+		target = jumpPtr->jump.codeOffset + 2;
+		if (TclFixupForwardJumpToHere(envPtr, &jumpPtr->jump, 127)) {
+		    target += 3;
+		}
+		freePtr = jumpPtr;
+		jumpPtr = jumpPtr->next;
+		TclStackFree(interp, freePtr);
+		TclFixupForwardJump(envPtr, &jumpPtr->jump,
+			target - jumpPtr->jump.codeOffset, 127);
+
+		freePtr = jumpPtr;
+		jumpPtr = jumpPtr->next;
+		TclStackFree(interp, freePtr);
+		break;
+	    case AND:
+	    case OR:
+		CLANG_ASSERT(jumpPtr);
+		pc1 = CurrentOffset(envPtr);
+		TclEmitInstInt1((nodePtr->lexeme == AND) ? INST_JUMP_FALSE1
+			: INST_JUMP_TRUE1, 0, envPtr);
+		TclEmitPush(TclRegisterLiteral(envPtr,
+			(nodePtr->lexeme == AND) ? "1" : "0", 1, 0), envPtr);
+		pc2 = CurrentOffset(envPtr);
+		TclEmitInstInt1(INST_JUMP1, 0, envPtr);
+		TclAdjustStackDepth(-1, envPtr);
+		TclStoreInt1AtPtr(CurrentOffset(envPtr) - pc1,
+			envPtr->codeStart + pc1 + 1);
+		if (TclFixupForwardJumpToHere(envPtr, &jumpPtr->jump, 127)) {
+		    pc2 += 3;
+		}
+		TclEmitPush(TclRegisterLiteral(envPtr,
+			(nodePtr->lexeme == AND) ? "0" : "1", 1, 0), envPtr);
+		TclStoreInt1AtPtr(CurrentOffset(envPtr) - pc2,
+			envPtr->codeStart + pc2 + 1);
+		convert = 0;
+		freePtr = jumpPtr;
+		jumpPtr = jumpPtr->next;
+		TclStackFree(interp, freePtr);
+		break;
+	    default:
+		TclEmitOpcode(instruction[nodePtr->lexeme], envPtr);
+		convert = 0;
+		break;
+	    }
+	    if (nodePtr == rootPtr) {
+		/* We're done */
+
+		return;
+	    }
+	    nodePtr = nodes + nodePtr->p.parent;
+	    continue;
+	}
+
+	nodePtr->mark++;
+	switch (next) {
+	case OT_EMPTY:
+	    numWords = 1;	/* No arguments, so just the command */
+	    break;
+	case OT_LITERAL: {
+	    Tcl_Obj *const *litObjv = *litObjvPtr;
+	    Tcl_Obj *literal = *litObjv;
+
+	    if (optimize) {
+		int length;
+		const char *bytes = TclGetStringFromObj(literal, &length);
+		int index = TclRegisterLiteral(envPtr, bytes, length, 0);
+		Tcl_Obj *objPtr = TclFetchLiteral(envPtr, index);
+
+		if ((objPtr->typePtr == NULL) && (literal->typePtr != NULL)) {
+		    /*
+		     * Would like to do this:
+		     *
+		     * lePtr->objPtr = literal;
+		     * Tcl_IncrRefCount(literal);
+		     * Tcl_DecrRefCount(objPtr);
+		     *
+		     * However, the design of the "global" and "local"
+		     * LiteralTable does not permit the value of lePtr->objPtr
+		     * to change. So rather than replace lePtr->objPtr, we do
+		     * surgery to transfer our desired intrep into it.
+		     */
+
+		    objPtr->typePtr = literal->typePtr;
+		    objPtr->internalRep = literal->internalRep;
+		    literal->typePtr = NULL;
+		}
+		TclEmitPush(index, envPtr);
+	    } else {
+		/*
+		 * When optimize==0, we know the expression is a one-off and
+		 * there's nothing to be gained from sharing literals when
+		 * they won't live long, and the copies we have already have
+		 * an appropriate intrep. In this case, skip literal
+		 * registration that would enable sharing, and use the routine
+		 * that preserves intreps.
+		 */
+
+		TclEmitPush(TclAddLiteralObj(envPtr, literal, NULL), envPtr);
+	    }
+	    (*litObjvPtr)++;
+	    break;
+	}
+	case OT_TOKENS:
+	    CompileTokens(envPtr, tokenPtr, interp);
+	    tokenPtr += tokenPtr->numComponents + 1;
+	    break;
+	default:
+	    if (optimize && nodes[next].constant) {
+		Tcl_InterpState save = Tcl_SaveInterpState(interp, TCL_OK);
+
+		if (ExecConstantExprTree(interp, nodes, next, litObjvPtr)
+			== TCL_OK) {
+		    int index;
+		    Tcl_Obj *objPtr = Tcl_GetObjResult(interp);
+
+		    /*
+		     * Don't generate a string rep, but if we have one
+		     * already, then use it to share via the literal table.
+		     */
+
+		    if (objPtr->bytes) {
+			Tcl_Obj *tableValue;
+
+			index = TclRegisterLiteral(envPtr, objPtr->bytes,
+				objPtr->length, 0);
+			tableValue = TclFetchLiteral(envPtr, index);
+			if ((tableValue->typePtr == NULL) &&
+				(objPtr->typePtr != NULL)) {
+			    /*
+			     * Same intrep surgery as for OT_LITERAL.
+			     */
+
+			    tableValue->typePtr = objPtr->typePtr;
+			    tableValue->internalRep = objPtr->internalRep;
+			    objPtr->typePtr = NULL;
+			}
+		    } else {
+			index = TclAddLiteralObj(envPtr, objPtr, NULL);
+		    }
+		    TclEmitPush(index, envPtr);
+		} else {
+		    TclCompileSyntaxError(interp, envPtr);
+		}
+		Tcl_RestoreInterpState(interp, save);
+		convert = 0;
+	    } else {
+		nodePtr = nodes + next;
+	    }
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSingleOpCmd --
+ *
+ *	Implements the commands: ~, !, <<, >>, %, !=, ne, in, ni
+ *	in the ::tcl::mathop namespace.  These commands have no
+ *	extension to arbitrary arguments; they accept only exactly one
+ *	or exactly two arguments as suitable for the operator.
+ *
+ * Results:
+ *	A standard Tcl return code and result left in interp.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSingleOpCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    TclOpCmdClientData *occdPtr = clientData;
+    unsigned char lexeme;
+    OpNode nodes[2];
+    Tcl_Obj *const *litObjv = objv + 1;
+
+    if (objc != 1 + occdPtr->i.numArgs) {
+	Tcl_WrongNumArgs(interp, 1, objv, occdPtr->expected);
+	return TCL_ERROR;
+    }
+
+    ParseLexeme(occdPtr->op, strlen(occdPtr->op), &lexeme, NULL);
+    nodes[0].lexeme = START;
+    nodes[0].mark = MARK_RIGHT;
+    nodes[0].right = 1;
+    nodes[1].lexeme = lexeme;
+    if (objc == 2) {
+	nodes[1].mark = MARK_RIGHT;
+    } else {
+	nodes[1].mark = MARK_LEFT;
+	nodes[1].left = OT_LITERAL;
+    }
+    nodes[1].right = OT_LITERAL;
+    nodes[1].p.parent = 0;
+
+    return ExecConstantExprTree(interp, nodes, 0, &litObjv);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSortingOpCmd --
+ *	Implements the commands:
+ *		<, <=, >, >=, ==, eq
+ *	in the ::tcl::mathop namespace. These commands are defined for
+ *	arbitrary number of arguments by computing the AND of the base
+ *	operator applied to all neighbor argument pairs.
+ *
+ * Results:
+ *	A standard Tcl return code and result left in interp.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSortingOpCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int code = TCL_OK;
+
+    if (objc < 3) {
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(1));
+    } else {
+	TclOpCmdClientData *occdPtr = clientData;
+	Tcl_Obj **litObjv = TclStackAlloc(interp,
+		2 * (objc-2) * sizeof(Tcl_Obj *));
+	OpNode *nodes = TclStackAlloc(interp, 2 * (objc-2) * sizeof(OpNode));
+	unsigned char lexeme;
+	int i, lastAnd = 1;
+	Tcl_Obj *const *litObjPtrPtr = litObjv;
+
+	ParseLexeme(occdPtr->op, strlen(occdPtr->op), &lexeme, NULL);
+
+	litObjv[0] = objv[1];
+	nodes[0].lexeme = START;
+	nodes[0].mark = MARK_RIGHT;
+	for (i=2; i<objc-1; i++) {
+	    litObjv[2*(i-1)-1] = objv[i];
+	    nodes[2*(i-1)-1].lexeme = lexeme;
+	    nodes[2*(i-1)-1].mark = MARK_LEFT;
+	    nodes[2*(i-1)-1].left = OT_LITERAL;
+	    nodes[2*(i-1)-1].right = OT_LITERAL;
+
+	    litObjv[2*(i-1)] = objv[i];
+	    nodes[2*(i-1)].lexeme = AND;
+	    nodes[2*(i-1)].mark = MARK_LEFT;
+	    nodes[2*(i-1)].left = lastAnd;
+	    nodes[lastAnd].p.parent = 2*(i-1);
+
+	    nodes[2*(i-1)].right = 2*(i-1)+1;
+	    nodes[2*(i-1)+1].p.parent= 2*(i-1);
+
+	    lastAnd = 2*(i-1);
+	}
+	litObjv[2*(objc-2)-1] = objv[objc-1];
+
+	nodes[2*(objc-2)-1].lexeme = lexeme;
+	nodes[2*(objc-2)-1].mark = MARK_LEFT;
+	nodes[2*(objc-2)-1].left = OT_LITERAL;
+	nodes[2*(objc-2)-1].right = OT_LITERAL;
+
+	nodes[0].right = lastAnd;
+	nodes[lastAnd].p.parent = 0;
+
+	code = ExecConstantExprTree(interp, nodes, 0, &litObjPtrPtr);
+
+	TclStackFree(interp, nodes);
+	TclStackFree(interp, litObjv);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclVariadicOpCmd --
+ *	Implements the commands: +, *, &, |, ^, **
+ *	in the ::tcl::mathop namespace. These commands are defined for
+ *	arbitrary number of arguments by repeatedly applying the base
+ *	operator with suitable associative rules. When fewer than two
+ *	arguments are provided, suitable identity values are returned.
+ *
+ * Results:
+ *	A standard Tcl return code and result left in interp.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclVariadicOpCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    TclOpCmdClientData *occdPtr = clientData;
+    unsigned char lexeme;
+    int code;
+
+    if (objc < 2) {
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(occdPtr->i.identity));
+	return TCL_OK;
+    }
+
+    ParseLexeme(occdPtr->op, strlen(occdPtr->op), &lexeme, NULL);
+    lexeme |= BINARY;
+
+    if (objc == 2) {
+	Tcl_Obj *litObjv[2];
+	OpNode nodes[2];
+	int decrMe = 0;
+	Tcl_Obj *const *litObjPtrPtr = litObjv;
+
+	if (lexeme == EXPON) {
+	    litObjv[1] = Tcl_NewIntObj(occdPtr->i.identity);
+	    Tcl_IncrRefCount(litObjv[1]);
+	    decrMe = 1;
+	    litObjv[0] = objv[1];
+	    nodes[0].lexeme = START;
+	    nodes[0].mark = MARK_RIGHT;
+	    nodes[0].right = 1;
+	    nodes[1].lexeme = lexeme;
+	    nodes[1].mark = MARK_LEFT;
+	    nodes[1].left = OT_LITERAL;
+	    nodes[1].right = OT_LITERAL;
+	    nodes[1].p.parent = 0;
+	} else {
+	    if (lexeme == DIVIDE) {
+		litObjv[0] = Tcl_NewDoubleObj(1.0);
+	    } else {
+		litObjv[0] = Tcl_NewIntObj(occdPtr->i.identity);
+	    }
+	    Tcl_IncrRefCount(litObjv[0]);
+	    litObjv[1] = objv[1];
+	    nodes[0].lexeme = START;
+	    nodes[0].mark = MARK_RIGHT;
+	    nodes[0].right = 1;
+	    nodes[1].lexeme = lexeme;
+	    nodes[1].mark = MARK_LEFT;
+	    nodes[1].left = OT_LITERAL;
+	    nodes[1].right = OT_LITERAL;
+	    nodes[1].p.parent = 0;
+	}
+
+	code = ExecConstantExprTree(interp, nodes, 0, &litObjPtrPtr);
+
+	Tcl_DecrRefCount(litObjv[decrMe]);
+	return code;
+    } else {
+	Tcl_Obj *const *litObjv = objv + 1;
+	OpNode *nodes = TclStackAlloc(interp, (objc-1) * sizeof(OpNode));
+	int i, lastOp = OT_LITERAL;
+
+	nodes[0].lexeme = START;
+	nodes[0].mark = MARK_RIGHT;
+	if (lexeme == EXPON) {
+	    for (i=objc-2; i>0; i--) {
+		nodes[i].lexeme = lexeme;
+		nodes[i].mark = MARK_LEFT;
+		nodes[i].left = OT_LITERAL;
+		nodes[i].right = lastOp;
+		if (lastOp >= 0) {
+		    nodes[lastOp].p.parent = i;
+		}
+		lastOp = i;
+	    }
+	} else {
+	    for (i=1; i<objc-1; i++) {
+		nodes[i].lexeme = lexeme;
+		nodes[i].mark = MARK_LEFT;
+		nodes[i].left = lastOp;
+		if (lastOp >= 0) {
+		    nodes[lastOp].p.parent = i;
+		}
+		nodes[i].right = OT_LITERAL;
+		lastOp = i;
+	    }
+	}
+	nodes[0].right = lastOp;
+	nodes[lastOp].p.parent = 0;
+
+	code = ExecConstantExprTree(interp, nodes, 0, &litObjv);
+
+	TclStackFree(interp, nodes);
+	return code;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNoIdentOpCmd --
+ *	Implements the commands: -, /
+ *	in the ::tcl::mathop namespace. These commands are defined for
+ *	arbitrary non-zero number of arguments by repeatedly applying the base
+ *	operator with suitable associative rules. When no arguments are
+ *	provided, an error is raised.
+ *
+ * Results:
+ *	A standard Tcl return code and result left in interp.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNoIdentOpCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    TclOpCmdClientData *occdPtr = clientData;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, occdPtr->expected);
+	return TCL_ERROR;
+    }
+    return TclVariadicOpCmd(clientData, interp, objc, objv);
+}
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclCompile.c b/generic/tclCompile.c
new file mode 100644
index 0000000..b5de230
--- /dev/null
+++ b/generic/tclCompile.c
@@ -0,0 +1,4527 @@
+/*
+ * tclCompile.c --
+ *
+ *	This file contains procedures that compile Tcl commands or parts of
+ *	commands (like quoted strings or nested sub-commands) into a sequence
+ *	of instructions ("bytecodes").
+ *
+ * Copyright (c) 1996-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include <assert.h>
+
+/*
+ * Variable that controls whether compilation tracing is enabled and, if so,
+ * what level of tracing is desired:
+ *    0: no compilation tracing
+ *    1: summarize compilation of top level cmds and proc bodies
+ *    2: display all instructions of each ByteCode compiled
+ * This variable is linked to the Tcl variable "tcl_traceCompile".
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+int tclTraceCompile = 0;
+static int traceInitialized = 0;
+#endif
+
+/*
+ * A table describing the Tcl bytecode instructions. Entries in this table
+ * must correspond to the instruction opcode definitions in tclCompile.h. The
+ * names "op1" and "op4" refer to an instruction's one or four byte first
+ * operand. Similarly, "stktop" and "stknext" refer to the topmost and next to
+ * topmost stack elements.
+ *
+ * Note that the load, store, and incr instructions do not distinguish local
+ * from global variables; the bytecode interpreter at runtime uses the
+ * existence of a procedure call frame to distinguish these.
+ */
+
+InstructionDesc const tclInstructionTable[] = {
+    /* Name	      Bytes stackEffect #Opnds  Operand types */
+    {"done",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Finish ByteCode execution and return stktop (top stack item) */
+    {"push1",		  2,   +1,         1,	{OPERAND_LIT1}},
+	/* Push object at ByteCode objArray[op1] */
+    {"push4",		  5,   +1,         1,	{OPERAND_LIT4}},
+	/* Push object at ByteCode objArray[op4] */
+    {"pop",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Pop the topmost stack object */
+    {"dup",		  1,   +1,         0,	{OPERAND_NONE}},
+	/* Duplicate the topmost stack object and push the result */
+    {"strcat",		  2,   INT_MIN,    1,	{OPERAND_UINT1}},
+	/* Concatenate the top op1 items and push result */
+    {"invokeStk1",	  2,   INT_MIN,    1,	{OPERAND_UINT1}},
+	/* Invoke command named objv[0]; <objc,objv> = <op1,top op1> */
+    {"invokeStk4",	  5,   INT_MIN,    1,	{OPERAND_UINT4}},
+	/* Invoke command named objv[0]; <objc,objv> = <op4,top op4> */
+    {"evalStk",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Evaluate command in stktop using Tcl_EvalObj. */
+    {"exprStk",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Execute expression in stktop using Tcl_ExprStringObj. */
+
+    {"loadScalar1",	  2,   1,          1,	{OPERAND_LVT1}},
+	/* Load scalar variable at index op1 <= 255 in call frame */
+    {"loadScalar4",	  5,   1,          1,	{OPERAND_LVT4}},
+	/* Load scalar variable at index op1 >= 256 in call frame */
+    {"loadScalarStk",	  1,   0,          0,	{OPERAND_NONE}},
+	/* Load scalar variable; scalar's name is stktop */
+    {"loadArray1",	  2,   0,          1,	{OPERAND_LVT1}},
+	/* Load array element; array at slot op1<=255, element is stktop */
+    {"loadArray4",	  5,   0,          1,	{OPERAND_LVT4}},
+	/* Load array element; array at slot op1 > 255, element is stktop */
+    {"loadArrayStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Load array element; element is stktop, array name is stknext */
+    {"loadStk",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Load general variable; unparsed variable name is stktop */
+    {"storeScalar1",	  2,   0,          1,	{OPERAND_LVT1}},
+	/* Store scalar variable at op1<=255 in frame; value is stktop */
+    {"storeScalar4",	  5,   0,          1,	{OPERAND_LVT4}},
+	/* Store scalar variable at op1 > 255 in frame; value is stktop */
+    {"storeScalarStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Store scalar; value is stktop, scalar name is stknext */
+    {"storeArray1",	  2,   -1,         1,	{OPERAND_LVT1}},
+	/* Store array element; array at op1<=255, value is top then elem */
+    {"storeArray4",	  5,   -1,         1,	{OPERAND_LVT4}},
+	/* Store array element; array at op1>=256, value is top then elem */
+    {"storeArrayStk",	  1,   -2,         0,	{OPERAND_NONE}},
+	/* Store array element; value is stktop, then elem, array names */
+    {"storeStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Store general variable; value is stktop, then unparsed name */
+
+    {"incrScalar1",	  2,   0,          1,	{OPERAND_LVT1}},
+	/* Incr scalar at index op1<=255 in frame; incr amount is stktop */
+    {"incrScalarStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Incr scalar; incr amount is stktop, scalar's name is stknext */
+    {"incrArray1",	  2,   -1,         1,	{OPERAND_LVT1}},
+	/* Incr array elem; arr at slot op1<=255, amount is top then elem */
+    {"incrArrayStk",	  1,   -2,         0,	{OPERAND_NONE}},
+	/* Incr array element; amount is top then elem then array names */
+    {"incrStk",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Incr general variable; amount is stktop then unparsed var name */
+    {"incrScalar1Imm",	  3,   +1,         2,	{OPERAND_LVT1, OPERAND_INT1}},
+	/* Incr scalar at slot op1 <= 255; amount is 2nd operand byte */
+    {"incrScalarStkImm",  2,   0,          1,	{OPERAND_INT1}},
+	/* Incr scalar; scalar name is stktop; incr amount is op1 */
+    {"incrArray1Imm",	  3,   0,          2,	{OPERAND_LVT1, OPERAND_INT1}},
+	/* Incr array elem; array at slot op1 <= 255, elem is stktop,
+	 * amount is 2nd operand byte */
+    {"incrArrayStkImm",	  2,   -1,         1,	{OPERAND_INT1}},
+	/* Incr array element; elem is top then array name, amount is op1 */
+    {"incrStkImm",	  2,   0,	   1,	{OPERAND_INT1}},
+	/* Incr general variable; unparsed name is top, amount is op1 */
+
+    {"jump1",		  2,   0,          1,	{OPERAND_OFFSET1}},
+	/* Jump relative to (pc + op1) */
+    {"jump4",		  5,   0,          1,	{OPERAND_OFFSET4}},
+	/* Jump relative to (pc + op4) */
+    {"jumpTrue1",	  2,   -1,         1,	{OPERAND_OFFSET1}},
+	/* Jump relative to (pc + op1) if stktop expr object is true */
+    {"jumpTrue4",	  5,   -1,         1,	{OPERAND_OFFSET4}},
+	/* Jump relative to (pc + op4) if stktop expr object is true */
+    {"jumpFalse1",	  2,   -1,         1,	{OPERAND_OFFSET1}},
+	/* Jump relative to (pc + op1) if stktop expr object is false */
+    {"jumpFalse4",	  5,   -1,         1,	{OPERAND_OFFSET4}},
+	/* Jump relative to (pc + op4) if stktop expr object is false */
+
+    {"lor",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Logical or:	push (stknext || stktop) */
+    {"land",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Logical and:	push (stknext && stktop) */
+    {"bitor",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Bitwise or:	push (stknext | stktop) */
+    {"bitxor",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Bitwise xor	push (stknext ^ stktop) */
+    {"bitand",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Bitwise and:	push (stknext & stktop) */
+    {"eq",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Equal:	push (stknext == stktop) */
+    {"neq",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Not equal:	push (stknext != stktop) */
+    {"lt",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Less:	push (stknext < stktop) */
+    {"gt",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Greater:	push (stknext > stktop) */
+    {"le",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Less or equal: push (stknext <= stktop) */
+    {"ge",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Greater or equal: push (stknext >= stktop) */
+    {"lshift",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Left shift:	push (stknext << stktop) */
+    {"rshift",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Right shift:	push (stknext >> stktop) */
+    {"add",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Add:		push (stknext + stktop) */
+    {"sub",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Sub:		push (stkext - stktop) */
+    {"mult",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Multiply:	push (stknext * stktop) */
+    {"div",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Divide:	push (stknext / stktop) */
+    {"mod",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Mod:		push (stknext % stktop) */
+    {"uplus",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Unary plus:	push +stktop */
+    {"uminus",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Unary minus:	push -stktop */
+    {"bitnot",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Bitwise not:	push ~stktop */
+    {"not",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Logical not:	push !stktop */
+    {"callBuiltinFunc1",  2,   1,          1,	{OPERAND_UINT1}},
+	/* Call builtin math function with index op1; any args are on stk */
+    {"callFunc1",	  2,   INT_MIN,    1,	{OPERAND_UINT1}},
+	/* Call non-builtin func objv[0]; <objc,objv>=<op1,top op1> */
+    {"tryCvtToNumeric",	  1,   0,          0,	{OPERAND_NONE}},
+	/* Try converting stktop to first int then double if possible. */
+
+    {"break",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Abort closest enclosing loop; if none, return TCL_BREAK code. */
+    {"continue",	  1,   0,          0,	{OPERAND_NONE}},
+	/* Skip to next iteration of closest enclosing loop; if none, return
+	 * TCL_CONTINUE code. */
+
+    {"foreach_start4",	  5,   0,          1,	{OPERAND_AUX4}},
+	/* Initialize execution of a foreach loop. Operand is aux data index
+	 * of the ForeachInfo structure for the foreach command. */
+    {"foreach_step4",	  5,   +1,         1,	{OPERAND_AUX4}},
+	/* "Step" or begin next iteration of foreach loop. Push 0 if to
+	 * terminate loop, else push 1. */
+
+    {"beginCatch4",	  5,   0,          1,	{OPERAND_UINT4}},
+	/* Record start of catch with the operand's exception index. Push the
+	 * current stack depth onto a special catch stack. */
+    {"endCatch",	  1,   0,          0,	{OPERAND_NONE}},
+	/* End of last catch. Pop the bytecode interpreter's catch stack. */
+    {"pushResult",	  1,   +1,         0,	{OPERAND_NONE}},
+	/* Push the interpreter's object result onto the stack. */
+    {"pushReturnCode",	  1,   +1,         0,	{OPERAND_NONE}},
+	/* Push interpreter's return code (e.g. TCL_OK or TCL_ERROR) as a new
+	 * object onto the stack. */
+
+    {"streq",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Str Equal:	push (stknext eq stktop) */
+    {"strneq",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Str !Equal:	push (stknext neq stktop) */
+    {"strcmp",		  1,   -1,         0,	{OPERAND_NONE}},
+	/* Str Compare:	push (stknext cmp stktop) */
+    {"strlen",		  1,   0,          0,	{OPERAND_NONE}},
+	/* Str Length:	push (strlen stktop) */
+    {"strindex",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Str Index:	push (strindex stknext stktop) */
+    {"strmatch",	  2,   -1,         1,	{OPERAND_INT1}},
+	/* Str Match:	push (strmatch stknext stktop) opnd == nocase */
+
+    {"list",		  5,   INT_MIN,    1,	{OPERAND_UINT4}},
+	/* List:	push (stk1 stk2 ... stktop) */
+    {"listIndex",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* List Index:	push (listindex stknext stktop) */
+    {"listLength",	  1,   0,          0,	{OPERAND_NONE}},
+	/* List Len:	push (listlength stktop) */
+
+    {"appendScalar1",	  2,   0,          1,	{OPERAND_LVT1}},
+	/* Append scalar variable at op1<=255 in frame; value is stktop */
+    {"appendScalar4",	  5,   0,          1,	{OPERAND_LVT4}},
+	/* Append scalar variable at op1 > 255 in frame; value is stktop */
+    {"appendArray1",	  2,   -1,         1,	{OPERAND_LVT1}},
+	/* Append array element; array at op1<=255, value is top then elem */
+    {"appendArray4",	  5,   -1,         1,	{OPERAND_LVT4}},
+	/* Append array element; array at op1>=256, value is top then elem */
+    {"appendArrayStk",	  1,   -2,         0,	{OPERAND_NONE}},
+	/* Append array element; value is stktop, then elem, array names */
+    {"appendStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Append general variable; value is stktop, then unparsed name */
+    {"lappendScalar1",	  2,   0,          1,	{OPERAND_LVT1}},
+	/* Lappend scalar variable at op1<=255 in frame; value is stktop */
+    {"lappendScalar4",	  5,   0,          1,	{OPERAND_LVT4}},
+	/* Lappend scalar variable at op1 > 255 in frame; value is stktop */
+    {"lappendArray1",	  2,   -1,         1,	{OPERAND_LVT1}},
+	/* Lappend array element; array at op1<=255, value is top then elem */
+    {"lappendArray4",	  5,   -1,         1,	{OPERAND_LVT4}},
+	/* Lappend array element; array at op1>=256, value is top then elem */
+    {"lappendArrayStk",	  1,   -2,         0,	{OPERAND_NONE}},
+	/* Lappend array element; value is stktop, then elem, array names */
+    {"lappendStk",	  1,   -1,         0,	{OPERAND_NONE}},
+	/* Lappend general variable; value is stktop, then unparsed name */
+
+    {"lindexMulti",	  5,   INT_MIN,    1,	{OPERAND_UINT4}},
+	/* Lindex with generalized args, operand is number of stacked objs
+	 * used: (operand-1) entries from stktop are the indices; then list to
+	 * process. */
+    {"over",		  5,   +1,         1,	{OPERAND_UINT4}},
+	/* Duplicate the arg-th element from top of stack (TOS=0) */
+    {"lsetList",          1,   -2,         0,	{OPERAND_NONE}},
+	/* Four-arg version of 'lset'. stktop is old value; next is new
+	 * element value, next is the index list; pushes new value */
+    {"lsetFlat",          5,   INT_MIN,    1,	{OPERAND_UINT4}},
+	/* Three- or >=5-arg version of 'lset', operand is number of stacked
+	 * objs: stktop is old value, next is new element value, next come
+	 * (operand-2) indices; pushes the new value.
+	 */
+
+    {"returnImm",	  9,   -1,         2,	{OPERAND_INT4, OPERAND_UINT4}},
+	/* Compiled [return], code, level are operands; options and result
+	 * are on the stack. */
+    {"expon",		  1,   -1,	   0,	{OPERAND_NONE}},
+	/* Binary exponentiation operator: push (stknext ** stktop) */
+
+    /*
+     * NOTE: the stack effects of expandStkTop and invokeExpanded are wrong -
+     * but it cannot be done right at compile time, the stack effect is only
+     * known at run time. The value for invokeExpanded is estimated better at
+     * compile time.
+     * See the comments further down in this file, where INST_INVOKE_EXPANDED
+     * is emitted.
+     */
+    {"expandStart",       1,    0,          0,	{OPERAND_NONE}},
+	/* Start of command with {*} (expanded) arguments */
+    {"expandStkTop",      5,    0,          1,	{OPERAND_UINT4}},
+	/* Expand the list at stacktop: push its elements on the stack */
+    {"invokeExpanded",    1,    0,          0,	{OPERAND_NONE}},
+	/* Invoke the command marked by the last 'expandStart' */
+
+    {"listIndexImm",	  5,	0,	   1,	{OPERAND_IDX4}},
+	/* List Index:	push (lindex stktop op4) */
+    {"listRangeImm",	  9,	0,	   2,	{OPERAND_IDX4, OPERAND_IDX4}},
+	/* List Range:	push (lrange stktop op4 op4) */
+    {"startCommand",	  9,	0,	   2,	{OPERAND_OFFSET4, OPERAND_UINT4}},
+	/* Start of bytecoded command: op is the length of the cmd's code, op2
+	 * is number of commands here */
+
+    {"listIn",		  1,	-1,	   0,	{OPERAND_NONE}},
+	/* List containment: push [lsearch stktop stknext]>=0) */
+    {"listNotIn",	  1,	-1,	   0,	{OPERAND_NONE}},
+	/* List negated containment: push [lsearch stktop stknext]<0) */
+
+    {"pushReturnOpts",	  1,	+1,	   0,	{OPERAND_NONE}},
+	/* Push the interpreter's return option dictionary as an object on the
+	 * stack. */
+    {"returnStk",	  1,	-1,	   0,	{OPERAND_NONE}},
+	/* Compiled [return]; options and result are on the stack, code and
+	 * level are in the options. */
+
+    {"dictGet",		  5,	INT_MIN,   1,	{OPERAND_UINT4}},
+	/* The top op4 words (min 1) are a key path into the dictionary just
+	 * below the keys on the stack, and all those values are replaced by
+	 * the value read out of that key-path (like [dict get]).
+	 * Stack:  ... dict key1 ... keyN => ... value */
+    {"dictSet",		  9,	INT_MIN,   2,	{OPERAND_UINT4, OPERAND_LVT4}},
+	/* Update a dictionary value such that the keys are a path pointing to
+	 * the value. op4#1 = numKeys, op4#2 = LVTindex
+	 * Stack:  ... key1 ... keyN value => ... newDict */
+    {"dictUnset",	  9,	INT_MIN,   2,	{OPERAND_UINT4, OPERAND_LVT4}},
+	/* Update a dictionary value such that the keys are not a path pointing
+	 * to any value. op4#1 = numKeys, op4#2 = LVTindex
+	 * Stack:  ... key1 ... keyN => ... newDict */
+    {"dictIncrImm",	  9,	0,	   2,	{OPERAND_INT4, OPERAND_LVT4}},
+	/* Update a dictionary value such that the value pointed to by key is
+	 * incremented by some value (or set to it if the key isn't in the
+	 * dictionary at all). op4#1 = incrAmount, op4#2 = LVTindex
+	 * Stack:  ... key => ... newDict */
+    {"dictAppend",	  5,	-1,	   1,	{OPERAND_LVT4}},
+	/* Update a dictionary value such that the value pointed to by key has
+	 * some value string-concatenated onto it. op4 = LVTindex
+	 * Stack:  ... key valueToAppend => ... newDict */
+    {"dictLappend",	  5,	-1,	   1,	{OPERAND_LVT4}},
+	/* Update a dictionary value such that the value pointed to by key has
+	 * some value list-appended onto it. op4 = LVTindex
+	 * Stack:  ... key valueToAppend => ... newDict */
+    {"dictFirst",	  5,	+2,	   1,	{OPERAND_LVT4}},
+	/* Begin iterating over the dictionary, using the local scalar
+	 * indicated by op4 to hold the iterator state. The local scalar
+	 * should not refer to a named variable as the value is not wholly
+	 * managed correctly.
+	 * Stack:  ... dict => ... value key doneBool */
+    {"dictNext",	  5,	+3,	   1,	{OPERAND_LVT4}},
+	/* Get the next iteration from the iterator in op4's local scalar.
+	 * Stack:  ... => ... value key doneBool */
+    {"dictDone",	  5,	0,	   1,	{OPERAND_LVT4}},
+	/* Terminate the iterator in op4's local scalar. Use unsetScalar
+	 * instead (with 0 for flags). */
+    {"dictUpdateStart",   9,    0,	   2,	{OPERAND_LVT4, OPERAND_AUX4}},
+	/* Create the variables (described in the aux data referred to by the
+	 * second immediate argument) to mirror the state of the dictionary in
+	 * the variable referred to by the first immediate argument. The list
+	 * of keys (top of the stack, not popped) must be the same length as
+	 * the list of variables.
+	 * Stack:  ... keyList => ... keyList */
+    {"dictUpdateEnd",	  9,    -1,	   2,	{OPERAND_LVT4, OPERAND_AUX4}},
+	/* Reflect the state of local variables (described in the aux data
+	 * referred to by the second immediate argument) back to the state of
+	 * the dictionary in the variable referred to by the first immediate
+	 * argument. The list of keys (popped from the stack) must be the same
+	 * length as the list of variables.
+	 * Stack:  ... keyList => ... */
+    {"jumpTable",	 5,	-1,	   1,	{OPERAND_AUX4}},
+	/* Jump according to the jump-table (in AuxData as indicated by the
+	 * operand) and the argument popped from the list. Always executes the
+	 * next instruction if no match against the table's entries was found.
+	 * Stack:  ... value => ...
+	 * Note that the jump table contains offsets relative to the PC when
+	 * it points to this instruction; the code is relocatable. */
+    {"upvar",            5,    -1,        1,   {OPERAND_LVT4}},
+	/* finds level and otherName in stack, links to local variable at
+	 * index op1. Leaves the level on stack. */
+    {"nsupvar",          5,    -1,        1,   {OPERAND_LVT4}},
+	/* finds namespace and otherName in stack, links to local variable at
+	 * index op1. Leaves the namespace on stack. */
+    {"variable",         5,    -1,        1,   {OPERAND_LVT4}},
+	/* finds namespace and otherName in stack, links to local variable at
+	 * index op1. Leaves the namespace on stack. */
+    {"syntax",		 9,   -1,         2,	{OPERAND_INT4, OPERAND_UINT4}},
+	/* Compiled bytecodes to signal syntax error. Equivalent to returnImm
+	 * except for the ERR_ALREADY_LOGGED flag in the interpreter. */
+    {"reverse",		 5,    0,         1,	{OPERAND_UINT4}},
+	/* Reverse the order of the arg elements at the top of stack */
+
+    {"regexp",		 2,   -1,         1,	{OPERAND_INT1}},
+	/* Regexp:	push (regexp stknext stktop) opnd == nocase */
+
+    {"existScalar",	 5,    1,         1,	{OPERAND_LVT4}},
+	/* Test if scalar variable at index op1 in call frame exists */
+    {"existArray",	 5,    0,         1,	{OPERAND_LVT4}},
+	/* Test if array element exists; array at slot op1, element is
+	 * stktop */
+    {"existArrayStk",	 1,    -1,        0,	{OPERAND_NONE}},
+	/* Test if array element exists; element is stktop, array name is
+	 * stknext */
+    {"existStk",	 1,    0,         0,	{OPERAND_NONE}},
+	/* Test if general variable exists; unparsed variable name is stktop*/
+
+    {"nop",		 1,    0,         0,	{OPERAND_NONE}},
+	/* Do nothing */
+    {"returnCodeBranch", 1,   -1,	  0,	{OPERAND_NONE}},
+	/* Jump to next instruction based on the return code on top of stack
+	 * ERROR: +1;	RETURN: +3;	BREAK: +5;	CONTINUE: +7;
+	 * Other non-OK: +9
+	 */
+
+    {"unsetScalar",	 6,    0,         2,	{OPERAND_UINT1, OPERAND_LVT4}},
+	/* Make scalar variable at index op2 in call frame cease to exist;
+	 * op1 is 1 for errors on problems, 0 otherwise */
+    {"unsetArray",	 6,    -1,        2,	{OPERAND_UINT1, OPERAND_LVT4}},
+	/* Make array element cease to exist; array at slot op2, element is
+	 * stktop; op1 is 1 for errors on problems, 0 otherwise */
+    {"unsetArrayStk",	 2,    -2,        1,	{OPERAND_UINT1}},
+	/* Make array element cease to exist; element is stktop, array name is
+	 * stknext; op1 is 1 for errors on problems, 0 otherwise */
+    {"unsetStk",	 2,    -1,        1,	{OPERAND_UINT1}},
+	/* Make general variable cease to exist; unparsed variable name is
+	 * stktop; op1 is 1 for errors on problems, 0 otherwise */
+
+    {"dictExpand",       1,    -1,        0,    {OPERAND_NONE}},
+        /* Probe into a dict and extract it (or a subdict of it) into
+         * variables with matched names. Produces list of keys bound as
+         * result. Part of [dict with].
+	 * Stack:  ... dict path => ... keyList */
+    {"dictRecombineStk", 1,    -3,        0,    {OPERAND_NONE}},
+        /* Map variable contents back into a dictionary in a variable. Part of
+         * [dict with].
+	 * Stack:  ... dictVarName path keyList => ... */
+    {"dictRecombineImm", 5,    -2,        1,    {OPERAND_LVT4}},
+        /* Map variable contents back into a dictionary in the local variable
+         * indicated by the LVT index. Part of [dict with].
+	 * Stack:  ... path keyList => ... */
+    {"dictExists",	 5,	INT_MIN,  1,	{OPERAND_UINT4}},
+	/* The top op4 words (min 1) are a key path into the dictionary just
+	 * below the keys on the stack, and all those values are replaced by a
+	 * boolean indicating whether it is possible to read out a value from
+	 * that key-path (like [dict exists]).
+	 * Stack:  ... dict key1 ... keyN => ... boolean */
+    {"verifyDict",	 1,    -1,	  0,	{OPERAND_NONE}},
+	/* Verifies that the word on the top of the stack is a dictionary,
+	 * popping it if it is and throwing an error if it is not.
+	 * Stack:  ... value => ... */
+
+    {"strmap",		 1,    -2,	  0,	{OPERAND_NONE}},
+	/* Simplified version of [string map] that only applies one change
+	 * string, and only case-sensitively.
+	 * Stack:  ... from to string => ... changedString */
+    {"strfind",		 1,    -1,	  0,	{OPERAND_NONE}},
+	/* Find the first index of a needle string in a haystack string,
+	 * producing the index (integer) or -1 if nothing found.
+	 * Stack:  ... needle haystack => ... index */
+    {"strrfind",	 1,    -1,	  0,	{OPERAND_NONE}},
+	/* Find the last index of a needle string in a haystack string,
+	 * producing the index (integer) or -1 if nothing found.
+	 * Stack:  ... needle haystack => ... index */
+    {"strrangeImm",	 9,	0,	  2,	{OPERAND_IDX4, OPERAND_IDX4}},
+	/* String Range: push (string range stktop op4 op4) */
+    {"strrange",	 1,    -2,	  0,	{OPERAND_NONE}},
+	/* String Range with non-constant arguments.
+	 * Stack:  ... string idxA idxB => ... substring */
+
+    {"yield",		 1,	0,	  0,	{OPERAND_NONE}},
+	/* Makes the current coroutine yield the value at the top of the
+	 * stack, and places the response back on top of the stack when it
+	 * resumes.
+	 * Stack:  ... valueToYield => ... resumeValue */
+    {"coroName",         1,    +1,	  0,	{OPERAND_NONE}},
+	/* Push the name of the interpreter's current coroutine as an object
+	 * on the stack. */
+    {"tailcall",	 2,    INT_MIN,	  1,	{OPERAND_UINT1}},
+	/* Do a tailcall with the opnd items on the stack as the thing to
+	 * tailcall to; opnd must be greater than 0 for the semantics to work
+	 * right. */
+
+    {"currentNamespace", 1,    +1,	  0,	{OPERAND_NONE}},
+	/* Push the name of the interpreter's current namespace as an object
+	 * on the stack. */
+    {"infoLevelNumber",  1,    +1,	  0,	{OPERAND_NONE}},
+	/* Push the stack depth (i.e., [info level]) of the interpreter as an
+	 * object on the stack. */
+    {"infoLevelArgs",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Push the argument words to a stack depth (i.e., [info level <n>])
+	 * of the interpreter as an object on the stack.
+	 * Stack:  ... depth => ... argList */
+    {"resolveCmd",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Resolves the command named on the top of the stack to its fully
+	 * qualified version, or produces the empty string if no such command
+	 * exists. Never generates errors.
+	 * Stack:  ... cmdName => ... fullCmdName */
+
+    {"tclooSelf",	 1,	+1,	  0,	{OPERAND_NONE}},
+	/* Push the identity of the current TclOO object (i.e., the name of
+	 * its current public access command) on the stack. */
+    {"tclooClass",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Push the class of the TclOO object named at the top of the stack
+	 * onto the stack.
+	 * Stack:  ... object => ... class */
+    {"tclooNamespace",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Push the namespace of the TclOO object named at the top of the
+	 * stack onto the stack.
+	 * Stack:  ... object => ... namespace */
+    {"tclooIsObject",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Push whether the value named at the top of the stack is a TclOO
+	 * object (i.e., a boolean). Can corrupt the interpreter result
+	 * despite not throwing, so not safe for use in a post-exception
+	 * context.
+	 * Stack:  ... value => ... boolean */
+
+    {"arrayExistsStk",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Looks up the element on the top of the stack and tests whether it
+	 * is an array. Pushes a boolean describing whether this is the
+	 * case. Also runs the whole-array trace on the named variable, so can
+	 * throw anything.
+	 * Stack:  ... varName => ... boolean */
+    {"arrayExistsImm",	 5,	+1,	  1,	{OPERAND_LVT4}},
+	/* Looks up the variable indexed by opnd and tests whether it is an
+	 * array. Pushes a boolean describing whether this is the case. Also
+	 * runs the whole-array trace on the named variable, so can throw
+	 * anything.
+	 * Stack:  ... => ... boolean */
+    {"arrayMakeStk",	 1,	-1,	  0,	{OPERAND_NONE}},
+	/* Forces the element on the top of the stack to be the name of an
+	 * array.
+	 * Stack:  ... varName => ... */
+    {"arrayMakeImm",	 5,	0,	  1,	{OPERAND_LVT4}},
+	/* Forces the variable indexed by opnd to be an array. Does not touch
+	 * the stack. */
+
+    {"invokeReplace",	 6,	INT_MIN,  2,	{OPERAND_UINT4,OPERAND_UINT1}},
+	/* Invoke command named objv[0], replacing the first two words with
+	 * the word at the top of the stack;
+	 * <objc,objv> = <op4,top op4 after popping 1> */
+
+    {"listConcat",	 1,	-1,	  0,	{OPERAND_NONE}},
+	/* Concatenates the two lists at the top of the stack into a single
+	 * list and pushes that resulting list onto the stack.
+	 * Stack: ... list1 list2 => ... [lconcat list1 list2] */
+
+    {"expandDrop",       1,    0,          0,	{OPERAND_NONE}},
+	/* Drops an element from the auxiliary stack, popping stack elements
+	 * until the matching stack depth is reached. */
+
+    /* New foreach implementation */
+    {"foreach_start",	 5,	+2,	  1,	{OPERAND_AUX4}},
+	/* Initialize execution of a foreach loop. Operand is aux data index
+	 * of the ForeachInfo structure for the foreach command. It pushes 2
+	 * elements which hold runtime params for foreach_step, they are later
+	 * dropped by foreach_end together with the value lists. NOTE that the
+	 * iterator-tracker and info reference must not be passed to bytecodes
+	 * that handle normal Tcl values. NOTE that this instruction jumps to
+	 * the foreach_step instruction paired with it; the stack info below
+	 * is only nominal.
+	 * Stack: ... listObjs... => ... listObjs... iterTracker info */
+    {"foreach_step",	 1,	 0,	  0,	{OPERAND_NONE}},
+	/* "Step" or begin next iteration of foreach loop. Assigns to foreach
+	 * iteration variables. May jump to straight after the foreach_start
+	 * that pushed the iterTracker and info values. MUST be followed
+	 * immediately by a foreach_end.
+	 * Stack: ... listObjs... iterTracker info =>
+	 *				... listObjs... iterTracker info */
+    {"foreach_end",	 1,	 0,	  0,	{OPERAND_NONE}},
+	/* Clean up a foreach loop by dropping the info value, the tracker
+	 * value and the lists that were being iterated over.
+	 * Stack: ... listObjs... iterTracker info => ... */
+    {"lmap_collect",	 1,	-1,	  0,	{OPERAND_NONE}},
+	/* Appends the value at the top of the stack to the list located on
+	 * the stack the "other side" of the foreach-related values.
+	 * Stack: ... collector listObjs... iterTracker info value =>
+	 *			... collector listObjs... iterTracker info */
+
+    {"strtrim",		 1,	-1,	  0,	{OPERAND_NONE}},
+	/* [string trim] core: removes the characters (designated by the value
+	 * at the top of the stack) from both ends of the string and pushes
+	 * the resulting string.
+	 * Stack: ... string charset => ... trimmedString */
+    {"strtrimLeft",	 1,	-1,	  0,	{OPERAND_NONE}},
+	/* [string trimleft] core: removes the characters (designated by the
+	 * value at the top of the stack) from the left of the string and
+	 * pushes the resulting string.
+	 * Stack: ... string charset => ... trimmedString */
+    {"strtrimRight",	 1,	-1,	  0,	{OPERAND_NONE}},
+	/* [string trimright] core: removes the characters (designated by the
+	 * value at the top of the stack) from the right of the string and
+	 * pushes the resulting string.
+	 * Stack: ... string charset => ... trimmedString */
+
+    {"concatStk",	 5,	INT_MIN,  1,	{OPERAND_UINT4}},
+	/* Wrapper round Tcl_ConcatObj(), used for [concat] and [eval]. opnd
+	 * is number of values to concatenate.
+	 * Operation:	push concat(stk1 stk2 ... stktop) */
+
+    {"strcaseUpper",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* [string toupper] core: converts whole string to upper case using
+	 * the default (extended "C" locale) rules.
+	 * Stack: ... string => ... newString */
+    {"strcaseLower",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* [string tolower] core: converts whole string to upper case using
+	 * the default (extended "C" locale) rules.
+	 * Stack: ... string => ... newString */
+    {"strcaseTitle",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* [string totitle] core: converts whole string to upper case using
+	 * the default (extended "C" locale) rules.
+	 * Stack: ... string => ... newString */
+    {"strreplace",	 1,	-3,	  0,	{OPERAND_NONE}},
+	/* [string replace] core: replaces a non-empty range of one string
+	 * with the contents of another.
+	 * Stack: ... string fromIdx toIdx replacement => ... newString */
+
+    {"originCmd",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Reports which command was the origin (via namespace import chain)
+	 * of the command named on the top of the stack.
+	 * Stack:  ... cmdName => ... fullOriginalCmdName */
+
+    {"tclooNext",	 2,	INT_MIN,  1,	{OPERAND_UINT1}},
+	/* Call the next item on the TclOO call chain, passing opnd arguments
+	 * (min 1, max 255, *includes* "next").  The result of the invoked
+	 * method implementation will be pushed on the stack in place of the
+	 * arguments (similar to invokeStk).
+	 * Stack:  ... "next" arg2 arg3 -- argN => ... result */
+    {"tclooNextClass",	 2,	INT_MIN,  1,	{OPERAND_UINT1}},
+	/* Call the following item on the TclOO call chain defined by class
+	 * className, passing opnd arguments (min 2, max 255, *includes*
+	 * "nextto" and the class name). The result of the invoked method
+	 * implementation will be pushed on the stack in place of the
+	 * arguments (similar to invokeStk).
+	 * Stack:  ... "nextto" className arg3 arg4 -- argN => ... result */
+
+    {"yieldToInvoke",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Makes the current coroutine yield the value at the top of the
+	 * stack, invoking the given command/args with resolution in the given
+	 * namespace (all packed into a list), and places the list of values
+	 * that are the response back on top of the stack when it resumes.
+	 * Stack:  ... [list ns cmd arg1 ... argN] => ... resumeList */
+
+    {"numericType",	 1,	0,	  0,	{OPERAND_NONE}},
+	/* Pushes the numeric type code of the word at the top of the stack.
+	 * Stack:  ... value => ... typeCode */
+    {"tryCvtToBoolean",	 1,	+1,	  0,	{OPERAND_NONE}},
+	/* Try converting stktop to boolean if possible. No errors.
+	 * Stack:  ... value => ... value isStrictBool */
+    {"strclass",	 2,	0,	  1,	{OPERAND_SCLS1}},
+	/* See if all the characters of the given string are a member of the
+	 * specified (by opnd) character class. Note that an empty string will
+	 * satisfy the class check (standard definition of "all").
+	 * Stack:  ... stringValue => ... boolean */
+
+    {"lappendList",	 5,	0,	1,	{OPERAND_LVT4}},
+	/* Lappend list to scalar variable at op4 in frame.
+	 * Stack:  ... list => ... listVarContents */
+    {"lappendListArray", 5,	-1,	1,	{OPERAND_LVT4}},
+	/* Lappend list to array element; array at op4.
+	 * Stack:  ... elem list => ... listVarContents */
+    {"lappendListArrayStk", 1,	-2,	0,	{OPERAND_NONE}},
+	/* Lappend list to array element.
+	 * Stack:  ... arrayName elem list => ... listVarContents */
+    {"lappendListStk",	 1,	-1,	0,	{OPERAND_NONE}},
+	/* Lappend list to general variable.
+	 * Stack:  ... varName list => ... listVarContents */
+
+    {"clockRead",	 2,	+1,	1,	{OPERAND_UINT1}},
+        /* Read clock out to the stack. Operand is which clock to read
+	 * 0=clicks, 1=microseconds, 2=milliseconds, 3=seconds.
+	 * Stack: ... => ... time */
+
+    {NULL, 0, 0, 0, {OPERAND_NONE}}
+};
+
+/*
+ * Prototypes for procedures defined later in this file:
+ */
+
+static void		CleanupByteCode(ByteCode *codePtr);
+static ByteCode *	CompileSubstObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    int flags);
+static void		DupByteCodeInternalRep(Tcl_Obj *srcPtr,
+			    Tcl_Obj *copyPtr);
+static unsigned char *	EncodeCmdLocMap(CompileEnv *envPtr,
+			    ByteCode *codePtr, unsigned char *startPtr);
+static void		EnterCmdExtentData(CompileEnv *envPtr,
+			    int cmdNumber, int numSrcBytes, int numCodeBytes);
+static void		EnterCmdStartData(CompileEnv *envPtr,
+			    int cmdNumber, int srcOffset, int codeOffset);
+static void		FreeByteCodeInternalRep(Tcl_Obj *objPtr);
+static void		FreeSubstCodeInternalRep(Tcl_Obj *objPtr);
+static int		GetCmdLocEncodingSize(CompileEnv *envPtr);
+static int		IsCompactibleCompileEnv(Tcl_Interp *interp,
+			    CompileEnv *envPtr);
+static void		PreventCycle(Tcl_Obj *objPtr, CompileEnv *envPtr);
+#ifdef TCL_COMPILE_STATS
+static void		RecordByteCodeStats(ByteCode *codePtr);
+#endif /* TCL_COMPILE_STATS */
+static int		SetByteCodeFromAny(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static void		StartExpanding(CompileEnv *envPtr);
+
+/*
+ * TIP #280: Helper for building the per-word line information of all compiled
+ * commands.
+ */
+static void		EnterCmdWordData(ExtCmdLoc *eclPtr, int srcOffset,
+			    Tcl_Token *tokenPtr, const char *cmd, int len,
+			    int numWords, int line, int *clNext, int **lines,
+			    CompileEnv *envPtr);
+static void		ReleaseCmdWordData(ExtCmdLoc *eclPtr);
+
+/*
+ * The structure below defines the bytecode Tcl object type by means of
+ * procedures that can be invoked by generic object code.
+ */
+
+const Tcl_ObjType tclByteCodeType = {
+    "bytecode",			/* name */
+    FreeByteCodeInternalRep,	/* freeIntRepProc */
+    DupByteCodeInternalRep,	/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    SetByteCodeFromAny		/* setFromAnyProc */
+};
+
+/*
+ * The structure below defines a bytecode Tcl object type to hold the
+ * compiled bytecode for the [subst]itution of Tcl values.
+ */
+
+static const Tcl_ObjType substCodeType = {
+    "substcode",		/* name */
+    FreeSubstCodeInternalRep,	/* freeIntRepProc */
+    DupByteCodeInternalRep,	/* dupIntRepProc - shared with bytecode */
+    NULL,			/* updateStringProc */
+    NULL,			/* setFromAnyProc */
+};
+
+/*
+ * Helper macros.
+ */
+
+#define TclIncrUInt4AtPtr(ptr, delta) \
+    TclStoreInt4AtPtr(TclGetUInt4AtPtr(ptr)+(delta), (ptr));
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetByteCodeFromAny --
+ *
+ *	Part of the bytecode Tcl object type implementation. Attempts to
+ *	generate an byte code internal form for the Tcl object "objPtr" by
+ *	compiling its string representation. This function also takes a hook
+ *	procedure that will be invoked to perform any needed post processing
+ *	on the compilation results before generating byte codes. interp is
+ *	compilation context and may not be NULL.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during compilation, an error message is left in the interpreter's
+ *	result.
+ *
+ * Side effects:
+ *	Frees the old internal representation. If no error occurs, then the
+ *	compiled code is stored as "objPtr"s bytecode representation. Also, if
+ *	debugging, initializes the "tcl_traceCompile" Tcl variable used to
+ *	trace compilations.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSetByteCodeFromAny(
+    Tcl_Interp *interp,		/* The interpreter for which the code is being
+				 * compiled. Must not be NULL. */
+    Tcl_Obj *objPtr,		/* The object to make a ByteCode object. */
+    CompileHookProc *hookProc,	/* Procedure to invoke after compilation. */
+    ClientData clientData)	/* Hook procedure private data. */
+{
+    Interp *iPtr = (Interp *) interp;
+    CompileEnv compEnv;		/* Compilation environment structure allocated
+				 * in frame. */
+    size_t length;
+    int result = TCL_OK;
+    const char *stringPtr;
+    Proc *procPtr = iPtr->compiledProcPtr;
+    ContLineLoc *clLocPtr;
+
+#ifdef TCL_COMPILE_DEBUG
+    if (!traceInitialized) {
+	if (Tcl_LinkVar(interp, "tcl_traceCompile",
+		(char *) &tclTraceCompile, TCL_LINK_INT) != TCL_OK) {
+	    Tcl_Panic("SetByteCodeFromAny: unable to create link for tcl_traceCompile variable");
+	}
+	traceInitialized = 1;
+    }
+#endif
+
+    stringPtr = TclGetString(objPtr);
+    length = objPtr->length;
+
+    /*
+     * TIP #280: Pick up the CmdFrame in which the BC compiler was invoked and
+     * use to initialize the tracking in the compiler. This information was
+     * stored by TclCompEvalObj and ProcCompileProc.
+     */
+
+    TclInitCompileEnv(interp, &compEnv, stringPtr, length,
+	    iPtr->invokeCmdFramePtr, iPtr->invokeWord);
+
+    /*
+     * Now we check if we have data about invisible continuation lines for the
+     * script, and make it available to the compile environment, if so.
+     *
+     * It is not clear if the script Tcl_Obj* can be free'd while the compiler
+     * is using it, leading to the release of the associated ContLineLoc
+     * structure as well. To ensure that the latter doesn't happen we set a
+     * lock on it. We release this lock in the function TclFreeCompileEnv(),
+     * found in this file. The "lineCLPtr" hashtable is managed in the file
+     * "tclObj.c".
+     */
+
+    clLocPtr = TclContinuationsGet(objPtr);
+    if (clLocPtr) {
+	compEnv.clNext = &clLocPtr->loc[0];
+    }
+
+    TclCompileScript(interp, stringPtr, length, &compEnv);
+
+    /*
+     * Successful compilation. Add a "done" instruction at the end.
+     */
+
+    TclEmitOpcode(INST_DONE, &compEnv);
+
+    /*
+     * Check for optimizations!
+     *
+     * Test if the generated code is free of most hazards; if so, recompile
+     * but with generation of INST_START_CMD disabled. This produces somewhat
+     * faster code in some cases, and more compact code in more.
+     */
+
+    if (Tcl_GetMaster(interp) == NULL &&
+	    !Tcl_LimitTypeEnabled(interp, TCL_LIMIT_COMMANDS|TCL_LIMIT_TIME)
+	    && IsCompactibleCompileEnv(interp, &compEnv)) {
+	TclFreeCompileEnv(&compEnv);
+	iPtr->compiledProcPtr = procPtr;
+	TclInitCompileEnv(interp, &compEnv, stringPtr, length,
+		iPtr->invokeCmdFramePtr, iPtr->invokeWord);
+	if (clLocPtr) {
+	    compEnv.clNext = &clLocPtr->loc[0];
+	}
+	compEnv.atCmdStart = 2;		/* The disabling magic. */
+	TclCompileScript(interp, stringPtr, length, &compEnv);
+	assert (compEnv.atCmdStart > 1);
+	TclEmitOpcode(INST_DONE, &compEnv);
+	assert (compEnv.atCmdStart > 1);
+    }
+
+    /*
+     * Apply some peephole optimizations that can cross specific/generic
+     * instruction generator boundaries.
+     */
+
+    if (iPtr->extra.optimizer) {
+	(iPtr->extra.optimizer)(&compEnv);
+    }
+
+    /*
+     * Invoke the compilation hook procedure if one exists.
+     */
+
+    if (hookProc) {
+	result = hookProc(interp, &compEnv, clientData);
+    }
+
+    /*
+     * Change the object into a ByteCode object. Ownership of the literal
+     * objects and aux data items is given to the ByteCode object.
+     */
+
+#ifdef TCL_COMPILE_DEBUG
+    TclVerifyLocalLiteralTable(&compEnv);
+#endif /*TCL_COMPILE_DEBUG*/
+
+    if (result == TCL_OK) {
+	(void) TclInitByteCodeObj(objPtr, &tclByteCodeType, &compEnv);
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceCompile >= 2) {
+	    TclPrintByteCodeObj(interp, objPtr);
+	    fflush(stdout);
+	}
+#endif /* TCL_COMPILE_DEBUG */
+    }
+
+    TclFreeCompileEnv(&compEnv);
+    return result;
+}
+
+/*
+ *-----------------------------------------------------------------------
+ *
+ * SetByteCodeFromAny --
+ *
+ *	Part of the bytecode Tcl object type implementation. Attempts to
+ *	generate an byte code internal form for the Tcl object "objPtr" by
+ *	compiling its string representation.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during compilation, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	Frees the old internal representation. If no error occurs, then the
+ *	compiled code is stored as "objPtr"s bytecode representation. Also, if
+ *	debugging, initializes the "tcl_traceCompile" Tcl variable used to
+ *	trace compilations.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetByteCodeFromAny(
+    Tcl_Interp *interp,		/* The interpreter for which the code is being
+				 * compiled. Must not be NULL. */
+    Tcl_Obj *objPtr)		/* The object to make a ByteCode object. */
+{
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+    return TclSetByteCodeFromAny(interp, objPtr, NULL, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupByteCodeInternalRep --
+ *
+ *	Part of the bytecode Tcl object type implementation. However, it does
+ *	not copy the internal representation of a bytecode Tcl_Obj, but
+ *	instead leaves the new object untyped (with a NULL type pointer).
+ *	Code will be compiled for the new object only if necessary.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupByteCodeInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    Tcl_Obj *copyPtr)		/* Object with internal rep to set. */
+{
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeByteCodeInternalRep --
+ *
+ *	Part of the bytecode Tcl object type implementation. Frees the storage
+ *	associated with a bytecode object's internal representation unless its
+ *	code is actively being executed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The bytecode object's internal rep is marked invalid and its code gets
+ *	freed unless the code is actively being executed. In that case the
+ *	cleanup is delayed until the last execution of the code completes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeByteCodeInternalRep(
+    register Tcl_Obj *objPtr)	/* Object whose internal rep to free. */
+{
+    register ByteCode *codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclReleaseByteCode(codePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclReleaseByteCode --
+ *
+ *	This procedure does all the real work of freeing up a bytecode
+ *	object's ByteCode structure. It's called only when the structure's
+ *	reference count becomes zero.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees objPtr's bytecode internal representation and sets its type NULL
+ *	Also releases its literals and frees its auxiliary data items.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclPreserveByteCode(
+    register ByteCode *codePtr)
+{
+    codePtr->refCount++;
+}
+
+void
+TclReleaseByteCode(
+    register ByteCode *codePtr)
+{
+    if (codePtr->refCount-- > 1) {
+	return;
+    }
+
+    /* Just dropped to refcount==0.  Clean up. */
+    CleanupByteCode(codePtr);
+}
+
+static void
+CleanupByteCode(
+    register ByteCode *codePtr)	/* Points to the ByteCode to free. */
+{
+    Tcl_Interp *interp = (Tcl_Interp *) *codePtr->interpHandle;
+    Interp *iPtr = (Interp *) interp;
+    int numLitObjects = codePtr->numLitObjects;
+    int numAuxDataItems = codePtr->numAuxDataItems;
+    register Tcl_Obj **objArrayPtr, *objPtr;
+    register const AuxData *auxDataPtr;
+    int i;
+#ifdef TCL_COMPILE_STATS
+
+    if (interp != NULL) {
+	ByteCodeStats *statsPtr;
+	Tcl_Time destroyTime;
+	int lifetimeSec, lifetimeMicroSec, log2;
+
+	statsPtr = &iPtr->stats;
+
+	statsPtr->numByteCodesFreed++;
+	statsPtr->currentSrcBytes -= (double) codePtr->numSrcBytes;
+	statsPtr->currentByteCodeBytes -= (double) codePtr->structureSize;
+
+	statsPtr->currentInstBytes -= (double) codePtr->numCodeBytes;
+	statsPtr->currentLitBytes -= (double)
+		codePtr->numLitObjects * sizeof(Tcl_Obj *);
+	statsPtr->currentExceptBytes -= (double)
+		codePtr->numExceptRanges * sizeof(ExceptionRange);
+	statsPtr->currentAuxBytes -= (double)
+		codePtr->numAuxDataItems * sizeof(AuxData);
+	statsPtr->currentCmdMapBytes -= (double) codePtr->numCmdLocBytes;
+
+	Tcl_GetTime(&destroyTime);
+	lifetimeSec = destroyTime.sec - codePtr->createTime.sec;
+	if (lifetimeSec > 2000) {	/* avoid overflow */
+	    lifetimeSec = 2000;
+	}
+	lifetimeMicroSec = 1000000 * lifetimeSec +
+		(destroyTime.usec - codePtr->createTime.usec);
+
+	log2 = TclLog2(lifetimeMicroSec);
+	if (log2 > 31) {
+	    log2 = 31;
+	}
+	statsPtr->lifetimeCount[log2]++;
+    }
+#endif /* TCL_COMPILE_STATS */
+
+    /*
+     * A single heap object holds the ByteCode structure and its code, object,
+     * command location, and auxiliary data arrays. This means we only need to
+     * 1) decrement the ref counts of the LiteralEntry's in its literal array,
+     * 2) call the free procs for the auxiliary data items, 3) free the
+     * localCache if it is unused, and finally 4) free the ByteCode
+     * structure's heap object.
+     *
+     * The case for TCL_BYTECODE_PRECOMPILED (precompiled ByteCodes, like
+     * those generated from tbcload) is special, as they doesn't make use of
+     * the global literal table. They instead maintain private references to
+     * their literals which must be decremented.
+     *
+     * In order to insure a proper and efficient cleanup of the literal array
+     * when it contains non-shared literals [Bug 983660], we also distinguish
+     * the case of an interpreter being deleted (signaled by interp == NULL).
+     * Also, as the interp deletion will remove the global literal table
+     * anyway, we avoid the extra cost of updating it for each literal being
+     * released.
+     */
+
+    if (codePtr->flags & TCL_BYTECODE_PRECOMPILED) {
+
+	objArrayPtr = codePtr->objArrayPtr;
+	for (i = 0;  i < numLitObjects;  i++) {
+	    objPtr = *objArrayPtr;
+	    if (objPtr) {
+		Tcl_DecrRefCount(objPtr);
+	    }
+	    objArrayPtr++;
+	}
+	codePtr->numLitObjects = 0;
+    } else {
+	objArrayPtr = codePtr->objArrayPtr;
+	while (numLitObjects--) {
+	    /* TclReleaseLiteral calls Tcl_DecrRefCount() for us */
+	    TclReleaseLiteral(interp, *objArrayPtr++);
+	}
+    }
+
+    auxDataPtr = codePtr->auxDataArrayPtr;
+    for (i = 0;  i < numAuxDataItems;  i++) {
+	if (auxDataPtr->type->freeProc != NULL) {
+	    auxDataPtr->type->freeProc(auxDataPtr->clientData);
+	}
+	auxDataPtr++;
+    }
+
+    /*
+     * TIP #280. Release the location data associated with this byte code
+     * structure, if any. NOTE: The interp we belong to may be gone already,
+     * and the data with it.
+     *
+     * See also tclBasic.c, DeleteInterpProc
+     */
+
+    if (iPtr) {
+	Tcl_HashEntry *hePtr = Tcl_FindHashEntry(iPtr->lineBCPtr,
+		(char *) codePtr);
+
+	if (hePtr) {
+	    ReleaseCmdWordData(Tcl_GetHashValue(hePtr));
+	    Tcl_DeleteHashEntry(hePtr);
+	}
+    }
+
+    if (codePtr->localCachePtr && (--codePtr->localCachePtr->refCount == 0)) {
+	TclFreeLocalCache(interp, codePtr->localCachePtr);
+    }
+
+    TclHandleRelease(codePtr->interpHandle);
+    ckfree(codePtr);
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * IsCompactibleCompileEnv --
+ *
+ *	Checks to see if we may apply some basic compaction optimizations to a
+ *	piece of bytecode. Idempotent.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+static int
+IsCompactibleCompileEnv(
+    Tcl_Interp *interp,
+    CompileEnv *envPtr)
+{
+    unsigned char *pc;
+    int size;
+
+    /*
+     * Special: procedures in the '::tcl' namespace (or its children) are
+     * considered to be well-behaved and so can have compaction applied even
+     * if it would otherwise be invalid.
+     */
+
+    if (envPtr->procPtr != NULL && envPtr->procPtr->cmdPtr != NULL
+	    && envPtr->procPtr->cmdPtr->nsPtr != NULL) {
+	Namespace *nsPtr = envPtr->procPtr->cmdPtr->nsPtr;
+
+	if (strcmp(nsPtr->fullName, "::tcl") == 0
+		|| strncmp(nsPtr->fullName, "::tcl::", 7) == 0) {
+	    return 1;
+	}
+    }
+
+    /*
+     * Go through and ensure that no operation involved can cause a desired
+     * change of bytecode sequence during running. This comes down to ensuring
+     * that there are no mapped variables (due to traces) or calls to external
+     * commands (traces, [uplevel] trickery). This is actually a very
+     * conservative check; it turns down a lot of code that is OK in practice.
+     */
+
+    for (pc = envPtr->codeStart ; pc < envPtr->codeNext ; pc += size) {
+	switch (*pc) {
+	    /* Invokes */
+	case INST_INVOKE_STK1:
+	case INST_INVOKE_STK4:
+	case INST_INVOKE_EXPANDED:
+	case INST_INVOKE_REPLACE:
+	    return 0;
+	    /* Runtime evals */
+	case INST_EVAL_STK:
+	case INST_EXPR_STK:
+	case INST_YIELD:
+	    return 0;
+	    /* Upvars */
+	case INST_UPVAR:
+	case INST_NSUPVAR:
+	case INST_VARIABLE:
+	    return 0;
+	default:
+	    size = tclInstructionTable[*pc].numBytes;
+	    assert (size > 0);
+	    break;
+	}
+    }
+
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SubstObj --
+ *
+ *	This function performs the substitutions specified on the given string
+ *	as described in the user documentation for the "subst" Tcl command.
+ *
+ * Results:
+ *	A Tcl_Obj* containing the substituted string, or NULL to indicate that
+ *	an error occurred.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_SubstObj(
+    Tcl_Interp *interp,		/* Interpreter in which substitution occurs */
+    Tcl_Obj *objPtr,		/* The value to be substituted. */
+    int flags)			/* What substitutions to do. */
+{
+    NRE_callback *rootPtr = TOP_CB(interp);
+
+    if (TclNRRunCallbacks(interp, Tcl_NRSubstObj(interp, objPtr, flags),
+	    rootPtr) != TCL_OK) {
+	return NULL;
+    }
+    return Tcl_GetObjResult(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NRSubstObj --
+ *
+ *	Request substitution of a Tcl value by the NR stack.
+ *
+ * Results:
+ *	Returns TCL_OK.
+ *
+ * Side effects:
+ *	Compiles objPtr into bytecode that performs the substitutions as
+ *	governed by flags and places callbacks on the NR stack to execute
+ *	the bytecode and store the result in the interp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_NRSubstObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    int flags)
+{
+    ByteCode *codePtr = CompileSubstObj(interp, objPtr, flags);
+
+    /* TODO: Confirm we do not need this. */
+    /* Tcl_ResetResult(interp); */
+    return TclNRExecuteByteCode(interp, codePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileSubstObj --
+ *
+ *	Compile a Tcl value into ByteCode implementing its substitution, as
+ *	governed by flags.
+ *
+ * Results:
+ *	A (ByteCode *) is returned pointing to the resulting ByteCode.
+ *
+ * Side effects:
+ *	The Tcl_ObjType of objPtr is changed to the "substcode" type, and the
+ *	ByteCode and governing flags value are kept in the internal rep for
+ *	faster operations the next time CompileSubstObj is called on the same
+ *	value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ByteCode *
+CompileSubstObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    int flags)
+{
+    Interp *iPtr = (Interp *) interp;
+    ByteCode *codePtr = NULL;
+
+    if (objPtr->typePtr == &substCodeType) {
+	Namespace *nsPtr = iPtr->varFramePtr->nsPtr;
+
+	codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	if (flags != PTR2INT(objPtr->internalRep.twoPtrValue.ptr2)
+		|| ((Interp *) *codePtr->interpHandle != iPtr)
+		|| (codePtr->compileEpoch != iPtr->compileEpoch)
+		|| (codePtr->nsPtr != nsPtr)
+		|| (codePtr->nsEpoch != nsPtr->resolverEpoch)
+		|| (codePtr->localCachePtr !=
+		iPtr->varFramePtr->localCachePtr)) {
+	    TclFreeIntRep(objPtr);
+	}
+    }
+    if (objPtr->typePtr != &substCodeType) {
+	CompileEnv compEnv;
+	int numBytes;
+	const char *bytes = TclGetStringFromObj(objPtr, &numBytes);
+
+	/* TODO: Check for more TIP 280 */
+	TclInitCompileEnv(interp, &compEnv, bytes, numBytes, NULL, 0);
+
+	TclSubstCompile(interp, bytes, numBytes, flags, 1, &compEnv);
+
+	TclEmitOpcode(INST_DONE, &compEnv);
+	codePtr = TclInitByteCodeObj(objPtr, &substCodeType, &compEnv);
+	TclFreeCompileEnv(&compEnv);
+
+	objPtr->internalRep.twoPtrValue.ptr1 = codePtr;
+	objPtr->internalRep.twoPtrValue.ptr2 = INT2PTR(flags);
+	if (iPtr->varFramePtr->localCachePtr) {
+	    codePtr->localCachePtr = iPtr->varFramePtr->localCachePtr;
+	    codePtr->localCachePtr->refCount++;
+	}
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceCompile >= 2) {
+	    TclPrintByteCodeObj(interp, objPtr);
+	    fflush(stdout);
+	}
+#endif /* TCL_COMPILE_DEBUG */
+    }
+    return codePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeSubstCodeInternalRep --
+ *
+ *	Part of the substcode Tcl object type implementation. Frees the
+ *	storage associated with a substcode object's internal representation
+ *	unless its code is actively being executed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The substcode object's internal rep is marked invalid and its code
+ *	gets freed unless the code is actively being executed. In that case
+ *	the cleanup is delayed until the last execution of the code completes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeSubstCodeInternalRep(
+    register Tcl_Obj *objPtr)	/* Object whose internal rep to free. */
+{
+    register ByteCode *codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclReleaseByteCode(codePtr);
+}
+
+static void
+ReleaseCmdWordData(
+    ExtCmdLoc *eclPtr)
+{
+    int i;
+
+    if (eclPtr->type == TCL_LOCATION_SOURCE) {
+	Tcl_DecrRefCount(eclPtr->path);
+    }
+    for (i=0 ; i<eclPtr->nuloc ; i++) {
+	ckfree(eclPtr->loc[i].line);
+    }
+
+    if (eclPtr->loc != NULL) {
+	ckfree(eclPtr->loc);
+    }
+
+    ckfree(eclPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitCompileEnv --
+ *
+ *	Initializes a CompileEnv compilation environment structure for the
+ *	compilation of a string in an interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The CompileEnv structure is initialized.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitCompileEnv(
+    Tcl_Interp *interp,		/* The interpreter for which a CompileEnv
+				 * structure is initialized. */
+    register CompileEnv *envPtr,/* Points to the CompileEnv structure to
+				 * initialize. */
+    const char *stringPtr,	/* The source string to be compiled. */
+    int numBytes,		/* Number of bytes in source string. */
+    const CmdFrame *invoker,	/* Location context invoking the bcc */
+    int word)			/* Index of the word in that context getting
+				 * compiled */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    assert(tclInstructionTable[LAST_INST_OPCODE+1].name == NULL);
+
+    envPtr->iPtr = iPtr;
+    envPtr->source = stringPtr;
+    envPtr->numSrcBytes = numBytes;
+    envPtr->procPtr = iPtr->compiledProcPtr;
+    iPtr->compiledProcPtr = NULL;
+    envPtr->numCommands = 0;
+    envPtr->exceptDepth = 0;
+    envPtr->maxExceptDepth = 0;
+    envPtr->maxStackDepth = 0;
+    envPtr->currStackDepth = 0;
+    TclInitLiteralTable(&envPtr->localLitTable);
+
+    envPtr->codeStart = envPtr->staticCodeSpace;
+    envPtr->codeNext = envPtr->codeStart;
+    envPtr->codeEnd = envPtr->codeStart + COMPILEENV_INIT_CODE_BYTES;
+    envPtr->mallocedCodeArray = 0;
+
+    envPtr->literalArrayPtr = envPtr->staticLiteralSpace;
+    envPtr->literalArrayNext = 0;
+    envPtr->literalArrayEnd = COMPILEENV_INIT_NUM_OBJECTS;
+    envPtr->mallocedLiteralArray = 0;
+
+    envPtr->exceptArrayPtr = envPtr->staticExceptArraySpace;
+    envPtr->exceptAuxArrayPtr = envPtr->staticExAuxArraySpace;
+    envPtr->exceptArrayNext = 0;
+    envPtr->exceptArrayEnd = COMPILEENV_INIT_EXCEPT_RANGES;
+    envPtr->mallocedExceptArray = 0;
+
+    envPtr->cmdMapPtr = envPtr->staticCmdMapSpace;
+    envPtr->cmdMapEnd = COMPILEENV_INIT_CMD_MAP_SIZE;
+    envPtr->mallocedCmdMap = 0;
+    envPtr->atCmdStart = 1;
+    envPtr->expandCount = 0;
+
+    /*
+     * TIP #280: Set up the extended command location information, based on
+     * the context invoking the byte code compiler. This structure is used to
+     * keep the per-word line information for all compiled commands.
+     *
+     * See also tclBasic.c, TclEvalObjEx, for the equivalent code in the
+     * non-compiling evaluator
+     */
+
+    envPtr->extCmdMapPtr = ckalloc(sizeof(ExtCmdLoc));
+    envPtr->extCmdMapPtr->loc = NULL;
+    envPtr->extCmdMapPtr->nloc = 0;
+    envPtr->extCmdMapPtr->nuloc = 0;
+    envPtr->extCmdMapPtr->path = NULL;
+
+    if (invoker == NULL) {
+	/*
+	 * Initialize the compiler for relative counting in case of a
+	 * dynamic context.
+	 */
+
+	envPtr->line = 1;
+	if (iPtr->evalFlags & TCL_EVAL_FILE) {
+	    iPtr->evalFlags &= ~TCL_EVAL_FILE;
+	    envPtr->extCmdMapPtr->type = TCL_LOCATION_SOURCE;
+
+	    if (iPtr->scriptFile) {
+		/*
+		 * Normalization here, to have the correct pwd. Should have
+		 * negligible impact on performance, as the norm should have
+		 * been done already by the 'source' invoking us, and it
+		 * caches the result.
+		 */
+
+		Tcl_Obj *norm =
+			Tcl_FSGetNormalizedPath(interp, iPtr->scriptFile);
+
+		if (norm == NULL) {
+		    /*
+		     * Error message in the interp result. No place to put it.
+		     * And no place to serve the error itself to either. Fake
+		     * a path, empty string.
+		     */
+
+		    TclNewLiteralStringObj(envPtr->extCmdMapPtr->path, "");
+		} else {
+		    envPtr->extCmdMapPtr->path = norm;
+		}
+	    } else {
+		TclNewLiteralStringObj(envPtr->extCmdMapPtr->path, "");
+	    }
+
+	    Tcl_IncrRefCount(envPtr->extCmdMapPtr->path);
+	} else {
+	    envPtr->extCmdMapPtr->type =
+		(envPtr->procPtr ? TCL_LOCATION_PROC : TCL_LOCATION_BC);
+	}
+    } else {
+	/*
+	 * Initialize the compiler using the context, making counting absolute
+	 * to that context. Note that the context can be byte code execution.
+	 * In that case we have to fill out the missing pieces (line, path,
+	 * ...) which may make change the type as well.
+	 */
+
+	CmdFrame *ctxPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+	int pc = 0;
+
+	*ctxPtr = *invoker;
+	if (invoker->type == TCL_LOCATION_BC) {
+	    /*
+	     * Note: Type BC => ctx.data.eval.path    is not used.
+	     *			ctx.data.tebc.codePtr is used instead.
+	     */
+
+	    TclGetSrcInfoForPc(ctxPtr);
+	    pc = 1;
+	}
+
+	if ((ctxPtr->nline <= word) || (ctxPtr->line[word] < 0)) {
+	    /*
+	     * Word is not a literal, relative counting.
+	     */
+
+	    envPtr->line = 1;
+	    envPtr->extCmdMapPtr->type =
+		    (envPtr->procPtr ? TCL_LOCATION_PROC : TCL_LOCATION_BC);
+
+	    if (pc && (ctxPtr->type == TCL_LOCATION_SOURCE)) {
+		/*
+		 * The reference made by 'TclGetSrcInfoForPc' is dead.
+		 */
+
+		Tcl_DecrRefCount(ctxPtr->data.eval.path);
+	    }
+	} else {
+	    envPtr->line = ctxPtr->line[word];
+	    envPtr->extCmdMapPtr->type = ctxPtr->type;
+
+	    if (ctxPtr->type == TCL_LOCATION_SOURCE) {
+		envPtr->extCmdMapPtr->path = ctxPtr->data.eval.path;
+
+		if (pc) {
+		    /*
+		     * The reference 'TclGetSrcInfoForPc' made is transfered.
+		     */
+
+		    ctxPtr->data.eval.path = NULL;
+		} else {
+		    /*
+		     * We have a new reference here.
+		     */
+
+		    Tcl_IncrRefCount(envPtr->extCmdMapPtr->path);
+		}
+	    }
+	}
+
+	TclStackFree(interp, ctxPtr);
+    }
+
+    envPtr->extCmdMapPtr->start = envPtr->line;
+
+    /*
+     * Initialize the data about invisible continuation lines as empty, i.e.
+     * not used. The caller (TclSetByteCodeFromAny) will set this up, if such
+     * data is available.
+     */
+
+    envPtr->clNext = NULL;
+
+    envPtr->auxDataArrayPtr = envPtr->staticAuxDataArraySpace;
+    envPtr->auxDataArrayNext = 0;
+    envPtr->auxDataArrayEnd = COMPILEENV_INIT_AUX_DATA_SIZE;
+    envPtr->mallocedAuxDataArray = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreeCompileEnv --
+ *
+ *	Free the storage allocated in a CompileEnv compilation environment
+ *	structure.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocated storage in the CompileEnv structure is freed. Note that its
+ *	local literal table is not deleted and its literal objects are not
+ *	released. In addition, storage referenced by its auxiliary data items
+ *	is not freed. This is done so that, when compilation is successful,
+ *	"ownership" of these objects and aux data items is handed over to the
+ *	corresponding ByteCode structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFreeCompileEnv(
+    register CompileEnv *envPtr)/* Points to the CompileEnv structure. */
+{
+    if (envPtr->localLitTable.buckets != envPtr->localLitTable.staticBuckets){
+	ckfree(envPtr->localLitTable.buckets);
+	envPtr->localLitTable.buckets = envPtr->localLitTable.staticBuckets;
+    }
+    if (envPtr->iPtr) {
+	/*
+	 * We never converted to Bytecode, so free the things we would
+	 * have transferred to it.
+	 */
+
+	int i;
+	LiteralEntry *entryPtr = envPtr->literalArrayPtr;
+	AuxData *auxDataPtr = envPtr->auxDataArrayPtr;
+
+	for (i = 0;  i < envPtr->literalArrayNext;  i++) {
+	    TclReleaseLiteral((Tcl_Interp *)envPtr->iPtr, entryPtr->objPtr);
+	    entryPtr++;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	TclVerifyGlobalLiteralTable(envPtr->iPtr);
+#endif /*TCL_COMPILE_DEBUG*/
+
+	for (i = 0;  i < envPtr->auxDataArrayNext;  i++) {
+	    if (auxDataPtr->type->freeProc != NULL) {
+		auxDataPtr->type->freeProc(auxDataPtr->clientData);
+	    }
+	    auxDataPtr++;
+	}
+    }
+    if (envPtr->mallocedCodeArray) {
+	ckfree(envPtr->codeStart);
+    }
+    if (envPtr->mallocedLiteralArray) {
+	ckfree(envPtr->literalArrayPtr);
+    }
+    if (envPtr->mallocedExceptArray) {
+	ckfree(envPtr->exceptArrayPtr);
+	ckfree(envPtr->exceptAuxArrayPtr);
+    }
+    if (envPtr->mallocedCmdMap) {
+	ckfree(envPtr->cmdMapPtr);
+    }
+    if (envPtr->mallocedAuxDataArray) {
+	ckfree(envPtr->auxDataArrayPtr);
+    }
+    if (envPtr->extCmdMapPtr) {
+	ReleaseCmdWordData(envPtr->extCmdMapPtr);
+	envPtr->extCmdMapPtr = NULL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclWordKnownAtCompileTime --
+ *
+ *	Test whether the value of a token is completely known at compile time.
+ *
+ * Results:
+ *	Returns true if the tokenPtr argument points to a word value that is
+ *	completely known at compile time. Generally, values that are known at
+ *	compile time can be compiled to their values, while values that cannot
+ *	be known until substitution at runtime must be compiled to bytecode
+ *	instructions that perform that substitution. For several commands,
+ *	whether or not arguments are known at compile time determine whether
+ *	it is worthwhile to compile at all.
+ *
+ * Side effects:
+ *	When returning true, appends the known value of the word to the
+ *	unshared Tcl_Obj (*valuePtr), unless valuePtr is NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclWordKnownAtCompileTime(
+    Tcl_Token *tokenPtr,	/* Points to Tcl_Token we should check */
+    Tcl_Obj *valuePtr)		/* If not NULL, points to an unshared Tcl_Obj
+				 * to which we should append the known value
+				 * of the word. */
+{
+    int numComponents = tokenPtr->numComponents;
+    Tcl_Obj *tempPtr = NULL;
+
+    if (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	if (valuePtr != NULL) {
+	    Tcl_AppendToObj(valuePtr, tokenPtr[1].start, tokenPtr[1].size);
+	}
+	return 1;
+    }
+    if (tokenPtr->type != TCL_TOKEN_WORD) {
+	return 0;
+    }
+    tokenPtr++;
+    if (valuePtr != NULL) {
+	tempPtr = Tcl_NewObj();
+	Tcl_IncrRefCount(tempPtr);
+    }
+    while (numComponents--) {
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_TEXT:
+	    if (tempPtr != NULL) {
+		Tcl_AppendToObj(tempPtr, tokenPtr->start, tokenPtr->size);
+	    }
+	    break;
+
+	case TCL_TOKEN_BS:
+	    if (tempPtr != NULL) {
+		char utfBuf[TCL_UTF_MAX];
+		int length = TclParseBackslash(tokenPtr->start,
+			tokenPtr->size, NULL, utfBuf);
+
+		Tcl_AppendToObj(tempPtr, utfBuf, length);
+	    }
+	    break;
+
+	default:
+	    if (tempPtr != NULL) {
+		Tcl_DecrRefCount(tempPtr);
+	    }
+	    return 0;
+	}
+	tokenPtr++;
+    }
+    if (valuePtr != NULL) {
+	Tcl_AppendObjToObj(valuePtr, tempPtr);
+	Tcl_DecrRefCount(tempPtr);
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileScript --
+ *
+ *	Compile a Tcl script in a string.
+ *
+ * Results:
+ *	The return value is TCL_OK on a successful compilation and TCL_ERROR
+ *	on failure. If TCL_ERROR is returned, then the interpreter's result
+ *	contains an error message.
+ *
+ * Side effects:
+ *	Adds instructions to envPtr to evaluate the script at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ExpandRequested(
+    Tcl_Token *tokenPtr,
+    int numWords)
+{
+    /* Determine whether any words of the command require expansion */
+    while (numWords--) {
+	if (tokenPtr->type == TCL_TOKEN_EXPAND_WORD) {
+	    return 1;
+	}
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+    return 0;
+}
+
+static void
+CompileCmdLiteral(
+    Tcl_Interp *interp,
+    Tcl_Obj *cmdObj,
+    CompileEnv *envPtr)
+{
+    int numBytes;
+    const char *bytes;
+    Command *cmdPtr;
+    int cmdLitIdx, extraLiteralFlags = LITERAL_CMD_NAME;
+
+    cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, cmdObj);
+    if ((cmdPtr != NULL) && (cmdPtr->flags & CMD_VIA_RESOLVER)) {
+	extraLiteralFlags |= LITERAL_UNSHARED;
+    }
+
+    bytes = TclGetStringFromObj(cmdObj, &numBytes);
+    cmdLitIdx = TclRegisterLiteral(envPtr, bytes, numBytes, extraLiteralFlags);
+
+    if (cmdPtr) {
+	TclSetCmdNameObj(interp, TclFetchLiteral(envPtr, cmdLitIdx), cmdPtr);
+    }
+    TclEmitPush(cmdLitIdx, envPtr);
+}
+
+void
+TclCompileInvocation(
+    Tcl_Interp *interp,
+    Tcl_Token *tokenPtr,
+    Tcl_Obj *cmdObj,
+    int numWords,
+    CompileEnv *envPtr)
+{
+    int wordIdx = 0, depth = TclGetStackDepth(envPtr);
+    DefineLineInformation;
+
+    if (cmdObj) {
+	CompileCmdLiteral(interp, cmdObj, envPtr);
+	wordIdx = 1;
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    for (; wordIdx < numWords; wordIdx++, tokenPtr = TokenAfter(tokenPtr)) {
+	int objIdx;
+
+	SetLineInformation(wordIdx);
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    CompileTokens(envPtr, tokenPtr, interp);
+	    continue;
+	}
+
+	objIdx = TclRegisterLiteral(envPtr,
+		tokenPtr[1].start, tokenPtr[1].size, 0);
+	if (envPtr->clNext) {
+	    TclContinuationsEnterDerived(TclFetchLiteral(envPtr, objIdx),
+		    tokenPtr[1].start - envPtr->source, envPtr->clNext);
+	}
+	TclEmitPush(objIdx, envPtr);
+    }
+
+    if (wordIdx <= 255) {
+	TclEmitInvoke(envPtr, INST_INVOKE_STK1, wordIdx);
+    } else {
+	TclEmitInvoke(envPtr, INST_INVOKE_STK4, wordIdx);
+    }
+    TclCheckStackDepth(depth+1, envPtr);
+}
+
+static void
+CompileExpanded(
+    Tcl_Interp *interp,
+    Tcl_Token *tokenPtr,
+    Tcl_Obj *cmdObj,
+    int numWords,
+    CompileEnv *envPtr)
+{
+    int wordIdx = 0;
+    DefineLineInformation;
+    int depth = TclGetStackDepth(envPtr);
+
+    StartExpanding(envPtr);
+    if (cmdObj) {
+	CompileCmdLiteral(interp, cmdObj, envPtr);
+	wordIdx = 1;
+	tokenPtr = TokenAfter(tokenPtr);
+    }
+
+    for (; wordIdx < numWords; wordIdx++, tokenPtr = TokenAfter(tokenPtr)) {
+	int objIdx;
+
+	SetLineInformation(wordIdx);
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    CompileTokens(envPtr, tokenPtr, interp);
+	    if (tokenPtr->type == TCL_TOKEN_EXPAND_WORD) {
+		TclEmitInstInt4(INST_EXPAND_STKTOP,
+			envPtr->currStackDepth, envPtr);
+	    }
+	    continue;
+	}
+
+	objIdx = TclRegisterLiteral(envPtr,
+		tokenPtr[1].start, tokenPtr[1].size, 0);
+	if (envPtr->clNext) {
+	    TclContinuationsEnterDerived(TclFetchLiteral(envPtr, objIdx),
+		    tokenPtr[1].start - envPtr->source, envPtr->clNext);
+	}
+	TclEmitPush(objIdx, envPtr);
+    }
+
+    /*
+     * The stack depth during argument expansion can only be managed at
+     * runtime, as the number of elements in the expanded lists is not known
+     * at compile time. We adjust here the stack depth estimate so that it is
+     * correct after the command with expanded arguments returns.
+     *
+     * The end effect of this command's invocation is that all the words of
+     * the command are popped from the stack, and the result is pushed: the
+     * stack top changes by (1-wordIdx).
+     *
+     * Note that the estimates are not correct while the command is being
+     * prepared and run, INST_EXPAND_STKTOP is not stack-neutral in general.
+     */
+
+    TclEmitInvoke(envPtr, INST_INVOKE_EXPANDED, wordIdx);
+    TclCheckStackDepth(depth+1, envPtr);
+}
+
+static int
+CompileCmdCompileProc(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Command *cmdPtr,
+    CompileEnv *envPtr)
+{
+    int unwind = 0, incrOffset = -1;
+    DefineLineInformation;
+    int depth = TclGetStackDepth(envPtr);
+
+    /*
+     * Emit of the INST_START_CMD instruction is controlled by the value of
+     * envPtr->atCmdStart:
+     *
+     * atCmdStart == 2	: We are not using the INST_START_CMD instruction.
+     * atCmdStart == 1	: INST_START_CMD was the last instruction emitted.
+     *			: We do not need to emit another.  Instead we
+     *			: increment the number of cmds started at it (except
+     *			: for the special case at the start of a script.)
+     * atCmdStart == 0	: The last instruction was something else.  We need
+     *			: to emit INST_START_CMD here.
+     */
+
+    switch (envPtr->atCmdStart) {
+    case 0:
+	unwind = tclInstructionTable[INST_START_CMD].numBytes;
+	TclEmitInstInt4(INST_START_CMD, 0, envPtr);
+	incrOffset = envPtr->codeNext - envPtr->codeStart;
+	TclEmitInt4(0, envPtr);
+	break;
+    case 1:
+	if (envPtr->codeNext > envPtr->codeStart) {
+	    incrOffset = envPtr->codeNext - 4 - envPtr->codeStart;
+	}
+	break;
+    case 2:
+	/* Nothing to do */
+	;
+    }
+
+    if (TCL_OK == TclAttemptCompileProc(interp, parsePtr, 1, cmdPtr, envPtr)) {
+	if (incrOffset >= 0) {
+	    /*
+	     * We successfully compiled a command.  Increment the number of
+	     * commands that start at the currently active INST_START_CMD.
+	     */
+
+	    unsigned char *incrPtr = envPtr->codeStart + incrOffset;
+	    unsigned char *startPtr = incrPtr - 5;
+
+	    TclIncrUInt4AtPtr(incrPtr, 1);
+	    if (unwind) {
+		/* We started the INST_START_CMD.  Record the code length. */
+		TclStoreInt4AtPtr(envPtr->codeNext - startPtr, startPtr + 1);
+	    }
+	}
+	TclCheckStackDepth(depth+1, envPtr);
+	return TCL_OK;
+    }
+
+    envPtr->codeNext -= unwind; /* Unwind INST_START_CMD */
+
+    /*
+     * Throw out any line information generated by the failed compile attempt.
+     */
+
+    while (mapPtr->nuloc - 1 > eclIndex) {
+	mapPtr->nuloc--;
+	ckfree(mapPtr->loc[mapPtr->nuloc].line);
+	mapPtr->loc[mapPtr->nuloc].line = NULL;
+    }
+
+    /*
+     * Reset the index of next command.  Toss out any from failed nested
+     * partial compiles.
+     */
+
+    envPtr->numCommands = mapPtr->nuloc;
+    return TCL_ERROR;
+}
+
+static int
+CompileCommandTokens(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    CompileEnv *envPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Token *tokenPtr = parsePtr->tokenPtr;
+    ExtCmdLoc *eclPtr = envPtr->extCmdMapPtr;
+    Tcl_Obj *cmdObj = Tcl_NewObj();
+    Command *cmdPtr = NULL;
+    int code = TCL_ERROR;
+    int cmdKnown, expand = -1;
+    int *wlines, wlineat;
+    int cmdLine = envPtr->line;
+    int *clNext = envPtr->clNext;
+    int cmdIdx = envPtr->numCommands;
+    int startCodeOffset = envPtr->codeNext - envPtr->codeStart;
+    int depth = TclGetStackDepth(envPtr);
+
+    assert (parsePtr->numWords > 0);
+
+    /* Pre-Compile */
+
+    envPtr->numCommands++;
+    EnterCmdStartData(envPtr, cmdIdx,
+	    parsePtr->commandStart - envPtr->source, startCodeOffset);
+
+    /*
+     * TIP #280. Scan the words and compute the extended location information.
+     * The map first contain full per-word line information for use by the
+     * compiler. This is later replaced by a reduced form which signals
+     * non-literal words, stored in 'wlines'.
+     */
+
+    EnterCmdWordData(eclPtr, parsePtr->commandStart - envPtr->source,
+	    parsePtr->tokenPtr, parsePtr->commandStart,
+	    parsePtr->commandSize, parsePtr->numWords, cmdLine,
+	    clNext, &wlines, envPtr);
+    wlineat = eclPtr->nuloc - 1;
+
+    envPtr->line = eclPtr->loc[wlineat].line[0];
+    envPtr->clNext = eclPtr->loc[wlineat].next[0];
+
+    /* Do we know the command word? */
+    Tcl_IncrRefCount(cmdObj);
+    tokenPtr = parsePtr->tokenPtr;
+    cmdKnown = TclWordKnownAtCompileTime(tokenPtr, cmdObj);
+
+    /* Is this a command we should (try to) compile with a compileProc ? */
+    if (cmdKnown && !(iPtr->flags & DONT_COMPILE_CMDS_INLINE)) {
+	cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, cmdObj);
+	if (cmdPtr) {
+	    /*
+	     * Found a command.  Test the ways we can be told not to attempt
+	     * to compile it.
+	     */
+	    if ((cmdPtr->compileProc == NULL)
+		    || (cmdPtr->nsPtr->flags & NS_SUPPRESS_COMPILATION)
+		    || (cmdPtr->flags & CMD_HAS_EXEC_TRACES)) {
+		cmdPtr = NULL;
+	    }
+	}
+	if (cmdPtr && !(cmdPtr->flags & CMD_COMPILES_EXPANDED)) {
+	    expand = ExpandRequested(parsePtr->tokenPtr, parsePtr->numWords);
+	    if (expand) {
+		/* We need to expand, but compileProc cannot. */
+		cmdPtr = NULL;
+	    }
+	}
+    }
+
+    /* If cmdPtr != NULL, we will try to call cmdPtr->compileProc */
+    if (cmdPtr) {
+	code = CompileCmdCompileProc(interp, parsePtr, cmdPtr, envPtr);
+    }
+
+    if (code == TCL_ERROR) {
+	if (expand < 0) {
+	    expand = ExpandRequested(parsePtr->tokenPtr, parsePtr->numWords);
+	}
+
+	if (expand) {
+	    CompileExpanded(interp, parsePtr->tokenPtr,
+		    cmdKnown ? cmdObj : NULL, parsePtr->numWords, envPtr);
+	} else {
+	    TclCompileInvocation(interp, parsePtr->tokenPtr,
+		    cmdKnown ? cmdObj : NULL, parsePtr->numWords, envPtr);
+	}
+    }
+
+    Tcl_DecrRefCount(cmdObj);
+
+    TclEmitOpcode(INST_POP, envPtr);
+    EnterCmdExtentData(envPtr, cmdIdx,
+	    parsePtr->term - parsePtr->commandStart,
+	    (envPtr->codeNext-envPtr->codeStart) - startCodeOffset);
+
+    /*
+     * TIP #280: Free full form of per-word line data and insert the reduced
+     * form now
+     */
+
+    envPtr->line = cmdLine;
+    envPtr->clNext = clNext;
+    ckfree(eclPtr->loc[wlineat].line);
+    ckfree(eclPtr->loc[wlineat].next);
+    eclPtr->loc[wlineat].line = wlines;
+    eclPtr->loc[wlineat].next = NULL;
+
+    TclCheckStackDepth(depth, envPtr);
+    return cmdIdx;
+}
+
+void
+TclCompileScript(
+    Tcl_Interp *interp,		/* Used for error and status reporting. Also
+				 * serves as context for finding and compiling
+				 * commands. May not be NULL. */
+    const char *script,		/* The source script to compile. */
+    int numBytes,		/* Number of bytes in script. If < 0, the
+				 * script consists of all bytes up to the
+				 * first null character. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    int lastCmdIdx = -1;	/* Index into envPtr->cmdMapPtr of the last
+				 * command this routine compiles into bytecode.
+				 * Initial value of -1 indicates this routine
+				 * has not yet generated any bytecode. */
+    const char *p = script;	/* Where we are in our compile. */
+    int depth = TclGetStackDepth(envPtr);
+
+    if (envPtr->iPtr == NULL) {
+	Tcl_Panic("TclCompileScript() called on uninitialized CompileEnv");
+    }
+
+    /* Each iteration compiles one command from the script. */
+
+    while (numBytes > 0) {
+	Tcl_Parse parse;
+	const char *next;
+
+	if (TCL_OK != Tcl_ParseCommand(interp, p, numBytes, 0, &parse)) {
+	    /*
+	     * Compile bytecodes to report the parse error at runtime.
+	     */
+
+	    Tcl_LogCommandInfo(interp, script, parse.commandStart,
+		    parse.term + 1 - parse.commandStart);
+	    TclCompileSyntaxError(interp, envPtr);
+	    return;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	/*
+	 * If tracing, print a line for each top level command compiled.
+	 * TODO: Suppress when numWords == 0 ?
+	 */
+
+	if ((tclTraceCompile >= 1) && (envPtr->procPtr == NULL)) {
+	    int commandLength = parse.term - parse.commandStart;
+	    fprintf(stdout, "  Compiling: ");
+	    TclPrintSource(stdout, parse.commandStart,
+		    TclMin(commandLength, 55));
+	    fprintf(stdout, "\n");
+	}
+#endif
+
+	/*
+	 * TIP #280: Count newlines before the command start.
+	 * (See test info-30.33).
+	 */
+
+	TclAdvanceLines(&envPtr->line, p, parse.commandStart);
+	TclAdvanceContinuations(&envPtr->line, &envPtr->clNext,
+		parse.commandStart - envPtr->source);
+
+	/*
+	 * Advance parser to the next command in the script.
+	 */
+
+	next = parse.commandStart + parse.commandSize;
+	numBytes -= next - p;
+	p = next;
+
+	if (parse.numWords == 0) {
+	    /*
+	     * The "command" parsed has no words.  In this case we can skip
+	     * the rest of the loop body.  With no words, clearly
+	     * CompileCommandTokens() has nothing to do.  Since the parser
+	     * aggressively sucks up leading comment and white space,
+	     * including newlines, parse.commandStart must be pointing at
+	     * either the end of script, or a command-terminating semi-colon.
+	     * In either case, the TclAdvance*() calls have nothing to do.
+	     * Finally, when no words are parsed, no tokens have been
+	     * allocated at parse.tokenPtr so there's also nothing for
+	     * Tcl_FreeParse() to do.
+	     *
+	     * The advantage of this shortcut is that CompileCommandTokens()
+	     * can be written with an assumption that parse.numWords > 0, with
+	     * the implication the CCT() always generates bytecode.
+	     */
+	    continue;
+	}
+
+	lastCmdIdx = CompileCommandTokens(interp, &parse, envPtr);
+
+	/*
+	 * TIP #280: Track lines in the just compiled command.
+	 */
+
+	TclAdvanceLines(&envPtr->line, parse.commandStart, p);
+	TclAdvanceContinuations(&envPtr->line, &envPtr->clNext,
+		p - envPtr->source);
+	Tcl_FreeParse(&parse);
+    }
+
+    if (lastCmdIdx == -1) {
+	/*
+	 * Compiling the script yielded no bytecode.  The script must be all
+	 * whitespace, comments, and empty commands.  Such scripts are defined
+	 * to successfully produce the empty string result, so we emit the
+	 * simple bytecode that makes that happen.
+	 */
+
+	PushStringLiteral(envPtr, "");
+    } else {
+	/*
+	 * We compiled at least one command to bytecode.  The routine
+	 * CompileCommandTokens() follows the bytecode of each compiled
+	 * command with an INST_POP, so that stack balance is maintained when
+	 * several commands are in sequence.  (The result of each command is
+	 * thrown away before moving on to the next command).  For the last
+	 * command compiled, we need to undo that INST_POP so that the result
+	 * of the last command becomes the result of the script.  The code
+	 * here removes that trailing INST_POP.
+	 */
+
+	envPtr->cmdMapPtr[lastCmdIdx].numCodeBytes--;
+	envPtr->codeNext--;
+	envPtr->currStackDepth++;
+    }
+    TclCheckStackDepth(depth+1, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileTokens --
+ *
+ *	Given an array of tokens parsed from a Tcl command (e.g., the tokens
+ *	that make up a word) this procedure emits instructions to evaluate the
+ *	tokens and concatenate their values to form a single result value on
+ *	the interpreter's runtime evaluation stack.
+ *
+ * Results:
+ *	The return value is a standard Tcl result. If an error occurs, an
+ *	error message is left in the interpreter's result.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to push and evaluate the tokens at
+ *	runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCompileVarSubst(
+    Tcl_Interp *interp,
+    Tcl_Token *tokenPtr,
+    CompileEnv *envPtr)
+{
+    const char *p, *name = tokenPtr[1].start;
+    int nameBytes = tokenPtr[1].size;
+    int i, localVar, localVarName = 1;
+
+    /*
+     * Determine how the variable name should be handled: if it contains any
+     * namespace qualifiers it is not a local variable (localVarName=-1); if
+     * it looks like an array element and the token has a single component, it
+     * should not be created here [Bug 569438] (localVarName=0); otherwise,
+     * the local variable can safely be created (localVarName=1).
+     */
+
+    for (i = 0, p = name;  i < nameBytes;  i++, p++) {
+	if ((*p == ':') && (i < nameBytes-1) && (*(p+1) == ':')) {
+	    localVarName = -1;
+	    break;
+	} else if ((*p == '(')
+		&& (tokenPtr->numComponents == 1)
+		&& (*(name + nameBytes - 1) == ')')) {
+	    localVarName = 0;
+	    break;
+	}
+    }
+
+    /*
+     * Either push the variable's name, or find its index in the array
+     * of local variables in a procedure frame.
+     */
+
+    localVar = -1;
+    if (localVarName != -1) {
+	localVar = TclFindCompiledLocal(name, nameBytes, localVarName, envPtr);
+    }
+    if (localVar < 0) {
+	PushLiteral(envPtr, name, nameBytes);
+    }
+
+    /*
+     * Emit instructions to load the variable.
+     */
+
+    TclAdvanceLines(&envPtr->line, tokenPtr[1].start,
+	    tokenPtr[1].start + tokenPtr[1].size);
+
+    if (tokenPtr->numComponents == 1) {
+	if (localVar < 0) {
+	    TclEmitOpcode(INST_LOAD_STK, envPtr);
+	} else if (localVar <= 255) {
+	    TclEmitInstInt1(INST_LOAD_SCALAR1, localVar, envPtr);
+	} else {
+	    TclEmitInstInt4(INST_LOAD_SCALAR4, localVar, envPtr);
+	}
+    } else {
+	TclCompileTokens(interp, tokenPtr+2, tokenPtr->numComponents-1, envPtr);
+	if (localVar < 0) {
+	    TclEmitOpcode(INST_LOAD_ARRAY_STK, envPtr);
+	} else if (localVar <= 255) {
+	    TclEmitInstInt1(INST_LOAD_ARRAY1, localVar, envPtr);
+	} else {
+	    TclEmitInstInt4(INST_LOAD_ARRAY4, localVar, envPtr);
+	}
+    }
+}
+
+void
+TclCompileTokens(
+    Tcl_Interp *interp,		/* Used for error and status reporting. */
+    Tcl_Token *tokenPtr,	/* Pointer to first in an array of tokens to
+				 * compile. */
+    int count,			/* Number of tokens to consider at tokenPtr.
+				 * Must be at least 1. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_DString textBuffer;	/* Holds concatenated chars from adjacent
+				 * TCL_TOKEN_TEXT, TCL_TOKEN_BS tokens. */
+    char buffer[TCL_UTF_MAX];
+    int i, numObjsToConcat, length, adjust;
+    unsigned char *entryCodeNext = envPtr->codeNext;
+#define NUM_STATIC_POS 20
+    int isLiteral, maxNumCL, numCL;
+    int *clPosition = NULL;
+    int depth = TclGetStackDepth(envPtr);
+
+    /*
+     * For the handling of continuation lines in literals we first check if
+     * this is actually a literal. For if not we can forego the additional
+     * processing. Otherwise we pre-allocate a small table to store the
+     * locations of all continuation lines we find in this literal, if any.
+     * The table is extended if needed.
+     *
+     * Note: Different to the equivalent code in function 'TclSubstTokens()'
+     * (see file "tclParse.c") we do not seem to need the 'adjust' variable.
+     * We also do not seem to need code which merges continuation line
+     * information of multiple words which concat'd at runtime. Either that or
+     * I have not managed to find a test case for these two possibilities yet.
+     * It might be a difference between compile- versus run-time processing.
+     */
+
+    numCL = 0;
+    maxNumCL = 0;
+    isLiteral = 1;
+    for (i=0 ; i < count; i++) {
+	if ((tokenPtr[i].type != TCL_TOKEN_TEXT)
+		&& (tokenPtr[i].type != TCL_TOKEN_BS)) {
+	    isLiteral = 0;
+	    break;
+	}
+    }
+
+    if (isLiteral) {
+	maxNumCL = NUM_STATIC_POS;
+	clPosition = ckalloc(maxNumCL * sizeof(int));
+    }
+
+    adjust = 0;
+    Tcl_DStringInit(&textBuffer);
+    numObjsToConcat = 0;
+    for ( ;  count > 0;  count--, tokenPtr++) {
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_TEXT:
+	    TclDStringAppendToken(&textBuffer, tokenPtr);
+	    TclAdvanceLines(&envPtr->line, tokenPtr->start,
+		    tokenPtr->start + tokenPtr->size);
+	    break;
+
+	case TCL_TOKEN_BS:
+	    length = TclParseBackslash(tokenPtr->start, tokenPtr->size,
+		    NULL, buffer);
+	    Tcl_DStringAppend(&textBuffer, buffer, length);
+
+	    /*
+	     * If the backslash sequence we found is in a literal, and
+	     * represented a continuation line, we compute and store its
+	     * location (as char offset to the beginning of the _result_
+	     * script). We may have to extend the table of locations.
+	     *
+	     * Note that the continuation line information is relevant even if
+	     * the word we are processing is not a literal, as it can affect
+	     * nested commands. See the branch for TCL_TOKEN_COMMAND below,
+	     * where the adjustment we are tracking here is taken into
+	     * account. The good thing is that we do not need a table of
+	     * everything, just the number of lines we have to add as
+	     * correction.
+	     */
+
+	    if ((length == 1) && (buffer[0] == ' ') &&
+		(tokenPtr->start[1] == '\n')) {
+		if (isLiteral) {
+		    int clPos = Tcl_DStringLength(&textBuffer);
+
+		    if (numCL >= maxNumCL) {
+			maxNumCL *= 2;
+			clPosition = ckrealloc(clPosition,
+                                maxNumCL * sizeof(int));
+		    }
+		    clPosition[numCL] = clPos;
+		    numCL ++;
+		}
+		adjust++;
+	    }
+	    break;
+
+	case TCL_TOKEN_COMMAND:
+	    /*
+	     * Push any accumulated chars appearing before the command.
+	     */
+
+	    if (Tcl_DStringLength(&textBuffer) > 0) {
+		int literal = TclRegisterDStringLiteral(envPtr, &textBuffer);
+
+		TclEmitPush(literal, envPtr);
+		numObjsToConcat++;
+		Tcl_DStringFree(&textBuffer);
+
+		if (numCL) {
+		    TclContinuationsEnter(TclFetchLiteral(envPtr, literal),
+			    numCL, clPosition);
+		}
+		numCL = 0;
+	    }
+
+	    envPtr->line += adjust;
+	    TclCompileScript(interp, tokenPtr->start+1,
+		    tokenPtr->size-2, envPtr);
+	    envPtr->line -= adjust;
+	    numObjsToConcat++;
+	    break;
+
+	case TCL_TOKEN_VARIABLE:
+	    /*
+	     * Push any accumulated chars appearing before the $<var>.
+	     */
+
+	    if (Tcl_DStringLength(&textBuffer) > 0) {
+		int literal;
+
+		literal = TclRegisterDStringLiteral(envPtr, &textBuffer);
+		TclEmitPush(literal, envPtr);
+		numObjsToConcat++;
+		Tcl_DStringFree(&textBuffer);
+	    }
+
+	    TclCompileVarSubst(interp, tokenPtr, envPtr);
+	    numObjsToConcat++;
+	    count -= tokenPtr->numComponents;
+	    tokenPtr += tokenPtr->numComponents;
+	    break;
+
+	default:
+	    Tcl_Panic("Unexpected token type in TclCompileTokens: %d; %.*s",
+		    tokenPtr->type, tokenPtr->size, tokenPtr->start);
+	}
+    }
+
+    /*
+     * Push any accumulated characters appearing at the end.
+     */
+
+    if (Tcl_DStringLength(&textBuffer) > 0) {
+	int literal = TclRegisterDStringLiteral(envPtr, &textBuffer);
+
+	TclEmitPush(literal, envPtr);
+	numObjsToConcat++;
+	if (numCL) {
+	    TclContinuationsEnter(TclFetchLiteral(envPtr, literal),
+		    numCL, clPosition);
+	}
+	numCL = 0;
+    }
+
+    /*
+     * If necessary, concatenate the parts of the word.
+     */
+
+    while (numObjsToConcat > 255) {
+	TclEmitInstInt1(INST_STR_CONCAT1, 255, envPtr);
+	numObjsToConcat -= 254;	/* concat pushes 1 obj, the result */
+    }
+    if (numObjsToConcat > 1) {
+	TclEmitInstInt1(INST_STR_CONCAT1, numObjsToConcat, envPtr);
+    }
+
+    /*
+     * If the tokens yielded no instructions, push an empty string.
+     */
+
+    if (envPtr->codeNext == entryCodeNext) {
+	PushStringLiteral(envPtr, "");
+    }
+    Tcl_DStringFree(&textBuffer);
+
+    /*
+     * Release the temp table we used to collect the locations of continuation
+     * lines, if any.
+     */
+
+    if (maxNumCL) {
+	ckfree(clPosition);
+    }
+    TclCheckStackDepth(depth+1, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileCmdWord --
+ *
+ *	Given an array of parse tokens for a word containing one or more Tcl
+ *	commands, emit inline instructions to execute them. This procedure
+ *	differs from TclCompileTokens in that a simple word such as a loop
+ *	body enclosed in braces is not just pushed as a string, but is itself
+ *	parsed into tokens and compiled.
+ *
+ * Results:
+ *	The return value is a standard Tcl result. If an error occurs, an
+ *	error message is left in the interpreter's result.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the tokens at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCompileCmdWord(
+    Tcl_Interp *interp,		/* Used for error and status reporting. */
+    Tcl_Token *tokenPtr,	/* Pointer to first in an array of tokens for
+				 * a command word to compile inline. */
+    int count,			/* Number of tokens to consider at tokenPtr.
+				 * Must be at least 1. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    if ((count == 1) && (tokenPtr->type == TCL_TOKEN_TEXT)) {
+	/*
+	 * Handle the common case: if there is a single text token, compile it
+	 * into an inline sequence of instructions.
+	 */
+
+	TclCompileScript(interp, tokenPtr->start, tokenPtr->size, envPtr);
+    } else {
+	/*
+	 * Multiple tokens or the single token involves substitutions. Emit
+	 * instructions to invoke the eval command procedure at runtime on the
+	 * result of evaluating the tokens.
+	 */
+
+	TclCompileTokens(interp, tokenPtr, count, envPtr);
+	TclEmitInvoke(envPtr, INST_EVAL_STK);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileExprWords --
+ *
+ *	Given an array of parse tokens representing one or more words that
+ *	contain a Tcl expression, emit inline instructions to execute the
+ *	expression. This procedure differs from TclCompileExpr in that it
+ *	supports Tcl's two-level substitution semantics for expressions that
+ *	appear as command words.
+ *
+ * Results:
+ *	The return value is a standard Tcl result. If an error occurs, an
+ *	error message is left in the interpreter's result.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the expression.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCompileExprWords(
+    Tcl_Interp *interp,		/* Used for error and status reporting. */
+    Tcl_Token *tokenPtr,	/* Points to first in an array of word tokens
+				 * tokens for the expression to compile
+				 * inline. */
+    int numWords,		/* Number of word tokens starting at tokenPtr.
+				 * Must be at least 1. Each word token
+				 * contains one or more subtokens. */
+    CompileEnv *envPtr)		/* Holds the resulting instructions. */
+{
+    Tcl_Token *wordPtr;
+    int i, concatItems;
+
+    /*
+     * If the expression is a single word that doesn't require substitutions,
+     * just compile its string into inline instructions.
+     */
+
+    if ((numWords == 1) && (tokenPtr->type == TCL_TOKEN_SIMPLE_WORD)) {
+	TclCompileExpr(interp, tokenPtr[1].start,tokenPtr[1].size, envPtr, 1);
+	return;
+    }
+
+    /*
+     * Emit code to call the expr command proc at runtime. Concatenate the
+     * (already substituted once) expr tokens with a space between each.
+     */
+
+    wordPtr = tokenPtr;
+    for (i = 0;  i < numWords;  i++) {
+	CompileTokens(envPtr, wordPtr, interp);
+	if (i < (numWords - 1)) {
+	    PushStringLiteral(envPtr, " ");
+	}
+	wordPtr += wordPtr->numComponents + 1;
+    }
+    concatItems = 2*numWords - 1;
+    while (concatItems > 255) {
+	TclEmitInstInt1(INST_STR_CONCAT1, 255, envPtr);
+	concatItems -= 254;
+    }
+    if (concatItems > 1) {
+	TclEmitInstInt1(INST_STR_CONCAT1, concatItems, envPtr);
+    }
+    TclEmitOpcode(INST_EXPR_STK, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileNoOp --
+ *
+ *	Function called to compile no-op's
+ *
+ * Results:
+ *	The return value is TCL_OK, indicating successful compilation.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute a no-op at runtime. No
+ *	result is pushed onto the stack: the compiler has to take care of this
+ *	itself if the last compiled command is a NoOp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileNoOp(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr;
+    int i;
+
+    tokenPtr = parsePtr->tokenPtr;
+    for (i = 1; i < parsePtr->numWords; i++) {
+	tokenPtr = tokenPtr + tokenPtr->numComponents + 1;
+
+	if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    CompileTokens(envPtr, tokenPtr, interp);
+	    TclEmitOpcode(INST_POP, envPtr);
+	}
+    }
+    PushStringLiteral(envPtr, "");
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitByteCodeObj --
+ *
+ *	Create a ByteCode structure and initialize it from a CompileEnv
+ *	compilation environment structure. The ByteCode structure is smaller
+ *	and contains just that information needed to execute the bytecode
+ *	instructions resulting from compiling a Tcl script. The resulting
+ *	structure is placed in the specified object.
+ *
+ * Results:
+ *	A newly constructed ByteCode object is stored in the internal
+ *	representation of the objPtr.
+ *
+ * Side effects:
+ *	A single heap object is allocated to hold the new ByteCode structure
+ *	and its code, object, command location, and aux data arrays. Note that
+ *	"ownership" (i.e., the pointers to) the Tcl objects and aux data items
+ *	will be handed over to the new ByteCode structure from the CompileEnv
+ *	structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PreventCycle(
+    Tcl_Obj *objPtr,
+    CompileEnv *envPtr)
+{
+    int i;
+
+    for (i = 0;  i < envPtr->literalArrayNext; i++) {
+	if (objPtr == TclFetchLiteral(envPtr, i)) {
+	    /*
+	     * Prevent circular reference where the bytecode intrep of
+	     * a value contains a literal which is that same value.
+	     * If this is allowed to happen, refcount decrements may not
+	     * reach zero, and memory may leak.  Bugs 467523, 3357771
+	     *
+	     * NOTE:  [Bugs 3392070, 3389764] We make a copy based completely
+	     * on the string value, and do not call Tcl_DuplicateObj() so we
+             * can be sure we do not have any lingering cycles hiding in
+	     * the intrep.
+	     */
+	    int numBytes;
+	    const char *bytes = TclGetStringFromObj(objPtr, &numBytes);
+	    Tcl_Obj *copyPtr = Tcl_NewStringObj(bytes, numBytes);
+
+	    Tcl_IncrRefCount(copyPtr);
+	    TclReleaseLiteral((Tcl_Interp *)envPtr->iPtr, objPtr);
+
+	    envPtr->literalArrayPtr[i].objPtr = copyPtr;
+	}
+    }
+}
+
+ByteCode *
+TclInitByteCode(
+    register CompileEnv *envPtr)/* Points to the CompileEnv structure from
+				 * which to create a ByteCode structure. */
+{
+    register ByteCode *codePtr;
+    size_t codeBytes, objArrayBytes, exceptArrayBytes, cmdLocBytes;
+    size_t auxDataArrayBytes, structureSize;
+    register unsigned char *p;
+#ifdef TCL_COMPILE_DEBUG
+    unsigned char *nextPtr;
+#endif
+    int numLitObjects = envPtr->literalArrayNext;
+    Namespace *namespacePtr;
+    int i, isNew;
+    Interp *iPtr;
+
+    if (envPtr->iPtr == NULL) {
+	Tcl_Panic("TclInitByteCodeObj() called on uninitialized CompileEnv");
+    }
+
+    iPtr = envPtr->iPtr;
+
+    codeBytes = envPtr->codeNext - envPtr->codeStart;
+    objArrayBytes = envPtr->literalArrayNext * sizeof(Tcl_Obj *);
+    exceptArrayBytes = envPtr->exceptArrayNext * sizeof(ExceptionRange);
+    auxDataArrayBytes = envPtr->auxDataArrayNext * sizeof(AuxData);
+    cmdLocBytes = GetCmdLocEncodingSize(envPtr);
+
+    /*
+     * Compute the total number of bytes needed for this bytecode.
+     */
+
+    structureSize = sizeof(ByteCode);
+    structureSize += TCL_ALIGN(codeBytes);	  /* align object array */
+    structureSize += TCL_ALIGN(objArrayBytes);	  /* align exc range arr */
+    structureSize += TCL_ALIGN(exceptArrayBytes); /* align AuxData array */
+    structureSize += auxDataArrayBytes;
+    structureSize += cmdLocBytes;
+
+    if (envPtr->iPtr->varFramePtr != NULL) {
+	namespacePtr = envPtr->iPtr->varFramePtr->nsPtr;
+    } else {
+	namespacePtr = envPtr->iPtr->globalNsPtr;
+    }
+
+    p = ckalloc(structureSize);
+    codePtr = (ByteCode *) p;
+    codePtr->interpHandle = TclHandlePreserve(iPtr->handle);
+    codePtr->compileEpoch = iPtr->compileEpoch;
+    codePtr->nsPtr = namespacePtr;
+    codePtr->nsEpoch = namespacePtr->resolverEpoch;
+    codePtr->refCount = 0;
+    TclPreserveByteCode(codePtr);
+    if (namespacePtr->compiledVarResProc || iPtr->resolverPtr) {
+	codePtr->flags = TCL_BYTECODE_RESOLVE_VARS;
+    } else {
+	codePtr->flags = 0;
+    }
+    codePtr->source = envPtr->source;
+    codePtr->procPtr = envPtr->procPtr;
+
+    codePtr->numCommands = envPtr->numCommands;
+    codePtr->numSrcBytes = envPtr->numSrcBytes;
+    codePtr->numCodeBytes = codeBytes;
+    codePtr->numLitObjects = numLitObjects;
+    codePtr->numExceptRanges = envPtr->exceptArrayNext;
+    codePtr->numAuxDataItems = envPtr->auxDataArrayNext;
+    codePtr->numCmdLocBytes = cmdLocBytes;
+    codePtr->maxExceptDepth = envPtr->maxExceptDepth;
+    codePtr->maxStackDepth = envPtr->maxStackDepth;
+
+    p += sizeof(ByteCode);
+    codePtr->codeStart = p;
+    memcpy(p, envPtr->codeStart, (size_t) codeBytes);
+
+    p += TCL_ALIGN(codeBytes);		/* align object array */
+    codePtr->objArrayPtr = (Tcl_Obj **) p;
+    for (i = 0;  i < numLitObjects;  i++) {
+	codePtr->objArrayPtr[i] = TclFetchLiteral(envPtr, i);
+    }
+
+    p += TCL_ALIGN(objArrayBytes);	/* align exception range array */
+    if (exceptArrayBytes > 0) {
+	codePtr->exceptArrayPtr = (ExceptionRange *) p;
+	memcpy(p, envPtr->exceptArrayPtr, (size_t) exceptArrayBytes);
+    } else {
+	codePtr->exceptArrayPtr = NULL;
+    }
+
+    p += TCL_ALIGN(exceptArrayBytes);	/* align AuxData array */
+    if (auxDataArrayBytes > 0) {
+	codePtr->auxDataArrayPtr = (AuxData *) p;
+	memcpy(p, envPtr->auxDataArrayPtr, (size_t) auxDataArrayBytes);
+    } else {
+	codePtr->auxDataArrayPtr = NULL;
+    }
+
+    p += auxDataArrayBytes;
+#ifndef TCL_COMPILE_DEBUG
+    EncodeCmdLocMap(envPtr, codePtr, (unsigned char *) p);
+#else
+    nextPtr = EncodeCmdLocMap(envPtr, codePtr, (unsigned char *) p);
+    if (((size_t)(nextPtr - p)) != cmdLocBytes) {
+	Tcl_Panic("TclInitByteCodeObj: encoded cmd location bytes %lu != expected size %lu", (unsigned long)(nextPtr - p), (unsigned long)cmdLocBytes);
+    }
+#endif
+
+    /*
+     * Record various compilation-related statistics about the new ByteCode
+     * structure. Don't include overhead for statistics-related fields.
+     */
+
+#ifdef TCL_COMPILE_STATS
+    codePtr->structureSize = structureSize
+	    - (sizeof(size_t) + sizeof(Tcl_Time));
+    Tcl_GetTime(&codePtr->createTime);
+
+    RecordByteCodeStats(codePtr);
+#endif /* TCL_COMPILE_STATS */
+
+    /*
+     * TIP #280. Associate the extended per-word line information with the
+     * byte code object (internal rep), for use with the bc compiler.
+     */
+
+    Tcl_SetHashValue(Tcl_CreateHashEntry(iPtr->lineBCPtr, codePtr,
+	    &isNew), envPtr->extCmdMapPtr);
+    envPtr->extCmdMapPtr = NULL;
+
+    /* We've used up the CompileEnv.  Mark as uninitialized. */
+    envPtr->iPtr = NULL;
+
+    codePtr->localCachePtr = NULL;
+    return codePtr;
+}
+
+ByteCode *
+TclInitByteCodeObj(
+    Tcl_Obj *objPtr,		/* Points object that should be initialized,
+				 * and whose string rep contains the source
+				 * code. */
+    const Tcl_ObjType *typePtr,
+    register CompileEnv *envPtr)/* Points to the CompileEnv structure from
+				 * which to create a ByteCode structure. */
+{
+    ByteCode *codePtr;
+
+    PreventCycle(objPtr, envPtr);
+
+    codePtr = TclInitByteCode(envPtr);
+
+    /*
+     * Free the old internal rep then convert the object to a bytecode object
+     * by making its internal rep point to the just compiled ByteCode.
+     */
+
+    TclFreeIntRep(objPtr);
+    objPtr->internalRep.twoPtrValue.ptr1 = codePtr;
+    objPtr->typePtr = typePtr;
+    return codePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFindCompiledLocal --
+ *
+ *	This procedure is called at compile time to look up and optionally
+ *	allocate an entry ("slot") for a variable in a procedure's array of
+ *	local variables. If the variable's name is NULL, a new temporary
+ *	variable is always created. (Such temporary variables can only be
+ *	referenced using their slot index.)
+ *
+ * Results:
+ *	If create is 0 and the name is non-NULL, then if the variable is
+ *	found, the index of its entry in the procedure's array of local
+ *	variables is returned; otherwise -1 is returned. If name is NULL, the
+ *	index of a new temporary variable is returned. Finally, if create is 1
+ *	and name is non-NULL, the index of a new entry is returned.
+ *
+ * Side effects:
+ *	Creates and registers a new local variable if create is 1 and the
+ *	variable is unknown, or if the name is NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFindCompiledLocal(
+    register const char *name,	/* Points to first character of the name of a
+				 * scalar or array variable. If NULL, a
+				 * temporary var should be created. */
+    int nameBytes,		/* Number of bytes in the name. */
+    int create,			/* If 1, allocate a local frame entry for the
+				 * variable if it is new. */
+    CompileEnv *envPtr)		/* Points to the current compile environment*/
+{
+    register CompiledLocal *localPtr;
+    int localVar = -1;
+    register int i;
+    Proc *procPtr;
+
+    /*
+     * If not creating a temporary, does a local variable of the specified
+     * name already exist?
+     */
+
+    procPtr = envPtr->procPtr;
+
+    if (procPtr == NULL) {
+	/*
+	 * Compiling a non-body script: give it read access to the LVT in the
+	 * current localCache
+	 */
+
+	LocalCache *cachePtr = envPtr->iPtr->varFramePtr->localCachePtr;
+	const char *localName;
+	Tcl_Obj **varNamePtr;
+	int len;
+
+	if (!cachePtr || !name) {
+	    return -1;
+	}
+
+	varNamePtr = &cachePtr->varName0;
+	for (i=0; i < cachePtr->numVars; varNamePtr++, i++) {
+	    if (*varNamePtr) {
+		localName = TclGetString(*varNamePtr);
+		len = (*varNamePtr)->length;
+		if ((len == nameBytes) && !strncmp(name, localName, len)) {
+		    return i;
+		}
+	    }
+	}
+	return -1;
+    }
+
+    if (name != NULL) {
+	int localCt = procPtr->numCompiledLocals;
+
+	localPtr = procPtr->firstLocalPtr;
+	for (i = 0;  i < localCt;  i++) {
+	    if (!TclIsVarTemporary(localPtr)) {
+		char *localName = localPtr->name;
+
+		if ((nameBytes == localPtr->nameLength) &&
+			(strncmp(name,localName,(unsigned)nameBytes) == 0)) {
+		    return i;
+		}
+	    }
+	    localPtr = localPtr->nextPtr;
+	}
+    }
+
+    /*
+     * Create a new variable if appropriate.
+     */
+
+    if (create || (name == NULL)) {
+	localVar = procPtr->numCompiledLocals;
+	localPtr = ckalloc(TclOffset(CompiledLocal, name) + nameBytes + 1);
+	if (procPtr->firstLocalPtr == NULL) {
+	    procPtr->firstLocalPtr = procPtr->lastLocalPtr = localPtr;
+	} else {
+	    procPtr->lastLocalPtr->nextPtr = localPtr;
+	    procPtr->lastLocalPtr = localPtr;
+	}
+	localPtr->nextPtr = NULL;
+	localPtr->nameLength = nameBytes;
+	localPtr->frameIndex = localVar;
+	localPtr->flags = 0;
+	if (name == NULL) {
+	    localPtr->flags |= VAR_TEMPORARY;
+	}
+	localPtr->defValuePtr = NULL;
+	localPtr->resolveInfo = NULL;
+
+	if (name != NULL) {
+	    memcpy(localPtr->name, name, (size_t) nameBytes);
+	}
+	localPtr->name[nameBytes] = '\0';
+	procPtr->numCompiledLocals++;
+    }
+    return localVar;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclExpandCodeArray --
+ *
+ *	Procedure that uses malloc to allocate more storage for a CompileEnv's
+ *	code array.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The byte code array in *envPtr is reallocated to a new array of double
+ *	the size, and if envPtr->mallocedCodeArray is non-zero the old array
+ *	is freed. Byte codes are copied from the old array to the new one.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclExpandCodeArray(
+    void *envArgPtr)		/* Points to the CompileEnv whose code array
+				 * must be enlarged. */
+{
+    CompileEnv *envPtr = envArgPtr;
+				/* The CompileEnv containing the code array to
+				 * be doubled in size. */
+
+    /*
+     * envPtr->codeNext is equal to envPtr->codeEnd. The currently defined
+     * code bytes are stored between envPtr->codeStart and envPtr->codeNext-1
+     * [inclusive].
+     */
+
+    size_t currBytes = envPtr->codeNext - envPtr->codeStart;
+    size_t newBytes = 2 * (envPtr->codeEnd - envPtr->codeStart);
+
+    if (envPtr->mallocedCodeArray) {
+	envPtr->codeStart = ckrealloc(envPtr->codeStart, newBytes);
+    } else {
+	/*
+	 * envPtr->codeStart isn't a ckalloc'd pointer, so we must code a
+	 * ckrealloc equivalent for ourselves.
+	 */
+
+	unsigned char *newPtr = ckalloc(newBytes);
+
+	memcpy(newPtr, envPtr->codeStart, currBytes);
+	envPtr->codeStart = newPtr;
+	envPtr->mallocedCodeArray = 1;
+    }
+
+    envPtr->codeNext = envPtr->codeStart + currBytes;
+    envPtr->codeEnd = envPtr->codeStart + newBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EnterCmdStartData --
+ *
+ *	Registers the starting source and bytecode location of a command. This
+ *	information is used at runtime to map between instruction pc and
+ *	source locations.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Inserts source and code location information into the compilation
+ *	environment envPtr for the command at index cmdIndex. The compilation
+ *	environment's CmdLocation array is grown if necessary.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+EnterCmdStartData(
+    CompileEnv *envPtr,		/* Points to the compilation environment
+				 * structure in which to enter command
+				 * location information. */
+    int cmdIndex,		/* Index of the command whose start data is
+				 * being set. */
+    int srcOffset,		/* Offset of first char of the command. */
+    int codeOffset)		/* Offset of first byte of command code. */
+{
+    CmdLocation *cmdLocPtr;
+
+    if ((cmdIndex < 0) || (cmdIndex >= envPtr->numCommands)) {
+	Tcl_Panic("EnterCmdStartData: bad command index %d", cmdIndex);
+    }
+
+    if (cmdIndex >= envPtr->cmdMapEnd) {
+	/*
+	 * Expand the command location array by allocating more storage from
+	 * the heap. The currently allocated CmdLocation entries are stored
+	 * from cmdMapPtr[0] up to cmdMapPtr[envPtr->cmdMapEnd] (inclusive).
+	 */
+
+	size_t currElems = envPtr->cmdMapEnd;
+	size_t newElems = 2 * currElems;
+	size_t currBytes = currElems * sizeof(CmdLocation);
+	size_t newBytes = newElems * sizeof(CmdLocation);
+
+	if (envPtr->mallocedCmdMap) {
+	    envPtr->cmdMapPtr = ckrealloc(envPtr->cmdMapPtr, newBytes);
+	} else {
+	    /*
+	     * envPtr->cmdMapPtr isn't a ckalloc'd pointer, so we must code a
+	     * ckrealloc equivalent for ourselves.
+	     */
+
+	    CmdLocation *newPtr = ckalloc(newBytes);
+
+	    memcpy(newPtr, envPtr->cmdMapPtr, currBytes);
+	    envPtr->cmdMapPtr = newPtr;
+	    envPtr->mallocedCmdMap = 1;
+	}
+	envPtr->cmdMapEnd = newElems;
+    }
+
+    if (cmdIndex > 0) {
+	if (codeOffset < envPtr->cmdMapPtr[cmdIndex-1].codeOffset) {
+	    Tcl_Panic("EnterCmdStartData: cmd map not sorted by code offset");
+	}
+    }
+
+    cmdLocPtr = &envPtr->cmdMapPtr[cmdIndex];
+    cmdLocPtr->codeOffset = codeOffset;
+    cmdLocPtr->srcOffset = srcOffset;
+    cmdLocPtr->numSrcBytes = -1;
+    cmdLocPtr->numCodeBytes = -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EnterCmdExtentData --
+ *
+ *	Registers the source and bytecode length for a command. This
+ *	information is used at runtime to map between instruction pc and
+ *	source locations.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Inserts source and code length information into the compilation
+ *	environment envPtr for the command at index cmdIndex. Starting source
+ *	and bytecode information for the command must already have been
+ *	registered.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+EnterCmdExtentData(
+    CompileEnv *envPtr,		/* Points to the compilation environment
+				 * structure in which to enter command
+				 * location information. */
+    int cmdIndex,		/* Index of the command whose source and code
+				 * length data is being set. */
+    int numSrcBytes,		/* Number of command source chars. */
+    int numCodeBytes)		/* Offset of last byte of command code. */
+{
+    CmdLocation *cmdLocPtr;
+
+    if ((cmdIndex < 0) || (cmdIndex >= envPtr->numCommands)) {
+	Tcl_Panic("EnterCmdExtentData: bad command index %d", cmdIndex);
+    }
+
+    if (cmdIndex > envPtr->cmdMapEnd) {
+	Tcl_Panic("EnterCmdExtentData: missing start data for command %d",
+		cmdIndex);
+    }
+
+    cmdLocPtr = &envPtr->cmdMapPtr[cmdIndex];
+    cmdLocPtr->numSrcBytes = numSrcBytes;
+    cmdLocPtr->numCodeBytes = numCodeBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * TIP #280
+ *
+ * EnterCmdWordData --
+ *
+ *	Registers the lines for the words of a command. This information is
+ *	used at runtime by 'info frame'.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Inserts word location information into the compilation environment
+ *	envPtr for the command at index cmdIndex. The compilation
+ *	environment's ExtCmdLoc.ECL array is grown if necessary.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+EnterCmdWordData(
+    ExtCmdLoc *eclPtr,		/* Points to the map environment structure in
+				 * which to enter command location
+				 * information. */
+    int srcOffset,		/* Offset of first char of the command. */
+    Tcl_Token *tokenPtr,
+    const char *cmd,
+    int len,
+    int numWords,
+    int line,
+    int *clNext,
+    int **wlines,
+    CompileEnv *envPtr)
+{
+    ECL *ePtr;
+    const char *last;
+    int wordIdx, wordLine, *wwlines, *wordNext;
+
+    if (eclPtr->nuloc >= eclPtr->nloc) {
+	/*
+	 * Expand the ECL array by allocating more storage from the heap. The
+	 * currently allocated ECL entries are stored from eclPtr->loc[0] up
+	 * to eclPtr->loc[eclPtr->nuloc-1] (inclusive).
+	 */
+
+	size_t currElems = eclPtr->nloc;
+	size_t newElems = (currElems ? 2*currElems : 1);
+	size_t newBytes = newElems * sizeof(ECL);
+
+	eclPtr->loc = ckrealloc(eclPtr->loc, newBytes);
+	eclPtr->nloc = newElems;
+    }
+
+    ePtr = &eclPtr->loc[eclPtr->nuloc];
+    ePtr->srcOffset = srcOffset;
+    ePtr->line = ckalloc(numWords * sizeof(int));
+    ePtr->next = ckalloc(numWords * sizeof(int *));
+    ePtr->nline = numWords;
+    wwlines = ckalloc(numWords * sizeof(int));
+
+    last = cmd;
+    wordLine = line;
+    wordNext = clNext;
+    for (wordIdx=0 ; wordIdx<numWords;
+	    wordIdx++, tokenPtr += tokenPtr->numComponents + 1) {
+	TclAdvanceLines(&wordLine, last, tokenPtr->start);
+	TclAdvanceContinuations(&wordLine, &wordNext,
+		tokenPtr->start - envPtr->source);
+	/* See Ticket 4b61afd660 */
+	wwlines[wordIdx] =
+		((wordIdx == 0) || TclWordKnownAtCompileTime(tokenPtr, NULL))
+		? wordLine : -1;
+	ePtr->line[wordIdx] = wordLine;
+	ePtr->next[wordIdx] = wordNext;
+	last = tokenPtr->start;
+    }
+
+    *wlines = wwlines;
+    eclPtr->nuloc ++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateExceptRange --
+ *
+ *	Procedure that allocates and initializes a new ExceptionRange
+ *	structure of the specified kind in a CompileEnv.
+ *
+ * Results:
+ *	Returns the index for the newly created ExceptionRange.
+ *
+ * Side effects:
+ *	If there is not enough room in the CompileEnv's ExceptionRange array,
+ *	the array in expanded: a new array of double the size is allocated, if
+ *	envPtr->mallocedExceptArray is non-zero the old array is freed, and
+ *	ExceptionRange entries are copied from the old array to the new one.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCreateExceptRange(
+    ExceptionRangeType type,	/* The kind of ExceptionRange desired. */
+    register CompileEnv *envPtr)/* Points to CompileEnv for which to create a
+				 * new ExceptionRange structure. */
+{
+    register ExceptionRange *rangePtr;
+    register ExceptionAux *auxPtr;
+    int index = envPtr->exceptArrayNext;
+
+    if (index >= envPtr->exceptArrayEnd) {
+	/*
+	 * Expand the ExceptionRange array. The currently allocated entries
+	 * are stored between elements 0 and (envPtr->exceptArrayNext - 1)
+	 * [inclusive].
+	 */
+
+	size_t currBytes =
+		envPtr->exceptArrayNext * sizeof(ExceptionRange);
+	size_t currBytes2 = envPtr->exceptArrayNext * sizeof(ExceptionAux);
+	int newElems = 2*envPtr->exceptArrayEnd;
+	size_t newBytes = newElems * sizeof(ExceptionRange);
+	size_t newBytes2 = newElems * sizeof(ExceptionAux);
+
+	if (envPtr->mallocedExceptArray) {
+	    envPtr->exceptArrayPtr =
+		    ckrealloc(envPtr->exceptArrayPtr, newBytes);
+	    envPtr->exceptAuxArrayPtr =
+		    ckrealloc(envPtr->exceptAuxArrayPtr, newBytes2);
+	} else {
+	    /*
+	     * envPtr->exceptArrayPtr isn't a ckalloc'd pointer, so we must
+	     * code a ckrealloc equivalent for ourselves.
+	     */
+
+	    ExceptionRange *newPtr = ckalloc(newBytes);
+	    ExceptionAux *newPtr2 = ckalloc(newBytes2);
+
+	    memcpy(newPtr, envPtr->exceptArrayPtr, currBytes);
+	    memcpy(newPtr2, envPtr->exceptAuxArrayPtr, currBytes2);
+	    envPtr->exceptArrayPtr = newPtr;
+	    envPtr->exceptAuxArrayPtr = newPtr2;
+	    envPtr->mallocedExceptArray = 1;
+	}
+	envPtr->exceptArrayEnd = newElems;
+    }
+    envPtr->exceptArrayNext++;
+
+    rangePtr = &envPtr->exceptArrayPtr[index];
+    rangePtr->type = type;
+    rangePtr->nestingLevel = envPtr->exceptDepth;
+    rangePtr->codeOffset = -1;
+    rangePtr->numCodeBytes = -1;
+    rangePtr->breakOffset = -1;
+    rangePtr->continueOffset = -1;
+    rangePtr->catchOffset = -1;
+    auxPtr = &envPtr->exceptAuxArrayPtr[index];
+    auxPtr->supportsContinue = 1;
+    auxPtr->stackDepth = envPtr->currStackDepth;
+    auxPtr->expandTarget = envPtr->expandCount;
+    auxPtr->expandTargetDepth = -1;
+    auxPtr->numBreakTargets = 0;
+    auxPtr->breakTargets = NULL;
+    auxPtr->allocBreakTargets = 0;
+    auxPtr->numContinueTargets = 0;
+    auxPtr->continueTargets = NULL;
+    auxPtr->allocContinueTargets = 0;
+    return index;
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * TclGetInnermostExceptionRange --
+ *
+ *	Returns the innermost exception range that covers the current code
+ *	creation point, and (optionally) the stack depth that is expected at
+ *	that point. Relies on the fact that the range has a numCodeBytes = -1
+ *	when it is being populated and that inner ranges come after outer
+ *	ranges.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+ExceptionRange *
+TclGetInnermostExceptionRange(
+    CompileEnv *envPtr,
+    int returnCode,
+    ExceptionAux **auxPtrPtr)
+{
+    int i = envPtr->exceptArrayNext;
+    ExceptionRange *rangePtr = envPtr->exceptArrayPtr + i;
+
+    while (i > 0) {
+	rangePtr--; i--;
+
+	if (CurrentOffset(envPtr) >= rangePtr->codeOffset &&
+		(rangePtr->numCodeBytes == -1 || CurrentOffset(envPtr) <
+			rangePtr->codeOffset+rangePtr->numCodeBytes) &&
+		(returnCode != TCL_CONTINUE ||
+			envPtr->exceptAuxArrayPtr[i].supportsContinue)) {
+
+	    if (auxPtrPtr) {
+		*auxPtrPtr = envPtr->exceptAuxArrayPtr + i;
+	    }
+	    return rangePtr;
+	}
+    }
+    return NULL;
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * TclAddLoopBreakFixup, TclAddLoopContinueFixup --
+ *
+ *	Adds a place that wants to break/continue to the loop exception range
+ *	tracking that will be fixed up once the loop can be finalized. These
+ *	functions will generate an INST_JUMP4 that will be fixed up during the
+ *	loop finalization.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+void
+TclAddLoopBreakFixup(
+    CompileEnv *envPtr,
+    ExceptionAux *auxPtr)
+{
+    int range = auxPtr - envPtr->exceptAuxArrayPtr;
+
+    if (envPtr->exceptArrayPtr[range].type != LOOP_EXCEPTION_RANGE) {
+	Tcl_Panic("trying to add 'break' fixup to full exception range");
+    }
+
+    if (++auxPtr->numBreakTargets > auxPtr->allocBreakTargets) {
+	auxPtr->allocBreakTargets *= 2;
+	auxPtr->allocBreakTargets += 2;
+	if (auxPtr->breakTargets) {
+	    auxPtr->breakTargets = ckrealloc(auxPtr->breakTargets,
+		    sizeof(int) * auxPtr->allocBreakTargets);
+	} else {
+	    auxPtr->breakTargets =
+		    ckalloc(sizeof(int) * auxPtr->allocBreakTargets);
+	}
+    }
+    auxPtr->breakTargets[auxPtr->numBreakTargets - 1] = CurrentOffset(envPtr);
+    TclEmitInstInt4(INST_JUMP4, 0, envPtr);
+}
+
+void
+TclAddLoopContinueFixup(
+    CompileEnv *envPtr,
+    ExceptionAux *auxPtr)
+{
+    int range = auxPtr - envPtr->exceptAuxArrayPtr;
+
+    if (envPtr->exceptArrayPtr[range].type != LOOP_EXCEPTION_RANGE) {
+	Tcl_Panic("trying to add 'continue' fixup to full exception range");
+    }
+
+    if (++auxPtr->numContinueTargets > auxPtr->allocContinueTargets) {
+	auxPtr->allocContinueTargets *= 2;
+	auxPtr->allocContinueTargets += 2;
+	if (auxPtr->continueTargets) {
+	    auxPtr->continueTargets = ckrealloc(auxPtr->continueTargets,
+		    sizeof(int) * auxPtr->allocContinueTargets);
+	} else {
+	    auxPtr->continueTargets =
+		    ckalloc(sizeof(int) * auxPtr->allocContinueTargets);
+	}
+    }
+    auxPtr->continueTargets[auxPtr->numContinueTargets - 1] =
+	    CurrentOffset(envPtr);
+    TclEmitInstInt4(INST_JUMP4, 0, envPtr);
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * TclCleanupStackForBreakContinue --
+ *
+ *	Ditch the extra elements from the auxiliary stack and the main stack.
+ *	How to do this exactly depends on whether there are any elements on
+ *	the auxiliary stack to pop.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+void
+TclCleanupStackForBreakContinue(
+    CompileEnv *envPtr,
+    ExceptionAux *auxPtr)
+{
+    int savedStackDepth = envPtr->currStackDepth;
+    int toPop = envPtr->expandCount - auxPtr->expandTarget;
+
+    if (toPop > 0) {
+	while (toPop --> 0) {
+	    TclEmitOpcode(INST_EXPAND_DROP, envPtr);
+	}
+	TclAdjustStackDepth(auxPtr->expandTargetDepth - envPtr->currStackDepth,
+		envPtr);
+	envPtr->currStackDepth = auxPtr->expandTargetDepth;
+    }
+    toPop = envPtr->currStackDepth - auxPtr->stackDepth;
+    while (toPop --> 0) {
+	TclEmitOpcode(INST_POP, envPtr);
+    }
+    envPtr->currStackDepth = savedStackDepth;
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * StartExpanding --
+ *
+ *	Pushes an INST_EXPAND_START and does some additional housekeeping so
+ *	that the [break] and [continue] compilers can use an exception-free
+ *	issue to discard it.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+static void
+StartExpanding(
+    CompileEnv *envPtr)
+{
+    int i;
+
+    TclEmitOpcode(INST_EXPAND_START, envPtr);
+
+    /*
+     * Update inner exception ranges with information about the environment
+     * where this expansion started.
+     */
+
+    for (i=0 ; i<envPtr->exceptArrayNext ; i++) {
+	ExceptionRange *rangePtr = &envPtr->exceptArrayPtr[i];
+	ExceptionAux *auxPtr = &envPtr->exceptAuxArrayPtr[i];
+
+	/*
+	 * Ignore loops unless they're still being built.
+	 */
+
+	if (rangePtr->codeOffset > CurrentOffset(envPtr)) {
+	    continue;
+	}
+	if (rangePtr->numCodeBytes != -1) {
+	    continue;
+	}
+
+	/*
+	 * Adequate condition: further out loops and further in exceptions
+	 * don't actually need this information.
+	 */
+
+	if (auxPtr->expandTarget == envPtr->expandCount) {
+	    auxPtr->expandTargetDepth = envPtr->currStackDepth;
+	}
+    }
+
+    /*
+     * There's now one more expansion being processed on the auxiliary stack.
+     */
+
+    envPtr->expandCount++;
+}
+
+/*
+ * ---------------------------------------------------------------------
+ *
+ * TclFinalizeLoopExceptionRange --
+ *
+ *	Finalizes a loop exception range, binding the registered [break] and
+ *	[continue] implementations so that they jump to the correct place.
+ *	Note that this must only be called after *all* the exception range
+ *	target offsets have been set.
+ *
+ * ---------------------------------------------------------------------
+ */
+
+void
+TclFinalizeLoopExceptionRange(
+    CompileEnv *envPtr,
+    int range)
+{
+    ExceptionRange *rangePtr = &envPtr->exceptArrayPtr[range];
+    ExceptionAux *auxPtr = &envPtr->exceptAuxArrayPtr[range];
+    int i, offset;
+    unsigned char *site;
+
+    if (rangePtr->type != LOOP_EXCEPTION_RANGE) {
+	Tcl_Panic("trying to finalize a loop exception range");
+    }
+
+    /*
+     * Do the jump fixups. Note that these are always issued as INST_JUMP4 so
+     * there is no need to fuss around with updating code offsets.
+     */
+
+    for (i=0 ; i<auxPtr->numBreakTargets ; i++) {
+	site = envPtr->codeStart + auxPtr->breakTargets[i];
+	offset = rangePtr->breakOffset - auxPtr->breakTargets[i];
+	TclUpdateInstInt4AtPc(INST_JUMP4, offset, site);
+    }
+    for (i=0 ; i<auxPtr->numContinueTargets ; i++) {
+	site = envPtr->codeStart + auxPtr->continueTargets[i];
+	if (rangePtr->continueOffset == -1) {
+	    int j;
+
+	    /*
+	     * WTF? Can't bind, so revert to an INST_CONTINUE. Not enough
+	     * space to do anything else.
+	     */
+
+	    *site = INST_CONTINUE;
+	    for (j=0 ; j<4 ; j++) {
+		*++site = INST_NOP;
+	    }
+	} else {
+	    offset = rangePtr->continueOffset - auxPtr->continueTargets[i];
+	    TclUpdateInstInt4AtPc(INST_JUMP4, offset, site);
+	}
+    }
+
+    /*
+     * Drop the arrays we were holding the only reference to.
+     */
+
+    if (auxPtr->breakTargets) {
+	ckfree(auxPtr->breakTargets);
+	auxPtr->breakTargets = NULL;
+	auxPtr->numBreakTargets = 0;
+    }
+    if (auxPtr->continueTargets) {
+	ckfree(auxPtr->continueTargets);
+	auxPtr->continueTargets = NULL;
+	auxPtr->numContinueTargets = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateAuxData --
+ *
+ *	Procedure that allocates and initializes a new AuxData structure in a
+ *	CompileEnv's array of compilation auxiliary data records. These
+ *	AuxData records hold information created during compilation by
+ *	CompileProcs and used by instructions during execution.
+ *
+ * Results:
+ *	Returns the index for the newly created AuxData structure.
+ *
+ * Side effects:
+ *	If there is not enough room in the CompileEnv's AuxData array, the
+ *	AuxData array in expanded: a new array of double the size is
+ *	allocated, if envPtr->mallocedAuxDataArray is non-zero the old array
+ *	is freed, and AuxData entries are copied from the old array to the new
+ *	one.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCreateAuxData(
+    ClientData clientData,	/* The compilation auxiliary data to store in
+				 * the new aux data record. */
+    const AuxDataType *typePtr,	/* Pointer to the type to attach to this
+				 * AuxData */
+    register CompileEnv *envPtr)/* Points to the CompileEnv for which a new
+				 * aux data structure is to be allocated. */
+{
+    int index;			/* Index for the new AuxData structure. */
+    register AuxData *auxDataPtr;
+				/* Points to the new AuxData structure */
+
+    index = envPtr->auxDataArrayNext;
+    if (index >= envPtr->auxDataArrayEnd) {
+	/*
+	 * Expand the AuxData array. The currently allocated entries are
+	 * stored between elements 0 and (envPtr->auxDataArrayNext - 1)
+	 * [inclusive].
+	 */
+
+	size_t currBytes = envPtr->auxDataArrayNext * sizeof(AuxData);
+	int newElems = 2*envPtr->auxDataArrayEnd;
+	size_t newBytes = newElems * sizeof(AuxData);
+
+	if (envPtr->mallocedAuxDataArray) {
+	    envPtr->auxDataArrayPtr =
+		    ckrealloc(envPtr->auxDataArrayPtr, newBytes);
+	} else {
+	    /*
+	     * envPtr->auxDataArrayPtr isn't a ckalloc'd pointer, so we must
+	     * code a ckrealloc equivalent for ourselves.
+	     */
+
+	    AuxData *newPtr = ckalloc(newBytes);
+
+	    memcpy(newPtr, envPtr->auxDataArrayPtr, currBytes);
+	    envPtr->auxDataArrayPtr = newPtr;
+	    envPtr->mallocedAuxDataArray = 1;
+	}
+	envPtr->auxDataArrayEnd = newElems;
+    }
+    envPtr->auxDataArrayNext++;
+
+    auxDataPtr = &envPtr->auxDataArrayPtr[index];
+    auxDataPtr->clientData = clientData;
+    auxDataPtr->type = typePtr;
+    return index;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitJumpFixupArray --
+ *
+ *	Initializes a JumpFixupArray structure to hold some number of jump
+ *	fixup entries.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The JumpFixupArray structure is initialized.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitJumpFixupArray(
+    register JumpFixupArray *fixupArrayPtr)
+				/* Points to the JumpFixupArray structure to
+				 * initialize. */
+{
+    fixupArrayPtr->fixup = fixupArrayPtr->staticFixupSpace;
+    fixupArrayPtr->next = 0;
+    fixupArrayPtr->end = JUMPFIXUP_INIT_ENTRIES - 1;
+    fixupArrayPtr->mallocedArray = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclExpandJumpFixupArray --
+ *
+ *	Procedure that uses malloc to allocate more storage for a jump fixup
+ *	array.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The jump fixup array in *fixupArrayPtr is reallocated to a new array
+ *	of double the size, and if fixupArrayPtr->mallocedArray is non-zero
+ *	the old array is freed. Jump fixup structures are copied from the old
+ *	array to the new one.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclExpandJumpFixupArray(
+    register JumpFixupArray *fixupArrayPtr)
+				/* Points to the JumpFixupArray structure to
+				 * enlarge. */
+{
+    /*
+     * The currently allocated jump fixup entries are stored from fixup[0] up
+     * to fixup[fixupArrayPtr->fixupNext] (*not* inclusive). We assume
+     * fixupArrayPtr->fixupNext is equal to fixupArrayPtr->fixupEnd.
+     */
+
+    size_t currBytes = fixupArrayPtr->next * sizeof(JumpFixup);
+    int newElems = 2*(fixupArrayPtr->end + 1);
+    size_t newBytes = newElems * sizeof(JumpFixup);
+
+    if (fixupArrayPtr->mallocedArray) {
+	fixupArrayPtr->fixup = ckrealloc(fixupArrayPtr->fixup, newBytes);
+    } else {
+	/*
+	 * fixupArrayPtr->fixup isn't a ckalloc'd pointer, so we must code a
+	 * ckrealloc equivalent for ourselves.
+	 */
+
+	JumpFixup *newPtr = ckalloc(newBytes);
+
+	memcpy(newPtr, fixupArrayPtr->fixup, currBytes);
+	fixupArrayPtr->fixup = newPtr;
+	fixupArrayPtr->mallocedArray = 1;
+    }
+    fixupArrayPtr->end = newElems;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreeJumpFixupArray --
+ *
+ *	Free any storage allocated in a jump fixup array structure.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocated storage in the JumpFixupArray structure is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFreeJumpFixupArray(
+    register JumpFixupArray *fixupArrayPtr)
+				/* Points to the JumpFixupArray structure to
+				 * free. */
+{
+    if (fixupArrayPtr->mallocedArray) {
+	ckfree(fixupArrayPtr->fixup);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclEmitForwardJump --
+ *
+ *	Procedure to emit a two-byte forward jump of kind "jumpType". Since
+ *	the jump may later have to be grown to five bytes if the jump target
+ *	is more than, say, 127 bytes away, this procedure also initializes a
+ *	JumpFixup record with information about the jump.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The JumpFixup record pointed to by "jumpFixupPtr" is initialized with
+ *	information needed later if the jump is to be grown. Also, a two byte
+ *	jump of the designated type is emitted at the current point in the
+ *	bytecode stream.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclEmitForwardJump(
+    CompileEnv *envPtr,		/* Points to the CompileEnv structure that
+				 * holds the resulting instruction. */
+    TclJumpType jumpType,	/* Indicates the kind of jump: if true or
+				 * false or unconditional. */
+    JumpFixup *jumpFixupPtr)	/* Points to the JumpFixup structure to
+				 * initialize with information about this
+				 * forward jump. */
+{
+    /*
+     * Initialize the JumpFixup structure:
+     *    - codeOffset is offset of first byte of jump below
+     *    - cmdIndex is index of the command after the current one
+     *    - exceptIndex is the index of the first ExceptionRange after the
+     *	    current one.
+     */
+
+    jumpFixupPtr->jumpType = jumpType;
+    jumpFixupPtr->codeOffset = envPtr->codeNext - envPtr->codeStart;
+    jumpFixupPtr->cmdIndex = envPtr->numCommands;
+    jumpFixupPtr->exceptIndex = envPtr->exceptArrayNext;
+
+    switch (jumpType) {
+    case TCL_UNCONDITIONAL_JUMP:
+	TclEmitInstInt1(INST_JUMP1, 0, envPtr);
+	break;
+    case TCL_TRUE_JUMP:
+	TclEmitInstInt1(INST_JUMP_TRUE1, 0, envPtr);
+	break;
+    default:
+	TclEmitInstInt1(INST_JUMP_FALSE1, 0, envPtr);
+	break;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFixupForwardJump --
+ *
+ *	Procedure that updates a previously-emitted forward jump to jump a
+ *	specified number of bytes, "jumpDist". If necessary, the jump is grown
+ *	from two to five bytes; this is done if the jump distance is greater
+ *	than "distThreshold" (normally 127 bytes). The jump is described by a
+ *	JumpFixup record previously initialized by TclEmitForwardJump.
+ *
+ * Results:
+ *	1 if the jump was grown and subsequent instructions had to be moved;
+ *	otherwise 0. This result is returned to allow callers to update any
+ *	additional code offsets they may hold.
+ *
+ * Side effects:
+ *	The jump may be grown and subsequent instructions moved. If this
+ *	happens, the code offsets for any commands and any ExceptionRange
+ *	records between the jump and the current code address will be updated
+ *	to reflect the moved code. Also, the bytecode instruction array in the
+ *	CompileEnv structure may be grown and reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFixupForwardJump(
+    CompileEnv *envPtr,		/* Points to the CompileEnv structure that
+				 * holds the resulting instruction. */
+    JumpFixup *jumpFixupPtr,	/* Points to the JumpFixup structure that
+				 * describes the forward jump. */
+    int jumpDist,		/* Jump distance to set in jump instr. */
+    int distThreshold)		/* Maximum distance before the two byte jump
+				 * is grown to five bytes. */
+{
+    unsigned char *jumpPc, *p;
+    int firstCmd, lastCmd, firstRange, lastRange, k;
+    unsigned numBytes;
+
+    if (jumpDist <= distThreshold) {
+	jumpPc = envPtr->codeStart + jumpFixupPtr->codeOffset;
+	switch (jumpFixupPtr->jumpType) {
+	case TCL_UNCONDITIONAL_JUMP:
+	    TclUpdateInstInt1AtPc(INST_JUMP1, jumpDist, jumpPc);
+	    break;
+	case TCL_TRUE_JUMP:
+	    TclUpdateInstInt1AtPc(INST_JUMP_TRUE1, jumpDist, jumpPc);
+	    break;
+	default:
+	    TclUpdateInstInt1AtPc(INST_JUMP_FALSE1, jumpDist, jumpPc);
+	    break;
+	}
+	return 0;
+    }
+
+    /*
+     * We must grow the jump then move subsequent instructions down. Note that
+     * if we expand the space for generated instructions, code addresses might
+     * change; be careful about updating any of these addresses held in
+     * variables.
+     */
+
+    if ((envPtr->codeNext + 3) > envPtr->codeEnd) {
+	TclExpandCodeArray(envPtr);
+    }
+    jumpPc = envPtr->codeStart + jumpFixupPtr->codeOffset;
+    numBytes = envPtr->codeNext-jumpPc-2;
+    p = jumpPc+2;
+    memmove(p+3, p, numBytes);
+
+    envPtr->codeNext += 3;
+    jumpDist += 3;
+    switch (jumpFixupPtr->jumpType) {
+    case TCL_UNCONDITIONAL_JUMP:
+	TclUpdateInstInt4AtPc(INST_JUMP4, jumpDist, jumpPc);
+	break;
+    case TCL_TRUE_JUMP:
+	TclUpdateInstInt4AtPc(INST_JUMP_TRUE4, jumpDist, jumpPc);
+	break;
+    default:
+	TclUpdateInstInt4AtPc(INST_JUMP_FALSE4, jumpDist, jumpPc);
+	break;
+    }
+
+    /*
+     * Adjust the code offsets for any commands and any ExceptionRange records
+     * between the jump and the current code address.
+     */
+
+    firstCmd = jumpFixupPtr->cmdIndex;
+    lastCmd = envPtr->numCommands - 1;
+    if (firstCmd < lastCmd) {
+	for (k = firstCmd;  k <= lastCmd;  k++) {
+	    envPtr->cmdMapPtr[k].codeOffset += 3;
+	}
+    }
+
+    firstRange = jumpFixupPtr->exceptIndex;
+    lastRange = envPtr->exceptArrayNext - 1;
+    for (k = firstRange;  k <= lastRange;  k++) {
+	ExceptionRange *rangePtr = &envPtr->exceptArrayPtr[k];
+
+	rangePtr->codeOffset += 3;
+	switch (rangePtr->type) {
+	case LOOP_EXCEPTION_RANGE:
+	    rangePtr->breakOffset += 3;
+	    if (rangePtr->continueOffset != -1) {
+		rangePtr->continueOffset += 3;
+	    }
+	    break;
+	case CATCH_EXCEPTION_RANGE:
+	    rangePtr->catchOffset += 3;
+	    break;
+	default:
+	    Tcl_Panic("TclFixupForwardJump: bad ExceptionRange type %d",
+		    rangePtr->type);
+	}
+    }
+
+    for (k = 0 ; k < envPtr->exceptArrayNext ; k++) {
+	ExceptionAux *auxPtr = &envPtr->exceptAuxArrayPtr[k];
+	int i;
+
+	for (i=0 ; i<auxPtr->numBreakTargets ; i++) {
+	    if (jumpFixupPtr->codeOffset < auxPtr->breakTargets[i]) {
+		auxPtr->breakTargets[i] += 3;
+	    }
+	}
+	for (i=0 ; i<auxPtr->numContinueTargets ; i++) {
+	    if (jumpFixupPtr->codeOffset < auxPtr->continueTargets[i]) {
+		auxPtr->continueTargets[i] += 3;
+	    }
+	}
+    }
+
+    return 1;			/* the jump was grown */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclEmitInvoke --
+ *
+ *	Emit one of the invoke-related instructions, wrapping it if necessary
+ *	in code that ensures that any break or continue operation passing
+ *	through it gets the stack unwinding correct, converting it into an
+ *	internal jump if in an appropriate context.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Issues the jump with all correct stack management. May create another
+ *	loop exception range; pointers to ExceptionRange and ExceptionAux
+ *	structures should not be held across this call.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclEmitInvoke(
+    CompileEnv *envPtr,
+    int opcode,
+    ...)
+{
+    va_list argList;
+    ExceptionRange *rangePtr;
+    ExceptionAux *auxBreakPtr, *auxContinuePtr;
+    int arg1, arg2, wordCount = 0, expandCount = 0;
+    int loopRange = 0, breakRange = 0, continueRange = 0;
+    int cleanup, depth = TclGetStackDepth(envPtr);
+
+    /*
+     * Parse the arguments.
+     */
+
+    va_start(argList, opcode);
+    switch (opcode) {
+    case INST_INVOKE_STK1:
+	wordCount = arg1 = cleanup = va_arg(argList, int);
+	arg2 = 0;
+	break;
+    case INST_INVOKE_STK4:
+	wordCount = arg1 = cleanup = va_arg(argList, int);
+	arg2 = 0;
+	break;
+    case INST_INVOKE_REPLACE:
+	arg1 = va_arg(argList, int);
+	arg2 = va_arg(argList, int);
+	wordCount = arg1 + arg2 - 1;
+	cleanup = arg1 + 1;
+	break;
+    default:
+	Tcl_Panic("unexpected opcode");
+    case INST_EVAL_STK:
+	wordCount = cleanup = 1;
+	arg1 = arg2 = 0;
+	break;
+    case INST_RETURN_STK:
+	wordCount = cleanup = 2;
+	arg1 = arg2 = 0;
+	break;
+    case INST_INVOKE_EXPANDED:
+	wordCount = arg1 = cleanup = va_arg(argList, int);
+	arg2 = 0;
+	expandCount = 1;
+	break;
+    }
+    va_end(argList);
+
+    /*
+     * Determine if we need to handle break and continue exceptions with a
+     * special handling exception range (so that we can correctly unwind the
+     * stack).
+     *
+     * These must be done separately; they can be different (especially for
+     * calls from inside a [for] increment clause).
+     */
+
+    rangePtr = TclGetInnermostExceptionRange(envPtr, TCL_CONTINUE,
+	    &auxContinuePtr);
+    if (rangePtr == NULL || rangePtr->type != LOOP_EXCEPTION_RANGE) {
+	auxContinuePtr = NULL;
+    } else if (auxContinuePtr->stackDepth == envPtr->currStackDepth-wordCount
+	    && auxContinuePtr->expandTarget == envPtr->expandCount-expandCount) {
+	auxContinuePtr = NULL;
+    } else {
+	continueRange = auxContinuePtr - envPtr->exceptAuxArrayPtr;
+    }
+
+    rangePtr = TclGetInnermostExceptionRange(envPtr, TCL_BREAK, &auxBreakPtr);
+    if (rangePtr == NULL || rangePtr->type != LOOP_EXCEPTION_RANGE) {
+	auxBreakPtr = NULL;
+    } else if (auxContinuePtr == NULL
+	    && auxBreakPtr->stackDepth == envPtr->currStackDepth-wordCount
+	    && auxBreakPtr->expandTarget == envPtr->expandCount-expandCount) {
+	auxBreakPtr = NULL;
+    } else {
+	breakRange = auxBreakPtr - envPtr->exceptAuxArrayPtr;
+    }
+
+    if (auxBreakPtr != NULL || auxContinuePtr != NULL) {
+	loopRange = TclCreateExceptRange(LOOP_EXCEPTION_RANGE, envPtr);
+	ExceptionRangeStarts(envPtr, loopRange);
+    }
+
+    /*
+     * Issue the invoke itself.
+     */
+
+    switch (opcode) {
+    case INST_INVOKE_STK1:
+	TclEmitInstInt1(INST_INVOKE_STK1, arg1, envPtr);
+	break;
+    case INST_INVOKE_STK4:
+	TclEmitInstInt4(INST_INVOKE_STK4, arg1, envPtr);
+	break;
+    case INST_INVOKE_EXPANDED:
+	TclEmitOpcode(INST_INVOKE_EXPANDED, envPtr);
+	envPtr->expandCount--;
+	TclAdjustStackDepth(1 - arg1, envPtr);
+	break;
+    case INST_EVAL_STK:
+	TclEmitOpcode(INST_EVAL_STK, envPtr);
+	break;
+    case INST_RETURN_STK:
+	TclEmitOpcode(INST_RETURN_STK, envPtr);
+	break;
+    case INST_INVOKE_REPLACE:
+	TclEmitInstInt4(INST_INVOKE_REPLACE, arg1, envPtr);
+	TclEmitInt1(arg2, envPtr);
+	TclAdjustStackDepth(-1, envPtr); /* Correction to stack depth calcs */
+	break;
+    }
+
+    /*
+     * If we're generating a special wrapper exception range, we need to
+     * finish that up now.
+     */
+
+    if (auxBreakPtr != NULL || auxContinuePtr != NULL) {
+	int savedStackDepth = envPtr->currStackDepth;
+	int savedExpandCount = envPtr->expandCount;
+	JumpFixup nonTrapFixup;
+
+	if (auxBreakPtr != NULL) {
+	    auxBreakPtr = envPtr->exceptAuxArrayPtr + breakRange;
+	}
+	if (auxContinuePtr != NULL) {
+	    auxContinuePtr = envPtr->exceptAuxArrayPtr + continueRange;
+	}
+
+	ExceptionRangeEnds(envPtr, loopRange);
+	TclEmitForwardJump(envPtr, TCL_UNCONDITIONAL_JUMP, &nonTrapFixup);
+
+	/*
+	 * Careful! When generating these stack unwinding sequences, the depth
+	 * of stack in the cases where they are taken is not the same as if
+	 * the exception is not taken.
+	 */
+
+	if (auxBreakPtr != NULL) {
+	    TclAdjustStackDepth(-1, envPtr);
+
+	    ExceptionRangeTarget(envPtr, loopRange, breakOffset);
+	    TclCleanupStackForBreakContinue(envPtr, auxBreakPtr);
+	    TclAddLoopBreakFixup(envPtr, auxBreakPtr);
+	    TclAdjustStackDepth(1, envPtr);
+
+	    envPtr->currStackDepth = savedStackDepth;
+	    envPtr->expandCount = savedExpandCount;
+	}
+
+	if (auxContinuePtr != NULL) {
+	    TclAdjustStackDepth(-1, envPtr);
+
+	    ExceptionRangeTarget(envPtr, loopRange, continueOffset);
+	    TclCleanupStackForBreakContinue(envPtr, auxContinuePtr);
+	    TclAddLoopContinueFixup(envPtr, auxContinuePtr);
+	    TclAdjustStackDepth(1, envPtr);
+
+	    envPtr->currStackDepth = savedStackDepth;
+	    envPtr->expandCount = savedExpandCount;
+	}
+
+	TclFinalizeLoopExceptionRange(envPtr, loopRange);
+	TclFixupForwardJumpToHere(envPtr, &nonTrapFixup, 127);
+    }
+    TclCheckStackDepth(depth+1-cleanup, envPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetInstructionTable --
+ *
+ *	Returns a pointer to the table describing Tcl bytecode instructions.
+ *	This procedure is defined so that clients can access the pointer from
+ *	outside the TCL DLLs.
+ *
+ * Results:
+ *	Returns a pointer to the global instruction table, same as the
+ *	expression (&tclInstructionTable[0]).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const void * /* == InstructionDesc* == */
+TclGetInstructionTable(void)
+{
+    return &tclInstructionTable[0];
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetCmdLocEncodingSize --
+ *
+ *	Computes the total number of bytes needed to encode the command
+ *	location information for some compiled code.
+ *
+ * Results:
+ *	The byte count needed to encode the compiled location information.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetCmdLocEncodingSize(
+    CompileEnv *envPtr)		/* Points to compilation environment structure
+				 * containing the CmdLocation structure to
+				 * encode. */
+{
+    register CmdLocation *mapPtr = envPtr->cmdMapPtr;
+    int numCmds = envPtr->numCommands;
+    int codeDelta, codeLen, srcDelta, srcLen;
+    int codeDeltaNext, codeLengthNext, srcDeltaNext, srcLengthNext;
+				/* The offsets in their respective byte
+				 * sequences where the next encoded offset or
+				 * length should go. */
+    int prevCodeOffset, prevSrcOffset, i;
+
+    codeDeltaNext = codeLengthNext = srcDeltaNext = srcLengthNext = 0;
+    prevCodeOffset = prevSrcOffset = 0;
+    for (i = 0;  i < numCmds;  i++) {
+	codeDelta = mapPtr[i].codeOffset - prevCodeOffset;
+	if (codeDelta < 0) {
+	    Tcl_Panic("GetCmdLocEncodingSize: bad code offset");
+	} else if (codeDelta <= 127) {
+	    codeDeltaNext++;
+	} else {
+	    codeDeltaNext += 5;	/* 1 byte for 0xFF, 4 for positive delta */
+	}
+	prevCodeOffset = mapPtr[i].codeOffset;
+
+	codeLen = mapPtr[i].numCodeBytes;
+	if (codeLen < 0) {
+	    Tcl_Panic("GetCmdLocEncodingSize: bad code length");
+	} else if (codeLen <= 127) {
+	    codeLengthNext++;
+	} else {
+	    codeLengthNext += 5;/* 1 byte for 0xFF, 4 for length */
+	}
+
+	srcDelta = mapPtr[i].srcOffset - prevSrcOffset;
+	if ((-127 <= srcDelta) && (srcDelta <= 127) && (srcDelta != -1)) {
+	    srcDeltaNext++;
+	} else {
+	    srcDeltaNext += 5;	/* 1 byte for 0xFF, 4 for delta */
+	}
+	prevSrcOffset = mapPtr[i].srcOffset;
+
+	srcLen = mapPtr[i].numSrcBytes;
+	if (srcLen < 0) {
+	    Tcl_Panic("GetCmdLocEncodingSize: bad source length");
+	} else if (srcLen <= 127) {
+	    srcLengthNext++;
+	} else {
+	    srcLengthNext += 5;	/* 1 byte for 0xFF, 4 for length */
+	}
+    }
+
+    return (codeDeltaNext + codeLengthNext + srcDeltaNext + srcLengthNext);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EncodeCmdLocMap --
+ *
+ *	Encode the command location information for some compiled code into a
+ *	ByteCode structure. The encoded command location map is stored as
+ *	three adjacent byte sequences.
+ *
+ * Results:
+ *	Pointer to the first byte after the encoded command location
+ *	information.
+ *
+ * Side effects:
+ *	The encoded information is stored into the block of memory headed by
+ *	codePtr. Also records pointers to the start of the four byte sequences
+ *	in fields in codePtr's ByteCode header structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static unsigned char *
+EncodeCmdLocMap(
+    CompileEnv *envPtr,		/* Points to compilation environment structure
+				 * containing the CmdLocation structure to
+				 * encode. */
+    ByteCode *codePtr,		/* ByteCode in which to encode envPtr's
+				 * command location information. */
+    unsigned char *startPtr)	/* Points to the first byte in codePtr's
+				 * memory block where the location information
+				 * is to be stored. */
+{
+    register CmdLocation *mapPtr = envPtr->cmdMapPtr;
+    int numCmds = envPtr->numCommands;
+    register unsigned char *p = startPtr;
+    int codeDelta, codeLen, srcDelta, srcLen, prevOffset;
+    register int i;
+
+    /*
+     * Encode the code offset for each command as a sequence of deltas.
+     */
+
+    codePtr->codeDeltaStart = p;
+    prevOffset = 0;
+    for (i = 0;  i < numCmds;  i++) {
+	codeDelta = mapPtr[i].codeOffset - prevOffset;
+	if (codeDelta < 0) {
+	    Tcl_Panic("EncodeCmdLocMap: bad code offset");
+	} else if (codeDelta <= 127) {
+	    TclStoreInt1AtPtr(codeDelta, p);
+	    p++;
+	} else {
+	    TclStoreInt1AtPtr(0xFF, p);
+	    p++;
+	    TclStoreInt4AtPtr(codeDelta, p);
+	    p += 4;
+	}
+	prevOffset = mapPtr[i].codeOffset;
+    }
+
+    /*
+     * Encode the code length for each command.
+     */
+
+    codePtr->codeLengthStart = p;
+    for (i = 0;  i < numCmds;  i++) {
+	codeLen = mapPtr[i].numCodeBytes;
+	if (codeLen < 0) {
+	    Tcl_Panic("EncodeCmdLocMap: bad code length");
+	} else if (codeLen <= 127) {
+	    TclStoreInt1AtPtr(codeLen, p);
+	    p++;
+	} else {
+	    TclStoreInt1AtPtr(0xFF, p);
+	    p++;
+	    TclStoreInt4AtPtr(codeLen, p);
+	    p += 4;
+	}
+    }
+
+    /*
+     * Encode the source offset for each command as a sequence of deltas.
+     */
+
+    codePtr->srcDeltaStart = p;
+    prevOffset = 0;
+    for (i = 0;  i < numCmds;  i++) {
+	srcDelta = mapPtr[i].srcOffset - prevOffset;
+	if ((-127 <= srcDelta) && (srcDelta <= 127) && (srcDelta != -1)) {
+	    TclStoreInt1AtPtr(srcDelta, p);
+	    p++;
+	} else {
+	    TclStoreInt1AtPtr(0xFF, p);
+	    p++;
+	    TclStoreInt4AtPtr(srcDelta, p);
+	    p += 4;
+	}
+	prevOffset = mapPtr[i].srcOffset;
+    }
+
+    /*
+     * Encode the source length for each command.
+     */
+
+    codePtr->srcLengthStart = p;
+    for (i = 0;  i < numCmds;  i++) {
+	srcLen = mapPtr[i].numSrcBytes;
+	if (srcLen < 0) {
+	    Tcl_Panic("EncodeCmdLocMap: bad source length");
+	} else if (srcLen <= 127) {
+	    TclStoreInt1AtPtr(srcLen, p);
+	    p++;
+	} else {
+	    TclStoreInt1AtPtr(0xFF, p);
+	    p++;
+	    TclStoreInt4AtPtr(srcLen, p);
+	    p += 4;
+	}
+    }
+
+    return p;
+}
+
+#ifdef TCL_COMPILE_STATS
+/*
+ *----------------------------------------------------------------------
+ *
+ * RecordByteCodeStats --
+ *
+ *	Accumulates various compilation-related statistics for each newly
+ *	compiled ByteCode. Called by the TclInitByteCodeObj when Tcl is
+ *	compiled with the -DTCL_COMPILE_STATS flag
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Accumulates aggregate code-related statistics in the interpreter's
+ *	ByteCodeStats structure. Records statistics specific to a ByteCode in
+ *	its ByteCode structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+RecordByteCodeStats(
+    ByteCode *codePtr)		/* Points to ByteCode structure with info
+				 * to add to accumulated statistics. */
+{
+    Interp *iPtr = (Interp *) *codePtr->interpHandle;
+    register ByteCodeStats *statsPtr;
+
+    if (iPtr == NULL) {
+	/* Avoid segfaulting in case we're called in a deleted interp */
+	return;
+    }
+    statsPtr = &(iPtr->stats);
+
+    statsPtr->numCompilations++;
+    statsPtr->totalSrcBytes += (double) codePtr->numSrcBytes;
+    statsPtr->totalByteCodeBytes += (double) codePtr->structureSize;
+    statsPtr->currentSrcBytes += (double) codePtr->numSrcBytes;
+    statsPtr->currentByteCodeBytes += (double) codePtr->structureSize;
+
+    statsPtr->srcCount[TclLog2(codePtr->numSrcBytes)]++;
+    statsPtr->byteCodeCount[TclLog2((int) codePtr->structureSize)]++;
+
+    statsPtr->currentInstBytes += (double) codePtr->numCodeBytes;
+    statsPtr->currentLitBytes += (double)
+	    codePtr->numLitObjects * sizeof(Tcl_Obj *);
+    statsPtr->currentExceptBytes += (double)
+	    codePtr->numExceptRanges * sizeof(ExceptionRange);
+    statsPtr->currentAuxBytes += (double)
+	    codePtr->numAuxDataItems * sizeof(AuxData);
+    statsPtr->currentCmdMapBytes += (double) codePtr->numCmdLocBytes;
+}
+#endif /* TCL_COMPILE_STATS */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * End:
+ */
diff --git a/generic/tclCompile.h b/generic/tclCompile.h
new file mode 100644
index 0000000..bd7aaab
--- /dev/null
+++ b/generic/tclCompile.h
@@ -0,0 +1,1890 @@
+/*
+ * tclCompile.h --
+ *
+ * Copyright (c) 1996-1998 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 by Scriptics Corporation.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLCOMPILATION
+#define _TCLCOMPILATION 1
+
+#include "tclInt.h"
+
+struct ByteCode;		/* Forward declaration. */
+
+/*
+ *------------------------------------------------------------------------
+ * Variables related to compilation. These are used in tclCompile.c,
+ * tclExecute.c, tclBasic.c, and their clients.
+ *------------------------------------------------------------------------
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ * Variable that controls whether compilation tracing is enabled and, if so,
+ * what level of tracing is desired:
+ *    0: no compilation tracing
+ *    1: summarize compilation of top level cmds and proc bodies
+ *    2: display all instructions of each ByteCode compiled
+ * This variable is linked to the Tcl variable "tcl_traceCompile".
+ */
+
+MODULE_SCOPE int 	tclTraceCompile;
+
+/*
+ * Variable that controls whether execution tracing is enabled and, if so,
+ * what level of tracing is desired:
+ *    0: no execution tracing
+ *    1: trace invocations of Tcl procs only
+ *    2: trace invocations of all (not compiled away) commands
+ *    3: display each instruction executed
+ * This variable is linked to the Tcl variable "tcl_traceExec".
+ */
+
+MODULE_SCOPE int 	tclTraceExec;
+#endif
+
+/*
+ * The type of lambda expressions. Note that every lambda will *always* have a
+ * string representation.
+ */
+
+MODULE_SCOPE const Tcl_ObjType tclLambdaType;
+
+/*
+ *------------------------------------------------------------------------
+ * Data structures related to compilation.
+ *------------------------------------------------------------------------
+ */
+
+/*
+ * The structure used to implement Tcl "exceptions" (exceptional returns): for
+ * example, those generated in loops by the break and continue commands, and
+ * those generated by scripts and caught by the catch command. This
+ * ExceptionRange structure describes a range of code (e.g., a loop body), the
+ * kind of exceptions (e.g., a break or continue) that might occur, and the PC
+ * offsets to jump to if a matching exception does occur. Exception ranges can
+ * nest so this structure includes a nesting level that is used at runtime to
+ * find the closest exception range surrounding a PC. For example, when a
+ * break command is executed, the ExceptionRange structure for the most deeply
+ * nested loop, if any, is found and used. These structures are also generated
+ * for the "next" subcommands of for loops since a break there terminates the
+ * for command. This means a for command actually generates two LoopInfo
+ * structures.
+ */
+
+typedef enum {
+    LOOP_EXCEPTION_RANGE,	/* Exception's range is part of a loop. Break
+				 * and continue "exceptions" cause jumps to
+				 * appropriate PC offsets. */
+    CATCH_EXCEPTION_RANGE	/* Exception's range is controlled by a catch
+				 * command. Errors in the range cause a jump
+				 * to a catch PC offset. */
+} ExceptionRangeType;
+
+typedef struct ExceptionRange {
+    ExceptionRangeType type;	/* The kind of ExceptionRange. */
+    int nestingLevel;		/* Static depth of the exception range. Used
+				 * to find the most deeply-nested range
+				 * surrounding a PC at runtime. */
+    int codeOffset;		/* Offset of the first instruction byte of the
+				 * code range. */
+    int numCodeBytes;		/* Number of bytes in the code range. */
+    int breakOffset;		/* If LOOP_EXCEPTION_RANGE, the target PC
+				 * offset for a break command in the range. */
+    int continueOffset;		/* If LOOP_EXCEPTION_RANGE and not -1, the
+				 * target PC offset for a continue command in
+				 * the code range. Otherwise, ignore this
+				 * range when processing a continue
+				 * command. */
+    int catchOffset;		/* If a CATCH_EXCEPTION_RANGE, the target PC
+				 * offset for any "exception" in range. */
+} ExceptionRange;
+
+/*
+ * Auxiliary data used when issuing (currently just loop) exception ranges,
+ * but which is not required during execution.
+ */
+
+typedef struct ExceptionAux {
+    int supportsContinue;	/* Whether this exception range will have a
+				 * continueOffset created for it; if it is a
+				 * loop exception range that *doesn't* have
+				 * one (see [for] next-clause) then we must
+				 * not pick up the range when scanning for a
+				 * target to continue to. */
+    int stackDepth;		/* The stack depth at the point where the
+				 * exception range was created. This is used
+				 * to calculate the number of POPs required to
+				 * restore the stack to its prior state. */
+    int expandTarget;		/* The number of expansions expected on the
+				 * auxData stack at the time the loop starts;
+				 * we can't currently discard them except by
+				 * doing INST_INVOKE_EXPANDED; this is a known
+				 * problem. */
+    int expandTargetDepth;	/* The stack depth expected at the outermost
+				 * expansion within the loop. Not meaningful
+				 * if there are no open expansions between the
+				 * looping level and the point of jump
+				 * issue. */
+    int numBreakTargets;	/* The number of [break]s that want to be
+				 * targeted to the place where this loop
+				 * exception will be bound to. */
+    unsigned int *breakTargets;	/* The offsets of the INST_JUMP4 instructions
+				 * issued by the [break]s that we must
+				 * update. Note that resizing a jump (via
+				 * TclFixupForwardJump) can cause the contents
+				 * of this array to be updated. When
+				 * numBreakTargets==0, this is NULL. */
+    int allocBreakTargets;	/* The size of the breakTargets array. */
+    int numContinueTargets;	/* The number of [continue]s that want to be
+				 * targeted to the place where this loop
+				 * exception will be bound to. */
+    unsigned int *continueTargets; /* The offsets of the INST_JUMP4 instructions
+				 * issued by the [continue]s that we must
+				 * update. Note that resizing a jump (via
+				 * TclFixupForwardJump) can cause the contents
+				 * of this array to be updated. When
+				 * numContinueTargets==0, this is NULL. */
+    int allocContinueTargets;	/* The size of the continueTargets array. */
+} ExceptionAux;
+
+/*
+ * Structure used to map between instruction pc and source locations. It
+ * defines for each compiled Tcl command its code's starting offset and its
+ * source's starting offset and length. Note that the code offset increases
+ * monotonically: that is, the table is sorted in code offset order. The
+ * source offset is not monotonic.
+ */
+
+typedef struct CmdLocation {
+    int codeOffset;		/* Offset of first byte of command code. */
+    int numCodeBytes;		/* Number of bytes for command's code. */
+    int srcOffset;		/* Offset of first char of the command. */
+    int numSrcBytes;		/* Number of command source chars. */
+} CmdLocation;
+
+/*
+ * TIP #280
+ * Structure to record additional location information for byte code. This
+ * information is internal and not saved. i.e. tbcload'ed code will not have
+ * this information. It records the lines for all words of all commands found
+ * in the byte code. The association with a ByteCode structure BC is done
+ * through the 'lineBCPtr' HashTable in Interp, keyed by the address of BC.
+ * Also recorded is information coming from the context, i.e. type of the
+ * frame and associated information, like the path of a sourced file.
+ */
+
+typedef struct ECL {
+    int srcOffset;		/* Command location to find the entry. */
+    int nline;			/* Number of words in the command */
+    int *line;			/* Line information for all words in the
+				 * command. */
+    int **next;			/* Transient information used by the compiler
+				 * for tracking of hidden continuation
+				 * lines. */
+} ECL;
+
+typedef struct ExtCmdLoc {
+    int type;			/* Context type. */
+    int start;			/* Starting line for compiled script. Needed
+				 * for the extended recompile check in
+				 * tclCompileObj. */
+    Tcl_Obj *path;		/* Path of the sourced file the command is
+				 * in. */
+    ECL *loc;			/* Command word locations (lines). */
+    int nloc;			/* Number of allocated entries in 'loc'. */
+    int nuloc;			/* Number of used entries in 'loc'. */
+} ExtCmdLoc;
+
+/*
+ * CompileProcs need the ability to record information during compilation that
+ * can be used by bytecode instructions during execution. The AuxData
+ * structure provides this "auxiliary data" mechanism. An arbitrary number of
+ * these structures can be stored in the ByteCode record (during compilation
+ * they are stored in a CompileEnv structure). Each AuxData record holds one
+ * word of client-specified data (often a pointer) and is given an index that
+ * instructions can later use to look up the structure and its data.
+ *
+ * The following definitions declare the types of procedures that are called
+ * to duplicate or free this auxiliary data when the containing ByteCode
+ * objects are duplicated and freed. Pointers to these procedures are kept in
+ * the AuxData structure.
+ */
+
+typedef ClientData (AuxDataDupProc)  (ClientData clientData);
+typedef void	   (AuxDataFreeProc) (ClientData clientData);
+typedef void	   (AuxDataPrintProc)(ClientData clientData,
+			    Tcl_Obj *appendObj, struct ByteCode *codePtr,
+			    unsigned int pcOffset);
+
+/*
+ * We define a separate AuxDataType struct to hold type-related information
+ * for the AuxData structure. This separation makes it possible for clients
+ * outside of the TCL core to manipulate (in a limited fashion!) AuxData; for
+ * example, it makes it possible to pickle and unpickle AuxData structs.
+ */
+
+typedef struct AuxDataType {
+    const char *name;		/* The name of the type. Types can be
+				 * registered and found by name */
+    AuxDataDupProc *dupProc;	/* Callback procedure to invoke when the aux
+				 * data is duplicated (e.g., when the ByteCode
+				 * structure containing the aux data is
+				 * duplicated). NULL means just copy the
+				 * source clientData bits; no proc need be
+				 * called. */
+    AuxDataFreeProc *freeProc;	/* Callback procedure to invoke when the aux
+				 * data is freed. NULL means no proc need be
+				 * called. */
+    AuxDataPrintProc *printProc;/* Callback function to invoke when printing
+				 * the aux data as part of debugging. NULL
+				 * means that the data can't be printed. */
+    AuxDataPrintProc *disassembleProc;
+				/* Callback function to invoke when doing a
+				 * disassembly of the aux data (like the
+				 * printProc, except that the output is
+				 * intended to be script-readable). The
+				 * appendObj argument should be filled in with
+				 * a descriptive dictionary; it will start out
+				 * with "name" mapped to the content of the
+				 * name field. NULL means that the printProc
+				 * should be used instead. */
+} AuxDataType;
+
+/*
+ * The definition of the AuxData structure that holds information created
+ * during compilation by CompileProcs and used by instructions during
+ * execution.
+ */
+
+typedef struct AuxData {
+    const AuxDataType *type;	/* Pointer to the AuxData type associated with
+				 * this ClientData. */
+    ClientData clientData;	/* The compilation data itself. */
+} AuxData;
+
+/*
+ * Structure defining the compilation environment. After compilation, fields
+ * describing bytecode instructions are copied out into the more compact
+ * ByteCode structure defined below.
+ */
+
+#define COMPILEENV_INIT_CODE_BYTES    250
+#define COMPILEENV_INIT_NUM_OBJECTS    60
+#define COMPILEENV_INIT_EXCEPT_RANGES   5
+#define COMPILEENV_INIT_CMD_MAP_SIZE   40
+#define COMPILEENV_INIT_AUX_DATA_SIZE   5
+
+typedef struct CompileEnv {
+    Interp *iPtr;		/* Interpreter containing the code being
+				 * compiled. Commands and their compile procs
+				 * are specific to an interpreter so the code
+				 * emitted will depend on the interpreter. */
+    const char *source;		/* The source string being compiled by
+				 * SetByteCodeFromAny. This pointer is not
+				 * owned by the CompileEnv and must not be
+				 * freed or changed by it. */
+    int numSrcBytes;		/* Number of bytes in source. */
+    Proc *procPtr;		/* If a procedure is being compiled, a pointer
+				 * to its Proc structure; otherwise NULL. Used
+				 * to compile local variables. Set from
+				 * information provided by ObjInterpProc in
+				 * tclProc.c. */
+    int numCommands;		/* Number of commands compiled. */
+    int exceptDepth;		/* Current exception range nesting level; -1
+				 * if not in any range currently. */
+    int maxExceptDepth;		/* Max nesting level of exception ranges; -1
+				 * if no ranges have been compiled. */
+    int maxStackDepth;		/* Maximum number of stack elements needed to
+				 * execute the code. Set by compilation
+				 * procedures before returning. */
+    int currStackDepth;		/* Current stack depth. */
+    LiteralTable localLitTable;	/* Contains LiteralEntry's describing all Tcl
+				 * objects referenced by this compiled code.
+				 * Indexed by the string representations of
+				 * the literals. Used to avoid creating
+				 * duplicate objects. */
+    unsigned char *codeStart;	/* Points to the first byte of the code. */
+    unsigned char *codeNext;	/* Points to next code array byte to use. */
+    unsigned char *codeEnd;	/* Points just after the last allocated code
+				 * array byte. */
+    int mallocedCodeArray;	/* Set 1 if code array was expanded and
+				 * codeStart points into the heap.*/
+    LiteralEntry *literalArrayPtr;
+    				/* Points to start of LiteralEntry array. */
+    int literalArrayNext;	/* Index of next free object array entry. */
+    int literalArrayEnd;	/* Index just after last obj array entry. */
+    int mallocedLiteralArray;	/* 1 if object array was expanded and objArray
+				 * points into the heap, else 0. */
+    ExceptionRange *exceptArrayPtr;
+    				/* Points to start of the ExceptionRange
+				 * array. */
+    int exceptArrayNext;	/* Next free ExceptionRange array index.
+				 * exceptArrayNext is the number of ranges and
+				 * (exceptArrayNext-1) is the index of the
+				 * current range's array entry. */
+    int exceptArrayEnd;		/* Index after the last ExceptionRange array
+				 * entry. */
+    int mallocedExceptArray;	/* 1 if ExceptionRange array was expanded and
+				 * exceptArrayPtr points in heap, else 0. */
+    ExceptionAux *exceptAuxArrayPtr;
+				/* Array of information used to restore the
+				 * state when processing BREAK/CONTINUE
+				 * exceptions. Must be the same size as the
+				 * exceptArrayPtr. */
+    CmdLocation *cmdMapPtr;	/* Points to start of CmdLocation array.
+				 * numCommands is the index of the next entry
+				 * to use; (numCommands-1) is the entry index
+				 * for the last command. */
+    int cmdMapEnd;		/* Index after last CmdLocation entry. */
+    int mallocedCmdMap;		/* 1 if command map array was expanded and
+				 * cmdMapPtr points in the heap, else 0. */
+    AuxData *auxDataArrayPtr;	/* Points to auxiliary data array start. */
+    int auxDataArrayNext;	/* Next free compile aux data array index.
+				 * auxDataArrayNext is the number of aux data
+				 * items and (auxDataArrayNext-1) is index of
+				 * current aux data array entry. */
+    int auxDataArrayEnd;	/* Index after last aux data array entry. */
+    int mallocedAuxDataArray;	/* 1 if aux data array was expanded and
+				 * auxDataArrayPtr points in heap else 0. */
+    unsigned char staticCodeSpace[COMPILEENV_INIT_CODE_BYTES];
+				/* Initial storage for code. */
+    LiteralEntry staticLiteralSpace[COMPILEENV_INIT_NUM_OBJECTS];
+				/* Initial storage of LiteralEntry array. */
+    ExceptionRange staticExceptArraySpace[COMPILEENV_INIT_EXCEPT_RANGES];
+				/* Initial ExceptionRange array storage. */
+    ExceptionAux staticExAuxArraySpace[COMPILEENV_INIT_EXCEPT_RANGES];
+				/* Initial static except auxiliary info array
+				 * storage. */
+    CmdLocation staticCmdMapSpace[COMPILEENV_INIT_CMD_MAP_SIZE];
+				/* Initial storage for cmd location map. */
+    AuxData staticAuxDataArraySpace[COMPILEENV_INIT_AUX_DATA_SIZE];
+				/* Initial storage for aux data array. */
+    /* TIP #280 */
+    ExtCmdLoc *extCmdMapPtr;	/* Extended command location information for
+				 * 'info frame'. */
+    int line;			/* First line of the script, based on the
+				 * invoking context, then the line of the
+				 * command currently compiled. */
+    int atCmdStart;		/* Flag to say whether an INST_START_CMD
+				 * should be issued; they should never be
+				 * issued repeatedly, as that is significantly
+				 * inefficient. If set to 2, that instruction
+				 * should not be issued at all (by the generic
+				 * part of the command compiler). */
+    int expandCount;		/* Number of INST_EXPAND_START instructions
+				 * encountered that have not yet been paired
+				 * with a corresponding
+				 * INST_INVOKE_EXPANDED. */
+    int *clNext;		/* If not NULL, it refers to the next slot in
+				 * clLoc to check for an invisible
+				 * continuation line. */
+} CompileEnv;
+
+/*
+ * The structure defining the bytecode instructions resulting from compiling a
+ * Tcl script. Note that this structure is variable length: a single heap
+ * object is allocated to hold the ByteCode structure immediately followed by
+ * the code bytes, the literal object array, the ExceptionRange array, the
+ * CmdLocation map, and the compilation AuxData array.
+ */
+
+/*
+ * A PRECOMPILED bytecode struct is one that was generated from a compiled
+ * image rather than implicitly compiled from source
+ */
+
+#define TCL_BYTECODE_PRECOMPILED		0x0001
+
+/*
+ * When a bytecode is compiled, interp or namespace resolvers have not been
+ * applied yet: this is indicated by the TCL_BYTECODE_RESOLVE_VARS flag.
+ */
+
+#define TCL_BYTECODE_RESOLVE_VARS		0x0002
+
+#define TCL_BYTECODE_RECOMPILE			0x0004
+
+typedef struct ByteCode {
+    TclHandle interpHandle;	/* Handle for interpreter containing the
+				 * compiled code. Commands and their compile
+				 * procs are specific to an interpreter so the
+				 * code emitted will depend on the
+				 * interpreter. */
+    unsigned int compileEpoch;	/* Value of iPtr->compileEpoch when this
+				 * ByteCode was compiled. Used to invalidate
+				 * code when, e.g., commands with compile
+				 * procs are redefined. */
+    Namespace *nsPtr;		/* Namespace context in which this code was
+				 * compiled. If the code is executed if a
+				 * different namespace, it must be
+				 * recompiled. */
+    size_t nsEpoch;		/* Value of nsPtr->resolverEpoch when this
+				 * ByteCode was compiled. Used to invalidate
+				 * code when new namespace resolution rules
+				 * are put into effect. */
+    int refCount;		/* Reference count: set 1 when created plus 1
+				 * for each execution of the code currently
+				 * active. This structure can be freed when
+				 * refCount becomes zero. */
+    unsigned int flags;		/* flags describing state for the codebyte.
+				 * this variable holds ORed values from the
+				 * TCL_BYTECODE_ masks defined above */
+    const char *source;		/* The source string from which this ByteCode
+				 * was compiled. Note that this pointer is not
+				 * owned by the ByteCode and must not be freed
+				 * or modified by it. */
+    Proc *procPtr;		/* If the ByteCode was compiled from a
+				 * procedure body, this is a pointer to its
+				 * Proc structure; otherwise NULL. This
+				 * pointer is also not owned by the ByteCode
+				 * and must not be freed by it. */
+    size_t structureSize;	/* Number of bytes in the ByteCode structure
+				 * itself. Does not include heap space for
+				 * literal Tcl objects or storage referenced
+				 * by AuxData entries. */
+    int numCommands;		/* Number of commands compiled. */
+    int numSrcBytes;		/* Number of source bytes compiled. */
+    int numCodeBytes;		/* Number of code bytes. */
+    int numLitObjects;		/* Number of objects in literal array. */
+    int numExceptRanges;	/* Number of ExceptionRange array elems. */
+    int numAuxDataItems;	/* Number of AuxData items. */
+    int numCmdLocBytes;		/* Number of bytes needed for encoded command
+				 * location information. */
+    int maxExceptDepth;		/* Maximum nesting level of ExceptionRanges;
+				 * -1 if no ranges were compiled. */
+    int maxStackDepth;		/* Maximum number of stack elements needed to
+				 * execute the code. */
+    unsigned char *codeStart;	/* Points to the first byte of the code. This
+				 * is just after the final ByteCode member
+				 * cmdMapPtr. */
+    Tcl_Obj **objArrayPtr;	/* Points to the start of the literal object
+				 * array. This is just after the last code
+				 * byte. */
+    ExceptionRange *exceptArrayPtr;
+    				/* Points to the start of the ExceptionRange
+				 * array. This is just after the last object
+				 * in the object array. */
+    AuxData *auxDataArrayPtr;	/* Points to the start of the auxiliary data
+				 * array. This is just after the last entry in
+				 * the ExceptionRange array. */
+    unsigned char *codeDeltaStart;
+				/* Points to the first of a sequence of bytes
+				 * that encode the change in the starting
+				 * offset of each command's code. If -127 <=
+				 * delta <= 127, it is encoded as 1 byte,
+				 * otherwise 0xFF (128) appears and the delta
+				 * is encoded by the next 4 bytes. Code deltas
+				 * are always positive. This sequence is just
+				 * after the last entry in the AuxData
+				 * array. */
+    unsigned char *codeLengthStart;
+				/* Points to the first of a sequence of bytes
+				 * that encode the length of each command's
+				 * code. The encoding is the same as for code
+				 * deltas. Code lengths are always positive.
+				 * This sequence is just after the last entry
+				 * in the code delta sequence. */
+    unsigned char *srcDeltaStart;
+				/* Points to the first of a sequence of bytes
+				 * that encode the change in the starting
+				 * offset of each command's source. The
+				 * encoding is the same as for code deltas.
+				 * Source deltas can be negative. This
+				 * sequence is just after the last byte in the
+				 * code length sequence. */
+    unsigned char *srcLengthStart;
+				/* Points to the first of a sequence of bytes
+				 * that encode the length of each command's
+				 * source. The encoding is the same as for
+				 * code deltas. Source lengths are always
+				 * positive. This sequence is just after the
+				 * last byte in the source delta sequence. */
+    LocalCache *localCachePtr;	/* Pointer to the start of the cached variable
+				 * names and initialisation data for local
+				 * variables. */
+#ifdef TCL_COMPILE_STATS
+    Tcl_Time createTime;	/* Absolute time when the ByteCode was
+				 * created. */
+#endif /* TCL_COMPILE_STATS */
+} ByteCode;
+
+/*
+ * Opcodes for the Tcl bytecode instructions. These must correspond to the
+ * entries in the table of instruction descriptions, tclInstructionTable, in
+ * tclCompile.c. Also, the order and number of the expression opcodes (e.g.,
+ * INST_LOR) must match the entries in the array operatorStrings in
+ * tclExecute.c.
+ */
+
+/* Opcodes 0 to 9 */
+#define INST_DONE			0
+#define INST_PUSH1			1
+#define INST_PUSH4			2
+#define INST_POP			3
+#define INST_DUP			4
+#define INST_STR_CONCAT1		5
+#define INST_INVOKE_STK1		6
+#define INST_INVOKE_STK4		7
+#define INST_EVAL_STK			8
+#define INST_EXPR_STK			9
+
+/* Opcodes 10 to 23 */
+#define INST_LOAD_SCALAR1		10
+#define INST_LOAD_SCALAR4		11
+#define INST_LOAD_SCALAR_STK		12
+#define INST_LOAD_ARRAY1		13
+#define INST_LOAD_ARRAY4		14
+#define INST_LOAD_ARRAY_STK		15
+#define INST_LOAD_STK			16
+#define INST_STORE_SCALAR1		17
+#define INST_STORE_SCALAR4		18
+#define INST_STORE_SCALAR_STK		19
+#define INST_STORE_ARRAY1		20
+#define INST_STORE_ARRAY4		21
+#define INST_STORE_ARRAY_STK		22
+#define INST_STORE_STK			23
+
+/* Opcodes 24 to 33 */
+#define INST_INCR_SCALAR1		24
+#define INST_INCR_SCALAR_STK		25
+#define INST_INCR_ARRAY1		26
+#define INST_INCR_ARRAY_STK		27
+#define INST_INCR_STK			28
+#define INST_INCR_SCALAR1_IMM		29
+#define INST_INCR_SCALAR_STK_IMM	30
+#define INST_INCR_ARRAY1_IMM		31
+#define INST_INCR_ARRAY_STK_IMM		32
+#define INST_INCR_STK_IMM		33
+
+/* Opcodes 34 to 39 */
+#define INST_JUMP1			34
+#define INST_JUMP4			35
+#define INST_JUMP_TRUE1			36
+#define INST_JUMP_TRUE4			37
+#define INST_JUMP_FALSE1		38
+#define INST_JUMP_FALSE4		39
+
+/* Opcodes 40 to 64 */
+#define INST_LOR			40
+#define INST_LAND			41
+#define INST_BITOR			42
+#define INST_BITXOR			43
+#define INST_BITAND			44
+#define INST_EQ				45
+#define INST_NEQ			46
+#define INST_LT				47
+#define INST_GT				48
+#define INST_LE				49
+#define INST_GE				50
+#define INST_LSHIFT			51
+#define INST_RSHIFT			52
+#define INST_ADD			53
+#define INST_SUB			54
+#define INST_MULT			55
+#define INST_DIV			56
+#define INST_MOD			57
+#define INST_UPLUS			58
+#define INST_UMINUS			59
+#define INST_BITNOT			60
+#define INST_LNOT			61
+#define INST_CALL_BUILTIN_FUNC1		62
+#define INST_CALL_FUNC1			63
+#define INST_TRY_CVT_TO_NUMERIC		64
+
+/* Opcodes 65 to 66 */
+#define INST_BREAK			65
+#define INST_CONTINUE			66
+
+/* Opcodes 67 to 68 */
+#define INST_FOREACH_START4		67 /* DEPRECATED */
+#define INST_FOREACH_STEP4		68 /* DEPRECATED */
+
+/* Opcodes 69 to 72 */
+#define INST_BEGIN_CATCH4		69
+#define INST_END_CATCH			70
+#define INST_PUSH_RESULT		71
+#define INST_PUSH_RETURN_CODE		72
+
+/* Opcodes 73 to 78 */
+#define INST_STR_EQ			73
+#define INST_STR_NEQ			74
+#define INST_STR_CMP			75
+#define INST_STR_LEN			76
+#define INST_STR_INDEX			77
+#define INST_STR_MATCH			78
+
+/* Opcodes 78 to 81 */
+#define INST_LIST			79
+#define INST_LIST_INDEX			80
+#define INST_LIST_LENGTH		81
+
+/* Opcodes 82 to 87 */
+#define INST_APPEND_SCALAR1		82
+#define INST_APPEND_SCALAR4		83
+#define INST_APPEND_ARRAY1		84
+#define INST_APPEND_ARRAY4		85
+#define INST_APPEND_ARRAY_STK		86
+#define INST_APPEND_STK			87
+
+/* Opcodes 88 to 93 */
+#define INST_LAPPEND_SCALAR1		88
+#define INST_LAPPEND_SCALAR4		89
+#define INST_LAPPEND_ARRAY1		90
+#define INST_LAPPEND_ARRAY4		91
+#define INST_LAPPEND_ARRAY_STK		92
+#define INST_LAPPEND_STK		93
+
+/* TIP #22 - LINDEX operator with flat arg list */
+
+#define INST_LIST_INDEX_MULTI		94
+
+/*
+ * TIP #33 - 'lset' command. Code gen also required a Forth-like
+ *	     OVER operation.
+ */
+
+#define INST_OVER			95
+#define INST_LSET_LIST			96
+#define INST_LSET_FLAT			97
+
+/* TIP#90 - 'return' command. */
+
+#define INST_RETURN_IMM			98
+
+/* TIP#123 - exponentiation operator. */
+
+#define INST_EXPON			99
+
+/* TIP #157 - {*}... (word expansion) language syntax support. */
+
+#define INST_EXPAND_START		100
+#define INST_EXPAND_STKTOP		101
+#define INST_INVOKE_EXPANDED		102
+
+/*
+ * TIP #57 - 'lassign' command. Code generation requires immediate
+ *	     LINDEX and LRANGE operators.
+ */
+
+#define INST_LIST_INDEX_IMM		103
+#define INST_LIST_RANGE_IMM		104
+
+#define INST_START_CMD			105
+
+#define INST_LIST_IN			106
+#define INST_LIST_NOT_IN		107
+
+#define INST_PUSH_RETURN_OPTIONS	108
+#define INST_RETURN_STK			109
+
+/*
+ * Dictionary (TIP#111) related commands.
+ */
+
+#define INST_DICT_GET			110
+#define INST_DICT_SET			111
+#define INST_DICT_UNSET			112
+#define INST_DICT_INCR_IMM		113
+#define INST_DICT_APPEND		114
+#define INST_DICT_LAPPEND		115
+#define INST_DICT_FIRST			116
+#define INST_DICT_NEXT			117
+#define INST_DICT_DONE			118
+#define INST_DICT_UPDATE_START		119
+#define INST_DICT_UPDATE_END		120
+
+/*
+ * Instruction to support jumps defined by tables (instead of the classic
+ * [switch] technique of chained comparisons).
+ */
+
+#define INST_JUMP_TABLE			121
+
+/*
+ * Instructions to support compilation of global, variable, upvar and
+ * [namespace upvar].
+ */
+
+#define INST_UPVAR			122
+#define INST_NSUPVAR			123
+#define INST_VARIABLE			124
+
+/* Instruction to support compiling syntax error to bytecode */
+
+#define INST_SYNTAX			125
+
+/* Instruction to reverse N items on top of stack */
+
+#define INST_REVERSE			126
+
+/* regexp instruction */
+
+#define INST_REGEXP			127
+
+/* For [info exists] compilation */
+#define INST_EXIST_SCALAR		128
+#define INST_EXIST_ARRAY		129
+#define INST_EXIST_ARRAY_STK		130
+#define INST_EXIST_STK			131
+
+/* For [subst] compilation */
+#define INST_NOP			132
+#define INST_RETURN_CODE_BRANCH		133
+
+/* For [unset] compilation */
+#define INST_UNSET_SCALAR		134
+#define INST_UNSET_ARRAY		135
+#define INST_UNSET_ARRAY_STK		136
+#define INST_UNSET_STK			137
+
+/* For [dict with], [dict exists], [dict create] and [dict merge] */
+#define INST_DICT_EXPAND		138
+#define INST_DICT_RECOMBINE_STK		139
+#define INST_DICT_RECOMBINE_IMM		140
+#define INST_DICT_EXISTS		141
+#define INST_DICT_VERIFY		142
+
+/* For [string map] and [regsub] compilation */
+#define INST_STR_MAP			143
+#define INST_STR_FIND			144
+#define INST_STR_FIND_LAST		145
+#define INST_STR_RANGE_IMM		146
+#define INST_STR_RANGE			147
+
+/* For operations to do with coroutines and other NRE-manipulators */
+#define INST_YIELD			148
+#define INST_COROUTINE_NAME		149
+#define INST_TAILCALL			150
+
+/* For compilation of basic information operations */
+#define INST_NS_CURRENT			151
+#define INST_INFO_LEVEL_NUM		152
+#define INST_INFO_LEVEL_ARGS		153
+#define INST_RESOLVE_COMMAND		154
+
+/* For compilation relating to TclOO */
+#define INST_TCLOO_SELF			155
+#define INST_TCLOO_CLASS		156
+#define INST_TCLOO_NS			157
+#define INST_TCLOO_IS_OBJECT		158
+
+/* For compilation of [array] subcommands */
+#define INST_ARRAY_EXISTS_STK		159
+#define INST_ARRAY_EXISTS_IMM		160
+#define INST_ARRAY_MAKE_STK		161
+#define INST_ARRAY_MAKE_IMM		162
+
+#define INST_INVOKE_REPLACE		163
+
+#define INST_LIST_CONCAT		164
+
+#define INST_EXPAND_DROP		165
+
+/* New foreach implementation */
+#define INST_FOREACH_START              166
+#define INST_FOREACH_STEP               167
+#define INST_FOREACH_END                168
+#define INST_LMAP_COLLECT               169
+
+/* For compilation of [string trim] and related */
+#define INST_STR_TRIM			170
+#define INST_STR_TRIM_LEFT		171
+#define INST_STR_TRIM_RIGHT		172
+
+#define INST_CONCAT_STK			173
+
+#define INST_STR_UPPER			174
+#define INST_STR_LOWER			175
+#define INST_STR_TITLE			176
+#define INST_STR_REPLACE		177
+
+#define INST_ORIGIN_COMMAND		178
+
+#define INST_TCLOO_NEXT			179
+#define INST_TCLOO_NEXT_CLASS		180
+
+#define INST_YIELD_TO_INVOKE		181
+
+#define INST_NUM_TYPE			182
+#define INST_TRY_CVT_TO_BOOLEAN		183
+#define INST_STR_CLASS			184
+
+#define INST_LAPPEND_LIST		185
+#define INST_LAPPEND_LIST_ARRAY		186
+#define INST_LAPPEND_LIST_ARRAY_STK	187
+#define INST_LAPPEND_LIST_STK		188
+
+#define INST_CLOCK_READ			189
+
+/* The last opcode */
+#define LAST_INST_OPCODE		189
+
+/*
+ * Table describing the Tcl bytecode instructions: their name (for displaying
+ * code), total number of code bytes required (including operand bytes), and a
+ * description of the type of each operand. These operand types include signed
+ * and unsigned integers of length one and four bytes. The unsigned integers
+ * are used for indexes or for, e.g., the count of objects to push in a "push"
+ * instruction.
+ */
+
+#define MAX_INSTRUCTION_OPERANDS 2
+
+typedef enum InstOperandType {
+    OPERAND_NONE,
+    OPERAND_INT1,		/* One byte signed integer. */
+    OPERAND_INT4,		/* Four byte signed integer. */
+    OPERAND_UINT1,		/* One byte unsigned integer. */
+    OPERAND_UINT4,		/* Four byte unsigned integer. */
+    OPERAND_IDX4,		/* Four byte signed index (actually an
+				 * integer, but displayed differently.) */
+    OPERAND_LVT1,		/* One byte unsigned index into the local
+				 * variable table. */
+    OPERAND_LVT4,		/* Four byte unsigned index into the local
+				 * variable table. */
+    OPERAND_AUX4,		/* Four byte unsigned index into the aux data
+				 * table. */
+    OPERAND_OFFSET1,		/* One byte signed jump offset. */
+    OPERAND_OFFSET4,		/* Four byte signed jump offset. */
+    OPERAND_LIT1,		/* One byte unsigned index into table of
+				 * literals. */
+    OPERAND_LIT4,		/* Four byte unsigned index into table of
+				 * literals. */
+    OPERAND_SCLS1		/* Index into tclStringClassTable. */
+} InstOperandType;
+
+typedef struct InstructionDesc {
+    const char *name;		/* Name of instruction. */
+    int numBytes;		/* Total number of bytes for instruction. */
+    int stackEffect;		/* The worst-case balance stack effect of the
+				 * instruction, used for stack requirements
+				 * computations. The value INT_MIN signals
+				 * that the instruction's worst case effect is
+				 * (1-opnd1). */
+    int numOperands;		/* Number of operands. */
+    InstOperandType opTypes[MAX_INSTRUCTION_OPERANDS];
+				/* The type of each operand. */
+} InstructionDesc;
+
+MODULE_SCOPE InstructionDesc const tclInstructionTable[];
+
+/*
+ * Constants used by INST_STRING_CLASS to indicate character classes. These
+ * correspond closely by name with what [string is] can support, but there is
+ * no requirement to keep the values the same.
+ */
+
+typedef enum InstStringClassType {
+    STR_CLASS_ALNUM,		/* Unicode alphabet or digit characters. */
+    STR_CLASS_ALPHA,		/* Unicode alphabet characters. */
+    STR_CLASS_ASCII,		/* Characters in range U+000000..U+00007F. */
+    STR_CLASS_CONTROL,		/* Unicode control characters. */
+    STR_CLASS_DIGIT,		/* Unicode digit characters. */
+    STR_CLASS_GRAPH,		/* Unicode printing characters, excluding
+				 * space. */
+    STR_CLASS_LOWER,		/* Unicode lower-case alphabet characters. */
+    STR_CLASS_PRINT,		/* Unicode printing characters, including
+				 * spaces. */
+    STR_CLASS_PUNCT,		/* Unicode punctuation characters. */
+    STR_CLASS_SPACE,		/* Unicode space characters. */
+    STR_CLASS_UPPER,		/* Unicode upper-case alphabet characters. */
+    STR_CLASS_WORD,		/* Unicode word (alphabetic, digit, connector
+				 * punctuation) characters. */
+    STR_CLASS_XDIGIT		/* Characters that can be used as digits in
+				 * hexadecimal numbers ([0-9A-Fa-f]). */
+} InstStringClassType;
+
+typedef struct StringClassDesc {
+    const char *name;		/* Name of the class. */
+    int (*comparator)(int);	/* Function to test if a single unicode
+				 * character is a member of the class. */
+} StringClassDesc;
+
+MODULE_SCOPE StringClassDesc const tclStringClassTable[];
+
+/*
+ * Compilation of some Tcl constructs such as if commands and the logical or
+ * (||) and logical and (&&) operators in expressions requires the generation
+ * of forward jumps. Since the PC target of these jumps isn't known when the
+ * jumps are emitted, we record the offset of each jump in an array of
+ * JumpFixup structures. There is one array for each sequence of jumps to one
+ * target PC. When we learn the target PC, we update the jumps with the
+ * correct distance. Also, if the distance is too great (> 127 bytes), we
+ * replace the single-byte jump with a four byte jump instruction, move the
+ * instructions after the jump down, and update the code offsets for any
+ * commands between the jump and the target.
+ */
+
+typedef enum {
+    TCL_UNCONDITIONAL_JUMP,
+    TCL_TRUE_JUMP,
+    TCL_FALSE_JUMP
+} TclJumpType;
+
+typedef struct JumpFixup {
+    TclJumpType jumpType;	/* Indicates the kind of jump. */
+    unsigned int codeOffset;	/* Offset of the first byte of the one-byte
+				 * forward jump's code. */
+    int cmdIndex;		/* Index of the first command after the one
+				 * for which the jump was emitted. Used to
+				 * update the code offsets for subsequent
+				 * commands if the two-byte jump at jumpPc
+				 * must be replaced with a five-byte one. */
+    int exceptIndex;		/* Index of the first range entry in the
+				 * ExceptionRange array after the current one.
+				 * This field is used to adjust the code
+				 * offsets in subsequent ExceptionRange
+				 * records when a jump is grown from 2 bytes
+				 * to 5 bytes. */
+} JumpFixup;
+
+#define JUMPFIXUP_INIT_ENTRIES	10
+
+typedef struct JumpFixupArray {
+    JumpFixup *fixup;		/* Points to start of jump fixup array. */
+    int next;			/* Index of next free array entry. */
+    int end;			/* Index of last usable entry in array. */
+    int mallocedArray;		/* 1 if array was expanded and fixups points
+				 * into the heap, else 0. */
+    JumpFixup staticFixupSpace[JUMPFIXUP_INIT_ENTRIES];
+				/* Initial storage for jump fixup array. */
+} JumpFixupArray;
+
+/*
+ * The structure describing one variable list of a foreach command. Note that
+ * only foreach commands inside procedure bodies are compiled inline so a
+ * ForeachVarList structure always describes local variables. Furthermore,
+ * only scalar variables are supported for inline-compiled foreach loops.
+ */
+
+typedef struct ForeachVarList {
+    int numVars;		/* The number of variables in the list. */
+    int varIndexes[1];		/* An array of the indexes ("slot numbers")
+				 * for each variable in the procedure's array
+				 * of local variables. Only scalar variables
+				 * are supported. The actual size of this
+				 * field will be large enough to numVars
+				 * indexes. THIS MUST BE THE LAST FIELD IN THE
+				 * STRUCTURE! */
+} ForeachVarList;
+
+/*
+ * Structure used to hold information about a foreach command that is needed
+ * during program execution. These structures are stored in CompileEnv and
+ * ByteCode structures as auxiliary data.
+ */
+
+typedef struct ForeachInfo {
+    int numLists;		/* The number of both the variable and value
+				 * lists of the foreach command. */
+    int firstValueTemp;		/* Index of the first temp var in a proc frame
+				 * used to point to a value list. */
+    int loopCtTemp;		/* Index of temp var in a proc frame holding
+				 * the loop's iteration count. Used to
+				 * determine next value list element to assign
+				 * each loop var. */
+    ForeachVarList *varLists[1];/* An array of pointers to ForeachVarList
+				 * structures describing each var list. The
+				 * actual size of this field will be large
+				 * enough to numVars indexes. THIS MUST BE THE
+				 * LAST FIELD IN THE STRUCTURE! */
+} ForeachInfo;
+
+/*
+ * Structure used to hold information about a switch command that is needed
+ * during program execution. These structures are stored in CompileEnv and
+ * ByteCode structures as auxiliary data.
+ */
+
+typedef struct JumptableInfo {
+    Tcl_HashTable hashTable;	/* Hash that maps strings to signed ints (PC
+				 * offsets). */
+} JumptableInfo;
+
+MODULE_SCOPE const AuxDataType tclJumptableInfoType;
+
+#define JUMPTABLEINFO(envPtr, index) \
+    ((JumptableInfo*)((envPtr)->auxDataArrayPtr[TclGetUInt4AtPtr(index)].clientData))
+
+/*
+ * Structure used to hold information about a [dict update] command that is
+ * needed during program execution. These structures are stored in CompileEnv
+ * and ByteCode structures as auxiliary data.
+ */
+
+typedef struct {
+    int length;			/* Size of array */
+    int varIndices[1];		/* Array of variable indices to manage when
+				 * processing the start and end of a [dict
+				 * update]. There is really more than one
+				 * entry, and the structure is allocated to
+				 * take account of this. MUST BE LAST FIELD IN
+				 * STRUCTURE. */
+} DictUpdateInfo;
+
+/*
+ * ClientData type used by the math operator commands.
+ */
+
+typedef struct {
+    const char *op;		/* Do not call it 'operator': C++ reserved */
+    const char *expected;
+    union {
+	int numArgs;
+	int identity;
+    } i;
+} TclOpCmdClientData;
+
+/*
+ *----------------------------------------------------------------
+ * Procedures exported by tclBasic.c to be used within the engine.
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE Tcl_ObjCmdProc	TclNRInterpCoroutine;
+
+/*
+ *----------------------------------------------------------------
+ * Procedures exported by the engine to be used by tclBasic.c
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE ByteCode *	TclCompileObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    const CmdFrame *invoker, int word);
+
+/*
+ *----------------------------------------------------------------
+ * Procedures shared among Tcl bytecode compilation and execution modules but
+ * not used outside:
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE int	TclAttemptCompileProc(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, int depth, Command *cmdPtr,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCleanupStackForBreakContinue(CompileEnv *envPtr,
+			    ExceptionAux *auxPtr);
+MODULE_SCOPE void	TclCompileCmdWord(Tcl_Interp *interp,
+			    Tcl_Token *tokenPtr, int count,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileExpr(Tcl_Interp *interp, const char *script,
+			    int numBytes, CompileEnv *envPtr, int optimize);
+MODULE_SCOPE void	TclCompileExprWords(Tcl_Interp *interp,
+			    Tcl_Token *tokenPtr, int numWords,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileInvocation(Tcl_Interp *interp,
+			    Tcl_Token *tokenPtr, Tcl_Obj *cmdObj, int numWords,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileScript(Tcl_Interp *interp,
+			    const char *script, int numBytes,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileSyntaxError(Tcl_Interp *interp,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileTokens(Tcl_Interp *interp,
+			    Tcl_Token *tokenPtr, int count,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclCompileVarSubst(Tcl_Interp *interp,
+			    Tcl_Token *tokenPtr, CompileEnv *envPtr);
+MODULE_SCOPE int	TclCreateAuxData(ClientData clientData,
+			    const AuxDataType *typePtr, CompileEnv *envPtr);
+MODULE_SCOPE int	TclCreateExceptRange(ExceptionRangeType type,
+			    CompileEnv *envPtr);
+MODULE_SCOPE ExecEnv *	TclCreateExecEnv(Tcl_Interp *interp, int size);
+MODULE_SCOPE Tcl_Obj *	TclCreateLiteral(Interp *iPtr, const char *bytes,
+			    int length, unsigned int hash, int *newPtr,
+			    Namespace *nsPtr, int flags,
+			    LiteralEntry **globalPtrPtr);
+MODULE_SCOPE void	TclDeleteExecEnv(ExecEnv *eePtr);
+MODULE_SCOPE void	TclDeleteLiteralTable(Tcl_Interp *interp,
+			    LiteralTable *tablePtr);
+MODULE_SCOPE void	TclEmitForwardJump(CompileEnv *envPtr,
+			    TclJumpType jumpType, JumpFixup *jumpFixupPtr);
+MODULE_SCOPE void	TclEmitInvoke(CompileEnv *envPtr, int opcode, ...);
+MODULE_SCOPE ExceptionRange * TclGetExceptionRangeForPc(unsigned char *pc,
+			    int catchOnly, ByteCode *codePtr);
+MODULE_SCOPE void	TclExpandJumpFixupArray(JumpFixupArray *fixupArrayPtr);
+MODULE_SCOPE int	TclNRExecuteByteCode(Tcl_Interp *interp,
+			    ByteCode *codePtr);
+MODULE_SCOPE Tcl_Obj *	TclFetchLiteral(CompileEnv *envPtr, unsigned int index);
+MODULE_SCOPE int	TclFindCompiledLocal(const char *name, int nameChars,
+			    int create, CompileEnv *envPtr);
+MODULE_SCOPE int	TclFixupForwardJump(CompileEnv *envPtr,
+			    JumpFixup *jumpFixupPtr, int jumpDist,
+			    int distThreshold);
+MODULE_SCOPE void	TclFreeCompileEnv(CompileEnv *envPtr);
+MODULE_SCOPE void	TclFreeJumpFixupArray(JumpFixupArray *fixupArrayPtr);
+MODULE_SCOPE ByteCode *	TclInitByteCode(CompileEnv *envPtr);
+MODULE_SCOPE ByteCode *	TclInitByteCodeObj(Tcl_Obj *objPtr,
+			    const Tcl_ObjType *typePtr, CompileEnv *envPtr);
+MODULE_SCOPE void	TclInitCompileEnv(Tcl_Interp *interp,
+			    CompileEnv *envPtr, const char *string,
+			    int numBytes, const CmdFrame *invoker, int word);
+MODULE_SCOPE void	TclInitJumpFixupArray(JumpFixupArray *fixupArrayPtr);
+MODULE_SCOPE void	TclInitLiteralTable(LiteralTable *tablePtr);
+MODULE_SCOPE ExceptionRange *TclGetInnermostExceptionRange(CompileEnv *envPtr,
+			    int returnCode, ExceptionAux **auxPtrPtr);
+MODULE_SCOPE void	TclAddLoopBreakFixup(CompileEnv *envPtr,
+			    ExceptionAux *auxPtr);
+MODULE_SCOPE void	TclAddLoopContinueFixup(CompileEnv *envPtr,
+			    ExceptionAux *auxPtr);
+MODULE_SCOPE void	TclFinalizeLoopExceptionRange(CompileEnv *envPtr,
+			    int range);
+#ifdef TCL_COMPILE_STATS
+MODULE_SCOPE char *	TclLiteralStats(LiteralTable *tablePtr);
+MODULE_SCOPE int	TclLog2(int value);
+#endif
+MODULE_SCOPE int	TclLocalScalar(const char *bytes, int numBytes,
+			    CompileEnv *envPtr);
+MODULE_SCOPE int	TclLocalScalarFromToken(Tcl_Token *tokenPtr,
+			    CompileEnv *envPtr);
+MODULE_SCOPE void	TclOptimizeBytecode(void *envPtr);
+#ifdef TCL_COMPILE_DEBUG
+MODULE_SCOPE void	TclPrintByteCodeObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+#endif
+MODULE_SCOPE int	TclPrintInstruction(ByteCode *codePtr,
+			    const unsigned char *pc);
+MODULE_SCOPE void	TclPrintObject(FILE *outFile,
+			    Tcl_Obj *objPtr, int maxChars);
+MODULE_SCOPE void	TclPrintSource(FILE *outFile,
+			    const char *string, int maxChars);
+MODULE_SCOPE void	TclPushVarName(Tcl_Interp *interp,
+			    Tcl_Token *varTokenPtr, CompileEnv *envPtr,
+			    int flags, int *localIndexPtr,
+			    int *isScalarPtr);
+MODULE_SCOPE void	TclPreserveByteCode(ByteCode *codePtr);
+MODULE_SCOPE void	TclReleaseByteCode(ByteCode *codePtr);
+MODULE_SCOPE void	TclReleaseLiteral(Tcl_Interp *interp, Tcl_Obj *objPtr);
+MODULE_SCOPE void	TclInvalidateCmdLiteral(Tcl_Interp *interp,
+			    const char *name, Namespace *nsPtr);
+MODULE_SCOPE int	TclSingleOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclSortingOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclVariadicOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclNoIdentOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#ifdef TCL_COMPILE_DEBUG
+MODULE_SCOPE void	TclVerifyGlobalLiteralTable(Interp *iPtr);
+MODULE_SCOPE void	TclVerifyLocalLiteralTable(CompileEnv *envPtr);
+#endif
+MODULE_SCOPE int	TclWordKnownAtCompileTime(Tcl_Token *tokenPtr,
+			    Tcl_Obj *valuePtr);
+MODULE_SCOPE void	TclLogCommandInfo(Tcl_Interp *interp,
+			    const char *script, const char *command,
+			    int length, const unsigned char *pc,
+			    Tcl_Obj **tosPtr);
+MODULE_SCOPE Tcl_Obj	*TclGetInnerContext(Tcl_Interp *interp,
+			    const unsigned char *pc, Tcl_Obj **tosPtr);
+MODULE_SCOPE Tcl_Obj	*TclNewInstNameObj(unsigned char inst);
+MODULE_SCOPE int	TclPushProcCallFrame(ClientData clientData,
+			    register Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[], int isLambda);
+
+
+/*
+ *----------------------------------------------------------------
+ * Macros and flag values used by Tcl bytecode compilation and execution
+ * modules inside the Tcl core but not used outside.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * Simplified form to access AuxData.
+ *
+ * ClientData TclFetchAuxData(CompileEng *envPtr, int index);
+ */
+
+#define TclFetchAuxData(envPtr, index) \
+    (envPtr)->auxDataArrayPtr[(index)].clientData
+
+#define LITERAL_ON_HEAP		0x01
+#define LITERAL_CMD_NAME	0x02
+#define LITERAL_UNSHARED	0x04
+
+/*
+ * Macro used to manually adjust the stack requirements; used in cases where
+ * the stack effect cannot be computed from the opcode and its operands, but
+ * is still known at compile time.
+ *
+ * void TclAdjustStackDepth(int delta, CompileEnv *envPtr);
+ */
+
+#define TclAdjustStackDepth(delta, envPtr) \
+    do {								\
+	if ((delta) < 0) {						\
+	    if ((envPtr)->maxStackDepth < (envPtr)->currStackDepth) {	\
+		(envPtr)->maxStackDepth = (envPtr)->currStackDepth;	\
+	    }								\
+	}								\
+	(envPtr)->currStackDepth += (delta);				\
+    } while (0)
+
+#define TclGetStackDepth(envPtr)		\
+    ((envPtr)->currStackDepth)
+
+#define TclSetStackDepth(depth, envPtr)		\
+    (envPtr)->currStackDepth = (depth)
+
+#define TclCheckStackDepth(depth, envPtr)				\
+    do {								\
+	int _dd = (depth);						\
+	if (_dd != (envPtr)->currStackDepth) {				\
+	    Tcl_Panic("bad stack depth computations: is %i, should be %i", \
+		    (envPtr)->currStackDepth, _dd);		\
+	}								\
+    } while (0)
+
+/*
+ * Macro used to update the stack requirements. It is called by the macros
+ * TclEmitOpCode, TclEmitInst1 and TclEmitInst4.
+ * Remark that the very last instruction of a bytecode always reduces the
+ * stack level: INST_DONE or INST_POP, so that the maxStackdepth is always
+ * updated.
+ *
+ * void TclUpdateStackReqs(unsigned char op, int i, CompileEnv *envPtr);
+ */
+
+#define TclUpdateStackReqs(op, i, envPtr) \
+    do {							\
+	int _delta = tclInstructionTable[(op)].stackEffect;	\
+	if (_delta) {						\
+	    if (_delta == INT_MIN) {				\
+		_delta = 1 - (i);				\
+	    }							\
+	    TclAdjustStackDepth(_delta, envPtr);			\
+	}							\
+    } while (0)
+
+/*
+ * Macros used to update the flag that indicates if we are at the start of a
+ * command, based on whether the opcode is INST_START_COMMAND.
+ *
+ * void TclUpdateAtCmdStart(unsigned char op, CompileEnv *envPtr);
+ */
+
+#define TclUpdateAtCmdStart(op, envPtr) \
+    if ((envPtr)->atCmdStart < 2) {				     \
+	(envPtr)->atCmdStart = ((op) == INST_START_CMD ? 1 : 0);     \
+    }
+
+/*
+ * Macro to emit an opcode byte into a CompileEnv's code array. The ANSI C
+ * "prototype" for this macro is:
+ *
+ * void TclEmitOpcode(unsigned char op, CompileEnv *envPtr);
+ */
+
+#define TclEmitOpcode(op, envPtr) \
+    do {							\
+	if ((envPtr)->codeNext == (envPtr)->codeEnd) {		\
+	    TclExpandCodeArray(envPtr);				\
+	}							\
+	*(envPtr)->codeNext++ = (unsigned char) (op);		\
+	TclUpdateAtCmdStart(op, envPtr);			\
+	TclUpdateStackReqs(op, 0, envPtr);			\
+    } while (0)
+
+/*
+ * Macros to emit an integer operand. The ANSI C "prototype" for these macros
+ * are:
+ *
+ * void TclEmitInt1(int i, CompileEnv *envPtr);
+ * void TclEmitInt4(int i, CompileEnv *envPtr);
+ */
+
+#define TclEmitInt1(i, envPtr) \
+    do {								\
+	if ((envPtr)->codeNext == (envPtr)->codeEnd) {			\
+	    TclExpandCodeArray(envPtr);					\
+	}								\
+	*(envPtr)->codeNext++ = (unsigned char) ((unsigned int) (i));	\
+    } while (0)
+
+#define TclEmitInt4(i, envPtr) \
+    do {								\
+	if (((envPtr)->codeNext + 4) > (envPtr)->codeEnd) {		\
+	    TclExpandCodeArray(envPtr);					\
+	}								\
+	*(envPtr)->codeNext++ =						\
+		(unsigned char) ((unsigned int) (i) >> 24);		\
+	*(envPtr)->codeNext++ =						\
+		(unsigned char) ((unsigned int) (i) >> 16);		\
+	*(envPtr)->codeNext++ =						\
+		(unsigned char) ((unsigned int) (i) >>  8);		\
+	*(envPtr)->codeNext++ =						\
+		(unsigned char) ((unsigned int) (i)      );		\
+    } while (0)
+
+/*
+ * Macros to emit an instruction with signed or unsigned integer operands.
+ * Four byte integers are stored in "big-endian" order with the high order
+ * byte stored at the lowest address. The ANSI C "prototypes" for these macros
+ * are:
+ *
+ * void TclEmitInstInt1(unsigned char op, int i, CompileEnv *envPtr);
+ * void TclEmitInstInt4(unsigned char op, int i, CompileEnv *envPtr);
+ */
+
+#define TclEmitInstInt1(op, i, envPtr) \
+    do {								\
+	if (((envPtr)->codeNext + 2) > (envPtr)->codeEnd) {		\
+	    TclExpandCodeArray(envPtr);					\
+	}								\
+	*(envPtr)->codeNext++ = (unsigned char) (op);			\
+	*(envPtr)->codeNext++ = (unsigned char) ((unsigned int) (i));	\
+	TclUpdateAtCmdStart(op, envPtr);				\
+	TclUpdateStackReqs(op, i, envPtr);				\
+    } while (0)
+
+#define TclEmitInstInt4(op, i, envPtr) \
+    do {							\
+	if (((envPtr)->codeNext + 5) > (envPtr)->codeEnd) {	\
+	    TclExpandCodeArray(envPtr);				\
+	}							\
+	*(envPtr)->codeNext++ = (unsigned char) (op);		\
+	*(envPtr)->codeNext++ =					\
+		(unsigned char) ((unsigned int) (i) >> 24);	\
+	*(envPtr)->codeNext++ =					\
+		(unsigned char) ((unsigned int) (i) >> 16);	\
+	*(envPtr)->codeNext++ =					\
+		(unsigned char) ((unsigned int) (i) >>  8);	\
+	*(envPtr)->codeNext++ =					\
+		(unsigned char) ((unsigned int) (i)      );	\
+	TclUpdateAtCmdStart(op, envPtr);			\
+	TclUpdateStackReqs(op, i, envPtr);			\
+    } while (0)
+
+/*
+ * Macro to push a Tcl object onto the Tcl evaluation stack. It emits the
+ * object's one or four byte array index into the CompileEnv's code array.
+ * These support, respectively, a maximum of 256 (2**8) and 2**32 objects in a
+ * CompileEnv. The ANSI C "prototype" for this macro is:
+ *
+ * void	TclEmitPush(int objIndex, CompileEnv *envPtr);
+ */
+
+#define TclEmitPush(objIndex, envPtr) \
+    do {							 \
+	register int _objIndexCopy = (objIndex);			 \
+	if (_objIndexCopy <= 255) {				 \
+	    TclEmitInstInt1(INST_PUSH1, _objIndexCopy, (envPtr)); \
+	} else {						 \
+	    TclEmitInstInt4(INST_PUSH4, _objIndexCopy, (envPtr)); \
+	}							 \
+    } while (0)
+
+/*
+ * Macros to update a (signed or unsigned) integer starting at a pointer. The
+ * two variants depend on the number of bytes. The ANSI C "prototypes" for
+ * these macros are:
+ *
+ * void TclStoreInt1AtPtr(int i, unsigned char *p);
+ * void TclStoreInt4AtPtr(int i, unsigned char *p);
+ */
+
+#define TclStoreInt1AtPtr(i, p) \
+    *(p)   = (unsigned char) ((unsigned int) (i))
+
+#define TclStoreInt4AtPtr(i, p) \
+    do {							\
+	*(p)   = (unsigned char) ((unsigned int) (i) >> 24);	\
+	*(p+1) = (unsigned char) ((unsigned int) (i) >> 16);	\
+	*(p+2) = (unsigned char) ((unsigned int) (i) >>  8);	\
+	*(p+3) = (unsigned char) ((unsigned int) (i)      );	\
+    } while (0)
+
+/*
+ * Macros to update instructions at a particular pc with a new op code and a
+ * (signed or unsigned) int operand. The ANSI C "prototypes" for these macros
+ * are:
+ *
+ * void TclUpdateInstInt1AtPc(unsigned char op, int i, unsigned char *pc);
+ * void TclUpdateInstInt4AtPc(unsigned char op, int i, unsigned char *pc);
+ */
+
+#define TclUpdateInstInt1AtPc(op, i, pc) \
+    do {					\
+	*(pc) = (unsigned char) (op);		\
+	TclStoreInt1AtPtr((i), ((pc)+1));	\
+    } while (0)
+
+#define TclUpdateInstInt4AtPc(op, i, pc) \
+    do {					\
+	*(pc) = (unsigned char) (op);		\
+	TclStoreInt4AtPtr((i), ((pc)+1));	\
+    } while (0)
+
+/*
+ * Macro to fix up a forward jump to point to the current code-generation
+ * position in the bytecode being created (the most common case). The ANSI C
+ * "prototypes" for this macro is:
+ *
+ * int TclFixupForwardJumpToHere(CompileEnv *envPtr, JumpFixup *fixupPtr,
+ *				 int threshold);
+ */
+
+#define TclFixupForwardJumpToHere(envPtr, fixupPtr, threshold) \
+    TclFixupForwardJump((envPtr), (fixupPtr),				\
+	    (envPtr)->codeNext-(envPtr)->codeStart-(fixupPtr)->codeOffset, \
+	    (threshold))
+
+/*
+ * Macros to get a signed integer (GET_INT{1,2}) or an unsigned int
+ * (GET_UINT{1,2}) from a pointer. There are two variants for each return type
+ * that depend on the number of bytes fetched. The ANSI C "prototypes" for
+ * these macros are:
+ *
+ * int TclGetInt1AtPtr(unsigned char *p);
+ * int TclGetInt4AtPtr(unsigned char *p);
+ * unsigned int TclGetUInt1AtPtr(unsigned char *p);
+ * unsigned int TclGetUInt4AtPtr(unsigned char *p);
+ */
+
+/*
+ * The TclGetInt1AtPtr macro is tricky because we want to do sign extension on
+ * the 1-byte value. Unfortunately the "char" type isn't signed on all
+ * platforms so sign-extension doesn't always happen automatically. Sometimes
+ * we can explicitly declare the pointer to be signed, but other times we have
+ * to explicitly sign-extend the value in software.
+ */
+
+#ifndef __CHAR_UNSIGNED__
+#   define TclGetInt1AtPtr(p) ((int) *((char *) p))
+#elif defined(HAVE_SIGNED_CHAR)
+#   define TclGetInt1AtPtr(p) ((int) *((signed char *) p))
+#else
+#   define TclGetInt1AtPtr(p) \
+    (((int) *((char *) p)) | ((*(p) & 0200) ? (-256) : 0))
+#endif
+
+#define TclGetInt4AtPtr(p) \
+    (((int) TclGetInt1AtPtr(p) << 24) |				\
+		     (*((p)+1) << 16) |				\
+		     (*((p)+2) <<  8) |				\
+		     (*((p)+3)))
+
+#define TclGetUInt1AtPtr(p) \
+    ((unsigned int) *(p))
+#define TclGetUInt4AtPtr(p) \
+    ((unsigned int) (*(p)     << 24) |				\
+		    (*((p)+1) << 16) |				\
+		    (*((p)+2) <<  8) |				\
+		    (*((p)+3)))
+
+/*
+ * Macros used to compute the minimum and maximum of two integers. The ANSI C
+ * "prototypes" for these macros are:
+ *
+ * int TclMin(int i, int j);
+ * int TclMax(int i, int j);
+ */
+
+#define TclMin(i, j)	((((int) i) < ((int) j))? (i) : (j))
+#define TclMax(i, j)	((((int) i) > ((int) j))? (i) : (j))
+
+/*
+ * Convenience macros for use when compiling bodies of commands. The ANSI C
+ * "prototype" for these macros are:
+ *
+ * static void		BODY(Tcl_Token *tokenPtr, int word);
+ */
+
+#define BODY(tokenPtr, word)						\
+    SetLineInformation((word));						\
+    TclCompileCmdWord(interp, (tokenPtr)+1, (tokenPtr)->numComponents,	\
+	    envPtr)
+
+/*
+ * Convenience macro for use when compiling tokens to be pushed. The ANSI C
+ * "prototype" for this macro is:
+ *
+ * static void		CompileTokens(CompileEnv *envPtr, Tcl_Token *tokenPtr,
+ *			    Tcl_Interp *interp);
+ */
+
+#define CompileTokens(envPtr, tokenPtr, interp) \
+    TclCompileTokens((interp), (tokenPtr)+1, (tokenPtr)->numComponents, \
+	    (envPtr));
+/*
+ * Convenience macros for use when pushing literals. The ANSI C "prototype" for
+ * these macros are:
+ *
+ * static void		PushLiteral(CompileEnv *envPtr,
+ *			    const char *string, int length);
+ * static void		PushStringLiteral(CompileEnv *envPtr,
+ *			    const char *string);
+ */
+
+#define PushLiteral(envPtr, string, length) \
+    TclEmitPush(TclRegisterLiteral(envPtr, string, length, 0), (envPtr))
+#define PushStringLiteral(envPtr, string) \
+    PushLiteral(envPtr, string, (int) (sizeof(string "") - 1))
+
+/*
+ * Macro to advance to the next token; it is more mnemonic than the address
+ * arithmetic that it replaces. The ANSI C "prototype" for this macro is:
+ *
+ * static Tcl_Token *	TokenAfter(Tcl_Token *tokenPtr);
+ */
+
+#define TokenAfter(tokenPtr) \
+    ((tokenPtr) + ((tokenPtr)->numComponents + 1))
+
+/*
+ * Macro to get the offset to the next instruction to be issued. The ANSI C
+ * "prototype" for this macro is:
+ *
+ * static int	CurrentOffset(CompileEnv *envPtr);
+ */
+
+#define CurrentOffset(envPtr) \
+    ((envPtr)->codeNext - (envPtr)->codeStart)
+
+/*
+ * Note: the exceptDepth is a bit of a misnomer: TEBC only needs the
+ * maximal depth of nested CATCH ranges in order to alloc runtime
+ * memory. These macros should compute precisely that? OTOH, the nesting depth
+ * of LOOP ranges is an interesting datum for debugging purposes, and that is
+ * what we compute now.
+ *
+ * static int	ExceptionRangeStarts(CompileEnv *envPtr, int index);
+ * static void	ExceptionRangeEnds(CompileEnv *envPtr, int index);
+ * static void	ExceptionRangeTarget(CompileEnv *envPtr, int index, LABEL);
+ */
+
+#define ExceptionRangeStarts(envPtr, index) \
+    (((envPtr)->exceptDepth++),						\
+    ((envPtr)->maxExceptDepth =						\
+	    TclMax((envPtr)->exceptDepth, (envPtr)->maxExceptDepth)),	\
+    ((envPtr)->exceptArrayPtr[(index)].codeOffset = CurrentOffset(envPtr)))
+#define ExceptionRangeEnds(envPtr, index) \
+    (((envPtr)->exceptDepth--),						\
+    ((envPtr)->exceptArrayPtr[(index)].numCodeBytes =			\
+	CurrentOffset(envPtr) - (envPtr)->exceptArrayPtr[(index)].codeOffset))
+#define ExceptionRangeTarget(envPtr, index, targetType) \
+    ((envPtr)->exceptArrayPtr[(index)].targetType = CurrentOffset(envPtr))
+
+/*
+ * Check if there is an LVT for compiled locals
+ */
+
+#define EnvHasLVT(envPtr) \
+    (envPtr->procPtr || envPtr->iPtr->varFramePtr->localCachePtr)
+
+/*
+ * Macros for making it easier to deal with tokens and DStrings.
+ */
+
+#define TclDStringAppendToken(dsPtr, tokenPtr) \
+    Tcl_DStringAppend((dsPtr), (tokenPtr)->start, (tokenPtr)->size)
+#define TclRegisterDStringLiteral(envPtr, dsPtr) \
+    TclRegisterLiteral(envPtr, Tcl_DStringValue(dsPtr), \
+	    Tcl_DStringLength(dsPtr), /*flags*/ 0)
+
+/*
+ * Macro that encapsulates an efficiency trick that avoids a function call for
+ * the simplest of compiles. The ANSI C "prototype" for this macro is:
+ *
+ * static void		CompileWord(CompileEnv *envPtr, Tcl_Token *tokenPtr,
+ *			    Tcl_Interp *interp, int word);
+ */
+
+#define CompileWord(envPtr, tokenPtr, interp, word) \
+    if ((tokenPtr)->type == TCL_TOKEN_SIMPLE_WORD) {			\
+	PushLiteral((envPtr), (tokenPtr)[1].start, (tokenPtr)[1].size);	\
+    } else {								\
+	SetLineInformation((word));					\
+	CompileTokens((envPtr), (tokenPtr), (interp));			\
+    }
+
+/*
+ * TIP #280: Remember the per-word line information of the current command. An
+ * index is used instead of a pointer as recursive compilation may reallocate,
+ * i.e. move, the array. This is also the reason to save the nuloc now, it may
+ * change during the course of the function.
+ *
+ * Macro to encapsulate the variable definition and setup.
+ */
+
+#define DefineLineInformation \
+    ExtCmdLoc *mapPtr = envPtr->extCmdMapPtr;				\
+    int eclIndex = mapPtr->nuloc - 1
+
+#define SetLineInformation(word) \
+    envPtr->line = mapPtr->loc[eclIndex].line[(word)];			\
+    envPtr->clNext = mapPtr->loc[eclIndex].next[(word)]
+
+#define PushVarNameWord(i,v,e,f,l,sc,word) \
+    SetLineInformation(word);						\
+    TclPushVarName(i,v,e,f,l,sc)
+
+/*
+ * Often want to issue one of two versions of an instruction based on whether
+ * the argument will fit in a single byte or not. This makes it much clearer.
+ */
+
+#define Emit14Inst(nm,idx,envPtr) \
+    if (idx <= 255) {							\
+	TclEmitInstInt1(nm##1,idx,envPtr);				\
+    } else {								\
+	TclEmitInstInt4(nm##4,idx,envPtr);				\
+    }
+
+/*
+ * How to get an anonymous local variable (used for holding temporary values
+ * off the stack) or a local simple scalar.
+ */
+
+#define AnonymousLocal(envPtr) \
+    (TclFindCompiledLocal(NULL, /*nameChars*/ 0, /*create*/ 1, (envPtr)))
+#define LocalScalar(chars,len,envPtr) \
+    TclLocalScalar(chars, len, envPtr)
+#define LocalScalarFromToken(tokenPtr,envPtr) \
+    TclLocalScalarFromToken(tokenPtr, envPtr)
+
+/*
+ * Flags bits used by TclPushVarName.
+ */
+
+#define TCL_NO_LARGE_INDEX 1	/* Do not return localIndex value > 255 */
+#define TCL_NO_ELEMENT 2	/* Do not push the array element. */
+
+/*
+ * DTrace probe macros (NOPs if DTrace support is not enabled).
+ */
+
+/*
+ * Define the following macros to enable debug logging of the DTrace proc,
+ * cmd, and inst probes. Note that this does _not_ require a platform with
+ * DTrace, it simply logs all probe output to /tmp/tclDTraceDebug-[pid].log.
+ *
+ * If the second macro is defined, logging to file starts immediately,
+ * otherwise only after the first call to [tcl::dtrace]. Note that the debug
+ * probe data is always computed, even when it is not logged to file.
+ *
+ * Defining the third macro enables debug logging of inst probes (disabled
+ * by default due to the significant performance impact).
+ */
+
+/*
+#define TCL_DTRACE_DEBUG 1
+#define TCL_DTRACE_DEBUG_LOG_ENABLED 1
+#define TCL_DTRACE_DEBUG_INST_PROBES 1
+*/
+
+#if !(defined(TCL_DTRACE_DEBUG) && defined(__GNUC__))
+
+#ifdef USE_DTRACE
+
+#if defined(__GNUC__) && __GNUC__ > 2
+/*
+ * Use gcc branch prediction hint to minimize cost of DTrace ENABLED checks.
+ */
+#define unlikely(x) (__builtin_expect((x), 0))
+#else
+#define unlikely(x) (x)
+#endif
+
+#define TCL_DTRACE_PROC_ENTRY_ENABLED()	    unlikely(TCL_PROC_ENTRY_ENABLED())
+#define TCL_DTRACE_PROC_RETURN_ENABLED()    unlikely(TCL_PROC_RETURN_ENABLED())
+#define TCL_DTRACE_PROC_RESULT_ENABLED()    unlikely(TCL_PROC_RESULT_ENABLED())
+#define TCL_DTRACE_PROC_ARGS_ENABLED()	    unlikely(TCL_PROC_ARGS_ENABLED())
+#define TCL_DTRACE_PROC_INFO_ENABLED()	    unlikely(TCL_PROC_INFO_ENABLED())
+#define TCL_DTRACE_PROC_ENTRY(a0, a1, a2)   TCL_PROC_ENTRY(a0, a1, a2)
+#define TCL_DTRACE_PROC_RETURN(a0, a1)	    TCL_PROC_RETURN(a0, a1)
+#define TCL_DTRACE_PROC_RESULT(a0, a1, a2, a3) TCL_PROC_RESULT(a0, a1, a2, a3)
+#define TCL_DTRACE_PROC_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+	TCL_PROC_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9)
+#define TCL_DTRACE_PROC_INFO(a0, a1, a2, a3, a4, a5, a6, a7) \
+	TCL_PROC_INFO(a0, a1, a2, a3, a4, a5, a6, a7)
+
+#define TCL_DTRACE_CMD_ENTRY_ENABLED()	    unlikely(TCL_CMD_ENTRY_ENABLED())
+#define TCL_DTRACE_CMD_RETURN_ENABLED()	    unlikely(TCL_CMD_RETURN_ENABLED())
+#define TCL_DTRACE_CMD_RESULT_ENABLED()	    unlikely(TCL_CMD_RESULT_ENABLED())
+#define TCL_DTRACE_CMD_ARGS_ENABLED()	    unlikely(TCL_CMD_ARGS_ENABLED())
+#define TCL_DTRACE_CMD_INFO_ENABLED()	    unlikely(TCL_CMD_INFO_ENABLED())
+#define TCL_DTRACE_CMD_ENTRY(a0, a1, a2)    TCL_CMD_ENTRY(a0, a1, a2)
+#define TCL_DTRACE_CMD_RETURN(a0, a1)	    TCL_CMD_RETURN(a0, a1)
+#define TCL_DTRACE_CMD_RESULT(a0, a1, a2, a3) TCL_CMD_RESULT(a0, a1, a2, a3)
+#define TCL_DTRACE_CMD_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+	TCL_CMD_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9)
+#define TCL_DTRACE_CMD_INFO(a0, a1, a2, a3, a4, a5, a6, a7) \
+	TCL_CMD_INFO(a0, a1, a2, a3, a4, a5, a6, a7)
+
+#define TCL_DTRACE_INST_START_ENABLED()	    unlikely(TCL_INST_START_ENABLED())
+#define TCL_DTRACE_INST_DONE_ENABLED()	    unlikely(TCL_INST_DONE_ENABLED())
+#define TCL_DTRACE_INST_START(a0, a1, a2)   TCL_INST_START(a0, a1, a2)
+#define TCL_DTRACE_INST_DONE(a0, a1, a2)    TCL_INST_DONE(a0, a1, a2)
+
+#define TCL_DTRACE_TCL_PROBE_ENABLED()	    unlikely(TCL_TCL_PROBE_ENABLED())
+#define TCL_DTRACE_TCL_PROBE(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+	TCL_TCL_PROBE(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9)
+
+#define TCL_DTRACE_DEBUG_LOG()
+
+MODULE_SCOPE void	TclDTraceInfo(Tcl_Obj *info, const char **args,
+			    int *argsi);
+
+#else /* USE_DTRACE */
+
+#define TCL_DTRACE_PROC_ENTRY_ENABLED()	    0
+#define TCL_DTRACE_PROC_RETURN_ENABLED()    0
+#define TCL_DTRACE_PROC_RESULT_ENABLED()    0
+#define TCL_DTRACE_PROC_ARGS_ENABLED()	    0
+#define TCL_DTRACE_PROC_INFO_ENABLED()	    0
+#define TCL_DTRACE_PROC_ENTRY(a0, a1, a2)   {if (a0) {}}
+#define TCL_DTRACE_PROC_RETURN(a0, a1)	    {if (a0) {}}
+#define TCL_DTRACE_PROC_RESULT(a0, a1, a2, a3) {if (a0) {}; if (a3) {}}
+#define TCL_DTRACE_PROC_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) {}
+#define TCL_DTRACE_PROC_INFO(a0, a1, a2, a3, a4, a5, a6, a7) {}
+
+#define TCL_DTRACE_CMD_ENTRY_ENABLED()	    0
+#define TCL_DTRACE_CMD_RETURN_ENABLED()	    0
+#define TCL_DTRACE_CMD_RESULT_ENABLED()	    0
+#define TCL_DTRACE_CMD_ARGS_ENABLED()	    0
+#define TCL_DTRACE_CMD_INFO_ENABLED()	    0
+#define TCL_DTRACE_CMD_ENTRY(a0, a1, a2)    {}
+#define TCL_DTRACE_CMD_RETURN(a0, a1)	    {}
+#define TCL_DTRACE_CMD_RESULT(a0, a1, a2, a3) {}
+#define TCL_DTRACE_CMD_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) {}
+#define TCL_DTRACE_CMD_INFO(a0, a1, a2, a3, a4, a5, a6, a7) {}
+
+#define TCL_DTRACE_INST_START_ENABLED()	    0
+#define TCL_DTRACE_INST_DONE_ENABLED()	    0
+#define TCL_DTRACE_INST_START(a0, a1, a2)   {}
+#define TCL_DTRACE_INST_DONE(a0, a1, a2)    {}
+
+#define TCL_DTRACE_TCL_PROBE_ENABLED()	    0
+#define TCL_DTRACE_TCL_PROBE(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) {}
+
+#define TclDTraceInfo(info, args, argsi)    {*args = ""; *argsi = 0;}
+
+#endif /* USE_DTRACE */
+
+#else /* TCL_DTRACE_DEBUG */
+
+#define USE_DTRACE 1
+
+#if !defined(TCL_DTRACE_DEBUG_LOG_ENABLED) || !(TCL_DTRACE_DEBUG_LOG_ENABLED)
+#undef TCL_DTRACE_DEBUG_LOG_ENABLED
+#define TCL_DTRACE_DEBUG_LOG_ENABLED 0
+#endif
+
+#if !defined(TCL_DTRACE_DEBUG_INST_PROBES) || !(TCL_DTRACE_DEBUG_INST_PROBES)
+#undef TCL_DTRACE_DEBUG_INST_PROBES
+#define TCL_DTRACE_DEBUG_INST_PROBES 0
+#endif
+
+MODULE_SCOPE int tclDTraceDebugEnabled, tclDTraceDebugIndent;
+MODULE_SCOPE FILE *tclDTraceDebugLog;
+MODULE_SCOPE void TclDTraceOpenDebugLog(void);
+MODULE_SCOPE void TclDTraceInfo(Tcl_Obj *info, const char **args, int *argsi);
+
+#define TCL_DTRACE_DEBUG_LOG() \
+    int tclDTraceDebugEnabled = TCL_DTRACE_DEBUG_LOG_ENABLED;	\
+    int tclDTraceDebugIndent = 0;				\
+    FILE *tclDTraceDebugLog = NULL;				\
+    void TclDTraceOpenDebugLog(void) {				\
+	char n[35];						\
+	sprintf(n, "/tmp/tclDTraceDebug-%lu.log",		\
+		(unsigned long) getpid());			\
+	tclDTraceDebugLog = fopen(n, "a");			\
+    }
+
+#define TclDTraceDbgMsg(p, m, ...) \
+    do {								\
+	if (tclDTraceDebugEnabled) {					\
+	    int _l, _t = 0;						\
+	    if (!tclDTraceDebugLog) { TclDTraceOpenDebugLog(); }	\
+	    fprintf(tclDTraceDebugLog, "%.12s:%.4d:%n",			\
+		    strrchr(__FILE__, '/')+1, __LINE__, &_l); _t += _l; \
+	    fprintf(tclDTraceDebugLog, " %.*s():%n",			\
+		    (_t < 18 ? 18 - _t : 0) + 18, __func__, &_l); _t += _l; \
+	    fprintf(tclDTraceDebugLog, "%*s" p "%n",			\
+		    (_t < 40 ? 40 - _t : 0) + 2 * tclDTraceDebugIndent, \
+		    "", &_l); _t += _l;					\
+	    fprintf(tclDTraceDebugLog, "%*s" m "\n",			\
+		    (_t < 64 ? 64 - _t : 1), "", ##__VA_ARGS__);	\
+	    fflush(tclDTraceDebugLog);					\
+	}								\
+    } while (0)
+
+#define TCL_DTRACE_PROC_ENTRY_ENABLED()	    1
+#define TCL_DTRACE_PROC_RETURN_ENABLED()    1
+#define TCL_DTRACE_PROC_RESULT_ENABLED()    1
+#define TCL_DTRACE_PROC_ARGS_ENABLED()	    1
+#define TCL_DTRACE_PROC_INFO_ENABLED()	    1
+#define TCL_DTRACE_PROC_ENTRY(a0, a1, a2) \
+	tclDTraceDebugIndent++; \
+	TclDTraceDbgMsg("-> proc-entry", "%s %d %p", a0, a1, a2)
+#define TCL_DTRACE_PROC_RETURN(a0, a1) \
+	TclDTraceDbgMsg("<- proc-return", "%s %d", a0, a1); \
+	tclDTraceDebugIndent--
+#define TCL_DTRACE_PROC_RESULT(a0, a1, a2, a3) \
+	TclDTraceDbgMsg(" | proc-result", "%s %d %s %p", a0, a1, a2, a3)
+#define TCL_DTRACE_PROC_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+	TclDTraceDbgMsg(" | proc-args", "%s %s %s %s %s %s %s %s %s %s", a0, \
+		a1, a2, a3, a4, a5, a6, a7, a8, a9)
+#define TCL_DTRACE_PROC_INFO(a0, a1, a2, a3, a4, a5, a6, a7) \
+	TclDTraceDbgMsg(" | proc-info", "%s %s %s %s %d %d %s %s", a0, a1, \
+		a2, a3, a4, a5, a6, a7)
+
+#define TCL_DTRACE_CMD_ENTRY_ENABLED()	    1
+#define TCL_DTRACE_CMD_RETURN_ENABLED()	    1
+#define TCL_DTRACE_CMD_RESULT_ENABLED()	    1
+#define TCL_DTRACE_CMD_ARGS_ENABLED()	    1
+#define TCL_DTRACE_CMD_INFO_ENABLED()	    1
+#define TCL_DTRACE_CMD_ENTRY(a0, a1, a2) \
+	tclDTraceDebugIndent++; \
+	TclDTraceDbgMsg("-> cmd-entry", "%s %d %p", a0, a1, a2)
+#define TCL_DTRACE_CMD_RETURN(a0, a1) \
+	TclDTraceDbgMsg("<- cmd-return", "%s %d", a0, a1); \
+	tclDTraceDebugIndent--
+#define TCL_DTRACE_CMD_RESULT(a0, a1, a2, a3) \
+	TclDTraceDbgMsg(" | cmd-result", "%s %d %s %p", a0, a1, a2, a3)
+#define TCL_DTRACE_CMD_ARGS(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+	TclDTraceDbgMsg(" | cmd-args", "%s %s %s %s %s %s %s %s %s %s", a0, \
+		a1, a2, a3, a4, a5, a6, a7, a8, a9)
+#define TCL_DTRACE_CMD_INFO(a0, a1, a2, a3, a4, a5, a6, a7) \
+	TclDTraceDbgMsg(" | cmd-info", "%s %s %s %s %d %d %s %s", a0, a1, \
+		a2, a3, a4, a5, a6, a7)
+
+#define TCL_DTRACE_INST_START_ENABLED()	    TCL_DTRACE_DEBUG_INST_PROBES
+#define TCL_DTRACE_INST_DONE_ENABLED()	    TCL_DTRACE_DEBUG_INST_PROBES
+#define TCL_DTRACE_INST_START(a0, a1, a2) \
+	TclDTraceDbgMsg(" | inst-start", "%s %d %p", a0, a1, a2)
+#define TCL_DTRACE_INST_DONE(a0, a1, a2) \
+	TclDTraceDbgMsg(" | inst-end", "%s %d %p", a0, a1, a2)
+
+#define TCL_DTRACE_TCL_PROBE_ENABLED()	    1
+#define TCL_DTRACE_TCL_PROBE(a0, a1, a2, a3, a4, a5, a6, a7, a8, a9) \
+    do {								\
+	tclDTraceDebugEnabled = 1;					\
+	TclDTraceDbgMsg(" | tcl-probe", "%s %s %s %s %s %s %s %s %s %s", a0, \
+		a1, a2, a3, a4, a5, a6, a7, a8, a9);			\
+    } while (0)
+
+#endif /* TCL_DTRACE_DEBUG */
+
+#endif /* _TCLCOMPILATION */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclConfig.c b/generic/tclConfig.c
new file mode 100644
index 0000000..eb6807c
--- /dev/null
+++ b/generic/tclConfig.c
@@ -0,0 +1,408 @@
+/*
+ * tclConfig.c --
+ *
+ *	This file provides the facilities which allow Tcl and other packages
+ *	to embed configuration information into their binary libraries.
+ *
+ * Copyright (c) 2002 Andreas Kupries <andreas_kupries@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Internal structure to hold embedded configuration information.
+ *
+ * Our structure is a two-level dictionary associated with the 'interp'. The
+ * first level is keyed with the package name and maps to the dictionary for
+ * that package. The package dictionary is keyed with metadata keys and maps
+ * to the metadata value for that key. This is package specific. The metadata
+ * values are in UTF-8, converted from the external representation given to us
+ * by the caller.
+ */
+
+#define ASSOC_KEY	"tclPackageAboutDict"
+
+/*
+ * A ClientData struct for the QueryConfig command.  Store the three bits
+ * of data we need; the package name for which we store a config dict,
+ * the (Tcl_Interp *) in which it is stored, and the encoding.
+ */
+
+typedef struct QCCD {
+    Tcl_Obj *pkg;
+    Tcl_Interp *interp;
+    char *encoding;
+} QCCD;
+
+/*
+ * Static functions in this file:
+ */
+
+static int		QueryConfigObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    struct Tcl_Obj *const *objv);
+static void		QueryConfigDelete(ClientData clientData);
+static Tcl_Obj *	GetConfigDict(Tcl_Interp *interp);
+static void		ConfigDictDeleteProc(ClientData clientData,
+			    Tcl_Interp *interp);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegisterConfig --
+ *
+ *	See TIP#59 for details on what this function does.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Creates namespace and cfg query command in it as per TIP #59.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_RegisterConfig(
+    Tcl_Interp *interp,		/* Interpreter the configuration command is
+				 * registered in. */
+    const char *pkgName,	/* Name of the package registering the
+				 * embedded configuration. ASCII, thus in
+				 * UTF-8 too. */
+    const Tcl_Config *configuration,	/* Embedded configuration. */
+    const char *valEncoding)	/* Name of the encoding used to store the
+				 * configuration values, ASCII, thus UTF-8. */
+{
+    Tcl_Obj *pDB, *pkgDict;
+    Tcl_DString cmdName;
+    const Tcl_Config *cfg;
+    QCCD *cdPtr = ckalloc(sizeof(QCCD));
+
+    cdPtr->interp = interp;
+    if (valEncoding) {
+	cdPtr->encoding = ckalloc(strlen(valEncoding)+1);
+	strcpy(cdPtr->encoding, valEncoding);
+    } else {
+	cdPtr->encoding = NULL;
+    }
+    cdPtr->pkg = Tcl_NewStringObj(pkgName, -1);
+
+    /*
+     * Phase I: Adding the provided information to the internal database of
+     * package meta data.
+     *
+     * Phase II: Create a command for querying this database, specific to the
+     * package registering its configuration. This is the approved interface
+     * in TIP 59. In the future a more general interface should be done, as
+     * follow-up to TIP 59. Simply because our database is now general across
+     * packages, and not a structure tied to one package.
+     *
+     * Note, the created command will have a reference through its clientdata.
+     */
+
+    Tcl_IncrRefCount(cdPtr->pkg);
+
+    /*
+     * For venc == NULL aka bogus encoding we skip the step setting up the
+     * dictionaries visible at Tcl level. I.e. they are not filled
+     */
+
+    pDB = GetConfigDict(interp);
+
+    /*
+     * Retrieve package specific configuration...
+     */
+
+    if (Tcl_DictObjGet(interp, pDB, cdPtr->pkg, &pkgDict) != TCL_OK
+	    || (pkgDict == NULL)) {
+	pkgDict = Tcl_NewDictObj();
+    } else if (Tcl_IsShared(pkgDict)) {
+	pkgDict = Tcl_DuplicateObj(pkgDict);
+    }
+
+    /*
+     * Extend the package configuration...
+     * We cannot assume that the encodings are initialized, therefore
+     * store the value as-is in a byte array. See Bug [9b2e636361].
+     */
+
+    for (cfg=configuration ; cfg->key!=NULL && cfg->key[0]!='\0' ; cfg++) {
+	Tcl_DictObjPut(interp, pkgDict, Tcl_NewStringObj(cfg->key, -1),
+		Tcl_NewByteArrayObj((unsigned char *)cfg->value, strlen(cfg->value)));
+    }
+
+    /*
+     * Write the changes back into the overall database.
+     */
+
+    Tcl_DictObjPut(interp, pDB, cdPtr->pkg, pkgDict);
+
+    /*
+     * Now create the interface command for retrieval of the package
+     * information.
+     */
+
+    Tcl_DStringInit(&cmdName);
+    TclDStringAppendLiteral(&cmdName, "::");
+    Tcl_DStringAppend(&cmdName, pkgName, -1);
+
+    /*
+     * The incomplete command name is the name of the namespace to place it
+     * in.
+     */
+
+    if (Tcl_FindNamespace(interp, Tcl_DStringValue(&cmdName), NULL,
+	    TCL_GLOBAL_ONLY) == NULL) {
+	if (Tcl_CreateNamespace(interp, Tcl_DStringValue(&cmdName),
+		NULL, NULL) == NULL) {
+	    Tcl_Panic("%s.\n%s: %s",
+		    Tcl_GetStringResult(interp), "Tcl_RegisterConfig",
+		    "Unable to create namespace for package configuration.");
+	}
+    }
+
+    TclDStringAppendLiteral(&cmdName, "::pkgconfig");
+
+    if (Tcl_CreateObjCommand(interp, Tcl_DStringValue(&cmdName),
+	    QueryConfigObjCmd, cdPtr, QueryConfigDelete) == NULL) {
+	Tcl_Panic("%s: %s", "Tcl_RegisterConfig",
+		"Unable to create query command for package configuration");
+    }
+
+    Tcl_DStringFree(&cmdName);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * QueryConfigObjCmd --
+ *
+ *	Implementation of "::<package>::pkgconfig", the command to query
+ *	configuration information embedded into a binary library.
+ *
+ * Results:
+ *	A standard tcl result.
+ *
+ * Side effects:
+ *	See the manual for what this command does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+QueryConfigObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    struct Tcl_Obj *const *objv)
+{
+    QCCD *cdPtr = clientData;
+    Tcl_Obj *pkgName = cdPtr->pkg;
+    Tcl_Obj *pDB, *pkgDict, *val, *listPtr;
+    int n, index;
+    static const char *const subcmdStrings[] = {
+	"get", "list", NULL
+    };
+    enum subcmds {
+	CFG_GET, CFG_LIST
+    };
+    Tcl_DString conv;
+    Tcl_Encoding venc = NULL;
+    const char *value;
+
+    if ((objc < 2) || (objc > 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "subcommand ?arg?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], subcmdStrings, "subcommand", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    pDB = GetConfigDict(interp);
+    if (Tcl_DictObjGet(interp, pDB, pkgName, &pkgDict) != TCL_OK
+	    || pkgDict == NULL) {
+	/*
+	 * Maybe a Tcl_Panic is better, because the package data has to be
+	 * present.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("package not known", -1));
+	Tcl_SetErrorCode(interp, "TCL", "FATAL", "PKGCFG_BASE",
+		TclGetString(pkgName), NULL);
+	return TCL_ERROR;
+    }
+
+    switch ((enum subcmds) index) {
+    case CFG_GET:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "key");
+	    return TCL_ERROR;
+	}
+
+	if (Tcl_DictObjGet(interp, pkgDict, objv[2], &val) != TCL_OK
+		|| val == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("key not known", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CONFIG",
+		    TclGetString(objv[2]), NULL);
+	    return TCL_ERROR;
+	}
+
+	if (cdPtr->encoding) {
+	    venc = Tcl_GetEncoding(interp, cdPtr->encoding);
+	    if (!venc) {
+		return TCL_ERROR;
+	    }
+	}
+	/*
+	 * Value is stored as-is in a byte array, see Bug [9b2e636361],
+	 * so we have to decode it first.
+	 */
+	value = (const char *) Tcl_GetByteArrayFromObj(val, &n);
+	value = Tcl_ExternalToUtfDString(venc, value, n, &conv);
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(value,
+		Tcl_DStringLength(&conv)));
+	Tcl_DStringFree(&conv);
+	return TCL_OK;
+
+    case CFG_LIST:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+
+	Tcl_DictObjSize(interp, pkgDict, &n);
+	listPtr = Tcl_NewListObj(n, NULL);
+
+	if (!listPtr) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "insufficient memory to create list", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	    return TCL_ERROR;
+	}
+
+	if (n) {
+	    Tcl_DictSearch s;
+	    Tcl_Obj *key;
+	    int done;
+
+	    for (Tcl_DictObjFirst(interp, pkgDict, &s, &key, NULL, &done);
+		    !done; Tcl_DictObjNext(&s, &key, NULL, &done)) {
+		Tcl_ListObjAppendElement(NULL, listPtr, key);
+	    }
+	}
+
+	Tcl_SetObjResult(interp, listPtr);
+	return TCL_OK;
+
+    default:
+	Tcl_Panic("QueryConfigObjCmd: Unknown subcommand to 'pkgconfig'. This can't happen");
+	break;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * QueryConfigDelete --
+ *
+ *	Command delete function. Cleans up after the configuration query
+ *	command when it is deleted by the user or during finalization.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deallocates all non-transient memory allocated by Tcl_RegisterConfig.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static void
+QueryConfigDelete(
+    ClientData clientData)
+{
+    QCCD *cdPtr = clientData;
+    Tcl_Obj *pkgName = cdPtr->pkg;
+    Tcl_Obj *pDB = GetConfigDict(cdPtr->interp);
+
+    Tcl_DictObjRemove(NULL, pDB, pkgName);
+    Tcl_DecrRefCount(pkgName);
+    if (cdPtr->encoding) {
+	ckfree(cdPtr->encoding);
+    }
+    ckfree(cdPtr);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * GetConfigDict --
+ *
+ *	Retrieve the package metadata database from the interpreter.
+ *	Initializes it, if not present yet.
+ *
+ * Results:
+ *	A Tcl_Obj reference
+ *
+ * Side effects:
+ *	May allocate a Tcl_Obj.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+GetConfigDict(
+    Tcl_Interp *interp)
+{
+    Tcl_Obj *pDB = Tcl_GetAssocData(interp, ASSOC_KEY, NULL);
+
+    if (pDB == NULL) {
+	pDB = Tcl_NewDictObj();
+	Tcl_IncrRefCount(pDB);
+	Tcl_SetAssocData(interp, ASSOC_KEY, ConfigDictDeleteProc, pDB);
+    }
+
+    return pDB;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConfigDictDeleteProc --
+ *
+ *	This function is associated with the "Package About dict" assoc data
+ *	for an interpreter; it is invoked when the interpreter is deleted in
+ *	order to free the information associated with any pending error
+ *	reports.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The package metadata database is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ConfigDictDeleteProc(
+    ClientData clientData,	/* Pointer to Tcl_Obj. */
+    Tcl_Interp *interp)		/* Interpreter being deleted. */
+{
+    Tcl_Obj *pDB = clientData;
+
+    Tcl_DecrRefCount(pDB);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclDTrace.d b/generic/tclDTrace.d
new file mode 100644
index 0000000..360bdff
--- /dev/null
+++ b/generic/tclDTrace.d
@@ -0,0 +1,225 @@
+/*
+ * tclDTrace.d --
+ *
+ *	Tcl DTrace provider.
+ *
+ * Copyright (c) 2007-2008 Daniel A. Steffen <das@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+typedef struct Tcl_Obj Tcl_Obj;
+typedef const char* TclDTraceStr;
+
+/*
+ * Tcl DTrace probes
+ */
+
+provider tcl {
+    /***************************** proc probes *****************************/
+    /*
+     *	tcl*:::proc-entry probe
+     *	    triggered immediately before proc bytecode execution
+     *		arg0: proc name				(string)
+     *		arg1: number of arguments		(int)
+     *		arg2: array of proc argument objects	(Tcl_Obj**)
+     */
+    probe proc__entry(TclDTraceStr name, int objc, struct Tcl_Obj **objv);
+    /*
+     *	tcl*:::proc-return probe
+     *	    triggered immediately after proc bytecode execution
+     *		arg0: proc name				(string)
+     *		arg1: return code			(int)
+     */
+    probe proc__return(TclDTraceStr name, int code);
+    /*
+     *	tcl*:::proc-result probe
+     *	    triggered after proc-return probe and result processing
+     *		arg0: proc name				(string)
+     *		arg1: return code			(int)
+     *		arg2: proc result			(string)
+     *		arg3: proc result object		(Tcl_Obj*)
+     */
+    probe proc__result(TclDTraceStr name, int code, TclDTraceStr result,
+	    struct Tcl_Obj *resultobj);
+    /*
+     *	tcl*:::proc-args probe
+     *	    triggered before proc-entry probe, gives access to string
+     *	    representation of proc arguments
+     *		arg0: proc name				(string)
+     *		arg1-arg9: proc arguments or NULL	(strings)
+     */
+    probe proc__args(TclDTraceStr name, TclDTraceStr arg1, TclDTraceStr arg2,
+	    TclDTraceStr arg3, TclDTraceStr arg4, TclDTraceStr arg5,
+	    TclDTraceStr arg6, TclDTraceStr arg7, TclDTraceStr arg8,
+	    TclDTraceStr arg9);
+    /*
+     *	tcl*:::proc-info probe
+     *	    triggered before proc-entry probe, gives access to TIP 280
+     *	    information for the proc invocation (i.e. [info frame 0])
+     *		arg0: TIP 280 cmd			(string)
+     *		arg1: TIP 280 type			(string)
+     *		arg2: TIP 280 proc			(string)
+     *		arg3: TIP 280 file			(string)
+     *		arg4: TIP 280 line			(int)
+     *		arg5: TIP 280 level			(int)
+     *		arg6: TclOO method			(string)
+     *		arg7: TclOO class/object		(string)
+     */
+    probe proc__info(TclDTraceStr cmd, TclDTraceStr type, TclDTraceStr proc,
+	    TclDTraceStr file, int line, int level, TclDTraceStr method,
+	    TclDTraceStr class);
+
+    /***************************** cmd probes ******************************/
+    /*
+     *	tcl*:::cmd-entry probe
+     *	    triggered immediately before commmand execution
+     *		arg0: command name			(string)
+     *		arg1: number of arguments		(int)
+     *		arg2: array of command argument objects	(Tcl_Obj**)
+     */
+    probe cmd__entry(TclDTraceStr name, int objc, struct Tcl_Obj **objv);
+    /*
+     *	tcl*:::cmd-return probe
+     *	    triggered immediately after commmand execution
+     *		arg0: command name			(string)
+     *		arg1: return code			(int)
+     */
+    probe cmd__return(TclDTraceStr name, int code);
+    /*
+     *	tcl*:::cmd-result probe
+     *	    triggered after cmd-return probe and result processing
+     *		arg0: command name			(string)
+     *		arg1: return code			(int)
+     *		arg2: command result			(string)
+     *		arg3: command result object		(Tcl_Obj*)
+     */
+    probe cmd__result(TclDTraceStr name, int code, TclDTraceStr result,
+	    struct Tcl_Obj *resultobj);
+    /*
+     *	tcl*:::cmd-args probe
+     *	    triggered before cmd-entry probe, gives access to string
+     *	    representation of command arguments
+     *		arg0: command name			(string)
+     *		arg1-arg9: command arguments or NULL	(strings)
+     */
+    probe cmd__args(TclDTraceStr name, TclDTraceStr arg1, TclDTraceStr arg2,
+	    TclDTraceStr arg3, TclDTraceStr arg4, TclDTraceStr arg5,
+	    TclDTraceStr arg6, TclDTraceStr arg7, TclDTraceStr arg8,
+	    TclDTraceStr arg9);
+    /*
+     *	tcl*:::cmd-info probe
+     *	    triggered before cmd-entry probe, gives access to TIP 280
+     *	    information for the command invocation (i.e. [info frame 0])
+     *		arg0: TIP 280 cmd			(string)
+     *		arg1: TIP 280 type			(string)
+     *		arg2: TIP 280 proc			(string)
+     *		arg3: TIP 280 file			(string)
+     *		arg4: TIP 280 line			(int)
+     *		arg5: TIP 280 level			(int)
+     *		arg6: TclOO method			(string)
+     *		arg7: TclOO class/object		(string)
+     */
+    probe cmd__info(TclDTraceStr cmd, TclDTraceStr type, TclDTraceStr proc,
+	    TclDTraceStr file, int line, int level, TclDTraceStr method,
+	    TclDTraceStr class);
+
+    /***************************** inst probes *****************************/
+    /*
+     *	tcl*:::inst-start probe
+     *	    triggered immediately before execution of a bytecode
+     *		arg0: bytecode name			(string)
+     *		arg1: depth of stack			(int)
+     *		arg2: top of stack			(Tcl_Obj**)
+     */
+    probe inst__start(TclDTraceStr name, int depth, struct Tcl_Obj **stack);
+    /*
+     *	tcl*:::inst-done probe
+     *	    triggered immediately after execution of a bytecode
+     *		arg0: bytecode name			(string)
+     *		arg1: depth of stack			(int)
+     *		arg2: top of stack			(Tcl_Obj**)
+     */
+    probe inst__done(TclDTraceStr name, int depth, struct Tcl_Obj **stack);
+
+    /***************************** obj probes ******************************/
+    /*
+     *	tcl*:::obj-create probe
+     *	    triggered immediately after a new Tcl_Obj has been created
+     *		arg0: object created			(Tcl_Obj*)
+     */
+    probe obj__create(struct Tcl_Obj* obj);
+    /*
+     *	tcl*:::obj-free probe
+     *	    triggered immediately before a Tcl_Obj is freed
+     *		arg0: object to be freed		(Tcl_Obj*)
+     */
+    probe obj__free(struct Tcl_Obj* obj);
+
+    /***************************** tcl probes ******************************/
+    /*
+     *	tcl*:::tcl-probe probe
+     *	    triggered when the ::tcl::dtrace command is called
+     *		arg0-arg9: command arguments		(strings)
+     */
+    probe tcl__probe(TclDTraceStr arg0, TclDTraceStr arg1, TclDTraceStr arg2,
+	    TclDTraceStr arg3, TclDTraceStr arg4, TclDTraceStr arg5,
+	    TclDTraceStr arg6, TclDTraceStr arg7, TclDTraceStr arg8,
+	    TclDTraceStr arg9);
+};
+
+/*
+ * Tcl types and constants for use in DTrace scripts
+ */
+
+typedef struct Tcl_ObjType {
+    char *name;
+    void *freeIntRepProc;
+    void *dupIntRepProc;
+    void *updateStringProc;
+    void *setFromAnyProc;
+} Tcl_ObjType;
+
+struct Tcl_Obj {
+    int refCount;
+    char *bytes;
+    int length;
+    Tcl_ObjType *typePtr;
+    union {
+	long longValue;
+	double doubleValue;
+	void *otherValuePtr;
+	int64_t wideValue;
+	struct {
+	    void *ptr1;
+	    void *ptr2;
+	} twoPtrValue;
+	struct {
+	    void *ptr;
+	    unsigned long value;
+	} ptrAndLongRep;
+    } internalRep;
+};
+
+enum return_codes {
+    TCL_OK = 0,
+    TCL_ERROR,
+    TCL_RETURN,
+    TCL_BREAK,
+    TCL_CONTINUE
+};
+
+#pragma D attributes Evolving/Evolving/Common provider tcl provider
+#pragma D attributes Private/Private/Common provider tcl module
+#pragma D attributes Private/Private/Common provider tcl function
+#pragma D attributes Evolving/Evolving/Common provider tcl name
+#pragma D attributes Evolving/Evolving/Common provider tcl args
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclDate.c b/generic/tclDate.c
new file mode 100644
index 0000000..e4dd000
--- /dev/null
+++ b/generic/tclDate.c
@@ -0,0 +1,2914 @@
+/* A Bison parser, made by GNU Bison 2.3.  */
+
+/* Skeleton implementation for Bison's Yacc-like parsers in C
+
+   Copyright (C) 1984, 1989, 1990, 2000, 2001, 2002, 2003, 2004, 2005, 2006
+   Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2, or (at your option)
+   any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software
+   Foundation, Inc., 51 Franklin Street, Fifth Floor,
+   Boston, MA 02110-1301, USA.  */
+
+/* As a special exception, you may create a larger work that contains
+   part or all of the Bison parser skeleton and distribute that work
+   under terms of your choice, so long as that work isn't itself a
+   parser generator using the skeleton or a modified version thereof
+   as a parser skeleton.  Alternatively, if you modify or redistribute
+   the parser skeleton itself, you may (at your option) remove this
+   special exception, which will cause the skeleton and the resulting
+   Bison output files to be licensed under the GNU General Public
+   License without this special exception.
+
+   This special exception was added by the Free Software Foundation in
+   version 2.2 of Bison.  */
+
+/* C LALR(1) parser skeleton written by Richard Stallman, by
+   simplifying the original so-called "semantic" parser.  */
+
+/* All symbols defined below should begin with yy or YY, to avoid
+   infringing on user name space.  This should be done even for local
+   variables, as they might otherwise be expanded by user macros.
+   There are some unavoidable exceptions within include files to
+   define necessary library symbols; they are noted "INFRINGES ON
+   USER NAME SPACE" below.  */
+
+/* Identify Bison output.  */
+#define YYBISON 1
+
+/* Bison version.  */
+#define YYBISON_VERSION "2.3"
+
+/* Skeleton name.  */
+#define YYSKELETON_NAME "yacc.c"
+
+/* Pure parsers.  */
+#define YYPURE 1
+
+/* Using locations.  */
+#define YYLSP_NEEDED 1
+
+/* Substitute the variable and function names.  */
+#define yyparse TclDateparse
+#define yylex   TclDatelex
+#define yyerror TclDateerror
+#define yylval  TclDatelval
+#define yychar  TclDatechar
+#define yydebug TclDatedebug
+#define yynerrs TclDatenerrs
+#define yylloc TclDatelloc
+
+/* Tokens.  */
+#ifndef YYTOKENTYPE
+# define YYTOKENTYPE
+   /* Put the tokens into the symbol table, so that GDB and other debuggers
+      know about them.  */
+   enum yytokentype {
+     tAGO = 258,
+     tDAY = 259,
+     tDAYZONE = 260,
+     tID = 261,
+     tMERIDIAN = 262,
+     tMONTH = 263,
+     tMONTH_UNIT = 264,
+     tSTARDATE = 265,
+     tSEC_UNIT = 266,
+     tSNUMBER = 267,
+     tUNUMBER = 268,
+     tZONE = 269,
+     tEPOCH = 270,
+     tDST = 271,
+     tISOBASE = 272,
+     tDAY_UNIT = 273,
+     tNEXT = 274
+   };
+#endif
+/* Tokens.  */
+#define tAGO 258
+#define tDAY 259
+#define tDAYZONE 260
+#define tID 261
+#define tMERIDIAN 262
+#define tMONTH 263
+#define tMONTH_UNIT 264
+#define tSTARDATE 265
+#define tSEC_UNIT 266
+#define tSNUMBER 267
+#define tUNUMBER 268
+#define tZONE 269
+#define tEPOCH 270
+#define tDST 271
+#define tISOBASE 272
+#define tDAY_UNIT 273
+#define tNEXT 274
+
+
+
+
+/* Copy the first part of user declarations.  */
+
+
+/*
+ * tclDate.c --
+ *
+ *	This file is generated from a yacc grammar defined in the file
+ *	tclGetDate.y. It should not be edited directly.
+ *
+ * Copyright (c) 1992-1995 Karl Lehenbauer and Mark Diekhans.
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+#include "tclInt.h"
+
+/*
+ * Bison generates several labels that happen to be unused. MS Visual C++
+ * doesn't like that, and complains. Tell it to shut up.
+ */
+
+#ifdef _MSC_VER
+#pragma warning( disable : 4102 )
+#endif /* _MSC_VER */
+
+/*
+ * yyparse will accept a 'struct DateInfo' as its parameter; that's where the
+ * parsed fields will be returned.
+ */
+
+typedef struct DateInfo {
+
+    Tcl_Obj* messages;		/* Error messages */
+    const char* separatrix;	/* String separating messages */
+
+    time_t dateYear;
+    time_t dateMonth;
+    time_t dateDay;
+    int dateHaveDate;
+
+    time_t dateHour;
+    time_t dateMinutes;
+    time_t dateSeconds;
+    int dateMeridian;
+    int dateHaveTime;
+
+    time_t dateTimezone;
+    int dateDSTmode;
+    int dateHaveZone;
+
+    time_t dateRelMonth;
+    time_t dateRelDay;
+    time_t dateRelSeconds;
+    int dateHaveRel;
+
+    time_t dateMonthOrdinal;
+    int dateHaveOrdinalMonth;
+
+    time_t dateDayOrdinal;
+    time_t dateDayNumber;
+    int dateHaveDay;
+
+    const char *dateStart;
+    const char *dateInput;
+    time_t *dateRelPointer;
+
+    int dateDigitCount;
+} DateInfo;
+
+#define YYMALLOC	ckalloc
+#define YYFREE(x)	(ckfree((void*) (x)))
+
+#define yyDSTmode	(info->dateDSTmode)
+#define yyDayOrdinal	(info->dateDayOrdinal)
+#define yyDayNumber	(info->dateDayNumber)
+#define yyMonthOrdinal	(info->dateMonthOrdinal)
+#define yyHaveDate	(info->dateHaveDate)
+#define yyHaveDay	(info->dateHaveDay)
+#define yyHaveOrdinalMonth (info->dateHaveOrdinalMonth)
+#define yyHaveRel	(info->dateHaveRel)
+#define yyHaveTime	(info->dateHaveTime)
+#define yyHaveZone	(info->dateHaveZone)
+#define yyTimezone	(info->dateTimezone)
+#define yyDay		(info->dateDay)
+#define yyMonth		(info->dateMonth)
+#define yyYear		(info->dateYear)
+#define yyHour		(info->dateHour)
+#define yyMinutes	(info->dateMinutes)
+#define yySeconds	(info->dateSeconds)
+#define yyMeridian	(info->dateMeridian)
+#define yyRelMonth	(info->dateRelMonth)
+#define yyRelDay	(info->dateRelDay)
+#define yyRelSeconds	(info->dateRelSeconds)
+#define yyRelPointer	(info->dateRelPointer)
+#define yyInput		(info->dateInput)
+#define yyDigitCount	(info->dateDigitCount)
+
+#define EPOCH		1970
+#define START_OF_TIME	1902
+#define END_OF_TIME	2037
+
+/*
+ * The offset of tm_year of struct tm returned by localtime, gmtime, etc.
+ * Posix requires 1900.
+ */
+
+#define TM_YEAR_BASE	1900
+
+#define HOUR(x)		((int) (60 * x))
+#define SECSPERDAY	(24L * 60L * 60L)
+#define IsLeapYear(x)	((x % 4 == 0) && (x % 100 != 0 || x % 400 == 0))
+
+/*
+ * An entry in the lexical lookup table.
+ */
+
+typedef struct _TABLE {
+    const char *name;
+    int type;
+    time_t value;
+} TABLE;
+
+/*
+ * Daylight-savings mode: on, off, or not yet known.
+ */
+
+typedef enum _DSTMODE {
+    DSTon, DSToff, DSTmaybe
+} DSTMODE;
+
+/*
+ * Meridian: am, pm, or 24-hour style.
+ */
+
+typedef enum _MERIDIAN {
+    MERam, MERpm, MER24
+} MERIDIAN;
+
+
+
+/* Enabling traces.  */
+#ifndef YYDEBUG
+# define YYDEBUG 0
+#endif
+
+/* Enabling verbose error messages.  */
+#ifdef YYERROR_VERBOSE
+# undef YYERROR_VERBOSE
+# define YYERROR_VERBOSE 1
+#else
+# define YYERROR_VERBOSE 0
+#endif
+
+/* Enabling the token table.  */
+#ifndef YYTOKEN_TABLE
+# define YYTOKEN_TABLE 0
+#endif
+
+#if ! defined YYSTYPE && ! defined YYSTYPE_IS_DECLARED
+typedef union YYSTYPE
+
+{
+    time_t Number;
+    enum _MERIDIAN Meridian;
+}
+/* Line 187 of yacc.c.  */
+
+	YYSTYPE;
+# define yystype YYSTYPE /* obsolescent; will be withdrawn */
+# define YYSTYPE_IS_DECLARED 1
+# define YYSTYPE_IS_TRIVIAL 1
+#endif
+
+#if ! defined YYLTYPE && ! defined YYLTYPE_IS_DECLARED
+typedef struct YYLTYPE
+{
+  int first_line;
+  int first_column;
+  int last_line;
+  int last_column;
+} YYLTYPE;
+# define yyltype YYLTYPE /* obsolescent; will be withdrawn */
+# define YYLTYPE_IS_DECLARED 1
+# define YYLTYPE_IS_TRIVIAL 1
+#endif
+
+
+/* Copy the second part of user declarations.  */
+
+
+
+/*
+ * Prototypes of internal functions.
+ */
+
+static int		LookupWord(YYSTYPE* yylvalPtr, char *buff);
+ static void		TclDateerror(YYLTYPE* location,
+				     DateInfo* info, const char *s);
+ static int		TclDatelex(YYSTYPE* yylvalPtr, YYLTYPE* location,
+				   DateInfo* info);
+static time_t		ToSeconds(time_t Hours, time_t Minutes,
+			    time_t Seconds, MERIDIAN Meridian);
+MODULE_SCOPE int	yyparse(DateInfo*);
+
+
+
+/* Line 216 of yacc.c.  */
+
+
+#ifdef short
+# undef short
+#endif
+
+#ifdef YYTYPE_UINT8
+typedef YYTYPE_UINT8 yytype_uint8;
+#else
+typedef unsigned char yytype_uint8;
+#endif
+
+#ifdef YYTYPE_INT8
+typedef YYTYPE_INT8 yytype_int8;
+#elif (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+typedef signed char yytype_int8;
+#else
+typedef short int yytype_int8;
+#endif
+
+#ifdef YYTYPE_UINT16
+typedef YYTYPE_UINT16 yytype_uint16;
+#else
+typedef unsigned short int yytype_uint16;
+#endif
+
+#ifdef YYTYPE_INT16
+typedef YYTYPE_INT16 yytype_int16;
+#else
+typedef short int yytype_int16;
+#endif
+
+#ifndef YYSIZE_T
+# ifdef __SIZE_TYPE__
+#  define YYSIZE_T __SIZE_TYPE__
+# else
+#  define YYSIZE_T size_t
+# endif
+#endif
+
+#define YYSIZE_MAXIMUM ((YYSIZE_T) -1)
+
+#ifndef YY_
+# if YYENABLE_NLS
+#  if ENABLE_NLS
+#   include <libintl.h> /* INFRINGES ON USER NAME SPACE */
+#   define YY_(msgid) dgettext ("bison-runtime", msgid)
+#  endif
+# endif
+# ifndef YY_
+#  define YY_(msgid) msgid
+# endif
+#endif
+
+/* Suppress unused-variable warnings by "using" E.  */
+#if ! defined lint || defined __GNUC__
+# define YYUSE(e) ((void) (e))
+#else
+# define YYUSE(e) /* empty */
+#endif
+
+/* Identity function, used to suppress warnings about constant conditions.  */
+#ifndef lint
+# define YYID(n) (n)
+#else
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static int
+YYID (int i)
+#else
+static int
+YYID (i)
+    int i;
+#endif
+{
+  return i;
+}
+#endif
+
+#if ! defined yyoverflow || YYERROR_VERBOSE
+
+/* The parser invokes alloca or malloc; define the necessary symbols.  */
+
+# ifdef YYSTACK_USE_ALLOCA
+#  if YYSTACK_USE_ALLOCA
+#   ifdef __GNUC__
+#    define YYSTACK_ALLOC __builtin_alloca
+#   elif defined __BUILTIN_VA_ARG_INCR
+#    include <alloca.h> /* INFRINGES ON USER NAME SPACE */
+#   elif defined _AIX
+#    define YYSTACK_ALLOC __alloca
+#   elif defined _MSC_VER
+#    include <malloc.h> /* INFRINGES ON USER NAME SPACE */
+#    define alloca _alloca
+#   else
+#    define YYSTACK_ALLOC alloca
+#    if ! defined _ALLOCA_H && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+#     include <stdlib.h> /* INFRINGES ON USER NAME SPACE */
+#     ifndef _STDLIB_H
+#      define _STDLIB_H 1
+#     endif
+#    endif
+#   endif
+#  endif
+# endif
+
+# ifdef YYSTACK_ALLOC
+   /* Pacify GCC's `empty if-body' warning.  */
+#  define YYSTACK_FREE(Ptr) do { /* empty */; } while (YYID (0))
+#  ifndef YYSTACK_ALLOC_MAXIMUM
+    /* The OS might guarantee only one guard page at the bottom of the stack,
+       and a page size can be as small as 4096 bytes.  So we cannot safely
+       invoke alloca (N) if N exceeds 4096.  Use a slightly smaller number
+       to allow for a few compiler-allocated temporary stack slots.  */
+#   define YYSTACK_ALLOC_MAXIMUM 4032 /* reasonable circa 2006 */
+#  endif
+# else
+#  define YYSTACK_ALLOC YYMALLOC
+#  define YYSTACK_FREE YYFREE
+#  ifndef YYSTACK_ALLOC_MAXIMUM
+#   define YYSTACK_ALLOC_MAXIMUM YYSIZE_MAXIMUM
+#  endif
+#  if (defined __cplusplus && ! defined _STDLIB_H \
+       && ! ((defined YYMALLOC || defined malloc) \
+	     && (defined YYFREE || defined free)))
+#   include <stdlib.h> /* INFRINGES ON USER NAME SPACE */
+#   ifndef _STDLIB_H
+#    define _STDLIB_H 1
+#   endif
+#  endif
+#  ifndef YYMALLOC
+#   define YYMALLOC malloc
+#   if ! defined malloc && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+void *malloc (YYSIZE_T); /* INFRINGES ON USER NAME SPACE */
+#   endif
+#  endif
+#  ifndef YYFREE
+#   define YYFREE free
+#   if ! defined free && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+void free (void *); /* INFRINGES ON USER NAME SPACE */
+#   endif
+#  endif
+# endif
+#endif /* ! defined yyoverflow || YYERROR_VERBOSE */
+
+
+#if (! defined yyoverflow \
+     && (! defined __cplusplus \
+	 || (defined YYLTYPE_IS_TRIVIAL && YYLTYPE_IS_TRIVIAL \
+	     && defined YYSTYPE_IS_TRIVIAL && YYSTYPE_IS_TRIVIAL)))
+
+/* A type that is properly aligned for any stack member.  */
+union yyalloc
+{
+  yytype_int16 yyss;
+  YYSTYPE yyvs;
+    YYLTYPE yyls;
+};
+
+/* The size of the maximum gap between one aligned stack and the next.  */
+# define YYSTACK_GAP_MAXIMUM (sizeof (union yyalloc) - 1)
+
+/* The size of an array large to enough to hold all stacks, each with
+   N elements.  */
+# define YYSTACK_BYTES(N) \
+     ((N) * (sizeof (yytype_int16) + sizeof (YYSTYPE) + sizeof (YYLTYPE)) \
+      + 2 * YYSTACK_GAP_MAXIMUM)
+
+/* Copy COUNT objects from FROM to TO.  The source and destination do
+   not overlap.  */
+# ifndef YYCOPY
+#  if defined __GNUC__ && 1 < __GNUC__
+#   define YYCOPY(To, From, Count) \
+      __builtin_memcpy (To, From, (Count) * sizeof (*(From)))
+#  else
+#   define YYCOPY(To, From, Count)		\
+      do					\
+	{					\
+	  YYSIZE_T yyi;				\
+	  for (yyi = 0; yyi < (Count); yyi++)	\
+	    (To)[yyi] = (From)[yyi];		\
+	}					\
+      while (YYID (0))
+#  endif
+# endif
+
+/* Relocate STACK from its old location to the new one.  The
+   local variables YYSIZE and YYSTACKSIZE give the old and new number of
+   elements in the stack, and YYPTR gives the new location of the
+   stack.  Advance YYPTR to a properly aligned location for the next
+   stack.  */
+# define YYSTACK_RELOCATE(Stack)					\
+    do									\
+      {									\
+	YYSIZE_T yynewbytes;						\
+	YYCOPY (&yyptr->Stack, Stack, yysize);				\
+	Stack = &yyptr->Stack;						\
+	yynewbytes = yystacksize * sizeof (*Stack) + YYSTACK_GAP_MAXIMUM; \
+	yyptr += yynewbytes / sizeof (*yyptr);				\
+      }									\
+    while (YYID (0))
+
+#endif
+
+/* YYFINAL -- State number of the termination state.  */
+#define YYFINAL  2
+/* YYLAST -- Last index in YYTABLE.  */
+#define YYLAST   79
+
+/* YYNTOKENS -- Number of terminals.  */
+#define YYNTOKENS  26
+/* YYNNTS -- Number of nonterminals.  */
+#define YYNNTS  16
+/* YYNRULES -- Number of rules.  */
+#define YYNRULES  56
+/* YYNRULES -- Number of states.  */
+#define YYNSTATES  83
+
+/* YYTRANSLATE(YYLEX) -- Bison symbol number corresponding to YYLEX.  */
+#define YYUNDEFTOK  2
+#define YYMAXUTOK   274
+
+#define YYTRANSLATE(YYX)						\
+  ((unsigned int) (YYX) <= YYMAXUTOK ? yytranslate[YYX] : YYUNDEFTOK)
+
+/* YYTRANSLATE[YYLEX] -- Bison symbol number corresponding to YYLEX.  */
+static const yytype_uint8 yytranslate[] =
+{
+       0,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,    25,    22,    21,    24,    23,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,    20,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     2,     2,     2,     2,
+       2,     2,     2,     2,     2,     2,     1,     2,     3,     4,
+       5,     6,     7,     8,     9,    10,    11,    12,    13,    14,
+      15,    16,    17,    18,    19
+};
+
+#if YYDEBUG
+/* YYPRHS[YYN] -- Index of the first RHS symbol of rule number YYN in
+   YYRHS.  */
+static const yytype_uint8 yyprhs[] =
+{
+       0,     0,     3,     4,     7,     9,    11,    13,    15,    17,
+      19,    21,    23,    25,    28,    33,    39,    46,    54,    57,
+      59,    61,    63,    66,    69,    73,    76,    80,    86,    88,
+      94,   100,   103,   108,   111,   113,   117,   120,   124,   128,
+     136,   139,   144,   147,   149,   153,   156,   159,   163,   165,
+     167,   169,   171,   173,   175,   177,   178
+};
+
+/* YYRHS -- A `-1'-separated list of the rules' RHS.  */
+static const yytype_int8 yyrhs[] =
+{
+      27,     0,    -1,    -1,    27,    28,    -1,    29,    -1,    30,
+      -1,    32,    -1,    33,    -1,    31,    -1,    36,    -1,    34,
+      -1,    35,    -1,    40,    -1,    13,     7,    -1,    13,    20,
+      13,    41,    -1,    13,    20,    13,    21,    13,    -1,    13,
+      20,    13,    20,    13,    41,    -1,    13,    20,    13,    20,
+      13,    21,    13,    -1,    14,    16,    -1,    14,    -1,     5,
+      -1,     4,    -1,     4,    22,    -1,    13,     4,    -1,    38,
+      13,     4,    -1,    19,     4,    -1,    13,    23,    13,    -1,
+      13,    23,    13,    23,    13,    -1,    17,    -1,    13,    21,
+       8,    21,    13,    -1,    13,    21,    13,    21,    13,    -1,
+       8,    13,    -1,     8,    13,    22,    13,    -1,    13,     8,
+      -1,    15,    -1,    13,     8,    13,    -1,    19,     8,    -1,
+      19,    13,     8,    -1,    17,    14,    17,    -1,    17,    14,
+      13,    20,    13,    20,    13,    -1,    17,    17,    -1,    10,
+      13,    24,    13,    -1,    37,     3,    -1,    37,    -1,    38,
+      13,    39,    -1,    13,    39,    -1,    19,    39,    -1,    19,
+      13,    39,    -1,    39,    -1,    21,    -1,    25,    -1,    11,
+      -1,    18,    -1,     9,    -1,    13,    -1,    -1,     7,    -1
+};
+
+/* YYRLINE[YYN] -- source line where rule number YYN was defined.  */
+static const yytype_uint16 yyrline[] =
+{
+       0,   225,   225,   226,   229,   232,   235,   238,   241,   244,
+     247,   251,   256,   259,   265,   271,   279,   285,   296,   300,
+     304,   310,   314,   318,   322,   326,   332,   336,   341,   346,
+     351,   356,   360,   365,   369,   374,   381,   385,   391,   400,
+     409,   419,   433,   438,   441,   444,   447,   450,   453,   458,
+     461,   466,   470,   474,   480,   498,   501
+};
+#endif
+
+#if YYDEBUG || YYERROR_VERBOSE || YYTOKEN_TABLE
+/* YYTNAME[SYMBOL-NUM] -- String name of the symbol SYMBOL-NUM.
+   First, the terminals, then, starting at YYNTOKENS, nonterminals.  */
+static const char *const yytname[] =
+{
+  "$end", "error", "$undefined", "tAGO", "tDAY", "tDAYZONE", "tID",
+  "tMERIDIAN", "tMONTH", "tMONTH_UNIT", "tSTARDATE", "tSEC_UNIT",
+  "tSNUMBER", "tUNUMBER", "tZONE", "tEPOCH", "tDST", "tISOBASE",
+  "tDAY_UNIT", "tNEXT", "':'", "'-'", "','", "'/'", "'.'", "'+'",
+  "$accept", "spec", "item", "time", "zone", "day", "date", "ordMonth",
+  "iso", "trek", "relspec", "relunits", "sign", "unit", "number",
+  "o_merid", 0
+};
+#endif
+
+# ifdef YYPRINT
+/* YYTOKNUM[YYLEX-NUM] -- Internal token number corresponding to
+   token YYLEX-NUM.  */
+static const yytype_uint16 yytoknum[] =
+{
+       0,   256,   257,   258,   259,   260,   261,   262,   263,   264,
+     265,   266,   267,   268,   269,   270,   271,   272,   273,   274,
+      58,    45,    44,    47,    46,    43
+};
+# endif
+
+/* YYR1[YYN] -- Symbol number of symbol that rule YYN derives.  */
+static const yytype_uint8 yyr1[] =
+{
+       0,    26,    27,    27,    28,    28,    28,    28,    28,    28,
+      28,    28,    28,    29,    29,    29,    29,    29,    30,    30,
+      30,    31,    31,    31,    31,    31,    32,    32,    32,    32,
+      32,    32,    32,    32,    32,    32,    33,    33,    34,    34,
+      34,    35,    36,    36,    37,    37,    37,    37,    37,    38,
+      38,    39,    39,    39,    40,    41,    41
+};
+
+/* YYR2[YYN] -- Number of symbols composing right hand side of rule YYN.  */
+static const yytype_uint8 yyr2[] =
+{
+       0,     2,     0,     2,     1,     1,     1,     1,     1,     1,
+       1,     1,     1,     2,     4,     5,     6,     7,     2,     1,
+       1,     1,     2,     2,     3,     2,     3,     5,     1,     5,
+       5,     2,     4,     2,     1,     3,     2,     3,     3,     7,
+       2,     4,     2,     1,     3,     2,     2,     3,     1,     1,
+       1,     1,     1,     1,     1,     0,     1
+};
+
+/* YYDEFACT[STATE-NAME] -- Default rule to reduce with in state
+   STATE-NUM when YYTABLE doesn't specify something else to do.  Zero
+   means the default is an error.  */
+static const yytype_uint8 yydefact[] =
+{
+       2,     0,     1,    21,    20,     0,    53,     0,    51,    54,
+      19,    34,    28,    52,     0,    49,    50,     3,     4,     5,
+       8,     6,     7,    10,    11,     9,    43,     0,    48,    12,
+      22,    31,     0,    23,    13,    33,     0,     0,     0,    45,
+      18,     0,    40,    25,    36,     0,    46,    42,     0,     0,
+       0,    35,    55,     0,     0,    26,     0,    38,    37,    47,
+      24,    44,    32,    41,    56,     0,     0,    14,     0,     0,
+       0,     0,    55,    15,    29,    30,    27,     0,     0,    16,
+       0,    17,    39
+};
+
+/* YYDEFGOTO[NTERM-NUM].  */
+static const yytype_int8 yydefgoto[] =
+{
+      -1,     1,    17,    18,    19,    20,    21,    22,    23,    24,
+      25,    26,    27,    28,    29,    67
+};
+
+/* YYPACT[STATE-NUM] -- Index in YYTABLE of the portion describing
+   STATE-NUM.  */
+#define YYPACT_NINF -22
+static const yytype_int8 yypact[] =
+{
+     -22,     2,   -22,   -21,   -22,    -4,   -22,     1,   -22,    22,
+      18,   -22,     8,   -22,    40,   -22,   -22,   -22,   -22,   -22,
+     -22,   -22,   -22,   -22,   -22,   -22,    32,    28,   -22,   -22,
+     -22,    24,    26,   -22,   -22,    42,    47,    -5,    49,   -22,
+     -22,    15,   -22,   -22,   -22,    48,   -22,   -22,    43,    50,
+      51,   -22,    17,    44,    46,    45,    52,   -22,   -22,   -22,
+     -22,   -22,   -22,   -22,   -22,    56,    57,   -22,    58,    60,
+      61,    62,    -3,   -22,   -22,   -22,   -22,    59,    63,   -22,
+      64,   -22,   -22
+};
+
+/* YYPGOTO[NTERM-NUM].  */
+static const yytype_int8 yypgoto[] =
+{
+     -22,   -22,   -22,   -22,   -22,   -22,   -22,   -22,   -22,   -22,
+     -22,   -22,   -22,    -9,   -22,     6
+};
+
+/* YYTABLE[YYPACT[STATE-NUM]].  What to do in state STATE-NUM.  If
+   positive, shift that token.  If negative, reduce the rule which
+   number is the opposite.  If zero, do what YYDEFACT says.
+   If YYTABLE_NINF, syntax error.  */
+#define YYTABLE_NINF -1
+static const yytype_uint8 yytable[] =
+{
+      39,    30,     2,    53,    64,    46,     3,     4,    54,    31,
+       5,     6,     7,     8,    32,     9,    10,    11,    78,    12,
+      13,    14,    41,    15,    64,    42,    33,    16,    56,    34,
+      35,     6,    57,     8,    40,    47,    59,    65,    66,    61,
+      13,    48,    36,    37,    43,    38,    49,    60,    44,     6,
+      50,     8,     6,    45,     8,    51,    58,     6,    13,     8,
+      52,    13,    55,    62,    63,    68,    13,    69,    70,    72,
+      73,    74,    71,    75,    76,    77,    81,    82,    79,    80
+};
+
+static const yytype_uint8 yycheck[] =
+{
+       9,    22,     0,     8,     7,    14,     4,     5,    13,    13,
+       8,     9,    10,    11,    13,    13,    14,    15,    21,    17,
+      18,    19,    14,    21,     7,    17,     4,    25,    13,     7,
+       8,     9,    17,    11,    16,     3,    45,    20,    21,    48,
+      18,    13,    20,    21,     4,    23,    22,     4,     8,     9,
+      24,    11,     9,    13,    11,    13,     8,     9,    18,    11,
+      13,    18,    13,    13,    13,    21,    18,    21,    23,    13,
+      13,    13,    20,    13,    13,    13,    13,    13,    72,    20
+};
+
+/* YYSTOS[STATE-NUM] -- The (internal number of the) accessing
+   symbol of state STATE-NUM.  */
+static const yytype_uint8 yystos[] =
+{
+       0,    27,     0,     4,     5,     8,     9,    10,    11,    13,
+      14,    15,    17,    18,    19,    21,    25,    28,    29,    30,
+      31,    32,    33,    34,    35,    36,    37,    38,    39,    40,
+      22,    13,    13,     4,     7,     8,    20,    21,    23,    39,
+      16,    14,    17,     4,     8,    13,    39,     3,    13,    22,
+      24,    13,    13,     8,    13,    13,    13,    17,     8,    39,
+       4,    39,    13,    13,     7,    20,    21,    41,    21,    21,
+      23,    20,    13,    13,    13,    13,    13,    13,    21,    41,
+      20,    13,    13
+};
+
+#define yyerrok		(yyerrstatus = 0)
+#define yyclearin	(yychar = YYEMPTY)
+#define YYEMPTY		(-2)
+#define YYEOF		0
+
+#define YYACCEPT	goto yyacceptlab
+#define YYABORT		goto yyabortlab
+#define YYERROR		goto yyerrorlab
+
+
+/* Like YYERROR except do call yyerror.  This remains here temporarily
+   to ease the transition to the new meaning of YYERROR, for GCC.
+   Once GCC version 2 has supplanted version 1, this can go.  */
+
+#define YYFAIL		goto yyerrlab
+
+#define YYRECOVERING()  (!!yyerrstatus)
+
+#define YYBACKUP(Token, Value)					\
+do								\
+  if (yychar == YYEMPTY && yylen == 1)				\
+    {								\
+      yychar = (Token);						\
+      yylval = (Value);						\
+      yytoken = YYTRANSLATE (yychar);				\
+      YYPOPSTACK (1);						\
+      goto yybackup;						\
+    }								\
+  else								\
+    {								\
+      yyerror (&yylloc, info, YY_("syntax error: cannot back up")); \
+      YYERROR;							\
+    }								\
+while (YYID (0))
+
+
+#define YYTERROR	1
+#define YYERRCODE	256
+
+
+/* YYLLOC_DEFAULT -- Set CURRENT to span from RHS[1] to RHS[N].
+   If N is 0, then set CURRENT to the empty location which ends
+   the previous symbol: RHS[0] (always defined).  */
+
+#define YYRHSLOC(Rhs, K) ((Rhs)[K])
+#ifndef YYLLOC_DEFAULT
+# define YYLLOC_DEFAULT(Current, Rhs, N)				\
+    do									\
+      if (YYID (N))                                                    \
+	{								\
+	  (Current).first_line   = YYRHSLOC (Rhs, 1).first_line;	\
+	  (Current).first_column = YYRHSLOC (Rhs, 1).first_column;	\
+	  (Current).last_line    = YYRHSLOC (Rhs, N).last_line;		\
+	  (Current).last_column  = YYRHSLOC (Rhs, N).last_column;	\
+	}								\
+      else								\
+	{								\
+	  (Current).first_line   = (Current).last_line   =		\
+	    YYRHSLOC (Rhs, 0).last_line;				\
+	  (Current).first_column = (Current).last_column =		\
+	    YYRHSLOC (Rhs, 0).last_column;				\
+	}								\
+    while (YYID (0))
+#endif
+
+
+/* YY_LOCATION_PRINT -- Print the location on the stream.
+   This macro was not mandated originally: define only if we know
+   we won't break user code: when these are the locations we know.  */
+
+#ifndef YY_LOCATION_PRINT
+# if YYLTYPE_IS_TRIVIAL
+#  define YY_LOCATION_PRINT(File, Loc)			\
+     fprintf (File, "%d.%d-%d.%d",			\
+	      (Loc).first_line, (Loc).first_column,	\
+	      (Loc).last_line,  (Loc).last_column)
+# else
+#  define YY_LOCATION_PRINT(File, Loc) ((void) 0)
+# endif
+#endif
+
+
+/* YYLEX -- calling `yylex' with the right arguments.  */
+
+#ifdef YYLEX_PARAM
+# define YYLEX yylex (&yylval, &yylloc, YYLEX_PARAM)
+#else
+# define YYLEX yylex (&yylval, &yylloc, info)
+#endif
+
+/* Enable debugging if requested.  */
+#if YYDEBUG
+
+# ifndef YYFPRINTF
+#  include <stdio.h> /* INFRINGES ON USER NAME SPACE */
+#  define YYFPRINTF fprintf
+# endif
+
+# define YYDPRINTF(Args)			\
+do {						\
+  if (yydebug)					\
+    YYFPRINTF Args;				\
+} while (YYID (0))
+
+# define YY_SYMBOL_PRINT(Title, Type, Value, Location)			  \
+do {									  \
+  if (yydebug)								  \
+    {									  \
+      YYFPRINTF (stderr, "%s ", Title);					  \
+      yy_symbol_print (stderr,						  \
+		  Type, Value, Location, info); \
+      YYFPRINTF (stderr, "\n");						  \
+    }									  \
+} while (YYID (0))
+
+
+/*--------------------------------.
+| Print this symbol on YYOUTPUT.  |
+`--------------------------------*/
+
+/*ARGSUSED*/
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static void
+yy_symbol_value_print (FILE *yyoutput, int yytype, YYSTYPE const * const yyvaluep, YYLTYPE const * const yylocationp, DateInfo* info)
+#else
+static void
+yy_symbol_value_print (yyoutput, yytype, yyvaluep, yylocationp, info)
+    FILE *yyoutput;
+    int yytype;
+    YYSTYPE const * const yyvaluep;
+    YYLTYPE const * const yylocationp;
+    DateInfo* info;
+#endif
+{
+  if (!yyvaluep)
+    return;
+  YYUSE (yylocationp);
+  YYUSE (info);
+# ifdef YYPRINT
+  if (yytype < YYNTOKENS)
+    YYPRINT (yyoutput, yytoknum[yytype], *yyvaluep);
+# else
+  YYUSE (yyoutput);
+# endif
+  switch (yytype)
+    {
+      default:
+	break;
+    }
+}
+
+
+/*--------------------------------.
+| Print this symbol on YYOUTPUT.  |
+`--------------------------------*/
+
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static void
+yy_symbol_print (FILE *yyoutput, int yytype, YYSTYPE const * const yyvaluep, YYLTYPE const * const yylocationp, DateInfo* info)
+#else
+static void
+yy_symbol_print (yyoutput, yytype, yyvaluep, yylocationp, info)
+    FILE *yyoutput;
+    int yytype;
+    YYSTYPE const * const yyvaluep;
+    YYLTYPE const * const yylocationp;
+    DateInfo* info;
+#endif
+{
+  if (yytype < YYNTOKENS)
+    YYFPRINTF (yyoutput, "token %s (", yytname[yytype]);
+  else
+    YYFPRINTF (yyoutput, "nterm %s (", yytname[yytype]);
+
+  YY_LOCATION_PRINT (yyoutput, *yylocationp);
+  YYFPRINTF (yyoutput, ": ");
+  yy_symbol_value_print (yyoutput, yytype, yyvaluep, yylocationp, info);
+  YYFPRINTF (yyoutput, ")");
+}
+
+/*------------------------------------------------------------------.
+| yy_stack_print -- Print the state stack from its BOTTOM up to its |
+| TOP (included).                                                   |
+`------------------------------------------------------------------*/
+
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static void
+yy_stack_print (yytype_int16 *bottom, yytype_int16 *top)
+#else
+static void
+yy_stack_print (bottom, top)
+    yytype_int16 *bottom;
+    yytype_int16 *top;
+#endif
+{
+  YYFPRINTF (stderr, "Stack now");
+  for (; bottom <= top; ++bottom)
+    YYFPRINTF (stderr, " %d", *bottom);
+  YYFPRINTF (stderr, "\n");
+}
+
+# define YY_STACK_PRINT(Bottom, Top)				\
+do {								\
+  if (yydebug)							\
+    yy_stack_print ((Bottom), (Top));				\
+} while (YYID (0))
+
+
+/*------------------------------------------------.
+| Report that the YYRULE is going to be reduced.  |
+`------------------------------------------------*/
+
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static void
+yy_reduce_print (YYSTYPE *yyvsp, YYLTYPE *yylsp, int yyrule, DateInfo* info)
+#else
+static void
+yy_reduce_print (yyvsp, yylsp, yyrule, info)
+    YYSTYPE *yyvsp;
+    YYLTYPE *yylsp;
+    int yyrule;
+    DateInfo* info;
+#endif
+{
+  int yynrhs = yyr2[yyrule];
+  int yyi;
+  unsigned long int yylno = yyrline[yyrule];
+  YYFPRINTF (stderr, "Reducing stack by rule %d (line %lu):\n",
+	     yyrule - 1, yylno);
+  /* The symbols being reduced.  */
+  for (yyi = 0; yyi < yynrhs; yyi++)
+    {
+      fprintf (stderr, "   $%d = ", yyi + 1);
+      yy_symbol_print (stderr, yyrhs[yyprhs[yyrule] + yyi],
+		       &(yyvsp[(yyi + 1) - (yynrhs)])
+		       , &(yylsp[(yyi + 1) - (yynrhs)])		       , info);
+      fprintf (stderr, "\n");
+    }
+}
+
+# define YY_REDUCE_PRINT(Rule)		\
+do {					\
+  if (yydebug)				\
+    yy_reduce_print (yyvsp, yylsp, Rule, info); \
+} while (YYID (0))
+
+/* Nonzero means print parse trace.  It is left uninitialized so that
+   multiple parsers can coexist.  */
+int yydebug;
+#else /* !YYDEBUG */
+# define YYDPRINTF(Args)
+# define YY_SYMBOL_PRINT(Title, Type, Value, Location)
+# define YY_STACK_PRINT(Bottom, Top)
+# define YY_REDUCE_PRINT(Rule)
+#endif /* !YYDEBUG */
+
+
+/* YYINITDEPTH -- initial size of the parser's stacks.  */
+#ifndef	YYINITDEPTH
+# define YYINITDEPTH 200
+#endif
+
+/* YYMAXDEPTH -- maximum size the stacks can grow to (effective only
+   if the built-in stack extension method is used).
+
+   Do not make this value too large; the results are undefined if
+   YYSTACK_ALLOC_MAXIMUM < YYSTACK_BYTES (YYMAXDEPTH)
+   evaluated with infinite-precision integer arithmetic.  */
+
+#ifndef YYMAXDEPTH
+# define YYMAXDEPTH 10000
+#endif
+
+
+
+#if YYERROR_VERBOSE
+
+# ifndef yystrlen
+#  if defined __GLIBC__ && defined _STRING_H
+#   define yystrlen strlen
+#  else
+/* Return the length of YYSTR.  */
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static YYSIZE_T
+yystrlen (const char *yystr)
+#else
+static YYSIZE_T
+yystrlen (yystr)
+    const char *yystr;
+#endif
+{
+  YYSIZE_T yylen;
+  for (yylen = 0; yystr[yylen]; yylen++)
+    continue;
+  return yylen;
+}
+#  endif
+# endif
+
+# ifndef yystpcpy
+#  if defined __GLIBC__ && defined _STRING_H && defined _GNU_SOURCE
+#   define yystpcpy stpcpy
+#  else
+/* Copy YYSRC to YYDEST, returning the address of the terminating '\0' in
+   YYDEST.  */
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static char *
+yystpcpy (char *yydest, const char *yysrc)
+#else
+static char *
+yystpcpy (yydest, yysrc)
+    char *yydest;
+    const char *yysrc;
+#endif
+{
+  char *yyd = yydest;
+  const char *yys = yysrc;
+
+  while ((*yyd++ = *yys++) != '\0')
+    continue;
+
+  return yyd - 1;
+}
+#  endif
+# endif
+
+# ifndef yytnamerr
+/* Copy to YYRES the contents of YYSTR after stripping away unnecessary
+   quotes and backslashes, so that it's suitable for yyerror.  The
+   heuristic is that double-quoting is unnecessary unless the string
+   contains an apostrophe, a comma, or backslash (other than
+   backslash-backslash).  YYSTR is taken from yytname.  If YYRES is
+   null, do not copy; instead, return the length of what the result
+   would have been.  */
+static YYSIZE_T
+yytnamerr (char *yyres, const char *yystr)
+{
+  if (*yystr == '"')
+    {
+      YYSIZE_T yyn = 0;
+      char const *yyp = yystr;
+
+      for (;;)
+	switch (*++yyp)
+	  {
+	  case '\'':
+	  case ',':
+	    goto do_not_strip_quotes;
+
+	  case '\\':
+	    if (*++yyp != '\\')
+	      goto do_not_strip_quotes;
+	    /* Fall through.  */
+	  default:
+	    if (yyres)
+	      yyres[yyn] = *yyp;
+	    yyn++;
+	    break;
+
+	  case '"':
+	    if (yyres)
+	      yyres[yyn] = '\0';
+	    return yyn;
+	  }
+    do_not_strip_quotes: ;
+    }
+
+  if (! yyres)
+    return yystrlen (yystr);
+
+  return yystpcpy (yyres, yystr) - yyres;
+}
+# endif
+
+/* Copy into YYRESULT an error message about the unexpected token
+   YYCHAR while in state YYSTATE.  Return the number of bytes copied,
+   including the terminating null byte.  If YYRESULT is null, do not
+   copy anything; just return the number of bytes that would be
+   copied.  As a special case, return 0 if an ordinary "syntax error"
+   message will do.  Return YYSIZE_MAXIMUM if overflow occurs during
+   size calculation.  */
+static YYSIZE_T
+yysyntax_error (char *yyresult, int yystate, int yychar)
+{
+  int yyn = yypact[yystate];
+
+  if (! (YYPACT_NINF < yyn && yyn <= YYLAST))
+    return 0;
+  else
+    {
+      int yytype = YYTRANSLATE (yychar);
+      YYSIZE_T yysize0 = yytnamerr (0, yytname[yytype]);
+      YYSIZE_T yysize = yysize0;
+      YYSIZE_T yysize1;
+      int yysize_overflow = 0;
+      enum { YYERROR_VERBOSE_ARGS_MAXIMUM = 5 };
+      char const *yyarg[YYERROR_VERBOSE_ARGS_MAXIMUM];
+      int yyx;
+
+# if 0
+      /* This is so xgettext sees the translatable formats that are
+	 constructed on the fly.  */
+      YY_("syntax error, unexpected %s");
+      YY_("syntax error, unexpected %s, expecting %s");
+      YY_("syntax error, unexpected %s, expecting %s or %s");
+      YY_("syntax error, unexpected %s, expecting %s or %s or %s");
+      YY_("syntax error, unexpected %s, expecting %s or %s or %s or %s");
+# endif
+      char *yyfmt;
+      char const *yyf;
+      static char const yyunexpected[] = "syntax error, unexpected %s";
+      static char const yyexpecting[] = ", expecting %s";
+      static char const yyor[] = " or %s";
+      char yyformat[sizeof yyunexpected
+		    + sizeof yyexpecting - 1
+		    + ((YYERROR_VERBOSE_ARGS_MAXIMUM - 2)
+		       * (sizeof yyor - 1))];
+      char const *yyprefix = yyexpecting;
+
+      /* Start YYX at -YYN if negative to avoid negative indexes in
+	 YYCHECK.  */
+      int yyxbegin = yyn < 0 ? -yyn : 0;
+
+      /* Stay within bounds of both yycheck and yytname.  */
+      int yychecklim = YYLAST - yyn + 1;
+      int yyxend = yychecklim < YYNTOKENS ? yychecklim : YYNTOKENS;
+      int yycount = 1;
+
+      yyarg[0] = yytname[yytype];
+      yyfmt = yystpcpy (yyformat, yyunexpected);
+
+      for (yyx = yyxbegin; yyx < yyxend; ++yyx)
+	if (yycheck[yyx + yyn] == yyx && yyx != YYTERROR)
+	  {
+	    if (yycount == YYERROR_VERBOSE_ARGS_MAXIMUM)
+	      {
+		yycount = 1;
+		yysize = yysize0;
+		yyformat[sizeof yyunexpected - 1] = '\0';
+		break;
+	      }
+	    yyarg[yycount++] = yytname[yyx];
+	    yysize1 = yysize + yytnamerr (0, yytname[yyx]);
+	    yysize_overflow |= (yysize1 < yysize);
+	    yysize = yysize1;
+	    yyfmt = yystpcpy (yyfmt, yyprefix);
+	    yyprefix = yyor;
+	  }
+
+      yyf = YY_(yyformat);
+      yysize1 = yysize + yystrlen (yyf);
+      yysize_overflow |= (yysize1 < yysize);
+      yysize = yysize1;
+
+      if (yysize_overflow)
+	return YYSIZE_MAXIMUM;
+
+      if (yyresult)
+	{
+	  /* Avoid sprintf, as that infringes on the user's name space.
+	     Don't have undefined behavior even if the translation
+	     produced a string with the wrong number of "%s"s.  */
+	  char *yyp = yyresult;
+	  int yyi = 0;
+	  while ((*yyp = *yyf) != '\0')
+	    {
+	      if (*yyp == '%' && yyf[1] == 's' && yyi < yycount)
+		{
+		  yyp += yytnamerr (yyp, yyarg[yyi++]);
+		  yyf += 2;
+		}
+	      else
+		{
+		  yyp++;
+		  yyf++;
+		}
+	    }
+	}
+      return yysize;
+    }
+}
+#endif /* YYERROR_VERBOSE */
+
+
+/*-----------------------------------------------.
+| Release the memory associated to this symbol.  |
+`-----------------------------------------------*/
+
+/*ARGSUSED*/
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+static void
+yydestruct (const char *yymsg, int yytype, YYSTYPE *yyvaluep, YYLTYPE *yylocationp, DateInfo* info)
+#else
+static void
+yydestruct (yymsg, yytype, yyvaluep, yylocationp, info)
+    const char *yymsg;
+    int yytype;
+    YYSTYPE *yyvaluep;
+    YYLTYPE *yylocationp;
+    DateInfo* info;
+#endif
+{
+  YYUSE (yyvaluep);
+  YYUSE (yylocationp);
+  YYUSE (info);
+
+  if (!yymsg)
+    yymsg = "Deleting";
+  YY_SYMBOL_PRINT (yymsg, yytype, yyvaluep, yylocationp);
+
+  switch (yytype)
+    {
+
+      default:
+	break;
+    }
+}
+
+
+/* Prevent warnings from -Wmissing-prototypes.  */
+
+#ifdef YYPARSE_PARAM
+#if defined __STDC__ || defined __cplusplus
+int yyparse (void *YYPARSE_PARAM);
+#else
+int yyparse ();
+#endif
+#else /* ! YYPARSE_PARAM */
+#if defined __STDC__ || defined __cplusplus
+int yyparse (DateInfo* info);
+#else
+int yyparse ();
+#endif
+#endif /* ! YYPARSE_PARAM */
+
+
+
+
+
+
+/*----------.
+| yyparse.  |
+`----------*/
+
+#ifdef YYPARSE_PARAM
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+int
+yyparse (void *YYPARSE_PARAM)
+#else
+int
+yyparse (YYPARSE_PARAM)
+    void *YYPARSE_PARAM;
+#endif
+#else /* ! YYPARSE_PARAM */
+#if (defined __STDC__ || defined __C99__FUNC__ \
+     || defined __cplusplus || defined _MSC_VER)
+int
+yyparse (DateInfo* info)
+#else
+int
+yyparse (info)
+    DateInfo* info;
+#endif
+#endif
+{
+  /* The look-ahead symbol.  */
+int yychar;
+
+/* The semantic value of the look-ahead symbol.  */
+YYSTYPE yylval;
+
+/* Number of syntax errors so far.  */
+int yynerrs;
+/* Location data for the look-ahead symbol.  */
+YYLTYPE yylloc;
+
+  int yystate;
+  int yyn;
+  int yyresult;
+  /* Number of tokens to shift before error messages enabled.  */
+  int yyerrstatus;
+  /* Look-ahead token as an internal (translated) token number.  */
+  int yytoken = 0;
+#if YYERROR_VERBOSE
+  /* Buffer for error messages, and its allocated size.  */
+  char yymsgbuf[128];
+  char *yymsg = yymsgbuf;
+  YYSIZE_T yymsg_alloc = sizeof yymsgbuf;
+#endif
+
+  /* Three stacks and their tools:
+     `yyss': related to states,
+     `yyvs': related to semantic values,
+     `yyls': related to locations.
+
+     Refer to the stacks thru separate pointers, to allow yyoverflow
+     to reallocate them elsewhere.  */
+
+  /* The state stack.  */
+  yytype_int16 yyssa[YYINITDEPTH];
+  yytype_int16 *yyss = yyssa;
+  yytype_int16 *yyssp;
+
+  /* The semantic value stack.  */
+  YYSTYPE yyvsa[YYINITDEPTH];
+  YYSTYPE *yyvs = yyvsa;
+  YYSTYPE *yyvsp;
+
+  /* The location stack.  */
+  YYLTYPE yylsa[YYINITDEPTH];
+  YYLTYPE *yyls = yylsa;
+  YYLTYPE *yylsp;
+  /* The locations where the error started and ended.  */
+  YYLTYPE yyerror_range[2];
+
+#define YYPOPSTACK(N)   (yyvsp -= (N), yyssp -= (N), yylsp -= (N))
+
+  YYSIZE_T yystacksize = YYINITDEPTH;
+
+  /* The variables used to return semantic value and location from the
+     action routines.  */
+  YYSTYPE yyval;
+  YYLTYPE yyloc;
+
+  /* The number of symbols on the RHS of the reduced rule.
+     Keep to zero when no symbol should be popped.  */
+  int yylen = 0;
+
+  YYDPRINTF ((stderr, "Starting parse\n"));
+
+  yystate = 0;
+  yyerrstatus = 0;
+  yynerrs = 0;
+  yychar = YYEMPTY;		/* Cause a token to be read.  */
+
+  /* Initialize stack pointers.
+     Waste one element of value and location stack
+     so that they stay on the same level as the state stack.
+     The wasted elements are never initialized.  */
+
+  yyssp = yyss;
+  yyvsp = yyvs;
+  yylsp = yyls;
+#if YYLTYPE_IS_TRIVIAL
+  /* Initialize the default location before parsing starts.  */
+  yylloc.first_line   = yylloc.last_line   = 1;
+  yylloc.first_column = yylloc.last_column = 0;
+#endif
+
+  goto yysetstate;
+
+/*------------------------------------------------------------.
+| yynewstate -- Push a new state, which is found in yystate.  |
+`------------------------------------------------------------*/
+ yynewstate:
+  /* In all cases, when you get here, the value and location stacks
+     have just been pushed.  So pushing a state here evens the stacks.  */
+  yyssp++;
+
+ yysetstate:
+  *yyssp = yystate;
+
+  if (yyss + yystacksize - 1 <= yyssp)
+    {
+      /* Get the current used size of the three stacks, in elements.  */
+      YYSIZE_T yysize = yyssp - yyss + 1;
+
+#ifdef yyoverflow
+      {
+	/* Give user a chance to reallocate the stack.  Use copies of
+	   these so that the &'s don't force the real ones into
+	   memory.  */
+	YYSTYPE *yyvs1 = yyvs;
+	yytype_int16 *yyss1 = yyss;
+	YYLTYPE *yyls1 = yyls;
+
+	/* Each stack pointer address is followed by the size of the
+	   data in use in that stack, in bytes.  This used to be a
+	   conditional around just the two extra args, but that might
+	   be undefined if yyoverflow is a macro.  */
+	yyoverflow (YY_("memory exhausted"),
+		    &yyss1, yysize * sizeof (*yyssp),
+		    &yyvs1, yysize * sizeof (*yyvsp),
+		    &yyls1, yysize * sizeof (*yylsp),
+		    &yystacksize);
+	yyls = yyls1;
+	yyss = yyss1;
+	yyvs = yyvs1;
+      }
+#else /* no yyoverflow */
+# ifndef YYSTACK_RELOCATE
+      goto yyexhaustedlab;
+# else
+      /* Extend the stack our own way.  */
+      if (YYMAXDEPTH <= yystacksize)
+	goto yyexhaustedlab;
+      yystacksize *= 2;
+      if (YYMAXDEPTH < yystacksize)
+	yystacksize = YYMAXDEPTH;
+
+      {
+	yytype_int16 *yyss1 = yyss;
+	union yyalloc *yyptr =
+	  (union yyalloc *) YYSTACK_ALLOC (YYSTACK_BYTES (yystacksize));
+	if (! yyptr)
+	  goto yyexhaustedlab;
+	YYSTACK_RELOCATE (yyss);
+	YYSTACK_RELOCATE (yyvs);
+	YYSTACK_RELOCATE (yyls);
+#  undef YYSTACK_RELOCATE
+	if (yyss1 != yyssa)
+	  YYSTACK_FREE (yyss1);
+      }
+# endif
+#endif /* no yyoverflow */
+
+      yyssp = yyss + yysize - 1;
+      yyvsp = yyvs + yysize - 1;
+      yylsp = yyls + yysize - 1;
+
+      YYDPRINTF ((stderr, "Stack size increased to %lu\n",
+		  (unsigned long int) yystacksize));
+
+      if (yyss + yystacksize - 1 <= yyssp)
+	YYABORT;
+    }
+
+  YYDPRINTF ((stderr, "Entering state %d\n", yystate));
+
+  goto yybackup;
+
+/*-----------.
+| yybackup.  |
+`-----------*/
+yybackup:
+
+  /* Do appropriate processing given the current state.  Read a
+     look-ahead token if we need one and don't already have one.  */
+
+  /* First try to decide what to do without reference to look-ahead token.  */
+  yyn = yypact[yystate];
+  if (yyn == YYPACT_NINF)
+    goto yydefault;
+
+  /* Not known => get a look-ahead token if don't already have one.  */
+
+  /* YYCHAR is either YYEMPTY or YYEOF or a valid look-ahead symbol.  */
+  if (yychar == YYEMPTY)
+    {
+      YYDPRINTF ((stderr, "Reading a token: "));
+      yychar = YYLEX;
+    }
+
+  if (yychar <= YYEOF)
+    {
+      yychar = yytoken = YYEOF;
+      YYDPRINTF ((stderr, "Now at end of input.\n"));
+    }
+  else
+    {
+      yytoken = YYTRANSLATE (yychar);
+      YY_SYMBOL_PRINT ("Next token is", yytoken, &yylval, &yylloc);
+    }
+
+  /* If the proper action on seeing token YYTOKEN is to reduce or to
+     detect an error, take that action.  */
+  yyn += yytoken;
+  if (yyn < 0 || YYLAST < yyn || yycheck[yyn] != yytoken)
+    goto yydefault;
+  yyn = yytable[yyn];
+  if (yyn <= 0)
+    {
+      if (yyn == 0 || yyn == YYTABLE_NINF)
+	goto yyerrlab;
+      yyn = -yyn;
+      goto yyreduce;
+    }
+
+  if (yyn == YYFINAL)
+    YYACCEPT;
+
+  /* Count tokens shifted since error; after three, turn off error
+     status.  */
+  if (yyerrstatus)
+    yyerrstatus--;
+
+  /* Shift the look-ahead token.  */
+  YY_SYMBOL_PRINT ("Shifting", yytoken, &yylval, &yylloc);
+
+  /* Discard the shifted token unless it is eof.  */
+  if (yychar != YYEOF)
+    yychar = YYEMPTY;
+
+  yystate = yyn;
+  *++yyvsp = yylval;
+  *++yylsp = yylloc;
+  goto yynewstate;
+
+
+/*-----------------------------------------------------------.
+| yydefault -- do the default action for the current state.  |
+`-----------------------------------------------------------*/
+yydefault:
+  yyn = yydefact[yystate];
+  if (yyn == 0)
+    goto yyerrlab;
+  goto yyreduce;
+
+
+/*-----------------------------.
+| yyreduce -- Do a reduction.  |
+`-----------------------------*/
+yyreduce:
+  /* yyn is the number of a rule to reduce with.  */
+  yylen = yyr2[yyn];
+
+  /* If YYLEN is nonzero, implement the default value of the action:
+     `$$ = $1'.
+
+     Otherwise, the following line sets YYVAL to garbage.
+     This behavior is undocumented and Bison
+     users should not rely upon it.  Assigning to YYVAL
+     unconditionally makes the parser a bit smaller, and it avoids a
+     GCC warning that YYVAL may be used uninitialized.  */
+  yyval = yyvsp[1-yylen];
+
+  /* Default location.  */
+  YYLLOC_DEFAULT (yyloc, (yylsp - yylen), yylen);
+  YY_REDUCE_PRINT (yyn);
+  switch (yyn)
+    {
+        case 4:
+
+    {
+	    yyHaveTime++;
+	;}
+    break;
+
+  case 5:
+
+    {
+	    yyHaveZone++;
+	;}
+    break;
+
+  case 6:
+
+    {
+	    yyHaveDate++;
+	;}
+    break;
+
+  case 7:
+
+    {
+	    yyHaveOrdinalMonth++;
+	;}
+    break;
+
+  case 8:
+
+    {
+	    yyHaveDay++;
+	;}
+    break;
+
+  case 9:
+
+    {
+	    yyHaveRel++;
+	;}
+    break;
+
+  case 10:
+
+    {
+	    yyHaveTime++;
+	    yyHaveDate++;
+	;}
+    break;
+
+  case 11:
+
+    {
+	    yyHaveTime++;
+	    yyHaveDate++;
+	    yyHaveRel++;
+	;}
+    break;
+
+  case 13:
+
+    {
+	    yyHour = (yyvsp[(1) - (2)].Number);
+	    yyMinutes = 0;
+	    yySeconds = 0;
+	    yyMeridian = (yyvsp[(2) - (2)].Meridian);
+	;}
+    break;
+
+  case 14:
+
+    {
+	    yyHour = (yyvsp[(1) - (4)].Number);
+	    yyMinutes = (yyvsp[(3) - (4)].Number);
+	    yySeconds = 0;
+	    yyMeridian = (yyvsp[(4) - (4)].Meridian);
+	;}
+    break;
+
+  case 15:
+
+    {
+	    yyHour = (yyvsp[(1) - (5)].Number);
+	    yyMinutes = (yyvsp[(3) - (5)].Number);
+	    yyMeridian = MER24;
+	    yyDSTmode = DSToff;
+	    yyTimezone = ((yyvsp[(5) - (5)].Number) % 100 + ((yyvsp[(5) - (5)].Number) / 100) * 60);
+	    ++yyHaveZone;
+	;}
+    break;
+
+  case 16:
+
+    {
+	    yyHour = (yyvsp[(1) - (6)].Number);
+	    yyMinutes = (yyvsp[(3) - (6)].Number);
+	    yySeconds = (yyvsp[(5) - (6)].Number);
+	    yyMeridian = (yyvsp[(6) - (6)].Meridian);
+	;}
+    break;
+
+  case 17:
+
+    {
+	    yyHour = (yyvsp[(1) - (7)].Number);
+	    yyMinutes = (yyvsp[(3) - (7)].Number);
+	    yySeconds = (yyvsp[(5) - (7)].Number);
+	    yyMeridian = MER24;
+	    yyDSTmode = DSToff;
+	    yyTimezone = ((yyvsp[(7) - (7)].Number) % 100 + ((yyvsp[(7) - (7)].Number) / 100) * 60);
+	    ++yyHaveZone;
+	;}
+    break;
+
+  case 18:
+
+    {
+	    yyTimezone = (yyvsp[(1) - (2)].Number);
+	    yyDSTmode = DSTon;
+	;}
+    break;
+
+  case 19:
+
+    {
+	    yyTimezone = (yyvsp[(1) - (1)].Number);
+	    yyDSTmode = DSToff;
+	;}
+    break;
+
+  case 20:
+
+    {
+	    yyTimezone = (yyvsp[(1) - (1)].Number);
+	    yyDSTmode = DSTon;
+	;}
+    break;
+
+  case 21:
+
+    {
+	    yyDayOrdinal = 1;
+	    yyDayNumber = (yyvsp[(1) - (1)].Number);
+	;}
+    break;
+
+  case 22:
+
+    {
+	    yyDayOrdinal = 1;
+	    yyDayNumber = (yyvsp[(1) - (2)].Number);
+	;}
+    break;
+
+  case 23:
+
+    {
+	    yyDayOrdinal = (yyvsp[(1) - (2)].Number);
+	    yyDayNumber = (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 24:
+
+    {
+	    yyDayOrdinal = (yyvsp[(1) - (3)].Number) * (yyvsp[(2) - (3)].Number);
+	    yyDayNumber = (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 25:
+
+    {
+	    yyDayOrdinal = 2;
+	    yyDayNumber = (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 26:
+
+    {
+	    yyMonth = (yyvsp[(1) - (3)].Number);
+	    yyDay = (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 27:
+
+    {
+	    yyMonth = (yyvsp[(1) - (5)].Number);
+	    yyDay = (yyvsp[(3) - (5)].Number);
+	    yyYear = (yyvsp[(5) - (5)].Number);
+	;}
+    break;
+
+  case 28:
+
+    {
+	    yyYear = (yyvsp[(1) - (1)].Number) / 10000;
+	    yyMonth = ((yyvsp[(1) - (1)].Number) % 10000)/100;
+	    yyDay = (yyvsp[(1) - (1)].Number) % 100;
+	;}
+    break;
+
+  case 29:
+
+    {
+	    yyDay = (yyvsp[(1) - (5)].Number);
+	    yyMonth = (yyvsp[(3) - (5)].Number);
+	    yyYear = (yyvsp[(5) - (5)].Number);
+	;}
+    break;
+
+  case 30:
+
+    {
+	    yyMonth = (yyvsp[(3) - (5)].Number);
+	    yyDay = (yyvsp[(5) - (5)].Number);
+	    yyYear = (yyvsp[(1) - (5)].Number);
+	;}
+    break;
+
+  case 31:
+
+    {
+	    yyMonth = (yyvsp[(1) - (2)].Number);
+	    yyDay = (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 32:
+
+    {
+	    yyMonth = (yyvsp[(1) - (4)].Number);
+	    yyDay = (yyvsp[(2) - (4)].Number);
+	    yyYear = (yyvsp[(4) - (4)].Number);
+	;}
+    break;
+
+  case 33:
+
+    {
+	    yyMonth = (yyvsp[(2) - (2)].Number);
+	    yyDay = (yyvsp[(1) - (2)].Number);
+	;}
+    break;
+
+  case 34:
+
+    {
+	    yyMonth = 1;
+	    yyDay = 1;
+	    yyYear = EPOCH;
+	;}
+    break;
+
+  case 35:
+
+    {
+	    yyMonth = (yyvsp[(2) - (3)].Number);
+	    yyDay = (yyvsp[(1) - (3)].Number);
+	    yyYear = (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 36:
+
+    {
+	    yyMonthOrdinal = 1;
+	    yyMonth = (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 37:
+
+    {
+	    yyMonthOrdinal = (yyvsp[(2) - (3)].Number);
+	    yyMonth = (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 38:
+
+    {
+	    if ((yyvsp[(2) - (3)].Number) != HOUR( 7)) YYABORT;
+	    yyYear = (yyvsp[(1) - (3)].Number) / 10000;
+	    yyMonth = ((yyvsp[(1) - (3)].Number) % 10000)/100;
+	    yyDay = (yyvsp[(1) - (3)].Number) % 100;
+	    yyHour = (yyvsp[(3) - (3)].Number) / 10000;
+	    yyMinutes = ((yyvsp[(3) - (3)].Number) % 10000)/100;
+	    yySeconds = (yyvsp[(3) - (3)].Number) % 100;
+	;}
+    break;
+
+  case 39:
+
+    {
+	    if ((yyvsp[(2) - (7)].Number) != HOUR( 7)) YYABORT;
+	    yyYear = (yyvsp[(1) - (7)].Number) / 10000;
+	    yyMonth = ((yyvsp[(1) - (7)].Number) % 10000)/100;
+	    yyDay = (yyvsp[(1) - (7)].Number) % 100;
+	    yyHour = (yyvsp[(3) - (7)].Number);
+	    yyMinutes = (yyvsp[(5) - (7)].Number);
+	    yySeconds = (yyvsp[(7) - (7)].Number);
+	;}
+    break;
+
+  case 40:
+
+    {
+	    yyYear = (yyvsp[(1) - (2)].Number) / 10000;
+	    yyMonth = ((yyvsp[(1) - (2)].Number) % 10000)/100;
+	    yyDay = (yyvsp[(1) - (2)].Number) % 100;
+	    yyHour = (yyvsp[(2) - (2)].Number) / 10000;
+	    yyMinutes = ((yyvsp[(2) - (2)].Number) % 10000)/100;
+	    yySeconds = (yyvsp[(2) - (2)].Number) % 100;
+	;}
+    break;
+
+  case 41:
+
+    {
+	    /*
+	     * Offset computed year by -377 so that the returned years will be
+	     * in a range accessible with a 32 bit clock seconds value.
+	     */
+
+	    yyYear = (yyvsp[(2) - (4)].Number)/1000 + 2323 - 377;
+	    yyDay  = 1;
+	    yyMonth = 1;
+	    yyRelDay += (((yyvsp[(2) - (4)].Number)%1000)*(365 + IsLeapYear(yyYear)))/1000;
+	    yyRelSeconds += (yyvsp[(4) - (4)].Number) * 144 * 60;
+	;}
+    break;
+
+  case 42:
+
+    {
+	    yyRelSeconds *= -1;
+	    yyRelMonth *= -1;
+	    yyRelDay *= -1;
+	;}
+    break;
+
+  case 44:
+
+    {
+	    *yyRelPointer += (yyvsp[(1) - (3)].Number) * (yyvsp[(2) - (3)].Number) * (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 45:
+
+    {
+	    *yyRelPointer += (yyvsp[(1) - (2)].Number) * (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 46:
+
+    {
+	    *yyRelPointer += (yyvsp[(2) - (2)].Number);
+	;}
+    break;
+
+  case 47:
+
+    {
+	    *yyRelPointer += (yyvsp[(2) - (3)].Number) * (yyvsp[(3) - (3)].Number);
+	;}
+    break;
+
+  case 48:
+
+    {
+	    *yyRelPointer += (yyvsp[(1) - (1)].Number);
+	;}
+    break;
+
+  case 49:
+
+    {
+	    (yyval.Number) = -1;
+	;}
+    break;
+
+  case 50:
+
+    {
+	    (yyval.Number) =  1;
+	;}
+    break;
+
+  case 51:
+
+    {
+	    (yyval.Number) = (yyvsp[(1) - (1)].Number);
+	    yyRelPointer = &yyRelSeconds;
+	;}
+    break;
+
+  case 52:
+
+    {
+	    (yyval.Number) = (yyvsp[(1) - (1)].Number);
+	    yyRelPointer = &yyRelDay;
+	;}
+    break;
+
+  case 53:
+
+    {
+	    (yyval.Number) = (yyvsp[(1) - (1)].Number);
+	    yyRelPointer = &yyRelMonth;
+	;}
+    break;
+
+  case 54:
+
+    {
+	    if (yyHaveTime && yyHaveDate && !yyHaveRel) {
+		yyYear = (yyvsp[(1) - (1)].Number);
+	    } else {
+		yyHaveTime++;
+		if (yyDigitCount <= 2) {
+		    yyHour = (yyvsp[(1) - (1)].Number);
+		    yyMinutes = 0;
+		} else {
+		    yyHour = (yyvsp[(1) - (1)].Number) / 100;
+		    yyMinutes = (yyvsp[(1) - (1)].Number) % 100;
+		}
+		yySeconds = 0;
+		yyMeridian = MER24;
+	    }
+	;}
+    break;
+
+  case 55:
+
+    {
+	    (yyval.Meridian) = MER24;
+	;}
+    break;
+
+  case 56:
+
+    {
+	    (yyval.Meridian) = (yyvsp[(1) - (1)].Meridian);
+	;}
+    break;
+
+
+/* Line 1267 of yacc.c.  */
+
+      default: break;
+    }
+  YY_SYMBOL_PRINT ("-> $$ =", yyr1[yyn], &yyval, &yyloc);
+
+  YYPOPSTACK (yylen);
+  yylen = 0;
+  YY_STACK_PRINT (yyss, yyssp);
+
+  *++yyvsp = yyval;
+  *++yylsp = yyloc;
+
+  /* Now `shift' the result of the reduction.  Determine what state
+     that goes to, based on the state we popped back to and the rule
+     number reduced by.  */
+
+  yyn = yyr1[yyn];
+
+  yystate = yypgoto[yyn - YYNTOKENS] + *yyssp;
+  if (0 <= yystate && yystate <= YYLAST && yycheck[yystate] == *yyssp)
+    yystate = yytable[yystate];
+  else
+    yystate = yydefgoto[yyn - YYNTOKENS];
+
+  goto yynewstate;
+
+
+/*------------------------------------.
+| yyerrlab -- here on detecting error |
+`------------------------------------*/
+yyerrlab:
+  /* If not already recovering from an error, report this error.  */
+  if (!yyerrstatus)
+    {
+      ++yynerrs;
+#if ! YYERROR_VERBOSE
+      yyerror (&yylloc, info, YY_("syntax error"));
+#else
+      {
+	YYSIZE_T yysize = yysyntax_error (0, yystate, yychar);
+	if (yymsg_alloc < yysize && yymsg_alloc < YYSTACK_ALLOC_MAXIMUM)
+	  {
+	    YYSIZE_T yyalloc = 2 * yysize;
+	    if (! (yysize <= yyalloc && yyalloc <= YYSTACK_ALLOC_MAXIMUM))
+	      yyalloc = YYSTACK_ALLOC_MAXIMUM;
+	    if (yymsg != yymsgbuf)
+	      YYSTACK_FREE (yymsg);
+	    yymsg = (char *) YYSTACK_ALLOC (yyalloc);
+	    if (yymsg)
+	      yymsg_alloc = yyalloc;
+	    else
+	      {
+		yymsg = yymsgbuf;
+		yymsg_alloc = sizeof yymsgbuf;
+	      }
+	  }
+
+	if (0 < yysize && yysize <= yymsg_alloc)
+	  {
+	    (void) yysyntax_error (yymsg, yystate, yychar);
+	    yyerror (&yylloc, info, yymsg);
+	  }
+	else
+	  {
+	    yyerror (&yylloc, info, YY_("syntax error"));
+	    if (yysize != 0)
+	      goto yyexhaustedlab;
+	  }
+      }
+#endif
+    }
+
+  yyerror_range[0] = yylloc;
+
+  if (yyerrstatus == 3)
+    {
+      /* If just tried and failed to reuse look-ahead token after an
+	 error, discard it.  */
+
+      if (yychar <= YYEOF)
+	{
+	  /* Return failure if at end of input.  */
+	  if (yychar == YYEOF)
+	    YYABORT;
+	}
+      else
+	{
+	  yydestruct ("Error: discarding",
+		      yytoken, &yylval, &yylloc, info);
+	  yychar = YYEMPTY;
+	}
+    }
+
+  /* Else will try to reuse look-ahead token after shifting the error
+     token.  */
+  goto yyerrlab1;
+
+
+/*---------------------------------------------------.
+| yyerrorlab -- error raised explicitly by YYERROR.  |
+`---------------------------------------------------*/
+yyerrorlab:
+
+  /* Pacify compilers like GCC when the user code never invokes
+     YYERROR and the label yyerrorlab therefore never appears in user
+     code.  */
+  if (/*CONSTCOND*/ 0)
+     goto yyerrorlab;
+
+  yyerror_range[0] = yylsp[1-yylen];
+  /* Do not reclaim the symbols of the rule which action triggered
+     this YYERROR.  */
+  YYPOPSTACK (yylen);
+  yylen = 0;
+  YY_STACK_PRINT (yyss, yyssp);
+  yystate = *yyssp;
+  goto yyerrlab1;
+
+
+/*-------------------------------------------------------------.
+| yyerrlab1 -- common code for both syntax error and YYERROR.  |
+`-------------------------------------------------------------*/
+yyerrlab1:
+  yyerrstatus = 3;	/* Each real token shifted decrements this.  */
+
+  for (;;)
+    {
+      yyn = yypact[yystate];
+      if (yyn != YYPACT_NINF)
+	{
+	  yyn += YYTERROR;
+	  if (0 <= yyn && yyn <= YYLAST && yycheck[yyn] == YYTERROR)
+	    {
+	      yyn = yytable[yyn];
+	      if (0 < yyn)
+		break;
+	    }
+	}
+
+      /* Pop the current state because it cannot handle the error token.  */
+      if (yyssp == yyss)
+	YYABORT;
+
+      yyerror_range[0] = *yylsp;
+      yydestruct ("Error: popping",
+		  yystos[yystate], yyvsp, yylsp, info);
+      YYPOPSTACK (1);
+      yystate = *yyssp;
+      YY_STACK_PRINT (yyss, yyssp);
+    }
+
+  if (yyn == YYFINAL)
+    YYACCEPT;
+
+  *++yyvsp = yylval;
+
+  yyerror_range[1] = yylloc;
+  /* Using YYLLOC is tempting, but would change the location of
+     the look-ahead.  YYLOC is available though.  */
+  YYLLOC_DEFAULT (yyloc, (yyerror_range - 1), 2);
+  *++yylsp = yyloc;
+
+  /* Shift the error token.  */
+  YY_SYMBOL_PRINT ("Shifting", yystos[yyn], yyvsp, yylsp);
+
+  yystate = yyn;
+  goto yynewstate;
+
+
+/*-------------------------------------.
+| yyacceptlab -- YYACCEPT comes here.  |
+`-------------------------------------*/
+yyacceptlab:
+  yyresult = 0;
+  goto yyreturn;
+
+/*-----------------------------------.
+| yyabortlab -- YYABORT comes here.  |
+`-----------------------------------*/
+yyabortlab:
+  yyresult = 1;
+  goto yyreturn;
+
+#ifndef yyoverflow
+/*-------------------------------------------------.
+| yyexhaustedlab -- memory exhaustion comes here.  |
+`-------------------------------------------------*/
+yyexhaustedlab:
+  yyerror (&yylloc, info, YY_("memory exhausted"));
+  yyresult = 2;
+  /* Fall through.  */
+#endif
+
+yyreturn:
+  if (yychar != YYEOF && yychar != YYEMPTY)
+     yydestruct ("Cleanup: discarding lookahead",
+		 yytoken, &yylval, &yylloc, info);
+  /* Do not reclaim the symbols of the rule which action triggered
+     this YYABORT or YYACCEPT.  */
+  YYPOPSTACK (yylen);
+  YY_STACK_PRINT (yyss, yyssp);
+  while (yyssp != yyss)
+    {
+      yydestruct ("Cleanup: popping",
+		  yystos[*yyssp], yyvsp, yylsp, info);
+      YYPOPSTACK (1);
+    }
+#ifndef yyoverflow
+  if (yyss != yyssa)
+    YYSTACK_FREE (yyss);
+#endif
+#if YYERROR_VERBOSE
+  if (yymsg != yymsgbuf)
+    YYSTACK_FREE (yymsg);
+#endif
+  /* Make sure YYID is used.  */
+  return YYID (yyresult);
+}
+
+
+
+
+/*
+ * Month and day table.
+ */
+
+static const TABLE MonthDayTable[] = {
+    { "january",	tMONTH,	 1 },
+    { "february",	tMONTH,	 2 },
+    { "march",		tMONTH,	 3 },
+    { "april",		tMONTH,	 4 },
+    { "may",		tMONTH,	 5 },
+    { "june",		tMONTH,	 6 },
+    { "july",		tMONTH,	 7 },
+    { "august",		tMONTH,	 8 },
+    { "september",	tMONTH,	 9 },
+    { "sept",		tMONTH,	 9 },
+    { "october",	tMONTH, 10 },
+    { "november",	tMONTH, 11 },
+    { "december",	tMONTH, 12 },
+    { "sunday",		tDAY, 0 },
+    { "monday",		tDAY, 1 },
+    { "tuesday",	tDAY, 2 },
+    { "tues",		tDAY, 2 },
+    { "wednesday",	tDAY, 3 },
+    { "wednes",		tDAY, 3 },
+    { "thursday",	tDAY, 4 },
+    { "thur",		tDAY, 4 },
+    { "thurs",		tDAY, 4 },
+    { "friday",		tDAY, 5 },
+    { "saturday",	tDAY, 6 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Time units table.
+ */
+
+static const TABLE UnitsTable[] = {
+    { "year",		tMONTH_UNIT,	12 },
+    { "month",		tMONTH_UNIT,	 1 },
+    { "fortnight",	tDAY_UNIT,	14 },
+    { "week",		tDAY_UNIT,	 7 },
+    { "day",		tDAY_UNIT,	 1 },
+    { "hour",		tSEC_UNIT, 60 * 60 },
+    { "minute",		tSEC_UNIT,	60 },
+    { "min",		tSEC_UNIT,	60 },
+    { "second",		tSEC_UNIT,	 1 },
+    { "sec",		tSEC_UNIT,	 1 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Assorted relative-time words.
+ */
+
+static const TABLE OtherTable[] = {
+    { "tomorrow",	tDAY_UNIT,	1 },
+    { "yesterday",	tDAY_UNIT,	-1 },
+    { "today",		tDAY_UNIT,	0 },
+    { "now",		tSEC_UNIT,	0 },
+    { "last",		tUNUMBER,	-1 },
+    { "this",		tSEC_UNIT,	0 },
+    { "next",		tNEXT,		1 },
+#if 0
+    { "first",		tUNUMBER,	1 },
+    { "second",		tUNUMBER,	2 },
+    { "third",		tUNUMBER,	3 },
+    { "fourth",		tUNUMBER,	4 },
+    { "fifth",		tUNUMBER,	5 },
+    { "sixth",		tUNUMBER,	6 },
+    { "seventh",	tUNUMBER,	7 },
+    { "eighth",		tUNUMBER,	8 },
+    { "ninth",		tUNUMBER,	9 },
+    { "tenth",		tUNUMBER,	10 },
+    { "eleventh",	tUNUMBER,	11 },
+    { "twelfth",	tUNUMBER,	12 },
+#endif
+    { "ago",		tAGO,		1 },
+    { "epoch",		tEPOCH,		0 },
+    { "stardate",	tSTARDATE,	0 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * The timezone table. (Note: This table was modified to not use any floating
+ * point constants to work around an SGI compiler bug).
+ */
+
+static const TABLE TimezoneTable[] = {
+    { "gmt",	tZONE,	   HOUR( 0) },	    /* Greenwich Mean */
+    { "ut",	tZONE,	   HOUR( 0) },	    /* Universal (Coordinated) */
+    { "utc",	tZONE,	   HOUR( 0) },
+    { "uct",	tZONE,	   HOUR( 0) },	    /* Universal Coordinated Time */
+    { "wet",	tZONE,	   HOUR( 0) },	    /* Western European */
+    { "bst",	tDAYZONE,  HOUR( 0) },	    /* British Summer */
+    { "wat",	tZONE,	   HOUR( 1) },	    /* West Africa */
+    { "at",	tZONE,	   HOUR( 2) },	    /* Azores */
+#if	0
+    /* For completeness.  BST is also British Summer, and GST is
+     * also Guam Standard. */
+    { "bst",	tZONE,	   HOUR( 3) },	    /* Brazil Standard */
+    { "gst",	tZONE,	   HOUR( 3) },	    /* Greenland Standard */
+#endif
+    { "nft",	tZONE,	   HOUR( 7/2) },    /* Newfoundland */
+    { "nst",	tZONE,	   HOUR( 7/2) },    /* Newfoundland Standard */
+    { "ndt",	tDAYZONE,  HOUR( 7/2) },    /* Newfoundland Daylight */
+    { "ast",	tZONE,	   HOUR( 4) },	    /* Atlantic Standard */
+    { "adt",	tDAYZONE,  HOUR( 4) },	    /* Atlantic Daylight */
+    { "est",	tZONE,	   HOUR( 5) },	    /* Eastern Standard */
+    { "edt",	tDAYZONE,  HOUR( 5) },	    /* Eastern Daylight */
+    { "cst",	tZONE,	   HOUR( 6) },	    /* Central Standard */
+    { "cdt",	tDAYZONE,  HOUR( 6) },	    /* Central Daylight */
+    { "mst",	tZONE,	   HOUR( 7) },	    /* Mountain Standard */
+    { "mdt",	tDAYZONE,  HOUR( 7) },	    /* Mountain Daylight */
+    { "pst",	tZONE,	   HOUR( 8) },	    /* Pacific Standard */
+    { "pdt",	tDAYZONE,  HOUR( 8) },	    /* Pacific Daylight */
+    { "yst",	tZONE,	   HOUR( 9) },	    /* Yukon Standard */
+    { "ydt",	tDAYZONE,  HOUR( 9) },	    /* Yukon Daylight */
+    { "hst",	tZONE,	   HOUR(10) },	    /* Hawaii Standard */
+    { "hdt",	tDAYZONE,  HOUR(10) },	    /* Hawaii Daylight */
+    { "cat",	tZONE,	   HOUR(10) },	    /* Central Alaska */
+    { "ahst",	tZONE,	   HOUR(10) },	    /* Alaska-Hawaii Standard */
+    { "nt",	tZONE,	   HOUR(11) },	    /* Nome */
+    { "idlw",	tZONE,	   HOUR(12) },	    /* International Date Line West */
+    { "cet",	tZONE,	  -HOUR( 1) },	    /* Central European */
+    { "cest",	tDAYZONE, -HOUR( 1) },	    /* Central European Summer */
+    { "met",	tZONE,	  -HOUR( 1) },	    /* Middle European */
+    { "mewt",	tZONE,	  -HOUR( 1) },	    /* Middle European Winter */
+    { "mest",	tDAYZONE, -HOUR( 1) },	    /* Middle European Summer */
+    { "swt",	tZONE,	  -HOUR( 1) },	    /* Swedish Winter */
+    { "sst",	tDAYZONE, -HOUR( 1) },	    /* Swedish Summer */
+    { "fwt",	tZONE,	  -HOUR( 1) },	    /* French Winter */
+    { "fst",	tDAYZONE, -HOUR( 1) },	    /* French Summer */
+    { "eet",	tZONE,	  -HOUR( 2) },	    /* Eastern Europe, USSR Zone 1 */
+    { "bt",	tZONE,	  -HOUR( 3) },	    /* Baghdad, USSR Zone 2 */
+    { "it",	tZONE,	  -HOUR( 7/2) },    /* Iran */
+    { "zp4",	tZONE,	  -HOUR( 4) },	    /* USSR Zone 3 */
+    { "zp5",	tZONE,	  -HOUR( 5) },	    /* USSR Zone 4 */
+    { "ist",	tZONE,	  -HOUR(11/2) },    /* Indian Standard */
+    { "zp6",	tZONE,	  -HOUR( 6) },	    /* USSR Zone 5 */
+#if	0
+    /* For completeness.  NST is also Newfoundland Stanard, nad SST is
+     * also Swedish Summer. */
+    { "nst",	tZONE,	  -HOUR(13/2) },    /* North Sumatra */
+    { "sst",	tZONE,	  -HOUR( 7) },	    /* South Sumatra, USSR Zone 6 */
+#endif	/* 0 */
+    { "wast",	tZONE,	  -HOUR( 7) },	    /* West Australian Standard */
+    { "wadt",	tDAYZONE, -HOUR( 7) },	    /* West Australian Daylight */
+    { "jt",	tZONE,	  -HOUR(15/2) },    /* Java (3pm in Cronusland!) */
+    { "cct",	tZONE,	  -HOUR( 8) },	    /* China Coast, USSR Zone 7 */
+    { "jst",	tZONE,	  -HOUR( 9) },	    /* Japan Standard, USSR Zone 8 */
+    { "jdt",	tDAYZONE, -HOUR( 9) },	    /* Japan Daylight */
+    { "kst",	tZONE,	  -HOUR( 9) },	    /* Korea Standard */
+    { "kdt",	tDAYZONE, -HOUR( 9) },	    /* Korea Daylight */
+    { "cast",	tZONE,	  -HOUR(19/2) },    /* Central Australian Standard */
+    { "cadt",	tDAYZONE, -HOUR(19/2) },    /* Central Australian Daylight */
+    { "east",	tZONE,	  -HOUR(10) },	    /* Eastern Australian Standard */
+    { "eadt",	tDAYZONE, -HOUR(10) },	    /* Eastern Australian Daylight */
+    { "gst",	tZONE,	  -HOUR(10) },	    /* Guam Standard, USSR Zone 9 */
+    { "nzt",	tZONE,	  -HOUR(12) },	    /* New Zealand */
+    { "nzst",	tZONE,	  -HOUR(12) },	    /* New Zealand Standard */
+    { "nzdt",	tDAYZONE, -HOUR(12) },	    /* New Zealand Daylight */
+    { "idle",	tZONE,	  -HOUR(12) },	    /* International Date Line East */
+    /* ADDED BY Marco Nijdam */
+    { "dst",	tDST,	  HOUR( 0) },	    /* DST on (hour is ignored) */
+    /* End ADDED */
+    { NULL, 0, 0 }
+};
+
+/*
+ * Military timezone table.
+ */
+
+static const TABLE MilitaryTable[] = {
+    { "a",	tZONE,	-HOUR( 1) },
+    { "b",	tZONE,	-HOUR( 2) },
+    { "c",	tZONE,	-HOUR( 3) },
+    { "d",	tZONE,	-HOUR( 4) },
+    { "e",	tZONE,	-HOUR( 5) },
+    { "f",	tZONE,	-HOUR( 6) },
+    { "g",	tZONE,	-HOUR( 7) },
+    { "h",	tZONE,	-HOUR( 8) },
+    { "i",	tZONE,	-HOUR( 9) },
+    { "k",	tZONE,	-HOUR(10) },
+    { "l",	tZONE,	-HOUR(11) },
+    { "m",	tZONE,	-HOUR(12) },
+    { "n",	tZONE,	HOUR(  1) },
+    { "o",	tZONE,	HOUR(  2) },
+    { "p",	tZONE,	HOUR(  3) },
+    { "q",	tZONE,	HOUR(  4) },
+    { "r",	tZONE,	HOUR(  5) },
+    { "s",	tZONE,	HOUR(  6) },
+    { "t",	tZONE,	HOUR(  7) },
+    { "u",	tZONE,	HOUR(  8) },
+    { "v",	tZONE,	HOUR(  9) },
+    { "w",	tZONE,	HOUR( 10) },
+    { "x",	tZONE,	HOUR( 11) },
+    { "y",	tZONE,	HOUR( 12) },
+    { "z",	tZONE,	HOUR( 0) },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Dump error messages in the bit bucket.
+ */
+
+static void
+TclDateerror(
+    YYLTYPE* location,
+    DateInfo* infoPtr,
+    const char *s)
+{
+    Tcl_Obj* t;
+    Tcl_AppendToObj(infoPtr->messages, infoPtr->separatrix, -1);
+    Tcl_AppendToObj(infoPtr->messages, s, -1);
+    Tcl_AppendToObj(infoPtr->messages, " (characters ", -1);
+    t = Tcl_NewIntObj(location->first_column);
+    Tcl_IncrRefCount(t);
+    Tcl_AppendObjToObj(infoPtr->messages, t);
+    Tcl_DecrRefCount(t);
+    Tcl_AppendToObj(infoPtr->messages, "-", -1);
+    t = Tcl_NewIntObj(location->last_column);
+    Tcl_IncrRefCount(t);
+    Tcl_AppendObjToObj(infoPtr->messages, t);
+    Tcl_DecrRefCount(t);
+    Tcl_AppendToObj(infoPtr->messages, ")", -1);
+    infoPtr->separatrix = "\n";
+}
+
+static time_t
+ToSeconds(
+    time_t Hours,
+    time_t Minutes,
+    time_t Seconds,
+    MERIDIAN Meridian)
+{
+    if (Minutes < 0 || Minutes > 59 || Seconds < 0 || Seconds > 59) {
+	return -1;
+    }
+    switch (Meridian) {
+    case MER24:
+	if (Hours < 0 || Hours > 23) {
+	    return -1;
+	}
+	return (Hours * 60L + Minutes) * 60L + Seconds;
+    case MERam:
+	if (Hours < 1 || Hours > 12) {
+	    return -1;
+	}
+	return ((Hours % 12) * 60L + Minutes) * 60L + Seconds;
+    case MERpm:
+	if (Hours < 1 || Hours > 12) {
+	    return -1;
+	}
+	return (((Hours % 12) + 12) * 60L + Minutes) * 60L + Seconds;
+    }
+    return -1;			/* Should never be reached */
+}
+
+static int
+LookupWord(
+    YYSTYPE* yylvalPtr,
+    char *buff)
+{
+    register char *p;
+    register char *q;
+    register const TABLE *tp;
+    int i, abbrev;
+
+    /*
+     * Make it lowercase.
+     */
+
+    Tcl_UtfToLower(buff);
+
+    if (strcmp(buff, "am") == 0 || strcmp(buff, "a.m.") == 0) {
+	yylvalPtr->Meridian = MERam;
+	return tMERIDIAN;
+    }
+    if (strcmp(buff, "pm") == 0 || strcmp(buff, "p.m.") == 0) {
+	yylvalPtr->Meridian = MERpm;
+	return tMERIDIAN;
+    }
+
+    /*
+     * See if we have an abbreviation for a month.
+     */
+
+    if (strlen(buff) == 3) {
+	abbrev = 1;
+    } else if (strlen(buff) == 4 && buff[3] == '.') {
+	abbrev = 1;
+	buff[3] = '\0';
+    } else {
+	abbrev = 0;
+    }
+
+    for (tp = MonthDayTable; tp->name; tp++) {
+	if (abbrev) {
+	    if (strncmp(buff, tp->name, 3) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	} else if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    for (tp = TimezoneTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    for (tp = UnitsTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    /*
+     * Strip off any plural and try the units table again.
+     */
+
+    i = strlen(buff) - 1;
+    if (i > 0 && buff[i] == 's') {
+	buff[i] = '\0';
+	for (tp = UnitsTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    for (tp = OtherTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    /*
+     * Military timezones.
+     */
+
+    if (buff[1] == '\0' && !(*buff & 0x80)
+	    && isalpha(UCHAR(*buff))) {			/* INTL: ISO only */
+	for (tp = MilitaryTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    /*
+     * Drop out any periods and try the timezone table again.
+     */
+
+    for (i = 0, p = q = buff; *q; q++) {
+	if (*q != '.') {
+	    *p++ = *q;
+	} else {
+	    i++;
+	}
+    }
+    *p = '\0';
+    if (i) {
+	for (tp = TimezoneTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    return tID;
+}
+
+static int
+TclDatelex(
+    YYSTYPE* yylvalPtr,
+    YYLTYPE* location,
+    DateInfo *info)
+{
+    register char c;
+    register char *p;
+    char buff[20];
+    int Count;
+
+    location->first_column = yyInput - info->dateStart;
+    for ( ; ; ) {
+	while (TclIsSpaceProc(*yyInput)) {
+	    yyInput++;
+	}
+
+	if (isdigit(UCHAR(c = *yyInput))) { /* INTL: digit */
+	    /*
+	     * Convert the string into a number; count the number of digits.
+	     */
+
+	    Count = 0;
+	    for (yylvalPtr->Number = 0;
+		    isdigit(UCHAR(c = *yyInput++)); ) {	  /* INTL: digit */
+		yylvalPtr->Number = 10 * yylvalPtr->Number + c - '0';
+		Count++;
+	    }
+	    yyInput--;
+	    yyDigitCount = Count;
+
+	    /*
+	     * A number with 6 or more digits is considered an ISO 8601 base.
+	     */
+
+	    if (Count >= 6) {
+		location->last_column = yyInput - info->dateStart - 1;
+		return tISOBASE;
+	    } else {
+		location->last_column = yyInput - info->dateStart - 1;
+		return tUNUMBER;
+	    }
+	}
+	if (!(c & 0x80) && isalpha(UCHAR(c))) {		  /* INTL: ISO only. */
+	    for (p = buff; isalpha(UCHAR(c = *yyInput++)) /* INTL: ISO only. */
+		     || c == '.'; ) {
+		if (p < &buff[sizeof buff - 1]) {
+		    *p++ = c;
+		}
+	    }
+	    *p = '\0';
+	    yyInput--;
+	    location->last_column = yyInput - info->dateStart - 1;
+	    return LookupWord(yylvalPtr, buff);
+	}
+	if (c != '(') {
+	    location->last_column = yyInput - info->dateStart;
+	    return *yyInput++;
+	}
+	Count = 0;
+	do {
+	    c = *yyInput++;
+	    if (c == '\0') {
+		location->last_column = yyInput - info->dateStart - 1;
+		return c;
+	    } else if (c == '(') {
+		Count++;
+	    } else if (c == ')') {
+		Count--;
+	    }
+	} while (Count > 0);
+    }
+}
+
+int
+TclClockOldscanObjCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Count of paraneters */
+    Tcl_Obj *const *objv)	/* Parameters */
+{
+    Tcl_Obj *result, *resultElement;
+    int yr, mo, da;
+    DateInfo dateInfo;
+    DateInfo* info = &dateInfo;
+    int status;
+
+    if (objc != 5) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"stringToParse baseYear baseMonth baseDay" );
+	return TCL_ERROR;
+    }
+
+    yyInput = Tcl_GetString( objv[1] );
+    dateInfo.dateStart = yyInput;
+
+    yyHaveDate = 0;
+    if (Tcl_GetIntFromObj(interp, objv[2], &yr) != TCL_OK
+	    || Tcl_GetIntFromObj(interp, objv[3], &mo) != TCL_OK
+	    || Tcl_GetIntFromObj(interp, objv[4], &da) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    yyYear = yr; yyMonth = mo; yyDay = da;
+
+    yyHaveTime = 0;
+    yyHour = 0; yyMinutes = 0; yySeconds = 0; yyMeridian = MER24;
+
+    yyHaveZone = 0;
+    yyTimezone = 0; yyDSTmode = DSTmaybe;
+
+    yyHaveOrdinalMonth = 0;
+    yyMonthOrdinal = 0;
+
+    yyHaveDay = 0;
+    yyDayOrdinal = 0; yyDayNumber = 0;
+
+    yyHaveRel = 0;
+    yyRelMonth = 0; yyRelDay = 0; yyRelSeconds = 0; yyRelPointer = NULL;
+
+    dateInfo.messages = Tcl_NewObj();
+    dateInfo.separatrix = "";
+    Tcl_IncrRefCount(dateInfo.messages);
+
+    status = yyparse(&dateInfo);
+    if (status == 1) {
+	Tcl_SetObjResult(interp, dateInfo.messages);
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "PARSE", NULL);
+	return TCL_ERROR;
+    } else if (status == 2) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("memory exhausted", -1));
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	return TCL_ERROR;
+    } else if (status != 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("Unknown status returned "
+						  "from date parser. Please "
+						  "report this error as a "
+						  "bug in Tcl.", -1));
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "BUG", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_DecrRefCount(dateInfo.messages);
+
+    if (yyHaveDate > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one date in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveTime > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one time of day in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveZone > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one time zone in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveDay > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one weekday in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveOrdinalMonth > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one ordinal month in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+
+    result = Tcl_NewObj();
+    resultElement = Tcl_NewObj();
+    if (yyHaveDate) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyYear));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonth));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDay));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    if (yyHaveTime) {
+	Tcl_ListObjAppendElement(interp, result, Tcl_NewIntObj((int)
+		ToSeconds(yyHour, yyMinutes, yySeconds, yyMeridian)));
+    } else {
+	Tcl_ListObjAppendElement(interp, result, Tcl_NewObj());
+    }
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveZone) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) -yyTimezone));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj(1 - yyDSTmode));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveRel) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelMonth));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelDay));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelSeconds));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveDay && !yyHaveDate) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDayOrdinal));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDayNumber));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveOrdinalMonth) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonthOrdinal));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonth));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    Tcl_SetObjResult(interp, result);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
+
diff --git a/generic/tclDecls.h b/generic/tclDecls.h
new file mode 100644
index 0000000..d543238
--- /dev/null
+++ b/generic/tclDecls.h
@@ -0,0 +1,3971 @@
+/*
+ * tclDecls.h --
+ *
+ *	Declarations of functions in the platform independent public Tcl API.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLDECLS
+#define _TCLDECLS
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/*
+ * WARNING: This file is automatically generated by the tools/genStubs.tcl
+ * script.  Any modifications to the function declarations below should be made
+ * in the generic/tcl.decls script.
+ */
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+/* 0 */
+EXTERN int		Tcl_PkgProvideEx(Tcl_Interp *interp,
+				const char *name, const char *version,
+				const void *clientData);
+/* 1 */
+EXTERN CONST84_RETURN char * Tcl_PkgRequireEx(Tcl_Interp *interp,
+				const char *name, const char *version,
+				int exact, void *clientDataPtr);
+/* 2 */
+EXTERN TCL_NORETURN void Tcl_Panic(const char *format, ...) TCL_FORMAT_PRINTF(1, 2);
+/* 3 */
+EXTERN char *		Tcl_Alloc(unsigned int size);
+/* 4 */
+EXTERN void		Tcl_Free(char *ptr);
+/* 5 */
+EXTERN char *		Tcl_Realloc(char *ptr, unsigned int size);
+/* 6 */
+EXTERN char *		Tcl_DbCkalloc(unsigned int size, const char *file,
+				int line);
+/* 7 */
+EXTERN void		Tcl_DbCkfree(char *ptr, const char *file, int line);
+/* 8 */
+EXTERN char *		Tcl_DbCkrealloc(char *ptr, unsigned int size,
+				const char *file, int line);
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+/* 9 */
+EXTERN void		Tcl_CreateFileHandler(int fd, int mask,
+				Tcl_FileProc *proc, ClientData clientData);
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+/* 9 */
+EXTERN void		Tcl_CreateFileHandler(int fd, int mask,
+				Tcl_FileProc *proc, ClientData clientData);
+#endif /* MACOSX */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+/* 10 */
+EXTERN void		Tcl_DeleteFileHandler(int fd);
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+/* 10 */
+EXTERN void		Tcl_DeleteFileHandler(int fd);
+#endif /* MACOSX */
+/* 11 */
+EXTERN void		Tcl_SetTimer(const Tcl_Time *timePtr);
+/* 12 */
+EXTERN void		Tcl_Sleep(int ms);
+/* 13 */
+EXTERN int		Tcl_WaitForEvent(const Tcl_Time *timePtr);
+/* 14 */
+EXTERN int		Tcl_AppendAllObjTypes(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 15 */
+EXTERN void		Tcl_AppendStringsToObj(Tcl_Obj *objPtr, ...);
+/* 16 */
+EXTERN void		Tcl_AppendToObj(Tcl_Obj *objPtr, const char *bytes,
+				int length);
+/* 17 */
+EXTERN Tcl_Obj *	Tcl_ConcatObj(int objc, Tcl_Obj *const objv[]);
+/* 18 */
+EXTERN int		Tcl_ConvertToType(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, const Tcl_ObjType *typePtr);
+/* 19 */
+EXTERN void		Tcl_DbDecrRefCount(Tcl_Obj *objPtr, const char *file,
+				int line);
+/* 20 */
+EXTERN void		Tcl_DbIncrRefCount(Tcl_Obj *objPtr, const char *file,
+				int line);
+/* 21 */
+EXTERN int		Tcl_DbIsShared(Tcl_Obj *objPtr, const char *file,
+				int line);
+/* 22 */
+EXTERN Tcl_Obj *	Tcl_DbNewBooleanObj(int boolValue, const char *file,
+				int line);
+/* 23 */
+EXTERN Tcl_Obj *	Tcl_DbNewByteArrayObj(const unsigned char *bytes,
+				int length, const char *file, int line);
+/* 24 */
+EXTERN Tcl_Obj *	Tcl_DbNewDoubleObj(double doubleValue,
+				const char *file, int line);
+/* 25 */
+EXTERN Tcl_Obj *	Tcl_DbNewListObj(int objc, Tcl_Obj *const *objv,
+				const char *file, int line);
+/* 26 */
+EXTERN Tcl_Obj *	Tcl_DbNewLongObj(long longValue, const char *file,
+				int line);
+/* 27 */
+EXTERN Tcl_Obj *	Tcl_DbNewObj(const char *file, int line);
+/* 28 */
+EXTERN Tcl_Obj *	Tcl_DbNewStringObj(const char *bytes, int length,
+				const char *file, int line);
+/* 29 */
+EXTERN Tcl_Obj *	Tcl_DuplicateObj(Tcl_Obj *objPtr);
+/* 30 */
+EXTERN void		TclFreeObj(Tcl_Obj *objPtr);
+/* 31 */
+EXTERN int		Tcl_GetBoolean(Tcl_Interp *interp, const char *src,
+				int *boolPtr);
+/* 32 */
+EXTERN int		Tcl_GetBooleanFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, int *boolPtr);
+/* 33 */
+EXTERN unsigned char *	Tcl_GetByteArrayFromObj(Tcl_Obj *objPtr,
+				int *lengthPtr);
+/* 34 */
+EXTERN int		Tcl_GetDouble(Tcl_Interp *interp, const char *src,
+				double *doublePtr);
+/* 35 */
+EXTERN int		Tcl_GetDoubleFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, double *doublePtr);
+/* 36 */
+EXTERN int		Tcl_GetIndexFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr,
+				CONST84 char *const *tablePtr,
+				const char *msg, int flags, int *indexPtr);
+/* 37 */
+EXTERN int		Tcl_GetInt(Tcl_Interp *interp, const char *src,
+				int *intPtr);
+/* 38 */
+EXTERN int		Tcl_GetIntFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, int *intPtr);
+/* 39 */
+EXTERN int		Tcl_GetLongFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, long *longPtr);
+/* 40 */
+EXTERN CONST86 Tcl_ObjType * Tcl_GetObjType(const char *typeName);
+/* 41 */
+EXTERN char *		Tcl_GetStringFromObj(Tcl_Obj *objPtr, int *lengthPtr);
+/* 42 */
+EXTERN void		Tcl_InvalidateStringRep(Tcl_Obj *objPtr);
+/* 43 */
+EXTERN int		Tcl_ListObjAppendList(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, Tcl_Obj *elemListPtr);
+/* 44 */
+EXTERN int		Tcl_ListObjAppendElement(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, Tcl_Obj *objPtr);
+/* 45 */
+EXTERN int		Tcl_ListObjGetElements(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, int *objcPtr,
+				Tcl_Obj ***objvPtr);
+/* 46 */
+EXTERN int		Tcl_ListObjIndex(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, int index,
+				Tcl_Obj **objPtrPtr);
+/* 47 */
+EXTERN int		Tcl_ListObjLength(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, int *lengthPtr);
+/* 48 */
+EXTERN int		Tcl_ListObjReplace(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, int first, int count,
+				int objc, Tcl_Obj *const objv[]);
+/* 49 */
+EXTERN Tcl_Obj *	Tcl_NewBooleanObj(int boolValue);
+/* 50 */
+EXTERN Tcl_Obj *	Tcl_NewByteArrayObj(const unsigned char *bytes,
+				int length);
+/* 51 */
+EXTERN Tcl_Obj *	Tcl_NewDoubleObj(double doubleValue);
+/* 52 */
+EXTERN Tcl_Obj *	Tcl_NewIntObj(int intValue);
+/* 53 */
+EXTERN Tcl_Obj *	Tcl_NewListObj(int objc, Tcl_Obj *const objv[]);
+/* 54 */
+EXTERN Tcl_Obj *	Tcl_NewLongObj(long longValue);
+/* 55 */
+EXTERN Tcl_Obj *	Tcl_NewObj(void);
+/* 56 */
+EXTERN Tcl_Obj *	Tcl_NewStringObj(const char *bytes, int length);
+/* 57 */
+EXTERN void		Tcl_SetBooleanObj(Tcl_Obj *objPtr, int boolValue);
+/* 58 */
+EXTERN unsigned char *	Tcl_SetByteArrayLength(Tcl_Obj *objPtr, int length);
+/* 59 */
+EXTERN void		Tcl_SetByteArrayObj(Tcl_Obj *objPtr,
+				const unsigned char *bytes, int length);
+/* 60 */
+EXTERN void		Tcl_SetDoubleObj(Tcl_Obj *objPtr, double doubleValue);
+/* 61 */
+EXTERN void		Tcl_SetIntObj(Tcl_Obj *objPtr, int intValue);
+/* 62 */
+EXTERN void		Tcl_SetListObj(Tcl_Obj *objPtr, int objc,
+				Tcl_Obj *const objv[]);
+/* 63 */
+EXTERN void		Tcl_SetLongObj(Tcl_Obj *objPtr, long longValue);
+/* 64 */
+EXTERN void		Tcl_SetObjLength(Tcl_Obj *objPtr, int length);
+/* 65 */
+EXTERN void		Tcl_SetStringObj(Tcl_Obj *objPtr, const char *bytes,
+				int length);
+/* 66 */
+EXTERN void		Tcl_AddErrorInfo(Tcl_Interp *interp,
+				const char *message);
+/* 67 */
+EXTERN void		Tcl_AddObjErrorInfo(Tcl_Interp *interp,
+				const char *message, int length);
+/* 68 */
+EXTERN void		Tcl_AllowExceptions(Tcl_Interp *interp);
+/* 69 */
+EXTERN void		Tcl_AppendElement(Tcl_Interp *interp,
+				const char *element);
+/* 70 */
+EXTERN void		Tcl_AppendResult(Tcl_Interp *interp, ...);
+/* 71 */
+EXTERN Tcl_AsyncHandler	 Tcl_AsyncCreate(Tcl_AsyncProc *proc,
+				ClientData clientData);
+/* 72 */
+EXTERN void		Tcl_AsyncDelete(Tcl_AsyncHandler async);
+/* 73 */
+EXTERN int		Tcl_AsyncInvoke(Tcl_Interp *interp, int code);
+/* 74 */
+EXTERN void		Tcl_AsyncMark(Tcl_AsyncHandler async);
+/* 75 */
+EXTERN int		Tcl_AsyncReady(void);
+/* 76 */
+EXTERN void		Tcl_BackgroundError(Tcl_Interp *interp);
+/* 77 */
+EXTERN char		Tcl_Backslash(const char *src, int *readPtr);
+/* 78 */
+EXTERN int		Tcl_BadChannelOption(Tcl_Interp *interp,
+				const char *optionName,
+				const char *optionList);
+/* 79 */
+EXTERN void		Tcl_CallWhenDeleted(Tcl_Interp *interp,
+				Tcl_InterpDeleteProc *proc,
+				ClientData clientData);
+/* 80 */
+EXTERN void		Tcl_CancelIdleCall(Tcl_IdleProc *idleProc,
+				ClientData clientData);
+/* 81 */
+EXTERN int		Tcl_Close(Tcl_Interp *interp, Tcl_Channel chan);
+/* 82 */
+EXTERN int		Tcl_CommandComplete(const char *cmd);
+/* 83 */
+EXTERN char *		Tcl_Concat(int argc, CONST84 char *const *argv);
+/* 84 */
+EXTERN int		Tcl_ConvertElement(const char *src, char *dst,
+				int flags);
+/* 85 */
+EXTERN int		Tcl_ConvertCountedElement(const char *src,
+				int length, char *dst, int flags);
+/* 86 */
+EXTERN int		Tcl_CreateAlias(Tcl_Interp *slave,
+				const char *slaveCmd, Tcl_Interp *target,
+				const char *targetCmd, int argc,
+				CONST84 char *const *argv);
+/* 87 */
+EXTERN int		Tcl_CreateAliasObj(Tcl_Interp *slave,
+				const char *slaveCmd, Tcl_Interp *target,
+				const char *targetCmd, int objc,
+				Tcl_Obj *const objv[]);
+/* 88 */
+EXTERN Tcl_Channel	Tcl_CreateChannel(const Tcl_ChannelType *typePtr,
+				const char *chanName,
+				ClientData instanceData, int mask);
+/* 89 */
+EXTERN void		Tcl_CreateChannelHandler(Tcl_Channel chan, int mask,
+				Tcl_ChannelProc *proc, ClientData clientData);
+/* 90 */
+EXTERN void		Tcl_CreateCloseHandler(Tcl_Channel chan,
+				Tcl_CloseProc *proc, ClientData clientData);
+/* 91 */
+EXTERN Tcl_Command	Tcl_CreateCommand(Tcl_Interp *interp,
+				const char *cmdName, Tcl_CmdProc *proc,
+				ClientData clientData,
+				Tcl_CmdDeleteProc *deleteProc);
+/* 92 */
+EXTERN void		Tcl_CreateEventSource(Tcl_EventSetupProc *setupProc,
+				Tcl_EventCheckProc *checkProc,
+				ClientData clientData);
+/* 93 */
+EXTERN void		Tcl_CreateExitHandler(Tcl_ExitProc *proc,
+				ClientData clientData);
+/* 94 */
+EXTERN Tcl_Interp *	Tcl_CreateInterp(void);
+/* 95 */
+EXTERN void		Tcl_CreateMathFunc(Tcl_Interp *interp,
+				const char *name, int numArgs,
+				Tcl_ValueType *argTypes, Tcl_MathProc *proc,
+				ClientData clientData);
+/* 96 */
+EXTERN Tcl_Command	Tcl_CreateObjCommand(Tcl_Interp *interp,
+				const char *cmdName, Tcl_ObjCmdProc *proc,
+				ClientData clientData,
+				Tcl_CmdDeleteProc *deleteProc);
+/* 97 */
+EXTERN Tcl_Interp *	Tcl_CreateSlave(Tcl_Interp *interp,
+				const char *slaveName, int isSafe);
+/* 98 */
+EXTERN Tcl_TimerToken	Tcl_CreateTimerHandler(int milliseconds,
+				Tcl_TimerProc *proc, ClientData clientData);
+/* 99 */
+EXTERN Tcl_Trace	Tcl_CreateTrace(Tcl_Interp *interp, int level,
+				Tcl_CmdTraceProc *proc,
+				ClientData clientData);
+/* 100 */
+EXTERN void		Tcl_DeleteAssocData(Tcl_Interp *interp,
+				const char *name);
+/* 101 */
+EXTERN void		Tcl_DeleteChannelHandler(Tcl_Channel chan,
+				Tcl_ChannelProc *proc, ClientData clientData);
+/* 102 */
+EXTERN void		Tcl_DeleteCloseHandler(Tcl_Channel chan,
+				Tcl_CloseProc *proc, ClientData clientData);
+/* 103 */
+EXTERN int		Tcl_DeleteCommand(Tcl_Interp *interp,
+				const char *cmdName);
+/* 104 */
+EXTERN int		Tcl_DeleteCommandFromToken(Tcl_Interp *interp,
+				Tcl_Command command);
+/* 105 */
+EXTERN void		Tcl_DeleteEvents(Tcl_EventDeleteProc *proc,
+				ClientData clientData);
+/* 106 */
+EXTERN void		Tcl_DeleteEventSource(Tcl_EventSetupProc *setupProc,
+				Tcl_EventCheckProc *checkProc,
+				ClientData clientData);
+/* 107 */
+EXTERN void		Tcl_DeleteExitHandler(Tcl_ExitProc *proc,
+				ClientData clientData);
+/* 108 */
+EXTERN void		Tcl_DeleteHashEntry(Tcl_HashEntry *entryPtr);
+/* 109 */
+EXTERN void		Tcl_DeleteHashTable(Tcl_HashTable *tablePtr);
+/* 110 */
+EXTERN void		Tcl_DeleteInterp(Tcl_Interp *interp);
+/* 111 */
+EXTERN void		Tcl_DetachPids(int numPids, Tcl_Pid *pidPtr);
+/* 112 */
+EXTERN void		Tcl_DeleteTimerHandler(Tcl_TimerToken token);
+/* 113 */
+EXTERN void		Tcl_DeleteTrace(Tcl_Interp *interp, Tcl_Trace trace);
+/* 114 */
+EXTERN void		Tcl_DontCallWhenDeleted(Tcl_Interp *interp,
+				Tcl_InterpDeleteProc *proc,
+				ClientData clientData);
+/* 115 */
+EXTERN int		Tcl_DoOneEvent(int flags);
+/* 116 */
+EXTERN void		Tcl_DoWhenIdle(Tcl_IdleProc *proc,
+				ClientData clientData);
+/* 117 */
+EXTERN char *		Tcl_DStringAppend(Tcl_DString *dsPtr,
+				const char *bytes, int length);
+/* 118 */
+EXTERN char *		Tcl_DStringAppendElement(Tcl_DString *dsPtr,
+				const char *element);
+/* 119 */
+EXTERN void		Tcl_DStringEndSublist(Tcl_DString *dsPtr);
+/* 120 */
+EXTERN void		Tcl_DStringFree(Tcl_DString *dsPtr);
+/* 121 */
+EXTERN void		Tcl_DStringGetResult(Tcl_Interp *interp,
+				Tcl_DString *dsPtr);
+/* 122 */
+EXTERN void		Tcl_DStringInit(Tcl_DString *dsPtr);
+/* 123 */
+EXTERN void		Tcl_DStringResult(Tcl_Interp *interp,
+				Tcl_DString *dsPtr);
+/* 124 */
+EXTERN void		Tcl_DStringSetLength(Tcl_DString *dsPtr, int length);
+/* 125 */
+EXTERN void		Tcl_DStringStartSublist(Tcl_DString *dsPtr);
+/* 126 */
+EXTERN int		Tcl_Eof(Tcl_Channel chan);
+/* 127 */
+EXTERN CONST84_RETURN char * Tcl_ErrnoId(void);
+/* 128 */
+EXTERN CONST84_RETURN char * Tcl_ErrnoMsg(int err);
+/* 129 */
+EXTERN int		Tcl_Eval(Tcl_Interp *interp, const char *script);
+/* 130 */
+EXTERN int		Tcl_EvalFile(Tcl_Interp *interp,
+				const char *fileName);
+/* 131 */
+EXTERN int		Tcl_EvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr);
+/* 132 */
+EXTERN void		Tcl_EventuallyFree(ClientData clientData,
+				Tcl_FreeProc *freeProc);
+/* 133 */
+EXTERN TCL_NORETURN void Tcl_Exit(int status);
+/* 134 */
+EXTERN int		Tcl_ExposeCommand(Tcl_Interp *interp,
+				const char *hiddenCmdToken,
+				const char *cmdName);
+/* 135 */
+EXTERN int		Tcl_ExprBoolean(Tcl_Interp *interp, const char *expr,
+				int *ptr);
+/* 136 */
+EXTERN int		Tcl_ExprBooleanObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, int *ptr);
+/* 137 */
+EXTERN int		Tcl_ExprDouble(Tcl_Interp *interp, const char *expr,
+				double *ptr);
+/* 138 */
+EXTERN int		Tcl_ExprDoubleObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, double *ptr);
+/* 139 */
+EXTERN int		Tcl_ExprLong(Tcl_Interp *interp, const char *expr,
+				long *ptr);
+/* 140 */
+EXTERN int		Tcl_ExprLongObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				long *ptr);
+/* 141 */
+EXTERN int		Tcl_ExprObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				Tcl_Obj **resultPtrPtr);
+/* 142 */
+EXTERN int		Tcl_ExprString(Tcl_Interp *interp, const char *expr);
+/* 143 */
+EXTERN void		Tcl_Finalize(void);
+/* 144 */
+EXTERN void		Tcl_FindExecutable(const char *argv0);
+/* 145 */
+EXTERN Tcl_HashEntry *	Tcl_FirstHashEntry(Tcl_HashTable *tablePtr,
+				Tcl_HashSearch *searchPtr);
+/* 146 */
+EXTERN int		Tcl_Flush(Tcl_Channel chan);
+/* 147 */
+EXTERN void		Tcl_FreeResult(Tcl_Interp *interp);
+/* 148 */
+EXTERN int		Tcl_GetAlias(Tcl_Interp *interp,
+				const char *slaveCmd,
+				Tcl_Interp **targetInterpPtr,
+				CONST84 char **targetCmdPtr, int *argcPtr,
+				CONST84 char ***argvPtr);
+/* 149 */
+EXTERN int		Tcl_GetAliasObj(Tcl_Interp *interp,
+				const char *slaveCmd,
+				Tcl_Interp **targetInterpPtr,
+				CONST84 char **targetCmdPtr, int *objcPtr,
+				Tcl_Obj ***objv);
+/* 150 */
+EXTERN ClientData	Tcl_GetAssocData(Tcl_Interp *interp,
+				const char *name,
+				Tcl_InterpDeleteProc **procPtr);
+/* 151 */
+EXTERN Tcl_Channel	Tcl_GetChannel(Tcl_Interp *interp,
+				const char *chanName, int *modePtr);
+/* 152 */
+EXTERN int		Tcl_GetChannelBufferSize(Tcl_Channel chan);
+/* 153 */
+EXTERN int		Tcl_GetChannelHandle(Tcl_Channel chan, int direction,
+				ClientData *handlePtr);
+/* 154 */
+EXTERN ClientData	Tcl_GetChannelInstanceData(Tcl_Channel chan);
+/* 155 */
+EXTERN int		Tcl_GetChannelMode(Tcl_Channel chan);
+/* 156 */
+EXTERN CONST84_RETURN char * Tcl_GetChannelName(Tcl_Channel chan);
+/* 157 */
+EXTERN int		Tcl_GetChannelOption(Tcl_Interp *interp,
+				Tcl_Channel chan, const char *optionName,
+				Tcl_DString *dsPtr);
+/* 158 */
+EXTERN CONST86 Tcl_ChannelType * Tcl_GetChannelType(Tcl_Channel chan);
+/* 159 */
+EXTERN int		Tcl_GetCommandInfo(Tcl_Interp *interp,
+				const char *cmdName, Tcl_CmdInfo *infoPtr);
+/* 160 */
+EXTERN CONST84_RETURN char * Tcl_GetCommandName(Tcl_Interp *interp,
+				Tcl_Command command);
+/* 161 */
+EXTERN int		Tcl_GetErrno(void);
+/* 162 */
+EXTERN CONST84_RETURN char * Tcl_GetHostName(void);
+/* 163 */
+EXTERN int		Tcl_GetInterpPath(Tcl_Interp *askInterp,
+				Tcl_Interp *slaveInterp);
+/* 164 */
+EXTERN Tcl_Interp *	Tcl_GetMaster(Tcl_Interp *interp);
+/* 165 */
+EXTERN const char *	Tcl_GetNameOfExecutable(void);
+/* 166 */
+EXTERN Tcl_Obj *	Tcl_GetObjResult(Tcl_Interp *interp);
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+/* 167 */
+EXTERN int		Tcl_GetOpenFile(Tcl_Interp *interp,
+				const char *chanID, int forWriting,
+				int checkUsage, ClientData *filePtr);
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+/* 167 */
+EXTERN int		Tcl_GetOpenFile(Tcl_Interp *interp,
+				const char *chanID, int forWriting,
+				int checkUsage, ClientData *filePtr);
+#endif /* MACOSX */
+/* 168 */
+EXTERN Tcl_PathType	Tcl_GetPathType(const char *path);
+/* 169 */
+EXTERN int		Tcl_Gets(Tcl_Channel chan, Tcl_DString *dsPtr);
+/* 170 */
+EXTERN int		Tcl_GetsObj(Tcl_Channel chan, Tcl_Obj *objPtr);
+/* 171 */
+EXTERN int		Tcl_GetServiceMode(void);
+/* 172 */
+EXTERN Tcl_Interp *	Tcl_GetSlave(Tcl_Interp *interp,
+				const char *slaveName);
+/* 173 */
+EXTERN Tcl_Channel	Tcl_GetStdChannel(int type);
+/* 174 */
+EXTERN CONST84_RETURN char * Tcl_GetStringResult(Tcl_Interp *interp);
+/* 175 */
+EXTERN CONST84_RETURN char * Tcl_GetVar(Tcl_Interp *interp,
+				const char *varName, int flags);
+/* 176 */
+EXTERN CONST84_RETURN char * Tcl_GetVar2(Tcl_Interp *interp,
+				const char *part1, const char *part2,
+				int flags);
+/* 177 */
+EXTERN int		Tcl_GlobalEval(Tcl_Interp *interp,
+				const char *command);
+/* 178 */
+EXTERN int		Tcl_GlobalEvalObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 179 */
+EXTERN int		Tcl_HideCommand(Tcl_Interp *interp,
+				const char *cmdName,
+				const char *hiddenCmdToken);
+/* 180 */
+EXTERN int		Tcl_Init(Tcl_Interp *interp);
+/* 181 */
+EXTERN void		Tcl_InitHashTable(Tcl_HashTable *tablePtr,
+				int keyType);
+/* 182 */
+EXTERN int		Tcl_InputBlocked(Tcl_Channel chan);
+/* 183 */
+EXTERN int		Tcl_InputBuffered(Tcl_Channel chan);
+/* 184 */
+EXTERN int		Tcl_InterpDeleted(Tcl_Interp *interp);
+/* 185 */
+EXTERN int		Tcl_IsSafe(Tcl_Interp *interp);
+/* 186 */
+EXTERN char *		Tcl_JoinPath(int argc, CONST84 char *const *argv,
+				Tcl_DString *resultPtr);
+/* 187 */
+EXTERN int		Tcl_LinkVar(Tcl_Interp *interp, const char *varName,
+				char *addr, int type);
+/* Slot 188 is reserved */
+/* 189 */
+EXTERN Tcl_Channel	Tcl_MakeFileChannel(ClientData handle, int mode);
+/* 190 */
+EXTERN int		Tcl_MakeSafe(Tcl_Interp *interp);
+/* 191 */
+EXTERN Tcl_Channel	Tcl_MakeTcpClientChannel(ClientData tcpSocket);
+/* 192 */
+EXTERN char *		Tcl_Merge(int argc, CONST84 char *const *argv);
+/* 193 */
+EXTERN Tcl_HashEntry *	Tcl_NextHashEntry(Tcl_HashSearch *searchPtr);
+/* 194 */
+EXTERN void		Tcl_NotifyChannel(Tcl_Channel channel, int mask);
+/* 195 */
+EXTERN Tcl_Obj *	Tcl_ObjGetVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, int flags);
+/* 196 */
+EXTERN Tcl_Obj *	Tcl_ObjSetVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr,
+				int flags);
+/* 197 */
+EXTERN Tcl_Channel	Tcl_OpenCommandChannel(Tcl_Interp *interp, int argc,
+				CONST84 char **argv, int flags);
+/* 198 */
+EXTERN Tcl_Channel	Tcl_OpenFileChannel(Tcl_Interp *interp,
+				const char *fileName, const char *modeString,
+				int permissions);
+/* 199 */
+EXTERN Tcl_Channel	Tcl_OpenTcpClient(Tcl_Interp *interp, int port,
+				const char *address, const char *myaddr,
+				int myport, int async);
+/* 200 */
+EXTERN Tcl_Channel	Tcl_OpenTcpServer(Tcl_Interp *interp, int port,
+				const char *host,
+				Tcl_TcpAcceptProc *acceptProc,
+				ClientData callbackData);
+/* 201 */
+EXTERN void		Tcl_Preserve(ClientData data);
+/* 202 */
+EXTERN void		Tcl_PrintDouble(Tcl_Interp *interp, double value,
+				char *dst);
+/* 203 */
+EXTERN int		Tcl_PutEnv(const char *assignment);
+/* 204 */
+EXTERN CONST84_RETURN char * Tcl_PosixError(Tcl_Interp *interp);
+/* 205 */
+EXTERN void		Tcl_QueueEvent(Tcl_Event *evPtr,
+				Tcl_QueuePosition position);
+/* 206 */
+EXTERN int		Tcl_Read(Tcl_Channel chan, char *bufPtr, int toRead);
+/* 207 */
+EXTERN void		Tcl_ReapDetachedProcs(void);
+/* 208 */
+EXTERN int		Tcl_RecordAndEval(Tcl_Interp *interp,
+				const char *cmd, int flags);
+/* 209 */
+EXTERN int		Tcl_RecordAndEvalObj(Tcl_Interp *interp,
+				Tcl_Obj *cmdPtr, int flags);
+/* 210 */
+EXTERN void		Tcl_RegisterChannel(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 211 */
+EXTERN void		Tcl_RegisterObjType(const Tcl_ObjType *typePtr);
+/* 212 */
+EXTERN Tcl_RegExp	Tcl_RegExpCompile(Tcl_Interp *interp,
+				const char *pattern);
+/* 213 */
+EXTERN int		Tcl_RegExpExec(Tcl_Interp *interp, Tcl_RegExp regexp,
+				const char *text, const char *start);
+/* 214 */
+EXTERN int		Tcl_RegExpMatch(Tcl_Interp *interp, const char *text,
+				const char *pattern);
+/* 215 */
+EXTERN void		Tcl_RegExpRange(Tcl_RegExp regexp, int index,
+				CONST84 char **startPtr,
+				CONST84 char **endPtr);
+/* 216 */
+EXTERN void		Tcl_Release(ClientData clientData);
+/* 217 */
+EXTERN void		Tcl_ResetResult(Tcl_Interp *interp);
+/* 218 */
+EXTERN int		Tcl_ScanElement(const char *src, int *flagPtr);
+/* 219 */
+EXTERN int		Tcl_ScanCountedElement(const char *src, int length,
+				int *flagPtr);
+/* 220 */
+EXTERN int		Tcl_SeekOld(Tcl_Channel chan, int offset, int mode);
+/* 221 */
+EXTERN int		Tcl_ServiceAll(void);
+/* 222 */
+EXTERN int		Tcl_ServiceEvent(int flags);
+/* 223 */
+EXTERN void		Tcl_SetAssocData(Tcl_Interp *interp,
+				const char *name, Tcl_InterpDeleteProc *proc,
+				ClientData clientData);
+/* 224 */
+EXTERN void		Tcl_SetChannelBufferSize(Tcl_Channel chan, int sz);
+/* 225 */
+EXTERN int		Tcl_SetChannelOption(Tcl_Interp *interp,
+				Tcl_Channel chan, const char *optionName,
+				const char *newValue);
+/* 226 */
+EXTERN int		Tcl_SetCommandInfo(Tcl_Interp *interp,
+				const char *cmdName,
+				const Tcl_CmdInfo *infoPtr);
+/* 227 */
+EXTERN void		Tcl_SetErrno(int err);
+/* 228 */
+EXTERN void		Tcl_SetErrorCode(Tcl_Interp *interp, ...);
+/* 229 */
+EXTERN void		Tcl_SetMaxBlockTime(const Tcl_Time *timePtr);
+/* 230 */
+EXTERN void		Tcl_SetPanicProc(
+				TCL_NORETURN1 Tcl_PanicProc *panicProc);
+/* 231 */
+EXTERN int		Tcl_SetRecursionLimit(Tcl_Interp *interp, int depth);
+/* 232 */
+EXTERN void		Tcl_SetResult(Tcl_Interp *interp, char *result,
+				Tcl_FreeProc *freeProc);
+/* 233 */
+EXTERN int		Tcl_SetServiceMode(int mode);
+/* 234 */
+EXTERN void		Tcl_SetObjErrorCode(Tcl_Interp *interp,
+				Tcl_Obj *errorObjPtr);
+/* 235 */
+EXTERN void		Tcl_SetObjResult(Tcl_Interp *interp,
+				Tcl_Obj *resultObjPtr);
+/* 236 */
+EXTERN void		Tcl_SetStdChannel(Tcl_Channel channel, int type);
+/* 237 */
+EXTERN CONST84_RETURN char * Tcl_SetVar(Tcl_Interp *interp,
+				const char *varName, const char *newValue,
+				int flags);
+/* 238 */
+EXTERN CONST84_RETURN char * Tcl_SetVar2(Tcl_Interp *interp,
+				const char *part1, const char *part2,
+				const char *newValue, int flags);
+/* 239 */
+EXTERN CONST84_RETURN char * Tcl_SignalId(int sig);
+/* 240 */
+EXTERN CONST84_RETURN char * Tcl_SignalMsg(int sig);
+/* 241 */
+EXTERN void		Tcl_SourceRCFile(Tcl_Interp *interp);
+/* 242 */
+EXTERN int		Tcl_SplitList(Tcl_Interp *interp,
+				const char *listStr, int *argcPtr,
+				CONST84 char ***argvPtr);
+/* 243 */
+EXTERN void		Tcl_SplitPath(const char *path, int *argcPtr,
+				CONST84 char ***argvPtr);
+/* 244 */
+EXTERN void		Tcl_StaticPackage(Tcl_Interp *interp,
+				const char *pkgName,
+				Tcl_PackageInitProc *initProc,
+				Tcl_PackageInitProc *safeInitProc);
+/* 245 */
+EXTERN int		Tcl_StringMatch(const char *str, const char *pattern);
+/* 246 */
+EXTERN int		Tcl_TellOld(Tcl_Channel chan);
+/* 247 */
+EXTERN int		Tcl_TraceVar(Tcl_Interp *interp, const char *varName,
+				int flags, Tcl_VarTraceProc *proc,
+				ClientData clientData);
+/* 248 */
+EXTERN int		Tcl_TraceVar2(Tcl_Interp *interp, const char *part1,
+				const char *part2, int flags,
+				Tcl_VarTraceProc *proc,
+				ClientData clientData);
+/* 249 */
+EXTERN char *		Tcl_TranslateFileName(Tcl_Interp *interp,
+				const char *name, Tcl_DString *bufferPtr);
+/* 250 */
+EXTERN int		Tcl_Ungets(Tcl_Channel chan, const char *str,
+				int len, int atHead);
+/* 251 */
+EXTERN void		Tcl_UnlinkVar(Tcl_Interp *interp,
+				const char *varName);
+/* 252 */
+EXTERN int		Tcl_UnregisterChannel(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 253 */
+EXTERN int		Tcl_UnsetVar(Tcl_Interp *interp, const char *varName,
+				int flags);
+/* 254 */
+EXTERN int		Tcl_UnsetVar2(Tcl_Interp *interp, const char *part1,
+				const char *part2, int flags);
+/* 255 */
+EXTERN void		Tcl_UntraceVar(Tcl_Interp *interp,
+				const char *varName, int flags,
+				Tcl_VarTraceProc *proc,
+				ClientData clientData);
+/* 256 */
+EXTERN void		Tcl_UntraceVar2(Tcl_Interp *interp,
+				const char *part1, const char *part2,
+				int flags, Tcl_VarTraceProc *proc,
+				ClientData clientData);
+/* 257 */
+EXTERN void		Tcl_UpdateLinkedVar(Tcl_Interp *interp,
+				const char *varName);
+/* 258 */
+EXTERN int		Tcl_UpVar(Tcl_Interp *interp, const char *frameName,
+				const char *varName, const char *localName,
+				int flags);
+/* 259 */
+EXTERN int		Tcl_UpVar2(Tcl_Interp *interp, const char *frameName,
+				const char *part1, const char *part2,
+				const char *localName, int flags);
+/* 260 */
+EXTERN int		Tcl_VarEval(Tcl_Interp *interp, ...);
+/* 261 */
+EXTERN ClientData	Tcl_VarTraceInfo(Tcl_Interp *interp,
+				const char *varName, int flags,
+				Tcl_VarTraceProc *procPtr,
+				ClientData prevClientData);
+/* 262 */
+EXTERN ClientData	Tcl_VarTraceInfo2(Tcl_Interp *interp,
+				const char *part1, const char *part2,
+				int flags, Tcl_VarTraceProc *procPtr,
+				ClientData prevClientData);
+/* 263 */
+EXTERN int		Tcl_Write(Tcl_Channel chan, const char *s, int slen);
+/* 264 */
+EXTERN void		Tcl_WrongNumArgs(Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[], const char *message);
+/* 265 */
+EXTERN int		Tcl_DumpActiveMemory(const char *fileName);
+/* 266 */
+EXTERN void		Tcl_ValidateAllMemory(const char *file, int line);
+/* 267 */
+EXTERN void		Tcl_AppendResultVA(Tcl_Interp *interp,
+				va_list argList);
+/* 268 */
+EXTERN void		Tcl_AppendStringsToObjVA(Tcl_Obj *objPtr,
+				va_list argList);
+/* 269 */
+EXTERN char *		Tcl_HashStats(Tcl_HashTable *tablePtr);
+/* 270 */
+EXTERN CONST84_RETURN char * Tcl_ParseVar(Tcl_Interp *interp,
+				const char *start, CONST84 char **termPtr);
+/* 271 */
+EXTERN CONST84_RETURN char * Tcl_PkgPresent(Tcl_Interp *interp,
+				const char *name, const char *version,
+				int exact);
+/* 272 */
+EXTERN CONST84_RETURN char * Tcl_PkgPresentEx(Tcl_Interp *interp,
+				const char *name, const char *version,
+				int exact, void *clientDataPtr);
+/* 273 */
+EXTERN int		Tcl_PkgProvide(Tcl_Interp *interp, const char *name,
+				const char *version);
+/* 274 */
+EXTERN CONST84_RETURN char * Tcl_PkgRequire(Tcl_Interp *interp,
+				const char *name, const char *version,
+				int exact);
+/* 275 */
+EXTERN void		Tcl_SetErrorCodeVA(Tcl_Interp *interp,
+				va_list argList);
+/* 276 */
+EXTERN int		Tcl_VarEvalVA(Tcl_Interp *interp, va_list argList);
+/* 277 */
+EXTERN Tcl_Pid		Tcl_WaitPid(Tcl_Pid pid, int *statPtr, int options);
+/* 278 */
+EXTERN TCL_NORETURN void Tcl_PanicVA(const char *format, va_list argList);
+/* 279 */
+EXTERN void		Tcl_GetVersion(int *major, int *minor,
+				int *patchLevel, int *type);
+/* 280 */
+EXTERN void		Tcl_InitMemory(Tcl_Interp *interp);
+/* 281 */
+EXTERN Tcl_Channel	Tcl_StackChannel(Tcl_Interp *interp,
+				const Tcl_ChannelType *typePtr,
+				ClientData instanceData, int mask,
+				Tcl_Channel prevChan);
+/* 282 */
+EXTERN int		Tcl_UnstackChannel(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 283 */
+EXTERN Tcl_Channel	Tcl_GetStackedChannel(Tcl_Channel chan);
+/* 284 */
+EXTERN void		Tcl_SetMainLoop(Tcl_MainLoopProc *proc);
+/* Slot 285 is reserved */
+/* 286 */
+EXTERN void		Tcl_AppendObjToObj(Tcl_Obj *objPtr,
+				Tcl_Obj *appendObjPtr);
+/* 287 */
+EXTERN Tcl_Encoding	Tcl_CreateEncoding(const Tcl_EncodingType *typePtr);
+/* 288 */
+EXTERN void		Tcl_CreateThreadExitHandler(Tcl_ExitProc *proc,
+				ClientData clientData);
+/* 289 */
+EXTERN void		Tcl_DeleteThreadExitHandler(Tcl_ExitProc *proc,
+				ClientData clientData);
+/* 290 */
+EXTERN void		Tcl_DiscardResult(Tcl_SavedResult *statePtr);
+/* 291 */
+EXTERN int		Tcl_EvalEx(Tcl_Interp *interp, const char *script,
+				int numBytes, int flags);
+/* 292 */
+EXTERN int		Tcl_EvalObjv(Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[], int flags);
+/* 293 */
+EXTERN int		Tcl_EvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags);
+/* 294 */
+EXTERN TCL_NORETURN void Tcl_ExitThread(int status);
+/* 295 */
+EXTERN int		Tcl_ExternalToUtf(Tcl_Interp *interp,
+				Tcl_Encoding encoding, const char *src,
+				int srcLen, int flags,
+				Tcl_EncodingState *statePtr, char *dst,
+				int dstLen, int *srcReadPtr,
+				int *dstWrotePtr, int *dstCharsPtr);
+/* 296 */
+EXTERN char *		Tcl_ExternalToUtfDString(Tcl_Encoding encoding,
+				const char *src, int srcLen,
+				Tcl_DString *dsPtr);
+/* 297 */
+EXTERN void		Tcl_FinalizeThread(void);
+/* 298 */
+EXTERN void		Tcl_FinalizeNotifier(ClientData clientData);
+/* 299 */
+EXTERN void		Tcl_FreeEncoding(Tcl_Encoding encoding);
+/* 300 */
+EXTERN Tcl_ThreadId	Tcl_GetCurrentThread(void);
+/* 301 */
+EXTERN Tcl_Encoding	Tcl_GetEncoding(Tcl_Interp *interp, const char *name);
+/* 302 */
+EXTERN CONST84_RETURN char * Tcl_GetEncodingName(Tcl_Encoding encoding);
+/* 303 */
+EXTERN void		Tcl_GetEncodingNames(Tcl_Interp *interp);
+/* 304 */
+EXTERN int		Tcl_GetIndexFromObjStruct(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, const void *tablePtr,
+				int offset, const char *msg, int flags,
+				int *indexPtr);
+/* 305 */
+EXTERN void *		Tcl_GetThreadData(Tcl_ThreadDataKey *keyPtr,
+				int size);
+/* 306 */
+EXTERN Tcl_Obj *	Tcl_GetVar2Ex(Tcl_Interp *interp, const char *part1,
+				const char *part2, int flags);
+/* 307 */
+EXTERN ClientData	Tcl_InitNotifier(void);
+/* 308 */
+EXTERN void		Tcl_MutexLock(Tcl_Mutex *mutexPtr);
+/* 309 */
+EXTERN void		Tcl_MutexUnlock(Tcl_Mutex *mutexPtr);
+/* 310 */
+EXTERN void		Tcl_ConditionNotify(Tcl_Condition *condPtr);
+/* 311 */
+EXTERN void		Tcl_ConditionWait(Tcl_Condition *condPtr,
+				Tcl_Mutex *mutexPtr, const Tcl_Time *timePtr);
+/* 312 */
+EXTERN int		Tcl_NumUtfChars(const char *src, int length);
+/* 313 */
+EXTERN int		Tcl_ReadChars(Tcl_Channel channel, Tcl_Obj *objPtr,
+				int charsToRead, int appendFlag);
+/* 314 */
+EXTERN void		Tcl_RestoreResult(Tcl_Interp *interp,
+				Tcl_SavedResult *statePtr);
+/* 315 */
+EXTERN void		Tcl_SaveResult(Tcl_Interp *interp,
+				Tcl_SavedResult *statePtr);
+/* 316 */
+EXTERN int		Tcl_SetSystemEncoding(Tcl_Interp *interp,
+				const char *name);
+/* 317 */
+EXTERN Tcl_Obj *	Tcl_SetVar2Ex(Tcl_Interp *interp, const char *part1,
+				const char *part2, Tcl_Obj *newValuePtr,
+				int flags);
+/* 318 */
+EXTERN void		Tcl_ThreadAlert(Tcl_ThreadId threadId);
+/* 319 */
+EXTERN void		Tcl_ThreadQueueEvent(Tcl_ThreadId threadId,
+				Tcl_Event *evPtr, Tcl_QueuePosition position);
+/* 320 */
+EXTERN Tcl_UniChar	Tcl_UniCharAtIndex(const char *src, int index);
+/* 321 */
+EXTERN Tcl_UniChar	Tcl_UniCharToLower(int ch);
+/* 322 */
+EXTERN Tcl_UniChar	Tcl_UniCharToTitle(int ch);
+/* 323 */
+EXTERN Tcl_UniChar	Tcl_UniCharToUpper(int ch);
+/* 324 */
+EXTERN int		Tcl_UniCharToUtf(int ch, char *buf);
+/* 325 */
+EXTERN CONST84_RETURN char * Tcl_UtfAtIndex(const char *src, int index);
+/* 326 */
+EXTERN int		Tcl_UtfCharComplete(const char *src, int length);
+/* 327 */
+EXTERN int		Tcl_UtfBackslash(const char *src, int *readPtr,
+				char *dst);
+/* 328 */
+EXTERN CONST84_RETURN char * Tcl_UtfFindFirst(const char *src, int ch);
+/* 329 */
+EXTERN CONST84_RETURN char * Tcl_UtfFindLast(const char *src, int ch);
+/* 330 */
+EXTERN CONST84_RETURN char * Tcl_UtfNext(const char *src);
+/* 331 */
+EXTERN CONST84_RETURN char * Tcl_UtfPrev(const char *src, const char *start);
+/* 332 */
+EXTERN int		Tcl_UtfToExternal(Tcl_Interp *interp,
+				Tcl_Encoding encoding, const char *src,
+				int srcLen, int flags,
+				Tcl_EncodingState *statePtr, char *dst,
+				int dstLen, int *srcReadPtr,
+				int *dstWrotePtr, int *dstCharsPtr);
+/* 333 */
+EXTERN char *		Tcl_UtfToExternalDString(Tcl_Encoding encoding,
+				const char *src, int srcLen,
+				Tcl_DString *dsPtr);
+/* 334 */
+EXTERN int		Tcl_UtfToLower(char *src);
+/* 335 */
+EXTERN int		Tcl_UtfToTitle(char *src);
+/* 336 */
+EXTERN int		Tcl_UtfToUniChar(const char *src, Tcl_UniChar *chPtr);
+/* 337 */
+EXTERN int		Tcl_UtfToUpper(char *src);
+/* 338 */
+EXTERN int		Tcl_WriteChars(Tcl_Channel chan, const char *src,
+				int srcLen);
+/* 339 */
+EXTERN int		Tcl_WriteObj(Tcl_Channel chan, Tcl_Obj *objPtr);
+/* 340 */
+EXTERN char *		Tcl_GetString(Tcl_Obj *objPtr);
+/* 341 */
+EXTERN CONST84_RETURN char * Tcl_GetDefaultEncodingDir(void);
+/* 342 */
+EXTERN void		Tcl_SetDefaultEncodingDir(const char *path);
+/* 343 */
+EXTERN void		Tcl_AlertNotifier(ClientData clientData);
+/* 344 */
+EXTERN void		Tcl_ServiceModeHook(int mode);
+/* 345 */
+EXTERN int		Tcl_UniCharIsAlnum(int ch);
+/* 346 */
+EXTERN int		Tcl_UniCharIsAlpha(int ch);
+/* 347 */
+EXTERN int		Tcl_UniCharIsDigit(int ch);
+/* 348 */
+EXTERN int		Tcl_UniCharIsLower(int ch);
+/* 349 */
+EXTERN int		Tcl_UniCharIsSpace(int ch);
+/* 350 */
+EXTERN int		Tcl_UniCharIsUpper(int ch);
+/* 351 */
+EXTERN int		Tcl_UniCharIsWordChar(int ch);
+/* 352 */
+EXTERN int		Tcl_UniCharLen(const Tcl_UniChar *uniStr);
+/* 353 */
+EXTERN int		Tcl_UniCharNcmp(const Tcl_UniChar *ucs,
+				const Tcl_UniChar *uct,
+				unsigned long numChars);
+/* 354 */
+EXTERN char *		Tcl_UniCharToUtfDString(const Tcl_UniChar *uniStr,
+				int uniLength, Tcl_DString *dsPtr);
+/* 355 */
+EXTERN Tcl_UniChar *	Tcl_UtfToUniCharDString(const char *src, int length,
+				Tcl_DString *dsPtr);
+/* 356 */
+EXTERN Tcl_RegExp	Tcl_GetRegExpFromObj(Tcl_Interp *interp,
+				Tcl_Obj *patObj, int flags);
+/* 357 */
+EXTERN Tcl_Obj *	Tcl_EvalTokens(Tcl_Interp *interp,
+				Tcl_Token *tokenPtr, int count);
+/* 358 */
+EXTERN void		Tcl_FreeParse(Tcl_Parse *parsePtr);
+/* 359 */
+EXTERN void		Tcl_LogCommandInfo(Tcl_Interp *interp,
+				const char *script, const char *command,
+				int length);
+/* 360 */
+EXTERN int		Tcl_ParseBraces(Tcl_Interp *interp,
+				const char *start, int numBytes,
+				Tcl_Parse *parsePtr, int append,
+				CONST84 char **termPtr);
+/* 361 */
+EXTERN int		Tcl_ParseCommand(Tcl_Interp *interp,
+				const char *start, int numBytes, int nested,
+				Tcl_Parse *parsePtr);
+/* 362 */
+EXTERN int		Tcl_ParseExpr(Tcl_Interp *interp, const char *start,
+				int numBytes, Tcl_Parse *parsePtr);
+/* 363 */
+EXTERN int		Tcl_ParseQuotedString(Tcl_Interp *interp,
+				const char *start, int numBytes,
+				Tcl_Parse *parsePtr, int append,
+				CONST84 char **termPtr);
+/* 364 */
+EXTERN int		Tcl_ParseVarName(Tcl_Interp *interp,
+				const char *start, int numBytes,
+				Tcl_Parse *parsePtr, int append);
+/* 365 */
+EXTERN char *		Tcl_GetCwd(Tcl_Interp *interp, Tcl_DString *cwdPtr);
+/* 366 */
+EXTERN int		Tcl_Chdir(const char *dirName);
+/* 367 */
+EXTERN int		Tcl_Access(const char *path, int mode);
+/* 368 */
+EXTERN int		Tcl_Stat(const char *path, struct stat *bufPtr);
+/* 369 */
+EXTERN int		Tcl_UtfNcmp(const char *s1, const char *s2,
+				unsigned long n);
+/* 370 */
+EXTERN int		Tcl_UtfNcasecmp(const char *s1, const char *s2,
+				unsigned long n);
+/* 371 */
+EXTERN int		Tcl_StringCaseMatch(const char *str,
+				const char *pattern, int nocase);
+/* 372 */
+EXTERN int		Tcl_UniCharIsControl(int ch);
+/* 373 */
+EXTERN int		Tcl_UniCharIsGraph(int ch);
+/* 374 */
+EXTERN int		Tcl_UniCharIsPrint(int ch);
+/* 375 */
+EXTERN int		Tcl_UniCharIsPunct(int ch);
+/* 376 */
+EXTERN int		Tcl_RegExpExecObj(Tcl_Interp *interp,
+				Tcl_RegExp regexp, Tcl_Obj *textObj,
+				int offset, int nmatches, int flags);
+/* 377 */
+EXTERN void		Tcl_RegExpGetInfo(Tcl_RegExp regexp,
+				Tcl_RegExpInfo *infoPtr);
+/* 378 */
+EXTERN Tcl_Obj *	Tcl_NewUnicodeObj(const Tcl_UniChar *unicode,
+				int numChars);
+/* 379 */
+EXTERN void		Tcl_SetUnicodeObj(Tcl_Obj *objPtr,
+				const Tcl_UniChar *unicode, int numChars);
+/* 380 */
+EXTERN int		Tcl_GetCharLength(Tcl_Obj *objPtr);
+/* 381 */
+EXTERN Tcl_UniChar	Tcl_GetUniChar(Tcl_Obj *objPtr, int index);
+/* 382 */
+EXTERN Tcl_UniChar *	Tcl_GetUnicode(Tcl_Obj *objPtr);
+/* 383 */
+EXTERN Tcl_Obj *	Tcl_GetRange(Tcl_Obj *objPtr, int first, int last);
+/* 384 */
+EXTERN void		Tcl_AppendUnicodeToObj(Tcl_Obj *objPtr,
+				const Tcl_UniChar *unicode, int length);
+/* 385 */
+EXTERN int		Tcl_RegExpMatchObj(Tcl_Interp *interp,
+				Tcl_Obj *textObj, Tcl_Obj *patternObj);
+/* 386 */
+EXTERN void		Tcl_SetNotifier(Tcl_NotifierProcs *notifierProcPtr);
+/* 387 */
+EXTERN Tcl_Mutex *	Tcl_GetAllocMutex(void);
+/* 388 */
+EXTERN int		Tcl_GetChannelNames(Tcl_Interp *interp);
+/* 389 */
+EXTERN int		Tcl_GetChannelNamesEx(Tcl_Interp *interp,
+				const char *pattern);
+/* 390 */
+EXTERN int		Tcl_ProcObjCmd(ClientData clientData,
+				Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[]);
+/* 391 */
+EXTERN void		Tcl_ConditionFinalize(Tcl_Condition *condPtr);
+/* 392 */
+EXTERN void		Tcl_MutexFinalize(Tcl_Mutex *mutex);
+/* 393 */
+EXTERN int		Tcl_CreateThread(Tcl_ThreadId *idPtr,
+				Tcl_ThreadCreateProc *proc,
+				ClientData clientData, int stackSize,
+				int flags);
+/* 394 */
+EXTERN int		Tcl_ReadRaw(Tcl_Channel chan, char *dst,
+				int bytesToRead);
+/* 395 */
+EXTERN int		Tcl_WriteRaw(Tcl_Channel chan, const char *src,
+				int srcLen);
+/* 396 */
+EXTERN Tcl_Channel	Tcl_GetTopChannel(Tcl_Channel chan);
+/* 397 */
+EXTERN int		Tcl_ChannelBuffered(Tcl_Channel chan);
+/* 398 */
+EXTERN CONST84_RETURN char * Tcl_ChannelName(
+				const Tcl_ChannelType *chanTypePtr);
+/* 399 */
+EXTERN Tcl_ChannelTypeVersion Tcl_ChannelVersion(
+				const Tcl_ChannelType *chanTypePtr);
+/* 400 */
+EXTERN Tcl_DriverBlockModeProc * Tcl_ChannelBlockModeProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 401 */
+EXTERN Tcl_DriverCloseProc * Tcl_ChannelCloseProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 402 */
+EXTERN Tcl_DriverClose2Proc * Tcl_ChannelClose2Proc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 403 */
+EXTERN Tcl_DriverInputProc * Tcl_ChannelInputProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 404 */
+EXTERN Tcl_DriverOutputProc * Tcl_ChannelOutputProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 405 */
+EXTERN Tcl_DriverSeekProc * Tcl_ChannelSeekProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 406 */
+EXTERN Tcl_DriverSetOptionProc * Tcl_ChannelSetOptionProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 407 */
+EXTERN Tcl_DriverGetOptionProc * Tcl_ChannelGetOptionProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 408 */
+EXTERN Tcl_DriverWatchProc * Tcl_ChannelWatchProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 409 */
+EXTERN Tcl_DriverGetHandleProc * Tcl_ChannelGetHandleProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 410 */
+EXTERN Tcl_DriverFlushProc * Tcl_ChannelFlushProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 411 */
+EXTERN Tcl_DriverHandlerProc * Tcl_ChannelHandlerProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 412 */
+EXTERN int		Tcl_JoinThread(Tcl_ThreadId threadId, int *result);
+/* 413 */
+EXTERN int		Tcl_IsChannelShared(Tcl_Channel channel);
+/* 414 */
+EXTERN int		Tcl_IsChannelRegistered(Tcl_Interp *interp,
+				Tcl_Channel channel);
+/* 415 */
+EXTERN void		Tcl_CutChannel(Tcl_Channel channel);
+/* 416 */
+EXTERN void		Tcl_SpliceChannel(Tcl_Channel channel);
+/* 417 */
+EXTERN void		Tcl_ClearChannelHandlers(Tcl_Channel channel);
+/* 418 */
+EXTERN int		Tcl_IsChannelExisting(const char *channelName);
+/* 419 */
+EXTERN int		Tcl_UniCharNcasecmp(const Tcl_UniChar *ucs,
+				const Tcl_UniChar *uct,
+				unsigned long numChars);
+/* 420 */
+EXTERN int		Tcl_UniCharCaseMatch(const Tcl_UniChar *uniStr,
+				const Tcl_UniChar *uniPattern, int nocase);
+/* 421 */
+EXTERN Tcl_HashEntry *	Tcl_FindHashEntry(Tcl_HashTable *tablePtr,
+				const void *key);
+/* 422 */
+EXTERN Tcl_HashEntry *	Tcl_CreateHashEntry(Tcl_HashTable *tablePtr,
+				const void *key, int *newPtr);
+/* 423 */
+EXTERN void		Tcl_InitCustomHashTable(Tcl_HashTable *tablePtr,
+				int keyType, const Tcl_HashKeyType *typePtr);
+/* 424 */
+EXTERN void		Tcl_InitObjHashTable(Tcl_HashTable *tablePtr);
+/* 425 */
+EXTERN ClientData	Tcl_CommandTraceInfo(Tcl_Interp *interp,
+				const char *varName, int flags,
+				Tcl_CommandTraceProc *procPtr,
+				ClientData prevClientData);
+/* 426 */
+EXTERN int		Tcl_TraceCommand(Tcl_Interp *interp,
+				const char *varName, int flags,
+				Tcl_CommandTraceProc *proc,
+				ClientData clientData);
+/* 427 */
+EXTERN void		Tcl_UntraceCommand(Tcl_Interp *interp,
+				const char *varName, int flags,
+				Tcl_CommandTraceProc *proc,
+				ClientData clientData);
+/* 428 */
+EXTERN char *		Tcl_AttemptAlloc(unsigned int size);
+/* 429 */
+EXTERN char *		Tcl_AttemptDbCkalloc(unsigned int size,
+				const char *file, int line);
+/* 430 */
+EXTERN char *		Tcl_AttemptRealloc(char *ptr, unsigned int size);
+/* 431 */
+EXTERN char *		Tcl_AttemptDbCkrealloc(char *ptr, unsigned int size,
+				const char *file, int line);
+/* 432 */
+EXTERN int		Tcl_AttemptSetObjLength(Tcl_Obj *objPtr, int length);
+/* 433 */
+EXTERN Tcl_ThreadId	Tcl_GetChannelThread(Tcl_Channel channel);
+/* 434 */
+EXTERN Tcl_UniChar *	Tcl_GetUnicodeFromObj(Tcl_Obj *objPtr,
+				int *lengthPtr);
+/* 435 */
+EXTERN int		Tcl_GetMathFuncInfo(Tcl_Interp *interp,
+				const char *name, int *numArgsPtr,
+				Tcl_ValueType **argTypesPtr,
+				Tcl_MathProc **procPtr,
+				ClientData *clientDataPtr);
+/* 436 */
+EXTERN Tcl_Obj *	Tcl_ListMathFuncs(Tcl_Interp *interp,
+				const char *pattern);
+/* 437 */
+EXTERN Tcl_Obj *	Tcl_SubstObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags);
+/* 438 */
+EXTERN int		Tcl_DetachChannel(Tcl_Interp *interp,
+				Tcl_Channel channel);
+/* 439 */
+EXTERN int		Tcl_IsStandardChannel(Tcl_Channel channel);
+/* 440 */
+EXTERN int		Tcl_FSCopyFile(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr);
+/* 441 */
+EXTERN int		Tcl_FSCopyDirectory(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr);
+/* 442 */
+EXTERN int		Tcl_FSCreateDirectory(Tcl_Obj *pathPtr);
+/* 443 */
+EXTERN int		Tcl_FSDeleteFile(Tcl_Obj *pathPtr);
+/* 444 */
+EXTERN int		Tcl_FSLoadFile(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+				const char *sym1, const char *sym2,
+				Tcl_PackageInitProc **proc1Ptr,
+				Tcl_PackageInitProc **proc2Ptr,
+				Tcl_LoadHandle *handlePtr,
+				Tcl_FSUnloadFileProc **unloadProcPtr);
+/* 445 */
+EXTERN int		Tcl_FSMatchInDirectory(Tcl_Interp *interp,
+				Tcl_Obj *result, Tcl_Obj *pathPtr,
+				const char *pattern, Tcl_GlobTypeData *types);
+/* 446 */
+EXTERN Tcl_Obj *	Tcl_FSLink(Tcl_Obj *pathPtr, Tcl_Obj *toPtr,
+				int linkAction);
+/* 447 */
+EXTERN int		Tcl_FSRemoveDirectory(Tcl_Obj *pathPtr,
+				int recursive, Tcl_Obj **errorPtr);
+/* 448 */
+EXTERN int		Tcl_FSRenameFile(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr);
+/* 449 */
+EXTERN int		Tcl_FSLstat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+/* 450 */
+EXTERN int		Tcl_FSUtime(Tcl_Obj *pathPtr, struct utimbuf *tval);
+/* 451 */
+EXTERN int		Tcl_FSFileAttrsGet(Tcl_Interp *interp, int index,
+				Tcl_Obj *pathPtr, Tcl_Obj **objPtrRef);
+/* 452 */
+EXTERN int		Tcl_FSFileAttrsSet(Tcl_Interp *interp, int index,
+				Tcl_Obj *pathPtr, Tcl_Obj *objPtr);
+/* 453 */
+EXTERN const char *CONST86 * Tcl_FSFileAttrStrings(Tcl_Obj *pathPtr,
+				Tcl_Obj **objPtrRef);
+/* 454 */
+EXTERN int		Tcl_FSStat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+/* 455 */
+EXTERN int		Tcl_FSAccess(Tcl_Obj *pathPtr, int mode);
+/* 456 */
+EXTERN Tcl_Channel	Tcl_FSOpenFileChannel(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr, const char *modeString,
+				int permissions);
+/* 457 */
+EXTERN Tcl_Obj *	Tcl_FSGetCwd(Tcl_Interp *interp);
+/* 458 */
+EXTERN int		Tcl_FSChdir(Tcl_Obj *pathPtr);
+/* 459 */
+EXTERN int		Tcl_FSConvertToPathType(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr);
+/* 460 */
+EXTERN Tcl_Obj *	Tcl_FSJoinPath(Tcl_Obj *listObj, int elements);
+/* 461 */
+EXTERN Tcl_Obj *	Tcl_FSSplitPath(Tcl_Obj *pathPtr, int *lenPtr);
+/* 462 */
+EXTERN int		Tcl_FSEqualPaths(Tcl_Obj *firstPtr,
+				Tcl_Obj *secondPtr);
+/* 463 */
+EXTERN Tcl_Obj *	Tcl_FSGetNormalizedPath(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr);
+/* 464 */
+EXTERN Tcl_Obj *	Tcl_FSJoinToPath(Tcl_Obj *pathPtr, int objc,
+				Tcl_Obj *const objv[]);
+/* 465 */
+EXTERN ClientData	Tcl_FSGetInternalRep(Tcl_Obj *pathPtr,
+				const Tcl_Filesystem *fsPtr);
+/* 466 */
+EXTERN Tcl_Obj *	Tcl_FSGetTranslatedPath(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr);
+/* 467 */
+EXTERN int		Tcl_FSEvalFile(Tcl_Interp *interp, Tcl_Obj *fileName);
+/* 468 */
+EXTERN Tcl_Obj *	Tcl_FSNewNativePath(
+				const Tcl_Filesystem *fromFilesystem,
+				ClientData clientData);
+/* 469 */
+EXTERN const void *	Tcl_FSGetNativePath(Tcl_Obj *pathPtr);
+/* 470 */
+EXTERN Tcl_Obj *	Tcl_FSFileSystemInfo(Tcl_Obj *pathPtr);
+/* 471 */
+EXTERN Tcl_Obj *	Tcl_FSPathSeparator(Tcl_Obj *pathPtr);
+/* 472 */
+EXTERN Tcl_Obj *	Tcl_FSListVolumes(void);
+/* 473 */
+EXTERN int		Tcl_FSRegister(ClientData clientData,
+				const Tcl_Filesystem *fsPtr);
+/* 474 */
+EXTERN int		Tcl_FSUnregister(const Tcl_Filesystem *fsPtr);
+/* 475 */
+EXTERN ClientData	Tcl_FSData(const Tcl_Filesystem *fsPtr);
+/* 476 */
+EXTERN const char *	Tcl_FSGetTranslatedStringPath(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr);
+/* 477 */
+EXTERN CONST86 Tcl_Filesystem * Tcl_FSGetFileSystemForPath(Tcl_Obj *pathPtr);
+/* 478 */
+EXTERN Tcl_PathType	Tcl_FSGetPathType(Tcl_Obj *pathPtr);
+/* 479 */
+EXTERN int		Tcl_OutputBuffered(Tcl_Channel chan);
+/* 480 */
+EXTERN void		Tcl_FSMountsChanged(const Tcl_Filesystem *fsPtr);
+/* 481 */
+EXTERN int		Tcl_EvalTokensStandard(Tcl_Interp *interp,
+				Tcl_Token *tokenPtr, int count);
+/* 482 */
+EXTERN void		Tcl_GetTime(Tcl_Time *timeBuf);
+/* 483 */
+EXTERN Tcl_Trace	Tcl_CreateObjTrace(Tcl_Interp *interp, int level,
+				int flags, Tcl_CmdObjTraceProc *objProc,
+				ClientData clientData,
+				Tcl_CmdObjTraceDeleteProc *delProc);
+/* 484 */
+EXTERN int		Tcl_GetCommandInfoFromToken(Tcl_Command token,
+				Tcl_CmdInfo *infoPtr);
+/* 485 */
+EXTERN int		Tcl_SetCommandInfoFromToken(Tcl_Command token,
+				const Tcl_CmdInfo *infoPtr);
+/* 486 */
+EXTERN Tcl_Obj *	Tcl_DbNewWideIntObj(Tcl_WideInt wideValue,
+				const char *file, int line);
+/* 487 */
+EXTERN int		Tcl_GetWideIntFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, Tcl_WideInt *widePtr);
+/* 488 */
+EXTERN Tcl_Obj *	Tcl_NewWideIntObj(Tcl_WideInt wideValue);
+/* 489 */
+EXTERN void		Tcl_SetWideIntObj(Tcl_Obj *objPtr,
+				Tcl_WideInt wideValue);
+/* 490 */
+EXTERN Tcl_StatBuf *	Tcl_AllocStatBuf(void);
+/* 491 */
+EXTERN Tcl_WideInt	Tcl_Seek(Tcl_Channel chan, Tcl_WideInt offset,
+				int mode);
+/* 492 */
+EXTERN Tcl_WideInt	Tcl_Tell(Tcl_Channel chan);
+/* 493 */
+EXTERN Tcl_DriverWideSeekProc * Tcl_ChannelWideSeekProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 494 */
+EXTERN int		Tcl_DictObjPut(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+				Tcl_Obj *keyPtr, Tcl_Obj *valuePtr);
+/* 495 */
+EXTERN int		Tcl_DictObjGet(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+				Tcl_Obj *keyPtr, Tcl_Obj **valuePtrPtr);
+/* 496 */
+EXTERN int		Tcl_DictObjRemove(Tcl_Interp *interp,
+				Tcl_Obj *dictPtr, Tcl_Obj *keyPtr);
+/* 497 */
+EXTERN int		Tcl_DictObjSize(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+				int *sizePtr);
+/* 498 */
+EXTERN int		Tcl_DictObjFirst(Tcl_Interp *interp,
+				Tcl_Obj *dictPtr, Tcl_DictSearch *searchPtr,
+				Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr,
+				int *donePtr);
+/* 499 */
+EXTERN void		Tcl_DictObjNext(Tcl_DictSearch *searchPtr,
+				Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr,
+				int *donePtr);
+/* 500 */
+EXTERN void		Tcl_DictObjDone(Tcl_DictSearch *searchPtr);
+/* 501 */
+EXTERN int		Tcl_DictObjPutKeyList(Tcl_Interp *interp,
+				Tcl_Obj *dictPtr, int keyc,
+				Tcl_Obj *const *keyv, Tcl_Obj *valuePtr);
+/* 502 */
+EXTERN int		Tcl_DictObjRemoveKeyList(Tcl_Interp *interp,
+				Tcl_Obj *dictPtr, int keyc,
+				Tcl_Obj *const *keyv);
+/* 503 */
+EXTERN Tcl_Obj *	Tcl_NewDictObj(void);
+/* 504 */
+EXTERN Tcl_Obj *	Tcl_DbNewDictObj(const char *file, int line);
+/* 505 */
+EXTERN void		Tcl_RegisterConfig(Tcl_Interp *interp,
+				const char *pkgName,
+				const Tcl_Config *configuration,
+				const char *valEncoding);
+/* 506 */
+EXTERN Tcl_Namespace *	Tcl_CreateNamespace(Tcl_Interp *interp,
+				const char *name, ClientData clientData,
+				Tcl_NamespaceDeleteProc *deleteProc);
+/* 507 */
+EXTERN void		Tcl_DeleteNamespace(Tcl_Namespace *nsPtr);
+/* 508 */
+EXTERN int		Tcl_AppendExportList(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr, Tcl_Obj *objPtr);
+/* 509 */
+EXTERN int		Tcl_Export(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+				const char *pattern, int resetListFirst);
+/* 510 */
+EXTERN int		Tcl_Import(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+				const char *pattern, int allowOverwrite);
+/* 511 */
+EXTERN int		Tcl_ForgetImport(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr, const char *pattern);
+/* 512 */
+EXTERN Tcl_Namespace *	Tcl_GetCurrentNamespace(Tcl_Interp *interp);
+/* 513 */
+EXTERN Tcl_Namespace *	Tcl_GetGlobalNamespace(Tcl_Interp *interp);
+/* 514 */
+EXTERN Tcl_Namespace *	Tcl_FindNamespace(Tcl_Interp *interp,
+				const char *name,
+				Tcl_Namespace *contextNsPtr, int flags);
+/* 515 */
+EXTERN Tcl_Command	Tcl_FindCommand(Tcl_Interp *interp, const char *name,
+				Tcl_Namespace *contextNsPtr, int flags);
+/* 516 */
+EXTERN Tcl_Command	Tcl_GetCommandFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 517 */
+EXTERN void		Tcl_GetCommandFullName(Tcl_Interp *interp,
+				Tcl_Command command, Tcl_Obj *objPtr);
+/* 518 */
+EXTERN int		Tcl_FSEvalFileEx(Tcl_Interp *interp,
+				Tcl_Obj *fileName, const char *encodingName);
+/* 519 */
+EXTERN Tcl_ExitProc *	Tcl_SetExitProc(TCL_NORETURN1 Tcl_ExitProc *proc);
+/* 520 */
+EXTERN void		Tcl_LimitAddHandler(Tcl_Interp *interp, int type,
+				Tcl_LimitHandlerProc *handlerProc,
+				ClientData clientData,
+				Tcl_LimitHandlerDeleteProc *deleteProc);
+/* 521 */
+EXTERN void		Tcl_LimitRemoveHandler(Tcl_Interp *interp, int type,
+				Tcl_LimitHandlerProc *handlerProc,
+				ClientData clientData);
+/* 522 */
+EXTERN int		Tcl_LimitReady(Tcl_Interp *interp);
+/* 523 */
+EXTERN int		Tcl_LimitCheck(Tcl_Interp *interp);
+/* 524 */
+EXTERN int		Tcl_LimitExceeded(Tcl_Interp *interp);
+/* 525 */
+EXTERN void		Tcl_LimitSetCommands(Tcl_Interp *interp,
+				int commandLimit);
+/* 526 */
+EXTERN void		Tcl_LimitSetTime(Tcl_Interp *interp,
+				Tcl_Time *timeLimitPtr);
+/* 527 */
+EXTERN void		Tcl_LimitSetGranularity(Tcl_Interp *interp, int type,
+				int granularity);
+/* 528 */
+EXTERN int		Tcl_LimitTypeEnabled(Tcl_Interp *interp, int type);
+/* 529 */
+EXTERN int		Tcl_LimitTypeExceeded(Tcl_Interp *interp, int type);
+/* 530 */
+EXTERN void		Tcl_LimitTypeSet(Tcl_Interp *interp, int type);
+/* 531 */
+EXTERN void		Tcl_LimitTypeReset(Tcl_Interp *interp, int type);
+/* 532 */
+EXTERN int		Tcl_LimitGetCommands(Tcl_Interp *interp);
+/* 533 */
+EXTERN void		Tcl_LimitGetTime(Tcl_Interp *interp,
+				Tcl_Time *timeLimitPtr);
+/* 534 */
+EXTERN int		Tcl_LimitGetGranularity(Tcl_Interp *interp, int type);
+/* 535 */
+EXTERN Tcl_InterpState	Tcl_SaveInterpState(Tcl_Interp *interp, int status);
+/* 536 */
+EXTERN int		Tcl_RestoreInterpState(Tcl_Interp *interp,
+				Tcl_InterpState state);
+/* 537 */
+EXTERN void		Tcl_DiscardInterpState(Tcl_InterpState state);
+/* 538 */
+EXTERN int		Tcl_SetReturnOptions(Tcl_Interp *interp,
+				Tcl_Obj *options);
+/* 539 */
+EXTERN Tcl_Obj *	Tcl_GetReturnOptions(Tcl_Interp *interp, int result);
+/* 540 */
+EXTERN int		Tcl_IsEnsemble(Tcl_Command token);
+/* 541 */
+EXTERN Tcl_Command	Tcl_CreateEnsemble(Tcl_Interp *interp,
+				const char *name,
+				Tcl_Namespace *namespacePtr, int flags);
+/* 542 */
+EXTERN Tcl_Command	Tcl_FindEnsemble(Tcl_Interp *interp,
+				Tcl_Obj *cmdNameObj, int flags);
+/* 543 */
+EXTERN int		Tcl_SetEnsembleSubcommandList(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj *subcmdList);
+/* 544 */
+EXTERN int		Tcl_SetEnsembleMappingDict(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj *mapDict);
+/* 545 */
+EXTERN int		Tcl_SetEnsembleUnknownHandler(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj *unknownList);
+/* 546 */
+EXTERN int		Tcl_SetEnsembleFlags(Tcl_Interp *interp,
+				Tcl_Command token, int flags);
+/* 547 */
+EXTERN int		Tcl_GetEnsembleSubcommandList(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj **subcmdListPtr);
+/* 548 */
+EXTERN int		Tcl_GetEnsembleMappingDict(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj **mapDictPtr);
+/* 549 */
+EXTERN int		Tcl_GetEnsembleUnknownHandler(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj **unknownListPtr);
+/* 550 */
+EXTERN int		Tcl_GetEnsembleFlags(Tcl_Interp *interp,
+				Tcl_Command token, int *flagsPtr);
+/* 551 */
+EXTERN int		Tcl_GetEnsembleNamespace(Tcl_Interp *interp,
+				Tcl_Command token,
+				Tcl_Namespace **namespacePtrPtr);
+/* 552 */
+EXTERN void		Tcl_SetTimeProc(Tcl_GetTimeProc *getProc,
+				Tcl_ScaleTimeProc *scaleProc,
+				ClientData clientData);
+/* 553 */
+EXTERN void		Tcl_QueryTimeProc(Tcl_GetTimeProc **getProc,
+				Tcl_ScaleTimeProc **scaleProc,
+				ClientData *clientData);
+/* 554 */
+EXTERN Tcl_DriverThreadActionProc * Tcl_ChannelThreadActionProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 555 */
+EXTERN Tcl_Obj *	Tcl_NewBignumObj(mp_int *value);
+/* 556 */
+EXTERN Tcl_Obj *	Tcl_DbNewBignumObj(mp_int *value, const char *file,
+				int line);
+/* 557 */
+EXTERN void		Tcl_SetBignumObj(Tcl_Obj *obj, mp_int *value);
+/* 558 */
+EXTERN int		Tcl_GetBignumFromObj(Tcl_Interp *interp,
+				Tcl_Obj *obj, mp_int *value);
+/* 559 */
+EXTERN int		Tcl_TakeBignumFromObj(Tcl_Interp *interp,
+				Tcl_Obj *obj, mp_int *value);
+/* 560 */
+EXTERN int		Tcl_TruncateChannel(Tcl_Channel chan,
+				Tcl_WideInt length);
+/* 561 */
+EXTERN Tcl_DriverTruncateProc * Tcl_ChannelTruncateProc(
+				const Tcl_ChannelType *chanTypePtr);
+/* 562 */
+EXTERN void		Tcl_SetChannelErrorInterp(Tcl_Interp *interp,
+				Tcl_Obj *msg);
+/* 563 */
+EXTERN void		Tcl_GetChannelErrorInterp(Tcl_Interp *interp,
+				Tcl_Obj **msg);
+/* 564 */
+EXTERN void		Tcl_SetChannelError(Tcl_Channel chan, Tcl_Obj *msg);
+/* 565 */
+EXTERN void		Tcl_GetChannelError(Tcl_Channel chan, Tcl_Obj **msg);
+/* 566 */
+EXTERN int		Tcl_InitBignumFromDouble(Tcl_Interp *interp,
+				double initval, mp_int *toInit);
+/* 567 */
+EXTERN Tcl_Obj *	Tcl_GetNamespaceUnknownHandler(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr);
+/* 568 */
+EXTERN int		Tcl_SetNamespaceUnknownHandler(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr, Tcl_Obj *handlerPtr);
+/* 569 */
+EXTERN int		Tcl_GetEncodingFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, Tcl_Encoding *encodingPtr);
+/* 570 */
+EXTERN Tcl_Obj *	Tcl_GetEncodingSearchPath(void);
+/* 571 */
+EXTERN int		Tcl_SetEncodingSearchPath(Tcl_Obj *searchPath);
+/* 572 */
+EXTERN const char *	Tcl_GetEncodingNameFromEnvironment(
+				Tcl_DString *bufPtr);
+/* 573 */
+EXTERN int		Tcl_PkgRequireProc(Tcl_Interp *interp,
+				const char *name, int objc,
+				Tcl_Obj *const objv[], void *clientDataPtr);
+/* 574 */
+EXTERN void		Tcl_AppendObjToErrorInfo(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 575 */
+EXTERN void		Tcl_AppendLimitedToObj(Tcl_Obj *objPtr,
+				const char *bytes, int length, int limit,
+				const char *ellipsis);
+/* 576 */
+EXTERN Tcl_Obj *	Tcl_Format(Tcl_Interp *interp, const char *format,
+				int objc, Tcl_Obj *const objv[]);
+/* 577 */
+EXTERN int		Tcl_AppendFormatToObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, const char *format,
+				int objc, Tcl_Obj *const objv[]);
+/* 578 */
+EXTERN Tcl_Obj *	Tcl_ObjPrintf(const char *format, ...) TCL_FORMAT_PRINTF(1, 2);
+/* 579 */
+EXTERN void		Tcl_AppendPrintfToObj(Tcl_Obj *objPtr,
+				const char *format, ...) TCL_FORMAT_PRINTF(2, 3);
+/* 580 */
+EXTERN int		Tcl_CancelEval(Tcl_Interp *interp,
+				Tcl_Obj *resultObjPtr, ClientData clientData,
+				int flags);
+/* 581 */
+EXTERN int		Tcl_Canceled(Tcl_Interp *interp, int flags);
+/* 582 */
+EXTERN int		Tcl_CreatePipe(Tcl_Interp *interp,
+				Tcl_Channel *rchan, Tcl_Channel *wchan,
+				int flags);
+/* 583 */
+EXTERN Tcl_Command	Tcl_NRCreateCommand(Tcl_Interp *interp,
+				const char *cmdName, Tcl_ObjCmdProc *proc,
+				Tcl_ObjCmdProc *nreProc,
+				ClientData clientData,
+				Tcl_CmdDeleteProc *deleteProc);
+/* 584 */
+EXTERN int		Tcl_NREvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags);
+/* 585 */
+EXTERN int		Tcl_NREvalObjv(Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[], int flags);
+/* 586 */
+EXTERN int		Tcl_NRCmdSwap(Tcl_Interp *interp, Tcl_Command cmd,
+				int objc, Tcl_Obj *const objv[], int flags);
+/* 587 */
+EXTERN void		Tcl_NRAddCallback(Tcl_Interp *interp,
+				Tcl_NRPostProc *postProcPtr,
+				ClientData data0, ClientData data1,
+				ClientData data2, ClientData data3);
+/* 588 */
+EXTERN int		Tcl_NRCallObjProc(Tcl_Interp *interp,
+				Tcl_ObjCmdProc *objProc,
+				ClientData clientData, int objc,
+				Tcl_Obj *const objv[]);
+/* 589 */
+EXTERN unsigned		Tcl_GetFSDeviceFromStat(const Tcl_StatBuf *statPtr);
+/* 590 */
+EXTERN unsigned		Tcl_GetFSInodeFromStat(const Tcl_StatBuf *statPtr);
+/* 591 */
+EXTERN unsigned		Tcl_GetModeFromStat(const Tcl_StatBuf *statPtr);
+/* 592 */
+EXTERN int		Tcl_GetLinkCountFromStat(const Tcl_StatBuf *statPtr);
+/* 593 */
+EXTERN int		Tcl_GetUserIdFromStat(const Tcl_StatBuf *statPtr);
+/* 594 */
+EXTERN int		Tcl_GetGroupIdFromStat(const Tcl_StatBuf *statPtr);
+/* 595 */
+EXTERN int		Tcl_GetDeviceTypeFromStat(const Tcl_StatBuf *statPtr);
+/* 596 */
+EXTERN Tcl_WideInt	Tcl_GetAccessTimeFromStat(const Tcl_StatBuf *statPtr);
+/* 597 */
+EXTERN Tcl_WideInt	Tcl_GetModificationTimeFromStat(
+				const Tcl_StatBuf *statPtr);
+/* 598 */
+EXTERN Tcl_WideInt	Tcl_GetChangeTimeFromStat(const Tcl_StatBuf *statPtr);
+/* 599 */
+EXTERN Tcl_WideUInt	Tcl_GetSizeFromStat(const Tcl_StatBuf *statPtr);
+/* 600 */
+EXTERN Tcl_WideUInt	Tcl_GetBlocksFromStat(const Tcl_StatBuf *statPtr);
+/* 601 */
+EXTERN unsigned		Tcl_GetBlockSizeFromStat(const Tcl_StatBuf *statPtr);
+/* 602 */
+EXTERN int		Tcl_SetEnsembleParameterList(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj *paramList);
+/* 603 */
+EXTERN int		Tcl_GetEnsembleParameterList(Tcl_Interp *interp,
+				Tcl_Command token, Tcl_Obj **paramListPtr);
+/* 604 */
+EXTERN int		Tcl_ParseArgsObjv(Tcl_Interp *interp,
+				const Tcl_ArgvInfo *argTable, int *objcPtr,
+				Tcl_Obj *const *objv, Tcl_Obj ***remObjv);
+/* 605 */
+EXTERN int		Tcl_GetErrorLine(Tcl_Interp *interp);
+/* 606 */
+EXTERN void		Tcl_SetErrorLine(Tcl_Interp *interp, int lineNum);
+/* 607 */
+EXTERN void		Tcl_TransferResult(Tcl_Interp *sourceInterp,
+				int result, Tcl_Interp *targetInterp);
+/* 608 */
+EXTERN int		Tcl_InterpActive(Tcl_Interp *interp);
+/* 609 */
+EXTERN void		Tcl_BackgroundException(Tcl_Interp *interp, int code);
+/* 610 */
+EXTERN int		Tcl_ZlibDeflate(Tcl_Interp *interp, int format,
+				Tcl_Obj *data, int level,
+				Tcl_Obj *gzipHeaderDictObj);
+/* 611 */
+EXTERN int		Tcl_ZlibInflate(Tcl_Interp *interp, int format,
+				Tcl_Obj *data, int buffersize,
+				Tcl_Obj *gzipHeaderDictObj);
+/* 612 */
+EXTERN unsigned int	Tcl_ZlibCRC32(unsigned int crc,
+				const unsigned char *buf, int len);
+/* 613 */
+EXTERN unsigned int	Tcl_ZlibAdler32(unsigned int adler,
+				const unsigned char *buf, int len);
+/* 614 */
+EXTERN int		Tcl_ZlibStreamInit(Tcl_Interp *interp, int mode,
+				int format, int level, Tcl_Obj *dictObj,
+				Tcl_ZlibStream *zshandle);
+/* 615 */
+EXTERN Tcl_Obj *	Tcl_ZlibStreamGetCommandName(Tcl_ZlibStream zshandle);
+/* 616 */
+EXTERN int		Tcl_ZlibStreamEof(Tcl_ZlibStream zshandle);
+/* 617 */
+EXTERN int		Tcl_ZlibStreamChecksum(Tcl_ZlibStream zshandle);
+/* 618 */
+EXTERN int		Tcl_ZlibStreamPut(Tcl_ZlibStream zshandle,
+				Tcl_Obj *data, int flush);
+/* 619 */
+EXTERN int		Tcl_ZlibStreamGet(Tcl_ZlibStream zshandle,
+				Tcl_Obj *data, int count);
+/* 620 */
+EXTERN int		Tcl_ZlibStreamClose(Tcl_ZlibStream zshandle);
+/* 621 */
+EXTERN int		Tcl_ZlibStreamReset(Tcl_ZlibStream zshandle);
+/* 622 */
+EXTERN void		Tcl_SetStartupScript(Tcl_Obj *path,
+				const char *encoding);
+/* 623 */
+EXTERN Tcl_Obj *	Tcl_GetStartupScript(const char **encodingPtr);
+/* 624 */
+EXTERN int		Tcl_CloseEx(Tcl_Interp *interp, Tcl_Channel chan,
+				int flags);
+/* 625 */
+EXTERN int		Tcl_NRExprObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				Tcl_Obj *resultPtr);
+/* 626 */
+EXTERN int		Tcl_NRSubstObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags);
+/* 627 */
+EXTERN int		Tcl_LoadFile(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+				const char *const symv[], int flags,
+				void *procPtrs, Tcl_LoadHandle *handlePtr);
+/* 628 */
+EXTERN void *		Tcl_FindSymbol(Tcl_Interp *interp,
+				Tcl_LoadHandle handle, const char *symbol);
+/* 629 */
+EXTERN int		Tcl_FSUnloadFile(Tcl_Interp *interp,
+				Tcl_LoadHandle handlePtr);
+/* 630 */
+EXTERN void		Tcl_ZlibStreamSetCompressionDictionary(
+				Tcl_ZlibStream zhandle,
+				Tcl_Obj *compressionDictionaryObj);
+/* 631 */
+EXTERN Tcl_Channel	Tcl_OpenTcpServerEx(Tcl_Interp *interp,
+				const char *service, const char *host,
+				unsigned int flags,
+				Tcl_TcpAcceptProc *acceptProc,
+				ClientData callbackData);
+
+typedef struct {
+    const struct TclPlatStubs *tclPlatStubs;
+    const struct TclIntStubs *tclIntStubs;
+    const struct TclIntPlatStubs *tclIntPlatStubs;
+} TclStubHooks;
+
+typedef struct TclStubs {
+    int magic;
+    const TclStubHooks *hooks;
+
+    int (*tcl_PkgProvideEx) (Tcl_Interp *interp, const char *name, const char *version, const void *clientData); /* 0 */
+    CONST84_RETURN char * (*tcl_PkgRequireEx) (Tcl_Interp *interp, const char *name, const char *version, int exact, void *clientDataPtr); /* 1 */
+    TCL_NORETURN1 void (*tcl_Panic) (const char *format, ...) TCL_FORMAT_PRINTF(1, 2); /* 2 */
+    char * (*tcl_Alloc) (unsigned int size); /* 3 */
+    void (*tcl_Free) (char *ptr); /* 4 */
+    char * (*tcl_Realloc) (char *ptr, unsigned int size); /* 5 */
+    char * (*tcl_DbCkalloc) (unsigned int size, const char *file, int line); /* 6 */
+    void (*tcl_DbCkfree) (char *ptr, const char *file, int line); /* 7 */
+    char * (*tcl_DbCkrealloc) (char *ptr, unsigned int size, const char *file, int line); /* 8 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    void (*tcl_CreateFileHandler) (int fd, int mask, Tcl_FileProc *proc, ClientData clientData); /* 9 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    void (*reserved9)(void);
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    void (*tcl_CreateFileHandler) (int fd, int mask, Tcl_FileProc *proc, ClientData clientData); /* 9 */
+#endif /* MACOSX */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    void (*tcl_DeleteFileHandler) (int fd); /* 10 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    void (*reserved10)(void);
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    void (*tcl_DeleteFileHandler) (int fd); /* 10 */
+#endif /* MACOSX */
+    void (*tcl_SetTimer) (const Tcl_Time *timePtr); /* 11 */
+    void (*tcl_Sleep) (int ms); /* 12 */
+    int (*tcl_WaitForEvent) (const Tcl_Time *timePtr); /* 13 */
+    int (*tcl_AppendAllObjTypes) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 14 */
+    void (*tcl_AppendStringsToObj) (Tcl_Obj *objPtr, ...); /* 15 */
+    void (*tcl_AppendToObj) (Tcl_Obj *objPtr, const char *bytes, int length); /* 16 */
+    Tcl_Obj * (*tcl_ConcatObj) (int objc, Tcl_Obj *const objv[]); /* 17 */
+    int (*tcl_ConvertToType) (Tcl_Interp *interp, Tcl_Obj *objPtr, const Tcl_ObjType *typePtr); /* 18 */
+    void (*tcl_DbDecrRefCount) (Tcl_Obj *objPtr, const char *file, int line); /* 19 */
+    void (*tcl_DbIncrRefCount) (Tcl_Obj *objPtr, const char *file, int line); /* 20 */
+    int (*tcl_DbIsShared) (Tcl_Obj *objPtr, const char *file, int line); /* 21 */
+    Tcl_Obj * (*tcl_DbNewBooleanObj) (int boolValue, const char *file, int line); /* 22 */
+    Tcl_Obj * (*tcl_DbNewByteArrayObj) (const unsigned char *bytes, int length, const char *file, int line); /* 23 */
+    Tcl_Obj * (*tcl_DbNewDoubleObj) (double doubleValue, const char *file, int line); /* 24 */
+    Tcl_Obj * (*tcl_DbNewListObj) (int objc, Tcl_Obj *const *objv, const char *file, int line); /* 25 */
+    Tcl_Obj * (*tcl_DbNewLongObj) (long longValue, const char *file, int line); /* 26 */
+    Tcl_Obj * (*tcl_DbNewObj) (const char *file, int line); /* 27 */
+    Tcl_Obj * (*tcl_DbNewStringObj) (const char *bytes, int length, const char *file, int line); /* 28 */
+    Tcl_Obj * (*tcl_DuplicateObj) (Tcl_Obj *objPtr); /* 29 */
+    void (*tclFreeObj) (Tcl_Obj *objPtr); /* 30 */
+    int (*tcl_GetBoolean) (Tcl_Interp *interp, const char *src, int *boolPtr); /* 31 */
+    int (*tcl_GetBooleanFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int *boolPtr); /* 32 */
+    unsigned char * (*tcl_GetByteArrayFromObj) (Tcl_Obj *objPtr, int *lengthPtr); /* 33 */
+    int (*tcl_GetDouble) (Tcl_Interp *interp, const char *src, double *doublePtr); /* 34 */
+    int (*tcl_GetDoubleFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, double *doublePtr); /* 35 */
+    int (*tcl_GetIndexFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, CONST84 char *const *tablePtr, const char *msg, int flags, int *indexPtr); /* 36 */
+    int (*tcl_GetInt) (Tcl_Interp *interp, const char *src, int *intPtr); /* 37 */
+    int (*tcl_GetIntFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int *intPtr); /* 38 */
+    int (*tcl_GetLongFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, long *longPtr); /* 39 */
+    CONST86 Tcl_ObjType * (*tcl_GetObjType) (const char *typeName); /* 40 */
+    char * (*tcl_GetStringFromObj) (Tcl_Obj *objPtr, int *lengthPtr); /* 41 */
+    void (*tcl_InvalidateStringRep) (Tcl_Obj *objPtr); /* 42 */
+    int (*tcl_ListObjAppendList) (Tcl_Interp *interp, Tcl_Obj *listPtr, Tcl_Obj *elemListPtr); /* 43 */
+    int (*tcl_ListObjAppendElement) (Tcl_Interp *interp, Tcl_Obj *listPtr, Tcl_Obj *objPtr); /* 44 */
+    int (*tcl_ListObjGetElements) (Tcl_Interp *interp, Tcl_Obj *listPtr, int *objcPtr, Tcl_Obj ***objvPtr); /* 45 */
+    int (*tcl_ListObjIndex) (Tcl_Interp *interp, Tcl_Obj *listPtr, int index, Tcl_Obj **objPtrPtr); /* 46 */
+    int (*tcl_ListObjLength) (Tcl_Interp *interp, Tcl_Obj *listPtr, int *lengthPtr); /* 47 */
+    int (*tcl_ListObjReplace) (Tcl_Interp *interp, Tcl_Obj *listPtr, int first, int count, int objc, Tcl_Obj *const objv[]); /* 48 */
+    Tcl_Obj * (*tcl_NewBooleanObj) (int boolValue); /* 49 */
+    Tcl_Obj * (*tcl_NewByteArrayObj) (const unsigned char *bytes, int length); /* 50 */
+    Tcl_Obj * (*tcl_NewDoubleObj) (double doubleValue); /* 51 */
+    Tcl_Obj * (*tcl_NewIntObj) (int intValue); /* 52 */
+    Tcl_Obj * (*tcl_NewListObj) (int objc, Tcl_Obj *const objv[]); /* 53 */
+    Tcl_Obj * (*tcl_NewLongObj) (long longValue); /* 54 */
+    Tcl_Obj * (*tcl_NewObj) (void); /* 55 */
+    Tcl_Obj * (*tcl_NewStringObj) (const char *bytes, int length); /* 56 */
+    void (*tcl_SetBooleanObj) (Tcl_Obj *objPtr, int boolValue); /* 57 */
+    unsigned char * (*tcl_SetByteArrayLength) (Tcl_Obj *objPtr, int length); /* 58 */
+    void (*tcl_SetByteArrayObj) (Tcl_Obj *objPtr, const unsigned char *bytes, int length); /* 59 */
+    void (*tcl_SetDoubleObj) (Tcl_Obj *objPtr, double doubleValue); /* 60 */
+    void (*tcl_SetIntObj) (Tcl_Obj *objPtr, int intValue); /* 61 */
+    void (*tcl_SetListObj) (Tcl_Obj *objPtr, int objc, Tcl_Obj *const objv[]); /* 62 */
+    void (*tcl_SetLongObj) (Tcl_Obj *objPtr, long longValue); /* 63 */
+    void (*tcl_SetObjLength) (Tcl_Obj *objPtr, int length); /* 64 */
+    void (*tcl_SetStringObj) (Tcl_Obj *objPtr, const char *bytes, int length); /* 65 */
+    void (*tcl_AddErrorInfo) (Tcl_Interp *interp, const char *message); /* 66 */
+    void (*tcl_AddObjErrorInfo) (Tcl_Interp *interp, const char *message, int length); /* 67 */
+    void (*tcl_AllowExceptions) (Tcl_Interp *interp); /* 68 */
+    void (*tcl_AppendElement) (Tcl_Interp *interp, const char *element); /* 69 */
+    void (*tcl_AppendResult) (Tcl_Interp *interp, ...); /* 70 */
+    Tcl_AsyncHandler (*tcl_AsyncCreate) (Tcl_AsyncProc *proc, ClientData clientData); /* 71 */
+    void (*tcl_AsyncDelete) (Tcl_AsyncHandler async); /* 72 */
+    int (*tcl_AsyncInvoke) (Tcl_Interp *interp, int code); /* 73 */
+    void (*tcl_AsyncMark) (Tcl_AsyncHandler async); /* 74 */
+    int (*tcl_AsyncReady) (void); /* 75 */
+    void (*tcl_BackgroundError) (Tcl_Interp *interp); /* 76 */
+    char (*tcl_Backslash) (const char *src, int *readPtr); /* 77 */
+    int (*tcl_BadChannelOption) (Tcl_Interp *interp, const char *optionName, const char *optionList); /* 78 */
+    void (*tcl_CallWhenDeleted) (Tcl_Interp *interp, Tcl_InterpDeleteProc *proc, ClientData clientData); /* 79 */
+    void (*tcl_CancelIdleCall) (Tcl_IdleProc *idleProc, ClientData clientData); /* 80 */
+    int (*tcl_Close) (Tcl_Interp *interp, Tcl_Channel chan); /* 81 */
+    int (*tcl_CommandComplete) (const char *cmd); /* 82 */
+    char * (*tcl_Concat) (int argc, CONST84 char *const *argv); /* 83 */
+    int (*tcl_ConvertElement) (const char *src, char *dst, int flags); /* 84 */
+    int (*tcl_ConvertCountedElement) (const char *src, int length, char *dst, int flags); /* 85 */
+    int (*tcl_CreateAlias) (Tcl_Interp *slave, const char *slaveCmd, Tcl_Interp *target, const char *targetCmd, int argc, CONST84 char *const *argv); /* 86 */
+    int (*tcl_CreateAliasObj) (Tcl_Interp *slave, const char *slaveCmd, Tcl_Interp *target, const char *targetCmd, int objc, Tcl_Obj *const objv[]); /* 87 */
+    Tcl_Channel (*tcl_CreateChannel) (const Tcl_ChannelType *typePtr, const char *chanName, ClientData instanceData, int mask); /* 88 */
+    void (*tcl_CreateChannelHandler) (Tcl_Channel chan, int mask, Tcl_ChannelProc *proc, ClientData clientData); /* 89 */
+    void (*tcl_CreateCloseHandler) (Tcl_Channel chan, Tcl_CloseProc *proc, ClientData clientData); /* 90 */
+    Tcl_Command (*tcl_CreateCommand) (Tcl_Interp *interp, const char *cmdName, Tcl_CmdProc *proc, ClientData clientData, Tcl_CmdDeleteProc *deleteProc); /* 91 */
+    void (*tcl_CreateEventSource) (Tcl_EventSetupProc *setupProc, Tcl_EventCheckProc *checkProc, ClientData clientData); /* 92 */
+    void (*tcl_CreateExitHandler) (Tcl_ExitProc *proc, ClientData clientData); /* 93 */
+    Tcl_Interp * (*tcl_CreateInterp) (void); /* 94 */
+    void (*tcl_CreateMathFunc) (Tcl_Interp *interp, const char *name, int numArgs, Tcl_ValueType *argTypes, Tcl_MathProc *proc, ClientData clientData); /* 95 */
+    Tcl_Command (*tcl_CreateObjCommand) (Tcl_Interp *interp, const char *cmdName, Tcl_ObjCmdProc *proc, ClientData clientData, Tcl_CmdDeleteProc *deleteProc); /* 96 */
+    Tcl_Interp * (*tcl_CreateSlave) (Tcl_Interp *interp, const char *slaveName, int isSafe); /* 97 */
+    Tcl_TimerToken (*tcl_CreateTimerHandler) (int milliseconds, Tcl_TimerProc *proc, ClientData clientData); /* 98 */
+    Tcl_Trace (*tcl_CreateTrace) (Tcl_Interp *interp, int level, Tcl_CmdTraceProc *proc, ClientData clientData); /* 99 */
+    void (*tcl_DeleteAssocData) (Tcl_Interp *interp, const char *name); /* 100 */
+    void (*tcl_DeleteChannelHandler) (Tcl_Channel chan, Tcl_ChannelProc *proc, ClientData clientData); /* 101 */
+    void (*tcl_DeleteCloseHandler) (Tcl_Channel chan, Tcl_CloseProc *proc, ClientData clientData); /* 102 */
+    int (*tcl_DeleteCommand) (Tcl_Interp *interp, const char *cmdName); /* 103 */
+    int (*tcl_DeleteCommandFromToken) (Tcl_Interp *interp, Tcl_Command command); /* 104 */
+    void (*tcl_DeleteEvents) (Tcl_EventDeleteProc *proc, ClientData clientData); /* 105 */
+    void (*tcl_DeleteEventSource) (Tcl_EventSetupProc *setupProc, Tcl_EventCheckProc *checkProc, ClientData clientData); /* 106 */
+    void (*tcl_DeleteExitHandler) (Tcl_ExitProc *proc, ClientData clientData); /* 107 */
+    void (*tcl_DeleteHashEntry) (Tcl_HashEntry *entryPtr); /* 108 */
+    void (*tcl_DeleteHashTable) (Tcl_HashTable *tablePtr); /* 109 */
+    void (*tcl_DeleteInterp) (Tcl_Interp *interp); /* 110 */
+    void (*tcl_DetachPids) (int numPids, Tcl_Pid *pidPtr); /* 111 */
+    void (*tcl_DeleteTimerHandler) (Tcl_TimerToken token); /* 112 */
+    void (*tcl_DeleteTrace) (Tcl_Interp *interp, Tcl_Trace trace); /* 113 */
+    void (*tcl_DontCallWhenDeleted) (Tcl_Interp *interp, Tcl_InterpDeleteProc *proc, ClientData clientData); /* 114 */
+    int (*tcl_DoOneEvent) (int flags); /* 115 */
+    void (*tcl_DoWhenIdle) (Tcl_IdleProc *proc, ClientData clientData); /* 116 */
+    char * (*tcl_DStringAppend) (Tcl_DString *dsPtr, const char *bytes, int length); /* 117 */
+    char * (*tcl_DStringAppendElement) (Tcl_DString *dsPtr, const char *element); /* 118 */
+    void (*tcl_DStringEndSublist) (Tcl_DString *dsPtr); /* 119 */
+    void (*tcl_DStringFree) (Tcl_DString *dsPtr); /* 120 */
+    void (*tcl_DStringGetResult) (Tcl_Interp *interp, Tcl_DString *dsPtr); /* 121 */
+    void (*tcl_DStringInit) (Tcl_DString *dsPtr); /* 122 */
+    void (*tcl_DStringResult) (Tcl_Interp *interp, Tcl_DString *dsPtr); /* 123 */
+    void (*tcl_DStringSetLength) (Tcl_DString *dsPtr, int length); /* 124 */
+    void (*tcl_DStringStartSublist) (Tcl_DString *dsPtr); /* 125 */
+    int (*tcl_Eof) (Tcl_Channel chan); /* 126 */
+    CONST84_RETURN char * (*tcl_ErrnoId) (void); /* 127 */
+    CONST84_RETURN char * (*tcl_ErrnoMsg) (int err); /* 128 */
+    int (*tcl_Eval) (Tcl_Interp *interp, const char *script); /* 129 */
+    int (*tcl_EvalFile) (Tcl_Interp *interp, const char *fileName); /* 130 */
+    int (*tcl_EvalObj) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 131 */
+    void (*tcl_EventuallyFree) (ClientData clientData, Tcl_FreeProc *freeProc); /* 132 */
+    TCL_NORETURN1 void (*tcl_Exit) (int status); /* 133 */
+    int (*tcl_ExposeCommand) (Tcl_Interp *interp, const char *hiddenCmdToken, const char *cmdName); /* 134 */
+    int (*tcl_ExprBoolean) (Tcl_Interp *interp, const char *expr, int *ptr); /* 135 */
+    int (*tcl_ExprBooleanObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int *ptr); /* 136 */
+    int (*tcl_ExprDouble) (Tcl_Interp *interp, const char *expr, double *ptr); /* 137 */
+    int (*tcl_ExprDoubleObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, double *ptr); /* 138 */
+    int (*tcl_ExprLong) (Tcl_Interp *interp, const char *expr, long *ptr); /* 139 */
+    int (*tcl_ExprLongObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, long *ptr); /* 140 */
+    int (*tcl_ExprObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_Obj **resultPtrPtr); /* 141 */
+    int (*tcl_ExprString) (Tcl_Interp *interp, const char *expr); /* 142 */
+    void (*tcl_Finalize) (void); /* 143 */
+    void (*tcl_FindExecutable) (const char *argv0); /* 144 */
+    Tcl_HashEntry * (*tcl_FirstHashEntry) (Tcl_HashTable *tablePtr, Tcl_HashSearch *searchPtr); /* 145 */
+    int (*tcl_Flush) (Tcl_Channel chan); /* 146 */
+    void (*tcl_FreeResult) (Tcl_Interp *interp); /* 147 */
+    int (*tcl_GetAlias) (Tcl_Interp *interp, const char *slaveCmd, Tcl_Interp **targetInterpPtr, CONST84 char **targetCmdPtr, int *argcPtr, CONST84 char ***argvPtr); /* 148 */
+    int (*tcl_GetAliasObj) (Tcl_Interp *interp, const char *slaveCmd, Tcl_Interp **targetInterpPtr, CONST84 char **targetCmdPtr, int *objcPtr, Tcl_Obj ***objv); /* 149 */
+    ClientData (*tcl_GetAssocData) (Tcl_Interp *interp, const char *name, Tcl_InterpDeleteProc **procPtr); /* 150 */
+    Tcl_Channel (*tcl_GetChannel) (Tcl_Interp *interp, const char *chanName, int *modePtr); /* 151 */
+    int (*tcl_GetChannelBufferSize) (Tcl_Channel chan); /* 152 */
+    int (*tcl_GetChannelHandle) (Tcl_Channel chan, int direction, ClientData *handlePtr); /* 153 */
+    ClientData (*tcl_GetChannelInstanceData) (Tcl_Channel chan); /* 154 */
+    int (*tcl_GetChannelMode) (Tcl_Channel chan); /* 155 */
+    CONST84_RETURN char * (*tcl_GetChannelName) (Tcl_Channel chan); /* 156 */
+    int (*tcl_GetChannelOption) (Tcl_Interp *interp, Tcl_Channel chan, const char *optionName, Tcl_DString *dsPtr); /* 157 */
+    CONST86 Tcl_ChannelType * (*tcl_GetChannelType) (Tcl_Channel chan); /* 158 */
+    int (*tcl_GetCommandInfo) (Tcl_Interp *interp, const char *cmdName, Tcl_CmdInfo *infoPtr); /* 159 */
+    CONST84_RETURN char * (*tcl_GetCommandName) (Tcl_Interp *interp, Tcl_Command command); /* 160 */
+    int (*tcl_GetErrno) (void); /* 161 */
+    CONST84_RETURN char * (*tcl_GetHostName) (void); /* 162 */
+    int (*tcl_GetInterpPath) (Tcl_Interp *askInterp, Tcl_Interp *slaveInterp); /* 163 */
+    Tcl_Interp * (*tcl_GetMaster) (Tcl_Interp *interp); /* 164 */
+    const char * (*tcl_GetNameOfExecutable) (void); /* 165 */
+    Tcl_Obj * (*tcl_GetObjResult) (Tcl_Interp *interp); /* 166 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    int (*tcl_GetOpenFile) (Tcl_Interp *interp, const char *chanID, int forWriting, int checkUsage, ClientData *filePtr); /* 167 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    void (*reserved167)(void);
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    int (*tcl_GetOpenFile) (Tcl_Interp *interp, const char *chanID, int forWriting, int checkUsage, ClientData *filePtr); /* 167 */
+#endif /* MACOSX */
+    Tcl_PathType (*tcl_GetPathType) (const char *path); /* 168 */
+    int (*tcl_Gets) (Tcl_Channel chan, Tcl_DString *dsPtr); /* 169 */
+    int (*tcl_GetsObj) (Tcl_Channel chan, Tcl_Obj *objPtr); /* 170 */
+    int (*tcl_GetServiceMode) (void); /* 171 */
+    Tcl_Interp * (*tcl_GetSlave) (Tcl_Interp *interp, const char *slaveName); /* 172 */
+    Tcl_Channel (*tcl_GetStdChannel) (int type); /* 173 */
+    CONST84_RETURN char * (*tcl_GetStringResult) (Tcl_Interp *interp); /* 174 */
+    CONST84_RETURN char * (*tcl_GetVar) (Tcl_Interp *interp, const char *varName, int flags); /* 175 */
+    CONST84_RETURN char * (*tcl_GetVar2) (Tcl_Interp *interp, const char *part1, const char *part2, int flags); /* 176 */
+    int (*tcl_GlobalEval) (Tcl_Interp *interp, const char *command); /* 177 */
+    int (*tcl_GlobalEvalObj) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 178 */
+    int (*tcl_HideCommand) (Tcl_Interp *interp, const char *cmdName, const char *hiddenCmdToken); /* 179 */
+    int (*tcl_Init) (Tcl_Interp *interp); /* 180 */
+    void (*tcl_InitHashTable) (Tcl_HashTable *tablePtr, int keyType); /* 181 */
+    int (*tcl_InputBlocked) (Tcl_Channel chan); /* 182 */
+    int (*tcl_InputBuffered) (Tcl_Channel chan); /* 183 */
+    int (*tcl_InterpDeleted) (Tcl_Interp *interp); /* 184 */
+    int (*tcl_IsSafe) (Tcl_Interp *interp); /* 185 */
+    char * (*tcl_JoinPath) (int argc, CONST84 char *const *argv, Tcl_DString *resultPtr); /* 186 */
+    int (*tcl_LinkVar) (Tcl_Interp *interp, const char *varName, char *addr, int type); /* 187 */
+    void (*reserved188)(void);
+    Tcl_Channel (*tcl_MakeFileChannel) (ClientData handle, int mode); /* 189 */
+    int (*tcl_MakeSafe) (Tcl_Interp *interp); /* 190 */
+    Tcl_Channel (*tcl_MakeTcpClientChannel) (ClientData tcpSocket); /* 191 */
+    char * (*tcl_Merge) (int argc, CONST84 char *const *argv); /* 192 */
+    Tcl_HashEntry * (*tcl_NextHashEntry) (Tcl_HashSearch *searchPtr); /* 193 */
+    void (*tcl_NotifyChannel) (Tcl_Channel channel, int mask); /* 194 */
+    Tcl_Obj * (*tcl_ObjGetVar2) (Tcl_Interp *interp, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, int flags); /* 195 */
+    Tcl_Obj * (*tcl_ObjSetVar2) (Tcl_Interp *interp, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr, int flags); /* 196 */
+    Tcl_Channel (*tcl_OpenCommandChannel) (Tcl_Interp *interp, int argc, CONST84 char **argv, int flags); /* 197 */
+    Tcl_Channel (*tcl_OpenFileChannel) (Tcl_Interp *interp, const char *fileName, const char *modeString, int permissions); /* 198 */
+    Tcl_Channel (*tcl_OpenTcpClient) (Tcl_Interp *interp, int port, const char *address, const char *myaddr, int myport, int async); /* 199 */
+    Tcl_Channel (*tcl_OpenTcpServer) (Tcl_Interp *interp, int port, const char *host, Tcl_TcpAcceptProc *acceptProc, ClientData callbackData); /* 200 */
+    void (*tcl_Preserve) (ClientData data); /* 201 */
+    void (*tcl_PrintDouble) (Tcl_Interp *interp, double value, char *dst); /* 202 */
+    int (*tcl_PutEnv) (const char *assignment); /* 203 */
+    CONST84_RETURN char * (*tcl_PosixError) (Tcl_Interp *interp); /* 204 */
+    void (*tcl_QueueEvent) (Tcl_Event *evPtr, Tcl_QueuePosition position); /* 205 */
+    int (*tcl_Read) (Tcl_Channel chan, char *bufPtr, int toRead); /* 206 */
+    void (*tcl_ReapDetachedProcs) (void); /* 207 */
+    int (*tcl_RecordAndEval) (Tcl_Interp *interp, const char *cmd, int flags); /* 208 */
+    int (*tcl_RecordAndEvalObj) (Tcl_Interp *interp, Tcl_Obj *cmdPtr, int flags); /* 209 */
+    void (*tcl_RegisterChannel) (Tcl_Interp *interp, Tcl_Channel chan); /* 210 */
+    void (*tcl_RegisterObjType) (const Tcl_ObjType *typePtr); /* 211 */
+    Tcl_RegExp (*tcl_RegExpCompile) (Tcl_Interp *interp, const char *pattern); /* 212 */
+    int (*tcl_RegExpExec) (Tcl_Interp *interp, Tcl_RegExp regexp, const char *text, const char *start); /* 213 */
+    int (*tcl_RegExpMatch) (Tcl_Interp *interp, const char *text, const char *pattern); /* 214 */
+    void (*tcl_RegExpRange) (Tcl_RegExp regexp, int index, CONST84 char **startPtr, CONST84 char **endPtr); /* 215 */
+    void (*tcl_Release) (ClientData clientData); /* 216 */
+    void (*tcl_ResetResult) (Tcl_Interp *interp); /* 217 */
+    int (*tcl_ScanElement) (const char *src, int *flagPtr); /* 218 */
+    int (*tcl_ScanCountedElement) (const char *src, int length, int *flagPtr); /* 219 */
+    int (*tcl_SeekOld) (Tcl_Channel chan, int offset, int mode); /* 220 */
+    int (*tcl_ServiceAll) (void); /* 221 */
+    int (*tcl_ServiceEvent) (int flags); /* 222 */
+    void (*tcl_SetAssocData) (Tcl_Interp *interp, const char *name, Tcl_InterpDeleteProc *proc, ClientData clientData); /* 223 */
+    void (*tcl_SetChannelBufferSize) (Tcl_Channel chan, int sz); /* 224 */
+    int (*tcl_SetChannelOption) (Tcl_Interp *interp, Tcl_Channel chan, const char *optionName, const char *newValue); /* 225 */
+    int (*tcl_SetCommandInfo) (Tcl_Interp *interp, const char *cmdName, const Tcl_CmdInfo *infoPtr); /* 226 */
+    void (*tcl_SetErrno) (int err); /* 227 */
+    void (*tcl_SetErrorCode) (Tcl_Interp *interp, ...); /* 228 */
+    void (*tcl_SetMaxBlockTime) (const Tcl_Time *timePtr); /* 229 */
+    void (*tcl_SetPanicProc) (TCL_NORETURN1 Tcl_PanicProc *panicProc); /* 230 */
+    int (*tcl_SetRecursionLimit) (Tcl_Interp *interp, int depth); /* 231 */
+    void (*tcl_SetResult) (Tcl_Interp *interp, char *result, Tcl_FreeProc *freeProc); /* 232 */
+    int (*tcl_SetServiceMode) (int mode); /* 233 */
+    void (*tcl_SetObjErrorCode) (Tcl_Interp *interp, Tcl_Obj *errorObjPtr); /* 234 */
+    void (*tcl_SetObjResult) (Tcl_Interp *interp, Tcl_Obj *resultObjPtr); /* 235 */
+    void (*tcl_SetStdChannel) (Tcl_Channel channel, int type); /* 236 */
+    CONST84_RETURN char * (*tcl_SetVar) (Tcl_Interp *interp, const char *varName, const char *newValue, int flags); /* 237 */
+    CONST84_RETURN char * (*tcl_SetVar2) (Tcl_Interp *interp, const char *part1, const char *part2, const char *newValue, int flags); /* 238 */
+    CONST84_RETURN char * (*tcl_SignalId) (int sig); /* 239 */
+    CONST84_RETURN char * (*tcl_SignalMsg) (int sig); /* 240 */
+    void (*tcl_SourceRCFile) (Tcl_Interp *interp); /* 241 */
+    int (*tcl_SplitList) (Tcl_Interp *interp, const char *listStr, int *argcPtr, CONST84 char ***argvPtr); /* 242 */
+    void (*tcl_SplitPath) (const char *path, int *argcPtr, CONST84 char ***argvPtr); /* 243 */
+    void (*tcl_StaticPackage) (Tcl_Interp *interp, const char *pkgName, Tcl_PackageInitProc *initProc, Tcl_PackageInitProc *safeInitProc); /* 244 */
+    int (*tcl_StringMatch) (const char *str, const char *pattern); /* 245 */
+    int (*tcl_TellOld) (Tcl_Channel chan); /* 246 */
+    int (*tcl_TraceVar) (Tcl_Interp *interp, const char *varName, int flags, Tcl_VarTraceProc *proc, ClientData clientData); /* 247 */
+    int (*tcl_TraceVar2) (Tcl_Interp *interp, const char *part1, const char *part2, int flags, Tcl_VarTraceProc *proc, ClientData clientData); /* 248 */
+    char * (*tcl_TranslateFileName) (Tcl_Interp *interp, const char *name, Tcl_DString *bufferPtr); /* 249 */
+    int (*tcl_Ungets) (Tcl_Channel chan, const char *str, int len, int atHead); /* 250 */
+    void (*tcl_UnlinkVar) (Tcl_Interp *interp, const char *varName); /* 251 */
+    int (*tcl_UnregisterChannel) (Tcl_Interp *interp, Tcl_Channel chan); /* 252 */
+    int (*tcl_UnsetVar) (Tcl_Interp *interp, const char *varName, int flags); /* 253 */
+    int (*tcl_UnsetVar2) (Tcl_Interp *interp, const char *part1, const char *part2, int flags); /* 254 */
+    void (*tcl_UntraceVar) (Tcl_Interp *interp, const char *varName, int flags, Tcl_VarTraceProc *proc, ClientData clientData); /* 255 */
+    void (*tcl_UntraceVar2) (Tcl_Interp *interp, const char *part1, const char *part2, int flags, Tcl_VarTraceProc *proc, ClientData clientData); /* 256 */
+    void (*tcl_UpdateLinkedVar) (Tcl_Interp *interp, const char *varName); /* 257 */
+    int (*tcl_UpVar) (Tcl_Interp *interp, const char *frameName, const char *varName, const char *localName, int flags); /* 258 */
+    int (*tcl_UpVar2) (Tcl_Interp *interp, const char *frameName, const char *part1, const char *part2, const char *localName, int flags); /* 259 */
+    int (*tcl_VarEval) (Tcl_Interp *interp, ...); /* 260 */
+    ClientData (*tcl_VarTraceInfo) (Tcl_Interp *interp, const char *varName, int flags, Tcl_VarTraceProc *procPtr, ClientData prevClientData); /* 261 */
+    ClientData (*tcl_VarTraceInfo2) (Tcl_Interp *interp, const char *part1, const char *part2, int flags, Tcl_VarTraceProc *procPtr, ClientData prevClientData); /* 262 */
+    int (*tcl_Write) (Tcl_Channel chan, const char *s, int slen); /* 263 */
+    void (*tcl_WrongNumArgs) (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[], const char *message); /* 264 */
+    int (*tcl_DumpActiveMemory) (const char *fileName); /* 265 */
+    void (*tcl_ValidateAllMemory) (const char *file, int line); /* 266 */
+    void (*tcl_AppendResultVA) (Tcl_Interp *interp, va_list argList); /* 267 */
+    void (*tcl_AppendStringsToObjVA) (Tcl_Obj *objPtr, va_list argList); /* 268 */
+    char * (*tcl_HashStats) (Tcl_HashTable *tablePtr); /* 269 */
+    CONST84_RETURN char * (*tcl_ParseVar) (Tcl_Interp *interp, const char *start, CONST84 char **termPtr); /* 270 */
+    CONST84_RETURN char * (*tcl_PkgPresent) (Tcl_Interp *interp, const char *name, const char *version, int exact); /* 271 */
+    CONST84_RETURN char * (*tcl_PkgPresentEx) (Tcl_Interp *interp, const char *name, const char *version, int exact, void *clientDataPtr); /* 272 */
+    int (*tcl_PkgProvide) (Tcl_Interp *interp, const char *name, const char *version); /* 273 */
+    CONST84_RETURN char * (*tcl_PkgRequire) (Tcl_Interp *interp, const char *name, const char *version, int exact); /* 274 */
+    void (*tcl_SetErrorCodeVA) (Tcl_Interp *interp, va_list argList); /* 275 */
+    int (*tcl_VarEvalVA) (Tcl_Interp *interp, va_list argList); /* 276 */
+    Tcl_Pid (*tcl_WaitPid) (Tcl_Pid pid, int *statPtr, int options); /* 277 */
+    TCL_NORETURN1 void (*tcl_PanicVA) (const char *format, va_list argList); /* 278 */
+    void (*tcl_GetVersion) (int *major, int *minor, int *patchLevel, int *type); /* 279 */
+    void (*tcl_InitMemory) (Tcl_Interp *interp); /* 280 */
+    Tcl_Channel (*tcl_StackChannel) (Tcl_Interp *interp, const Tcl_ChannelType *typePtr, ClientData instanceData, int mask, Tcl_Channel prevChan); /* 281 */
+    int (*tcl_UnstackChannel) (Tcl_Interp *interp, Tcl_Channel chan); /* 282 */
+    Tcl_Channel (*tcl_GetStackedChannel) (Tcl_Channel chan); /* 283 */
+    void (*tcl_SetMainLoop) (Tcl_MainLoopProc *proc); /* 284 */
+    void (*reserved285)(void);
+    void (*tcl_AppendObjToObj) (Tcl_Obj *objPtr, Tcl_Obj *appendObjPtr); /* 286 */
+    Tcl_Encoding (*tcl_CreateEncoding) (const Tcl_EncodingType *typePtr); /* 287 */
+    void (*tcl_CreateThreadExitHandler) (Tcl_ExitProc *proc, ClientData clientData); /* 288 */
+    void (*tcl_DeleteThreadExitHandler) (Tcl_ExitProc *proc, ClientData clientData); /* 289 */
+    void (*tcl_DiscardResult) (Tcl_SavedResult *statePtr); /* 290 */
+    int (*tcl_EvalEx) (Tcl_Interp *interp, const char *script, int numBytes, int flags); /* 291 */
+    int (*tcl_EvalObjv) (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[], int flags); /* 292 */
+    int (*tcl_EvalObjEx) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags); /* 293 */
+    TCL_NORETURN1 void (*tcl_ExitThread) (int status); /* 294 */
+    int (*tcl_ExternalToUtf) (Tcl_Interp *interp, Tcl_Encoding encoding, const char *src, int srcLen, int flags, Tcl_EncodingState *statePtr, char *dst, int dstLen, int *srcReadPtr, int *dstWrotePtr, int *dstCharsPtr); /* 295 */
+    char * (*tcl_ExternalToUtfDString) (Tcl_Encoding encoding, const char *src, int srcLen, Tcl_DString *dsPtr); /* 296 */
+    void (*tcl_FinalizeThread) (void); /* 297 */
+    void (*tcl_FinalizeNotifier) (ClientData clientData); /* 298 */
+    void (*tcl_FreeEncoding) (Tcl_Encoding encoding); /* 299 */
+    Tcl_ThreadId (*tcl_GetCurrentThread) (void); /* 300 */
+    Tcl_Encoding (*tcl_GetEncoding) (Tcl_Interp *interp, const char *name); /* 301 */
+    CONST84_RETURN char * (*tcl_GetEncodingName) (Tcl_Encoding encoding); /* 302 */
+    void (*tcl_GetEncodingNames) (Tcl_Interp *interp); /* 303 */
+    int (*tcl_GetIndexFromObjStruct) (Tcl_Interp *interp, Tcl_Obj *objPtr, const void *tablePtr, int offset, const char *msg, int flags, int *indexPtr); /* 304 */
+    void * (*tcl_GetThreadData) (Tcl_ThreadDataKey *keyPtr, int size); /* 305 */
+    Tcl_Obj * (*tcl_GetVar2Ex) (Tcl_Interp *interp, const char *part1, const char *part2, int flags); /* 306 */
+    ClientData (*tcl_InitNotifier) (void); /* 307 */
+    void (*tcl_MutexLock) (Tcl_Mutex *mutexPtr); /* 308 */
+    void (*tcl_MutexUnlock) (Tcl_Mutex *mutexPtr); /* 309 */
+    void (*tcl_ConditionNotify) (Tcl_Condition *condPtr); /* 310 */
+    void (*tcl_ConditionWait) (Tcl_Condition *condPtr, Tcl_Mutex *mutexPtr, const Tcl_Time *timePtr); /* 311 */
+    int (*tcl_NumUtfChars) (const char *src, int length); /* 312 */
+    int (*tcl_ReadChars) (Tcl_Channel channel, Tcl_Obj *objPtr, int charsToRead, int appendFlag); /* 313 */
+    void (*tcl_RestoreResult) (Tcl_Interp *interp, Tcl_SavedResult *statePtr); /* 314 */
+    void (*tcl_SaveResult) (Tcl_Interp *interp, Tcl_SavedResult *statePtr); /* 315 */
+    int (*tcl_SetSystemEncoding) (Tcl_Interp *interp, const char *name); /* 316 */
+    Tcl_Obj * (*tcl_SetVar2Ex) (Tcl_Interp *interp, const char *part1, const char *part2, Tcl_Obj *newValuePtr, int flags); /* 317 */
+    void (*tcl_ThreadAlert) (Tcl_ThreadId threadId); /* 318 */
+    void (*tcl_ThreadQueueEvent) (Tcl_ThreadId threadId, Tcl_Event *evPtr, Tcl_QueuePosition position); /* 319 */
+    Tcl_UniChar (*tcl_UniCharAtIndex) (const char *src, int index); /* 320 */
+    Tcl_UniChar (*tcl_UniCharToLower) (int ch); /* 321 */
+    Tcl_UniChar (*tcl_UniCharToTitle) (int ch); /* 322 */
+    Tcl_UniChar (*tcl_UniCharToUpper) (int ch); /* 323 */
+    int (*tcl_UniCharToUtf) (int ch, char *buf); /* 324 */
+    CONST84_RETURN char * (*tcl_UtfAtIndex) (const char *src, int index); /* 325 */
+    int (*tcl_UtfCharComplete) (const char *src, int length); /* 326 */
+    int (*tcl_UtfBackslash) (const char *src, int *readPtr, char *dst); /* 327 */
+    CONST84_RETURN char * (*tcl_UtfFindFirst) (const char *src, int ch); /* 328 */
+    CONST84_RETURN char * (*tcl_UtfFindLast) (const char *src, int ch); /* 329 */
+    CONST84_RETURN char * (*tcl_UtfNext) (const char *src); /* 330 */
+    CONST84_RETURN char * (*tcl_UtfPrev) (const char *src, const char *start); /* 331 */
+    int (*tcl_UtfToExternal) (Tcl_Interp *interp, Tcl_Encoding encoding, const char *src, int srcLen, int flags, Tcl_EncodingState *statePtr, char *dst, int dstLen, int *srcReadPtr, int *dstWrotePtr, int *dstCharsPtr); /* 332 */
+    char * (*tcl_UtfToExternalDString) (Tcl_Encoding encoding, const char *src, int srcLen, Tcl_DString *dsPtr); /* 333 */
+    int (*tcl_UtfToLower) (char *src); /* 334 */
+    int (*tcl_UtfToTitle) (char *src); /* 335 */
+    int (*tcl_UtfToUniChar) (const char *src, Tcl_UniChar *chPtr); /* 336 */
+    int (*tcl_UtfToUpper) (char *src); /* 337 */
+    int (*tcl_WriteChars) (Tcl_Channel chan, const char *src, int srcLen); /* 338 */
+    int (*tcl_WriteObj) (Tcl_Channel chan, Tcl_Obj *objPtr); /* 339 */
+    char * (*tcl_GetString) (Tcl_Obj *objPtr); /* 340 */
+    CONST84_RETURN char * (*tcl_GetDefaultEncodingDir) (void); /* 341 */
+    void (*tcl_SetDefaultEncodingDir) (const char *path); /* 342 */
+    void (*tcl_AlertNotifier) (ClientData clientData); /* 343 */
+    void (*tcl_ServiceModeHook) (int mode); /* 344 */
+    int (*tcl_UniCharIsAlnum) (int ch); /* 345 */
+    int (*tcl_UniCharIsAlpha) (int ch); /* 346 */
+    int (*tcl_UniCharIsDigit) (int ch); /* 347 */
+    int (*tcl_UniCharIsLower) (int ch); /* 348 */
+    int (*tcl_UniCharIsSpace) (int ch); /* 349 */
+    int (*tcl_UniCharIsUpper) (int ch); /* 350 */
+    int (*tcl_UniCharIsWordChar) (int ch); /* 351 */
+    int (*tcl_UniCharLen) (const Tcl_UniChar *uniStr); /* 352 */
+    int (*tcl_UniCharNcmp) (const Tcl_UniChar *ucs, const Tcl_UniChar *uct, unsigned long numChars); /* 353 */
+    char * (*tcl_UniCharToUtfDString) (const Tcl_UniChar *uniStr, int uniLength, Tcl_DString *dsPtr); /* 354 */
+    Tcl_UniChar * (*tcl_UtfToUniCharDString) (const char *src, int length, Tcl_DString *dsPtr); /* 355 */
+    Tcl_RegExp (*tcl_GetRegExpFromObj) (Tcl_Interp *interp, Tcl_Obj *patObj, int flags); /* 356 */
+    Tcl_Obj * (*tcl_EvalTokens) (Tcl_Interp *interp, Tcl_Token *tokenPtr, int count); /* 357 */
+    void (*tcl_FreeParse) (Tcl_Parse *parsePtr); /* 358 */
+    void (*tcl_LogCommandInfo) (Tcl_Interp *interp, const char *script, const char *command, int length); /* 359 */
+    int (*tcl_ParseBraces) (Tcl_Interp *interp, const char *start, int numBytes, Tcl_Parse *parsePtr, int append, CONST84 char **termPtr); /* 360 */
+    int (*tcl_ParseCommand) (Tcl_Interp *interp, const char *start, int numBytes, int nested, Tcl_Parse *parsePtr); /* 361 */
+    int (*tcl_ParseExpr) (Tcl_Interp *interp, const char *start, int numBytes, Tcl_Parse *parsePtr); /* 362 */
+    int (*tcl_ParseQuotedString) (Tcl_Interp *interp, const char *start, int numBytes, Tcl_Parse *parsePtr, int append, CONST84 char **termPtr); /* 363 */
+    int (*tcl_ParseVarName) (Tcl_Interp *interp, const char *start, int numBytes, Tcl_Parse *parsePtr, int append); /* 364 */
+    char * (*tcl_GetCwd) (Tcl_Interp *interp, Tcl_DString *cwdPtr); /* 365 */
+    int (*tcl_Chdir) (const char *dirName); /* 366 */
+    int (*tcl_Access) (const char *path, int mode); /* 367 */
+    int (*tcl_Stat) (const char *path, struct stat *bufPtr); /* 368 */
+    int (*tcl_UtfNcmp) (const char *s1, const char *s2, unsigned long n); /* 369 */
+    int (*tcl_UtfNcasecmp) (const char *s1, const char *s2, unsigned long n); /* 370 */
+    int (*tcl_StringCaseMatch) (const char *str, const char *pattern, int nocase); /* 371 */
+    int (*tcl_UniCharIsControl) (int ch); /* 372 */
+    int (*tcl_UniCharIsGraph) (int ch); /* 373 */
+    int (*tcl_UniCharIsPrint) (int ch); /* 374 */
+    int (*tcl_UniCharIsPunct) (int ch); /* 375 */
+    int (*tcl_RegExpExecObj) (Tcl_Interp *interp, Tcl_RegExp regexp, Tcl_Obj *textObj, int offset, int nmatches, int flags); /* 376 */
+    void (*tcl_RegExpGetInfo) (Tcl_RegExp regexp, Tcl_RegExpInfo *infoPtr); /* 377 */
+    Tcl_Obj * (*tcl_NewUnicodeObj) (const Tcl_UniChar *unicode, int numChars); /* 378 */
+    void (*tcl_SetUnicodeObj) (Tcl_Obj *objPtr, const Tcl_UniChar *unicode, int numChars); /* 379 */
+    int (*tcl_GetCharLength) (Tcl_Obj *objPtr); /* 380 */
+    Tcl_UniChar (*tcl_GetUniChar) (Tcl_Obj *objPtr, int index); /* 381 */
+    Tcl_UniChar * (*tcl_GetUnicode) (Tcl_Obj *objPtr); /* 382 */
+    Tcl_Obj * (*tcl_GetRange) (Tcl_Obj *objPtr, int first, int last); /* 383 */
+    void (*tcl_AppendUnicodeToObj) (Tcl_Obj *objPtr, const Tcl_UniChar *unicode, int length); /* 384 */
+    int (*tcl_RegExpMatchObj) (Tcl_Interp *interp, Tcl_Obj *textObj, Tcl_Obj *patternObj); /* 385 */
+    void (*tcl_SetNotifier) (Tcl_NotifierProcs *notifierProcPtr); /* 386 */
+    Tcl_Mutex * (*tcl_GetAllocMutex) (void); /* 387 */
+    int (*tcl_GetChannelNames) (Tcl_Interp *interp); /* 388 */
+    int (*tcl_GetChannelNamesEx) (Tcl_Interp *interp, const char *pattern); /* 389 */
+    int (*tcl_ProcObjCmd) (ClientData clientData, Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]); /* 390 */
+    void (*tcl_ConditionFinalize) (Tcl_Condition *condPtr); /* 391 */
+    void (*tcl_MutexFinalize) (Tcl_Mutex *mutex); /* 392 */
+    int (*tcl_CreateThread) (Tcl_ThreadId *idPtr, Tcl_ThreadCreateProc *proc, ClientData clientData, int stackSize, int flags); /* 393 */
+    int (*tcl_ReadRaw) (Tcl_Channel chan, char *dst, int bytesToRead); /* 394 */
+    int (*tcl_WriteRaw) (Tcl_Channel chan, const char *src, int srcLen); /* 395 */
+    Tcl_Channel (*tcl_GetTopChannel) (Tcl_Channel chan); /* 396 */
+    int (*tcl_ChannelBuffered) (Tcl_Channel chan); /* 397 */
+    CONST84_RETURN char * (*tcl_ChannelName) (const Tcl_ChannelType *chanTypePtr); /* 398 */
+    Tcl_ChannelTypeVersion (*tcl_ChannelVersion) (const Tcl_ChannelType *chanTypePtr); /* 399 */
+    Tcl_DriverBlockModeProc * (*tcl_ChannelBlockModeProc) (const Tcl_ChannelType *chanTypePtr); /* 400 */
+    Tcl_DriverCloseProc * (*tcl_ChannelCloseProc) (const Tcl_ChannelType *chanTypePtr); /* 401 */
+    Tcl_DriverClose2Proc * (*tcl_ChannelClose2Proc) (const Tcl_ChannelType *chanTypePtr); /* 402 */
+    Tcl_DriverInputProc * (*tcl_ChannelInputProc) (const Tcl_ChannelType *chanTypePtr); /* 403 */
+    Tcl_DriverOutputProc * (*tcl_ChannelOutputProc) (const Tcl_ChannelType *chanTypePtr); /* 404 */
+    Tcl_DriverSeekProc * (*tcl_ChannelSeekProc) (const Tcl_ChannelType *chanTypePtr); /* 405 */
+    Tcl_DriverSetOptionProc * (*tcl_ChannelSetOptionProc) (const Tcl_ChannelType *chanTypePtr); /* 406 */
+    Tcl_DriverGetOptionProc * (*tcl_ChannelGetOptionProc) (const Tcl_ChannelType *chanTypePtr); /* 407 */
+    Tcl_DriverWatchProc * (*tcl_ChannelWatchProc) (const Tcl_ChannelType *chanTypePtr); /* 408 */
+    Tcl_DriverGetHandleProc * (*tcl_ChannelGetHandleProc) (const Tcl_ChannelType *chanTypePtr); /* 409 */
+    Tcl_DriverFlushProc * (*tcl_ChannelFlushProc) (const Tcl_ChannelType *chanTypePtr); /* 410 */
+    Tcl_DriverHandlerProc * (*tcl_ChannelHandlerProc) (const Tcl_ChannelType *chanTypePtr); /* 411 */
+    int (*tcl_JoinThread) (Tcl_ThreadId threadId, int *result); /* 412 */
+    int (*tcl_IsChannelShared) (Tcl_Channel channel); /* 413 */
+    int (*tcl_IsChannelRegistered) (Tcl_Interp *interp, Tcl_Channel channel); /* 414 */
+    void (*tcl_CutChannel) (Tcl_Channel channel); /* 415 */
+    void (*tcl_SpliceChannel) (Tcl_Channel channel); /* 416 */
+    void (*tcl_ClearChannelHandlers) (Tcl_Channel channel); /* 417 */
+    int (*tcl_IsChannelExisting) (const char *channelName); /* 418 */
+    int (*tcl_UniCharNcasecmp) (const Tcl_UniChar *ucs, const Tcl_UniChar *uct, unsigned long numChars); /* 419 */
+    int (*tcl_UniCharCaseMatch) (const Tcl_UniChar *uniStr, const Tcl_UniChar *uniPattern, int nocase); /* 420 */
+    Tcl_HashEntry * (*tcl_FindHashEntry) (Tcl_HashTable *tablePtr, const void *key); /* 421 */
+    Tcl_HashEntry * (*tcl_CreateHashEntry) (Tcl_HashTable *tablePtr, const void *key, int *newPtr); /* 422 */
+    void (*tcl_InitCustomHashTable) (Tcl_HashTable *tablePtr, int keyType, const Tcl_HashKeyType *typePtr); /* 423 */
+    void (*tcl_InitObjHashTable) (Tcl_HashTable *tablePtr); /* 424 */
+    ClientData (*tcl_CommandTraceInfo) (Tcl_Interp *interp, const char *varName, int flags, Tcl_CommandTraceProc *procPtr, ClientData prevClientData); /* 425 */
+    int (*tcl_TraceCommand) (Tcl_Interp *interp, const char *varName, int flags, Tcl_CommandTraceProc *proc, ClientData clientData); /* 426 */
+    void (*tcl_UntraceCommand) (Tcl_Interp *interp, const char *varName, int flags, Tcl_CommandTraceProc *proc, ClientData clientData); /* 427 */
+    char * (*tcl_AttemptAlloc) (unsigned int size); /* 428 */
+    char * (*tcl_AttemptDbCkalloc) (unsigned int size, const char *file, int line); /* 429 */
+    char * (*tcl_AttemptRealloc) (char *ptr, unsigned int size); /* 430 */
+    char * (*tcl_AttemptDbCkrealloc) (char *ptr, unsigned int size, const char *file, int line); /* 431 */
+    int (*tcl_AttemptSetObjLength) (Tcl_Obj *objPtr, int length); /* 432 */
+    Tcl_ThreadId (*tcl_GetChannelThread) (Tcl_Channel channel); /* 433 */
+    Tcl_UniChar * (*tcl_GetUnicodeFromObj) (Tcl_Obj *objPtr, int *lengthPtr); /* 434 */
+    int (*tcl_GetMathFuncInfo) (Tcl_Interp *interp, const char *name, int *numArgsPtr, Tcl_ValueType **argTypesPtr, Tcl_MathProc **procPtr, ClientData *clientDataPtr); /* 435 */
+    Tcl_Obj * (*tcl_ListMathFuncs) (Tcl_Interp *interp, const char *pattern); /* 436 */
+    Tcl_Obj * (*tcl_SubstObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags); /* 437 */
+    int (*tcl_DetachChannel) (Tcl_Interp *interp, Tcl_Channel channel); /* 438 */
+    int (*tcl_IsStandardChannel) (Tcl_Channel channel); /* 439 */
+    int (*tcl_FSCopyFile) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr); /* 440 */
+    int (*tcl_FSCopyDirectory) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr); /* 441 */
+    int (*tcl_FSCreateDirectory) (Tcl_Obj *pathPtr); /* 442 */
+    int (*tcl_FSDeleteFile) (Tcl_Obj *pathPtr); /* 443 */
+    int (*tcl_FSLoadFile) (Tcl_Interp *interp, Tcl_Obj *pathPtr, const char *sym1, const char *sym2, Tcl_PackageInitProc **proc1Ptr, Tcl_PackageInitProc **proc2Ptr, Tcl_LoadHandle *handlePtr, Tcl_FSUnloadFileProc **unloadProcPtr); /* 444 */
+    int (*tcl_FSMatchInDirectory) (Tcl_Interp *interp, Tcl_Obj *result, Tcl_Obj *pathPtr, const char *pattern, Tcl_GlobTypeData *types); /* 445 */
+    Tcl_Obj * (*tcl_FSLink) (Tcl_Obj *pathPtr, Tcl_Obj *toPtr, int linkAction); /* 446 */
+    int (*tcl_FSRemoveDirectory) (Tcl_Obj *pathPtr, int recursive, Tcl_Obj **errorPtr); /* 447 */
+    int (*tcl_FSRenameFile) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr); /* 448 */
+    int (*tcl_FSLstat) (Tcl_Obj *pathPtr, Tcl_StatBuf *buf); /* 449 */
+    int (*tcl_FSUtime) (Tcl_Obj *pathPtr, struct utimbuf *tval); /* 450 */
+    int (*tcl_FSFileAttrsGet) (Tcl_Interp *interp, int index, Tcl_Obj *pathPtr, Tcl_Obj **objPtrRef); /* 451 */
+    int (*tcl_FSFileAttrsSet) (Tcl_Interp *interp, int index, Tcl_Obj *pathPtr, Tcl_Obj *objPtr); /* 452 */
+    const char *CONST86 * (*tcl_FSFileAttrStrings) (Tcl_Obj *pathPtr, Tcl_Obj **objPtrRef); /* 453 */
+    int (*tcl_FSStat) (Tcl_Obj *pathPtr, Tcl_StatBuf *buf); /* 454 */
+    int (*tcl_FSAccess) (Tcl_Obj *pathPtr, int mode); /* 455 */
+    Tcl_Channel (*tcl_FSOpenFileChannel) (Tcl_Interp *interp, Tcl_Obj *pathPtr, const char *modeString, int permissions); /* 456 */
+    Tcl_Obj * (*tcl_FSGetCwd) (Tcl_Interp *interp); /* 457 */
+    int (*tcl_FSChdir) (Tcl_Obj *pathPtr); /* 458 */
+    int (*tcl_FSConvertToPathType) (Tcl_Interp *interp, Tcl_Obj *pathPtr); /* 459 */
+    Tcl_Obj * (*tcl_FSJoinPath) (Tcl_Obj *listObj, int elements); /* 460 */
+    Tcl_Obj * (*tcl_FSSplitPath) (Tcl_Obj *pathPtr, int *lenPtr); /* 461 */
+    int (*tcl_FSEqualPaths) (Tcl_Obj *firstPtr, Tcl_Obj *secondPtr); /* 462 */
+    Tcl_Obj * (*tcl_FSGetNormalizedPath) (Tcl_Interp *interp, Tcl_Obj *pathPtr); /* 463 */
+    Tcl_Obj * (*tcl_FSJoinToPath) (Tcl_Obj *pathPtr, int objc, Tcl_Obj *const objv[]); /* 464 */
+    ClientData (*tcl_FSGetInternalRep) (Tcl_Obj *pathPtr, const Tcl_Filesystem *fsPtr); /* 465 */
+    Tcl_Obj * (*tcl_FSGetTranslatedPath) (Tcl_Interp *interp, Tcl_Obj *pathPtr); /* 466 */
+    int (*tcl_FSEvalFile) (Tcl_Interp *interp, Tcl_Obj *fileName); /* 467 */
+    Tcl_Obj * (*tcl_FSNewNativePath) (const Tcl_Filesystem *fromFilesystem, ClientData clientData); /* 468 */
+    const void * (*tcl_FSGetNativePath) (Tcl_Obj *pathPtr); /* 469 */
+    Tcl_Obj * (*tcl_FSFileSystemInfo) (Tcl_Obj *pathPtr); /* 470 */
+    Tcl_Obj * (*tcl_FSPathSeparator) (Tcl_Obj *pathPtr); /* 471 */
+    Tcl_Obj * (*tcl_FSListVolumes) (void); /* 472 */
+    int (*tcl_FSRegister) (ClientData clientData, const Tcl_Filesystem *fsPtr); /* 473 */
+    int (*tcl_FSUnregister) (const Tcl_Filesystem *fsPtr); /* 474 */
+    ClientData (*tcl_FSData) (const Tcl_Filesystem *fsPtr); /* 475 */
+    const char * (*tcl_FSGetTranslatedStringPath) (Tcl_Interp *interp, Tcl_Obj *pathPtr); /* 476 */
+    CONST86 Tcl_Filesystem * (*tcl_FSGetFileSystemForPath) (Tcl_Obj *pathPtr); /* 477 */
+    Tcl_PathType (*tcl_FSGetPathType) (Tcl_Obj *pathPtr); /* 478 */
+    int (*tcl_OutputBuffered) (Tcl_Channel chan); /* 479 */
+    void (*tcl_FSMountsChanged) (const Tcl_Filesystem *fsPtr); /* 480 */
+    int (*tcl_EvalTokensStandard) (Tcl_Interp *interp, Tcl_Token *tokenPtr, int count); /* 481 */
+    void (*tcl_GetTime) (Tcl_Time *timeBuf); /* 482 */
+    Tcl_Trace (*tcl_CreateObjTrace) (Tcl_Interp *interp, int level, int flags, Tcl_CmdObjTraceProc *objProc, ClientData clientData, Tcl_CmdObjTraceDeleteProc *delProc); /* 483 */
+    int (*tcl_GetCommandInfoFromToken) (Tcl_Command token, Tcl_CmdInfo *infoPtr); /* 484 */
+    int (*tcl_SetCommandInfoFromToken) (Tcl_Command token, const Tcl_CmdInfo *infoPtr); /* 485 */
+    Tcl_Obj * (*tcl_DbNewWideIntObj) (Tcl_WideInt wideValue, const char *file, int line); /* 486 */
+    int (*tcl_GetWideIntFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_WideInt *widePtr); /* 487 */
+    Tcl_Obj * (*tcl_NewWideIntObj) (Tcl_WideInt wideValue); /* 488 */
+    void (*tcl_SetWideIntObj) (Tcl_Obj *objPtr, Tcl_WideInt wideValue); /* 489 */
+    Tcl_StatBuf * (*tcl_AllocStatBuf) (void); /* 490 */
+    Tcl_WideInt (*tcl_Seek) (Tcl_Channel chan, Tcl_WideInt offset, int mode); /* 491 */
+    Tcl_WideInt (*tcl_Tell) (Tcl_Channel chan); /* 492 */
+    Tcl_DriverWideSeekProc * (*tcl_ChannelWideSeekProc) (const Tcl_ChannelType *chanTypePtr); /* 493 */
+    int (*tcl_DictObjPut) (Tcl_Interp *interp, Tcl_Obj *dictPtr, Tcl_Obj *keyPtr, Tcl_Obj *valuePtr); /* 494 */
+    int (*tcl_DictObjGet) (Tcl_Interp *interp, Tcl_Obj *dictPtr, Tcl_Obj *keyPtr, Tcl_Obj **valuePtrPtr); /* 495 */
+    int (*tcl_DictObjRemove) (Tcl_Interp *interp, Tcl_Obj *dictPtr, Tcl_Obj *keyPtr); /* 496 */
+    int (*tcl_DictObjSize) (Tcl_Interp *interp, Tcl_Obj *dictPtr, int *sizePtr); /* 497 */
+    int (*tcl_DictObjFirst) (Tcl_Interp *interp, Tcl_Obj *dictPtr, Tcl_DictSearch *searchPtr, Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr, int *donePtr); /* 498 */
+    void (*tcl_DictObjNext) (Tcl_DictSearch *searchPtr, Tcl_Obj **keyPtrPtr, Tcl_Obj **valuePtrPtr, int *donePtr); /* 499 */
+    void (*tcl_DictObjDone) (Tcl_DictSearch *searchPtr); /* 500 */
+    int (*tcl_DictObjPutKeyList) (Tcl_Interp *interp, Tcl_Obj *dictPtr, int keyc, Tcl_Obj *const *keyv, Tcl_Obj *valuePtr); /* 501 */
+    int (*tcl_DictObjRemoveKeyList) (Tcl_Interp *interp, Tcl_Obj *dictPtr, int keyc, Tcl_Obj *const *keyv); /* 502 */
+    Tcl_Obj * (*tcl_NewDictObj) (void); /* 503 */
+    Tcl_Obj * (*tcl_DbNewDictObj) (const char *file, int line); /* 504 */
+    void (*tcl_RegisterConfig) (Tcl_Interp *interp, const char *pkgName, const Tcl_Config *configuration, const char *valEncoding); /* 505 */
+    Tcl_Namespace * (*tcl_CreateNamespace) (Tcl_Interp *interp, const char *name, ClientData clientData, Tcl_NamespaceDeleteProc *deleteProc); /* 506 */
+    void (*tcl_DeleteNamespace) (Tcl_Namespace *nsPtr); /* 507 */
+    int (*tcl_AppendExportList) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, Tcl_Obj *objPtr); /* 508 */
+    int (*tcl_Export) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern, int resetListFirst); /* 509 */
+    int (*tcl_Import) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern, int allowOverwrite); /* 510 */
+    int (*tcl_ForgetImport) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern); /* 511 */
+    Tcl_Namespace * (*tcl_GetCurrentNamespace) (Tcl_Interp *interp); /* 512 */
+    Tcl_Namespace * (*tcl_GetGlobalNamespace) (Tcl_Interp *interp); /* 513 */
+    Tcl_Namespace * (*tcl_FindNamespace) (Tcl_Interp *interp, const char *name, Tcl_Namespace *contextNsPtr, int flags); /* 514 */
+    Tcl_Command (*tcl_FindCommand) (Tcl_Interp *interp, const char *name, Tcl_Namespace *contextNsPtr, int flags); /* 515 */
+    Tcl_Command (*tcl_GetCommandFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 516 */
+    void (*tcl_GetCommandFullName) (Tcl_Interp *interp, Tcl_Command command, Tcl_Obj *objPtr); /* 517 */
+    int (*tcl_FSEvalFileEx) (Tcl_Interp *interp, Tcl_Obj *fileName, const char *encodingName); /* 518 */
+    Tcl_ExitProc * (*tcl_SetExitProc) (TCL_NORETURN1 Tcl_ExitProc *proc); /* 519 */
+    void (*tcl_LimitAddHandler) (Tcl_Interp *interp, int type, Tcl_LimitHandlerProc *handlerProc, ClientData clientData, Tcl_LimitHandlerDeleteProc *deleteProc); /* 520 */
+    void (*tcl_LimitRemoveHandler) (Tcl_Interp *interp, int type, Tcl_LimitHandlerProc *handlerProc, ClientData clientData); /* 521 */
+    int (*tcl_LimitReady) (Tcl_Interp *interp); /* 522 */
+    int (*tcl_LimitCheck) (Tcl_Interp *interp); /* 523 */
+    int (*tcl_LimitExceeded) (Tcl_Interp *interp); /* 524 */
+    void (*tcl_LimitSetCommands) (Tcl_Interp *interp, int commandLimit); /* 525 */
+    void (*tcl_LimitSetTime) (Tcl_Interp *interp, Tcl_Time *timeLimitPtr); /* 526 */
+    void (*tcl_LimitSetGranularity) (Tcl_Interp *interp, int type, int granularity); /* 527 */
+    int (*tcl_LimitTypeEnabled) (Tcl_Interp *interp, int type); /* 528 */
+    int (*tcl_LimitTypeExceeded) (Tcl_Interp *interp, int type); /* 529 */
+    void (*tcl_LimitTypeSet) (Tcl_Interp *interp, int type); /* 530 */
+    void (*tcl_LimitTypeReset) (Tcl_Interp *interp, int type); /* 531 */
+    int (*tcl_LimitGetCommands) (Tcl_Interp *interp); /* 532 */
+    void (*tcl_LimitGetTime) (Tcl_Interp *interp, Tcl_Time *timeLimitPtr); /* 533 */
+    int (*tcl_LimitGetGranularity) (Tcl_Interp *interp, int type); /* 534 */
+    Tcl_InterpState (*tcl_SaveInterpState) (Tcl_Interp *interp, int status); /* 535 */
+    int (*tcl_RestoreInterpState) (Tcl_Interp *interp, Tcl_InterpState state); /* 536 */
+    void (*tcl_DiscardInterpState) (Tcl_InterpState state); /* 537 */
+    int (*tcl_SetReturnOptions) (Tcl_Interp *interp, Tcl_Obj *options); /* 538 */
+    Tcl_Obj * (*tcl_GetReturnOptions) (Tcl_Interp *interp, int result); /* 539 */
+    int (*tcl_IsEnsemble) (Tcl_Command token); /* 540 */
+    Tcl_Command (*tcl_CreateEnsemble) (Tcl_Interp *interp, const char *name, Tcl_Namespace *namespacePtr, int flags); /* 541 */
+    Tcl_Command (*tcl_FindEnsemble) (Tcl_Interp *interp, Tcl_Obj *cmdNameObj, int flags); /* 542 */
+    int (*tcl_SetEnsembleSubcommandList) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj *subcmdList); /* 543 */
+    int (*tcl_SetEnsembleMappingDict) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj *mapDict); /* 544 */
+    int (*tcl_SetEnsembleUnknownHandler) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj *unknownList); /* 545 */
+    int (*tcl_SetEnsembleFlags) (Tcl_Interp *interp, Tcl_Command token, int flags); /* 546 */
+    int (*tcl_GetEnsembleSubcommandList) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj **subcmdListPtr); /* 547 */
+    int (*tcl_GetEnsembleMappingDict) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj **mapDictPtr); /* 548 */
+    int (*tcl_GetEnsembleUnknownHandler) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj **unknownListPtr); /* 549 */
+    int (*tcl_GetEnsembleFlags) (Tcl_Interp *interp, Tcl_Command token, int *flagsPtr); /* 550 */
+    int (*tcl_GetEnsembleNamespace) (Tcl_Interp *interp, Tcl_Command token, Tcl_Namespace **namespacePtrPtr); /* 551 */
+    void (*tcl_SetTimeProc) (Tcl_GetTimeProc *getProc, Tcl_ScaleTimeProc *scaleProc, ClientData clientData); /* 552 */
+    void (*tcl_QueryTimeProc) (Tcl_GetTimeProc **getProc, Tcl_ScaleTimeProc **scaleProc, ClientData *clientData); /* 553 */
+    Tcl_DriverThreadActionProc * (*tcl_ChannelThreadActionProc) (const Tcl_ChannelType *chanTypePtr); /* 554 */
+    Tcl_Obj * (*tcl_NewBignumObj) (mp_int *value); /* 555 */
+    Tcl_Obj * (*tcl_DbNewBignumObj) (mp_int *value, const char *file, int line); /* 556 */
+    void (*tcl_SetBignumObj) (Tcl_Obj *obj, mp_int *value); /* 557 */
+    int (*tcl_GetBignumFromObj) (Tcl_Interp *interp, Tcl_Obj *obj, mp_int *value); /* 558 */
+    int (*tcl_TakeBignumFromObj) (Tcl_Interp *interp, Tcl_Obj *obj, mp_int *value); /* 559 */
+    int (*tcl_TruncateChannel) (Tcl_Channel chan, Tcl_WideInt length); /* 560 */
+    Tcl_DriverTruncateProc * (*tcl_ChannelTruncateProc) (const Tcl_ChannelType *chanTypePtr); /* 561 */
+    void (*tcl_SetChannelErrorInterp) (Tcl_Interp *interp, Tcl_Obj *msg); /* 562 */
+    void (*tcl_GetChannelErrorInterp) (Tcl_Interp *interp, Tcl_Obj **msg); /* 563 */
+    void (*tcl_SetChannelError) (Tcl_Channel chan, Tcl_Obj *msg); /* 564 */
+    void (*tcl_GetChannelError) (Tcl_Channel chan, Tcl_Obj **msg); /* 565 */
+    int (*tcl_InitBignumFromDouble) (Tcl_Interp *interp, double initval, mp_int *toInit); /* 566 */
+    Tcl_Obj * (*tcl_GetNamespaceUnknownHandler) (Tcl_Interp *interp, Tcl_Namespace *nsPtr); /* 567 */
+    int (*tcl_SetNamespaceUnknownHandler) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, Tcl_Obj *handlerPtr); /* 568 */
+    int (*tcl_GetEncodingFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_Encoding *encodingPtr); /* 569 */
+    Tcl_Obj * (*tcl_GetEncodingSearchPath) (void); /* 570 */
+    int (*tcl_SetEncodingSearchPath) (Tcl_Obj *searchPath); /* 571 */
+    const char * (*tcl_GetEncodingNameFromEnvironment) (Tcl_DString *bufPtr); /* 572 */
+    int (*tcl_PkgRequireProc) (Tcl_Interp *interp, const char *name, int objc, Tcl_Obj *const objv[], void *clientDataPtr); /* 573 */
+    void (*tcl_AppendObjToErrorInfo) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 574 */
+    void (*tcl_AppendLimitedToObj) (Tcl_Obj *objPtr, const char *bytes, int length, int limit, const char *ellipsis); /* 575 */
+    Tcl_Obj * (*tcl_Format) (Tcl_Interp *interp, const char *format, int objc, Tcl_Obj *const objv[]); /* 576 */
+    int (*tcl_AppendFormatToObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, const char *format, int objc, Tcl_Obj *const objv[]); /* 577 */
+    Tcl_Obj * (*tcl_ObjPrintf) (const char *format, ...) TCL_FORMAT_PRINTF(1, 2); /* 578 */
+    void (*tcl_AppendPrintfToObj) (Tcl_Obj *objPtr, const char *format, ...) TCL_FORMAT_PRINTF(2, 3); /* 579 */
+    int (*tcl_CancelEval) (Tcl_Interp *interp, Tcl_Obj *resultObjPtr, ClientData clientData, int flags); /* 580 */
+    int (*tcl_Canceled) (Tcl_Interp *interp, int flags); /* 581 */
+    int (*tcl_CreatePipe) (Tcl_Interp *interp, Tcl_Channel *rchan, Tcl_Channel *wchan, int flags); /* 582 */
+    Tcl_Command (*tcl_NRCreateCommand) (Tcl_Interp *interp, const char *cmdName, Tcl_ObjCmdProc *proc, Tcl_ObjCmdProc *nreProc, ClientData clientData, Tcl_CmdDeleteProc *deleteProc); /* 583 */
+    int (*tcl_NREvalObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags); /* 584 */
+    int (*tcl_NREvalObjv) (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[], int flags); /* 585 */
+    int (*tcl_NRCmdSwap) (Tcl_Interp *interp, Tcl_Command cmd, int objc, Tcl_Obj *const objv[], int flags); /* 586 */
+    void (*tcl_NRAddCallback) (Tcl_Interp *interp, Tcl_NRPostProc *postProcPtr, ClientData data0, ClientData data1, ClientData data2, ClientData data3); /* 587 */
+    int (*tcl_NRCallObjProc) (Tcl_Interp *interp, Tcl_ObjCmdProc *objProc, ClientData clientData, int objc, Tcl_Obj *const objv[]); /* 588 */
+    unsigned (*tcl_GetFSDeviceFromStat) (const Tcl_StatBuf *statPtr); /* 589 */
+    unsigned (*tcl_GetFSInodeFromStat) (const Tcl_StatBuf *statPtr); /* 590 */
+    unsigned (*tcl_GetModeFromStat) (const Tcl_StatBuf *statPtr); /* 591 */
+    int (*tcl_GetLinkCountFromStat) (const Tcl_StatBuf *statPtr); /* 592 */
+    int (*tcl_GetUserIdFromStat) (const Tcl_StatBuf *statPtr); /* 593 */
+    int (*tcl_GetGroupIdFromStat) (const Tcl_StatBuf *statPtr); /* 594 */
+    int (*tcl_GetDeviceTypeFromStat) (const Tcl_StatBuf *statPtr); /* 595 */
+    Tcl_WideInt (*tcl_GetAccessTimeFromStat) (const Tcl_StatBuf *statPtr); /* 596 */
+    Tcl_WideInt (*tcl_GetModificationTimeFromStat) (const Tcl_StatBuf *statPtr); /* 597 */
+    Tcl_WideInt (*tcl_GetChangeTimeFromStat) (const Tcl_StatBuf *statPtr); /* 598 */
+    Tcl_WideUInt (*tcl_GetSizeFromStat) (const Tcl_StatBuf *statPtr); /* 599 */
+    Tcl_WideUInt (*tcl_GetBlocksFromStat) (const Tcl_StatBuf *statPtr); /* 600 */
+    unsigned (*tcl_GetBlockSizeFromStat) (const Tcl_StatBuf *statPtr); /* 601 */
+    int (*tcl_SetEnsembleParameterList) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj *paramList); /* 602 */
+    int (*tcl_GetEnsembleParameterList) (Tcl_Interp *interp, Tcl_Command token, Tcl_Obj **paramListPtr); /* 603 */
+    int (*tcl_ParseArgsObjv) (Tcl_Interp *interp, const Tcl_ArgvInfo *argTable, int *objcPtr, Tcl_Obj *const *objv, Tcl_Obj ***remObjv); /* 604 */
+    int (*tcl_GetErrorLine) (Tcl_Interp *interp); /* 605 */
+    void (*tcl_SetErrorLine) (Tcl_Interp *interp, int lineNum); /* 606 */
+    void (*tcl_TransferResult) (Tcl_Interp *sourceInterp, int result, Tcl_Interp *targetInterp); /* 607 */
+    int (*tcl_InterpActive) (Tcl_Interp *interp); /* 608 */
+    void (*tcl_BackgroundException) (Tcl_Interp *interp, int code); /* 609 */
+    int (*tcl_ZlibDeflate) (Tcl_Interp *interp, int format, Tcl_Obj *data, int level, Tcl_Obj *gzipHeaderDictObj); /* 610 */
+    int (*tcl_ZlibInflate) (Tcl_Interp *interp, int format, Tcl_Obj *data, int buffersize, Tcl_Obj *gzipHeaderDictObj); /* 611 */
+    unsigned int (*tcl_ZlibCRC32) (unsigned int crc, const unsigned char *buf, int len); /* 612 */
+    unsigned int (*tcl_ZlibAdler32) (unsigned int adler, const unsigned char *buf, int len); /* 613 */
+    int (*tcl_ZlibStreamInit) (Tcl_Interp *interp, int mode, int format, int level, Tcl_Obj *dictObj, Tcl_ZlibStream *zshandle); /* 614 */
+    Tcl_Obj * (*tcl_ZlibStreamGetCommandName) (Tcl_ZlibStream zshandle); /* 615 */
+    int (*tcl_ZlibStreamEof) (Tcl_ZlibStream zshandle); /* 616 */
+    int (*tcl_ZlibStreamChecksum) (Tcl_ZlibStream zshandle); /* 617 */
+    int (*tcl_ZlibStreamPut) (Tcl_ZlibStream zshandle, Tcl_Obj *data, int flush); /* 618 */
+    int (*tcl_ZlibStreamGet) (Tcl_ZlibStream zshandle, Tcl_Obj *data, int count); /* 619 */
+    int (*tcl_ZlibStreamClose) (Tcl_ZlibStream zshandle); /* 620 */
+    int (*tcl_ZlibStreamReset) (Tcl_ZlibStream zshandle); /* 621 */
+    void (*tcl_SetStartupScript) (Tcl_Obj *path, const char *encoding); /* 622 */
+    Tcl_Obj * (*tcl_GetStartupScript) (const char **encodingPtr); /* 623 */
+    int (*tcl_CloseEx) (Tcl_Interp *interp, Tcl_Channel chan, int flags); /* 624 */
+    int (*tcl_NRExprObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_Obj *resultPtr); /* 625 */
+    int (*tcl_NRSubstObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags); /* 626 */
+    int (*tcl_LoadFile) (Tcl_Interp *interp, Tcl_Obj *pathPtr, const char *const symv[], int flags, void *procPtrs, Tcl_LoadHandle *handlePtr); /* 627 */
+    void * (*tcl_FindSymbol) (Tcl_Interp *interp, Tcl_LoadHandle handle, const char *symbol); /* 628 */
+    int (*tcl_FSUnloadFile) (Tcl_Interp *interp, Tcl_LoadHandle handlePtr); /* 629 */
+    void (*tcl_ZlibStreamSetCompressionDictionary) (Tcl_ZlibStream zhandle, Tcl_Obj *compressionDictionaryObj); /* 630 */
+    Tcl_Channel (*tcl_OpenTcpServerEx) (Tcl_Interp *interp, const char *service, const char *host, unsigned int flags, Tcl_TcpAcceptProc *acceptProc, ClientData callbackData); /* 631 */
+} TclStubs;
+
+extern const TclStubs *tclStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCL_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#define Tcl_PkgProvideEx \
+	(tclStubsPtr->tcl_PkgProvideEx) /* 0 */
+#define Tcl_PkgRequireEx \
+	(tclStubsPtr->tcl_PkgRequireEx) /* 1 */
+#define Tcl_Panic \
+	(tclStubsPtr->tcl_Panic) /* 2 */
+#define Tcl_Alloc \
+	(tclStubsPtr->tcl_Alloc) /* 3 */
+#define Tcl_Free \
+	(tclStubsPtr->tcl_Free) /* 4 */
+#define Tcl_Realloc \
+	(tclStubsPtr->tcl_Realloc) /* 5 */
+#define Tcl_DbCkalloc \
+	(tclStubsPtr->tcl_DbCkalloc) /* 6 */
+#define Tcl_DbCkfree \
+	(tclStubsPtr->tcl_DbCkfree) /* 7 */
+#define Tcl_DbCkrealloc \
+	(tclStubsPtr->tcl_DbCkrealloc) /* 8 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+#define Tcl_CreateFileHandler \
+	(tclStubsPtr->tcl_CreateFileHandler) /* 9 */
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+#define Tcl_CreateFileHandler \
+	(tclStubsPtr->tcl_CreateFileHandler) /* 9 */
+#endif /* MACOSX */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+#define Tcl_DeleteFileHandler \
+	(tclStubsPtr->tcl_DeleteFileHandler) /* 10 */
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+#define Tcl_DeleteFileHandler \
+	(tclStubsPtr->tcl_DeleteFileHandler) /* 10 */
+#endif /* MACOSX */
+#define Tcl_SetTimer \
+	(tclStubsPtr->tcl_SetTimer) /* 11 */
+#define Tcl_Sleep \
+	(tclStubsPtr->tcl_Sleep) /* 12 */
+#define Tcl_WaitForEvent \
+	(tclStubsPtr->tcl_WaitForEvent) /* 13 */
+#define Tcl_AppendAllObjTypes \
+	(tclStubsPtr->tcl_AppendAllObjTypes) /* 14 */
+#define Tcl_AppendStringsToObj \
+	(tclStubsPtr->tcl_AppendStringsToObj) /* 15 */
+#define Tcl_AppendToObj \
+	(tclStubsPtr->tcl_AppendToObj) /* 16 */
+#define Tcl_ConcatObj \
+	(tclStubsPtr->tcl_ConcatObj) /* 17 */
+#define Tcl_ConvertToType \
+	(tclStubsPtr->tcl_ConvertToType) /* 18 */
+#define Tcl_DbDecrRefCount \
+	(tclStubsPtr->tcl_DbDecrRefCount) /* 19 */
+#define Tcl_DbIncrRefCount \
+	(tclStubsPtr->tcl_DbIncrRefCount) /* 20 */
+#define Tcl_DbIsShared \
+	(tclStubsPtr->tcl_DbIsShared) /* 21 */
+#define Tcl_DbNewBooleanObj \
+	(tclStubsPtr->tcl_DbNewBooleanObj) /* 22 */
+#define Tcl_DbNewByteArrayObj \
+	(tclStubsPtr->tcl_DbNewByteArrayObj) /* 23 */
+#define Tcl_DbNewDoubleObj \
+	(tclStubsPtr->tcl_DbNewDoubleObj) /* 24 */
+#define Tcl_DbNewListObj \
+	(tclStubsPtr->tcl_DbNewListObj) /* 25 */
+#define Tcl_DbNewLongObj \
+	(tclStubsPtr->tcl_DbNewLongObj) /* 26 */
+#define Tcl_DbNewObj \
+	(tclStubsPtr->tcl_DbNewObj) /* 27 */
+#define Tcl_DbNewStringObj \
+	(tclStubsPtr->tcl_DbNewStringObj) /* 28 */
+#define Tcl_DuplicateObj \
+	(tclStubsPtr->tcl_DuplicateObj) /* 29 */
+#define TclFreeObj \
+	(tclStubsPtr->tclFreeObj) /* 30 */
+#define Tcl_GetBoolean \
+	(tclStubsPtr->tcl_GetBoolean) /* 31 */
+#define Tcl_GetBooleanFromObj \
+	(tclStubsPtr->tcl_GetBooleanFromObj) /* 32 */
+#define Tcl_GetByteArrayFromObj \
+	(tclStubsPtr->tcl_GetByteArrayFromObj) /* 33 */
+#define Tcl_GetDouble \
+	(tclStubsPtr->tcl_GetDouble) /* 34 */
+#define Tcl_GetDoubleFromObj \
+	(tclStubsPtr->tcl_GetDoubleFromObj) /* 35 */
+#define Tcl_GetIndexFromObj \
+	(tclStubsPtr->tcl_GetIndexFromObj) /* 36 */
+#define Tcl_GetInt \
+	(tclStubsPtr->tcl_GetInt) /* 37 */
+#define Tcl_GetIntFromObj \
+	(tclStubsPtr->tcl_GetIntFromObj) /* 38 */
+#define Tcl_GetLongFromObj \
+	(tclStubsPtr->tcl_GetLongFromObj) /* 39 */
+#define Tcl_GetObjType \
+	(tclStubsPtr->tcl_GetObjType) /* 40 */
+#define Tcl_GetStringFromObj \
+	(tclStubsPtr->tcl_GetStringFromObj) /* 41 */
+#define Tcl_InvalidateStringRep \
+	(tclStubsPtr->tcl_InvalidateStringRep) /* 42 */
+#define Tcl_ListObjAppendList \
+	(tclStubsPtr->tcl_ListObjAppendList) /* 43 */
+#define Tcl_ListObjAppendElement \
+	(tclStubsPtr->tcl_ListObjAppendElement) /* 44 */
+#define Tcl_ListObjGetElements \
+	(tclStubsPtr->tcl_ListObjGetElements) /* 45 */
+#define Tcl_ListObjIndex \
+	(tclStubsPtr->tcl_ListObjIndex) /* 46 */
+#define Tcl_ListObjLength \
+	(tclStubsPtr->tcl_ListObjLength) /* 47 */
+#define Tcl_ListObjReplace \
+	(tclStubsPtr->tcl_ListObjReplace) /* 48 */
+#define Tcl_NewBooleanObj \
+	(tclStubsPtr->tcl_NewBooleanObj) /* 49 */
+#define Tcl_NewByteArrayObj \
+	(tclStubsPtr->tcl_NewByteArrayObj) /* 50 */
+#define Tcl_NewDoubleObj \
+	(tclStubsPtr->tcl_NewDoubleObj) /* 51 */
+#define Tcl_NewIntObj \
+	(tclStubsPtr->tcl_NewIntObj) /* 52 */
+#define Tcl_NewListObj \
+	(tclStubsPtr->tcl_NewListObj) /* 53 */
+#define Tcl_NewLongObj \
+	(tclStubsPtr->tcl_NewLongObj) /* 54 */
+#define Tcl_NewObj \
+	(tclStubsPtr->tcl_NewObj) /* 55 */
+#define Tcl_NewStringObj \
+	(tclStubsPtr->tcl_NewStringObj) /* 56 */
+#define Tcl_SetBooleanObj \
+	(tclStubsPtr->tcl_SetBooleanObj) /* 57 */
+#define Tcl_SetByteArrayLength \
+	(tclStubsPtr->tcl_SetByteArrayLength) /* 58 */
+#define Tcl_SetByteArrayObj \
+	(tclStubsPtr->tcl_SetByteArrayObj) /* 59 */
+#define Tcl_SetDoubleObj \
+	(tclStubsPtr->tcl_SetDoubleObj) /* 60 */
+#define Tcl_SetIntObj \
+	(tclStubsPtr->tcl_SetIntObj) /* 61 */
+#define Tcl_SetListObj \
+	(tclStubsPtr->tcl_SetListObj) /* 62 */
+#define Tcl_SetLongObj \
+	(tclStubsPtr->tcl_SetLongObj) /* 63 */
+#define Tcl_SetObjLength \
+	(tclStubsPtr->tcl_SetObjLength) /* 64 */
+#define Tcl_SetStringObj \
+	(tclStubsPtr->tcl_SetStringObj) /* 65 */
+#define Tcl_AddErrorInfo \
+	(tclStubsPtr->tcl_AddErrorInfo) /* 66 */
+#define Tcl_AddObjErrorInfo \
+	(tclStubsPtr->tcl_AddObjErrorInfo) /* 67 */
+#define Tcl_AllowExceptions \
+	(tclStubsPtr->tcl_AllowExceptions) /* 68 */
+#define Tcl_AppendElement \
+	(tclStubsPtr->tcl_AppendElement) /* 69 */
+#define Tcl_AppendResult \
+	(tclStubsPtr->tcl_AppendResult) /* 70 */
+#define Tcl_AsyncCreate \
+	(tclStubsPtr->tcl_AsyncCreate) /* 71 */
+#define Tcl_AsyncDelete \
+	(tclStubsPtr->tcl_AsyncDelete) /* 72 */
+#define Tcl_AsyncInvoke \
+	(tclStubsPtr->tcl_AsyncInvoke) /* 73 */
+#define Tcl_AsyncMark \
+	(tclStubsPtr->tcl_AsyncMark) /* 74 */
+#define Tcl_AsyncReady \
+	(tclStubsPtr->tcl_AsyncReady) /* 75 */
+#define Tcl_BackgroundError \
+	(tclStubsPtr->tcl_BackgroundError) /* 76 */
+#define Tcl_Backslash \
+	(tclStubsPtr->tcl_Backslash) /* 77 */
+#define Tcl_BadChannelOption \
+	(tclStubsPtr->tcl_BadChannelOption) /* 78 */
+#define Tcl_CallWhenDeleted \
+	(tclStubsPtr->tcl_CallWhenDeleted) /* 79 */
+#define Tcl_CancelIdleCall \
+	(tclStubsPtr->tcl_CancelIdleCall) /* 80 */
+#define Tcl_Close \
+	(tclStubsPtr->tcl_Close) /* 81 */
+#define Tcl_CommandComplete \
+	(tclStubsPtr->tcl_CommandComplete) /* 82 */
+#define Tcl_Concat \
+	(tclStubsPtr->tcl_Concat) /* 83 */
+#define Tcl_ConvertElement \
+	(tclStubsPtr->tcl_ConvertElement) /* 84 */
+#define Tcl_ConvertCountedElement \
+	(tclStubsPtr->tcl_ConvertCountedElement) /* 85 */
+#define Tcl_CreateAlias \
+	(tclStubsPtr->tcl_CreateAlias) /* 86 */
+#define Tcl_CreateAliasObj \
+	(tclStubsPtr->tcl_CreateAliasObj) /* 87 */
+#define Tcl_CreateChannel \
+	(tclStubsPtr->tcl_CreateChannel) /* 88 */
+#define Tcl_CreateChannelHandler \
+	(tclStubsPtr->tcl_CreateChannelHandler) /* 89 */
+#define Tcl_CreateCloseHandler \
+	(tclStubsPtr->tcl_CreateCloseHandler) /* 90 */
+#define Tcl_CreateCommand \
+	(tclStubsPtr->tcl_CreateCommand) /* 91 */
+#define Tcl_CreateEventSource \
+	(tclStubsPtr->tcl_CreateEventSource) /* 92 */
+#define Tcl_CreateExitHandler \
+	(tclStubsPtr->tcl_CreateExitHandler) /* 93 */
+#define Tcl_CreateInterp \
+	(tclStubsPtr->tcl_CreateInterp) /* 94 */
+#define Tcl_CreateMathFunc \
+	(tclStubsPtr->tcl_CreateMathFunc) /* 95 */
+#define Tcl_CreateObjCommand \
+	(tclStubsPtr->tcl_CreateObjCommand) /* 96 */
+#define Tcl_CreateSlave \
+	(tclStubsPtr->tcl_CreateSlave) /* 97 */
+#define Tcl_CreateTimerHandler \
+	(tclStubsPtr->tcl_CreateTimerHandler) /* 98 */
+#define Tcl_CreateTrace \
+	(tclStubsPtr->tcl_CreateTrace) /* 99 */
+#define Tcl_DeleteAssocData \
+	(tclStubsPtr->tcl_DeleteAssocData) /* 100 */
+#define Tcl_DeleteChannelHandler \
+	(tclStubsPtr->tcl_DeleteChannelHandler) /* 101 */
+#define Tcl_DeleteCloseHandler \
+	(tclStubsPtr->tcl_DeleteCloseHandler) /* 102 */
+#define Tcl_DeleteCommand \
+	(tclStubsPtr->tcl_DeleteCommand) /* 103 */
+#define Tcl_DeleteCommandFromToken \
+	(tclStubsPtr->tcl_DeleteCommandFromToken) /* 104 */
+#define Tcl_DeleteEvents \
+	(tclStubsPtr->tcl_DeleteEvents) /* 105 */
+#define Tcl_DeleteEventSource \
+	(tclStubsPtr->tcl_DeleteEventSource) /* 106 */
+#define Tcl_DeleteExitHandler \
+	(tclStubsPtr->tcl_DeleteExitHandler) /* 107 */
+#define Tcl_DeleteHashEntry \
+	(tclStubsPtr->tcl_DeleteHashEntry) /* 108 */
+#define Tcl_DeleteHashTable \
+	(tclStubsPtr->tcl_DeleteHashTable) /* 109 */
+#define Tcl_DeleteInterp \
+	(tclStubsPtr->tcl_DeleteInterp) /* 110 */
+#define Tcl_DetachPids \
+	(tclStubsPtr->tcl_DetachPids) /* 111 */
+#define Tcl_DeleteTimerHandler \
+	(tclStubsPtr->tcl_DeleteTimerHandler) /* 112 */
+#define Tcl_DeleteTrace \
+	(tclStubsPtr->tcl_DeleteTrace) /* 113 */
+#define Tcl_DontCallWhenDeleted \
+	(tclStubsPtr->tcl_DontCallWhenDeleted) /* 114 */
+#define Tcl_DoOneEvent \
+	(tclStubsPtr->tcl_DoOneEvent) /* 115 */
+#define Tcl_DoWhenIdle \
+	(tclStubsPtr->tcl_DoWhenIdle) /* 116 */
+#define Tcl_DStringAppend \
+	(tclStubsPtr->tcl_DStringAppend) /* 117 */
+#define Tcl_DStringAppendElement \
+	(tclStubsPtr->tcl_DStringAppendElement) /* 118 */
+#define Tcl_DStringEndSublist \
+	(tclStubsPtr->tcl_DStringEndSublist) /* 119 */
+#define Tcl_DStringFree \
+	(tclStubsPtr->tcl_DStringFree) /* 120 */
+#define Tcl_DStringGetResult \
+	(tclStubsPtr->tcl_DStringGetResult) /* 121 */
+#define Tcl_DStringInit \
+	(tclStubsPtr->tcl_DStringInit) /* 122 */
+#define Tcl_DStringResult \
+	(tclStubsPtr->tcl_DStringResult) /* 123 */
+#define Tcl_DStringSetLength \
+	(tclStubsPtr->tcl_DStringSetLength) /* 124 */
+#define Tcl_DStringStartSublist \
+	(tclStubsPtr->tcl_DStringStartSublist) /* 125 */
+#define Tcl_Eof \
+	(tclStubsPtr->tcl_Eof) /* 126 */
+#define Tcl_ErrnoId \
+	(tclStubsPtr->tcl_ErrnoId) /* 127 */
+#define Tcl_ErrnoMsg \
+	(tclStubsPtr->tcl_ErrnoMsg) /* 128 */
+#define Tcl_Eval \
+	(tclStubsPtr->tcl_Eval) /* 129 */
+#define Tcl_EvalFile \
+	(tclStubsPtr->tcl_EvalFile) /* 130 */
+#define Tcl_EvalObj \
+	(tclStubsPtr->tcl_EvalObj) /* 131 */
+#define Tcl_EventuallyFree \
+	(tclStubsPtr->tcl_EventuallyFree) /* 132 */
+#define Tcl_Exit \
+	(tclStubsPtr->tcl_Exit) /* 133 */
+#define Tcl_ExposeCommand \
+	(tclStubsPtr->tcl_ExposeCommand) /* 134 */
+#define Tcl_ExprBoolean \
+	(tclStubsPtr->tcl_ExprBoolean) /* 135 */
+#define Tcl_ExprBooleanObj \
+	(tclStubsPtr->tcl_ExprBooleanObj) /* 136 */
+#define Tcl_ExprDouble \
+	(tclStubsPtr->tcl_ExprDouble) /* 137 */
+#define Tcl_ExprDoubleObj \
+	(tclStubsPtr->tcl_ExprDoubleObj) /* 138 */
+#define Tcl_ExprLong \
+	(tclStubsPtr->tcl_ExprLong) /* 139 */
+#define Tcl_ExprLongObj \
+	(tclStubsPtr->tcl_ExprLongObj) /* 140 */
+#define Tcl_ExprObj \
+	(tclStubsPtr->tcl_ExprObj) /* 141 */
+#define Tcl_ExprString \
+	(tclStubsPtr->tcl_ExprString) /* 142 */
+#define Tcl_Finalize \
+	(tclStubsPtr->tcl_Finalize) /* 143 */
+#define Tcl_FindExecutable \
+	(tclStubsPtr->tcl_FindExecutable) /* 144 */
+#define Tcl_FirstHashEntry \
+	(tclStubsPtr->tcl_FirstHashEntry) /* 145 */
+#define Tcl_Flush \
+	(tclStubsPtr->tcl_Flush) /* 146 */
+#define Tcl_FreeResult \
+	(tclStubsPtr->tcl_FreeResult) /* 147 */
+#define Tcl_GetAlias \
+	(tclStubsPtr->tcl_GetAlias) /* 148 */
+#define Tcl_GetAliasObj \
+	(tclStubsPtr->tcl_GetAliasObj) /* 149 */
+#define Tcl_GetAssocData \
+	(tclStubsPtr->tcl_GetAssocData) /* 150 */
+#define Tcl_GetChannel \
+	(tclStubsPtr->tcl_GetChannel) /* 151 */
+#define Tcl_GetChannelBufferSize \
+	(tclStubsPtr->tcl_GetChannelBufferSize) /* 152 */
+#define Tcl_GetChannelHandle \
+	(tclStubsPtr->tcl_GetChannelHandle) /* 153 */
+#define Tcl_GetChannelInstanceData \
+	(tclStubsPtr->tcl_GetChannelInstanceData) /* 154 */
+#define Tcl_GetChannelMode \
+	(tclStubsPtr->tcl_GetChannelMode) /* 155 */
+#define Tcl_GetChannelName \
+	(tclStubsPtr->tcl_GetChannelName) /* 156 */
+#define Tcl_GetChannelOption \
+	(tclStubsPtr->tcl_GetChannelOption) /* 157 */
+#define Tcl_GetChannelType \
+	(tclStubsPtr->tcl_GetChannelType) /* 158 */
+#define Tcl_GetCommandInfo \
+	(tclStubsPtr->tcl_GetCommandInfo) /* 159 */
+#define Tcl_GetCommandName \
+	(tclStubsPtr->tcl_GetCommandName) /* 160 */
+#define Tcl_GetErrno \
+	(tclStubsPtr->tcl_GetErrno) /* 161 */
+#define Tcl_GetHostName \
+	(tclStubsPtr->tcl_GetHostName) /* 162 */
+#define Tcl_GetInterpPath \
+	(tclStubsPtr->tcl_GetInterpPath) /* 163 */
+#define Tcl_GetMaster \
+	(tclStubsPtr->tcl_GetMaster) /* 164 */
+#define Tcl_GetNameOfExecutable \
+	(tclStubsPtr->tcl_GetNameOfExecutable) /* 165 */
+#define Tcl_GetObjResult \
+	(tclStubsPtr->tcl_GetObjResult) /* 166 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+#define Tcl_GetOpenFile \
+	(tclStubsPtr->tcl_GetOpenFile) /* 167 */
+#endif /* UNIX */
+#ifdef MAC_OSX_TCL /* MACOSX */
+#define Tcl_GetOpenFile \
+	(tclStubsPtr->tcl_GetOpenFile) /* 167 */
+#endif /* MACOSX */
+#define Tcl_GetPathType \
+	(tclStubsPtr->tcl_GetPathType) /* 168 */
+#define Tcl_Gets \
+	(tclStubsPtr->tcl_Gets) /* 169 */
+#define Tcl_GetsObj \
+	(tclStubsPtr->tcl_GetsObj) /* 170 */
+#define Tcl_GetServiceMode \
+	(tclStubsPtr->tcl_GetServiceMode) /* 171 */
+#define Tcl_GetSlave \
+	(tclStubsPtr->tcl_GetSlave) /* 172 */
+#define Tcl_GetStdChannel \
+	(tclStubsPtr->tcl_GetStdChannel) /* 173 */
+#define Tcl_GetStringResult \
+	(tclStubsPtr->tcl_GetStringResult) /* 174 */
+#define Tcl_GetVar \
+	(tclStubsPtr->tcl_GetVar) /* 175 */
+#define Tcl_GetVar2 \
+	(tclStubsPtr->tcl_GetVar2) /* 176 */
+#define Tcl_GlobalEval \
+	(tclStubsPtr->tcl_GlobalEval) /* 177 */
+#define Tcl_GlobalEvalObj \
+	(tclStubsPtr->tcl_GlobalEvalObj) /* 178 */
+#define Tcl_HideCommand \
+	(tclStubsPtr->tcl_HideCommand) /* 179 */
+#define Tcl_Init \
+	(tclStubsPtr->tcl_Init) /* 180 */
+#define Tcl_InitHashTable \
+	(tclStubsPtr->tcl_InitHashTable) /* 181 */
+#define Tcl_InputBlocked \
+	(tclStubsPtr->tcl_InputBlocked) /* 182 */
+#define Tcl_InputBuffered \
+	(tclStubsPtr->tcl_InputBuffered) /* 183 */
+#define Tcl_InterpDeleted \
+	(tclStubsPtr->tcl_InterpDeleted) /* 184 */
+#define Tcl_IsSafe \
+	(tclStubsPtr->tcl_IsSafe) /* 185 */
+#define Tcl_JoinPath \
+	(tclStubsPtr->tcl_JoinPath) /* 186 */
+#define Tcl_LinkVar \
+	(tclStubsPtr->tcl_LinkVar) /* 187 */
+/* Slot 188 is reserved */
+#define Tcl_MakeFileChannel \
+	(tclStubsPtr->tcl_MakeFileChannel) /* 189 */
+#define Tcl_MakeSafe \
+	(tclStubsPtr->tcl_MakeSafe) /* 190 */
+#define Tcl_MakeTcpClientChannel \
+	(tclStubsPtr->tcl_MakeTcpClientChannel) /* 191 */
+#define Tcl_Merge \
+	(tclStubsPtr->tcl_Merge) /* 192 */
+#define Tcl_NextHashEntry \
+	(tclStubsPtr->tcl_NextHashEntry) /* 193 */
+#define Tcl_NotifyChannel \
+	(tclStubsPtr->tcl_NotifyChannel) /* 194 */
+#define Tcl_ObjGetVar2 \
+	(tclStubsPtr->tcl_ObjGetVar2) /* 195 */
+#define Tcl_ObjSetVar2 \
+	(tclStubsPtr->tcl_ObjSetVar2) /* 196 */
+#define Tcl_OpenCommandChannel \
+	(tclStubsPtr->tcl_OpenCommandChannel) /* 197 */
+#define Tcl_OpenFileChannel \
+	(tclStubsPtr->tcl_OpenFileChannel) /* 198 */
+#define Tcl_OpenTcpClient \
+	(tclStubsPtr->tcl_OpenTcpClient) /* 199 */
+#define Tcl_OpenTcpServer \
+	(tclStubsPtr->tcl_OpenTcpServer) /* 200 */
+#define Tcl_Preserve \
+	(tclStubsPtr->tcl_Preserve) /* 201 */
+#define Tcl_PrintDouble \
+	(tclStubsPtr->tcl_PrintDouble) /* 202 */
+#define Tcl_PutEnv \
+	(tclStubsPtr->tcl_PutEnv) /* 203 */
+#define Tcl_PosixError \
+	(tclStubsPtr->tcl_PosixError) /* 204 */
+#define Tcl_QueueEvent \
+	(tclStubsPtr->tcl_QueueEvent) /* 205 */
+#define Tcl_Read \
+	(tclStubsPtr->tcl_Read) /* 206 */
+#define Tcl_ReapDetachedProcs \
+	(tclStubsPtr->tcl_ReapDetachedProcs) /* 207 */
+#define Tcl_RecordAndEval \
+	(tclStubsPtr->tcl_RecordAndEval) /* 208 */
+#define Tcl_RecordAndEvalObj \
+	(tclStubsPtr->tcl_RecordAndEvalObj) /* 209 */
+#define Tcl_RegisterChannel \
+	(tclStubsPtr->tcl_RegisterChannel) /* 210 */
+#define Tcl_RegisterObjType \
+	(tclStubsPtr->tcl_RegisterObjType) /* 211 */
+#define Tcl_RegExpCompile \
+	(tclStubsPtr->tcl_RegExpCompile) /* 212 */
+#define Tcl_RegExpExec \
+	(tclStubsPtr->tcl_RegExpExec) /* 213 */
+#define Tcl_RegExpMatch \
+	(tclStubsPtr->tcl_RegExpMatch) /* 214 */
+#define Tcl_RegExpRange \
+	(tclStubsPtr->tcl_RegExpRange) /* 215 */
+#define Tcl_Release \
+	(tclStubsPtr->tcl_Release) /* 216 */
+#define Tcl_ResetResult \
+	(tclStubsPtr->tcl_ResetResult) /* 217 */
+#define Tcl_ScanElement \
+	(tclStubsPtr->tcl_ScanElement) /* 218 */
+#define Tcl_ScanCountedElement \
+	(tclStubsPtr->tcl_ScanCountedElement) /* 219 */
+#define Tcl_SeekOld \
+	(tclStubsPtr->tcl_SeekOld) /* 220 */
+#define Tcl_ServiceAll \
+	(tclStubsPtr->tcl_ServiceAll) /* 221 */
+#define Tcl_ServiceEvent \
+	(tclStubsPtr->tcl_ServiceEvent) /* 222 */
+#define Tcl_SetAssocData \
+	(tclStubsPtr->tcl_SetAssocData) /* 223 */
+#define Tcl_SetChannelBufferSize \
+	(tclStubsPtr->tcl_SetChannelBufferSize) /* 224 */
+#define Tcl_SetChannelOption \
+	(tclStubsPtr->tcl_SetChannelOption) /* 225 */
+#define Tcl_SetCommandInfo \
+	(tclStubsPtr->tcl_SetCommandInfo) /* 226 */
+#define Tcl_SetErrno \
+	(tclStubsPtr->tcl_SetErrno) /* 227 */
+#define Tcl_SetErrorCode \
+	(tclStubsPtr->tcl_SetErrorCode) /* 228 */
+#define Tcl_SetMaxBlockTime \
+	(tclStubsPtr->tcl_SetMaxBlockTime) /* 229 */
+#define Tcl_SetPanicProc \
+	(tclStubsPtr->tcl_SetPanicProc) /* 230 */
+#define Tcl_SetRecursionLimit \
+	(tclStubsPtr->tcl_SetRecursionLimit) /* 231 */
+#define Tcl_SetResult \
+	(tclStubsPtr->tcl_SetResult) /* 232 */
+#define Tcl_SetServiceMode \
+	(tclStubsPtr->tcl_SetServiceMode) /* 233 */
+#define Tcl_SetObjErrorCode \
+	(tclStubsPtr->tcl_SetObjErrorCode) /* 234 */
+#define Tcl_SetObjResult \
+	(tclStubsPtr->tcl_SetObjResult) /* 235 */
+#define Tcl_SetStdChannel \
+	(tclStubsPtr->tcl_SetStdChannel) /* 236 */
+#define Tcl_SetVar \
+	(tclStubsPtr->tcl_SetVar) /* 237 */
+#define Tcl_SetVar2 \
+	(tclStubsPtr->tcl_SetVar2) /* 238 */
+#define Tcl_SignalId \
+	(tclStubsPtr->tcl_SignalId) /* 239 */
+#define Tcl_SignalMsg \
+	(tclStubsPtr->tcl_SignalMsg) /* 240 */
+#define Tcl_SourceRCFile \
+	(tclStubsPtr->tcl_SourceRCFile) /* 241 */
+#define Tcl_SplitList \
+	(tclStubsPtr->tcl_SplitList) /* 242 */
+#define Tcl_SplitPath \
+	(tclStubsPtr->tcl_SplitPath) /* 243 */
+#define Tcl_StaticPackage \
+	(tclStubsPtr->tcl_StaticPackage) /* 244 */
+#define Tcl_StringMatch \
+	(tclStubsPtr->tcl_StringMatch) /* 245 */
+#define Tcl_TellOld \
+	(tclStubsPtr->tcl_TellOld) /* 246 */
+#define Tcl_TraceVar \
+	(tclStubsPtr->tcl_TraceVar) /* 247 */
+#define Tcl_TraceVar2 \
+	(tclStubsPtr->tcl_TraceVar2) /* 248 */
+#define Tcl_TranslateFileName \
+	(tclStubsPtr->tcl_TranslateFileName) /* 249 */
+#define Tcl_Ungets \
+	(tclStubsPtr->tcl_Ungets) /* 250 */
+#define Tcl_UnlinkVar \
+	(tclStubsPtr->tcl_UnlinkVar) /* 251 */
+#define Tcl_UnregisterChannel \
+	(tclStubsPtr->tcl_UnregisterChannel) /* 252 */
+#define Tcl_UnsetVar \
+	(tclStubsPtr->tcl_UnsetVar) /* 253 */
+#define Tcl_UnsetVar2 \
+	(tclStubsPtr->tcl_UnsetVar2) /* 254 */
+#define Tcl_UntraceVar \
+	(tclStubsPtr->tcl_UntraceVar) /* 255 */
+#define Tcl_UntraceVar2 \
+	(tclStubsPtr->tcl_UntraceVar2) /* 256 */
+#define Tcl_UpdateLinkedVar \
+	(tclStubsPtr->tcl_UpdateLinkedVar) /* 257 */
+#define Tcl_UpVar \
+	(tclStubsPtr->tcl_UpVar) /* 258 */
+#define Tcl_UpVar2 \
+	(tclStubsPtr->tcl_UpVar2) /* 259 */
+#define Tcl_VarEval \
+	(tclStubsPtr->tcl_VarEval) /* 260 */
+#define Tcl_VarTraceInfo \
+	(tclStubsPtr->tcl_VarTraceInfo) /* 261 */
+#define Tcl_VarTraceInfo2 \
+	(tclStubsPtr->tcl_VarTraceInfo2) /* 262 */
+#define Tcl_Write \
+	(tclStubsPtr->tcl_Write) /* 263 */
+#define Tcl_WrongNumArgs \
+	(tclStubsPtr->tcl_WrongNumArgs) /* 264 */
+#define Tcl_DumpActiveMemory \
+	(tclStubsPtr->tcl_DumpActiveMemory) /* 265 */
+#define Tcl_ValidateAllMemory \
+	(tclStubsPtr->tcl_ValidateAllMemory) /* 266 */
+#define Tcl_AppendResultVA \
+	(tclStubsPtr->tcl_AppendResultVA) /* 267 */
+#define Tcl_AppendStringsToObjVA \
+	(tclStubsPtr->tcl_AppendStringsToObjVA) /* 268 */
+#define Tcl_HashStats \
+	(tclStubsPtr->tcl_HashStats) /* 269 */
+#define Tcl_ParseVar \
+	(tclStubsPtr->tcl_ParseVar) /* 270 */
+#define Tcl_PkgPresent \
+	(tclStubsPtr->tcl_PkgPresent) /* 271 */
+#define Tcl_PkgPresentEx \
+	(tclStubsPtr->tcl_PkgPresentEx) /* 272 */
+#define Tcl_PkgProvide \
+	(tclStubsPtr->tcl_PkgProvide) /* 273 */
+#define Tcl_PkgRequire \
+	(tclStubsPtr->tcl_PkgRequire) /* 274 */
+#define Tcl_SetErrorCodeVA \
+	(tclStubsPtr->tcl_SetErrorCodeVA) /* 275 */
+#define Tcl_VarEvalVA \
+	(tclStubsPtr->tcl_VarEvalVA) /* 276 */
+#define Tcl_WaitPid \
+	(tclStubsPtr->tcl_WaitPid) /* 277 */
+#define Tcl_PanicVA \
+	(tclStubsPtr->tcl_PanicVA) /* 278 */
+#define Tcl_GetVersion \
+	(tclStubsPtr->tcl_GetVersion) /* 279 */
+#define Tcl_InitMemory \
+	(tclStubsPtr->tcl_InitMemory) /* 280 */
+#define Tcl_StackChannel \
+	(tclStubsPtr->tcl_StackChannel) /* 281 */
+#define Tcl_UnstackChannel \
+	(tclStubsPtr->tcl_UnstackChannel) /* 282 */
+#define Tcl_GetStackedChannel \
+	(tclStubsPtr->tcl_GetStackedChannel) /* 283 */
+#define Tcl_SetMainLoop \
+	(tclStubsPtr->tcl_SetMainLoop) /* 284 */
+/* Slot 285 is reserved */
+#define Tcl_AppendObjToObj \
+	(tclStubsPtr->tcl_AppendObjToObj) /* 286 */
+#define Tcl_CreateEncoding \
+	(tclStubsPtr->tcl_CreateEncoding) /* 287 */
+#define Tcl_CreateThreadExitHandler \
+	(tclStubsPtr->tcl_CreateThreadExitHandler) /* 288 */
+#define Tcl_DeleteThreadExitHandler \
+	(tclStubsPtr->tcl_DeleteThreadExitHandler) /* 289 */
+#define Tcl_DiscardResult \
+	(tclStubsPtr->tcl_DiscardResult) /* 290 */
+#define Tcl_EvalEx \
+	(tclStubsPtr->tcl_EvalEx) /* 291 */
+#define Tcl_EvalObjv \
+	(tclStubsPtr->tcl_EvalObjv) /* 292 */
+#define Tcl_EvalObjEx \
+	(tclStubsPtr->tcl_EvalObjEx) /* 293 */
+#define Tcl_ExitThread \
+	(tclStubsPtr->tcl_ExitThread) /* 294 */
+#define Tcl_ExternalToUtf \
+	(tclStubsPtr->tcl_ExternalToUtf) /* 295 */
+#define Tcl_ExternalToUtfDString \
+	(tclStubsPtr->tcl_ExternalToUtfDString) /* 296 */
+#define Tcl_FinalizeThread \
+	(tclStubsPtr->tcl_FinalizeThread) /* 297 */
+#define Tcl_FinalizeNotifier \
+	(tclStubsPtr->tcl_FinalizeNotifier) /* 298 */
+#define Tcl_FreeEncoding \
+	(tclStubsPtr->tcl_FreeEncoding) /* 299 */
+#define Tcl_GetCurrentThread \
+	(tclStubsPtr->tcl_GetCurrentThread) /* 300 */
+#define Tcl_GetEncoding \
+	(tclStubsPtr->tcl_GetEncoding) /* 301 */
+#define Tcl_GetEncodingName \
+	(tclStubsPtr->tcl_GetEncodingName) /* 302 */
+#define Tcl_GetEncodingNames \
+	(tclStubsPtr->tcl_GetEncodingNames) /* 303 */
+#define Tcl_GetIndexFromObjStruct \
+	(tclStubsPtr->tcl_GetIndexFromObjStruct) /* 304 */
+#define Tcl_GetThreadData \
+	(tclStubsPtr->tcl_GetThreadData) /* 305 */
+#define Tcl_GetVar2Ex \
+	(tclStubsPtr->tcl_GetVar2Ex) /* 306 */
+#define Tcl_InitNotifier \
+	(tclStubsPtr->tcl_InitNotifier) /* 307 */
+#define Tcl_MutexLock \
+	(tclStubsPtr->tcl_MutexLock) /* 308 */
+#define Tcl_MutexUnlock \
+	(tclStubsPtr->tcl_MutexUnlock) /* 309 */
+#define Tcl_ConditionNotify \
+	(tclStubsPtr->tcl_ConditionNotify) /* 310 */
+#define Tcl_ConditionWait \
+	(tclStubsPtr->tcl_ConditionWait) /* 311 */
+#define Tcl_NumUtfChars \
+	(tclStubsPtr->tcl_NumUtfChars) /* 312 */
+#define Tcl_ReadChars \
+	(tclStubsPtr->tcl_ReadChars) /* 313 */
+#define Tcl_RestoreResult \
+	(tclStubsPtr->tcl_RestoreResult) /* 314 */
+#define Tcl_SaveResult \
+	(tclStubsPtr->tcl_SaveResult) /* 315 */
+#define Tcl_SetSystemEncoding \
+	(tclStubsPtr->tcl_SetSystemEncoding) /* 316 */
+#define Tcl_SetVar2Ex \
+	(tclStubsPtr->tcl_SetVar2Ex) /* 317 */
+#define Tcl_ThreadAlert \
+	(tclStubsPtr->tcl_ThreadAlert) /* 318 */
+#define Tcl_ThreadQueueEvent \
+	(tclStubsPtr->tcl_ThreadQueueEvent) /* 319 */
+#define Tcl_UniCharAtIndex \
+	(tclStubsPtr->tcl_UniCharAtIndex) /* 320 */
+#define Tcl_UniCharToLower \
+	(tclStubsPtr->tcl_UniCharToLower) /* 321 */
+#define Tcl_UniCharToTitle \
+	(tclStubsPtr->tcl_UniCharToTitle) /* 322 */
+#define Tcl_UniCharToUpper \
+	(tclStubsPtr->tcl_UniCharToUpper) /* 323 */
+#define Tcl_UniCharToUtf \
+	(tclStubsPtr->tcl_UniCharToUtf) /* 324 */
+#define Tcl_UtfAtIndex \
+	(tclStubsPtr->tcl_UtfAtIndex) /* 325 */
+#define Tcl_UtfCharComplete \
+	(tclStubsPtr->tcl_UtfCharComplete) /* 326 */
+#define Tcl_UtfBackslash \
+	(tclStubsPtr->tcl_UtfBackslash) /* 327 */
+#define Tcl_UtfFindFirst \
+	(tclStubsPtr->tcl_UtfFindFirst) /* 328 */
+#define Tcl_UtfFindLast \
+	(tclStubsPtr->tcl_UtfFindLast) /* 329 */
+#define Tcl_UtfNext \
+	(tclStubsPtr->tcl_UtfNext) /* 330 */
+#define Tcl_UtfPrev \
+	(tclStubsPtr->tcl_UtfPrev) /* 331 */
+#define Tcl_UtfToExternal \
+	(tclStubsPtr->tcl_UtfToExternal) /* 332 */
+#define Tcl_UtfToExternalDString \
+	(tclStubsPtr->tcl_UtfToExternalDString) /* 333 */
+#define Tcl_UtfToLower \
+	(tclStubsPtr->tcl_UtfToLower) /* 334 */
+#define Tcl_UtfToTitle \
+	(tclStubsPtr->tcl_UtfToTitle) /* 335 */
+#define Tcl_UtfToUniChar \
+	(tclStubsPtr->tcl_UtfToUniChar) /* 336 */
+#define Tcl_UtfToUpper \
+	(tclStubsPtr->tcl_UtfToUpper) /* 337 */
+#define Tcl_WriteChars \
+	(tclStubsPtr->tcl_WriteChars) /* 338 */
+#define Tcl_WriteObj \
+	(tclStubsPtr->tcl_WriteObj) /* 339 */
+#define Tcl_GetString \
+	(tclStubsPtr->tcl_GetString) /* 340 */
+#define Tcl_GetDefaultEncodingDir \
+	(tclStubsPtr->tcl_GetDefaultEncodingDir) /* 341 */
+#define Tcl_SetDefaultEncodingDir \
+	(tclStubsPtr->tcl_SetDefaultEncodingDir) /* 342 */
+#define Tcl_AlertNotifier \
+	(tclStubsPtr->tcl_AlertNotifier) /* 343 */
+#define Tcl_ServiceModeHook \
+	(tclStubsPtr->tcl_ServiceModeHook) /* 344 */
+#define Tcl_UniCharIsAlnum \
+	(tclStubsPtr->tcl_UniCharIsAlnum) /* 345 */
+#define Tcl_UniCharIsAlpha \
+	(tclStubsPtr->tcl_UniCharIsAlpha) /* 346 */
+#define Tcl_UniCharIsDigit \
+	(tclStubsPtr->tcl_UniCharIsDigit) /* 347 */
+#define Tcl_UniCharIsLower \
+	(tclStubsPtr->tcl_UniCharIsLower) /* 348 */
+#define Tcl_UniCharIsSpace \
+	(tclStubsPtr->tcl_UniCharIsSpace) /* 349 */
+#define Tcl_UniCharIsUpper \
+	(tclStubsPtr->tcl_UniCharIsUpper) /* 350 */
+#define Tcl_UniCharIsWordChar \
+	(tclStubsPtr->tcl_UniCharIsWordChar) /* 351 */
+#define Tcl_UniCharLen \
+	(tclStubsPtr->tcl_UniCharLen) /* 352 */
+#define Tcl_UniCharNcmp \
+	(tclStubsPtr->tcl_UniCharNcmp) /* 353 */
+#define Tcl_UniCharToUtfDString \
+	(tclStubsPtr->tcl_UniCharToUtfDString) /* 354 */
+#define Tcl_UtfToUniCharDString \
+	(tclStubsPtr->tcl_UtfToUniCharDString) /* 355 */
+#define Tcl_GetRegExpFromObj \
+	(tclStubsPtr->tcl_GetRegExpFromObj) /* 356 */
+#define Tcl_EvalTokens \
+	(tclStubsPtr->tcl_EvalTokens) /* 357 */
+#define Tcl_FreeParse \
+	(tclStubsPtr->tcl_FreeParse) /* 358 */
+#define Tcl_LogCommandInfo \
+	(tclStubsPtr->tcl_LogCommandInfo) /* 359 */
+#define Tcl_ParseBraces \
+	(tclStubsPtr->tcl_ParseBraces) /* 360 */
+#define Tcl_ParseCommand \
+	(tclStubsPtr->tcl_ParseCommand) /* 361 */
+#define Tcl_ParseExpr \
+	(tclStubsPtr->tcl_ParseExpr) /* 362 */
+#define Tcl_ParseQuotedString \
+	(tclStubsPtr->tcl_ParseQuotedString) /* 363 */
+#define Tcl_ParseVarName \
+	(tclStubsPtr->tcl_ParseVarName) /* 364 */
+#define Tcl_GetCwd \
+	(tclStubsPtr->tcl_GetCwd) /* 365 */
+#define Tcl_Chdir \
+	(tclStubsPtr->tcl_Chdir) /* 366 */
+#define Tcl_Access \
+	(tclStubsPtr->tcl_Access) /* 367 */
+#define Tcl_Stat \
+	(tclStubsPtr->tcl_Stat) /* 368 */
+#define Tcl_UtfNcmp \
+	(tclStubsPtr->tcl_UtfNcmp) /* 369 */
+#define Tcl_UtfNcasecmp \
+	(tclStubsPtr->tcl_UtfNcasecmp) /* 370 */
+#define Tcl_StringCaseMatch \
+	(tclStubsPtr->tcl_StringCaseMatch) /* 371 */
+#define Tcl_UniCharIsControl \
+	(tclStubsPtr->tcl_UniCharIsControl) /* 372 */
+#define Tcl_UniCharIsGraph \
+	(tclStubsPtr->tcl_UniCharIsGraph) /* 373 */
+#define Tcl_UniCharIsPrint \
+	(tclStubsPtr->tcl_UniCharIsPrint) /* 374 */
+#define Tcl_UniCharIsPunct \
+	(tclStubsPtr->tcl_UniCharIsPunct) /* 375 */
+#define Tcl_RegExpExecObj \
+	(tclStubsPtr->tcl_RegExpExecObj) /* 376 */
+#define Tcl_RegExpGetInfo \
+	(tclStubsPtr->tcl_RegExpGetInfo) /* 377 */
+#define Tcl_NewUnicodeObj \
+	(tclStubsPtr->tcl_NewUnicodeObj) /* 378 */
+#define Tcl_SetUnicodeObj \
+	(tclStubsPtr->tcl_SetUnicodeObj) /* 379 */
+#define Tcl_GetCharLength \
+	(tclStubsPtr->tcl_GetCharLength) /* 380 */
+#define Tcl_GetUniChar \
+	(tclStubsPtr->tcl_GetUniChar) /* 381 */
+#define Tcl_GetUnicode \
+	(tclStubsPtr->tcl_GetUnicode) /* 382 */
+#define Tcl_GetRange \
+	(tclStubsPtr->tcl_GetRange) /* 383 */
+#define Tcl_AppendUnicodeToObj \
+	(tclStubsPtr->tcl_AppendUnicodeToObj) /* 384 */
+#define Tcl_RegExpMatchObj \
+	(tclStubsPtr->tcl_RegExpMatchObj) /* 385 */
+#define Tcl_SetNotifier \
+	(tclStubsPtr->tcl_SetNotifier) /* 386 */
+#define Tcl_GetAllocMutex \
+	(tclStubsPtr->tcl_GetAllocMutex) /* 387 */
+#define Tcl_GetChannelNames \
+	(tclStubsPtr->tcl_GetChannelNames) /* 388 */
+#define Tcl_GetChannelNamesEx \
+	(tclStubsPtr->tcl_GetChannelNamesEx) /* 389 */
+#define Tcl_ProcObjCmd \
+	(tclStubsPtr->tcl_ProcObjCmd) /* 390 */
+#define Tcl_ConditionFinalize \
+	(tclStubsPtr->tcl_ConditionFinalize) /* 391 */
+#define Tcl_MutexFinalize \
+	(tclStubsPtr->tcl_MutexFinalize) /* 392 */
+#define Tcl_CreateThread \
+	(tclStubsPtr->tcl_CreateThread) /* 393 */
+#define Tcl_ReadRaw \
+	(tclStubsPtr->tcl_ReadRaw) /* 394 */
+#define Tcl_WriteRaw \
+	(tclStubsPtr->tcl_WriteRaw) /* 395 */
+#define Tcl_GetTopChannel \
+	(tclStubsPtr->tcl_GetTopChannel) /* 396 */
+#define Tcl_ChannelBuffered \
+	(tclStubsPtr->tcl_ChannelBuffered) /* 397 */
+#define Tcl_ChannelName \
+	(tclStubsPtr->tcl_ChannelName) /* 398 */
+#define Tcl_ChannelVersion \
+	(tclStubsPtr->tcl_ChannelVersion) /* 399 */
+#define Tcl_ChannelBlockModeProc \
+	(tclStubsPtr->tcl_ChannelBlockModeProc) /* 400 */
+#define Tcl_ChannelCloseProc \
+	(tclStubsPtr->tcl_ChannelCloseProc) /* 401 */
+#define Tcl_ChannelClose2Proc \
+	(tclStubsPtr->tcl_ChannelClose2Proc) /* 402 */
+#define Tcl_ChannelInputProc \
+	(tclStubsPtr->tcl_ChannelInputProc) /* 403 */
+#define Tcl_ChannelOutputProc \
+	(tclStubsPtr->tcl_ChannelOutputProc) /* 404 */
+#define Tcl_ChannelSeekProc \
+	(tclStubsPtr->tcl_ChannelSeekProc) /* 405 */
+#define Tcl_ChannelSetOptionProc \
+	(tclStubsPtr->tcl_ChannelSetOptionProc) /* 406 */
+#define Tcl_ChannelGetOptionProc \
+	(tclStubsPtr->tcl_ChannelGetOptionProc) /* 407 */
+#define Tcl_ChannelWatchProc \
+	(tclStubsPtr->tcl_ChannelWatchProc) /* 408 */
+#define Tcl_ChannelGetHandleProc \
+	(tclStubsPtr->tcl_ChannelGetHandleProc) /* 409 */
+#define Tcl_ChannelFlushProc \
+	(tclStubsPtr->tcl_ChannelFlushProc) /* 410 */
+#define Tcl_ChannelHandlerProc \
+	(tclStubsPtr->tcl_ChannelHandlerProc) /* 411 */
+#define Tcl_JoinThread \
+	(tclStubsPtr->tcl_JoinThread) /* 412 */
+#define Tcl_IsChannelShared \
+	(tclStubsPtr->tcl_IsChannelShared) /* 413 */
+#define Tcl_IsChannelRegistered \
+	(tclStubsPtr->tcl_IsChannelRegistered) /* 414 */
+#define Tcl_CutChannel \
+	(tclStubsPtr->tcl_CutChannel) /* 415 */
+#define Tcl_SpliceChannel \
+	(tclStubsPtr->tcl_SpliceChannel) /* 416 */
+#define Tcl_ClearChannelHandlers \
+	(tclStubsPtr->tcl_ClearChannelHandlers) /* 417 */
+#define Tcl_IsChannelExisting \
+	(tclStubsPtr->tcl_IsChannelExisting) /* 418 */
+#define Tcl_UniCharNcasecmp \
+	(tclStubsPtr->tcl_UniCharNcasecmp) /* 419 */
+#define Tcl_UniCharCaseMatch \
+	(tclStubsPtr->tcl_UniCharCaseMatch) /* 420 */
+#define Tcl_FindHashEntry \
+	(tclStubsPtr->tcl_FindHashEntry) /* 421 */
+#define Tcl_CreateHashEntry \
+	(tclStubsPtr->tcl_CreateHashEntry) /* 422 */
+#define Tcl_InitCustomHashTable \
+	(tclStubsPtr->tcl_InitCustomHashTable) /* 423 */
+#define Tcl_InitObjHashTable \
+	(tclStubsPtr->tcl_InitObjHashTable) /* 424 */
+#define Tcl_CommandTraceInfo \
+	(tclStubsPtr->tcl_CommandTraceInfo) /* 425 */
+#define Tcl_TraceCommand \
+	(tclStubsPtr->tcl_TraceCommand) /* 426 */
+#define Tcl_UntraceCommand \
+	(tclStubsPtr->tcl_UntraceCommand) /* 427 */
+#define Tcl_AttemptAlloc \
+	(tclStubsPtr->tcl_AttemptAlloc) /* 428 */
+#define Tcl_AttemptDbCkalloc \
+	(tclStubsPtr->tcl_AttemptDbCkalloc) /* 429 */
+#define Tcl_AttemptRealloc \
+	(tclStubsPtr->tcl_AttemptRealloc) /* 430 */
+#define Tcl_AttemptDbCkrealloc \
+	(tclStubsPtr->tcl_AttemptDbCkrealloc) /* 431 */
+#define Tcl_AttemptSetObjLength \
+	(tclStubsPtr->tcl_AttemptSetObjLength) /* 432 */
+#define Tcl_GetChannelThread \
+	(tclStubsPtr->tcl_GetChannelThread) /* 433 */
+#define Tcl_GetUnicodeFromObj \
+	(tclStubsPtr->tcl_GetUnicodeFromObj) /* 434 */
+#define Tcl_GetMathFuncInfo \
+	(tclStubsPtr->tcl_GetMathFuncInfo) /* 435 */
+#define Tcl_ListMathFuncs \
+	(tclStubsPtr->tcl_ListMathFuncs) /* 436 */
+#define Tcl_SubstObj \
+	(tclStubsPtr->tcl_SubstObj) /* 437 */
+#define Tcl_DetachChannel \
+	(tclStubsPtr->tcl_DetachChannel) /* 438 */
+#define Tcl_IsStandardChannel \
+	(tclStubsPtr->tcl_IsStandardChannel) /* 439 */
+#define Tcl_FSCopyFile \
+	(tclStubsPtr->tcl_FSCopyFile) /* 440 */
+#define Tcl_FSCopyDirectory \
+	(tclStubsPtr->tcl_FSCopyDirectory) /* 441 */
+#define Tcl_FSCreateDirectory \
+	(tclStubsPtr->tcl_FSCreateDirectory) /* 442 */
+#define Tcl_FSDeleteFile \
+	(tclStubsPtr->tcl_FSDeleteFile) /* 443 */
+#define Tcl_FSLoadFile \
+	(tclStubsPtr->tcl_FSLoadFile) /* 444 */
+#define Tcl_FSMatchInDirectory \
+	(tclStubsPtr->tcl_FSMatchInDirectory) /* 445 */
+#define Tcl_FSLink \
+	(tclStubsPtr->tcl_FSLink) /* 446 */
+#define Tcl_FSRemoveDirectory \
+	(tclStubsPtr->tcl_FSRemoveDirectory) /* 447 */
+#define Tcl_FSRenameFile \
+	(tclStubsPtr->tcl_FSRenameFile) /* 448 */
+#define Tcl_FSLstat \
+	(tclStubsPtr->tcl_FSLstat) /* 449 */
+#define Tcl_FSUtime \
+	(tclStubsPtr->tcl_FSUtime) /* 450 */
+#define Tcl_FSFileAttrsGet \
+	(tclStubsPtr->tcl_FSFileAttrsGet) /* 451 */
+#define Tcl_FSFileAttrsSet \
+	(tclStubsPtr->tcl_FSFileAttrsSet) /* 452 */
+#define Tcl_FSFileAttrStrings \
+	(tclStubsPtr->tcl_FSFileAttrStrings) /* 453 */
+#define Tcl_FSStat \
+	(tclStubsPtr->tcl_FSStat) /* 454 */
+#define Tcl_FSAccess \
+	(tclStubsPtr->tcl_FSAccess) /* 455 */
+#define Tcl_FSOpenFileChannel \
+	(tclStubsPtr->tcl_FSOpenFileChannel) /* 456 */
+#define Tcl_FSGetCwd \
+	(tclStubsPtr->tcl_FSGetCwd) /* 457 */
+#define Tcl_FSChdir \
+	(tclStubsPtr->tcl_FSChdir) /* 458 */
+#define Tcl_FSConvertToPathType \
+	(tclStubsPtr->tcl_FSConvertToPathType) /* 459 */
+#define Tcl_FSJoinPath \
+	(tclStubsPtr->tcl_FSJoinPath) /* 460 */
+#define Tcl_FSSplitPath \
+	(tclStubsPtr->tcl_FSSplitPath) /* 461 */
+#define Tcl_FSEqualPaths \
+	(tclStubsPtr->tcl_FSEqualPaths) /* 462 */
+#define Tcl_FSGetNormalizedPath \
+	(tclStubsPtr->tcl_FSGetNormalizedPath) /* 463 */
+#define Tcl_FSJoinToPath \
+	(tclStubsPtr->tcl_FSJoinToPath) /* 464 */
+#define Tcl_FSGetInternalRep \
+	(tclStubsPtr->tcl_FSGetInternalRep) /* 465 */
+#define Tcl_FSGetTranslatedPath \
+	(tclStubsPtr->tcl_FSGetTranslatedPath) /* 466 */
+#define Tcl_FSEvalFile \
+	(tclStubsPtr->tcl_FSEvalFile) /* 467 */
+#define Tcl_FSNewNativePath \
+	(tclStubsPtr->tcl_FSNewNativePath) /* 468 */
+#define Tcl_FSGetNativePath \
+	(tclStubsPtr->tcl_FSGetNativePath) /* 469 */
+#define Tcl_FSFileSystemInfo \
+	(tclStubsPtr->tcl_FSFileSystemInfo) /* 470 */
+#define Tcl_FSPathSeparator \
+	(tclStubsPtr->tcl_FSPathSeparator) /* 471 */
+#define Tcl_FSListVolumes \
+	(tclStubsPtr->tcl_FSListVolumes) /* 472 */
+#define Tcl_FSRegister \
+	(tclStubsPtr->tcl_FSRegister) /* 473 */
+#define Tcl_FSUnregister \
+	(tclStubsPtr->tcl_FSUnregister) /* 474 */
+#define Tcl_FSData \
+	(tclStubsPtr->tcl_FSData) /* 475 */
+#define Tcl_FSGetTranslatedStringPath \
+	(tclStubsPtr->tcl_FSGetTranslatedStringPath) /* 476 */
+#define Tcl_FSGetFileSystemForPath \
+	(tclStubsPtr->tcl_FSGetFileSystemForPath) /* 477 */
+#define Tcl_FSGetPathType \
+	(tclStubsPtr->tcl_FSGetPathType) /* 478 */
+#define Tcl_OutputBuffered \
+	(tclStubsPtr->tcl_OutputBuffered) /* 479 */
+#define Tcl_FSMountsChanged \
+	(tclStubsPtr->tcl_FSMountsChanged) /* 480 */
+#define Tcl_EvalTokensStandard \
+	(tclStubsPtr->tcl_EvalTokensStandard) /* 481 */
+#define Tcl_GetTime \
+	(tclStubsPtr->tcl_GetTime) /* 482 */
+#define Tcl_CreateObjTrace \
+	(tclStubsPtr->tcl_CreateObjTrace) /* 483 */
+#define Tcl_GetCommandInfoFromToken \
+	(tclStubsPtr->tcl_GetCommandInfoFromToken) /* 484 */
+#define Tcl_SetCommandInfoFromToken \
+	(tclStubsPtr->tcl_SetCommandInfoFromToken) /* 485 */
+#define Tcl_DbNewWideIntObj \
+	(tclStubsPtr->tcl_DbNewWideIntObj) /* 486 */
+#define Tcl_GetWideIntFromObj \
+	(tclStubsPtr->tcl_GetWideIntFromObj) /* 487 */
+#define Tcl_NewWideIntObj \
+	(tclStubsPtr->tcl_NewWideIntObj) /* 488 */
+#define Tcl_SetWideIntObj \
+	(tclStubsPtr->tcl_SetWideIntObj) /* 489 */
+#define Tcl_AllocStatBuf \
+	(tclStubsPtr->tcl_AllocStatBuf) /* 490 */
+#define Tcl_Seek \
+	(tclStubsPtr->tcl_Seek) /* 491 */
+#define Tcl_Tell \
+	(tclStubsPtr->tcl_Tell) /* 492 */
+#define Tcl_ChannelWideSeekProc \
+	(tclStubsPtr->tcl_ChannelWideSeekProc) /* 493 */
+#define Tcl_DictObjPut \
+	(tclStubsPtr->tcl_DictObjPut) /* 494 */
+#define Tcl_DictObjGet \
+	(tclStubsPtr->tcl_DictObjGet) /* 495 */
+#define Tcl_DictObjRemove \
+	(tclStubsPtr->tcl_DictObjRemove) /* 496 */
+#define Tcl_DictObjSize \
+	(tclStubsPtr->tcl_DictObjSize) /* 497 */
+#define Tcl_DictObjFirst \
+	(tclStubsPtr->tcl_DictObjFirst) /* 498 */
+#define Tcl_DictObjNext \
+	(tclStubsPtr->tcl_DictObjNext) /* 499 */
+#define Tcl_DictObjDone \
+	(tclStubsPtr->tcl_DictObjDone) /* 500 */
+#define Tcl_DictObjPutKeyList \
+	(tclStubsPtr->tcl_DictObjPutKeyList) /* 501 */
+#define Tcl_DictObjRemoveKeyList \
+	(tclStubsPtr->tcl_DictObjRemoveKeyList) /* 502 */
+#define Tcl_NewDictObj \
+	(tclStubsPtr->tcl_NewDictObj) /* 503 */
+#define Tcl_DbNewDictObj \
+	(tclStubsPtr->tcl_DbNewDictObj) /* 504 */
+#define Tcl_RegisterConfig \
+	(tclStubsPtr->tcl_RegisterConfig) /* 505 */
+#define Tcl_CreateNamespace \
+	(tclStubsPtr->tcl_CreateNamespace) /* 506 */
+#define Tcl_DeleteNamespace \
+	(tclStubsPtr->tcl_DeleteNamespace) /* 507 */
+#define Tcl_AppendExportList \
+	(tclStubsPtr->tcl_AppendExportList) /* 508 */
+#define Tcl_Export \
+	(tclStubsPtr->tcl_Export) /* 509 */
+#define Tcl_Import \
+	(tclStubsPtr->tcl_Import) /* 510 */
+#define Tcl_ForgetImport \
+	(tclStubsPtr->tcl_ForgetImport) /* 511 */
+#define Tcl_GetCurrentNamespace \
+	(tclStubsPtr->tcl_GetCurrentNamespace) /* 512 */
+#define Tcl_GetGlobalNamespace \
+	(tclStubsPtr->tcl_GetGlobalNamespace) /* 513 */
+#define Tcl_FindNamespace \
+	(tclStubsPtr->tcl_FindNamespace) /* 514 */
+#define Tcl_FindCommand \
+	(tclStubsPtr->tcl_FindCommand) /* 515 */
+#define Tcl_GetCommandFromObj \
+	(tclStubsPtr->tcl_GetCommandFromObj) /* 516 */
+#define Tcl_GetCommandFullName \
+	(tclStubsPtr->tcl_GetCommandFullName) /* 517 */
+#define Tcl_FSEvalFileEx \
+	(tclStubsPtr->tcl_FSEvalFileEx) /* 518 */
+#define Tcl_SetExitProc \
+	(tclStubsPtr->tcl_SetExitProc) /* 519 */
+#define Tcl_LimitAddHandler \
+	(tclStubsPtr->tcl_LimitAddHandler) /* 520 */
+#define Tcl_LimitRemoveHandler \
+	(tclStubsPtr->tcl_LimitRemoveHandler) /* 521 */
+#define Tcl_LimitReady \
+	(tclStubsPtr->tcl_LimitReady) /* 522 */
+#define Tcl_LimitCheck \
+	(tclStubsPtr->tcl_LimitCheck) /* 523 */
+#define Tcl_LimitExceeded \
+	(tclStubsPtr->tcl_LimitExceeded) /* 524 */
+#define Tcl_LimitSetCommands \
+	(tclStubsPtr->tcl_LimitSetCommands) /* 525 */
+#define Tcl_LimitSetTime \
+	(tclStubsPtr->tcl_LimitSetTime) /* 526 */
+#define Tcl_LimitSetGranularity \
+	(tclStubsPtr->tcl_LimitSetGranularity) /* 527 */
+#define Tcl_LimitTypeEnabled \
+	(tclStubsPtr->tcl_LimitTypeEnabled) /* 528 */
+#define Tcl_LimitTypeExceeded \
+	(tclStubsPtr->tcl_LimitTypeExceeded) /* 529 */
+#define Tcl_LimitTypeSet \
+	(tclStubsPtr->tcl_LimitTypeSet) /* 530 */
+#define Tcl_LimitTypeReset \
+	(tclStubsPtr->tcl_LimitTypeReset) /* 531 */
+#define Tcl_LimitGetCommands \
+	(tclStubsPtr->tcl_LimitGetCommands) /* 532 */
+#define Tcl_LimitGetTime \
+	(tclStubsPtr->tcl_LimitGetTime) /* 533 */
+#define Tcl_LimitGetGranularity \
+	(tclStubsPtr->tcl_LimitGetGranularity) /* 534 */
+#define Tcl_SaveInterpState \
+	(tclStubsPtr->tcl_SaveInterpState) /* 535 */
+#define Tcl_RestoreInterpState \
+	(tclStubsPtr->tcl_RestoreInterpState) /* 536 */
+#define Tcl_DiscardInterpState \
+	(tclStubsPtr->tcl_DiscardInterpState) /* 537 */
+#define Tcl_SetReturnOptions \
+	(tclStubsPtr->tcl_SetReturnOptions) /* 538 */
+#define Tcl_GetReturnOptions \
+	(tclStubsPtr->tcl_GetReturnOptions) /* 539 */
+#define Tcl_IsEnsemble \
+	(tclStubsPtr->tcl_IsEnsemble) /* 540 */
+#define Tcl_CreateEnsemble \
+	(tclStubsPtr->tcl_CreateEnsemble) /* 541 */
+#define Tcl_FindEnsemble \
+	(tclStubsPtr->tcl_FindEnsemble) /* 542 */
+#define Tcl_SetEnsembleSubcommandList \
+	(tclStubsPtr->tcl_SetEnsembleSubcommandList) /* 543 */
+#define Tcl_SetEnsembleMappingDict \
+	(tclStubsPtr->tcl_SetEnsembleMappingDict) /* 544 */
+#define Tcl_SetEnsembleUnknownHandler \
+	(tclStubsPtr->tcl_SetEnsembleUnknownHandler) /* 545 */
+#define Tcl_SetEnsembleFlags \
+	(tclStubsPtr->tcl_SetEnsembleFlags) /* 546 */
+#define Tcl_GetEnsembleSubcommandList \
+	(tclStubsPtr->tcl_GetEnsembleSubcommandList) /* 547 */
+#define Tcl_GetEnsembleMappingDict \
+	(tclStubsPtr->tcl_GetEnsembleMappingDict) /* 548 */
+#define Tcl_GetEnsembleUnknownHandler \
+	(tclStubsPtr->tcl_GetEnsembleUnknownHandler) /* 549 */
+#define Tcl_GetEnsembleFlags \
+	(tclStubsPtr->tcl_GetEnsembleFlags) /* 550 */
+#define Tcl_GetEnsembleNamespace \
+	(tclStubsPtr->tcl_GetEnsembleNamespace) /* 551 */
+#define Tcl_SetTimeProc \
+	(tclStubsPtr->tcl_SetTimeProc) /* 552 */
+#define Tcl_QueryTimeProc \
+	(tclStubsPtr->tcl_QueryTimeProc) /* 553 */
+#define Tcl_ChannelThreadActionProc \
+	(tclStubsPtr->tcl_ChannelThreadActionProc) /* 554 */
+#define Tcl_NewBignumObj \
+	(tclStubsPtr->tcl_NewBignumObj) /* 555 */
+#define Tcl_DbNewBignumObj \
+	(tclStubsPtr->tcl_DbNewBignumObj) /* 556 */
+#define Tcl_SetBignumObj \
+	(tclStubsPtr->tcl_SetBignumObj) /* 557 */
+#define Tcl_GetBignumFromObj \
+	(tclStubsPtr->tcl_GetBignumFromObj) /* 558 */
+#define Tcl_TakeBignumFromObj \
+	(tclStubsPtr->tcl_TakeBignumFromObj) /* 559 */
+#define Tcl_TruncateChannel \
+	(tclStubsPtr->tcl_TruncateChannel) /* 560 */
+#define Tcl_ChannelTruncateProc \
+	(tclStubsPtr->tcl_ChannelTruncateProc) /* 561 */
+#define Tcl_SetChannelErrorInterp \
+	(tclStubsPtr->tcl_SetChannelErrorInterp) /* 562 */
+#define Tcl_GetChannelErrorInterp \
+	(tclStubsPtr->tcl_GetChannelErrorInterp) /* 563 */
+#define Tcl_SetChannelError \
+	(tclStubsPtr->tcl_SetChannelError) /* 564 */
+#define Tcl_GetChannelError \
+	(tclStubsPtr->tcl_GetChannelError) /* 565 */
+#define Tcl_InitBignumFromDouble \
+	(tclStubsPtr->tcl_InitBignumFromDouble) /* 566 */
+#define Tcl_GetNamespaceUnknownHandler \
+	(tclStubsPtr->tcl_GetNamespaceUnknownHandler) /* 567 */
+#define Tcl_SetNamespaceUnknownHandler \
+	(tclStubsPtr->tcl_SetNamespaceUnknownHandler) /* 568 */
+#define Tcl_GetEncodingFromObj \
+	(tclStubsPtr->tcl_GetEncodingFromObj) /* 569 */
+#define Tcl_GetEncodingSearchPath \
+	(tclStubsPtr->tcl_GetEncodingSearchPath) /* 570 */
+#define Tcl_SetEncodingSearchPath \
+	(tclStubsPtr->tcl_SetEncodingSearchPath) /* 571 */
+#define Tcl_GetEncodingNameFromEnvironment \
+	(tclStubsPtr->tcl_GetEncodingNameFromEnvironment) /* 572 */
+#define Tcl_PkgRequireProc \
+	(tclStubsPtr->tcl_PkgRequireProc) /* 573 */
+#define Tcl_AppendObjToErrorInfo \
+	(tclStubsPtr->tcl_AppendObjToErrorInfo) /* 574 */
+#define Tcl_AppendLimitedToObj \
+	(tclStubsPtr->tcl_AppendLimitedToObj) /* 575 */
+#define Tcl_Format \
+	(tclStubsPtr->tcl_Format) /* 576 */
+#define Tcl_AppendFormatToObj \
+	(tclStubsPtr->tcl_AppendFormatToObj) /* 577 */
+#define Tcl_ObjPrintf \
+	(tclStubsPtr->tcl_ObjPrintf) /* 578 */
+#define Tcl_AppendPrintfToObj \
+	(tclStubsPtr->tcl_AppendPrintfToObj) /* 579 */
+#define Tcl_CancelEval \
+	(tclStubsPtr->tcl_CancelEval) /* 580 */
+#define Tcl_Canceled \
+	(tclStubsPtr->tcl_Canceled) /* 581 */
+#define Tcl_CreatePipe \
+	(tclStubsPtr->tcl_CreatePipe) /* 582 */
+#define Tcl_NRCreateCommand \
+	(tclStubsPtr->tcl_NRCreateCommand) /* 583 */
+#define Tcl_NREvalObj \
+	(tclStubsPtr->tcl_NREvalObj) /* 584 */
+#define Tcl_NREvalObjv \
+	(tclStubsPtr->tcl_NREvalObjv) /* 585 */
+#define Tcl_NRCmdSwap \
+	(tclStubsPtr->tcl_NRCmdSwap) /* 586 */
+#define Tcl_NRAddCallback \
+	(tclStubsPtr->tcl_NRAddCallback) /* 587 */
+#define Tcl_NRCallObjProc \
+	(tclStubsPtr->tcl_NRCallObjProc) /* 588 */
+#define Tcl_GetFSDeviceFromStat \
+	(tclStubsPtr->tcl_GetFSDeviceFromStat) /* 589 */
+#define Tcl_GetFSInodeFromStat \
+	(tclStubsPtr->tcl_GetFSInodeFromStat) /* 590 */
+#define Tcl_GetModeFromStat \
+	(tclStubsPtr->tcl_GetModeFromStat) /* 591 */
+#define Tcl_GetLinkCountFromStat \
+	(tclStubsPtr->tcl_GetLinkCountFromStat) /* 592 */
+#define Tcl_GetUserIdFromStat \
+	(tclStubsPtr->tcl_GetUserIdFromStat) /* 593 */
+#define Tcl_GetGroupIdFromStat \
+	(tclStubsPtr->tcl_GetGroupIdFromStat) /* 594 */
+#define Tcl_GetDeviceTypeFromStat \
+	(tclStubsPtr->tcl_GetDeviceTypeFromStat) /* 595 */
+#define Tcl_GetAccessTimeFromStat \
+	(tclStubsPtr->tcl_GetAccessTimeFromStat) /* 596 */
+#define Tcl_GetModificationTimeFromStat \
+	(tclStubsPtr->tcl_GetModificationTimeFromStat) /* 597 */
+#define Tcl_GetChangeTimeFromStat \
+	(tclStubsPtr->tcl_GetChangeTimeFromStat) /* 598 */
+#define Tcl_GetSizeFromStat \
+	(tclStubsPtr->tcl_GetSizeFromStat) /* 599 */
+#define Tcl_GetBlocksFromStat \
+	(tclStubsPtr->tcl_GetBlocksFromStat) /* 600 */
+#define Tcl_GetBlockSizeFromStat \
+	(tclStubsPtr->tcl_GetBlockSizeFromStat) /* 601 */
+#define Tcl_SetEnsembleParameterList \
+	(tclStubsPtr->tcl_SetEnsembleParameterList) /* 602 */
+#define Tcl_GetEnsembleParameterList \
+	(tclStubsPtr->tcl_GetEnsembleParameterList) /* 603 */
+#define Tcl_ParseArgsObjv \
+	(tclStubsPtr->tcl_ParseArgsObjv) /* 604 */
+#define Tcl_GetErrorLine \
+	(tclStubsPtr->tcl_GetErrorLine) /* 605 */
+#define Tcl_SetErrorLine \
+	(tclStubsPtr->tcl_SetErrorLine) /* 606 */
+#define Tcl_TransferResult \
+	(tclStubsPtr->tcl_TransferResult) /* 607 */
+#define Tcl_InterpActive \
+	(tclStubsPtr->tcl_InterpActive) /* 608 */
+#define Tcl_BackgroundException \
+	(tclStubsPtr->tcl_BackgroundException) /* 609 */
+#define Tcl_ZlibDeflate \
+	(tclStubsPtr->tcl_ZlibDeflate) /* 610 */
+#define Tcl_ZlibInflate \
+	(tclStubsPtr->tcl_ZlibInflate) /* 611 */
+#define Tcl_ZlibCRC32 \
+	(tclStubsPtr->tcl_ZlibCRC32) /* 612 */
+#define Tcl_ZlibAdler32 \
+	(tclStubsPtr->tcl_ZlibAdler32) /* 613 */
+#define Tcl_ZlibStreamInit \
+	(tclStubsPtr->tcl_ZlibStreamInit) /* 614 */
+#define Tcl_ZlibStreamGetCommandName \
+	(tclStubsPtr->tcl_ZlibStreamGetCommandName) /* 615 */
+#define Tcl_ZlibStreamEof \
+	(tclStubsPtr->tcl_ZlibStreamEof) /* 616 */
+#define Tcl_ZlibStreamChecksum \
+	(tclStubsPtr->tcl_ZlibStreamChecksum) /* 617 */
+#define Tcl_ZlibStreamPut \
+	(tclStubsPtr->tcl_ZlibStreamPut) /* 618 */
+#define Tcl_ZlibStreamGet \
+	(tclStubsPtr->tcl_ZlibStreamGet) /* 619 */
+#define Tcl_ZlibStreamClose \
+	(tclStubsPtr->tcl_ZlibStreamClose) /* 620 */
+#define Tcl_ZlibStreamReset \
+	(tclStubsPtr->tcl_ZlibStreamReset) /* 621 */
+#define Tcl_SetStartupScript \
+	(tclStubsPtr->tcl_SetStartupScript) /* 622 */
+#define Tcl_GetStartupScript \
+	(tclStubsPtr->tcl_GetStartupScript) /* 623 */
+#define Tcl_CloseEx \
+	(tclStubsPtr->tcl_CloseEx) /* 624 */
+#define Tcl_NRExprObj \
+	(tclStubsPtr->tcl_NRExprObj) /* 625 */
+#define Tcl_NRSubstObj \
+	(tclStubsPtr->tcl_NRSubstObj) /* 626 */
+#define Tcl_LoadFile \
+	(tclStubsPtr->tcl_LoadFile) /* 627 */
+#define Tcl_FindSymbol \
+	(tclStubsPtr->tcl_FindSymbol) /* 628 */
+#define Tcl_FSUnloadFile \
+	(tclStubsPtr->tcl_FSUnloadFile) /* 629 */
+#define Tcl_ZlibStreamSetCompressionDictionary \
+	(tclStubsPtr->tcl_ZlibStreamSetCompressionDictionary) /* 630 */
+#define Tcl_OpenTcpServerEx \
+	(tclStubsPtr->tcl_OpenTcpServerEx) /* 631 */
+
+#endif /* defined(USE_TCL_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#if defined(USE_TCL_STUBS)
+#   undef Tcl_CreateInterp
+#   undef Tcl_FindExecutable
+#   undef Tcl_GetStringResult
+#   undef Tcl_Init
+#   undef Tcl_SetPanicProc
+#   undef Tcl_SetVar
+#   undef Tcl_ObjSetVar2
+#   undef Tcl_StaticPackage
+#   define Tcl_CreateInterp() (tclStubsPtr->tcl_CreateInterp())
+#   define Tcl_GetStringResult(interp) (tclStubsPtr->tcl_GetStringResult(interp))
+#   define Tcl_Init(interp) (tclStubsPtr->tcl_Init(interp))
+#   define Tcl_SetPanicProc(proc) (tclStubsPtr->tcl_SetPanicProc(proc))
+#   define Tcl_SetVar(interp, varName, newValue, flags) \
+	    (tclStubsPtr->tcl_SetVar(interp, varName, newValue, flags))
+#   define Tcl_ObjSetVar2(interp, part1, part2, newValue, flags) \
+	    (tclStubsPtr->tcl_ObjSetVar2(interp, part1, part2, newValue, flags))
+#endif
+
+#if defined(_WIN32) && defined(UNICODE)
+#   define Tcl_FindExecutable(arg) ((Tcl_FindExecutable)((const char *)(arg)))
+#   define Tcl_MainEx Tcl_MainExW
+    EXTERN void Tcl_MainExW(int argc, wchar_t **argv,
+	    Tcl_AppInitProc *appInitProc, Tcl_Interp *interp);
+#endif
+
+#undef TCL_STORAGE_CLASS
+#define TCL_STORAGE_CLASS DLLIMPORT
+
+#undef Tcl_SeekOld
+#undef Tcl_TellOld
+
+#undef Tcl_PkgPresent
+#define Tcl_PkgPresent(interp, name, version, exact) \
+	Tcl_PkgPresentEx(interp, name, version, exact, NULL)
+#undef Tcl_PkgProvide
+#define Tcl_PkgProvide(interp, name, version) \
+	Tcl_PkgProvideEx(interp, name, version, NULL)
+#undef Tcl_PkgRequire
+#define Tcl_PkgRequire(interp, name, version, exact) \
+	Tcl_PkgRequireEx(interp, name, version, exact, NULL)
+#undef Tcl_GetIndexFromObj
+#define Tcl_GetIndexFromObj(interp, objPtr, tablePtr, msg, flags, indexPtr) \
+	Tcl_GetIndexFromObjStruct(interp, objPtr, tablePtr, \
+	sizeof(char *), msg, flags, indexPtr)
+#undef Tcl_NewBooleanObj
+#define Tcl_NewBooleanObj(boolValue) \
+	Tcl_NewLongObj((boolValue)!=0)
+#undef Tcl_DbNewBooleanObj
+#define Tcl_DbNewBooleanObj(boolValue, file, line) \
+	Tcl_DbNewLongObj((boolValue)!=0, file, line)
+#undef Tcl_SetBooleanObj
+#define Tcl_SetBooleanObj(objPtr, boolValue) \
+	Tcl_SetLongObj(objPtr, (boolValue)!=0)
+#undef Tcl_SetVar
+#define Tcl_SetVar(interp, varName, newValue, flags) \
+	Tcl_SetVar2(interp, varName, NULL, newValue, flags)
+#undef Tcl_UnsetVar
+#define Tcl_UnsetVar(interp, varName, flags) \
+	Tcl_UnsetVar2(interp, varName, NULL, flags)
+#undef Tcl_GetVar
+#define Tcl_GetVar(interp, varName, flags) \
+	Tcl_GetVar2(interp, varName, NULL, flags)
+#undef Tcl_TraceVar
+#define Tcl_TraceVar(interp, varName, flags, proc, clientData) \
+	Tcl_TraceVar2(interp, varName, NULL, flags, proc, clientData)
+#undef Tcl_UntraceVar
+#define Tcl_UntraceVar(interp, varName, flags, proc, clientData) \
+	Tcl_UntraceVar2(interp, varName, NULL, flags, proc, clientData)
+#undef Tcl_VarTraceInfo
+#define Tcl_VarTraceInfo(interp, varName, flags, proc, prevClientData) \
+	Tcl_VarTraceInfo2(interp, varName, NULL, flags, proc, prevClientData)
+#undef Tcl_UpVar
+#define Tcl_UpVar(interp, frameName, varName, localName, flags) \
+	Tcl_UpVar2(interp, frameName, varName, NULL, localName, flags)
+#undef Tcl_AddErrorInfo
+#define Tcl_AddErrorInfo(interp, message) \
+	Tcl_AppendObjToErrorInfo(interp, Tcl_NewStringObj(message, -1))
+#undef Tcl_AddObjErrorInfo
+#define Tcl_AddObjErrorInfo(interp, message, length) \
+	Tcl_AppendObjToErrorInfo(interp, Tcl_NewStringObj(message, length))
+#ifdef TCL_NO_DEPRECATED
+#undef Tcl_Eval
+#define Tcl_Eval(interp, objPtr) \
+	Tcl_EvalEx(interp, objPtr, -1, 0)
+#undef Tcl_GlobalEval
+#define Tcl_GlobalEval(interp, objPtr) \
+	Tcl_EvalEx(interp, objPtr, -1, TCL_EVAL_GLOBAL)
+#undef Tcl_SaveResult
+#define Tcl_SaveResult(interp, statePtr) \
+	do { \
+	    (statePtr)->objResultPtr = Tcl_GetObjResult(interp); \
+	    Tcl_IncrRefCount((statePtr)->objResultPtr); \
+	    Tcl_SetObjResult(interp, Tcl_NewObj()); \
+	} while(0)
+#undef Tcl_RestoreResult
+#define Tcl_RestoreResult(interp, statePtr) \
+	do { \
+	    Tcl_ResetResult(interp); \
+   	    Tcl_SetObjResult(interp, (statePtr)->objResultPtr); \
+   	    Tcl_DecrRefCount((statePtr)->objResultPtr); \
+	} while(0)
+#undef Tcl_DiscardResult
+#define Tcl_DiscardResult(statePtr) \
+	Tcl_DecrRefCount((statePtr)->objResultPtr)
+#undef Tcl_SetResult
+#define Tcl_SetResult(interp, result, freeProc) \
+	do { \
+	    char *__result = result; \
+	    Tcl_FreeProc *__freeProc = freeProc; \
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(__result, -1)); \
+	    if (__result != NULL && __freeProc != NULL && __freeProc != TCL_VOLATILE) { \
+		if (__freeProc == TCL_DYNAMIC) { \
+		    ckfree(__result); \
+		} else { \
+		    (*__freeProc)(__result); \
+		} \
+	    } \
+	} while(0)
+#endif /* TCL_NO_DEPRECATED */
+
+#if defined(USE_TCL_STUBS) && !defined(USE_TCL_STUB_PROCS)
+#   if defined(__CYGWIN__) && defined(TCL_WIDE_INT_IS_LONG)
+/* On Cygwin64, long is 64-bit while on Win64 long is 32-bit. Therefore
+ * we have to make sure that all stub entries on Cygwin64 follow the
+ * Win64 signature. Cygwin64 stubbed extensions cannot use those stub
+ * entries any more, they should use the 64-bit alternatives where
+ * possible. Tcl 9 must find a better solution, but that cannot be done
+ * without introducing a binary incompatibility.
+ */
+#	undef Tcl_DbNewLongObj
+#	undef Tcl_GetLongFromObj
+#	undef Tcl_NewLongObj
+#	undef Tcl_SetLongObj
+#	undef Tcl_ExprLong
+#	undef Tcl_ExprLongObj
+#	undef Tcl_UniCharNcmp
+#	undef Tcl_UtfNcmp
+#	undef Tcl_UtfNcasecmp
+#	undef Tcl_UniCharNcasecmp
+#	define Tcl_DbNewLongObj ((Tcl_Obj*(*)(long,const char*,int))Tcl_DbNewWideIntObj)
+#	define Tcl_GetLongFromObj ((int(*)(Tcl_Interp*,Tcl_Obj*,long*))Tcl_GetWideIntFromObj)
+#	define Tcl_NewLongObj ((Tcl_Obj*(*)(long))Tcl_NewWideIntObj)
+#	define Tcl_SetLongObj ((void(*)(Tcl_Obj*,long))Tcl_SetWideIntObj)
+#	define Tcl_ExprLong TclExprLong
+	static inline int TclExprLong(Tcl_Interp *interp, const char *string, long *ptr){
+	    int intValue;
+	    int result = tclStubsPtr->tcl_ExprLong(interp, string, (long *)&intValue);
+	    if (result == TCL_OK) *ptr = (long)intValue;
+	    return result;
+	}
+#	define Tcl_ExprLongObj TclExprLongObj
+	static inline int TclExprLongObj(Tcl_Interp *interp, Tcl_Obj *obj, long *ptr){
+	    int intValue;
+	    int result = tclStubsPtr->tcl_ExprLongObj(interp, obj, (long *)&intValue);
+	    if (result == TCL_OK) *ptr = (long)intValue;
+	    return result;
+	}
+#	define Tcl_UniCharNcmp(ucs,uct,n) \
+		((int(*)(const Tcl_UniChar*,const Tcl_UniChar*,unsigned int))tclStubsPtr->tcl_UniCharNcmp)(ucs,uct,(unsigned int)(n))
+#	define Tcl_UtfNcmp(s1,s2,n) \
+		((int(*)(const char*,const char*,unsigned int))tclStubsPtr->tcl_UtfNcmp)(s1,s2,(unsigned int)(n))
+#	define Tcl_UtfNcasecmp(s1,s2,n) \
+		((int(*)(const char*,const char*,unsigned int))tclStubsPtr->tcl_UtfNcasecmp)(s1,s2,(unsigned int)(n))
+#	define Tcl_UniCharNcasecmp(ucs,uct,n) \
+		((int(*)(const Tcl_UniChar*,const Tcl_UniChar*,unsigned int))tclStubsPtr->tcl_UniCharNcasecmp)(ucs,uct,(unsigned int)(n))
+#   endif
+#endif
+
+/*
+ * Deprecated Tcl procedures:
+ */
+
+#undef Tcl_EvalObj
+#define Tcl_EvalObj(interp, objPtr) \
+    Tcl_EvalObjEx(interp, objPtr, 0)
+#undef Tcl_GlobalEvalObj
+#define Tcl_GlobalEvalObj(interp, objPtr) \
+    Tcl_EvalObjEx(interp, objPtr, TCL_EVAL_GLOBAL)
+
+#endif /* _TCLDECLS */
diff --git a/generic/tclDictObj.c b/generic/tclDictObj.c
new file mode 100644
index 0000000..c82f88a
--- /dev/null
+++ b/generic/tclDictObj.c
@@ -0,0 +1,3689 @@
+/*
+ * tclDictObj.c --
+ *
+ *	This file contains functions that implement the Tcl dict object type
+ *	and its accessor command.
+ *
+ * Copyright (c) 2002-2010 by Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+
+/*
+ * Forward declaration.
+ */
+struct Dict;
+
+/*
+ * Prototypes for functions defined later in this file:
+ */
+
+static void		DeleteDict(struct Dict *dict);
+static int		DictAppendCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictCreateCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictExistsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictFilterCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictGetCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictIncrCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictInfoCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictKeysCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictLappendCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictMergeCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictRemoveCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictReplaceCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictSetCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictSizeCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictUnsetCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictUpdateCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictValuesCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static int		DictWithCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const *objv);
+static void		DupDictInternalRep(Tcl_Obj *srcPtr, Tcl_Obj *copyPtr);
+static void		FreeDictInternalRep(Tcl_Obj *dictPtr);
+static void		InvalidateDictChain(Tcl_Obj *dictObj);
+static int		SetDictFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		UpdateStringOfDict(Tcl_Obj *dictPtr);
+static Tcl_HashEntry *	AllocChainEntry(Tcl_HashTable *tablePtr,void *keyPtr);
+static inline void	InitChainTable(struct Dict *dict);
+static inline void	DeleteChainTable(struct Dict *dict);
+static inline Tcl_HashEntry *CreateChainEntry(struct Dict *dict,
+			    Tcl_Obj *keyPtr, int *newPtr);
+static inline int	DeleteChainEntry(struct Dict *dict, Tcl_Obj *keyPtr);
+static Tcl_NRPostProc	FinalizeDictUpdate;
+static Tcl_NRPostProc	FinalizeDictWith;
+static Tcl_ObjCmdProc	DictForNRCmd;
+static Tcl_ObjCmdProc	DictMapNRCmd;
+static Tcl_NRPostProc	DictForLoopCallback;
+static Tcl_NRPostProc	DictMapLoopCallback;
+
+/*
+ * Table of dict subcommand names and implementations.
+ */
+
+static const EnsembleImplMap implementationMap[] = {
+    {"append",	DictAppendCmd,	TclCompileDictAppendCmd, NULL, NULL, 0 },
+    {"create",	DictCreateCmd,	TclCompileDictCreateCmd, NULL, NULL, 0 },
+    {"exists",	DictExistsCmd,	TclCompileDictExistsCmd, NULL, NULL, 0 },
+    {"filter",	DictFilterCmd,	NULL, NULL, NULL, 0 },
+    {"for",	NULL,		TclCompileDictForCmd, DictForNRCmd, NULL, 0 },
+    {"get",	DictGetCmd,	TclCompileDictGetCmd, NULL, NULL, 0 },
+    {"incr",	DictIncrCmd,	TclCompileDictIncrCmd, NULL, NULL, 0 },
+    {"info",	DictInfoCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0 },
+    {"keys",	DictKeysCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0 },
+    {"lappend",	DictLappendCmd,	TclCompileDictLappendCmd, NULL, NULL, 0 },
+    {"map", 	NULL,       	TclCompileDictMapCmd, DictMapNRCmd, NULL, 0 },
+    {"merge",	DictMergeCmd,	TclCompileDictMergeCmd, NULL, NULL, 0 },
+    {"remove",	DictRemoveCmd,	TclCompileBasicMin1ArgCmd, NULL, NULL, 0 },
+    {"replace",	DictReplaceCmd, NULL, NULL, NULL, 0 },
+    {"set",	DictSetCmd,	TclCompileDictSetCmd, NULL, NULL, 0 },
+    {"size",	DictSizeCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0 },
+    {"unset",	DictUnsetCmd,	TclCompileDictUnsetCmd, NULL, NULL, 0 },
+    {"update",	DictUpdateCmd,	TclCompileDictUpdateCmd, NULL, NULL, 0 },
+    {"values",	DictValuesCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0 },
+    {"with",	DictWithCmd,	TclCompileDictWithCmd, NULL, NULL, 0 },
+    {NULL, NULL, NULL, NULL, NULL, 0}
+};
+
+/*
+ * Internal representation of the entries in the hash table that backs a
+ * dictionary.
+ */
+
+typedef struct ChainEntry {
+    Tcl_HashEntry entry;
+    struct ChainEntry *prevPtr;
+    struct ChainEntry *nextPtr;
+} ChainEntry;
+
+/*
+ * Internal representation of a dictionary.
+ *
+ * The internal representation of a dictionary object is a hash table (with
+ * Tcl_Objs for both keys and values), a reference count and epoch number for
+ * detecting concurrent modifications of the dictionary, and a pointer to the
+ * parent object (used when invalidating string reps of pathed dictionary
+ * trees) which is NULL in normal use. The fact that hash tables know (with
+ * appropriate initialisation) already about objects makes key management /so/
+ * much easier!
+ *
+ * Reference counts are used to enable safe iteration across hashes while
+ * allowing the type of the containing object to be modified.
+ */
+
+typedef struct Dict {
+    Tcl_HashTable table;	/* Object hash table to store mapping in. */
+    ChainEntry *entryChainHead;	/* Linked list of all entries in the
+				 * dictionary. Used for doing traversal of the
+				 * entries in the order that they are
+				 * created. */
+    ChainEntry *entryChainTail;	/* Other end of linked list of all entries in
+				 * the dictionary. Used for doing traversal of
+				 * the entries in the order that they are
+				 * created. */
+    int epoch;			/* Epoch counter */
+    size_t refCount;		/* Reference counter (see above) */
+    Tcl_Obj *chain;		/* Linked list used for invalidating the
+				 * string representations of updated nested
+				 * dictionaries. */
+} Dict;
+
+/*
+ * Accessor macro for converting between a Tcl_Obj* and a Dict. Note that this
+ * must be assignable as well as readable.
+ */
+
+#define DICT(dictObj)   (*((Dict **)&(dictObj)->internalRep.twoPtrValue.ptr1))
+
+/*
+ * The structure below defines the dictionary object type by means of
+ * functions that can be invoked by generic object code.
+ */
+
+const Tcl_ObjType tclDictType = {
+    "dict",
+    FreeDictInternalRep,		/* freeIntRepProc */
+    DupDictInternalRep,			/* dupIntRepProc */
+    UpdateStringOfDict,			/* updateStringProc */
+    SetDictFromAny			/* setFromAnyProc */
+};
+
+/*
+ * The type of the specially adapted version of the Tcl_Obj*-containing hash
+ * table defined in the tclObj.c code. This version differs in that it
+ * allocates a bit more space in each hash entry in order to hold the pointers
+ * used to keep the hash entries in a linked list.
+ *
+ * Note that this type of hash table is *only* suitable for direct use in
+ * *this* file. Everything else should use the dict iterator API.
+ */
+
+static const Tcl_HashKeyType chainHashType = {
+    TCL_HASH_KEY_TYPE_VERSION,
+    0,
+    TclHashObjKey,
+    TclCompareObjKeys,
+    AllocChainEntry,
+    TclFreeObjEntry
+};
+
+/*
+ * Structure used in implementation of 'dict map' to hold the state that gets
+ * passed between parts of the implementation.
+ */
+
+typedef struct {
+    Tcl_Obj *keyVarObj;		/* The name of the variable that will have
+				 * keys assigned to it. */
+    Tcl_Obj *valueVarObj;	/* The name of the variable that will have
+				 * values assigned to it. */
+    Tcl_DictSearch search;	/* The dictionary search structure. */
+    Tcl_Obj *scriptObj;		/* The script to evaluate each time through
+				 * the loop. */
+    Tcl_Obj *accumulatorObj;	/* The dictionary used to accumulate the
+				 * results. */
+} DictMapStorage;
+
+/***** START OF FUNCTIONS IMPLEMENTING DICT CORE API *****/
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AllocChainEntry --
+ *
+ *	Allocate space for a Tcl_HashEntry containing the Tcl_Obj * key, and
+ *	which has a bit of extra space afterwards for storing pointers to the
+ *	rest of the chain of entries (the extra pointers are left NULL).
+ *
+ * Results:
+ *	The return value is a pointer to the created entry.
+ *
+ * Side effects:
+ *	Increments the reference count on the object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashEntry *
+AllocChainEntry(
+    Tcl_HashTable *tablePtr,
+    void *keyPtr)
+{
+    Tcl_Obj *objPtr = keyPtr;
+    ChainEntry *cPtr;
+
+    cPtr = ckalloc(sizeof(ChainEntry));
+    cPtr->entry.key.objPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);
+    cPtr->entry.clientData = NULL;
+    cPtr->prevPtr = cPtr->nextPtr = NULL;
+
+    return &cPtr->entry;
+}
+
+/*
+ * Helper functions that disguise most of the details relating to how the
+ * linked list of hash entries is managed. In particular, these manage the
+ * creation of the table and initializing of the chain, the deletion of the
+ * table and chain, the adding of an entry to the chain, and the removal of an
+ * entry from the chain.
+ */
+
+static inline void
+InitChainTable(
+    Dict *dict)
+{
+    Tcl_InitCustomHashTable(&dict->table, TCL_CUSTOM_PTR_KEYS,
+	    &chainHashType);
+    dict->entryChainHead = dict->entryChainTail = NULL;
+}
+
+static inline void
+DeleteChainTable(
+    Dict *dict)
+{
+    ChainEntry *cPtr;
+
+    for (cPtr=dict->entryChainHead ; cPtr!=NULL ; cPtr=cPtr->nextPtr) {
+	Tcl_Obj *valuePtr = Tcl_GetHashValue(&cPtr->entry);
+
+	TclDecrRefCount(valuePtr);
+    }
+    Tcl_DeleteHashTable(&dict->table);
+}
+
+static inline Tcl_HashEntry *
+CreateChainEntry(
+    Dict *dict,
+    Tcl_Obj *keyPtr,
+    int *newPtr)
+{
+    ChainEntry *cPtr = (ChainEntry *)
+	    Tcl_CreateHashEntry(&dict->table, keyPtr, newPtr);
+
+    /*
+     * If this is a new entry in the hash table, stitch it into the chain.
+     */
+
+    if (*newPtr) {
+	cPtr->nextPtr = NULL;
+	if (dict->entryChainHead == NULL) {
+	    cPtr->prevPtr = NULL;
+	    dict->entryChainHead = cPtr;
+	    dict->entryChainTail = cPtr;
+	} else {
+	    cPtr->prevPtr = dict->entryChainTail;
+	    dict->entryChainTail->nextPtr = cPtr;
+	    dict->entryChainTail = cPtr;
+	}
+    }
+
+    return &cPtr->entry;
+}
+
+static inline int
+DeleteChainEntry(
+    Dict *dict,
+    Tcl_Obj *keyPtr)
+{
+    ChainEntry *cPtr = (ChainEntry *)
+	    Tcl_FindHashEntry(&dict->table, keyPtr);
+
+    if (cPtr == NULL) {
+	return 0;
+    } else {
+	Tcl_Obj *valuePtr = Tcl_GetHashValue(&cPtr->entry);
+
+	TclDecrRefCount(valuePtr);
+    }
+
+    /*
+     * Unstitch from the chain.
+     */
+
+    if (cPtr->nextPtr) {
+	cPtr->nextPtr->prevPtr = cPtr->prevPtr;
+    } else {
+	dict->entryChainTail = cPtr->prevPtr;
+    }
+    if (cPtr->prevPtr) {
+	cPtr->prevPtr->nextPtr = cPtr->nextPtr;
+    } else {
+	dict->entryChainHead = cPtr->nextPtr;
+    }
+
+    Tcl_DeleteHashEntry(&cPtr->entry);
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupDictInternalRep --
+ *
+ *	Initialize the internal representation of a dictionary Tcl_Obj to a
+ *	copy of the internal representation of an existing dictionary object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	"srcPtr"s dictionary internal rep pointer should not be NULL and we
+ *	assume it is not NULL. We set "copyPtr"s internal rep to a pointer to
+ *	a newly allocated dictionary rep that, in turn, points to "srcPtr"s
+ *	key and value objects. Those objects are not actually copied but are
+ *	shared between "srcPtr" and "copyPtr". The ref count of each key and
+ *	value object is incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupDictInternalRep(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *copyPtr)
+{
+    Dict *oldDict = DICT(srcPtr);
+    Dict *newDict = ckalloc(sizeof(Dict));
+    ChainEntry *cPtr;
+
+    /*
+     * Copy values across from the old hash table.
+     */
+
+    InitChainTable(newDict);
+    for (cPtr=oldDict->entryChainHead ; cPtr!=NULL ; cPtr=cPtr->nextPtr) {
+	Tcl_Obj *key = Tcl_GetHashKey(&oldDict->table, &cPtr->entry);
+	Tcl_Obj *valuePtr = Tcl_GetHashValue(&cPtr->entry);
+	int n;
+	Tcl_HashEntry *hPtr = CreateChainEntry(newDict, key, &n);
+
+	/*
+	 * Fill in the contents.
+	 */
+
+	Tcl_SetHashValue(hPtr, valuePtr);
+	Tcl_IncrRefCount(valuePtr);
+    }
+
+    /*
+     * Initialise other fields.
+     */
+
+    newDict->epoch = 0;
+    newDict->chain = NULL;
+    newDict->refCount = 1;
+
+    /*
+     * Store in the object.
+     */
+
+    DICT(copyPtr) = newDict;
+    copyPtr->internalRep.twoPtrValue.ptr2 = NULL;
+    copyPtr->typePtr = &tclDictType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeDictInternalRep --
+ *
+ *	Deallocate the storage associated with a dictionary object's internal
+ *	representation.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Frees the memory holding the dictionary's internal hash table unless
+ *	it is locked by an iteration going over it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeDictInternalRep(
+    Tcl_Obj *dictPtr)
+{
+    Dict *dict = DICT(dictPtr);
+
+    if (dict->refCount-- <= 1) {
+	DeleteDict(dict);
+    }
+    dictPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteDict --
+ *
+ *	Delete the structure that is used to implement a dictionary's internal
+ *	representation. Called when either the dictionary object loses its
+ *	internal representation or when the last iteration over the dictionary
+ *	completes.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Decrements the reference count of all key and value objects in the
+ *	dictionary, which may free them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteDict(
+    Dict *dict)
+{
+    DeleteChainTable(dict);
+    ckfree(dict);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfDict --
+ *
+ *	Update the string representation for a dictionary object. Note: This
+ *	function does not invalidate an existing old string rep so storage
+ *	will be lost if this has not already been done. This code is based on
+ *	UpdateStringOfList in tclListObj.c
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	dict-to-string conversion. This string will be empty if the dictionary
+ *	has no key/value pairs. The dictionary internal representation should
+ *	not be NULL and we assume it is not NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfDict(
+    Tcl_Obj *dictPtr)
+{
+#define LOCAL_SIZE 20
+    int localFlags[LOCAL_SIZE], *flagPtr = NULL;
+    Dict *dict = DICT(dictPtr);
+    ChainEntry *cPtr;
+    Tcl_Obj *keyPtr, *valuePtr;
+    int i, length, bytesNeeded = 0;
+    const char *elem;
+    char *dst;
+    const int maxFlags = UINT_MAX / sizeof(int);
+
+    /*
+     * This field is the most useful one in the whole hash structure, and it
+     * is not exposed by any API function...
+     */
+
+    int numElems = dict->table.numEntries * 2;
+
+    /* Handle empty list case first, simplifies what follows */
+    if (numElems == 0) {
+	dictPtr->bytes = &tclEmptyString;
+	dictPtr->length = 0;
+	return;
+    }
+
+    /*
+     * Pass 1: estimate space, gather flags.
+     */
+
+    if (numElems <= LOCAL_SIZE) {
+	flagPtr = localFlags;
+    } else if (numElems > maxFlags) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    } else {
+	flagPtr = ckalloc(numElems * sizeof(int));
+    }
+    for (i=0,cPtr=dict->entryChainHead; i<numElems; i+=2,cPtr=cPtr->nextPtr) {
+	/*
+	 * Assume that cPtr is never NULL since we know the number of array
+	 * elements already.
+	 */
+
+	flagPtr[i] = ( i ? TCL_DONT_QUOTE_HASH : 0 );
+	keyPtr = Tcl_GetHashKey(&dict->table, &cPtr->entry);
+	elem = TclGetStringFromObj(keyPtr, &length);
+	bytesNeeded += TclScanElement(elem, length, flagPtr+i);
+	if (bytesNeeded < 0) {
+	    Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+	}
+
+	flagPtr[i+1] = TCL_DONT_QUOTE_HASH;
+	valuePtr = Tcl_GetHashValue(&cPtr->entry);
+	elem = TclGetStringFromObj(valuePtr, &length);
+	bytesNeeded += TclScanElement(elem, length, flagPtr+i+1);
+	if (bytesNeeded < 0) {
+	    Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+	}
+    }
+    if (bytesNeeded > INT_MAX - numElems + 1) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+    bytesNeeded += numElems;
+
+    /*
+     * Pass 2: copy into string rep buffer.
+     */
+
+    dictPtr->length = bytesNeeded - 1;
+    dictPtr->bytes = ckalloc(bytesNeeded);
+    dst = dictPtr->bytes;
+    for (i=0,cPtr=dict->entryChainHead; i<numElems; i+=2,cPtr=cPtr->nextPtr) {
+	flagPtr[i] |= ( i ? TCL_DONT_QUOTE_HASH : 0 );
+	keyPtr = Tcl_GetHashKey(&dict->table, &cPtr->entry);
+	elem = TclGetStringFromObj(keyPtr, &length);
+	dst += TclConvertElement(elem, length, dst, flagPtr[i]);
+	*dst++ = ' ';
+
+	flagPtr[i+1] |= TCL_DONT_QUOTE_HASH;
+	valuePtr = Tcl_GetHashValue(&cPtr->entry);
+	elem = TclGetStringFromObj(valuePtr, &length);
+	dst += TclConvertElement(elem, length, dst, flagPtr[i+1]);
+	*dst++ = ' ';
+    }
+    dictPtr->bytes[dictPtr->length] = '\0';
+
+    if (flagPtr != localFlags) {
+	ckfree(flagPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetDictFromAny --
+ *
+ *	Convert a non-dictionary object into a dictionary object. This code is
+ *	very closely related to SetListFromAny in tclListObj.c but does not
+ *	actually guarantee that a dictionary object will have a string rep (as
+ *	conversions from lists are handled with a special case.)
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	If the string can be converted, it loses any old internal
+ *	representation that it had and gains a dictionary's internalRep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetDictFromAny(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)
+{
+    Tcl_HashEntry *hPtr;
+    int isNew;
+    Dict *dict = ckalloc(sizeof(Dict));
+
+    InitChainTable(dict);
+
+    /*
+     * Since lists and dictionaries have very closely-related string
+     * representations (i.e. the same parsing code) we can safely special-case
+     * the conversion from lists to dictionaries.
+     */
+
+    if (objPtr->typePtr == &tclListType) {
+	int objc, i;
+	Tcl_Obj **objv;
+
+	/* Cannot fail, we already know the Tcl_ObjType is "list". */
+	TclListObjGetElements(NULL, objPtr, &objc, &objv);
+	if (objc & 1) {
+	    goto missingValue;
+	}
+
+	for (i=0 ; i<objc ; i+=2) {
+
+	    /* Store key and value in the hash table we're building. */
+	    hPtr = CreateChainEntry(dict, objv[i], &isNew);
+	    if (!isNew) {
+		Tcl_Obj *discardedValue = Tcl_GetHashValue(hPtr);
+
+		/*
+		 * Not really a well-formed dictionary as there are duplicate
+		 * keys, so better get the string rep here so that we can
+		 * convert back.
+		 */
+
+		(void) Tcl_GetString(objPtr);
+
+		TclDecrRefCount(discardedValue);
+	    }
+	    Tcl_SetHashValue(hPtr, objv[i+1]);
+	    Tcl_IncrRefCount(objv[i+1]); /* Since hash now holds ref to it */
+	}
+    } else {
+	int length;
+	const char *nextElem = TclGetStringFromObj(objPtr, &length);
+	const char *limit = (nextElem + length);
+
+	while (nextElem < limit) {
+	    Tcl_Obj *keyPtr, *valuePtr;
+	    const char *elemStart;
+	    int elemSize, literal;
+
+	    if (TclFindDictElement(interp, nextElem, (limit - nextElem),
+		    &elemStart, &nextElem, &elemSize, &literal) != TCL_OK) {
+		goto errorInFindDictElement;
+	    }
+	    if (elemStart == limit) {
+		break;
+	    }
+	    if (nextElem == limit) {
+		goto missingValue;
+	    }
+
+	    if (literal) {
+		TclNewStringObj(keyPtr, elemStart, elemSize);
+	    } else {
+		/* Avoid double copy */
+		TclNewObj(keyPtr);
+		keyPtr->bytes = ckalloc((unsigned) elemSize + 1);
+		keyPtr->length = TclCopyAndCollapse(elemSize, elemStart,
+			keyPtr->bytes);
+	    }
+
+	    if (TclFindDictElement(interp, nextElem, (limit - nextElem),
+		    &elemStart, &nextElem, &elemSize, &literal) != TCL_OK) {
+		TclDecrRefCount(keyPtr);
+		goto errorInFindDictElement;
+	    }
+
+	    if (literal) {
+		TclNewStringObj(valuePtr, elemStart, elemSize);
+	    } else {
+		/* Avoid double copy */
+		TclNewObj(valuePtr);
+		valuePtr->bytes = ckalloc((unsigned) elemSize + 1);
+		valuePtr->length = TclCopyAndCollapse(elemSize, elemStart,
+			valuePtr->bytes);
+	    }
+
+	    /* Store key and value in the hash table we're building. */
+	    hPtr = CreateChainEntry(dict, keyPtr, &isNew);
+	    if (!isNew) {
+		Tcl_Obj *discardedValue = Tcl_GetHashValue(hPtr);
+
+		TclDecrRefCount(keyPtr);
+		TclDecrRefCount(discardedValue);
+	    }
+	    Tcl_SetHashValue(hPtr, valuePtr);
+	    Tcl_IncrRefCount(valuePtr); /* since hash now holds ref to it */
+	}
+    }
+
+    /*
+     * Free the old internalRep before setting the new one. We do this as late
+     * as possible to allow the conversion code, in particular
+     * Tcl_GetStringFromObj, to use that old internalRep.
+     */
+
+    TclFreeIntRep(objPtr);
+    dict->epoch = 0;
+    dict->chain = NULL;
+    dict->refCount = 1;
+    DICT(objPtr) = dict;
+    objPtr->internalRep.twoPtrValue.ptr2 = NULL;
+    objPtr->typePtr = &tclDictType;
+    return TCL_OK;
+
+  missingValue:
+    if (interp != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"missing value to go with key", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DICTIONARY", NULL);
+    }
+  errorInFindDictElement:
+    DeleteChainTable(dict);
+    ckfree(dict);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTraceDictPath --
+ *
+ *	Trace through a tree of dictionaries using the array of keys given. If
+ *	the flags argument has the DICT_PATH_UPDATE flag is set, a
+ *	backward-pointing chain of dictionaries is also built (in the Dict's
+ *	chain field) and the chained dictionaries are made into unshared
+ *	dictionaries (if they aren't already.)
+ *
+ * Results:
+ *	The object at the end of the path, or NULL if there was an error. Note
+ *	that this it is an error for an intermediate dictionary on the path to
+ *	not exist. If the flags argument has the DICT_PATH_EXISTS set, a
+ *	non-existent path gives a DICT_PATH_NON_EXISTENT result.
+ *
+ * Side effects:
+ *	If the flags argument is zero or DICT_PATH_EXISTS, there are no side
+ *	effects (other than potential conversion of objects to dictionaries.)
+ *	If the flags argument is DICT_PATH_UPDATE, the following additional
+ *	side effects occur. Shared dictionaries along the path are converted
+ *	into unshared objects, and a backward-pointing chain is built using
+ *	the chain fields of the dictionaries (for easy invalidation of string
+ *	representations using InvalidateDictChain). If the flags argument has
+ *	the DICT_PATH_CREATE bits set (and not the DICT_PATH_EXISTS bit),
+ *	non-existant keys will be inserted with a value of an empty
+ *	dictionary, resulting in the path being built.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclTraceDictPath(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    int keyc,
+    Tcl_Obj *const keyv[],
+    int flags)
+{
+    Dict *dict, *newDict;
+    int i;
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return NULL;
+    }
+    dict = DICT(dictPtr);
+    if (flags & DICT_PATH_UPDATE) {
+	dict->chain = NULL;
+    }
+
+    for (i=0 ; i<keyc ; i++) {
+	Tcl_HashEntry *hPtr = Tcl_FindHashEntry(&dict->table, keyv[i]);
+	Tcl_Obj *tmpObj;
+
+	if (hPtr == NULL) {
+	    int isNew;			/* Dummy */
+
+	    if (flags & DICT_PATH_EXISTS) {
+		return DICT_PATH_NON_EXISTENT;
+	    }
+	    if ((flags & DICT_PATH_CREATE) != DICT_PATH_CREATE) {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "key \"%s\" not known in dictionary",
+			    TclGetString(keyv[i])));
+		    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "DICT",
+			    TclGetString(keyv[i]), NULL);
+		}
+		return NULL;
+	    }
+
+	    /*
+	     * The next line should always set isNew to 1.
+	     */
+
+	    hPtr = CreateChainEntry(dict, keyv[i], &isNew);
+	    tmpObj = Tcl_NewDictObj();
+	    Tcl_IncrRefCount(tmpObj);
+	    Tcl_SetHashValue(hPtr, tmpObj);
+	} else {
+	    tmpObj = Tcl_GetHashValue(hPtr);
+	    if (tmpObj->typePtr != &tclDictType
+		    && SetDictFromAny(interp, tmpObj) != TCL_OK) {
+		return NULL;
+	    }
+	}
+
+	newDict = DICT(tmpObj);
+	if (flags & DICT_PATH_UPDATE) {
+	    if (Tcl_IsShared(tmpObj)) {
+		TclDecrRefCount(tmpObj);
+		tmpObj = Tcl_DuplicateObj(tmpObj);
+		Tcl_IncrRefCount(tmpObj);
+		Tcl_SetHashValue(hPtr, tmpObj);
+		dict->epoch++;
+		newDict = DICT(tmpObj);
+	    }
+
+	    newDict->chain = dictPtr;
+	}
+	dict = newDict;
+	dictPtr = tmpObj;
+    }
+    return dictPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InvalidateDictChain --
+ *
+ *	Go through a dictionary chain (built by an updating invokation of
+ *	TclTraceDictPath) and invalidate the string representations of all the
+ *	dictionaries on the chain.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	String reps are invalidated and epoch counters (for detecting illegal
+ *	concurrent modifications) are updated through the chain of updated
+ *	dictionaries.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+InvalidateDictChain(
+    Tcl_Obj *dictObj)
+{
+    Dict *dict = DICT(dictObj);
+
+    do {
+	TclInvalidateStringRep(dictObj);
+	dict->epoch++;
+	dictObj = dict->chain;
+	if (dictObj == NULL) {
+	    break;
+	}
+	dict->chain = NULL;
+	dict = DICT(dictObj);
+    } while (dict != NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjPut --
+ *
+ *	Add a key,value pair to a dictionary, or update the value for a key if
+ *	that key already has a mapping in the dictionary.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	The object pointed to by dictPtr is converted to a dictionary if it is
+ *	not already one, and any string representation that it has is
+ *	invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjPut(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    Tcl_Obj *keyPtr,
+    Tcl_Obj *valuePtr)
+{
+    Dict *dict;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (Tcl_IsShared(dictPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_DictObjPut");
+    }
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (dictPtr->bytes != NULL) {
+	TclInvalidateStringRep(dictPtr);
+    }
+    dict = DICT(dictPtr);
+    hPtr = CreateChainEntry(dict, keyPtr, &isNew);
+    Tcl_IncrRefCount(valuePtr);
+    if (!isNew) {
+	Tcl_Obj *oldValuePtr = Tcl_GetHashValue(hPtr);
+
+	TclDecrRefCount(oldValuePtr);
+    }
+    Tcl_SetHashValue(hPtr, valuePtr);
+    dict->epoch++;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjGet --
+ *
+ *	Given a key, get its value from the dictionary (or NULL if key is not
+ *	found in dictionary.)
+ *
+ * Results:
+ *	A standard Tcl result. The variable pointed to by valuePtrPtr is
+ *	updated with the value for the key. Note that it is not an error for
+ *	the key to have no mapping in the dictionary.
+ *
+ * Side effects:
+ *	The object pointed to by dictPtr is converted to a dictionary if it is
+ *	not already one.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjGet(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    Tcl_Obj *keyPtr,
+    Tcl_Obj **valuePtrPtr)
+{
+    Dict *dict;
+    Tcl_HashEntry *hPtr;
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	*valuePtrPtr = NULL;
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    hPtr = Tcl_FindHashEntry(&dict->table, keyPtr);
+    if (hPtr == NULL) {
+	*valuePtrPtr = NULL;
+    } else {
+	*valuePtrPtr = Tcl_GetHashValue(hPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjRemove --
+ *
+ *	Remove the key,value pair with the given key from the dictionary; the
+ *	key does not need to be present in the dictionary.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	The object pointed to by dictPtr is converted to a dictionary if it is
+ *	not already one, and any string representation that it has is
+ *	invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjRemove(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    Tcl_Obj *keyPtr)
+{
+    Dict *dict;
+
+    if (Tcl_IsShared(dictPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_DictObjRemove");
+    }
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    if (DeleteChainEntry(dict, keyPtr)) {
+	if (dictPtr->bytes != NULL) {
+	    TclInvalidateStringRep(dictPtr);
+	}
+	dict->epoch++;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjSize --
+ *
+ *	How many key,value pairs are there in the dictionary?
+ *
+ * Results:
+ *	A standard Tcl result. Updates the variable pointed to by sizePtr with
+ *	the number of key,value pairs in the dictionary.
+ *
+ * Side effects:
+ *	The dictPtr object is converted to a dictionary type if it is not a
+ *	dictionary already.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjSize(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    int *sizePtr)
+{
+    Dict *dict;
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    *sizePtr = dict->table.numEntries;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjFirst --
+ *
+ *	Start a traversal of the dictionary. Caller must supply the search
+ *	context, pointers for returning key and value, and a pointer to allow
+ *	indication of whether the dictionary has been traversed (i.e. the
+ *	dictionary is empty). The order of traversal is undefined.
+ *
+ * Results:
+ *	A standard Tcl result. Updates the variables pointed to by keyPtrPtr,
+ *	valuePtrPtr and donePtr. Either of keyPtrPtr and valuePtrPtr may be
+ *	NULL, in which case the key/value is not made available to the caller.
+ *
+ * Side effects:
+ *	The dictPtr object is converted to a dictionary type if it is not a
+ *	dictionary already. The search context is initialised if the search
+ *	has not finished. The dictionary's internal rep is Tcl_Preserve()d if
+ *	the dictionary has at least one element.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjFirst(
+    Tcl_Interp *interp,		/* For error messages, or NULL if no error
+				 * messages desired. */
+    Tcl_Obj *dictPtr,		/* Dictionary to traverse. */
+    Tcl_DictSearch *searchPtr,	/* Pointer to a dict search context. */
+    Tcl_Obj **keyPtrPtr,	/* Pointer to a variable to have the first key
+				 * written into, or NULL. */
+    Tcl_Obj **valuePtrPtr,	/* Pointer to a variable to have the first
+				 * value written into, or NULL.*/
+    int *donePtr)		/* Pointer to a variable which will have a 1
+				 * written into when there are no further
+				 * values in the dictionary, or a 0
+				 * otherwise. */
+{
+    Dict *dict;
+    ChainEntry *cPtr;
+
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    cPtr = dict->entryChainHead;
+    if (cPtr == NULL) {
+	searchPtr->epoch = -1;
+	*donePtr = 1;
+    } else {
+	*donePtr = 0;
+	searchPtr->dictionaryPtr = (Tcl_Dict) dict;
+	searchPtr->epoch = dict->epoch;
+	searchPtr->next = cPtr->nextPtr;
+	dict->refCount++;
+	if (keyPtrPtr != NULL) {
+	    *keyPtrPtr = Tcl_GetHashKey(&dict->table, &cPtr->entry);
+	}
+	if (valuePtrPtr != NULL) {
+	    *valuePtrPtr = Tcl_GetHashValue(&cPtr->entry);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjNext --
+ *
+ *	Continue a traversal of a dictionary previously started with
+ *	Tcl_DictObjFirst. This function is safe against concurrent
+ *	modification of the underlying object (including type shimmering),
+ *	treating such situations as if the search has terminated, though it is
+ *	up to the caller to ensure that the object itself is not disposed
+ *	until the search has finished. It is _not_ safe against modifications
+ *	from other threads.
+ *
+ * Results:
+ *	Updates the variables pointed to by keyPtrPtr, valuePtrPtr and
+ *	donePtr. Either of keyPtrPtr and valuePtrPtr may be NULL, in which
+ *	case the key/value is not made available to the caller.
+ *
+ * Side effects:
+ *	Removes a reference to the dictionary's internal rep if the search
+ *	terminates.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DictObjNext(
+    Tcl_DictSearch *searchPtr,	/* Pointer to a hash search context. */
+    Tcl_Obj **keyPtrPtr,	/* Pointer to a variable to have the first key
+				 * written into, or NULL. */
+    Tcl_Obj **valuePtrPtr,	/* Pointer to a variable to have the first
+				 * value written into, or NULL.*/
+    int *donePtr)		/* Pointer to a variable which will have a 1
+				 * written into when there are no further
+				 * values in the dictionary, or a 0
+				 * otherwise. */
+{
+    ChainEntry *cPtr;
+
+    /*
+     * If the searh is done; we do no work.
+     */
+
+    if (searchPtr->epoch == -1) {
+	*donePtr = 1;
+	return;
+    }
+
+    /*
+     * Bail out if the dictionary has had any elements added, modified or
+     * removed. This *shouldn't* happen, but...
+     */
+
+    if (((Dict *)searchPtr->dictionaryPtr)->epoch != searchPtr->epoch) {
+	Tcl_Panic("concurrent dictionary modification and search");
+    }
+
+    cPtr = searchPtr->next;
+    if (cPtr == NULL) {
+	Tcl_DictObjDone(searchPtr);
+	*donePtr = 1;
+	return;
+    }
+
+    searchPtr->next = cPtr->nextPtr;
+    *donePtr = 0;
+    if (keyPtrPtr != NULL) {
+	*keyPtrPtr = Tcl_GetHashKey(
+		&((Dict *)searchPtr->dictionaryPtr)->table, &cPtr->entry);
+    }
+    if (valuePtrPtr != NULL) {
+	*valuePtrPtr = Tcl_GetHashValue(&cPtr->entry);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjDone --
+ *
+ *	Call this if you want to stop a search before you reach the end of the
+ *	dictionary (e.g. because of abnormal termination of the search). It
+ *	need not be used if the search reaches its natural end (i.e. if either
+ *	Tcl_DictObjFirst or Tcl_DictObjNext sets its donePtr variable to 1).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes a reference to the dictionary's internal rep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DictObjDone(
+    Tcl_DictSearch *searchPtr)		/* Pointer to a hash search context. */
+{
+    Dict *dict;
+
+    if (searchPtr->epoch != -1) {
+	searchPtr->epoch = -1;
+	dict = (Dict *) searchPtr->dictionaryPtr;
+	if (dict->refCount-- <= 1) {
+	    DeleteDict(dict);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjPutKeyList --
+ *
+ *	Add a key...key,value pair to a dictionary tree. The main dictionary
+ *	value must not be shared, though sub-dictionaries may be. All
+ *	intermediate dictionaries on the path must exist.
+ *
+ * Results:
+ *	A standard Tcl result. Note that in the error case, a message is left
+ *	in interp unless that is NULL.
+ *
+ * Side effects:
+ *	If the dictionary and any of its sub-dictionaries on the path have
+ *	string representations, these are invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjPutKeyList(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    int keyc,
+    Tcl_Obj *const keyv[],
+    Tcl_Obj *valuePtr)
+{
+    Dict *dict;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (Tcl_IsShared(dictPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_DictObjPutKeyList");
+    }
+    if (keyc < 1) {
+	Tcl_Panic("%s called with empty key list", "Tcl_DictObjPutKeyList");
+    }
+
+    dictPtr = TclTraceDictPath(interp, dictPtr, keyc-1,keyv, DICT_PATH_CREATE);
+    if (dictPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    hPtr = CreateChainEntry(dict, keyv[keyc-1], &isNew);
+    Tcl_IncrRefCount(valuePtr);
+    if (!isNew) {
+	Tcl_Obj *oldValuePtr = Tcl_GetHashValue(hPtr);
+
+	TclDecrRefCount(oldValuePtr);
+    }
+    Tcl_SetHashValue(hPtr, valuePtr);
+    InvalidateDictChain(dictPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DictObjRemoveKeyList --
+ *
+ *	Remove a key...key,value pair from a dictionary tree (the value
+ *	removed is implicit in the key path). The main dictionary value must
+ *	not be shared, though sub-dictionaries may be. It is not an error if
+ *	there is no value associated with the given key list, but all
+ *	intermediate dictionaries on the key path must exist.
+ *
+ * Results:
+ *	A standard Tcl result. Note that in the error case, a message is left
+ *	in interp unless that is NULL.
+ *
+ * Side effects:
+ *	If the dictionary and any of its sub-dictionaries on the key path have
+ *	string representations, these are invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DictObjRemoveKeyList(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    int keyc,
+    Tcl_Obj *const keyv[])
+{
+    Dict *dict;
+
+    if (Tcl_IsShared(dictPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_DictObjRemoveKeyList");
+    }
+    if (keyc < 1) {
+	Tcl_Panic("%s called with empty key list", "Tcl_DictObjRemoveKeyList");
+    }
+
+    dictPtr = TclTraceDictPath(interp, dictPtr, keyc-1,keyv, DICT_PATH_UPDATE);
+    if (dictPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    dict = DICT(dictPtr);
+    DeleteChainEntry(dict, keyv[keyc-1]);
+    InvalidateDictChain(dictPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewDictObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates a new dict object without any
+ *	content.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewDictObj.
+ *
+ * Results:
+ *	A new dict object is returned; it has no keys defined in it. The new
+ *	object's string representation is left NULL, and the ref count of the
+ *	object is 0.
+ *
+ * Side Effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_NewDictObj(void)
+{
+#ifdef TCL_MEM_DEBUG
+    return Tcl_DbNewDictObj("unknown", 0);
+#else /* !TCL_MEM_DEBUG */
+
+    Tcl_Obj *dictPtr;
+    Dict *dict;
+
+    TclNewObj(dictPtr);
+    TclInvalidateStringRep(dictPtr);
+    dict = ckalloc(sizeof(Dict));
+    InitChainTable(dict);
+    dict->epoch = 0;
+    dict->chain = NULL;
+    dict->refCount = 1;
+    DICT(dictPtr) = dict;
+    dictPtr->internalRep.twoPtrValue.ptr2 = NULL;
+    dictPtr->typePtr = &tclDictType;
+    return dictPtr;
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewDictObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new dict objects. It is the same
+ *	as the Tcl_NewDictObj function above except that it calls
+ *	Tcl_DbCkalloc directly with the file name and line number from its
+ *	caller. This simplifies debugging since then the [memory active]
+ *	command will report the correct file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewDictObj.
+ *
+ * Results:
+ *	A new dict object is returned; it has no keys defined in it. The new
+ *	object's string representation is left NULL, and the ref count of the
+ *	object is 0.
+ *
+ * Side Effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_DbNewDictObj(
+    const char *file,
+    int line)
+{
+#ifdef TCL_MEM_DEBUG
+    Tcl_Obj *dictPtr;
+    Dict *dict;
+
+    TclDbNewObj(dictPtr, file, line);
+    TclInvalidateStringRep(dictPtr);
+    dict = ckalloc(sizeof(Dict));
+    InitChainTable(dict);
+    dict->epoch = 0;
+    dict->chain = NULL;
+    dict->refCount = 1;
+    DICT(dictPtr) = dict;
+    dictPtr->internalRep.twoPtrValue.ptr2 = NULL;
+    dictPtr->typePtr = &tclDictType;
+    return dictPtr;
+#else /* !TCL_MEM_DEBUG */
+    return Tcl_NewDictObj();
+#endif
+}
+
+/***** START OF FUNCTIONS IMPLEMENTING TCL COMMANDS *****/
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictCreateCmd --
+ *
+ *	This function implements the "dict create" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictCreateCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictObj;
+    int i;
+
+    /*
+     * Must have an even number of arguments; note that number of preceding
+     * arguments (i.e. "dict create" is also even, which makes this much
+     * easier.)
+     */
+
+    if ((objc & 1) == 0) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?key value ...?");
+	return TCL_ERROR;
+    }
+
+    dictObj = Tcl_NewDictObj();
+    for (i=1 ; i<objc ; i+=2) {
+	/*
+	 * The next command is assumed to never fail...
+	 */
+	Tcl_DictObjPut(NULL, dictObj, objv[i], objv[i+1]);
+    }
+    Tcl_SetObjResult(interp, dictObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictGetCmd --
+ *
+ *	This function implements the "dict get" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictGetCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *valuePtr = NULL;
+    int result;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary ?key ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Test for the special case of no keys, which returns a *list* of all
+     * key,value pairs. We produce a copy here because that makes subsequent
+     * list handling more efficient.
+     */
+
+    if (objc == 2) {
+	Tcl_Obj *keyPtr = NULL, *listPtr;
+	Tcl_DictSearch search;
+	int done;
+
+	result = Tcl_DictObjFirst(interp, objv[1], &search,
+		&keyPtr, &valuePtr, &done);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	listPtr = Tcl_NewListObj(0, NULL);
+	while (!done) {
+	    /*
+	     * Assume these won't fail as we have complete control over the
+	     * types of things here.
+	     */
+
+	    Tcl_ListObjAppendElement(interp, listPtr, keyPtr);
+	    Tcl_ListObjAppendElement(interp, listPtr, valuePtr);
+
+	    Tcl_DictObjNext(&search, &keyPtr, &valuePtr, &done);
+	}
+	Tcl_SetObjResult(interp, listPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Loop through the list of keys, looking up the key at the current index
+     * in the current dictionary each time. Once we've done the lookup, we set
+     * the current dictionary to be the value we looked up (in case the value
+     * was not the last one and we are going through a chain of searches.)
+     * Note that this loop always executes at least once.
+     */
+
+    dictPtr = TclTraceDictPath(interp, objv[1], objc-3,objv+2, DICT_PATH_READ);
+    if (dictPtr == NULL) {
+	return TCL_ERROR;
+    }
+    result = Tcl_DictObjGet(interp, dictPtr, objv[objc-1], &valuePtr);
+    if (result != TCL_OK) {
+	return result;
+    }
+    if (valuePtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"key \"%s\" not known in dictionary",
+		TclGetString(objv[objc-1])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "DICT",
+		TclGetString(objv[objc-1]), NULL);
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, valuePtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictReplaceCmd --
+ *
+ *	This function implements the "dict replace" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictReplaceCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr;
+    int i;
+
+    if ((objc < 2) || (objc & 1)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary ?key value ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = objv[1];
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_IsShared(dictPtr)) {
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+    if (dictPtr->bytes != NULL) {
+	TclInvalidateStringRep(dictPtr);
+    }
+    for (i=2 ; i<objc ; i+=2) {
+	Tcl_DictObjPut(NULL, dictPtr, objv[i], objv[i+1]);
+    }
+    Tcl_SetObjResult(interp, dictPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictRemoveCmd --
+ *
+ *	This function implements the "dict remove" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictRemoveCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr;
+    int i;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary ?key ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = objv[1];
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_IsShared(dictPtr)) {
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+    if (dictPtr->bytes != NULL) {
+	TclInvalidateStringRep(dictPtr);
+    }
+    for (i=2 ; i<objc ; i++) {
+	Tcl_DictObjRemove(NULL, dictPtr, objv[i]);
+    }
+    Tcl_SetObjResult(interp, dictPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictMergeCmd --
+ *
+ *	This function implements the "dict merge" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#163 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictMergeCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *targetObj, *keyObj = NULL, *valueObj = NULL;
+    int allocatedDict = 0;
+    int i, done;
+    Tcl_DictSearch search;
+
+    if (objc == 1) {
+	/*
+	 * No dictionary arguments; return default (empty value).
+	 */
+
+	return TCL_OK;
+    }
+
+    /*
+     * Make sure first argument is a dictionary.
+     */
+
+    targetObj = objv[1];
+    if (targetObj->typePtr != &tclDictType
+	    && SetDictFromAny(interp, targetObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	/*
+	 * Single argument, return it.
+	 */
+
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    /*
+     * Normal behaviour: combining two (or more) dictionaries.
+     */
+
+    if (Tcl_IsShared(targetObj)) {
+	targetObj = Tcl_DuplicateObj(targetObj);
+	allocatedDict = 1;
+    }
+    for (i=2 ; i<objc ; i++) {
+	if (Tcl_DictObjFirst(interp, objv[i], &search, &keyObj, &valueObj,
+		&done) != TCL_OK) {
+	    if (allocatedDict) {
+		TclDecrRefCount(targetObj);
+	    }
+	    return TCL_ERROR;
+	}
+	while (!done) {
+	    /*
+	     * Next line can't fail; already know we have a dictionary in
+	     * targetObj.
+	     */
+
+	    Tcl_DictObjPut(NULL, targetObj, keyObj, valueObj);
+	    Tcl_DictObjNext(&search, &keyObj, &valueObj, &done);
+	}
+	Tcl_DictObjDone(&search);
+    }
+    Tcl_SetObjResult(interp, targetObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictKeysCmd --
+ *
+ *	This function implements the "dict keys" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictKeysCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *listPtr;
+    const char *pattern = NULL;
+
+    if (objc!=2 && objc!=3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary ?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * A direct check that we have a dictionary. We don't start the iteration
+     * yet because that might allocate memory or set locks that we do not
+     * need. [Bug 1705778, leak K04]
+     */
+
+    if (objv[1]->typePtr != &tclDictType
+	    && SetDictFromAny(interp, objv[1]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	pattern = TclGetString(objv[2]);
+    }
+    listPtr = Tcl_NewListObj(0, NULL);
+    if ((pattern != NULL) && TclMatchIsTrivial(pattern)) {
+	Tcl_Obj *valuePtr = NULL;
+
+	Tcl_DictObjGet(interp, objv[1], objv[2], &valuePtr);
+	if (valuePtr != NULL) {
+	    Tcl_ListObjAppendElement(NULL, listPtr, objv[2]);
+	}
+    } else {
+	Tcl_DictSearch search;
+	Tcl_Obj *keyPtr = NULL;
+	int done = 0;
+
+	/*
+	 * At this point, we know we have a dictionary (or at least something
+	 * that can be represented; it could theoretically have shimmered away
+	 * when the pattern was fetched, but that shouldn't be damaging) so we
+	 * can start the iteration process without checking for failures.
+	 */
+
+	Tcl_DictObjFirst(NULL, objv[1], &search, &keyPtr, NULL, &done);
+	for (; !done ; Tcl_DictObjNext(&search, &keyPtr, NULL, &done)) {
+	    if (!pattern || Tcl_StringMatch(TclGetString(keyPtr), pattern)) {
+		Tcl_ListObjAppendElement(NULL, listPtr, keyPtr);
+	    }
+	}
+	Tcl_DictObjDone(&search);
+    }
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictValuesCmd --
+ *
+ *	This function implements the "dict values" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictValuesCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *valuePtr = NULL, *listPtr;
+    Tcl_DictSearch search;
+    int done;
+    const char *pattern;
+
+    if (objc!=2 && objc!=3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary ?pattern?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_DictObjFirst(interp, objv[1], &search, NULL, &valuePtr,
+	    &done) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (objc == 3) {
+	pattern = TclGetString(objv[2]);
+    } else {
+	pattern = NULL;
+    }
+    listPtr = Tcl_NewListObj(0, NULL);
+    for (; !done ; Tcl_DictObjNext(&search, NULL, &valuePtr, &done)) {
+	if (pattern==NULL || Tcl_StringMatch(TclGetString(valuePtr),pattern)) {
+	    /*
+	     * Assume this operation always succeeds.
+	     */
+
+	    Tcl_ListObjAppendElement(interp, listPtr, valuePtr);
+	}
+    }
+    Tcl_DictObjDone(&search);
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictSizeCmd --
+ *
+ *	This function implements the "dict size" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictSizeCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int result, size;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary");
+	return TCL_ERROR;
+    }
+    result = Tcl_DictObjSize(interp, objv[1], &size);
+    if (result == TCL_OK) {
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(size));
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictExistsCmd --
+ *
+ *	This function implements the "dict exists" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictExistsCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *valuePtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary key ?key ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = TclTraceDictPath(interp, objv[1], objc-3, objv+2,
+	    DICT_PATH_EXISTS);
+    if (dictPtr == NULL || dictPtr == DICT_PATH_NON_EXISTENT
+	    || Tcl_DictObjGet(interp, dictPtr, objv[objc-1],
+		    &valuePtr) != TCL_OK) {
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(0));
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(valuePtr != NULL));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictInfoCmd --
+ *
+ *	This function implements the "dict info" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictInfoCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr;
+    Dict *dict;
+    char *statsStr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary");
+	return TCL_ERROR;
+    }
+
+    dictPtr = objv[1];
+    if (dictPtr->typePtr != &tclDictType
+	    && SetDictFromAny(interp, dictPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    dict = DICT(dictPtr);
+
+    statsStr = Tcl_HashStats(&dict->table);
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(statsStr, -1));
+    ckfree(statsStr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictIncrCmd --
+ *
+ *	This function implements the "dict incr" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictIncrCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int code = TCL_OK;
+    Tcl_Obj *dictPtr, *valuePtr = NULL;
+
+    if (objc < 3 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName key ?increment?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+    if (dictPtr == NULL) {
+	/*
+	 * Variable didn't yet exist. Create new dictionary value.
+	 */
+
+	dictPtr = Tcl_NewDictObj();
+    } else if (Tcl_DictObjGet(interp, dictPtr, objv[2], &valuePtr) != TCL_OK) {
+	/*
+	 * Variable contents are not a dict, report error.
+	 */
+
+	return TCL_ERROR;
+    }
+    if (Tcl_IsShared(dictPtr)) {
+	/*
+	 * A little internals surgery to avoid copying a string rep that will
+	 * soon be no good.
+	 */
+
+	char *saved = dictPtr->bytes;
+	Tcl_Obj *oldPtr = dictPtr;
+
+	dictPtr->bytes = NULL;
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+	oldPtr->bytes = saved;
+    }
+    if (valuePtr == NULL) {
+	/*
+	 * Key not in dictionary. Create new key with increment as value.
+	 */
+
+	if (objc == 4) {
+	    /*
+	     * Verify increment is an integer.
+	     */
+
+	    mp_int increment;
+
+	    code = Tcl_GetBignumFromObj(interp, objv[3], &increment);
+	    if (code != TCL_OK) {
+		Tcl_AddErrorInfo(interp, "\n    (reading increment)");
+	    } else {
+		/*
+		 * Remember to dispose with the bignum as we're not actually
+		 * using it directly. [Bug 2874678]
+		 */
+
+		mp_clear(&increment);
+		Tcl_DictObjPut(NULL, dictPtr, objv[2], objv[3]);
+	    }
+	} else {
+	    Tcl_DictObjPut(NULL, dictPtr, objv[2], Tcl_NewIntObj(1));
+	}
+    } else {
+	/*
+	 * Key in dictionary. Increment its value with minimum dup.
+	 */
+
+	if (Tcl_IsShared(valuePtr)) {
+	    valuePtr = Tcl_DuplicateObj(valuePtr);
+	    Tcl_DictObjPut(NULL, dictPtr, objv[2], valuePtr);
+	}
+	if (objc == 4) {
+	    code = TclIncrObj(interp, valuePtr, objv[3]);
+	} else {
+	    Tcl_Obj *incrPtr = Tcl_NewIntObj(1);
+
+	    Tcl_IncrRefCount(incrPtr);
+	    code = TclIncrObj(interp, valuePtr, incrPtr);
+	    TclDecrRefCount(incrPtr);
+	}
+    }
+    if (code == TCL_OK) {
+	TclInvalidateStringRep(dictPtr);
+	valuePtr = Tcl_ObjSetVar2(interp, objv[1], NULL,
+		dictPtr, TCL_LEAVE_ERR_MSG);
+	if (valuePtr == NULL) {
+	    code = TCL_ERROR;
+	} else {
+	    Tcl_SetObjResult(interp, valuePtr);
+	}
+    } else if (dictPtr->refCount == 0) {
+	TclDecrRefCount(dictPtr);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictLappendCmd --
+ *
+ *	This function implements the "dict lappend" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictLappendCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *valuePtr, *resultPtr;
+    int i, allocatedDict = 0, allocatedValue = 0;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName key ?value ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+    if (dictPtr == NULL) {
+	allocatedDict = 1;
+	dictPtr = Tcl_NewDictObj();
+    } else if (Tcl_IsShared(dictPtr)) {
+	allocatedDict = 1;
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+
+    if (Tcl_DictObjGet(interp, dictPtr, objv[2], &valuePtr) != TCL_OK) {
+	if (allocatedDict) {
+	    TclDecrRefCount(dictPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    if (valuePtr == NULL) {
+	valuePtr = Tcl_NewListObj(objc-3, objv+3);
+	allocatedValue = 1;
+    } else {
+	if (Tcl_IsShared(valuePtr)) {
+	    allocatedValue = 1;
+	    valuePtr = Tcl_DuplicateObj(valuePtr);
+	}
+
+	for (i=3 ; i<objc ; i++) {
+	    if (Tcl_ListObjAppendElement(interp, valuePtr,
+		    objv[i]) != TCL_OK) {
+		if (allocatedValue) {
+		    TclDecrRefCount(valuePtr);
+		}
+		if (allocatedDict) {
+		    TclDecrRefCount(dictPtr);
+		}
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    if (allocatedValue) {
+	Tcl_DictObjPut(NULL, dictPtr, objv[2], valuePtr);
+    } else if (dictPtr->bytes != NULL) {
+	TclInvalidateStringRep(dictPtr);
+    }
+
+    resultPtr = Tcl_ObjSetVar2(interp, objv[1], NULL, dictPtr,
+	    TCL_LEAVE_ERR_MSG);
+    if (resultPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictAppendCmd --
+ *
+ *	This function implements the "dict append" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictAppendCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *valuePtr, *resultPtr;
+    int allocatedDict = 0;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName key ?value ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+    if (dictPtr == NULL) {
+	allocatedDict = 1;
+	dictPtr = Tcl_NewDictObj();
+    } else if (Tcl_IsShared(dictPtr)) {
+	allocatedDict = 1;
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+
+    if (Tcl_DictObjGet(interp, dictPtr, objv[2], &valuePtr) != TCL_OK) {
+	if (allocatedDict) {
+	    TclDecrRefCount(dictPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    if ((objc > 3) || (valuePtr == NULL)) {
+	/* Only go through append activites when something will change. */
+	Tcl_Obj *appendObjPtr = NULL;
+
+	if (objc > 3) {
+	    /* Something to append */
+
+	    if (objc == 4) {
+		appendObjPtr = objv[3];
+	    } else if (TCL_OK != TclStringCatObjv(interp, /* inPlace */ 1,
+		    objc-3, objv+3, &appendObjPtr)) {
+		return TCL_ERROR;
+	    }
+	}
+
+	if (appendObjPtr == NULL) {
+	    /* => (objc == 3) => (valuePtr == NULL) */
+	    TclNewObj(valuePtr);
+	} else if (valuePtr == NULL) {
+	    valuePtr = appendObjPtr;
+	    appendObjPtr = NULL;
+	}
+
+	if (appendObjPtr) {
+	    if (Tcl_IsShared(valuePtr)) {
+		valuePtr = Tcl_DuplicateObj(valuePtr);
+	    }
+
+	    Tcl_AppendObjToObj(valuePtr, appendObjPtr);
+	}
+
+	Tcl_DictObjPut(NULL, dictPtr, objv[2], valuePtr);
+    }
+
+    /*
+     * Even if nothing changed, we still overwrite so that variable
+     * trace expectations are met.
+     */
+
+    resultPtr = Tcl_ObjSetVar2(interp, objv[1], NULL, dictPtr,
+	    TCL_LEAVE_ERR_MSG);
+    if (resultPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictForNRCmd --
+ *
+ *	These functions implement the "dict for" Tcl command.  See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictForNRCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *scriptObj, *keyVarObj, *valueVarObj;
+    Tcl_Obj **varv, *keyObj, *valueObj;
+    Tcl_DictSearch *searchPtr;
+    int varc, done;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"{keyVarName valueVarName} dictionary script");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse arguments.
+     */
+
+    if (TclListObjGetElements(interp, objv[1], &varc, &varv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (varc != 2) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"must have exactly two variable names", -1));
+	Tcl_SetErrorCode(interp, "TCL", "SYNTAX", "dict", "for", NULL);
+	return TCL_ERROR;
+    }
+    searchPtr = TclStackAlloc(interp, sizeof(Tcl_DictSearch));
+    if (Tcl_DictObjFirst(interp, objv[2], searchPtr, &keyObj, &valueObj,
+	    &done) != TCL_OK) {
+	TclStackFree(interp, searchPtr);
+	return TCL_ERROR;
+    }
+    if (done) {
+	TclStackFree(interp, searchPtr);
+	return TCL_OK;
+    }
+    TclListObjGetElements(NULL, objv[1], &varc, &varv);
+    keyVarObj = varv[0];
+    valueVarObj = varv[1];
+    scriptObj = objv[3];
+
+    /*
+     * Make sure that these objects (which we need throughout the body of the
+     * loop) don't vanish. Note that the dictionary internal rep is locked
+     * internally so that updates, shimmering, etc are not a problem.
+     */
+
+    Tcl_IncrRefCount(keyVarObj);
+    Tcl_IncrRefCount(valueVarObj);
+    Tcl_IncrRefCount(scriptObj);
+
+    /*
+     * Stop the value from getting hit in any way by any traces on the key
+     * variable.
+     */
+
+    Tcl_IncrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, keyVarObj, NULL, keyObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	goto error;
+    }
+    TclDecrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, valueVarObj, NULL, valueObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	goto error;
+    }
+
+    /*
+     * Run the script.
+     */
+
+    TclNRAddCallback(interp, DictForLoopCallback, searchPtr, keyVarObj,
+	    valueVarObj, scriptObj);
+    return TclNREvalObjEx(interp, scriptObj, 0, iPtr->cmdFramePtr, 3);
+
+    /*
+     * For unwinding everything on error.
+     */
+
+  error:
+    TclDecrRefCount(keyVarObj);
+    TclDecrRefCount(valueVarObj);
+    TclDecrRefCount(scriptObj);
+    Tcl_DictObjDone(searchPtr);
+    TclStackFree(interp, searchPtr);
+    return TCL_ERROR;
+}
+
+static int
+DictForLoopCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_DictSearch *searchPtr = data[0];
+    Tcl_Obj *keyVarObj = data[1];
+    Tcl_Obj *valueVarObj = data[2];
+    Tcl_Obj *scriptObj = data[3];
+    Tcl_Obj *keyObj, *valueObj;
+    int done;
+
+    /*
+     * Process the result from the previous execution of the script body.
+     */
+
+    if (result == TCL_CONTINUE) {
+	result = TCL_OK;
+    } else if (result != TCL_OK) {
+	if (result == TCL_BREAK) {
+	    Tcl_ResetResult(interp);
+	    result = TCL_OK;
+	} else if (result == TCL_ERROR) {
+	    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		    "\n    (\"dict for\" body line %d)",
+		    Tcl_GetErrorLine(interp)));
+	}
+	goto done;
+    }
+
+    /*
+     * Get the next mapping from the dictionary.
+     */
+
+    Tcl_DictObjNext(searchPtr, &keyObj, &valueObj, &done);
+    if (done) {
+	Tcl_ResetResult(interp);
+	goto done;
+    }
+
+    /*
+     * Stop the value from getting hit in any way by any traces on the key
+     * variable.
+     */
+
+    Tcl_IncrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, keyVarObj, NULL, keyObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	result = TCL_ERROR;
+	goto done;
+    }
+    TclDecrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, valueVarObj, NULL, valueObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	result = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Run the script.
+     */
+
+    TclNRAddCallback(interp, DictForLoopCallback, searchPtr, keyVarObj,
+	    valueVarObj, scriptObj);
+    return TclNREvalObjEx(interp, scriptObj, 0, iPtr->cmdFramePtr, 3);
+
+    /*
+     * For unwinding everything once the iterating is done.
+     */
+
+  done:
+    TclDecrRefCount(keyVarObj);
+    TclDecrRefCount(valueVarObj);
+    TclDecrRefCount(scriptObj);
+    Tcl_DictObjDone(searchPtr);
+    TclStackFree(interp, searchPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictMapNRCmd --
+ *
+ *	These functions implement the "dict map" Tcl command.  See the user
+ *	documentation for details on what it does, and TIP#405 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictMapNRCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj **varv, *keyObj, *valueObj;
+    DictMapStorage *storagePtr;
+    int varc, done;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"{keyVarName valueVarName} dictionary script");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse arguments.
+     */
+
+    if (TclListObjGetElements(interp, objv[1], &varc, &varv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (varc != 2) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"must have exactly two variable names", -1));
+	Tcl_SetErrorCode(interp, "TCL", "SYNTAX", "dict", "map", NULL);
+	return TCL_ERROR;
+    }
+    storagePtr = TclStackAlloc(interp, sizeof(DictMapStorage));
+    if (Tcl_DictObjFirst(interp, objv[2], &storagePtr->search, &keyObj,
+	    &valueObj, &done) != TCL_OK) {
+	TclStackFree(interp, storagePtr);
+	return TCL_ERROR;
+    }
+    if (done) {
+	/*
+	 * Note that this exit leaves an empty value in the result (due to
+	 * command calling conventions) but that is OK since an empty value is
+	 * an empty dictionary.
+	 */
+
+	TclStackFree(interp, storagePtr);
+	return TCL_OK;
+    }
+    TclNewObj(storagePtr->accumulatorObj);
+    TclListObjGetElements(NULL, objv[1], &varc, &varv);
+    storagePtr->keyVarObj = varv[0];
+    storagePtr->valueVarObj = varv[1];
+    storagePtr->scriptObj = objv[3];
+
+    /*
+     * Make sure that these objects (which we need throughout the body of the
+     * loop) don't vanish. Note that the dictionary internal rep is locked
+     * internally so that updates, shimmering, etc are not a problem.
+     */
+
+    Tcl_IncrRefCount(storagePtr->accumulatorObj);
+    Tcl_IncrRefCount(storagePtr->keyVarObj);
+    Tcl_IncrRefCount(storagePtr->valueVarObj);
+    Tcl_IncrRefCount(storagePtr->scriptObj);
+
+    /*
+     * Stop the value from getting hit in any way by any traces on the key
+     * variable.
+     */
+
+    Tcl_IncrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, storagePtr->keyVarObj, NULL, keyObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	goto error;
+    }
+    if (Tcl_ObjSetVar2(interp, storagePtr->valueVarObj, NULL, valueObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	goto error;
+    }
+    TclDecrRefCount(valueObj);
+
+    /*
+     * Run the script.
+     */
+
+    TclNRAddCallback(interp, DictMapLoopCallback, storagePtr, NULL,NULL,NULL);
+    return TclNREvalObjEx(interp, storagePtr->scriptObj, 0,
+	    iPtr->cmdFramePtr, 3);
+
+    /*
+     * For unwinding everything on error.
+     */
+
+  error:
+    TclDecrRefCount(storagePtr->keyVarObj);
+    TclDecrRefCount(storagePtr->valueVarObj);
+    TclDecrRefCount(storagePtr->scriptObj);
+    TclDecrRefCount(storagePtr->accumulatorObj);
+    Tcl_DictObjDone(&storagePtr->search);
+    TclStackFree(interp, storagePtr);
+    return TCL_ERROR;
+}
+
+static int
+DictMapLoopCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    DictMapStorage *storagePtr = data[0];
+    Tcl_Obj *keyObj, *valueObj;
+    int done;
+
+    /*
+     * Process the result from the previous execution of the script body.
+     */
+
+    if (result == TCL_CONTINUE) {
+	result = TCL_OK;
+    } else if (result != TCL_OK) {
+	if (result == TCL_BREAK) {
+	    Tcl_ResetResult(interp);
+	    result = TCL_OK;
+	} else if (result == TCL_ERROR) {
+	    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		    "\n    (\"dict map\" body line %d)",
+		    Tcl_GetErrorLine(interp)));
+	}
+	goto done;
+    } else {
+	keyObj = Tcl_ObjGetVar2(interp, storagePtr->keyVarObj, NULL,
+		TCL_LEAVE_ERR_MSG);
+	if (keyObj == NULL) {
+	    result = TCL_ERROR;
+	    goto done;
+	}
+	Tcl_DictObjPut(NULL, storagePtr->accumulatorObj, keyObj,
+		Tcl_GetObjResult(interp));
+    }
+
+    /*
+     * Get the next mapping from the dictionary.
+     */
+
+    Tcl_DictObjNext(&storagePtr->search, &keyObj, &valueObj, &done);
+    if (done) {
+	Tcl_SetObjResult(interp, storagePtr->accumulatorObj);
+	goto done;
+    }
+
+    /*
+     * Stop the value from getting hit in any way by any traces on the key
+     * variable.
+     */
+
+    Tcl_IncrRefCount(valueObj);
+    if (Tcl_ObjSetVar2(interp, storagePtr->keyVarObj, NULL, keyObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	result = TCL_ERROR;
+	goto done;
+    }
+    if (Tcl_ObjSetVar2(interp, storagePtr->valueVarObj, NULL, valueObj,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	TclDecrRefCount(valueObj);
+	result = TCL_ERROR;
+	goto done;
+    }
+    TclDecrRefCount(valueObj);
+
+    /*
+     * Run the script.
+     */
+
+    TclNRAddCallback(interp, DictMapLoopCallback, storagePtr, NULL,NULL,NULL);
+    return TclNREvalObjEx(interp, storagePtr->scriptObj, 0,
+	    iPtr->cmdFramePtr, 3);
+
+    /*
+     * For unwinding everything once the iterating is done.
+     */
+
+  done:
+    TclDecrRefCount(storagePtr->keyVarObj);
+    TclDecrRefCount(storagePtr->valueVarObj);
+    TclDecrRefCount(storagePtr->scriptObj);
+    TclDecrRefCount(storagePtr->accumulatorObj);
+    Tcl_DictObjDone(&storagePtr->search);
+    TclStackFree(interp, storagePtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictSetCmd --
+ *
+ *	This function implements the "dict set" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictSetCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *resultPtr;
+    int result, allocatedDict = 0;
+
+    if (objc < 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName key ?key ...? value");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+    if (dictPtr == NULL) {
+	allocatedDict = 1;
+	dictPtr = Tcl_NewDictObj();
+    } else if (Tcl_IsShared(dictPtr)) {
+	allocatedDict = 1;
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+
+    result = Tcl_DictObjPutKeyList(interp, dictPtr, objc-3, objv+2,
+	    objv[objc-1]);
+    if (result != TCL_OK) {
+	if (allocatedDict) {
+	    TclDecrRefCount(dictPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    resultPtr = Tcl_ObjSetVar2(interp, objv[1], NULL, dictPtr,
+	    TCL_LEAVE_ERR_MSG);
+    if (resultPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictUnsetCmd --
+ *
+ *	This function implements the "dict unset" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictUnsetCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *dictPtr, *resultPtr;
+    int result, allocatedDict = 0;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName key ?key ...?");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+    if (dictPtr == NULL) {
+	allocatedDict = 1;
+	dictPtr = Tcl_NewDictObj();
+    } else if (Tcl_IsShared(dictPtr)) {
+	allocatedDict = 1;
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+
+    result = Tcl_DictObjRemoveKeyList(interp, dictPtr, objc-2, objv+2);
+    if (result != TCL_OK) {
+	if (allocatedDict) {
+	    TclDecrRefCount(dictPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    resultPtr = Tcl_ObjSetVar2(interp, objv[1], NULL, dictPtr,
+	    TCL_LEAVE_ERR_MSG);
+    if (resultPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictFilterCmd --
+ *
+ *	This function implements the "dict filter" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictFilterCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    static const char *const filters[] = {
+	"key", "script", "value", NULL
+    };
+    enum FilterTypes {
+	FILTER_KEYS, FILTER_SCRIPT, FILTER_VALUES
+    };
+    Tcl_Obj *scriptObj, *keyVarObj, *valueVarObj;
+    Tcl_Obj **varv, *keyObj = NULL, *valueObj = NULL, *resultObj, *boolObj;
+    Tcl_DictSearch search;
+    int index, varc, done, result, satisfied;
+    const char *pattern;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictionary filterType ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[2], filters, "filterType",
+	     0, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum FilterTypes) index) {
+    case FILTER_KEYS:
+	/*
+	 * Create a dictionary whose keys all match a certain pattern.
+	 */
+
+	if (Tcl_DictObjFirst(interp, objv[1], &search,
+		&keyObj, &valueObj, &done) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (objc == 3) {
+	    /*
+	     * Nothing to match, so return nothing (== empty dictionary).
+	     */
+
+	    Tcl_DictObjDone(&search);
+	    return TCL_OK;
+	} else if (objc == 4) {
+	    pattern = TclGetString(objv[3]);
+	    resultObj = Tcl_NewDictObj();
+	    if (TclMatchIsTrivial(pattern)) {
+		/*
+		 * Must release the search lock here to prevent a memory leak
+		 * since we are not exhausing the search. [Bug 1705778, leak
+		 * K05]
+		 */
+
+		Tcl_DictObjDone(&search);
+		Tcl_DictObjGet(interp, objv[1], objv[3], &valueObj);
+		if (valueObj != NULL) {
+		    Tcl_DictObjPut(NULL, resultObj, objv[3], valueObj);
+		}
+	    } else {
+		while (!done) {
+		    if (Tcl_StringMatch(TclGetString(keyObj), pattern)) {
+			Tcl_DictObjPut(NULL, resultObj, keyObj, valueObj);
+		    }
+		    Tcl_DictObjNext(&search, &keyObj, &valueObj, &done);
+		}
+	    }
+	} else {
+	    /*
+	     * Can't optimize this match for trivial globbing: would disturb
+	     * order.
+	     */
+
+	    resultObj = Tcl_NewDictObj();
+	    while (!done) {
+		int i;
+
+		for (i=3 ; i<objc ; i++) {
+		    pattern = TclGetString(objv[i]);
+		    if (Tcl_StringMatch(TclGetString(keyObj), pattern)) {
+			Tcl_DictObjPut(NULL, resultObj, keyObj, valueObj);
+			break;		/* stop inner loop */
+		    }
+		}
+		Tcl_DictObjNext(&search, &keyObj, &valueObj, &done);
+	    }
+	}
+	Tcl_SetObjResult(interp, resultObj);
+	return TCL_OK;
+
+    case FILTER_VALUES:
+	/*
+	 * Create a dictionary whose values all match a certain pattern.
+	 */
+
+	if (Tcl_DictObjFirst(interp, objv[1], &search,
+		&keyObj, &valueObj, &done) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	resultObj = Tcl_NewDictObj();
+	while (!done) {
+	    int i;
+
+	    for (i=3 ; i<objc ; i++) {
+		pattern = TclGetString(objv[i]);
+		if (Tcl_StringMatch(TclGetString(valueObj), pattern)) {
+		    Tcl_DictObjPut(NULL, resultObj, keyObj, valueObj);
+		    break;		/* stop inner loop */
+		}
+	    }
+	    Tcl_DictObjNext(&search, &keyObj, &valueObj, &done);
+	}
+	Tcl_SetObjResult(interp, resultObj);
+	return TCL_OK;
+
+    case FILTER_SCRIPT:
+	if (objc != 5) {
+	    Tcl_WrongNumArgs(interp, 1, objv,
+		    "dictionary script {keyVarName valueVarName} filterScript");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Create a dictionary whose key,value pairs all satisfy a script
+	 * (i.e. get a true boolean result from its evaluation). Massive
+	 * copying from the "dict for" implementation has occurred!
+	 */
+
+	if (TclListObjGetElements(interp, objv[3], &varc, &varv) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (varc != 2) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "must have exactly two variable names", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "SYNTAX", "dict", "filter", NULL);
+	    return TCL_ERROR;
+	}
+	keyVarObj = varv[0];
+	valueVarObj = varv[1];
+	scriptObj = objv[4];
+
+	/*
+	 * Make sure that these objects (which we need throughout the body of
+	 * the loop) don't vanish. Note that the dictionary internal rep is
+	 * locked internally so that updates, shimmering, etc are not a
+	 * problem.
+	 */
+
+	Tcl_IncrRefCount(keyVarObj);
+	Tcl_IncrRefCount(valueVarObj);
+	Tcl_IncrRefCount(scriptObj);
+
+	result = Tcl_DictObjFirst(interp, objv[1],
+		&search, &keyObj, &valueObj, &done);
+	if (result != TCL_OK) {
+	    TclDecrRefCount(keyVarObj);
+	    TclDecrRefCount(valueVarObj);
+	    TclDecrRefCount(scriptObj);
+	    return TCL_ERROR;
+	}
+
+	resultObj = Tcl_NewDictObj();
+
+	while (!done) {
+	    /*
+	     * Stop the value from getting hit in any way by any traces on the
+	     * key variable.
+	     */
+
+	    Tcl_IncrRefCount(keyObj);
+	    Tcl_IncrRefCount(valueObj);
+	    if (Tcl_ObjSetVar2(interp, keyVarObj, NULL, keyObj,
+		    TCL_LEAVE_ERR_MSG) == NULL) {
+		Tcl_AddErrorInfo(interp,
+			"\n    (\"dict filter\" filter script key variable)");
+		result = TCL_ERROR;
+		goto abnormalResult;
+	    }
+	    if (Tcl_ObjSetVar2(interp, valueVarObj, NULL, valueObj,
+		    TCL_LEAVE_ERR_MSG) == NULL) {
+		Tcl_AddErrorInfo(interp,
+			"\n    (\"dict filter\" filter script value variable)");
+		result = TCL_ERROR;
+		goto abnormalResult;
+	    }
+
+	    /*
+	     * TIP #280. Make invoking context available to loop body.
+	     */
+
+	    result = TclEvalObjEx(interp, scriptObj, 0, iPtr->cmdFramePtr, 4);
+	    switch (result) {
+	    case TCL_OK:
+		boolObj = Tcl_GetObjResult(interp);
+		Tcl_IncrRefCount(boolObj);
+		Tcl_ResetResult(interp);
+		if (Tcl_GetBooleanFromObj(interp, boolObj,
+			&satisfied) != TCL_OK) {
+		    TclDecrRefCount(boolObj);
+		    result = TCL_ERROR;
+		    goto abnormalResult;
+		}
+		TclDecrRefCount(boolObj);
+		if (satisfied) {
+		    Tcl_DictObjPut(NULL, resultObj, keyObj, valueObj);
+		}
+		break;
+	    case TCL_BREAK:
+		/*
+		 * Force loop termination by calling Tcl_DictObjDone; this
+		 * makes the next Tcl_DictObjNext say there is nothing more to
+		 * do.
+		 */
+
+		Tcl_ResetResult(interp);
+		Tcl_DictObjDone(&search);
+	    case TCL_CONTINUE:
+		result = TCL_OK;
+		break;
+	    case TCL_ERROR:
+		Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			"\n    (\"dict filter\" script line %d)",
+			Tcl_GetErrorLine(interp)));
+	    default:
+		goto abnormalResult;
+	    }
+
+	    TclDecrRefCount(keyObj);
+	    TclDecrRefCount(valueObj);
+
+	    Tcl_DictObjNext(&search, &keyObj, &valueObj, &done);
+	}
+
+	/*
+	 * Stop holding a reference to these objects.
+	 */
+
+	TclDecrRefCount(keyVarObj);
+	TclDecrRefCount(valueVarObj);
+	TclDecrRefCount(scriptObj);
+	Tcl_DictObjDone(&search);
+
+	if (result == TCL_OK) {
+	    Tcl_SetObjResult(interp, resultObj);
+	} else {
+	    TclDecrRefCount(resultObj);
+	}
+	return result;
+
+    abnormalResult:
+	Tcl_DictObjDone(&search);
+	TclDecrRefCount(keyObj);
+	TclDecrRefCount(valueObj);
+	TclDecrRefCount(keyVarObj);
+	TclDecrRefCount(valueVarObj);
+	TclDecrRefCount(scriptObj);
+	TclDecrRefCount(resultObj);
+	return result;
+    }
+    Tcl_Panic("unexpected fallthrough");
+    /* Control never reaches this point. */
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictUpdateCmd --
+ *
+ *	This function implements the "dict update" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#212 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictUpdateCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *dictPtr, *objPtr;
+    int i, dummy;
+
+    if (objc < 5 || !(objc & 1)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"dictVarName key varName ?key varName ...? script");
+	return TCL_ERROR;
+    }
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, TCL_LEAVE_ERR_MSG);
+    if (dictPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (Tcl_DictObjSize(interp, dictPtr, &dummy) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    Tcl_IncrRefCount(dictPtr);
+    for (i=2 ; i+2<objc ; i+=2) {
+	if (Tcl_DictObjGet(interp, dictPtr, objv[i], &objPtr) != TCL_OK) {
+	    TclDecrRefCount(dictPtr);
+	    return TCL_ERROR;
+	}
+	if (objPtr == NULL) {
+	    /* ??? */
+	    Tcl_UnsetVar(interp, Tcl_GetString(objv[i+1]), 0);
+	} else if (Tcl_ObjSetVar2(interp, objv[i+1], NULL, objPtr,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    TclDecrRefCount(dictPtr);
+	    return TCL_ERROR;
+	}
+    }
+    TclDecrRefCount(dictPtr);
+
+    /*
+     * Execute the body after setting up the NRE handler to process the
+     * results.
+     */
+
+    objPtr = Tcl_NewListObj(objc-3, objv+2);
+    Tcl_IncrRefCount(objPtr);
+    Tcl_IncrRefCount(objv[1]);
+    TclNRAddCallback(interp, FinalizeDictUpdate, objv[1], objPtr, NULL,NULL);
+
+    return TclNREvalObjEx(interp, objv[objc-1], 0, iPtr->cmdFramePtr, objc-1);
+}
+
+static int
+FinalizeDictUpdate(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj *dictPtr, *objPtr, **objv;
+    Tcl_InterpState state;
+    int i, objc;
+    Tcl_Obj *varName = data[0];
+    Tcl_Obj *argsObj = data[1];
+
+    /*
+     * ErrorInfo handling.
+     */
+
+    if (result == TCL_ERROR) {
+	Tcl_AddErrorInfo(interp, "\n    (body of \"dict update\")");
+    }
+
+    /*
+     * If the dictionary variable doesn't exist, drop everything silently.
+     */
+
+    dictPtr = Tcl_ObjGetVar2(interp, varName, NULL, 0);
+    if (dictPtr == NULL) {
+	TclDecrRefCount(varName);
+	TclDecrRefCount(argsObj);
+	return result;
+    }
+
+    /*
+     * Double-check that it is still a dictionary.
+     */
+
+    state = Tcl_SaveInterpState(interp, result);
+    if (Tcl_DictObjSize(interp, dictPtr, &objc) != TCL_OK) {
+	Tcl_DiscardInterpState(state);
+	TclDecrRefCount(varName);
+	TclDecrRefCount(argsObj);
+	return TCL_ERROR;
+    }
+
+    if (Tcl_IsShared(dictPtr)) {
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+    }
+
+    /*
+     * Write back the values from the variables, treating failure to read as
+     * an instruction to remove the key.
+     */
+
+    Tcl_ListObjGetElements(NULL, argsObj, &objc, &objv);
+    for (i=0 ; i<objc ; i+=2) {
+	objPtr = Tcl_ObjGetVar2(interp, objv[i+1], NULL, 0);
+	if (objPtr == NULL) {
+	    Tcl_DictObjRemove(NULL, dictPtr, objv[i]);
+	} else if (objPtr == dictPtr) {
+	    /*
+	     * Someone is messing us around, trying to build a recursive
+	     * structure. [Bug 1786481]
+	     */
+
+	    Tcl_DictObjPut(NULL, dictPtr, objv[i], Tcl_DuplicateObj(objPtr));
+	} else {
+	    /* Shouldn't fail */
+	    Tcl_DictObjPut(NULL, dictPtr, objv[i], objPtr);
+	}
+    }
+    TclDecrRefCount(argsObj);
+
+    /*
+     * Write the dictionary back to its variable.
+     */
+
+    if (Tcl_ObjSetVar2(interp, varName, NULL, dictPtr,
+	    TCL_LEAVE_ERR_MSG) == NULL) {
+	Tcl_DiscardInterpState(state);
+	TclDecrRefCount(varName);
+	return TCL_ERROR;
+    }
+
+    TclDecrRefCount(varName);
+    return Tcl_RestoreInterpState(interp, state);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DictWithCmd --
+ *
+ *	This function implements the "dict with" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#212 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DictWithCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *dictPtr, *keysPtr, *pathPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "dictVarName ?key ...? script");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the dictionary to open out.
+     */
+
+    dictPtr = Tcl_ObjGetVar2(interp, objv[1], NULL, TCL_LEAVE_ERR_MSG);
+    if (dictPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    keysPtr = TclDictWithInit(interp, dictPtr, objc-3, objv+2);
+    if (keysPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_IncrRefCount(keysPtr);
+
+    /*
+     * Execute the body, while making the invoking context available to the
+     * loop body (TIP#280) and postponing the cleanup until later (NRE).
+     */
+
+    pathPtr = NULL;
+    if (objc > 3) {
+	pathPtr = Tcl_NewListObj(objc-3, objv+2);
+	Tcl_IncrRefCount(pathPtr);
+    }
+    Tcl_IncrRefCount(objv[1]);
+    TclNRAddCallback(interp, FinalizeDictWith, objv[1], keysPtr, pathPtr,
+	    NULL);
+
+    return TclNREvalObjEx(interp, objv[objc-1], 0, iPtr->cmdFramePtr, objc-1);
+}
+
+static int
+FinalizeDictWith(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj **pathv;
+    int pathc;
+    Tcl_InterpState state;
+    Tcl_Obj *varName = data[0];
+    Tcl_Obj *keysPtr = data[1];
+    Tcl_Obj *pathPtr = data[2];
+    Var *varPtr, *arrayPtr;
+
+    if (result == TCL_ERROR) {
+	Tcl_AddErrorInfo(interp, "\n    (body of \"dict with\")");
+    }
+
+    /*
+     * Save the result state; TDWF doesn't guarantee to not modify that on
+     * TCL_OK result.
+     */
+
+    state = Tcl_SaveInterpState(interp, result);
+    if (pathPtr != NULL) {
+	Tcl_ListObjGetElements(NULL, pathPtr, &pathc, &pathv);
+    } else {
+	pathc = 0;
+	pathv = NULL;
+    }
+
+    /*
+     * Pack from local variables back into the dictionary.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varName, NULL, TCL_LEAVE_ERR_MSG, "set",
+	    /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+    if (varPtr == NULL) {
+	result = TCL_ERROR;
+    } else {
+	result = TclDictWithFinish(interp, varPtr, arrayPtr, varName, NULL, -1,
+		pathc, pathv, keysPtr);
+    }
+
+    /*
+     * Tidy up and return the real result (unless we had an error).
+     */
+
+    TclDecrRefCount(varName);
+    TclDecrRefCount(keysPtr);
+    if (pathPtr != NULL) {
+	TclDecrRefCount(pathPtr);
+    }
+    if (result != TCL_OK) {
+	Tcl_DiscardInterpState(state);
+	return TCL_ERROR;
+    }
+    return Tcl_RestoreInterpState(interp, state);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDictWithInit --
+ *
+ *	Part of the core of [dict with]. Pokes into a dictionary and converts
+ *	the mappings there into assignments to (presumably) local variables.
+ *	Returns a list of all the names that were mapped so that removal of
+ *	either the variable or the dictionary entry won't surprise us when we
+ *	come to stuffing everything back.
+ *
+ * Result:
+ *	List of mapped names, or NULL if there was an error.
+ *
+ * Side effects:
+ *	Assigns to variables, so potentially legion due to traces.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclDictWithInit(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictPtr,
+    int pathc,
+    Tcl_Obj *const pathv[])
+{
+    Tcl_DictSearch s;
+    Tcl_Obj *keyPtr, *valPtr, *keysPtr;
+    int done;
+
+    if (pathc > 0) {
+	dictPtr = TclTraceDictPath(interp, dictPtr, pathc, pathv,
+		DICT_PATH_READ);
+	if (dictPtr == NULL) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * Go over the list of keys and write each corresponding value to a
+     * variable in the current context with the same name. Also keep a copy of
+     * the keys so we can write back properly later on even if the dictionary
+     * has been structurally modified.
+     */
+
+    if (Tcl_DictObjFirst(interp, dictPtr, &s, &keyPtr, &valPtr,
+	    &done) != TCL_OK) {
+	return NULL;
+    }
+
+    TclNewObj(keysPtr);
+
+    for (; !done ; Tcl_DictObjNext(&s, &keyPtr, &valPtr, &done)) {
+	Tcl_ListObjAppendElement(NULL, keysPtr, keyPtr);
+	if (Tcl_ObjSetVar2(interp, keyPtr, NULL, valPtr,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    TclDecrRefCount(keysPtr);
+	    Tcl_DictObjDone(&s);
+	    return NULL;
+	}
+    }
+
+    return keysPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDictWithFinish --
+ *
+ *	Part of the core of [dict with]. Reassembles the piece of the dict (in
+ *	varName, location given by pathc/pathv) from the variables named in
+ *	the keysPtr argument. NB, does not try to preserve errors or manage
+ *	argument lifetimes.
+ *
+ * Result:
+ *	TCL_OK if we succeeded, or TCL_ERROR if we failed.
+ *
+ * Side effects:
+ *	Assigns to a variable, so potentially legion due to traces. Updates
+ *	the dictionary in the named variable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclDictWithFinish(
+    Tcl_Interp *interp,		/* Command interpreter in which variable
+				 * exists. Used for state management, traces
+				 * and error reporting. */
+    Var *varPtr,		/* Reference to the variable holding the
+				 * dictionary. */
+    Var *arrayPtr,		/* Reference to the array containing the
+				 * variable, or NULL if the variable is a
+				 * scalar. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. NULL if the 'index'
+				 * parameter is >= 0 */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    int index,			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+    int pathc,			/* The number of elements in the path into the
+				 * dictionary. */
+    Tcl_Obj *const pathv[],	/* The elements of the path to the subdict. */
+    Tcl_Obj *keysPtr)		/* List of keys to be synchronized. This is
+				 * the result value from TclDictWithInit. */
+{
+    Tcl_Obj *dictPtr, *leafPtr, *valPtr;
+    int i, allocdict, keyc;
+    Tcl_Obj **keyv;
+
+    /*
+     * If the dictionary variable doesn't exist, drop everything silently.
+     */
+
+    dictPtr = TclPtrGetVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    TCL_LEAVE_ERR_MSG, index);
+    if (dictPtr == NULL) {
+	return TCL_OK;
+    }
+
+    /*
+     * Double-check that it is still a dictionary.
+     */
+
+    if (Tcl_DictObjSize(interp, dictPtr, &i) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (Tcl_IsShared(dictPtr)) {
+	dictPtr = Tcl_DuplicateObj(dictPtr);
+	allocdict = 1;
+    } else {
+	allocdict = 0;
+    }
+
+    if (pathc > 0) {
+	/*
+	 * Want to get to the dictionary which we will update; need to do
+	 * prepare-for-update de-sharing along the path *but* avoid generating
+	 * an error on a non-existant path (we'll treat that the same as a
+	 * non-existant variable. Luckily, the de-sharing operation isn't
+	 * deeply damaging if we don't go on to update; it's just less than
+	 * perfectly efficient (but no memory should be leaked).
+	 */
+
+	leafPtr = TclTraceDictPath(interp, dictPtr, pathc, pathv,
+		DICT_PATH_EXISTS | DICT_PATH_UPDATE);
+	if (leafPtr == NULL) {
+	    if (allocdict) {
+		TclDecrRefCount(dictPtr);
+	    }
+	    return TCL_ERROR;
+	}
+	if (leafPtr == DICT_PATH_NON_EXISTENT) {
+	    if (allocdict) {
+		TclDecrRefCount(dictPtr);
+	    }
+	    return TCL_OK;
+	}
+    } else {
+	leafPtr = dictPtr;
+    }
+
+    /*
+     * Now process our updates on the leaf dictionary.
+     */
+
+    TclListObjGetElements(NULL, keysPtr, &keyc, &keyv);
+    for (i=0 ; i<keyc ; i++) {
+	valPtr = Tcl_ObjGetVar2(interp, keyv[i], NULL, 0);
+	if (valPtr == NULL) {
+	    Tcl_DictObjRemove(NULL, leafPtr, keyv[i]);
+	} else if (leafPtr == valPtr) {
+	    /*
+	     * Someone is messing us around, trying to build a recursive
+	     * structure. [Bug 1786481]
+	     */
+
+	    Tcl_DictObjPut(NULL, leafPtr, keyv[i], Tcl_DuplicateObj(valPtr));
+	} else {
+	    Tcl_DictObjPut(NULL, leafPtr, keyv[i], valPtr);
+	}
+    }
+
+    /*
+     * Ensure that none of the dictionaries in the chain still have a string
+     * rep.
+     */
+
+    if (pathc > 0) {
+	InvalidateDictChain(leafPtr);
+    }
+
+    /*
+     * Write back the outermost dictionary to the variable.
+     */
+
+    if (TclPtrSetVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    dictPtr, TCL_LEAVE_ERR_MSG, index) == NULL) {
+	if (allocdict) {
+	    TclDecrRefCount(dictPtr);
+	}
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitDictCmd --
+ *
+ *	This function is create the "dict" Tcl command. See the user
+ *	documentation for details on what it does, and TIP#111 for the formal
+ *	specification.
+ *
+ * Results:
+ *	A Tcl command handle.
+ *
+ * Side effects:
+ *	May advance compilation epoch.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitDictCmd(
+    Tcl_Interp *interp)
+{
+    return TclMakeEnsemble(interp, "dict", implementationMap);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclDisassemble.c b/generic/tclDisassemble.c
new file mode 100644
index 0000000..d61ed42
--- /dev/null
+++ b/generic/tclDisassemble.c
@@ -0,0 +1,1630 @@
+/*
+ * tclDisassemble.c --
+ *
+ *	This file contains procedures that disassemble bytecode into either
+ *	human-readable or Tcl-processable forms.
+ *
+ * Copyright (c) 1996-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ * Copyright (c) 2013-2016 Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include "tclOOInt.h"
+#include <assert.h>
+
+/*
+ * Prototypes for procedures defined later in this file:
+ */
+
+static Tcl_Obj *	DisassembleByteCodeAsDicts(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static Tcl_Obj *	DisassembleByteCodeObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static int		FormatInstruction(ByteCode *codePtr,
+			    const unsigned char *pc, Tcl_Obj *bufferObj);
+static void		GetLocationInformation(Proc *procPtr,
+			    Tcl_Obj **fileObjPtr, int *linePtr);
+static void		PrintSourceToObj(Tcl_Obj *appendObj,
+			    const char *stringPtr, int maxChars);
+static void		UpdateStringOfInstName(Tcl_Obj *objPtr);
+
+/*
+ * The structure below defines an instruction name Tcl object to allow
+ * reporting of inner contexts in errorstack without string allocation.
+ */
+
+static const Tcl_ObjType tclInstNameType = {
+    "instname",			/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    UpdateStringOfInstName,	/* updateStringProc */
+    NULL,			/* setFromAnyProc */
+};
+
+/*
+ * How to get the bytecode out of a Tcl_Obj.
+ */
+
+#define BYTECODE(objPtr)					\
+    ((ByteCode *) (objPtr)->internalRep.twoPtrValue.ptr1)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetLocationInformation --
+ *
+ *	This procedure looks up the information about where a procedure was
+ *	originally declared.
+ *
+ * Results:
+ *	Writes to the variables pointed at by fileObjPtr and linePtr.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+GetLocationInformation(
+    Proc *procPtr,		/* What to look up the information for. */
+    Tcl_Obj **fileObjPtr,	/* Where to write the information about what
+				 * file the code came from. Will be written
+				 * to, either with the object (assume shared!)
+				 * that describes what the file was, or with
+				 * NULL if the information is not
+				 * available. */
+    int *linePtr)		/* Where to write the information about what
+				 * line number represented the start of the
+				 * code in question. Will be written to,
+				 * either with the line number or with -1 if
+				 * the information is not available. */
+{
+    CmdFrame *cfPtr = TclGetCmdFrameForProcedure(procPtr);
+
+    *fileObjPtr = NULL;
+    *linePtr = -1;
+    if (cfPtr == NULL) {
+	return;
+    }
+
+    /*
+     * Get the source location data out of the CmdFrame.
+     */
+
+    *linePtr = cfPtr->line[0];
+    if (cfPtr->type == TCL_LOCATION_SOURCE) {
+	*fileObjPtr = cfPtr->data.eval.path;
+    }
+}
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPrintByteCodeObj --
+ *
+ *	This procedure prints ("disassembles") the instructions of a bytecode
+ *	object to stdout.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclPrintByteCodeObj(
+    Tcl_Interp *interp,		/* Used only for getting location info. */
+    Tcl_Obj *objPtr)		/* The bytecode object to disassemble. */
+{
+    Tcl_Obj *bufPtr = DisassembleByteCodeObj(interp, objPtr);
+
+    fprintf(stdout, "\n%s", TclGetString(bufPtr));
+    Tcl_DecrRefCount(bufPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPrintInstruction --
+ *
+ *	This procedure prints ("disassembles") one instruction from a bytecode
+ *	object to stdout.
+ *
+ * Results:
+ *	Returns the length in bytes of the current instruiction.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPrintInstruction(
+    ByteCode *codePtr,		/* Bytecode containing the instruction. */
+    const unsigned char *pc)	/* Points to first byte of instruction. */
+{
+    Tcl_Obj *bufferObj;
+    int numBytes;
+
+    TclNewObj(bufferObj);
+    numBytes = FormatInstruction(codePtr, pc, bufferObj);
+    fprintf(stdout, "%s", TclGetString(bufferObj));
+    Tcl_DecrRefCount(bufferObj);
+    return numBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPrintObject --
+ *
+ *	This procedure prints up to a specified number of characters from the
+ *	argument Tcl object's string representation to a specified file.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Outputs characters to the specified file.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclPrintObject(
+    FILE *outFile,		/* The file to print the source to. */
+    Tcl_Obj *objPtr,		/* Points to the Tcl object whose string
+				 * representation should be printed. */
+    int maxChars)		/* Maximum number of chars to print. */
+{
+    char *bytes;
+    int length;
+
+    bytes = TclGetStringFromObj(objPtr, &length);
+    TclPrintSource(outFile, bytes, TclMin(length, maxChars));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPrintSource --
+ *
+ *	This procedure prints up to a specified number of characters from the
+ *	argument string to a specified file. It tries to produce legible
+ *	output by adding backslashes as necessary.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Outputs characters to the specified file.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclPrintSource(
+    FILE *outFile,		/* The file to print the source to. */
+    const char *stringPtr,	/* The string to print. */
+    int maxChars)		/* Maximum number of chars to print. */
+{
+    Tcl_Obj *bufferObj;
+
+    TclNewObj(bufferObj);
+    PrintSourceToObj(bufferObj, stringPtr, maxChars);
+    fprintf(outFile, "%s", TclGetString(bufferObj));
+    Tcl_DecrRefCount(bufferObj);
+}
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DisassembleByteCodeObj --
+ *
+ *	Given an object which is of bytecode type, return a disassembled
+ *	version of the bytecode (in a new refcount 0 object). No guarantees
+ *	are made about the details of the contents of the result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+DisassembleByteCodeObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)		/* The bytecode object to disassemble. */
+{
+    ByteCode *codePtr = BYTECODE(objPtr);
+    unsigned char *codeStart, *codeLimit, *pc;
+    unsigned char *codeDeltaNext, *codeLengthNext;
+    unsigned char *srcDeltaNext, *srcLengthNext;
+    int codeOffset, codeLen, srcOffset, srcLen, numCmds, delta, i, line;
+    Interp *iPtr = (Interp *) *codePtr->interpHandle;
+    Tcl_Obj *bufferObj, *fileObj;
+
+    TclNewObj(bufferObj);
+    if (codePtr->refCount <= 0) {
+	return bufferObj;	/* Already freed. */
+    }
+
+    codeStart = codePtr->codeStart;
+    codeLimit = codeStart + codePtr->numCodeBytes;
+    numCmds = codePtr->numCommands;
+
+    /*
+     * Print header lines describing the ByteCode.
+     */
+
+    Tcl_AppendPrintfToObj(bufferObj,
+	    "ByteCode %p, refCt %u, epoch %u, interp %p (epoch %u)\n",
+	    codePtr, codePtr->refCount, codePtr->compileEpoch, iPtr,
+	    iPtr->compileEpoch);
+    Tcl_AppendToObj(bufferObj, "  Source ", -1);
+    PrintSourceToObj(bufferObj, codePtr->source,
+	    TclMin(codePtr->numSrcBytes, 55));
+    GetLocationInformation(codePtr->procPtr, &fileObj, &line);
+    if (line > -1 && fileObj != NULL) {
+	Tcl_AppendPrintfToObj(bufferObj, "\n  File \"%s\" Line %d",
+		Tcl_GetString(fileObj), line);
+    }
+    Tcl_AppendPrintfToObj(bufferObj,
+	    "\n  Cmds %d, src %d, inst %d, litObjs %u, aux %d, stkDepth %u, code/src %.2f\n",
+	    numCmds, codePtr->numSrcBytes, codePtr->numCodeBytes,
+	    codePtr->numLitObjects, codePtr->numAuxDataItems,
+	    codePtr->maxStackDepth,
+#ifdef TCL_COMPILE_STATS
+	    codePtr->numSrcBytes?
+		    codePtr->structureSize/(float)codePtr->numSrcBytes :
+#endif
+	    0.0);
+
+#ifdef TCL_COMPILE_STATS
+    Tcl_AppendPrintfToObj(bufferObj,
+	    "  Code %lu = header %lu+inst %d+litObj %lu+exc %lu+aux %lu+cmdMap %d\n",
+	    (unsigned long) codePtr->structureSize,
+	    (unsigned long) (sizeof(ByteCode) - sizeof(size_t) - sizeof(Tcl_Time)),
+	    codePtr->numCodeBytes,
+	    (unsigned long) (codePtr->numLitObjects * sizeof(Tcl_Obj *)),
+	    (unsigned long) (codePtr->numExceptRanges*sizeof(ExceptionRange)),
+	    (unsigned long) (codePtr->numAuxDataItems * sizeof(AuxData)),
+	    codePtr->numCmdLocBytes);
+#endif /* TCL_COMPILE_STATS */
+
+    /*
+     * If the ByteCode is the compiled body of a Tcl procedure, print
+     * information about that procedure. Note that we don't know the
+     * procedure's name since ByteCode's can be shared among procedures.
+     */
+
+    if (codePtr->procPtr != NULL) {
+	Proc *procPtr = codePtr->procPtr;
+	int numCompiledLocals = procPtr->numCompiledLocals;
+
+	Tcl_AppendPrintfToObj(bufferObj,
+		"  Proc %p, refCt %d, args %d, compiled locals %d\n",
+		procPtr, procPtr->refCount, procPtr->numArgs,
+		numCompiledLocals);
+	if (numCompiledLocals > 0) {
+	    CompiledLocal *localPtr = procPtr->firstLocalPtr;
+
+	    for (i = 0;  i < numCompiledLocals;  i++) {
+		Tcl_AppendPrintfToObj(bufferObj,
+			"      slot %d%s%s%s%s%s%s", i,
+			(localPtr->flags & (VAR_ARRAY|VAR_LINK)) ? "" : ", scalar",
+			(localPtr->flags & VAR_ARRAY) ? ", array" : "",
+			(localPtr->flags & VAR_LINK) ? ", link" : "",
+			(localPtr->flags & VAR_ARGUMENT) ? ", arg" : "",
+			(localPtr->flags & VAR_TEMPORARY) ? ", temp" : "",
+			(localPtr->flags & VAR_RESOLVED) ? ", resolved" : "");
+		if (TclIsVarTemporary(localPtr)) {
+		    Tcl_AppendToObj(bufferObj, "\n", -1);
+		} else {
+		    Tcl_AppendPrintfToObj(bufferObj, ", \"%s\"\n",
+			    localPtr->name);
+		}
+		localPtr = localPtr->nextPtr;
+	    }
+	}
+    }
+
+    /*
+     * Print the ExceptionRange array.
+     */
+
+    if (codePtr->numExceptRanges > 0) {
+	Tcl_AppendPrintfToObj(bufferObj, "  Exception ranges %d, depth %d:\n",
+		codePtr->numExceptRanges, codePtr->maxExceptDepth);
+	for (i = 0;  i < codePtr->numExceptRanges;  i++) {
+	    ExceptionRange *rangePtr = &codePtr->exceptArrayPtr[i];
+
+	    Tcl_AppendPrintfToObj(bufferObj,
+		    "      %d: level %d, %s, pc %d-%d, ",
+		    i, rangePtr->nestingLevel,
+		    (rangePtr->type==LOOP_EXCEPTION_RANGE ? "loop" : "catch"),
+		    rangePtr->codeOffset,
+		    (rangePtr->codeOffset + rangePtr->numCodeBytes - 1));
+	    switch (rangePtr->type) {
+	    case LOOP_EXCEPTION_RANGE:
+		Tcl_AppendPrintfToObj(bufferObj, "continue %d, break %d\n",
+			rangePtr->continueOffset, rangePtr->breakOffset);
+		break;
+	    case CATCH_EXCEPTION_RANGE:
+		Tcl_AppendPrintfToObj(bufferObj, "catch %d\n",
+			rangePtr->catchOffset);
+		break;
+	    default:
+		Tcl_Panic("DisassembleByteCodeObj: bad ExceptionRange type %d",
+			rangePtr->type);
+	    }
+	}
+    }
+
+    /*
+     * If there were no commands (e.g., an expression or an empty string was
+     * compiled), just print all instructions and return.
+     */
+
+    if (numCmds == 0) {
+	pc = codeStart;
+	while (pc < codeLimit) {
+	    Tcl_AppendToObj(bufferObj, "    ", -1);
+	    pc += FormatInstruction(codePtr, pc, bufferObj);
+	}
+	return bufferObj;
+    }
+
+    /*
+     * Print table showing the code offset, source offset, and source length
+     * for each command. These are encoded as a sequence of bytes.
+     */
+
+    Tcl_AppendPrintfToObj(bufferObj, "  Commands %d:", numCmds);
+    codeDeltaNext = codePtr->codeDeltaStart;
+    codeLengthNext = codePtr->codeLengthStart;
+    srcDeltaNext = codePtr->srcDeltaStart;
+    srcLengthNext = codePtr->srcLengthStart;
+    codeOffset = srcOffset = 0;
+    for (i = 0;  i < numCmds;  i++) {
+	if ((unsigned) *codeDeltaNext == (unsigned) 0xFF) {
+	    codeDeltaNext++;
+	    delta = TclGetInt4AtPtr(codeDeltaNext);
+	    codeDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(codeDeltaNext);
+	    codeDeltaNext++;
+	}
+	codeOffset += delta;
+
+	if ((unsigned) *codeLengthNext == (unsigned) 0xFF) {
+	    codeLengthNext++;
+	    codeLen = TclGetInt4AtPtr(codeLengthNext);
+	    codeLengthNext += 4;
+	} else {
+	    codeLen = TclGetInt1AtPtr(codeLengthNext);
+	    codeLengthNext++;
+	}
+
+	if ((unsigned) *srcDeltaNext == (unsigned) 0xFF) {
+	    srcDeltaNext++;
+	    delta = TclGetInt4AtPtr(srcDeltaNext);
+	    srcDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(srcDeltaNext);
+	    srcDeltaNext++;
+	}
+	srcOffset += delta;
+
+	if ((unsigned) *srcLengthNext == (unsigned) 0xFF) {
+	    srcLengthNext++;
+	    srcLen = TclGetInt4AtPtr(srcLengthNext);
+	    srcLengthNext += 4;
+	} else {
+	    srcLen = TclGetInt1AtPtr(srcLengthNext);
+	    srcLengthNext++;
+	}
+
+	Tcl_AppendPrintfToObj(bufferObj, "%s%4d: pc %d-%d, src %d-%d",
+		((i % 2)? "     " : "\n   "),
+		(i+1), codeOffset, (codeOffset + codeLen - 1),
+		srcOffset, (srcOffset + srcLen - 1));
+    }
+    if (numCmds > 0) {
+	Tcl_AppendToObj(bufferObj, "\n", -1);
+    }
+
+    /*
+     * Print each instruction. If the instruction corresponds to the start of
+     * a command, print the command's source. Note that we don't need the code
+     * length here.
+     */
+
+    codeDeltaNext = codePtr->codeDeltaStart;
+    srcDeltaNext = codePtr->srcDeltaStart;
+    srcLengthNext = codePtr->srcLengthStart;
+    codeOffset = srcOffset = 0;
+    pc = codeStart;
+    for (i = 0;  i < numCmds;  i++) {
+	if ((unsigned) *codeDeltaNext == (unsigned) 0xFF) {
+	    codeDeltaNext++;
+	    delta = TclGetInt4AtPtr(codeDeltaNext);
+	    codeDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(codeDeltaNext);
+	    codeDeltaNext++;
+	}
+	codeOffset += delta;
+
+	if ((unsigned) *srcDeltaNext == (unsigned) 0xFF) {
+	    srcDeltaNext++;
+	    delta = TclGetInt4AtPtr(srcDeltaNext);
+	    srcDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(srcDeltaNext);
+	    srcDeltaNext++;
+	}
+	srcOffset += delta;
+
+	if ((unsigned) *srcLengthNext == (unsigned) 0xFF) {
+	    srcLengthNext++;
+	    srcLen = TclGetInt4AtPtr(srcLengthNext);
+	    srcLengthNext += 4;
+	} else {
+	    srcLen = TclGetInt1AtPtr(srcLengthNext);
+	    srcLengthNext++;
+	}
+
+	/*
+	 * Print instructions before command i.
+	 */
+
+	while ((pc-codeStart) < codeOffset) {
+	    Tcl_AppendToObj(bufferObj, "    ", -1);
+	    pc += FormatInstruction(codePtr, pc, bufferObj);
+	}
+
+	Tcl_AppendPrintfToObj(bufferObj, "  Command %d: ", i+1);
+	PrintSourceToObj(bufferObj, (codePtr->source + srcOffset),
+		TclMin(srcLen, 55));
+	Tcl_AppendToObj(bufferObj, "\n", -1);
+    }
+    if (pc < codeLimit) {
+	/*
+	 * Print instructions after the last command.
+	 */
+
+	while (pc < codeLimit) {
+	    Tcl_AppendToObj(bufferObj, "    ", -1);
+	    pc += FormatInstruction(codePtr, pc, bufferObj);
+	}
+    }
+    return bufferObj;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FormatInstruction --
+ *
+ *	Appends a representation of a bytecode instruction to a Tcl_Obj.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FormatInstruction(
+    ByteCode *codePtr,		/* Bytecode containing the instruction. */
+    const unsigned char *pc,	/* Points to first byte of instruction. */
+    Tcl_Obj *bufferObj)		/* Object to append instruction info to. */
+{
+    Proc *procPtr = codePtr->procPtr;
+    unsigned char opCode = *pc;
+    register const InstructionDesc *instDesc = &tclInstructionTable[opCode];
+    unsigned char *codeStart = codePtr->codeStart;
+    unsigned pcOffset = pc - codeStart;
+    int opnd = 0, i, j, numBytes = 1;
+    int localCt = procPtr ? procPtr->numCompiledLocals : 0;
+    CompiledLocal *localPtr = procPtr ? procPtr->firstLocalPtr : NULL;
+    char suffixBuffer[128];	/* Additional info to print after main opcode
+				 * and immediates. */
+    char *suffixSrc = NULL;
+    Tcl_Obj *suffixObj = NULL;
+    AuxData *auxPtr = NULL;
+
+    suffixBuffer[0] = '\0';
+    Tcl_AppendPrintfToObj(bufferObj, "(%u) %s ", pcOffset, instDesc->name);
+    for (i = 0;  i < instDesc->numOperands;  i++) {
+	switch (instDesc->opTypes[i]) {
+	case OPERAND_INT1:
+	    opnd = TclGetInt1AtPtr(pc+numBytes); numBytes++;
+	    Tcl_AppendPrintfToObj(bufferObj, "%+d ", opnd);
+	    break;
+	case OPERAND_INT4:
+	    opnd = TclGetInt4AtPtr(pc+numBytes); numBytes += 4;
+	    Tcl_AppendPrintfToObj(bufferObj, "%+d ", opnd);
+	    break;
+	case OPERAND_UINT1:
+	    opnd = TclGetUInt1AtPtr(pc+numBytes); numBytes++;
+	    Tcl_AppendPrintfToObj(bufferObj, "%u ", (unsigned) opnd);
+	    break;
+	case OPERAND_UINT4:
+	    opnd = TclGetUInt4AtPtr(pc+numBytes); numBytes += 4;
+	    if (opCode == INST_START_CMD) {
+		sprintf(suffixBuffer+strlen(suffixBuffer),
+			", %u cmds start here", opnd);
+	    }
+	    Tcl_AppendPrintfToObj(bufferObj, "%u ", (unsigned) opnd);
+	    break;
+	case OPERAND_OFFSET1:
+	    opnd = TclGetInt1AtPtr(pc+numBytes); numBytes++;
+	    sprintf(suffixBuffer, "pc %u", pcOffset+opnd);
+	    Tcl_AppendPrintfToObj(bufferObj, "%+d ", opnd);
+	    break;
+	case OPERAND_OFFSET4:
+	    opnd = TclGetInt4AtPtr(pc+numBytes); numBytes += 4;
+	    if (opCode == INST_START_CMD) {
+		sprintf(suffixBuffer, "next cmd at pc %u", pcOffset+opnd);
+	    } else {
+		sprintf(suffixBuffer, "pc %u", pcOffset+opnd);
+	    }
+	    Tcl_AppendPrintfToObj(bufferObj, "%+d ", opnd);
+	    break;
+	case OPERAND_LIT1:
+	    opnd = TclGetUInt1AtPtr(pc+numBytes); numBytes++;
+	    suffixObj = codePtr->objArrayPtr[opnd];
+	    Tcl_AppendPrintfToObj(bufferObj, "%u ", (unsigned) opnd);
+	    break;
+	case OPERAND_LIT4:
+	    opnd = TclGetUInt4AtPtr(pc+numBytes); numBytes += 4;
+	    suffixObj = codePtr->objArrayPtr[opnd];
+	    Tcl_AppendPrintfToObj(bufferObj, "%u ", (unsigned) opnd);
+	    break;
+	case OPERAND_AUX4:
+	    opnd = TclGetUInt4AtPtr(pc+numBytes); numBytes += 4;
+	    Tcl_AppendPrintfToObj(bufferObj, "%u ", (unsigned) opnd);
+	    auxPtr = &codePtr->auxDataArrayPtr[opnd];
+	    break;
+	case OPERAND_IDX4:
+	    opnd = TclGetInt4AtPtr(pc+numBytes); numBytes += 4;
+	    if (opnd >= -1) {
+		Tcl_AppendPrintfToObj(bufferObj, "%d ", opnd);
+	    } else if (opnd == -2) {
+		Tcl_AppendPrintfToObj(bufferObj, "end ");
+	    } else {
+		Tcl_AppendPrintfToObj(bufferObj, "end-%d ", -2-opnd);
+	    }
+	    break;
+	case OPERAND_LVT1:
+	    opnd = TclGetUInt1AtPtr(pc+numBytes);
+	    numBytes++;
+	    goto printLVTindex;
+	case OPERAND_LVT4:
+	    opnd = TclGetUInt4AtPtr(pc+numBytes);
+	    numBytes += 4;
+	printLVTindex:
+	    if (localPtr != NULL) {
+		if (opnd >= localCt) {
+		    Tcl_Panic("FormatInstruction: bad local var index %u (%u locals)",
+			    (unsigned) opnd, localCt);
+		}
+		for (j = 0;  j < opnd;  j++) {
+		    localPtr = localPtr->nextPtr;
+		}
+		if (TclIsVarTemporary(localPtr)) {
+		    sprintf(suffixBuffer, "temp var %u", (unsigned) opnd);
+		} else {
+		    sprintf(suffixBuffer, "var ");
+		    suffixSrc = localPtr->name;
+		}
+	    }
+	    Tcl_AppendPrintfToObj(bufferObj, "%%v%u ", (unsigned) opnd);
+	    break;
+	case OPERAND_SCLS1:
+	    opnd = TclGetUInt1AtPtr(pc+numBytes); numBytes++;
+	    Tcl_AppendPrintfToObj(bufferObj, "%s ",
+		    tclStringClassTable[opnd].name);
+	    break;
+	case OPERAND_NONE:
+	default:
+	    break;
+	}
+    }
+    if (suffixObj) {
+	const char *bytes;
+	int length;
+
+	Tcl_AppendToObj(bufferObj, "\t# ", -1);
+	bytes = TclGetStringFromObj(codePtr->objArrayPtr[opnd], &length);
+	PrintSourceToObj(bufferObj, bytes, TclMin(length, 40));
+    } else if (suffixBuffer[0]) {
+	Tcl_AppendPrintfToObj(bufferObj, "\t# %s", suffixBuffer);
+	if (suffixSrc) {
+	    PrintSourceToObj(bufferObj, suffixSrc, 40);
+	}
+    }
+    Tcl_AppendToObj(bufferObj, "\n", -1);
+    if (auxPtr && auxPtr->type->printProc) {
+	Tcl_AppendToObj(bufferObj, "\t\t[", -1);
+	auxPtr->type->printProc(auxPtr->clientData, bufferObj, codePtr,
+		pcOffset);
+	Tcl_AppendToObj(bufferObj, "]\n", -1);
+    }
+    return numBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetInnerContext --
+ *
+ *	If possible, returns a list capturing the inner context. Otherwise
+ *	return NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetInnerContext(
+    Tcl_Interp *interp,
+    const unsigned char *pc,
+    Tcl_Obj **tosPtr)
+{
+    int objc = 0, off = 0;
+    Tcl_Obj *result;
+    Interp *iPtr = (Interp *) interp;
+
+    switch (*pc) {
+    case INST_STR_LEN:
+    case INST_LNOT:
+    case INST_BITNOT:
+    case INST_UMINUS:
+    case INST_UPLUS:
+    case INST_TRY_CVT_TO_NUMERIC:
+    case INST_EXPAND_STKTOP:
+    case INST_EXPR_STK:
+        objc = 1;
+        break;
+
+    case INST_LIST_IN:
+    case INST_LIST_NOT_IN:	/* Basic list containment operators. */
+    case INST_STR_EQ:
+    case INST_STR_NEQ:		/* String (in)equality check */
+    case INST_STR_CMP:		/* String compare. */
+    case INST_STR_INDEX:
+    case INST_STR_MATCH:
+    case INST_REGEXP:
+    case INST_EQ:
+    case INST_NEQ:
+    case INST_LT:
+    case INST_GT:
+    case INST_LE:
+    case INST_GE:
+    case INST_MOD:
+    case INST_LSHIFT:
+    case INST_RSHIFT:
+    case INST_BITOR:
+    case INST_BITXOR:
+    case INST_BITAND:
+    case INST_EXPON:
+    case INST_ADD:
+    case INST_SUB:
+    case INST_DIV:
+    case INST_MULT:
+        objc = 2;
+        break;
+
+    case INST_RETURN_STK:
+        /* early pop. TODO: dig out opt dict too :/ */
+        objc = 1;
+        break;
+
+    case INST_SYNTAX:
+    case INST_RETURN_IMM:
+        objc = 2;
+        break;
+
+    case INST_INVOKE_STK4:
+	objc = TclGetUInt4AtPtr(pc+1);
+        break;
+
+    case INST_INVOKE_STK1:
+	objc = TclGetUInt1AtPtr(pc+1);
+	break;
+    }
+
+    result = iPtr->innerContext;
+    if (Tcl_IsShared(result)) {
+        Tcl_DecrRefCount(result);
+        iPtr->innerContext = result = Tcl_NewListObj(objc + 1, NULL);
+        Tcl_IncrRefCount(result);
+    } else {
+        int len;
+
+        /*
+         * Reset while keeping the list intrep as much as possible.
+         */
+
+	Tcl_ListObjLength(interp, result, &len);
+        Tcl_ListObjReplace(interp, result, 0, len, 0, NULL);
+    }
+    Tcl_ListObjAppendElement(NULL, result, TclNewInstNameObj(*pc));
+
+    for (; objc>0 ; objc--) {
+        Tcl_Obj *objPtr;
+
+        objPtr = tosPtr[1 - objc + off];
+        if (!objPtr) {
+            Tcl_Panic("InnerContext: bad tos -- appending null object");
+        }
+        if ((objPtr->refCount<=0)
+#ifdef TCL_MEM_DEBUG
+                || (objPtr->refCount==0x61616161)
+#endif
+        ) {
+            Tcl_Panic("InnerContext: bad tos -- appending freed object %p",
+                    objPtr);
+        }
+        Tcl_ListObjAppendElement(NULL, result, objPtr);
+    }
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNewInstNameObj --
+ *
+ *	Creates a new InstName Tcl_Obj based on the given instruction
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclNewInstNameObj(
+    unsigned char inst)
+{
+    Tcl_Obj *objPtr = Tcl_NewObj();
+
+    objPtr->typePtr = &tclInstNameType;
+    objPtr->internalRep.longValue = (long) inst;
+    objPtr->bytes = NULL;
+
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfInstName --
+ *
+ *	Update the string representation for an instruction name object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfInstName(
+    Tcl_Obj *objPtr)
+{
+    int inst = objPtr->internalRep.longValue;
+    char *s, buf[20];
+    int len;
+
+    if ((inst < 0) || (inst > LAST_INST_OPCODE)) {
+        sprintf(buf, "inst_%d", inst);
+        s = buf;
+    } else {
+        s = (char *) tclInstructionTable[objPtr->internalRep.longValue].name;
+    }
+    len = strlen(s);
+    objPtr->bytes = ckalloc(len + 1);
+    memcpy(objPtr->bytes, s, len + 1);
+    objPtr->length = len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PrintSourceToObj --
+ *
+ *	Appends a quoted representation of a string to a Tcl_Obj.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PrintSourceToObj(
+    Tcl_Obj *appendObj,		/* The object to print the source to. */
+    const char *stringPtr,	/* The string to print. */
+    int maxChars)		/* Maximum number of chars to print. */
+{
+    register const char *p;
+    register int i = 0, len;
+    Tcl_UniChar ch = 0;
+
+    if (stringPtr == NULL) {
+	Tcl_AppendToObj(appendObj, "\"\"", -1);
+	return;
+    }
+
+    Tcl_AppendToObj(appendObj, "\"", -1);
+    p = stringPtr;
+    for (;  (*p != '\0') && (i < maxChars);  p+=len) {
+
+	len = TclUtfToUniChar(p, &ch);
+	switch (ch) {
+	case '"':
+	    Tcl_AppendToObj(appendObj, "\\\"", -1);
+	    i += 2;
+	    continue;
+	case '\f':
+	    Tcl_AppendToObj(appendObj, "\\f", -1);
+	    i += 2;
+	    continue;
+	case '\n':
+	    Tcl_AppendToObj(appendObj, "\\n", -1);
+	    i += 2;
+	    continue;
+	case '\r':
+	    Tcl_AppendToObj(appendObj, "\\r", -1);
+	    i += 2;
+	    continue;
+	case '\t':
+	    Tcl_AppendToObj(appendObj, "\\t", -1);
+	    i += 2;
+	    continue;
+	case '\v':
+	    Tcl_AppendToObj(appendObj, "\\v", -1);
+	    i += 2;
+	    continue;
+	default:
+#if TCL_UTF_MAX > 4
+	    if (ch > 0xffff) {
+		Tcl_AppendPrintfToObj(appendObj, "\\U%08x", ch);
+		i += 10;
+	    } else
+#elif TCL_UTF_MAX > 3
+	    /* If len == 0, this means we have a char > 0xffff, resulting in
+	     * TclUtfToUniChar producing a surrogate pair. We want to output
+	     * this pair as a single Unicode character.
+	     */
+	    if (len == 0) {
+		int upper = ((ch & 0x3ff) + 1) << 10;
+		len = TclUtfToUniChar(p, &ch);
+		Tcl_AppendPrintfToObj(appendObj, "\\U%08x", upper + (ch & 0x3ff));
+		i += 10;
+	    } else
+#endif
+	    if (ch < 0x20 || ch >= 0x7f) {
+		Tcl_AppendPrintfToObj(appendObj, "\\u%04x", ch);
+		i += 6;
+	    } else {
+		Tcl_AppendPrintfToObj(appendObj, "%c", ch);
+		i++;
+	    }
+	    continue;
+	}
+    }
+    if (*p != '\0') {
+	Tcl_AppendToObj(appendObj, "...", -1);
+    }
+    Tcl_AppendToObj(appendObj, "\"", -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DisassembleByteCodeAsDicts --
+ *
+ *	Given an object which is of bytecode type, return a disassembled
+ *	version of the bytecode (in a new refcount 0 object) in a dictionary.
+ *	No guarantees are made about the details of the contents of the
+ *	result, but it is intended to be more readable than the old output
+ *	format.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+DisassembleByteCodeAsDicts(
+    Tcl_Interp *interp,		/* Used for looking up the CmdFrame for the
+				 * procedure, if one exists. */
+    Tcl_Obj *objPtr)		/* The bytecode-holding value to take apart */
+{
+    ByteCode *codePtr = BYTECODE(objPtr);
+    Tcl_Obj *description, *literals, *variables, *instructions, *inst;
+    Tcl_Obj *aux, *exn, *commands, *file;
+    unsigned char *pc, *opnd, *codeOffPtr, *codeLenPtr, *srcOffPtr, *srcLenPtr;
+    int codeOffset, codeLength, sourceOffset, sourceLength;
+    int i, val, line;
+
+    /*
+     * Get the literals from the bytecode.
+     */
+
+    literals = Tcl_NewObj();
+    for (i=0 ; i<codePtr->numLitObjects ; i++) {
+	Tcl_ListObjAppendElement(NULL, literals, codePtr->objArrayPtr[i]);
+    }
+
+    /*
+     * Get the variables from the bytecode.
+     */
+
+    variables = Tcl_NewObj();
+    if (codePtr->procPtr) {
+	int localCount = codePtr->procPtr->numCompiledLocals;
+	CompiledLocal *localPtr = codePtr->procPtr->firstLocalPtr;
+
+	for (i=0 ; i<localCount ; i++,localPtr=localPtr->nextPtr) {
+	    Tcl_Obj *descriptor[2];
+
+	    descriptor[0] = Tcl_NewObj();
+	    if (!(localPtr->flags & (VAR_ARRAY|VAR_LINK))) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("scalar", -1));
+	    }
+	    if (localPtr->flags & VAR_ARRAY) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("array", -1));
+	    }
+	    if (localPtr->flags & VAR_LINK) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("link", -1));
+	    }
+	    if (localPtr->flags & VAR_ARGUMENT) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("arg", -1));
+	    }
+	    if (localPtr->flags & VAR_TEMPORARY) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("temp", -1));
+	    }
+	    if (localPtr->flags & VAR_RESOLVED) {
+		Tcl_ListObjAppendElement(NULL, descriptor[0],
+			Tcl_NewStringObj("resolved", -1));
+	    }
+	    if (localPtr->flags & VAR_TEMPORARY) {
+		Tcl_ListObjAppendElement(NULL, variables,
+			Tcl_NewListObj(1, descriptor));
+	    } else {
+		descriptor[1] = Tcl_NewStringObj(localPtr->name, -1);
+		Tcl_ListObjAppendElement(NULL, variables,
+			Tcl_NewListObj(2, descriptor));
+	    }
+	}
+    }
+
+    /*
+     * Get the instructions from the bytecode.
+     */
+
+    instructions = Tcl_NewObj();
+    for (pc=codePtr->codeStart; pc<codePtr->codeStart+codePtr->numCodeBytes;){
+	const InstructionDesc *instDesc = &tclInstructionTable[*pc];
+	int address = pc - codePtr->codeStart;
+
+	inst = Tcl_NewObj();
+	Tcl_ListObjAppendElement(NULL, inst, Tcl_NewStringObj(
+		instDesc->name, -1));
+	opnd = pc + 1;
+	for (i=0 ; i<instDesc->numOperands ; i++) {
+	    switch (instDesc->opTypes[i]) {
+	    case OPERAND_INT1:
+		val = TclGetInt1AtPtr(opnd);
+		opnd += 1;
+		goto formatNumber;
+	    case OPERAND_UINT1:
+		val = TclGetUInt1AtPtr(opnd);
+		opnd += 1;
+		goto formatNumber;
+	    case OPERAND_INT4:
+		val = TclGetInt4AtPtr(opnd);
+		opnd += 4;
+		goto formatNumber;
+	    case OPERAND_UINT4:
+		val = TclGetUInt4AtPtr(opnd);
+		opnd += 4;
+	    formatNumber:
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_NewIntObj(val));
+		break;
+
+	    case OPERAND_OFFSET1:
+		val = TclGetInt1AtPtr(opnd);
+		opnd += 1;
+		goto formatAddress;
+	    case OPERAND_OFFSET4:
+		val = TclGetInt4AtPtr(opnd);
+		opnd += 4;
+	    formatAddress:
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			"pc %d", address + val));
+		break;
+
+	    case OPERAND_LIT1:
+		val = TclGetUInt1AtPtr(opnd);
+		opnd += 1;
+		goto formatLiteral;
+	    case OPERAND_LIT4:
+		val = TclGetUInt4AtPtr(opnd);
+		opnd += 4;
+	    formatLiteral:
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			"@%d", val));
+		break;
+
+	    case OPERAND_LVT1:
+		val = TclGetUInt1AtPtr(opnd);
+		opnd += 1;
+		goto formatVariable;
+	    case OPERAND_LVT4:
+		val = TclGetUInt4AtPtr(opnd);
+		opnd += 4;
+	    formatVariable:
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			"%%%d", val));
+		break;
+	    case OPERAND_IDX4:
+		val = TclGetInt4AtPtr(opnd);
+		opnd += 4;
+		if (val >= -1) {
+		    Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			    ".%d", val));
+		} else if (val == -2) {
+		    Tcl_ListObjAppendElement(NULL, inst, Tcl_NewStringObj(
+			    ".end", -1));
+		} else {
+		    Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			    ".end-%d", -2-val));
+		}
+		break;
+	    case OPERAND_AUX4:
+		val = TclGetInt4AtPtr(opnd);
+		opnd += 4;
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			"?%d", val));
+		break;
+	    case OPERAND_SCLS1:
+		val = TclGetUInt1AtPtr(opnd);
+		opnd++;
+		Tcl_ListObjAppendElement(NULL, inst, Tcl_ObjPrintf(
+			"=%s", tclStringClassTable[val].name));
+		break;
+	    case OPERAND_NONE:
+		Tcl_Panic("opcode %d with more than zero 'no' operands", *pc);
+	    }
+	}
+	Tcl_DictObjPut(NULL, instructions, Tcl_NewIntObj(address), inst);
+	pc += instDesc->numBytes;
+    }
+
+    /*
+     * Get the auxiliary data from the bytecode.
+     */
+
+    aux = Tcl_NewObj();
+    for (i=0 ; i<codePtr->numAuxDataItems ; i++) {
+	AuxData *auxData = &codePtr->auxDataArrayPtr[i];
+	Tcl_Obj *auxDesc = Tcl_NewStringObj(auxData->type->name, -1);
+
+	if (auxData->type->disassembleProc) {
+	    Tcl_Obj *desc = Tcl_NewObj();
+
+	    Tcl_DictObjPut(NULL, desc, Tcl_NewStringObj("name", -1), auxDesc);
+	    auxDesc = desc;
+	    auxData->type->disassembleProc(auxData->clientData, auxDesc,
+		    codePtr, 0);
+	} else if (auxData->type->printProc) {
+	    Tcl_Obj *desc = Tcl_NewObj();
+
+	    auxData->type->printProc(auxData->clientData, desc, codePtr, 0);
+	    Tcl_ListObjAppendElement(NULL, auxDesc, desc);
+	}
+	Tcl_ListObjAppendElement(NULL, aux, auxDesc);
+    }
+
+    /*
+     * Get the exception ranges from the bytecode.
+     */
+
+    exn = Tcl_NewObj();
+    for (i=0 ; i<codePtr->numExceptRanges ; i++) {
+	ExceptionRange *rangePtr = &codePtr->exceptArrayPtr[i];
+
+	switch (rangePtr->type) {
+	case LOOP_EXCEPTION_RANGE:
+	    Tcl_ListObjAppendElement(NULL, exn, Tcl_ObjPrintf(
+		    "type %s level %d from %d to %d break %d continue %d",
+		    "loop", rangePtr->nestingLevel, rangePtr->codeOffset,
+		    rangePtr->codeOffset + rangePtr->numCodeBytes - 1,
+		    rangePtr->breakOffset, rangePtr->continueOffset));
+	    break;
+	case CATCH_EXCEPTION_RANGE:
+	    Tcl_ListObjAppendElement(NULL, exn, Tcl_ObjPrintf(
+		    "type %s level %d from %d to %d catch %d",
+		    "catch", rangePtr->nestingLevel, rangePtr->codeOffset,
+		    rangePtr->codeOffset + rangePtr->numCodeBytes - 1,
+		    rangePtr->catchOffset));
+	    break;
+	}
+    }
+
+    /*
+     * Get the command information from the bytecode.
+     *
+     * The way these are encoded in the bytecode is non-trivial; the Decode
+     * macro (which updates its argument and returns the next decoded value)
+     * handles this so that the rest of the code does not.
+     */
+
+#define Decode(ptr) \
+    ((TclGetUInt1AtPtr(ptr) == 0xFF)			\
+	? ((ptr)+=5 , TclGetInt4AtPtr((ptr)-4))		\
+	: ((ptr)+=1 , TclGetInt1AtPtr((ptr)-1)))
+
+    commands = Tcl_NewObj();
+    codeOffPtr = codePtr->codeDeltaStart;
+    codeLenPtr = codePtr->codeLengthStart;
+    srcOffPtr = codePtr->srcDeltaStart;
+    srcLenPtr = codePtr->srcLengthStart;
+    codeOffset = sourceOffset = 0;
+    for (i=0 ; i<codePtr->numCommands ; i++) {
+	Tcl_Obj *cmd;
+
+	codeOffset += Decode(codeOffPtr);
+	codeLength = Decode(codeLenPtr);
+	sourceOffset += Decode(srcOffPtr);
+	sourceLength = Decode(srcLenPtr);
+	cmd = Tcl_NewObj();
+	Tcl_DictObjPut(NULL, cmd, Tcl_NewStringObj("codefrom", -1),
+		Tcl_NewIntObj(codeOffset));
+	Tcl_DictObjPut(NULL, cmd, Tcl_NewStringObj("codeto", -1),
+		Tcl_NewIntObj(codeOffset + codeLength - 1));
+
+	/*
+	 * Convert byte offsets to character offsets; important if multibyte
+	 * characters are present in the source!
+	 */
+
+	Tcl_DictObjPut(NULL, cmd, Tcl_NewStringObj("scriptfrom", -1),
+		Tcl_NewIntObj(Tcl_NumUtfChars(codePtr->source,
+			sourceOffset)));
+	Tcl_DictObjPut(NULL, cmd, Tcl_NewStringObj("scriptto", -1),
+		Tcl_NewIntObj(Tcl_NumUtfChars(codePtr->source,
+			sourceOffset + sourceLength - 1)));
+	Tcl_DictObjPut(NULL, cmd, Tcl_NewStringObj("script", -1),
+		Tcl_NewStringObj(codePtr->source+sourceOffset, sourceLength));
+	Tcl_ListObjAppendElement(NULL, commands, cmd);
+    }
+
+#undef Decode
+
+    /*
+     * Get the source file and line number information from the CmdFrame
+     * system if it is available.
+     */
+
+    GetLocationInformation(codePtr->procPtr, &file, &line);
+
+    /*
+     * Build the overall result.
+     */
+
+    description = Tcl_NewObj();
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("literals", -1),
+	    literals);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("variables", -1),
+	    variables);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("exception", -1), exn);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("instructions", -1),
+	    instructions);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("auxiliary", -1), aux);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("commands", -1),
+	    commands);
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("script", -1),
+	    Tcl_NewStringObj(codePtr->source, codePtr->numSrcBytes));
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("namespace", -1),
+	    Tcl_NewStringObj(codePtr->nsPtr->fullName, -1));
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("stackdepth", -1),
+	    Tcl_NewIntObj(codePtr->maxStackDepth));
+    Tcl_DictObjPut(NULL, description, Tcl_NewStringObj("exceptdepth", -1),
+	    Tcl_NewIntObj(codePtr->maxExceptDepth));
+    if (line > -1) {
+	Tcl_DictObjPut(NULL, description,
+		Tcl_NewStringObj("initiallinenumber", -1),
+		Tcl_NewIntObj(line));
+    }
+    if (file) {
+	Tcl_DictObjPut(NULL, description,
+		Tcl_NewStringObj("sourcefile", -1), file);
+    }
+    return description;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DisassembleObjCmd --
+ *
+ *	Implementation of the "::tcl::unsupported::disassemble" command. This
+ *	command is not documented, but will disassemble procedures, lambda
+ *	terms and general scripts. Note that will compile terms if necessary
+ *	in order to disassemble them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DisassembleObjCmd(
+    ClientData clientData,	/* What type of operation. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const types[] = {
+	"constructor", "destructor",
+	"lambda", "method", "objmethod", "proc", "script", NULL
+    };
+    enum Types {
+	DISAS_CLASS_CONSTRUCTOR, DISAS_CLASS_DESTRUCTOR,
+	DISAS_LAMBDA, DISAS_CLASS_METHOD, DISAS_OBJECT_METHOD, DISAS_PROC,
+	DISAS_SCRIPT
+    };
+    int idx, result;
+    Tcl_Obj *codeObjPtr = NULL;
+    Proc *procPtr = NULL;
+    Tcl_HashEntry *hPtr;
+    Object *oPtr;
+    Method *methodPtr;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "type ...");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], types, "type", 0, &idx)!=TCL_OK){
+	return TCL_ERROR;
+    }
+
+    switch ((enum Types) idx) {
+    case DISAS_LAMBDA: {
+	Command cmd;
+	Tcl_Obj *nsObjPtr;
+	Tcl_Namespace *nsPtr;
+
+	/*
+	 * Compile (if uncompiled) and disassemble a lambda term.
+	 *
+	 * WARNING! Pokes inside the lambda objtype.
+	 */
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "lambdaTerm");
+	    return TCL_ERROR;
+	}
+	if (objv[2]->typePtr == &tclLambdaType) {
+	    procPtr = objv[2]->internalRep.twoPtrValue.ptr1;
+	}
+	if (procPtr == NULL || procPtr->iPtr != (Interp *) interp) {
+	    result = tclLambdaType.setFromAnyProc(interp, objv[2]);
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	    procPtr = objv[2]->internalRep.twoPtrValue.ptr1;
+	}
+
+	memset(&cmd, 0, sizeof(Command));
+	nsObjPtr = objv[2]->internalRep.twoPtrValue.ptr2;
+	result = TclGetNamespaceFromObj(interp, nsObjPtr, &nsPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	cmd.nsPtr = (Namespace *) nsPtr;
+	procPtr->cmdPtr = &cmd;
+	result = TclPushProcCallFrame(procPtr, interp, objc, objv, 1);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	TclPopStackFrame(interp);
+	codeObjPtr = procPtr->bodyPtr;
+	break;
+    }
+    case DISAS_PROC:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "procName");
+	    return TCL_ERROR;
+	}
+
+	procPtr = TclFindProc((Interp *) interp, TclGetString(objv[2]));
+	if (procPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" isn't a procedure", TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PROC",
+		    TclGetString(objv[2]), NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Compile (if uncompiled) and disassemble a procedure.
+	 */
+
+	result = TclPushProcCallFrame(procPtr, interp, 2, objv+1, 1);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	TclPopStackFrame(interp);
+	codeObjPtr = procPtr->bodyPtr;
+	break;
+    case DISAS_SCRIPT:
+	/*
+	 * Compile and disassemble a script.
+	 */
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "script");
+	    return TCL_ERROR;
+	}
+	if ((objv[2]->typePtr != &tclByteCodeType)
+		&& (TclSetByteCodeFromAny(interp, objv[2], NULL, NULL) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+	codeObjPtr = objv[2];
+	break;
+
+    case DISAS_CLASS_CONSTRUCTOR:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "className");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Look up the body of a constructor.
+	 */
+
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[2]);
+	if (oPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if (oPtr->classPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" is not a class", TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		    TclGetString(objv[2]), NULL);
+	    return TCL_ERROR;
+	}
+
+	methodPtr = oPtr->classPtr->constructorPtr;
+	if (methodPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" has no defined constructor",
+		    TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		    "CONSRUCTOR", NULL);
+	    return TCL_ERROR;
+	}
+	procPtr = TclOOGetProcFromMethod(methodPtr);
+	if (procPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "body not available for this kind of constructor", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		    "METHODTYPE", NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Compile if necessary.
+	 */
+
+	if (procPtr->bodyPtr->typePtr != &tclByteCodeType) {
+	    Command cmd;
+
+	    /*
+	     * Yes, this is ugly, but we need to pass the namespace in to the
+	     * compiler in two places.
+	     */
+
+	    cmd.nsPtr = (Namespace *) oPtr->namespacePtr;
+	    procPtr->cmdPtr = &cmd;
+	    result = TclProcCompileProc(interp, procPtr, procPtr->bodyPtr,
+		    (Namespace *) oPtr->namespacePtr, "body of constructor",
+		    TclGetString(objv[2]));
+	    procPtr->cmdPtr = NULL;
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+	codeObjPtr = procPtr->bodyPtr;
+	break;
+
+    case DISAS_CLASS_DESTRUCTOR:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "className");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Look up the body of a destructor.
+	 */
+
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[2]);
+	if (oPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if (oPtr->classPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" is not a class", TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		    TclGetString(objv[2]), NULL);
+	    return TCL_ERROR;
+	}
+
+	methodPtr = oPtr->classPtr->destructorPtr;
+	if (methodPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" has no defined destructor",
+		    TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		    "DESRUCTOR", NULL);
+	    return TCL_ERROR;
+	}
+	procPtr = TclOOGetProcFromMethod(methodPtr);
+	if (procPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "body not available for this kind of destructor", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		    "METHODTYPE", NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Compile if necessary.
+	 */
+
+	if (procPtr->bodyPtr->typePtr != &tclByteCodeType) {
+	    Command cmd;
+
+	    /*
+	     * Yes, this is ugly, but we need to pass the namespace in to the
+	     * compiler in two places.
+	     */
+
+	    cmd.nsPtr = (Namespace *) oPtr->namespacePtr;
+	    procPtr->cmdPtr = &cmd;
+	    result = TclProcCompileProc(interp, procPtr, procPtr->bodyPtr,
+		    (Namespace *) oPtr->namespacePtr, "body of destructor",
+		    TclGetString(objv[2]));
+	    procPtr->cmdPtr = NULL;
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+	codeObjPtr = procPtr->bodyPtr;
+	break;
+
+    case DISAS_CLASS_METHOD:
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "className methodName");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Look up the body of a class method.
+	 */
+
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[2]);
+	if (oPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if (oPtr->classPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "\"%s\" is not a class", TclGetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		    TclGetString(objv[2]), NULL);
+	    return TCL_ERROR;
+	}
+	hPtr = Tcl_FindHashEntry(&oPtr->classPtr->classMethods,
+		(char *) objv[3]);
+	goto methodBody;
+    case DISAS_OBJECT_METHOD:
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "objectName methodName");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Look up the body of an instance method.
+	 */
+
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[2]);
+	if (oPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if (oPtr->methodsPtr == NULL) {
+	    goto unknownMethod;
+	}
+	hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char *) objv[3]);
+
+	/*
+	 * Compile (if necessary) and disassemble a method body.
+	 */
+
+    methodBody:
+	if (hPtr == NULL) {
+	unknownMethod:
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "unknown method \"%s\"", TclGetString(objv[3])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		    TclGetString(objv[3]), NULL);
+	    return TCL_ERROR;
+	}
+	procPtr = TclOOGetProcFromMethod(Tcl_GetHashValue(hPtr));
+	if (procPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "body not available for this kind of method", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		    "METHODTYPE", NULL);
+	    return TCL_ERROR;
+	}
+	if (procPtr->bodyPtr->typePtr != &tclByteCodeType) {
+	    Command cmd;
+
+	    /*
+	     * Yes, this is ugly, but we need to pass the namespace in to the
+	     * compiler in two places.
+	     */
+
+	    cmd.nsPtr = (Namespace *) oPtr->namespacePtr;
+	    procPtr->cmdPtr = &cmd;
+	    result = TclProcCompileProc(interp, procPtr, procPtr->bodyPtr,
+		    (Namespace *) oPtr->namespacePtr, "body of method",
+		    TclGetString(objv[3]));
+	    procPtr->cmdPtr = NULL;
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+	codeObjPtr = procPtr->bodyPtr;
+	break;
+    default:
+	CLANG_ASSERT(0);
+    }
+
+    /*
+     * Do the actual disassembly.
+     */
+
+    if (BYTECODE(codeObjPtr)->flags & TCL_BYTECODE_PRECOMPILED) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"may not disassemble prebuilt bytecode", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "DISASSEMBLE",
+		"BYTECODE", NULL);
+	return TCL_ERROR;
+    }
+    if (PTR2INT(clientData)) {
+	Tcl_SetObjResult(interp,
+		DisassembleByteCodeAsDicts(interp, codeObjPtr));
+    } else {
+	Tcl_SetObjResult(interp,
+		DisassembleByteCodeObj(interp, codeObjPtr));
+    }
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * End:
+ */
diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c
new file mode 100644
index 0000000..7ab33f8
--- /dev/null
+++ b/generic/tclEncoding.c
@@ -0,0 +1,3647 @@
+/*
+ * tclEncoding.c --
+ *
+ *	Contains the implementation of the encoding conversion package.
+ *
+ * Copyright (c) 1996-1998 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+typedef size_t (LengthProc)(const char *src);
+
+/*
+ * The following data structure represents an encoding, which describes how to
+ * convert between various character sets and UTF-8.
+ */
+
+typedef struct Encoding {
+    char *name;			/* Name of encoding. Malloced because (1) hash
+				 * table entry that owns this encoding may be
+				 * freed prior to this encoding being freed,
+				 * (2) string passed in the Tcl_EncodingType
+				 * structure may not be persistent. */
+    Tcl_EncodingConvertProc *toUtfProc;
+				/* Function to convert from external encoding
+				 * into UTF-8. */
+    Tcl_EncodingConvertProc *fromUtfProc;
+				/* Function to convert from UTF-8 into
+				 * external encoding. */
+    Tcl_EncodingFreeProc *freeProc;
+				/* If non-NULL, function to call when this
+				 * encoding is deleted. */
+    int nullSize;		/* Number of 0x00 bytes that signify
+				 * end-of-string in this encoding. This number
+				 * is used to determine the source string
+				 * length when the srcLen argument is
+				 * negative. This number can be 1 or 2. */
+    ClientData clientData;	/* Arbitrary value associated with encoding
+				 * type. Passed to conversion functions. */
+    LengthProc *lengthProc;	/* Function to compute length of
+				 * null-terminated strings in this encoding.
+				 * If nullSize is 1, this is strlen; if
+				 * nullSize is 2, this is a function that
+				 * returns the number of bytes in a 0x0000
+				 * terminated string. */
+    size_t refCount;		/* Number of uses of this structure. */
+    Tcl_HashEntry *hPtr;	/* Hash table entry that owns this encoding. */
+} Encoding;
+
+/*
+ * The following structure is the clientData for a dynamically-loaded,
+ * table-driven encoding created by LoadTableEncoding(). It maps between
+ * Unicode and a single-byte, double-byte, or multibyte (1 or 2 bytes only)
+ * encoding.
+ */
+
+typedef struct TableEncodingData {
+    int fallback;		/* Character (in this encoding) to substitute
+				 * when this encoding cannot represent a UTF-8
+				 * character. */
+    char prefixBytes[256];	/* If a byte in the input stream is a lead
+				 * byte for a 2-byte sequence, the
+				 * corresponding entry in this array is 1,
+				 * otherwise it is 0. */
+    unsigned short **toUnicode;	/* Two dimensional sparse matrix to map
+				 * characters from the encoding to Unicode.
+				 * Each element of the toUnicode array points
+				 * to an array of 256 shorts. If there is no
+				 * corresponding character in Unicode, the
+				 * value in the matrix is 0x0000.
+				 * malloc'd. */
+    unsigned short **fromUnicode;
+				/* Two dimensional sparse matrix to map
+				 * characters from Unicode to the encoding.
+				 * Each element of the fromUnicode array
+				 * points to an array of 256 shorts. If there
+				 * is no corresponding character the encoding,
+				 * the value in the matrix is 0x0000.
+				 * malloc'd. */
+} TableEncodingData;
+
+/*
+ * The following structures is the clientData for a dynamically-loaded,
+ * escape-driven encoding that is itself comprised of other simpler encodings.
+ * An example is "iso-2022-jp", which uses escape sequences to switch between
+ * ascii, jis0208, jis0212, gb2312, and ksc5601. Note that "escape-driven"
+ * does not necessarily mean that the ESCAPE character is the character used
+ * for switching character sets.
+ */
+
+typedef struct EscapeSubTable {
+    unsigned sequenceLen;	/* Length of following string. */
+    char sequence[16];		/* Escape code that marks this encoding. */
+    char name[32];		/* Name for encoding. */
+    Encoding *encodingPtr;	/* Encoding loaded using above name, or NULL
+				 * if this sub-encoding has not been needed
+				 * yet. */
+} EscapeSubTable;
+
+typedef struct EscapeEncodingData {
+    int fallback;		/* Character (in this encoding) to substitute
+				 * when this encoding cannot represent a UTF-8
+				 * character. */
+    unsigned initLen;		/* Length of following string. */
+    char init[16];		/* String to emit or expect before first char
+				 * in conversion. */
+    unsigned finalLen;		/* Length of following string. */
+    char final[16];		/* String to emit or expect after last char in
+				 * conversion. */
+    char prefixBytes[256];	/* If a byte in the input stream is the first
+				 * character of one of the escape sequences in
+				 * the following array, the corresponding
+				 * entry in this array is 1, otherwise it is
+				 * 0. */
+    int numSubTables;		/* Length of following array. */
+    EscapeSubTable subTables[1];/* Information about each EscapeSubTable used
+				 * by this encoding type. The actual size will
+				 * be as large as necessary to hold all
+				 * EscapeSubTables. */
+} EscapeEncodingData;
+
+/*
+ * Constants used when loading an encoding file to identify the type of the
+ * file.
+ */
+
+#define ENCODING_SINGLEBYTE	0
+#define ENCODING_DOUBLEBYTE	1
+#define ENCODING_MULTIBYTE	2
+#define ENCODING_ESCAPE		3
+
+/*
+ * A list of directories in which Tcl should look for *.enc files. This list
+ * is shared by all threads. Access is governed by a mutex lock.
+ */
+
+static TclInitProcessGlobalValueProc InitializeEncodingSearchPath;
+static ProcessGlobalValue encodingSearchPath = {
+    0, 0, NULL, NULL, InitializeEncodingSearchPath, NULL, NULL
+};
+
+/*
+ * A map from encoding names to the directories in which their data files have
+ * been seen. The string value of the map is shared by all threads. Access to
+ * the shared string is governed by a mutex lock.
+ */
+
+static ProcessGlobalValue encodingFileMap = {
+    0, 0, NULL, NULL, NULL, NULL, NULL
+};
+
+/*
+ * A list of directories making up the "library path". Historically this
+ * search path has served many uses, but the only one remaining is a base for
+ * the encodingSearchPath above. If the application does not explicitly set
+ * the encodingSearchPath, then it will be initialized by appending /encoding
+ * to each directory in this "libraryPath".
+ */
+
+static ProcessGlobalValue libraryPath = {
+    0, 0, NULL, NULL, TclpInitLibraryPath, NULL, NULL
+};
+
+static int encodingsInitialized = 0;
+
+/*
+ * Hash table that keeps track of all loaded Encodings. Keys are the string
+ * names that represent the encoding, values are (Encoding *).
+ */
+
+static Tcl_HashTable encodingTable;
+TCL_DECLARE_MUTEX(encodingMutex)
+
+/*
+ * The following are used to hold the default and current system encodings.
+ * If NULL is passed to one of the conversion routines, the current setting of
+ * the system encoding will be used to perform the conversion.
+ */
+
+static Tcl_Encoding defaultEncoding = NULL;
+static Tcl_Encoding systemEncoding = NULL;
+Tcl_Encoding tclIdentityEncoding = NULL;
+
+/*
+ * The following variable is used in the sparse matrix code for a
+ * TableEncoding to represent a page in the table that has no entries.
+ */
+
+static unsigned short emptyPage[256];
+
+/*
+ * Functions used only in this module.
+ */
+
+static int		BinaryProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static void		DupEncodingIntRep(Tcl_Obj *srcPtr, Tcl_Obj *dupPtr);
+static void		EscapeFreeProc(ClientData clientData);
+static int		EscapeFromUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		EscapeToUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static void		FillEncodingFileMap(void);
+static void		FreeEncoding(Tcl_Encoding encoding);
+static void		FreeEncodingIntRep(Tcl_Obj *objPtr);
+static Encoding *	GetTableEncoding(EscapeEncodingData *dataPtr,
+			    int state);
+static Tcl_Encoding	LoadEncodingFile(Tcl_Interp *interp, const char *name);
+static Tcl_Encoding	LoadTableEncoding(const char *name, int type,
+			    Tcl_Channel chan);
+static Tcl_Encoding	LoadEscapeEncoding(const char *name, Tcl_Channel chan);
+static Tcl_Channel	OpenEncodingFileChannel(Tcl_Interp *interp,
+			    const char *name);
+static void		TableFreeProc(ClientData clientData);
+static int		TableFromUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		TableToUtfProc(ClientData clientData, const char *src,
+			    int srcLen, int flags, Tcl_EncodingState *statePtr,
+			    char *dst, int dstLen, int *srcReadPtr,
+			    int *dstWrotePtr, int *dstCharsPtr);
+static size_t		unilen(const char *src);
+static int		UnicodeToUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		UtfToUnicodeProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		UtfToUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr, int pureNullMode);
+static int		UtfIntToUtfExtProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		UtfExtToUtfIntProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		Iso88591FromUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst, int dstLen,
+			    int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		Iso88591ToUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst,
+			    int dstLen, int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+
+/*
+ * A Tcl_ObjType for holding a cached Tcl_Encoding in the twoPtrValue.ptr1 field
+ * of the intrep. This should help the lifetime of encodings be more useful.
+ * See concerns raised in [Bug 1077262].
+ */
+
+static const Tcl_ObjType encodingType = {
+    "encoding", FreeEncodingIntRep, DupEncodingIntRep, NULL, NULL
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEncodingFromObj --
+ *
+ *	Writes to (*encodingPtr) the Tcl_Encoding value of (*objPtr), if
+ *	possible, and returns TCL_OK. If no such encoding exists, TCL_ERROR is
+ *	returned, and if interp is non-NULL, an error message is written
+ *	there.
+ *
+ * Results:
+ *	Standard Tcl return code.
+ *
+ * Side effects:
+ *	Caches the Tcl_Encoding value as the internal rep of (*objPtr).
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEncodingFromObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    Tcl_Encoding *encodingPtr)
+{
+    const char *name = TclGetString(objPtr);
+
+    if (objPtr->typePtr != &encodingType) {
+	Tcl_Encoding encoding = Tcl_GetEncoding(interp, name);
+
+	if (encoding == NULL) {
+	    return TCL_ERROR;
+	}
+	TclFreeIntRep(objPtr);
+	objPtr->internalRep.twoPtrValue.ptr1 = encoding;
+	objPtr->typePtr = &encodingType;
+    }
+    *encodingPtr = Tcl_GetEncoding(NULL, name);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeEncodingIntRep --
+ *
+ *	The Tcl_FreeInternalRepProc for the "encoding" Tcl_ObjType.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeEncodingIntRep(
+    Tcl_Obj *objPtr)
+{
+    Tcl_FreeEncoding(objPtr->internalRep.twoPtrValue.ptr1);
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupEncodingIntRep --
+ *
+ *	The Tcl_DupInternalRepProc for the "encoding" Tcl_ObjType.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupEncodingIntRep(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *dupPtr)
+{
+    dupPtr->internalRep.twoPtrValue.ptr1 = Tcl_GetEncoding(NULL, srcPtr->bytes);
+    dupPtr->typePtr = &encodingType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEncodingSearchPath --
+ *
+ *	Keeps the per-thread copy of the encoding search path current with
+ *	changes to the global copy.
+ *
+ * Results:
+ *	Returns a "list" (Tcl_Obj *) that contains the encoding search path.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetEncodingSearchPath(void)
+{
+    return TclGetProcessGlobalValue(&encodingSearchPath);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEncodingSearchPath --
+ *
+ *	Keeps the per-thread copy of the encoding search path current with
+ *	changes to the global copy.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEncodingSearchPath(
+    Tcl_Obj *searchPath)
+{
+    int dummy;
+
+    if (TCL_ERROR == Tcl_ListObjLength(NULL, searchPath, &dummy)) {
+	return TCL_ERROR;
+    }
+    TclSetProcessGlobalValue(&encodingSearchPath, searchPath, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetLibraryPath --
+ *
+ *	Keeps the per-thread copy of the library path current with changes to
+ *	the global copy.
+ *
+ * Results:
+ *	Returns a "list" (Tcl_Obj *) that contains the library path.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetLibraryPath(void)
+{
+    return TclGetProcessGlobalValue(&libraryPath);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetLibraryPath --
+ *
+ *	Keeps the per-thread copy of the library path current with changes to
+ *	the global copy.
+ *
+ *	NOTE: this routine returns void, so there's no way to report the error
+ *	that searchPath is not a valid list. In that case, this routine will
+ *	silently do nothing.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetLibraryPath(
+    Tcl_Obj *path)
+{
+    int dummy;
+
+    if (TCL_ERROR == Tcl_ListObjLength(NULL, path, &dummy)) {
+	return;
+    }
+    TclSetProcessGlobalValue(&libraryPath, path, NULL);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FillEncodingFileMap --
+ *
+ *	Called to bring the encoding file map in sync with the current value
+ *	of the encoding search path.
+ *
+ *	Scan the directories on the encoding search path, find the *.enc
+ *	files, and store the found pathnames in a map associated with the
+ *	encoding name.
+ *
+ *	In particular, if $dir is on the encoding search path, and the file
+ *	$dir/foo.enc is found, then store a "foo" -> $dir entry in the map.
+ *	Later, any need for the "foo" encoding will quickly * be able to
+ *	construct the $dir/foo.enc pathname for reading the encoding data.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Entries are added to the encoding file map.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+FillEncodingFileMap(void)
+{
+    int i, numDirs = 0;
+    Tcl_Obj *map, *searchPath;
+
+    searchPath = Tcl_GetEncodingSearchPath();
+    Tcl_IncrRefCount(searchPath);
+    Tcl_ListObjLength(NULL, searchPath, &numDirs);
+    map = Tcl_NewDictObj();
+    Tcl_IncrRefCount(map);
+
+    for (i = numDirs-1; i >= 0; i--) {
+	/*
+	 * Iterate backwards through the search path so as we overwrite
+	 * entries found, we favor files earlier on the search path.
+	 */
+
+	int j, numFiles;
+	Tcl_Obj *directory, *matchFileList = Tcl_NewObj();
+	Tcl_Obj **filev;
+	Tcl_GlobTypeData readableFiles = {
+	    TCL_GLOB_TYPE_FILE, TCL_GLOB_PERM_R, NULL, NULL
+	};
+
+	Tcl_ListObjIndex(NULL, searchPath, i, &directory);
+	Tcl_IncrRefCount(directory);
+	Tcl_IncrRefCount(matchFileList);
+	Tcl_FSMatchInDirectory(NULL, matchFileList, directory, "*.enc",
+		&readableFiles);
+
+	Tcl_ListObjGetElements(NULL, matchFileList, &numFiles, &filev);
+	for (j=0; j<numFiles; j++) {
+	    Tcl_Obj *encodingName, *fileObj;
+
+	    fileObj = TclPathPart(NULL, filev[j], TCL_PATH_TAIL);
+	    encodingName = TclPathPart(NULL, fileObj, TCL_PATH_ROOT);
+	    Tcl_DictObjPut(NULL, map, encodingName, directory);
+	    Tcl_DecrRefCount(fileObj);
+	    Tcl_DecrRefCount(encodingName);
+	}
+	Tcl_DecrRefCount(matchFileList);
+	Tcl_DecrRefCount(directory);
+    }
+    Tcl_DecrRefCount(searchPath);
+    TclSetProcessGlobalValue(&encodingFileMap, map, NULL);
+    Tcl_DecrRefCount(map);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclInitEncodingSubsystem --
+ *
+ *	Initialize all resources used by this subsystem on a per-process
+ *	basis.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on the memory, object, and IO subsystems.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclInitEncodingSubsystem(void)
+{
+    Tcl_EncodingType type;
+    TableEncodingData *dataPtr;
+    unsigned size;
+    unsigned short i;
+
+    if (encodingsInitialized) {
+	return;
+    }
+
+    Tcl_MutexLock(&encodingMutex);
+    Tcl_InitHashTable(&encodingTable, TCL_STRING_KEYS);
+    Tcl_MutexUnlock(&encodingMutex);
+
+    /*
+     * Create a few initial encodings. Note that the UTF-8 to UTF-8
+     * translation is not a no-op, because it will turn a stream of improperly
+     * formed UTF-8 into a properly formed stream.
+     */
+
+    type.encodingName	= "identity";
+    type.toUtfProc	= BinaryProc;
+    type.fromUtfProc	= BinaryProc;
+    type.freeProc	= NULL;
+    type.nullSize	= 1;
+    type.clientData	= NULL;
+    tclIdentityEncoding = Tcl_CreateEncoding(&type);
+
+    type.encodingName	= "utf-8";
+    type.toUtfProc	= UtfExtToUtfIntProc;
+    type.fromUtfProc	= UtfIntToUtfExtProc;
+    type.freeProc	= NULL;
+    type.nullSize	= 1;
+    type.clientData	= NULL;
+    Tcl_CreateEncoding(&type);
+
+    type.encodingName   = "unicode";
+    type.toUtfProc	= UnicodeToUtfProc;
+    type.fromUtfProc    = UtfToUnicodeProc;
+    type.freeProc	= NULL;
+    type.nullSize	= 2;
+    type.clientData	= NULL;
+    Tcl_CreateEncoding(&type);
+
+    /*
+     * Need the iso8859-1 encoding in order to process binary data, so force
+     * it to always be embedded. Note that this encoding *must* be a proper
+     * table encoding or some of the escape encodings crash! Hence the ugly
+     * code to duplicate the structure of a table encoding here.
+     */
+
+    dataPtr = ckalloc(sizeof(TableEncodingData));
+    memset(dataPtr, 0, sizeof(TableEncodingData));
+    dataPtr->fallback = '?';
+
+    size = 256*(sizeof(unsigned short *) + sizeof(unsigned short));
+    dataPtr->toUnicode = ckalloc(size);
+    memset(dataPtr->toUnicode, 0, size);
+    dataPtr->fromUnicode = ckalloc(size);
+    memset(dataPtr->fromUnicode, 0, size);
+
+    dataPtr->toUnicode[0] = (unsigned short *) (dataPtr->toUnicode + 256);
+    dataPtr->fromUnicode[0] = (unsigned short *) (dataPtr->fromUnicode + 256);
+    for (i=1 ; i<256 ; i++) {
+	dataPtr->toUnicode[i] = emptyPage;
+	dataPtr->fromUnicode[i] = emptyPage;
+    }
+
+    for (i=0 ; i<256 ; i++) {
+	dataPtr->toUnicode[0][i] = i;
+	dataPtr->fromUnicode[0][i] = i;
+    }
+
+    type.encodingName	= "iso8859-1";
+    type.toUtfProc	= Iso88591ToUtfProc;
+    type.fromUtfProc	= Iso88591FromUtfProc;
+    type.freeProc	= TableFreeProc;
+    type.nullSize	= 1;
+    type.clientData	= dataPtr;
+    defaultEncoding	= Tcl_CreateEncoding(&type);
+    systemEncoding	= Tcl_GetEncoding(NULL, type.encodingName);
+
+    encodingsInitialized = 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeEncodingSubsystem --
+ *
+ *	Release the state associated with the encoding subsystem.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees all of the encodings.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeEncodingSubsystem(void)
+{
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hPtr;
+
+    Tcl_MutexLock(&encodingMutex);
+    encodingsInitialized = 0;
+    FreeEncoding(systemEncoding);
+    systemEncoding = NULL;
+    defaultEncoding = NULL;
+    FreeEncoding(tclIdentityEncoding);
+    tclIdentityEncoding = NULL;
+
+    hPtr = Tcl_FirstHashEntry(&encodingTable, &search);
+    while (hPtr != NULL) {
+	/*
+	 * Call FreeEncoding instead of doing it directly to handle refcounts
+	 * like escape encodings use. [Bug 524674] Make sure to call
+	 * Tcl_FirstHashEntry repeatedly so that all encodings are eventually
+	 * cleaned up.
+	 */
+
+	FreeEncoding(Tcl_GetHashValue(hPtr));
+	hPtr = Tcl_FirstHashEntry(&encodingTable, &search);
+    }
+
+    Tcl_DeleteHashTable(&encodingTable);
+    Tcl_MutexUnlock(&encodingMutex);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_GetDefaultEncodingDir --
+ *
+ *	Legacy public interface to retrieve first directory in the encoding
+ *	searchPath.
+ *
+ * Results:
+ *	The directory pathname, as a string, or NULL for an empty encoding
+ *	search path.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetDefaultEncodingDir(void)
+{
+    int numDirs;
+    Tcl_Obj *first, *searchPath = Tcl_GetEncodingSearchPath();
+
+    Tcl_ListObjLength(NULL, searchPath, &numDirs);
+    if (numDirs == 0) {
+	return NULL;
+    }
+    Tcl_ListObjIndex(NULL, searchPath, 0, &first);
+
+    return TclGetString(first);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_SetDefaultEncodingDir --
+ *
+ *	Legacy public interface to set the first directory in the encoding
+ *	search path.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Modifies the encoding search path.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+Tcl_SetDefaultEncodingDir(
+    const char *path)
+{
+    Tcl_Obj *searchPath = Tcl_GetEncodingSearchPath();
+    Tcl_Obj *directory = Tcl_NewStringObj(path, -1);
+
+    searchPath = Tcl_DuplicateObj(searchPath);
+    Tcl_ListObjReplace(NULL, searchPath, 0, 0, 1, &directory);
+    Tcl_SetEncodingSearchPath(searchPath);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_GetEncoding --
+ *
+ *	Given the name of a encoding, find the corresponding Tcl_Encoding
+ *	token. If the encoding did not already exist, Tcl attempts to
+ *	dynamically load an encoding by that name.
+ *
+ * Results:
+ *	Returns a token that represents the encoding. If the name didn't refer
+ *	to any known or loadable encoding, NULL is returned. If NULL was
+ *	returned, an error message is left in interp's result object, unless
+ *	interp was NULL.
+ *
+ * Side effects:
+ *	The new encoding type is entered into a table visible to all
+ *	interpreters, keyed off the encoding's name. For each call to this
+ *	function, there should eventually be a call to Tcl_FreeEncoding, so
+ *	that the database can be cleaned up when encodings aren't needed
+ *	anymore.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+Tcl_Encoding
+Tcl_GetEncoding(
+    Tcl_Interp *interp,		/* Interp for error reporting, if not NULL. */
+    const char *name)		/* The name of the desired encoding. */
+{
+    Tcl_HashEntry *hPtr;
+    Encoding *encodingPtr;
+
+    Tcl_MutexLock(&encodingMutex);
+    if (name == NULL) {
+	encodingPtr = (Encoding *) systemEncoding;
+	encodingPtr->refCount++;
+	Tcl_MutexUnlock(&encodingMutex);
+	return systemEncoding;
+    }
+
+    hPtr = Tcl_FindHashEntry(&encodingTable, name);
+    if (hPtr != NULL) {
+	encodingPtr = Tcl_GetHashValue(hPtr);
+	encodingPtr->refCount++;
+	Tcl_MutexUnlock(&encodingMutex);
+	return (Tcl_Encoding) encodingPtr;
+    }
+    Tcl_MutexUnlock(&encodingMutex);
+
+    return LoadEncodingFile(interp, name);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FreeEncoding --
+ *
+ *	This function is called to release an encoding allocated by
+ *	Tcl_CreateEncoding() or Tcl_GetEncoding().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The reference count associated with the encoding is decremented and
+ *	the encoding may be deleted if nothing is using it anymore.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+Tcl_FreeEncoding(
+    Tcl_Encoding encoding)
+{
+    Tcl_MutexLock(&encodingMutex);
+    FreeEncoding(encoding);
+    Tcl_MutexUnlock(&encodingMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeEncoding --
+ *
+ *	This function is called to release an encoding by functions that
+ *	already have the encodingMutex.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The reference count associated with the encoding is decremented and
+ *	the encoding may be deleted if nothing is using it anymore.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeEncoding(
+    Tcl_Encoding encoding)
+{
+    Encoding *encodingPtr = (Encoding *) encoding;
+
+    if (encodingPtr == NULL) {
+	return;
+    }
+    if (encodingPtr->refCount-- <= 1) {
+	if (encodingPtr->freeProc != NULL) {
+	    encodingPtr->freeProc(encodingPtr->clientData);
+	}
+	if (encodingPtr->hPtr != NULL) {
+	    Tcl_DeleteHashEntry(encodingPtr->hPtr);
+	}
+	ckfree(encodingPtr->name);
+	ckfree(encodingPtr);
+    }
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_GetEncodingName --
+ *
+ *	Given an encoding, return the name that was used to constuct the
+ *	encoding.
+ *
+ * Results:
+ *	The name of the encoding.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetEncodingName(
+    Tcl_Encoding encoding)	/* The encoding whose name to fetch. */
+{
+    if (encoding == NULL) {
+	encoding = systemEncoding;
+    }
+
+    return ((Encoding *) encoding)->name;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_GetEncodingNames --
+ *
+ *	Get the list of all known encodings, including the ones stored as
+ *	files on disk in the encoding path.
+ *
+ * Results:
+ *	Modifies interp's result object to hold a list of all the available
+ *	encodings.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+Tcl_GetEncodingNames(
+    Tcl_Interp *interp)		/* Interp to hold result. */
+{
+    Tcl_HashTable table;
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hPtr;
+    Tcl_Obj *map, *name, *result = Tcl_NewObj();
+    Tcl_DictSearch mapSearch;
+    int dummy, done = 0;
+
+    Tcl_InitObjHashTable(&table);
+
+    /*
+     * Copy encoding names from loaded encoding table to table.
+     */
+
+    Tcl_MutexLock(&encodingMutex);
+    for (hPtr = Tcl_FirstHashEntry(&encodingTable, &search); hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	Encoding *encodingPtr = Tcl_GetHashValue(hPtr);
+
+	Tcl_CreateHashEntry(&table,
+		Tcl_NewStringObj(encodingPtr->name, -1), &dummy);
+    }
+    Tcl_MutexUnlock(&encodingMutex);
+
+    FillEncodingFileMap();
+    map = TclGetProcessGlobalValue(&encodingFileMap);
+
+    /*
+     * Copy encoding names from encoding file map to table.
+     */
+
+    Tcl_DictObjFirst(NULL, map, &mapSearch, &name, NULL, &done);
+    for (; !done; Tcl_DictObjNext(&mapSearch, &name, NULL, &done)) {
+	Tcl_CreateHashEntry(&table, name, &dummy);
+    }
+
+    /*
+     * Pull all encoding names from table into the result list.
+     */
+
+    for (hPtr = Tcl_FirstHashEntry(&table, &search); hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	Tcl_ListObjAppendElement(NULL, result,
+		(Tcl_Obj *) Tcl_GetHashKey(&table, hPtr));
+    }
+    Tcl_SetObjResult(interp, result);
+    Tcl_DeleteHashTable(&table);
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * Tcl_SetSystemEncoding --
+ *
+ *	Sets the default encoding that should be used whenever the user passes
+ *	a NULL value in to one of the conversion routines. If the supplied
+ *	name is NULL, the system encoding is reset to the default system
+ *	encoding.
+ *
+ * Results:
+ *	The return value is TCL_OK if the system encoding was successfully set
+ *	to the encoding specified by name, TCL_ERROR otherwise. If TCL_ERROR
+ *	is returned, an error message is left in interp's result object,
+ *	unless interp was NULL.
+ *
+ * Side effects:
+ *	The reference count of the new system encoding is incremented. The
+ *	reference count of the old system encoding is decremented and it may
+ *	be freed. All VFS cached information is invalidated.
+ *
+ *------------------------------------------------------------------------
+ */
+
+int
+Tcl_SetSystemEncoding(
+    Tcl_Interp *interp,		/* Interp for error reporting, if not NULL. */
+    const char *name)		/* The name of the desired encoding, or NULL/""
+				 * to reset to default encoding. */
+{
+    Tcl_Encoding encoding;
+    Encoding *encodingPtr;
+
+    if (!name || !*name) {
+	Tcl_MutexLock(&encodingMutex);
+	encoding = defaultEncoding;
+	encodingPtr = (Encoding *) encoding;
+	encodingPtr->refCount++;
+	Tcl_MutexUnlock(&encodingMutex);
+    } else {
+	encoding = Tcl_GetEncoding(interp, name);
+	if (encoding == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_MutexLock(&encodingMutex);
+    FreeEncoding(systemEncoding);
+    systemEncoding = encoding;
+    Tcl_MutexUnlock(&encodingMutex);
+    Tcl_FSMountsChanged(NULL);
+
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_CreateEncoding --
+ *
+ *	This function is called to define a new encoding and the functions
+ *	that are used to convert between the specified encoding and Unicode.
+ *
+ * Results:
+ *	Returns a token that represents the encoding. If an encoding with the
+ *	same name already existed, the old encoding token remains valid and
+ *	continues to behave as it used to, and will eventually be garbage
+ *	collected when the last reference to it goes away. Any subsequent
+ *	calls to Tcl_GetEncoding with the specified name will retrieve the
+ *	most recent encoding token.
+ *
+ * Side effects:
+ *	The new encoding type is entered into a table visible to all
+ *	interpreters, keyed off the encoding's name. For each call to this
+ *	function, there should eventually be a call to Tcl_FreeEncoding, so
+ *	that the database can be cleaned up when encodings aren't needed
+ *	anymore.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Encoding
+Tcl_CreateEncoding(
+    const Tcl_EncodingType *typePtr)
+				/* The encoding type. */
+{
+    Tcl_HashEntry *hPtr;
+    int isNew;
+    Encoding *encodingPtr;
+    char *name;
+
+    Tcl_MutexLock(&encodingMutex);
+    hPtr = Tcl_CreateHashEntry(&encodingTable, typePtr->encodingName, &isNew);
+    if (isNew == 0) {
+	/*
+	 * Remove old encoding from hash table, but don't delete it until last
+	 * reference goes away.
+	 */
+
+	encodingPtr = Tcl_GetHashValue(hPtr);
+	encodingPtr->hPtr = NULL;
+    }
+
+    name = ckalloc(strlen(typePtr->encodingName) + 1);
+
+    encodingPtr = ckalloc(sizeof(Encoding));
+    encodingPtr->name		= strcpy(name, typePtr->encodingName);
+    encodingPtr->toUtfProc	= typePtr->toUtfProc;
+    encodingPtr->fromUtfProc	= typePtr->fromUtfProc;
+    encodingPtr->freeProc	= typePtr->freeProc;
+    encodingPtr->nullSize	= typePtr->nullSize;
+    encodingPtr->clientData	= typePtr->clientData;
+    if (typePtr->nullSize == 1) {
+	encodingPtr->lengthProc = (LengthProc *) strlen;
+    } else {
+	encodingPtr->lengthProc = (LengthProc *) unilen;
+    }
+    encodingPtr->refCount	= 1;
+    encodingPtr->hPtr		= hPtr;
+    Tcl_SetHashValue(hPtr, encodingPtr);
+
+    Tcl_MutexUnlock(&encodingMutex);
+
+    return (Tcl_Encoding) encodingPtr;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_ExternalToUtfDString --
+ *
+ *	Convert a source buffer from the specified encoding into UTF-8. If any
+ *	of the bytes in the source buffer are invalid or cannot be represented
+ *	in the target encoding, a default fallback character will be
+ *	substituted.
+ *
+ * Results:
+ *	The converted bytes are stored in the DString, which is then NULL
+ *	terminated. The return value is a pointer to the value stored in the
+ *	DString.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+char *
+Tcl_ExternalToUtfDString(
+    Tcl_Encoding encoding,	/* The encoding for the source string, or NULL
+				 * for the default system encoding. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes, or < 0 for
+				 * encoding-specific string length. */
+    Tcl_DString *dstPtr)	/* Uninitialized or free DString in which the
+				 * converted string is stored. */
+{
+    char *dst;
+    Tcl_EncodingState state;
+    const Encoding *encodingPtr;
+    int flags, dstLen, result, soFar, srcRead, dstWrote, dstChars;
+
+    Tcl_DStringInit(dstPtr);
+    dst = Tcl_DStringValue(dstPtr);
+    dstLen = dstPtr->spaceAvl - 1;
+
+    if (encoding == NULL) {
+	encoding = systemEncoding;
+    }
+    encodingPtr = (Encoding *) encoding;
+
+    if (src == NULL) {
+	srcLen = 0;
+    } else if (srcLen < 0) {
+	srcLen = encodingPtr->lengthProc(src);
+    }
+
+    flags = TCL_ENCODING_START | TCL_ENCODING_END;
+
+    while (1) {
+	result = encodingPtr->toUtfProc(encodingPtr->clientData, src, srcLen,
+		flags, &state, dst, dstLen, &srcRead, &dstWrote, &dstChars);
+	soFar = dst + dstWrote - Tcl_DStringValue(dstPtr);
+
+	if (result != TCL_CONVERT_NOSPACE) {
+	    Tcl_DStringSetLength(dstPtr, soFar);
+	    return Tcl_DStringValue(dstPtr);
+	}
+
+	flags &= ~TCL_ENCODING_START;
+	src += srcRead;
+	srcLen -= srcRead;
+	if (Tcl_DStringLength(dstPtr) == 0) {
+	    Tcl_DStringSetLength(dstPtr, dstLen);
+	}
+	Tcl_DStringSetLength(dstPtr, 2 * Tcl_DStringLength(dstPtr) + 1);
+	dst = Tcl_DStringValue(dstPtr) + soFar;
+	dstLen = Tcl_DStringLength(dstPtr) - soFar - 1;
+    }
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_ExternalToUtf --
+ *
+ *	Convert a source buffer from the specified encoding into UTF-8.
+ *
+ * Results:
+ *	The return value is one of TCL_OK, TCL_CONVERT_MULTIBYTE,
+ *	TCL_CONVERT_SYNTAX, TCL_CONVERT_UNKNOWN, or TCL_CONVERT_NOSPACE, as
+ *	documented in tcl.h.
+ *
+ * Side effects:
+ *	The converted bytes are stored in the output buffer.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+int
+Tcl_ExternalToUtf(
+    Tcl_Interp *interp,		/* Interp for error return, if not NULL. */
+    Tcl_Encoding encoding,	/* The encoding for the source string, or NULL
+				 * for the default system encoding. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes, or < 0 for
+				 * encoding-specific string length. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const Encoding *encodingPtr;
+    int result, srcRead, dstWrote, dstChars = 0;
+    int noTerminate = flags & TCL_ENCODING_NO_TERMINATE;
+    int charLimited = (flags & TCL_ENCODING_CHAR_LIMIT) && dstCharsPtr;
+    int maxChars = INT_MAX;
+    Tcl_EncodingState state;
+
+    if (encoding == NULL) {
+	encoding = systemEncoding;
+    }
+    encodingPtr = (Encoding *) encoding;
+
+    if (src == NULL) {
+	srcLen = 0;
+    } else if (srcLen < 0) {
+	srcLen = encodingPtr->lengthProc(src);
+    }
+    if (statePtr == NULL) {
+	flags |= TCL_ENCODING_START | TCL_ENCODING_END;
+	statePtr = &state;
+    }
+    if (srcReadPtr == NULL) {
+	srcReadPtr = &srcRead;
+    }
+    if (dstWrotePtr == NULL) {
+	dstWrotePtr = &dstWrote;
+    }
+    if (dstCharsPtr == NULL) {
+	dstCharsPtr = &dstChars;
+	flags &= ~TCL_ENCODING_CHAR_LIMIT;
+    } else if (charLimited) {
+	maxChars = *dstCharsPtr;
+    }
+
+    if (!noTerminate) {
+	/*
+	 * If there are any null characters in the middle of the buffer,
+	 * they will converted to the UTF-8 null character (\xC080). To get
+	 * the actual \0 at the end of the destination buffer, we need to
+	 * append it manually.  First make room for it...
+	 */
+
+	dstLen--;
+    }
+    do {
+	int savedFlags = flags;
+	Tcl_EncodingState savedState = *statePtr;
+
+	result = encodingPtr->toUtfProc(encodingPtr->clientData, src, srcLen,
+		flags, statePtr, dst, dstLen, srcReadPtr, dstWrotePtr,
+		dstCharsPtr);
+	if (*dstCharsPtr <= maxChars) {
+	    break;
+	}
+	dstLen = Tcl_UtfAtIndex(dst, maxChars) - 1 - dst + TCL_UTF_MAX;
+	flags = savedFlags;
+	*statePtr = savedState;
+    } while (1);
+    if (!noTerminate) {
+	/* ...and then append it */
+
+	dst[*dstWrotePtr] = '\0';
+    }
+
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_UtfToExternalDString --
+ *
+ *	Convert a source buffer from UTF-8 into the specified encoding. If any
+ *	of the bytes in the source buffer are invalid or cannot be represented
+ *	in the target encoding, a default fallback character will be
+ *	substituted.
+ *
+ * Results:
+ *	The converted bytes are stored in the DString, which is then NULL
+ *	terminated in an encoding-specific manner. The return value is a
+ *	pointer to the value stored in the DString.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+char *
+Tcl_UtfToExternalDString(
+    Tcl_Encoding encoding,	/* The encoding for the converted string, or
+				 * NULL for the default system encoding. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes, or < 0 for
+				 * strlen(). */
+    Tcl_DString *dstPtr)	/* Uninitialized or free DString in which the
+				 * converted string is stored. */
+{
+    char *dst;
+    Tcl_EncodingState state;
+    const Encoding *encodingPtr;
+    int flags, dstLen, result, soFar, srcRead, dstWrote, dstChars;
+
+    Tcl_DStringInit(dstPtr);
+    dst = Tcl_DStringValue(dstPtr);
+    dstLen = dstPtr->spaceAvl - 1;
+
+    if (encoding == NULL) {
+	encoding = systemEncoding;
+    }
+    encodingPtr = (Encoding *) encoding;
+
+    if (src == NULL) {
+	srcLen = 0;
+    } else if (srcLen < 0) {
+	srcLen = strlen(src);
+    }
+    flags = TCL_ENCODING_START | TCL_ENCODING_END;
+    while (1) {
+	result = encodingPtr->fromUtfProc(encodingPtr->clientData, src,
+		srcLen, flags, &state, dst, dstLen, &srcRead, &dstWrote,
+		&dstChars);
+	soFar = dst + dstWrote - Tcl_DStringValue(dstPtr);
+
+	if (result != TCL_CONVERT_NOSPACE) {
+	    if (encodingPtr->nullSize == 2) {
+		Tcl_DStringSetLength(dstPtr, soFar + 1);
+	    }
+	    Tcl_DStringSetLength(dstPtr, soFar);
+	    return Tcl_DStringValue(dstPtr);
+	}
+
+	flags &= ~TCL_ENCODING_START;
+	src += srcRead;
+	srcLen -= srcRead;
+	if (Tcl_DStringLength(dstPtr) == 0) {
+	    Tcl_DStringSetLength(dstPtr, dstLen);
+	}
+	Tcl_DStringSetLength(dstPtr, 2 * Tcl_DStringLength(dstPtr) + 1);
+	dst = Tcl_DStringValue(dstPtr) + soFar;
+	dstLen = Tcl_DStringLength(dstPtr) - soFar - 1;
+    }
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_UtfToExternal --
+ *
+ *	Convert a buffer from UTF-8 into the specified encoding.
+ *
+ * Results:
+ *	The return value is one of TCL_OK, TCL_CONVERT_MULTIBYTE,
+ *	TCL_CONVERT_SYNTAX, TCL_CONVERT_UNKNOWN, or TCL_CONVERT_NOSPACE, as
+ *	documented in tcl.h.
+ *
+ * Side effects:
+ *	The converted bytes are stored in the output buffer.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfToExternal(
+    Tcl_Interp *interp,		/* Interp for error return, if not NULL. */
+    Tcl_Encoding encoding,	/* The encoding for the converted string, or
+				 * NULL for the default system encoding. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes, or < 0 for
+				 * strlen(). */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string
+				 * is stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const Encoding *encodingPtr;
+    int result, srcRead, dstWrote, dstChars;
+    Tcl_EncodingState state;
+
+    if (encoding == NULL) {
+	encoding = systemEncoding;
+    }
+    encodingPtr = (Encoding *) encoding;
+
+    if (src == NULL) {
+	srcLen = 0;
+    } else if (srcLen < 0) {
+	srcLen = strlen(src);
+    }
+    if (statePtr == NULL) {
+	flags |= TCL_ENCODING_START | TCL_ENCODING_END;
+	statePtr = &state;
+    }
+    if (srcReadPtr == NULL) {
+	srcReadPtr = &srcRead;
+    }
+    if (dstWrotePtr == NULL) {
+	dstWrotePtr = &dstWrote;
+    }
+    if (dstCharsPtr == NULL) {
+	dstCharsPtr = &dstChars;
+    }
+
+    dstLen -= encodingPtr->nullSize;
+    result = encodingPtr->fromUtfProc(encodingPtr->clientData, src, srcLen,
+	    flags, statePtr, dst, dstLen, srcReadPtr, dstWrotePtr,
+	    dstCharsPtr);
+    if (encodingPtr->nullSize == 2) {
+	dst[*dstWrotePtr + 1] = '\0';
+    }
+    dst[*dstWrotePtr] = '\0';
+
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FindExecutable --
+ *
+ *	This function computes the absolute path name of the current
+ *	application, given its argv[0] value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The absolute pathname for the application is computed and stored to be
+ *	returned later by [info nameofexecutable].
+ *
+ *---------------------------------------------------------------------------
+ */
+#undef Tcl_FindExecutable
+void
+Tcl_FindExecutable(
+    const char *argv0)		/* The value of the application's argv[0]
+				 * (native). */
+{
+    TclInitSubsystems();
+    TclpSetInitialEncodings();
+    TclpFindExecutable(argv0);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * OpenEncodingFileChannel --
+ *
+ *	Open the file believed to hold data for the encoding, "name".
+ *
+ * Results:
+ *	Returns the readable Tcl_Channel from opening the file, or NULL if the
+ *	file could not be successfully opened. If NULL was returned, an error
+ *	message is left in interp's result object, unless interp was NULL.
+ *
+ * Side effects:
+ *	Channel may be opened. Information about the filesystem may be cached
+ *	to speed later calls.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Channel
+OpenEncodingFileChannel(
+    Tcl_Interp *interp,		/* Interp for error reporting, if not NULL. */
+    const char *name)		/* The name of the encoding file on disk and
+				 * also the name for new encoding. */
+{
+    Tcl_Obj *nameObj = Tcl_NewStringObj(name, -1);
+    Tcl_Obj *fileNameObj = Tcl_DuplicateObj(nameObj);
+    Tcl_Obj *searchPath = Tcl_DuplicateObj(Tcl_GetEncodingSearchPath());
+    Tcl_Obj *map = TclGetProcessGlobalValue(&encodingFileMap);
+    Tcl_Obj **dir, *path, *directory = NULL;
+    Tcl_Channel chan = NULL;
+    int i, numDirs;
+
+    Tcl_ListObjGetElements(NULL, searchPath, &numDirs, &dir);
+    Tcl_IncrRefCount(nameObj);
+    Tcl_AppendToObj(fileNameObj, ".enc", -1);
+    Tcl_IncrRefCount(fileNameObj);
+    Tcl_DictObjGet(NULL, map, nameObj, &directory);
+
+    /*
+     * Check that any cached directory is still on the encoding search path.
+     */
+
+    if (NULL != directory) {
+	int verified = 0;
+
+	for (i=0; i<numDirs && !verified; i++) {
+	    if (dir[i] == directory) {
+		verified = 1;
+	    }
+	}
+	if (!verified) {
+	    const char *dirString = TclGetString(directory);
+
+	    for (i=0; i<numDirs && !verified; i++) {
+		if (strcmp(dirString, TclGetString(dir[i])) == 0) {
+		    verified = 1;
+		}
+	    }
+	}
+	if (!verified) {
+	    /*
+	     * Directory no longer on the search path. Remove from cache.
+	     */
+
+	    map = Tcl_DuplicateObj(map);
+	    Tcl_DictObjRemove(NULL, map, nameObj);
+	    TclSetProcessGlobalValue(&encodingFileMap, map, NULL);
+	    directory = NULL;
+	}
+    }
+
+    if (NULL != directory) {
+	/*
+	 * Got a directory from the cache. Try to use it first.
+	 */
+
+	Tcl_IncrRefCount(directory);
+	path = Tcl_FSJoinToPath(directory, 1, &fileNameObj);
+	Tcl_IncrRefCount(path);
+	Tcl_DecrRefCount(directory);
+	chan = Tcl_FSOpenFileChannel(NULL, path, "r", 0);
+	Tcl_DecrRefCount(path);
+    }
+
+    /*
+     * Scan the search path until we find it.
+     */
+
+    for (i=0; i<numDirs && (chan == NULL); i++) {
+	path = Tcl_FSJoinToPath(dir[i], 1, &fileNameObj);
+	Tcl_IncrRefCount(path);
+	chan = Tcl_FSOpenFileChannel(NULL, path, "r", 0);
+	Tcl_DecrRefCount(path);
+	if (chan != NULL) {
+	    /*
+	     * Save directory in the cache.
+	     */
+
+	    map = Tcl_DuplicateObj(TclGetProcessGlobalValue(&encodingFileMap));
+	    Tcl_DictObjPut(NULL, map, nameObj, dir[i]);
+	    TclSetProcessGlobalValue(&encodingFileMap, map, NULL);
+	}
+    }
+
+    if ((NULL == chan) && (interp != NULL)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown encoding \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ENCODING", name, NULL);
+    }
+    Tcl_DecrRefCount(fileNameObj);
+    Tcl_DecrRefCount(nameObj);
+    Tcl_DecrRefCount(searchPath);
+
+    return chan;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * LoadEncodingFile --
+ *
+ *	Read a file that describes an encoding and create a new Encoding from
+ *	the data.
+ *
+ * Results:
+ *	The return value is the newly loaded Encoding, or NULL if the file
+ *	didn't exist of was in the incorrect format. If NULL was returned, an
+ *	error message is left in interp's result object, unless interp was
+ *	NULL.
+ *
+ * Side effects:
+ *	File read from disk.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Encoding
+LoadEncodingFile(
+    Tcl_Interp *interp,		/* Interp for error reporting, if not NULL. */
+    const char *name)		/* The name of the encoding file on disk and
+				 * also the name for new encoding. */
+{
+    Tcl_Channel chan = NULL;
+    Tcl_Encoding encoding = NULL;
+    int ch;
+
+    chan = OpenEncodingFileChannel(interp, name);
+    if (chan == NULL) {
+	return NULL;
+    }
+
+    Tcl_SetChannelOption(NULL, chan, "-encoding", "utf-8");
+
+    while (1) {
+	Tcl_DString ds;
+
+	Tcl_DStringInit(&ds);
+	Tcl_Gets(chan, &ds);
+	ch = Tcl_DStringValue(&ds)[0];
+	Tcl_DStringFree(&ds);
+	if (ch != '#') {
+	    break;
+	}
+    }
+
+    switch (ch) {
+    case 'S':
+	encoding = LoadTableEncoding(name, ENCODING_SINGLEBYTE, chan);
+	break;
+    case 'D':
+	encoding = LoadTableEncoding(name, ENCODING_DOUBLEBYTE, chan);
+	break;
+    case 'M':
+	encoding = LoadTableEncoding(name, ENCODING_MULTIBYTE, chan);
+	break;
+    case 'E':
+	encoding = LoadEscapeEncoding(name, chan);
+	break;
+    }
+    if ((encoding == NULL) && (interp != NULL)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"invalid encoding file \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ENCODING", name, NULL);
+    }
+    Tcl_Close(NULL, chan);
+
+    return encoding;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * LoadTableEncoding --
+ *
+ *	Helper function for LoadEncodingTable(). Loads a table to that
+ *	converts between Unicode and some other encoding and creates an
+ *	encoding (using a TableEncoding structure) from that information.
+ *
+ *	File contains binary data, but begins with a marker to indicate
+ *	byte-ordering, so that same binary file can be read on either endian
+ *	platforms.
+ *
+ * Results:
+ *	The return value is the new encoding, or NULL if the encoding could
+ *	not be created (because the file contained invalid data).
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static Tcl_Encoding
+LoadTableEncoding(
+    const char *name,		/* Name for new encoding. */
+    int type,			/* Type of encoding (ENCODING_?????). */
+    Tcl_Channel chan)		/* File containing new encoding. */
+{
+    Tcl_DString lineString;
+    Tcl_Obj *objPtr;
+    char *line;
+    int i, hi, lo, numPages, symbol, fallback, len;
+    unsigned char used[256];
+    unsigned size;
+    TableEncodingData *dataPtr;
+    unsigned short *pageMemPtr, *page;
+    Tcl_EncodingType encType;
+
+    /*
+     * Speed over memory. Use a full 256 character table to decode hex
+     * sequences in the encoding files.
+     */
+
+    static const char staticHex[] = {
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*   0 ...  15 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*  16 ...  31 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*  32 ...  47 */
+      0,  1,  2,  3,  4,  5,  6, 7, 8, 9, 0, 0, 0, 0, 0, 0, /*  48 ...  63 */
+      0, 10, 11, 12, 13, 14, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*  64 ...  79 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*  80 ...  95 */
+      0, 10, 11, 12, 13, 14, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, /*  96 ... 111 */
+      0,  1,  2,  3,  4,  5,  6, 7, 8, 9, 0, 0, 0, 0, 0, 0, /* 112 ... 127 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 128 ... 143 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 144 ... 159 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 160 ... 175 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 176 ... 191 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 192 ... 207 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 208 ... 223 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 224 ... 239 */
+      0,  0,  0,  0,  0,  0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 240 ... 255 */
+    };
+
+    Tcl_DStringInit(&lineString);
+    Tcl_Gets(chan, &lineString);
+    line = Tcl_DStringValue(&lineString);
+
+    fallback = (int) strtol(line, &line, 16);
+    symbol = (int) strtol(line, &line, 10);
+    numPages = (int) strtol(line, &line, 10);
+    Tcl_DStringFree(&lineString);
+
+    if (numPages < 0) {
+	numPages = 0;
+    } else if (numPages > 256) {
+	numPages = 256;
+    }
+
+    memset(used, 0, sizeof(used));
+
+#undef PAGESIZE
+#define PAGESIZE    (256 * sizeof(unsigned short))
+
+    dataPtr = ckalloc(sizeof(TableEncodingData));
+    memset(dataPtr, 0, sizeof(TableEncodingData));
+
+    dataPtr->fallback = fallback;
+
+    /*
+     * Read the table that maps characters to Unicode. Performs a single
+     * malloc to get the memory for the array and all the pages needed by the
+     * array.
+     */
+
+    size = 256 * sizeof(unsigned short *) + numPages * PAGESIZE;
+    dataPtr->toUnicode = ckalloc(size);
+    memset(dataPtr->toUnicode, 0, size);
+    pageMemPtr = (unsigned short *) (dataPtr->toUnicode + 256);
+
+    TclNewObj(objPtr);
+    Tcl_IncrRefCount(objPtr);
+    for (i = 0; i < numPages; i++) {
+	int ch;
+	const char *p;
+
+	Tcl_ReadChars(chan, objPtr, 3 + 16 * (16 * 4 + 1), 0);
+	p = TclGetString(objPtr);
+	hi = (staticHex[UCHAR(p[0])] << 4) + staticHex[UCHAR(p[1])];
+	dataPtr->toUnicode[hi] = pageMemPtr;
+	p += 2;
+	for (lo = 0; lo < 256; lo++) {
+	    if ((lo & 0x0f) == 0) {
+		p++;
+	    }
+	    ch = (staticHex[UCHAR(p[0])] << 12) + (staticHex[UCHAR(p[1])] << 8)
+		    + (staticHex[UCHAR(p[2])] << 4) + staticHex[UCHAR(p[3])];
+	    if (ch != 0) {
+		used[ch >> 8] = 1;
+	    }
+	    *pageMemPtr = (unsigned short) ch;
+	    pageMemPtr++;
+	    p += 4;
+	}
+    }
+    TclDecrRefCount(objPtr);
+
+    if (type == ENCODING_DOUBLEBYTE) {
+	memset(dataPtr->prefixBytes, 1, sizeof(dataPtr->prefixBytes));
+    } else {
+	for (hi = 1; hi < 256; hi++) {
+	    if (dataPtr->toUnicode[hi] != NULL) {
+		dataPtr->prefixBytes[hi] = 1;
+	    }
+	}
+    }
+
+    /*
+     * Invert toUnicode array to produce the fromUnicode array. Performs a
+     * single malloc to get the memory for the array and all the pages needed
+     * by the array. While reading in the toUnicode array, we remembered what
+     * pages that would be needed for the fromUnicode array.
+     */
+
+    if (symbol) {
+	used[0] = 1;
+    }
+    numPages = 0;
+    for (hi = 0; hi < 256; hi++) {
+	if (used[hi]) {
+	    numPages++;
+	}
+    }
+    size = 256 * sizeof(unsigned short *) + numPages * PAGESIZE;
+    dataPtr->fromUnicode = ckalloc(size);
+    memset(dataPtr->fromUnicode, 0, size);
+    pageMemPtr = (unsigned short *) (dataPtr->fromUnicode + 256);
+
+    for (hi = 0; hi < 256; hi++) {
+	if (dataPtr->toUnicode[hi] == NULL) {
+	    dataPtr->toUnicode[hi] = emptyPage;
+	    continue;
+	}
+	for (lo = 0; lo < 256; lo++) {
+	    int ch = dataPtr->toUnicode[hi][lo];
+
+	    if (ch != 0) {
+		page = dataPtr->fromUnicode[ch >> 8];
+		if (page == NULL) {
+		    page = pageMemPtr;
+		    pageMemPtr += 256;
+		    dataPtr->fromUnicode[ch >> 8] = page;
+		}
+		page[ch & 0xff] = (unsigned short) ((hi << 8) + lo);
+	    }
+	}
+    }
+    if (type == ENCODING_MULTIBYTE) {
+	/*
+	 * If multibyte encodings don't have a backslash character, define
+	 * one. Otherwise, on Windows, native file names won't work because
+	 * the backslash in the file name will map to the unknown character
+	 * (question mark) when converting from UTF-8 to external encoding.
+	 */
+
+	if (dataPtr->fromUnicode[0] != NULL) {
+	    if (dataPtr->fromUnicode[0]['\\'] == '\0') {
+		dataPtr->fromUnicode[0]['\\'] = '\\';
+	    }
+	}
+    }
+    if (symbol) {
+	/*
+	 * Make a special symbol encoding that not only maps the symbol
+	 * characters from their Unicode code points down into page 0, but
+	 * also ensure that the characters on page 0 map to themselves. This
+	 * is so that a symbol font can be used to display a simple string
+	 * like "abcd" and have alpha, beta, chi, delta show up, rather than
+	 * have "unknown" chars show up because strictly speaking the symbol
+	 * font doesn't have glyphs for those low ASCII chars.
+	 */
+
+	page = dataPtr->fromUnicode[0];
+	if (page == NULL) {
+	    page = pageMemPtr;
+	    dataPtr->fromUnicode[0] = page;
+	}
+	for (lo = 0; lo < 256; lo++) {
+	    if (dataPtr->toUnicode[0][lo] != 0) {
+		page[lo] = (unsigned short) lo;
+	    }
+	}
+    }
+    for (hi = 0; hi < 256; hi++) {
+	if (dataPtr->fromUnicode[hi] == NULL) {
+	    dataPtr->fromUnicode[hi] = emptyPage;
+	}
+    }
+
+    /*
+     * For trailing 'R'everse encoding, see [Patch 689341]
+     */
+
+    Tcl_DStringInit(&lineString);
+
+    /*
+     * Skip leading empty lines.
+     */
+
+    while ((len = Tcl_Gets(chan, &lineString)) == 0) {
+	/* empty body */
+    }
+    if (len < 0) {
+	goto doneParse;
+    }
+
+    /*
+     * Require that it starts with an 'R'.
+     */
+
+    line = Tcl_DStringValue(&lineString);
+    if (line[0] != 'R') {
+	goto doneParse;
+    }
+
+    /*
+     * Read lines from the encoding until EOF.
+     */
+
+    for (TclDStringClear(&lineString);
+	    (len = Tcl_Gets(chan, &lineString)) >= 0;
+	    TclDStringClear(&lineString)) {
+	const unsigned char *p;
+	int to, from;
+
+	/*
+	 * Skip short lines.
+	 */
+
+	if (len < 5) {
+	    continue;
+	}
+
+	/*
+	 * Parse the line as a sequence of hex digits.
+	 */
+
+	p = (const unsigned char *) Tcl_DStringValue(&lineString);
+	to = (staticHex[p[0]] << 12) + (staticHex[p[1]] << 8)
+		+ (staticHex[p[2]] << 4) + staticHex[p[3]];
+	if (to == 0) {
+	    continue;
+	}
+	for (p += 5, len -= 5; len >= 0 && *p; p += 5, len -= 5) {
+	    from = (staticHex[p[0]] << 12) + (staticHex[p[1]] << 8)
+		    + (staticHex[p[2]] << 4) + staticHex[p[3]];
+	    if (from == 0) {
+		continue;
+	    }
+	    dataPtr->fromUnicode[from >> 8][from & 0xff] = to;
+	}
+    }
+  doneParse:
+    Tcl_DStringFree(&lineString);
+
+    /*
+     * Package everything into an encoding structure.
+     */
+
+    encType.encodingName    = name;
+    encType.toUtfProc	    = TableToUtfProc;
+    encType.fromUtfProc	    = TableFromUtfProc;
+    encType.freeProc	    = TableFreeProc;
+    encType.nullSize	    = (type == ENCODING_DOUBLEBYTE) ? 2 : 1;
+    encType.clientData	    = dataPtr;
+
+    return Tcl_CreateEncoding(&encType);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * LoadEscapeEncoding --
+ *
+ *	Helper function for LoadEncodingTable(). Loads a state machine that
+ *	converts between Unicode and some other encoding.
+ *
+ *	File contains text data that describes the escape sequences that are
+ *	used to choose an encoding and the associated names for the
+ *	sub-encodings.
+ *
+ * Results:
+ *	The return value is the new encoding, or NULL if the encoding could
+ *	not be created (because the file contained invalid data).
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static Tcl_Encoding
+LoadEscapeEncoding(
+    const char *name,		/* Name for new encoding. */
+    Tcl_Channel chan)		/* File containing new encoding. */
+{
+    int i;
+    unsigned size;
+    Tcl_DString escapeData;
+    char init[16], final[16];
+    EscapeEncodingData *dataPtr;
+    Tcl_EncodingType type;
+
+    init[0] = '\0';
+    final[0] = '\0';
+    Tcl_DStringInit(&escapeData);
+
+    while (1) {
+	int argc;
+	const char **argv;
+	char *line;
+	Tcl_DString lineString;
+
+	Tcl_DStringInit(&lineString);
+	if (Tcl_Gets(chan, &lineString) < 0) {
+	    break;
+	}
+	line = Tcl_DStringValue(&lineString);
+	if (Tcl_SplitList(NULL, line, &argc, &argv) != TCL_OK) {
+	    Tcl_DStringFree(&lineString);
+	    continue;
+	}
+	if (argc >= 2) {
+	    if (strcmp(argv[0], "name") == 0) {
+		/* do nothing */
+	    } else if (strcmp(argv[0], "init") == 0) {
+		strncpy(init, argv[1], sizeof(init));
+		init[sizeof(init) - 1] = '\0';
+	    } else if (strcmp(argv[0], "final") == 0) {
+		strncpy(final, argv[1], sizeof(final));
+		final[sizeof(final) - 1] = '\0';
+	    } else {
+		EscapeSubTable est;
+		Encoding *e;
+
+		strncpy(est.sequence, argv[1], sizeof(est.sequence));
+		est.sequence[sizeof(est.sequence) - 1] = '\0';
+		est.sequenceLen = strlen(est.sequence);
+
+		strncpy(est.name, argv[0], sizeof(est.name));
+		est.name[sizeof(est.name) - 1] = '\0';
+
+		/*
+		 * To avoid infinite recursion in [encoding system iso2022-*]
+		 */
+
+		e = (Encoding *) Tcl_GetEncoding(NULL, est.name);
+		if ((e != NULL) && (e->toUtfProc != TableToUtfProc)
+			&& (e->toUtfProc != Iso88591ToUtfProc)) {
+		   Tcl_FreeEncoding((Tcl_Encoding) e);
+		   e = NULL;
+		}
+		est.encodingPtr = e;
+		Tcl_DStringAppend(&escapeData, (char *) &est, sizeof(est));
+	    }
+	}
+	ckfree(argv);
+	Tcl_DStringFree(&lineString);
+    }
+
+    size = sizeof(EscapeEncodingData) - sizeof(EscapeSubTable)
+	    + Tcl_DStringLength(&escapeData);
+    dataPtr = ckalloc(size);
+    dataPtr->initLen = strlen(init);
+    memcpy(dataPtr->init, init, (unsigned) dataPtr->initLen + 1);
+    dataPtr->finalLen = strlen(final);
+    memcpy(dataPtr->final, final, (unsigned) dataPtr->finalLen + 1);
+    dataPtr->numSubTables =
+	    Tcl_DStringLength(&escapeData) / sizeof(EscapeSubTable);
+    memcpy(dataPtr->subTables, Tcl_DStringValue(&escapeData),
+	    (size_t) Tcl_DStringLength(&escapeData));
+    Tcl_DStringFree(&escapeData);
+
+    memset(dataPtr->prefixBytes, 0, sizeof(dataPtr->prefixBytes));
+    for (i = 0; i < dataPtr->numSubTables; i++) {
+	dataPtr->prefixBytes[UCHAR(dataPtr->subTables[i].sequence[0])] = 1;
+    }
+    if (dataPtr->init[0] != '\0') {
+	dataPtr->prefixBytes[UCHAR(dataPtr->init[0])] = 1;
+    }
+    if (dataPtr->final[0] != '\0') {
+	dataPtr->prefixBytes[UCHAR(dataPtr->final[0])] = 1;
+    }
+
+    /*
+     * Package everything into an encoding structure.
+     */
+
+    type.encodingName	= name;
+    type.toUtfProc	= EscapeToUtfProc;
+    type.fromUtfProc    = EscapeFromUtfProc;
+    type.freeProc	= EscapeFreeProc;
+    type.nullSize	= 1;
+    type.clientData	= dataPtr;
+
+    return Tcl_CreateEncoding(&type);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * BinaryProc --
+ *
+ *	The default conversion when no other conversion is specified. No
+ *	translation is done; source bytes are copied directly to destination
+ *	bytes.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+BinaryProc(
+    ClientData clientData,	/* Not used. */
+    const char *src,		/* Source string (unknown encoding). */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    int result;
+
+    result = TCL_OK;
+    dstLen -= TCL_UTF_MAX - 1;
+    if (dstLen < 0) {
+	dstLen = 0;
+    }
+    if ((flags & TCL_ENCODING_CHAR_LIMIT) && srcLen > *dstCharsPtr) {
+	srcLen = *dstCharsPtr;
+    }
+    if (srcLen > dstLen) {
+	srcLen = dstLen;
+	result = TCL_CONVERT_NOSPACE;
+    }
+
+    *srcReadPtr = srcLen;
+    *dstWrotePtr = srcLen;
+    *dstCharsPtr = srcLen;
+    memcpy(dst, src, (size_t) srcLen);
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * UtfExtToUtfIntProc --
+ *
+ *	Convert from UTF-8 to UTF-8. While converting null-bytes from the
+ *	Tcl's internal representation (0xc0, 0x80) to the official
+ *	representation (0x00). See UtfToUtfProc for details.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+UtfIntToUtfExtProc(
+    ClientData clientData,	/* Not used. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string
+				 * is stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    return UtfToUtfProc(clientData, src, srcLen, flags, statePtr, dst, dstLen,
+	    srcReadPtr, dstWrotePtr, dstCharsPtr, 1);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * UtfExtToUtfIntProc --
+ *
+ *	Convert from UTF-8 to UTF-8 while converting null-bytes from the
+ *	official representation (0x00) to Tcl's internal representation (0xc0,
+ *	0x80). See UtfToUtfProc for details.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+UtfExtToUtfIntProc(
+    ClientData clientData,	/* Not used. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    return UtfToUtfProc(clientData, src, srcLen, flags, statePtr, dst, dstLen,
+	    srcReadPtr, dstWrotePtr, dstCharsPtr, 0);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * UtfToUtfProc --
+ *
+ *	Convert from UTF-8 to UTF-8. Note that the UTF-8 to UTF-8 translation
+ *	is not a no-op, because it will turn a stream of improperly formed
+ *	UTF-8 into a properly formed stream.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+UtfToUtfProc(
+    ClientData clientData,	/* Not used. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr,		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+    int pureNullMode)		/* Convert embedded nulls from internal
+				 * representation to real null-bytes or vice
+				 * versa. */
+{
+    const char *srcStart, *srcEnd, *srcClose;
+    const char *dstStart, *dstEnd;
+    int result, numChars, charLimit = INT_MAX;
+    Tcl_UniChar ch = 0;
+
+    result = TCL_OK;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+    srcClose = srcEnd;
+    if ((flags & TCL_ENCODING_END) == 0) {
+	srcClose -= TCL_UTF_MAX;
+    }
+    if (flags & TCL_ENCODING_CHAR_LIMIT) {
+	charLimit = *dstCharsPtr;
+    }
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - TCL_UTF_MAX;
+
+    for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
+	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
+	    /*
+	     * If there is more string to follow, this will ensure that the
+	     * last UTF-8 character in the source buffer hasn't been cut off.
+	     */
+
+	    result = TCL_CONVERT_MULTIBYTE;
+	    break;
+	}
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	if (UCHAR(*src) < 0x80 && !(UCHAR(*src) == 0 && pureNullMode == 0)) {
+	    /*
+	     * Copy 7bit chatacters, but skip null-bytes when we are in input
+	     * mode, so that they get converted to 0xc080.
+	     */
+
+	    *dst++ = *src++;
+	} else if (pureNullMode == 1 && UCHAR(*src) == 0xc0 &&
+		(src + 1 < srcEnd) && UCHAR(*(src+1)) == 0x80) {
+	    /*
+	     * Convert 0xc080 to real nulls when we are in output mode.
+	     */
+
+	    *dst++ = 0;
+	    src += 2;
+	} else if (!Tcl_UtfCharComplete(src, srcEnd - src)) {
+	    /*
+	     * Always check before using TclUtfToUniChar. Not doing can so
+	     * cause it run beyond the end of the buffer! If we happen such an
+	     * incomplete char its bytes are made to represent themselves.
+	     */
+
+	    ch = (unsigned char) *src;
+	    src += 1;
+	    dst += Tcl_UniCharToUtf(ch, dst);
+	} else {
+	    src += TclUtfToUniChar(src, &ch);
+	    dst += Tcl_UniCharToUtf(ch, dst);
+	}
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * UnicodeToUtfProc --
+ *
+ *	Convert from Unicode to UTF-8.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+UnicodeToUtfProc(
+    ClientData clientData,	/* Not used. */
+    const char *src,		/* Source string in Unicode. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd;
+    const char *dstEnd, *dstStart;
+    int result, numChars, charLimit = INT_MAX;
+    Tcl_UniChar ch = 0;
+
+    if (flags & TCL_ENCODING_CHAR_LIMIT) {
+	charLimit = *dstCharsPtr;
+    }
+    result = TCL_OK;
+    if ((srcLen % sizeof(Tcl_UniChar)) != 0) {
+	result = TCL_CONVERT_MULTIBYTE;
+	srcLen /= sizeof(Tcl_UniChar);
+	srcLen *= sizeof(Tcl_UniChar);
+    }
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - TCL_UTF_MAX;
+
+    for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+
+	/*
+	 * Special case for 1-byte utf chars for speed. Make sure we work with
+	 * Tcl_UniChar-size data.
+	 */
+
+	ch = *(Tcl_UniChar *)src;
+	if (ch && ch < 0x80) {
+	    *dst++ = (ch & 0xFF);
+	} else {
+	    dst += Tcl_UniCharToUtf(ch, dst);
+	}
+	src += sizeof(Tcl_UniChar);
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * UtfToUnicodeProc --
+ *
+ *	Convert from UTF-8 to Unicode.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+UtfToUnicodeProc(
+    ClientData clientData,	/* TableEncodingData that specifies
+				 * encoding. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd, *srcClose, *dstStart, *dstEnd;
+    int result, numChars;
+    Tcl_UniChar ch = 0;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+    srcClose = srcEnd;
+    if ((flags & TCL_ENCODING_END) == 0) {
+	srcClose -= TCL_UTF_MAX;
+    }
+
+    dstStart = dst;
+    dstEnd   = dst + dstLen - sizeof(Tcl_UniChar);
+
+    result = TCL_OK;
+    for (numChars = 0; src < srcEnd; numChars++) {
+	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
+	    /*
+	     * If there is more string to follow, this will ensure that the
+	     * last UTF-8 character in the source buffer hasn't been cut off.
+	     */
+
+	    result = TCL_CONVERT_MULTIBYTE;
+	    break;
+	}
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	src += TclUtfToUniChar(src, &ch);
+
+	/*
+	 * Need to handle this in a way that won't cause misalignment by
+	 * casting dst to a Tcl_UniChar. [Bug 1122671]
+	 */
+
+#ifdef WORDS_BIGENDIAN
+#if TCL_UTF_MAX > 4
+	*dst++ = (ch >> 24);
+	*dst++ = ((ch >> 16) & 0xFF);
+	*dst++ = ((ch >> 8) & 0xFF);
+	*dst++ = (ch & 0xFF);
+#else
+	*dst++ = (ch >> 8);
+	*dst++ = (ch & 0xFF);
+#endif
+#else
+#if TCL_UTF_MAX > 4
+	*dst++ = (ch & 0xFF);
+	*dst++ = ((ch >> 8) & 0xFF);
+	*dst++ = ((ch >> 16) & 0xFF);
+	*dst++ = (ch >> 24);
+#else
+	*dst++ = (ch & 0xFF);
+	*dst++ = (ch >> 8);
+#endif
+#endif
+    }
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TableToUtfProc --
+ *
+ *	Convert from the encoding specified by the TableEncodingData into
+ *	UTF-8.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+TableToUtfProc(
+    ClientData clientData,	/* TableEncodingData that specifies
+				 * encoding. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd;
+    const char *dstEnd, *dstStart, *prefixBytes;
+    int result, byte, numChars, charLimit = INT_MAX;
+    Tcl_UniChar ch = 0;
+    const unsigned short *const *toUnicode;
+    const unsigned short *pageZero;
+    TableEncodingData *dataPtr = clientData;
+
+    if (flags & TCL_ENCODING_CHAR_LIMIT) {
+	charLimit = *dstCharsPtr;
+    }
+    srcStart = src;
+    srcEnd = src + srcLen;
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - TCL_UTF_MAX;
+
+    toUnicode = (const unsigned short *const *) dataPtr->toUnicode;
+    prefixBytes = dataPtr->prefixBytes;
+    pageZero = toUnicode[0];
+
+    result = TCL_OK;
+    for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	byte = *((unsigned char *) src);
+	if (prefixBytes[byte]) {
+	    src++;
+	    if (src >= srcEnd) {
+		src--;
+		result = TCL_CONVERT_MULTIBYTE;
+		break;
+	    }
+	    ch = toUnicode[byte][*((unsigned char *) src)];
+	} else {
+	    ch = pageZero[byte];
+	}
+	if ((ch == 0) && (byte != 0)) {
+	    if (flags & TCL_ENCODING_STOPONERROR) {
+		result = TCL_CONVERT_SYNTAX;
+		break;
+	    }
+	    if (prefixBytes[byte]) {
+		src--;
+	    }
+	    ch = (Tcl_UniChar) byte;
+	}
+
+	/*
+	 * Special case for 1-byte utf chars for speed.
+	 */
+
+	if (ch && ch < 0x80) {
+	    *dst++ = (char) ch;
+	} else {
+	    dst += Tcl_UniCharToUtf(ch, dst);
+	}
+	src++;
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TableFromUtfProc --
+ *
+ *	Convert from UTF-8 into the encoding specified by the
+ *	TableEncodingData.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+TableFromUtfProc(
+    ClientData clientData,	/* TableEncodingData that specifies
+				 * encoding. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd, *srcClose;
+    const char *dstStart, *dstEnd, *prefixBytes;
+    Tcl_UniChar ch = 0;
+    int result, len, word, numChars;
+    TableEncodingData *dataPtr = clientData;
+    const unsigned short *const *fromUnicode;
+
+    result = TCL_OK;
+
+    prefixBytes = dataPtr->prefixBytes;
+    fromUnicode = (const unsigned short *const *) dataPtr->fromUnicode;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+    srcClose = srcEnd;
+    if ((flags & TCL_ENCODING_END) == 0) {
+	srcClose -= TCL_UTF_MAX;
+    }
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - 1;
+
+    for (numChars = 0; src < srcEnd; numChars++) {
+	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
+	    /*
+	     * If there is more string to follow, this will ensure that the
+	     * last UTF-8 character in the source buffer hasn't been cut off.
+	     */
+
+	    result = TCL_CONVERT_MULTIBYTE;
+	    break;
+	}
+	len = TclUtfToUniChar(src, &ch);
+
+#if TCL_UTF_MAX > 3
+	/*
+	 * This prevents a crash condition. More evaluation is required for
+	 * full support of int Tcl_UniChar. [Bug 1004065]
+	 */
+
+	if (ch & 0xffff0000) {
+	    word = 0;
+	} else
+#endif
+	    word = fromUnicode[(ch >> 8)][ch & 0xff];
+
+	if ((word == 0) && (ch != 0)) {
+	    if (flags & TCL_ENCODING_STOPONERROR) {
+		result = TCL_CONVERT_UNKNOWN;
+		break;
+	    }
+	    word = dataPtr->fallback;
+	}
+	if (prefixBytes[(word >> 8)] != 0) {
+	    if (dst + 1 > dstEnd) {
+		result = TCL_CONVERT_NOSPACE;
+		break;
+	    }
+	    dst[0] = (char) (word >> 8);
+	    dst[1] = (char) word;
+	    dst += 2;
+	} else {
+	    if (dst > dstEnd) {
+		result = TCL_CONVERT_NOSPACE;
+		break;
+	    }
+	    dst[0] = (char) word;
+	    dst++;
+	}
+	src += len;
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Iso88591ToUtfProc --
+ *
+ *	Convert from the "iso8859-1" encoding into UTF-8.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+Iso88591ToUtfProc(
+    ClientData clientData,	/* Ignored. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd;
+    const char *dstEnd, *dstStart;
+    int result, numChars, charLimit = INT_MAX;
+
+    if (flags & TCL_ENCODING_CHAR_LIMIT) {
+	charLimit = *dstCharsPtr;
+    }
+    srcStart = src;
+    srcEnd = src + srcLen;
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - TCL_UTF_MAX;
+
+    result = TCL_OK;
+    for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
+	Tcl_UniChar ch = 0;
+
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	ch = (Tcl_UniChar) *((unsigned char *) src);
+
+	/*
+	 * Special case for 1-byte utf chars for speed.
+	 */
+
+	if (ch && ch < 0x80) {
+	    *dst++ = (char) ch;
+	} else {
+	    dst += Tcl_UniCharToUtf(ch, dst);
+	}
+	src++;
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Iso88591FromUtfProc --
+ *
+ *	Convert from UTF-8 into the encoding "iso8859-1".
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+Iso88591FromUtfProc(
+    ClientData clientData,	/* Ignored. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    const char *srcStart, *srcEnd, *srcClose;
+    const char *dstStart, *dstEnd;
+    int result, numChars;
+
+    result = TCL_OK;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+    srcClose = srcEnd;
+    if ((flags & TCL_ENCODING_END) == 0) {
+	srcClose -= TCL_UTF_MAX;
+    }
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - 1;
+
+    for (numChars = 0; src < srcEnd; numChars++) {
+	Tcl_UniChar ch = 0;
+	int len;
+
+	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
+	    /*
+	     * If there is more string to follow, this will ensure that the
+	     * last UTF-8 character in the source buffer hasn't been cut off.
+	     */
+
+	    result = TCL_CONVERT_MULTIBYTE;
+	    break;
+	}
+	len = TclUtfToUniChar(src, &ch);
+
+	/*
+	 * Check for illegal characters.
+	 */
+
+	if (ch > 0xff) {
+	    if (flags & TCL_ENCODING_STOPONERROR) {
+		result = TCL_CONVERT_UNKNOWN;
+		break;
+	    }
+
+	    /*
+	     * Plunge on, using '?' as a fallback character.
+	     */
+
+	    ch = (Tcl_UniChar) '?';
+	}
+
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	*(dst++) = (char) ch;
+	src += len;
+    }
+
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TableFreeProc --
+ *
+ *	This function is invoked when an encoding is deleted. It deletes the
+ *	memory used by the TableEncodingData.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory freed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+TableFreeProc(
+    ClientData clientData)	/* TableEncodingData that specifies
+				 * encoding. */
+{
+    TableEncodingData *dataPtr = clientData;
+
+    /*
+     * Make sure we aren't freeing twice on shutdown. [Bug 219314]
+     */
+
+    ckfree(dataPtr->toUnicode);
+    dataPtr->toUnicode = NULL;
+    ckfree(dataPtr->fromUnicode);
+    dataPtr->fromUnicode = NULL;
+    ckfree(dataPtr);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * EscapeToUtfProc --
+ *
+ *	Convert from the encoding specified by the EscapeEncodingData into
+ *	UTF-8.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+EscapeToUtfProc(
+    ClientData clientData,	/* EscapeEncodingData that specifies
+				 * encoding. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    EscapeEncodingData *dataPtr = clientData;
+    const char *prefixBytes, *tablePrefixBytes, *srcStart, *srcEnd;
+    const unsigned short *const *tableToUnicode;
+    const Encoding *encodingPtr;
+    int state, result, numChars, charLimit = INT_MAX;
+    const char *dstStart, *dstEnd;
+
+    if (flags & TCL_ENCODING_CHAR_LIMIT) {
+	charLimit = *dstCharsPtr;
+    }
+    result = TCL_OK;
+    tablePrefixBytes = NULL;	/* lint. */
+    tableToUnicode = NULL;	/* lint. */
+    prefixBytes = dataPtr->prefixBytes;
+    encodingPtr = NULL;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - TCL_UTF_MAX;
+
+    state = PTR2INT(*statePtr);
+    if (flags & TCL_ENCODING_START) {
+	state = 0;
+    }
+
+    for (numChars = 0; src < srcEnd && numChars <= charLimit; ) {
+	int byte, hi, lo, ch;
+
+	if (dst > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	    break;
+	}
+	byte = *((unsigned char *) src);
+	if (prefixBytes[byte]) {
+	    unsigned left, len, longest;
+	    int checked, i;
+	    const EscapeSubTable *subTablePtr;
+
+	    /*
+	     * Saw the beginning of an escape sequence.
+	     */
+
+	    left = srcEnd - src;
+	    len = dataPtr->initLen;
+	    longest = len;
+	    checked = 0;
+
+	    if (len <= left) {
+		checked++;
+		if ((len > 0) && (memcmp(src, dataPtr->init, len) == 0)) {
+		    /*
+		     * If we see initialization string, skip it, even if we're
+		     * not at the beginning of the buffer.
+		     */
+
+		    src += len;
+		    continue;
+		}
+	    }
+
+	    len = dataPtr->finalLen;
+	    if (len > longest) {
+		longest = len;
+	    }
+
+	    if (len <= left) {
+		checked++;
+		if ((len > 0) && (memcmp(src, dataPtr->final, len) == 0)) {
+		    /*
+		     * If we see finalization string, skip it, even if we're
+		     * not at the end of the buffer.
+		     */
+
+		    src += len;
+		    continue;
+		}
+	    }
+
+	    subTablePtr = dataPtr->subTables;
+	    for (i = 0; i < dataPtr->numSubTables; i++) {
+		len = subTablePtr->sequenceLen;
+		if (len > longest) {
+		    longest = len;
+		}
+		if (len <= left) {
+		    checked++;
+		    if ((len > 0) &&
+			    (memcmp(src, subTablePtr->sequence, len) == 0)) {
+			state = i;
+			encodingPtr = NULL;
+			subTablePtr = NULL;
+			src += len;
+			break;
+		    }
+		}
+		subTablePtr++;
+	    }
+
+	    if (subTablePtr == NULL) {
+		/*
+		 * A match was found, the escape sequence was consumed, and
+		 * the state was updated.
+		 */
+
+		continue;
+	    }
+
+	    /*
+	     * We have a split-up or unrecognized escape sequence. If we
+	     * checked all the sequences, then it's a syntax error, otherwise
+	     * we need more bytes to determine a match.
+	     */
+
+	    if ((checked == dataPtr->numSubTables + 2)
+		    || (flags & TCL_ENCODING_END)) {
+		if ((flags & TCL_ENCODING_STOPONERROR) == 0) {
+		    /*
+		     * Skip the unknown escape sequence.
+		     */
+
+		    src += longest;
+		    continue;
+		}
+		result = TCL_CONVERT_SYNTAX;
+	    } else {
+		result = TCL_CONVERT_MULTIBYTE;
+	    }
+	    break;
+	}
+
+	if (encodingPtr == NULL) {
+	    TableEncodingData *tableDataPtr;
+
+	    encodingPtr = GetTableEncoding(dataPtr, state);
+	    tableDataPtr = encodingPtr->clientData;
+	    tablePrefixBytes = tableDataPtr->prefixBytes;
+	    tableToUnicode = (const unsigned short *const*)
+		    tableDataPtr->toUnicode;
+	}
+
+	if (tablePrefixBytes[byte]) {
+	    src++;
+	    if (src >= srcEnd) {
+		src--;
+		result = TCL_CONVERT_MULTIBYTE;
+		break;
+	    }
+	    hi = byte;
+	    lo = *((unsigned char *) src);
+	} else {
+	    hi = 0;
+	    lo = byte;
+	}
+
+	ch = tableToUnicode[hi][lo];
+	dst += Tcl_UniCharToUtf(ch, dst);
+	src++;
+	numChars++;
+    }
+
+    *statePtr = (Tcl_EncodingState) INT2PTR(state);
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * EscapeFromUtfProc --
+ *
+ *	Convert from UTF-8 into the encoding specified by the
+ *	EscapeEncodingData.
+ *
+ * Results:
+ *	Returns TCL_OK if conversion was successful.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static int
+EscapeFromUtfProc(
+    ClientData clientData,	/* EscapeEncodingData that specifies
+				 * encoding. */
+    const char *src,		/* Source string in UTF-8. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Place for conversion routine to store state
+				 * information used during a piecewise
+				 * conversion. Contents of statePtr are
+				 * initialized and/or reset by conversion
+				 * routine under control of flags argument. */
+    char *dst,			/* Output buffer in which converted string is
+				 * stored. */
+    int dstLen,			/* The maximum length of output buffer in
+				 * bytes. */
+    int *srcReadPtr,		/* Filled with the number of bytes from the
+				 * source string that were converted. This may
+				 * be less than the original source length if
+				 * there was a problem converting some source
+				 * characters. */
+    int *dstWrotePtr,		/* Filled with the number of bytes that were
+				 * stored in the output buffer as a result of
+				 * the conversion. */
+    int *dstCharsPtr)		/* Filled with the number of characters that
+				 * correspond to the bytes stored in the
+				 * output buffer. */
+{
+    EscapeEncodingData *dataPtr = clientData;
+    const Encoding *encodingPtr;
+    const char *srcStart, *srcEnd, *srcClose;
+    const char *dstStart, *dstEnd;
+    int state, result, numChars;
+    const TableEncodingData *tableDataPtr;
+    const char *tablePrefixBytes;
+    const unsigned short *const *tableFromUnicode;
+
+    result = TCL_OK;
+
+    srcStart = src;
+    srcEnd = src + srcLen;
+    srcClose = srcEnd;
+    if ((flags & TCL_ENCODING_END) == 0) {
+	srcClose -= TCL_UTF_MAX;
+    }
+
+    dstStart = dst;
+    dstEnd = dst + dstLen - 1;
+
+    /*
+     * RFC 1468 states that the text starts in ASCII, and switches to Japanese
+     * characters, and that the text must end in ASCII. [Patch 474358]
+     */
+
+    if (flags & TCL_ENCODING_START) {
+	state = 0;
+	if ((dst + dataPtr->initLen) > dstEnd) {
+	    *srcReadPtr = 0;
+	    *dstWrotePtr = 0;
+	    return TCL_CONVERT_NOSPACE;
+	}
+	memcpy(dst, dataPtr->init, (size_t)dataPtr->initLen);
+	dst += dataPtr->initLen;
+    } else {
+	state = PTR2INT(*statePtr);
+    }
+
+    encodingPtr = GetTableEncoding(dataPtr, state);
+    tableDataPtr = encodingPtr->clientData;
+    tablePrefixBytes = tableDataPtr->prefixBytes;
+    tableFromUnicode = (const unsigned short *const *)
+	    tableDataPtr->fromUnicode;
+
+    for (numChars = 0; src < srcEnd; numChars++) {
+	unsigned len;
+	int word;
+	Tcl_UniChar ch = 0;
+
+	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
+	    /*
+	     * If there is more string to follow, this will ensure that the
+	     * last UTF-8 character in the source buffer hasn't been cut off.
+	     */
+
+	    result = TCL_CONVERT_MULTIBYTE;
+	    break;
+	}
+	len = TclUtfToUniChar(src, &ch);
+	word = tableFromUnicode[(ch >> 8)][ch & 0xff];
+
+	if ((word == 0) && (ch != 0)) {
+	    int oldState;
+	    const EscapeSubTable *subTablePtr;
+
+	    oldState = state;
+	    for (state = 0; state < dataPtr->numSubTables; state++) {
+		encodingPtr = GetTableEncoding(dataPtr, state);
+		tableDataPtr = encodingPtr->clientData;
+		word = tableDataPtr->fromUnicode[(ch >> 8)][ch & 0xff];
+		if (word != 0) {
+		    break;
+		}
+	    }
+
+	    if (word == 0) {
+		state = oldState;
+		if (flags & TCL_ENCODING_STOPONERROR) {
+		    result = TCL_CONVERT_UNKNOWN;
+		    break;
+		}
+		encodingPtr = GetTableEncoding(dataPtr, state);
+		tableDataPtr = encodingPtr->clientData;
+		word = tableDataPtr->fallback;
+	    }
+
+	    tablePrefixBytes = (const char *) tableDataPtr->prefixBytes;
+	    tableFromUnicode = (const unsigned short *const *)
+		    tableDataPtr->fromUnicode;
+
+	    /*
+	     * The state variable has the value of oldState when word is 0.
+	     * In this case, the escape sequense should not be copied to dst
+	     * because the current character set is not changed.
+	     */
+
+	    if (state != oldState) {
+		subTablePtr = &dataPtr->subTables[state];
+		if ((dst + subTablePtr->sequenceLen) > dstEnd) {
+		    /*
+		     * If there is no space to write the escape sequence, the
+		     * state variable must be changed to the value of oldState
+		     * variable because this escape sequence must be written
+		     * in the next conversion.
+		     */
+
+		    state = oldState;
+		    result = TCL_CONVERT_NOSPACE;
+		    break;
+		}
+		memcpy(dst, subTablePtr->sequence,
+			(size_t) subTablePtr->sequenceLen);
+		dst += subTablePtr->sequenceLen;
+	    }
+	}
+
+	if (tablePrefixBytes[(word >> 8)] != 0) {
+	    if (dst + 1 > dstEnd) {
+		result = TCL_CONVERT_NOSPACE;
+		break;
+	    }
+	    dst[0] = (char) (word >> 8);
+	    dst[1] = (char) word;
+	    dst += 2;
+	} else {
+	    if (dst > dstEnd) {
+		result = TCL_CONVERT_NOSPACE;
+		break;
+	    }
+	    dst[0] = (char) word;
+	    dst++;
+	}
+	src += len;
+    }
+
+    if ((result == TCL_OK) && (flags & TCL_ENCODING_END)) {
+	unsigned len = dataPtr->subTables[0].sequenceLen;
+
+	/*
+	 * Certain encodings like iso2022-jp need to write an escape sequence
+	 * after all characters have been converted. This logic checks that
+	 * enough room is available in the buffer for the escape bytes. The
+	 * TCL_ENCODING_END flag is cleared after a final escape sequence has
+	 * been added to the buffer so that another call to this method does
+	 * not attempt to append escape bytes a second time.
+	 */
+
+	if ((dst + dataPtr->finalLen + (state?len:0)) > dstEnd) {
+	    result = TCL_CONVERT_NOSPACE;
+	} else {
+	    if (state) {
+		memcpy(dst, dataPtr->subTables[0].sequence, len);
+		dst += len;
+	    }
+	    memcpy(dst, dataPtr->final, (size_t) dataPtr->finalLen);
+	    dst += dataPtr->finalLen;
+	    state &= ~TCL_ENCODING_END;
+	}
+    }
+
+    *statePtr = (Tcl_EncodingState) INT2PTR(state);
+    *srcReadPtr = src - srcStart;
+    *dstWrotePtr = dst - dstStart;
+    *dstCharsPtr = numChars;
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * EscapeFreeProc --
+ *
+ *	This function is invoked when an EscapeEncodingData encoding is
+ *	deleted. It deletes the memory used by the encoding.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory freed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+EscapeFreeProc(
+    ClientData clientData)	/* EscapeEncodingData that specifies
+				 * encoding. */
+{
+    EscapeEncodingData *dataPtr = clientData;
+    EscapeSubTable *subTablePtr;
+    int i;
+
+    if (dataPtr == NULL) {
+	return;
+    }
+
+    /*
+     * The subTables should be freed recursively in normal operation but not
+     * during TclFinalizeEncodingSubsystem because they are also present as a
+     * weak reference in the toplevel encodingTable (i.e., they don't have a
+     * +1 refcount for this), and unpredictable nuking order could remove them
+     * from under the following loop's feet. [Bug 2891556]
+     *
+     * The encodingsInitialized flag, being reset on entry to TFES, can serve
+     * as a "not in finalization" test.
+     */
+
+    if (encodingsInitialized) {
+	subTablePtr = dataPtr->subTables;
+	for (i = 0; i < dataPtr->numSubTables; i++) {
+	    FreeEncoding((Tcl_Encoding) subTablePtr->encodingPtr);
+	    subTablePtr->encodingPtr = NULL;
+	    subTablePtr++;
+	}
+    }
+    ckfree(dataPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * GetTableEncoding --
+ *
+ *	Helper function for the EscapeEncodingData conversions. Gets the
+ *	encoding (of type TextEncodingData) that represents the specified
+ *	state.
+ *
+ * Results:
+ *	The return value is the encoding.
+ *
+ * Side effects:
+ *	If the encoding that represents the specified state has not already
+ *	been used by this EscapeEncoding, it will be loaded and cached in the
+ *	dataPtr.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Encoding *
+GetTableEncoding(
+    EscapeEncodingData *dataPtr,/* Contains names of encodings. */
+    int state)			/* Index in dataPtr of desired Encoding. */
+{
+    EscapeSubTable *subTablePtr = &dataPtr->subTables[state];
+    Encoding *encodingPtr = subTablePtr->encodingPtr;
+
+    if (encodingPtr == NULL) {
+	encodingPtr = (Encoding *) Tcl_GetEncoding(NULL, subTablePtr->name);
+	if ((encodingPtr == NULL)
+		|| (encodingPtr->toUtfProc != TableToUtfProc
+		&& encodingPtr->toUtfProc != Iso88591ToUtfProc)) {
+	    Tcl_Panic("EscapeToUtfProc: invalid sub table");
+	}
+	subTablePtr->encodingPtr = encodingPtr;
+    }
+
+    return encodingPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * unilen --
+ *
+ *	A helper function for the Tcl_ExternalToUtf functions. This function
+ *	is similar to strlen for double-byte characters: it returns the number
+ *	of bytes in a 0x0000 terminated string.
+ *
+ * Results:
+ *	As above.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static size_t
+unilen(
+    const char *src)
+{
+    unsigned short *p;
+
+    p = (unsigned short *) src;
+    while (*p != 0x0000) {
+	p++;
+    }
+    return (char *) p - src;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * InitializeEncodingSearchPath	--
+ *
+ *	This is the fallback routine that sets the default value of the
+ *	encoding search path if the application has not set one via a call to
+ *	Tcl_SetEncodingSearchPath() by the first time the search path is needed
+ *	to load encoding data.
+ *
+ *	The default encoding search path is produced by taking each directory
+ *	in the library path, appending a subdirectory named "encoding", and if
+ *	the resulting directory exists, adding it to the encoding search path.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Sets the encoding search path to an initial value.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+static void
+InitializeEncodingSearchPath(
+    char **valuePtr,
+    size_t *lengthPtr,
+    Tcl_Encoding *encodingPtr)
+{
+    const char *bytes;
+    int i, numDirs;
+    Tcl_Obj *libPathObj, *encodingObj, *searchPathObj;
+
+    TclNewLiteralStringObj(encodingObj, "encoding");
+    TclNewObj(searchPathObj);
+    Tcl_IncrRefCount(encodingObj);
+    Tcl_IncrRefCount(searchPathObj);
+    libPathObj = TclGetLibraryPath();
+    Tcl_IncrRefCount(libPathObj);
+    Tcl_ListObjLength(NULL, libPathObj, &numDirs);
+
+    for (i = 0; i < numDirs; i++) {
+	Tcl_Obj *directoryObj, *pathObj;
+	Tcl_StatBuf stat;
+
+	Tcl_ListObjIndex(NULL, libPathObj, i, &directoryObj);
+	pathObj = Tcl_FSJoinToPath(directoryObj, 1, &encodingObj);
+	Tcl_IncrRefCount(pathObj);
+	if ((0 == Tcl_FSStat(pathObj, &stat)) && S_ISDIR(stat.st_mode)) {
+	    Tcl_ListObjAppendElement(NULL, searchPathObj, pathObj);
+	}
+	Tcl_DecrRefCount(pathObj);
+    }
+
+    Tcl_DecrRefCount(libPathObj);
+    Tcl_DecrRefCount(encodingObj);
+    *encodingPtr = libraryPath.encoding;
+    if (*encodingPtr) {
+	((Encoding *)(*encodingPtr))->refCount++;
+    }
+    bytes = TclGetString(searchPathObj);
+
+    *lengthPtr = searchPathObj->length;
+    *valuePtr = ckalloc(*lengthPtr + 1);
+    memcpy(*valuePtr, bytes, *lengthPtr + 1);
+    Tcl_DecrRefCount(searchPathObj);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclEnsemble.c b/generic/tclEnsemble.c
new file mode 100644
index 0000000..f3e8187
--- /dev/null
+++ b/generic/tclEnsemble.c
@@ -0,0 +1,3668 @@
+/*
+ * tclEnsemble.c --
+ *
+ *	Contains support for ensembles (see TIP#112), which provide simple
+ *	mechanism for creating composite commands on top of namespaces.
+ *
+ * Copyright (c) 2005-2013 Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+
+/*
+ * Declarations for functions local to this file:
+ */
+
+static inline Tcl_Obj *	NewNsObj(Tcl_Namespace *namespacePtr);
+static inline int	EnsembleUnknownCallback(Tcl_Interp *interp,
+			    EnsembleConfig *ensemblePtr, int objc,
+			    Tcl_Obj *const objv[], Tcl_Obj **prefixObjPtr);
+static int		NsEnsembleImplementationCmd(ClientData clientData,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NsEnsembleImplementationCmdNR(ClientData clientData,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static void		BuildEnsembleConfig(EnsembleConfig *ensemblePtr);
+static int		NsEnsembleStringOrder(const void *strPtr1,
+			    const void *strPtr2);
+static void		DeleteEnsembleConfig(ClientData clientData);
+static void		MakeCachedEnsembleCommand(Tcl_Obj *objPtr,
+			    EnsembleConfig *ensemblePtr, Tcl_HashEntry *hPtr,
+			    Tcl_Obj *fix);
+static void		FreeEnsembleCmdRep(Tcl_Obj *objPtr);
+static void		DupEnsembleCmdRep(Tcl_Obj *objPtr, Tcl_Obj *copyPtr);
+static void		CompileToInvokedCommand(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Tcl_Obj *replacements,
+			    Command *cmdPtr, CompileEnv *envPtr);
+static int		CompileBasicNArgCommand(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    CompileEnv *envPtr);
+
+static Tcl_NRPostProc	FreeER;
+
+/*
+ * The lists of subcommands and options for the [namespace ensemble] command.
+ */
+
+static const char *const ensembleSubcommands[] = {
+    "configure", "create", "exists", NULL
+};
+enum EnsSubcmds {
+    ENS_CONFIG, ENS_CREATE, ENS_EXISTS
+};
+
+static const char *const ensembleCreateOptions[] = {
+    "-command", "-map", "-parameters", "-prefixes", "-subcommands",
+    "-unknown", NULL
+};
+enum EnsCreateOpts {
+    CRT_CMD, CRT_MAP, CRT_PARAM, CRT_PREFIX, CRT_SUBCMDS, CRT_UNKNOWN
+};
+
+static const char *const ensembleConfigOptions[] = {
+    "-map", "-namespace", "-parameters", "-prefixes", "-subcommands",
+    "-unknown", NULL
+};
+enum EnsConfigOpts {
+    CONF_MAP, CONF_NAMESPACE, CONF_PARAM, CONF_PREFIX, CONF_SUBCMDS,
+    CONF_UNKNOWN
+};
+
+/*
+ * This structure defines a Tcl object type that contains a reference to an
+ * ensemble subcommand (e.g. the "length" in [string length ab]). It is used
+ * to cache the mapping between the subcommand itself and the real command
+ * that implements it.
+ */
+
+static const Tcl_ObjType ensembleCmdType = {
+    "ensembleCommand",		/* the type's name */
+    FreeEnsembleCmdRep,		/* freeIntRepProc */
+    DupEnsembleCmdRep,		/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    NULL			/* setFromAnyProc */
+};
+
+/*
+ * The internal rep for caching ensemble subcommand lookups and
+ * spell corrections.
+ */
+
+typedef struct {
+    size_t epoch;               /* Used to confirm when the data in this
+                                 * really structure matches up with the
+                                 * ensemble. */
+    Command *token;             /* Reference to the command for which this
+                                 * structure is a cache of the resolution. */
+    Tcl_Obj *fix;               /* Corrected spelling, if needed. */
+    Tcl_HashEntry *hPtr;        /* Direct link to entry in the subcommand
+                                 * hash table. */
+} EnsembleCmdRep;
+
+
+static inline Tcl_Obj *
+NewNsObj(
+    Tcl_Namespace *namespacePtr)
+{
+    register Namespace *nsPtr = (Namespace *) namespacePtr;
+
+    if (namespacePtr == TclGetGlobalNamespace(nsPtr->interp)) {
+	return Tcl_NewStringObj("::", 2);
+    } else {
+	return Tcl_NewStringObj(nsPtr->fullName, -1);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNamespaceEnsembleCmd --
+ *
+ *	Invoked to implement the "namespace ensemble" command that creates and
+ *	manipulates ensembles built on top of namespaces. Handles the
+ *	following syntax:
+ *
+ *	    namespace ensemble name ?dictionary?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Creates the ensemble for the namespace if one did not previously
+ *	exist. Alternatively, alters the way that the ensemble's subcommand =>
+ *	implementation prefix is configured.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNamespaceEnsembleCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Namespace *namespacePtr;
+    Namespace *nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    Tcl_Command token;
+    Tcl_DictSearch search;
+    Tcl_Obj *listObj;
+    int index, done;
+
+    if (nsPtr == NULL || nsPtr->flags & NS_DYING) {
+	if (!Tcl_InterpDeleted(interp)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "tried to manipulate ensemble of deleted namespace",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "DEAD", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "subcommand ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], ensembleSubcommands,
+	    "subcommand", 0, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum EnsSubcmds) index) {
+    case ENS_CREATE: {
+	const char *name;
+	int len, allocatedMapFlag = 0;
+	/*
+	 * Defaults
+	 */
+	Tcl_Obj *subcmdObj = NULL;
+	Tcl_Obj *mapObj = NULL;
+	int permitPrefix = 1;
+	Tcl_Obj *unknownObj = NULL;
+	Tcl_Obj *paramObj = NULL;
+
+	/*
+	 * Check that we've got option-value pairs... [Bug 1558654]
+	 */
+
+	if (objc & 1) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?option value ...?");
+	    return TCL_ERROR;
+	}
+	objv += 2;
+	objc -= 2;
+
+	/*
+	 * Work out what name to use for the command to create. If supplied,
+	 * it is either fully specified or relative to the current namespace.
+	 * If not supplied, it is exactly the name of the current namespace.
+	 */
+
+	name = nsPtr->fullName;
+
+	/*
+	 * Parse the option list, applying type checks as we go. Note that we
+	 * are not incrementing any reference counts in the objects at this
+	 * stage, so the presence of an option multiple times won't cause any
+	 * memory leaks.
+	 */
+
+	for (; objc>1 ; objc-=2,objv+=2) {
+	    if (Tcl_GetIndexFromObj(interp, objv[0], ensembleCreateOptions,
+		    "option", 0, &index) != TCL_OK) {
+		if (allocatedMapFlag) {
+		    Tcl_DecrRefCount(mapObj);
+		}
+		return TCL_ERROR;
+	    }
+	    switch ((enum EnsCreateOpts) index) {
+	    case CRT_CMD:
+		name = TclGetString(objv[1]);
+		continue;
+	    case CRT_SUBCMDS:
+		if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		subcmdObj = (len > 0 ? objv[1] : NULL);
+		continue;
+	    case CRT_PARAM:
+		if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		paramObj = (len > 0 ? objv[1] : NULL);
+		continue;
+	    case CRT_MAP: {
+		Tcl_Obj *patchedDict = NULL, *subcmdWordsObj;
+
+		/*
+		 * Verify that the map is sensible.
+		 */
+
+		if (Tcl_DictObjFirst(interp, objv[1], &search,
+			&subcmdWordsObj, &listObj, &done) != TCL_OK) {
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		if (done) {
+		    mapObj = NULL;
+		    continue;
+		}
+		do {
+		    Tcl_Obj **listv;
+		    const char *cmd;
+
+		    if (TclListObjGetElements(interp, listObj, &len,
+			    &listv) != TCL_OK) {
+			Tcl_DictObjDone(&search);
+			if (patchedDict) {
+			    Tcl_DecrRefCount(patchedDict);
+			}
+			if (allocatedMapFlag) {
+			    Tcl_DecrRefCount(mapObj);
+			}
+			return TCL_ERROR;
+		    }
+		    if (len < 1) {
+			Tcl_SetObjResult(interp, Tcl_NewStringObj(
+				"ensemble subcommand implementations "
+				"must be non-empty lists", -1));
+			Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE",
+				"EMPTY_TARGET", NULL);
+			Tcl_DictObjDone(&search);
+			if (patchedDict) {
+			    Tcl_DecrRefCount(patchedDict);
+			}
+			if (allocatedMapFlag) {
+			    Tcl_DecrRefCount(mapObj);
+			}
+			return TCL_ERROR;
+		    }
+		    cmd = TclGetString(listv[0]);
+		    if (!(cmd[0] == ':' && cmd[1] == ':')) {
+			Tcl_Obj *newList = Tcl_NewListObj(len, listv);
+			Tcl_Obj *newCmd = NewNsObj((Tcl_Namespace *) nsPtr);
+
+			if (nsPtr->parentPtr) {
+			    Tcl_AppendStringsToObj(newCmd, "::", NULL);
+			}
+			Tcl_AppendObjToObj(newCmd, listv[0]);
+			Tcl_ListObjReplace(NULL, newList, 0, 1, 1, &newCmd);
+			if (patchedDict == NULL) {
+			    patchedDict = Tcl_DuplicateObj(objv[1]);
+			}
+			Tcl_DictObjPut(NULL, patchedDict, subcmdWordsObj,
+				newList);
+		    }
+		    Tcl_DictObjNext(&search, &subcmdWordsObj,&listObj, &done);
+		} while (!done);
+
+		if (allocatedMapFlag) {
+		    Tcl_DecrRefCount(mapObj);
+		}
+		mapObj = (patchedDict ? patchedDict : objv[1]);
+		if (patchedDict) {
+		    allocatedMapFlag = 1;
+		}
+		continue;
+	    }
+	    case CRT_PREFIX:
+		if (Tcl_GetBooleanFromObj(interp, objv[1],
+			&permitPrefix) != TCL_OK) {
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		continue;
+	    case CRT_UNKNOWN:
+		if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		unknownObj = (len > 0 ? objv[1] : NULL);
+		continue;
+	    }
+	}
+
+	/*
+	 * Create the ensemble. Note that this might delete another ensemble
+	 * linked to the same namespace, so we must be careful. However, we
+	 * should be OK because we only link the namespace into the list once
+	 * we've created it (and after any deletions have occurred.)
+	 */
+
+	token = Tcl_CreateEnsemble(interp, name, NULL,
+		(permitPrefix ? TCL_ENSEMBLE_PREFIX : 0));
+	Tcl_SetEnsembleSubcommandList(interp, token, subcmdObj);
+	Tcl_SetEnsembleMappingDict(interp, token, mapObj);
+	Tcl_SetEnsembleUnknownHandler(interp, token, unknownObj);
+	Tcl_SetEnsembleParameterList(interp, token, paramObj);
+
+	/*
+	 * Tricky! Must ensure that the result is not shared (command delete
+	 * traces could have corrupted the pristine object that we started
+	 * with). [Snit test rename-1.5]
+	 */
+
+	Tcl_ResetResult(interp);
+	Tcl_GetCommandFullName(interp, token, Tcl_GetObjResult(interp));
+	return TCL_OK;
+    }
+
+    case ENS_EXISTS:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "cmdname");
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(
+		Tcl_FindEnsemble(interp, objv[2], 0) != NULL));
+	return TCL_OK;
+
+    case ENS_CONFIG:
+	if (objc < 3 || (objc != 4 && !(objc & 1))) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "cmdname ?-option value ...? ?arg ...?");
+	    return TCL_ERROR;
+	}
+	token = Tcl_FindEnsemble(interp, objv[2], TCL_LEAVE_ERR_MSG);
+	if (token == NULL) {
+	    return TCL_ERROR;
+	}
+
+	if (objc == 4) {
+	    Tcl_Obj *resultObj = NULL;		/* silence gcc 4 warning */
+
+	    if (Tcl_GetIndexFromObj(interp, objv[3], ensembleConfigOptions,
+		    "option", 0, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum EnsConfigOpts) index) {
+	    case CONF_SUBCMDS:
+		Tcl_GetEnsembleSubcommandList(NULL, token, &resultObj);
+		if (resultObj != NULL) {
+		    Tcl_SetObjResult(interp, resultObj);
+		}
+		break;
+	    case CONF_PARAM:
+		Tcl_GetEnsembleParameterList(NULL, token, &resultObj);
+		if (resultObj != NULL) {
+		    Tcl_SetObjResult(interp, resultObj);
+		}
+		break;
+	    case CONF_MAP:
+		Tcl_GetEnsembleMappingDict(NULL, token, &resultObj);
+		if (resultObj != NULL) {
+		    Tcl_SetObjResult(interp, resultObj);
+		}
+		break;
+	    case CONF_NAMESPACE:
+		namespacePtr = NULL;		/* silence gcc 4 warning */
+		Tcl_GetEnsembleNamespace(NULL, token, &namespacePtr);
+		Tcl_SetObjResult(interp, NewNsObj(namespacePtr));
+		break;
+	    case CONF_PREFIX: {
+		int flags = 0;			/* silence gcc 4 warning */
+
+		Tcl_GetEnsembleFlags(NULL, token, &flags);
+		Tcl_SetObjResult(interp,
+			Tcl_NewBooleanObj(flags & TCL_ENSEMBLE_PREFIX));
+		break;
+	    }
+	    case CONF_UNKNOWN:
+		Tcl_GetEnsembleUnknownHandler(NULL, token, &resultObj);
+		if (resultObj != NULL) {
+		    Tcl_SetObjResult(interp, resultObj);
+		}
+		break;
+	    }
+	} else if (objc == 3) {
+	    /*
+	     * Produce list of all information.
+	     */
+
+	    Tcl_Obj *resultObj, *tmpObj = NULL;	/* silence gcc 4 warning */
+	    int flags = 0;			/* silence gcc 4 warning */
+
+	    TclNewObj(resultObj);
+
+	    /* -map option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_MAP], -1));
+	    Tcl_GetEnsembleMappingDict(NULL, token, &tmpObj);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    (tmpObj != NULL) ? tmpObj : Tcl_NewObj());
+
+	    /* -namespace option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_NAMESPACE],
+			    -1));
+	    namespacePtr = NULL;		/* silence gcc 4 warning */
+	    Tcl_GetEnsembleNamespace(NULL, token, &namespacePtr);
+	    Tcl_ListObjAppendElement(NULL, resultObj, NewNsObj(namespacePtr));
+
+	    /* -parameters option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_PARAM], -1));
+	    Tcl_GetEnsembleParameterList(NULL, token, &tmpObj);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    (tmpObj != NULL) ? tmpObj : Tcl_NewObj());
+
+	    /* -prefix option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_PREFIX], -1));
+	    Tcl_GetEnsembleFlags(NULL, token, &flags);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewBooleanObj(flags & TCL_ENSEMBLE_PREFIX));
+
+	    /* -subcommands option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_SUBCMDS],-1));
+	    Tcl_GetEnsembleSubcommandList(NULL, token, &tmpObj);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    (tmpObj != NULL) ? tmpObj : Tcl_NewObj());
+
+	    /* -unknown option */
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(ensembleConfigOptions[CONF_UNKNOWN],-1));
+	    Tcl_GetEnsembleUnknownHandler(NULL, token, &tmpObj);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    (tmpObj != NULL) ? tmpObj : Tcl_NewObj());
+
+	    Tcl_SetObjResult(interp, resultObj);
+	} else {
+	    int len, allocatedMapFlag = 0;
+	    Tcl_Obj *subcmdObj = NULL, *mapObj = NULL, *paramObj = NULL,
+		    *unknownObj = NULL; /* Defaults, silence gcc 4 warnings */
+	    int permitPrefix, flags = 0;	/* silence gcc 4 warning */
+
+	    Tcl_GetEnsembleSubcommandList(NULL, token, &subcmdObj);
+	    Tcl_GetEnsembleMappingDict(NULL, token, &mapObj);
+	    Tcl_GetEnsembleParameterList(NULL, token, &paramObj);
+	    Tcl_GetEnsembleUnknownHandler(NULL, token, &unknownObj);
+	    Tcl_GetEnsembleFlags(NULL, token, &flags);
+	    permitPrefix = (flags & TCL_ENSEMBLE_PREFIX) != 0;
+
+	    objv += 3;
+	    objc -= 3;
+
+	    /*
+	     * Parse the option list, applying type checks as we go. Note that
+	     * we are not incrementing any reference counts in the objects at
+	     * this stage, so the presence of an option multiple times won't
+	     * cause any memory leaks.
+	     */
+
+	    for (; objc>0 ; objc-=2,objv+=2) {
+		if (Tcl_GetIndexFromObj(interp, objv[0],ensembleConfigOptions,
+			"option", 0, &index) != TCL_OK) {
+		freeMapAndError:
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    return TCL_ERROR;
+		}
+		switch ((enum EnsConfigOpts) index) {
+		case CONF_SUBCMDS:
+		    if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+			goto freeMapAndError;
+		    }
+		    subcmdObj = (len > 0 ? objv[1] : NULL);
+		    continue;
+		case CONF_PARAM:
+		    if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+			goto freeMapAndError;
+		    }
+		    paramObj = (len > 0 ? objv[1] : NULL);
+		    continue;
+		case CONF_MAP: {
+		    Tcl_Obj *patchedDict = NULL, *subcmdWordsObj, **listv;
+		    const char *cmd;
+
+		    /*
+		     * Verify that the map is sensible.
+		     */
+
+		    if (Tcl_DictObjFirst(interp, objv[1], &search,
+			    &subcmdWordsObj, &listObj, &done) != TCL_OK) {
+			goto freeMapAndError;
+		    }
+		    if (done) {
+			mapObj = NULL;
+			continue;
+		    }
+		    do {
+			if (TclListObjGetElements(interp, listObj, &len,
+				&listv) != TCL_OK) {
+			    Tcl_DictObjDone(&search);
+			    if (patchedDict) {
+				Tcl_DecrRefCount(patchedDict);
+			    }
+			    goto freeMapAndError;
+			}
+			if (len < 1) {
+			    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+				    "ensemble subcommand implementations "
+				    "must be non-empty lists", -1));
+			    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE",
+				    "EMPTY_TARGET", NULL);
+			    Tcl_DictObjDone(&search);
+			    if (patchedDict) {
+				Tcl_DecrRefCount(patchedDict);
+			    }
+			    goto freeMapAndError;
+			}
+			cmd = TclGetString(listv[0]);
+			if (!(cmd[0] == ':' && cmd[1] == ':')) {
+			    Tcl_Obj *newList = Tcl_DuplicateObj(listObj);
+			    Tcl_Obj *newCmd = NewNsObj((Tcl_Namespace*)nsPtr);
+
+			    if (nsPtr->parentPtr) {
+				Tcl_AppendStringsToObj(newCmd, "::", NULL);
+			    }
+			    Tcl_AppendObjToObj(newCmd, listv[0]);
+			    Tcl_ListObjReplace(NULL, newList, 0,1, 1,&newCmd);
+			    if (patchedDict == NULL) {
+				patchedDict = Tcl_DuplicateObj(objv[1]);
+			    }
+			    Tcl_DictObjPut(NULL, patchedDict, subcmdWordsObj,
+				    newList);
+			}
+			Tcl_DictObjNext(&search, &subcmdWordsObj, &listObj,
+				&done);
+		    } while (!done);
+		    if (allocatedMapFlag) {
+			Tcl_DecrRefCount(mapObj);
+		    }
+		    mapObj = (patchedDict ? patchedDict : objv[1]);
+		    if (patchedDict) {
+			allocatedMapFlag = 1;
+		    }
+		    continue;
+		}
+		case CONF_NAMESPACE:
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "option -namespace is read-only", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "READ_ONLY",
+			    NULL);
+		    goto freeMapAndError;
+		case CONF_PREFIX:
+		    if (Tcl_GetBooleanFromObj(interp, objv[1],
+			    &permitPrefix) != TCL_OK) {
+			goto freeMapAndError;
+		    }
+		    continue;
+		case CONF_UNKNOWN:
+		    if (TclListObjLength(interp, objv[1], &len) != TCL_OK) {
+			goto freeMapAndError;
+		    }
+		    unknownObj = (len > 0 ? objv[1] : NULL);
+		    continue;
+		}
+	    }
+
+	    /*
+	     * Update the namespace now that we've finished the parsing stage.
+	     */
+
+	    flags = (permitPrefix ? flags|TCL_ENSEMBLE_PREFIX
+		    : flags&~TCL_ENSEMBLE_PREFIX);
+	    Tcl_SetEnsembleSubcommandList(interp, token, subcmdObj);
+	    Tcl_SetEnsembleMappingDict(interp, token, mapObj);
+	    Tcl_SetEnsembleParameterList(interp, token, paramObj);
+	    Tcl_SetEnsembleUnknownHandler(interp, token, unknownObj);
+	    Tcl_SetEnsembleFlags(interp, token, flags);
+	}
+	return TCL_OK;
+
+    default:
+	Tcl_Panic("unexpected ensemble command");
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateEnsemble --
+ *
+ *	Create a simple ensemble attached to the given namespace.
+ *
+ * Results:
+ *	The token for the command created.
+ *
+ * Side effects:
+ *	The ensemble is created and marked for compilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_CreateEnsemble(
+    Tcl_Interp *interp,
+    const char *name,
+    Tcl_Namespace *namespacePtr,
+    int flags)
+{
+    Namespace *nsPtr = (Namespace *) namespacePtr;
+    EnsembleConfig *ensemblePtr = ckalloc(sizeof(EnsembleConfig));
+    Tcl_Obj *nameObj = NULL;
+
+    if (nsPtr == NULL) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    }
+
+    /*
+     * Make the name of the ensemble into a fully qualified name. This might
+     * allocate a temporary object.
+     */
+
+    if (!(name[0] == ':' && name[1] == ':')) {
+	nameObj = NewNsObj((Tcl_Namespace *) nsPtr);
+	if (nsPtr->parentPtr == NULL) {
+	    Tcl_AppendStringsToObj(nameObj, name, NULL);
+	} else {
+	    Tcl_AppendStringsToObj(nameObj, "::", name, NULL);
+	}
+	Tcl_IncrRefCount(nameObj);
+	name = TclGetString(nameObj);
+    }
+
+    ensemblePtr->nsPtr = nsPtr;
+    ensemblePtr->epoch = 0;
+    Tcl_InitHashTable(&ensemblePtr->subcommandTable, TCL_STRING_KEYS);
+    ensemblePtr->subcommandArrayPtr = NULL;
+    ensemblePtr->subcmdList = NULL;
+    ensemblePtr->subcommandDict = NULL;
+    ensemblePtr->flags = flags;
+    ensemblePtr->numParameters = 0;
+    ensemblePtr->parameterList = NULL;
+    ensemblePtr->unknownHandler = NULL;
+    ensemblePtr->token = Tcl_NRCreateCommand(interp, name,
+	    NsEnsembleImplementationCmd, NsEnsembleImplementationCmdNR,
+	    ensemblePtr, DeleteEnsembleConfig);
+    ensemblePtr->next = (EnsembleConfig *) nsPtr->ensembles;
+    nsPtr->ensembles = (Tcl_Ensemble *) ensemblePtr;
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    nsPtr->exportLookupEpoch++;
+
+    if (flags & ENSEMBLE_COMPILE) {
+	((Command *) ensemblePtr->token)->compileProc = TclCompileEnsemble;
+    }
+
+    if (nameObj != NULL) {
+	TclDecrRefCount(nameObj);
+    }
+    return ensemblePtr->token;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEnsembleSubcommandList --
+ *
+ *	Set the subcommand list for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an ensemble
+ *	or the subcommand list - if non-NULL - is not a list).
+ *
+ * Side effects:
+ *	The ensemble is updated and marked for recompilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEnsembleSubcommandList(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj *subcmdList)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+    Tcl_Obj *oldList;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"command is not an ensemble", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	return TCL_ERROR;
+    }
+    if (subcmdList != NULL) {
+	int length;
+
+	if (TclListObjLength(interp, subcmdList, &length) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (length < 1) {
+	    subcmdList = NULL;
+	}
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    oldList = ensemblePtr->subcmdList;
+    ensemblePtr->subcmdList = subcmdList;
+    if (subcmdList != NULL) {
+	Tcl_IncrRefCount(subcmdList);
+    }
+    if (oldList != NULL) {
+	TclDecrRefCount(oldList);
+    }
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    ensemblePtr->nsPtr->exportLookupEpoch++;
+
+    /*
+     * Special hack to make compiling of [info exists] work when the
+     * dictionary is modified.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	((Interp *) interp)->compileEpoch++;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEnsembleParameterList --
+ *
+ *	Set the parameter list for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an ensemble
+ *	or the parameter list - if non-NULL - is not a list).
+ *
+ * Side effects:
+ *	The ensemble is updated and marked for recompilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEnsembleParameterList(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj *paramList)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+    Tcl_Obj *oldList;
+    int length;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"command is not an ensemble", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	return TCL_ERROR;
+    }
+    if (paramList == NULL) {
+	length = 0;
+    } else {
+	if (TclListObjLength(interp, paramList, &length) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (length < 1) {
+	    paramList = NULL;
+	}
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    oldList = ensemblePtr->parameterList;
+    ensemblePtr->parameterList = paramList;
+    if (paramList != NULL) {
+	Tcl_IncrRefCount(paramList);
+    }
+    if (oldList != NULL) {
+	TclDecrRefCount(oldList);
+    }
+    ensemblePtr->numParameters = length;
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    ensemblePtr->nsPtr->exportLookupEpoch++;
+
+    /*
+     * Special hack to make compiling of [info exists] work when the
+     * dictionary is modified.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	((Interp *) interp)->compileEpoch++;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEnsembleMappingDict --
+ *
+ *	Set the mapping dictionary for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an ensemble
+ *	or the mapping - if non-NULL - is not a dict).
+ *
+ * Side effects:
+ *	The ensemble is updated and marked for recompilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEnsembleMappingDict(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj *mapDict)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+    Tcl_Obj *oldDict;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"command is not an ensemble", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	return TCL_ERROR;
+    }
+    if (mapDict != NULL) {
+	int size, done;
+	Tcl_DictSearch search;
+	Tcl_Obj *valuePtr;
+
+	if (Tcl_DictObjSize(interp, mapDict, &size) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	for (Tcl_DictObjFirst(NULL, mapDict, &search, NULL, &valuePtr, &done);
+		!done; Tcl_DictObjNext(&search, NULL, &valuePtr, &done)) {
+	    Tcl_Obj *cmdObjPtr;
+	    const char *bytes;
+
+	    if (Tcl_ListObjIndex(interp, valuePtr, 0, &cmdObjPtr) != TCL_OK) {
+		Tcl_DictObjDone(&search);
+		return TCL_ERROR;
+	    }
+	    bytes = TclGetString(cmdObjPtr);
+	    if (bytes[0] != ':' || bytes[1] != ':') {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"ensemble target is not a fully-qualified command",
+			-1));
+		Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE",
+			"UNQUALIFIED_TARGET", NULL);
+		Tcl_DictObjDone(&search);
+		return TCL_ERROR;
+	    }
+	}
+
+	if (size < 1) {
+	    mapDict = NULL;
+	}
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    oldDict = ensemblePtr->subcommandDict;
+    ensemblePtr->subcommandDict = mapDict;
+    if (mapDict != NULL) {
+	Tcl_IncrRefCount(mapDict);
+    }
+    if (oldDict != NULL) {
+	TclDecrRefCount(oldDict);
+    }
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    ensemblePtr->nsPtr->exportLookupEpoch++;
+
+    /*
+     * Special hack to make compiling of [info exists] work when the
+     * dictionary is modified.
+     */
+
+    if (cmdPtr->compileProc != NULL) {
+	((Interp *) interp)->compileEpoch++;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEnsembleUnknownHandler --
+ *
+ *	Set the unknown handler for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an ensemble
+ *	or the unknown handler - if non-NULL - is not a list).
+ *
+ * Side effects:
+ *	The ensemble is updated and marked for recompilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEnsembleUnknownHandler(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj *unknownList)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+    Tcl_Obj *oldList;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"command is not an ensemble", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	return TCL_ERROR;
+    }
+    if (unknownList != NULL) {
+	int length;
+
+	if (TclListObjLength(interp, unknownList, &length) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (length < 1) {
+	    unknownList = NULL;
+	}
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    oldList = ensemblePtr->unknownHandler;
+    ensemblePtr->unknownHandler = unknownList;
+    if (unknownList != NULL) {
+	Tcl_IncrRefCount(unknownList);
+    }
+    if (oldList != NULL) {
+	TclDecrRefCount(oldList);
+    }
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    ensemblePtr->nsPtr->exportLookupEpoch++;
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetEnsembleFlags --
+ *
+ *	Set the flags for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble).
+ *
+ * Side effects:
+ *	The ensemble is updated and marked for recompilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetEnsembleFlags(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    int flags)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+    int wasCompiled;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"command is not an ensemble", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    wasCompiled = ensemblePtr->flags & ENSEMBLE_COMPILE;
+
+    /*
+     * This API refuses to set the ENSEMBLE_DEAD flag...
+     */
+
+    ensemblePtr->flags &= ENSEMBLE_DEAD;
+    ensemblePtr->flags |= flags & ~ENSEMBLE_DEAD;
+
+    /*
+     * Trigger an eventual recomputation of the ensemble command set. Note
+     * that this is slightly tricky, as it means that we are not actually
+     * counting the number of namespace export actions, but it is the simplest
+     * way to go!
+     */
+
+    ensemblePtr->nsPtr->exportLookupEpoch++;
+
+    /*
+     * If the ENSEMBLE_COMPILE flag status was changed, install or remove the
+     * compiler function and bump the interpreter's compilation epoch so that
+     * bytecode gets regenerated.
+     */
+
+    if (flags & ENSEMBLE_COMPILE) {
+	if (!wasCompiled) {
+	    ((Command*) ensemblePtr->token)->compileProc = TclCompileEnsemble;
+	    ((Interp *) interp)->compileEpoch++;
+	}
+    } else {
+	if (wasCompiled) {
+	    ((Command *) ensemblePtr->token)->compileProc = NULL;
+	    ((Interp *) interp)->compileEpoch++;
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleSubcommandList --
+ *
+ *	Get the list of subcommands associated with a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). The list of subcommands is returned by updating the
+ *	variable pointed to by the last parameter (NULL if this is to be
+ *	derived from the mapping dictionary or the associated namespace's
+ *	exported commands).
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleSubcommandList(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj **subcmdListPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *subcmdListPtr = ensemblePtr->subcmdList;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleParameterList --
+ *
+ *	Get the list of parameters associated with a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). The list of parameters is returned by updating the
+ *	variable pointed to by the last parameter (NULL if there are
+ *	no parameters).
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleParameterList(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj **paramListPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *paramListPtr = ensemblePtr->parameterList;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleMappingDict --
+ *
+ *	Get the command mapping dictionary associated with a particular
+ *	ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). The mapping dict is returned by updating the variable
+ *	pointed to by the last parameter (NULL if none is installed).
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleMappingDict(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj **mapDictPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *mapDictPtr = ensemblePtr->subcommandDict;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleUnknownHandler --
+ *
+ *	Get the unknown handler associated with a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). The unknown handler is returned by updating the variable
+ *	pointed to by the last parameter (NULL if no handler is installed).
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleUnknownHandler(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Obj **unknownListPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *unknownListPtr = ensemblePtr->unknownHandler;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleFlags --
+ *
+ *	Get the flags for a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). The flags are returned by updating the variable pointed to
+ *	by the last parameter.
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleFlags(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    int *flagsPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *flagsPtr = ensemblePtr->flags;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetEnsembleNamespace --
+ *
+ *	Get the namespace associated with a particular ensemble.
+ *
+ * Results:
+ *	Tcl result code (error if command token does not indicate an
+ *	ensemble). Namespace is returned by updating the variable pointed to
+ *	by the last parameter.
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetEnsembleNamespace(
+    Tcl_Interp *interp,
+    Tcl_Command token,
+    Tcl_Namespace **namespacePtrPtr)
+{
+    Command *cmdPtr = (Command *) token;
+    EnsembleConfig *ensemblePtr;
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command is not an ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "NOT_ENSEMBLE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    ensemblePtr = cmdPtr->objClientData;
+    *namespacePtrPtr = (Tcl_Namespace *) ensemblePtr->nsPtr;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindEnsemble --
+ *
+ *	Given a command name, get the ensemble token for it, allowing for
+ *	[namespace import]s. [Bug 1017022]
+ *
+ * Results:
+ *	The token for the ensemble command with the given name, or NULL if the
+ *	command either does not exist or is not an ensemble (when an error
+ *	message will be written into the interp if thats non-NULL).
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_FindEnsemble(
+    Tcl_Interp *interp,		/* Where to do the lookup, and where to write
+				 * the errors if TCL_LEAVE_ERR_MSG is set in
+				 * the flags. */
+    Tcl_Obj *cmdNameObj,	/* Name of command to look up. */
+    int flags)			/* Either 0 or TCL_LEAVE_ERR_MSG; other flags
+				 * are probably not useful. */
+{
+    Command *cmdPtr;
+
+    cmdPtr = (Command *)
+	    Tcl_FindCommand(interp, TclGetString(cmdNameObj), NULL, flags);
+    if (cmdPtr == NULL) {
+	return NULL;
+    }
+
+    if (cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	/*
+	 * Reuse existing infrastructure for following import link chains
+	 * rather than duplicating it.
+	 */
+
+	cmdPtr = (Command *) TclGetOriginalCommand((Tcl_Command) cmdPtr);
+
+	if (cmdPtr == NULL || cmdPtr->objProc != NsEnsembleImplementationCmd){
+	    if (flags & TCL_LEAVE_ERR_MSG) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"\"%s\" is not an ensemble command",
+			TclGetString(cmdNameObj)));
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ENSEMBLE",
+			TclGetString(cmdNameObj), NULL);
+	    }
+	    return NULL;
+	}
+    }
+
+    return (Tcl_Command) cmdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsEnsemble --
+ *
+ *	Simple test for ensemble-hood that takes into account imported
+ *	ensemble commands as well.
+ *
+ * Results:
+ *	Boolean value
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsEnsemble(
+    Tcl_Command token)
+{
+    Command *cmdPtr = (Command *) token;
+
+    if (cmdPtr->objProc == NsEnsembleImplementationCmd) {
+	return 1;
+    }
+    cmdPtr = (Command *) TclGetOriginalCommand((Tcl_Command) cmdPtr);
+    if (cmdPtr == NULL || cmdPtr->objProc != NsEnsembleImplementationCmd) {
+	return 0;
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclMakeEnsemble --
+ *
+ *	Create an ensemble from a table of implementation commands. The
+ *	ensemble will be subject to (limited) compilation if any of the
+ *	implementation commands are compilable.
+ *
+ *	The 'name' parameter may be a single command name or a list if
+ *	creating an ensemble subcommand (see the binary implementation).
+ *
+ *	Currently, the TCL_ENSEMBLE_PREFIX ensemble flag is only used on
+ *	top-level ensemble commands.
+ *
+ * Results:
+ *	Handle for the new ensemble, or NULL on failure.
+ *
+ * Side effects:
+ *	May advance the bytecode compilation epoch.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclMakeEnsemble(
+    Tcl_Interp *interp,
+    const char *name,		 /* The ensemble name (as explained above) */
+    const EnsembleImplMap map[]) /* The subcommands to create */
+{
+    Tcl_Command ensemble;
+    Tcl_Namespace *ns;
+    Tcl_DString buf, hiddenBuf;
+    const char **nameParts = NULL;
+    const char *cmdName = NULL;
+    int i, nameCount = 0, ensembleFlags = 0, hiddenLen;
+
+    /*
+     * Construct the path for the ensemble namespace and create it.
+     */
+
+    Tcl_DStringInit(&buf);
+    Tcl_DStringInit(&hiddenBuf);
+    TclDStringAppendLiteral(&hiddenBuf, "tcl:");
+    Tcl_DStringAppend(&hiddenBuf, name, -1);
+    TclDStringAppendLiteral(&hiddenBuf, ":");
+    hiddenLen = Tcl_DStringLength(&hiddenBuf);
+    if (name[0] == ':' && name[1] == ':') {
+	/*
+	 * An absolute name, so use it directly.
+	 */
+
+	cmdName = name;
+	Tcl_DStringAppend(&buf, name, -1);
+	ensembleFlags = TCL_ENSEMBLE_PREFIX;
+    } else {
+	/*
+	 * Not an absolute name, so do munging of it. Note that this treats a
+	 * multi-word list differently to a single word.
+	 */
+
+	TclDStringAppendLiteral(&buf, "::tcl");
+
+	if (Tcl_SplitList(NULL, name, &nameCount, &nameParts) != TCL_OK) {
+	    Tcl_Panic("invalid ensemble name '%s'", name);
+	}
+
+	for (i = 0; i < nameCount; ++i) {
+	    TclDStringAppendLiteral(&buf, "::");
+	    Tcl_DStringAppend(&buf, nameParts[i], -1);
+	}
+    }
+
+    ns = Tcl_FindNamespace(interp, Tcl_DStringValue(&buf), NULL,
+	    TCL_CREATE_NS_IF_UNKNOWN);
+    if (!ns) {
+	Tcl_Panic("unable to find or create %s namespace!",
+		Tcl_DStringValue(&buf));
+    }
+
+    /*
+     * Create the named ensemble in the correct namespace
+     */
+
+    if (cmdName == NULL) {
+	if (nameCount == 1) {
+	    ensembleFlags = TCL_ENSEMBLE_PREFIX;
+	    cmdName = Tcl_DStringValue(&buf) + 5;
+	} else {
+	    ns = ns->parentPtr;
+	    cmdName = nameParts[nameCount - 1];
+	}
+    }
+
+    /*
+     * Switch on compilation always for core ensembles now that we can do
+     * nice bytecode things with them.  Do it now.  Waiting until later will
+     * just cause pointless epoch bumps.
+     */
+
+    ensembleFlags |= ENSEMBLE_COMPILE;
+    ensemble = Tcl_CreateEnsemble(interp, cmdName, ns, ensembleFlags);
+
+    /*
+     * Create the ensemble mapping dictionary and the ensemble command procs.
+     */
+
+    if (ensemble != NULL) {
+	Tcl_Obj *mapDict, *fromObj, *toObj;
+	Command *cmdPtr;
+
+	TclDStringAppendLiteral(&buf, "::");
+	TclNewObj(mapDict);
+	for (i=0 ; map[i].name != NULL ; i++) {
+	    fromObj = Tcl_NewStringObj(map[i].name, -1);
+	    TclNewStringObj(toObj, Tcl_DStringValue(&buf),
+		    Tcl_DStringLength(&buf));
+	    Tcl_AppendToObj(toObj, map[i].name, -1);
+	    Tcl_DictObjPut(NULL, mapDict, fromObj, toObj);
+
+	    if (map[i].proc || map[i].nreProc) {
+		/*
+		 * If the command is unsafe, hide it when we're in a safe
+		 * interpreter. The code to do this is really hokey! It also
+		 * doesn't work properly yet; this function is always
+		 * currently called before the safe-interp flag is set so the
+		 * Tcl_IsSafe check fails.
+		 */
+
+		if (map[i].unsafe && Tcl_IsSafe(interp)) {
+		    cmdPtr = (Command *)
+			    Tcl_NRCreateCommand(interp, "___tmp", map[i].proc,
+			    map[i].nreProc, map[i].clientData, NULL);
+		    Tcl_DStringSetLength(&hiddenBuf, hiddenLen);
+		    if (Tcl_HideCommand(interp, "___tmp",
+			    Tcl_DStringAppend(&hiddenBuf, map[i].name, -1))) {
+			Tcl_Panic("%s", Tcl_GetString(Tcl_GetObjResult(interp)));
+		    }
+		} else {
+		    /*
+		     * Not hidden, so just create it. Yay!
+		     */
+
+		    cmdPtr = (Command *)
+			    Tcl_NRCreateCommand(interp, TclGetString(toObj),
+			    map[i].proc, map[i].nreProc, map[i].clientData,
+			    NULL);
+		}
+		cmdPtr->compileProc = map[i].compileProc;
+	    }
+	}
+	Tcl_SetEnsembleMappingDict(interp, ensemble, mapDict);
+    }
+
+    Tcl_DStringFree(&buf);
+    Tcl_DStringFree(&hiddenBuf);
+    if (nameParts != NULL) {
+	ckfree(nameParts);
+    }
+    return ensemble;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NsEnsembleImplementationCmd --
+ *
+ *	Implements an ensemble of commands (being those exported by a
+ *	namespace other than the global namespace) as a command with the same
+ *	(short) name as the namespace in the parent namespace.
+ *
+ * Results:
+ *	A standard Tcl result code. Will be TCL_ERROR if the command is not an
+ *	unambiguous prefix of any command exported by the ensemble's
+ *	namespace.
+ *
+ * Side effects:
+ *	Depends on the command within the namespace that gets executed. If the
+ *	ensemble itself returns TCL_ERROR, a descriptive error message will be
+ *	placed in the interpreter's result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NsEnsembleImplementationCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    return Tcl_NRCallObjProc(interp, NsEnsembleImplementationCmdNR,
+	    clientData, objc, objv);
+}
+
+static int
+NsEnsembleImplementationCmdNR(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    EnsembleConfig *ensemblePtr = clientData;
+				/* The ensemble itself. */
+    Tcl_Obj *prefixObj;		/* An object containing the prefix words of
+				 * the command that implements the
+				 * subcommand. */
+    Tcl_HashEntry *hPtr;	/* Used for efficient lookup of fully
+				 * specified but not yet cached command
+				 * names. */
+    int reparseCount = 0;	/* Number of reparses. */
+    Tcl_Obj *errorObj;		/* Used for building error messages. */
+    Tcl_Obj *subObj;
+    int subIdx;
+
+    /*
+     * Must recheck objc, since numParameters might have changed. Cf. test
+     * namespace-53.9.
+     */
+
+  restartEnsembleParse:
+    subIdx = 1 + ensemblePtr->numParameters;
+    if (objc < subIdx + 1) {
+	/*
+	 * We don't have a subcommand argument. Make error message.
+	 */
+
+	Tcl_DString buf;	/* Message being built */
+
+	Tcl_DStringInit(&buf);
+	if (ensemblePtr->parameterList) {
+	    Tcl_DStringAppend(&buf,
+		    TclGetString(ensemblePtr->parameterList), -1);
+	    TclDStringAppendLiteral(&buf, " ");
+	}
+	TclDStringAppendLiteral(&buf, "subcommand ?arg ...?");
+	Tcl_WrongNumArgs(interp, 1, objv, Tcl_DStringValue(&buf));
+	Tcl_DStringFree(&buf);
+
+	return TCL_ERROR;
+    }
+
+    if (ensemblePtr->nsPtr->flags & NS_DYING) {
+	/*
+	 * Don't know how we got here, but make things give up quickly.
+	 */
+
+	if (!Tcl_InterpDeleted(interp)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "ensemble activated for deleted namespace", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "DEAD", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Determine if the table of subcommands is right. If so, we can just look
+     * up in there and go straight to dispatch.
+     */
+
+    subObj = objv[subIdx];
+
+    if (ensemblePtr->epoch == ensemblePtr->nsPtr->exportLookupEpoch) {
+	/*
+	 * Table of subcommands is still valid; therefore there might be a
+	 * valid cache of discovered information which we can reuse. Do the
+	 * check here, and if we're still valid, we can jump straight to the
+	 * part where we do the invocation of the subcommand.
+	 */
+
+	if (subObj->typePtr==&ensembleCmdType){
+	    EnsembleCmdRep *ensembleCmd = subObj->internalRep.twoPtrValue.ptr1;
+
+	    if (ensembleCmd->epoch == ensemblePtr->epoch &&
+		    ensembleCmd->token == (Command *)ensemblePtr->token) {
+		prefixObj = Tcl_GetHashValue(ensembleCmd->hPtr);
+		Tcl_IncrRefCount(prefixObj);
+		if (ensembleCmd->fix) {
+		    TclSpellFix(interp, objv, objc, subIdx, subObj, ensembleCmd->fix);
+		}
+		goto runResultingSubcommand;
+	    }
+	}
+    } else {
+	BuildEnsembleConfig(ensemblePtr);
+	ensemblePtr->epoch = ensemblePtr->nsPtr->exportLookupEpoch;
+    }
+
+    /*
+     * Look in the hashtable for the subcommand name; this is the fastest way
+     * of all if there is no cache in operation.
+     */
+
+    hPtr = Tcl_FindHashEntry(&ensemblePtr->subcommandTable,
+	    TclGetString(subObj));
+    if (hPtr != NULL) {
+
+	/*
+	 * Cache for later in the subcommand object.
+	 */
+
+	MakeCachedEnsembleCommand(subObj, ensemblePtr, hPtr, NULL);
+    } else if (!(ensemblePtr->flags & TCL_ENSEMBLE_PREFIX)) {
+	/*
+	 * Could not map, no prefixing, go to unknown/error handling.
+	 */
+
+	goto unknownOrAmbiguousSubcommand;
+    } else {
+	/*
+	 * If we've not already confirmed the command with the hash as part of
+	 * building our export table, we need to scan the sorted array for
+	 * matches.
+	 */
+
+	const char *subcmdName; /* Name of the subcommand, or unique prefix of
+				 * it (will be an error for a non-unique
+				 * prefix). */
+	char *fullName = NULL;	/* Full name of the subcommand. */
+	int stringLength, i;
+	int tableLength = ensemblePtr->subcommandTable.numEntries;
+	Tcl_Obj *fix;
+
+	subcmdName = TclGetStringFromObj(subObj, &stringLength);
+	for (i=0 ; i<tableLength ; i++) {
+	    register int cmp = strncmp(subcmdName,
+		    ensemblePtr->subcommandArrayPtr[i],
+		    (unsigned) stringLength);
+
+	    if (cmp == 0) {
+		if (fullName != NULL) {
+		    /*
+		     * Since there's never the exact-match case to worry about
+		     * (hash search filters this), getting here indicates that
+		     * our subcommand is an ambiguous prefix of (at least) two
+		     * exported subcommands, which is an error case.
+		     */
+
+		    goto unknownOrAmbiguousSubcommand;
+		}
+		fullName = ensemblePtr->subcommandArrayPtr[i];
+	    } else if (cmp < 0) {
+		/*
+		 * Because we are searching a sorted table, we can now stop
+		 * searching because we have gone past anything that could
+		 * possibly match.
+		 */
+
+		break;
+	    }
+	}
+	if (fullName == NULL) {
+	    /*
+	     * The subcommand is not a prefix of anything, so bail out!
+	     */
+
+	    goto unknownOrAmbiguousSubcommand;
+	}
+	hPtr = Tcl_FindHashEntry(&ensemblePtr->subcommandTable, fullName);
+	if (hPtr == NULL) {
+	    Tcl_Panic("full name %s not found in supposedly synchronized hash",
+		    fullName);
+	}
+
+	/*
+	 * Record the spelling correction for usage message.
+	 */
+
+	fix = Tcl_NewStringObj(fullName, -1);
+
+	/*
+	 * Cache for later in the subcommand object.
+	 */
+
+	MakeCachedEnsembleCommand(subObj, ensemblePtr, hPtr, fix);
+	TclSpellFix(interp, objv, objc, subIdx, subObj, fix);
+    }
+
+    prefixObj = Tcl_GetHashValue(hPtr);
+    Tcl_IncrRefCount(prefixObj);
+  runResultingSubcommand:
+
+    /*
+     * Do the real work of execution of the subcommand by building an array of
+     * objects (note that this is potentially not the same length as the
+     * number of arguments to this ensemble command), populating it and then
+     * feeding it back through the main command-lookup engine. In theory, we
+     * could look up the command in the namespace ourselves, as we already
+     * have the namespace in which it is guaranteed to exist,
+     *
+     *   ((Q: That's not true if the -map option is used, is it?))
+     *
+     * but we don't do that (the cacheing of the command object used should
+     * help with that.)
+     */
+
+    {
+	Tcl_Obj *copyPtr;	/* The actual list of words to dispatch to.
+				 * Will be freed by the dispatch engine. */
+	Tcl_Obj **copyObjv;
+	int copyObjc, prefixObjc;
+
+	Tcl_ListObjLength(NULL, prefixObj, &prefixObjc);
+
+	if (objc == 2) {
+	    copyPtr = TclListObjCopy(NULL, prefixObj);
+	} else {
+	    copyPtr = Tcl_NewListObj(objc - 2 + prefixObjc, NULL);
+	    Tcl_ListObjAppendList(NULL, copyPtr, prefixObj);
+	    Tcl_ListObjReplace(NULL, copyPtr, LIST_MAX, 0,
+		    ensemblePtr->numParameters, objv + 1);
+	    Tcl_ListObjReplace(NULL, copyPtr, LIST_MAX, 0,
+		    objc - 2 - ensemblePtr->numParameters,
+		    objv + 2 + ensemblePtr->numParameters);
+	}
+	Tcl_IncrRefCount(copyPtr);
+	TclNRAddCallback(interp, TclNRReleaseValues, copyPtr, NULL, NULL, NULL);
+	TclDecrRefCount(prefixObj);
+
+	/*
+	 * Record what arguments the script sent in so that things like
+	 * Tcl_WrongNumArgs can give the correct error message. Parameters
+	 * count both as inserted and removed arguments.
+	 */
+
+	if (TclInitRewriteEnsemble(interp, 2 + ensemblePtr->numParameters,
+		prefixObjc + ensemblePtr->numParameters, objv)) {
+	    TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL,
+		    NULL);
+	}
+
+	/*
+	 * Hand off to the target command.
+	 */
+
+	TclSkipTailcall(interp);
+	Tcl_ListObjGetElements(NULL, copyPtr, &copyObjc, &copyObjv);
+	return TclNREvalObjv(interp, copyObjc, copyObjv, TCL_EVAL_INVOKE, NULL);
+    }
+
+  unknownOrAmbiguousSubcommand:
+    /*
+     * Have not been able to match the subcommand asked for with a real
+     * subcommand that we export. See whether a handler has been registered
+     * for dealing with this situation. Will only call (at most) once for any
+     * particular ensemble invocation.
+     */
+
+    if (ensemblePtr->unknownHandler != NULL && reparseCount++ < 1) {
+	switch (EnsembleUnknownCallback(interp, ensemblePtr, objc, objv,
+		&prefixObj)) {
+	case TCL_OK:
+	    goto runResultingSubcommand;
+	case TCL_ERROR:
+	    return TCL_ERROR;
+	case TCL_CONTINUE:
+	    goto restartEnsembleParse;
+	}
+    }
+
+    /*
+     * We cannot determine what subcommand to hand off to, so generate a
+     * (standard) failure message. Note the one odd case compared with
+     * standard ensemble-like command, which is where a namespace has no
+     * exported commands at all...
+     */
+
+    Tcl_ResetResult(interp);
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "SUBCOMMAND",
+	    TclGetString(subObj), NULL);
+    if (ensemblePtr->subcommandTable.numEntries == 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown subcommand \"%s\": namespace %s does not"
+		" export any commands", TclGetString(subObj),
+		ensemblePtr->nsPtr->fullName));
+	return TCL_ERROR;
+    }
+    errorObj = Tcl_ObjPrintf("unknown%s subcommand \"%s\": must be ",
+	    (ensemblePtr->flags & TCL_ENSEMBLE_PREFIX ? " or ambiguous" : ""),
+	    TclGetString(subObj));
+    if (ensemblePtr->subcommandTable.numEntries == 1) {
+	Tcl_AppendToObj(errorObj, ensemblePtr->subcommandArrayPtr[0], -1);
+    } else {
+	int i;
+
+	for (i=0 ; i<ensemblePtr->subcommandTable.numEntries-1 ; i++) {
+	    Tcl_AppendToObj(errorObj, ensemblePtr->subcommandArrayPtr[i], -1);
+	    Tcl_AppendToObj(errorObj, ", ", 2);
+	}
+	Tcl_AppendPrintfToObj(errorObj, "or %s",
+		ensemblePtr->subcommandArrayPtr[i]);
+    }
+    Tcl_SetObjResult(interp, errorObj);
+    return TCL_ERROR;
+}
+
+int
+TclClearRootEnsemble(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    TclResetRewriteEnsemble(interp, 1);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitRewriteEnsemble --
+ *
+ *	Applies a rewrite of arguments so that an ensemble subcommand will
+ *	report error messages correctly for the overall command.
+ *
+ * Results:
+ *	Whether this is the first rewrite applied, a value which must be
+ *	passed to TclResetRewriteEnsemble when undoing this command's
+ *	behaviour.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInitRewriteEnsemble(
+    Tcl_Interp *interp,
+    int numRemoved,
+    int numInserted,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    int isRootEnsemble = (iPtr->ensembleRewrite.sourceObjs == NULL);
+
+    if (isRootEnsemble) {
+	iPtr->ensembleRewrite.sourceObjs = objv;
+	iPtr->ensembleRewrite.numRemovedObjs = numRemoved;
+	iPtr->ensembleRewrite.numInsertedObjs = numInserted;
+    } else {
+	int numIns = iPtr->ensembleRewrite.numInsertedObjs;
+
+	if (numIns < numRemoved) {
+	    iPtr->ensembleRewrite.numRemovedObjs += numRemoved - numIns;
+	    iPtr->ensembleRewrite.numInsertedObjs = numInserted;
+	} else {
+	    iPtr->ensembleRewrite.numInsertedObjs += numInserted - numRemoved;
+	}
+    }
+    return isRootEnsemble;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclResetRewriteEnsemble --
+ *
+ *	Removes any rewrites applied to support proper reporting of error
+ *	messages used in ensembles. Should be paired with
+ *	TclInitRewriteEnsemble.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclResetRewriteEnsemble(
+    Tcl_Interp *interp,
+    int isRootEnsemble)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (isRootEnsemble) {
+	iPtr->ensembleRewrite.sourceObjs = NULL;
+	iPtr->ensembleRewrite.numRemovedObjs = 0;
+	iPtr->ensembleRewrite.numInsertedObjs = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSpellFix --
+ *
+ *	Record a spelling correction that needs making in the
+ *	generation of the WrongNumArgs usage message.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Can create an alternative ensemble rewrite structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FreeER(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj **tmp = (Tcl_Obj **)data[0];
+
+    ckfree(tmp[2]);
+    ckfree(tmp);
+    return result;
+}
+
+void
+TclSpellFix(
+    Tcl_Interp *interp,
+    Tcl_Obj *const *objv,
+    int objc,
+    int badIdx,
+    Tcl_Obj *bad,
+    Tcl_Obj *fix)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *const *search;
+    Tcl_Obj **store;
+    int idx;
+    int size;
+
+    if (iPtr->ensembleRewrite.sourceObjs == NULL) {
+	iPtr->ensembleRewrite.sourceObjs = objv;
+	iPtr->ensembleRewrite.numRemovedObjs = 0;
+	iPtr->ensembleRewrite.numInsertedObjs = 0;
+    }
+
+    /* Compute the valid length of the ensemble root */
+
+    size = iPtr->ensembleRewrite.numRemovedObjs + objc
+		- iPtr->ensembleRewrite.numInsertedObjs;
+
+    search = iPtr->ensembleRewrite.sourceObjs;
+    if (search[0] == NULL) {
+	/* Awful casting abuse here */
+	search = (Tcl_Obj *const *) search[1];
+    }
+
+    if (badIdx < iPtr->ensembleRewrite.numInsertedObjs) {
+	/*
+	 * Misspelled value was inserted. We cannot directly jump
+	 * to the bad value, but have to search.
+	 */
+	idx = 1;
+	while (idx < size) {
+	    if (search[idx] == bad) {
+		break;
+	    }
+	    idx++;
+	}
+	if (idx == size) {
+	    return;
+	}
+    } else {
+	/* Jump to the misspelled value. */
+	idx = iPtr->ensembleRewrite.numRemovedObjs + badIdx
+		- iPtr->ensembleRewrite.numInsertedObjs;
+
+	/* Verify */
+	if (search[idx] != bad) {
+	    Tcl_Panic("SpellFix: programming error");
+	}
+    }
+
+    search = iPtr->ensembleRewrite.sourceObjs;
+    if (search[0] == NULL) {
+	store = (Tcl_Obj **)search[2];
+    }  else {
+	Tcl_Obj **tmp = ckalloc(3 * sizeof(Tcl_Obj *));
+	tmp[0] = NULL;
+	tmp[1] = (Tcl_Obj *)iPtr->ensembleRewrite.sourceObjs;
+	tmp[2] = (Tcl_Obj *)ckalloc(size * sizeof(Tcl_Obj *));
+	memcpy(tmp[2], tmp[1], size*sizeof(Tcl_Obj *));
+
+	iPtr->ensembleRewrite.sourceObjs = (Tcl_Obj *const *) tmp;
+	TclNRAddCallback(interp, FreeER, tmp, NULL, NULL, NULL);
+	store = (Tcl_Obj **)tmp[2];
+    }
+
+    store[idx] = fix;
+    Tcl_IncrRefCount(fix);
+    TclNRAddCallback(interp, TclNRReleaseValues, fix, NULL, NULL, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFetchEnsembleRoot --
+ *
+ *	Returns the root of ensemble rewriting, if any.
+ *	If no root exists, returns objv instead.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *const *
+TclFetchEnsembleRoot(
+    Tcl_Interp *interp,
+    Tcl_Obj *const *objv,
+    int objc,
+    int *objcPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (iPtr->ensembleRewrite.sourceObjs) {
+	*objcPtr = objc + iPtr->ensembleRewrite.numRemovedObjs
+		- iPtr->ensembleRewrite.numInsertedObjs;
+	return iPtr->ensembleRewrite.sourceObjs;
+    }
+    *objcPtr = objc;
+    return objv;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * EnsmebleUnknownCallback --
+ *
+ *	Helper for the ensemble engine that handles the procesing of unknown
+ *	callbacks. See the user documentation of the ensemble unknown handler
+ *	for details; this function is only ever called when such a function is
+ *	defined, and is only ever called once per ensemble dispatch (i.e. if a
+ *	reparse still fails, this isn't called again).
+ *
+ * Results:
+ *	TCL_OK -	*prefixObjPtr contains the command words to dispatch
+ *			to.
+ *	TCL_CONTINUE -	Need to reparse (*prefixObjPtr is invalid).
+ *	TCL_ERROR -	Something went wrong! Error message in interpreter.
+ *
+ * Side effects:
+ *	Calls the Tcl interpreter, so arbitrary.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline int
+EnsembleUnknownCallback(
+    Tcl_Interp *interp,
+    EnsembleConfig *ensemblePtr,
+    int objc,
+    Tcl_Obj *const objv[],
+    Tcl_Obj **prefixObjPtr)
+{
+    int paramc, i, result, prefixObjc;
+    Tcl_Obj **paramv, *unknownCmd, *ensObj;
+
+    /*
+     * Create the unknown command callback to determine what to do.
+     */
+
+    unknownCmd = Tcl_DuplicateObj(ensemblePtr->unknownHandler);
+    TclNewObj(ensObj);
+    Tcl_GetCommandFullName(interp, ensemblePtr->token, ensObj);
+    Tcl_ListObjAppendElement(NULL, unknownCmd, ensObj);
+    for (i=1 ; i<objc ; i++) {
+	Tcl_ListObjAppendElement(NULL, unknownCmd, objv[i]);
+    }
+    TclListObjGetElements(NULL, unknownCmd, &paramc, &paramv);
+    Tcl_IncrRefCount(unknownCmd);
+
+    /*
+     * Now call the unknown handler. (We don't bother NRE-enabling this; deep
+     * recursing through unknown handlers is horribly perverse.) Note that it
+     * is always an error for an unknown handler to delete its ensemble; don't
+     * do that!
+     */
+
+    Tcl_Preserve(ensemblePtr);
+    TclSkipTailcall(interp);
+    result = Tcl_EvalObjv(interp, paramc, paramv, 0);
+    if ((result == TCL_OK) && (ensemblePtr->flags & ENSEMBLE_DEAD)) {
+	if (!Tcl_InterpDeleted(interp)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unknown subcommand handler deleted its ensemble", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "UNKNOWN_DELETED",
+		    NULL);
+	}
+	result = TCL_ERROR;
+    }
+    Tcl_Release(ensemblePtr);
+
+    /*
+     * If we succeeded, we should either have a list of words that form the
+     * command to be executed, or an empty list. In the empty-list case, the
+     * ensemble is believed to be updated so we should ask the ensemble engine
+     * to reparse the original command.
+     */
+
+    if (result == TCL_OK) {
+	*prefixObjPtr = Tcl_GetObjResult(interp);
+	Tcl_IncrRefCount(*prefixObjPtr);
+	TclDecrRefCount(unknownCmd);
+	Tcl_ResetResult(interp);
+
+	/*
+	 * Namespace is still there. Check if the result is a valid list. If
+	 * it is, and it is non-empty, that list is what we are using as our
+	 * replacement.
+	 */
+
+	if (TclListObjLength(interp, *prefixObjPtr, &prefixObjc) != TCL_OK) {
+	    TclDecrRefCount(*prefixObjPtr);
+	    Tcl_AddErrorInfo(interp, "\n    while parsing result of "
+		    "ensemble unknown subcommand handler");
+	    return TCL_ERROR;
+	}
+	if (prefixObjc > 0) {
+	    return TCL_OK;
+	}
+
+	/*
+	 * Namespace alive & empty result => reparse.
+	 */
+
+	TclDecrRefCount(*prefixObjPtr);
+	return TCL_CONTINUE;
+    }
+
+    /*
+     * Oh no! An exceptional result. Convert to an error.
+     */
+
+    if (!Tcl_InterpDeleted(interp)) {
+	if (result != TCL_ERROR) {
+	    Tcl_ResetResult(interp);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unknown subcommand handler returned bad code: ", -1));
+	    switch (result) {
+	    case TCL_RETURN:
+		Tcl_AppendToObj(Tcl_GetObjResult(interp), "return", -1);
+		break;
+	    case TCL_BREAK:
+		Tcl_AppendToObj(Tcl_GetObjResult(interp), "break", -1);
+		break;
+	    case TCL_CONTINUE:
+		Tcl_AppendToObj(Tcl_GetObjResult(interp), "continue", -1);
+		break;
+	    default:
+		Tcl_AppendPrintfToObj(Tcl_GetObjResult(interp), "%d", result);
+	    }
+	    Tcl_AddErrorInfo(interp, "\n    result of "
+		    "ensemble unknown subcommand handler: ");
+	    Tcl_AppendObjToErrorInfo(interp, unknownCmd);
+	    Tcl_SetErrorCode(interp, "TCL", "ENSEMBLE", "UNKNOWN_RESULT",
+		    NULL);
+	} else {
+	    Tcl_AddErrorInfo(interp,
+		    "\n    (ensemble unknown subcommand handler)");
+	}
+    }
+    TclDecrRefCount(unknownCmd);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeCachedEnsembleCommand --
+ *
+ *	Cache what we've computed so far; it's not nice to repeatedly copy
+ *	strings about. Note that to do this, we start by deleting any old
+ *	representation that there was (though if it was an out of date
+ *	ensemble rep, we can skip some of the deallocation process.)
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Alters the internal representation of the first object parameter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MakeCachedEnsembleCommand(
+    Tcl_Obj *objPtr,
+    EnsembleConfig *ensemblePtr,
+    Tcl_HashEntry *hPtr,
+    Tcl_Obj *fix)
+{
+    register EnsembleCmdRep *ensembleCmd;
+
+    if (objPtr->typePtr == &ensembleCmdType) {
+	ensembleCmd = objPtr->internalRep.twoPtrValue.ptr1;
+	TclCleanupCommandMacro(ensembleCmd->token);
+	if (ensembleCmd->fix) {
+	    Tcl_DecrRefCount(ensembleCmd->fix);
+	}
+    } else {
+	/*
+	 * Kill the old internal rep, and replace it with a brand new one of
+	 * our own.
+	 */
+
+	TclFreeIntRep(objPtr);
+	ensembleCmd = ckalloc(sizeof(EnsembleCmdRep));
+	objPtr->internalRep.twoPtrValue.ptr1 = ensembleCmd;
+	objPtr->typePtr = &ensembleCmdType;
+    }
+
+    /*
+     * Populate the internal rep.
+     */
+
+    ensembleCmd->epoch = ensemblePtr->epoch;
+    ensembleCmd->token = (Command *) ensemblePtr->token;
+    ensembleCmd->token->refCount++;
+    if (fix) {
+	Tcl_IncrRefCount(fix);
+    }
+    ensembleCmd->fix = fix;
+    ensembleCmd->hPtr = hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteEnsembleConfig --
+ *
+ *	Destroys the data structure used to represent an ensemble. This is
+ *	called when the ensemble's command is deleted (which happens
+ *	automatically if the ensemble's namespace is deleted.) Maintainers
+ *	should note that ensembles should be deleted by deleting their
+ *	commands.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is (eventually) deallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteEnsembleConfig(
+    ClientData clientData)
+{
+    EnsembleConfig *ensemblePtr = clientData;
+    Namespace *nsPtr = ensemblePtr->nsPtr;
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hEnt;
+
+    /*
+     * Unlink from the ensemble chain if it has not been marked as having been
+     * done already.
+     */
+
+    if (ensemblePtr->next != ensemblePtr) {
+	EnsembleConfig *ensPtr = (EnsembleConfig *) nsPtr->ensembles;
+
+	if (ensPtr == ensemblePtr) {
+	    nsPtr->ensembles = (Tcl_Ensemble *) ensemblePtr->next;
+	} else {
+	    while (ensPtr != NULL) {
+		if (ensPtr->next == ensemblePtr) {
+		    ensPtr->next = ensemblePtr->next;
+		    break;
+		}
+		ensPtr = ensPtr->next;
+	    }
+	}
+    }
+
+    /*
+     * Mark the namespace as dead so code that uses Tcl_Preserve() can tell
+     * whether disaster happened anyway.
+     */
+
+    ensemblePtr->flags |= ENSEMBLE_DEAD;
+
+    /*
+     * Kill the pointer-containing fields.
+     */
+
+    if (ensemblePtr->subcommandTable.numEntries != 0) {
+	ckfree(ensemblePtr->subcommandArrayPtr);
+    }
+    hEnt = Tcl_FirstHashEntry(&ensemblePtr->subcommandTable, &search);
+    while (hEnt != NULL) {
+	Tcl_Obj *prefixObj = Tcl_GetHashValue(hEnt);
+
+	Tcl_DecrRefCount(prefixObj);
+	hEnt = Tcl_NextHashEntry(&search);
+    }
+    Tcl_DeleteHashTable(&ensemblePtr->subcommandTable);
+    if (ensemblePtr->subcmdList != NULL) {
+	Tcl_DecrRefCount(ensemblePtr->subcmdList);
+    }
+    if (ensemblePtr->parameterList != NULL) {
+	Tcl_DecrRefCount(ensemblePtr->parameterList);
+    }
+    if (ensemblePtr->subcommandDict != NULL) {
+	Tcl_DecrRefCount(ensemblePtr->subcommandDict);
+    }
+    if (ensemblePtr->unknownHandler != NULL) {
+	Tcl_DecrRefCount(ensemblePtr->unknownHandler);
+    }
+
+    /*
+     * Arrange for the structure to be reclaimed. Note that this is complex
+     * because we have to make sure that we can react sensibly when an
+     * ensemble is deleted during the process of initialising the ensemble
+     * (especially the unknown callback.)
+     */
+
+    Tcl_EventuallyFree(ensemblePtr, TCL_DYNAMIC);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BuildEnsembleConfig --
+ *
+ *	Create the internal data structures that describe how an ensemble
+ *	looks, being a hash mapping from the full command name to the Tcl list
+ *	that describes the implementation prefix words, and a sorted array of
+ *	all the full command names to allow for reasonably efficient
+ *	unambiguous prefix handling.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Reallocates and rebuilds the hash table and array stored at the
+ *	ensemblePtr argument. For large ensembles or large namespaces, this is
+ *	a potentially expensive operation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+BuildEnsembleConfig(
+    EnsembleConfig *ensemblePtr)
+{
+    Tcl_HashSearch search;	/* Used for scanning the set of commands in
+				 * the namespace that backs up this
+				 * ensemble. */
+    int i, j, isNew;
+    Tcl_HashTable *hash = &ensemblePtr->subcommandTable;
+    Tcl_HashEntry *hPtr;
+
+    if (hash->numEntries != 0) {
+	/*
+	 * Remove pre-existing table.
+	 */
+
+	ckfree(ensemblePtr->subcommandArrayPtr);
+	hPtr = Tcl_FirstHashEntry(hash, &search);
+	while (hPtr != NULL) {
+	    Tcl_Obj *prefixObj = Tcl_GetHashValue(hPtr);
+
+	    Tcl_DecrRefCount(prefixObj);
+	    hPtr = Tcl_NextHashEntry(&search);
+	}
+	Tcl_DeleteHashTable(hash);
+	Tcl_InitHashTable(hash, TCL_STRING_KEYS);
+    }
+
+    /*
+     * See if we've got an export list. If so, we will only export exactly
+     * those commands, which may be either implemented by the prefix in the
+     * subcommandDict or mapped directly onto the namespace's commands.
+     */
+
+    if (ensemblePtr->subcmdList != NULL) {
+	Tcl_Obj **subcmdv, *target, *cmdObj, *cmdPrefixObj;
+	int subcmdc;
+
+	TclListObjGetElements(NULL, ensemblePtr->subcmdList, &subcmdc,
+		&subcmdv);
+	for (i=0 ; i<subcmdc ; i++) {
+	    const char *name = TclGetString(subcmdv[i]);
+
+	    hPtr = Tcl_CreateHashEntry(hash, name, &isNew);
+
+	    /*
+	     * Skip non-unique cases.
+	     */
+
+	    if (!isNew) {
+		continue;
+	    }
+
+	    /*
+	     * Look in our dictionary (if present) for the command.
+	     */
+
+	    if (ensemblePtr->subcommandDict != NULL) {
+		Tcl_DictObjGet(NULL, ensemblePtr->subcommandDict, subcmdv[i],
+			&target);
+		if (target != NULL) {
+		    Tcl_SetHashValue(hPtr, target);
+		    Tcl_IncrRefCount(target);
+		    continue;
+		}
+	    }
+
+	    /*
+	     * Not there, so map onto the namespace. Note in this case that we
+	     * do not guarantee that the command is actually there; that is
+	     * the programmer's responsibility (or [::unknown] of course).
+	     */
+
+	    cmdObj = NewNsObj((Tcl_Namespace *) ensemblePtr->nsPtr);
+	    if (ensemblePtr->nsPtr->parentPtr != NULL) {
+		Tcl_AppendStringsToObj(cmdObj, "::", name, NULL);
+	    } else {
+		Tcl_AppendStringsToObj(cmdObj, name, NULL);
+	    }
+	    cmdPrefixObj = Tcl_NewListObj(1, &cmdObj);
+	    Tcl_SetHashValue(hPtr, cmdPrefixObj);
+	    Tcl_IncrRefCount(cmdPrefixObj);
+	}
+    } else if (ensemblePtr->subcommandDict != NULL) {
+	/*
+	 * No subcmd list, but we do have a mapping dictionary so we should
+	 * use the keys of that. Convert the dictionary's contents into the
+	 * form required for the ensemble's internal hashtable.
+	 */
+
+	Tcl_DictSearch dictSearch;
+	Tcl_Obj *keyObj, *valueObj;
+	int done;
+
+	Tcl_DictObjFirst(NULL, ensemblePtr->subcommandDict, &dictSearch,
+		&keyObj, &valueObj, &done);
+	while (!done) {
+	    const char *name = TclGetString(keyObj);
+
+	    hPtr = Tcl_CreateHashEntry(hash, name, &isNew);
+	    Tcl_SetHashValue(hPtr, valueObj);
+	    Tcl_IncrRefCount(valueObj);
+	    Tcl_DictObjNext(&dictSearch, &keyObj, &valueObj, &done);
+	}
+    } else {
+	/*
+	 * Discover what commands are actually exported by the namespace.
+	 * What we have is an array of patterns and a hash table whose keys
+	 * are the command names exported by the namespace (the contents do
+	 * not matter here.) We must find out what commands are actually
+	 * exported by filtering each command in the namespace against each of
+	 * the patterns in the export list. Note that we use an intermediate
+	 * hash table to make memory management easier, and because that makes
+	 * exact matching far easier too.
+	 *
+	 * Suggestion for future enhancement: compute the unique prefixes and
+	 * place them in the hash too, which should make for even faster
+	 * matching.
+	 */
+
+	hPtr = Tcl_FirstHashEntry(&ensemblePtr->nsPtr->cmdTable, &search);
+	for (; hPtr!= NULL ; hPtr=Tcl_NextHashEntry(&search)) {
+	    char *nsCmdName =		/* Name of command in namespace. */
+		    Tcl_GetHashKey(&ensemblePtr->nsPtr->cmdTable, hPtr);
+
+	    for (i=0 ; i<ensemblePtr->nsPtr->numExportPatterns ; i++) {
+		if (Tcl_StringMatch(nsCmdName,
+			ensemblePtr->nsPtr->exportArrayPtr[i])) {
+		    hPtr = Tcl_CreateHashEntry(hash, nsCmdName, &isNew);
+
+		    /*
+		     * Remember, hash entries have a full reference to the
+		     * substituted part of the command (as a list) as their
+		     * content!
+		     */
+
+		    if (isNew) {
+			Tcl_Obj *cmdObj, *cmdPrefixObj;
+
+			TclNewObj(cmdObj);
+			Tcl_AppendStringsToObj(cmdObj,
+				ensemblePtr->nsPtr->fullName,
+				(ensemblePtr->nsPtr->parentPtr ? "::" : ""),
+				nsCmdName, NULL);
+			cmdPrefixObj = Tcl_NewListObj(1, &cmdObj);
+			Tcl_SetHashValue(hPtr, cmdPrefixObj);
+			Tcl_IncrRefCount(cmdPrefixObj);
+		    }
+		    break;
+		}
+	    }
+	}
+    }
+
+    if (hash->numEntries == 0) {
+	ensemblePtr->subcommandArrayPtr = NULL;
+	return;
+    }
+
+    /*
+     * Create a sorted array of all subcommands in the ensemble; hash tables
+     * are all very well for a quick look for an exact match, but they can't
+     * determine things like whether a string is a prefix of another (not
+     * without lots of preparation anyway) and they're no good for when we're
+     * generating the error message either.
+     *
+     * We do this by filling an array with the names (we use the hash keys
+     * directly to save a copy, since any time we change the array we change
+     * the hash too, and vice versa) and running quicksort over the array.
+     */
+
+    ensemblePtr->subcommandArrayPtr =
+	    ckalloc(sizeof(char *) * hash->numEntries);
+
+    /*
+     * Fill array from both ends as this makes us less likely to end up with
+     * performance problems in qsort(), which is good. Note that doing this
+     * makes this code much more opaque, but the naive alternatve:
+     *
+     * for (hPtr=Tcl_FirstHashEntry(hash,&search),i=0 ;
+     *	       hPtr!=NULL ; hPtr=Tcl_NextHashEntry(&search),i++) {
+     *     ensemblePtr->subcommandArrayPtr[i] = Tcl_GetHashKey(hash, &hPtr);
+     * }
+     *
+     * can produce long runs of precisely ordered table entries when the
+     * commands in the namespace are declared in a sorted fashion (an ordering
+     * some people like) and the hashing functions (or the command names
+     * themselves) are fairly unfortunate. By filling from both ends, it
+     * requires active malice (and probably a debugger) to get qsort() to have
+     * awful runtime behaviour.
+     */
+
+    i = 0;
+    j = hash->numEntries;
+    hPtr = Tcl_FirstHashEntry(hash, &search);
+    while (hPtr != NULL) {
+	ensemblePtr->subcommandArrayPtr[i++] = Tcl_GetHashKey(hash, hPtr);
+	hPtr = Tcl_NextHashEntry(&search);
+	if (hPtr == NULL) {
+	    break;
+	}
+	ensemblePtr->subcommandArrayPtr[--j] = Tcl_GetHashKey(hash, hPtr);
+	hPtr = Tcl_NextHashEntry(&search);
+    }
+    if (hash->numEntries > 1) {
+	qsort(ensemblePtr->subcommandArrayPtr, (unsigned) hash->numEntries,
+		sizeof(char *), NsEnsembleStringOrder);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NsEnsembleStringOrder --
+ *
+ *	Helper function to compare two pointers to two strings for use with
+ *	qsort().
+ *
+ * Results:
+ *	-1 if the first string is smaller, 1 if the second string is smaller,
+ *	and 0 if they are equal.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NsEnsembleStringOrder(
+    const void *strPtr1,
+    const void *strPtr2)
+{
+    return strcmp(*(const char **)strPtr1, *(const char **)strPtr2);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeEnsembleCmdRep --
+ *
+ *	Destroys the internal representation of a Tcl_Obj that has been
+ *	holding information about a command in an ensemble.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is deallocated. If this held the last reference to a
+ *	namespace's main structure, that main structure will also be
+ *	destroyed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeEnsembleCmdRep(
+    Tcl_Obj *objPtr)
+{
+    EnsembleCmdRep *ensembleCmd = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclCleanupCommandMacro(ensembleCmd->token);
+    if (ensembleCmd->fix) {
+	Tcl_DecrRefCount(ensembleCmd->fix);
+    }
+    ckfree(ensembleCmd);
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupEnsembleCmdRep --
+ *
+ *	Makes one Tcl_Obj into a copy of another that is a subcommand of an
+ *	ensemble.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is allocated, and the namespace that the ensemble is built on
+ *	top of gains another reference.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupEnsembleCmdRep(
+    Tcl_Obj *objPtr,
+    Tcl_Obj *copyPtr)
+{
+    EnsembleCmdRep *ensembleCmd = objPtr->internalRep.twoPtrValue.ptr1;
+    EnsembleCmdRep *ensembleCopy = ckalloc(sizeof(EnsembleCmdRep));
+
+    copyPtr->typePtr = &ensembleCmdType;
+    copyPtr->internalRep.twoPtrValue.ptr1 = ensembleCopy;
+    ensembleCopy->epoch = ensembleCmd->epoch;
+    ensembleCopy->token = ensembleCmd->token;
+    ensembleCopy->token->refCount++;
+    ensembleCopy->fix = ensembleCmd->fix;
+    if (ensembleCopy->fix) {
+	Tcl_IncrRefCount(ensembleCopy->fix);
+    }
+    ensembleCopy->hPtr = ensembleCmd->hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileEnsemble --
+ *
+ *	Procedure called to compile an ensemble command. Note that most
+ *	ensembles are not compiled, since modifying a compiled ensemble causes
+ *	a invalidation of all existing bytecode (expensive!) which is not
+ *	normally warranted.
+ *
+ * Results:
+ *	Returns TCL_OK for a successful compile. Returns TCL_ERROR to defer
+ *	evaluation to runtime.
+ *
+ * Side effects:
+ *	Instructions are added to envPtr to execute the subcommands of the
+ *	ensemble at runtime if a compile-time mapping is possible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompileEnsemble(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    Tcl_Obj *mapObj, *subcmdObj, *targetCmdObj, *listObj, **elems;
+    Tcl_Obj *replaced = Tcl_NewObj(), *replacement;
+    Tcl_Command ensemble = (Tcl_Command) cmdPtr;
+    Command *oldCmdPtr = cmdPtr, *newCmdPtr;
+    int len, result, flags = 0, i, depth = 1, invokeAnyway = 0;
+    int ourResult = TCL_ERROR;
+    unsigned numBytes;
+    const char *word;
+    DefineLineInformation;
+
+    Tcl_IncrRefCount(replaced);
+    if (parsePtr->numWords < depth + 1) {
+	goto failed;
+    }
+    if (tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	/*
+	 * Too hard.
+	 */
+
+	goto failed;
+    }
+
+    /*
+     * This is where we return to if we are parsing multiple nested compiled
+     * ensembles. [info object] is such a beast.
+     */
+
+  checkNextWord:
+    word = tokenPtr[1].start;
+    numBytes = tokenPtr[1].size;
+
+    /*
+     * There's a sporting chance we'll be able to compile this. But now we
+     * must check properly. To do that, check that we're compiling an ensemble
+     * that has a compilable command as its appropriate subcommand.
+     */
+
+    if (Tcl_GetEnsembleMappingDict(NULL, ensemble, &mapObj) != TCL_OK
+	    || mapObj == NULL) {
+	/*
+	 * Either not an ensemble or a mapping isn't installed. Crud. Too hard
+	 * to proceed.
+	 */
+
+	goto failed;
+    }
+
+    /*
+     * Also refuse to compile anything that uses a formal parameter list for
+     * now, on the grounds that it is too complex.
+     */
+
+    if (Tcl_GetEnsembleParameterList(NULL, ensemble, &listObj) != TCL_OK
+	    || listObj != NULL) {
+	/*
+	 * Figuring out how to compile this has become too much. Bail out.
+	 */
+
+	goto failed;
+    }
+
+    /*
+     * Next, get the flags. We need them on several code paths so that we can
+     * know whether we're to do prefix matching.
+     */
+
+    (void) Tcl_GetEnsembleFlags(NULL, ensemble, &flags);
+
+    /*
+     * Check to see if there's also a subcommand list; must check to see if
+     * the subcommand we are calling is in that list if it exists, since that
+     * list filters the entries in the map.
+     */
+
+    (void) Tcl_GetEnsembleSubcommandList(NULL, ensemble, &listObj);
+    if (listObj != NULL) {
+	int sclen;
+	const char *str;
+	Tcl_Obj *matchObj = NULL;
+
+	if (Tcl_ListObjGetElements(NULL, listObj, &len, &elems) != TCL_OK) {
+	    goto failed;
+	}
+	for (i=0 ; i<len ; i++) {
+	    str = TclGetStringFromObj(elems[i], &sclen);
+	    if ((sclen == (int) numBytes) && !memcmp(word, str, numBytes)) {
+		/*
+		 * Exact match! Excellent!
+		 */
+
+		result = Tcl_DictObjGet(NULL, mapObj,elems[i], &targetCmdObj);
+		if (result != TCL_OK || targetCmdObj == NULL) {
+		    goto failed;
+		}
+		replacement = elems[i];
+		goto doneMapLookup;
+	    }
+
+	    /*
+	     * Check to see if we've got a prefix match. A single prefix match
+	     * is fine, and allows us to refine our dictionary lookup, but
+	     * multiple prefix matches is a Bad Thing and will prevent us from
+	     * making progress. Note that we cannot do the lookup immediately
+	     * in the prefix case; might be another entry later in the list
+	     * that causes things to fail.
+	     */
+
+	    if ((flags & TCL_ENSEMBLE_PREFIX)
+		    && strncmp(word, str, numBytes) == 0) {
+		if (matchObj != NULL) {
+		    goto failed;
+		}
+		matchObj = elems[i];
+	    }
+	}
+	if (matchObj == NULL) {
+	    goto failed;
+	}
+	result = Tcl_DictObjGet(NULL, mapObj, matchObj, &targetCmdObj);
+	if (result != TCL_OK || targetCmdObj == NULL) {
+	    goto failed;
+	}
+	replacement = matchObj;
+    } else {
+	Tcl_DictSearch s;
+	int done, matched;
+	Tcl_Obj *tmpObj;
+
+	/*
+	 * No map, so check the dictionary directly.
+	 */
+
+	TclNewStringObj(subcmdObj, word, (int) numBytes);
+	result = Tcl_DictObjGet(NULL, mapObj, subcmdObj, &targetCmdObj);
+	if (result == TCL_OK && targetCmdObj != NULL) {
+	    /*
+	     * Got it. Skip the fiddling around with prefixes.
+	     */
+
+	    replacement = subcmdObj;
+	    goto doneMapLookup;
+	}
+	TclDecrRefCount(subcmdObj);
+
+	/*
+	 * We've not literally got a valid subcommand. But maybe we have a
+	 * prefix. Check if prefix matches are allowed.
+	 */
+
+	if (!(flags & TCL_ENSEMBLE_PREFIX)) {
+	    goto failed;
+	}
+
+	/*
+	 * Iterate over the keys in the dictionary, checking to see if we're a
+	 * prefix.
+	 */
+
+	Tcl_DictObjFirst(NULL, mapObj, &s, &subcmdObj, &tmpObj, &done);
+	matched = 0;
+	replacement = NULL;		/* Silence, fool compiler! */
+	while (!done) {
+	    if (strncmp(TclGetString(subcmdObj), word, numBytes) == 0) {
+		if (matched++) {
+		    /*
+		     * Must have matched twice! Not unique, so no point
+		     * looking further.
+		     */
+
+		    break;
+		}
+		replacement = subcmdObj;
+		targetCmdObj = tmpObj;
+	    }
+	    Tcl_DictObjNext(&s, &subcmdObj, &tmpObj, &done);
+	}
+	Tcl_DictObjDone(&s);
+
+	/*
+	 * If we have anything other than a single match, we've failed the
+	 * unique prefix check.
+	 */
+
+	if (matched != 1) {
+	    invokeAnyway = 1;
+	    goto failed;
+	}
+    }
+
+    /*
+     * OK, we definitely map to something. But what?
+     *
+     * The command we map to is the first word out of the map element. Note
+     * that we also reject dealing with multi-element rewrites if we are in a
+     * safe interpreter, as there is otherwise a (highly gnarly!) way to make
+     * Tcl crash open to exploit.
+     */
+
+  doneMapLookup:
+    Tcl_ListObjAppendElement(NULL, replaced, replacement);
+    if (Tcl_ListObjGetElements(NULL, targetCmdObj, &len, &elems) != TCL_OK) {
+	goto failed;
+    } else if (len != 1) {
+	/*
+	 * Note that at this point we know we can't issue any special
+	 * instruction sequence as the mapping isn't one that we support at
+	 * the compiled level.
+	 */
+
+	goto cleanup;
+    }
+    targetCmdObj = elems[0];
+
+    oldCmdPtr = cmdPtr;
+    Tcl_IncrRefCount(targetCmdObj);
+    newCmdPtr = (Command *) Tcl_GetCommandFromObj(interp, targetCmdObj);
+    TclDecrRefCount(targetCmdObj);
+    if (newCmdPtr == NULL || Tcl_IsSafe(interp)
+	    || newCmdPtr->nsPtr->flags & NS_SUPPRESS_COMPILATION
+	    || newCmdPtr->flags & CMD_HAS_EXEC_TRACES
+	    || ((Interp *)interp)->flags & DONT_COMPILE_CMDS_INLINE) {
+	/*
+	 * Maps to an undefined command or a command without a compiler.
+	 * Cannot compile.
+	 */
+
+	goto cleanup;
+    }
+    cmdPtr = newCmdPtr;
+    depth++;
+
+    /*
+     * See whether we have a nested ensemble. If we do, we can go round the
+     * mulberry bush again, consuming the next word.
+     */
+
+    if (cmdPtr->compileProc == TclCompileEnsemble) {
+	tokenPtr = TokenAfter(tokenPtr);
+	if (parsePtr->numWords < depth + 1
+		|| tokenPtr->type != TCL_TOKEN_SIMPLE_WORD) {
+	    /*
+	     * Too hard because the user has done something unpleasant like
+	     * omitting the sub-ensemble's command name or used a non-constant
+	     * name for a sub-ensemble's command name; we respond by bailing
+	     * out completely (this is a rare case). [Bug 6d2f249a01]
+	     */
+
+	    goto cleanup;
+	}
+	ensemble = (Tcl_Command) cmdPtr;
+	goto checkNextWord;
+    }
+
+    /*
+     * Now we've done the mapping process, can now actually try to compile.
+     * If there is a subcommand compiler and that successfully produces code,
+     * we'll use that. Otherwise, we fall back to generating opcodes to do the
+     * invoke at runtime.
+     */
+
+    invokeAnyway = 1;
+    if (TCL_OK == TclAttemptCompileProc(interp, parsePtr, depth, cmdPtr,
+	    envPtr)) {
+	ourResult = TCL_OK;
+	goto cleanup;
+    }
+
+    /*
+     * Throw out any line information generated by the failed compile attempt.
+     */
+
+    while (mapPtr->nuloc - 1 > eclIndex) {
+        mapPtr->nuloc--;
+        ckfree(mapPtr->loc[mapPtr->nuloc].line);
+        mapPtr->loc[mapPtr->nuloc].line = NULL;
+    }
+
+    /*
+     * Reset the index of next command.  Toss out any from failed nested
+     * partial compiles.
+     */
+
+    envPtr->numCommands = mapPtr->nuloc;
+
+    /*
+     * Failed to do a full compile for some reason. Try to do a direct invoke
+     * instead of going through the ensemble lookup process again.
+     */
+
+  failed:
+    if (depth < 250) {
+	if (depth > 1) {
+	    if (!invokeAnyway) {
+		cmdPtr = oldCmdPtr;
+		depth--;
+	    }
+	}
+	/*
+	 * The length of the "replaced" list must be depth-1.  Trim back
+	 * any extra elements that might have been appended by failing
+	 * pathways above.
+	 */
+	(void) Tcl_ListObjReplace(NULL, replaced, depth-1, LIST_MAX, 0, NULL);
+
+	/*
+	 * TODO: Reconsider whether we ought to call CompileToInvokedCommand()
+	 * when depth==1.  In that case we are choosing to emit the
+	 * INST_INVOKE_REPLACE bytecode when there is in fact no replacing
+	 * to be done.  It would be equally functional and presumably more
+	 * performant to fall through to cleanup below, return TCL_ERROR,
+	 * and let the compiler harness emit the INST_INVOKE_STK
+	 * implementation for us.
+	 */
+
+	CompileToInvokedCommand(interp, parsePtr, replaced, cmdPtr, envPtr);
+	ourResult = TCL_OK;
+    }
+
+    /*
+     * Release the memory we allocated. If we've got here, we've either done
+     * something useful or we're in a case that we can't compile at all and
+     * we're just giving up.
+     */
+
+  cleanup:
+    Tcl_DecrRefCount(replaced);
+    return ourResult;
+}
+
+int
+TclAttemptCompileProc(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    int depth,
+    Command *cmdPtr,
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    int result, i;
+    Tcl_Token *saveTokenPtr = parsePtr->tokenPtr;
+    int savedStackDepth = envPtr->currStackDepth;
+    unsigned savedCodeNext = envPtr->codeNext - envPtr->codeStart;
+    int savedAuxDataArrayNext = envPtr->auxDataArrayNext;
+    int savedExceptArrayNext = envPtr->exceptArrayNext;
+#ifdef TCL_COMPILE_DEBUG
+    int savedExceptDepth = envPtr->exceptDepth;
+#endif
+    DefineLineInformation;
+
+    if (cmdPtr->compileProc == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Advance parsePtr->tokenPtr so that it points at the last subcommand.
+     * This will be wrong, but it will not matter, and it will put the
+     * tokens for the arguments in the right place without the needed to
+     * allocate a synthetic Tcl_Parse struct, or copy tokens around.
+     */
+
+    for (i = 0; i < depth - 1; i++) {
+	parsePtr->tokenPtr = TokenAfter(parsePtr->tokenPtr);
+    }
+    parsePtr->numWords -= (depth - 1);
+
+    /*
+     * Shift the line information arrays to account for different word
+     * index values.
+     */
+
+    mapPtr->loc[eclIndex].line += (depth - 1);
+    mapPtr->loc[eclIndex].next += (depth - 1);
+
+    /*
+     * Hand off compilation to the subcommand compiler. At last!
+     */
+
+    result = cmdPtr->compileProc(interp, parsePtr, cmdPtr, envPtr);
+
+    /*
+     * Undo the shift.
+     */
+
+    mapPtr->loc[eclIndex].line -= (depth - 1);
+    mapPtr->loc[eclIndex].next -= (depth - 1);
+
+    parsePtr->numWords += (depth - 1);
+    parsePtr->tokenPtr = saveTokenPtr;
+
+    /*
+     * If our target failed to compile, revert any data from failed partial
+     * compiles.  Note that envPtr->numCommands need not be checked because
+     * we avoid compiling subcommands that recursively call TclCompileScript().
+     */
+
+#ifdef TCL_COMPILE_DEBUG
+    if (envPtr->exceptDepth != savedExceptDepth) {
+	Tcl_Panic("ExceptionRange Starts and Ends do not balance");
+    }
+#endif
+
+    if (result != TCL_OK) {
+	ExceptionAux *auxPtr = envPtr->exceptAuxArrayPtr;
+
+	for (i = 0; i < savedExceptArrayNext; i++) {
+	    while (auxPtr->numBreakTargets > 0
+		    && auxPtr->breakTargets[auxPtr->numBreakTargets - 1]
+		    >= savedCodeNext) {
+		auxPtr->numBreakTargets--;
+	    }
+	    while (auxPtr->numContinueTargets > 0
+		    && auxPtr->continueTargets[auxPtr->numContinueTargets - 1]
+		    >= savedCodeNext) {
+		auxPtr->numContinueTargets--;
+	    }
+	    auxPtr++;
+	}
+	envPtr->exceptArrayNext = savedExceptArrayNext;
+
+	if (savedAuxDataArrayNext != envPtr->auxDataArrayNext) {
+	    AuxData *auxDataPtr = envPtr->auxDataArrayPtr;
+	    AuxData *auxDataEnd = auxDataPtr;
+
+	    auxDataPtr += savedAuxDataArrayNext;
+	    auxDataEnd += envPtr->auxDataArrayNext;
+
+	    while (auxDataPtr < auxDataEnd) {
+		if (auxDataPtr->type->freeProc != NULL) {
+		    auxDataPtr->type->freeProc(auxDataPtr->clientData);
+		}
+		auxDataPtr++;
+	    }
+	    envPtr->auxDataArrayNext = savedAuxDataArrayNext;
+	}
+	envPtr->currStackDepth = savedStackDepth;
+	envPtr->codeNext = envPtr->codeStart + savedCodeNext;
+#ifdef TCL_COMPILE_DEBUG
+    } else {
+	/*
+	 * Confirm that the command compiler generated a single value on
+	 * the stack as its result. This is only done in debugging mode,
+	 * as it *should* be correct and normal users have no reasonable
+	 * way to fix it anyway.
+	 */
+
+	int diff = envPtr->currStackDepth - savedStackDepth;
+
+	if (diff != 1) {
+	    Tcl_Panic("bad stack adjustment when compiling"
+		    " %.*s (was %d instead of 1)", parsePtr->tokenPtr->size,
+		    parsePtr->tokenPtr->start, diff);
+	}
+#endif
+    }
+
+    return result;
+}
+
+/*
+ * How to compile a subcommand to a _replacing_ invoke of its implementation
+ * command.
+ */
+
+static void
+CompileToInvokedCommand(
+    Tcl_Interp *interp,
+    Tcl_Parse *parsePtr,
+    Tcl_Obj *replacements,
+    Command *cmdPtr,
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Token *tokPtr;
+    Tcl_Obj *objPtr, **words;
+    char *bytes;
+    int i, numWords, cmdLit, extraLiteralFlags = LITERAL_CMD_NAME;
+    DefineLineInformation;
+
+    /*
+     * Push the words of the command. Take care; the command words may be
+     * scripts that have backslashes in them, and [info frame 0] can see the
+     * difference. Hence the call to TclContinuationsEnterDerived...
+     */
+
+    Tcl_ListObjGetElements(NULL, replacements, &numWords, &words);
+    for (i = 0, tokPtr = parsePtr->tokenPtr; i < parsePtr->numWords;
+	    i++, tokPtr = TokenAfter(tokPtr)) {
+	if (i > 0 && i < numWords+1) {
+	    bytes = TclGetString(words[i-1]);
+	    PushLiteral(envPtr, bytes, words[i-1]->length);
+	    continue;
+	}
+
+	SetLineInformation(i);
+	if (tokPtr->type == TCL_TOKEN_SIMPLE_WORD) {
+	    int literal = TclRegisterLiteral(envPtr,
+		    tokPtr[1].start, tokPtr[1].size, 0);
+
+	    if (envPtr->clNext) {
+		TclContinuationsEnterDerived(
+			TclFetchLiteral(envPtr, literal),
+			tokPtr[1].start - envPtr->source,
+			envPtr->clNext);
+	    }
+	    TclEmitPush(literal, envPtr);
+	} else {
+	    CompileTokens(envPtr, tokPtr, interp);
+	}
+    }
+
+    /*
+     * Push the name of the command we're actually dispatching to as part of
+     * the implementation.
+     */
+
+    objPtr = Tcl_NewObj();
+    Tcl_GetCommandFullName(interp, (Tcl_Command) cmdPtr, objPtr);
+    bytes = TclGetString(objPtr);
+    if ((cmdPtr != NULL) && (cmdPtr->flags & CMD_VIA_RESOLVER)) {
+	extraLiteralFlags |= LITERAL_UNSHARED;
+    }
+    cmdLit = TclRegisterLiteral(envPtr, bytes, objPtr->length, extraLiteralFlags);
+    TclSetCmdNameObj(interp, TclFetchLiteral(envPtr, cmdLit), cmdPtr);
+    TclEmitPush(cmdLit, envPtr);
+    TclDecrRefCount(objPtr);
+
+    /*
+     * Do the replacing dispatch.
+     */
+
+    TclEmitInvoke(envPtr, INST_INVOKE_REPLACE, parsePtr->numWords,numWords+1);
+}
+
+/*
+ * Helpers that do issuing of instructions for commands that "don't have
+ * compilers" (well, they do; these). They all work by just generating base
+ * code to invoke the command; they're intended for ensemble subcommands so
+ * that the costs of INST_INVOKE_REPLACE can be avoided where we can work out
+ * that they're not needed.
+ *
+ * Note that these are NOT suitable for commands where there's an argument
+ * that is a script, as an [info level] or [info frame] in the inner context
+ * can see the difference.
+ */
+
+static int
+CompileBasicNArgCommand(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    Tcl_Obj *objPtr = Tcl_NewObj();
+
+    Tcl_IncrRefCount(objPtr);
+    Tcl_GetCommandFullName(interp, (Tcl_Command) cmdPtr, objPtr);
+    TclCompileInvocation(interp, parsePtr->tokenPtr, objPtr,
+	    parsePtr->numWords, envPtr);
+    Tcl_DecrRefCount(objPtr);
+    return TCL_OK;
+}
+
+int
+TclCompileBasic0ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 1) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic1ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic2ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic3ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic0Or1ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 1 && parsePtr->numWords != 2) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic1Or2ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 2 && parsePtr->numWords != 3) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic2Or3ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords != 3 && parsePtr->numWords != 4) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic0To2ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords < 1 || parsePtr->numWords > 3) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasic1To3ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords < 2 || parsePtr->numWords > 4) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasicMin0ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords < 1) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasicMin1ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords < 2) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+int
+TclCompileBasicMin2ArgCmd(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Parse *parsePtr,	/* Points to a parse structure for the command
+				 * created by Tcl_ParseCommand. */
+    Command *cmdPtr,		/* Points to defintion of command being
+				 * compiled. */
+    CompileEnv *envPtr)		/* Holds resulting instructions. */
+{
+    /*
+     * Verify that the number of arguments is correct; that's the only case
+     * that we know will avoid the call to Tcl_WrongNumArgs() at invoke time,
+     * which is the only code that sees the shenanigans of ensemble dispatch.
+     */
+
+    if (parsePtr->numWords < 3) {
+	return TCL_ERROR;
+    }
+
+    return CompileBasicNArgCommand(interp, parsePtr, cmdPtr, envPtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclEnv.c b/generic/tclEnv.c
new file mode 100644
index 0000000..66ddb57
--- /dev/null
+++ b/generic/tclEnv.c
@@ -0,0 +1,744 @@
+/*
+ * tclEnv.c --
+ *
+ *	Tcl support for environment variables, including a setenv function.
+ *	This file contains the generic portion of the environment module. It
+ *	is primarily responsible for keeping the "env" arrays in sync with the
+ *	system environment variables.
+ *
+ * Copyright (c) 1991-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+TCL_DECLARE_MUTEX(envMutex)	/* To serialize access to environ. */
+
+static struct {
+    int cacheSize;		/* Number of env strings in cache. */
+    char **cache;		/* Array containing all of the environment
+				 * strings that Tcl has allocated. */
+#ifndef USE_PUTENV
+    char **ourEnviron;		/* Cache of the array that we allocate. We
+				 * need to track this in case another
+				 * subsystem swaps around the environ array
+				 * like we do. */
+    int ourEnvironSize;		/* Non-zero means that the environ array was
+				 * malloced and has this many total entries
+				 * allocated to it (not all may be in use at
+				 * once). Zero means that the environment
+				 * array is in its original static state. */
+#endif
+} env;
+
+/*
+ * Declarations for local functions defined in this file:
+ */
+
+static char *		EnvTraceProc(ClientData clientData, Tcl_Interp *interp,
+			    const char *name1, const char *name2, int flags);
+static void		ReplaceString(const char *oldStr, char *newStr);
+MODULE_SCOPE void	TclSetEnv(const char *name, const char *value);
+MODULE_SCOPE void	TclUnsetEnv(const char *name);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetupEnv --
+ *
+ *	This function is invoked for an interpreter to make environment
+ *	variables accessible from that interpreter via the "env" associative
+ *	array.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interpreter is added to a list of interpreters managed by us, so
+ *	that its view of envariables can be kept consistent with the view in
+ *	other interpreters. If this is the first call to TclSetupEnv, then
+ *	additional initialization happens, such as copying the environment to
+ *	dynamically-allocated space for ease of management.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetupEnv(
+    Tcl_Interp *interp)		/* Interpreter whose "env" array is to be
+				 * managed. */
+{
+    Var *varPtr, *arrayPtr;
+    Tcl_Obj *varNamePtr;
+    Tcl_DString envString;
+    Tcl_HashTable namesHash;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+
+    /*
+     * Synchronize the values in the environ array with the contents of the
+     * Tcl "env" variable. To do this:
+     *    1) Remove the trace that fires when the "env" var is updated.
+     *    2) Find the existing contents of the "env", storing in a hash table.
+     *    3) Create/update elements for each environ variable, removing
+     *	     elements from the hash table as we go.
+     *    4) Remove the elements for each remaining entry in the hash table,
+     *	     which must have existed before yet have no analog in the environ
+     *	     variable.
+     *    5) Add a trace that synchronizes the "env" array.
+     */
+
+    Tcl_UntraceVar2(interp, "env", NULL,
+	    TCL_GLOBAL_ONLY | TCL_TRACE_WRITES | TCL_TRACE_UNSETS |
+	    TCL_TRACE_READS | TCL_TRACE_ARRAY, EnvTraceProc, NULL);
+
+    /*
+     * Find out what elements are currently in the global env array.
+     */
+
+    TclNewLiteralStringObj(varNamePtr, "env");
+    Tcl_IncrRefCount(varNamePtr);
+    Tcl_InitObjHashTable(&namesHash);
+    varPtr = TclObjLookupVarEx(interp, varNamePtr, NULL, TCL_GLOBAL_ONLY,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+    TclFindArrayPtrElements(varPtr, &namesHash);
+
+    /*
+     * Go through the environment array and transfer its values into Tcl. At
+     * the same time, remove those elements we add/update from the hash table
+     * of existing elements, so that after this part processes, that table
+     * will hold just the parts to remove.
+     */
+
+    if (environ[0] != NULL) {
+	int i;
+
+	Tcl_MutexLock(&envMutex);
+	for (i = 0; environ[i] != NULL; i++) {
+	    Tcl_Obj *obj1, *obj2;
+	    char *p1, *p2;
+
+	    p1 = Tcl_ExternalToUtfDString(NULL, environ[i], -1, &envString);
+	    p2 = strchr(p1, '=');
+	    if (p2 == NULL) {
+		/*
+		 * This condition seem to happen occasionally under some
+		 * versions of Solaris, or when encoding accidents swallow the
+		 * '='; ignore the entry.
+		 */
+
+		continue;
+	    }
+	    p2++;
+	    p2[-1] = '\0';
+	    obj1 = Tcl_NewStringObj(p1, -1);
+	    obj2 = Tcl_NewStringObj(p2, -1);
+	    Tcl_DStringFree(&envString);
+
+	    Tcl_IncrRefCount(obj1);
+	    Tcl_IncrRefCount(obj2);
+	    Tcl_ObjSetVar2(interp, varNamePtr, obj1, obj2, TCL_GLOBAL_ONLY);
+	    hPtr = Tcl_FindHashEntry(&namesHash, obj1);
+	    if (hPtr != NULL) {
+		Tcl_DeleteHashEntry(hPtr);
+	    }
+	    Tcl_DecrRefCount(obj1);
+	    Tcl_DecrRefCount(obj2);
+	}
+	Tcl_MutexUnlock(&envMutex);
+    }
+
+    /*
+     * Delete those elements that existed in the array but which had no
+     * counterparts in the environment array.
+     */
+
+    for (hPtr=Tcl_FirstHashEntry(&namesHash, &search); hPtr!=NULL;
+	    hPtr=Tcl_NextHashEntry(&search)) {
+	Tcl_Obj *elemName = Tcl_GetHashValue(hPtr);
+
+	TclObjUnsetVar2(interp, varNamePtr, elemName, TCL_GLOBAL_ONLY);
+    }
+    Tcl_DeleteHashTable(&namesHash);
+    Tcl_DecrRefCount(varNamePtr);
+
+    /*
+     * Re-establish the trace.
+     */
+
+    Tcl_TraceVar2(interp, "env", NULL,
+	    TCL_GLOBAL_ONLY | TCL_TRACE_WRITES | TCL_TRACE_UNSETS |
+	    TCL_TRACE_READS | TCL_TRACE_ARRAY, EnvTraceProc, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetEnv --
+ *
+ *	Set an environment variable, replacing an existing value or creating a
+ *	new variable if there doesn't exist a variable by the given name. This
+ *	function is intended to be a stand-in for the UNIX "setenv" function
+ *	so that applications using that function will interface properly to
+ *	Tcl. To make it a stand-in, the Makefile must define "TclSetEnv" to
+ *	"setenv".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The environ array gets updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetEnv(
+    const char *name,		/* Name of variable whose value is to be set
+				 * (UTF-8). */
+    const char *value)		/* New value for variable (UTF-8). */
+{
+    Tcl_DString envString;
+    unsigned nameLength, valueLength;
+    int index, length;
+    char *p, *oldValue;
+    const char *p2;
+
+    /*
+     * Figure out where the entry is going to go. If the name doesn't already
+     * exist, enlarge the array if necessary to make room. If the name exists,
+     * free its old entry.
+     */
+
+    Tcl_MutexLock(&envMutex);
+    index = TclpFindVariable(name, &length);
+
+    if (index == -1) {
+#ifndef USE_PUTENV
+	/*
+	 * We need to handle the case where the environment may be changed
+	 * outside our control. ourEnvironSize is only valid if the current
+	 * environment is the one we allocated. [Bug 979640]
+	 */
+
+	if ((env.ourEnviron != environ) || (length+2 > env.ourEnvironSize)) {
+	    char **newEnviron = ckalloc((length + 5) * sizeof(char *));
+
+	    memcpy(newEnviron, environ, length * sizeof(char *));
+	    if ((env.ourEnvironSize != 0) && (env.ourEnviron != NULL)) {
+		ckfree(env.ourEnviron);
+	    }
+	    environ = env.ourEnviron = newEnviron;
+	    env.ourEnvironSize = length + 5;
+	}
+	index = length;
+	environ[index + 1] = NULL;
+#endif /* USE_PUTENV */
+	oldValue = NULL;
+	nameLength = strlen(name);
+    } else {
+	const char *env;
+
+	/*
+	 * Compare the new value to the existing value. If they're the same
+	 * then quit immediately (e.g. don't rewrite the value or propagate it
+	 * to other interpreters). Otherwise, when there are N interpreters
+	 * there will be N! propagations of the same value among the
+	 * interpreters.
+	 */
+
+	env = Tcl_ExternalToUtfDString(NULL, environ[index], -1, &envString);
+	if (strcmp(value, env + (length + 1)) == 0) {
+	    Tcl_DStringFree(&envString);
+	    Tcl_MutexUnlock(&envMutex);
+	    return;
+	}
+	Tcl_DStringFree(&envString);
+
+	oldValue = environ[index];
+	nameLength = (unsigned) length;
+    }
+
+    /*
+     * Create a new entry. Build a complete UTF string that contains a
+     * "name=value" pattern. Then convert the string to the native encoding,
+     * and set the environ array value.
+     */
+
+    valueLength = strlen(value);
+    p = ckalloc(nameLength + valueLength + 2);
+    memcpy(p, name, nameLength);
+    p[nameLength] = '=';
+    memcpy(p+nameLength+1, value, valueLength+1);
+    p2 = Tcl_UtfToExternalDString(NULL, p, -1, &envString);
+
+    /*
+     * Copy the native string to heap memory.
+     */
+
+    p = ckrealloc(p, Tcl_DStringLength(&envString) + 1);
+    memcpy(p, p2, (unsigned) Tcl_DStringLength(&envString) + 1);
+    Tcl_DStringFree(&envString);
+
+#ifdef USE_PUTENV
+    /*
+     * Update the system environment.
+     */
+
+    putenv(p);
+    index = TclpFindVariable(name, &length);
+#else
+    environ[index] = p;
+#endif /* USE_PUTENV */
+
+    /*
+     * Watch out for versions of putenv that copy the string (e.g. VC++). In
+     * this case we need to free the string immediately. Otherwise update the
+     * string in the cache.
+     */
+
+    if ((index != -1) && (environ[index] == p)) {
+	ReplaceString(oldValue, p);
+#ifdef HAVE_PUTENV_THAT_COPIES
+    } else {
+	/*
+	 * This putenv() copies instead of taking ownership.
+	 */
+
+	ckfree(p);
+#endif /* HAVE_PUTENV_THAT_COPIES */
+    }
+
+    Tcl_MutexUnlock(&envMutex);
+
+    if (!strcmp(name, "HOME")) {
+	/*
+	 * If the user's home directory has changed, we must invalidate the
+	 * filesystem cache, because '~' expansions will now be incorrect.
+	 */
+
+	Tcl_FSMountsChanged(NULL);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PutEnv --
+ *
+ *	Set an environment variable. Similar to setenv except that the
+ *	information is passed in a single string of the form NAME=value,
+ *	rather than as separate name strings. This function is intended to be
+ *	a stand-in for the UNIX "putenv" function so that applications using
+ *	that function will interface properly to Tcl. To make it a stand-in,
+ *	the Makefile will define "Tcl_PutEnv" to "putenv".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The environ array gets updated, as do all of the interpreters that we
+ *	manage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_PutEnv(
+    const char *assignment)	/* Info about environment variable in the form
+				 * NAME=value. (native) */
+{
+    Tcl_DString nameString;
+    const char *name;
+    char *value;
+
+    if (assignment == NULL) {
+	return 0;
+    }
+
+    /*
+     * First convert the native string to UTF. Then separate the string into
+     * name and value parts, and call TclSetEnv to do all of the real work.
+     */
+
+    name = Tcl_ExternalToUtfDString(NULL, assignment, -1, &nameString);
+    value = strchr(name, '=');
+
+    if ((value != NULL) && (value != name)) {
+	value[0] = '\0';
+	TclSetEnv(name, value+1);
+    }
+
+    Tcl_DStringFree(&nameString);
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclUnsetEnv --
+ *
+ *	Remove an environment variable, updating the "env" arrays in all
+ *	interpreters managed by us. This function is intended to replace the
+ *	UNIX "unsetenv" function (but to do this the Makefile must be modified
+ *	to redefine "TclUnsetEnv" to "unsetenv".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Interpreters are updated, as is environ.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclUnsetEnv(
+    const char *name)		/* Name of variable to remove (UTF-8). */
+{
+    char *oldValue;
+    int length;
+    int index;
+#ifdef USE_PUTENV_FOR_UNSET
+    Tcl_DString envString;
+    char *string;
+#else
+    char **envPtr;
+#endif /* USE_PUTENV_FOR_UNSET */
+
+    Tcl_MutexLock(&envMutex);
+    index = TclpFindVariable(name, &length);
+
+    /*
+     * First make sure that the environment variable exists to avoid doing
+     * needless work and to avoid recursion on the unset.
+     */
+
+    if (index == -1) {
+	Tcl_MutexUnlock(&envMutex);
+	return;
+    }
+
+    /*
+     * Remember the old value so we can free it if Tcl created the string.
+     */
+
+    oldValue = environ[index];
+
+    /*
+     * Update the system environment. This must be done before we update the
+     * interpreters or we will recurse.
+     */
+
+#ifdef USE_PUTENV_FOR_UNSET
+    /*
+     * For those platforms that support putenv to unset, Linux indicates
+     * that no = should be included, and Windows requires it.
+     */
+
+#if defined(_WIN32)
+    string = ckalloc(length + 2);
+    memcpy(string, name, (size_t) length);
+    string[length] = '=';
+    string[length+1] = '\0';
+#else
+    string = ckalloc(length + 1);
+    memcpy(string, name, (size_t) length);
+    string[length] = '\0';
+#endif /* _WIN32 */
+
+    Tcl_UtfToExternalDString(NULL, string, -1, &envString);
+    string = ckrealloc(string, Tcl_DStringLength(&envString) + 1);
+    memcpy(string, Tcl_DStringValue(&envString),
+	    (unsigned) Tcl_DStringLength(&envString)+1);
+    Tcl_DStringFree(&envString);
+
+    putenv(string);
+
+    /*
+     * Watch out for versions of putenv that copy the string (e.g. VC++). In
+     * this case we need to free the string immediately. Otherwise update the
+     * string in the cache.
+     */
+
+    if (environ[index] == string) {
+	ReplaceString(oldValue, string);
+#ifdef HAVE_PUTENV_THAT_COPIES
+    } else {
+	/*
+	 * This putenv() copies instead of taking ownership.
+	 */
+
+	ckfree(string);
+#endif /* HAVE_PUTENV_THAT_COPIES */
+    }
+#else /* !USE_PUTENV_FOR_UNSET */
+    for (envPtr = environ+index+1; ; envPtr++) {
+	envPtr[-1] = *envPtr;
+	if (*envPtr == NULL) {
+	    break;
+	}
+    }
+    ReplaceString(oldValue, NULL);
+#endif /* USE_PUTENV_FOR_UNSET */
+
+    Tcl_MutexUnlock(&envMutex);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetEnv --
+ *
+ *	Retrieve the value of an environment variable.
+ *
+ * Results:
+ *	The result is a pointer to a string specifying the value of the
+ *	environment variable, or NULL if that environment variable does not
+ *	exist. Storage for the result string is allocated in valuePtr; the
+ *	caller must call Tcl_DStringFree() when the result is no longer
+ *	needed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+TclGetEnv(
+    const char *name,		/* Name of environment variable to find
+				 * (UTF-8). */
+    Tcl_DString *valuePtr)	/* Uninitialized or free DString in which the
+				 * value of the environment variable is
+				 * stored. */
+{
+    int length, index;
+    const char *result;
+
+    Tcl_MutexLock(&envMutex);
+    index = TclpFindVariable(name, &length);
+    result = NULL;
+    if (index != -1) {
+	Tcl_DString envStr;
+
+	result = Tcl_ExternalToUtfDString(NULL, environ[index], -1, &envStr);
+	result += length;
+	if (*result == '=') {
+	    result++;
+	    Tcl_DStringInit(valuePtr);
+	    Tcl_DStringAppend(valuePtr, result, -1);
+	    result = Tcl_DStringValue(valuePtr);
+	} else {
+	    result = NULL;
+	}
+	Tcl_DStringFree(&envStr);
+    }
+    Tcl_MutexUnlock(&envMutex);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EnvTraceProc --
+ *
+ *	This function is invoked whenever an environment variable is read,
+ *	modified or deleted. It propagates the change to the global "environ"
+ *	array.
+ *
+ * Results:
+ *	Returns NULL to indicate success, or an error-message if the array
+ *	element being handled doesn't exist.
+ *
+ * Side effects:
+ *	Environment variable changes get propagated. If the whole "env" array
+ *	is deleted, then we stop managing things for this interpreter (usually
+ *	this happens because the whole interpreter is being deleted).
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static char *
+EnvTraceProc(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter whose "env" variable is being
+				 * modified. */
+    const char *name1,		/* Better be "env". */
+    const char *name2,		/* Name of variable being modified, or NULL if
+				 * whole array is being deleted (UTF-8). */
+    int flags)			/* Indicates what's happening. */
+{
+    /*
+     * For array traces, let TclSetupEnv do all the work.
+     */
+
+    if (flags & TCL_TRACE_ARRAY) {
+	TclSetupEnv(interp);
+	return NULL;
+    }
+
+    /*
+     * If name2 is NULL, then return and do nothing.
+     */
+
+    if (name2 == NULL) {
+	return NULL;
+    }
+
+    /*
+     * If a value is being set, call TclSetEnv to do all of the work.
+     */
+
+    if (flags & TCL_TRACE_WRITES) {
+	const char *value;
+
+	value = Tcl_GetVar2(interp, "env", name2, TCL_GLOBAL_ONLY);
+	TclSetEnv(name2, value);
+    }
+
+    /*
+     * If a value is being read, call TclGetEnv to do all of the work.
+     */
+
+    if (flags & TCL_TRACE_READS) {
+	Tcl_DString valueString;
+	const char *value = TclGetEnv(name2, &valueString);
+
+	if (value == NULL) {
+	    return (char *) "no such variable";
+	}
+	Tcl_SetVar2(interp, name1, name2, value, 0);
+	Tcl_DStringFree(&valueString);
+    }
+
+    /*
+     * For unset traces, let TclUnsetEnv do all the work.
+     */
+
+    if (flags & TCL_TRACE_UNSETS) {
+	TclUnsetEnv(name2);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReplaceString --
+ *
+ *	Replace one string with another in the environment variable cache. The
+ *	cache keeps track of all of the environment variables that Tcl has
+ *	modified so they can be freed later.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May free the old string.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReplaceString(
+    const char *oldStr,		/* Old environment string. */
+    char *newStr)		/* New environment string. */
+{
+    int i;
+
+    /*
+     * Check to see if the old value was allocated by Tcl. If so, it needs to
+     * be deallocated to avoid memory leaks. Note that this algorithm is O(n),
+     * not O(1). This will result in n-squared behavior if lots of environment
+     * changes are being made.
+     */
+
+    for (i = 0; i < env.cacheSize; i++) {
+	if (env.cache[i]==oldStr || env.cache[i]==NULL) {
+	    break;
+	}
+    }
+    if (i < env.cacheSize) {
+	/*
+	 * Replace or delete the old value.
+	 */
+
+	if (env.cache[i]) {
+	    ckfree(env.cache[i]);
+	}
+
+	if (newStr) {
+	    env.cache[i] = newStr;
+	} else {
+	    for (; i < env.cacheSize-1; i++) {
+		env.cache[i] = env.cache[i+1];
+	    }
+	    env.cache[env.cacheSize-1] = NULL;
+	}
+    } else {
+	/*
+	 * We need to grow the cache in order to hold the new string.
+	 */
+
+	const int growth = 5;
+
+	env.cache = ckrealloc(env.cache,
+		(env.cacheSize + growth) * sizeof(char *));
+	env.cache[env.cacheSize] = newStr;
+	(void) memset(env.cache+env.cacheSize+1, 0,
+		(size_t) (growth-1) * sizeof(char *));
+	env.cacheSize += growth;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeEnvironment --
+ *
+ *	This function releases any storage allocated by this module that isn't
+ *	still in use by the global environment. Any strings that are still in
+ *	the environment will be leaked.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May deallocate storage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeEnvironment(void)
+{
+    /*
+     * For now we just deallocate the cache array and none of the environment
+     * strings. This may leak more memory that strictly necessary, since some
+     * of the strings may no longer be in the environment. However,
+     * determining which ones are ok to delete is n-squared, and is pretty
+     * unlikely, so we don't bother.
+     */
+
+    if (env.cache) {
+	ckfree(env.cache);
+	env.cache = NULL;
+	env.cacheSize = 0;
+#ifndef USE_PUTENV
+	env.ourEnvironSize = 0;
+#endif
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclEvent.c b/generic/tclEvent.c
new file mode 100644
index 0000000..49fd2ae
--- /dev/null
+++ b/generic/tclEvent.c
@@ -0,0 +1,1627 @@
+/*
+ * tclEvent.c --
+ *
+ *	This file implements some general event related interfaces including
+ *	background errors, exit handlers, and the "vwait" and "update" command
+ *	functions.
+ *
+ * Copyright (c) 1990-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2004 by Zoran Vasiljevic.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * The data structure below is used to report background errors. One such
+ * structure is allocated for each error; it holds information about the
+ * interpreter and the error until an idle handler command can be invoked.
+ */
+
+typedef struct BgError {
+    Tcl_Obj *errorMsg;		/* Copy of the error message (the interp's
+				 * result when the error occurred). */
+    Tcl_Obj *returnOpts;	/* Active return options when the error
+				 * occurred */
+    struct BgError *nextPtr;	/* Next in list of all pending error reports
+				 * for this interpreter, or NULL for end of
+				 * list. */
+} BgError;
+
+/*
+ * One of the structures below is associated with the "tclBgError" assoc data
+ * for each interpreter. It keeps track of the head and tail of the list of
+ * pending background errors for the interpreter.
+ */
+
+typedef struct {
+    Tcl_Interp *interp;		/* Interpreter in which error occurred. */
+    Tcl_Obj *cmdPrefix;		/* First word(s) of the handler command */
+    BgError *firstBgPtr;	/* First in list of all background errors
+				 * waiting to be processed for this
+				 * interpreter (NULL if none). */
+    BgError *lastBgPtr;		/* Last in list of all background errors
+				 * waiting to be processed for this
+				 * interpreter (NULL if none). */
+} ErrAssocData;
+
+/*
+ * For each exit handler created with a call to Tcl_Create(Late)ExitHandler
+ * there is a structure of the following type:
+ */
+
+typedef struct ExitHandler {
+    Tcl_ExitProc *proc;		/* Function to call when process exits. */
+    ClientData clientData;	/* One word of information to pass to proc. */
+    struct ExitHandler *nextPtr;/* Next in list of all exit handlers for this
+				 * application, or NULL for end of list. */
+} ExitHandler;
+
+/*
+ * There is both per-process and per-thread exit handlers. The first list is
+ * controlled by a mutex. The other is in thread local storage.
+ */
+
+static ExitHandler *firstExitPtr = NULL;
+				/* First in list of all exit handlers for
+				 * application. */
+static ExitHandler *firstLateExitPtr = NULL;
+				/* First in list of all late exit handlers for
+				 * application. */
+TCL_DECLARE_MUTEX(exitMutex)
+
+/*
+ * This variable is set to 1 when Tcl_Exit is called. The variable is checked
+ * by TclInExit() to allow different behavior for exit-time processing, e.g.,
+ * in closing of files and pipes.
+ */
+
+static int inExit = 0;
+
+static int subsystemsInitialized = 0;
+
+/*
+ * This variable contains the application wide exit handler. It will be called
+ * by Tcl_Exit instead of the C-runtime exit if this variable is set to a
+ * non-NULL value.
+ */
+
+static TCL_NORETURN1 Tcl_ExitProc *appExitPtr = NULL;
+
+typedef struct ThreadSpecificData {
+    ExitHandler *firstExitPtr;	/* First in list of all exit handlers for this
+				 * thread. */
+    int inExit;			/* True when this thread is exiting. This is
+				 * used as a hack to decide to close the
+				 * standard channels. */
+} ThreadSpecificData;
+static Tcl_ThreadDataKey dataKey;
+
+#ifdef TCL_THREADS
+typedef struct {
+    Tcl_ThreadCreateProc *proc;	/* Main() function of the thread */
+    ClientData clientData;	/* The one argument to Main() */
+} ThreadClientData;
+static Tcl_ThreadCreateType NewThreadProc(ClientData clientData);
+#endif /* TCL_THREADS */
+
+/*
+ * Prototypes for functions referenced only in this file:
+ */
+
+static void		BgErrorDeleteProc(ClientData clientData,
+			    Tcl_Interp *interp);
+static void		HandleBgErrors(ClientData clientData);
+static char *		VwaitVarProc(ClientData clientData,
+			    Tcl_Interp *interp, const char *name1,
+			    const char *name2, int flags);
+static void		InvokeExitHandlers(void);
+static void		FinalizeThread(int quick);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_BackgroundError --
+ *
+ *	This function is invoked to handle errors that occur in Tcl commands
+ *	that are invoked in "background" (e.g. from event or timer bindings).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	A handler command is invoked later as an idle handler to process the
+ *	error, passing it the interp result and return options.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_BackgroundError(
+    Tcl_Interp *interp)		/* Interpreter in which an error has
+				 * occurred. */
+{
+    Tcl_BackgroundException(interp, TCL_ERROR);
+}
+
+void
+Tcl_BackgroundException(
+    Tcl_Interp *interp,		/* Interpreter in which an exception has
+				 * occurred. */
+    int code)			/* The exception code value */
+{
+    BgError *errPtr;
+    ErrAssocData *assocPtr;
+
+    if (code == TCL_OK) {
+	return;
+    }
+
+    errPtr = ckalloc(sizeof(BgError));
+    errPtr->errorMsg = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(errPtr->errorMsg);
+    errPtr->returnOpts = Tcl_GetReturnOptions(interp, code);
+    Tcl_IncrRefCount(errPtr->returnOpts);
+    errPtr->nextPtr = NULL;
+
+    (void) TclGetBgErrorHandler(interp);
+    assocPtr = Tcl_GetAssocData(interp, "tclBgError", NULL);
+    if (assocPtr->firstBgPtr == NULL) {
+	assocPtr->firstBgPtr = errPtr;
+	Tcl_DoWhenIdle(HandleBgErrors, assocPtr);
+    } else {
+	assocPtr->lastBgPtr->nextPtr = errPtr;
+    }
+    assocPtr->lastBgPtr = errPtr;
+    Tcl_ResetResult(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * HandleBgErrors --
+ *
+ *	This function is invoked as an idle handler to process all of the
+ *	accumulated background errors.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on what actions the handler command takes for the errors.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+HandleBgErrors(
+    ClientData clientData)	/* Pointer to ErrAssocData structure. */
+{
+    ErrAssocData *assocPtr = clientData;
+    Tcl_Interp *interp = assocPtr->interp;
+    BgError *errPtr;
+
+    /*
+     * Not bothering to save/restore the interp state. Assume that any code
+     * that has interp state it needs to keep will make its own
+     * Tcl_SaveInterpState call before calling something like Tcl_DoOneEvent()
+     * that could lead us here.
+     */
+
+    Tcl_Preserve(assocPtr);
+    Tcl_Preserve(interp);
+    while (assocPtr->firstBgPtr != NULL) {
+	int code, prefixObjc;
+	Tcl_Obj **prefixObjv, **tempObjv;
+
+	/*
+	 * Note we copy the handler command prefix each pass through, so we do
+	 * support one handler setting another handler.
+	 */
+
+	Tcl_Obj *copyObj = TclListObjCopy(NULL, assocPtr->cmdPrefix);
+
+	errPtr = assocPtr->firstBgPtr;
+
+	Tcl_ListObjGetElements(NULL, copyObj, &prefixObjc, &prefixObjv);
+	tempObjv = ckalloc((prefixObjc+2) * sizeof(Tcl_Obj *));
+	memcpy(tempObjv, prefixObjv, prefixObjc*sizeof(Tcl_Obj *));
+	tempObjv[prefixObjc] = errPtr->errorMsg;
+	tempObjv[prefixObjc+1] = errPtr->returnOpts;
+	Tcl_AllowExceptions(interp);
+	code = Tcl_EvalObjv(interp, prefixObjc+2, tempObjv, TCL_EVAL_GLOBAL);
+
+	/*
+	 * Discard the command and the information about the error report.
+	 */
+
+	Tcl_DecrRefCount(copyObj);
+	Tcl_DecrRefCount(errPtr->errorMsg);
+	Tcl_DecrRefCount(errPtr->returnOpts);
+	assocPtr->firstBgPtr = errPtr->nextPtr;
+	ckfree(errPtr);
+	ckfree(tempObjv);
+
+	if (code == TCL_BREAK) {
+	    /*
+	     * Break means cancel any remaining error reports for this
+	     * interpreter.
+	     */
+
+	    while (assocPtr->firstBgPtr != NULL) {
+		errPtr = assocPtr->firstBgPtr;
+		assocPtr->firstBgPtr = errPtr->nextPtr;
+		Tcl_DecrRefCount(errPtr->errorMsg);
+		Tcl_DecrRefCount(errPtr->returnOpts);
+		ckfree(errPtr);
+	    }
+	} else if ((code == TCL_ERROR) && !Tcl_IsSafe(interp)) {
+	    Tcl_Channel errChannel = Tcl_GetStdChannel(TCL_STDERR);
+
+	    if (errChannel != NULL) {
+		Tcl_Obj *options = Tcl_GetReturnOptions(interp, code);
+		Tcl_Obj *keyPtr, *valuePtr = NULL;
+
+		TclNewLiteralStringObj(keyPtr, "-errorinfo");
+		Tcl_IncrRefCount(keyPtr);
+		Tcl_DictObjGet(NULL, options, keyPtr, &valuePtr);
+		Tcl_DecrRefCount(keyPtr);
+
+		Tcl_WriteChars(errChannel,
+			"error in background error handler:\n", -1);
+		if (valuePtr) {
+		    Tcl_WriteObj(errChannel, valuePtr);
+		} else {
+		    Tcl_WriteObj(errChannel, Tcl_GetObjResult(interp));
+		}
+		Tcl_WriteChars(errChannel, "\n", 1);
+		Tcl_Flush(errChannel);
+		Tcl_DecrRefCount(options);
+	    }
+	}
+    }
+    assocPtr->lastBgPtr = NULL;
+    Tcl_Release(interp);
+    Tcl_Release(assocPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDefaultBgErrorHandlerObjCmd --
+ *
+ *	This function is invoked to process the "::tcl::Bgerror" Tcl command.
+ *	It is the default handler command registered with [interp bgerror] for
+ *	the sake of compatibility with older Tcl releases.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Depends on what actions the "bgerror" command takes for the errors.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclDefaultBgErrorHandlerObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *keyPtr, *valuePtr;
+    Tcl_Obj *tempObjv[2];
+    int result, code, level;
+    Tcl_InterpState saved;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "msg options");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check for a valid return options dictionary.
+     */
+
+    TclNewLiteralStringObj(keyPtr, "-level");
+    Tcl_IncrRefCount(keyPtr);
+    result = Tcl_DictObjGet(NULL, objv[2], keyPtr, &valuePtr);
+    Tcl_DecrRefCount(keyPtr);
+    if (result != TCL_OK || valuePtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"missing return option \"-level\"", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIntFromObj(interp, valuePtr, &level) == TCL_ERROR) {
+	return TCL_ERROR;
+    }
+    TclNewLiteralStringObj(keyPtr, "-code");
+    Tcl_IncrRefCount(keyPtr);
+    result = Tcl_DictObjGet(NULL, objv[2], keyPtr, &valuePtr);
+    Tcl_DecrRefCount(keyPtr);
+    if (result != TCL_OK || valuePtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"missing return option \"-code\"", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIntFromObj(interp, valuePtr, &code) == TCL_ERROR) {
+	return TCL_ERROR;
+    }
+
+    if (level != 0) {
+	/*
+	 * We're handling a TCL_RETURN exception.
+	 */
+
+	code = TCL_RETURN;
+    }
+    if (code == TCL_OK) {
+	/*
+	 * Somehow we got to exception handling with no exception. (Pass
+	 * TCL_OK to Tcl_BackgroundException()?) Just return without doing
+	 * anything.
+	 */
+
+	return TCL_OK;
+    }
+
+    /*
+     * Construct the bgerror command.
+     */
+
+    TclNewLiteralStringObj(tempObjv[0], "bgerror");
+    Tcl_IncrRefCount(tempObjv[0]);
+
+    /*
+     * Determine error message argument.  Check the return options in case
+     * a non-error exception brought us here.
+     */
+
+    switch (code) {
+    case TCL_ERROR:
+	tempObjv[1] = objv[1];
+	break;
+    case TCL_BREAK:
+	TclNewLiteralStringObj(tempObjv[1],
+		"invoked \"break\" outside of a loop");
+	break;
+    case TCL_CONTINUE:
+	TclNewLiteralStringObj(tempObjv[1],
+		"invoked \"continue\" outside of a loop");
+	break;
+    default:
+	tempObjv[1] = Tcl_ObjPrintf("command returned bad code: %d", code);
+	break;
+    }
+    Tcl_IncrRefCount(tempObjv[1]);
+
+    if (code != TCL_ERROR) {
+	Tcl_SetObjResult(interp, tempObjv[1]);
+    }
+
+    TclNewLiteralStringObj(keyPtr, "-errorcode");
+    Tcl_IncrRefCount(keyPtr);
+    result = Tcl_DictObjGet(NULL, objv[2], keyPtr, &valuePtr);
+    Tcl_DecrRefCount(keyPtr);
+    if (result == TCL_OK && valuePtr != NULL) {
+	Tcl_SetObjErrorCode(interp, valuePtr);
+    }
+
+    TclNewLiteralStringObj(keyPtr, "-errorinfo");
+    Tcl_IncrRefCount(keyPtr);
+    result = Tcl_DictObjGet(NULL, objv[2], keyPtr, &valuePtr);
+    Tcl_DecrRefCount(keyPtr);
+    if (result == TCL_OK && valuePtr != NULL) {
+	Tcl_AppendObjToErrorInfo(interp, valuePtr);
+    }
+
+    if (code == TCL_ERROR) {
+	Tcl_SetObjResult(interp, tempObjv[1]);
+    }
+
+    /*
+     * Save interpreter state so we can restore it if multiple handler
+     * attempts are needed.
+     */
+
+    saved = Tcl_SaveInterpState(interp, code);
+
+    /*
+     * Invoke the bgerror command.
+     */
+
+    Tcl_AllowExceptions(interp);
+    code = Tcl_EvalObjv(interp, 2, tempObjv, TCL_EVAL_GLOBAL);
+    if (code == TCL_ERROR) {
+	/*
+	 * If the interpreter is safe, we look for a hidden command named
+	 * "bgerror" and call that with the error information. Otherwise,
+	 * simply ignore the error. The rationale is that this could be an
+	 * error caused by a malicious applet trying to cause an infinite
+	 * barrage of error messages. The hidden "bgerror" command can be used
+	 * by a security policy to interpose on such attacks and e.g. kill the
+	 * applet after a few attempts.
+	 */
+
+	if (Tcl_IsSafe(interp)) {
+	    Tcl_RestoreInterpState(interp, saved);
+	    TclObjInvoke(interp, 2, tempObjv, TCL_INVOKE_HIDDEN);
+	} else {
+	    Tcl_Channel errChannel = Tcl_GetStdChannel(TCL_STDERR);
+
+	    if (errChannel != NULL) {
+		Tcl_Obj *resultPtr = Tcl_GetObjResult(interp);
+
+		Tcl_IncrRefCount(resultPtr);
+		if (Tcl_FindCommand(interp, "bgerror", NULL,
+			TCL_GLOBAL_ONLY) == NULL) {
+		    Tcl_RestoreInterpState(interp, saved);
+		    Tcl_WriteObj(errChannel, Tcl_GetVar2Ex(interp,
+			    "errorInfo", NULL, TCL_GLOBAL_ONLY));
+		    Tcl_WriteChars(errChannel, "\n", -1);
+		} else {
+		    Tcl_DiscardInterpState(saved);
+		    Tcl_WriteChars(errChannel,
+			    "bgerror failed to handle background error.\n",-1);
+		    Tcl_WriteChars(errChannel, "    Original error: ", -1);
+		    Tcl_WriteObj(errChannel, tempObjv[1]);
+		    Tcl_WriteChars(errChannel, "\n", -1);
+		    Tcl_WriteChars(errChannel, "    Error in bgerror: ", -1);
+		    Tcl_WriteObj(errChannel, resultPtr);
+		    Tcl_WriteChars(errChannel, "\n", -1);
+		}
+		Tcl_DecrRefCount(resultPtr);
+		Tcl_Flush(errChannel);
+	    } else {
+		Tcl_DiscardInterpState(saved);
+	    }
+	}
+	code = TCL_OK;
+    } else {
+	Tcl_DiscardInterpState(saved);
+    }
+
+    Tcl_DecrRefCount(tempObjv[0]);
+    Tcl_DecrRefCount(tempObjv[1]);
+    Tcl_ResetResult(interp);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetBgErrorHandler --
+ *
+ *	This function sets the command prefix to be used to handle background
+ *	errors in interp.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Error handler is registered.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetBgErrorHandler(
+    Tcl_Interp *interp,
+    Tcl_Obj *cmdPrefix)
+{
+    ErrAssocData *assocPtr = Tcl_GetAssocData(interp, "tclBgError", NULL);
+
+    if (cmdPrefix == NULL) {
+	Tcl_Panic("TclSetBgErrorHandler: NULL cmdPrefix argument");
+    }
+    if (assocPtr == NULL) {
+	/*
+	 * First access: initialize.
+	 */
+
+	assocPtr = ckalloc(sizeof(ErrAssocData));
+	assocPtr->interp = interp;
+	assocPtr->cmdPrefix = NULL;
+	assocPtr->firstBgPtr = NULL;
+	assocPtr->lastBgPtr = NULL;
+	Tcl_SetAssocData(interp, "tclBgError", BgErrorDeleteProc, assocPtr);
+    }
+    if (assocPtr->cmdPrefix) {
+	Tcl_DecrRefCount(assocPtr->cmdPrefix);
+    }
+    assocPtr->cmdPrefix = cmdPrefix;
+    Tcl_IncrRefCount(assocPtr->cmdPrefix);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetBgErrorHandler --
+ *
+ *	This function retrieves the command prefix currently used to handle
+ *	background errors in interp.
+ *
+ * Results:
+ *	A (Tcl_Obj *) to a list of words (command prefix).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetBgErrorHandler(
+    Tcl_Interp *interp)
+{
+    ErrAssocData *assocPtr = Tcl_GetAssocData(interp, "tclBgError", NULL);
+
+    if (assocPtr == NULL) {
+	Tcl_Obj *bgerrorObj;
+
+	TclNewLiteralStringObj(bgerrorObj, "::tcl::Bgerror");
+	TclSetBgErrorHandler(interp, bgerrorObj);
+	assocPtr = Tcl_GetAssocData(interp, "tclBgError", NULL);
+    }
+    return assocPtr->cmdPrefix;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BgErrorDeleteProc --
+ *
+ *	This function is associated with the "tclBgError" assoc data for an
+ *	interpreter; it is invoked when the interpreter is deleted in order to
+ *	free the information assoicated with any pending error reports.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Background error information is freed: if there were any pending error
+ *	reports, they are canceled.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+BgErrorDeleteProc(
+    ClientData clientData,	/* Pointer to ErrAssocData structure. */
+    Tcl_Interp *interp)		/* Interpreter being deleted. */
+{
+    ErrAssocData *assocPtr = clientData;
+    BgError *errPtr;
+
+    while (assocPtr->firstBgPtr != NULL) {
+	errPtr = assocPtr->firstBgPtr;
+	assocPtr->firstBgPtr = errPtr->nextPtr;
+	Tcl_DecrRefCount(errPtr->errorMsg);
+	Tcl_DecrRefCount(errPtr->returnOpts);
+	ckfree(errPtr);
+    }
+    Tcl_CancelIdleCall(HandleBgErrors, assocPtr);
+    Tcl_DecrRefCount(assocPtr->cmdPrefix);
+    Tcl_EventuallyFree(assocPtr, TCL_DYNAMIC);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateExitHandler --
+ *
+ *	Arrange for a given function to be invoked just before the application
+ *	exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Proc will be invoked with clientData as argument when the application
+ *	exits.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateExitHandler(
+    Tcl_ExitProc *proc,		/* Function to invoke. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr = ckalloc(sizeof(ExitHandler));
+
+    exitPtr->proc = proc;
+    exitPtr->clientData = clientData;
+    Tcl_MutexLock(&exitMutex);
+    exitPtr->nextPtr = firstExitPtr;
+    firstExitPtr = exitPtr;
+    Tcl_MutexUnlock(&exitMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateLateExitHandler --
+ *
+ *	Arrange for a given function to be invoked after all pre-thread
+ *	cleanups.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Proc will be invoked with clientData as argument when the application
+ *	exits.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclCreateLateExitHandler(
+    Tcl_ExitProc *proc,		/* Function to invoke. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr = ckalloc(sizeof(ExitHandler));
+
+    exitPtr->proc = proc;
+    exitPtr->clientData = clientData;
+    Tcl_MutexLock(&exitMutex);
+    exitPtr->nextPtr = firstLateExitPtr;
+    firstLateExitPtr = exitPtr;
+    Tcl_MutexUnlock(&exitMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteExitHandler --
+ *
+ *	This function cancels an existing exit handler matching proc and
+ *	clientData, if such a handler exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there is an exit handler corresponding to proc and clientData then
+ *	it is canceled; if no such handler exists then nothing happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteExitHandler(
+    Tcl_ExitProc *proc,		/* Function that was previously registered. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr, *prevPtr;
+
+    Tcl_MutexLock(&exitMutex);
+    for (prevPtr = NULL, exitPtr = firstExitPtr; exitPtr != NULL;
+	    prevPtr = exitPtr, exitPtr = exitPtr->nextPtr) {
+	if ((exitPtr->proc == proc)
+		&& (exitPtr->clientData == clientData)) {
+	    if (prevPtr == NULL) {
+		firstExitPtr = exitPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = exitPtr->nextPtr;
+	    }
+	    ckfree(exitPtr);
+	    break;
+	}
+    }
+    Tcl_MutexUnlock(&exitMutex);
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteLateExitHandler --
+ *
+ *	This function cancels an existing late exit handler matching proc and
+ *	clientData, if such a handler exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there is a late exit handler corresponding to proc and clientData
+ *	then it is canceled; if no such handler exists then nothing happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDeleteLateExitHandler(
+    Tcl_ExitProc *proc,		/* Function that was previously registered. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr, *prevPtr;
+
+    Tcl_MutexLock(&exitMutex);
+    for (prevPtr = NULL, exitPtr = firstLateExitPtr; exitPtr != NULL;
+	    prevPtr = exitPtr, exitPtr = exitPtr->nextPtr) {
+	if ((exitPtr->proc == proc)
+		&& (exitPtr->clientData == clientData)) {
+	    if (prevPtr == NULL) {
+		firstLateExitPtr = exitPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = exitPtr->nextPtr;
+	    }
+	    ckfree(exitPtr);
+	    break;
+	}
+    }
+    Tcl_MutexUnlock(&exitMutex);
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateThreadExitHandler --
+ *
+ *	Arrange for a given function to be invoked just before the current
+ *	thread exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Proc will be invoked with clientData as argument when the application
+ *	exits.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateThreadExitHandler(
+    Tcl_ExitProc *proc,		/* Function to invoke. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    exitPtr = ckalloc(sizeof(ExitHandler));
+    exitPtr->proc = proc;
+    exitPtr->clientData = clientData;
+    exitPtr->nextPtr = tsdPtr->firstExitPtr;
+    tsdPtr->firstExitPtr = exitPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteThreadExitHandler --
+ *
+ *	This function cancels an existing exit handler matching proc and
+ *	clientData, if such a handler exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there is an exit handler corresponding to proc and clientData then
+ *	it is canceled; if no such handler exists then nothing happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteThreadExitHandler(
+    Tcl_ExitProc *proc,		/* Function that was previously registered. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    ExitHandler *exitPtr, *prevPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    for (prevPtr = NULL, exitPtr = tsdPtr->firstExitPtr; exitPtr != NULL;
+	    prevPtr = exitPtr, exitPtr = exitPtr->nextPtr) {
+	if ((exitPtr->proc == proc)
+		&& (exitPtr->clientData == clientData)) {
+	    if (prevPtr == NULL) {
+		tsdPtr->firstExitPtr = exitPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = exitPtr->nextPtr;
+	    }
+	    ckfree(exitPtr);
+	    return;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetExitProc --
+ *
+ *	This function sets the application wide exit handler that will be
+ *	called by Tcl_Exit in place of the C-runtime exit. If the application
+ *	wide exit handler is NULL, the C-runtime exit will be used instead.
+ *
+ * Results:
+ *	The previously set application wide exit handler.
+ *
+ * Side effects:
+ *	Sets the application wide exit handler to the specified value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_ExitProc *
+Tcl_SetExitProc(
+    TCL_NORETURN1 Tcl_ExitProc *proc)		/* New exit handler for app or NULL */
+{
+    Tcl_ExitProc *prevExitProc;
+
+    /*
+     * Swap the old exit proc for the new one, saving the old one for our
+     * return value.
+     */
+
+    Tcl_MutexLock(&exitMutex);
+    prevExitProc = appExitPtr;
+    appExitPtr = proc;
+    Tcl_MutexUnlock(&exitMutex);
+
+    return prevExitProc;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InvokeExitHandlers --
+ *
+ *      Call the registered exit handlers.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The exit handlers are invoked, and the ExitHandler struct is
+ *      freed.
+ *
+ *----------------------------------------------------------------------
+ */
+static void
+InvokeExitHandlers(void)
+{
+    ExitHandler *exitPtr;
+
+    Tcl_MutexLock(&exitMutex);
+    inExit = 1;
+
+    for (exitPtr = firstExitPtr; exitPtr != NULL; exitPtr = firstExitPtr) {
+	/*
+	 * Be careful to remove the handler from the list before invoking its
+	 * callback. This protects us against double-freeing if the callback
+	 * should call Tcl_DeleteExitHandler on itself.
+	 */
+
+	firstExitPtr = exitPtr->nextPtr;
+	Tcl_MutexUnlock(&exitMutex);
+	exitPtr->proc(exitPtr->clientData);
+	ckfree(exitPtr);
+	Tcl_MutexLock(&exitMutex);
+    }
+    firstExitPtr = NULL;
+    Tcl_MutexUnlock(&exitMutex);
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Exit --
+ *
+ *	This function is called to terminate the application.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	All existing exit handlers are invoked, then the application ends.
+ *
+ *----------------------------------------------------------------------
+ */
+
+TCL_NORETURN void
+Tcl_Exit(
+    int status)			/* Exit status for application; typically 0
+				 * for normal return, 1 for error return. */
+{
+    TCL_NORETURN1 Tcl_ExitProc *currentAppExitPtr;
+
+    Tcl_MutexLock(&exitMutex);
+    currentAppExitPtr = appExitPtr;
+    Tcl_MutexUnlock(&exitMutex);
+
+    if (currentAppExitPtr) {
+	/*
+	 * Warning: this code SHOULD NOT return, as there is code that depends
+	 * on Tcl_Exit never returning. In fact, we will Tcl_Panic if anyone
+	 * returns, so critical is this dependcy.
+	 */
+
+	currentAppExitPtr(INT2PTR(status));
+	Tcl_Panic("AppExitProc returned unexpectedly");
+    } else {
+
+	if (TclFullFinalizationRequested()) {
+
+	    /*
+	     * Thorough finalization for Valgrind et al.
+	     */
+
+	    Tcl_Finalize();
+
+	} else {
+
+	    /*
+	     * Fast and deterministic exit (default behavior)
+	     */
+
+	    InvokeExitHandlers();
+
+	    /*
+	     * Ensure the thread-specific data is initialised as it is used in
+	     * Tcl_FinalizeThread()
+	     */
+
+	    (void) TCL_TSD_INIT(&dataKey);
+
+	    /*
+	     * Now finalize the calling thread only (others are not safely
+	     * reachable).  Among other things, this triggers a flush of the
+	     * Tcl_Channels that may have data enqueued.
+	     */
+
+	    FinalizeThread(/* quick */ 1);
+	}
+	TclpExit(status);
+	Tcl_Panic("OS exit failed!");
+    }
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclInitSubsystems --
+ *
+ *	Initialize various subsytems in Tcl. This should be called the first
+ *	time an interp is created, or before any of the subsystems are used.
+ *	This function ensures an order for the initialization of subsystems:
+ *
+ *	1. that cannot be initialized in lazy order because they are mutually
+ *	dependent.
+ *
+ *	2. so that they can be finalized in a known order w/o causing the
+ *	subsequent re-initialization of a subsystem in the act of shutting
+ *	down another.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Varied, see the respective initialization routines.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+TclInitSubsystems(void)
+{
+    if (inExit != 0) {
+	Tcl_Panic("TclInitSubsystems called while exiting");
+    }
+
+    if (subsystemsInitialized == 0) {
+	/*
+	 * Double check inside the mutex. There are definitly calls back into
+	 * this routine from some of the functions below.
+	 */
+
+	TclpInitLock();
+	if (subsystemsInitialized == 0) {
+
+		/*
+	     * Initialize locks used by the memory allocators before anything
+	     * interesting happens so we can use the allocators in the
+	     * implementation of self-initializing locks.
+	     */
+
+	    TclInitThreadStorage();     /* Creates master hash table for
+					 * thread local storage */
+#if USE_TCLALLOC
+	    TclInitAlloc();		/* Process wide mutex init */
+#endif
+#if defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+	    TclInitThreadAlloc();	/* Setup thread allocator caches */
+#endif
+#ifdef TCL_MEM_DEBUG
+	    TclInitDbCkalloc();		/* Process wide mutex init */
+#endif
+
+	    TclpInitPlatform();		/* Creates signal handler(s) */
+	    TclInitDoubleConversion();	/* Initializes constants for
+					 * converting to/from double. */
+	    TclInitObjSubsystem();	/* Register obj types, create
+					 * mutexes. */
+	    TclInitIOSubsystem();	/* Inits a tsd key (noop). */
+	    TclInitEncodingSubsystem();	/* Process wide encoding init. */
+	    TclpSetInterfaces();
+	    TclInitNamespaceSubsystem();/* Register ns obj type (mutexed). */
+	    subsystemsInitialized = 1;
+	}
+	TclpInitUnlock();
+    }
+    TclInitNotifier();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Finalize --
+ *
+ *	Shut down Tcl. First calls registered exit handlers, then carefully
+ *	shuts down various subsystems.  Should be invoked by user before the
+ *	Tcl shared library is being unloaded in an embedded context.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Varied, see the respective finalization routines.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_Finalize(void)
+{
+    ExitHandler *exitPtr;
+
+    /*
+     * Invoke exit handlers first.
+     */
+
+    InvokeExitHandlers();
+
+    TclpInitLock();
+    if (subsystemsInitialized == 0) {
+	goto alreadyFinalized;
+    }
+    subsystemsInitialized = 0;
+
+    /*
+     * Ensure the thread-specific data is initialised as it is used in
+     * Tcl_FinalizeThread()
+     */
+
+    (void) TCL_TSD_INIT(&dataKey);
+
+    /*
+     * Clean up after the current thread now, after exit handlers. In
+     * particular, the testexithandler command sets up something that writes
+     * to standard output, which gets closed. Note that there is no
+     * thread-local storage or IO subsystem after this call.
+     */
+
+    Tcl_FinalizeThread();
+
+    /*
+     * Now invoke late (process-wide) exit handlers.
+     */
+
+    Tcl_MutexLock(&exitMutex);
+    for (exitPtr = firstLateExitPtr; exitPtr != NULL;
+	    exitPtr = firstLateExitPtr) {
+	/*
+	 * Be careful to remove the handler from the list before invoking its
+	 * callback. This protects us against double-freeing if the callback
+	 * should call Tcl_DeleteLateExitHandler on itself.
+	 */
+
+	firstLateExitPtr = exitPtr->nextPtr;
+	Tcl_MutexUnlock(&exitMutex);
+	exitPtr->proc(exitPtr->clientData);
+	ckfree(exitPtr);
+	Tcl_MutexLock(&exitMutex);
+    }
+    firstLateExitPtr = NULL;
+    Tcl_MutexUnlock(&exitMutex);
+
+    /*
+     * Now finalize the Tcl execution environment. Note that this must be done
+     * after the exit handlers, because there are order dependencies.
+     */
+
+    TclFinalizeEvaluation();
+    TclFinalizeExecution();
+    TclFinalizeEnvironment();
+
+    /*
+     * Finalizing the filesystem must come after anything which might
+     * conceivably interact with the 'Tcl_FS' API.
+     */
+
+    TclFinalizeFilesystem();
+
+    /*
+     * Undo all Tcl_ObjType registrations, and reset the master list of free
+     * Tcl_Obj's. After this returns, no more Tcl_Obj's should be allocated or
+     * freed.
+     *
+     * Note in particular that TclFinalizeObjects() must follow
+     * TclFinalizeFilesystem() because TclFinalizeFilesystem free's the
+     * Tcl_Obj that holds the path of the current working directory.
+     */
+
+    TclFinalizeObjects();
+
+    /*
+     * We must be sure the encoding finalization doesn't need to examine the
+     * filesystem in any way. Since it only needs to clean up internal data
+     * structures, this is fine.
+     */
+
+    TclFinalizeEncodingSubsystem();
+
+    /*
+     * Repeat finalization of the thread local storage once more. Although
+     * this step is already done by the Tcl_FinalizeThread call above, series
+     * of events happening afterwards may re-initialize TSD slots. Those need
+     * to be finalized again, otherwise we're leaking memory chunks. Very
+     * important to note is that things happening afterwards should not
+     * reference anything which may re-initialize TSD's. This includes freeing
+     * Tcl_Objs's, among other things.
+     *
+     * This fixes the Tcl Bug #990552.
+     */
+
+    TclFinalizeThreadData(/* quick */ 0);
+
+    /*
+     * Now we can free constants for conversions to/from double.
+     */
+
+    TclFinalizeDoubleConversion();
+
+    /*
+     * There have been several bugs in the past that cause exit handlers to be
+     * established during Tcl_Finalize processing. Such exit handlers leave
+     * malloc'ed memory, and Tcl_FinalizeMemorySubsystem or
+     * Tcl_FinalizeThreadAlloc will result in a corrupted heap. The result can
+     * be a mysterious crash on process exit. Check here that nobody's done
+     * this.
+     */
+
+    if (firstExitPtr != NULL) {
+	Tcl_Panic("exit handlers were created during Tcl_Finalize");
+    }
+
+    TclFinalizePreserve();
+
+    /*
+     * Free synchronization objects. There really should only be one thread
+     * alive at this moment.
+     */
+
+    TclFinalizeSynchronization();
+
+    /*
+     * Close down the thread-specific object allocator.
+     */
+
+#if defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+    TclFinalizeThreadAlloc();
+#endif
+
+    /*
+     * We defer unloading of packages until very late to avoid memory access
+     * issues. Both exit callbacks and synchronization variables may be stored
+     * in packages.
+     *
+     * Note that TclFinalizeLoad unloads packages in the reverse of the order
+     * they were loaded in (i.e. last to be loaded is the first to be
+     * unloaded). This can be important for correct unloading when
+     * dependencies exist.
+     *
+     * Once load has been finalized, we will have deleted any temporary copies
+     * of shared libraries and can therefore reset the filesystem to its
+     * original state.
+     */
+
+    TclFinalizeLoad();
+    TclResetFilesystem();
+
+    /*
+     * At this point, there should no longer be any ckalloc'ed memory.
+     */
+
+    TclFinalizeMemorySubsystem();
+
+  alreadyFinalized:
+    TclFinalizeLock();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FinalizeThread --
+ *
+ *	Runs the exit handlers to allow Tcl to clean up its state about a
+ *	particular thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Varied, see the respective finalization routines.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_FinalizeThread(void)
+{
+    FinalizeThread(/* quick */ 0);
+}
+
+void
+FinalizeThread(
+    int quick)
+{
+    ExitHandler *exitPtr;
+    ThreadSpecificData *tsdPtr;
+
+    /*
+     * We use TclThreadDataKeyGet here, rather than Tcl_GetThreadData, because
+     * we don't want to initialize the data block if it hasn't been
+     * initialized already.
+     */
+
+    tsdPtr = TclThreadDataKeyGet(&dataKey);
+    if (tsdPtr != NULL) {
+	tsdPtr->inExit = 1;
+
+	for (exitPtr = tsdPtr->firstExitPtr; exitPtr != NULL;
+		exitPtr = tsdPtr->firstExitPtr) {
+	    /*
+	     * Be careful to remove the handler from the list before invoking
+	     * its callback. This protects us against double-freeing if the
+	     * callback should call Tcl_DeleteThreadExitHandler on itself.
+	     */
+
+	    tsdPtr->firstExitPtr = exitPtr->nextPtr;
+	    exitPtr->proc(exitPtr->clientData);
+	    ckfree(exitPtr);
+	}
+	TclFinalizeIOSubsystem();
+	TclFinalizeNotifier();
+	TclFinalizeAsync();
+	TclFinalizeThreadObjects();
+    }
+
+    /*
+     * Blow away all thread local storage blocks.
+     *
+     * Note that Tcl API allows creation of threads which do not use any Tcl
+     * interp or other Tcl subsytems. Those threads might, however, use thread
+     * local storage, so we must unconditionally finalize it.
+     *
+     * Fix [Bug #571002]
+     */
+    TclFinalizeThreadData(quick);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInExit --
+ *
+ *	Determines if we are in the middle of exit-time cleanup.
+ *
+ * Results:
+ *	If we are in the middle of exiting, 1, otherwise 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInExit(void)
+{
+    return inExit;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInThreadExit --
+ *
+ *	Determines if we are in the middle of thread exit-time cleanup.
+ *
+ * Results:
+ *	If we are in the middle of exiting this thread, 1, otherwise 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInThreadExit(void)
+{
+    ThreadSpecificData *tsdPtr = TclThreadDataKeyGet(&dataKey);
+
+    if (tsdPtr == NULL) {
+	return 0;
+    }
+    return tsdPtr->inExit;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_VwaitObjCmd --
+ *
+ *	This function is invoked to process the "vwait" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_VwaitObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int done, foundEvent;
+    const char *nameString;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+    nameString = Tcl_GetString(objv[1]);
+    if (Tcl_TraceVar2(interp, nameString, NULL,
+	    TCL_GLOBAL_ONLY|TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    VwaitVarProc, &done) != TCL_OK) {
+	return TCL_ERROR;
+    };
+    done = 0;
+    foundEvent = 1;
+    while (!done && foundEvent) {
+	foundEvent = Tcl_DoOneEvent(TCL_ALL_EVENTS);
+	if (Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG) == TCL_ERROR) {
+	    break;
+	}
+	if (Tcl_LimitExceeded(interp)) {
+	    Tcl_ResetResult(interp);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("limit exceeded", -1));
+	    break;
+	}
+    }
+    Tcl_UntraceVar2(interp, nameString, NULL,
+	    TCL_GLOBAL_ONLY|TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    VwaitVarProc, &done);
+
+    if (!foundEvent) {
+	Tcl_ResetResult(interp);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't wait for variable \"%s\": would wait forever",
+		nameString));
+	Tcl_SetErrorCode(interp, "TCL", "EVENT", "NO_SOURCES", NULL);
+	return TCL_ERROR;
+    }
+    if (!done) {
+	/*
+	 * The interpreter's result was already set to the right error message
+	 * prior to exiting the loop above.
+	 */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Clear out the interpreter's result, since it may have been set by event
+     * handlers.
+     */
+
+    Tcl_ResetResult(interp);
+    return TCL_OK;
+}
+
+	/* ARGSUSED */
+static char *
+VwaitVarProc(
+    ClientData clientData,	/* Pointer to integer to set to 1. */
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *name1,		/* Name of variable. */
+    const char *name2,		/* Second part of variable name. */
+    int flags)			/* Information about what happened. */
+{
+    int *donePtr = clientData;
+
+    *donePtr = 1;
+    Tcl_UntraceVar(interp, name1, TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    VwaitVarProc, clientData);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UpdateObjCmd --
+ *
+ *	This function is invoked to process the "update" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_UpdateObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int optionIndex;
+    int flags = 0;		/* Initialized to avoid compiler warning. */
+    static const char *const updateOptions[] = {"idletasks", NULL};
+    enum updateOptions {OPT_IDLETASKS};
+
+    if (objc == 1) {
+	flags = TCL_ALL_EVENTS|TCL_DONT_WAIT;
+    } else if (objc == 2) {
+	if (Tcl_GetIndexFromObj(interp, objv[1], updateOptions,
+		"option", 0, &optionIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum updateOptions) optionIndex) {
+	case OPT_IDLETASKS:
+	    flags = TCL_WINDOW_EVENTS|TCL_IDLE_EVENTS|TCL_DONT_WAIT;
+	    break;
+	default:
+	    Tcl_Panic("Tcl_UpdateObjCmd: bad option index to UpdateOptions");
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?idletasks?");
+	return TCL_ERROR;
+    }
+
+    while (Tcl_DoOneEvent(flags) != 0) {
+	if (Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_LimitExceeded(interp)) {
+	    Tcl_ResetResult(interp);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("limit exceeded", -1));
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Must clear the interpreter's result because event handlers could have
+     * executed commands.
+     */
+
+    Tcl_ResetResult(interp);
+    return TCL_OK;
+}
+
+#ifdef TCL_THREADS
+/*
+ *----------------------------------------------------------------------
+ *
+ * NewThreadProc --
+ *
+ *	Bootstrap function of a new Tcl thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side Effects:
+ *	Initializes Tcl notifier for the current thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_ThreadCreateType
+NewThreadProc(
+    ClientData clientData)
+{
+    ThreadClientData *cdPtr = clientData;
+    ClientData threadClientData;
+    Tcl_ThreadCreateProc *threadProc;
+
+    threadProc = cdPtr->proc;
+    threadClientData = cdPtr->clientData;
+    ckfree(clientData);		/* Allocated in Tcl_CreateThread() */
+
+    threadProc(threadClientData);
+
+    TCL_THREAD_CREATE_RETURN;
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateThread --
+ *
+ *	This function creates a new thread. This actually belongs to the
+ *	tclThread.c file but since we use some private data structures local
+ *	to this file, it is placed here.
+ *
+ * Results:
+ *	TCL_OK if the thread could be created. The thread ID is returned in a
+ *	parameter.
+ *
+ * Side effects:
+ *	A new thread is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_CreateThread(
+    Tcl_ThreadId *idPtr,	/* Return, the ID of the thread */
+    Tcl_ThreadCreateProc *proc,	/* Main() function of the thread */
+    ClientData clientData,	/* The one argument to Main() */
+    int stackSize,		/* Size of stack for the new thread */
+    int flags)			/* Flags controlling behaviour of the new
+				 * thread. */
+{
+#ifdef TCL_THREADS
+    ThreadClientData *cdPtr = ckalloc(sizeof(ThreadClientData));
+    int result;
+
+    cdPtr->proc = proc;
+    cdPtr->clientData = clientData;
+    result = TclpThreadCreate(idPtr, NewThreadProc, cdPtr, stackSize, flags);
+    if (result != TCL_OK) {
+	ckfree(cdPtr);
+    }
+    return result;
+#else
+    return TCL_ERROR;
+#endif /* TCL_THREADS */
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclExecute.c b/generic/tclExecute.c
new file mode 100644
index 0000000..0113b28
--- /dev/null
+++ b/generic/tclExecute.c
@@ -0,0 +1,10546 @@
+/*
+ * tclExecute.c --
+ *
+ *	This file contains procedures that execute byte-compiled Tcl commands.
+ *
+ * Copyright (c) 1996-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 by Scriptics Corporation.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ * Copyright (c) 2002-2010 by Miguel Sofer.
+ * Copyright (c) 2005-2007 by Donal K. Fellows.
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ * Copyright (c) 2006-2008 by Joe Mistachkin.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include "tclOOInt.h"
+#include "tommath.h"
+#include <math.h>
+#include <assert.h>
+
+/*
+ * Hack to determine whether we may expect IEEE floating point. The hack is
+ * formally incorrect in that non-IEEE platforms might have the same precision
+ * and range, but VAX, IBM, and Cray do not; are there any other floating
+ * point units that we might care about?
+ */
+
+#if (FLT_RADIX == 2) && (DBL_MANT_DIG == 53) && (DBL_MAX_EXP == 1024)
+#define IEEE_FLOATING_POINT
+#endif
+
+/*
+ * A counter that is used to work out when the bytecode engine should call
+ * Tcl_AsyncReady() to see whether there is a signal that needs handling, and
+ * other expensive periodic operations.
+ */
+
+#ifndef ASYNC_CHECK_COUNT
+#   define ASYNC_CHECK_COUNT	64
+#endif /* !ASYNC_CHECK_COUNT */
+
+/*
+ * Boolean flag indicating whether the Tcl bytecode interpreter has been
+ * initialized.
+ */
+
+static int execInitialized = 0;
+TCL_DECLARE_MUTEX(execMutex)
+
+static int cachedInExit = 0;
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ * Variable that controls whether execution tracing is enabled and, if so,
+ * what level of tracing is desired:
+ *    0: no execution tracing
+ *    1: trace invocations of Tcl procs only
+ *    2: trace invocations of all (not compiled away) commands
+ *    3: display each instruction executed
+ * This variable is linked to the Tcl variable "tcl_traceExec".
+ */
+
+int tclTraceExec = 0;
+#endif
+
+/*
+ * Mapping from expression instruction opcodes to strings; used for error
+ * messages. Note that these entries must match the order and number of the
+ * expression opcodes (e.g., INST_LOR) in tclCompile.h.
+ *
+ * Does not include the string for INST_EXPON (and beyond), as that is
+ * disjoint for backward-compatability reasons.
+ */
+
+static const char *const operatorStrings[] = {
+    "||", "&&", "|", "^", "&", "==", "!=", "<", ">", "<=", ">=", "<<", ">>",
+    "+", "-", "*", "/", "%", "+", "-", "~", "!"
+};
+
+/*
+ * Mapping from Tcl result codes to strings; used for error and debugging
+ * messages.
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+static const char *const resultStrings[] = {
+    "TCL_OK", "TCL_ERROR", "TCL_RETURN", "TCL_BREAK", "TCL_CONTINUE"
+};
+#endif
+
+/*
+ * These are used by evalstats to monitor object usage in Tcl.
+ */
+
+#ifdef TCL_COMPILE_STATS
+long		tclObjsAlloced = 0;
+long		tclObjsFreed = 0;
+long		tclObjsShared[TCL_MAX_SHARED_OBJ_STATS] = { 0, 0, 0, 0, 0 };
+#endif /* TCL_COMPILE_STATS */
+
+/*
+ * Support pre-8.5 bytecodes unless specifically requested otherwise.
+ */
+
+#ifndef TCL_SUPPORT_84_BYTECODE
+#define TCL_SUPPORT_84_BYTECODE 1
+#endif
+
+#if TCL_SUPPORT_84_BYTECODE
+/*
+ * We need to know the tclBuiltinFuncTable to support translation of pre-8.5
+ * math functions to the namespace-based ::tcl::mathfunc::op in 8.5+.
+ */
+
+typedef struct {
+    const char *name;		/* Name of function. */
+    int numArgs;		/* Number of arguments for function. */
+} BuiltinFunc;
+
+/*
+ * Table describing the built-in math functions. Entries in this table are
+ * indexed by the values of the INST_CALL_BUILTIN_FUNC instruction's
+ * operand byte.
+ */
+
+static BuiltinFunc const tclBuiltinFuncTable[] = {
+    {"acos", 1},
+    {"asin", 1},
+    {"atan", 1},
+    {"atan2", 2},
+    {"ceil", 1},
+    {"cos", 1},
+    {"cosh", 1},
+    {"exp", 1},
+    {"floor", 1},
+    {"fmod", 2},
+    {"hypot", 2},
+    {"log", 1},
+    {"log10", 1},
+    {"pow", 2},
+    {"sin", 1},
+    {"sinh", 1},
+    {"sqrt", 1},
+    {"tan", 1},
+    {"tanh", 1},
+    {"abs", 1},
+    {"double", 1},
+    {"int", 1},
+    {"rand", 0},
+    {"round", 1},
+    {"srand", 1},
+    {"wide", 1},
+    {NULL, 0},
+};
+
+#define LAST_BUILTIN_FUNC	25
+#endif
+
+/*
+ * NR_TEBC
+ * Helpers for NR - non-recursive calls to TEBC
+ * Minimal data required to fully reconstruct the execution state.
+ */
+
+typedef struct TEBCdata {
+    ByteCode *codePtr;		/* Constant until the BC returns */
+				/* -----------------------------------------*/
+    ptrdiff_t *catchTop;	/* These fields are used on return TO this */
+    Tcl_Obj *auxObjList;	/* this level: they record the state when a */
+    CmdFrame cmdFrame;		/* new codePtr was received for NR */
+                                /* execution. */
+    void *stack[1];		/* Start of the actual combined catch and obj
+				 * stacks; the struct will be expanded as
+				 * necessary */
+} TEBCdata;
+
+#define TEBC_YIELD() \
+    do {						\
+	esPtr->tosPtr = tosPtr;				\
+	TclNRAddCallback(interp, TEBCresume,		\
+		TD, pc, INT2PTR(cleanup), NULL);	\
+    } while (0)
+
+#define TEBC_DATA_DIG() \
+    do {					\
+	tosPtr = esPtr->tosPtr;			\
+    } while (0)
+
+#define PUSH_TAUX_OBJ(objPtr) \
+    do {							\
+	if (auxObjList) {					\
+	    objPtr->length += auxObjList->length;		\
+	}							\
+	objPtr->internalRep.twoPtrValue.ptr1 = auxObjList;	\
+	auxObjList = objPtr;					\
+    } while (0)
+
+#define POP_TAUX_OBJ() \
+    do {							\
+	tmpPtr = auxObjList;					\
+	auxObjList = tmpPtr->internalRep.twoPtrValue.ptr1;	\
+	Tcl_DecrRefCount(tmpPtr);				\
+    } while (0)
+
+/*
+ * These variable-access macros have to coincide with those in tclVar.c
+ */
+
+#define VarHashGetValue(hPtr) \
+    ((Var *) ((char *)hPtr - TclOffset(VarInHash, entry)))
+
+static inline Var *
+VarHashCreateVar(
+    TclVarHashTable *tablePtr,
+    Tcl_Obj *key,
+    int *newPtr)
+{
+    Tcl_HashEntry *hPtr = Tcl_CreateHashEntry(&tablePtr->table,
+	    key, newPtr);
+
+    if (!hPtr) {
+	return NULL;
+    }
+    return VarHashGetValue(hPtr);
+}
+
+#define VarHashFindVar(tablePtr, key) \
+    VarHashCreateVar((tablePtr), (key), NULL)
+
+/*
+ * The new macro for ending an instruction; note that a reasonable C-optimiser
+ * will resolve all branches at compile time. (result) is always a constant;
+ * the macro NEXT_INST_F handles constant (nCleanup), NEXT_INST_V is resolved
+ * at runtime for variable (nCleanup).
+ *
+ * ARGUMENTS:
+ *    pcAdjustment: how much to increment pc
+ *    nCleanup: how many objects to remove from the stack
+ *    resultHandling: 0 indicates no object should be pushed on the stack;
+ *	otherwise, push objResultPtr. If (result < 0), objResultPtr already
+ *	has the correct reference count.
+ *
+ * We use the new compile-time assertions to check that nCleanup is constant
+ * and within range.
+ */
+
+/* Verify the stack depth, only when no expansion is in progress */
+
+#ifdef TCL_COMPILE_DEBUG
+#define CHECK_STACK()							\
+    do {								\
+	ValidatePcAndStackTop(codePtr, pc, CURR_DEPTH,			\
+		/*checkStack*/ !(starting || auxObjList));		\
+	starting = 0;							\
+    } while (0)
+#else
+#define CHECK_STACK()
+#endif
+
+#define NEXT_INST_F(pcAdjustment, nCleanup, resultHandling)	\
+    do {							\
+	TCL_CT_ASSERT((nCleanup >= 0) && (nCleanup <= 2));	\
+	CHECK_STACK();						\
+	if (nCleanup == 0) {					\
+	    if (resultHandling != 0) {				\
+		if ((resultHandling) > 0) {			\
+		    PUSH_OBJECT(objResultPtr);			\
+		} else {					\
+		    *(++tosPtr) = objResultPtr;			\
+		}						\
+	    }							\
+	    pc += (pcAdjustment);				\
+	    goto cleanup0;					\
+	} else if (resultHandling != 0) {			\
+	    if ((resultHandling) > 0) {				\
+		Tcl_IncrRefCount(objResultPtr);			\
+	    }							\
+	    pc += (pcAdjustment);				\
+	    switch (nCleanup) {					\
+	    case 1: goto cleanup1_pushObjResultPtr;		\
+	    case 2: goto cleanup2_pushObjResultPtr;		\
+	    case 0: break;					\
+	    }							\
+	} else {						\
+	    pc += (pcAdjustment);				\
+	    switch (nCleanup) {					\
+	    case 1: goto cleanup1;				\
+	    case 2: goto cleanup2;				\
+	    case 0: break;					\
+	    }							\
+	}							\
+    } while (0)
+
+#define NEXT_INST_V(pcAdjustment, nCleanup, resultHandling)	\
+    CHECK_STACK();						\
+    do {							\
+	pc += (pcAdjustment);					\
+	cleanup = (nCleanup);					\
+	if (resultHandling) {					\
+	    if ((resultHandling) > 0) {				\
+		Tcl_IncrRefCount(objResultPtr);			\
+	    }							\
+	    goto cleanupV_pushObjResultPtr;			\
+	} else {						\
+	    goto cleanupV;					\
+	}							\
+    } while (0)
+
+#ifndef TCL_COMPILE_DEBUG
+#define JUMP_PEEPHOLE_F(condition, pcAdjustment, cleanup) \
+    do {								\
+	pc += (pcAdjustment);						\
+	switch (*pc) {							\
+	case INST_JUMP_FALSE1:						\
+	    NEXT_INST_F(((condition)? 2 : TclGetInt1AtPtr(pc+1)), (cleanup), 0); \
+	case INST_JUMP_TRUE1:						\
+	    NEXT_INST_F(((condition)? TclGetInt1AtPtr(pc+1) : 2), (cleanup), 0); \
+	case INST_JUMP_FALSE4:						\
+	    NEXT_INST_F(((condition)? 5 : TclGetInt4AtPtr(pc+1)), (cleanup), 0); \
+	case INST_JUMP_TRUE4:						\
+	    NEXT_INST_F(((condition)? TclGetInt4AtPtr(pc+1) : 5), (cleanup), 0); \
+	default:							\
+	    if ((condition) < 0) {					\
+		TclNewLongObj(objResultPtr, -1);				\
+	    } else {							\
+		objResultPtr = TCONST((condition) > 0);			\
+	    }								\
+	    NEXT_INST_F(0, (cleanup), 1);				\
+	}								\
+    } while (0)
+#define JUMP_PEEPHOLE_V(condition, pcAdjustment, cleanup) \
+    do {								\
+	pc += (pcAdjustment);						\
+	switch (*pc) {							\
+	case INST_JUMP_FALSE1:						\
+	    NEXT_INST_V(((condition)? 2 : TclGetInt1AtPtr(pc+1)), (cleanup), 0); \
+	case INST_JUMP_TRUE1:						\
+	    NEXT_INST_V(((condition)? TclGetInt1AtPtr(pc+1) : 2), (cleanup), 0); \
+	case INST_JUMP_FALSE4:						\
+	    NEXT_INST_V(((condition)? 5 : TclGetInt4AtPtr(pc+1)), (cleanup), 0); \
+	case INST_JUMP_TRUE4:						\
+	    NEXT_INST_V(((condition)? TclGetInt4AtPtr(pc+1) : 5), (cleanup), 0); \
+	default:							\
+	    if ((condition) < 0) {					\
+		TclNewLongObj(objResultPtr, -1);				\
+	    } else {							\
+		objResultPtr = TCONST((condition) > 0);			\
+	    }								\
+	    NEXT_INST_V(0, (cleanup), 1);				\
+	}								\
+    } while (0)
+#else /* TCL_COMPILE_DEBUG */
+#define JUMP_PEEPHOLE_F(condition, pcAdjustment, cleanup) \
+    do{									\
+	if ((condition) < 0) {						\
+	    TclNewLongObj(objResultPtr, -1);				\
+	} else {							\
+	    objResultPtr = TCONST((condition) > 0);			\
+	}								\
+	NEXT_INST_F((pcAdjustment), (cleanup), 1);			\
+    } while (0)
+#define JUMP_PEEPHOLE_V(condition, pcAdjustment, cleanup) \
+    do{									\
+	if ((condition) < 0) {						\
+	    TclNewLongObj(objResultPtr, -1);				\
+	} else {							\
+	    objResultPtr = TCONST((condition) > 0);			\
+	}								\
+	NEXT_INST_V((pcAdjustment), (cleanup), 1);			\
+    } while (0)
+#endif
+
+/*
+ * Macros used to cache often-referenced Tcl evaluation stack information
+ * in local variables. Note that a DECACHE_STACK_INFO()-CACHE_STACK_INFO()
+ * pair must surround any call inside TclNRExecuteByteCode (and a few other
+ * procedures that use this scheme) that could result in a recursive call
+ * to TclNRExecuteByteCode.
+ */
+
+#define CACHE_STACK_INFO() \
+    checkInterp = 1
+
+#define DECACHE_STACK_INFO() \
+    esPtr->tosPtr = tosPtr
+
+/*
+ * Macros used to access items on the Tcl evaluation stack. PUSH_OBJECT
+ * increments the object's ref count since it makes the stack have another
+ * reference pointing to the object. However, POP_OBJECT does not decrement
+ * the ref count. This is because the stack may hold the only reference to the
+ * object, so the object would be destroyed if its ref count were decremented
+ * before the caller had a chance to, e.g., store it in a variable. It is the
+ * caller's responsibility to decrement the ref count when it is finished with
+ * an object.
+ *
+ * WARNING! It is essential that objPtr only appear once in the PUSH_OBJECT
+ * macro. The actual parameter might be an expression with side effects, and
+ * this ensures that it will be executed only once.
+ */
+
+#define PUSH_OBJECT(objPtr) \
+    Tcl_IncrRefCount(*(++tosPtr) = (objPtr))
+
+#define POP_OBJECT()	*(tosPtr--)
+
+#define OBJ_AT_TOS	*tosPtr
+
+#define OBJ_UNDER_TOS	*(tosPtr-1)
+
+#define OBJ_AT_DEPTH(n)	*(tosPtr-(n))
+
+#define CURR_DEPTH	((ptrdiff_t) (tosPtr - initTosPtr))
+
+#define STACK_BASE(esPtr) ((esPtr)->stackWords - 1)
+
+/*
+ * Macros used to trace instruction execution. The macros TRACE,
+ * TRACE_WITH_OBJ, and O2S are only used inside TclNRExecuteByteCode. O2S is
+ * only used in TRACE* calls to get a string from an object.
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+#   define TRACE(a) \
+    while (traceInstructions) {					\
+	fprintf(stdout, "%2d: %2d (%u) %s ", iPtr->numLevels,	\
+		(int) CURR_DEPTH,				\
+		(unsigned) (pc - codePtr->codeStart),		\
+		GetOpcodeName(pc));				\
+	printf a;						\
+	break;							\
+    }
+#   define TRACE_APPEND(a) \
+    while (traceInstructions) {		\
+	printf a;			\
+	break;				\
+    }
+#   define TRACE_ERROR(interp) \
+    TRACE_APPEND(("ERROR: %.30s\n", O2S(Tcl_GetObjResult(interp))));
+#   define TRACE_WITH_OBJ(a, objPtr) \
+    while (traceInstructions) {					\
+	fprintf(stdout, "%2d: %2d (%u) %s ", iPtr->numLevels,	\
+		(int) CURR_DEPTH,				\
+		(unsigned) (pc - codePtr->codeStart),		\
+		GetOpcodeName(pc));				\
+	printf a;						\
+	TclPrintObject(stdout, objPtr, 30);			\
+	fprintf(stdout, "\n");					\
+	break;							\
+    }
+#   define O2S(objPtr) \
+    (objPtr ? TclGetString(objPtr) : "")
+#else /* !TCL_COMPILE_DEBUG */
+#   define TRACE(a)
+#   define TRACE_APPEND(a)
+#   define TRACE_ERROR(interp)
+#   define TRACE_WITH_OBJ(a, objPtr)
+#   define O2S(objPtr)
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ * DTrace instruction probe macros.
+ */
+
+#define TCL_DTRACE_INST_NEXT() \
+    do {								\
+	if (TCL_DTRACE_INST_DONE_ENABLED()) {				\
+	    if (curInstName) {						\
+		TCL_DTRACE_INST_DONE(curInstName, (int) CURR_DEPTH,	\
+			tosPtr);					\
+	    }								\
+	    curInstName = tclInstructionTable[*pc].name;		\
+	    if (TCL_DTRACE_INST_START_ENABLED()) {			\
+		TCL_DTRACE_INST_START(curInstName, (int) CURR_DEPTH,	\
+			tosPtr);					\
+	    }								\
+	} else if (TCL_DTRACE_INST_START_ENABLED()) {			\
+	    TCL_DTRACE_INST_START(tclInstructionTable[*pc].name,	\
+			(int) CURR_DEPTH, tosPtr);			\
+	}								\
+    } while (0)
+#define TCL_DTRACE_INST_LAST() \
+    do {								\
+	if (TCL_DTRACE_INST_DONE_ENABLED() && curInstName) {		\
+	    TCL_DTRACE_INST_DONE(curInstName, (int) CURR_DEPTH, tosPtr);\
+	}								\
+    } while (0)
+
+/*
+ * Macro used in this file to save a function call for common uses of
+ * TclGetNumberFromObj(). The ANSI C "prototype" is:
+ *
+ * MODULE_SCOPE int GetNumberFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+ *			ClientData *ptrPtr, int *tPtr);
+ */
+
+#ifdef TCL_WIDE_INT_IS_LONG
+#define GetNumberFromObj(interp, objPtr, ptrPtr, tPtr) \
+    (((objPtr)->typePtr == &tclIntType)					\
+	?	(*(tPtr) = TCL_NUMBER_LONG,				\
+		*(ptrPtr) = (ClientData)				\
+		    (&((objPtr)->internalRep.longValue)), TCL_OK) :	\
+    ((objPtr)->typePtr == &tclDoubleType)				\
+	?	(((TclIsNaN((objPtr)->internalRep.doubleValue))		\
+		    ?	(*(tPtr) = TCL_NUMBER_NAN)			\
+		    :	(*(tPtr) = TCL_NUMBER_DOUBLE)),			\
+		*(ptrPtr) = (ClientData)				\
+		    (&((objPtr)->internalRep.doubleValue)), TCL_OK) :	\
+    (((objPtr)->bytes != NULL) && ((objPtr)->length == 0))		\
+	? TCL_ERROR :			\
+    TclGetNumberFromObj((interp), (objPtr), (ptrPtr), (tPtr)))
+#else /* !TCL_WIDE_INT_IS_LONG */
+#define GetNumberFromObj(interp, objPtr, ptrPtr, tPtr) \
+    (((objPtr)->typePtr == &tclIntType)					\
+	?	(*(tPtr) = TCL_NUMBER_LONG,				\
+		*(ptrPtr) = (ClientData)				\
+		    (&((objPtr)->internalRep.longValue)), TCL_OK) :	\
+    ((objPtr)->typePtr == &tclWideIntType)				\
+	?	(*(tPtr) = TCL_NUMBER_WIDE,				\
+		*(ptrPtr) = (ClientData)				\
+		    (&((objPtr)->internalRep.wideValue)), TCL_OK) :	\
+    ((objPtr)->typePtr == &tclDoubleType)				\
+	?	(((TclIsNaN((objPtr)->internalRep.doubleValue))		\
+		    ?	(*(tPtr) = TCL_NUMBER_NAN)			\
+		    :	(*(tPtr) = TCL_NUMBER_DOUBLE)),			\
+		*(ptrPtr) = (ClientData)				\
+		    (&((objPtr)->internalRep.doubleValue)), TCL_OK) :	\
+    (((objPtr)->bytes != NULL) && ((objPtr)->length == 0))		\
+	? TCL_ERROR :			\
+    TclGetNumberFromObj((interp), (objPtr), (ptrPtr), (tPtr)))
+#endif /* TCL_WIDE_INT_IS_LONG */
+
+/*
+ * Macro used in this file to save a function call for common uses of
+ * Tcl_GetBooleanFromObj(). The ANSI C "prototype" is:
+ *
+ * MODULE_SCOPE int TclGetBooleanFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+ *			int *boolPtr);
+ */
+
+#define TclGetBooleanFromObj(interp, objPtr, boolPtr) \
+    ((((objPtr)->typePtr == &tclIntType)				\
+	|| ((objPtr)->typePtr == &tclBooleanType))			\
+	? (*(boolPtr) = ((objPtr)->internalRep.longValue!=0), TCL_OK)	\
+	: Tcl_GetBooleanFromObj((interp), (objPtr), (boolPtr)))
+
+/*
+ * Macro used to make the check for type overflow more mnemonic. This works by
+ * comparing sign bits; the rest of the word is irrelevant. The ANSI C
+ * "prototype" (where inttype_t is any integer type) is:
+ *
+ * MODULE_SCOPE int Overflowing(inttype_t a, inttype_t b, inttype_t sum);
+ *
+ * Check first the condition most likely to fail in usual code (at least for
+ * usage in [incr]: do the first summand and the sum have != signs?
+ */
+
+#define Overflowing(a,b,sum) ((((a)^(sum)) < 0) && (((a)^(b)) >= 0))
+
+/*
+ * Macro for checking whether the type is NaN, used when we're thinking about
+ * throwing an error for supplying a non-number number.
+ */
+
+#ifndef ACCEPT_NAN
+#define IsErroringNaNType(type)		((type) == TCL_NUMBER_NAN)
+#else
+#define IsErroringNaNType(type)		0
+#endif
+
+/*
+ * Auxiliary tables used to compute powers of small integers.
+ */
+
+#if (LONG_MAX == 0x7fffffff)
+
+/*
+ * Maximum base that, when raised to powers 2, 3, ... 8, fits in a 32-bit
+ * signed integer.
+ */
+
+static const long MaxBase32[] = {46340, 1290, 215, 73, 35, 21, 14};
+static const size_t MaxBase32Size = sizeof(MaxBase32)/sizeof(long);
+
+/*
+ * Table giving 3, 4, ..., 11, raised to the powers 9, 10, ..., as far as they
+ * fit in a 32-bit signed integer. Exp32Index[i] gives the starting index of
+ * powers of i+3; Exp32Value[i] gives the corresponding powers.
+ */
+
+static const unsigned short Exp32Index[] = {
+    0, 11, 18, 23, 26, 29, 31, 32, 33
+};
+static const size_t Exp32IndexSize =
+    sizeof(Exp32Index) / sizeof(unsigned short);
+static const long Exp32Value[] = {
+    19683, 59049, 177147, 531441, 1594323, 4782969, 14348907, 43046721,
+    129140163, 387420489, 1162261467, 262144, 1048576, 4194304,
+    16777216, 67108864, 268435456, 1073741824, 1953125, 9765625,
+    48828125, 244140625, 1220703125, 10077696, 60466176, 362797056,
+    40353607, 282475249, 1977326743, 134217728, 1073741824, 387420489,
+    1000000000
+};
+static const size_t Exp32ValueSize = sizeof(Exp32Value)/sizeof(long);
+#endif /* LONG_MAX == 0x7fffffff -- 32 bit machine */
+
+#if (LONG_MAX > 0x7fffffff) || !defined(TCL_WIDE_INT_IS_LONG)
+
+/*
+ * Maximum base that, when raised to powers 2, 3, ..., 16, fits in a
+ * Tcl_WideInt.
+ */
+
+static const Tcl_WideInt MaxBase64[] = {
+    (Tcl_WideInt)46340*65536+62259,	/* 3037000499 == isqrt(2**63-1) */
+    (Tcl_WideInt)2097151, (Tcl_WideInt)55108, (Tcl_WideInt)6208,
+    (Tcl_WideInt)1448, (Tcl_WideInt)511, (Tcl_WideInt)234, (Tcl_WideInt)127,
+    (Tcl_WideInt)78, (Tcl_WideInt)52, (Tcl_WideInt)38, (Tcl_WideInt)28,
+    (Tcl_WideInt)22, (Tcl_WideInt)18, (Tcl_WideInt)15
+};
+static const size_t MaxBase64Size = sizeof(MaxBase64)/sizeof(Tcl_WideInt);
+
+/*
+ * Table giving 3, 4, ..., 13 raised to powers greater than 16 when the
+ * results fit in a 64-bit signed integer.
+ */
+
+static const unsigned short Exp64Index[] = {
+    0, 23, 38, 49, 57, 63, 67, 70, 72, 74, 75, 76
+};
+static const size_t Exp64IndexSize =
+    sizeof(Exp64Index) / sizeof(unsigned short);
+static const Tcl_WideInt Exp64Value[] = {
+    (Tcl_WideInt)243*243*243*3*3,
+    (Tcl_WideInt)243*243*243*3*3*3,
+    (Tcl_WideInt)243*243*243*3*3*3*3,
+    (Tcl_WideInt)243*243*243*243,
+    (Tcl_WideInt)243*243*243*243*3,
+    (Tcl_WideInt)243*243*243*243*3*3,
+    (Tcl_WideInt)243*243*243*243*3*3*3,
+    (Tcl_WideInt)243*243*243*243*3*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243,
+    (Tcl_WideInt)243*243*243*243*243*3,
+    (Tcl_WideInt)243*243*243*243*243*3*3,
+    (Tcl_WideInt)243*243*243*243*243*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243*3*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243,
+    (Tcl_WideInt)243*243*243*243*243*243*3,
+    (Tcl_WideInt)243*243*243*243*243*243*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243*3*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243*243,
+    (Tcl_WideInt)243*243*243*243*243*243*243*3,
+    (Tcl_WideInt)243*243*243*243*243*243*243*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243*243*3*3*3,
+    (Tcl_WideInt)243*243*243*243*243*243*243*3*3*3*3,
+    (Tcl_WideInt)1024*1024*1024*4*4,
+    (Tcl_WideInt)1024*1024*1024*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*4*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024,
+    (Tcl_WideInt)1024*1024*1024*1024*4,
+    (Tcl_WideInt)1024*1024*1024*1024*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*4*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*1024,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*4,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*4*4*4*4,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*1024,
+    (Tcl_WideInt)1024*1024*1024*1024*1024*1024*4,
+    (Tcl_WideInt)3125*3125*3125*5*5,
+    (Tcl_WideInt)3125*3125*3125*5*5*5,
+    (Tcl_WideInt)3125*3125*3125*5*5*5*5,
+    (Tcl_WideInt)3125*3125*3125*3125,
+    (Tcl_WideInt)3125*3125*3125*3125*5,
+    (Tcl_WideInt)3125*3125*3125*3125*5*5,
+    (Tcl_WideInt)3125*3125*3125*3125*5*5*5,
+    (Tcl_WideInt)3125*3125*3125*3125*5*5*5*5,
+    (Tcl_WideInt)3125*3125*3125*3125*3125,
+    (Tcl_WideInt)3125*3125*3125*3125*3125*5,
+    (Tcl_WideInt)3125*3125*3125*3125*3125*5*5,
+    (Tcl_WideInt)7776*7776*7776*6*6,
+    (Tcl_WideInt)7776*7776*7776*6*6*6,
+    (Tcl_WideInt)7776*7776*7776*6*6*6*6,
+    (Tcl_WideInt)7776*7776*7776*7776,
+    (Tcl_WideInt)7776*7776*7776*7776*6,
+    (Tcl_WideInt)7776*7776*7776*7776*6*6,
+    (Tcl_WideInt)7776*7776*7776*7776*6*6*6,
+    (Tcl_WideInt)7776*7776*7776*7776*6*6*6*6,
+    (Tcl_WideInt)16807*16807*16807*7*7,
+    (Tcl_WideInt)16807*16807*16807*7*7*7,
+    (Tcl_WideInt)16807*16807*16807*7*7*7*7,
+    (Tcl_WideInt)16807*16807*16807*16807,
+    (Tcl_WideInt)16807*16807*16807*16807*7,
+    (Tcl_WideInt)16807*16807*16807*16807*7*7,
+    (Tcl_WideInt)32768*32768*32768*8*8,
+    (Tcl_WideInt)32768*32768*32768*8*8*8,
+    (Tcl_WideInt)32768*32768*32768*8*8*8*8,
+    (Tcl_WideInt)32768*32768*32768*32768,
+    (Tcl_WideInt)59049*59049*59049*9*9,
+    (Tcl_WideInt)59049*59049*59049*9*9*9,
+    (Tcl_WideInt)59049*59049*59049*9*9*9*9,
+    (Tcl_WideInt)100000*100000*100000*10*10,
+    (Tcl_WideInt)100000*100000*100000*10*10*10,
+    (Tcl_WideInt)161051*161051*161051*11*11,
+    (Tcl_WideInt)161051*161051*161051*11*11*11,
+    (Tcl_WideInt)248832*248832*248832*12*12,
+    (Tcl_WideInt)371293*371293*371293*13*13
+};
+static const size_t Exp64ValueSize = sizeof(Exp64Value) / sizeof(Tcl_WideInt);
+#endif /* (LONG_MAX > 0x7fffffff) || !defined(TCL_WIDE_INT_IS_LONG) */
+
+/*
+ * Markers for ExecuteExtendedBinaryMathOp.
+ */
+
+#define DIVIDED_BY_ZERO		((Tcl_Obj *) -1)
+#define EXPONENT_OF_ZERO	((Tcl_Obj *) -2)
+#define GENERAL_ARITHMETIC_ERROR ((Tcl_Obj *) -3)
+
+/*
+ * Declarations for local procedures to this file:
+ */
+
+#ifdef TCL_COMPILE_STATS
+static int		EvalStatsCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#endif /* TCL_COMPILE_STATS */
+#ifdef TCL_COMPILE_DEBUG
+static const char *	GetOpcodeName(const unsigned char *pc);
+static void		PrintByteCodeInfo(ByteCode *codePtr);
+static const char *	StringForResultCode(int result);
+static void		ValidatePcAndStackTop(ByteCode *codePtr,
+			    const unsigned char *pc, int stackTop,
+			    int checkStack);
+#endif /* TCL_COMPILE_DEBUG */
+static ByteCode *	CompileExprObj(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		DeleteExecStack(ExecStack *esPtr);
+static void		DupExprCodeInternalRep(Tcl_Obj *srcPtr,
+			    Tcl_Obj *copyPtr);
+MODULE_SCOPE int	TclCompareTwoNumbers(Tcl_Obj *valuePtr,
+			    Tcl_Obj *value2Ptr);
+static Tcl_Obj *	ExecuteExtendedBinaryMathOp(Tcl_Interp *interp,
+			    int opcode, Tcl_Obj **constants,
+			    Tcl_Obj *valuePtr, Tcl_Obj *value2Ptr);
+static Tcl_Obj *	ExecuteExtendedUnaryMathOp(int opcode,
+			    Tcl_Obj *valuePtr);
+static void		FreeExprCodeInternalRep(Tcl_Obj *objPtr);
+static ExceptionRange *	GetExceptRangeForPc(const unsigned char *pc,
+			    int searchMode, ByteCode *codePtr);
+static const char *	GetSrcInfoForPc(const unsigned char *pc,
+			    ByteCode *codePtr, int *lengthPtr,
+			    const unsigned char **pcBeg, int *cmdIdxPtr);
+static Tcl_Obj **	GrowEvaluationStack(ExecEnv *eePtr, int growth,
+			    int move);
+static void		IllegalExprOperandType(Tcl_Interp *interp,
+			    const unsigned char *pc, Tcl_Obj *opndPtr);
+static void		InitByteCodeExecution(Tcl_Interp *interp);
+static inline int	wordSkip(void *ptr);
+static void		ReleaseDictIterator(Tcl_Obj *objPtr);
+/* Useful elsewhere, make available in tclInt.h or stubs? */
+static Tcl_Obj **	StackAllocWords(Tcl_Interp *interp, int numWords);
+static Tcl_Obj **	StackReallocWords(Tcl_Interp *interp, int numWords);
+static Tcl_NRPostProc	CopyCallback;
+static Tcl_NRPostProc	ExprObjCallback;
+static Tcl_NRPostProc	FinalizeOONext;
+static Tcl_NRPostProc	FinalizeOONextFilter;
+static Tcl_NRPostProc   TEBCresume;
+
+/*
+ * The structure below defines a bytecode Tcl object type to hold the
+ * compiled bytecode for Tcl expressions.
+ */
+
+static const Tcl_ObjType exprCodeType = {
+    "exprcode",
+    FreeExprCodeInternalRep,	/* freeIntRepProc */
+    DupExprCodeInternalRep,	/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    NULL			/* setFromAnyProc */
+};
+
+/*
+ * Custom object type only used in this file; values of its type should never
+ * be seen by user scripts.
+ */
+
+static const Tcl_ObjType dictIteratorType = {
+    "dictIterator",
+    ReleaseDictIterator,
+    NULL, NULL, NULL
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReleaseDictIterator --
+ *
+ *	This takes apart a dictionary iterator that is stored in the given Tcl
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deallocates memory, marks the object as being untyped.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReleaseDictIterator(
+    Tcl_Obj *objPtr)
+{
+    Tcl_DictSearch *searchPtr;
+    Tcl_Obj *dictPtr;
+
+    /*
+     * First kill the search, and then release the reference to the dictionary
+     * that we were holding.
+     */
+
+    searchPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    Tcl_DictObjDone(searchPtr);
+    ckfree(searchPtr);
+
+    dictPtr = objPtr->internalRep.twoPtrValue.ptr2;
+    TclDecrRefCount(dictPtr);
+
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InitByteCodeExecution --
+ *
+ *	This procedure is called once to initialize the Tcl bytecode
+ *	interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	This procedure initializes the array of instruction names. If
+ *	compiling with the TCL_COMPILE_STATS flag, it initializes the array
+ *	that counts the executions of each instruction and it creates the
+ *	"evalstats" command. It also establishes the link between the Tcl
+ *	"tcl_traceExec" and C "tclTraceExec" variables.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+InitByteCodeExecution(
+    Tcl_Interp *interp)		/* Interpreter for which the Tcl variable
+				 * "tcl_traceExec" is linked to control
+				 * instruction tracing. */
+{
+#ifdef TCL_COMPILE_DEBUG
+    if (Tcl_LinkVar(interp, "tcl_traceExec", (char *) &tclTraceExec,
+	    TCL_LINK_INT) != TCL_OK) {
+	Tcl_Panic("InitByteCodeExecution: can't create link for tcl_traceExec variable");
+    }
+#endif
+#ifdef TCL_COMPILE_STATS
+    Tcl_CreateObjCommand(interp, "evalstats", EvalStatsCmd, NULL, NULL);
+#endif /* TCL_COMPILE_STATS */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateExecEnv --
+ *
+ *	This procedure creates a new execution environment for Tcl bytecode
+ *	execution. An ExecEnv points to a Tcl evaluation stack. An ExecEnv is
+ *	typically created once for each Tcl interpreter (Interp structure) and
+ *	recursively passed to TclNRExecuteByteCode to execute ByteCode sequences
+ *	for nested commands.
+ *
+ * Results:
+ *	A newly allocated ExecEnv is returned. This points to an empty
+ *	evaluation stack of the standard initial size.
+ *
+ * Side effects:
+ *	The bytecode interpreter is also initialized here, as this procedure
+ *	will be called before any call to TclNRExecuteByteCode.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ExecEnv *
+TclCreateExecEnv(
+    Tcl_Interp *interp,		/* Interpreter for which the execution
+				 * environment is being created. */
+    int size)			/* The initial stack size, in number of words
+				 * [sizeof(Tcl_Obj*)] */
+{
+    ExecEnv *eePtr = ckalloc(sizeof(ExecEnv));
+    ExecStack *esPtr = ckalloc(sizeof(ExecStack)
+	    + (size_t) (size-1) * sizeof(Tcl_Obj *));
+
+    eePtr->execStackPtr = esPtr;
+    TclNewLongObj(eePtr->constants[0], 0);
+    Tcl_IncrRefCount(eePtr->constants[0]);
+    TclNewLongObj(eePtr->constants[1], 1);
+    Tcl_IncrRefCount(eePtr->constants[1]);
+    eePtr->interp = interp;
+    eePtr->callbackPtr = NULL;
+    eePtr->corPtr = NULL;
+    eePtr->rewind = 0;
+
+    esPtr->prevPtr = NULL;
+    esPtr->nextPtr = NULL;
+    esPtr->markerPtr = NULL;
+    esPtr->endPtr = &esPtr->stackWords[size-1];
+    esPtr->tosPtr = STACK_BASE(esPtr);
+
+    Tcl_MutexLock(&execMutex);
+    if (!execInitialized) {
+	InitByteCodeExecution(interp);
+	execInitialized = 1;
+    }
+    Tcl_MutexUnlock(&execMutex);
+
+    return eePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteExecEnv --
+ *
+ *	Frees the storage for an ExecEnv.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Storage for an ExecEnv and its contained storage (e.g. the evaluation
+ *	stack) is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteExecStack(
+    ExecStack *esPtr)
+{
+    if (esPtr->markerPtr && !cachedInExit) {
+	Tcl_Panic("freeing an execStack which is still in use");
+    }
+
+    if (esPtr->prevPtr) {
+	esPtr->prevPtr->nextPtr = esPtr->nextPtr;
+    }
+    if (esPtr->nextPtr) {
+	esPtr->nextPtr->prevPtr = esPtr->prevPtr;
+    }
+    ckfree(esPtr);
+}
+
+void
+TclDeleteExecEnv(
+    ExecEnv *eePtr)		/* Execution environment to free. */
+{
+    ExecStack *esPtr = eePtr->execStackPtr, *tmpPtr;
+
+	cachedInExit = TclInExit();
+
+    /*
+     * Delete all stacks in this exec env.
+     */
+
+    while (esPtr->nextPtr) {
+	esPtr = esPtr->nextPtr;
+    }
+    while (esPtr) {
+	tmpPtr = esPtr;
+	esPtr = tmpPtr->prevPtr;
+	DeleteExecStack(tmpPtr);
+    }
+
+    TclDecrRefCount(eePtr->constants[0]);
+    TclDecrRefCount(eePtr->constants[1]);
+    if (eePtr->callbackPtr && !cachedInExit) {
+	Tcl_Panic("Deleting execEnv with pending TEOV callbacks!");
+    }
+    if (eePtr->corPtr && !cachedInExit) {
+	Tcl_Panic("Deleting execEnv with existing coroutine");
+    }
+    ckfree(eePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeExecution --
+ *
+ *	Finalizes the execution environment setup so that it can be later
+ *	reinitialized.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	After this call, the next time TclCreateExecEnv will be called it will
+ *	call InitByteCodeExecution.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeExecution(void)
+{
+    Tcl_MutexLock(&execMutex);
+    execInitialized = 0;
+    Tcl_MutexUnlock(&execMutex);
+}
+
+/*
+ * Auxiliary code to insure that GrowEvaluationStack always returns correctly
+ * aligned memory.
+ *
+ * WALLOCALIGN represents the alignment reqs in words, just as TCL_ALLOCALIGN
+ * represents the reqs in bytes. This assumes that TCL_ALLOCALIGN is a
+ * multiple of the wordsize 'sizeof(Tcl_Obj *)'.
+ */
+
+#define WALLOCALIGN \
+    (TCL_ALLOCALIGN/sizeof(Tcl_Obj *))
+
+/*
+ * wordSkip computes how many words have to be skipped until the next aligned
+ * word. Note that we are only interested in the low order bits of ptr, so
+ * that any possible information loss in PTR2INT is of no consequence.
+ */
+
+static inline int
+wordSkip(
+    void *ptr)
+{
+    int mask = TCL_ALLOCALIGN-1;
+    int base = PTR2INT(ptr) & mask;
+    return (TCL_ALLOCALIGN - base)/sizeof(Tcl_Obj *);
+}
+
+/*
+ * Given a marker, compute where the following aligned memory starts.
+ */
+
+#define MEMSTART(markerPtr) \
+    ((markerPtr) + wordSkip(markerPtr))
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GrowEvaluationStack --
+ *
+ *	This procedure grows a Tcl evaluation stack stored in an ExecEnv,
+ *	copying over the words since the last mark if so requested. A mark is
+ *	set at the beginning of the new area when no copying is requested.
+ *
+ * Results:
+ *	Returns a pointer to the first usable word in the (possibly) grown
+ *	stack.
+ *
+ * Side effects:
+ *	The size of the evaluation stack may be grown, a marker is set
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj **
+GrowEvaluationStack(
+    ExecEnv *eePtr,		/* Points to the ExecEnv with an evaluation
+				 * stack to enlarge. */
+    int growth,			/* How much larger than the current used
+				 * size. */
+    int move)			/* 1 if move words since last marker. */
+{
+    ExecStack *esPtr = eePtr->execStackPtr, *oldPtr = NULL;
+    int newBytes, newElems, currElems;
+    int needed = growth - (esPtr->endPtr - esPtr->tosPtr);
+    Tcl_Obj **markerPtr = esPtr->markerPtr, **memStart;
+    int moveWords = 0;
+
+    if (move) {
+	if (!markerPtr) {
+	    Tcl_Panic("STACK: Reallocating with no previous alloc");
+	}
+	if (needed <= 0) {
+	    return MEMSTART(markerPtr);
+	}
+    } else {
+#ifndef PURIFY
+	Tcl_Obj **tmpMarkerPtr = esPtr->tosPtr + 1;
+	int offset = wordSkip(tmpMarkerPtr);
+
+	if (needed + offset < 0) {
+	    /*
+	     * Put a marker pointing to the previous marker in this stack, and
+	     * store it in esPtr as the current marker. Return a pointer to
+	     * the start of aligned memory.
+	     */
+
+	    esPtr->markerPtr = tmpMarkerPtr;
+	    memStart = tmpMarkerPtr + offset;
+	    esPtr->tosPtr = memStart - 1;
+	    *esPtr->markerPtr = (Tcl_Obj *) markerPtr;
+	    return memStart;
+	}
+#endif
+    }
+
+    /*
+     * Reset move to hold the number of words to be moved to new stack (if
+     * any) and growth to hold the complete stack requirements: add one for
+     * the marker, (WALLOCALIGN-1) for the maximal possible offset.
+     */
+
+    if (move) {
+	moveWords = esPtr->tosPtr - MEMSTART(markerPtr) + 1;
+    }
+    needed = growth + moveWords + WALLOCALIGN;
+
+
+    /*
+     * Check if there is enough room in the next stack (if there is one, it
+     * should be both empty and the last one!)
+     */
+
+    if (esPtr->nextPtr) {
+	oldPtr = esPtr;
+	esPtr = oldPtr->nextPtr;
+	currElems = esPtr->endPtr - STACK_BASE(esPtr);
+	if (esPtr->markerPtr || (esPtr->tosPtr != STACK_BASE(esPtr))) {
+	    Tcl_Panic("STACK: Stack after current is in use");
+	}
+	if (esPtr->nextPtr) {
+	    Tcl_Panic("STACK: Stack after current is not last");
+	}
+	if (needed <= currElems) {
+	    goto newStackReady;
+	}
+	DeleteExecStack(esPtr);
+	esPtr = oldPtr;
+    } else {
+	currElems = esPtr->endPtr - STACK_BASE(esPtr);
+    }
+
+    /*
+     * We need to allocate a new stack! It needs to store 'growth' words,
+     * including the elements to be copied over and the new marker.
+     */
+
+#ifndef PURIFY
+    newElems = 2*currElems;
+    while (needed > newElems) {
+	newElems *= 2;
+    }
+#else
+    newElems = needed;
+#endif
+
+    newBytes = sizeof(ExecStack) + (newElems-1) * sizeof(Tcl_Obj *);
+
+    oldPtr = esPtr;
+    esPtr = ckalloc(newBytes);
+
+    oldPtr->nextPtr = esPtr;
+    esPtr->prevPtr = oldPtr;
+    esPtr->nextPtr = NULL;
+    esPtr->endPtr = &esPtr->stackWords[newElems-1];
+
+  newStackReady:
+    eePtr->execStackPtr = esPtr;
+
+    /*
+     * Store a NULL marker at the beginning of the stack, to indicate that
+     * this is the first marker in this stack and that rewinding to here
+     * should actually be a return to the previous stack.
+     */
+
+    esPtr->stackWords[0] = NULL;
+    esPtr->markerPtr = &esPtr->stackWords[0];
+    memStart = MEMSTART(esPtr->markerPtr);
+    esPtr->tosPtr = memStart - 1;
+
+    if (move) {
+	memcpy(memStart, MEMSTART(markerPtr), moveWords*sizeof(Tcl_Obj *));
+	esPtr->tosPtr += moveWords;
+	oldPtr->markerPtr = (Tcl_Obj **) *markerPtr;
+	oldPtr->tosPtr = markerPtr-1;
+    }
+
+    /*
+     * Free the old stack if it is now unused.
+     */
+
+    if (!oldPtr->markerPtr) {
+	DeleteExecStack(oldPtr);
+    }
+
+    return memStart;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * TclStackAlloc, TclStackRealloc, TclStackFree --
+ *
+ *	Allocate memory from the execution stack; it has to be returned later
+ *	with a call to TclStackFree.
+ *
+ * Results:
+ *	A pointer to the first byte allocated, or panics if the allocation did
+ *	not succeed.
+ *
+ * Side effects:
+ *	The execution stack may be grown.
+ *
+ *--------------------------------------------------------------
+ */
+
+static Tcl_Obj **
+StackAllocWords(
+    Tcl_Interp *interp,
+    int numWords)
+{
+    /*
+     * Note that GrowEvaluationStack sets a marker in the stack. This marker
+     * is read when rewinding, e.g., by TclStackFree.
+     */
+
+    Interp *iPtr = (Interp *) interp;
+    ExecEnv *eePtr = iPtr->execEnvPtr;
+    Tcl_Obj **resPtr = GrowEvaluationStack(eePtr, numWords, 0);
+
+    eePtr->execStackPtr->tosPtr += numWords;
+    return resPtr;
+}
+
+static Tcl_Obj **
+StackReallocWords(
+    Tcl_Interp *interp,
+    int numWords)
+{
+    Interp *iPtr = (Interp *) interp;
+    ExecEnv *eePtr = iPtr->execEnvPtr;
+    Tcl_Obj **resPtr = GrowEvaluationStack(eePtr, numWords, 1);
+
+    eePtr->execStackPtr->tosPtr += numWords;
+    return resPtr;
+}
+
+void
+TclStackFree(
+    Tcl_Interp *interp,
+    void *freePtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    ExecEnv *eePtr;
+    ExecStack *esPtr;
+    Tcl_Obj **markerPtr, *marker;
+
+    if (iPtr == NULL || iPtr->execEnvPtr == NULL) {
+	ckfree(freePtr);
+	return;
+    }
+
+    /*
+     * Rewind the stack to the previous marker position. The current marker,
+     * as set in the last call to GrowEvaluationStack, contains a pointer to
+     * the previous marker.
+     */
+
+    eePtr = iPtr->execEnvPtr;
+    esPtr = eePtr->execStackPtr;
+    markerPtr = esPtr->markerPtr;
+    marker = *markerPtr;
+
+    if ((freePtr != NULL) && (MEMSTART(markerPtr) != (Tcl_Obj **)freePtr)) {
+	Tcl_Panic("TclStackFree: incorrect freePtr (%p != %p). Call out of sequence?",
+		freePtr, MEMSTART(markerPtr));
+    }
+
+    esPtr->tosPtr = markerPtr - 1;
+    esPtr->markerPtr = (Tcl_Obj **) marker;
+    if (marker) {
+	return;
+    }
+
+    /*
+     * Return to previous active stack. Note that repeated expansions or
+     * reallocs could have generated several unused intervening stacks: free
+     * them too.
+     */
+
+    while (esPtr->nextPtr) {
+	esPtr = esPtr->nextPtr;
+    }
+    esPtr->tosPtr = STACK_BASE(esPtr);
+    while (esPtr->prevPtr) {
+	ExecStack *tmpPtr = esPtr->prevPtr;
+	if (tmpPtr->tosPtr == STACK_BASE(tmpPtr)) {
+	    DeleteExecStack(tmpPtr);
+	} else {
+	    break;
+	}
+    }
+    if (esPtr->prevPtr) {
+	eePtr->execStackPtr = esPtr->prevPtr;
+#ifdef PURIFY
+	eePtr->execStackPtr->nextPtr = NULL;
+	DeleteExecStack(esPtr);
+#endif
+    } else {
+	eePtr->execStackPtr = esPtr;
+    }
+}
+
+void *
+TclStackAlloc(
+    Tcl_Interp *interp,
+    int numBytes)
+{
+    Interp *iPtr = (Interp *) interp;
+    int numWords = (numBytes + (sizeof(Tcl_Obj *) - 1))/sizeof(Tcl_Obj *);
+
+    if (iPtr == NULL || iPtr->execEnvPtr == NULL) {
+	return (void *) ckalloc(numBytes);
+    }
+
+    return (void *) StackAllocWords(interp, numWords);
+}
+
+void *
+TclStackRealloc(
+    Tcl_Interp *interp,
+    void *ptr,
+    int numBytes)
+{
+    Interp *iPtr = (Interp *) interp;
+    ExecEnv *eePtr;
+    ExecStack *esPtr;
+    Tcl_Obj **markerPtr;
+    int numWords;
+
+    if (iPtr == NULL || iPtr->execEnvPtr == NULL) {
+	return (void *) ckrealloc((char *) ptr, numBytes);
+    }
+
+    eePtr = iPtr->execEnvPtr;
+    esPtr = eePtr->execStackPtr;
+    markerPtr = esPtr->markerPtr;
+
+    if (MEMSTART(markerPtr) != (Tcl_Obj **)ptr) {
+	Tcl_Panic("TclStackRealloc: incorrect ptr. Call out of sequence?");
+    }
+
+    numWords = (numBytes + (sizeof(Tcl_Obj *) - 1))/sizeof(Tcl_Obj *);
+    return (void *) StackReallocWords(interp, numWords);
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_ExprObj --
+ *
+ *	Evaluate an expression in a Tcl_Obj.
+ *
+ * Results:
+ *	A standard Tcl object result. If the result is other than TCL_OK, then
+ *	the interpreter's result contains an error message. If the result is
+ *	TCL_OK, then a pointer to the expression's result value object is
+ *	stored in resultPtrPtr. In that case, the object's ref count is
+ *	incremented to reflect the reference returned to the caller; the
+ *	caller is then responsible for the resulting object and must, for
+ *	example, decrement the ref count when it is finished with the object.
+ *
+ * Side effects:
+ *	Any side effects caused by subcommands in the expression, if any. The
+ *	interpreter result is not modified unless there is an error.
+ *
+ *--------------------------------------------------------------
+ */
+
+int
+Tcl_ExprObj(
+    Tcl_Interp *interp,		/* Context in which to evaluate the
+				 * expression. */
+    register Tcl_Obj *objPtr,	/* Points to Tcl object containing expression
+				 * to evaluate. */
+    Tcl_Obj **resultPtrPtr)	/* Where the Tcl_Obj* that is the expression
+				 * result is stored if no errors occur. */
+{
+    NRE_callback *rootPtr = TOP_CB(interp);
+    Tcl_Obj *resultPtr;
+
+    TclNewObj(resultPtr);
+    TclNRAddCallback(interp, CopyCallback, resultPtrPtr, resultPtr,
+	    NULL, NULL);
+    Tcl_NRExprObj(interp, objPtr, resultPtr);
+    return TclNRRunCallbacks(interp, TCL_OK, rootPtr);
+}
+
+static int
+CopyCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj **resultPtrPtr = data[0];
+    Tcl_Obj *resultPtr = data[1];
+
+    if (result == TCL_OK) {
+	*resultPtrPtr = resultPtr;
+	Tcl_IncrRefCount(resultPtr);
+    } else {
+	Tcl_DecrRefCount(resultPtr);
+    }
+    return result;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_NRExprObj --
+ *
+ *	Request evaluation of the expression in a Tcl_Obj by the NR stack.
+ *
+ * Results:
+ *	Returns TCL_OK.
+ *
+ * Side effects:
+ *	Compiles objPtr as a Tcl expression and places callbacks on the
+ *	NR stack to execute the bytecode and store the result in resultPtr.
+ *	If bytecode execution raises an exception, nothing is written
+ *	to resultPtr, and the exceptional return code flows up the NR
+ *	stack.  If the exception is TCL_ERROR, an error message is left
+ *	in the interp result and the interp's return options dictionary
+ *	holds additional error information too.  Execution of the bytecode
+ *	may have other side effects, depending on the expression.
+ *
+ *--------------------------------------------------------------
+ */
+
+int
+Tcl_NRExprObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    Tcl_Obj *resultPtr)
+{
+    ByteCode *codePtr;
+    Tcl_InterpState state = Tcl_SaveInterpState(interp, TCL_OK);
+
+    Tcl_ResetResult(interp);
+    codePtr = CompileExprObj(interp, objPtr);
+
+    Tcl_NRAddCallback(interp, ExprObjCallback, state, resultPtr,
+	    NULL, NULL);
+    return TclNRExecuteByteCode(interp, codePtr);
+}
+
+static int
+ExprObjCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_InterpState state = data[0];
+    Tcl_Obj *resultPtr = data[1];
+
+    if (result == TCL_OK) {
+	TclSetDuplicateObj(resultPtr, Tcl_GetObjResult(interp));
+	(void) Tcl_RestoreInterpState(interp, state);
+    } else {
+	Tcl_DiscardInterpState(state);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompileExprObj --
+ *	Compile a Tcl expression value into ByteCode.
+ *
+ * Results:
+ *	A (ByteCode *) is returned pointing to the resulting ByteCode.
+ *
+ * Side effects:
+ *	The Tcl_ObjType of objPtr is changed to the "exprcode" type,
+ *	and the ByteCode is kept in the internal rep (along with context
+ *	data for checking validity) for faster operations the next time
+ *	CompileExprObj is called on the same value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ByteCode *
+CompileExprObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    CompileEnv compEnv;		/* Compilation environment structure allocated
+				 * in frame. */
+    register ByteCode *codePtr = NULL;
+				/* Tcl Internal type of bytecode. Initialized
+				 * to avoid compiler warning. */
+
+    /*
+     * Get the expression ByteCode from the object. If it exists, make sure it
+     * is valid in the current context.
+     */
+    if (objPtr->typePtr == &exprCodeType) {
+	Namespace *namespacePtr = iPtr->varFramePtr->nsPtr;
+
+	codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	if (((Interp *) *codePtr->interpHandle != iPtr)
+		|| (codePtr->compileEpoch != iPtr->compileEpoch)
+		|| (codePtr->nsPtr != namespacePtr)
+		|| (codePtr->nsEpoch != namespacePtr->resolverEpoch)
+		|| (codePtr->localCachePtr != iPtr->varFramePtr->localCachePtr)) {
+	    TclFreeIntRep(objPtr);
+	}
+    }
+    if (objPtr->typePtr != &exprCodeType) {
+	/*
+	 * TIP #280: No invoker (yet) - Expression compilation.
+	 */
+
+	const char *string = TclGetString(objPtr);
+
+	TclInitCompileEnv(interp, &compEnv, string, objPtr->length, NULL, 0);
+	TclCompileExpr(interp, string, objPtr->length, &compEnv, 0);
+
+	/*
+	 * Successful compilation. If the expression yielded no instructions,
+	 * push an zero object as the expression's result.
+	 */
+
+	if (compEnv.codeNext == compEnv.codeStart) {
+	    TclEmitPush(TclRegisterLiteral(&compEnv, "0", 1, 0),
+		    &compEnv);
+	}
+
+	/*
+	 * Add a "done" instruction as the last instruction and change the
+	 * object into a ByteCode object. Ownership of the literal objects and
+	 * aux data items is given to the ByteCode object.
+	 */
+
+	TclEmitOpcode(INST_DONE, &compEnv);
+	codePtr = TclInitByteCodeObj(objPtr, &exprCodeType, &compEnv);
+	TclFreeCompileEnv(&compEnv);
+	if (iPtr->varFramePtr->localCachePtr) {
+	    codePtr->localCachePtr = iPtr->varFramePtr->localCachePtr;
+	    codePtr->localCachePtr->refCount++;
+	}
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceCompile == 2) {
+	    TclPrintByteCodeObj(interp, objPtr);
+	    fflush(stdout);
+	}
+#endif /* TCL_COMPILE_DEBUG */
+    }
+    return codePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupExprCodeInternalRep --
+ *
+ *	Part of the Tcl object type implementation for Tcl expression
+ *	bytecode. We do not copy the bytecode intrep. Instead, we return
+ *	without setting copyPtr->typePtr, so the copy is a plain string copy
+ *	of the expression value, and if it is to be used as a compiled
+ *	expression, it will just need a recompile.
+ *
+ *	This makes sense, because with Tcl's copy-on-write practices, the
+ *	usual (only?) time Tcl_DuplicateObj() will be called is when the copy
+ *	is about to be modified, which would invalidate any copied bytecode
+ *	anyway. The only reason it might make sense to copy the bytecode is if
+ *	we had some modifying routines that operated directly on the intrep,
+ *	like we do for lists and dicts.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupExprCodeInternalRep(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *copyPtr)
+{
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeExprCodeInternalRep --
+ *
+ *	Part of the Tcl object type implementation for Tcl expression
+ *	bytecode. Frees the storage allocated to hold the internal rep, unless
+ *	ref counts indicate bytecode execution is still in progress.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May free allocated memory. Leaves objPtr untyped.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeExprCodeInternalRep(
+    Tcl_Obj *objPtr)
+{
+    ByteCode *codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclReleaseByteCode(codePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompileObj --
+ *
+ *	This procedure compiles the script contained in a Tcl_Obj.
+ *
+ * Results:
+ *	A pointer to the corresponding ByteCode, never NULL.
+ *
+ * Side effects:
+ *	The object is shimmered to bytecode type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ByteCode *
+TclCompileObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    const CmdFrame *invoker,
+    int word)
+{
+    register Interp *iPtr = (Interp *) interp;
+    register ByteCode *codePtr;	/* Tcl Internal type of bytecode. */
+    Namespace *namespacePtr = iPtr->varFramePtr->nsPtr;
+
+    /*
+     * If the object is not already of tclByteCodeType, compile it (and reset
+     * the compilation flags in the interpreter; this should be done after any
+     * compilation). Otherwise, check that it is "fresh" enough.
+     */
+
+    if (objPtr->typePtr == &tclByteCodeType) {
+	/*
+	 * Make sure the Bytecode hasn't been invalidated by, e.g., someone
+	 * redefining a command with a compile procedure (this might make the
+	 * compiled code wrong). The object needs to be recompiled if it was
+	 * compiled in/for a different interpreter, or for a different
+	 * namespace, or for the same namespace but with different name
+	 * resolution rules. Precompiled objects, however, are immutable and
+	 * therefore they are not recompiled, even if the epoch has changed.
+	 *
+	 * To be pedantically correct, we should also check that the
+	 * originating procPtr is the same as the current context procPtr
+	 * (assuming one exists at all - none for global level). This code is
+	 * #def'ed out because [info body] was changed to never return a
+	 * bytecode type object, which should obviate us from the extra checks
+	 * here.
+	 */
+
+	codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	if (((Interp *) *codePtr->interpHandle != iPtr)
+		|| (codePtr->compileEpoch != iPtr->compileEpoch)
+		|| (codePtr->nsPtr != namespacePtr)
+		|| (codePtr->nsEpoch != namespacePtr->resolverEpoch)) {
+	    if (!(codePtr->flags & TCL_BYTECODE_PRECOMPILED)) {
+		goto recompileObj;
+	    }
+	    if ((Interp *) *codePtr->interpHandle != iPtr) {
+		Tcl_Panic("Tcl_EvalObj: compiled script jumped interps");
+	    }
+	    codePtr->compileEpoch = iPtr->compileEpoch;
+	}
+
+	/*
+	 * Check that any compiled locals do refer to the current proc
+	 * environment! If not, recompile.
+	 */
+
+	if (!(codePtr->flags & TCL_BYTECODE_PRECOMPILED) &&
+		(codePtr->procPtr == NULL) &&
+		(codePtr->localCachePtr != iPtr->varFramePtr->localCachePtr)){
+	    goto recompileObj;
+	}
+
+	/*
+	 * #280.
+	 * Literal sharing fix. This part of the fix is not required by 8.4
+	 * nor 8.5, because they eval-direct any literals, so just saving the
+	 * argument locations per command in bytecode is enough, embedded
+	 * 'eval' commands, etc. get the correct information.
+	 *
+	 * But in 8.6 all the embedded script are compiled, and the resulting
+	 * bytecode stored in the literal. Now the shared literal has bytecode
+	 * with location data for _one_ particular location this literal is
+	 * found at. If we get executed from a different location the bytecode
+	 * has to be recompiled to get the correct locations. Not doing this
+	 * will execute the saved bytecode with data for a different location,
+	 * causing 'info frame' to point to the wrong place in the sources.
+	 *
+	 * Future optimizations ...
+	 * (1) Save the location data (ExtCmdLoc) keyed by start line. In that
+	 *     case we recompile once per location of the literal, but not
+	 *     continously, because the moment we have all locations we do not
+	 *     need to recompile any longer.
+	 *
+	 * (2) Alternative: Do not recompile, tell the execution engine the
+	 *     offset between saved starting line and actual one. Then modify
+	 *     the users to adjust the locations they have by this offset.
+	 *
+	 * (3) Alternative 2: Do not fully recompile, adjust just the location
+	 *     information.
+	 */
+
+	if (invoker == NULL) {
+	    return codePtr;
+	} else {
+	    Tcl_HashEntry *hePtr =
+		    Tcl_FindHashEntry(iPtr->lineBCPtr, codePtr);
+	    ExtCmdLoc *eclPtr;
+	    CmdFrame *ctxCopyPtr;
+	    int redo;
+
+	    if (!hePtr) {
+		return codePtr;
+	    }
+
+	    eclPtr = Tcl_GetHashValue(hePtr);
+	    redo = 0;
+	    ctxCopyPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+	    *ctxCopyPtr = *invoker;
+
+	    if (invoker->type == TCL_LOCATION_BC) {
+		/*
+		 * Note: Type BC => ctx.data.eval.path    is not used.
+		 *		    ctx.data.tebc.codePtr used instead
+		 */
+
+		TclGetSrcInfoForPc(ctxCopyPtr);
+		if (ctxCopyPtr->type == TCL_LOCATION_SOURCE) {
+		    /*
+		     * The reference made by 'TclGetSrcInfoForPc' is dead.
+		     */
+
+		    Tcl_DecrRefCount(ctxCopyPtr->data.eval.path);
+		    ctxCopyPtr->data.eval.path = NULL;
+		}
+	    }
+
+	    if (word < ctxCopyPtr->nline) {
+		/*
+		 * Note: We do not care if the line[word] is -1. This is a
+		 * difference and requires a recompile (location changed from
+		 * absolute to relative, literal is used fixed and through
+		 * variable)
+		 *
+		 * Example:
+		 * test info-32.0 using literal of info-24.8
+		 *     (dict with ... vs           set body ...).
+		 */
+
+		redo = ((eclPtr->type == TCL_LOCATION_SOURCE)
+			    && (eclPtr->start != ctxCopyPtr->line[word]))
+			|| ((eclPtr->type == TCL_LOCATION_BC)
+			    && (ctxCopyPtr->type == TCL_LOCATION_SOURCE));
+	    }
+
+	    TclStackFree(interp, ctxCopyPtr);
+	    if (!redo) {
+		return codePtr;
+	    }
+	}
+    }
+
+  recompileObj:
+    iPtr->errorLine = 1;
+
+    /*
+     * TIP #280. Remember the invoker for a moment in the interpreter
+     * structures so that the byte code compiler can pick it up when
+     * initializing the compilation environment, i.e. the extended location
+     * information.
+     */
+
+    iPtr->invokeCmdFramePtr = invoker;
+    iPtr->invokeWord = word;
+    TclSetByteCodeFromAny(interp, objPtr, NULL, NULL);
+    iPtr->invokeCmdFramePtr = NULL;
+    codePtr = objPtr->internalRep.twoPtrValue.ptr1;
+    if (iPtr->varFramePtr->localCachePtr) {
+	codePtr->localCachePtr = iPtr->varFramePtr->localCachePtr;
+	codePtr->localCachePtr->refCount++;
+    }
+    return codePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclIncrObj --
+ *
+ *	Increment an integeral value in a Tcl_Obj by an integeral value held
+ *	in another Tcl_Obj. Caller is responsible for making sure we can
+ *	update the first object.
+ *
+ * Results:
+ *	TCL_ERROR if either object is non-integer, and TCL_OK otherwise. On
+ *	error, an error message is left in the interpreter (if it is not NULL,
+ *	of course).
+ *
+ * Side effects:
+ *	valuePtr gets the new incrmented value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclIncrObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *valuePtr,
+    Tcl_Obj *incrPtr)
+{
+    ClientData ptr1, ptr2;
+    int type1, type2;
+    mp_int value, incr;
+
+    if (Tcl_IsShared(valuePtr)) {
+	Tcl_Panic("%s called with shared object", "TclIncrObj");
+    }
+
+    if (GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK) {
+	/*
+	 * Produce error message (reparse?!)
+	 */
+
+	return TclGetIntFromObj(interp, valuePtr, &type1);
+    }
+    if (GetNumberFromObj(NULL, incrPtr, &ptr2, &type2) != TCL_OK) {
+	/*
+	 * Produce error message (reparse?!)
+	 */
+
+	TclGetIntFromObj(interp, incrPtr, &type1);
+	Tcl_AddErrorInfo(interp, "\n    (reading increment)");
+	return TCL_ERROR;
+    }
+
+    if ((type1 == TCL_NUMBER_LONG) && (type2 == TCL_NUMBER_LONG)) {
+	long augend = *((const long *) ptr1);
+	long addend = *((const long *) ptr2);
+	long sum = augend + addend;
+
+	/*
+	 * Overflow when (augend and sum have different sign) and (augend and
+	 * addend have the same sign). This is encapsulated in the Overflowing
+	 * macro.
+	 */
+
+	if (!Overflowing(augend, addend, sum)) {
+	    TclSetLongObj(valuePtr, sum);
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	{
+	    Tcl_WideInt w1 = (Tcl_WideInt) augend;
+	    Tcl_WideInt w2 = (Tcl_WideInt) addend;
+
+	    /*
+	     * We know the sum value is outside the long range, so we use the
+	     * macro form that doesn't range test again.
+	     */
+
+	    TclSetWideIntObj(valuePtr, w1 + w2);
+	    return TCL_OK;
+	}
+#endif
+    }
+
+    if ((type1 == TCL_NUMBER_DOUBLE) || (type1 == TCL_NUMBER_NAN)) {
+	/*
+	 * Produce error message (reparse?!)
+	 */
+
+	return TclGetIntFromObj(interp, valuePtr, &type1);
+    }
+    if ((type2 == TCL_NUMBER_DOUBLE) || (type2 == TCL_NUMBER_NAN)) {
+	/*
+	 * Produce error message (reparse?!)
+	 */
+
+	TclGetIntFromObj(interp, incrPtr, &type1);
+	Tcl_AddErrorInfo(interp, "\n    (reading increment)");
+	return TCL_ERROR;
+    }
+
+#ifndef TCL_WIDE_INT_IS_LONG
+    if ((type1 != TCL_NUMBER_BIG) && (type2 != TCL_NUMBER_BIG)) {
+	Tcl_WideInt w1, w2, sum;
+
+	TclGetWideIntFromObj(NULL, valuePtr, &w1);
+	TclGetWideIntFromObj(NULL, incrPtr, &w2);
+	sum = w1 + w2;
+
+	/*
+	 * Check for overflow.
+	 */
+
+	if (!Overflowing(w1, w2, sum)) {
+	    Tcl_SetWideIntObj(valuePtr, sum);
+	    return TCL_OK;
+	}
+    }
+#endif
+
+    Tcl_TakeBignumFromObj(interp, valuePtr, &value);
+    Tcl_GetBignumFromObj(interp, incrPtr, &incr);
+    mp_add(&value, &incr, &value);
+    mp_clear(&incr);
+    Tcl_SetBignumObj(valuePtr, &value);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArgumentBCEnter --
+ *
+ *	This is a helper for TclNRExecuteByteCode/TEBCresume that encapsulates
+ *	a code sequence that is fairly common in the code but *not* commonly
+ *	called.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	May register information about the bytecode in the command frame.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ArgumentBCEnter(
+    Tcl_Interp *interp,
+    ByteCode *codePtr,
+    TEBCdata *tdPtr,
+    const unsigned char *pc,
+    int objc,
+    Tcl_Obj **objv)
+{
+    int cmd;
+
+    if (GetSrcInfoForPc(pc, codePtr, NULL, NULL, &cmd)) {
+	TclArgumentBCEnter(interp, objv, objc, codePtr, &tdPtr->cmdFrame, cmd,
+		pc - codePtr->codeStart);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRExecuteByteCode --
+ *
+ *	This procedure executes the instructions of a ByteCode structure. It
+ *	returns when a "done" instruction is executed or an error occurs.
+ *
+ * Results:
+ *	The return value is one of the return codes defined in tcl.h (such as
+ *	TCL_OK), and interp->objResultPtr refers to a Tcl object that either
+ *	contains the result of executing the code or an error message.
+ *
+ * Side effects:
+ *	Almost certainly, depending on the ByteCode's instructions.
+ *
+ *----------------------------------------------------------------------
+ */
+#define	bcFramePtr	(&TD->cmdFrame)
+#define	initCatchTop	((ptrdiff_t *) (&TD->stack[-1]))
+#define	initTosPtr	((Tcl_Obj **) (initCatchTop+codePtr->maxExceptDepth))
+#define esPtr		(iPtr->execEnvPtr->execStackPtr)
+
+int
+TclNRExecuteByteCode(
+    Tcl_Interp *interp,		/* Token for command interpreter. */
+    ByteCode *codePtr)		/* The bytecode sequence to interpret. */
+{
+    Interp *iPtr = (Interp *) interp;
+    TEBCdata *TD;
+    int size = sizeof(TEBCdata) - 1
+	    + (codePtr->maxStackDepth + codePtr->maxExceptDepth)
+		* sizeof(void *);
+    int numWords = (size + sizeof(Tcl_Obj *) - 1) / sizeof(Tcl_Obj *);
+
+    TclPreserveByteCode(codePtr);
+
+    /*
+     * Reserve the stack, setup the TEBCdataPtr (TD) and CallFrame
+     *
+     * The execution uses a unified stack: first a TEBCdata, immediately
+     * above it a CmdFrame, then the catch stack, then the execution stack.
+     *
+     * Make sure the catch stack is large enough to hold the maximum number of
+     * catch commands that could ever be executing at the same time (this will
+     * be no more than the exception range array's depth). Make sure the
+     * execution stack is large enough to execute this ByteCode.
+     */
+
+    TD = (TEBCdata *) GrowEvaluationStack(iPtr->execEnvPtr, numWords, 0);
+    esPtr->tosPtr = initTosPtr;
+
+    TD->codePtr     = codePtr;
+    TD->catchTop    = initCatchTop;
+    TD->auxObjList  = NULL;
+
+    /*
+     * TIP #280: Initialize the frame. Do not push it yet: it will be pushed
+     * every time that we call out from this TD, popped when we return to it.
+     */
+
+    bcFramePtr->type = ((codePtr->flags & TCL_BYTECODE_PRECOMPILED)
+	    ? TCL_LOCATION_PREBC : TCL_LOCATION_BC);
+    bcFramePtr->level = (iPtr->cmdFramePtr ? iPtr->cmdFramePtr->level+1 : 1);
+    bcFramePtr->framePtr = iPtr->framePtr;
+    bcFramePtr->nextPtr = iPtr->cmdFramePtr;
+    bcFramePtr->nline = 0;
+    bcFramePtr->line = NULL;
+    bcFramePtr->litarg = NULL;
+    bcFramePtr->data.tebc.codePtr = codePtr;
+    bcFramePtr->data.tebc.pc = NULL;
+    bcFramePtr->cmdObj = NULL;
+    bcFramePtr->cmd = NULL;
+    bcFramePtr->len = 0;
+
+#ifdef TCL_COMPILE_STATS
+    iPtr->stats.numExecutions++;
+#endif
+
+    /*
+     * Test namespace-50.9 demonstrates the need for this call.
+     * Use a --enable-symbols=mem bug to see.
+     */
+
+    TclResetRewriteEnsemble(interp, 1);
+
+    /*
+     * Push the callback for bytecode execution
+     */
+
+    TclNRAddCallback(interp, TEBCresume, TD, /* pc */ NULL,
+	    /* cleanup */ INT2PTR(0), NULL);
+    return TCL_OK;
+}
+
+static int
+TEBCresume(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    /*
+     * Compiler cast directive - not a real variable.
+     *	   Interp *iPtr = (Interp *) interp;
+     */
+#define iPtr ((Interp *) interp)
+
+    /*
+     * Check just the read-traced/write-traced bit of a variable.
+     */
+
+#define ReadTraced(varPtr) ((varPtr)->flags & VAR_TRACED_READ)
+#define WriteTraced(varPtr) ((varPtr)->flags & VAR_TRACED_WRITE)
+#define UnsetTraced(varPtr) ((varPtr)->flags & VAR_TRACED_UNSET)
+
+    /*
+     * Bottom of allocated stack holds the NR data
+     */
+
+    /*
+     * Constants: variables that do not change during the execution, used
+     * sporadically: no special need for speed.
+     */
+
+    unsigned interruptCounter = 1;
+				/* Counter that is used to work out when to
+				 * call Tcl_AsyncReady(). This must be 1
+				 * initially so that we call the async-check
+				 * stanza early, otherwise there are command
+				 * sequences that can make the interpreter
+				 * busy-loop without an opportunity to
+				 * recognise an interrupt. */
+    const char *curInstName;
+#ifdef TCL_COMPILE_DEBUG
+    int traceInstructions;	/* Whether we are doing instruction-level
+				 * tracing or not. */
+#endif
+
+    Var *compiledLocals = iPtr->varFramePtr->compiledLocals;
+    Tcl_Obj **constants = &iPtr->execEnvPtr->constants[0];
+
+#define LOCAL(i)	(&compiledLocals[(i)])
+#define TCONST(i)	(constants[(i)])
+
+    /*
+     * These macros are just meant to save some global variables that are not
+     * used too frequently
+     */
+
+    TEBCdata *TD = data[0];
+#define auxObjList	(TD->auxObjList)
+#define catchTop	(TD->catchTop)
+#define codePtr		(TD->codePtr)
+
+    /*
+     * Globals: variables that store state, must remain valid at all times.
+     */
+
+    Tcl_Obj **tosPtr;		/* Cached pointer to top of evaluation
+				 * stack. */
+    const unsigned char *pc = data[1];
+                                /* The current program counter. */
+    unsigned char inst;         /* The currently running instruction */
+
+    /*
+     * Transfer variables - needed only between opcodes, but not while
+     * executing an instruction.
+     */
+
+    int cleanup = PTR2INT(data[2]);
+    Tcl_Obj *objResultPtr;
+    int checkInterp;            /* Indicates when a check of interp readyness
+				 * is necessary. Set by CACHE_STACK_INFO() */
+
+    /*
+     * Locals - variables that are used within opcodes or bounded sections of
+     * the file (jumps between opcodes within a family).
+     * NOTE: These are now mostly defined locally where needed.
+     */
+
+    Tcl_Obj *objPtr, *valuePtr, *value2Ptr, *part1Ptr, *part2Ptr, *tmpPtr;
+    Tcl_Obj **objv;
+    int objc = 0;
+    int opnd, length, pcAdjustment;
+    Var *varPtr, *arrayPtr;
+#ifdef TCL_COMPILE_DEBUG
+    char cmdNameBuf[21];
+#endif
+
+#ifdef TCL_COMPILE_DEBUG
+    int starting = 1;
+    traceInstructions = (tclTraceExec == 3);
+#endif
+
+    TEBC_DATA_DIG();
+
+#ifdef TCL_COMPILE_DEBUG
+    if (!pc && (tclTraceExec >= 2)) {
+	PrintByteCodeInfo(codePtr);
+	fprintf(stdout, "  Starting stack top=%d\n", (int) CURR_DEPTH);
+	fflush(stdout);
+    }
+#endif
+
+    if (!pc) {
+	/* bytecode is starting from scratch */
+	checkInterp = 0;
+	pc = codePtr->codeStart;
+	goto cleanup0;
+    } else {
+        /* resume from invocation */
+	CACHE_STACK_INFO();
+
+	NRE_ASSERT(iPtr->cmdFramePtr == bcFramePtr);
+	if (bcFramePtr->cmdObj) {
+	    Tcl_DecrRefCount(bcFramePtr->cmdObj);
+	    bcFramePtr->cmdObj = NULL;
+	    bcFramePtr->cmd = NULL;
+	}
+	iPtr->cmdFramePtr = bcFramePtr->nextPtr;
+	if (iPtr->flags & INTERP_DEBUG_FRAME) {
+	    TclArgumentBCRelease(interp, bcFramePtr);
+	}
+	if (iPtr->execEnvPtr->rewind) {
+	    result = TCL_ERROR;
+	    goto abnormalReturn;
+	}
+	if (codePtr->flags & TCL_BYTECODE_RECOMPILE) {
+	    iPtr->flags |= ERR_ALREADY_LOGGED;
+	    codePtr->flags &= ~TCL_BYTECODE_RECOMPILE;
+	}
+
+	if (result != TCL_OK) {
+	    pc--;
+	    goto processExceptionReturn;
+	}
+
+	/*
+	 * Push the call's object result and continue execution with the next
+	 * instruction.
+	 */
+
+	TRACE_WITH_OBJ(("%u => ... after \"%.20s\": TCL_OK, result=",
+		objc, cmdNameBuf), Tcl_GetObjResult(interp));
+
+	/*
+	 * Reset the interp's result to avoid possible duplications of large
+	 * objects [Bug 781585]. We do not call Tcl_ResetResult to avoid any
+	 * side effects caused by the resetting of errorInfo and errorCode
+	 * [Bug 804681], which are not needed here. We chose instead to
+	 * manipulate the interp's object result directly.
+	 *
+	 * Note that the result object is now in objResultPtr, it keeps the
+	 * refCount it had in its role of iPtr->objResultPtr.
+	 */
+
+	objResultPtr = Tcl_GetObjResult(interp);
+	TclNewObj(objPtr);
+	Tcl_IncrRefCount(objPtr);
+	iPtr->objResultPtr = objPtr;
+#ifndef TCL_COMPILE_DEBUG
+	if (*pc == INST_POP) {
+	    TclDecrRefCount(objResultPtr);
+	    NEXT_INST_V(1, cleanup, 0);
+	}
+#endif
+	NEXT_INST_V(0, cleanup, -1);
+    }
+
+    /*
+     * Targets for standard instruction endings; unrolled for speed in the
+     * most frequent cases (instructions that consume up to two stack
+     * elements).
+     *
+     * This used to be a "for(;;)" loop, with each instruction doing its own
+     * cleanup.
+     */
+
+  cleanupV_pushObjResultPtr:
+    switch (cleanup) {
+    case 0:
+	*(++tosPtr) = (objResultPtr);
+	goto cleanup0;
+    default:
+	cleanup -= 2;
+	while (cleanup--) {
+	    objPtr = POP_OBJECT();
+	    TclDecrRefCount(objPtr);
+	}
+    case 2:
+    cleanup2_pushObjResultPtr:
+	objPtr = POP_OBJECT();
+	TclDecrRefCount(objPtr);
+    case 1:
+    cleanup1_pushObjResultPtr:
+	objPtr = OBJ_AT_TOS;
+	TclDecrRefCount(objPtr);
+    }
+    OBJ_AT_TOS = objResultPtr;
+    goto cleanup0;
+
+  cleanupV:
+    switch (cleanup) {
+    default:
+	cleanup -= 2;
+	while (cleanup--) {
+	    objPtr = POP_OBJECT();
+	    TclDecrRefCount(objPtr);
+	}
+    case 2:
+    cleanup2:
+	objPtr = POP_OBJECT();
+	TclDecrRefCount(objPtr);
+    case 1:
+    cleanup1:
+	objPtr = POP_OBJECT();
+	TclDecrRefCount(objPtr);
+    case 0:
+	/*
+	 * We really want to do nothing now, but this is needed for some
+	 * compilers (SunPro CC).
+	 */
+
+	break;
+    }
+  cleanup0:
+
+    /*
+     * Check for asynchronous handlers [Bug 746722]; we do the check every
+     * ASYNC_CHECK_COUNT instructions.
+     */
+
+    if ((--interruptCounter) == 0) {
+	interruptCounter = ASYNC_CHECK_COUNT;
+	DECACHE_STACK_INFO();
+	if (TclAsyncReady(iPtr)) {
+	    result = Tcl_AsyncInvoke(interp, result);
+	    if (result == TCL_ERROR) {
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+	}
+
+	if (TclCanceled(iPtr)) {
+	    if (Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG) == TCL_ERROR) {
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+	}
+
+	if (TclLimitReady(iPtr->limit)) {
+	    if (Tcl_LimitCheck(interp) == TCL_ERROR) {
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+	}
+	CACHE_STACK_INFO();
+    }
+
+    /*
+     * These two instructions account for 26% of all instructions (according
+     * to measurements on tclbench by Ben Vitale
+     * [http://www.cs.toronto.edu/syslab/pubs/tcl2005-vitale-zaleski.pdf]
+     * Resolving them before the switch reduces the cost of branch
+     * mispredictions, seems to improve runtime by 5% to 15%, and (amazingly!)
+     * reduces total obj size.
+     */
+
+    inst = *pc;
+
+    peepholeStart:
+#ifdef TCL_COMPILE_STATS
+    iPtr->stats.instructionCount[*pc]++;
+#endif
+
+#ifdef TCL_COMPILE_DEBUG
+    /*
+     * Skip the stack depth check if an expansion is in progress.
+     */
+
+    CHECK_STACK();
+    if (traceInstructions) {
+	fprintf(stdout, "%2d: %2d ", iPtr->numLevels, (int) CURR_DEPTH);
+	TclPrintInstruction(codePtr, pc);
+	fflush(stdout);
+    }
+#endif /* TCL_COMPILE_DEBUG */
+
+    TCL_DTRACE_INST_NEXT();
+
+    if (inst == INST_LOAD_SCALAR1) {
+	goto instLoadScalar1;
+    } else if (inst == INST_PUSH1) {
+	PUSH_OBJECT(codePtr->objArrayPtr[TclGetUInt1AtPtr(pc+1)]);
+	TRACE_WITH_OBJ(("%u => ", TclGetUInt1AtPtr(pc+1)), OBJ_AT_TOS);
+	inst = *(pc += 2);
+	goto peepholeStart;
+    } else if (inst == INST_START_CMD) {
+	/*
+	 * Peephole: do not run INST_START_CMD, just skip it
+	 */
+
+	iPtr->cmdCount += TclGetUInt4AtPtr(pc+5);
+	if (checkInterp) {
+	    checkInterp = 0;
+	    if (((codePtr->compileEpoch != iPtr->compileEpoch) ||
+		 (codePtr->nsEpoch != iPtr->varFramePtr->nsPtr->resolverEpoch)) &&
+		!(codePtr->flags & TCL_BYTECODE_PRECOMPILED)) {
+		goto instStartCmdFailed;
+	    }
+	}
+	inst = *(pc += 9);
+	goto peepholeStart;
+    } else if (inst == INST_NOP) {
+#ifndef TCL_COMPILE_DEBUG
+	while (inst == INST_NOP)
+#endif
+	{
+	    inst = *++pc;
+	}
+	goto peepholeStart;
+    }
+
+    switch (inst) {
+    case INST_SYNTAX:
+    case INST_RETURN_IMM: {
+	int code = TclGetInt4AtPtr(pc+1);
+	int level = TclGetUInt4AtPtr(pc+5);
+
+	/*
+	 * OBJ_AT_TOS is returnOpts, OBJ_UNDER_TOS is resultObjPtr.
+	 */
+
+	TRACE(("%u %u => ", code, level));
+	result = TclProcessReturn(interp, code, level, OBJ_AT_TOS);
+	if (result == TCL_OK) {
+	    TRACE_APPEND(("continuing to next instruction (result=\"%.30s\")\n",
+		    O2S(objResultPtr)));
+	    NEXT_INST_F(9, 1, 0);
+	}
+	Tcl_SetObjResult(interp, OBJ_UNDER_TOS);
+	if (*pc == INST_SYNTAX) {
+	    iPtr->flags &= ~ERR_ALREADY_LOGGED;
+	}
+	cleanup = 2;
+	TRACE_APPEND(("\n"));
+	goto processExceptionReturn;
+    }
+
+    case INST_RETURN_STK:
+	TRACE(("=> "));
+	objResultPtr = POP_OBJECT();
+	result = Tcl_SetReturnOptions(interp, OBJ_AT_TOS);
+	if (result == TCL_OK) {
+	    Tcl_DecrRefCount(OBJ_AT_TOS);
+	    OBJ_AT_TOS = objResultPtr;
+	    TRACE_APPEND(("continuing to next instruction (result=\"%.30s\")\n",
+		    O2S(objResultPtr)));
+	    NEXT_INST_F(1, 0, 0);
+	} else if (result == TCL_ERROR) {
+	    /*
+	     * BEWARE! Must do this in this order, because an error in the
+	     * option dictionary overrides the result (and can be verified by
+	     * test).
+	     */
+
+	    Tcl_SetObjResult(interp, objResultPtr);
+	    Tcl_SetReturnOptions(interp, OBJ_AT_TOS);
+	    Tcl_DecrRefCount(OBJ_AT_TOS);
+	    OBJ_AT_TOS = objResultPtr;
+	} else {
+	    Tcl_DecrRefCount(OBJ_AT_TOS);
+	    OBJ_AT_TOS = objResultPtr;
+	    Tcl_SetObjResult(interp, objResultPtr);
+	}
+	cleanup = 1;
+	TRACE_APPEND(("\n"));
+	goto processExceptionReturn;
+
+    {
+	CoroutineData *corPtr;
+	int yieldParameter;
+
+    case INST_YIELD:
+	corPtr = iPtr->execEnvPtr->corPtr;
+	TRACE(("%.30s => ", O2S(OBJ_AT_TOS)));
+	if (!corPtr) {
+	    TRACE_APPEND(("ERROR: yield outside coroutine\n"));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "yield can only be called in a coroutine", -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "ILLEGAL_YIELD",
+		    NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceExec >= 2) {
+	    if (traceInstructions) {
+		TRACE_APPEND(("YIELD...\n"));
+	    } else {
+		fprintf(stdout, "%d: (%u) yielding value \"%.30s\"\n",
+			iPtr->numLevels, (unsigned)(pc - codePtr->codeStart),
+			Tcl_GetString(OBJ_AT_TOS));
+	    }
+	    fflush(stdout);
+	}
+#endif
+	yieldParameter = 0;
+	Tcl_SetObjResult(interp, OBJ_AT_TOS);
+	goto doYield;
+
+    case INST_YIELD_TO_INVOKE:
+	corPtr = iPtr->execEnvPtr->corPtr;
+	valuePtr = OBJ_AT_TOS;
+	if (!corPtr) {
+	    TRACE(("[%.30s] => ERROR: yield outside coroutine\n",
+		    O2S(valuePtr)));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "yieldto can only be called in a coroutine", -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "ILLEGAL_YIELD",
+		    NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	if (((Namespace *)TclGetCurrentNamespace(interp))->flags & NS_DYING) {
+	    TRACE(("[%.30s] => ERROR: yield in deleted\n",
+		    O2S(valuePtr)));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "yieldto called in deleted namespace", -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "COROUTINE", "YIELDTO_IN_DELETED",
+		    NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceExec >= 2) {
+	    if (traceInstructions) {
+		TRACE(("[%.30s] => YIELD...\n", O2S(valuePtr)));
+	    } else {
+		/* FIXME: What is the right thing to trace? */
+		fprintf(stdout, "%d: (%u) yielding to [%.30s]\n",
+			iPtr->numLevels, (unsigned)(pc - codePtr->codeStart),
+			TclGetString(valuePtr));
+	    }
+	    fflush(stdout);
+	}
+#endif
+
+	/*
+	 * Install a tailcall record in the caller and continue with the
+	 * yield. The yield is switched into multi-return mode (via the
+	 * 'yieldParameter').
+	 */
+
+	Tcl_IncrRefCount(valuePtr);
+	iPtr->execEnvPtr = corPtr->callerEEPtr;
+	TclSetTailcall(interp, valuePtr);
+	iPtr->execEnvPtr = corPtr->eePtr;
+	yieldParameter = (PTR2INT(NULL)+1);	/*==CORO_ACTIVATE_YIELDM*/
+
+    doYield:
+	/* TIP #280: Record the last piece of info needed by
+	 * 'TclGetSrcInfoForPc', and push the frame.
+	 */
+
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+
+	if (iPtr->flags & INTERP_DEBUG_FRAME) {
+	    ArgumentBCEnter(interp, codePtr, TD, pc, objc, objv);
+	}
+
+	pc++;
+	cleanup = 1;
+	TEBC_YIELD();
+	TclNRAddCallback(interp, TclNRCoroutineActivateCallback, corPtr,
+		INT2PTR(yieldParameter), NULL, NULL);
+	return TCL_OK;
+    }
+
+    case INST_TAILCALL: {
+	Tcl_Obj *listPtr, *nsObjPtr;
+
+	opnd = TclGetUInt1AtPtr(pc+1);
+
+	if (!(iPtr->varFramePtr->isProcCallFrame & 1)) {
+	    TRACE(("%d => ERROR: tailcall in non-proc context\n", opnd));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "tailcall can only be called from a proc or lambda", -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "TAILCALL", "ILLEGAL", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	/* FIXME: What is the right thing to trace? */
+	{
+	    register int i;
+
+	    TRACE(("%d [", opnd));
+	    for (i=opnd-1 ; i>=0 ; i--) {
+		TRACE_APPEND(("\"%.30s\"", O2S(OBJ_AT_DEPTH(i))));
+		if (i > 0) {
+		    TRACE_APPEND((" "));
+		}
+	    }
+	    TRACE_APPEND(("] => RETURN..."));
+	}
+#endif
+
+	/*
+	 * Push the evaluation of the called command into the NR callback
+	 * stack.
+	 */
+
+	listPtr = Tcl_NewListObj(opnd, &OBJ_AT_DEPTH(opnd-1));
+	nsObjPtr = Tcl_NewStringObj(iPtr->varFramePtr->nsPtr->fullName, -1);
+	TclListObjSetElement(interp, listPtr, 0, nsObjPtr);
+	if (iPtr->varFramePtr->tailcallPtr) {
+	    Tcl_DecrRefCount(iPtr->varFramePtr->tailcallPtr);
+	}
+	iPtr->varFramePtr->tailcallPtr = listPtr;
+
+	result = TCL_RETURN;
+	cleanup = opnd;
+	goto processExceptionReturn;
+    }
+
+    case INST_DONE:
+	if (tosPtr > initTosPtr) {
+	    /*
+	     * Set the interpreter's object result to point to the topmost
+	     * object from the stack, and check for a possible [catch]. The
+	     * stackTop's level and refCount will be handled by "processCatch"
+	     * or "abnormalReturn".
+	     */
+
+	    Tcl_SetObjResult(interp, OBJ_AT_TOS);
+#ifdef TCL_COMPILE_DEBUG
+	    TRACE_WITH_OBJ(("=> return code=%d, result=", result),
+		    iPtr->objResultPtr);
+	    if (traceInstructions) {
+		fprintf(stdout, "\n");
+	    }
+#endif
+	    goto checkForCatch;
+	}
+	(void) POP_OBJECT();
+	goto abnormalReturn;
+
+    case INST_PUSH4:
+	objResultPtr = codePtr->objArrayPtr[TclGetUInt4AtPtr(pc+1)];
+	TRACE_WITH_OBJ(("%u => ", TclGetUInt4AtPtr(pc+1)), objResultPtr);
+	NEXT_INST_F(5, 0, 1);
+
+    case INST_POP:
+	TRACE_WITH_OBJ(("=> discarding "), OBJ_AT_TOS);
+	objPtr = POP_OBJECT();
+	TclDecrRefCount(objPtr);
+	NEXT_INST_F(1, 0, 0);
+
+    case INST_DUP:
+	objResultPtr = OBJ_AT_TOS;
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_OVER:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	objResultPtr = OBJ_AT_DEPTH(opnd);
+	TRACE_WITH_OBJ(("%u => ", opnd), objResultPtr);
+	NEXT_INST_F(5, 0, 1);
+
+    case INST_REVERSE: {
+	Tcl_Obj **a, **b;
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	a = tosPtr-(opnd-1);
+	b = tosPtr;
+	while (a<b) {
+	    tmpPtr = *a;
+	    *a = *b;
+	    *b = tmpPtr;
+	    a++; b--;
+	}
+	TRACE(("%u => OK\n", opnd));
+	NEXT_INST_F(5, 0, 0);
+    }
+
+    case INST_STR_CONCAT1:
+
+	opnd = TclGetUInt1AtPtr(pc+1);
+
+	if (TCL_OK != TclStringCatObjv(interp, /* inPlace */ 1,
+		opnd, &OBJ_AT_DEPTH(opnd-1), &objResultPtr)) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	TRACE_WITH_OBJ(("%u => ", opnd), objResultPtr);
+	NEXT_INST_V(2, opnd, 1);
+
+    case INST_CONCAT_STK:
+	/*
+	 * Pop the opnd (objc) top stack elements, run through Tcl_ConcatObj,
+	 * and then decrement their ref counts.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	objResultPtr = Tcl_ConcatObj(opnd, &OBJ_AT_DEPTH(opnd-1));
+	TRACE_WITH_OBJ(("%u => ", opnd), objResultPtr);
+	NEXT_INST_V(5, opnd, 1);
+
+    case INST_EXPAND_START:
+	/*
+	 * Push an element to the auxObjList. This records the current
+	 * stack depth - i.e., the point in the stack where the expanded
+	 * command starts.
+	 *
+	 * Use a Tcl_Obj as linked list element; slight mem waste, but faster
+	 * allocation than ckalloc. This also abuses the Tcl_Obj structure, as
+	 * we do not define a special tclObjType for it. It is not dangerous
+	 * as the obj is never passed anywhere, so that all manipulations are
+	 * performed here and in INST_INVOKE_EXPANDED (in case of an expansion
+	 * error, also in INST_EXPAND_STKTOP).
+	 */
+
+	TclNewObj(objPtr);
+	objPtr->internalRep.twoPtrValue.ptr2 = INT2PTR(CURR_DEPTH);
+	objPtr->length = 0;
+	PUSH_TAUX_OBJ(objPtr);
+	TRACE(("=> mark depth as %d\n", (int) CURR_DEPTH));
+	NEXT_INST_F(1, 0, 0);
+
+    case INST_EXPAND_DROP:
+	/*
+	 * Drops an element of the auxObjList, popping stack elements to
+	 * restore the stack to the state before the point where the aux
+	 * element was created.
+	 */
+
+	CLANG_ASSERT(auxObjList);
+	objc = CURR_DEPTH - PTR2INT(auxObjList->internalRep.twoPtrValue.ptr2);
+	POP_TAUX_OBJ();
+#ifdef TCL_COMPILE_DEBUG
+	/* Ugly abuse! */
+	starting = 1;
+#endif
+	TRACE(("=> drop %d items\n", objc));
+	NEXT_INST_V(1, objc, 0);
+
+    case INST_EXPAND_STKTOP: {
+	int i;
+	ptrdiff_t moved;
+
+	/*
+	 * Make sure that the element at stackTop is a list; if not, just
+	 * leave with an error. Note that the element from the expand list
+	 * will be removed at checkForCatch.
+	 */
+
+	objPtr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" => ", O2S(objPtr)));
+	if (TclListObjGetElements(interp, objPtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	(void) POP_OBJECT();
+
+	/*
+	 * Make sure there is enough room in the stack to expand this list
+	 * *and* process the rest of the command (at least up to the next
+	 * argument expansion or command end). The operand is the current
+	 * stack depth, as seen by the compiler.
+	 */
+
+	auxObjList->length += objc - 1;
+	if ((objc > 1) && (auxObjList->length > 0)) {
+	    length = auxObjList->length /* Total expansion room we need */
+		    + codePtr->maxStackDepth /* Beyond the original max */
+		    - CURR_DEPTH;	/* Relative to where we are */
+	    DECACHE_STACK_INFO();
+	    moved = GrowEvaluationStack(iPtr->execEnvPtr, length, 1)
+		    - (Tcl_Obj **) TD;
+	    if (moved) {
+		/*
+		 * Change the global data to point to the new stack: move the
+		 * TEBCdataPtr TD, recompute the position of every other
+		 * stack-allocated parameter, update the stack pointers.
+		 */
+
+		TD = (TEBCdata *) (((Tcl_Obj **)TD) + moved);
+
+		catchTop += moved;
+		tosPtr += moved;
+	    }
+	}
+
+	/*
+	 * Expand the list at stacktop onto the stack; free the list. Knowing
+	 * that it has a freeIntRepProc we use Tcl_DecrRefCount().
+	 */
+
+	for (i = 0; i < objc; i++) {
+	    PUSH_OBJECT(objv[i]);
+	}
+
+	TRACE_APPEND(("OK\n"));
+	Tcl_DecrRefCount(objPtr);
+	NEXT_INST_F(5, 0, 0);
+    }
+
+    case INST_EXPR_STK: {
+	ByteCode *newCodePtr;
+
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+	DECACHE_STACK_INFO();
+	newCodePtr = CompileExprObj(interp, OBJ_AT_TOS);
+	CACHE_STACK_INFO();
+	cleanup = 1;
+	pc++;
+	TEBC_YIELD();
+	return TclNRExecuteByteCode(interp, newCodePtr);
+    }
+
+	/*
+	 * INVOCATION BLOCK
+	 */
+
+    instEvalStk:
+    case INST_EVAL_STK:
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+
+	cleanup = 1;
+	pc += 1;
+	TEBC_YIELD();
+	return TclNREvalObjEx(interp, OBJ_AT_TOS, 0, NULL, 0);
+
+    case INST_INVOKE_EXPANDED:
+	CLANG_ASSERT(auxObjList);
+	objc = CURR_DEPTH - PTR2INT(auxObjList->internalRep.twoPtrValue.ptr2);
+	POP_TAUX_OBJ();
+	if (objc) {
+	    pcAdjustment = 1;
+	    goto doInvocation;
+	}
+
+	/*
+	 * Nothing was expanded, return {}.
+	 */
+
+	TclNewObj(objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_INVOKE_STK4:
+	objc = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	goto doInvocation;
+
+    case INST_INVOKE_STK1:
+	objc = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+
+    doInvocation:
+	objv = &OBJ_AT_DEPTH(objc-1);
+	cleanup = objc;
+
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceExec >= 2) {
+	    int i;
+
+	    if (traceInstructions) {
+		strncpy(cmdNameBuf, TclGetString(objv[0]), 20);
+		TRACE(("%u => call ", objc));
+	    } else {
+		fprintf(stdout, "%d: (%u) invoking ", iPtr->numLevels,
+			(unsigned)(pc - codePtr->codeStart));
+	    }
+	    for (i = 0;  i < objc;  i++) {
+		TclPrintObject(stdout, objv[i], 15);
+		fprintf(stdout, " ");
+	    }
+	    fprintf(stdout, "\n");
+	    fflush(stdout);
+	}
+#endif /*TCL_COMPILE_DEBUG*/
+
+	/*
+	 * Finally, let TclEvalObjv handle the command.
+	 *
+	 * TIP #280: Record the last piece of info needed by
+	 * 'TclGetSrcInfoForPc', and push the frame.
+	 */
+
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+
+	if (iPtr->flags & INTERP_DEBUG_FRAME) {
+	    ArgumentBCEnter(interp, codePtr, TD, pc, objc, objv);
+	}
+
+	DECACHE_STACK_INFO();
+
+	pc += pcAdjustment;
+	TEBC_YIELD();
+	return TclNREvalObjv(interp, objc, objv,
+		TCL_EVAL_NOERR | TCL_EVAL_SOURCE_IN_FRAME, NULL);
+
+#if TCL_SUPPORT_84_BYTECODE
+    case INST_CALL_BUILTIN_FUNC1:
+	/*
+	 * Call one of the built-in pre-8.5 Tcl math functions. This
+	 * translates to INST_INVOKE_STK1 with the first argument of
+	 * ::tcl::mathfunc::$objv[0]. We need to insert the named math
+	 * function into the stack.
+	 */
+
+	opnd = TclGetUInt1AtPtr(pc+1);
+	if ((opnd < 0) || (opnd > LAST_BUILTIN_FUNC)) {
+	    TRACE(("UNRECOGNIZED BUILTIN FUNC CODE %d\n", opnd));
+	    Tcl_Panic("TclNRExecuteByteCode: unrecognized builtin function code %d", opnd);
+	}
+
+	TclNewLiteralStringObj(objPtr, "::tcl::mathfunc::");
+	Tcl_AppendToObj(objPtr, tclBuiltinFuncTable[opnd].name, -1);
+
+	/*
+	 * Only 0, 1 or 2 args.
+	 */
+
+	{
+	    int numArgs = tclBuiltinFuncTable[opnd].numArgs;
+	    Tcl_Obj *tmpPtr1, *tmpPtr2;
+
+	    if (numArgs == 0) {
+		PUSH_OBJECT(objPtr);
+	    } else if (numArgs == 1) {
+		tmpPtr1 = POP_OBJECT();
+		PUSH_OBJECT(objPtr);
+		PUSH_OBJECT(tmpPtr1);
+		Tcl_DecrRefCount(tmpPtr1);
+	    } else {
+		tmpPtr2 = POP_OBJECT();
+		tmpPtr1 = POP_OBJECT();
+		PUSH_OBJECT(objPtr);
+		PUSH_OBJECT(tmpPtr1);
+		PUSH_OBJECT(tmpPtr2);
+		Tcl_DecrRefCount(tmpPtr1);
+		Tcl_DecrRefCount(tmpPtr2);
+	    }
+	    objc = numArgs + 1;
+	}
+	pcAdjustment = 2;
+	goto doInvocation;
+
+    case INST_CALL_FUNC1:
+	/*
+	 * Call a non-builtin Tcl math function previously registered by a
+	 * call to Tcl_CreateMathFunc pre-8.5. This is essentially
+	 * INST_INVOKE_STK1 converting the first arg to
+	 * ::tcl::mathfunc::$objv[0].
+	 */
+
+	objc = TclGetUInt1AtPtr(pc+1);	/* Number of arguments. The function
+					 * name is the 0-th argument. */
+
+	objPtr = OBJ_AT_DEPTH(objc-1);
+	TclNewLiteralStringObj(tmpPtr, "::tcl::mathfunc::");
+	Tcl_AppendObjToObj(tmpPtr, objPtr);
+	Tcl_DecrRefCount(objPtr);
+
+	/*
+	 * Variation of PUSH_OBJECT.
+	 */
+
+	OBJ_AT_DEPTH(objc-1) = tmpPtr;
+	Tcl_IncrRefCount(tmpPtr);
+
+	pcAdjustment = 2;
+	goto doInvocation;
+#else
+    /*
+     * INST_CALL_BUILTIN_FUNC1 and INST_CALL_FUNC1 were made obsolete by the
+     * changes to add a ::tcl::mathfunc namespace in 8.5. Optional support
+     * remains for existing bytecode precompiled files.
+     */
+
+    case INST_CALL_BUILTIN_FUNC1:
+	Tcl_Panic("TclNRExecuteByteCode: obsolete INST_CALL_BUILTIN_FUNC1 found");
+    case INST_CALL_FUNC1:
+	Tcl_Panic("TclNRExecuteByteCode: obsolete INST_CALL_FUNC1 found");
+#endif
+
+    case INST_INVOKE_REPLACE:
+	objc = TclGetUInt4AtPtr(pc+1);
+	opnd = TclGetUInt1AtPtr(pc+5);
+	objPtr = POP_OBJECT();
+	objv = &OBJ_AT_DEPTH(objc-1);
+	cleanup = objc;
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceExec >= 2) {
+	    int i;
+
+	    if (traceInstructions) {
+		strncpy(cmdNameBuf, TclGetString(objv[0]), 20);
+		TRACE(("%u => call (implementation %s) ", objc, O2S(objPtr)));
+	    } else {
+		fprintf(stdout,
+			"%d: (%u) invoking (using implementation %s) ",
+			iPtr->numLevels, (unsigned)(pc - codePtr->codeStart),
+			O2S(objPtr));
+	    }
+	    for (i = 0;  i < objc;  i++) {
+		if (i < opnd) {
+		    fprintf(stdout, "<");
+		    TclPrintObject(stdout, objv[i], 15);
+		    fprintf(stdout, ">");
+		} else {
+		    TclPrintObject(stdout, objv[i], 15);
+		}
+		fprintf(stdout, " ");
+	    }
+	    fprintf(stdout, "\n");
+	    fflush(stdout);
+	}
+#endif /*TCL_COMPILE_DEBUG*/
+
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+	if (iPtr->flags & INTERP_DEBUG_FRAME) {
+	    ArgumentBCEnter(interp, codePtr, TD, pc, objc, objv);
+	}
+
+	TclInitRewriteEnsemble(interp, opnd, 1, objv);
+
+	{
+	    Tcl_Obj *copyPtr = Tcl_NewListObj(objc - opnd + 1, NULL);
+
+	    Tcl_ListObjAppendElement(NULL, copyPtr, objPtr);
+	    Tcl_ListObjReplace(NULL, copyPtr, LIST_MAX, 0,
+		    objc - opnd, objv + opnd);
+	    Tcl_DecrRefCount(objPtr);
+	    objPtr = copyPtr;
+	}
+
+	DECACHE_STACK_INFO();
+	pc += 6;
+	TEBC_YIELD();
+
+	TclMarkTailcall(interp);
+	TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL, NULL);
+	Tcl_ListObjGetElements(NULL, objPtr, &objc, &objv);
+	TclNRAddCallback(interp, TclNRReleaseValues, objPtr, NULL, NULL, NULL);
+	return TclNREvalObjv(interp, objc, objv, TCL_EVAL_INVOKE, NULL);
+
+    /*
+     * -----------------------------------------------------------------
+     *	   Start of INST_LOAD instructions.
+     *
+     * WARNING: more 'goto' here than your doctor recommended! The different
+     * instructions set the value of some variables and then jump to some
+     * common execution code.
+     */
+
+    case INST_LOAD_SCALAR1:
+    instLoadScalar1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u => ", opnd));
+	if (TclIsVarDirectReadable(varPtr)) {
+	    /*
+	     * No errors, no traces: just get the value.
+	     */
+
+	    objResultPtr = varPtr->value.objPtr;
+	    TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(2, 0, 1);
+	}
+	pcAdjustment = 2;
+	cleanup = 0;
+	arrayPtr = NULL;
+	part1Ptr = part2Ptr = NULL;
+	goto doCallPtrGetVar;
+
+    case INST_LOAD_SCALAR4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u => ", opnd));
+	if (TclIsVarDirectReadable(varPtr)) {
+	    /*
+	     * No errors, no traces: just get the value.
+	     */
+
+	    objResultPtr = varPtr->value.objPtr;
+	    TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(5, 0, 1);
+	}
+	pcAdjustment = 5;
+	cleanup = 0;
+	arrayPtr = NULL;
+	part1Ptr = part2Ptr = NULL;
+	goto doCallPtrGetVar;
+
+    case INST_LOAD_ARRAY4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	goto doLoadArray;
+
+    case INST_LOAD_ARRAY1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+
+    doLoadArray:
+	part1Ptr = NULL;
+	part2Ptr = OBJ_AT_TOS;
+	arrayPtr = LOCAL(opnd);
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	TRACE(("%u \"%.30s\" => ", opnd, O2S(part2Ptr)));
+	if (TclIsVarArray(arrayPtr) && !ReadTraced(arrayPtr)) {
+	    varPtr = VarHashFindVar(arrayPtr->value.tablePtr, part2Ptr);
+	    if (varPtr && TclIsVarDirectReadable(varPtr)) {
+		/*
+		 * No errors, no traces: just get the value.
+		 */
+
+		objResultPtr = varPtr->value.objPtr;
+		TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+		NEXT_INST_F(pcAdjustment, 1, 1);
+	    }
+	}
+	varPtr = TclLookupArrayElement(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "read", 0, 1, arrayPtr, opnd);
+	if (varPtr == NULL) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	cleanup = 1;
+	goto doCallPtrGetVar;
+
+    case INST_LOAD_ARRAY_STK:
+	cleanup = 2;
+	part2Ptr = OBJ_AT_TOS;		/* element name */
+	objPtr = OBJ_UNDER_TOS;		/* array name */
+	TRACE(("\"%.30s(%.30s)\" => ", O2S(objPtr), O2S(part2Ptr)));
+	goto doLoadStk;
+
+    case INST_LOAD_STK:
+    case INST_LOAD_SCALAR_STK:
+	cleanup = 1;
+	part2Ptr = NULL;
+	objPtr = OBJ_AT_TOS;		/* variable name */
+	TRACE(("\"%.30s\" => ", O2S(objPtr)));
+
+    doLoadStk:
+	part1Ptr = objPtr;
+	varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "read", /*createPart1*/0, /*createPart2*/1,
+		&arrayPtr);
+	if (!varPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	if (TclIsVarDirectReadable2(varPtr, arrayPtr)) {
+	    /*
+	     * No errors, no traces: just get the value.
+	     */
+
+	    objResultPtr = varPtr->value.objPtr;
+	    TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	    NEXT_INST_V(1, cleanup, 1);
+	}
+	pcAdjustment = 1;
+	opnd = -1;
+
+    doCallPtrGetVar:
+	/*
+	 * There are either errors or the variable is traced: call
+	 * TclPtrGetVar to process fully.
+	 */
+
+	DECACHE_STACK_INFO();
+	objResultPtr = TclPtrGetVarIdx(interp, varPtr, arrayPtr,
+		part1Ptr, part2Ptr, TCL_LEAVE_ERR_MSG, opnd);
+	CACHE_STACK_INFO();
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+
+    /*
+     *	   End of INST_LOAD instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_STORE and related instructions.
+     *
+     * WARNING: more 'goto' here than your doctor recommended! The different
+     * instructions set the value of some variables and then jump to somme
+     * common execution code.
+     */
+
+    {
+	int storeFlags, len;
+
+    case INST_STORE_ARRAY4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	goto doStoreArrayDirect;
+
+    case INST_STORE_ARRAY1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+
+    doStoreArrayDirect:
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = OBJ_UNDER_TOS;
+	arrayPtr = LOCAL(opnd);
+	TRACE(("%u \"%.30s\" <- \"%.30s\" => ", opnd, O2S(part2Ptr),
+		O2S(valuePtr)));
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	if (TclIsVarArray(arrayPtr) && !WriteTraced(arrayPtr)) {
+	    varPtr = VarHashFindVar(arrayPtr->value.tablePtr, part2Ptr);
+	    if (varPtr && TclIsVarDirectWritable(varPtr)) {
+		tosPtr--;
+		Tcl_DecrRefCount(OBJ_AT_TOS);
+		OBJ_AT_TOS = valuePtr;
+		goto doStoreVarDirect;
+	    }
+	}
+	cleanup = 2;
+	storeFlags = TCL_LEAVE_ERR_MSG;
+	part1Ptr = NULL;
+	goto doStoreArrayDirectFailed;
+
+    case INST_STORE_SCALAR4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	goto doStoreScalarDirect;
+
+    case INST_STORE_SCALAR1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+
+    doStoreScalarDirect:
+	valuePtr = OBJ_AT_TOS;
+	varPtr = LOCAL(opnd);
+	TRACE(("%u <- \"%.30s\" => ", opnd, O2S(valuePtr)));
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	if (!TclIsVarDirectWritable(varPtr)) {
+	    storeFlags = TCL_LEAVE_ERR_MSG;
+	    part1Ptr = NULL;
+	    goto doStoreScalar;
+	}
+
+	/*
+	 * No traces, no errors, plain 'set': we can safely inline. The value
+	 * *will* be set to what's requested, so that the stack top remains
+	 * pointing to the same Tcl_Obj.
+	 */
+
+    doStoreVarDirect:
+	valuePtr = varPtr->value.objPtr;
+	if (valuePtr != NULL) {
+	    TclDecrRefCount(valuePtr);
+	}
+	objResultPtr = OBJ_AT_TOS;
+	varPtr->value.objPtr = objResultPtr;
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+pcAdjustment) == INST_POP) {
+	    tosPtr--;
+	    NEXT_INST_F((pcAdjustment+1), 0, 0);
+	}
+#else
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+#endif
+	Tcl_IncrRefCount(objResultPtr);
+	NEXT_INST_F(pcAdjustment, 0, 0);
+
+    case INST_LAPPEND_STK:
+	valuePtr = OBJ_AT_TOS; /* value to append */
+	part2Ptr = NULL;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreStk;
+
+    case INST_LAPPEND_ARRAY_STK:
+	valuePtr = OBJ_AT_TOS; /* value to append */
+	part2Ptr = OBJ_UNDER_TOS;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreStk;
+
+    case INST_APPEND_STK:
+	valuePtr = OBJ_AT_TOS; /* value to append */
+	part2Ptr = NULL;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreStk;
+
+    case INST_APPEND_ARRAY_STK:
+	valuePtr = OBJ_AT_TOS; /* value to append */
+	part2Ptr = OBJ_UNDER_TOS;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreStk;
+
+    case INST_STORE_ARRAY_STK:
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = OBJ_UNDER_TOS;
+	storeFlags = TCL_LEAVE_ERR_MSG;
+	goto doStoreStk;
+
+    case INST_STORE_STK:
+    case INST_STORE_SCALAR_STK:
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = NULL;
+	storeFlags = TCL_LEAVE_ERR_MSG;
+
+    doStoreStk:
+	objPtr = OBJ_AT_DEPTH(1 + (part2Ptr != NULL)); /* variable name */
+	part1Ptr = objPtr;
+#ifdef TCL_COMPILE_DEBUG
+	if (part2Ptr == NULL) {
+	    TRACE(("\"%.30s\" <- \"%.30s\" =>", O2S(part1Ptr),O2S(valuePtr)));
+	} else {
+	    TRACE(("\"%.30s(%.30s)\" <- \"%.30s\" => ",
+		    O2S(part1Ptr), O2S(part2Ptr), O2S(valuePtr)));
+	}
+#endif
+	varPtr = TclObjLookupVarEx(interp, objPtr,part2Ptr, TCL_LEAVE_ERR_MSG,
+		"set", /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+	if (!varPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	cleanup = ((part2Ptr == NULL)? 2 : 3);
+	pcAdjustment = 1;
+	opnd = -1;
+	goto doCallPtrSetVar;
+
+    case INST_LAPPEND_ARRAY4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreArray;
+
+    case INST_LAPPEND_ARRAY1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreArray;
+
+    case INST_APPEND_ARRAY4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreArray;
+
+    case INST_APPEND_ARRAY1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreArray;
+
+    doStoreArray:
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = OBJ_UNDER_TOS;
+	arrayPtr = LOCAL(opnd);
+	TRACE(("%u \"%.30s\" <- \"%.30s\" => ", opnd, O2S(part2Ptr),
+		O2S(valuePtr)));
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	cleanup = 2;
+	part1Ptr = NULL;
+
+    doStoreArrayDirectFailed:
+	varPtr = TclLookupArrayElement(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "set", 1, 1, arrayPtr, opnd);
+	if (!varPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	goto doCallPtrSetVar;
+
+    case INST_LAPPEND_SCALAR4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreScalar;
+
+    case INST_LAPPEND_SCALAR1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE
+		| TCL_LIST_ELEMENT);
+	goto doStoreScalar;
+
+    case INST_APPEND_SCALAR4:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreScalar;
+
+    case INST_APPEND_SCALAR1:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	pcAdjustment = 2;
+	storeFlags = (TCL_LEAVE_ERR_MSG | TCL_APPEND_VALUE);
+	goto doStoreScalar;
+
+    doStoreScalar:
+	valuePtr = OBJ_AT_TOS;
+	varPtr = LOCAL(opnd);
+	TRACE(("%u <- \"%.30s\" => ", opnd, O2S(valuePtr)));
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	cleanup = 1;
+	arrayPtr = NULL;
+	part1Ptr = part2Ptr = NULL;
+
+    doCallPtrSetVar:
+	DECACHE_STACK_INFO();
+	objResultPtr = TclPtrSetVarIdx(interp, varPtr, arrayPtr,
+		part1Ptr, part2Ptr, valuePtr, storeFlags, opnd);
+	CACHE_STACK_INFO();
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+pcAdjustment) == INST_POP) {
+	    NEXT_INST_V((pcAdjustment+1), cleanup, 0);
+	}
+#endif
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+
+    case INST_LAPPEND_LIST:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	valuePtr = OBJ_AT_TOS;
+	varPtr = LOCAL(opnd);
+	cleanup = 1;
+	pcAdjustment = 5;
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u <- \"%.30s\" => ", opnd, O2S(valuePtr)));
+	if (TclListObjGetElements(interp, valuePtr, &objc, &objv)
+		!= TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	if (TclIsVarDirectReadable(varPtr)
+		&& TclIsVarDirectWritable(varPtr)) {
+	    goto lappendListDirect;
+	}
+	arrayPtr = NULL;
+	part1Ptr = part2Ptr = NULL;
+	goto lappendListPtr;
+
+    case INST_LAPPEND_LIST_ARRAY:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	valuePtr = OBJ_AT_TOS;
+	part1Ptr = NULL;
+	part2Ptr = OBJ_UNDER_TOS;
+	arrayPtr = LOCAL(opnd);
+	cleanup = 2;
+	pcAdjustment = 5;
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	TRACE(("%u \"%.30s\" \"%.30s\" => ",
+		opnd, O2S(part2Ptr), O2S(valuePtr)));
+	if (TclListObjGetElements(interp, valuePtr, &objc, &objv)
+		!= TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	if (TclIsVarArray(arrayPtr) && !ReadTraced(arrayPtr)
+		&& !WriteTraced(arrayPtr)) {
+	    varPtr = VarHashFindVar(arrayPtr->value.tablePtr, part2Ptr);
+	    if (varPtr && TclIsVarDirectReadable(varPtr)
+		    && TclIsVarDirectWritable(varPtr)) {
+		goto lappendListDirect;
+	    }
+	}
+	varPtr = TclLookupArrayElement(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "set", 1, 1, arrayPtr, opnd);
+	if (varPtr == NULL) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	goto lappendListPtr;
+
+    case INST_LAPPEND_LIST_ARRAY_STK:
+	pcAdjustment = 1;
+	cleanup = 3;
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = OBJ_UNDER_TOS;	/* element name */
+	part1Ptr = OBJ_AT_DEPTH(2);	/* array name */
+	TRACE(("\"%.30s(%.30s)\" \"%.30s\" => ",
+		O2S(part1Ptr), O2S(part2Ptr), O2S(valuePtr)));
+	goto lappendList;
+
+    case INST_LAPPEND_LIST_STK:
+	pcAdjustment = 1;
+	cleanup = 2;
+	valuePtr = OBJ_AT_TOS;
+	part2Ptr = NULL;
+	part1Ptr = OBJ_UNDER_TOS;	/* variable name */
+	TRACE(("\"%.30s\" \"%.30s\" => ", O2S(part1Ptr), O2S(valuePtr)));
+	goto lappendList;
+
+    lappendListDirect:
+	objResultPtr = varPtr->value.objPtr;
+	if (TclListObjLength(interp, objResultPtr, &len) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	if (Tcl_IsShared(objResultPtr)) {
+	    Tcl_Obj *newValue = Tcl_DuplicateObj(objResultPtr);
+
+	    TclDecrRefCount(objResultPtr);
+	    varPtr->value.objPtr = objResultPtr = newValue;
+	    Tcl_IncrRefCount(newValue);
+	}
+	if (Tcl_ListObjReplace(interp, objResultPtr, len, 0, objc, objv)
+		!= TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+
+    lappendList:
+	opnd = -1;
+	if (TclListObjGetElements(interp, valuePtr, &objc, &objv)
+		!= TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	DECACHE_STACK_INFO();
+	varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "set", 1, 1, &arrayPtr);
+	CACHE_STACK_INFO();
+	if (!varPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+    lappendListPtr:
+	if (TclIsVarInHash(varPtr)) {
+	    VarHashRefCount(varPtr)++;
+	}
+	if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	    VarHashRefCount(arrayPtr)++;
+	}
+	DECACHE_STACK_INFO();
+	objResultPtr = TclPtrGetVarIdx(interp, varPtr, arrayPtr,
+		part1Ptr, part2Ptr, TCL_LEAVE_ERR_MSG, opnd);
+	CACHE_STACK_INFO();
+	if (TclIsVarInHash(varPtr)) {
+	    VarHashRefCount(varPtr)--;
+	}
+	if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	    VarHashRefCount(arrayPtr)--;
+	}
+
+	{
+	    int createdNewObj = 0;
+
+	    if (!objResultPtr) {
+		objResultPtr = valuePtr;
+	    } else if (TclListObjLength(interp, objResultPtr, &len)!=TCL_OK) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    } else {
+		if (Tcl_IsShared(objResultPtr)) {
+		    objResultPtr = Tcl_DuplicateObj(objResultPtr);
+		    createdNewObj = 1;
+		}
+		if (Tcl_ListObjReplace(interp, objResultPtr, len,0, objc,objv)
+			!= TCL_OK) {
+		    goto errorInLappendListPtr;
+		}
+	    }
+	    DECACHE_STACK_INFO();
+	    objResultPtr = TclPtrSetVarIdx(interp, varPtr, arrayPtr, part1Ptr,
+		    part2Ptr, objResultPtr, TCL_LEAVE_ERR_MSG, opnd);
+	    CACHE_STACK_INFO();
+	    if (!objResultPtr) {
+	    errorInLappendListPtr:
+		if (createdNewObj) {
+		    TclDecrRefCount(objResultPtr);
+		}
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+    }
+
+    /*
+     *	   End of INST_STORE and related instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_INCR instructions.
+     *
+     * WARNING: more 'goto' here than your doctor recommended! The different
+     * instructions set the value of some variables and then jump to somme
+     * common execution code.
+     */
+
+/*TODO: Consider more untangling here; merge with LOAD and STORE ? */
+
+    {
+	Tcl_Obj *incrPtr;
+#ifndef TCL_WIDE_INT_IS_LONG
+	Tcl_WideInt w;
+#endif
+	long increment;
+
+    case INST_INCR_SCALAR1:
+    case INST_INCR_ARRAY1:
+    case INST_INCR_ARRAY_STK:
+    case INST_INCR_SCALAR_STK:
+    case INST_INCR_STK:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	incrPtr = POP_OBJECT();
+	switch (*pc) {
+	case INST_INCR_SCALAR1:
+	    pcAdjustment = 2;
+	    goto doIncrScalar;
+	case INST_INCR_ARRAY1:
+	    pcAdjustment = 2;
+	    goto doIncrArray;
+	default:
+	    pcAdjustment = 1;
+	    goto doIncrStk;
+	}
+
+    case INST_INCR_ARRAY_STK_IMM:
+    case INST_INCR_SCALAR_STK_IMM:
+    case INST_INCR_STK_IMM:
+	increment = TclGetInt1AtPtr(pc+1);
+	incrPtr = Tcl_NewIntObj(increment);
+	Tcl_IncrRefCount(incrPtr);
+	pcAdjustment = 2;
+
+    doIncrStk:
+	if ((*pc == INST_INCR_ARRAY_STK_IMM)
+		|| (*pc == INST_INCR_ARRAY_STK)) {
+	    part2Ptr = OBJ_AT_TOS;
+	    objPtr = OBJ_UNDER_TOS;
+	    TRACE(("\"%.30s(%.30s)\" (by %ld) => ",
+		    O2S(objPtr), O2S(part2Ptr), increment));
+	} else {
+	    part2Ptr = NULL;
+	    objPtr = OBJ_AT_TOS;
+	    TRACE(("\"%.30s\" (by %ld) => ", O2S(objPtr), increment));
+	}
+	part1Ptr = objPtr;
+	opnd = -1;
+	varPtr = TclObjLookupVarEx(interp, objPtr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "read", 1, 1, &arrayPtr);
+	if (!varPtr) {
+	    DECACHE_STACK_INFO();
+	    Tcl_AddErrorInfo(interp,
+		    "\n    (reading value of variable to increment)");
+	    CACHE_STACK_INFO();
+	    TRACE_ERROR(interp);
+	    Tcl_DecrRefCount(incrPtr);
+	    goto gotError;
+	}
+	cleanup = ((part2Ptr == NULL)? 1 : 2);
+	goto doIncrVar;
+
+    case INST_INCR_ARRAY1_IMM:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	increment = TclGetInt1AtPtr(pc+2);
+	incrPtr = Tcl_NewIntObj(increment);
+	Tcl_IncrRefCount(incrPtr);
+	pcAdjustment = 3;
+
+    doIncrArray:
+	part1Ptr = NULL;
+	part2Ptr = OBJ_AT_TOS;
+	arrayPtr = LOCAL(opnd);
+	cleanup = 1;
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	TRACE(("%u \"%.30s\" (by %ld) => ", opnd, O2S(part2Ptr), increment));
+	varPtr = TclLookupArrayElement(interp, part1Ptr, part2Ptr,
+		TCL_LEAVE_ERR_MSG, "read", 1, 1, arrayPtr, opnd);
+	if (!varPtr) {
+	    TRACE_ERROR(interp);
+	    Tcl_DecrRefCount(incrPtr);
+	    goto gotError;
+	}
+	goto doIncrVar;
+
+    case INST_INCR_SCALAR1_IMM:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	increment = TclGetInt1AtPtr(pc+2);
+	pcAdjustment = 3;
+	cleanup = 0;
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+
+	if (TclIsVarDirectModifyable(varPtr)) {
+	    ClientData ptr;
+	    int type;
+
+	    objPtr = varPtr->value.objPtr;
+	    if (GetNumberFromObj(NULL, objPtr, &ptr, &type) == TCL_OK) {
+		if (type == TCL_NUMBER_LONG) {
+		    long augend = *((const long *)ptr);
+		    long sum = augend + increment;
+
+		    /*
+		     * Overflow when (augend and sum have different sign) and
+		     * (augend and increment have the same sign). This is
+		     * encapsulated in the Overflowing macro.
+		     */
+
+		    if (!Overflowing(augend, increment, sum)) {
+			TRACE(("%u %ld => ", opnd, increment));
+			if (Tcl_IsShared(objPtr)) {
+			    objPtr->refCount--;	/* We know it's shared. */
+			    TclNewLongObj(objResultPtr, sum);
+			    Tcl_IncrRefCount(objResultPtr);
+			    varPtr->value.objPtr = objResultPtr;
+			} else {
+			    objResultPtr = objPtr;
+			    TclSetLongObj(objPtr, sum);
+			}
+			goto doneIncr;
+		    }
+#ifndef TCL_WIDE_INT_IS_LONG
+		    w = (Tcl_WideInt)augend;
+
+		    TRACE(("%u %ld => ", opnd, increment));
+		    if (Tcl_IsShared(objPtr)) {
+			objPtr->refCount--;	/* We know it's shared. */
+			objResultPtr = Tcl_NewWideIntObj(w+increment);
+			Tcl_IncrRefCount(objResultPtr);
+			varPtr->value.objPtr = objResultPtr;
+		    } else {
+			objResultPtr = objPtr;
+
+			/*
+			 * We know the sum value is outside the long range;
+			 * use macro form that doesn't range test again.
+			 */
+
+			TclSetWideIntObj(objPtr, w+increment);
+		    }
+		    goto doneIncr;
+#endif
+		}	/* end if (type == TCL_NUMBER_LONG) */
+#ifndef TCL_WIDE_INT_IS_LONG
+		if (type == TCL_NUMBER_WIDE) {
+		    Tcl_WideInt sum;
+
+		    w = *((const Tcl_WideInt *) ptr);
+		    sum = w + increment;
+
+		    /*
+		     * Check for overflow.
+		     */
+
+		    if (!Overflowing(w, increment, sum)) {
+			TRACE(("%u %ld => ", opnd, increment));
+			if (Tcl_IsShared(objPtr)) {
+			    objPtr->refCount--;	/* We know it's shared. */
+			    objResultPtr = Tcl_NewWideIntObj(sum);
+			    Tcl_IncrRefCount(objResultPtr);
+			    varPtr->value.objPtr = objResultPtr;
+			} else {
+			    objResultPtr = objPtr;
+
+			    /*
+			     * We *do not* know the sum value is outside the
+			     * long range (wide + long can yield long); use
+			     * the function call that checks range.
+			     */
+
+			    Tcl_SetWideIntObj(objPtr, sum);
+			}
+			goto doneIncr;
+		    }
+		}
+#endif
+	    }
+	    if (Tcl_IsShared(objPtr)) {
+		objPtr->refCount--;	/* We know it's shared */
+		objResultPtr = Tcl_DuplicateObj(objPtr);
+		Tcl_IncrRefCount(objResultPtr);
+		varPtr->value.objPtr = objResultPtr;
+	    } else {
+		objResultPtr = objPtr;
+	    }
+	    TclNewLongObj(incrPtr, increment);
+	    if (TclIncrObj(interp, objResultPtr, incrPtr) != TCL_OK) {
+		Tcl_DecrRefCount(incrPtr);
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    Tcl_DecrRefCount(incrPtr);
+	    goto doneIncr;
+	}
+
+	/*
+	 * All other cases, flow through to generic handling.
+	 */
+
+	TclNewLongObj(incrPtr, increment);
+	Tcl_IncrRefCount(incrPtr);
+
+    doIncrScalar:
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	arrayPtr = NULL;
+	part1Ptr = part2Ptr = NULL;
+	cleanup = 0;
+	TRACE(("%u %s => ", opnd, Tcl_GetString(incrPtr)));
+
+    doIncrVar:
+	if (TclIsVarDirectModifyable2(varPtr, arrayPtr)) {
+	    objPtr = varPtr->value.objPtr;
+	    if (Tcl_IsShared(objPtr)) {
+		objPtr->refCount--;	/* We know it's shared */
+		objResultPtr = Tcl_DuplicateObj(objPtr);
+		Tcl_IncrRefCount(objResultPtr);
+		varPtr->value.objPtr = objResultPtr;
+	    } else {
+		objResultPtr = objPtr;
+	    }
+	    if (TclIncrObj(interp, objResultPtr, incrPtr) != TCL_OK) {
+		Tcl_DecrRefCount(incrPtr);
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    Tcl_DecrRefCount(incrPtr);
+	} else {
+	    DECACHE_STACK_INFO();
+	    objResultPtr = TclPtrIncrObjVarIdx(interp, varPtr, arrayPtr,
+		    part1Ptr, part2Ptr, incrPtr, TCL_LEAVE_ERR_MSG, opnd);
+	    CACHE_STACK_INFO();
+	    Tcl_DecrRefCount(incrPtr);
+	    if (objResultPtr == NULL) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+    doneIncr:
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+pcAdjustment) == INST_POP) {
+	    NEXT_INST_V((pcAdjustment+1), cleanup, 0);
+	}
+#endif
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+    }
+
+    /*
+     *	   End of INST_INCR instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_EXIST instructions.
+     */
+
+    case INST_EXIST_SCALAR:
+	cleanup = 0;
+	pcAdjustment = 5;
+	opnd = TclGetUInt4AtPtr(pc+1);
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u => ", opnd));
+	if (ReadTraced(varPtr)) {
+	    DECACHE_STACK_INFO();
+	    TclObjCallVarTraces(iPtr, NULL, varPtr, NULL, NULL,
+		    TCL_TRACE_READS, 0, opnd);
+	    CACHE_STACK_INFO();
+	    if (TclIsVarUndefined(varPtr)) {
+		TclCleanupVar(varPtr, NULL);
+		varPtr = NULL;
+	    }
+	}
+	goto afterExistsPeephole;
+
+    case INST_EXIST_ARRAY:
+	cleanup = 1;
+	pcAdjustment = 5;
+	opnd = TclGetUInt4AtPtr(pc+1);
+	part2Ptr = OBJ_AT_TOS;
+	arrayPtr = LOCAL(opnd);
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	TRACE(("%u \"%.30s\" => ", opnd, O2S(part2Ptr)));
+	if (TclIsVarArray(arrayPtr) && !ReadTraced(arrayPtr)) {
+	    varPtr = VarHashFindVar(arrayPtr->value.tablePtr, part2Ptr);
+	    if (!varPtr || !ReadTraced(varPtr)) {
+		goto afterExistsPeephole;
+	    }
+	}
+	varPtr = TclLookupArrayElement(interp, NULL, part2Ptr, 0, "access",
+		0, 1, arrayPtr, opnd);
+	if (varPtr) {
+	    if (ReadTraced(varPtr) || (arrayPtr && ReadTraced(arrayPtr))) {
+		DECACHE_STACK_INFO();
+		TclObjCallVarTraces(iPtr, arrayPtr, varPtr, NULL, part2Ptr,
+			TCL_TRACE_READS, 0, opnd);
+		CACHE_STACK_INFO();
+	    }
+	    if (TclIsVarUndefined(varPtr)) {
+		TclCleanupVar(varPtr, arrayPtr);
+		varPtr = NULL;
+	    }
+	}
+	goto afterExistsPeephole;
+
+    case INST_EXIST_ARRAY_STK:
+	cleanup = 2;
+	pcAdjustment = 1;
+	part2Ptr = OBJ_AT_TOS;		/* element name */
+	part1Ptr = OBJ_UNDER_TOS;	/* array name */
+	TRACE(("\"%.30s(%.30s)\" => ", O2S(part1Ptr), O2S(part2Ptr)));
+	goto doExistStk;
+
+    case INST_EXIST_STK:
+	cleanup = 1;
+	pcAdjustment = 1;
+	part2Ptr = NULL;
+	part1Ptr = OBJ_AT_TOS;		/* variable name */
+	TRACE(("\"%.30s\" => ", O2S(part1Ptr)));
+
+    doExistStk:
+	varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr, 0, "access",
+		/*createPart1*/0, /*createPart2*/1, &arrayPtr);
+	if (varPtr) {
+	    if (ReadTraced(varPtr) || (arrayPtr && ReadTraced(arrayPtr))) {
+		DECACHE_STACK_INFO();
+		TclObjCallVarTraces(iPtr, arrayPtr, varPtr, part1Ptr,part2Ptr,
+			TCL_TRACE_READS, 0, -1);
+		CACHE_STACK_INFO();
+	    }
+	    if (TclIsVarUndefined(varPtr)) {
+		TclCleanupVar(varPtr, arrayPtr);
+		varPtr = NULL;
+	    }
+	}
+
+	/*
+	 * Peep-hole optimisation: if you're about to jump, do jump from here.
+	 */
+
+    afterExistsPeephole: {
+	int found = (varPtr && !TclIsVarUndefined(varPtr));
+
+	TRACE_APPEND(("%d\n", found ? 1 : 0));
+	JUMP_PEEPHOLE_V(found, pcAdjustment, cleanup);
+    }
+
+    /*
+     *	   End of INST_EXIST instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_UNSET instructions.
+     */
+
+    {
+	int flags;
+
+    case INST_UNSET_SCALAR:
+	flags = TclGetUInt1AtPtr(pc+1) ? TCL_LEAVE_ERR_MSG : 0;
+	opnd = TclGetUInt4AtPtr(pc+2);
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%s %u => ", (flags ? "normal" : "noerr"), opnd));
+	if (TclIsVarDirectUnsettable(varPtr) && !TclIsVarInHash(varPtr)) {
+	    /*
+	     * No errors, no traces, no searches: just make the variable cease
+	     * to exist.
+	     */
+
+	    if (!TclIsVarUndefined(varPtr)) {
+		TclDecrRefCount(varPtr->value.objPtr);
+	    } else if (flags & TCL_LEAVE_ERR_MSG) {
+		goto slowUnsetScalar;
+	    }
+	    varPtr->value.objPtr = NULL;
+	    TRACE_APPEND(("OK\n"));
+	    NEXT_INST_F(6, 0, 0);
+	}
+
+    slowUnsetScalar:
+	DECACHE_STACK_INFO();
+	if (TclPtrUnsetVarIdx(interp, varPtr, NULL, NULL, NULL, flags,
+		opnd) != TCL_OK && flags) {
+	    goto errorInUnset;
+	}
+	CACHE_STACK_INFO();
+	NEXT_INST_F(6, 0, 0);
+
+    case INST_UNSET_ARRAY:
+	flags = TclGetUInt1AtPtr(pc+1) ? TCL_LEAVE_ERR_MSG : 0;
+	opnd = TclGetUInt4AtPtr(pc+2);
+	part2Ptr = OBJ_AT_TOS;
+	arrayPtr = LOCAL(opnd);
+	while (TclIsVarLink(arrayPtr)) {
+	    arrayPtr = arrayPtr->value.linkPtr;
+	}
+	TRACE(("%s %u \"%.30s\" => ",
+		(flags ? "normal" : "noerr"), opnd, O2S(part2Ptr)));
+	if (TclIsVarArray(arrayPtr) && !UnsetTraced(arrayPtr)) {
+	    varPtr = VarHashFindVar(arrayPtr->value.tablePtr, part2Ptr);
+	    if (varPtr && TclIsVarDirectUnsettable(varPtr)) {
+		/*
+		 * No nasty traces and element exists, so we can proceed to
+		 * unset it. Might still not exist though...
+		 */
+
+		if (!TclIsVarUndefined(varPtr)) {
+		    TclDecrRefCount(varPtr->value.objPtr);
+		    TclSetVarUndefined(varPtr);
+		    TclClearVarNamespaceVar(varPtr);
+		    TclCleanupVar(varPtr, arrayPtr);
+		} else if (flags & TCL_LEAVE_ERR_MSG) {
+		    goto slowUnsetArray;
+		}
+		TRACE_APPEND(("OK\n"));
+		NEXT_INST_F(6, 1, 0);
+	    } else if (!varPtr && !(flags & TCL_LEAVE_ERR_MSG)) {
+		/*
+		 * Don't need to do anything here.
+		 */
+
+		TRACE_APPEND(("OK\n"));
+		NEXT_INST_F(6, 1, 0);
+	    }
+	}
+    slowUnsetArray:
+	DECACHE_STACK_INFO();
+	varPtr = TclLookupArrayElement(interp, NULL, part2Ptr, flags, "unset",
+		0, 0, arrayPtr, opnd);
+	if (!varPtr) {
+	    if (flags & TCL_LEAVE_ERR_MSG) {
+		goto errorInUnset;
+	    }
+	} else if (TclPtrUnsetVarIdx(interp, varPtr, arrayPtr, NULL, part2Ptr,
+		flags, opnd) != TCL_OK && (flags & TCL_LEAVE_ERR_MSG)) {
+	    goto errorInUnset;
+	}
+	CACHE_STACK_INFO();
+	NEXT_INST_F(6, 1, 0);
+
+    case INST_UNSET_ARRAY_STK:
+	flags = TclGetUInt1AtPtr(pc+1) ? TCL_LEAVE_ERR_MSG : 0;
+	cleanup = 2;
+	part2Ptr = OBJ_AT_TOS;		/* element name */
+	part1Ptr = OBJ_UNDER_TOS;	/* array name */
+	TRACE(("%s \"%.30s(%.30s)\" => ", (flags ? "normal" : "noerr"),
+		O2S(part1Ptr), O2S(part2Ptr)));
+	goto doUnsetStk;
+
+    case INST_UNSET_STK:
+	flags = TclGetUInt1AtPtr(pc+1) ? TCL_LEAVE_ERR_MSG : 0;
+	cleanup = 1;
+	part2Ptr = NULL;
+	part1Ptr = OBJ_AT_TOS;		/* variable name */
+	TRACE(("%s \"%.30s\" => ", (flags ? "normal" : "noerr"),
+		O2S(part1Ptr)));
+
+    doUnsetStk:
+	DECACHE_STACK_INFO();
+	if (TclObjUnsetVar2(interp, part1Ptr, part2Ptr, flags) != TCL_OK
+		&& (flags & TCL_LEAVE_ERR_MSG)) {
+	    goto errorInUnset;
+	}
+	CACHE_STACK_INFO();
+	TRACE_APPEND(("OK\n"));
+	NEXT_INST_V(2, cleanup, 0);
+
+    errorInUnset:
+	CACHE_STACK_INFO();
+	TRACE_ERROR(interp);
+	goto gotError;
+
+	/*
+	 * This is really an unset operation these days. Do not issue.
+	 */
+
+    case INST_DICT_DONE:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	TRACE(("%u => OK\n", opnd));
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	if (TclIsVarDirectUnsettable(varPtr) && !TclIsVarInHash(varPtr)) {
+	    if (!TclIsVarUndefined(varPtr)) {
+		TclDecrRefCount(varPtr->value.objPtr);
+	    }
+	    varPtr->value.objPtr = NULL;
+	} else {
+	    DECACHE_STACK_INFO();
+	    TclPtrUnsetVarIdx(interp, varPtr, NULL, NULL, NULL, 0, opnd);
+	    CACHE_STACK_INFO();
+	}
+	NEXT_INST_F(5, 0, 0);
+    }
+
+    /*
+     *	   End of INST_UNSET instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_ARRAY instructions.
+     */
+
+    case INST_ARRAY_EXISTS_IMM:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	cleanup = 0;
+	part1Ptr = NULL;
+	arrayPtr = NULL;
+	TRACE(("%u => ", opnd));
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	goto doArrayExists;
+    case INST_ARRAY_EXISTS_STK:
+	opnd = -1;
+	pcAdjustment = 1;
+	cleanup = 1;
+	part1Ptr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" => ", O2S(part1Ptr)));
+	varPtr = TclObjLookupVarEx(interp, part1Ptr, NULL, 0, NULL,
+		/*createPart1*/0, /*createPart2*/0, &arrayPtr);
+    doArrayExists:
+	if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+		&& (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	    DECACHE_STACK_INFO();
+	    result = TclObjCallVarTraces(iPtr, arrayPtr, varPtr, part1Ptr,
+		    NULL, (TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|
+		    TCL_GLOBAL_ONLY|TCL_TRACE_ARRAY), 1, opnd);
+	    CACHE_STACK_INFO();
+	    if (result == TCL_ERROR) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+	if (varPtr && TclIsVarArray(varPtr) && !TclIsVarUndefined(varPtr)) {
+	    objResultPtr = TCONST(1);
+	} else {
+	    objResultPtr = TCONST(0);
+	}
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_V(pcAdjustment, cleanup, 1);
+
+    case INST_ARRAY_MAKE_IMM:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	pcAdjustment = 5;
+	cleanup = 0;
+	part1Ptr = NULL;
+	arrayPtr = NULL;
+	TRACE(("%u => ", opnd));
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	goto doArrayMake;
+    case INST_ARRAY_MAKE_STK:
+	opnd = -1;
+	pcAdjustment = 1;
+	cleanup = 1;
+	part1Ptr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" => ", O2S(part1Ptr)));
+	varPtr = TclObjLookupVarEx(interp, part1Ptr, NULL, TCL_LEAVE_ERR_MSG,
+		"set", /*createPart1*/1, /*createPart2*/0, &arrayPtr);
+	if (varPtr == NULL) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+    doArrayMake:
+	if (varPtr && !TclIsVarArray(varPtr)) {
+	    if (TclIsVarArrayElement(varPtr) || !TclIsVarUndefined(varPtr)) {
+		/*
+		 * Either an array element, or a scalar: lose!
+		 */
+
+		TclObjVarErrMsg(interp, part1Ptr, NULL, "array set",
+			"variable isn't array", opnd);
+		DECACHE_STACK_INFO();
+		Tcl_SetErrorCode(interp, "TCL", "WRITE", "ARRAY", NULL);
+		CACHE_STACK_INFO();
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    TclSetVarArray(varPtr);
+	    varPtr->value.tablePtr = ckalloc(sizeof(TclVarHashTable));
+	    TclInitVarHashTable(varPtr->value.tablePtr,
+		    TclGetVarNsPtr(varPtr));
+#ifdef TCL_COMPILE_DEBUG
+	    TRACE_APPEND(("done\n"));
+	} else {
+	    TRACE_APPEND(("nothing to do\n"));
+#endif
+	}
+	NEXT_INST_V(pcAdjustment, cleanup, 0);
+
+    /*
+     *	   End of INST_ARRAY instructions.
+     * -----------------------------------------------------------------
+     *	   Start of variable linking instructions.
+     */
+
+    {
+	Var *otherPtr;
+	CallFrame *framePtr, *savedFramePtr;
+	Tcl_Namespace *nsPtr;
+	Namespace *savedNsPtr;
+
+    case INST_UPVAR:
+	TRACE(("%d %.30s %.30s => ", TclGetInt4AtPtr(pc+1),
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS)));
+
+	if (TclObjGetFrame(interp, OBJ_UNDER_TOS, &framePtr) == -1) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Locate the other variable.
+	 */
+
+	savedFramePtr = iPtr->varFramePtr;
+	iPtr->varFramePtr = framePtr;
+	otherPtr = TclObjLookupVarEx(interp, OBJ_AT_TOS, NULL,
+		TCL_LEAVE_ERR_MSG, "access", /*createPart1*/ 1,
+		/*createPart2*/ 1, &varPtr);
+	iPtr->varFramePtr = savedFramePtr;
+	if (!otherPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	goto doLinkVars;
+
+    case INST_NSUPVAR:
+	TRACE(("%d %.30s %.30s => ", TclGetInt4AtPtr(pc+1),
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS)));
+	if (TclGetNamespaceFromObj(interp, OBJ_UNDER_TOS, &nsPtr) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Locate the other variable.
+	 */
+
+	savedNsPtr = iPtr->varFramePtr->nsPtr;
+	iPtr->varFramePtr->nsPtr = (Namespace *) nsPtr;
+	otherPtr = TclObjLookupVarEx(interp, OBJ_AT_TOS, NULL,
+		(TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG|TCL_AVOID_RESOLVERS),
+		"access", /*createPart1*/ 1, /*createPart2*/ 1, &varPtr);
+	iPtr->varFramePtr->nsPtr = savedNsPtr;
+	if (!otherPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	goto doLinkVars;
+
+    case INST_VARIABLE:
+	TRACE(("%d, %.30s => ", TclGetInt4AtPtr(pc+1), O2S(OBJ_AT_TOS)));
+	otherPtr = TclObjLookupVarEx(interp, OBJ_AT_TOS, NULL,
+		(TCL_NAMESPACE_ONLY | TCL_LEAVE_ERR_MSG), "access",
+		/*createPart1*/ 1, /*createPart2*/ 1, &varPtr);
+	if (!otherPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Do the [variable] magic.
+	 */
+
+	TclSetVarNamespaceVar(otherPtr);
+
+    doLinkVars:
+
+	/*
+	 * If we are here, the local variable has already been created: do the
+	 * little work of TclPtrMakeUpvar that remains to be done right here
+	 * if there are no errors; otherwise, let it handle the case.
+	 */
+
+	opnd = TclGetInt4AtPtr(pc+1);
+	varPtr = LOCAL(opnd);
+	if ((varPtr != otherPtr) && !TclIsVarTraced(varPtr)
+		&& (TclIsVarUndefined(varPtr) || TclIsVarLink(varPtr))) {
+	    if (!TclIsVarUndefined(varPtr)) {
+		/*
+		 * Then it is a defined link.
+		 */
+
+		Var *linkPtr = varPtr->value.linkPtr;
+
+		if (linkPtr == otherPtr) {
+		    TRACE_APPEND(("already linked\n"));
+		    NEXT_INST_F(5, 1, 0);
+		}
+		if (TclIsVarInHash(linkPtr)) {
+		    VarHashRefCount(linkPtr)--;
+		    if (TclIsVarUndefined(linkPtr)) {
+			TclCleanupVar(linkPtr, NULL);
+		    }
+		}
+	    }
+	    TclSetVarLink(varPtr);
+	    varPtr->value.linkPtr = otherPtr;
+	    if (TclIsVarInHash(otherPtr)) {
+		VarHashRefCount(otherPtr)++;
+	    }
+	} else if (TclPtrObjMakeUpvarIdx(interp, otherPtr, NULL, 0,
+		opnd) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Do not pop the namespace or frame index, it may be needed for other
+	 * variables - and [variable] did not push it at all.
+	 */
+
+	TRACE_APPEND(("link made\n"));
+	NEXT_INST_F(5, 1, 0);
+    }
+
+    /*
+     *	   End of variable linking instructions.
+     * -----------------------------------------------------------------
+     */
+
+    case INST_JUMP1:
+	opnd = TclGetInt1AtPtr(pc+1);
+	TRACE(("%d => new pc %u\n", opnd,
+		(unsigned)(pc + opnd - codePtr->codeStart)));
+	NEXT_INST_F(opnd, 0, 0);
+
+    case INST_JUMP4:
+	opnd = TclGetInt4AtPtr(pc+1);
+	TRACE(("%d => new pc %u\n", opnd,
+		(unsigned)(pc + opnd - codePtr->codeStart)));
+	NEXT_INST_F(opnd, 0, 0);
+
+    {
+	int jmpOffset[2], b;
+
+	/* TODO: consider rewrite so we don't compute the offset we're not
+	 * going to take. */
+    case INST_JUMP_FALSE4:
+	jmpOffset[0] = TclGetInt4AtPtr(pc+1);	/* FALSE offset */
+	jmpOffset[1] = 5;			/* TRUE offset */
+	goto doCondJump;
+
+    case INST_JUMP_TRUE4:
+	jmpOffset[0] = 5;
+	jmpOffset[1] = TclGetInt4AtPtr(pc+1);
+	goto doCondJump;
+
+    case INST_JUMP_FALSE1:
+	jmpOffset[0] = TclGetInt1AtPtr(pc+1);
+	jmpOffset[1] = 2;
+	goto doCondJump;
+
+    case INST_JUMP_TRUE1:
+	jmpOffset[0] = 2;
+	jmpOffset[1] = TclGetInt1AtPtr(pc+1);
+
+    doCondJump:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("%d => ", jmpOffset[
+		(*pc==INST_JUMP_FALSE1 || *pc==INST_JUMP_FALSE4) ? 0 : 1]));
+
+	/* TODO - check claim that taking address of b harms performance */
+	/* TODO - consider optimization search for constants */
+	if (TclGetBooleanFromObj(interp, valuePtr, &b) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+#ifdef TCL_COMPILE_DEBUG
+	if (b) {
+	    if ((*pc == INST_JUMP_TRUE1) || (*pc == INST_JUMP_TRUE4)) {
+		TRACE_APPEND(("%.20s true, new pc %u\n", O2S(valuePtr),
+			(unsigned)(pc + jmpOffset[1] - codePtr->codeStart)));
+	    } else {
+		TRACE_APPEND(("%.20s true\n", O2S(valuePtr)));
+	    }
+	} else {
+	    if ((*pc == INST_JUMP_TRUE1) || (*pc == INST_JUMP_TRUE4)) {
+		TRACE_APPEND(("%.20s false\n", O2S(valuePtr)));
+	    } else {
+		TRACE_APPEND(("%.20s false, new pc %u\n", O2S(valuePtr),
+			(unsigned)(pc + jmpOffset[0] - codePtr->codeStart)));
+	    }
+	}
+#endif
+	NEXT_INST_F(jmpOffset[b], 1, 0);
+    }
+
+    case INST_JUMP_TABLE: {
+	Tcl_HashEntry *hPtr;
+	JumptableInfo *jtPtr;
+
+	/*
+	 * Jump to location looked up in a hashtable; fall through to next
+	 * instr if lookup fails.
+	 */
+
+	opnd = TclGetInt4AtPtr(pc+1);
+	jtPtr = (JumptableInfo *) codePtr->auxDataArrayPtr[opnd].clientData;
+	TRACE(("%d \"%.20s\" => ", opnd, O2S(OBJ_AT_TOS)));
+	hPtr = Tcl_FindHashEntry(&jtPtr->hashTable, TclGetString(OBJ_AT_TOS));
+	if (hPtr != NULL) {
+	    int jumpOffset = PTR2INT(Tcl_GetHashValue(hPtr));
+
+	    TRACE_APPEND(("found in table, new pc %u\n",
+		    (unsigned)(pc - codePtr->codeStart + jumpOffset)));
+	    NEXT_INST_F(jumpOffset, 1, 0);
+	} else {
+	    TRACE_APPEND(("not found in table\n"));
+	    NEXT_INST_F(5, 1, 0);
+	}
+    }
+
+    /*
+     * These two instructions are now redundant: the complete logic of the LOR
+     * and LAND is now handled by the expression compiler.
+     */
+
+    case INST_LOR:
+    case INST_LAND: {
+	/*
+	 * Operands must be boolean or numeric. No int->double conversions are
+	 * performed.
+	 */
+
+	int i1, i2, iResult;
+
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+	if (TclGetBooleanFromObj(NULL, valuePtr, &i1) != TCL_OK) {
+	    TRACE(("\"%.20s\" => ILLEGAL TYPE %s \n", O2S(valuePtr),
+		    (valuePtr->typePtr? valuePtr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+	if (TclGetBooleanFromObj(NULL, value2Ptr, &i2) != TCL_OK) {
+	    TRACE(("\"%.20s\" => ILLEGAL TYPE %s \n", O2S(value2Ptr),
+		    (value2Ptr->typePtr? value2Ptr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, value2Ptr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+	if (*pc == INST_LOR) {
+	    iResult = (i1 || i2);
+	} else {
+	    iResult = (i1 && i2);
+	}
+	objResultPtr = TCONST(iResult);
+	TRACE(("%.20s %.20s => %d\n", O2S(valuePtr),O2S(value2Ptr),iResult));
+	NEXT_INST_F(1, 2, 1);
+    }
+
+    /*
+     * -----------------------------------------------------------------
+     *	   Start of general introspector instructions.
+     */
+
+    case INST_NS_CURRENT: {
+	Namespace *currNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+
+	if (currNsPtr == (Namespace *) TclGetGlobalNamespace(interp)) {
+	    TclNewLiteralStringObj(objResultPtr, "::");
+	} else {
+	    TclNewStringObj(objResultPtr, currNsPtr->fullName,
+		    strlen(currNsPtr->fullName));
+	}
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+    }
+    case INST_COROUTINE_NAME: {
+	CoroutineData *corPtr = iPtr->execEnvPtr->corPtr;
+
+	TclNewObj(objResultPtr);
+	if (corPtr && !(corPtr->cmdPtr->flags & CMD_IS_DELETED)) {
+	    Tcl_GetCommandFullName(interp, (Tcl_Command) corPtr->cmdPtr,
+		    objResultPtr);
+	}
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+    }
+    case INST_INFO_LEVEL_NUM:
+	TclNewLongObj(objResultPtr, iPtr->varFramePtr->level);
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+    case INST_INFO_LEVEL_ARGS: {
+	int level;
+	register CallFrame *framePtr = iPtr->varFramePtr;
+	register CallFrame *rootFramePtr = iPtr->rootFramePtr;
+
+	TRACE(("\"%.30s\" => ", O2S(OBJ_AT_TOS)));
+	if (TclGetIntFromObj(interp, OBJ_AT_TOS, &level) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	if (level <= 0) {
+	    level += framePtr->level;
+	}
+	for (; (framePtr->level!=level) && (framePtr!=rootFramePtr) ;
+		framePtr = framePtr->callerVarPtr) {
+	    /* Empty loop body */
+	}
+	if (framePtr == rootFramePtr) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad level \"%s\"", TclGetString(OBJ_AT_TOS)));
+	    TRACE_ERROR(interp);
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "STACK_LEVEL",
+		    TclGetString(OBJ_AT_TOS), NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	objResultPtr = Tcl_NewListObj(framePtr->objc, framePtr->objv);
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 1, 1);
+    }
+    {
+	Tcl_Command cmd, origCmd;
+
+    case INST_RESOLVE_COMMAND:
+	cmd = Tcl_GetCommandFromObj(interp, OBJ_AT_TOS);
+	TclNewObj(objResultPtr);
+	if (cmd != NULL) {
+	    Tcl_GetCommandFullName(interp, cmd, objResultPtr);
+	}
+	TRACE_WITH_OBJ(("\"%.20s\" => ", O2S(OBJ_AT_TOS)), objResultPtr);
+	NEXT_INST_F(1, 1, 1);
+
+    case INST_ORIGIN_COMMAND:
+	TRACE(("\"%.30s\" => ", O2S(OBJ_AT_TOS)));
+	cmd = Tcl_GetCommandFromObj(interp, OBJ_AT_TOS);
+	if (cmd == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "invalid command name \"%s\"", TclGetString(OBJ_AT_TOS)));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND",
+		    TclGetString(OBJ_AT_TOS), NULL);
+	    CACHE_STACK_INFO();
+	    TRACE_APPEND(("ERROR: not command\n"));
+	    goto gotError;
+	}
+	origCmd = TclGetOriginalCommand(cmd);
+	if (origCmd == NULL) {
+	    origCmd = cmd;
+	}
+	TclNewObj(objResultPtr);
+	Tcl_GetCommandFullName(interp, origCmd, objResultPtr);
+	TRACE_APPEND(("\"%.30s\"", O2S(OBJ_AT_TOS)));
+	NEXT_INST_F(1, 1, 1);
+    }
+
+    /*
+     * -----------------------------------------------------------------
+     *	   Start of TclOO support instructions.
+     */
+
+    {
+	Object *oPtr;
+	CallFrame *framePtr;
+	CallContext *contextPtr;
+	int skip, newDepth;
+
+    case INST_TCLOO_SELF:
+	framePtr = iPtr->varFramePtr;
+	if (framePtr == NULL ||
+		!(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	    TRACE(("=> ERROR: no TclOO call context\n"));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "self may only be called from inside a method",
+		    -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	contextPtr = framePtr->clientData;
+
+	/*
+	 * Call out to get the name; it's expensive to compute but cached.
+	 */
+
+	objResultPtr = TclOOObjectName(interp, contextPtr->oPtr);
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_TCLOO_NEXT_CLASS:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	framePtr = iPtr->varFramePtr;
+	valuePtr = OBJ_AT_DEPTH(opnd - 2);
+	objv = &OBJ_AT_DEPTH(opnd - 1);
+	skip = 2;
+	TRACE(("%d => ", opnd));
+	if (framePtr == NULL ||
+		!(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	    TRACE_APPEND(("ERROR: no TclOO call context\n"));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "nextto may only be called from inside a method",
+		    -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	contextPtr = framePtr->clientData;
+
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, valuePtr);
+	if (oPtr == NULL) {
+	    TRACE_APPEND(("ERROR: \"%.30s\" not object\n", O2S(valuePtr)));
+	    goto gotError;
+	} else {
+	    Class *classPtr = oPtr->classPtr;
+	    struct MInvoke *miPtr;
+	    int i;
+	    const char *methodType;
+
+	    if (classPtr == NULL) {
+		TRACE_APPEND(("ERROR: \"%.30s\" not class\n", O2S(valuePtr)));
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"\"%s\" is not a class", TclGetString(valuePtr)));
+		DECACHE_STACK_INFO();
+		Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_REQUIRED", NULL);
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+
+	    for (i=contextPtr->index+1 ; i<contextPtr->callPtr->numChain ; i++) {
+		miPtr = contextPtr->callPtr->chain + i;
+		if (!miPtr->isFilter &&
+			miPtr->mPtr->declaringClassPtr == classPtr) {
+		    newDepth = i;
+#ifdef TCL_COMPILE_DEBUG
+		    if (tclTraceExec >= 2) {
+			if (traceInstructions) {
+			    strncpy(cmdNameBuf, TclGetString(objv[0]), 20);
+			} else {
+			    fprintf(stdout, "%d: (%u) invoking ",
+				    iPtr->numLevels,
+				    (unsigned)(pc - codePtr->codeStart));
+			}
+			for (i = 0;  i < opnd;  i++) {
+			    TclPrintObject(stdout, objv[i], 15);
+			    fprintf(stdout, " ");
+			}
+			fprintf(stdout, "\n");
+			fflush(stdout);
+		    }
+#endif /*TCL_COMPILE_DEBUG*/
+		    goto doInvokeNext;
+		}
+	    }
+
+	    if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+		methodType = "constructor";
+	    } else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+		methodType = "destructor";
+	    } else {
+		methodType = "method";
+	    }
+
+	    TRACE_APPEND(("ERROR: \"%.30s\" not on reachable chain\n",
+		    O2S(valuePtr)));
+	    for (i=contextPtr->index ; i>=0 ; i--) {
+		miPtr = contextPtr->callPtr->chain + i;
+		if (miPtr->isFilter
+			|| miPtr->mPtr->declaringClassPtr != classPtr) {
+		    continue;
+		}
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"%s implementation by \"%s\" not reachable from here",
+			methodType, TclGetString(valuePtr)));
+		DECACHE_STACK_INFO();
+		Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_NOT_REACHABLE",
+			NULL);
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "%s has no non-filter implementation by \"%s\"",
+		    methodType, TclGetString(valuePtr)));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_NOT_THERE", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+    case INST_TCLOO_NEXT:
+	opnd = TclGetUInt1AtPtr(pc+1);
+	objv = &OBJ_AT_DEPTH(opnd - 1);
+	framePtr = iPtr->varFramePtr;
+	skip = 1;
+	TRACE(("%d => ", opnd));
+	if (framePtr == NULL ||
+		!(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	    TRACE_APPEND(("ERROR: no TclOO call context\n"));
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "next may only be called from inside a method",
+		    -1));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	contextPtr = framePtr->clientData;
+
+	newDepth = contextPtr->index + 1;
+	if (newDepth >= contextPtr->callPtr->numChain) {
+	    /*
+	     * We're at the end of the chain; generate an error message unless
+	     * the interpreter is being torn down, in which case we might be
+	     * getting here because of methods/destructors doing a [next] (or
+	     * equivalent) unexpectedly.
+	     */
+
+	    const char *methodType;
+
+	    if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+		methodType = "constructor";
+	    } else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+		methodType = "destructor";
+	    } else {
+		methodType = "method";
+	    }
+
+	    TRACE_APPEND(("ERROR: no TclOO next impl\n"));
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "no next %s implementation", methodType));
+	    DECACHE_STACK_INFO();
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "NOTHING_NEXT", NULL);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+#ifdef TCL_COMPILE_DEBUG
+	} else if (tclTraceExec >= 2) {
+	    int i;
+
+	    if (traceInstructions) {
+		strncpy(cmdNameBuf, TclGetString(objv[0]), 20);
+	    } else {
+		fprintf(stdout, "%d: (%u) invoking ",
+			iPtr->numLevels, (unsigned)(pc - codePtr->codeStart));
+	    }
+	    for (i = 0;  i < opnd;  i++) {
+		TclPrintObject(stdout, objv[i], 15);
+		fprintf(stdout, " ");
+	    }
+	    fprintf(stdout, "\n");
+	    fflush(stdout);
+#endif /*TCL_COMPILE_DEBUG*/
+	}
+
+    doInvokeNext:
+	bcFramePtr->data.tebc.pc = (char *) pc;
+	iPtr->cmdFramePtr = bcFramePtr;
+
+	if (iPtr->flags & INTERP_DEBUG_FRAME) {
+	    ArgumentBCEnter(interp, codePtr, TD, pc, opnd, objv);
+	}
+
+	pcAdjustment = 2;
+	cleanup = opnd;
+	DECACHE_STACK_INFO();
+	iPtr->varFramePtr = framePtr->callerVarPtr;
+	pc += pcAdjustment;
+	TEBC_YIELD();
+
+	TclPushTailcallPoint(interp);
+	oPtr = contextPtr->oPtr;
+	if (oPtr->flags & FILTER_HANDLING) {
+	    TclNRAddCallback(interp, FinalizeOONextFilter,
+		    framePtr, contextPtr, INT2PTR(contextPtr->index),
+		    INT2PTR(contextPtr->skip));
+	} else {
+	    TclNRAddCallback(interp, FinalizeOONext,
+		    framePtr, contextPtr, INT2PTR(contextPtr->index),
+		    INT2PTR(contextPtr->skip));
+	}
+	contextPtr->skip = skip;
+	contextPtr->index = newDepth;
+	if (contextPtr->callPtr->chain[newDepth].isFilter
+		|| contextPtr->callPtr->flags & FILTER_HANDLING) {
+	    oPtr->flags |= FILTER_HANDLING;
+	} else {
+	    oPtr->flags &= ~FILTER_HANDLING;
+	}
+
+	{
+	    register Method *const mPtr =
+		    contextPtr->callPtr->chain[newDepth].mPtr;
+
+	    return mPtr->typePtr->callProc(mPtr->clientData, interp,
+		    (Tcl_ObjectContext) contextPtr, opnd, objv);
+	}
+
+    case INST_TCLOO_IS_OBJECT:
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, OBJ_AT_TOS);
+	objResultPtr = TCONST(oPtr != NULL ? 1 : 0);
+	TRACE_WITH_OBJ(("%.30s => ", O2S(OBJ_AT_TOS)), objResultPtr);
+	NEXT_INST_F(1, 1, 1);
+    case INST_TCLOO_CLASS:
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, OBJ_AT_TOS);
+	if (oPtr == NULL) {
+	    TRACE(("%.30s => ERROR: not object\n", O2S(OBJ_AT_TOS)));
+	    goto gotError;
+	}
+	objResultPtr = TclOOObjectName(interp, oPtr->selfCls->thisPtr);
+	TRACE_WITH_OBJ(("%.30s => ", O2S(OBJ_AT_TOS)), objResultPtr);
+	NEXT_INST_F(1, 1, 1);
+    case INST_TCLOO_NS:
+	oPtr = (Object *) Tcl_GetObjectFromObj(interp, OBJ_AT_TOS);
+	if (oPtr == NULL) {
+	    TRACE(("%.30s => ERROR: not object\n", O2S(OBJ_AT_TOS)));
+	    goto gotError;
+	}
+
+	/*
+	 * TclOO objects *never* have the global namespace as their NS.
+	 */
+
+	TclNewStringObj(objResultPtr, oPtr->namespacePtr->fullName,
+		strlen(oPtr->namespacePtr->fullName));
+	TRACE_WITH_OBJ(("%.30s => ", O2S(OBJ_AT_TOS)), objResultPtr);
+	NEXT_INST_F(1, 1, 1);
+    }
+
+    /*
+     *     End of TclOO support instructions.
+     * -----------------------------------------------------------------
+     *	   Start of INST_LIST and related instructions.
+     */
+
+    {
+	int index, numIndices, fromIdx, toIdx;
+	int nocase, match, length2, cflags, s1len, s2len;
+	const char *s1, *s2;
+
+    case INST_LIST:
+	/*
+	 * Pop the opnd (objc) top stack elements into a new list obj and then
+	 * decrement their ref counts.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	objResultPtr = Tcl_NewListObj(opnd, &OBJ_AT_DEPTH(opnd-1));
+	TRACE_WITH_OBJ(("%u => ", opnd), objResultPtr);
+	NEXT_INST_V(5, opnd, 1);
+
+    case INST_LIST_LENGTH:
+	TRACE(("\"%.30s\" => ", O2S(OBJ_AT_TOS)));
+	if (TclListObjLength(interp, OBJ_AT_TOS, &length) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TclNewLongObj(objResultPtr, length);
+	TRACE_APPEND(("%d\n", length));
+	NEXT_INST_F(1, 1, 1);
+
+    case INST_LIST_INDEX:	/* lindex with objc == 3 */
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+	TRACE(("\"%.30s\" \"%.30s\" => ", O2S(valuePtr), O2S(value2Ptr)));
+
+	/*
+	 * Extract the desired list element.
+	 */
+
+	if ((TclListObjGetElements(interp, valuePtr, &objc, &objv) == TCL_OK)
+		&& (value2Ptr->typePtr != &tclListType)
+		&& (TclGetIntForIndexM(NULL , value2Ptr, objc-1,
+			&index) == TCL_OK)) {
+	    TclDecrRefCount(value2Ptr);
+	    tosPtr--;
+	    pcAdjustment = 1;
+	    goto lindexFastPath;
+	}
+
+	objResultPtr = TclLindexList(interp, valuePtr, value2Ptr);
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Stash the list element on the stack.
+	 */
+
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 2, -1);	/* Already has the correct refCount */
+
+    case INST_LIST_INDEX_IMM:	/* lindex with objc==3 and index in bytecode
+				 * stream */
+
+	/*
+	 * Pop the list and get the index.
+	 */
+
+	valuePtr = OBJ_AT_TOS;
+	opnd = TclGetInt4AtPtr(pc+1);
+	TRACE(("\"%.30s\" %d => ", O2S(valuePtr), opnd));
+
+	/*
+	 * Get the contents of the list, making sure that it really is a list
+	 * in the process.
+	 */
+
+	if (TclListObjGetElements(interp, valuePtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Select the list item based on the index. Negative operand means
+	 * end-based indexing.
+	 */
+
+	if (opnd < -1) {
+	    index = opnd+1 + objc;
+	} else {
+	    index = opnd;
+	}
+	pcAdjustment = 5;
+
+    lindexFastPath:
+	if (index >= 0 && index < objc) {
+	    objResultPtr = objv[index];
+	} else {
+	    TclNewObj(objResultPtr);
+	}
+
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(pcAdjustment, 1, 1);
+
+    case INST_LIST_INDEX_MULTI:	/* 'lindex' with multiple index args */
+	/*
+	 * Determine the count of index args.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	numIndices = opnd-1;
+
+	/*
+	 * Do the 'lindex' operation.
+	 */
+
+	TRACE(("%d => ", opnd));
+	objResultPtr = TclLindexFlat(interp, OBJ_AT_DEPTH(numIndices),
+		numIndices, &OBJ_AT_DEPTH(numIndices - 1));
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Set result.
+	 */
+
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_V(5, opnd, -1);
+
+    case INST_LSET_FLAT:
+	/*
+	 * Lset with 3, 5, or more args. Get the number of index args.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc + 1);
+	numIndices = opnd - 2;
+	TRACE(("%d => ", opnd));
+
+	/*
+	 * Get the old value of variable, and remove the stack ref. This is
+	 * safe because the variable still references the object; the ref
+	 * count will never go zero here - we can use the smaller macro
+	 * Tcl_DecrRefCount.
+	 */
+
+	valuePtr = POP_OBJECT();
+	Tcl_DecrRefCount(valuePtr); /* This one should be done here */
+
+	/*
+	 * Compute the new variable value.
+	 */
+
+	objResultPtr = TclLsetFlat(interp, valuePtr, numIndices,
+		&OBJ_AT_DEPTH(numIndices), OBJ_AT_TOS);
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Set result.
+	 */
+
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_V(5, numIndices+1, -1);
+
+    case INST_LSET_LIST:	/* 'lset' with 4 args */
+	/*
+	 * Get the old value of variable, and remove the stack ref. This is
+	 * safe because the variable still references the object; the ref
+	 * count will never go zero here - we can use the smaller macro
+	 * Tcl_DecrRefCount.
+	 */
+
+	objPtr = POP_OBJECT();
+	Tcl_DecrRefCount(objPtr);	/* This one should be done here. */
+
+	/*
+	 * Get the new element value, and the index list.
+	 */
+
+	valuePtr = OBJ_AT_TOS;
+	value2Ptr = OBJ_UNDER_TOS;
+	TRACE(("\"%.30s\" \"%.30s\" \"%.30s\" => ",
+		O2S(value2Ptr), O2S(valuePtr), O2S(objPtr)));
+
+	/*
+	 * Compute the new variable value.
+	 */
+
+	objResultPtr = TclLsetList(interp, objPtr, value2Ptr, valuePtr);
+	if (!objResultPtr) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Set result.
+	 */
+
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 2, -1);
+
+    case INST_LIST_RANGE_IMM:	/* lrange with objc==4 and both indices in
+				 * bytecode stream */
+
+	/*
+	 * Pop the list and get the indices.
+	 */
+
+	valuePtr = OBJ_AT_TOS;
+	fromIdx = TclGetInt4AtPtr(pc+1);
+	toIdx = TclGetInt4AtPtr(pc+5);
+	TRACE(("\"%.30s\" %d %d => ", O2S(valuePtr), TclGetInt4AtPtr(pc+1),
+		TclGetInt4AtPtr(pc+5)));
+
+	/*
+	 * Get the contents of the list, making sure that it really is a list
+	 * in the process.
+	 */
+
+	if (TclListObjGetElements(interp, valuePtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Skip a lot of work if we're about to throw the result away (common
+	 * with uses of [lassign]).
+	 */
+
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+9) == INST_POP) {
+	    NEXT_INST_F(10, 1, 0);
+	}
+#endif
+
+	/*
+	 * Adjust the indices for end-based handling.
+	 */
+
+	if (fromIdx < -1) {
+	    fromIdx += 1+objc;
+	    if (fromIdx < -1) {
+		fromIdx = -1;
+	    }
+	} else if (fromIdx > objc) {
+	    fromIdx = objc;
+	}
+	if (toIdx < -1) {
+	    toIdx += 1 + objc;
+	    if (toIdx < -1) {
+		toIdx = -1;
+	    }
+	} else if (toIdx > objc) {
+	    toIdx = objc;
+	}
+
+	/*
+	 * Check if we are referring to a valid, non-empty list range, and if
+	 * so, build the list of elements in that range.
+	 */
+
+	if (fromIdx<=toIdx && fromIdx<objc && toIdx>=0) {
+	    if (fromIdx < 0) {
+		fromIdx = 0;
+	    }
+	    if (toIdx >= objc) {
+		toIdx = objc-1;
+	    }
+	    if (fromIdx == 0 && toIdx != objc-1 && !Tcl_IsShared(valuePtr)) {
+		Tcl_ListObjReplace(interp, valuePtr,
+			toIdx + 1, LIST_MAX, 0, NULL);
+		TRACE_APPEND(("%.30s\n", O2S(valuePtr)));
+		NEXT_INST_F(9, 0, 0);
+	    }
+	    objResultPtr = Tcl_NewListObj(toIdx-fromIdx+1, objv+fromIdx);
+	} else {
+	    TclNewObj(objResultPtr);
+	}
+
+	TRACE_APPEND(("\"%.30s\"", O2S(objResultPtr)));
+	NEXT_INST_F(9, 1, 1);
+
+    case INST_LIST_IN:
+    case INST_LIST_NOT_IN:	/* Basic list containment operators. */
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+
+	s1 = TclGetStringFromObj(valuePtr, &s1len);
+	TRACE(("\"%.30s\" \"%.30s\" => ", O2S(valuePtr), O2S(value2Ptr)));
+	if (TclListObjLength(interp, value2Ptr, &length) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	match = 0;
+	if (length > 0) {
+	    int i = 0;
+	    Tcl_Obj *o;
+
+	    /*
+	     * An empty list doesn't match anything.
+	     */
+
+	    do {
+		Tcl_ListObjIndex(NULL, value2Ptr, i, &o);
+		if (o != NULL) {
+		    s2 = TclGetStringFromObj(o, &s2len);
+		} else {
+		    s2 = "";
+		    s2len = 0;
+		}
+		if (s1len == s2len) {
+		    match = (memcmp(s1, s2, s1len) == 0);
+		}
+		i++;
+	    } while (i < length && match == 0);
+	}
+
+	if (*pc == INST_LIST_NOT_IN) {
+	    match = !match;
+	}
+
+	TRACE_APPEND(("%d\n", match));
+
+	/*
+	 * Peep-hole optimisation: if you're about to jump, do jump from here.
+	 * We're saving the effort of pushing a boolean value only to pop it
+	 * for branching.
+	 */
+
+	JUMP_PEEPHOLE_F(match, 1, 2);
+
+    case INST_LIST_CONCAT:
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+	TRACE(("\"%.30s\" \"%.30s\" => ", O2S(valuePtr), O2S(value2Ptr)));
+	if (Tcl_IsShared(valuePtr)) {
+	    objResultPtr = Tcl_DuplicateObj(valuePtr);
+	    if (Tcl_ListObjAppendList(interp, objResultPtr,
+		    value2Ptr) != TCL_OK) {
+		TRACE_ERROR(interp);
+		TclDecrRefCount(objResultPtr);
+		goto gotError;
+	    }
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 2, 1);
+	} else {
+	    if (Tcl_ListObjAppendList(interp, valuePtr, value2Ptr) != TCL_OK){
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 1, 0);
+	}
+
+    /*
+     *	   End of INST_LIST and related instructions.
+     * -----------------------------------------------------------------
+     *	   Start of string-related instructions.
+     */
+
+    case INST_STR_EQ:
+    case INST_STR_NEQ:		/* String (in)equality check */
+    case INST_STR_CMP:		/* String compare. */
+    stringCompare:
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+
+	if (valuePtr == value2Ptr) {
+	    match = 0;
+	} else {
+	    /*
+	     * We only need to check (in)equality when we have equal length
+	     * strings.  We can use memcmp in all (n)eq cases because we
+	     * don't need to worry about lexical LE/BE variance.
+	     */
+
+	    typedef int (*memCmpFn_t)(const void*, const void*, size_t);
+	    memCmpFn_t memCmpFn;
+	    int checkEq = ((*pc == INST_EQ) || (*pc == INST_NEQ)
+		    || (*pc == INST_STR_EQ) || (*pc == INST_STR_NEQ));
+
+	    if (TclIsPureByteArray(valuePtr)
+		    && TclIsPureByteArray(value2Ptr)) {
+		s1 = (char *) Tcl_GetByteArrayFromObj(valuePtr, &s1len);
+		s2 = (char *) Tcl_GetByteArrayFromObj(value2Ptr, &s2len);
+		memCmpFn = memcmp;
+	    } else if ((valuePtr->typePtr == &tclStringType)
+		    && (value2Ptr->typePtr == &tclStringType)) {
+		/*
+		 * Do a unicode-specific comparison if both of the args are of
+		 * String type. If the char length == byte length, we can do a
+		 * memcmp. In benchmark testing this proved the most efficient
+		 * check between the unicode and string comparison operations.
+		 */
+
+		s1len = Tcl_GetCharLength(valuePtr);
+		s2len = Tcl_GetCharLength(value2Ptr);
+		if ((s1len == valuePtr->length)
+			&& (valuePtr->bytes != NULL)
+			&& (s2len == value2Ptr->length)
+			&& (value2Ptr->bytes != NULL)) {
+		    s1 = valuePtr->bytes;
+		    s2 = value2Ptr->bytes;
+		    memCmpFn = memcmp;
+		} else {
+		    s1 = (char *) Tcl_GetUnicode(valuePtr);
+		    s2 = (char *) Tcl_GetUnicode(value2Ptr);
+		    if (
+#ifdef WORDS_BIGENDIAN
+			1
+#else
+			checkEq
+#endif
+			) {
+			memCmpFn = memcmp;
+			s1len *= sizeof(Tcl_UniChar);
+			s2len *= sizeof(Tcl_UniChar);
+		    } else {
+			memCmpFn = (memCmpFn_t) Tcl_UniCharNcmp;
+		    }
+		}
+	    } else {
+		/*
+		 * strcmp can't do a simple memcmp in order to handle the
+		 * special Tcl \xC0\x80 null encoding for utf-8.
+		 */
+
+		s1 = TclGetStringFromObj(valuePtr, &s1len);
+		s2 = TclGetStringFromObj(value2Ptr, &s2len);
+		if (checkEq) {
+		    memCmpFn = memcmp;
+		} else {
+		    memCmpFn = (memCmpFn_t) TclpUtfNcmp2;
+		}
+	    }
+
+	    if (checkEq && (s1len != s2len)) {
+		match = 1;
+	    } else {
+		/*
+		 * The comparison function should compare up to the minimum
+		 * byte length only.
+		 */
+		match = memCmpFn(s1, s2,
+			(size_t) ((s1len < s2len) ? s1len : s2len));
+		if (match == 0) {
+		    match = s1len - s2len;
+		}
+	    }
+	}
+
+	/*
+	 * Make sure only -1,0,1 is returned
+	 * TODO: consider peephole opt.
+	 */
+
+	if (*pc != INST_STR_CMP) {
+	    /*
+	     * Take care of the opcodes that goto'ed into here.
+	     */
+
+	    switch (*pc) {
+	    case INST_STR_EQ:
+	    case INST_EQ:
+		match = (match == 0);
+		break;
+	    case INST_STR_NEQ:
+	    case INST_NEQ:
+		match = (match != 0);
+		break;
+	    case INST_LT:
+		match = (match < 0);
+		break;
+	    case INST_GT:
+		match = (match > 0);
+		break;
+	    case INST_LE:
+		match = (match <= 0);
+		break;
+	    case INST_GE:
+		match = (match >= 0);
+		break;
+	    }
+	}
+
+	TRACE(("\"%.20s\" \"%.20s\" => %d\n", O2S(valuePtr), O2S(value2Ptr),
+		(match < 0 ? -1 : match > 0 ? 1 : 0)));
+	JUMP_PEEPHOLE_F(match, 1, 2);
+
+    case INST_STR_LEN:
+	valuePtr = OBJ_AT_TOS;
+	length = Tcl_GetCharLength(valuePtr);
+	TclNewLongObj(objResultPtr, length);
+	TRACE(("\"%.20s\" => %d\n", O2S(valuePtr), length));
+	NEXT_INST_F(1, 1, 1);
+
+    case INST_STR_UPPER:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+	if (Tcl_IsShared(valuePtr)) {
+	    s1 = TclGetStringFromObj(valuePtr, &length);
+	    TclNewStringObj(objResultPtr, s1, length);
+	    length = Tcl_UtfToUpper(TclGetString(objResultPtr));
+	    Tcl_SetObjLength(objResultPtr, length);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 1, 1);
+	} else {
+	    length = Tcl_UtfToUpper(TclGetString(valuePtr));
+	    Tcl_SetObjLength(valuePtr, length);
+	    TclFreeIntRep(valuePtr);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+    case INST_STR_LOWER:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+	if (Tcl_IsShared(valuePtr)) {
+	    s1 = TclGetStringFromObj(valuePtr, &length);
+	    TclNewStringObj(objResultPtr, s1, length);
+	    length = Tcl_UtfToLower(TclGetString(objResultPtr));
+	    Tcl_SetObjLength(objResultPtr, length);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 1, 1);
+	} else {
+	    length = Tcl_UtfToLower(TclGetString(valuePtr));
+	    Tcl_SetObjLength(valuePtr, length);
+	    TclFreeIntRep(valuePtr);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+    case INST_STR_TITLE:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+	if (Tcl_IsShared(valuePtr)) {
+	    s1 = TclGetStringFromObj(valuePtr, &length);
+	    TclNewStringObj(objResultPtr, s1, length);
+	    length = Tcl_UtfToTitle(TclGetString(objResultPtr));
+	    Tcl_SetObjLength(objResultPtr, length);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 1, 1);
+	} else {
+	    length = Tcl_UtfToTitle(TclGetString(valuePtr));
+	    Tcl_SetObjLength(valuePtr, length);
+	    TclFreeIntRep(valuePtr);
+	    TRACE_APPEND(("\"%.20s\"\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+
+    case INST_STR_INDEX:
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+	TRACE(("\"%.20s\" %.20s => ", O2S(valuePtr), O2S(value2Ptr)));
+
+	/*
+	 * Get char length to calulate what 'end' means.
+	 */
+
+	length = Tcl_GetCharLength(valuePtr);
+	if (TclGetIntForIndexM(interp, value2Ptr, length-1, &index)!=TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	if ((index < 0) || (index >= length)) {
+	    TclNewObj(objResultPtr);
+	} else if (TclIsPureByteArray(valuePtr)) {
+	    objResultPtr = Tcl_NewByteArrayObj(
+		    Tcl_GetByteArrayFromObj(valuePtr, NULL)+index, 1);
+	} else if (valuePtr->bytes && length == valuePtr->length) {
+	    objResultPtr = Tcl_NewStringObj((const char *)
+		    valuePtr->bytes+index, 1);
+	} else {
+	    char buf[TCL_UTF_MAX];
+	    Tcl_UniChar ch = Tcl_GetUniChar(valuePtr, index);
+
+	    /*
+	     * This could be: Tcl_NewUnicodeObj((const Tcl_UniChar *)&ch, 1)
+	     * but creating the object as a string seems to be faster in
+	     * practical use.
+	     */
+
+	    length = Tcl_UniCharToUtf(ch, buf);
+	    objResultPtr = Tcl_NewStringObj(buf, length);
+	}
+
+	TRACE_APPEND(("\"%s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 2, 1);
+
+    case INST_STR_RANGE:
+	TRACE(("\"%.20s\" %.20s %.20s =>",
+		O2S(OBJ_AT_DEPTH(2)), O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS)));
+	length = Tcl_GetCharLength(OBJ_AT_DEPTH(2)) - 1;
+	if (TclGetIntForIndexM(interp, OBJ_UNDER_TOS, length,
+		    &fromIdx) != TCL_OK
+	    || TclGetIntForIndexM(interp, OBJ_AT_TOS, length,
+		    &toIdx) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	if (fromIdx < 0) {
+	    fromIdx = 0;
+	}
+	if (toIdx >= length) {
+	    toIdx = length;
+	}
+	if (toIdx >= fromIdx) {
+	    objResultPtr = Tcl_GetRange(OBJ_AT_DEPTH(2), fromIdx, toIdx);
+	} else {
+	    TclNewObj(objResultPtr);
+	}
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_V(1, 3, 1);
+
+    case INST_STR_RANGE_IMM:
+	valuePtr = OBJ_AT_TOS;
+	fromIdx = TclGetInt4AtPtr(pc+1);
+	toIdx = TclGetInt4AtPtr(pc+5);
+	length = Tcl_GetCharLength(valuePtr);
+	TRACE(("\"%.20s\" %d %d => ", O2S(valuePtr), fromIdx, toIdx));
+
+	/*
+	 * Adjust indices for end-based indexing.
+	 */
+
+	if (fromIdx < -1) {
+	    fromIdx += 1 + length;
+	    if (fromIdx < 0) {
+		fromIdx = 0;
+	    }
+	} else if (fromIdx >= length) {
+	    fromIdx = length;
+	}
+	if (toIdx < -1) {
+	    toIdx += 1 + length;
+	} else if (toIdx >= length) {
+	    toIdx = length - 1;
+	}
+
+	/*
+	 * Check if we can do a sane substring.
+	 */
+
+	if (fromIdx <= toIdx) {
+	    objResultPtr = Tcl_GetRange(valuePtr, fromIdx, toIdx);
+	} else {
+	    TclNewObj(objResultPtr);
+	}
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_F(9, 1, 1);
+
+    {
+	Tcl_UniChar *ustring1, *ustring2, *ustring3, *end, *p;
+	int length3;
+	Tcl_Obj *value3Ptr;
+
+    case INST_STR_REPLACE:
+	value3Ptr = POP_OBJECT();
+	valuePtr = OBJ_AT_DEPTH(2);
+	length = Tcl_GetCharLength(valuePtr) - 1;
+	TRACE(("\"%.20s\" %s %s \"%.20s\" => ", O2S(valuePtr),
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS), O2S(value3Ptr)));
+	if (TclGetIntForIndexM(interp, OBJ_UNDER_TOS, length,
+		    &fromIdx) != TCL_OK
+	    || TclGetIntForIndexM(interp, OBJ_AT_TOS, length,
+		    &toIdx) != TCL_OK) {
+	    TclDecrRefCount(value3Ptr);
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TclDecrRefCount(OBJ_AT_TOS);
+	(void) POP_OBJECT();
+	TclDecrRefCount(OBJ_AT_TOS);
+	(void) POP_OBJECT();
+	if (fromIdx < 0) {
+	    fromIdx = 0;
+	}
+
+	if (fromIdx > toIdx || fromIdx > length) {
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(valuePtr)));
+	    TclDecrRefCount(value3Ptr);
+	    NEXT_INST_F(1, 0, 0);
+	}
+
+	if (toIdx > length) {
+	    toIdx = length;
+	}
+
+	if (fromIdx == 0 && toIdx == length) {
+	    TclDecrRefCount(OBJ_AT_TOS);
+	    OBJ_AT_TOS = value3Ptr;
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(value3Ptr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+
+	length3 = Tcl_GetCharLength(value3Ptr);
+
+	/*
+	 * See if we can splice in place. This happens when the number of
+	 * characters being replaced is the same as the number of characters
+	 * in the string to be inserted.
+	 */
+
+	if (length3 - 1 == toIdx - fromIdx) {
+	    unsigned char *bytes1, *bytes2;
+
+	    if (Tcl_IsShared(valuePtr)) {
+		objResultPtr = Tcl_DuplicateObj(valuePtr);
+	    } else {
+		objResultPtr = valuePtr;
+	    }
+	    if (TclIsPureByteArray(objResultPtr)
+		    && TclIsPureByteArray(value3Ptr)) {
+		bytes1 = Tcl_GetByteArrayFromObj(objResultPtr, NULL);
+		bytes2 = Tcl_GetByteArrayFromObj(value3Ptr, NULL);
+		memcpy(bytes1 + fromIdx, bytes2, length3);
+	    } else {
+		ustring1 = Tcl_GetUnicodeFromObj(objResultPtr, NULL);
+		ustring2 = Tcl_GetUnicodeFromObj(value3Ptr, NULL);
+		memcpy(ustring1 + fromIdx, ustring2,
+			length3 * sizeof(Tcl_UniChar));
+	    }
+	    Tcl_InvalidateStringRep(objResultPtr);
+	    TclDecrRefCount(value3Ptr);
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	    if (objResultPtr == valuePtr) {
+		NEXT_INST_F(1, 0, 0);
+	    } else {
+		NEXT_INST_F(1, 1, 1);
+	    }
+	}
+
+	/*
+	 * Get the unicode representation; this is where we guarantee to lose
+	 * bytearrays.
+	 */
+
+	ustring1 = Tcl_GetUnicodeFromObj(valuePtr, &length);
+	length--;
+
+	/*
+	 * Remove substring using copying.
+	 */
+
+	objResultPtr = NULL;
+	if (fromIdx > 0) {
+	    objResultPtr = Tcl_NewUnicodeObj(ustring1, fromIdx);
+	}
+	if (length3 > 0) {
+	    if (objResultPtr) {
+		Tcl_AppendObjToObj(objResultPtr, value3Ptr);
+	    } else if (Tcl_IsShared(value3Ptr)) {
+		objResultPtr = Tcl_DuplicateObj(value3Ptr);
+	    } else {
+		objResultPtr = value3Ptr;
+	    }
+	}
+	if (toIdx < length) {
+	    if (objResultPtr) {
+		Tcl_AppendUnicodeToObj(objResultPtr, ustring1 + toIdx + 1,
+			length - toIdx);
+	    } else {
+		objResultPtr = Tcl_NewUnicodeObj(ustring1 + toIdx + 1,
+			length - toIdx);
+	    }
+	}
+	if (objResultPtr == NULL) {
+	    /* This has to be the case [string replace $s 0 end {}] */
+	    /* which has result {} which is same as value3Ptr. */
+	    objResultPtr = value3Ptr;
+	}
+	if (objResultPtr == value3Ptr) {
+	    /* See [Bug 82e7f67325] */
+	    TclDecrRefCount(OBJ_AT_TOS);
+	    OBJ_AT_TOS = value3Ptr;
+	    TRACE_APPEND(("\"%.30s\"\n", O2S(value3Ptr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+	TclDecrRefCount(value3Ptr);
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 1, 1);
+
+    case INST_STR_MAP:
+	valuePtr = OBJ_AT_TOS;		/* "Main" string. */
+	value3Ptr = OBJ_UNDER_TOS;	/* "Target" string. */
+	value2Ptr = OBJ_AT_DEPTH(2);	/* "Source" string. */
+	if (value3Ptr == value2Ptr) {
+	    objResultPtr = valuePtr;
+	    goto doneStringMap;
+	} else if (valuePtr == value2Ptr) {
+	    objResultPtr = value3Ptr;
+	    goto doneStringMap;
+	}
+	ustring1 = Tcl_GetUnicodeFromObj(valuePtr, &length);
+	if (length == 0) {
+	    objResultPtr = valuePtr;
+	    goto doneStringMap;
+	}
+	ustring2 = Tcl_GetUnicodeFromObj(value2Ptr, &length2);
+	if (length2 > length || length2 == 0) {
+	    objResultPtr = valuePtr;
+	    goto doneStringMap;
+	} else if (length2 == length) {
+	    if (memcmp(ustring1, ustring2, sizeof(Tcl_UniChar) * length)) {
+		objResultPtr = valuePtr;
+	    } else {
+		objResultPtr = value3Ptr;
+	    }
+	    goto doneStringMap;
+	}
+	ustring3 = Tcl_GetUnicodeFromObj(value3Ptr, &length3);
+
+	objResultPtr = Tcl_NewUnicodeObj(ustring1, 0);
+	p = ustring1;
+	end = ustring1 + length;
+	for (; ustring1 < end; ustring1++) {
+	    if ((*ustring1 == *ustring2) && (length2==1 ||
+		    memcmp(ustring1, ustring2, sizeof(Tcl_UniChar) * length2)
+			    == 0)) {
+		if (p != ustring1) {
+		    Tcl_AppendUnicodeToObj(objResultPtr, p, ustring1-p);
+		    p = ustring1 + length2;
+		} else {
+		    p += length2;
+		}
+		ustring1 = p - 1;
+
+		Tcl_AppendUnicodeToObj(objResultPtr, ustring3, length3);
+	    }
+	}
+	if (p != ustring1) {
+	    /*
+	     * Put the rest of the unmapped chars onto result.
+	     */
+
+	    Tcl_AppendUnicodeToObj(objResultPtr, p, ustring1 - p);
+	}
+    doneStringMap:
+	TRACE_WITH_OBJ(("%.20s %.20s %.20s => ",
+		O2S(value2Ptr), O2S(value3Ptr), O2S(valuePtr)), objResultPtr);
+	NEXT_INST_V(1, 3, 1);
+
+    case INST_STR_FIND:
+	match = TclStringFind(OBJ_UNDER_TOS, OBJ_AT_TOS, 0);
+
+	TRACE(("%.20s %.20s => %d\n",
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS), match));
+	TclNewLongObj(objResultPtr, match);
+	NEXT_INST_F(1, 2, 1);
+
+    case INST_STR_FIND_LAST:
+	match = TclStringLast(OBJ_UNDER_TOS, OBJ_AT_TOS, INT_MAX - 1);
+
+	TRACE(("%.20s %.20s => %d\n",
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS), match));
+	TclNewLongObj(objResultPtr, match);
+	NEXT_INST_F(1, 2, 1);
+
+    case INST_STR_CLASS:
+	opnd = TclGetInt1AtPtr(pc+1);
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("%s \"%.30s\" => ", tclStringClassTable[opnd].name,
+		O2S(valuePtr)));
+	ustring1 = Tcl_GetUnicodeFromObj(valuePtr, &length);
+	match = 1;
+	if (length > 0) {
+	    end = ustring1 + length;
+	    for (p=ustring1 ; p<end ; p++) {
+		if (!tclStringClassTable[opnd].comparator(*p)) {
+		    match = 0;
+		    break;
+		}
+	    }
+	}
+	TRACE_APPEND(("%d\n", match));
+	JUMP_PEEPHOLE_F(match, 2, 1);
+    }
+
+    case INST_STR_MATCH:
+	nocase = TclGetInt1AtPtr(pc+1);
+	valuePtr = OBJ_AT_TOS;		/* String */
+	value2Ptr = OBJ_UNDER_TOS;	/* Pattern */
+
+	/*
+	 * Check that at least one of the objects is Unicode before promoting
+	 * both.
+	 */
+
+	if ((valuePtr->typePtr == &tclStringType)
+		|| (value2Ptr->typePtr == &tclStringType)) {
+	    Tcl_UniChar *ustring1, *ustring2;
+
+	    ustring1 = Tcl_GetUnicodeFromObj(valuePtr, &length);
+	    ustring2 = Tcl_GetUnicodeFromObj(value2Ptr, &length2);
+	    match = TclUniCharMatch(ustring1, length, ustring2, length2,
+		    nocase);
+	} else if (TclIsPureByteArray(valuePtr) && !nocase) {
+	    unsigned char *bytes1, *bytes2;
+
+	    bytes1 = Tcl_GetByteArrayFromObj(valuePtr, &length);
+	    bytes2 = Tcl_GetByteArrayFromObj(value2Ptr, &length2);
+	    match = TclByteArrayMatch(bytes1, length, bytes2, length2, 0);
+	} else {
+	    match = Tcl_StringCaseMatch(TclGetString(valuePtr),
+		    TclGetString(value2Ptr), nocase);
+	}
+
+	/*
+	 * Reuse value2Ptr object already on stack if possible. Adjustment is
+	 * 2 due to the nocase byte
+	 */
+
+	TRACE(("%.20s %.20s => %d\n", O2S(valuePtr), O2S(value2Ptr), match));
+
+	/*
+	 * Peep-hole optimisation: if you're about to jump, do jump from here.
+	 */
+
+	JUMP_PEEPHOLE_F(match, 2, 2);
+
+    {
+	const char *string1, *string2;
+	int trim1, trim2;
+
+    case INST_STR_TRIM_LEFT:
+	valuePtr = OBJ_UNDER_TOS;	/* String */
+	value2Ptr = OBJ_AT_TOS;		/* TrimSet */
+	string2 = TclGetStringFromObj(value2Ptr, &length2);
+	string1 = TclGetStringFromObj(valuePtr, &length);
+	trim1 = TclTrimLeft(string1, length, string2, length2);
+	trim2 = 0;
+	goto createTrimmedString;
+    case INST_STR_TRIM_RIGHT:
+	valuePtr = OBJ_UNDER_TOS;	/* String */
+	value2Ptr = OBJ_AT_TOS;		/* TrimSet */
+	string2 = TclGetStringFromObj(value2Ptr, &length2);
+	string1 = TclGetStringFromObj(valuePtr, &length);
+	trim2 = TclTrimRight(string1, length, string2, length2);
+	trim1 = 0;
+	goto createTrimmedString;
+    case INST_STR_TRIM:
+	valuePtr = OBJ_UNDER_TOS;	/* String */
+	value2Ptr = OBJ_AT_TOS;		/* TrimSet */
+	string2 = TclGetStringFromObj(value2Ptr, &length2);
+	string1 = TclGetStringFromObj(valuePtr, &length);
+	trim1 = TclTrimLeft(string1, length, string2, length2);
+	if (trim1 < length) {
+	    trim2 = TclTrimRight(string1, length, string2, length2);
+	} else {
+	    trim2 = 0;
+	}
+    createTrimmedString:
+	/*
+	 * Careful here; trim set often contains non-ASCII characters so we
+	 * take care when printing. [Bug 971cb4f1db]
+	 */
+
+#ifdef TCL_COMPILE_DEBUG
+	if (traceInstructions) {
+	    TRACE(("\"%.30s\" ", O2S(valuePtr)));
+	    TclPrintObject(stdout, value2Ptr, 30);
+	    printf(" => ");
+	}
+#endif
+	if (trim1 == 0 && trim2 == 0) {
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		TclPrintObject(stdout, valuePtr, 30);
+		printf("\n");
+	    }
+#endif
+	    NEXT_INST_F(1, 1, 0);
+	} else {
+	    objResultPtr = Tcl_NewStringObj(string1+trim1, length-trim1-trim2);
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		TclPrintObject(stdout, objResultPtr, 30);
+		printf("\n");
+	    }
+#endif
+	    NEXT_INST_F(1, 2, 1);
+	}
+    }
+
+    case INST_REGEXP:
+	cflags = TclGetInt1AtPtr(pc+1); /* RE compile flages like NOCASE */
+	valuePtr = OBJ_AT_TOS;		/* String */
+	value2Ptr = OBJ_UNDER_TOS;	/* Pattern */
+	TRACE(("\"%.30s\" \"%.30s\" => ", O2S(valuePtr), O2S(value2Ptr)));
+
+	/*
+	 * Compile and match the regular expression.
+	 */
+
+	{
+	    Tcl_RegExp regExpr =
+		    Tcl_GetRegExpFromObj(interp, value2Ptr, cflags);
+
+	    if (regExpr == NULL) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    match = Tcl_RegExpExecObj(interp, regExpr, valuePtr, 0, 0, 0);
+	    if (match < 0) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+
+	TRACE_APPEND(("%d\n", match));
+
+	/*
+	 * Peep-hole optimisation: if you're about to jump, do jump from here.
+	 * Adjustment is 2 due to the nocase byte.
+	 */
+
+	JUMP_PEEPHOLE_F(match, 2, 2);
+    }
+
+    /*
+     *	   End of string-related instructions.
+     * -----------------------------------------------------------------
+     *	   Start of numeric operator instructions.
+     */
+
+    {
+	ClientData ptr1, ptr2;
+	int type1, type2;
+	long l1, l2, lResult;
+
+    case INST_NUM_TYPE:
+	if (GetNumberFromObj(NULL, OBJ_AT_TOS, &ptr1, &type1) != TCL_OK) {
+	    type1 = 0;
+	} else if (type1 == TCL_NUMBER_LONG) {
+	    /* value is between LONG_MIN and LONG_MAX */
+	    /* [string is integer] is -UINT_MAX to UINT_MAX range */
+	    int i;
+
+	    if (Tcl_GetIntFromObj(NULL, OBJ_AT_TOS, &i) != TCL_OK) {
+		type1 = TCL_NUMBER_WIDE;
+	    }
+#ifndef TCL_WIDE_INT_IS_LONG
+	} else if (type1 == TCL_NUMBER_WIDE) {
+	    /* value is between WIDE_MIN and WIDE_MAX */
+	    /* [string is wideinteger] is -UWIDE_MAX to UWIDE_MAX range */
+	    int i;
+	    if (Tcl_GetIntFromObj(NULL, OBJ_AT_TOS, &i) == TCL_OK) {
+		type1 = TCL_NUMBER_LONG;
+	    }
+#endif
+	} else if (type1 == TCL_NUMBER_BIG) {
+	    /* value is an integer outside the WIDE_MIN to WIDE_MAX range */
+	    /* [string is wideinteger] is -UWIDE_MAX to UWIDE_MAX range */
+	    Tcl_WideInt w;
+
+	    if (Tcl_GetWideIntFromObj(NULL, OBJ_AT_TOS, &w) == TCL_OK) {
+		type1 = TCL_NUMBER_WIDE;
+	    }
+	}
+	TclNewLongObj(objResultPtr, type1);
+	TRACE(("\"%.20s\" => %d\n", O2S(OBJ_AT_TOS), type1));
+	NEXT_INST_F(1, 1, 1);
+
+    case INST_EQ:
+    case INST_NEQ:
+    case INST_LT:
+    case INST_GT:
+    case INST_LE:
+    case INST_GE: {
+	int iResult = 0, compare = 0;
+
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+
+	if (GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK
+		|| GetNumberFromObj(NULL, value2Ptr, &ptr2, &type2) != TCL_OK) {
+	    /*
+	     * At least one non-numeric argument - compare as strings.
+	     */
+
+	    goto stringCompare;
+	}
+	if (type1 == TCL_NUMBER_NAN || type2 == TCL_NUMBER_NAN) {
+	    /*
+	     * NaN arg: NaN != to everything, other compares are false.
+	     */
+
+	    iResult = (*pc == INST_NEQ);
+	    goto foundResult;
+	}
+	if (valuePtr == value2Ptr) {
+	    compare = MP_EQ;
+	    goto convertComparison;
+	}
+	if ((type1 == TCL_NUMBER_LONG) && (type2 == TCL_NUMBER_LONG)) {
+	    l1 = *((const long *)ptr1);
+	    l2 = *((const long *)ptr2);
+	    compare = (l1 < l2) ? MP_LT : ((l1 > l2) ? MP_GT : MP_EQ);
+	} else {
+	    compare = TclCompareTwoNumbers(valuePtr, value2Ptr);
+	}
+
+	/*
+	 * Turn comparison outcome into appropriate result for opcode.
+	 */
+
+    convertComparison:
+	switch (*pc) {
+	case INST_EQ:
+	    iResult = (compare == MP_EQ);
+	    break;
+	case INST_NEQ:
+	    iResult = (compare != MP_EQ);
+	    break;
+	case INST_LT:
+	    iResult = (compare == MP_LT);
+	    break;
+	case INST_GT:
+	    iResult = (compare == MP_GT);
+	    break;
+	case INST_LE:
+	    iResult = (compare != MP_GT);
+	    break;
+	case INST_GE:
+	    iResult = (compare != MP_LT);
+	    break;
+	}
+
+	/*
+	 * Peep-hole optimisation: if you're about to jump, do jump from here.
+	 */
+
+    foundResult:
+	TRACE(("\"%.20s\" \"%.20s\" => %d\n", O2S(valuePtr), O2S(value2Ptr),
+		iResult));
+	JUMP_PEEPHOLE_F(iResult, 1, 2);
+    }
+
+    case INST_MOD:
+    case INST_LSHIFT:
+    case INST_RSHIFT:
+    case INST_BITOR:
+    case INST_BITXOR:
+    case INST_BITAND:
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+
+	if ((GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK)
+		|| (type1==TCL_NUMBER_DOUBLE) || (type1==TCL_NUMBER_NAN)) {
+	    TRACE(("%.20s %.20s => ILLEGAL 1st TYPE %s\n", O2S(valuePtr),
+		    O2S(value2Ptr), (valuePtr->typePtr?
+		    valuePtr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+	if ((GetNumberFromObj(NULL, value2Ptr, &ptr2, &type2) != TCL_OK)
+		|| (type2==TCL_NUMBER_DOUBLE) || (type2==TCL_NUMBER_NAN)) {
+	    TRACE(("%.20s %.20s => ILLEGAL 2nd TYPE %s\n", O2S(valuePtr),
+		    O2S(value2Ptr), (value2Ptr->typePtr?
+		    value2Ptr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, value2Ptr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+	/*
+	 * Check for common, simple case.
+	 */
+
+	if ((type1 == TCL_NUMBER_LONG) && (type2 == TCL_NUMBER_LONG)) {
+	    l1 = *((const long *)ptr1);
+	    l2 = *((const long *)ptr2);
+
+	    switch (*pc) {
+	    case INST_MOD:
+		if (l2 == 0) {
+		    TRACE(("%s %s => DIVIDE BY ZERO\n", O2S(valuePtr),
+			    O2S(value2Ptr)));
+		    goto divideByZero;
+		} else if ((l2 == 1) || (l2 == -1)) {
+		    /*
+		     * Div. by |1| always yields remainder of 0.
+		     */
+
+		    TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		    objResultPtr = TCONST(0);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		} else if (l1 == 0) {
+		    /*
+		     * 0 % (non-zero) always yields remainder of 0.
+		     */
+
+		    TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		    objResultPtr = TCONST(0);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		} else {
+		    lResult = l1 / l2;
+
+		    /*
+		     * Force Tcl's integer division rules.
+		     * TODO: examine for logic simplification
+		     */
+
+		    if ((lResult < 0 || (lResult == 0 &&
+			    ((l1 < 0 && l2 > 0) || (l1 > 0 && l2 < 0)))) &&
+			    (lResult * l2 != l1)) {
+			lResult -= 1;
+		    }
+		    lResult = l1 - l2*lResult;
+		    goto longResultOfArithmetic;
+		}
+
+	    case INST_RSHIFT:
+		if (l2 < 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "negative shift argument", -1));
+#ifdef ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR
+		    DECACHE_STACK_INFO();
+		    Tcl_SetErrorCode(interp, "ARITH", "DOMAIN",
+			    "domain error: argument not in valid range",
+			    NULL);
+		    CACHE_STACK_INFO();
+#endif /* ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR */
+		    goto gotError;
+		} else if (l1 == 0) {
+		    TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		    objResultPtr = TCONST(0);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		} else {
+		    /*
+		     * Quickly force large right shifts to 0 or -1.
+		     */
+
+		    if (l2 >= (long)(CHAR_BIT*sizeof(long))) {
+			/*
+			 * We assume that INT_MAX is much larger than the
+			 * number of bits in a long. This is a pretty safe
+			 * assumption, given that the former is usually around
+			 * 4e9 and the latter 32 or 64...
+			 */
+
+			TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+			if (l1 > 0L) {
+			    objResultPtr = TCONST(0);
+			} else {
+			    TclNewLongObj(objResultPtr, -1);
+			}
+			TRACE(("%s\n", O2S(objResultPtr)));
+			NEXT_INST_F(1, 2, 1);
+		    }
+
+		    /*
+		     * Handle shifts within the native long range.
+		     */
+
+		    lResult = l1 >> ((int) l2);
+		    goto longResultOfArithmetic;
+		}
+
+	    case INST_LSHIFT:
+		if (l2 < 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "negative shift argument", -1));
+#ifdef ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR
+		    DECACHE_STACK_INFO();
+		    Tcl_SetErrorCode(interp, "ARITH", "DOMAIN",
+			    "domain error: argument not in valid range",
+			    NULL);
+		    CACHE_STACK_INFO();
+#endif /* ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR */
+		    goto gotError;
+		} else if (l1 == 0) {
+		    TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		    objResultPtr = TCONST(0);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		} else if (l2 > (long) INT_MAX) {
+		    /*
+		     * Technically, we could hold the value (1 << (INT_MAX+1))
+		     * in an mp_int, but since we're using mp_mul_2d() to do
+		     * the work, and it takes only an int argument, that's a
+		     * good place to draw the line.
+		     */
+
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "integer value too large to represent", -1));
+#ifdef ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR
+		    DECACHE_STACK_INFO();
+		    Tcl_SetErrorCode(interp, "ARITH", "IOVERFLOW",
+			    "integer value too large to represent", NULL);
+		    CACHE_STACK_INFO();
+#endif /* ERROR_CODE_FOR_EARLY_DETECTED_ARITH_ERROR */
+		    goto gotError;
+		} else {
+		    int shift = (int) l2;
+
+		    /*
+		     * Handle shifts within the native long range.
+		     */
+
+		    if ((size_t) shift < CHAR_BIT*sizeof(long) && (l1 != 0)
+			    && !((l1>0 ? l1 : ~l1) &
+				-(1L<<(CHAR_BIT*sizeof(long) - 1 - shift)))) {
+			lResult = l1 << shift;
+			goto longResultOfArithmetic;
+		    }
+		}
+
+		/*
+		 * Too large; need to use the broken-out function.
+		 */
+
+		TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		break;
+
+	    case INST_BITAND:
+		lResult = l1 & l2;
+		goto longResultOfArithmetic;
+	    case INST_BITOR:
+		lResult = l1 | l2;
+		goto longResultOfArithmetic;
+	    case INST_BITXOR:
+		lResult = l1 ^ l2;
+	    longResultOfArithmetic:
+		TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		if (Tcl_IsShared(valuePtr)) {
+		    TclNewLongObj(objResultPtr, lResult);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		}
+		TclSetLongObj(valuePtr, lResult);
+		TRACE(("%s\n", O2S(valuePtr)));
+		NEXT_INST_F(1, 1, 0);
+	    }
+	}
+
+	/*
+	 * DO NOT MERGE THIS WITH THE EQUIVALENT SECTION LATER! That would
+	 * encourage the compiler to inline ExecuteExtendedBinaryMathOp, which
+	 * is highly undesirable due to the overall impact on size.
+	 */
+
+	TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+	objResultPtr = ExecuteExtendedBinaryMathOp(interp, *pc, &TCONST(0),
+		valuePtr, value2Ptr);
+	if (objResultPtr == DIVIDED_BY_ZERO) {
+	    TRACE_APPEND(("DIVIDE BY ZERO\n"));
+	    goto divideByZero;
+	} else if (objResultPtr == GENERAL_ARITHMETIC_ERROR) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	} else if (objResultPtr == NULL) {
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 1, 0);
+	} else {
+	    TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 2, 1);
+	}
+
+    case INST_EXPON:
+    case INST_ADD:
+    case INST_SUB:
+    case INST_DIV:
+    case INST_MULT:
+	value2Ptr = OBJ_AT_TOS;
+	valuePtr = OBJ_UNDER_TOS;
+
+	if ((GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK)
+		|| IsErroringNaNType(type1)) {
+	    TRACE(("%.20s %.20s => ILLEGAL 1st TYPE %s\n",
+		    O2S(value2Ptr), O2S(valuePtr),
+		    (valuePtr->typePtr? valuePtr->typePtr->name: "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+#ifdef ACCEPT_NAN
+	if (type1 == TCL_NUMBER_NAN) {
+	    /*
+	     * NaN first argument -> result is also NaN.
+	     */
+
+	    NEXT_INST_F(1, 1, 0);
+	}
+#endif
+
+	if ((GetNumberFromObj(NULL, value2Ptr, &ptr2, &type2) != TCL_OK)
+		|| IsErroringNaNType(type2)) {
+	    TRACE(("%.20s %.20s => ILLEGAL 2nd TYPE %s\n",
+		    O2S(value2Ptr), O2S(valuePtr),
+		    (value2Ptr->typePtr? value2Ptr->typePtr->name: "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, value2Ptr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+
+#ifdef ACCEPT_NAN
+	if (type2 == TCL_NUMBER_NAN) {
+	    /*
+	     * NaN second argument -> result is also NaN.
+	     */
+
+	    objResultPtr = value2Ptr;
+	    NEXT_INST_F(1, 2, 1);
+	}
+#endif
+
+	/*
+	 * Handle (long,long) arithmetic as best we can without going out to
+	 * an external function.
+	 */
+
+	if ((type1 == TCL_NUMBER_LONG) && (type2 == TCL_NUMBER_LONG)) {
+	    Tcl_WideInt w1, w2, wResult;
+
+	    l1 = *((const long *)ptr1);
+	    l2 = *((const long *)ptr2);
+
+	    switch (*pc) {
+	    case INST_ADD:
+		w1 = (Tcl_WideInt) l1;
+		w2 = (Tcl_WideInt) l2;
+		wResult = w1 + w2;
+#ifdef TCL_WIDE_INT_IS_LONG
+		/*
+		 * Check for overflow.
+		 */
+
+		if (Overflowing(w1, w2, wResult)) {
+		    goto overflow;
+		}
+#endif
+		goto wideResultOfArithmetic;
+
+	    case INST_SUB:
+		w1 = (Tcl_WideInt) l1;
+		w2 = (Tcl_WideInt) l2;
+		wResult = w1 - w2;
+#ifdef TCL_WIDE_INT_IS_LONG
+		/*
+		 * Must check for overflow. The macro tests for overflows in
+		 * sums by looking at the sign bits. As we have a subtraction
+		 * here, we are adding -w2. As -w2 could in turn overflow, we
+		 * test with ~w2 instead: it has the opposite sign bit to w2
+		 * so it does the job. Note that the only "bad" case (w2==0)
+		 * is irrelevant for this macro, as in that case w1 and
+		 * wResult have the same sign and there is no overflow anyway.
+		 */
+
+		if (Overflowing(w1, ~w2, wResult)) {
+		    goto overflow;
+		}
+#endif
+	    wideResultOfArithmetic:
+		TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+		if (Tcl_IsShared(valuePtr)) {
+		    objResultPtr = Tcl_NewWideIntObj(wResult);
+		    TRACE(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 2, 1);
+		}
+		Tcl_SetWideIntObj(valuePtr, wResult);
+		TRACE(("%s\n", O2S(valuePtr)));
+		NEXT_INST_F(1, 1, 0);
+
+	    case INST_DIV:
+		if (l2 == 0) {
+		    TRACE(("%s %s => DIVIDE BY ZERO\n",
+			    O2S(valuePtr), O2S(value2Ptr)));
+		    goto divideByZero;
+		} else if ((l1 == LONG_MIN) && (l2 == -1)) {
+		    /*
+		     * Can't represent (-LONG_MIN) as a long.
+		     */
+
+		    goto overflow;
+		}
+		lResult = l1 / l2;
+
+		/*
+		 * Force Tcl's integer division rules.
+		 * TODO: examine for logic simplification
+		 */
+
+		if (((lResult < 0) || ((lResult == 0) &&
+			((l1 < 0 && l2 > 0) || (l1 > 0 && l2 < 0)))) &&
+			((lResult * l2) != l1)) {
+		    lResult -= 1;
+		}
+		goto longResultOfArithmetic;
+
+	    case INST_MULT:
+		if (((sizeof(long) >= 2*sizeof(int))
+			&& (l1 <= INT_MAX) && (l1 >= INT_MIN)
+			&& (l2 <= INT_MAX) && (l2 >= INT_MIN))
+			|| ((sizeof(long) >= 2*sizeof(short))
+			&& (l1 <= SHRT_MAX) && (l1 >= SHRT_MIN)
+			&& (l2 <= SHRT_MAX) && (l2 >= SHRT_MIN))) {
+		    lResult = l1 * l2;
+		    goto longResultOfArithmetic;
+		}
+	    }
+
+	    /*
+	     * Fall through with INST_EXPON, INST_DIV and large multiplies.
+	     */
+	}
+
+    overflow:
+	TRACE(("%s %s => ", O2S(valuePtr), O2S(value2Ptr)));
+	objResultPtr = ExecuteExtendedBinaryMathOp(interp, *pc, &TCONST(0),
+		valuePtr, value2Ptr);
+	if (objResultPtr == DIVIDED_BY_ZERO) {
+	    TRACE_APPEND(("DIVIDE BY ZERO\n"));
+	    goto divideByZero;
+	} else if (objResultPtr == EXPONENT_OF_ZERO) {
+	    TRACE_APPEND(("EXPONENT OF ZERO\n"));
+	    goto exponOfZero;
+	} else if (objResultPtr == GENERAL_ARITHMETIC_ERROR) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	} else if (objResultPtr == NULL) {
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 1, 0);
+	} else {
+	    TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 2, 1);
+	}
+
+    case INST_LNOT: {
+	int b;
+
+	valuePtr = OBJ_AT_TOS;
+
+	/* TODO - check claim that taking address of b harms performance */
+	/* TODO - consider optimization search for constants */
+	if (TclGetBooleanFromObj(NULL, valuePtr, &b) != TCL_OK) {
+	    TRACE(("\"%.20s\" => ERROR: illegal type %s\n", O2S(valuePtr),
+		    (valuePtr->typePtr? valuePtr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	/* TODO: Consider peephole opt. */
+	objResultPtr = TCONST(!b);
+	TRACE_WITH_OBJ(("%s => ", O2S(valuePtr)), objResultPtr);
+	NEXT_INST_F(1, 1, 1);
+    }
+
+    case INST_BITNOT:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+	if ((GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK)
+		|| (type1==TCL_NUMBER_NAN) || (type1==TCL_NUMBER_DOUBLE)) {
+	    /*
+	     * ... ~$NonInteger => raise an error.
+	     */
+
+	    TRACE_APPEND(("ERROR: illegal type %s\n",
+		    (valuePtr->typePtr? valuePtr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	if (type1 == TCL_NUMBER_LONG) {
+	    l1 = *((const long *) ptr1);
+	    if (Tcl_IsShared(valuePtr)) {
+		TclNewLongObj(objResultPtr, ~l1);
+		TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+		NEXT_INST_F(1, 1, 1);
+	    }
+	    TclSetLongObj(valuePtr, ~l1);
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+	objResultPtr = ExecuteExtendedUnaryMathOp(*pc, valuePtr);
+	if (objResultPtr != NULL) {
+	    TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 1, 1);
+	} else {
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+
+    case INST_UMINUS:
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+	if ((GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK)
+		|| IsErroringNaNType(type1)) {
+	    TRACE_APPEND(("ERROR: illegal type %s \n",
+		    (valuePtr->typePtr? valuePtr->typePtr->name : "null")));
+	    DECACHE_STACK_INFO();
+	    IllegalExprOperandType(interp, pc, valuePtr);
+	    CACHE_STACK_INFO();
+	    goto gotError;
+	}
+	switch (type1) {
+	case TCL_NUMBER_NAN:
+	    /* -NaN => NaN */
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	case TCL_NUMBER_LONG:
+	    l1 = *((const long *) ptr1);
+	    if (l1 != LONG_MIN) {
+		if (Tcl_IsShared(valuePtr)) {
+		    TclNewLongObj(objResultPtr, -l1);
+		    TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+		    NEXT_INST_F(1, 1, 1);
+		}
+		TclSetLongObj(valuePtr, -l1);
+		TRACE_APPEND(("%s\n", O2S(valuePtr)));
+		NEXT_INST_F(1, 0, 0);
+	    }
+	    /* FALLTHROUGH */
+	}
+	objResultPtr = ExecuteExtendedUnaryMathOp(*pc, valuePtr);
+	if (objResultPtr != NULL) {
+	    TRACE_APPEND(("%s\n", O2S(objResultPtr)));
+	    NEXT_INST_F(1, 1, 1);
+	} else {
+	    TRACE_APPEND(("%s\n", O2S(valuePtr)));
+	    NEXT_INST_F(1, 0, 0);
+	}
+
+    case INST_UPLUS:
+    case INST_TRY_CVT_TO_NUMERIC:
+	/*
+	 * Try to convert the topmost stack object to numeric object. This is
+	 * done in order to support [expr]'s policy of interpreting operands
+	 * if at all possible as numbers first, then strings.
+	 */
+
+	valuePtr = OBJ_AT_TOS;
+	TRACE(("\"%.20s\" => ", O2S(valuePtr)));
+
+	if (GetNumberFromObj(NULL, valuePtr, &ptr1, &type1) != TCL_OK) {
+	    if (*pc == INST_UPLUS) {
+		/*
+		 * ... +$NonNumeric => raise an error.
+		 */
+
+		TRACE_APPEND(("ERROR: illegal type %s\n",
+			(valuePtr->typePtr? valuePtr->typePtr->name:"null")));
+		DECACHE_STACK_INFO();
+		IllegalExprOperandType(interp, pc, valuePtr);
+		CACHE_STACK_INFO();
+		goto gotError;
+	    }
+
+	    /* ... TryConvertToNumeric($NonNumeric) is acceptable */
+	    TRACE_APPEND(("not numeric\n"));
+	    NEXT_INST_F(1, 0, 0);
+	}
+	if (IsErroringNaNType(type1)) {
+	    if (*pc == INST_UPLUS) {
+		/*
+		 * ... +$NonNumeric => raise an error.
+		 */
+
+		TRACE_APPEND(("ERROR: illegal type %s\n",
+			(valuePtr->typePtr? valuePtr->typePtr->name:"null")));
+		DECACHE_STACK_INFO();
+		IllegalExprOperandType(interp, pc, valuePtr);
+		CACHE_STACK_INFO();
+	    } else {
+		/*
+		 * Numeric conversion of NaN -> error.
+		 */
+
+		TRACE_APPEND(("ERROR: IEEE floating pt error\n"));
+		DECACHE_STACK_INFO();
+		TclExprFloatError(interp, *((const double *) ptr1));
+		CACHE_STACK_INFO();
+	    }
+	    goto gotError;
+	}
+
+	/*
+	 * Ensure that the numeric value has a string rep the same as the
+	 * formatted version of its internal rep. This is used, e.g., to make
+	 * sure that "expr {0001}" yields "1", not "0001". We implement this
+	 * by _discarding_ the string rep since we know it will be
+	 * regenerated, if needed later, by formatting the internal rep's
+	 * value.
+	 */
+
+	if (valuePtr->bytes == NULL) {
+	    TRACE_APPEND(("numeric, same Tcl_Obj\n"));
+	    NEXT_INST_F(1, 0, 0);
+	}
+	if (Tcl_IsShared(valuePtr)) {
+	    /*
+	     * Here we do some surgery within the Tcl_Obj internals. We want
+	     * to copy the intrep, but not the string, so we temporarily hide
+	     * the string so we do not copy it.
+	     */
+
+	    char *savedString = valuePtr->bytes;
+
+	    valuePtr->bytes = NULL;
+	    objResultPtr = Tcl_DuplicateObj(valuePtr);
+	    valuePtr->bytes = savedString;
+	    TRACE_APPEND(("numeric, new Tcl_Obj\n"));
+	    NEXT_INST_F(1, 1, 1);
+	}
+	TclInvalidateStringRep(valuePtr);
+	TRACE_APPEND(("numeric, same Tcl_Obj\n"));
+	NEXT_INST_F(1, 0, 0);
+    }
+
+    /*
+     *	   End of numeric operator instructions.
+     * -----------------------------------------------------------------
+     */
+
+    case INST_TRY_CVT_TO_BOOLEAN:
+	valuePtr = OBJ_AT_TOS;
+	if (valuePtr->typePtr == &tclBooleanType) {
+	    objResultPtr = TCONST(1);
+	} else {
+	    int result = (TclSetBooleanFromAny(NULL, valuePtr) == TCL_OK);
+	    objResultPtr = TCONST(result);
+	}
+	TRACE_WITH_OBJ(("\"%.30s\" => ", O2S(valuePtr)), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_BREAK:
+	/*
+	DECACHE_STACK_INFO();
+	Tcl_ResetResult(interp);
+	CACHE_STACK_INFO();
+	*/
+	result = TCL_BREAK;
+	cleanup = 0;
+	TRACE(("=> BREAK!\n"));
+	goto processExceptionReturn;
+
+    case INST_CONTINUE:
+	/*
+	DECACHE_STACK_INFO();
+	Tcl_ResetResult(interp);
+	CACHE_STACK_INFO();
+	*/
+	result = TCL_CONTINUE;
+	cleanup = 0;
+	TRACE(("=> CONTINUE!\n"));
+	goto processExceptionReturn;
+
+    {
+	ForeachInfo *infoPtr;
+	Var *iterVarPtr, *listVarPtr;
+	Tcl_Obj *oldValuePtr, *listPtr, **elements;
+	ForeachVarList *varListPtr;
+	int numLists, iterNum, listTmpIndex, listLen, numVars;
+	int varIndex, valIndex, continueLoop, j, iterTmpIndex;
+	long i;
+
+    case INST_FOREACH_START4: /* DEPRECATED */
+	/*
+	 * Initialize the temporary local var that holds the count of the
+	 * number of iterations of the loop body to -1.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	infoPtr = codePtr->auxDataArrayPtr[opnd].clientData;
+	iterTmpIndex = infoPtr->loopCtTemp;
+	iterVarPtr = LOCAL(iterTmpIndex);
+	oldValuePtr = iterVarPtr->value.objPtr;
+
+	if (oldValuePtr == NULL) {
+	    TclNewLongObj(iterVarPtr->value.objPtr, -1);
+	    Tcl_IncrRefCount(iterVarPtr->value.objPtr);
+	} else {
+	    TclSetLongObj(oldValuePtr, -1);
+	}
+	TRACE(("%u => loop iter count temp %d\n", opnd, iterTmpIndex));
+
+#ifndef TCL_COMPILE_DEBUG
+	/*
+	 * Remark that the compiler ALWAYS sets INST_FOREACH_STEP4 immediately
+	 * after INST_FOREACH_START4 - let us just fall through instead of
+	 * jumping back to the top.
+	 */
+
+	pc += 5;
+	TCL_DTRACE_INST_NEXT();
+#else
+	NEXT_INST_F(5, 0, 0);
+#endif
+
+    case INST_FOREACH_STEP4: /* DEPRECATED */
+	/*
+	 * "Step" a foreach loop (i.e., begin its next iteration) by assigning
+	 * the next value list element to each loop var.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	TRACE(("%u => ", opnd));
+	infoPtr = codePtr->auxDataArrayPtr[opnd].clientData;
+	numLists = infoPtr->numLists;
+
+	/*
+	 * Increment the temp holding the loop iteration number.
+	 */
+
+	iterVarPtr = LOCAL(infoPtr->loopCtTemp);
+	valuePtr = iterVarPtr->value.objPtr;
+	iterNum = valuePtr->internalRep.longValue + 1;
+	TclSetLongObj(valuePtr, iterNum);
+
+	/*
+	 * Check whether all value lists are exhausted and we should stop the
+	 * loop.
+	 */
+
+	continueLoop = 0;
+	listTmpIndex = infoPtr->firstValueTemp;
+	for (i = 0;  i < numLists;  i++) {
+	    varListPtr = infoPtr->varLists[i];
+	    numVars = varListPtr->numVars;
+
+	    listVarPtr = LOCAL(listTmpIndex);
+	    listPtr = listVarPtr->value.objPtr;
+	    if (TclListObjLength(interp, listPtr, &listLen) != TCL_OK) {
+		TRACE_APPEND(("ERROR converting list %ld, \"%.30s\": %s\n",
+			i, O2S(listPtr), O2S(Tcl_GetObjResult(interp))));
+		goto gotError;
+	    }
+	    if (listLen > iterNum * numVars) {
+		continueLoop = 1;
+	    }
+	    listTmpIndex++;
+	}
+
+	/*
+	 * If some var in some var list still has a remaining list element
+	 * iterate one more time. Assign to var the next element from its
+	 * value list. We already checked above that each list temp holds a
+	 * valid list object (by calling Tcl_ListObjLength), but cannot rely
+	 * on that check remaining valid: one list could have been shimmered
+	 * as a side effect of setting a traced variable.
+	 */
+
+	if (continueLoop) {
+	    listTmpIndex = infoPtr->firstValueTemp;
+	    for (i = 0;  i < numLists;  i++) {
+		varListPtr = infoPtr->varLists[i];
+		numVars = varListPtr->numVars;
+
+		listVarPtr = LOCAL(listTmpIndex);
+		listPtr = TclListObjCopy(NULL, listVarPtr->value.objPtr);
+		TclListObjGetElements(interp, listPtr, &listLen, &elements);
+
+		valIndex = (iterNum * numVars);
+		for (j = 0;  j < numVars;  j++) {
+		    if (valIndex >= listLen) {
+			TclNewObj(valuePtr);
+		    } else {
+			valuePtr = elements[valIndex];
+		    }
+
+		    varIndex = varListPtr->varIndexes[j];
+		    varPtr = LOCAL(varIndex);
+		    while (TclIsVarLink(varPtr)) {
+			varPtr = varPtr->value.linkPtr;
+		    }
+		    if (TclIsVarDirectWritable(varPtr)) {
+			value2Ptr = varPtr->value.objPtr;
+			if (valuePtr != value2Ptr) {
+			    if (value2Ptr != NULL) {
+				TclDecrRefCount(value2Ptr);
+			    }
+			    varPtr->value.objPtr = valuePtr;
+			    Tcl_IncrRefCount(valuePtr);
+			}
+		    } else {
+			DECACHE_STACK_INFO();
+			if (TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+				valuePtr, TCL_LEAVE_ERR_MSG, varIndex)==NULL){
+			    CACHE_STACK_INFO();
+			    TRACE_APPEND((
+				    "ERROR init. index temp %d: %s\n",
+				    varIndex, O2S(Tcl_GetObjResult(interp))));
+			    TclDecrRefCount(listPtr);
+			    goto gotError;
+			}
+			CACHE_STACK_INFO();
+		    }
+		    valIndex++;
+		}
+		TclDecrRefCount(listPtr);
+		listTmpIndex++;
+	    }
+	}
+	TRACE_APPEND(("%d lists, iter %d, %s loop\n",
+		numLists, iterNum, (continueLoop? "continue" : "exit")));
+
+	/*
+	 * Run-time peep-hole optimisation: the compiler ALWAYS follows
+	 * INST_FOREACH_STEP4 with an INST_JUMP_FALSE. We just skip that
+	 * instruction and jump direct from here.
+	 */
+
+	pc += 5;
+	if (*pc == INST_JUMP_FALSE1) {
+	    NEXT_INST_F((continueLoop? 2 : TclGetInt1AtPtr(pc+1)), 0, 0);
+	} else {
+	    NEXT_INST_F((continueLoop? 5 : TclGetInt4AtPtr(pc+1)), 0, 0);
+	}
+
+    }
+    {
+	ForeachInfo *infoPtr;
+	Tcl_Obj *listPtr, **elements, *tmpPtr;
+	ForeachVarList *varListPtr;
+	int numLists, iterMax, listLen, numVars;
+	int iterTmp, iterNum, listTmpDepth;
+	int varIndex, valIndex, j;
+	long i;
+
+    case INST_FOREACH_START:
+	/*
+	 * Initialize the data for the looping construct, pushing the
+	 * corresponding Tcl_Objs to the stack.
+	 */
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	infoPtr = codePtr->auxDataArrayPtr[opnd].clientData;
+	numLists = infoPtr->numLists;
+	TRACE(("%u => ", opnd));
+
+	/*
+	 * Compute the number of iterations that will be run: iterMax
+	 */
+
+	iterMax = 0;
+	listTmpDepth = numLists-1;
+	for (i = 0;  i < numLists;  i++) {
+	    varListPtr = infoPtr->varLists[i];
+	    numVars = varListPtr->numVars;
+	    listPtr = OBJ_AT_DEPTH(listTmpDepth);
+	    if (TclListObjLength(interp, listPtr, &listLen) != TCL_OK) {
+		TRACE_APPEND(("ERROR converting list %ld, \"%s\": %s",
+			i, O2S(listPtr), O2S(Tcl_GetObjResult(interp))));
+		goto gotError;
+	    }
+	    if (Tcl_IsShared(listPtr)) {
+		objPtr = TclListObjCopy(NULL, listPtr);
+		Tcl_IncrRefCount(objPtr);
+		Tcl_DecrRefCount(listPtr);
+		OBJ_AT_DEPTH(listTmpDepth) = objPtr;
+	    }
+	    iterTmp = (listLen + (numVars - 1))/numVars;
+	    if (iterTmp > iterMax) {
+		iterMax = iterTmp;
+	    }
+	    listTmpDepth--;
+	}
+
+	/*
+	 * Store the iterNum and iterMax in a single Tcl_Obj; we keep a
+	 * nul-string obj with the pointer stored in the ptrValue so that the
+	 * thing is properly garbage collected. THIS OBJ MAKES NO SENSE, but
+	 * it will never leave this scope and is read-only.
+	 */
+
+	TclNewObj(tmpPtr);
+	tmpPtr->internalRep.twoPtrValue.ptr1 = INT2PTR(0);
+	tmpPtr->internalRep.twoPtrValue.ptr2 = INT2PTR(iterMax);
+	PUSH_OBJECT(tmpPtr); /* iterCounts object */
+
+	/*
+	 * Store a pointer to the ForeachInfo struct; same dirty trick
+	 * as above
+	 */
+
+	TclNewObj(tmpPtr);
+	tmpPtr->internalRep.twoPtrValue.ptr1 = infoPtr;
+	PUSH_OBJECT(tmpPtr); /* infoPtr object */
+	TRACE_APPEND(("jump to loop step\n"));
+
+	/*
+	 * Jump directly to the INST_FOREACH_STEP instruction; the C code just
+	 * falls through.
+	 */
+
+	pc += 5 - infoPtr->loopCtTemp;
+
+    case INST_FOREACH_STEP:
+	/*
+	 * "Step" a foreach loop (i.e., begin its next iteration) by assigning
+	 * the next value list element to each loop var.
+	 */
+
+	tmpPtr = OBJ_AT_TOS;
+	infoPtr = tmpPtr->internalRep.twoPtrValue.ptr1;
+	numLists = infoPtr->numLists;
+	TRACE(("=> "));
+
+	tmpPtr = OBJ_AT_DEPTH(1);
+	iterNum = PTR2INT(tmpPtr->internalRep.twoPtrValue.ptr1);
+	iterMax = PTR2INT(tmpPtr->internalRep.twoPtrValue.ptr2);
+
+	/*
+	 * If some list still has a remaining list element iterate one more
+	 * time. Assign to var the next element from its value list.
+	 */
+
+	if (iterNum < iterMax) {
+	    /*
+	     * Set the variables and jump back to run the body
+	     */
+
+	    tmpPtr->internalRep.twoPtrValue.ptr1 = INT2PTR(iterNum + 1);
+
+	    listTmpDepth = numLists + 1;
+
+	    for (i = 0;  i < numLists;  i++) {
+		varListPtr = infoPtr->varLists[i];
+		numVars = varListPtr->numVars;
+
+		listPtr = OBJ_AT_DEPTH(listTmpDepth);
+		TclListObjGetElements(interp, listPtr, &listLen, &elements);
+
+		valIndex = (iterNum * numVars);
+		for (j = 0;  j < numVars;  j++) {
+		    if (valIndex >= listLen) {
+			TclNewObj(valuePtr);
+		    } else {
+			valuePtr = elements[valIndex];
+		    }
+
+		    varIndex = varListPtr->varIndexes[j];
+		    varPtr = LOCAL(varIndex);
+		    while (TclIsVarLink(varPtr)) {
+			varPtr = varPtr->value.linkPtr;
+		    }
+		    if (TclIsVarDirectWritable(varPtr)) {
+			value2Ptr = varPtr->value.objPtr;
+			if (valuePtr != value2Ptr) {
+			    if (value2Ptr != NULL) {
+				TclDecrRefCount(value2Ptr);
+			    }
+			    varPtr->value.objPtr = valuePtr;
+			    Tcl_IncrRefCount(valuePtr);
+			}
+		    } else {
+			DECACHE_STACK_INFO();
+			if (TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+				valuePtr, TCL_LEAVE_ERR_MSG, varIndex)==NULL){
+			    CACHE_STACK_INFO();
+			    TRACE_APPEND(("ERROR init. index temp %d: %.30s",
+				    varIndex, O2S(Tcl_GetObjResult(interp))));
+			    goto gotError;
+			}
+			CACHE_STACK_INFO();
+		    }
+		    valIndex++;
+		}
+		listTmpDepth--;
+	    }
+	    TRACE_APPEND(("jump to loop start\n"));
+	    /* loopCtTemp being 'misused' for storing the jump size */
+	    NEXT_INST_F(infoPtr->loopCtTemp, 0, 0);
+	}
+
+	TRACE_APPEND(("loop has no more iterations\n"));
+#ifdef TCL_COMPILE_DEBUG
+	NEXT_INST_F(1, 0, 0);
+#else
+	/*
+	 * FALL THROUGH
+	 */
+	pc++;
+#endif
+
+    case INST_FOREACH_END:
+	/* THIS INSTRUCTION IS ONLY CALLED AS A BREAK TARGET */
+	tmpPtr = OBJ_AT_TOS;
+	infoPtr = tmpPtr->internalRep.twoPtrValue.ptr1;
+	numLists = infoPtr->numLists;
+	TRACE(("=> loop terminated\n"));
+	NEXT_INST_V(1, numLists+2, 0);
+
+    case INST_LMAP_COLLECT:
+	/*
+	 * This instruction is only issued by lmap. The stack is:
+	 *   - result
+	 *   - infoPtr
+	 *   - loop counters
+	 *   - valLists
+	 *   - collecting obj (unshared)
+	 * The instruction lappends the result to the collecting obj.
+	 */
+
+	tmpPtr = OBJ_AT_DEPTH(1);
+	infoPtr = tmpPtr->internalRep.twoPtrValue.ptr1;
+	numLists = infoPtr->numLists;
+	TRACE_APPEND(("=> appending to list at depth %d\n", 3 + numLists));
+
+	objPtr = OBJ_AT_DEPTH(3 + numLists);
+	Tcl_ListObjAppendElement(NULL, objPtr, OBJ_AT_TOS);
+	NEXT_INST_F(1, 1, 0);
+    }
+
+    case INST_BEGIN_CATCH4:
+	/*
+	 * Record start of the catch command with exception range index equal
+	 * to the operand. Push the current stack depth onto the special catch
+	 * stack.
+	 */
+
+	*(++catchTop) = CURR_DEPTH;
+	TRACE(("%u => catchTop=%d, stackTop=%d\n",
+		TclGetUInt4AtPtr(pc+1), (int) (catchTop - initCatchTop - 1),
+		(int) CURR_DEPTH));
+	NEXT_INST_F(5, 0, 0);
+
+    case INST_END_CATCH:
+	catchTop--;
+	DECACHE_STACK_INFO();
+	Tcl_ResetResult(interp);
+	CACHE_STACK_INFO();
+	result = TCL_OK;
+	TRACE(("=> catchTop=%d\n", (int) (catchTop - initCatchTop - 1)));
+	NEXT_INST_F(1, 0, 0);
+
+    case INST_PUSH_RESULT:
+	objResultPtr = Tcl_GetObjResult(interp);
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+
+	/*
+	 * See the comments at INST_INVOKE_STK
+	 */
+
+	TclNewObj(objPtr);
+	Tcl_IncrRefCount(objPtr);
+	iPtr->objResultPtr = objPtr;
+	NEXT_INST_F(1, 0, -1);
+
+    case INST_PUSH_RETURN_CODE:
+	TclNewLongObj(objResultPtr, result);
+	TRACE(("=> %u\n", result));
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_PUSH_RETURN_OPTIONS:
+	DECACHE_STACK_INFO();
+	objResultPtr = Tcl_GetReturnOptions(interp, result);
+	CACHE_STACK_INFO();
+	TRACE_WITH_OBJ(("=> "), objResultPtr);
+	NEXT_INST_F(1, 0, 1);
+
+    case INST_RETURN_CODE_BRANCH: {
+	int code;
+
+	if (TclGetIntFromObj(NULL, OBJ_AT_TOS, &code) != TCL_OK) {
+	    Tcl_Panic("INST_RETURN_CODE_BRANCH: TOS not a return code!");
+	}
+	if (code == TCL_OK) {
+	    Tcl_Panic("INST_RETURN_CODE_BRANCH: TOS is TCL_OK!");
+	}
+	if (code < TCL_ERROR || code > TCL_CONTINUE) {
+	    code = TCL_CONTINUE + 1;
+	}
+	TRACE(("\"%s\" => jump offset %d\n", O2S(OBJ_AT_TOS), 2*code-1));
+	NEXT_INST_F(2*code-1, 1, 0);
+    }
+
+    /*
+     * -----------------------------------------------------------------
+     *	   Start of dictionary-related instructions.
+     */
+
+    {
+	int opnd2, allocateDict, done, i, allocdict;
+	Tcl_Obj *dictPtr, *statePtr, *keyPtr, *listPtr, *varNamePtr, *keysPtr;
+	Tcl_Obj *emptyPtr, **keyPtrPtr;
+	Tcl_DictSearch *searchPtr;
+	DictUpdateInfo *duiPtr;
+
+    case INST_DICT_VERIFY:
+	dictPtr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" => ", O2S(dictPtr)));
+	if (Tcl_DictObjSize(interp, dictPtr, &done) != TCL_OK) {
+	    TRACE_APPEND(("ERROR verifying dictionary nature of \"%.30s\": %s\n",
+		    O2S(dictPtr), O2S(Tcl_GetObjResult(interp))));
+	    goto gotError;
+	}
+	TRACE_APPEND(("OK\n"));
+	NEXT_INST_F(1, 1, 0);
+
+    case INST_DICT_GET:
+    case INST_DICT_EXISTS: {
+	register Tcl_Interp *interp2 = interp;
+	register int found;
+
+	opnd = TclGetUInt4AtPtr(pc+1);
+	TRACE(("%u => ", opnd));
+	dictPtr = OBJ_AT_DEPTH(opnd);
+	if (*pc == INST_DICT_EXISTS) {
+	    interp2 = NULL;
+	}
+	if (opnd > 1) {
+	    dictPtr = TclTraceDictPath(interp2, dictPtr, opnd-1,
+		    &OBJ_AT_DEPTH(opnd-1), DICT_PATH_READ);
+	    if (dictPtr == NULL) {
+		if (*pc == INST_DICT_EXISTS) {
+		    found = 0;
+		    goto afterDictExists;
+		}
+		TRACE_WITH_OBJ((
+			"ERROR tracing dictionary path into \"%.30s\": ",
+			O2S(OBJ_AT_DEPTH(opnd))),
+			Tcl_GetObjResult(interp));
+		goto gotError;
+	    }
+	}
+	if (Tcl_DictObjGet(interp2, dictPtr, OBJ_AT_TOS,
+		&objResultPtr) == TCL_OK) {
+	    if (*pc == INST_DICT_EXISTS) {
+		found = (objResultPtr ? 1 : 0);
+		goto afterDictExists;
+	    }
+	    if (!objResultPtr) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"key \"%s\" not known in dictionary",
+			TclGetString(OBJ_AT_TOS)));
+		DECACHE_STACK_INFO();
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "DICT",
+			TclGetString(OBJ_AT_TOS), NULL);
+		CACHE_STACK_INFO();
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	    TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	    NEXT_INST_V(5, opnd+1, 1);
+	} else if (*pc != INST_DICT_EXISTS) {
+	    TRACE_APPEND(("ERROR reading leaf dictionary key \"%.30s\": %s",
+		    O2S(dictPtr), O2S(Tcl_GetObjResult(interp))));
+	    goto gotError;
+	} else {
+	    found = 0;
+	}
+    afterDictExists:
+	TRACE_APPEND(("%d\n", found));
+
+	/*
+	 * The INST_DICT_EXISTS instruction is usually followed by a
+	 * conditional jump, so we can take advantage of this to do some
+	 * peephole optimization (note that we're careful to not close out
+	 * someone doing something else).
+	 */
+
+	JUMP_PEEPHOLE_V(found, 5, opnd+1);
+    }
+
+    case INST_DICT_SET:
+    case INST_DICT_UNSET:
+    case INST_DICT_INCR_IMM:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	opnd2 = TclGetUInt4AtPtr(pc+5);
+
+	varPtr = LOCAL(opnd2);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u %u => ", opnd, opnd2));
+	if (TclIsVarDirectReadable(varPtr)) {
+	    dictPtr = varPtr->value.objPtr;
+	} else {
+	    DECACHE_STACK_INFO();
+	    dictPtr = TclPtrGetVarIdx(interp, varPtr, NULL, NULL, NULL, 0,
+		    opnd2);
+	    CACHE_STACK_INFO();
+	}
+	if (dictPtr == NULL) {
+	    TclNewObj(dictPtr);
+	    allocateDict = 1;
+	} else {
+	    allocateDict = Tcl_IsShared(dictPtr);
+	    if (allocateDict) {
+		dictPtr = Tcl_DuplicateObj(dictPtr);
+	    }
+	}
+
+	switch (*pc) {
+	case INST_DICT_SET:
+	    cleanup = opnd + 1;
+	    result = Tcl_DictObjPutKeyList(interp, dictPtr, opnd,
+		    &OBJ_AT_DEPTH(opnd), OBJ_AT_TOS);
+	    break;
+	case INST_DICT_INCR_IMM:
+	    cleanup = 1;
+	    opnd = TclGetInt4AtPtr(pc+1);
+	    result = Tcl_DictObjGet(interp, dictPtr, OBJ_AT_TOS, &valuePtr);
+	    if (result != TCL_OK) {
+		break;
+	    }
+	    if (valuePtr == NULL) {
+		Tcl_DictObjPut(NULL, dictPtr, OBJ_AT_TOS,Tcl_NewIntObj(opnd));
+	    } else {
+		value2Ptr = Tcl_NewIntObj(opnd);
+		Tcl_IncrRefCount(value2Ptr);
+		if (Tcl_IsShared(valuePtr)) {
+		    valuePtr = Tcl_DuplicateObj(valuePtr);
+		    Tcl_DictObjPut(NULL, dictPtr, OBJ_AT_TOS, valuePtr);
+		}
+		result = TclIncrObj(interp, valuePtr, value2Ptr);
+		if (result == TCL_OK) {
+		    TclInvalidateStringRep(dictPtr);
+		}
+		TclDecrRefCount(value2Ptr);
+	    }
+	    break;
+	case INST_DICT_UNSET:
+	    cleanup = opnd;
+	    result = Tcl_DictObjRemoveKeyList(interp, dictPtr, opnd,
+		    &OBJ_AT_DEPTH(opnd-1));
+	    break;
+	default:
+	    cleanup = 0; /* stop compiler warning */
+	    Tcl_Panic("Should not happen!");
+	}
+
+	if (result != TCL_OK) {
+	    if (allocateDict) {
+		TclDecrRefCount(dictPtr);
+	    }
+	    TRACE_APPEND(("ERROR updating dictionary: %s\n",
+		    O2S(Tcl_GetObjResult(interp))));
+	    goto checkForCatch;
+	}
+
+	if (TclIsVarDirectWritable(varPtr)) {
+	    if (allocateDict) {
+		value2Ptr = varPtr->value.objPtr;
+		Tcl_IncrRefCount(dictPtr);
+		if (value2Ptr != NULL) {
+		    TclDecrRefCount(value2Ptr);
+		}
+		varPtr->value.objPtr = dictPtr;
+	    }
+	    objResultPtr = dictPtr;
+	} else {
+	    Tcl_IncrRefCount(dictPtr);
+	    DECACHE_STACK_INFO();
+	    objResultPtr = TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+		    dictPtr, TCL_LEAVE_ERR_MSG, opnd2);
+	    CACHE_STACK_INFO();
+	    TclDecrRefCount(dictPtr);
+	    if (objResultPtr == NULL) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+9) == INST_POP) {
+	    NEXT_INST_V(10, cleanup, 0);
+	}
+#endif
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_V(9, cleanup, 1);
+
+    case INST_DICT_APPEND:
+    case INST_DICT_LAPPEND:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	varPtr = LOCAL(opnd);
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	TRACE(("%u => ", opnd));
+	if (TclIsVarDirectReadable(varPtr)) {
+	    dictPtr = varPtr->value.objPtr;
+	} else {
+	    DECACHE_STACK_INFO();
+	    dictPtr = TclPtrGetVarIdx(interp, varPtr, NULL, NULL, NULL, 0,
+		    opnd);
+	    CACHE_STACK_INFO();
+	}
+	if (dictPtr == NULL) {
+	    TclNewObj(dictPtr);
+	    allocateDict = 1;
+	} else {
+	    allocateDict = Tcl_IsShared(dictPtr);
+	    if (allocateDict) {
+		dictPtr = Tcl_DuplicateObj(dictPtr);
+	    }
+	}
+
+	if (Tcl_DictObjGet(interp, dictPtr, OBJ_UNDER_TOS,
+		&valuePtr) != TCL_OK) {
+	    if (allocateDict) {
+		TclDecrRefCount(dictPtr);
+	    }
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+
+	/*
+	 * Note that a non-existent key results in a NULL valuePtr, which is a
+	 * case handled separately below. What we *can* say at this point is
+	 * that the write-back will always succeed.
+	 */
+
+	switch (*pc) {
+	case INST_DICT_APPEND:
+	    if (valuePtr == NULL) {
+		Tcl_DictObjPut(NULL, dictPtr, OBJ_UNDER_TOS, OBJ_AT_TOS);
+	    } else if (Tcl_IsShared(valuePtr)) {
+		valuePtr = Tcl_DuplicateObj(valuePtr);
+		Tcl_AppendObjToObj(valuePtr, OBJ_AT_TOS);
+		Tcl_DictObjPut(NULL, dictPtr, OBJ_UNDER_TOS, valuePtr);
+	    } else {
+		Tcl_AppendObjToObj(valuePtr, OBJ_AT_TOS);
+
+		/*
+		 * Must invalidate the string representation of dictionary
+		 * here because we have directly updated the internal
+		 * representation; if we don't, callers could see the wrong
+		 * string rep despite the internal version of the dictionary
+		 * having the correct value. [Bug 3079830]
+		 */
+
+		TclInvalidateStringRep(dictPtr);
+	    }
+	    break;
+	case INST_DICT_LAPPEND:
+	    /*
+	     * More complex because list-append can fail.
+	     */
+
+	    if (valuePtr == NULL) {
+		Tcl_DictObjPut(NULL, dictPtr, OBJ_UNDER_TOS,
+			Tcl_NewListObj(1, &OBJ_AT_TOS));
+		break;
+	    } else if (Tcl_IsShared(valuePtr)) {
+		valuePtr = Tcl_DuplicateObj(valuePtr);
+		if (Tcl_ListObjAppendElement(interp, valuePtr,
+			OBJ_AT_TOS) != TCL_OK) {
+		    TclDecrRefCount(valuePtr);
+		    if (allocateDict) {
+			TclDecrRefCount(dictPtr);
+		    }
+		    TRACE_ERROR(interp);
+		    goto gotError;
+		}
+		Tcl_DictObjPut(NULL, dictPtr, OBJ_UNDER_TOS, valuePtr);
+	    } else {
+		if (Tcl_ListObjAppendElement(interp, valuePtr,
+			OBJ_AT_TOS) != TCL_OK) {
+		    if (allocateDict) {
+			TclDecrRefCount(dictPtr);
+		    }
+		    TRACE_ERROR(interp);
+		    goto gotError;
+		}
+
+		/*
+		 * Must invalidate the string representation of dictionary
+		 * here because we have directly updated the internal
+		 * representation; if we don't, callers could see the wrong
+		 * string rep despite the internal version of the dictionary
+		 * having the correct value. [Bug 3079830]
+		 */
+
+		TclInvalidateStringRep(dictPtr);
+	    }
+	    break;
+	default:
+	    Tcl_Panic("Should not happen!");
+	}
+
+	if (TclIsVarDirectWritable(varPtr)) {
+	    if (allocateDict) {
+		value2Ptr = varPtr->value.objPtr;
+		Tcl_IncrRefCount(dictPtr);
+		if (value2Ptr != NULL) {
+		    TclDecrRefCount(value2Ptr);
+		}
+		varPtr->value.objPtr = dictPtr;
+	    }
+	    objResultPtr = dictPtr;
+	} else {
+	    Tcl_IncrRefCount(dictPtr);
+	    DECACHE_STACK_INFO();
+	    objResultPtr = TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+		    dictPtr, TCL_LEAVE_ERR_MSG, opnd);
+	    CACHE_STACK_INFO();
+	    TclDecrRefCount(dictPtr);
+	    if (objResultPtr == NULL) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+#ifndef TCL_COMPILE_DEBUG
+	if (*(pc+5) == INST_POP) {
+	    NEXT_INST_F(6, 2, 0);
+	}
+#endif
+	TRACE_APPEND(("%.30s\n", O2S(objResultPtr)));
+	NEXT_INST_F(5, 2, 1);
+
+    case INST_DICT_FIRST:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	TRACE(("%u => ", opnd));
+	dictPtr = POP_OBJECT();
+	searchPtr = ckalloc(sizeof(Tcl_DictSearch));
+	if (Tcl_DictObjFirst(interp, dictPtr, searchPtr, &keyPtr,
+		&valuePtr, &done) != TCL_OK) {
+
+	    /*
+	     * dictPtr is no longer on the stack, and we're not
+	     * moving it into the intrep of an iterator.  We need
+	     * to drop the refcount [Tcl Bug 9b352768e6].
+	     */
+
+	    Tcl_DecrRefCount(dictPtr);
+	    ckfree(searchPtr);
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TclNewObj(statePtr);
+	statePtr->typePtr = &dictIteratorType;
+	statePtr->internalRep.twoPtrValue.ptr1 = searchPtr;
+	statePtr->internalRep.twoPtrValue.ptr2 = dictPtr;
+	varPtr = LOCAL(opnd);
+	if (varPtr->value.objPtr) {
+	    if (varPtr->value.objPtr->typePtr == &dictIteratorType) {
+		Tcl_Panic("mis-issued dictFirst!");
+	    }
+	    TclDecrRefCount(varPtr->value.objPtr);
+	}
+	varPtr->value.objPtr = statePtr;
+	Tcl_IncrRefCount(statePtr);
+	goto pushDictIteratorResult;
+
+    case INST_DICT_NEXT:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	TRACE(("%u => ", opnd));
+	statePtr = (*LOCAL(opnd)).value.objPtr;
+	if (statePtr == NULL || statePtr->typePtr != &dictIteratorType) {
+	    Tcl_Panic("mis-issued dictNext!");
+	}
+	searchPtr = statePtr->internalRep.twoPtrValue.ptr1;
+	Tcl_DictObjNext(searchPtr, &keyPtr, &valuePtr, &done);
+    pushDictIteratorResult:
+	if (done) {
+	    TclNewObj(emptyPtr);
+	    PUSH_OBJECT(emptyPtr);
+	    PUSH_OBJECT(emptyPtr);
+	} else {
+	    PUSH_OBJECT(valuePtr);
+	    PUSH_OBJECT(keyPtr);
+	}
+	TRACE_APPEND(("\"%.30s\" \"%.30s\" %d\n",
+		O2S(OBJ_UNDER_TOS), O2S(OBJ_AT_TOS), done));
+
+	/*
+	 * The INST_DICT_FIRST and INST_DICT_NEXT instructsions are always
+	 * followed by a conditional jump, so we can take advantage of this to
+	 * do some peephole optimization (note that we're careful to not close
+	 * out someone doing something else).
+	 */
+
+	JUMP_PEEPHOLE_F(done, 5, 0);
+
+    case INST_DICT_UPDATE_START:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	opnd2 = TclGetUInt4AtPtr(pc+5);
+	TRACE(("%u => ", opnd));
+	varPtr = LOCAL(opnd);
+	duiPtr = codePtr->auxDataArrayPtr[opnd2].clientData;
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	if (TclIsVarDirectReadable(varPtr)) {
+	    dictPtr = varPtr->value.objPtr;
+	} else {
+	    DECACHE_STACK_INFO();
+	    dictPtr = TclPtrGetVarIdx(interp, varPtr, NULL, NULL, NULL,
+		    TCL_LEAVE_ERR_MSG, opnd);
+	    CACHE_STACK_INFO();
+	    if (dictPtr == NULL) {
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+	Tcl_IncrRefCount(dictPtr);
+	if (TclListObjGetElements(interp, OBJ_AT_TOS, &length,
+		&keyPtrPtr) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	if (length != duiPtr->length) {
+	    Tcl_Panic("dictUpdateStart argument length mismatch");
+	}
+	for (i=0 ; i<length ; i++) {
+	    if (Tcl_DictObjGet(interp, dictPtr, keyPtrPtr[i],
+		    &valuePtr) != TCL_OK) {
+		TRACE_ERROR(interp);
+		Tcl_DecrRefCount(dictPtr);
+		goto gotError;
+	    }
+	    varPtr = LOCAL(duiPtr->varIndices[i]);
+	    while (TclIsVarLink(varPtr)) {
+		varPtr = varPtr->value.linkPtr;
+	    }
+	    DECACHE_STACK_INFO();
+	    if (valuePtr == NULL) {
+		TclObjUnsetVar2(interp,
+			localName(iPtr->varFramePtr, duiPtr->varIndices[i]),
+			NULL, 0);
+	    } else if (TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+		    valuePtr, TCL_LEAVE_ERR_MSG,
+		    duiPtr->varIndices[i]) == NULL) {
+		CACHE_STACK_INFO();
+		TRACE_ERROR(interp);
+		Tcl_DecrRefCount(dictPtr);
+		goto gotError;
+	    }
+	    CACHE_STACK_INFO();
+	}
+	TclDecrRefCount(dictPtr);
+	TRACE_APPEND(("OK\n"));
+	NEXT_INST_F(9, 0, 0);
+
+    case INST_DICT_UPDATE_END:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	opnd2 = TclGetUInt4AtPtr(pc+5);
+	TRACE(("%u => ", opnd));
+	varPtr = LOCAL(opnd);
+	duiPtr = codePtr->auxDataArrayPtr[opnd2].clientData;
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	if (TclIsVarDirectReadable(varPtr)) {
+	    dictPtr = varPtr->value.objPtr;
+	} else {
+	    DECACHE_STACK_INFO();
+	    dictPtr = TclPtrGetVarIdx(interp, varPtr, NULL, NULL, NULL, 0,
+		    opnd);
+	    CACHE_STACK_INFO();
+	}
+	if (dictPtr == NULL) {
+	    TRACE_APPEND(("storage was unset\n"));
+	    NEXT_INST_F(9, 1, 0);
+	}
+	if (Tcl_DictObjSize(interp, dictPtr, &length) != TCL_OK
+		|| TclListObjGetElements(interp, OBJ_AT_TOS, &length,
+			&keyPtrPtr) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	allocdict = Tcl_IsShared(dictPtr);
+	if (allocdict) {
+	    dictPtr = Tcl_DuplicateObj(dictPtr);
+	}
+	if (length > 0) {
+	    TclInvalidateStringRep(dictPtr);
+	}
+	for (i=0 ; i<length ; i++) {
+	    Var *var2Ptr = LOCAL(duiPtr->varIndices[i]);
+
+	    while (TclIsVarLink(var2Ptr)) {
+		var2Ptr = var2Ptr->value.linkPtr;
+	    }
+	    if (TclIsVarDirectReadable(var2Ptr)) {
+		valuePtr = var2Ptr->value.objPtr;
+	    } else {
+		DECACHE_STACK_INFO();
+		valuePtr = TclPtrGetVarIdx(interp, var2Ptr, NULL, NULL, NULL,
+			0, duiPtr->varIndices[i]);
+		CACHE_STACK_INFO();
+	    }
+	    if (valuePtr == NULL) {
+		Tcl_DictObjRemove(interp, dictPtr, keyPtrPtr[i]);
+	    } else if (dictPtr == valuePtr) {
+		Tcl_DictObjPut(interp, dictPtr, keyPtrPtr[i],
+			Tcl_DuplicateObj(valuePtr));
+	    } else {
+		Tcl_DictObjPut(interp, dictPtr, keyPtrPtr[i], valuePtr);
+	    }
+	}
+	if (TclIsVarDirectWritable(varPtr)) {
+	    Tcl_IncrRefCount(dictPtr);
+	    TclDecrRefCount(varPtr->value.objPtr);
+	    varPtr->value.objPtr = dictPtr;
+	} else {
+	    DECACHE_STACK_INFO();
+	    objResultPtr = TclPtrSetVarIdx(interp, varPtr, NULL, NULL, NULL,
+		    dictPtr, TCL_LEAVE_ERR_MSG, opnd);
+	    CACHE_STACK_INFO();
+	    if (objResultPtr == NULL) {
+		if (allocdict) {
+		    TclDecrRefCount(dictPtr);
+		}
+		TRACE_ERROR(interp);
+		goto gotError;
+	    }
+	}
+	TRACE_APPEND(("written back\n"));
+	NEXT_INST_F(9, 1, 0);
+
+    case INST_DICT_EXPAND:
+	dictPtr = OBJ_UNDER_TOS;
+	listPtr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" \"%.30s\" =>", O2S(dictPtr), O2S(listPtr)));
+	if (TclListObjGetElements(interp, listPtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	objResultPtr = TclDictWithInit(interp, dictPtr, objc, objv);
+	if (objResultPtr == NULL) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TRACE_APPEND(("\"%.30s\"\n", O2S(objResultPtr)));
+	NEXT_INST_F(1, 2, 1);
+
+    case INST_DICT_RECOMBINE_STK:
+	keysPtr = POP_OBJECT();
+	varNamePtr = OBJ_UNDER_TOS;
+	listPtr = OBJ_AT_TOS;
+	TRACE(("\"%.30s\" \"%.30s\" \"%.30s\" => ",
+		O2S(varNamePtr), O2S(valuePtr), O2S(keysPtr)));
+	if (TclListObjGetElements(interp, listPtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    TclDecrRefCount(keysPtr);
+	    goto gotError;
+	}
+	varPtr = TclObjLookupVarEx(interp, varNamePtr, NULL,
+		TCL_LEAVE_ERR_MSG, "set", 1, 1, &arrayPtr);
+	if (varPtr == NULL) {
+	    TRACE_ERROR(interp);
+	    TclDecrRefCount(keysPtr);
+	    goto gotError;
+	}
+	DECACHE_STACK_INFO();
+	result = TclDictWithFinish(interp, varPtr,arrayPtr,varNamePtr,NULL,-1,
+		objc, objv, keysPtr);
+	CACHE_STACK_INFO();
+	TclDecrRefCount(keysPtr);
+	if (result != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TRACE_APPEND(("OK\n"));
+	NEXT_INST_F(1, 2, 0);
+
+    case INST_DICT_RECOMBINE_IMM:
+	opnd = TclGetUInt4AtPtr(pc+1);
+	listPtr = OBJ_UNDER_TOS;
+	keysPtr = OBJ_AT_TOS;
+	varPtr = LOCAL(opnd);
+	TRACE(("%u <- \"%.30s\" \"%.30s\" => ", opnd, O2S(valuePtr),
+		O2S(keysPtr)));
+	if (TclListObjGetElements(interp, listPtr, &objc, &objv) != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	while (TclIsVarLink(varPtr)) {
+	    varPtr = varPtr->value.linkPtr;
+	}
+	DECACHE_STACK_INFO();
+	result = TclDictWithFinish(interp, varPtr, NULL, NULL, NULL, opnd,
+		objc, objv, keysPtr);
+	CACHE_STACK_INFO();
+	if (result != TCL_OK) {
+	    TRACE_ERROR(interp);
+	    goto gotError;
+	}
+	TRACE_APPEND(("OK\n"));
+	NEXT_INST_F(5, 2, 0);
+    }
+
+    /*
+     *	   End of dictionary-related instructions.
+     * -----------------------------------------------------------------
+     */
+
+    case INST_CLOCK_READ:
+	{			/* Read the wall clock */
+	    Tcl_WideInt wval;
+	    Tcl_Time now;
+	    switch(TclGetUInt1AtPtr(pc+1)) {
+	    case 0:		/* clicks */
+#ifdef TCL_WIDE_CLICKS
+		wval = TclpGetWideClicks();
+#else
+		wval = (Tcl_WideInt) TclpGetClicks();
+#endif
+		break;
+	    case 1:		/* microseconds */
+		Tcl_GetTime(&now);
+		wval = (Tcl_WideInt) now.sec * 1000000 + now.usec;
+		break;
+	    case 2:		/* milliseconds */
+		Tcl_GetTime(&now);
+		wval = (Tcl_WideInt) now.sec * 1000 + now.usec / 1000;
+		break;
+	    case 3:		/* seconds */
+		Tcl_GetTime(&now);
+		wval = (Tcl_WideInt) now.sec;
+		break;
+	    default:
+		Tcl_Panic("clockRead instruction with unknown clock#");
+	    }
+	    /* TclNewWideObj(objResultPtr, wval); doesn't exist */
+	    objResultPtr = Tcl_NewWideIntObj(wval);
+	    TRACE_WITH_OBJ(("=> "), objResultPtr);
+	    NEXT_INST_F(2, 0, 1);
+	}
+
+    default:
+	Tcl_Panic("TclNRExecuteByteCode: unrecognized opCode %u", *pc);
+    } /* end of switch on opCode */
+
+    /*
+     * Block for variables needed to process exception returns.
+     */
+
+    {
+	ExceptionRange *rangePtr;
+				/* Points to closest loop or catch exception
+				 * range enclosing the pc. Used by various
+				 * instructions and processCatch to process
+				 * break, continue, and errors. */
+	const char *bytes;
+
+	/*
+	 * An external evaluation (INST_INVOKE or INST_EVAL) returned
+	 * something different from TCL_OK, or else INST_BREAK or
+	 * INST_CONTINUE were called.
+	 */
+
+    processExceptionReturn:
+#ifdef TCL_COMPILE_DEBUG
+	switch (*pc) {
+	case INST_INVOKE_STK1:
+	    opnd = TclGetUInt1AtPtr(pc+1);
+	    TRACE(("%u => ... after \"%.20s\": ", opnd, cmdNameBuf));
+	    break;
+	case INST_INVOKE_STK4:
+	    opnd = TclGetUInt4AtPtr(pc+1);
+	    TRACE(("%u => ... after \"%.20s\": ", opnd, cmdNameBuf));
+	    break;
+	case INST_EVAL_STK:
+	    /*
+	     * Note that the object at stacktop has to be used before doing
+	     * the cleanup.
+	     */
+
+	    TRACE(("\"%.30s\" => ", O2S(OBJ_AT_TOS)));
+	    break;
+	default:
+	    TRACE(("=> "));
+	}
+#endif
+	if ((result == TCL_CONTINUE) || (result == TCL_BREAK)) {
+	    rangePtr = GetExceptRangeForPc(pc, result, codePtr);
+	    if (rangePtr == NULL) {
+		TRACE_APPEND(("no encl. loop or catch, returning %s\n",
+			StringForResultCode(result)));
+		goto abnormalReturn;
+	    }
+	    if (rangePtr->type == CATCH_EXCEPTION_RANGE) {
+		TRACE_APPEND(("%s ...\n", StringForResultCode(result)));
+		goto processCatch;
+	    }
+	    while (cleanup--) {
+		valuePtr = POP_OBJECT();
+		TclDecrRefCount(valuePtr);
+	    }
+	    if (result == TCL_BREAK) {
+		result = TCL_OK;
+		pc = (codePtr->codeStart + rangePtr->breakOffset);
+		TRACE_APPEND(("%s, range at %d, new pc %d\n",
+			StringForResultCode(result),
+			rangePtr->codeOffset, rangePtr->breakOffset));
+		NEXT_INST_F(0, 0, 0);
+	    }
+	    if (rangePtr->continueOffset == -1) {
+		TRACE_APPEND(("%s, loop w/o continue, checking for catch\n",
+			StringForResultCode(result)));
+		goto checkForCatch;
+	    }
+	    result = TCL_OK;
+	    pc = (codePtr->codeStart + rangePtr->continueOffset);
+	    TRACE_APPEND(("%s, range at %d, new pc %d\n",
+		    StringForResultCode(result),
+		    rangePtr->codeOffset, rangePtr->continueOffset));
+	    NEXT_INST_F(0, 0, 0);
+	}
+#ifdef TCL_COMPILE_DEBUG
+	if (traceInstructions) {
+	    objPtr = Tcl_GetObjResult(interp);
+	    if ((result != TCL_ERROR) && (result != TCL_RETURN)) {
+		TRACE_APPEND(("OTHER RETURN CODE %d, result=\"%.30s\"\n ",
+			result, O2S(objPtr)));
+	    } else {
+		TRACE_APPEND(("%s, result=\"%.30s\"\n",
+			StringForResultCode(result), O2S(objPtr)));
+	    }
+	}
+#endif
+	goto checkForCatch;
+
+	/*
+	 * Division by zero in an expression. Control only reaches this point
+	 * by "goto divideByZero".
+	 */
+
+    divideByZero:
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("divide by zero", -1));
+	DECACHE_STACK_INFO();
+	Tcl_SetErrorCode(interp, "ARITH", "DIVZERO", "divide by zero", NULL);
+	CACHE_STACK_INFO();
+	goto gotError;
+
+	/*
+	 * Exponentiation of zero by negative number in an expression. Control
+	 * only reaches this point by "goto exponOfZero".
+	 */
+
+    exponOfZero:
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"exponentiation of zero by negative power", -1));
+	DECACHE_STACK_INFO();
+	Tcl_SetErrorCode(interp, "ARITH", "DOMAIN",
+		"exponentiation of zero by negative power", NULL);
+	CACHE_STACK_INFO();
+
+	/*
+	 * Almost all error paths feed through here rather than assigning to
+	 * result themselves (for a small but consistent saving).
+	 */
+
+    gotError:
+	result = TCL_ERROR;
+
+	/*
+	 * Execution has generated an "exception" such as TCL_ERROR. If the
+	 * exception is an error, record information about what was being
+	 * executed when the error occurred. Find the closest enclosing catch
+	 * range, if any. If no enclosing catch range is found, stop execution
+	 * and return the "exception" code.
+	 */
+
+    checkForCatch:
+	if (iPtr->execEnvPtr->rewind) {
+	    goto abnormalReturn;
+	}
+	if ((result == TCL_ERROR) && !(iPtr->flags & ERR_ALREADY_LOGGED)) {
+	    const unsigned char *pcBeg;
+
+	    bytes = GetSrcInfoForPc(pc, codePtr, &length, &pcBeg, NULL);
+	    DECACHE_STACK_INFO();
+	    TclLogCommandInfo(interp, codePtr->source, bytes,
+		    bytes ? length : 0, pcBeg, tosPtr);
+	    CACHE_STACK_INFO();
+	}
+	iPtr->flags &= ~ERR_ALREADY_LOGGED;
+
+	/*
+	 * Clear all expansions that may have started after the last
+	 * INST_BEGIN_CATCH.
+	 */
+
+	while (auxObjList) {
+	    if ((catchTop != initCatchTop)
+		    && (*catchTop > (ptrdiff_t)
+			auxObjList->internalRep.twoPtrValue.ptr2)) {
+		break;
+	    }
+	    POP_TAUX_OBJ();
+	}
+
+	/*
+	 * We must not catch if the script in progress has been canceled with
+	 * the TCL_CANCEL_UNWIND flag. Instead, it blows outwards until we
+	 * either hit another interpreter (presumably where the script in
+	 * progress has not been canceled) or we get to the top-level. We do
+	 * NOT modify the interpreter result here because we know it will
+	 * already be set prior to vectoring down to this point in the code.
+	 */
+
+	if (TclCanceled(iPtr) && (Tcl_Canceled(interp, 0) == TCL_ERROR)) {
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		fprintf(stdout, "   ... cancel with unwind, returning %s\n",
+			StringForResultCode(result));
+	    }
+#endif
+	    goto abnormalReturn;
+	}
+
+	/*
+	 * We must not catch an exceeded limit. Instead, it blows outwards
+	 * until we either hit another interpreter (presumably where the limit
+	 * is not exceeded) or we get to the top-level.
+	 */
+
+	if (TclLimitExceeded(iPtr->limit)) {
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		fprintf(stdout, "   ... limit exceeded, returning %s\n",
+			StringForResultCode(result));
+	    }
+#endif
+	    goto abnormalReturn;
+	}
+	if (catchTop == initCatchTop) {
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		fprintf(stdout, "   ... no enclosing catch, returning %s\n",
+			StringForResultCode(result));
+	    }
+#endif
+	    goto abnormalReturn;
+	}
+	rangePtr = GetExceptRangeForPc(pc, TCL_ERROR, codePtr);
+	if (rangePtr == NULL) {
+	    /*
+	     * This is only possible when compiling a [catch] that sends its
+	     * script to INST_EVAL. Cannot correct the compiler without
+	     * breaking compat with previous .tbc compiled scripts.
+	     */
+
+#ifdef TCL_COMPILE_DEBUG
+	    if (traceInstructions) {
+		fprintf(stdout, "   ... no enclosing catch, returning %s\n",
+			StringForResultCode(result));
+	    }
+#endif
+	    goto abnormalReturn;
+	}
+
+	/*
+	 * A catch exception range (rangePtr) was found to handle an
+	 * "exception". It was found either by checkForCatch just above or by
+	 * an instruction during break, continue, or error processing. Jump to
+	 * its catchOffset after unwinding the operand stack to the depth it
+	 * had when starting to execute the range's catch command.
+	 */
+
+    processCatch:
+	while (CURR_DEPTH > *catchTop) {
+	    valuePtr = POP_OBJECT();
+	    TclDecrRefCount(valuePtr);
+	}
+#ifdef TCL_COMPILE_DEBUG
+	if (traceInstructions) {
+	    fprintf(stdout, "  ... found catch at %d, catchTop=%d, "
+		    "unwound to %ld, new pc %u\n",
+		    rangePtr->codeOffset, (int) (catchTop - initCatchTop - 1),
+		    (long) *catchTop, (unsigned) rangePtr->catchOffset);
+	}
+#endif
+	pc = (codePtr->codeStart + rangePtr->catchOffset);
+	NEXT_INST_F(0, 0, 0);	/* Restart the execution loop at pc. */
+
+	/*
+	 * end of infinite loop dispatching on instructions.
+	 */
+
+	/*
+	 * Abnormal return code. Restore the stack to state it had when
+	 * starting to execute the ByteCode. Panic if the stack is below the
+	 * initial level.
+	 */
+
+    abnormalReturn:
+	TCL_DTRACE_INST_LAST();
+
+	/*
+	 * Clear all expansions and same-level NR calls.
+	 *
+	 * Note that expansion markers have a NULL type; avoid removing other
+	 * markers.
+	 */
+
+	while (auxObjList) {
+	    POP_TAUX_OBJ();
+	}
+	while (tosPtr > initTosPtr) {
+	    objPtr = POP_OBJECT();
+	    Tcl_DecrRefCount(objPtr);
+	}
+
+	if (tosPtr < initTosPtr) {
+	    fprintf(stderr,
+		    "\nTclNRExecuteByteCode: abnormal return at pc %u: "
+		    "stack top %d < entry stack top %d\n",
+		    (unsigned)(pc - codePtr->codeStart),
+		    (unsigned) CURR_DEPTH, (unsigned) 0);
+	    Tcl_Panic("TclNRExecuteByteCode execution failure: end stack top < start stack top");
+	}
+	CLANG_ASSERT(bcFramePtr);
+    }
+
+    iPtr->cmdFramePtr = bcFramePtr->nextPtr;
+    TclReleaseByteCode(codePtr);
+    TclStackFree(interp, TD);	/* free my stack */
+
+    return result;
+
+    /*
+     * INST_START_CMD failure case removed where it doesn't bother that much
+     *
+     * Remark that if the interpreter is marked for deletion its
+     * compileEpoch is modified, so that the epoch check also verifies
+     * that the interp is not deleted. If no outside call has been made
+     * since the last check, it is safe to omit the check.
+
+     * case INST_START_CMD:
+     */
+
+	instStartCmdFailed:
+	{
+	    const char *bytes;
+
+	    checkInterp = 1;
+	    length = 0;
+
+	    /*
+	     * We used to switch to direct eval; for NRE-awareness we now
+	     * compile and eval the command so that this evaluation does not
+	     * add a new TEBC instance. [Bug 2910748]
+	     */
+
+	    if (TclInterpReady(interp) == TCL_ERROR) {
+		goto gotError;
+	    }
+
+	    codePtr->flags |= TCL_BYTECODE_RECOMPILE;
+	    bytes = GetSrcInfoForPc(pc, codePtr, &length, NULL, NULL);
+	    opnd = TclGetUInt4AtPtr(pc+1);
+	    pc += (opnd-1);
+	    assert(bytes);
+	    PUSH_OBJECT(Tcl_NewStringObj(bytes, length));
+	    goto instEvalStk;
+	}
+}
+
+#undef codePtr
+#undef iPtr
+#undef bcFramePtr
+#undef initCatchTop
+#undef initTosPtr
+#undef auxObjList
+#undef catchTop
+#undef TCONST
+#undef esPtr
+
+static int
+FinalizeOONext(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallContext *contextPtr = data[1];
+
+    /*
+     * Reset the variable lookup frame.
+     */
+
+    iPtr->varFramePtr = data[0];
+
+    /*
+     * Restore the call chain context index as we've finished the inner invoke
+     * and want to operate in the outer context again.
+     */
+
+    contextPtr->index = PTR2INT(data[2]);
+    contextPtr->skip = PTR2INT(data[3]);
+    contextPtr->oPtr->flags &= ~FILTER_HANDLING;
+    return result;
+}
+
+static int
+FinalizeOONextFilter(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallContext *contextPtr = data[1];
+
+    /*
+     * Reset the variable lookup frame.
+     */
+
+    iPtr->varFramePtr = data[0];
+
+    /*
+     * Restore the call chain context index as we've finished the inner invoke
+     * and want to operate in the outer context again.
+     */
+
+    contextPtr->index = PTR2INT(data[2]);
+    contextPtr->skip = PTR2INT(data[3]);
+    contextPtr->oPtr->flags |= FILTER_HANDLING;
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExecuteExtendedBinaryMathOp, ExecuteExtendedUnaryMathOp --
+ *
+ *	These functions do advanced math for binary and unary operators
+ *	respectively, so that the main TEBC code does not bear the cost of
+ *	them.
+ *
+ * Results:
+ *	A Tcl_Obj* result, or a NULL (in which case valuePtr is updated to
+ *	hold the result value), or one of the special flag values
+ *	GENERAL_ARITHMETIC_ERROR, EXPONENT_OF_ZERO or DIVIDED_BY_ZERO. The
+ *	latter two signify a zero value raised to a negative power or a value
+ *	divided by zero, respectively. With GENERAL_ARITHMETIC_ERROR, all
+ *	error information will have already been reported in the interpreter
+ *	result.
+ *
+ * Side effects:
+ *	May update the Tcl_Obj indicated valuePtr if it is unshared. Will
+ *	return a NULL when that happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+ExecuteExtendedBinaryMathOp(
+    Tcl_Interp *interp,		/* Where to report errors. */
+    int opcode,			/* What operation to perform. */
+    Tcl_Obj **constants,	/* The execution environment's constants. */
+    Tcl_Obj *valuePtr,		/* The first operand on the stack. */
+    Tcl_Obj *value2Ptr)		/* The second operand on the stack. */
+{
+#define LONG_RESULT(l) \
+    if (Tcl_IsShared(valuePtr)) {		\
+	TclNewLongObj(objResultPtr, l);		\
+	return objResultPtr;			\
+    } else {					\
+	Tcl_SetLongObj(valuePtr, l);		\
+	return NULL;				\
+    }
+#define WIDE_RESULT(w) \
+    if (Tcl_IsShared(valuePtr)) {		\
+	return Tcl_NewWideIntObj(w);		\
+    } else {					\
+	Tcl_SetWideIntObj(valuePtr, w);		\
+	return NULL;				\
+    }
+#define BIG_RESULT(b) \
+    if (Tcl_IsShared(valuePtr)) {		\
+	return Tcl_NewBignumObj(b);		\
+    } else {					\
+	Tcl_SetBignumObj(valuePtr, b);		\
+	return NULL;				\
+    }
+#define DOUBLE_RESULT(d) \
+    if (Tcl_IsShared(valuePtr)) {		\
+	TclNewDoubleObj(objResultPtr, (d));	\
+	return objResultPtr;			\
+    } else {					\
+	Tcl_SetDoubleObj(valuePtr, (d));	\
+	return NULL;				\
+    }
+
+    int type1, type2;
+    ClientData ptr1, ptr2;
+    double d1, d2, dResult;
+    long l1, l2, lResult;
+    Tcl_WideInt w1, w2, wResult;
+    mp_int big1, big2, bigResult, bigRemainder;
+    Tcl_Obj *objResultPtr;
+    int invalid, numPos, zero;
+    long shift;
+
+    (void) GetNumberFromObj(NULL, valuePtr, &ptr1, &type1);
+    (void) GetNumberFromObj(NULL, value2Ptr, &ptr2, &type2);
+
+    switch (opcode) {
+    case INST_MOD:
+	/* TODO: Attempts to re-use unshared operands on stack */
+
+	l2 = 0;			/* silence gcc warning */
+	if (type2 == TCL_NUMBER_LONG) {
+	    l2 = *((const long *)ptr2);
+	    if (l2 == 0) {
+		return DIVIDED_BY_ZERO;
+	    }
+	    if ((l2 == 1) || (l2 == -1)) {
+		/*
+		 * Div. by |1| always yields remainder of 0.
+		 */
+
+		return constants[0];
+	    }
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (type1 == TCL_NUMBER_WIDE) {
+	    w1 = *((const Tcl_WideInt *)ptr1);
+	    if (type2 != TCL_NUMBER_BIG) {
+		Tcl_WideInt wQuotient, wRemainder;
+		Tcl_GetWideIntFromObj(NULL, value2Ptr, &w2);
+		wQuotient = w1 / w2;
+
+		/*
+		 * Force Tcl's integer division rules.
+		 * TODO: examine for logic simplification
+		 */
+
+		if (((wQuotient < (Tcl_WideInt) 0)
+			|| ((wQuotient == (Tcl_WideInt) 0)
+			&& ((w1 < (Tcl_WideInt)0 && w2 > (Tcl_WideInt)0)
+			|| (w1 > (Tcl_WideInt)0 && w2 < (Tcl_WideInt)0))))
+			&& (wQuotient * w2 != w1)) {
+		    wQuotient -= (Tcl_WideInt) 1;
+		}
+		wRemainder = w1 - w2*wQuotient;
+		WIDE_RESULT(wRemainder);
+	    }
+
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+
+	    /* TODO: internals intrusion */
+	    if ((w1 > ((Tcl_WideInt) 0)) ^ (big2.sign == MP_ZPOS)) {
+		/*
+		 * Arguments are opposite sign; remainder is sum.
+		 */
+
+		TclBNInitBignumFromWideInt(&big1, w1);
+		mp_add(&big2, &big1, &big2);
+		mp_clear(&big1);
+		BIG_RESULT(&big2);
+	    }
+
+	    /*
+	     * Arguments are same sign; remainder is first operand.
+	     */
+
+	    mp_clear(&big2);
+	    return NULL;
+	}
+#endif
+	Tcl_GetBignumFromObj(NULL, valuePtr, &big1);
+	Tcl_GetBignumFromObj(NULL, value2Ptr, &big2);
+	mp_init(&bigResult);
+	mp_init(&bigRemainder);
+	mp_div(&big1, &big2, &bigResult, &bigRemainder);
+	if (!mp_iszero(&bigRemainder) && (bigRemainder.sign != big2.sign)) {
+	    /*
+	     * Convert to Tcl's integer division rules.
+	     */
+
+	    mp_sub_d(&bigResult, 1, &bigResult);
+	    mp_add(&bigRemainder, &big2, &bigRemainder);
+	}
+	mp_copy(&bigRemainder, &bigResult);
+	mp_clear(&bigRemainder);
+	mp_clear(&big1);
+	mp_clear(&big2);
+	BIG_RESULT(&bigResult);
+
+    case INST_LSHIFT:
+    case INST_RSHIFT: {
+	/*
+	 * Reject negative shift argument.
+	 */
+
+	switch (type2) {
+	case TCL_NUMBER_LONG:
+	    invalid = (*((const long *)ptr2) < 0L);
+	    break;
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+	    invalid = (*((const Tcl_WideInt *)ptr2) < (Tcl_WideInt)0);
+	    break;
+#endif
+	case TCL_NUMBER_BIG:
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	    invalid = (mp_cmp_d(&big2, 0) == MP_LT);
+	    mp_clear(&big2);
+	    break;
+	default:
+	    /* Unused, here to silence compiler warning */
+	    invalid = 0;
+	}
+	if (invalid) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "negative shift argument", -1));
+	    return GENERAL_ARITHMETIC_ERROR;
+	}
+
+	/*
+	 * Zero shifted any number of bits is still zero.
+	 */
+
+	if ((type1==TCL_NUMBER_LONG) && (*((const long *)ptr1) == (long)0)) {
+	    return constants[0];
+	}
+
+	if (opcode == INST_LSHIFT) {
+	    /*
+	     * Large left shifts create integer overflow.
+	     *
+	     * BEWARE! Can't use Tcl_GetIntFromObj() here because that
+	     * converts values in the (unsigned) range to their signed int
+	     * counterparts, leading to incorrect results.
+	     */
+
+	    if ((type2 != TCL_NUMBER_LONG)
+		    || (*((const long *)ptr2) > (long) INT_MAX)) {
+		/*
+		 * Technically, we could hold the value (1 << (INT_MAX+1)) in
+		 * an mp_int, but since we're using mp_mul_2d() to do the
+		 * work, and it takes only an int argument, that's a good
+		 * place to draw the line.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"integer value too large to represent", -1));
+		return GENERAL_ARITHMETIC_ERROR;
+	    }
+	    shift = (int)(*((const long *)ptr2));
+
+	    /*
+	     * Handle shifts within the native wide range.
+	     */
+
+	    if ((type1 != TCL_NUMBER_BIG)
+		    && ((size_t)shift < CHAR_BIT*sizeof(Tcl_WideInt))) {
+		TclGetWideIntFromObj(NULL, valuePtr, &w1);
+		if (!((w1>0 ? w1 : ~w1)
+			& -(((Tcl_WideInt)1)
+			<< (CHAR_BIT*sizeof(Tcl_WideInt) - 1 - shift)))) {
+		    WIDE_RESULT(w1 << shift);
+		}
+	    }
+	} else {
+	    /*
+	     * Quickly force large right shifts to 0 or -1.
+	     */
+
+	    if ((type2 != TCL_NUMBER_LONG)
+		    || (*(const long *)ptr2 > INT_MAX)) {
+		/*
+		 * Again, technically, the value to be shifted could be an
+		 * mp_int so huge that a right shift by (INT_MAX+1) bits could
+		 * not take us to the result of 0 or -1, but since we're using
+		 * mp_div_2d to do the work, and it takes only an int
+		 * argument, we draw the line there.
+		 */
+
+		switch (type1) {
+		case TCL_NUMBER_LONG:
+		    zero = (*(const long *)ptr1 > 0L);
+		    break;
+#ifndef TCL_WIDE_INT_IS_LONG
+		case TCL_NUMBER_WIDE:
+		    zero = (*(const Tcl_WideInt *)ptr1 > (Tcl_WideInt)0);
+		    break;
+#endif
+		case TCL_NUMBER_BIG:
+		    Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+		    zero = (mp_cmp_d(&big1, 0) == MP_GT);
+		    mp_clear(&big1);
+		    break;
+		default:
+		    /* Unused, here to silence compiler warning. */
+		    zero = 0;
+		}
+		if (zero) {
+		    return constants[0];
+		}
+		LONG_RESULT(-1);
+	    }
+	    shift = (int)(*(const long *)ptr2);
+
+#ifndef TCL_WIDE_INT_IS_LONG
+	    /*
+	     * Handle shifts within the native wide range.
+	     */
+
+	    if (type1 == TCL_NUMBER_WIDE) {
+		w1 = *(const Tcl_WideInt *)ptr1;
+		if ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideInt)) {
+		    if (w1 >= (Tcl_WideInt)0) {
+			return constants[0];
+		    }
+		    LONG_RESULT(-1);
+		}
+		WIDE_RESULT(w1 >> shift);
+	    }
+#endif
+	}
+
+	Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+
+	mp_init(&bigResult);
+	if (opcode == INST_LSHIFT) {
+	    mp_mul_2d(&big1, shift, &bigResult);
+	} else {
+	    mp_init(&bigRemainder);
+	    mp_div_2d(&big1, shift, &bigResult, &bigRemainder);
+	    if (mp_cmp_d(&bigRemainder, 0) == MP_LT) {
+		/*
+		 * Convert to Tcl's integer division rules.
+		 */
+
+		mp_sub_d(&bigResult, 1, &bigResult);
+	    }
+	    mp_clear(&bigRemainder);
+	}
+	mp_clear(&big1);
+	BIG_RESULT(&bigResult);
+    }
+
+    case INST_BITOR:
+    case INST_BITXOR:
+    case INST_BITAND:
+	if ((type1 == TCL_NUMBER_BIG) || (type2 == TCL_NUMBER_BIG)) {
+	    mp_int *First, *Second;
+
+	    Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+
+	    /*
+	     * Count how many positive arguments we have. If only one of the
+	     * arguments is negative, store it in 'Second'.
+	     */
+
+	    if (mp_cmp_d(&big1, 0) != MP_LT) {
+		numPos = 1 + (mp_cmp_d(&big2, 0) != MP_LT);
+		First = &big1;
+		Second = &big2;
+	    } else {
+		First = &big2;
+		Second = &big1;
+		numPos = (mp_cmp_d(First, 0) != MP_LT);
+	    }
+	    mp_init(&bigResult);
+
+	    switch (opcode) {
+	    case INST_BITAND:
+		switch (numPos) {
+		case 2:
+		    /*
+		     * Both arguments positive, base case.
+		     */
+
+		    mp_and(First, Second, &bigResult);
+		    break;
+		case 1:
+		    /*
+		     * First is positive; second negative:
+		     * P & N = P & ~~N = P&~(-N-1) = P & (P ^ (-N-1))
+		     */
+
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_xor(First, Second, &bigResult);
+		    mp_and(First, &bigResult, &bigResult);
+		    break;
+		case 0:
+		    /*
+		     * Both arguments negative:
+		     * a & b = ~ (~a | ~b) = -(-a-1|-b-1)-1
+		     */
+
+		    mp_neg(First, First);
+		    mp_sub_d(First, 1, First);
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_or(First, Second, &bigResult);
+		    mp_neg(&bigResult, &bigResult);
+		    mp_sub_d(&bigResult, 1, &bigResult);
+		    break;
+		}
+		break;
+
+	    case INST_BITOR:
+		switch (numPos) {
+		case 2:
+		    /*
+		     * Both arguments positive, base case.
+		     */
+
+		    mp_or(First, Second, &bigResult);
+		    break;
+		case 1:
+		    /*
+		     * First is positive; second negative:
+		     * N|P = ~(~N&~P) = ~((-N-1)&~P) = -((-N-1)&((-N-1)^P))-1
+		     */
+
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_xor(First, Second, &bigResult);
+		    mp_and(Second, &bigResult, &bigResult);
+		    mp_neg(&bigResult, &bigResult);
+		    mp_sub_d(&bigResult, 1, &bigResult);
+		    break;
+		case 0:
+		    /*
+		     * Both arguments negative:
+		     * a | b = ~ (~a & ~b) = -(-a-1&-b-1)-1
+		     */
+
+		    mp_neg(First, First);
+		    mp_sub_d(First, 1, First);
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_and(First, Second, &bigResult);
+		    mp_neg(&bigResult, &bigResult);
+		    mp_sub_d(&bigResult, 1, &bigResult);
+		    break;
+		}
+		break;
+
+	    case INST_BITXOR:
+		switch (numPos) {
+		case 2:
+		    /*
+		     * Both arguments positive, base case.
+		     */
+
+		    mp_xor(First, Second, &bigResult);
+		    break;
+		case 1:
+		    /*
+		     * First is positive; second negative:
+		     * P^N = ~(P^~N) = -(P^(-N-1))-1
+		     */
+
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_xor(First, Second, &bigResult);
+		    mp_neg(&bigResult, &bigResult);
+		    mp_sub_d(&bigResult, 1, &bigResult);
+		    break;
+		case 0:
+		    /*
+		     * Both arguments negative:
+		     * a ^ b = (~a ^ ~b) = (-a-1^-b-1)
+		     */
+
+		    mp_neg(First, First);
+		    mp_sub_d(First, 1, First);
+		    mp_neg(Second, Second);
+		    mp_sub_d(Second, 1, Second);
+		    mp_xor(First, Second, &bigResult);
+		    break;
+		}
+		break;
+	    }
+
+	    mp_clear(&big1);
+	    mp_clear(&big2);
+	    BIG_RESULT(&bigResult);
+	}
+
+#ifndef TCL_WIDE_INT_IS_LONG
+	if ((type1 == TCL_NUMBER_WIDE) || (type2 == TCL_NUMBER_WIDE)) {
+	    TclGetWideIntFromObj(NULL, valuePtr, &w1);
+	    TclGetWideIntFromObj(NULL, value2Ptr, &w2);
+
+	    switch (opcode) {
+	    case INST_BITAND:
+		wResult = w1 & w2;
+		break;
+	    case INST_BITOR:
+		wResult = w1 | w2;
+		break;
+	    case INST_BITXOR:
+		wResult = w1 ^ w2;
+		break;
+	    default:
+		/* Unused, here to silence compiler warning. */
+		wResult = 0;
+	    }
+	    WIDE_RESULT(wResult);
+	}
+#endif
+	l1 = *((const long *)ptr1);
+	l2 = *((const long *)ptr2);
+
+	switch (opcode) {
+	case INST_BITAND:
+	    lResult = l1 & l2;
+	    break;
+	case INST_BITOR:
+	    lResult = l1 | l2;
+	    break;
+	case INST_BITXOR:
+	    lResult = l1 ^ l2;
+	    break;
+	default:
+	    /* Unused, here to silence compiler warning. */
+	    lResult = 0;
+	}
+	LONG_RESULT(lResult);
+
+    case INST_EXPON: {
+	int oddExponent = 0, negativeExponent = 0;
+	unsigned short base;
+
+	if ((type1 == TCL_NUMBER_DOUBLE) || (type2 == TCL_NUMBER_DOUBLE)) {
+	    Tcl_GetDoubleFromObj(NULL, valuePtr, &d1);
+	    Tcl_GetDoubleFromObj(NULL, value2Ptr, &d2);
+
+	    if (d1==0.0 && d2<0.0) {
+		return EXPONENT_OF_ZERO;
+	    }
+	    dResult = pow(d1, d2);
+	    goto doubleResult;
+	}
+	l1 = l2 = 0;
+	if (type2 == TCL_NUMBER_LONG) {
+	    l2 = *((const long *) ptr2);
+	    if (l2 == 0) {
+		/*
+		 * Anything to the zero power is 1.
+		 */
+
+		return constants[1];
+	    } else if (l2 == 1) {
+		/*
+		 * Anything to the first power is itself
+		 */
+
+		return NULL;
+	    }
+	}
+
+	switch (type2) {
+	case TCL_NUMBER_LONG:
+	    negativeExponent = (l2 < 0);
+	    oddExponent = (int) (l2 & 1);
+	    break;
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+	    w2 = *((const Tcl_WideInt *)ptr2);
+	    negativeExponent = (w2 < 0);
+	    oddExponent = (int) (w2 & (Tcl_WideInt)1);
+	    break;
+#endif
+	case TCL_NUMBER_BIG:
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	    negativeExponent = (mp_cmp_d(&big2, 0) == MP_LT);
+	    mp_mod_2d(&big2, 1, &big2);
+	    oddExponent = !mp_iszero(&big2);
+	    mp_clear(&big2);
+	    break;
+	}
+
+	if (type1 == TCL_NUMBER_LONG) {
+	    l1 = *((const long *)ptr1);
+	}
+	if (negativeExponent) {
+	    if (type1 == TCL_NUMBER_LONG) {
+		switch (l1) {
+		case 0:
+		    /*
+		     * Zero to a negative power is div by zero error.
+		     */
+
+		    return EXPONENT_OF_ZERO;
+		case -1:
+		    if (oddExponent) {
+			LONG_RESULT(-1);
+		    }
+		    /* fallthrough */
+		case 1:
+		    /*
+		     * 1 to any power is 1.
+		     */
+
+		    return constants[1];
+		}
+	    }
+
+	    /*
+	     * Integers with magnitude greater than 1 raise to a negative
+	     * power yield the answer zero (see TIP 123).
+	     */
+
+	    return constants[0];
+	}
+
+	if (type1 == TCL_NUMBER_LONG) {
+	    switch (l1) {
+	    case 0:
+		/*
+		 * Zero to a positive power is zero.
+		 */
+
+		return constants[0];
+	    case 1:
+		/*
+		 * 1 to any power is 1.
+		 */
+
+		return constants[1];
+	    case -1:
+		if (!oddExponent) {
+		    return constants[1];
+		}
+		LONG_RESULT(-1);
+	    }
+	}
+
+	/*
+	 * We refuse to accept exponent arguments that exceed one mp_digit
+	 * which means the max exponent value is 2**28-1 = 0x0fffffff =
+	 * 268435455, which fits into a signed 32 bit int which is within the
+	 * range of the long int type. This means any numeric Tcl_Obj value
+	 * not using TCL_NUMBER_LONG type must hold a value larger than we
+	 * accept.
+	 */
+
+	if (type2 != TCL_NUMBER_LONG) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "exponent too large", -1));
+	    return GENERAL_ARITHMETIC_ERROR;
+	}
+
+	if (type1 == TCL_NUMBER_LONG) {
+	    if (l1 == 2) {
+		/*
+		 * Reduce small powers of 2 to shifts.
+		 */
+
+		if ((unsigned long) l2 < CHAR_BIT * sizeof(long) - 1) {
+		    LONG_RESULT(1L << l2);
+		}
+#if !defined(TCL_WIDE_INT_IS_LONG)
+		if ((unsigned long)l2 < CHAR_BIT*sizeof(Tcl_WideInt) - 1) {
+		    WIDE_RESULT(((Tcl_WideInt) 1) << l2);
+		}
+#endif
+		goto overflowExpon;
+	    }
+	    if (l1 == -2) {
+		int signum = oddExponent ? -1 : 1;
+
+		/*
+		 * Reduce small powers of 2 to shifts.
+		 */
+
+		if ((unsigned long) l2 < CHAR_BIT * sizeof(long) - 1) {
+		    LONG_RESULT(signum * (1L << l2));
+		}
+#if !defined(TCL_WIDE_INT_IS_LONG)
+		if ((unsigned long)l2 < CHAR_BIT*sizeof(Tcl_WideInt) - 1){
+		    WIDE_RESULT(signum * (((Tcl_WideInt) 1) << l2));
+		}
+#endif
+		goto overflowExpon;
+	    }
+#if (LONG_MAX == 0x7fffffff)
+	    if (l2 - 2 < (long)MaxBase32Size
+		    && l1 <= MaxBase32[l2 - 2]
+		    && l1 >= -MaxBase32[l2 - 2]) {
+		/*
+		 * Small powers of 32-bit integers.
+		 */
+
+		lResult = l1 * l1;		/* b**2 */
+		switch (l2) {
+		case 2:
+		    break;
+		case 3:
+		    lResult *= l1;		/* b**3 */
+		    break;
+		case 4:
+		    lResult *= lResult;		/* b**4 */
+		    break;
+		case 5:
+		    lResult *= lResult;		/* b**4 */
+		    lResult *= l1;		/* b**5 */
+		    break;
+		case 6:
+		    lResult *= l1;		/* b**3 */
+		    lResult *= lResult;		/* b**6 */
+		    break;
+		case 7:
+		    lResult *= l1;		/* b**3 */
+		    lResult *= lResult;		/* b**6 */
+		    lResult *= l1;		/* b**7 */
+		    break;
+		case 8:
+		    lResult *= lResult;		/* b**4 */
+		    lResult *= lResult;		/* b**8 */
+		    break;
+		}
+		LONG_RESULT(lResult);
+	    }
+
+	    if (l1 - 3 >= 0 && l1 -2 < (long)Exp32IndexSize
+		    && l2 - 2 < (long)(Exp32ValueSize + MaxBase32Size)) {
+		base = Exp32Index[l1 - 3]
+			+ (unsigned short) (l2 - 2 - MaxBase32Size);
+		if (base < Exp32Index[l1 - 2]) {
+		    /*
+		     * 32-bit number raised to intermediate power, done by
+		     * table lookup.
+		     */
+
+		    LONG_RESULT(Exp32Value[base]);
+		}
+	    }
+	    if (-l1 - 3 >= 0 && -l1 - 2 < (long)Exp32IndexSize
+		    && l2 - 2 < (long)(Exp32ValueSize + MaxBase32Size)) {
+		base = Exp32Index[-l1 - 3]
+			+ (unsigned short) (l2 - 2 - MaxBase32Size);
+		if (base < Exp32Index[-l1 - 2]) {
+		    /*
+		     * 32-bit number raised to intermediate power, done by
+		     * table lookup.
+		     */
+
+		    lResult = (oddExponent) ?
+			    -Exp32Value[base] : Exp32Value[base];
+		    LONG_RESULT(lResult);
+		}
+	    }
+#endif
+	}
+#if (LONG_MAX > 0x7fffffff) || !defined(TCL_WIDE_INT_IS_LONG)
+	if (type1 == TCL_NUMBER_LONG) {
+	    w1 = l1;
+#ifndef TCL_WIDE_INT_IS_LONG
+	} else if (type1 == TCL_NUMBER_WIDE) {
+	    w1 = *((const Tcl_WideInt *) ptr1);
+#endif
+	} else {
+	    goto overflowExpon;
+	}
+	if (l2 - 2 < (long)MaxBase64Size
+		&& w1 <=  MaxBase64[l2 - 2]
+		&& w1 >= -MaxBase64[l2 - 2]) {
+	    /*
+	     * Small powers of integers whose result is wide.
+	     */
+
+	    wResult = w1 * w1;		/* b**2 */
+	    switch (l2) {
+	    case 2:
+		break;
+	    case 3:
+		wResult *= l1;		/* b**3 */
+		break;
+	    case 4:
+		wResult *= wResult;	/* b**4 */
+		break;
+	    case 5:
+		wResult *= wResult;	/* b**4 */
+		wResult *= w1;		/* b**5 */
+		break;
+	    case 6:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		break;
+	    case 7:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		wResult *= w1;		/* b**7 */
+		break;
+	    case 8:
+		wResult *= wResult;	/* b**4 */
+		wResult *= wResult;	/* b**8 */
+		break;
+	    case 9:
+		wResult *= wResult;	/* b**4 */
+		wResult *= wResult;	/* b**8 */
+		wResult *= w1;		/* b**9 */
+		break;
+	    case 10:
+		wResult *= wResult;	/* b**4 */
+		wResult *= w1;		/* b**5 */
+		wResult *= wResult;	/* b**10 */
+		break;
+	    case 11:
+		wResult *= wResult;	/* b**4 */
+		wResult *= w1;		/* b**5 */
+		wResult *= wResult;	/* b**10 */
+		wResult *= w1;		/* b**11 */
+		break;
+	    case 12:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		wResult *= wResult;	/* b**12 */
+		break;
+	    case 13:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		wResult *= wResult;	/* b**12 */
+		wResult *= w1;		/* b**13 */
+		break;
+	    case 14:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		wResult *= w1;		/* b**7 */
+		wResult *= wResult;	/* b**14 */
+		break;
+	    case 15:
+		wResult *= w1;		/* b**3 */
+		wResult *= wResult;	/* b**6 */
+		wResult *= w1;		/* b**7 */
+		wResult *= wResult;	/* b**14 */
+		wResult *= w1;		/* b**15 */
+		break;
+	    case 16:
+		wResult *= wResult;	/* b**4 */
+		wResult *= wResult;	/* b**8 */
+		wResult *= wResult;	/* b**16 */
+		break;
+	    }
+	    WIDE_RESULT(wResult);
+	}
+
+	/*
+	 * Handle cases of powers > 16 that still fit in a 64-bit word by
+	 * doing table lookup.
+	 */
+
+	if (w1 - 3 >= 0 && w1 - 2 < (long)Exp64IndexSize
+		&& l2 - 2 < (long)(Exp64ValueSize + MaxBase64Size)) {
+	    base = Exp64Index[w1 - 3]
+		    + (unsigned short) (l2 - 2 - MaxBase64Size);
+	    if (base < Exp64Index[w1 - 2]) {
+		/*
+		 * 64-bit number raised to intermediate power, done by
+		 * table lookup.
+		 */
+
+		WIDE_RESULT(Exp64Value[base]);
+	    }
+	}
+
+	if (-w1 - 3 >= 0 && -w1 - 2 < (long)Exp64IndexSize
+		&& l2 - 2 < (long)(Exp64ValueSize + MaxBase64Size)) {
+	    base = Exp64Index[-w1 - 3]
+		    + (unsigned short) (l2 - 2 - MaxBase64Size);
+	    if (base < Exp64Index[-w1 - 2]) {
+		/*
+		 * 64-bit number raised to intermediate power, done by
+		 * table lookup.
+		 */
+
+		wResult = oddExponent ? -Exp64Value[base] : Exp64Value[base];
+		WIDE_RESULT(wResult);
+	    }
+	}
+#endif
+
+    overflowExpon:
+	Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	if (big2.used > 1) {
+	    mp_clear(&big2);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "exponent too large", -1));
+	    return GENERAL_ARITHMETIC_ERROR;
+	}
+	Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+	mp_init(&bigResult);
+	mp_expt_d_ex(&big1, big2.dp[0], &bigResult, 1);
+	mp_clear(&big1);
+	mp_clear(&big2);
+	BIG_RESULT(&bigResult);
+    }
+
+    case INST_ADD:
+    case INST_SUB:
+    case INST_MULT:
+    case INST_DIV:
+	if ((type1 == TCL_NUMBER_DOUBLE) || (type2 == TCL_NUMBER_DOUBLE)) {
+	    /*
+	     * At least one of the values is floating-point, so perform
+	     * floating point calculations.
+	     */
+
+	    Tcl_GetDoubleFromObj(NULL, valuePtr, &d1);
+	    Tcl_GetDoubleFromObj(NULL, value2Ptr, &d2);
+
+	    switch (opcode) {
+	    case INST_ADD:
+		dResult = d1 + d2;
+		break;
+	    case INST_SUB:
+		dResult = d1 - d2;
+		break;
+	    case INST_MULT:
+		dResult = d1 * d2;
+		break;
+	    case INST_DIV:
+#ifndef IEEE_FLOATING_POINT
+		if (d2 == 0.0) {
+		    return DIVIDED_BY_ZERO;
+		}
+#endif
+		/*
+		 * We presume that we are running with zero-divide unmasked if
+		 * we're on an IEEE box. Otherwise, this statement might cause
+		 * demons to fly out our noses.
+		 */
+
+		dResult = d1 / d2;
+		break;
+	    default:
+		/* Unused, here to silence compiler warning. */
+		dResult = 0;
+	    }
+
+	doubleResult:
+#ifndef ACCEPT_NAN
+	    /*
+	     * Check now for IEEE floating-point error.
+	     */
+
+	    if (TclIsNaN(dResult)) {
+		TclExprFloatError(interp, dResult);
+		return GENERAL_ARITHMETIC_ERROR;
+	    }
+#endif
+	    DOUBLE_RESULT(dResult);
+	}
+	if ((type1 != TCL_NUMBER_BIG) && (type2 != TCL_NUMBER_BIG)) {
+	    TclGetWideIntFromObj(NULL, valuePtr, &w1);
+	    TclGetWideIntFromObj(NULL, value2Ptr, &w2);
+
+	    switch (opcode) {
+	    case INST_ADD:
+		wResult = w1 + w2;
+#ifndef TCL_WIDE_INT_IS_LONG
+		if ((type1 == TCL_NUMBER_WIDE) || (type2 == TCL_NUMBER_WIDE))
+#endif
+		{
+		    /*
+		     * Check for overflow.
+		     */
+
+		    if (Overflowing(w1, w2, wResult)) {
+			goto overflowBasic;
+		    }
+		}
+		break;
+
+	    case INST_SUB:
+		wResult = w1 - w2;
+#ifndef TCL_WIDE_INT_IS_LONG
+		if ((type1 == TCL_NUMBER_WIDE) || (type2 == TCL_NUMBER_WIDE))
+#endif
+		{
+		    /*
+		     * Must check for overflow. The macro tests for overflows
+		     * in sums by looking at the sign bits. As we have a
+		     * subtraction here, we are adding -w2. As -w2 could in
+		     * turn overflow, we test with ~w2 instead: it has the
+		     * opposite sign bit to w2 so it does the job. Note that
+		     * the only "bad" case (w2==0) is irrelevant for this
+		     * macro, as in that case w1 and wResult have the same
+		     * sign and there is no overflow anyway.
+		     */
+
+		    if (Overflowing(w1, ~w2, wResult)) {
+			goto overflowBasic;
+		    }
+		}
+		break;
+
+	    case INST_MULT:
+		if ((type1 != TCL_NUMBER_LONG) || (type2 != TCL_NUMBER_LONG)
+			|| (sizeof(Tcl_WideInt) < 2*sizeof(long))) {
+		    goto overflowBasic;
+		}
+		wResult = w1 * w2;
+		break;
+
+	    case INST_DIV:
+		if (w2 == 0) {
+		    return DIVIDED_BY_ZERO;
+		}
+
+		/*
+		 * Need a bignum to represent (LLONG_MIN / -1)
+		 */
+
+		if ((w1 == LLONG_MIN) && (w2 == -1)) {
+		    goto overflowBasic;
+		}
+		wResult = w1 / w2;
+
+		/*
+		 * Force Tcl's integer division rules.
+		 * TODO: examine for logic simplification
+		 */
+
+		if (((wResult < 0) || ((wResult == 0) &&
+			((w1 < 0 && w2 > 0) || (w1 > 0 && w2 < 0)))) &&
+			(wResult*w2 != w1)) {
+		    wResult -= 1;
+		}
+		break;
+
+	    default:
+		/*
+		 * Unused, here to silence compiler warning.
+		 */
+
+		wResult = 0;
+	    }
+
+	    WIDE_RESULT(wResult);
+	}
+
+    overflowBasic:
+	Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+	Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	mp_init(&bigResult);
+	switch (opcode) {
+	case INST_ADD:
+	    mp_add(&big1, &big2, &bigResult);
+	    break;
+	case INST_SUB:
+	    mp_sub(&big1, &big2, &bigResult);
+	    break;
+	case INST_MULT:
+	    mp_mul(&big1, &big2, &bigResult);
+	    break;
+	case INST_DIV:
+	    if (mp_iszero(&big2)) {
+		mp_clear(&big1);
+		mp_clear(&big2);
+		mp_clear(&bigResult);
+		return DIVIDED_BY_ZERO;
+	    }
+	    mp_init(&bigRemainder);
+	    mp_div(&big1, &big2, &bigResult, &bigRemainder);
+	    /* TODO: internals intrusion */
+	    if (!mp_iszero(&bigRemainder)
+		    && (bigRemainder.sign != big2.sign)) {
+		/*
+		 * Convert to Tcl's integer division rules.
+		 */
+
+		mp_sub_d(&bigResult, 1, &bigResult);
+		mp_add(&bigRemainder, &big2, &bigRemainder);
+	    }
+	    mp_clear(&bigRemainder);
+	    break;
+	}
+	mp_clear(&big1);
+	mp_clear(&big2);
+	BIG_RESULT(&bigResult);
+    }
+
+    Tcl_Panic("unexpected opcode");
+    return NULL;
+}
+
+static Tcl_Obj *
+ExecuteExtendedUnaryMathOp(
+    int opcode,			/* What operation to perform. */
+    Tcl_Obj *valuePtr)		/* The operand on the stack. */
+{
+    ClientData ptr;
+    int type;
+    Tcl_WideInt w;
+    mp_int big;
+    Tcl_Obj *objResultPtr;
+
+    (void) GetNumberFromObj(NULL, valuePtr, &ptr, &type);
+
+    switch (opcode) {
+    case INST_BITNOT:
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (type == TCL_NUMBER_WIDE) {
+	    w = *((const Tcl_WideInt *) ptr);
+	    WIDE_RESULT(~w);
+	}
+#endif
+	Tcl_TakeBignumFromObj(NULL, valuePtr, &big);
+	/* ~a = - a - 1 */
+	mp_neg(&big, &big);
+	mp_sub_d(&big, 1, &big);
+	BIG_RESULT(&big);
+    case INST_UMINUS:
+	switch (type) {
+	case TCL_NUMBER_DOUBLE:
+	    DOUBLE_RESULT(-(*((const double *) ptr)));
+	case TCL_NUMBER_LONG:
+	    w = (Tcl_WideInt) (*((const long *) ptr));
+	    if (w != LLONG_MIN) {
+		WIDE_RESULT(-w);
+	    }
+	    TclBNInitBignumFromLong(&big, *(const long *) ptr);
+	    break;
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+	    w = *((const Tcl_WideInt *) ptr);
+	    if (w != LLONG_MIN) {
+		WIDE_RESULT(-w);
+	    }
+	    TclBNInitBignumFromWideInt(&big, w);
+	    break;
+#endif
+	default:
+	    Tcl_TakeBignumFromObj(NULL, valuePtr, &big);
+	}
+	mp_neg(&big, &big);
+	BIG_RESULT(&big);
+    }
+
+    Tcl_Panic("unexpected opcode");
+    return NULL;
+}
+#undef LONG_RESULT
+#undef WIDE_RESULT
+#undef BIG_RESULT
+#undef DOUBLE_RESULT
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompareTwoNumbers --
+ *
+ *	This function compares a pair of numbers in Tcl_Objs. Each argument
+ *	must already be known to be numeric and not NaN.
+ *
+ * Results:
+ *	One of MP_LT, MP_EQ or MP_GT, depending on whether valuePtr is less
+ *	than, equal to, or greater than value2Ptr (respectively).
+ *
+ * Side effects:
+ *	None, provided both values are numeric.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompareTwoNumbers(
+    Tcl_Obj *valuePtr,
+    Tcl_Obj *value2Ptr)
+{
+    int type1 = TCL_NUMBER_NAN, type2 = TCL_NUMBER_NAN, compare;
+    ClientData ptr1, ptr2;
+    mp_int big1, big2;
+    double d1, d2, tmp;
+    long l1, l2;
+#ifndef TCL_WIDE_INT_IS_LONG
+    Tcl_WideInt w1, w2;
+#endif
+
+    (void) GetNumberFromObj(NULL, valuePtr, &ptr1, &type1);
+    (void) GetNumberFromObj(NULL, value2Ptr, &ptr2, &type2);
+
+    switch (type1) {
+    case TCL_NUMBER_LONG:
+	l1 = *((const long *)ptr1);
+	switch (type2) {
+	case TCL_NUMBER_LONG:
+	    l2 = *((const long *)ptr2);
+	longCompare:
+	    return (l1 < l2) ? MP_LT : ((l1 > l2) ? MP_GT : MP_EQ);
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+	    w2 = *((const Tcl_WideInt *)ptr2);
+	    w1 = (Tcl_WideInt)l1;
+	    goto wideCompare;
+#endif
+	case TCL_NUMBER_DOUBLE:
+	    d2 = *((const double *)ptr2);
+	    d1 = (double) l1;
+
+	    /*
+	     * If the double has a fractional part, or if the long can be
+	     * converted to double without loss of precision, then compare as
+	     * doubles.
+	     */
+
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(long) || l1 == (long) d1
+		    || modf(d2, &tmp) != 0.0) {
+		goto doubleCompare;
+	    }
+
+	    /*
+	     * Otherwise, to make comparision based on full precision, need to
+	     * convert the double to a suitably sized integer.
+	     *
+	     * Need this to get comparsions like
+	     *	  expr 20000000000000003 < 20000000000000004.0
+	     * right. Converting the first argument to double will yield two
+	     * double values that are equivalent within double precision.
+	     * Converting the double to an integer gets done exactly, then
+	     * integer comparison can tell the difference.
+	     */
+
+	    if (d2 < (double)LONG_MIN) {
+		return MP_GT;
+	    }
+	    if (d2 > (double)LONG_MAX) {
+		return MP_LT;
+	    }
+	    l2 = (long) d2;
+	    goto longCompare;
+	case TCL_NUMBER_BIG:
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	    if (mp_cmp_d(&big2, 0) == MP_LT) {
+		compare = MP_GT;
+	    } else {
+		compare = MP_LT;
+	    }
+	    mp_clear(&big2);
+	    return compare;
+	}
+
+#ifndef TCL_WIDE_INT_IS_LONG
+    case TCL_NUMBER_WIDE:
+	w1 = *((const Tcl_WideInt *)ptr1);
+	switch (type2) {
+	case TCL_NUMBER_WIDE:
+	    w2 = *((const Tcl_WideInt *)ptr2);
+	wideCompare:
+	    return (w1 < w2) ? MP_LT : ((w1 > w2) ? MP_GT : MP_EQ);
+	case TCL_NUMBER_LONG:
+	    l2 = *((const long *)ptr2);
+	    w2 = (Tcl_WideInt)l2;
+	    goto wideCompare;
+	case TCL_NUMBER_DOUBLE:
+	    d2 = *((const double *)ptr2);
+	    d1 = (double) w1;
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(Tcl_WideInt)
+		    || w1 == (Tcl_WideInt) d1 || modf(d2, &tmp) != 0.0) {
+		goto doubleCompare;
+	    }
+	    if (d2 < (double)LLONG_MIN) {
+		return MP_GT;
+	    }
+	    if (d2 > (double)LLONG_MAX) {
+		return MP_LT;
+	    }
+	    w2 = (Tcl_WideInt) d2;
+	    goto wideCompare;
+	case TCL_NUMBER_BIG:
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	    if (mp_cmp_d(&big2, 0) == MP_LT) {
+		compare = MP_GT;
+	    } else {
+		compare = MP_LT;
+	    }
+	    mp_clear(&big2);
+	    return compare;
+	}
+#endif
+
+    case TCL_NUMBER_DOUBLE:
+	d1 = *((const double *)ptr1);
+	switch (type2) {
+	case TCL_NUMBER_DOUBLE:
+	    d2 = *((const double *)ptr2);
+	doubleCompare:
+	    return (d1 < d2) ? MP_LT : ((d1 > d2) ? MP_GT : MP_EQ);
+	case TCL_NUMBER_LONG:
+	    l2 = *((const long *)ptr2);
+	    d2 = (double) l2;
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(long) || l2 == (long) d2
+		    || modf(d1, &tmp) != 0.0) {
+		goto doubleCompare;
+	    }
+	    if (d1 < (double)LONG_MIN) {
+		return MP_LT;
+	    }
+	    if (d1 > (double)LONG_MAX) {
+		return MP_GT;
+	    }
+	    l1 = (long) d1;
+	    goto longCompare;
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+	    w2 = *((const Tcl_WideInt *)ptr2);
+	    d2 = (double) w2;
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(Tcl_WideInt)
+		    || w2 == (Tcl_WideInt) d2 || modf(d1, &tmp) != 0.0) {
+		goto doubleCompare;
+	    }
+	    if (d1 < (double)LLONG_MIN) {
+		return MP_LT;
+	    }
+	    if (d1 > (double)LLONG_MAX) {
+		return MP_GT;
+	    }
+	    w1 = (Tcl_WideInt) d1;
+	    goto wideCompare;
+#endif
+	case TCL_NUMBER_BIG:
+	    if (TclIsInfinite(d1)) {
+		return (d1 > 0.0) ? MP_GT : MP_LT;
+	    }
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	    if ((d1 < (double)LONG_MAX) && (d1 > (double)LONG_MIN)) {
+		if (mp_cmp_d(&big2, 0) == MP_LT) {
+		    compare = MP_GT;
+		} else {
+		    compare = MP_LT;
+		}
+		mp_clear(&big2);
+		return compare;
+	    }
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(long)
+		    && modf(d1, &tmp) != 0.0) {
+		d2 = TclBignumToDouble(&big2);
+		mp_clear(&big2);
+		goto doubleCompare;
+	    }
+	    Tcl_InitBignumFromDouble(NULL, d1, &big1);
+	    goto bigCompare;
+	}
+
+    case TCL_NUMBER_BIG:
+	Tcl_TakeBignumFromObj(NULL, valuePtr, &big1);
+	switch (type2) {
+#ifndef TCL_WIDE_INT_IS_LONG
+	case TCL_NUMBER_WIDE:
+#endif
+	case TCL_NUMBER_LONG:
+	    compare = mp_cmp_d(&big1, 0);
+	    mp_clear(&big1);
+	    return compare;
+	case TCL_NUMBER_DOUBLE:
+	    d2 = *((const double *)ptr2);
+	    if (TclIsInfinite(d2)) {
+		compare = (d2 > 0.0) ? MP_LT : MP_GT;
+		mp_clear(&big1);
+		return compare;
+	    }
+	    if ((d2 < (double)LONG_MAX) && (d2 > (double)LONG_MIN)) {
+		compare = mp_cmp_d(&big1, 0);
+		mp_clear(&big1);
+		return compare;
+	    }
+	    if (DBL_MANT_DIG > CHAR_BIT*sizeof(long)
+		    && modf(d2, &tmp) != 0.0) {
+		d1 = TclBignumToDouble(&big1);
+		mp_clear(&big1);
+		goto doubleCompare;
+	    }
+	    Tcl_InitBignumFromDouble(NULL, d2, &big2);
+	    goto bigCompare;
+	case TCL_NUMBER_BIG:
+	    Tcl_TakeBignumFromObj(NULL, value2Ptr, &big2);
+	bigCompare:
+	    compare = mp_cmp(&big1, &big2);
+	    mp_clear(&big1);
+	    mp_clear(&big2);
+	    return compare;
+	}
+    default:
+	Tcl_Panic("unexpected number type");
+	return TCL_ERROR;
+    }
+}
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ *----------------------------------------------------------------------
+ *
+ * PrintByteCodeInfo --
+ *
+ *	This procedure prints a summary about a bytecode object to stdout. It
+ *	is called by TclNRExecuteByteCode when starting to execute the bytecode
+ *	object if tclTraceExec has the value 2 or more.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PrintByteCodeInfo(
+    register ByteCode *codePtr)	/* The bytecode whose summary is printed to
+				 * stdout. */
+{
+    Proc *procPtr = codePtr->procPtr;
+    Interp *iPtr = (Interp *) *codePtr->interpHandle;
+
+    fprintf(stdout, "\nExecuting ByteCode 0x%p, refCt %" TCL_LL_MODIFIER "u, epoch %" TCL_LL_MODIFIER "u, interp 0x%p (epoch %" TCL_LL_MODIFIER "u)\n",
+	    codePtr, (Tcl_WideInt)codePtr->refCount, (Tcl_WideInt)codePtr->compileEpoch, iPtr,
+	    (Tcl_WideInt)iPtr->compileEpoch);
+
+    fprintf(stdout, "  Source: ");
+    TclPrintSource(stdout, codePtr->source, 60);
+
+    fprintf(stdout, "\n  Cmds %d, src %d, inst %u, litObjs %u, aux %d, stkDepth %u, code/src %.2f\n",
+	    codePtr->numCommands, codePtr->numSrcBytes,
+	    codePtr->numCodeBytes, codePtr->numLitObjects,
+	    codePtr->numAuxDataItems, codePtr->maxStackDepth,
+#ifdef TCL_COMPILE_STATS
+	    codePtr->numSrcBytes?
+		    ((float)codePtr->structureSize)/codePtr->numSrcBytes :
+#endif
+	    0.0);
+
+#ifdef TCL_COMPILE_STATS
+    fprintf(stdout, "  Code %lu = header %lu+inst %d+litObj %lu+exc %lu+aux %lu+cmdMap %d\n",
+	    (unsigned long) codePtr->structureSize,
+	    (unsigned long) (sizeof(ByteCode)-sizeof(size_t)-sizeof(Tcl_Time)),
+	    codePtr->numCodeBytes,
+	    (unsigned long) (codePtr->numLitObjects * sizeof(Tcl_Obj *)),
+	    (unsigned long) (codePtr->numExceptRanges*sizeof(ExceptionRange)),
+	    (unsigned long) (codePtr->numAuxDataItems * sizeof(AuxData)),
+	    codePtr->numCmdLocBytes);
+#endif /* TCL_COMPILE_STATS */
+    if (procPtr != NULL) {
+	fprintf(stdout,
+		"  Proc 0x%p, refCt %d, args %d, compiled locals %d\n",
+		procPtr, procPtr->refCount, procPtr->numArgs,
+		procPtr->numCompiledLocals);
+    }
+}
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ValidatePcAndStackTop --
+ *
+ *	This procedure is called by TclNRExecuteByteCode when debugging to
+ *	verify that the program counter and stack top are valid during
+ *	execution.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Prints a message to stderr and panics if either the pc or stack top
+ *	are invalid.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+static void
+ValidatePcAndStackTop(
+    register ByteCode *codePtr,	/* The bytecode whose summary is printed to
+				 * stdout. */
+    const unsigned char *pc,	/* Points to first byte of a bytecode
+				 * instruction. The program counter. */
+    int stackTop,		/* Current stack top. Must be between
+				 * stackLowerBound and stackUpperBound
+				 * (inclusive). */
+    int checkStack)		/* 0 if the stack depth check should be
+				 * skipped. */
+{
+    int stackUpperBound = codePtr->maxStackDepth;
+				/* Greatest legal value for stackTop. */
+    unsigned relativePc = (unsigned) (pc - codePtr->codeStart);
+    unsigned long codeStart = (unsigned long) codePtr->codeStart;
+    unsigned long codeEnd = (unsigned long)
+	    (codePtr->codeStart + codePtr->numCodeBytes);
+    unsigned char opCode = *pc;
+
+    if (((unsigned long) pc < codeStart) || ((unsigned long) pc > codeEnd)) {
+	fprintf(stderr, "\nBad instruction pc 0x%p in TclNRExecuteByteCode\n",
+		pc);
+	Tcl_Panic("TclNRExecuteByteCode execution failure: bad pc");
+    }
+    if ((unsigned) opCode > LAST_INST_OPCODE) {
+	fprintf(stderr, "\nBad opcode %d at pc %u in TclNRExecuteByteCode\n",
+		(unsigned) opCode, relativePc);
+	Tcl_Panic("TclNRExecuteByteCode execution failure: bad opcode");
+    }
+    if (checkStack &&
+	    ((stackTop < 0) || (stackTop > stackUpperBound))) {
+	int numChars;
+	const char *cmd = GetSrcInfoForPc(pc, codePtr, &numChars, NULL, NULL);
+
+	fprintf(stderr, "\nBad stack top %d at pc %u in TclNRExecuteByteCode (min 0, max %i)",
+		stackTop, relativePc, stackUpperBound);
+	if (cmd != NULL) {
+	    Tcl_Obj *message;
+
+	    TclNewLiteralStringObj(message, "\n executing ");
+	    Tcl_IncrRefCount(message);
+	    Tcl_AppendLimitedToObj(message, cmd, numChars, 100, NULL);
+	    fprintf(stderr,"%s\n", TclGetString(message));
+	    Tcl_DecrRefCount(message);
+	} else {
+	    fprintf(stderr, "\n");
+	}
+	Tcl_Panic("TclNRExecuteByteCode execution failure: bad stack top");
+    }
+}
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * IllegalExprOperandType --
+ *
+ *	Used by TclNRExecuteByteCode to append an error message to the interp
+ *	result when an illegal operand type is detected by an expression
+ *	instruction. The argument opndPtr holds the operand object in error.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	An error message is appended to the interp result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+IllegalExprOperandType(
+    Tcl_Interp *interp,		/* Interpreter to which error information
+				 * pertains. */
+    const unsigned char *pc, /* Points to the instruction being executed
+				 * when the illegal type was found. */
+    Tcl_Obj *opndPtr)		/* Points to the operand holding the value
+				 * with the illegal type. */
+{
+    ClientData ptr;
+    int type;
+    const unsigned char opcode = *pc;
+    const char *description, *operator = "unknown";
+
+    if (opcode == INST_EXPON) {
+	operator = "**";
+    } else if (opcode <= INST_LNOT) {
+	operator = operatorStrings[opcode - INST_LOR];
+    }
+
+    if (GetNumberFromObj(NULL, opndPtr, &ptr, &type) != TCL_OK) {
+	int numBytes;
+	const char *bytes = TclGetStringFromObj(opndPtr, &numBytes);
+
+	if (numBytes == 0) {
+	    description = "empty string";
+	} else if (TclCheckBadOctal(NULL, bytes)) {
+	    description = "invalid octal number";
+	} else {
+	    description = "non-numeric string";
+	}
+    } else if (type == TCL_NUMBER_NAN) {
+	description = "non-numeric floating-point value";
+    } else if (type == TCL_NUMBER_DOUBLE) {
+	description = "floating-point value";
+    } else {
+	/* TODO: No caller needs this. Eliminate? */
+	description = "(big) integer";
+    }
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "can't use %s as operand of \"%s\"", description, operator));
+    Tcl_SetErrorCode(interp, "ARITH", "DOMAIN", description, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetSrcInfoForPc, GetSrcInfoForPc, TclGetSourceFromFrame --
+ *
+ *	Given a program counter value, finds the closest command in the
+ *	bytecode code unit's CmdLocation array and returns information about
+ *	that command's source: a pointer to its first byte and the number of
+ *	characters.
+ *
+ * Results:
+ *	If a command is found that encloses the program counter value, a
+ *	pointer to the command's source is returned and the length of the
+ *	source is stored at *lengthPtr. If multiple commands resulted in code
+ *	at pc, information about the closest enclosing command is returned. If
+ *	no matching command is found, NULL is returned and *lengthPtr is
+ *	unchanged.
+ *
+ * Side effects:
+ *	The CmdFrame at *cfPtr is updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetSourceFromFrame(
+    CmdFrame *cfPtr,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (cfPtr == NULL) {
+        return Tcl_NewListObj(objc, objv);
+    }
+    if (cfPtr->cmdObj == NULL) {
+        if (cfPtr->cmd == NULL) {
+	    ByteCode *codePtr = (ByteCode *) cfPtr->data.tebc.codePtr;
+
+            cfPtr->cmd = GetSrcInfoForPc((unsigned char *)
+		    cfPtr->data.tebc.pc, codePtr, &cfPtr->len, NULL, NULL);
+        }
+	if (cfPtr->cmd) {
+	    cfPtr->cmdObj = Tcl_NewStringObj(cfPtr->cmd, cfPtr->len);
+	} else {
+	    cfPtr->cmdObj = Tcl_NewListObj(objc, objv);
+	}
+        Tcl_IncrRefCount(cfPtr->cmdObj);
+    }
+    return cfPtr->cmdObj;
+}
+
+void
+TclGetSrcInfoForPc(
+    CmdFrame *cfPtr)
+{
+    ByteCode *codePtr = (ByteCode *) cfPtr->data.tebc.codePtr;
+
+    assert(cfPtr->type == TCL_LOCATION_BC);
+
+    if (cfPtr->cmd == NULL) {
+
+	cfPtr->cmd = GetSrcInfoForPc(
+		(unsigned char *) cfPtr->data.tebc.pc, codePtr,
+		&cfPtr->len, NULL, NULL);
+    }
+
+    if (cfPtr->cmd != NULL) {
+	/*
+	 * We now have the command. We can get the srcOffset back and from
+	 * there find the list of word locations for this command.
+	 */
+
+	ExtCmdLoc *eclPtr;
+	ECL *locPtr = NULL;
+	int srcOffset, i;
+	Interp *iPtr = (Interp *) *codePtr->interpHandle;
+	Tcl_HashEntry *hePtr =
+		Tcl_FindHashEntry(iPtr->lineBCPtr, codePtr);
+
+	if (!hePtr) {
+	    return;
+	}
+
+	srcOffset = cfPtr->cmd - codePtr->source;
+	eclPtr = Tcl_GetHashValue(hePtr);
+
+	for (i=0; i < eclPtr->nuloc; i++) {
+	    if (eclPtr->loc[i].srcOffset == srcOffset) {
+		locPtr = eclPtr->loc+i;
+		break;
+	    }
+	}
+	if (locPtr == NULL) {
+	    Tcl_Panic("LocSearch failure");
+	}
+
+	cfPtr->line = locPtr->line;
+	cfPtr->nline = locPtr->nline;
+	cfPtr->type = eclPtr->type;
+
+	if (eclPtr->type == TCL_LOCATION_SOURCE) {
+	    cfPtr->data.eval.path = eclPtr->path;
+	    Tcl_IncrRefCount(cfPtr->data.eval.path);
+	}
+
+	/*
+	 * Do not set cfPtr->data.eval.path NULL for non-SOURCE. Needed for
+	 * cfPtr->data.tebc.codePtr.
+	 */
+    }
+}
+
+static const char *
+GetSrcInfoForPc(
+    const unsigned char *pc,	/* The program counter value for which to
+				 * return the closest command's source info.
+				 * This points within a bytecode instruction
+				 * in codePtr's code. */
+    ByteCode *codePtr,		/* The bytecode sequence in which to look up
+				 * the command source for the pc. */
+    int *lengthPtr,		/* If non-NULL, the location where the length
+				 * of the command's source should be stored.
+				 * If NULL, no length is stored. */
+    const unsigned char **pcBeg,/* If non-NULL, the bytecode location
+				 * where the current instruction starts.
+				 * If NULL; no pointer is stored. */
+    int *cmdIdxPtr)		/* If non-NULL, the location where the index
+				 * of the command containing the pc should
+				 * be stored. */
+{
+    register int pcOffset = (pc - codePtr->codeStart);
+    int numCmds = codePtr->numCommands;
+    unsigned char *codeDeltaNext, *codeLengthNext;
+    unsigned char *srcDeltaNext, *srcLengthNext;
+    int codeOffset, codeLen, codeEnd, srcOffset, srcLen, delta, i;
+    int bestDist = INT_MAX;	/* Distance of pc to best cmd's start pc. */
+    int bestSrcOffset = -1;	/* Initialized to avoid compiler warning. */
+    int bestSrcLength = -1;	/* Initialized to avoid compiler warning. */
+    int bestCmdIdx = -1;
+
+    /* The pc must point within the bytecode */
+    assert ((pcOffset >= 0) && (pcOffset < codePtr->numCodeBytes));
+
+    /*
+     * Decode the code and source offset and length for each command. The
+     * closest enclosing command is the last one whose code started before
+     * pcOffset.
+     */
+
+    codeDeltaNext = codePtr->codeDeltaStart;
+    codeLengthNext = codePtr->codeLengthStart;
+    srcDeltaNext = codePtr->srcDeltaStart;
+    srcLengthNext = codePtr->srcLengthStart;
+    codeOffset = srcOffset = 0;
+    for (i = 0;  i < numCmds;  i++) {
+	if ((unsigned) *codeDeltaNext == (unsigned) 0xFF) {
+	    codeDeltaNext++;
+	    delta = TclGetInt4AtPtr(codeDeltaNext);
+	    codeDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(codeDeltaNext);
+	    codeDeltaNext++;
+	}
+	codeOffset += delta;
+
+	if ((unsigned) *codeLengthNext == (unsigned) 0xFF) {
+	    codeLengthNext++;
+	    codeLen = TclGetInt4AtPtr(codeLengthNext);
+	    codeLengthNext += 4;
+	} else {
+	    codeLen = TclGetInt1AtPtr(codeLengthNext);
+	    codeLengthNext++;
+	}
+	codeEnd = (codeOffset + codeLen - 1);
+
+	if ((unsigned) *srcDeltaNext == (unsigned) 0xFF) {
+	    srcDeltaNext++;
+	    delta = TclGetInt4AtPtr(srcDeltaNext);
+	    srcDeltaNext += 4;
+	} else {
+	    delta = TclGetInt1AtPtr(srcDeltaNext);
+	    srcDeltaNext++;
+	}
+	srcOffset += delta;
+
+	if ((unsigned) *srcLengthNext == (unsigned) 0xFF) {
+	    srcLengthNext++;
+	    srcLen = TclGetInt4AtPtr(srcLengthNext);
+	    srcLengthNext += 4;
+	} else {
+	    srcLen = TclGetInt1AtPtr(srcLengthNext);
+	    srcLengthNext++;
+	}
+
+	if (codeOffset > pcOffset) {	/* Best cmd already found */
+	    break;
+	}
+	if (pcOffset <= codeEnd) {	/* This cmd's code encloses pc */
+	    int dist = (pcOffset - codeOffset);
+
+	    if (dist <= bestDist) {
+		bestDist = dist;
+		bestSrcOffset = srcOffset;
+		bestSrcLength = srcLen;
+		bestCmdIdx = i;
+	    }
+	}
+    }
+
+    if (pcBeg != NULL) {
+	const unsigned char *curr, *prev;
+
+	/*
+	 * Walk from beginning of command or BC to pc, by complete
+	 * instructions. Stop when crossing pc; keep previous.
+	 */
+
+	curr = ((bestDist == INT_MAX) ? codePtr->codeStart : pc - bestDist);
+	prev = curr;
+	while (curr <= pc) {
+	    prev = curr;
+	    curr += tclInstructionTable[*curr].numBytes;
+	}
+	*pcBeg = prev;
+    }
+
+    if (bestDist == INT_MAX) {
+	return NULL;
+    }
+
+    if (lengthPtr != NULL) {
+	*lengthPtr = bestSrcLength;
+    }
+
+    if (cmdIdxPtr != NULL) {
+	*cmdIdxPtr = bestCmdIdx;
+    }
+
+    return (codePtr->source + bestSrcOffset);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetExceptRangeForPc --
+ *
+ *	Given a program counter value, return the closest enclosing
+ *	ExceptionRange.
+ *
+ * Results:
+ *	If the searchMode is TCL_ERROR, this procedure ignores loop exception
+ *	ranges and returns a pointer to the closest catch range. If the
+ *	searchMode is TCL_BREAK, this procedure returns a pointer to the most
+ *	closely enclosing ExceptionRange regardless of whether it is a loop or
+ *	catch exception range. If the searchMode is TCL_CONTINUE, this
+ *	procedure returns a pointer to the most closely enclosing
+ *	ExceptionRange (of any type) skipping only loop exception ranges if
+ *	they don't have a sensible continueOffset defined. If no matching
+ *	ExceptionRange is found that encloses pc, a NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ExceptionRange *
+GetExceptRangeForPc(
+    const unsigned char *pc,	/* The program counter value for which to
+				 * search for a closest enclosing exception
+				 * range. This points to a bytecode
+				 * instruction in codePtr's code. */
+    int searchMode,		/* If TCL_BREAK, consider either loop or catch
+				 * ExceptionRanges in search. If TCL_ERROR
+				 * consider only catch ranges (and ignore any
+				 * closer loop ranges). If TCL_CONTINUE, look
+				 * for loop ranges that define a continue
+				 * point or a catch range. */
+    ByteCode *codePtr)		/* Points to the ByteCode in which to search
+				 * for the enclosing ExceptionRange. */
+{
+    ExceptionRange *rangeArrayPtr;
+    int numRanges = codePtr->numExceptRanges;
+    register ExceptionRange *rangePtr;
+    int pcOffset = pc - codePtr->codeStart;
+    register int start;
+
+    if (numRanges == 0) {
+	return NULL;
+    }
+
+    /*
+     * This exploits peculiarities of our compiler: nested ranges are always
+     * *after* their containing ranges, so that by scanning backwards we are
+     * sure that the first matching range is indeed the deepest.
+     */
+
+    rangeArrayPtr = codePtr->exceptArrayPtr;
+    rangePtr = rangeArrayPtr + numRanges;
+    while (--rangePtr >= rangeArrayPtr) {
+	start = rangePtr->codeOffset;
+	if ((start <= pcOffset) &&
+		(pcOffset < (start + rangePtr->numCodeBytes))) {
+	    if (rangePtr->type == CATCH_EXCEPTION_RANGE) {
+		return rangePtr;
+	    }
+	    if (searchMode == TCL_BREAK) {
+		return rangePtr;
+	    }
+	    if (searchMode == TCL_CONTINUE && rangePtr->continueOffset != -1){
+		return rangePtr;
+	    }
+	}
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetOpcodeName --
+ *
+ *	This procedure is called by the TRACE and TRACE_WITH_OBJ macros used
+ *	in TclNRExecuteByteCode when debugging. It returns the name of the
+ *	bytecode instruction at a specified instruction pc.
+ *
+ * Results:
+ *	A character string for the instruction.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_COMPILE_DEBUG
+static const char *
+GetOpcodeName(
+    const unsigned char *pc)	/* Points to the instruction whose name should
+				 * be returned. */
+{
+    unsigned char opCode = *pc;
+
+    return tclInstructionTable[opCode].name;
+}
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclExprFloatError --
+ *
+ *	This procedure is called when an error occurs during a floating-point
+ *	operation. It reads errno and sets interp->objResultPtr accordingly.
+ *
+ * Results:
+ *	interp->objResultPtr is set to hold an error message.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclExprFloatError(
+    Tcl_Interp *interp,		/* Where to store error message. */
+    double value)		/* Value returned after error; used to
+				 * distinguish underflows from overflows. */
+{
+    const char *s;
+
+    if ((errno == EDOM) || TclIsNaN(value)) {
+	s = "domain error: argument not in valid range";
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(s, -1));
+	Tcl_SetErrorCode(interp, "ARITH", "DOMAIN", s, NULL);
+    } else if ((errno == ERANGE) || TclIsInfinite(value)) {
+	if (value == 0.0) {
+	    s = "floating-point value too small to represent";
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(s, -1));
+	    Tcl_SetErrorCode(interp, "ARITH", "UNDERFLOW", s, NULL);
+	} else {
+	    s = "floating-point value too large to represent";
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(s, -1));
+	    Tcl_SetErrorCode(interp, "ARITH", "OVERFLOW", s, NULL);
+	}
+    } else {
+	Tcl_Obj *objPtr = Tcl_ObjPrintf(
+		"unknown floating-point error, errno = %d", errno);
+
+	Tcl_SetErrorCode(interp, "ARITH", "UNKNOWN",
+		TclGetString(objPtr), NULL);
+	Tcl_SetObjResult(interp, objPtr);
+    }
+}
+
+#ifdef TCL_COMPILE_STATS
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLog2 --
+ *
+ *	Procedure used while collecting compilation statistics to determine
+ *	the log base 2 of an integer.
+ *
+ * Results:
+ *	Returns the log base 2 of the operand. If the argument is less than or
+ *	equal to zero, a zero is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclLog2(
+    register int value)		/* The integer for which to compute the log
+				 * base 2. */
+{
+    register int n = value;
+    register int result = 0;
+
+    while (n > 1) {
+	n = n >> 1;
+	result++;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EvalStatsCmd --
+ *
+ *	Implements the "evalstats" command that prints instruction execution
+ *	counts to stdout.
+ *
+ * Results:
+ *	Standard Tcl results.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+EvalStatsCmd(
+    ClientData unused,		/* Unused. */
+    Tcl_Interp *interp,		/* The current interpreter. */
+    int objc,			/* The number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument strings. */
+{
+    Interp *iPtr = (Interp *) interp;
+    LiteralTable *globalTablePtr = &iPtr->literalTable;
+    ByteCodeStats *statsPtr = &iPtr->stats;
+    double totalCodeBytes, currentCodeBytes;
+    double totalLiteralBytes, currentLiteralBytes;
+    double objBytesIfUnshared, strBytesIfUnshared, sharingBytesSaved;
+    double strBytesSharedMultX, strBytesSharedOnce;
+    double numInstructions, currentHeaderBytes;
+    long numCurrentByteCodes, numByteCodeLits;
+    long refCountSum, literalMgmtBytes, sum;
+    int numSharedMultX, numSharedOnce;
+    int decadeHigh, minSizeDecade, maxSizeDecade, length, i;
+    char *litTableStats;
+    LiteralEntry *entryPtr;
+    Tcl_Obj *objPtr;
+
+#define Percent(a,b) ((a) * 100.0 / (b))
+
+    objPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(objPtr);
+
+    numInstructions = 0.0;
+    for (i = 0;  i < 256;  i++) {
+	if (statsPtr->instructionCount[i] != 0) {
+	    numInstructions += statsPtr->instructionCount[i];
+	}
+    }
+
+    totalLiteralBytes = sizeof(LiteralTable)
+	    + iPtr->literalTable.numBuckets * sizeof(LiteralEntry *)
+	    + (statsPtr->numLiteralsCreated * sizeof(LiteralEntry))
+	    + (statsPtr->numLiteralsCreated * sizeof(Tcl_Obj))
+	    + statsPtr->totalLitStringBytes;
+    totalCodeBytes = statsPtr->totalByteCodeBytes + totalLiteralBytes;
+
+    numCurrentByteCodes =
+	    statsPtr->numCompilations - statsPtr->numByteCodesFreed;
+    currentHeaderBytes = numCurrentByteCodes
+	    * (sizeof(ByteCode) - sizeof(size_t) - sizeof(Tcl_Time));
+    literalMgmtBytes = sizeof(LiteralTable)
+	    + (iPtr->literalTable.numBuckets * sizeof(LiteralEntry *))
+	    + (iPtr->literalTable.numEntries * sizeof(LiteralEntry));
+    currentLiteralBytes = literalMgmtBytes
+	    + iPtr->literalTable.numEntries * sizeof(Tcl_Obj)
+	    + statsPtr->currentLitStringBytes;
+    currentCodeBytes = statsPtr->currentByteCodeBytes + currentLiteralBytes;
+
+    /*
+     * Summary statistics, total and current source and ByteCode sizes.
+     */
+
+    Tcl_AppendPrintfToObj(objPtr, "\n----------------------------------------------------------------\n");
+    Tcl_AppendPrintfToObj(objPtr,
+	    "Compilation and execution statistics for interpreter %#lx\n",
+	    (long int)iPtr);
+
+    Tcl_AppendPrintfToObj(objPtr, "\nNumber ByteCodes executed\t%ld\n",
+	    statsPtr->numExecutions);
+    Tcl_AppendPrintfToObj(objPtr, "Number ByteCodes compiled\t%ld\n",
+	    statsPtr->numCompilations);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean executions/compile\t%.1f\n",
+	    statsPtr->numExecutions / (float)statsPtr->numCompilations);
+
+    Tcl_AppendPrintfToObj(objPtr, "\nInstructions executed\t\t%.0f\n",
+	    numInstructions);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean inst/compile\t\t%.0f\n",
+	    numInstructions / statsPtr->numCompilations);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean inst/execution\t\t%.0f\n",
+	    numInstructions / statsPtr->numExecutions);
+
+    Tcl_AppendPrintfToObj(objPtr, "\nTotal ByteCodes\t\t\t%ld\n",
+	    statsPtr->numCompilations);
+    Tcl_AppendPrintfToObj(objPtr, "  Source bytes\t\t\t%.6g\n",
+	    statsPtr->totalSrcBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Code bytes\t\t\t%.6g\n",
+	    totalCodeBytes);
+    Tcl_AppendPrintfToObj(objPtr, "    ByteCode bytes\t\t%.6g\n",
+	    statsPtr->totalByteCodeBytes);
+    Tcl_AppendPrintfToObj(objPtr, "    Literal bytes\t\t%.6g\n",
+	    totalLiteralBytes);
+    Tcl_AppendPrintfToObj(objPtr, "      table %lu + bkts %lu + entries %lu + objects %lu + strings %.6g\n",
+	    (unsigned long) sizeof(LiteralTable),
+	    (unsigned long) (iPtr->literalTable.numBuckets * sizeof(LiteralEntry *)),
+	    (unsigned long) (statsPtr->numLiteralsCreated * sizeof(LiteralEntry)),
+	    (unsigned long) (statsPtr->numLiteralsCreated * sizeof(Tcl_Obj)),
+	    statsPtr->totalLitStringBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean code/compile\t\t%.1f\n",
+	    totalCodeBytes / statsPtr->numCompilations);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean code/source\t\t%.1f\n",
+	    totalCodeBytes / statsPtr->totalSrcBytes);
+
+    Tcl_AppendPrintfToObj(objPtr, "\nCurrent (active) ByteCodes\t%ld\n",
+	    numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "  Source bytes\t\t\t%.6g\n",
+	    statsPtr->currentSrcBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Code bytes\t\t\t%.6g\n",
+	    currentCodeBytes);
+    Tcl_AppendPrintfToObj(objPtr, "    ByteCode bytes\t\t%.6g\n",
+	    statsPtr->currentByteCodeBytes);
+    Tcl_AppendPrintfToObj(objPtr, "    Literal bytes\t\t%.6g\n",
+	    currentLiteralBytes);
+    Tcl_AppendPrintfToObj(objPtr, "      table %lu + bkts %lu + entries %lu + objects %lu + strings %.6g\n",
+	    (unsigned long) sizeof(LiteralTable),
+	    (unsigned long) (iPtr->literalTable.numBuckets * sizeof(LiteralEntry *)),
+	    (unsigned long) (iPtr->literalTable.numEntries * sizeof(LiteralEntry)),
+	    (unsigned long) (iPtr->literalTable.numEntries * sizeof(Tcl_Obj)),
+	    statsPtr->currentLitStringBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean code/source\t\t%.1f\n",
+	    currentCodeBytes / statsPtr->currentSrcBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Code + source bytes\t\t%.6g (%0.1f mean code/src)\n",
+	    (currentCodeBytes + statsPtr->currentSrcBytes),
+	    (currentCodeBytes / statsPtr->currentSrcBytes) + 1.0);
+
+    /*
+     * Tcl_IsShared statistics check
+     *
+     * This gives the refcount of each obj as Tcl_IsShared was called for it.
+     * Shared objects must be duplicated before they can be modified.
+     */
+
+    numSharedMultX = 0;
+    Tcl_AppendPrintfToObj(objPtr, "\nTcl_IsShared object check (all objects):\n");
+    Tcl_AppendPrintfToObj(objPtr, "  Object had refcount <=1 (not shared)\t%ld\n",
+	    tclObjsShared[1]);
+    for (i = 2;  i < TCL_MAX_SHARED_OBJ_STATS;  i++) {
+	Tcl_AppendPrintfToObj(objPtr, "  refcount ==%d\t\t%ld\n",
+		i, tclObjsShared[i]);
+	numSharedMultX += tclObjsShared[i];
+    }
+    Tcl_AppendPrintfToObj(objPtr, "  refcount >=%d\t\t%ld\n",
+	    i, tclObjsShared[0]);
+    numSharedMultX += tclObjsShared[0];
+    Tcl_AppendPrintfToObj(objPtr, "  Total shared objects\t\t\t%d\n",
+	    numSharedMultX);
+
+    /*
+     * Literal table statistics.
+     */
+
+    numByteCodeLits = 0;
+    refCountSum = 0;
+    numSharedMultX = 0;
+    numSharedOnce = 0;
+    objBytesIfUnshared = 0.0;
+    strBytesIfUnshared = 0.0;
+    strBytesSharedMultX = 0.0;
+    strBytesSharedOnce = 0.0;
+    for (i = 0;  i < globalTablePtr->numBuckets;  i++) {
+	for (entryPtr = globalTablePtr->buckets[i];  entryPtr != NULL;
+		entryPtr = entryPtr->nextPtr) {
+	    if (entryPtr->objPtr->typePtr == &tclByteCodeType) {
+		numByteCodeLits++;
+	    }
+	    (void) TclGetStringFromObj(entryPtr->objPtr, &length);
+	    refCountSum += entryPtr->refCount;
+	    objBytesIfUnshared += (entryPtr->refCount * sizeof(Tcl_Obj));
+	    strBytesIfUnshared += (entryPtr->refCount * (length+1));
+	    if (entryPtr->refCount > 1) {
+		numSharedMultX++;
+		strBytesSharedMultX += (length+1);
+	    } else {
+		numSharedOnce++;
+		strBytesSharedOnce += (length+1);
+	    }
+	}
+    }
+    sharingBytesSaved = (objBytesIfUnshared + strBytesIfUnshared)
+	    - currentLiteralBytes;
+
+    Tcl_AppendPrintfToObj(objPtr, "\nTotal objects (all interps)\t%ld\n",
+	    tclObjsAlloced);
+    Tcl_AppendPrintfToObj(objPtr, "Current objects\t\t\t%ld\n",
+	    (tclObjsAlloced - tclObjsFreed));
+    Tcl_AppendPrintfToObj(objPtr, "Total literal objects\t\t%ld\n",
+	    statsPtr->numLiteralsCreated);
+
+    Tcl_AppendPrintfToObj(objPtr, "\nCurrent literal objects\t\t%d (%0.1f%% of current objects)\n",
+	    globalTablePtr->numEntries,
+	    Percent(globalTablePtr->numEntries, tclObjsAlloced-tclObjsFreed));
+    Tcl_AppendPrintfToObj(objPtr, "  ByteCode literals\t\t%ld (%0.1f%% of current literals)\n",
+	    numByteCodeLits,
+	    Percent(numByteCodeLits, globalTablePtr->numEntries));
+    Tcl_AppendPrintfToObj(objPtr, "  Literals reused > 1x\t\t%d\n",
+	    numSharedMultX);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean reference count\t\t%.2f\n",
+	    ((double) refCountSum) / globalTablePtr->numEntries);
+    Tcl_AppendPrintfToObj(objPtr, "  Mean len, str reused >1x \t%.2f\n",
+	    (numSharedMultX ? strBytesSharedMultX/numSharedMultX : 0.0));
+    Tcl_AppendPrintfToObj(objPtr, "  Mean len, str used 1x\t\t%.2f\n",
+	    (numSharedOnce ? strBytesSharedOnce/numSharedOnce : 0.0));
+    Tcl_AppendPrintfToObj(objPtr, "  Total sharing savings\t\t%.6g (%0.1f%% of bytes if no sharing)\n",
+	    sharingBytesSaved,
+	    Percent(sharingBytesSaved, objBytesIfUnshared+strBytesIfUnshared));
+    Tcl_AppendPrintfToObj(objPtr, "    Bytes with sharing\t\t%.6g\n",
+	    currentLiteralBytes);
+    Tcl_AppendPrintfToObj(objPtr, "      table %lu + bkts %lu + entries %lu + objects %lu + strings %.6g\n",
+	    (unsigned long) sizeof(LiteralTable),
+	    (unsigned long) (iPtr->literalTable.numBuckets * sizeof(LiteralEntry *)),
+	    (unsigned long) (iPtr->literalTable.numEntries * sizeof(LiteralEntry)),
+	    (unsigned long) (iPtr->literalTable.numEntries * sizeof(Tcl_Obj)),
+	    statsPtr->currentLitStringBytes);
+    Tcl_AppendPrintfToObj(objPtr, "    Bytes if no sharing\t\t%.6g = objects %.6g + strings %.6g\n",
+	    (objBytesIfUnshared + strBytesIfUnshared),
+	    objBytesIfUnshared, strBytesIfUnshared);
+    Tcl_AppendPrintfToObj(objPtr, "  String sharing savings \t%.6g = unshared %.6g - shared %.6g\n",
+	    (strBytesIfUnshared - statsPtr->currentLitStringBytes),
+	    strBytesIfUnshared, statsPtr->currentLitStringBytes);
+    Tcl_AppendPrintfToObj(objPtr, "  Literal mgmt overhead\t\t%ld (%0.1f%% of bytes with sharing)\n",
+	    literalMgmtBytes,
+	    Percent(literalMgmtBytes, currentLiteralBytes));
+    Tcl_AppendPrintfToObj(objPtr, "    table %lu + buckets %lu + entries %lu\n",
+	    (unsigned long) sizeof(LiteralTable),
+	    (unsigned long) (iPtr->literalTable.numBuckets * sizeof(LiteralEntry *)),
+	    (unsigned long) (iPtr->literalTable.numEntries * sizeof(LiteralEntry)));
+
+    /*
+     * Breakdown of current ByteCode space requirements.
+     */
+
+    Tcl_AppendPrintfToObj(objPtr, "\nBreakdown of current ByteCode requirements:\n");
+    Tcl_AppendPrintfToObj(objPtr, "                         Bytes      Pct of    Avg per\n");
+    Tcl_AppendPrintfToObj(objPtr, "                                     total    ByteCode\n");
+    Tcl_AppendPrintfToObj(objPtr, "Total             %12.6g     100.00%%   %8.1f\n",
+	    statsPtr->currentByteCodeBytes,
+	    statsPtr->currentByteCodeBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Header            %12.6g   %8.1f%%   %8.1f\n",
+	    currentHeaderBytes,
+	    Percent(currentHeaderBytes, statsPtr->currentByteCodeBytes),
+	    currentHeaderBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Instructions      %12.6g   %8.1f%%   %8.1f\n",
+	    statsPtr->currentInstBytes,
+	    Percent(statsPtr->currentInstBytes,statsPtr->currentByteCodeBytes),
+	    statsPtr->currentInstBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Literal ptr array %12.6g   %8.1f%%   %8.1f\n",
+	    statsPtr->currentLitBytes,
+	    Percent(statsPtr->currentLitBytes,statsPtr->currentByteCodeBytes),
+	    statsPtr->currentLitBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Exception table   %12.6g   %8.1f%%   %8.1f\n",
+	    statsPtr->currentExceptBytes,
+	    Percent(statsPtr->currentExceptBytes,statsPtr->currentByteCodeBytes),
+	    statsPtr->currentExceptBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Auxiliary data    %12.6g   %8.1f%%   %8.1f\n",
+	    statsPtr->currentAuxBytes,
+	    Percent(statsPtr->currentAuxBytes,statsPtr->currentByteCodeBytes),
+	    statsPtr->currentAuxBytes / numCurrentByteCodes);
+    Tcl_AppendPrintfToObj(objPtr, "Command map       %12.6g   %8.1f%%   %8.1f\n",
+	    statsPtr->currentCmdMapBytes,
+	    Percent(statsPtr->currentCmdMapBytes,statsPtr->currentByteCodeBytes),
+	    statsPtr->currentCmdMapBytes / numCurrentByteCodes);
+
+    /*
+     * Detailed literal statistics.
+     */
+
+    Tcl_AppendPrintfToObj(objPtr, "\nLiteral string sizes:\n");
+    Tcl_AppendPrintfToObj(objPtr, "\t Up to length\t\tPercentage\n");
+    maxSizeDecade = 0;
+    for (i = 31;  i >= 0;  i--) {
+	if (statsPtr->literalCount[i] > 0) {
+	    maxSizeDecade = i;
+	    break;
+	}
+    }
+    sum = 0;
+    for (i = 0;  i <= maxSizeDecade;  i++) {
+	decadeHigh = (1 << (i+1)) - 1;
+	sum += statsPtr->literalCount[i];
+	Tcl_AppendPrintfToObj(objPtr, "\t%10d\t\t%8.0f%%\n",
+		decadeHigh, Percent(sum, statsPtr->numLiteralsCreated));
+    }
+
+    litTableStats = TclLiteralStats(globalTablePtr);
+    Tcl_AppendPrintfToObj(objPtr, "\nCurrent literal table statistics:\n%s\n",
+	    litTableStats);
+    ckfree(litTableStats);
+
+    /*
+     * Source and ByteCode size distributions.
+     */
+
+    Tcl_AppendPrintfToObj(objPtr, "\nSource sizes:\n");
+    Tcl_AppendPrintfToObj(objPtr, "\t Up to size\t\tPercentage\n");
+    minSizeDecade = maxSizeDecade = 0;
+    for (i = 0;  i < 31;  i++) {
+	if (statsPtr->srcCount[i] > 0) {
+	    minSizeDecade = i;
+	    break;
+	}
+    }
+    for (i = 31;  i >= 0;  i--) {
+	if (statsPtr->srcCount[i] > 0) {
+	    maxSizeDecade = i;
+	    break;
+	}
+    }
+    sum = 0;
+    for (i = minSizeDecade;  i <= maxSizeDecade;  i++) {
+	decadeHigh = (1 << (i+1)) - 1;
+	sum += statsPtr->srcCount[i];
+	Tcl_AppendPrintfToObj(objPtr, "\t%10d\t\t%8.0f%%\n",
+		decadeHigh, Percent(sum, statsPtr->numCompilations));
+    }
+
+    Tcl_AppendPrintfToObj(objPtr, "\nByteCode sizes:\n");
+    Tcl_AppendPrintfToObj(objPtr, "\t Up to size\t\tPercentage\n");
+    minSizeDecade = maxSizeDecade = 0;
+    for (i = 0;  i < 31;  i++) {
+	if (statsPtr->byteCodeCount[i] > 0) {
+	    minSizeDecade = i;
+	    break;
+	}
+    }
+    for (i = 31;  i >= 0;  i--) {
+	if (statsPtr->byteCodeCount[i] > 0) {
+	    maxSizeDecade = i;
+	    break;
+	}
+    }
+    sum = 0;
+    for (i = minSizeDecade;  i <= maxSizeDecade;  i++) {
+	decadeHigh = (1 << (i+1)) - 1;
+	sum += statsPtr->byteCodeCount[i];
+	Tcl_AppendPrintfToObj(objPtr, "\t%10d\t\t%8.0f%%\n",
+		decadeHigh, Percent(sum, statsPtr->numCompilations));
+    }
+
+    Tcl_AppendPrintfToObj(objPtr, "\nByteCode longevity (excludes Current ByteCodes):\n");
+    Tcl_AppendPrintfToObj(objPtr, "\t       Up to ms\t\tPercentage\n");
+    minSizeDecade = maxSizeDecade = 0;
+    for (i = 0;  i < 31;  i++) {
+	if (statsPtr->lifetimeCount[i] > 0) {
+	    minSizeDecade = i;
+	    break;
+	}
+    }
+    for (i = 31;  i >= 0;  i--) {
+	if (statsPtr->lifetimeCount[i] > 0) {
+	    maxSizeDecade = i;
+	    break;
+	}
+    }
+    sum = 0;
+    for (i = minSizeDecade;  i <= maxSizeDecade;  i++) {
+	decadeHigh = (1 << (i+1)) - 1;
+	sum += statsPtr->lifetimeCount[i];
+	Tcl_AppendPrintfToObj(objPtr, "\t%12.3f\t\t%8.0f%%\n",
+		decadeHigh/1000.0, Percent(sum, statsPtr->numByteCodesFreed));
+    }
+
+    /*
+     * Instruction counts.
+     */
+
+    Tcl_AppendPrintfToObj(objPtr, "\nInstruction counts:\n");
+    for (i = 0;  i <= LAST_INST_OPCODE;  i++) {
+	Tcl_AppendPrintfToObj(objPtr, "%20s %8ld ",
+		tclInstructionTable[i].name, statsPtr->instructionCount[i]);
+	if (statsPtr->instructionCount[i]) {
+	    Tcl_AppendPrintfToObj(objPtr, "%6.1f%%\n",
+		    Percent(statsPtr->instructionCount[i], numInstructions));
+	} else {
+	    Tcl_AppendPrintfToObj(objPtr, "0\n");
+	}
+    }
+
+#ifdef TCL_MEM_DEBUG
+    Tcl_AppendPrintfToObj(objPtr, "\nHeap Statistics:\n");
+    TclDumpMemoryInfo((ClientData) objPtr, 1);
+#endif
+    Tcl_AppendPrintfToObj(objPtr, "\n----------------------------------------------------------------\n");
+
+    if (objc == 1) {
+	Tcl_SetObjResult(interp, objPtr);
+    } else {
+	Tcl_Channel outChan;
+	char *str = TclGetStringFromObj(objv[1], &length);
+
+	if (length) {
+	    if (strcmp(str, "stdout") == 0) {
+		outChan = Tcl_GetStdChannel(TCL_STDOUT);
+	    } else if (strcmp(str, "stderr") == 0) {
+		outChan = Tcl_GetStdChannel(TCL_STDERR);
+	    } else {
+		outChan = Tcl_OpenFileChannel(NULL, str, "w", 0664);
+	    }
+	} else {
+	    outChan = Tcl_GetStdChannel(TCL_STDOUT);
+	}
+	if (outChan != NULL) {
+	    Tcl_WriteObj(outChan, objPtr);
+	}
+    }
+    Tcl_DecrRefCount(objPtr);
+    return TCL_OK;
+}
+#endif /* TCL_COMPILE_STATS */
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringForResultCode --
+ *
+ *	Procedure that returns a human-readable string representing a Tcl
+ *	result code such as TCL_ERROR.
+ *
+ * Results:
+ *	If the result code is one of the standard Tcl return codes, the result
+ *	is a string representing that code such as "TCL_ERROR". Otherwise, the
+ *	result string is that code formatted as a sequence of decimal digit
+ *	characters. Note that the resulting string must not be modified by the
+ *	caller.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *
+StringForResultCode(
+    int result)			/* The Tcl result code for which to generate a
+				 * string. */
+{
+    static char buf[TCL_INTEGER_SPACE];
+
+    if ((result >= TCL_OK) && (result <= TCL_CONTINUE)) {
+	return resultStrings[result];
+    }
+    TclFormatInt(buf, result);
+    return buf;
+}
+#endif /* TCL_COMPILE_DEBUG */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclFCmd.c b/generic/tclFCmd.c
new file mode 100644
index 0000000..80898fc
--- /dev/null
+++ b/generic/tclFCmd.c
@@ -0,0 +1,1507 @@
+/*
+ * tclFCmd.c
+ *
+ *	This file implements the generic portion of file manipulation
+ *	subcommands of the "file" command.
+ *
+ * Copyright (c) 1996-1998 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclFileSystem.h"
+
+/*
+ * Declarations for local functions defined in this file:
+ */
+
+static int		CopyRenameOneFile(Tcl_Interp *interp,
+			    Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr,
+			    int copyFlag, int force);
+static Tcl_Obj *	FileBasename(Tcl_Interp *interp, Tcl_Obj *pathPtr);
+static int		FileCopyRename(Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[], int copyFlag);
+static int		FileForceOption(Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[], int *forcePtr);
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFileRenameCmd
+ *
+ *	This function implements the "rename" subcommand of the "file"
+ *	command. Filename arguments need to be translated to native format
+ *	before being passed to platform-specific code that implements rename
+ *	functionality.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclFileRenameCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Interp for error reporting or recursive
+				 * calls in the case of a tricky rename. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument strings passed to Tcl_FileCmd. */
+{
+    return FileCopyRename(interp, objc, objv, 0);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFileCopyCmd
+ *
+ *	This function implements the "copy" subcommand of the "file" command.
+ *	Filename arguments need to be translated to native format before being
+ *	passed to platform-specific code that implements copy functionality.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclFileCopyCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Used for error reporting or recursive calls
+				 * in the case of a tricky copy. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument strings passed to Tcl_FileCmd. */
+{
+    return FileCopyRename(interp, objc, objv, 1);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FileCopyRename --
+ *
+ *	Performs the work of TclFileRenameCmd and TclFileCopyCmd. See
+ *	comments for those functions.
+ *
+ * Results:
+ *	See above.
+ *
+ * Side effects:
+ *	See above.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+FileCopyRename(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[],	/* Argument strings passed to Tcl_FileCmd. */
+    int copyFlag)		/* If non-zero, copy source(s). Otherwise,
+				 * rename them. */
+{
+    int i, result, force;
+    Tcl_StatBuf statBuf;
+    Tcl_Obj *target;
+
+    i = FileForceOption(interp, objc - 1, objv + 1, &force);
+    if (i < 0) {
+	return TCL_ERROR;
+    }
+    i++;
+    if ((objc - i) < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-option value ...? source ?source ...? target");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If target doesn't exist or isn't a directory, try the copy/rename. More
+     * than 2 arguments is only valid if the target is an existing directory.
+     */
+
+    target = objv[objc - 1];
+    if (Tcl_FSConvertToPathType(interp, target) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    result = TCL_OK;
+
+    /*
+     * Call Tcl_FSStat() so that if target is a symlink that points to a
+     * directory we will put the sources in that directory instead of
+     * overwriting the symlink.
+     */
+
+    if ((Tcl_FSStat(target, &statBuf) != 0) || !S_ISDIR(statBuf.st_mode)) {
+	if ((objc - i) > 2) {
+	    errno = ENOTDIR;
+	    Tcl_PosixError(interp);
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error %s: target \"%s\" is not a directory",
+		    (copyFlag?"copying":"renaming"), TclGetString(target)));
+	    result = TCL_ERROR;
+	} else {
+	    /*
+	     * Even though already have target == translated(objv[i+1]), pass
+	     * the original argument down, so if there's an error, the error
+	     * message will reflect the original arguments.
+	     */
+
+	    result = CopyRenameOneFile(interp, objv[i], objv[i + 1], copyFlag,
+		    force);
+	}
+	return result;
+    }
+
+    /*
+     * Move each source file into target directory. Extract the basename from
+     * each source, and append it to the end of the target path.
+     */
+
+    for ( ; i<objc-1 ; i++) {
+	Tcl_Obj *jargv[2];
+	Tcl_Obj *source, *newFileName;
+
+	source = FileBasename(interp, objv[i]);
+	if (source == NULL) {
+	    result = TCL_ERROR;
+	    break;
+	}
+	jargv[0] = objv[objc - 1];
+	jargv[1] = source;
+	newFileName = TclJoinPath(2, jargv);
+	Tcl_IncrRefCount(newFileName);
+	result = CopyRenameOneFile(interp, objv[i], newFileName, copyFlag,
+		force);
+	Tcl_DecrRefCount(newFileName);
+	Tcl_DecrRefCount(source);
+
+	if (result == TCL_ERROR) {
+	    break;
+	}
+    }
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFileMakeDirsCmd
+ *
+ *	This function implements the "mkdir" subcommand of the "file" command.
+ *	Filename arguments need to be translated to native format before being
+ *	passed to platform-specific code that implements mkdir functionality.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFileMakeDirsCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    int objc,			/* Number of arguments */
+    Tcl_Obj *const objv[])	/* Argument strings passed to Tcl_FileCmd. */
+{
+    Tcl_Obj *errfile = NULL;
+    int result, i, j, pobjc;
+    Tcl_Obj *split = NULL;
+    Tcl_Obj *target = NULL;
+    Tcl_StatBuf statBuf;
+
+    result = TCL_OK;
+    for (i = 1; i < objc; i++) {
+	if (Tcl_FSConvertToPathType(interp, objv[i]) != TCL_OK) {
+	    result = TCL_ERROR;
+	    break;
+	}
+
+	split = Tcl_FSSplitPath(objv[i], &pobjc);
+	Tcl_IncrRefCount(split);
+	if (pobjc == 0) {
+	    errno = ENOENT;
+	    errfile = objv[i];
+	    break;
+	}
+	for (j = 0; j < pobjc; j++) {
+	    target = Tcl_FSJoinPath(split, j + 1);
+	    Tcl_IncrRefCount(target);
+
+	    /*
+	     * Call Tcl_FSStat() so that if target is a symlink that points to
+	     * a directory we will create subdirectories in that directory.
+	     */
+
+	    if (Tcl_FSStat(target, &statBuf) == 0) {
+		if (!S_ISDIR(statBuf.st_mode)) {
+		    errno = EEXIST;
+		    errfile = target;
+		    goto done;
+		}
+	    } else if (errno != ENOENT) {
+		/*
+		 * If Tcl_FSStat() failed and the error is anything other than
+		 * non-existence of the target, throw the error.
+		 */
+
+		errfile = target;
+		goto done;
+	    } else if (Tcl_FSCreateDirectory(target) != TCL_OK) {
+		/*
+		 * Create might have failed because of being in a race
+		 * condition with another process trying to create the same
+		 * subdirectory.
+		 */
+
+		if (errno != EEXIST) {
+		    errfile = target;
+		    goto done;
+		} else if ((Tcl_FSStat(target, &statBuf) == 0)
+			&& S_ISDIR(statBuf.st_mode)) {
+		    /*
+		     * It is a directory that wasn't there before, so keep
+		     * going without error.
+		     */
+
+		    Tcl_ResetResult(interp);
+		} else {
+		    errfile = target;
+		    goto done;
+		}
+	    }
+
+	    /*
+	     * Forget about this sub-path.
+	     */
+
+	    Tcl_DecrRefCount(target);
+	    target = NULL;
+	}
+	Tcl_DecrRefCount(split);
+	split = NULL;
+    }
+
+  done:
+    if (errfile != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create directory \"%s\": %s",
+		TclGetString(errfile), Tcl_PosixError(interp)));
+	result = TCL_ERROR;
+    }
+    if (split != NULL) {
+	Tcl_DecrRefCount(split);
+    }
+    if (target != NULL) {
+	Tcl_DecrRefCount(target);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFileDeleteCmd
+ *
+ *	This function implements the "delete" subcommand of the "file"
+ *	command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFileDeleteCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Used for error reporting */
+    int objc,			/* Number of arguments */
+    Tcl_Obj *const objv[])	/* Argument strings passed to Tcl_FileCmd. */
+{
+    int i, force, result;
+    Tcl_Obj *errfile;
+    Tcl_Obj *errorBuffer = NULL;
+
+    i = FileForceOption(interp, objc - 1, objv + 1, &force);
+    if (i < 0) {
+	return TCL_ERROR;
+    }
+
+    errfile = NULL;
+    result = TCL_OK;
+
+    for (i++ ; i < objc; i++) {
+	Tcl_StatBuf statBuf;
+
+	errfile = objv[i];
+	if (Tcl_FSConvertToPathType(interp, objv[i]) != TCL_OK) {
+	    result = TCL_ERROR;
+	    goto done;
+	}
+
+	/*
+	 * Call lstat() to get info so can delete symbolic link itself.
+	 */
+
+	if (Tcl_FSLstat(objv[i], &statBuf) != 0) {
+	    /*
+	     * Trying to delete a file that does not exist is not considered
+	     * an error, just a no-op
+	     */
+
+	    if (errno != ENOENT) {
+		result = TCL_ERROR;
+	    }
+	} else if (S_ISDIR(statBuf.st_mode)) {
+	    /*
+	     * We own a reference count on errorBuffer, if it was set as a
+	     * result of this call.
+	     */
+
+	    result = Tcl_FSRemoveDirectory(objv[i], force, &errorBuffer);
+	    if (result != TCL_OK) {
+		if ((force == 0) && (errno == EEXIST)) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "error deleting \"%s\": directory not empty",
+			    TclGetString(objv[i])));
+		    Tcl_PosixError(interp);
+		    goto done;
+		}
+
+		/*
+		 * If possible, use the untranslated name for the file.
+		 */
+
+		errfile = errorBuffer;
+
+		/*
+		 * FS supposed to check between translated objv and errfile.
+		 */
+
+		if (Tcl_FSEqualPaths(objv[i], errfile)) {
+		    errfile = objv[i];
+		}
+	    }
+	} else {
+	    result = Tcl_FSDeleteFile(objv[i]);
+	}
+
+	if (result != TCL_OK) {
+	    result = TCL_ERROR;
+
+	    /*
+	     * It is important that we break on error, otherwise we might end
+	     * up owning reference counts on numerous errorBuffers.
+	     */
+
+	    break;
+	}
+    }
+    if (result != TCL_OK) {
+	if (errfile == NULL) {
+	    /*
+	     * We try to accomodate poor error results from our Tcl_FS calls.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error deleting unknown file: %s",
+		    Tcl_PosixError(interp)));
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error deleting \"%s\": %s",
+		    TclGetString(errfile), Tcl_PosixError(interp)));
+	}
+    }
+
+  done:
+    if (errorBuffer != NULL) {
+	Tcl_DecrRefCount(errorBuffer);
+    }
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * CopyRenameOneFile
+ *
+ *	Copies or renames specified source file or directory hierarchy to the
+ *	specified target.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Target is overwritten if the force flag is set. Attempting to
+ *	copy/rename a file onto a directory or a directory onto a file will
+ *	always result in an error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CopyRenameOneFile(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Obj *source,		/* Pathname of file to copy. May need to be
+				 * translated. */
+    Tcl_Obj *target,		/* Pathname of file to create/overwrite. May
+				 * need to be translated. */
+    int copyFlag,		/* If non-zero, copy files. Otherwise, rename
+				 * them. */
+    int force)			/* If non-zero, overwrite target file if it
+				 * exists. Otherwise, error if target already
+				 * exists. */
+{
+    int result;
+    Tcl_Obj *errfile, *errorBuffer;
+    Tcl_Obj *actualSource=NULL;	/* If source is a link, then this is the real
+				 * file/directory. */
+    Tcl_StatBuf sourceStatBuf, targetStatBuf;
+
+    if (Tcl_FSConvertToPathType(interp, source) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_FSConvertToPathType(interp, target) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    errfile = NULL;
+    errorBuffer = NULL;
+    result = TCL_ERROR;
+
+    /*
+     * We want to copy/rename links and not the files they point to, so we use
+     * lstat(). If target is a link, we also want to replace the link and not
+     * the file it points to, so we also use lstat() on the target.
+     */
+
+    if (Tcl_FSLstat(source, &sourceStatBuf) != 0) {
+	errfile = source;
+	goto done;
+    }
+    if (Tcl_FSLstat(target, &targetStatBuf) != 0) {
+	if (errno != ENOENT) {
+	    errfile = target;
+	    goto done;
+	}
+    } else {
+	if (force == 0) {
+	    errno = EEXIST;
+	    errfile = target;
+	    goto done;
+	}
+
+	/*
+	 * Prevent copying or renaming a file onto itself. On Windows since
+	 * 8.5 we do get an inode number, however the unsigned short field is
+	 * insufficient to accept the Win32 API file id so it is truncated to
+	 * 16 bits and we get collisions. See bug #2015723.
+	 */
+
+#if !defined(_WIN32) && !defined(__CYGWIN__)
+	if ((sourceStatBuf.st_ino != 0) && (targetStatBuf.st_ino != 0)) {
+	    if ((sourceStatBuf.st_ino == targetStatBuf.st_ino) &&
+		    (sourceStatBuf.st_dev == targetStatBuf.st_dev)) {
+		result = TCL_OK;
+		goto done;
+	    }
+	}
+#endif
+
+	/*
+	 * Prevent copying/renaming a file onto a directory and vice-versa.
+	 * This is a policy decision based on the fact that existing
+	 * implementations of copy and rename on all platforms also prevent
+	 * this.
+	 */
+
+	if (S_ISDIR(sourceStatBuf.st_mode)
+		&& !S_ISDIR(targetStatBuf.st_mode)) {
+	    errno = EISDIR;
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "can't overwrite file \"%s\" with directory \"%s\"",
+		    TclGetString(target), TclGetString(source)));
+	    goto done;
+	}
+	if (!S_ISDIR(sourceStatBuf.st_mode)
+		&& S_ISDIR(targetStatBuf.st_mode)) {
+	    errno = EISDIR;
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "can't overwrite directory \"%s\" with file \"%s\"",
+		    TclGetString(target), TclGetString(source)));
+	    goto done;
+	}
+
+	/*
+	 * The destination exists, but appears to be ok to over-write, and
+	 * -force is given. We now try to adjust permissions to ensure the
+	 * operation succeeds. If we can't adjust permissions, we'll let the
+	 * actual copy/rename return an error later.
+	 */
+
+	{
+	    Tcl_Obj *perm;
+	    int index;
+
+	    TclNewLiteralStringObj(perm, "u+w");
+	    Tcl_IncrRefCount(perm);
+	    if (TclFSFileAttrIndex(target, "-permissions", &index) == TCL_OK) {
+		Tcl_FSFileAttrsSet(NULL, index, target, perm);
+	    }
+	    Tcl_DecrRefCount(perm);
+	}
+    }
+
+    if (copyFlag == 0) {
+	result = Tcl_FSRenameFile(source, target);
+	if (result == TCL_OK) {
+	    goto done;
+	}
+
+	if (errno == EINVAL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error renaming \"%s\" to \"%s\": trying to rename a"
+		    " volume or move a directory into itself",
+		    TclGetString(source), TclGetString(target)));
+	    goto done;
+	} else if (errno != EXDEV) {
+	    errfile = target;
+	    goto done;
+	}
+
+	/*
+	 * The rename failed because the move was across file systems. Fall
+	 * through to copy file and then remove original. Note that the
+	 * low-level Tcl_FSRenameFileProc in the filesystem is allowed to
+	 * implement cross-filesystem moves itself, if it desires.
+	 */
+    }
+
+    actualSource = source;
+    Tcl_IncrRefCount(actualSource);
+
+    /*
+     * Activate the following block to copy files instead of links. However
+     * Tcl's semantics currently say we should copy links, so any such change
+     * should be the subject of careful study on the consequences.
+     *
+     * Perhaps there could be an optional flag to 'file copy' to dictate which
+     * approach to use, with the default being _not_ to have this block
+     * active.
+     */
+
+#if 0
+#ifdef S_ISLNK
+    if (copyFlag && S_ISLNK(sourceStatBuf.st_mode)) {
+	/*
+	 * We want to copy files not links. Therefore we must follow the link.
+	 * There are two purposes to this 'stat' call here. First we want to
+	 * know if the linked-file/dir actually exists, and second, in the
+	 * block of code which follows, some 20 lines down, we want to check
+	 * if the thing is a file or directory.
+	 */
+
+	if (Tcl_FSStat(source, &sourceStatBuf) != 0) {
+	    /*
+	     * Actual file doesn't exist.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error copying \"%s\": the target of this link doesn't"
+		    " exist", TclGetString(source)));
+	    goto done;
+	} else {
+	    int counter = 0;
+
+	    while (1) {
+		Tcl_Obj *path = Tcl_FSLink(actualSource, NULL, 0);
+		if (path == NULL) {
+		    break;
+		}
+
+		/*
+		 * Now we want to check if this is a relative path, and if so,
+		 * to make it absolute.
+		 */
+
+		if (Tcl_FSGetPathType(path) == TCL_PATH_RELATIVE) {
+		    Tcl_Obj *abs = Tcl_FSJoinToPath(actualSource, 1, &path);
+
+		    if (abs == NULL) {
+			break;
+		    }
+		    Tcl_IncrRefCount(abs);
+		    Tcl_DecrRefCount(path);
+		    path = abs;
+		}
+		Tcl_DecrRefCount(actualSource);
+		actualSource = path;
+		counter++;
+
+		/*
+		 * Arbitrary limit of 20 links to follow.
+		 */
+
+		if (counter > 20) {
+		    /*
+		     * Too many links.
+		     */
+
+		    Tcl_SetErrno(EMLINK);
+		    errfile = source;
+		    goto done;
+		}
+	    }
+	    /* Now 'actualSource' is the correct file */
+	}
+    }
+#endif /* S_ISLNK */
+#endif
+
+    if (S_ISDIR(sourceStatBuf.st_mode)) {
+	result = Tcl_FSCopyDirectory(actualSource, target, &errorBuffer);
+	if (result != TCL_OK) {
+	    if (errno == EXDEV) {
+		/*
+		 * The copy failed because we're trying to do a
+		 * cross-filesystem copy. We do this through our Tcl library.
+		 */
+
+		Tcl_Obj *copyCommand, *cmdObj, *opObj;
+
+		TclNewObj(copyCommand);
+		TclNewLiteralStringObj(cmdObj, "::tcl::CopyDirectory");
+		Tcl_ListObjAppendElement(interp, copyCommand, cmdObj);
+		if (copyFlag) {
+		    TclNewLiteralStringObj(opObj, "copying");
+		} else {
+		    TclNewLiteralStringObj(opObj, "renaming");
+		}
+		Tcl_ListObjAppendElement(interp, copyCommand, opObj);
+		Tcl_ListObjAppendElement(interp, copyCommand, source);
+		Tcl_ListObjAppendElement(interp, copyCommand, target);
+		Tcl_IncrRefCount(copyCommand);
+		result = Tcl_EvalObjEx(interp, copyCommand,
+			TCL_EVAL_GLOBAL | TCL_EVAL_DIRECT);
+		Tcl_DecrRefCount(copyCommand);
+		if (result != TCL_OK) {
+		    /*
+		     * There was an error in the Tcl-level copy. We will pass
+		     * on the Tcl error message and can ensure this by setting
+		     * errfile to NULL
+		     */
+
+		    errfile = NULL;
+		}
+	    } else {
+		errfile = errorBuffer;
+		if (Tcl_FSEqualPaths(errfile, source)) {
+		    errfile = source;
+		} else if (Tcl_FSEqualPaths(errfile, target)) {
+		    errfile = target;
+		}
+	    }
+	}
+    } else {
+	result = Tcl_FSCopyFile(actualSource, target);
+	if ((result != TCL_OK) && (errno == EXDEV)) {
+	    result = TclCrossFilesystemCopy(interp, source, target);
+	}
+	if (result != TCL_OK) {
+	    /*
+	     * We could examine 'errno' to double-check if the problem was
+	     * with the target, but we checked the source above, so it should
+	     * be quite clear
+	     */
+
+	    errfile = target;
+	}
+	/*
+	 * We now need to reset the result, because the above call,
+	 * may have left set it.  (Ideally we would prefer not to pass
+	 * an interpreter in above, but the channel IO code used by
+	 * TclCrossFilesystemCopy currently requires one)
+	 */
+	Tcl_ResetResult(interp);
+    }
+    if ((copyFlag == 0) && (result == TCL_OK)) {
+	if (S_ISDIR(sourceStatBuf.st_mode)) {
+	    result = Tcl_FSRemoveDirectory(source, 1, &errorBuffer);
+	    if (result != TCL_OK) {
+		errfile = errorBuffer;
+		if (Tcl_FSEqualPaths(errfile, source) == 0) {
+		    errfile = source;
+		}
+	    }
+	} else {
+	    result = Tcl_FSDeleteFile(source);
+	    if (result != TCL_OK) {
+		errfile = source;
+	    }
+	}
+	if (result != TCL_OK) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf("can't unlink \"%s\": %s",
+		    TclGetString(errfile), Tcl_PosixError(interp)));
+	    errfile = NULL;
+	}
+    }
+
+  done:
+    if (errfile != NULL) {
+	Tcl_Obj *errorMsg = Tcl_ObjPrintf("error %s \"%s\"",
+		(copyFlag ? "copying" : "renaming"), TclGetString(source));
+
+	if (errfile != source) {
+	    Tcl_AppendPrintfToObj(errorMsg, " to \"%s\"",
+		    TclGetString(target));
+	    if (errfile != target) {
+		Tcl_AppendPrintfToObj(errorMsg, ": \"%s\"",
+			TclGetString(errfile));
+	    }
+	}
+	Tcl_AppendPrintfToObj(errorMsg, ": %s", Tcl_PosixError(interp));
+	Tcl_SetObjResult(interp, errorMsg);
+    }
+    if (errorBuffer != NULL) {
+	Tcl_DecrRefCount(errorBuffer);
+    }
+    if (actualSource != NULL) {
+	Tcl_DecrRefCount(actualSource);
+    }
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FileForceOption --
+ *
+ *	Helps parse command line options for file commands that take the
+ *	"-force" and "--" options.
+ *
+ * Results:
+ *	The return value is how many arguments from argv were consumed by this
+ *	function, or -1 if there was an error parsing the options. If an error
+ *	occurred, an error message is left in the interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+FileForceOption(
+    Tcl_Interp *interp,		/* Interp, for error return. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[],	/* Argument strings.  First command line
+				 * option, if it exists, begins at 0. */
+    int *forcePtr)		/* If the "-force" was specified, *forcePtr is
+				 * filled with 1, otherwise with 0. */
+{
+    int force, i, idx;
+    static const char *const options[] = {
+	"-force", "--", NULL
+    };
+
+    force = 0;
+    for (i = 0; i < objc; i++) {
+	if (TclGetString(objv[i])[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", TCL_EXACT,
+		&idx) != TCL_OK) {
+	    return -1;
+	}
+	if (idx == 0 /* -force */) {
+	    force = 1;
+	} else { /* -- */
+	    i++;
+	    break;
+	}
+    }
+    *forcePtr = force;
+    return i;
+}
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FileBasename --
+ *
+ *	Given a path in either tcl format (with / separators), or in the
+ *	platform-specific format for the current platform, return all the
+ *	characters in the path after the last directory separator. But, if
+ *	path is the root directory, returns no characters.
+ *
+ * Results:
+ *	Returns the string object that represents the basename. If there is an
+ *	error, an error message is left in interp, and NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+FileBasename(
+    Tcl_Interp *interp,		/* Interp, for error return. */
+    Tcl_Obj *pathPtr)		/* Path whose basename to extract. */
+{
+    int objc;
+    Tcl_Obj *splitPtr;
+    Tcl_Obj *resultPtr = NULL;
+
+    splitPtr = Tcl_FSSplitPath(pathPtr, &objc);
+    Tcl_IncrRefCount(splitPtr);
+
+    if (objc != 0) {
+	if ((objc == 1) && (*TclGetString(pathPtr) == '~')) {
+	    Tcl_DecrRefCount(splitPtr);
+	    if (Tcl_FSConvertToPathType(interp, pathPtr) != TCL_OK) {
+		return NULL;
+	    }
+	    splitPtr = Tcl_FSSplitPath(pathPtr, &objc);
+	    Tcl_IncrRefCount(splitPtr);
+	}
+
+	/*
+	 * Return the last component, unless it is the only component, and it
+	 * is the root of an absolute path.
+	 */
+
+	if (objc > 0) {
+	    Tcl_ListObjIndex(NULL, splitPtr, objc-1, &resultPtr);
+	    if ((objc == 1) &&
+		    (Tcl_FSGetPathType(resultPtr) != TCL_PATH_RELATIVE)) {
+		resultPtr = NULL;
+	    }
+	}
+    }
+    if (resultPtr == NULL) {
+	resultPtr = Tcl_NewObj();
+    }
+    Tcl_IncrRefCount(resultPtr);
+    Tcl_DecrRefCount(splitPtr);
+    return resultPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFileAttrsCmd --
+ *
+ *	Sets or gets the platform-specific attributes of a file. The objc-objv
+ *	points to the file name with the rest of the command line following.
+ *	This routine uses platform-specific tables of option strings and
+ *	callbacks. The callback to get the attributes take three parameters:
+ *	    Tcl_Interp *interp;	    The interp to report errors with. Since
+ *				    this is an object-based API, the object
+ *				    form of the result should be used.
+ *	    const char *fileName;   This is extracted using
+ *				    Tcl_TranslateFileName.
+ *	    TclObj **attrObjPtrPtr; A new object to hold the attribute is
+ *				    allocated and put here.
+ *	The first two parameters of the callback used to write out the
+ *	attributes are the same. The third parameter is:
+ *	    const *attrObjPtr;	    A pointer to the object that has the new
+ *				    attribute.
+ *	They both return standard TCL errors; if the routine to get an
+ *	attribute fails, no object is allocated and *attrObjPtrPtr is
+ *	unchanged.
+ *
+ * Results:
+ *	Standard TCL error.
+ *
+ * Side effects:
+ *	May set file attributes for the file name.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFileAttrsCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int objc,			/* Number of command line arguments. */
+    Tcl_Obj *const objv[])	/* The command line objects. */
+{
+    int result;
+    const char *const *attributeStrings;
+    const char **attributeStringsAllocated = NULL;
+    Tcl_Obj *objStrings = NULL;
+    int numObjStrings = -1;
+    Tcl_Obj *filePtr;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?-option value ...?");
+	return TCL_ERROR;
+    }
+
+    filePtr = objv[1];
+    if (Tcl_FSConvertToPathType(interp, filePtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    objc -= 2;
+    objv += 2;
+    result = TCL_ERROR;
+    Tcl_SetErrno(0);
+
+    /*
+     * Get the set of attribute names from the filesystem.
+     */
+
+    attributeStrings = Tcl_FSFileAttrStrings(filePtr, &objStrings);
+    if (attributeStrings == NULL) {
+	int index;
+	Tcl_Obj *objPtr;
+
+	if (objStrings == NULL) {
+	    if (Tcl_GetErrno() != 0) {
+		/*
+		 * There was an error, probably that the filePtr is not
+		 * accepted by any filesystem
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"could not read \"%s\": %s",
+			TclGetString(filePtr), Tcl_PosixError(interp)));
+	    }
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * We own the object now.
+	 */
+
+	Tcl_IncrRefCount(objStrings);
+
+	/*
+	 * Use objStrings as a list object.
+	 */
+
+	if (Tcl_ListObjLength(interp, objStrings, &numObjStrings) != TCL_OK) {
+	    goto end;
+	}
+	attributeStringsAllocated = (const char **)
+		TclStackAlloc(interp, (1+numObjStrings) * sizeof(char *));
+	for (index = 0; index < numObjStrings; index++) {
+	    Tcl_ListObjIndex(interp, objStrings, index, &objPtr);
+	    attributeStringsAllocated[index] = TclGetString(objPtr);
+	}
+	attributeStringsAllocated[index] = NULL;
+	attributeStrings = attributeStringsAllocated;
+    } else if (objStrings != NULL) {
+	Tcl_Panic("must not update objPtrRef's variable and return non-NULL");
+    }
+
+    /*
+     * Process the attributes to produce a list of all of them, the value of a
+     * particular attribute, or to set one or more attributes (depending on
+     * the number of arguments).
+     */
+
+    if (objc == 0) {
+	/*
+	 * Get all attributes.
+	 */
+
+	int index, res = TCL_OK, nbAtts = 0;
+	Tcl_Obj *listPtr;
+
+	listPtr = Tcl_NewListObj(0, NULL);
+	for (index = 0; attributeStrings[index] != NULL; index++) {
+	    Tcl_Obj *objPtrAttr;
+
+	    if (res != TCL_OK) {
+		/*
+		 * Clear the error from the last iteration.
+		 */
+
+		Tcl_ResetResult(interp);
+	    }
+
+	    res = Tcl_FSFileAttrsGet(interp, index, filePtr, &objPtrAttr);
+	    if (res == TCL_OK) {
+		Tcl_Obj *objPtr =
+			Tcl_NewStringObj(attributeStrings[index], -1);
+
+		Tcl_ListObjAppendElement(interp, listPtr, objPtr);
+		Tcl_ListObjAppendElement(interp, listPtr, objPtrAttr);
+		nbAtts++;
+	    }
+	}
+
+	if (index > 0 && nbAtts == 0) {
+	    /*
+	     * Error: no valid attributes found.
+	     */
+
+	    Tcl_DecrRefCount(listPtr);
+	    goto end;
+	}
+
+	Tcl_SetObjResult(interp, listPtr);
+    } else if (objc == 1) {
+	/*
+	 * Get one attribute.
+	 */
+
+	int index;
+	Tcl_Obj *objPtr = NULL;
+
+	if (numObjStrings == 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\", there are no file attributes in this"
+		    " filesystem", TclGetString(objv[0])));
+	    Tcl_SetErrorCode(interp, "TCL","OPERATION","FATTR","NONE", NULL);
+	    goto end;
+	}
+
+	if (Tcl_GetIndexFromObj(interp, objv[0], attributeStrings,
+		"option", INDEX_TEMP_TABLE, &index) != TCL_OK) {
+	    goto end;
+	}
+	if (Tcl_FSFileAttrsGet(interp, index, filePtr,
+		&objPtr) != TCL_OK) {
+	    goto end;
+	}
+	Tcl_SetObjResult(interp, objPtr);
+    } else {
+	/*
+	 * Set option/value pairs.
+	 */
+
+	int i, index;
+
+	if (numObjStrings == 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad option \"%s\", there are no file attributes in this"
+		    " filesystem", TclGetString(objv[0])));
+	    Tcl_SetErrorCode(interp, "TCL","OPERATION","FATTR","NONE", NULL);
+	    goto end;
+	}
+
+	for (i = 0; i < objc ; i += 2) {
+	    if (Tcl_GetIndexFromObj(interp, objv[i], attributeStrings,
+		    "option", INDEX_TEMP_TABLE, &index) != TCL_OK) {
+		goto end;
+	    }
+	    if (i + 1 == objc) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"value for \"%s\" missing", TclGetString(objv[i])));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "FATTR",
+			"NOVALUE", NULL);
+		goto end;
+	    }
+	    if (Tcl_FSFileAttrsSet(interp, index, filePtr,
+		    objv[i + 1]) != TCL_OK) {
+		goto end;
+	    }
+	}
+    }
+    result = TCL_OK;
+
+    /*
+     * Free up the array we allocated and drop our reference to any list of
+     * attribute names issued by the filesystem.
+     */
+
+  end:
+    if (attributeStringsAllocated != NULL) {
+	TclStackFree(interp, (void *) attributeStringsAllocated);
+    }
+    if (objStrings != NULL) {
+	Tcl_DecrRefCount(objStrings);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFileLinkCmd --
+ *
+ *	This function is invoked to process the "file link" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May create a new link.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFileLinkCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *contents;
+    int index;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-linktype? linkname ?target?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Index of the 'source' argument.
+     */
+
+    if (objc == 4) {
+	index = 2;
+    } else {
+	index = 1;
+    }
+
+    if (objc > 2) {
+	int linkAction;
+
+	if (objc == 4) {
+	    /*
+	     * We have a '-linktype' argument.
+	     */
+
+	    static const char *const linkTypes[] = {
+		"-symbolic", "-hard", NULL
+	    };
+	    if (Tcl_GetIndexFromObj(interp, objv[1], linkTypes, "option", 0,
+		    &linkAction) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (linkAction == 0) {
+		linkAction = TCL_CREATE_SYMBOLIC_LINK;
+	    } else {
+		linkAction = TCL_CREATE_HARD_LINK;
+	    }
+	} else {
+	    linkAction = TCL_CREATE_SYMBOLIC_LINK | TCL_CREATE_HARD_LINK;
+	}
+	if (Tcl_FSConvertToPathType(interp, objv[index]) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Create link from source to target.
+	 */
+
+	contents = Tcl_FSLink(objv[index], objv[index+1], linkAction);
+	if (contents == NULL) {
+	    /*
+	     * We handle three common error cases specially, and for all other
+	     * errors, we use the standard posix error message.
+	     */
+
+	    if (errno == EEXIST) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"could not create new link \"%s\": that path already"
+			" exists", TclGetString(objv[index])));
+		Tcl_PosixError(interp);
+	    } else if (errno == ENOENT) {
+		/*
+		 * There are two cases here: either the target doesn't exist,
+		 * or the directory of the src doesn't exist.
+		 */
+
+		int access;
+		Tcl_Obj *dirPtr = TclPathPart(interp, objv[index],
+			TCL_PATH_DIRNAME);
+
+		if (dirPtr == NULL) {
+		    return TCL_ERROR;
+		}
+		access = Tcl_FSAccess(dirPtr, F_OK);
+		Tcl_DecrRefCount(dirPtr);
+		if (access != 0) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "could not create new link \"%s\": no such file"
+			    " or directory", TclGetString(objv[index])));
+		    Tcl_PosixError(interp);
+		} else {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "could not create new link \"%s\": target \"%s\" "
+			    "doesn't exist", TclGetString(objv[index]),
+			    TclGetString(objv[index+1])));
+		    errno = ENOENT;
+		    Tcl_PosixError(interp);
+		}
+	    } else {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"could not create new link \"%s\" pointing to \"%s\": %s",
+			TclGetString(objv[index]),
+			TclGetString(objv[index+1]), Tcl_PosixError(interp)));
+	    }
+	    return TCL_ERROR;
+	}
+    } else {
+	if (Tcl_FSConvertToPathType(interp, objv[index]) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Read link
+	 */
+
+	contents = Tcl_FSLink(objv[index], NULL, 0);
+	if (contents == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "could not read link \"%s\": %s",
+		    TclGetString(objv[index]), Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_SetObjResult(interp, contents);
+    if (objc == 2) {
+	/*
+	 * If we are reading a link, we need to free this result refCount. If
+	 * we are creating a link, this will just be objv[index+1], and so we
+	 * don't own it.
+	 */
+
+	Tcl_DecrRefCount(contents);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFileReadLinkCmd --
+ *
+ *	This function is invoked to process the "file readlink" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFileReadLinkCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *contents;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_FSConvertToPathType(interp, objv[1]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    contents = Tcl_FSLink(objv[1], NULL, 0);
+
+    if (contents == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"could not read link \"%s\": %s",
+		TclGetString(objv[1]), Tcl_PosixError(interp)));
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, contents);
+    Tcl_DecrRefCount(contents);
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFileTemporaryCmd
+ *
+ *	This function implements the "tempfile" subcommand of the "file"
+ *	command.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a temporary file. Opens a channel to that file and puts the
+ *	name of that channel in the result. *Might* register suitable exit
+ *	handlers to ensure that the temporary file gets deleted. Might write
+ *	to a variable, so reentrancy is a potential issue.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclFileTemporaryCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *nameVarObj = NULL;	/* Variable to store the name of the temporary
+				 * file in. */
+    Tcl_Obj *nameObj = NULL;	/* Object that will contain the filename. */
+    Tcl_Channel chan;		/* The channel opened (RDWR) on the temporary
+				 * file, or NULL if there's an error. */
+    Tcl_Obj *tempDirObj = NULL, *tempBaseObj = NULL, *tempExtObj = NULL;
+				/* Pieces of template. Each piece is NULL if
+				 * it is omitted. The platform temporary file
+				 * engine might ignore some pieces. */
+
+    if (objc < 1 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?nameVar? ?template?");
+	return TCL_ERROR;
+    }
+
+    if (objc > 1) {
+	nameVarObj = objv[1];
+	TclNewObj(nameObj);
+    }
+    if (objc > 2) {
+	int length;
+	Tcl_Obj *templateObj = objv[2];
+	const char *string = TclGetStringFromObj(templateObj, &length);
+
+	/*
+	 * Treat an empty string as if it wasn't there.
+	 */
+
+	if (length == 0) {
+	    goto makeTemporary;
+	}
+
+	/*
+	 * The template only gives a directory if there is a directory
+	 * separator in it.
+	 */
+
+	if (strchr(string, '/') != NULL
+		|| (tclPlatform == TCL_PLATFORM_WINDOWS
+		    && strchr(string, '\\') != NULL)) {
+	    tempDirObj = TclPathPart(interp, templateObj, TCL_PATH_DIRNAME);
+
+	    /*
+	     * Only allow creation of temporary files in the native filesystem
+	     * since they are frequently used for integration with external
+	     * tools or system libraries. [Bug 2388866]
+	     */
+
+	    if (tempDirObj != NULL && Tcl_FSGetFileSystemForPath(tempDirObj)
+		    != &tclNativeFilesystem) {
+		TclDecrRefCount(tempDirObj);
+		tempDirObj = NULL;
+	    }
+	}
+
+	/*
+	 * The template only gives the filename if the last character isn't a
+	 * directory separator.
+	 */
+
+	if (string[length-1] != '/' && (tclPlatform != TCL_PLATFORM_WINDOWS
+		|| string[length-1] != '\\')) {
+	    Tcl_Obj *tailObj = TclPathPart(interp, templateObj,
+		    TCL_PATH_TAIL);
+
+	    if (tailObj != NULL) {
+		tempBaseObj = TclPathPart(interp, tailObj, TCL_PATH_ROOT);
+		tempExtObj = TclPathPart(interp, tailObj, TCL_PATH_EXTENSION);
+		TclDecrRefCount(tailObj);
+	    }
+	}
+    }
+
+    /*
+     * Convert empty parts of the template into unspecified parts.
+     */
+
+    if (tempDirObj && !TclGetString(tempDirObj)[0]) {
+	TclDecrRefCount(tempDirObj);
+	tempDirObj = NULL;
+    }
+    if (tempBaseObj && !TclGetString(tempBaseObj)[0]) {
+	TclDecrRefCount(tempBaseObj);
+	tempBaseObj = NULL;
+    }
+    if (tempExtObj && !TclGetString(tempExtObj)[0]) {
+	TclDecrRefCount(tempExtObj);
+	tempExtObj = NULL;
+    }
+
+    /*
+     * Create and open the temporary file.
+     */
+
+  makeTemporary:
+    chan = TclpOpenTemporaryFile(tempDirObj,tempBaseObj,tempExtObj, nameObj);
+
+    /*
+     * If we created pieces of template, get rid of them now.
+     */
+
+    if (tempDirObj) {
+	TclDecrRefCount(tempDirObj);
+    }
+    if (tempBaseObj) {
+	TclDecrRefCount(tempBaseObj);
+    }
+    if (tempExtObj) {
+	TclDecrRefCount(tempExtObj);
+    }
+
+    /*
+     * Deal with results.
+     */
+
+    if (chan == NULL) {
+	if (nameVarObj) {
+	    TclDecrRefCount(nameObj);
+	}
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create temporary file: %s", Tcl_PosixError(interp)));
+	return TCL_ERROR;
+    }
+    Tcl_RegisterChannel(interp, chan);
+    if (nameVarObj != NULL) {
+	if (Tcl_ObjSetVar2(interp, nameVarObj, NULL, nameObj,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    Tcl_UnregisterChannel(interp, chan);
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(Tcl_GetChannelName(chan), -1));
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclFileName.c b/generic/tclFileName.c
new file mode 100644
index 0000000..150fb8c
--- /dev/null
+++ b/generic/tclFileName.c
@@ -0,0 +1,2658 @@
+/*
+ * tclFileName.c --
+ *
+ *	This file contains routines for converting file names betwen native
+ *	and network form.
+ *
+ * Copyright (c) 1995-1998 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclRegexp.h"
+#include "tclFileSystem.h" /* For TclGetPathType() */
+
+/*
+ * The following variable is set in the TclPlatformInit call to one of:
+ * TCL_PLATFORM_UNIX or TCL_PLATFORM_WINDOWS.
+ */
+
+TclPlatformType tclPlatform = TCL_PLATFORM_UNIX;
+
+/*
+ * Prototypes for local procedures defined in this file:
+ */
+
+static const char *	DoTildeSubst(Tcl_Interp *interp,
+			    const char *user, Tcl_DString *resultPtr);
+static const char *	ExtractWinRoot(const char *path,
+			    Tcl_DString *resultPtr, int offset,
+			    Tcl_PathType *typePtr);
+static int		SkipToChar(char **stringPtr, int match);
+static Tcl_Obj *	SplitWinPath(const char *path);
+static Tcl_Obj *	SplitUnixPath(const char *path);
+static int		DoGlob(Tcl_Interp *interp, Tcl_Obj *resultPtr,
+			    const char *separators, Tcl_Obj *pathPtr, int flags,
+			    char *pattern, Tcl_GlobTypeData *types);
+
+/*
+ * When there is no support for getting the block size of a file in a stat()
+ * call, use this as a guess. Allow it to be overridden in the platform-
+ * specific files.
+ */
+
+#if (!defined(HAVE_STRUCT_STAT_ST_BLKSIZE) && !defined(GUESSED_BLOCK_SIZE))
+#define GUESSED_BLOCK_SIZE	1024
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetResultLength --
+ *
+ *	Resets the result DString for ExtractWinRoot to accommodate
+ *	any NT extended path prefixes.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May modify the Tcl_DString.
+ *----------------------------------------------------------------------
+ */
+
+static void
+SetResultLength(
+    Tcl_DString *resultPtr,
+    int offset,
+    int extended)
+{
+    Tcl_DStringSetLength(resultPtr, offset);
+    if (extended == 2) {
+	TclDStringAppendLiteral(resultPtr, "//?/UNC/");
+    } else if (extended == 1) {
+	TclDStringAppendLiteral(resultPtr, "//?/");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExtractWinRoot --
+ *
+ *	Matches the root portion of a Windows path and appends it to the
+ *	specified Tcl_DString.
+ *
+ * Results:
+ *	Returns the position in the path immediately after the root including
+ *	any trailing slashes. Appends a cleaned up version of the root to the
+ *	Tcl_DString at the specified offest.
+ *
+ * Side effects:
+ *	Modifies the specified Tcl_DString.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *
+ExtractWinRoot(
+    const char *path,		/* Path to parse. */
+    Tcl_DString *resultPtr,	/* Buffer to hold result. */
+    int offset,			/* Offset in buffer where result should be
+				 * stored. */
+    Tcl_PathType *typePtr)	/* Where to store pathType result */
+{
+    int extended = 0;
+
+    if (   (path[0] == '/' || path[0] == '\\')
+	&& (path[1] == '/' || path[1] == '\\')
+	&& (path[2] == '?')
+	&& (path[3] == '/' || path[3] == '\\')) {
+	extended = 1;
+	path = path + 4;
+	if (path[0] == 'U' && path[1] == 'N' && path[2] == 'C'
+	    && (path[3] == '/' || path[3] == '\\')) {
+	    extended = 2;
+	    path = path + 4;
+	}
+    }
+
+    if (path[0] == '/' || path[0] == '\\') {
+	/*
+	 * Might be a UNC or Vol-Relative path.
+	 */
+
+	const char *host, *share, *tail;
+	int hlen, slen;
+
+	if (path[1] != '/' && path[1] != '\\') {
+	    SetResultLength(resultPtr, offset, extended);
+	    *typePtr = TCL_PATH_VOLUME_RELATIVE;
+	    TclDStringAppendLiteral(resultPtr, "/");
+	    return &path[1];
+	}
+	host = &path[2];
+
+	/*
+	 * Skip separators.
+	 */
+
+	while (host[0] == '/' || host[0] == '\\') {
+	    host++;
+	}
+
+	for (hlen = 0; host[hlen];hlen++) {
+	    if (host[hlen] == '/' || host[hlen] == '\\') {
+		break;
+	    }
+	}
+	if (host[hlen] == 0 || host[hlen+1] == 0) {
+	    /*
+	     * The path given is simply of the form '/foo', '//foo',
+	     * '/////foo' or the same with backslashes. If there is exactly
+	     * one leading '/' the path is volume relative (see filename man
+	     * page). If there are more than one, we are simply assuming they
+	     * are superfluous and we trim them away. (An alternative
+	     * interpretation would be that it is a host name, but we have
+	     * been documented that that is not the case).
+	     */
+
+	    *typePtr = TCL_PATH_VOLUME_RELATIVE;
+	    TclDStringAppendLiteral(resultPtr, "/");
+	    return &path[2];
+	}
+	SetResultLength(resultPtr, offset, extended);
+	share = &host[hlen];
+
+	/*
+	 * Skip separators.
+	 */
+
+	while (share[0] == '/' || share[0] == '\\') {
+	    share++;
+	}
+
+	for (slen=0; share[slen]; slen++) {
+	    if (share[slen] == '/' || share[slen] == '\\') {
+		break;
+	    }
+	}
+	TclDStringAppendLiteral(resultPtr, "//");
+	Tcl_DStringAppend(resultPtr, host, hlen);
+	TclDStringAppendLiteral(resultPtr, "/");
+	Tcl_DStringAppend(resultPtr, share, slen);
+
+	tail = &share[slen];
+
+	/*
+	 * Skip separators.
+	 */
+
+	while (tail[0] == '/' || tail[0] == '\\') {
+	    tail++;
+	}
+
+	*typePtr = TCL_PATH_ABSOLUTE;
+	return tail;
+    } else if (*path && path[1] == ':') {
+	/*
+	 * Might be a drive separator.
+	 */
+
+	SetResultLength(resultPtr, offset, extended);
+
+	if (path[2] != '/' && path[2] != '\\') {
+	    *typePtr = TCL_PATH_VOLUME_RELATIVE;
+	    Tcl_DStringAppend(resultPtr, path, 2);
+	    return &path[2];
+	} else {
+	    const char *tail = &path[3];
+
+	    /*
+	     * Skip separators.
+	     */
+
+	    while (*tail && (tail[0] == '/' || tail[0] == '\\')) {
+		tail++;
+	    }
+
+	    *typePtr = TCL_PATH_ABSOLUTE;
+	    Tcl_DStringAppend(resultPtr, path, 2);
+	    TclDStringAppendLiteral(resultPtr, "/");
+
+	    return tail;
+	}
+    } else {
+	int abs = 0;
+
+	/*
+	 * Check for Windows devices.
+	 */
+
+	if ((path[0] == 'c' || path[0] == 'C')
+		&& (path[1] == 'o' || path[1] == 'O')) {
+	    if ((path[2] == 'm' || path[2] == 'M')
+		    && path[3] >= '1' && path[3] <= '9') {
+		/*
+		 * May have match for 'com[1-9]:?', which is a serial port.
+		 */
+
+		if (path[4] == '\0') {
+		    abs = 4;
+		} else if (path [4] == ':' && path[5] == '\0') {
+		    abs = 5;
+		}
+
+	    } else if ((path[2] == 'n' || path[2] == 'N') && path[3] == '\0') {
+		/*
+		 * Have match for 'con'.
+		 */
+
+		abs = 3;
+	    }
+
+	} else if ((path[0] == 'l' || path[0] == 'L')
+		&& (path[1] == 'p' || path[1] == 'P')
+		&& (path[2] == 't' || path[2] == 'T')) {
+	    if (path[3] >= '1' && path[3] <= '9') {
+		/*
+		 * May have match for 'lpt[1-9]:?'
+		 */
+
+		if (path[4] == '\0') {
+		    abs = 4;
+		} else if (path [4] == ':' && path[5] == '\0') {
+		    abs = 5;
+		}
+	    }
+
+	} else if ((path[0] == 'p' || path[0] == 'P')
+		&& (path[1] == 'r' || path[1] == 'R')
+		&& (path[2] == 'n' || path[2] == 'N')
+		&& path[3] == '\0') {
+	    /*
+	     * Have match for 'prn'.
+	     */
+	    abs = 3;
+
+	} else if ((path[0] == 'n' || path[0] == 'N')
+		&& (path[1] == 'u' || path[1] == 'U')
+		&& (path[2] == 'l' || path[2] == 'L')
+		&& path[3] == '\0') {
+	    /*
+	     * Have match for 'nul'.
+	     */
+
+	    abs = 3;
+
+	} else if ((path[0] == 'a' || path[0] == 'A')
+		&& (path[1] == 'u' || path[1] == 'U')
+		&& (path[2] == 'x' || path[2] == 'X')
+		&& path[3] == '\0') {
+	    /*
+	     * Have match for 'aux'.
+	     */
+
+	    abs = 3;
+	}
+
+	if (abs != 0) {
+	    *typePtr = TCL_PATH_ABSOLUTE;
+	    SetResultLength(resultPtr, offset, extended);
+	    Tcl_DStringAppend(resultPtr, path, abs);
+	    return path + abs;
+	}
+    }
+
+    /*
+     * Anything else is treated as relative.
+     */
+
+    *typePtr = TCL_PATH_RELATIVE;
+    return path;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetPathType --
+ *
+ *	Determines whether a given path is relative to the current directory,
+ *	relative to the current volume, or absolute.
+ *
+ *	The objectified Tcl_FSGetPathType should be used in preference to this
+ *	function (as you can see below, this is just a wrapper around that
+ *	other function).
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE, TCL_PATH_RELATIVE, or
+ *	TCL_PATH_VOLUME_RELATIVE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+Tcl_GetPathType(
+    const char *path)
+{
+    Tcl_PathType type;
+    Tcl_Obj *tempObj = Tcl_NewStringObj(path,-1);
+
+    Tcl_IncrRefCount(tempObj);
+    type = Tcl_FSGetPathType(tempObj);
+    Tcl_DecrRefCount(tempObj);
+    return type;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpGetNativePathType --
+ *
+ *	Determines whether a given path is relative to the current directory,
+ *	relative to the current volume, or absolute, but ONLY FOR THE NATIVE
+ *	FILESYSTEM. This function is called from tclIOUtil.c (but needs to be
+ *	here due to its dependence on static variables/functions in this
+ *	file). The exported function Tcl_FSGetPathType should be used by
+ *	extensions.
+ *
+ *	Note that '~' paths are always considered TCL_PATH_ABSOLUTE, even
+ *	though expanding the '~' could lead to any possible path type. This
+ *	function should therefore be considered a low-level, string
+ *	manipulation function only -- it doesn't actually do any expansion in
+ *	making its determination.
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE, TCL_PATH_RELATIVE, or
+ *	TCL_PATH_VOLUME_RELATIVE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+TclpGetNativePathType(
+    Tcl_Obj *pathPtr,		/* Native path of interest */
+    int *driveNameLengthPtr,	/* Returns length of drive, if non-NULL and
+				 * path was absolute */
+    Tcl_Obj **driveNameRef)
+{
+    Tcl_PathType type = TCL_PATH_ABSOLUTE;
+    int pathLen;
+    const char *path = TclGetStringFromObj(pathPtr, &pathLen);
+
+    if (path[0] == '~') {
+	/*
+	 * This case is common to all platforms. Paths that begin with ~ are
+	 * absolute.
+	 */
+
+	if (driveNameLengthPtr != NULL) {
+	    const char *end = path + 1;
+	    while ((*end != '\0') && (*end != '/')) {
+		end++;
+	    }
+	    *driveNameLengthPtr = end - path;
+	}
+    } else {
+	switch (tclPlatform) {
+	case TCL_PLATFORM_UNIX: {
+	    const char *origPath = path;
+
+	    /*
+	     * Paths that begin with / are absolute.
+	     */
+
+	    if (path[0] == '/') {
+		++path;
+#if defined(__CYGWIN__) || defined(__QNX__)
+		/*
+		 * Check for "//" network path prefix
+		 */
+		if ((*path == '/') && path[1] && (path[1] != '/')) {
+		    path += 2;
+		    while (*path && *path != '/') {
+			++path;
+		    }
+#if defined(__CYGWIN__)
+		    /* UNC paths need to be followed by a share name */
+		    if (*path++ && (*path && *path != '/')) {
+			++path;
+			while (*path && *path != '/') {
+			    ++path;
+			}
+		    } else {
+			path = origPath + 1;
+		    }
+#endif
+		}
+#endif
+		if (driveNameLengthPtr != NULL) {
+		    /*
+		     * We need this addition in case the QNX or Cygwin code was used.
+		     */
+
+		    *driveNameLengthPtr = (path - origPath);
+		}
+	    } else {
+		type = TCL_PATH_RELATIVE;
+	    }
+	    break;
+	}
+	case TCL_PLATFORM_WINDOWS: {
+	    Tcl_DString ds;
+	    const char *rootEnd;
+
+	    Tcl_DStringInit(&ds);
+	    rootEnd = ExtractWinRoot(path, &ds, 0, &type);
+	    if ((rootEnd != path) && (driveNameLengthPtr != NULL)) {
+		*driveNameLengthPtr = rootEnd - path;
+		if (driveNameRef != NULL) {
+		    *driveNameRef = TclDStringToObj(&ds);
+		    Tcl_IncrRefCount(*driveNameRef);
+		}
+	    }
+	    Tcl_DStringFree(&ds);
+	    break;
+	}
+	}
+    }
+    return type;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclpNativeSplitPath --
+ *
+ *	This function takes the given Tcl_Obj, which should be a valid path,
+ *	and returns a Tcl List object containing each segment of that path as
+ *	an element.
+ *
+ *	Note this function currently calls the older Split(Plat)Path
+ *	functions, which require more memory allocation than is desirable.
+ *
+ * Results:
+ *	Returns list object with refCount of zero. If the passed in lenPtr is
+ *	non-NULL, we use it to return the number of elements in the returned
+ *	list.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclpNativeSplitPath(
+    Tcl_Obj *pathPtr,		/* Path to split. */
+    int *lenPtr)		/* int to store number of path elements. */
+{
+    Tcl_Obj *resultPtr = NULL;	/* Needed only to prevent gcc warnings. */
+
+    /*
+     * Perform platform specific splitting.
+     */
+
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	resultPtr = SplitUnixPath(Tcl_GetString(pathPtr));
+	break;
+
+    case TCL_PLATFORM_WINDOWS:
+	resultPtr = SplitWinPath(Tcl_GetString(pathPtr));
+	break;
+    }
+
+    /*
+     * Compute the number of elements in the result.
+     */
+
+    if (lenPtr != NULL) {
+	Tcl_ListObjLength(NULL, resultPtr, lenPtr);
+    }
+    return resultPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SplitPath --
+ *
+ *	Split a path into a list of path components. The first element of the
+ *	list will have the same path type as the original path.
+ *
+ * Results:
+ *	Returns a standard Tcl result. The interpreter result contains a list
+ *	of path components. *argvPtr will be filled in with the address of an
+ *	array whose elements point to the elements of path, in order.
+ *	*argcPtr will get filled in with the number of valid elements in the
+ *	array. A single block of memory is dynamically allocated to hold both
+ *	the argv array and a copy of the path elements. The caller must
+ *	eventually free this memory by calling ckfree() on *argvPtr. Note:
+ *	*argvPtr and *argcPtr are only modified if the procedure returns
+ *	normally.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SplitPath(
+    const char *path,		/* Pointer to string containing a path. */
+    int *argcPtr,		/* Pointer to location to fill in with the
+				 * number of elements in the path. */
+    const char ***argvPtr)	/* Pointer to place to store pointer to array
+				 * of pointers to path elements. */
+{
+    Tcl_Obj *resultPtr = NULL;	/* Needed only to prevent gcc warnings. */
+    Tcl_Obj *tmpPtr, *eltPtr;
+    int i, size, len;
+    char *p;
+    const char *str;
+
+    /*
+     * Perform the splitting, using objectified, vfs-aware code.
+     */
+
+    tmpPtr = Tcl_NewStringObj(path, -1);
+    Tcl_IncrRefCount(tmpPtr);
+    resultPtr = Tcl_FSSplitPath(tmpPtr, argcPtr);
+    Tcl_IncrRefCount(resultPtr);
+    Tcl_DecrRefCount(tmpPtr);
+
+    /*
+     * Calculate space required for the result.
+     */
+
+    size = 1;
+    for (i = 0; i < *argcPtr; i++) {
+	Tcl_ListObjIndex(NULL, resultPtr, i, &eltPtr);
+	TclGetStringFromObj(eltPtr, &len);
+	size += len + 1;
+    }
+
+    /*
+     * Allocate a buffer large enough to hold the contents of all of the list
+     * plus the argv pointers and the terminating NULL pointer.
+     */
+
+    *argvPtr = ckalloc((((*argcPtr) + 1) * sizeof(char *)) + size);
+
+    /*
+     * Position p after the last argv pointer and copy the contents of the
+     * list in, piece by piece.
+     */
+
+    p = (char *) &(*argvPtr)[(*argcPtr) + 1];
+    for (i = 0; i < *argcPtr; i++) {
+	Tcl_ListObjIndex(NULL, resultPtr, i, &eltPtr);
+	str = TclGetStringFromObj(eltPtr, &len);
+	memcpy(p, str, (size_t) len+1);
+	p += len+1;
+    }
+
+    /*
+     * Now set up the argv pointers.
+     */
+
+    p = (char *) &(*argvPtr)[(*argcPtr) + 1];
+
+    for (i = 0; i < *argcPtr; i++) {
+	(*argvPtr)[i] = p;
+	for (; *(p++)!='\0'; );
+    }
+    (*argvPtr)[i] = NULL;
+
+    /*
+     * Free the result ptr given to us by Tcl_FSSplitPath
+     */
+
+    Tcl_DecrRefCount(resultPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SplitUnixPath --
+ *
+ *	This routine is used by Tcl_(FS)SplitPath to handle splitting Unix
+ *	paths.
+ *
+ * Results:
+ *	Returns a newly allocated Tcl list object.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+SplitUnixPath(
+    const char *path)		/* Pointer to string containing a path. */
+{
+    int length;
+    const char *origPath = path, *elementStart;
+    Tcl_Obj *result = Tcl_NewObj();
+
+    /*
+     * Deal with the root directory as a special case.
+     */
+
+    if (*path == '/') {
+	Tcl_Obj *rootElt;
+	++path;
+#if defined(__CYGWIN__) || defined(__QNX__)
+	/*
+	 * Check for "//" network path prefix
+	 */
+	if ((*path == '/') && path[1] && (path[1] != '/')) {
+	    path += 2;
+	    while (*path && *path != '/') {
+		++path;
+	    }
+#if defined(__CYGWIN__)
+	    /* UNC paths need to be followed by a share name */
+	    if (*path++ && (*path && *path != '/')) {
+		++path;
+		while (*path && *path != '/') {
+		    ++path;
+		}
+	    } else {
+		path = origPath + 1;
+	    }
+#endif
+	}
+#endif
+	rootElt = Tcl_NewStringObj(origPath, path - origPath);
+	Tcl_ListObjAppendElement(NULL, result, rootElt);
+	while (*path == '/') {
+	    ++path;
+	}
+    }
+
+    /*
+     * Split on slashes. Embedded elements that start with tilde will be
+     * prefixed with "./" so they are not affected by tilde substitution.
+     */
+
+    for (;;) {
+	elementStart = path;
+	while ((*path != '\0') && (*path != '/')) {
+	    path++;
+	}
+	length = path - elementStart;
+	if (length > 0) {
+	    Tcl_Obj *nextElt;
+	    if ((elementStart[0] == '~') && (elementStart != origPath)) {
+		TclNewLiteralStringObj(nextElt, "./");
+		Tcl_AppendToObj(nextElt, elementStart, length);
+	    } else {
+		nextElt = Tcl_NewStringObj(elementStart, length);
+	    }
+	    Tcl_ListObjAppendElement(NULL, result, nextElt);
+	}
+	if (*path++ == '\0') {
+	    break;
+	}
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SplitWinPath --
+ *
+ *	This routine is used by Tcl_(FS)SplitPath to handle splitting Windows
+ *	paths.
+ *
+ * Results:
+ *	Returns a newly allocated Tcl list object.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+SplitWinPath(
+    const char *path)		/* Pointer to string containing a path. */
+{
+    int length;
+    const char *p, *elementStart;
+    Tcl_PathType type = TCL_PATH_ABSOLUTE;
+    Tcl_DString buf;
+    Tcl_Obj *result = Tcl_NewObj();
+    Tcl_DStringInit(&buf);
+
+    p = ExtractWinRoot(path, &buf, 0, &type);
+
+    /*
+     * Terminate the root portion, if we matched something.
+     */
+
+    if (p != path) {
+	Tcl_ListObjAppendElement(NULL, result, TclDStringToObj(&buf));
+    }
+    Tcl_DStringFree(&buf);
+
+    /*
+     * Split on slashes. Embedded elements that start with tilde or a drive
+     * letter will be prefixed with "./" so they are not affected by tilde
+     * substitution.
+     */
+
+    do {
+	elementStart = p;
+	while ((*p != '\0') && (*p != '/') && (*p != '\\')) {
+	    p++;
+	}
+	length = p - elementStart;
+	if (length > 0) {
+	    Tcl_Obj *nextElt;
+	    if ((elementStart != path) && ((elementStart[0] == '~')
+		    || (isalpha(UCHAR(elementStart[0]))
+			&& elementStart[1] == ':'))) {
+		TclNewLiteralStringObj(nextElt, "./");
+		Tcl_AppendToObj(nextElt, elementStart, length);
+	    } else {
+		nextElt = Tcl_NewStringObj(elementStart, length);
+	    }
+	    Tcl_ListObjAppendElement(NULL, result, nextElt);
+	}
+    } while (*p++ != '\0');
+
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSJoinToPath --
+ *
+ *	This function takes the given object, which should usually be a valid
+ *	path or NULL, and joins onto it the array of paths segments given.
+ *
+ *	The objects in the array given will temporarily have their refCount
+ *	increased by one, and then decreased by one when this function exits
+ *	(which means if they had zero refCount when we were called, they will
+ *	be freed).
+ *
+ * Results:
+ *	Returns object owned by the caller (which should increment its
+ *	refCount) - typically an object with refCount of zero.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSJoinToPath(
+    Tcl_Obj *pathPtr,		/* Valid path or NULL. */
+    int objc,			/* Number of array elements to join */
+    Tcl_Obj *const objv[])	/* Path elements to join. */
+{
+    if (pathPtr == NULL) {
+	return TclJoinPath(objc, objv);
+    }
+    if (objc == 0) {
+	return TclJoinPath(1, &pathPtr);
+    }
+    if (objc == 1) {
+	Tcl_Obj *pair[2];
+
+	pair[0] = pathPtr;
+	pair[1] = objv[0];
+	return TclJoinPath(2, pair);
+    } else {
+	int elemc = objc + 1;
+	Tcl_Obj *ret, **elemv = ckalloc(elemc*sizeof(Tcl_Obj *));
+
+	elemv[0] = pathPtr;
+	memcpy(elemv+1, objv, objc*sizeof(Tcl_Obj *));
+	ret = TclJoinPath(elemc, elemv);
+	ckfree(elemv);
+	return ret;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclpNativeJoinPath --
+ *
+ *	'prefix' is absolute, 'joining' is relative to prefix.
+ *
+ * Results:
+ *	modifies prefix
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclpNativeJoinPath(
+    Tcl_Obj *prefix,
+    const char *joining)
+{
+    int length, needsSep;
+    char *dest;
+    const char *p;
+    const char *start;
+
+    start = TclGetStringFromObj(prefix, &length);
+
+    /*
+     * Remove the ./ from tilde prefixed elements, and drive-letter prefixed
+     * elements on Windows, unless it is the first component.
+     */
+
+    p = joining;
+
+    if (length != 0) {
+	if ((p[0] == '.') && (p[1] == '/') && ((p[2] == '~')
+		|| (tclPlatform==TCL_PLATFORM_WINDOWS && isalpha(UCHAR(p[2]))
+		&& (p[3] == ':')))) {
+	    p += 2;
+	}
+    }
+    if (*p == '\0') {
+	return;
+    }
+
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	/*
+	 * Append a separator if needed.
+	 */
+
+	if (length > 0 && (start[length-1] != '/')) {
+	    Tcl_AppendToObj(prefix, "/", 1);
+	    TclGetStringFromObj(prefix, &length);
+	}
+	needsSep = 0;
+
+	/*
+	 * Append the element, eliminating duplicate and trailing slashes.
+	 */
+
+	Tcl_SetObjLength(prefix, length + (int) strlen(p));
+
+	dest = Tcl_GetString(prefix) + length;
+	for (; *p != '\0'; p++) {
+	    if (*p == '/') {
+		while (p[1] == '/') {
+		    p++;
+		}
+		if (p[1] != '\0' && needsSep) {
+		    *dest++ = '/';
+		}
+	    } else {
+		*dest++ = *p;
+		needsSep = 1;
+	    }
+	}
+	length = dest - Tcl_GetString(prefix);
+	Tcl_SetObjLength(prefix, length);
+	break;
+
+    case TCL_PLATFORM_WINDOWS:
+	/*
+	 * Check to see if we need to append a separator.
+	 */
+
+	if ((length > 0) &&
+		(start[length-1] != '/') && (start[length-1] != ':')) {
+	    Tcl_AppendToObj(prefix, "/", 1);
+	    TclGetStringFromObj(prefix, &length);
+	}
+	needsSep = 0;
+
+	/*
+	 * Append the element, eliminating duplicate and trailing slashes.
+	 */
+
+	Tcl_SetObjLength(prefix, length + (int) strlen(p));
+	dest = Tcl_GetString(prefix) + length;
+	for (; *p != '\0'; p++) {
+	    if ((*p == '/') || (*p == '\\')) {
+		while ((p[1] == '/') || (p[1] == '\\')) {
+		    p++;
+		}
+		if ((p[1] != '\0') && needsSep) {
+		    *dest++ = '/';
+		}
+	    } else {
+		*dest++ = *p;
+		needsSep = 1;
+	    }
+	}
+	length = dest - Tcl_GetString(prefix);
+	Tcl_SetObjLength(prefix, length);
+	break;
+    }
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_JoinPath --
+ *
+ *	Combine a list of paths in a platform specific manner. The function
+ *	'Tcl_FSJoinPath' should be used in preference where possible.
+ *
+ * Results:
+ *	Appends the joined path to the end of the specified Tcl_DString
+ *	returning a pointer to the resulting string. Note that the
+ *	Tcl_DString must already be initialized.
+ *
+ * Side effects:
+ *	Modifies the Tcl_DString.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_JoinPath(
+    int argc,
+    const char *const *argv,
+    Tcl_DString *resultPtr)	/* Pointer to previously initialized DString */
+{
+    int i, len;
+    Tcl_Obj *listObj = Tcl_NewObj();
+    Tcl_Obj *resultObj;
+    const char *resultStr;
+
+    /*
+     * Build the list of paths.
+     */
+
+    for (i = 0; i < argc; i++) {
+	Tcl_ListObjAppendElement(NULL, listObj,
+		Tcl_NewStringObj(argv[i], -1));
+    }
+
+    /*
+     * Ask the objectified code to join the paths.
+     */
+
+    Tcl_IncrRefCount(listObj);
+    resultObj = Tcl_FSJoinPath(listObj, argc);
+    Tcl_IncrRefCount(resultObj);
+    Tcl_DecrRefCount(listObj);
+
+    /*
+     * Store the result.
+     */
+
+    resultStr = TclGetStringFromObj(resultObj, &len);
+    Tcl_DStringAppend(resultPtr, resultStr, len);
+    Tcl_DecrRefCount(resultObj);
+
+    /*
+     * Return a pointer to the result.
+     */
+
+    return Tcl_DStringValue(resultPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_TranslateFileName --
+ *
+ *	Converts a file name into a form usable by the native system
+ *	interfaces. If the name starts with a tilde, it will produce a name
+ *	where the tilde and following characters have been replaced by the
+ *	home directory location for the named user.
+ *
+ * Results:
+ *	The return value is a pointer to a string containing the name after
+ *	tilde substitution. If there was no tilde substitution, the return
+ *	value is a pointer to a copy of the original string. If there was an
+ *	error in processing the name, then an error message is left in the
+ *	interp's result (if interp was not NULL) and the return value is NULL.
+ *	Space for the return value is allocated in bufferPtr; the caller must
+ *	call Tcl_DStringFree() to free the space if the return value was not
+ *	NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_TranslateFileName(
+    Tcl_Interp *interp,		/* Interpreter in which to store error message
+				 * (if necessary). */
+    const char *name,		/* File name, which may begin with "~" (to
+				 * indicate current user's home directory) or
+				 * "~<user>" (to indicate any user's home
+				 * directory). */
+    Tcl_DString *bufferPtr)	/* Uninitialized or free DString filled with
+				 * name after tilde substitution. */
+{
+    Tcl_Obj *path = Tcl_NewStringObj(name, -1);
+    Tcl_Obj *transPtr;
+
+    Tcl_IncrRefCount(path);
+    transPtr = Tcl_FSGetTranslatedPath(interp, path);
+    if (transPtr == NULL) {
+	Tcl_DecrRefCount(path);
+	return NULL;
+    }
+
+    Tcl_DStringInit(bufferPtr);
+    TclDStringAppendObj(bufferPtr, transPtr);
+    Tcl_DecrRefCount(path);
+    Tcl_DecrRefCount(transPtr);
+
+    /*
+     * Convert forward slashes to backslashes in Windows paths because some
+     * system interfaces don't accept forward slashes.
+     */
+
+    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+	register char *p;
+	for (p = Tcl_DStringValue(bufferPtr); *p != '\0'; p++) {
+	    if (*p == '/') {
+		*p = '\\';
+	    }
+	}
+    }
+
+    return Tcl_DStringValue(bufferPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetExtension --
+ *
+ *	This function returns a pointer to the beginning of the extension part
+ *	of a file name.
+ *
+ * Results:
+ *	Returns a pointer into name which indicates where the extension
+ *	starts. If there is no extension, returns NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+TclGetExtension(
+    const char *name)		/* File name to parse. */
+{
+    const char *p, *lastSep;
+
+    /*
+     * First find the last directory separator.
+     */
+
+    lastSep = NULL;		/* Needed only to prevent gcc warnings. */
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	lastSep = strrchr(name, '/');
+	break;
+
+    case TCL_PLATFORM_WINDOWS:
+	lastSep = NULL;
+	for (p = name; *p != '\0'; p++) {
+	    if (strchr("/\\:", *p) != NULL) {
+		lastSep = p;
+	    }
+	}
+	break;
+    }
+    p = strrchr(name, '.');
+    if ((p != NULL) && (lastSep != NULL) && (lastSep > p)) {
+	p = NULL;
+    }
+
+    /*
+     * In earlier versions, we used to back up to the first period in a series
+     * so that "foo..o" would be split into "foo" and "..o". This is a
+     * confusing and usually incorrect behavior, so now we split at the last
+     * period in the name.
+     */
+
+    return p;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DoTildeSubst --
+ *
+ *	Given a string following a tilde, this routine returns the
+ *	corresponding home directory.
+ *
+ * Results:
+ *	The result is a pointer to a static string containing the home
+ *	directory in native format. If there was an error in processing the
+ *	substitution, then an error message is left in the interp's result and
+ *	the return value is NULL. On success, the results are appended to
+ *	resultPtr, and the contents of resultPtr are returned.
+ *
+ * Side effects:
+ *	Information may be left in resultPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *
+DoTildeSubst(
+    Tcl_Interp *interp,		/* Interpreter in which to store error message
+				 * (if necessary). */
+    const char *user,		/* Name of user whose home directory should be
+				 * substituted, or "" for current user. */
+    Tcl_DString *resultPtr)	/* Initialized DString filled with name after
+				 * tilde substitution. */
+{
+    const char *dir;
+
+    if (*user == '\0') {
+	Tcl_DString dirString;
+
+	dir = TclGetEnv("HOME", &dirString);
+	if (dir == NULL) {
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"couldn't find HOME environment "
+			"variable to expand path", -1));
+		Tcl_SetErrorCode(interp, "TCL", "FILENAME", "NO_HOME", NULL);
+	    }
+	    return NULL;
+	}
+	Tcl_JoinPath(1, &dir, resultPtr);
+	Tcl_DStringFree(&dirString);
+    } else if (TclpGetUserHome(user, resultPtr) == NULL) {
+	if (interp) {
+	    Tcl_ResetResult(interp);
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "user \"%s\" doesn't exist", user));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "USER", user, NULL);
+	}
+	return NULL;
+    }
+    return Tcl_DStringValue(resultPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GlobObjCmd --
+ *
+ *	This procedure is invoked to process the "glob" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_GlobObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int index, i, globFlags, length, join, dir, result;
+    char *string;
+    const char *separators;
+    Tcl_Obj *typePtr, *look;
+    Tcl_Obj *pathOrDir = NULL;
+    Tcl_DString prefix;
+    static const char *const options[] = {
+	"-directory", "-join", "-nocomplain", "-path", "-tails",
+	"-types", "--", NULL
+    };
+    enum options {
+	GLOB_DIR, GLOB_JOIN, GLOB_NOCOMPLAIN, GLOB_PATH, GLOB_TAILS,
+	GLOB_TYPE, GLOB_LAST
+    };
+    enum pathDirOptions {PATH_NONE = -1 , PATH_GENERAL = 0, PATH_DIR = 1};
+    Tcl_GlobTypeData *globTypes = NULL;
+
+    globFlags = 0;
+    join = 0;
+    dir = PATH_NONE;
+    typePtr = NULL;
+    for (i = 1; i < objc; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		&index) != TCL_OK) {
+	    string = TclGetStringFromObj(objv[i], &length);
+	    if (string[0] == '-') {
+		/*
+		 * It looks like the command contains an option so signal an
+		 * error.
+		 */
+
+		return TCL_ERROR;
+	    } else {
+		/*
+		 * This clearly isn't an option; assume it's the first glob
+		 * pattern. We must clear the error.
+		 */
+
+		Tcl_ResetResult(interp);
+		break;
+	    }
+	}
+
+	switch (index) {
+	case GLOB_NOCOMPLAIN:			/* -nocomplain */
+	    globFlags |= TCL_GLOBMODE_NO_COMPLAIN;
+	    break;
+	case GLOB_DIR:				/* -dir */
+	    if (i == (objc-1)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing argument to \"-directory\"", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		return TCL_ERROR;
+	    }
+	    if (dir != PATH_NONE) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-directory\" cannot be used with \"-path\"", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB",
+			"BADOPTIONCOMBINATION", NULL);
+		return TCL_ERROR;
+	    }
+	    dir = PATH_DIR;
+	    globFlags |= TCL_GLOBMODE_DIR;
+	    pathOrDir = objv[i+1];
+	    i++;
+	    break;
+	case GLOB_JOIN:				/* -join */
+	    join = 1;
+	    break;
+	case GLOB_TAILS:				/* -tails */
+	    globFlags |= TCL_GLOBMODE_TAILS;
+	    break;
+	case GLOB_PATH:				/* -path */
+	    if (i == (objc-1)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing argument to \"-path\"", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		return TCL_ERROR;
+	    }
+	    if (dir != PATH_NONE) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-path\" cannot be used with \"-directory\"", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB",
+			"BADOPTIONCOMBINATION", NULL);
+		return TCL_ERROR;
+	    }
+	    dir = PATH_GENERAL;
+	    pathOrDir = objv[i+1];
+	    i++;
+	    break;
+	case GLOB_TYPE:				/* -types */
+	    if (i == (objc-1)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing argument to \"-types\"", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "MISSING", NULL);
+		return TCL_ERROR;
+	    }
+	    typePtr = objv[i+1];
+	    if (Tcl_ListObjLength(interp, typePtr, &length) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    i++;
+	    break;
+	case GLOB_LAST:				/* -- */
+	    i++;
+	    goto endOfForLoop;
+	}
+    }
+
+  endOfForLoop:
+    if ((globFlags & TCL_GLOBMODE_TAILS) && (pathOrDir == NULL)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"\"-tails\" must be used with either "
+		"\"-directory\" or \"-path\"", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB",
+		"BADOPTIONCOMBINATION", NULL);
+	return TCL_ERROR;
+    }
+
+    separators = NULL;		/* lint. */
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	separators = "/";
+	break;
+    case TCL_PLATFORM_WINDOWS:
+	separators = "/\\:";
+	break;
+    }
+
+    if (dir == PATH_GENERAL) {
+	int pathlength;
+	const char *last;
+	const char *first = TclGetStringFromObj(pathOrDir,&pathlength);
+
+	/*
+	 * Find the last path separator in the path
+	 */
+
+	last = first + pathlength;
+	for (; last != first; last--) {
+	    if (strchr(separators, *(last-1)) != NULL) {
+		break;
+	    }
+	}
+
+	if (last == first + pathlength) {
+	    /*
+	     * It's really a directory.
+	     */
+
+	    dir = PATH_DIR;
+
+	} else {
+	    Tcl_DString pref;
+	    char *search, *find;
+	    Tcl_DStringInit(&pref);
+	    if (last == first) {
+		/*
+		 * The whole thing is a prefix. This means we must remove any
+		 * 'tails' flag too, since it is irrelevant now (the same
+		 * effect will happen without it), but in particular its use
+		 * in TclGlob requires a non-NULL pathOrDir.
+		 */
+
+		Tcl_DStringAppend(&pref, first, -1);
+		globFlags &= ~TCL_GLOBMODE_TAILS;
+		pathOrDir = NULL;
+	    } else {
+		/*
+		 * Have to split off the end.
+		 */
+
+		Tcl_DStringAppend(&pref, last, first+pathlength-last);
+		pathOrDir = Tcl_NewStringObj(first, last-first-1);
+
+		/*
+		 * We must ensure that we haven't cut off too much, and turned
+		 * a valid path like '/' or 'C:/' into an incorrect path like
+		 * '' or 'C:'. The way we do this is to add a separator if
+		 * there are none presently in the prefix.
+		 */
+
+		if (strpbrk(Tcl_GetString(pathOrDir), "\\/") == NULL) {
+		    Tcl_AppendToObj(pathOrDir, last-1, 1);
+		}
+	    }
+
+	    /*
+	     * Need to quote 'prefix'.
+	     */
+
+	    Tcl_DStringInit(&prefix);
+	    search = Tcl_DStringValue(&pref);
+	    while ((find = (strpbrk(search, "\\[]*?{}"))) != NULL) {
+		Tcl_DStringAppend(&prefix, search, find-search);
+		TclDStringAppendLiteral(&prefix, "\\");
+		Tcl_DStringAppend(&prefix, find, 1);
+		search = find+1;
+		if (*search == '\0') {
+		    break;
+		}
+	    }
+	    if (*search != '\0') {
+		Tcl_DStringAppend(&prefix, search, -1);
+	    }
+	    Tcl_DStringFree(&pref);
+	}
+    }
+
+    if (pathOrDir != NULL) {
+	Tcl_IncrRefCount(pathOrDir);
+    }
+
+    if (typePtr != NULL) {
+	/*
+	 * The rest of the possible type arguments (except 'd') are platform
+	 * specific. We don't complain when they are used on an incompatible
+	 * platform.
+	 */
+
+	Tcl_ListObjLength(interp, typePtr, &length);
+	if (length <= 0) {
+	    goto skipTypes;
+	}
+	globTypes = TclStackAlloc(interp, sizeof(Tcl_GlobTypeData));
+	globTypes->type = 0;
+	globTypes->perm = 0;
+	globTypes->macType = NULL;
+	globTypes->macCreator = NULL;
+
+	while (--length >= 0) {
+	    int len;
+	    const char *str;
+
+	    Tcl_ListObjIndex(interp, typePtr, length, &look);
+	    str = TclGetStringFromObj(look, &len);
+	    if (strcmp("readonly", str) == 0) {
+		globTypes->perm |= TCL_GLOB_PERM_RONLY;
+	    } else if (strcmp("hidden", str) == 0) {
+		globTypes->perm |= TCL_GLOB_PERM_HIDDEN;
+	    } else if (len == 1) {
+		switch (str[0]) {
+		case 'r':
+		    globTypes->perm |= TCL_GLOB_PERM_R;
+		    break;
+		case 'w':
+		    globTypes->perm |= TCL_GLOB_PERM_W;
+		    break;
+		case 'x':
+		    globTypes->perm |= TCL_GLOB_PERM_X;
+		    break;
+		case 'b':
+		    globTypes->type |= TCL_GLOB_TYPE_BLOCK;
+		    break;
+		case 'c':
+		    globTypes->type |= TCL_GLOB_TYPE_CHAR;
+		    break;
+		case 'd':
+		    globTypes->type |= TCL_GLOB_TYPE_DIR;
+		    break;
+		case 'p':
+		    globTypes->type |= TCL_GLOB_TYPE_PIPE;
+		    break;
+		case 'f':
+		    globTypes->type |= TCL_GLOB_TYPE_FILE;
+		    break;
+		case 'l':
+		    globTypes->type |= TCL_GLOB_TYPE_LINK;
+		    break;
+		case 's':
+		    globTypes->type |= TCL_GLOB_TYPE_SOCK;
+		    break;
+		default:
+		    goto badTypesArg;
+		}
+
+	    } else if (len == 4) {
+		/*
+		 * This is assumed to be a MacOS file type.
+		 */
+
+		if (globTypes->macType != NULL) {
+		    goto badMacTypesArg;
+		}
+		globTypes->macType = look;
+		Tcl_IncrRefCount(look);
+
+	    } else {
+		Tcl_Obj *item;
+
+		if ((Tcl_ListObjLength(NULL, look, &len) == TCL_OK)
+			&& (len == 3)) {
+		    Tcl_ListObjIndex(interp, look, 0, &item);
+		    if (!strcmp("macintosh", Tcl_GetString(item))) {
+			Tcl_ListObjIndex(interp, look, 1, &item);
+			if (!strcmp("type", Tcl_GetString(item))) {
+			    Tcl_ListObjIndex(interp, look, 2, &item);
+			    if (globTypes->macType != NULL) {
+				goto badMacTypesArg;
+			    }
+			    globTypes->macType = item;
+			    Tcl_IncrRefCount(item);
+			    continue;
+			} else if (!strcmp("creator", Tcl_GetString(item))) {
+			    Tcl_ListObjIndex(interp, look, 2, &item);
+			    if (globTypes->macCreator != NULL) {
+				goto badMacTypesArg;
+			    }
+			    globTypes->macCreator = item;
+			    Tcl_IncrRefCount(item);
+			    continue;
+			}
+		    }
+		}
+
+		/*
+		 * Error cases. We reset the 'join' flag to zero, since we
+		 * haven't yet made use of it.
+		 */
+
+	    badTypesArg:
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"bad argument to \"-types\": %s",
+			Tcl_GetString(look)));
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "BAD", NULL);
+		result = TCL_ERROR;
+		join = 0;
+		goto endOfGlob;
+
+	    badMacTypesArg:
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"only one MacOS type or creator argument"
+			" to \"-types\" allowed", -1));
+		result = TCL_ERROR;
+		Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "BAD", NULL);
+		join = 0;
+		goto endOfGlob;
+	    }
+	}
+    }
+
+  skipTypes:
+    /*
+     * Now we perform the actual glob below. This may involve joining together
+     * the pattern arguments, dealing with particular file types etc. We use a
+     * 'goto' to ensure we free any memory allocated along the way.
+     */
+
+    objc -= i;
+    objv += i;
+    result = TCL_OK;
+
+    if (join) {
+	if (dir != PATH_GENERAL) {
+	    Tcl_DStringInit(&prefix);
+	}
+	for (i = 0; i < objc; i++) {
+	    TclDStringAppendObj(&prefix, objv[i]);
+	    if (i != objc -1) {
+		Tcl_DStringAppend(&prefix, separators, 1);
+	    }
+	}
+	if (TclGlob(interp, Tcl_DStringValue(&prefix), pathOrDir, globFlags,
+		globTypes) != TCL_OK) {
+	    result = TCL_ERROR;
+	    goto endOfGlob;
+	}
+    } else if (dir == PATH_GENERAL) {
+	Tcl_DString str;
+
+	Tcl_DStringInit(&str);
+	for (i = 0; i < objc; i++) {
+	    Tcl_DStringSetLength(&str, 0);
+	    if (dir == PATH_GENERAL) {
+		TclDStringAppendDString(&str, &prefix);
+	    }
+	    TclDStringAppendObj(&str, objv[i]);
+	    if (TclGlob(interp, Tcl_DStringValue(&str), pathOrDir, globFlags,
+		    globTypes) != TCL_OK) {
+		result = TCL_ERROR;
+		Tcl_DStringFree(&str);
+		goto endOfGlob;
+	    }
+	}
+	Tcl_DStringFree(&str);
+    } else {
+	for (i = 0; i < objc; i++) {
+	    string = Tcl_GetString(objv[i]);
+	    if (TclGlob(interp, string, pathOrDir, globFlags,
+		    globTypes) != TCL_OK) {
+		result = TCL_ERROR;
+		goto endOfGlob;
+	    }
+	}
+    }
+
+    if ((globFlags & TCL_GLOBMODE_NO_COMPLAIN) == 0) {
+	if (Tcl_ListObjLength(interp, Tcl_GetObjResult(interp),
+		&length) != TCL_OK) {
+	    /*
+	     * This should never happen. Maybe we should be more dramatic.
+	     */
+
+	    result = TCL_ERROR;
+	    goto endOfGlob;
+	}
+
+	if (length == 0) {
+	    Tcl_Obj *errorMsg =
+		    Tcl_ObjPrintf("no files matched glob pattern%s \"",
+			    (join || (objc == 1)) ? "" : "s");
+
+	    if (join) {
+		Tcl_AppendToObj(errorMsg, Tcl_DStringValue(&prefix), -1);
+	    } else {
+		const char *sep = "";
+
+		for (i = 0; i < objc; i++) {
+		    Tcl_AppendPrintfToObj(errorMsg, "%s%s",
+			    sep, Tcl_GetString(objv[i]));
+		    sep = " ";
+		}
+	    }
+	    Tcl_AppendToObj(errorMsg, "\"", -1);
+	    Tcl_SetObjResult(interp, errorMsg);
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB", "NOMATCH",
+		    NULL);
+	    result = TCL_ERROR;
+	}
+    }
+
+  endOfGlob:
+    if (join || (dir == PATH_GENERAL)) {
+	Tcl_DStringFree(&prefix);
+    }
+    if (pathOrDir != NULL) {
+	Tcl_DecrRefCount(pathOrDir);
+    }
+    if (globTypes != NULL) {
+	if (globTypes->macType != NULL) {
+	    Tcl_DecrRefCount(globTypes->macType);
+	}
+	if (globTypes->macCreator != NULL) {
+	    Tcl_DecrRefCount(globTypes->macCreator);
+	}
+	TclStackFree(interp, globTypes);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGlob --
+ *
+ *	This procedure prepares arguments for the DoGlob call. It sets the
+ *	separator string based on the platform, performs * tilde substitution,
+ *	and calls DoGlob.
+ *
+ *	The interpreter's result, on entry to this function, must be a valid
+ *	Tcl list (e.g. it could be empty), since we will lappend any new
+ *	results to that list. If it is not a valid list, this function will
+ *	fail to do anything very meaningful.
+ *
+ *	Note that if globFlags contains 'TCL_GLOBMODE_TAILS' then pathPrefix
+ *	cannot be NULL (it is only allowed with -dir or -path).
+ *
+ * Results:
+ *	The return value is a standard Tcl result indicating whether an error
+ *	occurred in globbing. After a normal return the result in interp (set
+ *	by DoGlob) holds all of the file names given by the pattern and
+ *	pathPrefix arguments. After an error the result in interp will hold
+ *	an error message.
+ *
+ * Side effects:
+ *	The 'pattern' is written to.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+TclGlob(
+    Tcl_Interp *interp,		/* Interpreter for returning error message or
+				 * appending list of matching file names. */
+    char *pattern,		/* Glob pattern to match. Must not refer to a
+				 * static string. */
+    Tcl_Obj *pathPrefix,	/* Path prefix to glob pattern, if non-null,
+				 * which is considered literally. */
+    int globFlags,		/* Stores or'ed combination of flags */
+    Tcl_GlobTypeData *types)	/* Struct containing acceptable types. May be
+				 * NULL. */
+{
+    const char *separators;
+    const char *head;
+    char *tail, *start;
+    int result;
+    Tcl_Obj *filenamesObj, *savedResultObj;
+
+    separators = NULL;		/* lint. */
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	separators = "/";
+	break;
+    case TCL_PLATFORM_WINDOWS:
+	separators = "/\\:";
+	break;
+    }
+
+    if (pathPrefix == NULL) {
+	char c;
+	Tcl_DString buffer;
+	Tcl_DStringInit(&buffer);
+
+	start = pattern;
+
+	/*
+	 * Perform tilde substitution, if needed.
+	 */
+
+	if (start[0] == '~') {
+	    /*
+	     * Find the first path separator after the tilde.
+	     */
+
+	    for (tail = start; *tail != '\0'; tail++) {
+		if (*tail == '\\') {
+		    if (strchr(separators, tail[1]) != NULL) {
+			break;
+		    }
+		} else if (strchr(separators, *tail) != NULL) {
+		    break;
+		}
+	    }
+
+	    /*
+	     * Determine the home directory for the specified user.
+	     */
+
+	    c = *tail;
+	    *tail = '\0';
+	    head = DoTildeSubst(interp, start+1, &buffer);
+	    *tail = c;
+	    if (head == NULL) {
+		return TCL_ERROR;
+	    }
+	    if (head != Tcl_DStringValue(&buffer)) {
+		Tcl_DStringAppend(&buffer, head, -1);
+	    }
+	    pathPrefix = TclDStringToObj(&buffer);
+	    Tcl_IncrRefCount(pathPrefix);
+	    globFlags |= TCL_GLOBMODE_DIR;
+	    if (c != '\0') {
+		tail++;
+	    }
+	    Tcl_DStringFree(&buffer);
+	} else {
+	    tail = pattern;
+	}
+    } else {
+	Tcl_IncrRefCount(pathPrefix);
+	tail = pattern;
+    }
+
+    /*
+     * Handling empty path prefixes with glob patterns like 'C:' or
+     * 'c:////////' is a pain on Windows if we leave it too late, since these
+     * aren't really patterns at all! We therefore check the head of the
+     * pattern now for such cases, if we don't have an unquoted prefix yet.
+     *
+     * Similarly on Unix with '/' at the head of the pattern -- it just
+     * indicates the root volume, so we treat it as such.
+     */
+
+    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+	if (pathPrefix == NULL && tail[0] != '\0' && tail[1] == ':') {
+	    char *p = tail + 1;
+	    pathPrefix = Tcl_NewStringObj(tail, 1);
+	    while (*p != '\0') {
+		char c = p[1];
+		if (*p == '\\') {
+		    if (strchr(separators, c) != NULL) {
+			if (c == '\\') {
+			    c = '/';
+			}
+			Tcl_AppendToObj(pathPrefix, &c, 1);
+			p++;
+		    } else {
+			break;
+		    }
+		} else if (strchr(separators, *p) != NULL) {
+		    Tcl_AppendToObj(pathPrefix, p, 1);
+		} else {
+		    break;
+		}
+		p++;
+	    }
+	    tail = p;
+	    Tcl_IncrRefCount(pathPrefix);
+	} else if (pathPrefix == NULL && (tail[0] == '/'
+		|| (tail[0] == '\\' && tail[1] == '\\'))) {
+	    int driveNameLen;
+	    Tcl_Obj *driveName;
+	    Tcl_Obj *temp = Tcl_NewStringObj(tail, -1);
+	    Tcl_IncrRefCount(temp);
+
+	    switch (TclGetPathType(temp, NULL, &driveNameLen, &driveName)) {
+	    case TCL_PATH_VOLUME_RELATIVE: {
+		/*
+		 * Volume relative path which is equivalent to a path in the
+		 * root of the cwd's volume. We will actually return
+		 * non-volume-relative paths here. i.e. 'glob /foo*' will
+		 * return 'C:/foobar'. This is much the same as globbing for a
+		 * path with '\\' will return one with '/' on Windows.
+		 */
+
+		Tcl_Obj *cwd = Tcl_FSGetCwd(interp);
+
+		if (cwd == NULL) {
+		    Tcl_DecrRefCount(temp);
+		    return TCL_ERROR;
+		}
+		pathPrefix = Tcl_NewStringObj(Tcl_GetString(cwd), 3);
+		Tcl_DecrRefCount(cwd);
+		if (tail[0] == '/') {
+		    tail++;
+		} else {
+		    tail += 2;
+		}
+		Tcl_IncrRefCount(pathPrefix);
+		break;
+	    }
+	    case TCL_PATH_ABSOLUTE:
+		/*
+		 * Absolute, possibly network path //Machine/Share. Use that
+		 * as the path prefix (it already has a refCount).
+		 */
+
+		pathPrefix = driveName;
+		tail += driveNameLen;
+		break;
+	    case TCL_PATH_RELATIVE:
+		/* Do nothing */
+		break;
+	    }
+	    Tcl_DecrRefCount(temp);
+	}
+
+	/*
+	 * ':' no longer needed as a separator. It is only relevant to the
+	 * beginning of the path.
+	 */
+
+	separators = "/\\";
+
+    } else if (tclPlatform == TCL_PLATFORM_UNIX) {
+	if (pathPrefix == NULL && tail[0] == '/') {
+	    pathPrefix = Tcl_NewStringObj(tail, 1);
+	    tail++;
+	    Tcl_IncrRefCount(pathPrefix);
+	}
+    }
+
+    /*
+     * Finally if we still haven't managed to generate a path prefix, check if
+     * the path starts with a current volume.
+     */
+
+    if (pathPrefix == NULL) {
+	int driveNameLen;
+	Tcl_Obj *driveName;
+	if (TclFSNonnativePathType(tail, (int) strlen(tail), NULL,
+		&driveNameLen, &driveName) == TCL_PATH_ABSOLUTE) {
+	    pathPrefix = driveName;
+	    tail += driveNameLen;
+	}
+    }
+
+    /*
+     * To process a [glob] invokation, this function may be called multiple
+     * times. Each time, the previously discovered filenames are in the
+     * interpreter result. We stash that away here so the result is free for
+     * error messsages.
+     */
+
+    savedResultObj = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(savedResultObj);
+    Tcl_ResetResult(interp);
+    TclNewObj(filenamesObj);
+    Tcl_IncrRefCount(filenamesObj);
+
+    /*
+     * Now we do the actual globbing, adding filenames as we go to buffer in
+     * filenamesObj
+     */
+
+    if (*tail == '\0' && pathPrefix != NULL) {
+	/*
+	 * An empty pattern. This means 'pathPrefix' is actually a full path
+	 * of a file/directory we want to simply check for existence and type.
+	 */
+
+	if (types == NULL) {
+	    /*
+	     * We just want to check for existence. In this case we make it
+	     * easy on Tcl_FSMatchInDirectory and its sub-implementations by
+	     * not bothering them (even though they should support this
+	     * situation) and we just use the simple existence check with
+	     * Tcl_FSAccess.
+	     */
+
+	    if (Tcl_FSAccess(pathPrefix, F_OK) == 0) {
+		Tcl_ListObjAppendElement(interp, filenamesObj, pathPrefix);
+	    }
+	    result = TCL_OK;
+	} else {
+	    /*
+	     * We want to check for the correct type. Tcl_FSMatchInDirectory
+	     * is documented to do this for us, if we give it a NULL pattern.
+	     */
+
+	    result = Tcl_FSMatchInDirectory(interp, filenamesObj, pathPrefix,
+		    NULL, types);
+	}
+    } else {
+	result = DoGlob(interp, filenamesObj, separators, pathPrefix,
+		globFlags & TCL_GLOBMODE_DIR, tail, types);
+    }
+
+    /*
+     * Check for errors...
+     */
+
+    if (result != TCL_OK) {
+	TclDecrRefCount(filenamesObj);
+	TclDecrRefCount(savedResultObj);
+	if (pathPrefix != NULL) {
+	    Tcl_DecrRefCount(pathPrefix);
+	}
+	return result;
+    }
+
+    /*
+     * If we only want the tails, we must strip off the prefix now. It may
+     * seem more efficient to pass the tails flag down into DoGlob,
+     * Tcl_FSMatchInDirectory, but those functions are continually adjusting
+     * the prefix as the various pieces of the pattern are assimilated, so
+     * that would add a lot of complexity to the code. This way is a little
+     * slower (when the -tails flag is given), but much simpler to code.
+     *
+     * We do it by rewriting the result list in-place.
+     */
+
+    if (globFlags & TCL_GLOBMODE_TAILS) {
+	int objc, i;
+	Tcl_Obj **objv;
+	int prefixLen;
+	const char *pre;
+
+	/*
+	 * If this length has never been set, set it here.
+	 */
+
+	if (pathPrefix == NULL) {
+	    Tcl_Panic("Called TclGlob with TCL_GLOBMODE_TAILS and pathPrefix==NULL");
+	}
+
+	pre = TclGetStringFromObj(pathPrefix, &prefixLen);
+	if (prefixLen > 0
+		&& (strchr(separators, pre[prefixLen-1]) == NULL)) {
+	    /*
+	     * If we're on Windows and the prefix is a volume relative one
+	     * like 'C:', then there won't be a path separator in between, so
+	     * no need to skip it here.
+	     */
+
+	    if ((tclPlatform != TCL_PLATFORM_WINDOWS) || (prefixLen != 2)
+		    || (pre[1] != ':')) {
+		prefixLen++;
+	    }
+	}
+
+	Tcl_ListObjGetElements(NULL, filenamesObj, &objc, &objv);
+	for (i = 0; i< objc; i++) {
+	    int len;
+	    const char *oldStr = TclGetStringFromObj(objv[i], &len);
+	    Tcl_Obj *elem;
+
+	    if (len == prefixLen) {
+		if ((pattern[0] == '\0')
+			|| (strchr(separators, pattern[0]) == NULL)) {
+		    TclNewLiteralStringObj(elem, ".");
+		} else {
+		    TclNewLiteralStringObj(elem, "/");
+		}
+	    } else {
+		elem = Tcl_NewStringObj(oldStr+prefixLen, len-prefixLen);
+	    }
+	    Tcl_ListObjReplace(interp, filenamesObj, i, 1, 1, &elem);
+	}
+    }
+
+    /*
+     * Now we have a list of discovered filenames in filenamesObj and a list
+     * of previously discovered (saved earlier from the interpreter result) in
+     * savedResultObj. Merge them and put them back in the interpreter result.
+     */
+
+    if (Tcl_IsShared(savedResultObj)) {
+	TclDecrRefCount(savedResultObj);
+	savedResultObj = Tcl_DuplicateObj(savedResultObj);
+	Tcl_IncrRefCount(savedResultObj);
+    }
+    if (Tcl_ListObjAppendList(interp, savedResultObj, filenamesObj) != TCL_OK){
+	result = TCL_ERROR;
+    } else {
+	Tcl_SetObjResult(interp, savedResultObj);
+    }
+    TclDecrRefCount(savedResultObj);
+    TclDecrRefCount(filenamesObj);
+    if (pathPrefix != NULL) {
+	Tcl_DecrRefCount(pathPrefix);
+    }
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SkipToChar --
+ *
+ *	This function traverses a glob pattern looking for the next unquoted
+ *	occurance of the specified character at the same braces nesting level.
+ *
+ * Results:
+ *	Updates stringPtr to point to the matching character, or to the end of
+ *	the string if nothing matched. The return value is 1 if a match was
+ *	found at the top level, otherwise it is 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SkipToChar(
+    char **stringPtr,		/* Pointer string to check. */
+    int match)			/* Character to find. */
+{
+    int quoted, level;
+    register char *p;
+
+    quoted = 0;
+    level = 0;
+
+    for (p = *stringPtr; *p != '\0'; p++) {
+	if (quoted) {
+	    quoted = 0;
+	    continue;
+	}
+	if ((level == 0) && (*p == match)) {
+	    *stringPtr = p;
+	    return 1;
+	}
+	if (*p == '{') {
+	    level++;
+	} else if (*p == '}') {
+	    level--;
+	} else if (*p == '\\') {
+	    quoted = 1;
+	}
+    }
+    *stringPtr = p;
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DoGlob --
+ *
+ *	This recursive procedure forms the heart of the globbing code. It
+ *	performs a depth-first traversal of the tree given by the path name to
+ *	be globbed and the pattern. The directory and remainder are assumed to
+ *	be native format paths. The prefix contained in 'pathPtr' is either a
+ *	directory or path from which to start the search (or NULL). If pathPtr
+ *	is NULL, then the pattern must not start with an absolute path
+ *	specification (that case should be handled by moving the absolute path
+ *	prefix into pathPtr before calling DoGlob).
+ *
+ * Results:
+ *	The return value is a standard Tcl result indicating whether an error
+ *	occurred in globbing. After a normal return the result in interp will
+ *	be set to hold all of the file names given by the dir and remaining
+ *	arguments. After an error the result in interp will hold an error
+ *	message.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DoGlob(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting
+				 * (e.g. unmatched brace). */
+    Tcl_Obj *matchesObj,	/* Unshared list object in which to place all
+				 * resulting filenames. Caller allocates and
+				 * deallocates; DoGlob must not touch the
+				 * refCount of this object. */
+    const char *separators,	/* String containing separator characters that
+				 * should be used to identify globbing
+				 * boundaries. */
+    Tcl_Obj *pathPtr,		/* Completely expanded prefix. */
+    int flags,			/* If non-zero then pathPtr is a directory */
+    char *pattern,		/* The pattern to match against. Must not be a
+				 * pointer to a static string. */
+    Tcl_GlobTypeData *types)	/* List object containing list of acceptable
+				 * types. May be NULL. */
+{
+    int baseLength, quoted, count;
+    int result = TCL_OK;
+    char *name, *p, *openBrace, *closeBrace, *firstSpecialChar;
+    Tcl_Obj *joinedPtr;
+
+    /*
+     * Consume any leading directory separators, leaving pattern pointing just
+     * past the last initial separator.
+     */
+
+    count = 0;
+    name = pattern;
+    for (; *pattern != '\0'; pattern++) {
+	if (*pattern == '\\') {
+	    /*
+	     * If the first character is escaped, either we have a directory
+	     * separator, or we have any other character. In the latter case
+	     * the rest is a pattern, and we must break from the loop. This
+	     * is particularly important on Windows where '\' is both the
+	     * escaping character and a directory separator.
+	     */
+
+	    if (strchr(separators, pattern[1]) != NULL) {
+		pattern++;
+	    } else {
+		break;
+	    }
+	} else if (strchr(separators, *pattern) == NULL) {
+	    break;
+	}
+	count++;
+    }
+
+    /*
+     * Look for the first matching pair of braces or the first directory
+     * separator that is not inside a pair of braces.
+     */
+
+    openBrace = closeBrace = NULL;
+    quoted = 0;
+    for (p = pattern; *p != '\0'; p++) {
+	if (quoted) {
+	    quoted = 0;
+
+	} else if (*p == '\\') {
+	    quoted = 1;
+	    if (strchr(separators, p[1]) != NULL) {
+		/*
+		 * Quoted directory separator.
+		 */
+		break;
+	    }
+
+	} else if (strchr(separators, *p) != NULL) {
+	    /*
+	     * Unquoted directory separator.
+	     */
+	    break;
+
+	} else if (*p == '{') {
+	    openBrace = p;
+	    p++;
+	    if (SkipToChar(&p, '}')) {
+		/*
+		 * Balanced braces.
+		 */
+
+		closeBrace = p;
+		break;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unmatched open-brace in file name", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB", "BALANCE",
+		    NULL);
+	    return TCL_ERROR;
+
+	} else if (*p == '}') {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unmatched close-brace in file name", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "GLOB", "BALANCE",
+		    NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Substitute the alternate patterns from the braces and recurse.
+     */
+
+    if (openBrace != NULL) {
+	char *element;
+	Tcl_DString newName;
+
+	Tcl_DStringInit(&newName);
+
+	/*
+	 * For each element within in the outermost pair of braces, append the
+	 * element and the remainder to the fixed portion before the first
+	 * brace and recursively call DoGlob.
+	 */
+
+	Tcl_DStringAppend(&newName, pattern, openBrace-pattern);
+	baseLength = Tcl_DStringLength(&newName);
+	*closeBrace = '\0';
+	for (p = openBrace; p != closeBrace; ) {
+	    p++;
+	    element = p;
+	    SkipToChar(&p, ',');
+	    Tcl_DStringSetLength(&newName, baseLength);
+	    Tcl_DStringAppend(&newName, element, p-element);
+	    Tcl_DStringAppend(&newName, closeBrace+1, -1);
+	    result = DoGlob(interp, matchesObj, separators, pathPtr, flags,
+		    Tcl_DStringValue(&newName), types);
+	    if (result != TCL_OK) {
+		break;
+	    }
+	}
+	*closeBrace = '}';
+	Tcl_DStringFree(&newName);
+	return result;
+    }
+
+    /*
+     * At this point, there are no more brace substitutions to perform on this
+     * path component. The variable p is pointing at a quoted or unquoted
+     * directory separator or the end of the string. So we need to check for
+     * special globbing characters in the current pattern. We avoid modifying
+     * pattern if p is pointing at the end of the string.
+     *
+     * If we find any globbing characters, then we must call
+     * Tcl_FSMatchInDirectory. If we're at the end of the string, then that's
+     * all we need to do. If we're not at the end of the string, then we must
+     * recurse, so we do that below.
+     *
+     * Alternatively, if there are no globbing characters then again there are
+     * two cases. If we're at the end of the string, we just need to check for
+     * the given path's existence and type. If we're not at the end of the
+     * string, we recurse.
+     */
+
+    if (*p != '\0') {
+	char savedChar = *p;
+
+	/*
+	 * Note that we are modifying the string in place. This won't work if
+	 * the string is a static.
+	 */
+
+	*p = '\0';
+	firstSpecialChar = strpbrk(pattern, "*[]?\\");
+	*p = savedChar;
+    } else {
+	firstSpecialChar = strpbrk(pattern, "*[]?\\");
+    }
+
+    if (firstSpecialChar != NULL) {
+	/*
+	 * Look for matching files in the given directory. The implementation
+	 * of this function is filesystem specific. For each file that
+	 * matches, it will add the match onto the resultPtr given.
+	 */
+
+	static Tcl_GlobTypeData dirOnly = {
+	    TCL_GLOB_TYPE_DIR, 0, NULL, NULL
+	};
+	char save = *p;
+	Tcl_Obj *subdirsPtr;
+
+	if (*p == '\0') {
+	    return Tcl_FSMatchInDirectory(interp, matchesObj, pathPtr,
+		    pattern, types);
+	}
+
+	/*
+	 * We do the recursion ourselves. This makes implementing
+	 * Tcl_FSMatchInDirectory for each filesystem much easier.
+	 */
+
+	*p = '\0';
+	TclNewObj(subdirsPtr);
+	Tcl_IncrRefCount(subdirsPtr);
+	result = Tcl_FSMatchInDirectory(interp, subdirsPtr, pathPtr,
+		pattern, &dirOnly);
+	*p = save;
+	if (result == TCL_OK) {
+	    int subdirc, i, repair = -1;
+	    Tcl_Obj **subdirv;
+
+	    result = Tcl_ListObjGetElements(interp, subdirsPtr,
+		    &subdirc, &subdirv);
+	    for (i=0; result==TCL_OK && i<subdirc; i++) {
+		Tcl_Obj *copy = NULL;
+
+		if (pathPtr == NULL && Tcl_GetString(subdirv[i])[0] == '~') {
+		    Tcl_ListObjLength(NULL, matchesObj, &repair);
+		    copy = subdirv[i];
+		    subdirv[i] = Tcl_NewStringObj("./", 2);
+		    Tcl_AppendObjToObj(subdirv[i], copy);
+		    Tcl_IncrRefCount(subdirv[i]);
+		}
+		result = DoGlob(interp, matchesObj, separators, subdirv[i],
+			1, p+1, types);
+		if (copy) {
+		    int end;
+
+		    Tcl_DecrRefCount(subdirv[i]);
+		    subdirv[i] = copy;
+		    Tcl_ListObjLength(NULL, matchesObj, &end);
+		    while (repair < end) {
+			const char *bytes;
+			int numBytes;
+			Tcl_Obj *fixme, *newObj;
+
+			Tcl_ListObjIndex(NULL, matchesObj, repair, &fixme);
+			bytes = TclGetStringFromObj(fixme, &numBytes);
+			newObj = Tcl_NewStringObj(bytes+2, numBytes-2);
+			Tcl_ListObjReplace(NULL, matchesObj, repair, 1,
+				1, &newObj);
+			repair++;
+		    }
+		    repair = -1;
+		}
+	    }
+	}
+	TclDecrRefCount(subdirsPtr);
+	return result;
+    }
+
+    /*
+     * We reach here with no pattern char in current section
+     */
+
+    if (*p == '\0') {
+	int length;
+	Tcl_DString append;
+
+	/*
+	 * This is the code path reached by a command like 'glob foo'.
+	 *
+	 * There are no more wildcards in the pattern and no more unprocessed
+	 * characters in the pattern, so now we can construct the path, and
+	 * pass it to Tcl_FSMatchInDirectory with an empty pattern to verify
+	 * the existence of the file and check it is of the correct type (if a
+	 * 'types' flag it given -- if no such flag was given, we could just
+	 * use 'Tcl_FSLStat', but for simplicity we keep to a common
+	 * approach).
+	 */
+
+	Tcl_DStringInit(&append);
+	Tcl_DStringAppend(&append, pattern, p-pattern);
+
+	if (pathPtr != NULL) {
+	    (void) TclGetStringFromObj(pathPtr, &length);
+	} else {
+	    length = 0;
+	}
+
+	switch (tclPlatform) {
+	case TCL_PLATFORM_WINDOWS:
+	    if (length == 0 && (Tcl_DStringLength(&append) == 0)) {
+		if (((*name == '\\') && (name[1] == '/' ||
+			name[1] == '\\')) || (*name == '/')) {
+		    TclDStringAppendLiteral(&append, "/");
+		} else {
+		    TclDStringAppendLiteral(&append, ".");
+		}
+	    }
+
+	    break;
+
+	case TCL_PLATFORM_UNIX:
+	    if (length == 0 && (Tcl_DStringLength(&append) == 0)) {
+		if ((*name == '\\' && name[1] == '/') || (*name == '/')) {
+		    TclDStringAppendLiteral(&append, "/");
+		} else {
+		    TclDStringAppendLiteral(&append, ".");
+		}
+	    }
+	    break;
+	}
+
+	/*
+	 * Common for all platforms.
+	 */
+
+	if (pathPtr == NULL) {
+	    joinedPtr = TclDStringToObj(&append);
+	} else if (flags) {
+	    joinedPtr = TclNewFSPathObj(pathPtr, Tcl_DStringValue(&append),
+		    Tcl_DStringLength(&append));
+	} else {
+	    joinedPtr = Tcl_DuplicateObj(pathPtr);
+	    if (strchr(separators, Tcl_DStringValue(&append)[0]) == NULL) {
+		/*
+		 * The current prefix must end in a separator.
+		 */
+
+		int len;
+		const char *joined = TclGetStringFromObj(joinedPtr,&len);
+
+		if (strchr(separators, joined[len-1]) == NULL) {
+		    Tcl_AppendToObj(joinedPtr, "/", 1);
+		}
+	    }
+	    Tcl_AppendToObj(joinedPtr, Tcl_DStringValue(&append),
+		    Tcl_DStringLength(&append));
+	}
+	Tcl_IncrRefCount(joinedPtr);
+	Tcl_DStringFree(&append);
+	result = Tcl_FSMatchInDirectory(interp, matchesObj, joinedPtr, NULL,
+		types);
+	Tcl_DecrRefCount(joinedPtr);
+	return result;
+    }
+
+    /*
+     * If it's not the end of the string, we must recurse
+     */
+
+    if (pathPtr == NULL) {
+	joinedPtr = Tcl_NewStringObj(pattern, p-pattern);
+    } else if (flags) {
+	joinedPtr = TclNewFSPathObj(pathPtr, pattern, p-pattern);
+    } else {
+	joinedPtr = Tcl_DuplicateObj(pathPtr);
+	if (strchr(separators, pattern[0]) == NULL) {
+	    /*
+	     * The current prefix must end in a separator, unless this is a
+	     * volume-relative path. In particular globbing in Windows shares,
+	     * when not using -dir or -path, e.g. 'glob [file join
+	     * //machine/share/subdir *]' requires adding a separator here.
+	     * This behaviour is not currently tested for in the test suite.
+	     */
+
+	    int len;
+	    const char *joined = TclGetStringFromObj(joinedPtr,&len);
+
+	    if (strchr(separators, joined[len-1]) == NULL) {
+		if (Tcl_FSGetPathType(pathPtr) != TCL_PATH_VOLUME_RELATIVE) {
+		    Tcl_AppendToObj(joinedPtr, "/", 1);
+		}
+	    }
+	}
+	Tcl_AppendToObj(joinedPtr, pattern, p-pattern);
+    }
+
+    Tcl_IncrRefCount(joinedPtr);
+    result = DoGlob(interp, matchesObj, separators, joinedPtr, 1, p, types);
+    Tcl_DecrRefCount(joinedPtr);
+
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_AllocStatBuf --
+ *
+ *	This procedure allocates a Tcl_StatBuf on the heap. It exists so that
+ *	extensions may be used unchanged on systems where largefile support is
+ *	optional.
+ *
+ * Results:
+ *	A pointer to a Tcl_StatBuf which may be deallocated by being passed to
+ *	ckfree().
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_StatBuf *
+Tcl_AllocStatBuf(void)
+{
+    return ckalloc(sizeof(Tcl_StatBuf));
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Access functions for Tcl_StatBuf --
+ *
+ *	These functions provide portable read-only access to the portable
+ *	fields of the Tcl_StatBuf structure (really a 'struct stat', 'struct
+ *	stat64' or something else related). [TIP #316]
+ *
+ * Results:
+ *	The value from the field being retrieved.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+unsigned
+Tcl_GetFSDeviceFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (unsigned) statPtr->st_dev;
+}
+
+unsigned
+Tcl_GetFSInodeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (unsigned) statPtr->st_ino;
+}
+
+unsigned
+Tcl_GetModeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (unsigned) statPtr->st_mode;
+}
+
+int
+Tcl_GetLinkCountFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (int)statPtr->st_nlink;
+}
+
+int
+Tcl_GetUserIdFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (int) statPtr->st_uid;
+}
+
+int
+Tcl_GetGroupIdFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (int) statPtr->st_gid;
+}
+
+int
+Tcl_GetDeviceTypeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (int) statPtr->st_rdev;
+}
+
+Tcl_WideInt
+Tcl_GetAccessTimeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (Tcl_WideInt) statPtr->st_atime;
+}
+
+Tcl_WideInt
+Tcl_GetModificationTimeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (Tcl_WideInt) statPtr->st_mtime;
+}
+
+Tcl_WideInt
+Tcl_GetChangeTimeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (Tcl_WideInt) statPtr->st_ctime;
+}
+
+Tcl_WideUInt
+Tcl_GetSizeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+    return (Tcl_WideUInt) statPtr->st_size;
+}
+
+Tcl_WideUInt
+Tcl_GetBlocksFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+#ifdef HAVE_STRUCT_STAT_ST_BLOCKS
+    return (Tcl_WideUInt) statPtr->st_blocks;
+#else
+    register unsigned blksize = Tcl_GetBlockSizeFromStat(statPtr);
+
+    return ((Tcl_WideUInt) statPtr->st_size + blksize - 1) / blksize;
+#endif
+}
+
+unsigned
+Tcl_GetBlockSizeFromStat(
+    const Tcl_StatBuf *statPtr)
+{
+#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
+    return (unsigned) statPtr->st_blksize;
+#else
+    /*
+     * Not a great guess, but will do...
+     */
+
+    return GUESSED_BLOCK_SIZE;
+#endif
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclFileSystem.h b/generic/tclFileSystem.h
new file mode 100644
index 0000000..1eec7ff
--- /dev/null
+++ b/generic/tclFileSystem.h
@@ -0,0 +1,74 @@
+/*
+ * tclFileSystem.h --
+ *
+ *	This file contains the common defintions and prototypes for use by
+ *	Tcl's filesystem and path handling layers.
+ *
+ * Copyright (c) 2003 Vince Darley.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLFILESYSTEM
+#define _TCLFILESYSTEM
+
+#include "tcl.h"
+
+/*
+ * The internal TclFS API provides routines for handling and manipulating
+ * paths efficiently, taking direct advantage of the "path" Tcl_Obj type.
+ *
+ * These functions are not exported at all at present.
+ */
+
+MODULE_SCOPE int	TclFSCwdPointerEquals(Tcl_Obj **pathPtrPtr);
+MODULE_SCOPE int	TclFSNormalizeToUniquePath(Tcl_Interp *interp,
+			    Tcl_Obj *pathPtr, int startAt);
+MODULE_SCOPE Tcl_Obj *	TclFSMakePathRelative(Tcl_Interp *interp,
+			    Tcl_Obj *pathPtr, Tcl_Obj *cwdPtr);
+MODULE_SCOPE int	TclFSEnsureEpochOk(Tcl_Obj *pathPtr,
+			    const Tcl_Filesystem **fsPtrPtr);
+MODULE_SCOPE void	TclFSSetPathDetails(Tcl_Obj *pathPtr,
+			    const Tcl_Filesystem *fsPtr, ClientData clientData);
+MODULE_SCOPE Tcl_Obj *	TclFSNormalizeAbsolutePath(Tcl_Interp *interp,
+			    Tcl_Obj *pathPtr);
+MODULE_SCOPE size_t	TclFSEpoch(void);
+
+/*
+ * Private shared variables for use by tclIOUtil.c and tclPathObj.c
+ */
+
+MODULE_SCOPE const Tcl_Filesystem tclNativeFilesystem;
+
+/*
+ * Private shared functions for use by tclIOUtil.c, tclPathObj.c and
+ * tclFileName.c, and any platform-specific filesystem code.
+ */
+
+MODULE_SCOPE Tcl_PathType TclFSGetPathType(Tcl_Obj *pathPtr,
+			    const Tcl_Filesystem **filesystemPtrPtr,
+			    int *driveNameLengthPtr);
+MODULE_SCOPE Tcl_PathType TclFSNonnativePathType(const char *pathPtr,
+			    int pathLen, const Tcl_Filesystem **filesystemPtrPtr,
+			    int *driveNameLengthPtr, Tcl_Obj **driveNameRef);
+MODULE_SCOPE Tcl_PathType TclGetPathType(Tcl_Obj *pathPtr,
+			    const Tcl_Filesystem **filesystemPtrPtr,
+			    int *driveNameLengthPtr, Tcl_Obj **driveNameRef);
+MODULE_SCOPE int	TclFSEpochOk(size_t filesystemEpoch);
+MODULE_SCOPE int	TclFSCwdIsNative(void);
+MODULE_SCOPE Tcl_Obj *	TclWinVolumeRelativeNormalize(Tcl_Interp *interp,
+			    const char *path, Tcl_Obj **useThisCwdPtr);
+
+MODULE_SCOPE Tcl_FSPathInFilesystemProc TclNativePathInFilesystem;
+MODULE_SCOPE Tcl_FSCreateInternalRepProc TclNativeCreateNativeRep;
+
+#endif /* _TCLFILESYSTEM */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclGet.c b/generic/tclGet.c
new file mode 100644
index 0000000..97e8c7b
--- /dev/null
+++ b/generic/tclGet.c
@@ -0,0 +1,156 @@
+/*
+ * tclGet.c --
+ *
+ *	This file contains functions to convert strings into other forms, like
+ *	integers or floating-point numbers or booleans, doing syntax checking
+ *	along the way.
+ *
+ * Copyright (c) 1990-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetInt --
+ *
+ *	Given a string, produce the corresponding integer value.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case *intPtr will be set
+ *	to the integer value equivalent to src.  If src is improperly formed
+ *	then TCL_ERROR is returned and an error message will be left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetInt(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    const char *src,		/* String containing a (possibly signed)
+				 * integer in a form acceptable to
+				 * Tcl_GetIntFromObj(). */
+    int *intPtr)		/* Place to store converted result. */
+{
+    Tcl_Obj obj;
+    int code;
+
+    obj.refCount = 1;
+    obj.bytes = (char *) src;
+    obj.length = strlen(src);
+    obj.typePtr = NULL;
+
+    code = Tcl_GetIntFromObj(interp, &obj, intPtr);
+    if (obj.refCount > 1) {
+	Tcl_Panic("invalid sharing of Tcl_Obj on C stack");
+    }
+    TclFreeIntRep(&obj);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetDouble --
+ *
+ *	Given a string, produce the corresponding double-precision
+ *	floating-point value.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case *doublePtr will be
+ *	set to the double-precision value equivalent to src. If src is
+ *	improperly formed then TCL_ERROR is returned and an error message will
+ *	be left in the interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetDouble(
+    Tcl_Interp *interp,		/* Interpreter used for error reporting. */
+    const char *src,		/* String containing a floating-point number
+				 * in a form acceptable to
+				 * Tcl_GetDoubleFromObj(). */
+    double *doublePtr)		/* Place to store converted result. */
+{
+    Tcl_Obj obj;
+    int code;
+
+    obj.refCount = 1;
+    obj.bytes = (char *) src;
+    obj.length = strlen(src);
+    obj.typePtr = NULL;
+
+    code = Tcl_GetDoubleFromObj(interp, &obj, doublePtr);
+    if (obj.refCount > 1) {
+	Tcl_Panic("invalid sharing of Tcl_Obj on C stack");
+    }
+    TclFreeIntRep(&obj);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetBoolean --
+ *
+ *	Given a string, return a 0/1 boolean value corresponding to the
+ *	string.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case *boolPtr will be set
+ *	to the 0/1 value equivalent to src. If src is improperly formed then
+ *	TCL_ERROR is returned and an error message will be left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetBoolean(
+    Tcl_Interp *interp,		/* Interpreter used for error reporting. */
+    const char *src,		/* String containing one of the boolean values
+				 * 1, 0, true, false, yes, no, on, off. */
+    int *boolPtr)		/* Place to store converted result, which will
+				 * be 0 or 1. */
+{
+    Tcl_Obj obj;
+    int code;
+
+    obj.refCount = 1;
+    obj.bytes = (char *) src;
+    obj.length = strlen(src);
+    obj.typePtr = NULL;
+
+    code = TclSetBooleanFromAny(interp, &obj);
+    if (obj.refCount > 1) {
+	Tcl_Panic("invalid sharing of Tcl_Obj on C stack");
+    }
+    if (code == TCL_OK) {
+	*boolPtr = obj.internalRep.longValue;
+    }
+    return code;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclGetDate.y b/generic/tclGetDate.y
new file mode 100644
index 0000000..da4c3fd
--- /dev/null
+++ b/generic/tclGetDate.y
@@ -0,0 +1,1130 @@
+/*
+ * tclGetDate.y --
+ *
+ *	Contains yacc grammar for parsing date and time strings. The output of
+ *	this file should be the file tclDate.c which is used directly in the
+ *	Tcl sources. Note that this file is largely obsolete in Tcl 8.5; it is
+ *	only used when doing free-form date parsing, an ill-defined process
+ *	anyway.
+ *
+ * Copyright (c) 1992-1995 Karl Lehenbauer and Mark Diekhans.
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+%parse-param {DateInfo* info}
+%lex-param {DateInfo* info}
+%pure-parser
+ /* %error-verbose would be nice, but our token names are meaningless */
+%locations
+
+%{
+/*
+ * tclDate.c --
+ *
+ *	This file is generated from a yacc grammar defined in the file
+ *	tclGetDate.y. It should not be edited directly.
+ *
+ * Copyright (c) 1992-1995 Karl Lehenbauer and Mark Diekhans.
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ *
+ */
+#include "tclInt.h"
+
+/*
+ * Bison generates several labels that happen to be unused. MS Visual C++
+ * doesn't like that, and complains. Tell it to shut up.
+ */
+
+#ifdef _MSC_VER
+#pragma warning( disable : 4102 )
+#endif /* _MSC_VER */
+
+/*
+ * yyparse will accept a 'struct DateInfo' as its parameter; that's where the
+ * parsed fields will be returned.
+ */
+
+typedef struct DateInfo {
+
+    Tcl_Obj* messages;		/* Error messages */
+    const char* separatrix;	/* String separating messages */
+
+    time_t dateYear;
+    time_t dateMonth;
+    time_t dateDay;
+    int dateHaveDate;
+
+    time_t dateHour;
+    time_t dateMinutes;
+    time_t dateSeconds;
+    int dateMeridian;
+    int dateHaveTime;
+
+    time_t dateTimezone;
+    int dateDSTmode;
+    int dateHaveZone;
+
+    time_t dateRelMonth;
+    time_t dateRelDay;
+    time_t dateRelSeconds;
+    int dateHaveRel;
+
+    time_t dateMonthOrdinal;
+    int dateHaveOrdinalMonth;
+
+    time_t dateDayOrdinal;
+    time_t dateDayNumber;
+    int dateHaveDay;
+
+    const char *dateStart;
+    const char *dateInput;
+    time_t *dateRelPointer;
+
+    int dateDigitCount;
+} DateInfo;
+
+#define YYMALLOC	ckalloc
+#define YYFREE(x)	(ckfree((void*) (x)))
+
+#define yyDSTmode	(info->dateDSTmode)
+#define yyDayOrdinal	(info->dateDayOrdinal)
+#define yyDayNumber	(info->dateDayNumber)
+#define yyMonthOrdinal	(info->dateMonthOrdinal)
+#define yyHaveDate	(info->dateHaveDate)
+#define yyHaveDay	(info->dateHaveDay)
+#define yyHaveOrdinalMonth (info->dateHaveOrdinalMonth)
+#define yyHaveRel	(info->dateHaveRel)
+#define yyHaveTime	(info->dateHaveTime)
+#define yyHaveZone	(info->dateHaveZone)
+#define yyTimezone	(info->dateTimezone)
+#define yyDay		(info->dateDay)
+#define yyMonth		(info->dateMonth)
+#define yyYear		(info->dateYear)
+#define yyHour		(info->dateHour)
+#define yyMinutes	(info->dateMinutes)
+#define yySeconds	(info->dateSeconds)
+#define yyMeridian	(info->dateMeridian)
+#define yyRelMonth	(info->dateRelMonth)
+#define yyRelDay	(info->dateRelDay)
+#define yyRelSeconds	(info->dateRelSeconds)
+#define yyRelPointer	(info->dateRelPointer)
+#define yyInput		(info->dateInput)
+#define yyDigitCount	(info->dateDigitCount)
+
+#define EPOCH		1970
+#define START_OF_TIME	1902
+#define END_OF_TIME	2037
+
+/*
+ * The offset of tm_year of struct tm returned by localtime, gmtime, etc.
+ * Posix requires 1900.
+ */
+
+#define TM_YEAR_BASE	1900
+
+#define HOUR(x)		((int) (60 * x))
+#define SECSPERDAY	(24L * 60L * 60L)
+#define IsLeapYear(x)	((x % 4 == 0) && (x % 100 != 0 || x % 400 == 0))
+
+/*
+ * An entry in the lexical lookup table.
+ */
+
+typedef struct _TABLE {
+    const char *name;
+    int type;
+    time_t value;
+} TABLE;
+
+/*
+ * Daylight-savings mode: on, off, or not yet known.
+ */
+
+typedef enum _DSTMODE {
+    DSTon, DSToff, DSTmaybe
+} DSTMODE;
+
+/*
+ * Meridian: am, pm, or 24-hour style.
+ */
+
+typedef enum _MERIDIAN {
+    MERam, MERpm, MER24
+} MERIDIAN;
+
+%}
+
+%union {
+    time_t Number;
+    enum _MERIDIAN Meridian;
+}
+
+%{
+
+/*
+ * Prototypes of internal functions.
+ */
+
+static int		LookupWord(YYSTYPE* yylvalPtr, char *buff);
+ static void		TclDateerror(YYLTYPE* location,
+				     DateInfo* info, const char *s);
+ static int		TclDatelex(YYSTYPE* yylvalPtr, YYLTYPE* location,
+				   DateInfo* info);
+static time_t		ToSeconds(time_t Hours, time_t Minutes,
+			    time_t Seconds, MERIDIAN Meridian);
+MODULE_SCOPE int	yyparse(DateInfo*);
+
+%}
+
+%token	tAGO
+%token	tDAY
+%token	tDAYZONE
+%token	tID
+%token	tMERIDIAN
+%token	tMONTH
+%token	tMONTH_UNIT
+%token	tSTARDATE
+%token	tSEC_UNIT
+%token	tSNUMBER
+%token	tUNUMBER
+%token	tZONE
+%token	tEPOCH
+%token	tDST
+%token	tISOBASE
+%token	tDAY_UNIT
+%token	tNEXT
+
+%type	<Number>	tDAY
+%type	<Number>	tDAYZONE
+%type	<Number>	tMONTH
+%type	<Number>	tMONTH_UNIT
+%type	<Number>	tDST
+%type	<Number>	tSEC_UNIT
+%type	<Number>	tSNUMBER
+%type	<Number>	tUNUMBER
+%type	<Number>	tZONE
+%type	<Number>	tISOBASE
+%type	<Number>	tDAY_UNIT
+%type	<Number>	unit
+%type	<Number>	sign
+%type	<Number>	tNEXT
+%type	<Number>	tSTARDATE
+%type	<Meridian>	tMERIDIAN
+%type	<Meridian>	o_merid
+
+%%
+
+spec	: /* NULL */
+	| spec item
+	;
+
+item	: time {
+	    yyHaveTime++;
+	}
+	| zone {
+	    yyHaveZone++;
+	}
+	| date {
+	    yyHaveDate++;
+	}
+	| ordMonth {
+	    yyHaveOrdinalMonth++;
+	}
+	| day {
+	    yyHaveDay++;
+	}
+	| relspec {
+	    yyHaveRel++;
+	}
+	| iso {
+	    yyHaveTime++;
+	    yyHaveDate++;
+	}
+	| trek {
+	    yyHaveTime++;
+	    yyHaveDate++;
+	    yyHaveRel++;
+	}
+	| number
+	;
+
+time	: tUNUMBER tMERIDIAN {
+	    yyHour = $1;
+	    yyMinutes = 0;
+	    yySeconds = 0;
+	    yyMeridian = $2;
+	}
+	| tUNUMBER ':' tUNUMBER o_merid {
+	    yyHour = $1;
+	    yyMinutes = $3;
+	    yySeconds = 0;
+	    yyMeridian = $4;
+	}
+	| tUNUMBER ':' tUNUMBER '-' tUNUMBER {
+	    yyHour = $1;
+	    yyMinutes = $3;
+	    yyMeridian = MER24;
+	    yyDSTmode = DSToff;
+	    yyTimezone = ($5 % 100 + ($5 / 100) * 60);
+	    ++yyHaveZone;
+	}
+	| tUNUMBER ':' tUNUMBER ':' tUNUMBER o_merid {
+	    yyHour = $1;
+	    yyMinutes = $3;
+	    yySeconds = $5;
+	    yyMeridian = $6;
+	}
+	| tUNUMBER ':' tUNUMBER ':' tUNUMBER '-' tUNUMBER {
+	    yyHour = $1;
+	    yyMinutes = $3;
+	    yySeconds = $5;
+	    yyMeridian = MER24;
+	    yyDSTmode = DSToff;
+	    yyTimezone = ($7 % 100 + ($7 / 100) * 60);
+	    ++yyHaveZone;
+	}
+	;
+
+zone	: tZONE tDST {
+	    yyTimezone = $1;
+	    yyDSTmode = DSTon;
+	}
+	| tZONE {
+	    yyTimezone = $1;
+	    yyDSTmode = DSToff;
+	}
+	| tDAYZONE {
+	    yyTimezone = $1;
+	    yyDSTmode = DSTon;
+	}
+	;
+
+day	: tDAY {
+	    yyDayOrdinal = 1;
+	    yyDayNumber = $1;
+	}
+	| tDAY ',' {
+	    yyDayOrdinal = 1;
+	    yyDayNumber = $1;
+	}
+	| tUNUMBER tDAY {
+	    yyDayOrdinal = $1;
+	    yyDayNumber = $2;
+	}
+	| sign tUNUMBER tDAY {
+	    yyDayOrdinal = $1 * $2;
+	    yyDayNumber = $3;
+	}
+	| tNEXT tDAY {
+	    yyDayOrdinal = 2;
+	    yyDayNumber = $2;
+	}
+	;
+
+date	: tUNUMBER '/' tUNUMBER {
+	    yyMonth = $1;
+	    yyDay = $3;
+	}
+	| tUNUMBER '/' tUNUMBER '/' tUNUMBER {
+	    yyMonth = $1;
+	    yyDay = $3;
+	    yyYear = $5;
+	}
+	| tISOBASE {
+	    yyYear = $1 / 10000;
+	    yyMonth = ($1 % 10000)/100;
+	    yyDay = $1 % 100;
+	}
+	| tUNUMBER '-' tMONTH '-' tUNUMBER {
+	    yyDay = $1;
+	    yyMonth = $3;
+	    yyYear = $5;
+	}
+	| tUNUMBER '-' tUNUMBER '-' tUNUMBER {
+	    yyMonth = $3;
+	    yyDay = $5;
+	    yyYear = $1;
+	}
+	| tMONTH tUNUMBER {
+	    yyMonth = $1;
+	    yyDay = $2;
+	}
+	| tMONTH tUNUMBER ',' tUNUMBER {
+	    yyMonth = $1;
+	    yyDay = $2;
+	    yyYear = $4;
+	}
+	| tUNUMBER tMONTH {
+	    yyMonth = $2;
+	    yyDay = $1;
+	}
+	| tEPOCH {
+	    yyMonth = 1;
+	    yyDay = 1;
+	    yyYear = EPOCH;
+	}
+	| tUNUMBER tMONTH tUNUMBER {
+	    yyMonth = $2;
+	    yyDay = $1;
+	    yyYear = $3;
+	}
+	;
+
+ordMonth: tNEXT tMONTH {
+	    yyMonthOrdinal = 1;
+	    yyMonth = $2;
+	}
+	| tNEXT tUNUMBER tMONTH {
+	    yyMonthOrdinal = $2;
+	    yyMonth = $3;
+	}
+	;
+
+iso	: tISOBASE tZONE tISOBASE {
+	    if ($2 != HOUR( 7)) YYABORT;
+	    yyYear = $1 / 10000;
+	    yyMonth = ($1 % 10000)/100;
+	    yyDay = $1 % 100;
+	    yyHour = $3 / 10000;
+	    yyMinutes = ($3 % 10000)/100;
+	    yySeconds = $3 % 100;
+	}
+	| tISOBASE tZONE tUNUMBER ':' tUNUMBER ':' tUNUMBER {
+	    if ($2 != HOUR( 7)) YYABORT;
+	    yyYear = $1 / 10000;
+	    yyMonth = ($1 % 10000)/100;
+	    yyDay = $1 % 100;
+	    yyHour = $3;
+	    yyMinutes = $5;
+	    yySeconds = $7;
+	}
+	| tISOBASE tISOBASE {
+	    yyYear = $1 / 10000;
+	    yyMonth = ($1 % 10000)/100;
+	    yyDay = $1 % 100;
+	    yyHour = $2 / 10000;
+	    yyMinutes = ($2 % 10000)/100;
+	    yySeconds = $2 % 100;
+	}
+	;
+
+trek	: tSTARDATE tUNUMBER '.' tUNUMBER {
+	    /*
+	     * Offset computed year by -377 so that the returned years will be
+	     * in a range accessible with a 32 bit clock seconds value.
+	     */
+
+	    yyYear = $2/1000 + 2323 - 377;
+	    yyDay  = 1;
+	    yyMonth = 1;
+	    yyRelDay += (($2%1000)*(365 + IsLeapYear(yyYear)))/1000;
+	    yyRelSeconds += $4 * 144 * 60;
+	}
+	;
+
+relspec : relunits tAGO {
+	    yyRelSeconds *= -1;
+	    yyRelMonth *= -1;
+	    yyRelDay *= -1;
+	}
+	| relunits
+	;
+
+relunits : sign tUNUMBER unit {
+	    *yyRelPointer += $1 * $2 * $3;
+	}
+	| tUNUMBER unit {
+	    *yyRelPointer += $1 * $2;
+	}
+	| tNEXT unit {
+	    *yyRelPointer += $2;
+	}
+	| tNEXT tUNUMBER unit {
+	    *yyRelPointer += $2 * $3;
+	}
+	| unit {
+	    *yyRelPointer += $1;
+	}
+	;
+
+sign	: '-' {
+	    $$ = -1;
+	}
+	| '+' {
+	    $$ =  1;
+	}
+	;
+
+unit	: tSEC_UNIT {
+	    $$ = $1;
+	    yyRelPointer = &yyRelSeconds;
+	}
+	| tDAY_UNIT {
+	    $$ = $1;
+	    yyRelPointer = &yyRelDay;
+	}
+	| tMONTH_UNIT {
+	    $$ = $1;
+	    yyRelPointer = &yyRelMonth;
+	}
+	;
+
+number	: tUNUMBER {
+	    if (yyHaveTime && yyHaveDate && !yyHaveRel) {
+		yyYear = $1;
+	    } else {
+		yyHaveTime++;
+		if (yyDigitCount <= 2) {
+		    yyHour = $1;
+		    yyMinutes = 0;
+		} else {
+		    yyHour = $1 / 100;
+		    yyMinutes = $1 % 100;
+		}
+		yySeconds = 0;
+		yyMeridian = MER24;
+	    }
+	}
+	;
+
+o_merid : /* NULL */ {
+	    $$ = MER24;
+	}
+	| tMERIDIAN {
+	    $$ = $1;
+	}
+	;
+
+%%
+/*
+ * Month and day table.
+ */
+
+static const TABLE MonthDayTable[] = {
+    { "january",	tMONTH,	 1 },
+    { "february",	tMONTH,	 2 },
+    { "march",		tMONTH,	 3 },
+    { "april",		tMONTH,	 4 },
+    { "may",		tMONTH,	 5 },
+    { "june",		tMONTH,	 6 },
+    { "july",		tMONTH,	 7 },
+    { "august",		tMONTH,	 8 },
+    { "september",	tMONTH,	 9 },
+    { "sept",		tMONTH,	 9 },
+    { "october",	tMONTH, 10 },
+    { "november",	tMONTH, 11 },
+    { "december",	tMONTH, 12 },
+    { "sunday",		tDAY, 0 },
+    { "monday",		tDAY, 1 },
+    { "tuesday",	tDAY, 2 },
+    { "tues",		tDAY, 2 },
+    { "wednesday",	tDAY, 3 },
+    { "wednes",		tDAY, 3 },
+    { "thursday",	tDAY, 4 },
+    { "thur",		tDAY, 4 },
+    { "thurs",		tDAY, 4 },
+    { "friday",		tDAY, 5 },
+    { "saturday",	tDAY, 6 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Time units table.
+ */
+
+static const TABLE UnitsTable[] = {
+    { "year",		tMONTH_UNIT,	12 },
+    { "month",		tMONTH_UNIT,	 1 },
+    { "fortnight",	tDAY_UNIT,	14 },
+    { "week",		tDAY_UNIT,	 7 },
+    { "day",		tDAY_UNIT,	 1 },
+    { "hour",		tSEC_UNIT, 60 * 60 },
+    { "minute",		tSEC_UNIT,	60 },
+    { "min",		tSEC_UNIT,	60 },
+    { "second",		tSEC_UNIT,	 1 },
+    { "sec",		tSEC_UNIT,	 1 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Assorted relative-time words.
+ */
+
+static const TABLE OtherTable[] = {
+    { "tomorrow",	tDAY_UNIT,	1 },
+    { "yesterday",	tDAY_UNIT,	-1 },
+    { "today",		tDAY_UNIT,	0 },
+    { "now",		tSEC_UNIT,	0 },
+    { "last",		tUNUMBER,	-1 },
+    { "this",		tSEC_UNIT,	0 },
+    { "next",		tNEXT,		1 },
+#if 0
+    { "first",		tUNUMBER,	1 },
+    { "second",		tUNUMBER,	2 },
+    { "third",		tUNUMBER,	3 },
+    { "fourth",		tUNUMBER,	4 },
+    { "fifth",		tUNUMBER,	5 },
+    { "sixth",		tUNUMBER,	6 },
+    { "seventh",	tUNUMBER,	7 },
+    { "eighth",		tUNUMBER,	8 },
+    { "ninth",		tUNUMBER,	9 },
+    { "tenth",		tUNUMBER,	10 },
+    { "eleventh",	tUNUMBER,	11 },
+    { "twelfth",	tUNUMBER,	12 },
+#endif
+    { "ago",		tAGO,		1 },
+    { "epoch",		tEPOCH,		0 },
+    { "stardate",	tSTARDATE,	0 },
+    { NULL, 0, 0 }
+};
+
+/*
+ * The timezone table. (Note: This table was modified to not use any floating
+ * point constants to work around an SGI compiler bug).
+ */
+
+static const TABLE TimezoneTable[] = {
+    { "gmt",	tZONE,	   HOUR( 0) },	    /* Greenwich Mean */
+    { "ut",	tZONE,	   HOUR( 0) },	    /* Universal (Coordinated) */
+    { "utc",	tZONE,	   HOUR( 0) },
+    { "uct",	tZONE,	   HOUR( 0) },	    /* Universal Coordinated Time */
+    { "wet",	tZONE,	   HOUR( 0) },	    /* Western European */
+    { "bst",	tDAYZONE,  HOUR( 0) },	    /* British Summer */
+    { "wat",	tZONE,	   HOUR( 1) },	    /* West Africa */
+    { "at",	tZONE,	   HOUR( 2) },	    /* Azores */
+#if	0
+    /* For completeness.  BST is also British Summer, and GST is
+     * also Guam Standard. */
+    { "bst",	tZONE,	   HOUR( 3) },	    /* Brazil Standard */
+    { "gst",	tZONE,	   HOUR( 3) },	    /* Greenland Standard */
+#endif
+    { "nft",	tZONE,	   HOUR( 7/2) },    /* Newfoundland */
+    { "nst",	tZONE,	   HOUR( 7/2) },    /* Newfoundland Standard */
+    { "ndt",	tDAYZONE,  HOUR( 7/2) },    /* Newfoundland Daylight */
+    { "ast",	tZONE,	   HOUR( 4) },	    /* Atlantic Standard */
+    { "adt",	tDAYZONE,  HOUR( 4) },	    /* Atlantic Daylight */
+    { "est",	tZONE,	   HOUR( 5) },	    /* Eastern Standard */
+    { "edt",	tDAYZONE,  HOUR( 5) },	    /* Eastern Daylight */
+    { "cst",	tZONE,	   HOUR( 6) },	    /* Central Standard */
+    { "cdt",	tDAYZONE,  HOUR( 6) },	    /* Central Daylight */
+    { "mst",	tZONE,	   HOUR( 7) },	    /* Mountain Standard */
+    { "mdt",	tDAYZONE,  HOUR( 7) },	    /* Mountain Daylight */
+    { "pst",	tZONE,	   HOUR( 8) },	    /* Pacific Standard */
+    { "pdt",	tDAYZONE,  HOUR( 8) },	    /* Pacific Daylight */
+    { "yst",	tZONE,	   HOUR( 9) },	    /* Yukon Standard */
+    { "ydt",	tDAYZONE,  HOUR( 9) },	    /* Yukon Daylight */
+    { "hst",	tZONE,	   HOUR(10) },	    /* Hawaii Standard */
+    { "hdt",	tDAYZONE,  HOUR(10) },	    /* Hawaii Daylight */
+    { "cat",	tZONE,	   HOUR(10) },	    /* Central Alaska */
+    { "ahst",	tZONE,	   HOUR(10) },	    /* Alaska-Hawaii Standard */
+    { "nt",	tZONE,	   HOUR(11) },	    /* Nome */
+    { "idlw",	tZONE,	   HOUR(12) },	    /* International Date Line West */
+    { "cet",	tZONE,	  -HOUR( 1) },	    /* Central European */
+    { "cest",	tDAYZONE, -HOUR( 1) },	    /* Central European Summer */
+    { "met",	tZONE,	  -HOUR( 1) },	    /* Middle European */
+    { "mewt",	tZONE,	  -HOUR( 1) },	    /* Middle European Winter */
+    { "mest",	tDAYZONE, -HOUR( 1) },	    /* Middle European Summer */
+    { "swt",	tZONE,	  -HOUR( 1) },	    /* Swedish Winter */
+    { "sst",	tDAYZONE, -HOUR( 1) },	    /* Swedish Summer */
+    { "fwt",	tZONE,	  -HOUR( 1) },	    /* French Winter */
+    { "fst",	tDAYZONE, -HOUR( 1) },	    /* French Summer */
+    { "eet",	tZONE,	  -HOUR( 2) },	    /* Eastern Europe, USSR Zone 1 */
+    { "bt",	tZONE,	  -HOUR( 3) },	    /* Baghdad, USSR Zone 2 */
+    { "it",	tZONE,	  -HOUR( 7/2) },    /* Iran */
+    { "zp4",	tZONE,	  -HOUR( 4) },	    /* USSR Zone 3 */
+    { "zp5",	tZONE,	  -HOUR( 5) },	    /* USSR Zone 4 */
+    { "ist",	tZONE,	  -HOUR(11/2) },    /* Indian Standard */
+    { "zp6",	tZONE,	  -HOUR( 6) },	    /* USSR Zone 5 */
+#if	0
+    /* For completeness.  NST is also Newfoundland Stanard, nad SST is
+     * also Swedish Summer. */
+    { "nst",	tZONE,	  -HOUR(13/2) },    /* North Sumatra */
+    { "sst",	tZONE,	  -HOUR( 7) },	    /* South Sumatra, USSR Zone 6 */
+#endif	/* 0 */
+    { "wast",	tZONE,	  -HOUR( 7) },	    /* West Australian Standard */
+    { "wadt",	tDAYZONE, -HOUR( 7) },	    /* West Australian Daylight */
+    { "jt",	tZONE,	  -HOUR(15/2) },    /* Java (3pm in Cronusland!) */
+    { "cct",	tZONE,	  -HOUR( 8) },	    /* China Coast, USSR Zone 7 */
+    { "jst",	tZONE,	  -HOUR( 9) },	    /* Japan Standard, USSR Zone 8 */
+    { "jdt",	tDAYZONE, -HOUR( 9) },	    /* Japan Daylight */
+    { "kst",	tZONE,	  -HOUR( 9) },	    /* Korea Standard */
+    { "kdt",	tDAYZONE, -HOUR( 9) },	    /* Korea Daylight */
+    { "cast",	tZONE,	  -HOUR(19/2) },    /* Central Australian Standard */
+    { "cadt",	tDAYZONE, -HOUR(19/2) },    /* Central Australian Daylight */
+    { "east",	tZONE,	  -HOUR(10) },	    /* Eastern Australian Standard */
+    { "eadt",	tDAYZONE, -HOUR(10) },	    /* Eastern Australian Daylight */
+    { "gst",	tZONE,	  -HOUR(10) },	    /* Guam Standard, USSR Zone 9 */
+    { "nzt",	tZONE,	  -HOUR(12) },	    /* New Zealand */
+    { "nzst",	tZONE,	  -HOUR(12) },	    /* New Zealand Standard */
+    { "nzdt",	tDAYZONE, -HOUR(12) },	    /* New Zealand Daylight */
+    { "idle",	tZONE,	  -HOUR(12) },	    /* International Date Line East */
+    /* ADDED BY Marco Nijdam */
+    { "dst",	tDST,	  HOUR( 0) },	    /* DST on (hour is ignored) */
+    /* End ADDED */
+    { NULL, 0, 0 }
+};
+
+/*
+ * Military timezone table.
+ */
+
+static const TABLE MilitaryTable[] = {
+    { "a",	tZONE,	-HOUR( 1) },
+    { "b",	tZONE,	-HOUR( 2) },
+    { "c",	tZONE,	-HOUR( 3) },
+    { "d",	tZONE,	-HOUR( 4) },
+    { "e",	tZONE,	-HOUR( 5) },
+    { "f",	tZONE,	-HOUR( 6) },
+    { "g",	tZONE,	-HOUR( 7) },
+    { "h",	tZONE,	-HOUR( 8) },
+    { "i",	tZONE,	-HOUR( 9) },
+    { "k",	tZONE,	-HOUR(10) },
+    { "l",	tZONE,	-HOUR(11) },
+    { "m",	tZONE,	-HOUR(12) },
+    { "n",	tZONE,	HOUR(  1) },
+    { "o",	tZONE,	HOUR(  2) },
+    { "p",	tZONE,	HOUR(  3) },
+    { "q",	tZONE,	HOUR(  4) },
+    { "r",	tZONE,	HOUR(  5) },
+    { "s",	tZONE,	HOUR(  6) },
+    { "t",	tZONE,	HOUR(  7) },
+    { "u",	tZONE,	HOUR(  8) },
+    { "v",	tZONE,	HOUR(  9) },
+    { "w",	tZONE,	HOUR( 10) },
+    { "x",	tZONE,	HOUR( 11) },
+    { "y",	tZONE,	HOUR( 12) },
+    { "z",	tZONE,	HOUR( 0) },
+    { NULL, 0, 0 }
+};
+
+/*
+ * Dump error messages in the bit bucket.
+ */
+
+static void
+TclDateerror(
+    YYLTYPE* location,
+    DateInfo* infoPtr,
+    const char *s)
+{
+    Tcl_Obj* t;
+    Tcl_AppendToObj(infoPtr->messages, infoPtr->separatrix, -1);
+    Tcl_AppendToObj(infoPtr->messages, s, -1);
+    Tcl_AppendToObj(infoPtr->messages, " (characters ", -1);
+    t = Tcl_NewIntObj(location->first_column);
+    Tcl_IncrRefCount(t);
+    Tcl_AppendObjToObj(infoPtr->messages, t);
+    Tcl_DecrRefCount(t);
+    Tcl_AppendToObj(infoPtr->messages, "-", -1);
+    t = Tcl_NewIntObj(location->last_column);
+    Tcl_IncrRefCount(t);
+    Tcl_AppendObjToObj(infoPtr->messages, t);
+    Tcl_DecrRefCount(t);
+    Tcl_AppendToObj(infoPtr->messages, ")", -1);
+    infoPtr->separatrix = "\n";
+}
+
+static time_t
+ToSeconds(
+    time_t Hours,
+    time_t Minutes,
+    time_t Seconds,
+    MERIDIAN Meridian)
+{
+    if (Minutes < 0 || Minutes > 59 || Seconds < 0 || Seconds > 59) {
+	return -1;
+    }
+    switch (Meridian) {
+    case MER24:
+	if (Hours < 0 || Hours > 23) {
+	    return -1;
+	}
+	return (Hours * 60L + Minutes) * 60L + Seconds;
+    case MERam:
+	if (Hours < 1 || Hours > 12) {
+	    return -1;
+	}
+	return ((Hours % 12) * 60L + Minutes) * 60L + Seconds;
+    case MERpm:
+	if (Hours < 1 || Hours > 12) {
+	    return -1;
+	}
+	return (((Hours % 12) + 12) * 60L + Minutes) * 60L + Seconds;
+    }
+    return -1;			/* Should never be reached */
+}
+
+static int
+LookupWord(
+    YYSTYPE* yylvalPtr,
+    char *buff)
+{
+    register char *p;
+    register char *q;
+    register const TABLE *tp;
+    int i, abbrev;
+
+    /*
+     * Make it lowercase.
+     */
+
+    Tcl_UtfToLower(buff);
+
+    if (strcmp(buff, "am") == 0 || strcmp(buff, "a.m.") == 0) {
+	yylvalPtr->Meridian = MERam;
+	return tMERIDIAN;
+    }
+    if (strcmp(buff, "pm") == 0 || strcmp(buff, "p.m.") == 0) {
+	yylvalPtr->Meridian = MERpm;
+	return tMERIDIAN;
+    }
+
+    /*
+     * See if we have an abbreviation for a month.
+     */
+
+    if (strlen(buff) == 3) {
+	abbrev = 1;
+    } else if (strlen(buff) == 4 && buff[3] == '.') {
+	abbrev = 1;
+	buff[3] = '\0';
+    } else {
+	abbrev = 0;
+    }
+
+    for (tp = MonthDayTable; tp->name; tp++) {
+	if (abbrev) {
+	    if (strncmp(buff, tp->name, 3) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	} else if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    for (tp = TimezoneTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    for (tp = UnitsTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    /*
+     * Strip off any plural and try the units table again.
+     */
+
+    i = strlen(buff) - 1;
+    if (i > 0 && buff[i] == 's') {
+	buff[i] = '\0';
+	for (tp = UnitsTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    for (tp = OtherTable; tp->name; tp++) {
+	if (strcmp(buff, tp->name) == 0) {
+	    yylvalPtr->Number = tp->value;
+	    return tp->type;
+	}
+    }
+
+    /*
+     * Military timezones.
+     */
+
+    if (buff[1] == '\0' && !(*buff & 0x80)
+	    && isalpha(UCHAR(*buff))) {			/* INTL: ISO only */
+	for (tp = MilitaryTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    /*
+     * Drop out any periods and try the timezone table again.
+     */
+
+    for (i = 0, p = q = buff; *q; q++) {
+	if (*q != '.') {
+	    *p++ = *q;
+	} else {
+	    i++;
+	}
+    }
+    *p = '\0';
+    if (i) {
+	for (tp = TimezoneTable; tp->name; tp++) {
+	    if (strcmp(buff, tp->name) == 0) {
+		yylvalPtr->Number = tp->value;
+		return tp->type;
+	    }
+	}
+    }
+
+    return tID;
+}
+
+static int
+TclDatelex(
+    YYSTYPE* yylvalPtr,
+    YYLTYPE* location,
+    DateInfo *info)
+{
+    register char c;
+    register char *p;
+    char buff[20];
+    int Count;
+
+    location->first_column = yyInput - info->dateStart;
+    for ( ; ; ) {
+	while (isspace(UCHAR(*yyInput))) {
+	    yyInput++;
+	}
+
+	if (isdigit(UCHAR(c = *yyInput))) { /* INTL: digit */
+	    /*
+	     * Convert the string into a number; count the number of digits.
+	     */
+
+	    Count = 0;
+	    for (yylvalPtr->Number = 0;
+		    isdigit(UCHAR(c = *yyInput++)); ) {	  /* INTL: digit */
+		yylvalPtr->Number = 10 * yylvalPtr->Number + c - '0';
+		Count++;
+	    }
+	    yyInput--;
+	    yyDigitCount = Count;
+
+	    /*
+	     * A number with 6 or more digits is considered an ISO 8601 base.
+	     */
+
+	    if (Count >= 6) {
+		location->last_column = yyInput - info->dateStart - 1;
+		return tISOBASE;
+	    } else {
+		location->last_column = yyInput - info->dateStart - 1;
+		return tUNUMBER;
+	    }
+	}
+	if (!(c & 0x80) && isalpha(UCHAR(c))) {		  /* INTL: ISO only. */
+	    for (p = buff; isalpha(UCHAR(c = *yyInput++)) /* INTL: ISO only. */
+		     || c == '.'; ) {
+		if (p < &buff[sizeof buff - 1]) {
+		    *p++ = c;
+		}
+	    }
+	    *p = '\0';
+	    yyInput--;
+	    location->last_column = yyInput - info->dateStart - 1;
+	    return LookupWord(yylvalPtr, buff);
+	}
+	if (c != '(') {
+	    location->last_column = yyInput - info->dateStart;
+	    return *yyInput++;
+	}
+	Count = 0;
+	do {
+	    c = *yyInput++;
+	    if (c == '\0') {
+		location->last_column = yyInput - info->dateStart - 1;
+		return c;
+	    } else if (c == '(') {
+		Count++;
+	    } else if (c == ')') {
+		Count--;
+	    }
+	} while (Count > 0);
+    }
+}
+
+int
+TclClockOldscanObjCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Count of paraneters */
+    Tcl_Obj *const *objv)	/* Parameters */
+{
+    Tcl_Obj *result, *resultElement;
+    int yr, mo, da;
+    DateInfo dateInfo;
+    DateInfo* info = &dateInfo;
+    int status;
+
+    if (objc != 5) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"stringToParse baseYear baseMonth baseDay" );
+	return TCL_ERROR;
+    }
+
+    yyInput = Tcl_GetString( objv[1] );
+    dateInfo.dateStart = yyInput;
+
+    yyHaveDate = 0;
+    if (Tcl_GetIntFromObj(interp, objv[2], &yr) != TCL_OK
+	    || Tcl_GetIntFromObj(interp, objv[3], &mo) != TCL_OK
+	    || Tcl_GetIntFromObj(interp, objv[4], &da) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    yyYear = yr; yyMonth = mo; yyDay = da;
+
+    yyHaveTime = 0;
+    yyHour = 0; yyMinutes = 0; yySeconds = 0; yyMeridian = MER24;
+
+    yyHaveZone = 0;
+    yyTimezone = 0; yyDSTmode = DSTmaybe;
+
+    yyHaveOrdinalMonth = 0;
+    yyMonthOrdinal = 0;
+
+    yyHaveDay = 0;
+    yyDayOrdinal = 0; yyDayNumber = 0;
+
+    yyHaveRel = 0;
+    yyRelMonth = 0; yyRelDay = 0; yyRelSeconds = 0; yyRelPointer = NULL;
+
+    dateInfo.messages = Tcl_NewObj();
+    dateInfo.separatrix = "";
+    Tcl_IncrRefCount(dateInfo.messages);
+
+    status = yyparse(&dateInfo);
+    if (status == 1) {
+	Tcl_SetObjResult(interp, dateInfo.messages);
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "PARSE", NULL);
+	return TCL_ERROR;
+    } else if (status == 2) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("memory exhausted", -1));
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	return TCL_ERROR;
+    } else if (status != 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("Unknown status returned "
+						  "from date parser. Please "
+						  "report this error as a "
+						  "bug in Tcl.", -1));
+	Tcl_DecrRefCount(dateInfo.messages);
+	Tcl_SetErrorCode(interp, "TCL", "BUG", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_DecrRefCount(dateInfo.messages);
+
+    if (yyHaveDate > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one date in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveTime > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one time of day in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveZone > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one time zone in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveDay > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one weekday in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+    if (yyHaveOrdinalMonth > 1) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("more than one ordinal month in string", -1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "DATE", "MULTIPLE", NULL);
+	return TCL_ERROR;
+    }
+
+    result = Tcl_NewObj();
+    resultElement = Tcl_NewObj();
+    if (yyHaveDate) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyYear));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonth));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDay));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    if (yyHaveTime) {
+	Tcl_ListObjAppendElement(interp, result, Tcl_NewIntObj((int)
+		ToSeconds(yyHour, yyMinutes, yySeconds, yyMeridian)));
+    } else {
+	Tcl_ListObjAppendElement(interp, result, Tcl_NewObj());
+    }
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveZone) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) -yyTimezone));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj(1 - yyDSTmode));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveRel) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelMonth));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelDay));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyRelSeconds));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveDay && !yyHaveDate) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDayOrdinal));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyDayNumber));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    resultElement = Tcl_NewObj();
+    if (yyHaveOrdinalMonth) {
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonthOrdinal));
+	Tcl_ListObjAppendElement(interp, resultElement,
+		Tcl_NewIntObj((int) yyMonth));
+    }
+    Tcl_ListObjAppendElement(interp, result, resultElement);
+
+    Tcl_SetObjResult(interp, result);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclHash.c b/generic/tclHash.c
new file mode 100644
index 0000000..78ad514
--- /dev/null
+++ b/generic/tclHash.c
@@ -0,0 +1,1079 @@
+/*
+ * tclHash.c --
+ *
+ *	Implementation of in-memory hash tables for Tcl and Tcl-based
+ *	applications.
+ *
+ * Copyright (c) 1991-1993 The Regents of the University of California.
+ * Copyright (c) 1994 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Prevent macros from clashing with function definitions.
+ */
+
+#undef Tcl_FindHashEntry
+#undef Tcl_CreateHashEntry
+
+/*
+ * When there are this many entries per bucket, on average, rebuild the hash
+ * table to make it larger.
+ */
+
+#define REBUILD_MULTIPLIER	3
+
+/*
+ * The following macro takes a preliminary integer hash value and produces an
+ * index into a hash tables bucket list. The idea is to make it so that
+ * preliminary values that are arbitrarily similar will end up in different
+ * buckets. The hash function was taken from a random-number generator.
+ */
+
+#define RANDOM_INDEX(tablePtr, i) \
+    ((((i)*1103515245L) >> (tablePtr)->downShift) & (tablePtr)->mask)
+
+/*
+ * Prototypes for the array hash key methods.
+ */
+
+static Tcl_HashEntry *	AllocArrayEntry(Tcl_HashTable *tablePtr, void *keyPtr);
+static int		CompareArrayKeys(void *keyPtr, Tcl_HashEntry *hPtr);
+static TCL_HASH_TYPE	HashArrayKey(Tcl_HashTable *tablePtr, void *keyPtr);
+
+/*
+ * Prototypes for the one word hash key methods. Not actually declared because
+ * this is a critical path that is implemented in the core hash table access
+ * function.
+ */
+
+#if 0
+static Tcl_HashEntry *	AllocOneWordEntry(Tcl_HashTable *tablePtr,
+			    void *keyPtr);
+static int		CompareOneWordKeys(void *keyPtr, Tcl_HashEntry *hPtr);
+static unsigned int	HashOneWordKey(Tcl_HashTable *tablePtr, void *keyPtr);
+#endif
+
+/*
+ * Prototypes for the string hash key methods.
+ */
+
+static Tcl_HashEntry *	AllocStringEntry(Tcl_HashTable *tablePtr,
+			    void *keyPtr);
+static int		CompareStringKeys(void *keyPtr, Tcl_HashEntry *hPtr);
+static TCL_HASH_TYPE	HashStringKey(Tcl_HashTable *tablePtr, void *keyPtr);
+
+/*
+ * Function prototypes for static functions in this file:
+ */
+
+static Tcl_HashEntry *	BogusFind(Tcl_HashTable *tablePtr, const char *key);
+static Tcl_HashEntry *	BogusCreate(Tcl_HashTable *tablePtr, const char *key,
+			    int *newPtr);
+static Tcl_HashEntry *	CreateHashEntry(Tcl_HashTable *tablePtr, const char *key,
+			    int *newPtr);
+static Tcl_HashEntry *	FindHashEntry(Tcl_HashTable *tablePtr, const char *key);
+static void		RebuildTable(Tcl_HashTable *tablePtr);
+
+const Tcl_HashKeyType tclArrayHashKeyType = {
+    TCL_HASH_KEY_TYPE_VERSION,		/* version */
+    TCL_HASH_KEY_RANDOMIZE_HASH,	/* flags */
+    HashArrayKey,			/* hashKeyProc */
+    CompareArrayKeys,			/* compareKeysProc */
+    AllocArrayEntry,			/* allocEntryProc */
+    NULL				/* freeEntryProc */
+};
+
+const Tcl_HashKeyType tclOneWordHashKeyType = {
+    TCL_HASH_KEY_TYPE_VERSION,		/* version */
+    0,					/* flags */
+    NULL, /* HashOneWordKey, */		/* hashProc */
+    NULL, /* CompareOneWordKey, */	/* compareProc */
+    NULL, /* AllocOneWordKey, */	/* allocEntryProc */
+    NULL  /* FreeOneWordKey, */		/* freeEntryProc */
+};
+
+const Tcl_HashKeyType tclStringHashKeyType = {
+    TCL_HASH_KEY_TYPE_VERSION,		/* version */
+    0,					/* flags */
+    HashStringKey,			/* hashKeyProc */
+    CompareStringKeys,			/* compareKeysProc */
+    AllocStringEntry,			/* allocEntryProc */
+    NULL				/* freeEntryProc */
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitHashTable --
+ *
+ *	Given storage for a hash table, set up the fields to prepare the hash
+ *	table for use.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	TablePtr is now ready to be passed to Tcl_FindHashEntry and
+ *	Tcl_CreateHashEntry.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_InitHashTable(
+    register Tcl_HashTable *tablePtr,
+				/* Pointer to table record, which is supplied
+				 * by the caller. */
+    int keyType)		/* Type of keys to use in table:
+				 * TCL_STRING_KEYS, TCL_ONE_WORD_KEYS, or an
+				 * integer >= 2. */
+{
+    /*
+     * Use a special value to inform the extended version that it must not
+     * access any of the new fields in the Tcl_HashTable. If an extension is
+     * rebuilt then any calls to this function will be redirected to the
+     * extended version by a macro.
+     */
+
+    Tcl_InitCustomHashTable(tablePtr, keyType, (const Tcl_HashKeyType *) -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitCustomHashTable --
+ *
+ *	Given storage for a hash table, set up the fields to prepare the hash
+ *	table for use. This is an extended version of Tcl_InitHashTable which
+ *	supports user defined keys.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	TablePtr is now ready to be passed to Tcl_FindHashEntry and
+ *	Tcl_CreateHashEntry.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_InitCustomHashTable(
+    register Tcl_HashTable *tablePtr,
+				/* Pointer to table record, which is supplied
+				 * by the caller. */
+    int keyType,		/* Type of keys to use in table:
+				 * TCL_STRING_KEYS, TCL_ONE_WORD_KEYS,
+				 * TCL_CUSTOM_TYPE_KEYS, TCL_CUSTOM_PTR_KEYS,
+				 * or an integer >= 2. */
+    const Tcl_HashKeyType *typePtr) /* Pointer to structure which defines the
+				 * behaviour of this table. */
+{
+#if (TCL_SMALL_HASH_TABLE != 4)
+    Tcl_Panic("Tcl_InitCustomHashTable: TCL_SMALL_HASH_TABLE is %d, not 4",
+	    TCL_SMALL_HASH_TABLE);
+#endif
+
+    tablePtr->buckets = tablePtr->staticBuckets;
+    tablePtr->staticBuckets[0] = tablePtr->staticBuckets[1] = 0;
+    tablePtr->staticBuckets[2] = tablePtr->staticBuckets[3] = 0;
+    tablePtr->numBuckets = TCL_SMALL_HASH_TABLE;
+    tablePtr->numEntries = 0;
+    tablePtr->rebuildSize = TCL_SMALL_HASH_TABLE*REBUILD_MULTIPLIER;
+    tablePtr->downShift = 28;
+    tablePtr->mask = 3;
+    tablePtr->keyType = keyType;
+    tablePtr->findProc = FindHashEntry;
+    tablePtr->createProc = CreateHashEntry;
+
+    if (typePtr == NULL) {
+	/*
+	 * The caller has been rebuilt so the hash table is an extended
+	 * version.
+	 */
+    } else if (typePtr != (Tcl_HashKeyType *) -1) {
+	/*
+	 * The caller is requesting a customized hash table so it must be an
+	 * extended version.
+	 */
+
+	tablePtr->typePtr = typePtr;
+    } else {
+	/*
+	 * The caller has not been rebuilt so the hash table is not extended.
+	 */
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindHashEntry --
+ *
+ *	Given a hash table find the entry with a matching key.
+ *
+ * Results:
+ *	The return value is a token for the matching entry in the hash table,
+ *	or NULL if there was no matching entry.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashEntry *
+Tcl_FindHashEntry(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const void *key)		/* Key to use to find matching entry. */
+{
+    return (*((tablePtr)->findProc))(tablePtr, key);
+}
+
+static Tcl_HashEntry *
+FindHashEntry(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const char *key)		/* Key to use to find matching entry. */
+{
+    return CreateHashEntry(tablePtr, key, NULL);
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateHashEntry --
+ *
+ *	Given a hash table with string keys, and a string key, find the entry
+ *	with a matching key. If there is no matching entry, then create a new
+ *	entry that does match.
+ *
+ * Results:
+ *	The return value is a pointer to the matching entry. If this is a
+ *	newly-created entry, then *newPtr will be set to a non-zero value;
+ *	otherwise *newPtr will be set to 0. If this is a new entry the value
+ *	stored in the entry will initially be 0.
+ *
+ * Side effects:
+ *	A new entry may be added to the hash table.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashEntry *
+Tcl_CreateHashEntry(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const void *key,		/* Key to use to find or create matching
+				 * entry. */
+    int *newPtr)		/* Store info here telling whether a new entry
+				 * was created. */
+{
+    return (*((tablePtr)->createProc))(tablePtr, key, newPtr);
+}
+
+static Tcl_HashEntry *
+CreateHashEntry(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const char *key,		/* Key to use to find or create matching
+				 * entry. */
+    int *newPtr)		/* Store info here telling whether a new entry
+				 * was created. */
+{
+    register Tcl_HashEntry *hPtr;
+    const Tcl_HashKeyType *typePtr;
+    unsigned int hash;
+    int index;
+
+    if (tablePtr->keyType == TCL_STRING_KEYS) {
+	typePtr = &tclStringHashKeyType;
+    } else if (tablePtr->keyType == TCL_ONE_WORD_KEYS) {
+	typePtr = &tclOneWordHashKeyType;
+    } else if (tablePtr->keyType == TCL_CUSTOM_TYPE_KEYS
+	    || tablePtr->keyType == TCL_CUSTOM_PTR_KEYS) {
+	typePtr = tablePtr->typePtr;
+    } else {
+	typePtr = &tclArrayHashKeyType;
+    }
+
+    if (typePtr->hashKeyProc) {
+	hash = typePtr->hashKeyProc(tablePtr, (void *) key);
+	if (typePtr->flags & TCL_HASH_KEY_RANDOMIZE_HASH) {
+	    index = RANDOM_INDEX(tablePtr, hash);
+	} else {
+	    index = hash & tablePtr->mask;
+	}
+    } else {
+	hash = PTR2UINT(key);
+	index = RANDOM_INDEX(tablePtr, hash);
+    }
+
+    /*
+     * Search all of the entries in the appropriate bucket.
+     */
+
+    if (typePtr->compareKeysProc) {
+	Tcl_CompareHashKeysProc *compareKeysProc = typePtr->compareKeysProc;
+
+	for (hPtr = tablePtr->buckets[index]; hPtr != NULL;
+		hPtr = hPtr->nextPtr) {
+	    if (hash != PTR2UINT(hPtr->hash)) {
+		continue;
+	    }
+	    if (((void *) key == hPtr) || compareKeysProc((void *) key, hPtr)) {
+		if (newPtr) {
+		    *newPtr = 0;
+		}
+		return hPtr;
+	    }
+	}
+    } else {
+	for (hPtr = tablePtr->buckets[index]; hPtr != NULL;
+		hPtr = hPtr->nextPtr) {
+	    if (hash != PTR2UINT(hPtr->hash)) {
+		continue;
+	    }
+	    if (key == hPtr->key.oneWordValue) {
+		if (newPtr) {
+		    *newPtr = 0;
+		}
+		return hPtr;
+	    }
+	}
+    }
+
+    if (!newPtr) {
+	return NULL;
+    }
+
+    /*
+     * Entry not found. Add a new one to the bucket.
+     */
+
+    *newPtr = 1;
+    if (typePtr->allocEntryProc) {
+	hPtr = typePtr->allocEntryProc(tablePtr, (void *) key);
+    } else {
+	hPtr = ckalloc(sizeof(Tcl_HashEntry));
+	hPtr->key.oneWordValue = (char *) key;
+	hPtr->clientData = 0;
+    }
+
+    hPtr->tablePtr = tablePtr;
+    hPtr->hash = UINT2PTR(hash);
+    hPtr->nextPtr = tablePtr->buckets[index];
+    tablePtr->buckets[index] = hPtr;
+    tablePtr->numEntries++;
+
+    /*
+     * If the table has exceeded a decent size, rebuild it with many more
+     * buckets.
+     */
+
+    if (tablePtr->numEntries >= tablePtr->rebuildSize) {
+	RebuildTable(tablePtr);
+    }
+    return hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteHashEntry --
+ *
+ *	Remove a single entry from a hash table.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The entry given by entryPtr is deleted from its table and should never
+ *	again be used by the caller. It is up to the caller to free the
+ *	clientData field of the entry, if that is relevant.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteHashEntry(
+    Tcl_HashEntry *entryPtr)
+{
+    register Tcl_HashEntry *prevPtr;
+    const Tcl_HashKeyType *typePtr;
+    Tcl_HashTable *tablePtr;
+    Tcl_HashEntry **bucketPtr;
+    int index;
+
+    tablePtr = entryPtr->tablePtr;
+
+    if (tablePtr->keyType == TCL_STRING_KEYS) {
+	typePtr = &tclStringHashKeyType;
+    } else if (tablePtr->keyType == TCL_ONE_WORD_KEYS) {
+	typePtr = &tclOneWordHashKeyType;
+    } else if (tablePtr->keyType == TCL_CUSTOM_TYPE_KEYS
+	    || tablePtr->keyType == TCL_CUSTOM_PTR_KEYS) {
+	typePtr = tablePtr->typePtr;
+    } else {
+	typePtr = &tclArrayHashKeyType;
+    }
+
+    if (typePtr->hashKeyProc == NULL
+	    || typePtr->flags & TCL_HASH_KEY_RANDOMIZE_HASH) {
+	index = RANDOM_INDEX(tablePtr, PTR2INT(entryPtr->hash));
+    } else {
+	index = PTR2UINT(entryPtr->hash) & tablePtr->mask;
+    }
+
+    bucketPtr = &tablePtr->buckets[index];
+
+    if (*bucketPtr == entryPtr) {
+	*bucketPtr = entryPtr->nextPtr;
+    } else {
+	for (prevPtr = *bucketPtr; ; prevPtr = prevPtr->nextPtr) {
+	    if (prevPtr == NULL) {
+		Tcl_Panic("malformed bucket chain in Tcl_DeleteHashEntry");
+	    }
+	    if (prevPtr->nextPtr == entryPtr) {
+		prevPtr->nextPtr = entryPtr->nextPtr;
+		break;
+	    }
+	}
+    }
+
+    tablePtr->numEntries--;
+    if (typePtr->freeEntryProc) {
+	typePtr->freeEntryProc(entryPtr);
+    } else {
+	ckfree(entryPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteHashTable --
+ *
+ *	Free up everything associated with a hash table except for the record
+ *	for the table itself.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The hash table is no longer useable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteHashTable(
+    register Tcl_HashTable *tablePtr)	/* Table to delete. */
+{
+    register Tcl_HashEntry *hPtr, *nextPtr;
+    const Tcl_HashKeyType *typePtr;
+    int i;
+
+    if (tablePtr->keyType == TCL_STRING_KEYS) {
+	typePtr = &tclStringHashKeyType;
+    } else if (tablePtr->keyType == TCL_ONE_WORD_KEYS) {
+	typePtr = &tclOneWordHashKeyType;
+    } else if (tablePtr->keyType == TCL_CUSTOM_TYPE_KEYS
+	    || tablePtr->keyType == TCL_CUSTOM_PTR_KEYS) {
+	typePtr = tablePtr->typePtr;
+    } else {
+	typePtr = &tclArrayHashKeyType;
+    }
+
+    /*
+     * Free up all the entries in the table.
+     */
+
+    for (i = 0; i < tablePtr->numBuckets; i++) {
+	hPtr = tablePtr->buckets[i];
+	while (hPtr != NULL) {
+	    nextPtr = hPtr->nextPtr;
+	    if (typePtr->freeEntryProc) {
+		typePtr->freeEntryProc(hPtr);
+	    } else {
+		ckfree(hPtr);
+	    }
+	    hPtr = nextPtr;
+	}
+    }
+
+    /*
+     * Free up the bucket array, if it was dynamically allocated.
+     */
+
+    if (tablePtr->buckets != tablePtr->staticBuckets) {
+	if (typePtr->flags & TCL_HASH_KEY_SYSTEM_HASH) {
+	    TclpSysFree((char *) tablePtr->buckets);
+	} else {
+	    ckfree(tablePtr->buckets);
+	}
+    }
+
+    /*
+     * Arrange for panics if the table is used again without
+     * re-initialization.
+     */
+
+    tablePtr->findProc = BogusFind;
+    tablePtr->createProc = BogusCreate;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FirstHashEntry --
+ *
+ *	Locate the first entry in a hash table and set up a record that can be
+ *	used to step through all the remaining entries of the table.
+ *
+ * Results:
+ *	The return value is a pointer to the first entry in tablePtr, or NULL
+ *	if tablePtr has no entries in it. The memory at *searchPtr is
+ *	initialized so that subsequent calls to Tcl_NextHashEntry will return
+ *	all of the entries in the table, one at a time.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashEntry *
+Tcl_FirstHashEntry(
+    Tcl_HashTable *tablePtr,	/* Table to search. */
+    Tcl_HashSearch *searchPtr)	/* Place to store information about progress
+				 * through the table. */
+{
+    searchPtr->tablePtr = tablePtr;
+    searchPtr->nextIndex = 0;
+    searchPtr->nextEntryPtr = NULL;
+    return Tcl_NextHashEntry(searchPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NextHashEntry --
+ *
+ *	Once a hash table enumeration has been initiated by calling
+ *	Tcl_FirstHashEntry, this function may be called to return successive
+ *	elements of the table.
+ *
+ * Results:
+ *	The return value is the next entry in the hash table being enumerated,
+ *	or NULL if the end of the table is reached.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashEntry *
+Tcl_NextHashEntry(
+    register Tcl_HashSearch *searchPtr)
+				/* Place to store information about progress
+				 * through the table. Must have been
+				 * initialized by calling
+				 * Tcl_FirstHashEntry. */
+{
+    Tcl_HashEntry *hPtr;
+    Tcl_HashTable *tablePtr = searchPtr->tablePtr;
+
+    while (searchPtr->nextEntryPtr == NULL) {
+	if (searchPtr->nextIndex >= tablePtr->numBuckets) {
+	    return NULL;
+	}
+	searchPtr->nextEntryPtr =
+		tablePtr->buckets[searchPtr->nextIndex];
+	searchPtr->nextIndex++;
+    }
+    hPtr = searchPtr->nextEntryPtr;
+    searchPtr->nextEntryPtr = hPtr->nextPtr;
+    return hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_HashStats --
+ *
+ *	Return statistics describing the layout of the hash table in its hash
+ *	buckets.
+ *
+ * Results:
+ *	The return value is a malloc-ed string containing information about
+ *	tablePtr. It is the caller's responsibility to free this string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_HashStats(
+    Tcl_HashTable *tablePtr)	/* Table for which to produce stats. */
+{
+#define NUM_COUNTERS 10
+    int count[NUM_COUNTERS], overflow, i, j;
+    double average, tmp;
+    register Tcl_HashEntry *hPtr;
+    char *result, *p;
+
+    /*
+     * Compute a histogram of bucket usage.
+     */
+
+    for (i = 0; i < NUM_COUNTERS; i++) {
+	count[i] = 0;
+    }
+    overflow = 0;
+    average = 0.0;
+    for (i = 0; i < tablePtr->numBuckets; i++) {
+	j = 0;
+	for (hPtr = tablePtr->buckets[i]; hPtr != NULL; hPtr = hPtr->nextPtr) {
+	    j++;
+	}
+	if (j < NUM_COUNTERS) {
+	    count[j]++;
+	} else {
+	    overflow++;
+	}
+	tmp = j;
+	if (tablePtr->numEntries != 0) {
+	    average += (tmp+1.0)*(tmp/tablePtr->numEntries)/2.0;
+	}
+    }
+
+    /*
+     * Print out the histogram and a few other pieces of information.
+     */
+
+    result = ckalloc((NUM_COUNTERS * 60) + 300);
+    sprintf(result, "%d entries in table, %d buckets\n",
+	    tablePtr->numEntries, tablePtr->numBuckets);
+    p = result + strlen(result);
+    for (i = 0; i < NUM_COUNTERS; i++) {
+	sprintf(p, "number of buckets with %d entries: %d\n",
+		i, count[i]);
+	p += strlen(p);
+    }
+    sprintf(p, "number of buckets with %d or more entries: %d\n",
+	    NUM_COUNTERS, overflow);
+    p += strlen(p);
+    sprintf(p, "average search distance for entry: %.1f", average);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AllocArrayEntry --
+ *
+ *	Allocate space for a Tcl_HashEntry containing the array key.
+ *
+ * Results:
+ *	The return value is a pointer to the created entry.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashEntry *
+AllocArrayEntry(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key to store in the hash table entry. */
+{
+    int *array = (int *) keyPtr;
+    register int *iPtr1, *iPtr2;
+    Tcl_HashEntry *hPtr;
+    int count;
+    unsigned int size;
+
+    count = tablePtr->keyType;
+
+    size = sizeof(Tcl_HashEntry) + (count*sizeof(int)) - sizeof(hPtr->key);
+    if (size < sizeof(Tcl_HashEntry)) {
+	size = sizeof(Tcl_HashEntry);
+    }
+    hPtr = ckalloc(size);
+
+    for (iPtr1 = array, iPtr2 = hPtr->key.words;
+	    count > 0; count--, iPtr1++, iPtr2++) {
+	*iPtr2 = *iPtr1;
+    }
+    hPtr->clientData = 0;
+
+    return hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompareArrayKeys --
+ *
+ *	Compares two array keys.
+ *
+ * Results:
+ *	The return value is 0 if they are different and 1 if they are the
+ *	same.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompareArrayKeys(
+    void *keyPtr,		/* New key to compare. */
+    Tcl_HashEntry *hPtr)	/* Existing key to compare. */
+{
+    register const int *iPtr1 = (const int *) keyPtr;
+    register const int *iPtr2 = (const int *) hPtr->key.words;
+    Tcl_HashTable *tablePtr = hPtr->tablePtr;
+    int count;
+
+    for (count = tablePtr->keyType; ; count--, iPtr1++, iPtr2++) {
+	if (count == 0) {
+	    return 1;
+	}
+	if (*iPtr1 != *iPtr2) {
+	    break;
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * HashArrayKey --
+ *
+ *	Compute a one-word summary of an array, which can be used to generate
+ *	a hash index.
+ *
+ * Results:
+ *	The return value is a one-word summary of the information in
+ *	string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static TCL_HASH_TYPE
+HashArrayKey(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key from which to compute hash value. */
+{
+    register const int *array = (const int *) keyPtr;
+    register unsigned int result;
+    int count;
+
+    for (result = 0, count = tablePtr->keyType; count > 0;
+	    count--, array++) {
+	result += *array;
+    }
+    return (TCL_HASH_TYPE) result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AllocStringEntry --
+ *
+ *	Allocate space for a Tcl_HashEntry containing the string key.
+ *
+ * Results:
+ *	The return value is a pointer to the created entry.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashEntry *
+AllocStringEntry(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key to store in the hash table entry. */
+{
+    const char *string = (const char *) keyPtr;
+    Tcl_HashEntry *hPtr;
+    unsigned int size, allocsize;
+
+    allocsize = size = strlen(string) + 1;
+    if (size < sizeof(hPtr->key)) {
+	allocsize = sizeof(hPtr->key);
+    }
+    hPtr = ckalloc(TclOffset(Tcl_HashEntry, key) + allocsize);
+    memcpy(hPtr->key.string, string, size);
+    hPtr->clientData = 0;
+    return hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompareStringKeys --
+ *
+ *	Compares two string keys.
+ *
+ * Results:
+ *	The return value is 0 if they are different and 1 if they are the
+ *	same.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompareStringKeys(
+    void *keyPtr,		/* New key to compare. */
+    Tcl_HashEntry *hPtr)	/* Existing key to compare. */
+{
+    register const char *p1 = (const char *) keyPtr;
+    register const char *p2 = (const char *) hPtr->key.string;
+
+    return !strcmp(p1, p2);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * HashStringKey --
+ *
+ *	Compute a one-word summary of a text string, which can be used to
+ *	generate a hash index.
+ *
+ * Results:
+ *	The return value is a one-word summary of the information in string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static TCL_HASH_TYPE
+HashStringKey(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key from which to compute hash value. */
+{
+    register const char *string = keyPtr;
+    register unsigned int result;
+    register char c;
+
+    /*
+     * I tried a zillion different hash functions and asked many other people
+     * for advice. Many people had their own favorite functions, all
+     * different, but no-one had much idea why they were good ones. I chose
+     * the one below (multiply by 9 and add new character) because of the
+     * following reasons:
+     *
+     * 1. Multiplying by 10 is perfect for keys that are decimal strings, and
+     *    multiplying by 9 is just about as good.
+     * 2. Times-9 is (shift-left-3) plus (old). This means that each
+     *    character's bits hang around in the low-order bits of the hash value
+     *    for ever, plus they spread fairly rapidly up to the high-order bits
+     *    to fill out the hash value. This seems works well both for decimal
+     *    and non-decimal strings, but isn't strong against maliciously-chosen
+     *    keys.
+     *
+     * Note that this function is very weak against malicious strings; it's
+     * very easy to generate multiple keys that have the same hashcode. On the
+     * other hand, that hardly ever actually occurs and this function *is*
+     * very cheap, even by comparison with industry-standard hashes like FNV.
+     * If real strength of hash is required though, use a custom hash based on
+     * Bob Jenkins's lookup3(), but be aware that it's significantly slower.
+     * Since Tcl command and namespace names are usually reasonably-named (the
+     * main use for string hashes in modern Tcl) speed is far more important
+     * than strength.
+     *
+     * See also HashString in tclLiteral.c.
+     * See also TclObjHashKey in tclObj.c.
+     *
+     * See [tcl-Feature Request #2958832]
+     */
+
+    if ((result = UCHAR(*string)) != 0) {
+	while ((c = *++string) != 0) {
+	    result += (result << 3) + UCHAR(c);
+	}
+    }
+    return (TCL_HASH_TYPE) result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BogusFind --
+ *
+ *	This function is invoked when Tcl_FindHashEntry is called on a
+ *	table that has been deleted.
+ *
+ * Results:
+ *	If Tcl_Panic returns (which it shouldn't) this function returns NULL.
+ *
+ * Side effects:
+ *	Generates a panic.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static Tcl_HashEntry *
+BogusFind(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const char *key)		/* Key to use to find matching entry. */
+{
+    Tcl_Panic("called %s on deleted table", "Tcl_FindHashEntry");
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BogusCreate --
+ *
+ *	This function is invoked when Tcl_CreateHashEntry is called on a
+ *	table that has been deleted.
+ *
+ * Results:
+ *	If panic returns (which it shouldn't) this function returns NULL.
+ *
+ * Side effects:
+ *	Generates a panic.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static Tcl_HashEntry *
+BogusCreate(
+    Tcl_HashTable *tablePtr,	/* Table in which to lookup entry. */
+    const char *key,		/* Key to use to find or create matching
+				 * entry. */
+    int *newPtr)		/* Store info here telling whether a new entry
+				 * was created. */
+{
+    Tcl_Panic("called %s on deleted table", "Tcl_CreateHashEntry");
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RebuildTable --
+ *
+ *	This function is invoked when the ratio of entries to hash buckets
+ *	becomes too large. It creates a new table with a larger bucket array
+ *	and moves all of the entries into the new table.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory gets reallocated and entries get re-hashed to new buckets.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RebuildTable(
+    register Tcl_HashTable *tablePtr)	/* Table to enlarge. */
+{
+    int oldSize, count, index;
+    Tcl_HashEntry **oldBuckets;
+    register Tcl_HashEntry **oldChainPtr, **newChainPtr;
+    register Tcl_HashEntry *hPtr;
+    const Tcl_HashKeyType *typePtr;
+
+    if (tablePtr->keyType == TCL_STRING_KEYS) {
+	typePtr = &tclStringHashKeyType;
+    } else if (tablePtr->keyType == TCL_ONE_WORD_KEYS) {
+	typePtr = &tclOneWordHashKeyType;
+    } else if (tablePtr->keyType == TCL_CUSTOM_TYPE_KEYS
+	    || tablePtr->keyType == TCL_CUSTOM_PTR_KEYS) {
+	typePtr = tablePtr->typePtr;
+    } else {
+	typePtr = &tclArrayHashKeyType;
+    }
+
+    oldSize = tablePtr->numBuckets;
+    oldBuckets = tablePtr->buckets;
+
+    /*
+     * Allocate and initialize the new bucket array, and set up hashing
+     * constants for new array size.
+     */
+
+    tablePtr->numBuckets *= 4;
+    if (typePtr->flags & TCL_HASH_KEY_SYSTEM_HASH) {
+	tablePtr->buckets = (Tcl_HashEntry **) TclpSysAlloc((unsigned)
+		(tablePtr->numBuckets * sizeof(Tcl_HashEntry *)), 0);
+    } else {
+	tablePtr->buckets =
+		ckalloc(tablePtr->numBuckets * sizeof(Tcl_HashEntry *));
+    }
+    for (count = tablePtr->numBuckets, newChainPtr = tablePtr->buckets;
+	    count > 0; count--, newChainPtr++) {
+	*newChainPtr = NULL;
+    }
+    tablePtr->rebuildSize *= 4;
+    tablePtr->downShift -= 2;
+    tablePtr->mask = (tablePtr->mask << 2) + 3;
+
+    /*
+     * Rehash all of the existing entries into the new bucket array.
+     */
+
+    for (oldChainPtr = oldBuckets; oldSize > 0; oldSize--, oldChainPtr++) {
+	for (hPtr = *oldChainPtr; hPtr != NULL; hPtr = *oldChainPtr) {
+	    *oldChainPtr = hPtr->nextPtr;
+	    if (typePtr->hashKeyProc == NULL
+		    || typePtr->flags & TCL_HASH_KEY_RANDOMIZE_HASH) {
+		index = RANDOM_INDEX(tablePtr, PTR2INT(hPtr->hash));
+	    } else {
+		index = PTR2UINT(hPtr->hash) & tablePtr->mask;
+	    }
+	    hPtr->nextPtr = tablePtr->buckets[index];
+	    tablePtr->buckets[index] = hPtr;
+	}
+    }
+
+    /*
+     * Free up the old bucket array, if it was dynamically allocated.
+     */
+
+    if (oldBuckets != tablePtr->staticBuckets) {
+	if (typePtr->flags & TCL_HASH_KEY_SYSTEM_HASH) {
+	    TclpSysFree((char *) oldBuckets);
+	} else {
+	    ckfree(oldBuckets);
+	}
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclHistory.c b/generic/tclHistory.c
new file mode 100644
index 0000000..47806d4
--- /dev/null
+++ b/generic/tclHistory.c
@@ -0,0 +1,229 @@
+/*
+ * tclHistory.c --
+ *
+ *	This module and the Tcl library file history.tcl together implement
+ *	Tcl command history. Tcl_RecordAndEval(Obj) can be called to record
+ *	commands ("events") before they are executed. Commands defined in
+ *	history.tcl may be used to perform history substitutions.
+ *
+ * Copyright (c) 1990-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Type of the assocData structure used to hold the reference to the [history
+ * add] subcommand, used in Tcl_RecordAndEvalObj.
+ */
+
+typedef struct {
+    Tcl_Obj *historyObj;	/* == "::history" */
+    Tcl_Obj *addObj;		/* == "add" */
+} HistoryObjs;
+
+#define HISTORY_OBJS_KEY	"::tcl::HistoryObjs"
+
+/*
+ * Static functions in this file.
+ */
+
+static Tcl_InterpDeleteProc DeleteHistoryObjs;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RecordAndEval --
+ *
+ *	This procedure adds its command argument to the current list of
+ *	recorded events and then executes the command by calling Tcl_Eval.
+ *
+ * Results:
+ *	The return value is a standard Tcl return value, the result of
+ *	executing cmd.
+ *
+ * Side effects:
+ *	The command is recorded and executed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RecordAndEval(
+    Tcl_Interp *interp,		/* Token for interpreter in which command will
+				 * be executed. */
+    const char *cmd,		/* Command to record. */
+    int flags)			/* Additional flags. TCL_NO_EVAL means only
+				 * record: don't execute command.
+				 * TCL_EVAL_GLOBAL means use Tcl_GlobalEval
+				 * instead of Tcl_Eval. */
+{
+    register Tcl_Obj *cmdPtr;
+    int result;
+
+    if (cmd[0]) {
+	/*
+	 * Call Tcl_RecordAndEvalObj to do the actual work.
+	 */
+
+	cmdPtr = Tcl_NewStringObj(cmd, -1);
+	Tcl_IncrRefCount(cmdPtr);
+	result = Tcl_RecordAndEvalObj(interp, cmdPtr, flags);
+
+	/*
+	 * Move the interpreter's object result to the string result, then
+	 * reset the object result.
+	 */
+
+	(void) Tcl_GetStringResult(interp);
+
+	/*
+	 * Discard the Tcl object created to hold the command.
+	 */
+
+	Tcl_DecrRefCount(cmdPtr);
+    } else {
+	/*
+	 * An empty string. Just reset the interpreter's result.
+	 */
+
+	Tcl_ResetResult(interp);
+	result = TCL_OK;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RecordAndEvalObj --
+ *
+ *	This procedure adds the command held in its argument object to the
+ *	current list of recorded events and then executes the command by
+ *	calling Tcl_EvalObj.
+ *
+ * Results:
+ *	The return value is a standard Tcl return value, the result of
+ *	executing the command.
+ *
+ * Side effects:
+ *	The command is recorded and executed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RecordAndEvalObj(
+    Tcl_Interp *interp,		/* Token for interpreter in which command will
+				 * be executed. */
+    Tcl_Obj *cmdPtr,		/* Points to object holding the command to
+				 * record and execute. */
+    int flags)			/* Additional flags. TCL_NO_EVAL means record
+				 * only: don't execute the command.
+				 * TCL_EVAL_GLOBAL means evaluate the script
+				 * in global variable context instead of the
+				 * current procedure. */
+{
+    int result, call = 1;
+    Tcl_CmdInfo info;
+    HistoryObjs *histObjsPtr =
+	    Tcl_GetAssocData(interp, HISTORY_OBJS_KEY, NULL);
+
+    /*
+     * Create the references to the [::history add] command if necessary.
+     */
+
+    if (histObjsPtr == NULL) {
+	histObjsPtr = ckalloc(sizeof(HistoryObjs));
+	TclNewLiteralStringObj(histObjsPtr->historyObj, "::history");
+	TclNewLiteralStringObj(histObjsPtr->addObj, "add");
+	Tcl_IncrRefCount(histObjsPtr->historyObj);
+	Tcl_IncrRefCount(histObjsPtr->addObj);
+	Tcl_SetAssocData(interp, HISTORY_OBJS_KEY, DeleteHistoryObjs,
+		histObjsPtr);
+    }
+
+    /*
+     * Do not call [history] if it has been replaced by an empty proc
+     */
+
+    result = Tcl_GetCommandInfo(interp, "::history", &info);
+    if (result && (info.deleteProc == TclProcDeleteProc)) {
+	Proc *procPtr = (Proc *) info.objClientData;
+	call = (procPtr->cmdPtr->compileProc != TclCompileNoOp);
+    }
+
+    if (call) {
+	Tcl_Obj *list[3];
+
+	/*
+	 * Do recording by eval'ing a tcl history command: history add $cmd.
+	 */
+
+	list[0] = histObjsPtr->historyObj;
+	list[1] = histObjsPtr->addObj;
+	list[2] = cmdPtr;
+
+	Tcl_IncrRefCount(cmdPtr);
+	(void) Tcl_EvalObjv(interp, 3, list, TCL_EVAL_GLOBAL);
+	Tcl_DecrRefCount(cmdPtr);
+
+	/*
+	 * One possible failure mode above: exceeding a resource limit.
+	 */
+
+	if (Tcl_LimitExceeded(interp)) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Execute the command.
+     */
+
+    result = TCL_OK;
+    if (!(flags & TCL_NO_EVAL)) {
+	result = Tcl_EvalObjEx(interp, cmdPtr, flags & TCL_EVAL_GLOBAL);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteHistoryObjs --
+ *
+ *	Called to delete the references to the constant words used when adding
+ *	to the history.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The constant words may be deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteHistoryObjs(
+    ClientData clientData,
+    Tcl_Interp *interp)
+{
+    register HistoryObjs *histObjsPtr = clientData;
+
+    TclDecrRefCount(histObjsPtr->historyObj);
+    TclDecrRefCount(histObjsPtr->addObj);
+    ckfree(histObjsPtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIO.c b/generic/tclIO.c
new file mode 100644
index 0000000..1460392
--- /dev/null
+++ b/generic/tclIO.c
@@ -0,0 +1,11278 @@
+/*
+ * tclIO.c --
+ *
+ *	This file provides the generic portions (those that are the same on
+ *	all platforms and for all channel types) of Tcl's IO facilities.
+ *
+ * Copyright (c) 1998-2000 Ajuba Solutions
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Contributions from Don Porter, NIST, 2014. (not subject to US copyright)
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclIO.h"
+#include <assert.h>
+
+/*
+ * For each channel handler registered in a call to Tcl_CreateChannelHandler,
+ * there is one record of the following type. All of records for a specific
+ * channel are chained together in a singly linked list which is stored in
+ * the channel structure.
+ */
+
+typedef struct ChannelHandler {
+    Channel *chanPtr;		/* The channel structure for this channel. */
+    int mask;			/* Mask of desired events. */
+    Tcl_ChannelProc *proc;	/* Procedure to call in the type of
+				 * Tcl_CreateChannelHandler. */
+    ClientData clientData;	/* Argument to pass to procedure. */
+    struct ChannelHandler *nextPtr;
+				/* Next one in list of registered handlers. */
+} ChannelHandler;
+
+/*
+ * This structure keeps track of the current ChannelHandler being invoked in
+ * the current invocation of Tcl_NotifyChannel. There is a potential
+ * problem if a ChannelHandler is deleted while it is the current one, since
+ * Tcl_NotifyChannel needs to look at the nextPtr field. To handle this
+ * problem, structures of the type below indicate the next handler to be
+ * processed for any (recursively nested) dispatches in progress. The
+ * nextHandlerPtr field is updated if the handler being pointed to is deleted.
+ * The nestedHandlerPtr field is used to chain together all recursive
+ * invocations, so that Tcl_DeleteChannelHandler can find all the recursively
+ * nested invocations of Tcl_NotifyChannel and compare the handler being
+ * deleted against the NEXT handler to be invoked in that invocation; when it
+ * finds such a situation, Tcl_DeleteChannelHandler updates the nextHandlerPtr
+ * field of the structure to the next handler.
+ */
+
+typedef struct NextChannelHandler {
+    ChannelHandler *nextHandlerPtr;	/* The next handler to be invoked in
+					 * this invocation. */
+    struct NextChannelHandler *nestedHandlerPtr;
+					/* Next nested invocation of
+					 * Tcl_NotifyChannel. */
+} NextChannelHandler;
+
+/*
+ * The following structure is used by Tcl_GetsObj() to encapsulates the
+ * state for a "gets" operation.
+ */
+
+typedef struct GetsState {
+    Tcl_Obj *objPtr;		/* The object to which UTF-8 characters
+				 * will be appended. */
+    char **dstPtr;		/* Pointer into objPtr's string rep where
+				 * next character should be stored. */
+    Tcl_Encoding encoding;	/* The encoding to use to convert raw bytes
+				 * to UTF-8.  */
+    ChannelBuffer *bufPtr;	/* The current buffer of raw bytes being
+				 * emptied. */
+    Tcl_EncodingState state;	/* The encoding state just before the last
+				 * external to UTF-8 conversion in
+				 * FilterInputBytes(). */
+    int rawRead;		/* The number of bytes removed from bufPtr
+				 * in the last call to FilterInputBytes(). */
+    int bytesWrote;		/* The number of bytes of UTF-8 data
+				 * appended to objPtr during the last call to
+				 * FilterInputBytes(). */
+    int charsWrote;		/* The corresponding number of UTF-8
+				 * characters appended to objPtr during the
+				 * last call to FilterInputBytes(). */
+    int totalChars;		/* The total number of UTF-8 characters
+				 * appended to objPtr so far, just before the
+				 * last call to FilterInputBytes(). */
+} GetsState;
+
+/*
+ * The following structure encapsulates the state for a background channel
+ * copy.  Note that the data buffer for the copy will be appended to this
+ * structure.
+ */
+
+typedef struct CopyState {
+    struct Channel *readPtr;	/* Pointer to input channel. */
+    struct Channel *writePtr;	/* Pointer to output channel. */
+    int readFlags;		/* Original read channel flags. */
+    int writeFlags;		/* Original write channel flags. */
+    Tcl_WideInt toRead;		/* Number of bytes to copy, or -1. */
+    Tcl_WideInt total;		/* Total bytes transferred (written). */
+    Tcl_Interp *interp;		/* Interp that started the copy. */
+    Tcl_Obj *cmdPtr;		/* Command to be invoked at completion. */
+    int bufSize;		/* Size of appended buffer. */
+    char buffer[1];		/* Copy buffer, this must be the last
+                                 * field. */
+} CopyState;
+
+/*
+ * All static variables used in this file are collected into a single instance
+ * of the following structure. For multi-threaded implementations, there is
+ * one instance of this structure for each thread.
+ *
+ * Notice that different structures with the same name appear in other files.
+ * The structure defined below is used in this file only.
+ */
+
+typedef struct {
+    NextChannelHandler *nestedHandlerPtr;
+				/* This variable holds the list of nested
+				 * Tcl_NotifyChannel invocations. */
+    ChannelState *firstCSPtr;	/* List of all channels currently open,
+				 * indexed by ChannelState, as only one
+				 * ChannelState exists per set of stacked
+				 * channels. */
+    Tcl_Channel stdinChannel;	/* Static variable for the stdin channel. */
+    int stdinInitialized;
+    Tcl_Channel stdoutChannel;	/* Static variable for the stdout channel. */
+    int stdoutInitialized;
+    Tcl_Channel stderrChannel;	/* Static variable for the stderr channel. */
+    int stderrInitialized;
+    Tcl_Encoding binaryEncoding;
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * Structure to record a close callback. One such record exists for
+ * each close callback registered for a channel.
+ */
+
+typedef struct CloseCallback {
+    Tcl_CloseProc *proc;		/* The procedure to call. */
+    ClientData clientData;		/* Arbitrary one-word data to pass
+					 * to the callback. */
+    struct CloseCallback *nextPtr;	/* For chaining close callbacks. */
+} CloseCallback;
+
+/*
+ * Static functions in this file:
+ */
+
+static ChannelBuffer *	AllocChannelBuffer(int length);
+static void		PreserveChannelBuffer(ChannelBuffer *bufPtr);
+static void		ReleaseChannelBuffer(ChannelBuffer *bufPtr);
+static int		IsShared(ChannelBuffer *bufPtr);
+static void		ChannelFree(Channel *chanPtr);
+static void		ChannelTimerProc(ClientData clientData);
+static int		ChanRead(Channel *chanPtr, char *dst, int dstSize);
+static int		CheckChannelErrors(ChannelState *statePtr,
+			    int direction);
+static int		CheckForDeadChannel(Tcl_Interp *interp,
+			    ChannelState *statePtr);
+static void		CheckForStdChannelsBeingClosed(Tcl_Channel chan);
+static void		CleanupChannelHandlers(Tcl_Interp *interp,
+			    Channel *chanPtr);
+static int		CloseChannel(Tcl_Interp *interp, Channel *chanPtr,
+			    int errorCode);
+static int		CloseChannelPart(Tcl_Interp *interp, Channel *chanPtr,
+			    int errorCode, int flags);
+static int		CloseWrite(Tcl_Interp *interp, Channel *chanPtr);
+static void		CommonGetsCleanup(Channel *chanPtr);
+static int		CopyData(CopyState *csPtr, int mask);
+static int		MoveBytes(CopyState *csPtr);
+
+static void		MBCallback(CopyState *csPtr, Tcl_Obj *errObj);
+static void		MBError(CopyState *csPtr, int mask, int errorCode);
+static int		MBRead(CopyState *csPtr);
+static int		MBWrite(CopyState *csPtr);
+static void		MBEvent(ClientData clientData, int mask);
+
+static void		CopyEventProc(ClientData clientData, int mask);
+static void		CreateScriptRecord(Tcl_Interp *interp,
+			    Channel *chanPtr, int mask, Tcl_Obj *scriptPtr);
+static void		DeleteChannelTable(ClientData clientData,
+			    Tcl_Interp *interp);
+static void		DeleteScriptRecord(Tcl_Interp *interp,
+			    Channel *chanPtr, int mask);
+static int		DetachChannel(Tcl_Interp *interp, Tcl_Channel chan);
+static void		DiscardInputQueued(ChannelState *statePtr,
+			    int discardSavedBuffers);
+static void		DiscardOutputQueued(ChannelState *chanPtr);
+static int		DoRead(Channel *chanPtr, char *dst, int bytesToRead,
+			    int allowShortReads);
+static int		DoReadChars(Channel *chan, Tcl_Obj *objPtr, int toRead,
+			    int appendFlag);
+static int		FilterInputBytes(Channel *chanPtr,
+			    GetsState *statePtr);
+static int		FlushChannel(Tcl_Interp *interp, Channel *chanPtr,
+			    int calledFromAsyncFlush);
+static int		TclGetsObjBinary(Tcl_Channel chan, Tcl_Obj *objPtr);
+static Tcl_Encoding	GetBinaryEncoding();
+static void		FreeBinaryEncoding(ClientData clientData);
+static Tcl_HashTable *	GetChannelTable(Tcl_Interp *interp);
+static int		GetInput(Channel *chanPtr);
+static int		HaveVersion(const Tcl_ChannelType *typePtr,
+			    Tcl_ChannelTypeVersion minimumVersion);
+static void		PeekAhead(Channel *chanPtr, char **dstEndPtr,
+			    GetsState *gsPtr);
+static int		ReadBytes(ChannelState *statePtr, Tcl_Obj *objPtr,
+			    int charsLeft);
+static int		ReadChars(ChannelState *statePtr, Tcl_Obj *objPtr,
+			    int charsLeft, int *factorPtr);
+static void		RecycleBuffer(ChannelState *statePtr,
+			    ChannelBuffer *bufPtr, int mustDiscard);
+static int		StackSetBlockMode(Channel *chanPtr, int mode);
+static int		SetBlockMode(Tcl_Interp *interp, Channel *chanPtr,
+			    int mode);
+static void		StopCopy(CopyState *csPtr);
+static void		TranslateInputEOL(ChannelState *statePtr, char *dst,
+			    const char *src, int *dstLenPtr, int *srcLenPtr);
+static void		UpdateInterest(Channel *chanPtr);
+static int		Write(Channel *chanPtr, const char *src,
+			    int srcLen, Tcl_Encoding encoding);
+static Tcl_Obj *	FixLevelCode(Tcl_Obj *msg);
+static void		SpliceChannel(Tcl_Channel chan);
+static void		CutChannel(Tcl_Channel chan);
+static int              WillRead(Channel *chanPtr);
+
+#define WriteChars(chanPtr, src, srcLen) \
+			Write(chanPtr, src, srcLen, chanPtr->state->encoding)
+#define WriteBytes(chanPtr, src, srcLen) \
+			Write(chanPtr, src, srcLen, tclIdentityEncoding)
+
+/*
+ * Simplifying helper macros. All may use their argument(s) multiple times.
+ * The ANSI C "prototypes" for the macros are listed below, together with a
+ * short description of what the macro does.
+ *
+ * --------------------------------------------------------------------------
+ * int BytesLeft(ChannelBuffer *bufPtr)
+ *
+ *	Returns the number of bytes of data remaining in the buffer.
+ *
+ * int SpaceLeft(ChannelBuffer *bufPtr)
+ *
+ *	Returns the number of bytes of space remaining at the end of the
+ *	buffer.
+ *
+ * int IsBufferReady(ChannelBuffer *bufPtr)
+ *
+ *	Returns whether a buffer has bytes available within it.
+ *
+ * int IsBufferEmpty(ChannelBuffer *bufPtr)
+ *
+ *	Returns whether a buffer is entirely empty. Note that this is not the
+ *	inverse of the above operation; trying to merge the two seems to lead
+ *	to occasional crashes...
+ *
+ * int IsBufferFull(ChannelBuffer *bufPtr)
+ *
+ *	Returns whether more data can be added to a buffer.
+ *
+ * int IsBufferOverflowing(ChannelBuffer *bufPtr)
+ *
+ *	Returns whether a buffer has more data in it than it should.
+ *
+ * char *InsertPoint(ChannelBuffer *bufPtr)
+ *
+ *	Returns a pointer to where characters should be added to the buffer.
+ *
+ * char *RemovePoint(ChannelBuffer *bufPtr)
+ *
+ *	Returns a pointer to where characters should be removed from the
+ *	buffer.
+ * --------------------------------------------------------------------------
+ */
+
+#define BytesLeft(bufPtr)	((bufPtr)->nextAdded - (bufPtr)->nextRemoved)
+
+#define SpaceLeft(bufPtr)	((bufPtr)->bufLength - (bufPtr)->nextAdded)
+
+#define IsBufferReady(bufPtr)	((bufPtr)->nextAdded > (bufPtr)->nextRemoved)
+
+#define IsBufferEmpty(bufPtr)	((bufPtr)->nextAdded == (bufPtr)->nextRemoved)
+
+#define IsBufferFull(bufPtr)	((bufPtr) && (bufPtr)->nextAdded >= (bufPtr)->bufLength)
+
+#define IsBufferOverflowing(bufPtr) ((bufPtr)->nextAdded>(bufPtr)->bufLength)
+
+#define InsertPoint(bufPtr)	((bufPtr)->buf + (bufPtr)->nextAdded)
+
+#define RemovePoint(bufPtr)	((bufPtr)->buf + (bufPtr)->nextRemoved)
+
+/*
+ * For working with channel state flag bits.
+ */
+
+#define SetFlag(statePtr, flag)		((statePtr)->flags |= (flag))
+#define ResetFlag(statePtr, flag)	((statePtr)->flags &= ~(flag))
+#define GotFlag(statePtr, flag)		((statePtr)->flags & (flag))
+
+/*
+ * Macro for testing whether a string (in optionName, length len) matches a
+ * value (prefix matching rules). Arguments are the minimum length to match
+ * and the value to match against. (Can't use Tcl_GetIndexFromObj as this is
+ * used in a situation where no objects are available.)
+ */
+
+#define HaveOpt(minLength, nameString) \
+	((len > (minLength)) && (optionName[1] == (nameString)[1]) \
+		&& (strncmp(optionName, (nameString), len) == 0))
+
+/*
+ * The ChannelObjType type.  Used to store the result of looking up
+ * a channel name in the context of an interp.  Saves the lookup
+ * result and values needed to check its continued validity.
+ */
+
+typedef struct ResolvedChanName {
+    ChannelState *statePtr;	/* The saved lookup result */
+    Tcl_Interp *interp;		/* The interp in which the lookup was done. */
+    size_t epoch;		/* The epoch of the channel when the lookup
+				 * was done. Use to verify validity. */
+    size_t refCount;		/* Share this struct among many Tcl_Obj. */
+} ResolvedChanName;
+
+static void		DupChannelIntRep(Tcl_Obj *objPtr, Tcl_Obj *copyPtr);
+static void		FreeChannelIntRep(Tcl_Obj *objPtr);
+
+static const Tcl_ObjType chanObjType = {
+    "channel",			/* name for this type */
+    FreeChannelIntRep,		/* freeIntRepProc */
+    DupChannelIntRep,		/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    NULL			/* setFromAnyProc */
+};
+
+#define BUSY_STATE(st, fl) \
+     ((((st)->csPtrR) && ((fl) & TCL_READABLE)) || \
+      (((st)->csPtrW) && ((fl) & TCL_WRITABLE)))
+
+#define MAX_CHANNEL_BUFFER_SIZE (1024*1024)
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * ChanClose, ChanRead, ChanSeek, ChanThreadAction, ChanWatch, ChanWrite --
+ *
+ *	Simplify the access to selected channel driver "methods" that are used
+ *	in multiple places in a stereotypical fashion. These are just thin
+ *	wrappers around the driver functions.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static inline int
+ChanClose(
+    Channel *chanPtr,
+    Tcl_Interp *interp)
+{
+    if (chanPtr->typePtr->closeProc != TCL_CLOSE2PROC) {
+	return chanPtr->typePtr->closeProc(chanPtr->instanceData, interp);
+    } else {
+	return chanPtr->typePtr->close2Proc(chanPtr->instanceData, interp, 0);
+    }
+}
+
+static inline int
+ChanCloseHalf(
+    Channel *chanPtr,
+    Tcl_Interp *interp,
+    int flags)
+{
+    return chanPtr->typePtr->close2Proc(chanPtr->instanceData, interp, flags);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * ChanRead --
+ *
+ *	Read up to dstSize bytes using the inputProc of chanPtr, store them at
+ *	dst, and return the number of bytes stored.
+ *
+ * Results:
+ *	The return value of the driver inputProc,
+ *	  - number of bytes stored at dst, ot
+ *	  - -1 on error, with a Posix error code available to the caller by
+ *	    calling Tcl_GetErrno().
+ *
+ * Side effects:
+ *	The CHANNEL_BLOCKED and CHANNEL_EOF flags of the channel state are set
+ *	as appropriate.  On EOF, the inputEncodingFlags are set to perform
+ *	ending operations on decoding.
+ *
+ *	TODO - Is this really the right place for that?
+ *
+ *---------------------------------------------------------------------------
+ */
+static int
+ChanRead(
+    Channel *chanPtr,
+    char *dst,
+    int dstSize)
+{
+    int bytesRead, result;
+
+    /*
+     * If the caller asked for zero bytes, we'd force the inputProc to return
+     * zero bytes, and then misinterpret that as EOF.
+     */
+
+    assert(dstSize > 0);
+
+    /*
+     * Each read op must set the blocked and eof states anew, not let
+     * the effect of prior reads leak through.
+     */
+
+    if (GotFlag(chanPtr->state, CHANNEL_EOF)) {
+        chanPtr->state->inputEncodingFlags |= TCL_ENCODING_START;
+    }
+    ResetFlag(chanPtr->state, CHANNEL_BLOCKED | CHANNEL_EOF);
+    chanPtr->state->inputEncodingFlags &= ~TCL_ENCODING_END;
+    if (WillRead(chanPtr) < 0) {
+        return -1;
+    }
+
+    bytesRead = chanPtr->typePtr->inputProc(chanPtr->instanceData,
+	    dst, dstSize, &result);
+
+    /*
+     * Stop any flag leakage through stacked channel levels.
+     */
+
+    if (GotFlag(chanPtr->state, CHANNEL_EOF)) {
+        chanPtr->state->inputEncodingFlags |= TCL_ENCODING_START;
+    }
+    ResetFlag(chanPtr->state, CHANNEL_BLOCKED | CHANNEL_EOF);
+    chanPtr->state->inputEncodingFlags &= ~TCL_ENCODING_END;
+    if (bytesRead > 0) {
+	/*
+	 * If we get a short read, signal up that we may be BLOCKED. We should
+	 * avoid calling the driver because on some platforms we will block in
+	 * the low level reading code even though the channel is set into
+	 * nonblocking mode.
+	 */
+
+	if (bytesRead < dstSize) {
+	    SetFlag(chanPtr->state, CHANNEL_BLOCKED);
+	}
+    } else if (bytesRead == 0) {
+	SetFlag(chanPtr->state, CHANNEL_EOF);
+	chanPtr->state->inputEncodingFlags |= TCL_ENCODING_END;
+    } else if (bytesRead < 0) {
+	if ((result == EWOULDBLOCK) || (result == EAGAIN)) {
+	    SetFlag(chanPtr->state, CHANNEL_BLOCKED);
+	    result = EAGAIN;
+	}
+	Tcl_SetErrno(result);
+    }
+    return bytesRead;
+}
+
+static inline Tcl_WideInt
+ChanSeek(
+    Channel *chanPtr,
+    Tcl_WideInt offset,
+    int mode,
+    int *errnoPtr)
+{
+    /*
+     * Note that we prefer the wideSeekProc if that field is available in the
+     * type and non-NULL.
+     */
+
+    if (HaveVersion(chanPtr->typePtr, TCL_CHANNEL_VERSION_3) &&
+	    chanPtr->typePtr->wideSeekProc != NULL) {
+	return chanPtr->typePtr->wideSeekProc(chanPtr->instanceData,
+		offset, mode, errnoPtr);
+    }
+
+    if (offset<Tcl_LongAsWide(LONG_MIN) || offset>Tcl_LongAsWide(LONG_MAX)) {
+	*errnoPtr = EOVERFLOW;
+	return Tcl_LongAsWide(-1);
+    }
+
+    return Tcl_LongAsWide(chanPtr->typePtr->seekProc(chanPtr->instanceData,
+	    Tcl_WideAsLong(offset), mode, errnoPtr));
+}
+
+static inline void
+ChanThreadAction(
+    Channel *chanPtr,
+    int action)
+{
+    Tcl_DriverThreadActionProc *threadActionProc =
+	    Tcl_ChannelThreadActionProc(chanPtr->typePtr);
+
+    if (threadActionProc != NULL) {
+	threadActionProc(chanPtr->instanceData, action);
+    }
+}
+
+static inline void
+ChanWatch(
+    Channel *chanPtr,
+    int mask)
+{
+    chanPtr->typePtr->watchProc(chanPtr->instanceData, mask);
+}
+
+static inline int
+ChanWrite(
+    Channel *chanPtr,
+    const char *src,
+    int srcLen,
+    int *errnoPtr)
+{
+    return chanPtr->typePtr->outputProc(chanPtr->instanceData, src, srcLen,
+	    errnoPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclInitIOSubsystem --
+ *
+ *	Initialize all resources used by this subsystem on a per-process
+ *	basis.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on the memory subsystems.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclInitIOSubsystem(void)
+{
+    /*
+     * By fetching thread local storage we take care of allocating it for each
+     * thread.
+     */
+
+    (void) TCL_TSD_INIT(&dataKey);
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclFinalizeIOSubsystem --
+ *
+ *	Releases all resources used by this subsystem on a per-process basis.
+ *	Closes all extant channels that have not already been closed because
+ *	they were not owned by any interp.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on encoding and memory subsystems.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+void
+TclFinalizeIOSubsystem(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    Channel *chanPtr = NULL;	/* Iterates over open channels. */
+    ChannelState *statePtr;	/* State of channel stack */
+    int active = 1;		/* Flag == 1 while there's still work to do */
+    int doflushnb;
+
+    /*
+     * Fetch the pre-TIP#398 compatibility flag.
+     */
+
+    {
+        const char *s;
+        Tcl_DString ds;
+
+        s = TclGetEnv("TCL_FLUSH_NONBLOCKING_ON_EXIT", &ds);
+        doflushnb = ((s != NULL) && strcmp(s, "0"));
+        if (s != NULL) {
+            Tcl_DStringFree(&ds);
+        }
+    }
+
+    /*
+     * Walk all channel state structures known to this thread and close
+     * corresponding channels.
+     */
+
+    while (active) {
+	/*
+	 * Iterate through the open channel list, and find the first channel
+	 * that isn't dead. We start from the head of the list each time,
+	 * because the close action on one channel can close others.
+	 */
+
+	active = 0;
+	for (statePtr = tsdPtr->firstCSPtr;
+		statePtr != NULL;
+		statePtr = statePtr->nextCSPtr) {
+	    chanPtr = statePtr->topChanPtr;
+            if (GotFlag(statePtr, CHANNEL_DEAD)) {
+                continue;
+            }
+	    if (!GotFlag(statePtr, CHANNEL_INCLOSE | CHANNEL_CLOSED )
+                || GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+                ResetFlag(statePtr, BG_FLUSH_SCHEDULED);
+		active = 1;
+		break;
+	    }
+	}
+
+	/*
+	 * We've found a live (or bg-closing) channel. Close it.
+	 */
+
+	if (active) {
+	    TclChannelPreserve((Tcl_Channel)chanPtr);
+
+	    /*
+	     * TIP #398: by default, we no longer set the channel back into
+             * blocking mode.  To restore the old blocking behavior, the
+             * environment variable TCL_FLUSH_NONBLOCKING_ON_EXIT must be set
+             * and not be "0".
+	     */
+
+            if (doflushnb) {
+                /*
+                 * Set the channel back into blocking mode to ensure that we
+                 * wait for all data to flush out.
+                 */
+
+                (void) Tcl_SetChannelOption(NULL, (Tcl_Channel) chanPtr,
+                                            "-blocking", "on");
+            }
+
+	    if ((chanPtr == (Channel *) tsdPtr->stdinChannel) ||
+		    (chanPtr == (Channel *) tsdPtr->stdoutChannel) ||
+		    (chanPtr == (Channel *) tsdPtr->stderrChannel)) {
+		/*
+		 * Decrement the refcount which was earlier artificially
+		 * bumped up to keep the channel from being closed.
+		 */
+
+		statePtr->refCount--;
+	    }
+
+	    if (statePtr->refCount <= 0) {
+		/*
+		 * Close it only if the refcount indicates that the channel is
+		 * not referenced from any interpreter. If it is, that
+		 * interpreter will close the channel when it gets destroyed.
+		 */
+
+		(void) Tcl_Close(NULL, (Tcl_Channel) chanPtr);
+	    } else {
+		/*
+		 * The refcount is greater than zero, so flush the channel.
+		 */
+
+		Tcl_Flush((Tcl_Channel) chanPtr);
+
+		/*
+		 * Call the device driver to actually close the underlying
+		 * device for this channel.
+		 */
+
+		(void) ChanClose(chanPtr, NULL);
+
+		/*
+		 * Finally, we clean up the fields in the channel data
+		 * structure since all of them have been deleted already. We
+		 * mark the channel with CHANNEL_DEAD to prevent any further
+		 * IO operations on it.
+		 */
+
+		chanPtr->instanceData = NULL;
+		SetFlag(statePtr, CHANNEL_DEAD);
+	    }
+	    TclChannelRelease((Tcl_Channel)chanPtr);
+	}
+    }
+
+    TclpFinalizeSockets();
+    TclpFinalizePipes();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetStdChannel --
+ *
+ *	This function is used to change the channels that are used for
+ *	stdin/stdout/stderr in new interpreters.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetStdChannel(
+    Tcl_Channel channel,
+    int type)			/* One of TCL_STDIN, TCL_STDOUT, TCL_STDERR. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    switch (type) {
+    case TCL_STDIN:
+	tsdPtr->stdinInitialized = 1;
+	tsdPtr->stdinChannel = channel;
+	break;
+    case TCL_STDOUT:
+	tsdPtr->stdoutInitialized = 1;
+	tsdPtr->stdoutChannel = channel;
+	break;
+    case TCL_STDERR:
+	tsdPtr->stderrInitialized = 1;
+	tsdPtr->stderrChannel = channel;
+	break;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetStdChannel --
+ *
+ *	Returns the specified standard channel.
+ *
+ * Results:
+ *	Returns the specified standard channel, or NULL.
+ *
+ * Side effects:
+ *	May cause the creation of a standard channel and the underlying file.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_GetStdChannel(
+    int type)			/* One of TCL_STDIN, TCL_STDOUT, TCL_STDERR. */
+{
+    Tcl_Channel channel = NULL;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    /*
+     * If the channels were not created yet, create them now and store them in
+     * the static variables.
+     */
+
+    switch (type) {
+    case TCL_STDIN:
+	if (!tsdPtr->stdinInitialized) {
+	    tsdPtr->stdinChannel = TclpGetDefaultStdChannel(TCL_STDIN);
+	    tsdPtr->stdinInitialized = 1;
+
+	    /*
+	     * Artificially bump the refcount to ensure that the channel is
+	     * only closed on exit.
+	     *
+	     * NOTE: Must only do this if stdinChannel is not NULL. It can be
+	     * NULL in situations where Tcl is unable to connect to the
+	     * standard input.
+	     */
+
+	    if (tsdPtr->stdinChannel != NULL) {
+		Tcl_RegisterChannel(NULL, tsdPtr->stdinChannel);
+	    }
+	}
+	channel = tsdPtr->stdinChannel;
+	break;
+    case TCL_STDOUT:
+	if (!tsdPtr->stdoutInitialized) {
+	    tsdPtr->stdoutChannel = TclpGetDefaultStdChannel(TCL_STDOUT);
+	    tsdPtr->stdoutInitialized = 1;
+	    if (tsdPtr->stdoutChannel != NULL) {
+		Tcl_RegisterChannel(NULL, tsdPtr->stdoutChannel);
+	    }
+	}
+	channel = tsdPtr->stdoutChannel;
+	break;
+    case TCL_STDERR:
+	if (!tsdPtr->stderrInitialized) {
+	    tsdPtr->stderrChannel = TclpGetDefaultStdChannel(TCL_STDERR);
+	    tsdPtr->stderrInitialized = 1;
+	    if (tsdPtr->stderrChannel != NULL) {
+		Tcl_RegisterChannel(NULL, tsdPtr->stderrChannel);
+	    }
+	}
+	channel = tsdPtr->stderrChannel;
+	break;
+    }
+    return channel;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateCloseHandler
+ *
+ *	Creates a close callback which will be called when the channel is
+ *	closed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Causes the callback to be called in the future when the channel will
+ *	be closed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateCloseHandler(
+    Tcl_Channel chan,		/* The channel for which to create the close
+				 * callback. */
+    Tcl_CloseProc *proc,	/* The callback routine to call when the
+				 * channel will be closed. */
+    ClientData clientData)	/* Arbitrary data to pass to the close
+				 * callback. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+    CloseCallback *cbPtr;
+
+    cbPtr = ckalloc(sizeof(CloseCallback));
+    cbPtr->proc = proc;
+    cbPtr->clientData = clientData;
+
+    cbPtr->nextPtr = statePtr->closeCbPtr;
+    statePtr->closeCbPtr = cbPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteCloseHandler --
+ *
+ *	Removes a callback that would have been called on closing the channel.
+ *	If there is no matching callback then this function has no effect.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The callback will not be called in the future when the channel is
+ *	eventually closed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteCloseHandler(
+    Tcl_Channel chan,		/* The channel for which to cancel the close
+				 * callback. */
+    Tcl_CloseProc *proc,	/* The procedure for the callback to
+				 * remove. */
+    ClientData clientData)	/* The callback data for the callback to
+				 * remove. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+    CloseCallback *cbPtr, *cbPrevPtr;
+
+    for (cbPtr = statePtr->closeCbPtr, cbPrevPtr = NULL;
+	    cbPtr != NULL; cbPtr = cbPtr->nextPtr) {
+	if ((cbPtr->proc == proc) && (cbPtr->clientData == clientData)) {
+	    if (cbPrevPtr == NULL) {
+		statePtr->closeCbPtr = cbPtr->nextPtr;
+	    } else {
+		cbPrevPtr->nextPtr = cbPtr->nextPtr;
+	    }
+	    ckfree(cbPtr);
+	    break;
+	}
+	cbPrevPtr = cbPtr;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetChannelTable --
+ *
+ *	Gets and potentially initializes the channel table for an interpreter.
+ *	If it is initializing the table it also inserts channels for stdin,
+ *	stdout and stderr if the interpreter is trusted.
+ *
+ * Results:
+ *	A pointer to the hash table created, for use by the caller.
+ *
+ * Side effects:
+ *	Initializes the channel table for an interpreter. May create channels
+ *	for stdin, stdout and stderr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashTable *
+GetChannelTable(
+    Tcl_Interp *interp)
+{
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_Channel stdinChan, stdoutChan, stderrChan;
+
+    hTblPtr = Tcl_GetAssocData(interp, "tclIO", NULL);
+    if (hTblPtr == NULL) {
+	hTblPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(hTblPtr, TCL_STRING_KEYS);
+	Tcl_SetAssocData(interp, "tclIO",
+		(Tcl_InterpDeleteProc *) DeleteChannelTable, hTblPtr);
+
+	/*
+	 * If the interpreter is trusted (not "safe"), insert channels for
+	 * stdin, stdout and stderr (possibly creating them in the process).
+	 */
+
+	if (Tcl_IsSafe(interp) == 0) {
+	    stdinChan = Tcl_GetStdChannel(TCL_STDIN);
+	    if (stdinChan != NULL) {
+		Tcl_RegisterChannel(interp, stdinChan);
+	    }
+	    stdoutChan = Tcl_GetStdChannel(TCL_STDOUT);
+	    if (stdoutChan != NULL) {
+		Tcl_RegisterChannel(interp, stdoutChan);
+	    }
+	    stderrChan = Tcl_GetStdChannel(TCL_STDERR);
+	    if (stderrChan != NULL) {
+		Tcl_RegisterChannel(interp, stderrChan);
+	    }
+	}
+    }
+    return hTblPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteChannelTable --
+ *
+ *	Deletes the channel table for an interpreter, closing any open
+ *	channels whose refcount reaches zero. This procedure is invoked when
+ *	an interpreter is deleted, via the AssocData cleanup mechanism.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the hash table of channels. May close channels. May flush
+ *	output on closed channels. Removes any channeEvent handlers that were
+ *	registered in this interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteChannelTable(
+    ClientData clientData,	/* The per-interpreter data structure. */
+    Tcl_Interp *interp)		/* The interpreter being deleted. */
+{
+    Tcl_HashTable *hTblPtr;	/* The hash table. */
+    Tcl_HashSearch hSearch;	/* Search variable. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    Channel *chanPtr;		/* Channel being deleted. */
+    ChannelState *statePtr;	/* State of Channel being deleted. */
+    EventScriptRecord *sPtr, *prevPtr, *nextPtr;
+				/* Variables to loop over all channel events
+				 * registered, to delete the ones that refer
+				 * to the interpreter being deleted. */
+
+    /*
+     * Delete all the registered channels - this will close channels whose
+     * refcount reaches zero.
+     */
+
+    hTblPtr = clientData;
+    for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch); hPtr != NULL;
+	    hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch)) {
+	chanPtr = Tcl_GetHashValue(hPtr);
+	statePtr = chanPtr->state;
+
+	/*
+	 * Remove any fileevents registered in this interpreter.
+	 */
+
+	for (sPtr = statePtr->scriptRecordPtr, prevPtr = NULL;
+		sPtr != NULL; sPtr = nextPtr) {
+	    nextPtr = sPtr->nextPtr;
+	    if (sPtr->interp == interp) {
+		if (prevPtr == NULL) {
+		    statePtr->scriptRecordPtr = nextPtr;
+		} else {
+		    prevPtr->nextPtr = nextPtr;
+		}
+
+		Tcl_DeleteChannelHandler((Tcl_Channel) chanPtr,
+			TclChannelEventScriptInvoker, sPtr);
+
+		TclDecrRefCount(sPtr->scriptPtr);
+		ckfree(sPtr);
+	    } else {
+		prevPtr = sPtr;
+	    }
+	}
+
+	/*
+	 * Cannot call Tcl_UnregisterChannel because that procedure calls
+	 * Tcl_GetAssocData to get the channel table, which might already be
+	 * inaccessible from the interpreter structure. Instead, we emulate
+	 * the behavior of Tcl_UnregisterChannel directly here.
+	 */
+
+	Tcl_DeleteHashEntry(hPtr);
+	statePtr->epoch++;
+	if (statePtr->refCount-- <= 1) {
+	    if (!GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+		(void) Tcl_Close(interp, (Tcl_Channel) chanPtr);
+	    }
+	}
+
+    }
+    Tcl_DeleteHashTable(hTblPtr);
+    ckfree(hTblPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckForStdChannelsBeingClosed --
+ *
+ *	Perform special handling for standard channels being closed. When
+ *	given a standard channel, if the refcount is now 1, it means that the
+ *	last reference to the standard channel is being explicitly closed. Now
+ *	bump the refcount artificially down to 0, to ensure the normal
+ *	handling of channels being closed will occur. Also reset the static
+ *	pointer to the channel to NULL, to avoid dangling references.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Manipulates the refcount on standard channels. May smash the global
+ *	static pointer to a standard channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CheckForStdChannelsBeingClosed(
+    Tcl_Channel chan)
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->stdinInitialized
+	    && tsdPtr->stdinChannel != NULL
+	    && statePtr == ((Channel *)tsdPtr->stdinChannel)->state) {
+	if (statePtr->refCount < 2) {
+	    statePtr->refCount = 0;
+	    tsdPtr->stdinChannel = NULL;
+	    return;
+	}
+    } else if (tsdPtr->stdoutInitialized
+	    && tsdPtr->stdoutChannel != NULL
+	    && statePtr == ((Channel *)tsdPtr->stdoutChannel)->state) {
+	if (statePtr->refCount < 2) {
+	    statePtr->refCount = 0;
+	    tsdPtr->stdoutChannel = NULL;
+	    return;
+	}
+    } else if (tsdPtr->stderrInitialized
+	    && tsdPtr->stderrChannel != NULL
+	    && statePtr == ((Channel *)tsdPtr->stderrChannel)->state) {
+	if (statePtr->refCount < 2) {
+	    statePtr->refCount = 0;
+	    tsdPtr->stderrChannel = NULL;
+	    return;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsStandardChannel --
+ *
+ *	Test if the given channel is a standard channel. No attempt is made to
+ *	check if the channel or the standard channels are initialized or
+ *	otherwise valid.
+ *
+ * Results:
+ *	Returns 1 if true, 0 if false.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsStandardChannel(
+    Tcl_Channel chan)		/* Channel to check. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if ((chan == tsdPtr->stdinChannel)
+	    || (chan == tsdPtr->stdoutChannel)
+	    || (chan == tsdPtr->stderrChannel)) {
+	return 1;
+    } else {
+	return 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegisterChannel --
+ *
+ *	Adds an already-open channel to the channel table of an interpreter.
+ *	If the interpreter passed as argument is NULL, it only increments the
+ *	channel refCount.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May increment the reference count of a channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_RegisterChannel(
+    Tcl_Interp *interp,		/* Interpreter in which to add the channel. */
+    Tcl_Channel chan)		/* The channel to add to this interpreter
+				 * channel table. */
+{
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    int isNew;			/* Is the hash entry new or does it exist? */
+    Channel *chanPtr;		/* The actual channel. */
+    ChannelState *statePtr;	/* State of the actual channel. */
+
+    /*
+     * Always (un)register bottom-most channel in the stack. This makes
+     * management of the channel list easier because no manipulation is
+     * necessary during (un)stack operation.
+     */
+
+    chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    statePtr = chanPtr->state;
+
+    if (statePtr->channelName == NULL) {
+	Tcl_Panic("Tcl_RegisterChannel: channel without name");
+    }
+    if (interp != NULL) {
+	hTblPtr = GetChannelTable(interp);
+	hPtr = Tcl_CreateHashEntry(hTblPtr, statePtr->channelName, &isNew);
+	if (!isNew) {
+	    if (chan == Tcl_GetHashValue(hPtr)) {
+		return;
+	    }
+
+	    Tcl_Panic("Tcl_RegisterChannel: duplicate channel names");
+	}
+	Tcl_SetHashValue(hPtr, chanPtr);
+    }
+    statePtr->refCount++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnregisterChannel --
+ *
+ *	Deletes the hash entry for a channel associated with an interpreter.
+ *	If the interpreter given as argument is NULL, it only decrements the
+ *	reference count. (This all happens in the Tcl_DetachChannel helper
+ *	function).
+ *
+ *	Finally, if the reference count of the channel drops to zero, it is
+ *	deleted.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Calls Tcl_DetachChannel which deletes the hash entry for a channel
+ *	associated with an interpreter.
+ *
+ *	May delete the channel, which can have a variety of consequences,
+ *	especially if we are forced to close the channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UnregisterChannel(
+    Tcl_Interp *interp,		/* Interpreter in which channel is defined. */
+    Tcl_Channel chan)		/* Channel to delete. */
+{
+    ChannelState *statePtr;	/* State of the real channel. */
+
+    statePtr = ((Channel *) chan)->state->bottomChanPtr->state;
+
+    if (GotFlag(statePtr, CHANNEL_INCLOSE)) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "illegal recursive call to close through close-handler"
+                    " of channel", -1));
+	}
+	return TCL_ERROR;
+    }
+
+    if (DetachChannel(interp, chan) != TCL_OK) {
+	return TCL_OK;
+    }
+
+    statePtr = ((Channel *) chan)->state->bottomChanPtr->state;
+
+    /*
+     * Perform special handling for standard channels being closed. If the
+     * refCount is now 1 it means that the last reference to the standard
+     * channel is being explicitly closed, so bump the refCount down
+     * artificially to 0. This will ensure that the channel is actually
+     * closed, below. Also set the static pointer to NULL for the channel.
+     */
+
+    CheckForStdChannelsBeingClosed(chan);
+
+    /*
+     * If the refCount reached zero, close the actual channel.
+     */
+
+    if (statePtr->refCount <= 0) {
+	Tcl_Preserve(statePtr);
+	if (!GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	    /*
+	     * We don't want to re-enter Tcl_Close().
+	     */
+
+	    if (!GotFlag(statePtr, CHANNEL_CLOSED)) {
+		if (Tcl_Close(interp, chan) != TCL_OK) {
+		    SetFlag(statePtr, CHANNEL_CLOSED);
+		    Tcl_Release(statePtr);
+		    return TCL_ERROR;
+		}
+	    }
+	}
+	SetFlag(statePtr, CHANNEL_CLOSED);
+	Tcl_Release(statePtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DetachChannel --
+ *
+ *	Deletes the hash entry for a channel associated with an interpreter.
+ *	If the interpreter given as argument is NULL, it only decrements the
+ *	reference count. Even if the ref count drops to zero, the channel is
+ *	NOT closed or cleaned up. This allows a channel to be detached from an
+ *	interpreter and left in the same state it was in when it was
+ *	originally returned by 'Tcl_OpenFileChannel', for example.
+ *
+ *	This function cannot be used on the standard channels, and will return
+ *	TCL_ERROR if that is attempted.
+ *
+ *	This function should only be necessary for special purposes in which
+ *	you need to generate a pristine channel from one that has already been
+ *	used. All ordinary purposes will almost always want to use
+ *	Tcl_UnregisterChannel instead.
+ *
+ *	Provided the channel is not attached to any other interpreter, it can
+ *	then be closed with Tcl_Close, rather than with Tcl_UnregisterChannel.
+ *
+ * Results:
+ *	A standard Tcl result. If the channel is not currently registered with
+ *	the given interpreter, TCL_ERROR is returned, otherwise TCL_OK.
+ *	However no error messages are left in the interp's result.
+ *
+ * Side effects:
+ *	Deletes the hash entry for a channel associated with an interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DetachChannel(
+    Tcl_Interp *interp,		/* Interpreter in which channel is defined. */
+    Tcl_Channel chan)		/* Channel to delete. */
+{
+    if (Tcl_IsStandardChannel(chan)) {
+	return TCL_ERROR;
+    }
+
+    return DetachChannel(interp, chan);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DetachChannel --
+ *
+ *	Deletes the hash entry for a channel associated with an interpreter.
+ *	If the interpreter given as argument is NULL, it only decrements the
+ *	reference count. Even if the ref count drops to zero, the channel is
+ *	NOT closed or cleaned up. This allows a channel to be detached from an
+ *	interpreter and left in the same state it was in when it was
+ *	originally returned by 'Tcl_OpenFileChannel', for example.
+ *
+ * Results:
+ *	A standard Tcl result. If the channel is not currently registered with
+ *	the given interpreter, TCL_ERROR is returned, otherwise TCL_OK.
+ *	However no error messages are left in the interp's result.
+ *
+ * Side effects:
+ *	Deletes the hash entry for a channel associated with an interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DetachChannel(
+    Tcl_Interp *interp,		/* Interpreter in which channel is defined. */
+    Tcl_Channel chan)		/* Channel to delete. */
+{
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    Channel *chanPtr;		/* The real IO channel. */
+    ChannelState *statePtr;	/* State of the real channel. */
+
+    /*
+     * Always (un)register bottom-most channel in the stack. This makes
+     * management of the channel list easier because no manipulation is
+     * necessary during (un)stack operation.
+     */
+
+    chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    statePtr = chanPtr->state;
+
+    if (interp != NULL) {
+	hTblPtr = Tcl_GetAssocData(interp, "tclIO", NULL);
+	if (hTblPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	hPtr = Tcl_FindHashEntry(hTblPtr, statePtr->channelName);
+	if (hPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if ((Channel *) Tcl_GetHashValue(hPtr) != chanPtr) {
+	    return TCL_ERROR;
+	}
+	Tcl_DeleteHashEntry(hPtr);
+	statePtr->epoch++;
+
+	/*
+	 * Remove channel handlers that refer to this interpreter, so that
+	 * they will not be present if the actual close is delayed and more
+	 * events happen on the channel. This may occur if the channel is
+	 * shared between several interpreters, or if the channel has async
+	 * flushing active.
+	 */
+
+	CleanupChannelHandlers(interp, chanPtr);
+    }
+
+    statePtr->refCount--;
+
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_GetChannel --
+ *
+ *	Finds an existing Tcl_Channel structure by name in a given
+ *	interpreter. This function is public because it is used by
+ *	channel-type-specific functions.
+ *
+ * Results:
+ *	A Tcl_Channel or NULL on failure. If failed, interp's result object
+ *	contains an error message. *modePtr is filled with the modes in which
+ *	the channel was opened.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_GetChannel(
+    Tcl_Interp *interp,		/* Interpreter in which to find or create the
+				 * channel. */
+    const char *chanName,	/* The name of the channel. */
+    int *modePtr)		/* Where to store the mode in which the
+				 * channel was opened? Will contain an ORed
+				 * combination of TCL_READABLE and
+				 * TCL_WRITABLE, if non-NULL. */
+{
+    Channel *chanPtr;		/* The actual channel. */
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    const char *name;		/* Translated name. */
+
+    /*
+     * Substitute "stdin", etc. Note that even though we immediately find the
+     * channel using Tcl_GetStdChannel, we still need to look it up in the
+     * specified interpreter to ensure that it is present in the channel
+     * table. Otherwise, safe interpreters would always have access to the
+     * standard channels.
+     */
+
+    name = chanName;
+    if ((chanName[0] == 's') && (chanName[1] == 't')) {
+	chanPtr = NULL;
+	if (strcmp(chanName, "stdin") == 0) {
+	    chanPtr = (Channel *) Tcl_GetStdChannel(TCL_STDIN);
+	} else if (strcmp(chanName, "stdout") == 0) {
+	    chanPtr = (Channel *) Tcl_GetStdChannel(TCL_STDOUT);
+	} else if (strcmp(chanName, "stderr") == 0) {
+	    chanPtr = (Channel *) Tcl_GetStdChannel(TCL_STDERR);
+	}
+	if (chanPtr != NULL) {
+	    name = chanPtr->state->channelName;
+	}
+    }
+
+    hTblPtr = GetChannelTable(interp);
+    hPtr = Tcl_FindHashEntry(hTblPtr, name);
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can not find channel named \"%s\"", chanName));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CHANNEL", chanName, NULL);
+	return NULL;
+    }
+
+    /*
+     * Always return bottom-most channel in the stack. This one lives the
+     * longest - other channels may go away unnoticed. The other APIs
+     * compensate where necessary to retrieve the topmost channel again.
+     */
+
+    chanPtr = Tcl_GetHashValue(hPtr);
+    chanPtr = chanPtr->state->bottomChanPtr;
+    if (modePtr != NULL) {
+	*modePtr = chanPtr->state->flags & (TCL_READABLE|TCL_WRITABLE);
+    }
+
+    return (Tcl_Channel) chanPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetChannelFromObj --
+ *
+ *	Finds an existing Tcl_Channel structure by name in a given
+ *	interpreter. This function is public because it is used by
+ *	channel-type-specific functions.
+ *
+ * Results:
+ *	A Tcl_Channel or NULL on failure. If failed, interp's result object
+ *	contains an error message. *modePtr is filled with the modes in which
+ *	the channel was opened.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclGetChannelFromObj(
+    Tcl_Interp *interp,		/* Interpreter in which to find or create the
+				 * channel. */
+    Tcl_Obj *objPtr,
+    Tcl_Channel *channelPtr,
+    int *modePtr,		/* Where to store the mode in which the
+				 * channel was opened? Will contain an ORed
+				 * combination of TCL_READABLE and
+				 * TCL_WRITABLE, if non-NULL. */
+    int flags)
+{
+    ChannelState *statePtr;
+    ResolvedChanName *resPtr = NULL;
+    Tcl_Channel chan;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (objPtr->typePtr == &chanObjType) {
+	/*
+ 	 * Confirm validity of saved lookup results.
+ 	 */
+
+	resPtr = (ResolvedChanName *) objPtr->internalRep.twoPtrValue.ptr1;
+	statePtr = resPtr->statePtr;
+	if ((resPtr->interp == interp)		/* Same interp context */
+			/* No epoch change in channel since lookup */
+		&& (resPtr->epoch == statePtr->epoch)) {
+	    /*
+             * Have a valid saved lookup. Jump to end to return it.
+             */
+
+	    goto valid;
+	}
+    }
+
+    chan = Tcl_GetChannel(interp, TclGetString(objPtr), NULL);
+
+    if (chan == NULL) {
+	if (resPtr) {
+	    FreeChannelIntRep(objPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    if (resPtr && resPtr->refCount == 1) {
+	/*
+         * Re-use the ResolvedCmdName struct.
+         */
+
+	Tcl_Release((ClientData) resPtr->statePtr);
+
+    } else {
+	TclFreeIntRep(objPtr);
+
+	resPtr = (ResolvedChanName *) ckalloc(sizeof(ResolvedChanName));
+	resPtr->refCount = 1;
+	objPtr->internalRep.twoPtrValue.ptr1 = (ClientData) resPtr;
+	objPtr->typePtr = &chanObjType;
+    }
+    statePtr = ((Channel *)chan)->state;
+    resPtr->statePtr = statePtr;
+    Tcl_Preserve((ClientData) statePtr);
+    resPtr->interp = interp;
+    resPtr->epoch = statePtr->epoch;
+
+  valid:
+    *channelPtr = (Tcl_Channel) statePtr->bottomChanPtr;
+
+    if (modePtr != NULL) {
+	*modePtr = statePtr->flags & (TCL_READABLE|TCL_WRITABLE);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateChannel --
+ *
+ *	Creates a new entry in the hash table for a Tcl_Channel record.
+ *
+ * Results:
+ *	Returns the new Tcl_Channel.
+ *
+ * Side effects:
+ *	Creates a new Tcl_Channel instance and inserts it into the hash table.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_CreateChannel(
+    const Tcl_ChannelType *typePtr, /* The channel type record. */
+    const char *chanName,	/* Name of channel to record. */
+    ClientData instanceData,	/* Instance specific data. */
+    int mask)			/* TCL_READABLE & TCL_WRITABLE to indicate if
+				 * the channel is readable, writable. */
+{
+    Channel *chanPtr;		/* The channel structure newly created. */
+    ChannelState *statePtr;	/* The stack-level independent state info for
+				 * the channel. */
+    const char *name;
+    char *tmp;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    /*
+     * With the change of the Tcl_ChannelType structure to use a version in
+     * 8.3.2+, we have to make sure that our assumption that the structure
+     * remains a binary compatible size is true.
+     *
+     * If this assertion fails on some system, then it can be removed only if
+     * the user recompiles code with older channel drivers in the new system
+     * as well.
+     */
+
+    assert(sizeof(Tcl_ChannelTypeVersion) == sizeof(Tcl_DriverBlockModeProc *));
+    assert(typePtr->typeName != NULL);
+    if (NULL == typePtr->closeProc) {
+	Tcl_Panic("channel type %s must define closeProc", typePtr->typeName);
+    }
+    if ((TCL_READABLE & mask) && (NULL == typePtr->inputProc)) {
+	Tcl_Panic("channel type %s must define inputProc when used for reader channel", typePtr->typeName);
+    }
+    if ((TCL_WRITABLE & mask) &&  (NULL == typePtr->outputProc)) {
+	Tcl_Panic("channel type %s must define outputProc when used for writer channel", typePtr->typeName);
+    }
+    if (NULL == typePtr->watchProc) {
+	Tcl_Panic("channel type %s must define watchProc", typePtr->typeName);
+    }
+    if ((NULL!=typePtr->wideSeekProc) && (NULL == typePtr->seekProc)) {
+	Tcl_Panic("channel type %s must define seekProc if defining wideSeekProc", typePtr->typeName);
+    }
+
+    /*
+     * JH: We could subsequently memset these to 0 to avoid the numerous
+     * assignments to 0/NULL below.
+     */
+
+    chanPtr = ckalloc(sizeof(Channel));
+    statePtr = ckalloc(sizeof(ChannelState));
+    chanPtr->state = statePtr;
+
+    chanPtr->instanceData = instanceData;
+    chanPtr->typePtr = typePtr;
+
+    /*
+     * Set all the bits that are part of the stack-independent state
+     * information for the channel.
+     */
+
+    if (chanName != NULL) {
+	unsigned len = strlen(chanName) + 1;
+
+	/*
+         * Make sure we allocate at least 7 bytes, so it fits for "stdout"
+         * later.
+         */
+
+	tmp = ckalloc((len < 7) ? 7 : len);
+	strcpy(tmp, chanName);
+    } else {
+	tmp = ckalloc(7);
+	tmp[0] = '\0';
+    }
+    statePtr->channelName = tmp;
+    statePtr->flags = mask;
+
+    /*
+     * Set the channel to system default encoding.
+     *
+     * Note the strange bit of protection taking place here. If the system
+     * encoding name is reported back as "binary", something weird is
+     * happening. Tcl provides no "binary" encoding, so someone else has
+     * provided one. We ignore it so as not to interfere with the "magic"
+     * interpretation that Tcl_Channels give to the "-encoding binary" option.
+     */
+
+    statePtr->encoding = NULL;
+    name = Tcl_GetEncodingName(NULL);
+    if (strcmp(name, "binary") != 0) {
+	statePtr->encoding = Tcl_GetEncoding(NULL, name);
+    }
+    statePtr->inputEncodingState  = NULL;
+    statePtr->inputEncodingFlags  = TCL_ENCODING_START;
+    statePtr->outputEncodingState = NULL;
+    statePtr->outputEncodingFlags = TCL_ENCODING_START;
+
+    /*
+     * Set the channel up initially in AUTO input translation mode to accept
+     * "\n", "\r" and "\r\n". Output translation mode is set to a platform
+     * specific default value. The eofChar is set to 0 for both input and
+     * output, so that Tcl does not look for an in-file EOF indicator (e.g.,
+     * ^Z) and does not append an EOF indicator to files.
+     */
+
+    statePtr->inputTranslation	= TCL_TRANSLATE_AUTO;
+    statePtr->outputTranslation	= TCL_PLATFORM_TRANSLATION;
+    statePtr->inEofChar		= 0;
+    statePtr->outEofChar	= 0;
+
+    statePtr->unreportedError	= 0;
+    statePtr->refCount		= 0;
+    statePtr->closeCbPtr	= NULL;
+    statePtr->curOutPtr		= NULL;
+    statePtr->outQueueHead	= NULL;
+    statePtr->outQueueTail	= NULL;
+    statePtr->saveInBufPtr	= NULL;
+    statePtr->inQueueHead	= NULL;
+    statePtr->inQueueTail	= NULL;
+    statePtr->chPtr		= NULL;
+    statePtr->interestMask	= 0;
+    statePtr->scriptRecordPtr	= NULL;
+    statePtr->bufSize		= CHANNELBUFFER_DEFAULT_SIZE;
+    statePtr->timer		= NULL;
+    statePtr->csPtrR		= NULL;
+    statePtr->csPtrW		= NULL;
+    statePtr->outputStage	= NULL;
+
+    /*
+     * As we are creating the channel, it is obviously the top for now.
+     */
+
+    statePtr->topChanPtr	= chanPtr;
+    statePtr->bottomChanPtr	= chanPtr;
+    chanPtr->downChanPtr	= NULL;
+    chanPtr->upChanPtr		= NULL;
+    chanPtr->inQueueHead	= NULL;
+    chanPtr->inQueueTail	= NULL;
+    chanPtr->refCount		= 0;
+
+    /*
+     * TIP #219, Tcl Channel Reflection API
+     */
+
+    statePtr->chanMsg		= NULL;
+    statePtr->unreportedMsg	= NULL;
+
+    statePtr->epoch		= 0;
+
+    /*
+     * Link the channel into the list of all channels; create an on-exit
+     * handler if there is not one already, to close off all the channels in
+     * the list on exit.
+     *
+     * JH: Could call Tcl_SpliceChannel, but need to avoid NULL check.
+     *
+     * TIP #218.
+     * AK: Just initialize the field to NULL before invoking Tcl_SpliceChannel
+     *	   We need Tcl_SpliceChannel, for the threadAction calls. There is no
+     *	   real reason to duplicate all of this.
+     * NOTE: All drivers using thread actions now have to perform their TSD
+     *	     manipulation only in their thread action proc. Doing it when
+     *	     creating their instance structures will collide with the thread
+     *	     action activity and lead to damaged lists.
+     */
+
+    statePtr->nextCSPtr = NULL;
+    SpliceChannel((Tcl_Channel) chanPtr);
+
+    /*
+     * Install this channel in the first empty standard channel slot, if the
+     * channel was previously closed explicitly.
+     */
+
+    if ((tsdPtr->stdinChannel == NULL) && (tsdPtr->stdinInitialized == 1)) {
+	strcpy(tmp, "stdin");
+	Tcl_SetStdChannel((Tcl_Channel) chanPtr, TCL_STDIN);
+	Tcl_RegisterChannel(NULL, (Tcl_Channel) chanPtr);
+    } else if ((tsdPtr->stdoutChannel == NULL) &&
+	    (tsdPtr->stdoutInitialized == 1)) {
+	strcpy(tmp, "stdout");
+	Tcl_SetStdChannel((Tcl_Channel) chanPtr, TCL_STDOUT);
+	Tcl_RegisterChannel(NULL, (Tcl_Channel) chanPtr);
+    } else if ((tsdPtr->stderrChannel == NULL) &&
+	    (tsdPtr->stderrInitialized == 1)) {
+	strcpy(tmp, "stderr");
+	Tcl_SetStdChannel((Tcl_Channel) chanPtr, TCL_STDERR);
+	Tcl_RegisterChannel(NULL, (Tcl_Channel) chanPtr);
+    }
+    return (Tcl_Channel) chanPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_StackChannel --
+ *
+ *	Replaces an entry in the hash table for a Tcl_Channel record. The
+ *	replacement is a new channel with same name, it supercedes the
+ *	replaced channel. Input and output of the superceded channel is now
+ *	going through the newly created channel and allows the arbitrary
+ *	filtering/manipulation of the dataflow.
+ *
+ *	Andreas Kupries <a.kupries@westend.com>, 12/13/1998 "Trf-Patch for
+ *	filtering channels"
+ *
+ * Results:
+ *	Returns the new Tcl_Channel, which actually contains the saved
+ *	information about prevChan.
+ *
+ * Side effects:
+ *	A new channel structure is allocated and linked below the existing
+ *	channel. The channel operations and client data of the existing
+ *	channel are copied down to the newly created channel, and the current
+ *	channel has its operations replaced by the new typePtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_StackChannel(
+    Tcl_Interp *interp,		/* The interpreter we are working in */
+    const Tcl_ChannelType *typePtr,
+				/* The channel type record for the new
+				 * channel. */
+    ClientData instanceData,	/* Instance specific data for the new
+				 * channel. */
+    int mask,			/* TCL_READABLE & TCL_WRITABLE to indicate if
+				 * the channel is readable, writable. */
+    Tcl_Channel prevChan)	/* The channel structure to replace */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    Channel *chanPtr, *prevChanPtr;
+    ChannelState *statePtr;
+
+    /*
+     * Find the given channel (prevChan) in the list of all channels. If we do
+     * not find it, then it was never registered correctly.
+     *
+     * This operation should occur at the top of a channel stack.
+     */
+
+    statePtr = (ChannelState *) tsdPtr->firstCSPtr;
+    prevChanPtr = ((Channel *) prevChan)->state->topChanPtr;
+
+    while ((statePtr != NULL) && (statePtr->topChanPtr != prevChanPtr)) {
+	statePtr = statePtr->nextCSPtr;
+    }
+
+    if (statePtr == NULL) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "couldn't find state for channel \"%s\"",
+		    Tcl_GetChannelName(prevChan)));
+	}
+	return NULL;
+    }
+
+    /*
+     * Here we check if the given "mask" matches the "flags" of the already
+     * existing channel.
+     *
+     *	  | - | R | W | RW |
+     *	--+---+---+---+----+	<=>  0 != (chan->mask & prevChan->mask)
+     *	- |   |   |   |    |
+     *	R |   | + |   | +  |	The superceding channel is allowed to restrict
+     *	W |   |   | + | +  |	the capabilities of the superceded one!
+     *	RW|   | + | + | +  |
+     *	--+---+---+---+----+
+     */
+
+    if ((mask & (statePtr->flags & (TCL_READABLE | TCL_WRITABLE))) == 0) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "reading and writing both disallowed for channel \"%s\"",
+		    Tcl_GetChannelName(prevChan)));
+	}
+	return NULL;
+    }
+
+    /*
+     * Flush the buffers. This ensures that any data still in them at this
+     * time is not handled by the new transformation. Restrict this to
+     * writable channels. Take care to hide a possible bg-copy in progress
+     * from Tcl_Flush and the CheckForChannelErrors inside.
+     */
+
+    if ((mask & TCL_WRITABLE) != 0) {
+	CopyState *csPtrR = statePtr->csPtrR;
+	CopyState *csPtrW = statePtr->csPtrW;
+
+	statePtr->csPtrR = NULL;
+	statePtr->csPtrW = NULL;
+
+	/*
+	 * TODO: Examine what can go wrong if Tcl_Flush() call disturbs
+	 * the stacking state of this channel during its operations.
+	 */
+	if (Tcl_Flush((Tcl_Channel) prevChanPtr) != TCL_OK) {
+	    statePtr->csPtrR = csPtrR;
+	    statePtr->csPtrW = csPtrW;
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "could not flush channel \"%s\"",
+			Tcl_GetChannelName(prevChan)));
+	    }
+	    return NULL;
+	}
+
+	statePtr->csPtrR = csPtrR;
+	statePtr->csPtrW = csPtrW;
+    }
+
+    /*
+     * Discard any input in the buffers. They are not yet read by the user of
+     * the channel, so they have to go through the new transformation before
+     * reading. As the buffers contain the untransformed form their contents
+     * are not only useless but actually distorts our view of the system.
+     *
+     * To preserve the information without having to read them again and to
+     * avoid problems with the location in the channel (seeking might be
+     * impossible) we move the buffers from the common state structure into
+     * the channel itself. We use the buffers in the channel below the new
+     * transformation to hold the data. In the future this allows us to write
+     * transformations which pre-read data and push the unused part back when
+     * they are going away.
+     */
+
+    if (((mask & TCL_READABLE) != 0) && (statePtr->inQueueHead != NULL)) {
+	/*
+	 * When statePtr->inQueueHead is not NULL, we know
+	 * prevChanPtr->inQueueHead must be NULL.
+	 */
+
+	assert(prevChanPtr->inQueueHead == NULL);
+	assert(prevChanPtr->inQueueTail == NULL);
+
+	prevChanPtr->inQueueHead = statePtr->inQueueHead;
+	prevChanPtr->inQueueTail = statePtr->inQueueTail;
+
+	statePtr->inQueueHead = NULL;
+	statePtr->inQueueTail = NULL;
+    }
+
+    chanPtr = ckalloc(sizeof(Channel));
+
+    /*
+     * Save some of the current state into the new structure, reinitialize the
+     * parts which will stay with the transformation.
+     *
+     * Remarks:
+     */
+
+    chanPtr->state		= statePtr;
+    chanPtr->instanceData	= instanceData;
+    chanPtr->typePtr		= typePtr;
+    chanPtr->downChanPtr	= prevChanPtr;
+    chanPtr->upChanPtr		= NULL;
+    chanPtr->inQueueHead	= NULL;
+    chanPtr->inQueueTail	= NULL;
+    chanPtr->refCount		= 0;
+
+    /*
+     * Place new block at the head of a possibly existing list of previously
+     * stacked channels.
+     */
+
+    prevChanPtr->upChanPtr	= chanPtr;
+    statePtr->topChanPtr	= chanPtr;
+
+    /*
+     * TIP #218, Channel Thread Actions.
+     *
+     * We call the thread actions for the new channel directly. We _cannot_
+     * use SpliceChannel, because the (thread-)global list of all channels
+     * always contains the _ChannelState_ for a stack of channels, not the
+     * individual channels. And SpliceChannel would not only call the thread
+     * actions, but also add the shared ChannelState to this list a second
+     * time, mangling it.
+     */
+
+    ChanThreadAction(chanPtr, TCL_CHANNEL_THREAD_INSERT);
+
+    return (Tcl_Channel) chanPtr;
+}
+
+void
+TclChannelPreserve(
+    Tcl_Channel chan)
+{
+    ((Channel *)chan)->refCount++;
+}
+
+void
+TclChannelRelease(
+    Tcl_Channel chan)
+{
+    Channel *chanPtr = (Channel *) chan;
+
+    if (chanPtr->refCount == 0) {
+	Tcl_Panic("Channel released more than preserved");
+    }
+    if (--chanPtr->refCount) {
+	return;
+    }
+    if (chanPtr->typePtr == NULL) {
+	ckfree(chanPtr);
+    }
+}
+
+static void
+ChannelFree(
+    Channel *chanPtr)
+{
+    if (chanPtr->refCount == 0) {
+	ckfree(chanPtr);
+	return;
+    }
+    chanPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnstackChannel --
+ *
+ *	Unstacks an entry in the hash table for a Tcl_Channel record. This is
+ *	the reverse to 'Tcl_StackChannel'.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	If TCL_ERROR is returned, the posix error code will be set with
+ *	Tcl_SetErrno. May leave a message in interp result as well.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UnstackChannel(
+    Tcl_Interp *interp,		/* The interpreter we are working in */
+    Tcl_Channel chan)		/* The channel to unstack */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+    int result = 0;
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    if (chanPtr->downChanPtr != NULL) {
+	/*
+	 * Instead of manipulating the per-thread / per-interp list/hashtable
+	 * of registered channels we wind down the state of the
+	 * transformation, and then restore the state of underlying channel
+	 * into the old structure.
+	 *
+	 * TODO: Figure out how to handle the situation where the chan
+	 * operations called below by this unstacking operation cause
+	 * another unstacking recursively.  In that case the downChanPtr
+	 * value we're holding on to will not be the right thing.
+	 */
+
+	Channel *downChanPtr = chanPtr->downChanPtr;
+
+	/*
+	 * Flush the buffers. This ensures that any data still in them at this
+	 * time _is_ handled by the transformation we are unstacking right
+	 * now. Restrict this to writable channels. Take care to hide a
+	 * possible bg-copy in progress from Tcl_Flush and the
+	 * CheckForChannelErrors inside.
+	 */
+
+	if (GotFlag(statePtr, TCL_WRITABLE)) {
+	    CopyState *csPtrR = statePtr->csPtrR;
+	    CopyState *csPtrW = statePtr->csPtrW;
+
+	    statePtr->csPtrR = NULL;
+	    statePtr->csPtrW = NULL;
+
+	    if (Tcl_Flush((Tcl_Channel) chanPtr) != TCL_OK) {
+		statePtr->csPtrR = csPtrR;
+		statePtr->csPtrW = csPtrW;
+
+		/*
+		 * TIP #219, Tcl Channel Reflection API.
+		 * Move error messages put by the driver into the chan/ip
+		 * bypass area into the regular interpreter result. Fall back
+		 * to the regular message if nothing was found in the
+		 * bypasses.
+		 */
+
+		if (!TclChanCaughtErrorBypass(interp, chan) && interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                            "could not flush channel \"%s\"",
+			    Tcl_GetChannelName((Tcl_Channel) chanPtr)));
+		}
+		return TCL_ERROR;
+	    }
+
+	    statePtr->csPtrR  = csPtrR;
+	    statePtr->csPtrW = csPtrW;
+	}
+
+	/*
+	 * Anything in the input queue and the push-back buffers of the
+	 * transformation going away is transformed data, but not yet read. As
+	 * unstacking means that the caller does not want to see transformed
+	 * data any more we have to discard these bytes. To avoid writing an
+	 * analogue to 'DiscardInputQueued' we move the information in the
+	 * push back buffers to the input queue and then call
+	 * 'DiscardInputQueued' on that.
+	 */
+
+	if (GotFlag(statePtr, TCL_READABLE) &&
+		((statePtr->inQueueHead != NULL) ||
+		(chanPtr->inQueueHead != NULL))) {
+	    if ((statePtr->inQueueHead != NULL) &&
+		    (chanPtr->inQueueHead != NULL)) {
+		statePtr->inQueueTail->nextPtr = chanPtr->inQueueHead;
+		statePtr->inQueueTail = chanPtr->inQueueTail;
+		statePtr->inQueueHead = statePtr->inQueueTail;
+	    } else if (chanPtr->inQueueHead != NULL) {
+		statePtr->inQueueHead = chanPtr->inQueueHead;
+		statePtr->inQueueTail = chanPtr->inQueueTail;
+	    }
+
+	    chanPtr->inQueueHead = NULL;
+	    chanPtr->inQueueTail = NULL;
+
+	    DiscardInputQueued(statePtr, 0);
+	}
+
+	/*
+	 * TIP #218, Channel Thread Actions.
+	 *
+	 * We call the thread actions for the new channel directly. We
+	 * _cannot_ use CutChannel, because the (thread-)global list of all
+	 * channels always contains the _ChannelState_ for a stack of
+	 * channels, not the individual channels. And SpliceChannel would not
+	 * only call the thread actions, but also remove the shared
+	 * ChannelState from this list despite there being more channels for
+	 * the state which are still active.
+	 */
+
+	ChanThreadAction(chanPtr, TCL_CHANNEL_THREAD_REMOVE);
+
+	statePtr->topChanPtr = downChanPtr;
+	downChanPtr->upChanPtr = NULL;
+
+	/*
+	 * Leave this link intact for closeproc
+	 *  chanPtr->downChanPtr = NULL;
+	 */
+
+	/*
+	 * Close and free the channel driver state.
+	 */
+
+	result = ChanClose(chanPtr, interp);
+	ChannelFree(chanPtr);
+
+	UpdateInterest(statePtr->topChanPtr);
+
+	if (result != 0) {
+	    Tcl_SetErrno(result);
+
+	    /*
+	     * TIP #219, Tcl Channel Reflection API.
+	     * Move error messages put by the driver into the chan/ip bypass
+	     * area into the regular interpreter result.
+	     */
+
+	    TclChanCaughtErrorBypass(interp, chan);
+	    return TCL_ERROR;
+	}
+    } else {
+	/*
+	 * This channel does not cover another one. Simply do a close, if
+	 * necessary.
+	 */
+
+	if (statePtr->refCount <= 0) {
+	    if (Tcl_Close(interp, chan) != TCL_OK) {
+		/*
+		 * TIP #219, Tcl Channel Reflection API.
+		 * "TclChanCaughtErrorBypass" is not required here, it was
+		 * done already by "Tcl_Close".
+		 */
+
+		return TCL_ERROR;
+	    }
+	}
+
+	/*
+	 * TIP #218, Channel Thread Actions.
+	 * Not required in this branch, this is done by Tcl_Close. If
+	 * Tcl_Close is not called then the ChannelState is still active in
+	 * the thread and no action has to be taken either.
+	 */
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetStackedChannel --
+ *
+ *	Determines whether the specified channel is stacked upon another.
+ *
+ * Results:
+ *	NULL if the channel is not stacked upon another one, or a reference to
+ *	the channel it is stacked upon. This reference can be used in queries,
+ *	but modification is not allowed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_GetStackedChannel(
+    Tcl_Channel chan)
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+
+    return (Tcl_Channel) chanPtr->downChanPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetTopChannel --
+ *
+ *	Returns the top channel of a channel stack.
+ *
+ * Results:
+ *	NULL if the channel is not stacked upon another one, or a reference to
+ *	the channel it is stacked upon. This reference can be used in queries,
+ *	but modification is not allowed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_GetTopChannel(
+    Tcl_Channel chan)
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+
+    return (Tcl_Channel) chanPtr->state->topChanPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelInstanceData --
+ *
+ *	Returns the client data associated with a channel.
+ *
+ * Results:
+ *	The client data.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_GetChannelInstanceData(
+    Tcl_Channel chan)		/* Channel for which to return client data. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+
+    return chanPtr->instanceData;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelThread --
+ *
+ *	Given a channel structure, returns the thread managing it. TIP #10
+ *
+ * Results:
+ *	Returns the id of the thread managing the channel.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_ThreadId
+Tcl_GetChannelThread(
+    Tcl_Channel chan)		/* The channel to return the managing thread
+				 * for. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+
+    return chanPtr->state->managingThread;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelType --
+ *
+ *	Given a channel structure, returns the channel type structure.
+ *
+ * Results:
+ *	Returns a pointer to the channel type structure.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const Tcl_ChannelType *
+Tcl_GetChannelType(
+    Tcl_Channel chan)		/* The channel to return type for. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+
+    return chanPtr->typePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelMode --
+ *
+ *	Computes a mask indicating whether the channel is open for reading and
+ *	writing.
+ *
+ * Results:
+ *	An OR-ed combination of TCL_READABLE and TCL_WRITABLE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelMode(
+    Tcl_Channel chan)		/* The channel for which the mode is being
+				 * computed. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of actual channel. */
+
+    return (statePtr->flags & (TCL_READABLE | TCL_WRITABLE));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelName --
+ *
+ *	Returns the string identifying the channel name.
+ *
+ * Results:
+ *	The string containing the channel name. This memory is owned by the
+ *	generic layer and should not be modified by the caller.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetChannelName(
+    Tcl_Channel chan)		/* The channel for which to return the name. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of actual channel. */
+
+    return statePtr->channelName;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelHandle --
+ *
+ *	Returns an OS handle associated with a channel.
+ *
+ * Results:
+ *	Returns TCL_OK and places the handle in handlePtr, or returns
+ *	TCL_ERROR on failure.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelHandle(
+    Tcl_Channel chan,		/* The channel to get file from. */
+    int direction,		/* TCL_WRITABLE or TCL_READABLE. */
+    ClientData *handlePtr)	/* Where to store handle */
+{
+    Channel *chanPtr;		/* The actual channel. */
+    ClientData handle;
+    int result;
+
+    chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    if (!chanPtr->typePtr->getHandleProc) {
+        Tcl_SetChannelError(chan, Tcl_ObjPrintf(
+                "channel \"%s\" does not support OS handles",
+                Tcl_GetChannelName(chan)));
+	return TCL_ERROR;
+    }
+    result = chanPtr->typePtr->getHandleProc(chanPtr->instanceData, direction,
+	    &handle);
+    if (handlePtr) {
+	*handlePtr = handle;
+    }
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * AllocChannelBuffer --
+ *
+ *	A channel buffer has BUFFER_PADDING bytes extra at beginning to hold
+ *	any bytes of a native-encoding character that got split by the end of
+ *	the previous buffer and need to be moved to the beginning of the next
+ *	buffer to make a contiguous string so it can be converted to UTF-8.
+ *
+ *	A channel buffer has BUFFER_PADDING bytes extra at the end to hold any
+ *	bytes of a native-encoding character (generated from a UTF-8
+ *	character) that overflow past the end of the buffer and need to be
+ *	moved to the next buffer.
+ *
+ * Results:
+ *	A newly allocated channel buffer.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static ChannelBuffer *
+AllocChannelBuffer(
+    int length)			/* Desired length of channel buffer. */
+{
+    ChannelBuffer *bufPtr;
+    int n;
+
+    n = length + CHANNELBUFFER_HEADER_SIZE + BUFFER_PADDING + BUFFER_PADDING;
+    bufPtr = ckalloc(n);
+    bufPtr->nextAdded	= BUFFER_PADDING;
+    bufPtr->nextRemoved	= BUFFER_PADDING;
+    bufPtr->bufLength	= length + BUFFER_PADDING;
+    bufPtr->nextPtr	= NULL;
+    bufPtr->refCount	= 1;
+    return bufPtr;
+}
+
+static void
+PreserveChannelBuffer(
+    ChannelBuffer *bufPtr)
+{
+    if (bufPtr->refCount == 0) {
+	Tcl_Panic("Reuse of ChannelBuffer! %p", bufPtr);
+    }
+    bufPtr->refCount++;
+}
+
+static void
+ReleaseChannelBuffer(
+    ChannelBuffer *bufPtr)
+{
+    if (--bufPtr->refCount) {
+	return;
+    }
+    ckfree(bufPtr);
+}
+
+static int
+IsShared(
+    ChannelBuffer *bufPtr)
+{
+    return bufPtr->refCount > 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RecycleBuffer --
+ *
+ *	Helper function to recycle input and output buffers. Ensures that two
+ *	input buffers are saved (one in the input queue and another in the
+ *	saveInBufPtr field) and that curOutPtr is set to a buffer. Only if
+ *	these conditions are met is the buffer freed to the OS.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May free a buffer to the OS.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RecycleBuffer(
+    ChannelState *statePtr,	/* ChannelState in which to recycle buffers. */
+    ChannelBuffer *bufPtr,	/* The buffer to recycle. */
+    int mustDiscard)		/* If nonzero, free the buffer to the OS,
+				 * always. */
+{
+    /*
+     * Do we have to free the buffer to the OS?
+     */
+
+    if (IsShared(bufPtr)) {
+	mustDiscard = 1;
+    }
+
+    if (mustDiscard) {
+	ReleaseChannelBuffer(bufPtr);
+	return;
+    }
+
+    /*
+     * Only save buffers which have the requested buffersize for the channel.
+     * This is to honor dynamic changes of the buffersize made by the user.
+     */
+
+    if ((bufPtr->bufLength - BUFFER_PADDING) != statePtr->bufSize) {
+	ReleaseChannelBuffer(bufPtr);
+	return;
+    }
+
+    /*
+     * Only save buffers for the input queue if the channel is readable.
+     */
+
+    if (GotFlag(statePtr, TCL_READABLE)) {
+	if (statePtr->inQueueHead == NULL) {
+	    statePtr->inQueueHead = bufPtr;
+	    statePtr->inQueueTail = bufPtr;
+	    goto keepBuffer;
+	}
+	if (statePtr->saveInBufPtr == NULL) {
+	    statePtr->saveInBufPtr = bufPtr;
+	    goto keepBuffer;
+	}
+    }
+
+    /*
+     * Only save buffers for the output queue if the channel is writable.
+     */
+
+    if (GotFlag(statePtr, TCL_WRITABLE)) {
+	if (statePtr->curOutPtr == NULL) {
+	    statePtr->curOutPtr = bufPtr;
+	    goto keepBuffer;
+	}
+    }
+
+    /*
+     * If we reached this code we return the buffer to the OS.
+     */
+
+    ReleaseChannelBuffer(bufPtr);
+    return;
+
+  keepBuffer:
+    bufPtr->nextRemoved = BUFFER_PADDING;
+    bufPtr->nextAdded = BUFFER_PADDING;
+    bufPtr->nextPtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DiscardOutputQueued --
+ *
+ *	Discards all output queued in the output queue of a channel.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Recycles buffers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DiscardOutputQueued(
+    ChannelState *statePtr)	/* ChannelState for which to discard output. */
+{
+    ChannelBuffer *bufPtr;
+
+    while (statePtr->outQueueHead != NULL) {
+	bufPtr = statePtr->outQueueHead;
+	statePtr->outQueueHead = bufPtr->nextPtr;
+	RecycleBuffer(statePtr, bufPtr, 0);
+    }
+    statePtr->outQueueHead = NULL;
+    statePtr->outQueueTail = NULL;
+    bufPtr = statePtr->curOutPtr;
+    if (bufPtr && BytesLeft(bufPtr)) {
+	statePtr->curOutPtr = NULL;
+	RecycleBuffer(statePtr, bufPtr, 0);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckForDeadChannel --
+ *
+ *	This function checks is a given channel is Dead (a channel that has
+ *	been closed but not yet deallocated.)
+ *
+ * Results:
+ *	True (1) if channel is Dead, False (0) if channel is Ok
+ *
+ * Side effects:
+ *	None
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckForDeadChannel(
+    Tcl_Interp *interp,		/* For error reporting (can be NULL) */
+    ChannelState *statePtr)	/* The channel state to check. */
+{
+    if (!GotFlag(statePtr, CHANNEL_DEAD)) {
+	return 0;
+    }
+
+    Tcl_SetErrno(EINVAL);
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                "unable to access channel: invalid channel", -1));
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FlushChannel --
+ *
+ *	This function flushes as much of the queued output as is possible now.
+ *	If calledFromAsyncFlush is nonzero, it is being called in an event
+ *	handler to flush channel output asynchronously.
+ *
+ * Results:
+ *	0 if successful, else the error code that was returned by the channel
+ *	type operation. May leave a message in the interp result.
+ *
+ * Side effects:
+ *	May produce output on a channel. May block indefinitely if the channel
+ *	is synchronous. May schedule an async flush on the channel. May
+ *	recycle memory for buffers in the output queue.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+FlushChannel(
+    Tcl_Interp *interp,		/* For error reporting during close. */
+    Channel *chanPtr,		/* The channel to flush on. */
+    int calledFromAsyncFlush)	/* If nonzero then we are being called from an
+				 * asynchronous flush callback. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State of the channel stack. */
+    ChannelBuffer *bufPtr;	/* Iterates over buffered output queue. */
+    int written;		/* Amount of output data actually written in
+				 * current round. */
+    int errorCode = 0;		/* Stores POSIX error codes from channel
+				 * driver operations. */
+    int wroteSome = 0;		/* Set to one if any data was written to the
+				 * driver. */
+
+    /*
+     * Prevent writing on a dead channel -- a channel that has been closed but
+     * not yet deallocated. This can occur if the exit handler for the channel
+     * deallocation runs before all channels are deregistered in all
+     * interpreters.
+     */
+
+    if (CheckForDeadChannel(interp, statePtr)) {
+	return -1;
+    }
+
+    /*
+     * Should we shift the current output buffer over to the output queue?
+     * First check that there are bytes in it.  If so then...
+     *
+     * If the output queue is empty, then yes, trusting the caller called us
+     * only when written bytes ought to be flushed.
+     *
+     * If the current output buffer is full, then yes, so we can meet the
+     * post-condition that on a successful return to caller we've left space
+     * in the current output buffer for more writing (the flush call was to
+     * make new room).
+     *
+     * If the channel is blocking, then yes, so we guarantee that blocking
+     * flushes actually flush all pending data.
+     *
+     * Otherwise, no.  Keep the current output buffer where it is so more
+     * can be written to it, possibly filling it, to promote more efficient
+     * buffer usage.
+     */
+
+    bufPtr = statePtr->curOutPtr;
+    if (bufPtr && BytesLeft(bufPtr) && /* Keep empties off queue */
+	    (statePtr->outQueueHead == NULL || IsBufferFull(bufPtr)
+		    || !GotFlag(statePtr, CHANNEL_NONBLOCKING))) {
+	if (statePtr->outQueueHead == NULL) {
+	    statePtr->outQueueHead = bufPtr;
+	} else {
+	    statePtr->outQueueTail->nextPtr = bufPtr;
+	}
+	statePtr->outQueueTail = bufPtr;
+	statePtr->curOutPtr = NULL;
+    }
+
+    assert(!IsBufferFull(statePtr->curOutPtr));
+
+    /*
+     * If we are not being called from an async flush and an async flush
+     * is active, we just return without producing any output.
+     */
+
+    if (!calledFromAsyncFlush && GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	return 0;
+    }
+
+    /*
+     * Loop over the queued buffers and attempt to flush as much as possible
+     * of the queued output to the channel.
+     */
+
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+    while (statePtr->outQueueHead) {
+	bufPtr = statePtr->outQueueHead;
+
+	/*
+	 * Produce the output on the channel.
+	 */
+
+	PreserveChannelBuffer(bufPtr);
+	written = ChanWrite(chanPtr, RemovePoint(bufPtr), BytesLeft(bufPtr),
+		&errorCode);
+
+	/*
+	 * If the write failed completely attempt to start the asynchronous
+	 * flush mechanism and break out of this loop - do not attempt to
+	 * write any more output at this time.
+	 */
+
+	if (written < 0) {
+	    /*
+	     * If the last attempt to write was interrupted, simply retry.
+	     */
+
+	    if (errorCode == EINTR) {
+		errorCode = 0;
+		ReleaseChannelBuffer(bufPtr);
+		continue;
+	    }
+
+	    /*
+	     * If the channel is non-blocking and we would have blocked, start
+	     * a background flushing handler and break out of the loop.
+	     */
+
+	    if ((errorCode == EWOULDBLOCK) || (errorCode == EAGAIN)) {
+		/*
+		 * This used to check for CHANNEL_NONBLOCKING, and panic if
+		 * the channel was blocking. However, it appears that setting
+		 * stdin to -blocking 0 has some effect on the stdout when
+		 * it's a tty channel (dup'ed underneath)
+		 */
+
+		if (!GotFlag(statePtr, BG_FLUSH_SCHEDULED) && !TclInExit()) {
+		    SetFlag(statePtr, BG_FLUSH_SCHEDULED);
+		    UpdateInterest(chanPtr);
+		}
+		errorCode = 0;
+		ReleaseChannelBuffer(bufPtr);
+		break;
+	    }
+
+	    /*
+	     * Decide whether to report the error upwards or defer it.
+	     */
+
+	    if (calledFromAsyncFlush) {
+		/*
+		 * TIP #219, Tcl Channel Reflection API.
+		 * When defering the error copy a message from the bypass into
+		 * the unreported area. Or discard it if the new error is to
+		 * be ignored in favor of an earlier defered error.
+		 */
+
+		Tcl_Obj *msg = statePtr->chanMsg;
+
+		if (statePtr->unreportedError == 0) {
+		    statePtr->unreportedError = errorCode;
+		    statePtr->unreportedMsg = msg;
+		    if (msg != NULL) {
+			Tcl_IncrRefCount(msg);
+		    }
+		} else {
+		    /*
+		     * An old unreported error is kept, and this error thrown
+		     * away.
+		     */
+
+		    statePtr->chanMsg = NULL;
+		    if (msg != NULL) {
+			TclDecrRefCount(msg);
+		    }
+		}
+	    } else {
+		/*
+		 * TIP #219, Tcl Channel Reflection API.
+		 * Move error messages put by the driver into the chan bypass
+		 * area into the regular interpreter result. Fall back to the
+		 * regular message if nothing was found in the bypasses.
+		 */
+
+		Tcl_SetErrno(errorCode);
+		if (interp != NULL && !TclChanCaughtErrorBypass(interp,
+			(Tcl_Channel) chanPtr)) {
+		    Tcl_SetObjResult(interp,
+			    Tcl_NewStringObj(Tcl_PosixError(interp), -1));
+		}
+
+		/*
+		 * An unreportable bypassed message is kept, for the caller of
+		 * Tcl_Seek, Tcl_Write, etc.
+		 */
+	    }
+
+	    /*
+	     * When we get an error we throw away all the output currently
+	     * queued.
+	     */
+
+	    DiscardOutputQueued(statePtr);
+	    ReleaseChannelBuffer(bufPtr);
+	    break;
+	} else {
+	    /*
+             * TODO: Consider detecting and reacting to short writes on
+	     * blocking channels.  Ought not happen.  See iocmd-24.2.
+             */
+
+	    wroteSome = 1;
+	}
+
+	bufPtr->nextRemoved += written;
+
+	/*
+	 * If this buffer is now empty, recycle it.
+	 */
+
+	if (IsBufferEmpty(bufPtr)) {
+	    statePtr->outQueueHead = bufPtr->nextPtr;
+	    if (statePtr->outQueueHead == NULL) {
+		statePtr->outQueueTail = NULL;
+	    }
+	    RecycleBuffer(statePtr, bufPtr, 0);
+	}
+	ReleaseChannelBuffer(bufPtr);
+    }	/* Closes "while". */
+
+    /*
+     * If we wrote some data while flushing in the background, we are done.
+     * We can't finish the background flush until we run out of data and the
+     * channel becomes writable again. This ensures that all of the pending
+     * data has been flushed at the system level.
+     */
+
+    if (GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	if (wroteSome) {
+	    goto done;
+	} else if (statePtr->outQueueHead == NULL) {
+	    ResetFlag(statePtr, BG_FLUSH_SCHEDULED);
+	    ChanWatch(chanPtr, statePtr->interestMask);
+	} else {
+	    /*
+	     * When we are calledFromAsyncFlush, that means a writable
+	     * state on the channel triggered the call, so we should be
+	     * able to write something.  Either we did write something
+	     * and wroteSome should be set, or there was nothing left to
+	     * write in this call, and we've completed the BG flush.
+	     * These are the two cases above.  If we get here, that means
+	     * there is some kind failure in the writable event machinery.
+	     *
+	     * The tls extension indeed suffers from flaws in its channel
+	     * event mgmt.  See http://core.tcl.tk/tcl/info/c31ca233ca.
+	     * Until that patch is broadly distributed, disable the
+	     * assertion checking here, so that programs using Tcl and
+	     * tls can be debugged.
+
+	    assert(!calledFromAsyncFlush);
+	     */
+	}
+    }
+
+    /*
+     * If the channel is flagged as closed, delete it when the refCount drops
+     * to zero, the output queue is empty and there is no output in the
+     * current output buffer.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_CLOSED) && (statePtr->refCount <= 0) &&
+	    (statePtr->outQueueHead == NULL) &&
+	    ((statePtr->curOutPtr == NULL) ||
+	    IsBufferEmpty(statePtr->curOutPtr))) {
+	errorCode = CloseChannel(interp, chanPtr, errorCode);
+	goto done;
+    }
+
+    /*
+     * If the write-side of the channel is flagged as closed, delete it when
+     * the output queue is empty and there is no output in the current output
+     * buffer.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_CLOSEDWRITE) &&
+	    (statePtr->outQueueHead == NULL) &&
+	    ((statePtr->curOutPtr == NULL) ||
+	    IsBufferEmpty(statePtr->curOutPtr))) {
+	errorCode = CloseChannelPart(interp, chanPtr, errorCode,
+                TCL_CLOSE_WRITE);
+	goto done;
+    }
+
+  done:
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    return errorCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CloseChannel --
+ *
+ *	Utility procedure to close a channel and free associated resources.
+ *
+ *	If the channel was stacked, then the it will copy the necessary
+ *	elements of the NEXT channel into the TOP channel, in essence
+ *	unstacking the channel. The NEXT channel will then be freed.
+ *
+ *	If the channel was not stacked, then we will free all the bits for the
+ *	TOP channel, including the data structure itself.
+ *
+ * Results:
+ *	Error code from an unreported error or the driver close operation.
+ *
+ * Side effects:
+ *	May close the actual channel, may free memory, may change the value of
+ *	errno.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CloseChannel(
+    Tcl_Interp *interp,		/* For error reporting. */
+    Channel *chanPtr,		/* The channel to close. */
+    int errorCode)		/* Status of operation so far. */
+{
+    int result = 0;		/* Of calling driver close operation. */
+    ChannelState *statePtr;	/* State of the channel stack. */
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (chanPtr == NULL) {
+	return result;
+    }
+    statePtr = chanPtr->state;
+
+    /*
+     * No more input can be consumed so discard any leftover input.
+     */
+
+    DiscardInputQueued(statePtr, 1);
+
+    /*
+     * Discard a leftover buffer in the current output buffer field.
+     */
+
+    if (statePtr->curOutPtr != NULL) {
+	ReleaseChannelBuffer(statePtr->curOutPtr);
+	statePtr->curOutPtr = NULL;
+    }
+
+    /*
+     * The caller guarantees that there are no more buffers queued for output.
+     */
+
+    if (statePtr->outQueueHead != NULL) {
+	Tcl_Panic("TclFlush, closed channel: queued output left");
+    }
+
+    /*
+     * If the EOF character is set in the channel, append that to the output
+     * device.
+     */
+
+    if ((statePtr->outEofChar != 0) && GotFlag(statePtr, TCL_WRITABLE)) {
+	int dummy;
+	char c = (char) statePtr->outEofChar;
+
+	(void) ChanWrite(chanPtr, &c, 1, &dummy);
+    }
+
+    /*
+     * TIP #219, Tcl Channel Reflection API.
+     * Move a leftover error message in the channel bypass into the
+     * interpreter bypass. Just clear it if there is no interpreter.
+     */
+
+    if (statePtr->chanMsg != NULL) {
+	if (interp != NULL) {
+	    Tcl_SetChannelErrorInterp(interp, statePtr->chanMsg);
+	}
+	TclDecrRefCount(statePtr->chanMsg);
+	statePtr->chanMsg = NULL;
+    }
+
+    /*
+     * Remove this channel from of the list of all channels.
+     */
+
+    CutChannel((Tcl_Channel) chanPtr);
+
+    /*
+     * Close and free the channel driver state.
+     * This may leave a TIP #219 error message in the interp.
+     */
+
+    result = ChanClose(chanPtr, interp);
+
+    /*
+     * Some resources can be cleared only if the bottom channel in a stack is
+     * closed. All the other channels in the stack are not allowed to remove.
+     */
+
+    if (chanPtr == statePtr->bottomChanPtr) {
+	if (statePtr->channelName != NULL) {
+	    ckfree(statePtr->channelName);
+	    statePtr->channelName = NULL;
+	}
+
+	Tcl_FreeEncoding(statePtr->encoding);
+    }
+
+    /*
+     * If we are being called synchronously, report either any latent error on
+     * the channel or the current error.
+     */
+
+    if (statePtr->unreportedError != 0) {
+	errorCode = statePtr->unreportedError;
+
+	/*
+	 * TIP #219, Tcl Channel Reflection API.
+	 * Move an error message found in the unreported area into the regular
+	 * bypass (interp). This kills any message in the channel bypass area.
+	 */
+
+	if (statePtr->chanMsg != NULL) {
+	    TclDecrRefCount(statePtr->chanMsg);
+	    statePtr->chanMsg = NULL;
+	}
+	if (interp) {
+	    Tcl_SetChannelErrorInterp(interp, statePtr->unreportedMsg);
+	}
+    }
+    if (errorCode == 0) {
+	errorCode = result;
+	if (errorCode != 0) {
+	    Tcl_SetErrno(errorCode);
+	}
+    }
+
+    /*
+     * Cancel any outstanding timer.
+     */
+
+    Tcl_DeleteTimerHandler(statePtr->timer);
+
+    /*
+     * Mark the channel as deleted by clearing the type structure.
+     */
+
+    if (chanPtr->downChanPtr != NULL) {
+	Channel *downChanPtr = chanPtr->downChanPtr;
+
+	statePtr->nextCSPtr = tsdPtr->firstCSPtr;
+	tsdPtr->firstCSPtr = statePtr;
+
+	statePtr->topChanPtr = downChanPtr;
+	downChanPtr->upChanPtr = NULL;
+
+	ChannelFree(chanPtr);
+
+	return Tcl_Close(interp, (Tcl_Channel) downChanPtr);
+    }
+
+    /*
+     * There is only the TOP Channel, so we free the remaining pointers we
+     * have and then ourselves. Since this is the last of the channels in the
+     * stack, make sure to free the ChannelState structure associated with it.
+     */
+
+    ChannelFree(chanPtr);
+
+    Tcl_EventuallyFree(statePtr, TCL_DYNAMIC);
+
+    return errorCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CutChannel --
+ * CutChannel --
+ *
+ *	Removes a channel from the (thread-)global list of all channels (in
+ *	that thread). This is actually the statePtr for the stack of channel.
+ *
+ * Results:
+ *	Nothing.
+ *
+ * Side effects:
+ *	Resets the field 'nextCSPtr' of the specified channel state to NULL.
+ *
+ * NOTE:
+ *	The channel to cut out of the list must not be referenced in any
+ *	interpreter. This is something this procedure cannot check (despite
+ *	the refcount) because the caller usually wants fiddle with the channel
+ *	(like transfering it to a different thread) and thus keeps the
+ *	refcount artifically high to prevent its destruction.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CutChannel(
+    Tcl_Channel chan)		/* The channel being removed. Must not be
+				 * referenced in any interpreter. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelState *prevCSPtr;	/* Preceding channel state in list of all
+				 * states - used to splice a channel out of
+				 * the list on close. */
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of the channel stack. */
+
+    /*
+     * Remove this channel from of the list of all channels (in the current
+     * thread).
+     */
+
+    if (tsdPtr->firstCSPtr && (statePtr == tsdPtr->firstCSPtr)) {
+	tsdPtr->firstCSPtr = statePtr->nextCSPtr;
+    } else {
+	for (prevCSPtr = tsdPtr->firstCSPtr;
+		prevCSPtr && (prevCSPtr->nextCSPtr != statePtr);
+		prevCSPtr = prevCSPtr->nextCSPtr) {
+	    /* Empty loop body. */
+	}
+	if (prevCSPtr == NULL) {
+	    Tcl_Panic("FlushChannel: damaged channel list");
+	}
+	prevCSPtr->nextCSPtr = statePtr->nextCSPtr;
+    }
+
+    statePtr->nextCSPtr = NULL;
+
+    /*
+     * TIP #218, Channel Thread Actions
+     */
+
+    ChanThreadAction((Channel *) chan, TCL_CHANNEL_THREAD_REMOVE);
+}
+
+void
+Tcl_CutChannel(
+    Tcl_Channel chan)		/* The channel being added. Must not be
+				 * referenced in any interpreter. */
+{
+    Channel *chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelState *prevCSPtr;	/* Preceding channel state in list of all
+				 * states - used to splice a channel out of
+				 * the list on close. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State of the channel stack. */
+
+    /*
+     * Remove this channel from of the list of all channels (in the current
+     * thread).
+     */
+
+    if (tsdPtr->firstCSPtr && (statePtr == tsdPtr->firstCSPtr)) {
+	tsdPtr->firstCSPtr = statePtr->nextCSPtr;
+    } else {
+	for (prevCSPtr = tsdPtr->firstCSPtr;
+		prevCSPtr && (prevCSPtr->nextCSPtr != statePtr);
+		prevCSPtr = prevCSPtr->nextCSPtr) {
+	    /* Empty loop body. */
+	}
+	if (prevCSPtr == NULL) {
+	    Tcl_Panic("FlushChannel: damaged channel list");
+	}
+	prevCSPtr->nextCSPtr = statePtr->nextCSPtr;
+    }
+
+    statePtr->nextCSPtr = NULL;
+
+    /*
+     * TIP #218, Channel Thread Actions
+     * For all transformations and the base channel.
+     */
+
+    for (; chanPtr != NULL ; chanPtr = chanPtr->upChanPtr) {
+	ChanThreadAction(chanPtr, TCL_CHANNEL_THREAD_REMOVE);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SpliceChannel --
+ * SpliceChannel --
+ *
+ *	Adds a channel to the (thread-)global list of all channels (in that
+ *	thread). Expects that the field 'nextChanPtr' in the channel is set to
+ *	NULL.
+ *
+ * Results:
+ *	Nothing.
+ *
+ * Side effects:
+ *	Nothing.
+ *
+ * NOTE:
+ *	The channel to splice into the list must not be referenced in any
+ *	interpreter. This is something this procedure cannot check (despite
+ *	the refcount) because the caller usually wants figgle with the channel
+ *	(like transfering it to a different thread) and thus keeps the
+ *	refcount artifically high to prevent its destruction.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+SpliceChannel(
+    Tcl_Channel chan)		/* The channel being added. Must not be
+				 * referenced in any interpreter. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelState *statePtr = ((Channel *) chan)->state;
+
+    if (statePtr->nextCSPtr != NULL) {
+	Tcl_Panic("SpliceChannel: trying to add channel used in different list");
+    }
+
+    statePtr->nextCSPtr = tsdPtr->firstCSPtr;
+    tsdPtr->firstCSPtr = statePtr;
+
+    /*
+     * TIP #10. Mark the current thread as the new one managing this channel.
+     *		Note: 'Tcl_GetCurrentThread' returns sensible values even for
+     *		a non-threaded core.
+     */
+
+    statePtr->managingThread = Tcl_GetCurrentThread();
+
+    /*
+     * TIP #218, Channel Thread Actions
+     */
+
+    ChanThreadAction((Channel *) chan, TCL_CHANNEL_THREAD_INSERT);
+}
+
+void
+Tcl_SpliceChannel(
+    Tcl_Channel chan)		/* The channel being added. Must not be
+				 * referenced in any interpreter. */
+{
+    Channel *chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelState *statePtr = chanPtr->state;
+
+    if (statePtr->nextCSPtr != NULL) {
+	Tcl_Panic("SpliceChannel: trying to add channel used in different list");
+    }
+
+    statePtr->nextCSPtr = tsdPtr->firstCSPtr;
+    tsdPtr->firstCSPtr = statePtr;
+
+    /*
+     * TIP #10. Mark the current thread as the new one managing this channel.
+     *		Note: 'Tcl_GetCurrentThread' returns sensible values even for
+     *		a non-threaded core.
+     */
+
+    statePtr->managingThread = Tcl_GetCurrentThread();
+
+    /*
+     * TIP #218, Channel Thread Actions
+     * For all transformations and the base channel.
+     */
+
+    for (; chanPtr != NULL ; chanPtr = chanPtr->upChanPtr) {
+	ChanThreadAction(chanPtr, TCL_CHANNEL_THREAD_INSERT);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Close --
+ *
+ *	Closes a channel.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Closes the channel if this is the last reference.
+ *
+ * NOTE:
+ *	Tcl_Close removes the channel as far as the user is concerned.
+ *	However, it may continue to exist for a while longer if it has a
+ *	background flush scheduled. The device itself is eventually closed and
+ *	the channel record removed, in CloseChannel, above.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_Close(
+    Tcl_Interp *interp,		/* Interpreter for errors. */
+    Tcl_Channel chan)		/* The channel being closed. Must not be
+				 * referenced in any interpreter. */
+{
+    CloseCallback *cbPtr;	/* Iterate over close callbacks for this
+				 * channel. */
+    Channel *chanPtr;		/* The real IO channel. */
+    ChannelState *statePtr;	/* State of real IO channel. */
+    int result;			/* Of calling FlushChannel. */
+    int flushcode;
+    int stickyError;
+
+    if (chan == NULL) {
+	return TCL_OK;
+    }
+
+    /*
+     * Perform special handling for standard channels being closed. If the
+     * refCount is now 1 it means that the last reference to the standard
+     * channel is being explicitly closed, so bump the refCount down
+     * artificially to 0. This will ensure that the channel is actually
+     * closed, below. Also set the static pointer to NULL for the channel.
+     */
+
+    CheckForStdChannelsBeingClosed(chan);
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = (Channel *) chan;
+    statePtr = chanPtr->state;
+    chanPtr = statePtr->topChanPtr;
+
+    if (statePtr->refCount > 0) {
+	Tcl_Panic("called Tcl_Close on channel with refCount > 0");
+    }
+
+    if (GotFlag(statePtr, CHANNEL_INCLOSE)) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "illegal recursive call to close through close-handler"
+                    " of channel", -1));
+	}
+	return TCL_ERROR;
+    }
+    SetFlag(statePtr, CHANNEL_INCLOSE);
+
+    /*
+     * When the channel has an escape sequence driven encoding such as
+     * iso2022, the terminated escape sequence must write to the buffer.
+     */
+
+    stickyError = 0;
+
+    if (GotFlag(statePtr, TCL_WRITABLE) && (statePtr->encoding != NULL)
+	    && !(statePtr->outputEncodingFlags & TCL_ENCODING_START)) {
+	int code = CheckChannelErrors(statePtr, TCL_WRITABLE);
+
+	if (code == 0) {
+	    statePtr->outputEncodingFlags |= TCL_ENCODING_END;
+	    code = WriteChars(chanPtr, "", 0);
+	    statePtr->outputEncodingFlags &= ~TCL_ENCODING_END;
+	    statePtr->outputEncodingFlags |= TCL_ENCODING_START;
+	}
+	if (code < 0) {
+	    stickyError = Tcl_GetErrno();
+	}
+
+	/*
+	 * TIP #219, Tcl Channel Reflection API.
+	 * Move an error message found in the channel bypass into the
+	 * interpreter bypass. Just clear it if there is no interpreter.
+	 */
+
+	if (statePtr->chanMsg != NULL) {
+	    if (interp != NULL) {
+		Tcl_SetChannelErrorInterp(interp, statePtr->chanMsg);
+	    }
+	    TclDecrRefCount(statePtr->chanMsg);
+	    statePtr->chanMsg = NULL;
+	}
+    }
+
+    Tcl_ClearChannelHandlers(chan);
+
+    /*
+     * Invoke the registered close callbacks and delete their records.
+     */
+
+    while (statePtr->closeCbPtr != NULL) {
+	cbPtr = statePtr->closeCbPtr;
+	statePtr->closeCbPtr = cbPtr->nextPtr;
+	cbPtr->proc(cbPtr->clientData);
+	ckfree(cbPtr);
+    }
+
+    ResetFlag(statePtr, CHANNEL_INCLOSE);
+
+    /*
+     * If this channel supports it, close the read side, since we don't need
+     * it anymore and this will help avoid deadlocks on some channel types.
+     */
+
+    if (chanPtr->typePtr->closeProc == TCL_CLOSE2PROC) {
+	result = chanPtr->typePtr->close2Proc(chanPtr->instanceData, interp,
+		TCL_CLOSE_READ);
+    } else {
+	result = 0;
+    }
+
+    /*
+     * The call to FlushChannel will flush any queued output and invoke the
+     * close function of the channel driver, or it will set up the channel to
+     * be flushed and closed asynchronously.
+     */
+
+    SetFlag(statePtr, CHANNEL_CLOSED);
+
+    flushcode = FlushChannel(interp, chanPtr, 0);
+
+    /*
+     * TIP #219.
+     * Capture error messages put by the driver into the bypass area and put
+     * them into the regular interpreter result.
+     *
+     * Notes: Due to the assertion of CHANNEL_CLOSED in the flags
+     * FlushChannel() has called CloseChannel() and thus freed all the channel
+     * structures. We must not try to access "chan" anymore, hence the NULL
+     * argument in the call below. The only place which may still contain a
+     * message is the interpreter itself, and "CloseChannel" made sure to lift
+     * any channel message it generated into it.
+     */
+
+    if (TclChanCaughtErrorBypass(interp, NULL)) {
+	result = EINVAL;
+    }
+
+    if (stickyError != 0) {
+	Tcl_SetErrno(stickyError);
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp,
+			     Tcl_NewStringObj(Tcl_PosixError(interp), -1));
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Bug 97069ea11a: set error message if a flush code is set and no error
+     * message set up to now.
+     */
+
+    if (flushcode != 0 && interp != NULL
+	    && 0 == Tcl_GetCharLength(Tcl_GetObjResult(interp))) {
+	Tcl_SetErrno(flushcode);
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj(Tcl_PosixError(interp), -1));
+    }
+    if ((flushcode != 0) || (result != 0)) {
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CloseEx --
+ *
+ *	Closes one side of a channel, read or write.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Closes one direction of the channel.
+ *
+ * NOTE:
+ *	Tcl_CloseEx closes the specified direction of the channel as far as
+ *	the user is concerned. The channel keeps existing however. You cannot
+ *	calls this function to close the last possible direction of the
+ *	channel. Use Tcl_Close for that.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_CloseEx(
+    Tcl_Interp *interp,		/* Interpreter for errors. */
+    Tcl_Channel chan,		/* The channel being closed. May still be used
+				 * by some interpreter. */
+    int flags)			/* Flags telling us which side to close. */
+{
+    Channel *chanPtr;		/* The real IO channel. */
+    ChannelState *statePtr;	/* State of real IO channel. */
+
+    if (chan == NULL) {
+	return TCL_OK;
+    }
+
+    /* TODO: assert flags validity ? */
+
+    chanPtr = (Channel *) chan;
+    statePtr = chanPtr->state;
+
+    /*
+     * Does the channel support half-close anyway? Error if not.
+     */
+
+    if (!chanPtr->typePtr->close2Proc) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "half-close of channels not supported by %ss",
+		chanPtr->typePtr->typeName));
+	return TCL_ERROR;
+    }
+
+    /*
+     * Is the channel unstacked ? If not we fail.
+     */
+
+    if (chanPtr != statePtr->topChanPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"half-close not applicable to stack of transformations", -1));
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check direction against channel mode. It is an error if we try to close
+     * a direction not supported by the channel (already closed, or never
+     * opened for that direction).
+     */
+
+    if (!(statePtr->flags & (TCL_READABLE | TCL_WRITABLE) & flags)) {
+	const char *msg;
+
+	if (flags & TCL_CLOSE_READ) {
+	    msg = "read";
+	} else {
+	    msg = "write";
+	}
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "Half-close of %s-side not possible, side not opened or"
+                " already closed", msg));
+	return TCL_ERROR;
+    }
+
+    /*
+     * A user may try to call half-close from within a channel close handler.
+     * That won't do.
+     */
+
+    if (statePtr->flags & CHANNEL_INCLOSE) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "illegal recursive call to close through close-handler"
+                    " of channel", -1));
+	}
+	return TCL_ERROR;
+    }
+
+    if (flags & TCL_CLOSE_READ) {
+	/*
+	 * Call the finalization code directly. There are no events to handle,
+	 * there cannot be for the read-side.
+	 */
+
+	return CloseChannelPart(interp, chanPtr, 0, flags);
+    } else if (flags & TCL_CLOSE_WRITE) {
+	Tcl_Preserve(statePtr);
+	if (!GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	    /*
+	     * We don't want to re-enter CloseWrite().
+	     */
+
+	    if (!GotFlag(statePtr, CHANNEL_CLOSEDWRITE)) {
+		if (CloseWrite(interp, chanPtr) != TCL_OK) {
+		    SetFlag(statePtr, CHANNEL_CLOSEDWRITE);
+		    Tcl_Release(statePtr);
+		    return TCL_ERROR;
+		}
+	    }
+	}
+	SetFlag(statePtr, CHANNEL_CLOSEDWRITE);
+	Tcl_Release(statePtr);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CloseWrite --
+ *
+ *	Closes the write side a channel.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Closes the write side of the channel.
+ *
+ * NOTE:
+ *	CloseWrite removes the channel as far as the user is concerned.
+ *	However, the ooutput data structures may continue to exist for a while
+ *	longer if it has a background flush scheduled. The device itself is
+ *	eventually closed and the channel structures modified, in
+ *	CloseChannelPart, below.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CloseWrite(
+    Tcl_Interp *interp,		/* Interpreter for errors. */
+    Channel *chanPtr)		/* The channel whose write side is being
+                                 * closed. May still be used by some
+                                 * interpreter */
+{
+    /*
+     * Notes: clear-channel-handlers - write side only ? or keep around, just
+     * not called.
+     *
+     * No close callbacks are run - channel is still open (read side)
+     */
+
+    ChannelState *statePtr = chanPtr->state;
+                                /* State of real IO channel. */
+    int flushcode;
+    int result = 0;
+
+    /*
+     * The call to FlushChannel will flush any queued output and invoke the
+     * close function of the channel driver, or it will set up the channel to
+     * be flushed and closed asynchronously.
+     */
+
+    SetFlag(statePtr, CHANNEL_CLOSEDWRITE);
+
+    flushcode = FlushChannel(interp, chanPtr, 0);
+
+    /*
+     * TIP #219.
+     * Capture error messages put by the driver into the bypass area and put
+     * them into the regular interpreter result.
+     *
+     * Notes: Due to the assertion of CHANNEL_CLOSEDWRITE in the flags
+     * FlushChannel() has called CloseChannelPart(). While we can still access
+     * "chan" (no structures were freed), the only place which may still
+     * contain a message is the interpreter itself, and "CloseChannelPart"
+     * made sure to lift any channel message it generated into it. Hence the
+     * NULL argument in the call below.
+     */
+
+    if (TclChanCaughtErrorBypass(interp, NULL)) {
+	result = EINVAL;
+    }
+
+    if ((flushcode != 0) || (result != 0)) {
+	return TCL_ERROR;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CloseChannelPart --
+ *
+ *	Utility procedure to close a channel partially and free associated
+ *	resources. If the channel was stacked it will never be run (The higher
+ *	level forbid this). If the channel was not stacked, then we will free
+ *	all the bits of the chosen side (read, or write) for the TOP channel.
+ *
+ * Results:
+ *	Error code from an unreported error or the driver close2 operation.
+ *
+ * Side effects:
+ *	May free memory, may change the value of errno.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CloseChannelPart(
+    Tcl_Interp *interp,		/* Interpreter for errors. */
+    Channel *chanPtr,		/* The channel being closed. May still be used
+				 * by some interpreter. */
+    int errorCode,		/* Status of operation so far. */
+    int flags)			/* Flags telling us which side to close. */
+{
+    ChannelState *statePtr;	/* State of real IO channel. */
+    int result;			/* Of calling the close2proc. */
+
+    statePtr = chanPtr->state;
+
+    if (flags & TCL_CLOSE_READ) {
+	/*
+	 * No more input can be consumed so discard any leftover input.
+	 */
+
+	DiscardInputQueued(statePtr, 1);
+    } else if (flags & TCL_CLOSE_WRITE) {
+	/*
+	 * The caller guarantees that there are no more buffers queued for
+	 * output.
+	 */
+
+	if (statePtr->outQueueHead != NULL) {
+	    Tcl_Panic("ClosechanHalf, closed write-side of channel: "
+		    "queued output left");
+	}
+
+	/*
+	 * If the EOF character is set in the channel, append that to the
+	 * output device.
+	 */
+
+	if ((statePtr->outEofChar != 0) && GotFlag(statePtr, TCL_WRITABLE)) {
+	    int dummy;
+	    char c = (char) statePtr->outEofChar;
+
+	    (void) ChanWrite(chanPtr, &c, 1, &dummy);
+	}
+
+	/*
+	 * TIP #219, Tcl Channel Reflection API.
+	 * Move a leftover error message in the channel bypass into the
+	 * interpreter bypass. Just clear it if there is no interpreter.
+	 */
+
+	if (statePtr->chanMsg != NULL) {
+	    if (interp != NULL) {
+		Tcl_SetChannelErrorInterp(interp, statePtr->chanMsg);
+	    }
+	    TclDecrRefCount(statePtr->chanMsg);
+	    statePtr->chanMsg = NULL;
+	}
+    }
+
+    /*
+     * Finally do what is asked of us. Close and free the channel driver state
+     * for the chosen side of the channel. This may leave a TIP #219 error
+     * message in the interp.
+     */
+
+    result = ChanCloseHalf(chanPtr, interp, flags);
+
+    /*
+     * If we are being called synchronously, report either any latent error on
+     * the channel or the current error.
+     */
+
+    if (statePtr->unreportedError != 0) {
+	errorCode = statePtr->unreportedError;
+
+	/*
+	 * TIP #219, Tcl Channel Reflection API.
+	 * Move an error message found in the unreported area into the regular
+	 * bypass (interp). This kills any message in the channel bypass area.
+	 */
+
+	if (statePtr->chanMsg != NULL) {
+	    TclDecrRefCount(statePtr->chanMsg);
+	    statePtr->chanMsg = NULL;
+	}
+	if (interp) {
+	    Tcl_SetChannelErrorInterp(interp, statePtr->unreportedMsg);
+	}
+    }
+    if (errorCode == 0) {
+	errorCode = result;
+	if (errorCode != 0) {
+	    Tcl_SetErrno(errorCode);
+	}
+    }
+
+    /*
+     * TIP #219.
+     * Capture error messages put by the driver into the bypass area and put
+     * them into the regular interpreter result. See also the bottom of
+     * CloseWrite().
+     */
+
+    if (TclChanCaughtErrorBypass(interp, (Tcl_Channel) chanPtr)) {
+	result = EINVAL;
+    }
+
+    if (result != 0) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Remove the closed side from the channel mode/flags.
+     */
+
+    ResetFlag(statePtr, flags & (TCL_READABLE | TCL_WRITABLE));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ClearChannelHandlers --
+ *
+ *	Removes all channel handlers and event scripts from the channel,
+ *	cancels all background copies involving the channel and any interest
+ *	in events.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	See above. Deallocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ClearChannelHandlers(
+    Tcl_Channel channel)
+{
+    ChannelHandler *chPtr, *chNext;	/* Iterate over channel handlers. */
+    EventScriptRecord *ePtr, *eNextPtr;	/* Iterate over eventscript records. */
+    Channel *chanPtr;			/* The real IO channel. */
+    ChannelState *statePtr;		/* State of real IO channel. */
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    NextChannelHandler *nhPtr;
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = (Channel *) channel;
+    statePtr = chanPtr->state;
+    chanPtr = statePtr->topChanPtr;
+
+    /*
+     * Cancel any outstanding timer.
+     */
+
+    Tcl_DeleteTimerHandler(statePtr->timer);
+
+    /*
+     * Remove any references to channel handlers for this channel that may be
+     * about to be invoked.
+     */
+
+    for (nhPtr = tsdPtr->nestedHandlerPtr; nhPtr != NULL;
+	    nhPtr = nhPtr->nestedHandlerPtr) {
+	if (nhPtr->nextHandlerPtr &&
+		(nhPtr->nextHandlerPtr->chanPtr == chanPtr)) {
+	    nhPtr->nextHandlerPtr = NULL;
+	}
+    }
+
+    /*
+     * Remove all the channel handler records attached to the channel itself.
+     */
+
+    for (chPtr = statePtr->chPtr; chPtr != NULL; chPtr = chNext) {
+	chNext = chPtr->nextPtr;
+	ckfree(chPtr);
+    }
+    statePtr->chPtr = NULL;
+
+    /*
+     * Cancel any pending copy operation.
+     */
+
+    StopCopy(statePtr->csPtrR);
+    StopCopy(statePtr->csPtrW);
+
+    /*
+     * Must set the interest mask now to 0, otherwise infinite loops will
+     * occur if Tcl_DoOneEvent is called before the channel is finally deleted
+     * in FlushChannel. This can happen if the channel has a background flush
+     * active.
+     */
+
+    statePtr->interestMask = 0;
+
+    /*
+     * Remove any EventScript records for this channel.
+     */
+
+    for (ePtr = statePtr->scriptRecordPtr; ePtr != NULL; ePtr = eNextPtr) {
+	eNextPtr = ePtr->nextPtr;
+	TclDecrRefCount(ePtr->scriptPtr);
+	ckfree(ePtr);
+    }
+    statePtr->scriptRecordPtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Write --
+ *
+ *	Puts a sequence of bytes into an output buffer, may queue the buffer
+ *	for output if it gets full, and also remembers whether the current
+ *	buffer is ready e.g. if it contains a newline and we are in line
+ *	buffering mode. Compensates stacking, i.e. will redirect the data from
+ *	the specified channel to the topmost channel in a stack.
+ *
+ *	No encoding conversions are applied to the bytes being read.
+ *
+ * Results:
+ *	The number of bytes written or -1 in case of error. If -1,
+ *	Tcl_GetErrno will return the error code.
+ *
+ * Side effects:
+ *	May buffer up output and may cause output to be produced on the
+ *	channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Write(
+    Tcl_Channel chan,		/* The channel to buffer output for. */
+    const char *src,		/* Data to queue in output buffer. */
+    int srcLen)			/* Length of data in bytes, or < 0 for
+				 * strlen(). */
+{
+    /*
+     * Always use the topmost channel of the stack
+     */
+
+    Channel *chanPtr;
+    ChannelState *statePtr;	/* State info for channel */
+
+    statePtr = ((Channel *) chan)->state;
+    chanPtr = statePtr->topChanPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE) != 0) {
+	return -1;
+    }
+
+    if (srcLen < 0) {
+	srcLen = strlen(src);
+    }
+    if (WriteBytes(chanPtr, src, srcLen) < 0) {
+	return -1;
+    }
+    return srcLen;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_WriteRaw --
+ *
+ *	Puts a sequence of bytes into an output buffer, may queue the buffer
+ *	for output if it gets full, and also remembers whether the current
+ *	buffer is ready e.g. if it contains a newline and we are in line
+ *	buffering mode. Writes directly to the driver of the channel, does not
+ *	compensate for stacking.
+ *
+ *	No encoding conversions are applied to the bytes being read.
+ *
+ * Results:
+ *	The number of bytes written or -1 in case of error. If -1,
+ *	Tcl_GetErrno will return the error code.
+ *
+ * Side effects:
+ *	May buffer up output and may cause output to be produced on the
+ *	channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_WriteRaw(
+    Tcl_Channel chan,		/* The channel to buffer output for. */
+    const char *src,		/* Data to queue in output buffer. */
+    int srcLen)			/* Length of data in bytes, or < 0 for
+				 * strlen(). */
+{
+    Channel *chanPtr = ((Channel *) chan);
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int errorCode, written;
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE | CHANNEL_RAW_MODE) != 0) {
+	return -1;
+    }
+
+    if (srcLen < 0) {
+	srcLen = strlen(src);
+    }
+
+    /*
+     * Go immediately to the driver, do all the error handling by ourselves.
+     * The code was stolen from 'FlushChannel'.
+     */
+
+    written = ChanWrite(chanPtr, src, srcLen, &errorCode);
+    if (written < 0) {
+	Tcl_SetErrno(errorCode);
+    }
+
+    return written;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_WriteChars --
+ *
+ *	Takes a sequence of UTF-8 characters and converts them for output
+ *	using the channel's current encoding, may queue the buffer for output
+ *	if it gets full, and also remembers whether the current buffer is
+ *	ready e.g. if it contains a newline and we are in line buffering
+ *	mode. Compensates stacking, i.e. will redirect the data from the
+ *	specified channel to the topmost channel in a stack.
+ *
+ * Results:
+ *	The number of bytes written or -1 in case of error. If -1,
+ *	Tcl_GetErrno will return the error code.
+ *
+ * Side effects:
+ *	May buffer up output and may cause output to be produced on the
+ *	channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_WriteChars(
+    Tcl_Channel chan,		/* The channel to buffer output for. */
+    const char *src,		/* UTF-8 characters to queue in output
+				 * buffer. */
+    int len)			/* Length of string in bytes, or < 0 for
+				 * strlen(). */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;	/* State info for channel */
+    int result;
+    Tcl_Obj *objPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE) != 0) {
+	return -1;
+    }
+
+    chanPtr = statePtr->topChanPtr;
+
+    if (len < 0) {
+	len = strlen(src);
+    }
+    if (statePtr->encoding) {
+	return WriteChars(chanPtr, src, len);
+    }
+
+    /*
+     * Inefficient way to convert UTF-8 to byte-array, but the code
+     * parallels the way it is done for objects.  Special case for 1-byte
+     * (used by eg [puts] for the \n) could be extended to more efficient
+     * translation of the src string.
+     */
+
+    if ((len == 1) && (UCHAR(*src) < 0xC0)) {
+	return WriteBytes(chanPtr, src, len);
+    }
+
+    objPtr = Tcl_NewStringObj(src, len);
+    src = (char *) Tcl_GetByteArrayFromObj(objPtr, &len);
+    result = WriteBytes(chanPtr, src, len);
+    TclDecrRefCount(objPtr);
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_WriteObj --
+ *
+ *	Takes the Tcl object and queues its contents for output. If the
+ *	encoding of the channel is NULL, takes the byte-array representation
+ *	of the object and queues those bytes for output. Otherwise, takes the
+ *	characters in the UTF-8 (string) representation of the object and
+ *	converts them for output using the channel's current encoding. May
+ *	flush internal buffers to output if one becomes full or is ready for
+ *	some other reason, e.g. if it contains a newline and the channel is in
+ *	line buffering mode.
+ *
+ * Results:
+ *	The number of bytes written or -1 in case of error. If -1,
+ *	Tcl_GetErrno() will return the error code.
+ *
+ * Side effects:
+ *	May buffer up output and may cause output to be produced on the
+ *	channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_WriteObj(
+    Tcl_Channel chan,		/* The channel to buffer output for. */
+    Tcl_Obj *objPtr)		/* The object to write. */
+{
+    /*
+     * Always use the topmost channel of the stack
+     */
+
+    Channel *chanPtr;
+    ChannelState *statePtr;	/* State info for channel */
+    const char *src;
+    int srcLen;
+
+    statePtr = ((Channel *) chan)->state;
+    chanPtr = statePtr->topChanPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE) != 0) {
+	return -1;
+    }
+    if (statePtr->encoding == NULL) {
+	src = (char *) Tcl_GetByteArrayFromObj(objPtr, &srcLen);
+	return WriteBytes(chanPtr, src, srcLen);
+    } else {
+	src = TclGetStringFromObj(objPtr, &srcLen);
+	return WriteChars(chanPtr, src, srcLen);
+    }
+}
+
+static void
+WillWrite(
+    Channel *chanPtr)
+{
+    int inputBuffered;
+
+    if ((chanPtr->typePtr->seekProc != NULL) &&
+            ((inputBuffered = Tcl_InputBuffered((Tcl_Channel) chanPtr)) > 0)){
+        int ignore;
+
+        DiscardInputQueued(chanPtr->state, 0);
+        ChanSeek(chanPtr, -inputBuffered, SEEK_CUR, &ignore);
+    }
+}
+
+static int
+WillRead(
+    Channel *chanPtr)
+{
+    if (chanPtr->typePtr == NULL) {
+	/*
+         * Prevent read attempts on a closed channel.
+         */
+
+        DiscardInputQueued(chanPtr->state, 0);
+	Tcl_SetErrno(EINVAL);
+	return -1;
+    }
+    if ((chanPtr->typePtr->seekProc != NULL)
+            && (Tcl_OutputBuffered((Tcl_Channel) chanPtr) > 0)) {
+	/*
+	 * CAVEAT - The assumption here is that FlushChannel() will push out
+	 * the bytes of any writes that are in progress.  Since this is a
+	 * seekable channel, we assume it is not one that can block and force
+	 * bg flushing.  Channels we know that can do that - sockets, pipes -
+	 * are not seekable. If the assumption is wrong, more drastic measures
+	 * may be required here like temporarily setting the channel into
+	 * blocking mode.
+	 */
+
+        if (FlushChannel(NULL, chanPtr, 0) != 0) {
+            return -1;
+        }
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Write --
+ *
+ *	Convert srcLen bytes starting at src according to encoding and write
+ *	produced bytes into an output buffer, may queue the buffer for output
+ *	if it gets full, and also remembers whether the current buffer is
+ *	ready e.g. if it contains a newline and we are in line buffering mode.
+ *
+ * Results:
+ *	The number of bytes written or -1 in case of error. If -1,
+ *	Tcl_GetErrno will return the error code.
+ *
+ * Side effects:
+ *	May buffer up output and may cause output to be produced on the
+ *	channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+Write(
+    Channel *chanPtr,		/* The channel to buffer output for. */
+    const char *src,		/* UTF-8 string to write. */
+    int srcLen,			/* Length of UTF-8 string in bytes. */
+    Tcl_Encoding encoding)
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    char *nextNewLine = NULL;
+    int endEncoding, saved = 0, total = 0, flushed = 0, needNlFlush = 0;
+
+    if (srcLen) {
+        WillWrite(chanPtr);
+    }
+
+    /*
+     * Write the terminated escape sequence even if srcLen is 0.
+     */
+
+    endEncoding = ((statePtr->outputEncodingFlags & TCL_ENCODING_END) != 0);
+
+    if (GotFlag(statePtr, CHANNEL_LINEBUFFERED)
+	    || (statePtr->outputTranslation != TCL_TRANSLATE_LF)) {
+	nextNewLine = memchr(src, '\n', srcLen);
+    }
+
+    while (srcLen + saved + endEncoding > 0) {
+	ChannelBuffer *bufPtr;
+	char *dst, safe[BUFFER_PADDING];
+	int result, srcRead, dstLen, dstWrote, srcLimit = srcLen;
+
+	if (nextNewLine) {
+	    srcLimit = nextNewLine - src;
+	}
+
+	/* Get space to write into */
+	bufPtr = statePtr->curOutPtr;
+	if (bufPtr == NULL) {
+	    bufPtr = AllocChannelBuffer(statePtr->bufSize);
+	    statePtr->curOutPtr = bufPtr;
+	}
+	if (saved) {
+	    /*
+	     * Here's some translated bytes left over from the last buffer
+	     * that we need to stick at the beginning of this buffer.
+	     */
+
+	    memcpy(InsertPoint(bufPtr), safe, (size_t) saved);
+	    bufPtr->nextAdded += saved;
+	    saved = 0;
+	}
+	PreserveChannelBuffer(bufPtr);
+	dst = InsertPoint(bufPtr);
+	dstLen = SpaceLeft(bufPtr);
+
+	result = Tcl_UtfToExternal(NULL, encoding, src, srcLimit,
+		statePtr->outputEncodingFlags,
+		&statePtr->outputEncodingState, dst,
+		dstLen + BUFFER_PADDING, &srcRead, &dstWrote, NULL);
+
+	/*
+         * See chan-io-1.[89]. Tcl Bug 506297.
+         */
+
+	statePtr->outputEncodingFlags &= ~TCL_ENCODING_START;
+
+	if ((result != TCL_OK) && (srcRead + dstWrote == 0)) {
+	    /*
+             * We're reading from invalid/incomplete UTF-8.
+             */
+
+	    ReleaseChannelBuffer(bufPtr);
+	    if (total == 0) {
+		Tcl_SetErrno(EINVAL);
+		return -1;
+	    }
+	    break;
+	}
+
+	bufPtr->nextAdded += dstWrote;
+	src += srcRead;
+	srcLen -= srcRead;
+	total += dstWrote;
+	dst += dstWrote;
+	dstLen -= dstWrote;
+
+	if (src == nextNewLine && dstLen > 0) {
+	    static char crln[3] = "\r\n";
+	    char *nl = NULL;
+	    int nlLen = 0;
+
+	    switch (statePtr->outputTranslation) {
+	    case TCL_TRANSLATE_LF:
+		nl = crln + 1;
+		nlLen = 1;
+		break;
+	    case TCL_TRANSLATE_CR:
+		nl = crln;
+		nlLen = 1;
+		break;
+	    case TCL_TRANSLATE_CRLF:
+		nl = crln;
+		nlLen = 2;
+		break;
+	    default:
+		Tcl_Panic("unknown output translation requested");
+		break;
+	    }
+
+	    result |= Tcl_UtfToExternal(NULL, encoding, nl, nlLen,
+		    statePtr->outputEncodingFlags,
+		    &statePtr->outputEncodingState, dst,
+		    dstLen + BUFFER_PADDING, &srcRead, &dstWrote, NULL);
+	    assert(srcRead == nlLen);
+
+	    bufPtr->nextAdded += dstWrote;
+	    src++;
+	    srcLen--;
+	    total += dstWrote;
+	    dst += dstWrote;
+	    dstLen -= dstWrote;
+	    nextNewLine = memchr(src, '\n', srcLen);
+	    needNlFlush = 1;
+	}
+
+	if (IsBufferOverflowing(bufPtr)) {
+	    /*
+	     * When translating from UTF-8 to external encoding, we allowed
+	     * the translation to produce a character that crossed the end of
+	     * the output buffer, so that we would get a completely full
+	     * buffer before flushing it. The extra bytes will be moved to the
+	     * beginning of the next buffer.
+	     */
+
+	    saved = -SpaceLeft(bufPtr);
+	    memcpy(safe, dst + dstLen, (size_t) saved);
+	    bufPtr->nextAdded = bufPtr->bufLength;
+	}
+
+	if ((srcLen + saved == 0) && (result == TCL_OK)) {
+	    endEncoding = 0;
+	}
+
+	if (IsBufferFull(bufPtr)) {
+	    if (FlushChannel(NULL, chanPtr, 0) != 0) {
+		ReleaseChannelBuffer(bufPtr);
+		return -1;
+	    }
+	    flushed += statePtr->bufSize;
+
+	    /*
+ 	     * We just flushed.  So if we have needNlFlush set to record that
+ 	     * we need to flush because theres a (translated) newline in the
+ 	     * buffer, that's likely not true any more.  But there is a tricky
+ 	     * exception.  If we have saved bytes that did not really get
+ 	     * flushed and those bytes came from a translation of a newline as
+ 	     * the last thing taken from the src array, then needNlFlush needs
+ 	     * to remain set to flag that the next buffer still needs a
+ 	     * newline flush.
+ 	     */
+
+	    if (needNlFlush && (saved == 0 || src[-1] != '\n')) {
+		needNlFlush = 0;
+	    }
+	}
+	ReleaseChannelBuffer(bufPtr);
+    }
+    if ((flushed < total) && (GotFlag(statePtr, CHANNEL_UNBUFFERED) ||
+	    (needNlFlush && GotFlag(statePtr, CHANNEL_LINEBUFFERED)))) {
+	if (FlushChannel(NULL, chanPtr, 0) != 0) {
+	    return -1;
+	}
+    }
+
+    return total;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_Gets --
+ *
+ *	Reads a complete line of input from the channel into a Tcl_DString.
+ *
+ * Results:
+ *	Length of line read (in characters) or -1 if error, EOF, or blocked.
+ *	If -1, use Tcl_GetErrno() to retrieve the POSIX error code for the
+ *	error or condition that occurred.
+ *
+ * Side effects:
+ *	May flush output on the channel. May cause input to be consumed from
+ *	the channel.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_Gets(
+    Tcl_Channel chan,		/* Channel from which to read. */
+    Tcl_DString *lineRead)	/* The line read will be appended to this
+				 * DString as UTF-8 characters. The caller
+				 * must have initialized it and is responsible
+				 * for managing the storage. */
+{
+    Tcl_Obj *objPtr;
+    int charsStored;
+
+    TclNewObj(objPtr);
+    charsStored = Tcl_GetsObj(chan, objPtr);
+    if (charsStored > 0) {
+	TclDStringAppendObj(lineRead, objPtr);
+    }
+    TclDecrRefCount(objPtr);
+    return charsStored;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_GetsObj --
+ *
+ *	Accumulate input from the input channel until end-of-line or
+ *	end-of-file has been seen. Bytes read from the input channel are
+ *	converted to UTF-8 using the encoding specified by the channel.
+ *
+ * Results:
+ *	Number of characters accumulated in the object or -1 if error,
+ *	blocked, or EOF. If -1, use Tcl_GetErrno() to retrieve the POSIX error
+ *	code for the error or condition that occurred.
+ *
+ * Side effects:
+ *	Consumes input from the channel.
+ *
+ *	On reading EOF, leave channel pointing at EOF char. On reading EOL,
+ *	leave channel pointing after EOL, but don't return EOL in dst buffer.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_GetsObj(
+    Tcl_Channel chan,		/* Channel from which to read. */
+    Tcl_Obj *objPtr)		/* The line read will be appended to this
+				 * object as UTF-8 characters. */
+{
+    GetsState gs;
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr;
+    int inEofChar, skip, copiedTotal, oldLength, oldFlags, oldRemoved;
+    Tcl_Encoding encoding;
+    char *dst, *dstEnd, *eol, *eof;
+    Tcl_EncodingState oldState;
+
+    if (CheckChannelErrors(statePtr, TCL_READABLE) != 0) {
+	return -1;
+    }
+
+    /*
+     * If we're sitting ready to read the eofchar, there's no need to
+     * do it.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_STICKY_EOF)) {
+	SetFlag(statePtr, CHANNEL_EOF);
+	assert(statePtr->inputEncodingFlags & TCL_ENCODING_END);
+	assert(!GotFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR));
+
+	/* TODO: Do we need this? */
+	UpdateInterest(chanPtr);
+	return -1;
+    }
+
+    /*
+     * A binary version of Tcl_GetsObj. This could also handle encodings that
+     * are ascii-7 pure (iso8859, utf-8, ...) with a final encoding conversion
+     * done on objPtr.
+     */
+
+    if ((statePtr->encoding == NULL)
+	    && ((statePtr->inputTranslation == TCL_TRANSLATE_LF)
+		    || (statePtr->inputTranslation == TCL_TRANSLATE_CR))) {
+	return TclGetsObjBinary(chan, objPtr);
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+
+    bufPtr = statePtr->inQueueHead;
+    encoding = statePtr->encoding;
+
+    /*
+     * Preserved so we can restore the channel's state in case we don't find a
+     * newline in the available input.
+     */
+
+    TclGetStringFromObj(objPtr, &oldLength);
+    oldFlags = statePtr->inputEncodingFlags;
+    oldState = statePtr->inputEncodingState;
+    oldRemoved = BUFFER_PADDING;
+    if (bufPtr != NULL) {
+	oldRemoved = bufPtr->nextRemoved;
+    }
+
+    /*
+     * If there is no encoding, use "iso8859-1" -- Tcl_GetsObj() doesn't
+     * produce ByteArray objects.
+     */
+
+    if (encoding == NULL) {
+	encoding = GetBinaryEncoding();
+    }
+
+    /*
+     * Object used by FilterInputBytes to keep track of how much data has been
+     * consumed from the channel buffers.
+     */
+
+    gs.objPtr		= objPtr;
+    gs.dstPtr		= &dst;
+    gs.encoding		= encoding;
+    gs.bufPtr		= bufPtr;
+    gs.state		= oldState;
+    gs.rawRead		= 0;
+    gs.bytesWrote	= 0;
+    gs.charsWrote	= 0;
+    gs.totalChars	= 0;
+
+    dst = objPtr->bytes + oldLength;
+    dstEnd = dst;
+
+    skip = 0;
+    eof = NULL;
+    inEofChar = statePtr->inEofChar;
+
+    ResetFlag(statePtr, CHANNEL_BLOCKED);
+    while (1) {
+	if (dst >= dstEnd) {
+	    if (FilterInputBytes(chanPtr, &gs) != 0) {
+		goto restore;
+	    }
+	    dstEnd = dst + gs.bytesWrote;
+	}
+
+	/*
+	 * Remember if EOF char is seen, then look for EOL anyhow, because the
+	 * EOL might be before the EOF char.
+	 */
+
+	if (inEofChar != '\0') {
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == inEofChar) {
+		    dstEnd = eol;
+		    eof = eol;
+		    break;
+		}
+	    }
+	}
+
+	/*
+	 * On EOL, leave current file position pointing after the EOL, but
+	 * don't store the EOL in the output string.
+	 */
+
+	switch (statePtr->inputTranslation) {
+	case TCL_TRANSLATE_LF:
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == '\n') {
+		    skip = 1;
+		    goto gotEOL;
+		}
+	    }
+	    break;
+	case TCL_TRANSLATE_CR:
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == '\r') {
+		    skip = 1;
+		    goto gotEOL;
+		}
+	    }
+	    break;
+	case TCL_TRANSLATE_CRLF:
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == '\r') {
+		    eol++;
+
+		    /*
+		     * If a CR is at the end of the buffer, then check for a
+		     * LF at the begining of the next buffer, unless EOF char
+		     * was found already.
+		     */
+
+		    if (eol >= dstEnd) {
+			int offset;
+
+			if (eol != eof) {
+			    offset = eol - objPtr->bytes;
+			    dst = dstEnd;
+			    if (FilterInputBytes(chanPtr, &gs) != 0) {
+				goto restore;
+			    }
+			    dstEnd = dst + gs.bytesWrote;
+			    eol = objPtr->bytes + offset;
+			}
+			if (eol >= dstEnd) {
+			    skip = 0;
+			    goto gotEOL;
+			}
+		    }
+		    if (*eol == '\n') {
+			eol--;
+			skip = 2;
+			goto gotEOL;
+		    }
+		}
+	    }
+	    break;
+	case TCL_TRANSLATE_AUTO:
+	    eol = dst;
+	    skip = 1;
+	    if (GotFlag(statePtr, INPUT_SAW_CR)) {
+		ResetFlag(statePtr, INPUT_SAW_CR);
+		if ((eol < dstEnd) && (*eol == '\n')) {
+		    /*
+		     * Skip the raw bytes that make up the '\n'.
+		     */
+
+		    char tmp[TCL_UTF_MAX];
+		    int rawRead;
+
+		    bufPtr = gs.bufPtr;
+		    Tcl_ExternalToUtf(NULL, gs.encoding, RemovePoint(bufPtr),
+			    gs.rawRead, statePtr->inputEncodingFlags
+				| TCL_ENCODING_NO_TERMINATE, &gs.state, tmp,
+			    TCL_UTF_MAX, &rawRead, NULL, NULL);
+		    bufPtr->nextRemoved += rawRead;
+		    gs.rawRead -= rawRead;
+		    gs.bytesWrote--;
+		    gs.charsWrote--;
+		    memmove(dst, dst + 1, (size_t) (dstEnd - dst));
+		    dstEnd--;
+		}
+	    }
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == '\r') {
+		    eol++;
+		    if (eol == dstEnd) {
+			/*
+			 * If buffer ended on \r, peek ahead to see if a \n is
+			 * available, unless EOF char was found already.
+			 */
+
+			if (eol != eof) {
+			    int offset;
+
+			    offset = eol - objPtr->bytes;
+			    dst = dstEnd;
+			    PeekAhead(chanPtr, &dstEnd, &gs);
+			    eol = objPtr->bytes + offset;
+			}
+
+			if (eol >= dstEnd) {
+			    eol--;
+			    SetFlag(statePtr, INPUT_SAW_CR);
+			    goto gotEOL;
+			}
+		    }
+		    if (*eol == '\n') {
+			skip++;
+		    }
+		    eol--;
+		    goto gotEOL;
+		} else if (*eol == '\n') {
+		    goto gotEOL;
+		}
+	    }
+	}
+	if (eof != NULL) {
+	    /*
+	     * EOF character was seen. On EOF, leave current file position
+	     * pointing at the EOF character, but don't store the EOF
+	     * character in the output string.
+	     */
+
+	    dstEnd = eof;
+	    SetFlag(statePtr, CHANNEL_EOF | CHANNEL_STICKY_EOF);
+	    statePtr->inputEncodingFlags |= TCL_ENCODING_END;
+	    ResetFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR);
+	}
+	if (GotFlag(statePtr, CHANNEL_EOF)) {
+	    skip = 0;
+	    eol = dstEnd;
+	    if (eol == objPtr->bytes + oldLength) {
+		/*
+		 * If we didn't append any bytes before encountering EOF,
+		 * caller needs to see -1.
+		 */
+
+		Tcl_SetObjLength(objPtr, oldLength);
+		CommonGetsCleanup(chanPtr);
+		copiedTotal = -1;
+		ResetFlag(statePtr, CHANNEL_BLOCKED);
+		goto done;
+	    }
+	    goto gotEOL;
+	}
+	dst = dstEnd;
+    }
+
+    /*
+     * Found EOL or EOF, but the output buffer may now contain too many UTF-8
+     * characters. We need to know how many raw bytes correspond to the number
+     * of UTF-8 characters we want, plus how many raw bytes correspond to the
+     * character(s) making up EOL (if any), so we can remove the correct
+     * number of bytes from the channel buffer.
+     */
+
+  gotEOL:
+    /*
+     * Regenerate the top channel, in case it was changed due to
+     * self-modifying reflected transforms.
+     */
+
+    if (chanPtr != statePtr->topChanPtr) {
+	TclChannelRelease((Tcl_Channel)chanPtr);
+	chanPtr = statePtr->topChanPtr;
+	TclChannelPreserve((Tcl_Channel)chanPtr);
+    }
+
+    bufPtr = gs.bufPtr;
+    if (bufPtr == NULL) {
+	Tcl_Panic("Tcl_GetsObj: gotEOL reached with bufPtr==NULL");
+    }
+    statePtr->inputEncodingState = gs.state;
+    Tcl_ExternalToUtf(NULL, gs.encoding, RemovePoint(bufPtr), gs.rawRead,
+	    statePtr->inputEncodingFlags | TCL_ENCODING_NO_TERMINATE,
+	    &statePtr->inputEncodingState, dst,
+	    eol - dst + skip + TCL_UTF_MAX - 1, &gs.rawRead, NULL,
+	    &gs.charsWrote);
+    bufPtr->nextRemoved += gs.rawRead;
+
+    /*
+     * Recycle all the emptied buffers.
+     */
+
+    Tcl_SetObjLength(objPtr, eol - objPtr->bytes);
+    CommonGetsCleanup(chanPtr);
+    ResetFlag(statePtr, CHANNEL_BLOCKED);
+    copiedTotal = gs.totalChars + gs.charsWrote - skip;
+    goto done;
+
+    /*
+     * Couldn't get a complete line. This only happens if we get a error
+     * reading from the channel or we are non-blocking and there wasn't an EOL
+     * or EOF in the data available.
+     */
+
+  restore:
+    /*
+     * Regenerate the top channel, in case it was changed due to
+     * self-modifying reflected transforms.
+     */
+    if (chanPtr != statePtr->topChanPtr) {
+	TclChannelRelease((Tcl_Channel)chanPtr);
+	chanPtr = statePtr->topChanPtr;
+	TclChannelPreserve((Tcl_Channel)chanPtr);
+    }
+    bufPtr = statePtr->inQueueHead;
+    if (bufPtr != NULL) {
+	bufPtr->nextRemoved = oldRemoved;
+	bufPtr = bufPtr->nextPtr;
+    }
+
+    for ( ; bufPtr != NULL; bufPtr = bufPtr->nextPtr) {
+	bufPtr->nextRemoved = BUFFER_PADDING;
+    }
+    CommonGetsCleanup(chanPtr);
+
+    statePtr->inputEncodingState = oldState;
+    statePtr->inputEncodingFlags = oldFlags;
+    Tcl_SetObjLength(objPtr, oldLength);
+
+    /*
+     * We didn't get a complete line so we need to indicate to UpdateInterest
+     * that the gets blocked. It will wait for more data instead of firing a
+     * timer, avoiding a busy wait. This is where we are assuming that the
+     * next operation is a gets. No more file events will be delivered on this
+     * channel until new data arrives or some operation is performed on the
+     * channel (e.g. gets, read, fconfigure) that changes the blocking state.
+     * Note that this means a file event will not be delivered even though a
+     * read would be able to consume the buffered data.
+     */
+
+    SetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+    copiedTotal = -1;
+
+    /*
+     * Update the notifier state so we don't block while there is still data
+     * in the buffers.
+     */
+
+  done:
+    assert(!GotFlag(statePtr, CHANNEL_EOF)
+	    || GotFlag(statePtr, CHANNEL_STICKY_EOF)
+	    || Tcl_InputBuffered((Tcl_Channel)chanPtr) == 0);
+    assert(!(GotFlag(statePtr, CHANNEL_EOF|CHANNEL_BLOCKED)
+	    == (CHANNEL_EOF|CHANNEL_BLOCKED)));
+
+    /*
+     * Regenerate the top channel, in case it was changed due to
+     * self-modifying reflected transforms.
+     */
+
+    if (chanPtr != statePtr->topChanPtr) {
+	TclChannelRelease((Tcl_Channel)chanPtr);
+	chanPtr = statePtr->topChanPtr;
+	TclChannelPreserve((Tcl_Channel)chanPtr);
+    }
+    UpdateInterest(chanPtr);
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    return copiedTotal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetsObjBinary --
+ *
+ *	A variation of Tcl_GetsObj that works directly on the buffers until
+ *	end-of-line or end-of-file has been seen. Bytes read from the input
+ *	channel return as a ByteArray obj.
+ *
+ *	WARNING!  The notion of "binary" used here is different from notions
+ *	of "binary" used in other places. In particular, this "binary" routine
+ *	may be called when an -eofchar is set on the channel.
+ *
+ * Results:
+ *	Number of characters accumulated in the object or -1 if error,
+ *	blocked, or EOF. If -1, use Tcl_GetErrno() to retrieve the POSIX error
+ *	code for the error or condition that occurred.
+ *
+ * Side effects:
+ *	Consumes input from the channel.
+ *
+ *	On reading EOF, leave channel pointing at EOF char. On reading EOL,
+ *	leave channel pointing after EOL, but don't return EOL in dst buffer.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+TclGetsObjBinary(
+    Tcl_Channel chan,		/* Channel from which to read. */
+    Tcl_Obj *objPtr)		/* The line read will be appended to this
+				 * object as UTF-8 characters. */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr;
+    int inEofChar, skip, copiedTotal, oldLength, oldFlags, oldRemoved;
+    int rawLen, byteLen, eolChar;
+    unsigned char *dst, *dstEnd, *eol, *eof, *byteArray;
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+
+    bufPtr = statePtr->inQueueHead;
+
+    /*
+     * Preserved so we can restore the channel's state in case we don't find a
+     * newline in the available input.
+     */
+
+    byteArray = Tcl_GetByteArrayFromObj(objPtr, &byteLen);
+    oldFlags = statePtr->inputEncodingFlags;
+    oldRemoved = BUFFER_PADDING;
+    oldLength = byteLen;
+    if (bufPtr != NULL) {
+	oldRemoved = bufPtr->nextRemoved;
+    }
+
+    rawLen = 0;
+    skip = 0;
+    eof = NULL;
+    inEofChar = statePtr->inEofChar;
+
+    /*
+     * Only handle TCL_TRANSLATE_LF and TCL_TRANSLATE_CR.
+     */
+
+    eolChar = (statePtr->inputTranslation == TCL_TRANSLATE_LF) ? '\n' : '\r';
+
+    ResetFlag(statePtr, CHANNEL_BLOCKED);
+    while (1) {
+	/*
+	 * Subtract the number of bytes that were removed from channel buffer
+	 * during last call.
+	 */
+
+	if (bufPtr != NULL) {
+	    bufPtr->nextRemoved += rawLen;
+	    if (!IsBufferReady(bufPtr)) {
+		bufPtr = bufPtr->nextPtr;
+	    }
+	}
+
+	if ((bufPtr == NULL) || (bufPtr->nextAdded == BUFFER_PADDING)) {
+	    /*
+	     * All channel buffers were exhausted and the caller still hasn't
+	     * seen EOL. Need to read more bytes from the channel device. Side
+	     * effect is to allocate another channel buffer.
+	     */
+
+	    if (GetInput(chanPtr) != 0) {
+		goto restore;
+	    }
+	    bufPtr = statePtr->inQueueTail;
+	    if (bufPtr == NULL) {
+		goto restore;
+	    }
+	} else {
+	    /*
+	     * Incoming CHANNEL_STICKY_EOF is filtered out on entry.  A new
+	     * CHANNEL_STICKY_EOF set in this routine leads to return before
+	     * coming back here.  When we are not dealing with
+	     * CHANNEL_STICKY_EOF, a CHANNEL_EOF implies an empty buffer.
+	     * Here the buffer is non-empty so we know we're a non-EOF.
+             */
+
+	    assert(!GotFlag(statePtr, CHANNEL_STICKY_EOF));
+	    assert(!GotFlag(statePtr, CHANNEL_EOF));
+	}
+
+	dst = (unsigned char *) RemovePoint(bufPtr);
+	dstEnd = dst + BytesLeft(bufPtr);
+
+	/*
+	 * Remember if EOF char is seen, then look for EOL anyhow, because the
+	 * EOL might be before the EOF char.
+	 * XXX - in the binary case, consider coincident search for eol/eof.
+	 */
+
+	if (inEofChar != '\0') {
+	    for (eol = dst; eol < dstEnd; eol++) {
+		if (*eol == inEofChar) {
+		    dstEnd = eol;
+		    eof = eol;
+		    break;
+		}
+	    }
+	}
+
+	/*
+	 * On EOL, leave current file position pointing after the EOL, but
+	 * don't store the EOL in the output string.
+	 */
+
+	for (eol = dst; eol < dstEnd; eol++) {
+	    if (*eol == eolChar) {
+		skip = 1;
+		goto gotEOL;
+	    }
+	}
+	if (eof != NULL) {
+	    /*
+	     * EOF character was seen. On EOF, leave current file position
+	     * pointing at the EOF character, but don't store the EOF
+	     * character in the output string.
+	     */
+
+	    SetFlag(statePtr, CHANNEL_EOF | CHANNEL_STICKY_EOF);
+	    statePtr->inputEncodingFlags |= TCL_ENCODING_END;
+	    ResetFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR);
+	}
+	if (GotFlag(statePtr, CHANNEL_EOF)) {
+	    skip = 0;
+	    eol = dstEnd;
+	    if ((dst == dstEnd) && (byteLen == oldLength)) {
+		/*
+		 * If we didn't append any bytes before encountering EOF,
+		 * caller needs to see -1.
+		 */
+
+		byteArray = Tcl_SetByteArrayLength(objPtr, oldLength);
+		CommonGetsCleanup(chanPtr);
+		copiedTotal = -1;
+		ResetFlag(statePtr, CHANNEL_BLOCKED);
+		goto done;
+	    }
+	    goto gotEOL;
+	}
+	if (GotFlag(statePtr, CHANNEL_BLOCKED|CHANNEL_NONBLOCKING)
+		== (CHANNEL_BLOCKED|CHANNEL_NONBLOCKING)) {
+	    goto restore;
+	}
+
+	/*
+	 * Copy bytes from the channel buffer to the ByteArray. This may
+	 * realloc space, so keep track of result.
+	 */
+
+	rawLen = dstEnd - dst;
+	byteArray = Tcl_SetByteArrayLength(objPtr, byteLen + rawLen);
+	memcpy(byteArray + byteLen, dst, (size_t) rawLen);
+	byteLen += rawLen;
+    }
+
+    /*
+     * Found EOL or EOF, but the output buffer may now contain too many bytes.
+     * We need to know how many bytes correspond to the number we want, so we
+     * can remove the correct number of bytes from the channel buffer.
+     */
+
+  gotEOL:
+    if (bufPtr == NULL) {
+	Tcl_Panic("TclGetsObjBinary: gotEOL reached with bufPtr==NULL");
+    }
+
+    rawLen = eol - dst;
+    byteArray = Tcl_SetByteArrayLength(objPtr, byteLen + rawLen);
+    memcpy(byteArray + byteLen, dst, (size_t) rawLen);
+    byteLen += rawLen;
+    bufPtr->nextRemoved += rawLen + skip;
+
+    /*
+     * Convert the buffer if there was an encoding.
+     * XXX - unimplemented.
+     */
+
+    if (statePtr->encoding != NULL) {
+    }
+
+    /*
+     * Recycle all the emptied buffers.
+     */
+
+    CommonGetsCleanup(chanPtr);
+    ResetFlag(statePtr, CHANNEL_BLOCKED);
+    copiedTotal = byteLen;
+    goto done;
+
+    /*
+     * Couldn't get a complete line. This only happens if we get a error
+     * reading from the channel or we are non-blocking and there wasn't an EOL
+     * or EOF in the data available.
+     */
+
+  restore:
+    bufPtr = statePtr->inQueueHead;
+    if (bufPtr) {
+	bufPtr->nextRemoved = oldRemoved;
+	bufPtr = bufPtr->nextPtr;
+    }
+
+    for ( ; bufPtr != NULL; bufPtr = bufPtr->nextPtr) {
+	bufPtr->nextRemoved = BUFFER_PADDING;
+    }
+    CommonGetsCleanup(chanPtr);
+
+    statePtr->inputEncodingFlags = oldFlags;
+    byteArray = Tcl_SetByteArrayLength(objPtr, oldLength);
+
+    /*
+     * We didn't get a complete line so we need to indicate to UpdateInterest
+     * that the gets blocked. It will wait for more data instead of firing a
+     * timer, avoiding a busy wait. This is where we are assuming that the
+     * next operation is a gets. No more file events will be delivered on this
+     * channel until new data arrives or some operation is performed on the
+     * channel (e.g. gets, read, fconfigure) that changes the blocking state.
+     * Note that this means a file event will not be delivered even though a
+     * read would be able to consume the buffered data.
+     */
+
+    SetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+    copiedTotal = -1;
+
+    /*
+     * Update the notifier state so we don't block while there is still data
+     * in the buffers.
+     */
+
+  done:
+    assert(!GotFlag(statePtr, CHANNEL_EOF)
+	    || GotFlag(statePtr, CHANNEL_STICKY_EOF)
+	    || Tcl_InputBuffered((Tcl_Channel)chanPtr) == 0);
+    assert(!(GotFlag(statePtr, CHANNEL_EOF|CHANNEL_BLOCKED)
+	    == (CHANNEL_EOF|CHANNEL_BLOCKED)));
+    UpdateInterest(chanPtr);
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    return copiedTotal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FreeBinaryEncoding --
+ *
+ *	Frees any "iso8859-1" Tcl_Encoding created by [gets] on a binary
+ *	channel in a thread as part of that thread's finalization.
+ *
+ * Results:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+FreeBinaryEncoding(
+    ClientData dummy)	/* Not used */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->binaryEncoding != NULL) {
+	Tcl_FreeEncoding(tsdPtr->binaryEncoding);
+	tsdPtr->binaryEncoding = NULL;
+    }
+}
+
+static Tcl_Encoding
+GetBinaryEncoding()
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->binaryEncoding == NULL) {
+	tsdPtr->binaryEncoding = Tcl_GetEncoding(NULL, "iso8859-1");
+	Tcl_CreateThreadExitHandler(FreeBinaryEncoding, NULL);
+    }
+    if (tsdPtr->binaryEncoding == NULL) {
+	Tcl_Panic("binary encoding is not available");
+    }
+    return tsdPtr->binaryEncoding;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FilterInputBytes --
+ *
+ *	Helper function for Tcl_GetsObj. Produces UTF-8 characters from raw
+ *	bytes read from the channel.
+ *
+ *	Consumes available bytes from channel buffers. When channel buffers
+ *	are exhausted, reads more bytes from channel device into a new channel
+ *	buffer. It is the caller's responsibility to free the channel buffers
+ *	that have been exhausted.
+ *
+ * Results:
+ *	The return value is -1 if there was an error reading from the channel,
+ *	0 otherwise.
+ *
+ * Side effects:
+ *	Status object keeps track of how much data from channel buffers has
+ *	been consumed and where UTF-8 bytes should be stored.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+FilterInputBytes(
+    Channel *chanPtr,		/* Channel to read. */
+    GetsState *gsPtr)		/* Current state of gets operation. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr;
+    char *raw, *dst;
+    int offset, toRead, dstNeeded, spaceLeft, result, rawLen;
+    Tcl_Obj *objPtr;
+#define ENCODING_LINESIZE 20	/* Lower bound on how many bytes to convert at
+				 * a time. Since we don't know a priori how
+				 * many bytes of storage this many source
+				 * bytes will use, we actually need at least
+				 * ENCODING_LINESIZE * TCL_MAX_UTF bytes of
+				 * room. */
+
+    objPtr = gsPtr->objPtr;
+
+    /*
+     * Subtract the number of bytes that were removed from channel buffer
+     * during last call.
+     */
+
+    bufPtr = gsPtr->bufPtr;
+    if (bufPtr != NULL) {
+	bufPtr->nextRemoved += gsPtr->rawRead;
+	if (!IsBufferReady(bufPtr)) {
+	    bufPtr = bufPtr->nextPtr;
+	}
+    }
+    gsPtr->totalChars += gsPtr->charsWrote;
+
+    if ((bufPtr == NULL) || (bufPtr->nextAdded == BUFFER_PADDING)) {
+	/*
+	 * All channel buffers were exhausted and the caller still hasn't seen
+	 * EOL. Need to read more bytes from the channel device. Side effect
+	 * is to allocate another channel buffer.
+	 */
+
+    read:
+	if (GotFlag(statePtr, CHANNEL_NONBLOCKING|CHANNEL_BLOCKED)
+		== (CHANNEL_NONBLOCKING|CHANNEL_BLOCKED)) {
+	    gsPtr->charsWrote = 0;
+	    gsPtr->rawRead = 0;
+	    return -1;
+	}
+	if (GetInput(chanPtr) != 0) {
+	    gsPtr->charsWrote = 0;
+	    gsPtr->rawRead = 0;
+	    return -1;
+	}
+	bufPtr = statePtr->inQueueTail;
+	gsPtr->bufPtr = bufPtr;
+	if (bufPtr == NULL) {
+	    gsPtr->charsWrote = 0;
+	    gsPtr->rawRead = 0;
+	    return -1;
+	}
+    } else {
+	/*
+	 * Incoming CHANNEL_STICKY_EOF is filtered out on entry.  A new
+	 * CHANNEL_STICKY_EOF set in this routine leads to return before
+	 * coming back here.  When we are not dealing with CHANNEL_STICKY_EOF,
+	 * a CHANNEL_EOF implies an empty buffer.  Here the buffer is
+	 * non-empty so we know we're a non-EOF.
+         */
+
+	assert(!GotFlag(statePtr, CHANNEL_STICKY_EOF));
+	assert(!GotFlag(statePtr, CHANNEL_EOF));
+    }
+
+    /*
+     * Convert some of the bytes from the channel buffer to UTF-8. Space in
+     * objPtr's string rep is used to hold the UTF-8 characters. Grow the
+     * string rep if we need more space.
+     */
+
+    raw = RemovePoint(bufPtr);
+    rawLen = BytesLeft(bufPtr);
+
+    dst = *gsPtr->dstPtr;
+    offset = dst - objPtr->bytes;
+    toRead = ENCODING_LINESIZE;
+    if (toRead > rawLen) {
+	toRead = rawLen;
+    }
+    dstNeeded = toRead * TCL_UTF_MAX;
+    spaceLeft = objPtr->length - offset;
+    if (dstNeeded > spaceLeft) {
+	int length = offset + ((offset < dstNeeded) ? dstNeeded : offset);
+
+	if (Tcl_AttemptSetObjLength(objPtr, length) == 0) {
+	    length = offset + dstNeeded;
+	    if (Tcl_AttemptSetObjLength(objPtr, length) == 0) {
+		dstNeeded = TCL_UTF_MAX - 1 + toRead;
+		length = offset + dstNeeded;
+		Tcl_SetObjLength(objPtr, length);
+	    }
+	}
+	spaceLeft = length - offset;
+	dst = objPtr->bytes + offset;
+	*gsPtr->dstPtr = dst;
+    }
+    gsPtr->state = statePtr->inputEncodingState;
+    result = Tcl_ExternalToUtf(NULL, gsPtr->encoding, raw, rawLen,
+	    statePtr->inputEncodingFlags | TCL_ENCODING_NO_TERMINATE,
+	    &statePtr->inputEncodingState, dst, spaceLeft, &gsPtr->rawRead,
+	    &gsPtr->bytesWrote, &gsPtr->charsWrote);
+
+    /*
+     * Make sure that if we go through 'gets', that we reset the
+     * TCL_ENCODING_START flag still. [Bug #523988]
+     */
+
+    statePtr->inputEncodingFlags &= ~TCL_ENCODING_START;
+
+    if (result == TCL_CONVERT_MULTIBYTE) {
+	/*
+	 * The last few bytes in this channel buffer were the start of a
+	 * multibyte sequence. If this buffer was full, then move them to the
+	 * next buffer so the bytes will be contiguous.
+	 */
+
+	ChannelBuffer *nextPtr;
+	int extra;
+
+	nextPtr = bufPtr->nextPtr;
+	if (!IsBufferFull(bufPtr)) {
+	    if (gsPtr->rawRead > 0) {
+		/*
+		 * Some raw bytes were converted to UTF-8. Fall through,
+		 * returning those UTF-8 characters because a EOL might be
+		 * present in them.
+		 */
+	    } else if (GotFlag(statePtr, CHANNEL_EOF)) {
+		/*
+		 * There was a partial character followed by EOF on the
+		 * device. Fall through, returning that nothing was found.
+		 */
+
+		bufPtr->nextRemoved = bufPtr->nextAdded;
+	    } else {
+		/*
+		 * There are no more cached raw bytes left. See if we can get
+		 * some more, but avoid blocking on a non-blocking channel.
+		 */
+
+		goto read;
+	    }
+	} else {
+	    if (nextPtr == NULL) {
+		nextPtr = AllocChannelBuffer(statePtr->bufSize);
+		bufPtr->nextPtr = nextPtr;
+		statePtr->inQueueTail = nextPtr;
+	    }
+	    extra = rawLen - gsPtr->rawRead;
+	    memcpy(nextPtr->buf + (BUFFER_PADDING - extra),
+		    raw + gsPtr->rawRead, (size_t) extra);
+	    nextPtr->nextRemoved -= extra;
+	    bufPtr->nextAdded -= extra;
+	}
+    }
+
+    gsPtr->bufPtr = bufPtr;
+    return 0;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * PeekAhead --
+ *
+ *	Helper function used by Tcl_GetsObj(). Called when we've seen a \r at
+ *	the end of the UTF-8 string and want to look ahead one character to
+ *	see if it is a \n.
+ *
+ * Results:
+ *	*gsPtr->dstPtr is filled with a pointer to the start of the range of
+ *	UTF-8 characters that were found by peeking and *dstEndPtr is filled
+ *	with a pointer to the bytes just after the end of the range.
+ *
+ * Side effects:
+ *	If no more raw bytes were available in one of the channel buffers,
+ *	tries to perform a non-blocking read to get more bytes from the
+ *	channel device.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+PeekAhead(
+    Channel *chanPtr,		/* The channel to read. */
+    char **dstEndPtr,		/* Filled with pointer to end of new range of
+				 * UTF-8 characters. */
+    GetsState *gsPtr)		/* Current state of gets operation. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr;
+    Tcl_DriverBlockModeProc *blockModeProc;
+    int bytesLeft;
+
+    bufPtr = gsPtr->bufPtr;
+
+    /*
+     * If there's any more raw input that's still buffered, we'll peek into
+     * that. Otherwise, only get more data from the channel driver if it looks
+     * like there might actually be more data. The assumption is that if the
+     * channel buffer is filled right up to the end, then there might be more
+     * data to read.
+     */
+
+    blockModeProc = NULL;
+    if (bufPtr->nextPtr == NULL) {
+	bytesLeft = BytesLeft(bufPtr) - gsPtr->rawRead;
+	if (bytesLeft == 0) {
+	    if (!IsBufferFull(bufPtr)) {
+		/*
+		 * Don't peek ahead if last read was short read.
+		 */
+
+		goto cleanup;
+	    }
+	    if (!GotFlag(statePtr, CHANNEL_NONBLOCKING)) {
+		blockModeProc = Tcl_ChannelBlockModeProc(chanPtr->typePtr);
+		if (blockModeProc == NULL) {
+		    /*
+		     * Don't peek ahead if cannot set non-blocking mode.
+		     */
+
+		    goto cleanup;
+		}
+		StackSetBlockMode(chanPtr, TCL_MODE_NONBLOCKING);
+	    }
+	}
+    }
+    if (FilterInputBytes(chanPtr, gsPtr) == 0) {
+	*dstEndPtr = *gsPtr->dstPtr + gsPtr->bytesWrote;
+    }
+    if (blockModeProc != NULL) {
+	StackSetBlockMode(chanPtr, TCL_MODE_BLOCKING);
+    }
+    return;
+
+  cleanup:
+    bufPtr->nextRemoved += gsPtr->rawRead;
+    gsPtr->rawRead = 0;
+    gsPtr->totalChars += gsPtr->charsWrote;
+    gsPtr->bytesWrote = 0;
+    gsPtr->charsWrote = 0;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * CommonGetsCleanup --
+ *
+ *	Helper function for Tcl_GetsObj() to restore the channel after a
+ *	"gets" operation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Encoding may be freed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+CommonGetsCleanup(
+    Channel *chanPtr)
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr, *nextPtr;
+
+    bufPtr = statePtr->inQueueHead;
+    for ( ; bufPtr != NULL; bufPtr = nextPtr) {
+	nextPtr = bufPtr->nextPtr;
+	if (IsBufferReady(bufPtr)) {
+	    break;
+	}
+	RecycleBuffer(statePtr, bufPtr, 0);
+    }
+    statePtr->inQueueHead = bufPtr;
+    if (bufPtr == NULL) {
+	statePtr->inQueueTail = NULL;
+    } else {
+	/*
+	 * If any multi-byte characters were split across channel buffer
+	 * boundaries, the split-up bytes were moved to the next channel
+	 * buffer by FilterInputBytes(). Move the bytes back to their original
+	 * buffer because the caller could change the channel's encoding which
+	 * could change the interpretation of whether those bytes really made
+	 * up multi-byte characters after all.
+	 */
+
+	nextPtr = bufPtr->nextPtr;
+	for ( ; nextPtr != NULL; nextPtr = bufPtr->nextPtr) {
+	    int extra;
+
+	    extra = SpaceLeft(bufPtr);
+	    if (extra > 0) {
+		memcpy(InsertPoint(bufPtr),
+			nextPtr->buf + (BUFFER_PADDING - extra),
+			(size_t) extra);
+		bufPtr->nextAdded += extra;
+		nextPtr->nextRemoved = BUFFER_PADDING;
+	    }
+	    bufPtr = nextPtr;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Read --
+ *
+ *	Reads a given number of bytes from a channel. EOL and EOF translation
+ *	is done on the bytes being read, so the number of bytes consumed from
+ *	the channel may not be equal to the number of bytes stored in the
+ *	destination buffer.
+ *
+ *	No encoding conversions are applied to the bytes being read.
+ *
+ * Results:
+ *	The number of bytes read, or -1 on error. Use Tcl_GetErrno() to
+ *	retrieve the error code for the error that occurred.
+ *
+ * Side effects:
+ *	May cause input to be buffered.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Read(
+    Tcl_Channel chan,		/* The channel from which to read. */
+    char *dst,			/* Where to store input read. */
+    int bytesToRead)		/* Maximum number of bytes to read. */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_READABLE) != 0) {
+	return -1;
+    }
+
+    return DoRead(chanPtr, dst, bytesToRead, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ReadRaw --
+ *
+ *	Reads a given number of bytes from a channel. EOL and EOF translation
+ *	is done on the bytes being read, so the number of bytes consumed from
+ *	the channel may not be equal to the number of bytes stored in the
+ *	destination buffer.
+ *
+ *	No encoding conversions are applied to the bytes being read.
+ *
+ * Results:
+ *	The number of bytes read, or -1 on error. Use Tcl_GetErrno() to
+ *	retrieve the error code for the error that occurred.
+ *
+ * Side effects:
+ *	May cause input to be buffered.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ReadRaw(
+    Tcl_Channel chan,		/* The channel from which to read. */
+    char *readBuf,		/* Where to store input read. */
+    int bytesToRead)		/* Maximum number of bytes to read. */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int copied = 0;
+
+    assert(bytesToRead > 0);
+    if (CheckChannelErrors(statePtr, TCL_READABLE | CHANNEL_RAW_MODE) != 0) {
+	return -1;
+    }
+
+    /*
+     * First read bytes from the push-back buffers.
+     */
+
+    while (chanPtr->inQueueHead && bytesToRead > 0) {
+	ChannelBuffer *bufPtr = chanPtr->inQueueHead;
+	int bytesInBuffer = BytesLeft(bufPtr);
+	int toCopy = (bytesInBuffer < bytesToRead) ? bytesInBuffer
+		: bytesToRead;
+
+	/*
+         * Copy the current chunk into the read buffer.
+         */
+
+	memcpy(readBuf, RemovePoint(bufPtr), (size_t) toCopy);
+	bufPtr->nextRemoved += toCopy;
+	copied += toCopy;
+	readBuf += toCopy;
+	bytesToRead -= toCopy;
+
+	/*
+         * If the current buffer is empty recycle it.
+         */
+
+	if (IsBufferEmpty(bufPtr)) {
+	    chanPtr->inQueueHead = bufPtr->nextPtr;
+	    if (chanPtr->inQueueHead == NULL) {
+		chanPtr->inQueueTail = NULL;
+	    }
+	    RecycleBuffer(chanPtr->state, bufPtr, 0);
+	}
+    }
+
+    /*
+     * Go to the driver only if we got nothing from pushback.  Have to do it
+     * this way to avoid EOF mis-timings when we consider the ability that EOF
+     * may not be a permanent condition in the driver, and in that case we
+     * have to synchronize.
+     */
+
+    if (copied) {
+	return copied;
+    }
+
+    /*
+     * This test not needed.
+     */
+
+    if (bytesToRead > 0) {
+	int nread = ChanRead(chanPtr, readBuf, bytesToRead);
+
+	if (nread > 0) {
+	    /*
+             * Successful read (short is OK) - add to bytes copied.
+             */
+
+	    copied += nread;
+	} else if (nread < 0) {
+	    /*
+	     * An error signaled.  If CHANNEL_BLOCKED, then the error is not
+	     * real, but an indication of blocked state.  In that case, retain
+	     * the flag and let caller receive the short read of copied bytes
+	     * from the pushback.  HOWEVER, if copied==0 bytes from pushback
+	     * then repeat signalling the blocked state as an error to caller
+	     * so there is no false report of an EOF.  When !CHANNEL_BLOCKED,
+	     * the error is real and passes on to caller.
+	     */
+
+	    if (!GotFlag(statePtr, CHANNEL_BLOCKED) || copied == 0) {
+		copied = -1;
+	    }
+	} else {
+	    /*
+	     * nread == 0.  Driver is at EOF. Let that state filter up.
+	     */
+	}
+    }
+    return copied;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_ReadChars --
+ *
+ *	Reads from the channel until the requested number of characters have
+ *	been seen, EOF is seen, or the channel would block. EOL and EOF
+ *	translation is done. If reading binary data, the raw bytes are wrapped
+ *	in a Tcl byte array object. Otherwise, the raw bytes are converted to
+ *	UTF-8 using the channel's current encoding and stored in a Tcl string
+ *	object.
+ *
+ * Results:
+ *	The number of characters read, or -1 on error. Use Tcl_GetErrno() to
+ *	retrieve the error code for the error that occurred.
+ *
+ * Side effects:
+ *	May cause input to be buffered.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_ReadChars(
+    Tcl_Channel chan,		/* The channel to read. */
+    Tcl_Obj *objPtr,		/* Input data is stored in this object. */
+    int toRead,			/* Maximum number of characters to store, or
+				 * -1 to read all available data (up to EOF or
+				 * when channel blocks). */
+    int appendFlag)		/* If non-zero, data read from the channel
+				 * will be appended to the object. Otherwise,
+				 * the data will replace the existing contents
+				 * of the object. */
+{
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_READABLE) != 0) {
+	/*
+	 * Update the notifier state so we don't block while there is still
+	 * data in the buffers.
+	 */
+
+	UpdateInterest(chanPtr);
+	return -1;
+    }
+
+    return DoReadChars(chanPtr, objPtr, toRead, appendFlag);
+}
+/*
+ *---------------------------------------------------------------------------
+ *
+ * DoReadChars --
+ *
+ *	Reads from the channel until the requested number of characters have
+ *	been seen, EOF is seen, or the channel would block. EOL and EOF
+ *	translation is done. If reading binary data, the raw bytes are wrapped
+ *	in a Tcl byte array object. Otherwise, the raw bytes are converted to
+ *	UTF-8 using the channel's current encoding and stored in a Tcl string
+ *	object.
+ *
+ * Results:
+ *	The number of characters read, or -1 on error. Use Tcl_GetErrno() to
+ *	retrieve the error code for the error that occurred.
+ *
+ * Side effects:
+ *	May cause input to be buffered.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+DoReadChars(
+    Channel *chanPtr,		/* The channel to read. */
+    Tcl_Obj *objPtr,		/* Input data is stored in this object. */
+    int toRead,			/* Maximum number of characters to store, or
+				 * -1 to read all available data (up to EOF or
+				 * when channel blocks). */
+    int appendFlag)		/* If non-zero, data read from the channel
+				 * will be appended to the object. Otherwise,
+				 * the data will replace the existing contents
+				 * of the object. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelBuffer *bufPtr;
+    int copied, copiedNow, result;
+    Tcl_Encoding encoding = statePtr->encoding;
+    int binaryMode;
+#define UTF_EXPANSION_FACTOR	1024
+    int factor = UTF_EXPANSION_FACTOR;
+
+    binaryMode = (encoding == NULL)
+	    && (statePtr->inputTranslation == TCL_TRANSLATE_LF)
+	    && (statePtr->inEofChar == '\0');
+
+    if (appendFlag == 0) {
+	if (binaryMode) {
+	    Tcl_SetByteArrayLength(objPtr, 0);
+	} else {
+	    Tcl_SetObjLength(objPtr, 0);
+
+	    /*
+	     * We're going to access objPtr->bytes directly, so we must ensure
+	     * that this is actually a string object (otherwise it might have
+	     * been pure Unicode).
+	     *
+	     * Probably not needed anymore.
+	     */
+
+	    TclGetString(objPtr);
+	}
+    }
+
+    /*
+     * Early out when next read will see eofchar.
+     *
+     * NOTE: See DoRead for argument that it's a bug (one we're keeping) to
+     * have this escape before the one for zero-char read request.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_STICKY_EOF)) {
+	SetFlag(statePtr, CHANNEL_EOF);
+	assert(statePtr->inputEncodingFlags & TCL_ENCODING_END);
+	assert(!GotFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR));
+
+	/* TODO: We don't need this call? */
+	UpdateInterest(chanPtr);
+	return 0;
+    }
+
+    /*
+     * Special handling for zero-char read request.
+     */
+    if (toRead == 0) {
+	if (GotFlag(statePtr, CHANNEL_EOF)) {
+	    statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+	}
+	ResetFlag(statePtr, CHANNEL_BLOCKED|CHANNEL_EOF);
+	statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+	/* TODO: We don't need this call? */
+	UpdateInterest(chanPtr);
+	return 0;
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+
+    /*
+     * Must clear the BLOCKED|EOF flags here since we check before reading.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_EOF)) {
+	statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+    }
+    ResetFlag(statePtr, CHANNEL_BLOCKED|CHANNEL_EOF);
+    statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+    for (copied = 0; (unsigned) toRead > 0; ) {
+	copiedNow = -1;
+	if (statePtr->inQueueHead != NULL) {
+	    if (binaryMode) {
+		copiedNow = ReadBytes(statePtr, objPtr, toRead);
+	    } else {
+		copiedNow = ReadChars(statePtr, objPtr, toRead, &factor);
+	    }
+
+	    /*
+	     * If the current buffer is empty recycle it.
+	     */
+
+	    bufPtr = statePtr->inQueueHead;
+	    if (IsBufferEmpty(bufPtr)) {
+		ChannelBuffer *nextPtr = bufPtr->nextPtr;
+
+		RecycleBuffer(statePtr, bufPtr, 0);
+		statePtr->inQueueHead = nextPtr;
+		if (nextPtr == NULL) {
+		    statePtr->inQueueTail = NULL;
+		}
+	    }
+	}
+
+	if (copiedNow < 0) {
+	    if (GotFlag(statePtr, CHANNEL_EOF)) {
+		break;
+	    }
+	    if (GotFlag(statePtr, CHANNEL_NONBLOCKING|CHANNEL_BLOCKED)
+		    == (CHANNEL_NONBLOCKING|CHANNEL_BLOCKED)) {
+		break;
+	    }
+	    result = GetInput(chanPtr);
+	    if (chanPtr != statePtr->topChanPtr) {
+		TclChannelRelease((Tcl_Channel)chanPtr);
+		chanPtr = statePtr->topChanPtr;
+		TclChannelPreserve((Tcl_Channel)chanPtr);
+	    }
+	    if (result != 0) {
+		if (!GotFlag(statePtr, CHANNEL_BLOCKED)) {
+		    copied = -1;
+		}
+		break;
+	    }
+	} else {
+	    copied += copiedNow;
+	    toRead -= copiedNow;
+	}
+    }
+
+    /*
+     * Failure to fill a channel buffer may have left channel reporting a
+     * "blocked" state, but so long as we fulfilled the request here, the
+     * caller does not consider us blocked.
+     */
+
+    if (toRead == 0) {
+	ResetFlag(statePtr, CHANNEL_BLOCKED);
+    }
+
+    /*
+     * Regenerate the top channel, in case it was changed due to
+     * self-modifying reflected transforms.
+     */
+
+    if (chanPtr != statePtr->topChanPtr) {
+	TclChannelRelease((Tcl_Channel)chanPtr);
+	chanPtr = statePtr->topChanPtr;
+	TclChannelPreserve((Tcl_Channel)chanPtr);
+    }
+
+    /*
+     * Update the notifier state so we don't block while there is still data
+     * in the buffers.
+     */
+
+    assert(!GotFlag(statePtr, CHANNEL_EOF)
+	    || GotFlag(statePtr, CHANNEL_STICKY_EOF)
+	    || Tcl_InputBuffered((Tcl_Channel)chanPtr) == 0);
+    assert(!(GotFlag(statePtr, CHANNEL_EOF|CHANNEL_BLOCKED)
+            == (CHANNEL_EOF|CHANNEL_BLOCKED)));
+    UpdateInterest(chanPtr);
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    return copied;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * ReadBytes --
+ *
+ *	Reads from the channel until the requested number of bytes have been
+ *	seen, EOF is seen, or the channel would block. Bytes from the channel
+ *	are stored in objPtr as a ByteArray object. EOL and EOF translation
+ *	are done.
+ *
+ *	'bytesToRead' can safely be a very large number because space is only
+ *	allocated to hold data read from the channel as needed.
+ *
+ * Results:
+ *	The return value is the number of bytes appended to the object, or
+ *	-1 to indicate that zero bytes were read due to an EOF.
+ *
+ * Side effects:
+ *	The storage of bytes in objPtr can cause (re-)allocation of memory.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+ReadBytes(
+    ChannelState *statePtr,	/* State of the channel to read. */
+    Tcl_Obj *objPtr,		/* Input data is appended to this ByteArray
+				 * object. Its length is how much space has
+				 * been allocated to hold data, not how many
+				 * bytes of data have been stored in the
+				 * object. */
+    int bytesToRead)		/* Maximum number of bytes to store, or < 0 to
+				 * get all available bytes. Bytes are obtained
+				 * from the first buffer in the queue - even
+				 * if this number is larger than the number of
+				 * bytes available in the first buffer, only
+				 * the bytes from the first buffer are
+				 * returned. */
+{
+    ChannelBuffer *bufPtr = statePtr->inQueueHead;
+    int srcLen = BytesLeft(bufPtr);
+    int toRead = bytesToRead>srcLen || bytesToRead<0 ? srcLen : bytesToRead;
+
+    TclAppendBytesToByteArray(objPtr, (unsigned char *) RemovePoint(bufPtr),
+	    toRead);
+    bufPtr->nextRemoved += toRead;
+    return toRead;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * ReadChars --
+ *
+ *	Reads from the channel until the requested number of UTF-8 characters
+ *	have been seen, EOF is seen, or the channel would block. Raw bytes
+ *	from the channel are converted to UTF-8 and stored in objPtr. EOL and
+ *	EOF translation is done.
+ *
+ *	'charsToRead' can safely be a very large number because space is only
+ *	allocated to hold data read from the channel as needed.
+ *
+ *	'charsToRead' may *not* be 0.
+ *
+ * Results:
+ *	The return value is the number of characters appended to the object,
+ *	*offsetPtr is filled with the number of bytes that were appended, and
+ *	*factorPtr is filled with the expansion factor used to guess how many
+ *	bytes of UTF-8 to allocate to hold N source bytes.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+ReadChars(
+    ChannelState *statePtr,	/* State of channel to read. */
+    Tcl_Obj *objPtr,		/* Input data is appended to this object.
+				 * objPtr->length is how much space has been
+				 * allocated to hold data, not how many bytes
+				 * of data have been stored in the object. */
+    int charsToRead,		/* Maximum number of characters to store, or
+				 * -1 to get all available characters.
+				 * Characters are obtained from the first
+				 * buffer in the queue -- even if this number
+				 * is larger than the number of characters
+				 * available in the first buffer, only the
+				 * characters from the first buffer are
+				 * returned. The execption is when there is
+				 * not any complete character in the first
+				 * buffer.  In that case, a recursive call
+				 * effectively obtains chars from the
+				 * second buffer. */
+    int *factorPtr)		/* On input, contains a guess of how many
+				 * bytes need to be allocated to hold the
+				 * result of converting N source bytes to
+				 * UTF-8. On output, contains another guess
+				 * based on the data seen so far. */
+{
+    Tcl_Encoding encoding = statePtr->encoding? statePtr->encoding
+	    : GetBinaryEncoding();
+    Tcl_EncodingState savedState = statePtr->inputEncodingState;
+    ChannelBuffer *bufPtr = statePtr->inQueueHead;
+    int savedIEFlags = statePtr->inputEncodingFlags;
+    int savedFlags = statePtr->flags;
+    char *dst, *src = RemovePoint(bufPtr);
+    int numBytes, srcLen = BytesLeft(bufPtr);
+
+    /*
+     * One src byte can yield at most one character.  So when the number of
+     * src bytes we plan to read is less than the limit on character count to
+     * be read, clearly we will remain within that limit, and we can use the
+     * value of "srcLen" as a tighter limit for sizing receiving buffers.
+     */
+
+    int toRead = ((charsToRead<0)||(charsToRead > srcLen)) ? srcLen : charsToRead;
+
+    /*
+     * 'factor' is how much we guess that the bytes in the source buffer will
+     * expand when converted to UTF-8 chars. This guess comes from analyzing
+     * how many characters were produced by the previous pass.
+     */
+
+    int factor = *factorPtr;
+    int dstLimit = TCL_UTF_MAX - 1 + toRead * factor / UTF_EXPANSION_FACTOR;
+
+    (void) TclGetStringFromObj(objPtr, &numBytes);
+    Tcl_AppendToObj(objPtr, NULL, dstLimit);
+    if (toRead == srcLen) {
+	unsigned int size;
+
+	dst = TclGetStringStorage(objPtr, &size) + numBytes;
+	dstLimit = size - numBytes;
+    } else {
+	dst = TclGetString(objPtr) + numBytes;
+    }
+
+    /*
+     * This routine is burdened with satisfying several constraints. It cannot
+     * append more than 'charsToRead` chars onto objPtr. This is measured
+     * after encoding and translation transformations are completed. There is
+     * no precise number of src bytes that can be associated with the limit.
+     * Yet, when we are done, we must know precisely the number of src bytes
+     * that were consumed to produce the appended chars, so that all
+     * subsequent bytes are left in the buffers for future read operations.
+     *
+     * The consequence is that we have no choice but to implement a "trial and
+     * error" approach, where in general we may need to perform
+     * transformations and copies multiple times to achieve a consistent set
+     * of results.  This takes the shape of a loop.
+     */
+
+    while (1) {
+	int dstDecoded, dstRead, dstWrote, srcRead, numChars, code;
+	int flags = statePtr->inputEncodingFlags | TCL_ENCODING_NO_TERMINATE;
+
+	if (charsToRead > 0) {
+	    flags |= TCL_ENCODING_CHAR_LIMIT;
+	    numChars = charsToRead;
+	}
+
+	/*
+	 * Perform the encoding transformation.  Read no more than srcLen
+	 * bytes, write no more than dstLimit bytes.
+	 *
+	 * Some trickiness with encoding flags here.  We do not want the end
+	 * of a buffer to be treated as the end of all input when the presence
+	 * of bytes in a next buffer are already known to exist.  This is
+	 * checked with an assert() because so far no test case causing the
+	 * assertion to be false has been created.  The normal operations of
+	 * channel reading appear to cause EOF and TCL_ENCODING_END setting to
+	 * appear only in situations where there are no further bytes in any
+	 * buffers.
+	 */
+
+	assert(bufPtr->nextPtr == NULL || BytesLeft(bufPtr->nextPtr) == 0
+		|| (statePtr->inputEncodingFlags & TCL_ENCODING_END) == 0);
+
+	code = Tcl_ExternalToUtf(NULL, encoding, src, srcLen,
+		flags, &statePtr->inputEncodingState,
+		dst, dstLimit, &srcRead, &dstDecoded, &numChars);
+
+	/*
+	 * Perform the translation transformation in place.  Read no more than
+	 * the dstDecoded bytes the encoding transformation actually produced.
+	 * Capture the number of bytes written in dstWrote. Capture the number
+	 * of bytes actually consumed in dstRead.
+	 */
+
+	dstWrote = dstLimit;
+	dstRead = dstDecoded;
+	TranslateInputEOL(statePtr, dst, dst, &dstWrote, &dstRead);
+
+	if (dstRead < dstDecoded) {
+	    /*
+	     * The encoding transformation produced bytes that the translation
+	     * transformation did not consume.  Why did this happen?
+	     */
+
+	    if (statePtr->inEofChar && dst[dstRead] == statePtr->inEofChar) {
+		/*
+		 * 1) There's an eof char set on the channel, and
+		 *    we saw it and stopped translating at that point.
+		 *
+		 * NOTE the bizarre spec of TranslateInputEOL in this case.
+		 * Clearly the eof char had to be read in order to account for
+		 * the stopping, but the value of dstRead does not include it.
+		 *
+		 * Also rather bizarre, our caller can only notice an EOF
+		 * condition if we return the value -1 as the number of chars
+		 * read.  This forces us to perform a 2-call dance where the
+		 * first call can read all the chars up to the eof char, and
+		 * the second call is solely for consuming the encoded eof
+		 * char then pointed at by src so that we can return that
+		 * magic -1 value.  This seems really wasteful, especially
+		 * since the first decoding pass of each call is likely to
+		 * decode many bytes beyond that eof char that's all we care
+		 * about.
+		 */
+
+		if (dstRead == 0) {
+		    /*
+		     * Curious choice in the eof char handling.  We leave the
+		     * eof char in the buffer. So, no need to compute a proper
+		     * srcRead value. At this point, there are no chars before
+		     * the eof char in the buffer.
+		     */
+
+		    Tcl_SetObjLength(objPtr, numBytes);
+		    return -1;
+		}
+
+		{
+		    /*
+		     * There are chars leading the buffer before the eof char.
+		     * Adjust the dstLimit so we go back and read only those
+		     * and do not encounter the eof char this time.
+		     */
+
+		    dstLimit = dstRead - 1 + TCL_UTF_MAX;
+		    statePtr->flags = savedFlags;
+		    statePtr->inputEncodingFlags = savedIEFlags;
+		    statePtr->inputEncodingState = savedState;
+		    continue;
+		}
+	    }
+
+	    /*
+	     * 2) The other way to read fewer bytes than are decoded is when
+	     *    the final byte is \r and we're in a CRLF translation mode so
+	     *    we cannot decide whether to record \r or \n yet.
+	     */
+
+	    assert(dst[dstRead] == '\r');
+	    assert(statePtr->inputTranslation == TCL_TRANSLATE_CRLF);
+
+	    if (dstWrote > 0) {
+		/*
+		 * There are chars we can read before we hit the bare CR.  Go
+		 * back with a smaller dstLimit so we get them in the next
+		 * pass, compute a matching srcRead, and don't end up back
+		 * here in this call.
+		 */
+
+		dstLimit = dstRead - 1 + TCL_UTF_MAX;
+		statePtr->flags = savedFlags;
+		statePtr->inputEncodingFlags = savedIEFlags;
+		statePtr->inputEncodingState = savedState;
+		continue;
+	    }
+
+	    assert(dstWrote == 0);
+	    assert(dstRead == 0);
+
+	    /*
+	     * We decoded only the bare CR, and we cannot read a translated
+	     * char from that alone. We have to know what's next.  So why do
+	     * we only have the one decoded char?
+	     */
+
+	    if (code != TCL_OK) {
+		char buffer[TCL_UTF_MAX + 1];
+		int read, decoded, count;
+
+		/*
+		 * Didn't get everything the buffer could offer
+		 */
+
+		statePtr->flags = savedFlags;
+		statePtr->inputEncodingFlags = savedIEFlags;
+		statePtr->inputEncodingState = savedState;
+
+		assert(bufPtr->nextPtr == NULL
+			|| BytesLeft(bufPtr->nextPtr) == 0 || 0 ==
+			(statePtr->inputEncodingFlags & TCL_ENCODING_END));
+
+		Tcl_ExternalToUtf(NULL, encoding, src, srcLen,
+		(statePtr->inputEncodingFlags | TCL_ENCODING_NO_TERMINATE),
+		&statePtr->inputEncodingState, buffer, TCL_UTF_MAX + 1,
+		&read, &decoded, &count);
+
+		if (count == 2) {
+		    if (buffer[1] == '\n') {
+			/* \r\n translate to \n */
+			dst[0] = '\n';
+			bufPtr->nextRemoved += read;
+		    } else {
+			dst[0] = '\r';
+			bufPtr->nextRemoved += srcRead;
+		    }
+
+		    statePtr->inputEncodingFlags &= ~TCL_ENCODING_START;
+
+		    Tcl_SetObjLength(objPtr, numBytes + 1);
+		    return 1;
+		}
+
+	    } else if (statePtr->flags & CHANNEL_EOF) {
+		/*
+		 * The bare \r is the only char and we will never read a
+		 * subsequent char to make the determination.
+		 */
+
+		dst[0] = '\r';
+		bufPtr->nextRemoved = bufPtr->nextAdded;
+		Tcl_SetObjLength(objPtr, numBytes + 1);
+		return 1;
+	    }
+
+	    /*
+	     * Revise the dstRead value so that the numChars calc below
+	     * correctly computes zero characters read.
+	     */
+
+	    dstRead = numChars;
+
+	    /* FALL THROUGH - get more data (dstWrote == 0) */
+	}
+
+	/*
+	 * The translation transformation can only reduce the number of chars
+	 * when it converts \r\n into \n. The reduction in the number of chars
+	 * is the difference in bytes read and written.
+	 */
+
+	numChars -= (dstRead - dstWrote);
+
+	if (charsToRead > 0 && numChars > charsToRead) {
+
+	    /*
+	     * TODO: This cannot happen anymore.
+	     *
+	     * We read more chars than allowed.  Reset limits to prevent that
+	     * and try again.  Don't forget the extra padding of TCL_UTF_MAX
+	     * bytes demanded by the Tcl_ExternalToUtf() call!
+	     */
+
+	    dstLimit = Tcl_UtfAtIndex(dst, charsToRead) - 1 + TCL_UTF_MAX - dst;
+	    statePtr->flags = savedFlags;
+	    statePtr->inputEncodingFlags = savedIEFlags;
+	    statePtr->inputEncodingState = savedState;
+	    continue;
+	}
+
+	if (dstWrote == 0) {
+	    ChannelBuffer *nextPtr;
+
+	    /*
+             * We were not able to read any chars.
+             */
+
+	    assert(numChars == 0);
+
+	    /*
+	     * There is one situation where this is the correct final result.
+	     * If the src buffer contains only a single \n byte, and we are in
+	     * TCL_TRANSLATE_AUTO mode, and when the translation pass was made
+	     * the INPUT_SAW_CR flag was set on the channel. In that case, the
+	     * correct behavior is to consume that \n and produce the empty
+	     * string.
+	     */
+
+	    if (dstRead == 1 && dst[0] == '\n') {
+		assert(statePtr->inputTranslation == TCL_TRANSLATE_AUTO);
+
+		goto consume;
+	    }
+
+	    /*
+             * Otherwise, reading zero characters indicates there's something
+	     * incomplete at the end of the src buffer.  Maybe there were not
+	     * enough src bytes to decode into a char.  Maybe a lone \r could
+	     * not be translated (crlf mode).  Need to combine any unused src
+	     * bytes we have in the first buffer with subsequent bytes to try
+	     * again.
+	     */
+
+	    nextPtr = bufPtr->nextPtr;
+
+	    if (nextPtr == NULL) {
+		if (srcLen > 0) {
+		    SetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+		}
+		Tcl_SetObjLength(objPtr, numBytes);
+		return -1;
+	    }
+
+	    /*
+	     * Space is made at the beginning of the buffer to copy the
+	     * previous unused bytes there. Check first if the buffer we are
+	     * using actually has enough space at its beginning for the data
+	     * we are copying.  Because if not we will write over the buffer
+	     * management information, especially the 'nextPtr'.
+	     *
+	     * Note that the BUFFER_PADDING (See AllocChannelBuffer) is used
+	     * to prevent exactly this situation. I.e. it should never happen.
+	     * Therefore it is ok to panic should it happen despite the
+	     * precautions.
+	     */
+
+	    if (nextPtr->nextRemoved - srcLen < 0) {
+		Tcl_Panic("Buffer Underflow, BUFFER_PADDING not enough");
+	    }
+
+	    nextPtr->nextRemoved -= srcLen;
+	    memcpy(RemovePoint(nextPtr), src, (size_t) srcLen);
+	    RecycleBuffer(statePtr, bufPtr, 0);
+	    statePtr->inQueueHead = nextPtr;
+	    Tcl_SetObjLength(objPtr, numBytes);
+	    return ReadChars(statePtr, objPtr, charsToRead, factorPtr);
+	}
+
+	statePtr->inputEncodingFlags &= ~TCL_ENCODING_START;
+
+    consume:
+	bufPtr->nextRemoved += srcRead;
+
+	/*
+	 * If this read contained multibyte characters, revise factorPtr so
+	 * the next read will allocate bigger buffers.
+	 */
+
+	if (numChars && numChars < srcRead) {
+	    *factorPtr = srcRead * UTF_EXPANSION_FACTOR / numChars;
+	}
+	Tcl_SetObjLength(objPtr, numBytes + dstWrote);
+	return numChars;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TranslateInputEOL --
+ *
+ *	Perform input EOL and EOF translation on the source buffer, leaving
+ *	the translated result in the destination buffer.
+ *
+ * Results:
+ *	The return value is 1 if the EOF character was found when copying
+ *	bytes to the destination buffer, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+TranslateInputEOL(
+    ChannelState *statePtr,	/* Channel being read, for EOL translation and
+				 * EOF character. */
+    char *dstStart,		/* Output buffer filled with chars by applying
+				 * appropriate EOL translation to source
+				 * characters. */
+    const char *srcStart,	/* Source characters. */
+    int *dstLenPtr,		/* On entry, the maximum length of output
+				 * buffer in bytes. On exit, the number of
+				 * bytes actually used in output buffer. */
+    int *srcLenPtr)		/* On entry, the length of source buffer. On
+				 * exit, the number of bytes read from the
+				 * source buffer. */
+{
+    const char *eof = NULL;
+    int dstLen = *dstLenPtr;
+    int srcLen = *srcLenPtr;
+    int inEofChar = statePtr->inEofChar;
+
+    /*
+     * Depending on the translation mode in use, there's no need to scan more
+     * srcLen bytes at srcStart than can possibly transform to dstLen bytes.
+     * This keeps the scan for eof char below from being pointlessly long.
+     */
+
+    switch (statePtr->inputTranslation) {
+    case TCL_TRANSLATE_LF:
+    case TCL_TRANSLATE_CR:
+	if (srcLen > dstLen) {
+            /*
+             * In these modes, each src byte become a dst byte.
+             */
+
+	    srcLen = dstLen;
+	}
+	break;
+    default:
+	/*
+         * In other modes, at most 2 src bytes become a dst byte.
+         */
+
+	if (srcLen/2 > dstLen) {
+	    srcLen = 2 * dstLen;
+	}
+	break;
+    }
+
+    if (inEofChar != '\0') {
+	/*
+	 * Make sure we do not read past any logical end of channel input
+	 * created by the presence of the input eof char.
+	 */
+
+	if ((eof = memchr(srcStart, inEofChar, srcLen))) {
+	    srcLen = eof - srcStart;
+	}
+    }
+
+    switch (statePtr->inputTranslation) {
+    case TCL_TRANSLATE_LF:
+    case TCL_TRANSLATE_CR:
+	if (dstStart != srcStart) {
+	    memcpy(dstStart, srcStart, (size_t) srcLen);
+	}
+	if (statePtr->inputTranslation == TCL_TRANSLATE_CR) {
+	    char *dst = dstStart;
+	    char *dstEnd = dstStart + srcLen;
+
+	    while ((dst = memchr(dst, '\r', dstEnd - dst))) {
+		*dst++ = '\n';
+	    }
+	}
+	dstLen = srcLen;
+	break;
+    case TCL_TRANSLATE_CRLF: {
+	const char *crFound, *src = srcStart;
+	char *dst = dstStart;
+	int lesser = (dstLen < srcLen) ? dstLen : srcLen;
+
+	while ((crFound = memchr(src, '\r', lesser))) {
+	    int numBytes = crFound - src;
+	    memmove(dst, src, numBytes);
+
+	    dst += numBytes; dstLen -= numBytes;
+	    src += numBytes; srcLen -= numBytes;
+	    if (srcLen == 1) {
+		/* valid src bytes end in \r */
+		if (eof) {
+		    *dst++ = '\r';
+		    src++; srcLen--;
+		} else {
+		    lesser = 0;
+		    break;
+		}
+	    } else if (src[1] == '\n') {
+		*dst++ = '\n';
+		src += 2; srcLen -= 2;
+	    } else {
+		*dst++ = '\r';
+		src++; srcLen--;
+	    }
+	    dstLen--;
+	    lesser = (dstLen < srcLen) ? dstLen : srcLen;
+	}
+	memmove(dst, src, lesser);
+	srcLen = src + lesser - srcStart;
+	dstLen = dst + lesser - dstStart;
+	break;
+    }
+    case TCL_TRANSLATE_AUTO: {
+	const char *crFound, *src = srcStart;
+	char *dst = dstStart;
+	int lesser;
+
+	if ((statePtr->flags & INPUT_SAW_CR) && srcLen) {
+	    if (*src == '\n') { src++; srcLen--; }
+	    ResetFlag(statePtr, INPUT_SAW_CR);
+	}
+	lesser = (dstLen < srcLen) ? dstLen : srcLen;
+	while ((crFound = memchr(src, '\r', lesser))) {
+	    int numBytes = crFound - src;
+	    memmove(dst, src, numBytes);
+
+	    dst[numBytes] = '\n';
+	    dst += numBytes + 1; dstLen -= numBytes + 1;
+	    src += numBytes + 1; srcLen -= numBytes + 1;
+	    if (srcLen == 0) {
+		SetFlag(statePtr, INPUT_SAW_CR);
+	    } else if (*src == '\n') {
+		src++; srcLen--;
+	    }
+	    lesser = (dstLen < srcLen) ? dstLen : srcLen;
+	}
+	memmove(dst, src, lesser);
+	srcLen = src + lesser - srcStart;
+	dstLen = dst + lesser - dstStart;
+	break;
+    }
+    default:
+	Tcl_Panic("unknown input translation %d", statePtr->inputTranslation);
+    }
+    *dstLenPtr = dstLen;
+    *srcLenPtr = srcLen;
+
+    if (srcStart + srcLen == eof) {
+	/*
+	 * EOF character was seen in EOL translated range. Leave current file
+	 * position pointing at the EOF character, but don't store the EOF
+	 * character in the output string.
+	 */
+
+	SetFlag(statePtr, CHANNEL_EOF | CHANNEL_STICKY_EOF);
+	statePtr->inputEncodingFlags |= TCL_ENCODING_END;
+	ResetFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Ungets --
+ *
+ *	Causes the supplied string to be added to the input queue of the
+ *	channel, at either the head or tail of the queue.
+ *
+ * Results:
+ *	The number of bytes stored in the channel, or -1 on error.
+ *
+ * Side effects:
+ *	Adds input to the input queue of a channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Ungets(
+    Tcl_Channel chan,		/* The channel for which to add the input. */
+    const char *str,		/* The input itself. */
+    int len,			/* The length of the input. */
+    int atEnd)			/* If non-zero, add at end of queue; otherwise
+				 * add at head of queue. */
+{
+    Channel *chanPtr;		/* The real IO channel. */
+    ChannelState *statePtr;	/* State of actual channel. */
+    ChannelBuffer *bufPtr;	/* Buffer to contain the data. */
+    int flags;
+
+    chanPtr = (Channel *) chan;
+    statePtr = chanPtr->state;
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    /*
+     * CheckChannelErrors clears too many flag bits in this one case.
+     */
+
+    flags = statePtr->flags;
+    if (CheckChannelErrors(statePtr, TCL_READABLE) != 0) {
+	len = -1;
+	goto done;
+    }
+    statePtr->flags = flags;
+
+    /*
+     * Clear the EOF flags, and clear the BLOCKED bit.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_EOF)) {
+	statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+    }
+    ResetFlag(statePtr,
+	    CHANNEL_BLOCKED | CHANNEL_STICKY_EOF | CHANNEL_EOF | INPUT_SAW_CR);
+    statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+
+    bufPtr = AllocChannelBuffer(len);
+    memcpy(InsertPoint(bufPtr), str, (size_t) len);
+    bufPtr->nextAdded += len;
+
+    if (statePtr->inQueueHead == NULL) {
+	bufPtr->nextPtr = NULL;
+	statePtr->inQueueHead = bufPtr;
+	statePtr->inQueueTail = bufPtr;
+    } else if (atEnd) {
+	bufPtr->nextPtr = NULL;
+	statePtr->inQueueTail->nextPtr = bufPtr;
+	statePtr->inQueueTail = bufPtr;
+    } else {
+	bufPtr->nextPtr = statePtr->inQueueHead;
+	statePtr->inQueueHead = bufPtr;
+    }
+
+    /*
+     * Update the notifier state so we don't block while there is still data
+     * in the buffers.
+     */
+
+  done:
+    UpdateInterest(chanPtr);
+    return len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Flush --
+ *
+ *	Flushes output data on a channel.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May flush output queued on this channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Flush(
+    Tcl_Channel chan)		/* The Channel to flush. */
+{
+    int result;			/* Of calling FlushChannel. */
+    Channel *chanPtr = (Channel *) chan;
+				/* The actual channel. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State of actual channel. */
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE) != 0) {
+	return -1;
+    }
+
+    result = FlushChannel(NULL, chanPtr, 0);
+    if (result != 0) {
+	return TCL_ERROR;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DiscardInputQueued --
+ *
+ *	Discards any input read from the channel but not yet consumed by Tcl
+ *	reading commands.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May discard input from the channel. If discardLastBuffer is zero,
+ *	leaves one buffer in place for back-filling.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DiscardInputQueued(
+    ChannelState *statePtr,	/* Channel on which to discard the queued
+				 * input. */
+    int discardSavedBuffers)	/* If non-zero, discard all buffers including
+				 * last one. */
+{
+    ChannelBuffer *bufPtr, *nxtPtr;
+				/* Loop variables. */
+
+    bufPtr = statePtr->inQueueHead;
+    statePtr->inQueueHead = NULL;
+    statePtr->inQueueTail = NULL;
+    for (; bufPtr != NULL; bufPtr = nxtPtr) {
+	nxtPtr = bufPtr->nextPtr;
+	RecycleBuffer(statePtr, bufPtr, discardSavedBuffers);
+    }
+
+    /*
+     * If discardSavedBuffers is nonzero, must also discard any previously
+     * saved buffer in the saveInBufPtr field.
+     */
+
+    if (discardSavedBuffers && statePtr->saveInBufPtr != NULL) {
+	ReleaseChannelBuffer(statePtr->saveInBufPtr);
+	statePtr->saveInBufPtr = NULL;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * GetInput --
+ *
+ *	Reads input data from a device into a channel buffer.
+ *
+ *	IMPORTANT!  This routine is only called on a chanPtr argument
+ *	that is the top channel of a stack!
+ *
+ * Results:
+ *	The return value is the Posix error code if an error occurred while
+ *	reading from the file, or 0 otherwise.
+ *
+ * Side effects:
+ *	Reads from the underlying device.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+GetInput(
+    Channel *chanPtr)		/* Channel to read input from. */
+{
+    int toRead;			/* How much to read? */
+    int result;			/* Of calling driver. */
+    int nread;			/* How much was read from channel? */
+    ChannelBuffer *bufPtr;	/* New buffer to add to input queue. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    /*
+     * Verify that all callers know better than to call us when
+     * it's recorded that the next char waiting to be read is the
+     * eofchar.
+     */
+
+    assert(!GotFlag(statePtr, CHANNEL_STICKY_EOF));
+
+    /*
+     * Prevent reading from a dead channel -- a channel that has been closed
+     * but not yet deallocated, which can happen if the exit handler for
+     * channel cleanup has run but the channel is still registered in some
+     * interpreter.
+     */
+
+    if (CheckForDeadChannel(NULL, statePtr)) {
+	return EINVAL;
+    }
+
+    /*
+     * WARNING: There was once a comment here claiming that it was a bad idea
+     * to make another call to the inputproc of a channel driver when EOF has
+     * already been detected on the channel.  Through much of Tcl's history,
+     * this warning was then completely negated by having all (most?) read
+     * paths clear the EOF setting before reaching here.  So we had a guard
+     * that was never triggered.
+     *
+     * Don't be tempted to restore the guard.  Even if EOF is set on the
+     * channel, continue through and call the inputproc again.  This is the
+     * way to enable the ability to [read] again beyond the EOF, which seems a
+     * strange thing to do, but for which use cases exist [Tcl Bug 5adc350683]
+     * and which may even be essential for channels representing things like
+     * ttys or other devices where the stream might take the logical form of a
+     * series of 'files' separated by an EOF condition.
+     *
+     * First check for more buffers in the pushback area of the topmost
+     * channel in the stack and use them. They can be the result of a
+     * transformation which went away without reading all the information
+     * placed in the area when it was stacked.
+     */
+
+    if (chanPtr->inQueueHead != NULL) {
+	/* TODO: Tests to cover this. */
+	assert(statePtr->inQueueHead == NULL);
+
+	statePtr->inQueueHead = chanPtr->inQueueHead;
+	statePtr->inQueueTail = chanPtr->inQueueTail;
+	chanPtr->inQueueHead = NULL;
+	chanPtr->inQueueTail = NULL;
+	return 0;
+    }
+
+    /*
+     * Nothing in the pushback area, fall back to the usual handling (driver,
+     * etc.)
+     */
+
+    /*
+     * See if we can fill an existing buffer. If we can, read only as much as
+     * will fit in it. Otherwise allocate a new buffer, add it to the input
+     * queue and attempt to fill it to the max.
+     */
+
+    bufPtr = statePtr->inQueueTail;
+
+    if ((bufPtr == NULL) || IsBufferFull(bufPtr)) {
+	bufPtr = statePtr->saveInBufPtr;
+	statePtr->saveInBufPtr = NULL;
+
+	/*
+	 * Check the actual buffersize against the requested buffersize.
+	 * Saved buffers of the wrong size are squashed. This is done to honor
+	 * dynamic changes of the buffersize made by the user.
+         *
+	 * TODO: Tests to cover this.
+	 */
+
+	if ((bufPtr != NULL)
+		&& (bufPtr->bufLength - BUFFER_PADDING != statePtr->bufSize)) {
+	    ReleaseChannelBuffer(bufPtr);
+	    bufPtr = NULL;
+	}
+
+	if (bufPtr == NULL) {
+	    bufPtr = AllocChannelBuffer(statePtr->bufSize);
+	}
+	bufPtr->nextPtr = NULL;
+
+	toRead = SpaceLeft(bufPtr);
+	assert(toRead == statePtr->bufSize);
+
+	if (statePtr->inQueueTail == NULL) {
+	    statePtr->inQueueHead = bufPtr;
+	} else {
+	    statePtr->inQueueTail->nextPtr = bufPtr;
+	}
+	statePtr->inQueueTail = bufPtr;
+    } else {
+	toRead = SpaceLeft(bufPtr);
+    }
+
+    PreserveChannelBuffer(bufPtr);
+    nread = ChanRead(chanPtr, InsertPoint(bufPtr), toRead);
+
+    if (nread < 0) {
+	result = Tcl_GetErrno();
+    } else {
+	result = 0;
+	bufPtr->nextAdded += nread;
+    }
+
+    ReleaseChannelBuffer(bufPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Seek --
+ *
+ *	Implements seeking on Tcl Channels. This is a public function so that
+ *	other C facilities may be implemented on top of it.
+ *
+ * Results:
+ *	The new access point or -1 on error. If error, use Tcl_GetErrno() to
+ *	retrieve the POSIX error code for the error that occurred.
+ *
+ * Side effects:
+ *	May flush output on the channel. May discard queued input.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_WideInt
+Tcl_Seek(
+    Tcl_Channel chan,		/* The channel on which to seek. */
+    Tcl_WideInt offset,		/* Offset to seek to. */
+    int mode)			/* Relative to which location to seek? */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The real IO channel. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int inputBuffered, outputBuffered;
+				/* # bytes held in buffers. */
+    int result;			/* Of device driver operations. */
+    Tcl_WideInt curPos;		/* Position on the device. */
+    int wasAsync;		/* Was the channel nonblocking before the seek
+				 * operation? If so, must restore to
+				 * non-blocking mode after the seek. */
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE | TCL_READABLE) != 0) {
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * Disallow seek on dead channels - channels that have been closed but not
+     * yet been deallocated. Such channels can be found if the exit handler
+     * for channel cleanup has run but the channel is still registered in an
+     * interpreter.
+     */
+
+    if (CheckForDeadChannel(NULL, statePtr)) {
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    /*
+     * Disallow seek on channels whose type does not have a seek procedure
+     * defined. This means that the channel does not support seeking.
+     */
+
+    if (chanPtr->typePtr->seekProc == NULL) {
+	Tcl_SetErrno(EINVAL);
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * Compute how much input and output is buffered. If both input and output
+     * is buffered, cannot compute the current position.
+     */
+
+    inputBuffered = Tcl_InputBuffered(chan);
+    outputBuffered = Tcl_OutputBuffered(chan);
+
+    if ((inputBuffered != 0) && (outputBuffered != 0)) {
+	Tcl_SetErrno(EFAULT);
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * If we are seeking relative to the current position, compute the
+     * corrected offset taking into account the amount of unread input.
+     */
+
+    if (mode == SEEK_CUR) {
+	offset -= inputBuffered;
+    }
+
+    /*
+     * Discard any queued input - this input should not be read after the
+     * seek.
+     */
+
+    DiscardInputQueued(statePtr, 0);
+
+    /*
+     * Reset EOF and BLOCKED flags. We invalidate them by moving the access
+     * point. Also clear CR related flags.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_EOF)) {
+	statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+    }
+    ResetFlag(statePtr, CHANNEL_EOF | CHANNEL_STICKY_EOF | CHANNEL_BLOCKED |
+	    INPUT_SAW_CR);
+    statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+
+    /*
+     * If the channel is in asynchronous output mode, switch it back to
+     * synchronous mode and cancel any async flush that may be scheduled.
+     * After the flush, the channel will be put back into asynchronous output
+     * mode.
+     */
+
+    wasAsync = 0;
+    if (GotFlag(statePtr, CHANNEL_NONBLOCKING)) {
+	wasAsync = 1;
+	result = StackSetBlockMode(chanPtr, TCL_MODE_BLOCKING);
+	if (result != 0) {
+	    return Tcl_LongAsWide(-1);
+	}
+	ResetFlag(statePtr, CHANNEL_NONBLOCKING);
+	if (GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	    ResetFlag(statePtr, BG_FLUSH_SCHEDULED);
+	}
+    }
+
+    /*
+     * If the flush fails we cannot recover the original position. In that
+     * case the seek is not attempted because we do not know where the access
+     * position is - instead we return the error. FlushChannel has already
+     * called Tcl_SetErrno() to report the error upwards. If the flush
+     * succeeds we do the seek also.
+     */
+
+    if (FlushChannel(NULL, chanPtr, 0) != 0) {
+	curPos = -1;
+    } else {
+	/*
+	 * Now seek to the new position in the channel as requested by the
+	 * caller.
+	 */
+
+	curPos = ChanSeek(chanPtr, offset, mode, &result);
+	if (curPos == Tcl_LongAsWide(-1)) {
+	    Tcl_SetErrno(result);
+	}
+    }
+
+    /*
+     * Restore to nonblocking mode if that was the previous behavior.
+     *
+     * NOTE: Even if there was an async flush active we do not restore it now
+     * because we already flushed all the queued output, above.
+     */
+
+    if (wasAsync) {
+	SetFlag(statePtr, CHANNEL_NONBLOCKING);
+	result = StackSetBlockMode(chanPtr, TCL_MODE_NONBLOCKING);
+	if (result != 0) {
+	    return Tcl_LongAsWide(-1);
+	}
+    }
+
+    return curPos;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Tell --
+ *
+ *	Returns the position of the next character to be read/written on this
+ *	channel.
+ *
+ * Results:
+ *	A nonnegative integer on success, -1 on failure. If failed, use
+ *	Tcl_GetErrno() to retrieve the POSIX error code for the error that
+ *	occurred.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_WideInt
+Tcl_Tell(
+    Tcl_Channel chan)		/* The channel to return pos for. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The real IO channel. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int inputBuffered, outputBuffered;
+				/* # bytes held in buffers. */
+    int result;			/* Of calling device driver. */
+    Tcl_WideInt curPos;		/* Position on device. */
+
+    if (CheckChannelErrors(statePtr, TCL_WRITABLE | TCL_READABLE) != 0) {
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * Disallow tell on dead channels -- channels that have been closed but
+     * not yet been deallocated. Such channels can be found if the exit
+     * handler for channel cleanup has run but the channel is still registered
+     * in an interpreter.
+     */
+
+    if (CheckForDeadChannel(NULL, statePtr)) {
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    /*
+     * Disallow tell on channels whose type does not have a seek procedure
+     * defined. This means that the channel does not support seeking.
+     */
+
+    if (chanPtr->typePtr->seekProc == NULL) {
+	Tcl_SetErrno(EINVAL);
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * Compute how much input and output is buffered. If both input and output
+     * is buffered, cannot compute the current position.
+     */
+
+    inputBuffered = Tcl_InputBuffered(chan);
+    outputBuffered = Tcl_OutputBuffered(chan);
+
+    /*
+     * Get the current position in the device and compute the position where
+     * the next character will be read or written. Note that we prefer the
+     * wideSeekProc if that is available and non-NULL...
+     */
+
+    curPos = ChanSeek(chanPtr, Tcl_LongAsWide(0), SEEK_CUR, &result);
+    if (curPos == Tcl_LongAsWide(-1)) {
+	Tcl_SetErrno(result);
+	return Tcl_LongAsWide(-1);
+    }
+
+    if (inputBuffered != 0) {
+	return curPos - inputBuffered;
+    }
+    return curPos + outputBuffered;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_TruncateChannel --
+ *
+ *	Truncate a channel to the given length.
+ *
+ * Results:
+ *	TCL_OK on success, TCL_ERROR if the operation failed (e.g., is not
+ *	supported by the type of channel, or the underlying OS operation
+ *	failed in some way).
+ *
+ * Side effects:
+ *	Seeks the channel to the current location. Sets errno on OS error.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_TruncateChannel(
+    Tcl_Channel chan,		/* Channel to truncate. */
+    Tcl_WideInt length)		/* Length to truncate it to. */
+{
+    Channel *chanPtr = (Channel *) chan;
+    Tcl_DriverTruncateProc *truncateProc =
+	    Tcl_ChannelTruncateProc(chanPtr->typePtr);
+    int result;
+
+    if (truncateProc == NULL) {
+	/*
+	 * Feature not supported and it's not emulatable. Pretend it's
+	 * returned an EINVAL, a very generic error!
+	 */
+
+	Tcl_SetErrno(EINVAL);
+	return TCL_ERROR;
+    }
+
+    if (!GotFlag(chanPtr->state, TCL_WRITABLE)) {
+	/*
+	 * We require that the file was opened of writing. Do that check now
+	 * so that we only flush if we think we're going to succeed.
+	 */
+
+	Tcl_SetErrno(EINVAL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Seek first to force a total flush of all pending buffers and ditch any
+     * pre-read input data.
+     */
+
+    WillWrite(chanPtr);
+
+    if (WillRead(chanPtr) < 0) {
+        return TCL_ERROR;
+    }
+
+    /*
+     * We're all flushed to disk now and we also don't have any unfortunate
+     * input baggage around either; can truncate with impunity.
+     */
+
+    result = truncateProc(chanPtr->instanceData, length);
+    if (result != 0) {
+	Tcl_SetErrno(result);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * CheckChannelErrors --
+ *
+ *	See if the channel is in an ready state and can perform the desired
+ *	operation.
+ *
+ * Results:
+ *	The return value is 0 if the channel is OK, otherwise the return value
+ *	is -1 and errno is set to indicate the error.
+ *
+ * Side effects:
+ *	May clear the EOF and/or BLOCKED bits if reading from channel.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+CheckChannelErrors(
+    ChannelState *statePtr,	/* Channel to check. */
+    int flags)			/* Test if channel supports desired operation:
+				 * TCL_READABLE, TCL_WRITABLE. Also indicates
+				 * Raw read or write for special close
+				 * processing */
+{
+    int direction = flags & (TCL_READABLE|TCL_WRITABLE);
+
+    /*
+     * Check for unreported error.
+     */
+
+    if (statePtr->unreportedError != 0) {
+	Tcl_SetErrno(statePtr->unreportedError);
+	statePtr->unreportedError = 0;
+
+	/*
+	 * TIP #219, Tcl Channel Reflection API.
+	 * Move a defered error message back into the channel bypass.
+	 */
+
+	if (statePtr->chanMsg != NULL) {
+	    TclDecrRefCount(statePtr->chanMsg);
+	}
+	statePtr->chanMsg = statePtr->unreportedMsg;
+	statePtr->unreportedMsg = NULL;
+	return -1;
+    }
+
+    /*
+     * Only the raw read and write operations are allowed during close in
+     * order to drain data from stacked channels.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_CLOSED) && !(flags & CHANNEL_RAW_MODE)) {
+	Tcl_SetErrno(EACCES);
+	return -1;
+    }
+
+    /*
+     * Fail if the channel is not opened for desired operation.
+     */
+
+    if ((statePtr->flags & direction) == 0) {
+	Tcl_SetErrno(EACCES);
+	return -1;
+    }
+
+    /*
+     * Fail if the channel is in the middle of a background copy.
+     *
+     * Don't do this tests for raw channels here or else the chaining in the
+     * transformation drivers will fail with 'file busy' error instead of
+     * retrieving and transforming the data to copy.
+     */
+
+    if (BUSY_STATE(statePtr, flags) && ((flags & CHANNEL_RAW_MODE) == 0)) {
+	Tcl_SetErrno(EBUSY);
+	return -1;
+    }
+
+    if (direction == TCL_READABLE) {
+	ResetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+    }
+
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Eof --
+ *
+ *	Returns 1 if the channel is at EOF, 0 otherwise.
+ *
+ * Results:
+ *	1 or 0, always.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Eof(
+    Tcl_Channel chan)		/* Does this channel have EOF? */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+
+    return GotFlag(statePtr, CHANNEL_EOF) ? 1 : 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InputBlocked --
+ *
+ *	Returns 1 if input is blocked on this channel, 0 otherwise.
+ *
+ * Results:
+ *	0 or 1, always.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_InputBlocked(
+    Tcl_Channel chan)		/* Is this channel blocked? */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+
+    return GotFlag(statePtr, CHANNEL_BLOCKED) ? 1 : 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InputBuffered --
+ *
+ *	Returns the number of bytes of input currently buffered in the common
+ *	internal buffer of a channel.
+ *
+ * Results:
+ *	The number of input bytes buffered, or zero if the channel is not open
+ *	for reading.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_InputBuffered(
+    Tcl_Channel chan)		/* The channel to query. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+    ChannelBuffer *bufPtr;
+    int bytesBuffered;
+
+    for (bytesBuffered = 0, bufPtr = statePtr->inQueueHead; bufPtr != NULL;
+	    bufPtr = bufPtr->nextPtr) {
+	bytesBuffered += BytesLeft(bufPtr);
+    }
+
+    /*
+     * Don't forget the bytes in the topmost pushback area.
+     */
+
+    for (bufPtr = statePtr->topChanPtr->inQueueHead; bufPtr != NULL;
+	    bufPtr = bufPtr->nextPtr) {
+	bytesBuffered += BytesLeft(bufPtr);
+    }
+
+    return bytesBuffered;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_OutputBuffered --
+ *
+ *    Returns the number of bytes of output currently buffered in the common
+ *    internal buffer of a channel.
+ *
+ * Results:
+ *    The number of output bytes buffered, or zero if the channel is not open
+ *    for writing.
+ *
+ * Side effects:
+ *    None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_OutputBuffered(
+    Tcl_Channel chan)		/* The channel to query. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+    ChannelBuffer *bufPtr;
+    int bytesBuffered;
+
+    for (bytesBuffered = 0, bufPtr = statePtr->outQueueHead; bufPtr != NULL;
+	    bufPtr = bufPtr->nextPtr) {
+	bytesBuffered += BytesLeft(bufPtr);
+    }
+    if (statePtr->curOutPtr != NULL) {
+	register ChannelBuffer *curOutPtr = statePtr->curOutPtr;
+
+	if (IsBufferReady(curOutPtr)) {
+	    bytesBuffered += BytesLeft(curOutPtr);
+	}
+    }
+
+    return bytesBuffered;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelBuffered --
+ *
+ *	Returns the number of bytes of input currently buffered in the
+ *	internal buffer (push back area) of a channel.
+ *
+ * Results:
+ *	The number of input bytes buffered, or zero if the channel is not open
+ *	for reading.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ChannelBuffered(
+    Tcl_Channel chan)		/* The channel to query. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* Real channel structure. */
+    ChannelBuffer *bufPtr;
+    int bytesBuffered = 0;
+
+    for (bufPtr = chanPtr->inQueueHead; bufPtr != NULL;
+	    bufPtr = bufPtr->nextPtr) {
+	bytesBuffered += BytesLeft(bufPtr);
+    }
+
+    return bytesBuffered;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetChannelBufferSize --
+ *
+ *	Sets the size of buffers to allocate to store input or output in the
+ *	channel. The size must be between 1 byte and 1 MByte.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Sets the size of buffers subsequently allocated for this channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetChannelBufferSize(
+    Tcl_Channel chan,		/* The channel whose buffer size to set. */
+    int sz)			/* The size to set. */
+{
+    ChannelState *statePtr;	/* State of real channel structure. */
+
+    /*
+     * Clip the buffer size to force it into the [1,1M] range
+     */
+
+    if (sz < 1) {
+	sz = 1;
+    } else if (sz > MAX_CHANNEL_BUFFER_SIZE) {
+	sz = MAX_CHANNEL_BUFFER_SIZE;
+    }
+
+    statePtr = ((Channel *) chan)->state;
+
+    if (statePtr->bufSize == sz) {
+	return;
+    }
+    statePtr->bufSize = sz;
+
+    /*
+     * If bufsize changes, need to get rid of old utility buffer.
+     */
+
+    if (statePtr->saveInBufPtr != NULL) {
+	RecycleBuffer(statePtr, statePtr->saveInBufPtr, 1);
+	statePtr->saveInBufPtr = NULL;
+    }
+    if ((statePtr->inQueueHead != NULL)
+	    && (statePtr->inQueueHead->nextPtr == NULL)
+	    && IsBufferEmpty(statePtr->inQueueHead)) {
+	RecycleBuffer(statePtr, statePtr->inQueueHead, 1);
+	statePtr->inQueueHead = NULL;
+	statePtr->inQueueTail = NULL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelBufferSize --
+ *
+ *	Retrieves the size of buffers to allocate for this channel.
+ *
+ * Results:
+ *	The size.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelBufferSize(
+    Tcl_Channel chan)		/* The channel for which to find the buffer
+				 * size. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+
+    return statePtr->bufSize;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_BadChannelOption --
+ *
+ *	This procedure generates a "bad option" error message in an (optional)
+ *	interpreter. It is used by channel drivers when a invalid Set/Get
+ *	option is requested. Its purpose is to concatenate the generic options
+ *	list to the specific ones and factorize the generic options error
+ *	message string.
+ *
+ * Results:
+ *	TCL_ERROR.
+ *
+ * Side effects:
+
+ *	An error message is generated in interp's result object to indicate
+ *	that a command was invoked with the a bad option. The message has the
+ *	form:
+ *		bad option "blah": should be one of
+ *		<...generic options...>+<...specific options...>
+ *	"blah" is the optionName argument and "<specific options>" is a space
+ *	separated list of specific option words. The function takes good care
+ *	of inserting minus signs before each option, commas after, and an "or"
+ *	before the last option.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_BadChannelOption(
+    Tcl_Interp *interp,		/* Current interpreter (can be NULL).*/
+    const char *optionName,	/* 'bad option' name */
+    const char *optionList)	/* Specific options list to append to the
+				 * standard generic options. Can be NULL for
+				 * generic options only. */
+{
+    if (interp != NULL) {
+	const char *genericopt =
+		"blocking buffering buffersize encoding eofchar translation";
+	const char **argv;
+	int argc, i;
+	Tcl_DString ds;
+        Tcl_Obj *errObj;
+
+	Tcl_DStringInit(&ds);
+	Tcl_DStringAppend(&ds, genericopt, -1);
+	if (optionList && (*optionList)) {
+	    TclDStringAppendLiteral(&ds, " ");
+	    Tcl_DStringAppend(&ds, optionList, -1);
+	}
+	if (Tcl_SplitList(interp, Tcl_DStringValue(&ds),
+		&argc, &argv) != TCL_OK) {
+	    Tcl_Panic("malformed option list in channel driver");
+	}
+	Tcl_ResetResult(interp);
+	errObj = Tcl_ObjPrintf("bad option \"%s\": should be one of ",
+                optionName);
+	argc--;
+	for (i = 0; i < argc; i++) {
+	    Tcl_AppendPrintfToObj(errObj, "-%s, ", argv[i]);
+	}
+	Tcl_AppendPrintfToObj(errObj, "or -%s", argv[i]);
+        Tcl_SetObjResult(interp, errObj);
+	Tcl_DStringFree(&ds);
+	ckfree(argv);
+    }
+    Tcl_SetErrno(EINVAL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelOption --
+ *
+ *	Gets a mode associated with an IO channel. If the optionName arg is
+ *	non NULL, retrieves the value of that option. If the optionName arg is
+ *	NULL, retrieves a list of alternating option names and values for the
+ *	given channel.
+ *
+ * Results:
+ *	A standard Tcl result. Also sets the supplied DString to the string
+ *	value of the option(s) returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelOption(
+    Tcl_Interp *interp,		/* For error reporting - can be NULL. */
+    Tcl_Channel chan,		/* Channel on which to get option. */
+    const char *optionName,	/* Option to get. */
+    Tcl_DString *dsPtr)		/* Where to store value(s). */
+{
+    size_t len;			/* Length of optionName string. */
+    char optionVal[128];	/* Buffer for sprintf. */
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int flags;
+
+    /*
+     * Disallow options on dead channels -- channels that have been closed but
+     * not yet been deallocated. Such channels can be found if the exit
+     * handler for channel cleanup has run but the channel is still registered
+     * in an interpreter.
+     */
+
+    if (CheckForDeadChannel(interp, statePtr)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    /*
+     * If we are in the middle of a background copy, use the saved flags.
+     */
+
+    if (statePtr->csPtrR) {
+	flags = statePtr->csPtrR->readFlags;
+    } else if (statePtr->csPtrW) {
+	flags = statePtr->csPtrW->writeFlags;
+    } else {
+	flags = statePtr->flags;
+    }
+
+    /*
+     * If the optionName is NULL it means that we want a list of all options
+     * and values.
+     */
+
+    if (optionName == NULL) {
+	len = 0;
+    } else {
+	len = strlen(optionName);
+    }
+
+    if (len == 0 || HaveOpt(2, "-blocking")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-blocking");
+	}
+	Tcl_DStringAppendElement(dsPtr,
+		(flags & CHANNEL_NONBLOCKING) ? "0" : "1");
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+    if (len == 0 || HaveOpt(7, "-buffering")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-buffering");
+	}
+	if (flags & CHANNEL_LINEBUFFERED) {
+	    Tcl_DStringAppendElement(dsPtr, "line");
+	} else if (flags & CHANNEL_UNBUFFERED) {
+	    Tcl_DStringAppendElement(dsPtr, "none");
+	} else {
+	    Tcl_DStringAppendElement(dsPtr, "full");
+	}
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+    if (len == 0 || HaveOpt(7, "-buffersize")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-buffersize");
+	}
+	TclFormatInt(optionVal, statePtr->bufSize);
+	Tcl_DStringAppendElement(dsPtr, optionVal);
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+    if (len == 0 || HaveOpt(2, "-encoding")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-encoding");
+	}
+	if (statePtr->encoding == NULL) {
+	    Tcl_DStringAppendElement(dsPtr, "binary");
+	} else {
+	    Tcl_DStringAppendElement(dsPtr,
+		    Tcl_GetEncodingName(statePtr->encoding));
+	}
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+    if (len == 0 || HaveOpt(2, "-eofchar")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-eofchar");
+	}
+	if (((flags & (TCL_READABLE|TCL_WRITABLE)) ==
+		(TCL_READABLE|TCL_WRITABLE)) && (len == 0)) {
+	    Tcl_DStringStartSublist(dsPtr);
+	}
+	if (flags & TCL_READABLE) {
+	    if (statePtr->inEofChar == 0) {
+		Tcl_DStringAppendElement(dsPtr, "");
+	    } else {
+		char buf[4];
+
+		sprintf(buf, "%c", statePtr->inEofChar);
+		Tcl_DStringAppendElement(dsPtr, buf);
+	    }
+	}
+	if (flags & TCL_WRITABLE) {
+	    if (statePtr->outEofChar == 0) {
+		Tcl_DStringAppendElement(dsPtr, "");
+	    } else {
+		char buf[4];
+
+		sprintf(buf, "%c", statePtr->outEofChar);
+		Tcl_DStringAppendElement(dsPtr, buf);
+	    }
+	}
+	if (!(flags & (TCL_READABLE|TCL_WRITABLE))) {
+	    /*
+	     * Not readable or writable (e.g. server socket)
+	     */
+
+	    Tcl_DStringAppendElement(dsPtr, "");
+	}
+	if (((flags & (TCL_READABLE|TCL_WRITABLE)) ==
+		(TCL_READABLE|TCL_WRITABLE)) && (len == 0)) {
+	    Tcl_DStringEndSublist(dsPtr);
+	}
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+    if (len == 0 || HaveOpt(1, "-translation")) {
+	if (len == 0) {
+	    Tcl_DStringAppendElement(dsPtr, "-translation");
+	}
+	if (((flags & (TCL_READABLE|TCL_WRITABLE)) ==
+		(TCL_READABLE|TCL_WRITABLE)) && (len == 0)) {
+	    Tcl_DStringStartSublist(dsPtr);
+	}
+	if (flags & TCL_READABLE) {
+	    if (statePtr->inputTranslation == TCL_TRANSLATE_AUTO) {
+		Tcl_DStringAppendElement(dsPtr, "auto");
+	    } else if (statePtr->inputTranslation == TCL_TRANSLATE_CR) {
+		Tcl_DStringAppendElement(dsPtr, "cr");
+	    } else if (statePtr->inputTranslation == TCL_TRANSLATE_CRLF) {
+		Tcl_DStringAppendElement(dsPtr, "crlf");
+	    } else {
+		Tcl_DStringAppendElement(dsPtr, "lf");
+	    }
+	}
+	if (flags & TCL_WRITABLE) {
+	    if (statePtr->outputTranslation == TCL_TRANSLATE_AUTO) {
+		Tcl_DStringAppendElement(dsPtr, "auto");
+	    } else if (statePtr->outputTranslation == TCL_TRANSLATE_CR) {
+		Tcl_DStringAppendElement(dsPtr, "cr");
+	    } else if (statePtr->outputTranslation == TCL_TRANSLATE_CRLF) {
+		Tcl_DStringAppendElement(dsPtr, "crlf");
+	    } else {
+		Tcl_DStringAppendElement(dsPtr, "lf");
+	    }
+	}
+	if (!(flags & (TCL_READABLE|TCL_WRITABLE))) {
+	    /*
+	     * Not readable or writable (e.g. server socket)
+	     */
+
+	    Tcl_DStringAppendElement(dsPtr, "auto");
+	}
+	if (((flags & (TCL_READABLE|TCL_WRITABLE)) ==
+		(TCL_READABLE|TCL_WRITABLE)) && (len == 0)) {
+	    Tcl_DStringEndSublist(dsPtr);
+	}
+	if (len > 0) {
+	    return TCL_OK;
+	}
+    }
+
+    if (chanPtr->typePtr->getOptionProc != NULL) {
+	/*
+	 * Let the driver specific handle additional options and result code
+	 * and message.
+	 */
+
+	return chanPtr->typePtr->getOptionProc(chanPtr->instanceData, interp,
+		optionName, dsPtr);
+    } else {
+	/*
+	 * No driver specific options case.
+	 */
+
+	if (len == 0) {
+	    return TCL_OK;
+	}
+	return Tcl_BadChannelOption(interp, optionName, NULL);
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_SetChannelOption --
+ *
+ *	Sets an option on a channel.
+ *
+ * Results:
+ *	A standard Tcl result. On error, sets interp's result object if
+ *	interp is not NULL.
+ *
+ * Side effects:
+ *	May modify an option on a device.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_SetChannelOption(
+    Tcl_Interp *interp,		/* For error reporting - can be NULL. */
+    Tcl_Channel chan,		/* Channel on which to set mode. */
+    const char *optionName,	/* Which option to set? */
+    const char *newValue)	/* New value for option. */
+{
+    Channel *chanPtr = (Channel *) chan;
+				/* The real IO channel. */
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    size_t len;			/* Length of optionName string. */
+    int argc;
+    const char **argv;
+
+    /*
+     * If the channel is in the middle of a background copy, fail.
+     */
+
+    if (statePtr->csPtrR || statePtr->csPtrW) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "unable to set channel options: background copy in"
+                    " progress", -1));
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Disallow options on dead channels -- channels that have been closed but
+     * not yet been deallocated. Such channels can be found if the exit
+     * handler for channel cleanup has run but the channel is still registered
+     * in an interpreter.
+     */
+
+    if (CheckForDeadChannel(NULL, statePtr)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * This operation should occur at the top of a channel stack.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+
+    len = strlen(optionName);
+
+    if (HaveOpt(2, "-blocking")) {
+	int newMode;
+
+	if (Tcl_GetBoolean(interp, newValue, &newMode) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (newMode) {
+	    newMode = TCL_MODE_BLOCKING;
+	} else {
+	    newMode = TCL_MODE_NONBLOCKING;
+	}
+	return SetBlockMode(interp, chanPtr, newMode);
+    } else if (HaveOpt(7, "-buffering")) {
+	len = strlen(newValue);
+	if ((newValue[0] == 'f') && (strncmp(newValue, "full", len) == 0)) {
+	    ResetFlag(statePtr, CHANNEL_UNBUFFERED | CHANNEL_LINEBUFFERED);
+	} else if ((newValue[0] == 'l') &&
+		(strncmp(newValue, "line", len) == 0)) {
+	    ResetFlag(statePtr, CHANNEL_UNBUFFERED);
+	    SetFlag(statePtr, CHANNEL_LINEBUFFERED);
+	} else if ((newValue[0] == 'n') &&
+		(strncmp(newValue, "none", len) == 0)) {
+	    ResetFlag(statePtr, CHANNEL_LINEBUFFERED);
+	    SetFlag(statePtr, CHANNEL_UNBUFFERED);
+	} else if (interp) {
+            Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                    "bad value for -buffering: must be one of"
+                    " full, line, or none", -1));
+            return TCL_ERROR;
+	}
+	return TCL_OK;
+    } else if (HaveOpt(7, "-buffersize")) {
+	int newBufferSize;
+
+	if (Tcl_GetInt(interp, newValue, &newBufferSize) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetChannelBufferSize(chan, newBufferSize);
+	return TCL_OK;
+    } else if (HaveOpt(2, "-encoding")) {
+	Tcl_Encoding encoding;
+
+	if ((newValue[0] == '\0') || (strcmp(newValue, "binary") == 0)) {
+	    encoding = NULL;
+	} else {
+	    encoding = Tcl_GetEncoding(interp, newValue);
+	    if (encoding == NULL) {
+		return TCL_ERROR;
+	    }
+	}
+
+	/*
+	 * When the channel has an escape sequence driven encoding such as
+	 * iso2022, the terminated escape sequence must write to the buffer.
+	 */
+
+	if ((statePtr->encoding != NULL)
+		&& !(statePtr->outputEncodingFlags & TCL_ENCODING_START)
+		&& (CheckChannelErrors(statePtr, TCL_WRITABLE) == 0)) {
+	    statePtr->outputEncodingFlags |= TCL_ENCODING_END;
+	    WriteChars(chanPtr, "", 0);
+	}
+	Tcl_FreeEncoding(statePtr->encoding);
+	statePtr->encoding = encoding;
+	statePtr->inputEncodingState = NULL;
+	statePtr->inputEncodingFlags = TCL_ENCODING_START;
+	statePtr->outputEncodingState = NULL;
+	statePtr->outputEncodingFlags = TCL_ENCODING_START;
+	ResetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+	UpdateInterest(chanPtr);
+	return TCL_OK;
+    } else if (HaveOpt(2, "-eofchar")) {
+	if (Tcl_SplitList(interp, newValue, &argc, &argv) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (argc == 0) {
+	    statePtr->inEofChar = 0;
+	    statePtr->outEofChar = 0;
+	} else if (argc == 1 || argc == 2) {
+	    int outIndex = (argc - 1);
+	    int inValue = (int) argv[0][0];
+	    int outValue = (int) argv[outIndex][0];
+
+	    if (inValue & 0x80 || outValue & 0x80) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+                            "bad value for -eofchar: must be non-NUL ASCII"
+                            " character", -1));
+		}
+		ckfree(argv);
+		return TCL_ERROR;
+	    }
+	    if (GotFlag(statePtr, TCL_READABLE)) {
+		statePtr->inEofChar = inValue;
+	    }
+	    if (GotFlag(statePtr, TCL_WRITABLE)) {
+		statePtr->outEofChar = outValue;
+	    }
+	} else {
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"bad value for -eofchar: should be a list of zero,"
+			" one, or two elements", -1));
+	    }
+	    ckfree(argv);
+	    return TCL_ERROR;
+	}
+	if (argv != NULL) {
+	    ckfree(argv);
+	}
+
+	/*
+	 * [Bug 930851] Reset EOF and BLOCKED flags. Changing the character
+	 * which signals eof can transform a current eof condition into a 'go
+	 * ahead'. Ditto for blocked.
+	 */
+
+	if (GotFlag(statePtr, CHANNEL_EOF)) {
+	    statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+	}
+	ResetFlag(statePtr, CHANNEL_EOF|CHANNEL_STICKY_EOF|CHANNEL_BLOCKED);
+	statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+	return TCL_OK;
+    } else if (HaveOpt(1, "-translation")) {
+	const char *readMode, *writeMode;
+
+	if (Tcl_SplitList(interp, newValue, &argc, &argv) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+
+	if (argc == 1) {
+	    readMode = GotFlag(statePtr, TCL_READABLE) ? argv[0] : NULL;
+	    writeMode = GotFlag(statePtr, TCL_WRITABLE) ? argv[0] : NULL;
+	} else if (argc == 2) {
+	    readMode = GotFlag(statePtr, TCL_READABLE) ? argv[0] : NULL;
+	    writeMode = GotFlag(statePtr, TCL_WRITABLE) ? argv[1] : NULL;
+	} else {
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"bad value for -translation: must be a one or two"
+			" element list", -1));
+	    }
+	    ckfree(argv);
+	    return TCL_ERROR;
+	}
+
+	if (readMode) {
+	    TclEolTranslation translation;
+
+	    if (*readMode == '\0') {
+		translation = statePtr->inputTranslation;
+	    } else if (strcmp(readMode, "auto") == 0) {
+		translation = TCL_TRANSLATE_AUTO;
+	    } else if (strcmp(readMode, "binary") == 0) {
+		translation = TCL_TRANSLATE_LF;
+		statePtr->inEofChar = 0;
+		Tcl_FreeEncoding(statePtr->encoding);
+		statePtr->encoding = NULL;
+	    } else if (strcmp(readMode, "lf") == 0) {
+		translation = TCL_TRANSLATE_LF;
+	    } else if (strcmp(readMode, "cr") == 0) {
+		translation = TCL_TRANSLATE_CR;
+	    } else if (strcmp(readMode, "crlf") == 0) {
+		translation = TCL_TRANSLATE_CRLF;
+	    } else if (strcmp(readMode, "platform") == 0) {
+		translation = TCL_PLATFORM_TRANSLATION;
+	    } else {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "bad value for -translation: must be one of "
+                            "auto, binary, cr, lf, crlf, or platform", -1));
+		}
+		ckfree(argv);
+		return TCL_ERROR;
+	    }
+
+	    /*
+	     * Reset the EOL flags since we need to look at any buffered data
+	     * to see if the new translation mode allows us to complete the
+	     * line.
+	     */
+
+	    if (translation != statePtr->inputTranslation) {
+		statePtr->inputTranslation = translation;
+		ResetFlag(statePtr, INPUT_SAW_CR | CHANNEL_NEED_MORE_DATA);
+		UpdateInterest(chanPtr);
+	    }
+	}
+	if (writeMode) {
+	    if (*writeMode == '\0') {
+		/* Do nothing. */
+	    } else if (strcmp(writeMode, "auto") == 0) {
+		/*
+		 * This is a hack to get TCP sockets to produce output in CRLF
+		 * mode if they are being set into AUTO mode. A better
+		 * solution for achieving this effect will be coded later.
+		 */
+
+		if (strcmp(Tcl_ChannelName(chanPtr->typePtr), "tcp") == 0) {
+		    statePtr->outputTranslation = TCL_TRANSLATE_CRLF;
+		} else {
+		    statePtr->outputTranslation = TCL_PLATFORM_TRANSLATION;
+		}
+	    } else if (strcmp(writeMode, "binary") == 0) {
+		statePtr->outEofChar = 0;
+		statePtr->outputTranslation = TCL_TRANSLATE_LF;
+		Tcl_FreeEncoding(statePtr->encoding);
+		statePtr->encoding = NULL;
+	    } else if (strcmp(writeMode, "lf") == 0) {
+		statePtr->outputTranslation = TCL_TRANSLATE_LF;
+	    } else if (strcmp(writeMode, "cr") == 0) {
+		statePtr->outputTranslation = TCL_TRANSLATE_CR;
+	    } else if (strcmp(writeMode, "crlf") == 0) {
+		statePtr->outputTranslation = TCL_TRANSLATE_CRLF;
+	    } else if (strcmp(writeMode, "platform") == 0) {
+		statePtr->outputTranslation = TCL_PLATFORM_TRANSLATION;
+	    } else {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "bad value for -translation: must be one of "
+                            "auto, binary, cr, lf, crlf, or platform", -1));
+		}
+		ckfree(argv);
+		return TCL_ERROR;
+	    }
+	}
+	ckfree(argv);
+	return TCL_OK;
+    } else if (chanPtr->typePtr->setOptionProc != NULL) {
+	return chanPtr->typePtr->setOptionProc(chanPtr->instanceData, interp,
+		optionName, newValue);
+    } else {
+	return Tcl_BadChannelOption(interp, optionName, NULL);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CleanupChannelHandlers --
+ *
+ *	Removes channel handlers that refer to the supplied interpreter, so
+ *	that if the actual channel is not closed now, these handlers will not
+ *	run on subsequent events on the channel. This would be erroneous,
+ *	because the interpreter no longer has a reference to this channel.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes channel handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CleanupChannelHandlers(
+    Tcl_Interp *interp,
+    Channel *chanPtr)
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    EventScriptRecord *sPtr, *prevPtr, *nextPtr;
+
+    /*
+     * Remove fileevent records on this channel that refer to the given
+     * interpreter.
+     */
+
+    for (sPtr = statePtr->scriptRecordPtr, prevPtr = NULL;
+	    sPtr != NULL; sPtr = nextPtr) {
+	nextPtr = sPtr->nextPtr;
+	if (sPtr->interp == interp) {
+	    if (prevPtr == NULL) {
+		statePtr->scriptRecordPtr = nextPtr;
+	    } else {
+		prevPtr->nextPtr = nextPtr;
+	    }
+
+	    Tcl_DeleteChannelHandler((Tcl_Channel) chanPtr,
+		    TclChannelEventScriptInvoker, sPtr);
+
+	    TclDecrRefCount(sPtr->scriptPtr);
+	    ckfree(sPtr);
+	} else {
+	    prevPtr = sPtr;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NotifyChannel --
+ *
+ *	This procedure is called by a channel driver when a driver detects an
+ *	event on a channel. This procedure is responsible for actually
+ *	handling the event by invoking any channel handler callbacks.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Whatever the channel handler callback procedure does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_NotifyChannel(
+    Tcl_Channel channel,	/* Channel that detected an event. */
+    int mask)			/* OR'ed combination of TCL_READABLE,
+				 * TCL_WRITABLE, or TCL_EXCEPTION: indicates
+				 * which events were detected. */
+{
+    Channel *chanPtr = (Channel *) channel;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    ChannelHandler *chPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    NextChannelHandler nh;
+    Channel *upChanPtr;
+    const Tcl_ChannelType *upTypePtr;
+
+    /*
+     * In contrast to the other API functions this procedure walks towards the
+     * top of a stack and not down from it.
+     *
+     * The channel calling this procedure is the one who generated the event,
+     * and thus does not take part in handling it. IOW, its HandlerProc is not
+     * called, instead we begin with the channel above it.
+     *
+     * This behaviour also allows the transformation channels to generate
+     * their own events and pass them upward.
+     */
+
+    while (mask && (chanPtr->upChanPtr != NULL)) {
+	Tcl_DriverHandlerProc *upHandlerProc;
+
+	upChanPtr = chanPtr->upChanPtr;
+	upTypePtr = upChanPtr->typePtr;
+	upHandlerProc = Tcl_ChannelHandlerProc(upTypePtr);
+	if (upHandlerProc != NULL) {
+	    mask = upHandlerProc(upChanPtr->instanceData, mask);
+	}
+
+	/*
+	 * ELSE: Ignore transformations which are unable to handle the event
+	 * coming from below. Assume that they don't change the mask and pass
+	 * it on.
+	 */
+
+	chanPtr = upChanPtr;
+    }
+
+    channel = (Tcl_Channel) chanPtr;
+
+    /*
+     * Here we have either reached the top of the stack or the mask is empty.
+     * We break out of the procedure if it is the latter.
+     */
+
+    if (!mask) {
+	return;
+    }
+
+    /*
+     * We are now above the topmost channel in a stack and have events left.
+     * Now call the channel handlers as usual.
+     *
+     * Preserve the channel struct in case the script closes it.
+     */
+
+    TclChannelPreserve((Tcl_Channel)channel);
+    Tcl_Preserve(statePtr);
+
+    /*
+     * If we are flushing in the background, be sure to call FlushChannel for
+     * writable events. Note that we have to discard the writable event so we
+     * don't call any write handlers before the flush is complete.
+     */
+
+    if (GotFlag(statePtr, BG_FLUSH_SCHEDULED) && (mask & TCL_WRITABLE)) {
+	if (0 == FlushChannel(NULL, chanPtr, 1)) {
+	    mask &= ~TCL_WRITABLE;
+	}
+    }
+
+    /*
+     * Add this invocation to the list of recursive invocations of
+     * Tcl_NotifyChannel.
+     */
+
+    nh.nextHandlerPtr = NULL;
+    nh.nestedHandlerPtr = tsdPtr->nestedHandlerPtr;
+    tsdPtr->nestedHandlerPtr = &nh;
+
+    for (chPtr = statePtr->chPtr; chPtr != NULL; ) {
+	/*
+	 * If this channel handler is interested in any of the events that
+	 * have occurred on the channel, invoke its procedure.
+	 */
+
+	if ((chPtr->mask & mask) != 0) {
+	    nh.nextHandlerPtr = chPtr->nextPtr;
+	    chPtr->proc(chPtr->clientData, chPtr->mask & mask);
+	    chPtr = nh.nextHandlerPtr;
+	} else {
+	    chPtr = chPtr->nextPtr;
+	}
+    }
+
+    /*
+     * Update the notifier interest, since it may have changed after invoking
+     * event handlers. Skip that if the channel was deleted in the call to the
+     * channel handler.
+     */
+
+    if (chanPtr->typePtr != NULL) {
+	/*
+	 * TODO: This call may not be needed.  If a handler induced a
+	 * change in interest, that handler should have made its own
+	 * UpdateInterest() call, one would think.
+	 */
+	UpdateInterest(chanPtr);
+    }
+
+    Tcl_Release(statePtr);
+    TclChannelRelease(channel);
+
+    tsdPtr->nestedHandlerPtr = nh.nestedHandlerPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateInterest --
+ *
+ *	Arrange for the notifier to call us back at appropriate times based on
+ *	the current state of the channel.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May schedule a timer or driver handler.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateInterest(
+    Channel *chanPtr)		/* Channel to update. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    int mask = statePtr->interestMask;
+
+    if (chanPtr->typePtr == NULL) {
+	/* Do not update interest on a closed channel */
+	return;
+    }
+
+    /*
+     * If there are flushed buffers waiting to be written, then we need to
+     * watch for the channel to become writable.
+     */
+
+    if (GotFlag(statePtr, BG_FLUSH_SCHEDULED)) {
+	mask |= TCL_WRITABLE;
+    }
+
+    /*
+     * If there is data in the input queue, and we aren't waiting for more
+     * data, then we need to schedule a timer so we don't block in the
+     * notifier. Also, cancel the read interest so we don't get duplicate
+     * events.
+     */
+
+    if (mask & TCL_READABLE) {
+	if (!GotFlag(statePtr, CHANNEL_NEED_MORE_DATA)
+		&& (statePtr->inQueueHead != NULL)
+		&& IsBufferReady(statePtr->inQueueHead)) {
+	    mask &= ~TCL_READABLE;
+
+	    /*
+	     * Andreas Kupries, April 11, 2003
+	     *
+	     * Some operating systems (Solaris 2.6 and higher (but not Solaris
+	     * 2.5, go figure)) generate READABLE and EXCEPTION events when
+	     * select()'ing [*] on a plain file, even if EOF was not yet
+	     * reached. This is a problem in the following situation:
+	     *
+	     * - An extension asks to get both READABLE and EXCEPTION events.
+	     * - It reads data into a buffer smaller than the buffer used by
+	     *	 Tcl itself.
+	     * - It does not process all events in the event queue, but only
+	     *	 one, at least in some situations.
+	     *
+	     * In that case we can get into a situation where
+	     *
+	     * - Tcl drops READABLE here, because it has data in its own
+	     *	 buffers waiting to be read by the extension.
+	     * - A READABLE event is syntesized via timer.
+	     * - The OS still reports the EXCEPTION condition on the file.
+	     * - And the extension gets the EXCPTION event first, and handles
+	     *	 this as EOF.
+	     *
+	     * End result ==> Premature end of reading from a file.
+	     *
+	     * The concrete example is 'Expect', and its [expect] command
+	     * (and at the C-level, deep in the bowels of Expect,
+	     * 'exp_get_next_event'. See marker 'SunOS' for commentary in
+	     * that function too).
+	     *
+	     * [*] As the Tcl notifier does. See also for marker 'SunOS' in
+	     * file 'exp_event.c' of Expect.
+	     *
+	     * Our solution here is to drop the interest in the EXCEPTION
+	     * events too. This compiles on all platforms, and also passes the
+	     * testsuite on all of them.
+	     */
+
+	    mask &= ~TCL_EXCEPTION;
+
+	    if (!statePtr->timer) {
+		statePtr->timer = Tcl_CreateTimerHandler(SYNTHETIC_EVENT_TIME,
+                        ChannelTimerProc, chanPtr);
+	    }
+	}
+    }
+    ChanWatch(chanPtr, mask);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ChannelTimerProc --
+ *
+ *	Timer handler scheduled by UpdateInterest to monitor the channel
+ *	buffers until they are empty.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May invoke channel handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ChannelTimerProc(
+    ClientData clientData)
+{
+    Channel *chanPtr = clientData;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    if (!GotFlag(statePtr, CHANNEL_NEED_MORE_DATA)
+	    && (statePtr->interestMask & TCL_READABLE)
+	    && (statePtr->inQueueHead != NULL)
+	    && IsBufferReady(statePtr->inQueueHead)) {
+	/*
+	 * Restart the timer in case a channel handler reenters the event loop
+	 * before UpdateInterest gets called by Tcl_NotifyChannel.
+	 */
+
+	statePtr->timer = Tcl_CreateTimerHandler(SYNTHETIC_EVENT_TIME,
+                ChannelTimerProc,chanPtr);
+	Tcl_Preserve(statePtr);
+	Tcl_NotifyChannel((Tcl_Channel) chanPtr, TCL_READABLE);
+	Tcl_Release(statePtr);
+    } else {
+	statePtr->timer = NULL;
+	UpdateInterest(chanPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateChannelHandler --
+ *
+ *	Arrange for a given procedure to be invoked whenever the channel
+ *	indicated by the chanPtr arg becomes readable or writable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	From now on, whenever the I/O channel given by chanPtr becomes ready
+ *	in the way indicated by mask, proc will be invoked. See the manual
+ *	entry for details on the calling sequence to proc. If there is already
+ *	an event handler for chan, proc and clientData, then the mask will be
+ *	updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateChannelHandler(
+    Tcl_Channel chan,		/* The channel to create the handler for. */
+    int mask,			/* OR'ed combination of TCL_READABLE,
+				 * TCL_WRITABLE, and TCL_EXCEPTION: indicates
+				 * conditions under which proc should be
+				 * called. Use 0 to disable a registered
+				 * handler. */
+    Tcl_ChannelProc *proc,	/* Procedure to call for each selected
+				 * event. */
+    ClientData clientData)	/* Arbitrary data to pass to proc. */
+{
+    ChannelHandler *chPtr;
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    /*
+     * Check whether this channel handler is not already registered. If it is
+     * not, create a new record, else reuse existing record (smash current
+     * values).
+     */
+
+    for (chPtr = statePtr->chPtr; chPtr != NULL; chPtr = chPtr->nextPtr) {
+	if ((chPtr->chanPtr == chanPtr) && (chPtr->proc == proc) &&
+		(chPtr->clientData == clientData)) {
+	    break;
+	}
+    }
+    if (chPtr == NULL) {
+	chPtr = ckalloc(sizeof(ChannelHandler));
+	chPtr->mask = 0;
+	chPtr->proc = proc;
+	chPtr->clientData = clientData;
+	chPtr->chanPtr = chanPtr;
+	chPtr->nextPtr = statePtr->chPtr;
+	statePtr->chPtr = chPtr;
+    }
+
+    /*
+     * The remainder of the initialization below is done regardless of whether
+     * or not this is a new record or a modification of an old one.
+     */
+
+    chPtr->mask = mask;
+
+    /*
+     * Recompute the interest mask for the channel - this call may actually be
+     * disabling an existing handler.
+     */
+
+    statePtr->interestMask = 0;
+    for (chPtr = statePtr->chPtr; chPtr != NULL; chPtr = chPtr->nextPtr) {
+	statePtr->interestMask |= chPtr->mask;
+    }
+
+    UpdateInterest(statePtr->topChanPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteChannelHandler --
+ *
+ *	Cancel a previously arranged callback arrangement for an IO channel.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If a callback was previously registered for this chan, proc and
+ *	clientData, it is removed and the callback will no longer be called
+ *	when the channel becomes ready for IO.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteChannelHandler(
+    Tcl_Channel chan,		/* The channel for which to remove the
+				 * callback. */
+    Tcl_ChannelProc *proc,	/* The procedure in the callback to delete. */
+    ClientData clientData)	/* The client data in the callback to
+				 * delete. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelHandler *chPtr, *prevChPtr;
+    Channel *chanPtr = (Channel *) chan;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    NextChannelHandler *nhPtr;
+
+    /*
+     * Find the entry and the previous one in the list.
+     */
+
+    for (prevChPtr = NULL, chPtr = statePtr->chPtr; chPtr != NULL;
+	    chPtr = chPtr->nextPtr) {
+	if ((chPtr->chanPtr == chanPtr) && (chPtr->clientData == clientData)
+		&& (chPtr->proc == proc)) {
+	    break;
+	}
+	prevChPtr = chPtr;
+    }
+
+    /*
+     * If not found, return without doing anything.
+     */
+
+    if (chPtr == NULL) {
+	return;
+    }
+
+    /*
+     * If Tcl_NotifyChannel is about to process this handler, tell it to
+     * process the next one instead - we are going to delete *this* one.
+     */
+
+    for (nhPtr = tsdPtr->nestedHandlerPtr; nhPtr != NULL;
+	    nhPtr = nhPtr->nestedHandlerPtr) {
+	if (nhPtr->nextHandlerPtr == chPtr) {
+	    nhPtr->nextHandlerPtr = chPtr->nextPtr;
+	}
+    }
+
+    /*
+     * Splice it out of the list of channel handlers.
+     */
+
+    if (prevChPtr == NULL) {
+	statePtr->chPtr = chPtr->nextPtr;
+    } else {
+	prevChPtr->nextPtr = chPtr->nextPtr;
+    }
+    ckfree(chPtr);
+
+    /*
+     * Recompute the interest list for the channel, so that infinite loops
+     * will not result if Tcl_DeleteChannelHandler is called inside an event.
+     */
+
+    statePtr->interestMask = 0;
+    for (chPtr = statePtr->chPtr; chPtr != NULL; chPtr = chPtr->nextPtr) {
+	statePtr->interestMask |= chPtr->mask;
+    }
+
+    UpdateInterest(statePtr->topChanPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteScriptRecord --
+ *
+ *	Delete a script record for this combination of channel, interp and
+ *	mask.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes a script record and cancels a channel event handler.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteScriptRecord(
+    Tcl_Interp *interp,		/* Interpreter in which script was to be
+				 * executed. */
+    Channel *chanPtr,		/* The channel for which to delete the script
+				 * record (if any). */
+    int mask)			/* Events in mask must exactly match mask of
+				 * script to delete. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    EventScriptRecord *esPtr, *prevEsPtr;
+
+    for (esPtr = statePtr->scriptRecordPtr, prevEsPtr = NULL; esPtr != NULL;
+	    prevEsPtr = esPtr, esPtr = esPtr->nextPtr) {
+	if ((esPtr->interp == interp) && (esPtr->mask == mask)) {
+	    if (esPtr == statePtr->scriptRecordPtr) {
+		statePtr->scriptRecordPtr = esPtr->nextPtr;
+	    } else {
+		CLANG_ASSERT(prevEsPtr);
+		prevEsPtr->nextPtr = esPtr->nextPtr;
+	    }
+
+	    Tcl_DeleteChannelHandler((Tcl_Channel) chanPtr,
+		    TclChannelEventScriptInvoker, esPtr);
+
+	    TclDecrRefCount(esPtr->scriptPtr);
+	    ckfree(esPtr);
+
+	    break;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CreateScriptRecord --
+ *
+ *	Creates a record to store a script to be executed when a specific
+ *	event fires on a specific channel.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Causes the script to be stored for later execution.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CreateScriptRecord(
+    Tcl_Interp *interp,		/* Interpreter in which to execute the stored
+				 * script. */
+    Channel *chanPtr,		/* Channel for which script is to be stored */
+    int mask,			/* Set of events for which script will be
+				 * invoked. */
+    Tcl_Obj *scriptPtr)		/* Pointer to script object. */
+{
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+    EventScriptRecord *esPtr;
+    int makeCH;
+
+    for (esPtr=statePtr->scriptRecordPtr; esPtr!=NULL; esPtr=esPtr->nextPtr) {
+	if ((esPtr->interp == interp) && (esPtr->mask == mask)) {
+	    TclDecrRefCount(esPtr->scriptPtr);
+	    esPtr->scriptPtr = NULL;
+	    break;
+	}
+    }
+
+    makeCH = (esPtr == NULL);
+
+    if (makeCH) {
+	esPtr = ckalloc(sizeof(EventScriptRecord));
+    }
+
+    /*
+     * Initialize the structure before calling Tcl_CreateChannelHandler,
+     * because a reflected channel calling 'chan postevent' aka
+     * 'Tcl_NotifyChannel' in its 'watch'Proc will invoke
+     * 'TclChannelEventScriptInvoker' immediately, and we do not wish it to
+     * see uninitialized memory and crash. See [Bug 2918110].
+     */
+
+    esPtr->chanPtr = chanPtr;
+    esPtr->interp = interp;
+    esPtr->mask = mask;
+    Tcl_IncrRefCount(scriptPtr);
+    esPtr->scriptPtr = scriptPtr;
+
+    if (makeCH) {
+	esPtr->nextPtr = statePtr->scriptRecordPtr;
+	statePtr->scriptRecordPtr = esPtr;
+
+	Tcl_CreateChannelHandler((Tcl_Channel) chanPtr, mask,
+		TclChannelEventScriptInvoker, esPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChannelEventScriptInvoker --
+ *
+ *	Invokes a script scheduled by "fileevent" for when the channel becomes
+ *	ready for IO. This function is invoked by the channel handler which
+ *	was created by the Tcl "fileevent" command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Whatever the script does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclChannelEventScriptInvoker(
+    ClientData clientData,	/* The script+interp record. */
+    int mask)			/* Not used. */
+{
+    Tcl_Interp *interp;		/* Interpreter in which to eval the script. */
+    Channel *chanPtr;		/* The channel for which this handler is
+				 * registered. */
+    EventScriptRecord *esPtr;	/* The event script + interpreter to eval it
+				 * in. */
+    int result;			/* Result of call to eval script. */
+
+    esPtr = clientData;
+    chanPtr = esPtr->chanPtr;
+    mask = esPtr->mask;
+    interp = esPtr->interp;
+
+    /*
+     * We must preserve the interpreter so we can report errors on it later.
+     * Note that we do not need to preserve the channel because that is done
+     * by Tcl_NotifyChannel before calling channel handlers.
+     */
+
+    Tcl_Preserve(interp);
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+    result = Tcl_EvalObjEx(interp, esPtr->scriptPtr, TCL_EVAL_GLOBAL);
+
+    /*
+     * On error, cause a background error and remove the channel handler and
+     * the script record.
+     *
+     * NOTE: Must delete channel handler before causing the background error
+     * because the background error may want to reinstall the handler.
+     */
+
+    if (result != TCL_OK) {
+	if (chanPtr->typePtr != NULL) {
+	    DeleteScriptRecord(interp, chanPtr, mask);
+	}
+	Tcl_BackgroundException(interp, result);
+    }
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    Tcl_Release(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FileEventObjCmd --
+ *
+ *	This procedure implements the "fileevent" Tcl command. See the user
+ *	documentation for details on what it does. This command is based on
+ *	the Tk command "fileevent" which in turn is based on work contributed
+ *	by Mark Diekhans.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May create a channel handler for the specified channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_FileEventObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter in which the channel for which
+				 * to create the handler is found. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Channel *chanPtr;		/* The channel to create the handler for. */
+    ChannelState *statePtr;	/* State info for channel */
+    Tcl_Channel chan;		/* The opaque type for the channel. */
+    const char *chanName;
+    int modeIndex;		/* Index of mode argument. */
+    int mask;
+    static const char *const modeOptions[] = {"readable", "writable", NULL};
+    static const int maskArray[] = {TCL_READABLE, TCL_WRITABLE};
+
+    if ((objc != 3) && (objc != 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId event ?script?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[2], modeOptions, "event name", 0,
+	    &modeIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    mask = maskArray[modeIndex];
+
+    chanName = TclGetString(objv[1]);
+    chan = Tcl_GetChannel(interp, chanName, NULL);
+    if (chan == NULL) {
+	return TCL_ERROR;
+    }
+    chanPtr = (Channel *) chan;
+    statePtr = chanPtr->state;
+    if ((statePtr->flags & mask) == 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf("channel is not %s",
+		(mask == TCL_READABLE) ? "readable" : "writable"));
+	return TCL_ERROR;
+    }
+
+    /*
+     * If we are supposed to return the script, do so.
+     */
+
+    if (objc == 3) {
+	EventScriptRecord *esPtr;
+
+	for (esPtr = statePtr->scriptRecordPtr; esPtr != NULL;
+		esPtr = esPtr->nextPtr) {
+	    if ((esPtr->interp == interp) && (esPtr->mask == mask)) {
+		Tcl_SetObjResult(interp, esPtr->scriptPtr);
+		break;
+	    }
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * If we are supposed to delete a stored script, do so.
+     */
+
+    if (*(TclGetString(objv[3])) == '\0') {
+	DeleteScriptRecord(interp, chanPtr, mask);
+	return TCL_OK;
+    }
+
+    /*
+     * Make the script record that will link between the event and the script
+     * to invoke. This also creates a channel event handler which will
+     * evaluate the script in the supplied interpreter.
+     */
+
+    CreateScriptRecord(interp, chanPtr, mask, objv[3]);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZeroTransferTimerProc --
+ *
+ *	Timer handler scheduled by TclCopyChannel so that -command is
+ *	called asynchronously even when -size is 0.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Calls CopyData for -command invocation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ZeroTransferTimerProc(
+    ClientData clientData)
+{
+    /* calling CopyData with mask==0 still implies immediate invocation of the
+     *  -command callback, and completion of the fcopy.
+     */
+    CopyData(clientData, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCopyChannel --
+ *
+ *	This routine copies data from one channel to another, either
+ *	synchronously or asynchronously. If a command script is supplied, the
+ *	operation runs in the background. The script is invoked when the copy
+ *	completes. Otherwise the function waits until the copy is completed
+ *	before returning.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May schedule a background copy operation that causes both channels to
+ *	be marked busy.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCopyChannelOld(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Channel inChan,		/* Channel to read from. */
+    Tcl_Channel outChan,	/* Channel to write to. */
+    int toRead,			/* Amount of data to copy, or -1 for all. */
+    Tcl_Obj *cmdPtr)		/* Pointer to script to execute or NULL. */
+{
+    return TclCopyChannel(interp, inChan, outChan, (Tcl_WideInt) toRead,
+            cmdPtr);
+}
+
+int
+TclCopyChannel(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Channel inChan,		/* Channel to read from. */
+    Tcl_Channel outChan,	/* Channel to write to. */
+    Tcl_WideInt toRead,		/* Amount of data to copy, or -1 for all. */
+    Tcl_Obj *cmdPtr)		/* Pointer to script to execute or NULL. */
+{
+    Channel *inPtr = (Channel *) inChan;
+    Channel *outPtr = (Channel *) outChan;
+    ChannelState *inStatePtr, *outStatePtr;
+    int readFlags, writeFlags;
+    CopyState *csPtr;
+    int nonBlocking = (cmdPtr) ? CHANNEL_NONBLOCKING : 0;
+    int moveBytes;
+
+    inStatePtr = inPtr->state;
+    outStatePtr = outPtr->state;
+
+    if (BUSY_STATE(inStatePtr, TCL_READABLE)) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "channel \"%s\" is busy", Tcl_GetChannelName(inChan)));
+	}
+	return TCL_ERROR;
+    }
+    if (BUSY_STATE(outStatePtr, TCL_WRITABLE)) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "channel \"%s\" is busy", Tcl_GetChannelName(outChan)));
+	}
+	return TCL_ERROR;
+    }
+
+    readFlags = inStatePtr->flags;
+    writeFlags = outStatePtr->flags;
+
+    /*
+     * Set up the blocking mode appropriately. Background copies need
+     * non-blocking channels. Foreground copies need blocking channels. If
+     * there is an error, restore the old blocking mode.
+     */
+
+    if (nonBlocking != (readFlags & CHANNEL_NONBLOCKING)) {
+	if (SetBlockMode(interp, inPtr, nonBlocking ?
+		TCL_MODE_NONBLOCKING : TCL_MODE_BLOCKING) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+    if ((inPtr!=outPtr) && (nonBlocking!=(writeFlags&CHANNEL_NONBLOCKING)) &&
+	    (SetBlockMode(NULL, outPtr, nonBlocking ?
+		    TCL_MODE_NONBLOCKING : TCL_MODE_BLOCKING) != TCL_OK) &&
+	    (nonBlocking != (readFlags & CHANNEL_NONBLOCKING))) {
+	SetBlockMode(NULL, inPtr, (readFlags & CHANNEL_NONBLOCKING)
+		? TCL_MODE_NONBLOCKING : TCL_MODE_BLOCKING);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure the output side is unbuffered.
+     */
+
+    outStatePtr->flags = (outStatePtr->flags & ~CHANNEL_LINEBUFFERED)
+	    | CHANNEL_UNBUFFERED;
+
+    /*
+     * Test for conditions where we know we can just move bytes from input
+     * channel to output channel with no transformation or even examination
+     * of the bytes themselves.
+     */
+
+    moveBytes = inStatePtr->inEofChar == '\0'	/* No eofChar to stop input */
+	    && inStatePtr->inputTranslation == TCL_TRANSLATE_LF
+	    && outStatePtr->outputTranslation == TCL_TRANSLATE_LF
+	    && inStatePtr->encoding == outStatePtr->encoding;
+
+    /*
+     * Allocate a new CopyState to maintain info about the current copy in
+     * progress. This structure will be deallocated when the copy is
+     * completed.
+     */
+
+    csPtr = ckalloc(sizeof(CopyState) + !moveBytes * inStatePtr->bufSize);
+    csPtr->bufSize = !moveBytes * inStatePtr->bufSize;
+    csPtr->readPtr = inPtr;
+    csPtr->writePtr = outPtr;
+    csPtr->readFlags = readFlags;
+    csPtr->writeFlags = writeFlags;
+    csPtr->toRead = toRead;
+    csPtr->total = (Tcl_WideInt) 0;
+    csPtr->interp = interp;
+    if (cmdPtr) {
+	Tcl_IncrRefCount(cmdPtr);
+    }
+    csPtr->cmdPtr = cmdPtr;
+
+    inStatePtr->csPtrR  = csPtr;
+    outStatePtr->csPtrW = csPtr;
+
+    if (moveBytes) {
+	return MoveBytes(csPtr);
+    }
+
+    /*
+     * Special handling of -size 0 async transfers, so that the -command is
+     * still called asynchronously.
+     */
+
+    if ((nonBlocking == CHANNEL_NONBLOCKING) && (toRead == 0)) {
+        Tcl_CreateTimerHandler(0, ZeroTransferTimerProc, csPtr);
+        return 0;
+    }
+
+    /*
+     * Start copying data between the channels.
+     */
+
+    return CopyData(csPtr, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CopyData --
+ *
+ *	This function implements the lowest level of the copying mechanism for
+ *	TclCopyChannel.
+ *
+ * Results:
+ *	Returns TCL_OK on success, else TCL_ERROR.
+ *
+ * Side effects:
+ *	Moves data between channels, may create channel handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MBCallback(
+    CopyState *csPtr,
+    Tcl_Obj *errObj)
+{
+    Tcl_Obj *cmdPtr = Tcl_DuplicateObj(csPtr->cmdPtr);
+    Tcl_WideInt total = csPtr->total;
+    Tcl_Interp *interp = csPtr->interp;
+    int code;
+
+    Tcl_IncrRefCount(cmdPtr);
+    StopCopy(csPtr);
+
+    /* TODO: What if cmdPtr is not a list?! */
+
+    Tcl_ListObjAppendElement(NULL, cmdPtr, Tcl_NewWideIntObj(total));
+    if (errObj) {
+	Tcl_ListObjAppendElement(NULL, cmdPtr, errObj);
+    }
+
+    Tcl_Preserve(interp);
+    code = Tcl_EvalObjEx(interp, cmdPtr, TCL_EVAL_GLOBAL);
+    if (code != TCL_OK) {
+	Tcl_BackgroundException(interp, code);
+    }
+    Tcl_Release(interp);
+    TclDecrRefCount(cmdPtr);
+}
+
+static void
+MBError(
+    CopyState *csPtr,
+    int mask,
+    int errorCode)
+{
+    Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+    Tcl_Channel outChan = (Tcl_Channel) csPtr->writePtr;
+    Tcl_Obj *errObj;
+
+    Tcl_SetErrno(errorCode);
+
+    errObj = Tcl_ObjPrintf( "error %sing \"%s\": %s",
+	    (mask & TCL_READABLE) ? "read" : "writ",
+	    Tcl_GetChannelName((mask & TCL_READABLE) ? inChan : outChan),
+	    Tcl_PosixError(csPtr->interp));
+
+    if (csPtr->cmdPtr) {
+	MBCallback(csPtr, errObj);
+    } else {
+	Tcl_SetObjResult(csPtr->interp, errObj);
+	StopCopy(csPtr);
+    }
+}
+
+static void
+MBEvent(
+    ClientData clientData,
+    int mask)
+{
+    CopyState *csPtr = (CopyState *) clientData;
+    Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+    Tcl_Channel outChan = (Tcl_Channel) csPtr->writePtr;
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+
+    if (mask & TCL_WRITABLE) {
+	Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	Tcl_DeleteChannelHandler(outChan, MBEvent, csPtr);
+	switch (MBWrite(csPtr)) {
+	case TCL_OK:
+	    MBCallback(csPtr, NULL);
+	    break;
+	case TCL_CONTINUE:
+	    Tcl_CreateChannelHandler(inChan, TCL_READABLE, MBEvent, csPtr);
+	    break;
+	}
+    } else if (mask & TCL_READABLE) {
+	if (TCL_OK == MBRead(csPtr)) {
+	    /* When at least one full buffer is present, stop reading. */
+	    if (IsBufferFull(inStatePtr->inQueueHead)
+		    || !Tcl_InputBlocked(inChan)) {
+		Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	    }
+
+	    /* Successful read -- set up to write the bytes we read */
+	    Tcl_CreateChannelHandler(outChan, TCL_WRITABLE, MBEvent, csPtr);
+	}
+    }
+}
+
+static int
+MBRead(
+    CopyState *csPtr)
+{
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+    ChannelBuffer *bufPtr = inStatePtr->inQueueHead;
+    int code;
+
+    if (bufPtr && BytesLeft(bufPtr) > 0) {
+	return TCL_OK;
+    }
+
+    code = GetInput(inStatePtr->topChanPtr);
+    if (code == 0 || GotFlag(inStatePtr, CHANNEL_BLOCKED)) {
+	return TCL_OK;
+    } else {
+	MBError(csPtr, TCL_READABLE, code);
+	return TCL_ERROR;
+    }
+}
+
+static int
+MBWrite(
+    CopyState *csPtr)
+{
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+    ChannelState *outStatePtr = csPtr->writePtr->state;
+    ChannelBuffer *bufPtr = inStatePtr->inQueueHead;
+    ChannelBuffer *tail = NULL;
+    int code;
+    Tcl_WideInt inBytes = 0;
+
+    /* Count up number of bytes waiting in the input queue */
+    while (bufPtr) {
+	inBytes += BytesLeft(bufPtr);
+	tail = bufPtr;
+	if (csPtr->toRead != -1 && csPtr->toRead < inBytes) {
+	    /* Queue has enough bytes to complete the copy */
+	    break;
+	}
+	bufPtr = bufPtr->nextPtr;
+    }
+
+    if (bufPtr) {
+	/* Split the overflowing buffer in two */
+	int extra = (int) (inBytes - csPtr->toRead);
+        /* Note that going with int for extra assumes that inBytes is not too
+         * much over toRead to require a wide itself. If that gets violated
+         * then the calculations involving extra must be made wide too.
+         *
+         * Noted with Win32/MSVC debug build treating the warning (possible of
+         * data in __int64 to int conversion) as error.
+         */
+
+	bufPtr = AllocChannelBuffer(extra);
+
+	tail->nextAdded -= extra;
+	memcpy(InsertPoint(bufPtr), InsertPoint(tail), extra);
+	bufPtr->nextAdded += extra;
+	bufPtr->nextPtr = tail->nextPtr;
+	tail->nextPtr = NULL;
+	inBytes = csPtr->toRead;
+    }
+
+    /* Update the byte counts */
+    if (csPtr->toRead != -1) {
+	csPtr->toRead -= inBytes;
+    }
+    csPtr->total += inBytes;
+
+    /* Move buffers from input to output channels */
+    if (outStatePtr->outQueueTail) {
+	outStatePtr->outQueueTail->nextPtr = inStatePtr->inQueueHead;
+    } else {
+	outStatePtr->outQueueHead = inStatePtr->inQueueHead;
+    }
+    outStatePtr->outQueueTail = tail;
+    inStatePtr->inQueueHead = bufPtr;
+    if (inStatePtr->inQueueTail == tail) {
+	inStatePtr->inQueueTail = bufPtr;
+    }
+    if (bufPtr == NULL) {
+	inStatePtr->inQueueTail = NULL;
+    }
+
+    code = FlushChannel(csPtr->interp, outStatePtr->topChanPtr, 0);
+    if (code) {
+	MBError(csPtr, TCL_WRITABLE, code);
+	return TCL_ERROR;
+    }
+    if (csPtr->toRead == 0 || GotFlag(inStatePtr, CHANNEL_EOF)) {
+	return TCL_OK;
+    }
+    return TCL_CONTINUE;
+}
+
+static int
+MoveBytes(
+    CopyState *csPtr)		/* State of copy operation. */
+{
+    ChannelState *outStatePtr = csPtr->writePtr->state;
+    ChannelBuffer *bufPtr = outStatePtr->curOutPtr;
+    int errorCode;
+
+    if (bufPtr && BytesLeft(bufPtr)) {
+	/* If we start with unflushed bytes in the destination
+	 * channel, flush them out of the way first. */
+
+	errorCode = FlushChannel(csPtr->interp, outStatePtr->topChanPtr, 0);
+	if (errorCode != 0) {
+	    MBError(csPtr, TCL_WRITABLE, errorCode);
+	    return TCL_ERROR;
+	}
+    }
+
+    if (csPtr->cmdPtr) {
+	Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+	Tcl_CreateChannelHandler(inChan, TCL_READABLE, MBEvent, csPtr);
+	return TCL_OK;
+    }
+
+    while (1) {
+	int code;
+
+	if (TCL_ERROR == MBRead(csPtr)) {
+	    return TCL_ERROR;
+	}
+	code = MBWrite(csPtr);
+	if (code == TCL_OK) {
+	    Tcl_SetObjResult(csPtr->interp, Tcl_NewWideIntObj(csPtr->total));
+	    StopCopy(csPtr);
+	    return TCL_OK;
+	}
+	if (code == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	/* code == TCL_CONTINUE --> continue the loop */
+    }
+    return TCL_OK;	/* Silence compiler warnings */
+}
+
+static int
+CopyData(
+    CopyState *csPtr,		/* State of copy operation. */
+    int mask)			/* Current channel event flags. */
+{
+    Tcl_Interp *interp;
+    Tcl_Obj *cmdPtr, *errObj = NULL, *bufObj = NULL, *msg = NULL;
+    Tcl_Channel inChan, outChan;
+    ChannelState *inStatePtr, *outStatePtr;
+    int result = TCL_OK, size, sizeb;
+    Tcl_WideInt total;
+    const char *buffer;
+    int inBinary, outBinary, sameEncoding;
+				/* Encoding control */
+    int underflow;		/* Input underflow */
+
+    inChan	= (Tcl_Channel) csPtr->readPtr;
+    outChan	= (Tcl_Channel) csPtr->writePtr;
+    inStatePtr	= csPtr->readPtr->state;
+    outStatePtr	= csPtr->writePtr->state;
+    interp	= csPtr->interp;
+    cmdPtr	= csPtr->cmdPtr;
+
+    /*
+     * Copy the data the slow way, using the translation mechanism.
+     *
+     * Note: We have make sure that we use the topmost channel in a stack for
+     * the copying. The caller uses Tcl_GetChannel to access it, and thus gets
+     * the bottom of the stack.
+     */
+
+    inBinary = (inStatePtr->encoding == NULL);
+    outBinary = (outStatePtr->encoding == NULL);
+    sameEncoding = (inStatePtr->encoding == outStatePtr->encoding);
+
+    if (!(inBinary || sameEncoding)) {
+	TclNewObj(bufObj);
+	Tcl_IncrRefCount(bufObj);
+    }
+
+    while (csPtr->toRead != (Tcl_WideInt) 0) {
+	/*
+	 * Check for unreported background errors.
+	 */
+
+	Tcl_GetChannelError(inChan, &msg);
+	if ((inStatePtr->unreportedError != 0) || (msg != NULL)) {
+	    Tcl_SetErrno(inStatePtr->unreportedError);
+	    inStatePtr->unreportedError = 0;
+	    goto readError;
+	}
+	Tcl_GetChannelError(outChan, &msg);
+	if ((outStatePtr->unreportedError != 0) || (msg != NULL)) {
+	    Tcl_SetErrno(outStatePtr->unreportedError);
+	    outStatePtr->unreportedError = 0;
+	    goto writeError;
+	}
+
+	if (cmdPtr && (mask == 0)) {
+	    /*
+	     * In async mode, we skip reading synchronously and fake an
+	     * underflow instead to prime the readable fileevent.
+	     */
+
+	    size = 0;
+	    underflow = 1;
+	} else {
+	    /*
+	     * Read up to bufSize bytes.
+	     */
+
+	    if ((csPtr->toRead == (Tcl_WideInt) -1)
+                    || (csPtr->toRead > (Tcl_WideInt) csPtr->bufSize)) {
+		sizeb = csPtr->bufSize;
+	    } else {
+		sizeb = (int) csPtr->toRead;
+	    }
+
+	    if (inBinary || sameEncoding) {
+		size = DoRead(inStatePtr->topChanPtr, csPtr->buffer, sizeb,
+                              !GotFlag(inStatePtr, CHANNEL_NONBLOCKING));
+	    } else {
+		size = DoReadChars(inStatePtr->topChanPtr, bufObj, sizeb,
+			0 /* No append */);
+	    }
+	    underflow = (size >= 0) && (size < sizeb);	/* Input underflow */
+	}
+
+	if (size < 0) {
+	readError:
+	    if (interp) {
+		TclNewObj(errObj);
+		Tcl_AppendStringsToObj(errObj, "error reading \"",
+			Tcl_GetChannelName(inChan), "\": ", NULL);
+		if (msg != NULL) {
+		    Tcl_AppendObjToObj(errObj, msg);
+		} else {
+		    Tcl_AppendStringsToObj(errObj, Tcl_PosixError(interp),
+			    NULL);
+		}
+	    }
+	    if (msg != NULL) {
+		Tcl_DecrRefCount(msg);
+	    }
+	    break;
+	} else if (underflow) {
+	    /*
+	     * We had an underflow on the read side. If we are at EOF, and not
+	     * in the synchronous part of an asynchronous fcopy, then the
+	     * copying is done, otherwise set up a channel handler to detect
+	     * when the channel becomes readable again.
+	     */
+
+	    if ((size == 0) && Tcl_Eof(inChan) && !(cmdPtr && (mask == 0))) {
+		break;
+	    }
+	    if (cmdPtr && (!Tcl_Eof(inChan) || (mask == 0)) &&
+                !(mask & TCL_READABLE)) {
+		if (mask & TCL_WRITABLE) {
+		    Tcl_DeleteChannelHandler(outChan, CopyEventProc, csPtr);
+		}
+		Tcl_CreateChannelHandler(inChan, TCL_READABLE, CopyEventProc,
+			csPtr);
+	    }
+	    if (size == 0) {
+		if (!GotFlag(inStatePtr, CHANNEL_NONBLOCKING)) {
+		    /*
+                     * We allowed a short read.  Keep trying.
+                     */
+
+		    continue;
+		}
+		if (bufObj != NULL) {
+		    TclDecrRefCount(bufObj);
+		    bufObj = NULL;
+		}
+		return TCL_OK;
+	    }
+	}
+
+	/*
+	 * Now write the buffer out.
+	 */
+
+	if (inBinary || sameEncoding) {
+	    buffer = csPtr->buffer;
+	    sizeb = size;
+	} else {
+	    buffer = TclGetStringFromObj(bufObj, &sizeb);
+	}
+
+	if (outBinary || sameEncoding) {
+	    sizeb = WriteBytes(outStatePtr->topChanPtr, buffer, sizeb);
+	} else {
+	    sizeb = WriteChars(outStatePtr->topChanPtr, buffer, sizeb);
+	}
+
+	/*
+	 * [Bug 2895565]. At this point 'size' still contains the number of
+	 * bytes or characters which have been read. We keep this to later to
+	 * update the totals and toRead information, see marker (UP) below. We
+	 * must not overwrite it with 'sizeb', which is the number of written
+	 * bytes or characters, and both EOL translation and encoding
+	 * conversion may have changed this number unpredictably in relation
+	 * to 'size' (It can be smaller or larger, in the latter case able to
+	 * drive toRead below -1, causing infinite looping). Completely
+	 * unsuitable for updating totals and toRead.
+	 */
+
+	if (sizeb < 0) {
+	writeError:
+	    if (interp) {
+		TclNewObj(errObj);
+		Tcl_AppendStringsToObj(errObj, "error writing \"",
+			Tcl_GetChannelName(outChan), "\": ", NULL);
+		if (msg != NULL) {
+		    Tcl_AppendObjToObj(errObj, msg);
+		} else {
+		    Tcl_AppendStringsToObj(errObj, Tcl_PosixError(interp),
+			    NULL);
+		}
+	    }
+	    if (msg != NULL) {
+		Tcl_DecrRefCount(msg);
+	    }
+	    break;
+	}
+
+	/*
+	 * Update the current byte count. Do it now so the count is valid
+	 * before a return or break takes us out of the loop. The invariant at
+	 * the top of the loop should be that csPtr->toRead holds the number
+	 * of bytes left to copy.
+	 */
+
+	if (csPtr->toRead != -1) {
+	    csPtr->toRead -= size;
+	}
+	csPtr->total += size;
+
+	/*
+	 * Break loop if EOF && (size>0)
+	 */
+
+	if (Tcl_Eof(inChan)) {
+	    break;
+	}
+
+	/*
+	 * Check to see if the write is happening in the background. If so,
+	 * stop copying and wait for the channel to become writable again.
+	 * After input underflow we already installed a readable handler
+	 * therefore we don't need a writable handler.
+	 */
+
+	if (!underflow && GotFlag(outStatePtr, BG_FLUSH_SCHEDULED)) {
+	    if (!(mask & TCL_WRITABLE)) {
+		if (mask & TCL_READABLE) {
+		    Tcl_DeleteChannelHandler(inChan, CopyEventProc, csPtr);
+		}
+		Tcl_CreateChannelHandler(outChan, TCL_WRITABLE,
+			CopyEventProc, csPtr);
+	    }
+	    if (bufObj != NULL) {
+		TclDecrRefCount(bufObj);
+		bufObj = NULL;
+	    }
+	    return TCL_OK;
+	}
+
+	/*
+	 * For background copies, we only do one buffer per invocation so we
+	 * don't starve the rest of the system.
+	 */
+
+	if (cmdPtr && (csPtr->toRead != 0)) {
+	    /*
+	     * The first time we enter this code, there won't be a channel
+	     * handler established yet, so do it here.
+	     */
+
+	    if (mask == 0) {
+		Tcl_CreateChannelHandler(outChan, TCL_WRITABLE, CopyEventProc,
+			csPtr);
+	    }
+	    if (bufObj != NULL) {
+		TclDecrRefCount(bufObj);
+		bufObj = NULL;
+	    }
+	    return TCL_OK;
+	}
+    } /* while */
+
+    if (bufObj != NULL) {
+	TclDecrRefCount(bufObj);
+	bufObj = NULL;
+    }
+
+    /*
+     * Make the callback or return the number of bytes transferred. The local
+     * total is used because StopCopy frees csPtr.
+     */
+
+    total = csPtr->total;
+    if (cmdPtr && interp) {
+	int code;
+
+	/*
+	 * Get a private copy of the command so we can mutate it by adding
+	 * arguments. Note that StopCopy frees our saved reference to the
+	 * original command obj.
+	 */
+
+	cmdPtr = Tcl_DuplicateObj(cmdPtr);
+	Tcl_IncrRefCount(cmdPtr);
+	StopCopy(csPtr);
+	Tcl_Preserve(interp);
+
+	Tcl_ListObjAppendElement(interp, cmdPtr, Tcl_NewWideIntObj(total));
+	if (errObj) {
+	    Tcl_ListObjAppendElement(interp, cmdPtr, errObj);
+	}
+	code = Tcl_EvalObjEx(interp, cmdPtr, TCL_EVAL_GLOBAL);
+	if (code != TCL_OK) {
+	    Tcl_BackgroundException(interp, code);
+	    result = TCL_ERROR;
+	}
+	TclDecrRefCount(cmdPtr);
+	Tcl_Release(interp);
+    } else {
+	StopCopy(csPtr);
+	if (interp) {
+	    if (errObj) {
+		Tcl_SetObjResult(interp, errObj);
+		result = TCL_ERROR;
+	    } else {
+		Tcl_ResetResult(interp);
+		Tcl_SetObjResult(interp, Tcl_NewWideIntObj(total));
+	    }
+	}
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DoRead --
+ *
+ *	Stores up to "bytesToRead" bytes in memory pointed to by "dst".
+ *	These bytes come from reading the channel "chanPtr" and
+ *	performing the configured translations.  No encoding conversions
+ *	are applied to the bytes being read.
+ *
+ * Results:
+ *	The number of bytes actually stored (<= bytesToRead),
+ * 	or -1 if there is an error in reading the channel.  Use
+ * 	Tcl_GetErrno() to retrieve the error code for the error
+ *	that occurred.
+ *
+ *	The number of bytes stored can be less than the number
+ * 	requested when
+ *	  - EOF is reached on the channel; or
+ *	  - the channel is non-blocking, and we've read all we can
+ *	    without blocking.
+ *	  - a channel reading error occurs (and we return -1)
+ *
+ * Side effects:
+ *	May cause input to be buffered.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DoRead(
+    Channel *chanPtr,		/* The channel from which to read. */
+    char *dst,			/* Where to store input read. */
+    int bytesToRead,		/* Maximum number of bytes to read. */
+    int allowShortReads)	/* Allow half-blocking (pipes,sockets) */
+{
+    ChannelState *statePtr = chanPtr->state;
+    char *p = dst;
+
+    assert(bytesToRead >= 0);
+
+    /*
+     * Early out when we know a read will get the eofchar.
+     *
+     * NOTE: This seems to be a bug.  The special handling for
+     * a zero-char read request ought to come first.  As coded
+     * the EOF due to eofchar has distinguishing behavior from
+     * the EOF due to reported EOF on the underlying device, and
+     * that seems undesirable.  However recent history indicates
+     * that new inconsistent behavior in a patchlevel has problems
+     * too.  Keep on keeping on for now.
+     */
+
+    if (GotFlag(statePtr, CHANNEL_STICKY_EOF)) {
+	SetFlag(statePtr, CHANNEL_EOF);
+	assert(statePtr->inputEncodingFlags & TCL_ENCODING_END);
+	assert(!GotFlag(statePtr, CHANNEL_BLOCKED|INPUT_SAW_CR));
+
+	/* TODO: Don't need this call */
+	UpdateInterest(chanPtr);
+	return 0;
+    }
+
+    /*
+     * Special handling for zero-char read request.
+     */
+
+    if (bytesToRead == 0) {
+	if (GotFlag(statePtr, CHANNEL_EOF)) {
+	    statePtr->inputEncodingFlags |= TCL_ENCODING_START;
+	}
+	ResetFlag(statePtr, CHANNEL_BLOCKED|CHANNEL_EOF);
+	statePtr->inputEncodingFlags &= ~TCL_ENCODING_END;
+	/* TODO: Don't need this call */
+	UpdateInterest(chanPtr);
+	return 0;
+    }
+
+    TclChannelPreserve((Tcl_Channel)chanPtr);
+    while (bytesToRead) {
+	/*
+	 * Each pass through the loop is intended to process up to one channel
+	 * buffer.
+	 */
+
+	int bytesRead, bytesWritten;
+	ChannelBuffer *bufPtr = statePtr->inQueueHead;
+
+	/*
+	 * Don't read more data if we have what we need.
+	 */
+
+	while (!bufPtr ||			/* We got no buffer!   OR */
+		(!IsBufferFull(bufPtr) && 	/* Our buffer has room AND */
+		(BytesLeft(bufPtr) < bytesToRead))) {
+						/* Not enough bytes in it yet
+						 * to fill the dst */
+	    int code;
+
+	moreData:
+	    code = GetInput(chanPtr);
+	    bufPtr = statePtr->inQueueHead;
+
+	    assert(bufPtr != NULL);
+
+	    if (GotFlag(statePtr, CHANNEL_EOF|CHANNEL_BLOCKED)) {
+		/*
+                 * Further reads cannot do any more.
+                 */
+
+		break;
+	    }
+
+	    if (code) {
+		/*
+                 * Read error
+                 */
+
+		UpdateInterest(chanPtr);
+		TclChannelRelease((Tcl_Channel)chanPtr);
+		return -1;
+	    }
+
+	    assert(IsBufferFull(bufPtr));
+	}
+
+	assert(bufPtr != NULL);
+
+	bytesRead = BytesLeft(bufPtr);
+	bytesWritten = bytesToRead;
+
+	TranslateInputEOL(statePtr, p, RemovePoint(bufPtr),
+		&bytesWritten, &bytesRead);
+	bufPtr->nextRemoved += bytesRead;
+	p += bytesWritten;
+	bytesToRead -= bytesWritten;
+
+	if (!IsBufferEmpty(bufPtr)) {
+	    /*
+	     * Buffer is not empty.  How can that be?
+	     *
+	     * 0) We stopped early because we got all the bytes we were
+	     *    seeking. That's fine.
+	     */
+
+	    if (bytesToRead == 0) {
+		break;
+	    }
+
+	    /*
+	     * 1) We're @EOF because we saw eof char.
+	     */
+
+	    if (GotFlag(statePtr, CHANNEL_STICKY_EOF)) {
+		break;
+	    }
+
+	    /*
+	     * 2) The buffer holds a \r while in CRLF translation, followed by
+	     *    the end of the buffer.
+	     */
+
+	    assert(statePtr->inputTranslation == TCL_TRANSLATE_CRLF);
+	    assert(RemovePoint(bufPtr)[0] == '\r');
+	    assert(BytesLeft(bufPtr) == 1);
+
+	    if (bufPtr->nextPtr == NULL) {
+		/*
+                 * There's no more buffered data...
+                 */
+
+		if (statePtr->flags & CHANNEL_EOF) {
+		    /*
+                     * ...and there never will be.
+                     */
+
+		    *p++ = '\r';
+		    bytesToRead--;
+		    bufPtr->nextRemoved++;
+		} else if (statePtr->flags & CHANNEL_BLOCKED) {
+		    /*
+                     * ...and we cannot get more now.
+                     */
+
+		    SetFlag(statePtr, CHANNEL_NEED_MORE_DATA);
+		    break;
+		} else {
+		    /*
+                     * ...so we need to get some.
+                     */
+
+		    goto moreData;
+		}
+	    }
+
+	    if (bufPtr->nextPtr) {
+		/*
+                 * There's a next buffer.  Shift orphan \r to it.
+                 */
+
+		ChannelBuffer *nextPtr = bufPtr->nextPtr;
+
+		nextPtr->nextRemoved -= 1;
+		RemovePoint(nextPtr)[0] = '\r';
+		bufPtr->nextRemoved++;
+	    }
+	}
+
+	if (IsBufferEmpty(bufPtr)) {
+	    statePtr->inQueueHead = bufPtr->nextPtr;
+	    if (statePtr->inQueueHead == NULL) {
+		statePtr->inQueueTail = NULL;
+	    }
+	    RecycleBuffer(statePtr, bufPtr, 0);
+	    bufPtr = statePtr->inQueueHead;
+	}
+
+	if ((GotFlag(statePtr, CHANNEL_NONBLOCKING) || allowShortReads)
+		&& GotFlag(statePtr, CHANNEL_BLOCKED)) {
+	    break;
+	}
+
+	/*
+	 * When there's no buffered data to read, and we're at EOF, escape to
+	 * the caller.
+	 */
+
+	if (GotFlag(statePtr, CHANNEL_EOF)
+		&& (bufPtr == NULL || IsBufferEmpty(bufPtr))) {
+	    break;
+	}
+    }
+    if (bytesToRead == 0) {
+	ResetFlag(statePtr, CHANNEL_BLOCKED);
+    }
+
+    assert(!GotFlag(statePtr, CHANNEL_EOF)
+	    || GotFlag(statePtr, CHANNEL_STICKY_EOF)
+	    || Tcl_InputBuffered((Tcl_Channel)chanPtr) == 0);
+    assert(!(GotFlag(statePtr, CHANNEL_EOF|CHANNEL_BLOCKED)
+	    == (CHANNEL_EOF|CHANNEL_BLOCKED)));
+    UpdateInterest(chanPtr);
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    return (int)(p - dst);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CopyEventProc --
+ *
+ *	This routine is invoked as a channel event handler for the background
+ *	copy operation. It is just a trivial wrapper around the CopyData
+ *	routine.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CopyEventProc(
+    ClientData clientData,
+    int mask)
+{
+    (void) CopyData(clientData, mask);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StopCopy --
+ *
+ *	This routine halts a copy that is in progress.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes any pending channel handlers and restores the blocking and
+ *	buffering modes of the channels. The CopyState is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+StopCopy(
+    CopyState *csPtr)		/* State for bg copy to stop . */
+{
+    ChannelState *inStatePtr, *outStatePtr;
+    Tcl_Channel inChan, outChan;
+
+    int nonBlocking;
+
+    if (!csPtr) {
+	return;
+    }
+
+    inChan = (Tcl_Channel) csPtr->readPtr;
+    outChan = (Tcl_Channel) csPtr->writePtr;
+    inStatePtr = csPtr->readPtr->state;
+    outStatePtr = csPtr->writePtr->state;
+
+    /*
+     * Restore the old blocking mode and output buffering mode.
+     */
+
+    nonBlocking = csPtr->readFlags & CHANNEL_NONBLOCKING;
+    if (nonBlocking != (inStatePtr->flags & CHANNEL_NONBLOCKING)) {
+	SetBlockMode(NULL, csPtr->readPtr,
+		nonBlocking ? TCL_MODE_NONBLOCKING : TCL_MODE_BLOCKING);
+    }
+    if (csPtr->readPtr != csPtr->writePtr) {
+	nonBlocking = csPtr->writeFlags & CHANNEL_NONBLOCKING;
+	if (nonBlocking != (outStatePtr->flags & CHANNEL_NONBLOCKING)) {
+	    SetBlockMode(NULL, csPtr->writePtr,
+		    nonBlocking ? TCL_MODE_NONBLOCKING : TCL_MODE_BLOCKING);
+	}
+    }
+    ResetFlag(outStatePtr, CHANNEL_LINEBUFFERED | CHANNEL_UNBUFFERED);
+    outStatePtr->flags |=
+	    csPtr->writeFlags & (CHANNEL_LINEBUFFERED | CHANNEL_UNBUFFERED);
+
+    if (csPtr->cmdPtr) {
+	Tcl_DeleteChannelHandler(inChan, CopyEventProc, csPtr);
+	if (inChan != outChan) {
+	    Tcl_DeleteChannelHandler(outChan, CopyEventProc, csPtr);
+	}
+	Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	Tcl_DeleteChannelHandler(outChan, MBEvent, csPtr);
+	TclDecrRefCount(csPtr->cmdPtr);
+    }
+    inStatePtr->csPtrR = NULL;
+    outStatePtr->csPtrW = NULL;
+    ckfree(csPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StackSetBlockMode --
+ *
+ *	This function sets the blocking mode for a channel, iterating through
+ *	each channel in a stack and updates the state flags.
+ *
+ * Results:
+ *	0 if OK, result code from failed blockModeProc otherwise.
+ *
+ * Side effects:
+ *	Modifies the blocking mode of the channel and possibly generates an
+ *	error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StackSetBlockMode(
+    Channel *chanPtr,		/* Channel to modify. */
+    int mode)			/* One of TCL_MODE_BLOCKING or
+				 * TCL_MODE_NONBLOCKING. */
+{
+    int result = 0;
+    Tcl_DriverBlockModeProc *blockModeProc;
+    ChannelState *statePtr = chanPtr->state;
+
+    /*
+     * Start at the top of the channel stack
+     * TODO: Examine what can go wrong when blockModeProc calls
+     * disturb the stacking state of the channel.
+     */
+
+    chanPtr = statePtr->topChanPtr;
+    while (chanPtr != NULL) {
+	blockModeProc = Tcl_ChannelBlockModeProc(chanPtr->typePtr);
+	if (blockModeProc != NULL) {
+	    result = blockModeProc(chanPtr->instanceData, mode);
+	    if (result != 0) {
+		Tcl_SetErrno(result);
+		return result;
+	    }
+	}
+	chanPtr = chanPtr->downChanPtr;
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetBlockMode --
+ *
+ *	This function sets the blocking mode for a channel and updates the
+ *	state flags.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Modifies the blocking mode of the channel and possibly generates an
+ *	error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetBlockMode(
+    Tcl_Interp *interp,		/* Interp for error reporting. */
+    Channel *chanPtr,		/* Channel to modify. */
+    int mode)			/* One of TCL_MODE_BLOCKING or
+				 * TCL_MODE_NONBLOCKING. */
+{
+    int result = 0;
+    ChannelState *statePtr = chanPtr->state;
+				/* State info for channel */
+
+    result = StackSetBlockMode(chanPtr, mode);
+    if (result != 0) {
+	if (interp != NULL) {
+	    /*
+	     * TIP #219.
+	     * Move error messages put by the driver into the bypass area and
+	     * put them into the regular interpreter result. Fall back to the
+	     * regular message if nothing was found in the bypass.
+	     *
+	     * Note that we cannot have a message in the interpreter bypass
+	     * area, StackSetBlockMode is restricted to the channel bypass.
+	     * We still need the interp as the destination of the move.
+	     */
+
+	    if (!TclChanCaughtErrorBypass(interp, (Tcl_Channel) chanPtr)) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "error setting blocking mode: %s",
+			Tcl_PosixError(interp)));
+	    }
+	} else {
+	    /*
+	     * TIP #219.
+	     * If we have no interpreter to put a bypass message into we have
+	     * to clear it, to prevent its propagation and use in other places
+	     * unrelated to the actual occurence of the problem.
+	     */
+
+	    Tcl_SetChannelError((Tcl_Channel) chanPtr, NULL);
+	}
+	return TCL_ERROR;
+    }
+    if (mode == TCL_MODE_BLOCKING) {
+	ResetFlag(statePtr, CHANNEL_NONBLOCKING | BG_FLUSH_SCHEDULED);
+    } else {
+	SetFlag(statePtr, CHANNEL_NONBLOCKING);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelNames --
+ *
+ *	Return the names of all open channels in the interp.
+ *
+ * Results:
+ *	TCL_OK or TCL_ERROR.
+ *
+ * Side effects:
+ *	Interp result modified with list of channel names.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelNames(
+    Tcl_Interp *interp)		/* Interp for error reporting. */
+{
+    return Tcl_GetChannelNamesEx(interp, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelNamesEx --
+ *
+ *	Return the names of open channels in the interp filtered filtered
+ *	through a pattern. If pattern is NULL, it returns all the open
+ *	channels.
+ *
+ * Results:
+ *	TCL_OK or TCL_ERROR.
+ *
+ * Side effects:
+ *	Interp result modified with list of channel names.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetChannelNamesEx(
+    Tcl_Interp *interp,		/* Interp for error reporting. */
+    const char *pattern)	/* Pattern to filter on. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ChannelState *statePtr;
+    const char *name;		/* Name for channel */
+    Tcl_Obj *resultPtr;		/* Pointer to result object */
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    Tcl_HashSearch hSearch;	/* Search variable. */
+
+    if (interp == NULL) {
+	return TCL_OK;
+    }
+
+    /*
+     * Get the channel table that stores the channels registered for this
+     * interpreter.
+     */
+
+    hTblPtr = GetChannelTable(interp);
+    TclNewObj(resultPtr);
+    if ((pattern != NULL) && TclMatchIsTrivial(pattern)
+	    && !((pattern[0] == 's') && (pattern[1] == 't')
+	    && (pattern[2] == 'd'))) {
+	if ((Tcl_FindHashEntry(hTblPtr, pattern) != NULL)
+		&& (Tcl_ListObjAppendElement(interp, resultPtr,
+		Tcl_NewStringObj(pattern, -1)) != TCL_OK)) {
+	    goto error;
+	}
+	goto done;
+    }
+
+    for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch); hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&hSearch)) {
+	statePtr = ((Channel *) Tcl_GetHashValue(hPtr))->state;
+
+	if (statePtr->topChanPtr == (Channel *) tsdPtr->stdinChannel) {
+	    name = "stdin";
+	} else if (statePtr->topChanPtr == (Channel *) tsdPtr->stdoutChannel) {
+	    name = "stdout";
+	} else if (statePtr->topChanPtr == (Channel *) tsdPtr->stderrChannel) {
+	    name = "stderr";
+	} else {
+	    /*
+	     * This is also stored in Tcl_GetHashKey(hTblPtr, hPtr), but it's
+	     * simpler to just grab the name from the statePtr.
+	     */
+
+	    name = statePtr->channelName;
+	}
+
+	if (((pattern == NULL) || Tcl_StringMatch(name, pattern)) &&
+		(Tcl_ListObjAppendElement(interp, resultPtr,
+			Tcl_NewStringObj(name, -1)) != TCL_OK)) {
+	error:
+	    TclDecrRefCount(resultPtr);
+	    return TCL_ERROR;
+	}
+    }
+
+  done:
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsChannelRegistered --
+ *
+ *	Checks whether the channel is associated with the interp. See also
+ *	Tcl_RegisterChannel and Tcl_UnregisterChannel.
+ *
+ * Results:
+ *	0 if the channel is not registered in the interpreter, 1 else.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsChannelRegistered(
+    Tcl_Interp *interp,		/* The interp to query of the channel */
+    Tcl_Channel chan)		/* The channel to check */
+{
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    Channel *chanPtr;		/* The real IO channel. */
+    ChannelState *statePtr;	/* State of the real channel. */
+
+    /*
+     * Always check bottom-most channel in the stack. This is the one that
+     * gets registered.
+     */
+
+    chanPtr = ((Channel *) chan)->state->bottomChanPtr;
+    statePtr = chanPtr->state;
+
+    hTblPtr = Tcl_GetAssocData(interp, "tclIO", NULL);
+    if (hTblPtr == NULL) {
+	return 0;
+    }
+    hPtr = Tcl_FindHashEntry(hTblPtr, statePtr->channelName);
+    if (hPtr == NULL) {
+	return 0;
+    }
+    if ((Channel *) Tcl_GetHashValue(hPtr) != chanPtr) {
+	return 0;
+    }
+
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsChannelShared --
+ *
+ *	Checks whether the channel is shared by multiple interpreters.
+ *
+ * Results:
+ *	A boolean value (0 = Not shared, 1 = Shared).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsChannelShared(
+    Tcl_Channel chan)		/* The channel to query */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+				/* State of real channel structure. */
+
+    return ((statePtr->refCount > 1) ? 1 : 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsChannelExisting --
+ *
+ *	Checks whether a channel of the given name exists in the
+ *	(thread)-global list of all channels. See Tcl_GetChannelNamesEx for
+ *	function exposed at the Tcl level.
+ *
+ * Results:
+ *	A boolean value (0 = Does not exist, 1 = Does exist).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsChannelExisting(
+    const char *chanName)	/* The name of the channel to look for. */
+{
+    ChannelState *statePtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    const char *name;
+    int chanNameLen;
+
+    chanNameLen = strlen(chanName);
+    for (statePtr = tsdPtr->firstCSPtr; statePtr != NULL;
+	    statePtr = statePtr->nextCSPtr) {
+	if (statePtr->topChanPtr == (Channel *) tsdPtr->stdinChannel) {
+	    name = "stdin";
+	} else if (statePtr->topChanPtr == (Channel *) tsdPtr->stdoutChannel) {
+	    name = "stdout";
+	} else if (statePtr->topChanPtr == (Channel *) tsdPtr->stderrChannel) {
+	    name = "stderr";
+	} else {
+	    name = statePtr->channelName;
+	}
+
+	if ((*chanName == *name) &&
+		(memcmp(name, chanName, (size_t) chanNameLen + 1) == 0)) {
+	    return 1;
+	}
+    }
+
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelName --
+ *
+ *	Return the name of the channel type.
+ *
+ * Results:
+ *	A pointer the name of the channel type.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_ChannelName(
+    const Tcl_ChannelType *chanTypePtr) /* Pointer to channel type. */
+{
+    return chanTypePtr->typeName;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelVersion --
+ *
+ *	Return the of version of the channel type.
+ *
+ * Results:
+ *	One of the TCL_CHANNEL_VERSION_* constants from tcl.h
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_ChannelTypeVersion
+Tcl_ChannelVersion(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (chanTypePtr->version == TCL_CHANNEL_VERSION_2) {
+	return TCL_CHANNEL_VERSION_2;
+    } else if (chanTypePtr->version == TCL_CHANNEL_VERSION_3) {
+	return TCL_CHANNEL_VERSION_3;
+    } else if (chanTypePtr->version == TCL_CHANNEL_VERSION_4) {
+	return TCL_CHANNEL_VERSION_4;
+    } else if (chanTypePtr->version == TCL_CHANNEL_VERSION_5) {
+	return TCL_CHANNEL_VERSION_5;
+    } else {
+	/*
+	 * In <v2 channel versions, the version field is occupied by the
+	 * Tcl_DriverBlockModeProc
+	 */
+
+	return TCL_CHANNEL_VERSION_1;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * HaveVersion --
+ *
+ *	Return whether a channel type is (at least) of a given version.
+ *
+ * Results:
+ *	True if the minimum version is exceeded by the version actually
+ *	present.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+HaveVersion(
+    const Tcl_ChannelType *chanTypePtr,
+    Tcl_ChannelTypeVersion minimumVersion)
+{
+    Tcl_ChannelTypeVersion actualVersion = Tcl_ChannelVersion(chanTypePtr);
+
+    return (PTR2INT(actualVersion)) >= (PTR2INT(minimumVersion));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelBlockModeProc --
+ *
+ *	Return the Tcl_DriverBlockModeProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------- */
+
+Tcl_DriverBlockModeProc *
+Tcl_ChannelBlockModeProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_2)) {
+	return chanTypePtr->blockModeProc;
+    }
+
+    /*
+     * The v1 structure had the blockModeProc in a different place.
+     */
+
+    return (Tcl_DriverBlockModeProc *) chanTypePtr->version;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelCloseProc --
+ *
+ *	Return the Tcl_DriverCloseProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverCloseProc *
+Tcl_ChannelCloseProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->closeProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelClose2Proc --
+ *
+ *	Return the Tcl_DriverClose2Proc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverClose2Proc *
+Tcl_ChannelClose2Proc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->close2Proc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelInputProc --
+ *
+ *	Return the Tcl_DriverInputProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverInputProc *
+Tcl_ChannelInputProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->inputProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelOutputProc --
+ *
+ *	Return the Tcl_DriverOutputProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverOutputProc *
+Tcl_ChannelOutputProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->outputProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelSeekProc --
+ *
+ *	Return the Tcl_DriverSeekProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverSeekProc *
+Tcl_ChannelSeekProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->seekProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelSetOptionProc --
+ *
+ *	Return the Tcl_DriverSetOptionProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverSetOptionProc *
+Tcl_ChannelSetOptionProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->setOptionProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelGetOptionProc --
+ *
+ *	Return the Tcl_DriverGetOptionProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverGetOptionProc *
+Tcl_ChannelGetOptionProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->getOptionProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelWatchProc --
+ *
+ *	Return the Tcl_DriverWatchProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverWatchProc *
+Tcl_ChannelWatchProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->watchProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelGetHandleProc --
+ *
+ *	Return the Tcl_DriverGetHandleProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverGetHandleProc *
+Tcl_ChannelGetHandleProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    return chanTypePtr->getHandleProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelFlushProc --
+ *
+ *	Return the Tcl_DriverFlushProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverFlushProc *
+Tcl_ChannelFlushProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_2)) {
+	return chanTypePtr->flushProc;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelHandlerProc --
+ *
+ *	Return the Tcl_DriverHandlerProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverHandlerProc *
+Tcl_ChannelHandlerProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_2)) {
+	return chanTypePtr->handlerProc;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelWideSeekProc --
+ *
+ *	Return the Tcl_DriverWideSeekProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverWideSeekProc *
+Tcl_ChannelWideSeekProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_3)) {
+	return chanTypePtr->wideSeekProc;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelThreadActionProc --
+ *
+ *	TIP #218, Channel Thread Actions. Return the
+ *	Tcl_DriverThreadActionProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverThreadActionProc *
+Tcl_ChannelThreadActionProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_4)) {
+	return chanTypePtr->threadActionProc;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetChannelErrorInterp --
+ *
+ *	TIP #219, Tcl Channel Reflection API.
+ *	Store an error message for the I/O system.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Discards a previously stored message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetChannelErrorInterp(
+    Tcl_Interp *interp,		/* Interp to store the data into. */
+    Tcl_Obj *msg)		/* Error message to store. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (iPtr->chanMsg != NULL) {
+	TclDecrRefCount(iPtr->chanMsg);
+	iPtr->chanMsg = NULL;
+    }
+
+    if (msg != NULL) {
+	iPtr->chanMsg = FixLevelCode(msg);
+	Tcl_IncrRefCount(iPtr->chanMsg);
+    }
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetChannelError --
+ *
+ *	TIP #219, Tcl Channel Reflection API.
+ *	Store an error message for the I/O system.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Discards a previously stored message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetChannelError(
+    Tcl_Channel chan,		/* Channel to store the data into. */
+    Tcl_Obj *msg)		/* Error message to store. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+
+    if (statePtr->chanMsg != NULL) {
+	TclDecrRefCount(statePtr->chanMsg);
+	statePtr->chanMsg = NULL;
+    }
+
+    if (msg != NULL) {
+	statePtr->chanMsg = FixLevelCode(msg);
+	Tcl_IncrRefCount(statePtr->chanMsg);
+    }
+    return;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FixLevelCode --
+ *
+ *	TIP #219, Tcl Channel Reflection API.
+ *	Scans an error message for bad -code / -level directives. Returns a
+ *	modified copy with such directives corrected, and the input if it had
+ *	no problems.
+ *
+ * Results:
+ *	A Tcl_Obj*
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+FixLevelCode(
+    Tcl_Obj *msg)
+{
+    int explicitResult, numOptions, lc, lcn;
+    Tcl_Obj **lv, **lvn;
+    int res, i, j, val, lignore, cignore;
+    int newlevel = -1, newcode = -1;
+
+    /* ASSERT msg != NULL */
+
+    /*
+     * Process the caught message.
+     *
+     * Syntax = (option value)... ?message?
+     *
+     * Bad message syntax causes a panic, because the other side uses
+     * Tcl_GetReturnOptions and list construction functions to marshall the
+     * information. Hence an error means that we've got serious breakage.
+     */
+
+    res = Tcl_ListObjGetElements(NULL, msg, &lc, &lv);
+    if (res != TCL_OK) {
+	Tcl_Panic("Tcl_SetChannelError: bad syntax of message");
+    }
+
+    explicitResult = (1 == (lc % 2));
+    numOptions = lc - explicitResult;
+
+    /*
+     * No options, nothing to do.
+     */
+
+    if (numOptions == 0) {
+	return msg;
+    }
+
+    /*
+     * Check for -code x, x != 1|error, and -level x, x != 0
+     */
+
+    for (i = 0; i < numOptions; i += 2) {
+	if (0 == strcmp(TclGetString(lv[i]), "-code")) {
+	    /*
+	     * !"error", !integer, integer != 1 (numeric code for error)
+	     */
+
+	    res = TclGetIntFromObj(NULL, lv[i+1], &val);
+	    if (((res == TCL_OK) && (val != 1)) || ((res != TCL_OK) &&
+		    (0 != strcmp(TclGetString(lv[i+1]), "error")))) {
+		newcode = 1;
+	    }
+	} else if (0 == strcmp(TclGetString(lv[i]), "-level")) {
+	    /*
+	     * !integer, integer != 0
+	     */
+
+	    res = TclGetIntFromObj(NULL, lv [i+1], &val);
+	    if ((res != TCL_OK) || (val != 0)) {
+		newlevel = 0;
+	    }
+	}
+    }
+
+    /*
+     * -code, -level are either not present or ok. Nothing to do.
+     */
+
+    if ((newlevel < 0) && (newcode < 0)) {
+	return msg;
+    }
+
+    lcn = numOptions;
+    if (explicitResult) {
+	lcn ++;
+    }
+    if (newlevel >= 0) {
+	lcn += 2;
+    }
+    if (newcode >= 0) {
+	lcn += 2;
+    }
+
+    lvn = ckalloc(lcn * sizeof(Tcl_Obj *));
+
+    /*
+     * New level/code information is spliced into the first occurence of
+     * -level, -code, further occurences are ignored. The options cannot be
+     * not present, we would not come here. Options which are ok are simply
+     * copied over.
+     */
+
+    lignore = cignore = 0;
+    for (i=0, j=0; i<numOptions; i+=2) {
+	if (0 == strcmp(TclGetString(lv[i]), "-level")) {
+	    if (newlevel >= 0) {
+		lvn[j++] = lv[i];
+		lvn[j++] = Tcl_NewIntObj(newlevel);
+		newlevel = -1;
+		lignore = 1;
+		continue;
+	    } else if (lignore) {
+		continue;
+	    }
+	} else if (0 == strcmp(TclGetString(lv[i]), "-code")) {
+	    if (newcode >= 0) {
+		lvn[j++] = lv[i];
+		lvn[j++] = Tcl_NewIntObj(newcode);
+		newcode = -1;
+		cignore = 1;
+		continue;
+	    } else if (cignore) {
+		continue;
+	    }
+	}
+
+	/*
+	 * Keep everything else, possibly copied down.
+	 */
+
+	lvn[j++] = lv[i];
+	lvn[j++] = lv[i+1];
+    }
+    if (newlevel >= 0) {
+	Tcl_Panic("Defined newlevel not used in rewrite");
+    }
+    if (newcode >= 0) {
+	Tcl_Panic("Defined newcode not used in rewrite");
+    }
+
+    if (explicitResult) {
+	lvn[j++] = lv[i];
+    }
+
+    msg = Tcl_NewListObj(j, lvn);
+
+    ckfree(lvn);
+    return msg;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelErrorInterp --
+ *
+ *	TIP #219, Tcl Channel Reflection API.
+ *	Return the message stored by the channel driver.
+ *
+ * Results:
+ *	Tcl error message object.
+ *
+ * Side effects:
+ *	Resets the stored data to NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetChannelErrorInterp(
+    Tcl_Interp *interp,		/* Interp to query. */
+    Tcl_Obj **msg)		/* Place for error message. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    *msg = iPtr->chanMsg;
+    iPtr->chanMsg = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetChannelError --
+ *
+ *	TIP #219, Tcl Channel Reflection API.
+ *	Return the message stored by the channel driver.
+ *
+ * Results:
+ *	Tcl error message object.
+ *
+ * Side effects:
+ *	Resets the stored data to NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetChannelError(
+    Tcl_Channel chan,		/* Channel to query. */
+    Tcl_Obj **msg)		/* Place for error message. */
+{
+    ChannelState *statePtr = ((Channel *) chan)->state;
+
+    *msg = statePtr->chanMsg;
+    statePtr->chanMsg = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ChannelTruncateProc --
+ *
+ *	TIP #208 (subsection relating to truncation, based on TIP #206).
+ *	Return the Tcl_DriverTruncateProc of the channel type.
+ *
+ * Results:
+ *	A pointer to the proc.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_DriverTruncateProc *
+Tcl_ChannelTruncateProc(
+    const Tcl_ChannelType *chanTypePtr)
+				/* Pointer to channel type. */
+{
+    if (HaveVersion(chanTypePtr, TCL_CHANNEL_VERSION_5)) {
+	return chanTypePtr->truncateProc;
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupChannelIntRep --
+ *
+ *	Initialize the internal representation of a new Tcl_Obj to a copy of
+ *	the internal representation of an existing string object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	copyPtr's internal rep is set to a copy of srcPtr's internal
+ *	representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupChannelIntRep(
+    register Tcl_Obj *srcPtr,	/* Object with internal rep to copy. Must have
+				 * an internal rep of type "Channel". */
+    register Tcl_Obj *copyPtr)	/* Object with internal rep to set. Must not
+				 * currently have an internal rep.*/
+{
+    ResolvedChanName *resPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    resPtr->refCount++;
+    copyPtr->internalRep.twoPtrValue.ptr1 = resPtr;
+    copyPtr->typePtr = srcPtr->typePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeChannelIntRep --
+ *
+ *	Release statePtr storage.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May cause state to be freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeChannelIntRep(
+    Tcl_Obj *objPtr)		/* Object with internal rep to free. */
+{
+    ResolvedChanName *resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    objPtr->typePtr = NULL;
+    if (resPtr->refCount-- > 1) {
+	return;
+    }
+    Tcl_Release(resPtr->statePtr);
+    ckfree(resPtr);
+}
+
+#if 0
+/*
+ * For future debugging work, a simple function to print the flags of a
+ * channel in semi-readable form.
+ */
+
+static int
+DumpFlags(
+    char *str,
+    int flags)
+{
+    char buf[20];
+    int i = 0;
+
+#define ChanFlag(chr, bit)      (buf[i++] = ((flags & (bit)) ? (chr) : '_'))
+
+    ChanFlag('r', TCL_READABLE);
+    ChanFlag('w', TCL_WRITABLE);
+    ChanFlag('n', CHANNEL_NONBLOCKING);
+    ChanFlag('l', CHANNEL_LINEBUFFERED);
+    ChanFlag('u', CHANNEL_UNBUFFERED);
+    ChanFlag('F', BG_FLUSH_SCHEDULED);
+    ChanFlag('c', CHANNEL_CLOSED);
+    ChanFlag('E', CHANNEL_EOF);
+    ChanFlag('S', CHANNEL_STICKY_EOF);
+    ChanFlag('B', CHANNEL_BLOCKED);
+    ChanFlag('/', INPUT_SAW_CR);
+    ChanFlag('D', CHANNEL_DEAD);
+    ChanFlag('R', CHANNEL_RAW_MODE);
+    ChanFlag('x', CHANNEL_INCLOSE);
+
+    buf[i] ='\0';
+
+    fprintf(stderr, "%s: %s\n", str, buf);
+    return 0;
+}
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclIO.h b/generic/tclIO.h
new file mode 100644
index 0000000..07c54fa
--- /dev/null
+++ b/generic/tclIO.h
@@ -0,0 +1,297 @@
+/*
+ * tclIO.h --
+ *
+ *	This file provides the generic portions (those that are the same on
+ *	all platforms and for all channel types) of Tcl's IO facilities.
+ *
+ * Copyright (c) 1998-2000 Ajuba Solutions
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/*
+ * Make sure that both EAGAIN and EWOULDBLOCK are defined. This does not
+ * compile on systems where neither is defined. We want both defined so that
+ * we can test safely for both. In the code we still have to test for both
+ * because there may be systems on which both are defined and have different
+ * values.
+ */
+
+#if ((!defined(EWOULDBLOCK)) && (defined(EAGAIN)))
+#   define EWOULDBLOCK EAGAIN
+#endif
+#if ((!defined(EAGAIN)) && (defined(EWOULDBLOCK)))
+#   define EAGAIN EWOULDBLOCK
+#endif
+#if ((!defined(EAGAIN)) && (!defined(EWOULDBLOCK)))
+#error one of EWOULDBLOCK or EAGAIN must be defined
+#endif
+
+/*
+ * struct ChannelBuffer:
+ *
+ * Buffers data being sent to or from a channel.
+ */
+
+typedef struct ChannelBuffer {
+    int refCount;		/* Current uses count */
+    int nextAdded;		/* The next position into which a character
+				 * will be put in the buffer. */
+    int nextRemoved;		/* Position of next byte to be removed from
+				 * the buffer. */
+    int bufLength;		/* How big is the buffer? */
+    struct ChannelBuffer *nextPtr;
+    				/* Next buffer in chain. */
+    char buf[1];		/* Placeholder for real buffer. The real
+				 * buffer occuppies this space + bufSize-1
+				 * bytes. This must be the last field in the
+				 * structure. */
+} ChannelBuffer;
+
+#define CHANNELBUFFER_HEADER_SIZE	TclOffset(ChannelBuffer, buf)
+
+/*
+ * How much extra space to allocate in buffer to hold bytes from previous
+ * buffer (when converting to UTF-8) or to hold bytes that will go to next
+ * buffer (when converting from UTF-8).
+ */
+
+#define BUFFER_PADDING		16
+
+/*
+ * The following defines the *default* buffer size for channels.
+ */
+
+#define CHANNELBUFFER_DEFAULT_SIZE	(1024 * 4)
+
+/*
+ * The following structure describes the information saved from a call to
+ * "fileevent". This is used later when the event being waited for to invoke
+ * the saved script in the interpreter designed in this record.
+ */
+
+typedef struct EventScriptRecord {
+    struct Channel *chanPtr;	/* The channel for which this script is
+				 * registered. This is used only when an error
+				 * occurs during evaluation of the script, to
+				 * delete the handler. */
+    Tcl_Obj *scriptPtr;		/* Script to invoke. */
+    Tcl_Interp *interp;		/* In what interpreter to invoke script? */
+    int mask;			/* Events must overlap current mask for the
+				 * stored script to be invoked. */
+    struct EventScriptRecord *nextPtr;
+				/* Next in chain of records. */
+} EventScriptRecord;
+
+/*
+ * struct Channel:
+ *
+ * One of these structures is allocated for each open channel. It contains
+ * data specific to the channel but which belongs to the generic part of the
+ * Tcl channel mechanism, and it points at an instance specific (and type
+ * specific) instance data, and at a channel type structure.
+ */
+
+typedef struct Channel {
+    struct ChannelState *state; /* Split out state information */
+    ClientData instanceData;	/* Instance-specific data provided by creator
+				 * of channel. */
+    const Tcl_ChannelType *typePtr; /* Pointer to channel type structure. */
+    struct Channel *downChanPtr;/* Refers to channel this one was stacked
+				 * upon. This reference is NULL for normal
+				 * channels. See Tcl_StackChannel. */
+    struct Channel *upChanPtr;	/* Refers to the channel above stacked this
+				 * one. NULL for the top most channel. */
+
+    /*
+     * Intermediate buffers to hold pre-read data for consumption by a newly
+     * stacked transformation. See 'Tcl_StackChannel'.
+     */
+
+    ChannelBuffer *inQueueHead;	/* Points at first buffer in input queue. */
+    ChannelBuffer *inQueueTail;	/* Points at last buffer in input queue. */
+
+    int refCount;
+} Channel;
+
+/*
+ * struct ChannelState:
+ *
+ * One of these structures is allocated for each open channel. It contains
+ * data specific to the channel but which belongs to the generic part of the
+ * Tcl channel mechanism, and it points at an instance specific (and type
+ * specific) instance data, and at a channel type structure.
+ */
+
+typedef struct ChannelState {
+    char *channelName;		/* The name of the channel instance in Tcl
+				 * commands. Storage is owned by the generic
+				 * IO code, is dynamically allocated. */
+    int	flags;			/* ORed combination of the flags defined
+				 * below. */
+    Tcl_Encoding encoding;	/* Encoding to apply when reading or writing
+				 * data on this channel. NULL means no
+				 * encoding is applied to data. */
+    Tcl_EncodingState inputEncodingState;
+				/* Current encoding state, used when
+				 * converting input data bytes to UTF-8. */
+    int inputEncodingFlags;	/* Encoding flags to pass to conversion
+				 * routine when converting input data bytes to
+				 * UTF-8. May be TCL_ENCODING_START before
+				 * converting first byte and TCL_ENCODING_END
+				 * when EOF is seen. */
+    Tcl_EncodingState outputEncodingState;
+				/* Current encoding state, used when
+				 * converting UTF-8 to output data bytes. */
+    int outputEncodingFlags;	/* Encoding flags to pass to conversion
+				 * routine when converting UTF-8 to output
+				 * data bytes. May be TCL_ENCODING_START
+				 * before converting first byte and
+				 * TCL_ENCODING_END when EOF is seen. */
+    TclEolTranslation inputTranslation;
+				/* What translation to apply for end of line
+				 * sequences on input? */
+    TclEolTranslation outputTranslation;
+				/* What translation to use for generating end
+				 * of line sequences in output? */
+    int inEofChar;		/* If nonzero, use this as a signal of EOF on
+				 * input. */
+    int outEofChar;		/* If nonzero, append this to the channel when
+				 * it is closed if it is open for writing. */
+    int unreportedError;	/* Non-zero if an error report was deferred
+				 * because it happened in the background. The
+				 * value is the POSIX error code. */
+    int refCount;		/* How many interpreters hold references to
+				 * this IO channel? */
+    struct CloseCallback *closeCbPtr;
+				/* Callbacks registered to be called when the
+				 * channel is closed. */
+    char *outputStage;		/* Temporary staging buffer used when
+				 * translating EOL before converting from
+				 * UTF-8 to external form. */
+    ChannelBuffer *curOutPtr;	/* Current output buffer being filled. */
+    ChannelBuffer *outQueueHead;/* Points at first buffer in output queue. */
+    ChannelBuffer *outQueueTail;/* Points at last buffer in output queue. */
+    ChannelBuffer *saveInBufPtr;/* Buffer saved for input queue - eliminates
+				 * need to allocate a new buffer for "gets"
+				 * that crosses buffer boundaries. */
+    ChannelBuffer *inQueueHead;	/* Points at first buffer in input queue. */
+    ChannelBuffer *inQueueTail;	/* Points at last buffer in input queue. */
+    struct ChannelHandler *chPtr;/* List of channel handlers registered for
+				  * this channel. */
+    int interestMask;		/* Mask of all events this channel has
+				 * handlers for. */
+    EventScriptRecord *scriptRecordPtr;
+				/* Chain of all scripts registered for event
+				 * handlers ("fileevent") on this channel. */
+    int bufSize;		/* What size buffers to allocate? */
+    Tcl_TimerToken timer;	/* Handle to wakeup timer for this channel. */
+    struct CopyState *csPtrR;	/* State of background copy for which channel
+				 * is input, or NULL. */
+    struct CopyState *csPtrW;	/* State of background copy for which channel
+				 * is output, or NULL. */
+    Channel *topChanPtr;	/* Refers to topmost channel in a stack. Never
+				 * NULL. */
+    Channel *bottomChanPtr;	/* Refers to bottommost channel in a stack.
+				 * This channel can be relied on to live as
+				 * long as the channel state. Never NULL. */
+    struct ChannelState *nextCSPtr;
+				/* Next in list of channels currently open. */
+    Tcl_ThreadId managingThread;/* TIP #10: Id of the thread managing this
+				 * stack of channels. */
+
+    /*
+     * TIP #219 ... Info for the I/O system ...
+     * Error message set by channel drivers, for the propagation of arbitrary
+     * Tcl errors. This information, if present (chanMsg not NULL), takes
+     * precedence over a posix error code returned by a channel operation.
+     */
+
+    Tcl_Obj* chanMsg;
+    Tcl_Obj* unreportedMsg;     /* Non-NULL if an error report was deferred
+				 * because it happened in the background. The
+				 * value is the chanMg, if any. #219's
+				 * companion to 'unreportedError'. */
+    size_t epoch;		/* Used to test validity of stored channelname
+				 * lookup results. */
+} ChannelState;
+
+/*
+ * Values for the flags field in Channel. Any ORed combination of the
+ * following flags can be stored in the field. These flags record various
+ * options and state bits about the channel. In addition to the flags below,
+ * the channel can also have TCL_READABLE (1<<1) and TCL_WRITABLE (1<<2) set.
+ */
+
+#define CHANNEL_NONBLOCKING	(1<<3)	/* Channel is currently in nonblocking
+					 * mode. */
+#define CHANNEL_LINEBUFFERED	(1<<4)	/* Output to the channel must be
+					 * flushed after every newline. */
+#define CHANNEL_UNBUFFERED	(1<<5)	/* Output to the channel must always
+					 * be flushed immediately. */
+#define BG_FLUSH_SCHEDULED	(1<<7)	/* A background flush of the queued
+					 * output buffers has been
+					 * scheduled. */
+#define CHANNEL_CLOSED		(1<<8)	/* Channel has been closed. No further
+					 * Tcl-level IO on the channel is
+					 * allowed. */
+#define CHANNEL_EOF		(1<<9)	/* EOF occurred on this channel. This
+					 * bit is cleared before every input
+					 * operation. */
+#define CHANNEL_STICKY_EOF	(1<<10)	/* EOF occurred on this channel
+					 * because we saw the input
+					 * eofChar. This bit prevents clearing
+					 * of the EOF bit before every input
+					 * operation. */
+#define CHANNEL_BLOCKED		(1<<11)	/* EWOULDBLOCK or EAGAIN occurred on
+					 * this channel. This bit is cleared
+					 * before every input or output
+					 * operation. */
+#define INPUT_SAW_CR		(1<<12)	/* Channel is in CRLF eol input
+					 * translation mode and the last byte
+					 * seen was a "\r". */
+#define CHANNEL_DEAD		(1<<13)	/* The channel has been closed by the
+					 * exit handler (on exit) but not
+					 * deallocated. When any IO operation
+					 * sees this flag on a channel, it
+					 * does not call driver level
+					 * functions to avoid referring to
+					 * deallocated data. */
+#define CHANNEL_NEED_MORE_DATA	(1<<14)	/* The last input operation failed
+					 * because there was not enough data
+					 * to complete the operation. This
+					 * flag is set when gets fails to get
+					 * a complete line or when read fails
+					 * to get a complete character. When
+					 * set, file events will not be
+					 * delivered for buffered data until
+					 * the state of the channel
+					 * changes. */
+#define CHANNEL_RAW_MODE	(1<<16)	/* When set, notes that the Raw API is
+					 * being used. */
+
+#define CHANNEL_INCLOSE		(1<<19)	/* Channel is currently being closed.
+					 * Its structures are still live and
+					 * usable, but it may not be closed
+					 * again from within the close
+					 * handler. */
+#define CHANNEL_CLOSEDWRITE	(1<<21)	/* Channel write side has been closed.
+					 * No further Tcl-level write IO on
+					 * the channel is allowed. */
+
+/*
+ * The length of time to wait between synthetic timer events. Must be zero or
+ * bad things tend to happen.
+ */
+
+#define SYNTHETIC_EVENT_TIME	0
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIOCmd.c b/generic/tclIOCmd.c
new file mode 100644
index 0000000..6e8bd09
--- /dev/null
+++ b/generic/tclIOCmd.c
@@ -0,0 +1,2096 @@
+/*
+ * tclIOCmd.c --
+ *
+ *	Contains the definitions of most of the Tcl commands relating to IO.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Callback structure for accept callback in a TCP server.
+ */
+
+typedef struct AcceptCallback {
+    Tcl_Obj *script;		/* Script to invoke. */
+    Tcl_Interp *interp;		/* Interpreter in which to run it. */
+} AcceptCallback;
+
+/*
+ * Thread local storage used to maintain a per-thread stdout channel obj.
+ * It must be per-thread because of std channel limitations.
+ */
+
+typedef struct {
+    int initialized;		/* Set to 1 when the module is initialized. */
+    Tcl_Obj *stdoutObjPtr;	/* Cached stdout channel Tcl_Obj */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * Static functions for this file:
+ */
+
+static void		FinalizeIOCmdTSD(ClientData clientData);
+static Tcl_TcpAcceptProc AcceptCallbackProc;
+static int		ChanPendingObjCmd(ClientData unused,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		ChanTruncateObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		RegisterTcpServerInterpCleanup(Tcl_Interp *interp,
+			    AcceptCallback *acceptCallbackPtr);
+static void		TcpAcceptCallbacksDeleteProc(ClientData clientData,
+			    Tcl_Interp *interp);
+static void		TcpServerCloseProc(ClientData callbackData);
+static void		UnregisterTcpServerInterpCleanupProc(
+			    Tcl_Interp *interp,
+			    AcceptCallback *acceptCallbackPtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FinalizeIOCmdTSD --
+ *
+ *	Release the storage associated with the per-thread cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FinalizeIOCmdTSD(
+    ClientData clientData)	/* Not used. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->stdoutObjPtr != NULL) {
+	Tcl_DecrRefCount(tsdPtr->stdoutObjPtr);
+	tsdPtr->stdoutObjPtr = NULL;
+    }
+    tsdPtr->initialized = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PutsObjCmd --
+ *
+ *	This function is invoked to process the "puts" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Produces output on a channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_PutsObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to puts on. */
+    Tcl_Obj *string;		/* String to write. */
+    Tcl_Obj *chanObjPtr = NULL;	/* channel object. */
+    int newline;		/* Add a newline at end? */
+    int result;			/* Result of puts operation. */
+    int mode;			/* Mode in which channel is opened. */
+
+    switch (objc) {
+    case 2:			/* [puts $x] */
+	string = objv[1];
+	newline = 1;
+	break;
+
+    case 3:			/* [puts -nonewline $x] or [puts $chan $x] */
+	if (strcmp(TclGetString(objv[1]), "-nonewline") == 0) {
+	    newline = 0;
+	} else {
+	    newline = 1;
+	    chanObjPtr = objv[1];
+	}
+	string = objv[2];
+	break;
+
+    case 4:			/* [puts -nonewline $chan $x] or
+				 * [puts $chan $x nonewline] */
+	newline = 0;
+	if (strcmp(TclGetString(objv[1]), "-nonewline") == 0) {
+	    chanObjPtr = objv[2];
+	    string = objv[3];
+	    break;
+#if TCL_MAJOR_VERSION < 9
+	} else if (strcmp(TclGetString(objv[3]), "nonewline") == 0) {
+	    /*
+	     * The code below provides backwards compatibility with an old
+	     * form of the command that is no longer recommended or
+	     * documented. See also [Bug #3151675]. Will be removed in Tcl 9,
+	     * maybe even earlier.
+	     */
+
+	    chanObjPtr = objv[1];
+	    string = objv[2];
+	    break;
+#endif
+	}
+	/* Fall through */
+    default:			/* [puts] or
+				 * [puts some bad number of arguments...] */
+	Tcl_WrongNumArgs(interp, 1, objv, "?-nonewline? ?channelId? string");
+	return TCL_ERROR;
+    }
+
+    if (chanObjPtr == NULL) {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+	if (!tsdPtr->initialized) {
+	    tsdPtr->initialized = 1;
+	    TclNewLiteralStringObj(tsdPtr->stdoutObjPtr, "stdout");
+	    Tcl_IncrRefCount(tsdPtr->stdoutObjPtr);
+	    Tcl_CreateThreadExitHandler(FinalizeIOCmdTSD, NULL);
+	}
+	chanObjPtr = tsdPtr->stdoutObjPtr;
+    }
+    if (TclGetChannelFromObj(interp, chanObjPtr, &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_WRITABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for writing",
+		TclGetString(chanObjPtr)));
+	return TCL_ERROR;
+    }
+
+    TclChannelPreserve(chan);
+    result = Tcl_WriteObj(chan, string);
+    if (result < 0) {
+	goto error;
+    }
+    if (newline != 0) {
+	result = Tcl_WriteChars(chan, "\n", 1);
+	if (result < 0) {
+	    goto error;
+	}
+    }
+    TclChannelRelease(chan);
+    return TCL_OK;
+
+    /*
+     * TIP #219.
+     * Capture error messages put by the driver into the bypass area and put
+     * them into the regular interpreter result. Fall back to the regular
+     * message if nothing was found in the bypass.
+     */
+
+  error:
+    if (!TclChanCaughtErrorBypass(interp, chan)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf("error writing \"%s\": %s",
+		TclGetString(chanObjPtr), Tcl_PosixError(interp)));
+    }
+    TclChannelRelease(chan);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FlushObjCmd --
+ *
+ *	This function is called to process the Tcl "flush" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May cause output to appear on the specified channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_FlushObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *chanObjPtr;
+    Tcl_Channel chan;		/* The channel to flush on. */
+    int mode;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId");
+	return TCL_ERROR;
+    }
+    chanObjPtr = objv[1];
+    if (TclGetChannelFromObj(interp, chanObjPtr, &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_WRITABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for writing",
+		TclGetString(chanObjPtr)));
+	return TCL_ERROR;
+    }
+
+    TclChannelPreserve(chan);
+    if (Tcl_Flush(chan) != TCL_OK) {
+	/*
+	 * TIP #219.
+	 * Capture error messages put by the driver into the bypass area and
+	 * put them into the regular interpreter result. Fall back to the
+	 * regular message if nothing was found in the bypass.
+	 */
+
+	if (!TclChanCaughtErrorBypass(interp, chan)) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error flushing \"%s\": %s",
+		    TclGetString(chanObjPtr), Tcl_PosixError(interp)));
+	}
+	TclChannelRelease(chan);
+	return TCL_ERROR;
+    }
+    TclChannelRelease(chan);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetsObjCmd --
+ *
+ *	This function is called to process the Tcl "gets" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May consume input from channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_GetsObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to read from. */
+    int lineLen;		/* Length of line just read. */
+    int mode;			/* Mode in which channel is opened. */
+    Tcl_Obj *linePtr, *chanObjPtr;
+    int code = TCL_OK;
+
+    if ((objc != 2) && (objc != 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId ?varName?");
+	return TCL_ERROR;
+    }
+    chanObjPtr = objv[1];
+    if (TclGetChannelFromObj(interp, chanObjPtr, &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_READABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for reading",
+		TclGetString(chanObjPtr)));
+	return TCL_ERROR;
+    }
+
+    TclChannelPreserve(chan);
+    linePtr = Tcl_NewObj();
+    lineLen = Tcl_GetsObj(chan, linePtr);
+    if (lineLen < 0) {
+	if (!Tcl_Eof(chan) && !Tcl_InputBlocked(chan)) {
+	    Tcl_DecrRefCount(linePtr);
+
+	    /*
+	     * TIP #219.
+	     * Capture error messages put by the driver into the bypass area
+	     * and put them into the regular interpreter result. Fall back to
+	     * the regular message if nothing was found in the bypass.
+	     */
+
+	    if (!TclChanCaughtErrorBypass(interp, chan)) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error reading \"%s\": %s",
+			TclGetString(chanObjPtr), Tcl_PosixError(interp)));
+	    }
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	lineLen = -1;
+    }
+    if (objc == 3) {
+	if (Tcl_ObjSetVar2(interp, objv[2], NULL, linePtr,
+		TCL_LEAVE_ERR_MSG) == NULL) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(lineLen));
+    } else {
+	Tcl_SetObjResult(interp, linePtr);
+    }
+  done:
+    TclChannelRelease(chan);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ReadObjCmd --
+ *
+ *	This function is invoked to process the Tcl "read" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May consume input from channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ReadObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to read from. */
+    int newline, i;		/* Discard newline at end? */
+    int toRead;			/* How many bytes to read? */
+    int charactersRead;		/* How many characters were read? */
+    int mode;			/* Mode in which channel is opened. */
+    Tcl_Obj *resultPtr, *chanObjPtr;
+
+    if ((objc != 2) && (objc != 3)) {
+	Interp *iPtr;
+
+    argerror:
+	iPtr = (Interp *) interp;
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId ?numChars?");
+
+	/*
+	 * Do not append directly; that makes ensembles using this command as
+	 * a subcommand produce the wrong message.
+	 */
+
+	iPtr->flags |= INTERP_ALTERNATE_WRONG_ARGS;
+	Tcl_WrongNumArgs(interp, 1, objv, "?-nonewline? channelId");
+	return TCL_ERROR;
+    }
+
+    i = 1;
+    newline = 0;
+    if (strcmp(TclGetString(objv[1]), "-nonewline") == 0) {
+	newline = 1;
+	i++;
+    }
+
+    if (i == objc) {
+	goto argerror;
+    }
+
+    chanObjPtr = objv[i];
+    if (TclGetChannelFromObj(interp, chanObjPtr, &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_READABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for reading",
+		TclGetString(chanObjPtr)));
+	return TCL_ERROR;
+    }
+    i++;			/* Consumed channel name. */
+
+    /*
+     * Compute how many bytes to read.
+     */
+
+    toRead = -1;
+    if (i < objc) {
+	if ((TclGetIntFromObj(interp, objv[i], &toRead) != TCL_OK)
+		|| (toRead < 0)) {
+#if TCL_MAJOR_VERSION < 9
+	    /*
+	     * The code below provides backwards compatibility with an old
+	     * form of the command that is no longer recommended or
+	     * documented. See also [Bug #3151675]. Will be removed in Tcl 9,
+	     * maybe even earlier.
+	     */
+
+	    if (strcmp(TclGetString(objv[i]), "nonewline") != 0) {
+#endif
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"expected non-negative integer but got \"%s\"",
+			TclGetString(objv[i])));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "NUMBER", NULL);
+		return TCL_ERROR;
+#if TCL_MAJOR_VERSION < 9
+	    }
+	    newline = 1;
+#endif
+	}
+    }
+
+    resultPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(resultPtr);
+    TclChannelPreserve(chan);
+    charactersRead = Tcl_ReadChars(chan, resultPtr, toRead, 0);
+    if (charactersRead < 0) {
+	/*
+	 * TIP #219.
+	 * Capture error messages put by the driver into the bypass area and
+	 * put them into the regular interpreter result. Fall back to the
+	 * regular message if nothing was found in the bypass.
+	 */
+
+	if (!TclChanCaughtErrorBypass(interp, chan)) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error reading \"%s\": %s",
+		    TclGetString(chanObjPtr), Tcl_PosixError(interp)));
+	}
+	TclChannelRelease(chan);
+	Tcl_DecrRefCount(resultPtr);
+	return TCL_ERROR;
+    }
+
+    /*
+     * If requested, remove the last newline in the channel if at EOF.
+     */
+
+    if ((charactersRead > 0) && (newline != 0)) {
+	const char *result;
+	int length;
+
+	result = TclGetStringFromObj(resultPtr, &length);
+	if (result[length - 1] == '\n') {
+	    Tcl_SetObjLength(resultPtr, length - 1);
+	}
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    TclChannelRelease(chan);
+    Tcl_DecrRefCount(resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SeekObjCmd --
+ *
+ *	This function is invoked to process the Tcl "seek" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Moves the position of the access point on the specified channel.  May
+ *	flush queued output.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_SeekObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to tell on. */
+    Tcl_WideInt offset;		/* Where to seek? */
+    int mode;			/* How to seek? */
+    Tcl_WideInt result;		/* Of calling Tcl_Seek. */
+    int optionIndex;
+    static const char *const originOptions[] = {
+	"start", "current", "end", NULL
+    };
+    static const int modeArray[] = {SEEK_SET, SEEK_CUR, SEEK_END};
+
+    if ((objc != 3) && (objc != 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId offset ?origin?");
+	return TCL_ERROR;
+    }
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetWideIntFromObj(interp, objv[2], &offset) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    mode = SEEK_SET;
+    if (objc == 4) {
+	if (Tcl_GetIndexFromObj(interp, objv[3], originOptions, "origin", 0,
+		&optionIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	mode = modeArray[optionIndex];
+    }
+
+    TclChannelPreserve(chan);
+    result = Tcl_Seek(chan, offset, mode);
+    if (result == Tcl_LongAsWide(-1)) {
+	/*
+	 * TIP #219.
+	 * Capture error messages put by the driver into the bypass area and
+	 * put them into the regular interpreter result. Fall back to the
+	 * regular message if nothing was found in the bypass.
+	 */
+
+	if (!TclChanCaughtErrorBypass(interp, chan)) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "error during seek on \"%s\": %s",
+		    TclGetString(objv[1]), Tcl_PosixError(interp)));
+	}
+	TclChannelRelease(chan);
+	return TCL_ERROR;
+    }
+    TclChannelRelease(chan);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TellObjCmd --
+ *
+ *	This function is invoked to process the Tcl "tell" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_TellObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to tell on. */
+    Tcl_WideInt newLoc;
+    int code;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Try to find a channel with the right name and permissions in the IO
+     * channel table of this interpreter.
+     */
+
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    TclChannelPreserve(chan);
+    newLoc = Tcl_Tell(chan);
+
+    /*
+     * TIP #219.
+     * Capture error messages put by the driver into the bypass area and put
+     * them into the regular interpreter result.
+     */
+
+
+    code  = TclChanCaughtErrorBypass(interp, chan);
+    TclChannelRelease(chan);
+    if (code) {
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj(newLoc));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CloseObjCmd --
+ *
+ *	This function is invoked to process the Tcl "close" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May discard queued input; may flush queued output.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_CloseObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;		/* The channel to close. */
+    static const char *const dirOptions[] = {
+	"read", "write", NULL
+    };
+    static const int dirArray[] = {TCL_CLOSE_READ, TCL_CLOSE_WRITE};
+
+    if ((objc != 2) && (objc != 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId ?direction?");
+	return TCL_ERROR;
+    }
+
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	int index, dir;
+
+	/*
+	 * Get direction requested to close, and check syntax.
+	 */
+
+	if (Tcl_GetIndexFromObj(interp, objv[2], dirOptions, "direction", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	dir = dirArray[index];
+
+	/*
+	 * Check direction against channel mode. It is an error if we try to
+	 * close a direction not supported by the channel (already closed, or
+	 * never opened for that direction).
+	 */
+
+	if (!(dir & Tcl_GetChannelMode(chan))) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "Half-close of %s-side not possible, side not opened"
+		    " or already closed", dirOptions[index]));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Special handling is needed if and only if the channel mode supports
+	 * more than the direction to close. Because if the close the last
+	 * direction suppported we can and will go through the regular
+	 * process.
+	 */
+
+	if ((Tcl_GetChannelMode(chan) &
+		(TCL_CLOSE_READ|TCL_CLOSE_WRITE)) != dir) {
+	    return Tcl_CloseEx(interp, chan, dir);
+	}
+    }
+
+    if (Tcl_UnregisterChannel(interp, chan) != TCL_OK) {
+	/*
+	 * If there is an error message and it ends with a newline, remove the
+	 * newline. This is done for command pipeline channels where the error
+	 * output from the subprocesses is stored in interp's result.
+	 *
+	 * NOTE: This is likely to not have any effect on regular error
+	 * messages produced by drivers during the closing of a channel,
+	 * because the Tcl convention is that such error messages do not have
+	 * a terminating newline.
+	 */
+
+	Tcl_Obj *resultPtr = Tcl_GetObjResult(interp);
+	const char *string;
+	int len;
+
+	if (Tcl_IsShared(resultPtr)) {
+	    resultPtr = Tcl_DuplicateObj(resultPtr);
+	    Tcl_SetObjResult(interp, resultPtr);
+	}
+	string = TclGetStringFromObj(resultPtr, &len);
+	if ((len > 0) && (string[len - 1] == '\n')) {
+	    Tcl_SetObjLength(resultPtr, len - 1);
+	}
+	return TCL_ERROR;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FconfigureObjCmd --
+ *
+ *	This function is invoked to process the Tcl "fconfigure" command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May modify the behavior of an IO channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_FconfigureObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *optionName, *valueName;
+    Tcl_Channel chan;		/* The channel to set a mode on. */
+    int i;			/* Iterate over arg-value pairs. */
+
+    if ((objc < 2) || (((objc % 2) == 1) && (objc != 3))) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId ?-option value ...?");
+	return TCL_ERROR;
+    }
+
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	Tcl_DString ds;		/* DString to hold result of calling
+				 * Tcl_GetChannelOption. */
+
+	Tcl_DStringInit(&ds);
+	if (Tcl_GetChannelOption(interp, chan, NULL, &ds) != TCL_OK) {
+	    Tcl_DStringFree(&ds);
+	    return TCL_ERROR;
+	}
+	Tcl_DStringResult(interp, &ds);
+	return TCL_OK;
+    } else if (objc == 3) {
+	Tcl_DString ds;		/* DString to hold result of calling
+				 * Tcl_GetChannelOption. */
+
+	Tcl_DStringInit(&ds);
+	optionName = TclGetString(objv[2]);
+	if (Tcl_GetChannelOption(interp, chan, optionName, &ds) != TCL_OK) {
+	    Tcl_DStringFree(&ds);
+	    return TCL_ERROR;
+	}
+	Tcl_DStringResult(interp, &ds);
+	return TCL_OK;
+    }
+
+    for (i = 3; i < objc; i += 2) {
+	optionName = TclGetString(objv[i-1]);
+	valueName = TclGetString(objv[i]);
+	if (Tcl_SetChannelOption(interp, chan, optionName, valueName)
+		!= TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_EofObjCmd --
+ *
+ *	This function is invoked to process the Tcl "eof" command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Sets interp's result to boolean true or false depending on whether the
+ *	specified channel has an EOF condition.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_EofObjCmd(
+    ClientData unused,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId");
+	return TCL_ERROR;
+    }
+
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(Tcl_Eof(chan)));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ExecObjCmd --
+ *
+ *	This function is invoked to process the "exec" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ExecObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *resultPtr;
+    const char **argv;		/* An array for the string arguments. Stored
+				 * on the _Tcl_ stack. */
+    const char *string;
+    Tcl_Channel chan;
+    int argc, background, i, index, keepNewline, result, skip, length;
+    int ignoreStderr;
+    static const char *const options[] = {
+	"-ignorestderr", "-keepnewline", "--", NULL
+    };
+    enum options {
+	EXEC_IGNORESTDERR, EXEC_KEEPNEWLINE, EXEC_LAST
+    };
+
+    /*
+     * Check for any leading option arguments.
+     */
+
+    keepNewline = 0;
+    ignoreStderr = 0;
+    for (skip = 1; skip < objc; skip++) {
+	string = TclGetString(objv[skip]);
+	if (string[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[skip], options, "option",
+		TCL_EXACT, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (index == EXEC_KEEPNEWLINE) {
+	    keepNewline = 1;
+	} else if (index == EXEC_IGNORESTDERR) {
+	    ignoreStderr = 1;
+	} else {
+	    skip++;
+	    break;
+	}
+    }
+    if (objc <= skip) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-option ...? arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * See if the command is to be run in background.
+     */
+
+    background = 0;
+    string = TclGetString(objv[objc - 1]);
+    if ((string[0] == '&') && (string[1] == '\0')) {
+	objc--;
+	background = 1;
+    }
+
+    /*
+     * Create the string argument array "argv". Make sure argv is large enough
+     * to hold the argc arguments plus 1 extra for the zero end-of-argv word.
+     */
+
+    argc = objc - skip;
+    argv = TclStackAlloc(interp, (unsigned)(argc + 1) * sizeof(char *));
+
+    /*
+     * Copy the string conversions of each (post option) object into the
+     * argument vector.
+     */
+
+    for (i = 0; i < argc; i++) {
+	argv[i] = TclGetString(objv[i + skip]);
+    }
+    argv[argc] = NULL;
+    chan = Tcl_OpenCommandChannel(interp, argc, argv, (background ? 0 :
+	    ignoreStderr ? TCL_STDOUT : TCL_STDOUT|TCL_STDERR));
+
+    /*
+     * Free the argv array.
+     */
+
+    TclStackFree(interp, (void *) argv);
+
+    if (chan == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (background) {
+	/*
+	 * Store the list of PIDs from the pipeline in interp's result and
+	 * detach the PIDs (instead of waiting for them).
+	 */
+
+	TclGetAndDetachPids(interp, chan);
+	if (Tcl_Close(interp, chan) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+
+    resultPtr = Tcl_NewObj();
+    if (Tcl_GetChannelHandle(chan, TCL_READABLE, NULL) == TCL_OK) {
+	if (Tcl_ReadChars(chan, resultPtr, -1, 0) < 0) {
+	    /*
+	     * TIP #219.
+	     * Capture error messages put by the driver into the bypass area
+	     * and put them into the regular interpreter result. Fall back to
+	     * the regular message if nothing was found in the bypass.
+	     */
+
+	    if (!TclChanCaughtErrorBypass(interp, chan)) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error reading output from command: %s",
+			Tcl_PosixError(interp)));
+		Tcl_DecrRefCount(resultPtr);
+	    }
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * If the process produced anything on stderr, it will have been returned
+     * in the interpreter result. It needs to be appended to the result
+     * string.
+     */
+
+    result = Tcl_Close(interp, chan);
+    Tcl_AppendObjToObj(resultPtr, Tcl_GetObjResult(interp));
+
+    /*
+     * If the last character of the result is a newline, then remove the
+     * newline character.
+     */
+
+    if (keepNewline == 0) {
+	string = TclGetStringFromObj(resultPtr, &length);
+	if ((length > 0) && (string[length - 1] == '\n')) {
+	    Tcl_SetObjLength(resultPtr, length - 1);
+	}
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FblockedObjCmd --
+ *
+ *	This function is invoked to process the Tcl "fblocked" command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Sets interp's result to boolean true or false depending on whether the
+ *	preceeding input operation on the channel would have blocked.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_FblockedObjCmd(
+    ClientData unused,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;
+    int mode;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId");
+	return TCL_ERROR;
+    }
+
+    if (TclGetChannelFromObj(interp, objv[1], &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_READABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for reading",
+		TclGetString(objv[1])));
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(Tcl_InputBlocked(chan)));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_OpenObjCmd --
+ *
+ *	This function is invoked to process the "open" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_OpenObjCmd(
+    ClientData notUsed,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int pipeline, prot;
+    const char *modeString, *what;
+    Tcl_Channel chan;
+
+    if ((objc < 2) || (objc > 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "fileName ?access? ?permissions?");
+	return TCL_ERROR;
+    }
+    prot = 0666;
+    if (objc == 2) {
+	modeString = "r";
+    } else {
+	modeString = TclGetString(objv[2]);
+	if (objc == 4) {
+	    const char *permString = TclGetString(objv[3]);
+	    int code = TCL_ERROR;
+	    int scanned = TclParseAllWhiteSpace(permString, -1);
+
+	    /*
+	     * Support legacy octal numbers.
+	     */
+
+	    if ((permString[scanned] == '0')
+		    && (permString[scanned+1] >= '0')
+		    && (permString[scanned+1] <= '7')) {
+		Tcl_Obj *permObj;
+
+		TclNewLiteralStringObj(permObj, "0o");
+		Tcl_AppendToObj(permObj, permString+scanned+1, -1);
+		code = TclGetIntFromObj(NULL, permObj, &prot);
+		Tcl_DecrRefCount(permObj);
+	    }
+
+	    if ((code == TCL_ERROR)
+		    && TclGetIntFromObj(interp, objv[3], &prot) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    pipeline = 0;
+    what = TclGetString(objv[1]);
+    if (what[0] == '|') {
+	pipeline = 1;
+    }
+
+    /*
+     * Open the file or create a process pipeline.
+     */
+
+    if (!pipeline) {
+	chan = Tcl_FSOpenFileChannel(interp, objv[1], modeString, prot);
+    } else {
+	int mode, seekFlag, cmdObjc, binary;
+	const char **cmdArgv;
+
+	if (Tcl_SplitList(interp, what+1, &cmdObjc, &cmdArgv) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	mode = TclGetOpenModeEx(interp, modeString, &seekFlag, &binary);
+	if (mode == -1) {
+	    chan = NULL;
+	} else {
+	    int flags = TCL_STDERR | TCL_ENFORCE_MODE;
+
+	    switch (mode & (O_RDONLY | O_WRONLY | O_RDWR)) {
+	    case O_RDONLY:
+		flags |= TCL_STDOUT;
+		break;
+	    case O_WRONLY:
+		flags |= TCL_STDIN;
+		break;
+	    case O_RDWR:
+		flags |= (TCL_STDIN | TCL_STDOUT);
+		break;
+	    default:
+		Tcl_Panic("Tcl_OpenCmd: invalid mode value");
+		break;
+	    }
+	    chan = Tcl_OpenCommandChannel(interp, cmdObjc, cmdArgv, flags);
+	    if (binary && chan) {
+		Tcl_SetChannelOption(interp, chan, "-translation", "binary");
+	    }
+	}
+	ckfree(cmdArgv);
+    }
+    if (chan == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_RegisterChannel(interp, chan);
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(Tcl_GetChannelName(chan), -1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TcpAcceptCallbacksDeleteProc --
+ *
+ *	Assocdata cleanup routine called when an interpreter is being deleted
+ *	to set the interp field of all the accept callback records registered
+ *	with the interpreter to NULL. This will prevent the interpreter from
+ *	being used in the future to eval accept scripts.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deallocates memory and sets the interp field of all the accept
+ *	callback records to NULL to prevent this interpreter from being used
+ *	subsequently to eval accept scripts.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static void
+TcpAcceptCallbacksDeleteProc(
+    ClientData clientData,	/* Data which was passed when the assocdata
+				 * was registered. */
+    Tcl_Interp *interp)		/* Interpreter being deleted - not used. */
+{
+    Tcl_HashTable *hTblPtr = clientData;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch hSearch;
+
+    for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch);
+	    hPtr != NULL; hPtr = Tcl_NextHashEntry(&hSearch)) {
+	AcceptCallback *acceptCallbackPtr = Tcl_GetHashValue(hPtr);
+
+	acceptCallbackPtr->interp = NULL;
+    }
+    Tcl_DeleteHashTable(hTblPtr);
+    ckfree(hTblPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RegisterTcpServerInterpCleanup --
+ *
+ *	Registers an accept callback record to have its interp field set to
+ *	NULL when the interpreter is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	When, in the future, the interpreter is deleted, the interp field of
+ *	the accept callback data structure will be set to NULL. This will
+ *	prevent attempts to eval the accept script in a deleted interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RegisterTcpServerInterpCleanup(
+    Tcl_Interp *interp,		/* Interpreter for which we want to be
+				 * informed of deletion. */
+    AcceptCallback *acceptCallbackPtr)
+				/* The accept callback record whose interp
+				 * field we want set to NULL when the
+				 * interpreter is deleted. */
+{
+    Tcl_HashTable *hTblPtr;	/* Hash table for accept callback records to
+				 * smash when the interpreter will be
+				 * deleted. */
+    Tcl_HashEntry *hPtr;	/* Entry for this record. */
+    int isNew;			/* Is the entry new? */
+
+    hTblPtr = Tcl_GetAssocData(interp, "tclTCPAcceptCallbacks", NULL);
+
+    if (hTblPtr == NULL) {
+	hTblPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(hTblPtr, TCL_ONE_WORD_KEYS);
+	Tcl_SetAssocData(interp, "tclTCPAcceptCallbacks",
+		TcpAcceptCallbacksDeleteProc, hTblPtr);
+    }
+
+    hPtr = Tcl_CreateHashEntry(hTblPtr, acceptCallbackPtr, &isNew);
+    if (!isNew) {
+	Tcl_Panic("RegisterTcpServerCleanup: damaged accept record table");
+    }
+    Tcl_SetHashValue(hPtr, acceptCallbackPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UnregisterTcpServerInterpCleanupProc --
+ *
+ *	Unregister a previously registered accept callback record. The interp
+ *	field of this record will no longer be set to NULL in the future when
+ *	the interpreter is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Prevents the interp field of the accept callback record from being set
+ *	to NULL in the future when the interpreter is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UnregisterTcpServerInterpCleanupProc(
+    Tcl_Interp *interp,		/* Interpreter in which the accept callback
+				 * record was registered. */
+    AcceptCallback *acceptCallbackPtr)
+				/* The record for which to delete the
+				 * registration. */
+{
+    Tcl_HashTable *hTblPtr;
+    Tcl_HashEntry *hPtr;
+
+    hTblPtr = Tcl_GetAssocData(interp, "tclTCPAcceptCallbacks", NULL);
+    if (hTblPtr == NULL) {
+	return;
+    }
+
+    hPtr = Tcl_FindHashEntry(hTblPtr, (char *) acceptCallbackPtr);
+    if (hPtr != NULL) {
+	Tcl_DeleteHashEntry(hPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AcceptCallbackProc --
+ *
+ *	This callback is invoked by the TCP channel driver when it accepts a
+ *	new connection from a client on a server socket.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Whatever the script does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AcceptCallbackProc(
+    ClientData callbackData,	/* The data stored when the callback was
+				 * created in the call to
+				 * Tcl_OpenTcpServer. */
+    Tcl_Channel chan,		/* Channel for the newly accepted
+				 * connection. */
+    char *address,		/* Address of client that was accepted. */
+    int port)			/* Port of client that was accepted. */
+{
+    AcceptCallback *acceptCallbackPtr = callbackData;
+
+    /*
+     * Check if the callback is still valid; the interpreter may have gone
+     * away, this is signalled by setting the interp field of the callback
+     * data to NULL.
+     */
+
+    if (acceptCallbackPtr->interp != NULL) {
+	Tcl_Interp *interp = acceptCallbackPtr->interp;
+	Tcl_Obj *script, *objv[2];
+	int result = TCL_OK;
+
+	objv[0] = acceptCallbackPtr->script;
+	objv[1] = Tcl_NewListObj(3, NULL);
+	Tcl_ListObjAppendElement(NULL, objv[1], Tcl_NewStringObj(
+		Tcl_GetChannelName(chan), -1));
+	Tcl_ListObjAppendElement(NULL, objv[1], Tcl_NewStringObj(address, -1));
+	Tcl_ListObjAppendElement(NULL, objv[1], Tcl_NewIntObj(port));
+
+	script = Tcl_ConcatObj(2, objv);
+	Tcl_IncrRefCount(script);
+	Tcl_DecrRefCount(objv[1]);
+
+	Tcl_Preserve(interp);
+	Tcl_RegisterChannel(interp, chan);
+
+	/*
+	 * Artificially bump the refcount to protect the channel from being
+	 * deleted while the script is being evaluated.
+	 */
+
+	Tcl_RegisterChannel(NULL, chan);
+
+	result = Tcl_EvalObjEx(interp, script, TCL_EVAL_DIRECT|TCL_EVAL_GLOBAL);
+	Tcl_DecrRefCount(script);
+
+	if (result != TCL_OK) {
+	    Tcl_BackgroundException(interp, result);
+	    Tcl_UnregisterChannel(interp, chan);
+	}
+
+	/*
+	 * Decrement the artificially bumped refcount. After this it is not
+	 * safe anymore to use "chan", because it may now be deleted.
+	 */
+
+	Tcl_UnregisterChannel(NULL, chan);
+
+	Tcl_Release(interp);
+    } else {
+	/*
+	 * The interpreter has been deleted, so there is no useful way to use
+	 * the client socket - just close it.
+	 */
+
+	Tcl_Close(NULL, chan);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TcpServerCloseProc --
+ *
+ *	This callback is called when the TCP server channel for which it was
+ *	registered is being closed. It informs the interpreter in which the
+ *	accept script is evaluated (if that interpreter still exists) that
+ *	this channel no longer needs to be informed if the interpreter is
+ *	deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	In the future, if the interpreter is deleted this channel will no
+ *	longer be informed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TcpServerCloseProc(
+    ClientData callbackData)	/* The data passed in the call to
+				 * Tcl_CreateCloseHandler. */
+{
+    AcceptCallback *acceptCallbackPtr = callbackData;
+				/* The actual data. */
+
+    if (acceptCallbackPtr->interp != NULL) {
+	UnregisterTcpServerInterpCleanupProc(acceptCallbackPtr->interp,
+		acceptCallbackPtr);
+    }
+    Tcl_DecrRefCount(acceptCallbackPtr->script);
+    ckfree(acceptCallbackPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SocketObjCmd --
+ *
+ *	This function is invoked to process the "socket" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a socket based channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SocketObjCmd(
+    ClientData notUsed,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const socketOptions[] = {
+	"-async", "-myaddr", "-myport", "-reuseaddr", "-reuseport", "-server",
+	NULL
+    };
+    enum socketOptions {
+	SKT_ASYNC, SKT_MYADDR, SKT_MYPORT, SKT_REUSEADDR, SKT_REUSEPORT,
+	SKT_SERVER
+    };
+    int optionIndex, a, server = 0, myport = 0, async = 0, reusep = -1,
+	reusea = -1;
+    unsigned int flags = 0;
+    const char *host, *port, *myaddr = NULL;
+    Tcl_Obj *script = NULL;
+    Tcl_Channel chan;
+
+    if (TclpHasSockets(interp) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    for (a = 1; a < objc; a++) {
+	const char *arg = Tcl_GetString(objv[a]);
+
+	if (arg[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[a], socketOptions, "option",
+		TCL_EXACT, &optionIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum socketOptions) optionIndex) {
+	case SKT_ASYNC:
+	    if (server == 1) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"cannot set -async option for server sockets", -1));
+		return TCL_ERROR;
+	    }
+	    async = 1;
+	    break;
+	case SKT_MYADDR:
+	    a++;
+	    if (a >= objc) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"no argument given for -myaddr option", -1));
+		return TCL_ERROR;
+	    }
+	    myaddr = TclGetString(objv[a]);
+	    break;
+	case SKT_MYPORT: {
+	    const char *myPortName;
+
+	    a++;
+	    if (a >= objc) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"no argument given for -myport option", -1));
+		return TCL_ERROR;
+	    }
+	    myPortName = TclGetString(objv[a]);
+	    if (TclSockGetPort(interp, myPortName, "tcp", &myport) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    break;
+	}
+	case SKT_SERVER:
+	    if (async == 1) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"cannot set -async option for server sockets", -1));
+		return TCL_ERROR;
+	    }
+	    server = 1;
+	    a++;
+	    if (a >= objc) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"no argument given for -server option", -1));
+		return TCL_ERROR;
+	    }
+	    script = objv[a];
+	    break;
+	case SKT_REUSEADDR:
+	    a++;
+	    if (a >= objc) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"no argument given for -reuseaddr option", -1));
+		return TCL_ERROR;
+	    }
+	    if (Tcl_GetBooleanFromObj(interp, objv[a], &reusea) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    break;
+	case SKT_REUSEPORT:
+	    a++;
+	    if (a >= objc) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"no argument given for -reuseport option", -1));
+		return TCL_ERROR;
+	    }
+	    if (Tcl_GetBooleanFromObj(interp, objv[a], &reusep) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    break;
+	default:
+	    Tcl_Panic("Tcl_SocketObjCmd: bad option index to SocketOptions");
+	}
+    }
+    if (server) {
+	host = myaddr;		/* NULL implies INADDR_ANY */
+	if (myport != 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "option -myport is not valid for servers", -1));
+	    return TCL_ERROR;
+	}
+    } else if (a < objc) {
+	host = TclGetString(objv[a]);
+	a++;
+    } else {
+	Interp *iPtr;
+
+    wrongNumArgs:
+	iPtr = (Interp *) interp;
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-myaddr addr? ?-myport myport? ?-async? host port");
+	iPtr->flags |= INTERP_ALTERNATE_WRONG_ARGS;
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"-server command ?-reuseaddr boolean? ?-reuseport boolean? "
+		"?-myaddr addr? port");
+	return TCL_ERROR;
+    }
+
+    if (!server && (reusea != -1 || reusep != -1)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"options -reuseaddr and -reuseport are only valid for servers",
+		-1));
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set the options to their default value if the user didn't override
+     * their value.
+     */
+
+    if (reusep == -1) {
+	reusep = 0;
+    }
+    if (reusea == -1) {
+	reusea = 1;
+    }
+
+    /*
+     * Build the bitset with the flags values.
+     */
+
+    if (reusea) {
+	flags |= TCL_TCPSERVER_REUSEADDR;
+    }
+    if (reusep) {
+	flags |= TCL_TCPSERVER_REUSEPORT;
+    }
+
+    /*
+     * All the arguments should have been parsed by now, 'a' points to the
+     * last one, the port number.
+     */
+
+    if (a != objc-1) {
+	goto wrongNumArgs;
+    }
+
+    port = TclGetString(objv[a]);
+
+    if (server) {
+	AcceptCallback *acceptCallbackPtr = ckalloc(sizeof(AcceptCallback));
+
+	Tcl_IncrRefCount(script);
+	acceptCallbackPtr->script = script;
+	acceptCallbackPtr->interp = interp;
+
+	chan = Tcl_OpenTcpServerEx(interp, port, host, flags,
+		AcceptCallbackProc, acceptCallbackPtr);
+	if (chan == NULL) {
+	    Tcl_DecrRefCount(script);
+	    ckfree(acceptCallbackPtr);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Register with the interpreter to let us know when the interpreter
+	 * is deleted (by having the callback set the interp field of the
+	 * acceptCallbackPtr's structure to NULL). This is to avoid trying to
+	 * eval the script in a deleted interpreter.
+	 */
+
+	RegisterTcpServerInterpCleanup(interp, acceptCallbackPtr);
+
+	/*
+	 * Register a close callback. This callback will inform the
+	 * interpreter (if it still exists) that this channel does not need to
+	 * be informed when the interpreter is deleted.
+	 */
+
+	Tcl_CreateCloseHandler(chan, TcpServerCloseProc, acceptCallbackPtr);
+    } else {
+	int portNum;
+
+	if (TclSockGetPort(interp, port, "tcp", &portNum) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	chan = Tcl_OpenTcpClient(interp, portNum, host, myaddr, myport, async);
+	if (chan == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    Tcl_RegisterChannel(interp, chan);
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(Tcl_GetChannelName(chan), -1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FcopyObjCmd --
+ *
+ *	This function is invoked to process the "fcopy" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Moves data between two channels and possibly sets up a background copy
+ *	handler.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FcopyObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel inChan, outChan;
+    int mode, i, index;
+    Tcl_WideInt toRead;
+    Tcl_Obj *cmdPtr;
+    static const char *const switches[] = { "-size", "-command", NULL };
+    enum { FcopySize, FcopyCommand };
+
+    if ((objc < 3) || (objc > 7) || (objc == 4) || (objc == 6)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"input output ?-size size? ?-command callback?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the channel arguments and verify that they are readable or
+     * writable, as appropriate.
+     */
+
+    if (TclGetChannelFromObj(interp, objv[1], &inChan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_READABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for reading",
+		TclGetString(objv[1])));
+	return TCL_ERROR;
+    }
+    if (TclGetChannelFromObj(interp, objv[2], &outChan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (!(mode & TCL_WRITABLE)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"channel \"%s\" wasn't opened for writing",
+		TclGetString(objv[2])));
+	return TCL_ERROR;
+    }
+
+    toRead = -1;
+    cmdPtr = NULL;
+    for (i = 3; i < objc; i += 2) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], switches, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (index) {
+	case FcopySize:
+	    if (Tcl_GetWideIntFromObj(interp, objv[i+1], &toRead) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (toRead < 0) {
+		/*
+		 * Handle all negative sizes like -1, meaning 'copy all'. By
+		 * resetting toRead we avoid changes in the core copying
+		 * functions (which explicitly check for -1 and crash on any
+		 * other negative value).
+		 */
+
+		toRead = -1;
+	    }
+	    break;
+	case FcopyCommand:
+	    cmdPtr = objv[i+1];
+	    break;
+	}
+    }
+
+    return TclCopyChannel(interp, inChan, outChan, toRead, cmdPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * ChanPendingObjCmd --
+ *
+ *	This function is invoked to process the Tcl "chan pending" command
+ *	(TIP #287). See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Sets interp's result to the number of bytes of buffered input or
+ *	output (depending on whether the first argument is "input" or
+ *	"output"), or -1 if the channel wasn't opened for that mode.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ChanPendingObjCmd(
+    ClientData unused,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;
+    int index, mode;
+    static const char *const options[] = {"input", "output", NULL};
+    enum options {PENDING_INPUT, PENDING_OUTPUT};
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "mode channelId");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], options, "mode", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (TclGetChannelFromObj(interp, objv[2], &chan, &mode, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum options) index) {
+    case PENDING_INPUT:
+	if (!(mode & TCL_READABLE)) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(-1));
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(Tcl_InputBuffered(chan)));
+	}
+	break;
+    case PENDING_OUTPUT:
+	if (!(mode & TCL_WRITABLE)) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(-1));
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(Tcl_OutputBuffered(chan)));
+	}
+	break;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ChanTruncateObjCmd --
+ *
+ *	This function is invoked to process the "chan truncate" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Truncates a channel (or rather a file underlying a channel).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ChanTruncateObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel chan;
+    Tcl_WideInt length;
+
+    if ((objc < 2) || (objc > 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channelId ?length?");
+	return TCL_ERROR;
+    }
+    if (TclGetChannelFromObj(interp, objv[1], &chan, NULL, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	/*
+	 * User is supplying an explicit length.
+	 */
+
+	if (Tcl_GetWideIntFromObj(interp, objv[2], &length) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (length < 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "cannot truncate to negative length of file", -1));
+	    return TCL_ERROR;
+	}
+    } else {
+	/*
+	 * User wants to truncate to the current file position.
+	 */
+
+	length = Tcl_Tell(chan);
+	if (length == Tcl_WideAsLong(-1)) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "could not determine current location in \"%s\": %s",
+		    TclGetString(objv[1]), Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+    }
+
+    if (Tcl_TruncateChannel(chan, length) != TCL_OK) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"error during truncate on \"%s\": %s",
+		TclGetString(objv[1]), Tcl_PosixError(interp)));
+	return TCL_ERROR;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ChanPipeObjCmd --
+ *
+ *	This function is invoked to process the "chan pipe" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a pair of Tcl channels wrapping both ends of a new
+ *	anonymous pipe.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ChanPipeObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Channel rchan, wchan;
+    const char *channelNames[2];
+    Tcl_Obj *resultPtr;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_CreatePipe(interp, &rchan, &wchan, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    channelNames[0] = Tcl_GetChannelName(rchan);
+    channelNames[1] = Tcl_GetChannelName(wchan);
+
+    resultPtr = Tcl_NewObj();
+    Tcl_ListObjAppendElement(NULL, resultPtr,
+	    Tcl_NewStringObj(channelNames[0], -1));
+    Tcl_ListObjAppendElement(NULL, resultPtr,
+	    Tcl_NewStringObj(channelNames[1], -1));
+    Tcl_SetObjResult(interp, resultPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChannelNamesCmd --
+ *
+ *	This function is invoked to process the "chan names" and "file
+ *	channels" Tcl commands.  See the user documentation for details on
+ *	what they do.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclChannelNamesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc < 1 || objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+    return Tcl_GetChannelNamesEx(interp,
+	    ((objc == 1) ? NULL : TclGetString(objv[1])));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitChanCmd --
+ *
+ *	This function is invoked to create the "chan" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A Tcl command handle.
+ *
+ * Side effects:
+ *	None (since nothing is byte-compiled).
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitChanCmd(
+    Tcl_Interp *interp)
+{
+    /*
+     * Most commands are plugged directly together, but some are done via
+     * alias-like rewriting; [chan configure] is this way for security reasons
+     * (want overwriting of [fconfigure] to control that nicely), and [chan
+     * names] because the functionality isn't available as a separate command
+     * function at the moment.
+     */
+    static const EnsembleImplMap initMap[] = {
+	{"blocked",	Tcl_FblockedObjCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"close",	Tcl_CloseObjCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"copy",	Tcl_FcopyObjCmd,	NULL, NULL, NULL, 0},
+	{"create",	TclChanCreateObjCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},		/* TIP #219 */
+	{"eof",		Tcl_EofObjCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"event",	Tcl_FileEventObjCmd,	TclCompileBasic2Or3ArgCmd, NULL, NULL, 0},
+	{"flush",	Tcl_FlushObjCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"gets",	Tcl_GetsObjCmd,		TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"names",	TclChannelNamesCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+	{"pending",	ChanPendingObjCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},		/* TIP #287 */
+	{"pipe",	ChanPipeObjCmd,		TclCompileBasic0ArgCmd, NULL, NULL, 0},		/* TIP #304 */
+	{"pop",		TclChanPopObjCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},		/* TIP #230 */
+	{"postevent",	TclChanPostEventObjCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},	/* TIP #219 */
+	{"push",	TclChanPushObjCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},		/* TIP #230 */
+	{"puts",	Tcl_PutsObjCmd,		NULL, NULL, NULL, 0},
+	{"read",	Tcl_ReadObjCmd,		NULL, NULL, NULL, 0},
+	{"seek",	Tcl_SeekObjCmd,		TclCompileBasic2Or3ArgCmd, NULL, NULL, 0},
+	{"tell",	Tcl_TellObjCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"truncate",	ChanTruncateObjCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},		/* TIP #208 */
+	{NULL, NULL, NULL, NULL, NULL, 0}
+    };
+    static const char *const extras[] = {
+	"configure",	"::fconfigure",
+	NULL
+    };
+    Tcl_Command ensemble;
+    Tcl_Obj *mapObj;
+    int i;
+
+    ensemble = TclMakeEnsemble(interp, "chan", initMap);
+    Tcl_GetEnsembleMappingDict(NULL, ensemble, &mapObj);
+    for (i=0 ; extras[i] ; i+=2) {
+	/*
+	 * Can assume that reference counts are all incremented.
+	 */
+
+	Tcl_DictObjPut(NULL, mapObj, Tcl_NewStringObj(extras[i], -1),
+		Tcl_NewStringObj(extras[i+1], -1));
+    }
+    Tcl_SetEnsembleMappingDict(interp, ensemble, mapObj);
+    return ensemble;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIOGT.c b/generic/tclIOGT.c
new file mode 100644
index 0000000..c1e8c44
--- /dev/null
+++ b/generic/tclIOGT.c
@@ -0,0 +1,1441 @@
+/*
+ * tclIOGT.c --
+ *
+ *	Implements a generic transformation exposing the underlying API at the
+ *	script level. Contributed by Andreas Kupries.
+ *
+ * Copyright (c) 2000 Ajuba Solutions
+ * Copyright (c) 1999-2000 Andreas Kupries (a.kupries@westend.com)
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclIO.h"
+
+/*
+ * Forward declarations of internal procedures. First the driver procedures of
+ * the transformation.
+ */
+
+static int		TransformBlockModeProc(ClientData instanceData,
+			    int mode);
+static int		TransformCloseProc(ClientData instanceData,
+			    Tcl_Interp *interp);
+static int		TransformInputProc(ClientData instanceData, char *buf,
+			    int toRead, int *errorCodePtr);
+static int		TransformOutputProc(ClientData instanceData,
+			    const char *buf, int toWrite, int *errorCodePtr);
+static int		TransformSeekProc(ClientData instanceData, long offset,
+			    int mode, int *errorCodePtr);
+static int		TransformSetOptionProc(ClientData instanceData,
+			    Tcl_Interp *interp, const char *optionName,
+			    const char *value);
+static int		TransformGetOptionProc(ClientData instanceData,
+			    Tcl_Interp *interp, const char *optionName,
+			    Tcl_DString *dsPtr);
+static void		TransformWatchProc(ClientData instanceData, int mask);
+static int		TransformGetFileHandleProc(ClientData instanceData,
+			    int direction, ClientData *handlePtr);
+static int		TransformNotifyProc(ClientData instanceData, int mask);
+static Tcl_WideInt	TransformWideSeekProc(ClientData instanceData,
+			    Tcl_WideInt offset, int mode, int *errorCodePtr);
+
+/*
+ * Forward declarations of internal procedures. Secondly the procedures for
+ * handling and generating fileeevents.
+ */
+
+static void		TransformChannelHandlerTimer(ClientData clientData);
+
+/*
+ * Forward declarations of internal procedures. Third, helper procedures
+ * encapsulating essential tasks.
+ */
+
+typedef struct TransformChannelData TransformChannelData;
+
+static int		ExecuteCallback(TransformChannelData *ctrl,
+			    Tcl_Interp *interp, unsigned char *op,
+			    unsigned char *buf, int bufLen, int transmit,
+			    int preserve);
+
+/*
+ * Action codes to give to 'ExecuteCallback' (argument 'transmit'), telling
+ * the procedure what to do with the result of the script it calls.
+ */
+
+#define TRANSMIT_DONT	0	/* No transfer to do. */
+#define TRANSMIT_DOWN	1	/* Transfer to the underlying channel. */
+#define TRANSMIT_SELF	2	/* Transfer into our channel. */
+#define TRANSMIT_IBUF	3	/* Transfer to internal input buffer. */
+#define TRANSMIT_NUM	4	/* Transfer number to 'maxRead'. */
+
+/*
+ * Codes for 'preserve' of 'ExecuteCallback'.
+ */
+
+#define P_PRESERVE	1
+#define P_NO_PRESERVE	0
+
+/*
+ * Strings for the action codes delivered to the script implementing a
+ * transformation. Argument 'op' of 'ExecuteCallback'.
+ */
+
+#define A_CREATE_WRITE	(UCHARP("create/write"))
+#define A_DELETE_WRITE	(UCHARP("delete/write"))
+#define A_FLUSH_WRITE	(UCHARP("flush/write"))
+#define A_WRITE		(UCHARP("write"))
+
+#define A_CREATE_READ	(UCHARP("create/read"))
+#define A_DELETE_READ	(UCHARP("delete/read"))
+#define A_FLUSH_READ	(UCHARP("flush/read"))
+#define A_READ		(UCHARP("read"))
+
+#define A_QUERY_MAXREAD	(UCHARP("query/maxRead"))
+#define A_CLEAR_READ	(UCHARP("clear/read"))
+
+/*
+ * Management of a simple buffer.
+ */
+
+typedef struct ResultBuffer ResultBuffer;
+
+static inline void	ResultClear(ResultBuffer *r);
+static inline void	ResultInit(ResultBuffer *r);
+static inline int	ResultEmpty(ResultBuffer *r);
+static inline int	ResultCopy(ResultBuffer *r, unsigned char *buf,
+			    size_t toRead);
+static inline void	ResultAdd(ResultBuffer *r, unsigned char *buf,
+			    size_t toWrite);
+
+/*
+ * This structure describes the channel type structure for Tcl-based
+ * transformations.
+ */
+
+static const Tcl_ChannelType transformChannelType = {
+    "transform",		/* Type name. */
+    TCL_CHANNEL_VERSION_5,	/* v5 channel */
+    TransformCloseProc,		/* Close proc. */
+    TransformInputProc,		/* Input proc. */
+    TransformOutputProc,	/* Output proc. */
+    TransformSeekProc,		/* Seek proc. */
+    TransformSetOptionProc,	/* Set option proc. */
+    TransformGetOptionProc,	/* Get option proc. */
+    TransformWatchProc,		/* Initialize notifier. */
+    TransformGetFileHandleProc,	/* Get OS handles out of channel. */
+    NULL,			/* close2proc */
+    TransformBlockModeProc,	/* Set blocking/nonblocking mode.*/
+    NULL,			/* Flush proc. */
+    TransformNotifyProc,	/* Handling of events bubbling up. */
+    TransformWideSeekProc,	/* Wide seek proc. */
+    NULL,			/* Thread action. */
+    NULL			/* Truncate. */
+};
+
+/*
+ * Possible values for 'flags' field in control structure, see below.
+ */
+
+#define CHANNEL_ASYNC (1<<0)	/* Non-blocking mode. */
+
+/*
+ * Definition of the structure containing the information about the internal
+ * input buffer.
+ */
+
+struct ResultBuffer {
+    unsigned char *buf;		/* Reference to the buffer area. */
+    size_t allocated;		/* Allocated size of the buffer area. */
+    size_t used;		/* Number of bytes in the buffer, no more than
+				 * number allocated. */
+};
+
+/*
+ * Additional bytes to allocate during buffer expansion.
+ */
+
+#define INCREMENT	512
+
+/*
+ * Number of milliseconds to wait before firing an event to flush out
+ * information waiting in buffers (fileevent support).
+ */
+
+#define FLUSH_DELAY	5
+
+/*
+ * Convenience macro to make some casts easier to use.
+ */
+
+#define UCHARP(x)	((unsigned char *) (x))
+
+/*
+ * Definition of a structure used by all transformations generated here to
+ * maintain their local state.
+ */
+
+struct TransformChannelData {
+    /*
+     * General section. Data to integrate the transformation into the channel
+     * system.
+     */
+
+    Tcl_Channel self;		/* Our own Channel handle. */
+    int readIsFlushed;		/* Flag to note whether in.flushProc was
+				 * called or not. */
+    int eofPending;		/* Flag: EOF seen down, not raised up */
+    int flags;			/* Currently CHANNEL_ASYNC or zero. */
+    int watchMask;		/* Current watch/event/interest mask. */
+    int mode;			/* Mode of parent channel, OR'ed combination
+				 * of TCL_READABLE, TCL_WRITABLE. */
+    Tcl_TimerToken timer;	/* Timer for automatic flushing of information
+				 * sitting in an internal buffer. Required for
+				 * full fileevent support. */
+
+    /*
+     * Transformation specific data.
+     */
+
+    int maxRead;		/* Maximum allowed number of bytes to read, as
+				 * given to us by the Tcl script implementing
+				 * the transformation. */
+    Tcl_Interp *interp;		/* Reference to the interpreter which created
+				 * the transformation. Used to execute the
+				 * code below. */
+    Tcl_Obj *command;		/* Tcl code to execute for a buffer */
+    ResultBuffer result;	/* Internal buffer used to store the result of
+				 * a transformation of incoming data. Also
+				 * serves as buffer of all data not yet
+				 * consumed by the reader. */
+    size_t refCount;
+};
+
+static void
+PreserveData(
+    TransformChannelData *dataPtr)
+{
+    dataPtr->refCount++;
+}
+
+static void
+ReleaseData(
+    TransformChannelData *dataPtr)
+{
+    if (dataPtr->refCount-- > 1) {
+	return;
+    }
+    ResultClear(&dataPtr->result);
+    Tcl_DecrRefCount(dataPtr->command);
+    ckfree(dataPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChannelTransform --
+ *
+ *	Implements the Tcl "testchannel transform" debugging command. This is
+ *	part of the testing environment. This sets up a tcl script (cmdObjPtr)
+ *	to be used as a transform on the channel.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+TclChannelTransform(
+    Tcl_Interp *interp,		/* Interpreter for result. */
+    Tcl_Channel chan,		/* Channel to transform. */
+    Tcl_Obj *cmdObjPtr)		/* Script to use for transform. */
+{
+    Channel *chanPtr;		/* The actual channel. */
+    ChannelState *statePtr;	/* State info for channel. */
+    int mode;			/* Read/write mode of the channel. */
+    int objc;
+    TransformChannelData *dataPtr;
+    Tcl_DString ds;
+
+    if (chan == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (TCL_OK != Tcl_ListObjLength(interp, cmdObjPtr, &objc)) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("-command value is not a list", -1));
+	return TCL_ERROR;
+    }
+
+    chanPtr = (Channel *) chan;
+    statePtr = chanPtr->state;
+    chanPtr = statePtr->topChanPtr;
+    chan = (Tcl_Channel) chanPtr;
+    mode = (statePtr->flags & (TCL_READABLE|TCL_WRITABLE));
+
+    /*
+     * Now initialize the transformation state and stack it upon the specified
+     * channel. One of the necessary things to do is to retrieve the blocking
+     * regime of the underlying channel and to use the same for us too.
+     */
+
+    dataPtr = ckalloc(sizeof(TransformChannelData));
+
+    dataPtr->refCount = 1;
+    Tcl_DStringInit(&ds);
+    Tcl_GetChannelOption(interp, chan, "-blocking", &ds);
+    dataPtr->readIsFlushed = 0;
+    dataPtr->eofPending = 0;
+    dataPtr->flags = 0;
+    if (ds.string[0] == '0') {
+	dataPtr->flags |= CHANNEL_ASYNC;
+    }
+    Tcl_DStringFree(&ds);
+
+    dataPtr->watchMask = 0;
+    dataPtr->mode = mode;
+    dataPtr->timer = NULL;
+    dataPtr->maxRead = 4096;	/* Initial value not relevant. */
+    dataPtr->interp = interp;
+    dataPtr->command = cmdObjPtr;
+    Tcl_IncrRefCount(dataPtr->command);
+
+    ResultInit(&dataPtr->result);
+
+    dataPtr->self = Tcl_StackChannel(interp, &transformChannelType, dataPtr,
+	    mode, chan);
+    if (dataPtr->self == NULL) {
+	Tcl_AppendPrintfToObj(Tcl_GetObjResult(interp),
+		"\nfailed to stack channel \"%s\"", Tcl_GetChannelName(chan));
+	ReleaseData(dataPtr);
+	return TCL_ERROR;
+    }
+    Tcl_Preserve(dataPtr->self);
+
+    /*
+     * At last initialize the transformation at the script level.
+     */
+
+    PreserveData(dataPtr);
+    if ((dataPtr->mode & TCL_WRITABLE) && ExecuteCallback(dataPtr, NULL,
+	    A_CREATE_WRITE, NULL, 0, TRANSMIT_DONT, P_NO_PRESERVE) != TCL_OK){
+	Tcl_UnstackChannel(interp, chan);
+	ReleaseData(dataPtr);
+	return TCL_ERROR;
+    }
+
+    if ((dataPtr->mode & TCL_READABLE) && ExecuteCallback(dataPtr, NULL,
+	    A_CREATE_READ, NULL, 0, TRANSMIT_DONT, P_NO_PRESERVE) != TCL_OK) {
+	ExecuteCallback(dataPtr, NULL, A_DELETE_WRITE, NULL, 0, TRANSMIT_DONT,
+		P_NO_PRESERVE);
+	Tcl_UnstackChannel(interp, chan);
+	ReleaseData(dataPtr);
+	return TCL_ERROR;
+    }
+
+    ReleaseData(dataPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExecuteCallback --
+ *
+ *	Executes the defined callback for buffer and operation.
+ *
+ * Side effects:
+ *	As of the executed tcl script.
+ *
+ * Result:
+ *	A standard TCL error code. In case of an error a message is left in
+ *	the result area of the specified interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ExecuteCallback(
+    TransformChannelData *dataPtr,
+				/* Transformation with the callback. */
+    Tcl_Interp *interp,		/* Current interpreter, possibly NULL. */
+    unsigned char *op,		/* Operation invoking the callback. */
+    unsigned char *buf,		/* Buffer to give to the script. */
+    int bufLen,			/* And its length. */
+    int transmit,		/* Flag, determines whether the result of the
+				 * callback is sent to the underlying channel
+				 * or not. */
+    int preserve)		/* Flag. If true the procedure will preserve
+				 * the result state of all accessed
+				 * interpreters. */
+{
+    Tcl_Obj *resObj;		/* See below, switch (transmit). */
+    int resLen;
+    unsigned char *resBuf;
+    Tcl_InterpState state = NULL;
+    int res = TCL_OK;
+    Tcl_Obj *command = TclListObjCopy(NULL, dataPtr->command);
+    Tcl_Interp *eval = dataPtr->interp;
+
+    Tcl_Preserve(eval);
+
+    /*
+     * Step 1, create the complete command to execute. Do this by appending
+     * operation and buffer to operate upon to a copy of the callback
+     * definition. We *cannot* create a list containing 3 objects and then use
+     * 'Tcl_EvalObjv', because the command may contain additional prefixed
+     * arguments. Feather's curried commands would come in handy here.
+     */
+
+    if (preserve == P_PRESERVE) {
+	state = Tcl_SaveInterpState(eval, res);
+    }
+
+    Tcl_IncrRefCount(command);
+    Tcl_ListObjAppendElement(NULL, command, Tcl_NewStringObj((char *) op, -1));
+
+    /*
+     * Use a byte-array to prevent the misinterpretation of binary data coming
+     * through as UTF while at the tcl level.
+     */
+
+    Tcl_ListObjAppendElement(NULL, command, Tcl_NewByteArrayObj(buf, bufLen));
+
+    /*
+     * Step 2, execute the command at the global level of the interpreter used
+     * to create the transformation. Destroy the command afterward. If an
+     * error occured and the current interpreter is defined and not equal to
+     * the interpreter for the callback, then copy the error message into
+     * current interpreter. Don't copy if in preservation mode.
+     */
+
+    res = Tcl_EvalObjEx(eval, command, TCL_EVAL_GLOBAL);
+    Tcl_DecrRefCount(command);
+    command = NULL;
+
+    if ((res != TCL_OK) && (interp != NULL) && (eval != interp)
+	    && (preserve == P_NO_PRESERVE)) {
+	Tcl_SetObjResult(interp, Tcl_GetObjResult(eval));
+	Tcl_Release(eval);
+	return res;
+    }
+
+    /*
+     * Step 3, transmit a possible conversion result to the underlying
+     * channel, or ourselves.
+     */
+
+    switch (transmit) {
+    case TRANSMIT_DONT:
+	/* nothing to do */
+	break;
+
+    case TRANSMIT_DOWN:
+	if (dataPtr->self == NULL) {
+	    break;
+	}
+	resObj = Tcl_GetObjResult(eval);
+	resBuf = Tcl_GetByteArrayFromObj(resObj, &resLen);
+	Tcl_WriteRaw(Tcl_GetStackedChannel(dataPtr->self), (char *) resBuf,
+		resLen);
+	break;
+
+    case TRANSMIT_SELF:
+	if (dataPtr->self == NULL) {
+	    break;
+	}
+	resObj = Tcl_GetObjResult(eval);
+	resBuf = Tcl_GetByteArrayFromObj(resObj, &resLen);
+	Tcl_WriteRaw(dataPtr->self, (char *) resBuf, resLen);
+	break;
+
+    case TRANSMIT_IBUF:
+	resObj = Tcl_GetObjResult(eval);
+	resBuf = Tcl_GetByteArrayFromObj(resObj, &resLen);
+	ResultAdd(&dataPtr->result, resBuf, resLen);
+	break;
+
+    case TRANSMIT_NUM:
+	/*
+	 * Interpret result as integer number.
+	 */
+
+	resObj = Tcl_GetObjResult(eval);
+	TclGetIntFromObj(eval, resObj, &dataPtr->maxRead);
+	break;
+    }
+
+    Tcl_ResetResult(eval);
+    if (preserve == P_PRESERVE) {
+	(void) Tcl_RestoreInterpState(eval, state);
+    }
+    Tcl_Release(eval);
+    return res;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformBlockModeProc --
+ *
+ *	Trap handler. Called by the generic IO system during option processing
+ *	to change the blocking mode of the channel.
+ *
+ * Side effects:
+ *	Forwards the request to the underlying channel.
+ *
+ * Result:
+ *	0 if successful, errno when failed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformBlockModeProc(
+    ClientData instanceData,	/* State of transformation. */
+    int mode)			/* New blocking mode. */
+{
+    TransformChannelData *dataPtr = instanceData;
+
+    if (mode == TCL_MODE_NONBLOCKING) {
+	dataPtr->flags |= CHANNEL_ASYNC;
+    } else {
+	dataPtr->flags &= ~CHANNEL_ASYNC;
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformCloseProc --
+ *
+ *	Trap handler. Called by the generic IO system during destruction of
+ *	the transformation channel.
+ *
+ * Side effects:
+ *	Releases the memory allocated in 'Tcl_TransformObjCmd'.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformCloseProc(
+    ClientData instanceData,
+    Tcl_Interp *interp)
+{
+    TransformChannelData *dataPtr = instanceData;
+
+    /*
+     * Important: In this procedure 'dataPtr->self' already points to the
+     * underlying channel.
+     *
+     * There is no need to cancel an existing channel handler, this is already
+     * done. Either by 'Tcl_UnstackChannel' or by the general cleanup in
+     * 'Tcl_Close'.
+     *
+     * But we have to cancel an active timer to prevent it from firing on the
+     * removed channel.
+     */
+
+    if (dataPtr->timer != NULL) {
+	Tcl_DeleteTimerHandler(dataPtr->timer);
+	dataPtr->timer = NULL;
+    }
+
+    /*
+     * Now flush data waiting in internal buffers to output and input. The
+     * input must be done despite the fact that there is no real receiver for
+     * it anymore. But the scripts might have sideeffects other parts of the
+     * system rely on (f.e. signaling the close to interested parties).
+     */
+
+    PreserveData(dataPtr);
+    if (dataPtr->mode & TCL_WRITABLE) {
+	ExecuteCallback(dataPtr, interp, A_FLUSH_WRITE, NULL, 0,
+		TRANSMIT_DOWN, P_PRESERVE);
+    }
+
+    if ((dataPtr->mode & TCL_READABLE) && !dataPtr->readIsFlushed) {
+	dataPtr->readIsFlushed = 1;
+	ExecuteCallback(dataPtr, interp, A_FLUSH_READ, NULL, 0, TRANSMIT_IBUF,
+		P_PRESERVE);
+    }
+
+    if (dataPtr->mode & TCL_WRITABLE) {
+	ExecuteCallback(dataPtr, interp, A_DELETE_WRITE, NULL, 0,
+		TRANSMIT_DONT, P_PRESERVE);
+    }
+    if (dataPtr->mode & TCL_READABLE) {
+	ExecuteCallback(dataPtr, interp, A_DELETE_READ, NULL, 0,
+		TRANSMIT_DONT, P_PRESERVE);
+    }
+    ReleaseData(dataPtr);
+
+    /*
+     * General cleanup.
+     */
+
+    Tcl_Release(dataPtr->self);
+    dataPtr->self = NULL;
+    ReleaseData(dataPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformInputProc --
+ *
+ *	Called by the generic IO system to convert read data.
+ *
+ * Side effects:
+ *	As defined by the conversion.
+ *
+ * Result:
+ *	A transformed buffer.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformInputProc(
+    ClientData instanceData,
+    char *buf,
+    int toRead,
+    int *errorCodePtr)
+{
+    TransformChannelData *dataPtr = instanceData;
+    int gotBytes, read, copied;
+    Tcl_Channel downChan;
+
+    /*
+     * Should assert(dataPtr->mode & TCL_READABLE);
+     */
+
+    if (toRead == 0 || dataPtr->self == NULL) {
+	/*
+	 * Catch a no-op. TODO: Is this a panic()?
+	 */
+	return 0;
+    }
+
+    gotBytes = 0;
+    downChan = Tcl_GetStackedChannel(dataPtr->self);
+
+    PreserveData(dataPtr);
+    while (toRead > 0) {
+	/*
+	 * Loop until the request is satisfied (or no data is available from
+	 * below, possibly EOF).
+	 */
+
+	copied = ResultCopy(&dataPtr->result, UCHARP(buf), toRead);
+	toRead -= copied;
+	buf += copied;
+	gotBytes += copied;
+
+	if (toRead == 0) {
+	    /*
+	     * The request was completely satisfied from our buffers. We can
+	     * break out of the loop and return to the caller.
+	     */
+
+	    break;
+	}
+
+	/*
+	 * Length (dataPtr->result) == 0, toRead > 0 here. Use the incoming
+	 * 'buf'! as target to store the intermediary information read from
+	 * the underlying channel.
+	 *
+	 * Ask the tcl level how much data it allows us to read from the
+	 * underlying channel. This feature allows the transform to signal EOF
+	 * upstream although there is none downstream. Useful to control an
+	 * unbounded 'fcopy', either through counting bytes, or by pattern
+	 * matching.
+	 */
+
+	ExecuteCallback(dataPtr, NULL, A_QUERY_MAXREAD, NULL, 0,
+		TRANSMIT_NUM /* -> maxRead */, P_PRESERVE);
+
+	if (dataPtr->maxRead >= 0) {
+	    if (dataPtr->maxRead < toRead) {
+		toRead = dataPtr->maxRead;
+	    }
+	} /* else: 'maxRead < 0' == Accept the current value of toRead. */
+	if (toRead <= 0) {
+	    break;
+	}
+	if (dataPtr->eofPending) {
+	    /*
+	     * Already saw EOF from downChan; don't ask again.
+	     * NOTE: Could move this up to avoid the last maxRead
+	     * execution.  Believe this would still be correct behavior,
+	     * but the test suite tests the whole command callback
+	     * sequence, so leave it unchanged for now.
+	     */
+
+	    break;
+	}
+
+	/*
+	 * Get bytes from the underlying channel.
+	 */
+
+	read = Tcl_ReadRaw(downChan, buf, toRead);
+	if (read < 0) {
+	    if (Tcl_InputBlocked(downChan) && (gotBytes > 0)) {
+		/*
+		 * Zero bytes available from downChan because blocked.
+		 * But nonzero bytes already copied, so total is a
+		 * valid blocked short read. Return to caller.
+		 */
+
+		break;
+	    }
+
+	    /*
+	     * Either downChan is not blocked (there's a real error).
+	     * or it is and there are no bytes copied yet.  In either
+	     * case we want to pass the "error" along to the caller,
+	     * either to report an error, or to signal to the caller
+	     * that zero bytes are available because blocked.
+	     */
+
+	    *errorCodePtr = Tcl_GetErrno();
+	    gotBytes = -1;
+	    break;
+	} else if (read == 0) {
+
+	    /*
+	     * Zero returned from Tcl_ReadRaw() always indicates EOF
+	     * on the down channel.
+	     */
+
+	    dataPtr->eofPending = 1;
+	    dataPtr->readIsFlushed = 1;
+	    ExecuteCallback(dataPtr, NULL, A_FLUSH_READ, NULL, 0,
+		    TRANSMIT_IBUF, P_PRESERVE);
+
+	    if (ResultEmpty(&dataPtr->result)) {
+		/*
+		 * We had nothing to flush.
+		 */
+
+		break;
+	    }
+
+	    continue;		/* at: while (toRead > 0) */
+	} /* read == 0 */
+
+	/*
+	 * Transform the read chunk and add the result to our read buffer
+	 * (dataPtr->result).
+	 */
+
+	if (ExecuteCallback(dataPtr, NULL, A_READ, UCHARP(buf), read,
+		TRANSMIT_IBUF, P_PRESERVE) != TCL_OK) {
+	    *errorCodePtr = EINVAL;
+	    gotBytes = -1;
+	    break;
+	}
+    } /* while toRead > 0 */
+
+    if (gotBytes == 0) {
+	dataPtr->eofPending = 0;
+    }
+    ReleaseData(dataPtr);
+    return gotBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformOutputProc --
+ *
+ *	Called by the generic IO system to convert data waiting to be written.
+ *
+ * Side effects:
+ *	As defined by the transformation.
+ *
+ * Result:
+ *	A transformed buffer.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformOutputProc(
+    ClientData instanceData,
+    const char *buf,
+    int toWrite,
+    int *errorCodePtr)
+{
+    TransformChannelData *dataPtr = instanceData;
+
+    /*
+     * Should assert(dataPtr->mode & TCL_WRITABLE);
+     */
+
+    if (toWrite == 0) {
+	/*
+	 * Catch a no-op.
+	 */
+
+	return 0;
+    }
+
+    PreserveData(dataPtr);
+    if (ExecuteCallback(dataPtr, NULL, A_WRITE, UCHARP(buf), toWrite,
+	    TRANSMIT_DOWN, P_NO_PRESERVE) != TCL_OK) {
+	*errorCodePtr = EINVAL;
+	toWrite = -1;
+    }
+    ReleaseData(dataPtr);
+
+    return toWrite;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformSeekProc --
+ *
+ *	This procedure is called by the generic IO level to move the access
+ *	point in a channel.
+ *
+ * Side effects:
+ *	Moves the location at which the channel will be accessed in future
+ *	operations. Flushes all transformation buffers, then forwards it to
+ *	the underlying channel.
+ *
+ * Result:
+ *	-1 if failed, the new position if successful. An output argument
+ *	contains the POSIX error code if an error occurred, or zero.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformSeekProc(
+    ClientData instanceData,	/* The channel to manipulate. */
+    long offset,		/* Size of movement. */
+    int mode,			/* How to move. */
+    int *errorCodePtr)		/* Location of error flag. */
+{
+    TransformChannelData *dataPtr = instanceData;
+    Tcl_Channel parent = Tcl_GetStackedChannel(dataPtr->self);
+    const Tcl_ChannelType *parentType = Tcl_GetChannelType(parent);
+    Tcl_DriverSeekProc *parentSeekProc = Tcl_ChannelSeekProc(parentType);
+
+    if ((offset == 0) && (mode == SEEK_CUR)) {
+	/*
+	 * This is no seek but a request to tell the caller the current
+	 * location. Simply pass the request down.
+	 */
+
+	return parentSeekProc(Tcl_GetChannelInstanceData(parent), offset,
+		mode, errorCodePtr);
+    }
+
+    /*
+     * It is a real request to change the position. Flush all data waiting for
+     * output and discard everything in the input buffers. Then pass the
+     * request down, unchanged.
+     */
+
+    PreserveData(dataPtr);
+    if (dataPtr->mode & TCL_WRITABLE) {
+	ExecuteCallback(dataPtr, NULL, A_FLUSH_WRITE, NULL, 0, TRANSMIT_DOWN,
+		P_NO_PRESERVE);
+    }
+
+    if (dataPtr->mode & TCL_READABLE) {
+	ExecuteCallback(dataPtr, NULL, A_CLEAR_READ, NULL, 0, TRANSMIT_DONT,
+		P_NO_PRESERVE);
+	ResultClear(&dataPtr->result);
+	dataPtr->readIsFlushed = 0;
+	dataPtr->eofPending = 0;
+    }
+    ReleaseData(dataPtr);
+
+    return parentSeekProc(Tcl_GetChannelInstanceData(parent), offset, mode,
+	    errorCodePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformWideSeekProc --
+ *
+ *	This procedure is called by the generic IO level to move the access
+ *	point in a channel, with a (potentially) 64-bit offset.
+ *
+ * Side effects:
+ *	Moves the location at which the channel will be accessed in future
+ *	operations. Flushes all transformation buffers, then forwards it to
+ *	the underlying channel.
+ *
+ * Result:
+ *	-1 if failed, the new position if successful. An output argument
+ *	contains the POSIX error code if an error occurred, or zero.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_WideInt
+TransformWideSeekProc(
+    ClientData instanceData,	/* The channel to manipulate. */
+    Tcl_WideInt offset,		/* Size of movement. */
+    int mode,			/* How to move. */
+    int *errorCodePtr)		/* Location of error flag. */
+{
+    TransformChannelData *dataPtr = instanceData;
+    Tcl_Channel parent = Tcl_GetStackedChannel(dataPtr->self);
+    const Tcl_ChannelType *parentType	= Tcl_GetChannelType(parent);
+    Tcl_DriverSeekProc *parentSeekProc = Tcl_ChannelSeekProc(parentType);
+    Tcl_DriverWideSeekProc *parentWideSeekProc =
+	    Tcl_ChannelWideSeekProc(parentType);
+    ClientData parentData = Tcl_GetChannelInstanceData(parent);
+
+    if ((offset == Tcl_LongAsWide(0)) && (mode == SEEK_CUR)) {
+	/*
+	 * This is no seek but a request to tell the caller the current
+	 * location. Simply pass the request down.
+	 */
+
+	if (parentWideSeekProc != NULL) {
+	    return parentWideSeekProc(parentData, offset, mode, errorCodePtr);
+	}
+
+	return Tcl_LongAsWide(parentSeekProc(parentData, 0, mode,
+		errorCodePtr));
+    }
+
+    /*
+     * It is a real request to change the position. Flush all data waiting for
+     * output and discard everything in the input buffers. Then pass the
+     * request down, unchanged.
+     */
+
+    PreserveData(dataPtr);
+    if (dataPtr->mode & TCL_WRITABLE) {
+	ExecuteCallback(dataPtr, NULL, A_FLUSH_WRITE, NULL, 0, TRANSMIT_DOWN,
+		P_NO_PRESERVE);
+    }
+
+    if (dataPtr->mode & TCL_READABLE) {
+	ExecuteCallback(dataPtr, NULL, A_CLEAR_READ, NULL, 0, TRANSMIT_DONT,
+		P_NO_PRESERVE);
+	ResultClear(&dataPtr->result);
+	dataPtr->readIsFlushed = 0;
+	dataPtr->eofPending = 0;
+    }
+    ReleaseData(dataPtr);
+
+    /*
+     * If we have a wide seek capability, we should stick with that.
+     */
+
+    if (parentWideSeekProc != NULL) {
+	return parentWideSeekProc(parentData, offset, mode, errorCodePtr);
+    }
+
+    /*
+     * We're transferring to narrow seeks at this point; this is a bit complex
+     * because we have to check whether the seek is possible first (i.e.
+     * whether we are losing information in truncating the bits of the
+     * offset). Luckily, there's a defined error for what happens when trying
+     * to go out of the representable range.
+     */
+
+    if (offset<Tcl_LongAsWide(LONG_MIN) || offset>Tcl_LongAsWide(LONG_MAX)) {
+	*errorCodePtr = EOVERFLOW;
+	return Tcl_LongAsWide(-1);
+    }
+
+    return Tcl_LongAsWide(parentSeekProc(parentData, Tcl_WideAsLong(offset),
+	    mode, errorCodePtr));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformSetOptionProc --
+ *
+ *	Called by generic layer to handle the reconfiguration of channel
+ *	specific options. As this channel type does not have such, it simply
+ *	passes all requests downstream.
+ *
+ * Side effects:
+ *	As defined by the channel downstream.
+ *
+ * Result:
+ *	A standard TCL error code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformSetOptionProc(
+    ClientData instanceData,
+    Tcl_Interp *interp,
+    const char *optionName,
+    const char *value)
+{
+    TransformChannelData *dataPtr = instanceData;
+    Tcl_Channel downChan = Tcl_GetStackedChannel(dataPtr->self);
+    Tcl_DriverSetOptionProc *setOptionProc;
+
+    setOptionProc = Tcl_ChannelSetOptionProc(Tcl_GetChannelType(downChan));
+    if (setOptionProc == NULL) {
+	return TCL_ERROR;
+    }
+
+    return setOptionProc(Tcl_GetChannelInstanceData(downChan), interp,
+	    optionName, value);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformGetOptionProc --
+ *
+ *	Called by generic layer to handle requests for the values of channel
+ *	specific options. As this channel type does not have such, it simply
+ *	passes all requests downstream.
+ *
+ * Side effects:
+ *	As defined by the channel downstream.
+ *
+ * Result:
+ *	A standard TCL error code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformGetOptionProc(
+    ClientData instanceData,
+    Tcl_Interp *interp,
+    const char *optionName,
+    Tcl_DString *dsPtr)
+{
+    TransformChannelData *dataPtr = instanceData;
+    Tcl_Channel downChan = Tcl_GetStackedChannel(dataPtr->self);
+    Tcl_DriverGetOptionProc *getOptionProc;
+
+    getOptionProc = Tcl_ChannelGetOptionProc(Tcl_GetChannelType(downChan));
+    if (getOptionProc != NULL) {
+	return getOptionProc(Tcl_GetChannelInstanceData(downChan), interp,
+		optionName, dsPtr);
+    } else if (optionName == NULL) {
+	/*
+	 * Request is query for all options, this is ok.
+	 */
+
+	return TCL_OK;
+    }
+
+    /*
+     * Request for a specific option has to fail, since we don't have any.
+     */
+
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformWatchProc --
+ *
+ *	Initialize the notifier to watch for events from this channel.
+ *
+ * Side effects:
+ *	Sets up the notifier so that a future event on the channel will be
+ *	seen by Tcl.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static void
+TransformWatchProc(
+    ClientData instanceData,	/* Channel to watch. */
+    int mask)			/* Events of interest. */
+{
+    TransformChannelData *dataPtr = instanceData;
+    Tcl_Channel downChan;
+
+    /*
+     * The caller expressed interest in events occuring for this channel. We
+     * are forwarding the call to the underlying channel now.
+     */
+
+    dataPtr->watchMask = mask;
+
+    /*
+     * No channel handlers any more. We will be notified automatically about
+     * events on the channel below via a call to our 'TransformNotifyProc'.
+     * But we have to pass the interest down now. We are allowed to add
+     * additional 'interest' to the mask if we want to. But this
+     * transformation has no such interest. It just passes the request down,
+     * unchanged.
+     */
+
+    if (dataPtr->self == NULL) {
+	return;
+    }
+    downChan = Tcl_GetStackedChannel(dataPtr->self);
+
+    Tcl_GetChannelType(downChan)->watchProc(
+	    Tcl_GetChannelInstanceData(downChan), mask);
+
+    /*
+     * Management of the internal timer.
+     */
+
+    if ((dataPtr->timer != NULL) &&
+	    (!(mask & TCL_READABLE) || ResultEmpty(&dataPtr->result))) {
+	/*
+	 * A pending timer exists, but either is there no (more) interest in
+	 * the events it generates or nothing is available for reading, so
+	 * remove it.
+	 */
+
+	Tcl_DeleteTimerHandler(dataPtr->timer);
+	dataPtr->timer = NULL;
+    }
+
+    if ((dataPtr->timer == NULL) && (mask & TCL_READABLE)
+	    && !ResultEmpty(&dataPtr->result)) {
+	/*
+	 * There is no pending timer, but there is interest in readable events
+	 * and we actually have data waiting, so generate a timer to flush
+	 * that.
+	 */
+
+	dataPtr->timer = Tcl_CreateTimerHandler(FLUSH_DELAY,
+		TransformChannelHandlerTimer, dataPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformGetFileHandleProc --
+ *
+ *	Called from Tcl_GetChannelHandle to retrieve OS specific file handle
+ *	from inside this channel.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Result:
+ *	The appropriate Tcl_File or NULL if not present.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformGetFileHandleProc(
+    ClientData instanceData,	/* Channel to query. */
+    int direction,		/* Direction of interest. */
+    ClientData *handlePtr)	/* Place to store the handle into. */
+{
+    TransformChannelData *dataPtr = instanceData;
+
+    /*
+     * Return the handle belonging to parent channel. IOW, pass the request
+     * down and the result up.
+     */
+
+    return Tcl_GetChannelHandle(Tcl_GetStackedChannel(dataPtr->self),
+	    direction, handlePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformNotifyProc --
+ *
+ *	Handler called by Tcl to inform us of activity on the underlying
+ *	channel.
+ *
+ * Side effects:
+ *	May process the incoming event by itself.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TransformNotifyProc(
+    ClientData clientData,	/* The state of the notified
+				 * transformation. */
+    int mask)			/* The mask of occuring events. */
+{
+    TransformChannelData *dataPtr = clientData;
+
+    /*
+     * An event occured in the underlying channel. This transformation doesn't
+     * process such events thus returns the incoming mask unchanged.
+     */
+
+    if (dataPtr->timer != NULL) {
+	/*
+	 * Delete an existing timer. It was not fired, yet we are here, so the
+	 * channel below generated such an event and we don't have to. The
+	 * renewal of the interest after the execution of channel handlers
+	 * will eventually cause us to recreate the timer (in
+	 * TransformWatchProc).
+	 */
+
+	Tcl_DeleteTimerHandler(dataPtr->timer);
+	dataPtr->timer = NULL;
+    }
+    return mask;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TransformChannelHandlerTimer --
+ *
+ *	Called by the notifier (-> timer) to flush out information waiting in
+ *	the input buffer.
+ *
+ * Side effects:
+ *	As of 'Tcl_NotifyChannel'.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TransformChannelHandlerTimer(
+    ClientData clientData)	/* Transformation to query. */
+{
+    TransformChannelData *dataPtr = clientData;
+
+    dataPtr->timer = NULL;
+    if (!(dataPtr->watchMask&TCL_READABLE) || ResultEmpty(&dataPtr->result)) {
+	/*
+	 * The timer fired, but either is there no (more) interest in the
+	 * events it generates or nothing is available for reading, so ignore
+	 * it and don't recreate it.
+	 */
+
+	return;
+    }
+    Tcl_NotifyChannel(dataPtr->self, TCL_READABLE);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultClear --
+ *
+ *	Deallocates any memory allocated by 'ResultAdd'.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+ResultClear(
+    ResultBuffer *r)		/* Reference to the buffer to clear out. */
+{
+    r->used = 0;
+
+    if (r->allocated) {
+	ckfree(r->buf);
+	r->buf = NULL;
+	r->allocated = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultInit --
+ *
+ *	Initializes the specified buffer structure. The structure will contain
+ *	valid information for an emtpy buffer.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+ResultInit(
+    ResultBuffer *r)		/* Reference to the structure to
+				 * initialize. */
+{
+    r->used = 0;
+    r->allocated = 0;
+    r->buf = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultEmpty --
+ *
+ *	Returns whether the number of bytes stored in the buffer is zero.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Result:
+ *	A boolean.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ResultEmpty(
+    ResultBuffer *r)		/* The structure to query. */
+{
+    return r->used == 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultCopy --
+ *
+ *	Copies the requested number of bytes from the buffer into the
+ *	specified array and removes them from the buffer afterward. Copies
+ *	less if there is not enough data in the buffer.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	The number of actually copied bytes, possibly less than 'toRead'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ResultCopy(
+    ResultBuffer *r,		/* The buffer to read from. */
+    unsigned char *buf,		/* The buffer to copy into. */
+    size_t toRead)		/* Number of requested bytes. */
+{
+    if (r->used == 0) {
+	/*
+	 * Nothing to copy in the case of an empty buffer.
+	 */
+
+	return 0;
+    } else if (r->used == toRead) {
+	/*
+	 * We have just enough. Copy everything to the caller.
+	 */
+
+	memcpy(buf, r->buf, toRead);
+	r->used = 0;
+    } else if (r->used > toRead) {
+	/*
+	 * The internal buffer contains more than requested. Copy the
+	 * requested subset to the caller, and shift the remaining bytes down.
+	 */
+
+	memcpy(buf, r->buf, toRead);
+	memmove(r->buf, r->buf + toRead, r->used - toRead);
+	r->used -= toRead;
+    } else {
+	/*
+	 * There is not enough in the buffer to satisfy the caller, so take
+	 * everything.
+	 */
+
+	memcpy(buf, r->buf, r->used);
+	toRead = r->used;
+	r->used = 0;
+    }
+    return toRead;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultAdd --
+ *
+ *	Adds the bytes in the specified array to the buffer, by appending it.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+ResultAdd(
+    ResultBuffer *r,		/* The buffer to extend. */
+    unsigned char *buf,		/* The buffer to read from. */
+    size_t toWrite)		/* The number of bytes in 'buf'. */
+{
+    if (r->used + toWrite > r->allocated) {
+	/*
+	 * Extension of the internal buffer is required.
+	 */
+
+	if (r->allocated == 0) {
+	    r->allocated = toWrite + INCREMENT;
+	    r->buf = ckalloc(r->allocated);
+	} else {
+	    r->allocated += toWrite + INCREMENT;
+	    r->buf = ckrealloc(r->buf, r->allocated);
+	}
+    }
+
+    /*
+     * Now we may copy the data.
+     */
+
+    memcpy(r->buf + r->used, buf, toWrite);
+    r->used += toWrite;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIORChan.c b/generic/tclIORChan.c
new file mode 100644
index 0000000..8e1496d
--- /dev/null
+++ b/generic/tclIORChan.c
@@ -0,0 +1,3313 @@
+/*
+ * tclIORChan.c --
+ *
+ *	This file contains the implementation of Tcl's generic channel
+ *	reflection code, which allows the implementation of Tcl channels in
+ *	Tcl code.
+ *
+ *	Parts of this file are based on code contributed by Jean-Claude
+ *	Wippler.
+ *
+ *	See TIP #219 for the specification of this functionality.
+ *
+ * Copyright (c) 2004-2005 ActiveState, a divison of Sophos
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclIO.h"
+#include <assert.h>
+
+#ifndef EINVAL
+#define EINVAL	9
+#endif
+#ifndef EOK
+#define EOK	0
+#endif
+
+/*
+ * Signatures of all functions used in the C layer of the reflection.
+ */
+
+static int		ReflectClose(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		ReflectInput(ClientData clientData, char *buf,
+			    int toRead, int *errorCodePtr);
+static int		ReflectOutput(ClientData clientData, const char *buf,
+			    int toWrite, int *errorCodePtr);
+static void		ReflectWatch(ClientData clientData, int mask);
+static int		ReflectBlock(ClientData clientData, int mode);
+#ifdef TCL_THREADS
+static void		ReflectThread(ClientData clientData, int action);
+static int		ReflectEventRun(Tcl_Event *ev, int flags);
+static int		ReflectEventDelete(Tcl_Event *ev, ClientData cd);
+#endif
+static Tcl_WideInt	ReflectSeekWide(ClientData clientData,
+			    Tcl_WideInt offset, int mode, int *errorCodePtr);
+static int		ReflectSeek(ClientData clientData, long offset,
+			    int mode, int *errorCodePtr);
+static int		ReflectGetOption(ClientData clientData,
+			    Tcl_Interp *interp, const char *optionName,
+			    Tcl_DString *dsPtr);
+static int		ReflectSetOption(ClientData clientData,
+			    Tcl_Interp *interp, const char *optionName,
+			    const char *newValue);
+
+/*
+ * The C layer channel type/driver definition used by the reflection. This is
+ * a version 3 structure.
+ */
+
+static const Tcl_ChannelType tclRChannelType = {
+    "tclrchannel",	   /* Type name.				  */
+    TCL_CHANNEL_VERSION_5, /* v5 channel */
+    ReflectClose,	   /* Close channel, clean instance data	  */
+    ReflectInput,	   /* Handle read request			  */
+    ReflectOutput,	   /* Handle write request			  */
+    ReflectSeek,	   /* Move location of access point.	NULL'able */
+    ReflectSetOption,	   /* Set options.			NULL'able */
+    ReflectGetOption,	   /* Get options.			NULL'able */
+    ReflectWatch,	   /* Initialize notifier			  */
+    NULL,		   /* Get OS handle from the channel.	NULL'able */
+    NULL,		   /* No close2 support.		NULL'able */
+    ReflectBlock,	   /* Set blocking/nonblocking.		NULL'able */
+    NULL,		   /* Flush channel. Not used by core.	NULL'able */
+    NULL,		   /* Handle events.			NULL'able */
+    ReflectSeekWide,	   /* Move access point (64 bit).	NULL'able */
+#ifdef TCL_THREADS
+    ReflectThread,         /* thread action, tracking owner */
+#else
+    NULL,		   /* thread action */
+#endif
+    NULL		   /* truncate */
+};
+
+/*
+ * Instance data for a reflected channel. ===========================
+ */
+
+typedef struct {
+    Tcl_Channel chan;		/* Back reference to generic channel
+				 * structure. */
+    Tcl_Interp *interp;		/* Reference to the interpreter containing the
+				 * Tcl level part of the channel. NULL here
+				 * signals the channel is dead because the
+				 * interpreter/thread containing its Tcl
+				 * command is gone.
+				 */
+#ifdef TCL_THREADS
+    Tcl_ThreadId thread;	/* Thread the 'interp' belongs to. == Handler thread */
+    Tcl_ThreadId owner;         /* Thread owning the structure.    == Channel thread */
+#endif
+    Tcl_Obj *cmd;		/* Callback command prefix */
+    Tcl_Obj *methods;		/* Methods to append to command prefix */
+    Tcl_Obj *name;		/* Name of the channel as created */
+
+    int mode;			/* Mask of R/W mode */
+    int interest;		/* Mask of events the channel is interested
+				 * in. */
+
+    int dead;			/* Boolean signal that some operations
+				 * should no longer be attempted. */
+
+    /*
+     * Note regarding the usage of timers.
+     *
+     * Most channel implementations need a timer in the C level to ensure that
+     * data in buffers is flushed out through the generation of fake file
+     * events.
+     *
+     * See 'rechan', 'memchan', etc.
+     *
+     * Here this is _not_ required. Interest in events is posted to the Tcl
+     * level via 'watch'. And posting of events is possible from the Tcl level
+     * as well, via 'chan postevent'. This means that the generation of all
+     * events, fake or not, timer based or not, is completely in the hands of
+     * the Tcl level. Therefore no timer here.
+     */
+} ReflectedChannel;
+
+/*
+ * Structure of the table maping from channel handles to reflected
+ * channels. Each interpreter which has the handler command for one or more
+ * reflected channels records them in such a table, so that 'chan postevent'
+ * is able to find them even if the actual channel was moved to a different
+ * interpreter and/or thread.
+ *
+ * The table is reachable via the standard interpreter AssocData, the key is
+ * defined below.
+ */
+
+typedef struct {
+    Tcl_HashTable map;
+} ReflectedChannelMap;
+
+#define RCMKEY "ReflectedChannelMap"
+
+/*
+ * Event literals. ==================================================
+ */
+
+static const char *const eventOptions[] = {
+    "read", "write", NULL
+};
+typedef enum {
+    EVENT_READ, EVENT_WRITE
+} EventOption;
+
+/*
+ * Method literals. ==================================================
+ */
+
+static const char *const methodNames[] = {
+    "blocking",		/* OPT */
+    "cget",		/* OPT \/ Together or none */
+    "cgetall",		/* OPT /\ of these two     */
+    "configure",	/* OPT */
+    "finalize",		/*     */
+    "initialize",	/*     */
+    "read",		/* OPT */
+    "seek",		/* OPT */
+    "watch",		/*     */
+    "write",		/* OPT */
+    NULL
+};
+typedef enum {
+    METH_BLOCKING,
+    METH_CGET,
+    METH_CGETALL,
+    METH_CONFIGURE,
+    METH_FINAL,
+    METH_INIT,
+    METH_READ,
+    METH_SEEK,
+    METH_WATCH,
+    METH_WRITE
+} MethodName;
+
+#define FLAG(m) (1 << (m))
+#define REQUIRED_METHODS \
+	(FLAG(METH_INIT) | FLAG(METH_FINAL) | FLAG(METH_WATCH))
+#define NULLABLE_METHODS \
+	(FLAG(METH_BLOCKING) | FLAG(METH_SEEK) | \
+	FLAG(METH_CONFIGURE) | FLAG(METH_CGET) | FLAG(METH_CGETALL))
+
+#define RANDW \
+	(TCL_READABLE | TCL_WRITABLE)
+
+#define IMPLIES(a,b)	((!(a)) || (b))
+#define NEGIMPL(a,b)
+#define HAS(x,f)	(x & FLAG(f))
+
+#ifdef TCL_THREADS
+/*
+ * Thread specific types and structures.
+ *
+ * We are here essentially creating a very specific implementation of 'thread
+ * send'.
+ */
+
+/*
+ * Enumeration of all operations which can be forwarded.
+ */
+
+typedef enum {
+    ForwardedClose,
+    ForwardedInput,
+    ForwardedOutput,
+    ForwardedSeek,
+    ForwardedWatch,
+    ForwardedBlock,
+    ForwardedSetOpt,
+    ForwardedGetOpt,
+    ForwardedGetOptAll
+} ForwardedOperation;
+
+/*
+ * Event used to forward driver invocations to the thread actually managing
+ * the channel. We cannot construct the command to execute and forward that.
+ * Because then it will contain a mixture of Tcl_Obj's belonging to both the
+ * command handler thread (CT), and the thread managing the channel (MT),
+ * executed in CT. Tcl_Obj's are not allowed to cross thread boundaries. So we
+ * forward an operation code, the argument details, and reference to results.
+ * The command is assembled in the CT and belongs fully to that thread. No
+ * sharing problems.
+ */
+
+typedef struct {
+    int code;			/* O: Ok/Fail of the cmd handler */
+    char *msgStr;		/* O: Error message for handler failure */
+    int mustFree;		/* O: True if msgStr is allocated, false if
+				 * otherwise (static). */
+} ForwardParamBase;
+
+/*
+ * Operation specific parameter/result structures. (These are "subtypes" of
+ * ForwardParamBase. Where an operation does not need any special types, it
+ * has no "subtype" and just uses ForwardParamBase, as listed above.)
+ */
+
+struct ForwardParamInput {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    char *buf;			/* O: Where to store the read bytes */
+    int toRead;			/* I: #bytes to read,
+				 * O: #bytes actually read */
+};
+struct ForwardParamOutput {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    const char *buf;		/* I: Where the bytes to write come from */
+    int toWrite;		/* I: #bytes to write,
+				 * O: #bytes actually written */
+};
+struct ForwardParamSeek {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    int seekMode;		/* I: How to seek */
+    Tcl_WideInt offset;		/* I: Where to seek,
+				 * O: New location */
+};
+struct ForwardParamWatch {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    int mask;			/* I: What events to watch for */
+};
+struct ForwardParamBlock {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    int nonblocking;		/* I: What mode to activate */
+};
+struct ForwardParamSetOpt {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    const char *name;		/* Name of option to set */
+    const char *value;		/* Value to set */
+};
+struct ForwardParamGetOpt {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    const char *name;		/* Name of option to get, maybe NULL */
+    Tcl_DString *value;		/* Result */
+};
+
+/*
+ * Now join all these together in a single union for convenience.
+ */
+
+typedef union ForwardParam {
+    ForwardParamBase base;
+    struct ForwardParamInput input;
+    struct ForwardParamOutput output;
+    struct ForwardParamSeek seek;
+    struct ForwardParamWatch watch;
+    struct ForwardParamBlock block;
+    struct ForwardParamSetOpt setOpt;
+    struct ForwardParamGetOpt getOpt;
+} ForwardParam;
+
+/*
+ * Forward declaration.
+ */
+
+typedef struct ForwardingResult ForwardingResult;
+
+/*
+ * General event structure, with reference to operation specific data.
+ */
+
+typedef struct {
+    Tcl_Event event;		/* Basic event data, has to be first item */
+    ForwardingResult *resultPtr;
+    ForwardedOperation op;	/* Forwarded driver operation */
+    ReflectedChannel *rcPtr;	/* Channel instance */
+    ForwardParam *param;	/* Packaged arguments and return values, a
+				 * ForwardParam pointer. */
+} ForwardingEvent;
+
+/*
+ * Structure to manage the result of the forwarding. This is not the result of
+ * the operation itself, but about the success of the forward event itself.
+ * The event can be successful, even if the operation which was forwarded
+ * failed. It is also there to manage the synchronization between the involved
+ * threads.
+ */
+
+struct ForwardingResult {
+    Tcl_ThreadId src;		/* Originating thread. */
+    Tcl_ThreadId dst;		/* Thread the op was forwarded to. */
+    Tcl_Interp *dsti;		/* Interpreter in the thread the op was
+				 * forwarded to. */
+    /*
+     * Note regarding 'dsti' above: Its information is also available via the
+     * chain evPtr->rcPtr->interp, however, as can be seen, two more
+     * indirections are needed to retrieve it. And the evPtr may be gone,
+     * breaking the chain.
+     */
+    Tcl_Condition done;		/* Condition variable the forwarder blocks
+				 * on. */
+    int result;			/* TCL_OK or TCL_ERROR */
+    ForwardingEvent *evPtr;	/* Event the result belongs to. */
+    ForwardingResult *prevPtr, *nextPtr;
+				/* Links into the list of pending forwarded
+				 * results. */
+};
+
+typedef struct {
+    /*
+     * Table of all reflected channels owned by this thread. This is the
+     * per-thread version of the per-interpreter map.
+     */
+
+    ReflectedChannelMap *rcmPtr;
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * List of forwarded operations which have not completed yet, plus the mutex
+ * to protect the access to this process global list.
+ */
+
+static ForwardingResult *forwardList = NULL;
+TCL_DECLARE_MUTEX(rcForwardMutex)
+
+/*
+ * Function containing the generic code executing a forward, and wrapper
+ * macros for the actual operations we wish to forward. Uses ForwardProc as
+ * the event function executed by the thread receiving a forwarding event
+ * (which executes the appropriate function and collects the result, if any).
+ *
+ * The ExitProc ensures that things do not deadlock when the sending thread
+ * involved in the forwarding exits. It also clean things up so that we don't
+ * leak resources when threads go away.
+ */
+
+static void		ForwardOpToHandlerThread(ReflectedChannel *rcPtr,
+			    ForwardedOperation op, const void *param);
+static int		ForwardProc(Tcl_Event *evPtr, int mask);
+static void		SrcExitProc(ClientData clientData);
+
+#define FreeReceivedError(p) \
+	if ((p)->base.mustFree) {                               \
+	    ckfree((p)->base.msgStr);                           \
+	}
+#define PassReceivedErrorInterp(i,p) \
+	if ((i) != NULL) {                                      \
+	    Tcl_SetChannelErrorInterp((i),                      \
+		    Tcl_NewStringObj((p)->base.msgStr, -1));    \
+	}                                                       \
+	FreeReceivedError(p)
+#define PassReceivedError(c,p) \
+	Tcl_SetChannelError((c), Tcl_NewStringObj((p)->base.msgStr, -1)); \
+	FreeReceivedError(p)
+#define ForwardSetStaticError(p,emsg) \
+	(p)->base.code = TCL_ERROR;                             \
+	(p)->base.mustFree = 0;                                 \
+	(p)->base.msgStr = (char *) (emsg)
+#define ForwardSetDynamicError(p,emsg) \
+	(p)->base.code = TCL_ERROR;                             \
+	(p)->base.mustFree = 1;                                 \
+	(p)->base.msgStr = (char *) (emsg)
+
+static void		ForwardSetObjError(ForwardParam *p, Tcl_Obj *objPtr);
+
+static ReflectedChannelMap *	GetThreadReflectedChannelMap(void);
+static void		DeleteThreadReflectedChannelMap(ClientData clientData);
+
+#endif /* TCL_THREADS */
+
+#define SetChannelErrorStr(c,msgStr) \
+	Tcl_SetChannelError((c), Tcl_NewStringObj((msgStr), -1))
+
+static Tcl_Obj *	MarshallError(Tcl_Interp *interp);
+static void		UnmarshallErrorResult(Tcl_Interp *interp,
+			    Tcl_Obj *msgObj);
+
+/*
+ * Static functions for this file:
+ */
+
+static int		EncodeEventMask(Tcl_Interp *interp,
+			    const char *objName, Tcl_Obj *obj, int *mask);
+static Tcl_Obj *	DecodeEventMask(int mask);
+static ReflectedChannel * NewReflectedChannel(Tcl_Interp *interp,
+			    Tcl_Obj *cmdpfxObj, int mode, Tcl_Obj *handleObj);
+static Tcl_Obj *	NextHandle(void);
+static void		FreeReflectedChannel(ReflectedChannel *rcPtr);
+static int		InvokeTclMethod(ReflectedChannel *rcPtr,
+			    MethodName method, Tcl_Obj *argOneObj,
+			    Tcl_Obj *argTwoObj, Tcl_Obj **resultObjPtr);
+
+static ReflectedChannelMap *	GetReflectedChannelMap(Tcl_Interp *interp);
+static void		DeleteReflectedChannelMap(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		ErrnoReturn(ReflectedChannel *rcPtr, Tcl_Obj *resObj);
+static void		MarkDead(ReflectedChannel *rcPtr);
+
+/*
+ * Global constant strings (messages). ==================
+ * These string are used directly as bypass errors, thus they have to be valid
+ * Tcl lists where the last element is the message itself. Hence the
+ * list-quoting to keep the words of the message together. See also [x].
+ */
+
+static const char *msg_read_toomuch = "{read delivered more than requested}";
+static const char *msg_write_toomuch = "{write wrote more than requested}";
+static const char *msg_write_nothing = "{write wrote nothing}";
+static const char *msg_seek_beforestart = "{Tried to seek before origin}";
+#ifdef TCL_THREADS
+static const char *msg_send_originlost = "{Channel thread lost}";
+#endif /* TCL_THREADS */
+static const char *msg_send_dstlost    = "{Owner lost}";
+static const char *msg_dstlost    = "-code 1 -level 0 -errorcode NONE -errorinfo {} -errorline 1 {Owner lost}";
+
+/*
+ * Main methods to plug into the 'chan' ensemble'. ==================
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChanCreateObjCmd --
+ *
+ *	This function is invoked to process the "chan create" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result. The handle of the new channel is placed in the
+ *	interp result.
+ *
+ * Side effects:
+ *	Creates a new channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclChanCreateObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    ReflectedChannel *rcPtr;	/* Instance data of the new channel */
+    Tcl_Obj *rcId;		/* Handle of the new channel */
+    int mode;			/* R/W mode of new channel. Has to match
+				 * abilities of handler commands */
+    Tcl_Obj *cmdObj;		/* Command prefix, list of words */
+    Tcl_Obj *cmdNameObj;	/* Command name */
+    Tcl_Channel chan;		/* Token for the new channel */
+    Tcl_Obj *modeObj;		/* mode in obj form for method call */
+    int listc;			/* Result of 'initialize', and of */
+    Tcl_Obj **listv;		/* its sublist in the 2nd element */
+    int methIndex;		/* Encoded method name */
+    int result;			/* Result code for 'initialize' */
+    Tcl_Obj *resObj;		/* Result data for 'initialize' */
+    int methods;		/* Bitmask for supported methods. */
+    Channel *chanPtr;		/* 'chan' resolved to internal struct. */
+    Tcl_Obj *err;		/* Error message */
+    ReflectedChannelMap *rcmPtr;
+				/* Map of reflected channels with handlers in
+				 * this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+    int isNew;			/* Placeholder. */
+
+    /*
+     * Syntax:   chan create MODE CMDPREFIX
+     *           [0]  [1]    [2]  [3]
+     *
+     * Actually: rCreate MODE CMDPREFIX
+     *           [0]     [1]  [2]
+     */
+
+#define MODE	(1)
+#define CMD	(2)
+
+    /*
+     * Number of arguments...
+     */
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "mode cmdprefix");
+	return TCL_ERROR;
+    }
+
+    /*
+     * First argument is a list of modes. Allowed entries are "read", "write".
+     * Expect at least one list element. Abbreviations are ok.
+     */
+
+    modeObj = objv[MODE];
+    if (EncodeEventMask(interp, "mode", objv[MODE], &mode) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Second argument is command prefix, i.e. list of words, first word is
+     * name of handler command, other words are fixed arguments. Run the
+     * 'initialize' method to get the list of supported methods. Validate
+     * this.
+     */
+
+    cmdObj = objv[CMD];
+
+    /*
+     * Basic check that the command prefix truly is a list.
+     */
+
+    if (Tcl_ListObjIndex(interp, cmdObj, 0, &cmdNameObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now create the channel.
+     */
+
+    rcId = NextHandle();
+    rcPtr = NewReflectedChannel(interp, cmdObj, mode, rcId);
+
+    /*
+     * Invoke 'initialize' and validate that the handler is present and ok.
+     * Squash the channel if not.
+     *
+     * Note: The conversion of 'mode' back into a Tcl_Obj ensures that
+     * 'initialize' is invoked with canonical mode names, and no
+     * abbreviations. Using modeObj directly could feed abbreviations into the
+     * handler, and the handler is not specified to handle such.
+     */
+
+    modeObj = DecodeEventMask(mode);
+    /* assert modeObj.refCount == 1 */
+    result = InvokeTclMethod(rcPtr, METH_INIT, modeObj, NULL, &resObj);
+    Tcl_DecrRefCount(modeObj);
+
+    if (result != TCL_OK) {
+	UnmarshallErrorResult(interp, resObj);
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	goto error;
+    }
+
+    /*
+     * Verify the result.
+     * - List, of method names. Convert to mask.
+     *   Check for non-optionals through the mask.
+     *   Compare open mode against optional r/w.
+     */
+
+    if (Tcl_ListObjGetElements(NULL, resObj, &listc, &listv) != TCL_OK) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s initialize\" returned non-list: %s",
+                TclGetString(cmdObj), TclGetString(resObj)));
+	Tcl_DecrRefCount(resObj);
+	goto error;
+    }
+
+    methods = 0;
+    while (listc > 0) {
+	if (Tcl_GetIndexFromObj(interp, listv[listc-1], methodNames,
+		"method", TCL_EXACT, &methIndex) != TCL_OK) {
+	    TclNewLiteralStringObj(err, "chan handler \"");
+	    Tcl_AppendObjToObj(err, cmdObj);
+	    Tcl_AppendToObj(err, " initialize\" returned ", -1);
+	    Tcl_AppendObjToObj(err, Tcl_GetObjResult(interp));
+	    Tcl_SetObjResult(interp, err);
+	    Tcl_DecrRefCount(resObj);
+	    goto error;
+	}
+
+	methods |= FLAG(methIndex);
+	listc--;
+    }
+    Tcl_DecrRefCount(resObj);
+
+    if ((REQUIRED_METHODS & methods) != REQUIRED_METHODS) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" does not support all required methods",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    if ((mode & TCL_READABLE) && !HAS(methods, METH_READ)) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" lacks a \"read\" method",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    if ((mode & TCL_WRITABLE) && !HAS(methods, METH_WRITE)) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" lacks a \"write\" method",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    if (!IMPLIES(HAS(methods, METH_CGET), HAS(methods, METH_CGETALL))) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" supports \"cget\" but not \"cgetall\"",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    if (!IMPLIES(HAS(methods, METH_CGETALL), HAS(methods, METH_CGET))) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" supports \"cgetall\" but not \"cget\"",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    Tcl_ResetResult(interp);
+
+    /*
+     * Everything is fine now.
+     */
+
+    chan = Tcl_CreateChannel(&tclRChannelType, TclGetString(rcId), rcPtr,
+	    mode);
+    rcPtr->chan = chan;
+    TclChannelPreserve(chan);
+    chanPtr = (Channel *) chan;
+
+    if ((methods & NULLABLE_METHODS) != NULLABLE_METHODS) {
+	/*
+	 * Some of the nullable methods are not supported. We clone the
+	 * channel type, null the associated C functions, and use the result
+	 * as the actual channel type.
+	 */
+
+	Tcl_ChannelType *clonePtr = ckalloc(sizeof(Tcl_ChannelType));
+
+	memcpy(clonePtr, &tclRChannelType, sizeof(Tcl_ChannelType));
+
+	if (!(methods & FLAG(METH_CONFIGURE))) {
+	    clonePtr->setOptionProc = NULL;
+	}
+
+	if (!(methods & FLAG(METH_CGET)) && !(methods & FLAG(METH_CGETALL))) {
+	    clonePtr->getOptionProc = NULL;
+	}
+	if (!(methods & FLAG(METH_BLOCKING))) {
+	    clonePtr->blockModeProc = NULL;
+	}
+	if (!(methods & FLAG(METH_SEEK))) {
+	    clonePtr->seekProc = NULL;
+	    clonePtr->wideSeekProc = NULL;
+	}
+
+	chanPtr->typePtr = clonePtr;
+    }
+
+    /*
+     * Register the channel in the I/O system, and in our our map for 'chan
+     * postevent'.
+     */
+
+    Tcl_RegisterChannel(interp, chan);
+
+    rcmPtr = GetReflectedChannelMap(interp);
+    hPtr = Tcl_CreateHashEntry(&rcmPtr->map, chanPtr->state->channelName,
+	    &isNew);
+    if (!isNew && chanPtr != Tcl_GetHashValue(hPtr)) {
+	Tcl_Panic("TclChanCreateObjCmd: duplicate channel names");
+    }
+    Tcl_SetHashValue(hPtr, chan);
+#ifdef TCL_THREADS
+    rcmPtr = GetThreadReflectedChannelMap();
+    hPtr = Tcl_CreateHashEntry(&rcmPtr->map, chanPtr->state->channelName,
+	    &isNew);
+    Tcl_SetHashValue(hPtr, chan);
+#endif
+
+    /*
+     * Return handle as result of command.
+     */
+
+    Tcl_SetObjResult(interp,
+            Tcl_NewStringObj(chanPtr->state->channelName, -1));
+    return TCL_OK;
+
+  error:
+    Tcl_DecrRefCount(rcPtr->name);
+    Tcl_DecrRefCount(rcPtr->methods);
+    Tcl_DecrRefCount(rcPtr->cmd);
+    ckfree(rcPtr);
+    return TCL_ERROR;
+
+#undef MODE
+#undef CMD
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChanPostEventObjCmd --
+ *
+ *	This function is invoked to process the "chan postevent" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Posts events to a reflected channel, invokes event handlers. The
+ *	latter implies that arbitrary side effects are possible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_THREADS
+typedef struct {
+    Tcl_Event header;
+    ReflectedChannel *rcPtr;
+    int events;
+} ReflectEvent;
+
+static int
+ReflectEventRun(
+    Tcl_Event *ev,
+    int flags)
+{
+    /* OWNER thread
+     *
+     * Note: When the channel is closed any pending events of this type are
+     * deleted. See ReflectClose() for the Tcl_DeleteEvents() calls
+     * accomplishing that.
+     */
+
+    ReflectEvent *e = (ReflectEvent *) ev;
+
+    Tcl_NotifyChannel(e->rcPtr->chan, e->events);
+    return 1;
+}
+
+static int
+ReflectEventDelete(
+    Tcl_Event *ev,
+    ClientData cd)
+{
+    /* OWNER thread
+     *
+     * Invoked by DeleteThreadReflectedChannelMap() and ReflectClose(). The
+     * latter ensures that no pending events of this type are run on an
+     * invalid channel.
+     */
+
+    ReflectEvent *e = (ReflectEvent *) ev;
+
+    if ((ev->proc != ReflectEventRun) || ((cd != NULL) && (cd != e->rcPtr))) {
+        return 0;
+    }
+    return 1;
+}
+#endif
+
+int
+TclChanPostEventObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    /*
+     * Ensure -> HANDLER thread
+     *
+     * Syntax:   chan postevent CHANNEL EVENTSPEC
+     *           [0]  [1]       [2]     [3]
+     *
+     * Actually: rPostevent CHANNEL EVENTSPEC
+     *           [0]        [1]     [2]
+     *
+     * where EVENTSPEC = {read write ...} (Abbreviations allowed as well).
+     */
+
+#define CHAN	(1)
+#define EVENT	(2)
+
+    const char *chanId;		/* Tcl level channel handle */
+    Tcl_Channel chan;		/* Channel associated to the handle */
+    const Tcl_ChannelType *chanTypePtr;
+				/* Its associated driver structure */
+    ReflectedChannel *rcPtr;	/* Associated instance data */
+    int events;			/* Mask of events to post */
+    ReflectedChannelMap *rcmPtr;/* Map of reflected channels with handlers in
+				 * this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+
+    /*
+     * Number of arguments...
+     */
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channel eventspec");
+	return TCL_ERROR;
+    }
+
+    /*
+     * First argument is a channel, a reflected channel, and the call of this
+     * command is done from the interp defining the channel handler cmd.
+     */
+
+    chanId = TclGetString(objv[CHAN]);
+
+    rcmPtr = GetReflectedChannelMap(interp);
+    hPtr = Tcl_FindHashEntry(&rcmPtr->map, chanId);
+
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can not find reflected channel named \"%s\"", chanId));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CHANNEL", chanId, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Note that the search above subsumes several of the older checks,
+     * namely:
+     *
+     * (1) Does the channel handle refer to a reflected channel?
+     * (2) Is the post event issued from the interpreter holding the handler
+     *     of the reflected channel?
+     *
+     * A successful search answers yes to both. Because the map holds only
+     * handles of reflected channels, and only of such whose handler is
+     * defined in this interpreter.
+     *
+     * We keep the old checks for both, for paranioa, but abort now instead of
+     * throwing errors, as failure now means that our internal datastructures
+     * have gone seriously haywire.
+     */
+
+    chan = Tcl_GetHashValue(hPtr);
+    chanTypePtr = Tcl_GetChannelType(chan);
+
+    /*
+     * We use a function referenced by the channel type as our cookie to
+     * detect calls to non-reflecting channels. The channel type itself is not
+     * suitable, as it might not be the static definition in this file, but a
+     * clone thereof. And while we have reserved the name of the type nothing
+     * in the core checks against violation, so someone else might have
+     * created a channel type using our name, clashing with ourselves.
+     */
+
+    if (chanTypePtr->watchProc != &ReflectWatch) {
+	Tcl_Panic("TclChanPostEventObjCmd: channel is not a reflected channel");
+    }
+
+    rcPtr = Tcl_GetChannelInstanceData(chan);
+
+    if (rcPtr->interp != interp) {
+	Tcl_Panic("TclChanPostEventObjCmd: postevent accepted for call from outside interpreter");
+    }
+
+    /*
+     * Second argument is a list of events. Allowed entries are "read",
+     * "write". Expect at least one list element. Abbreviations are ok.
+     */
+
+    if (EncodeEventMask(interp, "event", objv[EVENT], &events) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check that the channel is actually interested in the provided events.
+     */
+
+    if (events & ~rcPtr->interest) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "tried to post events channel \"%s\" is not interested in",
+                chanId));
+	return TCL_ERROR;
+    }
+
+    /*
+     * We have the channel and the events to post.
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->owner == rcPtr->thread) {
+#endif
+        Tcl_NotifyChannel(chan, events);
+#ifdef TCL_THREADS
+    } else {
+        ReflectEvent *ev = ckalloc(sizeof(ReflectEvent));
+
+        ev->header.proc = ReflectEventRun;
+        ev->events = events;
+        ev->rcPtr = rcPtr;
+
+        /*
+         * We are not preserving the structure here. When the channel is
+         * closed any pending events are deleted, see ReflectClose(), and
+         * ReflectEventDelete(). Trying to preserve and later release when the
+         * event is run may generate a situation where the channel structure
+         * is deleted but not our structure, crashing in
+         * FreeReflectedChannel().
+         *
+         * Force creation of the RCM, for proper cleanup on thread teardown.
+         * The teardown of unprocessed events is currently coupled to the
+         * thread reflected channel map
+         */
+
+        (void) GetThreadReflectedChannelMap();
+
+        /*
+         * XXX Race condition !!
+         * XXX The destination thread may not exist anymore already.
+         * XXX (Delayed postevent executed after channel got removed).
+         * XXX Can we detect this ? (check the validity of the owner threadid ?)
+         * XXX Actually, in that case the channel should be dead also !
+         */
+
+        Tcl_ThreadQueueEvent(rcPtr->owner, (Tcl_Event *) ev, TCL_QUEUE_TAIL);
+        Tcl_ThreadAlert(rcPtr->owner);
+    }
+#endif
+
+    /*
+     * Squash interp results left by the event script.
+     */
+
+    Tcl_ResetResult(interp);
+    return TCL_OK;
+
+#undef CHAN
+#undef EVENT
+}
+
+/*
+ * Channel error message marshalling utilities.
+ */
+
+static Tcl_Obj *
+MarshallError(
+    Tcl_Interp *interp)
+{
+    /*
+     * Capture the result status of the interpreter into a string. => List of
+     * options and values, followed by the error message. The result has
+     * refCount 0.
+     */
+
+    Tcl_Obj *returnOpt = Tcl_GetReturnOptions(interp, TCL_ERROR);
+
+    /*
+     * => returnOpt.refCount == 0. We can append directly.
+     */
+
+    Tcl_ListObjAppendElement(NULL, returnOpt, Tcl_GetObjResult(interp));
+    return returnOpt;
+}
+
+static void
+UnmarshallErrorResult(
+    Tcl_Interp *interp,
+    Tcl_Obj *msgObj)
+{
+    int lc;
+    Tcl_Obj **lv;
+    int explicitResult;
+    int numOptions;
+
+    /*
+     * Process the caught message.
+     *
+     * Syntax = (option value)... ?message?
+     *
+     * Bad syntax causes a panic. This is OK because the other side uses
+     * Tcl_GetReturnOptions and list construction functions to marshall the
+     * information; if we panic here, something has gone badly wrong already.
+     */
+
+    if (Tcl_ListObjGetElements(interp, msgObj, &lc, &lv) != TCL_OK) {
+	Tcl_Panic("TclChanCaughtErrorBypass: Bad syntax of caught result");
+    }
+    if (interp == NULL) {
+	return;
+    }
+
+    explicitResult = lc & 1;		/* Odd number of values? */
+    numOptions = lc - explicitResult;
+
+    if (explicitResult) {
+	Tcl_SetObjResult(interp, lv[lc-1]);
+    }
+
+    (void) Tcl_SetReturnOptions(interp, Tcl_NewListObj(numOptions, lv));
+    ((Interp *) interp)->flags &= ~ERR_ALREADY_LOGGED;
+}
+
+int
+TclChanCaughtErrorBypass(
+    Tcl_Interp *interp,
+    Tcl_Channel chan)
+{
+    Tcl_Obj *chanMsgObj = NULL;
+    Tcl_Obj *interpMsgObj = NULL;
+    Tcl_Obj *msgObj = NULL;
+
+    /*
+     * Get a bypassed error message from channel and/or interpreter, save the
+     * reference, then kill the returned objects, if there were any. If there
+     * are messages in both the channel has preference.
+     */
+
+    if ((chan == NULL) && (interp == NULL)) {
+	return 0;
+    }
+
+    if (chan != NULL) {
+	Tcl_GetChannelError(chan, &chanMsgObj);
+    }
+    if (interp != NULL) {
+	Tcl_GetChannelErrorInterp(interp, &interpMsgObj);
+    }
+
+    if (chanMsgObj != NULL) {
+	msgObj = chanMsgObj;
+    } else if (interpMsgObj != NULL) {
+	msgObj = interpMsgObj;
+    }
+    if (msgObj != NULL) {
+	Tcl_IncrRefCount(msgObj);
+    }
+
+    if (chanMsgObj != NULL) {
+	Tcl_DecrRefCount(chanMsgObj);
+    }
+    if (interpMsgObj != NULL) {
+	Tcl_DecrRefCount(interpMsgObj);
+    }
+
+    /*
+     * No message returned, nothing caught.
+     */
+
+    if (msgObj == NULL) {
+	return 0;
+    }
+
+    UnmarshallErrorResult(interp, msgObj);
+
+    Tcl_DecrRefCount(msgObj);
+    return 1;
+}
+
+/*
+ * Driver functions. ================================================
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectClose --
+ *
+ *	This function is invoked when the channel is closed, to delete the
+ *	driver specific instance data.
+ *
+ * Results:
+ *	A posix error.
+ *
+ * Side effects:
+ *	Releases memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectClose(
+    ClientData clientData,
+    Tcl_Interp *interp)
+{
+    ReflectedChannel *rcPtr = clientData;
+    int result;			/* Result code for 'close' */
+    Tcl_Obj *resObj;		/* Result data for 'close' */
+    ReflectedChannelMap *rcmPtr;/* Map of reflected channels with handlers in
+				 * this interp */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+    const Tcl_ChannelType *tctPtr;
+
+    if (TclInThreadExit()) {
+	/*
+	 * This call comes from TclFinalizeIOSystem. There are no
+	 * interpreters, and therefore we cannot call upon the handler command
+	 * anymore. Threading is irrelevant as well. We simply clean up all
+	 * our C level data structures and leave the Tcl level to the other
+	 * finalization functions.
+	 */
+
+	/*
+	 * THREADED => Forward this to the origin thread
+	 *
+	 * Note: DeleteThreadReflectedChannelMap() is the thread exit handler
+	 * for the origin thread. Use this to clean up the structure? Except
+	 * if lost?
+	 */
+
+#ifdef TCL_THREADS
+	if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	    ForwardParam p;
+
+	    ForwardOpToHandlerThread(rcPtr, ForwardedClose, &p);
+	    result = p.base.code;
+
+            /*
+             * Now squash the pending reflection events for this channel.
+             */
+
+            Tcl_DeleteEvents(ReflectEventDelete, rcPtr);
+
+	    if (result != TCL_OK) {
+		FreeReceivedError(&p);
+	    }
+	}
+#endif
+
+	tctPtr = ((Channel *)rcPtr->chan)->typePtr;
+	if (tctPtr && tctPtr != &tclRChannelType) {
+	    ckfree(tctPtr);
+	    ((Channel *)rcPtr->chan)->typePtr = NULL;
+	}
+        Tcl_EventuallyFree(rcPtr, (Tcl_FreeProc *) FreeReflectedChannel);
+	return EOK;
+    }
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedClose, &p);
+	result = p.base.code;
+
+        /*
+         * Now squash the pending reflection events for this channel.
+         */
+
+        Tcl_DeleteEvents(ReflectEventDelete, rcPtr);
+
+	if (result != TCL_OK) {
+	    PassReceivedErrorInterp(interp, &p);
+	}
+    } else {
+#endif
+	result = InvokeTclMethod(rcPtr, METH_FINAL, NULL, NULL, &resObj);
+	if ((result != TCL_OK) && (interp != NULL)) {
+	    Tcl_SetChannelErrorInterp(interp, resObj);
+	}
+
+	Tcl_DecrRefCount(resObj);	/* Remove reference we held from the
+					 * invoke */
+
+	/*
+	 * Remove the channel from the map before releasing the memory, to
+	 * prevent future accesses (like by 'postevent') from finding and
+	 * dereferencing a dangling pointer.
+	 *
+	 * NOTE: The channel may not be in the map. This is ok, that happens
+	 * when the channel was created in a different interpreter and/or
+	 * thread and then was moved here.
+	 *
+	 * NOTE: The channel may have been removed from the map already via
+	 * the per-interp DeleteReflectedChannelMap exit-handler.
+	 */
+
+	if (!rcPtr->dead) {
+	    rcmPtr = GetReflectedChannelMap(rcPtr->interp);
+	    hPtr = Tcl_FindHashEntry(&rcmPtr->map,
+		    Tcl_GetChannelName(rcPtr->chan));
+	    if (hPtr) {
+		Tcl_DeleteHashEntry(hPtr);
+	    }
+	}
+#ifdef TCL_THREADS
+	rcmPtr = GetThreadReflectedChannelMap();
+	hPtr = Tcl_FindHashEntry(&rcmPtr->map,
+		Tcl_GetChannelName(rcPtr->chan));
+	if (hPtr) {
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+    }
+#endif
+    tctPtr = ((Channel *)rcPtr->chan)->typePtr;
+    if (tctPtr && tctPtr != &tclRChannelType) {
+	ckfree(tctPtr);
+	((Channel *)rcPtr->chan)->typePtr = NULL;
+    }
+    Tcl_EventuallyFree(rcPtr, (Tcl_FreeProc *) FreeReflectedChannel);
+    return (result == TCL_OK) ? EOK : EINVAL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectInput --
+ *
+ *	This function is invoked when more data is requested from the channel.
+ *
+ * Results:
+ *	The number of bytes read.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectInput(
+    ClientData clientData,
+    char *buf,
+    int toRead,
+    int *errorCodePtr)
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *toReadObj;
+    int bytec;			/* Number of returned bytes */
+    unsigned char *bytev;	/* Array of returned bytes */
+    Tcl_Obj *resObj;		/* Result data for 'read' */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.input.buf = buf;
+	p.input.toRead = toRead;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedInput, &p);
+
+	if (p.base.code != TCL_OK) {
+	    if (p.base.code < 0) {
+		/*
+                 * No error message, this is an errno signal.
+                 */
+
+		*errorCodePtr = -p.base.code;
+	    } else {
+		PassReceivedError(rcPtr->chan, &p);
+		*errorCodePtr = EINVAL;
+	    }
+	    p.input.toRead = -1;
+	} else {
+	    *errorCodePtr = EOK;
+	}
+
+	return p.input.toRead;
+    }
+#endif
+
+    /* ASSERT: rcPtr->method & FLAG(METH_READ) */
+    /* ASSERT: rcPtr->mode & TCL_READABLE */
+
+    Tcl_Preserve(rcPtr);
+
+    toReadObj = Tcl_NewIntObj(toRead);
+    Tcl_IncrRefCount(toReadObj);
+
+    if (InvokeTclMethod(rcPtr, METH_READ, toReadObj, NULL, &resObj)!=TCL_OK) {
+	int code = ErrnoReturn(rcPtr, resObj);
+
+	if (code < 0) {
+	    *errorCodePtr = -code;
+            goto error;
+	}
+
+	Tcl_SetChannelError(rcPtr->chan, resObj);
+        goto invalid;
+    }
+
+    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+    if (toRead < bytec) {
+	SetChannelErrorStr(rcPtr->chan, msg_read_toomuch);
+        goto invalid;
+    }
+
+    *errorCodePtr = EOK;
+
+    if (bytec > 0) {
+	memcpy(buf, bytev, (size_t) bytec);
+    }
+
+ stop:
+    Tcl_DecrRefCount(toReadObj);
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    Tcl_Release(rcPtr);
+    return bytec;
+ invalid:
+    *errorCodePtr = EINVAL;
+ error:
+    bytec = -1;
+    goto stop;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectOutput --
+ *
+ *	This function is invoked when data is writen to the channel.
+ *
+ * Results:
+ *	The number of bytes actually written.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectOutput(
+    ClientData clientData,
+    const char *buf,
+    int toWrite,
+    int *errorCodePtr)
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *bufObj;
+    Tcl_Obj *resObj;		/* Result data for 'write' */
+    int written;
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.output.buf = buf;
+	p.output.toWrite = toWrite;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedOutput, &p);
+
+	if (p.base.code != TCL_OK) {
+	    if (p.base.code < 0) {
+		/*
+                 * No error message, this is an errno signal.
+                 */
+
+		*errorCodePtr = -p.base.code;
+	    } else {
+                PassReceivedError(rcPtr->chan, &p);
+                *errorCodePtr = EINVAL;
+            }
+	    p.output.toWrite = -1;
+	} else {
+	    *errorCodePtr = EOK;
+	}
+
+	return p.output.toWrite;
+    }
+#endif
+
+    /* ASSERT: rcPtr->method & FLAG(METH_WRITE) */
+    /* ASSERT: rcPtr->mode & TCL_WRITABLE */
+
+    Tcl_Preserve(rcPtr);
+    Tcl_Preserve(rcPtr->interp);
+
+    bufObj = Tcl_NewByteArrayObj((unsigned char *) buf, toWrite);
+    Tcl_IncrRefCount(bufObj);
+
+    if (InvokeTclMethod(rcPtr, METH_WRITE, bufObj, NULL, &resObj) != TCL_OK) {
+	int code = ErrnoReturn(rcPtr, resObj);
+
+	if (code < 0) {
+	    *errorCodePtr = -code;
+            goto error;
+	}
+
+	Tcl_SetChannelError(rcPtr->chan, resObj);
+        goto invalid;
+    }
+
+    if (Tcl_InterpDeleted(rcPtr->interp)) {
+	/*
+	 * The interp was destroyed during InvokeTclMethod().
+	 */
+
+	SetChannelErrorStr(rcPtr->chan, msg_send_dstlost);
+        goto invalid;
+    }
+    if (Tcl_GetIntFromObj(rcPtr->interp, resObj, &written) != TCL_OK) {
+	Tcl_SetChannelError(rcPtr->chan, MarshallError(rcPtr->interp));
+        goto invalid;
+    }
+
+    if ((written == 0) && (toWrite > 0)) {
+	/*
+	 * The handler claims to have written nothing of what it was given.
+	 * That is bad.
+	 */
+
+	SetChannelErrorStr(rcPtr->chan, msg_write_nothing);
+        goto invalid;
+    }
+    if (toWrite < written) {
+	/*
+	 * The handler claims to have written more than it was given. That is
+	 * bad. Note that the I/O core would crash if we were to return this
+	 * information, trying to write -nnn bytes in the next iteration.
+	 */
+
+	SetChannelErrorStr(rcPtr->chan, msg_write_toomuch);
+        goto invalid;
+    }
+
+    *errorCodePtr = EOK;
+ stop:
+    Tcl_DecrRefCount(bufObj);
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    Tcl_Release(rcPtr->interp);
+    Tcl_Release(rcPtr);
+    return written;
+ invalid:
+    *errorCodePtr = EINVAL;
+ error:
+    written = -1;
+    goto stop;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectSeekWide / ReflectSeek --
+ *
+ *	This function is invoked when the user wishes to seek on the channel.
+ *
+ * Results:
+ *	The new location of the access point.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_WideInt
+ReflectSeekWide(
+    ClientData clientData,
+    Tcl_WideInt offset,
+    int seekMode,
+    int *errorCodePtr)
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *offObj, *baseObj;
+    Tcl_Obj *resObj;		/* Result for 'seek' */
+    Tcl_WideInt newLoc;
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.seek.seekMode = seekMode;
+	p.seek.offset = offset;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedSeek, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rcPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    p.seek.offset = -1;
+	} else {
+	    *errorCodePtr = EOK;
+	}
+
+	return p.seek.offset;
+    }
+#endif
+
+    /* ASSERT: rcPtr->method & FLAG(METH_SEEK) */
+
+    Tcl_Preserve(rcPtr);
+
+    offObj  = Tcl_NewWideIntObj(offset);
+    baseObj = Tcl_NewStringObj(
+            (seekMode == SEEK_SET) ? "start" :
+            (seekMode == SEEK_CUR) ? "current" : "end", -1);
+    Tcl_IncrRefCount(offObj);
+    Tcl_IncrRefCount(baseObj);
+
+    if (InvokeTclMethod(rcPtr, METH_SEEK, offObj, baseObj, &resObj)!=TCL_OK) {
+	Tcl_SetChannelError(rcPtr->chan, resObj);
+        goto invalid;
+    }
+
+    if (Tcl_GetWideIntFromObj(rcPtr->interp, resObj, &newLoc) != TCL_OK) {
+	Tcl_SetChannelError(rcPtr->chan, MarshallError(rcPtr->interp));
+        goto invalid;
+    }
+
+    if (newLoc < Tcl_LongAsWide(0)) {
+	SetChannelErrorStr(rcPtr->chan, msg_seek_beforestart);
+        goto invalid;
+    }
+
+    *errorCodePtr = EOK;
+ stop:
+    Tcl_DecrRefCount(offObj);
+    Tcl_DecrRefCount(baseObj);
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    Tcl_Release(rcPtr);
+    return newLoc;
+ invalid:
+    *errorCodePtr = EINVAL;
+    newLoc = -1;
+    goto stop;
+}
+
+static int
+ReflectSeek(
+    ClientData clientData,
+    long offset,
+    int seekMode,
+    int *errorCodePtr)
+{
+    /*
+     * This function can be invoked from a transformation which is based on
+     * standard seeking, i.e. non-wide. Because of this we have to implement
+     * it, a dummy is not enough. We simply delegate the call to the wide
+     * routine.
+     */
+
+    return (int) ReflectSeekWide(clientData, Tcl_LongAsWide(offset), seekMode,
+	    errorCodePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectWatch --
+ *
+ *	This function is invoked to tell the channel what events the I/O
+ *	system is interested in.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReflectWatch(
+    ClientData clientData,
+    int mask)
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *maskObj;
+
+    /*
+     * We restrict the interest to what the channel can support. IOW there
+     * will never be write events for a channel which is not writable.
+     * Analoguously for read events and non-readable channels.
+     */
+
+    mask &= rcPtr->mode;
+
+    if (mask == rcPtr->interest) {
+	/*
+	 * Same old, same old, why should we do something?
+	 */
+
+	return;
+    }
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.watch.mask = mask;
+	ForwardOpToHandlerThread(rcPtr, ForwardedWatch, &p);
+
+	/*
+	 * Any failure from the forward is ignored. We have no place to put
+	 * this.
+	 */
+
+	return;
+    }
+#endif
+
+    Tcl_Preserve(rcPtr);
+
+    rcPtr->interest = mask;
+    maskObj = DecodeEventMask(mask);
+    /* assert maskObj.refCount == 1 */
+    (void) InvokeTclMethod(rcPtr, METH_WATCH, maskObj, NULL, NULL);
+    Tcl_DecrRefCount(maskObj);
+
+    Tcl_Release(rcPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectBlock --
+ *
+ *	This function is invoked to tell the channel which blocking behaviour
+ *	is required of it.
+ *
+ * Results:
+ *	A posix error number.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectBlock(
+    ClientData clientData,
+    int nonblocking)
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *blockObj;
+    int errorNum;		/* EINVAL or EOK (success). */
+    Tcl_Obj *resObj;		/* Result data for 'blocking' */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.block.nonblocking = nonblocking;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedBlock, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rcPtr->chan, &p);
+	    return EINVAL;
+	}
+
+	return EOK;
+    }
+#endif
+
+    blockObj = Tcl_NewBooleanObj(!nonblocking);
+    Tcl_IncrRefCount(blockObj);
+
+    Tcl_Preserve(rcPtr);
+
+    if (InvokeTclMethod(rcPtr,METH_BLOCKING,blockObj,NULL,&resObj)!=TCL_OK) {
+	Tcl_SetChannelError(rcPtr->chan, resObj);
+	errorNum = EINVAL;
+    } else {
+	errorNum = EOK;
+    }
+
+    Tcl_DecrRefCount(blockObj);
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+
+    Tcl_Release(rcPtr);
+    return errorNum;
+}
+
+#ifdef TCL_THREADS
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectThread --
+ *
+ *	This function is invoked to tell the channel about thread movements.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReflectThread(
+    ClientData clientData,
+    int action)
+{
+    ReflectedChannel *rcPtr = clientData;
+
+    switch (action) {
+    case TCL_CHANNEL_THREAD_INSERT:
+        rcPtr->owner = Tcl_GetCurrentThread();
+        break;
+    case TCL_CHANNEL_THREAD_REMOVE:
+        rcPtr->owner = NULL;
+        break;
+    default:
+        Tcl_Panic("Unknown thread action code.");
+        break;
+    }
+}
+
+#endif
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectSetOption --
+ *
+ *	This function is invoked to configure a channel option.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, as it calls upon a Tcl script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectSetOption(
+    ClientData clientData,	/* Channel to query */
+    Tcl_Interp *interp,		/* Interpreter to leave error messages in */
+    const char *optionName,	/* Name of requested option */
+    const char *newValue)	/* The new value */
+{
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *optionObj, *valueObj;
+    int result;			/* Result code for 'configure' */
+    Tcl_Obj *resObj;		/* Result data for 'configure' */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.setOpt.name = optionName;
+	p.setOpt.value = newValue;
+
+	ForwardOpToHandlerThread(rcPtr, ForwardedSetOpt, &p);
+
+	if (p.base.code != TCL_OK) {
+	    Tcl_Obj *err = Tcl_NewStringObj(p.base.msgStr, -1);
+
+	    UnmarshallErrorResult(interp, err);
+	    Tcl_DecrRefCount(err);
+	    FreeReceivedError(&p);
+	}
+
+	return p.base.code;
+    }
+#endif
+    Tcl_Preserve(rcPtr);
+
+    optionObj = Tcl_NewStringObj(optionName, -1);
+    valueObj = Tcl_NewStringObj(newValue, -1);
+
+    Tcl_IncrRefCount(optionObj);
+    Tcl_IncrRefCount(valueObj);
+
+    result = InvokeTclMethod(rcPtr, METH_CONFIGURE,optionObj,valueObj, &resObj);
+    if (result != TCL_OK) {
+	UnmarshallErrorResult(interp, resObj);
+    }
+
+    Tcl_DecrRefCount(optionObj);
+    Tcl_DecrRefCount(valueObj);
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    Tcl_Release(rcPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectGetOption --
+ *
+ *	This function is invoked to retrieve all or a channel option.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, as it calls upon a Tcl script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectGetOption(
+    ClientData clientData,	/* Channel to query */
+    Tcl_Interp *interp,		/* Interpreter to leave error messages in */
+    const char *optionName,	/* Name of reuqested option */
+    Tcl_DString *dsPtr)		/* String to place the result into */
+{
+    /*
+     * This code is special. It has regular passing of Tcl result, and errors.
+     * The bypass functions are not required.
+     */
+
+    ReflectedChannel *rcPtr = clientData;
+    Tcl_Obj *optionObj;
+    Tcl_Obj *resObj;		/* Result data for 'configure' */
+    int listc, result = TCL_OK;
+    Tcl_Obj **listv;
+    MethodName method;
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rcPtr->thread != Tcl_GetCurrentThread()) {
+	int opcode;
+	ForwardParam p;
+
+	p.getOpt.name = optionName;
+	p.getOpt.value = dsPtr;
+
+	if (optionName == NULL) {
+	    opcode = ForwardedGetOptAll;
+	} else {
+	    opcode = ForwardedGetOpt;
+	}
+
+	ForwardOpToHandlerThread(rcPtr, opcode, &p);
+
+	if (p.base.code != TCL_OK) {
+	    Tcl_Obj *err = Tcl_NewStringObj(p.base.msgStr, -1);
+
+	    UnmarshallErrorResult(interp, err);
+	    Tcl_DecrRefCount(err);
+	    FreeReceivedError(&p);
+	}
+
+	return p.base.code;
+    }
+#endif
+
+    if (optionName == NULL) {
+	/*
+	 * Retrieve all options.
+	 */
+
+	method = METH_CGETALL;
+	optionObj = NULL;
+    } else {
+	/*
+	 * Retrieve the value of one option.
+	 */
+
+	method = METH_CGET;
+	optionObj = Tcl_NewStringObj(optionName, -1);
+        Tcl_IncrRefCount(optionObj);
+    }
+
+    Tcl_Preserve(rcPtr);
+
+    if (InvokeTclMethod(rcPtr, method, optionObj, NULL, &resObj)!=TCL_OK) {
+	UnmarshallErrorResult(interp, resObj);
+        goto error;
+    }
+
+    /*
+     * The result has to go into the 'dsPtr' for propagation to the caller of
+     * the driver.
+     */
+
+    if (optionObj != NULL) {
+	TclDStringAppendObj(dsPtr, resObj);
+        goto ok;
+    }
+
+    /*
+     * Extract the list and append each item as element.
+     */
+
+    /*
+     * NOTE (4): If we extract the string rep we can assume a properly quoted
+     * string. Together with a separating space this way of simply appending
+     * the whole string rep might be faster. It also doesn't check if the
+     * result is a valid list. Nor that the list has an even number elements.
+     */
+
+    if (Tcl_ListObjGetElements(interp, resObj, &listc, &listv) != TCL_OK) {
+        goto error;
+    }
+
+    if ((listc % 2) == 1) {
+	/*
+	 * Odd number of elements is wrong.
+	 */
+
+	Tcl_ResetResult(interp);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"Expected list with even number of "
+		"elements, got %d element%s instead", listc,
+		(listc == 1 ? "" : "s")));
+        goto error;
+    } else {
+	int len;
+	const char *str = TclGetStringFromObj(resObj, &len);
+
+	if (len) {
+	    TclDStringAppendLiteral(dsPtr, " ");
+	    Tcl_DStringAppend(dsPtr, str, len);
+	}
+        goto ok;
+    }
+
+ ok:
+    result = TCL_OK;
+ stop:
+    if (optionObj) {
+        Tcl_DecrRefCount(optionObj);
+    }
+    Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+    Tcl_Release(rcPtr);
+    return result;
+ error:
+    result = TCL_ERROR;
+    goto stop;
+}
+
+/*
+ * Helpers. =========================================================
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EncodeEventMask --
+ *
+ *	This function takes a list of event items and constructs the
+ *	equivalent internal bitmask. The list must contain at least one
+ *	element. Elements are "read", "write", or any unique abbreviation of
+ *	them. Note that the bitmask is not changed if problems are
+ *	encountered.
+ *
+ * Results:
+ *	A standard Tcl error code. A bitmask where TCL_READABLE and/or
+ *	TCL_WRITABLE can be set.
+ *
+ * Side effects:
+ *	May shimmer 'obj' to a list representation. May place an error message
+ *	into the interp result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+EncodeEventMask(
+    Tcl_Interp *interp,
+    const char *objName,
+    Tcl_Obj *obj,
+    int *mask)
+{
+    int events;			/* Mask of events to post */
+    int listc;			/* #elements in eventspec list */
+    Tcl_Obj **listv;		/* Elements of eventspec list */
+    int evIndex;		/* Id of event for an element of the eventspec
+				 * list. */
+
+    if (Tcl_ListObjGetElements(interp, obj, &listc, &listv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (listc < 1) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "bad %s list: is empty", objName));
+	return TCL_ERROR;
+    }
+
+    events = 0;
+    while (listc > 0) {
+	if (Tcl_GetIndexFromObj(interp, listv[listc-1], eventOptions,
+		objName, 0, &evIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch (evIndex) {
+	case EVENT_READ:
+	    events |= TCL_READABLE;
+	    break;
+	case EVENT_WRITE:
+	    events |= TCL_WRITABLE;
+	    break;
+	}
+	listc --;
+    }
+
+    *mask = events;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DecodeEventMask --
+ *
+ *	This function takes an internal bitmask of events and constructs the
+ *	equivalent list of event items.
+ *
+ * Results, Contract:
+ *	A Tcl_Obj reference. The object will have a refCount of one. The user
+ *	has to decrement it to release the object.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+DecodeEventMask(
+    int mask)
+{
+    register const char *eventStr;
+    Tcl_Obj *evObj;
+
+    switch (mask & RANDW) {
+    case RANDW:
+	eventStr = "read write";
+	break;
+    case TCL_READABLE:
+	eventStr = "read";
+	break;
+    case TCL_WRITABLE:
+	eventStr = "write";
+	break;
+    default:
+	eventStr = "";
+	break;
+    }
+
+    evObj = Tcl_NewStringObj(eventStr, -1);
+    Tcl_IncrRefCount(evObj);
+    /* assert evObj.refCount == 1 */
+    return evObj;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NewReflectedChannel --
+ *
+ *	This function is invoked to allocate and initialize the instance data
+ *	of a new reflected channel.
+ *
+ * Results:
+ *	A heap-allocated channel instance.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedChannel *
+NewReflectedChannel(
+    Tcl_Interp *interp,
+    Tcl_Obj *cmdpfxObj,
+    int mode,
+    Tcl_Obj *handleObj)
+{
+    ReflectedChannel *rcPtr;
+    MethodName mn = METH_BLOCKING;
+
+    rcPtr = ckalloc(sizeof(ReflectedChannel));
+
+    /* rcPtr->chan: Assigned by caller. Dummy data here. */
+
+    rcPtr->chan = NULL;
+    rcPtr->interp = interp;
+    rcPtr->dead = 0;
+#ifdef TCL_THREADS
+    rcPtr->thread = Tcl_GetCurrentThread();
+#endif
+    rcPtr->mode = mode;
+    rcPtr->interest = 0;		/* Initially no interest registered */
+
+    /* ASSERT: cmdpfxObj is a Tcl List */
+    rcPtr->cmd = TclListObjCopy(NULL, cmdpfxObj);
+    Tcl_IncrRefCount(rcPtr->cmd);
+    rcPtr->methods = Tcl_NewListObj(METH_WRITE + 1, NULL);
+    while (mn <= METH_WRITE) {
+	Tcl_ListObjAppendElement(NULL, rcPtr->methods,
+		Tcl_NewStringObj(methodNames[mn++], -1));
+    }
+    Tcl_IncrRefCount(rcPtr->methods);
+    rcPtr->name = handleObj;
+    Tcl_IncrRefCount(rcPtr->name);
+    return rcPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NextHandle --
+ *
+ *	This function is invoked to generate a channel handle for a new
+ *	reflected channel.
+ *
+ * Results:
+ *	A Tcl_Obj containing the string of the new channel handle. The
+ *	refcount of the returned object is -- zero --.
+ *
+ * Side effects:
+ *	May allocate memory. Mutex protected critical section locks out other
+ *	threads for a short time.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+NextHandle(void)
+{
+    /*
+     * Count number of generated reflected channels. Used for id generation.
+     * Ids are never reclaimed and there is no dealing with wrap around. On
+     * the other hand, "unsigned long" should be big enough except for
+     * absolute longrunners (generate a 100 ids per second => overflow will
+     * occur in 1 1/3 years).
+     */
+
+    TCL_DECLARE_MUTEX(rcCounterMutex)
+    static unsigned long rcCounter = 0;
+    Tcl_Obj *resObj;
+
+    Tcl_MutexLock(&rcCounterMutex);
+    resObj = Tcl_ObjPrintf("rc%lu", rcCounter);
+    rcCounter++;
+    Tcl_MutexUnlock(&rcCounterMutex);
+
+    return resObj;
+}
+
+static void
+FreeReflectedChannel(
+    ReflectedChannel *rcPtr)
+{
+    Channel *chanPtr = (Channel *) rcPtr->chan;
+
+    TclChannelRelease((Tcl_Channel)chanPtr);
+    if (rcPtr->name) {
+	Tcl_DecrRefCount(rcPtr->name);
+    }
+    if (rcPtr->methods) {
+	Tcl_DecrRefCount(rcPtr->methods);
+    }
+    if (rcPtr->cmd) {
+	Tcl_DecrRefCount(rcPtr->cmd);
+    }
+    ckfree(rcPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InvokeTclMethod --
+ *
+ *	This function is used to invoke the Tcl level of a reflected channel.
+ *	It handles all the command assembly, invokation, and generic state and
+ *	result mgmt. It does *not* handle thread redirection; that is the
+ *	responsibility of clients of this function.
+ *
+ * Results:
+ *	Result code and data as returned by the method.
+ *
+ * Side effects:
+ *	Arbitrary, as it calls upon a Tcl script.
+ *
+ * Contract:
+ *	argOneObj.refCount >= 1 on entry and exit, if argOneObj != NULL
+ *	argTwoObj.refCount >= 1 on entry and exit, if argTwoObj != NULL
+ *	resObj.refCount in {0, 1, ...}
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InvokeTclMethod(
+    ReflectedChannel *rcPtr,
+    MethodName method,
+    Tcl_Obj *argOneObj,		/* NULL'able */
+    Tcl_Obj *argTwoObj,		/* NULL'able */
+    Tcl_Obj **resultObjPtr)	/* NULL'able */
+{
+    Tcl_Obj *methObj = NULL;	/* Method name in object form */
+    Tcl_InterpState sr;		/* State of handler interp */
+    int result;			/* Result code of method invokation */
+    Tcl_Obj *resObj = NULL;	/* Result of method invokation. */
+    Tcl_Obj *cmd;
+
+    if (rcPtr->dead) {
+	/*
+	 * The channel is marked as dead. Bail out immediately, with an
+	 * appropriate error.
+	 */
+
+	if (resultObjPtr != NULL) {
+	    resObj = Tcl_NewStringObj(msg_dstlost,-1);
+	    *resultObjPtr = resObj;
+	    Tcl_IncrRefCount(resObj);
+	}
+
+        /*
+         * Not touching argOneObj, argTwoObj, they have not been used.
+         * See the contract as well.
+         */
+
+	return TCL_ERROR;
+    }
+
+    /*
+     * Insert method into the callback command, after the command prefix,
+     * before the channel id.
+     */
+
+    cmd = TclListObjCopy(NULL, rcPtr->cmd);
+
+    Tcl_ListObjIndex(NULL, rcPtr->methods, method, &methObj);
+    Tcl_ListObjAppendElement(NULL, cmd, methObj);
+    Tcl_ListObjAppendElement(NULL, cmd, rcPtr->name);
+
+    /*
+     * Append the additional argument containing method specific details
+     * behind the channel id. If specified.
+     *
+     * Because of the contract there is no need to increment the refcounts.
+     * The objects will survive the Tcl_EvalObjv without change.
+     */
+
+    if (argOneObj) {
+	Tcl_ListObjAppendElement(NULL, cmd, argOneObj);
+	if (argTwoObj) {
+	    Tcl_ListObjAppendElement(NULL, cmd, argTwoObj);
+	}
+    }
+
+    /*
+     * And run the handler... This is done in auch a manner which leaves any
+     * existing state intact.
+     */
+
+    Tcl_IncrRefCount(cmd);
+    sr = Tcl_SaveInterpState(rcPtr->interp, 0 /* Dummy */);
+    Tcl_Preserve(rcPtr->interp);
+    result = Tcl_EvalObjEx(rcPtr->interp, cmd, TCL_EVAL_GLOBAL);
+
+    /*
+     * We do not try to extract the result information if the caller has no
+     * interest in it. I.e. there is no need to put effort into creating
+     * something which is discarded immediately after.
+     */
+
+    if (resultObjPtr) {
+	if (result == TCL_OK) {
+	    /*
+	     * Ok result taken as is, also if the caller requests that there
+	     * is no capture.
+	     */
+
+	    resObj = Tcl_GetObjResult(rcPtr->interp);
+	} else {
+	    /*
+	     * Non-ok result is always treated as an error. We have to capture
+	     * the full state of the result, including additional options.
+	     *
+	     * This is complex and ugly, and would be completely unnecessary
+	     * if we only added support for a TCL_FORBID_EXCEPTIONS flag.
+	     */
+
+	    if (result != TCL_ERROR) {
+		int cmdLen;
+		const char *cmdString = TclGetStringFromObj(cmd, &cmdLen);
+
+		Tcl_IncrRefCount(cmd);
+		Tcl_ResetResult(rcPtr->interp);
+		Tcl_SetObjResult(rcPtr->interp, Tcl_ObjPrintf(
+			"chan handler returned bad code: %d", result));
+		Tcl_LogCommandInfo(rcPtr->interp, cmdString, cmdString,
+			cmdLen);
+		Tcl_DecrRefCount(cmd);
+		result = TCL_ERROR;
+	    }
+	    Tcl_AppendObjToErrorInfo(rcPtr->interp, Tcl_ObjPrintf(
+		    "\n    (chan handler subcommand \"%s\")",
+		    methodNames[method]));
+	    resObj = MarshallError(rcPtr->interp);
+	}
+	Tcl_IncrRefCount(resObj);
+    }
+    Tcl_DecrRefCount(cmd);
+    Tcl_RestoreInterpState(rcPtr->interp, sr);
+    Tcl_Release(rcPtr->interp);
+
+    /*
+     * The resObj has a ref count of 1 at this location. This means that the
+     * caller of InvokeTclMethod has to dispose of it (but only if it was
+     * returned to it).
+     */
+
+    if (resultObjPtr != NULL) {
+	*resultObjPtr = resObj;
+    }
+
+    /*
+     * There no need to handle the case where nothing is returned, because for
+     * that case resObj was not set anyway.
+     */
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ErrnoReturn --
+ *
+ *	Checks a method error result if it returned an 'errno'.
+ *
+ * Results:
+ *	The negative errno found in the error result, or 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Users:
+ *	ReflectInput/Output(), to enable the signaling of EAGAIN on 0-sized
+ *	short reads/writes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ErrnoReturn(
+    ReflectedChannel *rcPtr,
+    Tcl_Obj *resObj)
+{
+    int code;
+    Tcl_InterpState sr;		/* State of handler interp */
+
+    if (rcPtr->dead) {
+	return 0;
+    }
+
+    sr = Tcl_SaveInterpState(rcPtr->interp, 0 /* Dummy */);
+    UnmarshallErrorResult(rcPtr->interp, resObj);
+
+    resObj = Tcl_GetObjResult(rcPtr->interp);
+
+    if (((Tcl_GetIntFromObj(rcPtr->interp, resObj, &code) != TCL_OK)
+	    || (code >= 0))) {
+	if (strcmp("EAGAIN", TclGetString(resObj)) == 0) {
+	    code = -EAGAIN;
+	} else {
+	    code = 0;
+	}
+    }
+
+    Tcl_RestoreInterpState(rcPtr->interp, sr);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetReflectedChannelMap --
+ *
+ *	Gets and potentially initializes the reflected channel map for an
+ *	interpreter.
+ *
+ * Results:
+ *	A pointer to the map created, for use by the caller.
+ *
+ * Side effects:
+ *	Initializes the reflected channel map for an interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedChannelMap *
+GetReflectedChannelMap(
+    Tcl_Interp *interp)
+{
+    ReflectedChannelMap *rcmPtr = Tcl_GetAssocData(interp, RCMKEY, NULL);
+
+    if (rcmPtr == NULL) {
+	rcmPtr = ckalloc(sizeof(ReflectedChannelMap));
+	Tcl_InitHashTable(&rcmPtr->map, TCL_STRING_KEYS);
+	Tcl_SetAssocData(interp, RCMKEY,
+		(Tcl_InterpDeleteProc *) DeleteReflectedChannelMap, rcmPtr);
+    }
+    return rcmPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteReflectedChannelMap --
+ *
+ *	Deletes the channel table for an interpreter, closing any open
+ *	channels whose refcount reaches zero. This procedure is invoked when
+ *	an interpreter is deleted, via the AssocData cleanup mechanism.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the hash table of channels. May close channels. May flush
+ *	output on closed channels. Removes any channeEvent handlers that were
+ *	registered in this interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MarkDead(
+    ReflectedChannel *rcPtr)
+{
+    if (rcPtr->dead) {
+	return;
+    }
+    if (rcPtr->name) {
+	Tcl_DecrRefCount(rcPtr->name);
+	rcPtr->name = NULL;
+    }
+    if (rcPtr->methods) {
+	Tcl_DecrRefCount(rcPtr->methods);
+	rcPtr->methods = NULL;
+    }
+    if (rcPtr->cmd) {
+	Tcl_DecrRefCount(rcPtr->cmd);
+	rcPtr->cmd = NULL;
+    }
+    rcPtr->dead = 1;
+}
+
+static void
+DeleteReflectedChannelMap(
+    ClientData clientData,	/* The per-interpreter data structure. */
+    Tcl_Interp *interp)		/* The interpreter being deleted. */
+{
+    ReflectedChannelMap *rcmPtr = clientData;
+				/* The map */
+    Tcl_HashSearch hSearch;	 /* Search variable. */
+    Tcl_HashEntry *hPtr;	 /* Search variable. */
+    ReflectedChannel *rcPtr;
+    Tcl_Channel chan;
+#ifdef TCL_THREADS
+    ForwardingResult *resultPtr;
+    ForwardingEvent *evPtr;
+    ForwardParam *paramPtr;
+#endif
+
+    /*
+     * Delete all entries. The channels may have been closed already, or will
+     * be closed later, by the standard IO finalization of an interpreter
+     * under destruction. Except for the channels which were moved to a
+     * different interpreter and/or thread. They do not exist from the IO
+     * systems point of view and will not get closed. Therefore mark all as
+     * dead so that any future access will cause a proper error. For channels
+     * in a different thread we actually do the same as
+     * DeleteThreadReflectedChannelMap(), just restricted to the channels of
+     * this interp.
+     */
+
+    for (hPtr = Tcl_FirstHashEntry(&rcmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_FirstHashEntry(&rcmPtr->map, &hSearch)) {
+	chan = Tcl_GetHashValue(hPtr);
+	rcPtr = Tcl_GetChannelInstanceData(chan);
+
+	MarkDead(rcPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    Tcl_DeleteHashTable(&rcmPtr->map);
+    ckfree(&rcmPtr->map);
+
+#ifdef TCL_THREADS
+    /*
+     * The origin interpreter for one or more reflected channels is gone.
+     */
+
+    /*
+     * Go through the list of pending results and cancel all whose events were
+     * destined for this interpreter. While this is in progress we block any
+     * other access to the list of pending results.
+     */
+
+    Tcl_MutexLock(&rcForwardMutex);
+
+    for (resultPtr = forwardList;
+	    resultPtr != NULL;
+	    resultPtr = resultPtr->nextPtr) {
+	if (resultPtr->dsti != interp) {
+	    /*
+	     * Ignore results/events for other interpreters.
+	     */
+
+	    continue;
+	}
+
+	/*
+	 * The receiver for the event exited, before processing the event. We
+	 * detach the result now, wake the originator up and signal failure.
+         *
+         * Attention: Results may have been detached already, by either the
+         * receiver, or this thread, as part of other parts in the thread
+         * teardown. Such results are ignored. See ticket [b47b176adf] for the
+         * identical race condition in Tcl 8.6 IORTrans.
+	 */
+
+	evPtr = resultPtr->evPtr;
+
+	/*
+         * Basic crash safety until this routine can get revised [3411310]
+         */
+
+	if (evPtr == NULL) {
+	    continue;
+	}
+	paramPtr = evPtr->param;
+	if (!evPtr) {
+	    continue;
+	}
+
+	evPtr->resultPtr = NULL;
+	resultPtr->evPtr = NULL;
+	resultPtr->result = TCL_ERROR;
+
+	ForwardSetStaticError(paramPtr, msg_send_dstlost);
+
+	Tcl_ConditionNotify(&resultPtr->done);
+    }
+    Tcl_MutexUnlock(&rcForwardMutex);
+
+    /*
+     * Get the map of all channels handled by the current thread. This is a
+     * ReflectedChannelMap, but on a per-thread basis, not per-interp. Go
+     * through the channels and remove all which were handled by this
+     * interpreter. They have already been marked as dead.
+     */
+
+    rcmPtr = GetThreadReflectedChannelMap();
+    for (hPtr = Tcl_FirstHashEntry(&rcmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&hSearch)) {
+	chan = Tcl_GetHashValue(hPtr);
+	rcPtr = Tcl_GetChannelInstanceData(chan);
+
+	if (rcPtr->interp != interp) {
+	    /*
+	     * Ignore entries for other interpreters.
+	     */
+
+	    continue;
+	}
+
+	MarkDead(rcPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+#endif
+}
+
+#ifdef TCL_THREADS
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetThreadReflectedChannelMap --
+ *
+ *	Gets and potentially initializes the reflected channel map for a
+ *	thread.
+ *
+ * Results:
+ *	A pointer to the map created, for use by the caller.
+ *
+ * Side effects:
+ *	Initializes the reflected channel map for a thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedChannelMap *
+GetThreadReflectedChannelMap(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (!tsdPtr->rcmPtr) {
+	tsdPtr->rcmPtr = ckalloc(sizeof(ReflectedChannelMap));
+	Tcl_InitHashTable(&tsdPtr->rcmPtr->map, TCL_STRING_KEYS);
+	Tcl_CreateThreadExitHandler(DeleteThreadReflectedChannelMap, NULL);
+    }
+
+    return tsdPtr->rcmPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteThreadReflectedChannelMap --
+ *
+ *	Deletes the channel table for a thread. This procedure is invoked when
+ *	a thread is deleted. The channels have already been marked as dead, in
+ *	DeleteReflectedChannelMap().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the hash table of channels.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteThreadReflectedChannelMap(
+    ClientData clientData)	/* The per-thread data structure. */
+{
+    Tcl_HashSearch hSearch;	 /* Search variable. */
+    Tcl_HashEntry *hPtr;	 /* Search variable. */
+    Tcl_ThreadId self = Tcl_GetCurrentThread();
+    ReflectedChannelMap *rcmPtr; /* The map */
+    ForwardingResult *resultPtr;
+
+    /*
+     * The origin thread for one or more reflected channels is gone.
+     * NOTE: If this function is called due to a thread getting killed the
+     *       per-interp DeleteReflectedChannelMap is apparently not called.
+     */
+
+    /*
+     * Go through the list of pending results and cancel all whose events were
+     * destined for this thread. While this is in progress we block any other
+     * access to the list of pending results.
+     */
+
+    Tcl_MutexLock(&rcForwardMutex);
+
+    for (resultPtr = forwardList;
+	    resultPtr != NULL;
+	    resultPtr = resultPtr->nextPtr) {
+	ForwardingEvent *evPtr;
+	ForwardParam *paramPtr;
+
+	if (resultPtr->dst != self) {
+	    /*
+	     * Ignore results/events for other threads.
+	     */
+
+	    continue;
+	}
+
+	/*
+	 * The receiver for the event exited, before processing the event. We
+	 * detach the result now, wake the originator up and signal failure.
+         *
+         * Attention: Results may have been detached already, by either the
+         * receiver, or this thread, as part of other parts in the thread
+         * teardown. Such results are ignored. See ticket [b47b176adf] for the
+         * identical race condition in Tcl 8.6 IORTrans.
+	 */
+
+	evPtr = resultPtr->evPtr;
+
+	/*
+         * Basic crash safety until this routine can get revised [3411310]
+         */
+
+	if (evPtr == NULL ) {
+	    continue;
+	}
+	paramPtr = evPtr->param;
+	if (!evPtr) {
+	    continue;
+	}
+
+	evPtr->resultPtr = NULL;
+	resultPtr->evPtr = NULL;
+	resultPtr->result = TCL_ERROR;
+
+	ForwardSetStaticError(paramPtr, msg_send_dstlost);
+
+	Tcl_ConditionNotify(&resultPtr->done);
+    }
+    Tcl_MutexUnlock(&rcForwardMutex);
+
+    /*
+     * Run over the event queue of this thread and remove all ReflectEvent's
+     * still pending. These are inbound events for reflected channels this
+     * thread owns but doesn't handle. The inverse of the channel map
+     * actually.
+     */
+
+    Tcl_DeleteEvents(ReflectEventDelete, NULL);
+
+    /*
+     * Get the map of all channels handled by the current thread. This is a
+     * ReflectedChannelMap, but on a per-thread basis, not per-interp. Go
+     * through the channels, remove all, mark them as dead.
+     */
+
+    rcmPtr = GetThreadReflectedChannelMap();
+    for (hPtr = Tcl_FirstHashEntry(&rcmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_FirstHashEntry(&rcmPtr->map, &hSearch)) {
+	Tcl_Channel chan = Tcl_GetHashValue(hPtr);
+	ReflectedChannel *rcPtr = Tcl_GetChannelInstanceData(chan);
+
+	MarkDead(rcPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    ckfree(rcmPtr);
+}
+
+static void
+ForwardOpToHandlerThread(
+    ReflectedChannel *rcPtr,	/* Channel instance */
+    ForwardedOperation op,	/* Forwarded driver operation */
+    const void *param)		/* Arguments */
+{
+    /*
+     * Core of the communication from OWNER to HANDLER thread. The receiver is
+     * ForwardProc() below.
+     */
+
+    Tcl_ThreadId dst = rcPtr->thread;
+    ForwardingEvent *evPtr;
+    ForwardingResult *resultPtr;
+
+    /*
+     * We gather the lock early. This allows us to check the liveness of the
+     * channel without interference from DeleteThreadReflectedChannelMap().
+     */
+
+    Tcl_MutexLock(&rcForwardMutex);
+
+    if (rcPtr->dead) {
+	/*
+	 * The channel is marked as dead. Bail out immediately, with an
+	 * appropriate error. Do not forget to unlock the mutex on this path.
+	 */
+
+	ForwardSetStaticError((ForwardParam *) param, msg_send_dstlost);
+	Tcl_MutexUnlock(&rcForwardMutex);
+	return;
+    }
+
+    /*
+     * Create and initialize the event and data structures.
+     */
+
+    evPtr = ckalloc(sizeof(ForwardingEvent));
+    resultPtr = ckalloc(sizeof(ForwardingResult));
+
+    evPtr->event.proc = ForwardProc;
+    evPtr->resultPtr = resultPtr;
+    evPtr->op = op;
+    evPtr->rcPtr = rcPtr;
+    evPtr->param = (ForwardParam *) param;
+
+    resultPtr->src = Tcl_GetCurrentThread();
+    resultPtr->dst = dst;
+    resultPtr->dsti = rcPtr->interp;
+    resultPtr->done = NULL;
+    resultPtr->result = -1;
+    resultPtr->evPtr = evPtr;
+
+    /*
+     * Now execute the forward.
+     */
+
+    TclSpliceIn(resultPtr, forwardList);
+
+    /*
+     * Do not unlock here. That is done by the ConditionWait.
+     */
+
+    /*
+     * Ensure cleanup of the event if the origin thread exits while this event
+     * is pending or in progress. Exit of the destination thread is handled by
+     * DeleteThreadReflectedChannelMap(), this is set up by
+     * GetThreadReflectedChannelMap(). This is what we use the 'forwardList'
+     * (see above) for.
+     */
+
+    Tcl_CreateThreadExitHandler(SrcExitProc, evPtr);
+
+    /*
+     * Queue the event and poke the other thread's notifier.
+     */
+
+    Tcl_ThreadQueueEvent(dst, (Tcl_Event *) evPtr, TCL_QUEUE_TAIL);
+    Tcl_ThreadAlert(dst);
+
+    /*
+     * (*) Block until the handler thread has either processed the transfer or
+     * rejected it.
+     */
+
+    while (resultPtr->result < 0) {
+	/*
+	 * NOTE (1): Is it possible that the current thread goes away while
+	 * waiting here? IOW Is it possible that "SrcExitProc" is called while
+	 * we are here? See complementary note (2) in "SrcExitProc"
+	 *
+	 * The ConditionWait unlocks the mutex during the wait and relocks it
+	 * immediately after.
+	 */
+
+	Tcl_ConditionWait(&resultPtr->done, &rcForwardMutex, NULL);
+    }
+
+    /*
+     * Unlink result from the forwarder list. No need to lock. Either still
+     * locked, or locked by the ConditionWait
+     */
+
+    TclSpliceOut(resultPtr, forwardList);
+
+    resultPtr->nextPtr = NULL;
+    resultPtr->prevPtr = NULL;
+
+    Tcl_MutexUnlock(&rcForwardMutex);
+    Tcl_ConditionFinalize(&resultPtr->done);
+
+    /*
+     * Kill the cleanup handler now, and the result structure as well, before
+     * returning the success code.
+     *
+     * Note: The event structure has already been deleted.
+     */
+
+    Tcl_DeleteThreadExitHandler(SrcExitProc, evPtr);
+
+    ckfree(resultPtr);
+}
+
+static int
+ForwardProc(
+    Tcl_Event *evGPtr,
+    int mask)
+{
+    /*
+     * HANDLER thread.
+
+     * The receiver part for the operations coming from the OWNER thread.
+     * See ForwardOpToHandlerThread() for the transmitter.
+     *
+     * Notes regarding access to the referenced data.
+     *
+     * In principle the data belongs to the originating thread (see
+     * evPtr->src), however this thread is currently blocked at (*), i.e.,
+     * quiescent. Because of this we can treat the data as belonging to us,
+     * without fear of race conditions. I.e. we can read and write as we like.
+     *
+     * The only thing we cannot be sure of is the resultPtr. This can be be
+     * NULLed if the originating thread went away while the event is handled
+     * here now.
+     */
+
+    ForwardingEvent *evPtr = (ForwardingEvent *) evGPtr;
+    ForwardingResult *resultPtr = evPtr->resultPtr;
+    ReflectedChannel *rcPtr = evPtr->rcPtr;
+    Tcl_Interp *interp = rcPtr->interp;
+    ForwardParam *paramPtr = evPtr->param;
+    Tcl_Obj *resObj = NULL;	/* Interp result of InvokeTclMethod */
+    ReflectedChannelMap *rcmPtr;/* Map of reflected channels with handlers in
+                                 * this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+
+    /*
+     * Ignore the event if no one is waiting for its result anymore.
+     */
+
+    if (!resultPtr) {
+	return 1;
+    }
+
+    paramPtr->base.code = TCL_OK;
+    paramPtr->base.msgStr = NULL;
+    paramPtr->base.mustFree = 0;
+
+    switch (evPtr->op) {
+	/*
+	 * The destination thread for the following operations is
+	 * rcPtr->thread, which contains rcPtr->interp, the interp we have to
+	 * call upon for the driver.
+	 */
+
+    case ForwardedClose: {
+	/*
+	 * No parameters/results.
+	 */
+
+	if (InvokeTclMethod(rcPtr, METH_FINAL, NULL, NULL, &resObj)!=TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	}
+
+	/*
+	 * Freeing is done here, in the origin thread, callback command
+	 * objects belong to this thread. Deallocating them in a different
+	 * thread is not allowed
+	 *
+	 * We remove the channel from both interpreter and thread maps before
+	 * releasing the memory, to prevent future accesses (like by
+	 * 'postevent') from finding and dereferencing a dangling pointer.
+	 */
+
+	rcmPtr = GetReflectedChannelMap(interp);
+	hPtr = Tcl_FindHashEntry(&rcmPtr->map,
+                Tcl_GetChannelName(rcPtr->chan));
+	Tcl_DeleteHashEntry(hPtr);
+
+	rcmPtr = GetThreadReflectedChannelMap();
+	hPtr = Tcl_FindHashEntry(&rcmPtr->map,
+                Tcl_GetChannelName(rcPtr->chan));
+	Tcl_DeleteHashEntry(hPtr);
+	MarkDead(rcPtr);
+	break;
+    }
+
+    case ForwardedInput: {
+	Tcl_Obj *toReadObj = Tcl_NewIntObj(paramPtr->input.toRead);
+        Tcl_IncrRefCount(toReadObj);
+
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_READ, toReadObj, NULL, &resObj)!=TCL_OK){
+	    int code = ErrnoReturn(rcPtr, resObj);
+
+	    if (code < 0) {
+		paramPtr->base.code = code;
+	    } else {
+		ForwardSetObjError(paramPtr, resObj);
+	    }
+	    paramPtr->input.toRead = -1;
+	} else {
+	    /*
+	     * Process a regular result.
+	     */
+
+	    int bytec;			/* Number of returned bytes */
+	    unsigned char *bytev;	/* Array of returned bytes */
+
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+	    if (paramPtr->input.toRead < bytec) {
+		ForwardSetStaticError(paramPtr, msg_read_toomuch);
+		paramPtr->input.toRead = -1;
+	    } else {
+		if (bytec > 0) {
+		    memcpy(paramPtr->input.buf, bytev, (size_t) bytec);
+		}
+		paramPtr->input.toRead = bytec;
+	    }
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(toReadObj);
+	break;
+    }
+
+    case ForwardedOutput: {
+	Tcl_Obj *bufObj = Tcl_NewByteArrayObj((unsigned char *)
+                paramPtr->output.buf, paramPtr->output.toWrite);
+        Tcl_IncrRefCount(bufObj);
+
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_WRITE, bufObj, NULL, &resObj) != TCL_OK) {
+	    int code = ErrnoReturn(rcPtr, resObj);
+
+	    if (code < 0) {
+		paramPtr->base.code = code;
+	    } else {
+		ForwardSetObjError(paramPtr, resObj);
+	    }
+	    paramPtr->output.toWrite = -1;
+	} else {
+	    /*
+	     * Process a regular result.
+	     */
+
+	    int written;
+
+	    if (Tcl_GetIntFromObj(interp, resObj, &written) != TCL_OK) {
+		Tcl_DecrRefCount(resObj);
+		resObj = MarshallError(interp);
+		ForwardSetObjError(paramPtr, resObj);
+		paramPtr->output.toWrite = -1;
+	    } else if (written==0 || paramPtr->output.toWrite<written) {
+		ForwardSetStaticError(paramPtr, msg_write_toomuch);
+		paramPtr->output.toWrite = -1;
+	    } else {
+		paramPtr->output.toWrite = written;
+	    }
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(bufObj);
+	break;
+    }
+
+    case ForwardedSeek: {
+	Tcl_Obj *offObj = Tcl_NewWideIntObj(paramPtr->seek.offset);
+	Tcl_Obj *baseObj = Tcl_NewStringObj(
+                (paramPtr->seek.seekMode==SEEK_SET) ? "start" :
+                (paramPtr->seek.seekMode==SEEK_CUR) ? "current" : "end", -1);
+
+        Tcl_IncrRefCount(offObj);
+        Tcl_IncrRefCount(baseObj);
+
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_SEEK, offObj, baseObj, &resObj)!=TCL_OK){
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->seek.offset = -1;
+	} else {
+	    /*
+	     * Process a regular result. If the type is wrong this may change
+	     * into an error.
+	     */
+
+	    Tcl_WideInt newLoc;
+
+	    if (Tcl_GetWideIntFromObj(interp, resObj, &newLoc) == TCL_OK) {
+		if (newLoc < Tcl_LongAsWide(0)) {
+		    ForwardSetStaticError(paramPtr, msg_seek_beforestart);
+		    paramPtr->seek.offset = -1;
+		} else {
+		    paramPtr->seek.offset = newLoc;
+		}
+	    } else {
+		Tcl_DecrRefCount(resObj);
+		resObj = MarshallError(interp);
+		ForwardSetObjError(paramPtr, resObj);
+		paramPtr->seek.offset = -1;
+	    }
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(offObj);
+        Tcl_DecrRefCount(baseObj);
+	break;
+    }
+
+    case ForwardedWatch: {
+	Tcl_Obj *maskObj = DecodeEventMask(paramPtr->watch.mask);
+        /* assert maskObj.refCount == 1 */
+
+        Tcl_Preserve(rcPtr);
+	rcPtr->interest = paramPtr->watch.mask;
+	(void) InvokeTclMethod(rcPtr, METH_WATCH, maskObj, NULL, NULL);
+	Tcl_DecrRefCount(maskObj);
+        Tcl_Release(rcPtr);
+	break;
+    }
+
+    case ForwardedBlock: {
+	Tcl_Obj *blockObj = Tcl_NewBooleanObj(!paramPtr->block.nonblocking);
+
+        Tcl_IncrRefCount(blockObj);
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_BLOCKING, blockObj, NULL,
+                &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(blockObj);
+	break;
+    }
+
+    case ForwardedSetOpt: {
+	Tcl_Obj *optionObj = Tcl_NewStringObj(paramPtr->setOpt.name, -1);
+	Tcl_Obj *valueObj  = Tcl_NewStringObj(paramPtr->setOpt.value, -1);
+
+        Tcl_IncrRefCount(optionObj);
+        Tcl_IncrRefCount(valueObj);
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_CONFIGURE, optionObj, valueObj,
+                &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(optionObj);
+        Tcl_DecrRefCount(valueObj);
+	break;
+    }
+
+    case ForwardedGetOpt: {
+	/*
+	 * Retrieve the value of one option.
+	 */
+
+	Tcl_Obj *optionObj = Tcl_NewStringObj(paramPtr->getOpt.name, -1);
+
+        Tcl_IncrRefCount(optionObj);
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_CGET, optionObj, NULL, &resObj)!=TCL_OK){
+	    ForwardSetObjError(paramPtr, resObj);
+	} else {
+	    TclDStringAppendObj(paramPtr->getOpt.value, resObj);
+	}
+        Tcl_Release(rcPtr);
+        Tcl_DecrRefCount(optionObj);
+	break;
+    }
+
+    case ForwardedGetOptAll:
+	/*
+	 * Retrieve all options.
+	 */
+
+        Tcl_Preserve(rcPtr);
+	if (InvokeTclMethod(rcPtr, METH_CGETALL, NULL, NULL, &resObj) != TCL_OK){
+	    ForwardSetObjError(paramPtr, resObj);
+	} else {
+	    /*
+	     * Extract list, validate that it is a list, and #elements. See
+	     * NOTE (4) as well.
+	     */
+
+	    int listc;
+	    Tcl_Obj **listv;
+
+	    if (Tcl_ListObjGetElements(interp, resObj, &listc,
+                    &listv) != TCL_OK) {
+		Tcl_DecrRefCount(resObj);
+		resObj = MarshallError(interp);
+		ForwardSetObjError(paramPtr, resObj);
+	    } else if ((listc % 2) == 1) {
+		/*
+		 * Odd number of elements is wrong. [x].
+		 */
+
+		char *buf = ckalloc(200);
+		sprintf(buf,
+			"{Expected list with even number of elements, got %d %s instead}",
+			listc, (listc == 1 ? "element" : "elements"));
+
+		ForwardSetDynamicError(paramPtr, buf);
+	    } else {
+		int len;
+		const char *str = TclGetStringFromObj(resObj, &len);
+
+		if (len) {
+		    TclDStringAppendLiteral(paramPtr->getOpt.value, " ");
+		    Tcl_DStringAppend(paramPtr->getOpt.value, str, len);
+		}
+	    }
+	}
+        Tcl_Release(rcPtr);
+	break;
+
+    default:
+	/*
+	 * Bad operation code.
+	 */
+
+	Tcl_Panic("Bad operation code in ForwardProc");
+	break;
+    }
+
+    /*
+     * Remove the reference we held on the result of the invoke, if we had
+     * such.
+     */
+
+    if (resObj != NULL) {
+	Tcl_DecrRefCount(resObj);
+    }
+
+    if (resultPtr) {
+	/*
+	 * Report the forwarding result synchronously to the waiting caller.
+	 * This unblocks (*) as well. This is wrapped into a conditional
+	 * because the caller may have exited in the mean time.
+	 */
+
+	Tcl_MutexLock(&rcForwardMutex);
+	resultPtr->result = TCL_OK;
+	Tcl_ConditionNotify(&resultPtr->done);
+	Tcl_MutexUnlock(&rcForwardMutex);
+    }
+
+    return 1;
+}
+
+static void
+SrcExitProc(
+    ClientData clientData)
+{
+    ForwardingEvent *evPtr = clientData;
+    ForwardingResult *resultPtr;
+    ForwardParam *paramPtr;
+
+    /*
+     * NOTE (2): Can this handler be called with the originator blocked?
+     */
+
+    /*
+     * The originator for the event exited. It is not sure if this can happen,
+     * as the originator should be blocked at (*) while the event is in
+     * transit/pending.
+     *
+     * We make sure that the event cannot refer to the result anymore, remove
+     * it from the list of pending results and free the structure. Locking the
+     * access ensures that we cannot get in conflict with "ForwardProc",
+     * should it already execute the event.
+     */
+
+    Tcl_MutexLock(&rcForwardMutex);
+
+    resultPtr = evPtr->resultPtr;
+    paramPtr = evPtr->param;
+
+    evPtr->resultPtr = NULL;
+    resultPtr->evPtr = NULL;
+    resultPtr->result = TCL_ERROR;
+
+    ForwardSetStaticError(paramPtr, msg_send_originlost);
+
+    /*
+     * See below: TclSpliceOut(resultPtr, forwardList);
+     */
+
+    Tcl_MutexUnlock(&rcForwardMutex);
+
+    /*
+     * This unlocks (*). The structure will be spliced out and freed by
+     * "ForwardProc". Maybe.
+     */
+
+    Tcl_ConditionNotify(&resultPtr->done);
+}
+
+static void
+ForwardSetObjError(
+    ForwardParam *paramPtr,
+    Tcl_Obj *obj)
+{
+    int len;
+    const char *msgStr = TclGetStringFromObj(obj, &len);
+
+    len++;
+    ForwardSetDynamicError(paramPtr, ckalloc(len));
+    memcpy(paramPtr->base.msgStr, msgStr, (unsigned) len);
+}
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclIORTrans.c b/generic/tclIORTrans.c
new file mode 100644
index 0000000..f198c69
--- /dev/null
+++ b/generic/tclIORTrans.c
@@ -0,0 +1,3427 @@
+/*
+ * tclIORTrans.c --
+ *
+ *	This file contains the implementation of Tcl's generic transformation
+ *	reflection code, which allows the implementation of Tcl channel
+ *	transformations in Tcl code.
+ *
+ *	Parts of this file are based on code contributed by Jean-Claude
+ *	Wippler.
+ *
+ *	See TIP #230 for the specification of this functionality.
+ *
+ * Copyright (c) 2007-2008 ActiveState.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclIO.h"
+#include <assert.h>
+
+#ifndef EINVAL
+#define EINVAL	9
+#endif
+#ifndef EOK
+#define EOK	0
+#endif
+
+/* DUPLICATE of HaveVersion() in tclIO.c // TODO - MODULE_SCOPE */
+static int		HaveVersion(const Tcl_ChannelType *typePtr,
+			    Tcl_ChannelTypeVersion minimumVersion);
+
+/*
+ * Signatures of all functions used in the C layer of the reflection.
+ */
+
+static int		ReflectClose(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		ReflectInput(ClientData clientData, char *buf,
+			    int toRead, int *errorCodePtr);
+static int		ReflectOutput(ClientData clientData, const char *buf,
+			    int toWrite, int *errorCodePtr);
+static void		ReflectWatch(ClientData clientData, int mask);
+static int		ReflectBlock(ClientData clientData, int mode);
+static Tcl_WideInt	ReflectSeekWide(ClientData clientData,
+			    Tcl_WideInt offset, int mode, int *errorCodePtr);
+static int		ReflectSeek(ClientData clientData, long offset,
+			    int mode, int *errorCodePtr);
+static int		ReflectGetOption(ClientData clientData,
+			    Tcl_Interp *interp, const char *optionName,
+			    Tcl_DString *dsPtr);
+static int		ReflectSetOption(ClientData clientData,
+			    Tcl_Interp *interp, const char *optionName,
+			    const char *newValue);
+static int		ReflectHandle(ClientData clientData, int direction,
+			    ClientData *handle);
+static int		ReflectNotify(ClientData clientData, int mask);
+
+/*
+ * The C layer channel type/driver definition used by the reflection.
+ */
+
+static const Tcl_ChannelType tclRTransformType = {
+    "tclrtransform",		/* Type name. */
+    TCL_CHANNEL_VERSION_5,	/* v5 channel. */
+    ReflectClose,		/* Close channel, clean instance data. */
+    ReflectInput,		/* Handle read request. */
+    ReflectOutput,		/* Handle write request. */
+    ReflectSeek,		/* Move location of access point. */
+    ReflectSetOption,		/* Set options. */
+    ReflectGetOption,		/* Get options. */
+    ReflectWatch,		/* Initialize notifier. */
+    ReflectHandle,		/* Get OS handle from the channel. */
+    NULL,			/* No close2 support. NULL'able. */
+    ReflectBlock,		/* Set blocking/nonblocking. */
+    NULL,			/* Flush channel. Not used by core.
+				 * NULL'able. */
+    ReflectNotify,		/* Handle events. */
+    ReflectSeekWide,		/* Move access point (64 bit). */
+    NULL,			/* thread action */
+    NULL			/* truncate */
+};
+
+/*
+ * Structure of the buffer to hold transform results to be consumed by higher
+ * layers upon reading from the channel, plus the functions to manage such.
+ */
+
+typedef struct _ResultBuffer_ {
+    unsigned char *buf;		/* Reference to the buffer area. */
+    int allocated;		/* Allocated size of the buffer area. */
+    int used;			/* Number of bytes in the buffer,
+				 * <= allocated. */
+} ResultBuffer;
+
+#define ResultLength(r) ((r)->used)
+/* static int		ResultLength(ResultBuffer *r); */
+
+static void		ResultClear(ResultBuffer *r);
+static void		ResultInit(ResultBuffer *r);
+static void		ResultAdd(ResultBuffer *r, unsigned char *buf,
+			    int toWrite);
+static int		ResultCopy(ResultBuffer *r, unsigned char *buf,
+			    int toRead);
+
+#define RB_INCREMENT (512)
+
+/*
+ * Convenience macro to make some casts easier to use.
+ */
+
+#define UCHARP(x)	((unsigned char *) (x))
+
+/*
+ * Instance data for a reflected transformation. ===========================
+ */
+
+typedef struct {
+    Tcl_Channel chan;		/* Back reference to the channel of the
+				 * transformation itself. */
+    Tcl_Channel parent;		/* Reference to the channel the transformation
+				 * was pushed on. */
+    Tcl_Interp *interp;		/* Reference to the interpreter containing the
+				 * Tcl level part of the channel. */
+    Tcl_Obj *handle;		/* Reference to transform handle. Also stored
+				 * in the argv, see below. The separate field
+				 * gives us direct access, needed when working
+				 * with the reflection maps. */
+#ifdef TCL_THREADS
+    Tcl_ThreadId thread;	/* Thread the 'interp' belongs to. */
+#endif
+
+    Tcl_TimerToken timer;
+
+    /* See [==] as well.
+     * Storage for the command prefix and the additional words required for
+     * the invocation of methods in the command handler.
+     *
+     * argv [0] ... [.] | [argc-2] [argc-1] | [argc]  [argc+2]
+     *      cmd ... pfx | method   chan     | detail1 detail2
+     *      ~~~~ CT ~~~            ~~ CT ~~
+     *
+     * CT = Belongs to the 'Command handler Thread'.
+     */
+
+    int argc;			/* Number of preallocated words - 2. */
+    Tcl_Obj **argv;		/* Preallocated array for calling the handler.
+				 * args[0] is placeholder for cmd word.
+				 * Followed by the arguments in the prefix,
+				 * plus 4 placeholders for method, channel,
+				 * and at most two varying (method specific)
+				 * words. */
+    int methods;		/* Bitmask of supported methods. */
+
+    /*
+     * NOTE (9): Should we have predefined shared literals for the method
+     * names?
+     */
+
+    int mode;			/* Mask of R/W mode */
+    int nonblocking;		/* Flag: Channel is blocking or not. */
+    int readIsDrained;		/* Flag: Read buffers are flushed. */
+    int eofPending;		/* Flag: EOF seen down, but not raised up */
+    int dead;			/* Boolean signal that some operations
+				 * should no longer be attempted. */
+    ResultBuffer result;
+} ReflectedTransform;
+
+/*
+ * Structure of the table mapping from transform handles to reflected
+ * transform (channels). Each interpreter which has the handler command for
+ * one or more reflected transforms records them in such a table, so that we
+ * are able to find them during interpreter/thread cleanup even if the actual
+ * channel they belong to was moved to a different interpreter and/or thread.
+ *
+ * The table is reachable via the standard interpreter AssocData, the key is
+ * defined below.
+ */
+
+typedef struct {
+    Tcl_HashTable map;
+} ReflectedTransformMap;
+
+#define RTMKEY "ReflectedTransformMap"
+
+/*
+ * Method literals. ==================================================
+ */
+
+static const char *const methodNames[] = {
+    "clear",		/* OPT */
+    "drain",		/* OPT, drain => read */
+    "finalize",		/*     */
+    "flush",		/* OPT, flush => write */
+    "initialize",	/*     */
+    "limit?",		/* OPT */
+    "read",		/* OPT */
+    "write",		/* OPT */
+    NULL
+};
+typedef enum {
+    METH_CLEAR,
+    METH_DRAIN,
+    METH_FINAL,
+    METH_FLUSH,
+    METH_INIT,
+    METH_LIMIT,
+    METH_READ,
+    METH_WRITE
+} MethodName;
+
+#define FLAG(m) (1 << (m))
+#define REQUIRED_METHODS \
+	(FLAG(METH_INIT) | FLAG(METH_FINAL))
+#define RANDW \
+	(TCL_READABLE | TCL_WRITABLE)
+
+#define IMPLIES(a,b)	((!(a)) || (b))
+#define NEGIMPL(a,b)
+#define HAS(x,f)	(x & FLAG(f))
+
+#ifdef TCL_THREADS
+/*
+ * Thread specific types and structures.
+ *
+ * We are here essentially creating a very specific implementation of 'thread
+ * send'.
+ */
+
+/*
+ * Enumeration of all operations which can be forwarded.
+ */
+
+typedef enum {
+    ForwardedClear,
+    ForwardedClose,
+    ForwardedDrain,
+    ForwardedFlush,
+    ForwardedInput,
+    ForwardedLimit,
+    ForwardedOutput
+} ForwardedOperation;
+
+/*
+ * Event used to forward driver invocations to the thread actually managing
+ * the channel. We cannot construct the command to execute and forward that.
+ * Because then it will contain a mixture of Tcl_Obj's belonging to both the
+ * command handler thread (CT), and the thread managing the channel (MT),
+ * executed in CT. Tcl_Obj's are not allowed to cross thread boundaries. So we
+ * forward an operation code, the argument details, and reference to results.
+ * The command is assembled in the CT and belongs fully to that thread. No
+ * sharing problems.
+ */
+
+typedef struct ForwardParamBase {
+    int code;			/* O: Ok/Fail of the cmd handler */
+    char *msgStr;		/* O: Error message for handler failure */
+    int mustFree;		/* O: True if msgStr is allocated, false if
+				 * otherwise (static). */
+} ForwardParamBase;
+
+/*
+ * Operation specific parameter/result structures. (These are "subtypes" of
+ * ForwardParamBase. Where an operation does not need any special types, it
+ * has no "subtype" and just uses ForwardParamBase, as listed above.)
+ */
+
+struct ForwardParamTransform {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    char *buf;			/* I: Bytes to transform,
+				 * O: Bytes in transform result */
+    int size;			/* I: #bytes to transform,
+				 * O: #bytes in the transform result */
+};
+struct ForwardParamLimit {
+    ForwardParamBase base;	/* "Supertype". MUST COME FIRST. */
+    int max;			/* O: Character read limit */
+};
+
+/*
+ * Now join all these together in a single union for convenience.
+ */
+
+typedef union ForwardParam {
+    ForwardParamBase base;
+    struct ForwardParamTransform transform;
+    struct ForwardParamLimit limit;
+} ForwardParam;
+
+/*
+ * Forward declaration.
+ */
+
+typedef struct ForwardingResult ForwardingResult;
+
+/*
+ * General event structure, with reference to operation specific data.
+ */
+
+typedef struct ForwardingEvent {
+    Tcl_Event event;		/* Basic event data, has to be first item */
+    ForwardingResult *resultPtr;
+    ForwardedOperation op;	/* Forwarded driver operation */
+    ReflectedTransform *rtPtr;	/* Channel instance */
+    ForwardParam *param;	/* Packaged arguments and return values, a
+				 * ForwardParam pointer. */
+} ForwardingEvent;
+
+/*
+ * Structure to manage the result of the forwarding. This is not the result of
+ * the operation itself, but about the success of the forward event itself.
+ * The event can be successful, even if the operation which was forwarded
+ * failed. It is also there to manage the synchronization between the involved
+ * threads.
+ */
+
+struct ForwardingResult {
+    Tcl_ThreadId src;		/* Originating thread. */
+    Tcl_ThreadId dst;		/* Thread the op was forwarded to. */
+    Tcl_Interp *dsti;		/* Interpreter in the thread the op was
+				 * forwarded to. */
+    Tcl_Condition done;		/* Condition variable the forwarder blocks
+				 * on. */
+    int result;			/* TCL_OK or TCL_ERROR */
+    ForwardingEvent *evPtr;	/* Event the result belongs to. */
+    ForwardingResult *prevPtr, *nextPtr;
+				/* Links into the list of pending forwarded
+				 * results. */
+};
+
+typedef struct {
+    /*
+     * Table of all reflected transformations owned by this thread.
+     */
+
+    ReflectedTransformMap *rtmPtr;
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * List of forwarded operations which have not completed yet, plus the mutex
+ * to protect the access to this process global list.
+ */
+
+static ForwardingResult *forwardList = NULL;
+TCL_DECLARE_MUTEX(rtForwardMutex)
+
+/*
+ * Function containing the generic code executing a forward, and wrapper
+ * macros for the actual operations we wish to forward. Uses ForwardProc as
+ * the event function executed by the thread receiving a forwarding event
+ * (which executes the appropriate function and collects the result, if any).
+ *
+ * The two ExitProcs are handlers so that things do not deadlock when either
+ * thread involved in the forwarding exits. They also clean things up so that
+ * we don't leak resources when threads go away.
+ */
+
+static void		ForwardOpToOwnerThread(ReflectedTransform *rtPtr,
+			    ForwardedOperation op, const void *param);
+static int		ForwardProc(Tcl_Event *evPtr, int mask);
+static void		SrcExitProc(ClientData clientData);
+
+#define FreeReceivedError(p) \
+	do {								\
+	    if ((p)->base.mustFree) {					\
+		ckfree((p)->base.msgStr);				\
+	    }								\
+	} while (0)
+#define PassReceivedErrorInterp(i,p) \
+	do {								\
+	    if ((i) != NULL) {						\
+		Tcl_SetChannelErrorInterp((i),				\
+			Tcl_NewStringObj((p)->base.msgStr, -1));	\
+	    }								\
+	    FreeReceivedError(p);					\
+	} while (0)
+#define PassReceivedError(c,p) \
+	do {								\
+	    Tcl_SetChannelError((c),					\
+		    Tcl_NewStringObj((p)->base.msgStr, -1));		\
+	    FreeReceivedError(p);					\
+	} while (0)
+#define ForwardSetStaticError(p,emsg) \
+	do {								\
+	    (p)->base.code = TCL_ERROR;					\
+	    (p)->base.mustFree = 0;					\
+	    (p)->base.msgStr = (char *) (emsg);				\
+	} while (0)
+#define ForwardSetDynamicError(p,emsg) \
+	do {								\
+	    (p)->base.code = TCL_ERROR;					\
+	    (p)->base.mustFree = 1;					\
+	    (p)->base.msgStr = (char *) (emsg);				\
+	} while (0)
+
+static void		ForwardSetObjError(ForwardParam *p,
+			    Tcl_Obj *objPtr);
+static ReflectedTransformMap *	GetThreadReflectedTransformMap(void);
+static void		DeleteThreadReflectedTransformMap(
+			    ClientData clientData);
+#endif /* TCL_THREADS */
+
+#define SetChannelErrorStr(c,msgStr) \
+	Tcl_SetChannelError((c), Tcl_NewStringObj((msgStr), -1))
+
+static Tcl_Obj *	MarshallError(Tcl_Interp *interp);
+static void		UnmarshallErrorResult(Tcl_Interp *interp,
+			    Tcl_Obj *msgObj);
+
+/*
+ * Static functions for this file:
+ */
+
+static Tcl_Obj *	DecodeEventMask(int mask);
+static ReflectedTransform * NewReflectedTransform(Tcl_Interp *interp,
+			    Tcl_Obj *cmdpfxObj, int mode, Tcl_Obj *handleObj,
+			    Tcl_Channel parentChan);
+static Tcl_Obj *	NextHandle(void);
+static void		FreeReflectedTransform(ReflectedTransform *rtPtr);
+static void		FreeReflectedTransformArgs(ReflectedTransform *rtPtr);
+static int		InvokeTclMethod(ReflectedTransform *rtPtr,
+			    const char *method, Tcl_Obj *argOneObj,
+			    Tcl_Obj *argTwoObj, Tcl_Obj **resultObjPtr);
+
+static ReflectedTransformMap *	GetReflectedTransformMap(Tcl_Interp *interp);
+static void		DeleteReflectedTransformMap(ClientData clientData,
+			    Tcl_Interp *interp);
+
+/*
+ * Global constant strings (messages). ==================
+ * These string are used directly as bypass errors, thus they have to be valid
+ * Tcl lists where the last element is the message itself. Hence the
+ * list-quoting to keep the words of the message together. See also [x].
+ */
+
+static const char *msg_read_unsup = "{read not supported by Tcl driver}";
+static const char *msg_write_unsup = "{write not supported by Tcl driver}";
+#ifdef TCL_THREADS
+static const char *msg_send_originlost = "{Channel thread lost}";
+static const char *msg_send_dstlost = "{Owner lost}";
+#endif /* TCL_THREADS */
+static const char *msg_dstlost =
+    "-code 1 -level 0 -errorcode NONE -errorinfo {} -errorline 1 {Owner lost}";
+
+/*
+ * Timer management (flushing out buffered data via artificial events).
+ */
+
+/*
+ * Helper functions encapsulating some of the thread forwarding to make the
+ * control flow in callers easier.
+ */
+
+static void		TimerKill(ReflectedTransform *rtPtr);
+static void		TimerSetup(ReflectedTransform *rtPtr);
+static void		TimerRun(ClientData clientData);
+static int		TransformRead(ReflectedTransform *rtPtr,
+			    int *errorCodePtr, Tcl_Obj *bufObj);
+static int		TransformWrite(ReflectedTransform *rtPtr,
+			    int *errorCodePtr, unsigned char *buf,
+			    int toWrite);
+static int		TransformDrain(ReflectedTransform *rtPtr,
+			    int *errorCodePtr);
+static int		TransformFlush(ReflectedTransform *rtPtr,
+			    int *errorCodePtr, int op);
+static void		TransformClear(ReflectedTransform *rtPtr);
+static int		TransformLimit(ReflectedTransform *rtPtr,
+			    int *errorCodePtr, int *maxPtr);
+
+/*
+ * Operation codes for TransformFlush().
+ */
+
+#define FLUSH_WRITE	1
+#define FLUSH_DISCARD	0
+
+/*
+ * Main methods to plug into the 'chan' ensemble'. ==================
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChanPushObjCmd --
+ *
+ *	This function is invoked to process the "chan push" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result. The handle of the new channel is placed in the
+ *	interp result.
+ *
+ * Side effects:
+ *	Creates a new channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclChanPushObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    ReflectedTransform *rtPtr;	/* Instance data of the new (transform)
+				 * channel. */
+    Tcl_Obj *chanObj;		/* Handle of parent channel */
+    Tcl_Channel parentChan;	/* Token of parent channel */
+    int mode;			/* R/W mode of parent, later the new channel.
+				 * Has to match the abilities of the handler
+				 * commands */
+    Tcl_Obj *cmdObj;		/* Command prefix, list of words */
+    Tcl_Obj *cmdNameObj;	/* Command name */
+    Tcl_Obj *rtId;		/* Handle of the new transform (channel) */
+    Tcl_Obj *modeObj;		/* mode in obj form for method call */
+    int listc;			/* Result of 'initialize', and of */
+    Tcl_Obj **listv;		/* its sublist in the 2nd element */
+    int methIndex;		/* Encoded method name */
+    int result;			/* Result code for 'initialize' */
+    Tcl_Obj *resObj;		/* Result data for 'initialize' */
+    int methods;		/* Bitmask for supported methods. */
+    ReflectedTransformMap *rtmPtr;
+				/* Map of reflected transforms with handlers
+				 * in this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+    int isNew;			/* Placeholder. */
+
+    /*
+     * Syntax:   chan push CHANNEL CMDPREFIX
+     *           [0]  [1]  [2]     [3]
+     *
+     * Actually: rPush CHANNEL CMDPREFIX
+     *           [0]   [1]     [2]
+     */
+
+#define CHAN	(1)
+#define CMD	(2)
+
+    /*
+     * Number of arguments...
+     */
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channel cmdprefix");
+	return TCL_ERROR;
+    }
+
+    /*
+     * First argument is a channel handle.
+     */
+
+    chanObj = objv[CHAN];
+    parentChan = Tcl_GetChannel(interp, TclGetString(chanObj), &mode);
+    if (parentChan == NULL) {
+	return TCL_ERROR;
+    }
+    parentChan = Tcl_GetTopChannel(parentChan);
+
+    /*
+     * Second argument is command prefix, i.e. list of words, first word is
+     * name of handler command, other words are fixed arguments. Run the
+     * 'initialize' method to get the list of supported methods. Validate
+     * this.
+     */
+
+    cmdObj = objv[CMD];
+
+    /*
+     * Basic check that the command prefix truly is a list.
+     */
+
+    if (Tcl_ListObjIndex(interp, cmdObj, 0, &cmdNameObj) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now create the transformation (channel).
+     */
+
+    rtId = NextHandle();
+    rtPtr = NewReflectedTransform(interp, cmdObj, mode, rtId, parentChan);
+
+    /*
+     * Invoke 'initialize' and validate that the handler is present and ok.
+     * Squash the transformation if not.
+     */
+
+    modeObj = DecodeEventMask(mode);
+    /* assert modeObj.refCount == 1 */
+    result = InvokeTclMethod(rtPtr, "initialize", modeObj, NULL, &resObj);
+    Tcl_DecrRefCount(modeObj);
+    if (result != TCL_OK) {
+	UnmarshallErrorResult(interp, resObj);
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	goto error;
+    }
+
+    /*
+     * Verify the result.
+     * - List, of method names. Convert to mask. Check for non-optionals
+     *   through the mask. Compare open mode against optional r/w.
+     */
+
+    if (Tcl_ListObjGetElements(NULL, resObj, &listc, &listv) != TCL_OK) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s initialize\" returned non-list: %s",
+                TclGetString(cmdObj), TclGetString(resObj)));
+	Tcl_DecrRefCount(resObj);
+	goto error;
+    }
+
+    methods = 0;
+    while (listc > 0) {
+	if (Tcl_GetIndexFromObj(interp, listv[listc-1], methodNames,
+		"method", TCL_EXACT, &methIndex) != TCL_OK) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "chan handler \"%s initialize\" returned %s",
+		    TclGetString(cmdObj),
+		    Tcl_GetString(Tcl_GetObjResult(interp))));
+	    Tcl_DecrRefCount(resObj);
+	    goto error;
+	}
+
+	methods |= FLAG(methIndex);
+	listc--;
+    }
+    Tcl_DecrRefCount(resObj);
+
+    if ((REQUIRED_METHODS & methods) != REQUIRED_METHODS) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" does not support all required methods",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    /*
+     * Mode tell us what the parent channel supports. The methods tell us what
+     * the handler supports. We remove the non-supported bits from the mode
+     * and check that the channel is not completely inacessible. Afterward the
+     * mode tells us which methods are still required, and these methods will
+     * also be supported by the handler, by design of the check.
+     */
+
+    if (!HAS(methods, METH_READ)) {
+	mode &= ~TCL_READABLE;
+    }
+    if (!HAS(methods, METH_WRITE)) {
+	mode &= ~TCL_WRITABLE;
+    }
+
+    if (!mode) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" makes the channel inaccessible",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    /*
+     * The mode and support for it is ok, now check the internal constraints.
+     */
+
+    if (!IMPLIES(HAS(methods, METH_DRAIN), HAS(methods, METH_READ))) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" supports \"drain\" but not \"read\"",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    if (!IMPLIES(HAS(methods, METH_FLUSH), HAS(methods, METH_WRITE))) {
+        Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "chan handler \"%s\" supports \"flush\" but not \"write\"",
+                TclGetString(cmdObj)));
+	goto error;
+    }
+
+    Tcl_ResetResult(interp);
+
+    /*
+     * Everything is fine now.
+     */
+
+    rtPtr->methods = methods;
+    rtPtr->mode = mode;
+    rtPtr->chan = Tcl_StackChannel(interp, &tclRTransformType, rtPtr, mode,
+	    rtPtr->parent);
+
+    /*
+     * Register the transform in our our map for proper handling of deleted
+     * interpreters and/or threads.
+     */
+
+    rtmPtr = GetReflectedTransformMap(interp);
+    hPtr = Tcl_CreateHashEntry(&rtmPtr->map, TclGetString(rtId), &isNew);
+    if (!isNew && rtPtr != Tcl_GetHashValue(hPtr)) {
+	Tcl_Panic("TclChanPushObjCmd: duplicate transformation handle");
+    }
+    Tcl_SetHashValue(hPtr, rtPtr);
+#ifdef TCL_THREADS
+    rtmPtr = GetThreadReflectedTransformMap();
+    hPtr = Tcl_CreateHashEntry(&rtmPtr->map, TclGetString(rtId), &isNew);
+    Tcl_SetHashValue(hPtr, rtPtr);
+#endif /* TCL_THREADS */
+
+    /*
+     * Return the channel as the result of the command.
+     */
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    Tcl_GetChannelName(rtPtr->chan), -1));
+    return TCL_OK;
+
+  error:
+    /*
+     * We are not going through ReflectClose as we never had a channel
+     * structure.
+     */
+
+    Tcl_EventuallyFree(rtPtr, (Tcl_FreeProc *) FreeReflectedTransform);
+    return TCL_ERROR;
+
+#undef CHAN
+#undef CMD
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclChanPopObjCmd --
+ *
+ *	This function is invoked to process the "chan pop" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Posts events to a reflected channel, invokes event handlers. The
+ *	latter implies that arbitrary side effects are possible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclChanPopObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    /*
+     * Syntax:   chan pop CHANNEL
+     *           [0]  [1] [2]
+     *
+     * Actually: rPop CHANNEL
+     *           [0]  [1]
+     */
+
+#define CHAN	(1)
+
+    const char *chanId;		/* Tcl level channel handle */
+    Tcl_Channel chan;		/* Channel associated to the handle */
+    int mode;			/* Channel r/w mode */
+
+    /*
+     * Number of arguments...
+     */
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "channel");
+	return TCL_ERROR;
+    }
+
+    /*
+     * First argument is a channel, which may have a (reflected)
+     * transformation.
+     */
+
+    chanId = TclGetString(objv[CHAN]);
+    chan = Tcl_GetChannel(interp, chanId, &mode);
+
+    if (chan == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Removing transformations is generic, and not restricted to reflected
+     * transformations.
+     */
+
+    Tcl_UnstackChannel(interp, chan);
+    return TCL_OK;
+
+#undef CHAN
+}
+
+/*
+ * Channel error message marshalling utilities.
+ */
+
+static Tcl_Obj *
+MarshallError(
+    Tcl_Interp *interp)
+{
+    /*
+     * Capture the result status of the interpreter into a string. => List of
+     * options and values, followed by the error message. The result has
+     * refCount 0.
+     */
+
+    Tcl_Obj *returnOpt = Tcl_GetReturnOptions(interp, TCL_ERROR);
+
+    /*
+     * => returnOpt.refCount == 0. We can append directly.
+     */
+
+    Tcl_ListObjAppendElement(NULL, returnOpt, Tcl_GetObjResult(interp));
+    return returnOpt;
+}
+
+static void
+UnmarshallErrorResult(
+    Tcl_Interp *interp,
+    Tcl_Obj *msgObj)
+{
+    int lc;
+    Tcl_Obj **lv;
+    int explicitResult;
+    int numOptions;
+
+    /*
+     * Process the caught message.
+     *
+     * Syntax = (option value)... ?message?
+     *
+     * Bad syntax causes a panic. This is OK because the other side uses
+     * Tcl_GetReturnOptions and list construction functions to marshall the
+     * information; if we panic here, something has gone badly wrong already.
+     */
+
+    if (Tcl_ListObjGetElements(interp, msgObj, &lc, &lv) != TCL_OK) {
+	Tcl_Panic("TclChanCaughtErrorBypass: Bad syntax of caught result");
+    }
+    if (interp == NULL) {
+	return;
+    }
+
+    explicitResult = lc & 1;		/* Odd number of values? */
+    numOptions = lc - explicitResult;
+
+    if (explicitResult) {
+	Tcl_SetObjResult(interp, lv[lc-1]);
+    }
+
+    Tcl_SetReturnOptions(interp, Tcl_NewListObj(numOptions, lv));
+    ((Interp *) interp)->flags &= ~ERR_ALREADY_LOGGED;
+}
+
+/*
+ * Driver functions. ================================================
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectClose --
+ *
+ *	This function is invoked when the channel is closed, to delete the
+ *	driver specific instance data.
+ *
+ * Results:
+ *	A posix error.
+ *
+ * Side effects:
+ *	Releases memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectClose(
+    ClientData clientData,
+    Tcl_Interp *interp)
+{
+    ReflectedTransform *rtPtr = clientData;
+    int errorCode, errorCodeSet = 0;
+    int result = TCL_OK;	/* Result code for 'close' */
+    Tcl_Obj *resObj;		/* Result data for 'close' */
+    ReflectedTransformMap *rtmPtr;
+				/* Map of reflected transforms with handlers
+				 * in this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+
+    if (TclInThreadExit()) {
+	/*
+	 * This call comes from TclFinalizeIOSystem. There are no
+	 * interpreters, and therefore we cannot call upon the handler command
+	 * anymore. Threading is irrelevant as well. We simply clean up all
+	 * our C level data structures and leave the Tcl level to the other
+	 * finalization functions.
+	 */
+
+	/*
+	 * THREADED => Forward this to the origin thread
+	 *
+	 * Note: DeleteThreadReflectedTransformMap() is the thread exit handler
+	 * for the origin thread. Use this to clean up the structure? Except
+	 * if lost?
+	 */
+
+#ifdef TCL_THREADS
+	if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	    ForwardParam p;
+
+	    ForwardOpToOwnerThread(rtPtr, ForwardedClose, &p);
+	    result = p.base.code;
+
+	    if (result != TCL_OK) {
+		FreeReceivedError(&p);
+	    }
+	}
+#endif /* TCL_THREADS */
+
+	Tcl_EventuallyFree(rtPtr, (Tcl_FreeProc *) FreeReflectedTransform);
+	return EOK;
+    }
+
+    /*
+     * In the reflected channel implementation a cleaned method mask here
+     * implies that the channel creation was aborted, and "finalize" must not
+     * be called. for transformations however we are not going through here on
+     * such an abort, but directly through FreeReflectedTransform. So for us
+     * that check is not necessary. We always go through 'finalize'.
+     */
+
+    if (HAS(rtPtr->methods, METH_DRAIN) && !rtPtr->readIsDrained) {
+	if (!TransformDrain(rtPtr, &errorCode)) {
+#ifdef TCL_THREADS
+	    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+		Tcl_EventuallyFree(rtPtr,
+			(Tcl_FreeProc *) FreeReflectedTransform);
+		return errorCode;
+	    }
+#endif /* TCL_THREADS */
+	    errorCodeSet = 1;
+	    goto cleanup;
+	}
+    }
+
+    if (HAS(rtPtr->methods, METH_FLUSH)) {
+	if (!TransformFlush(rtPtr, &errorCode, FLUSH_WRITE)) {
+#ifdef TCL_THREADS
+	    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+		Tcl_EventuallyFree(rtPtr,
+			(Tcl_FreeProc *) FreeReflectedTransform);
+		return errorCode;
+	    }
+#endif /* TCL_THREADS */
+	    errorCodeSet = 1;
+	    goto cleanup;
+	}
+    }
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedClose, &p);
+	result = p.base.code;
+
+	Tcl_EventuallyFree(rtPtr, (Tcl_FreeProc *) FreeReflectedTransform);
+
+	if (result != TCL_OK) {
+	    PassReceivedErrorInterp(interp, &p);
+	    return EINVAL;
+	}
+	return EOK;
+    }
+#endif /* TCL_THREADS */
+
+    /*
+     * Do the actual invokation of "finalize" now; we're in the right thread.
+     */
+
+    result = InvokeTclMethod(rtPtr, "finalize", NULL, NULL, &resObj);
+    if ((result != TCL_OK) && (interp != NULL)) {
+	Tcl_SetChannelErrorInterp(interp, resObj);
+    }
+
+    Tcl_DecrRefCount(resObj);	/* Remove reference we held from the
+				 * invoke. */
+
+  cleanup:
+
+    /*
+     * Remove the transform from the map before releasing the memory, to
+     * prevent future accesses from finding and dereferencing a dangling
+     * pointer.
+     *
+     * NOTE: The transform may not be in the map. This is ok, that happens
+     * when the transform was created in a different interpreter and/or thread
+     * and then was moved here.
+     *
+     * NOTE: The channel may have been removed from the map already via
+     * the per-interp DeleteReflectedTransformMap exit-handler.
+     */
+
+    if (!rtPtr->dead) {
+	rtmPtr = GetReflectedTransformMap(rtPtr->interp);
+	hPtr = Tcl_FindHashEntry(&rtmPtr->map, Tcl_GetString(rtPtr->handle));
+	if (hPtr) {
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+
+	/*
+	 * In a threaded interpreter we manage a per-thread map as well,
+	 * to allow us to survive if the script level pulls the rug out
+	 * under a channel by deleting the owning thread.
+	 */
+
+#ifdef TCL_THREADS
+	rtmPtr = GetThreadReflectedTransformMap();
+	hPtr = Tcl_FindHashEntry(&rtmPtr->map, TclGetString(rtPtr->handle));
+	if (hPtr) {
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+#endif /* TCL_THREADS */
+    }
+
+    Tcl_EventuallyFree (rtPtr, (Tcl_FreeProc *) FreeReflectedTransform);
+    return errorCodeSet ? errorCode : ((result == TCL_OK) ? EOK : EINVAL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectInput --
+ *
+ *	This function is invoked when more data is requested from the channel.
+ *
+ * Results:
+ *	The number of bytes read.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectInput(
+    ClientData clientData,
+    char *buf,
+    int toRead,
+    int *errorCodePtr)
+{
+    ReflectedTransform *rtPtr = clientData;
+    int gotBytes, copied, readBytes;
+    Tcl_Obj *bufObj;
+
+    /*
+     * The following check can be done before thread redirection, because we
+     * are reading from an item which is readonly, i.e. will never change
+     * during the lifetime of the channel.
+     */
+
+    if (!(rtPtr->methods & FLAG(METH_READ))) {
+	SetChannelErrorStr(rtPtr->chan, msg_read_unsup);
+	*errorCodePtr = EINVAL;
+	return -1;
+    }
+
+    Tcl_Preserve(rtPtr);
+
+    /* TODO: Consider a more appropriate buffer size. */
+    bufObj = Tcl_NewByteArrayObj(NULL, toRead);
+    Tcl_IncrRefCount(bufObj);
+    gotBytes = 0;
+    if (rtPtr->eofPending) {
+	goto stop;
+    }
+    rtPtr->readIsDrained = 0;
+    while (toRead > 0) {
+	/*
+	 * Loop until the request is satisfied (or no data available from
+	 * below, possibly EOF).
+	 */
+
+	copied = ResultCopy(&rtPtr->result, UCHARP(buf), toRead);
+	toRead -= copied;
+	buf += copied;
+	gotBytes += copied;
+
+	if (toRead == 0) {
+	    goto stop;
+	}
+
+	if (rtPtr->eofPending) {
+	    goto stop;
+	}
+
+
+	/*
+	 * The buffer is exhausted, but the caller wants even more. We now
+	 * have to go to the underlying channel, get more bytes and then
+	 * transform them for delivery. We may not get what we want (full EOF
+	 * or temporarily out of data).
+	 *
+	 * Length (rtPtr->result) == 0, toRead > 0 here. Use 'buf'! as target
+	 * to store the intermediary information read from the parent channel.
+	 *
+	 * Ask the transform how much data it allows us to read from the
+	 * underlying channel. This feature allows the transform to signal EOF
+	 * upstream although there is none downstream. Useful to control an
+	 * unbounded 'fcopy' for example, either through counting bytes, or by
+	 * pattern matching.
+	 */
+
+	if ((rtPtr->methods & FLAG(METH_LIMIT))) {
+	    int maxRead = -1;
+
+	    if (!TransformLimit(rtPtr, errorCodePtr, &maxRead)) {
+		goto error;
+	    }
+	    if (maxRead == 0) {
+		goto stop;
+	    } else if (maxRead > 0) {
+		if (maxRead < toRead) {
+		    toRead = maxRead;
+		}
+	    } /* else: 'maxRead < 0' == Accept the current value of toRead */
+	}
+
+	if (toRead <= 0) {
+	    goto stop;
+	}
+
+
+	readBytes = Tcl_ReadRaw(rtPtr->parent,
+		(char *) Tcl_SetByteArrayLength(bufObj, toRead), toRead);
+	if (readBytes < 0) {
+	    if (Tcl_InputBlocked(rtPtr->parent) && (gotBytes > 0)) {
+
+		/*
+		 * Down channel is blocked and offers zero additional bytes.
+		 * The nonzero gotBytes already returned makes the total
+		 * operation a valid short read.  Return to caller.
+		 */
+
+		goto stop;
+	    }
+
+	    /*
+	     * Either the down channel is not blocked (a real error)
+	     * or it is and there are gotBytes==0 byte copied so far.
+	     * In either case, pass up the error, so we either report
+	     * any real error, or do not mistakenly signal EOF by
+	     * returning 0 to the caller.
+	     */
+
+	    *errorCodePtr = Tcl_GetErrno();
+	    goto error;
+	}
+
+	if (readBytes == 0) {
+
+	    /*
+	     * Zero returned from Tcl_ReadRaw() always indicates EOF
+	     * on the down channel.
+	     */
+
+	    rtPtr->eofPending = 1;
+
+		/*
+		 * Now this is a bit different. The partial data waiting is
+		 * converted and returned.
+		 */
+
+		if (HAS(rtPtr->methods, METH_DRAIN)) {
+		    if (!TransformDrain(rtPtr, errorCodePtr)) {
+			goto error;
+		    }
+		}
+
+		if (ResultLength(&rtPtr->result) == 0) {
+		    /*
+		     * The drain delivered nothing.
+		     */
+
+		    goto stop;
+		}
+
+		continue; /* at: while (toRead > 0) */
+	} /* readBytes == 0 */
+
+	/*
+	 * Transform the read chunk, which was not empty. Anything we got back
+	 * is a transformation result is put into our buffers, and the next
+	 * iteration will put it into the result.
+	 */
+
+	Tcl_SetByteArrayLength(bufObj, readBytes);
+	if (!TransformRead(rtPtr, errorCodePtr, bufObj)) {
+	    goto error;
+	}
+	if (Tcl_IsShared(bufObj)) {
+	    Tcl_DecrRefCount(bufObj);
+	    bufObj = Tcl_NewObj();
+	    Tcl_IncrRefCount(bufObj);
+	}
+	Tcl_SetByteArrayLength(bufObj, 0);
+    } /* while toRead > 0 */
+
+ stop:
+    if (gotBytes == 0) {
+	rtPtr->eofPending = 0;
+    }
+    Tcl_DecrRefCount(bufObj);
+    Tcl_Release(rtPtr);
+    return gotBytes;
+
+ error:
+    gotBytes = -1;
+    goto stop;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectOutput --
+ *
+ *	This function is invoked when data is written to the channel.
+ *
+ * Results:
+ *	The number of bytes actually written.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectOutput(
+    ClientData clientData,
+    const char *buf,
+    int toWrite,
+    int *errorCodePtr)
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * The following check can be done before thread redirection, because we
+     * are reading from an item which is readonly, i.e. will never change
+     * during the lifetime of the channel.
+     */
+
+    if (!(rtPtr->methods & FLAG(METH_WRITE))) {
+	SetChannelErrorStr(rtPtr->chan, msg_write_unsup);
+	*errorCodePtr = EINVAL;
+	return -1;
+    }
+
+    if (toWrite == 0) {
+	/*
+	 * Nothing came in to write, ignore the call
+	 */
+
+	return 0;
+    }
+
+    /*
+     * Discard partial data in the input buffers, i.e. on the read side. Like
+     * we do when explicitly seeking as well.
+     */
+
+    Tcl_Preserve(rtPtr);
+
+    if ((rtPtr->methods & FLAG(METH_CLEAR))) {
+	TransformClear(rtPtr);
+    }
+
+    /*
+     * Hand the data to the transformation itself. Anything it deigned to
+     * return to us is a (partial) transformation result and written to the
+     * parent channel for further processing.
+     */
+
+    if (!TransformWrite(rtPtr, errorCodePtr, UCHARP(buf), toWrite)) {
+	Tcl_Release(rtPtr);
+	return -1;
+    }
+
+    *errorCodePtr = EOK;
+    Tcl_Release(rtPtr);
+    return toWrite;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectSeekWide / ReflectSeek --
+ *
+ *	This function is invoked when the user wishes to seek on the channel.
+ *
+ * Results:
+ *	The new location of the access point.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, per the parent channel, and the called
+ *	scripts.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_WideInt
+ReflectSeekWide(
+    ClientData clientData,
+    Tcl_WideInt offset,
+    int seekMode,
+    int *errorCodePtr)
+{
+    ReflectedTransform *rtPtr = clientData;
+    Channel *parent = (Channel *) rtPtr->parent;
+    Tcl_WideInt curPos;		/* Position on the device. */
+
+    Tcl_DriverSeekProc *seekProc =
+	    Tcl_ChannelSeekProc(Tcl_GetChannelType(rtPtr->parent));
+
+    /*
+     * Fail if the parent channel is not seekable.
+     */
+
+    if (seekProc == NULL) {
+	Tcl_SetErrno(EINVAL);
+	return Tcl_LongAsWide(-1);
+    }
+
+    /*
+     * Check if we can leave out involving the Tcl level, i.e. transformation
+     * handler. This is true for tell requests, and transformations which
+     * support neither flush, nor drain. For these cases we can pass the
+     * request down and the result back up unchanged.
+     */
+
+    Tcl_Preserve(rtPtr);
+
+    if (((seekMode != SEEK_CUR) || (offset != 0))
+	    && (HAS(rtPtr->methods, METH_CLEAR)
+	    || HAS(rtPtr->methods, METH_FLUSH))) {
+	/*
+	 * Neither a tell request, nor clear/flush both not supported. We have
+	 * to go through the Tcl level to clear and/or flush the
+	 * transformation.
+	 */
+
+	if (rtPtr->methods & FLAG(METH_CLEAR)) {
+	    TransformClear(rtPtr);
+	}
+
+	/*
+	 * When flushing the transform for seeking the generated results are
+	 * irrelevant. We cannot put them into the channel, this would move
+	 * the location, throwing it off with regard to where we are and are
+	 * seeking to.
+	 */
+
+	if (HAS(rtPtr->methods, METH_FLUSH)) {
+	    if (!TransformFlush(rtPtr, errorCodePtr, FLUSH_DISCARD)) {
+		Tcl_Release(rtPtr);
+		return -1;
+	    }
+	}
+    }
+
+    /*
+     * Now seek to the new position in the channel as requested by the
+     * caller. Note that we prefer the wideSeekProc if that is available and
+     * non-NULL...
+     */
+
+    if (HaveVersion(parent->typePtr, TCL_CHANNEL_VERSION_3) &&
+	parent->typePtr->wideSeekProc != NULL) {
+	curPos = parent->typePtr->wideSeekProc(parent->instanceData, offset,
+		seekMode, errorCodePtr);
+    } else if (offset < Tcl_LongAsWide(LONG_MIN) ||
+	    offset > Tcl_LongAsWide(LONG_MAX)) {
+	*errorCodePtr = EOVERFLOW;
+	curPos = Tcl_LongAsWide(-1);
+    } else {
+	curPos = Tcl_LongAsWide(parent->typePtr->seekProc(
+		parent->instanceData, Tcl_WideAsLong(offset), seekMode,
+		errorCodePtr));
+    }
+    if (curPos == Tcl_LongAsWide(-1)) {
+	Tcl_SetErrno(*errorCodePtr);
+    }
+
+    *errorCodePtr = EOK;
+    Tcl_Release(rtPtr);
+    return curPos;
+}
+
+static int
+ReflectSeek(
+    ClientData clientData,
+    long offset,
+    int seekMode,
+    int *errorCodePtr)
+{
+    /*
+     * This function can be invoked from a transformation which is based on
+     * standard seeking, i.e. non-wide. Because of this we have to implement
+     * it, a dummy is not enough. We simply delegate the call to the wide
+     * routine.
+     */
+
+    return (int) ReflectSeekWide(clientData, Tcl_LongAsWide(offset), seekMode,
+	    errorCodePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectWatch --
+ *
+ *	This function is invoked to tell the channel what events the I/O
+ *	system is interested in.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReflectWatch(
+    ClientData clientData,
+    int mask)
+{
+    ReflectedTransform *rtPtr = clientData;
+    Tcl_DriverWatchProc *watchProc;
+
+    watchProc = Tcl_ChannelWatchProc(Tcl_GetChannelType(rtPtr->parent));
+    watchProc(Tcl_GetChannelInstanceData(rtPtr->parent), mask);
+
+    /*
+     * Management of the internal timer.
+     */
+
+    if (!(mask & TCL_READABLE) || (ResultLength(&rtPtr->result) == 0)) {
+	/*
+	 * A pending timer may exist, but either is there no (more) interest
+	 * in the events it generates or nothing is available for reading.
+	 * Remove it, if existing.
+	 */
+
+	TimerKill(rtPtr);
+    } else {
+	/*
+	 * There might be no pending timer, but there is interest in readable
+	 * events and we actually have data waiting, so generate a timer to
+	 * flush that if it does not exist.
+	 */
+
+	TimerSetup(rtPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectBlock --
+ *
+ *	This function is invoked to tell the channel which blocking behaviour
+ *	is required of it.
+ *
+ * Results:
+ *	A posix error number.
+ *
+ * Side effects:
+ *	Allocates memory. Arbitrary, as it calls upon a script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectBlock(
+    ClientData clientData,
+    int nonblocking)
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * Transformations simply record the blocking mode in their C level
+     * structure for use by --> ReflectInput. The Tcl level doesn't see this
+     * information or change. As such thread forwarding is not required.
+     */
+
+    rtPtr->nonblocking = nonblocking;
+    return EOK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectSetOption --
+ *
+ *	This function is invoked to configure a channel option.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, per the parent channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectSetOption(
+    ClientData clientData,	/* Channel to query */
+    Tcl_Interp *interp,		/* Interpreter to leave error messages in */
+    const char *optionName,	/* Name of requested option */
+    const char *newValue)	/* The new value */
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * Transformations have no options. Thus the call is passed down unchanged
+     * to the parent channel for processing. Its results are passed back
+     * unchanged as well. This all happens in the thread we are in. As the Tcl
+     * level is not involved there is no need for thread forwarding.
+     */
+
+    Tcl_DriverSetOptionProc *setOptionProc =
+	    Tcl_ChannelSetOptionProc(Tcl_GetChannelType(rtPtr->parent));
+
+    if (setOptionProc == NULL) {
+	return TCL_ERROR;
+    }
+    return setOptionProc(Tcl_GetChannelInstanceData(rtPtr->parent), interp,
+	    optionName, newValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectGetOption --
+ *
+ *	This function is invoked to retrieve all or a channel options.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, per the parent channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectGetOption(
+    ClientData clientData,	/* Channel to query */
+    Tcl_Interp *interp,		/* Interpreter to leave error messages in */
+    const char *optionName,	/* Name of reuqested option */
+    Tcl_DString *dsPtr)		/* String to place the result into */
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * Transformations have no options. Thus the call is passed down unchanged
+     * to the parent channel for processing. Its results are passed back
+     * unchanged as well. This all happens in the thread we are in. As the Tcl
+     * level is not involved there is no need for thread forwarding.
+     *
+     * Note that the parent not having a driver for option retrieval is not an
+     * immediate error. A query for all options is ok. Only a request for a
+     * specific option has to fail.
+     */
+
+    Tcl_DriverGetOptionProc *getOptionProc =
+	    Tcl_ChannelGetOptionProc(Tcl_GetChannelType(rtPtr->parent));
+
+    if (getOptionProc != NULL) {
+	return getOptionProc(Tcl_GetChannelInstanceData(rtPtr->parent),
+		interp, optionName, dsPtr);
+    } else if (optionName == NULL) {
+	return TCL_OK;
+    } else {
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectHandle --
+ *
+ *	This function is invoked to retrieve the associated file handle.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, per the parent channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectHandle(
+    ClientData clientData,
+    int direction,
+    ClientData *handlePtr)
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * Transformations have no handle of their own. As such we simply query
+     * the parent channel for it. This way the qery will ripple down through
+     * all transformations until reaches the base channel. Which then returns
+     * its handle, or fails. The former will then ripple up the stack.
+     *
+     * This all happens in the thread we are in. As the Tcl level is not
+     * involved no forwarding is required.
+     */
+
+    return Tcl_GetChannelHandle(rtPtr->parent, direction, handlePtr);
+}
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReflectNotify --
+ *
+ *	This function is invoked to reported incoming events.
+ *
+ * Results:
+ *	A standard Tcl result code.
+ *
+ * Side effects:
+ *	Arbitrary, per the parent channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ReflectNotify(
+    ClientData clientData,
+    int mask)
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    /*
+     * An event occured in the underlying channel.
+     *
+     * We delete our timer. It was not fired, yet we are here, so the channel
+     * below generated such an event and we don't have to. The renewal of the
+     * interest after the execution of channel handlers will eventually cause
+     * us to recreate the timer (in ReflectWatch).
+     */
+
+    TimerKill(rtPtr);
+
+    /*
+     * Pass to higher layers.
+     */
+
+    return mask;
+}
+
+/*
+ * Helpers. =========================================================
+ */
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DecodeEventMask --
+ *
+ *	This function takes an internal bitmask of events and constructs the
+ *	equivalent list of event items.
+ *
+ * Results:
+ *	A Tcl_Obj reference. The object will have a refCount of one. The user
+ *	has to decrement it to release the object.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ * DUPLICATE of 'DecodeEventMask' in tclIORChan.c
+ */
+
+static Tcl_Obj *
+DecodeEventMask(
+    int mask)
+{
+    register const char *eventStr;
+    Tcl_Obj *evObj;
+
+    switch (mask & RANDW) {
+    case RANDW:
+	eventStr = "read write";
+	break;
+    case TCL_READABLE:
+	eventStr = "read";
+	break;
+    case TCL_WRITABLE:
+	eventStr = "write";
+	break;
+    default:
+	eventStr = "";
+	break;
+    }
+
+    evObj = Tcl_NewStringObj(eventStr, -1);
+    Tcl_IncrRefCount(evObj);
+    return evObj;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NewReflectedTransform --
+ *
+ *	This function is invoked to allocate and initialize the instance data
+ *	of a new reflected channel.
+ *
+ * Results:
+ *	A heap-allocated channel instance.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedTransform *
+NewReflectedTransform(
+    Tcl_Interp *interp,
+    Tcl_Obj *cmdpfxObj,
+    int mode,
+    Tcl_Obj *handleObj,
+    Tcl_Channel parentChan)
+{
+    ReflectedTransform *rtPtr;
+    int listc;
+    Tcl_Obj **listv;
+    int i;
+
+    rtPtr = ckalloc(sizeof(ReflectedTransform));
+
+    /* rtPtr->chan: Assigned by caller. Dummy data here. */
+    /* rtPtr->methods: Assigned by caller. Dummy data here. */
+
+    rtPtr->chan = NULL;
+    rtPtr->methods = 0;
+#ifdef TCL_THREADS
+    rtPtr->thread = Tcl_GetCurrentThread();
+#endif
+    rtPtr->parent = parentChan;
+    rtPtr->interp = interp;
+    rtPtr->handle = handleObj;
+    Tcl_IncrRefCount(handleObj);
+    rtPtr->timer = NULL;
+    rtPtr->mode = 0;
+    rtPtr->readIsDrained = 0;
+    rtPtr->eofPending = 0;
+    rtPtr->nonblocking =
+	    (((Channel *) parentChan)->state->flags & CHANNEL_NONBLOCKING);
+    rtPtr->dead = 0;
+
+    /*
+     * Query parent for current blocking mode.
+     */
+
+    ResultInit(&rtPtr->result);
+
+    /*
+     * Method placeholder.
+     */
+
+    /* ASSERT: cmdpfxObj is a Tcl List */
+
+    Tcl_ListObjGetElements(interp, cmdpfxObj, &listc, &listv);
+
+    /*
+     * See [==] as well.
+     * Storage for the command prefix and the additional words required for
+     * the invocation of methods in the command handler.
+     *
+     * listv [0] [listc-1] | [listc]  [listc+1] |
+     * argv  [0]   ... [.] | [argc-2] [argc-1]  | [argc]  [argc+2]
+     *       cmd   ... pfx | method   chan      | detail1 detail2
+     */
+
+    rtPtr->argc = listc + 2;
+    rtPtr->argv = ckalloc(sizeof(Tcl_Obj *) * (listc+4));
+
+    /*
+     * Duplicate object references.
+     */
+
+    for (i=0; i<listc ; i++) {
+	Tcl_Obj *word = rtPtr->argv[i] = listv[i];
+
+	Tcl_IncrRefCount(word);
+    }
+
+    i++;				/* Skip placeholder for method */
+
+    /*
+     * See [x] in FreeReflectedTransform for release
+     */
+    rtPtr->argv[i] = handleObj;
+    Tcl_IncrRefCount(handleObj);
+
+    /*
+     * The next two objects are kept empty, varying arguments.
+     */
+
+    /*
+     * Initialization complete.
+     */
+
+    return rtPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NextHandle --
+ *
+ *	This function is invoked to generate a channel handle for a new
+ *	reflected channel.
+ *
+ * Results:
+ *	A Tcl_Obj containing the string of the new channel handle. The
+ *	refcount of the returned object is -- zero --.
+ *
+ * Side effects:
+ *	May allocate memory. Mutex protected critical section locks out other
+ *	threads for a short time.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+NextHandle(void)
+{
+    /*
+     * Count number of generated reflected channels. Used for id generation.
+     * Ids are never reclaimed and there is no dealing with wrap around. On
+     * the other hand, "unsigned long" should be big enough except for
+     * absolute longrunners (generate a 100 ids per second => overflow will
+     * occur in 1 1/3 years).
+     */
+
+    TCL_DECLARE_MUTEX(rtCounterMutex)
+    static unsigned long rtCounter = 0;
+    Tcl_Obj *resObj;
+
+    Tcl_MutexLock(&rtCounterMutex);
+    resObj = Tcl_ObjPrintf("rt%lu", rtCounter);
+    rtCounter++;
+    Tcl_MutexUnlock(&rtCounterMutex);
+
+    return resObj;
+}
+
+static void
+FreeReflectedTransformArgs(
+    ReflectedTransform *rtPtr)
+{
+    int i, n = rtPtr->argc - 2;
+
+    if (n < 0) {
+	return;
+    }
+
+    Tcl_DecrRefCount(rtPtr->handle);
+    rtPtr->handle = NULL;
+
+    for (i=0; i<n; i++) {
+	Tcl_DecrRefCount(rtPtr->argv[i]);
+    }
+
+    /*
+     * See [x] in NewReflectedTransform for lock
+     * n+1 = argc-1.
+     */
+    Tcl_DecrRefCount(rtPtr->argv[n+1]);
+
+    rtPtr->argc = 1;
+}
+
+static void
+FreeReflectedTransform(
+    ReflectedTransform *rtPtr)
+{
+    TimerKill(rtPtr);
+    ResultClear(&rtPtr->result);
+
+    FreeReflectedTransformArgs(rtPtr);
+
+    ckfree(rtPtr->argv);
+    ckfree(rtPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InvokeTclMethod --
+ *
+ *	This function is used to invoke the Tcl level of a reflected channel.
+ *	It handles all the command assembly, invokation, and generic state and
+ *	result mgmt. It does *not* handle thread redirection; that is the
+ *	responsibility of clients of this function.
+ *
+ * Results:
+ *	Result code and data as returned by the method.
+ *
+ * Side effects:
+ *	Arbitrary, as it calls upon a Tcl script.
+ *
+ * Contract:
+ *	argOneObj.refCount >= 1 on entry and exit, if argOneObj != NULL
+ *	argTwoObj.refCount >= 1 on entry and exit, if argTwoObj != NULL
+ *	resObj.refCount in {0, 1, ...}
+ *
+ *----------------------------------------------------------------------
+ * Semi-DUPLICATE of 'InvokeTclMethod' in tclIORChan.c
+ * - Semi because different structures are used.
+ * - Still possible to factor out the commonalities into a separate structure.
+ */
+
+static int
+InvokeTclMethod(
+    ReflectedTransform *rtPtr,
+    const char *method,
+    Tcl_Obj *argOneObj,		/* NULL'able */
+    Tcl_Obj *argTwoObj,		/* NULL'able */
+    Tcl_Obj **resultObjPtr)	/* NULL'able */
+{
+    int cmdc;			/* #words in constructed command */
+    Tcl_Obj *methObj = NULL;	/* Method name in object form */
+    Tcl_InterpState sr;		/* State of handler interp */
+    int result;			/* Result code of method invokation */
+    Tcl_Obj *resObj = NULL;	/* Result of method invokation. */
+
+    if (rtPtr->dead) {
+	/*
+	 * The transform is marked as dead. Bail out immediately, with an
+	 * appropriate error.
+	 */
+
+	if (resultObjPtr != NULL) {
+	    resObj = Tcl_NewStringObj(msg_dstlost,-1);
+	    *resultObjPtr = resObj;
+	    Tcl_IncrRefCount(resObj);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * NOTE (5): Decide impl. issue: Cache objects with method names?
+     * Requires TSD data as reflections can be created in many different
+     * threads.
+     * NO: Caching of command resolutions means storage per channel.
+     */
+
+    /*
+     * Insert method into the pre-allocated area, after the command prefix,
+     * before the channel id.
+     */
+
+    methObj = Tcl_NewStringObj(method, -1);
+    Tcl_IncrRefCount(methObj);
+    rtPtr->argv[rtPtr->argc - 2] = methObj;
+
+    /*
+     * Append the additional argument containing method specific details
+     * behind the channel id. If specified.
+     *
+     * Because of the contract there is no need to increment the refcounts.
+     * The objects will survive the Tcl_EvalObjv without change.
+     */
+
+    cmdc = rtPtr->argc;
+    if (argOneObj) {
+	rtPtr->argv[cmdc] = argOneObj;
+	cmdc++;
+	if (argTwoObj) {
+	    rtPtr->argv[cmdc] = argTwoObj;
+	    cmdc++;
+	}
+    }
+
+    /*
+     * And run the handler... This is done in auch a manner which leaves any
+     * existing state intact.
+     */
+
+    sr = Tcl_SaveInterpState(rtPtr->interp, 0 /* Dummy */);
+    Tcl_Preserve(rtPtr);
+    Tcl_Preserve(rtPtr->interp);
+    result = Tcl_EvalObjv(rtPtr->interp, cmdc, rtPtr->argv, TCL_EVAL_GLOBAL);
+
+    /*
+     * We do not try to extract the result information if the caller has no
+     * interest in it. I.e. there is no need to put effort into creating
+     * something which is discarded immediately after.
+     */
+
+    if (resultObjPtr) {
+	if (result == TCL_OK) {
+	    /*
+	     * Ok result taken as is, also if the caller requests that there
+	     * is no capture.
+	     */
+
+	    resObj = Tcl_GetObjResult(rtPtr->interp);
+	} else {
+	    /*
+	     * Non-ok result is always treated as an error. We have to capture
+	     * the full state of the result, including additional options.
+	     *
+	     * This is complex and ugly, and would be completely unnecessary
+	     * if we only added support for a TCL_FORBID_EXCEPTIONS flag.
+	     */
+	    if (result != TCL_ERROR) {
+		Tcl_Obj *cmd = Tcl_NewListObj(cmdc, rtPtr->argv);
+		int cmdLen;
+		const char *cmdString = TclGetStringFromObj(cmd, &cmdLen);
+
+		Tcl_IncrRefCount(cmd);
+		Tcl_ResetResult(rtPtr->interp);
+		Tcl_SetObjResult(rtPtr->interp, Tcl_ObjPrintf(
+			"chan handler returned bad code: %d", result));
+		Tcl_LogCommandInfo(rtPtr->interp, cmdString, cmdString, cmdLen);
+		Tcl_DecrRefCount(cmd);
+		result = TCL_ERROR;
+	    }
+	    Tcl_AppendObjToErrorInfo(rtPtr->interp, Tcl_ObjPrintf(
+		    "\n    (chan handler subcommand \"%s\")", method));
+	    resObj = MarshallError(rtPtr->interp);
+	}
+	Tcl_IncrRefCount(resObj);
+    }
+    Tcl_RestoreInterpState(rtPtr->interp, sr);
+    Tcl_Release(rtPtr->interp);
+    Tcl_Release(rtPtr);
+
+    /*
+     * Cleanup of the dynamic parts of the command.
+     *
+     * The detail objects survived the Tcl_EvalObjv without change because of
+     * the contract. Therefore there is no need to decrement the refcounts. Only
+     * the internal method object has to be disposed of.
+     */
+
+    Tcl_DecrRefCount(methObj);
+
+    /*
+     * The resObj has a ref count of 1 at this location. This means that the
+     * caller of InvokeTclMethod has to dispose of it (but only if it was
+     * returned to it).
+     */
+
+    if (resultObjPtr != NULL) {
+	*resultObjPtr = resObj;
+    }
+
+    /*
+     * There no need to handle the case where nothing is returned, because for
+     * that case resObj was not set anyway.
+     */
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetReflectedTransformMap --
+ *
+ *	Gets and potentially initializes the reflected channel map for an
+ *	interpreter.
+ *
+ * Results:
+ *	A pointer to the map created, for use by the caller.
+ *
+ * Side effects:
+ *	Initializes the reflected channel map for an interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedTransformMap *
+GetReflectedTransformMap(
+    Tcl_Interp *interp)
+{
+    ReflectedTransformMap *rtmPtr = Tcl_GetAssocData(interp, RTMKEY, NULL);
+
+    if (rtmPtr == NULL) {
+	rtmPtr = ckalloc(sizeof(ReflectedTransformMap));
+	Tcl_InitHashTable(&rtmPtr->map, TCL_STRING_KEYS);
+	Tcl_SetAssocData(interp, RTMKEY,
+		(Tcl_InterpDeleteProc *) DeleteReflectedTransformMap, rtmPtr);
+    }
+    return rtmPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteReflectedTransformMap --
+ *
+ *	Deletes the channel table for an interpreter, closing any open
+ *	channels whose refcount reaches zero. This procedure is invoked when
+ *	an interpreter is deleted, via the AssocData cleanup mechanism.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the hash table of channels. May close channels. May flush
+ *	output on closed channels. Removes any channeEvent handlers that were
+ *	registered in this interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteReflectedTransformMap(
+    ClientData clientData,	/* The per-interpreter data structure. */
+    Tcl_Interp *interp)		/* The interpreter being deleted. */
+{
+    ReflectedTransformMap *rtmPtr; /* The map */
+    Tcl_HashSearch hSearch;	 /* Search variable. */
+    Tcl_HashEntry *hPtr;	 /* Search variable. */
+    ReflectedTransform *rtPtr;
+#ifdef TCL_THREADS
+    ForwardingResult *resultPtr;
+    ForwardingEvent *evPtr;
+    ForwardParam *paramPtr;
+#endif /* TCL_THREADS */
+
+    /*
+     * Delete all entries. The channels may have been closed already, or will
+     * be closed later, by the standard IO finalization of an interpreter
+     * under destruction. Except for the channels which were moved to a
+     * different interpreter and/or thread. They do not exist from the IO
+     * systems point of view and will not get closed. Therefore mark all as
+     * dead so that any future access will cause a proper error. For channels
+     * in a different thread we actually do the same as
+     * DeleteThreadReflectedTransformMap(), just restricted to the channels of
+     * this interp.
+     */
+
+    rtmPtr = clientData;
+    for (hPtr = Tcl_FirstHashEntry(&rtmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_FirstHashEntry(&rtmPtr->map, &hSearch)) {
+	rtPtr = Tcl_GetHashValue(hPtr);
+
+	rtPtr->dead = 1;
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    Tcl_DeleteHashTable(&rtmPtr->map);
+    ckfree(&rtmPtr->map);
+
+#ifdef TCL_THREADS
+    /*
+     * The origin interpreter for one or more reflected channels is gone.
+     */
+
+    /*
+     * Get the map of all channels handled by the current thread. This is a
+     * ReflectedTransformMap, but on a per-thread basis, not per-interp. Go
+     * through the channels and remove all which were handled by this
+     * interpreter. They have already been marked as dead.
+     */
+
+    rtmPtr = GetThreadReflectedTransformMap();
+    for (hPtr = Tcl_FirstHashEntry(&rtmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&hSearch)) {
+	rtPtr = Tcl_GetHashValue(hPtr);
+
+	if (rtPtr->interp != interp) {
+	    /*
+	     * Ignore entries for other interpreters.
+	     */
+
+	    continue;
+	}
+
+	rtPtr->dead = 1;
+	FreeReflectedTransformArgs(rtPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+
+    /*
+     * Go through the list of pending results and cancel all whose events were
+     * destined for this interpreter. While this is in progress we block any
+     * other access to the list of pending results.
+     */
+
+    Tcl_MutexLock(&rtForwardMutex);
+
+    for (resultPtr = forwardList; resultPtr != NULL;
+	    resultPtr = resultPtr->nextPtr) {
+	if (resultPtr->dsti != interp) {
+	    /*
+	     * Ignore results/events for other interpreters.
+	     */
+
+	    continue;
+	}
+
+	/*
+	 * The receiver for the event exited, before processing the event. We
+	 * detach the result now, wake the originator up and signal failure.
+	 */
+
+	evPtr = resultPtr->evPtr;
+	if (evPtr == NULL) {
+	    continue;
+	}
+	paramPtr = evPtr->param;
+
+	evPtr->resultPtr = NULL;
+	resultPtr->evPtr = NULL;
+	resultPtr->result = TCL_ERROR;
+
+	ForwardSetStaticError(paramPtr, msg_send_dstlost);
+
+	Tcl_ConditionNotify(&resultPtr->done);
+    }
+    Tcl_MutexUnlock(&rtForwardMutex);
+#endif /* TCL_THREADS */
+}
+
+#ifdef TCL_THREADS
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetThreadReflectedTransformMap --
+ *
+ *	Gets and potentially initializes the reflected channel map for a
+ *	thread.
+ *
+ * Results:
+ *	A pointer to the map created, for use by the caller.
+ *
+ * Side effects:
+ *	Initializes the reflected channel map for a thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ReflectedTransformMap *
+GetThreadReflectedTransformMap(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (!tsdPtr->rtmPtr) {
+	tsdPtr->rtmPtr = ckalloc(sizeof(ReflectedTransformMap));
+	Tcl_InitHashTable(&tsdPtr->rtmPtr->map, TCL_STRING_KEYS);
+	Tcl_CreateThreadExitHandler(DeleteThreadReflectedTransformMap, NULL);
+    }
+
+    return tsdPtr->rtmPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteThreadReflectedTransformMap --
+ *
+ *	Deletes the channel table for a thread. This procedure is invoked when
+ *	a thread is deleted. The channels have already been marked as dead, in
+ *	DeleteReflectedTransformMap().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the hash table of channels.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteThreadReflectedTransformMap(
+    ClientData clientData)	/* The per-thread data structure. */
+{
+    Tcl_HashSearch hSearch;	 /* Search variable. */
+    Tcl_HashEntry *hPtr;	 /* Search variable. */
+    Tcl_ThreadId self = Tcl_GetCurrentThread();
+    ReflectedTransformMap *rtmPtr; /* The map */
+    ForwardingResult *resultPtr;
+
+    /*
+     * The origin thread for one or more reflected channels is gone.
+     * NOTE: If this function is called due to a thread getting killed the
+     *       per-interp DeleteReflectedTransformMap is apparently not called.
+     */
+
+    /*
+     * Get the map of all channels handled by the current thread. This is a
+     * ReflectedTransformMap, but on a per-thread basis, not per-interp. Go
+     * through the channels, remove all, mark them as dead.
+     */
+
+    rtmPtr = GetThreadReflectedTransformMap();
+    for (hPtr = Tcl_FirstHashEntry(&rtmPtr->map, &hSearch);
+	    hPtr != NULL;
+	    hPtr = Tcl_FirstHashEntry(&rtmPtr->map, &hSearch)) {
+	ReflectedTransform *rtPtr = Tcl_GetHashValue(hPtr);
+
+	rtPtr->dead = 1;
+	FreeReflectedTransformArgs(rtPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    ckfree(rtmPtr);
+
+    /*
+     * Go through the list of pending results and cancel all whose events were
+     * destined for this thread. While this is in progress we block any
+     * other access to the list of pending results.
+     */
+
+    Tcl_MutexLock(&rtForwardMutex);
+
+    for (resultPtr = forwardList; resultPtr != NULL;
+	    resultPtr = resultPtr->nextPtr) {
+	ForwardingEvent *evPtr;
+	ForwardParam *paramPtr;
+
+	if (resultPtr->dst != self) {
+	    /*
+	     * Ignore results/events for other threads.
+	     */
+
+	    continue;
+	}
+
+	/*
+	 * The receiver for the event exited, before processing the event. We
+	 * detach the result now, wake the originator up and signal failure.
+	 */
+
+	evPtr = resultPtr->evPtr;
+	if (evPtr == NULL) {
+	    continue;
+	}
+	paramPtr = evPtr->param;
+
+	evPtr->resultPtr = NULL;
+	resultPtr->evPtr = NULL;
+	resultPtr->result = TCL_ERROR;
+
+	ForwardSetStaticError(paramPtr, msg_send_dstlost);
+
+	Tcl_ConditionNotify(&resultPtr->done);
+    }
+    Tcl_MutexUnlock(&rtForwardMutex);
+}
+
+static void
+ForwardOpToOwnerThread(
+    ReflectedTransform *rtPtr,	/* Channel instance */
+    ForwardedOperation op,	/* Forwarded driver operation */
+    const void *param)		/* Arguments */
+{
+    Tcl_ThreadId dst = rtPtr->thread;
+    ForwardingEvent *evPtr;
+    ForwardingResult *resultPtr;
+
+    /*
+     * We gather the lock early. This allows us to check the liveness of the
+     * channel without interference from DeleteThreadReflectedTransformMap().
+     */
+
+    Tcl_MutexLock(&rtForwardMutex);
+
+    if (rtPtr->dead) {
+	/*
+	 * The channel is marked as dead. Bail out immediately, with an
+	 * appropriate error. Do not forget to unlock the mutex on this path.
+	 */
+
+	ForwardSetStaticError((ForwardParam *) param, msg_send_dstlost);
+	Tcl_MutexUnlock(&rtForwardMutex);
+	return;
+    }
+
+    /*
+     * Create and initialize the event and data structures.
+     */
+
+    evPtr = ckalloc(sizeof(ForwardingEvent));
+    resultPtr = ckalloc(sizeof(ForwardingResult));
+
+    evPtr->event.proc = ForwardProc;
+    evPtr->resultPtr = resultPtr;
+    evPtr->op = op;
+    evPtr->rtPtr = rtPtr;
+    evPtr->param = (ForwardParam *) param;
+
+    resultPtr->src = Tcl_GetCurrentThread();
+    resultPtr->dst = dst;
+    resultPtr->dsti = rtPtr->interp;
+    resultPtr->done = NULL;
+    resultPtr->result = -1;
+    resultPtr->evPtr = evPtr;
+
+    /*
+     * Now execute the forward.
+     */
+
+    TclSpliceIn(resultPtr, forwardList);
+    /* Do not unlock here. That is done by the ConditionWait */
+
+    /*
+     * Ensure cleanup of the event if the origin thread exits while this event
+     * is pending or in progress. Exit of the destination thread is handled by
+     * DeleteThreadReflectionChannelMap(), this is set up by
+     * GetThreadReflectedTransformMap(). This is what we use the 'forwardList'
+     * (see above) for.
+     */
+
+    Tcl_CreateThreadExitHandler(SrcExitProc, evPtr);
+
+    /*
+     * Queue the event and poke the other thread's notifier.
+     */
+
+    Tcl_ThreadQueueEvent(dst, (Tcl_Event *) evPtr, TCL_QUEUE_TAIL);
+    Tcl_ThreadAlert(dst);
+
+    /*
+     * (*) Block until the other thread has either processed the transfer or
+     * rejected it.
+     */
+
+    while (resultPtr->result < 0) {
+	/*
+	 * NOTE (1): Is it possible that the current thread goes away while
+	 * waiting here? IOW Is it possible that "SrcExitProc" is called
+	 * while we are here? See complementary note (2) in "SrcExitProc"
+	 *
+	 * The ConditionWait unlocks the mutex during the wait and relocks it
+	 * immediately after.
+	 */
+
+	Tcl_ConditionWait(&resultPtr->done, &rtForwardMutex, NULL);
+    }
+
+    /*
+     * Unlink result from the forwarder list. No need to lock. Either still
+     * locked, or locked by the ConditionWait
+     */
+
+    TclSpliceOut(resultPtr, forwardList);
+
+    resultPtr->nextPtr = NULL;
+    resultPtr->prevPtr = NULL;
+
+    Tcl_MutexUnlock(&rtForwardMutex);
+    Tcl_ConditionFinalize(&resultPtr->done);
+
+    /*
+     * Kill the cleanup handler now, and the result structure as well, before
+     * returning the success code.
+     *
+     * Note: The event structure has already been deleted by the destination
+     * notifier, after it serviced the event.
+     */
+
+    Tcl_DeleteThreadExitHandler(SrcExitProc, evPtr);
+
+    ckfree(resultPtr);
+}
+
+static int
+ForwardProc(
+    Tcl_Event *evGPtr,
+    int mask)
+{
+    /*
+     * Notes regarding access to the referenced data.
+     *
+     * In principle the data belongs to the originating thread (see
+     * evPtr->src), however this thread is currently blocked at (*), i.e.
+     * quiescent. Because of this we can treat the data as belonging to us,
+     * without fear of race conditions. I.e. we can read and write as we like.
+     *
+     * The only thing we cannot be sure of is the resultPtr. This can be be
+     * NULLed if the originating thread went away while the event is handled
+     * here now.
+     */
+
+    ForwardingEvent *evPtr = (ForwardingEvent *) evGPtr;
+    ForwardingResult *resultPtr = evPtr->resultPtr;
+    ReflectedTransform *rtPtr = evPtr->rtPtr;
+    Tcl_Interp *interp = rtPtr->interp;
+    ForwardParam *paramPtr = evPtr->param;
+    Tcl_Obj *resObj = NULL;	/* Interp result of InvokeTclMethod */
+    ReflectedTransformMap *rtmPtr;
+				/* Map of reflected channels with handlers in
+				 * this interp. */
+    Tcl_HashEntry *hPtr;	/* Entry in the above map */
+
+    /*
+     * Ignore the event if no one is waiting for its result anymore.
+     */
+
+    if (!resultPtr) {
+	return 1;
+    }
+
+    paramPtr->base.code = TCL_OK;
+    paramPtr->base.msgStr = NULL;
+    paramPtr->base.mustFree = 0;
+
+    switch (evPtr->op) {
+	/*
+	 * The destination thread for the following operations is
+	 * rtPtr->thread, which contains rtPtr->interp, the interp we have to
+	 * call upon for the driver.
+	 */
+
+    case ForwardedClose:
+	/*
+	 * No parameters/results.
+	 */
+
+	if (InvokeTclMethod(rtPtr, "finalize", NULL, NULL,
+		&resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	}
+
+	/*
+	 * Freeing is done here, in the origin thread, because the argv[]
+	 * objects belong to this thread. Deallocating them in a different
+	 * thread is not allowed
+	 */
+
+	/*
+	 * Remove the channel from the map before releasing the memory, to
+	 * prevent future accesses (like by 'postevent') from finding and
+	 * dereferencing a dangling pointer.
+	 */
+
+	rtmPtr = GetReflectedTransformMap(interp);
+	hPtr = Tcl_FindHashEntry(&rtmPtr->map, TclGetString(rtPtr->handle));
+	Tcl_DeleteHashEntry(hPtr);
+
+	/*
+	 * In a threaded interpreter we manage a per-thread map as well, to
+	 * allow us to survive if the script level pulls the rug out under a
+	 * channel by deleting the owning thread.
+	 */
+
+	rtmPtr = GetThreadReflectedTransformMap();
+	hPtr = Tcl_FindHashEntry(&rtmPtr->map, TclGetString(rtPtr->handle));
+	Tcl_DeleteHashEntry(hPtr);
+
+	FreeReflectedTransformArgs(rtPtr);
+	break;
+
+    case ForwardedInput: {
+	Tcl_Obj *bufObj = Tcl_NewByteArrayObj((unsigned char *)
+		paramPtr->transform.buf, paramPtr->transform.size);
+	Tcl_IncrRefCount(bufObj);
+
+	if (InvokeTclMethod(rtPtr, "read", bufObj, NULL, &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->transform.size = -1;
+	} else {
+	    /*
+	     * Process a regular return. Contains the transformation result.
+	     * Sent it back to the request originator.
+	     */
+
+	    int bytec;		/* Number of returned bytes */
+	    unsigned char *bytev;
+				/* Array of returned bytes */
+
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+	    paramPtr->transform.size = bytec;
+
+	    if (bytec > 0) {
+		paramPtr->transform.buf = ckalloc(bytec);
+		memcpy(paramPtr->transform.buf, bytev, (size_t)bytec);
+	    } else {
+		paramPtr->transform.buf = NULL;
+	    }
+	}
+
+	Tcl_DecrRefCount(bufObj);
+	break;
+    }
+
+    case ForwardedOutput: {
+	Tcl_Obj *bufObj = Tcl_NewByteArrayObj((unsigned char *)
+		paramPtr->transform.buf, paramPtr->transform.size);
+	Tcl_IncrRefCount(bufObj);
+
+	if (InvokeTclMethod(rtPtr, "write", bufObj, NULL, &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->transform.size = -1;
+	} else {
+	    /*
+	     * Process a regular return. Contains the transformation result.
+	     * Sent it back to the request originator.
+	     */
+
+	    int bytec;		/* Number of returned bytes */
+	    unsigned char *bytev;
+				/* Array of returned bytes */
+
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+	    paramPtr->transform.size = bytec;
+
+	    if (bytec > 0) {
+		paramPtr->transform.buf = ckalloc(bytec);
+		memcpy(paramPtr->transform.buf, bytev, (size_t)bytec);
+	    } else {
+		paramPtr->transform.buf = NULL;
+	    }
+	}
+
+	Tcl_DecrRefCount(bufObj);
+	break;
+    }
+
+    case ForwardedDrain:
+	if (InvokeTclMethod(rtPtr, "drain", NULL, NULL, &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->transform.size = -1;
+	} else {
+	    /*
+	     * Process a regular return. Contains the transformation result.
+	     * Sent it back to the request originator.
+	     */
+
+	    int bytec;		/* Number of returned bytes */
+	    unsigned char *bytev; /* Array of returned bytes */
+
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+	    paramPtr->transform.size = bytec;
+
+	    if (bytec > 0) {
+		paramPtr->transform.buf = ckalloc(bytec);
+		memcpy(paramPtr->transform.buf, bytev, (size_t)bytec);
+	    } else {
+		paramPtr->transform.buf = NULL;
+	    }
+	}
+	break;
+
+    case ForwardedFlush:
+	if (InvokeTclMethod(rtPtr, "flush", NULL, NULL, &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->transform.size = -1;
+	} else {
+	    /*
+	     * Process a regular return. Contains the transformation result.
+	     * Sent it back to the request originator.
+	     */
+
+	    int bytec;		/* Number of returned bytes */
+	    unsigned char *bytev;
+				/* Array of returned bytes */
+
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+
+	    paramPtr->transform.size = bytec;
+
+	    if (bytec > 0) {
+		paramPtr->transform.buf = ckalloc(bytec);
+		memcpy(paramPtr->transform.buf, bytev, (size_t)bytec);
+	    } else {
+		paramPtr->transform.buf = NULL;
+	    }
+	}
+	break;
+
+    case ForwardedClear:
+	(void) InvokeTclMethod(rtPtr, "clear", NULL, NULL, NULL);
+	break;
+
+    case ForwardedLimit:
+	if (InvokeTclMethod(rtPtr, "limit?", NULL, NULL, &resObj) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, resObj);
+	    paramPtr->limit.max = -1;
+	} else if (Tcl_GetIntFromObj(interp, resObj,
+		&paramPtr->limit.max) != TCL_OK) {
+	    ForwardSetObjError(paramPtr, MarshallError(interp));
+	    paramPtr->limit.max = -1;
+	}
+	break;
+
+    default:
+	/*
+	 * Bad operation code.
+	 */
+	Tcl_Panic("Bad operation code in ForwardProc");
+	break;
+    }
+
+    /*
+     * Remove the reference we held on the result of the invoke, if we had
+     * such.
+     */
+
+    if (resObj != NULL) {
+	Tcl_DecrRefCount(resObj);
+    }
+
+    if (resultPtr) {
+	/*
+	 * Report the forwarding result synchronously to the waiting caller.
+	 * This unblocks (*) as well. This is wrapped into a conditional
+	 * because the caller may have exited in the mean time.
+	 */
+
+	Tcl_MutexLock(&rtForwardMutex);
+	resultPtr->result = TCL_OK;
+	Tcl_ConditionNotify(&resultPtr->done);
+	Tcl_MutexUnlock(&rtForwardMutex);
+    }
+
+    return 1;
+}
+
+static void
+SrcExitProc(
+    ClientData clientData)
+{
+    ForwardingEvent *evPtr = clientData;
+    ForwardingResult *resultPtr;
+    ForwardParam *paramPtr;
+
+    /*
+     * NOTE (2): Can this handler be called with the originator blocked?
+     */
+
+    /*
+     * The originator for the event exited. It is not sure if this can happen,
+     * as the originator should be blocked at (*) while the event is in
+     * transit/pending.
+     *
+     * We make sure that the event cannot refer to the result anymore, remove
+     * it from the list of pending results and free the structure. Locking the
+     * access ensures that we cannot get in conflict with "ForwardProc",
+     * should it already execute the event.
+     */
+
+    Tcl_MutexLock(&rtForwardMutex);
+
+    resultPtr = evPtr->resultPtr;
+    paramPtr = evPtr->param;
+
+    evPtr->resultPtr = NULL;
+    resultPtr->evPtr = NULL;
+    resultPtr->result = TCL_ERROR;
+
+    ForwardSetStaticError(paramPtr, msg_send_originlost);
+
+    /*
+     * See below: TclSpliceOut(resultPtr, forwardList);
+     */
+
+    Tcl_MutexUnlock(&rtForwardMutex);
+
+    /*
+     * This unlocks (*). The structure will be spliced out and freed by
+     * "ForwardProc". Maybe.
+     */
+
+    Tcl_ConditionNotify(&resultPtr->done);
+}
+
+static void
+ForwardSetObjError(
+    ForwardParam *paramPtr,
+    Tcl_Obj *obj)
+{
+    int len;
+    const char *msgStr = TclGetStringFromObj(obj, &len);
+
+    len++;
+    ForwardSetDynamicError(paramPtr, ckalloc(len));
+    memcpy(paramPtr->base.msgStr, msgStr, (unsigned) len);
+}
+#endif /* TCL_THREADS */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerKill --
+ *
+ *	Timer management. Removes the internal timer if it exists.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerKill(
+    ReflectedTransform *rtPtr)
+{
+    if (rtPtr->timer == NULL) {
+	return;
+    }
+
+    /*
+     * Delete an existing flush-out timer, prevent it from firing on a
+     * removed/dead channel.
+     */
+
+    Tcl_DeleteTimerHandler(rtPtr->timer);
+    rtPtr->timer = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerSetup --
+ *
+ *	Timer management. Creates the internal timer if it does not exist.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerSetup(
+    ReflectedTransform *rtPtr)
+{
+    if (rtPtr->timer != NULL) {
+	return;
+    }
+
+    rtPtr->timer = Tcl_CreateTimerHandler(SYNTHETIC_EVENT_TIME,
+	    TimerRun, rtPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerRun --
+ *
+ *	Called by the notifier (-> timer) to flush out information waiting in
+ *	channel buffers.
+ *
+ * Side effects:
+ *	As of 'Tcl_NotifyChannel'.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerRun(
+    ClientData clientData)
+{
+    ReflectedTransform *rtPtr = clientData;
+
+    rtPtr->timer = NULL;
+    Tcl_NotifyChannel(rtPtr->chan, TCL_READABLE);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultInit --
+ *
+ *	Initializes the specified buffer structure. The structure will contain
+ *	valid information for an emtpy buffer.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ResultInit(
+    ResultBuffer *rPtr)		/* Reference to the structure to
+				 * initialize. */
+{
+    rPtr->used = 0;
+    rPtr->allocated = 0;
+    rPtr->buf = NULL;
+}
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultClear --
+ *
+ *	Deallocates any memory allocated by 'ResultAdd'.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ResultClear(
+    ResultBuffer *rPtr)		/* Reference to the buffer to clear out */
+{
+    rPtr->used = 0;
+
+    if (!rPtr->allocated) {
+	return;
+    }
+
+    ckfree(rPtr->buf);
+    rPtr->buf = NULL;
+    rPtr->allocated = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultAdd --
+ *
+ *	Adds the bytes in the specified array to the buffer, by appending it.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ResultAdd(
+    ResultBuffer *rPtr,		/* The buffer to extend */
+    unsigned char *buf,		/* The buffer to read from */
+    int toWrite)		/* The number of bytes in 'buf' */
+{
+    if ((rPtr->used + toWrite + 1) > rPtr->allocated) {
+	/*
+	 * Extension of the internal buffer is required.
+	 * NOTE: Currently linear. Should be doubling to amortize.
+	 */
+
+	if (rPtr->allocated == 0) {
+	    rPtr->allocated = toWrite + RB_INCREMENT;
+	    rPtr->buf = UCHARP(ckalloc(rPtr->allocated));
+	} else {
+	    rPtr->allocated += toWrite + RB_INCREMENT;
+	    rPtr->buf = UCHARP(ckrealloc((char *) rPtr->buf,
+		    rPtr->allocated));
+	}
+    }
+
+    /*
+     * Now copy data.
+     */
+
+    memcpy(rPtr->buf + rPtr->used, buf, toWrite);
+    rPtr->used += toWrite;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultCopy --
+ *
+ *	Copies the requested number of bytes from the buffer into the
+ *	specified array and removes them from the buffer afterward. Copies
+ *	less if there is not enough data in the buffer.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	The number of actually copied bytes, possibly less than 'toRead'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ResultCopy(
+    ResultBuffer *rPtr,		/* The buffer to read from */
+    unsigned char *buf,		/* The buffer to copy into */
+    int toRead)			/* Number of requested bytes */
+{
+    int copied;
+
+    if (rPtr->used == 0) {
+	/*
+	 * Nothing to copy in the case of an empty buffer.
+	 */
+
+	copied = 0;
+    } else if (rPtr->used == toRead) {
+	/*
+	 * We have just enough. Copy everything to the caller.
+	 */
+
+	memcpy(buf, rPtr->buf, toRead);
+	rPtr->used = 0;
+	copied = toRead;
+    } else if (rPtr->used > toRead) {
+	/*
+	 * The internal buffer contains more than requested. Copy the
+	 * requested subset to the caller, and shift the remaining bytes down.
+	 */
+
+	memcpy(buf, rPtr->buf, toRead);
+	memmove(rPtr->buf, rPtr->buf + toRead, rPtr->used - toRead);
+
+	rPtr->used -= toRead;
+	copied = toRead;
+    } else {
+	/*
+	 * There is not enough in the buffer to satisfy the caller, so take
+	 * everything.
+	 */
+
+	memcpy(buf, rPtr->buf, rPtr->used);
+	toRead = rPtr->used;
+	rPtr->used = 0;
+	copied = toRead;
+    }
+
+    /* -- common postwork code ------- */
+
+    return copied;
+}
+
+static int
+TransformRead(
+    ReflectedTransform *rtPtr,
+    int *errorCodePtr,
+    Tcl_Obj *bufObj)
+{
+    Tcl_Obj *resObj;
+    int bytec;			/* Number of returned bytes */
+    unsigned char *bytev;	/* Array of returned bytes */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.transform.buf = (char *) Tcl_GetByteArrayFromObj(bufObj,
+		&(p.transform.size));
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedInput, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rtPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+	ResultAdd(&rtPtr->result, UCHARP(p.transform.buf), p.transform.size);
+	ckfree(p.transform.buf);
+	return 1;
+    }
+#endif /* TCL_THREADS */
+
+    /* ASSERT: rtPtr->method & FLAG(METH_READ) */
+    /* ASSERT: rtPtr->mode & TCL_READABLE */
+
+    if (InvokeTclMethod(rtPtr, "read", bufObj, NULL, &resObj) != TCL_OK) {
+	Tcl_SetChannelError(rtPtr->chan, resObj);
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	*errorCodePtr = EINVAL;
+	return 0;
+    }
+
+    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+    ResultAdd(&rtPtr->result, bytev, bytec);
+
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    return 1;
+}
+
+static int
+TransformWrite(
+    ReflectedTransform *rtPtr,
+    int *errorCodePtr,
+    unsigned char *buf,
+    int toWrite)
+{
+    Tcl_Obj *bufObj;
+    Tcl_Obj *resObj;
+    int bytec;			/* Number of returned bytes */
+    unsigned char *bytev;	/* Array of returned bytes */
+    int res;
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	p.transform.buf = (char *) buf;
+	p.transform.size = toWrite;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedOutput, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rtPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+	res = Tcl_WriteRaw(rtPtr->parent, (char *) p.transform.buf,
+		p.transform.size);
+	ckfree(p.transform.buf);
+    } else
+#endif /* TCL_THREADS */
+    {
+	/* ASSERT: rtPtr->method & FLAG(METH_WRITE) */
+	/* ASSERT: rtPtr->mode & TCL_WRITABLE */
+
+	bufObj = Tcl_NewByteArrayObj((unsigned char *) buf, toWrite);
+	Tcl_IncrRefCount(bufObj);
+	if (InvokeTclMethod(rtPtr, "write", bufObj, NULL, &resObj) != TCL_OK) {
+	    *errorCodePtr = EINVAL;
+	    Tcl_SetChannelError(rtPtr->chan, resObj);
+
+	    Tcl_DecrRefCount(bufObj);
+	    Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+
+	bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+	res = Tcl_WriteRaw(rtPtr->parent, (char *) bytev, bytec);
+
+	Tcl_DecrRefCount(bufObj);
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+    }
+
+    if (res < 0) {
+	*errorCodePtr = Tcl_GetErrno();
+	return 0;
+    }
+
+    return 1;
+}
+
+static int
+TransformDrain(
+    ReflectedTransform *rtPtr,
+    int *errorCodePtr)
+{
+    Tcl_Obj *resObj;
+    int bytec;			/* Number of returned bytes */
+    unsigned char *bytev;	/* Array of returned bytes */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedDrain, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rtPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+	ResultAdd(&rtPtr->result, UCHARP(p.transform.buf), p.transform.size);
+	ckfree(p.transform.buf);
+    } else
+#endif /* TCL_THREADS */
+    {
+	if (InvokeTclMethod(rtPtr, "drain", NULL, NULL, &resObj)!=TCL_OK) {
+	    Tcl_SetChannelError(rtPtr->chan, resObj);
+	    Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+	ResultAdd(&rtPtr->result, bytev, bytec);
+
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+    }
+
+    rtPtr->readIsDrained = 1;
+    return 1;
+}
+
+static int
+TransformFlush(
+    ReflectedTransform *rtPtr,
+    int *errorCodePtr,
+    int op)
+{
+    Tcl_Obj *resObj;
+    int bytec;			/* Number of returned bytes */
+    unsigned char *bytev;	/* Array of returned bytes */
+    int res;
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedFlush, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rtPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+	if (op == FLUSH_WRITE) {
+	    res = Tcl_WriteRaw(rtPtr->parent, (char *) p.transform.buf,
+		    p.transform.size);
+	} else {
+	    res = 0;
+	}
+	ckfree(p.transform.buf);
+    } else
+#endif /* TCL_THREADS */
+    {
+	if (InvokeTclMethod(rtPtr, "flush", NULL, NULL, &resObj)!=TCL_OK) {
+	    Tcl_SetChannelError(rtPtr->chan, resObj);
+	    Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	if (op == FLUSH_WRITE) {
+	    bytev = Tcl_GetByteArrayFromObj(resObj, &bytec);
+	    res = Tcl_WriteRaw(rtPtr->parent, (char *) bytev, bytec);
+	} else {
+	    res = 0;
+	}
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+    }
+
+    if (res < 0) {
+	*errorCodePtr = Tcl_GetErrno();
+	return 0;
+    }
+
+    return 1;
+}
+
+static void
+TransformClear(
+    ReflectedTransform *rtPtr)
+{
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedClear, &p);
+	return;
+    }
+#endif /* TCL_THREADS */
+
+    /* ASSERT: rtPtr->method & FLAG(METH_READ) */
+    /* ASSERT: rtPtr->mode & TCL_READABLE */
+
+    (void) InvokeTclMethod(rtPtr, "clear", NULL, NULL, NULL);
+
+    rtPtr->readIsDrained = 0;
+    rtPtr->eofPending = 0;
+    ResultClear(&rtPtr->result);
+}
+
+static int
+TransformLimit(
+    ReflectedTransform *rtPtr,
+    int *errorCodePtr,
+    int *maxPtr)
+{
+    Tcl_Obj *resObj;
+    Tcl_InterpState sr;		/* State of handler interp */
+
+    /*
+     * Are we in the correct thread?
+     */
+
+#ifdef TCL_THREADS
+    if (rtPtr->thread != Tcl_GetCurrentThread()) {
+	ForwardParam p;
+
+	ForwardOpToOwnerThread(rtPtr, ForwardedLimit, &p);
+
+	if (p.base.code != TCL_OK) {
+	    PassReceivedError(rtPtr->chan, &p);
+	    *errorCodePtr = EINVAL;
+	    return 0;
+	}
+
+	*errorCodePtr = EOK;
+	*maxPtr = p.limit.max;
+	return 1;
+    }
+#endif
+
+    /* ASSERT: rtPtr->method & FLAG(METH_WRITE) */
+    /* ASSERT: rtPtr->mode & TCL_WRITABLE */
+
+    if (InvokeTclMethod(rtPtr, "limit?", NULL, NULL, &resObj) != TCL_OK) {
+	Tcl_SetChannelError(rtPtr->chan, resObj);
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	*errorCodePtr = EINVAL;
+	return 0;
+    }
+
+    sr = Tcl_SaveInterpState(rtPtr->interp, 0 /* Dummy */);
+
+    if (Tcl_GetIntFromObj(rtPtr->interp, resObj, maxPtr) != TCL_OK) {
+	Tcl_DecrRefCount(resObj);	/* Remove reference held from invoke */
+	Tcl_SetChannelError(rtPtr->chan, MarshallError(rtPtr->interp));
+	*errorCodePtr = EINVAL;
+
+	Tcl_RestoreInterpState(rtPtr->interp, sr);
+	return 0;
+    }
+
+    Tcl_DecrRefCount(resObj);		/* Remove reference held from invoke */
+    Tcl_RestoreInterpState(rtPtr->interp, sr);
+    return 1;
+}
+
+/* DUPLICATE of HaveVersion() in tclIO.c
+ *----------------------------------------------------------------------
+ *
+ * HaveVersion --
+ *
+ *	Return whether a channel type is (at least) of a given version.
+ *
+ * Results:
+ *	True if the minimum version is exceeded by the version actually
+ *	present.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+HaveVersion(
+    const Tcl_ChannelType *chanTypePtr,
+    Tcl_ChannelTypeVersion minimumVersion)
+{
+    Tcl_ChannelTypeVersion actualVersion = Tcl_ChannelVersion(chanTypePtr);
+
+    return PTR2INT(actualVersion) >= PTR2INT(minimumVersion);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIOSock.c b/generic/tclIOSock.c
new file mode 100644
index 0000000..6abfa60
--- /dev/null
+++ b/generic/tclIOSock.c
@@ -0,0 +1,330 @@
+/*
+ * tclIOSock.c --
+ *
+ *	Common routines used by all socket based channel types.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+#if defined(_WIN32) && defined(UNICODE)
+/*
+ * On Windows, we need to do proper Unicode->UTF-8 conversion.
+ */
+
+typedef struct {
+    int initialized;
+    Tcl_DString errorMsg; /* UTF-8 encoded error-message */
+} ThreadSpecificData;
+static Tcl_ThreadDataKey dataKey;
+
+#undef gai_strerror
+static const char *
+gai_strerror(
+    int code)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->initialized) {
+	Tcl_DStringFree(&tsdPtr->errorMsg);
+    } else {
+	tsdPtr->initialized = 1;
+    }
+    Tcl_WinTCharToUtf(gai_strerrorW(code), -1, &tsdPtr->errorMsg);
+    return Tcl_DStringValue(&tsdPtr->errorMsg);
+}
+#endif
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclSockGetPort --
+ *
+ *	Maps from a string, which could be a service name, to a port. Used by
+ *	socket creation code to get port numbers and resolve registered
+ *	service names to port numbers.
+ *
+ * Results:
+ *	A standard Tcl result. On success, the port number is returned in
+ *	portPtr. On failure, an error message is left in the interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclSockGetPort(
+    Tcl_Interp *interp,
+    const char *string,		/* Integer or service name */
+    const char *proto,		/* "tcp" or "udp", typically */
+    int *portPtr)		/* Return port number */
+{
+    struct servent *sp;		/* Protocol info for named services */
+    Tcl_DString ds;
+    const char *native;
+
+    if (Tcl_GetInt(NULL, string, portPtr) != TCL_OK) {
+	/*
+	 * Don't bother translating 'proto' to native.
+	 */
+
+	native = Tcl_UtfToExternalDString(NULL, string, -1, &ds);
+	sp = getservbyname(native, proto);		/* INTL: Native. */
+	Tcl_DStringFree(&ds);
+	if (sp != NULL) {
+	    *portPtr = ntohs((unsigned short) sp->s_port);
+	    return TCL_OK;
+	}
+    }
+    if (Tcl_GetInt(interp, string, portPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (*portPtr > 0xFFFF) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"couldn't open socket: port number too high", -1));
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSockMinimumBuffers --
+ *
+ *	Ensure minimum buffer sizes (non zero).
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Sets SO_SNDBUF and SO_RCVBUF sizes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#if !defined(_WIN32) && !defined(__CYGWIN__)
+#   define SOCKET int
+#endif
+
+int
+TclSockMinimumBuffers(
+    void *sock,			/* Socket file descriptor */
+    int size)			/* Minimum buffer size */
+{
+    int current;
+    socklen_t len;
+
+    len = sizeof(int);
+    getsockopt((SOCKET)(size_t) sock, SOL_SOCKET, SO_SNDBUF,
+	    (char *) &current, &len);
+    if (current < size) {
+	len = sizeof(int);
+	setsockopt((SOCKET)(size_t) sock, SOL_SOCKET, SO_SNDBUF,
+		(char *) &size, len);
+    }
+    len = sizeof(int);
+    getsockopt((SOCKET)(size_t) sock, SOL_SOCKET, SO_RCVBUF,
+	    (char *) &current, &len);
+    if (current < size) {
+	len = sizeof(int);
+	setsockopt((SOCKET)(size_t) sock, SOL_SOCKET, SO_RCVBUF,
+		(char *) &size, len);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateSocketAddress --
+ *
+ *	This function initializes a sockaddr structure for a host and port.
+ *
+ * Results:
+ *	1 if the host was valid, 0 if the host could not be converted to an IP
+ *	address.
+ *
+ * Side effects:
+ *	Fills in the *sockaddrPtr structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCreateSocketAddress(
+    Tcl_Interp *interp,		/* Interpreter for querying the desired socket
+				 * family */
+    struct addrinfo **addrlist,	/* Socket address list */
+    const char *host,		/* Host. NULL implies INADDR_ANY */
+    int port,			/* Port number */
+    int willBind,		/* Is this an address to bind() to or to
+				 * connect() to? */
+    const char **errorMsgPtr)	/* Place to store the error message detail, if
+				 * available. */
+{
+    struct addrinfo hints;
+    struct addrinfo *p;
+    struct addrinfo *v4head = NULL, *v4ptr = NULL;
+    struct addrinfo *v6head = NULL, *v6ptr = NULL;
+    char *native = NULL, portbuf[TCL_INTEGER_SPACE], *portstring;
+    const char *family = NULL;
+    Tcl_DString ds;
+    int result;
+
+    if (host != NULL) {
+	native = Tcl_UtfToExternalDString(NULL, host, -1, &ds);
+    }
+
+    /*
+     * Workaround for OSX's apparent inability to resolve "localhost", "0"
+     * when the loopback device is the only available network interface.
+     */
+
+    if (host != NULL && port == 0) {
+	portstring = NULL;
+    } else {
+	TclFormatInt(portbuf, port);
+	portstring = portbuf;
+    }
+
+    (void) memset(&hints, 0, sizeof(hints));
+    hints.ai_family = AF_UNSPEC;
+
+    /*
+     * Magic variable to enforce a certain address family; to be superseded
+     * by a TIP that adds explicit switches to [socket].
+     */
+
+    if (interp != NULL) {
+	family = Tcl_GetVar2(interp, "::tcl::unsupported::socketAF", NULL, 0);
+	if (family != NULL) {
+	    if (strcmp(family, "inet") == 0) {
+		hints.ai_family = AF_INET;
+	    } else if (strcmp(family, "inet6") == 0) {
+		hints.ai_family = AF_INET6;
+	    }
+	}
+    }
+
+    hints.ai_socktype = SOCK_STREAM;
+
+#if 0
+    /*
+     * We found some problems when using AI_ADDRCONFIG, e.g. on systems that
+     * have no networking besides the loopback interface and want to resolve
+     * localhost. See [Bugs 3385024, 3382419, 3382431]. As the advantage of
+     * using AI_ADDRCONFIG is probably low even in situations where it works,
+     * we'll leave it out for now. After all, it is just an optimisation.
+     *
+     * Missing on: OpenBSD, NetBSD.
+     * Causes failure when used on AIX 5.1 and HP-UX
+     */
+
+#if defined(AI_ADDRCONFIG) && !defined(_AIX) && !defined(__hpux)
+    hints.ai_flags |= AI_ADDRCONFIG;
+#endif /* AI_ADDRCONFIG && !_AIX && !__hpux */
+#endif /* 0 */
+
+    if (willBind) {
+	hints.ai_flags |= AI_PASSIVE;
+    }
+
+    result = getaddrinfo(native, portstring, &hints, addrlist);
+
+    if (host != NULL) {
+	Tcl_DStringFree(&ds);
+    }
+
+    if (result != 0) {
+	*errorMsgPtr =
+#ifdef EAI_SYSTEM	/* Doesn't exist on Windows */
+		(result == EAI_SYSTEM) ? Tcl_PosixError(interp) :
+#endif /* EAI_SYSTEM */
+		gai_strerror(result);
+        return 0;
+    }
+
+    /*
+     * Put IPv4 addresses before IPv6 addresses to maximize backwards
+     * compatibility of [fconfigure -sockname] output.
+     *
+     * There might be more elegant/efficient ways to do this.
+     */
+
+    if (willBind) {
+	for (p = *addrlist; p != NULL; p = p->ai_next) {
+	    if (p->ai_family == AF_INET) {
+		if (v4head == NULL) {
+		    v4head = p;
+		} else {
+		    v4ptr->ai_next = p;
+		}
+		v4ptr = p;
+	    } else {
+		if (v6head == NULL) {
+		    v6head = p;
+		} else {
+		    v6ptr->ai_next = p;
+		}
+		v6ptr = p;
+	    }
+	}
+	*addrlist = NULL;
+	if (v6head != NULL) {
+	    *addrlist = v6head;
+	    v6ptr->ai_next = NULL;
+	}
+	if (v4head != NULL) {
+	    v4ptr->ai_next = *addrlist;
+	    *addrlist = v4head;
+	}
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_OpenTcpServer --
+ *
+ *	Opens a TCP server socket and creates a channel around it.
+ *
+ * Results:
+ *	The channel or NULL if failed. If an error occurred, an error message
+ *	is left in the interp's result if interp is not NULL.
+ *
+ * Side effects:
+ *	Opens a server socket and creates a new channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_OpenTcpServer(
+    Tcl_Interp *interp,
+    int port,
+    const char *host,
+    Tcl_TcpAcceptProc *acceptProc,
+    ClientData callbackData)
+{
+    char portbuf[TCL_INTEGER_SPACE];
+
+    TclFormatInt(portbuf, port);
+    return Tcl_OpenTcpServerEx(interp, portbuf, host, TCL_TCPSERVER_REUSEADDR,
+	    acceptProc, callbackData);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIOUtil.c b/generic/tclIOUtil.c
new file mode 100644
index 0000000..2c389c6
--- /dev/null
+++ b/generic/tclIOUtil.c
@@ -0,0 +1,4882 @@
+/*
+ * tclIOUtil.c --
+ *
+ *	This file contains the implementation of Tcl's generic filesystem
+ *	code, which supports a pluggable filesystem architecture allowing both
+ *	platform specific filesystems and 'virtual filesystems'. All
+ *	filesystem access should go through the functions defined in this
+ *	file. Most of this code was contributed by Vince Darley.
+ *
+ *	Parts of this file are based on code contributed by Karl Lehenbauer,
+ *	Mark Diekhans and Peter da Silva.
+ *
+ * Copyright (c) 1991-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 2001-2004 Vincent Darley.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#ifdef _WIN32
+#   include "tclWinInt.h"
+#endif
+#include "tclFileSystem.h"
+
+#ifdef TCL_TEMPLOAD_NO_UNLINK
+#ifndef NO_FSTATFS
+#include <sys/statfs.h>
+#endif
+#endif
+
+/*
+ * struct FilesystemRecord --
+ *
+ * A filesystem record is used to keep track of each filesystem currently
+ * registered with the core, in a linked list.
+ */
+
+typedef struct FilesystemRecord {
+    ClientData clientData;	/* Client specific data for the new filesystem
+				 * (can be NULL) */
+    const Tcl_Filesystem *fsPtr;/* Pointer to filesystem dispatch table. */
+    struct FilesystemRecord *nextPtr;
+				/* The next filesystem registered to Tcl, or
+				 * NULL if no more. */
+    struct FilesystemRecord *prevPtr;
+				/* The previous filesystem registered to Tcl,
+				 * or NULL if no more. */
+} FilesystemRecord;
+
+/*
+ * This structure holds per-thread private copy of the current directory
+ * maintained by the global cwdPathPtr. This structure holds per-thread
+ * private copies of some global data. This way we avoid most of the
+ * synchronization calls which boosts performance, at cost of having to update
+ * this information each time the corresponding epoch counter changes.
+ */
+
+typedef struct {
+    int initialized;
+    size_t cwdPathEpoch;
+    size_t filesystemEpoch;
+    Tcl_Obj *cwdPathPtr;
+    ClientData cwdClientData;
+    FilesystemRecord *filesystemList;
+    size_t claims;
+} ThreadSpecificData;
+
+/*
+ * Prototypes for functions defined later in this file.
+ */
+
+static Tcl_NRPostProc	EvalFileCallback;
+static FilesystemRecord*FsGetFirstFilesystem(void);
+static void		FsThrExitProc(ClientData cd);
+static Tcl_Obj *	FsListMounts(Tcl_Obj *pathPtr, const char *pattern);
+static void		FsAddMountsToGlobResult(Tcl_Obj *resultPtr,
+			    Tcl_Obj *pathPtr, const char *pattern,
+			    Tcl_GlobTypeData *types);
+static void		FsUpdateCwd(Tcl_Obj *cwdObj, ClientData clientData);
+static void		FsRecacheFilesystemList(void);
+static void		Claim(void);
+static void		Disclaim(void);
+
+static void *		DivertFindSymbol(Tcl_Interp *interp,
+			    Tcl_LoadHandle loadHandle, const char *symbol);
+static void		DivertUnloadFile(Tcl_LoadHandle loadHandle);
+
+/*
+ * These form part of the native filesystem support. They are needed here
+ * because we have a few native filesystem functions (which are the same for
+ * win/unix) in this file. There is no need to place them in tclInt.h, because
+ * they are not (and should not be) used anywhere else.
+ */
+
+MODULE_SCOPE const char *const		tclpFileAttrStrings[];
+MODULE_SCOPE const TclFileAttrProcs	tclpFileAttrProcs[];
+
+/*
+ * Declare the native filesystem support. These functions should be considered
+ * private to Tcl, and should really not be called directly by any code other
+ * than this file (i.e. neither by Tcl's core nor by extensions). Similarly,
+ * the old string-based Tclp... native filesystem functions should not be
+ * called.
+ *
+ * The correct API to use now is the Tcl_FS... set of functions, which ensure
+ * correct and complete virtual filesystem support.
+ *
+ * We cannot make all of these static, since some of them are implemented in
+ * the platform-specific directories.
+ */
+
+static Tcl_FSFilesystemSeparatorProc NativeFilesystemSeparator;
+static Tcl_FSFreeInternalRepProc NativeFreeInternalRep;
+static Tcl_FSFileAttrStringsProc NativeFileAttrStrings;
+static Tcl_FSFileAttrsGetProc	NativeFileAttrsGet;
+static Tcl_FSFileAttrsSetProc	NativeFileAttrsSet;
+
+/*
+ * The only reason these functions are not static is that they are either
+ * called by code in the native (win/unix) directories or they are actually
+ * implemented in those directories. They should simply not be called by code
+ * outside Tcl's native filesystem core i.e. they should be considered
+ * 'static' to Tcl's filesystem code (if we ever built the native filesystem
+ * support into a separate code library, this could actually be enforced).
+ */
+
+Tcl_FSFilesystemPathTypeProc	TclpFilesystemPathType;
+Tcl_FSInternalToNormalizedProc	TclpNativeToNormalized;
+Tcl_FSStatProc			TclpObjStat;
+Tcl_FSAccessProc		TclpObjAccess;
+Tcl_FSMatchInDirectoryProc	TclpMatchInDirectory;
+Tcl_FSChdirProc			TclpObjChdir;
+Tcl_FSLstatProc			TclpObjLstat;
+Tcl_FSCopyFileProc		TclpObjCopyFile;
+Tcl_FSDeleteFileProc		TclpObjDeleteFile;
+Tcl_FSRenameFileProc		TclpObjRenameFile;
+Tcl_FSCreateDirectoryProc	TclpObjCreateDirectory;
+Tcl_FSCopyDirectoryProc		TclpObjCopyDirectory;
+Tcl_FSRemoveDirectoryProc	TclpObjRemoveDirectory;
+Tcl_FSUnloadFileProc		TclpUnloadFile;
+Tcl_FSLinkProc			TclpObjLink;
+Tcl_FSListVolumesProc		TclpObjListVolumes;
+
+/*
+ * Define the native filesystem dispatch table. If necessary, it is ok to make
+ * this non-static, but it should only be accessed by the functions actually
+ * listed within it (or perhaps other helper functions of them). Anything
+ * which is not part of this 'native filesystem implementation' should not be
+ * delving inside here!
+ */
+
+const Tcl_Filesystem tclNativeFilesystem = {
+    "native",
+    sizeof(Tcl_Filesystem),
+    TCL_FILESYSTEM_VERSION_2,
+    TclNativePathInFilesystem,
+    TclNativeDupInternalRep,
+    NativeFreeInternalRep,
+    TclpNativeToNormalized,
+    TclNativeCreateNativeRep,
+    TclpObjNormalizePath,
+    TclpFilesystemPathType,
+    NativeFilesystemSeparator,
+    TclpObjStat,
+    TclpObjAccess,
+    TclpOpenFileChannel,
+    TclpMatchInDirectory,
+    TclpUtime,
+#ifndef S_IFLNK
+    NULL,
+#else
+    TclpObjLink,
+#endif /* S_IFLNK */
+    TclpObjListVolumes,
+    NativeFileAttrStrings,
+    NativeFileAttrsGet,
+    NativeFileAttrsSet,
+    TclpObjCreateDirectory,
+    TclpObjRemoveDirectory,
+    TclpObjDeleteFile,
+    TclpObjCopyFile,
+    TclpObjRenameFile,
+    TclpObjCopyDirectory,
+    TclpObjLstat,
+    /* Needs casts since we're using version_2. */
+    (Tcl_FSLoadFileProc *) TclpDlopen,
+    (Tcl_FSGetCwdProc *) TclpGetNativeCwd,
+    TclpObjChdir
+};
+
+/*
+ * Define the tail of the linked list. Note that for unconventional uses of
+ * Tcl without a native filesystem, we may in the future wish to modify the
+ * current approach of hard-coding the native filesystem in the lookup list
+ * 'filesystemList' below.
+ *
+ * We initialize the record so that it thinks one file uses it. This means it
+ * will never be freed.
+ */
+
+static FilesystemRecord nativeFilesystemRecord = {
+    NULL,
+    &tclNativeFilesystem,
+    NULL,
+    NULL
+};
+
+/*
+ * This is incremented each time we modify the linked list of filesystems. Any
+ * time it changes, all cached filesystem representations are suspect and must
+ * be freed. For multithreading builds, change of the filesystem epoch will
+ * trigger cache cleanup in all threads.
+ */
+
+static size_t theFilesystemEpoch = 1;
+
+/*
+ * Stores the linked list of filesystems. A 1:1 copy of this list is also
+ * maintained in the TSD for each thread. This is to avoid synchronization
+ * issues.
+ */
+
+static FilesystemRecord *filesystemList = &nativeFilesystemRecord;
+TCL_DECLARE_MUTEX(filesystemMutex)
+
+/*
+ * Used to implement Tcl_FSGetCwd in a file-system independent way.
+ */
+
+static Tcl_Obj *cwdPathPtr = NULL;
+static size_t cwdPathEpoch = 0;
+static ClientData cwdClientData = NULL;
+TCL_DECLARE_MUTEX(cwdMutex)
+
+static Tcl_ThreadDataKey fsDataKey;
+
+/*
+ * One of these structures is used each time we successfully load a file from
+ * a file system by way of making a temporary copy of the file on the native
+ * filesystem. We need to store both the actual unloadProc/clientData
+ * combination which was used, and the original and modified filenames, so
+ * that we can correctly undo the entire operation when we want to unload the
+ * code.
+ */
+
+typedef struct FsDivertLoad {
+    Tcl_LoadHandle loadHandle;
+    Tcl_FSUnloadFileProc *unloadProcPtr;
+    Tcl_Obj *divertedFile;
+    const Tcl_Filesystem *divertedFilesystem;
+    ClientData divertedFileNativeRep;
+} FsDivertLoad;
+
+/*
+ * The following functions are obsolete string based APIs, and should be
+ * removed in a future release (Tcl 9 would be a good time).
+ */
+
+/* Obsolete */
+int
+Tcl_Stat(
+    const char *path,		/* Path of file to stat (in current CP). */
+    struct stat *oldStyleBuf)	/* Filled with results of stat call. */
+{
+    int ret;
+    Tcl_StatBuf buf;
+    Tcl_Obj *pathPtr = Tcl_NewStringObj(path,-1);
+
+    Tcl_IncrRefCount(pathPtr);
+    ret = Tcl_FSStat(pathPtr, &buf);
+    Tcl_DecrRefCount(pathPtr);
+    if (ret != -1) {
+#ifndef TCL_WIDE_INT_IS_LONG
+	Tcl_WideInt tmp1, tmp2, tmp3 = 0;
+
+# define OUT_OF_RANGE(x) \
+	(((Tcl_WideInt)(x)) < Tcl_LongAsWide(LONG_MIN) || \
+	 ((Tcl_WideInt)(x)) > Tcl_LongAsWide(LONG_MAX))
+# define OUT_OF_URANGE(x) \
+	(((Tcl_WideUInt)(x)) > ((Tcl_WideUInt)ULONG_MAX))
+
+	/*
+	 * Perform the result-buffer overflow check manually.
+	 *
+	 * Note that ino_t/ino64_t is unsigned...
+	 *
+	 * Workaround gcc warning of "comparison is always false due to
+	 * limited range of data type" by assigning to tmp var of type
+	 * Tcl_WideInt.
+	 */
+
+	tmp1 = (Tcl_WideInt) buf.st_ino;
+	tmp2 = (Tcl_WideInt) buf.st_size;
+#ifdef HAVE_STRUCT_STAT_ST_BLOCKS
+	tmp3 = (Tcl_WideInt) buf.st_blocks;
+#endif
+
+	if (OUT_OF_URANGE(tmp1) || OUT_OF_RANGE(tmp2) || OUT_OF_RANGE(tmp3)) {
+#if defined(EFBIG)
+	    errno = EFBIG;
+#elif defined(EOVERFLOW)
+	    errno = EOVERFLOW;
+#else
+#error "What status should be returned for file size out of range?"
+#endif
+	    return -1;
+	}
+
+#   undef OUT_OF_RANGE
+#   undef OUT_OF_URANGE
+#endif /* !TCL_WIDE_INT_IS_LONG */
+
+	/*
+	 * Copy across all supported fields, with possible type coercions on
+	 * those fields that change between the normal and lf64 versions of
+	 * the stat structure (on Solaris at least). This is slow when the
+	 * structure sizes coincide, but that's what you get for using an
+	 * obsolete interface.
+	 */
+
+	oldStyleBuf->st_mode	= buf.st_mode;
+	oldStyleBuf->st_ino	= (ino_t) buf.st_ino;
+	oldStyleBuf->st_dev	= buf.st_dev;
+	oldStyleBuf->st_rdev	= buf.st_rdev;
+	oldStyleBuf->st_nlink	= buf.st_nlink;
+	oldStyleBuf->st_uid	= buf.st_uid;
+	oldStyleBuf->st_gid	= buf.st_gid;
+	oldStyleBuf->st_size	= (off_t) buf.st_size;
+	oldStyleBuf->st_atime	= buf.st_atime;
+	oldStyleBuf->st_mtime	= buf.st_mtime;
+	oldStyleBuf->st_ctime	= buf.st_ctime;
+#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
+	oldStyleBuf->st_blksize	= buf.st_blksize;
+#endif
+#ifdef HAVE_STRUCT_STAT_ST_BLOCKS
+#ifdef HAVE_BLKCNT_T
+	oldStyleBuf->st_blocks	= (blkcnt_t) buf.st_blocks;
+#else
+	oldStyleBuf->st_blocks	= (unsigned long) buf.st_blocks;
+#endif
+#endif
+    }
+    return ret;
+}
+
+/* Obsolete */
+int
+Tcl_Access(
+    const char *path,		/* Path of file to access (in current CP). */
+    int mode)			/* Permission setting. */
+{
+    int ret;
+    Tcl_Obj *pathPtr = Tcl_NewStringObj(path,-1);
+
+    Tcl_IncrRefCount(pathPtr);
+    ret = Tcl_FSAccess(pathPtr,mode);
+    Tcl_DecrRefCount(pathPtr);
+
+    return ret;
+}
+
+/* Obsolete */
+Tcl_Channel
+Tcl_OpenFileChannel(
+    Tcl_Interp *interp,		/* Interpreter for error reporting; can be
+				 * NULL. */
+    const char *path,		/* Name of file to open. */
+    const char *modeString,	/* A list of POSIX open modes or a string such
+				 * as "rw". */
+    int permissions)		/* If the open involves creating a file, with
+				 * what modes to create it? */
+{
+    Tcl_Channel ret;
+    Tcl_Obj *pathPtr = Tcl_NewStringObj(path,-1);
+
+    Tcl_IncrRefCount(pathPtr);
+    ret = Tcl_FSOpenFileChannel(interp, pathPtr, modeString, permissions);
+    Tcl_DecrRefCount(pathPtr);
+
+    return ret;
+}
+
+/* Obsolete */
+int
+Tcl_Chdir(
+    const char *dirName)
+{
+    int ret;
+    Tcl_Obj *pathPtr = Tcl_NewStringObj(dirName,-1);
+    Tcl_IncrRefCount(pathPtr);
+    ret = Tcl_FSChdir(pathPtr);
+    Tcl_DecrRefCount(pathPtr);
+    return ret;
+}
+
+/* Obsolete */
+char *
+Tcl_GetCwd(
+    Tcl_Interp *interp,
+    Tcl_DString *cwdPtr)
+{
+    Tcl_Obj *cwd = Tcl_FSGetCwd(interp);
+
+    if (cwd == NULL) {
+	return NULL;
+    }
+    Tcl_DStringInit(cwdPtr);
+    TclDStringAppendObj(cwdPtr, cwd);
+    Tcl_DecrRefCount(cwd);
+    return Tcl_DStringValue(cwdPtr);
+}
+
+/* Obsolete */
+int
+Tcl_EvalFile(
+    Tcl_Interp *interp,		/* Interpreter in which to process file. */
+    const char *fileName)	/* Name of file to process. Tilde-substitution
+				 * will be performed on this name. */
+{
+    int ret;
+    Tcl_Obj *pathPtr = Tcl_NewStringObj(fileName,-1);
+
+    Tcl_IncrRefCount(pathPtr);
+    ret = Tcl_FSEvalFile(interp, pathPtr);
+    Tcl_DecrRefCount(pathPtr);
+    return ret;
+}
+
+/*
+ * Now move on to the basic filesystem implementation.
+ */
+
+static void
+FsThrExitProc(
+    ClientData cd)
+{
+    ThreadSpecificData *tsdPtr = cd;
+    FilesystemRecord *fsRecPtr = NULL, *tmpFsRecPtr = NULL;
+
+    /*
+     * Trash the cwd copy.
+     */
+
+    if (tsdPtr->cwdPathPtr != NULL) {
+	Tcl_DecrRefCount(tsdPtr->cwdPathPtr);
+	tsdPtr->cwdPathPtr = NULL;
+    }
+    if (tsdPtr->cwdClientData != NULL) {
+	NativeFreeInternalRep(tsdPtr->cwdClientData);
+    }
+
+    /*
+     * Trash the filesystems cache.
+     */
+
+    fsRecPtr = tsdPtr->filesystemList;
+    while (fsRecPtr != NULL) {
+	tmpFsRecPtr = fsRecPtr->nextPtr;
+	fsRecPtr->fsPtr = NULL;
+	ckfree(fsRecPtr);
+	fsRecPtr = tmpFsRecPtr;
+    }
+    tsdPtr->filesystemList = NULL;
+    tsdPtr->initialized = 0;
+}
+
+int
+TclFSCwdIsNative(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    if (tsdPtr->cwdClientData != NULL) {
+	return 1;
+    } else {
+	return 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFSCwdPointerEquals --
+ *
+ *	Check whether the current working directory is equal to the path
+ *	given.
+ *
+ * Results:
+ *	1 (equal) or 0 (un-equal) as appropriate.
+ *
+ * Side effects:
+ *	If the paths are equal, but are not the same object, this method will
+ *	modify the given pathPtrPtr to refer to the same object. In this case
+ *	the object pointed to by pathPtrPtr will have its refCount
+ *	decremented, and it will be adjusted to point to the cwd (with a new
+ *	refCount).
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFSCwdPointerEquals(
+    Tcl_Obj **pathPtrPtr)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    Tcl_MutexLock(&cwdMutex);
+    if (tsdPtr->cwdPathPtr == NULL
+	    || tsdPtr->cwdPathEpoch != cwdPathEpoch) {
+	if (tsdPtr->cwdPathPtr != NULL) {
+	    Tcl_DecrRefCount(tsdPtr->cwdPathPtr);
+	}
+	if (tsdPtr->cwdClientData != NULL) {
+	    NativeFreeInternalRep(tsdPtr->cwdClientData);
+	}
+	if (cwdPathPtr == NULL) {
+	    tsdPtr->cwdPathPtr = NULL;
+	} else {
+	    tsdPtr->cwdPathPtr = Tcl_DuplicateObj(cwdPathPtr);
+	    Tcl_IncrRefCount(tsdPtr->cwdPathPtr);
+	}
+	if (cwdClientData == NULL) {
+	    tsdPtr->cwdClientData = NULL;
+	} else {
+	    tsdPtr->cwdClientData = TclNativeDupInternalRep(cwdClientData);
+	}
+	tsdPtr->cwdPathEpoch = cwdPathEpoch;
+    }
+    Tcl_MutexUnlock(&cwdMutex);
+
+    if (tsdPtr->initialized == 0) {
+	Tcl_CreateThreadExitHandler(FsThrExitProc, tsdPtr);
+	tsdPtr->initialized = 1;
+    }
+
+    if (pathPtrPtr == NULL) {
+	return (tsdPtr->cwdPathPtr == NULL);
+    }
+
+    if (tsdPtr->cwdPathPtr == *pathPtrPtr) {
+	return 1;
+    } else {
+	int len1, len2;
+	const char *str1, *str2;
+
+	str1 = TclGetStringFromObj(tsdPtr->cwdPathPtr, &len1);
+	str2 = TclGetStringFromObj(*pathPtrPtr, &len2);
+	if ((len1 == len2) && !memcmp(str1, str2, len1)) {
+	    /*
+	     * They are equal, but different objects. Update so they will be
+	     * the same object in the future.
+	     */
+
+	    Tcl_DecrRefCount(*pathPtrPtr);
+	    *pathPtrPtr = tsdPtr->cwdPathPtr;
+	    Tcl_IncrRefCount(*pathPtrPtr);
+	    return 1;
+	} else {
+	    return 0;
+	}
+    }
+}
+
+static void
+FsRecacheFilesystemList(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+    FilesystemRecord *fsRecPtr, *tmpFsRecPtr = NULL, *toFree = NULL, *list;
+
+    /*
+     * Trash the current cache.
+     */
+
+    fsRecPtr = tsdPtr->filesystemList;
+    while (fsRecPtr != NULL) {
+	tmpFsRecPtr = fsRecPtr->nextPtr;
+	fsRecPtr->nextPtr = toFree;
+	toFree = fsRecPtr;
+	fsRecPtr = tmpFsRecPtr;
+    }
+
+    /*
+     * Locate tail of the global filesystem list.
+     */
+
+    Tcl_MutexLock(&filesystemMutex);
+    fsRecPtr = filesystemList;
+    while (fsRecPtr != NULL) {
+	tmpFsRecPtr = fsRecPtr;
+	fsRecPtr = fsRecPtr->nextPtr;
+    }
+
+    /*
+     * Refill the cache honouring the order.
+     */
+
+    list = NULL;
+    fsRecPtr = tmpFsRecPtr;
+    while (fsRecPtr != NULL) {
+	tmpFsRecPtr = ckalloc(sizeof(FilesystemRecord));
+	*tmpFsRecPtr = *fsRecPtr;
+	tmpFsRecPtr->nextPtr = list;
+	tmpFsRecPtr->prevPtr = NULL;
+	list = tmpFsRecPtr;
+	fsRecPtr = fsRecPtr->prevPtr;
+    }
+    tsdPtr->filesystemList = list;
+    tsdPtr->filesystemEpoch = theFilesystemEpoch;
+    Tcl_MutexUnlock(&filesystemMutex);
+
+    while (toFree) {
+	FilesystemRecord *next = toFree->nextPtr;
+
+	toFree->fsPtr = NULL;
+	ckfree(toFree);
+	toFree = next;
+    }
+
+    /*
+     * Make sure the above gets released on thread exit.
+     */
+
+    if (tsdPtr->initialized == 0) {
+	Tcl_CreateThreadExitHandler(FsThrExitProc, tsdPtr);
+	tsdPtr->initialized = 1;
+    }
+}
+
+static FilesystemRecord *
+FsGetFirstFilesystem(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+    if (tsdPtr->filesystemList == NULL || ((tsdPtr->claims == 0)
+	    && (tsdPtr->filesystemEpoch != theFilesystemEpoch))) {
+	FsRecacheFilesystemList();
+    }
+    return tsdPtr->filesystemList;
+}
+
+/*
+ * The epoch can be changed by filesystems being added or removed, by changing
+ * the "system encoding" and by env(HOME) changing.
+ */
+
+int
+TclFSEpochOk(
+    size_t filesystemEpoch)
+{
+    return (filesystemEpoch == 0 || filesystemEpoch == theFilesystemEpoch);
+}
+
+static void
+Claim(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    tsdPtr->claims++;
+}
+
+static void
+Disclaim(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    tsdPtr->claims--;
+}
+
+size_t
+TclFSEpoch(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    return tsdPtr->filesystemEpoch;
+}
+
+/*
+ * If non-NULL, clientData is owned by us and must be freed later.
+ */
+
+static void
+FsUpdateCwd(
+    Tcl_Obj *cwdObj,
+    ClientData clientData)
+{
+    int len;
+    const char *str = NULL;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    if (cwdObj != NULL) {
+	str = TclGetStringFromObj(cwdObj, &len);
+    }
+
+    Tcl_MutexLock(&cwdMutex);
+    if (cwdPathPtr != NULL) {
+	Tcl_DecrRefCount(cwdPathPtr);
+    }
+    if (cwdClientData != NULL) {
+	NativeFreeInternalRep(cwdClientData);
+    }
+
+    if (cwdObj == NULL) {
+	cwdPathPtr = NULL;
+	cwdClientData = NULL;
+    } else {
+	/*
+	 * This must be stored as string obj!
+	 */
+
+	cwdPathPtr = Tcl_NewStringObj(str, len);
+	Tcl_IncrRefCount(cwdPathPtr);
+	cwdClientData = TclNativeDupInternalRep(clientData);
+    }
+
+    if (++cwdPathEpoch == 0) {
+	++cwdPathEpoch;
+    }
+    tsdPtr->cwdPathEpoch = cwdPathEpoch;
+    Tcl_MutexUnlock(&cwdMutex);
+
+    if (tsdPtr->cwdPathPtr) {
+	Tcl_DecrRefCount(tsdPtr->cwdPathPtr);
+    }
+    if (tsdPtr->cwdClientData) {
+	NativeFreeInternalRep(tsdPtr->cwdClientData);
+    }
+
+    if (cwdObj == NULL) {
+	tsdPtr->cwdPathPtr = NULL;
+	tsdPtr->cwdClientData = NULL;
+    } else {
+	tsdPtr->cwdPathPtr = Tcl_NewStringObj(str, len);
+	tsdPtr->cwdClientData = clientData;
+	Tcl_IncrRefCount(tsdPtr->cwdPathPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeFilesystem --
+ *
+ *	Clean up the filesystem. After this, calls to all Tcl_FS... functions
+ *	will fail.
+ *
+ *	We will later call TclResetFilesystem to restore the FS to a pristine
+ *	state.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees any memory allocated by the filesystem.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeFilesystem(void)
+{
+    FilesystemRecord *fsRecPtr;
+
+    /*
+     * Assumption that only one thread is active now. Otherwise we would need
+     * to put various mutexes around this code.
+     */
+
+    if (cwdPathPtr != NULL) {
+	Tcl_DecrRefCount(cwdPathPtr);
+	cwdPathPtr = NULL;
+	cwdPathEpoch = 0;
+    }
+    if (cwdClientData != NULL) {
+	NativeFreeInternalRep(cwdClientData);
+	cwdClientData = NULL;
+    }
+
+    /*
+     * Remove all filesystems, freeing any allocated memory that is no longer
+     * needed.
+     */
+
+    fsRecPtr = filesystemList;
+    while (fsRecPtr != NULL) {
+	FilesystemRecord *tmpFsRecPtr = fsRecPtr->nextPtr;
+
+	/*
+	 * The native filesystem is static, so we don't free it.
+	 */
+
+	if (fsRecPtr != &nativeFilesystemRecord) {
+	    ckfree(fsRecPtr);
+	}
+	fsRecPtr = tmpFsRecPtr;
+    }
+    if (++theFilesystemEpoch == 0) {
+	++theFilesystemEpoch;
+    }
+    filesystemList = NULL;
+
+    /*
+     * Now filesystemList is NULL. This means that any attempt to use the
+     * filesystem is likely to fail.
+     */
+
+#ifdef _WIN32
+    TclWinEncodingsCleanup();
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclResetFilesystem --
+ *
+ *	Restore the filesystem to a pristine state.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclResetFilesystem(void)
+{
+    filesystemList = &nativeFilesystemRecord;
+    if (++theFilesystemEpoch == 0) {
+	++theFilesystemEpoch;
+    }
+
+#ifdef _WIN32
+    /*
+     * Cleans up the win32 API filesystem proc lookup table. This must happen
+     * very late in finalization so that deleting of copied dlls can occur.
+     */
+
+    TclWinResetInterfaces();
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSRegister --
+ *
+ *	Insert the filesystem function table at the head of the list of
+ *	functions which are used during calls to all file-system operations.
+ *	The filesystem will be added even if it is already in the list. (You
+ *	can use Tcl_FSData to check if it is in the list, provided the
+ *	ClientData used was not NULL).
+ *
+ *	Note that the filesystem handling is head-to-tail of the list. Each
+ *	filesystem is asked in turn whether it can handle a particular
+ *	request, until one of them says 'yes'. At that point no further
+ *	filesystems are asked.
+ *
+ *	In particular this means if you want to add a diagnostic filesystem
+ *	(which simply reports all fs activity), it must be at the head of the
+ *	list: i.e. it must be the last registered.
+ *
+ * Results:
+ *	Normally TCL_OK; TCL_ERROR if memory for a new node in the list could
+ *	not be allocated.
+ *
+ * Side effects:
+ *	Memory allocated and modifies the link list for filesystems.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSRegister(
+    ClientData clientData,	/* Client specific data for this fs. */
+    const Tcl_Filesystem *fsPtr)/* The filesystem record for the new fs. */
+{
+    FilesystemRecord *newFilesystemPtr;
+
+    if (fsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    newFilesystemPtr = ckalloc(sizeof(FilesystemRecord));
+
+    newFilesystemPtr->clientData = clientData;
+    newFilesystemPtr->fsPtr = fsPtr;
+
+    /*
+     * Is this lock and wait strictly speaking necessary? Since any iterators
+     * out there will have grabbed a copy of the head of the list and be
+     * iterating away from that, if we add a new element to the head of the
+     * list, it can't possibly have any effect on any of their loops. In fact
+     * it could be better not to wait, since we are adjusting the filesystem
+     * epoch, any cached representations calculated by existing iterators are
+     * going to have to be thrown away anyway.
+     *
+     * However, since registering and unregistering filesystems is a very rare
+     * action, this is not a very important point.
+     */
+
+    Tcl_MutexLock(&filesystemMutex);
+
+    newFilesystemPtr->nextPtr = filesystemList;
+    newFilesystemPtr->prevPtr = NULL;
+    if (filesystemList) {
+	filesystemList->prevPtr = newFilesystemPtr;
+    }
+    filesystemList = newFilesystemPtr;
+
+    /*
+     * Increment the filesystem epoch counter, since existing paths might
+     * conceivably now belong to different filesystems.
+     */
+
+    if (++theFilesystemEpoch == 0) {
+	++theFilesystemEpoch;
+    }
+    Tcl_MutexUnlock(&filesystemMutex);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSUnregister --
+ *
+ *	Remove the passed filesystem from the list of filesystem function
+ *	tables. It also ensures that the built-in (native) filesystem is not
+ *	removable, although we may wish to change that decision in the future
+ *	to allow a smaller Tcl core, in which the native filesystem is not
+ *	used at all (we could, say, initialise Tcl completely over a network
+ *	connection).
+ *
+ * Results:
+ *	TCL_OK if the function pointer was successfully removed, TCL_ERROR
+ *	otherwise.
+ *
+ * Side effects:
+ *	Memory may be deallocated (or will be later, once no "path" objects
+ *	refer to this filesystem), but the list of registered filesystems is
+ *	updated immediately.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSUnregister(
+    const Tcl_Filesystem *fsPtr)/* The filesystem record to remove. */
+{
+    int retVal = TCL_ERROR;
+    FilesystemRecord *fsRecPtr;
+
+    Tcl_MutexLock(&filesystemMutex);
+
+    /*
+     * Traverse the 'filesystemList' looking for the particular node whose
+     * 'fsPtr' member matches 'fsPtr' and remove that one from the list.
+     * Ensure that the "default" node cannot be removed.
+     */
+
+    fsRecPtr = filesystemList;
+    while ((retVal == TCL_ERROR) && (fsRecPtr != &nativeFilesystemRecord)) {
+	if (fsRecPtr->fsPtr == fsPtr) {
+	    if (fsRecPtr->prevPtr) {
+		fsRecPtr->prevPtr->nextPtr = fsRecPtr->nextPtr;
+	    } else {
+		filesystemList = fsRecPtr->nextPtr;
+	    }
+	    if (fsRecPtr->nextPtr) {
+		fsRecPtr->nextPtr->prevPtr = fsRecPtr->prevPtr;
+	    }
+
+	    /*
+	     * Increment the filesystem epoch counter, since existing paths
+	     * might conceivably now belong to different filesystems. This
+	     * should also ensure that paths which have cached the filesystem
+	     * which is about to be deleted do not reference that filesystem
+	     * (which would of course lead to memory exceptions).
+	     */
+
+	    if (++theFilesystemEpoch == 0) {
+		++theFilesystemEpoch;
+	    }
+
+	    ckfree(fsRecPtr);
+
+	    retVal = TCL_OK;
+	} else {
+	    fsRecPtr = fsRecPtr->nextPtr;
+	}
+    }
+
+    Tcl_MutexUnlock(&filesystemMutex);
+    return retVal;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSMatchInDirectory --
+ *
+ *	This routine is used by the globbing code to search a directory for
+ *	all files which match a given pattern. The appropriate function for
+ *	the filesystem to which pathPtr belongs will be called. If pathPtr
+ *	does not belong to any filesystem and if it is NULL or the empty
+ *	string, then we assume the pattern is to be matched in the current
+ *	working directory. To avoid have the Tcl_FSMatchInDirectoryProc for
+ *	each filesystem from having to deal with this issue, we create a
+ *	pathPtr on the fly (equal to the cwd), and then remove it from the
+ *	results returned. This makes filesystems easy to write, since they can
+ *	assume the pathPtr passed to them is an ordinary path. In fact this
+ *	means we could remove such special case handling from Tcl's native
+ *	filesystems.
+ *
+ *	If 'pattern' is NULL, then pathPtr is assumed to be a fully specified
+ *	path of a single file/directory which must be checked for existence
+ *	and correct type.
+ *
+ * Results:
+ *
+ *	The return value is a standard Tcl result indicating whether an error
+ *	occurred in globbing. Error messages are placed in interp, but good
+ *	results are placed in the resultPtr given.
+ *
+ *	Recursive searches, e.g.
+ *		glob -dir $dir -join * pkgIndex.tcl
+ *	which must recurse through each directory matching '*' are handled
+ *	internally by Tcl, by passing specific flags in a modified 'types'
+ *	parameter. This means the actual filesystem only ever sees patterns
+ *	which match in a single directory.
+ *
+ * Side effects:
+ *	The interpreter may have an error message inserted into it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSMatchInDirectory(
+    Tcl_Interp *interp,		/* Interpreter to receive error messages, but
+				 * may be NULL. */
+    Tcl_Obj *resultPtr,		/* List object to receive results. */
+    Tcl_Obj *pathPtr,		/* Contains path to directory to search. */
+    const char *pattern,	/* Pattern to match against. */
+    Tcl_GlobTypeData *types)	/* Object containing list of acceptable types.
+				 * May be NULL. In particular the directory
+				 * flag is very important. */
+{
+    const Tcl_Filesystem *fsPtr;
+    Tcl_Obj *cwd, *tmpResultPtr, **elemsPtr;
+    int resLength, i, ret = -1;
+
+    if (types != NULL && (types->type & TCL_GLOB_TYPE_MOUNT)) {
+	/*
+	 * We don't currently allow querying of mounts by external code (a
+	 * valuable future step), so since we're the only function that
+	 * actually knows about mounts, this means we're being called
+	 * recursively by ourself. Return no matches.
+	 */
+
+	return TCL_OK;
+    }
+
+    if (pathPtr != NULL) {
+	fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+    } else {
+	fsPtr = NULL;
+    }
+
+    /*
+     * Check if we've successfully mapped the path to a filesystem within
+     * which to search.
+     */
+
+    if (fsPtr != NULL) {
+	if (fsPtr->matchInDirectoryProc == NULL) {
+	    Tcl_SetErrno(ENOENT);
+	    return -1;
+	}
+	ret = fsPtr->matchInDirectoryProc(interp, resultPtr, pathPtr, pattern,
+		types);
+	if (ret == TCL_OK && pattern != NULL) {
+	    FsAddMountsToGlobResult(resultPtr, pathPtr, pattern, types);
+	}
+	return ret;
+    }
+
+    /*
+     * If the path isn't empty, we have no idea how to match files in a
+     * directory which belongs to no known filesystem.
+     */
+
+    if (pathPtr != NULL && TclGetString(pathPtr)[0] != '\0') {
+	Tcl_SetErrno(ENOENT);
+	return -1;
+    }
+
+    /*
+     * We have an empty or NULL path. This is defined to mean we must search
+     * for files within the current 'cwd'. We therefore use that, but then
+     * since the proc we call will return results which include the cwd we
+     * must then trim it off the front of each path in the result. We choose
+     * to deal with this here (in the generic code), since if we don't, every
+     * single filesystem's implementation of Tcl_FSMatchInDirectory will have
+     * to deal with it for us.
+     */
+
+    cwd = Tcl_FSGetCwd(NULL);
+    if (cwd == NULL) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "glob couldn't determine the current working directory",
+		    -1));
+	}
+	return TCL_ERROR;
+    }
+
+    fsPtr = Tcl_FSGetFileSystemForPath(cwd);
+    if (fsPtr != NULL && fsPtr->matchInDirectoryProc != NULL) {
+	TclNewObj(tmpResultPtr);
+	Tcl_IncrRefCount(tmpResultPtr);
+	ret = fsPtr->matchInDirectoryProc(interp, tmpResultPtr, cwd, pattern,
+		types);
+	if (ret == TCL_OK) {
+	    FsAddMountsToGlobResult(tmpResultPtr, cwd, pattern, types);
+
+	    /*
+	     * Note that we know resultPtr and tmpResultPtr are distinct.
+	     */
+
+	    ret = Tcl_ListObjGetElements(interp, tmpResultPtr,
+		    &resLength, &elemsPtr);
+	    for (i=0 ; ret==TCL_OK && i<resLength ; i++) {
+		ret = Tcl_ListObjAppendElement(interp, resultPtr,
+			TclFSMakePathRelative(interp, elemsPtr[i], cwd));
+	    }
+	}
+	TclDecrRefCount(tmpResultPtr);
+    }
+    Tcl_DecrRefCount(cwd);
+    return ret;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FsAddMountsToGlobResult --
+ *
+ *	This routine is used by the globbing code to take the results of a
+ *	directory listing and add any mounted paths to that listing. This is
+ *	required so that simple things like 'glob *' merge mounts and listings
+ *	correctly.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Modifies the resultPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FsAddMountsToGlobResult(
+    Tcl_Obj *resultPtr,		/* The current list of matching paths; must
+				 * not be shared! */
+    Tcl_Obj *pathPtr,		/* The directory in question. */
+    const char *pattern,	/* Pattern to match against. */
+    Tcl_GlobTypeData *types)	/* Object containing list of acceptable types.
+				 * May be NULL. In particular the directory
+				 * flag is very important. */
+{
+    int mLength, gLength, i;
+    int dir = (types == NULL || (types->type & TCL_GLOB_TYPE_DIR));
+    Tcl_Obj *mounts = FsListMounts(pathPtr, pattern);
+
+    if (mounts == NULL) {
+	return;
+    }
+
+    if (Tcl_ListObjLength(NULL, mounts, &mLength) != TCL_OK || mLength == 0) {
+	goto endOfMounts;
+    }
+    if (Tcl_ListObjLength(NULL, resultPtr, &gLength) != TCL_OK) {
+	goto endOfMounts;
+    }
+    for (i=0 ; i<mLength ; i++) {
+	Tcl_Obj *mElt;
+	int j;
+	int found = 0;
+
+	Tcl_ListObjIndex(NULL, mounts, i, &mElt);
+
+	for (j=0 ; j<gLength ; j++) {
+	    Tcl_Obj *gElt;
+
+	    Tcl_ListObjIndex(NULL, resultPtr, j, &gElt);
+	    if (Tcl_FSEqualPaths(mElt, gElt)) {
+		found = 1;
+		if (!dir) {
+		    /*
+		     * We don't want to list this.
+		     */
+
+		    Tcl_ListObjReplace(NULL, resultPtr, j, 1, 0, NULL);
+		    gLength--;
+		}
+		break;		/* Break out of for loop. */
+	    }
+	}
+	if (!found && dir) {
+	    Tcl_Obj *norm;
+	    int len, mlen;
+
+	    /*
+	     * We know mElt is absolute normalized and lies inside pathPtr, so
+	     * now we must add to the result the right representation of mElt,
+	     * i.e. the representation which is relative to pathPtr.
+	     */
+
+	    norm = Tcl_FSGetNormalizedPath(NULL, pathPtr);
+	    if (norm != NULL) {
+		const char *path, *mount;
+
+		mount = TclGetStringFromObj(mElt, &mlen);
+		path = TclGetStringFromObj(norm, &len);
+		if (path[len-1] == '/') {
+		    /*
+		     * Deal with the root of the volume.
+		     */
+
+		    len--;
+		}
+		len++;		/* account for '/' in the mElt [Bug 1602539] */
+		mElt = TclNewFSPathObj(pathPtr, mount + len, mlen - len);
+		Tcl_ListObjAppendElement(NULL, resultPtr, mElt);
+	    }
+	    /*
+	     * No need to increment gLength, since we don't want to compare
+	     * mounts against mounts.
+	     */
+	}
+    }
+
+  endOfMounts:
+    Tcl_DecrRefCount(mounts);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSMountsChanged --
+ *
+ *	Notify the filesystem that the available mounted filesystems (or
+ *	within any one filesystem type, the number or location of mount
+ *	points) have changed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The global filesystem variable 'theFilesystemEpoch' is incremented.
+ *	The effect of this is to make all cached path representations invalid.
+ *	Clearly it should only therefore be called when it is really required!
+ *	There are a few circumstances when it should be called:
+ *
+ *	(1) when a new filesystem is registered or unregistered. Strictly
+ *	speaking this is only necessary if the new filesystem accepts file
+ *	paths as is (normally the filesystem itself is really a shell which
+ *	hasn't yet had any mount points established and so its
+ *	'pathInFilesystem' proc will always fail). However, for safety, Tcl
+ *	always calls this for you in these circumstances.
+ *
+ *	(2) when additional mount points are established inside any existing
+ *	filesystem (except the native fs)
+ *
+ *	(3) when any filesystem (except the native fs) changes the list of
+ *	available volumes.
+ *
+ *	(4) when the mapping from a string representation of a file to a full,
+ *	normalized path changes. For example, if 'env(HOME)' is modified, then
+ *	any path containing '~' will map to a different filesystem location.
+ *	Therefore all such paths need to have their internal representation
+ *	invalidated.
+ *
+ *	Tcl has no control over (2) and (3), so any registered filesystem must
+ *	make sure it calls this function when those situations occur.
+ *
+ *	(Note: the reason for the exception in 2,3 for the native filesystem
+ *	is that the native filesystem by default claims all unknown files even
+ *	if it really doesn't understand them or if they don't exist).
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_FSMountsChanged(
+    const Tcl_Filesystem *fsPtr)
+{
+    /*
+     * We currently don't do anything with this parameter. We could in the
+     * future only invalidate files for this filesystem or otherwise take more
+     * advanced action.
+     */
+
+    (void)fsPtr;
+
+    /*
+     * Increment the filesystem epoch counter, since existing paths might now
+     * belong to different filesystems.
+     */
+
+    Tcl_MutexLock(&filesystemMutex);
+    if (++theFilesystemEpoch == 0) {
+	++theFilesystemEpoch;
+    }
+    Tcl_MutexUnlock(&filesystemMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSData --
+ *
+ *	Retrieve the clientData field for the filesystem given, or NULL if
+ *	that filesystem is not registered.
+ *
+ * Results:
+ *	A clientData value, or NULL. Note that if the filesystem was
+ *	registered with a NULL clientData field, this function will return
+ *	that NULL value.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_FSData(
+    const Tcl_Filesystem *fsPtr) /* The filesystem record to query. */
+{
+    ClientData retVal = NULL;
+    FilesystemRecord *fsRecPtr = FsGetFirstFilesystem();
+
+    /*
+     * Traverse the list of filesystems look for a particular one. If found,
+     * return that filesystem's clientData (originally provided when calling
+     * Tcl_FSRegister).
+     */
+
+    while ((retVal == NULL) && (fsRecPtr != NULL)) {
+	if (fsRecPtr->fsPtr == fsPtr) {
+	    retVal = fsRecPtr->clientData;
+	}
+	fsRecPtr = fsRecPtr->nextPtr;
+    }
+
+    return retVal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFSNormalizeToUniquePath --
+ *
+ *	Takes a path specification containing no ../, ./ sequences, and
+ *	converts it into a unique path for the given platform. On Unix, this
+ *	means the path must be free of symbolic links/aliases, and on Windows
+ *	it means we want the long form, with that long form's case-dependence
+ *	(which gives us a unique, case-dependent path).
+ *
+ * Results:
+ *	The pathPtr is modified in place. The return value is the last byte
+ *	offset which was recognised in the path string.
+ *
+ * Side effects:
+ *	None (beyond the memory allocation for the result).
+ *
+ * Special notes:
+ *	If the filesystem-specific normalizePathProcs can re-introduce ../, ./
+ *	sequences into the path, then this function will not return the
+ *	correct result. This may be possible with symbolic links on unix.
+ *
+ *	Important assumption: if startAt is non-zero, it must point to a
+ *	directory separator that we know exists and is already normalized (so
+ *	it is important not to point to the char just after the separator).
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclFSNormalizeToUniquePath(
+    Tcl_Interp *interp,		/* Used for error messages. */
+    Tcl_Obj *pathPtr,		/* The path to normalize in place. */
+    int startAt)		/* Start at this char-offset. */
+{
+    FilesystemRecord *fsRecPtr, *firstFsRecPtr;
+
+    /*
+     * Call each of the "normalise path" functions in succession. This is a
+     * special case, in which if we have a native filesystem handler, we call
+     * it first. This is because the root of Tcl's filesystem is always a
+     * native filesystem (i.e., '/' on unix is native).
+     */
+
+    firstFsRecPtr = FsGetFirstFilesystem();
+
+    Claim();
+    for (fsRecPtr=firstFsRecPtr; fsRecPtr!=NULL; fsRecPtr=fsRecPtr->nextPtr) {
+	if (fsRecPtr->fsPtr != &tclNativeFilesystem) {
+	    continue;
+	}
+
+	/*
+	 * TODO: Assume that we always find the native file system; it should
+	 * always be there...
+	 */
+
+	if (fsRecPtr->fsPtr->normalizePathProc != NULL) {
+	    startAt = fsRecPtr->fsPtr->normalizePathProc(interp, pathPtr,
+		    startAt);
+	}
+	break;
+    }
+
+    for (fsRecPtr=firstFsRecPtr; fsRecPtr!=NULL; fsRecPtr=fsRecPtr->nextPtr) {
+	/*
+	 * Skip the native system next time through.
+	 */
+
+	if (fsRecPtr->fsPtr == &tclNativeFilesystem) {
+	    continue;
+	}
+
+	if (fsRecPtr->fsPtr->normalizePathProc != NULL) {
+	    startAt = fsRecPtr->fsPtr->normalizePathProc(interp, pathPtr,
+		    startAt);
+	}
+
+	/*
+	 * We could add an efficiency check like this:
+	 *		if (retVal == length-of(pathPtr)) {break;}
+	 * but there's not much benefit.
+	 */
+    }
+    Disclaim();
+
+    return startAt;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetOpenMode --
+ *
+ *	This routine is an obsolete, limited version of TclGetOpenModeEx()
+ *	below. It exists only to satisfy any extensions imprudently using it
+ *	via Tcl's internal stubs table.
+ *
+ * Results:
+ *	Same as TclGetOpenModeEx().
+ *
+ * Side effects:
+ *	Same as TclGetOpenModeEx().
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclGetOpenMode(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting -
+				 * may be NULL. */
+    const char *modeString,	/* Mode string, e.g. "r+" or "RDONLY CREAT" */
+    int *seekFlagPtr)		/* Set this to 1 if the caller should seek to
+				 * EOF during the opening of the file. */
+{
+    int binary = 0;
+    return TclGetOpenModeEx(interp, modeString, seekFlagPtr, &binary);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetOpenModeEx --
+ *
+ *	Computes a POSIX mode mask for opening a file, from a given string,
+ *	and also sets flags to indicate whether the caller should seek to EOF
+ *	after opening the file, and whether the caller should configure the
+ *	channel for binary data.
+ *
+ * Results:
+ *	On success, returns mode to pass to "open". If an error occurs, the
+ *	return value is -1 and if interp is not NULL, sets interp's result
+ *	object to an error message.
+ *
+ * Side effects:
+ *	Sets the integer referenced by seekFlagPtr to 1 to tell the caller to
+ *	seek to EOF after opening the file, or to 0 otherwise. Sets the
+ *	integer referenced by binaryPtr to 1 to tell the caller to seek to
+ *	configure the channel for binary data, or to 0 otherwise.
+ *
+ * Special note:
+ *	This code is based on a prototype implementation contributed by Mark
+ *	Diekhans.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclGetOpenModeEx(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting -
+				 * may be NULL. */
+    const char *modeString,	/* Mode string, e.g. "r+" or "RDONLY CREAT" */
+    int *seekFlagPtr,		/* Set this to 1 if the caller should seek to
+				 * EOF during the opening of the file. */
+    int *binaryPtr)		/* Set this to 1 if the caller should
+				 * configure the opened channel for binary
+				 * operations. */
+{
+    int mode, modeArgc, c, i, gotRW;
+    const char **modeArgv, *flag;
+#define RW_MODES (O_RDONLY|O_WRONLY|O_RDWR)
+
+    /*
+     * Check for the simpler fopen-like access modes (e.g., "r"). They are
+     * distinguished from the POSIX access modes by the presence of a
+     * lower-case first letter.
+     */
+
+    *seekFlagPtr = 0;
+    *binaryPtr = 0;
+    mode = 0;
+
+    /*
+     * Guard against international characters before using byte oriented
+     * routines.
+     */
+
+    if (!(modeString[0] & 0x80)
+	    && islower(UCHAR(modeString[0]))) { /* INTL: ISO only. */
+	switch (modeString[0]) {
+	case 'r':
+	    mode = O_RDONLY;
+	    break;
+	case 'w':
+	    mode = O_WRONLY|O_CREAT|O_TRUNC;
+	    break;
+	case 'a':
+	    /*
+	     * Added O_APPEND for proper automatic seek-to-end-on-write by the
+	     * OS. [Bug 680143]
+	     */
+
+	    mode = O_WRONLY|O_CREAT|O_APPEND;
+	    *seekFlagPtr = 1;
+	    break;
+	default:
+	    goto error;
+	}
+	i = 1;
+	while (i<3 && modeString[i]) {
+	    if (modeString[i] == modeString[i-1]) {
+		goto error;
+	    }
+	    switch (modeString[i++]) {
+	    case '+':
+		/*
+		 * Must remove the O_APPEND flag so that the seek command
+		 * works. [Bug 1773127]
+		 */
+
+		mode &= ~(O_RDONLY|O_WRONLY|O_APPEND);
+		mode |= O_RDWR;
+		break;
+	    case 'b':
+		*binaryPtr = 1;
+		break;
+	    default:
+		goto error;
+	    }
+	}
+	if (modeString[i] != 0) {
+	    goto error;
+	}
+	return mode;
+
+    error:
+	*seekFlagPtr = 0;
+	*binaryPtr = 0;
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "illegal access mode \"%s\"", modeString));
+	}
+	return -1;
+    }
+
+    /*
+     * The access modes are specified using a list of POSIX modes such as
+     * O_CREAT.
+     *
+     * IMPORTANT NOTE: We rely on Tcl_SplitList working correctly when a NULL
+     * interpreter is passed in.
+     */
+
+    if (Tcl_SplitList(interp, modeString, &modeArgc, &modeArgv) != TCL_OK) {
+	if (interp != NULL) {
+	    Tcl_AddErrorInfo(interp,
+		    "\n    while processing open access modes \"");
+	    Tcl_AddErrorInfo(interp, modeString);
+	    Tcl_AddErrorInfo(interp, "\"");
+	}
+	return -1;
+    }
+
+    gotRW = 0;
+    for (i = 0; i < modeArgc; i++) {
+	flag = modeArgv[i];
+	c = flag[0];
+	if ((c == 'R') && (strcmp(flag, "RDONLY") == 0)) {
+	    mode = (mode & ~RW_MODES) | O_RDONLY;
+	    gotRW = 1;
+	} else if ((c == 'W') && (strcmp(flag, "WRONLY") == 0)) {
+	    mode = (mode & ~RW_MODES) | O_WRONLY;
+	    gotRW = 1;
+	} else if ((c == 'R') && (strcmp(flag, "RDWR") == 0)) {
+	    mode = (mode & ~RW_MODES) | O_RDWR;
+	    gotRW = 1;
+	} else if ((c == 'A') && (strcmp(flag, "APPEND") == 0)) {
+	    mode |= O_APPEND;
+	    *seekFlagPtr = 1;
+	} else if ((c == 'C') && (strcmp(flag, "CREAT") == 0)) {
+	    mode |= O_CREAT;
+	} else if ((c == 'E') && (strcmp(flag, "EXCL") == 0)) {
+	    mode |= O_EXCL;
+
+	} else if ((c == 'N') && (strcmp(flag, "NOCTTY") == 0)) {
+#ifdef O_NOCTTY
+	    mode |= O_NOCTTY;
+#else
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"access mode \"%s\" not supported by this system",
+			flag));
+	    }
+	    ckfree(modeArgv);
+	    return -1;
+#endif
+
+	} else if ((c == 'N') && (strcmp(flag, "NONBLOCK") == 0)) {
+#ifdef O_NONBLOCK
+	    mode |= O_NONBLOCK;
+#else
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"access mode \"%s\" not supported by this system",
+			flag));
+	    }
+	    ckfree(modeArgv);
+	    return -1;
+#endif
+
+	} else if ((c == 'T') && (strcmp(flag, "TRUNC") == 0)) {
+	    mode |= O_TRUNC;
+	} else if ((c == 'B') && (strcmp(flag, "BINARY") == 0)) {
+	    *binaryPtr = 1;
+	} else {
+
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"invalid access mode \"%s\": must be RDONLY, WRONLY, "
+			"RDWR, APPEND, BINARY, CREAT, EXCL, NOCTTY, NONBLOCK,"
+			" or TRUNC", flag));
+	    }
+	    ckfree(modeArgv);
+	    return -1;
+	}
+    }
+
+    ckfree(modeArgv);
+
+    if (!gotRW) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "access mode must include either RDONLY, WRONLY, or RDWR",
+		    -1));
+	}
+	return -1;
+    }
+    return mode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSEvalFile, Tcl_FSEvalFileEx, TclNREvalFile --
+ *
+ *	Read in a file and process the entire file as one gigantic Tcl
+ *	command. Tcl_FSEvalFile is Tcl_FSEvalFileEx without encoding argument.
+ *	TclNREvalFile is an NRE-enabled version of Tcl_FSEvalFileEx.
+ *
+ * Results:
+ *	A standard Tcl result, which is either the result of executing the
+ *	file or an error indicating why the file couldn't be read.
+ *
+ * Side effects:
+ *	Depends on the commands in the file. During the evaluation of the
+ *	contents of the file, iPtr->scriptFile is made to point to pathPtr
+ *	(the old value is cached and replaced when this function returns).
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSEvalFile(
+    Tcl_Interp *interp,		/* Interpreter in which to process file. */
+    Tcl_Obj *pathPtr)		/* Path of file to process. Tilde-substitution
+				 * will be performed on this name. */
+{
+    return Tcl_FSEvalFileEx(interp, pathPtr, NULL);
+}
+
+int
+Tcl_FSEvalFileEx(
+    Tcl_Interp *interp,		/* Interpreter in which to process file. */
+    Tcl_Obj *pathPtr,		/* Path of file to process. Tilde-substitution
+				 * will be performed on this name. */
+    const char *encodingName)	/* If non-NULL, then use this encoding for the
+				 * file. NULL means use the system encoding. */
+{
+    int length, result = TCL_ERROR;
+    Tcl_StatBuf statBuf;
+    Tcl_Obj *oldScriptFile;
+    Interp *iPtr;
+    const char *string;
+    Tcl_Channel chan;
+    Tcl_Obj *objPtr;
+
+    if (Tcl_FSGetNormalizedPath(interp, pathPtr) == NULL) {
+	return result;
+    }
+
+    if (Tcl_FSStat(pathPtr, &statBuf) == -1) {
+	Tcl_SetErrno(errno);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	return result;
+    }
+    chan = Tcl_FSOpenFileChannel(interp, pathPtr, "r", 0644);
+    if (chan == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	return result;
+    }
+
+    /*
+     * The eofchar is \32 (^Z). This is the usual on Windows, but we effect
+     * this cross-platform to allow for scripted documents. [Bug: 2040]
+     */
+
+    Tcl_SetChannelOption(interp, chan, "-eofchar", "\32 {}");
+
+    /*
+     * If the encoding is specified, set it for the channel. Else don't touch
+     * it (and use the system encoding) Report error on unknown encoding.
+     */
+
+    if (encodingName != NULL) {
+	if (Tcl_SetChannelOption(interp, chan, "-encoding", encodingName)
+		!= TCL_OK) {
+	    Tcl_Close(interp,chan);
+	    return result;
+	}
+    }
+
+    objPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(objPtr);
+
+    /*
+     * Try to read first character of stream, so we can check for utf-8 BOM to
+     * be handled especially.
+     */
+
+    if (Tcl_ReadChars(chan, objPtr, 1, 0) < 0) {
+	Tcl_Close(interp, chan);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	goto end;
+    }
+    string = Tcl_GetString(objPtr);
+
+    /*
+     * If first character is not a BOM, append the remaining characters,
+     * otherwise replace them. [Bug 3466099]
+     */
+
+    if (Tcl_ReadChars(chan, objPtr, -1,
+	    memcmp(string, "\xef\xbb\xbf", 3)) < 0) {
+	Tcl_Close(interp, chan);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	goto end;
+    }
+
+    if (Tcl_Close(interp, chan) != TCL_OK) {
+	goto end;
+    }
+
+    iPtr = (Interp *) interp;
+    oldScriptFile = iPtr->scriptFile;
+    iPtr->scriptFile = pathPtr;
+    Tcl_IncrRefCount(iPtr->scriptFile);
+    string = TclGetStringFromObj(objPtr, &length);
+
+    /*
+     * TIP #280 Force the evaluator to open a frame for a sourced file.
+     */
+
+    iPtr->evalFlags |= TCL_EVAL_FILE;
+    result = TclEvalEx(interp, string, length, 0, 1, NULL, string);
+
+    /*
+     * Now we have to be careful; the script may have changed the
+     * iPtr->scriptFile value, so we must reset it without assuming it still
+     * points to 'pathPtr'.
+     */
+
+    if (iPtr->scriptFile != NULL) {
+	Tcl_DecrRefCount(iPtr->scriptFile);
+    }
+    iPtr->scriptFile = oldScriptFile;
+
+    if (result == TCL_RETURN) {
+	result = TclUpdateReturnInfo(iPtr);
+    } else if (result == TCL_ERROR) {
+	/*
+	 * Record information telling where the error occurred.
+	 */
+
+	const char *pathString = TclGetStringFromObj(pathPtr, &length);
+	int limit = 150;
+	int overflow = (length > limit);
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (file \"%.*s%s\" line %d)",
+		(overflow ? limit : length), pathString,
+		(overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+    }
+
+  end:
+    Tcl_DecrRefCount(objPtr);
+    return result;
+}
+
+int
+TclNREvalFile(
+    Tcl_Interp *interp,		/* Interpreter in which to process file. */
+    Tcl_Obj *pathPtr,		/* Path of file to process. Tilde-substitution
+				 * will be performed on this name. */
+    const char *encodingName)	/* If non-NULL, then use this encoding for the
+				 * file. NULL means use the system encoding. */
+{
+    Tcl_StatBuf statBuf;
+    Tcl_Obj *oldScriptFile, *objPtr;
+    Interp *iPtr;
+    Tcl_Channel chan;
+    const char *string;
+
+    if (Tcl_FSGetNormalizedPath(interp, pathPtr) == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (Tcl_FSStat(pathPtr, &statBuf) == -1) {
+	Tcl_SetErrno(errno);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	return TCL_ERROR;
+    }
+    chan = Tcl_FSOpenFileChannel(interp, pathPtr, "r", 0644);
+    if (chan == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	return TCL_ERROR;
+    }
+    TclPkgFileSeen(interp, Tcl_GetString(pathPtr));
+
+    /*
+     * The eofchar is \32 (^Z). This is the usual on Windows, but we effect
+     * this cross-platform to allow for scripted documents. [Bug: 2040]
+     */
+
+    Tcl_SetChannelOption(interp, chan, "-eofchar", "\32 {}");
+
+    /*
+     * If the encoding is specified, set it for the channel. Else don't touch
+     * it (and use the system encoding) Report error on unknown encoding.
+     */
+
+    if (encodingName != NULL) {
+	if (Tcl_SetChannelOption(interp, chan, "-encoding", encodingName)
+		!= TCL_OK) {
+	    Tcl_Close(interp,chan);
+	    return TCL_ERROR;
+	}
+    }
+
+    objPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(objPtr);
+
+    /*
+     * Try to read first character of stream, so we can check for utf-8 BOM to
+     * be handled especially.
+     */
+
+    if (Tcl_ReadChars(chan, objPtr, 1, 0) < 0) {
+	Tcl_Close(interp, chan);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	Tcl_DecrRefCount(objPtr);
+	return TCL_ERROR;
+    }
+    string = Tcl_GetString(objPtr);
+
+    /*
+     * If first character is not a BOM, append the remaining characters,
+     * otherwise replace them. [Bug 3466099]
+     */
+
+    if (Tcl_ReadChars(chan, objPtr, -1,
+	    memcmp(string, "\xef\xbb\xbf", 3)) < 0) {
+	Tcl_Close(interp, chan);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't read file \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	Tcl_DecrRefCount(objPtr);
+	return TCL_ERROR;
+    }
+
+    if (Tcl_Close(interp, chan) != TCL_OK) {
+	Tcl_DecrRefCount(objPtr);
+	return TCL_ERROR;
+    }
+
+    iPtr = (Interp *) interp;
+    oldScriptFile = iPtr->scriptFile;
+    iPtr->scriptFile = pathPtr;
+    Tcl_IncrRefCount(iPtr->scriptFile);
+
+    /*
+     * TIP #280: Force the evaluator to open a frame for a sourced file.
+     */
+
+    iPtr->evalFlags |= TCL_EVAL_FILE;
+    TclNRAddCallback(interp, EvalFileCallback, oldScriptFile, pathPtr, objPtr,
+	    NULL);
+    return TclNREvalObjEx(interp, objPtr, 0, NULL, INT_MIN);
+}
+
+static int
+EvalFileCallback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *oldScriptFile = data[0];
+    Tcl_Obj *pathPtr = data[1];
+    Tcl_Obj *objPtr = data[2];
+
+    /*
+     * Now we have to be careful; the script may have changed the
+     * iPtr->scriptFile value, so we must reset it without assuming it still
+     * points to 'pathPtr'.
+     */
+
+    if (iPtr->scriptFile != NULL) {
+	Tcl_DecrRefCount(iPtr->scriptFile);
+    }
+    iPtr->scriptFile = oldScriptFile;
+
+    if (result == TCL_RETURN) {
+	result = TclUpdateReturnInfo(iPtr);
+    } else if (result == TCL_ERROR) {
+	/*
+	 * Record information telling where the error occurred.
+	 */
+
+	int length;
+	const char *pathString = TclGetStringFromObj(pathPtr, &length);
+	const int limit = 150;
+	int overflow = (length > limit);
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (file \"%.*s%s\" line %d)",
+		(overflow ? limit : length), pathString,
+		(overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+    }
+
+    Tcl_DecrRefCount(objPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetErrno --
+ *
+ *	Gets the current value of the Tcl error code variable. This is
+ *	currently the global variable "errno" but could in the future change
+ *	to something else.
+ *
+ * Results:
+ *	The value of the Tcl error code variable.
+ *
+ * Side effects:
+ *	None. Note that the value of the Tcl error code variable is UNDEFINED
+ *	if a call to Tcl_SetErrno did not precede this call.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetErrno(void)
+{
+    /*
+     * On some platforms, errno is really a thread local (implemented by the C
+     * library).
+     */
+
+    return errno;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetErrno --
+ *
+ *	Sets the Tcl error code variable to the supplied value. On some saner
+ *	platforms this is actually a thread-local (this is implemented in the
+ *	C library) but this is *really* unsafe to assume!
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Modifies the value of the Tcl error code variable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetErrno(
+    int err)			/* The new value. */
+{
+    /*
+     * On some platforms, errno is really a thread local (implemented by the C
+     * library).
+     */
+
+    errno = err;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PosixError --
+ *
+ *	This function is typically called after UNIX kernel calls return
+ *	errors. It stores machine-readable information about the error in
+ *	errorCode field of interp and returns an information string for the
+ *	caller's use.
+ *
+ * Results:
+ *	The return value is a human-readable string describing the error.
+ *
+ * Side effects:
+ *	The errorCode field of the interp is set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_PosixError(
+    Tcl_Interp *interp)		/* Interpreter whose errorCode field is to be
+				 * set. */
+{
+    const char *id, *msg;
+
+    msg = Tcl_ErrnoMsg(errno);
+    id = Tcl_ErrnoId();
+    if (interp) {
+	Tcl_SetErrorCode(interp, "POSIX", id, msg, NULL);
+    }
+    return msg;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSStat --
+ *
+ *	This function replaces the library version of stat and lsat.
+ *
+ *	The appropriate function for the filesystem to which pathPtr belongs
+ *	will be called.
+ *
+ * Results:
+ *	See stat documentation.
+ *
+ * Side effects:
+ *	See stat documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSStat(
+    Tcl_Obj *pathPtr,		/* Path of file to stat (in current CP). */
+    Tcl_StatBuf *buf)		/* Filled with results of stat call. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->statProc != NULL) {
+	return fsPtr->statProc(pathPtr, buf);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSLstat --
+ *
+ *	This function replaces the library version of lstat. The appropriate
+ *	function for the filesystem to which pathPtr belongs will be called.
+ *	If no 'lstat' function is listed, but a 'stat' function is, then Tcl
+ *	will fall back on the stat function.
+ *
+ * Results:
+ *	See lstat documentation.
+ *
+ * Side effects:
+ *	See lstat documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSLstat(
+    Tcl_Obj *pathPtr,		/* Path of file to stat (in current CP). */
+    Tcl_StatBuf *buf)		/* Filled with results of stat call. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL) {
+	if (fsPtr->lstatProc != NULL) {
+	    return fsPtr->lstatProc(pathPtr, buf);
+	}
+	if (fsPtr->statProc != NULL) {
+	    return fsPtr->statProc(pathPtr, buf);
+	}
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSAccess --
+ *
+ *	This function replaces the library version of access. The appropriate
+ *	function for the filesystem to which pathPtr belongs will be called.
+ *
+ * Results:
+ *	See access documentation.
+ *
+ * Side effects:
+ *	See access documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSAccess(
+    Tcl_Obj *pathPtr,		/* Path of file to access (in current CP). */
+    int mode)			/* Permission setting. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->accessProc != NULL) {
+	return fsPtr->accessProc(pathPtr, mode);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSOpenFileChannel --
+ *
+ *	The appropriate function for the filesystem to which pathPtr belongs
+ *	will be called.
+ *
+ * Results:
+ *	The new channel or NULL, if the named file could not be opened.
+ *
+ * Side effects:
+ *	May open the channel and may cause creation of a file on the file
+ *	system.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_FSOpenFileChannel(
+    Tcl_Interp *interp,		/* Interpreter for error reporting; can be
+				 * NULL. */
+    Tcl_Obj *pathPtr,		/* Name of file to open. */
+    const char *modeString,	/* A list of POSIX open modes or a string such
+				 * as "rw". */
+    int permissions)		/* If the open involves creating a file, with
+				 * what modes to create it? */
+{
+    const Tcl_Filesystem *fsPtr;
+    Tcl_Channel retVal = NULL;
+
+    /*
+     * We need this just to ensure we return the correct error messages under
+     * some circumstances.
+     */
+
+    if (Tcl_FSGetNormalizedPath(interp, pathPtr) == NULL) {
+	return NULL;
+    }
+
+    fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+    if (fsPtr != NULL && fsPtr->openFileChannelProc != NULL) {
+	int mode, seekFlag, binary;
+
+	/*
+	 * Parse the mode, picking up whether we want to seek to start with
+	 * and/or set the channel automatically into binary mode.
+	 */
+
+	mode = TclGetOpenModeEx(interp, modeString, &seekFlag, &binary);
+	if (mode == -1) {
+	    return NULL;
+	}
+
+	/*
+	 * Do the actual open() call.
+	 */
+
+	retVal = fsPtr->openFileChannelProc(interp, pathPtr, mode,
+		permissions);
+	if (retVal == NULL) {
+	    return NULL;
+	}
+
+	/*
+	 * Apply appropriate flags parsed out above.
+	 */
+
+	if (seekFlag && Tcl_Seek(retVal, (Tcl_WideInt) 0, SEEK_END)
+		< (Tcl_WideInt) 0) {
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"could not seek to end of file while opening \"%s\": %s",
+			Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	    }
+	    Tcl_Close(NULL, retVal);
+	    return NULL;
+	}
+	if (binary) {
+	    Tcl_SetChannelOption(interp, retVal, "-translation", "binary");
+	}
+	return retVal;
+    }
+
+    /*
+     * File doesn't belong to any filesystem that can open it.
+     */
+
+    Tcl_SetErrno(ENOENT);
+    if (interp != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't open \"%s\": %s",
+		Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSUtime --
+ *
+ *	This function replaces the library version of utime. The appropriate
+ *	function for the filesystem to which pathPtr belongs will be called.
+ *
+ * Results:
+ *	See utime documentation.
+ *
+ * Side effects:
+ *	See utime documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSUtime(
+    Tcl_Obj *pathPtr,		/* File to change access/modification
+				 * times. */
+    struct utimbuf *tval)	/* Structure containing access/modification
+				 * times to use. Should not be modified. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->utimeProc != NULL) {
+	return fsPtr->utimeProc(pathPtr, tval);
+    }
+    /* TODO: set errno here? Tcl_SetErrno(ENOENT); */
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NativeFileAttrStrings --
+ *
+ *	This function implements the platform dependent 'file attributes'
+ *	subcommand, for the native filesystem, for listing the set of possible
+ *	attribute strings. This function is part of Tcl's native filesystem
+ *	support, and is placed here because it is shared by Unix and Windows
+ *	code.
+ *
+ * Results:
+ *	An array of strings
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *const *
+NativeFileAttrStrings(
+    Tcl_Obj *pathPtr,
+    Tcl_Obj **objPtrRef)
+{
+    return tclpFileAttrStrings;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NativeFileAttrsGet --
+ *
+ *	This function implements the platform dependent 'file attributes'
+ *	subcommand, for the native filesystem, for 'get' operations. This
+ *	function is part of Tcl's native filesystem support, and is placed
+ *	here because it is shared by Unix and Windows code.
+ *
+ * Results:
+ *	Standard Tcl return code. The object placed in objPtrRef (if TCL_OK
+ *	was returned) is likely to have a refCount of zero. Either way we must
+ *	either store it somewhere (e.g. the Tcl result), or Incr/Decr its
+ *	refCount to ensure it is properly freed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NativeFileAttrsGet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *pathPtr,		/* path of file we are operating on. */
+    Tcl_Obj **objPtrRef)	/* for output. */
+{
+    return tclpFileAttrProcs[index].getProc(interp, index, pathPtr,objPtrRef);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NativeFileAttrsSet --
+ *
+ *	This function implements the platform dependent 'file attributes'
+ *	subcommand, for the native filesystem, for 'set' operations. This
+ *	function is part of Tcl's native filesystem support, and is placed
+ *	here because it is shared by Unix and Windows code.
+ *
+ * Results:
+ *	Standard Tcl return code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NativeFileAttrsSet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *pathPtr,		/* path of file we are operating on. */
+    Tcl_Obj *objPtr)		/* set to this value. */
+{
+    return tclpFileAttrProcs[index].setProc(interp, index, pathPtr, objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSFileAttrStrings --
+ *
+ *	This function implements part of the hookable 'file attributes'
+ *	subcommand. The appropriate function for the filesystem to which
+ *	pathPtr belongs will be called.
+ *
+ * Results:
+ *	The called function may either return an array of strings, or may
+ *	instead return NULL and place a Tcl list into the given objPtrRef.
+ *	Tcl will take that list and first increment its refCount before using
+ *	it. On completion of that use, Tcl will decrement its refCount. Hence
+ *	if the list should be disposed of by Tcl when done, it should have a
+ *	refCount of zero, and if the list should not be disposed of, the
+ *	filesystem should ensure it retains a refCount on the object.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *const *
+Tcl_FSFileAttrStrings(
+    Tcl_Obj *pathPtr,
+    Tcl_Obj **objPtrRef)
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->fileAttrStringsProc != NULL) {
+	return fsPtr->fileAttrStringsProc(pathPtr, objPtrRef);
+    }
+    Tcl_SetErrno(ENOENT);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFSFileAttrIndex --
+ *
+ *	Helper function for converting an attribute name to an index into the
+ *	attribute table.
+ *
+ * Results:
+ *	Tcl result code, index written to *indexPtr on result==TCL_OK
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFSFileAttrIndex(
+    Tcl_Obj *pathPtr,		/* File whose attributes are to be indexed
+				 * into. */
+    const char *attributeName,	/* The attribute being looked for. */
+    int *indexPtr)		/* Where to write the found index. */
+{
+    Tcl_Obj *listObj = NULL;
+    const char *const *attrTable;
+
+    /*
+     * Get the attribute table for the file.
+     */
+
+    attrTable = Tcl_FSFileAttrStrings(pathPtr, &listObj);
+    if (listObj != NULL) {
+	Tcl_IncrRefCount(listObj);
+    }
+
+    if (attrTable != NULL) {
+	/*
+	 * It's a constant attribute table, so use T_GIFO.
+	 */
+
+	Tcl_Obj *tmpObj = Tcl_NewStringObj(attributeName, -1);
+	int result;
+
+	result = Tcl_GetIndexFromObj(NULL, tmpObj, attrTable, NULL, TCL_EXACT,
+		indexPtr);
+	TclDecrRefCount(tmpObj);
+	if (listObj != NULL) {
+	    TclDecrRefCount(listObj);
+	}
+	return result;
+    } else if (listObj != NULL) {
+	/*
+	 * It's a non-constant attribute list, so do a literal search.
+	 */
+
+	int i, objc;
+	Tcl_Obj **objv;
+
+	if (Tcl_ListObjGetElements(NULL, listObj, &objc, &objv) != TCL_OK) {
+	    TclDecrRefCount(listObj);
+	    return TCL_ERROR;
+	}
+	for (i=0 ; i<objc ; i++) {
+	    if (!strcmp(attributeName, TclGetString(objv[i]))) {
+		TclDecrRefCount(listObj);
+		*indexPtr = i;
+		return TCL_OK;
+	    }
+	}
+	TclDecrRefCount(listObj);
+	return TCL_ERROR;
+    } else {
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSFileAttrsGet --
+ *
+ *	This function implements read access for the hookable 'file
+ *	attributes' subcommand. The appropriate function for the filesystem to
+ *	which pathPtr belongs will be called.
+ *
+ * Results:
+ *	Standard Tcl return code. The object placed in objPtrRef (if TCL_OK
+ *	was returned) is likely to have a refCount of zero. Either way we must
+ *	either store it somewhere (e.g. the Tcl result), or Incr/Decr its
+ *	refCount to ensure it is properly freed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSFileAttrsGet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *pathPtr,		/* filename we are operating on. */
+    Tcl_Obj **objPtrRef)	/* for output. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->fileAttrsGetProc != NULL) {
+	return fsPtr->fileAttrsGetProc(interp, index, pathPtr, objPtrRef);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSFileAttrsSet --
+ *
+ *	This function implements write access for the hookable 'file
+ *	attributes' subcommand. The appropriate function for the filesystem to
+ *	which pathPtr belongs will be called.
+ *
+ * Results:
+ *	Standard Tcl return code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSFileAttrsSet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *pathPtr,		/* filename we are operating on. */
+    Tcl_Obj *objPtr)		/* Input value. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->fileAttrsSetProc != NULL) {
+	return fsPtr->fileAttrsSetProc(interp, index, pathPtr, objPtr);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSGetCwd --
+ *
+ *	This function replaces the library version of getcwd().
+ *
+ *	Most VFS's will *not* implement a 'cwdProc'. Tcl now maintains its own
+ *	record (in a Tcl_Obj) of the cwd, and an attempt is made to synch this
+ *	with the cwd's containing filesystem, if that filesystem provides a
+ *	cwdProc (e.g. the native filesystem).
+ *
+ *	Note that if Tcl's cwd is not in the native filesystem, then of course
+ *	Tcl's cwd and the native cwd are different: extensions should
+ *	therefore ensure they only access the cwd through this function to
+ *	avoid confusion.
+ *
+ *	If a global cwdPathPtr already exists, it is cached in the thread's
+ *	private data structures and reference to the cached copy is returned,
+ *	subject to a synchronisation attempt in that cwdPathPtr's fs.
+ *
+ *	Otherwise, the chain of functions that have been "inserted" into the
+ *	filesystem will be called in succession until either a value other
+ *	than NULL is returned, or the entire list is visited.
+ *
+ * Results:
+ *	The result is a pointer to a Tcl_Obj specifying the current directory,
+ *	or NULL if the current directory could not be determined. If NULL is
+ *	returned, an error message is left in the interp's result.
+ *
+ *	The result already has its refCount incremented for the caller. When
+ *	it is no longer needed, that refCount should be decremented.
+ *
+ * Side effects:
+ *	Various objects may be freed and allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSGetCwd(
+    Tcl_Interp *interp)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+
+    if (TclFSCwdPointerEquals(NULL)) {
+	FilesystemRecord *fsRecPtr;
+	Tcl_Obj *retVal = NULL;
+
+	/*
+	 * We've never been called before, try to find a cwd. Call each of the
+	 * "Tcl_GetCwd" function in succession. A non-NULL return value
+	 * indicates the particular function has succeeded.
+	 */
+
+	fsRecPtr = FsGetFirstFilesystem();
+	Claim();
+	for (; (retVal == NULL) && (fsRecPtr != NULL);
+		fsRecPtr = fsRecPtr->nextPtr) {
+	    ClientData retCd;
+	    TclFSGetCwdProc2 *proc2;
+
+	    if (fsRecPtr->fsPtr->getCwdProc == NULL) {
+		continue;
+	    }
+
+	    if (fsRecPtr->fsPtr->version == TCL_FILESYSTEM_VERSION_1) {
+		retVal = fsRecPtr->fsPtr->getCwdProc(interp);
+		continue;
+	    }
+
+	    proc2 = (TclFSGetCwdProc2 *) fsRecPtr->fsPtr->getCwdProc;
+	    retCd = proc2(NULL);
+	    if (retCd != NULL) {
+		Tcl_Obj *norm;
+
+		/*
+		 * Looks like a new current directory.
+		 */
+
+		retVal = fsRecPtr->fsPtr->internalToNormalizedProc(retCd);
+		Tcl_IncrRefCount(retVal);
+		norm = TclFSNormalizeAbsolutePath(interp,retVal);
+		if (norm != NULL) {
+		    /*
+		     * We found a cwd, which is now in our global storage. We
+		     * must make a copy. Norm already has a refCount of 1.
+		     *
+		     * Threading issue: note that multiple threads at system
+		     * startup could in principle call this function
+		     * simultaneously. They will therefore each set the
+		     * cwdPathPtr independently. That behaviour is a bit
+		     * peculiar, but should be fine. Once we have a cwd, we'll
+		     * always be in the 'else' branch below which is simpler.
+		     */
+
+		    FsUpdateCwd(norm, retCd);
+		    Tcl_DecrRefCount(norm);
+		} else {
+		    fsRecPtr->fsPtr->freeInternalRepProc(retCd);
+		}
+		Tcl_DecrRefCount(retVal);
+		retVal = NULL;
+		Disclaim();
+		goto cdDidNotChange;
+	    } else if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error getting working directory name: %s",
+			Tcl_PosixError(interp)));
+	    }
+	}
+	Disclaim();
+
+	/*
+	 * Now the 'cwd' may NOT be normalized, at least on some platforms.
+	 * For the sake of efficiency, we want a completely normalized cwd at
+	 * all times.
+	 *
+	 * Finally, if retVal is NULL, we do not have a cwd, which could be
+	 * problematic.
+	 */
+
+	if (retVal != NULL) {
+	    Tcl_Obj *norm = TclFSNormalizeAbsolutePath(interp, retVal);
+
+	    if (norm != NULL) {
+		/*
+		 * We found a cwd, which is now in our global storage. We must
+		 * make a copy. Norm already has a refCount of 1.
+		 *
+		 * Threading issue: note that multiple threads at system
+		 * startup could in principle call this function
+		 * simultaneously. They will therefore each set the cwdPathPtr
+		 * independently. That behaviour is a bit peculiar, but should
+		 * be fine. Once we have a cwd, we'll always be in the 'else'
+		 * branch below which is simpler.
+		 */
+
+		ClientData cd = (ClientData) Tcl_FSGetNativePath(norm);
+
+		FsUpdateCwd(norm, TclNativeDupInternalRep(cd));
+		Tcl_DecrRefCount(norm);
+	    }
+	    Tcl_DecrRefCount(retVal);
+	}
+    } else {
+	/*
+	 * We already have a cwd cached, but we want to give the filesystem it
+	 * is in a chance to check whether that cwd has changed, or is perhaps
+	 * no longer accessible. This allows an error to be thrown if, say,
+	 * the permissions on that directory have changed.
+	 */
+
+	const Tcl_Filesystem *fsPtr =
+		Tcl_FSGetFileSystemForPath(tsdPtr->cwdPathPtr);
+	ClientData retCd = NULL;
+	Tcl_Obj *retVal, *norm;
+
+	/*
+	 * If the filesystem couldn't be found, or if no cwd function exists
+	 * for this filesystem, then we simply assume the cached cwd is ok.
+	 * If we do call a cwd, we must watch for errors (if the cwd returns
+	 * NULL). This ensures that, say, on Unix if the permissions of the
+	 * cwd change, 'pwd' does actually throw the correct error in Tcl.
+	 * (This is tested for in the test suite on unix).
+	 */
+
+	if (fsPtr == NULL || fsPtr->getCwdProc == NULL) {
+	    goto cdDidNotChange;
+	}
+
+	if (fsPtr->version == TCL_FILESYSTEM_VERSION_1) {
+	    retVal = fsPtr->getCwdProc(interp);
+	} else {
+	    /*
+	     * New API.
+	     */
+
+	    TclFSGetCwdProc2 *proc2 = (TclFSGetCwdProc2 *) fsPtr->getCwdProc;
+
+	    retCd = proc2(tsdPtr->cwdClientData);
+	    if (retCd == NULL && interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error getting working directory name: %s",
+			Tcl_PosixError(interp)));
+	    }
+
+	    if (retCd == tsdPtr->cwdClientData) {
+		goto cdDidNotChange;
+	    }
+
+	    /*
+	     * Looks like a new current directory.
+	     */
+
+	    retVal = fsPtr->internalToNormalizedProc(retCd);
+	    Tcl_IncrRefCount(retVal);
+	}
+
+	/*
+	 * Check if the 'cwd' function returned an error; if so, reset the
+	 * cwd.
+	 */
+
+	if (retVal == NULL) {
+	    FsUpdateCwd(NULL, NULL);
+	    goto cdDidNotChange;
+	}
+
+	/*
+	 * Normalize the path.
+	 */
+
+	norm = TclFSNormalizeAbsolutePath(interp, retVal);
+
+	/*
+	 * Check whether cwd has changed from the value previously stored in
+	 * cwdPathPtr. Really 'norm' shouldn't be NULL, but we are careful.
+	 */
+
+	if (norm == NULL) {
+	    /* Do nothing */
+	    if (retCd != NULL) {
+		fsPtr->freeInternalRepProc(retCd);
+	    }
+	} else if (norm == tsdPtr->cwdPathPtr) {
+	    goto cdEqual;
+	} else {
+	    /*
+	     * Note that both 'norm' and 'tsdPtr->cwdPathPtr' are normalized
+	     * paths. Therefore we can be more efficient than calling
+	     * 'Tcl_FSEqualPaths', and in addition avoid a nasty infinite loop
+	     * bug when trying to normalize tsdPtr->cwdPathPtr.
+	     */
+
+	    int len1, len2;
+	    const char *str1, *str2;
+
+	    str1 = TclGetStringFromObj(tsdPtr->cwdPathPtr, &len1);
+	    str2 = TclGetStringFromObj(norm, &len2);
+	    if ((len1 == len2) && (strcmp(str1, str2) == 0)) {
+		/*
+		 * If the paths were equal, we can be more efficient and
+		 * retain the old path object which will probably already be
+		 * shared. In this case we can simply free the normalized path
+		 * we just calculated.
+		 */
+
+	    cdEqual:
+		Tcl_DecrRefCount(norm);
+		if (retCd != NULL) {
+		    fsPtr->freeInternalRepProc(retCd);
+		}
+	    } else {
+		FsUpdateCwd(norm, retCd);
+		Tcl_DecrRefCount(norm);
+	    }
+	}
+	Tcl_DecrRefCount(retVal);
+    }
+
+  cdDidNotChange:
+    if (tsdPtr->cwdPathPtr != NULL) {
+	Tcl_IncrRefCount(tsdPtr->cwdPathPtr);
+    }
+
+    return tsdPtr->cwdPathPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSChdir --
+ *
+ *	This function replaces the library version of chdir().
+ *
+ *	The path is normalized and then passed to the filesystem which claims
+ *	it.
+ *
+ * Results:
+ *	See chdir() documentation. If successful, we keep a record of the
+ *	successful path in cwdPathPtr for subsequent calls to getcwd.
+ *
+ * Side effects:
+ *	See chdir() documentation. The global cwdPathPtr may change value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSChdir(
+    Tcl_Obj *pathPtr)
+{
+    const Tcl_Filesystem *fsPtr, *oldFsPtr = NULL;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&fsDataKey);
+    int retVal = -1;
+
+    if (tsdPtr->cwdPathPtr != NULL) {
+	oldFsPtr = Tcl_FSGetFileSystemForPath(tsdPtr->cwdPathPtr);
+    }
+    if (Tcl_FSGetNormalizedPath(NULL, pathPtr) == NULL) {
+	Tcl_SetErrno(ENOENT);
+	return retVal;
+    }
+
+    fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+    if (fsPtr != NULL) {
+	if (fsPtr->chdirProc != NULL) {
+	    /*
+	     * If this fails, an appropriate errno will have been stored using
+	     * 'Tcl_SetErrno()'.
+	     */
+
+	    retVal = fsPtr->chdirProc(pathPtr);
+	} else {
+	    /*
+	     * Fallback on stat-based implementation.
+	     */
+
+	    Tcl_StatBuf buf;
+
+	    /*
+	     * If the file can be stat'ed and is a directory and is readable,
+	     * then we can chdir. If any of these actions fail, then
+	     * 'Tcl_SetErrno()' should automatically have been called to set
+	     * an appropriate error code.
+	     */
+
+	    if ((Tcl_FSStat(pathPtr, &buf) == 0) && (S_ISDIR(buf.st_mode))
+		    && (Tcl_FSAccess(pathPtr, R_OK) == 0)) {
+		/*
+		 * We allow the chdir.
+		 */
+
+		retVal = 0;
+	    }
+	}
+    } else {
+	Tcl_SetErrno(ENOENT);
+    }
+
+    /*
+     * The cwd changed, or an error was thrown. If an error was thrown, we can
+     * just continue (and that will report the error to the user). If there
+     * was no error we must assume that the cwd was actually changed to the
+     * normalized value we calculated above, and we must therefore cache that
+     * information.
+     *
+     * If the filesystem in question has a getCwdProc, then the correct logic
+     * which performs the part below is already part of the Tcl_FSGetCwd()
+     * call, so no need to replicate it again. This will have a side effect
+     * though. The private authoritative representation of the current working
+     * directory stored in cwdPathPtr in static memory will be out-of-sync
+     * with the real OS-maintained value. The first call to Tcl_FSGetCwd will
+     * however recalculate the private copy to match the OS-value so
+     * everything will work right.
+     *
+     * However, if there is no getCwdProc, then we _must_ update our private
+     * storage of the cwd, since this is the only opportunity to do that!
+     *
+     * Note: We currently call this block of code irrespective of whether
+     * there was a getCwdProc or not, but the code should all in principle
+     * work if we only call this block if fsPtr->getCwdProc == NULL.
+     */
+
+    if (retVal == 0) {
+	/*
+	 * Note that this normalized path may be different to what we found
+	 * above (or at least a different object), if the filesystem epoch
+	 * changed recently. This can actually happen with scripted documents
+	 * very easily. Therefore we ask for the normalized path again (the
+	 * correct value will have been cached as a result of the
+	 * Tcl_FSGetFileSystemForPath call above anyway).
+	 */
+
+	Tcl_Obj *normDirName = Tcl_FSGetNormalizedPath(NULL, pathPtr);
+
+	if (normDirName == NULL) {
+	    /* Not really true, but what else to do? */
+	    Tcl_SetErrno(ENOENT);
+	    return -1;
+	}
+
+	if (fsPtr == &tclNativeFilesystem) {
+	    /*
+	     * For the native filesystem, we keep a cache of the native
+	     * representation of the cwd. But, we want to do that for the
+	     * exact format that is returned by 'getcwd' (so that we can later
+	     * compare the two representations for equality), which might not
+	     * be exactly the same char-string as the native representation of
+	     * the fully normalized path (e.g. on Windows there's a
+	     * forward-slash vs backslash difference). Hence we ask for this
+	     * again here. On Unix it might actually be true that we always
+	     * have the correct form in the native rep in which case we could
+	     * simply use:
+	     *		cd = Tcl_FSGetNativePath(pathPtr);
+	     * instead. This should be examined by someone on Unix.
+	     */
+
+	    ClientData cd;
+	    ClientData oldcd = tsdPtr->cwdClientData;
+
+	    /*
+	     * Assumption we are using a filesystem version 2.
+	     */
+
+	    TclFSGetCwdProc2 *proc2 = (TclFSGetCwdProc2 *) fsPtr->getCwdProc;
+
+	    cd = proc2(oldcd);
+	    if (cd != oldcd) {
+		FsUpdateCwd(normDirName, cd);
+	    }
+	} else {
+	    FsUpdateCwd(normDirName, NULL);
+	}
+
+	/*
+	 * If the filesystem changed between old and new cwd
+	 * force filesystem refresh on path objects.
+	 */
+	if (oldFsPtr != NULL && fsPtr != oldFsPtr) {
+	    Tcl_FSMountsChanged(NULL);
+	}
+    }
+
+    return retVal;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSLoadFile --
+ *
+ *	Dynamically loads a binary code file into memory and returns the
+ *	addresses of two functions within that file, if they are defined. The
+ *	appropriate function for the filesystem to which pathPtr belongs will
+ *	be called.
+ *
+ *	Note that the native filesystem doesn't actually assume 'pathPtr' is a
+ *	path. Rather it assumes pathPtr is either a path or just the name
+ *	(tail) of a file which can be found somewhere in the environment's
+ *	loadable path. This behaviour is not very compatible with virtual
+ *	filesystems (and has other problems documented in the load man-page),
+ *	so it is advised that full paths are always used.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs, an error message
+ *	is left in the interp's result.
+ *
+ * Side effects:
+ *	New code suddenly appears in memory. This may later be unloaded by
+ *	passing the clientData to the unloadProc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSLoadFile(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Obj *pathPtr,		/* Name of the file containing the desired
+				 * code. */
+    const char *sym1, const char *sym2,
+				/* Names of two functions to look up in the
+				 * file's symbol table. */
+    Tcl_PackageInitProc **proc1Ptr, Tcl_PackageInitProc **proc2Ptr,
+				/* Where to return the addresses corresponding
+				 * to sym1 and sym2. */
+    Tcl_LoadHandle *handlePtr,	/* Filled with token for dynamically loaded
+				 * file which will be passed back to
+				 * (*unloadProcPtr)() to unload the file. */
+    Tcl_FSUnloadFileProc **unloadProcPtr)
+				/* Filled with address of Tcl_FSUnloadFileProc
+				 * function which should be used for this
+				 * file. */
+{
+    const char *symbols[3];
+    void *procPtrs[2];
+    int res;
+
+    /*
+     * Initialize the arrays.
+     */
+
+    symbols[0] = sym1;
+    symbols[1] = sym2;
+    symbols[2] = NULL;
+
+    /*
+     * Perform the load.
+     */
+
+    res = Tcl_LoadFile(interp, pathPtr, symbols, 0, procPtrs, handlePtr);
+    if (res == TCL_OK) {
+	*proc1Ptr = (Tcl_PackageInitProc *) procPtrs[0];
+	*proc2Ptr = (Tcl_PackageInitProc *) procPtrs[1];
+    } else {
+	*proc1Ptr = *proc2Ptr = NULL;
+    }
+
+    return res;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LoadFile --
+ *
+ *	Dynamically loads a binary code file into memory and returns the
+ *	addresses of a number of given functions within that file, if they are
+ *	defined. The appropriate function for the filesystem to which pathPtr
+ *	belongs will be called.
+ *
+ *	Note that the native filesystem doesn't actually assume 'pathPtr' is a
+ *	path. Rather it assumes pathPtr is either a path or just the name
+ *	(tail) of a file which can be found somewhere in the environment's
+ *	loadable path. This behaviour is not very compatible with virtual
+ *	filesystems (and has other problems documented in the load man-page),
+ *	so it is advised that full paths are always used.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs, an error message
+ *	is left in the interp's result.
+ *
+ * Side effects:
+ *	New code suddenly appears in memory. This may later be unloaded by
+ *	calling TclFS_UnloadFile.
+ *
+ *----------------------------------------------------------------------
+ */
+
+/*
+ * Workaround for issue with modern HPUX which do allow the unlink (no ETXTBSY
+ * error) yet somehow trash some internal data structures which prevents the
+ * second and further shared libraries from getting properly loaded. Only the
+ * first is ok. We try to get around the issue by not unlinking, i.e.,
+ * emulating the behaviour of the older HPUX which denied removal.
+ *
+ * Doing the unlink is also an issue within docker containers, whose AUFS
+ * bungles this as well, see
+ *     https://github.com/dotcloud/docker/issues/1911
+ *
+ * For these situations the change below makes the execution of the unlink
+ * semi-controllable at runtime.
+ *
+ *     An AUFS filesystem (if it can be detected) will force avoidance of
+ *     unlink. The env variable TCL_TEMPLOAD_NO_UNLINK allows detection of a
+ *     users general request (unlink and not.
+ *
+ * By default the unlink is done (if not in AUFS). However if the variable is
+ * present and set to true (any integer > 0) then the unlink is skipped.
+ */
+
+int
+TclSkipUnlink(
+    Tcl_Obj *shlibFile)
+{
+    /*
+     * Order of testing:
+     * 1. On hpux we generally want to skip unlink in general
+     *
+     * Outside of hpux then:
+     * 2. For a general user request   (TCL_TEMPLOAD_NO_UNLINK present,
+     *					non-empty, => int)
+     * 3. For general AUFS environment (statfs, if available).
+     *
+     * Ad 2: This variable can disable/override the AUFS detection, i.e. for
+     *	     testing if a newer AUFS does not have the bug any more.
+     *
+     * Ad 3: This is conditionally compiled in. Condition currently must be
+     *	     set manually. This part needs proper tests in the configure(.in).
+     */
+
+#ifdef hpux
+    return 1;
+#else
+    char *skipstr = getenv("TCL_TEMPLOAD_NO_UNLINK");
+
+    if (skipstr && (skipstr[0] != '\0')) {
+	return atoi(skipstr);
+    }
+
+#ifdef TCL_TEMPLOAD_NO_UNLINK
+#ifndef NO_FSTATFS
+    {
+	struct statfs fs;
+	/*
+	 * Have fstatfs. May not have the AUFS super magic ... Indeed our build
+	 * box is too old to have it directly in the headers. Define taken from
+	 *     http://mooon.googlecode.com/svn/trunk/linux_include/linux/aufs_type.h
+	 *     http://aufs.sourceforge.net/
+	 * Better reference will be gladly taken.
+	 */
+#ifndef AUFS_SUPER_MAGIC
+#define AUFS_SUPER_MAGIC ('a' << 24 | 'u' << 16 | 'f' << 8 | 's')
+#endif /* AUFS_SUPER_MAGIC */
+	if ((statfs(Tcl_GetString (shlibFile), &fs) == 0) &&
+	    (fs.f_type == AUFS_SUPER_MAGIC)) {
+	    return 1;
+	}
+    }
+#endif /* ... NO_FSTATFS */
+#endif /* ... TCL_TEMPLOAD_NO_UNLINK */
+
+    /*
+     * Fallback: !hpux, no EV override, no AUFS (detection, nor detected):
+     * Don't skip
+     */
+    return 0;
+#endif /* hpux */
+}
+
+int
+Tcl_LoadFile(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Obj *pathPtr,		/* Name of the file containing the desired
+				 * code. */
+    const char *const symbols[],/* Names of functions to look up in the file's
+				 * symbol table. */
+    int flags,			/* Flags */
+    void *procVPtrs,		/* Where to return the addresses corresponding
+				 * to symbols[]. */
+    Tcl_LoadHandle *handlePtr)	/* Filled with token for shared library
+				 * information which can be used in
+				 * TclpFindSymbol. */
+{
+    void **procPtrs = (void **) procVPtrs;
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+    const Tcl_Filesystem *copyFsPtr;
+    Tcl_FSUnloadFileProc *unloadProcPtr;
+    Tcl_Obj *copyToPtr;
+    Tcl_LoadHandle newLoadHandle = NULL;
+    Tcl_LoadHandle divertedLoadHandle = NULL;
+    Tcl_FSUnloadFileProc *newUnloadProcPtr = NULL;
+    FsDivertLoad *tvdlPtr;
+    int retVal;
+    int i;
+
+    if (fsPtr == NULL) {
+	Tcl_SetErrno(ENOENT);
+	return TCL_ERROR;
+    }
+
+    if (fsPtr->loadFileProc != NULL) {
+	int retVal = ((Tcl_FSLoadFileProc2 *)(fsPtr->loadFileProc))
+		(interp, pathPtr, handlePtr, &unloadProcPtr, flags);
+
+	if (retVal == TCL_OK) {
+	    if (*handlePtr == NULL) {
+		return TCL_ERROR;
+	    }
+	    if (interp) {
+		Tcl_ResetResult(interp);
+	    }
+	    goto resolveSymbols;
+	}
+	if (Tcl_GetErrno() != EXDEV) {
+	    return retVal;
+	}
+    }
+
+    /*
+     * The filesystem doesn't support 'load', so we fall back on the following
+     * technique:
+     *
+     * First check if it is readable -- and exists!
+     */
+
+    if (Tcl_FSAccess(pathPtr, R_OK) != 0) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "couldn't load library \"%s\": %s",
+		    Tcl_GetString(pathPtr), Tcl_PosixError(interp)));
+	}
+	return TCL_ERROR;
+    }
+
+#ifdef TCL_LOAD_FROM_MEMORY
+    /*
+     * The platform supports loading code from memory, so ask for a buffer of
+     * the appropriate size, read the file into it and load the code from the
+     * buffer:
+     */
+
+    {
+	int ret, size;
+	void *buffer;
+	Tcl_StatBuf statBuf;
+	Tcl_Channel data;
+
+	ret = Tcl_FSStat(pathPtr, &statBuf);
+	if (ret < 0) {
+	    goto mustCopyToTempAnyway;
+	}
+	size = (int) statBuf.st_size;
+
+	/*
+	 * Tcl_Read takes an int: check that file size isn't wide.
+	 */
+
+	if (size != (Tcl_WideInt) statBuf.st_size) {
+	    goto mustCopyToTempAnyway;
+	}
+	data = Tcl_FSOpenFileChannel(interp, pathPtr, "rb", 0666);
+	if (!data) {
+	    goto mustCopyToTempAnyway;
+	}
+	buffer = TclpLoadMemoryGetBuffer(interp, size);
+	if (!buffer) {
+	    Tcl_Close(interp, data);
+	    goto mustCopyToTempAnyway;
+	}
+	ret = Tcl_Read(data, buffer, size);
+	Tcl_Close(interp, data);
+	ret = TclpLoadMemory(interp, buffer, size, ret, handlePtr,
+		&unloadProcPtr, flags);
+	if (ret == TCL_OK && *handlePtr != NULL) {
+	    goto resolveSymbols;
+	}
+    }
+
+  mustCopyToTempAnyway:
+    if (interp) {
+	Tcl_ResetResult(interp);
+    }
+#endif /* TCL_LOAD_FROM_MEMORY */
+
+    /*
+     * Get a temporary filename to use, first to copy the file into, and then
+     * to load.
+     */
+
+    copyToPtr = TclpTempFileNameForLibrary(interp, pathPtr);
+    if (copyToPtr == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_IncrRefCount(copyToPtr);
+
+    copyFsPtr = Tcl_FSGetFileSystemForPath(copyToPtr);
+    if ((copyFsPtr == NULL) || (copyFsPtr == fsPtr)) {
+	/*
+	 * We already know we can't use Tcl_FSLoadFile from this filesystem,
+	 * and we must avoid a possible infinite loop. Try to delete the file
+	 * we probably created, and then exit.
+	 */
+
+	Tcl_FSDeleteFile(copyToPtr);
+	Tcl_DecrRefCount(copyToPtr);
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "couldn't load from current filesystem", -1));
+	}
+	return TCL_ERROR;
+    }
+
+    if (TclCrossFilesystemCopy(interp, pathPtr, copyToPtr) != TCL_OK) {
+	/*
+	 * Cross-platform copy failed.
+	 */
+
+	Tcl_FSDeleteFile(copyToPtr);
+	Tcl_DecrRefCount(copyToPtr);
+	return TCL_ERROR;
+    }
+
+#ifndef _WIN32
+    /*
+     * Do we need to set appropriate permissions on the file? This may be
+     * required on some systems. On Unix we could loop over the file
+     * attributes, and set any that are called "-permissions" to 0700. However
+     * we just do this directly, like this:
+     */
+
+    {
+	int index;
+	Tcl_Obj *perm;
+
+	TclNewLiteralStringObj(perm, "0700");
+	Tcl_IncrRefCount(perm);
+	if (TclFSFileAttrIndex(copyToPtr, "-permissions", &index) == TCL_OK) {
+	    Tcl_FSFileAttrsSet(NULL, index, copyToPtr, perm);
+	}
+	Tcl_DecrRefCount(perm);
+    }
+#endif
+
+    /*
+     * We need to reset the result now, because the cross-filesystem copy may
+     * have stored the number of bytes in the result.
+     */
+
+    if (interp) {
+	Tcl_ResetResult(interp);
+    }
+
+    retVal = Tcl_LoadFile(interp, copyToPtr, symbols, flags, procPtrs,
+	    &newLoadHandle);
+    if (retVal != TCL_OK) {
+	/*
+	 * The file didn't load successfully.
+	 */
+
+	Tcl_FSDeleteFile(copyToPtr);
+	Tcl_DecrRefCount(copyToPtr);
+	return retVal;
+    }
+
+    /*
+     * Try to delete the file immediately - this is possible in some OSes, and
+     * avoids any worries about leaving the copy laying around on exit.
+     */
+
+    if (!TclSkipUnlink(copyToPtr) &&
+	    (Tcl_FSDeleteFile(copyToPtr) == TCL_OK)) {
+	Tcl_DecrRefCount(copyToPtr);
+
+	/*
+	 * We tell our caller about the real shared library which was loaded.
+	 * Note that this does mean that the package list maintained by 'load'
+	 * will store the original (vfs) path alongside the temporary load
+	 * handle and unload proc ptr.
+	 */
+
+	*handlePtr = newLoadHandle;
+	if (interp) {
+	    Tcl_ResetResult(interp);
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * When we unload this file, we need to divert the unloading so we can
+     * unload and cleanup the temporary file correctly.
+     */
+
+    tvdlPtr = ckalloc(sizeof(FsDivertLoad));
+
+    /*
+     * Remember three pieces of information. This allows us to cleanup the
+     * diverted load completely, on platforms which allow proper unloading of
+     * code.
+     */
+
+    tvdlPtr->loadHandle = newLoadHandle;
+    tvdlPtr->unloadProcPtr = newUnloadProcPtr;
+
+    if (copyFsPtr != &tclNativeFilesystem) {
+	/*
+	 * copyToPtr is already incremented for this reference.
+	 */
+
+	tvdlPtr->divertedFile = copyToPtr;
+
+	/*
+	 * This is the filesystem we loaded it into. Since we have a reference
+	 * to 'copyToPtr', we already have a refCount on this filesystem, so
+	 * we don't need to worry about it disappearing on us.
+	 */
+
+	tvdlPtr->divertedFilesystem = copyFsPtr;
+	tvdlPtr->divertedFileNativeRep = NULL;
+    } else {
+	/*
+	 * We need the native rep.
+	 */
+
+	tvdlPtr->divertedFileNativeRep = TclNativeDupInternalRep(
+		Tcl_FSGetInternalRep(copyToPtr, copyFsPtr));
+
+	/*
+	 * We don't need or want references to the copied Tcl_Obj or the
+	 * filesystem if it is the native one.
+	 */
+
+	tvdlPtr->divertedFile = NULL;
+	tvdlPtr->divertedFilesystem = NULL;
+	Tcl_DecrRefCount(copyToPtr);
+    }
+
+    copyToPtr = NULL;
+
+    divertedLoadHandle = ckalloc(sizeof(struct Tcl_LoadHandle_));
+    divertedLoadHandle->clientData = tvdlPtr;
+    divertedLoadHandle->findSymbolProcPtr = DivertFindSymbol;
+    divertedLoadHandle->unloadFileProcPtr = DivertUnloadFile;
+    *handlePtr = divertedLoadHandle;
+
+    if (interp) {
+	Tcl_ResetResult(interp);
+    }
+    return retVal;
+
+  resolveSymbols:
+    /*
+     * At this point, *handlePtr is already set up to the handle for the
+     * loaded library. We now try to resolve the symbols.
+     */
+
+    if (symbols != NULL) {
+	for (i=0 ; symbols[i] != NULL; i++) {
+	    procPtrs[i] = Tcl_FindSymbol(interp, *handlePtr, symbols[i]);
+	    if (procPtrs[i] == NULL) {
+		/*
+		 * At least one symbol in the list was not found.  Unload the
+		 * file, and report the problem back to the caller.
+		 * (Tcl_FindSymbol should already have left an appropriate
+		 * error message.)
+		 */
+
+		(*handlePtr)->unloadFileProcPtr(*handlePtr);
+		*handlePtr = NULL;
+		return TCL_ERROR;
+	    }
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DivertFindSymbol --
+ *
+ *	Find a symbol in a shared library loaded by copy-from-VFS.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void *
+DivertFindSymbol(
+    Tcl_Interp *interp, 	/* Tcl interpreter */
+    Tcl_LoadHandle loadHandle,	/* Handle to the diverted module */
+    const char *symbol)		/* Symbol to resolve */
+{
+    FsDivertLoad *tvdlPtr = (FsDivertLoad *) loadHandle->clientData;
+    Tcl_LoadHandle originalHandle = tvdlPtr->loadHandle;
+
+    return originalHandle->findSymbolProcPtr(interp, originalHandle, symbol);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DivertUnloadFile --
+ *
+ *	Unloads a file that has been loaded by copying from VFS to the native
+ *	filesystem.
+ *
+ * Parameters:
+ *	loadHandle -- Handle of the file to unload
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DivertUnloadFile(
+    Tcl_LoadHandle loadHandle)
+{
+    FsDivertLoad *tvdlPtr = (FsDivertLoad *) loadHandle->clientData;
+    Tcl_LoadHandle originalHandle;
+
+    /*
+     * This test should never trigger, since we give the client data in the
+     * function above.
+     */
+
+    if (tvdlPtr == NULL) {
+	return;
+    }
+    originalHandle = tvdlPtr->loadHandle;
+
+    /*
+     * Call the real 'unloadfile' proc we actually used. It is very important
+     * that we call this first, so that the shared library is actually
+     * unloaded by the OS. Otherwise, the following 'delete' may well fail
+     * because the shared library is still in use.
+     */
+
+    originalHandle->unloadFileProcPtr(originalHandle);
+
+    /*
+     * What filesystem contains the temp copy of the library?
+     */
+
+    if (tvdlPtr->divertedFilesystem == NULL) {
+	/*
+	 * It was the native filesystem, and we have a special function
+	 * available just for this purpose, which we know works even at this
+	 * late stage.
+	 */
+
+	TclpDeleteFile(tvdlPtr->divertedFileNativeRep);
+	NativeFreeInternalRep(tvdlPtr->divertedFileNativeRep);
+    } else {
+	/*
+	 * Remove the temporary file we created. Note, we may crash here
+	 * because encodings have been taken down already.
+	 */
+
+	if (tvdlPtr->divertedFilesystem->deleteFileProc(tvdlPtr->divertedFile)
+		!= TCL_OK) {
+	    /*
+	     * The above may have failed because the filesystem, or something
+	     * it depends upon (e.g. encodings) have been taken down because
+	     * Tcl is exiting.
+	     *
+	     * We may need to work out how to delete this file more robustly
+	     * (or give the filesystem the information it needs to delete the
+	     * file more robustly).
+	     *
+	     * In particular, one problem might be that the filesystem cannot
+	     * extract the information it needs from the above path object
+	     * because Tcl's entire filesystem apparatus (the code in this
+	     * file) has been finalized, and it refuses to pass the internal
+	     * representation to the filesystem.
+	     */
+	}
+
+	/*
+	 * And free up the allocations. This will also of course remove a
+	 * refCount from the Tcl_Filesystem to which this file belongs, which
+	 * could then free up the filesystem if we are exiting.
+	 */
+
+	Tcl_DecrRefCount(tvdlPtr->divertedFile);
+    }
+
+    ckfree(tvdlPtr);
+    ckfree(loadHandle);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindSymbol --
+ *
+ *	Find a symbol in a loaded library
+ *
+ * Results:
+ *	Returns a pointer to the symbol if found. If not found, returns NULL
+ *	and leaves an error message in the interpreter result.
+ *
+ * This function was once filesystem-specific, but has been made portable by
+ * having TclpDlopen return a structure that includes procedure pointers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void *
+Tcl_FindSymbol(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    Tcl_LoadHandle loadHandle,	/* Handle to the loaded library */
+    const char *symbol)		/* Name of the symbol to resolve */
+{
+    return loadHandle->findSymbolProcPtr(interp, loadHandle, symbol);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSUnloadFile --
+ *
+ *	Unloads a library given its handle. Checks first that the library
+ *	supports unloading.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_FSUnloadFile(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    Tcl_LoadHandle handle)	/* Handle of the file to unload */
+{
+    if (handle->unloadFileProcPtr == NULL) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "cannot unload: filesystem does not support unloading",
+		    -1));
+	}
+	return TCL_ERROR;
+    }
+    TclpUnloadFile(handle);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpUnloadFile --
+ *
+ *	Unloads a library given its handle
+ *
+ * This function was once filesystem-specific, but has been made portable by
+ * having TclpDlopen return a structure that includes procedure pointers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclpUnloadFile(
+    Tcl_LoadHandle handle)
+{
+    if (handle->unloadFileProcPtr != NULL) {
+	handle->unloadFileProcPtr(handle);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFSUnloadTempFile --
+ *
+ *	This function is called when we loaded a library of code via an
+ *	intermediate temporary file. This function ensures the library is
+ *	correctly unloaded and the temporary file is correctly deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The effects of the 'unload' function called, and of course the
+ *	temporary file will be deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFSUnloadTempFile(
+    Tcl_LoadHandle loadHandle)	/* loadHandle returned by a previous call to
+				 * Tcl_FSLoadFile(). The loadHandle is a token
+				 * that represents the loaded file. */
+{
+    FsDivertLoad *tvdlPtr = (FsDivertLoad *) loadHandle;
+
+    /*
+     * This test should never trigger, since we give the client data in the
+     * function above.
+     */
+
+    if (tvdlPtr == NULL) {
+	return;
+    }
+
+    /*
+     * Call the real 'unloadfile' proc we actually used. It is very important
+     * that we call this first, so that the shared library is actually
+     * unloaded by the OS. Otherwise, the following 'delete' may well fail
+     * because the shared library is still in use.
+     */
+
+    if (tvdlPtr->unloadProcPtr != NULL) {
+	tvdlPtr->unloadProcPtr(tvdlPtr->loadHandle);
+    }
+
+    if (tvdlPtr->divertedFilesystem == NULL) {
+	/*
+	 * It was the native filesystem, and we have a special function
+	 * available just for this purpose, which we know works even at this
+	 * late stage.
+	 */
+
+	TclpDeleteFile(tvdlPtr->divertedFileNativeRep);
+	NativeFreeInternalRep(tvdlPtr->divertedFileNativeRep);
+    } else {
+	/*
+	 * Remove the temporary file we created. Note, we may crash here
+	 * because encodings have been taken down already.
+	 */
+
+	if (tvdlPtr->divertedFilesystem->deleteFileProc(tvdlPtr->divertedFile)
+		!= TCL_OK) {
+	    /*
+	     * The above may have failed because the filesystem, or something
+	     * it depends upon (e.g. encodings) have been taken down because
+	     * Tcl is exiting.
+	     *
+	     * We may need to work out how to delete this file more robustly
+	     * (or give the filesystem the information it needs to delete the
+	     * file more robustly).
+	     *
+	     * In particular, one problem might be that the filesystem cannot
+	     * extract the information it needs from the above path object
+	     * because Tcl's entire filesystem apparatus (the code in this
+	     * file) has been finalized, and it refuses to pass the internal
+	     * representation to the filesystem.
+	     */
+	}
+
+	/*
+	 * And free up the allocations. This will also of course remove a
+	 * refCount from the Tcl_Filesystem to which this file belongs, which
+	 * could then free up the filesystem if we are exiting.
+	 */
+
+	Tcl_DecrRefCount(tvdlPtr->divertedFile);
+    }
+
+    ckfree(tvdlPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSLink --
+ *
+ *	This function replaces the library version of readlink() and can also
+ *	be used to make links. The appropriate function for the filesystem to
+ *	which pathPtr belongs will be called.
+ *
+ * Results:
+ *	If toPtr is NULL, then the result is a Tcl_Obj specifying the contents
+ *	of the symbolic link given by 'pathPtr', or NULL if the symbolic link
+ *	could not be read. The result is owned by the caller, which should
+ *	call Tcl_DecrRefCount when the result is no longer needed.
+ *
+ *	If toPtr is non-NULL, then the result is toPtr if the link action was
+ *	successful, or NULL if not. In this case the result has no additional
+ *	reference count, and need not be freed. The actual action to perform
+ *	is given by the 'linkAction' flags, which is an or'd combination of:
+ *
+ *		TCL_CREATE_SYMBOLIC_LINK
+ *		TCL_CREATE_HARD_LINK
+ *
+ *	Note that most filesystems will not support linking across to
+ *	different filesystems, so this function will usually fail unless toPtr
+ *	is in the same FS as pathPtr.
+ *
+ * Side effects:
+ *	See readlink() documentation. A new filesystem link object may appear.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSLink(
+    Tcl_Obj *pathPtr,		/* Path of file to readlink or link. */
+    Tcl_Obj *toPtr,		/* NULL or path to be linked to. */
+    int linkAction)		/* Action to perform. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->linkProc != NULL) {
+	return fsPtr->linkProc(pathPtr, toPtr, linkAction);
+    }
+
+    /*
+     * If S_IFLNK isn't defined it means that the machine doesn't support
+     * symbolic links, so the file can't possibly be a symbolic link. Generate
+     * an EINVAL error, which is what happens on machines that do support
+     * symbolic links when you invoke readlink on a file that isn't a symbolic
+     * link.
+     */
+
+#ifndef S_IFLNK
+    errno = EINVAL; /* TODO: Change to Tcl_SetErrno()? */
+#else
+    Tcl_SetErrno(ENOENT);
+#endif /* S_IFLNK */
+    return NULL;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSListVolumes --
+ *
+ *	Lists the currently mounted volumes. The chain of functions that have
+ *	been "inserted" into the filesystem will be called in succession; each
+ *	may return a list of volumes, all of which are added to the result
+ *	until all mounted file systems are listed.
+ *
+ *	Notice that we assume the lists returned by each filesystem (if non
+ *	NULL) have been given a refCount for us already. However, we are NOT
+ *	allowed to hang on to the list itself (it belongs to the filesystem we
+ *	called). Therefore we quite naturally add its contents to the result
+ *	we are building, and then decrement the refCount.
+ *
+ * Results:
+ *	The list of volumes, in an object which has refCount 0.
+ *
+ * Side effects:
+ *	None
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSListVolumes(void)
+{
+    FilesystemRecord *fsRecPtr;
+    Tcl_Obj *resultPtr = Tcl_NewObj();
+
+    /*
+     * Call each of the "listVolumes" function in succession. A non-NULL
+     * return value indicates the particular function has succeeded. We call
+     * all the functions registered, since we want a list of all drives from
+     * all filesystems.
+     */
+
+    fsRecPtr = FsGetFirstFilesystem();
+    Claim();
+    while (fsRecPtr != NULL) {
+	if (fsRecPtr->fsPtr->listVolumesProc != NULL) {
+	    Tcl_Obj *thisFsVolumes = fsRecPtr->fsPtr->listVolumesProc();
+
+	    if (thisFsVolumes != NULL) {
+		Tcl_ListObjAppendList(NULL, resultPtr, thisFsVolumes);
+		Tcl_DecrRefCount(thisFsVolumes);
+	    }
+	}
+	fsRecPtr = fsRecPtr->nextPtr;
+    }
+    Disclaim();
+
+    return resultPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FsListMounts --
+ *
+ *	List all mounts within the given directory, which match the given
+ *	pattern.
+ *
+ * Results:
+ *	The list of mounts, in a list object which has refCount 0, or NULL if
+ *	we didn't even find any filesystems to try to list mounts.
+ *
+ * Side effects:
+ *	None
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+FsListMounts(
+    Tcl_Obj *pathPtr,		/* Contains path to directory to search. */
+    const char *pattern)	/* Pattern to match against. */
+{
+    FilesystemRecord *fsRecPtr;
+    Tcl_GlobTypeData mountsOnly = { TCL_GLOB_TYPE_MOUNT, 0, NULL, NULL };
+    Tcl_Obj *resultPtr = NULL;
+
+    /*
+     * Call each of the "matchInDirectory" functions in succession, with the
+     * specific type information 'mountsOnly'. A non-NULL return value
+     * indicates the particular function has succeeded. We call all the
+     * functions registered, since we want a list from each filesystems.
+     */
+
+    fsRecPtr = FsGetFirstFilesystem();
+    Claim();
+    while (fsRecPtr != NULL) {
+	if (fsRecPtr->fsPtr != &tclNativeFilesystem &&
+		fsRecPtr->fsPtr->matchInDirectoryProc != NULL) {
+	    if (resultPtr == NULL) {
+		resultPtr = Tcl_NewObj();
+	    }
+	    fsRecPtr->fsPtr->matchInDirectoryProc(NULL, resultPtr, pathPtr,
+		    pattern, &mountsOnly);
+	}
+	fsRecPtr = fsRecPtr->nextPtr;
+    }
+    Disclaim();
+
+    return resultPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSSplitPath --
+ *
+ *	This function takes the given Tcl_Obj, which should be a valid path,
+ *	and returns a Tcl List object containing each segment of that path as
+ *	an element.
+ *
+ * Results:
+ *	Returns list object with refCount of zero. If the passed in lenPtr is
+ *	non-NULL, we use it to return the number of elements in the returned
+ *	list.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSSplitPath(
+    Tcl_Obj *pathPtr,		/* Path to split. */
+    int *lenPtr)		/* int to store number of path elements. */
+{
+    Tcl_Obj *result = NULL;	/* Needed only to prevent gcc warnings. */
+    const Tcl_Filesystem *fsPtr;
+    char separator = '/';
+    int driveNameLength;
+    const char *p;
+
+    /*
+     * Perform platform specific splitting.
+     */
+
+    if (TclFSGetPathType(pathPtr, &fsPtr,
+	    &driveNameLength) == TCL_PATH_ABSOLUTE) {
+	if (fsPtr == &tclNativeFilesystem) {
+	    return TclpNativeSplitPath(pathPtr, lenPtr);
+	}
+    } else {
+	return TclpNativeSplitPath(pathPtr, lenPtr);
+    }
+
+    /*
+     * We assume separators are single characters.
+     */
+
+    if (fsPtr->filesystemSeparatorProc != NULL) {
+	Tcl_Obj *sep = fsPtr->filesystemSeparatorProc(pathPtr);
+
+	if (sep != NULL) {
+	    Tcl_IncrRefCount(sep);
+	    separator = Tcl_GetString(sep)[0];
+	    Tcl_DecrRefCount(sep);
+	}
+    }
+
+    /*
+     * Place the drive name as first element of the result list. The drive
+     * name may contain strange characters, like colons and multiple forward
+     * slashes (for example 'ftp://' is a valid vfs drive name)
+     */
+
+    result = Tcl_NewObj();
+    p = Tcl_GetString(pathPtr);
+    Tcl_ListObjAppendElement(NULL, result,
+	    Tcl_NewStringObj(p, driveNameLength));
+    p += driveNameLength;
+
+    /*
+     * Add the remaining path elements to the list.
+     */
+
+    for (;;) {
+	const char *elementStart = p;
+	int length;
+
+	while ((*p != '\0') && (*p != separator)) {
+	    p++;
+	}
+	length = p - elementStart;
+	if (length > 0) {
+	    Tcl_Obj *nextElt;
+
+	    if (elementStart[0] == '~') {
+		TclNewLiteralStringObj(nextElt, "./");
+		Tcl_AppendToObj(nextElt, elementStart, length);
+	    } else {
+		nextElt = Tcl_NewStringObj(elementStart, length);
+	    }
+	    Tcl_ListObjAppendElement(NULL, result, nextElt);
+	}
+	if (*p++ == '\0') {
+	    break;
+	}
+    }
+
+    /*
+     * Compute the number of elements in the result.
+     */
+
+    if (lenPtr != NULL) {
+	TclListObjLength(NULL, result, lenPtr);
+    }
+    return result;
+}
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetPathType --
+ *
+ *	Helper function used by FSGetPathType.
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE, TCL_PATH_RELATIVE, or
+ *	TCL_PATH_VOLUME_RELATIVE. The filesystem reference will be set if and
+ *	only if it is non-NULL and the function's return value is
+ *	TCL_PATH_ABSOLUTE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+TclGetPathType(
+    Tcl_Obj *pathPtr,		/* Path to determine type for. */
+    const Tcl_Filesystem **filesystemPtrPtr,
+				/* If absolute path and this is not NULL, then
+				 * set to the filesystem which claims this
+				 * path. */
+    int *driveNameLengthPtr,	/* If the path is absolute, and this is
+				 * non-NULL, then set to the length of the
+				 * driveName. */
+    Tcl_Obj **driveNameRef)	/* If the path is absolute, and this is
+				 * non-NULL, then set to the name of the
+				 * drive, network-volume which contains the
+				 * path, already with a refCount for the
+				 * caller. */
+{
+    int pathLen;
+    const char *path = TclGetStringFromObj(pathPtr, &pathLen);
+    Tcl_PathType type;
+
+    type = TclFSNonnativePathType(path, pathLen, filesystemPtrPtr,
+	    driveNameLengthPtr, driveNameRef);
+
+    if (type != TCL_PATH_ABSOLUTE) {
+	type = TclpGetNativePathType(pathPtr, driveNameLengthPtr,
+		driveNameRef);
+	if ((type == TCL_PATH_ABSOLUTE) && (filesystemPtrPtr != NULL)) {
+	    *filesystemPtrPtr = &tclNativeFilesystem;
+	}
+    }
+    return type;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFSNonnativePathType --
+ *
+ *	Helper function used by TclGetPathType. Its purpose is to check
+ *	whether the given path starts with a string which corresponds to a
+ *	file volume in any registered filesystem except the native one. For
+ *	speed and historical reasons the native filesystem has special
+ *	hard-coded checks dotted here and there in the filesystem code.
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE or TCL_PATH_RELATIVE. The filesystem
+ *	reference will be set if and only if it is non-NULL and the function's
+ *	return value is TCL_PATH_ABSOLUTE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+TclFSNonnativePathType(
+    const char *path,		/* Path to determine type for. */
+    int pathLen,		/* Length of the path. */
+    const Tcl_Filesystem **filesystemPtrPtr,
+				/* If absolute path and this is not NULL, then
+				 * set to the filesystem which claims this
+				 * path. */
+    int *driveNameLengthPtr,	/* If the path is absolute, and this is
+				 * non-NULL, then set to the length of the
+				 * driveName. */
+    Tcl_Obj **driveNameRef)	/* If the path is absolute, and this is
+				 * non-NULL, then set to the name of the
+				 * drive, network-volume which contains the
+				 * path, already with a refCount for the
+				 * caller. */
+{
+    FilesystemRecord *fsRecPtr;
+    Tcl_PathType type = TCL_PATH_RELATIVE;
+
+    /*
+     * Call each of the "listVolumes" function in succession, checking whether
+     * the given path is an absolute path on any of the volumes returned (this
+     * is done by checking whether the path's prefix matches).
+     */
+
+    fsRecPtr = FsGetFirstFilesystem();
+    Claim();
+    while (fsRecPtr != NULL) {
+	/*
+	 * We want to skip the native filesystem in this loop because
+	 * otherwise we won't necessarily pass all the Tcl testsuite - this is
+	 * because some of the tests artificially change the current platform
+	 * (between win, unix) but the list of volumes we get by calling
+	 * fsRecPtr->fsPtr->listVolumesProc will reflect the current (real)
+	 * platform only and this may cause some tests to fail. In particular,
+	 * on Unix '/' will match the beginning of certain absolute Windows
+	 * paths starting '//' and those tests will go wrong.
+	 *
+	 * Besides these test-suite issues, there is one other reason to skip
+	 * the native filesystem - since the tclFilename.c code has nice fast
+	 * 'absolute path' checkers, we don't want to waste time repeating
+	 * that effort here, and this function is actually called quite often,
+	 * so if we can save the overhead of the native filesystem returning
+	 * us a list of volumes all the time, it is better.
+	 */
+
+	if ((fsRecPtr->fsPtr != &tclNativeFilesystem)
+		&& (fsRecPtr->fsPtr->listVolumesProc != NULL)) {
+	    int numVolumes;
+	    Tcl_Obj *thisFsVolumes = fsRecPtr->fsPtr->listVolumesProc();
+
+	    if (thisFsVolumes != NULL) {
+		if (Tcl_ListObjLength(NULL, thisFsVolumes, &numVolumes)
+			!= TCL_OK) {
+		    /*
+		     * This is VERY bad; the listVolumesProc didn't return a
+		     * valid list. Set numVolumes to -1 so that we skip the
+		     * while loop below and just return with the current value
+		     * of 'type'.
+		     *
+		     * It would be better if we could signal an error here
+		     * (but Tcl_Panic seems a bit excessive).
+		     */
+
+		    numVolumes = -1;
+		}
+		while (numVolumes > 0) {
+		    Tcl_Obj *vol;
+		    int len;
+		    const char *strVol;
+
+		    numVolumes--;
+		    Tcl_ListObjIndex(NULL, thisFsVolumes, numVolumes, &vol);
+		    strVol = TclGetStringFromObj(vol,&len);
+		    if (pathLen < len) {
+			continue;
+		    }
+		    if (strncmp(strVol, path, (size_t) len) == 0) {
+			type = TCL_PATH_ABSOLUTE;
+			if (filesystemPtrPtr != NULL) {
+			    *filesystemPtrPtr = fsRecPtr->fsPtr;
+			}
+			if (driveNameLengthPtr != NULL) {
+			    *driveNameLengthPtr = len;
+			}
+			if (driveNameRef != NULL) {
+			    *driveNameRef = vol;
+			    Tcl_IncrRefCount(vol);
+			}
+			break;
+		    }
+		}
+		Tcl_DecrRefCount(thisFsVolumes);
+		if (type == TCL_PATH_ABSOLUTE) {
+		    /*
+		     * We don't need to examine any more filesystems.
+		     */
+
+		    break;
+		}
+	    }
+	}
+	fsRecPtr = fsRecPtr->nextPtr;
+    }
+    Disclaim();
+    return type;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSRenameFile --
+ *
+ *	If the two paths given belong to the same filesystem, we call that
+ *	filesystems rename function. Otherwise we simply return the POSIX
+ *	error 'EXDEV', and -1.
+ *
+ * Results:
+ *	Standard Tcl error code if a function was called.
+ *
+ * Side effects:
+ *	A file may be renamed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSRenameFile(
+    Tcl_Obj *srcPathPtr,	/* Pathname of file or dir to be renamed
+				 * (UTF-8). */
+    Tcl_Obj *destPathPtr)	/* New pathname of file or directory
+				 * (UTF-8). */
+{
+    int retVal = -1;
+    const Tcl_Filesystem *fsPtr, *fsPtr2;
+
+    fsPtr = Tcl_FSGetFileSystemForPath(srcPathPtr);
+    fsPtr2 = Tcl_FSGetFileSystemForPath(destPathPtr);
+
+    if ((fsPtr == fsPtr2) && (fsPtr != NULL)
+	    && (fsPtr->renameFileProc != NULL)) {
+	retVal = fsPtr->renameFileProc(srcPathPtr, destPathPtr);
+    }
+    if (retVal == -1) {
+	Tcl_SetErrno(EXDEV);
+    }
+    return retVal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSCopyFile --
+ *
+ *	If the two paths given belong to the same filesystem, we call that
+ *	filesystem's copy function. Otherwise we simply return the POSIX error
+ *	'EXDEV', and -1.
+ *
+ *	Note that in the native filesystems, 'copyFileProc' is defined to copy
+ *	soft links (i.e. it copies the links themselves, not the things they
+ *	point to).
+ *
+ * Results:
+ *	Standard Tcl error code if a function was called.
+ *
+ * Side effects:
+ *	A file may be copied.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSCopyFile(
+    Tcl_Obj *srcPathPtr,	/* Pathname of file to be copied (UTF-8). */
+    Tcl_Obj *destPathPtr)	/* Pathname of file to copy to (UTF-8). */
+{
+    int retVal = -1;
+    const Tcl_Filesystem *fsPtr, *fsPtr2;
+
+    fsPtr = Tcl_FSGetFileSystemForPath(srcPathPtr);
+    fsPtr2 = Tcl_FSGetFileSystemForPath(destPathPtr);
+
+    if (fsPtr == fsPtr2 && fsPtr != NULL && fsPtr->copyFileProc != NULL) {
+	retVal = fsPtr->copyFileProc(srcPathPtr, destPathPtr);
+    }
+    if (retVal == -1) {
+	Tcl_SetErrno(EXDEV);
+    }
+    return retVal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclCrossFilesystemCopy --
+ *
+ *	Helper for above function, and for Tcl_FSLoadFile, to copy files from
+ *	one filesystem to another. This function will overwrite the target
+ *	file if it already exists.
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	A file may be created.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclCrossFilesystemCopy(
+    Tcl_Interp *interp,		/* For error messages. */
+    Tcl_Obj *source,		/* Pathname of file to be copied (UTF-8). */
+    Tcl_Obj *target)		/* Pathname of file to copy to (UTF-8). */
+{
+    int result = TCL_ERROR;
+    int prot = 0666;
+    Tcl_Channel in, out;
+    Tcl_StatBuf sourceStatBuf;
+    struct utimbuf tval;
+
+    out = Tcl_FSOpenFileChannel(interp, target, "wb", prot);
+    if (out == NULL) {
+	/*
+	 * It looks like we cannot copy it over. Bail out...
+	 */
+	goto done;
+    }
+
+    in = Tcl_FSOpenFileChannel(interp, source, "rb", prot);
+    if (in == NULL) {
+	/*
+	 * This is very strange, caller should have checked this...
+	 */
+
+	Tcl_Close(interp, out);
+	goto done;
+    }
+
+    /*
+     * Copy it synchronously. We might wish to add an asynchronous option to
+     * support vfs's which are slow (e.g. network sockets).
+     */
+
+    if (TclCopyChannel(interp, in, out, -1, NULL) == TCL_OK) {
+	result = TCL_OK;
+    }
+
+    /*
+     * If the copy failed, assume that copy channel left a good error message.
+     */
+
+    Tcl_Close(interp, in);
+    Tcl_Close(interp, out);
+
+    /*
+     * Set modification date of copied file.
+     */
+
+    if (Tcl_FSLstat(source, &sourceStatBuf) == 0) {
+	tval.actime = sourceStatBuf.st_atime;
+	tval.modtime = sourceStatBuf.st_mtime;
+	Tcl_FSUtime(target, &tval);
+    }
+
+  done:
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSDeleteFile --
+ *
+ *	The appropriate function for the filesystem to which pathPtr belongs
+ *	will be called.
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	A file may be deleted.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSDeleteFile(
+    Tcl_Obj *pathPtr)		/* Pathname of file to be removed (UTF-8). */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->deleteFileProc != NULL) {
+	return fsPtr->deleteFileProc(pathPtr);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSCreateDirectory --
+ *
+ *	The appropriate function for the filesystem to which pathPtr belongs
+ *	will be called.
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	A directory may be created.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSCreateDirectory(
+    Tcl_Obj *pathPtr)		/* Pathname of directory to create (UTF-8). */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr != NULL && fsPtr->createDirectoryProc != NULL) {
+	return fsPtr->createDirectoryProc(pathPtr);
+    }
+    Tcl_SetErrno(ENOENT);
+    return -1;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSCopyDirectory --
+ *
+ *	If the two paths given belong to the same filesystem, we call that
+ *	filesystems copy-directory function. Otherwise we simply return the
+ *	POSIX error 'EXDEV', and -1.
+ *
+ * Results:
+ *	Standard Tcl error code if a function was called.
+ *
+ * Side effects:
+ *	A directory may be copied.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSCopyDirectory(
+    Tcl_Obj *srcPathPtr,	/* Pathname of directory to be copied
+				 * (UTF-8). */
+    Tcl_Obj *destPathPtr,	/* Pathname of target directory (UTF-8). */
+    Tcl_Obj **errorPtr)		/* If non-NULL, then will be set to a new
+				 * object containing name of file causing
+				 * error, with refCount 1. */
+{
+    int retVal = -1;
+    const Tcl_Filesystem *fsPtr, *fsPtr2;
+
+    fsPtr = Tcl_FSGetFileSystemForPath(srcPathPtr);
+    fsPtr2 = Tcl_FSGetFileSystemForPath(destPathPtr);
+
+    if (fsPtr == fsPtr2 && fsPtr != NULL && fsPtr->copyDirectoryProc != NULL){
+	retVal = fsPtr->copyDirectoryProc(srcPathPtr, destPathPtr, errorPtr);
+    }
+    if (retVal == -1) {
+	Tcl_SetErrno(EXDEV);
+    }
+    return retVal;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSRemoveDirectory --
+ *
+ *	The appropriate function for the filesystem to which pathPtr belongs
+ *	will be called.
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	A directory may be deleted.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSRemoveDirectory(
+    Tcl_Obj *pathPtr,		/* Pathname of directory to be removed
+				 * (UTF-8). */
+    int recursive,		/* If non-zero, removes directories that are
+				 * nonempty. Otherwise, will only remove empty
+				 * directories. */
+    Tcl_Obj **errorPtr)		/* If non-NULL, then will be set to a new
+				 * object containing name of file causing
+				 * error, with refCount 1. */
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr == NULL || fsPtr->removeDirectoryProc == NULL) {
+	Tcl_SetErrno(ENOENT);
+	return -1;
+    }
+
+    /*
+     * When working recursively, we check whether the cwd lies inside this
+     * directory and move it if it does.
+     */
+
+    if (recursive) {
+	Tcl_Obj *cwdPtr = Tcl_FSGetCwd(NULL);
+
+	if (cwdPtr != NULL) {
+	    const char *cwdStr, *normPathStr;
+	    int cwdLen, normLen;
+	    Tcl_Obj *normPath = Tcl_FSGetNormalizedPath(NULL, pathPtr);
+
+	    if (normPath != NULL) {
+		normPathStr = TclGetStringFromObj(normPath, &normLen);
+		cwdStr = TclGetStringFromObj(cwdPtr, &cwdLen);
+		if ((cwdLen >= normLen) && (strncmp(normPathStr, cwdStr,
+			(size_t) normLen) == 0)) {
+		    /*
+		     * The cwd is inside the directory, so we perform a 'cd
+		     * [file dirname $path]'.
+		     */
+
+		    Tcl_Obj *dirPtr = TclPathPart(NULL, pathPtr,
+			    TCL_PATH_DIRNAME);
+
+		    Tcl_FSChdir(dirPtr);
+		    Tcl_DecrRefCount(dirPtr);
+		}
+	    }
+	    Tcl_DecrRefCount(cwdPtr);
+	}
+    }
+    return fsPtr->removeDirectoryProc(pathPtr, recursive, errorPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetFileSystemForPath --
+ *
+ *	This function determines which filesystem to use for a particular path
+ *	object, and returns the filesystem which accepts this file. If no
+ *	filesystem will accept this object as a valid file path, then NULL is
+ *	returned.
+ *
+ * Results:
+ *	NULL or a filesystem which will accept this path.
+ *
+ * Side effects:
+ *	The object may be converted to a path type.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const Tcl_Filesystem *
+Tcl_FSGetFileSystemForPath(
+    Tcl_Obj *pathPtr)
+{
+    FilesystemRecord *fsRecPtr;
+    const Tcl_Filesystem *retVal = NULL;
+
+    if (pathPtr == NULL) {
+	Tcl_Panic("Tcl_FSGetFileSystemForPath called with NULL object");
+	return NULL;
+    }
+
+    /*
+     * If the object has a refCount of zero, we reject it. This is to avoid
+     * possible segfaults or nondeterministic memory leaks (i.e. the user
+     * doesn't know if they should decrement the ref count on return or not).
+     */
+
+    if (pathPtr->refCount == 0) {
+	Tcl_Panic("Tcl_FSGetFileSystemForPath called with object with refCount == 0");
+	return NULL;
+    }
+
+    /*
+     * Check if the filesystem has changed in some way since this object's
+     * internal representation was calculated. Before doing that, assure we
+     * have the most up-to-date copy of the master filesystem. This is
+     * accomplished by the FsGetFirstFilesystem() call.
+     */
+
+    fsRecPtr = FsGetFirstFilesystem();
+    Claim();
+
+    if (TclFSEnsureEpochOk(pathPtr, &retVal) != TCL_OK) {
+	Disclaim();
+	return NULL;
+    } else if (retVal != NULL) {
+	/* TODO: Can this happen? */
+	Disclaim();
+	return retVal;
+    }
+
+    /*
+     * Call each of the "pathInFilesystem" functions in succession. A
+     * non-return value of -1 indicates the particular function has succeeded.
+     */
+
+    for (; fsRecPtr!=NULL ; fsRecPtr=fsRecPtr->nextPtr) {
+	ClientData clientData = NULL;
+
+	if (fsRecPtr->fsPtr->pathInFilesystemProc == NULL) {
+	    continue;
+	}
+
+	if (fsRecPtr->fsPtr->pathInFilesystemProc(pathPtr, &clientData)!=-1) {
+	    /*
+	     * We assume the type of pathPtr hasn't been changed by the above
+	     * call to the pathInFilesystemProc.
+	     */
+
+	    TclFSSetPathDetails(pathPtr, fsRecPtr->fsPtr, clientData);
+	    Disclaim();
+	    return fsRecPtr->fsPtr;
+	}
+    }
+    Disclaim();
+
+    return NULL;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetNativePath --
+ *
+ *	This function is for use by the Win/Unix native filesystems, so that
+ *	they can easily retrieve the native (char* or TCHAR*) representation
+ *	of a path. Other filesystems will probably want to implement similar
+ *	functions. They basically act as a safety net around
+ *	Tcl_FSGetInternalRep. Normally your file-system functions will always
+ *	be called with path objects already converted to the correct
+ *	filesystem, but if for some reason they are called directly (i.e. by
+ *	functions not in this file), then one cannot necessarily guarantee
+ *	that the path object pointer is from the correct filesystem.
+ *
+ *	Note: in the future it might be desirable to have separate versions
+ *	of this function with different signatures, for example
+ *	Tcl_FSGetNativeWinPath, Tcl_FSGetNativeUnixPath etc. Right now, since
+ *	native paths are all string based, we use just one function.
+ *
+ * Results:
+ *	NULL or a valid native path.
+ *
+ * Side effects:
+ *	See Tcl_FSGetInternalRep.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const void *
+Tcl_FSGetNativePath(
+    Tcl_Obj *pathPtr)
+{
+    return Tcl_FSGetInternalRep(pathPtr, &tclNativeFilesystem);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * NativeFreeInternalRep --
+ *
+ *	Free a native internal representation, which will be non-NULL.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is released.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+NativeFreeInternalRep(
+    ClientData clientData)
+{
+    ckfree(clientData);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSFileSystemInfo --
+ *
+ *	This function returns a list of two elements. The first element is the
+ *	name of the filesystem (e.g. "native" or "vfs"), and the second is the
+ *	particular type of the given path within that filesystem.
+ *
+ * Results:
+ *	A list of two elements.
+ *
+ * Side effects:
+ *	The object may be converted to a path type.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSFileSystemInfo(
+    Tcl_Obj *pathPtr)
+{
+    Tcl_Obj *resPtr;
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+
+    if (fsPtr == NULL) {
+	return NULL;
+    }
+
+    resPtr = Tcl_NewListObj(0, NULL);
+    Tcl_ListObjAppendElement(NULL, resPtr,
+	    Tcl_NewStringObj(fsPtr->typeName, -1));
+
+    if (fsPtr->filesystemPathTypeProc != NULL) {
+	Tcl_Obj *typePtr = fsPtr->filesystemPathTypeProc(pathPtr);
+
+	if (typePtr != NULL) {
+	    Tcl_ListObjAppendElement(NULL, resPtr, typePtr);
+	}
+    }
+
+    return resPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSPathSeparator --
+ *
+ *	This function returns the separator to be used for a given path. The
+ *	object returned should have a refCount of zero
+ *
+ * Results:
+ *	A Tcl object, with a refCount of zero. If the caller needs to retain a
+ *	reference to the object, it should call Tcl_IncrRefCount, and should
+ *	otherwise free the object.
+ *
+ * Side effects:
+ *	The path object may be converted to a path type.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSPathSeparator(
+    Tcl_Obj *pathPtr)
+{
+    const Tcl_Filesystem *fsPtr = Tcl_FSGetFileSystemForPath(pathPtr);
+    Tcl_Obj *resultObj;
+
+    if (fsPtr == NULL) {
+	return NULL;
+    }
+
+    if (fsPtr->filesystemSeparatorProc != NULL) {
+	return fsPtr->filesystemSeparatorProc(pathPtr);
+    }
+
+    /*
+     * Allow filesystems not to provide a filesystemSeparatorProc if they wish
+     * to use the standard forward slash.
+     */
+
+    TclNewLiteralStringObj(resultObj, "/");
+    return resultObj;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * NativeFilesystemSeparator --
+ *
+ *	This function is part of the native filesystem support, and returns
+ *	the separator for the given path.
+ *
+ * Results:
+ *	String object containing the separator character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+NativeFilesystemSeparator(
+    Tcl_Obj *pathPtr)
+{
+    const char *separator = NULL; /* lint */
+
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	separator = "/";
+	break;
+    case TCL_PLATFORM_WINDOWS:
+	separator = "\\";
+	break;
+    }
+    return Tcl_NewStringObj(separator,1);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIndexObj.c b/generic/tclIndexObj.c
new file mode 100644
index 0000000..9f38638
--- /dev/null
+++ b/generic/tclIndexObj.c
@@ -0,0 +1,1487 @@
+/*
+ * tclIndexObj.c --
+ *
+ *	This file implements objects of type "index". This object type is used
+ *	to lookup a keyword in a table of valid values and cache the index of
+ *	the matching entry. Also provides table-based argv/argc processing.
+ *
+ * Copyright (c) 1990-1994 The Regents of the University of California.
+ * Copyright (c) 1997 Sun Microsystems, Inc.
+ * Copyright (c) 2006 Sam Bromley.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Prototypes for functions defined later in this file:
+ */
+
+static int		GetIndexFromObjList(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr, Tcl_Obj *tableObjPtr,
+			    const char *msg, int flags, int *indexPtr);
+static int		SetIndexFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		UpdateStringOfIndex(Tcl_Obj *objPtr);
+static void		DupIndex(Tcl_Obj *srcPtr, Tcl_Obj *dupPtr);
+static void		FreeIndex(Tcl_Obj *objPtr);
+static int		PrefixAllObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		PrefixLongestObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		PrefixMatchObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		PrintUsage(Tcl_Interp *interp,
+			    const Tcl_ArgvInfo *argTable);
+
+/*
+ * The structure below defines the index Tcl object type by means of functions
+ * that can be invoked by generic object code.
+ */
+
+static const Tcl_ObjType indexType = {
+    "index",			/* name */
+    FreeIndex,			/* freeIntRepProc */
+    DupIndex,			/* dupIntRepProc */
+    UpdateStringOfIndex,	/* updateStringProc */
+    SetIndexFromAny		/* setFromAnyProc */
+};
+
+/*
+ * The definition of the internal representation of the "index" object; The
+ * internalRep.twoPtrValue.ptr1 field of an object of "index" type will be a
+ * pointer to one of these structures.
+ *
+ * Keep this structure declaration in sync with tclTestObj.c
+ */
+
+typedef struct {
+    void *tablePtr;		/* Pointer to the table of strings */
+    int offset;			/* Offset between table entries */
+    int index;			/* Selected index into table. */
+} IndexRep;
+
+/*
+ * The following macros greatly simplify moving through a table...
+ */
+
+#define STRING_AT(table, offset) \
+	(*((const char *const *)(((char *)(table)) + (offset))))
+#define NEXT_ENTRY(table, offset) \
+	(&(STRING_AT(table, offset)))
+#define EXPAND_OF(indexRep) \
+	STRING_AT((indexRep)->tablePtr, (indexRep)->offset*(indexRep)->index)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetIndexFromObj --
+ *
+ *	This function looks up an object's value in a table of strings and
+ *	returns the index of the matching string, if any.
+ *
+ * Results:
+ *	If the value of objPtr is identical to or a unique abbreviation for
+ *	one of the entries in tablePtr, then the return value is TCL_OK and the
+ *	index of the matching entry is stored at *indexPtr. If there isn't a
+ *	proper match, then TCL_ERROR is returned and an error message is left
+ *	in interp's result (unless interp is NULL). The msg argument is used
+ *	in the error message; for example, if msg has the value "option" then
+ *	the error message will say something flag 'bad option "foo": must be
+ *	...'
+ *
+ * Side effects:
+ *	The result of the lookup is cached as the internal rep of objPtr, so
+ *	that repeated lookups can be done quickly.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_GetIndexFromObj
+int
+Tcl_GetIndexFromObj(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr,		/* Object containing the string to lookup. */
+    const char *const*tablePtr,	/* Array of strings to compare against the
+				 * value of objPtr; last entry must be NULL
+				 * and there must not be duplicate entries. */
+    const char *msg,		/* Identifying word to use in error
+				 * messages. */
+    int flags,			/* 0 or TCL_EXACT */
+    int *indexPtr)		/* Place to store resulting integer index. */
+{
+    /*
+     * See if there is a valid cached result from a previous lookup (doing the
+     * check here saves the overhead of calling Tcl_GetIndexFromObjStruct in
+     * the common case where the result is cached).
+     */
+
+    if (!(flags & INDEX_TEMP_TABLE) && objPtr->typePtr == &indexType) {
+	IndexRep *indexRep = objPtr->internalRep.twoPtrValue.ptr1;
+
+	/*
+	 * Here's hoping we don't get hit by unfortunate packing constraints
+	 * on odd platforms like a Cray PVP...
+	 */
+
+	if (indexRep->tablePtr == (void *) tablePtr
+		&& indexRep->offset == sizeof(char *)) {
+	    *indexPtr = indexRep->index;
+	    return TCL_OK;
+	}
+    }
+    return Tcl_GetIndexFromObjStruct(interp, objPtr, tablePtr, sizeof(char *),
+	    msg, flags, indexPtr);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetIndexFromObjList --
+ *
+ *	This procedure looks up an object's value in a table of strings and
+ *	returns the index of the matching string, if any.
+ *
+ * Results:
+ *	If the value of objPtr is identical to or a unique abbreviation for
+ *	one of the entries in tableObjPtr, then the return value is TCL_OK and
+ *	the index of the matching entry is stored at *indexPtr. If there isn't
+ *	a proper match, then TCL_ERROR is returned and an error message is
+ *	left in interp's result (unless interp is NULL). The msg argument is
+ *	used in the error message; for example, if msg has the value "option"
+ *	then the error message will say something flag 'bad option "foo": must
+ *	be ...'
+ *
+ * Side effects:
+ *	Removes any internal representation that the object might have. (TODO:
+ *	find a way to cache the lookup.)
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+GetIndexFromObjList(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr,		/* Object containing the string to lookup. */
+    Tcl_Obj *tableObjPtr,	/* List of strings to compare against the
+				 * value of objPtr. */
+    const char *msg,		/* Identifying word to use in error
+				 * messages. */
+    int flags,			/* 0 or TCL_EXACT */
+    int *indexPtr)		/* Place to store resulting integer index. */
+{
+
+    int objc, result, t;
+    Tcl_Obj **objv;
+    const char **tablePtr;
+
+    /*
+     * Use Tcl_GetIndexFromObjStruct to do the work to avoid duplicating most
+     * of the code there. This is a bit ineffiecient but simpler.
+     */
+
+    result = Tcl_ListObjGetElements(interp, tableObjPtr, &objc, &objv);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    /*
+     * Build a string table from the list.
+     */
+
+    tablePtr = ckalloc((objc + 1) * sizeof(char *));
+    for (t = 0; t < objc; t++) {
+	if (objv[t] == objPtr) {
+	    /*
+	     * An exact match is always chosen, so we can stop here.
+	     */
+
+	    ckfree(tablePtr);
+	    *indexPtr = t;
+	    return TCL_OK;
+	}
+
+	tablePtr[t] = Tcl_GetString(objv[t]);
+    }
+    tablePtr[objc] = NULL;
+
+    result = Tcl_GetIndexFromObjStruct(interp, objPtr, tablePtr,
+	    sizeof(char *), msg, flags | INDEX_TEMP_TABLE, indexPtr);
+
+    ckfree(tablePtr);
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetIndexFromObjStruct --
+ *
+ *	This function looks up an object's value given a starting string and
+ *	an offset for the amount of space between strings. This is useful when
+ *	the strings are embedded in some other kind of array.
+ *
+ * Results:
+ *	If the value of objPtr is identical to or a unique abbreviation for
+ *	one of the entries in tablePtr, then the return value is TCL_OK and
+ *	the index of the matching entry is stored at *indexPtr. If there isn't
+ *	a proper match, then TCL_ERROR is returned and an error message is
+ *	left in interp's result (unless interp is NULL). The msg argument is
+ *	used in the error message; for example, if msg has the value "option"
+ *	then the error message will say something like 'bad option "foo": must
+ *	be ...'
+ *
+ * Side effects:
+ *	The result of the lookup is cached as the internal rep of objPtr, so
+ *	that repeated lookups can be done quickly.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetIndexFromObjStruct(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr,		/* Object containing the string to lookup. */
+    const void *tablePtr,	/* The first string in the table. The second
+				 * string will be at this address plus the
+				 * offset, the third plus the offset again,
+				 * etc. The last entry must be NULL and there
+				 * must not be duplicate entries. */
+    int offset,			/* The number of bytes between entries */
+    const char *msg,		/* Identifying word to use in error
+				 * messages. */
+    int flags,			/* 0 or TCL_EXACT */
+    int *indexPtr)		/* Place to store resulting integer index. */
+{
+    int index, idx, numAbbrev;
+    const char *key, *p1;
+    const char *p2;
+    const char *const *entryPtr;
+    Tcl_Obj *resultPtr;
+    IndexRep *indexRep;
+
+    /* Protect against invalid values, like -1 or 0. */
+    if (offset < (int)sizeof(char *)) {
+	offset = (int)sizeof(char *);
+    }
+    /*
+     * See if there is a valid cached result from a previous lookup.
+     */
+
+    if (!(flags & INDEX_TEMP_TABLE) && objPtr->typePtr == &indexType) {
+	indexRep = objPtr->internalRep.twoPtrValue.ptr1;
+	if (indexRep->tablePtr==tablePtr && indexRep->offset==offset) {
+	    *indexPtr = indexRep->index;
+	    return TCL_OK;
+	}
+    }
+
+    /*
+     * Lookup the value of the object in the table. Accept unique
+     * abbreviations unless TCL_EXACT is set in flags.
+     */
+
+    key = TclGetString(objPtr);
+    index = -1;
+    numAbbrev = 0;
+
+    /*
+     * Scan the table looking for one of:
+     *  - An exact match (always preferred)
+     *  - A single abbreviation (allowed depending on flags)
+     *  - Several abbreviations (never allowed, but overridden by exact match)
+     */
+
+    for (entryPtr = tablePtr, idx = 0; *entryPtr != NULL;
+	    entryPtr = NEXT_ENTRY(entryPtr, offset), idx++) {
+	for (p1 = key, p2 = *entryPtr; *p1 == *p2; p1++, p2++) {
+	    if (*p1 == '\0') {
+		index = idx;
+		goto done;
+	    }
+	}
+	if (*p1 == '\0') {
+	    /*
+	     * The value is an abbreviation for this entry. Continue checking
+	     * other entries to make sure it's unique. If we get more than one
+	     * unique abbreviation, keep searching to see if there is an exact
+	     * match, but remember the number of unique abbreviations and
+	     * don't allow either.
+	     */
+
+	    numAbbrev++;
+	    index = idx;
+	}
+    }
+
+    /*
+     * Check if we were instructed to disallow abbreviations.
+     */
+
+    if ((flags & TCL_EXACT) || (key[0] == '\0') || (numAbbrev != 1)) {
+	goto error;
+    }
+
+  done:
+    /*
+     * Cache the found representation. Note that we want to avoid allocating a
+     * new internal-rep if at all possible since that is potentially a slow
+     * operation.
+     */
+
+    if (!(flags & INDEX_TEMP_TABLE)) {
+	if (objPtr->typePtr == &indexType) {
+	    indexRep = objPtr->internalRep.twoPtrValue.ptr1;
+	} else {
+	    TclFreeIntRep(objPtr);
+	    indexRep = ckalloc(sizeof(IndexRep));
+	    objPtr->internalRep.twoPtrValue.ptr1 = indexRep;
+	    objPtr->typePtr = &indexType;
+	}
+	indexRep->tablePtr = (void *) tablePtr;
+	indexRep->offset = offset;
+	indexRep->index = index;
+    }
+
+    *indexPtr = index;
+    return TCL_OK;
+
+  error:
+    if (interp != NULL) {
+	/*
+	 * Produce a fancy error message.
+	 */
+
+	int count = 0;
+
+	TclNewObj(resultPtr);
+	entryPtr = tablePtr;
+	while ((*entryPtr != NULL) && !**entryPtr) {
+	    entryPtr = NEXT_ENTRY(entryPtr, offset);
+	}
+	Tcl_AppendStringsToObj(resultPtr,
+		(numAbbrev>1 && !(flags & TCL_EXACT) ? "ambiguous " : "bad "),
+		msg, " \"", key, NULL);
+	if (*entryPtr == NULL) {
+	    Tcl_AppendStringsToObj(resultPtr, "\": no valid options", NULL);
+	} else {
+	    Tcl_AppendStringsToObj(resultPtr, "\": must be ",
+		    *entryPtr, NULL);
+	    entryPtr = NEXT_ENTRY(entryPtr, offset);
+	    while (*entryPtr != NULL) {
+		if (*NEXT_ENTRY(entryPtr, offset) == NULL) {
+		    Tcl_AppendStringsToObj(resultPtr, (count > 0 ? "," : ""),
+			    " or ", *entryPtr, NULL);
+		} else if (**entryPtr) {
+		    Tcl_AppendStringsToObj(resultPtr, ", ", *entryPtr, NULL);
+		    count++;
+		}
+		entryPtr = NEXT_ENTRY(entryPtr, offset);
+	    }
+	}
+	Tcl_SetObjResult(interp, resultPtr);
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", msg, key, NULL);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetIndexFromAny --
+ *
+ *	This function is called to convert a Tcl object to index internal
+ *	form. However, this doesn't make sense (need to have a table of
+ *	keywords in order to do the conversion) so the function always
+ *	generates an error.
+ *
+ * Results:
+ *	The return value is always TCL_ERROR, and an error message is left in
+ *	interp's result if interp isn't NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetIndexFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+	    "can't convert value to index except via Tcl_GetIndexFromObj API",
+	    -1));
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfIndex --
+ *
+ *	This function is called to convert a Tcl object from index internal
+ *	form to its string form. No abbreviation is ever generated.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The string representation of the object is updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfIndex(
+    Tcl_Obj *objPtr)
+{
+    IndexRep *indexRep = objPtr->internalRep.twoPtrValue.ptr1;
+    register char *buf;
+    register unsigned len;
+    register const char *indexStr = EXPAND_OF(indexRep);
+
+    len = strlen(indexStr);
+    buf = ckalloc(len + 1);
+    memcpy(buf, indexStr, len+1);
+    objPtr->bytes = buf;
+    objPtr->length = len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupIndex --
+ *
+ *	This function is called to copy the internal rep of an index Tcl
+ *	object from to another object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The internal representation of the target object is updated and the
+ *	type is set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupIndex(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *dupPtr)
+{
+    IndexRep *srcIndexRep = srcPtr->internalRep.twoPtrValue.ptr1;
+    IndexRep *dupIndexRep = ckalloc(sizeof(IndexRep));
+
+    memcpy(dupIndexRep, srcIndexRep, sizeof(IndexRep));
+    dupPtr->internalRep.twoPtrValue.ptr1 = dupIndexRep;
+    dupPtr->typePtr = &indexType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeIndex --
+ *
+ *	This function is called to delete the internal rep of an index Tcl
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The internal representation of the target object is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeIndex(
+    Tcl_Obj *objPtr)
+{
+    ckfree(objPtr->internalRep.twoPtrValue.ptr1);
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitPrefixCmd --
+ *
+ *	This procedure creates the "prefix" Tcl command. See the user
+ *	documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitPrefixCmd(
+    Tcl_Interp *interp)		/* Current interpreter. */
+{
+    static const EnsembleImplMap prefixImplMap[] = {
+	{"all",	    PrefixAllObjCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"longest", PrefixLongestObjCmd,TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"match",   PrefixMatchObjCmd,	TclCompileBasicMin2ArgCmd, NULL, NULL, 0},
+	{NULL, NULL, NULL, NULL, NULL, 0}
+    };
+    Tcl_Command prefixCmd;
+
+    prefixCmd = TclMakeEnsemble(interp, "::tcl::prefix", prefixImplMap);
+    Tcl_Export(interp, Tcl_FindNamespace(interp, "::tcl", NULL, 0),
+	    "prefix", 0);
+    return prefixCmd;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * PrefixMatchObjCmd --
+ *
+ *	This function implements the 'prefix match' Tcl command. Refer to the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PrefixMatchObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int flags = 0, result, index;
+    int dummyLength, i, errorLength;
+    Tcl_Obj *errorPtr = NULL;
+    const char *message = "option";
+    Tcl_Obj *tablePtr, *objPtr, *resultPtr;
+    static const char *const matchOptions[] = {
+	"-error", "-exact", "-message", NULL
+    };
+    enum matchOptions {
+	PRFMATCH_ERROR, PRFMATCH_EXACT, PRFMATCH_MESSAGE
+    };
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?options? table string");
+	return TCL_ERROR;
+    }
+
+    for (i = 1; i < (objc - 2); i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], matchOptions, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum matchOptions) index) {
+	case PRFMATCH_EXACT:
+	    flags |= TCL_EXACT;
+	    break;
+	case PRFMATCH_MESSAGE:
+	    if (i > objc-4) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing value for -message", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "NOARG", NULL);
+		return TCL_ERROR;
+	    }
+	    i++;
+	    message = Tcl_GetString(objv[i]);
+	    break;
+	case PRFMATCH_ERROR:
+	    if (i > objc-4) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"missing value for -error", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "NOARG", NULL);
+		return TCL_ERROR;
+	    }
+	    i++;
+	    result = Tcl_ListObjLength(interp, objv[i], &errorLength);
+	    if (result != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if ((errorLength % 2) != 0) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"error options must have an even number of elements",
+			-1));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "DICTIONARY", NULL);
+		return TCL_ERROR;
+	    }
+	    errorPtr = objv[i];
+	    break;
+	}
+    }
+
+    tablePtr = objv[objc - 2];
+    objPtr = objv[objc - 1];
+
+    /*
+     * Check that table is a valid list first, since we want to handle that
+     * error case regardless of level.
+     */
+
+    result = Tcl_ListObjLength(interp, tablePtr, &dummyLength);
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    result = GetIndexFromObjList(interp, objPtr, tablePtr, message, flags,
+	    &index);
+    if (result != TCL_OK) {
+	if (errorPtr != NULL && errorLength == 0) {
+	    Tcl_ResetResult(interp);
+	    return TCL_OK;
+	} else if (errorPtr == NULL) {
+	    return TCL_ERROR;
+	}
+
+	if (Tcl_IsShared(errorPtr)) {
+	    errorPtr = Tcl_DuplicateObj(errorPtr);
+	}
+	Tcl_ListObjAppendElement(interp, errorPtr,
+		Tcl_NewStringObj("-code", 5));
+	Tcl_ListObjAppendElement(interp, errorPtr, Tcl_NewIntObj(result));
+
+	return Tcl_SetReturnOptions(interp, errorPtr);
+    }
+
+    result = Tcl_ListObjIndex(interp, tablePtr, index, &resultPtr);
+    if (result != TCL_OK) {
+	return result;
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * PrefixAllObjCmd --
+ *
+ *	This function implements the 'prefix all' Tcl command. Refer to the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PrefixAllObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int tableObjc, result, t, length, elemLength;
+    const char *string, *elemString;
+    Tcl_Obj **tableObjv, *resultPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "table string");
+	return TCL_ERROR;
+    }
+
+    result = Tcl_ListObjGetElements(interp, objv[1], &tableObjc, &tableObjv);
+    if (result != TCL_OK) {
+	return result;
+    }
+    resultPtr = Tcl_NewListObj(0, NULL);
+    string = TclGetStringFromObj(objv[2], &length);
+
+    for (t = 0; t < tableObjc; t++) {
+	elemString = TclGetStringFromObj(tableObjv[t], &elemLength);
+
+	/*
+	 * A prefix cannot match if it is longest.
+	 */
+
+	if (length <= elemLength) {
+	    if (TclpUtfNcmp2(elemString, string, length) == 0) {
+		Tcl_ListObjAppendElement(interp, resultPtr, tableObjv[t]);
+	    }
+	}
+    }
+
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * PrefixLongestObjCmd --
+ *
+ *	This function implements the 'prefix longest' Tcl command. Refer to
+ *	the user documentation for details on what it does.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+PrefixLongestObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int tableObjc, result, i, t, length, elemLength, resultLength;
+    const char *string, *elemString, *resultString;
+    Tcl_Obj **tableObjv;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "table string");
+	return TCL_ERROR;
+    }
+
+    result = Tcl_ListObjGetElements(interp, objv[1], &tableObjc, &tableObjv);
+    if (result != TCL_OK) {
+	return result;
+    }
+    string = TclGetStringFromObj(objv[2], &length);
+
+    resultString = NULL;
+    resultLength = 0;
+
+    for (t = 0; t < tableObjc; t++) {
+	elemString = TclGetStringFromObj(tableObjv[t], &elemLength);
+
+	/*
+	 * First check if the prefix string matches the element. A prefix
+	 * cannot match if it is longest.
+	 */
+
+	if ((length > elemLength) ||
+		TclpUtfNcmp2(elemString, string, length) != 0) {
+	    continue;
+	}
+
+	if (resultString == NULL) {
+	    /*
+	     * If this is the first match, the longest common substring this
+	     * far is the complete string. The result is part of this string
+	     * so we only need to adjust the length later.
+	     */
+
+	    resultString = elemString;
+	    resultLength = elemLength;
+	} else {
+	    /*
+	     * Longest common substring cannot be longer than shortest string.
+	     */
+
+	    if (elemLength < resultLength) {
+		resultLength = elemLength;
+	    }
+
+	    /*
+	     * Compare strings.
+	     */
+
+	    for (i = 0; i < resultLength; i++) {
+		if (resultString[i] != elemString[i]) {
+		    /*
+		     * Adjust in case we stopped in the middle of a UTF char.
+		     */
+
+		    resultLength = Tcl_UtfPrev(&resultString[i+1],
+			    resultString) - resultString;
+		    break;
+		}
+	    }
+	}
+    }
+    if (resultLength > 0) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj(resultString, resultLength));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_WrongNumArgs --
+ *
+ *	This function generates a "wrong # args" error message in an
+ *	interpreter. It is used as a utility function by many command
+ *	functions, including the function that implements procedures.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	An error message is generated in interp's result object to indicate
+ *	that a command was invoked with the wrong number of arguments. The
+ *	message has the form
+ *		wrong # args: should be "foo bar additional stuff"
+ *	where "foo" and "bar" are the initial objects in objv (objc determines
+ *	how many of these are printed) and "additional stuff" is the contents
+ *	of the message argument.
+ *
+ *	The message printed is modified somewhat if the command is wrapped
+ *	inside an ensemble. In that case, the error message generated is
+ *	rewritten in such a way that it appears to be generated from the
+ *	user-visible command and not how that command is actually implemented,
+ *	giving a better overall user experience.
+ *
+ *	Internally, the Tcl core may set the flag INTERP_ALTERNATE_WRONG_ARGS
+ *	in the interpreter to generate complex multi-part messages by calling
+ *	this function repeatedly. This allows the code that knows how to
+ *	handle ensemble-related error messages to be kept here while still
+ *	generating suitable error messages for commands like [read] and
+ *	[socket]. Ideally, this would be done through an extra flags argument,
+ *	but that wouldn't be source-compatible with the existing API and it's
+ *	a fairly rare requirement anyway.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_WrongNumArgs(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments to print from objv. */
+    Tcl_Obj *const objv[],	/* Initial argument objects, which should be
+				 * included in the error message. */
+    const char *message)	/* Error message to print after the leading
+				 * objects in objv. The message may be
+				 * NULL. */
+{
+    Tcl_Obj *objPtr;
+    int i, len, elemLen, flags;
+    Interp *iPtr = (Interp *) interp;
+    const char *elementStr;
+
+    /*
+     * [incr Tcl] does something fairly horrific when generating error
+     * messages for its ensembles; it passes the whole set of ensemble
+     * arguments as a list in the first argument. This means that this code
+     * causes a problem in iTcl if it attempts to correctly quote all
+     * arguments, which would be the correct thing to do. We work around this
+     * nasty behaviour for now, and hope that we can remove it all in the
+     * future...
+     */
+
+#ifndef AVOID_HACKS_FOR_ITCL
+    int isFirst = 1;		/* Special flag used to inhibit the treating
+				 * of the first word as a list element so the
+				 * hacky way Itcl generates error messages for
+				 * its ensembles will still work. [Bug
+				 * 1066837] */
+#   define MAY_QUOTE_WORD	(!isFirst)
+#   define AFTER_FIRST_WORD	(isFirst = 0)
+#else /* !AVOID_HACKS_FOR_ITCL */
+#   define MAY_QUOTE_WORD	1
+#   define AFTER_FIRST_WORD	(void) 0
+#endif /* AVOID_HACKS_FOR_ITCL */
+
+    TclNewObj(objPtr);
+    if (iPtr->flags & INTERP_ALTERNATE_WRONG_ARGS) {
+	iPtr->flags &= ~INTERP_ALTERNATE_WRONG_ARGS;
+	Tcl_AppendObjToObj(objPtr, Tcl_GetObjResult(interp));
+	Tcl_AppendToObj(objPtr, " or \"", -1);
+    } else {
+	Tcl_AppendToObj(objPtr, "wrong # args: should be \"", -1);
+    }
+
+    /*
+     * Check to see if we are processing an ensemble implementation, and if so
+     * rewrite the results in terms of how the ensemble was invoked.
+     */
+
+    if (iPtr->ensembleRewrite.sourceObjs != NULL) {
+	int toSkip = iPtr->ensembleRewrite.numInsertedObjs;
+	int toPrint = iPtr->ensembleRewrite.numRemovedObjs;
+	Tcl_Obj *const *origObjv = iPtr->ensembleRewrite.sourceObjs;
+
+	/*
+	 * Check for spelling fixes, and substitute the fixed values.
+	 */
+
+	if (origObjv[0] == NULL) {
+	    origObjv = (Tcl_Obj *const *)origObjv[2];
+	}
+
+	/*
+	 * We only know how to do rewriting if all the replaced objects are
+	 * actually arguments (in objv) to this function. Otherwise it just
+	 * gets too complicated and we'd be better off just giving a slightly
+	 * confusing error message...
+	 */
+
+	if (objc < toSkip) {
+	    goto addNormalArgumentsToMessage;
+	}
+
+	/*
+	 * Strip out the actual arguments that the ensemble inserted.
+	 */
+
+	objv += toSkip;
+	objc -= toSkip;
+
+	/*
+	 * We assume no object is of index type.
+	 */
+
+	for (i=0 ; i<toPrint ; i++) {
+	    /*
+	     * Add the element, quoting it if necessary.
+	     */
+
+	    if (origObjv[i]->typePtr == &indexType) {
+		register IndexRep *indexRep =
+			origObjv[i]->internalRep.twoPtrValue.ptr1;
+
+		elementStr = EXPAND_OF(indexRep);
+		elemLen = strlen(elementStr);
+	    } else {
+		elementStr = TclGetStringFromObj(origObjv[i], &elemLen);
+	    }
+	    flags = 0;
+	    len = TclScanElement(elementStr, elemLen, &flags);
+
+	    if (MAY_QUOTE_WORD && len != elemLen) {
+		char *quotedElementStr = TclStackAlloc(interp,
+			(unsigned)len + 1);
+
+		len = TclConvertElement(elementStr, elemLen,
+			quotedElementStr, flags);
+		Tcl_AppendToObj(objPtr, quotedElementStr, len);
+		TclStackFree(interp, quotedElementStr);
+	    } else {
+		Tcl_AppendToObj(objPtr, elementStr, elemLen);
+	    }
+
+	    AFTER_FIRST_WORD;
+
+	    /*
+	     * Add a space if the word is not the last one (which has a
+	     * moderately complex condition here).
+	     */
+
+	    if (i<toPrint-1 || objc!=0 || message!=NULL) {
+		Tcl_AppendStringsToObj(objPtr, " ", NULL);
+	    }
+	}
+    }
+
+    /*
+     * Now add the arguments (other than those rewritten) that the caller took
+     * from its calling context.
+     */
+
+  addNormalArgumentsToMessage:
+    for (i = 0; i < objc; i++) {
+	/*
+	 * If the object is an index type use the index table which allows for
+	 * the correct error message even if the subcommand was abbreviated.
+	 * Otherwise, just use the string rep.
+	 */
+
+	if (objv[i]->typePtr == &indexType) {
+	    register IndexRep *indexRep = objv[i]->internalRep.twoPtrValue.ptr1;
+
+	    Tcl_AppendStringsToObj(objPtr, EXPAND_OF(indexRep), NULL);
+	} else {
+	    /*
+	     * Quote the argument if it contains spaces (Bug 942757).
+	     */
+
+	    elementStr = TclGetStringFromObj(objv[i], &elemLen);
+	    flags = 0;
+	    len = TclScanElement(elementStr, elemLen, &flags);
+
+	    if (MAY_QUOTE_WORD && len != elemLen) {
+		char *quotedElementStr = TclStackAlloc(interp,
+			(unsigned) len + 1);
+
+		len = TclConvertElement(elementStr, elemLen,
+			quotedElementStr, flags);
+		Tcl_AppendToObj(objPtr, quotedElementStr, len);
+		TclStackFree(interp, quotedElementStr);
+	    } else {
+		Tcl_AppendToObj(objPtr, elementStr, elemLen);
+	    }
+	}
+
+	AFTER_FIRST_WORD;
+
+	/*
+	 * Append a space character (" ") if there is more text to follow
+	 * (either another element from objv, or the message string).
+	 */
+
+	if (i<objc-1 || message!=NULL) {
+	    Tcl_AppendStringsToObj(objPtr, " ", NULL);
+	}
+    }
+
+    /*
+     * Add any trailing message bits and set the resulting string as the
+     * interpreter result. Caller is responsible for reporting this as an
+     * actual error.
+     */
+
+    if (message != NULL) {
+	Tcl_AppendStringsToObj(objPtr, message, NULL);
+    }
+    Tcl_AppendStringsToObj(objPtr, "\"", NULL);
+    Tcl_SetErrorCode(interp, "TCL", "WRONGARGS", NULL);
+    Tcl_SetObjResult(interp, objPtr);
+#undef MAY_QUOTE_WORD
+#undef AFTER_FIRST_WORD
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseArgsObjv --
+ *
+ *	Process an objv array according to a table of expected command-line
+ *	options. See the manual page for more details.
+ *
+ * Results:
+ *	The return value is a standard Tcl return value. If an error occurs
+ *	then an error message is left in the interp's result. Under normal
+ *	conditions, both *objcPtr and *objv are modified to return the
+ *	arguments that couldn't be processed here (they didn't match the
+ *	option table, or followed an TCL_ARGV_REST argument).
+ *
+ * Side effects:
+ *	Variables may be modified, or procedures may be called. It all depends
+ *	on the arguments and their entries in argTable. See the user
+ *	documentation for details.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseArgsObjv(
+    Tcl_Interp *interp,		/* Place to store error message. */
+    const Tcl_ArgvInfo *argTable,
+				/* Array of option descriptions. */
+    int *objcPtr,		/* Number of arguments in objv. Modified to
+				 * hold # args left in objv at end. */
+    Tcl_Obj *const *objv,	/* Array of arguments to be parsed. */
+    Tcl_Obj ***remObjv)		/* Pointer to array of arguments that were not
+				 * processed here. Should be NULL if no return
+				 * of arguments is desired. */
+{
+    Tcl_Obj **leftovers;	/* Array to write back to remObjv on
+				 * successful exit. Will include the name of
+				 * the command. */
+    int nrem;			/* Size of leftovers.*/
+    register const Tcl_ArgvInfo *infoPtr;
+				/* Pointer to the current entry in the table
+				 * of argument descriptions. */
+    const Tcl_ArgvInfo *matchPtr;
+				/* Descriptor that matches current argument */
+    Tcl_Obj *curArg;		/* Current argument */
+    const char *str = NULL;
+    register char c;		/* Second character of current arg (used for
+				 * quick check for matching; use 2nd char.
+				 * because first char. will almost always be
+				 * '-'). */
+    int srcIndex;		/* Location from which to read next argument
+				 * from objv. */
+    int dstIndex;		/* Used to keep track of current arguments
+				 * being processed, primarily for error
+				 * reporting. */
+    int objc;			/* # arguments in objv still to process. */
+    int length;			/* Number of characters in current argument */
+
+    if (remObjv != NULL) {
+	/*
+	 * Then we should copy the name of the command (0th argument). The
+	 * upper bound on the number of elements is known, and (undocumented,
+	 * but historically true) there should be a NULL argument after the
+	 * last result. [Bug 3413857]
+	 */
+
+	nrem = 1;
+	leftovers = ckalloc((1 + *objcPtr) * sizeof(Tcl_Obj *));
+	leftovers[0] = objv[0];
+    } else {
+	nrem = 0;
+	leftovers = NULL;
+    }
+
+    /*
+     * OK, now start processing from the second element (1st argument).
+     */
+
+    srcIndex = dstIndex = 1;
+    objc = *objcPtr-1;
+
+    while (objc > 0) {
+	curArg = objv[srcIndex];
+	srcIndex++;
+	objc--;
+	str = TclGetStringFromObj(curArg, &length);
+	if (length > 0) {
+	    c = str[1];
+	} else {
+	    c = 0;
+	}
+
+	/*
+	 * Loop throught the argument descriptors searching for one with the
+	 * matching key string. If found, leave a pointer to it in matchPtr.
+	 */
+
+	matchPtr = NULL;
+	infoPtr = argTable;
+	for (; infoPtr != NULL && infoPtr->type != TCL_ARGV_END ; infoPtr++) {
+	    if (infoPtr->keyStr == NULL) {
+		continue;
+	    }
+	    if ((infoPtr->keyStr[1] != c)
+		    || (strncmp(infoPtr->keyStr, str, length) != 0)) {
+		continue;
+	    }
+	    if (infoPtr->keyStr[length] == 0) {
+		matchPtr = infoPtr;
+		goto gotMatch;
+	    }
+	    if (matchPtr != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"ambiguous option \"%s\"", str));
+		goto error;
+	    }
+	    matchPtr = infoPtr;
+	}
+	if (matchPtr == NULL) {
+	    /*
+	     * Unrecognized argument. Just copy it down, unless the caller
+	     * prefers an error to be registered.
+	     */
+
+	    if (remObjv == NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"unrecognized argument \"%s\"", str));
+		goto error;
+	    }
+
+	    dstIndex++;		/* This argument is now handled */
+	    leftovers[nrem++] = curArg;
+	    continue;
+	}
+
+	/*
+	 * Take the appropriate action based on the option type
+	 */
+
+    gotMatch:
+	infoPtr = matchPtr;
+	switch (infoPtr->type) {
+	case TCL_ARGV_CONSTANT:
+	    *((int *) infoPtr->dstPtr) = PTR2INT(infoPtr->srcPtr);
+	    break;
+	case TCL_ARGV_INT:
+	    if (objc == 0) {
+		goto missingArg;
+	    }
+	    if (Tcl_GetIntFromObj(interp, objv[srcIndex],
+		    (int *) infoPtr->dstPtr) == TCL_ERROR) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"expected integer argument for \"%s\" but got \"%s\"",
+			infoPtr->keyStr, Tcl_GetString(objv[srcIndex])));
+		goto error;
+	    }
+	    srcIndex++;
+	    objc--;
+	    break;
+	case TCL_ARGV_STRING:
+	    if (objc == 0) {
+		goto missingArg;
+	    }
+	    *((const char **) infoPtr->dstPtr) =
+		    Tcl_GetString(objv[srcIndex]);
+	    srcIndex++;
+	    objc--;
+	    break;
+	case TCL_ARGV_REST:
+	    /*
+	     * Only store the point where we got to if it's not to be written
+	     * to NULL, so that TCL_ARGV_AUTO_REST works.
+	     */
+
+	    if (infoPtr->dstPtr != NULL) {
+		*((int *) infoPtr->dstPtr) = dstIndex;
+	    }
+	    goto argsDone;
+	case TCL_ARGV_FLOAT:
+	    if (objc == 0) {
+		goto missingArg;
+	    }
+	    if (Tcl_GetDoubleFromObj(interp, objv[srcIndex],
+		    (double *) infoPtr->dstPtr) == TCL_ERROR) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"expected floating-point argument for \"%s\" but got \"%s\"",
+			infoPtr->keyStr, Tcl_GetString(objv[srcIndex])));
+		goto error;
+	    }
+	    srcIndex++;
+	    objc--;
+	    break;
+	case TCL_ARGV_FUNC: {
+	    Tcl_ArgvFuncProc *handlerProc = (Tcl_ArgvFuncProc *)
+		    infoPtr->srcPtr;
+	    Tcl_Obj *argObj;
+
+	    if (objc == 0) {
+		argObj = NULL;
+	    } else {
+		argObj = objv[srcIndex];
+	    }
+	    if (handlerProc(infoPtr->clientData, argObj, infoPtr->dstPtr)) {
+		srcIndex++;
+		objc--;
+	    }
+	    break;
+	}
+	case TCL_ARGV_GENFUNC: {
+	    Tcl_ArgvGenFuncProc *handlerProc = (Tcl_ArgvGenFuncProc *)
+		    infoPtr->srcPtr;
+
+	    objc = handlerProc(infoPtr->clientData, interp, objc,
+		    &objv[srcIndex], infoPtr->dstPtr);
+	    if (objc < 0) {
+		goto error;
+	    }
+	    break;
+	}
+	case TCL_ARGV_HELP:
+	    PrintUsage(interp, argTable);
+	    goto error;
+	default:
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad argument type %d in Tcl_ArgvInfo", infoPtr->type));
+	    goto error;
+	}
+    }
+
+    /*
+     * If we broke out of the loop because of an OPT_REST argument, copy the
+     * remaining arguments down. Note that there is always at least one
+     * argument left over - the command name - so we always have a result if
+     * our caller is willing to receive it. [Bug 3413857]
+     */
+
+  argsDone:
+    if (remObjv == NULL) {
+	/*
+	 * Nothing to do.
+	 */
+
+	return TCL_OK;
+    }
+
+    if (objc > 0) {
+	memcpy(leftovers+nrem, objv+srcIndex, objc*sizeof(Tcl_Obj *));
+	nrem += objc;
+    }
+    leftovers[nrem] = NULL;
+    *objcPtr = nrem++;
+    *remObjv = ckrealloc(leftovers, nrem * sizeof(Tcl_Obj *));
+    return TCL_OK;
+
+    /*
+     * Make sure to handle freeing any temporary space we've allocated on the
+     * way to an error.
+     */
+
+  missingArg:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "\"%s\" option requires an additional argument", str));
+  error:
+    if (leftovers != NULL) {
+	ckfree(leftovers);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PrintUsage --
+ *
+ *	Generate a help string describing command-line options.
+ *
+ * Results:
+ *	The interp's result will be modified to hold a help string describing
+ *	all the options in argTable.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PrintUsage(
+    Tcl_Interp *interp,		/* Place information in this interp's result
+				 * area. */
+    const Tcl_ArgvInfo *argTable)
+				/* Array of command-specific argument
+				 * descriptions. */
+{
+    register const Tcl_ArgvInfo *infoPtr;
+    int width, numSpaces;
+#define NUM_SPACES 20
+    static const char spaces[] = "                    ";
+    char tmp[TCL_DOUBLE_SPACE];
+    Tcl_Obj *msg;
+
+    /*
+     * First, compute the width of the widest option key, so that we can make
+     * everything line up.
+     */
+
+    width = 4;
+    for (infoPtr = argTable; infoPtr->type != TCL_ARGV_END; infoPtr++) {
+	int length;
+
+	if (infoPtr->keyStr == NULL) {
+	    continue;
+	}
+	length = strlen(infoPtr->keyStr);
+	if (length > width) {
+	    width = length;
+	}
+    }
+
+    /*
+     * Now add the option information, with pretty-printing.
+     */
+
+    msg = Tcl_NewStringObj("Command-specific options:", -1);
+    for (infoPtr = argTable; infoPtr->type != TCL_ARGV_END; infoPtr++) {
+	if ((infoPtr->type == TCL_ARGV_HELP) && (infoPtr->keyStr == NULL)) {
+	    Tcl_AppendPrintfToObj(msg, "\n%s", infoPtr->helpStr);
+	    continue;
+	}
+	Tcl_AppendPrintfToObj(msg, "\n %s:", infoPtr->keyStr);
+	numSpaces = width + 1 - strlen(infoPtr->keyStr);
+	while (numSpaces > 0) {
+	    if (numSpaces >= NUM_SPACES) {
+		Tcl_AppendToObj(msg, spaces, NUM_SPACES);
+	    } else {
+		Tcl_AppendToObj(msg, spaces, numSpaces);
+	    }
+	    numSpaces -= NUM_SPACES;
+	}
+	Tcl_AppendToObj(msg, infoPtr->helpStr, -1);
+	switch (infoPtr->type) {
+	case TCL_ARGV_INT:
+	    Tcl_AppendPrintfToObj(msg, "\n\t\tDefault value: %d",
+		    *((int *) infoPtr->dstPtr));
+	    break;
+	case TCL_ARGV_FLOAT:
+	    Tcl_AppendPrintfToObj(msg, "\n\t\tDefault value: %g",
+		    *((double *) infoPtr->dstPtr));
+	    sprintf(tmp, "%g", *((double *) infoPtr->dstPtr));
+	    break;
+	case TCL_ARGV_STRING: {
+	    char *string = *((char **) infoPtr->dstPtr);
+
+	    if (string != NULL) {
+		Tcl_AppendPrintfToObj(msg, "\n\t\tDefault value: \"%s\"",
+			string);
+	    }
+	    break;
+	}
+	default:
+	    break;
+	}
+    }
+    Tcl_SetObjResult(interp, msg);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetCompletionCodeFromObj --
+ *
+ *	Parses Completion code Code
+ *
+ * Results:
+ *	Returns TCL_ERROR if the value is an invalid completion code.
+ *	Otherwise, returns TCL_OK, and writes the completion code to the
+ *	pointer provided.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetCompletionCodeFromObj(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Obj *value,
+    int *codePtr)		/* Argument objects. */
+{
+    static const char *const returnCodes[] = {
+	"ok", "error", "return", "break", "continue", NULL
+    };
+
+    if ((value->typePtr != &indexType)
+	    && TclGetIntFromObj(NULL, value, codePtr) == TCL_OK) {
+	return TCL_OK;
+    }
+    if (Tcl_GetIndexFromObj(NULL, value, returnCodes, NULL, TCL_EXACT,
+	    codePtr) == TCL_OK) {
+	return TCL_OK;
+    }
+
+    /*
+     * Value is not a legal completion code.
+     */
+
+    if (interp != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad completion code \"%s\": must be"
+		" ok, error, return, break, continue, or an integer",
+		TclGetString(value)));
+	Tcl_SetErrorCode(interp, "TCL", "RESULT", "ILLEGAL_CODE", NULL);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclInt.decls b/generic/tclInt.decls
new file mode 100644
index 0000000..dea698c
--- /dev/null
+++ b/generic/tclInt.decls
@@ -0,0 +1,1287 @@
+# tclInt.decls --
+#
+#	This file contains the declarations for all unsupported
+#	functions that are exported by the Tcl library.  This file
+#	is used to generate the tclIntDecls.h, tclIntPlatDecls.h
+#	and tclStubInit.c files
+#
+# Copyright (c) 1998-1999 by Scriptics Corporation.
+# Copyright (c) 2001 by Kevin B. Kenny.  All rights reserved.
+# Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+#
+# See the file "license.terms" for information on usage and redistribution
+# of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+
+library tcl
+
+# Define the unsupported generic interfaces.
+
+interface tclInt
+
+# Declare each of the functions in the unsupported internal Tcl
+# interface.  These interfaces are allowed to changed between versions.
+# Use at your own risk.  Note that the position of functions should not
+# be changed between versions to avoid gratuitous incompatibilities.
+
+# Replaced by Tcl_FSAccess in 8.4:
+#declare 0 {
+#    int TclAccess(const char *path, int mode)
+#}
+#declare 1 {
+#    int TclAccessDeleteProc(TclAccessProc_ *proc)
+#}
+#declare 2 {
+#    int TclAccessInsertProc(TclAccessProc_ *proc)
+#}
+declare 3 {
+    void TclAllocateFreeObjects(void)
+}
+# Replaced by TclpChdir in 8.1:
+#  declare 4 {
+#      int TclChdir(Tcl_Interp *interp, char *dirName)
+#  }
+declare 5 {
+    int TclCleanupChildren(Tcl_Interp *interp, int numPids, Tcl_Pid *pidPtr,
+	    Tcl_Channel errorChan)
+}
+declare 6 {
+    void TclCleanupCommand(Command *cmdPtr)
+}
+declare 7 {
+    int TclCopyAndCollapse(int count, const char *src, char *dst)
+}
+declare 8 {
+    int TclCopyChannelOld(Tcl_Interp *interp, Tcl_Channel inChan,
+	    Tcl_Channel outChan, int toRead, Tcl_Obj *cmdPtr)
+}
+
+# TclCreatePipeline unofficially exported for use by BLT.
+
+declare 9 {
+    int TclCreatePipeline(Tcl_Interp *interp, int argc, const char **argv,
+	    Tcl_Pid **pidArrayPtr, TclFile *inPipePtr, TclFile *outPipePtr,
+	    TclFile *errFilePtr)
+}
+declare 10 {
+    int TclCreateProc(Tcl_Interp *interp, Namespace *nsPtr,
+	    const char *procName,
+	    Tcl_Obj *argsPtr, Tcl_Obj *bodyPtr, Proc **procPtrPtr)
+}
+declare 11 {
+    void TclDeleteCompiledLocalVars(Interp *iPtr, CallFrame *framePtr)
+}
+declare 12 {
+    void TclDeleteVars(Interp *iPtr, TclVarHashTable *tablePtr)
+}
+# Removed in 8.5
+#declare 13 {
+#    int TclDoGlob(Tcl_Interp *interp, char *separators,
+#	    Tcl_DString *headPtr, char *tail, Tcl_GlobTypeData *types)
+#}
+declare 14 {
+    int TclDumpMemoryInfo(ClientData clientData, int flags)
+}
+# Removed in 8.1:
+#  declare 15 {
+#      void TclExpandParseValue(ParseValue *pvPtr, int needed)
+#  }
+declare 16 {
+    void TclExprFloatError(Tcl_Interp *interp, double value)
+}
+# Removed in 8.4
+#declare 17 {
+#    int TclFileAttrsCmd(Tcl_Interp *interp, int objc, Tcl_Obj *const objv[])
+#}
+#declare 18 {
+#    int TclFileCopyCmd(Tcl_Interp *interp, int argc, char **argv)
+#}
+#declare 19 {
+#    int TclFileDeleteCmd(Tcl_Interp *interp, int argc, char **argv)
+#}
+#declare 20 {
+#    int TclFileMakeDirsCmd(Tcl_Interp *interp, int argc, char **argv)
+#}
+#declare 21 {
+#    int TclFileRenameCmd(Tcl_Interp *interp, int argc, char **argv)
+#}
+declare 22 {
+    int TclFindElement(Tcl_Interp *interp, const char *listStr,
+	    int listLength, const char **elementPtr, const char **nextPtr,
+	    int *sizePtr, int *bracePtr)
+}
+declare 23 {
+    Proc *TclFindProc(Interp *iPtr, const char *procName)
+}
+# Replaced with macro (see tclInt.h) in Tcl 8.5.0, restored in 8.5.10
+declare 24 {
+    int TclFormatInt(char *buffer, long n)
+}
+declare 25 {
+    void TclFreePackageInfo(Interp *iPtr)
+}
+# Removed in 8.1:
+#  declare 26 {
+#      char *TclGetCwd(Tcl_Interp *interp)
+#  }
+# Removed in 8.5
+#declare 27 {
+#    int TclGetDate(char *p, unsigned long now, long zone,
+#	    unsigned long *timePtr)
+#}
+declare 28 {
+    Tcl_Channel TclpGetDefaultStdChannel(int type)
+}
+# Removed in 8.4b2:
+#declare 29 {
+#    Tcl_Obj *TclGetElementOfIndexedArray(Tcl_Interp *interp,
+#	    int localIndex, Tcl_Obj *elemPtr, int flags)
+#}
+# Replaced by char *TclGetEnv(const char *name, Tcl_DString *valuePtr) in 8.1:
+#  declare 30 {
+#      char *TclGetEnv(const char *name)
+#  }
+declare 31 {
+    const char *TclGetExtension(const char *name)
+}
+declare 32 {
+    int TclGetFrame(Tcl_Interp *interp, const char *str,
+	    CallFrame **framePtrPtr)
+}
+# Removed in Tcl 8.5
+#declare 33 {
+#    TclCmdProcType TclGetInterpProc(void)
+#}
+declare 34 {
+    int TclGetIntForIndex(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    int endValue, int *indexPtr)
+}
+# Removed in 8.4b2:
+#declare 35 {
+#    Tcl_Obj *TclGetIndexedScalar(Tcl_Interp *interp, int localIndex,
+#	    int flags)
+#}
+# Removed in 8.6a2
+#declare 36 {
+#    int TclGetLong(Tcl_Interp *interp, const char *str, long *longPtr)
+#}
+declare 37 {
+    int TclGetLoadedPackages(Tcl_Interp *interp, const char *targetName)
+}
+declare 38 {
+    int TclGetNamespaceForQualName(Tcl_Interp *interp, const char *qualName,
+	    Namespace *cxtNsPtr, int flags, Namespace **nsPtrPtr,
+	    Namespace **altNsPtrPtr, Namespace **actualCxtPtrPtr,
+	    const char **simpleNamePtr)
+}
+declare 39 {
+    TclObjCmdProcType TclGetObjInterpProc(void)
+}
+declare 40 {
+    int TclGetOpenMode(Tcl_Interp *interp, const char *str, int *seekFlagPtr)
+}
+declare 41 {
+    Tcl_Command TclGetOriginalCommand(Tcl_Command command)
+}
+declare 42 {
+    CONST86 char *TclpGetUserHome(const char *name, Tcl_DString *bufferPtr)
+}
+# Removed in Tcl 8.5a2
+#declare 43 {
+#    int TclGlobalInvoke(Tcl_Interp *interp, int argc, CONST84 char **argv,
+#	    int flags)
+#}
+declare 44 {
+    int TclGuessPackageName(const char *fileName, Tcl_DString *bufPtr)
+}
+declare 45 {
+    int TclHideUnsafeCommands(Tcl_Interp *interp)
+}
+declare 46 {
+    int TclInExit(void)
+}
+# Removed in 8.4b2:
+#declare 47 {
+#    Tcl_Obj *TclIncrElementOfIndexedArray(Tcl_Interp *interp,
+#	    int localIndex, Tcl_Obj *elemPtr, long incrAmount)
+#}
+# Removed in 8.4b2:
+#declare 48 {
+#    Tcl_Obj *TclIncrIndexedScalar(Tcl_Interp *interp, int localIndex,
+#	    long incrAmount)
+#}
+#declare 49 {
+#    Tcl_Obj *TclIncrVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+#	    Tcl_Obj *part2Ptr, long incrAmount, int part1NotParsed)
+#}
+declare 50 {
+    void TclInitCompiledLocals(Tcl_Interp *interp, CallFrame *framePtr,
+	    Namespace *nsPtr)
+}
+declare 51 {
+    int TclInterpInit(Tcl_Interp *interp)
+}
+# Removed in Tcl 8.5a2
+#declare 52 {
+#    int TclInvoke(Tcl_Interp *interp, int argc, CONST84 char **argv,
+#	    int flags)
+#}
+declare 53 {
+    int TclInvokeObjectCommand(ClientData clientData, Tcl_Interp *interp,
+	    int argc, CONST84 char **argv)
+}
+declare 54 {
+    int TclInvokeStringCommand(ClientData clientData, Tcl_Interp *interp,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 55 {
+    Proc *TclIsProc(Command *cmdPtr)
+}
+# Replaced with TclpLoadFile in 8.1:
+#  declare 56 {
+#      int TclLoadFile(Tcl_Interp *interp, char *fileName, char *sym1,
+#  	    char *sym2, Tcl_PackageInitProc **proc1Ptr,
+#  	    Tcl_PackageInitProc **proc2Ptr)
+#  }
+# Signature changed to take a length in 8.1:
+#  declare 57 {
+#      int TclLooksLikeInt(char *p)
+#  }
+declare 58 {
+    Var *TclLookupVar(Tcl_Interp *interp, const char *part1, const char *part2,
+	    int flags, const char *msg, int createPart1, int createPart2,
+	    Var **arrayPtrPtr)
+}
+# Replaced by Tcl_FSMatchInDirectory in 8.4
+#declare 59 {
+#    int TclpMatchFiles(Tcl_Interp *interp, char *separators,
+#	    Tcl_DString *dirPtr, char *pattern, char *tail)
+#}
+declare 60 {
+    int TclNeedSpace(const char *start, const char *end)
+}
+declare 61 {
+    Tcl_Obj *TclNewProcBodyObj(Proc *procPtr)
+}
+declare 62 {
+    int TclObjCommandComplete(Tcl_Obj *cmdPtr)
+}
+declare 63 {
+    int TclObjInterpProc(ClientData clientData, Tcl_Interp *interp,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 64 {
+    int TclObjInvoke(Tcl_Interp *interp, int objc, Tcl_Obj *const objv[],
+	    int flags)
+}
+# Removed in Tcl 8.5a2
+#declare 65 {
+#    int TclObjInvokeGlobal(Tcl_Interp *interp, int objc,
+#	    Tcl_Obj *const objv[], int flags)
+#}
+#declare 66 {
+#    int TclOpenFileChannelDeleteProc(TclOpenFileChannelProc_ *proc)
+#}
+#declare 67 {
+#    int TclOpenFileChannelInsertProc(TclOpenFileChannelProc_ *proc)
+#}
+# Replaced by Tcl_FSAccess in 8.4:
+#declare 68 {
+#    int TclpAccess(const char *path, int mode)
+#}
+declare 69 {
+    char *TclpAlloc(unsigned int size)
+}
+#declare 70 {
+#    int TclpCopyFile(const char *source, const char *dest)
+#}
+#declare 71 {
+#    int TclpCopyDirectory(const char *source, const char *dest,
+#	    Tcl_DString *errorPtr)
+#}
+#declare 72 {
+#    int TclpCreateDirectory(const char *path)
+#}
+#declare 73 {
+#    int TclpDeleteFile(const char *path)
+#}
+declare 74 {
+    void TclpFree(char *ptr)
+}
+declare 75 {
+    unsigned long TclpGetClicks(void)
+}
+declare 76 {
+    unsigned long TclpGetSeconds(void)
+}
+
+# deprecated
+declare 77 {
+    void TclpGetTime(Tcl_Time *time)
+}
+# Removed in 8.6:
+#declare 78 {
+#    int TclpGetTimeZone(unsigned long time)
+#}
+# Replaced by Tcl_FSListVolumes in 8.4:
+#declare 79 {
+#    int TclpListVolumes(Tcl_Interp *interp)
+#}
+# Replaced by Tcl_FSOpenFileChannel in 8.4:
+#declare 80 {
+#    Tcl_Channel TclpOpenFileChannel(Tcl_Interp *interp, char *fileName,
+#	    char *modeString, int permissions)
+#}
+declare 81 {
+    char *TclpRealloc(char *ptr, unsigned int size)
+}
+#declare 82 {
+#    int TclpRemoveDirectory(const char *path, int recursive,
+#	    Tcl_DString *errorPtr)
+#}
+#declare 83 {
+#    int TclpRenameFile(const char *source, const char *dest)
+#}
+# Removed in 8.1:
+#  declare 84 {
+#      int TclParseBraces(Tcl_Interp *interp, char *str, char **termPtr,
+#  	    ParseValue *pvPtr)
+#  }
+#  declare 85 {
+#      int TclParseNestedCmd(Tcl_Interp *interp, char *str, int flags,
+#  	    char **termPtr, ParseValue *pvPtr)
+#  }
+#  declare 86 {
+#      int TclParseQuotes(Tcl_Interp *interp, char *str, int termChar,
+#  	    int flags, char **termPtr, ParseValue *pvPtr)
+#  }
+#  declare 87 {
+#      void TclPlatformInit(Tcl_Interp *interp)
+#  }
+declare 88 {
+    char *TclPrecTraceProc(ClientData clientData, Tcl_Interp *interp,
+	    const char *name1, const char *name2, int flags)
+}
+declare 89 {
+    int TclPreventAliasLoop(Tcl_Interp *interp, Tcl_Interp *cmdInterp,
+	    Tcl_Command cmd)
+}
+# Removed in 8.1 (only available if compiled with TCL_COMPILE_DEBUG):
+#  declare 90 {
+#      void TclPrintByteCodeObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+#  }
+declare 91 {
+    void TclProcCleanupProc(Proc *procPtr)
+}
+declare 92 {
+    int TclProcCompileProc(Tcl_Interp *interp, Proc *procPtr,
+	    Tcl_Obj *bodyPtr, Namespace *nsPtr, const char *description,
+	    const char *procName)
+}
+declare 93 {
+    void TclProcDeleteProc(ClientData clientData)
+}
+# Removed in Tcl 8.5:
+#declare 94 {
+#    int TclProcInterpProc(ClientData clientData, Tcl_Interp *interp,
+#	    int argc, const char **argv)
+#}
+# Replaced by Tcl_FSStat in 8.4:
+#declare 95 {
+#    int TclpStat(const char *path, Tcl_StatBuf *buf)
+#}
+declare 96 {
+    int TclRenameCommand(Tcl_Interp *interp, const char *oldName,
+            const char *newName)
+}
+declare 97 {
+    void TclResetShadowedCmdRefs(Tcl_Interp *interp, Command *newCmdPtr)
+}
+declare 98 {
+    int TclServiceIdle(void)
+}
+# Removed in 8.4b2:
+#declare 99 {
+#    Tcl_Obj *TclSetElementOfIndexedArray(Tcl_Interp *interp, int localIndex,
+#	    Tcl_Obj *elemPtr, Tcl_Obj *objPtr, int flags)
+#}
+# Removed in 8.4b2:
+#declare 100 {
+#    Tcl_Obj *TclSetIndexedScalar(Tcl_Interp *interp, int localIndex,
+#	    Tcl_Obj *objPtr, int flags)
+#}
+declare 101 {
+    CONST86 char *TclSetPreInitScript(const char *string)
+}
+declare 102 {
+    void TclSetupEnv(Tcl_Interp *interp)
+}
+declare 103 {
+    int TclSockGetPort(Tcl_Interp *interp, const char *str, const char *proto,
+	    int *portPtr)
+}
+declare 104 {
+    int TclSockMinimumBuffersOld(int sock, int size)
+}
+# Replaced by Tcl_FSStat in 8.4:
+#declare 105 {
+#    int TclStat(const char *path, Tcl_StatBuf *buf)
+#}
+#declare 106 {
+#    int TclStatDeleteProc(TclStatProc_ *proc)
+#}
+#declare 107 {
+#    int TclStatInsertProc(TclStatProc_ *proc)
+#}
+declare 108 {
+    void TclTeardownNamespace(Namespace *nsPtr)
+}
+declare 109 {
+    int TclUpdateReturnInfo(Interp *iPtr)
+}
+declare 110 {
+    int TclSockMinimumBuffers(void *sock, int size)
+}
+# Removed in 8.1:
+#  declare 110 {
+#      char *TclWordEnd(char *start, char *lastChar, int nested, int *semiPtr)
+#  }
+
+# Procedures used in conjunction with Tcl namespaces. They are
+# defined here instead of in tcl.decls since they are not stable yet.
+
+declare 111 {
+    void Tcl_AddInterpResolvers(Tcl_Interp *interp, const char *name,
+	    Tcl_ResolveCmdProc *cmdProc, Tcl_ResolveVarProc *varProc,
+	    Tcl_ResolveCompiledVarProc *compiledVarProc)
+}
+declare 112 {
+    int Tcl_AppendExportList(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    Tcl_Obj *objPtr)
+}
+declare 113 {
+    Tcl_Namespace *Tcl_CreateNamespace(Tcl_Interp *interp, const char *name,
+	    ClientData clientData, Tcl_NamespaceDeleteProc *deleteProc)
+}
+declare 114 {
+    void Tcl_DeleteNamespace(Tcl_Namespace *nsPtr)
+}
+declare 115 {
+    int Tcl_Export(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern, int resetListFirst)
+}
+declare 116 {
+    Tcl_Command Tcl_FindCommand(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *contextNsPtr, int flags)
+}
+declare 117 {
+    Tcl_Namespace *Tcl_FindNamespace(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *contextNsPtr, int flags)
+}
+declare 118 {
+    int Tcl_GetInterpResolvers(Tcl_Interp *interp, const char *name,
+	    Tcl_ResolverInfo *resInfo)
+}
+declare 119 {
+    int Tcl_GetNamespaceResolvers(Tcl_Namespace *namespacePtr,
+	    Tcl_ResolverInfo *resInfo)
+}
+declare 120 {
+    Tcl_Var Tcl_FindNamespaceVar(Tcl_Interp *interp, const char *name,
+	    Tcl_Namespace *contextNsPtr, int flags)
+}
+declare 121 {
+    int Tcl_ForgetImport(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern)
+}
+declare 122 {
+    Tcl_Command Tcl_GetCommandFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 123 {
+    void Tcl_GetCommandFullName(Tcl_Interp *interp, Tcl_Command command,
+	    Tcl_Obj *objPtr)
+}
+declare 124 {
+    Tcl_Namespace *Tcl_GetCurrentNamespace(Tcl_Interp *interp)
+}
+declare 125 {
+    Tcl_Namespace *Tcl_GetGlobalNamespace(Tcl_Interp *interp)
+}
+declare 126 {
+    void Tcl_GetVariableFullName(Tcl_Interp *interp, Tcl_Var variable,
+	    Tcl_Obj *objPtr)
+}
+declare 127 {
+    int Tcl_Import(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+	    const char *pattern, int allowOverwrite)
+}
+declare 128 {
+    void Tcl_PopCallFrame(Tcl_Interp *interp)
+}
+declare 129 {
+    int Tcl_PushCallFrame(Tcl_Interp *interp, Tcl_CallFrame *framePtr,
+	    Tcl_Namespace *nsPtr, int isProcCallFrame)
+}
+declare 130 {
+    int Tcl_RemoveInterpResolvers(Tcl_Interp *interp, const char *name)
+}
+declare 131 {
+    void Tcl_SetNamespaceResolvers(Tcl_Namespace *namespacePtr,
+	    Tcl_ResolveCmdProc *cmdProc, Tcl_ResolveVarProc *varProc,
+	    Tcl_ResolveCompiledVarProc *compiledVarProc)
+}
+declare 132 {
+    int TclpHasSockets(Tcl_Interp *interp)
+}
+declare 133 {
+    struct tm *TclpGetDate(const time_t *time, int useGMT)
+}
+# Removed in 8.5
+#declare 134 {
+#    size_t TclpStrftime(char *s, size_t maxsize, const char *format,
+#	    const struct tm *t, int useGMT)
+#}
+#declare 135 {
+#    int TclpCheckStackSpace(void)
+#}
+
+# Added in 8.1:
+
+#declare 137 {
+#   int TclpChdir(const char *dirName)
+#}
+declare 138 {
+    CONST84_RETURN char *TclGetEnv(const char *name, Tcl_DString *valuePtr)
+}
+#declare 139 {
+#    int TclpLoadFile(Tcl_Interp *interp, char *fileName, char *sym1,
+#	    char *sym2, Tcl_PackageInitProc **proc1Ptr,
+#	    Tcl_PackageInitProc **proc2Ptr, ClientData *clientDataPtr)
+#}
+#declare 140 {
+#    int TclLooksLikeInt(const char *bytes, int length)
+#}
+# This is used by TclX, but should otherwise be considered private
+declare 141 {
+    CONST84_RETURN char *TclpGetCwd(Tcl_Interp *interp, Tcl_DString *cwdPtr)
+}
+declare 142 {
+    int TclSetByteCodeFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    CompileHookProc *hookProc, ClientData clientData)
+}
+declare 143 {
+    int TclAddLiteralObj(struct CompileEnv *envPtr, Tcl_Obj *objPtr,
+	    LiteralEntry **litPtrPtr)
+}
+declare 144 {
+    void TclHideLiteral(Tcl_Interp *interp, struct CompileEnv *envPtr,
+	    int index)
+}
+declare 145 {
+    const struct AuxDataType *TclGetAuxDataType(const char *typeName)
+}
+declare 146 {
+    TclHandle TclHandleCreate(void *ptr)
+}
+declare 147 {
+    void TclHandleFree(TclHandle handle)
+}
+declare 148 {
+    TclHandle TclHandlePreserve(TclHandle handle)
+}
+declare 149 {
+    void TclHandleRelease(TclHandle handle)
+}
+
+# Added for Tcl 8.2
+
+declare 150 {
+    int TclRegAbout(Tcl_Interp *interp, Tcl_RegExp re)
+}
+declare 151 {
+    void TclRegExpRangeUniChar(Tcl_RegExp re, int index, int *startPtr,
+	    int *endPtr)
+}
+declare 152 {
+    void TclSetLibraryPath(Tcl_Obj *pathPtr)
+}
+declare 153 {
+    Tcl_Obj *TclGetLibraryPath(void)
+}
+
+# moved to tclTest.c (static) in 8.3.2/8.4a2
+#declare 154 {
+#    int TclTestChannelCmd(ClientData clientData,
+#    Tcl_Interp *interp, int argc, char **argv)
+#}
+#declare 155 {
+#    int TclTestChannelEventCmd(ClientData clientData,
+#	     Tcl_Interp *interp, int argc, char **argv)
+#}
+
+declare 156 {
+    void TclRegError(Tcl_Interp *interp, const char *msg,
+	    int status)
+}
+declare 157 {
+    Var *TclVarTraceExists(Tcl_Interp *interp, const char *varName)
+}
+# REMOVED (except from stub table) - use public Tcl_SetStartupScript()
+declare 158 {
+    void TclSetStartupScriptFileName(const char *filename)
+}
+# REMOVED (except from stub table) - use public Tcl_GetStartupScript()
+declare 159 {
+    const char *TclGetStartupScriptFileName(void)
+}
+#declare 160 {
+#    int TclpMatchFilesTypes(Tcl_Interp *interp, char *separators,
+#	    Tcl_DString *dirPtr, char *pattern, char *tail,
+#	    GlobTypeData *types)
+#}
+
+# new in 8.3.2/8.4a2
+declare 161 {
+    int TclChannelTransform(Tcl_Interp *interp, Tcl_Channel chan,
+	    Tcl_Obj *cmdObjPtr)
+}
+declare 162 {
+    void TclChannelEventScriptInvoker(ClientData clientData, int flags)
+}
+
+# ALERT: The result of 'TclGetInstructionTable' is actually a
+# "const InstructionDesc*" but we do not want to describe this structure in
+# "tclInt.h". It is described in "tclCompile.h". Use a cast to the
+# correct type when calling this procedure.
+
+declare 163 {
+    const void *TclGetInstructionTable(void)
+}
+
+# ALERT: The argument of 'TclExpandCodeArray' is actually a
+# "CompileEnv*" but we do not want to describe this structure in
+# "tclInt.h". It is described in "tclCompile.h".
+
+declare 164 {
+    void TclExpandCodeArray(void *envPtr)
+}
+
+# These functions are vfs aware, but are generally only useful internally.
+declare 165 {
+    void TclpSetInitialEncodings(void)
+}
+
+# New function due to TIP #33
+declare 166 {
+    int TclListObjSetElement(Tcl_Interp *interp, Tcl_Obj *listPtr,
+	    int index, Tcl_Obj *valuePtr)
+}
+
+# VFS-aware versions of Tcl*StartupScriptFileName (158 and 159 above)
+# REMOVED (except from stub table) - use public Tcl_SetStartupScript()
+declare 167 {
+    void TclSetStartupScriptPath(Tcl_Obj *pathPtr)
+}
+# REMOVED (except from stub table) - use public Tcl_GetStartupScript()
+declare 168 {
+    Tcl_Obj *TclGetStartupScriptPath(void)
+}
+# variant of Tcl_UtfNCmp that takes n as bytes, not chars
+declare 169 {
+    int TclpUtfNcmp2(const char *s1, const char *s2, unsigned long n)
+}
+declare 170 {
+    int TclCheckInterpTraces(Tcl_Interp *interp, const char *command,
+	    int numChars, Command *cmdPtr, int result, int traceFlags,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 171 {
+    int TclCheckExecutionTraces(Tcl_Interp *interp, const char *command,
+	    int numChars, Command *cmdPtr, int result, int traceFlags,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 172 {
+    int TclInThreadExit(void)
+}
+
+# added for 8.4.2
+
+declare 173 {
+    int TclUniCharMatch(const Tcl_UniChar *string, int strLen,
+	    const Tcl_UniChar *pattern, int ptnLen, int flags)
+}
+
+# added for 8.4.3
+
+#declare 174 {
+#    Tcl_Obj *TclIncrWideVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+#	    Tcl_Obj *part2Ptr, Tcl_WideInt wideIncrAmount, int part1NotParsed)
+#}
+
+# Factoring out of trace code
+
+declare 175 {
+    int TclCallVarTraces(Interp *iPtr, Var *arrayPtr, Var *varPtr,
+	    const char *part1, const char *part2, int flags, int leaveErrMsg)
+}
+declare 176 {
+    void TclCleanupVar(Var *varPtr, Var *arrayPtr)
+}
+declare 177 {
+    void TclVarErrMsg(Tcl_Interp *interp, const char *part1, const char *part2,
+	    const char *operation, const char *reason)
+}
+# TIP 338 made these public - now declared in tcl.h too
+declare 178 {
+    void Tcl_SetStartupScript(Tcl_Obj *pathPtr, const char *encodingName)
+}
+declare 179 {
+    Tcl_Obj *Tcl_GetStartupScript(const char **encodingNamePtr)
+}
+
+# REMOVED
+# Allocate lists without copying arrays
+# declare 180 {
+#    Tcl_Obj *TclNewListObjDirect(int objc, Tcl_Obj **objv)
+# }
+#declare 181 {
+#    Tcl_Obj *TclDbNewListObjDirect(int objc, Tcl_Obj **objv,
+#	    const char *file, int line)
+#}
+
+# TclpGmtime and TclpLocaltime promoted to the generic interface from unix
+
+declare 182 {
+     struct tm *TclpLocaltime(const time_t *clock)
+}
+declare 183 {
+     struct tm *TclpGmtime(const time_t *clock)
+}
+
+# For the new "Thread Storage" subsystem.
+
+### REMOVED on grounds it should never have been exposed. All these
+### functions are now either static in tclThreadStorage.c or
+### MODULE_SCOPE.
+# declare 184 {
+#      void TclThreadStorageLockInit(void)
+# }
+# declare 185 {
+#      void TclThreadStorageLock(void)
+# }
+# declare 186 {
+#      void TclThreadStorageUnlock(void)
+# }
+# declare 187 {
+#      void TclThreadStoragePrint(FILE *outFile, int flags)
+# }
+# declare 188 {
+#      Tcl_HashTable *TclThreadStorageGetHashTable(Tcl_ThreadId id)
+# }
+# declare 189 {
+#      Tcl_HashTable *TclThreadStorageInit(Tcl_ThreadId id, void *reserved)
+# }
+# declare 190 {
+#      void TclThreadStorageDataKeyInit(Tcl_ThreadDataKey *keyPtr)
+# }
+# declare 191 {
+#      void *TclThreadStorageDataKeyGet(Tcl_ThreadDataKey *keyPtr)
+# }
+# declare 192 {
+#      void TclThreadStorageDataKeySet(Tcl_ThreadDataKey *keyPtr, void *data)
+# }
+# declare 193 {
+#      void TclFinalizeThreadStorageThread(Tcl_ThreadId id)
+# }
+# declare 194 {
+#      void TclFinalizeThreadStorage(void)
+# }
+# declare 195 {
+#      void TclFinalizeThreadStorageData(Tcl_ThreadDataKey *keyPtr)
+# }
+# declare 196 {
+#      void TclFinalizeThreadStorageDataKey(Tcl_ThreadDataKey *keyPtr)
+# }
+
+#
+# Added in tcl8.5a5 for compiler/executor experimentation.
+# Disabled in Tcl 8.5.1; experiments terminated. :/
+#
+#declare 197 {
+#    int TclCompEvalObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+#		        const CmdFrame *invoker, int word)
+#}
+declare 198 {
+    int TclObjGetFrame(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    CallFrame **framePtrPtr)
+}
+
+#declare 199 {
+#    int TclMatchIsTrivial(const char *pattern)
+#}
+
+# 200-208 exported for use by the test suite [Bug 1054748]
+declare 200 {
+    int TclpObjRemoveDirectory(Tcl_Obj *pathPtr, int recursive,
+	Tcl_Obj **errorPtr)
+}
+declare 201 {
+    int TclpObjCopyDirectory(Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr,
+	Tcl_Obj **errorPtr)
+}
+declare 202 {
+    int TclpObjCreateDirectory(Tcl_Obj *pathPtr)
+}
+declare 203 {
+    int TclpObjDeleteFile(Tcl_Obj *pathPtr)
+}
+declare 204 {
+    int TclpObjCopyFile(Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr)
+}
+declare 205 {
+    int TclpObjRenameFile(Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr)
+}
+declare 206 {
+    int TclpObjStat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf)
+}
+declare 207 {
+    int TclpObjAccess(Tcl_Obj *pathPtr, int mode)
+}
+declare 208 {
+    Tcl_Channel TclpOpenFileChannel(Tcl_Interp *interp,
+	    Tcl_Obj *pathPtr, int mode, int permissions)
+}
+# Made public by TIP 258
+#declare 209 {
+#    Tcl_Obj *TclGetEncodingSearchPath(void)
+#}
+#declare 210 {
+#    int TclSetEncodingSearchPath(Tcl_Obj *searchPath)
+#}
+#declare 211 {
+#    const char *TclpGetEncodingNameFromEnvironment(Tcl_DString *bufPtr)
+#}
+declare 212 {
+    void TclpFindExecutable(const char *argv0)
+}
+declare 213 {
+    Tcl_Obj *TclGetObjNameOfExecutable(void)
+}
+declare 214 {
+    void TclSetObjNameOfExecutable(Tcl_Obj *name, Tcl_Encoding encoding)
+}
+declare 215 {
+    void *TclStackAlloc(Tcl_Interp *interp, int numBytes)
+}
+declare 216 {
+    void TclStackFree(Tcl_Interp *interp, void *freePtr)
+}
+declare 217 {
+    int TclPushStackFrame(Tcl_Interp *interp, Tcl_CallFrame **framePtrPtr,
+            Tcl_Namespace *namespacePtr, int isProcCallFrame)
+}
+declare 218 {
+    void TclPopStackFrame(Tcl_Interp *interp)
+}
+
+# for use in tclTest.c
+declare 224 {
+    TclPlatformType *TclGetPlatform(void)
+}
+
+#
+declare 225 {
+    Tcl_Obj *TclTraceDictPath(Tcl_Interp *interp, Tcl_Obj *rootPtr,
+	    int keyc, Tcl_Obj *const keyv[], int flags)
+}
+declare 226 {
+    int TclObjBeingDeleted(Tcl_Obj *objPtr)
+}
+declare 227 {
+    void TclSetNsPath(Namespace *nsPtr, int pathLength,
+            Tcl_Namespace *pathAry[])
+}
+#  Used to be needed for TclOO-extension; unneeded now that TclOO is in the
+#  core and NRE-enabled
+#  declare 228 {
+#      int TclObjInterpProcCore(register Tcl_Interp *interp, Tcl_Obj *procNameObj,
+#             int skip, ProcErrorProc *errorProc)
+#  }
+declare 229 {
+    int	TclPtrMakeUpvar(Tcl_Interp *interp, Var *otherP1Ptr,
+	    const char *myName, int myFlags, int index)
+}
+declare 230 {
+    Var *TclObjLookupVar(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+	    const char *part2, int flags, const char *msg,
+	    const int createPart1, const int createPart2, Var **arrayPtrPtr)
+}
+declare 231 {
+    int	TclGetNamespaceFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+	    Tcl_Namespace **nsPtrPtr)
+}
+
+# Bits and pieces of TIP#280's guts
+declare 232 {
+    int TclEvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags,
+	    const CmdFrame *invoker, int word)
+}
+declare 233 {
+    void TclGetSrcInfoForPc(CmdFrame *contextPtr)
+}
+
+# Exports for VarReform compat: Itcl, XOTcl like to peek into our varTables :(
+declare 234 {
+    Var *TclVarHashCreateVar(TclVarHashTable *tablePtr, const char *key,
+             int *newPtr)
+}
+declare 235 {
+    void TclInitVarHashTable(TclVarHashTable *tablePtr, Namespace *nsPtr)
+}
+
+
+# TIP 337 made this one public
+declare 236 {
+    void TclBackgroundException(Tcl_Interp *interp, int code)
+}
+
+# TIP #285: Script cancellation support.
+declare 237 {
+    int TclResetCancellation(Tcl_Interp *interp, int force)
+}
+
+# NRE functions for "rogue" extensions to exploit NRE; they will need to
+# include NRE.h too.
+declare 238 {
+    int TclNRInterpProc(ClientData clientData, Tcl_Interp *interp,
+	    int objc, Tcl_Obj *const objv[])
+}
+declare 239 {
+    int TclNRInterpProcCore(Tcl_Interp *interp, Tcl_Obj *procNameObj,
+			    int skip, ProcErrorProc *errorProc)
+}
+declare 240 {
+    int TclNRRunCallbacks(Tcl_Interp *interp, int result,
+	      struct NRE_callback *rootPtr)
+}
+declare 241 {
+    int TclNREvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr, int flags,
+	    const CmdFrame *invoker, int word)
+}
+declare 242 {
+    int TclNREvalObjv(Tcl_Interp *interp, int objc,
+	      Tcl_Obj *const objv[], int flags, Command *cmdPtr)
+}
+
+# Tcl_Obj leak detection support.
+declare 243 {
+    void TclDbDumpActiveObjects(FILE *outFile)
+}
+
+# Functions to make things better for itcl
+declare 244 {
+    Tcl_HashTable *TclGetNamespaceChildTable(Tcl_Namespace *nsPtr)
+}
+declare 245 {
+    Tcl_HashTable *TclGetNamespaceCommandTable(Tcl_Namespace *nsPtr)
+}
+declare 246 {
+    int TclInitRewriteEnsemble(Tcl_Interp *interp, int numRemoved,
+	    int numInserted, Tcl_Obj *const *objv)
+}
+declare 247 {
+    void TclResetRewriteEnsemble(Tcl_Interp *interp, int isRootEnsemble)
+}
+
+declare 248 {
+    int TclCopyChannel(Tcl_Interp *interp, Tcl_Channel inChan,
+	    Tcl_Channel outChan, Tcl_WideInt toRead, Tcl_Obj *cmdPtr)
+}
+
+declare 249 {
+    char *TclDoubleDigits(double dv, int ndigits, int flags,
+			  int *decpt, int *signum, char **endPtr)
+}
+# TIP #285: Script cancellation support.
+declare 250 {
+    void TclSetSlaveCancelFlags(Tcl_Interp *interp, int flags, int force)
+}
+
+# Allow extensions for optimization
+declare 251 {
+    int TclRegisterLiteral(void *envPtr,
+	    const char *bytes, int length, int flags)
+}
+
+# Exporting of the internal API to variables.
+
+declare 252 {
+    Tcl_Obj *TclPtrGetVar(Tcl_Interp *interp, Tcl_Var varPtr,
+	    Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr,
+	    const int flags)
+}
+declare 253 {
+    Tcl_Obj *TclPtrSetVar(Tcl_Interp *interp, Tcl_Var varPtr,
+	    Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr,
+	    Tcl_Obj *newValuePtr, const int flags)
+}
+declare 254 {
+    Tcl_Obj *TclPtrIncrObjVar(Tcl_Interp *interp, Tcl_Var varPtr,
+	    Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr,
+	    Tcl_Obj *incrPtr, const int flags)
+}
+declare 255 {
+    int	TclPtrObjMakeUpvar(Tcl_Interp *interp, Tcl_Var otherPtr,
+	    Tcl_Obj *myNamePtr, int myFlags)
+}
+declare 256 {
+    int	TclPtrUnsetVar(Tcl_Interp *interp, Tcl_Var varPtr, Tcl_Var arrayPtr,
+	    Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, const int flags)
+}
+
+##############################################################################
+
+# Define the platform specific internal Tcl interface. These functions are
+# only available on the designated platform.
+
+interface tclIntPlat
+
+################################
+# Windows specific functions
+
+declare 0 win {
+    void TclWinConvertError(DWORD errCode)
+}
+declare 1 win {
+    void TclWinConvertWSAError(DWORD errCode)
+}
+declare 2 win {
+    struct servent *TclWinGetServByName(const char *nm,
+	    const char *proto)
+}
+declare 3 win {
+    int TclWinGetSockOpt(SOCKET s, int level, int optname,
+	    char *optval, int *optlen)
+}
+declare 4 win {
+    HINSTANCE TclWinGetTclInstance(void)
+}
+# new for 8.4.20+/8.5.12+ Cygwin only
+declare 5 win {
+    int TclUnixWaitForFile(int fd, int mask, int timeout)
+}
+# Removed in 8.1:
+#  declare 5 win {
+#      HINSTANCE TclWinLoadLibrary(char *name)
+#  }
+declare 6 win {
+    unsigned short TclWinNToHS(unsigned short ns)
+}
+declare 7 win {
+    int TclWinSetSockOpt(SOCKET s, int level, int optname,
+	    const char *optval, int optlen)
+}
+declare 8 win {
+    int TclpGetPid(Tcl_Pid pid)
+}
+declare 9 win {
+    int TclWinGetPlatformId(void)
+}
+# new for 8.4.20+/8.5.12+ Cygwin only
+declare 10 win {
+    Tcl_DirEntry *TclpReaddir(DIR *dir)
+}
+# Removed in 8.3.1 (for Win32s only)
+#declare 10 win {
+#    int TclWinSynchSpawn(void *args, int type, void **trans, Tcl_Pid *pidPtr)
+#}
+
+# Pipe channel functions
+
+declare 11 win {
+    void TclGetAndDetachPids(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 12 win {
+    int TclpCloseFile(TclFile file)
+}
+declare 13 win {
+    Tcl_Channel TclpCreateCommandChannel(TclFile readFile,
+	    TclFile writeFile, TclFile errorFile, int numPids, Tcl_Pid *pidPtr)
+}
+declare 14 win {
+    int TclpCreatePipe(TclFile *readPipe, TclFile *writePipe)
+}
+declare 15 win {
+    int TclpCreateProcess(Tcl_Interp *interp, int argc,
+	    const char **argv, TclFile inputFile, TclFile outputFile,
+	    TclFile errorFile, Tcl_Pid *pidPtr)
+}
+# new for 8.4.20+/8.5.12+ Cygwin only
+declare 16 win {
+    int TclpIsAtty(int fd)
+}
+# Signature changed in 8.1:
+#  declare 16 win {
+#      TclFile TclpCreateTempFile(char *contents, Tcl_DString *namePtr)
+#  }
+#  declare 17 win {
+#      char *TclpGetTZName(void)
+#  }
+# new for 8.5.12+ Cygwin only
+declare 17 win {
+    int TclUnixCopyFile(const char *src, const char *dst,
+	    const Tcl_StatBuf *statBufPtr, int dontCopyAtts)
+}
+declare 18 win {
+    TclFile TclpMakeFile(Tcl_Channel channel, int direction)
+}
+declare 19 win {
+    TclFile TclpOpenFile(const char *fname, int mode)
+}
+declare 20 win {
+    void TclWinAddProcess(HANDLE hProcess, DWORD id)
+}
+# new for 8.4.20+/8.5.12+
+declare 21 win {
+    char *TclpInetNtoa(struct in_addr addr)
+}
+# removed permanently for 8.4
+#declare 21 win {
+#    void TclpAsyncMark(Tcl_AsyncHandler async)
+#}
+
+# Added in 8.1:
+declare 22 win {
+    TclFile TclpCreateTempFile(const char *contents)
+}
+# Removed in 8.6:
+#declare 23 win {
+#    char *TclpGetTZName(int isdst)
+#}
+declare 24 win {
+    char *TclWinNoBackslash(char *path)
+}
+# replaced by generic TclGetPlatform
+#declare 25 win {
+#    TclPlatformType *TclWinGetPlatform(void)
+#}
+declare 26 win {
+    void TclWinSetInterfaces(int wide)
+}
+
+# Added in Tcl 8.3.3 / 8.4
+
+declare 27 win {
+    void TclWinFlushDirtyChannels(void)
+}
+
+# Added in 8.4.2
+
+declare 28 win {
+    void TclWinResetInterfaces(void)
+}
+
+################################
+# Unix specific functions
+
+# Pipe channel functions
+
+declare 0 unix {
+    void TclGetAndDetachPids(Tcl_Interp *interp, Tcl_Channel chan)
+}
+declare 1 unix {
+    int TclpCloseFile(TclFile file)
+}
+declare 2 unix {
+    Tcl_Channel TclpCreateCommandChannel(TclFile readFile,
+	    TclFile writeFile, TclFile errorFile, int numPids, Tcl_Pid *pidPtr)
+}
+declare 3 unix {
+    int TclpCreatePipe(TclFile *readPipe, TclFile *writePipe)
+}
+declare 4 unix {
+    int TclpCreateProcess(Tcl_Interp *interp, int argc,
+	    const char **argv, TclFile inputFile, TclFile outputFile,
+	    TclFile errorFile, Tcl_Pid *pidPtr)
+}
+# Signature changed in 8.1:
+#  declare 5 unix {
+#      TclFile TclpCreateTempFile(char *contents, Tcl_DString *namePtr)
+#  }
+declare 6 unix {
+    TclFile TclpMakeFile(Tcl_Channel channel, int direction)
+}
+declare 7 unix {
+    TclFile TclpOpenFile(const char *fname, int mode)
+}
+declare 8 unix {
+    int TclUnixWaitForFile(int fd, int mask, int timeout)
+}
+
+# Added in 8.1:
+
+declare 9 unix {
+    TclFile TclpCreateTempFile(const char *contents)
+}
+
+# Added in 8.4:
+
+declare 10 unix {
+    Tcl_DirEntry *TclpReaddir(DIR *dir)
+}
+# Slots 11 and 12 are forwarders for functions that were promoted to
+# generic Stubs
+declare 11 unix {
+    struct tm *TclpLocaltime_unix(const time_t *clock)
+}
+declare 12 unix {
+    struct tm *TclpGmtime_unix(const time_t *clock)
+}
+declare 13 unix {
+    char *TclpInetNtoa(struct in_addr addr)
+}
+
+# Added in 8.5:
+
+declare 14 unix {
+    int TclUnixCopyFile(const char *src, const char *dst,
+	    const Tcl_StatBuf *statBufPtr, int dontCopyAtts)
+}
+
+################################
+# Mac OS X specific functions
+
+declare 15 macosx {
+    int TclMacOSXGetFileAttribute(Tcl_Interp *interp, int objIndex,
+	    Tcl_Obj *fileName, Tcl_Obj **attributePtrPtr)
+}
+declare 16 macosx {
+    int TclMacOSXSetFileAttribute(Tcl_Interp *interp, int objIndex,
+	    Tcl_Obj *fileName, Tcl_Obj *attributePtr)
+}
+declare 17 macosx {
+    int TclMacOSXCopyFileAttributes(const char *src, const char *dst,
+	    const Tcl_StatBuf *statBufPtr)
+}
+declare 18 macosx {
+    int TclMacOSXMatchType(Tcl_Interp *interp, const char *pathName,
+	    const char *fileName, Tcl_StatBuf *statBufPtr,
+	    Tcl_GlobTypeData *types)
+}
+declare 19 macosx {
+    void TclMacOSXNotifierAddRunLoopMode(const void *runLoopMode)
+}
+
+declare 29 {win unix} {
+    int TclWinCPUID(int index, int *regs)
+}
+# Added in 8.6; core of TclpOpenTemporaryFile
+declare 30 {win unix} {
+    int TclUnixOpenTemporaryFile(Tcl_Obj *dirObj, Tcl_Obj *basenameObj,
+	    Tcl_Obj *extensionObj, Tcl_Obj *resultingNameObj)
+}
+
+
+
+# Local Variables:
+# mode: tcl
+# End:
diff --git a/generic/tclInt.h b/generic/tclInt.h
new file mode 100644
index 0000000..118af85
--- /dev/null
+++ b/generic/tclInt.h
@@ -0,0 +1,4885 @@
+/*
+ * tclInt.h --
+ *
+ *	Declarations of things used internally by the Tcl interpreter.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1993-1997 Lucent Technologies.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 2001, 2002 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ * Copyright (c) 2006-2008 by Joe Mistachkin.  All rights reserved.
+ * Copyright (c) 2008 by Miguel Sofer. All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLINT
+#define _TCLINT
+
+/*
+ * Some numerics configuration options.
+ */
+
+#undef ACCEPT_NAN
+
+/*
+ * In Tcl 8.7, stop supporting special hacks for legacy Itcl 3.
+ * Itcl 4 doesn't need them. Itcl 3 can be updated to not need them
+ * using the Tcl(Init|Reset)RewriteEnsemble() routines in all Tcl 8.6+
+ * releases.  Perhaps Tcl 8.7 will add even better public interfaces
+ * supporting all the re-invocation mechanisms extensions like Itcl 3
+ * need.  As an absolute last resort, folks who must make Itcl 3 work
+ * unchanged with Tcl 8.7 can remove this line to regain the migration
+ * support.  Tcl 9 will no longer offer even that option.
+ */
+
+#define AVOID_HACKS_FOR_ITCL 1
+
+/*
+ * Common include files needed by most of the Tcl source files are included
+ * here, so that system-dependent personalizations for the include files only
+ * have to be made in once place. This results in a few extra includes, but
+ * greater modularity. The order of the three groups of #includes is
+ * important. For example, stdio.h is needed by tcl.h.
+ */
+
+#include "tclPort.h"
+
+#include <stdio.h>
+
+#include <ctype.h>
+#ifdef NO_STDLIB_H
+#   include "../compat/stdlib.h"
+#else
+#   include <stdlib.h>
+#endif
+#ifdef NO_STRING_H
+#include "../compat/string.h"
+#else
+#include <string.h>
+#endif
+#if defined(STDC_HEADERS) || defined(__STDC__) || defined(__C99__FUNC__) \
+     || defined(__cplusplus) || defined(_MSC_VER)
+#include <stddef.h>
+#else
+typedef int ptrdiff_t;
+#endif
+
+/*
+ * Ensure WORDS_BIGENDIAN is defined correctly:
+ * Needs to happen here in addition to configure to work with fat compiles on
+ * Darwin (where configure runs only once for multiple architectures).
+ */
+
+#ifdef HAVE_SYS_TYPES_H
+#    include <sys/types.h>
+#endif
+#ifdef HAVE_SYS_PARAM_H
+#    include <sys/param.h>
+#endif
+#ifdef BYTE_ORDER
+#    ifdef BIG_ENDIAN
+#	 if BYTE_ORDER == BIG_ENDIAN
+#	     undef WORDS_BIGENDIAN
+#	     define WORDS_BIGENDIAN 1
+#	 endif
+#    endif
+#    ifdef LITTLE_ENDIAN
+#	 if BYTE_ORDER == LITTLE_ENDIAN
+#	     undef WORDS_BIGENDIAN
+#	 endif
+#    endif
+#endif
+
+/*
+ * Used to tag functions that are only to be visible within the module being
+ * built and not outside it (where this is supported by the linker).
+ */
+
+#ifndef MODULE_SCOPE
+#   ifdef __cplusplus
+#	define MODULE_SCOPE extern "C"
+#   else
+#	define MODULE_SCOPE extern
+#   endif
+#endif
+
+/*
+ * Macros used to cast between pointers and integers (e.g. when storing an int
+ * in ClientData), on 64-bit architectures they avoid gcc warning about "cast
+ * to/from pointer from/to integer of different size".
+ */
+
+#if !defined(INT2PTR) && !defined(PTR2INT)
+#   if defined(HAVE_INTPTR_T) || defined(intptr_t)
+#	define INT2PTR(p) ((void *)(intptr_t)(p))
+#	define PTR2INT(p) ((int)(intptr_t)(p))
+#   else
+#	define INT2PTR(p) ((void *)(p))
+#	define PTR2INT(p) ((int)(p))
+#   endif
+#endif
+#if !defined(UINT2PTR) && !defined(PTR2UINT)
+#   if defined(HAVE_UINTPTR_T) || defined(uintptr_t)
+#	define UINT2PTR(p) ((void *)(uintptr_t)(p))
+#	define PTR2UINT(p) ((unsigned int)(uintptr_t)(p))
+#   else
+#	define UINT2PTR(p) ((void *)(p))
+#	define PTR2UINT(p) ((unsigned int)(p))
+#   endif
+#endif
+
+#if defined(_WIN32) && defined(_MSC_VER)
+#   define vsnprintf _vsnprintf
+#endif
+
+/*
+ * The following procedures allow namespaces to be customized to support
+ * special name resolution rules for commands/variables.
+ */
+
+struct Tcl_ResolvedVarInfo;
+
+typedef Tcl_Var (Tcl_ResolveRuntimeVarProc)(Tcl_Interp *interp,
+	struct Tcl_ResolvedVarInfo *vinfoPtr);
+
+typedef void (Tcl_ResolveVarDeleteProc)(struct Tcl_ResolvedVarInfo *vinfoPtr);
+
+/*
+ * The following structure encapsulates the routines needed to resolve a
+ * variable reference at runtime. Any variable specific state will typically
+ * be appended to this structure.
+ */
+
+typedef struct Tcl_ResolvedVarInfo {
+    Tcl_ResolveRuntimeVarProc *fetchProc;
+    Tcl_ResolveVarDeleteProc *deleteProc;
+} Tcl_ResolvedVarInfo;
+
+typedef int (Tcl_ResolveCompiledVarProc)(Tcl_Interp *interp,
+	CONST84 char *name, int length, Tcl_Namespace *context,
+	Tcl_ResolvedVarInfo **rPtr);
+
+typedef int (Tcl_ResolveVarProc)(Tcl_Interp *interp, CONST84 char *name,
+	Tcl_Namespace *context, int flags, Tcl_Var *rPtr);
+
+typedef int (Tcl_ResolveCmdProc)(Tcl_Interp *interp, CONST84 char *name,
+	Tcl_Namespace *context, int flags, Tcl_Command *rPtr);
+
+typedef struct Tcl_ResolverInfo {
+    Tcl_ResolveCmdProc *cmdResProc;
+				/* Procedure handling command name
+				 * resolution. */
+    Tcl_ResolveVarProc *varResProc;
+				/* Procedure handling variable name resolution
+				 * for variables that can only be handled at
+				 * runtime. */
+    Tcl_ResolveCompiledVarProc *compiledVarResProc;
+				/* Procedure handling variable name resolution
+				 * at compile time. */
+} Tcl_ResolverInfo;
+
+/*
+ * This flag bit should not interfere with TCL_GLOBAL_ONLY,
+ * TCL_NAMESPACE_ONLY, or TCL_LEAVE_ERR_MSG; it signals that the variable
+ * lookup is performed for upvar (or similar) purposes, with slightly
+ * different rules:
+ *    - Bug #696893 - variable is either proc-local or in the current
+ *	namespace; never follow the second (global) resolution path
+ *    - Bug #631741 - do not use special namespace or interp resolvers
+ *
+ * It should also not collide with the (deprecated) TCL_PARSE_PART1 flag
+ * (Bug #835020)
+ */
+
+#define TCL_AVOID_RESOLVERS 0x40000
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to namespaces.
+ *----------------------------------------------------------------
+ */
+
+typedef struct Tcl_Ensemble Tcl_Ensemble;
+typedef struct NamespacePathEntry NamespacePathEntry;
+
+/*
+ * Special hashtable for variables: this is just a Tcl_HashTable with an nsPtr
+ * field added at the end: in this way variables can find their namespace
+ * without having to copy a pointer in their struct: they can access it via
+ * their hPtr->tablePtr.
+ */
+
+typedef struct TclVarHashTable {
+    Tcl_HashTable table;
+    struct Namespace *nsPtr;
+} TclVarHashTable;
+
+/*
+ * This is for itcl - it likes to search our varTables directly :(
+ */
+
+#define TclVarHashFindVar(tablePtr, key) \
+    TclVarHashCreateVar((tablePtr), (key), NULL)
+
+/*
+ * Define this to reduce the amount of space that the average namespace
+ * consumes by only allocating the table of child namespaces when necessary.
+ * Defining it breaks compatibility for Tcl extensions (e.g., itcl) which
+ * reach directly into the Namespace structure.
+ */
+
+#undef BREAK_NAMESPACE_COMPAT
+
+/*
+ * The structure below defines a namespace.
+ * Note: the first five fields must match exactly the fields in a
+ * Tcl_Namespace structure (see tcl.h). If you change one, be sure to change
+ * the other.
+ */
+
+typedef struct Namespace {
+    char *name;			/* The namespace's simple (unqualified) name.
+				 * This contains no ::'s. The name of the
+				 * global namespace is "" although "::" is an
+				 * synonym. */
+    char *fullName;		/* The namespace's fully qualified name. This
+				 * starts with ::. */
+    ClientData clientData;	/* An arbitrary value associated with this
+				 * namespace. */
+    Tcl_NamespaceDeleteProc *deleteProc;
+				/* Procedure invoked when deleting the
+				 * namespace to, e.g., free clientData. */
+    struct Namespace *parentPtr;/* Points to the namespace that contains this
+				 * one. NULL if this is the global
+				 * namespace. */
+#ifndef BREAK_NAMESPACE_COMPAT
+    Tcl_HashTable childTable;	/* Contains any child namespaces. Indexed by
+				 * strings; values have type (Namespace *). */
+#else
+    Tcl_HashTable *childTablePtr;
+				/* Contains any child namespaces. Indexed by
+				 * strings; values have type (Namespace *). If
+				 * NULL, there are no children. */
+#endif
+    size_t nsId;		/* Unique id for the namespace. */
+    Tcl_Interp *interp;		/* The interpreter containing this
+				 * namespace. */
+    int flags;			/* OR-ed combination of the namespace status
+				 * flags NS_DYING and NS_DEAD listed below. */
+    int activationCount;	/* Number of "activations" or active call
+				 * frames for this namespace that are on the
+				 * Tcl call stack. The namespace won't be
+				 * freed until activationCount becomes zero. */
+    int refCount;		/* Count of references by namespaceName
+				 * objects. The namespace can't be freed until
+				 * refCount becomes zero. */
+    Tcl_HashTable cmdTable;	/* Contains all the commands currently
+				 * registered in the namespace. Indexed by
+				 * strings; values have type (Command *).
+				 * Commands imported by Tcl_Import have
+				 * Command structures that point (via an
+				 * ImportedCmdRef structure) to the Command
+				 * structure in the source namespace's command
+				 * table. */
+    TclVarHashTable varTable;	/* Contains all the (global) variables
+				 * currently in this namespace. Indexed by
+				 * strings; values have type (Var *). */
+    char **exportArrayPtr;	/* Points to an array of string patterns
+				 * specifying which commands are exported. A
+				 * pattern may include "string match" style
+				 * wildcard characters to specify multiple
+				 * commands; however, no namespace qualifiers
+				 * are allowed. NULL if no export patterns are
+				 * registered. */
+    int numExportPatterns;	/* Number of export patterns currently
+				 * registered using "namespace export". */
+    int maxExportPatterns;	/* Mumber of export patterns for which space
+				 * is currently allocated. */
+    size_t cmdRefEpoch;		/* Incremented if a newly added command
+				 * shadows a command for which this namespace
+				 * has already cached a Command* pointer; this
+				 * causes all its cached Command* pointers to
+				 * be invalidated. */
+    size_t resolverEpoch;	/* Incremented whenever (a) the name
+				 * resolution rules change for this namespace
+				 * or (b) a newly added command shadows a
+				 * command that is compiled to bytecodes. This
+				 * invalidates all byte codes compiled in the
+				 * namespace, causing the code to be
+				 * recompiled under the new rules.*/
+    Tcl_ResolveCmdProc *cmdResProc;
+				/* If non-null, this procedure overrides the
+				 * usual command resolution mechanism in Tcl.
+				 * This procedure is invoked within
+				 * Tcl_FindCommand to resolve all command
+				 * references within the namespace. */
+    Tcl_ResolveVarProc *varResProc;
+				/* If non-null, this procedure overrides the
+				 * usual variable resolution mechanism in Tcl.
+				 * This procedure is invoked within
+				 * Tcl_FindNamespaceVar to resolve all
+				 * variable references within the namespace at
+				 * runtime. */
+    Tcl_ResolveCompiledVarProc *compiledVarResProc;
+				/* If non-null, this procedure overrides the
+				 * usual variable resolution mechanism in Tcl.
+				 * This procedure is invoked within
+				 * LookupCompiledLocal to resolve variable
+				 * references within the namespace at compile
+				 * time. */
+    size_t exportLookupEpoch;	/* Incremented whenever a command is added to
+				 * a namespace, removed from a namespace or
+				 * the exports of a namespace are changed.
+				 * Allows TIP#112-driven command lists to be
+				 * validated efficiently. */
+    Tcl_Ensemble *ensembles;	/* List of structures that contain the details
+				 * of the ensembles that are implemented on
+				 * top of this namespace. */
+    Tcl_Obj *unknownHandlerPtr;	/* A script fragment to be used when command
+				 * resolution in this namespace fails. TIP
+				 * 181. */
+    int commandPathLength;	/* The length of the explicit path. */
+    NamespacePathEntry *commandPathArray;
+				/* The explicit path of the namespace as an
+				 * array. */
+    NamespacePathEntry *commandPathSourceList;
+				/* Linked list of path entries that point to
+				 * this namespace. */
+    Tcl_NamespaceDeleteProc *earlyDeleteProc;
+				/* Just like the deleteProc field (and called
+				 * with the same clientData) but called at the
+				 * start of the deletion process, so there is
+				 * a chance for code to do stuff inside the
+				 * namespace before deletion completes. */
+} Namespace;
+
+/*
+ * An entry on a namespace's command resolution path.
+ */
+
+struct NamespacePathEntry {
+    Namespace *nsPtr;		/* What does this path entry point to? If it
+				 * is NULL, this path entry points is
+				 * redundant and should be skipped. */
+    Namespace *creatorNsPtr;	/* Where does this path entry point from? This
+				 * allows for efficient invalidation of
+				 * references when the path entry's target
+				 * updates its current list of defined
+				 * commands. */
+    NamespacePathEntry *prevPtr, *nextPtr;
+				/* Linked list pointers or NULL at either end
+				 * of the list that hangs off Namespace's
+				 * commandPathSourceList field. */
+};
+
+/*
+ * Flags used to represent the status of a namespace:
+ *
+ * NS_DYING -	1 means Tcl_DeleteNamespace has been called to delete the
+ *		namespace but there are still active call frames on the Tcl
+ *		stack that refer to the namespace. When the last call frame
+ *		referring to it has been popped, it's variables and command
+ *		will be destroyed and it will be marked "dead" (NS_DEAD). The
+ *		namespace can no longer be looked up by name.
+ * NS_DEAD -	1 means Tcl_DeleteNamespace has been called to delete the
+ *		namespace and no call frames still refer to it. Its variables
+ *		and command have already been destroyed. This bit allows the
+ *		namespace resolution code to recognize that the namespace is
+ *		"deleted". When the last namespaceName object in any byte code
+ *		unit that refers to the namespace has been freed (i.e., when
+ *		the namespace's refCount is 0), the namespace's storage will
+ *		be freed.
+ * NS_KILLED -	1 means that TclTeardownNamespace has already been called on
+ *		this namespace and it should not be called again [Bug 1355942]
+ * NS_SUPPRESS_COMPILATION -
+ *		Marks the commands in this namespace for not being compiled,
+ *		forcing them to be looked up every time.
+ */
+
+#define NS_DYING	0x01
+#define NS_DEAD		0x02
+#define NS_KILLED	0x04
+#define NS_SUPPRESS_COMPILATION	0x08
+
+/*
+ * Flags passed to TclGetNamespaceForQualName:
+ *
+ * TCL_GLOBAL_ONLY		- (see tcl.h) Look only in the global ns.
+ * TCL_NAMESPACE_ONLY		- (see tcl.h) Look only in the context ns.
+ * TCL_CREATE_NS_IF_UNKNOWN	- Create unknown namespaces.
+ * TCL_FIND_ONLY_NS		- The name sought is a namespace name.
+ */
+
+#define TCL_CREATE_NS_IF_UNKNOWN	0x800
+#define TCL_FIND_ONLY_NS		0x1000
+
+/*
+ * The client data for an ensemble command. This consists of the table of
+ * commands that are actually exported by the namespace, and an epoch counter
+ * that, combined with the exportLookupEpoch field of the namespace structure,
+ * defines whether the table contains valid data or will need to be recomputed
+ * next time the ensemble command is called.
+ */
+
+typedef struct EnsembleConfig {
+    Namespace *nsPtr;		/* The namspace backing this ensemble up. */
+    Tcl_Command token;		/* The token for the command that provides
+				 * ensemble support for the namespace, or NULL
+				 * if the command has been deleted (or never
+				 * existed; the global namespace never has an
+				 * ensemble command.) */
+    size_t epoch;		/* The epoch at which this ensemble's table of
+				 * exported commands is valid. */
+    char **subcommandArrayPtr;	/* Array of ensemble subcommand names. At all
+				 * consistent points, this will have the same
+				 * number of entries as there are entries in
+				 * the subcommandTable hash. */
+    Tcl_HashTable subcommandTable;
+				/* Hash table of ensemble subcommand names,
+				 * which are its keys so this also provides
+				 * the storage management for those subcommand
+				 * names. The contents of the entry values are
+				 * object version the prefix lists to use when
+				 * substituting for the command/subcommand to
+				 * build the ensemble implementation command.
+				 * Has to be stored here as well as in
+				 * subcommandDict because that field is NULL
+				 * when we are deriving the ensemble from the
+				 * namespace exports list. FUTURE WORK: use
+				 * object hash table here. */
+    struct EnsembleConfig *next;/* The next ensemble in the linked list of
+				 * ensembles associated with a namespace. If
+				 * this field points to this ensemble, the
+				 * structure has already been unlinked from
+				 * all lists, and cannot be found by scanning
+				 * the list from the namespace's ensemble
+				 * field. */
+    int flags;			/* ORed combo of TCL_ENSEMBLE_PREFIX,
+				 * ENSEMBLE_DEAD and ENSEMBLE_COMPILE. */
+
+    /* OBJECT FIELDS FOR ENSEMBLE CONFIGURATION */
+
+    Tcl_Obj *subcommandDict;	/* Dictionary providing mapping from
+				 * subcommands to their implementing command
+				 * prefixes, or NULL if we are to build the
+				 * map automatically from the namespace
+				 * exports. */
+    Tcl_Obj *subcmdList;	/* List of commands that this ensemble
+				 * actually provides, and whose implementation
+				 * will be built using the subcommandDict (if
+				 * present and defined) and by simple mapping
+				 * to the namespace otherwise. If NULL,
+				 * indicates that we are using the (dynamic)
+				 * list of currently exported commands. */
+    Tcl_Obj *unknownHandler;	/* Script prefix used to handle the case when
+				 * no match is found (according to the rule
+				 * defined by flag bit TCL_ENSEMBLE_PREFIX) or
+				 * NULL to use the default error-generating
+				 * behaviour. The script execution gets all
+				 * the arguments to the ensemble command
+				 * (including objv[0]) and will have the
+				 * results passed directly back to the caller
+				 * (including the error code) unless the code
+				 * is TCL_CONTINUE in which case the
+				 * subcommand will be reparsed by the ensemble
+				 * core, presumably because the ensemble
+				 * itself has been updated. */
+    Tcl_Obj *parameterList;	/* List of ensemble parameter names. */
+    int numParameters;		/* Cached number of parameters. This is either
+				 * 0 (if the parameterList field is NULL) or
+				 * the length of the list in the parameterList
+				 * field. */
+} EnsembleConfig;
+
+/*
+ * Various bits for the EnsembleConfig.flags field.
+ */
+
+#define ENSEMBLE_DEAD	0x1	/* Flag value to say that the ensemble is dead
+				 * and on its way out. */
+#define ENSEMBLE_COMPILE 0x4	/* Flag to enable bytecode compilation of an
+				 * ensemble. */
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to variables. These are used primarily in tclVar.c
+ *----------------------------------------------------------------
+ */
+
+/*
+ * The following structure defines a variable trace, which is used to invoke a
+ * specific C procedure whenever certain operations are performed on a
+ * variable.
+ */
+
+typedef struct VarTrace {
+    Tcl_VarTraceProc *traceProc;/* Procedure to call when operations given by
+				 * flags are performed on variable. */
+    ClientData clientData;	/* Argument to pass to proc. */
+    int flags;			/* What events the trace procedure is
+				 * interested in: OR-ed combination of
+				 * TCL_TRACE_READS, TCL_TRACE_WRITES,
+				 * TCL_TRACE_UNSETS and TCL_TRACE_ARRAY. */
+    struct VarTrace *nextPtr;	/* Next in list of traces associated with a
+				 * particular variable. */
+} VarTrace;
+
+/*
+ * The following structure defines a command trace, which is used to invoke a
+ * specific C procedure whenever certain operations are performed on a
+ * command.
+ */
+
+typedef struct CommandTrace {
+    Tcl_CommandTraceProc *traceProc;
+				/* Procedure to call when operations given by
+				 * flags are performed on command. */
+    ClientData clientData;	/* Argument to pass to proc. */
+    int flags;			/* What events the trace procedure is
+				 * interested in: OR-ed combination of
+				 * TCL_TRACE_RENAME, TCL_TRACE_DELETE. */
+    struct CommandTrace *nextPtr;
+				/* Next in list of traces associated with a
+				 * particular command. */
+    int refCount;		/* Used to ensure this structure is not
+				 * deleted too early. Keeps track of how many
+				 * pieces of code have a pointer to this
+				 * structure. */
+} CommandTrace;
+
+/*
+ * When a command trace is active (i.e. its associated procedure is executing)
+ * one of the following structures is linked into a list associated with the
+ * command's interpreter. The information in the structure is needed in order
+ * for Tcl to behave reasonably if traces are deleted while traces are active.
+ */
+
+typedef struct ActiveCommandTrace {
+    struct Command *cmdPtr;	/* Command that's being traced. */
+    struct ActiveCommandTrace *nextPtr;
+				/* Next in list of all active command traces
+				 * for the interpreter, or NULL if no more. */
+    CommandTrace *nextTracePtr;	/* Next trace to check after current trace
+				 * procedure returns; if this trace gets
+				 * deleted, must update pointer to avoid using
+				 * free'd memory. */
+    int reverseScan;		/* Boolean set true when traces are scanning
+				 * in reverse order. */
+} ActiveCommandTrace;
+
+/*
+ * When a variable trace is active (i.e. its associated procedure is
+ * executing) one of the following structures is linked into a list associated
+ * with the variable's interpreter. The information in the structure is needed
+ * in order for Tcl to behave reasonably if traces are deleted while traces
+ * are active.
+ */
+
+typedef struct ActiveVarTrace {
+    struct Var *varPtr;		/* Variable that's being traced. */
+    struct ActiveVarTrace *nextPtr;
+				/* Next in list of all active variable traces
+				 * for the interpreter, or NULL if no more. */
+    VarTrace *nextTracePtr;	/* Next trace to check after current trace
+				 * procedure returns; if this trace gets
+				 * deleted, must update pointer to avoid using
+				 * free'd memory. */
+} ActiveVarTrace;
+
+/*
+ * The structure below defines a variable, which associates a string name with
+ * a Tcl_Obj value. These structures are kept in procedure call frames (for
+ * local variables recognized by the compiler) or in the heap (for global
+ * variables and any variable not known to the compiler). For each Var
+ * structure in the heap, a hash table entry holds the variable name and a
+ * pointer to the Var structure.
+ */
+
+typedef struct Var {
+    int flags;			/* Miscellaneous bits of information about
+				 * variable. See below for definitions. */
+    union {
+	Tcl_Obj *objPtr;	/* The variable's object value. Used for
+				 * scalar variables and array elements. */
+	TclVarHashTable *tablePtr;/* For array variables, this points to
+				 * information about the hash table used to
+				 * implement the associative array. Points to
+				 * ckalloc-ed data. */
+	struct Var *linkPtr;	/* If this is a global variable being referred
+				 * to in a procedure, or a variable created by
+				 * "upvar", this field points to the
+				 * referenced variable's Var struct. */
+    } value;
+} Var;
+
+typedef struct VarInHash {
+    Var var;
+    int refCount;		/* Counts number of active uses of this
+				 * variable: 1 for the entry in the hash
+				 * table, 1 for each additional variable whose
+				 * linkPtr points here, 1 for each nested
+				 * trace active on variable, and 1 if the
+				 * variable is a namespace variable. This
+				 * record can't be deleted until refCount
+				 * becomes 0. */
+    Tcl_HashEntry entry;	/* The hash table entry that refers to this
+				 * variable. This is used to find the name of
+				 * the variable and to delete it from its
+				 * hashtable if it is no longer needed. It
+				 * also holds the variable's name. */
+} VarInHash;
+
+/*
+ * Flag bits for variables. The first two (VAR_ARRAY and VAR_LINK) are
+ * mutually exclusive and give the "type" of the variable. If none is set,
+ * this is a scalar variable.
+ *
+ * VAR_ARRAY -			1 means this is an array variable rather than
+ *				a scalar variable or link. The "tablePtr"
+ *				field points to the array's hashtable for its
+ *				elements.
+ * VAR_LINK -			1 means this Var structure contains a pointer
+ *				to another Var structure that either has the
+ *				real value or is itself another VAR_LINK
+ *				pointer. Variables like this come about
+ *				through "upvar" and "global" commands, or
+ *				through references to variables in enclosing
+ *				namespaces.
+ *
+ * Flags that indicate the type and status of storage; none is set for
+ * compiled local variables (Var structs).
+ *
+ * VAR_IN_HASHTABLE -		1 means this variable is in a hashtable and
+ *				the Var structure is malloced. 0 if it is a
+ *				local variable that was assigned a slot in a
+ *				procedure frame by the compiler so the Var
+ *				storage is part of the call frame.
+ * VAR_DEAD_HASH		1 means that this var's entry in the hashtable
+ *				has already been deleted.
+ * VAR_ARRAY_ELEMENT -		1 means that this variable is an array
+ *				element, so it is not legal for it to be an
+ *				array itself (the VAR_ARRAY flag had better
+ *				not be set).
+ * VAR_NAMESPACE_VAR -		1 means that this variable was declared as a
+ *				namespace variable. This flag ensures it
+ *				persists until its namespace is destroyed or
+ *				until the variable is unset; it will persist
+ *				even if it has not been initialized and is
+ *				marked undefined. The variable's refCount is
+ *				incremented to reflect the "reference" from
+ *				its namespace.
+ *
+ * Flag values relating to the variable's trace and search status.
+ *
+ * VAR_TRACED_READ
+ * VAR_TRACED_WRITE
+ * VAR_TRACED_UNSET
+ * VAR_TRACED_ARRAY
+ * VAR_TRACE_ACTIVE -		1 means that trace processing is currently
+ *				underway for a read or write access, so new
+ *				read or write accesses should not cause trace
+ *				procedures to be called and the variable can't
+ *				be deleted.
+ * VAR_SEARCH_ACTIVE
+ *
+ * The following additional flags are used with the CompiledLocal type defined
+ * below:
+ *
+ * VAR_ARGUMENT -		1 means that this variable holds a procedure
+ *				argument.
+ * VAR_TEMPORARY -		1 if the local variable is an anonymous
+ *				temporary variable. Temporaries have a NULL
+ *				name.
+ * VAR_RESOLVED -		1 if name resolution has been done for this
+ *				variable.
+ * VAR_IS_ARGS			1 if this variable is the last argument and is
+ *				named "args".
+ */
+
+/*
+ * FLAGS RENUMBERED: everything breaks already, make things simpler.
+ *
+ * IMPORTANT: skip the values 0x10, 0x20, 0x40, 0x800 corresponding to
+ * TCL_TRACE_(READS/WRITES/UNSETS/ARRAY): makes code simpler in tclTrace.c
+ *
+ * Keep the flag values for VAR_ARGUMENT and VAR_TEMPORARY so that old values
+ * in precompiled scripts keep working.
+ */
+
+/* Type of value (0 is scalar) */
+#define VAR_ARRAY		0x1
+#define VAR_LINK		0x2
+
+/* Type of storage (0 is compiled local) */
+#define VAR_IN_HASHTABLE	0x4
+#define VAR_DEAD_HASH		0x8
+#define VAR_ARRAY_ELEMENT	0x1000
+#define VAR_NAMESPACE_VAR	0x80	/* KEEP OLD VALUE for Itcl */
+
+#define VAR_ALL_HASH \
+	(VAR_IN_HASHTABLE|VAR_DEAD_HASH|VAR_NAMESPACE_VAR|VAR_ARRAY_ELEMENT)
+
+/* Trace and search state. */
+
+#define VAR_TRACED_READ		0x10	/* TCL_TRACE_READS */
+#define VAR_TRACED_WRITE	0x20	/* TCL_TRACE_WRITES */
+#define VAR_TRACED_UNSET	0x40	/* TCL_TRACE_UNSETS */
+#define VAR_TRACED_ARRAY	0x800	/* TCL_TRACE_ARRAY */
+#define VAR_TRACE_ACTIVE	0x2000
+#define VAR_SEARCH_ACTIVE	0x4000
+#define VAR_ALL_TRACES \
+	(VAR_TRACED_READ|VAR_TRACED_WRITE|VAR_TRACED_ARRAY|VAR_TRACED_UNSET)
+
+/* Special handling on initialisation (only CompiledLocal). */
+#define VAR_ARGUMENT		0x100	/* KEEP OLD VALUE! See tclProc.c */
+#define VAR_TEMPORARY		0x200	/* KEEP OLD VALUE! See tclProc.c */
+#define VAR_IS_ARGS		0x400
+#define VAR_RESOLVED		0x8000
+
+/*
+ * Macros to ensure that various flag bits are set properly for variables.
+ * The ANSI C "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE void	TclSetVarScalar(Var *varPtr);
+ * MODULE_SCOPE void	TclSetVarArray(Var *varPtr);
+ * MODULE_SCOPE void	TclSetVarLink(Var *varPtr);
+ * MODULE_SCOPE void	TclSetVarArrayElement(Var *varPtr);
+ * MODULE_SCOPE void	TclSetVarUndefined(Var *varPtr);
+ * MODULE_SCOPE void	TclClearVarUndefined(Var *varPtr);
+ */
+
+#define TclSetVarScalar(varPtr) \
+    (varPtr)->flags &= ~(VAR_ARRAY|VAR_LINK)
+
+#define TclSetVarArray(varPtr) \
+    (varPtr)->flags = ((varPtr)->flags & ~VAR_LINK) | VAR_ARRAY
+
+#define TclSetVarLink(varPtr) \
+    (varPtr)->flags = ((varPtr)->flags & ~VAR_ARRAY) | VAR_LINK
+
+#define TclSetVarArrayElement(varPtr) \
+    (varPtr)->flags = ((varPtr)->flags & ~VAR_ARRAY) | VAR_ARRAY_ELEMENT
+
+#define TclSetVarUndefined(varPtr) \
+    (varPtr)->flags &= ~(VAR_ARRAY|VAR_LINK);\
+    (varPtr)->value.objPtr = NULL
+
+#define TclClearVarUndefined(varPtr)
+
+#define TclSetVarTraceActive(varPtr) \
+    (varPtr)->flags |= VAR_TRACE_ACTIVE
+
+#define TclClearVarTraceActive(varPtr) \
+    (varPtr)->flags &= ~VAR_TRACE_ACTIVE
+
+#define TclSetVarNamespaceVar(varPtr) \
+    if (!TclIsVarNamespaceVar(varPtr)) {\
+	(varPtr)->flags |= VAR_NAMESPACE_VAR;\
+	if (TclIsVarInHash(varPtr)) {\
+	    ((VarInHash *)(varPtr))->refCount++;\
+	}\
+    }
+
+#define TclClearVarNamespaceVar(varPtr) \
+    if (TclIsVarNamespaceVar(varPtr)) {\
+	(varPtr)->flags &= ~VAR_NAMESPACE_VAR;\
+	if (TclIsVarInHash(varPtr)) {\
+	    ((VarInHash *)(varPtr))->refCount--;\
+	}\
+    }
+
+/*
+ * Macros to read various flag bits of variables.
+ * The ANSI C "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE int	TclIsVarScalar(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarLink(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarArray(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarUndefined(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarArrayElement(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarTemporary(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarArgument(Var *varPtr);
+ * MODULE_SCOPE int	TclIsVarResolved(Var *varPtr);
+ */
+
+#define TclIsVarScalar(varPtr) \
+    !((varPtr)->flags & (VAR_ARRAY|VAR_LINK))
+
+#define TclIsVarLink(varPtr) \
+    ((varPtr)->flags & VAR_LINK)
+
+#define TclIsVarArray(varPtr) \
+    ((varPtr)->flags & VAR_ARRAY)
+
+#define TclIsVarUndefined(varPtr) \
+    ((varPtr)->value.objPtr == NULL)
+
+#define TclIsVarArrayElement(varPtr) \
+    ((varPtr)->flags & VAR_ARRAY_ELEMENT)
+
+#define TclIsVarNamespaceVar(varPtr) \
+    ((varPtr)->flags & VAR_NAMESPACE_VAR)
+
+#define TclIsVarTemporary(varPtr) \
+    ((varPtr)->flags & VAR_TEMPORARY)
+
+#define TclIsVarArgument(varPtr) \
+    ((varPtr)->flags & VAR_ARGUMENT)
+
+#define TclIsVarResolved(varPtr) \
+    ((varPtr)->flags & VAR_RESOLVED)
+
+#define TclIsVarTraceActive(varPtr) \
+    ((varPtr)->flags & VAR_TRACE_ACTIVE)
+
+#define TclIsVarTraced(varPtr) \
+    ((varPtr)->flags & VAR_ALL_TRACES)
+
+#define TclIsVarInHash(varPtr) \
+    ((varPtr)->flags & VAR_IN_HASHTABLE)
+
+#define TclIsVarDeadHash(varPtr) \
+    ((varPtr)->flags & VAR_DEAD_HASH)
+
+#define TclGetVarNsPtr(varPtr) \
+    (TclIsVarInHash(varPtr) \
+	? ((TclVarHashTable *) ((((VarInHash *) (varPtr))->entry.tablePtr)))->nsPtr \
+	: NULL)
+
+#define VarHashRefCount(varPtr) \
+    ((VarInHash *) (varPtr))->refCount
+
+/*
+ * Macros for direct variable access by TEBC.
+ */
+
+#define TclIsVarDirectReadable(varPtr) \
+    (   !((varPtr)->flags & (VAR_ARRAY|VAR_LINK|VAR_TRACED_READ)) \
+    &&  (varPtr)->value.objPtr)
+
+#define TclIsVarDirectWritable(varPtr) \
+    !((varPtr)->flags & (VAR_ARRAY|VAR_LINK|VAR_TRACED_WRITE|VAR_DEAD_HASH))
+
+#define TclIsVarDirectUnsettable(varPtr) \
+    !((varPtr)->flags & (VAR_ARRAY|VAR_LINK|VAR_TRACED_READ|VAR_TRACED_WRITE|VAR_TRACED_UNSET|VAR_DEAD_HASH))
+
+#define TclIsVarDirectModifyable(varPtr) \
+    (   !((varPtr)->flags & (VAR_ARRAY|VAR_LINK|VAR_TRACED_READ|VAR_TRACED_WRITE)) \
+    &&  (varPtr)->value.objPtr)
+
+#define TclIsVarDirectReadable2(varPtr, arrayPtr) \
+    (TclIsVarDirectReadable(varPtr) &&\
+	(!(arrayPtr) || !((arrayPtr)->flags & VAR_TRACED_READ)))
+
+#define TclIsVarDirectWritable2(varPtr, arrayPtr) \
+    (TclIsVarDirectWritable(varPtr) &&\
+	(!(arrayPtr) || !((arrayPtr)->flags & VAR_TRACED_WRITE)))
+
+#define TclIsVarDirectModifyable2(varPtr, arrayPtr) \
+    (TclIsVarDirectModifyable(varPtr) &&\
+	(!(arrayPtr) || !((arrayPtr)->flags & (VAR_TRACED_READ|VAR_TRACED_WRITE))))
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to procedures. These are used primarily in
+ * tclProc.c, tclCompile.c, and tclExecute.c.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * Forward declaration to prevent an error when the forward reference to
+ * Command is encountered in the Proc and ImportRef types declared below.
+ */
+
+struct Command;
+
+/*
+ * The variable-length structure below describes a local variable of a
+ * procedure that was recognized by the compiler. These variables have a name,
+ * an element in the array of compiler-assigned local variables in the
+ * procedure's call frame, and various other items of information. If the
+ * local variable is a formal argument, it may also have a default value. The
+ * compiler can't recognize local variables whose names are expressions (these
+ * names are only known at runtime when the expressions are evaluated) or
+ * local variables that are created as a result of an "upvar" or "uplevel"
+ * command. These other local variables are kept separately in a hash table in
+ * the call frame.
+ */
+
+typedef struct CompiledLocal {
+    struct CompiledLocal *nextPtr;
+				/* Next compiler-recognized local variable for
+				 * this procedure, or NULL if this is the last
+				 * local. */
+    int nameLength;		/* The number of characters in local
+				 * variable's name. Used to speed up variable
+				 * lookups. */
+    int frameIndex;		/* Index in the array of compiler-assigned
+				 * variables in the procedure call frame. */
+    int flags;			/* Flag bits for the local variable. Same as
+				 * the flags for the Var structure above,
+				 * although only VAR_ARGUMENT, VAR_TEMPORARY,
+				 * and VAR_RESOLVED make sense. */
+    Tcl_Obj *defValuePtr;	/* Pointer to the default value of an
+				 * argument, if any. NULL if not an argument
+				 * or, if an argument, no default value. */
+    Tcl_ResolvedVarInfo *resolveInfo;
+				/* Customized variable resolution info
+				 * supplied by the Tcl_ResolveCompiledVarProc
+				 * associated with a namespace. Each variable
+				 * is marked by a unique ClientData tag during
+				 * compilation, and that same tag is used to
+				 * find the variable at runtime. */
+    char name[1];		/* Name of the local variable starts here. If
+				 * the name is NULL, this will just be '\0'.
+				 * The actual size of this field will be large
+				 * enough to hold the name. MUST BE THE LAST
+				 * FIELD IN THE STRUCTURE! */
+} CompiledLocal;
+
+/*
+ * The structure below defines a command procedure, which consists of a
+ * collection of Tcl commands plus information about arguments and other local
+ * variables recognized at compile time.
+ */
+
+typedef struct Proc {
+    struct Interp *iPtr;	/* Interpreter for which this command is
+				 * defined. */
+    int refCount;		/* Reference count: 1 if still present in
+				 * command table plus 1 for each call to the
+				 * procedure that is currently active. This
+				 * structure can be freed when refCount
+				 * becomes zero. */
+    struct Command *cmdPtr;	/* Points to the Command structure for this
+				 * procedure. This is used to get the
+				 * namespace in which to execute the
+				 * procedure. */
+    Tcl_Obj *bodyPtr;		/* Points to the ByteCode object for
+				 * procedure's body command. */
+    int numArgs;		/* Number of formal parameters. */
+    int numCompiledLocals;	/* Count of local variables recognized by the
+				 * compiler including arguments and
+				 * temporaries. */
+    CompiledLocal *firstLocalPtr;
+				/* Pointer to first of the procedure's
+				 * compiler-allocated local variables, or NULL
+				 * if none. The first numArgs entries in this
+				 * list describe the procedure's formal
+				 * arguments. */
+    CompiledLocal *lastLocalPtr;/* Pointer to the last allocated local
+				 * variable or NULL if none. This has frame
+				 * index (numCompiledLocals-1). */
+} Proc;
+
+/*
+ * The type of functions called to process errors found during the execution
+ * of a procedure (or lambda term or ...).
+ */
+
+typedef void (ProcErrorProc)(Tcl_Interp *interp, Tcl_Obj *procNameObj);
+
+/*
+ * The structure below defines a command trace. This is used to allow Tcl
+ * clients to find out whenever a command is about to be executed.
+ */
+
+typedef struct Trace {
+    int level;			/* Only trace commands at nesting level less
+				 * than or equal to this. */
+    Tcl_CmdObjTraceProc *proc;	/* Procedure to call to trace command. */
+    ClientData clientData;	/* Arbitrary value to pass to proc. */
+    struct Trace *nextPtr;	/* Next in list of traces for this interp. */
+    int flags;			/* Flags governing the trace - see
+				 * Tcl_CreateObjTrace for details. */
+    Tcl_CmdObjTraceDeleteProc *delProc;
+				/* Procedure to call when trace is deleted. */
+} Trace;
+
+/*
+ * When an interpreter trace is active (i.e. its associated procedure is
+ * executing), one of the following structures is linked into a list
+ * associated with the interpreter. The information in the structure is needed
+ * in order for Tcl to behave reasonably if traces are deleted while traces
+ * are active.
+ */
+
+typedef struct ActiveInterpTrace {
+    struct ActiveInterpTrace *nextPtr;
+				/* Next in list of all active command traces
+				 * for the interpreter, or NULL if no more. */
+    Trace *nextTracePtr;	/* Next trace to check after current trace
+				 * procedure returns; if this trace gets
+				 * deleted, must update pointer to avoid using
+				 * free'd memory. */
+    int reverseScan;		/* Boolean set true when traces are scanning
+				 * in reverse order. */
+} ActiveInterpTrace;
+
+/*
+ * Flag values designating types of execution traces. See tclTrace.c for
+ * related flag values.
+ *
+ * TCL_TRACE_ENTER_EXEC		- triggers enter/enterstep traces.
+ * 				- passed to Tcl_CreateObjTrace to set up
+ *				  "enterstep" traces.
+ * TCL_TRACE_LEAVE_EXEC		- triggers leave/leavestep traces.
+ * 				- passed to Tcl_CreateObjTrace to set up
+ *				  "leavestep" traces.
+ */
+
+#define TCL_TRACE_ENTER_EXEC	1
+#define TCL_TRACE_LEAVE_EXEC	2
+
+/*
+ * The structure below defines an entry in the assocData hash table which is
+ * associated with an interpreter. The entry contains a pointer to a function
+ * to call when the interpreter is deleted, and a pointer to a user-defined
+ * piece of data.
+ */
+
+typedef struct AssocData {
+    Tcl_InterpDeleteProc *proc;	/* Proc to call when deleting. */
+    ClientData clientData;	/* Value to pass to proc. */
+} AssocData;
+
+/*
+ * The structure below defines a call frame. A call frame defines a naming
+ * context for a procedure call: its local naming scope (for local variables)
+ * and its global naming scope (a namespace, perhaps the global :: namespace).
+ * A call frame can also define the naming context for a namespace eval or
+ * namespace inscope command: the namespace in which the command's code should
+ * execute. The Tcl_CallFrame structures exist only while procedures or
+ * namespace eval/inscope's are being executed, and provide a kind of Tcl call
+ * stack.
+ *
+ * WARNING!! The structure definition must be kept consistent with the
+ * Tcl_CallFrame structure in tcl.h. If you change one, change the other.
+ */
+
+/*
+ * Will be grown to contain: pointers to the varnames (allocated at the end),
+ * plus the init values for each variable (suitable to be memcopied on init)
+ */
+
+typedef struct LocalCache {
+    int refCount;
+    int numVars;
+    Tcl_Obj *varName0;
+} LocalCache;
+
+#define localName(framePtr, i) \
+    ((&((framePtr)->localCachePtr->varName0))[(i)])
+
+MODULE_SCOPE void	TclFreeLocalCache(Tcl_Interp *interp,
+			    LocalCache *localCachePtr);
+
+typedef struct CallFrame {
+    Namespace *nsPtr;		/* Points to the namespace used to resolve
+				 * commands and global variables. */
+    int isProcCallFrame;	/* If 0, the frame was pushed to execute a
+				 * namespace command and var references are
+				 * treated as references to namespace vars;
+				 * varTablePtr and compiledLocals are ignored.
+				 * If FRAME_IS_PROC is set, the frame was
+				 * pushed to execute a Tcl procedure and may
+				 * have local vars. */
+    int objc;			/* This and objv below describe the arguments
+				 * for this procedure call. */
+    Tcl_Obj *const *objv;	/* Array of argument objects. */
+    struct CallFrame *callerPtr;
+				/* Value of interp->framePtr when this
+				 * procedure was invoked (i.e. next higher in
+				 * stack of all active procedures). */
+    struct CallFrame *callerVarPtr;
+				/* Value of interp->varFramePtr when this
+				 * procedure was invoked (i.e. determines
+				 * variable scoping within caller). Same as
+				 * callerPtr unless an "uplevel" command or
+				 * something equivalent was active in the
+				 * caller). */
+    int level;			/* Level of this procedure, for "uplevel"
+				 * purposes (i.e. corresponds to nesting of
+				 * callerVarPtr's, not callerPtr's). 1 for
+				 * outermost procedure, 0 for top-level. */
+    Proc *procPtr;		/* Points to the structure defining the called
+				 * procedure. Used to get information such as
+				 * the number of compiled local variables
+				 * (local variables assigned entries ["slots"]
+				 * in the compiledLocals array below). */
+    TclVarHashTable *varTablePtr;
+				/* Hash table containing local variables not
+				 * recognized by the compiler, or created at
+				 * execution time through, e.g., upvar.
+				 * Initially NULL and created if needed. */
+    int numCompiledLocals;	/* Count of local variables recognized by the
+				 * compiler including arguments. */
+    Var *compiledLocals;	/* Points to the array of local variables
+				 * recognized by the compiler. The compiler
+				 * emits code that refers to these variables
+				 * using an index into this array. */
+    ClientData clientData;	/* Pointer to some context that is used by
+				 * object systems. The meaning of the contents
+				 * of this field is defined by the code that
+				 * sets it, and it should only ever be set by
+				 * the code that is pushing the frame. In that
+				 * case, the code that sets it should also
+				 * have some means of discovering what the
+				 * meaning of the value is, which we do not
+				 * specify. */
+    LocalCache *localCachePtr;
+    Tcl_Obj    *tailcallPtr;
+				/* NULL if no tailcall is scheduled */
+} CallFrame;
+
+#define FRAME_IS_PROC	0x1
+#define FRAME_IS_LAMBDA 0x2
+#define FRAME_IS_METHOD	0x4	/* The frame is a method body, and the frame's
+				 * clientData field contains a CallContext
+				 * reference. Part of TIP#257. */
+#define FRAME_IS_OO_DEFINE 0x8	/* The frame is part of the inside workings of
+				 * the [oo::define] command; the clientData
+				 * field contains an Object reference that has
+				 * been confirmed to refer to a class. Part of
+				 * TIP#257. */
+
+/*
+ * TIP #280
+ * The structure below defines a command frame. A command frame provides
+ * location information for all commands executing a tcl script (source, eval,
+ * uplevel, procedure bodies, ...). The runtime structure essentially contains
+ * the stack trace as it would be if the currently executing command were to
+ * throw an error.
+ *
+ * For commands where it makes sense it refers to the associated CallFrame as
+ * well.
+ *
+ * The structures are chained in a single list, with the top of the stack
+ * anchored in the Interp structure.
+ *
+ * Instances can be allocated on the C stack, or the heap, the former making
+ * cleanup a bit simpler.
+ */
+
+typedef struct CmdFrame {
+    /*
+     * General data. Always available.
+     */
+
+    int type;			/* Values see below. */
+    int level;			/* Number of frames in stack, prevent O(n)
+				 * scan of list. */
+    int *line;			/* Lines the words of the command start on. */
+    int nline;
+    CallFrame *framePtr;	/* Procedure activation record, may be
+				 * NULL. */
+    struct CmdFrame *nextPtr;	/* Link to calling frame. */
+    /*
+     * Data needed for Eval vs TEBC
+     *
+     * EXECUTION CONTEXTS and usage of CmdFrame
+     *
+     * Field	  TEBC		  EvalEx
+     * =======	  ====		  ======
+     * level	  yes		  yes
+     * type	  BC/PREBC	  SRC/EVAL
+     * line0	  yes		  yes
+     * framePtr	  yes		  yes
+     * =======	  ====		  ======
+     *
+     * =======	  ====		  ========= union data
+     * line1	  -		  yes
+     * line3	  -		  yes
+     * path	  -		  yes
+     * -------	  ----		  ------
+     * codePtr	  yes		  -
+     * pc	  yes		  -
+     * =======	  ====		  ======
+     *
+     * =======	  ====		  ========= union cmd
+     * str.cmd	  yes		  yes
+     * str.len	  yes		  yes
+     * -------	  ----		  ------
+     */
+
+    union {
+	struct {
+	    Tcl_Obj *path;	/* Path of the sourced file the command is
+				 * in. */
+	} eval;
+	struct {
+	    const void *codePtr;/* Byte code currently executed... */
+	    const char *pc;	/* ... and instruction pointer. */
+	} tebc;
+    } data;
+    Tcl_Obj *cmdObj;
+    const char *cmd;		/* The executed command, if possible... */
+    int len;			/* ... and its length. */
+    const struct CFWordBC *litarg;
+				/* Link to set of literal arguments which have
+				 * ben pushed on the lineLABCPtr stack by
+				 * TclArgumentBCEnter(). These will be removed
+				 * by TclArgumentBCRelease. */
+} CmdFrame;
+
+typedef struct CFWord {
+    CmdFrame *framePtr;		/* CmdFrame to access. */
+    int word;			/* Index of the word in the command. */
+    int refCount;		/* Number of times the word is on the
+				 * stack. */
+} CFWord;
+
+typedef struct CFWordBC {
+    CmdFrame *framePtr;		/* CmdFrame to access. */
+    int pc;			/* Instruction pointer of a command in
+				 * ExtCmdLoc.loc[.] */
+    int word;			/* Index of word in
+				 * ExtCmdLoc.loc[cmd]->line[.] */
+    struct CFWordBC *prevPtr;	/* Previous entry in stack for same Tcl_Obj. */
+    struct CFWordBC *nextPtr;	/* Next entry for same command call. See
+				 * CmdFrame litarg field for the list start. */
+    Tcl_Obj *obj;		/* Back reference to hashtable key */
+} CFWordBC;
+
+/*
+ * Structure to record the locations of invisible continuation lines in
+ * literal scripts, as character offset from the beginning of the script. Both
+ * compiler and direct evaluator use this information to adjust their line
+ * counters when tracking through the script, because when it is invoked the
+ * continuation line marker as a whole has been removed already, meaning that
+ * the \n which was part of it is gone as well, breaking regular line
+ * tracking.
+ *
+ * These structures are allocated and filled by both the function
+ * TclSubstTokens() in the file "tclParse.c" and its caller TclEvalEx() in the
+ * file "tclBasic.c", and stored in the thread-global hashtable "lineCLPtr" in
+ * file "tclObj.c". They are used by the functions TclSetByteCodeFromAny() and
+ * TclCompileScript(), both found in the file "tclCompile.c". Their memory is
+ * released by the function TclFreeObj(), in the file "tclObj.c", and also by
+ * the function TclThreadFinalizeObjects(), in the same file.
+ */
+
+#define CLL_END		(-1)
+
+typedef struct ContLineLoc {
+    int num;			/* Number of entries in loc, not counting the
+				 * final -1 marker entry. */
+    int loc[1];			/* Table of locations, as character offsets.
+				 * The table is allocated as part of the
+				 * structure, extending behind the nominal end
+				 * of the structure. An entry containing the
+				 * value -1 is put after the last location, as
+				 * end-marker/sentinel. */
+} ContLineLoc;
+
+/*
+ * The following macros define the allowed values for the type field of the
+ * CmdFrame structure above. Some of the values occur only in the extended
+ * location data referenced via the 'baseLocPtr'.
+ *
+ * TCL_LOCATION_EVAL	  : Frame is for a script evaluated by EvalEx.
+ * TCL_LOCATION_BC	  : Frame is for bytecode.
+ * TCL_LOCATION_PREBC	  : Frame is for precompiled bytecode.
+ * TCL_LOCATION_SOURCE	  : Frame is for a script evaluated by EvalEx, from a
+ *			    sourced file.
+ * TCL_LOCATION_PROC	  : Frame is for bytecode of a procedure.
+ *
+ * A TCL_LOCATION_BC type in a frame can be overridden by _SOURCE and _PROC
+ * types, per the context of the byte code in execution.
+ */
+
+#define TCL_LOCATION_EVAL	(0) /* Location in a dynamic eval script. */
+#define TCL_LOCATION_BC		(2) /* Location in byte code. */
+#define TCL_LOCATION_PREBC	(3) /* Location in precompiled byte code, no
+				     * location. */
+#define TCL_LOCATION_SOURCE	(4) /* Location in a file. */
+#define TCL_LOCATION_PROC	(5) /* Location in a dynamic proc. */
+#define TCL_LOCATION_LAST	(6) /* Number of values in the enum. */
+
+/*
+ * Structure passed to describe procedure-like "procedures" that are not
+ * procedures (e.g. a lambda) so that their details can be reported correctly
+ * by [info frame]. Contains a sub-structure for each extra field.
+ */
+
+typedef Tcl_Obj * (GetFrameInfoValueProc)(ClientData clientData);
+typedef struct {
+    const char *name;		/* Name of this field. */
+    GetFrameInfoValueProc *proc;	/* Function to generate a Tcl_Obj* from the
+				 * clientData, or just use the clientData
+				 * directly (after casting) if NULL. */
+    ClientData clientData;	/* Context for above function, or Tcl_Obj* if
+				 * proc field is NULL. */
+} ExtraFrameInfoField;
+typedef struct {
+    int length;			/* Length of array. */
+    ExtraFrameInfoField fields[2];
+				/* Really as long as necessary, but this is
+				 * long enough for nearly anything. */
+} ExtraFrameInfo;
+
+/*
+ *----------------------------------------------------------------
+ * Data structures and procedures related to TclHandles, which are a very
+ * lightweight method of preserving enough information to determine if an
+ * arbitrary malloc'd block has been deleted.
+ *----------------------------------------------------------------
+ */
+
+typedef void **TclHandle;
+
+/*
+ *----------------------------------------------------------------
+ * Experimental flag value passed to Tcl_GetRegExpFromObj. Intended for use
+ * only by Expect. It will probably go away in a later release.
+ *----------------------------------------------------------------
+ */
+
+#define TCL_REG_BOSONLY 002000	/* Prepend \A to pattern so it only matches at
+				 * the beginning of the string. */
+
+/*
+ * These are a thin layer over TclpThreadKeyDataGet and TclpThreadKeyDataSet
+ * when threads are used, or an emulation if there are no threads. These are
+ * really internal and Tcl clients should use Tcl_GetThreadData.
+ */
+
+MODULE_SCOPE void *	TclThreadDataKeyGet(Tcl_ThreadDataKey *keyPtr);
+MODULE_SCOPE void	TclThreadDataKeySet(Tcl_ThreadDataKey *keyPtr,
+			    void *data);
+
+/*
+ * This is a convenience macro used to initialize a thread local storage ptr.
+ */
+
+#define TCL_TSD_INIT(keyPtr) \
+	Tcl_GetThreadData((keyPtr), sizeof(ThreadSpecificData))
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to bytecode compilation and execution. These are
+ * used primarily in tclCompile.c, tclExecute.c, and tclBasic.c.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * Forward declaration to prevent errors when the forward references to
+ * Tcl_Parse and CompileEnv are encountered in the procedure type CompileProc
+ * declared below.
+ */
+
+struct CompileEnv;
+
+/*
+ * The type of procedures called by the Tcl bytecode compiler to compile
+ * commands. Pointers to these procedures are kept in the Command structure
+ * describing each command. The integer value returned by a CompileProc must
+ * be one of the following:
+ *
+ * TCL_OK		Compilation completed normally.
+ * TCL_ERROR 		Compilation could not be completed. This can be just a
+ * 			judgment by the CompileProc that the command is too
+ * 			complex to compile effectively, or it can indicate
+ * 			that in the current state of the interp, the command
+ * 			would raise an error. The bytecode compiler will not
+ * 			do any error reporting at compiler time. Error
+ * 			reporting is deferred until the actual runtime,
+ * 			because by then changes in the interp state may allow
+ * 			the command to be successfully evaluated.
+ * TCL_OUT_LINE_COMPILE	A source-compatible alias for TCL_ERROR, kept for the
+ * 			sake of old code only.
+ */
+
+#define TCL_OUT_LINE_COMPILE	TCL_ERROR
+
+typedef int (CompileProc)(Tcl_Interp *interp, Tcl_Parse *parsePtr,
+	struct Command *cmdPtr, struct CompileEnv *compEnvPtr);
+
+/*
+ * The type of procedure called from the compilation hook point in
+ * SetByteCodeFromAny.
+ */
+
+typedef int (CompileHookProc)(Tcl_Interp *interp,
+	struct CompileEnv *compEnvPtr, ClientData clientData);
+
+/*
+ * The data structure for a (linked list of) execution stacks.
+ */
+
+typedef struct ExecStack {
+    struct ExecStack *prevPtr;
+    struct ExecStack *nextPtr;
+    Tcl_Obj **markerPtr;
+    Tcl_Obj **endPtr;
+    Tcl_Obj **tosPtr;
+    Tcl_Obj *stackWords[1];
+} ExecStack;
+
+/*
+ * The data structure defining the execution environment for ByteCode's.
+ * There is one ExecEnv structure per Tcl interpreter. It holds the evaluation
+ * stack that holds command operands and results. The stack grows towards
+ * increasing addresses. The member stackPtr points to the stackItems of the
+ * currently active execution stack.
+ */
+
+typedef struct CorContext {
+    struct CallFrame *framePtr;
+    struct CallFrame *varFramePtr;
+    struct CmdFrame *cmdFramePtr;  /* See Interp.cmdFramePtr */
+    Tcl_HashTable *lineLABCPtr;    /* See Interp.lineLABCPtr */
+} CorContext;
+
+typedef struct CoroutineData {
+    struct Command *cmdPtr;	/* The command handle for the coroutine. */
+    struct ExecEnv *eePtr;	/* The special execution environment (stacks,
+				 * etc.) for the coroutine. */
+    struct ExecEnv *callerEEPtr;/* The execution environment for the caller of
+				 * the coroutine, which might be the
+				 * interpreter global environment or another
+				 * coroutine. */
+    CorContext caller;
+    CorContext running;
+    Tcl_HashTable *lineLABCPtr;    /* See Interp.lineLABCPtr */
+    void *stackLevel;
+    int auxNumLevels;		/* While the coroutine is running the
+				 * numLevels of the create/resume command is
+				 * stored here; for suspended coroutines it
+				 * holds the nesting numLevels at yield. */
+    int nargs;                  /* Number of args required for resuming this
+				 * coroutine; -2 means "0 or 1" (default), -1
+				 * means "any" */
+} CoroutineData;
+
+typedef struct ExecEnv {
+    ExecStack *execStackPtr;	/* Points to the first item in the evaluation
+				 * stack on the heap. */
+    Tcl_Obj *constants[2];	/* Pointers to constant "0" and "1" objs. */
+    struct Tcl_Interp *interp;
+    struct NRE_callback *callbackPtr;
+				/* Top callback in NRE's stack. */
+    struct CoroutineData *corPtr;
+    int rewind;
+} ExecEnv;
+
+#define COR_IS_SUSPENDED(corPtr) \
+    ((corPtr)->stackLevel == NULL)
+
+/*
+ * The definitions for the LiteralTable and LiteralEntry structures. Each
+ * interpreter contains a LiteralTable. It is used to reduce the storage
+ * needed for all the Tcl objects that hold the literals of scripts compiled
+ * by the interpreter. A literal's object is shared by all the ByteCodes that
+ * refer to the literal. Each distinct literal has one LiteralEntry entry in
+ * the LiteralTable. A literal table is a specialized hash table that is
+ * indexed by the literal's string representation, which may contain null
+ * characters.
+ *
+ * Note that we reduce the space needed for literals by sharing literal
+ * objects both within a ByteCode (each ByteCode contains a local
+ * LiteralTable) and across all an interpreter's ByteCodes (with the
+ * interpreter's global LiteralTable).
+ */
+
+typedef struct LiteralEntry {
+    struct LiteralEntry *nextPtr;
+				/* Points to next entry in this hash bucket or
+				 * NULL if end of chain. */
+    Tcl_Obj *objPtr;		/* Points to Tcl object that holds the
+				 * literal's bytes and length. */
+    int refCount;		/* If in an interpreter's global literal
+				 * table, the number of ByteCode structures
+				 * that share the literal object; the literal
+				 * entry can be freed when refCount drops to
+				 * 0. If in a local literal table, -1. */
+    Namespace *nsPtr;		/* Namespace in which this literal is used. We
+				 * try to avoid sharing literal non-FQ command
+				 * names among different namespaces to reduce
+				 * shimmering. */
+} LiteralEntry;
+
+typedef struct LiteralTable {
+    LiteralEntry **buckets;	/* Pointer to bucket array. Each element
+				 * points to first entry in bucket's hash
+				 * chain, or NULL. */
+    LiteralEntry *staticBuckets[TCL_SMALL_HASH_TABLE];
+				/* Bucket array used for small tables to avoid
+				 * mallocs and frees. */
+    int numBuckets;		/* Total number of buckets allocated at
+				 * **buckets. */
+    int numEntries;		/* Total number of entries present in
+				 * table. */
+    int rebuildSize;		/* Enlarge table when numEntries gets to be
+				 * this large. */
+    int mask;			/* Mask value used in hashing function. */
+} LiteralTable;
+
+/*
+ * The following structure defines for each Tcl interpreter various
+ * statistics-related information about the bytecode compiler and
+ * interpreter's operation in that interpreter.
+ */
+
+#ifdef TCL_COMPILE_STATS
+typedef struct ByteCodeStats {
+    long numExecutions;		/* Number of ByteCodes executed. */
+    long numCompilations;	/* Number of ByteCodes created. */
+    long numByteCodesFreed;	/* Number of ByteCodes destroyed. */
+    long instructionCount[256];	/* Number of times each instruction was
+				 * executed. */
+
+    double totalSrcBytes;	/* Total source bytes ever compiled. */
+    double totalByteCodeBytes;	/* Total bytes for all ByteCodes. */
+    double currentSrcBytes;	/* Src bytes for all current ByteCodes. */
+    double currentByteCodeBytes;/* Code bytes in all current ByteCodes. */
+
+    long srcCount[32];		/* Source size distribution: # of srcs of
+				 * size [2**(n-1)..2**n), n in [0..32). */
+    long byteCodeCount[32];	/* ByteCode size distribution. */
+    long lifetimeCount[32];	/* ByteCode lifetime distribution (ms). */
+
+    double currentInstBytes;	/* Instruction bytes-current ByteCodes. */
+    double currentLitBytes;	/* Current literal bytes. */
+    double currentExceptBytes;	/* Current exception table bytes. */
+    double currentAuxBytes;	/* Current auxiliary information bytes. */
+    double currentCmdMapBytes;	/* Current src<->code map bytes. */
+
+    long numLiteralsCreated;	/* Total literal objects ever compiled. */
+    double totalLitStringBytes;	/* Total string bytes in all literals. */
+    double currentLitStringBytes;
+				/* String bytes in current literals. */
+    long literalCount[32];	/* Distribution of literal string sizes. */
+} ByteCodeStats;
+#endif /* TCL_COMPILE_STATS */
+
+/*
+ * Structure used in implementation of those core ensembles which are
+ * partially compiled. Used as an array of these, with a terminating field
+ * whose 'name' is NULL.
+ */
+
+typedef struct {
+    const char *name;		/* The name of the subcommand. */
+    Tcl_ObjCmdProc *proc;	/* The implementation of the subcommand. */
+    CompileProc *compileProc;	/* The compiler for the subcommand. */
+    Tcl_ObjCmdProc *nreProc;	/* NRE implementation of this command. */
+    ClientData clientData;	/* Any clientData to give the command. */
+    int unsafe;			/* Whether this command is to be hidden by
+				 * default in a safe interpreter. */
+} EnsembleImplMap;
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to commands.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * An imported command is created in an namespace when it imports a "real"
+ * command from another namespace. An imported command has a Command structure
+ * that points (via its ClientData value) to the "real" Command structure in
+ * the source namespace's command table. The real command records all the
+ * imported commands that refer to it in a list of ImportRef structures so
+ * that they can be deleted when the real command is deleted.
+ */
+
+typedef struct ImportRef {
+    struct Command *importedCmdPtr;
+				/* Points to the imported command created in
+				 * an importing namespace; this command
+				 * redirects its invocations to the "real"
+				 * command. */
+    struct ImportRef *nextPtr;	/* Next element on the linked list of imported
+				 * commands that refer to the "real" command.
+				 * The real command deletes these imported
+				 * commands on this list when it is
+				 * deleted. */
+} ImportRef;
+
+/*
+ * Data structure used as the ClientData of imported commands: commands
+ * created in an namespace when it imports a "real" command from another
+ * namespace.
+ */
+
+typedef struct ImportedCmdData {
+    struct Command *realCmdPtr;	/* "Real" command that this imported command
+				 * refers to. */
+    struct Command *selfPtr;	/* Pointer to this imported command. Needed
+				 * only when deleting it in order to remove it
+				 * from the real command's linked list of
+				 * imported commands that refer to it. */
+} ImportedCmdData;
+
+/*
+ * A Command structure exists for each command in a namespace. The Tcl_Command
+ * opaque type actually refers to these structures.
+ */
+
+typedef struct Command {
+    Tcl_HashEntry *hPtr;	/* Pointer to the hash table entry that refers
+				 * to this command. The hash table is either a
+				 * namespace's command table or an
+				 * interpreter's hidden command table. This
+				 * pointer is used to get a command's name
+				 * from its Tcl_Command handle. NULL means
+				 * that the hash table entry has been removed
+				 * already (this can happen if deleteProc
+				 * causes the command to be deleted or
+				 * recreated). */
+    Namespace *nsPtr;		/* Points to the namespace containing this
+				 * command. */
+    int refCount;		/* 1 if in command hashtable plus 1 for each
+				 * reference from a CmdName Tcl object
+				 * representing a command's name in a ByteCode
+				 * instruction sequence. This structure can be
+				 * freed when refCount becomes zero. */
+    size_t cmdEpoch;		/* Incremented to invalidate any references
+				 * that point to this command when it is
+				 * renamed, deleted, hidden, or exposed. */
+    CompileProc *compileProc;	/* Procedure called to compile command. NULL
+				 * if no compile proc exists for command. */
+    Tcl_ObjCmdProc *objProc;	/* Object-based command procedure. */
+    ClientData objClientData;	/* Arbitrary value passed to object proc. */
+    Tcl_CmdProc *proc;		/* String-based command procedure. */
+    ClientData clientData;	/* Arbitrary value passed to string proc. */
+    Tcl_CmdDeleteProc *deleteProc;
+				/* Procedure invoked when deleting command to,
+				 * e.g., free all client data. */
+    ClientData deleteData;	/* Arbitrary value passed to deleteProc. */
+    int flags;			/* Miscellaneous bits of information about
+				 * command. See below for definitions. */
+    ImportRef *importRefPtr;	/* List of each imported Command created in
+				 * another namespace when this command is
+				 * imported. These imported commands redirect
+				 * invocations back to this command. The list
+				 * is used to remove all those imported
+				 * commands when deleting this "real"
+				 * command. */
+    CommandTrace *tracePtr;	/* First in list of all traces set for this
+				 * command. */
+    Tcl_ObjCmdProc *nreProc;	/* NRE implementation of this command. */
+} Command;
+
+/*
+ * Flag bits for commands.
+ *
+ * CMD_IS_DELETED -		Means that the command is in the process of
+ *				being deleted (its deleteProc is currently
+ *				executing). Other attempts to delete the
+ *				command should be ignored.
+ * CMD_TRACE_ACTIVE -		1 means that trace processing is currently
+ *				underway for a rename/delete change. See the
+ *				two flags below for which is currently being
+ *				processed.
+ * CMD_HAS_EXEC_TRACES -	1 means that this command has at least one
+ *				execution trace (as opposed to simple
+ *				delete/rename traces) in its tracePtr list.
+ * CMD_COMPILES_EXPANDED -	1 means that this command has a compiler that
+ *				can handle expansion (provided it is not the
+ *				first word).
+ * TCL_TRACE_RENAME -		A rename trace is in progress. Further
+ *				recursive renames will not be traced.
+ * TCL_TRACE_DELETE -		A delete trace is in progress. Further
+ *				recursive deletes will not be traced.
+ * (these last two flags are defined in tcl.h)
+ */
+
+#define CMD_IS_DELETED		    0x01
+#define CMD_TRACE_ACTIVE	    0x02
+#define CMD_HAS_EXEC_TRACES	    0x04
+#define CMD_COMPILES_EXPANDED	    0x08
+#define CMD_REDEF_IN_PROGRESS	    0x10
+#define CMD_VIA_RESOLVER	    0x20
+
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to name resolution procedures.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * The interpreter keeps a linked list of name resolution schemes. The scheme
+ * for a namespace is consulted first, followed by the list of schemes in an
+ * interpreter, followed by the default name resolution in Tcl. Schemes are
+ * added/removed from the interpreter's list by calling Tcl_AddInterpResolver
+ * and Tcl_RemoveInterpResolver.
+ */
+
+typedef struct ResolverScheme {
+    char *name;			/* Name identifying this scheme. */
+    Tcl_ResolveCmdProc *cmdResProc;
+				/* Procedure handling command name
+				 * resolution. */
+    Tcl_ResolveVarProc *varResProc;
+				/* Procedure handling variable name resolution
+				 * for variables that can only be handled at
+				 * runtime. */
+    Tcl_ResolveCompiledVarProc *compiledVarResProc;
+				/* Procedure handling variable name resolution
+				 * at compile time. */
+
+    struct ResolverScheme *nextPtr;
+				/* Pointer to next record in linked list. */
+} ResolverScheme;
+
+/*
+ * Forward declaration of the TIP#143 limit handler structure.
+ */
+
+typedef struct LimitHandler LimitHandler;
+
+/*
+ * TIP #268.
+ * Values for the selection mode, i.e the package require preferences.
+ */
+
+enum PkgPreferOptions {
+    PKG_PREFER_LATEST, PKG_PREFER_STABLE
+};
+
+/*
+ *----------------------------------------------------------------
+ * This structure shadows the first few fields of the memory cache for the
+ * allocator defined in tclThreadAlloc.c; it has to be kept in sync with the
+ * definition there.
+ * Some macros require knowledge of some fields in the struct in order to
+ * avoid hitting the TSD unnecessarily. In order to facilitate this, a pointer
+ * to the relevant fields is kept in the allocCache field in struct Interp.
+ *----------------------------------------------------------------
+ */
+
+typedef struct AllocCache {
+    struct Cache *nextPtr;	/* Linked list of cache entries. */
+    Tcl_ThreadId owner;		/* Which thread's cache is this? */
+    Tcl_Obj *firstObjPtr;	/* List of free objects for thread. */
+    int numObjects;		/* Number of objects for thread. */
+} AllocCache;
+
+/*
+ *----------------------------------------------------------------
+ * This structure defines an interpreter, which is a collection of commands
+ * plus other state information related to interpreting commands, such as
+ * variable storage. Primary responsibility for this data structure is in
+ * tclBasic.c, but almost every Tcl source file uses something in here.
+ *----------------------------------------------------------------
+ */
+
+typedef struct Interp {
+    /*
+     * Note: the first three fields must match exactly the fields in a
+     * Tcl_Interp struct (see tcl.h). If you change one, be sure to change the
+     * other.
+     *
+     * The interpreter's result is held in both the string and the
+     * objResultPtr fields. These fields hold, respectively, the result's
+     * string or object value. The interpreter's result is always in the
+     * result field if that is non-empty, otherwise it is in objResultPtr.
+     * The two fields are kept consistent unless some C code sets
+     * interp->result directly. Programs should not access result and
+     * objResultPtr directly; instead, they should always get and set the
+     * result using procedures such as Tcl_SetObjResult, Tcl_GetObjResult, and
+     * Tcl_GetStringResult. See the SetResult man page for details.
+     */
+
+    char *result;		/* If the last command returned a string
+				 * result, this points to it. Should not be
+				 * accessed directly; see comment above. */
+    Tcl_FreeProc *freeProc;	/* Zero means a string result is statically
+				 * allocated. TCL_DYNAMIC means string result
+				 * was allocated with ckalloc and should be
+				 * freed with ckfree. Other values give
+				 * address of procedure to invoke to free the
+				 * string result. Tcl_Eval must free it before
+				 * executing next command. */
+    int errorLine;		/* When TCL_ERROR is returned, this gives the
+				 * line number in the command where the error
+				 * occurred (1 means first line). */
+    const struct TclStubs *stubTable;
+				/* Pointer to the exported Tcl stub table. On
+				 * previous versions of Tcl this is a pointer
+				 * to the objResultPtr or a pointer to a
+				 * buckets array in a hash table. We therefore
+				 * have to do some careful checking before we
+				 * can use this. */
+
+    TclHandle handle;		/* Handle used to keep track of when this
+				 * interp is deleted. */
+
+    Namespace *globalNsPtr;	/* The interpreter's global namespace. */
+    Tcl_HashTable *hiddenCmdTablePtr;
+				/* Hash table used by tclBasic.c to keep track
+				 * of hidden commands on a per-interp
+				 * basis. */
+    ClientData interpInfo;	/* Information used by tclInterp.c to keep
+				 * track of master/slave interps on a
+				 * per-interp basis. */
+    union {
+	void (*optimizer)(void *envPtr);
+	Tcl_HashTable unused2;	/* No longer used (was mathFuncTable). The
+				 * unused space in interp was repurposed for
+				 * pluggable bytecode optimizers. The core
+				 * contains one optimizer, which can be
+				 * selectively overriden by extensions. */
+    } extra;
+
+    /*
+     * Information related to procedures and variables. See tclProc.c and
+     * tclVar.c for usage.
+     */
+
+    int numLevels;		/* Keeps track of how many nested calls to
+				 * Tcl_Eval are in progress for this
+				 * interpreter. It's used to delay deletion of
+				 * the table until all Tcl_Eval invocations
+				 * are completed. */
+    int maxNestingDepth;	/* If numLevels exceeds this value then Tcl
+				 * assumes that infinite recursion has
+				 * occurred and it generates an error. */
+    CallFrame *framePtr;	/* Points to top-most in stack of all nested
+				 * procedure invocations. */
+    CallFrame *varFramePtr;	/* Points to the call frame whose variables
+				 * are currently in use (same as framePtr
+				 * unless an "uplevel" command is
+				 * executing). */
+    ActiveVarTrace *activeVarTracePtr;
+				/* First in list of active traces for interp,
+				 * or NULL if no active traces. */
+    int returnCode;		/* [return -code] parameter. */
+    CallFrame *rootFramePtr;	/* Global frame pointer for this
+				 * interpreter. */
+    Namespace *lookupNsPtr;	/* Namespace to use ONLY on the next
+				 * TCL_EVAL_INVOKE call to Tcl_EvalObjv. */
+
+    /*
+     * Information used by Tcl_AppendResult to keep track of partial results.
+     * See Tcl_AppendResult code for details.
+     */
+
+#ifndef TCL_NO_DEPRECATED
+    char *appendResult;		/* Storage space for results generated by
+				 * Tcl_AppendResult. Ckalloc-ed. NULL means
+				 * not yet allocated. */
+    int appendAvl;		/* Total amount of space available at
+				 * partialResult. */
+    int appendUsed;		/* Number of non-null bytes currently stored
+				 * at partialResult. */
+#else
+    char *appendResultDontUse;
+    int appendAvlDontUse;
+    int appendUsedDontUse;
+#endif
+
+    /*
+     * Information about packages. Used only in tclPkg.c.
+     */
+
+    Tcl_HashTable packageTable;	/* Describes all of the packages loaded in or
+				 * available to this interpreter. Keys are
+				 * package names, values are (Package *)
+				 * pointers. */
+    char *packageUnknown;	/* Command to invoke during "package require"
+				 * commands for packages that aren't described
+				 * in packageTable. Ckalloc'ed, may be
+				 * NULL. */
+    /*
+     * Miscellaneous information:
+     */
+
+    int cmdCount;		/* Total number of times a command procedure
+				 * has been called for this interpreter. */
+    int evalFlags;		/* Flags to control next call to Tcl_Eval.
+				 * Normally zero, but may be set before
+				 * calling Tcl_Eval. See below for valid
+				 * values. */
+    int unused1;		/* No longer used (was termOffset) */
+    LiteralTable literalTable;	/* Contains LiteralEntry's describing all Tcl
+				 * objects holding literals of scripts
+				 * compiled by the interpreter. Indexed by the
+				 * string representations of literals. Used to
+				 * avoid creating duplicate objects. */
+    unsigned int compileEpoch;	/* Holds the current "compilation epoch" for
+				 * this interpreter. This is incremented to
+				 * invalidate existing ByteCodes when, e.g., a
+				 * command with a compile procedure is
+				 * redefined. */
+    Proc *compiledProcPtr;	/* If a procedure is being compiled, a pointer
+				 * to its Proc structure; otherwise, this is
+				 * NULL. Set by ObjInterpProc in tclProc.c and
+				 * used by tclCompile.c to process local
+				 * variables appropriately. */
+    ResolverScheme *resolverPtr;
+				/* Linked list of name resolution schemes
+				 * added to this interpreter. Schemes are
+				 * added and removed by calling
+				 * Tcl_AddInterpResolvers and
+				 * Tcl_RemoveInterpResolver respectively. */
+    Tcl_Obj *scriptFile;	/* NULL means there is no nested source
+				 * command active; otherwise this points to
+				 * pathPtr of the file being sourced. */
+    int flags;			/* Various flag bits. See below. */
+    long randSeed;		/* Seed used for rand() function. */
+    Trace *tracePtr;		/* List of traces for this interpreter. */
+    Tcl_HashTable *assocData;	/* Hash table for associating data with this
+				 * interpreter. Cleaned up when this
+				 * interpreter is deleted. */
+    struct ExecEnv *execEnvPtr;	/* Execution environment for Tcl bytecode
+				 * execution. Contains a pointer to the Tcl
+				 * evaluation stack. */
+    Tcl_Obj *emptyObjPtr;	/* Points to an object holding an empty
+				 * string. Returned by Tcl_ObjSetVar2 when
+				 * variable traces change a variable in a
+				 * gross way. */
+#ifndef TCL_NO_DEPRECATED
+    char resultSpace[TCL_RESULT_SIZE+1];
+				/* Static space holding small results. */
+#else
+    char resultSpaceDontUse[TCL_RESULT_SIZE+1];
+#endif
+    Tcl_Obj *objResultPtr;	/* If the last command returned an object
+				 * result, this points to it. Should not be
+				 * accessed directly; see comment above. */
+    Tcl_ThreadId threadId;	/* ID of thread that owns the interpreter. */
+
+    ActiveCommandTrace *activeCmdTracePtr;
+				/* First in list of active command traces for
+				 * interp, or NULL if no active traces. */
+    ActiveInterpTrace *activeInterpTracePtr;
+				/* First in list of active traces for interp,
+				 * or NULL if no active traces. */
+
+    int tracesForbiddingInline;	/* Count of traces (in the list headed by
+				 * tracePtr) that forbid inline bytecode
+				 * compilation. */
+
+    /*
+     * Fields used to manage extensible return options (TIP 90).
+     */
+
+    Tcl_Obj *returnOpts;	/* A dictionary holding the options to the
+				 * last [return] command. */
+
+    Tcl_Obj *errorInfo;		/* errorInfo value (now as a Tcl_Obj). */
+    Tcl_Obj *eiVar;		/* cached ref to ::errorInfo variable. */
+    Tcl_Obj *errorCode;		/* errorCode value (now as a Tcl_Obj). */
+    Tcl_Obj *ecVar;		/* cached ref to ::errorInfo variable. */
+    int returnLevel;		/* [return -level] parameter. */
+
+    /*
+     * Resource limiting framework support (TIP#143).
+     */
+
+    struct {
+	int active;		/* Flag values defining which limits have been
+				 * set. */
+	int granularityTicker;	/* Counter used to determine how often to
+				 * check the limits. */
+	int exceeded;		/* Which limits have been exceeded, described
+				 * as flag values the same as the 'active'
+				 * field. */
+
+	int cmdCount;		/* Limit for how many commands to execute in
+				 * the interpreter. */
+	LimitHandler *cmdHandlers;
+				/* Handlers to execute when the limit is
+				 * reached. */
+	int cmdGranularity;	/* Mod factor used to determine how often to
+				 * evaluate the limit check. */
+
+	Tcl_Time time;		/* Time limit for execution within the
+				 * interpreter. */
+	LimitHandler *timeHandlers;
+				/* Handlers to execute when the limit is
+				 * reached. */
+	int timeGranularity;	/* Mod factor used to determine how often to
+				 * evaluate the limit check. */
+	Tcl_TimerToken timeEvent;
+				/* Handle for a timer callback that will occur
+				 * when the time-limit is exceeded. */
+
+	Tcl_HashTable callbacks;/* Mapping from (interp,type) pair to data
+				 * used to install a limit handler callback to
+				 * run in _this_ interp when the limit is
+				 * exceeded. */
+    } limit;
+
+    /*
+     * Information for improved default error generation from ensembles
+     * (TIP#112).
+     */
+
+    struct {
+	Tcl_Obj *const *sourceObjs;
+				/* What arguments were actually input into the
+				 * *root* ensemble command? (Nested ensembles
+				 * don't rewrite this.) NULL if we're not
+				 * processing an ensemble. */
+	int numRemovedObjs;	/* How many arguments have been stripped off
+				 * because of ensemble processing. */
+	int numInsertedObjs;	/* How many of the current arguments were
+				 * inserted by an ensemble. */
+    } ensembleRewrite;
+
+    /*
+     * TIP #219: Global info for the I/O system.
+     */
+
+    Tcl_Obj *chanMsg;		/* Error message set by channel drivers, for
+				 * the propagation of arbitrary Tcl errors.
+				 * This information, if present (chanMsg not
+				 * NULL), takes precedence over a POSIX error
+				 * code returned by a channel operation. */
+
+    /*
+     * Source code origin information (TIP #280).
+     */
+
+    CmdFrame *cmdFramePtr;	/* Points to the command frame containing the
+				 * location information for the current
+				 * command. */
+    const CmdFrame *invokeCmdFramePtr;
+				/* Points to the command frame which is the
+				 * invoking context of the bytecode compiler.
+				 * NULL when the byte code compiler is not
+				 * active. */
+    int invokeWord;		/* Index of the word in the command which
+				 * is getting compiled. */
+    Tcl_HashTable *linePBodyPtr;/* This table remembers for each statically
+				 * defined procedure the location information
+				 * for its body. It is keyed by the address of
+				 * the Proc structure for a procedure. The
+				 * values are "struct CmdFrame*". */
+    Tcl_HashTable *lineBCPtr;	/* This table remembers for each ByteCode
+				 * object the location information for its
+				 * body. It is keyed by the address of the
+				 * Proc structure for a procedure. The values
+				 * are "struct ExtCmdLoc*". (See
+				 * tclCompile.h) */
+    Tcl_HashTable *lineLABCPtr;
+    Tcl_HashTable *lineLAPtr;	/* This table remembers for each argument of a
+				 * command on the execution stack the index of
+				 * the argument in the command, and the
+				 * location data of the command. It is keyed
+				 * by the address of the Tcl_Obj containing
+				 * the argument. The values are "struct
+				 * CFWord*" (See tclBasic.c). This allows
+				 * commands like uplevel, eval, etc. to find
+				 * location information for their arguments,
+				 * if they are a proper literal argument to an
+				 * invoking command. Alt view: An index to the
+				 * CmdFrame stack keyed by command argument
+				 * holders. */
+    ContLineLoc *scriptCLLocPtr;/* This table points to the location data for
+				 * invisible continuation lines in the script,
+				 * if any. This pointer is set by the function
+				 * TclEvalObjEx() in file "tclBasic.c", and
+				 * used by function ...() in the same file.
+				 * It does for the eval/direct path of script
+				 * execution what CompileEnv.clLoc does for
+				 * the bytecode compiler.
+				 */
+    /*
+     * TIP #268. The currently active selection mode, i.e. the package require
+     * preferences.
+     */
+
+    int packagePrefer;		/* Current package selection mode. */
+
+    /*
+     * Hashtables for variable traces and searches.
+     */
+
+    Tcl_HashTable varTraces;	/* Hashtable holding the start of a variable's
+				 * active trace list; varPtr is the key. */
+    Tcl_HashTable varSearches;	/* Hashtable holding the start of a variable's
+				 * active searches list; varPtr is the key. */
+    /*
+     * The thread-specific data ekeko: cache pointers or values that
+     *  (a) do not change during the thread's lifetime
+     *  (b) require access to TSD to determine at runtime
+     *  (c) are accessed very often (e.g., at each command call)
+     *
+     * Note that these are the same for all interps in the same thread. They
+     * just have to be initialised for the thread's master interp, slaves
+     * inherit the value.
+     *
+     * They are used by the macros defined below.
+     */
+
+    AllocCache *allocCache;
+    void *pendingObjDataPtr;	/* Pointer to the Cache and PendingObjData
+				 * structs for this interp's thread; see
+				 * tclObj.c and tclThreadAlloc.c */
+    int *asyncReadyPtr;		/* Pointer to the asyncReady indicator for
+				 * this interp's thread; see tclAsync.c */
+    /*
+     * The pointer to the object system root ekeko. c.f. TIP #257.
+     */
+    void *objectFoundation;	/* Pointer to the Foundation structure of the
+				 * object system, which contains things like
+				 * references to key namespaces. See
+				 * tclOOInt.h and tclOO.c for real definition
+				 * and setup. */
+
+    struct NRE_callback *deferredCallbacks;
+				/* Callbacks that are set previous to a call
+				 * to some Eval function but that actually
+				 * belong to the command that is about to be
+				 * called - i.e., they should be run *before*
+				 * any tailcall is invoked. */
+
+    /*
+     * TIP #285, Script cancellation support.
+     */
+
+    Tcl_AsyncHandler asyncCancel;
+				/* Async handler token for Tcl_CancelEval. */
+    Tcl_Obj *asyncCancelMsg;	/* Error message set by async cancel handler
+				 * for the propagation of arbitrary Tcl
+				 * errors. This information, if present
+				 * (asyncCancelMsg not NULL), takes precedence
+				 * over the default error messages returned by
+				 * a script cancellation operation. */
+
+	/*
+	 * TIP #348 IMPLEMENTATION  -  Substituted error stack
+	 */
+    Tcl_Obj *errorStack;	/* [info errorstack] value (as a Tcl_Obj). */
+    Tcl_Obj *upLiteral;		/* "UP" literal for [info errorstack] */
+    Tcl_Obj *callLiteral;	/* "CALL" literal for [info errorstack] */
+    Tcl_Obj *innerLiteral;	/* "INNER" literal for [info errorstack] */
+    Tcl_Obj *innerContext;	/* cached list for fast reallocation */
+    int resetErrorStack;        /* controls cleaning up of ::errorStack */
+
+#ifdef TCL_COMPILE_STATS
+    /*
+     * Statistical information about the bytecode compiler and interpreter's
+     * operation. This should be the last field of Interp.
+     */
+
+    ByteCodeStats stats;	/* Holds compilation and execution statistics
+				 * for this interpreter. */
+#endif /* TCL_COMPILE_STATS */
+} Interp;
+
+/*
+ * Macros that use the TSD-ekeko.
+ */
+
+#define TclAsyncReady(iPtr) \
+    *((iPtr)->asyncReadyPtr)
+
+/*
+ * Macros for script cancellation support (TIP #285).
+ */
+
+#define TclCanceled(iPtr) \
+    (((iPtr)->flags & CANCELED) || ((iPtr)->flags & TCL_CANCEL_UNWIND))
+
+#define TclSetCancelFlags(iPtr, cancelFlags)   \
+    (iPtr)->flags |= CANCELED;                 \
+    if ((cancelFlags) & TCL_CANCEL_UNWIND) {   \
+        (iPtr)->flags |= TCL_CANCEL_UNWIND;    \
+    }
+
+#define TclUnsetCancelFlags(iPtr) \
+    (iPtr)->flags &= (~(CANCELED | TCL_CANCEL_UNWIND))
+
+/*
+ * Macros for splicing into and out of doubly linked lists. They assume
+ * existence of struct items 'prevPtr' and 'nextPtr'.
+ *
+ * a = element to add or remove.
+ * b = list head.
+ *
+ * TclSpliceIn adds to the head of the list.
+ */
+
+#define TclSpliceIn(a,b)			\
+    (a)->nextPtr = (b);				\
+    if ((b) != NULL) {				\
+	(b)->prevPtr = (a);			\
+    }						\
+    (a)->prevPtr = NULL, (b) = (a);
+
+#define TclSpliceOut(a,b)			\
+    if ((a)->prevPtr != NULL) {			\
+	(a)->prevPtr->nextPtr = (a)->nextPtr;	\
+    } else {					\
+	(b) = (a)->nextPtr;			\
+    }						\
+    if ((a)->nextPtr != NULL) {			\
+	(a)->nextPtr->prevPtr = (a)->prevPtr;	\
+    }
+
+/*
+ * EvalFlag bits for Interp structures:
+ *
+ * TCL_ALLOW_EXCEPTIONS	1 means it's OK for the script to terminate with a
+ *			code other than TCL_OK or TCL_ERROR; 0 means codes
+ *			other than these should be turned into errors.
+ */
+
+#define TCL_ALLOW_EXCEPTIONS		0x04
+#define TCL_EVAL_FILE			0x02
+#define TCL_EVAL_SOURCE_IN_FRAME	0x10
+#define TCL_EVAL_NORESOLVE		0x20
+
+/*
+ * Flag bits for Interp structures:
+ *
+ * DELETED:		Non-zero means the interpreter has been deleted:
+ *			don't process any more commands for it, and destroy
+ *			the structure as soon as all nested invocations of
+ *			Tcl_Eval are done.
+ * ERR_ALREADY_LOGGED:	Non-zero means information has already been logged in
+ *			iPtr->errorInfo for the current Tcl_Eval instance, so
+ *			Tcl_Eval needn't log it (used to implement the "error
+ *			message log" command).
+ * DONT_COMPILE_CMDS_INLINE: Non-zero means that the bytecode compiler should
+ *			not compile any commands into an inline sequence of
+ *			instructions. This is set 1, for example, when command
+ *			traces are requested.
+ * RAND_SEED_INITIALIZED: Non-zero means that the randSeed value of the interp
+ *			has not be initialized. This is set 1 when we first
+ *			use the rand() or srand() functions.
+ * SAFE_INTERP:		Non zero means that the current interp is a safe
+ *			interp (i.e. it has only the safe commands installed,
+ *			less priviledge than a regular interp).
+ * INTERP_DEBUG_FRAME:	Used for switching on various extra interpreter
+ *			debug/info mechanisms (e.g. info frame eval/uplevel
+ *			tracing) which are performance intensive.
+ * INTERP_TRACE_IN_PROGRESS: Non-zero means that an interp trace is currently
+ *			active; so no further trace callbacks should be
+ *			invoked.
+ * INTERP_ALTERNATE_WRONG_ARGS: Used for listing second and subsequent forms
+ *			of the wrong-num-args string in Tcl_WrongNumArgs.
+ *			Makes it append instead of replacing and uses
+ *			different intermediate text.
+ * CANCELED:		Non-zero means that the script in progress should be
+ *			canceled as soon as possible. This can be checked by
+ *			extensions (and the core itself) by calling
+ *			Tcl_Canceled and checking if TCL_ERROR is returned.
+ *			This is a one-shot flag that is reset immediately upon
+ *			being detected; however, if the TCL_CANCEL_UNWIND flag
+ *			is set Tcl_Canceled will continue to report that the
+ *			script in progress has been canceled thereby allowing
+ *			the evaluation stack for the interp to be fully
+ *			unwound.
+ *
+ * WARNING: For the sake of some extensions that have made use of former
+ * internal values, do not re-use the flag values 2 (formerly ERR_IN_PROGRESS)
+ * or 8 (formerly ERROR_CODE_SET).
+ */
+
+#define DELETED				     1
+#define ERR_ALREADY_LOGGED		     4
+#define INTERP_DEBUG_FRAME		  0x10
+#define DONT_COMPILE_CMDS_INLINE	  0x20
+#define RAND_SEED_INITIALIZED		  0x40
+#define SAFE_INTERP			  0x80
+#define INTERP_TRACE_IN_PROGRESS	 0x200
+#define INTERP_ALTERNATE_WRONG_ARGS	 0x400
+#define ERR_LEGACY_COPY			 0x800
+#define CANCELED			0x1000
+
+/*
+ * Maximum number of levels of nesting permitted in Tcl commands (used to
+ * catch infinite recursion).
+ */
+
+#define MAX_NESTING_DEPTH	1000
+
+/*
+ * The macro below is used to modify a "char" value (e.g. by casting it to an
+ * unsigned character) so that it can be used safely with macros such as
+ * isspace.
+ */
+
+#define UCHAR(c) ((unsigned char) (c))
+
+/*
+ * This macro is used to properly align the memory allocated by Tcl, giving
+ * the same alignment as the native malloc.
+ */
+
+#if defined(__APPLE__)
+#define TCL_ALLOCALIGN	16
+#else
+#define TCL_ALLOCALIGN	(2*sizeof(void *))
+#endif
+
+/*
+ * This macro is used to determine the offset needed to safely allocate any
+ * data structure in memory. Given a starting offset or size, it "rounds up"
+ * or "aligns" the offset to the next 8-byte boundary so that any data
+ * structure can be placed at the resulting offset without fear of an
+ * alignment error.
+ *
+ * WARNING!! DO NOT USE THIS MACRO TO ALIGN POINTERS: it will produce the
+ * wrong result on platforms that allocate addresses that are divisible by 4
+ * or 2. Only use it for offsets or sizes.
+ *
+ * This macro is only used by tclCompile.c in the core (Bug 926445). It
+ * however not be made file static, as extensions that touch bytecodes
+ * (notably tbcload) require it.
+ */
+
+#define TCL_ALIGN(x) (((int)(x) + 7) & ~7)
+
+/*
+ * The following enum values are used to specify the runtime platform setting
+ * of the tclPlatform variable.
+ */
+
+typedef enum {
+    TCL_PLATFORM_UNIX = 0,	/* Any Unix-like OS. */
+    TCL_PLATFORM_WINDOWS = 2	/* Any Microsoft Windows OS. */
+} TclPlatformType;
+
+/*
+ * The following enum values are used to indicate the translation of a Tcl
+ * channel. Declared here so that each platform can define
+ * TCL_PLATFORM_TRANSLATION to the native translation on that platform.
+ */
+
+typedef enum TclEolTranslation {
+    TCL_TRANSLATE_AUTO,		/* Eol == \r, \n and \r\n. */
+    TCL_TRANSLATE_CR,		/* Eol == \r. */
+    TCL_TRANSLATE_LF,		/* Eol == \n. */
+    TCL_TRANSLATE_CRLF		/* Eol == \r\n. */
+} TclEolTranslation;
+
+/*
+ * Flags for TclInvoke:
+ *
+ * TCL_INVOKE_HIDDEN		Invoke a hidden command; if not set, invokes
+ *				an exposed command.
+ * TCL_INVOKE_NO_UNKNOWN	If set, "unknown" is not invoked if the
+ *				command to be invoked is not found. Only has
+ *				an effect if invoking an exposed command,
+ *				i.e. if TCL_INVOKE_HIDDEN is not also set.
+ * TCL_INVOKE_NO_TRACEBACK	Does not record traceback information if the
+ *				invoked command returns an error. Used if the
+ *				caller plans on recording its own traceback
+ *				information.
+ */
+
+#define	TCL_INVOKE_HIDDEN	(1<<0)
+#define TCL_INVOKE_NO_UNKNOWN	(1<<1)
+#define TCL_INVOKE_NO_TRACEBACK	(1<<2)
+
+/*
+ * The structure used as the internal representation of Tcl list objects. This
+ * struct is grown (reallocated and copied) as necessary to hold all the
+ * list's element pointers. The struct might contain more slots than currently
+ * used to hold all element pointers. This is done to make append operations
+ * faster.
+ */
+
+typedef struct List {
+    int refCount;
+    int maxElemCount;		/* Total number of element array slots. */
+    int elemCount;		/* Current number of list elements. */
+    int canonicalFlag;		/* Set if the string representation was
+				 * derived from the list representation. May
+				 * be ignored if there is no string rep at
+				 * all.*/
+    Tcl_Obj *elements;		/* First list element; the struct is grown to
+				 * accomodate all elements. */
+} List;
+
+#define LIST_MAX \
+	(1 + (int)(((size_t)UINT_MAX - sizeof(List))/sizeof(Tcl_Obj *)))
+#define LIST_SIZE(numElems) \
+	(unsigned)(sizeof(List) + (((numElems) - 1) * sizeof(Tcl_Obj *)))
+
+/*
+ * Macro used to get the elements of a list object.
+ */
+
+#define ListRepPtr(listPtr) \
+    ((List *) (listPtr)->internalRep.twoPtrValue.ptr1)
+
+#define ListSetIntRep(objPtr, listRepPtr) \
+    (objPtr)->internalRep.twoPtrValue.ptr1 = (void *)(listRepPtr), \
+    (objPtr)->internalRep.twoPtrValue.ptr2 = NULL, \
+    (listRepPtr)->refCount++, \
+    (objPtr)->typePtr = &tclListType
+
+#define ListObjGetElements(listPtr, objc, objv) \
+    ((objv) = &(ListRepPtr(listPtr)->elements), \
+     (objc) = ListRepPtr(listPtr)->elemCount)
+
+#define ListObjLength(listPtr, len) \
+    ((len) = ListRepPtr(listPtr)->elemCount)
+
+#define ListObjIsCanonical(listPtr) \
+    (((listPtr)->bytes == NULL) || ListRepPtr(listPtr)->canonicalFlag)
+
+#define TclListObjGetElements(interp, listPtr, objcPtr, objvPtr) \
+    (((listPtr)->typePtr == &tclListType) \
+	    ? ((ListObjGetElements((listPtr), *(objcPtr), *(objvPtr))), TCL_OK)\
+	    : Tcl_ListObjGetElements((interp), (listPtr), (objcPtr), (objvPtr)))
+
+#define TclListObjLength(interp, listPtr, lenPtr) \
+    (((listPtr)->typePtr == &tclListType) \
+	    ? ((ListObjLength((listPtr), *(lenPtr))), TCL_OK)\
+	    : Tcl_ListObjLength((interp), (listPtr), (lenPtr)))
+
+#define TclListObjIsCanonical(listPtr) \
+    (((listPtr)->typePtr == &tclListType) ? ListObjIsCanonical((listPtr)) : 0)
+
+/*
+ * Modes for collecting (or not) in the implementations of TclNRForeachCmd,
+ * TclNRLmapCmd and their compilations.
+ */
+
+#define TCL_EACH_KEEP_NONE  0	/* Discard iteration result like [foreach] */
+#define TCL_EACH_COLLECT    1	/* Collect iteration result like [lmap] */
+
+/*
+ * Macros providing a faster path to integers: Tcl_GetLongFromObj everywhere,
+ * Tcl_GetIntFromObj and TclGetIntForIndex on platforms where longs are ints.
+ *
+ * WARNING: these macros eval their args more than once.
+ */
+
+#define TclGetLongFromObj(interp, objPtr, longPtr) \
+    (((objPtr)->typePtr == &tclIntType)	\
+	    ? ((*(longPtr) = (objPtr)->internalRep.longValue), TCL_OK) \
+	    : Tcl_GetLongFromObj((interp), (objPtr), (longPtr)))
+
+#if (LONG_MAX == INT_MAX)
+#define TclGetIntFromObj(interp, objPtr, intPtr) \
+    (((objPtr)->typePtr == &tclIntType)	\
+	    ? ((*(intPtr) = (objPtr)->internalRep.longValue), TCL_OK) \
+	    : Tcl_GetIntFromObj((interp), (objPtr), (intPtr)))
+#define TclGetIntForIndexM(interp, objPtr, endValue, idxPtr) \
+    (((objPtr)->typePtr == &tclIntType)	\
+	    ? ((*(idxPtr) = (objPtr)->internalRep.longValue), TCL_OK) \
+	    : TclGetIntForIndex((interp), (objPtr), (endValue), (idxPtr)))
+#else
+#define TclGetIntFromObj(interp, objPtr, intPtr) \
+    Tcl_GetIntFromObj((interp), (objPtr), (intPtr))
+#define TclGetIntForIndexM(interp, objPtr, ignore, idxPtr)	\
+    TclGetIntForIndex(interp, objPtr, ignore, idxPtr)
+#endif
+
+/*
+ * Macro used to save a function call for common uses of
+ * Tcl_GetWideIntFromObj(). The ANSI C "prototype" is:
+ *
+ * MODULE_SCOPE int TclGetWideIntFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+ *			Tcl_WideInt *wideIntPtr);
+ */
+
+#ifdef TCL_WIDE_INT_IS_LONG
+#define TclGetWideIntFromObj(interp, objPtr, wideIntPtr) \
+    (((objPtr)->typePtr == &tclIntType)					\
+	? (*(wideIntPtr) = (Tcl_WideInt)				\
+		((objPtr)->internalRep.longValue), TCL_OK) :		\
+	Tcl_GetWideIntFromObj((interp), (objPtr), (wideIntPtr)))
+#else /* !TCL_WIDE_INT_IS_LONG */
+#define TclGetWideIntFromObj(interp, objPtr, wideIntPtr)		\
+    (((objPtr)->typePtr == &tclWideIntType)				\
+	? (*(wideIntPtr) = (objPtr)->internalRep.wideValue, TCL_OK) :	\
+    ((objPtr)->typePtr == &tclIntType)					\
+	? (*(wideIntPtr) = (Tcl_WideInt)				\
+		((objPtr)->internalRep.longValue), TCL_OK) :		\
+	Tcl_GetWideIntFromObj((interp), (objPtr), (wideIntPtr)))
+#endif /* TCL_WIDE_INT_IS_LONG */
+
+/*
+ * Flag values for TclTraceDictPath().
+ *
+ * DICT_PATH_READ indicates that all entries on the path must exist but no
+ * updates will be needed.
+ *
+ * DICT_PATH_UPDATE indicates that we are going to be doing an update at the
+ * tip of the path, so duplication of shared objects should be done along the
+ * way.
+ *
+ * DICT_PATH_EXISTS indicates that we are performing an existance test and a
+ * lookup failure should therefore not be an error. If (and only if) this flag
+ * is set, TclTraceDictPath() will return the special value
+ * DICT_PATH_NON_EXISTENT if the path is not traceable.
+ *
+ * DICT_PATH_CREATE (which also requires the DICT_PATH_UPDATE bit to be set)
+ * indicates that we are to create non-existant dictionaries on the path.
+ */
+
+#define DICT_PATH_READ		0
+#define DICT_PATH_UPDATE	1
+#define DICT_PATH_EXISTS	2
+#define DICT_PATH_CREATE	5
+
+#define DICT_PATH_NON_EXISTENT	((Tcl_Obj *) (void *) 1)
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to the filesystem internals
+ *----------------------------------------------------------------
+ */
+
+/*
+ * The version_2 filesystem is private to Tcl. As and when these changes have
+ * been thoroughly tested and investigated a new public filesystem interface
+ * will be released. The aim is more versatile virtual filesystem interfaces,
+ * more efficiency in 'path' manipulation and usage, and cleaner filesystem
+ * code internally.
+ */
+
+#define TCL_FILESYSTEM_VERSION_2	((Tcl_FSVersion) 0x2)
+typedef ClientData (TclFSGetCwdProc2)(ClientData clientData);
+typedef int (Tcl_FSLoadFileProc2) (Tcl_Interp *interp, Tcl_Obj *pathPtr,
+	Tcl_LoadHandle *handlePtr, Tcl_FSUnloadFileProc **unloadProcPtr, int flags);
+
+/*
+ * The following types are used for getting and storing platform-specific file
+ * attributes in tclFCmd.c and the various platform-versions of that file.
+ * This is done to have as much common code as possible in the file attributes
+ * code. For more information about the callbacks, see TclFileAttrsCmd in
+ * tclFCmd.c.
+ */
+
+typedef int (TclGetFileAttrProc)(Tcl_Interp *interp, int objIndex,
+	Tcl_Obj *fileName, Tcl_Obj **attrObjPtrPtr);
+typedef int (TclSetFileAttrProc)(Tcl_Interp *interp, int objIndex,
+	Tcl_Obj *fileName, Tcl_Obj *attrObjPtr);
+
+typedef struct TclFileAttrProcs {
+    TclGetFileAttrProc *getProc;/* The procedure for getting attrs. */
+    TclSetFileAttrProc *setProc;/* The procedure for setting attrs. */
+} TclFileAttrProcs;
+
+/*
+ * Private flag value which controls Tcl_GetIndexFromObj*() routines
+ * to instruct them not to cache lookups because the table will not
+ * live long enough to make it worthwhile.  Must not clash with public
+ * flag value TCL_EXACT.
+ */
+
+#define INDEX_TEMP_TABLE 2
+
+/*
+ * Opaque handle used in pipeline routines to encapsulate platform-dependent
+ * state.
+ */
+
+typedef struct TclFile_ *TclFile;
+
+/*
+ * The "globParameters" argument of the function TclGlob is an or'ed
+ * combination of the following values:
+ */
+
+#define TCL_GLOBMODE_NO_COMPLAIN	1
+#define TCL_GLOBMODE_JOIN		2
+#define TCL_GLOBMODE_DIR		4
+#define TCL_GLOBMODE_TAILS		8
+
+typedef enum Tcl_PathPart {
+    TCL_PATH_DIRNAME,
+    TCL_PATH_TAIL,
+    TCL_PATH_EXTENSION,
+    TCL_PATH_ROOT
+} Tcl_PathPart;
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to obsolete filesystem hooks
+ *----------------------------------------------------------------
+ */
+
+typedef int (TclStatProc_)(const char *path, struct stat *buf);
+typedef int (TclAccessProc_)(const char *path, int mode);
+typedef Tcl_Channel (TclOpenFileChannelProc_)(Tcl_Interp *interp,
+	const char *fileName, const char *modeString, int permissions);
+
+/*
+ *----------------------------------------------------------------
+ * Data structures related to procedures
+ *----------------------------------------------------------------
+ */
+
+typedef Tcl_CmdProc *TclCmdProcType;
+typedef Tcl_ObjCmdProc *TclObjCmdProcType;
+
+/*
+ *----------------------------------------------------------------
+ * Data structures for process-global values.
+ *----------------------------------------------------------------
+ */
+
+typedef void (TclInitProcessGlobalValueProc)(char **valuePtr, size_t *lengthPtr,
+	Tcl_Encoding *encodingPtr);
+
+/*
+ * A ProcessGlobalValue struct exists for each internal value in Tcl that is
+ * to be shared among several threads. Each thread sees a (Tcl_Obj) copy of
+ * the value, and the master is kept as a counted string, with epoch and mutex
+ * control. Each ProcessGlobalValue struct should be a static variable in some
+ * file.
+ */
+
+typedef struct ProcessGlobalValue {
+    size_t epoch;		/* Epoch counter to detect changes in the
+				 * master value. */
+    size_t numBytes;		/* Length of the master string. */
+    char *value;		/* The master string value. */
+    Tcl_Encoding encoding;	/* system encoding when master string was
+				 * initialized. */
+    TclInitProcessGlobalValueProc *proc;
+    				/* A procedure to initialize the master string
+				 * copy when a "get" request comes in before
+				 * any "set" request has been received. */
+    Tcl_Mutex mutex;		/* Enforce orderly access from multiple
+				 * threads. */
+    Tcl_ThreadDataKey key;	/* Key for per-thread data holding the
+				 * (Tcl_Obj) copy for each thread. */
+} ProcessGlobalValue;
+
+/*
+ *----------------------------------------------------------------------
+ * Flags for TclParseNumber
+ *----------------------------------------------------------------------
+ */
+
+#define TCL_PARSE_DECIMAL_ONLY		1
+				/* Leading zero doesn't denote octal or
+				 * hex. */
+#define TCL_PARSE_OCTAL_ONLY		2
+				/* Parse octal even without prefix. */
+#define TCL_PARSE_HEXADECIMAL_ONLY	4
+				/* Parse hexadecimal even without prefix. */
+#define TCL_PARSE_INTEGER_ONLY		8
+				/* Disable floating point parsing. */
+#define TCL_PARSE_SCAN_PREFIXES		16
+				/* Use [scan] rules dealing with 0?
+				 * prefixes. */
+#define TCL_PARSE_NO_WHITESPACE		32
+				/* Reject leading/trailing whitespace. */
+#define TCL_PARSE_BINARY_ONLY	64
+				/* Parse binary even without prefix. */
+
+/*
+ *----------------------------------------------------------------------
+ * Type values TclGetNumberFromObj
+ *----------------------------------------------------------------------
+ */
+
+#define TCL_NUMBER_LONG		1
+#define TCL_NUMBER_WIDE		2
+#define TCL_NUMBER_BIG		3
+#define TCL_NUMBER_DOUBLE	4
+#define TCL_NUMBER_NAN		5
+
+/*
+ *----------------------------------------------------------------
+ * Variables shared among Tcl modules but not used by the outside world.
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE char *tclNativeExecutableName;
+MODULE_SCOPE int tclFindExecutableSearchDone;
+MODULE_SCOPE char *tclMemDumpFileName;
+MODULE_SCOPE TclPlatformType tclPlatform;
+MODULE_SCOPE Tcl_NotifierProcs tclNotifierHooks;
+
+MODULE_SCOPE Tcl_Encoding tclIdentityEncoding;
+
+/*
+ * TIP #233 (Virtualized Time)
+ * Data for the time hooks, if any.
+ */
+
+MODULE_SCOPE Tcl_GetTimeProc *tclGetTimeProcPtr;
+MODULE_SCOPE Tcl_ScaleTimeProc *tclScaleTimeProcPtr;
+MODULE_SCOPE ClientData tclTimeClientData;
+
+/*
+ * Variables denoting the Tcl object types defined in the core.
+ */
+
+MODULE_SCOPE const Tcl_ObjType tclBignumType;
+MODULE_SCOPE const Tcl_ObjType tclBooleanType;
+MODULE_SCOPE const Tcl_ObjType tclByteArrayType;
+MODULE_SCOPE const Tcl_ObjType tclByteCodeType;
+MODULE_SCOPE const Tcl_ObjType tclDoubleType;
+MODULE_SCOPE const Tcl_ObjType tclEndOffsetType;
+MODULE_SCOPE const Tcl_ObjType tclIntType;
+MODULE_SCOPE const Tcl_ObjType tclListType;
+MODULE_SCOPE const Tcl_ObjType tclDictType;
+MODULE_SCOPE const Tcl_ObjType tclProcBodyType;
+MODULE_SCOPE const Tcl_ObjType tclStringType;
+MODULE_SCOPE const Tcl_ObjType tclEnsembleCmdType;
+#ifndef TCL_WIDE_INT_IS_LONG
+MODULE_SCOPE const Tcl_ObjType tclWideIntType;
+#endif
+MODULE_SCOPE const Tcl_ObjType tclRegexpType;
+MODULE_SCOPE Tcl_ObjType tclCmdNameType;
+
+/*
+ * Variables denoting the hash key types defined in the core.
+ */
+
+MODULE_SCOPE const Tcl_HashKeyType tclArrayHashKeyType;
+MODULE_SCOPE const Tcl_HashKeyType tclOneWordHashKeyType;
+MODULE_SCOPE const Tcl_HashKeyType tclStringHashKeyType;
+MODULE_SCOPE const Tcl_HashKeyType tclObjHashKeyType;
+
+/*
+ * The head of the list of free Tcl objects, and the total number of Tcl
+ * objects ever allocated and freed.
+ */
+
+MODULE_SCOPE Tcl_Obj *	tclFreeObjList;
+
+#ifdef TCL_COMPILE_STATS
+MODULE_SCOPE long	tclObjsAlloced;
+MODULE_SCOPE long	tclObjsFreed;
+#define TCL_MAX_SHARED_OBJ_STATS 5
+MODULE_SCOPE long	tclObjsShared[TCL_MAX_SHARED_OBJ_STATS];
+#endif /* TCL_COMPILE_STATS */
+
+/*
+ * Pointer to a heap-allocated string of length zero that the Tcl core uses as
+ * the value of an empty string representation for an object. This value is
+ * shared by all new objects allocated by Tcl_NewObj.
+ */
+
+MODULE_SCOPE char	tclEmptyString;
+
+/*
+ *----------------------------------------------------------------
+ * Procedures shared among Tcl modules but not used by the outside world,
+ * introduced by/for NRE.
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE Tcl_ObjCmdProc TclNRApplyObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNREvalObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRCatchObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRExprObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRForObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRForeachCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRIfObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRLmapCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRSourceObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRSubstObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRSwitchObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRTryObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRUplevelObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRWhileObjCmd;
+
+MODULE_SCOPE Tcl_NRPostProc TclNRForIterCallback;
+MODULE_SCOPE Tcl_NRPostProc TclNRCoroutineActivateCallback;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRTailcallObjCmd;
+MODULE_SCOPE Tcl_NRPostProc TclNRTailcallEval;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRCoroutineObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRYieldObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRYieldmObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRYieldToObjCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclNRInvoke;
+MODULE_SCOPE Tcl_NRPostProc TclNRReleaseValues;
+
+MODULE_SCOPE void  TclSetTailcall(Tcl_Interp *interp, Tcl_Obj *tailcallPtr);
+MODULE_SCOPE void  TclPushTailcallPoint(Tcl_Interp *interp);
+
+/* These two can be considered for the public api */
+MODULE_SCOPE void  TclMarkTailcall(Tcl_Interp *interp);
+MODULE_SCOPE void  TclSkipTailcall(Tcl_Interp *interp);
+
+/*
+ * This structure holds the data for the various iteration callbacks used to
+ * NRE the 'for' and 'while' commands. We need a separate structure because we
+ * have more than the 4 client data entries we can provide directly thorugh
+ * the callback API. It is the 'word' information which puts us over the
+ * limit. It is needed because the loop body is argument 4 of 'for' and
+ * argument 2 of 'while'. Not providing the correct index confuses the #280
+ * code. We TclSmallAlloc/Free this.
+ */
+
+typedef struct ForIterData {
+    Tcl_Obj *cond;		/* Loop condition expression. */
+    Tcl_Obj *body;		/* Loop body. */
+    Tcl_Obj *next;		/* Loop step script, NULL for 'while'. */
+    const char *msg;		/* Error message part. */
+    int word;			/* Index of the body script in the command */
+} ForIterData;
+
+/* TIP #357 - Structure doing the bookkeeping of handles for Tcl_LoadFile
+ *            and Tcl_FindSymbol. This structure corresponds to an opaque
+ *            typedef in tcl.h */
+
+typedef void* TclFindSymbolProc(Tcl_Interp* interp, Tcl_LoadHandle loadHandle,
+				const char* symbol);
+struct Tcl_LoadHandle_ {
+    ClientData clientData;	/* Client data is the load handle in the
+				 * native filesystem if a module was loaded
+				 * there, or an opaque pointer to a structure
+				 * for further bookkeeping on load-from-VFS
+				 * and load-from-memory */
+    TclFindSymbolProc* findSymbolProcPtr;
+				/* Procedure that resolves symbols in a
+				 * loaded module */
+    Tcl_FSUnloadFileProc* unloadFileProcPtr;
+				/* Procedure that unloads a loaded module */
+};
+
+/* Flags for conversion of doubles to digit strings */
+
+#define TCL_DD_SHORTEST 		0x4
+				/* Use the shortest possible string */
+#define TCL_DD_STEELE   		0x5
+				/* Use the original Steele&White algorithm */
+#define TCL_DD_E_FORMAT 		0x2
+				/* Use a fixed-length string of digits,
+				 * suitable for E format*/
+#define TCL_DD_F_FORMAT 		0x3
+				/* Use a fixed number of digits after the
+				 * decimal point, suitable for F format */
+
+#define TCL_DD_SHORTEN_FLAG 		0x4
+				/* Allow return of a shorter digit string
+				 * if it converts losslessly */
+#define TCL_DD_NO_QUICK 		0x8
+				/* Debug flag: forbid quick FP conversion */
+
+#define TCL_DD_CONVERSION_TYPE_MASK	0x3
+				/* Mask to isolate the conversion type */
+#define TCL_DD_STEELE0 			0x1
+				/* 'Steele&White' after masking */
+#define TCL_DD_SHORTEST0		0x0
+				/* 'Shortest possible' after masking */
+
+/*
+ *----------------------------------------------------------------
+ * Procedures shared among Tcl modules but not used by the outside world:
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE void	TclAppendBytesToByteArray(Tcl_Obj *objPtr,
+			    const unsigned char *bytes, int len);
+MODULE_SCOPE int	TclNREvalCmd(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    int flags);
+MODULE_SCOPE void	TclAdvanceContinuations(int *line, int **next,
+			    int loc);
+MODULE_SCOPE void	TclAdvanceLines(int *line, const char *start,
+			    const char *end);
+MODULE_SCOPE void	TclArgumentEnter(Tcl_Interp *interp,
+			    Tcl_Obj *objv[], int objc, CmdFrame *cf);
+MODULE_SCOPE void	TclArgumentRelease(Tcl_Interp *interp,
+			    Tcl_Obj *objv[], int objc);
+MODULE_SCOPE void	TclArgumentBCEnter(Tcl_Interp *interp,
+			    Tcl_Obj *objv[], int objc,
+			    void *codePtr, CmdFrame *cfPtr, int cmd, int pc);
+MODULE_SCOPE void	TclArgumentBCRelease(Tcl_Interp *interp,
+			    CmdFrame *cfPtr);
+MODULE_SCOPE void	TclArgumentGet(Tcl_Interp *interp, Tcl_Obj *obj,
+			    CmdFrame **cfPtrPtr, int *wordPtr);
+MODULE_SCOPE int	TclArraySet(Tcl_Interp *interp,
+			    Tcl_Obj *arrayNameObj, Tcl_Obj *arrayElemObj);
+MODULE_SCOPE double	TclBignumToDouble(const mp_int *bignum);
+MODULE_SCOPE int	TclByteArrayMatch(const unsigned char *string,
+			    int strLen, const unsigned char *pattern,
+			    int ptnLen, int flags);
+MODULE_SCOPE double	TclCeil(const mp_int *a);
+MODULE_SCOPE void	TclChannelPreserve(Tcl_Channel chan);
+MODULE_SCOPE void	TclChannelRelease(Tcl_Channel chan);
+MODULE_SCOPE int	TclCheckBadOctal(Tcl_Interp *interp,
+			    const char *value);
+MODULE_SCOPE int	TclChanCaughtErrorBypass(Tcl_Interp *interp,
+			    Tcl_Channel chan);
+MODULE_SCOPE Tcl_ObjCmdProc TclChannelNamesCmd;
+MODULE_SCOPE Tcl_NRPostProc TclClearRootEnsemble;
+MODULE_SCOPE ContLineLoc *TclContinuationsEnter(Tcl_Obj *objPtr, int num,
+			    int *loc);
+MODULE_SCOPE void	TclContinuationsEnterDerived(Tcl_Obj *objPtr,
+			    int start, int *clNext);
+MODULE_SCOPE ContLineLoc *TclContinuationsGet(Tcl_Obj *objPtr);
+MODULE_SCOPE void	TclContinuationsCopy(Tcl_Obj *objPtr,
+			    Tcl_Obj *originObjPtr);
+MODULE_SCOPE int	TclConvertElement(const char *src, int length,
+			    char *dst, int flags);
+MODULE_SCOPE void	TclDeleteNamespaceVars(Namespace *nsPtr);
+MODULE_SCOPE int	TclFindDictElement(Tcl_Interp *interp,
+			    const char *dict, int dictLength,
+			    const char **elementPtr, const char **nextPtr,
+			    int *sizePtr, int *literalPtr);
+/* TIP #280 - Modified token based evulation, with line information. */
+MODULE_SCOPE int	TclEvalEx(Tcl_Interp *interp, const char *script,
+			    int numBytes, int flags, int line,
+			    int *clNextOuter, const char *outerScript);
+MODULE_SCOPE Tcl_ObjCmdProc TclFileAttrsCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileCopyCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileDeleteCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileLinkCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileMakeDirsCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileReadLinkCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileRenameCmd;
+MODULE_SCOPE Tcl_ObjCmdProc TclFileTemporaryCmd;
+MODULE_SCOPE void	TclCreateLateExitHandler(Tcl_ExitProc *proc,
+			    ClientData clientData);
+MODULE_SCOPE void	TclDeleteLateExitHandler(Tcl_ExitProc *proc,
+			    ClientData clientData);
+MODULE_SCOPE char *	TclDStringAppendObj(Tcl_DString *dsPtr,
+			    Tcl_Obj *objPtr);
+MODULE_SCOPE char *	TclDStringAppendDString(Tcl_DString *dsPtr,
+			    Tcl_DString *toAppendPtr);
+MODULE_SCOPE Tcl_Obj *	TclDStringToObj(Tcl_DString *dsPtr);
+MODULE_SCOPE Tcl_Obj *const *	TclFetchEnsembleRoot(Tcl_Interp *interp,
+			    Tcl_Obj *const *objv, int objc, int *objcPtr);
+MODULE_SCOPE void	TclFinalizeAllocSubsystem(void);
+MODULE_SCOPE void	TclFinalizeAsync(void);
+MODULE_SCOPE void	TclFinalizeDoubleConversion(void);
+MODULE_SCOPE void	TclFinalizeEncodingSubsystem(void);
+MODULE_SCOPE void	TclFinalizeEnvironment(void);
+MODULE_SCOPE void	TclFinalizeEvaluation(void);
+MODULE_SCOPE void	TclFinalizeExecution(void);
+MODULE_SCOPE void	TclFinalizeIOSubsystem(void);
+MODULE_SCOPE void	TclFinalizeFilesystem(void);
+MODULE_SCOPE void	TclResetFilesystem(void);
+MODULE_SCOPE void	TclFinalizeLoad(void);
+MODULE_SCOPE void	TclFinalizeLock(void);
+MODULE_SCOPE void	TclFinalizeMemorySubsystem(void);
+MODULE_SCOPE void	TclFinalizeNotifier(void);
+MODULE_SCOPE void	TclFinalizeObjects(void);
+MODULE_SCOPE void	TclFinalizePreserve(void);
+MODULE_SCOPE void	TclFinalizeSynchronization(void);
+MODULE_SCOPE void	TclInitThreadAlloc(void);
+MODULE_SCOPE void	TclFinalizeThreadAlloc(void);
+MODULE_SCOPE void	TclFinalizeThreadAllocThread(void);
+MODULE_SCOPE void	TclFinalizeThreadData(int quick);
+MODULE_SCOPE void	TclFinalizeThreadObjects(void);
+MODULE_SCOPE double	TclFloor(const mp_int *a);
+MODULE_SCOPE void	TclFormatNaN(double value, char *buffer);
+MODULE_SCOPE int	TclFSFileAttrIndex(Tcl_Obj *pathPtr,
+			    const char *attributeName, int *indexPtr);
+MODULE_SCOPE int	TclNREvalFile(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    const char *encodingName);
+MODULE_SCOPE void	TclFSUnloadTempFile(Tcl_LoadHandle loadHandle);
+MODULE_SCOPE int *	TclGetAsyncReadyPtr(void);
+MODULE_SCOPE Tcl_Obj *	TclGetBgErrorHandler(Tcl_Interp *interp);
+MODULE_SCOPE int	TclGetChannelFromObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr, Tcl_Channel *chanPtr,
+			    int *modePtr, int flags);
+MODULE_SCOPE CmdFrame *	TclGetCmdFrameForProcedure(Proc *procPtr);
+MODULE_SCOPE int	TclGetCompletionCodeFromObj(Tcl_Interp *interp,
+			    Tcl_Obj *value, int *code);
+MODULE_SCOPE int	TclGetNumberFromObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr, ClientData *clientDataPtr,
+			    int *typePtr);
+MODULE_SCOPE int	TclGetOpenModeEx(Tcl_Interp *interp,
+			    const char *modeString, int *seekFlagPtr,
+			    int *binaryPtr);
+MODULE_SCOPE Tcl_Obj *	TclGetProcessGlobalValue(ProcessGlobalValue *pgvPtr);
+MODULE_SCOPE Tcl_Obj *	TclGetSourceFromFrame(CmdFrame *cfPtr, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE char *	TclGetStringStorage(Tcl_Obj *objPtr,
+			    unsigned int *sizePtr);
+MODULE_SCOPE int	TclGetLoadedPackagesEx(Tcl_Interp *interp,
+				const char *targetName,
+				const char *packageName);
+MODULE_SCOPE int	TclGlob(Tcl_Interp *interp, char *pattern,
+			    Tcl_Obj *unquotedPrefix, int globFlags,
+			    Tcl_GlobTypeData *types);
+MODULE_SCOPE int	TclIncrObj(Tcl_Interp *interp, Tcl_Obj *valuePtr,
+			    Tcl_Obj *incrPtr);
+MODULE_SCOPE Tcl_Obj *	TclIncrObjVar2(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, Tcl_Obj *incrPtr, int flags);
+MODULE_SCOPE int	TclInfoExistsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclInfoCoroutineCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Obj *	TclInfoFrame(Tcl_Interp *interp, CmdFrame *framePtr);
+MODULE_SCOPE int	TclInfoGlobalsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclInfoLocalsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclInfoVarsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE void	TclInitAlloc(void);
+MODULE_SCOPE void	TclInitDbCkalloc(void);
+MODULE_SCOPE void	TclInitDoubleConversion(void);
+MODULE_SCOPE void	TclInitEmbeddedConfigurationInformation(
+			    Tcl_Interp *interp);
+MODULE_SCOPE void	TclInitEncodingSubsystem(void);
+MODULE_SCOPE void	TclInitIOSubsystem(void);
+MODULE_SCOPE void	TclInitLimitSupport(Tcl_Interp *interp);
+MODULE_SCOPE void	TclInitNamespaceSubsystem(void);
+MODULE_SCOPE void	TclInitNotifier(void);
+MODULE_SCOPE void	TclInitObjSubsystem(void);
+MODULE_SCOPE void	TclInitSubsystems(void);
+MODULE_SCOPE int	TclInterpReady(Tcl_Interp *interp);
+MODULE_SCOPE int	TclIsSpaceProc(char byte);
+MODULE_SCOPE int	TclIsBareword(char byte);
+MODULE_SCOPE Tcl_Obj *	TclJoinPath(int elements, Tcl_Obj * const objv[]);
+MODULE_SCOPE int	TclJoinThread(Tcl_ThreadId id, int *result);
+MODULE_SCOPE void	TclLimitRemoveAllHandlers(Tcl_Interp *interp);
+MODULE_SCOPE Tcl_Obj *	TclLindexList(Tcl_Interp *interp,
+			    Tcl_Obj *listPtr, Tcl_Obj *argPtr);
+MODULE_SCOPE Tcl_Obj *	TclLindexFlat(Tcl_Interp *interp, Tcl_Obj *listPtr,
+			    int indexCount, Tcl_Obj *const indexArray[]);
+/* TIP #280 */
+MODULE_SCOPE void	TclListLines(Tcl_Obj *listObj, int line, int n,
+			    int *lines, Tcl_Obj *const *elems);
+MODULE_SCOPE Tcl_Obj *	TclListObjCopy(Tcl_Interp *interp, Tcl_Obj *listPtr);
+MODULE_SCOPE Tcl_Obj *	TclLsetList(Tcl_Interp *interp, Tcl_Obj *listPtr,
+			    Tcl_Obj *indexPtr, Tcl_Obj *valuePtr);
+MODULE_SCOPE Tcl_Obj *	TclLsetFlat(Tcl_Interp *interp, Tcl_Obj *listPtr,
+			    int indexCount, Tcl_Obj *const indexArray[],
+			    Tcl_Obj *valuePtr);
+MODULE_SCOPE Tcl_Command TclMakeEnsemble(Tcl_Interp *interp, const char *name,
+			    const EnsembleImplMap map[]);
+MODULE_SCOPE int	TclMaxListLength(const char *bytes, int numBytes,
+			    const char **endPtr);
+MODULE_SCOPE int	TclMergeReturnOptions(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[], Tcl_Obj **optionsPtrPtr,
+			    int *codePtr, int *levelPtr);
+MODULE_SCOPE Tcl_Obj *  TclNoErrorStack(Tcl_Interp *interp, Tcl_Obj *options);
+MODULE_SCOPE int	TclNokia770Doubles(void);
+MODULE_SCOPE void	TclNsDecrRefCount(Namespace *nsPtr);
+MODULE_SCOPE void	TclObjVarErrMsg(Tcl_Interp *interp, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, const char *operation,
+			    const char *reason, int index);
+MODULE_SCOPE int	TclObjInvokeNamespace(Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[],
+			    Tcl_Namespace *nsPtr, int flags);
+MODULE_SCOPE int	TclObjUnsetVar2(Tcl_Interp *interp,
+			    Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, int flags);
+MODULE_SCOPE int	TclParseBackslash(const char *src,
+			    int numBytes, int *readPtr, char *dst);
+MODULE_SCOPE int	TclParseHex(const char *src, int numBytes,
+			    int *resultPtr);
+MODULE_SCOPE int	TclParseNumber(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    const char *expected, const char *bytes,
+			    int numBytes, const char **endPtrPtr, int flags);
+MODULE_SCOPE void	TclParseInit(Tcl_Interp *interp, const char *string,
+			    int numBytes, Tcl_Parse *parsePtr);
+MODULE_SCOPE int	TclParseAllWhiteSpace(const char *src, int numBytes);
+MODULE_SCOPE int	TclProcessReturn(Tcl_Interp *interp,
+			    int code, int level, Tcl_Obj *returnOpts);
+MODULE_SCOPE int	TclpObjLstat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+MODULE_SCOPE Tcl_Obj *	TclpTempFileName(void);
+MODULE_SCOPE Tcl_Obj *  TclpTempFileNameForLibrary(Tcl_Interp *interp, Tcl_Obj* pathPtr);
+MODULE_SCOPE Tcl_Obj *	TclNewFSPathObj(Tcl_Obj *dirPtr, const char *addStrRep,
+			    int len);
+MODULE_SCOPE int	TclpDeleteFile(const void *path);
+MODULE_SCOPE void	TclpFinalizeCondition(Tcl_Condition *condPtr);
+MODULE_SCOPE void	TclpFinalizeMutex(Tcl_Mutex *mutexPtr);
+MODULE_SCOPE void	TclpFinalizePipes(void);
+MODULE_SCOPE void	TclpFinalizeSockets(void);
+MODULE_SCOPE int	TclCreateSocketAddress(Tcl_Interp *interp,
+			    struct addrinfo **addrlist,
+			    const char *host, int port, int willBind,
+			    const char **errorMsgPtr);
+MODULE_SCOPE int	TclpThreadCreate(Tcl_ThreadId *idPtr,
+			    Tcl_ThreadCreateProc *proc, ClientData clientData,
+			    int stackSize, int flags);
+MODULE_SCOPE int	TclpFindVariable(const char *name, int *lengthPtr);
+MODULE_SCOPE void	TclpInitLibraryPath(char **valuePtr,
+			    size_t *lengthPtr, Tcl_Encoding *encodingPtr);
+MODULE_SCOPE void	TclpInitLock(void);
+MODULE_SCOPE void	TclpInitPlatform(void);
+MODULE_SCOPE void	TclpInitUnlock(void);
+MODULE_SCOPE Tcl_Obj *	TclpObjListVolumes(void);
+MODULE_SCOPE void	TclpMasterLock(void);
+MODULE_SCOPE void	TclpMasterUnlock(void);
+MODULE_SCOPE int	TclpMatchFiles(Tcl_Interp *interp, char *separators,
+			    Tcl_DString *dirPtr, char *pattern, char *tail);
+MODULE_SCOPE int	TclpObjNormalizePath(Tcl_Interp *interp,
+			    Tcl_Obj *pathPtr, int nextCheckpoint);
+MODULE_SCOPE void	TclpNativeJoinPath(Tcl_Obj *prefix, const char *joining);
+MODULE_SCOPE Tcl_Obj *	TclpNativeSplitPath(Tcl_Obj *pathPtr, int *lenPtr);
+MODULE_SCOPE Tcl_PathType TclpGetNativePathType(Tcl_Obj *pathPtr,
+			    int *driveNameLengthPtr, Tcl_Obj **driveNameRef);
+MODULE_SCOPE int	TclCrossFilesystemCopy(Tcl_Interp *interp,
+			    Tcl_Obj *source, Tcl_Obj *target);
+MODULE_SCOPE int	TclpMatchInDirectory(Tcl_Interp *interp,
+			    Tcl_Obj *resultPtr, Tcl_Obj *pathPtr,
+			    const char *pattern, Tcl_GlobTypeData *types);
+MODULE_SCOPE ClientData	TclpGetNativeCwd(ClientData clientData);
+MODULE_SCOPE Tcl_FSDupInternalRepProc TclNativeDupInternalRep;
+MODULE_SCOPE Tcl_Obj *	TclpObjLink(Tcl_Obj *pathPtr, Tcl_Obj *toPtr,
+			    int linkType);
+MODULE_SCOPE int	TclpObjChdir(Tcl_Obj *pathPtr);
+MODULE_SCOPE Tcl_Channel TclpOpenTemporaryFile(Tcl_Obj *dirObj,
+			    Tcl_Obj *basenameObj, Tcl_Obj *extensionObj,
+			    Tcl_Obj *resultingNameObj);
+MODULE_SCOPE void TclPkgFileSeen(Tcl_Interp *interp, const char *fileName);
+MODULE_SCOPE void *TclInitPkgFiles(Tcl_Interp *interp);
+MODULE_SCOPE Tcl_Obj *	TclPathPart(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    Tcl_PathPart portion);
+MODULE_SCOPE char *	TclpReadlink(const char *fileName,
+			    Tcl_DString *linkPtr);
+MODULE_SCOPE void	TclpSetInterfaces(void);
+MODULE_SCOPE void	TclpSetVariables(Tcl_Interp *interp);
+MODULE_SCOPE void *	TclThreadStorageKeyGet(Tcl_ThreadDataKey *keyPtr);
+MODULE_SCOPE void	TclThreadStorageKeySet(Tcl_ThreadDataKey *keyPtr,
+			    void *data);
+MODULE_SCOPE TCL_NORETURN void TclpThreadExit(int status);
+MODULE_SCOPE void	TclRememberCondition(Tcl_Condition *mutex);
+MODULE_SCOPE void	TclRememberJoinableThread(Tcl_ThreadId id);
+MODULE_SCOPE void	TclRememberMutex(Tcl_Mutex *mutex);
+MODULE_SCOPE void	TclRemoveScriptLimitCallbacks(Tcl_Interp *interp);
+MODULE_SCOPE int	TclReToGlob(Tcl_Interp *interp, const char *reStr,
+			    int reStrLen, Tcl_DString *dsPtr, int *flagsPtr,
+			    int *quantifiersFoundPtr);
+MODULE_SCOPE int	TclScanElement(const char *string, int length,
+			    int *flagPtr);
+MODULE_SCOPE void	TclSetBgErrorHandler(Tcl_Interp *interp,
+			    Tcl_Obj *cmdPrefix);
+MODULE_SCOPE void	TclSetBignumIntRep(Tcl_Obj *objPtr,
+			    mp_int *bignumValue);
+MODULE_SCOPE int	TclSetBooleanFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+MODULE_SCOPE void	TclSetCmdNameObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    Command *cmdPtr);
+MODULE_SCOPE void	TclSetDuplicateObj(Tcl_Obj *dupPtr, Tcl_Obj *objPtr);
+MODULE_SCOPE void	TclSetProcessGlobalValue(ProcessGlobalValue *pgvPtr,
+			    Tcl_Obj *newValue, Tcl_Encoding encoding);
+MODULE_SCOPE void	TclSignalExitThread(Tcl_ThreadId id, int result);
+MODULE_SCOPE void	TclSpellFix(Tcl_Interp *interp,
+			    Tcl_Obj *const *objv, int objc, int subIdx,
+			    Tcl_Obj *bad, Tcl_Obj *fix);
+MODULE_SCOPE void *	TclStackRealloc(Tcl_Interp *interp, void *ptr,
+			    int numBytes);
+MODULE_SCOPE int	TclStringCatObjv(Tcl_Interp *interp, int inPlace,
+			    int objc, Tcl_Obj *const objv[],
+			    Tcl_Obj **objPtrPtr);
+MODULE_SCOPE int	TclStringFind(Tcl_Obj *needle, Tcl_Obj *haystack,
+			    int start);
+MODULE_SCOPE int	TclStringLast(Tcl_Obj *needle, Tcl_Obj *haystack,
+			    int last);
+MODULE_SCOPE int	TclStringMatch(const char *str, int strLen,
+			    const char *pattern, int ptnLen, int flags);
+MODULE_SCOPE int	TclStringMatchObj(Tcl_Obj *stringObj,
+			    Tcl_Obj *patternObj, int flags);
+MODULE_SCOPE Tcl_Obj *	TclStringObjReverse(Tcl_Obj *objPtr);
+MODULE_SCOPE int	TclStringRepeat(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    int count, Tcl_Obj **objPtrPtr);
+MODULE_SCOPE void	TclSubstCompile(Tcl_Interp *interp, const char *bytes,
+			    int numBytes, int flags, int line,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclSubstOptions(Tcl_Interp *interp, int numOpts,
+			    Tcl_Obj *const opts[], int *flagPtr);
+MODULE_SCOPE void	TclSubstParse(Tcl_Interp *interp, const char *bytes,
+			    int numBytes, int flags, Tcl_Parse *parsePtr,
+			    Tcl_InterpState *statePtr);
+MODULE_SCOPE int	TclSubstTokens(Tcl_Interp *interp, Tcl_Token *tokenPtr,
+			    int count, int *tokensLeftPtr, int line,
+			    int *clNextOuter, const char *outerScript);
+MODULE_SCOPE int	TclTrimLeft(const char *bytes, int numBytes,
+			    const char *trim, int numTrim);
+MODULE_SCOPE int	TclTrimRight(const char *bytes, int numBytes,
+			    const char *trim, int numTrim);
+MODULE_SCOPE int	TclUtfCasecmp(const char *cs, const char *ct);
+MODULE_SCOPE int	TclUtfCount(int ch);
+MODULE_SCOPE Tcl_Obj *	TclpNativeToNormalized(ClientData clientData);
+MODULE_SCOPE Tcl_Obj *	TclpFilesystemPathType(Tcl_Obj *pathPtr);
+MODULE_SCOPE int	TclpDlopen(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    Tcl_LoadHandle *loadHandle,
+			    Tcl_FSUnloadFileProc **unloadProcPtr, int flags);
+MODULE_SCOPE int	TclpUtime(Tcl_Obj *pathPtr, struct utimbuf *tval);
+#ifdef TCL_LOAD_FROM_MEMORY
+MODULE_SCOPE void *	TclpLoadMemoryGetBuffer(Tcl_Interp *interp, int size);
+MODULE_SCOPE int	TclpLoadMemory(Tcl_Interp *interp, void *buffer,
+			    int size, int codeSize, Tcl_LoadHandle *loadHandle,
+			    Tcl_FSUnloadFileProc **unloadProcPtr, int flags);
+#endif
+MODULE_SCOPE void	TclInitThreadStorage(void);
+MODULE_SCOPE void	TclFinalizeThreadDataThread(void);
+MODULE_SCOPE void	TclFinalizeThreadStorage(void);
+#ifdef TCL_WIDE_CLICKS
+MODULE_SCOPE Tcl_WideInt TclpGetWideClicks(void);
+MODULE_SCOPE double	TclpWideClicksToNanoseconds(Tcl_WideInt clicks);
+#endif
+MODULE_SCOPE int	TclZlibInit(Tcl_Interp *interp);
+MODULE_SCOPE void *	TclpThreadCreateKey(void);
+MODULE_SCOPE void	TclpThreadDeleteKey(void *keyPtr);
+MODULE_SCOPE void	TclpThreadSetMasterTSD(void *tsdKeyPtr, void *ptr);
+MODULE_SCOPE void *	TclpThreadGetMasterTSD(void *tsdKeyPtr);
+
+MODULE_SCOPE void	TclErrorStackResetIf(Tcl_Interp *interp, const char *msg, int length);
+
+/*
+ *----------------------------------------------------------------
+ * Command procedures in the generic core:
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE int	Tcl_AfterObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_AppendObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ApplyObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitArrayCmd(Tcl_Interp *interp);
+MODULE_SCOPE Tcl_Command TclInitBinaryCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_BreakObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#ifndef TCL_NO_DEPRECATED
+MODULE_SCOPE int	Tcl_CaseObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#endif
+MODULE_SCOPE int	Tcl_CatchObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_CdObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitChanCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	TclChanCreateObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclChanPostEventObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclChanPopObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclChanPushObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE void	TclClockInit(Tcl_Interp *interp);
+MODULE_SCOPE int	TclClockOldscanObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_CloseObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ConcatObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ContinueObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_TimerToken TclCreateAbsoluteTimerHandler(
+			    Tcl_Time *timePtr, Tcl_TimerProc *proc,
+			    ClientData clientData);
+MODULE_SCOPE int	TclDefaultBgErrorHandlerObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitDictCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	TclDictWithFinish(Tcl_Interp *interp, Var *varPtr,
+			    Var *arrayPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, int index, int pathc,
+			    Tcl_Obj *const pathv[], Tcl_Obj *keysPtr);
+MODULE_SCOPE Tcl_Obj *	TclDictWithInit(Tcl_Interp *interp, Tcl_Obj *dictPtr,
+			    int pathc, Tcl_Obj *const pathv[]);
+MODULE_SCOPE int	Tcl_DisassembleObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+
+/* Assemble command function */
+MODULE_SCOPE int	Tcl_AssembleObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclNRAssembleObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitEncodingCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	TclMakeEncodingCommandSafe(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_EofObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ErrorObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_EvalObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ExecObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ExitObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ExprObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_FblockedObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_FconfigureObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_FcopyObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitFileCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	TclMakeFileCommandSafe(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_FileEventObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_FlushObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ForObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ForeachObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_FormatObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_GetsObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_GlobalObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_GlobObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_IfObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_IncrObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitInfoCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_InterpObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int argc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_JoinObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LappendObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LassignObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LindexObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LinsertObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LlengthObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ListObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LmapObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LoadObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LrangeObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LrepeatObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LreplaceObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LreverseObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LsearchObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LsetObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_LsortObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitNamespaceCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	TclNamespaceEnsembleCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_OpenObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_PackageObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_PidObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitPrefixCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_PutsObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_PwdObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ReadObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_RegexpObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_RegsubObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_RenameObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_RepresentationCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ReturnObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ScanObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SeekObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SetObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SplitObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SocketObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SourceObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE Tcl_Command TclInitStringCmd(Tcl_Interp *interp);
+MODULE_SCOPE int	Tcl_SubstObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_SwitchObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_TellObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_ThrowObjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_TimeObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_TraceObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_TryObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_UnloadObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_UnsetObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_UpdateObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_UplevelObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_UpvarObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_VariableObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_VwaitObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	Tcl_WhileObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+
+/*
+ *----------------------------------------------------------------
+ * Compilation procedures for commands in the generic core:
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE int	TclCompileAppendCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileArrayExistsCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileArraySetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileArrayUnsetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBreakCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileCatchCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileClockClicksCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileClockReadingCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileConcatCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileContinueCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictAppendCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictCreateCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictExistsCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictForCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictGetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictIncrCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictLappendCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictMapCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictMergeCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictSetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictUnsetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictUpdateCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileDictWithCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileEnsemble(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileErrorCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileExprCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileForCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileForeachCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileFormatCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileGlobalCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileIfCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoCommandsCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoCoroutineCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoExistsCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoLevelCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoObjectClassCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoObjectIsACmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileInfoObjectNamespaceCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileIncrCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLappendCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLassignCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLindexCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLinsertCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileListCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLlengthCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLmapCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLrangeCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLreplaceCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileLsetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceCodeCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceCurrentCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceOriginCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceQualifiersCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceTailCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceUpvarCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNamespaceWhichCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileNoOp(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileObjectNextCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileObjectNextToCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileObjectSelfCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileRegexpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileRegsubCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileReturnCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileSetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringCatCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringCmpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringEqualCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringFirstCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringIndexCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringIsCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringLastCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringLenCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringMapCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringMatchCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringRangeCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringReplaceCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringToLowerCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringToTitleCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringToUpperCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringTrimCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringTrimLCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileStringTrimRCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileSubstCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileSwitchCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileTailcallCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileThrowCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileTryCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileUnsetCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileUpvarCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileVariableCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileWhileCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileYieldCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileYieldToCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic0ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic1ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic2ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic3ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic0Or1ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic1Or2ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic2Or3ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic0To2ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasic1To3ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasicMin0ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasicMin1ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclCompileBasicMin2ArgCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+
+MODULE_SCOPE int	TclInvertOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileInvertOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclNotOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileNotOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclAddOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileAddOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclMulOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileMulOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclAndOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileAndOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclOrOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileOrOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclXorOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileXorOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclPowOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompilePowOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclLshiftOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileLshiftOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclRshiftOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileRshiftOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclModOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileModOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclNeqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileNeqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclStrneqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileStrneqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclInOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileInOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclNiOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileNiOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclMinusOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileMinusOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclDivOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileDivOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclLessOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileLessOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclLeqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileLeqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclGreaterOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileGreaterOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclGeqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileGeqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclEqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileEqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+MODULE_SCOPE int	TclStreqOpCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclCompileStreqOpCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+
+MODULE_SCOPE int	TclCompileAssembleCmd(Tcl_Interp *interp,
+			    Tcl_Parse *parsePtr, Command *cmdPtr,
+			    struct CompileEnv *envPtr);
+
+/*
+ * Functions defined in generic/tclVar.c and currently exported only for use
+ * by the bytecode compiler and engine. Some of these could later be placed in
+ * the public interface.
+ */
+
+MODULE_SCOPE Var *	TclObjLookupVarEx(Tcl_Interp * interp,
+			    Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, int flags,
+			    const char *msg, const int createPart1,
+			    const int createPart2, Var **arrayPtrPtr);
+MODULE_SCOPE Var *	TclLookupArrayElement(Tcl_Interp *interp,
+			    Tcl_Obj *arrayNamePtr, Tcl_Obj *elNamePtr,
+			    const int flags, const char *msg,
+			    const int createPart1, const int createPart2,
+			    Var *arrayPtr, int index);
+MODULE_SCOPE Tcl_Obj *	TclPtrGetVarIdx(Tcl_Interp *interp,
+			    Var *varPtr, Var *arrayPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, const int flags, int index);
+MODULE_SCOPE Tcl_Obj *	TclPtrSetVarIdx(Tcl_Interp *interp,
+			    Var *varPtr, Var *arrayPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr,
+			    const int flags, int index);
+MODULE_SCOPE Tcl_Obj *	TclPtrIncrObjVarIdx(Tcl_Interp *interp,
+			    Var *varPtr, Var *arrayPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, Tcl_Obj *incrPtr,
+			    const int flags, int index);
+MODULE_SCOPE int	TclPtrObjMakeUpvarIdx(Tcl_Interp *interp,
+			    Var *otherPtr, Tcl_Obj *myNamePtr, int myFlags,
+			    int index);
+MODULE_SCOPE int	TclPtrUnsetVarIdx(Tcl_Interp *interp, Var *varPtr,
+			    Var *arrayPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, const int flags,
+			    int index);
+MODULE_SCOPE void	TclInvalidateNsPath(Namespace *nsPtr);
+MODULE_SCOPE void	TclFindArrayPtrElements(Var *arrayPtr,
+			    Tcl_HashTable *tablePtr);
+
+/*
+ * The new extended interface to the variable traces.
+ */
+
+MODULE_SCOPE int	TclObjCallVarTraces(Interp *iPtr, Var *arrayPtr,
+			    Var *varPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr,
+			    int flags, int leaveErrMsg, int index);
+
+/*
+ * So tclObj.c and tclDictObj.c can share these implementations.
+ */
+
+MODULE_SCOPE int	TclCompareObjKeys(void *keyPtr, Tcl_HashEntry *hPtr);
+MODULE_SCOPE void	TclFreeObjEntry(Tcl_HashEntry *hPtr);
+MODULE_SCOPE TCL_HASH_TYPE TclHashObjKey(Tcl_HashTable *tablePtr, void *keyPtr);
+
+MODULE_SCOPE int	TclFullFinalizationRequested(void);
+
+/*
+ *----------------------------------------------------------------
+ * Macros used by the Tcl core to create and release Tcl objects.
+ * TclNewObj(objPtr) creates a new object denoting an empty string.
+ * TclDecrRefCount(objPtr) decrements the object's reference count, and frees
+ * the object if its reference count is zero. These macros are inline versions
+ * of Tcl_NewObj() and Tcl_DecrRefCount(). Notice that the names differ in not
+ * having a "_" after the "Tcl". Notice also that these macros reference their
+ * argument more than once, so you should avoid calling them with an
+ * expression that is expensive to compute or has side effects. The ANSI C
+ * "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE void	TclNewObj(Tcl_Obj *objPtr);
+ * MODULE_SCOPE void	TclDecrRefCount(Tcl_Obj *objPtr);
+ *
+ * These macros are defined in terms of two macros that depend on memory
+ * allocator in use: TclAllocObjStorage, TclFreeObjStorage. They are defined
+ * below.
+ *----------------------------------------------------------------
+ */
+
+/*
+ * DTrace object allocation probe macros.
+ */
+
+#ifdef USE_DTRACE
+#ifndef _TCLDTRACE_H
+typedef const char *TclDTraceStr;
+#include "tclDTrace.h"
+#endif
+#define	TCL_DTRACE_OBJ_CREATE(objPtr)	TCL_OBJ_CREATE(objPtr)
+#define	TCL_DTRACE_OBJ_FREE(objPtr)	TCL_OBJ_FREE(objPtr)
+#else /* USE_DTRACE */
+#define	TCL_DTRACE_OBJ_CREATE(objPtr)	{}
+#define	TCL_DTRACE_OBJ_FREE(objPtr)	{}
+#endif /* USE_DTRACE */
+
+#ifdef TCL_COMPILE_STATS
+#  define TclIncrObjsAllocated() \
+    tclObjsAlloced++
+#  define TclIncrObjsFreed() \
+    tclObjsFreed++
+#else
+#  define TclIncrObjsAllocated()
+#  define TclIncrObjsFreed()
+#endif /* TCL_COMPILE_STATS */
+
+#  define TclAllocObjStorage(objPtr)		\
+	TclAllocObjStorageEx(NULL, (objPtr))
+
+#  define TclFreeObjStorage(objPtr)		\
+	TclFreeObjStorageEx(NULL, (objPtr))
+
+#ifndef TCL_MEM_DEBUG
+# define TclNewObj(objPtr) \
+    TclIncrObjsAllocated(); \
+    TclAllocObjStorage(objPtr); \
+    (objPtr)->refCount = 0; \
+    (objPtr)->bytes    = &tclEmptyString; \
+    (objPtr)->length   = 0; \
+    (objPtr)->typePtr  = NULL; \
+    TCL_DTRACE_OBJ_CREATE(objPtr)
+
+/*
+ * Invalidate the string rep first so we can use the bytes value for our
+ * pointer chain, and signal an obj deletion (as opposed to shimmering) with
+ * 'length == -1'.
+ * Use empty 'if ; else' to handle use in unbraced outer if/else conditions.
+ */
+
+# define TclDecrRefCount(objPtr) \
+    if ((objPtr)->refCount-- > 1) ; else { \
+	if (!(objPtr)->typePtr || !(objPtr)->typePtr->freeIntRepProc) { \
+	    TCL_DTRACE_OBJ_FREE(objPtr); \
+	    if ((objPtr)->bytes \
+		    && ((objPtr)->bytes != &tclEmptyString)) { \
+		ckfree((objPtr)->bytes); \
+	    } \
+	    (objPtr)->length = -1; \
+	    TclFreeObjStorage(objPtr); \
+	    TclIncrObjsFreed(); \
+	} else { \
+	    TclFreeObj(objPtr); \
+	} \
+    }
+
+#if defined(PURIFY)
+
+/*
+ * The PURIFY mode is like the regular mode, but instead of doing block
+ * Tcl_Obj allocation and keeping a freed list for efficiency, it always
+ * allocates and frees a single Tcl_Obj so that tools like Purify can better
+ * track memory leaks.
+ */
+
+#  define TclAllocObjStorageEx(interp, objPtr) \
+	(objPtr) = (Tcl_Obj *) ckalloc(sizeof(Tcl_Obj))
+
+#  define TclFreeObjStorageEx(interp, objPtr) \
+	ckfree(objPtr)
+
+#undef USE_THREAD_ALLOC
+#undef USE_TCLALLOC
+#elif defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+
+/*
+ * The TCL_THREADS mode is like the regular mode but allocates Tcl_Obj's from
+ * per-thread caches.
+ */
+
+MODULE_SCOPE Tcl_Obj *	TclThreadAllocObj(void);
+MODULE_SCOPE void	TclThreadFreeObj(Tcl_Obj *);
+MODULE_SCOPE Tcl_Mutex *TclpNewAllocMutex(void);
+MODULE_SCOPE void	TclFreeAllocCache(void *);
+MODULE_SCOPE void *	TclpGetAllocCache(void);
+MODULE_SCOPE void	TclpSetAllocCache(void *);
+MODULE_SCOPE void	TclpFreeAllocMutex(Tcl_Mutex *mutex);
+MODULE_SCOPE void	TclpInitAllocCache(void);
+MODULE_SCOPE void	TclpFreeAllocCache(void *);
+
+/*
+ * These macros need to be kept in sync with the code of TclThreadAllocObj()
+ * and TclThreadFreeObj().
+ *
+ * Note that the optimiser should resolve the case (interp==NULL) at compile
+ * time.
+ */
+
+#  define ALLOC_NOBJHIGH 1200
+
+#  define TclAllocObjStorageEx(interp, objPtr)				\
+    do {								\
+	AllocCache *cachePtr;						\
+	if (((interp) == NULL) ||					\
+		((cachePtr = ((Interp *)(interp))->allocCache),		\
+			(cachePtr->numObjects == 0))) {			\
+	    (objPtr) = TclThreadAllocObj();				\
+	} else {							\
+	    (objPtr) = cachePtr->firstObjPtr;				\
+	    cachePtr->firstObjPtr = (objPtr)->internalRep.twoPtrValue.ptr1; \
+	    --cachePtr->numObjects;					\
+	}								\
+    } while (0)
+
+#  define TclFreeObjStorageEx(interp, objPtr)				\
+    do {								\
+	AllocCache *cachePtr;						\
+	if (((interp) == NULL) ||					\
+		((cachePtr = ((Interp *)(interp))->allocCache),		\
+			((cachePtr->numObjects == 0) ||			\
+			(cachePtr->numObjects >= ALLOC_NOBJHIGH)))) {	\
+	    TclThreadFreeObj(objPtr);					\
+	} else {							\
+	    (objPtr)->internalRep.twoPtrValue.ptr1 = cachePtr->firstObjPtr; \
+	    cachePtr->firstObjPtr = objPtr;				\
+	    ++cachePtr->numObjects;					\
+	}								\
+    } while (0)
+
+#else /* not PURIFY or USE_THREAD_ALLOC */
+
+#if defined(USE_TCLALLOC) && USE_TCLALLOC
+    MODULE_SCOPE void TclFinalizeAllocSubsystem();
+    MODULE_SCOPE void TclInitAlloc();
+#else
+#   define USE_TCLALLOC 0
+#endif
+
+#ifdef TCL_THREADS
+/* declared in tclObj.c */
+MODULE_SCOPE Tcl_Mutex	tclObjMutex;
+#endif
+
+#  define TclAllocObjStorageEx(interp, objPtr) \
+    do {								\
+	Tcl_MutexLock(&tclObjMutex);					\
+	if (tclFreeObjList == NULL) {					\
+	    TclAllocateFreeObjects();					\
+	}								\
+	(objPtr) = tclFreeObjList;					\
+	tclFreeObjList = (Tcl_Obj *)					\
+		tclFreeObjList->internalRep.twoPtrValue.ptr1;		\
+	Tcl_MutexUnlock(&tclObjMutex);					\
+    } while (0)
+
+#  define TclFreeObjStorageEx(interp, objPtr) \
+    do {							       \
+	Tcl_MutexLock(&tclObjMutex);				       \
+	(objPtr)->internalRep.twoPtrValue.ptr1 = (void *) tclFreeObjList; \
+	tclFreeObjList = (objPtr);				       \
+	Tcl_MutexUnlock(&tclObjMutex);				       \
+    } while (0)
+#endif
+
+#else /* TCL_MEM_DEBUG */
+MODULE_SCOPE void	TclDbInitNewObj(Tcl_Obj *objPtr, const char *file,
+			    int line);
+
+# define TclDbNewObj(objPtr, file, line) \
+    do { \
+	TclIncrObjsAllocated();						\
+	(objPtr) = (Tcl_Obj *)						\
+		Tcl_DbCkalloc(sizeof(Tcl_Obj), (file), (line));		\
+	TclDbInitNewObj((objPtr), (file), (line));			\
+	TCL_DTRACE_OBJ_CREATE(objPtr);					\
+    } while (0)
+
+# define TclNewObj(objPtr) \
+    TclDbNewObj(objPtr, __FILE__, __LINE__);
+
+# define TclDecrRefCount(objPtr) \
+    Tcl_DbDecrRefCount(objPtr, __FILE__, __LINE__)
+
+# define TclNewListObjDirect(objc, objv) \
+    TclDbNewListObjDirect(objc, objv, __FILE__, __LINE__)
+
+#undef USE_THREAD_ALLOC
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to set a Tcl_Obj's string representation to a
+ * copy of the "len" bytes starting at "bytePtr". This code works even if the
+ * byte array contains NULLs as long as the length is correct. Because "len"
+ * is referenced multiple times, it should be as simple an expression as
+ * possible. The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void TclInitStringRep(Tcl_Obj *objPtr, char *bytePtr, int len);
+ *
+ * This macro should only be called on an unshared objPtr where
+ *  objPtr->typePtr->freeIntRepProc == NULL
+ *----------------------------------------------------------------
+ */
+
+#define TclInitStringRep(objPtr, bytePtr, len) \
+    if ((len) == 0) { \
+	(objPtr)->bytes	 = &tclEmptyString; \
+	(objPtr)->length = 0; \
+    } else { \
+	(objPtr)->bytes = (char *) ckalloc((unsigned) ((len) + 1)); \
+	memcpy((objPtr)->bytes, (bytePtr), (unsigned) (len)); \
+	(objPtr)->bytes[len] = '\0'; \
+	(objPtr)->length = (len); \
+    }
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to get the string representation's byte array
+ * pointer from a Tcl_Obj. This is an inline version of Tcl_GetString(). The
+ * macro's expression result is the string rep's byte pointer which might be
+ * NULL. The bytes referenced by this pointer must not be modified by the
+ * caller. The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE char *	TclGetString(Tcl_Obj *objPtr);
+ *----------------------------------------------------------------
+ */
+
+#define TclGetString(objPtr) \
+    ((objPtr)->bytes? (objPtr)->bytes : Tcl_GetString(objPtr))
+
+#define TclGetStringFromObj(objPtr, lenPtr) \
+    ((objPtr)->bytes \
+	    ? (*(lenPtr) = (objPtr)->length, (objPtr)->bytes)	\
+	    : Tcl_GetStringFromObj((objPtr), (lenPtr)))
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to clean out an object's internal
+ * representation. Does not actually reset the rep's bytes. The ANSI C
+ * "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void	TclFreeIntRep(Tcl_Obj *objPtr);
+ *----------------------------------------------------------------
+ */
+
+#define TclFreeIntRep(objPtr) \
+    if ((objPtr)->typePtr != NULL) { \
+	if ((objPtr)->typePtr->freeIntRepProc != NULL) { \
+	    (objPtr)->typePtr->freeIntRepProc(objPtr); \
+	} \
+	(objPtr)->typePtr = NULL; \
+    }
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to clean out an object's string representation.
+ * The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void	TclInvalidateStringRep(Tcl_Obj *objPtr);
+ *----------------------------------------------------------------
+ */
+
+#define TclInvalidateStringRep(objPtr) \
+    if ((objPtr)->bytes != NULL) { \
+	if ((objPtr)->bytes != &tclEmptyString) { \
+	    ckfree((objPtr)->bytes); \
+	} \
+	(objPtr)->bytes = NULL; \
+    }
+
+/*
+ *----------------------------------------------------------------
+ * Macros used by the Tcl core to grow Tcl_Token arrays. They use the same
+ * growth algorithm as used in tclStringObj.c for growing strings. The ANSI C
+ * "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void	TclGrowTokenArray(Tcl_Token *tokenPtr, int used,
+ *				int available, int append,
+ *				Tcl_Token *staticPtr);
+ * MODULE_SCOPE void	TclGrowParseTokenArray(Tcl_Parse *parsePtr,
+ *				int append);
+ *----------------------------------------------------------------
+ */
+
+/* General tuning for minimum growth in Tcl growth algorithms */
+#ifndef TCL_MIN_GROWTH
+#  ifdef TCL_GROWTH_MIN_ALLOC
+     /* Support for any legacy tuners */
+#    define TCL_MIN_GROWTH TCL_GROWTH_MIN_ALLOC
+#  else
+#    define TCL_MIN_GROWTH 1024
+#  endif
+#endif
+
+/* Token growth tuning, default to the general value. */
+#ifndef TCL_MIN_TOKEN_GROWTH
+#define TCL_MIN_TOKEN_GROWTH TCL_MIN_GROWTH/sizeof(Tcl_Token)
+#endif
+
+#define TCL_MAX_TOKENS (int)(UINT_MAX / sizeof(Tcl_Token))
+#define TclGrowTokenArray(tokenPtr, used, available, append, staticPtr)	\
+    do {								\
+	int _needed = (used) + (append);					\
+	if (_needed > TCL_MAX_TOKENS) {					\
+	    Tcl_Panic("max # of tokens for a Tcl parse (%d) exceeded",	\
+		    TCL_MAX_TOKENS);					\
+	}								\
+	if (_needed > (available)) {					\
+	    int allocated = 2 * _needed;					\
+	    Tcl_Token *oldPtr = (tokenPtr);				\
+	    Tcl_Token *newPtr;						\
+	    if (oldPtr == (staticPtr)) {				\
+		oldPtr = NULL;						\
+	    }								\
+	    if (allocated > TCL_MAX_TOKENS) {				\
+		allocated = TCL_MAX_TOKENS;				\
+	    }								\
+	    newPtr = (Tcl_Token *) attemptckrealloc((char *) oldPtr,	\
+		    (unsigned int) (allocated * sizeof(Tcl_Token)));	\
+	    if (newPtr == NULL) {					\
+		allocated = _needed + (append) + TCL_MIN_TOKEN_GROWTH;	\
+		if (allocated > TCL_MAX_TOKENS) {			\
+		    allocated = TCL_MAX_TOKENS;				\
+		}							\
+		newPtr = (Tcl_Token *) ckrealloc((char *) oldPtr,	\
+			(unsigned int) (allocated * sizeof(Tcl_Token))); \
+	    }								\
+	    (available) = allocated;					\
+	    if (oldPtr == NULL) {					\
+		memcpy(newPtr, staticPtr,				\
+			(size_t) ((used) * sizeof(Tcl_Token)));		\
+	    }								\
+	    (tokenPtr) = newPtr;					\
+	}								\
+    } while (0)
+
+#define TclGrowParseTokenArray(parsePtr, append)			\
+    TclGrowTokenArray((parsePtr)->tokenPtr, (parsePtr)->numTokens,	\
+	    (parsePtr)->tokensAvailable, (append),			\
+	    (parsePtr)->staticTokens)
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core get a unicode char from a utf string. It checks
+ * to see if we have a one-byte utf char before calling the real
+ * Tcl_UtfToUniChar, as this will save a lot of time for primarily ASCII
+ * string handling. The macro's expression result is 1 for the 1-byte case or
+ * the result of Tcl_UtfToUniChar. The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE int	TclUtfToUniChar(const char *string, Tcl_UniChar *ch);
+ *----------------------------------------------------------------
+ */
+
+#define TclUtfToUniChar(str, chPtr) \
+	((((unsigned char) *(str)) < 0xC0) ?		\
+	    ((*(chPtr) = (unsigned char) *(str)), 1)	\
+	    : Tcl_UtfToUniChar(str, chPtr))
+
+/*
+ *----------------------------------------------------------------
+ * Macro counterpart of the Tcl_NumUtfChars() function. To be used in speed-
+ * -sensitive points where it pays to avoid a function call in the common case
+ * of counting along a string of all one-byte characters.  The ANSI C
+ * "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void	TclNumUtfChars(int numChars, const char *bytes,
+ *				int numBytes);
+ *----------------------------------------------------------------
+ */
+
+#define TclNumUtfChars(numChars, bytes, numBytes) \
+    do { \
+	int _count, _i = (numBytes); \
+	unsigned char *_str = (unsigned char *) (bytes); \
+	while (_i && (*_str < 0xC0)) { _i--; _str++; } \
+	_count = (numBytes) - _i; \
+	if (_i) { \
+	    _count += Tcl_NumUtfChars((bytes) + _count, _i); \
+	} \
+	(numChars) = _count; \
+    } while (0);
+
+/*
+ *----------------------------------------------------------------
+ * Macro that encapsulates the logic that determines when it is safe to
+ * interpret a string as a byte array directly. In summary, the object must be
+ * a byte array and must not have a string representation (as the operations
+ * that it is used in are defined on strings, not byte arrays). Theoretically
+ * it is possible to also be efficient in the case where the object's bytes
+ * field is filled by generation from the byte array (c.f. list canonicality)
+ * but we don't do that at the moment since this is purely about efficiency.
+ * The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE int	TclIsPureByteArray(Tcl_Obj *objPtr);
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE int	TclIsPureByteArray(Tcl_Obj *objPtr);
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to compare Unicode strings. On big-endian
+ * systems we can use the more efficient memcmp, but this would not be
+ * lexically correct on little-endian systems. The ANSI C "prototype" for
+ * this macro is:
+ *
+ * MODULE_SCOPE int	TclUniCharNcmp(const Tcl_UniChar *cs,
+ *			    const Tcl_UniChar *ct, unsigned long n);
+ *----------------------------------------------------------------
+ */
+
+#ifdef WORDS_BIGENDIAN
+#   define TclUniCharNcmp(cs,ct,n) memcmp((cs),(ct),(n)*sizeof(Tcl_UniChar))
+#else /* !WORDS_BIGENDIAN */
+#   define TclUniCharNcmp Tcl_UniCharNcmp
+#endif /* WORDS_BIGENDIAN */
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to increment a namespace's export export epoch
+ * counter. The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE void	TclInvalidateNsCmdLookup(Namespace *nsPtr);
+ *----------------------------------------------------------------
+ */
+
+#define TclInvalidateNsCmdLookup(nsPtr) \
+    if ((nsPtr)->numExportPatterns) {		\
+	(nsPtr)->exportLookupEpoch++;		\
+    }						\
+    if ((nsPtr)->commandPathLength) {		\
+	(nsPtr)->cmdRefEpoch++;			\
+    }
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Core procedure added to libtommath for bignum manipulation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+MODULE_SCOPE Tcl_PackageInitProc TclTommath_Init;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * External (platform specific) initialization routine, these declarations
+ * explicitly don't use EXTERN since this code does not get compiled into the
+ * library:
+ *
+ *----------------------------------------------------------------------
+ */
+
+MODULE_SCOPE Tcl_PackageInitProc TclplatformtestInit;
+MODULE_SCOPE Tcl_PackageInitProc TclObjTest_Init;
+MODULE_SCOPE Tcl_PackageInitProc TclThread_Init;
+MODULE_SCOPE Tcl_PackageInitProc Procbodytest_Init;
+MODULE_SCOPE Tcl_PackageInitProc Procbodytest_SafeInit;
+
+/*
+ *----------------------------------------------------------------
+ * Macro used by the Tcl core to check whether a pattern has any characters
+ * special to [string match]. The ANSI C "prototype" for this macro is:
+ *
+ * MODULE_SCOPE int	TclMatchIsTrivial(const char *pattern);
+ *----------------------------------------------------------------
+ */
+
+#define TclMatchIsTrivial(pattern) \
+    (strpbrk((pattern), "*[?\\") == NULL)
+
+/*
+ *----------------------------------------------------------------
+ * Macros used by the Tcl core to set a Tcl_Obj's numeric representation
+ * avoiding the corresponding function calls in time critical parts of the
+ * core. They should only be called on unshared objects. The ANSI C
+ * "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE void	TclSetLongObj(Tcl_Obj *objPtr, long longValue);
+ * MODULE_SCOPE void	TclSetWideIntObj(Tcl_Obj *objPtr, Tcl_WideInt w);
+ * MODULE_SCOPE void	TclSetDoubleObj(Tcl_Obj *objPtr, double d);
+ *----------------------------------------------------------------
+ */
+
+#define TclSetLongObj(objPtr, i) \
+    do {						\
+	TclInvalidateStringRep(objPtr);			\
+	TclFreeIntRep(objPtr);				\
+	(objPtr)->internalRep.longValue = (long)(i);	\
+	(objPtr)->typePtr = &tclIntType;		\
+    } while (0)
+
+#ifndef TCL_WIDE_INT_IS_LONG
+#define TclSetWideIntObj(objPtr, w) \
+    do {							\
+	TclInvalidateStringRep(objPtr);				\
+	TclFreeIntRep(objPtr);					\
+	(objPtr)->internalRep.wideValue = (Tcl_WideInt)(w);	\
+	(objPtr)->typePtr = &tclWideIntType;			\
+    } while (0)
+#endif
+
+#define TclSetDoubleObj(objPtr, d) \
+    do {							\
+	TclInvalidateStringRep(objPtr);				\
+	TclFreeIntRep(objPtr);					\
+	(objPtr)->internalRep.doubleValue = (double)(d);	\
+	(objPtr)->typePtr = &tclDoubleType;			\
+    } while (0)
+
+/*
+ *----------------------------------------------------------------
+ * Macros used by the Tcl core to create and initialise objects of standard
+ * types, avoiding the corresponding function calls in time critical parts of
+ * the core. The ANSI C "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE void	TclNewLongObj(Tcl_Obj *objPtr, long l);
+ * MODULE_SCOPE void	TclNewWideObj(Tcl_Obj *objPtr, Tcl_WideInt w);
+ * MODULE_SCOPE void	TclNewDoubleObj(Tcl_Obj *objPtr, double d);
+ * MODULE_SCOPE void	TclNewStringObj(Tcl_Obj *objPtr, const char *s, int len);
+ * MODULE_SCOPE void	TclNewLiteralStringObj(Tcl_Obj*objPtr, const char *sLiteral);
+ *
+ *----------------------------------------------------------------
+ */
+
+#ifndef TCL_MEM_DEBUG
+#define TclNewLongObj(objPtr, i) \
+    do {						\
+	TclIncrObjsAllocated();				\
+	TclAllocObjStorage(objPtr);			\
+	(objPtr)->refCount = 0;				\
+	(objPtr)->bytes = NULL;				\
+	(objPtr)->internalRep.longValue = (long)(i);	\
+	(objPtr)->typePtr = &tclIntType;		\
+	TCL_DTRACE_OBJ_CREATE(objPtr);			\
+    } while (0)
+
+#define TclNewDoubleObj(objPtr, d) \
+    do {							\
+	TclIncrObjsAllocated();					\
+	TclAllocObjStorage(objPtr);				\
+	(objPtr)->refCount = 0;					\
+	(objPtr)->bytes = NULL;					\
+	(objPtr)->internalRep.doubleValue = (double)(d);	\
+	(objPtr)->typePtr = &tclDoubleType;			\
+	TCL_DTRACE_OBJ_CREATE(objPtr);				\
+    } while (0)
+
+#define TclNewStringObj(objPtr, s, len) \
+    do {							\
+	TclIncrObjsAllocated();					\
+	TclAllocObjStorage(objPtr);				\
+	(objPtr)->refCount = 0;					\
+	TclInitStringRep((objPtr), (s), (len));			\
+	(objPtr)->typePtr = NULL;				\
+	TCL_DTRACE_OBJ_CREATE(objPtr);				\
+    } while (0)
+
+#else /* TCL_MEM_DEBUG */
+#define TclNewLongObj(objPtr, l) \
+    (objPtr) = Tcl_NewLongObj(l)
+
+#define TclNewDoubleObj(objPtr, d) \
+    (objPtr) = Tcl_NewDoubleObj(d)
+
+#define TclNewStringObj(objPtr, s, len) \
+    (objPtr) = Tcl_NewStringObj((s), (len))
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ * The sLiteral argument *must* be a string literal; the incantation with
+ * sizeof(sLiteral "") will fail to compile otherwise.
+ */
+#define TclNewLiteralStringObj(objPtr, sLiteral) \
+    TclNewStringObj((objPtr), (sLiteral), (int) (sizeof(sLiteral "") - 1))
+
+/*
+ *----------------------------------------------------------------
+ * Convenience macros for DStrings.
+ * The ANSI C "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE char * TclDStringAppendLiteral(Tcl_DString *dsPtr,
+ *			const char *sLiteral);
+ * MODULE_SCOPE void   TclDStringClear(Tcl_DString *dsPtr);
+ */
+
+#define TclDStringAppendLiteral(dsPtr, sLiteral) \
+    Tcl_DStringAppend((dsPtr), (sLiteral), (int) (sizeof(sLiteral "") - 1))
+#define TclDStringClear(dsPtr) \
+    Tcl_DStringSetLength((dsPtr), 0)
+
+/*
+ *----------------------------------------------------------------
+ * Macros used by the Tcl core to test for some special double values.
+ * The ANSI C "prototypes" for these macros are:
+ *
+ * MODULE_SCOPE int	TclIsInfinite(double d);
+ * MODULE_SCOPE int	TclIsNaN(double d);
+ */
+
+#ifdef _MSC_VER
+#    define TclIsInfinite(d)	(!(_finite((d))))
+#    define TclIsNaN(d)		(_isnan((d)))
+#else
+#    define TclIsInfinite(d)	((d) > DBL_MAX || (d) < -DBL_MAX)
+#    ifdef NO_ISNAN
+#	 define TclIsNaN(d)	((d) != (d))
+#    else
+#	 define TclIsNaN(d)	(isnan(d))
+#    endif
+#endif
+
+/*
+ * ----------------------------------------------------------------------
+ * Macro to use to find the offset of a field in a structure. Computes number
+ * of bytes from beginning of structure to a given field.
+ */
+
+#ifdef offsetof
+#define TclOffset(type, field) ((int) offsetof(type, field))
+#else
+#define TclOffset(type, field) ((int) ((char *) &((type *) 0)->field))
+#endif
+
+/*
+ *----------------------------------------------------------------
+ * Inline version of Tcl_GetCurrentNamespace and Tcl_GetGlobalNamespace.
+ */
+
+#define TclGetCurrentNamespace(interp) \
+    (Tcl_Namespace *) ((Interp *)(interp))->varFramePtr->nsPtr
+
+#define TclGetGlobalNamespace(interp) \
+    (Tcl_Namespace *) ((Interp *)(interp))->globalNsPtr
+
+/*
+ *----------------------------------------------------------------
+ * Inline version of TclCleanupCommand; still need the function as it is in
+ * the internal stubs, but the core can use the macro instead.
+ */
+
+#define TclCleanupCommandMacro(cmdPtr) \
+    if ((cmdPtr)->refCount-- <= 1) { \
+	ckfree(cmdPtr);\
+    }
+
+/*
+ *----------------------------------------------------------------
+ * Inline versions of Tcl_LimitReady() and Tcl_LimitExceeded to limit number
+ * of calls out of the critical path. Note that this code isn't particularly
+ * readable; the non-inline version (in tclInterp.c) is much easier to
+ * understand. Note also that these macros takes different args (iPtr->limit)
+ * to the non-inline version.
+ */
+
+#define TclLimitExceeded(limit) ((limit).exceeded != 0)
+
+#define TclLimitReady(limit)						\
+    (((limit).active == 0) ? 0 :					\
+    (++(limit).granularityTicker,					\
+    ((((limit).active & TCL_LIMIT_COMMANDS) &&				\
+	    (((limit).cmdGranularity == 1) ||				\
+	    ((limit).granularityTicker % (limit).cmdGranularity == 0)))	\
+	    ? 1 :							\
+    (((limit).active & TCL_LIMIT_TIME) &&				\
+	    (((limit).timeGranularity == 1) ||				\
+	    ((limit).granularityTicker % (limit).timeGranularity == 0)))\
+	    ? 1 : 0)))
+
+/*
+ * Compile-time assertions: these produce a compile time error if the
+ * expression is not known to be true at compile time. If the assertion is
+ * known to be false, the compiler (or optimizer?) will error out with
+ * "division by zero". If the assertion cannot be evaluated at compile time,
+ * the compiler will error out with "non-static initializer".
+ *
+ * Adapted with permission from
+ * http://www.pixelbeat.org/programming/gcc/static_assert.html
+ */
+
+#define TCL_CT_ASSERT(e) \
+    {enum { ct_assert_value = 1/(!!(e)) };}
+
+/*
+ *----------------------------------------------------------------
+ * Allocator for small structs (<=sizeof(Tcl_Obj)) using the Tcl_Obj pool.
+ * Only checked at compile time.
+ *
+ * ONLY USE FOR CONSTANT nBytes.
+ *
+ * DO NOT LET THEM CROSS THREAD BOUNDARIES
+ *----------------------------------------------------------------
+ */
+
+#define TclSmallAlloc(nbytes, memPtr) \
+    TclSmallAllocEx(NULL, (nbytes), (memPtr))
+
+#define TclSmallFree(memPtr) \
+    TclSmallFreeEx(NULL, (memPtr))
+
+#ifndef TCL_MEM_DEBUG
+#define TclSmallAllocEx(interp, nbytes, memPtr) \
+    do {								\
+	Tcl_Obj *_objPtr;						\
+	TCL_CT_ASSERT((nbytes)<=sizeof(Tcl_Obj));			\
+	TclIncrObjsAllocated();						\
+	TclAllocObjStorageEx((interp), (_objPtr));			\
+	memPtr = (ClientData) (_objPtr);					\
+    } while (0)
+
+#define TclSmallFreeEx(interp, memPtr) \
+    do {								\
+	TclFreeObjStorageEx((interp), (Tcl_Obj *) (memPtr));		\
+	TclIncrObjsFreed();						\
+    } while (0)
+
+#else    /* TCL_MEM_DEBUG */
+#define TclSmallAllocEx(interp, nbytes, memPtr) \
+    do {								\
+	Tcl_Obj *_objPtr;						\
+	TCL_CT_ASSERT((nbytes)<=sizeof(Tcl_Obj));			\
+	TclNewObj(_objPtr);						\
+	memPtr = (ClientData) _objPtr;					\
+    } while (0)
+
+#define TclSmallFreeEx(interp, memPtr) \
+    do {								\
+	Tcl_Obj *_objPtr = (Tcl_Obj *) memPtr;				\
+	_objPtr->bytes = NULL;						\
+	_objPtr->typePtr = NULL;					\
+	_objPtr->refCount = 1;						\
+	TclDecrRefCount(_objPtr);					\
+    } while (0)
+#endif   /* TCL_MEM_DEBUG */
+
+/*
+ * Support for Clang Static Analyzer <http://clang-analyzer.llvm.org>
+ */
+
+#if defined(PURIFY) && defined(__clang__)
+#if __has_feature(attribute_analyzer_noreturn) && \
+	!defined(Tcl_Panic) && defined(Tcl_Panic_TCL_DECLARED)
+void Tcl_Panic(const char *, ...) __attribute__((analyzer_noreturn));
+#endif
+#if !defined(CLANG_ASSERT)
+#include <assert.h>
+#define CLANG_ASSERT(x) assert(x)
+#endif
+#elif !defined(CLANG_ASSERT)
+#define CLANG_ASSERT(x)
+#endif /* PURIFY && __clang__ */
+
+/*
+ *----------------------------------------------------------------
+ * Parameters, structs and macros for the non-recursive engine (NRE)
+ *----------------------------------------------------------------
+ */
+
+#define NRE_USE_SMALL_ALLOC	1  /* Only turn off for debugging purposes. */
+#ifndef NRE_ENABLE_ASSERTS
+#define NRE_ENABLE_ASSERTS	0
+#endif
+
+/*
+ * This is the main data struct for representing NR commands. It is designed
+ * to fit in sizeof(Tcl_Obj) in order to exploit the fastest memory allocator
+ * available.
+ */
+
+typedef struct NRE_callback {
+    Tcl_NRPostProc *procPtr;
+    ClientData data[4];
+    struct NRE_callback *nextPtr;
+} NRE_callback;
+
+#define TOP_CB(iPtr) (((Interp *)(iPtr))->execEnvPtr->callbackPtr)
+
+/*
+ * Inline version of Tcl_NRAddCallback.
+ */
+
+#define TclNRAddCallback(interp,postProcPtr,data0,data1,data2,data3) \
+    do {								\
+	NRE_callback *_callbackPtr;					\
+	TCLNR_ALLOC((interp), (_callbackPtr));				\
+	_callbackPtr->procPtr = (postProcPtr);				\
+	_callbackPtr->data[0] = (ClientData)(data0);			\
+	_callbackPtr->data[1] = (ClientData)(data1);			\
+	_callbackPtr->data[2] = (ClientData)(data2);			\
+	_callbackPtr->data[3] = (ClientData)(data3);			\
+	_callbackPtr->nextPtr = TOP_CB(interp);				\
+	TOP_CB(interp) = _callbackPtr;					\
+    } while (0)
+
+#if NRE_USE_SMALL_ALLOC
+#define TCLNR_ALLOC(interp, ptr) \
+    TclSmallAllocEx(interp, sizeof(NRE_callback), (ptr))
+#define TCLNR_FREE(interp, ptr)  TclSmallFreeEx((interp), (ptr))
+#else
+#define TCLNR_ALLOC(interp, ptr) \
+    (ptr = ((ClientData) ckalloc(sizeof(NRE_callback))))
+#define TCLNR_FREE(interp, ptr)  ckfree(ptr)
+#endif
+
+#if NRE_ENABLE_ASSERTS
+#define NRE_ASSERT(expr) assert((expr))
+#else
+#define NRE_ASSERT(expr)
+#endif
+
+#include "tclIntDecls.h"
+#include "tclIntPlatDecls.h"
+#include "tclTomMathDecls.h"
+
+#if !defined(USE_TCL_STUBS) && !defined(TCL_MEM_DEBUG)
+#define Tcl_AttemptAlloc(size)        TclpAlloc(size)
+#define Tcl_AttemptRealloc(ptr, size) TclpRealloc((ptr), (size))
+#define Tcl_Free(ptr)                 TclpFree(ptr)
+#endif
+
+#endif /* _TCLINT */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclIntDecls.h b/generic/tclIntDecls.h
new file mode 100644
index 0000000..5bccfe5
--- /dev/null
+++ b/generic/tclIntDecls.h
@@ -0,0 +1,1407 @@
+/*
+ * tclIntDecls.h --
+ *
+ *	This file contains the declarations for all unsupported
+ *	functions that are exported by the Tcl library.  These
+ *	interfaces are not guaranteed to remain the same between
+ *	versions.  Use at your own risk.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLINTDECLS
+#define _TCLINTDECLS
+
+#include "tclPort.h"
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/* [Bug #803489] Tcl_FindNamespace problem in the Stubs table */
+#undef Tcl_CreateNamespace
+#undef Tcl_DeleteNamespace
+#undef Tcl_AppendExportList
+#undef Tcl_Export
+#undef Tcl_Import
+#undef Tcl_ForgetImport
+#undef Tcl_GetCurrentNamespace
+#undef Tcl_GetGlobalNamespace
+#undef Tcl_FindNamespace
+#undef Tcl_FindCommand
+#undef Tcl_GetCommandFromObj
+#undef Tcl_GetCommandFullName
+#undef Tcl_SetStartupScript
+#undef Tcl_GetStartupScript
+
+/*
+ * WARNING: This file is automatically generated by the tools/genStubs.tcl
+ * script.  Any modifications to the function declarations below should be made
+ * in the generic/tclInt.decls script.
+ */
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+/* Slot 0 is reserved */
+/* Slot 1 is reserved */
+/* Slot 2 is reserved */
+/* 3 */
+EXTERN void		TclAllocateFreeObjects(void);
+/* Slot 4 is reserved */
+/* 5 */
+EXTERN int		TclCleanupChildren(Tcl_Interp *interp, int numPids,
+				Tcl_Pid *pidPtr, Tcl_Channel errorChan);
+/* 6 */
+EXTERN void		TclCleanupCommand(Command *cmdPtr);
+/* 7 */
+EXTERN int		TclCopyAndCollapse(int count, const char *src,
+				char *dst);
+/* 8 */
+EXTERN int		TclCopyChannelOld(Tcl_Interp *interp,
+				Tcl_Channel inChan, Tcl_Channel outChan,
+				int toRead, Tcl_Obj *cmdPtr);
+/* 9 */
+EXTERN int		TclCreatePipeline(Tcl_Interp *interp, int argc,
+				const char **argv, Tcl_Pid **pidArrayPtr,
+				TclFile *inPipePtr, TclFile *outPipePtr,
+				TclFile *errFilePtr);
+/* 10 */
+EXTERN int		TclCreateProc(Tcl_Interp *interp, Namespace *nsPtr,
+				const char *procName, Tcl_Obj *argsPtr,
+				Tcl_Obj *bodyPtr, Proc **procPtrPtr);
+/* 11 */
+EXTERN void		TclDeleteCompiledLocalVars(Interp *iPtr,
+				CallFrame *framePtr);
+/* 12 */
+EXTERN void		TclDeleteVars(Interp *iPtr,
+				TclVarHashTable *tablePtr);
+/* Slot 13 is reserved */
+/* 14 */
+EXTERN int		TclDumpMemoryInfo(ClientData clientData, int flags);
+/* Slot 15 is reserved */
+/* 16 */
+EXTERN void		TclExprFloatError(Tcl_Interp *interp, double value);
+/* Slot 17 is reserved */
+/* Slot 18 is reserved */
+/* Slot 19 is reserved */
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+/* 22 */
+EXTERN int		TclFindElement(Tcl_Interp *interp,
+				const char *listStr, int listLength,
+				const char **elementPtr,
+				const char **nextPtr, int *sizePtr,
+				int *bracePtr);
+/* 23 */
+EXTERN Proc *		TclFindProc(Interp *iPtr, const char *procName);
+/* 24 */
+EXTERN int		TclFormatInt(char *buffer, long n);
+/* 25 */
+EXTERN void		TclFreePackageInfo(Interp *iPtr);
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+/* 28 */
+EXTERN Tcl_Channel	TclpGetDefaultStdChannel(int type);
+/* Slot 29 is reserved */
+/* Slot 30 is reserved */
+/* 31 */
+EXTERN const char *	TclGetExtension(const char *name);
+/* 32 */
+EXTERN int		TclGetFrame(Tcl_Interp *interp, const char *str,
+				CallFrame **framePtrPtr);
+/* Slot 33 is reserved */
+/* 34 */
+EXTERN int		TclGetIntForIndex(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, int endValue, int *indexPtr);
+/* Slot 35 is reserved */
+/* Slot 36 is reserved */
+/* 37 */
+EXTERN int		TclGetLoadedPackages(Tcl_Interp *interp,
+				const char *targetName);
+/* 38 */
+EXTERN int		TclGetNamespaceForQualName(Tcl_Interp *interp,
+				const char *qualName, Namespace *cxtNsPtr,
+				int flags, Namespace **nsPtrPtr,
+				Namespace **altNsPtrPtr,
+				Namespace **actualCxtPtrPtr,
+				const char **simpleNamePtr);
+/* 39 */
+EXTERN TclObjCmdProcType TclGetObjInterpProc(void);
+/* 40 */
+EXTERN int		TclGetOpenMode(Tcl_Interp *interp, const char *str,
+				int *seekFlagPtr);
+/* 41 */
+EXTERN Tcl_Command	TclGetOriginalCommand(Tcl_Command command);
+/* 42 */
+EXTERN CONST86 char *	TclpGetUserHome(const char *name,
+				Tcl_DString *bufferPtr);
+/* Slot 43 is reserved */
+/* 44 */
+EXTERN int		TclGuessPackageName(const char *fileName,
+				Tcl_DString *bufPtr);
+/* 45 */
+EXTERN int		TclHideUnsafeCommands(Tcl_Interp *interp);
+/* 46 */
+EXTERN int		TclInExit(void);
+/* Slot 47 is reserved */
+/* Slot 48 is reserved */
+/* Slot 49 is reserved */
+/* 50 */
+EXTERN void		TclInitCompiledLocals(Tcl_Interp *interp,
+				CallFrame *framePtr, Namespace *nsPtr);
+/* 51 */
+EXTERN int		TclInterpInit(Tcl_Interp *interp);
+/* Slot 52 is reserved */
+/* 53 */
+EXTERN int		TclInvokeObjectCommand(ClientData clientData,
+				Tcl_Interp *interp, int argc,
+				CONST84 char **argv);
+/* 54 */
+EXTERN int		TclInvokeStringCommand(ClientData clientData,
+				Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[]);
+/* 55 */
+EXTERN Proc *		TclIsProc(Command *cmdPtr);
+/* Slot 56 is reserved */
+/* Slot 57 is reserved */
+/* 58 */
+EXTERN Var *		TclLookupVar(Tcl_Interp *interp, const char *part1,
+				const char *part2, int flags,
+				const char *msg, int createPart1,
+				int createPart2, Var **arrayPtrPtr);
+/* Slot 59 is reserved */
+/* 60 */
+EXTERN int		TclNeedSpace(const char *start, const char *end);
+/* 61 */
+EXTERN Tcl_Obj *	TclNewProcBodyObj(Proc *procPtr);
+/* 62 */
+EXTERN int		TclObjCommandComplete(Tcl_Obj *cmdPtr);
+/* 63 */
+EXTERN int		TclObjInterpProc(ClientData clientData,
+				Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[]);
+/* 64 */
+EXTERN int		TclObjInvoke(Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[], int flags);
+/* Slot 65 is reserved */
+/* Slot 66 is reserved */
+/* Slot 67 is reserved */
+/* Slot 68 is reserved */
+/* 69 */
+EXTERN char *		TclpAlloc(unsigned int size);
+/* Slot 70 is reserved */
+/* Slot 71 is reserved */
+/* Slot 72 is reserved */
+/* Slot 73 is reserved */
+/* 74 */
+EXTERN void		TclpFree(char *ptr);
+/* 75 */
+EXTERN unsigned long	TclpGetClicks(void);
+/* 76 */
+EXTERN unsigned long	TclpGetSeconds(void);
+/* 77 */
+EXTERN void		TclpGetTime(Tcl_Time *time);
+/* Slot 78 is reserved */
+/* Slot 79 is reserved */
+/* Slot 80 is reserved */
+/* 81 */
+EXTERN char *		TclpRealloc(char *ptr, unsigned int size);
+/* Slot 82 is reserved */
+/* Slot 83 is reserved */
+/* Slot 84 is reserved */
+/* Slot 85 is reserved */
+/* Slot 86 is reserved */
+/* Slot 87 is reserved */
+/* 88 */
+EXTERN char *		TclPrecTraceProc(ClientData clientData,
+				Tcl_Interp *interp, const char *name1,
+				const char *name2, int flags);
+/* 89 */
+EXTERN int		TclPreventAliasLoop(Tcl_Interp *interp,
+				Tcl_Interp *cmdInterp, Tcl_Command cmd);
+/* Slot 90 is reserved */
+/* 91 */
+EXTERN void		TclProcCleanupProc(Proc *procPtr);
+/* 92 */
+EXTERN int		TclProcCompileProc(Tcl_Interp *interp, Proc *procPtr,
+				Tcl_Obj *bodyPtr, Namespace *nsPtr,
+				const char *description,
+				const char *procName);
+/* 93 */
+EXTERN void		TclProcDeleteProc(ClientData clientData);
+/* Slot 94 is reserved */
+/* Slot 95 is reserved */
+/* 96 */
+EXTERN int		TclRenameCommand(Tcl_Interp *interp,
+				const char *oldName, const char *newName);
+/* 97 */
+EXTERN void		TclResetShadowedCmdRefs(Tcl_Interp *interp,
+				Command *newCmdPtr);
+/* 98 */
+EXTERN int		TclServiceIdle(void);
+/* Slot 99 is reserved */
+/* Slot 100 is reserved */
+/* 101 */
+EXTERN CONST86 char *	TclSetPreInitScript(const char *string);
+/* 102 */
+EXTERN void		TclSetupEnv(Tcl_Interp *interp);
+/* 103 */
+EXTERN int		TclSockGetPort(Tcl_Interp *interp, const char *str,
+				const char *proto, int *portPtr);
+/* 104 */
+EXTERN int		TclSockMinimumBuffersOld(int sock, int size);
+/* Slot 105 is reserved */
+/* Slot 106 is reserved */
+/* Slot 107 is reserved */
+/* 108 */
+EXTERN void		TclTeardownNamespace(Namespace *nsPtr);
+/* 109 */
+EXTERN int		TclUpdateReturnInfo(Interp *iPtr);
+/* 110 */
+EXTERN int		TclSockMinimumBuffers(void *sock, int size);
+/* 111 */
+EXTERN void		Tcl_AddInterpResolvers(Tcl_Interp *interp,
+				const char *name,
+				Tcl_ResolveCmdProc *cmdProc,
+				Tcl_ResolveVarProc *varProc,
+				Tcl_ResolveCompiledVarProc *compiledVarProc);
+/* 112 */
+EXTERN int		Tcl_AppendExportList(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr, Tcl_Obj *objPtr);
+/* 113 */
+EXTERN Tcl_Namespace *	Tcl_CreateNamespace(Tcl_Interp *interp,
+				const char *name, ClientData clientData,
+				Tcl_NamespaceDeleteProc *deleteProc);
+/* 114 */
+EXTERN void		Tcl_DeleteNamespace(Tcl_Namespace *nsPtr);
+/* 115 */
+EXTERN int		Tcl_Export(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+				const char *pattern, int resetListFirst);
+/* 116 */
+EXTERN Tcl_Command	Tcl_FindCommand(Tcl_Interp *interp, const char *name,
+				Tcl_Namespace *contextNsPtr, int flags);
+/* 117 */
+EXTERN Tcl_Namespace *	Tcl_FindNamespace(Tcl_Interp *interp,
+				const char *name,
+				Tcl_Namespace *contextNsPtr, int flags);
+/* 118 */
+EXTERN int		Tcl_GetInterpResolvers(Tcl_Interp *interp,
+				const char *name, Tcl_ResolverInfo *resInfo);
+/* 119 */
+EXTERN int		Tcl_GetNamespaceResolvers(
+				Tcl_Namespace *namespacePtr,
+				Tcl_ResolverInfo *resInfo);
+/* 120 */
+EXTERN Tcl_Var		Tcl_FindNamespaceVar(Tcl_Interp *interp,
+				const char *name,
+				Tcl_Namespace *contextNsPtr, int flags);
+/* 121 */
+EXTERN int		Tcl_ForgetImport(Tcl_Interp *interp,
+				Tcl_Namespace *nsPtr, const char *pattern);
+/* 122 */
+EXTERN Tcl_Command	Tcl_GetCommandFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 123 */
+EXTERN void		Tcl_GetCommandFullName(Tcl_Interp *interp,
+				Tcl_Command command, Tcl_Obj *objPtr);
+/* 124 */
+EXTERN Tcl_Namespace *	Tcl_GetCurrentNamespace(Tcl_Interp *interp);
+/* 125 */
+EXTERN Tcl_Namespace *	Tcl_GetGlobalNamespace(Tcl_Interp *interp);
+/* 126 */
+EXTERN void		Tcl_GetVariableFullName(Tcl_Interp *interp,
+				Tcl_Var variable, Tcl_Obj *objPtr);
+/* 127 */
+EXTERN int		Tcl_Import(Tcl_Interp *interp, Tcl_Namespace *nsPtr,
+				const char *pattern, int allowOverwrite);
+/* 128 */
+EXTERN void		Tcl_PopCallFrame(Tcl_Interp *interp);
+/* 129 */
+EXTERN int		Tcl_PushCallFrame(Tcl_Interp *interp,
+				Tcl_CallFrame *framePtr,
+				Tcl_Namespace *nsPtr, int isProcCallFrame);
+/* 130 */
+EXTERN int		Tcl_RemoveInterpResolvers(Tcl_Interp *interp,
+				const char *name);
+/* 131 */
+EXTERN void		Tcl_SetNamespaceResolvers(
+				Tcl_Namespace *namespacePtr,
+				Tcl_ResolveCmdProc *cmdProc,
+				Tcl_ResolveVarProc *varProc,
+				Tcl_ResolveCompiledVarProc *compiledVarProc);
+/* 132 */
+EXTERN int		TclpHasSockets(Tcl_Interp *interp);
+/* 133 */
+EXTERN struct tm *	TclpGetDate(const time_t *time, int useGMT);
+/* Slot 134 is reserved */
+/* Slot 135 is reserved */
+/* Slot 136 is reserved */
+/* Slot 137 is reserved */
+/* 138 */
+EXTERN CONST84_RETURN char * TclGetEnv(const char *name,
+				Tcl_DString *valuePtr);
+/* Slot 139 is reserved */
+/* Slot 140 is reserved */
+/* 141 */
+EXTERN CONST84_RETURN char * TclpGetCwd(Tcl_Interp *interp,
+				Tcl_DString *cwdPtr);
+/* 142 */
+EXTERN int		TclSetByteCodeFromAny(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, CompileHookProc *hookProc,
+				ClientData clientData);
+/* 143 */
+EXTERN int		TclAddLiteralObj(struct CompileEnv *envPtr,
+				Tcl_Obj *objPtr, LiteralEntry **litPtrPtr);
+/* 144 */
+EXTERN void		TclHideLiteral(Tcl_Interp *interp,
+				struct CompileEnv *envPtr, int index);
+/* 145 */
+EXTERN const struct AuxDataType * TclGetAuxDataType(const char *typeName);
+/* 146 */
+EXTERN TclHandle	TclHandleCreate(void *ptr);
+/* 147 */
+EXTERN void		TclHandleFree(TclHandle handle);
+/* 148 */
+EXTERN TclHandle	TclHandlePreserve(TclHandle handle);
+/* 149 */
+EXTERN void		TclHandleRelease(TclHandle handle);
+/* 150 */
+EXTERN int		TclRegAbout(Tcl_Interp *interp, Tcl_RegExp re);
+/* 151 */
+EXTERN void		TclRegExpRangeUniChar(Tcl_RegExp re, int index,
+				int *startPtr, int *endPtr);
+/* 152 */
+EXTERN void		TclSetLibraryPath(Tcl_Obj *pathPtr);
+/* 153 */
+EXTERN Tcl_Obj *	TclGetLibraryPath(void);
+/* Slot 154 is reserved */
+/* Slot 155 is reserved */
+/* 156 */
+EXTERN void		TclRegError(Tcl_Interp *interp, const char *msg,
+				int status);
+/* 157 */
+EXTERN Var *		TclVarTraceExists(Tcl_Interp *interp,
+				const char *varName);
+/* 158 */
+EXTERN void		TclSetStartupScriptFileName(const char *filename);
+/* 159 */
+EXTERN const char *	TclGetStartupScriptFileName(void);
+/* Slot 160 is reserved */
+/* 161 */
+EXTERN int		TclChannelTransform(Tcl_Interp *interp,
+				Tcl_Channel chan, Tcl_Obj *cmdObjPtr);
+/* 162 */
+EXTERN void		TclChannelEventScriptInvoker(ClientData clientData,
+				int flags);
+/* 163 */
+EXTERN const void *	TclGetInstructionTable(void);
+/* 164 */
+EXTERN void		TclExpandCodeArray(void *envPtr);
+/* 165 */
+EXTERN void		TclpSetInitialEncodings(void);
+/* 166 */
+EXTERN int		TclListObjSetElement(Tcl_Interp *interp,
+				Tcl_Obj *listPtr, int index,
+				Tcl_Obj *valuePtr);
+/* 167 */
+EXTERN void		TclSetStartupScriptPath(Tcl_Obj *pathPtr);
+/* 168 */
+EXTERN Tcl_Obj *	TclGetStartupScriptPath(void);
+/* 169 */
+EXTERN int		TclpUtfNcmp2(const char *s1, const char *s2,
+				unsigned long n);
+/* 170 */
+EXTERN int		TclCheckInterpTraces(Tcl_Interp *interp,
+				const char *command, int numChars,
+				Command *cmdPtr, int result, int traceFlags,
+				int objc, Tcl_Obj *const objv[]);
+/* 171 */
+EXTERN int		TclCheckExecutionTraces(Tcl_Interp *interp,
+				const char *command, int numChars,
+				Command *cmdPtr, int result, int traceFlags,
+				int objc, Tcl_Obj *const objv[]);
+/* 172 */
+EXTERN int		TclInThreadExit(void);
+/* 173 */
+EXTERN int		TclUniCharMatch(const Tcl_UniChar *string,
+				int strLen, const Tcl_UniChar *pattern,
+				int ptnLen, int flags);
+/* Slot 174 is reserved */
+/* 175 */
+EXTERN int		TclCallVarTraces(Interp *iPtr, Var *arrayPtr,
+				Var *varPtr, const char *part1,
+				const char *part2, int flags,
+				int leaveErrMsg);
+/* 176 */
+EXTERN void		TclCleanupVar(Var *varPtr, Var *arrayPtr);
+/* 177 */
+EXTERN void		TclVarErrMsg(Tcl_Interp *interp, const char *part1,
+				const char *part2, const char *operation,
+				const char *reason);
+/* 178 */
+EXTERN void		Tcl_SetStartupScript(Tcl_Obj *pathPtr,
+				const char *encodingName);
+/* 179 */
+EXTERN Tcl_Obj *	Tcl_GetStartupScript(const char **encodingNamePtr);
+/* Slot 180 is reserved */
+/* Slot 181 is reserved */
+/* 182 */
+EXTERN struct tm *	TclpLocaltime(const time_t *clock);
+/* 183 */
+EXTERN struct tm *	TclpGmtime(const time_t *clock);
+/* Slot 184 is reserved */
+/* Slot 185 is reserved */
+/* Slot 186 is reserved */
+/* Slot 187 is reserved */
+/* Slot 188 is reserved */
+/* Slot 189 is reserved */
+/* Slot 190 is reserved */
+/* Slot 191 is reserved */
+/* Slot 192 is reserved */
+/* Slot 193 is reserved */
+/* Slot 194 is reserved */
+/* Slot 195 is reserved */
+/* Slot 196 is reserved */
+/* Slot 197 is reserved */
+/* 198 */
+EXTERN int		TclObjGetFrame(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				CallFrame **framePtrPtr);
+/* Slot 199 is reserved */
+/* 200 */
+EXTERN int		TclpObjRemoveDirectory(Tcl_Obj *pathPtr,
+				int recursive, Tcl_Obj **errorPtr);
+/* 201 */
+EXTERN int		TclpObjCopyDirectory(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr);
+/* 202 */
+EXTERN int		TclpObjCreateDirectory(Tcl_Obj *pathPtr);
+/* 203 */
+EXTERN int		TclpObjDeleteFile(Tcl_Obj *pathPtr);
+/* 204 */
+EXTERN int		TclpObjCopyFile(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr);
+/* 205 */
+EXTERN int		TclpObjRenameFile(Tcl_Obj *srcPathPtr,
+				Tcl_Obj *destPathPtr);
+/* 206 */
+EXTERN int		TclpObjStat(Tcl_Obj *pathPtr, Tcl_StatBuf *buf);
+/* 207 */
+EXTERN int		TclpObjAccess(Tcl_Obj *pathPtr, int mode);
+/* 208 */
+EXTERN Tcl_Channel	TclpOpenFileChannel(Tcl_Interp *interp,
+				Tcl_Obj *pathPtr, int mode, int permissions);
+/* Slot 209 is reserved */
+/* Slot 210 is reserved */
+/* Slot 211 is reserved */
+/* 212 */
+EXTERN void		TclpFindExecutable(const char *argv0);
+/* 213 */
+EXTERN Tcl_Obj *	TclGetObjNameOfExecutable(void);
+/* 214 */
+EXTERN void		TclSetObjNameOfExecutable(Tcl_Obj *name,
+				Tcl_Encoding encoding);
+/* 215 */
+EXTERN void *		TclStackAlloc(Tcl_Interp *interp, int numBytes);
+/* 216 */
+EXTERN void		TclStackFree(Tcl_Interp *interp, void *freePtr);
+/* 217 */
+EXTERN int		TclPushStackFrame(Tcl_Interp *interp,
+				Tcl_CallFrame **framePtrPtr,
+				Tcl_Namespace *namespacePtr,
+				int isProcCallFrame);
+/* 218 */
+EXTERN void		TclPopStackFrame(Tcl_Interp *interp);
+/* Slot 219 is reserved */
+/* Slot 220 is reserved */
+/* Slot 221 is reserved */
+/* Slot 222 is reserved */
+/* Slot 223 is reserved */
+/* 224 */
+EXTERN TclPlatformType * TclGetPlatform(void);
+/* 225 */
+EXTERN Tcl_Obj *	TclTraceDictPath(Tcl_Interp *interp,
+				Tcl_Obj *rootPtr, int keyc,
+				Tcl_Obj *const keyv[], int flags);
+/* 226 */
+EXTERN int		TclObjBeingDeleted(Tcl_Obj *objPtr);
+/* 227 */
+EXTERN void		TclSetNsPath(Namespace *nsPtr, int pathLength,
+				Tcl_Namespace *pathAry[]);
+/* Slot 228 is reserved */
+/* 229 */
+EXTERN int		TclPtrMakeUpvar(Tcl_Interp *interp, Var *otherP1Ptr,
+				const char *myName, int myFlags, int index);
+/* 230 */
+EXTERN Var *		TclObjLookupVar(Tcl_Interp *interp,
+				Tcl_Obj *part1Ptr, const char *part2,
+				int flags, const char *msg,
+				const int createPart1, const int createPart2,
+				Var **arrayPtrPtr);
+/* 231 */
+EXTERN int		TclGetNamespaceFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr, Tcl_Namespace **nsPtrPtr);
+/* 232 */
+EXTERN int		TclEvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags, const CmdFrame *invoker, int word);
+/* 233 */
+EXTERN void		TclGetSrcInfoForPc(CmdFrame *contextPtr);
+/* 234 */
+EXTERN Var *		TclVarHashCreateVar(TclVarHashTable *tablePtr,
+				const char *key, int *newPtr);
+/* 235 */
+EXTERN void		TclInitVarHashTable(TclVarHashTable *tablePtr,
+				Namespace *nsPtr);
+/* 236 */
+EXTERN void		TclBackgroundException(Tcl_Interp *interp, int code);
+/* 237 */
+EXTERN int		TclResetCancellation(Tcl_Interp *interp, int force);
+/* 238 */
+EXTERN int		TclNRInterpProc(ClientData clientData,
+				Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[]);
+/* 239 */
+EXTERN int		TclNRInterpProcCore(Tcl_Interp *interp,
+				Tcl_Obj *procNameObj, int skip,
+				ProcErrorProc *errorProc);
+/* 240 */
+EXTERN int		TclNRRunCallbacks(Tcl_Interp *interp, int result,
+				struct NRE_callback *rootPtr);
+/* 241 */
+EXTERN int		TclNREvalObjEx(Tcl_Interp *interp, Tcl_Obj *objPtr,
+				int flags, const CmdFrame *invoker, int word);
+/* 242 */
+EXTERN int		TclNREvalObjv(Tcl_Interp *interp, int objc,
+				Tcl_Obj *const objv[], int flags,
+				Command *cmdPtr);
+/* 243 */
+EXTERN void		TclDbDumpActiveObjects(FILE *outFile);
+/* 244 */
+EXTERN Tcl_HashTable *	TclGetNamespaceChildTable(Tcl_Namespace *nsPtr);
+/* 245 */
+EXTERN Tcl_HashTable *	TclGetNamespaceCommandTable(Tcl_Namespace *nsPtr);
+/* 246 */
+EXTERN int		TclInitRewriteEnsemble(Tcl_Interp *interp,
+				int numRemoved, int numInserted,
+				Tcl_Obj *const *objv);
+/* 247 */
+EXTERN void		TclResetRewriteEnsemble(Tcl_Interp *interp,
+				int isRootEnsemble);
+/* 248 */
+EXTERN int		TclCopyChannel(Tcl_Interp *interp,
+				Tcl_Channel inChan, Tcl_Channel outChan,
+				Tcl_WideInt toRead, Tcl_Obj *cmdPtr);
+/* 249 */
+EXTERN char *		TclDoubleDigits(double dv, int ndigits, int flags,
+				int *decpt, int *signum, char **endPtr);
+/* 250 */
+EXTERN void		TclSetSlaveCancelFlags(Tcl_Interp *interp, int flags,
+				int force);
+/* 251 */
+EXTERN int		TclRegisterLiteral(void *envPtr, const char *bytes,
+				int length, int flags);
+/* 252 */
+EXTERN Tcl_Obj *	TclPtrGetVar(Tcl_Interp *interp, Tcl_Var varPtr,
+				Tcl_Var arrayPtr, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, const int flags);
+/* 253 */
+EXTERN Tcl_Obj *	TclPtrSetVar(Tcl_Interp *interp, Tcl_Var varPtr,
+				Tcl_Var arrayPtr, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr,
+				const int flags);
+/* 254 */
+EXTERN Tcl_Obj *	TclPtrIncrObjVar(Tcl_Interp *interp, Tcl_Var varPtr,
+				Tcl_Var arrayPtr, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, Tcl_Obj *incrPtr,
+				const int flags);
+/* 255 */
+EXTERN int		TclPtrObjMakeUpvar(Tcl_Interp *interp,
+				Tcl_Var otherPtr, Tcl_Obj *myNamePtr,
+				int myFlags);
+/* 256 */
+EXTERN int		TclPtrUnsetVar(Tcl_Interp *interp, Tcl_Var varPtr,
+				Tcl_Var arrayPtr, Tcl_Obj *part1Ptr,
+				Tcl_Obj *part2Ptr, const int flags);
+
+typedef struct TclIntStubs {
+    int magic;
+    void *hooks;
+
+    void (*reserved0)(void);
+    void (*reserved1)(void);
+    void (*reserved2)(void);
+    void (*tclAllocateFreeObjects) (void); /* 3 */
+    void (*reserved4)(void);
+    int (*tclCleanupChildren) (Tcl_Interp *interp, int numPids, Tcl_Pid *pidPtr, Tcl_Channel errorChan); /* 5 */
+    void (*tclCleanupCommand) (Command *cmdPtr); /* 6 */
+    int (*tclCopyAndCollapse) (int count, const char *src, char *dst); /* 7 */
+    int (*tclCopyChannelOld) (Tcl_Interp *interp, Tcl_Channel inChan, Tcl_Channel outChan, int toRead, Tcl_Obj *cmdPtr); /* 8 */
+    int (*tclCreatePipeline) (Tcl_Interp *interp, int argc, const char **argv, Tcl_Pid **pidArrayPtr, TclFile *inPipePtr, TclFile *outPipePtr, TclFile *errFilePtr); /* 9 */
+    int (*tclCreateProc) (Tcl_Interp *interp, Namespace *nsPtr, const char *procName, Tcl_Obj *argsPtr, Tcl_Obj *bodyPtr, Proc **procPtrPtr); /* 10 */
+    void (*tclDeleteCompiledLocalVars) (Interp *iPtr, CallFrame *framePtr); /* 11 */
+    void (*tclDeleteVars) (Interp *iPtr, TclVarHashTable *tablePtr); /* 12 */
+    void (*reserved13)(void);
+    int (*tclDumpMemoryInfo) (ClientData clientData, int flags); /* 14 */
+    void (*reserved15)(void);
+    void (*tclExprFloatError) (Tcl_Interp *interp, double value); /* 16 */
+    void (*reserved17)(void);
+    void (*reserved18)(void);
+    void (*reserved19)(void);
+    void (*reserved20)(void);
+    void (*reserved21)(void);
+    int (*tclFindElement) (Tcl_Interp *interp, const char *listStr, int listLength, const char **elementPtr, const char **nextPtr, int *sizePtr, int *bracePtr); /* 22 */
+    Proc * (*tclFindProc) (Interp *iPtr, const char *procName); /* 23 */
+    int (*tclFormatInt) (char *buffer, long n); /* 24 */
+    void (*tclFreePackageInfo) (Interp *iPtr); /* 25 */
+    void (*reserved26)(void);
+    void (*reserved27)(void);
+    Tcl_Channel (*tclpGetDefaultStdChannel) (int type); /* 28 */
+    void (*reserved29)(void);
+    void (*reserved30)(void);
+    const char * (*tclGetExtension) (const char *name); /* 31 */
+    int (*tclGetFrame) (Tcl_Interp *interp, const char *str, CallFrame **framePtrPtr); /* 32 */
+    void (*reserved33)(void);
+    int (*tclGetIntForIndex) (Tcl_Interp *interp, Tcl_Obj *objPtr, int endValue, int *indexPtr); /* 34 */
+    void (*reserved35)(void);
+    void (*reserved36)(void);
+    int (*tclGetLoadedPackages) (Tcl_Interp *interp, const char *targetName); /* 37 */
+    int (*tclGetNamespaceForQualName) (Tcl_Interp *interp, const char *qualName, Namespace *cxtNsPtr, int flags, Namespace **nsPtrPtr, Namespace **altNsPtrPtr, Namespace **actualCxtPtrPtr, const char **simpleNamePtr); /* 38 */
+    TclObjCmdProcType (*tclGetObjInterpProc) (void); /* 39 */
+    int (*tclGetOpenMode) (Tcl_Interp *interp, const char *str, int *seekFlagPtr); /* 40 */
+    Tcl_Command (*tclGetOriginalCommand) (Tcl_Command command); /* 41 */
+    CONST86 char * (*tclpGetUserHome) (const char *name, Tcl_DString *bufferPtr); /* 42 */
+    void (*reserved43)(void);
+    int (*tclGuessPackageName) (const char *fileName, Tcl_DString *bufPtr); /* 44 */
+    int (*tclHideUnsafeCommands) (Tcl_Interp *interp); /* 45 */
+    int (*tclInExit) (void); /* 46 */
+    void (*reserved47)(void);
+    void (*reserved48)(void);
+    void (*reserved49)(void);
+    void (*tclInitCompiledLocals) (Tcl_Interp *interp, CallFrame *framePtr, Namespace *nsPtr); /* 50 */
+    int (*tclInterpInit) (Tcl_Interp *interp); /* 51 */
+    void (*reserved52)(void);
+    int (*tclInvokeObjectCommand) (ClientData clientData, Tcl_Interp *interp, int argc, CONST84 char **argv); /* 53 */
+    int (*tclInvokeStringCommand) (ClientData clientData, Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]); /* 54 */
+    Proc * (*tclIsProc) (Command *cmdPtr); /* 55 */
+    void (*reserved56)(void);
+    void (*reserved57)(void);
+    Var * (*tclLookupVar) (Tcl_Interp *interp, const char *part1, const char *part2, int flags, const char *msg, int createPart1, int createPart2, Var **arrayPtrPtr); /* 58 */
+    void (*reserved59)(void);
+    int (*tclNeedSpace) (const char *start, const char *end); /* 60 */
+    Tcl_Obj * (*tclNewProcBodyObj) (Proc *procPtr); /* 61 */
+    int (*tclObjCommandComplete) (Tcl_Obj *cmdPtr); /* 62 */
+    int (*tclObjInterpProc) (ClientData clientData, Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]); /* 63 */
+    int (*tclObjInvoke) (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[], int flags); /* 64 */
+    void (*reserved65)(void);
+    void (*reserved66)(void);
+    void (*reserved67)(void);
+    void (*reserved68)(void);
+    char * (*tclpAlloc) (unsigned int size); /* 69 */
+    void (*reserved70)(void);
+    void (*reserved71)(void);
+    void (*reserved72)(void);
+    void (*reserved73)(void);
+    void (*tclpFree) (char *ptr); /* 74 */
+    unsigned long (*tclpGetClicks) (void); /* 75 */
+    unsigned long (*tclpGetSeconds) (void); /* 76 */
+    void (*tclpGetTime) (Tcl_Time *time); /* 77 */
+    void (*reserved78)(void);
+    void (*reserved79)(void);
+    void (*reserved80)(void);
+    char * (*tclpRealloc) (char *ptr, unsigned int size); /* 81 */
+    void (*reserved82)(void);
+    void (*reserved83)(void);
+    void (*reserved84)(void);
+    void (*reserved85)(void);
+    void (*reserved86)(void);
+    void (*reserved87)(void);
+    char * (*tclPrecTraceProc) (ClientData clientData, Tcl_Interp *interp, const char *name1, const char *name2, int flags); /* 88 */
+    int (*tclPreventAliasLoop) (Tcl_Interp *interp, Tcl_Interp *cmdInterp, Tcl_Command cmd); /* 89 */
+    void (*reserved90)(void);
+    void (*tclProcCleanupProc) (Proc *procPtr); /* 91 */
+    int (*tclProcCompileProc) (Tcl_Interp *interp, Proc *procPtr, Tcl_Obj *bodyPtr, Namespace *nsPtr, const char *description, const char *procName); /* 92 */
+    void (*tclProcDeleteProc) (ClientData clientData); /* 93 */
+    void (*reserved94)(void);
+    void (*reserved95)(void);
+    int (*tclRenameCommand) (Tcl_Interp *interp, const char *oldName, const char *newName); /* 96 */
+    void (*tclResetShadowedCmdRefs) (Tcl_Interp *interp, Command *newCmdPtr); /* 97 */
+    int (*tclServiceIdle) (void); /* 98 */
+    void (*reserved99)(void);
+    void (*reserved100)(void);
+    CONST86 char * (*tclSetPreInitScript) (const char *string); /* 101 */
+    void (*tclSetupEnv) (Tcl_Interp *interp); /* 102 */
+    int (*tclSockGetPort) (Tcl_Interp *interp, const char *str, const char *proto, int *portPtr); /* 103 */
+    int (*tclSockMinimumBuffersOld) (int sock, int size); /* 104 */
+    void (*reserved105)(void);
+    void (*reserved106)(void);
+    void (*reserved107)(void);
+    void (*tclTeardownNamespace) (Namespace *nsPtr); /* 108 */
+    int (*tclUpdateReturnInfo) (Interp *iPtr); /* 109 */
+    int (*tclSockMinimumBuffers) (void *sock, int size); /* 110 */
+    void (*tcl_AddInterpResolvers) (Tcl_Interp *interp, const char *name, Tcl_ResolveCmdProc *cmdProc, Tcl_ResolveVarProc *varProc, Tcl_ResolveCompiledVarProc *compiledVarProc); /* 111 */
+    int (*tcl_AppendExportList) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, Tcl_Obj *objPtr); /* 112 */
+    Tcl_Namespace * (*tcl_CreateNamespace) (Tcl_Interp *interp, const char *name, ClientData clientData, Tcl_NamespaceDeleteProc *deleteProc); /* 113 */
+    void (*tcl_DeleteNamespace) (Tcl_Namespace *nsPtr); /* 114 */
+    int (*tcl_Export) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern, int resetListFirst); /* 115 */
+    Tcl_Command (*tcl_FindCommand) (Tcl_Interp *interp, const char *name, Tcl_Namespace *contextNsPtr, int flags); /* 116 */
+    Tcl_Namespace * (*tcl_FindNamespace) (Tcl_Interp *interp, const char *name, Tcl_Namespace *contextNsPtr, int flags); /* 117 */
+    int (*tcl_GetInterpResolvers) (Tcl_Interp *interp, const char *name, Tcl_ResolverInfo *resInfo); /* 118 */
+    int (*tcl_GetNamespaceResolvers) (Tcl_Namespace *namespacePtr, Tcl_ResolverInfo *resInfo); /* 119 */
+    Tcl_Var (*tcl_FindNamespaceVar) (Tcl_Interp *interp, const char *name, Tcl_Namespace *contextNsPtr, int flags); /* 120 */
+    int (*tcl_ForgetImport) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern); /* 121 */
+    Tcl_Command (*tcl_GetCommandFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 122 */
+    void (*tcl_GetCommandFullName) (Tcl_Interp *interp, Tcl_Command command, Tcl_Obj *objPtr); /* 123 */
+    Tcl_Namespace * (*tcl_GetCurrentNamespace) (Tcl_Interp *interp); /* 124 */
+    Tcl_Namespace * (*tcl_GetGlobalNamespace) (Tcl_Interp *interp); /* 125 */
+    void (*tcl_GetVariableFullName) (Tcl_Interp *interp, Tcl_Var variable, Tcl_Obj *objPtr); /* 126 */
+    int (*tcl_Import) (Tcl_Interp *interp, Tcl_Namespace *nsPtr, const char *pattern, int allowOverwrite); /* 127 */
+    void (*tcl_PopCallFrame) (Tcl_Interp *interp); /* 128 */
+    int (*tcl_PushCallFrame) (Tcl_Interp *interp, Tcl_CallFrame *framePtr, Tcl_Namespace *nsPtr, int isProcCallFrame); /* 129 */
+    int (*tcl_RemoveInterpResolvers) (Tcl_Interp *interp, const char *name); /* 130 */
+    void (*tcl_SetNamespaceResolvers) (Tcl_Namespace *namespacePtr, Tcl_ResolveCmdProc *cmdProc, Tcl_ResolveVarProc *varProc, Tcl_ResolveCompiledVarProc *compiledVarProc); /* 131 */
+    int (*tclpHasSockets) (Tcl_Interp *interp); /* 132 */
+    struct tm * (*tclpGetDate) (const time_t *time, int useGMT); /* 133 */
+    void (*reserved134)(void);
+    void (*reserved135)(void);
+    void (*reserved136)(void);
+    void (*reserved137)(void);
+    CONST84_RETURN char * (*tclGetEnv) (const char *name, Tcl_DString *valuePtr); /* 138 */
+    void (*reserved139)(void);
+    void (*reserved140)(void);
+    CONST84_RETURN char * (*tclpGetCwd) (Tcl_Interp *interp, Tcl_DString *cwdPtr); /* 141 */
+    int (*tclSetByteCodeFromAny) (Tcl_Interp *interp, Tcl_Obj *objPtr, CompileHookProc *hookProc, ClientData clientData); /* 142 */
+    int (*tclAddLiteralObj) (struct CompileEnv *envPtr, Tcl_Obj *objPtr, LiteralEntry **litPtrPtr); /* 143 */
+    void (*tclHideLiteral) (Tcl_Interp *interp, struct CompileEnv *envPtr, int index); /* 144 */
+    const struct AuxDataType * (*tclGetAuxDataType) (const char *typeName); /* 145 */
+    TclHandle (*tclHandleCreate) (void *ptr); /* 146 */
+    void (*tclHandleFree) (TclHandle handle); /* 147 */
+    TclHandle (*tclHandlePreserve) (TclHandle handle); /* 148 */
+    void (*tclHandleRelease) (TclHandle handle); /* 149 */
+    int (*tclRegAbout) (Tcl_Interp *interp, Tcl_RegExp re); /* 150 */
+    void (*tclRegExpRangeUniChar) (Tcl_RegExp re, int index, int *startPtr, int *endPtr); /* 151 */
+    void (*tclSetLibraryPath) (Tcl_Obj *pathPtr); /* 152 */
+    Tcl_Obj * (*tclGetLibraryPath) (void); /* 153 */
+    void (*reserved154)(void);
+    void (*reserved155)(void);
+    void (*tclRegError) (Tcl_Interp *interp, const char *msg, int status); /* 156 */
+    Var * (*tclVarTraceExists) (Tcl_Interp *interp, const char *varName); /* 157 */
+    void (*tclSetStartupScriptFileName) (const char *filename); /* 158 */
+    const char * (*tclGetStartupScriptFileName) (void); /* 159 */
+    void (*reserved160)(void);
+    int (*tclChannelTransform) (Tcl_Interp *interp, Tcl_Channel chan, Tcl_Obj *cmdObjPtr); /* 161 */
+    void (*tclChannelEventScriptInvoker) (ClientData clientData, int flags); /* 162 */
+    const void * (*tclGetInstructionTable) (void); /* 163 */
+    void (*tclExpandCodeArray) (void *envPtr); /* 164 */
+    void (*tclpSetInitialEncodings) (void); /* 165 */
+    int (*tclListObjSetElement) (Tcl_Interp *interp, Tcl_Obj *listPtr, int index, Tcl_Obj *valuePtr); /* 166 */
+    void (*tclSetStartupScriptPath) (Tcl_Obj *pathPtr); /* 167 */
+    Tcl_Obj * (*tclGetStartupScriptPath) (void); /* 168 */
+    int (*tclpUtfNcmp2) (const char *s1, const char *s2, unsigned long n); /* 169 */
+    int (*tclCheckInterpTraces) (Tcl_Interp *interp, const char *command, int numChars, Command *cmdPtr, int result, int traceFlags, int objc, Tcl_Obj *const objv[]); /* 170 */
+    int (*tclCheckExecutionTraces) (Tcl_Interp *interp, const char *command, int numChars, Command *cmdPtr, int result, int traceFlags, int objc, Tcl_Obj *const objv[]); /* 171 */
+    int (*tclInThreadExit) (void); /* 172 */
+    int (*tclUniCharMatch) (const Tcl_UniChar *string, int strLen, const Tcl_UniChar *pattern, int ptnLen, int flags); /* 173 */
+    void (*reserved174)(void);
+    int (*tclCallVarTraces) (Interp *iPtr, Var *arrayPtr, Var *varPtr, const char *part1, const char *part2, int flags, int leaveErrMsg); /* 175 */
+    void (*tclCleanupVar) (Var *varPtr, Var *arrayPtr); /* 176 */
+    void (*tclVarErrMsg) (Tcl_Interp *interp, const char *part1, const char *part2, const char *operation, const char *reason); /* 177 */
+    void (*tcl_SetStartupScript) (Tcl_Obj *pathPtr, const char *encodingName); /* 178 */
+    Tcl_Obj * (*tcl_GetStartupScript) (const char **encodingNamePtr); /* 179 */
+    void (*reserved180)(void);
+    void (*reserved181)(void);
+    struct tm * (*tclpLocaltime) (const time_t *clock); /* 182 */
+    struct tm * (*tclpGmtime) (const time_t *clock); /* 183 */
+    void (*reserved184)(void);
+    void (*reserved185)(void);
+    void (*reserved186)(void);
+    void (*reserved187)(void);
+    void (*reserved188)(void);
+    void (*reserved189)(void);
+    void (*reserved190)(void);
+    void (*reserved191)(void);
+    void (*reserved192)(void);
+    void (*reserved193)(void);
+    void (*reserved194)(void);
+    void (*reserved195)(void);
+    void (*reserved196)(void);
+    void (*reserved197)(void);
+    int (*tclObjGetFrame) (Tcl_Interp *interp, Tcl_Obj *objPtr, CallFrame **framePtrPtr); /* 198 */
+    void (*reserved199)(void);
+    int (*tclpObjRemoveDirectory) (Tcl_Obj *pathPtr, int recursive, Tcl_Obj **errorPtr); /* 200 */
+    int (*tclpObjCopyDirectory) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr, Tcl_Obj **errorPtr); /* 201 */
+    int (*tclpObjCreateDirectory) (Tcl_Obj *pathPtr); /* 202 */
+    int (*tclpObjDeleteFile) (Tcl_Obj *pathPtr); /* 203 */
+    int (*tclpObjCopyFile) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr); /* 204 */
+    int (*tclpObjRenameFile) (Tcl_Obj *srcPathPtr, Tcl_Obj *destPathPtr); /* 205 */
+    int (*tclpObjStat) (Tcl_Obj *pathPtr, Tcl_StatBuf *buf); /* 206 */
+    int (*tclpObjAccess) (Tcl_Obj *pathPtr, int mode); /* 207 */
+    Tcl_Channel (*tclpOpenFileChannel) (Tcl_Interp *interp, Tcl_Obj *pathPtr, int mode, int permissions); /* 208 */
+    void (*reserved209)(void);
+    void (*reserved210)(void);
+    void (*reserved211)(void);
+    void (*tclpFindExecutable) (const char *argv0); /* 212 */
+    Tcl_Obj * (*tclGetObjNameOfExecutable) (void); /* 213 */
+    void (*tclSetObjNameOfExecutable) (Tcl_Obj *name, Tcl_Encoding encoding); /* 214 */
+    void * (*tclStackAlloc) (Tcl_Interp *interp, int numBytes); /* 215 */
+    void (*tclStackFree) (Tcl_Interp *interp, void *freePtr); /* 216 */
+    int (*tclPushStackFrame) (Tcl_Interp *interp, Tcl_CallFrame **framePtrPtr, Tcl_Namespace *namespacePtr, int isProcCallFrame); /* 217 */
+    void (*tclPopStackFrame) (Tcl_Interp *interp); /* 218 */
+    void (*reserved219)(void);
+    void (*reserved220)(void);
+    void (*reserved221)(void);
+    void (*reserved222)(void);
+    void (*reserved223)(void);
+    TclPlatformType * (*tclGetPlatform) (void); /* 224 */
+    Tcl_Obj * (*tclTraceDictPath) (Tcl_Interp *interp, Tcl_Obj *rootPtr, int keyc, Tcl_Obj *const keyv[], int flags); /* 225 */
+    int (*tclObjBeingDeleted) (Tcl_Obj *objPtr); /* 226 */
+    void (*tclSetNsPath) (Namespace *nsPtr, int pathLength, Tcl_Namespace *pathAry[]); /* 227 */
+    void (*reserved228)(void);
+    int (*tclPtrMakeUpvar) (Tcl_Interp *interp, Var *otherP1Ptr, const char *myName, int myFlags, int index); /* 229 */
+    Var * (*tclObjLookupVar) (Tcl_Interp *interp, Tcl_Obj *part1Ptr, const char *part2, int flags, const char *msg, const int createPart1, const int createPart2, Var **arrayPtrPtr); /* 230 */
+    int (*tclGetNamespaceFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr, Tcl_Namespace **nsPtrPtr); /* 231 */
+    int (*tclEvalObjEx) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags, const CmdFrame *invoker, int word); /* 232 */
+    void (*tclGetSrcInfoForPc) (CmdFrame *contextPtr); /* 233 */
+    Var * (*tclVarHashCreateVar) (TclVarHashTable *tablePtr, const char *key, int *newPtr); /* 234 */
+    void (*tclInitVarHashTable) (TclVarHashTable *tablePtr, Namespace *nsPtr); /* 235 */
+    void (*tclBackgroundException) (Tcl_Interp *interp, int code); /* 236 */
+    int (*tclResetCancellation) (Tcl_Interp *interp, int force); /* 237 */
+    int (*tclNRInterpProc) (ClientData clientData, Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]); /* 238 */
+    int (*tclNRInterpProcCore) (Tcl_Interp *interp, Tcl_Obj *procNameObj, int skip, ProcErrorProc *errorProc); /* 239 */
+    int (*tclNRRunCallbacks) (Tcl_Interp *interp, int result, struct NRE_callback *rootPtr); /* 240 */
+    int (*tclNREvalObjEx) (Tcl_Interp *interp, Tcl_Obj *objPtr, int flags, const CmdFrame *invoker, int word); /* 241 */
+    int (*tclNREvalObjv) (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[], int flags, Command *cmdPtr); /* 242 */
+    void (*tclDbDumpActiveObjects) (FILE *outFile); /* 243 */
+    Tcl_HashTable * (*tclGetNamespaceChildTable) (Tcl_Namespace *nsPtr); /* 244 */
+    Tcl_HashTable * (*tclGetNamespaceCommandTable) (Tcl_Namespace *nsPtr); /* 245 */
+    int (*tclInitRewriteEnsemble) (Tcl_Interp *interp, int numRemoved, int numInserted, Tcl_Obj *const *objv); /* 246 */
+    void (*tclResetRewriteEnsemble) (Tcl_Interp *interp, int isRootEnsemble); /* 247 */
+    int (*tclCopyChannel) (Tcl_Interp *interp, Tcl_Channel inChan, Tcl_Channel outChan, Tcl_WideInt toRead, Tcl_Obj *cmdPtr); /* 248 */
+    char * (*tclDoubleDigits) (double dv, int ndigits, int flags, int *decpt, int *signum, char **endPtr); /* 249 */
+    void (*tclSetSlaveCancelFlags) (Tcl_Interp *interp, int flags, int force); /* 250 */
+    int (*tclRegisterLiteral) (void *envPtr, const char *bytes, int length, int flags); /* 251 */
+    Tcl_Obj * (*tclPtrGetVar) (Tcl_Interp *interp, Tcl_Var varPtr, Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, const int flags); /* 252 */
+    Tcl_Obj * (*tclPtrSetVar) (Tcl_Interp *interp, Tcl_Var varPtr, Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, Tcl_Obj *newValuePtr, const int flags); /* 253 */
+    Tcl_Obj * (*tclPtrIncrObjVar) (Tcl_Interp *interp, Tcl_Var varPtr, Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, Tcl_Obj *incrPtr, const int flags); /* 254 */
+    int (*tclPtrObjMakeUpvar) (Tcl_Interp *interp, Tcl_Var otherPtr, Tcl_Obj *myNamePtr, int myFlags); /* 255 */
+    int (*tclPtrUnsetVar) (Tcl_Interp *interp, Tcl_Var varPtr, Tcl_Var arrayPtr, Tcl_Obj *part1Ptr, Tcl_Obj *part2Ptr, const int flags); /* 256 */
+} TclIntStubs;
+
+extern const TclIntStubs *tclIntStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCL_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+/* Slot 0 is reserved */
+/* Slot 1 is reserved */
+/* Slot 2 is reserved */
+#define TclAllocateFreeObjects \
+	(tclIntStubsPtr->tclAllocateFreeObjects) /* 3 */
+/* Slot 4 is reserved */
+#define TclCleanupChildren \
+	(tclIntStubsPtr->tclCleanupChildren) /* 5 */
+#define TclCleanupCommand \
+	(tclIntStubsPtr->tclCleanupCommand) /* 6 */
+#define TclCopyAndCollapse \
+	(tclIntStubsPtr->tclCopyAndCollapse) /* 7 */
+#define TclCopyChannelOld \
+	(tclIntStubsPtr->tclCopyChannelOld) /* 8 */
+#define TclCreatePipeline \
+	(tclIntStubsPtr->tclCreatePipeline) /* 9 */
+#define TclCreateProc \
+	(tclIntStubsPtr->tclCreateProc) /* 10 */
+#define TclDeleteCompiledLocalVars \
+	(tclIntStubsPtr->tclDeleteCompiledLocalVars) /* 11 */
+#define TclDeleteVars \
+	(tclIntStubsPtr->tclDeleteVars) /* 12 */
+/* Slot 13 is reserved */
+#define TclDumpMemoryInfo \
+	(tclIntStubsPtr->tclDumpMemoryInfo) /* 14 */
+/* Slot 15 is reserved */
+#define TclExprFloatError \
+	(tclIntStubsPtr->tclExprFloatError) /* 16 */
+/* Slot 17 is reserved */
+/* Slot 18 is reserved */
+/* Slot 19 is reserved */
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+#define TclFindElement \
+	(tclIntStubsPtr->tclFindElement) /* 22 */
+#define TclFindProc \
+	(tclIntStubsPtr->tclFindProc) /* 23 */
+#define TclFormatInt \
+	(tclIntStubsPtr->tclFormatInt) /* 24 */
+#define TclFreePackageInfo \
+	(tclIntStubsPtr->tclFreePackageInfo) /* 25 */
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+#define TclpGetDefaultStdChannel \
+	(tclIntStubsPtr->tclpGetDefaultStdChannel) /* 28 */
+/* Slot 29 is reserved */
+/* Slot 30 is reserved */
+#define TclGetExtension \
+	(tclIntStubsPtr->tclGetExtension) /* 31 */
+#define TclGetFrame \
+	(tclIntStubsPtr->tclGetFrame) /* 32 */
+/* Slot 33 is reserved */
+#define TclGetIntForIndex \
+	(tclIntStubsPtr->tclGetIntForIndex) /* 34 */
+/* Slot 35 is reserved */
+/* Slot 36 is reserved */
+#define TclGetLoadedPackages \
+	(tclIntStubsPtr->tclGetLoadedPackages) /* 37 */
+#define TclGetNamespaceForQualName \
+	(tclIntStubsPtr->tclGetNamespaceForQualName) /* 38 */
+#define TclGetObjInterpProc \
+	(tclIntStubsPtr->tclGetObjInterpProc) /* 39 */
+#define TclGetOpenMode \
+	(tclIntStubsPtr->tclGetOpenMode) /* 40 */
+#define TclGetOriginalCommand \
+	(tclIntStubsPtr->tclGetOriginalCommand) /* 41 */
+#define TclpGetUserHome \
+	(tclIntStubsPtr->tclpGetUserHome) /* 42 */
+/* Slot 43 is reserved */
+#define TclGuessPackageName \
+	(tclIntStubsPtr->tclGuessPackageName) /* 44 */
+#define TclHideUnsafeCommands \
+	(tclIntStubsPtr->tclHideUnsafeCommands) /* 45 */
+#define TclInExit \
+	(tclIntStubsPtr->tclInExit) /* 46 */
+/* Slot 47 is reserved */
+/* Slot 48 is reserved */
+/* Slot 49 is reserved */
+#define TclInitCompiledLocals \
+	(tclIntStubsPtr->tclInitCompiledLocals) /* 50 */
+#define TclInterpInit \
+	(tclIntStubsPtr->tclInterpInit) /* 51 */
+/* Slot 52 is reserved */
+#define TclInvokeObjectCommand \
+	(tclIntStubsPtr->tclInvokeObjectCommand) /* 53 */
+#define TclInvokeStringCommand \
+	(tclIntStubsPtr->tclInvokeStringCommand) /* 54 */
+#define TclIsProc \
+	(tclIntStubsPtr->tclIsProc) /* 55 */
+/* Slot 56 is reserved */
+/* Slot 57 is reserved */
+#define TclLookupVar \
+	(tclIntStubsPtr->tclLookupVar) /* 58 */
+/* Slot 59 is reserved */
+#define TclNeedSpace \
+	(tclIntStubsPtr->tclNeedSpace) /* 60 */
+#define TclNewProcBodyObj \
+	(tclIntStubsPtr->tclNewProcBodyObj) /* 61 */
+#define TclObjCommandComplete \
+	(tclIntStubsPtr->tclObjCommandComplete) /* 62 */
+#define TclObjInterpProc \
+	(tclIntStubsPtr->tclObjInterpProc) /* 63 */
+#define TclObjInvoke \
+	(tclIntStubsPtr->tclObjInvoke) /* 64 */
+/* Slot 65 is reserved */
+/* Slot 66 is reserved */
+/* Slot 67 is reserved */
+/* Slot 68 is reserved */
+#define TclpAlloc \
+	(tclIntStubsPtr->tclpAlloc) /* 69 */
+/* Slot 70 is reserved */
+/* Slot 71 is reserved */
+/* Slot 72 is reserved */
+/* Slot 73 is reserved */
+#define TclpFree \
+	(tclIntStubsPtr->tclpFree) /* 74 */
+#define TclpGetClicks \
+	(tclIntStubsPtr->tclpGetClicks) /* 75 */
+#define TclpGetSeconds \
+	(tclIntStubsPtr->tclpGetSeconds) /* 76 */
+#define TclpGetTime \
+	(tclIntStubsPtr->tclpGetTime) /* 77 */
+/* Slot 78 is reserved */
+/* Slot 79 is reserved */
+/* Slot 80 is reserved */
+#define TclpRealloc \
+	(tclIntStubsPtr->tclpRealloc) /* 81 */
+/* Slot 82 is reserved */
+/* Slot 83 is reserved */
+/* Slot 84 is reserved */
+/* Slot 85 is reserved */
+/* Slot 86 is reserved */
+/* Slot 87 is reserved */
+#define TclPrecTraceProc \
+	(tclIntStubsPtr->tclPrecTraceProc) /* 88 */
+#define TclPreventAliasLoop \
+	(tclIntStubsPtr->tclPreventAliasLoop) /* 89 */
+/* Slot 90 is reserved */
+#define TclProcCleanupProc \
+	(tclIntStubsPtr->tclProcCleanupProc) /* 91 */
+#define TclProcCompileProc \
+	(tclIntStubsPtr->tclProcCompileProc) /* 92 */
+#define TclProcDeleteProc \
+	(tclIntStubsPtr->tclProcDeleteProc) /* 93 */
+/* Slot 94 is reserved */
+/* Slot 95 is reserved */
+#define TclRenameCommand \
+	(tclIntStubsPtr->tclRenameCommand) /* 96 */
+#define TclResetShadowedCmdRefs \
+	(tclIntStubsPtr->tclResetShadowedCmdRefs) /* 97 */
+#define TclServiceIdle \
+	(tclIntStubsPtr->tclServiceIdle) /* 98 */
+/* Slot 99 is reserved */
+/* Slot 100 is reserved */
+#define TclSetPreInitScript \
+	(tclIntStubsPtr->tclSetPreInitScript) /* 101 */
+#define TclSetupEnv \
+	(tclIntStubsPtr->tclSetupEnv) /* 102 */
+#define TclSockGetPort \
+	(tclIntStubsPtr->tclSockGetPort) /* 103 */
+#define TclSockMinimumBuffersOld \
+	(tclIntStubsPtr->tclSockMinimumBuffersOld) /* 104 */
+/* Slot 105 is reserved */
+/* Slot 106 is reserved */
+/* Slot 107 is reserved */
+#define TclTeardownNamespace \
+	(tclIntStubsPtr->tclTeardownNamespace) /* 108 */
+#define TclUpdateReturnInfo \
+	(tclIntStubsPtr->tclUpdateReturnInfo) /* 109 */
+#define TclSockMinimumBuffers \
+	(tclIntStubsPtr->tclSockMinimumBuffers) /* 110 */
+#define Tcl_AddInterpResolvers \
+	(tclIntStubsPtr->tcl_AddInterpResolvers) /* 111 */
+#define Tcl_AppendExportList \
+	(tclIntStubsPtr->tcl_AppendExportList) /* 112 */
+#define Tcl_CreateNamespace \
+	(tclIntStubsPtr->tcl_CreateNamespace) /* 113 */
+#define Tcl_DeleteNamespace \
+	(tclIntStubsPtr->tcl_DeleteNamespace) /* 114 */
+#define Tcl_Export \
+	(tclIntStubsPtr->tcl_Export) /* 115 */
+#define Tcl_FindCommand \
+	(tclIntStubsPtr->tcl_FindCommand) /* 116 */
+#define Tcl_FindNamespace \
+	(tclIntStubsPtr->tcl_FindNamespace) /* 117 */
+#define Tcl_GetInterpResolvers \
+	(tclIntStubsPtr->tcl_GetInterpResolvers) /* 118 */
+#define Tcl_GetNamespaceResolvers \
+	(tclIntStubsPtr->tcl_GetNamespaceResolvers) /* 119 */
+#define Tcl_FindNamespaceVar \
+	(tclIntStubsPtr->tcl_FindNamespaceVar) /* 120 */
+#define Tcl_ForgetImport \
+	(tclIntStubsPtr->tcl_ForgetImport) /* 121 */
+#define Tcl_GetCommandFromObj \
+	(tclIntStubsPtr->tcl_GetCommandFromObj) /* 122 */
+#define Tcl_GetCommandFullName \
+	(tclIntStubsPtr->tcl_GetCommandFullName) /* 123 */
+#define Tcl_GetCurrentNamespace \
+	(tclIntStubsPtr->tcl_GetCurrentNamespace) /* 124 */
+#define Tcl_GetGlobalNamespace \
+	(tclIntStubsPtr->tcl_GetGlobalNamespace) /* 125 */
+#define Tcl_GetVariableFullName \
+	(tclIntStubsPtr->tcl_GetVariableFullName) /* 126 */
+#define Tcl_Import \
+	(tclIntStubsPtr->tcl_Import) /* 127 */
+#define Tcl_PopCallFrame \
+	(tclIntStubsPtr->tcl_PopCallFrame) /* 128 */
+#define Tcl_PushCallFrame \
+	(tclIntStubsPtr->tcl_PushCallFrame) /* 129 */
+#define Tcl_RemoveInterpResolvers \
+	(tclIntStubsPtr->tcl_RemoveInterpResolvers) /* 130 */
+#define Tcl_SetNamespaceResolvers \
+	(tclIntStubsPtr->tcl_SetNamespaceResolvers) /* 131 */
+#define TclpHasSockets \
+	(tclIntStubsPtr->tclpHasSockets) /* 132 */
+#define TclpGetDate \
+	(tclIntStubsPtr->tclpGetDate) /* 133 */
+/* Slot 134 is reserved */
+/* Slot 135 is reserved */
+/* Slot 136 is reserved */
+/* Slot 137 is reserved */
+#define TclGetEnv \
+	(tclIntStubsPtr->tclGetEnv) /* 138 */
+/* Slot 139 is reserved */
+/* Slot 140 is reserved */
+#define TclpGetCwd \
+	(tclIntStubsPtr->tclpGetCwd) /* 141 */
+#define TclSetByteCodeFromAny \
+	(tclIntStubsPtr->tclSetByteCodeFromAny) /* 142 */
+#define TclAddLiteralObj \
+	(tclIntStubsPtr->tclAddLiteralObj) /* 143 */
+#define TclHideLiteral \
+	(tclIntStubsPtr->tclHideLiteral) /* 144 */
+#define TclGetAuxDataType \
+	(tclIntStubsPtr->tclGetAuxDataType) /* 145 */
+#define TclHandleCreate \
+	(tclIntStubsPtr->tclHandleCreate) /* 146 */
+#define TclHandleFree \
+	(tclIntStubsPtr->tclHandleFree) /* 147 */
+#define TclHandlePreserve \
+	(tclIntStubsPtr->tclHandlePreserve) /* 148 */
+#define TclHandleRelease \
+	(tclIntStubsPtr->tclHandleRelease) /* 149 */
+#define TclRegAbout \
+	(tclIntStubsPtr->tclRegAbout) /* 150 */
+#define TclRegExpRangeUniChar \
+	(tclIntStubsPtr->tclRegExpRangeUniChar) /* 151 */
+#define TclSetLibraryPath \
+	(tclIntStubsPtr->tclSetLibraryPath) /* 152 */
+#define TclGetLibraryPath \
+	(tclIntStubsPtr->tclGetLibraryPath) /* 153 */
+/* Slot 154 is reserved */
+/* Slot 155 is reserved */
+#define TclRegError \
+	(tclIntStubsPtr->tclRegError) /* 156 */
+#define TclVarTraceExists \
+	(tclIntStubsPtr->tclVarTraceExists) /* 157 */
+#define TclSetStartupScriptFileName \
+	(tclIntStubsPtr->tclSetStartupScriptFileName) /* 158 */
+#define TclGetStartupScriptFileName \
+	(tclIntStubsPtr->tclGetStartupScriptFileName) /* 159 */
+/* Slot 160 is reserved */
+#define TclChannelTransform \
+	(tclIntStubsPtr->tclChannelTransform) /* 161 */
+#define TclChannelEventScriptInvoker \
+	(tclIntStubsPtr->tclChannelEventScriptInvoker) /* 162 */
+#define TclGetInstructionTable \
+	(tclIntStubsPtr->tclGetInstructionTable) /* 163 */
+#define TclExpandCodeArray \
+	(tclIntStubsPtr->tclExpandCodeArray) /* 164 */
+#define TclpSetInitialEncodings \
+	(tclIntStubsPtr->tclpSetInitialEncodings) /* 165 */
+#define TclListObjSetElement \
+	(tclIntStubsPtr->tclListObjSetElement) /* 166 */
+#define TclSetStartupScriptPath \
+	(tclIntStubsPtr->tclSetStartupScriptPath) /* 167 */
+#define TclGetStartupScriptPath \
+	(tclIntStubsPtr->tclGetStartupScriptPath) /* 168 */
+#define TclpUtfNcmp2 \
+	(tclIntStubsPtr->tclpUtfNcmp2) /* 169 */
+#define TclCheckInterpTraces \
+	(tclIntStubsPtr->tclCheckInterpTraces) /* 170 */
+#define TclCheckExecutionTraces \
+	(tclIntStubsPtr->tclCheckExecutionTraces) /* 171 */
+#define TclInThreadExit \
+	(tclIntStubsPtr->tclInThreadExit) /* 172 */
+#define TclUniCharMatch \
+	(tclIntStubsPtr->tclUniCharMatch) /* 173 */
+/* Slot 174 is reserved */
+#define TclCallVarTraces \
+	(tclIntStubsPtr->tclCallVarTraces) /* 175 */
+#define TclCleanupVar \
+	(tclIntStubsPtr->tclCleanupVar) /* 176 */
+#define TclVarErrMsg \
+	(tclIntStubsPtr->tclVarErrMsg) /* 177 */
+#define Tcl_SetStartupScript \
+	(tclIntStubsPtr->tcl_SetStartupScript) /* 178 */
+#define Tcl_GetStartupScript \
+	(tclIntStubsPtr->tcl_GetStartupScript) /* 179 */
+/* Slot 180 is reserved */
+/* Slot 181 is reserved */
+#define TclpLocaltime \
+	(tclIntStubsPtr->tclpLocaltime) /* 182 */
+#define TclpGmtime \
+	(tclIntStubsPtr->tclpGmtime) /* 183 */
+/* Slot 184 is reserved */
+/* Slot 185 is reserved */
+/* Slot 186 is reserved */
+/* Slot 187 is reserved */
+/* Slot 188 is reserved */
+/* Slot 189 is reserved */
+/* Slot 190 is reserved */
+/* Slot 191 is reserved */
+/* Slot 192 is reserved */
+/* Slot 193 is reserved */
+/* Slot 194 is reserved */
+/* Slot 195 is reserved */
+/* Slot 196 is reserved */
+/* Slot 197 is reserved */
+#define TclObjGetFrame \
+	(tclIntStubsPtr->tclObjGetFrame) /* 198 */
+/* Slot 199 is reserved */
+#define TclpObjRemoveDirectory \
+	(tclIntStubsPtr->tclpObjRemoveDirectory) /* 200 */
+#define TclpObjCopyDirectory \
+	(tclIntStubsPtr->tclpObjCopyDirectory) /* 201 */
+#define TclpObjCreateDirectory \
+	(tclIntStubsPtr->tclpObjCreateDirectory) /* 202 */
+#define TclpObjDeleteFile \
+	(tclIntStubsPtr->tclpObjDeleteFile) /* 203 */
+#define TclpObjCopyFile \
+	(tclIntStubsPtr->tclpObjCopyFile) /* 204 */
+#define TclpObjRenameFile \
+	(tclIntStubsPtr->tclpObjRenameFile) /* 205 */
+#define TclpObjStat \
+	(tclIntStubsPtr->tclpObjStat) /* 206 */
+#define TclpObjAccess \
+	(tclIntStubsPtr->tclpObjAccess) /* 207 */
+#define TclpOpenFileChannel \
+	(tclIntStubsPtr->tclpOpenFileChannel) /* 208 */
+/* Slot 209 is reserved */
+/* Slot 210 is reserved */
+/* Slot 211 is reserved */
+#define TclpFindExecutable \
+	(tclIntStubsPtr->tclpFindExecutable) /* 212 */
+#define TclGetObjNameOfExecutable \
+	(tclIntStubsPtr->tclGetObjNameOfExecutable) /* 213 */
+#define TclSetObjNameOfExecutable \
+	(tclIntStubsPtr->tclSetObjNameOfExecutable) /* 214 */
+#define TclStackAlloc \
+	(tclIntStubsPtr->tclStackAlloc) /* 215 */
+#define TclStackFree \
+	(tclIntStubsPtr->tclStackFree) /* 216 */
+#define TclPushStackFrame \
+	(tclIntStubsPtr->tclPushStackFrame) /* 217 */
+#define TclPopStackFrame \
+	(tclIntStubsPtr->tclPopStackFrame) /* 218 */
+/* Slot 219 is reserved */
+/* Slot 220 is reserved */
+/* Slot 221 is reserved */
+/* Slot 222 is reserved */
+/* Slot 223 is reserved */
+#define TclGetPlatform \
+	(tclIntStubsPtr->tclGetPlatform) /* 224 */
+#define TclTraceDictPath \
+	(tclIntStubsPtr->tclTraceDictPath) /* 225 */
+#define TclObjBeingDeleted \
+	(tclIntStubsPtr->tclObjBeingDeleted) /* 226 */
+#define TclSetNsPath \
+	(tclIntStubsPtr->tclSetNsPath) /* 227 */
+/* Slot 228 is reserved */
+#define TclPtrMakeUpvar \
+	(tclIntStubsPtr->tclPtrMakeUpvar) /* 229 */
+#define TclObjLookupVar \
+	(tclIntStubsPtr->tclObjLookupVar) /* 230 */
+#define TclGetNamespaceFromObj \
+	(tclIntStubsPtr->tclGetNamespaceFromObj) /* 231 */
+#define TclEvalObjEx \
+	(tclIntStubsPtr->tclEvalObjEx) /* 232 */
+#define TclGetSrcInfoForPc \
+	(tclIntStubsPtr->tclGetSrcInfoForPc) /* 233 */
+#define TclVarHashCreateVar \
+	(tclIntStubsPtr->tclVarHashCreateVar) /* 234 */
+#define TclInitVarHashTable \
+	(tclIntStubsPtr->tclInitVarHashTable) /* 235 */
+#define TclBackgroundException \
+	(tclIntStubsPtr->tclBackgroundException) /* 236 */
+#define TclResetCancellation \
+	(tclIntStubsPtr->tclResetCancellation) /* 237 */
+#define TclNRInterpProc \
+	(tclIntStubsPtr->tclNRInterpProc) /* 238 */
+#define TclNRInterpProcCore \
+	(tclIntStubsPtr->tclNRInterpProcCore) /* 239 */
+#define TclNRRunCallbacks \
+	(tclIntStubsPtr->tclNRRunCallbacks) /* 240 */
+#define TclNREvalObjEx \
+	(tclIntStubsPtr->tclNREvalObjEx) /* 241 */
+#define TclNREvalObjv \
+	(tclIntStubsPtr->tclNREvalObjv) /* 242 */
+#define TclDbDumpActiveObjects \
+	(tclIntStubsPtr->tclDbDumpActiveObjects) /* 243 */
+#define TclGetNamespaceChildTable \
+	(tclIntStubsPtr->tclGetNamespaceChildTable) /* 244 */
+#define TclGetNamespaceCommandTable \
+	(tclIntStubsPtr->tclGetNamespaceCommandTable) /* 245 */
+#define TclInitRewriteEnsemble \
+	(tclIntStubsPtr->tclInitRewriteEnsemble) /* 246 */
+#define TclResetRewriteEnsemble \
+	(tclIntStubsPtr->tclResetRewriteEnsemble) /* 247 */
+#define TclCopyChannel \
+	(tclIntStubsPtr->tclCopyChannel) /* 248 */
+#define TclDoubleDigits \
+	(tclIntStubsPtr->tclDoubleDigits) /* 249 */
+#define TclSetSlaveCancelFlags \
+	(tclIntStubsPtr->tclSetSlaveCancelFlags) /* 250 */
+#define TclRegisterLiteral \
+	(tclIntStubsPtr->tclRegisterLiteral) /* 251 */
+#define TclPtrGetVar \
+	(tclIntStubsPtr->tclPtrGetVar) /* 252 */
+#define TclPtrSetVar \
+	(tclIntStubsPtr->tclPtrSetVar) /* 253 */
+#define TclPtrIncrObjVar \
+	(tclIntStubsPtr->tclPtrIncrObjVar) /* 254 */
+#define TclPtrObjMakeUpvar \
+	(tclIntStubsPtr->tclPtrObjMakeUpvar) /* 255 */
+#define TclPtrUnsetVar \
+	(tclIntStubsPtr->tclPtrUnsetVar) /* 256 */
+
+#endif /* defined(USE_TCL_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#undef TCL_STORAGE_CLASS
+#define TCL_STORAGE_CLASS DLLIMPORT
+
+#undef TclGetStartupScriptFileName
+#undef TclSetStartupScriptFileName
+#undef TclGetStartupScriptPath
+#undef TclSetStartupScriptPath
+#undef TclBackgroundException
+
+#if defined(USE_TCL_STUBS) && defined(TCL_NO_DEPRECATED)
+#   undef Tcl_SetStartupScript
+#   define Tcl_SetStartupScript \
+	    (tclStubsPtr->tcl_SetStartupScript) /* 622 */
+#   undef Tcl_GetStartupScript
+#   define Tcl_GetStartupScript \
+	    (tclStubsPtr->tcl_GetStartupScript) /* 623 */
+#   undef Tcl_CreateNamespace
+#   define Tcl_CreateNamespace \
+	   (tclStubsPtr->tcl_CreateNamespace) /* 506 */
+#   undef Tcl_DeleteNamespace
+#   define Tcl_DeleteNamespace \
+	   (tclStubsPtr->tcl_DeleteNamespace) /* 507 */
+#   undef Tcl_AppendExportList
+#   define Tcl_AppendExportList \
+	   (tclStubsPtr->tcl_AppendExportList) /* 508 */
+#   undef Tcl_Export
+#   define Tcl_Export \
+	   (tclStubsPtr->tcl_Export) /* 509 */
+#   undef Tcl_Import
+#   define Tcl_Import \
+	   (tclStubsPtr->tcl_Import) /* 510 */
+#   undef Tcl_ForgetImport
+#   define Tcl_ForgetImport \
+	   (tclStubsPtr->tcl_ForgetImport) /* 511 */
+#   undef Tcl_GetCurrentNamespace
+#   define Tcl_GetCurrentNamespace \
+	   (tclStubsPtr->tcl_GetCurrentNamespace) /* 512 */
+#   undef Tcl_GetGlobalNamespace
+#   define Tcl_GetGlobalNamespace \
+	   (tclStubsPtr->tcl_GetGlobalNamespace) /* 513 */
+#   undef Tcl_FindNamespace
+#   define Tcl_FindNamespace \
+	   (tclStubsPtr->tcl_FindNamespace) /* 514 */
+#   undef Tcl_FindCommand
+#   define Tcl_FindCommand \
+	   (tclStubsPtr->tcl_FindCommand) /* 515 */
+#   undef Tcl_GetCommandFromObj
+#   define Tcl_GetCommandFromObj \
+	   (tclStubsPtr->tcl_GetCommandFromObj) /* 516 */
+#   undef Tcl_GetCommandFullName
+#   define Tcl_GetCommandFullName \
+	   (tclStubsPtr->tcl_GetCommandFullName) /* 517 */
+#endif
+
+#undef TclCopyChannelOld
+#undef TclSockMinimumBuffersOld
+
+#endif /* _TCLINTDECLS */
diff --git a/generic/tclIntPlatDecls.h b/generic/tclIntPlatDecls.h
new file mode 100644
index 0000000..494d6f1
--- /dev/null
+++ b/generic/tclIntPlatDecls.h
@@ -0,0 +1,567 @@
+/*
+ * tclIntPlatDecls.h --
+ *
+ *	This file contains the declarations for all platform dependent
+ *	unsupported functions that are exported by the Tcl library.  These
+ *	interfaces are not guaranteed to remain the same between
+ *	versions.  Use at your own risk.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * All rights reserved.
+ */
+
+#ifndef _TCLINTPLATDECLS
+#define _TCLINTPLATDECLS
+
+#ifdef _WIN32
+#   define Tcl_DirEntry void
+#   define DIR void
+#endif
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/*
+ * WARNING: This file is automatically generated by the tools/genStubs.tcl
+ * script.  Any modifications to the function declarations below should be made
+ * in the generic/tclInt.decls script.
+ */
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+#if !defined(_WIN32) && !defined(__CYGWIN__) && !defined(MAC_OSX_TCL) /* UNIX */
+/* 0 */
+EXTERN void		TclGetAndDetachPids(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 1 */
+EXTERN int		TclpCloseFile(TclFile file);
+/* 2 */
+EXTERN Tcl_Channel	TclpCreateCommandChannel(TclFile readFile,
+				TclFile writeFile, TclFile errorFile,
+				int numPids, Tcl_Pid *pidPtr);
+/* 3 */
+EXTERN int		TclpCreatePipe(TclFile *readPipe, TclFile *writePipe);
+/* 4 */
+EXTERN int		TclpCreateProcess(Tcl_Interp *interp, int argc,
+				const char **argv, TclFile inputFile,
+				TclFile outputFile, TclFile errorFile,
+				Tcl_Pid *pidPtr);
+/* Slot 5 is reserved */
+/* 6 */
+EXTERN TclFile		TclpMakeFile(Tcl_Channel channel, int direction);
+/* 7 */
+EXTERN TclFile		TclpOpenFile(const char *fname, int mode);
+/* 8 */
+EXTERN int		TclUnixWaitForFile(int fd, int mask, int timeout);
+/* 9 */
+EXTERN TclFile		TclpCreateTempFile(const char *contents);
+/* 10 */
+EXTERN Tcl_DirEntry *	TclpReaddir(DIR *dir);
+/* 11 */
+EXTERN struct tm *	TclpLocaltime_unix(const time_t *clock);
+/* 12 */
+EXTERN struct tm *	TclpGmtime_unix(const time_t *clock);
+/* 13 */
+EXTERN char *		TclpInetNtoa(struct in_addr addr);
+/* 14 */
+EXTERN int		TclUnixCopyFile(const char *src, const char *dst,
+				const Tcl_StatBuf *statBufPtr,
+				int dontCopyAtts);
+/* Slot 15 is reserved */
+/* Slot 16 is reserved */
+/* Slot 17 is reserved */
+/* Slot 18 is reserved */
+/* Slot 19 is reserved */
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+/* Slot 22 is reserved */
+/* Slot 23 is reserved */
+/* Slot 24 is reserved */
+/* Slot 25 is reserved */
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+/* Slot 28 is reserved */
+/* 29 */
+EXTERN int		TclWinCPUID(int index, int *regs);
+/* 30 */
+EXTERN int		TclUnixOpenTemporaryFile(Tcl_Obj *dirObj,
+				Tcl_Obj *basenameObj, Tcl_Obj *extensionObj,
+				Tcl_Obj *resultingNameObj);
+#endif /* UNIX */
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+/* 0 */
+EXTERN void		TclWinConvertError(DWORD errCode);
+/* 1 */
+EXTERN void		TclWinConvertWSAError(DWORD errCode);
+/* 2 */
+EXTERN struct servent *	 TclWinGetServByName(const char *nm,
+				const char *proto);
+/* 3 */
+EXTERN int		TclWinGetSockOpt(SOCKET s, int level, int optname,
+				char *optval, int *optlen);
+/* 4 */
+EXTERN HINSTANCE	TclWinGetTclInstance(void);
+/* 5 */
+EXTERN int		TclUnixWaitForFile(int fd, int mask, int timeout);
+/* 6 */
+EXTERN unsigned short	TclWinNToHS(unsigned short ns);
+/* 7 */
+EXTERN int		TclWinSetSockOpt(SOCKET s, int level, int optname,
+				const char *optval, int optlen);
+/* 8 */
+EXTERN int		TclpGetPid(Tcl_Pid pid);
+/* 9 */
+EXTERN int		TclWinGetPlatformId(void);
+/* 10 */
+EXTERN Tcl_DirEntry *	TclpReaddir(DIR *dir);
+/* 11 */
+EXTERN void		TclGetAndDetachPids(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 12 */
+EXTERN int		TclpCloseFile(TclFile file);
+/* 13 */
+EXTERN Tcl_Channel	TclpCreateCommandChannel(TclFile readFile,
+				TclFile writeFile, TclFile errorFile,
+				int numPids, Tcl_Pid *pidPtr);
+/* 14 */
+EXTERN int		TclpCreatePipe(TclFile *readPipe, TclFile *writePipe);
+/* 15 */
+EXTERN int		TclpCreateProcess(Tcl_Interp *interp, int argc,
+				const char **argv, TclFile inputFile,
+				TclFile outputFile, TclFile errorFile,
+				Tcl_Pid *pidPtr);
+/* 16 */
+EXTERN int		TclpIsAtty(int fd);
+/* 17 */
+EXTERN int		TclUnixCopyFile(const char *src, const char *dst,
+				const Tcl_StatBuf *statBufPtr,
+				int dontCopyAtts);
+/* 18 */
+EXTERN TclFile		TclpMakeFile(Tcl_Channel channel, int direction);
+/* 19 */
+EXTERN TclFile		TclpOpenFile(const char *fname, int mode);
+/* 20 */
+EXTERN void		TclWinAddProcess(HANDLE hProcess, DWORD id);
+/* 21 */
+EXTERN char *		TclpInetNtoa(struct in_addr addr);
+/* 22 */
+EXTERN TclFile		TclpCreateTempFile(const char *contents);
+/* Slot 23 is reserved */
+/* 24 */
+EXTERN char *		TclWinNoBackslash(char *path);
+/* Slot 25 is reserved */
+/* 26 */
+EXTERN void		TclWinSetInterfaces(int wide);
+/* 27 */
+EXTERN void		TclWinFlushDirtyChannels(void);
+/* 28 */
+EXTERN void		TclWinResetInterfaces(void);
+/* 29 */
+EXTERN int		TclWinCPUID(int index, int *regs);
+/* 30 */
+EXTERN int		TclUnixOpenTemporaryFile(Tcl_Obj *dirObj,
+				Tcl_Obj *basenameObj, Tcl_Obj *extensionObj,
+				Tcl_Obj *resultingNameObj);
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+/* 0 */
+EXTERN void		TclGetAndDetachPids(Tcl_Interp *interp,
+				Tcl_Channel chan);
+/* 1 */
+EXTERN int		TclpCloseFile(TclFile file);
+/* 2 */
+EXTERN Tcl_Channel	TclpCreateCommandChannel(TclFile readFile,
+				TclFile writeFile, TclFile errorFile,
+				int numPids, Tcl_Pid *pidPtr);
+/* 3 */
+EXTERN int		TclpCreatePipe(TclFile *readPipe, TclFile *writePipe);
+/* 4 */
+EXTERN int		TclpCreateProcess(Tcl_Interp *interp, int argc,
+				const char **argv, TclFile inputFile,
+				TclFile outputFile, TclFile errorFile,
+				Tcl_Pid *pidPtr);
+/* Slot 5 is reserved */
+/* 6 */
+EXTERN TclFile		TclpMakeFile(Tcl_Channel channel, int direction);
+/* 7 */
+EXTERN TclFile		TclpOpenFile(const char *fname, int mode);
+/* 8 */
+EXTERN int		TclUnixWaitForFile(int fd, int mask, int timeout);
+/* 9 */
+EXTERN TclFile		TclpCreateTempFile(const char *contents);
+/* 10 */
+EXTERN Tcl_DirEntry *	TclpReaddir(DIR *dir);
+/* 11 */
+EXTERN struct tm *	TclpLocaltime_unix(const time_t *clock);
+/* 12 */
+EXTERN struct tm *	TclpGmtime_unix(const time_t *clock);
+/* 13 */
+EXTERN char *		TclpInetNtoa(struct in_addr addr);
+/* 14 */
+EXTERN int		TclUnixCopyFile(const char *src, const char *dst,
+				const Tcl_StatBuf *statBufPtr,
+				int dontCopyAtts);
+/* 15 */
+EXTERN int		TclMacOSXGetFileAttribute(Tcl_Interp *interp,
+				int objIndex, Tcl_Obj *fileName,
+				Tcl_Obj **attributePtrPtr);
+/* 16 */
+EXTERN int		TclMacOSXSetFileAttribute(Tcl_Interp *interp,
+				int objIndex, Tcl_Obj *fileName,
+				Tcl_Obj *attributePtr);
+/* 17 */
+EXTERN int		TclMacOSXCopyFileAttributes(const char *src,
+				const char *dst,
+				const Tcl_StatBuf *statBufPtr);
+/* 18 */
+EXTERN int		TclMacOSXMatchType(Tcl_Interp *interp,
+				const char *pathName, const char *fileName,
+				Tcl_StatBuf *statBufPtr,
+				Tcl_GlobTypeData *types);
+/* 19 */
+EXTERN void		TclMacOSXNotifierAddRunLoopMode(
+				const void *runLoopMode);
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+/* Slot 22 is reserved */
+/* Slot 23 is reserved */
+/* Slot 24 is reserved */
+/* Slot 25 is reserved */
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+/* Slot 28 is reserved */
+/* 29 */
+EXTERN int		TclWinCPUID(int index, int *regs);
+/* 30 */
+EXTERN int		TclUnixOpenTemporaryFile(Tcl_Obj *dirObj,
+				Tcl_Obj *basenameObj, Tcl_Obj *extensionObj,
+				Tcl_Obj *resultingNameObj);
+#endif /* MACOSX */
+
+typedef struct TclIntPlatStubs {
+    int magic;
+    void *hooks;
+
+#if !defined(_WIN32) && !defined(__CYGWIN__) && !defined(MAC_OSX_TCL) /* UNIX */
+    void (*tclGetAndDetachPids) (Tcl_Interp *interp, Tcl_Channel chan); /* 0 */
+    int (*tclpCloseFile) (TclFile file); /* 1 */
+    Tcl_Channel (*tclpCreateCommandChannel) (TclFile readFile, TclFile writeFile, TclFile errorFile, int numPids, Tcl_Pid *pidPtr); /* 2 */
+    int (*tclpCreatePipe) (TclFile *readPipe, TclFile *writePipe); /* 3 */
+    int (*tclpCreateProcess) (Tcl_Interp *interp, int argc, const char **argv, TclFile inputFile, TclFile outputFile, TclFile errorFile, Tcl_Pid *pidPtr); /* 4 */
+    void (*reserved5)(void);
+    TclFile (*tclpMakeFile) (Tcl_Channel channel, int direction); /* 6 */
+    TclFile (*tclpOpenFile) (const char *fname, int mode); /* 7 */
+    int (*tclUnixWaitForFile) (int fd, int mask, int timeout); /* 8 */
+    TclFile (*tclpCreateTempFile) (const char *contents); /* 9 */
+    Tcl_DirEntry * (*tclpReaddir) (DIR *dir); /* 10 */
+    struct tm * (*tclpLocaltime_unix) (const time_t *clock); /* 11 */
+    struct tm * (*tclpGmtime_unix) (const time_t *clock); /* 12 */
+    char * (*tclpInetNtoa) (struct in_addr addr); /* 13 */
+    int (*tclUnixCopyFile) (const char *src, const char *dst, const Tcl_StatBuf *statBufPtr, int dontCopyAtts); /* 14 */
+    void (*reserved15)(void);
+    void (*reserved16)(void);
+    void (*reserved17)(void);
+    void (*reserved18)(void);
+    void (*reserved19)(void);
+    void (*reserved20)(void);
+    void (*reserved21)(void);
+    void (*reserved22)(void);
+    void (*reserved23)(void);
+    void (*reserved24)(void);
+    void (*reserved25)(void);
+    void (*reserved26)(void);
+    void (*reserved27)(void);
+    void (*reserved28)(void);
+    int (*tclWinCPUID) (int index, int *regs); /* 29 */
+    int (*tclUnixOpenTemporaryFile) (Tcl_Obj *dirObj, Tcl_Obj *basenameObj, Tcl_Obj *extensionObj, Tcl_Obj *resultingNameObj); /* 30 */
+#endif /* UNIX */
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+    void (*tclWinConvertError) (DWORD errCode); /* 0 */
+    void (*tclWinConvertWSAError) (DWORD errCode); /* 1 */
+    struct servent * (*tclWinGetServByName) (const char *nm, const char *proto); /* 2 */
+    int (*tclWinGetSockOpt) (SOCKET s, int level, int optname, char *optval, int *optlen); /* 3 */
+    HINSTANCE (*tclWinGetTclInstance) (void); /* 4 */
+    int (*tclUnixWaitForFile) (int fd, int mask, int timeout); /* 5 */
+    unsigned short (*tclWinNToHS) (unsigned short ns); /* 6 */
+    int (*tclWinSetSockOpt) (SOCKET s, int level, int optname, const char *optval, int optlen); /* 7 */
+    int (*tclpGetPid) (Tcl_Pid pid); /* 8 */
+    int (*tclWinGetPlatformId) (void); /* 9 */
+    Tcl_DirEntry * (*tclpReaddir) (DIR *dir); /* 10 */
+    void (*tclGetAndDetachPids) (Tcl_Interp *interp, Tcl_Channel chan); /* 11 */
+    int (*tclpCloseFile) (TclFile file); /* 12 */
+    Tcl_Channel (*tclpCreateCommandChannel) (TclFile readFile, TclFile writeFile, TclFile errorFile, int numPids, Tcl_Pid *pidPtr); /* 13 */
+    int (*tclpCreatePipe) (TclFile *readPipe, TclFile *writePipe); /* 14 */
+    int (*tclpCreateProcess) (Tcl_Interp *interp, int argc, const char **argv, TclFile inputFile, TclFile outputFile, TclFile errorFile, Tcl_Pid *pidPtr); /* 15 */
+    int (*tclpIsAtty) (int fd); /* 16 */
+    int (*tclUnixCopyFile) (const char *src, const char *dst, const Tcl_StatBuf *statBufPtr, int dontCopyAtts); /* 17 */
+    TclFile (*tclpMakeFile) (Tcl_Channel channel, int direction); /* 18 */
+    TclFile (*tclpOpenFile) (const char *fname, int mode); /* 19 */
+    void (*tclWinAddProcess) (HANDLE hProcess, DWORD id); /* 20 */
+    char * (*tclpInetNtoa) (struct in_addr addr); /* 21 */
+    TclFile (*tclpCreateTempFile) (const char *contents); /* 22 */
+    void (*reserved23)(void);
+    char * (*tclWinNoBackslash) (char *path); /* 24 */
+    void (*reserved25)(void);
+    void (*tclWinSetInterfaces) (int wide); /* 26 */
+    void (*tclWinFlushDirtyChannels) (void); /* 27 */
+    void (*tclWinResetInterfaces) (void); /* 28 */
+    int (*tclWinCPUID) (int index, int *regs); /* 29 */
+    int (*tclUnixOpenTemporaryFile) (Tcl_Obj *dirObj, Tcl_Obj *basenameObj, Tcl_Obj *extensionObj, Tcl_Obj *resultingNameObj); /* 30 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    void (*tclGetAndDetachPids) (Tcl_Interp *interp, Tcl_Channel chan); /* 0 */
+    int (*tclpCloseFile) (TclFile file); /* 1 */
+    Tcl_Channel (*tclpCreateCommandChannel) (TclFile readFile, TclFile writeFile, TclFile errorFile, int numPids, Tcl_Pid *pidPtr); /* 2 */
+    int (*tclpCreatePipe) (TclFile *readPipe, TclFile *writePipe); /* 3 */
+    int (*tclpCreateProcess) (Tcl_Interp *interp, int argc, const char **argv, TclFile inputFile, TclFile outputFile, TclFile errorFile, Tcl_Pid *pidPtr); /* 4 */
+    void (*reserved5)(void);
+    TclFile (*tclpMakeFile) (Tcl_Channel channel, int direction); /* 6 */
+    TclFile (*tclpOpenFile) (const char *fname, int mode); /* 7 */
+    int (*tclUnixWaitForFile) (int fd, int mask, int timeout); /* 8 */
+    TclFile (*tclpCreateTempFile) (const char *contents); /* 9 */
+    Tcl_DirEntry * (*tclpReaddir) (DIR *dir); /* 10 */
+    struct tm * (*tclpLocaltime_unix) (const time_t *clock); /* 11 */
+    struct tm * (*tclpGmtime_unix) (const time_t *clock); /* 12 */
+    char * (*tclpInetNtoa) (struct in_addr addr); /* 13 */
+    int (*tclUnixCopyFile) (const char *src, const char *dst, const Tcl_StatBuf *statBufPtr, int dontCopyAtts); /* 14 */
+    int (*tclMacOSXGetFileAttribute) (Tcl_Interp *interp, int objIndex, Tcl_Obj *fileName, Tcl_Obj **attributePtrPtr); /* 15 */
+    int (*tclMacOSXSetFileAttribute) (Tcl_Interp *interp, int objIndex, Tcl_Obj *fileName, Tcl_Obj *attributePtr); /* 16 */
+    int (*tclMacOSXCopyFileAttributes) (const char *src, const char *dst, const Tcl_StatBuf *statBufPtr); /* 17 */
+    int (*tclMacOSXMatchType) (Tcl_Interp *interp, const char *pathName, const char *fileName, Tcl_StatBuf *statBufPtr, Tcl_GlobTypeData *types); /* 18 */
+    void (*tclMacOSXNotifierAddRunLoopMode) (const void *runLoopMode); /* 19 */
+    void (*reserved20)(void);
+    void (*reserved21)(void);
+    void (*reserved22)(void);
+    void (*reserved23)(void);
+    void (*reserved24)(void);
+    void (*reserved25)(void);
+    void (*reserved26)(void);
+    void (*reserved27)(void);
+    void (*reserved28)(void);
+    int (*tclWinCPUID) (int index, int *regs); /* 29 */
+    int (*tclUnixOpenTemporaryFile) (Tcl_Obj *dirObj, Tcl_Obj *basenameObj, Tcl_Obj *extensionObj, Tcl_Obj *resultingNameObj); /* 30 */
+#endif /* MACOSX */
+} TclIntPlatStubs;
+
+extern const TclIntPlatStubs *tclIntPlatStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCL_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#if !defined(_WIN32) && !defined(__CYGWIN__) && !defined(MAC_OSX_TCL) /* UNIX */
+#define TclGetAndDetachPids \
+	(tclIntPlatStubsPtr->tclGetAndDetachPids) /* 0 */
+#define TclpCloseFile \
+	(tclIntPlatStubsPtr->tclpCloseFile) /* 1 */
+#define TclpCreateCommandChannel \
+	(tclIntPlatStubsPtr->tclpCreateCommandChannel) /* 2 */
+#define TclpCreatePipe \
+	(tclIntPlatStubsPtr->tclpCreatePipe) /* 3 */
+#define TclpCreateProcess \
+	(tclIntPlatStubsPtr->tclpCreateProcess) /* 4 */
+/* Slot 5 is reserved */
+#define TclpMakeFile \
+	(tclIntPlatStubsPtr->tclpMakeFile) /* 6 */
+#define TclpOpenFile \
+	(tclIntPlatStubsPtr->tclpOpenFile) /* 7 */
+#define TclUnixWaitForFile \
+	(tclIntPlatStubsPtr->tclUnixWaitForFile) /* 8 */
+#define TclpCreateTempFile \
+	(tclIntPlatStubsPtr->tclpCreateTempFile) /* 9 */
+#define TclpReaddir \
+	(tclIntPlatStubsPtr->tclpReaddir) /* 10 */
+#define TclpLocaltime_unix \
+	(tclIntPlatStubsPtr->tclpLocaltime_unix) /* 11 */
+#define TclpGmtime_unix \
+	(tclIntPlatStubsPtr->tclpGmtime_unix) /* 12 */
+#define TclpInetNtoa \
+	(tclIntPlatStubsPtr->tclpInetNtoa) /* 13 */
+#define TclUnixCopyFile \
+	(tclIntPlatStubsPtr->tclUnixCopyFile) /* 14 */
+/* Slot 15 is reserved */
+/* Slot 16 is reserved */
+/* Slot 17 is reserved */
+/* Slot 18 is reserved */
+/* Slot 19 is reserved */
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+/* Slot 22 is reserved */
+/* Slot 23 is reserved */
+/* Slot 24 is reserved */
+/* Slot 25 is reserved */
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+/* Slot 28 is reserved */
+#define TclWinCPUID \
+	(tclIntPlatStubsPtr->tclWinCPUID) /* 29 */
+#define TclUnixOpenTemporaryFile \
+	(tclIntPlatStubsPtr->tclUnixOpenTemporaryFile) /* 30 */
+#endif /* UNIX */
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+#define TclWinConvertError \
+	(tclIntPlatStubsPtr->tclWinConvertError) /* 0 */
+#define TclWinConvertWSAError \
+	(tclIntPlatStubsPtr->tclWinConvertWSAError) /* 1 */
+#define TclWinGetServByName \
+	(tclIntPlatStubsPtr->tclWinGetServByName) /* 2 */
+#define TclWinGetSockOpt \
+	(tclIntPlatStubsPtr->tclWinGetSockOpt) /* 3 */
+#define TclWinGetTclInstance \
+	(tclIntPlatStubsPtr->tclWinGetTclInstance) /* 4 */
+#define TclUnixWaitForFile \
+	(tclIntPlatStubsPtr->tclUnixWaitForFile) /* 5 */
+#define TclWinNToHS \
+	(tclIntPlatStubsPtr->tclWinNToHS) /* 6 */
+#define TclWinSetSockOpt \
+	(tclIntPlatStubsPtr->tclWinSetSockOpt) /* 7 */
+#define TclpGetPid \
+	(tclIntPlatStubsPtr->tclpGetPid) /* 8 */
+#define TclWinGetPlatformId \
+	(tclIntPlatStubsPtr->tclWinGetPlatformId) /* 9 */
+#define TclpReaddir \
+	(tclIntPlatStubsPtr->tclpReaddir) /* 10 */
+#define TclGetAndDetachPids \
+	(tclIntPlatStubsPtr->tclGetAndDetachPids) /* 11 */
+#define TclpCloseFile \
+	(tclIntPlatStubsPtr->tclpCloseFile) /* 12 */
+#define TclpCreateCommandChannel \
+	(tclIntPlatStubsPtr->tclpCreateCommandChannel) /* 13 */
+#define TclpCreatePipe \
+	(tclIntPlatStubsPtr->tclpCreatePipe) /* 14 */
+#define TclpCreateProcess \
+	(tclIntPlatStubsPtr->tclpCreateProcess) /* 15 */
+#define TclpIsAtty \
+	(tclIntPlatStubsPtr->tclpIsAtty) /* 16 */
+#define TclUnixCopyFile \
+	(tclIntPlatStubsPtr->tclUnixCopyFile) /* 17 */
+#define TclpMakeFile \
+	(tclIntPlatStubsPtr->tclpMakeFile) /* 18 */
+#define TclpOpenFile \
+	(tclIntPlatStubsPtr->tclpOpenFile) /* 19 */
+#define TclWinAddProcess \
+	(tclIntPlatStubsPtr->tclWinAddProcess) /* 20 */
+#define TclpInetNtoa \
+	(tclIntPlatStubsPtr->tclpInetNtoa) /* 21 */
+#define TclpCreateTempFile \
+	(tclIntPlatStubsPtr->tclpCreateTempFile) /* 22 */
+/* Slot 23 is reserved */
+#define TclWinNoBackslash \
+	(tclIntPlatStubsPtr->tclWinNoBackslash) /* 24 */
+/* Slot 25 is reserved */
+#define TclWinSetInterfaces \
+	(tclIntPlatStubsPtr->tclWinSetInterfaces) /* 26 */
+#define TclWinFlushDirtyChannels \
+	(tclIntPlatStubsPtr->tclWinFlushDirtyChannels) /* 27 */
+#define TclWinResetInterfaces \
+	(tclIntPlatStubsPtr->tclWinResetInterfaces) /* 28 */
+#define TclWinCPUID \
+	(tclIntPlatStubsPtr->tclWinCPUID) /* 29 */
+#define TclUnixOpenTemporaryFile \
+	(tclIntPlatStubsPtr->tclUnixOpenTemporaryFile) /* 30 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+#define TclGetAndDetachPids \
+	(tclIntPlatStubsPtr->tclGetAndDetachPids) /* 0 */
+#define TclpCloseFile \
+	(tclIntPlatStubsPtr->tclpCloseFile) /* 1 */
+#define TclpCreateCommandChannel \
+	(tclIntPlatStubsPtr->tclpCreateCommandChannel) /* 2 */
+#define TclpCreatePipe \
+	(tclIntPlatStubsPtr->tclpCreatePipe) /* 3 */
+#define TclpCreateProcess \
+	(tclIntPlatStubsPtr->tclpCreateProcess) /* 4 */
+/* Slot 5 is reserved */
+#define TclpMakeFile \
+	(tclIntPlatStubsPtr->tclpMakeFile) /* 6 */
+#define TclpOpenFile \
+	(tclIntPlatStubsPtr->tclpOpenFile) /* 7 */
+#define TclUnixWaitForFile \
+	(tclIntPlatStubsPtr->tclUnixWaitForFile) /* 8 */
+#define TclpCreateTempFile \
+	(tclIntPlatStubsPtr->tclpCreateTempFile) /* 9 */
+#define TclpReaddir \
+	(tclIntPlatStubsPtr->tclpReaddir) /* 10 */
+#define TclpLocaltime_unix \
+	(tclIntPlatStubsPtr->tclpLocaltime_unix) /* 11 */
+#define TclpGmtime_unix \
+	(tclIntPlatStubsPtr->tclpGmtime_unix) /* 12 */
+#define TclpInetNtoa \
+	(tclIntPlatStubsPtr->tclpInetNtoa) /* 13 */
+#define TclUnixCopyFile \
+	(tclIntPlatStubsPtr->tclUnixCopyFile) /* 14 */
+#define TclMacOSXGetFileAttribute \
+	(tclIntPlatStubsPtr->tclMacOSXGetFileAttribute) /* 15 */
+#define TclMacOSXSetFileAttribute \
+	(tclIntPlatStubsPtr->tclMacOSXSetFileAttribute) /* 16 */
+#define TclMacOSXCopyFileAttributes \
+	(tclIntPlatStubsPtr->tclMacOSXCopyFileAttributes) /* 17 */
+#define TclMacOSXMatchType \
+	(tclIntPlatStubsPtr->tclMacOSXMatchType) /* 18 */
+#define TclMacOSXNotifierAddRunLoopMode \
+	(tclIntPlatStubsPtr->tclMacOSXNotifierAddRunLoopMode) /* 19 */
+/* Slot 20 is reserved */
+/* Slot 21 is reserved */
+/* Slot 22 is reserved */
+/* Slot 23 is reserved */
+/* Slot 24 is reserved */
+/* Slot 25 is reserved */
+/* Slot 26 is reserved */
+/* Slot 27 is reserved */
+/* Slot 28 is reserved */
+#define TclWinCPUID \
+	(tclIntPlatStubsPtr->tclWinCPUID) /* 29 */
+#define TclUnixOpenTemporaryFile \
+	(tclIntPlatStubsPtr->tclUnixOpenTemporaryFile) /* 30 */
+#endif /* MACOSX */
+
+#endif /* defined(USE_TCL_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#undef TCL_STORAGE_CLASS
+#define TCL_STORAGE_CLASS DLLIMPORT
+#undef TclpLocaltime_unix
+#undef TclpGmtime_unix
+#undef TclWinConvertWSAError
+#define TclWinConvertWSAError TclWinConvertError
+#undef TclpInetNtoa
+#define TclpInetNtoa inet_ntoa
+
+#if defined(_WIN32)
+#   undef TclWinNToHS
+#   undef TclWinGetServByName
+#   undef TclWinGetSockOpt
+#   undef TclWinSetSockOpt
+#   define TclWinNToHS ntohs
+#   define TclWinGetServByName getservbyname
+#   define TclWinGetSockOpt getsockopt
+#   define TclWinSetSockOpt setsockopt
+#else
+#   undef TclpGetPid
+#   define TclpGetPid(pid) ((unsigned long) (pid))
+#endif
+
+#endif /* _TCLINTPLATDECLS */
diff --git a/generic/tclInterp.c b/generic/tclInterp.c
new file mode 100644
index 0000000..d9dfd37
--- /dev/null
+++ b/generic/tclInterp.c
@@ -0,0 +1,4833 @@
+/*
+ * tclInterp.c --
+ *
+ *	This file implements the "interp" command which allows creation and
+ *	manipulation of Tcl interpreters from within Tcl scripts.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 2004 Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * A pointer to a string that holds an initialization script that if non-NULL
+ * is evaluated in Tcl_Init() prior to the built-in initialization script
+ * above. This variable can be modified by the function below.
+ */
+
+static const char *tclPreInitScript = NULL;
+
+/* Forward declaration */
+struct Target;
+
+/*
+ * struct Alias:
+ *
+ * Stores information about an alias. Is stored in the slave interpreter and
+ * used by the source command to find the target command in the master when
+ * the source command is invoked.
+ */
+
+typedef struct Alias {
+    Tcl_Obj *token;		/* Token for the alias command in the slave
+				 * interp. This used to be the command name in
+				 * the slave when the alias was first
+				 * created. */
+    Tcl_Interp *targetInterp;	/* Interp in which target command will be
+				 * invoked. */
+    Tcl_Command slaveCmd;	/* Source command in slave interpreter, bound
+				 * to command that invokes the target command
+				 * in the target interpreter. */
+    Tcl_HashEntry *aliasEntryPtr;
+				/* Entry for the alias hash table in slave.
+				 * This is used by alias deletion to remove
+				 * the alias from the slave interpreter alias
+				 * table. */
+    struct Target *targetPtr;	/* Entry for target command in master. This is
+				 * used in the master interpreter to map back
+				 * from the target command to aliases
+				 * redirecting to it. */
+    int objc;			/* Count of Tcl_Obj in the prefix of the
+				 * target command to be invoked in the target
+				 * interpreter. Additional arguments specified
+				 * when calling the alias in the slave interp
+				 * will be appended to the prefix before the
+				 * command is invoked. */
+    Tcl_Obj *objPtr;		/* The first actual prefix object - the target
+				 * command name; this has to be at the end of
+				 * the structure, which will be extended to
+				 * accomodate the remaining objects in the
+				 * prefix. */
+} Alias;
+
+/*
+ *
+ * struct Slave:
+ *
+ * Used by the "interp" command to record and find information about slave
+ * interpreters. Maps from a command name in the master to information about a
+ * slave interpreter, e.g. what aliases are defined in it.
+ */
+
+typedef struct Slave {
+    Tcl_Interp *masterInterp;	/* Master interpreter for this slave. */
+    Tcl_HashEntry *slaveEntryPtr;
+				/* Hash entry in masters slave table for this
+				 * slave interpreter. Used to find this
+				 * record, and used when deleting the slave
+				 * interpreter to delete it from the master's
+				 * table. */
+    Tcl_Interp	*slaveInterp;	/* The slave interpreter. */
+    Tcl_Command interpCmd;	/* Interpreter object command. */
+    Tcl_HashTable aliasTable;	/* Table which maps from names of commands in
+				 * slave interpreter to struct Alias defined
+				 * below. */
+} Slave;
+
+/*
+ * struct Target:
+ *
+ * Maps from master interpreter commands back to the source commands in slave
+ * interpreters. This is needed because aliases can be created between sibling
+ * interpreters and must be deleted when the target interpreter is deleted. In
+ * case they would not be deleted the source interpreter would be left with a
+ * "dangling pointer". One such record is stored in the Master record of the
+ * master interpreter with the master for each alias which directs to a
+ * command in the master. These records are used to remove the source command
+ * for an from a slave if/when the master is deleted. They are organized in a
+ * doubly-linked list attached to the master interpreter.
+ */
+
+typedef struct Target {
+    Tcl_Command	slaveCmd;	/* Command for alias in slave interp. */
+    Tcl_Interp *slaveInterp;	/* Slave Interpreter. */
+    struct Target *nextPtr;	/* Next in list of target records, or NULL if
+				 * at the end of the list of targets. */
+    struct Target *prevPtr;	/* Previous in list of target records, or NULL
+				 * if at the start of the list of targets. */
+} Target;
+
+/*
+ * struct Master:
+ *
+ * This record is used for two purposes: First, slaveTable (a hashtable) maps
+ * from names of commands to slave interpreters. This hashtable is used to
+ * store information about slave interpreters of this interpreter, to map over
+ * all slaves, etc. The second purpose is to store information about all
+ * aliases in slaves (or siblings) which direct to target commands in this
+ * interpreter (using the targetsPtr doubly-linked list).
+ *
+ * NB: the flags field in the interp structure, used with SAFE_INTERP mask
+ * denotes whether the interpreter is safe or not. Safe interpreters have
+ * restricted functionality, can only create safe slave interpreters and can
+ * only load safe extensions.
+ */
+
+typedef struct Master {
+    Tcl_HashTable slaveTable;	/* Hash table for slave interpreters. Maps
+				 * from command names to Slave records. */
+    Target *targetsPtr;		/* The head of a doubly-linked list of all the
+				 * target records which denote aliases from
+				 * slaves or sibling interpreters that direct
+				 * to commands in this interpreter. This list
+				 * is used to remove dangling pointers from
+				 * the slave (or sibling) interpreters when
+				 * this interpreter is deleted. */
+} Master;
+
+/*
+ * The following structure keeps track of all the Master and Slave information
+ * on a per-interp basis.
+ */
+
+typedef struct InterpInfo {
+    Master master;		/* Keeps track of all interps for which this
+				 * interp is the Master. */
+    Slave slave;		/* Information necessary for this interp to
+				 * function as a slave. */
+} InterpInfo;
+
+/*
+ * Limit callbacks handled by scripts are modelled as structures which are
+ * stored in hashes indexed by a two-word key. Note that the type of the
+ * 'type' field in the key is not int; this is to make sure that things are
+ * likely to work properly on 64-bit architectures.
+ */
+
+typedef struct ScriptLimitCallback {
+    Tcl_Interp *interp;		/* The interpreter in which to execute the
+				 * callback. */
+    Tcl_Obj *scriptObj;		/* The script to execute to perform the
+				 * user-defined part of the callback. */
+    int type;			/* What kind of callback is this. */
+    Tcl_HashEntry *entryPtr;	/* The entry in the hash table maintained by
+				 * the target interpreter that refers to this
+				 * callback record, or NULL if the entry has
+				 * already been deleted from that hash
+				 * table. */
+} ScriptLimitCallback;
+
+typedef struct ScriptLimitCallbackKey {
+    Tcl_Interp *interp;		/* The interpreter that the limit callback was
+				 * attached to. This is not the interpreter
+				 * that the callback runs in! */
+    long type;			/* The type of callback that this is. */
+} ScriptLimitCallbackKey;
+
+/*
+ * TIP#143 limit handler internal representation.
+ */
+
+struct LimitHandler {
+    int flags;			/* The state of this particular handler. */
+    Tcl_LimitHandlerProc *handlerProc;
+				/* The handler callback. */
+    ClientData clientData;	/* Opaque argument to the handler callback. */
+    Tcl_LimitHandlerDeleteProc *deleteProc;
+				/* How to delete the clientData. */
+    LimitHandler *prevPtr;	/* Previous item in linked list of
+				 * handlers. */
+    LimitHandler *nextPtr;	/* Next item in linked list of handlers. */
+};
+
+/*
+ * Values for the LimitHandler flags field.
+ *      LIMIT_HANDLER_ACTIVE - Whether the handler is currently being
+ *              processed; handlers are never to be entered reentrantly.
+ *      LIMIT_HANDLER_DELETED - Whether the handler has been deleted. This
+ *              should not normally be observed because when a handler is
+ *              deleted it is also spliced out of the list of handlers, but
+ *              even so we will be careful.
+ */
+
+#define LIMIT_HANDLER_ACTIVE    0x01
+#define LIMIT_HANDLER_DELETED   0x02
+
+
+
+/*
+ * Prototypes for local static functions:
+ */
+
+static int		AliasCreate(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, Tcl_Interp *masterInterp,
+			    Tcl_Obj *namePtr, Tcl_Obj *targetPtr, int objc,
+			    Tcl_Obj *const objv[]);
+static int		AliasDelete(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, Tcl_Obj *namePtr);
+static int		AliasDescribe(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, Tcl_Obj *objPtr);
+static int		AliasList(Tcl_Interp *interp, Tcl_Interp *slaveInterp);
+static int		AliasObjCmd(ClientData dummy,
+			    Tcl_Interp *currentInterp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		AliasNRCmd(ClientData dummy,
+			    Tcl_Interp *currentInterp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		AliasObjCmdDeleteProc(ClientData clientData);
+static Tcl_Interp *	GetInterp(Tcl_Interp *interp, Tcl_Obj *pathPtr);
+static Tcl_Interp *	GetInterp2(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		InterpInfoDeleteProc(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		SlaveBgerror(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, int objc,
+			    Tcl_Obj *const objv[]);
+static Tcl_Interp *	SlaveCreate(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+			    int safe);
+static int		SlaveDebugCmd(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SlaveEval(Tcl_Interp *interp, Tcl_Interp *slaveInterp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SlaveExpose(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		SlaveHide(Tcl_Interp *interp, Tcl_Interp *slaveInterp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SlaveHidden(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp);
+static int		SlaveInvokeHidden(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp,
+			    const char *namespaceName,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SlaveMarkTrusted(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp);
+static int		SlaveObjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static void		SlaveObjCmdDeleteProc(ClientData clientData);
+static int		SlaveRecursionLimit(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		SlaveCommandLimitCmd(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, int consumedObjc,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SlaveTimeLimitCmd(Tcl_Interp *interp,
+			    Tcl_Interp *slaveInterp, int consumedObjc,
+			    int objc, Tcl_Obj *const objv[]);
+static void		InheritLimitsFromMaster(Tcl_Interp *slaveInterp,
+			    Tcl_Interp *masterInterp);
+static void		SetScriptLimitCallback(Tcl_Interp *interp, int type,
+			    Tcl_Interp *targetInterp, Tcl_Obj *scriptObj);
+static void		CallScriptLimitCallback(ClientData clientData,
+			    Tcl_Interp *interp);
+static void		DeleteScriptLimitCallback(ClientData clientData);
+static void		RunLimitHandlers(LimitHandler *handlerPtr,
+			    Tcl_Interp *interp);
+static void		TimeLimitCallback(ClientData clientData);
+
+/* NRE enabling */
+static Tcl_NRPostProc	NRPostInvokeHidden;
+static Tcl_ObjCmdProc	NRInterpCmd;
+static Tcl_ObjCmdProc	NRSlaveCmd;
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetPreInitScript --
+ *
+ *	This routine is used to change the value of the internal variable,
+ *	tclPreInitScript.
+ *
+ * Results:
+ *	Returns the current value of tclPreInitScript.
+ *
+ * Side effects:
+ *	Changes the way Tcl_Init() routine behaves.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+TclSetPreInitScript(
+    const char *string)		/* Pointer to a script. */
+{
+    const char *prevString = tclPreInitScript;
+    tclPreInitScript = string;
+    return(prevString);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Init --
+ *
+ *	This function is typically invoked by Tcl_AppInit functions to find
+ *	and source the "init.tcl" script, which should exist somewhere on the
+ *	Tcl library path.
+ *
+ * Results:
+ *	Returns a standard Tcl completion code and sets the interp's result if
+ *	there is an error.
+ *
+ * Side effects:
+ *	Depends on what's in the init.tcl script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+typedef struct PkgName {
+    struct PkgName *nextPtr;	/* Next in list of package names being initialized. */
+    char name[4];
+} PkgName;
+
+int
+Tcl_Init(
+    Tcl_Interp *interp)		/* Interpreter to initialize. */
+{
+    PkgName pkgName = {NULL, "Tcl"};
+    PkgName **names = TclInitPkgFiles(interp);
+    int result = TCL_ERROR;
+
+    pkgName.nextPtr = *names;
+    *names = &pkgName;
+    if (tclPreInitScript != NULL) {
+	if (Tcl_EvalEx(interp, tclPreInitScript, -1, 0) == TCL_ERROR) {
+	    goto end;
+	}
+    }
+
+    /*
+     * In order to find init.tcl during initialization, the following script
+     * is invoked by Tcl_Init(). It looks in several different directories:
+     *
+     *	$tcl_library		- can specify a primary location, if set, no
+     *				  other locations will be checked. This is the
+     *				  recommended way for a program that embeds
+     *				  Tcl to specifically tell Tcl where to find
+     *				  an init.tcl file.
+     *
+     *	$env(TCL_LIBRARY)	- highest priority so user can always override
+     *				  the search path unless the application has
+     *				  specified an exact directory above
+     *
+     *	$tclDefaultLibrary	- INTERNAL: This variable is set by Tcl on
+     *				  those platforms where it can determine at
+     *				  runtime the directory where it expects the
+     *				  init.tcl file to be. After [tclInit] reads
+     *				  and uses this value, it [unset]s it.
+     *				  External users of Tcl should not make use of
+     *				  the variable to customize [tclInit].
+     *
+     *	$tcl_libPath		- OBSOLETE: This variable is no longer set by
+     *				  Tcl itself, but [tclInit] examines it in
+     *				  case some program that embeds Tcl is
+     *				  customizing [tclInit] by setting this
+     *				  variable to a list of directories in which
+     *				  to search.
+     *
+     *	[tcl::pkgconfig get scriptdir,runtime]
+     *				- the directory determined by configure to be
+     *				  the place where Tcl's script library is to
+     *				  be installed.
+     *
+     * The first directory on this path that contains a valid init.tcl script
+     * will be set as the value of tcl_library.
+     *
+     * Note that this entire search mechanism can be bypassed by defining an
+     * alternate tclInit command before calling Tcl_Init().
+     */
+
+    result = Tcl_EvalEx(interp,
+"if {[namespace which -command tclInit] eq \"\"} {\n"
+"  proc tclInit {} {\n"
+"    global tcl_libPath tcl_library env tclDefaultLibrary\n"
+"    rename tclInit {}\n"
+"    if {[info exists tcl_library]} {\n"
+"	set scripts {{set tcl_library}}\n"
+"    } else {\n"
+"	set scripts {}\n"
+"	if {[info exists env(TCL_LIBRARY)] && ($env(TCL_LIBRARY) ne {})} {\n"
+"	    lappend scripts {set env(TCL_LIBRARY)}\n"
+"	    lappend scripts {\n"
+"if {[regexp ^tcl(.*)$ [file tail $env(TCL_LIBRARY)] -> tail] == 0} continue\n"
+"if {$tail eq [info tclversion]} continue\n"
+"file join [file dirname $env(TCL_LIBRARY)] tcl[info tclversion]}\n"
+"	}\n"
+"	if {[info exists tclDefaultLibrary]} {\n"
+"	    lappend scripts {set tclDefaultLibrary}\n"
+"	} else {\n"
+"	    lappend scripts {::tcl::pkgconfig get scriptdir,runtime}\n"
+"	}\n"
+"	lappend scripts {\n"
+"set parentDir [file dirname [file dirname [info nameofexecutable]]]\n"
+"set grandParentDir [file dirname $parentDir]\n"
+"file join $parentDir lib tcl[info tclversion]} \\\n"
+"	{file join $grandParentDir lib tcl[info tclversion]} \\\n"
+"	{file join $parentDir library} \\\n"
+"	{file join $grandParentDir library} \\\n"
+"	{file join $grandParentDir tcl[info tclversion] library} \\\n"
+"	{file join $grandParentDir tcl[info patchlevel] library} \\\n"
+"	{\n"
+"file join [file dirname $grandParentDir] tcl[info patchlevel] library}\n"
+"	if {[info exists tcl_libPath]\n"
+"		&& [catch {llength $tcl_libPath} len] == 0} {\n"
+"	    for {set i 0} {$i < $len} {incr i} {\n"
+"		lappend scripts [list lindex \\$tcl_libPath $i]\n"
+"	    }\n"
+"	}\n"
+"    }\n"
+"    set dirs {}\n"
+"    set errors {}\n"
+"    foreach script $scripts {\n"
+"	lappend dirs [eval $script]\n"
+"	set tcl_library [lindex $dirs end]\n"
+"	set tclfile [file join $tcl_library init.tcl]\n"
+"	if {[file exists $tclfile]} {\n"
+"	    if {[catch {uplevel #0 [list source $tclfile]} msg opts]} {\n"
+"		append errors \"$tclfile: $msg\n\"\n"
+"		append errors \"[dict get $opts -errorinfo]\n\"\n"
+"		continue\n"
+"	    }\n"
+"	    unset -nocomplain tclDefaultLibrary\n"
+"	    return\n"
+"	}\n"
+"    }\n"
+"    unset -nocomplain tclDefaultLibrary\n"
+"    set msg \"Can't find a usable init.tcl in the following directories: \n\"\n"
+"    append msg \"    $dirs\n\n\"\n"
+"    append msg \"$errors\n\n\"\n"
+"    append msg \"This probably means that Tcl wasn't installed properly.\n\"\n"
+"    error $msg\n"
+"  }\n"
+"}\n"
+"tclInit", -1, 0);
+
+end:
+    *names = (*names)->nextPtr;
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclInterpInit --
+ *
+ *	Initializes the invoking interpreter for using the master, slave and
+ *	safe interp facilities. This is called from inside Tcl_CreateInterp().
+ *
+ * Results:
+ *	Always returns TCL_OK for backwards compatibility.
+ *
+ * Side effects:
+ *	Adds the "interp" command to an interpreter and initializes the
+ *	interpInfoPtr field of the invoking interpreter.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclInterpInit(
+    Tcl_Interp *interp)		/* Interpreter to initialize. */
+{
+    InterpInfo *interpInfoPtr;
+    Master *masterPtr;
+    Slave *slavePtr;
+
+    interpInfoPtr = ckalloc(sizeof(InterpInfo));
+    ((Interp *) interp)->interpInfo = interpInfoPtr;
+
+    masterPtr = &interpInfoPtr->master;
+    Tcl_InitHashTable(&masterPtr->slaveTable, TCL_STRING_KEYS);
+    masterPtr->targetsPtr = NULL;
+
+    slavePtr = &interpInfoPtr->slave;
+    slavePtr->masterInterp	= NULL;
+    slavePtr->slaveEntryPtr	= NULL;
+    slavePtr->slaveInterp	= interp;
+    slavePtr->interpCmd		= NULL;
+    Tcl_InitHashTable(&slavePtr->aliasTable, TCL_STRING_KEYS);
+
+    Tcl_NRCreateCommand(interp, "interp", Tcl_InterpObjCmd, NRInterpCmd,
+	    NULL, NULL);
+
+    Tcl_CallWhenDeleted(interp, InterpInfoDeleteProc, NULL);
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * InterpInfoDeleteProc --
+ *
+ *	Invoked when an interpreter is being deleted. It releases all storage
+ *	used by the master/slave/safe interpreter facilities.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Cleans up storage. Sets the interpInfoPtr field of the interp to NULL.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+InterpInfoDeleteProc(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp)		/* Interp being deleted. All commands for
+				 * slave interps should already be deleted. */
+{
+    InterpInfo *interpInfoPtr;
+    Slave *slavePtr;
+    Master *masterPtr;
+    Target *targetPtr;
+
+    interpInfoPtr = (InterpInfo *) ((Interp *) interp)->interpInfo;
+
+    /*
+     * There shouldn't be any commands left.
+     */
+
+    masterPtr = &interpInfoPtr->master;
+    if (masterPtr->slaveTable.numEntries != 0) {
+	Tcl_Panic("InterpInfoDeleteProc: still exist commands");
+    }
+    Tcl_DeleteHashTable(&masterPtr->slaveTable);
+
+    /*
+     * Tell any interps that have aliases to this interp that they should
+     * delete those aliases. If the other interp was already dead, it would
+     * have removed the target record already.
+     */
+
+    for (targetPtr = masterPtr->targetsPtr; targetPtr != NULL; ) {
+	Target *tmpPtr = targetPtr->nextPtr;
+	Tcl_DeleteCommandFromToken(targetPtr->slaveInterp,
+		targetPtr->slaveCmd);
+	targetPtr = tmpPtr;
+    }
+
+    slavePtr = &interpInfoPtr->slave;
+    if (slavePtr->interpCmd != NULL) {
+	/*
+	 * Tcl_DeleteInterp() was called on this interpreter, rather "interp
+	 * delete" or the equivalent deletion of the command in the master.
+	 * First ensure that the cleanup callback doesn't try to delete the
+	 * interp again.
+	 */
+
+	slavePtr->slaveInterp = NULL;
+	Tcl_DeleteCommandFromToken(slavePtr->masterInterp,
+		slavePtr->interpCmd);
+    }
+
+    /*
+     * There shouldn't be any aliases left.
+     */
+
+    if (slavePtr->aliasTable.numEntries != 0) {
+	Tcl_Panic("InterpInfoDeleteProc: still exist aliases");
+    }
+    Tcl_DeleteHashTable(&slavePtr->aliasTable);
+
+    ckfree(interpInfoPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InterpObjCmd --
+ *
+ *	This function is invoked to process the "interp" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+	/* ARGSUSED */
+int
+Tcl_InterpObjCmd(
+    ClientData clientData,		/* Unused. */
+    Tcl_Interp *interp,			/* Current interpreter. */
+    int objc,				/* Number of arguments. */
+    Tcl_Obj *const objv[])		/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, NRInterpCmd, clientData, objc, objv);
+}
+
+static int
+NRInterpCmd(
+    ClientData clientData,		/* Unused. */
+    Tcl_Interp *interp,			/* Current interpreter. */
+    int objc,				/* Number of arguments. */
+    Tcl_Obj *const objv[])		/* Argument objects. */
+{
+    Tcl_Interp *slaveInterp;
+    int index;
+    static const char *const options[] = {
+	"alias",	"aliases",	"bgerror",	"cancel",
+	"create",	"debug",	"delete",
+	"eval",		"exists",	"expose",
+	"hide",		"hidden",	"issafe",
+	"invokehidden",	"limit",	"marktrusted",	"recursionlimit",
+	"slaves",	"share",	"target",	"transfer",
+	NULL
+    };
+    enum option {
+	OPT_ALIAS,	OPT_ALIASES,	OPT_BGERROR,	OPT_CANCEL,
+	OPT_CREATE,	OPT_DEBUG,	OPT_DELETE,
+	OPT_EVAL,	OPT_EXISTS,	OPT_EXPOSE,
+	OPT_HIDE,	OPT_HIDDEN,	OPT_ISSAFE,
+	OPT_INVOKEHID,	OPT_LIMIT,	OPT_MARKTRUSTED,OPT_RECLIMIT,
+	OPT_SLAVES,	OPT_SHARE,	OPT_TARGET,	OPT_TRANSFER
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "cmd ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], options, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch ((enum option) index) {
+    case OPT_ALIAS: {
+	Tcl_Interp *masterInterp;
+
+	if (objc < 4) {
+	aliasArgs:
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "slavePath slaveCmd ?masterPath masterCmd? ?arg ...?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	if (objc == 4) {
+	    return AliasDescribe(interp, slaveInterp, objv[3]);
+	}
+	if ((objc == 5) && (TclGetString(objv[4])[0] == '\0')) {
+	    return AliasDelete(interp, slaveInterp, objv[3]);
+	}
+	if (objc > 5) {
+	    masterInterp = GetInterp(interp, objv[4]);
+	    if (masterInterp == NULL) {
+		return TCL_ERROR;
+	    }
+
+	    return AliasCreate(interp, slaveInterp, masterInterp, objv[3],
+		    objv[5], objc - 6, objv + 6);
+	}
+	goto aliasArgs;
+    }
+    case OPT_ALIASES:
+	slaveInterp = GetInterp2(interp, objc, objv);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return AliasList(interp, slaveInterp);
+    case OPT_BGERROR:
+	if (objc != 3 && objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path ?cmdPrefix?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveBgerror(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_CANCEL: {
+	int i, flags;
+	Tcl_Obj *resultObjPtr;
+	static const char *const cancelOptions[] = {
+	    "-unwind",	"--",	NULL
+	};
+	enum option {
+	    OPT_UNWIND,	OPT_LAST
+	};
+
+	flags = 0;
+
+	for (i = 2; i < objc; i++) {
+	    if (TclGetString(objv[i])[0] != '-') {
+		break;
+	    }
+	    if (Tcl_GetIndexFromObj(interp, objv[i], cancelOptions, "option",
+		    0, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+
+	    switch ((enum option) index) {
+	    case OPT_UNWIND:
+		/*
+		 * The evaluation stack in the target interp is to be unwound.
+		 */
+
+		flags |= TCL_CANCEL_UNWIND;
+		break;
+	    case OPT_LAST:
+		i++;
+		goto endOfForLoop;
+	    }
+	}
+
+    endOfForLoop:
+	if (i < objc - 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "?-unwind? ?--? ?path? ?result?");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Did they specify a slave interp to cancel the script in progress
+	 * in?  If not, use the current interp.
+	 */
+
+	if (i < objc) {
+	    slaveInterp = GetInterp(interp, objv[i]);
+	    if (slaveInterp == NULL) {
+		return TCL_ERROR;
+	    }
+	    i++;
+	} else {
+	    slaveInterp = interp;
+	}
+
+	if (i < objc) {
+	    resultObjPtr = objv[i];
+
+	    /*
+	     * Tcl_CancelEval removes this reference.
+	     */
+
+	    Tcl_IncrRefCount(resultObjPtr);
+	    i++;
+	} else {
+	    resultObjPtr = NULL;
+	}
+
+	return Tcl_CancelEval(slaveInterp, resultObjPtr, 0, flags);
+    }
+    case OPT_CREATE: {
+	int i, last, safe;
+	Tcl_Obj *slavePtr;
+	char buf[16 + TCL_INTEGER_SPACE];
+	static const char *const createOptions[] = {
+	    "-safe",	"--", NULL
+	};
+	enum option {
+	    OPT_SAFE,	OPT_LAST
+	};
+
+	safe = Tcl_IsSafe(interp);
+
+	/*
+	 * Weird historical rules: "-safe" is accepted at the end, too.
+	 */
+
+	slavePtr = NULL;
+	last = 0;
+	for (i = 2; i < objc; i++) {
+	    if ((last == 0) && (Tcl_GetString(objv[i])[0] == '-')) {
+		if (Tcl_GetIndexFromObj(interp, objv[i], createOptions,
+			"option", 0, &index) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (index == OPT_SAFE) {
+		    safe = 1;
+		    continue;
+		}
+		i++;
+		last = 1;
+	    }
+	    if (slavePtr != NULL) {
+		Tcl_WrongNumArgs(interp, 2, objv, "?-safe? ?--? ?path?");
+		return TCL_ERROR;
+	    }
+	    if (i < objc) {
+		slavePtr = objv[i];
+	    }
+	}
+	buf[0] = '\0';
+	if (slavePtr == NULL) {
+	    /*
+	     * Create an anonymous interpreter -- we choose its name and the
+	     * name of the command. We check that the command name that we use
+	     * for the interpreter does not collide with an existing command
+	     * in the master interpreter.
+	     */
+
+	    for (i = 0; ; i++) {
+		Tcl_CmdInfo cmdInfo;
+
+		sprintf(buf, "interp%d", i);
+		if (Tcl_GetCommandInfo(interp, buf, &cmdInfo) == 0) {
+		    break;
+		}
+	    }
+	    slavePtr = Tcl_NewStringObj(buf, -1);
+	}
+	if (SlaveCreate(interp, slavePtr, safe) == NULL) {
+	    if (buf[0] != '\0') {
+		Tcl_DecrRefCount(slavePtr);
+	    }
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, slavePtr);
+	return TCL_OK;
+    }
+    case OPT_DEBUG:		/* TIP #378 */
+	/*
+	 * Currently only -frame supported, otherwise ?-option ?value??
+	 */
+
+	if (objc < 3 || objc > 5) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path ?-frame ?bool??");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveDebugCmd(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_DELETE: {
+	int i;
+	InterpInfo *iiPtr;
+
+	for (i = 2; i < objc; i++) {
+	    slaveInterp = GetInterp(interp, objv[i]);
+	    if (slaveInterp == NULL) {
+		return TCL_ERROR;
+	    } else if (slaveInterp == interp) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"cannot delete the current interpreter", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			"DELETESELF", NULL);
+		return TCL_ERROR;
+	    }
+	    iiPtr = (InterpInfo *) ((Interp *) slaveInterp)->interpInfo;
+	    Tcl_DeleteCommandFromToken(iiPtr->slave.masterInterp,
+		    iiPtr->slave.interpCmd);
+	}
+	return TCL_OK;
+    }
+    case OPT_EVAL:
+	if (objc < 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path arg ?arg ...?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveEval(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_EXISTS: {
+	int exists = 1;
+
+	slaveInterp = GetInterp2(interp, objc, objv);
+	if (slaveInterp == NULL) {
+	    if (objc > 3) {
+		return TCL_ERROR;
+	    }
+	    Tcl_ResetResult(interp);
+	    exists = 0;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(exists));
+	return TCL_OK;
+    }
+    case OPT_EXPOSE:
+	if ((objc < 4) || (objc > 5)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path hiddenCmdName ?cmdName?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveExpose(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_HIDE:
+	if ((objc < 4) || (objc > 5)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path cmdName ?hiddenCmdName?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveHide(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_HIDDEN:
+	slaveInterp = GetInterp2(interp, objc, objv);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveHidden(interp, slaveInterp);
+    case OPT_ISSAFE:
+	slaveInterp = GetInterp2(interp, objc, objv);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(Tcl_IsSafe(slaveInterp)));
+	return TCL_OK;
+    case OPT_INVOKEHID: {
+	int i;
+	const char *namespaceName;
+	static const char *const hiddenOptions[] = {
+	    "-global",	"-namespace",	"--", NULL
+	};
+	enum hiddenOption {
+	    OPT_GLOBAL,	OPT_NAMESPACE,	OPT_LAST
+	};
+
+	namespaceName = NULL;
+	for (i = 3; i < objc; i++) {
+	    if (TclGetString(objv[i])[0] != '-') {
+		break;
+	    }
+	    if (Tcl_GetIndexFromObj(interp, objv[i], hiddenOptions, "option",
+		    0, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (index == OPT_GLOBAL) {
+		namespaceName = "::";
+	    } else if (index == OPT_NAMESPACE) {
+		if (++i == objc) { /* There must be more arguments. */
+		    break;
+		} else {
+		    namespaceName = TclGetString(objv[i]);
+		}
+	    } else {
+		i++;
+		break;
+	    }
+	}
+	if (objc - i < 1) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "path ?-namespace ns? ?-global? ?--? cmd ?arg ..?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveInvokeHidden(interp, slaveInterp, namespaceName, objc - i,
+		objv + i);
+    }
+    case OPT_LIMIT: {
+	static const char *const limitTypes[] = {
+	    "commands", "time", NULL
+	};
+	enum LimitTypes {
+	    LIMIT_TYPE_COMMANDS, LIMIT_TYPE_TIME
+	};
+	int limitType;
+
+	if (objc < 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "path limitType ?-option value ...?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[3], limitTypes, "limit type", 0,
+		&limitType) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum LimitTypes) limitType) {
+	case LIMIT_TYPE_COMMANDS:
+	    return SlaveCommandLimitCmd(interp, slaveInterp, 4, objc,objv);
+	case LIMIT_TYPE_TIME:
+	    return SlaveTimeLimitCmd(interp, slaveInterp, 4, objc, objv);
+	}
+    }
+    case OPT_MARKTRUSTED:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveMarkTrusted(interp, slaveInterp);
+    case OPT_RECLIMIT:
+	if (objc != 3 && objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path ?newlimit?");
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	return SlaveRecursionLimit(interp, slaveInterp, objc - 3, objv + 3);
+    case OPT_SLAVES: {
+	InterpInfo *iiPtr;
+	Tcl_Obj *resultPtr;
+	Tcl_HashEntry *hPtr;
+	Tcl_HashSearch hashSearch;
+	char *string;
+
+	slaveInterp = GetInterp2(interp, objc, objv);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	iiPtr = (InterpInfo *) ((Interp *) slaveInterp)->interpInfo;
+	resultPtr = Tcl_NewObj();
+	hPtr = Tcl_FirstHashEntry(&iiPtr->master.slaveTable, &hashSearch);
+	for ( ; hPtr != NULL; hPtr = Tcl_NextHashEntry(&hashSearch)) {
+	    string = Tcl_GetHashKey(&iiPtr->master.slaveTable, hPtr);
+	    Tcl_ListObjAppendElement(NULL, resultPtr,
+		    Tcl_NewStringObj(string, -1));
+	}
+	Tcl_SetObjResult(interp, resultPtr);
+	return TCL_OK;
+    }
+    case OPT_TRANSFER:
+    case OPT_SHARE: {
+	Tcl_Interp *masterInterp;	/* The master of the slave. */
+	Tcl_Channel chan;
+
+	if (objc != 5) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "srcPath channelId destPath");
+	    return TCL_ERROR;
+	}
+	masterInterp = GetInterp(interp, objv[2]);
+	if (masterInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	chan = Tcl_GetChannel(masterInterp, TclGetString(objv[3]), NULL);
+	if (chan == NULL) {
+	    Tcl_TransferResult(masterInterp, TCL_OK, interp);
+	    return TCL_ERROR;
+	}
+	slaveInterp = GetInterp(interp, objv[4]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_RegisterChannel(slaveInterp, chan);
+	if (index == OPT_TRANSFER) {
+	    /*
+	     * When transferring, as opposed to sharing, we must unhitch the
+	     * channel from the interpreter where it started.
+	     */
+
+	    if (Tcl_UnregisterChannel(masterInterp, chan) != TCL_OK) {
+		Tcl_TransferResult(masterInterp, TCL_OK, interp);
+		return TCL_ERROR;
+	    }
+	}
+	return TCL_OK;
+    }
+    case OPT_TARGET: {
+	InterpInfo *iiPtr;
+	Tcl_HashEntry *hPtr;
+	Alias *aliasPtr;
+	const char *aliasName;
+
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "path alias");
+	    return TCL_ERROR;
+	}
+
+	slaveInterp = GetInterp(interp, objv[2]);
+	if (slaveInterp == NULL) {
+	    return TCL_ERROR;
+	}
+
+	aliasName = TclGetString(objv[3]);
+
+	iiPtr = (InterpInfo *) ((Interp *) slaveInterp)->interpInfo;
+	hPtr = Tcl_FindHashEntry(&iiPtr->slave.aliasTable, aliasName);
+	if (hPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "alias \"%s\" in path \"%s\" not found",
+		    aliasName, Tcl_GetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ALIAS", aliasName,
+		    NULL);
+	    return TCL_ERROR;
+	}
+	aliasPtr = Tcl_GetHashValue(hPtr);
+	if (Tcl_GetInterpPath(interp, aliasPtr->targetInterp) != TCL_OK) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "target interpreter for alias \"%s\" in path \"%s\" is "
+		    "not my descendant", aliasName, Tcl_GetString(objv[2])));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+		    "TARGETSHROUDED", NULL);
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+    }
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * GetInterp2 --
+ *
+ *	Helper function for Tcl_InterpObjCmd() to convert the interp name
+ *	potentially specified on the command line to an Tcl_Interp.
+ *
+ * Results:
+ *	The return value is the interp specified on the command line, or the
+ *	interp argument itself if no interp was specified on the command line.
+ *	If the interp could not be found or the wrong number of arguments was
+ *	specified on the command line, the return value is NULL and an error
+ *	message is left in the interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static Tcl_Interp *
+GetInterp2(
+    Tcl_Interp *interp,		/* Default interp if no interp was specified
+				 * on the command line. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    if (objc == 2) {
+	return interp;
+    } else if (objc == 3) {
+	return GetInterp(interp, objv[2]);
+    } else {
+	Tcl_WrongNumArgs(interp, 2, objv, "?path?");
+	return NULL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateAlias --
+ *
+ *	Creates an alias between two interpreters.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a new alias, manipulates the result field of slaveInterp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_CreateAlias(
+    Tcl_Interp *slaveInterp,	/* Interpreter for source command. */
+    const char *slaveCmd,	/* Command to install in slave. */
+    Tcl_Interp *targetInterp,	/* Interpreter for target command. */
+    const char *targetCmd,	/* Name of target command. */
+    int argc,			/* How many additional arguments? */
+    const char *const *argv)	/* These are the additional args. */
+{
+    Tcl_Obj *slaveObjPtr, *targetObjPtr;
+    Tcl_Obj **objv;
+    int i;
+    int result;
+
+    objv = TclStackAlloc(slaveInterp, (unsigned) sizeof(Tcl_Obj *) * argc);
+    for (i = 0; i < argc; i++) {
+	objv[i] = Tcl_NewStringObj(argv[i], -1);
+	Tcl_IncrRefCount(objv[i]);
+    }
+
+    slaveObjPtr = Tcl_NewStringObj(slaveCmd, -1);
+    Tcl_IncrRefCount(slaveObjPtr);
+
+    targetObjPtr = Tcl_NewStringObj(targetCmd, -1);
+    Tcl_IncrRefCount(targetObjPtr);
+
+    result = AliasCreate(slaveInterp, slaveInterp, targetInterp, slaveObjPtr,
+	    targetObjPtr, argc, objv);
+
+    for (i = 0; i < argc; i++) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    TclStackFree(slaveInterp, objv);
+    Tcl_DecrRefCount(targetObjPtr);
+    Tcl_DecrRefCount(slaveObjPtr);
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateAliasObj --
+ *
+ *	Object version: Creates an alias between two interpreters.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a new alias.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_CreateAliasObj(
+    Tcl_Interp *slaveInterp,	/* Interpreter for source command. */
+    const char *slaveCmd,	/* Command to install in slave. */
+    Tcl_Interp *targetInterp,	/* Interpreter for target command. */
+    const char *targetCmd,	/* Name of target command. */
+    int objc,			/* How many additional arguments? */
+    Tcl_Obj *const objv[])	/* Argument vector. */
+{
+    Tcl_Obj *slaveObjPtr, *targetObjPtr;
+    int result;
+
+    slaveObjPtr = Tcl_NewStringObj(slaveCmd, -1);
+    Tcl_IncrRefCount(slaveObjPtr);
+
+    targetObjPtr = Tcl_NewStringObj(targetCmd, -1);
+    Tcl_IncrRefCount(targetObjPtr);
+
+    result = AliasCreate(slaveInterp, slaveInterp, targetInterp, slaveObjPtr,
+	    targetObjPtr, objc, objv);
+
+    Tcl_DecrRefCount(slaveObjPtr);
+    Tcl_DecrRefCount(targetObjPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetAlias --
+ *
+ *	Gets information about an alias.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetAlias(
+    Tcl_Interp *interp,		/* Interp to start search from. */
+    const char *aliasName,	/* Name of alias to find. */
+    Tcl_Interp **targetInterpPtr,
+				/* (Return) target interpreter. */
+    const char **targetNamePtr,	/* (Return) name of target command. */
+    int *argcPtr,		/* (Return) count of addnl args. */
+    const char ***argvPtr)	/* (Return) additional arguments. */
+{
+    InterpInfo *iiPtr = (InterpInfo *) ((Interp *) interp)->interpInfo;
+    Tcl_HashEntry *hPtr;
+    Alias *aliasPtr;
+    int i, objc;
+    Tcl_Obj **objv;
+
+    hPtr = Tcl_FindHashEntry(&iiPtr->slave.aliasTable, aliasName);
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"alias \"%s\" not found", aliasName));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ALIAS", aliasName, NULL);
+	return TCL_ERROR;
+    }
+    aliasPtr = Tcl_GetHashValue(hPtr);
+    objc = aliasPtr->objc;
+    objv = &aliasPtr->objPtr;
+
+    if (targetInterpPtr != NULL) {
+	*targetInterpPtr = aliasPtr->targetInterp;
+    }
+    if (targetNamePtr != NULL) {
+	*targetNamePtr = TclGetString(objv[0]);
+    }
+    if (argcPtr != NULL) {
+	*argcPtr = objc - 1;
+    }
+    if (argvPtr != NULL) {
+	*argvPtr = (const char **)
+		ckalloc(sizeof(const char *) * (objc - 1));
+	for (i = 1; i < objc; i++) {
+	    (*argvPtr)[i - 1] = TclGetString(objv[i]);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetAliasObj --
+ *
+ *	Object version: Gets information about an alias.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetAliasObj(
+    Tcl_Interp *interp,		/* Interp to start search from. */
+    const char *aliasName,	/* Name of alias to find. */
+    Tcl_Interp **targetInterpPtr,
+				/* (Return) target interpreter. */
+    const char **targetNamePtr,	/* (Return) name of target command. */
+    int *objcPtr,		/* (Return) count of addnl args. */
+    Tcl_Obj ***objvPtr)		/* (Return) additional args. */
+{
+    InterpInfo *iiPtr = (InterpInfo *) ((Interp *) interp)->interpInfo;
+    Tcl_HashEntry *hPtr;
+    Alias *aliasPtr;
+    int objc;
+    Tcl_Obj **objv;
+
+    hPtr = Tcl_FindHashEntry(&iiPtr->slave.aliasTable, aliasName);
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"alias \"%s\" not found", aliasName));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ALIAS", aliasName, NULL);
+	return TCL_ERROR;
+    }
+    aliasPtr = Tcl_GetHashValue(hPtr);
+    objc = aliasPtr->objc;
+    objv = &aliasPtr->objPtr;
+
+    if (targetInterpPtr != NULL) {
+	*targetInterpPtr = aliasPtr->targetInterp;
+    }
+    if (targetNamePtr != NULL) {
+	*targetNamePtr = TclGetString(objv[0]);
+    }
+    if (objcPtr != NULL) {
+	*objcPtr = objc - 1;
+    }
+    if (objvPtr != NULL) {
+	*objvPtr = objv + 1;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPreventAliasLoop --
+ *
+ *	When defining an alias or renaming a command, prevent an alias loop
+ *	from being formed.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	If TCL_ERROR is returned, the function also stores an error message in
+ *	the interpreter's result object.
+ *
+ * NOTE:
+ *	This function is public internal (instead of being static to this
+ *	file) because it is also used from TclRenameCommand.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPreventAliasLoop(
+    Tcl_Interp *interp,		/* Interp in which to report errors. */
+    Tcl_Interp *cmdInterp,	/* Interp in which the command is being
+				 * defined. */
+    Tcl_Command cmd)		/* Tcl command we are attempting to define. */
+{
+    Command *cmdPtr = (Command *) cmd;
+    Alias *aliasPtr, *nextAliasPtr;
+    Tcl_Command aliasCmd;
+    Command *aliasCmdPtr;
+
+    /*
+     * If we are not creating or renaming an alias, then it is always OK to
+     * create or rename the command.
+     */
+
+    if (cmdPtr->objProc != AliasObjCmd) {
+	return TCL_OK;
+    }
+
+    /*
+     * OK, we are dealing with an alias, so traverse the chain of aliases. If
+     * we encounter the alias we are defining (or renaming to) any in the
+     * chain then we have a loop.
+     */
+
+    aliasPtr = cmdPtr->objClientData;
+    nextAliasPtr = aliasPtr;
+    while (1) {
+	Tcl_Obj *cmdNamePtr;
+
+	/*
+	 * If the target of the next alias in the chain is the same as the
+	 * source alias, we have a loop.
+	 */
+
+	if (Tcl_InterpDeleted(nextAliasPtr->targetInterp)) {
+	    /*
+	     * The slave interpreter can be deleted while creating the alias.
+	     * [Bug #641195]
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "cannot define or rename alias \"%s\": interpreter deleted",
+		    Tcl_GetCommandName(cmdInterp, cmd)));
+	    return TCL_ERROR;
+	}
+	cmdNamePtr = nextAliasPtr->objPtr;
+	aliasCmd = Tcl_FindCommand(nextAliasPtr->targetInterp,
+		TclGetString(cmdNamePtr),
+		Tcl_GetGlobalNamespace(nextAliasPtr->targetInterp),
+		/*flags*/ 0);
+	if (aliasCmd == NULL) {
+	    return TCL_OK;
+	}
+	aliasCmdPtr = (Command *) aliasCmd;
+	if (aliasCmdPtr == cmdPtr) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "cannot define or rename alias \"%s\": would create a loop",
+		    Tcl_GetCommandName(cmdInterp, cmd)));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+		    "ALIASLOOP", NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Otherwise, follow the chain one step further. See if the target
+	 * command is an alias - if so, follow the loop to its target command.
+	 * Otherwise we do not have a loop.
+	 */
+
+	if (aliasCmdPtr->objProc != AliasObjCmd) {
+	    return TCL_OK;
+	}
+	nextAliasPtr = aliasCmdPtr->objClientData;
+    }
+
+    /* NOTREACHED */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasCreate --
+ *
+ *	Helper function to do the work to actually create an alias.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	An alias command is created and entered into the alias table for the
+ *	slave interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AliasCreate(
+    Tcl_Interp *interp,		/* Interp for error reporting. */
+    Tcl_Interp *slaveInterp,	/* Interp where alias cmd will live or from
+				 * which alias will be deleted. */
+    Tcl_Interp *masterInterp,	/* Interp in which target command will be
+				 * invoked. */
+    Tcl_Obj *namePtr,		/* Name of alias cmd. */
+    Tcl_Obj *targetNamePtr,	/* Name of target cmd. */
+    int objc,			/* Additional arguments to store */
+    Tcl_Obj *const objv[])	/* with alias. */
+{
+    Alias *aliasPtr;
+    Tcl_HashEntry *hPtr;
+    Target *targetPtr;
+    Slave *slavePtr;
+    Master *masterPtr;
+    Tcl_Obj **prefv;
+    int isNew, i;
+
+    aliasPtr = ckalloc(sizeof(Alias) + objc * sizeof(Tcl_Obj *));
+    aliasPtr->token = namePtr;
+    Tcl_IncrRefCount(aliasPtr->token);
+    aliasPtr->targetInterp = masterInterp;
+
+    aliasPtr->objc = objc + 1;
+    prefv = &aliasPtr->objPtr;
+
+    *prefv = targetNamePtr;
+    Tcl_IncrRefCount(targetNamePtr);
+    for (i = 0; i < objc; i++) {
+	*(++prefv) = objv[i];
+	Tcl_IncrRefCount(objv[i]);
+    }
+
+    Tcl_Preserve(slaveInterp);
+    Tcl_Preserve(masterInterp);
+
+    if (slaveInterp == masterInterp) {
+	aliasPtr->slaveCmd = Tcl_NRCreateCommand(slaveInterp,
+		TclGetString(namePtr), AliasObjCmd, AliasNRCmd, aliasPtr,
+		AliasObjCmdDeleteProc);
+    } else {
+    aliasPtr->slaveCmd = Tcl_CreateObjCommand(slaveInterp,
+	    TclGetString(namePtr), AliasObjCmd, aliasPtr,
+	    AliasObjCmdDeleteProc);
+    }
+
+    if (TclPreventAliasLoop(interp, slaveInterp,
+	    aliasPtr->slaveCmd) != TCL_OK) {
+	/*
+	 * Found an alias loop! The last call to Tcl_CreateObjCommand made the
+	 * alias point to itself. Delete the command and its alias record. Be
+	 * careful to wipe out its client data first, so the command doesn't
+	 * try to delete itself.
+	 */
+
+	Command *cmdPtr;
+
+	Tcl_DecrRefCount(aliasPtr->token);
+	Tcl_DecrRefCount(targetNamePtr);
+	for (i = 0; i < objc; i++) {
+	    Tcl_DecrRefCount(objv[i]);
+	}
+
+	cmdPtr = (Command *) aliasPtr->slaveCmd;
+	cmdPtr->clientData = NULL;
+	cmdPtr->deleteProc = NULL;
+	cmdPtr->deleteData = NULL;
+	Tcl_DeleteCommandFromToken(slaveInterp, aliasPtr->slaveCmd);
+
+	ckfree(aliasPtr);
+
+	/*
+	 * The result was already set by TclPreventAliasLoop.
+	 */
+
+	Tcl_Release(slaveInterp);
+	Tcl_Release(masterInterp);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make an entry in the alias table. If it already exists, retry.
+     */
+
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+    while (1) {
+	Tcl_Obj *newToken;
+	const char *string;
+
+	string = TclGetString(aliasPtr->token);
+	hPtr = Tcl_CreateHashEntry(&slavePtr->aliasTable, string, &isNew);
+	if (isNew != 0) {
+	    break;
+	}
+
+	/*
+	 * The alias name cannot be used as unique token, it is already taken.
+	 * We can produce a unique token by prepending "::" repeatedly. This
+	 * algorithm is a stop-gap to try to maintain the command name as
+	 * token for most use cases, fearful of possible backwards compat
+	 * problems. A better algorithm would produce unique tokens that need
+	 * not be related to the command name.
+	 *
+	 * ATTENTION: the tests in interp.test and possibly safe.test depend
+	 * on the precise definition of these tokens.
+	 */
+
+	TclNewLiteralStringObj(newToken, "::");
+	Tcl_AppendObjToObj(newToken, aliasPtr->token);
+	Tcl_DecrRefCount(aliasPtr->token);
+	aliasPtr->token = newToken;
+	Tcl_IncrRefCount(aliasPtr->token);
+    }
+
+    aliasPtr->aliasEntryPtr = hPtr;
+    Tcl_SetHashValue(hPtr, aliasPtr);
+
+    /*
+     * Create the new command. We must do it after deleting any old command,
+     * because the alias may be pointing at a renamed alias, as in:
+     *
+     * interp alias {} foo {} bar		# Create an alias "foo"
+     * rename foo zop				# Now rename the alias
+     * interp alias {} foo {} zop		# Now recreate "foo"...
+     */
+
+    targetPtr = ckalloc(sizeof(Target));
+    targetPtr->slaveCmd = aliasPtr->slaveCmd;
+    targetPtr->slaveInterp = slaveInterp;
+
+    masterPtr = &((InterpInfo*) ((Interp*) masterInterp)->interpInfo)->master;
+    targetPtr->nextPtr = masterPtr->targetsPtr;
+    targetPtr->prevPtr = NULL;
+    if (masterPtr->targetsPtr != NULL) {
+	masterPtr->targetsPtr->prevPtr = targetPtr;
+    }
+    masterPtr->targetsPtr = targetPtr;
+    aliasPtr->targetPtr = targetPtr;
+
+    Tcl_SetObjResult(interp, aliasPtr->token);
+
+    Tcl_Release(slaveInterp);
+    Tcl_Release(masterInterp);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasDelete --
+ *
+ *	Deletes the given alias from the slave interpreter given.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Deletes the alias from the slave interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AliasDelete(
+    Tcl_Interp *interp,		/* Interpreter for result & errors. */
+    Tcl_Interp *slaveInterp,	/* Interpreter containing alias. */
+    Tcl_Obj *namePtr)		/* Name of alias to delete. */
+{
+    Slave *slavePtr;
+    Alias *aliasPtr;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * If the alias has been renamed in the slave, the master can still use
+     * the original name (with which it was created) to find the alias to
+     * delete it.
+     */
+
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+    hPtr = Tcl_FindHashEntry(&slavePtr->aliasTable, TclGetString(namePtr));
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"alias \"%s\" not found", TclGetString(namePtr)));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ALIAS",
+		TclGetString(namePtr), NULL);
+	return TCL_ERROR;
+    }
+    aliasPtr = Tcl_GetHashValue(hPtr);
+    Tcl_DeleteCommandFromToken(slaveInterp, aliasPtr->slaveCmd);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasDescribe --
+ *
+ *	Sets the interpreter's result object to a Tcl list describing the
+ *	given alias in the given interpreter: its target command and the
+ *	additional arguments to prepend to any invocation of the alias.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AliasDescribe(
+    Tcl_Interp *interp,		/* Interpreter for result & errors. */
+    Tcl_Interp *slaveInterp,	/* Interpreter containing alias. */
+    Tcl_Obj *namePtr)		/* Name of alias to describe. */
+{
+    Slave *slavePtr;
+    Tcl_HashEntry *hPtr;
+    Alias *aliasPtr;
+    Tcl_Obj *prefixPtr;
+
+    /*
+     * If the alias has been renamed in the slave, the master can still use
+     * the original name (with which it was created) to find the alias to
+     * describe it.
+     */
+
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+    hPtr = Tcl_FindHashEntry(&slavePtr->aliasTable, Tcl_GetString(namePtr));
+    if (hPtr == NULL) {
+	return TCL_OK;
+    }
+    aliasPtr = Tcl_GetHashValue(hPtr);
+    prefixPtr = Tcl_NewListObj(aliasPtr->objc, &aliasPtr->objPtr);
+    Tcl_SetObjResult(interp, prefixPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasList --
+ *
+ *	Computes a list of aliases defined in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AliasList(
+    Tcl_Interp *interp,		/* Interp for data return. */
+    Tcl_Interp *slaveInterp)	/* Interp whose aliases to compute. */
+{
+    Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch hashSearch;
+    Tcl_Obj *resultPtr = Tcl_NewObj();
+    Alias *aliasPtr;
+    Slave *slavePtr;
+
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+
+    entryPtr = Tcl_FirstHashEntry(&slavePtr->aliasTable, &hashSearch);
+    for ( ; entryPtr != NULL; entryPtr = Tcl_NextHashEntry(&hashSearch)) {
+	aliasPtr = Tcl_GetHashValue(entryPtr);
+	Tcl_ListObjAppendElement(NULL, resultPtr, aliasPtr->token);
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasObjCmd --
+ *
+ *	This is the function that services invocations of aliases in a slave
+ *	interpreter. One such command exists for each alias. When invoked,
+ *	this function redirects the invocation to the target command in the
+ *	master interpreter as designated by the Alias record associated with
+ *	this command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Causes forwarding of the invocation; all possible side effects may
+ *	occur as a result of invoking the command to which the invocation is
+ *	forwarded.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AliasNRCmd(
+    ClientData clientData,	/* Alias record. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument vector. */
+{
+    Alias *aliasPtr = clientData;
+    int prefc, cmdc, i;
+    Tcl_Obj **prefv, **cmdv;
+    Tcl_Obj *listPtr;
+    List *listRep;
+    int flags = TCL_EVAL_INVOKE;
+
+    /*
+     * Append the arguments to the command prefix and invoke the command in
+     * the target interp's global namespace.
+     */
+
+    prefc = aliasPtr->objc;
+    prefv = &aliasPtr->objPtr;
+    cmdc = prefc + objc - 1;
+
+    listPtr = Tcl_NewListObj(cmdc, NULL);
+    listRep = listPtr->internalRep.twoPtrValue.ptr1;
+    listRep->elemCount = cmdc;
+    cmdv = &listRep->elements;
+
+    prefv = &aliasPtr->objPtr;
+    memcpy(cmdv, prefv, (size_t) (prefc * sizeof(Tcl_Obj *)));
+    memcpy(cmdv+prefc, objv+1, (size_t) ((objc-1) * sizeof(Tcl_Obj *)));
+
+    for (i=0; i<cmdc; i++) {
+	Tcl_IncrRefCount(cmdv[i]);
+    }
+
+    /*
+     * Use the ensemble rewriting machinery to ensure correct error messages:
+     * only the source command should show, not the full target prefix.
+     */
+
+    if (TclInitRewriteEnsemble(interp, 1, prefc, objv)) {
+	TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL, NULL);
+    }
+    TclSkipTailcall(interp);
+    return Tcl_NREvalObj(interp, listPtr, flags);
+}
+
+static int
+AliasObjCmd(
+    ClientData clientData,	/* Alias record. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument vector. */
+{
+#define ALIAS_CMDV_PREALLOC 10
+    Alias *aliasPtr = clientData;
+    Tcl_Interp *targetInterp = aliasPtr->targetInterp;
+    int result, prefc, cmdc, i;
+    Tcl_Obj **prefv, **cmdv;
+    Tcl_Obj *cmdArr[ALIAS_CMDV_PREALLOC];
+    Interp *tPtr = (Interp *) targetInterp;
+    int isRootEnsemble;
+
+    /*
+     * Append the arguments to the command prefix and invoke the command in
+     * the target interp's global namespace.
+     */
+
+    prefc = aliasPtr->objc;
+    prefv = &aliasPtr->objPtr;
+    cmdc = prefc + objc - 1;
+    if (cmdc <= ALIAS_CMDV_PREALLOC) {
+	cmdv = cmdArr;
+    } else {
+	cmdv = TclStackAlloc(interp, cmdc * sizeof(Tcl_Obj *));
+    }
+
+    memcpy(cmdv, prefv, (size_t) (prefc * sizeof(Tcl_Obj *)));
+    memcpy(cmdv+prefc, objv+1, (size_t) ((objc-1) * sizeof(Tcl_Obj *)));
+
+    Tcl_ResetResult(targetInterp);
+
+    for (i=0; i<cmdc; i++) {
+	Tcl_IncrRefCount(cmdv[i]);
+    }
+
+    /*
+     * Use the ensemble rewriting machinery to ensure correct error messages:
+     * only the source command should show, not the full target prefix.
+     */
+
+    isRootEnsemble = TclInitRewriteEnsemble((Tcl_Interp *)tPtr, 1, prefc, objv);
+
+    /*
+     * Protect the target interpreter if it isn't the same as the source
+     * interpreter so that we can continue to work with it after the target
+     * command completes.
+     */
+
+    if (targetInterp != interp) {
+	Tcl_Preserve(targetInterp);
+    }
+
+    /*
+     * Execute the target command in the target interpreter.
+     */
+
+    result = Tcl_EvalObjv(targetInterp, cmdc, cmdv, TCL_EVAL_INVOKE);
+
+    /*
+     * Clean up the ensemble rewrite info if we set it in the first place.
+     */
+
+    if (isRootEnsemble) {
+	TclResetRewriteEnsemble((Tcl_Interp *)tPtr, 1);
+    }
+
+    /*
+     * If it was a cross-interpreter alias, we need to transfer the result
+     * back to the source interpreter and release the lock we previously set
+     * on the target interpreter.
+     */
+
+    if (targetInterp != interp) {
+	Tcl_TransferResult(targetInterp, result, interp);
+	Tcl_Release(targetInterp);
+    }
+
+    for (i=0; i<cmdc; i++) {
+	Tcl_DecrRefCount(cmdv[i]);
+    }
+    if (cmdv != cmdArr) {
+	TclStackFree(interp, cmdv);
+    }
+    return result;
+#undef ALIAS_CMDV_PREALLOC
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AliasObjCmdDeleteProc --
+ *
+ *	Is invoked when an alias command is deleted in a slave. Cleans up all
+ *	storage associated with this alias.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deletes the alias record and its entry in the alias table for the
+ *	interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AliasObjCmdDeleteProc(
+    ClientData clientData)	/* The alias record for this alias. */
+{
+    Alias *aliasPtr = clientData;
+    Target *targetPtr;
+    int i;
+    Tcl_Obj **objv;
+
+    Tcl_DecrRefCount(aliasPtr->token);
+    objv = &aliasPtr->objPtr;
+    for (i = 0; i < aliasPtr->objc; i++) {
+	Tcl_DecrRefCount(objv[i]);
+    }
+    Tcl_DeleteHashEntry(aliasPtr->aliasEntryPtr);
+
+    /*
+     * Splice the target record out of the target interpreter's master list.
+     */
+
+    targetPtr = aliasPtr->targetPtr;
+    if (targetPtr->prevPtr != NULL) {
+	targetPtr->prevPtr->nextPtr = targetPtr->nextPtr;
+    } else {
+	Master *masterPtr = &((InterpInfo *) ((Interp *)
+		aliasPtr->targetInterp)->interpInfo)->master;
+
+	masterPtr->targetsPtr = targetPtr->nextPtr;
+    }
+    if (targetPtr->nextPtr != NULL) {
+	targetPtr->nextPtr->prevPtr = targetPtr->prevPtr;
+    }
+
+    ckfree(targetPtr);
+    ckfree(aliasPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateSlave --
+ *
+ *	Creates a slave interpreter. The slavePath argument denotes the name
+ *	of the new slave relative to the current interpreter; the slave is a
+ *	direct descendant of the one-before-last component of the path,
+ *	e.g. it is a descendant of the current interpreter if the slavePath
+ *	argument contains only one component. Optionally makes the slave
+ *	interpreter safe.
+ *
+ * Results:
+ *	Returns the interpreter structure created, or NULL if an error
+ *	occurred.
+ *
+ * Side effects:
+ *	Creates a new interpreter and a new interpreter object command in the
+ *	interpreter indicated by the slavePath argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Interp *
+Tcl_CreateSlave(
+    Tcl_Interp *interp,		/* Interpreter to start search at. */
+    const char *slavePath,	/* Name of slave to create. */
+    int isSafe)			/* Should new slave be "safe" ? */
+{
+    Tcl_Obj *pathPtr;
+    Tcl_Interp *slaveInterp;
+
+    pathPtr = Tcl_NewStringObj(slavePath, -1);
+    slaveInterp = SlaveCreate(interp, pathPtr, isSafe);
+    Tcl_DecrRefCount(pathPtr);
+
+    return slaveInterp;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetSlave --
+ *
+ *	Finds a slave interpreter by its path name.
+ *
+ * Results:
+ *	Returns a Tcl_Interp * for the named interpreter or NULL if not found.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Interp *
+Tcl_GetSlave(
+    Tcl_Interp *interp,		/* Interpreter to start search from. */
+    const char *slavePath)	/* Path of slave to find. */
+{
+    Tcl_Obj *pathPtr;
+    Tcl_Interp *slaveInterp;
+
+    pathPtr = Tcl_NewStringObj(slavePath, -1);
+    slaveInterp = GetInterp(interp, pathPtr);
+    Tcl_DecrRefCount(pathPtr);
+
+    return slaveInterp;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetMaster --
+ *
+ *	Finds the master interpreter of a slave interpreter.
+ *
+ * Results:
+ *	Returns a Tcl_Interp * for the master interpreter or NULL if none.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Interp *
+Tcl_GetMaster(
+    Tcl_Interp *interp)		/* Get the master of this interpreter. */
+{
+    Slave *slavePtr;		/* Slave record of this interpreter. */
+
+    if (interp == NULL) {
+	return NULL;
+    }
+    slavePtr = &((InterpInfo *) ((Interp *) interp)->interpInfo)->slave;
+    return slavePtr->masterInterp;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetSlaveCancelFlags --
+ *
+ *	This function marks all slave interpreters belonging to a given
+ *	interpreter as being canceled or not canceled, depending on the
+ *	provided flags.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetSlaveCancelFlags(
+    Tcl_Interp *interp,		/* Set cancel flags of this interpreter. */
+    int flags,			/* Collection of OR-ed bits that control
+				 * the cancellation of the script. Only
+				 * TCL_CANCEL_UNWIND is currently
+				 * supported. */
+    int force)			/* Non-zero to ignore numLevels for the purpose
+				 * of resetting the cancellation flags. */
+{
+    Master *masterPtr;		/* Master record of given interpreter. */
+    Tcl_HashEntry *hPtr;	/* Search element. */
+    Tcl_HashSearch hashSearch;	/* Search variable. */
+    Slave *slavePtr;		/* Slave record of interpreter. */
+    Interp *iPtr;
+
+    if (interp == NULL) {
+	return;
+    }
+
+    flags &= (CANCELED | TCL_CANCEL_UNWIND);
+
+    masterPtr = &((InterpInfo *) ((Interp *) interp)->interpInfo)->master;
+
+    hPtr = Tcl_FirstHashEntry(&masterPtr->slaveTable, &hashSearch);
+    for ( ; hPtr != NULL; hPtr = Tcl_NextHashEntry(&hashSearch)) {
+	slavePtr = Tcl_GetHashValue(hPtr);
+	iPtr = (Interp *) slavePtr->slaveInterp;
+
+	if (iPtr == NULL) {
+	    continue;
+	}
+
+	if (flags == 0) {
+	    TclResetCancellation((Tcl_Interp *) iPtr, force);
+	} else {
+	    TclSetCancelFlags(iPtr, flags);
+	}
+
+	/*
+	 * Now, recursively handle this for the slaves of this slave
+	 * interpreter.
+	 */
+
+	TclSetSlaveCancelFlags((Tcl_Interp *) iPtr, flags, force);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetInterpPath --
+ *
+ *	Sets the result of the asking interpreter to a proper Tcl list
+ *	containing the names of interpreters between the asking and target
+ *	interpreters. The target interpreter must be either the same as the
+ *	asking interpreter or one of its slaves (including recursively).
+ *
+ * Results:
+ *	TCL_OK if the target interpreter is the same as, or a descendant of,
+ *	the asking interpreter; TCL_ERROR else. This way one can distinguish
+ *	between the case where the asking and target interps are the same (an
+ *	empty list is the result, and TCL_OK is returned) and when the target
+ *	is not a descendant of the asking interpreter (in which case the Tcl
+ *	result is an error message and the function returns TCL_ERROR).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetInterpPath(
+    Tcl_Interp *askingInterp,	/* Interpreter to start search from. */
+    Tcl_Interp *targetInterp)	/* Interpreter to find. */
+{
+    InterpInfo *iiPtr;
+
+    if (targetInterp == askingInterp) {
+	Tcl_SetObjResult(askingInterp, Tcl_NewObj());
+	return TCL_OK;
+    }
+    if (targetInterp == NULL) {
+	return TCL_ERROR;
+    }
+    iiPtr = (InterpInfo *) ((Interp *) targetInterp)->interpInfo;
+    if (Tcl_GetInterpPath(askingInterp, iiPtr->slave.masterInterp) != TCL_OK){
+	return TCL_ERROR;
+    }
+    Tcl_ListObjAppendElement(NULL, Tcl_GetObjResult(askingInterp),
+	    Tcl_NewStringObj(Tcl_GetHashKey(&iiPtr->master.slaveTable,
+		    iiPtr->slave.slaveEntryPtr), -1));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetInterp --
+ *
+ *	Helper function to find a slave interpreter given a pathname.
+ *
+ * Results:
+ *	Returns the slave interpreter known by that name in the calling
+ *	interpreter, or NULL if no interpreter known by that name exists.
+ *
+ * Side effects:
+ *	Assigns to the pointer variable passed in, if not NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Interp *
+GetInterp(
+    Tcl_Interp *interp,		/* Interp. to start search from. */
+    Tcl_Obj *pathPtr)		/* List object containing name of interp. to
+				 * be found. */
+{
+    Tcl_HashEntry *hPtr;	/* Search element. */
+    Slave *slavePtr;		/* Interim slave record. */
+    Tcl_Obj **objv;
+    int objc, i;
+    Tcl_Interp *searchInterp;	/* Interim storage for interp. to find. */
+    InterpInfo *masterInfoPtr;
+
+    if (TclListObjGetElements(interp, pathPtr, &objc, &objv) != TCL_OK) {
+	return NULL;
+    }
+
+    searchInterp = interp;
+    for (i = 0; i < objc; i++) {
+	masterInfoPtr = (InterpInfo *) ((Interp *) searchInterp)->interpInfo;
+	hPtr = Tcl_FindHashEntry(&masterInfoPtr->master.slaveTable,
+		TclGetString(objv[i]));
+	if (hPtr == NULL) {
+	    searchInterp = NULL;
+	    break;
+	}
+	slavePtr = Tcl_GetHashValue(hPtr);
+	searchInterp = slavePtr->slaveInterp;
+	if (searchInterp == NULL) {
+	    break;
+	}
+    }
+    if (searchInterp == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"could not find interpreter \"%s\"", TclGetString(pathPtr)));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INTERP",
+		TclGetString(pathPtr), NULL);
+    }
+    return searchInterp;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveBgerror --
+ *
+ *	Helper function to set/query the background error handling command
+ *	prefix of an interp
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	When (objc == 1), slaveInterp will be set to a new background handler
+ *	of objv[0].
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveBgerror(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* Interp in which limit is set/queried. */
+    int objc,			/* Set or Query. */
+    Tcl_Obj *const objv[])	/* Argument strings. */
+{
+    if (objc) {
+	int length;
+
+	if (TCL_ERROR == TclListObjLength(NULL, objv[0], &length)
+		|| (length < 1)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "cmdPrefix must be list of length >= 1", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+		    "BGERRORFORMAT", NULL);
+	    return TCL_ERROR;
+	}
+	TclSetBgErrorHandler(slaveInterp, objv[0]);
+    }
+    Tcl_SetObjResult(interp, TclGetBgErrorHandler(slaveInterp));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveCreate --
+ *
+ *	Helper function to do the actual work of creating a slave interp and
+ *	new object command. Also optionally makes the new slave interpreter
+ *	"safe".
+ *
+ * Results:
+ *	Returns the new Tcl_Interp * if successful or NULL if not. If failed,
+ *	the result of the invoking interpreter contains an error message.
+ *
+ * Side effects:
+ *	Creates a new slave interpreter and a new object command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Interp *
+SlaveCreate(
+    Tcl_Interp *interp,		/* Interp. to start search from. */
+    Tcl_Obj *pathPtr,		/* Path (name) of slave to create. */
+    int safe)			/* Should we make it "safe"? */
+{
+    Tcl_Interp *masterInterp, *slaveInterp;
+    Slave *slavePtr;
+    InterpInfo *masterInfoPtr;
+    Tcl_HashEntry *hPtr;
+    const char *path;
+    int isNew, objc;
+    Tcl_Obj **objv;
+
+    if (Tcl_ListObjGetElements(interp, pathPtr, &objc, &objv) != TCL_OK) {
+	return NULL;
+    }
+    if (objc < 2) {
+	masterInterp = interp;
+	path = TclGetString(pathPtr);
+    } else {
+	Tcl_Obj *objPtr;
+
+	objPtr = Tcl_NewListObj(objc - 1, objv);
+	masterInterp = GetInterp(interp, objPtr);
+	Tcl_DecrRefCount(objPtr);
+	if (masterInterp == NULL) {
+	    return NULL;
+	}
+	path = TclGetString(objv[objc - 1]);
+    }
+    if (safe == 0) {
+	safe = Tcl_IsSafe(masterInterp);
+    }
+
+    masterInfoPtr = (InterpInfo *) ((Interp *) masterInterp)->interpInfo;
+    hPtr = Tcl_CreateHashEntry(&masterInfoPtr->master.slaveTable, path,
+	    &isNew);
+    if (isNew == 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"interpreter named \"%s\" already exists, cannot create",
+		path));
+	return NULL;
+    }
+
+    slaveInterp = Tcl_CreateInterp();
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+    slavePtr->masterInterp = masterInterp;
+    slavePtr->slaveEntryPtr = hPtr;
+    slavePtr->slaveInterp = slaveInterp;
+    slavePtr->interpCmd = Tcl_NRCreateCommand(masterInterp, path,
+	    SlaveObjCmd, NRSlaveCmd, slaveInterp, SlaveObjCmdDeleteProc);
+    Tcl_InitHashTable(&slavePtr->aliasTable, TCL_STRING_KEYS);
+    Tcl_SetHashValue(hPtr, slavePtr);
+    Tcl_SetVar2(slaveInterp, "tcl_interactive", NULL, "0", TCL_GLOBAL_ONLY);
+
+    /*
+     * Inherit the recursion limit.
+     */
+
+    ((Interp *) slaveInterp)->maxNestingDepth =
+	    ((Interp *) masterInterp)->maxNestingDepth;
+
+    if (safe) {
+	if (Tcl_MakeSafe(slaveInterp) == TCL_ERROR) {
+	    goto error;
+	}
+    } else {
+	if (Tcl_Init(slaveInterp) == TCL_ERROR) {
+	    goto error;
+	}
+
+	/*
+	 * This will create the "memory" command in slave interpreters if we
+	 * compiled with TCL_MEM_DEBUG, otherwise it does nothing.
+	 */
+
+	Tcl_InitMemory(slaveInterp);
+    }
+
+    /*
+     * Inherit the TIP#143 limits.
+     */
+
+    InheritLimitsFromMaster(slaveInterp, masterInterp);
+
+    /*
+     * The [clock] command presents a safe API, but uses unsafe features in
+     * its implementation. This means it has to be implemented in safe interps
+     * as an alias to a version in the (trusted) master.
+     */
+
+    if (safe) {
+	Tcl_Obj *clockObj;
+	int status;
+
+	TclNewLiteralStringObj(clockObj, "clock");
+	Tcl_IncrRefCount(clockObj);
+	status = AliasCreate(interp, slaveInterp, masterInterp, clockObj,
+		clockObj, 0, NULL);
+	Tcl_DecrRefCount(clockObj);
+	if (status != TCL_OK) {
+	    goto error2;
+	}
+    }
+
+    return slaveInterp;
+
+  error:
+    Tcl_TransferResult(slaveInterp, TCL_ERROR, interp);
+  error2:
+    Tcl_DeleteInterp(slaveInterp);
+
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveObjCmd --
+ *
+ *	Command to manipulate an interpreter, e.g. to send commands to it to
+ *	be evaluated. One such command exists for each slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See user documentation for details.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveObjCmd(
+    ClientData clientData,	/* Slave interpreter. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, NRSlaveCmd, clientData, objc, objv);
+}
+
+static int
+NRSlaveCmd(
+    ClientData clientData,	/* Slave interpreter. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Interp *slaveInterp = clientData;
+    int index;
+    static const char *const options[] = {
+	"alias",	"aliases",	"bgerror",	"debug",
+	"eval",		"expose",	"hide",		"hidden",
+	"issafe",	"invokehidden",	"limit",	"marktrusted",
+	"recursionlimit", NULL
+    };
+    enum options {
+	OPT_ALIAS,	OPT_ALIASES,	OPT_BGERROR,	OPT_DEBUG,
+	OPT_EVAL,	OPT_EXPOSE,	OPT_HIDE,	OPT_HIDDEN,
+	OPT_ISSAFE,	OPT_INVOKEHIDDEN, OPT_LIMIT,	OPT_MARKTRUSTED,
+	OPT_RECLIMIT
+    };
+
+    if (slaveInterp == NULL) {
+	Tcl_Panic("SlaveObjCmd: interpreter has been deleted");
+    }
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "cmd ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], options, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum options) index) {
+    case OPT_ALIAS:
+	if (objc > 2) {
+	    if (objc == 3) {
+		return AliasDescribe(interp, slaveInterp, objv[2]);
+	    }
+	    if (TclGetString(objv[3])[0] == '\0') {
+		if (objc == 4) {
+		    return AliasDelete(interp, slaveInterp, objv[2]);
+		}
+	    } else {
+		return AliasCreate(interp, slaveInterp, interp, objv[2],
+			objv[3], objc - 4, objv + 4);
+	    }
+	}
+	Tcl_WrongNumArgs(interp, 2, objv, "aliasName ?targetName? ?arg ...?");
+	return TCL_ERROR;
+    case OPT_ALIASES:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return AliasList(interp, slaveInterp);
+    case OPT_BGERROR:
+	if (objc != 2 && objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?cmdPrefix?");
+	    return TCL_ERROR;
+	}
+	return SlaveBgerror(interp, slaveInterp, objc - 2, objv + 2);
+    case OPT_DEBUG:
+	/*
+	 * TIP #378
+	 * Currently only -frame supported, otherwise ?-option ?value? ...?
+	 */
+	if (objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?-frame ?bool??");
+	    return TCL_ERROR;
+	}
+	return SlaveDebugCmd(interp, slaveInterp, objc - 2, objv + 2);
+    case OPT_EVAL:
+	if (objc < 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "arg ?arg ...?");
+	    return TCL_ERROR;
+	}
+	return SlaveEval(interp, slaveInterp, objc - 2, objv + 2);
+    case OPT_EXPOSE:
+	if ((objc < 3) || (objc > 4)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "hiddenCmdName ?cmdName?");
+	    return TCL_ERROR;
+	}
+	return SlaveExpose(interp, slaveInterp, objc - 2, objv + 2);
+    case OPT_HIDE:
+	if ((objc < 3) || (objc > 4)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "cmdName ?hiddenCmdName?");
+	    return TCL_ERROR;
+	}
+	return SlaveHide(interp, slaveInterp, objc - 2, objv + 2);
+    case OPT_HIDDEN:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return SlaveHidden(interp, slaveInterp);
+    case OPT_ISSAFE:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(Tcl_IsSafe(slaveInterp)));
+	return TCL_OK;
+    case OPT_INVOKEHIDDEN: {
+	int i;
+	const char *namespaceName;
+	static const char *const hiddenOptions[] = {
+	    "-global",	"-namespace",	"--", NULL
+	};
+	enum hiddenOption {
+	    OPT_GLOBAL,	OPT_NAMESPACE,	OPT_LAST
+	};
+
+	namespaceName = NULL;
+	for (i = 2; i < objc; i++) {
+	    if (TclGetString(objv[i])[0] != '-') {
+		break;
+	    }
+	    if (Tcl_GetIndexFromObj(interp, objv[i], hiddenOptions, "option",
+		    0, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (index == OPT_GLOBAL) {
+		namespaceName = "::";
+	    } else if (index == OPT_NAMESPACE) {
+		if (++i == objc) { /* There must be more arguments. */
+		    break;
+		} else {
+		    namespaceName = TclGetString(objv[i]);
+		}
+	    } else {
+		i++;
+		break;
+	    }
+	}
+	if (objc - i < 1) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "?-namespace ns? ?-global? ?--? cmd ?arg ..?");
+	    return TCL_ERROR;
+	}
+	return SlaveInvokeHidden(interp, slaveInterp, namespaceName,
+		objc - i, objv + i);
+    }
+    case OPT_LIMIT: {
+	static const char *const limitTypes[] = {
+	    "commands", "time", NULL
+	};
+	enum LimitTypes {
+	    LIMIT_TYPE_COMMANDS, LIMIT_TYPE_TIME
+	};
+	int limitType;
+
+	if (objc < 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "limitType ?-option value ...?");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[2], limitTypes, "limit type", 0,
+		&limitType) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum LimitTypes) limitType) {
+	case LIMIT_TYPE_COMMANDS:
+	    return SlaveCommandLimitCmd(interp, slaveInterp, 3, objc,objv);
+	case LIMIT_TYPE_TIME:
+	    return SlaveTimeLimitCmd(interp, slaveInterp, 3, objc, objv);
+	}
+    }
+    case OPT_MARKTRUSTED:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return SlaveMarkTrusted(interp, slaveInterp);
+    case OPT_RECLIMIT:
+	if (objc != 2 && objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?newlimit?");
+	    return TCL_ERROR;
+	}
+	return SlaveRecursionLimit(interp, slaveInterp, objc - 2, objv + 2);
+    }
+
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveObjCmdDeleteProc --
+ *
+ *	Invoked when an object command for a slave interpreter is deleted;
+ *	cleans up all state associated with the slave interpreter and destroys
+ *	the slave interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Cleans up all state associated with the slave interpreter and destroys
+ *	the slave interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+SlaveObjCmdDeleteProc(
+    ClientData clientData)	/* The SlaveRecord for the command. */
+{
+    Slave *slavePtr;		/* Interim storage for Slave record. */
+    Tcl_Interp *slaveInterp = clientData;
+				/* And for a slave interp. */
+
+    slavePtr = &((InterpInfo *) ((Interp *) slaveInterp)->interpInfo)->slave;
+
+    /*
+     * Unlink the slave from its master interpreter.
+     */
+
+    Tcl_DeleteHashEntry(slavePtr->slaveEntryPtr);
+
+    /*
+     * Set to NULL so that when the InterpInfo is cleaned up in the slave it
+     * does not try to delete the command causing all sorts of grief. See
+     * SlaveRecordDeleteProc().
+     */
+
+    slavePtr->interpCmd = NULL;
+
+    if (slavePtr->slaveInterp != NULL) {
+	Tcl_DeleteInterp(slavePtr->slaveInterp);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveDebugCmd -- TIP #378
+ *
+ *	Helper function to handle 'debug' command in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May modify INTERP_DEBUG_FRAME flag in the slave.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveDebugCmd(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* The slave interpreter in which command
+				 * will be evaluated. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const debugTypes[] = {
+	"-frame", NULL
+    };
+    enum DebugTypes {
+	DEBUG_TYPE_FRAME
+    };
+    int debugType;
+    Interp *iPtr;
+    Tcl_Obj *resultPtr;
+
+    iPtr = (Interp *) slaveInterp;
+    if (objc == 0) {
+	resultPtr = Tcl_NewObj();
+	Tcl_ListObjAppendElement(NULL, resultPtr,
+		Tcl_NewStringObj("-frame", -1));
+	Tcl_ListObjAppendElement(NULL, resultPtr,
+		Tcl_NewBooleanObj(iPtr->flags & INTERP_DEBUG_FRAME));
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	if (Tcl_GetIndexFromObj(interp, objv[0], debugTypes, "debug option",
+		0, &debugType) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (debugType == DEBUG_TYPE_FRAME) {
+	    if (objc == 2) { /* set */
+		if (Tcl_GetBooleanFromObj(interp, objv[1], &debugType)
+			!= TCL_OK) {
+		    return TCL_ERROR;
+		}
+
+		/*
+		 * Quietly ignore attempts to disable interp debugging.  This
+		 * is a one-way switch as frame debug info is maintained in a
+		 * stack that must be consistent once turned on.
+		 */
+
+		if (debugType) {
+		    iPtr->flags |= INTERP_DEBUG_FRAME;
+		}
+	    }
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewBooleanObj(iPtr->flags & INTERP_DEBUG_FRAME));
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveEval --
+ *
+ *	Helper function to evaluate a command in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Whatever the command does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveEval(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* The slave interpreter in which command
+				 * will be evaluated. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int result;
+
+    /*
+     * TIP #285: If necessary, reset the cancellation flags for the slave
+     * interpreter now; otherwise, canceling a script in a master interpreter
+     * can result in a situation where a slave interpreter can no longer
+     * evaluate any scripts unless somebody calls the TclResetCancellation
+     * function for that particular Tcl_Interp.
+     */
+
+    TclSetSlaveCancelFlags(slaveInterp, 0, 0);
+
+    Tcl_Preserve(slaveInterp);
+    Tcl_AllowExceptions(slaveInterp);
+
+    if (objc == 1) {
+	/*
+	 * TIP #280: Make actual argument location available to eval'd script.
+	 */
+
+	Interp *iPtr = (Interp *) interp;
+	CmdFrame *invoker = iPtr->cmdFramePtr;
+	int word = 0;
+
+	TclArgumentGet(interp, objv[0], &invoker, &word);
+
+	result = TclEvalObjEx(slaveInterp, objv[0], 0, invoker, word);
+    } else {
+	Tcl_Obj *objPtr = Tcl_ConcatObj(objc, objv);
+	Tcl_IncrRefCount(objPtr);
+	result = Tcl_EvalObjEx(slaveInterp, objPtr, 0);
+	Tcl_DecrRefCount(objPtr);
+    }
+    Tcl_TransferResult(slaveInterp, result, interp);
+
+    Tcl_Release(slaveInterp);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveExpose --
+ *
+ *	Helper function to expose a command in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	After this call scripts in the slave will be able to invoke the newly
+ *	exposed command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveExpose(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* Interp in which command will be exposed. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument strings. */
+{
+    const char *name;
+
+    if (Tcl_IsSafe(interp)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"permission denied: safe interpreter cannot expose commands",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "UNSAFE",
+		NULL);
+	return TCL_ERROR;
+    }
+
+    name = TclGetString(objv[(objc == 1) ? 0 : 1]);
+    if (Tcl_ExposeCommand(slaveInterp, TclGetString(objv[0]),
+	    name) != TCL_OK) {
+	Tcl_TransferResult(slaveInterp, TCL_ERROR, interp);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveRecursionLimit --
+ *
+ *	Helper function to set/query the Recursion limit of an interp
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	When (objc == 1), slaveInterp will be set to a new recursion limit of
+ *	objv[0].
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveRecursionLimit(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* Interp in which limit is set/queried. */
+    int objc,			/* Set or Query. */
+    Tcl_Obj *const objv[])	/* Argument strings. */
+{
+    Interp *iPtr;
+    int limit;
+
+    if (objc) {
+	if (Tcl_IsSafe(interp)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("permission denied: "
+		    "safe interpreters cannot change recursion limit", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "UNSAFE",
+		    NULL);
+	    return TCL_ERROR;
+	}
+	if (TclGetIntFromObj(interp, objv[0], &limit) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (limit <= 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "recursion limit must be > 0", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "BADLIMIT",
+		    NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetRecursionLimit(slaveInterp, limit);
+	iPtr = (Interp *) slaveInterp;
+	if (interp == slaveInterp && iPtr->numLevels > limit) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "falling back due to new recursion limit", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "RECURSION", NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, objv[0]);
+	return TCL_OK;
+    } else {
+	limit = Tcl_SetRecursionLimit(slaveInterp, 0);
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(limit));
+	return TCL_OK;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveHide --
+ *
+ *	Helper function to hide a command in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	After this call scripts in the slave will no longer be able to invoke
+ *	the named command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveHide(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* Interp in which command will be exposed. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument strings. */
+{
+    const char *name;
+
+    if (Tcl_IsSafe(interp)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"permission denied: safe interpreter cannot hide commands",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "UNSAFE",
+		NULL);
+	return TCL_ERROR;
+    }
+
+    name = TclGetString(objv[(objc == 1) ? 0 : 1]);
+    if (Tcl_HideCommand(slaveInterp, TclGetString(objv[0]), name) != TCL_OK) {
+	Tcl_TransferResult(slaveInterp, TCL_ERROR, interp);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveHidden --
+ *
+ *	Helper function to compute list of hidden commands in a slave
+ *	interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveHidden(
+    Tcl_Interp *interp,		/* Interp for data return. */
+    Tcl_Interp *slaveInterp)	/* Interp whose hidden commands to query. */
+{
+    Tcl_Obj *listObjPtr = Tcl_NewObj();	/* Local object pointer. */
+    Tcl_HashTable *hTblPtr;		/* For local searches. */
+    Tcl_HashEntry *hPtr;		/* For local searches. */
+    Tcl_HashSearch hSearch;		/* For local searches. */
+
+    hTblPtr = ((Interp *) slaveInterp)->hiddenCmdTablePtr;
+    if (hTblPtr != NULL) {
+	for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch);
+		hPtr != NULL;
+		hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    Tcl_ListObjAppendElement(NULL, listObjPtr,
+		    Tcl_NewStringObj(Tcl_GetHashKey(hTblPtr, hPtr), -1));
+	}
+    }
+    Tcl_SetObjResult(interp, listObjPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveInvokeHidden --
+ *
+ *	Helper function to invoke a hidden command in a slave interpreter.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Whatever the hidden command does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveInvokeHidden(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp,	/* The slave interpreter in which command will
+				 * be invoked. */
+    const char *namespaceName,	/* The namespace to use, if any. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int result;
+
+    if (Tcl_IsSafe(interp)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"not allowed to invoke hidden commands from safe interpreter",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "UNSAFE",
+		NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_Preserve(slaveInterp);
+    Tcl_AllowExceptions(slaveInterp);
+
+    if (namespaceName == NULL) {
+	NRE_callback *rootPtr = TOP_CB(slaveInterp);
+
+	Tcl_NRAddCallback(interp, NRPostInvokeHidden, slaveInterp,
+		rootPtr, NULL, NULL);
+	return TclNRInvoke(NULL, slaveInterp, objc, objv);
+    } else {
+	Namespace *nsPtr, *dummy1, *dummy2;
+	const char *tail;
+
+	result = TclGetNamespaceForQualName(slaveInterp, namespaceName, NULL,
+		TCL_FIND_ONLY_NS | TCL_GLOBAL_ONLY | TCL_LEAVE_ERR_MSG
+		| TCL_CREATE_NS_IF_UNKNOWN, &nsPtr, &dummy1, &dummy2, &tail);
+	if (result == TCL_OK) {
+	    result = TclObjInvokeNamespace(slaveInterp, objc, objv,
+		    (Tcl_Namespace *) nsPtr, TCL_INVOKE_HIDDEN);
+	}
+    }
+
+    Tcl_TransferResult(slaveInterp, result, interp);
+
+    Tcl_Release(slaveInterp);
+    return result;
+}
+
+static int
+NRPostInvokeHidden(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Interp *slaveInterp = (Tcl_Interp *)data[0];
+    NRE_callback *rootPtr = (NRE_callback *)data[1];
+
+    if (interp != slaveInterp) {
+	result = TclNRRunCallbacks(slaveInterp, result, rootPtr);
+	Tcl_TransferResult(slaveInterp, result, interp);
+    }
+    Tcl_Release(slaveInterp);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveMarkTrusted --
+ *
+ *	Helper function to mark a slave interpreter as trusted (unsafe).
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	After this call the hard-wired security checks in the core no longer
+ *	prevent the slave from performing certain operations.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveMarkTrusted(
+    Tcl_Interp *interp,		/* Interp for error return. */
+    Tcl_Interp *slaveInterp)	/* The slave interpreter which will be marked
+				 * trusted. */
+{
+    if (Tcl_IsSafe(interp)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"permission denied: safe interpreter cannot mark trusted",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "UNSAFE",
+		NULL);
+	return TCL_ERROR;
+    }
+    ((Interp *) slaveInterp)->flags &= ~SAFE_INTERP;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_IsSafe --
+ *
+ *	Determines whether an interpreter is safe
+ *
+ * Results:
+ *	1 if it is safe, 0 if it is not.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_IsSafe(
+    Tcl_Interp *interp)		/* Is this interpreter "safe" ? */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (iPtr == NULL) {
+	return 0;
+    }
+    return (iPtr->flags & SAFE_INTERP) ? 1 : 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_MakeSafe --
+ *
+ *	Makes its argument interpreter contain only functionality that is
+ *	defined to be part of Safe Tcl. Unsafe commands are hidden, the env
+ *	array is unset, and the standard channels are removed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Hides commands in its argument interpreter, and removes settings and
+ *	channels.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_MakeSafe(
+    Tcl_Interp *interp)		/* Interpreter to be made safe. */
+{
+    Tcl_Channel chan;		/* Channel to remove from safe interpreter. */
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Interp *master = ((InterpInfo*) iPtr->interpInfo)->slave.masterInterp;
+
+    TclHideUnsafeCommands(interp);
+
+    if (master != NULL) {
+	/*
+	 * Alias these function implementations in the slave to those in the
+	 * master; the overall implementations are safe, but they're normally
+	 * defined by init.tcl which is not sourced by safe interpreters.
+	 * Assume these functions all work. [Bug 2895741]
+	 */
+
+	(void) Tcl_EvalEx(interp,
+		"namespace eval ::tcl {namespace eval mathfunc {}}", -1, 0);
+	(void) Tcl_CreateAlias(interp, "::tcl::mathfunc::min", master,
+		"::tcl::mathfunc::min", 0, NULL);
+	(void) Tcl_CreateAlias(interp, "::tcl::mathfunc::max", master,
+		"::tcl::mathfunc::max", 0, NULL);
+    }
+
+    iPtr->flags |= SAFE_INTERP;
+
+    /*
+     * Unsetting variables : (which should not have been set in the first
+     * place, but...)
+     */
+
+    /*
+     * No env array in a safe slave.
+     */
+
+    Tcl_UnsetVar(interp, "env", TCL_GLOBAL_ONLY);
+
+    /*
+     * Remove unsafe parts of tcl_platform
+     */
+
+    Tcl_UnsetVar2(interp, "tcl_platform", "os", TCL_GLOBAL_ONLY);
+    Tcl_UnsetVar2(interp, "tcl_platform", "osVersion", TCL_GLOBAL_ONLY);
+    Tcl_UnsetVar2(interp, "tcl_platform", "machine", TCL_GLOBAL_ONLY);
+    Tcl_UnsetVar2(interp, "tcl_platform", "user", TCL_GLOBAL_ONLY);
+
+    /*
+     * Unset path informations variables (the only one remaining is [info
+     * nameofexecutable])
+     */
+
+    Tcl_UnsetVar(interp, "tclDefaultLibrary", TCL_GLOBAL_ONLY);
+    Tcl_UnsetVar(interp, "tcl_library", TCL_GLOBAL_ONLY);
+    Tcl_UnsetVar(interp, "tcl_pkgPath", TCL_GLOBAL_ONLY);
+
+    /*
+     * Remove the standard channels from the interpreter; safe interpreters do
+     * not ordinarily have access to stdin, stdout and stderr.
+     *
+     * NOTE: These channels are not added to the interpreter by the
+     * Tcl_CreateInterp call, but may be added later, by another I/O
+     * operation. We want to ensure that the interpreter does not have these
+     * channels even if it is being made safe after being used for some time..
+     */
+
+    chan = Tcl_GetStdChannel(TCL_STDIN);
+    if (chan != NULL) {
+	Tcl_UnregisterChannel(interp, chan);
+    }
+    chan = Tcl_GetStdChannel(TCL_STDOUT);
+    if (chan != NULL) {
+	Tcl_UnregisterChannel(interp, chan);
+    }
+    chan = Tcl_GetStdChannel(TCL_STDERR);
+    if (chan != NULL) {
+	Tcl_UnregisterChannel(interp, chan);
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitExceeded --
+ *
+ *	Tests whether any limit has been exceeded in the given interpreter
+ *	(i.e. whether the interpreter is currently unable to process further
+ *	scripts).
+ *
+ * Results:
+ *	A boolean value.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Notes:
+ *	If you change this function, you MUST also update TclLimitExceeded() in
+ *	tclInt.h.
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitExceeded(
+    Tcl_Interp *interp)
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    return iPtr->limit.exceeded != 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitReady --
+ *
+ *	Find out whether any limit has been set on the interpreter, and if so
+ *	check whether the granularity of that limit is such that the full
+ *	limit check should be carried out.
+ *
+ * Results:
+ *	A boolean value that indicates whether to call Tcl_LimitCheck.
+ *
+ * Side effects:
+ *	Increments the limit granularity counter.
+ *
+ * Notes:
+ *	If you change this function, you MUST also update TclLimitReady() in
+ *	tclInt.h.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitReady(
+    Tcl_Interp *interp)
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    if (iPtr->limit.active != 0) {
+	register int ticker = ++iPtr->limit.granularityTicker;
+
+	if ((iPtr->limit.active & TCL_LIMIT_COMMANDS) &&
+		((iPtr->limit.cmdGranularity == 1) ||
+		    (ticker % iPtr->limit.cmdGranularity == 0))) {
+	    return 1;
+	}
+	if ((iPtr->limit.active & TCL_LIMIT_TIME) &&
+		((iPtr->limit.timeGranularity == 1) ||
+		    (ticker % iPtr->limit.timeGranularity == 0))) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitCheck --
+ *
+ *	Check all currently set limits in the interpreter (where permitted by
+ *	granularity). If a limit is exceeded, call its callbacks and, if the
+ *	limit is still exceeded after the callbacks have run, make the
+ *	interpreter generate an error that cannot be caught within the limited
+ *	interpreter.
+ *
+ * Results:
+ *	A Tcl result value (TCL_OK if no limit is exceeded, and TCL_ERROR if a
+ *	limit has been exceeded).
+ *
+ * Side effects:
+ *	May invoke system calls. May invoke other interpreters. May be
+ *	reentrant. May put the interpreter into a state where it can no longer
+ *	execute commands without outside intervention.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitCheck(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+    register int ticker = iPtr->limit.granularityTicker;
+
+    if (Tcl_InterpDeleted(interp)) {
+	return TCL_OK;
+    }
+
+    if ((iPtr->limit.active & TCL_LIMIT_COMMANDS) &&
+	    ((iPtr->limit.cmdGranularity == 1) ||
+		    (ticker % iPtr->limit.cmdGranularity == 0)) &&
+	    (iPtr->limit.cmdCount < iPtr->cmdCount)) {
+	iPtr->limit.exceeded |= TCL_LIMIT_COMMANDS;
+	Tcl_Preserve(interp);
+	RunLimitHandlers(iPtr->limit.cmdHandlers, interp);
+	if (iPtr->limit.cmdCount >= iPtr->cmdCount) {
+	    iPtr->limit.exceeded &= ~TCL_LIMIT_COMMANDS;
+	} else if (iPtr->limit.exceeded & TCL_LIMIT_COMMANDS) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "command count limit exceeded", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "LIMIT", "COMMANDS", NULL);
+	    Tcl_Release(interp);
+	    return TCL_ERROR;
+	}
+	Tcl_Release(interp);
+    }
+
+    if ((iPtr->limit.active & TCL_LIMIT_TIME) &&
+	    ((iPtr->limit.timeGranularity == 1) ||
+		(ticker % iPtr->limit.timeGranularity == 0))) {
+	Tcl_Time now;
+
+	Tcl_GetTime(&now);
+	if (iPtr->limit.time.sec < now.sec ||
+		(iPtr->limit.time.sec == now.sec &&
+		iPtr->limit.time.usec < now.usec)) {
+	    iPtr->limit.exceeded |= TCL_LIMIT_TIME;
+	    Tcl_Preserve(interp);
+	    RunLimitHandlers(iPtr->limit.timeHandlers, interp);
+	    if (iPtr->limit.time.sec > now.sec ||
+		    (iPtr->limit.time.sec == now.sec &&
+		    iPtr->limit.time.usec >= now.usec)) {
+		iPtr->limit.exceeded &= ~TCL_LIMIT_TIME;
+	    } else if (iPtr->limit.exceeded & TCL_LIMIT_TIME) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"time limit exceeded", -1));
+		Tcl_SetErrorCode(interp, "TCL", "LIMIT", "TIME", NULL);
+		Tcl_Release(interp);
+		return TCL_ERROR;
+	    }
+	    Tcl_Release(interp);
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RunLimitHandlers --
+ *
+ *	Invoke all the limit handlers in a list (for a particular limit).
+ *	Note that no particular limit handler callback will be invoked
+ *	reentrantly.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on the limit handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RunLimitHandlers(
+    LimitHandler *handlerPtr,
+    Tcl_Interp *interp)
+{
+    LimitHandler *nextPtr;
+    for (; handlerPtr!=NULL ; handlerPtr=nextPtr) {
+	if (handlerPtr->flags & (LIMIT_HANDLER_DELETED|LIMIT_HANDLER_ACTIVE)) {
+	    /*
+	     * Reentrant call or something seriously strange in the delete
+	     * code.
+	     */
+
+	    nextPtr = handlerPtr->nextPtr;
+	    continue;
+	}
+
+	/*
+	 * Set the ACTIVE flag while running the limit handler itself so we
+	 * cannot reentrantly call this handler and know to use the alternate
+	 * method of deletion if necessary.
+	 */
+
+	handlerPtr->flags |= LIMIT_HANDLER_ACTIVE;
+	handlerPtr->handlerProc(handlerPtr->clientData, interp);
+	handlerPtr->flags &= ~LIMIT_HANDLER_ACTIVE;
+
+	/*
+	 * Rediscover this value; it might have changed during the processing
+	 * of a limit handler. We have to record it here because we might
+	 * delete the structure below, and reading a value out of a deleted
+	 * structure is unsafe (even if actually legal with some
+	 * malloc()/free() implementations.)
+	 */
+
+	nextPtr = handlerPtr->nextPtr;
+
+	/*
+	 * If we deleted the current handler while we were executing it, we
+	 * will have spliced it out of the list and set the
+	 * LIMIT_HANDLER_DELETED flag.
+	 */
+
+	if (handlerPtr->flags & LIMIT_HANDLER_DELETED) {
+	    if (handlerPtr->deleteProc != NULL) {
+		handlerPtr->deleteProc(handlerPtr->clientData);
+	    }
+	    ckfree(handlerPtr);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitAddHandler --
+ *
+ *	Add a callback handler for a particular resource limit.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Extends the internal linked list of handlers for a limit.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitAddHandler(
+    Tcl_Interp *interp,
+    int type,
+    Tcl_LimitHandlerProc *handlerProc,
+    ClientData clientData,
+    Tcl_LimitHandlerDeleteProc *deleteProc)
+{
+    Interp *iPtr = (Interp *) interp;
+    LimitHandler *handlerPtr;
+
+    /*
+     * Convert everything into a real deletion callback.
+     */
+
+    if (deleteProc == (Tcl_LimitHandlerDeleteProc *) TCL_DYNAMIC) {
+	deleteProc = (Tcl_LimitHandlerDeleteProc *) Tcl_Free;
+    }
+
+    /*
+     * Allocate a handler record.
+     */
+
+    handlerPtr = ckalloc(sizeof(LimitHandler));
+    handlerPtr->flags = 0;
+    handlerPtr->handlerProc = handlerProc;
+    handlerPtr->clientData = clientData;
+    handlerPtr->deleteProc = deleteProc;
+    handlerPtr->prevPtr = NULL;
+
+    /*
+     * Prepend onto the front of the correct linked list.
+     */
+
+    switch (type) {
+    case TCL_LIMIT_COMMANDS:
+	handlerPtr->nextPtr = iPtr->limit.cmdHandlers;
+	if (handlerPtr->nextPtr != NULL) {
+	    handlerPtr->nextPtr->prevPtr = handlerPtr;
+	}
+	iPtr->limit.cmdHandlers = handlerPtr;
+	return;
+
+    case TCL_LIMIT_TIME:
+	handlerPtr->nextPtr = iPtr->limit.timeHandlers;
+	if (handlerPtr->nextPtr != NULL) {
+	    handlerPtr->nextPtr->prevPtr = handlerPtr;
+	}
+	iPtr->limit.timeHandlers = handlerPtr;
+	return;
+    }
+
+    Tcl_Panic("unknown type of resource limit");
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitRemoveHandler --
+ *
+ *	Remove a callback handler for a particular resource limit.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The handler is spliced out of the internal linked list for the limit,
+ *	and if not currently being invoked, deleted. Otherwise it is just
+ *	marked for deletion and removed when the limit handler has finished
+ *	executing.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitRemoveHandler(
+    Tcl_Interp *interp,
+    int type,
+    Tcl_LimitHandlerProc *handlerProc,
+    ClientData clientData)
+{
+    Interp *iPtr = (Interp *) interp;
+    LimitHandler *handlerPtr;
+
+    switch (type) {
+    case TCL_LIMIT_COMMANDS:
+	handlerPtr = iPtr->limit.cmdHandlers;
+	break;
+    case TCL_LIMIT_TIME:
+	handlerPtr = iPtr->limit.timeHandlers;
+	break;
+    default:
+	Tcl_Panic("unknown type of resource limit");
+	return;
+    }
+
+    for (; handlerPtr!=NULL ; handlerPtr=handlerPtr->nextPtr) {
+	if ((handlerPtr->handlerProc != handlerProc) ||
+		(handlerPtr->clientData != clientData)) {
+	    continue;
+	}
+
+	/*
+	 * We've found the handler to delete; mark it as doomed if not already
+	 * so marked (which shouldn't actually happen).
+	 */
+
+	if (handlerPtr->flags & LIMIT_HANDLER_DELETED) {
+	    return;
+	}
+	handlerPtr->flags |= LIMIT_HANDLER_DELETED;
+
+	/*
+	 * Splice the handler out of the doubly-linked list.
+	 */
+
+	if (handlerPtr->prevPtr == NULL) {
+	    switch (type) {
+	    case TCL_LIMIT_COMMANDS:
+		iPtr->limit.cmdHandlers = handlerPtr->nextPtr;
+		break;
+	    case TCL_LIMIT_TIME:
+		iPtr->limit.timeHandlers = handlerPtr->nextPtr;
+		break;
+	    }
+	} else {
+	    handlerPtr->prevPtr->nextPtr = handlerPtr->nextPtr;
+	}
+	if (handlerPtr->nextPtr != NULL) {
+	    handlerPtr->nextPtr->prevPtr = handlerPtr->prevPtr;
+	}
+
+	/*
+	 * If nothing is currently executing the handler, delete its client
+	 * data and the overall handler structure now. Otherwise it will all
+	 * go away when the handler returns.
+	 */
+
+	if (!(handlerPtr->flags & LIMIT_HANDLER_ACTIVE)) {
+	    if (handlerPtr->deleteProc != NULL) {
+		handlerPtr->deleteProc(handlerPtr->clientData);
+	    }
+	    ckfree(handlerPtr);
+	}
+	return;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLimitRemoveAllHandlers --
+ *
+ *	Remove all limit callback handlers for an interpreter. This is invoked
+ *	as part of deleting the interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Limit handlers are deleted or marked for deletion (as with
+ *	Tcl_LimitRemoveHandler).
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclLimitRemoveAllHandlers(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+    LimitHandler *handlerPtr, *nextHandlerPtr;
+
+    /*
+     * Delete all command-limit handlers.
+     */
+
+    for (handlerPtr=iPtr->limit.cmdHandlers, iPtr->limit.cmdHandlers=NULL;
+	    handlerPtr!=NULL; handlerPtr=nextHandlerPtr) {
+	nextHandlerPtr = handlerPtr->nextPtr;
+
+	/*
+	 * Do not delete here if it has already been marked for deletion.
+	 */
+
+	if (handlerPtr->flags & LIMIT_HANDLER_DELETED) {
+	    continue;
+	}
+	handlerPtr->flags |= LIMIT_HANDLER_DELETED;
+	handlerPtr->prevPtr = NULL;
+	handlerPtr->nextPtr = NULL;
+
+	/*
+	 * If nothing is currently executing the handler, delete its client
+	 * data and the overall handler structure now. Otherwise it will all
+	 * go away when the handler returns.
+	 */
+
+	if (!(handlerPtr->flags & LIMIT_HANDLER_ACTIVE)) {
+	    if (handlerPtr->deleteProc != NULL) {
+		handlerPtr->deleteProc(handlerPtr->clientData);
+	    }
+	    ckfree(handlerPtr);
+	}
+    }
+
+    /*
+     * Delete all time-limit handlers.
+     */
+
+    for (handlerPtr=iPtr->limit.timeHandlers, iPtr->limit.timeHandlers=NULL;
+	    handlerPtr!=NULL; handlerPtr=nextHandlerPtr) {
+	nextHandlerPtr = handlerPtr->nextPtr;
+
+	/*
+	 * Do not delete here if it has already been marked for deletion.
+	 */
+
+	if (handlerPtr->flags & LIMIT_HANDLER_DELETED) {
+	    continue;
+	}
+	handlerPtr->flags |= LIMIT_HANDLER_DELETED;
+	handlerPtr->prevPtr = NULL;
+	handlerPtr->nextPtr = NULL;
+
+	/*
+	 * If nothing is currently executing the handler, delete its client
+	 * data and the overall handler structure now. Otherwise it will all
+	 * go away when the handler returns.
+	 */
+
+	if (!(handlerPtr->flags & LIMIT_HANDLER_ACTIVE)) {
+	    if (handlerPtr->deleteProc != NULL) {
+		handlerPtr->deleteProc(handlerPtr->clientData);
+	    }
+	    ckfree(handlerPtr);
+	}
+    }
+
+    /*
+     * Delete the timer callback that is used to trap limits that occur in
+     * [vwait]s...
+     */
+
+    if (iPtr->limit.timeEvent != NULL) {
+	Tcl_DeleteTimerHandler(iPtr->limit.timeEvent);
+	iPtr->limit.timeEvent = NULL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitTypeEnabled --
+ *
+ *	Check whether a particular limit has been enabled for an interpreter.
+ *
+ * Results:
+ *	A boolean value.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitTypeEnabled(
+    Tcl_Interp *interp,
+    int type)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    return (iPtr->limit.active & type) != 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitTypeExceeded --
+ *
+ *	Check whether a particular limit has been exceeded for an interpreter.
+ *
+ * Results:
+ *	A boolean value (note that Tcl_LimitExceeded will always return
+ *	non-zero when this function returns non-zero).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitTypeExceeded(
+    Tcl_Interp *interp,
+    int type)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    return (iPtr->limit.exceeded & type) != 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitTypeSet --
+ *
+ *	Enable a particular limit for an interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The limit is turned on and will be checked in future at an interval
+ *	determined by the frequency of calling of Tcl_LimitReady and the
+ *	granularity of the limit in question.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitTypeSet(
+    Tcl_Interp *interp,
+    int type)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->limit.active |= type;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitTypeReset --
+ *
+ *	Disable a particular limit for an interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The limit is disabled. If the limit was exceeded when this function
+ *	was called, the limit will no longer be exceeded afterwards and the
+ *	interpreter will be free to execute further scripts (assuming it isn't
+ *	also deleted, of course).
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitTypeReset(
+    Tcl_Interp *interp,
+    int type)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->limit.active &= ~type;
+    iPtr->limit.exceeded &= ~type;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitSetCommands --
+ *
+ *	Set the command limit for an interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Also resets whether the command limit was exceeded. This might permit
+ *	a small amount of further execution in the interpreter even if the
+ *	limit itself is theoretically exceeded.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitSetCommands(
+    Tcl_Interp *interp,
+    int commandLimit)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->limit.cmdCount = commandLimit;
+    iPtr->limit.exceeded &= ~TCL_LIMIT_COMMANDS;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitGetCommands --
+ *
+ *	Get the number of commands that may be executed in the interpreter
+ *	before the command-limit is reached.
+ *
+ * Results:
+ *	An upper bound on the number of commands.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitGetCommands(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    return iPtr->limit.cmdCount;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitSetTime --
+ *
+ *	Set the time limit for an interpreter by copying it from the value
+ *	pointed to by the timeLimitPtr argument.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Also resets whether the time limit was exceeded. This might permit a
+ *	small amount of further execution in the interpreter even if the limit
+ *	itself is theoretically exceeded.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitSetTime(
+    Tcl_Interp *interp,
+    Tcl_Time *timeLimitPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Time nextMoment;
+
+    memcpy(&iPtr->limit.time, timeLimitPtr, sizeof(Tcl_Time));
+    if (iPtr->limit.timeEvent != NULL) {
+	Tcl_DeleteTimerHandler(iPtr->limit.timeEvent);
+    }
+    nextMoment.sec = timeLimitPtr->sec;
+    nextMoment.usec = timeLimitPtr->usec+10;
+    if (nextMoment.usec >= 1000000) {
+	nextMoment.sec++;
+	nextMoment.usec -= 1000000;
+    }
+    iPtr->limit.timeEvent = TclCreateAbsoluteTimerHandler(&nextMoment,
+	    TimeLimitCallback, interp);
+    iPtr->limit.exceeded &= ~TCL_LIMIT_TIME;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimeLimitCallback --
+ *
+ *	Callback that allows time limits to be enforced even when doing a
+ *	blocking wait for events.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May put the interpreter into a state where it can no longer execute
+ *	commands. May make callbacks into other interpreters.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimeLimitCallback(
+    ClientData clientData)
+{
+    Tcl_Interp *interp = clientData;
+    Interp *iPtr = clientData;
+    int code;
+
+    Tcl_Preserve(interp);
+    iPtr->limit.timeEvent = NULL;
+
+    /*
+     * Must reset the granularity ticker here to force an immediate full
+     * check. This is OK because we're swallowing the cost in the overall cost
+     * of the event loop. [Bug 2891362]
+     */
+
+    iPtr->limit.granularityTicker = 0;
+
+    code = Tcl_LimitCheck(interp);
+    if (code != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (while waiting for event)");
+	Tcl_BackgroundException(interp, code);
+    }
+    Tcl_Release(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitGetTime --
+ *
+ *	Get the current time limit.
+ *
+ * Results:
+ *	The time limit (by it being copied into the variable pointed to by the
+ *	timeLimitPtr).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitGetTime(
+    Tcl_Interp *interp,
+    Tcl_Time *timeLimitPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    memcpy(timeLimitPtr, &iPtr->limit.time, sizeof(Tcl_Time));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitSetGranularity --
+ *
+ *	Set the granularity divisor (which must be positive) for a particular
+ *	limit.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The granularity is updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LimitSetGranularity(
+    Tcl_Interp *interp,
+    int type,
+    int granularity)
+{
+    Interp *iPtr = (Interp *) interp;
+    if (granularity < 1) {
+	Tcl_Panic("limit granularity must be positive");
+    }
+
+    switch (type) {
+    case TCL_LIMIT_COMMANDS:
+	iPtr->limit.cmdGranularity = granularity;
+	return;
+    case TCL_LIMIT_TIME:
+	iPtr->limit.timeGranularity = granularity;
+	return;
+    }
+    Tcl_Panic("unknown type of resource limit");
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LimitGetGranularity --
+ *
+ *	Get the granularity divisor for a particular limit.
+ *
+ * Results:
+ *	The granularity divisor for the given limit.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LimitGetGranularity(
+    Tcl_Interp *interp,
+    int type)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    switch (type) {
+    case TCL_LIMIT_COMMANDS:
+	return iPtr->limit.cmdGranularity;
+    case TCL_LIMIT_TIME:
+	return iPtr->limit.timeGranularity;
+    }
+    Tcl_Panic("unknown type of resource limit");
+    return -1; /* NOT REACHED */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteScriptLimitCallback --
+ *
+ *	Callback for when a script limit (a limit callback implemented as a
+ *	Tcl script in a master interpreter, as set up from Tcl) is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The reference to the script callback from the controlling interpreter
+ *	is removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteScriptLimitCallback(
+    ClientData clientData)
+{
+    ScriptLimitCallback *limitCBPtr = clientData;
+
+    Tcl_DecrRefCount(limitCBPtr->scriptObj);
+    if (limitCBPtr->entryPtr != NULL) {
+	Tcl_DeleteHashEntry(limitCBPtr->entryPtr);
+    }
+    ckfree(limitCBPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CallScriptLimitCallback --
+ *
+ *	Invoke a script limit callback. Used to implement limit callbacks set
+ *	at the Tcl level on child interpreters.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on the callback script. Errors are reported as background
+ *	errors.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CallScriptLimitCallback(
+    ClientData clientData,
+    Tcl_Interp *interp)		/* Interpreter which failed the limit */
+{
+    ScriptLimitCallback *limitCBPtr = clientData;
+    int code;
+
+    if (Tcl_InterpDeleted(limitCBPtr->interp)) {
+	return;
+    }
+    Tcl_Preserve(limitCBPtr->interp);
+    code = Tcl_EvalObjEx(limitCBPtr->interp, limitCBPtr->scriptObj,
+	    TCL_EVAL_GLOBAL);
+    if (code != TCL_OK && !Tcl_InterpDeleted(limitCBPtr->interp)) {
+	Tcl_BackgroundException(limitCBPtr->interp, code);
+    }
+    Tcl_Release(limitCBPtr->interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetScriptLimitCallback --
+ *
+ *	Install (or remove, if scriptObj is NULL) a limit callback script that
+ *	is called when the target interpreter exceeds the type of limit
+ *	specified. Each interpreter may only have one callback set on another
+ *	interpreter through this mechanism (though as many interpreters may be
+ *	limited as the programmer chooses overall).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	A limit callback implemented as an invokation of a Tcl script in
+ *	another interpreter is either installed or removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+SetScriptLimitCallback(
+    Tcl_Interp *interp,
+    int type,
+    Tcl_Interp *targetInterp,
+    Tcl_Obj *scriptObj)
+{
+    ScriptLimitCallback *limitCBPtr;
+    Tcl_HashEntry *hashPtr;
+    int isNew;
+    ScriptLimitCallbackKey key;
+    Interp *iPtr = (Interp *) interp;
+
+    if (interp == targetInterp) {
+	Tcl_Panic("installing limit callback to the limited interpreter");
+    }
+
+    key.interp = targetInterp;
+    key.type = type;
+
+    if (scriptObj == NULL) {
+	hashPtr = Tcl_FindHashEntry(&iPtr->limit.callbacks, (char *) &key);
+	if (hashPtr != NULL) {
+	    Tcl_LimitRemoveHandler(targetInterp, type, CallScriptLimitCallback,
+		    Tcl_GetHashValue(hashPtr));
+	}
+	return;
+    }
+
+    hashPtr = Tcl_CreateHashEntry(&iPtr->limit.callbacks, &key,
+	    &isNew);
+    if (!isNew) {
+	limitCBPtr = Tcl_GetHashValue(hashPtr);
+	limitCBPtr->entryPtr = NULL;
+	Tcl_LimitRemoveHandler(targetInterp, type, CallScriptLimitCallback,
+		limitCBPtr);
+    }
+
+    limitCBPtr = ckalloc(sizeof(ScriptLimitCallback));
+    limitCBPtr->interp = interp;
+    limitCBPtr->scriptObj = scriptObj;
+    limitCBPtr->entryPtr = hashPtr;
+    limitCBPtr->type = type;
+    Tcl_IncrRefCount(scriptObj);
+
+    Tcl_LimitAddHandler(targetInterp, type, CallScriptLimitCallback,
+	    limitCBPtr, DeleteScriptLimitCallback);
+    Tcl_SetHashValue(hashPtr, limitCBPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRemoveScriptLimitCallbacks --
+ *
+ *	Remove all script-implemented limit callbacks that make calls back
+ *	into the given interpreter. This invoked as part of deleting an
+ *	interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The script limit callbacks are removed or marked for later removal.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclRemoveScriptLimitCallbacks(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hashPtr;
+    Tcl_HashSearch search;
+    ScriptLimitCallbackKey *keyPtr;
+
+    hashPtr = Tcl_FirstHashEntry(&iPtr->limit.callbacks, &search);
+    while (hashPtr != NULL) {
+	keyPtr = (ScriptLimitCallbackKey *)
+		Tcl_GetHashKey(&iPtr->limit.callbacks, hashPtr);
+	Tcl_LimitRemoveHandler(keyPtr->interp, keyPtr->type,
+		CallScriptLimitCallback, Tcl_GetHashValue(hashPtr));
+	hashPtr = Tcl_NextHashEntry(&search);
+    }
+    Tcl_DeleteHashTable(&iPtr->limit.callbacks);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitLimitSupport --
+ *
+ *	Initialise all the parts of the interpreter relating to resource limit
+ *	management. This allows an interpreter to both have limits set upon
+ *	itself and set limits upon other interpreters.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The resource limit subsystem is initialised for the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitLimitSupport(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    iPtr->limit.active = 0;
+    iPtr->limit.granularityTicker = 0;
+    iPtr->limit.exceeded = 0;
+    iPtr->limit.cmdCount = 0;
+    iPtr->limit.cmdHandlers = NULL;
+    iPtr->limit.cmdGranularity = 1;
+    memset(&iPtr->limit.time, 0, sizeof(Tcl_Time));
+    iPtr->limit.timeHandlers = NULL;
+    iPtr->limit.timeEvent = NULL;
+    iPtr->limit.timeGranularity = 10;
+    Tcl_InitHashTable(&iPtr->limit.callbacks,
+	    sizeof(ScriptLimitCallbackKey)/sizeof(int));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InheritLimitsFromMaster --
+ *
+ *	Derive the interpreter limit configuration for a slave interpreter
+ *	from the limit config for the master.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The slave interpreter limits are set so that if the master has a
+ *	limit, it may not exceed it by handing off work to slave interpreters.
+ *	Note that this does not transfer limit callbacks from the master to
+ *	the slave.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+InheritLimitsFromMaster(
+    Tcl_Interp *slaveInterp,
+    Tcl_Interp *masterInterp)
+{
+    Interp *slavePtr = (Interp *) slaveInterp;
+    Interp *masterPtr = (Interp *) masterInterp;
+
+    if (masterPtr->limit.active & TCL_LIMIT_COMMANDS) {
+	slavePtr->limit.active |= TCL_LIMIT_COMMANDS;
+	slavePtr->limit.cmdCount = 0;
+	slavePtr->limit.cmdGranularity = masterPtr->limit.cmdGranularity;
+    }
+    if (masterPtr->limit.active & TCL_LIMIT_TIME) {
+	slavePtr->limit.active |= TCL_LIMIT_TIME;
+	memcpy(&slavePtr->limit.time, &masterPtr->limit.time,
+		sizeof(Tcl_Time));
+	slavePtr->limit.timeGranularity = masterPtr->limit.timeGranularity;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveCommandLimitCmd --
+ *
+ *	Implementation of the [interp limit $i commands] and [$i limit
+ *	commands] subcommands. See the interp manual page for a full
+ *	description.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Depends on the arguments.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveCommandLimitCmd(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Interp *slaveInterp,	/* Interpreter being adjusted. */
+    int consumedObjc,		/* Number of args already parsed. */
+    int objc,			/* Total number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const options[] = {
+	"-command", "-granularity", "-value", NULL
+    };
+    enum Options {
+	OPT_CMD, OPT_GRAN, OPT_VAL
+    };
+    Interp *iPtr = (Interp *) interp;
+    int index;
+    ScriptLimitCallbackKey key;
+    ScriptLimitCallback *limitCBPtr;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * First, ensure that we are not reading or writing the calling
+     * interpreter's limits; it may only manipulate its children. Note that
+     * the low level API enforces this with Tcl_Panic, which we want to
+     * avoid. [Bug 3398794]
+     */
+
+    if (interp == slaveInterp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"limits on current interpreter inaccessible", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "SELF", NULL);
+	return TCL_ERROR;
+    }
+
+    if (objc == consumedObjc) {
+	Tcl_Obj *dictPtr;
+
+	TclNewObj(dictPtr);
+	key.interp = slaveInterp;
+	key.type = TCL_LIMIT_COMMANDS;
+	hPtr = Tcl_FindHashEntry(&iPtr->limit.callbacks, (char *) &key);
+	if (hPtr != NULL) {
+	    limitCBPtr = Tcl_GetHashValue(hPtr);
+	    if (limitCBPtr != NULL && limitCBPtr->scriptObj != NULL) {
+		Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[0], -1),
+			limitCBPtr->scriptObj);
+	    } else {
+		goto putEmptyCommandInDict;
+	    }
+	} else {
+	    Tcl_Obj *empty;
+
+	putEmptyCommandInDict:
+	    TclNewObj(empty);
+	    Tcl_DictObjPut(NULL, dictPtr,
+		    Tcl_NewStringObj(options[0], -1), empty);
+	}
+	Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[1], -1),
+		Tcl_NewIntObj(Tcl_LimitGetGranularity(slaveInterp,
+		TCL_LIMIT_COMMANDS)));
+
+	if (Tcl_LimitTypeEnabled(slaveInterp, TCL_LIMIT_COMMANDS)) {
+	    Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[2], -1),
+		    Tcl_NewIntObj(Tcl_LimitGetCommands(slaveInterp)));
+	} else {
+	    Tcl_Obj *empty;
+
+	    TclNewObj(empty);
+	    Tcl_DictObjPut(NULL, dictPtr,
+		    Tcl_NewStringObj(options[2], -1), empty);
+	}
+	Tcl_SetObjResult(interp, dictPtr);
+	return TCL_OK;
+    } else if (objc == consumedObjc+1) {
+	if (Tcl_GetIndexFromObj(interp, objv[consumedObjc], options, "option",
+		0, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum Options) index) {
+	case OPT_CMD:
+	    key.interp = slaveInterp;
+	    key.type = TCL_LIMIT_COMMANDS;
+	    hPtr = Tcl_FindHashEntry(&iPtr->limit.callbacks, (char *) &key);
+	    if (hPtr != NULL) {
+		limitCBPtr = Tcl_GetHashValue(hPtr);
+		if (limitCBPtr != NULL && limitCBPtr->scriptObj != NULL) {
+		    Tcl_SetObjResult(interp, limitCBPtr->scriptObj);
+		}
+	    }
+	    break;
+	case OPT_GRAN:
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(
+		    Tcl_LimitGetGranularity(slaveInterp, TCL_LIMIT_COMMANDS)));
+	    break;
+	case OPT_VAL:
+	    if (Tcl_LimitTypeEnabled(slaveInterp, TCL_LIMIT_COMMANDS)) {
+		Tcl_SetObjResult(interp,
+			Tcl_NewIntObj(Tcl_LimitGetCommands(slaveInterp)));
+	    }
+	    break;
+	}
+	return TCL_OK;
+    } else if ((objc-consumedObjc) & 1 /* isOdd(objc-consumedObjc) */) {
+	Tcl_WrongNumArgs(interp, consumedObjc, objv, "?-option value ...?");
+	return TCL_ERROR;
+    } else {
+	int i, scriptLen = 0, limitLen = 0;
+	Tcl_Obj *scriptObj = NULL, *granObj = NULL, *limitObj = NULL;
+	int gran = 0, limit = 0;
+
+	for (i=consumedObjc ; i<objc ; i+=2) {
+	    if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		    &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum Options) index) {
+	    case OPT_CMD:
+		scriptObj = objv[i+1];
+		(void) TclGetStringFromObj(scriptObj, &scriptLen);
+		break;
+	    case OPT_GRAN:
+		granObj = objv[i+1];
+		if (TclGetIntFromObj(interp, objv[i+1], &gran) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (gran < 1) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "granularity must be at least 1", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADVALUE", NULL);
+		    return TCL_ERROR;
+		}
+		break;
+	    case OPT_VAL:
+		limitObj = objv[i+1];
+		(void) TclGetStringFromObj(objv[i+1], &limitLen);
+		if (limitLen == 0) {
+		    break;
+		}
+		if (TclGetIntFromObj(interp, objv[i+1], &limit) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (limit < 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "command limit value must be at least 0", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADVALUE", NULL);
+		    return TCL_ERROR;
+		}
+		break;
+	    }
+	}
+	if (scriptObj != NULL) {
+	    SetScriptLimitCallback(interp, TCL_LIMIT_COMMANDS, slaveInterp,
+		    (scriptLen > 0 ? scriptObj : NULL));
+	}
+	if (granObj != NULL) {
+	    Tcl_LimitSetGranularity(slaveInterp, TCL_LIMIT_COMMANDS, gran);
+	}
+	if (limitObj != NULL) {
+	    if (limitLen > 0) {
+		Tcl_LimitSetCommands(slaveInterp, limit);
+		Tcl_LimitTypeSet(slaveInterp, TCL_LIMIT_COMMANDS);
+	    } else {
+		Tcl_LimitTypeReset(slaveInterp, TCL_LIMIT_COMMANDS);
+	    }
+	}
+	return TCL_OK;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SlaveTimeLimitCmd --
+ *
+ *	Implementation of the [interp limit $i time] and [$i limit time]
+ *	subcommands. See the interp manual page for a full description.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Depends on the arguments.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SlaveTimeLimitCmd(
+    Tcl_Interp *interp,			/* Current interpreter. */
+    Tcl_Interp *slaveInterp,		/* Interpreter being adjusted. */
+    int consumedObjc,			/* Number of args already parsed. */
+    int objc,				/* Total number of arguments. */
+    Tcl_Obj *const objv[])		/* Argument objects. */
+{
+    static const char *const options[] = {
+	"-command", "-granularity", "-milliseconds", "-seconds", NULL
+    };
+    enum Options {
+	OPT_CMD, OPT_GRAN, OPT_MILLI, OPT_SEC
+    };
+    Interp *iPtr = (Interp *) interp;
+    int index;
+    ScriptLimitCallbackKey key;
+    ScriptLimitCallback *limitCBPtr;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * First, ensure that we are not reading or writing the calling
+     * interpreter's limits; it may only manipulate its children. Note that
+     * the low level API enforces this with Tcl_Panic, which we want to
+     * avoid. [Bug 3398794]
+     */
+
+    if (interp == slaveInterp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"limits on current interpreter inaccessible", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP", "SELF", NULL);
+	return TCL_ERROR;
+    }
+
+    if (objc == consumedObjc) {
+	Tcl_Obj *dictPtr;
+
+	TclNewObj(dictPtr);
+	key.interp = slaveInterp;
+	key.type = TCL_LIMIT_TIME;
+	hPtr = Tcl_FindHashEntry(&iPtr->limit.callbacks, (char *) &key);
+	if (hPtr != NULL) {
+	    limitCBPtr = Tcl_GetHashValue(hPtr);
+	    if (limitCBPtr != NULL && limitCBPtr->scriptObj != NULL) {
+		Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[0], -1),
+			limitCBPtr->scriptObj);
+	    } else {
+		goto putEmptyCommandInDict;
+	    }
+	} else {
+	    Tcl_Obj *empty;
+	putEmptyCommandInDict:
+	    TclNewObj(empty);
+	    Tcl_DictObjPut(NULL, dictPtr,
+		    Tcl_NewStringObj(options[0], -1), empty);
+	}
+	Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[1], -1),
+		Tcl_NewIntObj(Tcl_LimitGetGranularity(slaveInterp,
+		TCL_LIMIT_TIME)));
+
+	if (Tcl_LimitTypeEnabled(slaveInterp, TCL_LIMIT_TIME)) {
+	    Tcl_Time limitMoment;
+
+	    Tcl_LimitGetTime(slaveInterp, &limitMoment);
+	    Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[2], -1),
+		    Tcl_NewLongObj(limitMoment.usec/1000));
+	    Tcl_DictObjPut(NULL, dictPtr, Tcl_NewStringObj(options[3], -1),
+		    Tcl_NewLongObj(limitMoment.sec));
+	} else {
+	    Tcl_Obj *empty;
+
+	    TclNewObj(empty);
+	    Tcl_DictObjPut(NULL, dictPtr,
+		    Tcl_NewStringObj(options[2], -1), empty);
+	    Tcl_DictObjPut(NULL, dictPtr,
+		    Tcl_NewStringObj(options[3], -1), empty);
+	}
+	Tcl_SetObjResult(interp, dictPtr);
+	return TCL_OK;
+    } else if (objc == consumedObjc+1) {
+	if (Tcl_GetIndexFromObj(interp, objv[consumedObjc], options, "option",
+		0, &index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum Options) index) {
+	case OPT_CMD:
+	    key.interp = slaveInterp;
+	    key.type = TCL_LIMIT_TIME;
+	    hPtr = Tcl_FindHashEntry(&iPtr->limit.callbacks, (char *) &key);
+	    if (hPtr != NULL) {
+		limitCBPtr = Tcl_GetHashValue(hPtr);
+		if (limitCBPtr != NULL && limitCBPtr->scriptObj != NULL) {
+		    Tcl_SetObjResult(interp, limitCBPtr->scriptObj);
+		}
+	    }
+	    break;
+	case OPT_GRAN:
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(
+		    Tcl_LimitGetGranularity(slaveInterp, TCL_LIMIT_TIME)));
+	    break;
+	case OPT_MILLI:
+	    if (Tcl_LimitTypeEnabled(slaveInterp, TCL_LIMIT_TIME)) {
+		Tcl_Time limitMoment;
+
+		Tcl_LimitGetTime(slaveInterp, &limitMoment);
+		Tcl_SetObjResult(interp,
+			Tcl_NewLongObj(limitMoment.usec/1000));
+	    }
+	    break;
+	case OPT_SEC:
+	    if (Tcl_LimitTypeEnabled(slaveInterp, TCL_LIMIT_TIME)) {
+		Tcl_Time limitMoment;
+
+		Tcl_LimitGetTime(slaveInterp, &limitMoment);
+		Tcl_SetObjResult(interp, Tcl_NewLongObj(limitMoment.sec));
+	    }
+	    break;
+	}
+	return TCL_OK;
+    } else if ((objc-consumedObjc) & 1 /* isOdd(objc-consumedObjc) */) {
+	Tcl_WrongNumArgs(interp, consumedObjc, objv, "?-option value ...?");
+	return TCL_ERROR;
+    } else {
+	int i, scriptLen = 0, milliLen = 0, secLen = 0;
+	Tcl_Obj *scriptObj = NULL, *granObj = NULL;
+	Tcl_Obj *milliObj = NULL, *secObj = NULL;
+	int gran = 0;
+	Tcl_Time limitMoment;
+	int tmp;
+
+	Tcl_LimitGetTime(slaveInterp, &limitMoment);
+	for (i=consumedObjc ; i<objc ; i+=2) {
+	    if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		    &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum Options) index) {
+	    case OPT_CMD:
+		scriptObj = objv[i+1];
+		(void) TclGetStringFromObj(objv[i+1], &scriptLen);
+		break;
+	    case OPT_GRAN:
+		granObj = objv[i+1];
+		if (TclGetIntFromObj(interp, objv[i+1], &gran) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (gran < 1) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "granularity must be at least 1", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADVALUE", NULL);
+		    return TCL_ERROR;
+		}
+		break;
+	    case OPT_MILLI:
+		milliObj = objv[i+1];
+		(void) TclGetStringFromObj(objv[i+1], &milliLen);
+		if (milliLen == 0) {
+		    break;
+		}
+		if (TclGetIntFromObj(interp, objv[i+1], &tmp) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (tmp < 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "milliseconds must be at least 0", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADVALUE", NULL);
+		    return TCL_ERROR;
+		}
+		limitMoment.usec = ((long) tmp)*1000;
+		break;
+	    case OPT_SEC:
+		secObj = objv[i+1];
+		(void) TclGetStringFromObj(objv[i+1], &secLen);
+		if (secLen == 0) {
+		    break;
+		}
+		if (TclGetIntFromObj(interp, objv[i+1], &tmp) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (tmp < 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "seconds must be at least 0", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADVALUE", NULL);
+		    return TCL_ERROR;
+		}
+		limitMoment.sec = tmp;
+		break;
+	    }
+	}
+	if (milliObj != NULL || secObj != NULL) {
+	    if (milliObj != NULL) {
+		/*
+		 * Setting -milliseconds but clearing -seconds, or resetting
+		 * -milliseconds but not resetting -seconds? Bad voodoo!
+		 */
+
+		if (secObj != NULL && secLen == 0 && milliLen > 0) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "may only set -milliseconds if -seconds is not "
+			    "also being reset", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADUSAGE", NULL);
+		    return TCL_ERROR;
+		}
+		if (milliLen == 0 && (secObj == NULL || secLen > 0)) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "may only reset -milliseconds if -seconds is "
+			    "also being reset", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "INTERP",
+			    "BADUSAGE", NULL);
+		    return TCL_ERROR;
+		}
+	    }
+
+	    if (milliLen > 0 || secLen > 0) {
+		/*
+		 * Force usec to be in range [0..1000000), possibly
+		 * incrementing sec in the process. This makes it much easier
+		 * for people to write scripts that do small time increments.
+		 */
+
+		limitMoment.sec += limitMoment.usec / 1000000;
+		limitMoment.usec %= 1000000;
+
+		Tcl_LimitSetTime(slaveInterp, &limitMoment);
+		Tcl_LimitTypeSet(slaveInterp, TCL_LIMIT_TIME);
+	    } else {
+		Tcl_LimitTypeReset(slaveInterp, TCL_LIMIT_TIME);
+	    }
+	}
+	if (scriptObj != NULL) {
+	    SetScriptLimitCallback(interp, TCL_LIMIT_TIME, slaveInterp,
+		    (scriptLen > 0 ? scriptObj : NULL));
+	}
+	if (granObj != NULL) {
+	    Tcl_LimitSetGranularity(slaveInterp, TCL_LIMIT_TIME, gran);
+	}
+	return TCL_OK;
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclLink.c b/generic/tclLink.c
new file mode 100644
index 0000000..7366acc
--- /dev/null
+++ b/generic/tclLink.c
@@ -0,0 +1,758 @@
+/*
+ * tclLink.c --
+ *
+ *	This file implements linked variables (a C variable that is tied to a
+ *	Tcl variable). The idea of linked variables was first suggested by
+ *	Andreas Stolcke and this implementation is based heavily on a
+ *	prototype implementation provided by him.
+ *
+ * Copyright (c) 1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * For each linked variable there is a data structure of the following type,
+ * which describes the link and is the clientData for the trace set on the Tcl
+ * variable.
+ */
+
+typedef struct Link {
+    Tcl_Interp *interp;		/* Interpreter containing Tcl variable. */
+    Tcl_Obj *varName;		/* Name of variable (must be global). This is
+				 * needed during trace callbacks, since the
+				 * actual variable may be aliased at that time
+				 * via upvar. */
+    char *addr;			/* Location of C variable. */
+    int type;			/* Type of link (TCL_LINK_INT, etc.). */
+    union {
+	char c;
+	unsigned char uc;
+	int i;
+	unsigned int ui;
+	short s;
+	unsigned short us;
+#if !defined(TCL_WIDE_INT_IS_LONG) && !defined(_WIN32) && !defined(__CYGWIN__)
+	long l;
+	unsigned long ul;
+#endif
+	Tcl_WideInt w;
+	Tcl_WideUInt uw;
+	float f;
+	double d;
+    } lastValue;		/* Last known value of C variable; used to
+				 * avoid string conversions. */
+    int flags;			/* Miscellaneous one-bit values; see below for
+				 * definitions. */
+} Link;
+
+/*
+ * Definitions for flag bits:
+ * LINK_READ_ONLY -		1 means errors should be generated if Tcl
+ *				script attempts to write variable.
+ * LINK_BEING_UPDATED -		1 means that a call to Tcl_UpdateLinkedVar is
+ *				in progress for this variable, so trace
+ *				callbacks on the variable should be ignored.
+ */
+
+#define LINK_READ_ONLY		1
+#define LINK_BEING_UPDATED	2
+
+/*
+ * Forward references to functions defined later in this file:
+ */
+
+static char *		LinkTraceProc(ClientData clientData,Tcl_Interp *interp,
+			    const char *name1, const char *name2, int flags);
+static Tcl_Obj *	ObjValue(Link *linkPtr);
+static int		GetInvalidIntFromObj(Tcl_Obj *objPtr, int *intPtr);
+static int		GetInvalidWideFromObj(Tcl_Obj *objPtr, Tcl_WideInt *widePtr);
+static int		GetInvalidDoubleFromObj(Tcl_Obj *objPtr, double *doublePtr);
+
+/*
+ * Convenience macro for accessing the value of the C variable pointed to by a
+ * link. Note that this macro produces something that may be regarded as an
+ * lvalue or rvalue; it may be assigned to as well as read. Also note that
+ * this macro assumes the name of the variable being accessed (linkPtr); this
+ * is not strictly a good thing, but it keeps the code much shorter and
+ * cleaner.
+ */
+
+#define LinkedVar(type) (*(type *) linkPtr->addr)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LinkVar --
+ *
+ *	Link a C variable to a Tcl variable so that changes to either one
+ *	causes the other to change.
+ *
+ * Results:
+ *	The return value is TCL_OK if everything went well or TCL_ERROR if an
+ *	error occurred (the interp's result is also set after errors).
+ *
+ * Side effects:
+ *	The value at *addr is linked to the Tcl variable "varName", using
+ *	"type" to convert between string values for Tcl and binary values for
+ *	*addr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LinkVar(
+    Tcl_Interp *interp,		/* Interpreter in which varName exists. */
+    const char *varName,	/* Name of a global variable in interp. */
+    char *addr,			/* Address of a C variable to be linked to
+				 * varName. */
+    int type)			/* Type of C variable: TCL_LINK_INT, etc. Also
+				 * may have TCL_LINK_READ_ONLY OR'ed in. */
+{
+    Tcl_Obj *objPtr;
+    Link *linkPtr;
+    int code;
+
+    linkPtr = (Link *) Tcl_VarTraceInfo2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY, LinkTraceProc, (ClientData) NULL);
+    if (linkPtr != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"variable '%s' is already linked", varName));
+	return TCL_ERROR;
+    }
+
+    linkPtr = ckalloc(sizeof(Link));
+    linkPtr->interp = interp;
+    linkPtr->varName = Tcl_NewStringObj(varName, -1);
+    Tcl_IncrRefCount(linkPtr->varName);
+    linkPtr->addr = addr;
+    linkPtr->type = type & ~TCL_LINK_READ_ONLY;
+#if !defined(TCL_NO_DEPRECATED) && (defined(TCL_WIDE_INT_IS_LONG) \
+	|| defined(_WIN32) || defined(__CYGWIN__))
+    if (linkPtr->type == 11 /* legacy TCL_LINK_LONG */) {
+	linkPtr->type = TCL_LINK_LONG;
+    } else if (linkPtr->type == 12 /* legacy TCL_LINK_ULONG */) {
+	linkPtr->type = TCL_LINK_ULONG;
+    }
+#endif
+    if (type & TCL_LINK_READ_ONLY) {
+	linkPtr->flags = LINK_READ_ONLY;
+    } else {
+	linkPtr->flags = 0;
+    }
+    objPtr = ObjValue(linkPtr);
+    if (Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, objPtr,
+	    TCL_GLOBAL_ONLY|TCL_LEAVE_ERR_MSG) == NULL) {
+	Tcl_DecrRefCount(linkPtr->varName);
+	ckfree(linkPtr);
+	return TCL_ERROR;
+    }
+    code = Tcl_TraceVar2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY|TCL_TRACE_READS|TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    LinkTraceProc, linkPtr);
+    if (code != TCL_OK) {
+	Tcl_DecrRefCount(linkPtr->varName);
+	ckfree(linkPtr);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnlinkVar --
+ *
+ *	Destroy the link between a Tcl variable and a C variable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If "varName" was previously linked to a C variable, the link is broken
+ *	to make the variable independent. If there was no previous link for
+ *	"varName" then nothing happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_UnlinkVar(
+    Tcl_Interp *interp,		/* Interpreter containing variable to unlink */
+    const char *varName)	/* Global variable in interp to unlink. */
+{
+    Link *linkPtr = (Link *) Tcl_VarTraceInfo2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY, LinkTraceProc, NULL);
+
+    if (linkPtr == NULL) {
+	return;
+    }
+    Tcl_UntraceVar2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY|TCL_TRACE_READS|TCL_TRACE_WRITES|TCL_TRACE_UNSETS,
+	    LinkTraceProc, linkPtr);
+    Tcl_DecrRefCount(linkPtr->varName);
+    ckfree(linkPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UpdateLinkedVar --
+ *
+ *	This function is invoked after a linked variable has been changed by C
+ *	code. It updates the Tcl variable so that traces on the variable will
+ *	trigger.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The Tcl variable "varName" is updated from its C value, causing traces
+ *	on the variable to trigger.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_UpdateLinkedVar(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *varName)	/* Name of global variable that is linked. */
+{
+    Link *linkPtr = (Link *) Tcl_VarTraceInfo2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY, LinkTraceProc, NULL);
+    int savedFlag;
+
+    if (linkPtr == NULL) {
+	return;
+    }
+    savedFlag = linkPtr->flags & LINK_BEING_UPDATED;
+    linkPtr->flags |= LINK_BEING_UPDATED;
+    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+	    TCL_GLOBAL_ONLY);
+    /*
+     * Callback may have unlinked the variable. [Bug 1740631]
+     */
+    linkPtr = (Link *) Tcl_VarTraceInfo2(interp, varName, NULL,
+	    TCL_GLOBAL_ONLY, LinkTraceProc, NULL);
+    if (linkPtr != NULL) {
+	linkPtr->flags = (linkPtr->flags & ~LINK_BEING_UPDATED) | savedFlag;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * LinkTraceProc --
+ *
+ *	This function is invoked when a linked Tcl variable is read, written,
+ *	or unset from Tcl. It's responsible for keeping the C variable in sync
+ *	with the Tcl variable.
+ *
+ * Results:
+ *	If all goes well, NULL is returned; otherwise an error message is
+ *	returned.
+ *
+ * Side effects:
+ *	The C variable may be updated to make it consistent with the Tcl
+ *	variable, or the Tcl variable may be overwritten to reject a
+ *	modification.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+LinkTraceProc(
+    ClientData clientData,	/* Contains information about the link. */
+    Tcl_Interp *interp,		/* Interpreter containing Tcl variable. */
+    const char *name1,		/* First part of variable name. */
+    const char *name2,		/* Second part of variable name. */
+    int flags)			/* Miscellaneous additional information. */
+{
+    Link *linkPtr = clientData;
+    int changed;
+    size_t valueLength;
+    const char *value;
+    char **pp;
+    Tcl_Obj *valueObj;
+    int valueInt;
+    Tcl_WideInt valueWide;
+    double valueDouble;
+
+    /*
+     * If the variable is being unset, then just re-create it (with a trace)
+     * unless the whole interpreter is going away.
+     */
+
+    if (flags & TCL_TRACE_UNSETS) {
+	if (Tcl_InterpDeleted(interp)) {
+	    Tcl_DecrRefCount(linkPtr->varName);
+	    ckfree(linkPtr);
+	} else if (flags & TCL_TRACE_DESTROYED) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    Tcl_TraceVar2(interp, Tcl_GetString(linkPtr->varName), NULL,
+		    TCL_GLOBAL_ONLY|TCL_TRACE_READS|TCL_TRACE_WRITES
+		    |TCL_TRACE_UNSETS, LinkTraceProc, linkPtr);
+	}
+	return NULL;
+    }
+
+    /*
+     * If we were invoked because of a call to Tcl_UpdateLinkedVar, then don't
+     * do anything at all. In particular, we don't want to get upset that the
+     * variable is being modified, even if it is supposed to be read-only.
+     */
+
+    if (linkPtr->flags & LINK_BEING_UPDATED) {
+	return NULL;
+    }
+
+    /*
+     * For read accesses, update the Tcl variable if the C variable has
+     * changed since the last time we updated the Tcl variable.
+     */
+
+    if (flags & TCL_TRACE_READS) {
+	switch (linkPtr->type) {
+	case TCL_LINK_INT:
+	case TCL_LINK_BOOLEAN:
+	    changed = (LinkedVar(int) != linkPtr->lastValue.i);
+	    break;
+	case TCL_LINK_DOUBLE:
+	    changed = (LinkedVar(double) != linkPtr->lastValue.d);
+	    break;
+	case TCL_LINK_WIDE_INT:
+	    changed = (LinkedVar(Tcl_WideInt) != linkPtr->lastValue.w);
+	    break;
+	case TCL_LINK_WIDE_UINT:
+	    changed = (LinkedVar(Tcl_WideUInt) != linkPtr->lastValue.uw);
+	    break;
+	case TCL_LINK_CHAR:
+	    changed = (LinkedVar(char) != linkPtr->lastValue.c);
+	    break;
+	case TCL_LINK_UCHAR:
+	    changed = (LinkedVar(unsigned char) != linkPtr->lastValue.uc);
+	    break;
+	case TCL_LINK_SHORT:
+	    changed = (LinkedVar(short) != linkPtr->lastValue.s);
+	    break;
+	case TCL_LINK_USHORT:
+	    changed = (LinkedVar(unsigned short) != linkPtr->lastValue.us);
+	    break;
+	case TCL_LINK_UINT:
+	    changed = (LinkedVar(unsigned int) != linkPtr->lastValue.ui);
+	    break;
+#if !defined(TCL_WIDE_INT_IS_LONG) && !defined(_WIN32) && !defined(__CYGWIN__)
+	case TCL_LINK_LONG:
+	    changed = (LinkedVar(long) != linkPtr->lastValue.l);
+	    break;
+	case TCL_LINK_ULONG:
+	    changed = (LinkedVar(unsigned long) != linkPtr->lastValue.ul);
+	    break;
+#endif
+	case TCL_LINK_FLOAT:
+	    changed = (LinkedVar(float) != linkPtr->lastValue.f);
+	    break;
+	case TCL_LINK_STRING:
+	    changed = 1;
+	    break;
+	default:
+	    return (char *) "internal error: bad linked variable type";
+	}
+	if (changed) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	}
+	return NULL;
+    }
+
+    /*
+     * For writes, first make sure that the variable is writable. Then convert
+     * the Tcl value to C if possible. If the variable isn't writable or can't
+     * be converted, then restore the varaible's old value and return an
+     * error. Another tricky thing: we have to save and restore the interp's
+     * result, since the variable access could occur when the result has been
+     * partially set.
+     */
+
+    if (linkPtr->flags & LINK_READ_ONLY) {
+	Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		TCL_GLOBAL_ONLY);
+	return (char *) "linked variable is read-only";
+    }
+    valueObj = Tcl_ObjGetVar2(interp, linkPtr->varName,NULL, TCL_GLOBAL_ONLY);
+    if (valueObj == NULL) {
+	/*
+	 * This shouldn't ever happen.
+	 */
+
+	return (char *) "internal error: linked variable couldn't be read";
+    }
+
+    switch (linkPtr->type) {
+    case TCL_LINK_INT:
+	if (Tcl_GetIntFromObj(NULL, valueObj, &linkPtr->lastValue.i) != TCL_OK
+		&& GetInvalidIntFromObj(valueObj, &linkPtr->lastValue.i) != TCL_OK) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have integer value";
+	}
+	LinkedVar(int) = linkPtr->lastValue.i;
+	break;
+
+    case TCL_LINK_WIDE_INT:
+	if (Tcl_GetWideIntFromObj(NULL, valueObj, &linkPtr->lastValue.w) != TCL_OK
+		&& GetInvalidWideFromObj(valueObj, &linkPtr->lastValue.w) != TCL_OK) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have integer value";
+	}
+	LinkedVar(Tcl_WideInt) = linkPtr->lastValue.w;
+	break;
+
+    case TCL_LINK_DOUBLE:
+	if (Tcl_GetDoubleFromObj(NULL, valueObj, &linkPtr->lastValue.d) != TCL_OK) {
+#ifdef ACCEPT_NAN
+	    if (valueObj->typePtr != &tclDoubleType) {
+#endif
+		if (GetInvalidDoubleFromObj(valueObj, &linkPtr->lastValue.d) != TCL_OK) {
+		    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+			TCL_GLOBAL_ONLY);
+		    return (char *) "variable must have real value";
+		}
+#ifdef ACCEPT_NAN
+	    }
+	    linkPtr->lastValue.d = valueObj->internalRep.doubleValue;
+#endif
+	}
+	LinkedVar(double) = linkPtr->lastValue.d;
+	break;
+
+    case TCL_LINK_BOOLEAN:
+	if (Tcl_GetBooleanFromObj(NULL, valueObj, &linkPtr->lastValue.i) != TCL_OK) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have boolean value";
+	}
+	LinkedVar(int) = linkPtr->lastValue.i;
+	break;
+
+    case TCL_LINK_CHAR:
+	if ((Tcl_GetIntFromObj(NULL, valueObj, &valueInt) != TCL_OK
+		&& GetInvalidIntFromObj(valueObj, &valueInt) != TCL_OK)
+		|| valueInt < SCHAR_MIN || valueInt > SCHAR_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have char value";
+	}
+	LinkedVar(char) = linkPtr->lastValue.c = (char)valueInt;
+	break;
+
+    case TCL_LINK_UCHAR:
+	if ((Tcl_GetIntFromObj(NULL, valueObj, &valueInt) != TCL_OK
+		&& GetInvalidIntFromObj(valueObj, &valueInt) != TCL_OK)
+		|| valueInt < 0 || valueInt > UCHAR_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have unsigned char value";
+	}
+	LinkedVar(unsigned char) = linkPtr->lastValue.uc = (unsigned char) valueInt;
+	break;
+
+    case TCL_LINK_SHORT:
+	if ((Tcl_GetIntFromObj(NULL, valueObj, &valueInt) != TCL_OK
+		&& GetInvalidIntFromObj(valueObj, &valueInt) != TCL_OK)
+		|| valueInt < SHRT_MIN || valueInt > SHRT_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have short value";
+	}
+	LinkedVar(short) = linkPtr->lastValue.s = (short)valueInt;
+	break;
+
+    case TCL_LINK_USHORT:
+	if ((Tcl_GetIntFromObj(NULL, valueObj, &valueInt) != TCL_OK
+		&& GetInvalidIntFromObj(valueObj, &valueInt) != TCL_OK)
+		|| valueInt < 0 || valueInt > USHRT_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have unsigned short value";
+	}
+	LinkedVar(unsigned short) = linkPtr->lastValue.us = (unsigned short)valueInt;
+	break;
+
+    case TCL_LINK_UINT:
+	if ((Tcl_GetWideIntFromObj(NULL, valueObj, &valueWide) != TCL_OK
+		&& GetInvalidWideFromObj(valueObj, &valueWide) != TCL_OK)
+		|| valueWide < 0 || valueWide > UINT_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have unsigned int value";
+	}
+	LinkedVar(unsigned int) = linkPtr->lastValue.ui = (unsigned int)valueWide;
+	break;
+
+#if !defined(TCL_WIDE_INT_IS_LONG) && !defined(_WIN32) && !defined(__CYGWIN__)
+    case TCL_LINK_LONG:
+	if ((Tcl_GetWideIntFromObj(NULL, valueObj, &valueWide) != TCL_OK
+		&& GetInvalidWideFromObj(valueObj, &valueWide) != TCL_OK)
+		|| valueWide < LONG_MIN || valueWide > LONG_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have long value";
+	}
+	LinkedVar(long) = linkPtr->lastValue.l = (long)valueWide;
+	break;
+
+    case TCL_LINK_ULONG:
+	if ((Tcl_GetWideIntFromObj(NULL, valueObj, &valueWide) != TCL_OK
+		&& GetInvalidWideFromObj(valueObj, &valueWide) != TCL_OK)
+		|| valueWide < 0 || (Tcl_WideUInt) valueWide > ULONG_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have unsigned long value";
+	}
+	LinkedVar(unsigned long) = linkPtr->lastValue.ul = (unsigned long)valueWide;
+	break;
+#endif
+
+    case TCL_LINK_WIDE_UINT:
+	/*
+	 * FIXME: represent as a bignum.
+	 */
+	if (Tcl_GetWideIntFromObj(NULL, valueObj, &valueWide) != TCL_OK
+		&& GetInvalidWideFromObj(valueObj, &valueWide) != TCL_OK) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have unsigned wide int value";
+	}
+	LinkedVar(Tcl_WideUInt) = linkPtr->lastValue.uw = (Tcl_WideUInt)valueWide;
+	break;
+
+    case TCL_LINK_FLOAT:
+	if ((Tcl_GetDoubleFromObj(NULL, valueObj, &valueDouble) != TCL_OK
+		&& GetInvalidDoubleFromObj(valueObj, &valueDouble) != TCL_OK)
+		|| valueDouble < -FLT_MAX || valueDouble > FLT_MAX) {
+	    Tcl_ObjSetVar2(interp, linkPtr->varName, NULL, ObjValue(linkPtr),
+		    TCL_GLOBAL_ONLY);
+	    return (char *) "variable must have float value";
+	}
+	LinkedVar(float) = linkPtr->lastValue.f = (float)valueDouble;
+	break;
+
+    case TCL_LINK_STRING:
+	value = TclGetString(valueObj);
+	valueLength = valueObj->length + 1;
+	pp = (char **) linkPtr->addr;
+
+	*pp = ckrealloc(*pp, valueLength);
+	memcpy(*pp, value, valueLength);
+	break;
+
+    default:
+	return (char *) "internal error: bad linked variable type";
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ObjValue --
+ *
+ *	Converts the value of a C variable to a Tcl_Obj* for use in a Tcl
+ *	variable to which it is linked.
+ *
+ * Results:
+ *	The return value is a pointer to a Tcl_Obj that represents the value
+ *	of the C variable given by linkPtr.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+ObjValue(
+    Link *linkPtr)		/* Structure describing linked variable. */
+{
+    char *p;
+    Tcl_Obj *resultObj;
+
+    switch (linkPtr->type) {
+    case TCL_LINK_INT:
+	linkPtr->lastValue.i = LinkedVar(int);
+	return Tcl_NewIntObj(linkPtr->lastValue.i);
+    case TCL_LINK_WIDE_INT:
+	linkPtr->lastValue.w = LinkedVar(Tcl_WideInt);
+	return Tcl_NewWideIntObj(linkPtr->lastValue.w);
+    case TCL_LINK_DOUBLE:
+	linkPtr->lastValue.d = LinkedVar(double);
+	return Tcl_NewDoubleObj(linkPtr->lastValue.d);
+    case TCL_LINK_BOOLEAN:
+	linkPtr->lastValue.i = LinkedVar(int);
+	return Tcl_NewBooleanObj(linkPtr->lastValue.i);
+    case TCL_LINK_CHAR:
+	linkPtr->lastValue.c = LinkedVar(char);
+	return Tcl_NewIntObj(linkPtr->lastValue.c);
+    case TCL_LINK_UCHAR:
+	linkPtr->lastValue.uc = LinkedVar(unsigned char);
+	return Tcl_NewIntObj(linkPtr->lastValue.uc);
+    case TCL_LINK_SHORT:
+	linkPtr->lastValue.s = LinkedVar(short);
+	return Tcl_NewIntObj(linkPtr->lastValue.s);
+    case TCL_LINK_USHORT:
+	linkPtr->lastValue.us = LinkedVar(unsigned short);
+	return Tcl_NewIntObj(linkPtr->lastValue.us);
+    case TCL_LINK_UINT:
+	linkPtr->lastValue.ui = LinkedVar(unsigned int);
+	return Tcl_NewWideIntObj((Tcl_WideInt) linkPtr->lastValue.ui);
+#if !defined(TCL_WIDE_INT_IS_LONG) && !defined(_WIN32) && !defined(__CYGWIN__)
+    case TCL_LINK_LONG:
+	linkPtr->lastValue.l = LinkedVar(long);
+	return Tcl_NewWideIntObj((Tcl_WideInt) linkPtr->lastValue.l);
+    case TCL_LINK_ULONG:
+	linkPtr->lastValue.ul = LinkedVar(unsigned long);
+	return Tcl_NewWideIntObj((Tcl_WideInt) linkPtr->lastValue.ul);
+#endif
+    case TCL_LINK_FLOAT:
+	linkPtr->lastValue.f = LinkedVar(float);
+	return Tcl_NewDoubleObj(linkPtr->lastValue.f);
+    case TCL_LINK_WIDE_UINT:
+	linkPtr->lastValue.uw = LinkedVar(Tcl_WideUInt);
+	/*
+	 * FIXME: represent as a bignum.
+	 */
+	return Tcl_NewWideIntObj((Tcl_WideInt) linkPtr->lastValue.uw);
+    case TCL_LINK_STRING:
+	p = LinkedVar(char *);
+	if (p == NULL) {
+	    TclNewLiteralStringObj(resultObj, "NULL");
+	    return resultObj;
+	}
+	return Tcl_NewStringObj(p, -1);
+
+    /*
+     * This code only gets executed if the link type is unknown (shouldn't
+     * ever happen).
+     */
+
+    default:
+	TclNewLiteralStringObj(resultObj, "??");
+	return resultObj;
+    }
+}
+
+static int SetInvalidRealFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+
+static Tcl_ObjType invalidRealType = {
+    "invalidReal",			/* name */
+    NULL,				/* freeIntRepProc */
+    NULL,				/* dupIntRepProc */
+    NULL,				/* updateStringProc */
+    SetInvalidRealFromAny		/* setFromAnyProc */
+};
+
+static int
+SetInvalidRealFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr) {
+    const char *str;
+    const char *endPtr;
+
+    str = TclGetString(objPtr);
+    if ((objPtr->length == 1) && (str[0] == '.')){
+	objPtr->typePtr = &invalidRealType;
+	objPtr->internalRep.doubleValue = 0.0;
+	return TCL_OK;
+    }
+    if (TclParseNumber(NULL, objPtr, NULL, str, objPtr->length, &endPtr,
+	    TCL_PARSE_DECIMAL_ONLY) == TCL_OK) {
+	/* If number is followed by [eE][+-]?, then it is an invalid
+	 * double, but it could be the start of a valid double. */
+	if (*endPtr == 'e' || *endPtr == 'E') {
+	    ++endPtr;
+	    if (*endPtr == '+' || *endPtr == '-') ++endPtr;
+	    if (*endPtr == 0) {
+		double doubleValue = 0.0;
+		Tcl_GetDoubleFromObj(NULL, objPtr, &doubleValue);
+		TclFreeIntRep(objPtr);
+		objPtr->typePtr = &invalidRealType;
+		objPtr->internalRep.doubleValue = doubleValue;
+		return TCL_OK;
+	    }
+	}
+    }
+    return TCL_ERROR;
+}
+
+
+/*
+ * This function checks for integer representations, which are valid
+ * when linking with C variables, but which are invalid in other
+ * contexts in Tcl. Handled are "+", "-", "", "0x", "0b", "0d" and "0o"
+ * (upperand lowercase). See bug [39f6304c2e].
+ */
+int
+GetInvalidIntFromObj(Tcl_Obj *objPtr, int *intPtr)
+{
+    const char *str = TclGetString(objPtr);
+
+    if ((objPtr->length == 0) ||
+	    ((objPtr->length == 2) && (str[0] == '0') && strchr("xXbBoOdD", str[1]))) {
+	*intPtr = 0;
+	return TCL_OK;
+    } else if ((objPtr->length == 1) && strchr("+-", str[0])) {
+	*intPtr = (str[0] == '+');
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+int
+GetInvalidWideFromObj(Tcl_Obj *objPtr, Tcl_WideInt *widePtr)
+{
+    int intValue;
+
+    if (GetInvalidIntFromObj(objPtr, &intValue) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    *widePtr = intValue;
+    return TCL_OK;
+}
+
+/*
+ * This function checks for double representations, which are valid
+ * when linking with C variables, but which are invalid in other
+ * contexts in Tcl. Handled are "+", "-", "", ".", "0x", "0b" and "0o"
+ * (upper- and lowercase) and sequences like "1e-". See bug [39f6304c2e].
+ */
+int
+GetInvalidDoubleFromObj(Tcl_Obj *objPtr, double *doublePtr)
+{
+    int intValue;
+
+    if (objPtr->typePtr == &invalidRealType) {
+	goto gotdouble;
+    }
+    if (GetInvalidIntFromObj(objPtr, &intValue) == TCL_OK) {
+	*doublePtr = (double) intValue;
+	return TCL_OK;
+    }
+    if (SetInvalidRealFromAny(NULL, objPtr) == TCL_OK) {
+    gotdouble:
+	*doublePtr = objPtr->internalRep.doubleValue;
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclListObj.c b/generic/tclListObj.c
new file mode 100644
index 0000000..11374cc
--- /dev/null
+++ b/generic/tclListObj.c
@@ -0,0 +1,2040 @@
+/*
+ * tclListObj.c --
+ *
+ *	This file contains functions that implement the Tcl list object type.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998 by Scriptics Corporation.
+ * Copyright (c) 2001 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Prototypes for functions defined later in this file:
+ */
+
+static List *		AttemptNewList(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static List *		NewListIntRep(int objc, Tcl_Obj *const objv[], int p);
+static void		DupListInternalRep(Tcl_Obj *srcPtr, Tcl_Obj *copyPtr);
+static void		FreeListInternalRep(Tcl_Obj *listPtr);
+static int		SetListFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		UpdateStringOfList(Tcl_Obj *listPtr);
+
+/*
+ * The structure below defines the list Tcl object type by means of functions
+ * that can be invoked by generic object code.
+ *
+ * The internal representation of a list object is a two-pointer
+ * representation. The first pointer designates a List structure that contains
+ * an array of pointers to the element objects, together with integers that
+ * represent the current element count and the allocated size of the array.
+ * The second pointer is normally NULL; during execution of functions in this
+ * file that operate on nested sublists, it is occasionally used as working
+ * storage to avoid an auxiliary stack.
+ */
+
+const Tcl_ObjType tclListType = {
+    "list",			/* name */
+    FreeListInternalRep,	/* freeIntRepProc */
+    DupListInternalRep,		/* dupIntRepProc */
+    UpdateStringOfList,		/* updateStringProc */
+    SetListFromAny		/* setFromAnyProc */
+};
+
+#ifndef TCL_MIN_ELEMENT_GROWTH
+#define TCL_MIN_ELEMENT_GROWTH TCL_MIN_GROWTH/sizeof(Tcl_Obj *)
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NewListIntRep --
+ *
+ *	Creates a list internal rep with space for objc elements.  objc
+ *	must be > 0.  If objv!=NULL, initializes with the first objc values
+ *	in that array.  If objv==NULL, initalize list internal rep to have
+ *	0 elements, with space to add objc more.  Flag value "p" indicates
+ *	how to behave on failure.
+ *
+ * Results:
+ *	A new List struct with refCount 0 is returned. If some failure
+ *	prevents this then if p=0, NULL is returned and otherwise the
+ *	routine panics.
+ *
+ * Side effects:
+ *	The ref counts of the elements in objv are incremented since the
+ *	resulting list now refers to them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static List *
+NewListIntRep(
+    int objc,
+    Tcl_Obj *const objv[],
+    int p)
+{
+    List *listRepPtr;
+
+    if (objc <= 0) {
+	Tcl_Panic("NewListIntRep: expects postive element count");
+    }
+
+    /*
+     * First check to see if we'd overflow and try to allocate an object
+     * larger than our memory allocator allows. Note that this is actually a
+     * fairly small value when you're on a serious 64-bit machine, but that
+     * requires API changes to fix. See [Bug 219196] for a discussion.
+     */
+
+    if ((size_t)objc > LIST_MAX) {
+	if (p) {
+	    Tcl_Panic("max length of a Tcl list (%d elements) exceeded",
+		    LIST_MAX);
+	}
+	return NULL;
+    }
+
+    listRepPtr = attemptckalloc(LIST_SIZE(objc));
+    if (listRepPtr == NULL) {
+	if (p) {
+	    Tcl_Panic("list creation failed: unable to alloc %u bytes",
+		    LIST_SIZE(objc));
+	}
+	return NULL;
+    }
+
+    listRepPtr->canonicalFlag = 0;
+    listRepPtr->refCount = 0;
+    listRepPtr->maxElemCount = objc;
+
+    if (objv) {
+	Tcl_Obj **elemPtrs;
+	int i;
+
+	listRepPtr->elemCount = objc;
+	elemPtrs = &listRepPtr->elements;
+	for (i = 0;  i < objc;  i++) {
+	    elemPtrs[i] = objv[i];
+	    Tcl_IncrRefCount(elemPtrs[i]);
+	}
+    } else {
+	listRepPtr->elemCount = 0;
+    }
+    return listRepPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AttemptNewList --
+ *
+ *	Creates a list internal rep with space for objc elements.  objc
+ *	must be > 0.  If objv!=NULL, initializes with the first objc values
+ *	in that array.  If objv==NULL, initalize list internal rep to have
+ *	0 elements, with space to add objc more.
+ *
+ * Results:
+ *	A new List struct with refCount 0 is returned. If some failure
+ *	prevents this then NULL is returned, and an error message is left
+ *	in the interp result, unless interp is NULL.
+ *
+ * Side effects:
+ *	The ref counts of the elements in objv are incremented since the
+ *	resulting list now refers to them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static List *
+AttemptNewList(
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    List *listRepPtr = NewListIntRep(objc, objv, 0);
+
+    if (interp != NULL && listRepPtr == NULL) {
+	if (objc > LIST_MAX) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "max length of a Tcl list (%d elements) exceeded",
+		    LIST_MAX));
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "list creation failed: unable to alloc %u bytes",
+		    LIST_SIZE(objc)));
+	}
+	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+    }
+    return listRepPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewListObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates a new list object from an
+ *	(objc,objv) array: that is, each of the objc elements of the array
+ *	referenced by objv is inserted as an element into a new Tcl object.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewListObj.
+ *
+ * Results:
+ *	A new list object is returned that is initialized from the object
+ *	pointers in objv. If objc is less than or equal to zero, an empty
+ *	object is returned. The new object's string representation is left
+ *	NULL. The resulting new list object has ref count 0.
+ *
+ * Side effects:
+ *	The ref counts of the elements in objv are incremented since the
+ *	resulting list now refers to them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewListObj
+
+Tcl_Obj *
+Tcl_NewListObj(
+    int objc,			/* Count of objects referenced by objv. */
+    Tcl_Obj *const objv[])	/* An array of pointers to Tcl objects. */
+{
+    return Tcl_DbNewListObj(objc, objv, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewListObj(
+    int objc,			/* Count of objects referenced by objv. */
+    Tcl_Obj *const objv[])	/* An array of pointers to Tcl objects. */
+{
+    List *listRepPtr;
+    Tcl_Obj *listPtr;
+
+    TclNewObj(listPtr);
+
+    if (objc <= 0) {
+	return listPtr;
+    }
+
+    /*
+     * Create the internal rep.
+     */
+
+    listRepPtr = NewListIntRep(objc, objv, 1);
+
+    /*
+     * Now create the object.
+     */
+
+    TclInvalidateStringRep(listPtr);
+    ListSetIntRep(listPtr, listRepPtr);
+    return listPtr;
+}
+#endif /* if TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewListObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new list objects. It is the same
+ *	as the Tcl_NewListObj function above except that it calls
+ *	Tcl_DbCkalloc directly with the file name and line number from its
+ *	caller. This simplifies debugging since then the [memory active]
+ *	command will report the correct file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewListObj.
+ *
+ * Results:
+ *	A new list object is returned that is initialized from the object
+ *	pointers in objv. If objc is less than or equal to zero, an empty
+ *	object is returned. The new object's string representation is left
+ *	NULL. The new list object has ref count 0.
+ *
+ * Side effects:
+ *	The ref counts of the elements in objv are incremented since the
+ *	resulting list now refers to them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewListObj(
+    int objc,			/* Count of objects referenced by objv. */
+    Tcl_Obj *const objv[],	/* An array of pointers to Tcl objects. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    Tcl_Obj *listPtr;
+    List *listRepPtr;
+
+    TclDbNewObj(listPtr, file, line);
+
+    if (objc <= 0) {
+	return listPtr;
+    }
+
+    /*
+     * Create the internal rep.
+     */
+
+    listRepPtr = NewListIntRep(objc, objv, 1);
+
+    /*
+     * Now create the object.
+     */
+
+    TclInvalidateStringRep(listPtr);
+    ListSetIntRep(listPtr, listRepPtr);
+
+    return listPtr;
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewListObj(
+    int objc,			/* Count of objects referenced by objv. */
+    Tcl_Obj *const objv[],	/* An array of pointers to Tcl objects. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewListObj(objc, objv);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetListObj --
+ *
+ *	Modify an object to be a list containing each of the objc elements of
+ *	the object array referenced by objv.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object is made a list object and is initialized from the object
+ *	pointers in objv. If objc is less than or equal to zero, an empty
+ *	object is returned. The new object's string representation is left
+ *	NULL. The ref counts of the elements in objv are incremented since the
+ *	list now refers to them. The object's old string and internal
+ *	representations are freed and its type is set NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetListObj(
+    Tcl_Obj *objPtr,		/* Object whose internal rep to init. */
+    int objc,			/* Count of objects referenced by objv. */
+    Tcl_Obj *const objv[])	/* An array of pointers to Tcl objects. */
+{
+    List *listRepPtr;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetListObj");
+    }
+
+    /*
+     * Free any old string rep and any internal rep for the old type.
+     */
+
+    TclFreeIntRep(objPtr);
+    TclInvalidateStringRep(objPtr);
+
+    /*
+     * Set the object's type to "list" and initialize the internal rep.
+     * However, if there are no elements to put in the list, just give the
+     * object an empty string rep and a NULL type.
+     */
+
+    if (objc > 0) {
+	listRepPtr = NewListIntRep(objc, objv, 1);
+	ListSetIntRep(objPtr, listRepPtr);
+    } else {
+	objPtr->bytes = &tclEmptyString;
+	objPtr->length = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclListObjCopy --
+ *
+ *	Makes a "pure list" copy of a list value. This provides for the C
+ *	level a counterpart of the [lrange $list 0 end] command, while using
+ *	internals details to be as efficient as possible.
+ *
+ * Results:
+ *	Normally returns a pointer to a new Tcl_Obj, that contains the same
+ *	list value as *listPtr does. The returned Tcl_Obj has a refCount of
+ *	zero. If *listPtr does not hold a list, NULL is returned, and if
+ *	interp is non-NULL, an error message is recorded there.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclListObjCopy(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    Tcl_Obj *listPtr)		/* List object for which an element array is
+				 * to be returned. */
+{
+    Tcl_Obj *copyPtr;
+
+    if (listPtr->typePtr != &tclListType) {
+	if (SetListFromAny(interp, listPtr) != TCL_OK) {
+	    return NULL;
+	}
+    }
+
+    TclNewObj(copyPtr);
+    TclInvalidateStringRep(copyPtr);
+    DupListInternalRep(listPtr, copyPtr);
+    return copyPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjGetElements --
+ *
+ *	This function returns an (objc,objv) array of the elements in a list
+ *	object.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case *objcPtr is set to
+ *	the count of list elements and *objvPtr is set to a pointer to an
+ *	array of (*objcPtr) pointers to each list element. If listPtr does not
+ *	refer to a list object and the object can not be converted to one,
+ *	TCL_ERROR is returned and an error message will be left in the
+ *	interpreter's result if interp is not NULL.
+ *
+ *	The objects referenced by the returned array should be treated as
+ *	readonly and their ref counts are _not_ incremented; the caller must
+ *	do that if it holds on to a reference. Furthermore, the pointer and
+ *	length returned by this function may change as soon as any function is
+ *	called on the list object; be careful about retaining the pointer in a
+ *	local data structure.
+ *
+ * Side effects:
+ *	The possible conversion of the object referenced by listPtr
+ *	to a list object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjGetElements(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    register Tcl_Obj *listPtr,	/* List object for which an element array is
+				 * to be returned. */
+    int *objcPtr,		/* Where to store the count of objects
+				 * referenced by objv. */
+    Tcl_Obj ***objvPtr)		/* Where to store the pointer to an array of
+				 * pointers to the list's objects. */
+{
+    register List *listRepPtr;
+
+    if (listPtr->typePtr != &tclListType) {
+	int result;
+
+	if (listPtr->bytes == &tclEmptyString) {
+	    *objcPtr = 0;
+	    *objvPtr = NULL;
+	    return TCL_OK;
+	}
+	result = SetListFromAny(interp, listPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    listRepPtr = ListRepPtr(listPtr);
+    *objcPtr = listRepPtr->elemCount;
+    *objvPtr = &listRepPtr->elements;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjAppendList --
+ *
+ *	This function appends the elements in the list value referenced by
+ *	elemListPtr to the list value referenced by listPtr.
+ *
+ * Results:
+ *	The return value is normally TCL_OK. If listPtr or elemListPtr do not
+ *	refer to list values, TCL_ERROR is returned and an error message is
+ *	left in the interpreter's result if interp is not NULL.
+ *
+ * Side effects:
+ *	The reference counts of the elements in elemListPtr are incremented
+ *	since the list now refers to them. listPtr and elemListPtr are
+ *	converted, if necessary, to list objects. Also, appending the new
+ *	elements may cause listObj's array of element pointers to grow.
+ *	listPtr's old string representation, if any, is invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjAppendList(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    register Tcl_Obj *listPtr,	/* List object to append elements to. */
+    Tcl_Obj *elemListPtr)	/* List obj with elements to append. */
+{
+    int objc;
+    Tcl_Obj **objv;
+
+    if (Tcl_IsShared(listPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_ListObjAppendList");
+    }
+
+    /*
+     * Pull the elements to append from elemListPtr.
+     */
+
+    if (TCL_OK != TclListObjGetElements(interp, elemListPtr, &objc, &objv)) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Insert the new elements starting after the lists's last element.
+     * Delete zero existing elements.
+     */
+
+    return Tcl_ListObjReplace(interp, listPtr, LIST_MAX, 0, objc, objv);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjAppendElement --
+ *
+ *	This function is a special purpose version of Tcl_ListObjAppendList:
+ *	it appends a single object referenced by objPtr to the list object
+ *	referenced by listPtr. If listPtr is not already a list object, an
+ *	attempt will be made to convert it to one.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case objPtr is added to
+ *	the end of listPtr's list. If listPtr does not refer to a list object
+ *	and the object can not be converted to one, TCL_ERROR is returned and
+ *	an error message will be left in the interpreter's result if interp is
+ *	not NULL.
+ *
+ * Side effects:
+ *	The ref count of objPtr is incremented since the list now refers to
+ *	it. listPtr will be converted, if necessary, to a list object. Also,
+ *	appending the new element may cause listObj's array of element
+ *	pointers to grow. listPtr's old string representation, if any, is
+ *	invalidated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjAppendElement(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    Tcl_Obj *listPtr,		/* List object to append objPtr to. */
+    Tcl_Obj *objPtr)		/* Object to append to listPtr's list. */
+{
+    register List *listRepPtr, *newPtr = NULL;
+    int numElems, numRequired, needGrow, isShared, attempt;
+
+    if (Tcl_IsShared(listPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_ListObjAppendElement");
+    }
+    if (listPtr->typePtr != &tclListType) {
+	int result;
+
+	if (listPtr->bytes == &tclEmptyString) {
+	    Tcl_SetListObj(listPtr, 1, &objPtr);
+	    return TCL_OK;
+	}
+	result = SetListFromAny(interp, listPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+
+    listRepPtr = ListRepPtr(listPtr);
+    numElems = listRepPtr->elemCount;
+    numRequired = numElems + 1 ;
+    needGrow = (numRequired > listRepPtr->maxElemCount);
+    isShared = (listRepPtr->refCount > 1);
+
+    if (numRequired > LIST_MAX) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "max length of a Tcl list (%d elements) exceeded",
+		    LIST_MAX));
+	    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    if (needGrow && !isShared) {
+	/*
+	 * Need to grow + unshared intrep => try to realloc
+	 */
+
+	attempt = 2 * numRequired;
+	if (attempt <= LIST_MAX) {
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr == NULL) {
+	    attempt = numRequired + 1 + TCL_MIN_ELEMENT_GROWTH;
+	    if (attempt > LIST_MAX) {
+		attempt = LIST_MAX;
+	    }
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr == NULL) {
+	    attempt = numRequired;
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr) {
+	    listRepPtr = newPtr;
+	    listRepPtr->maxElemCount = attempt;
+	    needGrow = 0;
+	}
+    }
+    if (isShared || needGrow) {
+	Tcl_Obj **dst, **src = &listRepPtr->elements;
+
+	/*
+	 * Either we have a shared intrep and we must copy to write, or we
+	 * need to grow and realloc attempts failed.  Attempt intrep copy.
+	 */
+
+	attempt = 2 * numRequired;
+	newPtr = AttemptNewList(NULL, attempt, NULL);
+	if (newPtr == NULL) {
+	    attempt = numRequired + 1 + TCL_MIN_ELEMENT_GROWTH;
+	    if (attempt > LIST_MAX) {
+		attempt = LIST_MAX;
+	    }
+	    newPtr = AttemptNewList(NULL, attempt, NULL);
+	}
+	if (newPtr == NULL) {
+	    attempt = numRequired;
+	    newPtr = AttemptNewList(interp, attempt, NULL);
+	}
+	if (newPtr == NULL) {
+	    /*
+	     * All growth attempts failed; throw the error.
+	     */
+
+	    return TCL_ERROR;
+	}
+
+	dst = &newPtr->elements;
+	newPtr->refCount++;
+	newPtr->canonicalFlag = listRepPtr->canonicalFlag;
+	newPtr->elemCount = listRepPtr->elemCount;
+
+	if (isShared) {
+	    /*
+	     * The original intrep must remain undisturbed.  Copy into the new
+	     * one and bump refcounts
+	     */
+	    while (numElems--) {
+		*dst = *src++;
+		Tcl_IncrRefCount(*dst++);
+	    }
+	    listRepPtr->refCount--;
+	} else {
+	    /*
+	     * Old intrep to be freed, re-use refCounts.
+	     */
+
+	    memcpy(dst, src, (size_t) numElems * sizeof(Tcl_Obj *));
+	    ckfree(listRepPtr);
+	}
+	listRepPtr = newPtr;
+    }
+    listPtr->internalRep.twoPtrValue.ptr1 = listRepPtr;
+
+    /*
+     * Add objPtr to the end of listPtr's array of element pointers. Increment
+     * the ref count for the (now shared) objPtr.
+     */
+
+    *(&listRepPtr->elements + listRepPtr->elemCount) = objPtr;
+    Tcl_IncrRefCount(objPtr);
+    listRepPtr->elemCount++;
+
+    /*
+     * Invalidate any old string representation since the list's internal
+     * representation has changed.
+     */
+
+    TclInvalidateStringRep(listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjIndex --
+ *
+ *	This function returns a pointer to the index'th object from the list
+ *	referenced by listPtr. The first element has index 0. If index is
+ *	negative or greater than or equal to the number of elements in the
+ *	list, a NULL is returned. If listPtr is not a list object, an attempt
+ *	will be made to convert it to a list.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case objPtrPtr is set to
+ *	the Tcl_Obj pointer for the index'th list element or NULL if index is
+ *	out of range. This object should be treated as readonly and its ref
+ *	count is _not_ incremented; the caller must do that if it holds on to
+ *	the reference. If listPtr does not refer to a list and can't be
+ *	converted to one, TCL_ERROR is returned and an error message is left
+ *	in the interpreter's result if interp is not NULL.
+ *
+ * Side effects:
+ *	listPtr will be converted, if necessary, to a list object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjIndex(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    register Tcl_Obj *listPtr,	/* List object to index into. */
+    register int index,		/* Index of element to return. */
+    Tcl_Obj **objPtrPtr)	/* The resulting Tcl_Obj* is stored here. */
+{
+    register List *listRepPtr;
+
+    if (listPtr->typePtr != &tclListType) {
+	int result;
+
+	if (listPtr->bytes == &tclEmptyString) {
+	    *objPtrPtr = NULL;
+	    return TCL_OK;
+	}
+	result = SetListFromAny(interp, listPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+
+    listRepPtr = ListRepPtr(listPtr);
+    if ((index < 0) || (index >= listRepPtr->elemCount)) {
+	*objPtrPtr = NULL;
+    } else {
+	*objPtrPtr = (&listRepPtr->elements)[index];
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjLength --
+ *
+ *	This function returns the number of elements in a list object. If the
+ *	object is not already a list object, an attempt will be made to
+ *	convert it to one.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case *intPtr will be set
+ *	to the integer count of list elements. If listPtr does not refer to a
+ *	list object and the object can not be converted to one, TCL_ERROR is
+ *	returned and an error message will be left in the interpreter's result
+ *	if interp is not NULL.
+ *
+ * Side effects:
+ *	The possible conversion of the argument object to a list object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjLength(
+    Tcl_Interp *interp,		/* Used to report errors if not NULL. */
+    register Tcl_Obj *listPtr,	/* List object whose #elements to return. */
+    register int *intPtr)	/* The resulting int is stored here. */
+{
+    register List *listRepPtr;
+
+    if (listPtr->typePtr != &tclListType) {
+	int result;
+
+	if (listPtr->bytes == &tclEmptyString) {
+	    *intPtr = 0;
+	    return TCL_OK;
+	}
+	result = SetListFromAny(interp, listPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+
+    listRepPtr = ListRepPtr(listPtr);
+    *intPtr = listRepPtr->elemCount;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ListObjReplace --
+ *
+ *	This function replaces zero or more elements of the list referenced by
+ *	listPtr with the objects from an (objc,objv) array. The objc elements
+ *	of the array referenced by objv replace the count elements in listPtr
+ *	starting at first.
+ *
+ *	If the argument first is zero or negative, it refers to the first
+ *	element. If first is greater than or equal to the number of elements
+ *	in the list, then no elements are deleted; the new elements are
+ *	appended to the list. Count gives the number of elements to replace.
+ *	If count is zero or negative then no elements are deleted; the new
+ *	elements are simply inserted before first.
+ *
+ *	The argument objv refers to an array of objc pointers to the new
+ *	elements to be added to listPtr in place of those that were deleted.
+ *	If objv is NULL, no new elements are added. If listPtr is not a list
+ *	object, an attempt will be made to convert it to one.
+ *
+ * Results:
+ *	The return value is normally TCL_OK. If listPtr does not refer to a
+ *	list object and can not be converted to one, TCL_ERROR is returned and
+ *	an error message will be left in the interpreter's result if interp is
+ *	not NULL.
+ *
+ * Side effects:
+ *	The ref counts of the objc elements in objv are incremented since the
+ *	resulting list now refers to them. Similarly, the ref counts for
+ *	replaced objects are decremented. listPtr is converted, if necessary,
+ *	to a list object. listPtr's old string representation, if any, is
+ *	freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ListObjReplace(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *listPtr,		/* List object whose elements to replace. */
+    int first,			/* Index of first element to replace. */
+    int count,			/* Number of elements to replace. */
+    int objc,			/* Number of objects to insert. */
+    Tcl_Obj *const objv[])	/* An array of objc pointers to Tcl objects to
+				 * insert. */
+{
+    List *listRepPtr;
+    register Tcl_Obj **elemPtrs;
+    int needGrow, numElems, numRequired, numAfterLast, start, i, j, isShared;
+
+    if (Tcl_IsShared(listPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_ListObjReplace");
+    }
+    if (listPtr->typePtr != &tclListType) {
+	if (listPtr->bytes == &tclEmptyString) {
+	    if (!objc) {
+		return TCL_OK;
+	    }
+	    Tcl_SetListObj(listPtr, objc, NULL);
+	} else {
+	    int result = SetListFromAny(interp, listPtr);
+
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+    }
+
+    /*
+     * Note that when count == 0 and objc == 0, this routine is logically a
+     * no-op, removing and adding no elements to the list. However, by flowing
+     * through this routine anyway, we get the important side effect that the
+     * resulting listPtr is a list in canoncial form. This is important.
+     * Resist any temptation to optimize this case.
+     */
+
+    listRepPtr = ListRepPtr(listPtr);
+    elemPtrs = &listRepPtr->elements;
+    numElems = listRepPtr->elemCount;
+
+    if (first < 0) {
+	first = 0;
+    }
+    if (first >= numElems) {
+	first = numElems;	/* So we'll insert after last element. */
+    }
+    if (count < 0) {
+	count = 0;
+    } else if (first > INT_MAX - count /* Handle integer overflow */
+	    || numElems < first+count) {
+
+	count = numElems - first;
+    }
+
+    if (objc > LIST_MAX - (numElems - count)) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "max length of a Tcl list (%d elements) exceeded",
+		    LIST_MAX));
+	}
+	return TCL_ERROR;
+    }
+    isShared = (listRepPtr->refCount > 1);
+    numRequired = numElems - count + objc; /* Known <= LIST_MAX */
+    needGrow = numRequired > listRepPtr->maxElemCount;
+
+    for (i = 0;  i < objc;  i++) {
+	Tcl_IncrRefCount(objv[i]);
+    }
+
+    if (needGrow && !isShared) {
+	/* Try to use realloc */
+	List *newPtr = NULL;
+	int attempt = 2 * numRequired;
+	if (attempt <= LIST_MAX) {
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr == NULL) {
+	    attempt = numRequired + 1 + TCL_MIN_ELEMENT_GROWTH;
+	    if (attempt > LIST_MAX) {
+		attempt = LIST_MAX;
+	    }
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr == NULL) {
+	    attempt = numRequired;
+	    newPtr = attemptckrealloc(listRepPtr, LIST_SIZE(attempt));
+	}
+	if (newPtr) {
+	    listRepPtr = newPtr;
+	    listPtr->internalRep.twoPtrValue.ptr1 = listRepPtr;
+	    elemPtrs = &listRepPtr->elements;
+	    listRepPtr->maxElemCount = attempt;
+	    needGrow = numRequired > listRepPtr->maxElemCount;
+	}
+    }
+    if (!needGrow && !isShared) {
+	int shift;
+
+	/*
+	 * Can use the current List struct. First "delete" count elements
+	 * starting at first.
+	 */
+
+	for (j = first;  j < first + count;  j++) {
+	    Tcl_Obj *victimPtr = elemPtrs[j];
+
+	    TclDecrRefCount(victimPtr);
+	}
+
+	/*
+	 * Shift the elements after the last one removed to their new
+	 * locations.
+	 */
+
+	start = first + count;
+	numAfterLast = numElems - start;
+	shift = objc - count;	/* numNewElems - numDeleted */
+	if ((numAfterLast > 0) && (shift != 0)) {
+	    Tcl_Obj **src = elemPtrs + start;
+
+	    memmove(src+shift, src, (size_t) numAfterLast * sizeof(Tcl_Obj*));
+	}
+    } else {
+	/*
+	 * Cannot use the current List struct; it is shared, too small, or
+	 * both. Allocate a new struct and insert elements into it.
+	 */
+
+	List *oldListRepPtr = listRepPtr;
+	Tcl_Obj **oldPtrs = elemPtrs;
+	int newMax;
+
+	if (needGrow){
+	    newMax = 2 * numRequired;
+	} else {
+	    newMax = listRepPtr->maxElemCount;
+	}
+
+	listRepPtr = AttemptNewList(NULL, newMax, NULL);
+	if (listRepPtr == NULL) {
+	    unsigned int limit = LIST_MAX - numRequired;
+	    unsigned int extra = numRequired - numElems
+		    + TCL_MIN_ELEMENT_GROWTH;
+	    int growth = (int) ((extra > limit) ? limit : extra);
+
+	    listRepPtr = AttemptNewList(NULL, numRequired + growth, NULL);
+	    if (listRepPtr == NULL) {
+		listRepPtr = AttemptNewList(interp, numRequired, NULL);
+		if (listRepPtr == NULL) {
+		    for (i = 0;  i < objc;  i++) {
+			/* See bug 3598580 */
+#if TCL_MAJOR_VERSION > 8
+			Tcl_DecrRefCount(objv[i]);
+#else
+			objv[i]->refCount--;
+#endif
+		    }
+		    return TCL_ERROR;
+		}
+	    }
+	}
+
+	listPtr->internalRep.twoPtrValue.ptr1 = listRepPtr;
+	listRepPtr->refCount++;
+
+	elemPtrs = &listRepPtr->elements;
+
+	if (isShared) {
+	    /*
+	     * The old struct will remain in place; need new refCounts for the
+	     * new List struct references. Copy over only the surviving
+	     * elements.
+	     */
+
+	    for (i=0; i < first; i++) {
+		elemPtrs[i] = oldPtrs[i];
+		Tcl_IncrRefCount(elemPtrs[i]);
+	    }
+	    for (i = first + count, j = first + objc;
+		    j < numRequired; i++, j++) {
+		elemPtrs[j] = oldPtrs[i];
+		Tcl_IncrRefCount(elemPtrs[j]);
+	    }
+
+	    oldListRepPtr->refCount--;
+	} else {
+	    /*
+	     * The old struct will be removed; use its inherited refCounts.
+	     */
+
+	    if (first > 0) {
+		memcpy(elemPtrs, oldPtrs, (size_t) first * sizeof(Tcl_Obj *));
+	    }
+
+	    /*
+	     * "Delete" count elements starting at first.
+	     */
+
+	    for (j = first;  j < first + count;  j++) {
+		Tcl_Obj *victimPtr = oldPtrs[j];
+
+		TclDecrRefCount(victimPtr);
+	    }
+
+	    /*
+	     * Copy the elements after the last one removed, shifted to their
+	     * new locations.
+	     */
+
+	    start = first + count;
+	    numAfterLast = numElems - start;
+	    if (numAfterLast > 0) {
+		memcpy(elemPtrs + first + objc, oldPtrs + start,
+			(size_t) numAfterLast * sizeof(Tcl_Obj *));
+	    }
+
+	    ckfree(oldListRepPtr);
+	}
+    }
+
+    /*
+     * Insert the new elements into elemPtrs before "first".
+     */
+
+    for (i=0,j=first ; i<objc ; i++,j++) {
+	elemPtrs[j] = objv[i];
+    }
+
+    /*
+     * Update the count of elements.
+     */
+
+    listRepPtr->elemCount = numRequired;
+
+    /*
+     * Invalidate and free any old string representation since it no longer
+     * reflects the list's internal representation.
+     */
+
+    TclInvalidateStringRep(listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLindexList --
+ *
+ *	This procedure handles the 'lindex' command when objc==3.
+ *
+ * Results:
+ *	Returns a pointer to the object extracted, or NULL if an error
+ *	occurred. The returned object already includes one reference count for
+ *	the pointer returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Notes:
+ *	This procedure is implemented entirely as a wrapper around
+ *	TclLindexFlat. All it does is reconfigure the argument format into the
+ *	form required by TclLindexFlat, while taking care to manage shimmering
+ *	in such a way that we tend to keep the most useful intreps and/or
+ *	avoid the most expensive conversions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclLindexList(
+    Tcl_Interp *interp,		/* Tcl interpreter. */
+    Tcl_Obj *listPtr,		/* List being unpacked. */
+    Tcl_Obj *argPtr)		/* Index or index list. */
+{
+
+    int index;			/* Index into the list. */
+    Tcl_Obj *indexListCopy;
+
+    /*
+     * Determine whether argPtr designates a list or a single index. We have
+     * to be careful about the order of the checks to avoid repeated
+     * shimmering; see TIP#22 and TIP#33 for the details.
+     */
+
+    if (argPtr->typePtr != &tclListType
+	    && TclGetIntForIndexM(NULL , argPtr, 0, &index) == TCL_OK) {
+	/*
+	 * argPtr designates a single index.
+	 */
+
+	return TclLindexFlat(interp, listPtr, 1, &argPtr);
+    }
+
+    /*
+     * Here we make a private copy of the index list argument to avoid any
+     * shimmering issues that might invalidate the indices array below while
+     * we are still using it. This is probably unnecessary. It does not appear
+     * that any damaging shimmering is possible, and no test has been devised
+     * to show any error when this private copy is not made. But it's cheap,
+     * and it offers some future-proofing insurance in case the TclLindexFlat
+     * implementation changes in some unexpected way, or some new form of
+     * trace or callback permits things to happen that the current
+     * implementation does not.
+     */
+
+    indexListCopy = TclListObjCopy(NULL, argPtr);
+    if (indexListCopy == NULL) {
+	/*
+	 * argPtr designates something that is neither an index nor a
+	 * well-formed list. Report the error via TclLindexFlat.
+	 */
+
+	return TclLindexFlat(interp, listPtr, 1, &argPtr);
+    }
+
+    if (indexListCopy->typePtr == &tclListType) {
+	List *listRepPtr = ListRepPtr(indexListCopy);
+
+	listPtr = TclLindexFlat(interp, listPtr, listRepPtr->elemCount,
+		&listRepPtr->elements);
+    } else {
+	int indexCount = -1;	/* Size of the array of list indices. */
+	Tcl_Obj **indices = NULL;
+				/* Array of list indices. */
+
+	Tcl_ListObjGetElements(NULL, indexListCopy, &indexCount, &indices);
+	listPtr = TclLindexFlat(interp, listPtr, indexCount, indices);
+    }
+    Tcl_DecrRefCount(indexListCopy);
+    return listPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLindexFlat --
+ *
+ *	This procedure is the core of the 'lindex' command, with all index
+ *	arguments presented as a flat list.
+ *
+ * Results:
+ *	Returns a pointer to the object extracted, or NULL if an error
+ *	occurred. The returned object already includes one reference count for
+ *	the pointer returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Notes:
+ *	The reference count of the returned object includes one reference
+ *	corresponding to the pointer returned. Thus, the calling code will
+ *	usually do something like:
+ *		Tcl_SetObjResult(interp, result);
+ *		Tcl_DecrRefCount(result);
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclLindexFlat(
+    Tcl_Interp *interp,		/* Tcl interpreter. */
+    Tcl_Obj *listPtr,		/* Tcl object representing the list. */
+    int indexCount,		/* Count of indices. */
+    Tcl_Obj *const indexArray[])/* Array of pointers to Tcl objects that
+				 * represent the indices in the list. */
+{
+    int i;
+
+    Tcl_IncrRefCount(listPtr);
+
+    for (i=0 ; i<indexCount && listPtr ; i++) {
+	int index, listLen = 0;
+	Tcl_Obj **elemPtrs = NULL, *sublistCopy;
+
+	/*
+	 * Here we make a private copy of the current sublist, so we avoid any
+	 * shimmering issues that might invalidate the elemPtr array below
+	 * while we are still using it. See test lindex-8.4.
+	 */
+
+	sublistCopy = TclListObjCopy(interp, listPtr);
+	Tcl_DecrRefCount(listPtr);
+	listPtr = NULL;
+
+	if (sublistCopy == NULL) {
+	    /*
+	     * The sublist is not a list at all => error.
+	     */
+
+	    break;
+	}
+	TclListObjGetElements(NULL, sublistCopy, &listLen, &elemPtrs);
+
+	if (TclGetIntForIndexM(interp, indexArray[i], /*endValue*/ listLen-1,
+		&index) == TCL_OK) {
+	    if (index<0 || index>=listLen) {
+		/*
+		 * Index is out of range. Break out of loop with empty result.
+		 * First check remaining indices for validity
+		 */
+
+		while (++i < indexCount) {
+		    if (TclGetIntForIndexM(interp, indexArray[i], -1, &index)
+			!= TCL_OK) {
+			Tcl_DecrRefCount(sublistCopy);
+			return NULL;
+		    }
+		}
+		listPtr = Tcl_NewObj();
+	    } else {
+		/*
+		 * Extract the pointer to the appropriate element.
+		 */
+
+		listPtr = elemPtrs[index];
+	    }
+	    Tcl_IncrRefCount(listPtr);
+	}
+	Tcl_DecrRefCount(sublistCopy);
+    }
+
+    return listPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLsetList --
+ *
+ *	Core of the 'lset' command when objc == 4. Objv[2] may be either a
+ *	scalar index or a list of indices.
+ *
+ * Results:
+ *	Returns the new value of the list variable, or NULL if there was an
+ *	error. The returned object includes one reference count for the
+ *	pointer returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ * Notes:
+ *	This procedure is implemented entirely as a wrapper around
+ *	TclLsetFlat. All it does is reconfigure the argument format into the
+ *	form required by TclLsetFlat, while taking care to manage shimmering
+ *	in such a way that we tend to keep the most useful intreps and/or
+ *	avoid the most expensive conversions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclLsetList(
+    Tcl_Interp *interp,		/* Tcl interpreter. */
+    Tcl_Obj *listPtr,		/* Pointer to the list being modified. */
+    Tcl_Obj *indexArgPtr,	/* Index or index-list arg to 'lset'. */
+    Tcl_Obj *valuePtr)		/* Value arg to 'lset'. */
+{
+    int indexCount = 0;		/* Number of indices in the index list. */
+    Tcl_Obj **indices = NULL;	/* Vector of indices in the index list. */
+    Tcl_Obj *retValuePtr;	/* Pointer to the list to be returned. */
+    int index;			/* Current index in the list - discarded. */
+    Tcl_Obj *indexListCopy;
+
+    /*
+     * Determine whether the index arg designates a list or a single index.
+     * We have to be careful about the order of the checks to avoid repeated
+     * shimmering; see TIP #22 and #23 for details.
+     */
+
+    if (indexArgPtr->typePtr != &tclListType
+	    && TclGetIntForIndexM(NULL, indexArgPtr, 0, &index) == TCL_OK) {
+	/*
+	 * indexArgPtr designates a single index.
+	 */
+
+	return TclLsetFlat(interp, listPtr, 1, &indexArgPtr, valuePtr);
+
+    }
+
+    indexListCopy = TclListObjCopy(NULL, indexArgPtr);
+    if (indexListCopy == NULL) {
+	/*
+	 * indexArgPtr designates something that is neither an index nor a
+	 * well formed list. Report the error via TclLsetFlat.
+	 */
+
+	return TclLsetFlat(interp, listPtr, 1, &indexArgPtr, valuePtr);
+    }
+    TclListObjGetElements(NULL, indexArgPtr, &indexCount, &indices);
+
+    /*
+     * Let TclLsetFlat handle the actual lset'ting.
+     */
+
+    retValuePtr = TclLsetFlat(interp, listPtr, indexCount, indices, valuePtr);
+
+    Tcl_DecrRefCount(indexListCopy);
+    return retValuePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLsetFlat --
+ *
+ *	Core engine of the 'lset' command.
+ *
+ * Results:
+ *	Returns the new value of the list variable, or NULL if an error
+ *	occurred. The returned object includes one reference count for the
+ *	pointer returned.
+ *
+ * Side effects:
+ *	On entry, the reference count of the variable value does not reflect
+ *	any references held on the stack. The first action of this function is
+ *	to determine whether the object is shared, and to duplicate it if it
+ *	is. The reference count of the duplicate is incremented. At this
+ *	point, the reference count will be 1 for either case, so that the
+ *	object will appear to be unshared.
+ *
+ *	If an error occurs, and the object has been duplicated, the reference
+ *	count on the duplicate is decremented so that it is now 0: this
+ *	dismisses any memory that was allocated by this function.
+ *
+ *	If no error occurs, the reference count of the original object is
+ *	incremented if the object has not been duplicated, and nothing is done
+ *	to a reference count of the duplicate. Now the reference count of an
+ *	unduplicated object is 2 (the returned pointer, plus the one stored in
+ *	the variable). The reference count of a duplicate object is 1,
+ *	reflecting that the returned pointer is the only active reference. The
+ *	caller is expected to store the returned value back in the variable
+ *	and decrement its reference count. (INST_STORE_* does exactly this.)
+ *
+ *	Surgery is performed on the unshared list value to produce the result.
+ *	TclLsetFlat maintains a linked list of Tcl_Obj's whose string
+ *	representations must be spoilt by threading via 'ptr2' of the
+ *	two-pointer internal representation. On entry to TclLsetFlat, the
+ *	values of 'ptr2' are immaterial; on exit, the 'ptr2' field of any
+ *	Tcl_Obj that has been modified is set to NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclLsetFlat(
+    Tcl_Interp *interp,		/* Tcl interpreter. */
+    Tcl_Obj *listPtr,		/* Pointer to the list being modified. */
+    int indexCount,		/* Number of index args. */
+    Tcl_Obj *const indexArray[],
+				/* Index args. */
+    Tcl_Obj *valuePtr)		/* Value arg to 'lset'. */
+{
+    int index, result, len;
+    Tcl_Obj *subListPtr, *retValuePtr, *chainPtr;
+
+    /*
+     * If there are no indices, simply return the new value.  (Without
+     * indices, [lset] is a synonym for [set].
+     */
+
+    if (indexCount == 0) {
+	Tcl_IncrRefCount(valuePtr);
+	return valuePtr;
+    }
+
+    /*
+     * If the list is shared, make a copy we can modify (copy-on-write).  We
+     * use Tcl_DuplicateObj() instead of TclListObjCopy() for a few reasons:
+     * 1) we have not yet confirmed listPtr is actually a list; 2) We make a
+     * verbatim copy of any existing string rep, and when we combine that with
+     * the delayed invalidation of string reps of modified Tcl_Obj's
+     * implemented below, the outcome is that any error condition that causes
+     * this routine to return NULL, will leave the string rep of listPtr and
+     * all elements to be unchanged.
+     */
+
+    subListPtr = Tcl_IsShared(listPtr) ? Tcl_DuplicateObj(listPtr) : listPtr;
+
+    /*
+     * Anchor the linked list of Tcl_Obj's whose string reps must be
+     * invalidated if the operation succeeds.
+     */
+
+    retValuePtr = subListPtr;
+    chainPtr = NULL;
+    result = TCL_OK;
+
+    /*
+     * Loop through all the index arguments, and for each one dive into the
+     * appropriate sublist.
+     */
+
+    do {
+	int elemCount;
+	Tcl_Obj *parentList, **elemPtrs;
+
+	/*
+	 * Check for the possible error conditions...
+	 */
+
+	if (TclListObjGetElements(interp, subListPtr, &elemCount, &elemPtrs)
+		!= TCL_OK) {
+	    /* ...the sublist we're indexing into isn't a list at all. */
+	    result = TCL_ERROR;
+	    break;
+	}
+
+	/*
+	 * WARNING: the macro TclGetIntForIndexM is not safe for
+	 * post-increments, avoid '*indexArray++' here.
+	 */
+
+	if (TclGetIntForIndexM(interp, *indexArray, elemCount - 1, &index)
+		!= TCL_OK)  {
+	    /* ...the index we're trying to use isn't an index at all. */
+	    result = TCL_ERROR;
+	    indexArray++;
+	    break;
+	}
+	indexArray++;
+
+	if (index < 0 || index > elemCount) {
+	    /* ...the index points outside the sublist. */
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp,
+			Tcl_NewStringObj("list index out of range", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSET",
+			"BADINDEX", NULL);
+	    }
+	    result = TCL_ERROR;
+	    break;
+	}
+
+	/*
+	 * No error conditions.  As long as we're not yet on the last index,
+	 * determine the next sublist for the next pass through the loop, and
+	 * take steps to make sure it is an unshared copy, as we intend to
+	 * modify it.
+	 */
+
+	if (--indexCount) {
+	    parentList = subListPtr;
+	    if (index == elemCount) {
+		subListPtr = Tcl_NewObj();
+	    } else {
+		subListPtr = elemPtrs[index];
+	    }
+	    if (Tcl_IsShared(subListPtr)) {
+		subListPtr = Tcl_DuplicateObj(subListPtr);
+	    }
+
+	    /*
+	     * Replace the original elemPtr[index] in parentList with a copy
+	     * we know to be unshared.  This call will also deal with the
+	     * situation where parentList shares its intrep with other
+	     * Tcl_Obj's.  Dealing with the shared intrep case can cause
+	     * subListPtr to become shared again, so detect that case and make
+	     * and store another copy.
+	     */
+
+	    if (index == elemCount) {
+		Tcl_ListObjAppendElement(NULL, parentList, subListPtr);
+	    } else {
+		TclListObjSetElement(NULL, parentList, index, subListPtr);
+	    }
+	    if (Tcl_IsShared(subListPtr)) {
+		subListPtr = Tcl_DuplicateObj(subListPtr);
+		TclListObjSetElement(NULL, parentList, index, subListPtr);
+	    }
+
+	    /*
+	     * The TclListObjSetElement() calls do not spoil the string rep of
+	     * parentList, and that's fine for now, since all we've done so
+	     * far is replace a list element with an unshared copy.  The list
+	     * value remains the same, so the string rep. is still valid, and
+	     * unchanged, which is good because if this whole routine returns
+	     * NULL, we'd like to leave no change to the value of the lset
+	     * variable.  Later on, when we set valuePtr in its proper place,
+	     * then all containing lists will have their values changed, and
+	     * will need their string reps spoiled.  We maintain a list of all
+	     * those Tcl_Obj's (via a little intrep surgery) so we can spoil
+	     * them at that time.
+	     */
+
+	    parentList->internalRep.twoPtrValue.ptr2 = chainPtr;
+	    chainPtr = parentList;
+	}
+    } while (indexCount > 0);
+
+    /*
+     * Either we've detected and error condition, and exited the loop with
+     * result == TCL_ERROR, or we've successfully reached the last index, and
+     * we're ready to store valuePtr.  In either case, we need to clean up our
+     * string spoiling list of Tcl_Obj's.
+     */
+
+    while (chainPtr) {
+	Tcl_Obj *objPtr = chainPtr;
+
+	if (result == TCL_OK) {
+	    /*
+	     * We're going to store valuePtr, so spoil string reps of all
+	     * containing lists.
+	     */
+
+	    TclInvalidateStringRep(objPtr);
+	}
+
+	/*
+	 * Clear away our intrep surgery mess.
+	 */
+
+	chainPtr = objPtr->internalRep.twoPtrValue.ptr2;
+	objPtr->internalRep.twoPtrValue.ptr2 = NULL;
+    }
+
+    if (result != TCL_OK) {
+	/*
+	 * Error return; message is already in interp. Clean up any excess
+	 * memory.
+	 */
+
+	if (retValuePtr != listPtr) {
+	    Tcl_DecrRefCount(retValuePtr);
+	}
+	return NULL;
+    }
+
+    /*
+     * Store valuePtr in proper sublist and return. The -1 is to avoid a
+     * compiler warning (not a problem because we checked that we have a
+     * proper list - or something convertible to one - above).
+     */
+
+    len = -1;
+    TclListObjLength(NULL, subListPtr, &len);
+    if (index == len) {
+	Tcl_ListObjAppendElement(NULL, subListPtr, valuePtr);
+    } else {
+	TclListObjSetElement(NULL, subListPtr, index, valuePtr);
+    }
+    TclInvalidateStringRep(subListPtr);
+    Tcl_IncrRefCount(retValuePtr);
+    return retValuePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclListObjSetElement --
+ *
+ *	Set a single element of a list to a specified value
+ *
+ * Results:
+ *	The return value is normally TCL_OK. If listPtr does not refer to a
+ *	list object and cannot be converted to one, TCL_ERROR is returned and
+ *	an error message will be left in the interpreter result if interp is
+ *	not NULL. Similarly, if index designates an element outside the range
+ *	[0..listLength-1], where listLength is the count of elements in the
+ *	list object designated by listPtr, TCL_ERROR is returned and an error
+ *	message is left in the interpreter result.
+ *
+ * Side effects:
+ *	Tcl_Panic if listPtr designates a shared object. Otherwise, attempts
+ *	to convert it to a list with a non-shared internal rep. Decrements the
+ *	ref count of the object at the specified index within the list,
+ *	replaces with the object designated by valuePtr, and increments the
+ *	ref count of the replacement object.
+ *
+ *	It is the caller's responsibility to invalidate the string
+ *	representation of the object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclListObjSetElement(
+    Tcl_Interp *interp,		/* Tcl interpreter; used for error reporting
+				 * if not NULL. */
+    Tcl_Obj *listPtr,		/* List object in which element should be
+				 * stored. */
+    int index,			/* Index of element to store. */
+    Tcl_Obj *valuePtr)		/* Tcl object to store in the designated list
+				 * element. */
+{
+    List *listRepPtr;		/* Internal representation of the list being
+				 * modified. */
+    Tcl_Obj **elemPtrs;		/* Pointers to elements of the list. */
+    int elemCount;		/* Number of elements in the list. */
+
+    /*
+     * Ensure that the listPtr parameter designates an unshared list.
+     */
+
+    if (Tcl_IsShared(listPtr)) {
+	Tcl_Panic("%s called with shared object", "TclListObjSetElement");
+    }
+    if (listPtr->typePtr != &tclListType) {
+	int result;
+
+	if (listPtr->bytes == &tclEmptyString) {
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp,
+			Tcl_NewStringObj("list index out of range", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSET",
+			"BADINDEX", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	result = SetListFromAny(interp, listPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+
+    listRepPtr = ListRepPtr(listPtr);
+    elemCount = listRepPtr->elemCount;
+
+    /*
+     * Ensure that the index is in bounds.
+     */
+
+    if (index<0 || index>=elemCount) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("list index out of range", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LSET", "BADINDEX",
+		    NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the internal rep is shared, replace it with an unshared copy.
+     */
+
+    if (listRepPtr->refCount > 1) {
+	Tcl_Obj **dst, **src = &listRepPtr->elements;
+	List *newPtr = AttemptNewList(NULL, listRepPtr->maxElemCount, NULL);
+
+	if (newPtr == NULL) {
+	    newPtr = AttemptNewList(interp, elemCount, NULL);
+	    if (newPtr == NULL) {
+		return TCL_ERROR;
+	    }
+	}
+	newPtr->refCount++;
+	newPtr->elemCount = elemCount;
+	newPtr->canonicalFlag = listRepPtr->canonicalFlag;
+
+	dst = &newPtr->elements;
+	while (elemCount--) {
+	    *dst = *src++;
+	    Tcl_IncrRefCount(*dst++);
+	}
+
+	listRepPtr->refCount--;
+
+	listPtr->internalRep.twoPtrValue.ptr1 = listRepPtr = newPtr;
+    }
+    elemPtrs = &listRepPtr->elements;
+
+    /*
+     * Add a reference to the new list element.
+     */
+
+    Tcl_IncrRefCount(valuePtr);
+
+    /*
+     * Remove a reference from the old list element.
+     */
+
+    Tcl_DecrRefCount(elemPtrs[index]);
+
+    /*
+     * Stash the new object in the list.
+     */
+
+    elemPtrs[index] = valuePtr;
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeListInternalRep --
+ *
+ *	Deallocate the storage associated with a list object's internal
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees listPtr's List* internal representation and sets listPtr's
+ *	internalRep.twoPtrValue.ptr1 to NULL. Decrements the ref counts of all
+ *	element objects, which may free them.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeListInternalRep(
+    Tcl_Obj *listPtr)		/* List object with internal rep to free. */
+{
+    List *listRepPtr = ListRepPtr(listPtr);
+
+    if (listRepPtr->refCount-- <= 1) {
+	Tcl_Obj **elemPtrs = &listRepPtr->elements;
+	int i, numElems = listRepPtr->elemCount;
+
+	for (i = 0;  i < numElems;  i++) {
+	    Tcl_DecrRefCount(elemPtrs[i]);
+	}
+	ckfree(listRepPtr);
+    }
+
+    listPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupListInternalRep --
+ *
+ *	Initialize the internal representation of a list Tcl_Obj to share the
+ *	internal representation of an existing list object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The reference count of the List internal rep is incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupListInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    Tcl_Obj *copyPtr)		/* Object with internal rep to set. */
+{
+    List *listRepPtr = ListRepPtr(srcPtr);
+
+    ListSetIntRep(copyPtr, listRepPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetListFromAny --
+ *
+ *	Attempt to generate a list internal form for the Tcl object "objPtr".
+ *
+ * Results:
+ *	The return value is TCL_OK or TCL_ERROR. If an error occurs during
+ *	conversion, an error message is left in the interpreter's result
+ *	unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If no error occurs, a list is stored as "objPtr"s internal
+ *	representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetListFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr)		/* The object to convert. */
+{
+    List *listRepPtr;
+    Tcl_Obj **elemPtrs;
+
+    /*
+     * Dictionaries are a special case; they have a string representation such
+     * that *all* valid dictionaries are valid lists. Hence we can convert
+     * more directly. Only do this when there's no existing string rep; if
+     * there is, it is the string rep that's authoritative (because it could
+     * describe duplicate keys).
+     */
+
+    if (objPtr->typePtr == &tclDictType && !objPtr->bytes) {
+	Tcl_Obj *keyPtr, *valuePtr;
+	Tcl_DictSearch search;
+	int done, size;
+
+	/*
+	 * Create the new list representation. Note that we do not need to do
+	 * anything with the string representation as the transformation (and
+	 * the reverse back to a dictionary) are both order-preserving. Also
+	 * note that since we know we've got a valid dictionary (by
+	 * representation) we also know that fetching the size of the
+	 * dictionary or iterating over it will not fail.
+	 */
+
+	Tcl_DictObjSize(NULL, objPtr, &size);
+	listRepPtr = AttemptNewList(interp, size > 0 ? 2*size : 1, NULL);
+	if (!listRepPtr) {
+	    return TCL_ERROR;
+	}
+	listRepPtr->elemCount = 2 * size;
+
+	/*
+	 * Populate the list representation.
+	 */
+
+	elemPtrs = &listRepPtr->elements;
+	Tcl_DictObjFirst(NULL, objPtr, &search, &keyPtr, &valuePtr, &done);
+	while (!done) {
+	    *elemPtrs++ = keyPtr;
+	    *elemPtrs++ = valuePtr;
+	    Tcl_IncrRefCount(keyPtr);
+	    Tcl_IncrRefCount(valuePtr);
+	    Tcl_DictObjNext(&search, &keyPtr, &valuePtr, &done);
+	}
+    } else {
+	int estCount, length;
+	const char *limit, *nextElem = TclGetStringFromObj(objPtr, &length);
+
+	/*
+	 * Allocate enough space to hold a (Tcl_Obj *) for each
+	 * (possible) list element.
+	 */
+
+	estCount = TclMaxListLength(nextElem, length, &limit);
+	estCount += (estCount == 0);	/* Smallest list struct holds 1
+					 * element. */
+	listRepPtr = AttemptNewList(interp, estCount, NULL);
+	if (listRepPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	elemPtrs = &listRepPtr->elements;
+
+	/*
+	 * Each iteration, parse and store a list element.
+	 */
+
+	while (nextElem < limit) {
+	    const char *elemStart;
+	    int elemSize, literal;
+
+	    if (TCL_OK != TclFindElement(interp, nextElem, limit - nextElem,
+		    &elemStart, &nextElem, &elemSize, &literal)) {
+		while (--elemPtrs >= &listRepPtr->elements) {
+		    Tcl_DecrRefCount(*elemPtrs);
+		}
+		ckfree(listRepPtr);
+		return TCL_ERROR;
+	    }
+	    if (elemStart == limit) {
+		break;
+	    }
+
+	    /* TODO: replace panic with error on alloc failure? */
+	    if (literal) {
+		TclNewStringObj(*elemPtrs, elemStart, elemSize);
+	    } else {
+		TclNewObj(*elemPtrs);
+		(*elemPtrs)->bytes = ckalloc((unsigned) elemSize + 1);
+		(*elemPtrs)->length = TclCopyAndCollapse(elemSize, elemStart,
+			(*elemPtrs)->bytes);
+	    }
+
+	    Tcl_IncrRefCount(*elemPtrs++);/* Since list now holds ref to it. */
+	}
+
+ 	listRepPtr->elemCount = elemPtrs - &listRepPtr->elements;
+    }
+
+    /*
+     * Free the old internalRep before setting the new one. We do this as late
+     * as possible to allow the conversion code, in particular
+     * Tcl_GetStringFromObj, to use that old internalRep.
+     */
+
+    TclFreeIntRep(objPtr);
+    ListSetIntRep(objPtr, listRepPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfList --
+ *
+ *	Update the string representation for a list object. Note: This
+ *	function does not invalidate an existing old string rep so storage
+ *	will be lost if this has not already been done.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	list-to-string conversion. This string will be empty if the list has
+ *	no elements. The list internal representation should not be NULL and
+ *	we assume it is not NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfList(
+    Tcl_Obj *listPtr)		/* List object with string rep to update. */
+{
+#   define LOCAL_SIZE 20
+    int localFlags[LOCAL_SIZE], *flagPtr = NULL;
+    List *listRepPtr = ListRepPtr(listPtr);
+    int numElems = listRepPtr->elemCount;
+    int i, length, bytesNeeded = 0;
+    const char *elem;
+    char *dst;
+    Tcl_Obj **elemPtrs;
+
+    /*
+     * Mark the list as being canonical; although it will now have a string
+     * rep, it is one we derived through proper "canonical" quoting and so
+     * it's known to be free from nasties relating to [concat] and [eval].
+     */
+
+    listRepPtr->canonicalFlag = 1;
+
+    /*
+     * Handle empty list case first, so rest of the routine is simpler.
+     */
+
+    if (numElems == 0) {
+	listPtr->bytes = &tclEmptyString;
+	listPtr->length = 0;
+	return;
+    }
+
+    /*
+     * Pass 1: estimate space, gather flags.
+     */
+
+    if (numElems <= LOCAL_SIZE) {
+	flagPtr = localFlags;
+    } else {
+	/*
+	 * We know numElems <= LIST_MAX, so this is safe.
+	 */
+
+	flagPtr = ckalloc(numElems * sizeof(int));
+    }
+    elemPtrs = &listRepPtr->elements;
+    for (i = 0; i < numElems; i++) {
+	flagPtr[i] = (i ? TCL_DONT_QUOTE_HASH : 0);
+	elem = TclGetStringFromObj(elemPtrs[i], &length);
+	bytesNeeded += TclScanElement(elem, length, flagPtr+i);
+	if (bytesNeeded < 0) {
+	    Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+	}
+    }
+    if (bytesNeeded > INT_MAX - numElems + 1) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+    bytesNeeded += numElems;
+
+    /*
+     * Pass 2: copy into string rep buffer.
+     */
+
+    listPtr->length = bytesNeeded - 1;
+    listPtr->bytes = ckalloc(bytesNeeded);
+    dst = listPtr->bytes;
+    for (i = 0; i < numElems; i++) {
+	flagPtr[i] |= (i ? TCL_DONT_QUOTE_HASH : 0);
+	elem = TclGetStringFromObj(elemPtrs[i], &length);
+	dst += TclConvertElement(elem, length, dst, flagPtr[i]);
+	*dst++ = ' ';
+    }
+    listPtr->bytes[listPtr->length] = '\0';
+
+    if (flagPtr != localFlags) {
+	ckfree(flagPtr);
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclLiteral.c b/generic/tclLiteral.c
new file mode 100644
index 0000000..7acc9ad
--- /dev/null
+++ b/generic/tclLiteral.c
@@ -0,0 +1,1242 @@
+/*
+ * tclLiteral.c --
+ *
+ *	Implementation of the global and ByteCode-local literal tables used to
+ *	manage the Tcl objects created for literal values during compilation
+ *	of Tcl scripts. This implementation borrows heavily from the more
+ *	general hashtable implementation of Tcl hash tables that appears in
+ *	tclHash.c.
+ *
+ * Copyright (c) 1997-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2004 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+
+/*
+ * When there are this many entries per bucket, on average, rebuild a
+ * literal's hash table to make it larger.
+ */
+
+#define REBUILD_MULTIPLIER	3
+
+/*
+ * Function prototypes for static functions in this file:
+ */
+
+static int		AddLocalLiteralEntry(CompileEnv *envPtr,
+			    Tcl_Obj *objPtr, int localHash);
+static void		ExpandLocalLiteralArray(CompileEnv *envPtr);
+static unsigned		HashString(const char *string, int length);
+#ifdef TCL_COMPILE_DEBUG
+static LiteralEntry *	LookupLiteralEntry(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+#endif
+static void		RebuildLiteralTable(LiteralTable *tablePtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitLiteralTable --
+ *
+ *	This function is called to initialize the fields of a literal table
+ *	structure for either an interpreter or a compilation's CompileEnv
+ *	structure.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The literal table is made ready for use.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitLiteralTable(
+    register LiteralTable *tablePtr)
+				/* Pointer to table structure, which is
+				 * supplied by the caller. */
+{
+#if (TCL_SMALL_HASH_TABLE != 4)
+    Tcl_Panic("%s: TCL_SMALL_HASH_TABLE is %d, not 4", "TclInitLiteralTable",
+	    TCL_SMALL_HASH_TABLE);
+#endif
+
+    tablePtr->buckets = tablePtr->staticBuckets;
+    tablePtr->staticBuckets[0] = tablePtr->staticBuckets[1] = 0;
+    tablePtr->staticBuckets[2] = tablePtr->staticBuckets[3] = 0;
+    tablePtr->numBuckets = TCL_SMALL_HASH_TABLE;
+    tablePtr->numEntries = 0;
+    tablePtr->rebuildSize = TCL_SMALL_HASH_TABLE * REBUILD_MULTIPLIER;
+    tablePtr->mask = 3;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteLiteralTable --
+ *
+ *	This function frees up everything associated with a literal table
+ *	except for the table's structure itself. It is called when the
+ *	interpreter is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Each literal in the table is released: i.e., its reference count in
+ *	the global literal table is decremented and, if it becomes zero, the
+ *	literal is freed. In addition, the table's bucket array is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDeleteLiteralTable(
+    Tcl_Interp *interp,		/* Interpreter containing shared literals
+				 * referenced by the table to delete. */
+    LiteralTable *tablePtr)	/* Points to the literal table to delete. */
+{
+    LiteralEntry *entryPtr, *nextPtr;
+    Tcl_Obj *objPtr;
+    int i;
+
+    /*
+     * Release remaining literals in the table. Note that releasing a literal
+     * might release other literals, modifying the table, so we restart the
+     * search from the bucket chain we last found an entry.
+     */
+
+#ifdef TCL_COMPILE_DEBUG
+    TclVerifyGlobalLiteralTable((Interp *) interp);
+#endif /*TCL_COMPILE_DEBUG*/
+
+    /*
+     * We used to call TclReleaseLiteral for each literal in the table, which
+     * is rather inefficient as it causes one lookup-by-hash for each
+     * reference to the literal. We now rely at interp-deletion on each
+     * bytecode object to release its references to the literal Tcl_Obj
+     * without requiring that it updates the global table itself, and deal
+     * here only with the table.
+     */
+
+    for (i=0 ; i<tablePtr->numBuckets ; i++) {
+	entryPtr = tablePtr->buckets[i];
+	while (entryPtr != NULL) {
+	    objPtr = entryPtr->objPtr;
+	    TclDecrRefCount(objPtr);
+	    nextPtr = entryPtr->nextPtr;
+	    ckfree(entryPtr);
+	    entryPtr = nextPtr;
+	}
+    }
+
+    /*
+     * Free up the table's bucket array if it was dynamically allocated.
+     */
+
+    if (tablePtr->buckets != tablePtr->staticBuckets) {
+	ckfree(tablePtr->buckets);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateLiteral --
+ *
+ *	Find, or if necessary create, an object in the interpreter's literal
+ *	table that has a string representation matching the argument
+ *	string. If nsPtr!=NULL then only literals stored for the namespace are
+ *	considered.
+ *
+ * Results:
+ *	The literal object. If it was created in this call *newPtr is set to
+ *	1, else 0. NULL is returned if newPtr==NULL and no literal is found.
+ *
+ * Side effects:
+ *	Increments the ref count of the global LiteralEntry since the caller
+ *	now holds a reference. If LITERAL_ON_HEAP is set in flags, this
+ *	function is given ownership of the string: if an object is created
+ *	then its string representation is set directly from string, otherwise
+ *	the string is freed. Typically, a caller sets LITERAL_ON_HEAP if
+ *	"string" is an already heap-allocated buffer holding the result of
+ *	backslash substitutions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclCreateLiteral(
+    Interp *iPtr,
+    const char *bytes,	/* The start of the string. Note that this is
+				 * not a NUL-terminated string. */
+    int length,			/* Number of bytes in the string. */
+    unsigned hash,		/* The string's hash. If -1, it will be
+				 * computed here. */
+    int *newPtr,
+    Namespace *nsPtr,
+    int flags,
+    LiteralEntry **globalPtrPtr)
+{
+    LiteralTable *globalTablePtr = &iPtr->literalTable;
+    LiteralEntry *globalPtr;
+    int globalHash;
+    Tcl_Obj *objPtr;
+
+    /*
+     * Is it in the interpreter's global literal table?
+     */
+
+    if (hash == (unsigned) -1) {
+	hash = HashString(bytes, length);
+    }
+    globalHash = (hash & globalTablePtr->mask);
+    for (globalPtr=globalTablePtr->buckets[globalHash] ; globalPtr!=NULL;
+	    globalPtr = globalPtr->nextPtr) {
+	objPtr = globalPtr->objPtr;
+	if ((globalPtr->nsPtr == nsPtr)
+		&& (objPtr->length == length) && ((length == 0)
+		|| ((objPtr->bytes[0] == bytes[0])
+		&& (memcmp(objPtr->bytes, bytes, (unsigned) length) == 0)))) {
+	    /*
+	     * A literal was found: return it
+	     */
+
+	    if (newPtr) {
+		*newPtr = 0;
+	    }
+	    if (globalPtrPtr) {
+		*globalPtrPtr = globalPtr;
+	    }
+	    if ((flags & LITERAL_ON_HEAP)) {
+		ckfree(bytes);
+	    }
+	    globalPtr->refCount++;
+	    return objPtr;
+	}
+    }
+    if (!newPtr) {
+	if ((flags & LITERAL_ON_HEAP)) {
+	    ckfree(bytes);
+	}
+	return NULL;
+    }
+
+    /*
+     * The literal is new to the interpreter.
+     */
+
+    TclNewObj(objPtr);
+    if ((flags & LITERAL_ON_HEAP)) {
+	objPtr->bytes = (char *) bytes;
+	objPtr->length = length;
+    } else {
+	TclInitStringRep(objPtr, bytes, length);
+    }
+
+    /* Should the new literal be shared globally? */
+
+    if ((flags & LITERAL_UNSHARED)) {
+	/*
+	 * No, do *not* add it the global literal table
+	 * Make clear, that no global value is returned
+	 */
+	if (globalPtrPtr != NULL) {
+	    *globalPtrPtr = NULL;
+	}
+	return objPtr;
+    }
+
+    /*
+     * Yes, add it to the global literal table.
+     */
+#ifdef TCL_COMPILE_DEBUG
+    if (LookupLiteralEntry((Tcl_Interp *) iPtr, objPtr) != NULL) {
+	Tcl_Panic("%s: literal \"%.*s\" found globally but shouldn't be",
+		"TclRegisterLiteral", (length>60? 60 : length), bytes);
+    }
+#endif
+
+    globalPtr = ckalloc(sizeof(LiteralEntry));
+    globalPtr->objPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);
+    globalPtr->refCount = 1;
+    globalPtr->nsPtr = nsPtr;
+    globalPtr->nextPtr = globalTablePtr->buckets[globalHash];
+    globalTablePtr->buckets[globalHash] = globalPtr;
+    globalTablePtr->numEntries++;
+
+    /*
+     * If the global literal table has exceeded a decent size, rebuild it with
+     * more buckets.
+     */
+
+    if (globalTablePtr->numEntries >= globalTablePtr->rebuildSize) {
+	RebuildLiteralTable(globalTablePtr);
+    }
+
+#ifdef TCL_COMPILE_DEBUG
+    TclVerifyGlobalLiteralTable(iPtr);
+    {
+	LiteralEntry *entryPtr;
+	int found, i;
+
+	found = 0;
+	for (i=0 ; i<globalTablePtr->numBuckets ; i++) {
+	    for (entryPtr=globalTablePtr->buckets[i]; entryPtr!=NULL ;
+		    entryPtr=entryPtr->nextPtr) {
+		if ((entryPtr == globalPtr) && (entryPtr->objPtr == objPtr)) {
+		    found = 1;
+		}
+	    }
+	}
+	if (!found) {
+	    Tcl_Panic("%s: literal \"%.*s\" wasn't global",
+		    "TclRegisterLiteral", (length>60? 60 : length), bytes);
+	}
+    }
+#endif /*TCL_COMPILE_DEBUG*/
+
+#ifdef TCL_COMPILE_STATS
+    iPtr->stats.numLiteralsCreated++;
+    iPtr->stats.totalLitStringBytes += (double) (length + 1);
+    iPtr->stats.currentLitStringBytes += (double) (length + 1);
+    iPtr->stats.literalCount[TclLog2(length)]++;
+#endif /*TCL_COMPILE_STATS*/
+
+    if (globalPtrPtr) {
+	*globalPtrPtr = globalPtr;
+    }
+    *newPtr = 1;
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFetchLiteral --
+ *
+ *	Fetch from a CompileEnv the literal value identified by an index
+ *	value, as returned by a prior call to TclRegisterLiteral().
+ *
+ * Results:
+ *	The literal value, or NULL if the index is out of range.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclFetchLiteral(
+    CompileEnv *envPtr,		/* Points to the CompileEnv from which to
+				 * fetch the registered literal value. */
+    unsigned int index)		/* Index of the desired literal, as returned
+				 * by prior call to TclRegisterLiteral() */
+{
+    if (index >= (unsigned int) envPtr->literalArrayNext) {
+	return NULL;
+    }
+    return envPtr->literalArrayPtr[index].objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRegisterLiteral --
+ *
+ *	Find, or if necessary create, an object in a CompileEnv literal array
+ *	that has a string representation matching the argument string.
+ *
+ * Results:
+ *	The index in the CompileEnv's literal array that references a shared
+ *	literal matching the string. The object is created if necessary.
+ *
+ * Side effects:
+ *	To maximize sharing, we look up the string in the interpreter's global
+ *	literal table. If not found, we create a new shared literal in the
+ *	global table. We then add a reference to the shared literal in the
+ *	CompileEnv's literal array.
+ *
+ *	If LITERAL_ON_HEAP is set in flags, this function is given ownership
+ *	of the string: if an object is created then its string representation
+ *	is set directly from string, otherwise the string is freed. Typically,
+ *	a caller sets LITERAL_ON_HEAP if "string" is an already heap-allocated
+ *	buffer holding the result of backslash substitutions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclRegisterLiteral(
+    void *ePtr,		/* Points to the CompileEnv in whose object
+				 * array an object is found or created. */
+    register const char *bytes,	/* Points to string for which to find or
+				 * create an object in CompileEnv's object
+				 * array. */
+    int length,			/* Number of bytes in the string. If < 0, the
+				 * string consists of all bytes up to the
+				 * first null character. */
+    int flags)			/* If LITERAL_ON_HEAP then the caller already
+				 * malloc'd bytes and ownership is passed to
+				 * this function. If LITERAL_CMD_NAME then
+				 * the literal should not be shared accross
+				 * namespaces. */
+{
+    CompileEnv *envPtr = ePtr;
+    Interp *iPtr = envPtr->iPtr;
+    LiteralTable *localTablePtr = &envPtr->localLitTable;
+    LiteralEntry *globalPtr, *localPtr;
+    Tcl_Obj *objPtr;
+    unsigned hash;
+    int localHash, objIndex, new;
+    Namespace *nsPtr;
+
+    if (length < 0) {
+	length = (bytes ? strlen(bytes) : 0);
+    }
+    hash = HashString(bytes, length);
+
+    /*
+     * Is the literal already in the CompileEnv's local literal array? If so,
+     * just return its index.
+     */
+
+    localHash = (hash & localTablePtr->mask);
+    for (localPtr=localTablePtr->buckets[localHash] ; localPtr!=NULL;
+	    localPtr = localPtr->nextPtr) {
+	objPtr = localPtr->objPtr;
+	if ((objPtr->length == length) && ((length == 0)
+		|| ((objPtr->bytes[0] == bytes[0])
+		&& (memcmp(objPtr->bytes, bytes, (unsigned) length) == 0)))) {
+	    if ((flags & LITERAL_ON_HEAP)) {
+		ckfree(bytes);
+	    }
+	    objIndex = (localPtr - envPtr->literalArrayPtr);
+#ifdef TCL_COMPILE_DEBUG
+	    TclVerifyLocalLiteralTable(envPtr);
+#endif /*TCL_COMPILE_DEBUG*/
+
+	    return objIndex;
+	}
+    }
+
+    /*
+     * The literal is new to this CompileEnv. If it is a command name, avoid
+     * sharing it accross namespaces, and try not to share it with non-cmd
+     * literals. Note that FQ command names can be shared, so that we register
+     * the namespace as the interp's global NS.
+     */
+
+    if ((flags & LITERAL_CMD_NAME)) {
+	if ((length >= 2) && (bytes[0] == ':') && (bytes[1] == ':')) {
+	    nsPtr = iPtr->globalNsPtr;
+	} else {
+	    nsPtr = iPtr->varFramePtr->nsPtr;
+	}
+    } else {
+	nsPtr = NULL;
+    }
+
+    /*
+     * Is it in the interpreter's global literal table? If not, create it.
+     */
+
+    globalPtr = NULL;
+    objPtr = TclCreateLiteral(iPtr, bytes, length, hash, &new, nsPtr, flags,
+	    &globalPtr);
+    objIndex = AddLocalLiteralEntry(envPtr, objPtr, localHash);
+
+#ifdef TCL_COMPILE_DEBUG
+    if (globalPtr != NULL && globalPtr->refCount < 1) {
+	Tcl_Panic("%s: global literal \"%.*s\" had bad refCount %d",
+		"TclRegisterLiteral", (length>60? 60 : length), bytes,
+		globalPtr->refCount);
+    }
+    TclVerifyLocalLiteralTable(envPtr);
+#endif /*TCL_COMPILE_DEBUG*/
+    return objIndex;
+}
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ *----------------------------------------------------------------------
+ *
+ * LookupLiteralEntry --
+ *
+ *	Finds the LiteralEntry that corresponds to a literal Tcl object
+ *	holding a literal.
+ *
+ * Results:
+ *	Returns the matching LiteralEntry if found, otherwise NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static LiteralEntry *
+LookupLiteralEntry(
+    Tcl_Interp *interp,		/* Interpreter for which objPtr was created to
+				 * hold a literal. */
+    register Tcl_Obj *objPtr)	/* Points to a Tcl object holding a literal
+				 * that was previously created by a call to
+				 * TclRegisterLiteral. */
+{
+    Interp *iPtr = (Interp *) interp;
+    LiteralTable *globalTablePtr = &iPtr->literalTable;
+    register LiteralEntry *entryPtr;
+    const char *bytes;
+    int length, globalHash;
+
+    bytes = TclGetStringFromObj(objPtr, &length);
+    globalHash = (HashString(bytes, length) & globalTablePtr->mask);
+    for (entryPtr=globalTablePtr->buckets[globalHash] ; entryPtr!=NULL;
+	    entryPtr=entryPtr->nextPtr) {
+	if (entryPtr->objPtr == objPtr) {
+	    return entryPtr;
+	}
+    }
+    return NULL;
+}
+
+#endif
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclHideLiteral --
+ *
+ *	Remove a literal entry from the literal hash tables, leaving it in the
+ *	literal array so existing references continue to function. This makes
+ *	it possible to turn a shared literal into a private literal that
+ *	cannot be shared.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes the literal from the local hash table and decrements the
+ *	global hash entry's reference count.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclHideLiteral(
+    Tcl_Interp *interp,		/* Interpreter for which objPtr was created to
+				 * hold a literal. */
+    register CompileEnv *envPtr,/* Points to CompileEnv whose literal array
+				 * contains the entry being hidden. */
+    int index)			/* The index of the entry in the literal
+				 * array. */
+{
+    LiteralEntry **nextPtrPtr, *entryPtr, *lPtr;
+    LiteralTable *localTablePtr = &envPtr->localLitTable;
+    int localHash, length;
+    const char *bytes;
+    Tcl_Obj *newObjPtr;
+
+    lPtr = &envPtr->literalArrayPtr[index];
+
+    /*
+     * To avoid unwanted sharing we need to copy the object and remove it from
+     * the local and global literal tables. It still has a slot in the literal
+     * array so it can be referred to by byte codes, but it will not be
+     * matched by literal searches.
+     */
+
+    newObjPtr = Tcl_DuplicateObj(lPtr->objPtr);
+    Tcl_IncrRefCount(newObjPtr);
+    TclReleaseLiteral(interp, lPtr->objPtr);
+    lPtr->objPtr = newObjPtr;
+
+    bytes = TclGetStringFromObj(newObjPtr, &length);
+    localHash = (HashString(bytes, length) & localTablePtr->mask);
+    nextPtrPtr = &localTablePtr->buckets[localHash];
+
+    for (entryPtr=*nextPtrPtr ; entryPtr!=NULL ; entryPtr=*nextPtrPtr) {
+	if (entryPtr == lPtr) {
+	    *nextPtrPtr = lPtr->nextPtr;
+	    lPtr->nextPtr = NULL;
+	    localTablePtr->numEntries--;
+	    break;
+	}
+	nextPtrPtr = &entryPtr->nextPtr;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclAddLiteralObj --
+ *
+ *	Add a single literal object to the literal array. This function does
+ *	not add the literal to the local or global literal tables. The caller
+ *	is expected to add the entry to whatever tables are appropriate.
+ *
+ * Results:
+ *	The index in the CompileEnv's literal array that references the
+ *	literal. Stores the pointer to the new literal entry in the location
+ *	referenced by the localPtrPtr argument.
+ *
+ * Side effects:
+ *	Expands the literal array if necessary. Increments the refcount on the
+ *	literal object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclAddLiteralObj(
+    register CompileEnv *envPtr,/* Points to CompileEnv in whose literal array
+				 * the object is to be inserted. */
+    Tcl_Obj *objPtr,		/* The object to insert into the array. */
+    LiteralEntry **litPtrPtr)	/* The location where the pointer to the new
+				 * literal entry should be stored. May be
+				 * NULL. */
+{
+    register LiteralEntry *lPtr;
+    int objIndex;
+
+    if (envPtr->literalArrayNext >= envPtr->literalArrayEnd) {
+	ExpandLocalLiteralArray(envPtr);
+    }
+    objIndex = envPtr->literalArrayNext;
+    envPtr->literalArrayNext++;
+
+    lPtr = &envPtr->literalArrayPtr[objIndex];
+    lPtr->objPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);
+    lPtr->refCount = -1;	/* i.e., unused */
+    lPtr->nextPtr = NULL;
+
+    if (litPtrPtr) {
+	*litPtrPtr = lPtr;
+    }
+
+    return objIndex;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AddLocalLiteralEntry --
+ *
+ *	Insert a new literal into a CompileEnv's local literal array.
+ *
+ * Results:
+ *	The index in the CompileEnv's literal array that references the
+ *	literal.
+ *
+ * Side effects:
+ *	Expands the literal array if necessary. May rebuild the hash bucket
+ *	array of the CompileEnv's literal array if it becomes too large.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AddLocalLiteralEntry(
+    register CompileEnv *envPtr,/* Points to CompileEnv in whose literal array
+				 * the object is to be inserted. */
+    Tcl_Obj *objPtr,		/* The literal to add to the CompileEnv. */
+    int localHash)		/* Hash value for the literal's string. */
+{
+    register LiteralTable *localTablePtr = &envPtr->localLitTable;
+    LiteralEntry *localPtr;
+    int objIndex;
+
+    objIndex = TclAddLiteralObj(envPtr, objPtr, &localPtr);
+
+    /*
+     * Add the literal to the local table.
+     */
+
+    localPtr->nextPtr = localTablePtr->buckets[localHash];
+    localTablePtr->buckets[localHash] = localPtr;
+    localTablePtr->numEntries++;
+
+    /*
+     * If the CompileEnv's local literal table has exceeded a decent size,
+     * rebuild it with more buckets.
+     */
+
+    if (localTablePtr->numEntries >= localTablePtr->rebuildSize) {
+	RebuildLiteralTable(localTablePtr);
+    }
+
+#ifdef TCL_COMPILE_DEBUG
+    TclVerifyLocalLiteralTable(envPtr);
+    {
+	char *bytes;
+	int length, found, i;
+
+	found = 0;
+	for (i=0 ; i<localTablePtr->numBuckets ; i++) {
+	    for (localPtr=localTablePtr->buckets[i] ; localPtr!=NULL ;
+		    localPtr=localPtr->nextPtr) {
+		if (localPtr->objPtr == objPtr) {
+		    found = 1;
+		}
+	    }
+	}
+
+	if (!found) {
+	    bytes = TclGetStringFromObj(objPtr, &length);
+	    Tcl_Panic("%s: literal \"%.*s\" wasn't found locally",
+		    "AddLocalLiteralEntry", (length>60? 60 : length), bytes);
+	}
+    }
+#endif /*TCL_COMPILE_DEBUG*/
+
+    return objIndex;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExpandLocalLiteralArray --
+ *
+ *	Function that uses malloc to allocate more storage for a CompileEnv's
+ *	local literal array.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The literal array in *envPtr is reallocated to a new array of double
+ *	the size, and if envPtr->mallocedLiteralArray is non-zero the old
+ *	array is freed. Entries are copied from the old array to the new one.
+ *	The local literal table is updated to refer to the new entries.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ExpandLocalLiteralArray(
+    register CompileEnv *envPtr)/* Points to the CompileEnv whose object array
+				 * must be enlarged. */
+{
+    /*
+     * The current allocated local literal entries are stored between elements
+     * 0 and (envPtr->literalArrayNext - 1) [inclusive].
+     */
+
+    LiteralTable *localTablePtr = &envPtr->localLitTable;
+    int currElems = envPtr->literalArrayNext;
+    size_t currBytes = (currElems * sizeof(LiteralEntry));
+    LiteralEntry *currArrayPtr = envPtr->literalArrayPtr;
+    LiteralEntry *newArrayPtr;
+    int i;
+    unsigned int newSize = (currBytes <= UINT_MAX / 2) ? 2*currBytes : UINT_MAX;
+
+    if (currBytes == newSize) {
+	Tcl_Panic("max size of Tcl literal array (%d literals) exceeded",
+		currElems);
+    }
+
+    if (envPtr->mallocedLiteralArray) {
+	newArrayPtr = ckrealloc(currArrayPtr, newSize);
+    } else {
+	/*
+	 * envPtr->literalArrayPtr isn't a ckalloc'd pointer, so we must
+	 * code a ckrealloc equivalent for ourselves.
+	 */
+
+	newArrayPtr = ckalloc(newSize);
+	memcpy(newArrayPtr, currArrayPtr, currBytes);
+	envPtr->mallocedLiteralArray = 1;
+    }
+
+    /*
+     * Update the local literal table's bucket array.
+     */
+
+    if (currArrayPtr != newArrayPtr) {
+	for (i=0 ; i<currElems ; i++) {
+	    if (newArrayPtr[i].nextPtr != NULL) {
+		newArrayPtr[i].nextPtr = newArrayPtr
+			+ (newArrayPtr[i].nextPtr - currArrayPtr);
+	    }
+	}
+	for (i=0 ; i<localTablePtr->numBuckets ; i++) {
+	    if (localTablePtr->buckets[i] != NULL) {
+		localTablePtr->buckets[i] = newArrayPtr
+			+ (localTablePtr->buckets[i] - currArrayPtr);
+	    }
+	}
+    }
+
+    envPtr->literalArrayPtr = newArrayPtr;
+    envPtr->literalArrayEnd = newSize / sizeof(LiteralEntry);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclReleaseLiteral --
+ *
+ *	This function releases a reference to one of the shared Tcl objects
+ *	that hold literals. It is called to release the literals referenced by
+ *	a ByteCode that is being destroyed, and it is also called by
+ *	TclDeleteLiteralTable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The reference count for the global LiteralTable entry that corresponds
+ *	to the literal is decremented. If no other reference to a global
+ *	literal object remains, it is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclReleaseLiteral(
+    Tcl_Interp *interp,		/* Interpreter for which objPtr was created to
+				 * hold a literal. */
+    register Tcl_Obj *objPtr)	/* Points to a literal object that was
+				 * previously created by a call to
+				 * TclRegisterLiteral. */
+{
+    Interp *iPtr = (Interp *) interp;
+    LiteralTable *globalTablePtr;
+    register LiteralEntry *entryPtr, *prevPtr;
+    const char *bytes;
+    int length, index;
+
+    if (iPtr == NULL) {
+	goto done;
+    }
+
+    globalTablePtr = &iPtr->literalTable;
+    bytes = TclGetStringFromObj(objPtr, &length);
+    index = (HashString(bytes, length) & globalTablePtr->mask);
+
+    /*
+     * Check to see if the object is in the global literal table and remove
+     * this reference. The object may not be in the table if it is a hidden
+     * local literal.
+     */
+
+    for (prevPtr=NULL, entryPtr=globalTablePtr->buckets[index];
+	    entryPtr!=NULL ; prevPtr=entryPtr, entryPtr=entryPtr->nextPtr) {
+	if (entryPtr->objPtr == objPtr) {
+	    entryPtr->refCount--;
+
+	    /*
+	     * If the literal is no longer being used by any ByteCode, delete
+	     * the entry then remove the reference corresponding to the global
+	     * literal table entry (decrement the ref count of the object).
+	     */
+
+	    if (entryPtr->refCount == 0) {
+		if (prevPtr == NULL) {
+		    globalTablePtr->buckets[index] = entryPtr->nextPtr;
+		} else {
+		    prevPtr->nextPtr = entryPtr->nextPtr;
+		}
+		ckfree(entryPtr);
+		globalTablePtr->numEntries--;
+
+		TclDecrRefCount(objPtr);
+
+#ifdef TCL_COMPILE_STATS
+		iPtr->stats.currentLitStringBytes -= (double) (length + 1);
+#endif /*TCL_COMPILE_STATS*/
+	    }
+	    break;
+	}
+    }
+
+    /*
+     * Remove the reference corresponding to the local literal table entry.
+     */
+
+    done:
+    Tcl_DecrRefCount(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * HashString --
+ *
+ *	Compute a one-word summary of a text string, which can be used to
+ *	generate a hash index.
+ *
+ * Results:
+ *	The return value is a one-word summary of the information in string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static unsigned
+HashString(
+    register const char *string,	/* String for which to compute hash value. */
+    int length)			/* Number of bytes in the string. */
+{
+    register unsigned int result = 0;
+
+    /*
+     * I tried a zillion different hash functions and asked many other people
+     * for advice. Many people had their own favorite functions, all
+     * different, but no-one had much idea why they were good ones. I chose
+     * the one below (multiply by 9 and add new character) because of the
+     * following reasons:
+     *
+     * 1. Multiplying by 10 is perfect for keys that are decimal strings, and
+     *    multiplying by 9 is just about as good.
+     * 2. Times-9 is (shift-left-3) plus (old). This means that each
+     *    character's bits hang around in the low-order bits of the hash value
+     *    for ever, plus they spread fairly rapidly up to the high-order bits
+     *    to fill out the hash value. This seems works well both for decimal
+     *    and non-decimal strings.
+     *
+     * Note that this function is very weak against malicious strings; it's
+     * very easy to generate multiple keys that have the same hashcode. On the
+     * other hand, that hardly ever actually occurs and this function *is*
+     * very cheap, even by comparison with industry-standard hashes like FNV.
+     * If real strength of hash is required though, use a custom hash based on
+     * Bob Jenkins's lookup3(), but be aware that it's significantly slower.
+     * Tcl scripts tend to not have a big issue in this area, and literals
+     * mostly aren't looked up by name anyway.
+     *
+     * See also HashStringKey in tclHash.c.
+     * See also TclObjHashKey in tclObj.c.
+     *
+     * See [tcl-Feature Request #2958832]
+     */
+
+    if (length > 0) {
+	result = UCHAR(*string);
+	while (--length) {
+	    result += (result << 3) + UCHAR(*++string);
+	}
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RebuildLiteralTable --
+ *
+ *	This function is invoked when the ratio of entries to hash buckets
+ *	becomes too large in a local or global literal table. It allocates a
+ *	larger bucket array and moves the entries into the new buckets.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory gets reallocated and entries get rehashed into new buckets.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RebuildLiteralTable(
+    register LiteralTable *tablePtr)
+				/* Local or global table to enlarge. */
+{
+    LiteralEntry **oldBuckets;
+    register LiteralEntry **oldChainPtr, **newChainPtr;
+    register LiteralEntry *entryPtr;
+    LiteralEntry **bucketPtr;
+    const char *bytes;
+    unsigned int oldSize;
+    int count, index, length;
+
+    oldSize = tablePtr->numBuckets;
+    oldBuckets = tablePtr->buckets;
+
+    /*
+     * Allocate and initialize the new bucket array, and set up hashing
+     * constants for new array size.
+     */
+
+    if (oldSize > UINT_MAX/(4 * sizeof(LiteralEntry *))) {
+	/*
+	 * Memory allocator limitations will not let us create the
+	 * next larger table size.  Best option is to limp along
+	 * with what we have.
+	 */
+
+	return;
+    }
+
+    tablePtr->numBuckets *= 4;
+    tablePtr->buckets = ckalloc(tablePtr->numBuckets * sizeof(LiteralEntry*));
+    for (count=tablePtr->numBuckets, newChainPtr=tablePtr->buckets;
+	    count>0 ; count--, newChainPtr++) {
+	*newChainPtr = NULL;
+    }
+    tablePtr->rebuildSize *= 4;
+    tablePtr->mask = (tablePtr->mask << 2) + 3;
+
+    /*
+     * Rehash all of the existing entries into the new bucket array.
+     */
+
+    for (oldChainPtr=oldBuckets ; oldSize>0 ; oldSize--,oldChainPtr++) {
+	for (entryPtr=*oldChainPtr ; entryPtr!=NULL ; entryPtr=*oldChainPtr) {
+	    bytes = TclGetStringFromObj(entryPtr->objPtr, &length);
+	    index = (HashString(bytes, length) & tablePtr->mask);
+
+	    *oldChainPtr = entryPtr->nextPtr;
+	    bucketPtr = &tablePtr->buckets[index];
+	    entryPtr->nextPtr = *bucketPtr;
+	    *bucketPtr = entryPtr;
+	}
+    }
+
+    /*
+     * Free up the old bucket array, if it was dynamically allocated.
+     */
+
+    if (oldBuckets != tablePtr->staticBuckets) {
+	ckfree(oldBuckets);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInvalidateCmdLiteral --
+ *
+ *	Invalidate a command literal entry, if present in the literal hash
+ *	tables, by resetting its internal representation. This invalidation
+ *	leaves it in the literal tables and in existing literal arrays. As a
+ *	result, existing references continue to work but we force a fresh
+ *	command look-up upon the next use (see, in particular,
+ *	TclSetCmdNameObj()).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Resets the internal representation of the CmdName Tcl_Obj
+ *	using TclFreeIntRep().
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInvalidateCmdLiteral(
+    Tcl_Interp *interp,		/* Interpreter for which to invalidate a
+				 * command literal. */
+    const char *name,		/* Points to the start of the cmd literal
+				 * name. */
+    Namespace *nsPtr)		/* The namespace for which to lookup and
+				 * invalidate a cmd literal. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *literalObjPtr = TclCreateLiteral(iPtr, name,
+	    strlen(name), -1, NULL, nsPtr, 0, NULL);
+
+    if (literalObjPtr != NULL) {
+	if (literalObjPtr->typePtr == &tclCmdNameType) {
+	    TclFreeIntRep(literalObjPtr);
+	}
+	/* Balance the refcount effects of TclCreateLiteral() above */
+	Tcl_IncrRefCount(literalObjPtr);
+	TclReleaseLiteral(interp, literalObjPtr);
+    }
+}
+
+#ifdef TCL_COMPILE_STATS
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLiteralStats --
+ *
+ *	Return statistics describing the layout of the hash table in its hash
+ *	buckets.
+ *
+ * Results:
+ *	The return value is a malloc-ed string containing information about
+ *	tablePtr. It is the caller's responsibility to free this string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclLiteralStats(
+    LiteralTable *tablePtr)	/* Table for which to produce stats. */
+{
+#define NUM_COUNTERS 10
+    int count[NUM_COUNTERS], overflow, i, j;
+    double average, tmp;
+    register LiteralEntry *entryPtr;
+    char *result, *p;
+
+    /*
+     * Compute a histogram of bucket usage. For each bucket chain i, j is the
+     * number of entries in the chain.
+     */
+
+    for (i=0 ; i<NUM_COUNTERS ; i++) {
+	count[i] = 0;
+    }
+    overflow = 0;
+    average = 0.0;
+    for (i=0 ; i<tablePtr->numBuckets ; i++) {
+	j = 0;
+	for (entryPtr=tablePtr->buckets[i] ; entryPtr!=NULL;
+		entryPtr=entryPtr->nextPtr) {
+	    j++;
+	}
+	if (j < NUM_COUNTERS) {
+	    count[j]++;
+	} else {
+	    overflow++;
+	}
+	tmp = j;
+	average += (tmp+1.0)*(tmp/tablePtr->numEntries)/2.0;
+    }
+
+    /*
+     * Print out the histogram and a few other pieces of information.
+     */
+
+    result = ckalloc(NUM_COUNTERS*60 + 300);
+    sprintf(result, "%d entries in table, %d buckets\n",
+	    tablePtr->numEntries, tablePtr->numBuckets);
+    p = result + strlen(result);
+    for (i=0 ; i<NUM_COUNTERS ; i++) {
+	sprintf(p, "number of buckets with %d entries: %d\n",
+		i, count[i]);
+	p += strlen(p);
+    }
+    sprintf(p, "number of buckets with %d or more entries: %d\n",
+	    NUM_COUNTERS, overflow);
+    p += strlen(p);
+    sprintf(p, "average search distance for entry: %.1f", average);
+    return result;
+}
+#endif /*TCL_COMPILE_STATS*/
+
+#ifdef TCL_COMPILE_DEBUG
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclVerifyLocalLiteralTable --
+ *
+ *	Check a CompileEnv's local literal table for consistency.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Tcl_Panic if problems are found.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclVerifyLocalLiteralTable(
+    CompileEnv *envPtr)		/* Points to CompileEnv whose literal table is
+				 * to be validated. */
+{
+    register LiteralTable *localTablePtr = &envPtr->localLitTable;
+    register LiteralEntry *localPtr;
+    char *bytes;
+    register int i;
+    int length, count;
+
+    count = 0;
+    for (i=0 ; i<localTablePtr->numBuckets ; i++) {
+	for (localPtr=localTablePtr->buckets[i] ; localPtr!=NULL;
+		localPtr=localPtr->nextPtr) {
+	    count++;
+	    if (localPtr->refCount != -1) {
+		bytes = TclGetStringFromObj(localPtr->objPtr, &length);
+		Tcl_Panic("%s: local literal \"%.*s\" had bad refCount %d",
+			"TclVerifyLocalLiteralTable",
+			(length>60? 60 : length), bytes, localPtr->refCount);
+	    }
+	    if (localPtr->objPtr->bytes == NULL) {
+		Tcl_Panic("%s: literal has NULL string rep",
+			"TclVerifyLocalLiteralTable");
+	    }
+	}
+    }
+    if (count != localTablePtr->numEntries) {
+	Tcl_Panic("%s: local literal table had %d entries, should be %d",
+		"TclVerifyLocalLiteralTable", count,
+		localTablePtr->numEntries);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclVerifyGlobalLiteralTable --
+ *
+ *	Check an interpreter's global literal table literal for consistency.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Tcl_Panic if problems are found.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclVerifyGlobalLiteralTable(
+    Interp *iPtr)		/* Points to interpreter whose global literal
+				 * table is to be validated. */
+{
+    register LiteralTable *globalTablePtr = &iPtr->literalTable;
+    register LiteralEntry *globalPtr;
+    char *bytes;
+    register int i;
+    int length, count;
+
+    count = 0;
+    for (i=0 ; i<globalTablePtr->numBuckets ; i++) {
+	for (globalPtr=globalTablePtr->buckets[i] ; globalPtr!=NULL;
+		globalPtr=globalPtr->nextPtr) {
+	    count++;
+	    if (globalPtr->refCount < 1) {
+		bytes = TclGetStringFromObj(globalPtr->objPtr, &length);
+		Tcl_Panic("%s: global literal \"%.*s\" had bad refCount %d",
+			"TclVerifyGlobalLiteralTable",
+			(length>60? 60 : length), bytes, globalPtr->refCount);
+	    }
+	    if (globalPtr->objPtr->bytes == NULL) {
+		Tcl_Panic("%s: literal has NULL string rep",
+			"TclVerifyGlobalLiteralTable");
+	    }
+	}
+    }
+    if (count != globalTablePtr->numEntries) {
+	Tcl_Panic("%s: global literal table had %d entries, should be %d",
+		"TclVerifyGlobalLiteralTable", count,
+		globalTablePtr->numEntries);
+    }
+}
+#endif /*TCL_COMPILE_DEBUG*/
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclLoad.c b/generic/tclLoad.c
new file mode 100644
index 0000000..e0bb5ef
--- /dev/null
+++ b/generic/tclLoad.c
@@ -0,0 +1,1212 @@
+/*
+ * tclLoad.c --
+ *
+ *	This file provides the generic portion (those that are the same on all
+ *	platforms) of Tcl's dynamic loading facilities.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * The following structure describes a package that has been loaded either
+ * dynamically (with the "load" command) or statically (as indicated by a call
+ * to TclGetLoadedPackages). All such packages are linked together into a
+ * single list for the process. Packages are never unloaded, until the
+ * application exits, when TclFinalizeLoad is called, and these structures are
+ * freed.
+ */
+
+typedef struct LoadedPackage {
+    char *fileName;		/* Name of the file from which the package was
+				 * loaded. An empty string means the package
+				 * is loaded statically. Malloc-ed. */
+    char *packageName;		/* Name of package prefix for the package,
+				 * properly capitalized (first letter UC,
+				 * others LC), no "_", as in "Net".
+				 * Malloc-ed. */
+    Tcl_LoadHandle loadHandle;	/* Token for the loaded file which should be
+				 * passed to (*unLoadProcPtr)() when the file
+				 * is no longer needed. If fileName is NULL,
+				 * then this field is irrelevant. */
+    Tcl_PackageInitProc *initProc;
+				/* Initialization function to call to
+				 * incorporate this package into a trusted
+				 * interpreter. */
+    Tcl_PackageInitProc *safeInitProc;
+				/* Initialization function to call to
+				 * incorporate this package into a safe
+				 * interpreter (one that will execute
+				 * untrusted scripts). NULL means the package
+				 * can't be used in unsafe interpreters. */
+    Tcl_PackageUnloadProc *unloadProc;
+				/* Finalisation function to unload a package
+				 * from a trusted interpreter. NULL means that
+				 * the package cannot be unloaded. */
+    Tcl_PackageUnloadProc *safeUnloadProc;
+				/* Finalisation function to unload a package
+				 * from a safe interpreter. NULL means that
+				 * the package cannot be unloaded. */
+    int interpRefCount;		/* How many times the package has been loaded
+				 * in trusted interpreters. */
+    int safeInterpRefCount;	/* How many times the package has been loaded
+				 * in safe interpreters. */
+    struct LoadedPackage *nextPtr;
+				/* Next in list of all packages loaded into
+				 * this application process. NULL means end of
+				 * list. */
+} LoadedPackage;
+
+/*
+ * TCL_THREADS
+ * There is a global list of packages that is anchored at firstPackagePtr.
+ * Access to this list is governed by a mutex.
+ */
+
+static LoadedPackage *firstPackagePtr = NULL;
+				/* First in list of all packages loaded into
+				 * this process. */
+
+TCL_DECLARE_MUTEX(packageMutex)
+
+/*
+ * The following structure represents a particular package that has been
+ * incorporated into a particular interpreter (by calling its initialization
+ * function). There is a list of these structures for each interpreter, with
+ * an AssocData value (key "load") for the interpreter that points to the
+ * first package (if any).
+ */
+
+typedef struct InterpPackage {
+    LoadedPackage *pkgPtr;	/* Points to detailed information about
+				 * package. */
+    struct InterpPackage *nextPtr;
+				/* Next package in this interpreter, or NULL
+				 * for end of list. */
+} InterpPackage;
+
+/*
+ * Prototypes for functions that are private to this file:
+ */
+
+static void		LoadCleanupProc(ClientData clientData,
+			    Tcl_Interp *interp);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LoadObjCmd --
+ *
+ *	This function is invoked to process the "load" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_LoadObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Interp *target;
+    LoadedPackage *pkgPtr, *defaultPtr;
+    Tcl_DString pkgName, tmp, initName, safeInitName;
+    Tcl_DString unloadName, safeUnloadName;
+    InterpPackage *ipFirstPtr, *ipPtr;
+    int code, namesMatch, filesMatch, offset;
+    const char *symbols[2];
+    Tcl_PackageInitProc *initProc;
+    const char *p, *fullFileName, *packageName;
+    Tcl_LoadHandle loadHandle;
+    Tcl_UniChar ch = 0;
+    unsigned len;
+    int index, flags = 0;
+    Tcl_Obj *const *savedobjv = objv;
+    static const char *const options[] = {
+	"-global",		"-lazy",		"--",	NULL
+    };
+    enum options {
+	LOAD_GLOBAL,	LOAD_LAZY,	LOAD_LAST
+    };
+
+    while (objc > 2) {
+	if (TclGetString(objv[1])[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[1], options, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	++objv; --objc;
+	if (LOAD_GLOBAL == (enum options) index) {
+	    flags |= TCL_LOAD_GLOBAL;
+	} else if (LOAD_LAZY == (enum options) index) {
+	    flags |= TCL_LOAD_LAZY;
+	} else {
+		break;
+	}
+    }
+    if ((objc < 2) || (objc > 4)) {
+	Tcl_WrongNumArgs(interp, 1, savedobjv, "?-global? ?-lazy? ?--? fileName ?packageName? ?interp?");
+	return TCL_ERROR;
+    }
+    if (Tcl_FSConvertToPathType(interp, objv[1]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    fullFileName = Tcl_GetString(objv[1]);
+
+    Tcl_DStringInit(&pkgName);
+    Tcl_DStringInit(&initName);
+    Tcl_DStringInit(&safeInitName);
+    Tcl_DStringInit(&unloadName);
+    Tcl_DStringInit(&safeUnloadName);
+    Tcl_DStringInit(&tmp);
+
+    packageName = NULL;
+    if (objc >= 3) {
+	packageName = Tcl_GetString(objv[2]);
+	if (packageName[0] == '\0') {
+	    packageName = NULL;
+	}
+    }
+    if ((fullFileName[0] == 0) && (packageName == NULL)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"must specify either file name or package name", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD", "NOLIBRARY",
+		NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Figure out which interpreter we're going to load the package into.
+     */
+
+    target = interp;
+    if (objc == 4) {
+	const char *slaveIntName = Tcl_GetString(objv[3]);
+
+	target = Tcl_GetSlave(interp, slaveIntName);
+	if (target == NULL) {
+	    code = TCL_ERROR;
+	    goto done;
+	}
+    }
+
+    /*
+     * Scan through the packages that are currently loaded to see if the
+     * package we want is already loaded. We'll use a loaded package if it
+     * meets any of the following conditions:
+     *  - Its name and file match the once we're looking for.
+     *  - Its file matches, and we weren't given a name.
+     *  - Its name matches, the file name was specified as empty, and there is
+     *	  only no statically loaded package with the same name.
+     */
+
+    Tcl_MutexLock(&packageMutex);
+
+    defaultPtr = NULL;
+    for (pkgPtr = firstPackagePtr; pkgPtr != NULL; pkgPtr = pkgPtr->nextPtr) {
+	if (packageName == NULL) {
+	    namesMatch = 0;
+	} else {
+	    TclDStringClear(&pkgName);
+	    Tcl_DStringAppend(&pkgName, packageName, -1);
+	    TclDStringClear(&tmp);
+	    Tcl_DStringAppend(&tmp, pkgPtr->packageName, -1);
+	    Tcl_UtfToLower(Tcl_DStringValue(&pkgName));
+	    Tcl_UtfToLower(Tcl_DStringValue(&tmp));
+	    if (strcmp(Tcl_DStringValue(&tmp),
+		    Tcl_DStringValue(&pkgName)) == 0) {
+		namesMatch = 1;
+	    } else {
+		namesMatch = 0;
+	    }
+	}
+	TclDStringClear(&pkgName);
+
+	filesMatch = (strcmp(pkgPtr->fileName, fullFileName) == 0);
+	if (filesMatch && (namesMatch || (packageName == NULL))) {
+	    break;
+	}
+	if (namesMatch && (fullFileName[0] == 0)) {
+	    defaultPtr = pkgPtr;
+	}
+	if (filesMatch && !namesMatch && (fullFileName[0] != 0)) {
+	    /*
+	     * Can't have two different packages loaded from the same file.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "file \"%s\" is already loaded for package \"%s\"",
+		    fullFileName, pkgPtr->packageName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD",
+		    "SPLITPERSONALITY", NULL);
+	    code = TCL_ERROR;
+	    Tcl_MutexUnlock(&packageMutex);
+	    goto done;
+	}
+    }
+    Tcl_MutexUnlock(&packageMutex);
+    if (pkgPtr == NULL) {
+	pkgPtr = defaultPtr;
+    }
+
+    /*
+     * Scan through the list of packages already loaded in the target
+     * interpreter. If the package we want is already loaded there, then
+     * there's nothing for us to do.
+     */
+
+    if (pkgPtr != NULL) {
+	ipFirstPtr = Tcl_GetAssocData(target, "tclLoad", NULL);
+	for (ipPtr = ipFirstPtr; ipPtr != NULL; ipPtr = ipPtr->nextPtr) {
+	    if (ipPtr->pkgPtr == pkgPtr) {
+		code = TCL_OK;
+		goto done;
+	    }
+	}
+    }
+
+    if (pkgPtr == NULL) {
+	/*
+	 * The desired file isn't currently loaded, so load it. It's an error
+	 * if the desired package is a static one.
+	 */
+
+	if (fullFileName[0] == 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "package \"%s\" isn't loaded statically", packageName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD", "NOTSTATIC",
+		    NULL);
+	    code = TCL_ERROR;
+	    goto done;
+	}
+
+	/*
+	 * Figure out the module name if it wasn't provided explicitly.
+	 */
+
+	if (packageName != NULL) {
+	    Tcl_DStringAppend(&pkgName, packageName, -1);
+	} else {
+	    int retc;
+
+	    /*
+	     * Threading note - this call used to be protected by a mutex.
+	     */
+
+	    retc = TclGuessPackageName(fullFileName, &pkgName);
+	    if (!retc) {
+		Tcl_Obj *splitPtr, *pkgGuessPtr;
+		int pElements;
+		const char *pkgGuess;
+
+		/*
+		 * The platform-specific code couldn't figure out the module
+		 * name. Make a guess by taking the last element of the file
+		 * name, stripping off any leading "lib", and then using all
+		 * of the alphabetic and underline characters that follow
+		 * that.
+		 */
+
+		splitPtr = Tcl_FSSplitPath(objv[1], &pElements);
+		Tcl_ListObjIndex(NULL, splitPtr, pElements -1, &pkgGuessPtr);
+		pkgGuess = Tcl_GetString(pkgGuessPtr);
+		if ((pkgGuess[0] == 'l') && (pkgGuess[1] == 'i')
+			&& (pkgGuess[2] == 'b')) {
+		    pkgGuess += 3;
+		}
+#ifdef __CYGWIN__
+		if ((pkgGuess[0] == 'c') && (pkgGuess[1] == 'y')
+			&& (pkgGuess[2] == 'g')) {
+		    pkgGuess += 3;
+		}
+#endif /* __CYGWIN__ */
+		for (p = pkgGuess; *p != 0; p += offset) {
+		    offset = TclUtfToUniChar(p, &ch);
+		    if ((ch > 0x100)
+			    || !(isalpha(UCHAR(ch)) /* INTL: ISO only */
+				    || (UCHAR(ch) == '_'))) {
+			break;
+		    }
+		}
+		if (p == pkgGuess) {
+		    Tcl_DecrRefCount(splitPtr);
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "couldn't figure out package name for %s",
+			    fullFileName));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD",
+			    "WHATPACKAGE", NULL);
+		    code = TCL_ERROR;
+		    goto done;
+		}
+		Tcl_DStringAppend(&pkgName, pkgGuess, p - pkgGuess);
+		Tcl_DecrRefCount(splitPtr);
+	    }
+	}
+
+	/*
+	 * Fix the capitalization in the package name so that the first
+	 * character is in caps (or title case) but the others are all
+	 * lower-case.
+	 */
+
+	Tcl_DStringSetLength(&pkgName,
+		Tcl_UtfToTitle(Tcl_DStringValue(&pkgName)));
+
+	/*
+	 * Compute the names of the two initialization functions, based on the
+	 * package name.
+	 */
+
+	TclDStringAppendDString(&initName, &pkgName);
+	TclDStringAppendLiteral(&initName, "_Init");
+	TclDStringAppendDString(&safeInitName, &pkgName);
+	TclDStringAppendLiteral(&safeInitName, "_SafeInit");
+	TclDStringAppendDString(&unloadName, &pkgName);
+	TclDStringAppendLiteral(&unloadName, "_Unload");
+	TclDStringAppendDString(&safeUnloadName, &pkgName);
+	TclDStringAppendLiteral(&safeUnloadName, "_SafeUnload");
+
+	/*
+	 * Call platform-specific code to load the package and find the two
+	 * initialization functions.
+	 */
+
+	symbols[0] = Tcl_DStringValue(&initName);
+	symbols[1] = NULL;
+
+	Tcl_MutexLock(&packageMutex);
+	code = Tcl_LoadFile(interp, objv[1], symbols, flags, &initProc,
+		&loadHandle);
+	Tcl_MutexUnlock(&packageMutex);
+	if (code != TCL_OK) {
+	    goto done;
+	}
+
+	/*
+	 * Create a new record to describe this package.
+	 */
+
+	pkgPtr = ckalloc(sizeof(LoadedPackage));
+	len = strlen(fullFileName) + 1;
+	pkgPtr->fileName	   = ckalloc(len);
+	memcpy(pkgPtr->fileName, fullFileName, len);
+	len = (unsigned) Tcl_DStringLength(&pkgName) + 1;
+	pkgPtr->packageName	   = ckalloc(len);
+	memcpy(pkgPtr->packageName, Tcl_DStringValue(&pkgName), len);
+	pkgPtr->loadHandle	   = loadHandle;
+	pkgPtr->initProc	   = initProc;
+	pkgPtr->safeInitProc	   = (Tcl_PackageInitProc *)
+		Tcl_FindSymbol(interp, loadHandle,
+			Tcl_DStringValue(&safeInitName));
+	pkgPtr->unloadProc	   = (Tcl_PackageUnloadProc *)
+		Tcl_FindSymbol(interp, loadHandle,
+			Tcl_DStringValue(&unloadName));
+	pkgPtr->safeUnloadProc	   = (Tcl_PackageUnloadProc *)
+		Tcl_FindSymbol(interp, loadHandle,
+			Tcl_DStringValue(&safeUnloadName));
+	pkgPtr->interpRefCount	   = 0;
+	pkgPtr->safeInterpRefCount = 0;
+
+	Tcl_MutexLock(&packageMutex);
+	pkgPtr->nextPtr		   = firstPackagePtr;
+	firstPackagePtr		   = pkgPtr;
+	Tcl_MutexUnlock(&packageMutex);
+
+	/*
+	 * The Tcl_FindSymbol calls may have left a spurious error message in
+	 * the interpreter result.
+	 */
+
+	Tcl_ResetResult(interp);
+    }
+
+    /*
+     * Invoke the package's initialization function (either the normal one or
+     * the safe one, depending on whether or not the interpreter is safe).
+     */
+
+    if (Tcl_IsSafe(target)) {
+	if (pkgPtr->safeInitProc == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "can't use package in a safe interpreter: no"
+		    " %s_SafeInit procedure", pkgPtr->packageName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD", "UNSAFE",
+		    NULL);
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	code = pkgPtr->safeInitProc(target);
+    } else {
+	if (pkgPtr->initProc == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "can't attach package to interpreter: no %s_Init procedure",
+		    pkgPtr->packageName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "LOAD", "ENTRYPOINT",
+		    NULL);
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	code = pkgPtr->initProc(target);
+    }
+
+    /*
+     * Test for whether the initialization failed. If so, transfer the error
+     * from the target interpreter to the originating one.
+     */
+
+    if (code != TCL_OK) {
+	Tcl_TransferResult(target, code, interp);
+	goto done;
+    }
+
+    /*
+     * Record the fact that the package has been loaded in the target
+     * interpreter.
+     *
+     * Update the proper reference count.
+     */
+
+    Tcl_MutexLock(&packageMutex);
+    if (Tcl_IsSafe(target)) {
+	pkgPtr->safeInterpRefCount++;
+    } else {
+	pkgPtr->interpRefCount++;
+    }
+    Tcl_MutexUnlock(&packageMutex);
+
+    /*
+     * Refetch ipFirstPtr: loading the package may have introduced additional
+     * static packages at the head of the linked list!
+     */
+
+    ipFirstPtr = Tcl_GetAssocData(target, "tclLoad", NULL);
+    ipPtr = ckalloc(sizeof(InterpPackage));
+    ipPtr->pkgPtr = pkgPtr;
+    ipPtr->nextPtr = ipFirstPtr;
+    Tcl_SetAssocData(target, "tclLoad", LoadCleanupProc, ipPtr);
+
+  done:
+    Tcl_DStringFree(&pkgName);
+    Tcl_DStringFree(&initName);
+    Tcl_DStringFree(&safeInitName);
+    Tcl_DStringFree(&unloadName);
+    Tcl_DStringFree(&safeUnloadName);
+    Tcl_DStringFree(&tmp);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnloadObjCmd --
+ *
+ *	This function is invoked to process the "unload" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UnloadObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Interp *target;		/* Which interpreter to unload from. */
+    LoadedPackage *pkgPtr, *defaultPtr;
+    Tcl_DString pkgName, tmp;
+    Tcl_PackageUnloadProc *unloadProc;
+    InterpPackage *ipFirstPtr, *ipPtr;
+    int i, index, code, complain = 1, keepLibrary = 0;
+    int trustedRefCount = -1, safeRefCount = -1;
+    const char *fullFileName = "";
+    const char *packageName;
+    static const char *const options[] = {
+	"-nocomplain", "-keeplibrary", "--", NULL
+    };
+    enum options {
+	UNLOAD_NOCOMPLAIN, UNLOAD_KEEPLIB, UNLOAD_LAST
+    };
+
+    for (i = 1; i < objc; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		&index) != TCL_OK) {
+	    fullFileName = Tcl_GetString(objv[i]);
+	    if (fullFileName[0] == '-') {
+		/*
+		 * It looks like the command contains an option so signal an
+		 * error
+		 */
+
+		return TCL_ERROR;
+	    } else {
+		/*
+		 * This clearly isn't an option; assume it's the filename. We
+		 * must clear the error.
+		 */
+
+		Tcl_ResetResult(interp);
+		break;
+	    }
+	}
+	switch (index) {
+	case UNLOAD_NOCOMPLAIN:		/* -nocomplain */
+	    complain = 0;
+	    break;
+	case UNLOAD_KEEPLIB:		/* -keeplibrary */
+	    keepLibrary = 1;
+	    break;
+	case UNLOAD_LAST:		/* -- */
+	    i++;
+	    goto endOfForLoop;
+	}
+    }
+  endOfForLoop:
+    if ((objc-i < 1) || (objc-i > 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-switch ...? fileName ?packageName? ?interp?");
+	return TCL_ERROR;
+    }
+    if (Tcl_FSConvertToPathType(interp, objv[i]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    fullFileName = Tcl_GetString(objv[i]);
+    Tcl_DStringInit(&pkgName);
+    Tcl_DStringInit(&tmp);
+
+    packageName = NULL;
+    if (objc - i >= 2) {
+	packageName = Tcl_GetString(objv[i+1]);
+	if (packageName[0] == '\0') {
+	    packageName = NULL;
+	}
+    }
+    if ((fullFileName[0] == 0) && (packageName == NULL)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"must specify either file name or package name", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "NOLIBRARY",
+		NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Figure out which interpreter we're going to load the package into.
+     */
+
+    target = interp;
+    if (objc - i == 3) {
+	const char *slaveIntName = Tcl_GetString(objv[i + 2]);
+
+	target = Tcl_GetSlave(interp, slaveIntName);
+	if (target == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Scan through the packages that are currently loaded to see if the
+     * package we want is already loaded. We'll use a loaded package if it
+     * meets any of the following conditions:
+     *  - Its name and file match the once we're looking for.
+     *  - Its file matches, and we weren't given a name.
+     *  - Its name matches, the file name was specified as empty, and there is
+     *	  only no statically loaded package with the same name.
+     */
+
+    Tcl_MutexLock(&packageMutex);
+
+    defaultPtr = NULL;
+    for (pkgPtr = firstPackagePtr; pkgPtr != NULL; pkgPtr = pkgPtr->nextPtr) {
+	int namesMatch, filesMatch;
+
+	if (packageName == NULL) {
+	    namesMatch = 0;
+	} else {
+	    TclDStringClear(&pkgName);
+	    Tcl_DStringAppend(&pkgName, packageName, -1);
+	    TclDStringClear(&tmp);
+	    Tcl_DStringAppend(&tmp, pkgPtr->packageName, -1);
+	    Tcl_UtfToLower(Tcl_DStringValue(&pkgName));
+	    Tcl_UtfToLower(Tcl_DStringValue(&tmp));
+	    if (strcmp(Tcl_DStringValue(&tmp),
+		    Tcl_DStringValue(&pkgName)) == 0) {
+		namesMatch = 1;
+	    } else {
+		namesMatch = 0;
+	    }
+	}
+	TclDStringClear(&pkgName);
+
+	filesMatch = (strcmp(pkgPtr->fileName, fullFileName) == 0);
+	if (filesMatch && (namesMatch || (packageName == NULL))) {
+	    break;
+	}
+	if (namesMatch && (fullFileName[0] == 0)) {
+	    defaultPtr = pkgPtr;
+	}
+	if (filesMatch && !namesMatch && (fullFileName[0] != 0)) {
+	    break;
+	}
+    }
+    Tcl_MutexUnlock(&packageMutex);
+    if (fullFileName[0] == 0) {
+	/*
+	 * It's an error to try unload a static package.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"package \"%s\" is loaded statically and cannot be unloaded",
+		packageName));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "STATIC",
+		NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+    if (pkgPtr == NULL) {
+	/*
+	 * The DLL pointed by the provided filename has never been loaded.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"file \"%s\" has never been loaded", fullFileName));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "NEVERLOADED",
+		NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Scan through the list of packages already loaded in the target
+     * interpreter. If the package we want is already loaded there, then we
+     * should proceed with unloading.
+     */
+
+    code = TCL_ERROR;
+    if (pkgPtr != NULL) {
+	ipFirstPtr = Tcl_GetAssocData(target, "tclLoad", NULL);
+	for (ipPtr = ipFirstPtr; ipPtr != NULL; ipPtr = ipPtr->nextPtr) {
+	    if (ipPtr->pkgPtr == pkgPtr) {
+		code = TCL_OK;
+		break;
+	    }
+	}
+    }
+    if (code != TCL_OK) {
+	/*
+	 * The package has not been loaded in this interpreter.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"file \"%s\" has never been loaded in this interpreter",
+		fullFileName));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "NEVERLOADED",
+		NULL);
+	code = TCL_ERROR;
+	goto done;
+    }
+
+    /*
+     * Ensure that the DLL can be unloaded. If it is a trusted interpreter,
+     * pkgPtr->unloadProc must not be NULL for the DLL to be unloadable. If
+     * the interpreter is a safe one, pkgPtr->safeUnloadProc must be non-NULL.
+     */
+
+    if (Tcl_IsSafe(target)) {
+	if (pkgPtr->safeUnloadProc == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "file \"%s\" cannot be unloaded under a safe interpreter",
+		    fullFileName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "CANNOT",
+		    NULL);
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	unloadProc = pkgPtr->safeUnloadProc;
+    } else {
+	if (pkgPtr->unloadProc == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "file \"%s\" cannot be unloaded under a trusted interpreter",
+		    fullFileName));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "CANNOT",
+		    NULL);
+	    code = TCL_ERROR;
+	    goto done;
+	}
+	unloadProc = pkgPtr->unloadProc;
+    }
+
+    /*
+     * We are ready to unload the package. First, evaluate the unload
+     * function. If this fails, we cannot proceed with unload. Also, we must
+     * specify the proper flag to pass to the unload callback.
+     * TCL_UNLOAD_DETACH_FROM_INTERPRETER is defined when the callback should
+     * only remove itself from the interpreter; the library will be unloaded
+     * in a future call of unload. In case the library will be unloaded just
+     * after the callback returns, TCL_UNLOAD_DETACH_FROM_PROCESS is passed.
+     */
+
+    code = TCL_UNLOAD_DETACH_FROM_INTERPRETER;
+    if (!keepLibrary) {
+	Tcl_MutexLock(&packageMutex);
+	trustedRefCount = pkgPtr->interpRefCount;
+	safeRefCount = pkgPtr->safeInterpRefCount;
+	Tcl_MutexUnlock(&packageMutex);
+
+	if (Tcl_IsSafe(target)) {
+	    safeRefCount--;
+	} else {
+	    trustedRefCount--;
+	}
+
+	if (safeRefCount <= 0 && trustedRefCount <= 0) {
+	    code = TCL_UNLOAD_DETACH_FROM_PROCESS;
+	}
+    }
+    code = unloadProc(target, code);
+    if (code != TCL_OK) {
+	Tcl_TransferResult(target, code, interp);
+	goto done;
+    }
+
+    /*
+     * The unload function executed fine. Examine the reference count to see
+     * if we unload the DLL.
+     */
+
+    Tcl_MutexLock(&packageMutex);
+    if (Tcl_IsSafe(target)) {
+	pkgPtr->safeInterpRefCount--;
+
+	/*
+	 * Do not let counter get negative.
+	 */
+
+	if (pkgPtr->safeInterpRefCount < 0) {
+	    pkgPtr->safeInterpRefCount = 0;
+	}
+    } else {
+	pkgPtr->interpRefCount--;
+
+	/*
+	 * Do not let counter get negative.
+	 */
+
+	if (pkgPtr->interpRefCount < 0) {
+	    pkgPtr->interpRefCount = 0;
+	}
+    }
+    trustedRefCount = pkgPtr->interpRefCount;
+    safeRefCount = pkgPtr->safeInterpRefCount;
+    Tcl_MutexUnlock(&packageMutex);
+
+    code = TCL_OK;
+    if (pkgPtr->safeInterpRefCount <= 0 && pkgPtr->interpRefCount <= 0
+	    && !keepLibrary) {
+	/*
+	 * Unload the shared library from the application memory...
+	 */
+
+#if defined(TCL_UNLOAD_DLLS) || defined(_WIN32)
+	/*
+	 * Some Unix dlls are poorly behaved - registering things like atexit
+	 * calls that can't be unregistered. If you unload such dlls, you get
+	 * a core on exit because it wants to call a function in the dll after
+	 * it's been unloaded.
+	 */
+
+	if (pkgPtr->fileName[0] != '\0') {
+	    Tcl_MutexLock(&packageMutex);
+	    if (Tcl_FSUnloadFile(interp, pkgPtr->loadHandle) == TCL_OK) {
+		/*
+		 * Remove this library from the loaded library cache.
+		 */
+
+		defaultPtr = pkgPtr;
+		if (defaultPtr == firstPackagePtr) {
+		    firstPackagePtr = pkgPtr->nextPtr;
+		} else {
+		    for (pkgPtr = firstPackagePtr; pkgPtr != NULL;
+			    pkgPtr = pkgPtr->nextPtr) {
+			if (pkgPtr->nextPtr == defaultPtr) {
+			    pkgPtr->nextPtr = defaultPtr->nextPtr;
+			    break;
+			}
+		    }
+		}
+
+		/*
+		 * Remove this library from the interpreter's library cache.
+		 */
+
+		ipFirstPtr = Tcl_GetAssocData(target, "tclLoad", NULL);
+		ipPtr = ipFirstPtr;
+		if (ipPtr->pkgPtr == defaultPtr) {
+		    ipFirstPtr = ipFirstPtr->nextPtr;
+		} else {
+		    InterpPackage *ipPrevPtr;
+
+		    for (ipPrevPtr = ipPtr; ipPtr != NULL;
+			    ipPrevPtr = ipPtr, ipPtr = ipPtr->nextPtr) {
+			if (ipPtr->pkgPtr == pkgPtr) {
+			    ipPrevPtr->nextPtr = ipPtr->nextPtr;
+			    break;
+			}
+		    }
+		}
+		Tcl_SetAssocData(target, "tclLoad", LoadCleanupProc,
+			ipFirstPtr);
+		ckfree(defaultPtr->fileName);
+		ckfree(defaultPtr->packageName);
+		ckfree(defaultPtr);
+		ckfree(ipPtr);
+		Tcl_MutexUnlock(&packageMutex);
+	    } else {
+		code = TCL_ERROR;
+	    }
+	}
+#else
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"file \"%s\" cannot be unloaded: unloading disabled",
+		fullFileName));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "UNLOAD", "DISABLED",
+		NULL);
+	code = TCL_ERROR;
+#endif
+    }
+
+  done:
+    Tcl_DStringFree(&pkgName);
+    Tcl_DStringFree(&tmp);
+    if (!complain && (code != TCL_OK)) {
+	code = TCL_OK;
+	Tcl_ResetResult(interp);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_StaticPackage --
+ *
+ *	This function is invoked to indicate that a particular package has
+ *	been linked statically with an application.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Once this function completes, the package becomes loadable via the
+ *	"load" command with an empty file name.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_StaticPackage(
+    Tcl_Interp *interp,		/* If not NULL, it means that the package has
+				 * already been loaded into the given
+				 * interpreter by calling the appropriate init
+				 * proc. */
+    const char *pkgName,	/* Name of package (must be properly
+				 * capitalized: first letter upper case,
+				 * others lower case). */
+    Tcl_PackageInitProc *initProc,
+				/* Function to call to incorporate this
+				 * package into a trusted interpreter. */
+    Tcl_PackageInitProc *safeInitProc)
+				/* Function to call to incorporate this
+				 * package into a safe interpreter (one that
+				 * will execute untrusted scripts). NULL means
+				 * the package can't be used in safe
+				 * interpreters. */
+{
+    LoadedPackage *pkgPtr;
+    InterpPackage *ipPtr, *ipFirstPtr;
+
+    /*
+     * Check to see if someone else has already reported this package as
+     * statically loaded in the process.
+     */
+
+    Tcl_MutexLock(&packageMutex);
+    for (pkgPtr = firstPackagePtr; pkgPtr != NULL; pkgPtr = pkgPtr->nextPtr) {
+	if ((pkgPtr->initProc == initProc)
+		&& (pkgPtr->safeInitProc == safeInitProc)
+		&& (strcmp(pkgPtr->packageName, pkgName) == 0)) {
+	    break;
+	}
+    }
+    Tcl_MutexUnlock(&packageMutex);
+
+    /*
+     * If the package is not yet recorded as being loaded statically, add it
+     * to the list now.
+     */
+
+    if (pkgPtr == NULL) {
+	pkgPtr = ckalloc(sizeof(LoadedPackage));
+	pkgPtr->fileName	= ckalloc(1);
+	pkgPtr->fileName[0]	= 0;
+	pkgPtr->packageName	= ckalloc(strlen(pkgName) + 1);
+	strcpy(pkgPtr->packageName, pkgName);
+	pkgPtr->loadHandle	= NULL;
+	pkgPtr->initProc	= initProc;
+	pkgPtr->safeInitProc	= safeInitProc;
+	Tcl_MutexLock(&packageMutex);
+	pkgPtr->nextPtr		= firstPackagePtr;
+	firstPackagePtr		= pkgPtr;
+	Tcl_MutexUnlock(&packageMutex);
+    }
+
+    if (interp != NULL) {
+
+	/*
+	 * If we're loading the package into an interpreter, determine whether
+	 * it's already loaded.
+	 */
+
+	ipFirstPtr = Tcl_GetAssocData(interp, "tclLoad", NULL);
+	for (ipPtr = ipFirstPtr; ipPtr != NULL; ipPtr = ipPtr->nextPtr) {
+	    if (ipPtr->pkgPtr == pkgPtr) {
+		return;
+	    }
+	}
+
+	/*
+	 * Package isn't loaded in the current interp yet. Mark it as now being
+	 * loaded.
+	 */
+
+	ipPtr = ckalloc(sizeof(InterpPackage));
+	ipPtr->pkgPtr = pkgPtr;
+	ipPtr->nextPtr = ipFirstPtr;
+	Tcl_SetAssocData(interp, "tclLoad", LoadCleanupProc, ipPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetLoadedPackages, TclGetLoadedPackagesEx --
+ *
+ *	This function returns information about all of the files that are
+ *	loaded (either in a particular interpreter, or for all interpreters).
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code. If successful, a
+ *	list of lists is placed in the interp's result. Each sublist
+ *	corresponds to one loaded file; its first element is the name of the
+ *	file (or an empty string for something that's statically loaded) and
+ *	the second element is the name of the package in that file.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetLoadedPackages(
+    Tcl_Interp *interp,		/* Interpreter in which to return information
+				 * or error message. */
+    const char *targetName)	/* Name of target interpreter or NULL. If
+				 * NULL, return info about all interps;
+				 * otherwise, just return info about this
+				 * interpreter. */
+{
+    return TclGetLoadedPackagesEx(interp, targetName, NULL);
+}
+
+int
+TclGetLoadedPackagesEx(
+    Tcl_Interp *interp,		/* Interpreter in which to return information
+				 * or error message. */
+    const char *targetName,	/* Name of target interpreter or NULL. If
+				 * NULL, return info about all interps;
+				 * otherwise, just return info about this
+				 * interpreter. */
+    const char *packageName)	/* Package name or NULL. If NULL, return info
+				 * for all packages.
+				 */
+{
+    Tcl_Interp *target;
+    LoadedPackage *pkgPtr;
+    InterpPackage *ipPtr;
+    Tcl_Obj *resultObj, *pkgDesc[2];
+
+    if (targetName == NULL) {
+	resultObj = Tcl_NewObj();
+	Tcl_MutexLock(&packageMutex);
+	for (pkgPtr = firstPackagePtr; pkgPtr != NULL;
+		pkgPtr = pkgPtr->nextPtr) {
+	    pkgDesc[0] = Tcl_NewStringObj(pkgPtr->fileName, -1);
+	    pkgDesc[1] = Tcl_NewStringObj(pkgPtr->packageName, -1);
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewListObj(2, pkgDesc));
+	}
+	Tcl_MutexUnlock(&packageMutex);
+	Tcl_SetObjResult(interp, resultObj);
+	return TCL_OK;
+    }
+
+    target = Tcl_GetSlave(interp, targetName);
+    if (target == NULL) {
+	return TCL_ERROR;
+    }
+    ipPtr = Tcl_GetAssocData(target, "tclLoad", NULL);
+
+    /*
+     * Return information about all of the available packages.
+     */
+    if (packageName) {
+	resultObj = NULL;
+
+	for (; ipPtr != NULL; ipPtr = ipPtr->nextPtr) {
+	    pkgPtr = ipPtr->pkgPtr;
+
+	    if (!strcmp(packageName, pkgPtr->packageName)) {
+		resultObj = Tcl_NewStringObj(pkgPtr->fileName, -1);
+		break;
+	    }
+	}
+
+	if (resultObj) {
+	    Tcl_SetObjResult(interp, resultObj);
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * Return information about only the packages that are loaded in a given
+     * interpreter.
+     */
+
+    resultObj = Tcl_NewObj();
+    for (; ipPtr != NULL; ipPtr = ipPtr->nextPtr) {
+	pkgPtr = ipPtr->pkgPtr;
+	pkgDesc[0] = Tcl_NewStringObj(pkgPtr->fileName, -1);
+	pkgDesc[1] = Tcl_NewStringObj(pkgPtr->packageName, -1);
+	Tcl_ListObjAppendElement(NULL, resultObj, Tcl_NewListObj(2, pkgDesc));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * LoadCleanupProc --
+ *
+ *	This function is called to delete all of the InterpPackage structures
+ *	for an interpreter when the interpreter is deleted. It gets invoked
+ *	via the Tcl AssocData mechanism.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Storage for all of the InterpPackage functions for interp get deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+LoadCleanupProc(
+    ClientData clientData,	/* Pointer to first InterpPackage structure
+				 * for interp. */
+    Tcl_Interp *interp)		/* Interpreter that is being deleted. */
+{
+    InterpPackage *ipPtr, *nextPtr;
+
+    ipPtr = clientData;
+    while (ipPtr != NULL) {
+	nextPtr = ipPtr->nextPtr;
+	ckfree(ipPtr);
+	ipPtr = nextPtr;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeLoad --
+ *
+ *	This function is invoked just before the application exits. It frees
+ *	all of the LoadedPackage structures.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeLoad(void)
+{
+    LoadedPackage *pkgPtr;
+
+    /*
+     * No synchronization here because there should just be one thread alive
+     * at this point. Logically, packageMutex should be grabbed at this point,
+     * but the Mutexes get finalized before the call to this routine. The only
+     * subsystem left alive at this point is the memory allocator.
+     */
+
+    while (firstPackagePtr != NULL) {
+	pkgPtr = firstPackagePtr;
+	firstPackagePtr = pkgPtr->nextPtr;
+
+#if defined(TCL_UNLOAD_DLLS) || defined(_WIN32)
+	/*
+	 * Some Unix dlls are poorly behaved - registering things like atexit
+	 * calls that can't be unregistered. If you unload such dlls, you get
+	 * a core on exit because it wants to call a function in the dll after
+	 * it has been unloaded.
+	 */
+
+	if (pkgPtr->fileName[0] != '\0') {
+	    Tcl_FSUnloadFile(NULL, pkgPtr->loadHandle);
+	}
+#endif
+
+	ckfree(pkgPtr->fileName);
+	ckfree(pkgPtr->packageName);
+	ckfree(pkgPtr);
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclLoadNone.c b/generic/tclLoadNone.c
new file mode 100644
index 0000000..6af5c4f
--- /dev/null
+++ b/generic/tclLoadNone.c
@@ -0,0 +1,129 @@
+/*
+ * tclLoadNone.c --
+ *
+ *	This procedure provides a version of the TclpDlopen for use in
+ *	systems that don't support dynamic loading; it just returns an error.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpDlopen --
+ *
+ *	This procedure is called to carry out dynamic loading of binary code;
+ *	it is intended for use only on systems that don't support dynamic
+ *	loading (it returns an error).
+ *
+ * Results:
+ *	The result is TCL_ERROR, and an error message is left in the interp's
+ *	result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclpDlopen(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Obj *pathPtr,		/* Name of the file containing the desired
+				 * code (UTF-8). */
+    Tcl_LoadHandle *loadHandle,	/* Filled with token for dynamically loaded
+				 * file which will be passed back to
+				 * (*unloadProcPtr)() to unload the file. */
+    Tcl_FSUnloadFileProc **unloadProcPtr,
+				/* Filled with address of Tcl_FSUnloadFileProc
+				 * function which should be used for this
+				 * file. */
+    int flags)
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"dynamic loading is not currently available on this system",
+		-1));
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGuessPackageName --
+ *
+ *	If the "load" command is invoked without providing a package name,
+ *	this procedure is invoked to try to figure it out.
+ *
+ * Results:
+ *	Always returns 0 to indicate that we couldn't figure out a package
+ *	name; generic code will then try to guess the package from the file
+ *	name. A return value of 1 would have meant that we figured out the
+ *	package name and put it in bufPtr.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGuessPackageName(
+    const char *fileName,	/* Name of file containing package (already
+				 * translated to local form if needed). */
+    Tcl_DString *bufPtr)	/* Initialized empty dstring. Append package
+				 * name to this if possible. */
+{
+    return 0;
+}
+
+/*
+ * These functions are fallbacks if we somehow determine that the platform can
+ * do loading from memory but the user wishes to disable it. They just report
+ * (gracefully) that they fail.
+ */
+
+#ifdef TCL_LOAD_FROM_MEMORY
+
+MODULE_SCOPE void *
+TclpLoadMemoryGetBuffer(
+    Tcl_Interp *interp,		/* Dummy: unused by this implementation */
+    int size)			/* Dummy: unused by this implementation */
+{
+    return NULL;
+}
+
+MODULE_SCOPE int
+TclpLoadMemory(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    void *buffer,		/* Dummy: unused by this implementation */
+    int size,			/* Dummy: unused by this implementation */
+    int codeSize,		/* Dummy: unused by this implementation */
+    Tcl_LoadHandle *loadHandle,	/* Dummy: unused by this implementation */
+    Tcl_FSUnloadFileProc **unloadProcPtr,
+				/* Dummy: unused by this implementation */
+    int flags)
+				/* Dummy: unused by this implementation */
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("dynamic loading from memory "
+		"is not available on this system", -1));
+    }
+    return TCL_ERROR;
+}
+
+#endif /* TCL_LOAD_FROM_MEMORY */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclMain.c b/generic/tclMain.c
new file mode 100644
index 0000000..f89bd5e
--- /dev/null
+++ b/generic/tclMain.c
@@ -0,0 +1,950 @@
+/*
+ * tclMain.c --
+ *
+ *	Main program for Tcl shells and other Tcl-based applications.
+ *	This file contains a generic main program for Tcl shells and other
+ *	Tcl-based applications. It can be used as-is for many applications,
+ *	just by supplying a different appInitProc function for each specific
+ *	application. Or, it can be used as a template for creating new main
+ *	programs for Tcl applications.
+ *
+ * Copyright (c) 1988-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 2000 Ajuba Solutions.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/*
+ * On Windows, this file needs to be compiled twice, once with TCL_ASCII_MAIN
+ * defined. This way both Tcl_Main and Tcl_MainExW can be implemented, sharing
+ * the same source code.
+ */
+
+#if defined(TCL_ASCII_MAIN)
+#   ifdef UNICODE
+#	undef UNICODE
+#	undef _UNICODE
+#   else
+#	define UNICODE
+#	define _UNICODE
+#   endif
+#endif
+
+#include "tclInt.h"
+
+/*
+ * The default prompt used when the user has not overridden it.
+ */
+
+#define DEFAULT_PRIMARY_PROMPT	"% "
+
+/*
+ * This file can be compiled on Windows in UNICODE mode, as well as on all
+ * other platforms using the native encoding. This is done by using the normal
+ * Windows functions like _tcscmp, but on platforms which don't have <tchar.h>
+ * we have to translate that to strcmp here.
+ */
+
+#ifndef _WIN32
+#   define TCHAR char
+#   define TEXT(arg) arg
+#   define _tcscmp strcmp
+#endif
+
+/*
+ * Further on, in UNICODE mode we just use Tcl_NewUnicodeObj, otherwise
+ * NewNativeObj is needed (which provides proper conversion from native
+ * encoding to UTF-8).
+ */
+
+#ifdef UNICODE
+#   define NewNativeObj Tcl_NewUnicodeObj
+#else /* !UNICODE */
+static inline Tcl_Obj *
+NewNativeObj(
+    char *string,
+    int length)
+{
+    Tcl_DString ds;
+
+    Tcl_ExternalToUtfDString(NULL, string, length, &ds);
+    return TclDStringToObj(&ds);
+}
+#endif /* !UNICODE */
+
+/*
+ * Declarations for various library functions and variables (don't want to
+ * include tclPort.h here, because people might copy this file out of the Tcl
+ * source directory to make their own modified versions).
+ */
+
+#if defined _MSC_VER && _MSC_VER < 1900
+/* isatty is always defined on MSVC 14.0, but not necessarily as CRTIMPORT. */
+extern CRTIMPORT int	isatty(int fd);
+#endif
+
+/*
+ * The thread-local variables for this file's functions.
+ */
+
+typedef struct {
+    Tcl_Obj *path;		/* The filename of the script for *_Main()
+				 * routines to [source] as a startup script,
+				 * or NULL for none set, meaning enter
+				 * interactive mode. */
+    Tcl_Obj *encoding;		/* The encoding of the startup script file. */
+    Tcl_MainLoopProc *mainLoopProc;
+				/* Any installed main loop handler. The main
+				 * extension that installs these is Tk. */
+} ThreadSpecificData;
+
+/*
+ * Structure definition for information used to keep the state of an
+ * interactive command processor that reads lines from standard input and
+ * writes prompts and results to standard output.
+ */
+
+typedef enum {
+    PROMPT_NONE,		/* Print no prompt */
+    PROMPT_START,		/* Print prompt for command start */
+    PROMPT_CONTINUE		/* Print prompt for command continuation */
+} PromptType;
+
+typedef struct {
+    Tcl_Channel input;		/* The standard input channel from which lines
+				 * are read. */
+    int tty;			/* Non-zero means standard input is a
+				 * terminal-like device. Zero means it's a
+				 * file. */
+    Tcl_Obj *commandPtr;	/* Used to assemble lines of input into Tcl
+				 * commands. */
+    PromptType prompt;		/* Next prompt to print */
+    Tcl_Interp *interp;		/* Interpreter that evaluates interactive
+				 * commands. */
+} InteractiveState;
+
+/*
+ * Forward declarations for functions defined later in this file.
+ */
+
+MODULE_SCOPE Tcl_MainLoopProc *TclGetMainLoop(void);
+static void		Prompt(Tcl_Interp *interp, InteractiveState *isPtr);
+static void		StdinProc(ClientData clientData, int mask);
+static void		FreeMainInterp(ClientData clientData);
+
+#ifndef TCL_ASCII_MAIN
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetStartupScript --
+ *
+ *	Sets the path and encoding of the startup script to be evaluated by
+ *	Tcl_Main, used to override the command line processing.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetStartupScript(
+    Tcl_Obj *path,		/* Filesystem path of startup script file */
+    const char *encoding)	/* Encoding of the data in that file */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    Tcl_Obj *newEncoding = NULL;
+
+    if (encoding != NULL) {
+	newEncoding = Tcl_NewStringObj(encoding, -1);
+    }
+
+    if (tsdPtr->path != NULL) {
+	Tcl_DecrRefCount(tsdPtr->path);
+    }
+    tsdPtr->path = path;
+    if (tsdPtr->path != NULL) {
+	Tcl_IncrRefCount(tsdPtr->path);
+    }
+
+    if (tsdPtr->encoding != NULL) {
+	Tcl_DecrRefCount(tsdPtr->encoding);
+    }
+    tsdPtr->encoding = newEncoding;
+    if (tsdPtr->encoding != NULL) {
+	Tcl_IncrRefCount(tsdPtr->encoding);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetStartupScript --
+ *
+ *	Gets the path and encoding of the startup script to be evaluated by
+ *	Tcl_Main.
+ *
+ * Results:
+ *	The path of the startup script; NULL if none has been set.
+ *
+ * Side effects:
+ *	If encodingPtr is not NULL, stores a (const char *) in it pointing to
+ *	the encoding name registered for the startup script. Tcl retains
+ *	ownership of the string, and may free it. Caller should make a copy
+ *	for long-term use.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetStartupScript(
+    const char **encodingPtr)	/* When not NULL, points to storage for the
+				 * (const char *) that points to the
+				 * registered encoding name for the startup
+				 * script. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (encodingPtr != NULL) {
+	if (tsdPtr->encoding == NULL) {
+	    *encodingPtr = NULL;
+	} else {
+	    *encodingPtr = Tcl_GetString(tsdPtr->encoding);
+	}
+    }
+    return tsdPtr->path;
+}
+
+/*----------------------------------------------------------------------
+ *
+ * Tcl_SourceRCFile --
+ *
+ *	This function is typically invoked by Tcl_Main of Tk_Main function to
+ *	source an application specific rc file into the interpreter at startup
+ *	time.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on what's in the rc script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SourceRCFile(
+    Tcl_Interp *interp)		/* Interpreter to source rc file into. */
+{
+    Tcl_DString temp;
+    const char *fileName;
+    Tcl_Channel chan;
+
+    fileName = Tcl_GetVar2(interp, "tcl_rcFileName", NULL, TCL_GLOBAL_ONLY);
+    if (fileName != NULL) {
+	Tcl_Channel c;
+	const char *fullName;
+
+	Tcl_DStringInit(&temp);
+	fullName = Tcl_TranslateFileName(interp, fileName, &temp);
+	if (fullName == NULL) {
+	    /*
+	     * Couldn't translate the file name (e.g. it referred to a bogus
+	     * user or there was no HOME environment variable). Just do
+	     * nothing.
+	     */
+	} else {
+	    /*
+	     * Test for the existence of the rc file before trying to read it.
+	     */
+
+	    c = Tcl_OpenFileChannel(NULL, fullName, "r", 0);
+	    if (c != NULL) {
+		Tcl_Obj *fullNameObj = Tcl_NewStringObj(fullName, -1);
+
+		Tcl_Close(NULL, c);
+		Tcl_IncrRefCount(fullNameObj);
+		if (Tcl_FSEvalFileEx(interp, fullNameObj, NULL) != TCL_OK) {
+		    chan = Tcl_GetStdChannel(TCL_STDERR);
+		    if (chan) {
+			Tcl_WriteObj(chan, Tcl_GetObjResult(interp));
+			Tcl_WriteChars(chan, "\n", 1);
+		    }
+		}
+		Tcl_DecrRefCount(fullNameObj);
+	    }
+	}
+	Tcl_DStringFree(&temp);
+    }
+}
+#endif /* !TCL_ASCII_MAIN */
+
+/*----------------------------------------------------------------------
+ *
+ * Tcl_MainEx --
+ *
+ *	Main program for tclsh and most other Tcl-based applications.
+ *
+ * Results:
+ *	None. This function never returns (it exits the process when it's
+ *	done).
+ *
+ * Side effects:
+ *	This function initializes the Tcl world and then starts interpreting
+ *	commands; almost anything could happen, depending on the script being
+ *	interpreted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_MainEx(
+    int argc,			/* Number of arguments. */
+    TCHAR **argv,		/* Array of argument strings. */
+    Tcl_AppInitProc *appInitProc,
+				/* Application-specific initialization
+				 * function to call after most initialization
+				 * but before starting to execute commands. */
+    Tcl_Interp *interp)
+{
+    Tcl_Obj *path, *resultPtr, *argvPtr, *appName;
+    const char *encodingName = NULL;
+    int code, exitCode = 0;
+    Tcl_MainLoopProc *mainLoopProc;
+    Tcl_Channel chan;
+    InteractiveState is;
+
+    TclpSetInitialEncodings();
+    TclpFindExecutable((const char *)argv[0]);
+
+    Tcl_InitMemory(interp);
+
+    is.interp = interp;
+    is.prompt = PROMPT_START;
+    is.commandPtr = Tcl_NewObj();
+
+    /*
+     * If the application has not already set a startup script, parse the
+     * first few command line arguments to determine the script path and
+     * encoding.
+     */
+
+    if (NULL == Tcl_GetStartupScript(NULL)) {
+	/*
+	 * Check whether first 3 args (argv[1] - argv[3]) look like
+	 *  -encoding ENCODING FILENAME
+	 * or like
+	 *  FILENAME
+	 */
+
+	if ((argc > 3) && (0 == _tcscmp(TEXT("-encoding"), argv[1]))
+		&& ('-' != argv[3][0])) {
+	    Tcl_Obj *value = NewNativeObj(argv[2], -1);
+	    Tcl_SetStartupScript(NewNativeObj(argv[3], -1),
+		    Tcl_GetString(value));
+	    Tcl_DecrRefCount(value);
+	    argc -= 3;
+	    argv += 3;
+	} else if ((argc > 1) && ('-' != argv[1][0])) {
+	    Tcl_SetStartupScript(NewNativeObj(argv[1], -1), NULL);
+	    argc--;
+	    argv++;
+	}
+    }
+
+    path = Tcl_GetStartupScript(&encodingName);
+    if (path == NULL) {
+	appName = NewNativeObj(argv[0], -1);
+    } else {
+	appName = path;
+    }
+    Tcl_SetVar2Ex(interp, "argv0", NULL, appName, TCL_GLOBAL_ONLY);
+    argc--;
+    argv++;
+
+    Tcl_SetVar2Ex(interp, "argc", NULL, Tcl_NewIntObj(argc), TCL_GLOBAL_ONLY);
+
+    argvPtr = Tcl_NewListObj(0, NULL);
+    while (argc--) {
+	Tcl_ListObjAppendElement(NULL, argvPtr, NewNativeObj(*argv++, -1));
+    }
+    Tcl_SetVar2Ex(interp, "argv", NULL, argvPtr, TCL_GLOBAL_ONLY);
+
+    /*
+     * Set the "tcl_interactive" variable.
+     */
+
+    is.tty = isatty(0);
+    Tcl_SetVar2Ex(interp, "tcl_interactive", NULL,
+	    Tcl_NewIntObj(!path && is.tty), TCL_GLOBAL_ONLY);
+
+    /*
+     * Invoke application-specific initialization.
+     */
+
+    Tcl_Preserve(interp);
+    if (appInitProc(interp) != TCL_OK) {
+	chan = Tcl_GetStdChannel(TCL_STDERR);
+	if (chan) {
+	    Tcl_WriteChars(chan,
+		    "application-specific initialization failed: ", -1);
+	    Tcl_WriteObj(chan, Tcl_GetObjResult(interp));
+	    Tcl_WriteChars(chan, "\n", 1);
+	}
+    }
+    if (Tcl_InterpDeleted(interp)) {
+	goto done;
+    }
+    if (Tcl_LimitExceeded(interp)) {
+	goto done;
+    }
+    if (TclFullFinalizationRequested()) {
+	/*
+	 * Arrange for final deletion of the main interp
+	 */
+
+	/* ARGH Munchhausen effect */
+	Tcl_CreateExitHandler(FreeMainInterp, interp);
+    }
+
+    /*
+     * Invoke the script specified on the command line, if any. Must fetch it
+     * again, as the appInitProc might have reset it.
+     */
+
+    path = Tcl_GetStartupScript(&encodingName);
+    if (path != NULL) {
+	Tcl_ResetResult(interp);
+	code = Tcl_FSEvalFileEx(interp, path, encodingName);
+	if (code != TCL_OK) {
+	    chan = Tcl_GetStdChannel(TCL_STDERR);
+	    if (chan) {
+		Tcl_Obj *options = Tcl_GetReturnOptions(interp, code);
+		Tcl_Obj *keyPtr, *valuePtr;
+
+		TclNewLiteralStringObj(keyPtr, "-errorinfo");
+		Tcl_IncrRefCount(keyPtr);
+		Tcl_DictObjGet(NULL, options, keyPtr, &valuePtr);
+		Tcl_DecrRefCount(keyPtr);
+
+		if (valuePtr) {
+		    Tcl_WriteObj(chan, valuePtr);
+		}
+		Tcl_WriteChars(chan, "\n", 1);
+		Tcl_DecrRefCount(options);
+	    }
+	    exitCode = 1;
+	}
+	goto done;
+    }
+
+    /*
+     * We're running interactively. Source a user-specific startup file if the
+     * application specified one and if the file exists.
+     */
+
+    Tcl_SourceRCFile(interp);
+    if (Tcl_LimitExceeded(interp)) {
+	goto done;
+    }
+
+    /*
+     * Process commands from stdin until there's an end-of-file. Note that we
+     * need to fetch the standard channels again after every eval, since they
+     * may have been changed.
+     */
+
+    Tcl_IncrRefCount(is.commandPtr);
+
+    /*
+     * Get a new value for tty if anyone writes to ::tcl_interactive
+     */
+
+    Tcl_LinkVar(interp, "tcl_interactive", (char *) &is.tty, TCL_LINK_BOOLEAN);
+    is.input = Tcl_GetStdChannel(TCL_STDIN);
+    while ((is.input != NULL) && !Tcl_InterpDeleted(interp)) {
+	mainLoopProc = TclGetMainLoop();
+	if (mainLoopProc == NULL) {
+	    int length;
+
+	    if (is.tty) {
+		Prompt(interp, &is);
+		if (Tcl_InterpDeleted(interp)) {
+		    break;
+		}
+		if (Tcl_LimitExceeded(interp)) {
+		    break;
+		}
+		is.input = Tcl_GetStdChannel(TCL_STDIN);
+		if (is.input == NULL) {
+		    break;
+		}
+	    }
+	    if (Tcl_IsShared(is.commandPtr)) {
+		Tcl_DecrRefCount(is.commandPtr);
+		is.commandPtr = Tcl_DuplicateObj(is.commandPtr);
+		Tcl_IncrRefCount(is.commandPtr);
+	    }
+	    length = Tcl_GetsObj(is.input, is.commandPtr);
+	    if (length < 0) {
+		if (Tcl_InputBlocked(is.input)) {
+		    /*
+		     * This can only happen if stdin has been set to
+		     * non-blocking. In that case cycle back and try again.
+		     * This sets up a tight polling loop (since we have no
+		     * event loop running). If this causes bad CPU hogging, we
+		     * might try toggling the blocking on stdin instead.
+		     */
+
+		    continue;
+		}
+
+		/*
+		 * Either EOF, or an error on stdin; we're done
+		 */
+
+		break;
+	    }
+
+	    /*
+	     * Add the newline removed by Tcl_GetsObj back to the string. Have
+	     * to add it back before testing completeness, because it can make
+	     * a difference. [Bug 1775878]
+	     */
+
+	    if (Tcl_IsShared(is.commandPtr)) {
+		Tcl_DecrRefCount(is.commandPtr);
+		is.commandPtr = Tcl_DuplicateObj(is.commandPtr);
+		Tcl_IncrRefCount(is.commandPtr);
+	    }
+	    Tcl_AppendToObj(is.commandPtr, "\n", 1);
+	    if (!TclObjCommandComplete(is.commandPtr)) {
+		is.prompt = PROMPT_CONTINUE;
+		continue;
+	    }
+
+	    is.prompt = PROMPT_START;
+
+	    /*
+	     * The final newline is syntactically redundant, and causes some
+	     * error messages troubles deeper in, so lop it back off.
+	     */
+
+	    TclGetStringFromObj(is.commandPtr, &length);
+	    Tcl_SetObjLength(is.commandPtr, --length);
+	    code = Tcl_RecordAndEvalObj(interp, is.commandPtr,
+		    TCL_EVAL_GLOBAL);
+	    is.input = Tcl_GetStdChannel(TCL_STDIN);
+	    Tcl_DecrRefCount(is.commandPtr);
+	    is.commandPtr = Tcl_NewObj();
+	    Tcl_IncrRefCount(is.commandPtr);
+	    if (code != TCL_OK) {
+		chan = Tcl_GetStdChannel(TCL_STDERR);
+		if (chan) {
+		    Tcl_WriteObj(chan, Tcl_GetObjResult(interp));
+		    Tcl_WriteChars(chan, "\n", 1);
+		}
+	    } else if (is.tty) {
+		resultPtr = Tcl_GetObjResult(interp);
+		Tcl_IncrRefCount(resultPtr);
+		TclGetStringFromObj(resultPtr, &length);
+		chan = Tcl_GetStdChannel(TCL_STDOUT);
+		if ((length > 0) && chan) {
+		    Tcl_WriteObj(chan, resultPtr);
+		    Tcl_WriteChars(chan, "\n", 1);
+		}
+		Tcl_DecrRefCount(resultPtr);
+	    }
+	} else {	/* (mainLoopProc != NULL) */
+	    /*
+	     * If a main loop has been defined while running interactively, we
+	     * want to start a fileevent based prompt by establishing a
+	     * channel handler for stdin.
+	     */
+
+	    if (is.input) {
+		if (is.tty) {
+		    Prompt(interp, &is);
+		}
+
+		Tcl_CreateChannelHandler(is.input, TCL_READABLE,
+			StdinProc, &is);
+	    }
+
+	    mainLoopProc();
+	    Tcl_SetMainLoop(NULL);
+
+	    if (is.input) {
+		Tcl_DeleteChannelHandler(is.input, StdinProc, &is);
+	    }
+	    is.input = Tcl_GetStdChannel(TCL_STDIN);
+	}
+
+	/*
+	 * This code here only for the (unsupported and deprecated) [checkmem]
+	 * command.
+	 */
+
+#ifdef TCL_MEM_DEBUG
+	if (tclMemDumpFileName != NULL) {
+	    Tcl_SetMainLoop(NULL);
+	    Tcl_DeleteInterp(interp);
+	}
+#endif /* TCL_MEM_DEBUG */
+    }
+
+  done:
+    mainLoopProc = TclGetMainLoop();
+    if ((exitCode == 0) && mainLoopProc && !Tcl_LimitExceeded(interp)) {
+	/*
+	 * If everything has gone OK so far, call the main loop proc, if it
+	 * exists. Packages (like Tk) can set it to start processing events at
+	 * this point.
+	 */
+
+	mainLoopProc();
+	Tcl_SetMainLoop(NULL);
+    }
+    if (is.commandPtr != NULL) {
+	Tcl_DecrRefCount(is.commandPtr);
+    }
+
+    /*
+     * Rather than calling exit, invoke the "exit" command so that users can
+     * replace "exit" with some other command to do additional cleanup on
+     * exit. The Tcl_EvalObjEx call should never return.
+     */
+
+    if (!Tcl_InterpDeleted(interp) && !Tcl_LimitExceeded(interp)) {
+	Tcl_Obj *cmd = Tcl_ObjPrintf("exit %d", exitCode);
+
+	Tcl_IncrRefCount(cmd);
+	Tcl_EvalObjEx(interp, cmd, TCL_EVAL_GLOBAL);
+	Tcl_DecrRefCount(cmd);
+    }
+
+    /*
+     * If Tcl_EvalObjEx returns, trying to eval [exit], something unusual is
+     * happening. Maybe interp has been deleted; maybe [exit] was redefined,
+     * maybe we've blown up because of an exceeded limit. We still want to
+     * cleanup and exit.
+     */
+
+    Tcl_Exit(exitCode);
+}
+
+#ifndef TCL_ASCII_MAIN
+
+/*
+ *---------------------------------------------------------------
+ *
+ * Tcl_SetMainLoop --
+ *
+ *	Sets an alternative main loop function.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	This function will be called before Tcl exits, allowing for the
+ *	creation of an event loop.
+ *
+ *---------------------------------------------------------------
+ */
+
+void
+Tcl_SetMainLoop(
+    Tcl_MainLoopProc *proc)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    tsdPtr->mainLoopProc = proc;
+}
+
+/*
+ *---------------------------------------------------------------
+ *
+ * TclGetMainLoop --
+ *
+ *	Returns the current alternative main loop function.
+ *
+ * Results:
+ *	Returns the previously defined main loop function, or NULL to indicate
+ *	that no such function has been installed and standard tclsh behaviour
+ *	(i.e., exit once the script is evaluated if not interactive) is
+ *	requested..
+ *
+ * Side effects:
+ *	None (other than possible creation of this file's TSD block).
+ *
+ *---------------------------------------------------------------
+ */
+
+Tcl_MainLoopProc *
+TclGetMainLoop(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    return tsdPtr->mainLoopProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFullFinalizationRequested --
+ *
+ *	This function returns true when either -DPURIFY is specified, or the
+ *	environment variable TCL_FINALIZE_ON_EXIT is set and not "0". This
+ *	predicate is called at places affecting the exit sequence, so that the
+ *	default behavior is a fast and deadlock-free exit, and the modified
+ *	behavior is a more thorough finalization for debugging purposes (leak
+ *	hunting etc).
+ *
+ * Results:
+ *	A boolean.
+ *
+ *----------------------------------------------------------------------
+ */
+
+MODULE_SCOPE int
+TclFullFinalizationRequested(void)
+{
+#ifdef PURIFY
+    return 1;
+#else
+    const char *fin;
+    Tcl_DString ds;
+    int finalize = 0;
+
+    fin = TclGetEnv("TCL_FINALIZE_ON_EXIT", &ds);
+    finalize = ((fin != NULL) && strcmp(fin, "0"));
+    if (fin != NULL) {
+	Tcl_DStringFree(&ds);
+    }
+    return finalize;
+#endif /* PURIFY */
+}
+#endif /* !TCL_ASCII_MAIN */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StdinProc --
+ *
+ *	This function is invoked by the event dispatcher whenever standard
+ *	input becomes readable. It grabs the next line of input characters,
+ *	adds them to a command being assembled, and executes the command if
+ *	it's complete.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Could be almost arbitrary, depending on the command that's typed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+    /* ARGSUSED */
+static void
+StdinProc(
+    ClientData clientData,	/* The state of interactive cmd line */
+    int mask)			/* Not used. */
+{
+    int code, length;
+    InteractiveState *isPtr = clientData;
+    Tcl_Channel chan = isPtr->input;
+    Tcl_Obj *commandPtr = isPtr->commandPtr;
+    Tcl_Interp *interp = isPtr->interp;
+
+    if (Tcl_IsShared(commandPtr)) {
+	Tcl_DecrRefCount(commandPtr);
+	commandPtr = Tcl_DuplicateObj(commandPtr);
+	Tcl_IncrRefCount(commandPtr);
+    }
+    length = Tcl_GetsObj(chan, commandPtr);
+    if (length < 0) {
+	if (Tcl_InputBlocked(chan)) {
+	    return;
+	}
+	if (isPtr->tty) {
+	    /*
+	     * Would be better to find a way to exit the mainLoop? Or perhaps
+	     * evaluate [exit]? Leaving as is for now due to compatibility
+	     * concerns.
+	     */
+
+	    Tcl_Exit(0);
+	}
+	Tcl_DeleteChannelHandler(chan, StdinProc, isPtr);
+	return;
+    }
+
+    if (Tcl_IsShared(commandPtr)) {
+	Tcl_DecrRefCount(commandPtr);
+	commandPtr = Tcl_DuplicateObj(commandPtr);
+	Tcl_IncrRefCount(commandPtr);
+    }
+    Tcl_AppendToObj(commandPtr, "\n", 1);
+    if (!TclObjCommandComplete(commandPtr)) {
+	isPtr->prompt = PROMPT_CONTINUE;
+	goto prompt;
+    }
+    isPtr->prompt = PROMPT_START;
+    TclGetStringFromObj(commandPtr, &length);
+    Tcl_SetObjLength(commandPtr, --length);
+
+    /*
+     * Disable the stdin channel handler while evaluating the command;
+     * otherwise if the command re-enters the event loop we might process
+     * commands from stdin before the current command is finished. Among other
+     * things, this will trash the text of the command being evaluated.
+     */
+
+    Tcl_CreateChannelHandler(chan, 0, StdinProc, isPtr);
+    code = Tcl_RecordAndEvalObj(interp, commandPtr, TCL_EVAL_GLOBAL);
+    isPtr->input = chan = Tcl_GetStdChannel(TCL_STDIN);
+    Tcl_DecrRefCount(commandPtr);
+    isPtr->commandPtr = commandPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(commandPtr);
+    if (chan != NULL) {
+	Tcl_CreateChannelHandler(chan, TCL_READABLE, StdinProc, isPtr);
+    }
+    if (code != TCL_OK) {
+	chan = Tcl_GetStdChannel(TCL_STDERR);
+
+	if (chan != NULL) {
+	    Tcl_WriteObj(chan, Tcl_GetObjResult(interp));
+	    Tcl_WriteChars(chan, "\n", 1);
+	}
+    } else if (isPtr->tty) {
+	Tcl_Obj *resultPtr = Tcl_GetObjResult(interp);
+	chan = Tcl_GetStdChannel(TCL_STDOUT);
+
+	Tcl_IncrRefCount(resultPtr);
+	TclGetStringFromObj(resultPtr, &length);
+	if ((length > 0) && (chan != NULL)) {
+	    Tcl_WriteObj(chan, resultPtr);
+	    Tcl_WriteChars(chan, "\n", 1);
+	}
+	Tcl_DecrRefCount(resultPtr);
+    }
+
+    /*
+     * If a tty stdin is still around, output a prompt.
+     */
+
+  prompt:
+    if (isPtr->tty && (isPtr->input != NULL)) {
+	Prompt(interp, isPtr);
+	isPtr->input = Tcl_GetStdChannel(TCL_STDIN);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Prompt --
+ *
+ *	Issue a prompt on standard output, or invoke a script to issue the
+ *	prompt.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	A prompt gets output, and a Tcl script may be evaluated in interp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+Prompt(
+    Tcl_Interp *interp,		/* Interpreter to use for prompting. */
+    InteractiveState *isPtr)	/* InteractiveState. Filled with PROMPT_NONE
+				 * after a prompt is printed. */
+{
+    Tcl_Obj *promptCmdPtr;
+    int code;
+    Tcl_Channel chan;
+
+    if (isPtr->prompt == PROMPT_NONE) {
+	return;
+    }
+
+    promptCmdPtr = Tcl_GetVar2Ex(interp,
+	    (isPtr->prompt==PROMPT_CONTINUE ? "tcl_prompt2" : "tcl_prompt1"),
+	    NULL, TCL_GLOBAL_ONLY);
+
+    if (Tcl_InterpDeleted(interp)) {
+	return;
+    }
+    if (promptCmdPtr == NULL) {
+    defaultPrompt:
+	if (isPtr->prompt == PROMPT_START) {
+	    chan = Tcl_GetStdChannel(TCL_STDOUT);
+	    if (chan != NULL) {
+		Tcl_WriteChars(chan, DEFAULT_PRIMARY_PROMPT,
+			strlen(DEFAULT_PRIMARY_PROMPT));
+	    }
+	}
+    } else {
+	code = Tcl_EvalObjEx(interp, promptCmdPtr, TCL_EVAL_GLOBAL);
+	if (code != TCL_OK) {
+	    Tcl_AddErrorInfo(interp,
+		    "\n    (script that generates prompt)");
+	    chan = Tcl_GetStdChannel(TCL_STDERR);
+	    if (chan != NULL) {
+		Tcl_WriteObj(chan, Tcl_GetObjResult(interp));
+		Tcl_WriteChars(chan, "\n", 1);
+	    }
+	    goto defaultPrompt;
+	}
+    }
+
+    chan = Tcl_GetStdChannel(TCL_STDOUT);
+    if (chan != NULL) {
+	Tcl_Flush(chan);
+    }
+    isPtr->prompt = PROMPT_NONE;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeMainInterp --
+ *
+ *	Exit handler used to cleanup the main interpreter and ancillary
+ *	startup script storage at exit.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeMainInterp(
+    ClientData clientData)
+{
+    Tcl_Interp *interp = clientData;
+
+    /*if (TclInExit()) return;*/
+
+    if (!Tcl_InterpDeleted(interp)) {
+	Tcl_DeleteInterp(interp);
+    }
+    Tcl_SetStartupScript(NULL, NULL);
+    Tcl_Release(interp);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclNamesp.c b/generic/tclNamesp.c
new file mode 100644
index 0000000..e1bad0e
--- /dev/null
+++ b/generic/tclNamesp.c
@@ -0,0 +1,5102 @@
+/*
+ * tclNamesp.c --
+ *
+ *	Contains support for namespaces, which provide a separate context of
+ *	commands and global variables. The global :: namespace is the
+ *	traditional Tcl "global" scope. Other namespaces are created as
+ *	children of the global namespace. These other namespaces contain
+ *	special-purpose commands and variables for packages.
+ *
+ * Copyright (c) 1993-1997 Lucent Technologies.
+ * Copyright (c) 1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 2002-2005 Donal K. Fellows.
+ * Copyright (c) 2006 Neil Madden.
+ * Contributions from Don Porter, NIST, 2007. (not subject to US copyright)
+ *
+ * Originally implemented by
+ *   Michael J. McLennan
+ *   Bell Labs Innovations for Lucent Technologies
+ *   mmclennan@lucent.com
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h" /* for TclLogCommandInfo visibility */
+
+/*
+ * Thread-local storage used to avoid having a global lock on data that is not
+ * limited to a single interpreter.
+ */
+
+typedef struct {
+    size_t numNsCreated;	/* Count of the number of namespaces created
+				 * within the thread. This value is used as a
+				 * unique id for each namespace. Cannot be
+				 * per-interp because the nsId is used to
+				 * distinguish objects which can be passed
+				 * around between interps in the same thread,
+				 * but does not need to be global because
+				 * object internal reps are always per-thread
+				 * anyway. */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * This structure contains a cached pointer to a namespace that is the result
+ * of resolving the namespace's name in some other namespace. It is the
+ * internal representation for a nsName object. It contains the pointer along
+ * with some information that is used to check the cached pointer's validity.
+ */
+
+typedef struct ResolvedNsName {
+    Namespace *nsPtr;		/* A cached pointer to the Namespace that the
+				 * name resolved to. */
+    Namespace *refNsPtr;	/* Points to the namespace context in which
+				 * the name was resolved. NULL if the name is
+				 * fully qualified and thus the resolution
+				 * does not depend on the context. */
+    size_t refCount;		/* Reference count: 1 for each nsName object
+				 * that has a pointer to this ResolvedNsName
+				 * structure as its internal rep. This
+				 * structure can be freed when refCount
+				 * becomes zero. */
+} ResolvedNsName;
+
+/*
+ * Declarations for functions local to this file:
+ */
+
+static void		DeleteImportedCmd(ClientData clientData);
+static int		DoImport(Tcl_Interp *interp,
+			    Namespace *nsPtr, Tcl_HashEntry *hPtr,
+			    const char *cmdName, const char *pattern,
+			    Namespace *importNsPtr, int allowOverwrite);
+static void		DupNsNameInternalRep(Tcl_Obj *objPtr,Tcl_Obj *copyPtr);
+static char *		ErrorCodeRead(ClientData clientData,Tcl_Interp *interp,
+			    const char *name1, const char *name2, int flags);
+static char *		ErrorInfoRead(ClientData clientData,Tcl_Interp *interp,
+			    const char *name1, const char *name2, int flags);
+static char *		EstablishErrorCodeTraces(ClientData clientData,
+			    Tcl_Interp *interp, const char *name1,
+			    const char *name2, int flags);
+static char *		EstablishErrorInfoTraces(ClientData clientData,
+			    Tcl_Interp *interp, const char *name1,
+			    const char *name2, int flags);
+static void		FreeNsNameInternalRep(Tcl_Obj *objPtr);
+static int		GetNamespaceFromObj(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr, Tcl_Namespace **nsPtrPtr);
+static int		InvokeImportedCmd(ClientData clientData,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		InvokeImportedNRCmd(ClientData clientData,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceChildrenCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceCodeCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceCurrentCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceDeleteCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceEvalCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NRNamespaceEvalCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceExistsCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceExportCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceForgetCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static void		NamespaceFree(Namespace *nsPtr);
+static int		NamespaceImportCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceInscopeCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NRNamespaceInscopeCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceOriginCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceParentCmd(ClientData dummy,Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespacePathCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceQualifiersCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceTailCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceUpvarCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		NamespaceUnknownCmd(ClientData dummy,
+			    Tcl_Interp *interp,int objc,Tcl_Obj *const objv[]);
+static int		NamespaceWhichCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		SetNsNameFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		UnlinkNsPath(Namespace *nsPtr);
+
+static Tcl_NRPostProc NsEval_Callback;
+
+/*
+ * This structure defines a Tcl object type that contains a namespace
+ * reference. It is used in commands that take the name of a namespace as an
+ * argument. The namespace reference is resolved, and the result in cached in
+ * the object.
+ */
+
+static const Tcl_ObjType nsNameType = {
+    "nsName",			/* the type's name */
+    FreeNsNameInternalRep,	/* freeIntRepProc */
+    DupNsNameInternalRep,	/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    SetNsNameFromAny		/* setFromAnyProc */
+};
+
+/*
+ * Array of values describing how to implement each standard subcommand of the
+ * "namespace" command.
+ */
+
+static const EnsembleImplMap defaultNamespaceMap[] = {
+    {"children",   NamespaceChildrenCmd, TclCompileBasic0To2ArgCmd, NULL, NULL, 0},
+    {"code",	   NamespaceCodeCmd,	TclCompileNamespaceCodeCmd, NULL, NULL, 0},
+    {"current",	   NamespaceCurrentCmd,	TclCompileNamespaceCurrentCmd, NULL, NULL, 0},
+    {"delete",	   NamespaceDeleteCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+    {"ensemble",   TclNamespaceEnsembleCmd, NULL, NULL, NULL, 0},
+    {"eval",	   NamespaceEvalCmd,	NULL, NRNamespaceEvalCmd, NULL, 0},
+    {"exists",	   NamespaceExistsCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"export",	   NamespaceExportCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+    {"forget",	   NamespaceForgetCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+    {"import",	   NamespaceImportCmd,	TclCompileBasicMin0ArgCmd, NULL, NULL, 0},
+    {"inscope",	   NamespaceInscopeCmd,	NULL, NRNamespaceInscopeCmd, NULL, 0},
+    {"origin",	   NamespaceOriginCmd,	TclCompileNamespaceOriginCmd, NULL, NULL, 0},
+    {"parent",	   NamespaceParentCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"path",	   NamespacePathCmd,	TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"qualifiers", NamespaceQualifiersCmd, TclCompileNamespaceQualifiersCmd, NULL, NULL, 0},
+    {"tail",	   NamespaceTailCmd,	TclCompileNamespaceTailCmd, NULL, NULL, 0},
+    {"unknown",	   NamespaceUnknownCmd, TclCompileBasic0Or1ArgCmd, NULL, NULL, 0},
+    {"upvar",	   NamespaceUpvarCmd,	TclCompileNamespaceUpvarCmd, NULL, NULL, 0},
+    {"which",	   NamespaceWhichCmd,	TclCompileNamespaceWhichCmd, NULL, NULL, 0},
+    {NULL, NULL, NULL, NULL, NULL, 0}
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitNamespaceSubsystem --
+ *
+ *	This function is called to initialize all the structures that are used
+ *	by namespaces on a per-process basis.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitNamespaceSubsystem(void)
+{
+    /*
+     * Does nothing for now.
+     */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCurrentNamespace --
+ *
+ *	Returns a pointer to an interpreter's currently active namespace.
+ *
+ * Results:
+ *	Returns a pointer to the interpreter's current namespace.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Namespace *
+Tcl_GetCurrentNamespace(
+    register Tcl_Interp *interp)/* Interpreter whose current namespace is
+				 * being queried. */
+{
+    return TclGetCurrentNamespace(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetGlobalNamespace --
+ *
+ *	Returns a pointer to an interpreter's global :: namespace.
+ *
+ * Results:
+ *	Returns a pointer to the specified interpreter's global namespace.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Namespace *
+Tcl_GetGlobalNamespace(
+    register Tcl_Interp *interp)/* Interpreter whose global namespace should
+				 * be returned. */
+{
+    return TclGetGlobalNamespace(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PushCallFrame --
+ *
+ *	Pushes a new call frame onto the interpreter's Tcl call stack. Called
+ *	when executing a Tcl procedure or a "namespace eval" or "namespace
+ *	inscope" command.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, or TCL_ERROR (along with an error
+ *	message in the interpreter's result object) if something goes wrong.
+ *
+ * Side effects:
+ *	Modifies the interpreter's Tcl call stack.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_PushCallFrame(
+    Tcl_Interp *interp,		/* Interpreter in which the new call frame is
+				 * to be pushed. */
+    Tcl_CallFrame *callFramePtr,/* Points to a call frame structure to push.
+				 * Storage for this has already been allocated
+				 * by the caller; typically this is the
+				 * address of a CallFrame structure allocated
+				 * on the caller's C stack. The call frame
+				 * will be initialized by this function. The
+				 * caller can pop the frame later with
+				 * Tcl_PopCallFrame, and it is responsible for
+				 * freeing the frame's storage. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace in which the frame
+				 * will execute. If NULL, the interpreter's
+				 * current namespace will be used. */
+    int isProcCallFrame)	/* If nonzero, the frame represents a called
+				 * Tcl procedure and may have local vars. Vars
+				 * will ordinarily be looked up in the frame.
+				 * If new variables are created, they will be
+				 * created in the frame. If 0, the frame is
+				 * for a "namespace eval" or "namespace
+				 * inscope" command and var references are
+				 * treated as references to namespace
+				 * variables. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register CallFrame *framePtr = (CallFrame *) callFramePtr;
+    register Namespace *nsPtr;
+
+    if (namespacePtr == NULL) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    } else {
+	nsPtr = (Namespace *) namespacePtr;
+
+	/*
+	 * TODO: Examine whether it would be better to guard based on NS_DYING
+	 * or NS_KILLED. It appears that these are not tested because they can
+	 * be set in a global interp that has been [namespace delete]d, but
+	 * which never really completely goes away because of lingering global
+	 * things like ::errorInfo and [::unknown] and hidden commands.
+	 * Review of those designs might permit stricter checking here.
+	 */
+
+	if (nsPtr->flags & NS_DEAD) {
+	    Tcl_Panic("Trying to push call frame for dead namespace");
+	    /*NOTREACHED*/
+	}
+    }
+
+    nsPtr->activationCount++;
+    framePtr->nsPtr = nsPtr;
+    framePtr->isProcCallFrame = isProcCallFrame;
+    framePtr->objc = 0;
+    framePtr->objv = NULL;
+    framePtr->callerPtr = iPtr->framePtr;
+    framePtr->callerVarPtr = iPtr->varFramePtr;
+    if (iPtr->varFramePtr != NULL) {
+	framePtr->level = (iPtr->varFramePtr->level + 1);
+    } else {
+	framePtr->level = 0;
+    }
+    framePtr->procPtr = NULL;		/* no called procedure */
+    framePtr->varTablePtr = NULL;	/* and no local variables */
+    framePtr->numCompiledLocals = 0;
+    framePtr->compiledLocals = NULL;
+    framePtr->clientData = NULL;
+    framePtr->localCachePtr = NULL;
+    framePtr->tailcallPtr = NULL;
+
+    /*
+     * Push the new call frame onto the interpreter's stack of procedure call
+     * frames making it the current frame.
+     */
+
+    iPtr->framePtr = framePtr;
+    iPtr->varFramePtr = framePtr;
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PopCallFrame --
+ *
+ *	Removes a call frame from the Tcl call stack for the interpreter.
+ *	Called to remove a frame previously pushed by Tcl_PushCallFrame.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Modifies the call stack of the interpreter. Resets various fields of
+ *	the popped call frame. If a namespace has been deleted and has no more
+ *	activations on the call stack, the namespace is destroyed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_PopCallFrame(
+    Tcl_Interp *interp)		/* Interpreter with call frame to pop. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    register CallFrame *framePtr = iPtr->framePtr;
+    Namespace *nsPtr;
+
+    /*
+     * It's important to remove the call frame from the interpreter's stack of
+     * call frames before deleting local variables, so that traces invoked by
+     * the variable deletion don't see the partially-deleted frame.
+     */
+
+    if (framePtr->callerPtr) {
+	iPtr->framePtr = framePtr->callerPtr;
+	iPtr->varFramePtr = framePtr->callerVarPtr;
+    } else {
+	/* Tcl_PopCallFrame: trying to pop rootCallFrame! */
+    }
+
+    if (framePtr->varTablePtr != NULL) {
+	TclDeleteVars(iPtr, framePtr->varTablePtr);
+	ckfree(framePtr->varTablePtr);
+	framePtr->varTablePtr = NULL;
+    }
+    if (framePtr->numCompiledLocals > 0) {
+	TclDeleteCompiledLocalVars(iPtr, framePtr);
+	if (--framePtr->localCachePtr->refCount == 0) {
+	    TclFreeLocalCache(interp, framePtr->localCachePtr);
+	}
+	framePtr->localCachePtr = NULL;
+    }
+
+    /*
+     * Decrement the namespace's count of active call frames. If the namespace
+     * is "dying" and there are no more active call frames, call
+     * Tcl_DeleteNamespace to destroy it.
+     */
+
+    nsPtr = framePtr->nsPtr;
+    nsPtr->activationCount--;
+    if ((nsPtr->flags & NS_DYING)
+	    && (nsPtr->activationCount - (nsPtr == iPtr->globalNsPtr) == 0)) {
+	Tcl_DeleteNamespace((Tcl_Namespace *) nsPtr);
+    }
+    framePtr->nsPtr = NULL;
+
+    if (framePtr->tailcallPtr) {
+	TclSetTailcall(interp, framePtr->tailcallPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPushStackFrame --
+ *
+ *	Allocates a new call frame in the interpreter's execution stack, then
+ *	pushes it onto the interpreter's Tcl call stack. Called when executing
+ *	a Tcl procedure or a "namespace eval" or "namespace inscope" command.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, or TCL_ERROR (along with an error
+ *	message in the interpreter's result object) if something goes wrong.
+ *
+ * Side effects:
+ *	Modifies the interpreter's Tcl call stack.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPushStackFrame(
+    Tcl_Interp *interp,		/* Interpreter in which the new call frame is
+				 * to be pushed. */
+    Tcl_CallFrame **framePtrPtr,/* Place to store a pointer to the stack
+				 * allocated call frame. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace in which the frame
+				 * will execute. If NULL, the interpreter's
+				 * current namespace will be used. */
+    int isProcCallFrame)	/* If nonzero, the frame represents a called
+				 * Tcl procedure and may have local vars. Vars
+				 * will ordinarily be looked up in the frame.
+				 * If new variables are created, they will be
+				 * created in the frame. If 0, the frame is
+				 * for a "namespace eval" or "namespace
+				 * inscope" command and var references are
+				 * treated as references to namespace
+				 * variables. */
+{
+    *framePtrPtr = TclStackAlloc(interp, sizeof(CallFrame));
+    return Tcl_PushCallFrame(interp, *framePtrPtr, namespacePtr,
+	    isProcCallFrame);
+}
+
+void
+TclPopStackFrame(
+    Tcl_Interp *interp)		/* Interpreter with call frame to pop. */
+{
+    CallFrame *freePtr = ((Interp *) interp)->framePtr;
+
+    Tcl_PopCallFrame(interp);
+    TclStackFree(interp, freePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EstablishErrorCodeTraces --
+ *
+ *	Creates traces on the ::errorCode variable to keep its value
+ *	consistent with the expectations of legacy code.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Read and unset traces are established on ::errorCode.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+EstablishErrorCodeTraces(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    const char *name1,
+    const char *name2,
+    int flags)
+{
+    Tcl_TraceVar2(interp, "errorCode", NULL, TCL_GLOBAL_ONLY|TCL_TRACE_READS,
+	    ErrorCodeRead, NULL);
+    Tcl_TraceVar2(interp, "errorCode", NULL, TCL_GLOBAL_ONLY|TCL_TRACE_UNSETS,
+	    EstablishErrorCodeTraces, NULL);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ErrorCodeRead --
+ *
+ *	Called when the ::errorCode variable is read. Copies the current value
+ *	of the interp's errorCode field into ::errorCode.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+ErrorCodeRead(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    const char *name1,
+    const char *name2,
+    int flags)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (Tcl_InterpDeleted(interp) || !(iPtr->flags & ERR_LEGACY_COPY)) {
+	return NULL;
+    }
+    if (iPtr->errorCode) {
+	Tcl_ObjSetVar2(interp, iPtr->ecVar, NULL,
+		iPtr->errorCode, TCL_GLOBAL_ONLY);
+	return NULL;
+    }
+    if (NULL == Tcl_ObjGetVar2(interp, iPtr->ecVar, NULL, TCL_GLOBAL_ONLY)) {
+	Tcl_ObjSetVar2(interp, iPtr->ecVar, NULL,
+		Tcl_NewObj(), TCL_GLOBAL_ONLY);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * EstablishErrorInfoTraces --
+ *
+ *	Creates traces on the ::errorInfo variable to keep its value
+ *	consistent with the expectations of legacy code.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Read and unset traces are established on ::errorInfo.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+EstablishErrorInfoTraces(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    const char *name1,
+    const char *name2,
+    int flags)
+{
+    Tcl_TraceVar2(interp, "errorInfo", NULL, TCL_GLOBAL_ONLY|TCL_TRACE_READS,
+	    ErrorInfoRead, NULL);
+    Tcl_TraceVar2(interp, "errorInfo", NULL, TCL_GLOBAL_ONLY|TCL_TRACE_UNSETS,
+	    EstablishErrorInfoTraces, NULL);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ErrorInfoRead --
+ *
+ *	Called when the ::errorInfo variable is read. Copies the current value
+ *	of the interp's errorInfo field into ::errorInfo.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+ErrorInfoRead(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    const char *name1,
+    const char *name2,
+    int flags)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (Tcl_InterpDeleted(interp) || !(iPtr->flags & ERR_LEGACY_COPY)) {
+	return NULL;
+    }
+    if (iPtr->errorInfo) {
+	Tcl_ObjSetVar2(interp, iPtr->eiVar, NULL,
+		iPtr->errorInfo, TCL_GLOBAL_ONLY);
+	return NULL;
+    }
+    if (NULL == Tcl_ObjGetVar2(interp, iPtr->eiVar, NULL, TCL_GLOBAL_ONLY)) {
+	Tcl_ObjSetVar2(interp, iPtr->eiVar, NULL,
+		Tcl_NewObj(), TCL_GLOBAL_ONLY);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateNamespace --
+ *
+ *	Creates a new namespace with the given name. If there is no active
+ *	namespace (i.e., the interpreter is being initialized), the global ::
+ *	namespace is created and returned.
+ *
+ * Results:
+ *	Returns a pointer to the new namespace if successful. If the namespace
+ *	already exists or if another error occurs, this routine returns NULL,
+ *	along with an error message in the interpreter's result object.
+ *
+ * Side effects:
+ *	If the name contains "::" qualifiers and a parent namespace does not
+ *	already exist, it is automatically created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Namespace *
+Tcl_CreateNamespace(
+    Tcl_Interp *interp,		/* Interpreter in which a new namespace is
+				 * being created. Also used for error
+				 * reporting. */
+    const char *name,		/* Name for the new namespace. May be a
+				 * qualified name with names of ancestor
+				 * namespaces separated by "::"s. */
+    ClientData clientData,	/* One-word value to store with namespace. */
+    Tcl_NamespaceDeleteProc *deleteProc)
+				/* Function called to delete client data when
+				 * the namespace is deleted. NULL if no
+				 * function should be called. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register Namespace *nsPtr, *ancestorPtr;
+    Namespace *parentPtr, *dummy1Ptr, *dummy2Ptr;
+    Namespace *globalNsPtr = iPtr->globalNsPtr;
+    const char *simpleName;
+    Tcl_HashEntry *entryPtr;
+    Tcl_DString buffer1, buffer2;
+    Tcl_DString *namePtr, *buffPtr;
+    int newEntry, nameLen;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    const char *nameStr;
+    Tcl_DString tmpBuffer;
+
+    Tcl_DStringInit(&tmpBuffer);
+
+    /*
+     * If there is no active namespace, the interpreter is being initialized.
+     */
+
+    if ((globalNsPtr == NULL) && (iPtr->varFramePtr == NULL)) {
+	/*
+	 * Treat this namespace as the global namespace, and avoid looking for
+	 * a parent.
+	 */
+
+	parentPtr = NULL;
+	simpleName = "";
+	goto doCreate;
+    }
+
+    /*
+     * Ensure that there are no trailing colons as that causes chaos when a
+     * deleteProc is specified. [Bug d614d63989]
+     */
+
+    if (deleteProc != NULL) {
+	nameStr = name + strlen(name) - 2;
+	if (nameStr >= name && nameStr[1] == ':' && nameStr[0] == ':') {
+	    Tcl_DStringAppend(&tmpBuffer, name, -1);
+	    while ((nameLen = Tcl_DStringLength(&tmpBuffer)) > 0
+		    && Tcl_DStringValue(&tmpBuffer)[nameLen-1] == ':') {
+		Tcl_DStringSetLength(&tmpBuffer, nameLen-1);
+	    }
+	    name = Tcl_DStringValue(&tmpBuffer);
+	}
+    }
+
+    /*
+     * If we've ended up with an empty string now, we're attempting to create
+     * the global namespace despite the global namespace existing. That's
+     * naughty!
+     */
+
+    if (*name == '\0') {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("can't create namespace"
+                " \"\": only global namespace can have empty name", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "NAMESPACE",
+		"CREATEGLOBAL", NULL);
+	Tcl_DStringFree(&tmpBuffer);
+	return NULL;
+    }
+
+    /*
+     * Find the parent for the new namespace.
+     */
+
+    TclGetNamespaceForQualName(interp, name, NULL, TCL_CREATE_NS_IF_UNKNOWN,
+	    &parentPtr, &dummy1Ptr, &dummy2Ptr, &simpleName);
+
+    /*
+     * If the unqualified name at the end is empty, there were trailing "::"s
+     * after the namespace's name which we ignore. The new namespace was
+     * already (recursively) created and is pointed to by parentPtr.
+     */
+
+    if (*simpleName == '\0') {
+	Tcl_DStringFree(&tmpBuffer);
+	return (Tcl_Namespace *) parentPtr;
+    }
+
+    /*
+     * Check for a bad namespace name and make sure that the name does not
+     * already exist in the parent namespace.
+     */
+
+    if (
+#ifndef BREAK_NAMESPACE_COMPAT
+	Tcl_FindHashEntry(&parentPtr->childTable, simpleName) != NULL
+#else
+	parentPtr->childTablePtr != NULL &&
+	Tcl_FindHashEntry(parentPtr->childTablePtr, simpleName) != NULL
+#endif
+    ) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create namespace \"%s\": already exists", name));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "NAMESPACE",
+		"CREATEEXISTING", NULL);
+	Tcl_DStringFree(&tmpBuffer);
+	return NULL;
+    }
+
+    /*
+     * Create the new namespace and root it in its parent. Increment the count
+     * of namespaces created.
+     */
+
+  doCreate:
+    nsPtr = ckalloc(sizeof(Namespace));
+    nameLen = strlen(simpleName) + 1;
+    nsPtr->name = ckalloc(nameLen);
+    memcpy(nsPtr->name, simpleName, nameLen);
+    nsPtr->fullName = NULL;		/* Set below. */
+    nsPtr->clientData = clientData;
+    nsPtr->deleteProc = deleteProc;
+    nsPtr->parentPtr = parentPtr;
+#ifndef BREAK_NAMESPACE_COMPAT
+    Tcl_InitHashTable(&nsPtr->childTable, TCL_STRING_KEYS);
+#else
+    nsPtr->childTablePtr = NULL;
+#endif
+    nsPtr->nsId = ++(tsdPtr->numNsCreated);
+    nsPtr->interp = interp;
+    nsPtr->flags = 0;
+    nsPtr->activationCount = 0;
+    nsPtr->refCount = 0;
+    Tcl_InitHashTable(&nsPtr->cmdTable, TCL_STRING_KEYS);
+    TclInitVarHashTable(&nsPtr->varTable, nsPtr);
+    nsPtr->exportArrayPtr = NULL;
+    nsPtr->numExportPatterns = 0;
+    nsPtr->maxExportPatterns = 0;
+    nsPtr->cmdRefEpoch = 0;
+    nsPtr->resolverEpoch = 0;
+    nsPtr->cmdResProc = NULL;
+    nsPtr->varResProc = NULL;
+    nsPtr->compiledVarResProc = NULL;
+    nsPtr->exportLookupEpoch = 0;
+    nsPtr->ensembles = NULL;
+    nsPtr->unknownHandlerPtr = NULL;
+    nsPtr->commandPathLength = 0;
+    nsPtr->commandPathArray = NULL;
+    nsPtr->commandPathSourceList = NULL;
+    nsPtr->earlyDeleteProc = NULL;
+
+    if (parentPtr != NULL) {
+	entryPtr = Tcl_CreateHashEntry(
+		TclGetNamespaceChildTable((Tcl_Namespace *) parentPtr),
+		simpleName, &newEntry);
+	Tcl_SetHashValue(entryPtr, nsPtr);
+    } else {
+	/*
+	 * In the global namespace create traces to maintain the ::errorInfo
+	 * and ::errorCode variables.
+	 */
+
+	iPtr->globalNsPtr = nsPtr;
+	EstablishErrorInfoTraces(NULL, interp, NULL, NULL, 0);
+	EstablishErrorCodeTraces(NULL, interp, NULL, NULL, 0);
+    }
+
+    /*
+     * Build the fully qualified name for this namespace.
+     */
+
+    Tcl_DStringInit(&buffer1);
+    Tcl_DStringInit(&buffer2);
+    namePtr = &buffer1;
+    buffPtr = &buffer2;
+    for (ancestorPtr = nsPtr; ancestorPtr != NULL;
+	    ancestorPtr = ancestorPtr->parentPtr) {
+	if (ancestorPtr != globalNsPtr) {
+	    register Tcl_DString *tempPtr = namePtr;
+
+	    TclDStringAppendLiteral(buffPtr, "::");
+	    Tcl_DStringAppend(buffPtr, ancestorPtr->name, -1);
+	    TclDStringAppendDString(buffPtr, namePtr);
+
+	    /*
+	     * Clear the unwanted buffer or we end up appending to previous
+	     * results, making the namespace fullNames of nested namespaces
+	     * very wrong (and strange).
+	     */
+
+	    TclDStringClear(namePtr);
+
+	    /*
+	     * Now swap the buffer pointers so that we build in the other
+	     * buffer. This is faster than repeated copying back and forth
+	     * between buffers.
+	     */
+
+	    namePtr = buffPtr;
+	    buffPtr = tempPtr;
+	}
+    }
+
+    name = Tcl_DStringValue(namePtr);
+    nameLen = Tcl_DStringLength(namePtr);
+    nsPtr->fullName = ckalloc(nameLen + 1);
+    memcpy(nsPtr->fullName, name, (unsigned) nameLen + 1);
+
+    Tcl_DStringFree(&buffer1);
+    Tcl_DStringFree(&buffer2);
+    Tcl_DStringFree(&tmpBuffer);
+
+    /*
+     * If compilation of commands originating from the parent NS is
+     * suppressed, suppress it for commands originating in this one too.
+     */
+
+    if (nsPtr->parentPtr != NULL &&
+	    nsPtr->parentPtr->flags & NS_SUPPRESS_COMPILATION) {
+	nsPtr->flags |= NS_SUPPRESS_COMPILATION;
+    }
+
+    /*
+     * Return a pointer to the new namespace.
+     */
+
+    return (Tcl_Namespace *) nsPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteNamespace --
+ *
+ *	Deletes a namespace and all of the commands, variables, and other
+ *	namespaces within it.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	When a namespace is deleted, it is automatically removed as a child of
+ *	its parent namespace. Also, all its commands, variables and child
+ *	namespaces are deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteNamespace(
+    Tcl_Namespace *namespacePtr)/* Points to the namespace to delete. */
+{
+    register Namespace *nsPtr = (Namespace *) namespacePtr;
+    Interp *iPtr = (Interp *) nsPtr->interp;
+    Namespace *globalNsPtr = (Namespace *)
+	    TclGetGlobalNamespace((Tcl_Interp *) iPtr);
+    Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch search;
+    Command *cmdPtr;
+
+    /*
+     * Give anyone interested - notably TclOO - a chance to use this namespace
+     * normally despite the fact that the namespace is going to go. Allows the
+     * calling of destructors. Will only be called once (unless re-established
+     * by the called function). [Bug 2950259]
+     *
+     * Note that setting this field requires access to the internal definition
+     * of namespaces, so it should only be accessed by code that knows about
+     * being careful with reentrancy.
+     */
+
+    if (nsPtr->earlyDeleteProc != NULL) {
+	Tcl_NamespaceDeleteProc *earlyDeleteProc = nsPtr->earlyDeleteProc;
+
+	nsPtr->earlyDeleteProc = NULL;
+	nsPtr->activationCount++;
+	earlyDeleteProc(nsPtr->clientData);
+	nsPtr->activationCount--;
+    }
+
+    /*
+     * Delete all coroutine commands now: break the circular ref cycle between
+     * the namespace and the coroutine command [Bug 2724403]. This code is
+     * essentially duplicated in TclTeardownNamespace() for all other
+     * commands. Don't optimize to Tcl_NextHashEntry() because of traces.
+     *
+     * NOTE: we could avoid traversing the ns's command list by keeping a
+     * separate list of coros.
+     */
+
+    for (entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+	    entryPtr != NULL;) {
+	cmdPtr = Tcl_GetHashValue(entryPtr);
+	if (cmdPtr->nreProc == TclNRInterpCoroutine) {
+	    Tcl_DeleteCommandFromToken((Tcl_Interp *) iPtr,
+		    (Tcl_Command) cmdPtr);
+	    entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+	} else {
+	    entryPtr = Tcl_NextHashEntry(&search);
+	}
+    }
+
+    /*
+     * If the namespace has associated ensemble commands, delete them first.
+     * This leaves the actual contents of the namespace alone (unless they are
+     * linked ensemble commands, of course). Note that this code is actually
+     * reentrant so command delete traces won't purturb things badly.
+     */
+
+    while (nsPtr->ensembles != NULL) {
+	EnsembleConfig *ensemblePtr = (EnsembleConfig *) nsPtr->ensembles;
+
+	/*
+	 * Splice out and link to indicate that we've already been killed.
+	 */
+
+	nsPtr->ensembles = (Tcl_Ensemble *) ensemblePtr->next;
+	ensemblePtr->next = ensemblePtr;
+	Tcl_DeleteCommandFromToken(nsPtr->interp, ensemblePtr->token);
+    }
+
+    /*
+     * If the namespace has a registered unknown handler (TIP 181), then free
+     * it here.
+     */
+
+    if (nsPtr->unknownHandlerPtr != NULL) {
+	Tcl_DecrRefCount(nsPtr->unknownHandlerPtr);
+	nsPtr->unknownHandlerPtr = NULL;
+    }
+
+    /*
+     * If the namespace is on the call frame stack, it is marked as "dying"
+     * (NS_DYING is OR'd into its flags): the namespace can't be looked up by
+     * name but its commands and variables are still usable by those active
+     * call frames. When all active call frames referring to the namespace
+     * have been popped from the Tcl stack, Tcl_PopCallFrame will call this
+     * function again to delete everything in the namespace. If no nsName
+     * objects refer to the namespace (i.e., if its refCount is zero), its
+     * commands and variables are deleted and the storage for its namespace
+     * structure is freed. Otherwise, if its refCount is nonzero, the
+     * namespace's commands and variables are deleted but the structure isn't
+     * freed. Instead, NS_DEAD is OR'd into the structure's flags to allow the
+     * namespace resolution code to recognize that the namespace is "deleted".
+     * The structure's storage is freed by FreeNsNameInternalRep when its
+     * refCount reaches 0.
+     */
+
+    if (nsPtr->activationCount - (nsPtr == globalNsPtr) > 0) {
+	nsPtr->flags |= NS_DYING;
+	if (nsPtr->parentPtr != NULL) {
+	    entryPtr = Tcl_FindHashEntry(
+		    TclGetNamespaceChildTable((Tcl_Namespace *)
+			    nsPtr->parentPtr), nsPtr->name);
+	    if (entryPtr != NULL) {
+		Tcl_DeleteHashEntry(entryPtr);
+	    }
+	}
+	nsPtr->parentPtr = NULL;
+    } else if (!(nsPtr->flags & NS_KILLED)) {
+	/*
+	 * Delete the namespace and everything in it. If this is the global
+	 * namespace, then clear it but don't free its storage unless the
+	 * interpreter is being torn down. Set the NS_KILLED flag to avoid
+	 * recursive calls here - if the namespace is really in the process of
+	 * being deleted, ignore any second call.
+	 */
+
+	nsPtr->flags |= (NS_DYING|NS_KILLED);
+
+	TclTeardownNamespace(nsPtr);
+
+	if ((nsPtr != globalNsPtr) || (iPtr->flags & DELETED)) {
+	    /*
+	     * If this is the global namespace, then it may have residual
+	     * "errorInfo" and "errorCode" variables for errors that occurred
+	     * while it was being torn down. Try to clear the variable list
+	     * one last time.
+	     */
+
+	    TclDeleteNamespaceVars(nsPtr);
+
+#ifndef BREAK_NAMESPACE_COMPAT
+	    Tcl_DeleteHashTable(&nsPtr->childTable);
+#else
+	    if (nsPtr->childTablePtr != NULL) {
+		Tcl_DeleteHashTable(nsPtr->childTablePtr);
+		ckfree(nsPtr->childTablePtr);
+	    }
+#endif
+	    Tcl_DeleteHashTable(&nsPtr->cmdTable);
+
+	    /*
+	     * If the reference count is 0, then discard the namespace.
+	     * Otherwise, mark it as "dead" so that it can't be used.
+	     */
+
+	    if (nsPtr->refCount == 0) {
+		NamespaceFree(nsPtr);
+	    } else {
+		nsPtr->flags |= NS_DEAD;
+	    }
+	} else {
+	    /*
+	     * Restore the ::errorInfo and ::errorCode traces.
+	     */
+
+	    EstablishErrorInfoTraces(NULL, nsPtr->interp, NULL, NULL, 0);
+	    EstablishErrorCodeTraces(NULL, nsPtr->interp, NULL, NULL, 0);
+
+	    /*
+	     * We didn't really kill it, so remove the KILLED marks, so it can
+	     * get killed later, avoiding mem leaks.
+	     */
+
+	    nsPtr->flags &= ~(NS_DYING|NS_KILLED);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTeardownNamespace --
+ *
+ *	Used internally to dismantle and unlink a namespace when it is
+ *	deleted. Divorces the namespace from its parent, and deletes all
+ *	commands, variables, and child namespaces.
+ *
+ *	This is kept separate from Tcl_DeleteNamespace so that the global
+ *	namespace can be handled specially.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes this namespace from its parent's child namespace hashtable.
+ *	Deletes all commands, variables and namespaces in this namespace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclTeardownNamespace(
+    register Namespace *nsPtr)	/* Points to the namespace to be dismantled
+				 * and unlinked from its parent. */
+{
+    Interp *iPtr = (Interp *) nsPtr->interp;
+    register Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch search;
+    int i;
+
+    /*
+     * Start by destroying the namespace's variable table, since variables
+     * might trigger traces. Variable table should be cleared but not freed!
+     * TclDeleteNamespaceVars frees it, so we reinitialize it afterwards.
+     */
+
+    TclDeleteNamespaceVars(nsPtr);
+    TclInitVarHashTable(&nsPtr->varTable, nsPtr);
+
+    /*
+     * Delete all commands in this namespace. Be careful when traversing the
+     * hash table: when each command is deleted, it removes itself from the
+     * command table. Because of traces (and the desire to avoid the quadratic
+     * problems of just using Tcl_FirstHashEntry over and over, [Bug
+     * f97d4ee020]) we copy to a temporary array and then delete all those
+     * commands.
+     */
+
+    while (nsPtr->cmdTable.numEntries > 0) {
+	int length = nsPtr->cmdTable.numEntries;
+	Command **cmds = TclStackAlloc((Tcl_Interp *) iPtr,
+		sizeof(Command *) * length);
+
+	i = 0;
+	for (entryPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+		entryPtr != NULL;
+		entryPtr = Tcl_NextHashEntry(&search)) {
+	    cmds[i] = Tcl_GetHashValue(entryPtr);
+	    cmds[i]->refCount++;
+	    i++;
+	}
+	for (i = 0 ; i < length ; i++) {
+	    Tcl_DeleteCommandFromToken((Tcl_Interp *) iPtr,
+		    (Tcl_Command) cmds[i]);
+	    TclCleanupCommandMacro(cmds[i]);
+	}
+	TclStackFree((Tcl_Interp *) iPtr, cmds);
+    }
+    Tcl_DeleteHashTable(&nsPtr->cmdTable);
+    Tcl_InitHashTable(&nsPtr->cmdTable, TCL_STRING_KEYS);
+
+    /*
+     * Remove the namespace from its parent's child hashtable.
+     */
+
+    if (nsPtr->parentPtr != NULL) {
+	entryPtr = Tcl_FindHashEntry(
+		TclGetNamespaceChildTable((Tcl_Namespace *)
+			nsPtr->parentPtr), nsPtr->name);
+	if (entryPtr != NULL) {
+	    Tcl_DeleteHashEntry(entryPtr);
+	}
+    }
+    nsPtr->parentPtr = NULL;
+
+    /*
+     * Delete the namespace path if one is installed.
+     */
+
+    if (nsPtr->commandPathLength != 0) {
+	UnlinkNsPath(nsPtr);
+	nsPtr->commandPathLength = 0;
+    }
+    if (nsPtr->commandPathSourceList != NULL) {
+	NamespacePathEntry *nsPathPtr = nsPtr->commandPathSourceList;
+
+	do {
+	    if (nsPathPtr->nsPtr != NULL && nsPathPtr->creatorNsPtr != NULL) {
+		nsPathPtr->creatorNsPtr->cmdRefEpoch++;
+	    }
+	    nsPathPtr->nsPtr = NULL;
+	    nsPathPtr = nsPathPtr->nextPtr;
+	} while (nsPathPtr != NULL);
+	nsPtr->commandPathSourceList = NULL;
+    }
+
+    /*
+     * Delete all the child namespaces.
+     *
+     * BE CAREFUL: When each child is deleted, it will divorce itself from its
+     * parent. You can't traverse a hash table properly if its elements are
+     * being deleted.  Because of traces (and the desire to avoid the
+     * quadratic problems of just using Tcl_FirstHashEntry over and over, [Bug
+     * f97d4ee020]) we copy to a temporary array and then delete all those
+     * namespaces.
+     *
+     * Important: leave the hash table itself still live.
+     */
+
+#ifndef BREAK_NAMESPACE_COMPAT
+    while (nsPtr->childTable.numEntries > 0) {
+	int length = nsPtr->childTable.numEntries;
+	Namespace **children = TclStackAlloc((Tcl_Interp *) iPtr,
+		sizeof(Namespace *) * length);
+
+	i = 0;
+	for (entryPtr = Tcl_FirstHashEntry(&nsPtr->childTable, &search);
+		entryPtr != NULL;
+		entryPtr = Tcl_NextHashEntry(&search)) {
+	    children[i] = Tcl_GetHashValue(entryPtr);
+	    children[i]->refCount++;
+	    i++;
+	}
+	for (i = 0 ; i < length ; i++) {
+	    Tcl_DeleteNamespace((Tcl_Namespace *) children[i]);
+	    TclNsDecrRefCount(children[i]);
+	}
+	TclStackFree((Tcl_Interp *) iPtr, children);
+    }
+#else
+    if (nsPtr->childTablePtr != NULL) {
+	while (nsPtr->childTablePtr->numEntries > 0) {
+	    int length = nsPtr->childTablePtr->numEntries;
+	    Namespace **children = TclStackAlloc((Tcl_Interp *) iPtr,
+		    sizeof(Namespace *) * length);
+
+	    i = 0;
+	    for (entryPtr = Tcl_FirstHashEntry(nsPtr->childTablePtr, &search);
+		    entryPtr != NULL;
+		    entryPtr = Tcl_NextHashEntry(&search)) {
+		children[i] = Tcl_GetHashValue(entryPtr);
+		children[i]->refCount++;
+		i++;
+	    }
+	    for (i = 0 ; i < length ; i++) {
+		Tcl_DeleteNamespace((Tcl_Namespace *) children[i]);
+		TclNsDecrRefCount(children[i]);
+	    }
+	    TclStackFree((Tcl_Interp *) iPtr, children);
+	}
+    }
+#endif
+
+    /*
+     * Free the namespace's export pattern array.
+     */
+
+    if (nsPtr->exportArrayPtr != NULL) {
+	for (i = 0;  i < nsPtr->numExportPatterns;  i++) {
+	    ckfree(nsPtr->exportArrayPtr[i]);
+	}
+	ckfree(nsPtr->exportArrayPtr);
+	nsPtr->exportArrayPtr = NULL;
+	nsPtr->numExportPatterns = 0;
+	nsPtr->maxExportPatterns = 0;
+    }
+
+    /*
+     * Free any client data associated with the namespace.
+     */
+
+    if (nsPtr->deleteProc != NULL) {
+	nsPtr->deleteProc(nsPtr->clientData);
+    }
+    nsPtr->deleteProc = NULL;
+    nsPtr->clientData = NULL;
+
+    /*
+     * Reset the namespace's id field to ensure that this namespace won't be
+     * interpreted as valid by, e.g., the cache validation code for cached
+     * command references in Tcl_GetCommandFromObj.
+     */
+
+    nsPtr->nsId = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceFree --
+ *
+ *	Called after a namespace has been deleted, when its reference count
+ *	reaches 0. Frees the data structure representing the namespace.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+NamespaceFree(
+    register Namespace *nsPtr)	/* Points to the namespace to free. */
+{
+    /*
+     * Most of the namespace's contents are freed when the namespace is
+     * deleted by Tcl_DeleteNamespace. All that remains is to free its names
+     * (for error messages), and the structure itself.
+     */
+
+    ckfree(nsPtr->name);
+    ckfree(nsPtr->fullName);
+    ckfree(nsPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNsDecrRefCount --
+ *
+ *	Drops a reference to a namespace and frees it if the namespace has
+ *	been deleted and the last reference has just been dropped.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclNsDecrRefCount(
+    Namespace *nsPtr)
+{
+    if ((nsPtr->refCount-- <= 1) && (nsPtr->flags & NS_DEAD)) {
+	NamespaceFree(nsPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Export --
+ *
+ *	Makes all the commands matching a pattern available to later be
+ *	imported from the namespace specified by namespacePtr (or the current
+ *	namespace if namespacePtr is NULL). The specified pattern is appended
+ *	onto the namespace's export pattern list, which is optionally cleared
+ *	beforehand.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, or TCL_ERROR (along with an error
+ *	message in the interpreter's result) if something goes wrong.
+ *
+ * Side effects:
+ *	Appends the export pattern onto the namespace's export list.
+ *	Optionally reset the namespace's export pattern list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Export(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace from which commands
+				 * are to be exported. NULL for the current
+				 * namespace. */
+    const char *pattern,	/* String pattern indicating which commands to
+				 * export. This pattern may not include any
+				 * namespace qualifiers; only commands in the
+				 * specified namespace may be exported. */
+    int resetListFirst)		/* If nonzero, resets the namespace's export
+				 * list before appending. */
+{
+#define INIT_EXPORT_PATTERNS 5
+    Namespace *nsPtr, *exportNsPtr, *dummyPtr;
+    Namespace *currNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    const char *simplePattern;
+    char *patternCpy;
+    int neededElems, len, i;
+
+    /*
+     * If the specified namespace is NULL, use the current namespace.
+     */
+
+    if (namespacePtr == NULL) {
+	nsPtr = (Namespace *) currNsPtr;
+    } else {
+	nsPtr = (Namespace *) namespacePtr;
+    }
+
+    /*
+     * If resetListFirst is true (nonzero), clear the namespace's export
+     * pattern list.
+     */
+
+    if (resetListFirst) {
+	if (nsPtr->exportArrayPtr != NULL) {
+	    for (i = 0;  i < nsPtr->numExportPatterns;  i++) {
+		ckfree(nsPtr->exportArrayPtr[i]);
+	    }
+	    ckfree(nsPtr->exportArrayPtr);
+	    nsPtr->exportArrayPtr = NULL;
+	    TclInvalidateNsCmdLookup(nsPtr);
+	    nsPtr->numExportPatterns = 0;
+	    nsPtr->maxExportPatterns = 0;
+	}
+    }
+
+    /*
+     * Check that the pattern doesn't have namespace qualifiers.
+     */
+
+    TclGetNamespaceForQualName(interp, pattern, nsPtr, TCL_NAMESPACE_ONLY,
+	    &exportNsPtr, &dummyPtr, &dummyPtr, &simplePattern);
+
+    if ((exportNsPtr != nsPtr) || (strcmp(pattern, simplePattern) != 0)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf("invalid export pattern"
+                " \"%s\": pattern can't specify a namespace", pattern));
+	Tcl_SetErrorCode(interp, "TCL", "EXPORT", "INVALID", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure that we don't already have the pattern in the array
+     */
+
+    if (nsPtr->exportArrayPtr != NULL) {
+	for (i = 0;  i < nsPtr->numExportPatterns;  i++) {
+	    if (strcmp(pattern, nsPtr->exportArrayPtr[i]) == 0) {
+		/*
+		 * The pattern already exists in the list.
+		 */
+
+		return TCL_OK;
+	    }
+	}
+    }
+
+    /*
+     * Make sure there is room in the namespace's pattern array for the new
+     * pattern.
+     */
+
+    neededElems = nsPtr->numExportPatterns + 1;
+    if (neededElems > nsPtr->maxExportPatterns) {
+	nsPtr->maxExportPatterns = nsPtr->maxExportPatterns ?
+		2 * nsPtr->maxExportPatterns : INIT_EXPORT_PATTERNS;
+	nsPtr->exportArrayPtr = ckrealloc(nsPtr->exportArrayPtr,
+		sizeof(char *) * nsPtr->maxExportPatterns);
+    }
+
+    /*
+     * Add the pattern to the namespace's array of export patterns.
+     */
+
+    len = strlen(pattern);
+    patternCpy = ckalloc(len + 1);
+    memcpy(patternCpy, pattern, (unsigned) len + 1);
+
+    nsPtr->exportArrayPtr[nsPtr->numExportPatterns] = patternCpy;
+    nsPtr->numExportPatterns++;
+
+    /*
+     * The list of commands actually exported from the namespace might have
+     * changed (probably will have!) However, we do not need to recompute this
+     * just yet; next time we need the info will be soon enough.
+     */
+
+    TclInvalidateNsCmdLookup(nsPtr);
+
+    return TCL_OK;
+#undef INIT_EXPORT_PATTERNS
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendExportList --
+ *
+ *	Appends onto the argument object the list of export patterns for the
+ *	specified namespace.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case the object
+ *	referenced by objPtr has each export pattern appended to it. If an
+ *	error occurs, TCL_ERROR is returned and the interpreter's result holds
+ *	an error message.
+ *
+ * Side effects:
+ *	If necessary, the object referenced by objPtr is converted into a list
+ *	object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AppendExportList(
+    Tcl_Interp *interp,		/* Interpreter used for error reporting. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace whose export
+				 * pattern list is appended onto objPtr. NULL
+				 * for the current namespace. */
+    Tcl_Obj *objPtr)		/* Points to the Tcl object onto which the
+				 * export pattern list is appended. */
+{
+    Namespace *nsPtr;
+    int i, result;
+
+    /*
+     * If the specified namespace is NULL, use the current namespace.
+     */
+
+    if (namespacePtr == NULL) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    } else {
+	nsPtr = (Namespace *) namespacePtr;
+    }
+
+    /*
+     * Append the export pattern list onto objPtr.
+     */
+
+    for (i = 0;  i < nsPtr->numExportPatterns;  i++) {
+	result = Tcl_ListObjAppendElement(interp, objPtr,
+		Tcl_NewStringObj(nsPtr->exportArrayPtr[i], -1));
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Import --
+ *
+ *	Imports all of the commands matching a pattern into the namespace
+ *	specified by namespacePtr (or the current namespace if contextNsPtr is
+ *	NULL). This is done by creating a new command (the "imported command")
+ *	that points to the real command in its original namespace.
+ *
+ *	If matching commands are on the autoload path but haven't been loaded
+ *	yet, this command forces them to be loaded, then creates the links to
+ *	them.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, or TCL_ERROR (along with an error
+ *	message in the interpreter's result) if something goes wrong.
+ *
+ * Side effects:
+ *	Creates new commands in the importing namespace. These indirect calls
+ *	back to the real command and are deleted if the real commands are
+ *	deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_Import(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace into which the
+				 * commands are to be imported. NULL for the
+				 * current namespace. */
+    const char *pattern,	/* String pattern indicating which commands to
+				 * import. This pattern should be qualified by
+				 * the name of the namespace from which to
+				 * import the command(s). */
+    int allowOverwrite)		/* If nonzero, allow existing commands to be
+				 * overwritten by imported commands. If 0,
+				 * return an error if an imported cmd
+				 * conflicts with an existing one. */
+{
+    Namespace *nsPtr, *importNsPtr, *dummyPtr;
+    const char *simplePattern;
+    register Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+
+    /*
+     * If the specified namespace is NULL, use the current namespace.
+     */
+
+    if (namespacePtr == NULL) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    } else {
+	nsPtr = (Namespace *) namespacePtr;
+    }
+
+    /*
+     * First, invoke the "auto_import" command with the pattern being
+     * imported. This command is part of the Tcl library. It looks for
+     * imported commands in autoloaded libraries and loads them in. That way,
+     * they will be found when we try to create links below.
+     *
+     * Note that we don't just call Tcl_EvalObjv() directly because we do not
+     * want absence of the command to be a failure case.
+     */
+
+    if (Tcl_FindCommand(interp,"auto_import",NULL,TCL_GLOBAL_ONLY) != NULL) {
+	Tcl_Obj *objv[2];
+	int result;
+
+	TclNewLiteralStringObj(objv[0], "auto_import");
+	objv[1] = Tcl_NewStringObj(pattern, -1);
+
+	Tcl_IncrRefCount(objv[0]);
+	Tcl_IncrRefCount(objv[1]);
+	result = Tcl_EvalObjv(interp, 2, objv, TCL_GLOBAL_ONLY);
+	Tcl_DecrRefCount(objv[0]);
+	Tcl_DecrRefCount(objv[1]);
+
+	if (result != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_ResetResult(interp);
+    }
+
+    /*
+     * From the pattern, find the namespace from which we are importing and
+     * get the simple pattern (no namespace qualifiers or ::'s) at the end.
+     */
+
+    if (strlen(pattern) == 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("empty import pattern",-1));
+	Tcl_SetErrorCode(interp, "TCL", "IMPORT", "EMPTY", NULL);
+	return TCL_ERROR;
+    }
+    TclGetNamespaceForQualName(interp, pattern, nsPtr, TCL_NAMESPACE_ONLY,
+	    &importNsPtr, &dummyPtr, &dummyPtr, &simplePattern);
+
+    if (importNsPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "unknown namespace in import pattern \"%s\"", pattern));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE", pattern, NULL);
+	return TCL_ERROR;
+    }
+    if (importNsPtr == nsPtr) {
+	if (pattern == simplePattern) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "no namespace specified in import pattern \"%s\"",
+                    pattern));
+	    Tcl_SetErrorCode(interp, "TCL", "IMPORT", "ORIGIN", NULL);
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "import pattern \"%s\" tries to import from namespace"
+                    " \"%s\" into itself", pattern, importNsPtr->name));
+	    Tcl_SetErrorCode(interp, "TCL", "IMPORT", "SELF", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Scan through the command table in the source namespace and look for
+     * exported commands that match the string pattern. Create an "imported
+     * command" in the current namespace for each imported command; these
+     * commands redirect their invocations to the "real" command.
+     */
+
+    if ((simplePattern != NULL) && TclMatchIsTrivial(simplePattern)) {
+	hPtr = Tcl_FindHashEntry(&importNsPtr->cmdTable, simplePattern);
+	if (hPtr == NULL) {
+	    return TCL_OK;
+	}
+	return DoImport(interp, nsPtr, hPtr, simplePattern, pattern,
+		importNsPtr, allowOverwrite);
+    }
+    for (hPtr = Tcl_FirstHashEntry(&importNsPtr->cmdTable, &search);
+	    (hPtr != NULL); hPtr = Tcl_NextHashEntry(&search)) {
+	char *cmdName = Tcl_GetHashKey(&importNsPtr->cmdTable, hPtr);
+
+	if (Tcl_StringMatch(cmdName, simplePattern) &&
+		DoImport(interp, nsPtr, hPtr, cmdName, pattern, importNsPtr,
+		allowOverwrite) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DoImport --
+ *
+ *	Import a particular command from one namespace into another. Helper
+ *	for Tcl_Import().
+ *
+ * Results:
+ *	Standard Tcl result code. If TCL_ERROR, appends an error message to
+ *	the interpreter result.
+ *
+ * Side effects:
+ *	A new command is created in the target namespace unless this is a
+ *	reimport of exactly the same command as before.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+DoImport(
+    Tcl_Interp *interp,
+    Namespace *nsPtr,
+    Tcl_HashEntry *hPtr,
+    const char *cmdName,
+    const char *pattern,
+    Namespace *importNsPtr,
+    int allowOverwrite)
+{
+    int i = 0, exported = 0;
+    Tcl_HashEntry *found;
+
+    /*
+     * The command cmdName in the source namespace matches the pattern. Check
+     * whether it was exported. If it wasn't, we ignore it.
+     */
+
+    while (!exported && (i < importNsPtr->numExportPatterns)) {
+	exported |= Tcl_StringMatch(cmdName,
+		importNsPtr->exportArrayPtr[i++]);
+    }
+    if (!exported) {
+	return TCL_OK;
+    }
+
+    /*
+     * Unless there is a name clash, create an imported command in the current
+     * namespace that refers to cmdPtr.
+     */
+
+    found = Tcl_FindHashEntry(&nsPtr->cmdTable, cmdName);
+    if ((found == NULL) || allowOverwrite) {
+	/*
+	 * Create the imported command and its client data. To create the new
+	 * command in the current namespace, generate a fully qualified name
+	 * for it.
+	 */
+
+	Tcl_DString ds;
+	Tcl_Command importedCmd;
+	ImportedCmdData *dataPtr;
+	Command *cmdPtr;
+	ImportRef *refPtr;
+
+	Tcl_DStringInit(&ds);
+	Tcl_DStringAppend(&ds, nsPtr->fullName, -1);
+	if (nsPtr != ((Interp *) interp)->globalNsPtr) {
+	    TclDStringAppendLiteral(&ds, "::");
+	}
+	Tcl_DStringAppend(&ds, cmdName, -1);
+
+	/*
+	 * Check whether creating the new imported command in the current
+	 * namespace would create a cycle of imported command references.
+	 */
+
+	cmdPtr = Tcl_GetHashValue(hPtr);
+	if (found != NULL && cmdPtr->deleteProc == DeleteImportedCmd) {
+	    Command *overwrite = Tcl_GetHashValue(found);
+	    Command *linkCmd = cmdPtr;
+
+	    while (linkCmd->deleteProc == DeleteImportedCmd) {
+		dataPtr = linkCmd->objClientData;
+		linkCmd = dataPtr->realCmdPtr;
+		if (overwrite == linkCmd) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                            "import pattern \"%s\" would create a loop"
+                            " containing command \"%s\"",
+                            pattern, Tcl_DStringValue(&ds)));
+		    Tcl_DStringFree(&ds);
+		    Tcl_SetErrorCode(interp, "TCL", "IMPORT", "LOOP", NULL);
+		    return TCL_ERROR;
+		}
+	    }
+	}
+
+	dataPtr = ckalloc(sizeof(ImportedCmdData));
+	importedCmd = Tcl_NRCreateCommand(interp, Tcl_DStringValue(&ds),
+		InvokeImportedCmd, InvokeImportedNRCmd, dataPtr,
+		DeleteImportedCmd);
+	dataPtr->realCmdPtr = cmdPtr;
+	dataPtr->selfPtr = (Command *) importedCmd;
+	dataPtr->selfPtr->compileProc = cmdPtr->compileProc;
+	Tcl_DStringFree(&ds);
+
+	/*
+	 * Create an ImportRef structure describing this new import command
+	 * and add it to the import ref list in the "real" command.
+	 */
+
+	refPtr = ckalloc(sizeof(ImportRef));
+	refPtr->importedCmdPtr = (Command *) importedCmd;
+	refPtr->nextPtr = cmdPtr->importRefPtr;
+	cmdPtr->importRefPtr = refPtr;
+    } else {
+	Command *overwrite = Tcl_GetHashValue(found);
+
+	if (overwrite->deleteProc == DeleteImportedCmd) {
+	    ImportedCmdData *dataPtr = overwrite->objClientData;
+
+	    if (dataPtr->realCmdPtr == Tcl_GetHashValue(hPtr)) {
+		/*
+		 * Repeated import of same command is acceptable.
+		 */
+
+		return TCL_OK;
+	    }
+	}
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "can't import command \"%s\": already exists", cmdName));
+	Tcl_SetErrorCode(interp, "TCL", "IMPORT", "OVERWRITE", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ForgetImport --
+ *
+ *	Deletes commands previously imported into the namespace indicated.
+ *	The by namespacePtr, or the current namespace of interp, when
+ *	namespacePtr is NULL. The pattern controls which imported commands are
+ *	deleted. A simple pattern, one without namespace separators, matches
+ *	the current command names of imported commands in the namespace.
+ *	Matching imported commands are deleted. A qualified pattern is
+ *	interpreted as deletion selection on the basis of where the command is
+ *	imported from. The original command and "first link" command for each
+ *	imported command are determined, and they are matched against the
+ *	pattern. A match leads to deletion of the imported command.
+ *
+ * Results:
+ *	Returns TCL_ERROR and records an error message in the interp result if
+ *	a namespace qualified pattern refers to a namespace that does not
+ *	exist. Otherwise, returns TCL_OK.
+ *
+ * Side effects:
+ *	May delete commands.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ForgetImport(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Namespace *namespacePtr,/* Points to the namespace from which
+				 * previously imported commands should be
+				 * removed. NULL for current namespace. */
+    const char *pattern)	/* String pattern indicating which imported
+				 * commands to remove. */
+{
+    Namespace *nsPtr, *sourceNsPtr, *dummyPtr;
+    const char *simplePattern;
+    char *cmdName;
+    register Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+
+    /*
+     * If the specified namespace is NULL, use the current namespace.
+     */
+
+    if (namespacePtr == NULL) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    } else {
+	nsPtr = (Namespace *) namespacePtr;
+    }
+
+    /*
+     * Parse the pattern into its namespace-qualification (if any) and the
+     * simple pattern.
+     */
+
+    TclGetNamespaceForQualName(interp, pattern, nsPtr, TCL_NAMESPACE_ONLY,
+	    &sourceNsPtr, &dummyPtr, &dummyPtr, &simplePattern);
+
+    if (sourceNsPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown namespace in namespace forget pattern \"%s\"",
+		pattern));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE", pattern, NULL);
+	return TCL_ERROR;
+    }
+
+    if (strcmp(pattern, simplePattern) == 0) {
+	/*
+	 * The pattern is simple. Delete any imported commands that match it.
+	 */
+
+	if (TclMatchIsTrivial(simplePattern)) {
+	    hPtr = Tcl_FindHashEntry(&nsPtr->cmdTable, simplePattern);
+	    if (hPtr != NULL) {
+		Command *cmdPtr = Tcl_GetHashValue(hPtr);
+
+		if (cmdPtr && (cmdPtr->deleteProc == DeleteImportedCmd)) {
+		    Tcl_DeleteCommandFromToken(interp, (Tcl_Command) cmdPtr);
+		}
+	    }
+	    return TCL_OK;
+	}
+	for (hPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+		(hPtr != NULL); hPtr = Tcl_NextHashEntry(&search)) {
+	    Command *cmdPtr = Tcl_GetHashValue(hPtr);
+
+	    if (cmdPtr->deleteProc != DeleteImportedCmd) {
+		continue;
+	    }
+	    cmdName = Tcl_GetHashKey(&nsPtr->cmdTable, hPtr);
+	    if (Tcl_StringMatch(cmdName, simplePattern)) {
+		Tcl_DeleteCommandFromToken(interp, (Tcl_Command) cmdPtr);
+	    }
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * The pattern was namespace-qualified.
+     */
+
+    for (hPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search); (hPtr != NULL);
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	Tcl_CmdInfo info;
+	Tcl_Command token = Tcl_GetHashValue(hPtr);
+	Tcl_Command origin = TclGetOriginalCommand(token);
+
+	if (Tcl_GetCommandInfoFromToken(origin, &info) == 0) {
+	    continue;			/* Not an imported command. */
+	}
+	if (info.namespacePtr != (Tcl_Namespace *) sourceNsPtr) {
+	    /*
+	     * Original not in namespace we're matching. Check the first link
+	     * in the import chain.
+	     */
+
+	    Command *cmdPtr = (Command *) token;
+	    ImportedCmdData *dataPtr = cmdPtr->objClientData;
+	    Tcl_Command firstToken = (Tcl_Command) dataPtr->realCmdPtr;
+
+	    if (firstToken == origin) {
+		continue;
+	    }
+	    Tcl_GetCommandInfoFromToken(firstToken, &info);
+	    if (info.namespacePtr != (Tcl_Namespace *) sourceNsPtr) {
+		continue;
+	    }
+	    origin = firstToken;
+	}
+	if (Tcl_StringMatch(Tcl_GetCommandName(NULL, origin), simplePattern)){
+	    Tcl_DeleteCommandFromToken(interp, token);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetOriginalCommand --
+ *
+ *	An imported command is created in an namespace when a "real" command
+ *	is imported from another namespace. If the specified command is an
+ *	imported command, this function returns the original command it refers
+ *	to.
+ *
+ * Results:
+ *	If the command was imported into a sequence of namespaces a, b,...,n
+ *	where each successive namespace just imports the command from the
+ *	previous namespace, this function returns the Tcl_Command token in the
+ *	first namespace, a. Otherwise, if the specified command is not an
+ *	imported command, the function returns NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclGetOriginalCommand(
+    Tcl_Command command)	/* The imported command for which the original
+				 * command should be returned. */
+{
+    register Command *cmdPtr = (Command *) command;
+    ImportedCmdData *dataPtr;
+
+    if (cmdPtr->deleteProc != DeleteImportedCmd) {
+	return NULL;
+    }
+
+    while (cmdPtr->deleteProc == DeleteImportedCmd) {
+	dataPtr = cmdPtr->objClientData;
+	cmdPtr = dataPtr->realCmdPtr;
+    }
+    return (Tcl_Command) cmdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InvokeImportedCmd --
+ *
+ *	Invoked by Tcl whenever the user calls an imported command that was
+ *	created by Tcl_Import. Finds the "real" command (in another
+ *	namespace), and passes control to it.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result object is set to an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InvokeImportedNRCmd(
+    ClientData clientData,	/* Points to the imported command's
+				 * ImportedCmdData structure. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    ImportedCmdData *dataPtr = clientData;
+    Command *realCmdPtr = dataPtr->realCmdPtr;
+
+    TclSkipTailcall(interp);
+    return TclNREvalObjv(interp, objc, objv, TCL_EVAL_NOERR, realCmdPtr);
+}
+
+static int
+InvokeImportedCmd(
+    ClientData clientData,	/* Points to the imported command's
+				 * ImportedCmdData structure. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, InvokeImportedNRCmd, clientData,
+	    objc, objv);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteImportedCmd --
+ *
+ *	Invoked by Tcl whenever an imported command is deleted. The "real"
+ *	command keeps a list of all the imported commands that refer to it, so
+ *	those imported commands can be deleted when the real command is
+ *	deleted. This function removes the imported command reference from the
+ *	real command's list, and frees up the memory associated with the
+ *	imported command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes the imported command from the real command's import list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteImportedCmd(
+    ClientData clientData)	/* Points to the imported command's
+				 * ImportedCmdData structure. */
+{
+    ImportedCmdData *dataPtr = clientData;
+    Command *realCmdPtr = dataPtr->realCmdPtr;
+    Command *selfPtr = dataPtr->selfPtr;
+    register ImportRef *refPtr, *prevPtr;
+
+    prevPtr = NULL;
+    for (refPtr = realCmdPtr->importRefPtr; refPtr != NULL;
+	    refPtr = refPtr->nextPtr) {
+	if (refPtr->importedCmdPtr == selfPtr) {
+	    /*
+	     * Remove *refPtr from real command's list of imported commands
+	     * that refer to it.
+	     */
+
+	    if (prevPtr == NULL) { /* refPtr is first in list. */
+		realCmdPtr->importRefPtr = refPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = refPtr->nextPtr;
+	    }
+	    ckfree(refPtr);
+	    ckfree(dataPtr);
+	    return;
+	}
+	prevPtr = refPtr;
+    }
+
+    Tcl_Panic("DeleteImportedCmd: did not find cmd in real cmd's list of import references");
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetNamespaceForQualName --
+ *
+ *	Given a qualified name specifying a command, variable, or namespace,
+ *	and a namespace in which to resolve the name, this function returns a
+ *	pointer to the namespace that contains the item. A qualified name
+ *	consists of the "simple" name of an item qualified by the names of an
+ *	arbitrary number of containing namespace separated by "::"s. If the
+ *	qualified name starts with "::", it is interpreted absolutely from the
+ *	global namespace. Otherwise, it is interpreted relative to the
+ *	namespace specified by cxtNsPtr if it is non-NULL. If cxtNsPtr is
+ *	NULL, the name is interpreted relative to the current namespace.
+ *
+ *	A relative name like "foo::bar::x" can be found starting in either the
+ *	current namespace or in the global namespace. So each search usually
+ *	follows two tracks, and two possible namespaces are returned. If the
+ *	function sets either *nsPtrPtr or *altNsPtrPtr to NULL, then that path
+ *	failed.
+ *
+ *	If "flags" contains TCL_GLOBAL_ONLY, the relative qualified name is
+ *	sought only in the global :: namespace. The alternate search (also)
+ *	starting from the global namespace is ignored and *altNsPtrPtr is set
+ *	NULL.
+ *
+ *	If "flags" contains TCL_NAMESPACE_ONLY, the relative qualified name is
+ *	sought only in the namespace specified by cxtNsPtr. The alternate
+ *	search starting from the global namespace is ignored and *altNsPtrPtr
+ *	is set NULL. If both TCL_GLOBAL_ONLY and TCL_NAMESPACE_ONLY are
+ *	specified, TCL_GLOBAL_ONLY is ignored and the search starts from the
+ *	namespace specified by cxtNsPtr.
+ *
+ *	If "flags" contains TCL_CREATE_NS_IF_UNKNOWN, all namespace components
+ *	of the qualified name that cannot be found are automatically created
+ *	within their specified parent. This makes sure that functions like
+ *	Tcl_CreateCommand always succeed. There is no alternate search path,
+ *	so *altNsPtrPtr is set NULL.
+ *
+ *	If "flags" contains TCL_FIND_ONLY_NS, the qualified name is treated as
+ *	a reference to a namespace, and the entire qualified name is followed.
+ *	If the name is relative, the namespace is looked up only in the
+ *	current namespace. A pointer to the namespace is stored in *nsPtrPtr
+ *	and NULL is stored in *simpleNamePtr. Otherwise, if TCL_FIND_ONLY_NS
+ *	is not specified, only the leading components are treated as namespace
+ *	names, and a pointer to the simple name of the final component is
+ *	stored in *simpleNamePtr.
+ *
+ * Results:
+ *	It sets *nsPtrPtr and *altNsPtrPtr to point to the two possible
+ *	namespaces which represent the last (containing) namespace in the
+ *	qualified name. If the function sets either *nsPtrPtr or *altNsPtrPtr
+ *	to NULL, then the search along that path failed. The function also
+ *	stores a pointer to the simple name of the final component in
+ *	*simpleNamePtr. If the qualified name is "::" or was treated as a
+ *	namespace reference (TCL_FIND_ONLY_NS), the function stores a pointer
+ *	to the namespace in *nsPtrPtr, NULL in *altNsPtrPtr, and sets
+ *	*simpleNamePtr to point to an empty string.
+ *
+ *	If there is an error, this function returns TCL_ERROR. If "flags"
+ *	contains TCL_LEAVE_ERR_MSG, an error message is returned in the
+ *	interpreter's result object. Otherwise, the interpreter's result
+ *	object is left unchanged.
+ *
+ *	*actualCxtPtrPtr is set to the actual context namespace. It is set to
+ *	the input context namespace pointer in cxtNsPtr. If cxtNsPtr is NULL,
+ *	it is set to the current namespace context.
+ *
+ *	For backwards compatibility with the TclPro byte code loader, this
+ *	function always returns TCL_OK.
+ *
+ * Side effects:
+ *	If "flags" contains TCL_CREATE_NS_IF_UNKNOWN, new namespaces may be
+ *	created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetNamespaceForQualName(
+    Tcl_Interp *interp,		/* Interpreter in which to find the namespace
+				 * containing qualName. */
+    const char *qualName,	/* A namespace-qualified name of an command,
+				 * variable, or namespace. */
+    Namespace *cxtNsPtr,	/* The namespace in which to start the search
+				 * for qualName's namespace. If NULL start
+				 * from the current namespace. Ignored if
+				 * TCL_GLOBAL_ONLY is set. */
+    int flags,			/* Flags controlling the search: an OR'd
+				 * combination of TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY, TCL_FIND_ONLY_NS, and
+				 * TCL_CREATE_NS_IF_UNKNOWN. */
+    Namespace **nsPtrPtr,	/* Address where function stores a pointer to
+				 * containing namespace if qualName is found
+				 * starting from *cxtNsPtr or, if
+				 * TCL_GLOBAL_ONLY is set, if qualName is
+				 * found in the global :: namespace. NULL is
+				 * stored otherwise. */
+    Namespace **altNsPtrPtr,	/* Address where function stores a pointer to
+				 * containing namespace if qualName is found
+				 * starting from the global :: namespace.
+				 * NULL is stored if qualName isn't found
+				 * starting from :: or if the TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY, TCL_FIND_ONLY_NS,
+				 * TCL_CREATE_NS_IF_UNKNOWN flag is set. */
+    Namespace **actualCxtPtrPtr,/* Address where function stores a pointer to
+				 * the actual namespace from which the search
+				 * started. This is either cxtNsPtr, the ::
+				 * namespace if TCL_GLOBAL_ONLY was specified,
+				 * or the current namespace if cxtNsPtr was
+				 * NULL. */
+    const char **simpleNamePtr) /* Address where function stores the simple
+				 * name at end of the qualName, or NULL if
+				 * qualName is "::" or the flag
+				 * TCL_FIND_ONLY_NS was specified. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Namespace *nsPtr = cxtNsPtr;
+    Namespace *altNsPtr;
+    Namespace *globalNsPtr = iPtr->globalNsPtr;
+    const char *start, *end;
+    const char *nsName;
+    Tcl_HashEntry *entryPtr;
+    Tcl_DString buffer;
+    int len;
+
+    /*
+     * Determine the context namespace nsPtr in which to start the primary
+     * search. If the qualName name starts with a "::" or TCL_GLOBAL_ONLY was
+     * specified, search from the global namespace. Otherwise, use the
+     * namespace given in cxtNsPtr, or if that is NULL, use the current
+     * namespace context. Note that we always treat two or more adjacent ":"s
+     * as a namespace separator.
+     */
+
+    if (flags & TCL_GLOBAL_ONLY) {
+	nsPtr = globalNsPtr;
+    } else if (nsPtr == NULL) {
+	nsPtr = iPtr->varFramePtr->nsPtr;
+    }
+
+    start = qualName;			/* Points to start of qualifying
+					 * namespace. */
+    if ((*qualName == ':') && (*(qualName+1) == ':')) {
+	start = qualName+2;		/* Skip over the initial :: */
+	while (*start == ':') {
+	    start++;			/* Skip over a subsequent : */
+	}
+	nsPtr = globalNsPtr;
+	if (*start == '\0') {		/* qualName is just two or more
+					 * ":"s. */
+	    *nsPtrPtr = globalNsPtr;
+	    *altNsPtrPtr = NULL;
+	    *actualCxtPtrPtr = globalNsPtr;
+	    *simpleNamePtr = start;	/* Points to empty string. */
+	    return TCL_OK;
+	}
+    }
+    *actualCxtPtrPtr = nsPtr;
+
+    /*
+     * Start an alternate search path starting with the global namespace.
+     * However, if the starting context is the global namespace, or if the
+     * flag is set to search only the namespace *cxtNsPtr, ignore the
+     * alternate search path.
+     */
+
+    altNsPtr = globalNsPtr;
+    if ((nsPtr == globalNsPtr)
+	    || (flags & (TCL_NAMESPACE_ONLY | TCL_FIND_ONLY_NS))) {
+	altNsPtr = NULL;
+    }
+
+    /*
+     * Loop to resolve each namespace qualifier in qualName.
+     */
+
+    Tcl_DStringInit(&buffer);
+    end = start;
+    while (*start != '\0') {
+	/*
+	 * Find the next namespace qualifier (i.e., a name ending in "::") or
+	 * the end of the qualified name (i.e., a name ending in "\0"). Set
+	 * len to the number of characters, starting from start, in the name;
+	 * set end to point after the "::"s or at the "\0".
+	 */
+
+	len = 0;
+	for (end = start;  *end != '\0';  end++) {
+	    if ((*end == ':') && (*(end+1) == ':')) {
+		end += 2;		/* Skip over the initial :: */
+		while (*end == ':') {
+		    end++;		/* Skip over the subsequent : */
+		}
+		break;			/* Exit for loop; end is after ::'s */
+	    }
+	    len++;
+	}
+
+	if (*end=='\0' && !(end-start>=2 && *(end-1)==':' && *(end-2)==':')) {
+	    /*
+	     * qualName ended with a simple name at start. If TCL_FIND_ONLY_NS
+	     * was specified, look this up as a namespace. Otherwise, start is
+	     * the name of a cmd or var and we are done.
+	     */
+
+	    if (flags & TCL_FIND_ONLY_NS) {
+		nsName = start;
+	    } else {
+		*nsPtrPtr = nsPtr;
+		*altNsPtrPtr = altNsPtr;
+		*simpleNamePtr = start;
+		Tcl_DStringFree(&buffer);
+		return TCL_OK;
+	    }
+	} else {
+	    /*
+	     * start points to the beginning of a namespace qualifier ending
+	     * in "::". end points to the start of a name in that namespace
+	     * that might be empty. Copy the namespace qualifier to a buffer
+	     * so it can be null terminated. We can't modify the incoming
+	     * qualName since it may be a string constant.
+	     */
+
+	    TclDStringClear(&buffer);
+	    Tcl_DStringAppend(&buffer, start, len);
+	    nsName = Tcl_DStringValue(&buffer);
+	}
+
+	/*
+	 * Look up the namespace qualifier nsName in the current namespace
+	 * context. If it isn't found but TCL_CREATE_NS_IF_UNKNOWN is set,
+	 * create that qualifying namespace. This is needed for functions like
+	 * Tcl_CreateCommand that cannot fail.
+	 */
+
+	if (nsPtr != NULL) {
+#ifndef BREAK_NAMESPACE_COMPAT
+	    entryPtr = Tcl_FindHashEntry(&nsPtr->childTable, nsName);
+#else
+	    if (nsPtr->childTablePtr == NULL) {
+		entryPtr = NULL;
+	    } else {
+		entryPtr = Tcl_FindHashEntry(nsPtr->childTablePtr, nsName);
+	    }
+#endif
+	    if (entryPtr != NULL) {
+		nsPtr = Tcl_GetHashValue(entryPtr);
+	    } else if (flags & TCL_CREATE_NS_IF_UNKNOWN) {
+		Tcl_CallFrame *framePtr;
+
+		(void) TclPushStackFrame(interp, &framePtr,
+			(Tcl_Namespace *) nsPtr, /*isProcCallFrame*/ 0);
+
+		nsPtr = (Namespace *)
+			Tcl_CreateNamespace(interp, nsName, NULL, NULL);
+		TclPopStackFrame(interp);
+
+		if (nsPtr == NULL) {
+		    Tcl_Panic("Could not create namespace '%s'", nsName);
+		}
+	    } else {			/* Namespace not found and was not
+					 * created. */
+		nsPtr = NULL;
+	    }
+	}
+
+	/*
+	 * Look up the namespace qualifier in the alternate search path too.
+	 */
+
+	if (altNsPtr != NULL) {
+#ifndef BREAK_NAMESPACE_COMPAT
+	    entryPtr = Tcl_FindHashEntry(&altNsPtr->childTable, nsName);
+#else
+	    if (altNsPtr->childTablePtr != NULL) {
+		entryPtr = Tcl_FindHashEntry(altNsPtr->childTablePtr, nsName);
+	    } else {
+		entryPtr = NULL;
+	    }
+#endif
+	    if (entryPtr != NULL) {
+		altNsPtr = Tcl_GetHashValue(entryPtr);
+	    } else {
+		altNsPtr = NULL;
+	    }
+	}
+
+	/*
+	 * If both search paths have failed, return NULL results.
+	 */
+
+	if ((nsPtr == NULL) && (altNsPtr == NULL)) {
+	    *nsPtrPtr = NULL;
+	    *altNsPtrPtr = NULL;
+	    *simpleNamePtr = NULL;
+	    Tcl_DStringFree(&buffer);
+	    return TCL_OK;
+	}
+
+	start = end;
+    }
+
+    /*
+     * We ignore trailing "::"s in a namespace name, but in a command or
+     * variable name, trailing "::"s refer to the cmd or var named {}.
+     */
+
+    if ((flags & TCL_FIND_ONLY_NS) || (end>start && *(end-1)!=':')) {
+	*simpleNamePtr = NULL;		/* Found namespace name. */
+    } else {
+	*simpleNamePtr = end;		/* Found cmd/var: points to empty
+					 * string. */
+    }
+
+    /*
+     * As a special case, if we are looking for a namespace and qualName is ""
+     * and the current active namespace (nsPtr) is not the global namespace,
+     * return NULL (no namespace was found). This is because namespaces can
+     * not have empty names except for the global namespace.
+     */
+
+    if ((flags & TCL_FIND_ONLY_NS) && (*qualName == '\0')
+	    && (nsPtr != globalNsPtr)) {
+	nsPtr = NULL;
+    }
+
+    *nsPtrPtr = nsPtr;
+    *altNsPtrPtr = altNsPtr;
+    Tcl_DStringFree(&buffer);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindNamespace --
+ *
+ *	Searches for a namespace.
+ *
+ * Results:
+ *	Returns a pointer to the namespace if it is found. Otherwise, returns
+ *	NULL and leaves an error message in the interpreter's result object if
+ *	"flags" contains TCL_LEAVE_ERR_MSG.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Namespace *
+Tcl_FindNamespace(
+    Tcl_Interp *interp,		/* The interpreter in which to find the
+				 * namespace. */
+    const char *name,		/* Namespace name. If it starts with "::",
+				 * will be looked up in global namespace.
+				 * Else, looked up first in contextNsPtr
+				 * (current namespace if contextNsPtr is
+				 * NULL), then in global namespace. */
+    Tcl_Namespace *contextNsPtr,/* Ignored if TCL_GLOBAL_ONLY flag is set or
+				 * if the name starts with "::". Otherwise,
+				 * points to namespace in which to resolve
+				 * name; if NULL, look up name in the current
+				 * namespace. */
+    register int flags)		/* Flags controlling namespace lookup: an OR'd
+				 * combination of TCL_GLOBAL_ONLY and
+				 * TCL_LEAVE_ERR_MSG flags. */
+{
+    Namespace *nsPtr, *dummy1Ptr, *dummy2Ptr;
+    const char *dummy;
+
+    /*
+     * Find the namespace(s) that contain the specified namespace name. Add
+     * the TCL_FIND_ONLY_NS flag to resolve the name all the way down to its
+     * last component, a namespace.
+     */
+
+    TclGetNamespaceForQualName(interp, name, (Namespace *) contextNsPtr,
+	    flags|TCL_FIND_ONLY_NS, &nsPtr, &dummy1Ptr, &dummy2Ptr, &dummy);
+
+    if (nsPtr != NULL) {
+	return (Tcl_Namespace *) nsPtr;
+    }
+
+    if (flags & TCL_LEAVE_ERR_MSG) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "unknown namespace \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE", name, NULL);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindCommand --
+ *
+ *	Searches for a command.
+ *
+ * Results:
+ *	Returns a token for the command if it is found. Otherwise, if it can't
+ *	be found or there is an error, returns NULL and leaves an error
+ *	message in the interpreter's result object if "flags" contains
+ *	TCL_LEAVE_ERR_MSG.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_FindCommand(
+    Tcl_Interp *interp,		/* The interpreter in which to find the
+				 * command and to report errors. */
+    const char *name,		/* Command's name. If it starts with "::",
+				 * will be looked up in global namespace.
+				 * Else, looked up first in contextNsPtr
+				 * (current namespace if contextNsPtr is
+				 * NULL), then in global namespace. */
+    Tcl_Namespace *contextNsPtr,/* Ignored if TCL_GLOBAL_ONLY flag set.
+				 * Otherwise, points to namespace in which to
+				 * resolve name. If NULL, look up name in the
+				 * current namespace. */
+    int flags)			/* An OR'd combination of flags:
+				 * TCL_GLOBAL_ONLY (look up name only in
+				 * global namespace), TCL_NAMESPACE_ONLY (look
+				 * up only in contextNsPtr, or the current
+				 * namespace if contextNsPtr is NULL), and
+				 * TCL_LEAVE_ERR_MSG. If both TCL_GLOBAL_ONLY
+				 * and TCL_NAMESPACE_ONLY are given,
+				 * TCL_GLOBAL_ONLY is ignored. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Namespace *cxtNsPtr;
+    register Tcl_HashEntry *entryPtr;
+    register Command *cmdPtr;
+    const char *simpleName;
+    int result;
+
+    /*
+     * If this namespace has a command resolver, then give it first crack at
+     * the command resolution. If the interpreter has any command resolvers,
+     * consult them next. The command resolver functions may return a
+     * Tcl_Command value, they may signal to continue onward, or they may
+     * signal an error.
+     */
+
+    if ((flags & TCL_GLOBAL_ONLY) || !strncmp(name, "::", 2)) {
+	cxtNsPtr = (Namespace *) TclGetGlobalNamespace(interp);
+    } else if (contextNsPtr != NULL) {
+	cxtNsPtr = (Namespace *) contextNsPtr;
+    } else {
+	cxtNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    }
+
+    if (cxtNsPtr->cmdResProc != NULL || iPtr->resolverPtr != NULL) {
+	ResolverScheme *resPtr = iPtr->resolverPtr;
+	Tcl_Command cmd;
+
+	if (cxtNsPtr->cmdResProc) {
+	    result = cxtNsPtr->cmdResProc(interp, name,
+		    (Tcl_Namespace *) cxtNsPtr, flags, &cmd);
+	} else {
+	    result = TCL_CONTINUE;
+	}
+
+	while (result == TCL_CONTINUE && resPtr) {
+	    if (resPtr->cmdResProc) {
+		result = resPtr->cmdResProc(interp, name,
+			(Tcl_Namespace *) cxtNsPtr, flags, &cmd);
+	    }
+	    resPtr = resPtr->nextPtr;
+	}
+
+	if (result == TCL_OK) {
+	    ((Command *)cmd)->flags |= CMD_VIA_RESOLVER;
+	    return cmd;
+
+	} else if (result != TCL_CONTINUE) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * Find the namespace(s) that contain the command.
+     */
+
+    cmdPtr = NULL;
+    if (cxtNsPtr->commandPathLength!=0 && strncmp(name, "::", 2)
+	    && !(flags & TCL_NAMESPACE_ONLY)) {
+	int i;
+	Namespace *pathNsPtr, *realNsPtr, *dummyNsPtr;
+
+	(void) TclGetNamespaceForQualName(interp, name, cxtNsPtr,
+		TCL_NAMESPACE_ONLY, &realNsPtr, &dummyNsPtr, &dummyNsPtr,
+		&simpleName);
+	if ((realNsPtr != NULL) && (simpleName != NULL)) {
+	    if ((cxtNsPtr == realNsPtr)
+		    || !(realNsPtr->flags & NS_DYING)) {
+		entryPtr = Tcl_FindHashEntry(&realNsPtr->cmdTable, simpleName);
+		if (entryPtr != NULL) {
+		    cmdPtr = Tcl_GetHashValue(entryPtr);
+		}
+	    }
+	}
+
+	/*
+	 * Next, check along the path.
+	 */
+
+	for (i=0 ; i<cxtNsPtr->commandPathLength && cmdPtr==NULL ; i++) {
+	    pathNsPtr = cxtNsPtr->commandPathArray[i].nsPtr;
+	    if (pathNsPtr == NULL) {
+		continue;
+	    }
+	    (void) TclGetNamespaceForQualName(interp, name, pathNsPtr,
+		    TCL_NAMESPACE_ONLY, &realNsPtr, &dummyNsPtr, &dummyNsPtr,
+		    &simpleName);
+	    if ((realNsPtr != NULL) && (simpleName != NULL)
+		    && !(realNsPtr->flags & NS_DYING)) {
+		entryPtr = Tcl_FindHashEntry(&realNsPtr->cmdTable, simpleName);
+		if (entryPtr != NULL) {
+		    cmdPtr = Tcl_GetHashValue(entryPtr);
+		}
+	    }
+	}
+
+	/*
+	 * If we've still not found the command, look in the global namespace
+	 * as a last resort.
+	 */
+
+	if (cmdPtr == NULL) {
+	    (void) TclGetNamespaceForQualName(interp, name, NULL,
+		    TCL_GLOBAL_ONLY, &realNsPtr, &dummyNsPtr, &dummyNsPtr,
+		    &simpleName);
+	    if ((realNsPtr != NULL) && (simpleName != NULL)
+		    && !(realNsPtr->flags & NS_DYING)) {
+		entryPtr = Tcl_FindHashEntry(&realNsPtr->cmdTable, simpleName);
+		if (entryPtr != NULL) {
+		    cmdPtr = Tcl_GetHashValue(entryPtr);
+		}
+	    }
+	}
+    } else {
+	Namespace *nsPtr[2];
+	register int search;
+
+	TclGetNamespaceForQualName(interp, name, (Namespace *) contextNsPtr,
+		flags, &nsPtr[0], &nsPtr[1], &cxtNsPtr, &simpleName);
+
+	/*
+	 * Look for the command in the command table of its namespace. Be sure
+	 * to check both possible search paths: from the specified namespace
+	 * context and from the global namespace.
+	 */
+
+	for (search = 0;  (search < 2) && (cmdPtr == NULL);  search++) {
+	    if ((nsPtr[search] != NULL) && (simpleName != NULL)) {
+		entryPtr = Tcl_FindHashEntry(&nsPtr[search]->cmdTable,
+			simpleName);
+		if (entryPtr != NULL) {
+		    cmdPtr = Tcl_GetHashValue(entryPtr);
+		}
+	    }
+	}
+    }
+
+    if (cmdPtr != NULL) {
+	cmdPtr->flags  &= ~CMD_VIA_RESOLVER;
+	return (Tcl_Command) cmdPtr;
+    }
+
+    if (flags & TCL_LEAVE_ERR_MSG) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "unknown command \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND", name, NULL);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclResetShadowedCmdRefs --
+ *
+ *	Called when a command is added to a namespace to check for existing
+ *	command references that the new command may invalidate. Consider the
+ *	following cases that could happen when you add a command "foo" to a
+ *	namespace "b":
+ *	   1. It could shadow a command named "foo" at the global scope. If
+ *	      it does, all command references in the namespace "b" are
+ *	      suspect.
+ *	   2. Suppose the namespace "b" resides in a namespace "a". Then to
+ *	      "a" the new command "b::foo" could shadow another command
+ *	      "b::foo" in the global namespace. If so, then all command
+ *	      references in "a" * are suspect.
+ *	The same checks are applied to all parent namespaces, until we reach
+ *	the global :: namespace.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the new command shadows an existing command, the cmdRefEpoch
+ *	counter is incremented in each namespace that sees the shadow. This
+ *	invalidates all command references that were previously cached in that
+ *	namespace. The next time the commands are used, they are resolved from
+ *	scratch.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclResetShadowedCmdRefs(
+    Tcl_Interp *interp,		/* Interpreter containing the new command. */
+    Command *newCmdPtr)		/* Points to the new command. */
+{
+    char *cmdName;
+    Tcl_HashEntry *hPtr;
+    register Namespace *nsPtr;
+    Namespace *trailNsPtr, *shadowNsPtr;
+    Namespace *globalNsPtr = (Namespace *) TclGetGlobalNamespace(interp);
+    int found, i;
+    int trailFront = -1;
+    int trailSize = 5;		/* Formerly NUM_TRAIL_ELEMS. */
+    Namespace **trailPtr = TclStackAlloc(interp,
+	    trailSize * sizeof(Namespace *));
+
+    /*
+     * Start at the namespace containing the new command, and work up through
+     * the list of parents. Stop just before the global namespace, since the
+     * global namespace can't "shadow" its own entries.
+     *
+     * The namespace "trail" list we build consists of the names of each
+     * namespace that encloses the new command, in order from outermost to
+     * innermost: for example, "a" then "b". Each iteration of this loop
+     * eventually extends the trail upwards by one namespace, nsPtr. We use
+     * this trail list to see if nsPtr (e.g. "a" in 2. above) could have
+     * now-invalid cached command references. This will happen if nsPtr
+     * (e.g. "a") contains a sequence of child namespaces (e.g. "b") such that
+     * there is a identically-named sequence of child namespaces starting from
+     * :: (e.g. "::b") whose tail namespace contains a command also named
+     * cmdName.
+     */
+
+    cmdName = Tcl_GetHashKey(newCmdPtr->hPtr->tablePtr, newCmdPtr->hPtr);
+    for (nsPtr=newCmdPtr->nsPtr ; (nsPtr!=NULL) && (nsPtr!=globalNsPtr) ;
+	    nsPtr=nsPtr->parentPtr) {
+	/*
+	 * Find the maximal sequence of child namespaces contained in nsPtr
+	 * such that there is a identically-named sequence of child namespaces
+	 * starting from ::. shadowNsPtr will be the tail of this sequence, or
+	 * the deepest namespace under :: that might contain a command now
+	 * shadowed by cmdName. We check below if shadowNsPtr actually
+	 * contains a command cmdName.
+	 */
+
+	found = 1;
+	shadowNsPtr = globalNsPtr;
+
+	for (i = trailFront;  i >= 0;  i--) {
+	    trailNsPtr = trailPtr[i];
+#ifndef BREAK_NAMESPACE_COMPAT
+	    hPtr = Tcl_FindHashEntry(&shadowNsPtr->childTable,
+		    trailNsPtr->name);
+#else
+	    if (shadowNsPtr->childTablePtr != NULL) {
+		hPtr = Tcl_FindHashEntry(shadowNsPtr->childTablePtr,
+			trailNsPtr->name);
+	    } else {
+		hPtr = NULL;
+	    }
+#endif
+	    if (hPtr != NULL) {
+		shadowNsPtr = Tcl_GetHashValue(hPtr);
+	    } else {
+		found = 0;
+		break;
+	    }
+	}
+
+	/*
+	 * If shadowNsPtr contains a command named cmdName, we invalidate all
+	 * of the command refs cached in nsPtr. As a boundary case,
+	 * shadowNsPtr is initially :: and we check for case 1. above.
+	 */
+
+	if (found) {
+	    hPtr = Tcl_FindHashEntry(&shadowNsPtr->cmdTable, cmdName);
+	    if (hPtr != NULL) {
+		nsPtr->cmdRefEpoch++;
+		TclInvalidateNsPath(nsPtr);
+
+		/*
+		 * If the shadowed command was compiled to bytecodes, we
+		 * invalidate all the bytecodes in nsPtr, to force a new
+		 * compilation. We use the resolverEpoch to signal the need
+		 * for a fresh compilation of every bytecode.
+		 */
+
+		if (((Command *)Tcl_GetHashValue(hPtr))->compileProc != NULL){
+		    nsPtr->resolverEpoch++;
+		}
+	    }
+	}
+
+	/*
+	 * Insert nsPtr at the front of the trail list: i.e., at the end of
+	 * the trailPtr array.
+	 */
+
+	trailFront++;
+	if (trailFront == trailSize) {
+	    int newSize = 2 * trailSize;
+
+	    trailPtr = TclStackRealloc(interp, trailPtr,
+		    newSize * sizeof(Namespace *));
+	    trailSize = newSize;
+	}
+	trailPtr[trailFront] = nsPtr;
+    }
+    TclStackFree(interp, trailPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetNamespaceFromObj, GetNamespaceFromObj --
+ *
+ *	Gets the namespace specified by the name in a Tcl_Obj.
+ *
+ * Results:
+ *	Returns TCL_OK if the namespace was resolved successfully, and stores
+ *	a pointer to the namespace in the location specified by nsPtrPtr. If
+ *	the namespace can't be found, or anything else goes wrong, this
+ *	function returns TCL_ERROR and writes an error message to interp,
+ *	if non-NULL.
+ *
+ * Side effects:
+ *	May update the internal representation for the object, caching the
+ *	namespace reference. The next time this function is called, the
+ *	namespace value can be found quickly.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetNamespaceFromObj(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    Tcl_Obj *objPtr,		/* The object to be resolved as the name of a
+				 * namespace. */
+    Tcl_Namespace **nsPtrPtr)	/* Result namespace pointer goes here. */
+{
+    if (GetNamespaceFromObj(interp, objPtr, nsPtrPtr) == TCL_ERROR) {
+	const char *name = TclGetString(objPtr);
+
+	if ((name[0] == ':') && (name[1] == ':')) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "namespace \"%s\" not found", name));
+	} else {
+	    /*
+	     * Get the current namespace name.
+	     */
+
+	    NamespaceCurrentCmd(NULL, interp, 1, NULL);
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "namespace \"%s\" not found in \"%s\"", name,
+		    Tcl_GetStringResult(interp)));
+	}
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE", name, NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+static int
+GetNamespaceFromObj(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    Tcl_Obj *objPtr,		/* The object to be resolved as the name of a
+				 * namespace. */
+    Tcl_Namespace **nsPtrPtr)	/* Result namespace pointer goes here. */
+{
+    ResolvedNsName *resNamePtr;
+    Namespace *nsPtr, *refNsPtr;
+
+    if (objPtr->typePtr == &nsNameType) {
+	/*
+	 * Check that the ResolvedNsName is still valid; avoid letting the ref
+	 * cross interps.
+	 */
+
+	resNamePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	nsPtr = resNamePtr->nsPtr;
+	refNsPtr = resNamePtr->refNsPtr;
+	if (!(nsPtr->flags & NS_DYING) && (interp == nsPtr->interp)
+		&& (!refNsPtr || (refNsPtr ==
+		(Namespace *) TclGetCurrentNamespace(interp)))) {
+	    *nsPtrPtr = (Tcl_Namespace *) nsPtr;
+	    return TCL_OK;
+	}
+    }
+    if (SetNsNameFromAny(interp, objPtr) == TCL_OK) {
+	resNamePtr = objPtr->internalRep.twoPtrValue.ptr1;
+	*nsPtrPtr = (Tcl_Namespace *) resNamePtr->nsPtr;
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitNamespaceCmd --
+ *
+ *	This function is called to create the "namespace" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	Handle for the namespace command, or NULL on failure.
+ *
+ * Side effects:
+ *	none
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+TclInitNamespaceCmd(
+    Tcl_Interp *interp)		/* Current interpreter. */
+{
+    return TclMakeEnsemble(interp, "namespace", defaultNamespaceMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceChildrenCmd --
+ *
+ *	Invoked to implement the "namespace children" command that returns a
+ *	list containing the fully-qualified names of the child namespaces of a
+ *	given namespace. Handles the following syntax:
+ *
+ *	    namespace children ?name? ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceChildrenCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *namespacePtr;
+    Namespace *nsPtr, *childNsPtr;
+    Namespace *globalNsPtr = (Namespace *) TclGetGlobalNamespace(interp);
+    const char *pattern = NULL;
+    Tcl_DString buffer;
+    register Tcl_HashEntry *entryPtr;
+    Tcl_HashSearch search;
+    Tcl_Obj *listPtr, *elemPtr;
+
+    /*
+     * Get a pointer to the specified namespace, or the current namespace.
+     */
+
+    if (objc == 1) {
+	nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    } else if ((objc == 2) || (objc == 3)) {
+	if (TclGetNamespaceFromObj(interp, objv[1], &namespacePtr) != TCL_OK){
+	    return TCL_ERROR;
+	}
+	nsPtr = (Namespace *) namespacePtr;
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?name? ?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the glob-style pattern, if any, used to narrow the search.
+     */
+
+    Tcl_DStringInit(&buffer);
+    if (objc == 3) {
+	const char *name = TclGetString(objv[2]);
+
+	if ((*name == ':') && (*(name+1) == ':')) {
+	    pattern = name;
+	} else {
+	    Tcl_DStringAppend(&buffer, nsPtr->fullName, -1);
+	    if (nsPtr != globalNsPtr) {
+		TclDStringAppendLiteral(&buffer, "::");
+	    }
+	    Tcl_DStringAppend(&buffer, name, -1);
+	    pattern = Tcl_DStringValue(&buffer);
+	}
+    }
+
+    /*
+     * Create a list containing the full names of all child namespaces whose
+     * names match the specified pattern, if any.
+     */
+
+    listPtr = Tcl_NewListObj(0, NULL);
+    if ((pattern != NULL) && TclMatchIsTrivial(pattern)) {
+	unsigned int length = strlen(nsPtr->fullName);
+
+	if (strncmp(pattern, nsPtr->fullName, length) != 0) {
+	    goto searchDone;
+	}
+	if (
+#ifndef BREAK_NAMESPACE_COMPAT
+	    Tcl_FindHashEntry(&nsPtr->childTable, pattern+length) != NULL
+#else
+	    nsPtr->childTablePtr != NULL &&
+	    Tcl_FindHashEntry(nsPtr->childTablePtr, pattern+length) != NULL
+#endif
+	) {
+	    Tcl_ListObjAppendElement(interp, listPtr,
+		    Tcl_NewStringObj(pattern, -1));
+	}
+	goto searchDone;
+    }
+#ifndef BREAK_NAMESPACE_COMPAT
+    entryPtr = Tcl_FirstHashEntry(&nsPtr->childTable, &search);
+#else
+    if (nsPtr->childTablePtr == NULL) {
+	goto searchDone;
+    }
+    entryPtr = Tcl_FirstHashEntry(nsPtr->childTablePtr, &search);
+#endif
+    while (entryPtr != NULL) {
+	childNsPtr = Tcl_GetHashValue(entryPtr);
+	if ((pattern == NULL)
+		|| Tcl_StringMatch(childNsPtr->fullName, pattern)) {
+	    elemPtr = Tcl_NewStringObj(childNsPtr->fullName, -1);
+	    Tcl_ListObjAppendElement(interp, listPtr, elemPtr);
+	}
+	entryPtr = Tcl_NextHashEntry(&search);
+    }
+
+  searchDone:
+    Tcl_SetObjResult(interp, listPtr);
+    Tcl_DStringFree(&buffer);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceCodeCmd --
+ *
+ *	Invoked to implement the "namespace code" command to capture the
+ *	namespace context of a command. Handles the following syntax:
+ *
+ *	    namespace code arg
+ *
+ *	Here "arg" can be a list. "namespace code arg" produces a result
+ *	equivalent to that produced by the command
+ *
+ *	    list ::namespace inscope [namespace current] $arg
+ *
+ *	However, if "arg" is itself a scoped value starting with "::namespace
+ *	inscope", then the result is just "arg".
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	If anything goes wrong, this function returns an error message as the
+ *	result in the interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceCodeCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Namespace *currNsPtr;
+    Tcl_Obj *listPtr, *objPtr;
+    register const char *arg;
+    int length;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arg");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If "arg" is already a scoped value, then return it directly.
+     * Take care to only check for scoping in precisely the style that
+     * [::namespace code] generates it.  Anything more forgiving can have
+     * the effect of failing in namespaces that contain their own custom
+     " "namespace" command.  [Bug 3202171].
+     */
+
+    arg = TclGetStringFromObj(objv[1], &length);
+    if (*arg==':' && length > 20
+	    && strncmp(arg, "::namespace inscope ", 20) == 0) {
+	Tcl_SetObjResult(interp, objv[1]);
+	return TCL_OK;
+    }
+
+    /*
+     * Otherwise, construct a scoped command by building a list with
+     * "namespace inscope", the full name of the current namespace, and the
+     * argument "arg". By constructing a list, we ensure that scoped commands
+     * are interpreted properly when they are executed later, by the
+     * "namespace inscope" command.
+     */
+
+    TclNewObj(listPtr);
+    TclNewLiteralStringObj(objPtr, "::namespace");
+    Tcl_ListObjAppendElement(interp, listPtr, objPtr);
+    TclNewLiteralStringObj(objPtr, "inscope");
+    Tcl_ListObjAppendElement(interp, listPtr, objPtr);
+
+    currNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    if (currNsPtr == (Namespace *) TclGetGlobalNamespace(interp)) {
+	TclNewLiteralStringObj(objPtr, "::");
+    } else {
+	objPtr = Tcl_NewStringObj(currNsPtr->fullName, -1);
+    }
+    Tcl_ListObjAppendElement(interp, listPtr, objPtr);
+
+    Tcl_ListObjAppendElement(interp, listPtr, objv[1]);
+
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceCurrentCmd --
+ *
+ *	Invoked to implement the "namespace current" command which returns the
+ *	fully-qualified name of the current namespace. Handles the following
+ *	syntax:
+ *
+ *	    namespace current
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceCurrentCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Namespace *currNsPtr;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * The "real" name of the global namespace ("::") is the null string, but
+     * we return "::" for it as a convenience to programmers. Note that "" and
+     * "::" are treated as synonyms by the namespace code so that it is still
+     * easy to do things like:
+     *
+     *    namespace [namespace current]::bar { ... }
+     */
+
+    currNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    if (currNsPtr == (Namespace *) TclGetGlobalNamespace(interp)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("::", 2));
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(currNsPtr->fullName, -1));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceDeleteCmd --
+ *
+ *	Invoked to implement the "namespace delete" command to delete
+ *	namespace(s). Handles the following syntax:
+ *
+ *	    namespace delete ?name name...?
+ *
+ *	Each name identifies a namespace. It may include a sequence of
+ *	namespace qualifiers separated by "::"s. If a namespace is found, it
+ *	is deleted: all variables and procedures contained in that namespace
+ *	are deleted. If that namespace is being used on the call stack, it is
+ *	kept alive (but logically deleted) until it is removed from the call
+ *	stack: that is, it can no longer be referenced by name but any
+ *	currently executing procedure that refers to it is allowed to do so
+ *	until the procedure returns. If the namespace can't be found, this
+ *	function returns an error. If no namespaces are specified, this
+ *	command does nothing.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Deletes the specified namespaces. If anything goes wrong, this
+ *	function returns an error message in the interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceDeleteCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *namespacePtr;
+    const char *name;
+    register int i;
+
+    if (objc < 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?name name...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Destroying one namespace may cause another to be destroyed. Break this
+     * into two passes: first check to make sure that all namespaces on the
+     * command line are valid, and report any errors.
+     */
+
+    for (i = 1;  i < objc;  i++) {
+	name = TclGetString(objv[i]);
+	namespacePtr = Tcl_FindNamespace(interp, name, NULL, /*flags*/ 0);
+	if ((namespacePtr == NULL)
+		|| (((Namespace *) namespacePtr)->flags & NS_KILLED)) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "unknown namespace \"%s\" in namespace delete command",
+		    TclGetString(objv[i])));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "NAMESPACE",
+		    TclGetString(objv[i]), NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Okay, now delete each namespace.
+     */
+
+    for (i = 1;  i < objc;  i++) {
+	name = TclGetString(objv[i]);
+	namespacePtr = Tcl_FindNamespace(interp, name, NULL, /* flags */ 0);
+	if (namespacePtr) {
+	    Tcl_DeleteNamespace(namespacePtr);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceEvalCmd --
+ *
+ *	Invoked to implement the "namespace eval" command. Executes commands
+ *	in a namespace. If the namespace does not already exist, it is
+ *	created. Handles the following syntax:
+ *
+ *	    namespace eval name arg ?arg...?
+ *
+ *	If more than one arg argument is specified, the command that is
+ *	executed is the result of concatenating the arguments together with a
+ *	space between each argument.
+ *
+ * Results:
+ *	Returns TCL_OK if the namespace is found and the commands are executed
+ *	successfully. Returns TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns the result of the command in the interpreter's result object.
+ *	If anything goes wrong, this function returns an error message as the
+ *	result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceEvalCmd(
+    ClientData clientData,	/* Arbitrary value passed to cmd. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, NRNamespaceEvalCmd, clientData, objc,
+	    objv);
+}
+
+static int
+NRNamespaceEvalCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    CmdFrame *invoker;
+    int word;
+    Tcl_Namespace *namespacePtr;
+    CallFrame *framePtr, **framePtrPtr;
+    Tcl_Obj *objPtr;
+    int result;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name arg ?arg...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Try to resolve the namespace reference, caching the result in the
+     * namespace object along the way.
+     */
+
+    result = GetNamespaceFromObj(interp, objv[1], &namespacePtr);
+
+    /*
+     * If the namespace wasn't found, try to create it.
+     */
+
+    if (result == TCL_ERROR) {
+	const char *name = TclGetString(objv[1]);
+
+	namespacePtr = Tcl_CreateNamespace(interp, name, NULL, NULL);
+	if (namespacePtr == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Make the specified namespace the current namespace and evaluate the
+     * command(s).
+     */
+
+    /* This is needed to satisfy GCC 3.3's strict aliasing rules */
+    framePtrPtr = &framePtr;
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    namespacePtr, /*isProcCallFrame*/ 0);
+
+    framePtr->objv = TclFetchEnsembleRoot(interp, objv, objc, &framePtr->objc);
+
+    if (objc == 3) {
+	/*
+	 * TIP #280: Make actual argument location available to eval'd script.
+	 */
+
+	objPtr = objv[2];
+	invoker = iPtr->cmdFramePtr;
+	word = 3;
+	TclArgumentGet(interp, objPtr, &invoker, &word);
+    } else {
+	/*
+	 * More than one argument: concatenate them together with spaces
+	 * between, then evaluate the result. Tcl_EvalObjEx will delete the
+	 * object when it decrements its refcount after eval'ing it.
+	 */
+
+	objPtr = Tcl_ConcatObj(objc-2, objv+2);
+	invoker = NULL;
+	word = 0;
+    }
+
+    /*
+     * TIP #280: Make invoking context available to eval'd script.
+     */
+
+    TclNRAddCallback(interp, NsEval_Callback, namespacePtr, "eval",
+	    NULL, NULL);
+    return TclNREvalObjEx(interp, objPtr, 0, invoker, word);
+}
+
+static int
+NsEval_Callback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Namespace *namespacePtr = data[0];
+
+    if (result == TCL_ERROR) {
+	int length = strlen(namespacePtr->fullName);
+	int limit = 200;
+	int overflow = (length > limit);
+	char *cmd = data[1];
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (in namespace %s \"%.*s%s\" script line %d)",
+		cmd,
+		(overflow ? limit : length), namespacePtr->fullName,
+		(overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+    }
+
+    /*
+     * Restore the previous "current" namespace.
+     */
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceExistsCmd --
+ *
+ *	Invoked to implement the "namespace exists" command that returns true
+ *	if the given namespace currently exists, and false otherwise. Handles
+ *	the following syntax:
+ *
+ *	    namespace exists name
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceExistsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *namespacePtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(
+	    GetNamespaceFromObj(interp, objv[1], &namespacePtr) == TCL_OK));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceExportCmd --
+ *
+ *	Invoked to implement the "namespace export" command that specifies
+ *	which commands are exported from a namespace. The exported commands
+ *	are those that can be imported into another namespace using "namespace
+ *	import". Both commands defined in a namespace and commands the
+ *	namespace has imported can be exported by a namespace. This command
+ *	has the following syntax:
+ *
+ *	    namespace export ?-clear? ?pattern pattern...?
+ *
+ *	Each pattern may contain "string match"-style pattern matching special
+ *	characters, but the pattern may not include any namespace qualifiers:
+ *	that is, the pattern must specify commands in the current (exporting)
+ *	namespace. The specified patterns are appended onto the namespace's
+ *	list of export patterns.
+ *
+ *	To reset the namespace's export pattern list, specify the "-clear"
+ *	flag.
+ *
+ *	If there are no export patterns and the "-clear" flag isn't given,
+ *	this command returns the namespace's current export list.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceExportCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int firstArg, i;
+
+    if (objc < 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-clear? ?pattern pattern...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If no pattern arguments are given, and "-clear" isn't specified, return
+     * the namespace's current export pattern list.
+     */
+
+    if (objc == 1) {
+	Tcl_Obj *listPtr = Tcl_NewObj();
+
+	(void) Tcl_AppendExportList(interp, NULL, listPtr);
+	Tcl_SetObjResult(interp, listPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Process the optional "-clear" argument.
+     */
+
+    firstArg = 1;
+    if (strcmp("-clear", Tcl_GetString(objv[firstArg])) == 0) {
+	Tcl_Export(interp, NULL, "::", 1);
+	Tcl_ResetResult(interp);
+	firstArg++;
+    }
+
+    /*
+     * Add each pattern to the namespace's export pattern list.
+     */
+
+    for (i = firstArg;  i < objc;  i++) {
+	int result = Tcl_Export(interp, NULL, Tcl_GetString(objv[i]), 0);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceForgetCmd --
+ *
+ *	Invoked to implement the "namespace forget" command to remove imported
+ *	commands from a namespace. Handles the following syntax:
+ *
+ *	    namespace forget ?pattern pattern...?
+ *
+ *	Each pattern is a name like "foo::*" or "a::b::x*". That is, the
+ *	pattern may include the special pattern matching characters recognized
+ *	by the "string match" command, but only in the command name at the end
+ *	of the qualified name; the special pattern characters may not appear
+ *	in a namespace name. All of the commands that match that pattern are
+ *	checked to see if they have an imported command in the current
+ *	namespace that refers to the matched command. If there is an alias, it
+ *	is removed.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Imported commands are removed from the current namespace. If anything
+ *	goes wrong, this function returns an error message in the
+ *	interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceForgetCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *pattern;
+    register int i, result;
+
+    if (objc < 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern pattern...?");
+	return TCL_ERROR;
+    }
+
+    for (i = 1;  i < objc;  i++) {
+	pattern = TclGetString(objv[i]);
+	result = Tcl_ForgetImport(interp, NULL, pattern);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceImportCmd --
+ *
+ *	Invoked to implement the "namespace import" command that imports
+ *	commands into a namespace. Handles the following syntax:
+ *
+ *	    namespace import ?-force? ?pattern pattern...?
+ *
+ *	Each pattern is a namespace-qualified name like "foo::*", "a::b::x*",
+ *	or "bar::p". That is, the pattern may include the special pattern
+ *	matching characters recognized by the "string match" command, but only
+ *	in the command name at the end of the qualified name; the special
+ *	pattern characters may not appear in a namespace name. All of the
+ *	commands that match the pattern and which are exported from their
+ *	namespace are made accessible from the current namespace context. This
+ *	is done by creating a new "imported command" in the current namespace
+ *	that points to the real command in its original namespace; when the
+ *	imported command is called, it invokes the real command.
+ *
+ *	If an imported command conflicts with an existing command, it is
+ *	treated as an error. But if the "-force" option is included, then
+ *	existing commands are overwritten by the imported commands.
+ *
+ *	If there are no pattern arguments and the "-force" flag isn't given,
+ *	this command returns the list of commands currently imported in
+ *	the current namespace.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Adds imported commands to the current namespace. If anything goes
+ *	wrong, this function returns an error message in the interpreter's
+ *	result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceImportCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int allowOverwrite = 0;
+    const char *string, *pattern;
+    register int i, result;
+    int firstArg;
+
+    if (objc < 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?-force? ?pattern pattern...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Skip over the optional "-force" as the first argument.
+     */
+
+    firstArg = 1;
+    if (firstArg < objc) {
+	string = TclGetString(objv[firstArg]);
+	if ((*string == '-') && (strcmp(string, "-force") == 0)) {
+	    allowOverwrite = 1;
+	    firstArg++;
+	}
+    } else {
+	/*
+	 * When objc == 1, command is just [namespace import]. Introspection
+	 * form to return list of imported commands.
+	 */
+
+	Tcl_HashEntry *hPtr;
+	Tcl_HashSearch search;
+	Namespace *nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+	Tcl_Obj *listPtr;
+
+	TclNewObj(listPtr);
+	for (hPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+		hPtr != NULL; hPtr = Tcl_NextHashEntry(&search)) {
+	    Command *cmdPtr = Tcl_GetHashValue(hPtr);
+
+	    if (cmdPtr->deleteProc == DeleteImportedCmd) {
+		Tcl_ListObjAppendElement(NULL, listPtr, Tcl_NewStringObj(
+			Tcl_GetHashKey(&nsPtr->cmdTable, hPtr) ,-1));
+	    }
+	}
+	Tcl_SetObjResult(interp, listPtr);
+	return TCL_OK;
+    }
+
+    /*
+     * Handle the imports for each of the patterns.
+     */
+
+    for (i = firstArg;  i < objc;  i++) {
+	pattern = TclGetString(objv[i]);
+	result = Tcl_Import(interp, NULL, pattern, allowOverwrite);
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceInscopeCmd --
+ *
+ *	Invoked to implement the "namespace inscope" command that executes a
+ *	script in the context of a particular namespace. This command is not
+ *	expected to be used directly by programmers; calls to it are generated
+ *	implicitly when programs use "namespace code" commands to register
+ *	callback scripts. Handles the following syntax:
+ *
+ *	    namespace inscope name arg ?arg...?
+ *
+ *	The "namespace inscope" command is much like the "namespace eval"
+ *	command except that it has lappend semantics and the namespace must
+ *	already exist. It treats the first argument as a list, and appends any
+ *	arguments after the first onto the end as proper list elements. For
+ *	example,
+ *
+ *	    namespace inscope ::foo {a b} c d e
+ *
+ *	is equivalent to
+ *
+ *	    namespace eval ::foo [concat {a b} [list c d e]]
+ *
+ *	This lappend semantics is important because many callback scripts are
+ *	actually prefixes.
+ *
+ * Results:
+ *	Returns TCL_OK to indicate success, or TCL_ERROR to indicate failure.
+ *
+ * Side effects:
+ *	Returns a result in the Tcl interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceInscopeCmd(
+    ClientData clientData,	/* Arbitrary value passed to cmd. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, NRNamespaceInscopeCmd, clientData, objc,
+	    objv);
+}
+
+static int
+NRNamespaceInscopeCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *namespacePtr;
+    CallFrame *framePtr, **framePtrPtr;
+    int i;
+    Tcl_Obj *cmdObjPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name arg ?arg...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Resolve the namespace reference.
+     */
+
+    if (TclGetNamespaceFromObj(interp, objv[1], &namespacePtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the specified namespace the current namespace.
+     */
+
+    framePtrPtr = &framePtr;		/* This is needed to satisfy GCC's
+					 * strict aliasing rules. */
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    namespacePtr, /*isProcCallFrame*/ 0);
+
+    framePtr->objv = TclFetchEnsembleRoot(interp, objv, objc, &framePtr->objc);
+
+    /*
+     * Execute the command. If there is just one argument, just treat it as a
+     * script and evaluate it. Otherwise, create a list from the arguments
+     * after the first one, then concatenate the first argument and the list
+     * of extra arguments to form the command to evaluate.
+     */
+
+    if (objc == 3) {
+	cmdObjPtr = objv[2];
+    } else {
+	Tcl_Obj *concatObjv[2];
+	register Tcl_Obj *listPtr;
+
+	listPtr = Tcl_NewListObj(0, NULL);
+	for (i = 3;  i < objc;  i++) {
+	    if (Tcl_ListObjAppendElement(interp, listPtr, objv[i]) != TCL_OK){
+		Tcl_DecrRefCount(listPtr);	/* Free unneeded obj. */
+		return TCL_ERROR;
+	    }
+	}
+
+	concatObjv[0] = objv[2];
+	concatObjv[1] = listPtr;
+	cmdObjPtr = Tcl_ConcatObj(2, concatObjv);
+	Tcl_DecrRefCount(listPtr);    /* We're done with the list object. */
+    }
+
+    TclNRAddCallback(interp, NsEval_Callback, namespacePtr, "inscope",
+	    NULL, NULL);
+    return TclNREvalObjEx(interp, cmdObjPtr, 0, NULL, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceOriginCmd --
+ *
+ *	Invoked to implement the "namespace origin" command to return the
+ *	fully-qualified name of the "real" command to which the specified
+ *	"imported command" refers. Handles the following syntax:
+ *
+ *	    namespace origin name
+ *
+ * Results:
+ *	An imported command is created in an namespace when that namespace
+ *	imports a command from another namespace. If a command is imported
+ *	into a sequence of namespaces a, b,...,n where each successive
+ *	namespace just imports the command from the previous namespace, this
+ *	command returns the fully-qualified name of the original command in
+ *	the first namespace, a. If "name" does not refer to an alias, its
+ *	fully-qualified name is returned. The returned name is stored in the
+ *	interpreter's result object. This function returns TCL_OK if
+ *	successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	If anything goes wrong, this function returns an error message in the
+ *	interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceOriginCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Command command, origCommand;
+    Tcl_Obj *resultPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name");
+	return TCL_ERROR;
+    }
+
+    command = Tcl_GetCommandFromObj(interp, objv[1]);
+    if (command == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "invalid command name \"%s\"", TclGetString(objv[1])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND",
+		TclGetString(objv[1]), NULL);
+	return TCL_ERROR;
+    }
+    origCommand = TclGetOriginalCommand(command);
+    TclNewObj(resultPtr);
+    if (origCommand == NULL) {
+	/*
+	 * The specified command isn't an imported command. Return the
+	 * command's name qualified by the full name of the namespace it was
+	 * defined in.
+	 */
+
+	Tcl_GetCommandFullName(interp, command, resultPtr);
+    } else {
+	Tcl_GetCommandFullName(interp, origCommand, resultPtr);
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceParentCmd --
+ *
+ *	Invoked to implement the "namespace parent" command that returns the
+ *	fully-qualified name of the parent namespace for a specified
+ *	namespace. Handles the following syntax:
+ *
+ *	    namespace parent ?name?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceParentCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *nsPtr;
+
+    if (objc == 1) {
+	nsPtr = TclGetCurrentNamespace(interp);
+    } else if (objc == 2) {
+	if (TclGetNamespaceFromObj(interp, objv[1], &nsPtr) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?name?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Report the parent of the specified namespace.
+     */
+
+    if (nsPtr->parentPtr != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		nsPtr->parentPtr->fullName, -1));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespacePathCmd --
+ *
+ *	Invoked to implement the "namespace path" command that reads and
+ *	writes the current namespace's command resolution path. Has one
+ *	optional argument: if present, it is a list of named namespaces to set
+ *	the path to, and if absent, the current path should be returned.
+ *	Handles the following syntax:
+ *
+ *	    namespace path ?nsList?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong
+ *	(most notably if the namespace list contains the name of something
+ *	other than a namespace). In the successful-exit case, may set the
+ *	interpreter result to the list of names of the namespaces on the
+ *	current namespace's path.
+ *
+ * Side effects:
+ *	May update the namespace path (triggering a recomputing of all command
+ *	names that depend on the namespace for resolution).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespacePathCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Namespace *nsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    int i, nsObjc, result = TCL_ERROR;
+    Tcl_Obj **nsObjv;
+    Tcl_Namespace **namespaceList = NULL;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pathList?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If no path is given, return the current path.
+     */
+
+    if (objc == 1) {
+	Tcl_Obj *resultObj = Tcl_NewObj();
+
+	for (i=0 ; i<nsPtr->commandPathLength ; i++) {
+	    if (nsPtr->commandPathArray[i].nsPtr != NULL) {
+		Tcl_ListObjAppendElement(NULL, resultObj, Tcl_NewStringObj(
+			nsPtr->commandPathArray[i].nsPtr->fullName, -1));
+	    }
+	}
+	Tcl_SetObjResult(interp, resultObj);
+	return TCL_OK;
+    }
+
+    /*
+     * There is a path given, so parse it into an array of namespace pointers.
+     */
+
+    if (TclListObjGetElements(interp, objv[1], &nsObjc, &nsObjv) != TCL_OK) {
+	goto badNamespace;
+    }
+    if (nsObjc != 0) {
+	namespaceList = TclStackAlloc(interp,
+		sizeof(Tcl_Namespace *) * nsObjc);
+
+	for (i=0 ; i<nsObjc ; i++) {
+	    if (TclGetNamespaceFromObj(interp, nsObjv[i],
+		    &namespaceList[i]) != TCL_OK) {
+		goto badNamespace;
+	    }
+	}
+    }
+
+    /*
+     * Now we have the list of valid namespaces, install it as the path.
+     */
+
+    TclSetNsPath(nsPtr, nsObjc, namespaceList);
+
+    result = TCL_OK;
+  badNamespace:
+    if (namespaceList != NULL) {
+	TclStackFree(interp, namespaceList);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetNsPath --
+ *
+ *	Sets the namespace command name resolution path to the given list of
+ *	namespaces. If the list is empty (of zero length) the path is set to
+ *	empty and the default old-style behaviour of command name resolution
+ *	is used.
+ *
+ * Results:
+ *	nothing
+ *
+ * Side effects:
+ *	Invalidates the command name resolution caches for any command
+ *	resolved in the given namespace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetNsPath(
+    Namespace *nsPtr,		/* Namespace whose path is to be set. */
+    int pathLength,		/* Length of pathAry. */
+    Tcl_Namespace *pathAry[])	/* Array of namespaces that are the path. */
+{
+    if (pathLength != 0) {
+	NamespacePathEntry *tmpPathArray =
+		ckalloc(sizeof(NamespacePathEntry) * pathLength);
+	int i;
+
+	for (i=0 ; i<pathLength ; i++) {
+	    tmpPathArray[i].nsPtr = (Namespace *) pathAry[i];
+	    tmpPathArray[i].creatorNsPtr = nsPtr;
+	    tmpPathArray[i].prevPtr = NULL;
+	    tmpPathArray[i].nextPtr =
+		    tmpPathArray[i].nsPtr->commandPathSourceList;
+	    if (tmpPathArray[i].nextPtr != NULL) {
+		tmpPathArray[i].nextPtr->prevPtr = &tmpPathArray[i];
+	    }
+	    tmpPathArray[i].nsPtr->commandPathSourceList = &tmpPathArray[i];
+	}
+	if (nsPtr->commandPathLength != 0) {
+	    UnlinkNsPath(nsPtr);
+	}
+	nsPtr->commandPathArray = tmpPathArray;
+    } else {
+	if (nsPtr->commandPathLength != 0) {
+	    UnlinkNsPath(nsPtr);
+	}
+    }
+
+    nsPtr->commandPathLength = pathLength;
+    nsPtr->cmdRefEpoch++;
+    nsPtr->resolverEpoch++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UnlinkNsPath --
+ *
+ *	Delete the given namespace's command name resolution path. Only call
+ *	if the path is non-empty. Caller must reset the counter containing the
+ *	path size.
+ *
+ * Results:
+ *	nothing
+ *
+ * Side effects:
+ *	Deletes the array of path entries and unlinks those path entries from
+ *	the target namespace's list of interested namespaces.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UnlinkNsPath(
+    Namespace *nsPtr)
+{
+    int i;
+    for (i=0 ; i<nsPtr->commandPathLength ; i++) {
+	NamespacePathEntry *nsPathPtr = &nsPtr->commandPathArray[i];
+
+	if (nsPathPtr->prevPtr != NULL) {
+	    nsPathPtr->prevPtr->nextPtr = nsPathPtr->nextPtr;
+	}
+	if (nsPathPtr->nextPtr != NULL) {
+	    nsPathPtr->nextPtr->prevPtr = nsPathPtr->prevPtr;
+	}
+	if (nsPathPtr->nsPtr != NULL) {
+	    if (nsPathPtr->nsPtr->commandPathSourceList == nsPathPtr) {
+		nsPathPtr->nsPtr->commandPathSourceList = nsPathPtr->nextPtr;
+	    }
+	}
+    }
+    ckfree(nsPtr->commandPathArray);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInvalidateNsPath --
+ *
+ *	Invalidate the name resolution caches for all names looked up in
+ *	namespaces whose name path includes the given namespace.
+ *
+ * Results:
+ *	nothing
+ *
+ * Side effects:
+ *	Increments the command reference epoch in each namespace whose path
+ *	includes the given namespace. This causes any cached resolved names
+ *	whose root cacheing context starts at that namespace to be recomputed
+ *	the next time they are used.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInvalidateNsPath(
+    Namespace *nsPtr)
+{
+    NamespacePathEntry *nsPathPtr = nsPtr->commandPathSourceList;
+
+    while (nsPathPtr != NULL) {
+	if (nsPathPtr->nsPtr != NULL) {
+	    nsPathPtr->creatorNsPtr->cmdRefEpoch++;
+	}
+	nsPathPtr = nsPathPtr->nextPtr;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceQualifiersCmd --
+ *
+ *	Invoked to implement the "namespace qualifiers" command that returns
+ *	any leading namespace qualifiers in a string. These qualifiers are
+ *	namespace names separated by "::"s. For example, for "::foo::p" this
+ *	command returns "::foo", and for "::" it returns "". This command is
+ *	the complement of the "namespace tail" command. Note that this command
+ *	does not check whether the "namespace" names are, in fact, the names
+ *	of currently defined namespaces. Handles the following syntax:
+ *
+ *	    namespace qualifiers string
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceQualifiersCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register const char *name, *p;
+    int length;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the end of the string, then work backward and find the start of
+     * the last "::" qualifier.
+     */
+
+    name = TclGetString(objv[1]);
+    for (p = name;  *p != '\0';  p++) {
+	/* empty body */
+    }
+    while (--p >= name) {
+	if ((*p == ':') && (p > name) && (*(p-1) == ':')) {
+	    p -= 2;			/* Back up over the :: */
+	    while ((p >= name) && (*p == ':')) {
+		p--;			/* Back up over the preceeding : */
+	    }
+	    break;
+	}
+    }
+
+    if (p >= name) {
+	length = p-name+1;
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(name, length));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceUnknownCmd --
+ *
+ *	Invoked to implement the "namespace unknown" command (TIP 181) that
+ *	sets or queries a per-namespace unknown command handler. This handler
+ *	is called when command lookup fails (current and global ns). The
+ *	default handler for the global namespace is ::unknown. The default
+ *	handler for other namespaces is to call the global namespace unknown
+ *	handler. Passing an empty list results in resetting the handler to its
+ *	default.
+ *
+ *	    namespace unknown ?handler?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	If no handler is specified, returns a result in the interpreter's
+ *	result object, otherwise it sets the unknown handler pointer in the
+ *	current namespace to the script fragment provided. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceUnknownCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Namespace *currNsPtr;
+    Tcl_Obj *resultPtr;
+    int rc;
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "?script?");
+	return TCL_ERROR;
+    }
+
+    currNsPtr = TclGetCurrentNamespace(interp);
+
+    if (objc == 1) {
+	/*
+	 * Introspection - return the current namespace handler.
+	 */
+
+	resultPtr = Tcl_GetNamespaceUnknownHandler(interp, currNsPtr);
+	if (resultPtr == NULL) {
+	    TclNewObj(resultPtr);
+	}
+	Tcl_SetObjResult(interp, resultPtr);
+    } else {
+	rc = Tcl_SetNamespaceUnknownHandler(interp, currNsPtr, objv[1]);
+	if (rc == TCL_OK) {
+	    Tcl_SetObjResult(interp, objv[1]);
+	}
+	return rc;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetNamespaceUnknownHandler --
+ *
+ *	Returns the unknown command handler registered for the given
+ *	namespace.
+ *
+ * Results:
+ *	Returns the current unknown command handler, or NULL if none exists
+ *	for the namespace.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetNamespaceUnknownHandler(
+    Tcl_Interp *interp,		/* The interpreter in which the namespace
+				 * exists. */
+    Tcl_Namespace *nsPtr)	/* The namespace. */
+{
+    Namespace *currNsPtr = (Namespace *) nsPtr;
+
+    if (currNsPtr->unknownHandlerPtr == NULL &&
+	    currNsPtr == ((Interp *) interp)->globalNsPtr) {
+	/*
+	 * Default handler for global namespace is "::unknown". For all other
+	 * namespaces, it is NULL (which falls back on the global unknown
+	 * handler).
+	 */
+
+	TclNewLiteralStringObj(currNsPtr->unknownHandlerPtr, "::unknown");
+	Tcl_IncrRefCount(currNsPtr->unknownHandlerPtr);
+    }
+    return currNsPtr->unknownHandlerPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetNamespaceUnknownHandler --
+ *
+ *	Sets the unknown command handler for the given namespace to the
+ *	command prefix passed.
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Sets the namespace unknown command handler. If the passed in handler
+ *	is NULL or an empty list, then the handler is reset to its default. If
+ *	an error occurs, then an error message is left in the interpreter
+ *	result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetNamespaceUnknownHandler(
+    Tcl_Interp *interp,		/* Interpreter in which the namespace
+				 * exists. */
+    Tcl_Namespace *nsPtr,	/* Namespace which is being updated. */
+    Tcl_Obj *handlerPtr)	/* The new handler, or NULL to reset. */
+{
+    int lstlen = 0;
+    Namespace *currNsPtr = (Namespace *) nsPtr;
+
+    /*
+     * Ensure that we check for errors *first* before we change anything.
+     */
+
+    if (handlerPtr != NULL) {
+	if (TclListObjLength(interp, handlerPtr, &lstlen) != TCL_OK) {
+	    /*
+	     * Not a list.
+	     */
+
+	    return TCL_ERROR;
+	}
+	if (lstlen > 0) {
+	    /*
+	     * We are going to be saving this handler. Increment the reference
+	     * count before decrementing the refcount on the previous handler,
+	     * so that nothing strange can happen if we are told to set the
+	     * handler to the previous value.
+	     */
+
+	    Tcl_IncrRefCount(handlerPtr);
+	}
+    }
+
+    /*
+     * Remove old handler next.
+     */
+
+    if (currNsPtr->unknownHandlerPtr != NULL) {
+	Tcl_DecrRefCount(currNsPtr->unknownHandlerPtr);
+    }
+
+    /*
+     * Install the new handler.
+     */
+
+    if (lstlen > 0) {
+	/*
+	 * Just store the handler. It already has the correct reference count.
+	 */
+
+	currNsPtr->unknownHandlerPtr = handlerPtr;
+    } else {
+	/*
+	 * If NULL or an empty list is passed, this resets to the default
+	 * handler.
+	 */
+
+	currNsPtr->unknownHandlerPtr = NULL;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceTailCmd --
+ *
+ *	Invoked to implement the "namespace tail" command that returns the
+ *	trailing name at the end of a string with "::" namespace qualifiers.
+ *	These qualifiers are namespace names separated by "::"s. For example,
+ *	for "::foo::p" this command returns "p", and for "::" it returns "".
+ *	This command is the complement of the "namespace qualifiers" command.
+ *	Note that this command does not check whether the "namespace" names
+ *	are, in fact, the names of currently defined namespaces. Handles the
+ *	following syntax:
+ *
+ *	    namespace tail string
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceTailCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register const char *name, *p;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "string");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the end of the string, then work backward and find the last "::"
+     * qualifier.
+     */
+
+    name = TclGetString(objv[1]);
+    for (p = name;  *p != '\0';  p++) {
+	/* empty body */
+    }
+    while (--p > name) {
+	if ((*p == ':') && (*(p-1) == ':')) {
+	    p++;			/* Just after the last "::" */
+	    break;
+	}
+    }
+
+    if (p >= name) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(p, -1));
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceUpvarCmd --
+ *
+ *	Invoked to implement the "namespace upvar" command, that creates
+ *	variables in the current scope linked to variables in another
+ *	namespace. Handles the following syntax:
+ *
+ *	    namespace upvar ns otherVar myVar ?otherVar myVar ...?
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Creates new variables in the current scope, linked to the
+ *	corresponding variables in the stipulated nmamespace. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceUpvarCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Namespace *nsPtr, *savedNsPtr;
+    Var *otherPtr, *arrayPtr;
+    const char *myName;
+
+    if (objc < 2 || (objc & 1)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "ns ?otherVar myVar ...?");
+	return TCL_ERROR;
+    }
+
+    if (TclGetNamespaceFromObj(interp, objv[1], &nsPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    objc -= 2;
+    objv += 2;
+
+    for (; objc>0 ; objc-=2, objv+=2) {
+	/*
+	 * Locate the other variable.
+	 */
+
+	savedNsPtr = (Tcl_Namespace *) iPtr->varFramePtr->nsPtr;
+	iPtr->varFramePtr->nsPtr = (Namespace *) nsPtr;
+	otherPtr = TclObjLookupVarEx(interp, objv[0], NULL,
+		(TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG|TCL_AVOID_RESOLVERS),
+		"access", /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+	iPtr->varFramePtr->nsPtr = (Namespace *) savedNsPtr;
+	if (otherPtr == NULL) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Create the new variable and link it to otherPtr.
+	 */
+
+	myName = TclGetString(objv[1]);
+	if (TclPtrMakeUpvar(interp, otherPtr, myName, 0, -1) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NamespaceWhichCmd --
+ *
+ *	Invoked to implement the "namespace which" command that returns the
+ *	fully-qualified name of a command or variable. If the specified
+ *	command or variable does not exist, it returns "". Handles the
+ *	following syntax:
+ *
+ *	    namespace which ?-command? ?-variable? name
+ *
+ * Results:
+ *	Returns TCL_OK if successful, and TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If anything goes
+ *	wrong, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NamespaceWhichCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const opts[] = {
+	"-command", "-variable", NULL
+    };
+    int lookupType = 0;
+    Tcl_Obj *resultPtr;
+
+    if (objc < 2 || objc > 3) {
+    badArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "?-command? ?-variable? name");
+	return TCL_ERROR;
+    } else if (objc == 3) {
+	/*
+	 * Look for a flag controlling the lookup.
+	 */
+
+	if (Tcl_GetIndexFromObj(interp, objv[1], opts, "option", 0,
+		&lookupType) != TCL_OK) {
+	    /*
+	     * Preserve old style of error message!
+	     */
+
+	    Tcl_ResetResult(interp);
+	    goto badArgs;
+	}
+    }
+
+    TclNewObj(resultPtr);
+    switch (lookupType) {
+    case 0: {				/* -command */
+	Tcl_Command cmd = Tcl_GetCommandFromObj(interp, objv[objc-1]);
+
+	if (cmd != NULL) {
+	    Tcl_GetCommandFullName(interp, cmd, resultPtr);
+	}
+	break;
+    }
+    case 1: {				/* -variable */
+	Tcl_Var var = Tcl_FindNamespaceVar(interp,
+		TclGetString(objv[objc-1]), NULL, /*flags*/ 0);
+
+	if (var != NULL) {
+	    Tcl_GetVariableFullName(interp, var, resultPtr);
+	}
+	break;
+    }
+    }
+    Tcl_SetObjResult(interp, resultPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeNsNameInternalRep --
+ *
+ *	Frees the resources associated with a nsName object's internal
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Decrements the ref count of any Namespace structure pointed to by the
+ *	nsName's internal representation. If there are no more references to
+ *	the namespace, it's structure will be freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeNsNameInternalRep(
+    register Tcl_Obj *objPtr)	/* nsName object with internal representation
+				 * to free. */
+{
+    ResolvedNsName *resNamePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    /*
+     * Decrement the reference count of the namespace. If there are no more
+     * references, free it up.
+     */
+
+    if (resNamePtr->refCount-- <= 1) {
+	/*
+	 * Decrement the reference count for the cached namespace. If the
+	 * namespace is dead, and there are no more references to it, free
+	 * it.
+	 */
+
+	TclNsDecrRefCount(resNamePtr->nsPtr);
+	ckfree(resNamePtr);
+    }
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupNsNameInternalRep --
+ *
+ *	Initializes the internal representation of a nsName object to a copy
+ *	of the internal representation of another nsName object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	copyPtr's internal rep is set to refer to the same namespace
+ *	referenced by srcPtr's internal rep. Increments the ref count of the
+ *	ResolvedNsName structure used to hold the namespace reference.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupNsNameInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    register Tcl_Obj *copyPtr)	/* Object with internal rep to set. */
+{
+    ResolvedNsName *resNamePtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    copyPtr->internalRep.twoPtrValue.ptr1 = resNamePtr;
+    resNamePtr->refCount++;
+    copyPtr->typePtr = &nsNameType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetNsNameFromAny --
+ *
+ *	Attempt to generate a nsName internal representation for a Tcl object.
+ *
+ * Results:
+ *	Returns TCL_OK if the value could be converted to a proper namespace
+ *	reference. Otherwise, it returns TCL_ERROR, along with an error
+ *	message in the interpreter's result object.
+ *
+ * Side effects:
+ *	If successful, the object is made a nsName object. Its internal rep is
+ *	set to point to a ResolvedNsName, which contains a cached pointer to
+ *	the Namespace. Reference counts are kept on both the ResolvedNsName
+ *	and the Namespace, so we can keep track of their usage and free them
+ *	when appropriate.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetNsNameFromAny(
+    Tcl_Interp *interp,		/* Points to the namespace in which to resolve
+				 * name. Also used for error reporting if not
+				 * NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    const char *dummy;
+    Namespace *nsPtr, *dummy1Ptr, *dummy2Ptr;
+    register ResolvedNsName *resNamePtr;
+    const char *name;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+
+    name = TclGetString(objPtr);
+    TclGetNamespaceForQualName(interp, name, NULL, TCL_FIND_ONLY_NS,
+	     &nsPtr, &dummy1Ptr, &dummy2Ptr, &dummy);
+
+    /*
+     * If we found a namespace, then create a new ResolvedNsName structure
+     * that holds a reference to it.
+     */
+
+    if ((nsPtr == NULL) || (nsPtr->flags & NS_DYING)) {
+	/*
+	 * Our failed lookup proves any previously cached nsName intrep is no
+	 * longer valid. Get rid of it so we no longer waste memory storing
+	 * it, nor time determining its invalidity again and again.
+	 */
+
+	if (objPtr->typePtr == &nsNameType) {
+	    TclFreeIntRep(objPtr);
+	}
+	return TCL_ERROR;
+    }
+
+    nsPtr->refCount++;
+    resNamePtr = ckalloc(sizeof(ResolvedNsName));
+    resNamePtr->nsPtr = nsPtr;
+    if ((name[0] == ':') && (name[1] == ':')) {
+	resNamePtr->refNsPtr = NULL;
+    } else {
+	resNamePtr->refNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    }
+    resNamePtr->refCount = 1;
+    TclFreeIntRep(objPtr);
+    objPtr->internalRep.twoPtrValue.ptr1 = resNamePtr;
+    objPtr->typePtr = &nsNameType;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetNamespaceCommandTable --
+ *
+ *	Returns the hash table of commands.
+ *
+ * Results:
+ *	Pointer to the hash table.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashTable *
+TclGetNamespaceCommandTable(
+    Tcl_Namespace *nsPtr)
+{
+    return &((Namespace *) nsPtr)->cmdTable;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetNamespaceChildTable --
+ *
+ *	Returns the hash table of child namespaces.
+ *
+ * Results:
+ *	Pointer to the hash table.
+ *
+ * Side effects:
+ *	Might allocate memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_HashTable *
+TclGetNamespaceChildTable(
+    Tcl_Namespace *nsPtr)
+{
+    Namespace *nPtr = (Namespace *) nsPtr;
+#ifndef BREAK_NAMESPACE_COMPAT
+    return &nPtr->childTable;
+#else
+    if (nPtr->childTablePtr == NULL) {
+	nPtr->childTablePtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(nPtr->childTablePtr, TCL_STRING_KEYS);
+    }
+    return nPtr->childTablePtr;
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLogCommandInfo --
+ *
+ *	This function is invoked after an error occurs in an interpreter. It
+ *	adds information to iPtr->errorInfo/errorStack fields to describe the
+ *	command that was being executed when the error occurred. When pc and
+ *	tosPtr are non-NULL, conveying a bytecode execution "inner context",
+ *	and the offending instruction is suitable, that inner context is
+ *	recorded in errorStack.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Information about the command is added to errorInfo/errorStack and the
+ *	line number stored internally in the interpreter is set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclLogCommandInfo(
+    Tcl_Interp *interp,		/* Interpreter in which to log information. */
+    const char *script,		/* First character in script containing
+				 * command (must be <= command). */
+    const char *command,	/* First character in command that generated
+				 * the error. */
+    int length,			/* Number of bytes in command (-1 means use
+				 * all bytes up to first null byte). */
+    const unsigned char *pc,    /* Current pc of bytecode execution context */
+    Tcl_Obj **tosPtr)		/* Current stack of bytecode execution
+				 * context */
+{
+    register const char *p;
+    Interp *iPtr = (Interp *) interp;
+    int overflow, limit = 150;
+    Var *varPtr, *arrayPtr;
+
+    if (iPtr->flags & ERR_ALREADY_LOGGED) {
+	/*
+	 * Someone else has already logged error information for this command;
+	 * we shouldn't add anything more.
+	 */
+
+	return;
+    }
+
+    if (command != NULL) {
+	/*
+	 * Compute the line number where the error occurred.
+	 */
+
+	iPtr->errorLine = 1;
+	for (p = script; p != command; p++) {
+	    if (*p == '\n') {
+		iPtr->errorLine++;
+	    }
+	}
+
+	if (length < 0) {
+	    length = strlen(command);
+	}
+	overflow = (length > limit);
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    %s\n\"%.*s%s\"", ((iPtr->errorInfo == NULL)
+		? "while executing" : "invoked from within"),
+		(overflow ? limit : length), command,
+		(overflow ? "..." : "")));
+
+	varPtr = TclObjLookupVarEx(interp, iPtr->eiVar, NULL, TCL_GLOBAL_ONLY,
+		NULL, 0, 0, &arrayPtr);
+	if ((varPtr == NULL) || !TclIsVarTraced(varPtr)) {
+	    /*
+	     * Should not happen.
+	     */
+
+	    return;
+	} else {
+	    Tcl_HashEntry *hPtr
+		    = Tcl_FindHashEntry(&iPtr->varTraces, (char *) varPtr);
+	    VarTrace *tracePtr = Tcl_GetHashValue(hPtr);
+
+	    if (tracePtr->traceProc != EstablishErrorInfoTraces) {
+		/*
+		 * The most recent trace set on ::errorInfo is not the one the
+		 * core itself puts on last. This means some other code is
+		 * tracing the variable, and the additional trace(s) might be
+		 * write traces that expect the timing of writes to
+		 * ::errorInfo that existed Tcl releases before 8.5. To
+		 * satisfy that compatibility need, we write the current
+		 * -errorinfo value to the ::errorInfo variable.
+		 */
+
+		Tcl_ObjSetVar2(interp, iPtr->eiVar, NULL, iPtr->errorInfo,
+			TCL_GLOBAL_ONLY);
+	    }
+	}
+    }
+
+    /*
+     * TIP #348
+     */
+
+    if (Tcl_IsShared(iPtr->errorStack)) {
+	Tcl_Obj *newObj;
+
+	newObj = Tcl_DuplicateObj(iPtr->errorStack);
+	Tcl_DecrRefCount(iPtr->errorStack);
+	Tcl_IncrRefCount(newObj);
+	iPtr->errorStack = newObj;
+    }
+    if (iPtr->resetErrorStack) {
+	int len;
+
+	iPtr->resetErrorStack = 0;
+	Tcl_ListObjLength(interp, iPtr->errorStack, &len);
+
+	/*
+	 * Reset while keeping the list intrep as much as possible.
+	 */
+
+	Tcl_ListObjReplace(interp, iPtr->errorStack, 0, len, 0, NULL);
+	if (pc != NULL) {
+	    Tcl_Obj *innerContext;
+
+	    innerContext = TclGetInnerContext(interp, pc, tosPtr);
+	    if (innerContext != NULL) {
+		Tcl_ListObjAppendElement(NULL, iPtr->errorStack,
+			iPtr->innerLiteral);
+		Tcl_ListObjAppendElement(NULL, iPtr->errorStack, innerContext);
+	    }
+	} else if (command != NULL) {
+	    Tcl_ListObjAppendElement(NULL, iPtr->errorStack,
+		    iPtr->innerLiteral);
+	    Tcl_ListObjAppendElement(NULL, iPtr->errorStack,
+		    Tcl_NewStringObj(command, length));
+	}
+    }
+
+    if (!iPtr->framePtr->objc) {
+	/*
+	 * Special frame, nothing to report.
+	 */
+    } else if (iPtr->varFramePtr != iPtr->framePtr) {
+	/*
+	 * uplevel case, [lappend errorstack UP $relativelevel]
+	 */
+
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack, iPtr->upLiteral);
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack, Tcl_NewIntObj(
+		iPtr->framePtr->level - iPtr->varFramePtr->level));
+    } else if (iPtr->framePtr != iPtr->rootFramePtr) {
+	/*
+	 * normal case, [lappend errorstack CALL [info level 0]]
+	 */
+
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack, iPtr->callLiteral);
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack, Tcl_NewListObj(
+		iPtr->framePtr->objc, iPtr->framePtr->objv));
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclErrorStackResetIf --
+ *
+ *	The TIP 348 reset/no-bc part of TLCI, for specific use by
+ *	TclCompileSyntaxError.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Reset errorstack if it needs be, and in that case remember the
+ *	passed-in error message as inner context.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclErrorStackResetIf(
+    Tcl_Interp *interp,
+    const char *msg,
+    int length)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (Tcl_IsShared(iPtr->errorStack)) {
+	Tcl_Obj *newObj;
+
+	newObj = Tcl_DuplicateObj(iPtr->errorStack);
+	Tcl_DecrRefCount(iPtr->errorStack);
+	Tcl_IncrRefCount(newObj);
+	iPtr->errorStack = newObj;
+    }
+    if (iPtr->resetErrorStack) {
+	int len;
+
+	iPtr->resetErrorStack = 0;
+	Tcl_ListObjLength(interp, iPtr->errorStack, &len);
+
+	/*
+	 * Reset while keeping the list intrep as much as possible.
+	 */
+
+	Tcl_ListObjReplace(interp, iPtr->errorStack, 0, len, 0, NULL);
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack, iPtr->innerLiteral);
+	Tcl_ListObjAppendElement(NULL, iPtr->errorStack,
+		Tcl_NewStringObj(msg, length));
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LogCommandInfo --
+ *
+ *	This function is invoked after an error occurs in an interpreter. It
+ *	adds information to iPtr->errorInfo/errorStack fields to describe the
+ *	command that was being executed when the error occurred.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Information about the command is added to errorInfo/errorStack and the
+ *	line number stored internally in the interpreter is set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_LogCommandInfo(
+    Tcl_Interp *interp,		/* Interpreter in which to log information. */
+    const char *script,		/* First character in script containing
+				 * command (must be <= command). */
+    const char *command,	/* First character in command that generated
+				 * the error. */
+    int length)			/* Number of bytes in command (-1 means use
+				 * all bytes up to first null byte). */
+{
+    TclLogCommandInfo(interp, script, command, length, NULL, NULL);
+}
+
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * End:
+ */
diff --git a/generic/tclNotify.c b/generic/tclNotify.c
new file mode 100644
index 0000000..e76bca8
--- /dev/null
+++ b/generic/tclNotify.c
@@ -0,0 +1,1141 @@
+/*
+ * tclNotify.c --
+ *
+ *	This file implements the generic portion of the Tcl notifier. The
+ *	notifier is lowest-level part of the event system. It manages an event
+ *	queue that holds Tcl_Event structures. The platform specific portion
+ *	of the notifier is defined in the tcl*Notify.c files in each platform
+ *	directory.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998 by Scriptics Corporation.
+ * Copyright (c) 2003 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Module-scope struct of notifier hooks that are checked in the default
+ * notifier functions (for overriding via Tcl_SetNotifier).
+ */
+
+Tcl_NotifierProcs tclNotifierHooks = {
+    NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL
+};
+
+/*
+ * For each event source (created with Tcl_CreateEventSource) there is a
+ * structure of the following type:
+ */
+
+typedef struct EventSource {
+    Tcl_EventSetupProc *setupProc;
+    Tcl_EventCheckProc *checkProc;
+    ClientData clientData;
+    struct EventSource *nextPtr;
+} EventSource;
+
+/*
+ * The following structure keeps track of the state of the notifier on a
+ * per-thread basis. The first three elements keep track of the event queue.
+ * In addition to the first (next to be serviced) and last events in the
+ * queue, we keep track of a "marker" event. This provides a simple priority
+ * mechanism whereby events can be inserted at the front of the queue but
+ * behind all other high-priority events already in the queue (this is used
+ * for things like a sequence of Enter and Leave events generated during a
+ * grab in Tk). These elements are protected by the queueMutex so that any
+ * thread can queue an event on any notifier. Note that all of the values in
+ * this structure will be initialized to 0.
+ */
+
+typedef struct ThreadSpecificData {
+    Tcl_Event *firstEventPtr;	/* First pending event, or NULL if none. */
+    Tcl_Event *lastEventPtr;	/* Last pending event, or NULL if none. */
+    Tcl_Event *markerEventPtr;	/* Last high-priority event in queue, or NULL
+				 * if none. */
+    Tcl_Mutex queueMutex;	/* Mutex to protect access to the previous
+				 * three fields. */
+    int serviceMode;		/* One of TCL_SERVICE_NONE or
+				 * TCL_SERVICE_ALL. */
+    int blockTimeSet;		/* 0 means there is no maximum block time:
+				 * block forever. */
+    Tcl_Time blockTime;		/* If blockTimeSet is 1, gives the maximum
+				 * elapsed time for the next block. */
+    int inTraversal;		/* 1 if Tcl_SetMaxBlockTime is being called
+				 * during an event source traversal. */
+    EventSource *firstEventSourcePtr;
+				/* Pointer to first event source in list of
+				 * event sources for this thread. */
+    Tcl_ThreadId threadId;	/* Thread that owns this notifier instance. */
+    ClientData clientData;	/* Opaque handle for platform specific
+				 * notifier. */
+    int initialized;		/* 1 if notifier has been initialized. */
+    struct ThreadSpecificData *nextPtr;
+				/* Next notifier in global list of notifiers.
+				 * Access is controlled by the listLock global
+				 * mutex. */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * Global list of notifiers. Access to this list is controlled by the listLock
+ * mutex. If this becomes a performance bottleneck, this could be replaced
+ * with a hashtable.
+ */
+
+static ThreadSpecificData *firstNotifierPtr = NULL;
+TCL_DECLARE_MUTEX(listLock)
+
+/*
+ * Declarations for routines used only in this file.
+ */
+
+static void		QueueEvent(ThreadSpecificData *tsdPtr,
+			    Tcl_Event *evPtr, Tcl_QueuePosition position);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitNotifier --
+ *
+ *	Initialize the thread local data structures for the notifier
+ *	subsystem.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Adds the current thread to the global list of notifiers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitNotifier(void)
+{
+    ThreadSpecificData *tsdPtr;
+    Tcl_ThreadId threadId = Tcl_GetCurrentThread();
+
+    Tcl_MutexLock(&listLock);
+    for (tsdPtr = firstNotifierPtr; tsdPtr && tsdPtr->threadId != threadId;
+	    tsdPtr = tsdPtr->nextPtr) {
+	/* Empty loop body. */
+    }
+
+    if (NULL == tsdPtr) {
+	/*
+	 * Notifier not yet initialized in this thread.
+	 */
+
+	tsdPtr = TCL_TSD_INIT(&dataKey);
+	tsdPtr->threadId = threadId;
+	tsdPtr->clientData = Tcl_InitNotifier();
+	tsdPtr->initialized = 1;
+	tsdPtr->nextPtr = firstNotifierPtr;
+	firstNotifierPtr = tsdPtr;
+    }
+    Tcl_MutexUnlock(&listLock);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeNotifier --
+ *
+ *	Finalize the thread local data structures for the notifier subsystem.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes the notifier associated with the current thread from the
+ *	global notifier list. This is done only if the notifier was
+ *	initialized for this thread by call to TclInitNotifier(). This is
+ *	always true for threads which have been seeded with an Tcl
+ *	interpreter, since the call to Tcl_CreateInterp will, among other
+ *	things, call TclInitializeSubsystems() and this one will, in turn,
+ *	call the TclInitNotifier() for the thread. For threads created without
+ *	the Tcl interpreter, though, nobody is explicitly nor implicitly
+ *	calling the TclInitNotifier hence, TclFinalizeNotifier should not be
+ *	performed at all.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeNotifier(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ThreadSpecificData **prevPtrPtr;
+    Tcl_Event *evPtr, *hold;
+
+    if (!tsdPtr->initialized) {
+	return;		/* Notifier not initialized for the current thread */
+    }
+
+    Tcl_MutexLock(&(tsdPtr->queueMutex));
+    for (evPtr = tsdPtr->firstEventPtr; evPtr != NULL; ) {
+	hold = evPtr;
+	evPtr = evPtr->nextPtr;
+	ckfree(hold);
+    }
+    tsdPtr->firstEventPtr = NULL;
+    tsdPtr->lastEventPtr = NULL;
+    Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+
+    Tcl_MutexLock(&listLock);
+
+    Tcl_FinalizeNotifier(tsdPtr->clientData);
+    Tcl_MutexFinalize(&(tsdPtr->queueMutex));
+    for (prevPtrPtr = &firstNotifierPtr; *prevPtrPtr != NULL;
+	    prevPtrPtr = &((*prevPtrPtr)->nextPtr)) {
+	if (*prevPtrPtr == tsdPtr) {
+	    *prevPtrPtr = tsdPtr->nextPtr;
+	    break;
+	}
+    }
+    tsdPtr->initialized = 0;
+
+    Tcl_MutexUnlock(&listLock);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetNotifier --
+ *
+ *	Install a set of alternate functions for use with the notifier. In
+ *	particular, this can be used to install the Xt-based notifier for use
+ *	with the Browser plugin.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Set the tclNotifierHooks global, which is checked in the default
+ *	notifier functions.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetNotifier(
+    Tcl_NotifierProcs *notifierProcPtr)
+{
+    tclNotifierHooks = *notifierProcPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateEventSource --
+ *
+ *	This function is invoked to create a new source of events. The source
+ *	is identified by a function that gets invoked during Tcl_DoOneEvent to
+ *	check for events on that source and queue them.
+ *
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	SetupProc and checkProc will be invoked each time that Tcl_DoOneEvent
+ *	runs out of things to do. SetupProc will be invoked before
+ *	Tcl_DoOneEvent calls select or whatever else it uses to wait for
+ *	events. SetupProc typically calls functions like Tcl_SetMaxBlockTime
+ *	to indicate what to wait for.
+ *
+ *	CheckProc is called after select or whatever operation was actually
+ *	used to wait. It figures out whether anything interesting actually
+ *	happened (e.g. by calling Tcl_AsyncReady), and then calls
+ *	Tcl_QueueEvent to queue any events that are ready.
+ *
+ *	Each of these functions is passed two arguments, e.g.
+ *		(*checkProc)(ClientData clientData, int flags));
+ *	ClientData is the same as the clientData argument here, and flags is a
+ *	combination of things like TCL_FILE_EVENTS that indicates what events
+ *	are of interest: setupProc and checkProc use flags to figure out
+ *	whether their events are relevant or not.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CreateEventSource(
+    Tcl_EventSetupProc *setupProc,
+				/* Function to invoke to figure out what to
+				 * wait for. */
+    Tcl_EventCheckProc *checkProc,
+				/* Function to call after waiting to see what
+				 * happened. */
+    ClientData clientData)	/* One-word argument to pass to setupProc and
+				 * checkProc. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    EventSource *sourcePtr = ckalloc(sizeof(EventSource));
+
+    sourcePtr->setupProc = setupProc;
+    sourcePtr->checkProc = checkProc;
+    sourcePtr->clientData = clientData;
+    sourcePtr->nextPtr = tsdPtr->firstEventSourcePtr;
+    tsdPtr->firstEventSourcePtr = sourcePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteEventSource --
+ *
+ *	This function is invoked to delete the source of events given by proc
+ *	and clientData.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The given event source is canceled, so its function will never again
+ *	be called. If no such source exists, nothing happens.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteEventSource(
+    Tcl_EventSetupProc *setupProc,
+				/* Function to invoke to figure out what to
+				 * wait for. */
+    Tcl_EventCheckProc *checkProc,
+				/* Function to call after waiting to see what
+				 * happened. */
+    ClientData clientData)	/* One-word argument to pass to setupProc and
+				 * checkProc. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    EventSource *sourcePtr, *prevPtr;
+
+    for (sourcePtr = tsdPtr->firstEventSourcePtr, prevPtr = NULL;
+	    sourcePtr != NULL;
+	    prevPtr = sourcePtr, sourcePtr = sourcePtr->nextPtr) {
+	if ((sourcePtr->setupProc != setupProc)
+		|| (sourcePtr->checkProc != checkProc)
+		|| (sourcePtr->clientData != clientData)) {
+	    continue;
+	}
+	if (prevPtr == NULL) {
+	    tsdPtr->firstEventSourcePtr = sourcePtr->nextPtr;
+	} else {
+	    prevPtr->nextPtr = sourcePtr->nextPtr;
+	}
+	ckfree(sourcePtr);
+	return;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_QueueEvent --
+ *
+ *	Queue an event on the event queue associated with the current thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_QueueEvent(
+    Tcl_Event *evPtr,		/* Event to add to queue. The storage space
+				 * must have been allocated the caller with
+				 * malloc (ckalloc), and it becomes the
+				 * property of the event queue. It will be
+				 * freed after the event has been handled. */
+    Tcl_QueuePosition position)	/* One of TCL_QUEUE_TAIL, TCL_QUEUE_HEAD,
+				 * TCL_QUEUE_MARK. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    QueueEvent(tsdPtr, evPtr, position);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ThreadQueueEvent --
+ *
+ *	Queue an event on the specified thread's event queue.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ThreadQueueEvent(
+    Tcl_ThreadId threadId,	/* Identifier for thread to use. */
+    Tcl_Event *evPtr,		/* Event to add to queue. The storage space
+				 * must have been allocated the caller with
+				 * malloc (ckalloc), and it becomes the
+				 * property of the event queue. It will be
+				 * freed after the event has been handled. */
+    Tcl_QueuePosition position)	/* One of TCL_QUEUE_TAIL, TCL_QUEUE_HEAD,
+				 * TCL_QUEUE_MARK. */
+{
+    ThreadSpecificData *tsdPtr;
+
+    /*
+     * Find the notifier associated with the specified thread.
+     */
+
+    Tcl_MutexLock(&listLock);
+    for (tsdPtr = firstNotifierPtr; tsdPtr && tsdPtr->threadId != threadId;
+	    tsdPtr = tsdPtr->nextPtr) {
+	/* Empty loop body. */
+    }
+
+    /*
+     * Queue the event if there was a notifier associated with the thread.
+     */
+
+    if (tsdPtr) {
+	QueueEvent(tsdPtr, evPtr, position);
+    } else {
+	ckfree(evPtr);
+    }
+    Tcl_MutexUnlock(&listLock);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * QueueEvent --
+ *
+ *	Insert an event into the specified thread's event queue at one of
+ *	three positions: the head, the tail, or before a floating marker.
+ *	Events inserted before the marker will be processed in first-in-
+ *	first-out order, but before any events inserted at the tail of the
+ *	queue. Events inserted at the head of the queue will be processed in
+ *	last-in-first-out order.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+QueueEvent(
+    ThreadSpecificData *tsdPtr,	/* Handle to thread local data that indicates
+				 * which event queue to use. */
+    Tcl_Event *evPtr,		/* Event to add to queue.  The storage space
+				 * must have been allocated the caller with
+				 * malloc (ckalloc), and it becomes the
+				 * property of the event queue. It will be
+				 * freed after the event has been handled. */
+    Tcl_QueuePosition position)	/* One of TCL_QUEUE_TAIL, TCL_QUEUE_HEAD,
+				 * TCL_QUEUE_MARK. */
+{
+    Tcl_MutexLock(&(tsdPtr->queueMutex));
+    if (position == TCL_QUEUE_TAIL) {
+	/*
+	 * Append the event on the end of the queue.
+	 */
+
+	evPtr->nextPtr = NULL;
+	if (tsdPtr->firstEventPtr == NULL) {
+	    tsdPtr->firstEventPtr = evPtr;
+	} else {
+	    tsdPtr->lastEventPtr->nextPtr = evPtr;
+	}
+	tsdPtr->lastEventPtr = evPtr;
+    } else if (position == TCL_QUEUE_HEAD) {
+	/*
+	 * Push the event on the head of the queue.
+	 */
+
+	evPtr->nextPtr = tsdPtr->firstEventPtr;
+	if (tsdPtr->firstEventPtr == NULL) {
+	    tsdPtr->lastEventPtr = evPtr;
+	}
+	tsdPtr->firstEventPtr = evPtr;
+    } else if (position == TCL_QUEUE_MARK) {
+	/*
+	 * Insert the event after the current marker event and advance the
+	 * marker to the new event.
+	 */
+
+	if (tsdPtr->markerEventPtr == NULL) {
+	    evPtr->nextPtr = tsdPtr->firstEventPtr;
+	    tsdPtr->firstEventPtr = evPtr;
+	} else {
+	    evPtr->nextPtr = tsdPtr->markerEventPtr->nextPtr;
+	    tsdPtr->markerEventPtr->nextPtr = evPtr;
+	}
+	tsdPtr->markerEventPtr = evPtr;
+	if (evPtr->nextPtr == NULL) {
+	    tsdPtr->lastEventPtr = evPtr;
+	}
+    }
+    Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteEvents --
+ *
+ *	Calls a function for each event in the queue and deletes those for
+ *	which the function returns 1. Events for which the function returns 0
+ *	are left in the queue. Operates on the queue associated with the
+ *	current thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Potentially removes one or more events from the event queue.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteEvents(
+    Tcl_EventDeleteProc *proc,	/* The function to call. */
+    ClientData clientData)	/* The type-specific data. */
+{
+    Tcl_Event *evPtr;		/* Pointer to the event being examined */
+    Tcl_Event *prevPtr;		/* Pointer to evPtr's predecessor, or NULL if
+				 * evPtr designates the first event in the
+				 * queue for the thread. */
+    Tcl_Event *hold;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    Tcl_MutexLock(&(tsdPtr->queueMutex));
+
+    /*
+     * Walk the queue of events for the thread, applying 'proc' to each to
+     * decide whether to eliminate the event.
+     */
+
+    prevPtr = NULL;
+    evPtr = tsdPtr->firstEventPtr;
+    while (evPtr != NULL) {
+	if (proc(evPtr, clientData) == 1) {
+	    /*
+	     * This event should be deleted. Unlink it.
+	     */
+
+	    if (prevPtr == NULL) {
+		tsdPtr->firstEventPtr = evPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = evPtr->nextPtr;
+	    }
+
+	    /*
+	     * Update 'last' and 'marker' events if either has been deleted.
+	     */
+
+	    if (evPtr->nextPtr == NULL) {
+		tsdPtr->lastEventPtr = prevPtr;
+	    }
+	    if (tsdPtr->markerEventPtr == evPtr) {
+		tsdPtr->markerEventPtr = prevPtr;
+	    }
+
+	    /*
+	     * Delete the event data structure.
+	     */
+
+	    hold = evPtr;
+	    evPtr = evPtr->nextPtr;
+	    ckfree(hold);
+	} else {
+	    /*
+	     * Event is to be retained.
+	     */
+
+	    prevPtr = evPtr;
+	    evPtr = evPtr->nextPtr;
+	}
+    }
+    Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ServiceEvent --
+ *
+ *	Process one event from the event queue, or invoke an asynchronous
+ *	event handler. Operates on event queue for current thread.
+ *
+ * Results:
+ *	The return value is 1 if the function actually found an event to
+ *	process. If no processing occurred, then 0 is returned.
+ *
+ * Side effects:
+ *	Invokes all of the event handlers for the highest priority event in
+ *	the event queue. May collapse some events into a single event or
+ *	discard stale events.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ServiceEvent(
+    int flags)			/* Indicates what events should be processed.
+				 * May be any combination of TCL_WINDOW_EVENTS
+				 * TCL_FILE_EVENTS, TCL_TIMER_EVENTS, or other
+				 * flags defined elsewhere. Events not
+				 * matching this will be skipped for
+				 * processing later. */
+{
+    Tcl_Event *evPtr, *prevPtr;
+    Tcl_EventProc *proc;
+    int result;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    /*
+     * Asynchronous event handlers are considered to be the highest priority
+     * events, and so must be invoked before we process events on the event
+     * queue.
+     */
+
+    if (Tcl_AsyncReady()) {
+	(void) Tcl_AsyncInvoke(NULL, 0);
+	return 1;
+    }
+
+    /*
+     * No event flags is equivalent to TCL_ALL_EVENTS.
+     */
+
+    if ((flags & TCL_ALL_EVENTS) == 0) {
+	flags |= TCL_ALL_EVENTS;
+    }
+
+    /*
+     * Loop through all the events in the queue until we find one that can
+     * actually be handled.
+     */
+
+    Tcl_MutexLock(&(tsdPtr->queueMutex));
+    for (evPtr = tsdPtr->firstEventPtr; evPtr != NULL;
+	    evPtr = evPtr->nextPtr) {
+	/*
+	 * Call the handler for the event. If it actually handles the event
+	 * then free the storage for the event. There are two tricky things
+	 * here, both stemming from the fact that the event code may be
+	 * re-entered while servicing the event:
+	 *
+	 * 1. Set the "proc" field to NULL.  This is a signal to ourselves
+	 *    that we shouldn't reexecute the handler if the event loop is
+	 *    re-entered.
+	 * 2. When freeing the event, must search the queue again from the
+	 *    front to find it. This is because the event queue could change
+	 *    almost arbitrarily while handling the event, so we can't depend
+	 *    on pointers found now still being valid when the handler
+	 *    returns.
+	 */
+
+	proc = evPtr->proc;
+	if (proc == NULL) {
+	    continue;
+	}
+	evPtr->proc = NULL;
+
+	/*
+	 * Release the lock before calling the event function. This allows
+	 * other threads to post events if we enter a recursive event loop in
+	 * this thread. Note that we are making the assumption that if the
+	 * proc returns 0, the event is still in the list.
+	 */
+
+	Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+	result = proc(evPtr, flags);
+	Tcl_MutexLock(&(tsdPtr->queueMutex));
+
+	if (result) {
+	    /*
+	     * The event was processed, so remove it from the queue.
+	     */
+
+	    if (tsdPtr->firstEventPtr == evPtr) {
+		tsdPtr->firstEventPtr = evPtr->nextPtr;
+		if (evPtr->nextPtr == NULL) {
+		    tsdPtr->lastEventPtr = NULL;
+		}
+		if (tsdPtr->markerEventPtr == evPtr) {
+		    tsdPtr->markerEventPtr = NULL;
+		}
+	    } else {
+		for (prevPtr = tsdPtr->firstEventPtr;
+			prevPtr && prevPtr->nextPtr != evPtr;
+			prevPtr = prevPtr->nextPtr) {
+		    /* Empty loop body. */
+		}
+		if (prevPtr) {
+		    prevPtr->nextPtr = evPtr->nextPtr;
+		    if (evPtr->nextPtr == NULL) {
+			tsdPtr->lastEventPtr = prevPtr;
+		    }
+		    if (tsdPtr->markerEventPtr == evPtr) {
+			tsdPtr->markerEventPtr = prevPtr;
+		    }
+		} else {
+		    evPtr = NULL;
+		}
+	    }
+	    if (evPtr) {
+		ckfree(evPtr);
+	    }
+	    Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+	    return 1;
+	} else {
+	    /*
+	     * The event wasn't actually handled, so we have to restore the
+	     * proc field to allow the event to be attempted again.
+	     */
+
+	    evPtr->proc = proc;
+	}
+    }
+    Tcl_MutexUnlock(&(tsdPtr->queueMutex));
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetServiceMode --
+ *
+ *	This routine returns the current service mode of the notifier.
+ *
+ * Results:
+ *	Returns either TCL_SERVICE_ALL or TCL_SERVICE_NONE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetServiceMode(void)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    return tsdPtr->serviceMode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetServiceMode --
+ *
+ *	This routine sets the current service mode of the tsdPtr->
+ *
+ * Results:
+ *	Returns the previous service mode.
+ *
+ * Side effects:
+ *	Invokes the notifier service mode hook function.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SetServiceMode(
+    int mode)			/* New service mode: TCL_SERVICE_ALL or
+				 * TCL_SERVICE_NONE */
+{
+    int oldMode;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    oldMode = tsdPtr->serviceMode;
+    tsdPtr->serviceMode = mode;
+    Tcl_ServiceModeHook(mode);
+    return oldMode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetMaxBlockTime --
+ *
+ *	This function is invoked by event sources to tell the notifier how
+ *	long it may block the next time it blocks. The timePtr argument gives
+ *	a maximum time; the actual time may be less if some other event source
+ *	requested a smaller time.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May reduce the length of the next sleep in the tsdPtr->
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetMaxBlockTime(
+    const Tcl_Time *timePtr)		/* Specifies a maximum elapsed time for the
+				 * next blocking operation in the event
+				 * tsdPtr-> */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (!tsdPtr->blockTimeSet || (timePtr->sec < tsdPtr->blockTime.sec)
+	    || ((timePtr->sec == tsdPtr->blockTime.sec)
+	    && (timePtr->usec < tsdPtr->blockTime.usec))) {
+	tsdPtr->blockTime = *timePtr;
+	tsdPtr->blockTimeSet = 1;
+    }
+
+    /*
+     * If we are called outside an event source traversal, set the timeout
+     * immediately.
+     */
+
+    if (!tsdPtr->inTraversal) {
+	Tcl_SetTimer(&tsdPtr->blockTime);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DoOneEvent --
+ *
+ *	Process a single event of some sort. If there's no work to do, wait
+ *	for an event to occur, then process it.
+ *
+ * Results:
+ *	The return value is 1 if the function actually found an event to
+ *	process. If no processing occurred, then 0 is returned (this can
+ *	happen if the TCL_DONT_WAIT flag is set or if there are no event
+ *	handlers to wait for in the set specified by flags).
+ *
+ * Side effects:
+ *	May delay execution of process while waiting for an event, unless
+ *	TCL_DONT_WAIT is set in the flags argument. Event sources are invoked
+ *	to check for and queue events. Event handlers may produce arbitrary
+ *	side effects.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DoOneEvent(
+    int flags)			/* Miscellaneous flag values: may be any
+				 * combination of TCL_DONT_WAIT,
+				 * TCL_WINDOW_EVENTS, TCL_FILE_EVENTS,
+				 * TCL_TIMER_EVENTS, TCL_IDLE_EVENTS, or
+				 * others defined by event sources. */
+{
+    int result = 0, oldMode;
+    EventSource *sourcePtr;
+    Tcl_Time *timePtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    /*
+     * The first thing we do is to service any asynchronous event handlers.
+     */
+
+    if (Tcl_AsyncReady()) {
+	(void) Tcl_AsyncInvoke(NULL, 0);
+	return 1;
+    }
+
+    /*
+     * No event flags is equivalent to TCL_ALL_EVENTS.
+     */
+
+    if ((flags & TCL_ALL_EVENTS) == 0) {
+	flags |= TCL_ALL_EVENTS;
+    }
+
+    /*
+     * Set the service mode to none so notifier event routines won't try to
+     * service events recursively.
+     */
+
+    oldMode = tsdPtr->serviceMode;
+    tsdPtr->serviceMode = TCL_SERVICE_NONE;
+
+    /*
+     * The core of this function is an infinite loop, even though we only
+     * service one event. The reason for this is that we may be processing
+     * events that don't do anything inside of Tcl.
+     */
+
+    while (1) {
+	/*
+	 * If idle events are the only things to service, skip the main part
+	 * of the loop and go directly to handle idle events (i.e. don't wait
+	 * even if TCL_DONT_WAIT isn't set).
+	 */
+
+	if ((flags & TCL_ALL_EVENTS) == TCL_IDLE_EVENTS) {
+	    flags = TCL_IDLE_EVENTS | TCL_DONT_WAIT;
+	    goto idleEvents;
+	}
+
+	/*
+	 * Ask Tcl to service a queued event, if there are any.
+	 */
+
+	if (Tcl_ServiceEvent(flags)) {
+	    result = 1;
+	    break;
+	}
+
+	/*
+	 * If TCL_DONT_WAIT is set, be sure to poll rather than blocking,
+	 * otherwise reset the block time to infinity.
+	 */
+
+	if (flags & TCL_DONT_WAIT) {
+	    tsdPtr->blockTime.sec = 0;
+	    tsdPtr->blockTime.usec = 0;
+	    tsdPtr->blockTimeSet = 1;
+	} else {
+	    tsdPtr->blockTimeSet = 0;
+	}
+
+	/*
+	 * Set up all the event sources for new events. This will cause the
+	 * block time to be updated if necessary.
+	 */
+
+	tsdPtr->inTraversal = 1;
+	for (sourcePtr = tsdPtr->firstEventSourcePtr; sourcePtr != NULL;
+		sourcePtr = sourcePtr->nextPtr) {
+	    if (sourcePtr->setupProc) {
+		sourcePtr->setupProc(sourcePtr->clientData, flags);
+	    }
+	}
+	tsdPtr->inTraversal = 0;
+
+	if ((flags & TCL_DONT_WAIT) || tsdPtr->blockTimeSet) {
+	    timePtr = &tsdPtr->blockTime;
+	} else {
+	    timePtr = NULL;
+	}
+
+	/*
+	 * Wait for a new event or a timeout. If Tcl_WaitForEvent returns -1,
+	 * we should abort Tcl_DoOneEvent.
+	 */
+
+	result = Tcl_WaitForEvent(timePtr);
+	if (result < 0) {
+	    result = 0;
+	    break;
+	}
+
+	/*
+	 * Check all the event sources for new events.
+	 */
+
+	for (sourcePtr = tsdPtr->firstEventSourcePtr; sourcePtr != NULL;
+		sourcePtr = sourcePtr->nextPtr) {
+	    if (sourcePtr->checkProc) {
+		sourcePtr->checkProc(sourcePtr->clientData, flags);
+	    }
+	}
+
+	/*
+	 * Check for events queued by the notifier or event sources.
+	 */
+
+	if (Tcl_ServiceEvent(flags)) {
+	    result = 1;
+	    break;
+	}
+
+	/*
+	 * We've tried everything at this point, but nobody we know about had
+	 * anything to do. Check for idle events. If none, either quit or go
+	 * back to the top and try again.
+	 */
+
+    idleEvents:
+	if (flags & TCL_IDLE_EVENTS) {
+	    if (TclServiceIdle()) {
+		result = 1;
+		break;
+	    }
+	}
+	if (flags & TCL_DONT_WAIT) {
+	    break;
+	}
+
+	/*
+	 * If Tcl_WaitForEvent has returned 1, indicating that one system
+	 * event has been dispatched (and thus that some Tcl code might have
+	 * been indirectly executed), we break out of the loop. We do this to
+	 * give VwaitCmd for instance a chance to check if that system event
+	 * had the side effect of changing the variable (so the vwait can
+	 * return and unwind properly).
+	 *
+	 * NB: We will process idle events if any first, because otherwise we
+	 *     might never do the idle events if the notifier always gets
+	 *     system events.
+	 */
+
+	if (result) {
+	    break;
+	}
+    }
+
+    tsdPtr->serviceMode = oldMode;
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ServiceAll --
+ *
+ *	This routine checks all of the event sources, processes events that
+ *	are on the Tcl event queue, and then calls the any idle handlers.
+ *	Platform specific notifier callbacks that generate events should call
+ *	this routine before returning to the system in order to ensure that
+ *	Tcl gets a chance to process the new events.
+ *
+ * Results:
+ *	Returns 1 if an event or idle handler was invoked, else 0.
+ *
+ * Side effects:
+ *	Anything that an event or idle handler may do.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ServiceAll(void)
+{
+    int result = 0;
+    EventSource *sourcePtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->serviceMode == TCL_SERVICE_NONE) {
+	return result;
+    }
+
+    /*
+     * We need to turn off event servicing like we to in Tcl_DoOneEvent, to
+     * avoid recursive calls.
+     */
+
+    tsdPtr->serviceMode = TCL_SERVICE_NONE;
+
+    /*
+     * Check async handlers first.
+     */
+
+    if (Tcl_AsyncReady()) {
+	(void) Tcl_AsyncInvoke(NULL, 0);
+    }
+
+    /*
+     * Make a single pass through all event sources, queued events, and idle
+     * handlers. Note that we wait to update the notifier timer until the end
+     * so we can avoid multiple changes.
+     */
+
+    tsdPtr->inTraversal = 1;
+    tsdPtr->blockTimeSet = 0;
+
+    for (sourcePtr = tsdPtr->firstEventSourcePtr; sourcePtr != NULL;
+	    sourcePtr = sourcePtr->nextPtr) {
+	if (sourcePtr->setupProc) {
+	    sourcePtr->setupProc(sourcePtr->clientData, TCL_ALL_EVENTS);
+	}
+    }
+    for (sourcePtr = tsdPtr->firstEventSourcePtr; sourcePtr != NULL;
+	    sourcePtr = sourcePtr->nextPtr) {
+	if (sourcePtr->checkProc) {
+	    sourcePtr->checkProc(sourcePtr->clientData, TCL_ALL_EVENTS);
+	}
+    }
+
+    while (Tcl_ServiceEvent(0)) {
+	result = 1;
+    }
+    if (TclServiceIdle()) {
+	result = 1;
+    }
+
+    if (!tsdPtr->blockTimeSet) {
+	Tcl_SetTimer(NULL);
+    } else {
+	Tcl_SetTimer(&tsdPtr->blockTime);
+    }
+    tsdPtr->inTraversal = 0;
+    tsdPtr->serviceMode = TCL_SERVICE_ALL;
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ThreadAlert --
+ *
+ *	This function wakes up the notifier associated with the specified
+ *	thread (if there is one).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ThreadAlert(
+    Tcl_ThreadId threadId)	/* Identifier for thread to use. */
+{
+    ThreadSpecificData *tsdPtr;
+
+    /*
+     * Find the notifier associated with the specified thread. Note that we
+     * need to hold the listLock while calling Tcl_AlertNotifier to avoid a
+     * race condition where the specified thread might destroy its notifier.
+     */
+
+    Tcl_MutexLock(&listLock);
+    for (tsdPtr = firstNotifierPtr; tsdPtr; tsdPtr = tsdPtr->nextPtr) {
+	if (tsdPtr->threadId == threadId) {
+	    Tcl_AlertNotifier(tsdPtr->clientData);
+	    break;
+	}
+    }
+    Tcl_MutexUnlock(&listLock);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOO.c b/generic/tclOO.c
new file mode 100644
index 0000000..73acce8
--- /dev/null
+++ b/generic/tclOO.c
@@ -0,0 +1,3038 @@
+/*
+ * tclOO.c --
+ *
+ *	This file contains the object-system core (NB: not Tcl_Obj, but ::oo)
+ *
+ * Copyright (c) 2005-2012 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+/*
+ * Commands in oo::define.
+ */
+
+static const struct {
+    const char *name;
+    Tcl_ObjCmdProc *objProc;
+    int flag;
+} defineCmds[] = {
+    {"constructor", TclOODefineConstructorObjCmd, 0},
+    {"deletemethod", TclOODefineDeleteMethodObjCmd, 0},
+    {"destructor", TclOODefineDestructorObjCmd, 0},
+    {"export", TclOODefineExportObjCmd, 0},
+    {"forward", TclOODefineForwardObjCmd, 0},
+    {"method", TclOODefineMethodObjCmd, 0},
+    {"renamemethod", TclOODefineRenameMethodObjCmd, 0},
+    {"self", TclOODefineSelfObjCmd, 0},
+    {"unexport", TclOODefineUnexportObjCmd, 0},
+    {NULL, NULL, 0}
+}, objdefCmds[] = {
+    {"class", TclOODefineClassObjCmd, 1},
+    {"deletemethod", TclOODefineDeleteMethodObjCmd, 1},
+    {"export", TclOODefineExportObjCmd, 1},
+    {"forward", TclOODefineForwardObjCmd, 1},
+    {"method", TclOODefineMethodObjCmd, 1},
+    {"renamemethod", TclOODefineRenameMethodObjCmd, 1},
+    {"self", TclOODefineObjSelfObjCmd, 0},
+    {"unexport", TclOODefineUnexportObjCmd, 1},
+    {NULL, NULL, 0}
+};
+
+/*
+ * What sort of size of things we like to allocate.
+ */
+
+#define ALLOC_CHUNK 8
+
+/*
+ * Function declarations for things defined in this file.
+ */
+
+static Class *		AllocClass(Tcl_Interp *interp, Object *useThisObj);
+static Object *		AllocObject(Tcl_Interp *interp, const char *nameStr,
+			    const char *nsNameStr);
+static void		ClearMixins(Class *clsPtr);
+static void		ClearSuperclasses(Class *clsPtr);
+static int		CloneClassMethod(Tcl_Interp *interp, Class *clsPtr,
+			    Method *mPtr, Tcl_Obj *namePtr,
+			    Method **newMPtrPtr);
+static int		CloneObjectMethod(Tcl_Interp *interp, Object *oPtr,
+			    Method *mPtr, Tcl_Obj *namePtr);
+static void		DeletedDefineNamespace(ClientData clientData);
+static void		DeletedObjdefNamespace(ClientData clientData);
+static void		DeletedHelpersNamespace(ClientData clientData);
+static Tcl_NRPostProc	FinalizeAlloc;
+static Tcl_NRPostProc	FinalizeNext;
+static Tcl_NRPostProc	FinalizeObjectCall;
+static int		InitFoundation(Tcl_Interp *interp);
+static void		KillFoundation(ClientData clientData,
+			    Tcl_Interp *interp);
+static void		MyDeleted(ClientData clientData);
+static void		ObjectNamespaceDeleted(ClientData clientData);
+static void		ObjectRenamedTrace(ClientData clientData,
+			    Tcl_Interp *interp, const char *oldName,
+			    const char *newName, int flags);
+static void		ReleaseClassContents(Tcl_Interp *interp,Object *oPtr);
+static inline void	SquelchCachedName(Object *oPtr);
+static void		SquelchedNsFirst(ClientData clientData);
+
+static int		PublicObjectCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+static int		PublicNRObjectCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+static int		PrivateObjectCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+static int		PrivateNRObjectCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+
+/*
+ * Methods in the oo::object and oo::class classes. First, we define a helper
+ * macro that makes building the method type declaration structure a lot
+ * easier. No point in making life harder than it has to be!
+ *
+ * Note that the core methods don't need clone or free proc callbacks.
+ */
+
+#define DCM(name,visibility,proc) \
+    {name,visibility,\
+	{TCL_OO_METHOD_VERSION_CURRENT,"core method: "#name,proc,NULL,NULL}}
+
+static const DeclaredClassMethod objMethods[] = {
+    DCM("destroy", 1,	TclOO_Object_Destroy),
+    DCM("eval", 0,	TclOO_Object_Eval),
+    DCM("unknown", 0,	TclOO_Object_Unknown),
+    DCM("variable", 0,	TclOO_Object_LinkVar),
+    DCM("varname", 0,	TclOO_Object_VarName),
+    {NULL, 0, {0, NULL, NULL, NULL, NULL}}
+}, clsMethods[] = {
+    DCM("create", 1,	TclOO_Class_Create),
+    DCM("new", 1,	TclOO_Class_New),
+    DCM("createWithNamespace", 0, TclOO_Class_CreateNs),
+    {NULL, 0, {0, NULL, NULL, NULL, NULL}}
+};
+
+/*
+ * And for the oo::class constructor...
+ */
+
+static const Tcl_MethodType classConstructor = {
+    TCL_OO_METHOD_VERSION_CURRENT,
+    "oo::class constructor",
+    TclOO_Class_Constructor, NULL, NULL
+};
+
+/*
+ * Scripted parts of TclOO. First, the master script (cannot be outside this
+ * file).
+ */
+
+static const char *initScript =
+"package ifneeded TclOO " TCLOO_PATCHLEVEL " {# Already present, OK?};"
+"namespace eval ::oo { variable version " TCLOO_VERSION " };"
+"namespace eval ::oo { variable patchlevel " TCLOO_PATCHLEVEL " };";
+/* "tcl_findLibrary tcloo $oo::version $oo::version" */
+/* " tcloo.tcl OO_LIBRARY oo::library;"; */
+
+/*
+ * The scripted part of the definitions of slots.
+ */
+
+static const char *slotScript =
+"::oo::define ::oo::Slot {\n"
+"    method Get {} {error unimplemented}\n"
+"    method Set list {error unimplemented}\n"
+"    method -set args {\n"
+"        uplevel 1 [list [namespace which my] Set $args]\n"
+"    }\n"
+"    method -append args {\n"
+"        uplevel 1 [list [namespace which my] Set [list"
+"                {*}[uplevel 1 [list [namespace which my] Get]] {*}$args]]\n"
+"    }\n"
+"    method -clear {} {uplevel 1 [list [namespace which my] Set {}]}\n"
+"    forward --default-operation my -append\n"
+"    method unknown {args} {\n"
+"        set def --default-operation\n"
+"        if {[llength $args] == 0} {\n"
+"            return [uplevel 1 [list [namespace which my] $def]]\n"
+"        } elseif {![string match -* [lindex $args 0]]} {\n"
+"            return [uplevel 1 [list [namespace which my] $def {*}$args]]\n"
+"        }\n"
+"        next {*}$args\n"
+"    }\n"
+"    export -set -append -clear\n"
+"    unexport unknown destroy\n"
+"}\n"
+"::oo::objdefine ::oo::define::superclass forward --default-operation my -set\n"
+"::oo::objdefine ::oo::define::mixin forward --default-operation my -set\n"
+"::oo::objdefine ::oo::objdefine::mixin forward --default-operation my -set\n";
+
+/*
+ * The body of the <cloned> method of oo::object.
+ */
+
+static const char *clonedBody =
+"foreach p [info procs [info object namespace $originObject]::*] {"
+"    set args [info args $p];"
+"    set idx -1;"
+"    foreach a $args {"
+"        lset args [incr idx] "
+"            [if {[info default $p $a d]} {list $a $d} {list $a}]"
+"    };"
+"    set b [info body $p];"
+"    set p [namespace tail $p];"
+"    proc $p $args $b;"
+"};"
+"foreach v [info vars [info object namespace $originObject]::*] {"
+"    upvar 0 $v vOrigin;"
+"    namespace upvar [namespace current] [namespace tail $v] vNew;"
+"    if {[info exists vOrigin]} {"
+"        if {[array exists vOrigin]} {"
+"            array set vNew [array get vOrigin];"
+"        } else {"
+"            set vNew $vOrigin;"
+"        }"
+"    }"
+"}";
+
+/*
+ * The actual definition of the variable holding the TclOO stub table.
+ */
+
+MODULE_SCOPE const TclOOStubs tclOOStubs;
+
+/*
+ * Convenience macro for getting the foundation from an interpreter.
+ */
+
+#define GetFoundation(interp) \
+	((Foundation *)((Interp *)(interp))->objectFoundation)
+
+/*
+ * Macros to make inspecting into the guts of an object cleaner.
+ *
+ * The ocPtr parameter (only in these macros) is assumed to work fine with
+ * either an oPtr or a classPtr. Note that the roots oo::object and oo::class
+ * have _both_ their object and class flags tagged with ROOT_OBJECT and
+ * ROOT_CLASS respectively.
+ */
+
+#define Deleted(oPtr)		(((Object *)(oPtr))->command == NULL)
+#define IsRootObject(ocPtr)	((ocPtr)->flags & ROOT_OBJECT)
+#define IsRootClass(ocPtr)	((ocPtr)->flags & ROOT_CLASS)
+#define IsRoot(ocPtr)		((ocPtr)->flags & (ROOT_OBJECT|ROOT_CLASS))
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOInit --
+ *
+ *	Called to initialise the OO system within an interpreter.
+ *
+ * Result:
+ *	TCL_OK if the setup succeeded. Currently assumed to always work.
+ *
+ * Side effects:
+ *	Creates namespaces, commands, several classes and a number of
+ *	callbacks. Upon return, the OO system is ready for use.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOInit(
+    Tcl_Interp *interp)		/* The interpreter to install into. */
+{
+    /*
+     * Build the core of the OO system.
+     */
+
+    if (InitFoundation(interp) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Run our initialization script and, if that works, declare the package
+     * to be fully provided.
+     */
+
+    if (Tcl_EvalEx(interp, initScript, -1, 0) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    return Tcl_PkgProvideEx(interp, "TclOO", TCLOO_PATCHLEVEL,
+	    (ClientData) &tclOOStubs);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetFoundation --
+ *
+ *	Get a reference to the OO core class system.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Foundation *
+TclOOGetFoundation(
+    Tcl_Interp *interp)
+{
+    return GetFoundation(interp);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InitFoundation --
+ *
+ *	Set up the core of the OO core class system. This is a structure
+ *	holding references to the magical bits that need to be known about in
+ *	other places, plus the oo::object and oo::class classes.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InitFoundation(
+    Tcl_Interp *interp)
+{
+    static Tcl_ThreadDataKey tsdKey;
+    ThreadLocalData *tsdPtr =
+	    Tcl_GetThreadData(&tsdKey, sizeof(ThreadLocalData));
+    Foundation *fPtr = ckalloc(sizeof(Foundation));
+    Tcl_Obj *namePtr, *argsPtr, *bodyPtr;
+    Tcl_DString buffer;
+    Command *cmdPtr;
+    int i;
+
+    /*
+     * Initialize the structure that holds the OO system core. This is
+     * attached to the interpreter via an assocData entry; not very efficient,
+     * but the best we can do without hacking the core more.
+     */
+
+    memset(fPtr, 0, sizeof(Foundation));
+    ((Interp *) interp)->objectFoundation = fPtr;
+    fPtr->interp = interp;
+    fPtr->ooNs = Tcl_CreateNamespace(interp, "::oo", fPtr, NULL);
+    Tcl_Export(interp, fPtr->ooNs, "[a-z]*", 1);
+    fPtr->defineNs = Tcl_CreateNamespace(interp, "::oo::define", fPtr,
+	    DeletedDefineNamespace);
+    fPtr->objdefNs = Tcl_CreateNamespace(interp, "::oo::objdefine", fPtr,
+	    DeletedObjdefNamespace);
+    fPtr->helpersNs = Tcl_CreateNamespace(interp, "::oo::Helpers", fPtr,
+	    DeletedHelpersNamespace);
+    fPtr->epoch = 0;
+    fPtr->tsdPtr = tsdPtr;
+    TclNewLiteralStringObj(fPtr->unknownMethodNameObj, "unknown");
+    TclNewLiteralStringObj(fPtr->constructorName, "<constructor>");
+    TclNewLiteralStringObj(fPtr->destructorName, "<destructor>");
+    TclNewLiteralStringObj(fPtr->clonedName, "<cloned>");
+    TclNewLiteralStringObj(fPtr->defineName, "::oo::define");
+    Tcl_IncrRefCount(fPtr->unknownMethodNameObj);
+    Tcl_IncrRefCount(fPtr->constructorName);
+    Tcl_IncrRefCount(fPtr->destructorName);
+    Tcl_IncrRefCount(fPtr->clonedName);
+    Tcl_IncrRefCount(fPtr->defineName);
+    Tcl_CreateObjCommand(interp, "::oo::UnknownDefinition",
+	    TclOOUnknownDefinition, NULL, NULL);
+    TclNewLiteralStringObj(namePtr, "::oo::UnknownDefinition");
+    Tcl_SetNamespaceUnknownHandler(interp, fPtr->defineNs, namePtr);
+    Tcl_SetNamespaceUnknownHandler(interp, fPtr->objdefNs, namePtr);
+
+    /*
+     * Create the subcommands in the oo::define and oo::objdefine spaces.
+     */
+
+    Tcl_DStringInit(&buffer);
+    for (i=0 ; defineCmds[i].name ; i++) {
+	TclDStringAppendLiteral(&buffer, "::oo::define::");
+	Tcl_DStringAppend(&buffer, defineCmds[i].name, -1);
+	Tcl_CreateObjCommand(interp, Tcl_DStringValue(&buffer),
+		defineCmds[i].objProc, INT2PTR(defineCmds[i].flag), NULL);
+	Tcl_DStringFree(&buffer);
+    }
+    for (i=0 ; objdefCmds[i].name ; i++) {
+	TclDStringAppendLiteral(&buffer, "::oo::objdefine::");
+	Tcl_DStringAppend(&buffer, objdefCmds[i].name, -1);
+	Tcl_CreateObjCommand(interp, Tcl_DStringValue(&buffer),
+		objdefCmds[i].objProc, INT2PTR(objdefCmds[i].flag), NULL);
+	Tcl_DStringFree(&buffer);
+    }
+
+    Tcl_CallWhenDeleted(interp, KillFoundation, NULL);
+
+    /*
+     * Create the objects at the core of the object system. These need to be
+     * spliced manually.
+     */
+
+    fPtr->objectCls = AllocClass(interp,
+	    AllocObject(interp, "::oo::object", NULL));
+    fPtr->classCls = AllocClass(interp,
+	    AllocObject(interp, "::oo::class", NULL));
+    fPtr->objectCls->thisPtr->selfCls = fPtr->classCls;
+    fPtr->objectCls->thisPtr->flags |= ROOT_OBJECT;
+    fPtr->objectCls->flags |= ROOT_OBJECT;
+    fPtr->objectCls->superclasses.num = 0;
+    ckfree(fPtr->objectCls->superclasses.list);
+    fPtr->objectCls->superclasses.list = NULL;
+    fPtr->classCls->thisPtr->selfCls = fPtr->classCls;
+    fPtr->classCls->thisPtr->flags |= ROOT_CLASS;
+    fPtr->classCls->flags |= ROOT_CLASS;
+    TclOOAddToInstances(fPtr->objectCls->thisPtr, fPtr->classCls);
+    TclOOAddToInstances(fPtr->classCls->thisPtr, fPtr->classCls);
+    TclOOAddToSubclasses(fPtr->classCls, fPtr->objectCls);
+    AddRef(fPtr->objectCls->thisPtr);
+    AddRef(fPtr->objectCls);
+
+    /*
+     * Basic method declarations for the core classes.
+     */
+
+    for (i=0 ; objMethods[i].name ; i++) {
+	TclOONewBasicMethod(interp, fPtr->objectCls, &objMethods[i]);
+    }
+    for (i=0 ; clsMethods[i].name ; i++) {
+	TclOONewBasicMethod(interp, fPtr->classCls, &clsMethods[i]);
+    }
+
+    /*
+     * Create the default <cloned> method implementation, used when 'oo::copy'
+     * is called to finish the copying of one object to another.
+     */
+
+    TclNewLiteralStringObj(argsPtr, "originObject");
+    Tcl_IncrRefCount(argsPtr);
+    bodyPtr = Tcl_NewStringObj(clonedBody, -1);
+    TclOONewProcMethod(interp, fPtr->objectCls, 0, fPtr->clonedName, argsPtr,
+	    bodyPtr, NULL);
+    TclDecrRefCount(argsPtr);
+
+    /*
+     * Finish setting up the class of classes by marking the 'new' method as
+     * private; classes, unlike general objects, must have explicit names. We
+     * also need to create the constructor for classes.
+     */
+
+    TclNewLiteralStringObj(namePtr, "new");
+    Tcl_NewInstanceMethod(interp, (Tcl_Object) fPtr->classCls->thisPtr,
+	    namePtr /* keeps ref */, 0 /* ==private */, NULL, NULL);
+    fPtr->classCls->constructorPtr = (Method *) Tcl_NewMethod(interp,
+	    (Tcl_Class) fPtr->classCls, NULL, 0, &classConstructor, NULL);
+
+    /*
+     * Create non-object commands and plug ourselves into the Tcl [info]
+     * ensemble.
+     */
+
+    cmdPtr = (Command *) Tcl_NRCreateCommand(interp, "::oo::Helpers::next",
+	    NULL, TclOONextObjCmd, NULL, NULL);
+    cmdPtr->compileProc = TclCompileObjectNextCmd;
+    cmdPtr = (Command *) Tcl_NRCreateCommand(interp, "::oo::Helpers::nextto",
+	    NULL, TclOONextToObjCmd, NULL, NULL);
+    cmdPtr->compileProc = TclCompileObjectNextToCmd;
+    cmdPtr = (Command *) Tcl_CreateObjCommand(interp, "::oo::Helpers::self",
+	    TclOOSelfObjCmd, NULL, NULL);
+    cmdPtr->compileProc = TclCompileObjectSelfCmd;
+    Tcl_CreateObjCommand(interp, "::oo::define", TclOODefineObjCmd, NULL,
+	    NULL);
+    Tcl_CreateObjCommand(interp, "::oo::objdefine", TclOOObjDefObjCmd, NULL,
+	    NULL);
+    Tcl_CreateObjCommand(interp, "::oo::copy", TclOOCopyObjectCmd, NULL,NULL);
+    TclOOInitInfo(interp);
+
+    /*
+     * Now make the class of slots.
+     */
+
+    if (TclOODefineSlots(fPtr) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return Tcl_EvalEx(interp, slotScript, -1, 0);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * DeletedDefineNamespace, DeletedObjdefNamespace, DeletedHelpersNamespace --
+ *
+ *	Simple helpers used to clear fields of the foundation when they no
+ *	longer hold useful information.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+DeletedDefineNamespace(
+    ClientData clientData)
+{
+    Foundation *fPtr = clientData;
+
+    fPtr->defineNs = NULL;
+}
+
+static void
+DeletedObjdefNamespace(
+    ClientData clientData)
+{
+    Foundation *fPtr = clientData;
+
+    fPtr->objdefNs = NULL;
+}
+
+static void
+DeletedHelpersNamespace(
+    ClientData clientData)
+{
+    Foundation *fPtr = clientData;
+
+    fPtr->helpersNs = NULL;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * KillFoundation --
+ *
+ *	Delete those parts of the OO core that are not deleted automatically
+ *	when the objects and classes themselves are destroyed.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+KillFoundation(
+    ClientData clientData,	/* Pointer to the OO system foundation
+				 * structure. */
+    Tcl_Interp *interp)		/* The interpreter containing the OO system
+				 * foundation. */
+{
+    Foundation *fPtr = GetFoundation(interp);
+
+    DelRef(fPtr->objectCls->thisPtr);
+    DelRef(fPtr->objectCls);
+    TclDecrRefCount(fPtr->unknownMethodNameObj);
+    TclDecrRefCount(fPtr->constructorName);
+    TclDecrRefCount(fPtr->destructorName);
+    TclDecrRefCount(fPtr->clonedName);
+    TclDecrRefCount(fPtr->defineName);
+    ckfree(fPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AllocObject --
+ *
+ *	Allocate an object of basic type. Does not splice the object into its
+ *	class's instance list.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static Object *
+AllocObject(
+    Tcl_Interp *interp,		/* Interpreter within which to create the
+				 * object. */
+    const char *nameStr,	/* The name of the object to create, or NULL
+				 * if the OO system should pick the object
+				 * name itself (equal to the namespace
+				 * name). */
+    const char *nsNameStr)	/* The name of the namespace to create, or
+				 * NULL if the OO system should pick a unique
+				 * name itself. If this is non-NULL but names
+				 * a namespace that already exists, the effect
+				 * will be the same as if this was NULL. */
+{
+    Foundation *fPtr = GetFoundation(interp);
+    Object *oPtr;
+    Command *cmdPtr;
+    CommandTrace *tracePtr;
+    int creationEpoch, ignored;
+
+    oPtr = ckalloc(sizeof(Object));
+    memset(oPtr, 0, sizeof(Object));
+
+    /*
+     * Every object has a namespace; make one. Note that this also normally
+     * computes the creation epoch value for the object, a sequence number
+     * that is unique to the object (and which allows us to manage method
+     * caching without comparing pointers).
+     *
+     * When creating a namespace, we first check to see if the caller
+     * specified the name for the namespace. If not, we generate namespace
+     * names using the epoch until such time as a new namespace is actually
+     * created.
+     */
+
+    if (nsNameStr != NULL) {
+	oPtr->namespacePtr = Tcl_CreateNamespace(interp, nsNameStr, oPtr,
+		ObjectNamespaceDeleted);
+	if (oPtr->namespacePtr != NULL) {
+	    creationEpoch = ++fPtr->tsdPtr->nsCount;
+	    goto configNamespace;
+	}
+	Tcl_ResetResult(interp);
+    }
+
+    while (1) {
+	char objName[10 + TCL_INTEGER_SPACE];
+
+	sprintf(objName, "::oo::Obj%d", ++fPtr->tsdPtr->nsCount);
+	oPtr->namespacePtr = Tcl_CreateNamespace(interp, objName, oPtr,
+		ObjectNamespaceDeleted);
+	if (oPtr->namespacePtr != NULL) {
+	    creationEpoch = fPtr->tsdPtr->nsCount;
+	    break;
+	}
+
+	/*
+	 * Could not make that namespace, so we make another. But first we
+	 * have to get rid of the error message from Tcl_CreateNamespace,
+	 * since that's something that should not be exposed to the user.
+	 */
+
+	Tcl_ResetResult(interp);
+    }
+
+    /*
+     * Make the namespace know about the helper commands. This grants access
+     * to the [self] and [next] commands.
+     */
+
+  configNamespace:
+    if (fPtr->helpersNs != NULL) {
+	TclSetNsPath((Namespace *) oPtr->namespacePtr, 1, &fPtr->helpersNs);
+    }
+    TclOOSetupVariableResolver(oPtr->namespacePtr);
+
+    /*
+     * Suppress use of compiled versions of the commands in this object's
+     * namespace and its children; causes wrong behaviour without expensive
+     * recompilation. [Bug 2037727]
+     */
+
+    ((Namespace *) oPtr->namespacePtr)->flags |= NS_SUPPRESS_COMPILATION;
+
+    /*
+     * Set up a callback to get notification of the deletion of a namespace
+     * when enough of the namespace still remains to execute commands and
+     * access variables in it. [Bug 2950259]
+     */
+
+    ((Namespace *) oPtr->namespacePtr)->earlyDeleteProc = SquelchedNsFirst;
+
+    /*
+     * Fill in the rest of the non-zero/NULL parts of the structure.
+     */
+
+    oPtr->fPtr = fPtr;
+    oPtr->selfCls = fPtr->objectCls;
+    oPtr->creationEpoch = creationEpoch;
+    oPtr->refCount = 1;
+    oPtr->flags = USE_CLASS_CACHE;
+
+    /*
+     * Finally, create the object commands and initialize the trace on the
+     * public command (so that the object structures are deleted when the
+     * command is deleted).
+     */
+
+    if (!nameStr) {
+	oPtr->command = Tcl_CreateObjCommand(interp,
+		oPtr->namespacePtr->fullName, PublicObjectCmd, oPtr, NULL);
+    } else if (nameStr[0] == ':' && nameStr[1] == ':') {
+	oPtr->command = Tcl_CreateObjCommand(interp, nameStr,
+		PublicObjectCmd, oPtr, NULL);
+    } else {
+	Tcl_DString buffer;
+
+	Tcl_DStringInit(&buffer);
+	Tcl_DStringAppend(&buffer,
+		Tcl_GetCurrentNamespace(interp)->fullName, -1);
+	TclDStringAppendLiteral(&buffer, "::");
+	Tcl_DStringAppend(&buffer, nameStr, -1);
+	oPtr->command = Tcl_CreateObjCommand(interp,
+		Tcl_DStringValue(&buffer), PublicObjectCmd, oPtr, NULL);
+	Tcl_DStringFree(&buffer);
+    }
+
+    /*
+     * Add the NRE command and trace directly. While this breaks a number of
+     * abstractions, it is faster and we're inside Tcl here so we're allowed.
+     */
+
+    cmdPtr = (Command *) oPtr->command;
+    cmdPtr->nreProc = PublicNRObjectCmd;
+    cmdPtr->tracePtr = tracePtr = ckalloc(sizeof(CommandTrace));
+    tracePtr->traceProc = ObjectRenamedTrace;
+    tracePtr->clientData = oPtr;
+    tracePtr->flags = TCL_TRACE_RENAME|TCL_TRACE_DELETE;
+    tracePtr->nextPtr = NULL;
+    tracePtr->refCount = 1;
+
+    /*
+     * Access the namespace command table directly when creating "my" to avoid
+     * a bottleneck in string manipulation. Another abstraction-buster.
+     */
+
+    cmdPtr = ckalloc(sizeof(Command));
+    memset(cmdPtr, 0, sizeof(Command));
+    cmdPtr->nsPtr = (Namespace *) oPtr->namespacePtr;
+    cmdPtr->hPtr = Tcl_CreateHashEntry(&cmdPtr->nsPtr->cmdTable, "my",
+	    &ignored);
+    cmdPtr->refCount = 1;
+    cmdPtr->objProc = PrivateObjectCmd;
+    cmdPtr->deleteProc = MyDeleted;
+    cmdPtr->objClientData = cmdPtr->deleteData = oPtr;
+    cmdPtr->proc = TclInvokeObjectCommand;
+    cmdPtr->clientData = cmdPtr;
+    cmdPtr->nreProc = PrivateNRObjectCmd;
+    Tcl_SetHashValue(cmdPtr->hPtr, cmdPtr);
+    oPtr->myCommand = (Tcl_Command) cmdPtr;
+
+    return oPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * SquelchCachedName --
+ *
+ *	Encapsulates how to throw away a cached object name. Called from
+ *	object rename traces and at object destruction.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+SquelchCachedName(
+    Object *oPtr)
+{
+    if (oPtr->cachedNameObj) {
+	Tcl_DecrRefCount(oPtr->cachedNameObj);
+	oPtr->cachedNameObj = NULL;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * MyDeleted --
+ *
+ *	This callback is triggered when the object's [my] command is deleted
+ *	by any mechanism. It just marks the object as not having a [my]
+ *	command, and so prevents cleanup of that when the object itself is
+ *	deleted.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+MyDeleted(
+    ClientData clientData)	/* Reference to the object whose [my] has been
+				 * squelched. */
+{
+    register Object *oPtr = clientData;
+
+    oPtr->myCommand = NULL;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * SquelchedNsFirst --
+ *
+ *	This callback is triggered when the object's namespace is deleted by
+ *	any mechanism. It deletes the object's public command if it has not
+ *	already been deleted, so ensuring that destructors get run at an
+ *	appropriate time. [Bug 2950259]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+SquelchedNsFirst(
+    ClientData clientData)
+{
+    Object *oPtr = clientData;
+
+    if (oPtr->command) {
+	Tcl_DeleteCommandFromToken(oPtr->fPtr->interp, oPtr->command);
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ObjectRenamedTrace --
+ *
+ *	This callback is triggered when the object is deleted by any
+ *	mechanism. It runs the destructors and arranges for the actual cleanup
+ *	of the object's namespace, which in turn triggers cleansing of the
+ *	object data structures.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+ObjectRenamedTrace(
+    ClientData clientData,	/* The object being deleted. */
+    Tcl_Interp *interp,		/* The interpreter containing the object. */
+    const char *oldName,	/* What the object was (last) called. */
+    const char *newName,	/* What it's getting renamed to. (unused) */
+    int flags)			/* Why was the object deleted? */
+{
+    Object *oPtr = clientData;
+    Foundation *fPtr = oPtr->fPtr;
+
+    /*
+     * If this is a rename and not a delete of the object, we just flush the
+     * cache of the object name.
+     */
+
+    if (flags & TCL_TRACE_RENAME) {
+	SquelchCachedName(oPtr);
+	return;
+    }
+
+    /*
+     * Oh dear, the object really is being deleted. Handle this by running the
+     * destructors and deleting the object's namespace, which in turn causes
+     * the real object structures to be deleted.
+     *
+     * Note that it is possible for the namespace to be deleted before the
+     * command. Because of that case, we must take care here to mark the
+     * command as being deleted so that if we return here we don't run into
+     * reentrancy problems.
+     *
+     * We also do not run destructors on the core class objects when the
+     * interpreter is being deleted; their incestuous nature causes problems
+     * in that case when the destructor is partially deleted before the uses
+     * of it have gone. [Bug 2949397]
+     */
+
+    AddRef(oPtr);
+    AddRef(fPtr->classCls);
+    AddRef(fPtr->objectCls);
+    AddRef(fPtr->classCls->thisPtr);
+    AddRef(fPtr->objectCls->thisPtr);
+    oPtr->command = NULL;
+
+    if (!(oPtr->flags & DESTRUCTOR_CALLED) && !Tcl_InterpDeleted(interp)) {
+	CallContext *contextPtr =
+		TclOOGetCallContext(oPtr, NULL, DESTRUCTOR, NULL);
+	int result;
+	Tcl_InterpState state;
+
+	oPtr->flags |= DESTRUCTOR_CALLED;
+	if (contextPtr != NULL) {
+	    contextPtr->callPtr->flags |= DESTRUCTOR;
+	    contextPtr->skip = 0;
+	    state = Tcl_SaveInterpState(interp, TCL_OK);
+	    result = Tcl_NRCallObjProc(interp, TclOOInvokeContext,
+		    contextPtr, 0, NULL);
+	    if (result != TCL_OK) {
+		Tcl_BackgroundException(interp, result);
+	    }
+	    Tcl_RestoreInterpState(interp, state);
+	    TclOODeleteContext(contextPtr);
+	}
+    }
+
+    /*
+     * OK, the destructor's been run. Time to splat the class data (if any)
+     * and nuke the namespace (which triggers the final crushing of the object
+     * structure itself).
+     *
+     * The class of objects needs some special care; if it is deleted (and
+     * we're not killing the whole interpreter) we force the delete of the
+     * class of classes now as well. Due to the incestuous nature of those two
+     * classes, if one goes the other must too and yet the tangle can
+     * sometimes not go away automatically; we force it here. [Bug 2962664]
+     */
+
+    if (!Tcl_InterpDeleted(interp) && IsRootObject(oPtr)
+	    && !Deleted(fPtr->classCls->thisPtr)) {
+	Tcl_DeleteCommandFromToken(interp, fPtr->classCls->thisPtr->command);
+    }
+
+    if (oPtr->classPtr != NULL) {
+	AddRef(oPtr->classPtr);
+	ReleaseClassContents(interp, oPtr);
+    }
+
+    /*
+     * The namespace is only deleted if it hasn't already been deleted. [Bug
+     * 2950259]
+     */
+
+    if (((Namespace *) oPtr->namespacePtr)->earlyDeleteProc != NULL) {
+	Tcl_DeleteNamespace(oPtr->namespacePtr);
+    }
+    if (oPtr->classPtr) {
+	DelRef(oPtr->classPtr);
+    }
+    DelRef(fPtr->classCls->thisPtr);
+    DelRef(fPtr->objectCls->thisPtr);
+    DelRef(fPtr->classCls);
+    DelRef(fPtr->objectCls);
+    DelRef(oPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ClearMixins, ClearSuperclasses --
+ *
+ *	Utility functions for correctly clearing the list of mixins or
+ *	superclasses of a class. Will ckfree() the list storage.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+ClearMixins(
+    Class *clsPtr)
+{
+    int i;
+    Class *mixinPtr;
+
+    if (clsPtr->mixins.num == 0) {
+	return;
+    }
+
+    FOREACH(mixinPtr, clsPtr->mixins) {
+	TclOORemoveFromMixinSubs(clsPtr, mixinPtr);
+    }
+    ckfree(clsPtr->mixins.list);
+    clsPtr->mixins.list = NULL;
+    clsPtr->mixins.num = 0;
+}
+
+static void
+ClearSuperclasses(
+    Class *clsPtr)
+{
+    int i;
+    Class *superPtr;
+
+    if (clsPtr->superclasses.num == 0) {
+	return;
+    }
+
+    FOREACH(superPtr, clsPtr->superclasses) {
+	TclOORemoveFromSubclasses(clsPtr, superPtr);
+    }
+    ckfree(clsPtr->superclasses.list);
+    clsPtr->superclasses.list = NULL;
+    clsPtr->superclasses.num = 0;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ReleaseClassContents --
+ *
+ *	Tear down the special class data structure, including deleting all
+ *	dependent classes and objects.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+ReleaseClassContents(
+    Tcl_Interp *interp,		/* The interpreter containing the class. */
+    Object *oPtr)		/* The object representing the class. */
+{
+    FOREACH_HASH_DECLS;
+    int i;
+    Class *clsPtr = oPtr->classPtr, *mixinSubclassPtr, *subclassPtr;
+    Object *instancePtr;
+    Foundation *fPtr = oPtr->fPtr;
+
+    /*
+     * Sanity check!
+     */
+
+    if (!Deleted(oPtr)) {
+	if (IsRootClass(oPtr)) {
+	    Tcl_Panic("deleting class structure for non-deleted %s",
+		    "::oo::class");
+	} else if (IsRootObject(oPtr)) {
+	    Tcl_Panic("deleting class structure for non-deleted %s",
+		    "::oo::object");
+	} else {
+	    Tcl_Panic("deleting class structure for non-deleted %s",
+		    "general object");
+	}
+    }
+
+    /*
+     * Lock a number of dependent objects until we've stopped putting our
+     * fingers in them.
+     */
+
+    FOREACH(mixinSubclassPtr, clsPtr->mixinSubs) {
+	if (mixinSubclassPtr != NULL) {
+	    AddRef(mixinSubclassPtr);
+	    AddRef(mixinSubclassPtr->thisPtr);
+	}
+    }
+    FOREACH(subclassPtr, clsPtr->subclasses) {
+	if (subclassPtr != NULL && !IsRoot(subclassPtr)) {
+	    AddRef(subclassPtr);
+	    AddRef(subclassPtr->thisPtr);
+	}
+    }
+    if (!IsRootClass(oPtr)) {
+	FOREACH(instancePtr, clsPtr->instances) {
+	    int j;
+	    if (instancePtr->selfCls == clsPtr) {
+		instancePtr->flags |= CLASS_GONE;
+	    }
+	    for(j=0 ; j<instancePtr->mixins.num ; j++) {
+		Class *mixin = instancePtr->mixins.list[j];
+		if (mixin == clsPtr) {
+		    instancePtr->mixins.list[j] = NULL;
+		}
+	    }
+	    if (instancePtr != NULL && !IsRoot(instancePtr)) {
+		AddRef(instancePtr);
+	    }
+	}
+    }
+
+    /*
+     * Squelch classes that this class has been mixed into.
+     */
+
+    FOREACH(mixinSubclassPtr, clsPtr->mixinSubs) {
+	if (!Deleted(mixinSubclassPtr->thisPtr)) {
+	    Tcl_DeleteCommandFromToken(interp,
+		    mixinSubclassPtr->thisPtr->command);
+	}
+	ClearMixins(mixinSubclassPtr);
+	DelRef(mixinSubclassPtr->thisPtr);
+	DelRef(mixinSubclassPtr);
+    }
+    if (clsPtr->mixinSubs.list != NULL) {
+	ckfree(clsPtr->mixinSubs.list);
+	clsPtr->mixinSubs.list = NULL;
+	clsPtr->mixinSubs.num = 0;
+    }
+
+    /*
+     * Squelch subclasses of this class.
+     */
+
+    FOREACH(subclassPtr, clsPtr->subclasses) {
+	if (IsRoot(subclassPtr)) {
+	    continue;
+	}
+	if (!Deleted(subclassPtr->thisPtr)) {
+	    Tcl_DeleteCommandFromToken(interp, subclassPtr->thisPtr->command);
+	}
+	ClearSuperclasses(subclassPtr);
+	DelRef(subclassPtr->thisPtr);
+	DelRef(subclassPtr);
+    }
+    if (clsPtr->subclasses.list != NULL) {
+	ckfree(clsPtr->subclasses.list);
+	clsPtr->subclasses.list = NULL;
+	clsPtr->subclasses.num = 0;
+    }
+
+    /*
+     * Squelch instances of this class (includes objects we're mixed into).
+     */
+
+    if (!IsRootClass(oPtr)) {
+	FOREACH(instancePtr, clsPtr->instances) {
+	    if (instancePtr == NULL || IsRoot(instancePtr)) {
+		continue;
+	    }
+	    if (!Deleted(instancePtr)) {
+		Tcl_DeleteCommandFromToken(interp, instancePtr->command);
+		/*
+		 * Tcl_DeleteCommandFromToken() may have done to whole
+		 * job for us.  Roll back and check again.
+		 */
+		i--;
+		continue;
+	    }
+	    DelRef(instancePtr);
+	}
+    }
+    if (clsPtr->instances.list != NULL) {
+	ckfree(clsPtr->instances.list);
+	clsPtr->instances.list = NULL;
+	clsPtr->instances.num = 0;
+    }
+
+    /*
+     * Special: We delete these after everything else.
+     */
+
+    if (IsRootClass(oPtr) && !Deleted(fPtr->objectCls->thisPtr)) {
+	Tcl_DeleteCommandFromToken(interp, fPtr->objectCls->thisPtr->command);
+    }
+
+    /*
+     * Squelch method implementation chain caches.
+     */
+
+    if (clsPtr->constructorChainPtr) {
+	TclOODeleteChain(clsPtr->constructorChainPtr);
+	clsPtr->constructorChainPtr = NULL;
+    }
+    if (clsPtr->destructorChainPtr) {
+	TclOODeleteChain(clsPtr->destructorChainPtr);
+	clsPtr->destructorChainPtr = NULL;
+    }
+    if (clsPtr->classChainCache) {
+	CallChain *callPtr;
+
+	FOREACH_HASH_VALUE(callPtr, clsPtr->classChainCache) {
+	    TclOODeleteChain(callPtr);
+	}
+	Tcl_DeleteHashTable(clsPtr->classChainCache);
+	ckfree(clsPtr->classChainCache);
+	clsPtr->classChainCache = NULL;
+    }
+
+    /*
+     * Squelch our filter list.
+     */
+
+    if (clsPtr->filters.num) {
+	Tcl_Obj *filterObj;
+
+	FOREACH(filterObj, clsPtr->filters) {
+	    TclDecrRefCount(filterObj);
+	}
+	ckfree(clsPtr->filters.list);
+	clsPtr->filters.num = 0;
+    }
+
+    /*
+     * Squelch our metadata.
+     */
+
+    if (clsPtr->metadataPtr != NULL) {
+	Tcl_ObjectMetadataType *metadataTypePtr;
+	ClientData value;
+
+	FOREACH_HASH(metadataTypePtr, value, clsPtr->metadataPtr) {
+	    metadataTypePtr->deleteProc(value);
+	}
+	Tcl_DeleteHashTable(clsPtr->metadataPtr);
+	ckfree(clsPtr->metadataPtr);
+	clsPtr->metadataPtr = NULL;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ObjectNamespaceDeleted --
+ *
+ *	Callback when the object's namespace is deleted. Used to clean up the
+ *	data structures associated with the object. The complicated bit is
+ *	that this can sometimes happen before the object's command is deleted
+ *	(interpreter teardown is complex!)
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+ObjectNamespaceDeleted(
+    ClientData clientData)	/* Pointer to the class whose namespace is
+				 * being deleted. */
+{
+    Object *oPtr = clientData;
+    FOREACH_HASH_DECLS;
+    Class *clsPtr = oPtr->classPtr, *mixinPtr;
+    Method *mPtr;
+    Tcl_Obj *filterObj, *variableObj;
+    int i;
+
+    /*
+     * Instruct everyone to no longer use any allocated fields of the object.
+     * Also delete the commands that refer to the object at this point (if
+     * they still exist) because otherwise their references to the object
+     * point into freed memory, allowing crashes.
+     */
+
+    if (oPtr->command) {
+	Tcl_DeleteCommandFromToken(oPtr->fPtr->interp, oPtr->command);
+    }
+    if (oPtr->myCommand) {
+	Tcl_DeleteCommandFromToken(oPtr->fPtr->interp, oPtr->myCommand);
+    }
+
+    /*
+     * Splice the object out of its context. After this, we must *not* call
+     * methods on the object.
+     */
+
+    if (!IsRootObject(oPtr) && !(oPtr->flags & CLASS_GONE)) {
+	TclOORemoveFromInstances(oPtr, oPtr->selfCls);
+    }
+
+    FOREACH(mixinPtr, oPtr->mixins) {
+	if (mixinPtr) {
+	    TclOORemoveFromInstances(oPtr, mixinPtr);
+	}
+    }
+    if (i) {
+	ckfree(oPtr->mixins.list);
+    }
+
+    FOREACH(filterObj, oPtr->filters) {
+	TclDecrRefCount(filterObj);
+    }
+    if (i) {
+	ckfree(oPtr->filters.list);
+    }
+
+    if (oPtr->methodsPtr) {
+	FOREACH_HASH_VALUE(mPtr, oPtr->methodsPtr) {
+	    TclOODelMethodRef(mPtr);
+	}
+	Tcl_DeleteHashTable(oPtr->methodsPtr);
+	ckfree(oPtr->methodsPtr);
+    }
+
+    FOREACH(variableObj, oPtr->variables) {
+	TclDecrRefCount(variableObj);
+    }
+    if (i) {
+	ckfree(oPtr->variables.list);
+    }
+
+    if (oPtr->chainCache) {
+	TclOODeleteChainCache(oPtr->chainCache);
+    }
+
+    SquelchCachedName(oPtr);
+
+    if (oPtr->metadataPtr != NULL) {
+	Tcl_ObjectMetadataType *metadataTypePtr;
+	ClientData value;
+
+	FOREACH_HASH(metadataTypePtr, value, oPtr->metadataPtr) {
+	    metadataTypePtr->deleteProc(value);
+	}
+	Tcl_DeleteHashTable(oPtr->metadataPtr);
+	ckfree(oPtr->metadataPtr);
+	oPtr->metadataPtr = NULL;
+    }
+
+    /*
+     * If this was a class, there's additional deletion work to do.
+     */
+
+    if (clsPtr != NULL) {
+	Tcl_ObjectMetadataType *metadataTypePtr;
+	ClientData value;
+
+	if (clsPtr->metadataPtr != NULL) {
+	    FOREACH_HASH(metadataTypePtr, value, clsPtr->metadataPtr) {
+		metadataTypePtr->deleteProc(value);
+	    }
+	    Tcl_DeleteHashTable(clsPtr->metadataPtr);
+	    ckfree(clsPtr->metadataPtr);
+	    clsPtr->metadataPtr = NULL;
+	}
+
+	FOREACH(filterObj, clsPtr->filters) {
+	    TclDecrRefCount(filterObj);
+	}
+	if (i) {
+	    ckfree(clsPtr->filters.list);
+	    clsPtr->filters.num = 0;
+	}
+
+	ClearMixins(clsPtr);
+
+	ClearSuperclasses(clsPtr);
+
+	if (clsPtr->subclasses.list) {
+	    ckfree(clsPtr->subclasses.list);
+	    clsPtr->subclasses.num = 0;
+	}
+	if (clsPtr->instances.list) {
+	    ckfree(clsPtr->instances.list);
+	    clsPtr->instances.num = 0;
+	}
+	if (clsPtr->mixinSubs.list) {
+	    ckfree(clsPtr->mixinSubs.list);
+	    clsPtr->mixinSubs.num = 0;
+	}
+
+	FOREACH_HASH_VALUE(mPtr, &clsPtr->classMethods) {
+	    TclOODelMethodRef(mPtr);
+	}
+	Tcl_DeleteHashTable(&clsPtr->classMethods);
+	TclOODelMethodRef(clsPtr->constructorPtr);
+	TclOODelMethodRef(clsPtr->destructorPtr);
+
+	FOREACH(variableObj, clsPtr->variables) {
+	    TclDecrRefCount(variableObj);
+	}
+	if (i) {
+	    ckfree(clsPtr->variables.list);
+	}
+
+	DelRef(clsPtr);
+    }
+
+    /*
+     * Delete the object structure itself.
+     */
+
+    DelRef(oPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOORemoveFromInstances --
+ *
+ *	Utility function to remove an object from the list of instances within
+ *	a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOORemoveFromInstances(
+    Object *oPtr,		/* The instance to remove. */
+    Class *clsPtr)		/* The class (possibly) containing the
+				 * reference to the instance. */
+{
+    int i;
+    Object *instPtr;
+
+    FOREACH(instPtr, clsPtr->instances) {
+	if (oPtr == instPtr) {
+	    goto removeInstance;
+	}
+    }
+    return;
+
+  removeInstance:
+    if (Deleted(clsPtr->thisPtr)) {
+	if (!IsRootClass(clsPtr)) {
+	    DelRef(clsPtr->instances.list[i]);
+	}
+	clsPtr->instances.list[i] = NULL;
+    } else {
+	clsPtr->instances.num--;
+	if (i < clsPtr->instances.num) {
+	    clsPtr->instances.list[i] =
+		    clsPtr->instances.list[clsPtr->instances.num];
+	}
+	clsPtr->instances.list[clsPtr->instances.num] = NULL;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOAddToInstances --
+ *
+ *	Utility function to add an object to the list of instances within a
+ *	class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOAddToInstances(
+    Object *oPtr,		/* The instance to add. */
+    Class *clsPtr)		/* The class to add the instance to. It is
+				 * assumed that the class is not already
+				 * present as an instance in the class. */
+{
+    if (Deleted(clsPtr->thisPtr)) {
+	return;
+    }
+    if (clsPtr->instances.num >= clsPtr->instances.size) {
+	clsPtr->instances.size += ALLOC_CHUNK;
+	if (clsPtr->instances.size == ALLOC_CHUNK) {
+	    clsPtr->instances.list = ckalloc(sizeof(Object *) * ALLOC_CHUNK);
+	} else {
+	    clsPtr->instances.list = ckrealloc(clsPtr->instances.list,
+		    sizeof(Object *) * clsPtr->instances.size);
+	}
+    }
+    clsPtr->instances.list[clsPtr->instances.num++] = oPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOORemoveFromSubclasses --
+ *
+ *	Utility function to remove a class from the list of subclasses within
+ *	another class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOORemoveFromSubclasses(
+    Class *subPtr,		/* The subclass to remove. */
+    Class *superPtr)		/* The superclass to (possibly) remove the
+				 * subclass reference from. */
+{
+    int i;
+    Class *subclsPtr;
+
+    FOREACH(subclsPtr, superPtr->subclasses) {
+	if (subPtr == subclsPtr) {
+	    goto removeSubclass;
+	}
+    }
+    return;
+
+  removeSubclass:
+    if (!Deleted(superPtr->thisPtr)) {
+	superPtr->subclasses.num--;
+	if (i < superPtr->subclasses.num) {
+	    superPtr->subclasses.list[i] =
+		    superPtr->subclasses.list[superPtr->subclasses.num];
+	}
+	superPtr->subclasses.list[superPtr->subclasses.num] = NULL;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOAddToSubclasses --
+ *
+ *	Utility function to add a class to the list of subclasses within
+ *	another class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOAddToSubclasses(
+    Class *subPtr,		/* The subclass to add. */
+    Class *superPtr)		/* The superclass to add the subclass to. It
+				 * is assumed that the class is not already
+				 * present as a subclass in the superclass. */
+{
+    if (Deleted(superPtr->thisPtr)) {
+	return;
+    }
+    if (superPtr->subclasses.num >= superPtr->subclasses.size) {
+	superPtr->subclasses.size += ALLOC_CHUNK;
+	if (superPtr->subclasses.size == ALLOC_CHUNK) {
+	    superPtr->subclasses.list = ckalloc(sizeof(Class*) * ALLOC_CHUNK);
+	} else {
+	    superPtr->subclasses.list = ckrealloc(superPtr->subclasses.list,
+		    sizeof(Class *) * superPtr->subclasses.size);
+	}
+    }
+    superPtr->subclasses.list[superPtr->subclasses.num++] = subPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOORemoveFromMixinSubs --
+ *
+ *	Utility function to remove a class from the list of mixinSubs within
+ *	another class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOORemoveFromMixinSubs(
+    Class *subPtr,		/* The subclass to remove. */
+    Class *superPtr)		/* The superclass to (possibly) remove the
+				 * subclass reference from. */
+{
+    int i;
+    Class *subclsPtr;
+
+    FOREACH(subclsPtr, superPtr->mixinSubs) {
+	if (subPtr == subclsPtr) {
+	    goto removeSubclass;
+	}
+    }
+    return;
+
+  removeSubclass:
+    if (!Deleted(superPtr->thisPtr)) {
+	superPtr->mixinSubs.num--;
+	if (i < superPtr->mixinSubs.num) {
+	    superPtr->mixinSubs.list[i] =
+		    superPtr->mixinSubs.list[superPtr->mixinSubs.num];
+	}
+	superPtr->mixinSubs.list[superPtr->mixinSubs.num] = NULL;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOAddToMixinSubs --
+ *
+ *	Utility function to add a class to the list of mixinSubs within
+ *	another class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOAddToMixinSubs(
+    Class *subPtr,		/* The subclass to add. */
+    Class *superPtr)		/* The superclass to add the subclass to. It
+				 * is assumed that the class is not already
+				 * present as a subclass in the superclass. */
+{
+    if (Deleted(superPtr->thisPtr)) {
+	return;
+    }
+    if (superPtr->mixinSubs.num >= superPtr->mixinSubs.size) {
+	superPtr->mixinSubs.size += ALLOC_CHUNK;
+	if (superPtr->mixinSubs.size == ALLOC_CHUNK) {
+	    superPtr->mixinSubs.list = ckalloc(sizeof(Class *) * ALLOC_CHUNK);
+	} else {
+	    superPtr->mixinSubs.list = ckrealloc(superPtr->mixinSubs.list,
+		    sizeof(Class *) * superPtr->mixinSubs.size);
+	}
+    }
+    superPtr->mixinSubs.list[superPtr->mixinSubs.num++] = subPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AllocClass --
+ *
+ *	Allocate a basic class. Does not splice the class object into its
+ *	class's instance list.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static Class *
+AllocClass(
+    Tcl_Interp *interp,		/* Interpreter within which to allocate the
+				 * class. */
+    Object *useThisObj)		/* Object that is to act as the class
+				 * representation, or NULL if a new object
+				 * (with automatic name) is to be used. */
+{
+    Foundation *fPtr = GetFoundation(interp);
+    Class *clsPtr = ckalloc(sizeof(Class));
+
+    /*
+     * Make an object if we haven't been given one.
+     */
+
+    memset(clsPtr, 0, sizeof(Class));
+    if (useThisObj == NULL) {
+	clsPtr->thisPtr = AllocObject(interp, NULL, NULL);
+    } else {
+	clsPtr->thisPtr = useThisObj;
+    }
+
+    /*
+     * Configure the namespace path for the class's object.
+     */
+
+    if (fPtr->helpersNs != NULL) {
+	Tcl_Namespace *path[2];
+
+	path[0] = fPtr->helpersNs;
+	path[1] = fPtr->ooNs;
+	TclSetNsPath((Namespace *) clsPtr->thisPtr->namespacePtr, 2, path);
+    } else {
+	TclSetNsPath((Namespace *) clsPtr->thisPtr->namespacePtr, 1,
+		&fPtr->ooNs);
+    }
+
+    /*
+     * Class objects inherit from the class of classes unless they inherit
+     * from some subclass of it. Enforce this right now.
+     */
+
+    clsPtr->thisPtr->selfCls = fPtr->classCls;
+
+    /*
+     * Classes are subclasses of oo::object, i.e. the objects they create are
+     * objects.
+     */
+
+    clsPtr->superclasses.num = 1;
+    clsPtr->superclasses.list = ckalloc(sizeof(Class *));
+    clsPtr->superclasses.list[0] = fPtr->objectCls;
+
+    /*
+     * Finish connecting the class structure to the object structure.
+     */
+
+    clsPtr->thisPtr->classPtr = clsPtr;
+
+    /*
+     * That's the complicated bit. Now fill in the rest of the non-zero/NULL
+     * fields.
+     */
+
+    clsPtr->refCount = 1;
+    Tcl_InitObjHashTable(&clsPtr->classMethods);
+    return clsPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_NewObjectInstance --
+ *
+ *	Allocate a new instance of an object.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Object
+Tcl_NewObjectInstance(
+    Tcl_Interp *interp,		/* Interpreter context. */
+    Tcl_Class cls,		/* Class to create an instance of. */
+    const char *nameStr,	/* Name of object to create, or NULL to ask
+				 * the code to pick its own unique name. */
+    const char *nsNameStr,	/* Name of namespace to create inside object,
+				 * or NULL to ask the code to pick its own
+				 * unique name. */
+    int objc,			/* Number of arguments. Negative value means
+				 * do not call constructor. */
+    Tcl_Obj *const *objv,	/* Argument list. */
+    int skip)			/* Number of arguments to _not_ pass to the
+				 * constructor. */
+{
+    register Class *classPtr = (Class *) cls;
+    Foundation *fPtr = GetFoundation(interp);
+    Object *oPtr;
+
+    /*
+     * Check if we're going to create an object over an existing command;
+     * that's not allowed.
+     */
+
+    if (nameStr && Tcl_FindCommand(interp, nameStr, NULL,
+	    TCL_NAMESPACE_ONLY)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create object \"%s\": command already exists with"
+		" that name", nameStr));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "OVERWRITE_OBJECT", NULL);
+	return NULL;
+    }
+
+    /*
+     * Create the object.
+     */
+
+    oPtr = AllocObject(interp, nameStr, nsNameStr);
+    oPtr->selfCls = classPtr;
+    TclOOAddToInstances(oPtr, classPtr);
+
+    /*
+     * Check to see if we're really creating a class. If so, allocate the
+     * class structure as well.
+     */
+
+    if (TclOOIsReachable(fPtr->classCls, classPtr)) {
+	/*
+	 * Is a class, so attach a class structure. Note that the AllocClass
+	 * function splices the structure into the object, so we don't have
+	 * to. Once that's done, we need to repatch the object to have the
+	 * right class since AllocClass interferes with that.
+	 */
+
+	AllocClass(interp, oPtr);
+	oPtr->selfCls = classPtr;
+	TclOOAddToSubclasses(oPtr->classPtr, fPtr->objectCls);
+    }
+
+    /*
+     * Run constructors, except when objc < 0 (a special flag case used for
+     * object cloning only).
+     */
+
+    if (objc >= 0) {
+	CallContext *contextPtr =
+		TclOOGetCallContext(oPtr, NULL, CONSTRUCTOR, NULL);
+
+	if (contextPtr != NULL) {
+	    int isRoot, result;
+	    Tcl_InterpState state;
+
+	    state = Tcl_SaveInterpState(interp, TCL_OK);
+	    contextPtr->callPtr->flags |= CONSTRUCTOR;
+	    contextPtr->skip = skip;
+
+	    /*
+	     * Adjust the ensmble tracking record if necessary. [Bug 3514761]
+	     */
+
+	    isRoot = TclInitRewriteEnsemble(interp, skip, skip, objv);
+	    result = Tcl_NRCallObjProc(interp, TclOOInvokeContext, contextPtr,
+		    objc, objv);
+
+	    if (isRoot) {
+		TclResetRewriteEnsemble(interp, 1);
+	    }
+
+	    /*
+	     * It's an error if the object was whacked in the constructor.
+	     * Force this if it isn't already an error (don't want to lose
+	     * errors by accident...) [Bug 2903011]
+	     */
+
+	    if (result != TCL_ERROR && Deleted(oPtr)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"object deleted in constructor", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OO", "STILLBORN", NULL);
+		result = TCL_ERROR;
+	    }
+	    TclOODeleteContext(contextPtr);
+	    if (result != TCL_OK) {
+		Tcl_DiscardInterpState(state);
+
+		/*
+		 * Take care to not delete a deleted object; that would be
+		 * bad. [Bug 2903011] Also take care to make sure that we have
+		 * the name of the command before we delete it. [Bug
+		 * 9dd1bd7a74]
+		 */
+
+		if (!Deleted(oPtr)) {
+		    (void) TclOOObjectName(interp, oPtr);
+		    Tcl_DeleteCommandFromToken(interp, oPtr->command);
+		}
+		return NULL;
+	    }
+	    Tcl_RestoreInterpState(interp, state);
+	}
+    }
+
+    return (Tcl_Object) oPtr;
+}
+
+int
+TclNRNewObjectInstance(
+    Tcl_Interp *interp,		/* Interpreter context. */
+    Tcl_Class cls,		/* Class to create an instance of. */
+    const char *nameStr,	/* Name of object to create, or NULL to ask
+				 * the code to pick its own unique name. */
+    const char *nsNameStr,	/* Name of namespace to create inside object,
+				 * or NULL to ask the code to pick its own
+				 * unique name. */
+    int objc,			/* Number of arguments. Negative value means
+				 * do not call constructor. */
+    Tcl_Obj *const *objv,	/* Argument list. */
+    int skip,			/* Number of arguments to _not_ pass to the
+				 * constructor. */
+    Tcl_Object *objectPtr)	/* Place to write the object reference upon
+				 * successful allocation. */
+{
+    register Class *classPtr = (Class *) cls;
+    Foundation *fPtr = GetFoundation(interp);
+    CallContext *contextPtr;
+    Tcl_InterpState state;
+    Object *oPtr;
+
+    /*
+     * Check if we're going to create an object over an existing command;
+     * that's not allowed.
+     */
+
+    if (nameStr && Tcl_FindCommand(interp, nameStr, NULL,
+	    TCL_NAMESPACE_ONLY)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create object \"%s\": command already exists with"
+		" that name", nameStr));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "OVERWRITE_OBJECT", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create the object.
+     */
+
+    oPtr = AllocObject(interp, nameStr, nsNameStr);
+    oPtr->selfCls = classPtr;
+    TclOOAddToInstances(oPtr, classPtr);
+
+    /*
+     * Check to see if we're really creating a class. If so, allocate the
+     * class structure as well.
+     */
+
+    if (TclOOIsReachable(fPtr->classCls, classPtr)) {
+	/*
+	 * Is a class, so attach a class structure. Note that the AllocClass
+	 * function splices the structure into the object, so we don't have
+	 * to. Once that's done, we need to repatch the object to have the
+	 * right class since AllocClass interferes with that.
+	 */
+
+	AllocClass(interp, oPtr);
+	oPtr->selfCls = classPtr;
+	TclOOAddToSubclasses(oPtr->classPtr, fPtr->objectCls);
+    }
+
+    /*
+     * Run constructors, except when objc < 0 (a special flag case used for
+     * object cloning only). If there aren't any constructors, we do nothing.
+     */
+
+    if (objc < 0) {
+	*objectPtr = (Tcl_Object) oPtr;
+	return TCL_OK;
+    }
+    contextPtr = TclOOGetCallContext(oPtr, NULL, CONSTRUCTOR, NULL);
+    if (contextPtr == NULL) {
+	*objectPtr = (Tcl_Object) oPtr;
+	return TCL_OK;
+    }
+
+    state = Tcl_SaveInterpState(interp, TCL_OK);
+    contextPtr->callPtr->flags |= CONSTRUCTOR;
+    contextPtr->skip = skip;
+
+    /*
+     * Adjust the ensmble tracking record if necessary. [Bug 3514761]
+     */
+
+    if (TclInitRewriteEnsemble(interp, skip, skip, objv)) {
+	TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL, NULL);
+    }
+
+    /*
+     * Fire off the constructors non-recursively.
+     */
+
+    AddRef(oPtr);
+    TclNRAddCallback(interp, FinalizeAlloc, contextPtr, oPtr, state,
+	    objectPtr);
+    TclPushTailcallPoint(interp);
+    return TclOOInvokeContext(contextPtr, interp, objc, objv);
+}
+
+static int
+FinalizeAlloc(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+    Object *oPtr = data[1];
+    Tcl_InterpState state = data[2];
+    Tcl_Object *objectPtr = data[3];
+
+    /*
+     * It's an error if the object was whacked in the constructor. Force this
+     * if it isn't already an error (don't want to lose errors by accident...)
+     * [Bug 2903011]
+     */
+
+    if (result != TCL_ERROR && Deleted(oPtr)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"object deleted in constructor", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "STILLBORN", NULL);
+	result = TCL_ERROR;
+    }
+    TclOODeleteContext(contextPtr);
+    if (result != TCL_OK) {
+	Tcl_DiscardInterpState(state);
+
+	/*
+	 * Take care to not delete a deleted object; that would be bad. [Bug
+	 * 2903011] Also take care to make sure that we have the name of the
+	 * command before we delete it. [Bug 9dd1bd7a74]
+	 */
+
+	if (!Deleted(oPtr)) {
+	    (void) TclOOObjectName(interp, oPtr);
+	    Tcl_DeleteCommandFromToken(interp, oPtr->command);
+	}
+	DelRef(oPtr);
+	return TCL_ERROR;
+    }
+    Tcl_RestoreInterpState(interp, state);
+    *objectPtr = (Tcl_Object) oPtr;
+    DelRef(oPtr);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_CopyObjectInstance --
+ *
+ *	Creates a copy of an object. Does not copy the backing namespace,
+ *	since the correct way to do that (e.g., shallow/deep) depends on the
+ *	object/class's own policies.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Object
+Tcl_CopyObjectInstance(
+    Tcl_Interp *interp,
+    Tcl_Object sourceObject,
+    const char *targetName,
+    const char *targetNamespaceName)
+{
+    Object *oPtr = (Object *) sourceObject, *o2Ptr;
+    FOREACH_HASH_DECLS;
+    Method *mPtr;
+    Class *mixinPtr;
+    CallContext *contextPtr;
+    Tcl_Obj *keyPtr, *filterObj, *variableObj, *args[3];
+    int i, result;
+
+    /*
+     * Sanity check.
+     */
+
+    if (IsRootClass(oPtr)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"may not clone the class of classes", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "CLONING_CLASS", NULL);
+	return NULL;
+    }
+
+    /*
+     * Build the instance. Note that this does not run any constructors.
+     */
+
+    o2Ptr = (Object *) Tcl_NewObjectInstance(interp,
+	    (Tcl_Class) oPtr->selfCls, targetName, targetNamespaceName, -1,
+	    NULL, -1);
+    if (o2Ptr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * Copy the object-local methods to the new object.
+     */
+
+    if (oPtr->methodsPtr) {
+	FOREACH_HASH(keyPtr, mPtr, oPtr->methodsPtr) {
+	    if (CloneObjectMethod(interp, o2Ptr, mPtr, keyPtr) != TCL_OK) {
+		Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+		return NULL;
+	    }
+	}
+    }
+
+    /*
+     * Copy the object's mixin references to the new object.
+     */
+
+    FOREACH(mixinPtr, o2Ptr->mixins) {
+	if (mixinPtr && mixinPtr != o2Ptr->selfCls) {
+	    TclOORemoveFromInstances(o2Ptr, mixinPtr);
+	}
+    }
+    DUPLICATE(o2Ptr->mixins, oPtr->mixins, Class *);
+    FOREACH(mixinPtr, o2Ptr->mixins) {
+	if (mixinPtr && mixinPtr != o2Ptr->selfCls) {
+	    TclOOAddToInstances(o2Ptr, mixinPtr);
+	}
+    }
+
+    /*
+     * Copy the object's filter list to the new object.
+     */
+
+    DUPLICATE(o2Ptr->filters, oPtr->filters, Tcl_Obj *);
+    FOREACH(filterObj, o2Ptr->filters) {
+	Tcl_IncrRefCount(filterObj);
+    }
+
+    /*
+     * Copy the object's variable resolution list to the new object.
+     */
+
+    DUPLICATE(o2Ptr->variables, oPtr->variables, Tcl_Obj *);
+    FOREACH(variableObj, o2Ptr->variables) {
+	Tcl_IncrRefCount(variableObj);
+    }
+
+    /*
+     * Copy the object's flags to the new object, clearing those that must be
+     * kept object-local. The duplicate is never deleted at this point, nor is
+     * it the root of the object system or in the midst of processing a filter
+     * call.
+     */
+
+    o2Ptr->flags = oPtr->flags & ~(
+	    OBJECT_DELETED | ROOT_OBJECT | ROOT_CLASS | FILTER_HANDLING);
+
+    /*
+     * Copy the object's metadata.
+     */
+
+    if (oPtr->metadataPtr != NULL) {
+	Tcl_ObjectMetadataType *metadataTypePtr;
+	ClientData value, duplicate;
+
+	FOREACH_HASH(metadataTypePtr, value, oPtr->metadataPtr) {
+	    if (metadataTypePtr->cloneProc == NULL) {
+		duplicate = value;
+	    } else {
+		if (metadataTypePtr->cloneProc(interp, value,
+			&duplicate) != TCL_OK) {
+		    Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+		    return NULL;
+		}
+	    }
+	    if (duplicate != NULL) {
+		Tcl_ObjectSetMetadata((Tcl_Object) o2Ptr, metadataTypePtr,
+			duplicate);
+	    }
+	}
+    }
+
+    /*
+     * Copy the class, if present. Note that if there is a class present in
+     * the source object, there must also be one in the copy.
+     */
+
+    if (oPtr->classPtr != NULL) {
+	Class *clsPtr = oPtr->classPtr;
+	Class *cls2Ptr = o2Ptr->classPtr;
+	Class *superPtr;
+
+	/*
+	 * Copy the class flags across.
+	 */
+
+	cls2Ptr->flags = clsPtr->flags;
+
+	/*
+	 * Ensure that the new class's superclass structure is the same as the
+	 * old class's.
+	 */
+
+	FOREACH(superPtr, cls2Ptr->superclasses) {
+	    TclOORemoveFromSubclasses(cls2Ptr, superPtr);
+	}
+	if (cls2Ptr->superclasses.num) {
+	    cls2Ptr->superclasses.list = ckrealloc(cls2Ptr->superclasses.list,
+		    sizeof(Class *) * clsPtr->superclasses.num);
+	} else {
+	    cls2Ptr->superclasses.list =
+		    ckalloc(sizeof(Class *) * clsPtr->superclasses.num);
+	}
+	memcpy(cls2Ptr->superclasses.list, clsPtr->superclasses.list,
+		sizeof(Class *) * clsPtr->superclasses.num);
+	cls2Ptr->superclasses.num = clsPtr->superclasses.num;
+	FOREACH(superPtr, cls2Ptr->superclasses) {
+	    TclOOAddToSubclasses(cls2Ptr, superPtr);
+	}
+
+	/*
+	 * Duplicate the source class's filters.
+	 */
+
+	DUPLICATE(cls2Ptr->filters, clsPtr->filters, Tcl_Obj *);
+	FOREACH(filterObj, cls2Ptr->filters) {
+	    Tcl_IncrRefCount(filterObj);
+	}
+
+	/*
+	 * Copy the source class's variable resolution list.
+	 */
+
+	DUPLICATE(cls2Ptr->variables, clsPtr->variables, Tcl_Obj *);
+	FOREACH(variableObj, cls2Ptr->variables) {
+	    Tcl_IncrRefCount(variableObj);
+	}
+
+	/*
+	 * Duplicate the source class's mixins (which cannot be circular
+	 * references to the duplicate).
+	 */
+
+	FOREACH(mixinPtr, cls2Ptr->mixins) {
+	    TclOORemoveFromMixinSubs(cls2Ptr, mixinPtr);
+	}
+	if (cls2Ptr->mixins.num != 0) {
+	    ckfree(clsPtr->mixins.list);
+	}
+	DUPLICATE(cls2Ptr->mixins, clsPtr->mixins, Class *);
+	FOREACH(mixinPtr, cls2Ptr->mixins) {
+	    TclOOAddToMixinSubs(cls2Ptr, mixinPtr);
+	}
+
+	/*
+	 * Duplicate the source class's methods, constructor and destructor.
+	 */
+
+	FOREACH_HASH(keyPtr, mPtr, &clsPtr->classMethods) {
+	    if (CloneClassMethod(interp, cls2Ptr, mPtr, keyPtr,
+		    NULL) != TCL_OK) {
+		Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+		return NULL;
+	    }
+	}
+	if (clsPtr->constructorPtr) {
+	    if (CloneClassMethod(interp, cls2Ptr, clsPtr->constructorPtr,
+		    NULL, &cls2Ptr->constructorPtr) != TCL_OK) {
+		Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+		return NULL;
+	    }
+	}
+	if (clsPtr->destructorPtr) {
+	    if (CloneClassMethod(interp, cls2Ptr, clsPtr->destructorPtr, NULL,
+		    &cls2Ptr->destructorPtr) != TCL_OK) {
+		Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+		return NULL;
+	    }
+	}
+
+	/*
+	 * Duplicate the class's metadata.
+	 */
+
+	if (clsPtr->metadataPtr != NULL) {
+	    Tcl_ObjectMetadataType *metadataTypePtr;
+	    ClientData value, duplicate;
+
+	    FOREACH_HASH(metadataTypePtr, value, clsPtr->metadataPtr) {
+		if (metadataTypePtr->cloneProc == NULL) {
+		    duplicate = value;
+		} else {
+		    if (metadataTypePtr->cloneProc(interp, value,
+			    &duplicate) != TCL_OK) {
+			Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+			return NULL;
+		    }
+		}
+		if (duplicate != NULL) {
+		    Tcl_ClassSetMetadata((Tcl_Class) cls2Ptr, metadataTypePtr,
+			    duplicate);
+		}
+	    }
+	}
+    }
+
+    TclResetRewriteEnsemble(interp, 1);
+    contextPtr = TclOOGetCallContext(o2Ptr, oPtr->fPtr->clonedName, 0, NULL);
+    if (contextPtr) {
+	args[0] = TclOOObjectName(interp, o2Ptr);
+	args[1] = oPtr->fPtr->clonedName;
+	args[2] = TclOOObjectName(interp, oPtr);
+	Tcl_IncrRefCount(args[0]);
+	Tcl_IncrRefCount(args[1]);
+	Tcl_IncrRefCount(args[2]);
+	result = Tcl_NRCallObjProc(interp, TclOOInvokeContext, contextPtr, 3,
+		args);
+	TclDecrRefCount(args[0]);
+	TclDecrRefCount(args[1]);
+	TclDecrRefCount(args[2]);
+	TclOODeleteContext(contextPtr);
+	if (result == TCL_ERROR) {
+	    Tcl_AddErrorInfo(interp,
+		    "\n    (while performing post-copy callback)");
+	}
+	if (result != TCL_OK) {
+	    Tcl_DeleteCommandFromToken(interp, o2Ptr->command);
+	    return NULL;
+	}
+    }
+
+    return (Tcl_Object) o2Ptr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * CloneObjectMethod, CloneClassMethod --
+ *
+ *	Helper functions used for cloning methods. They work identically to
+ *	each other, except for the difference between them in how they
+ *	register the cloned method on a successful clone.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+CloneObjectMethod(
+    Tcl_Interp *interp,
+    Object *oPtr,
+    Method *mPtr,
+    Tcl_Obj *namePtr)
+{
+    if (mPtr->typePtr == NULL) {
+	Tcl_NewInstanceMethod(interp, (Tcl_Object) oPtr, namePtr,
+		mPtr->flags & PUBLIC_METHOD, NULL, NULL);
+    } else if (mPtr->typePtr->cloneProc) {
+	ClientData newClientData;
+
+	if (mPtr->typePtr->cloneProc(interp, mPtr->clientData,
+		&newClientData) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_NewInstanceMethod(interp, (Tcl_Object) oPtr, namePtr,
+		mPtr->flags & PUBLIC_METHOD, mPtr->typePtr, newClientData);
+    } else {
+	Tcl_NewInstanceMethod(interp, (Tcl_Object) oPtr, namePtr,
+		mPtr->flags & PUBLIC_METHOD, mPtr->typePtr, mPtr->clientData);
+    }
+    return TCL_OK;
+}
+
+static int
+CloneClassMethod(
+    Tcl_Interp *interp,
+    Class *clsPtr,
+    Method *mPtr,
+    Tcl_Obj *namePtr,
+    Method **m2PtrPtr)
+{
+    Method *m2Ptr;
+
+    if (mPtr->typePtr == NULL) {
+	m2Ptr = (Method *) Tcl_NewMethod(interp, (Tcl_Class) clsPtr,
+		namePtr, mPtr->flags & PUBLIC_METHOD, NULL, NULL);
+    } else if (mPtr->typePtr->cloneProc) {
+	ClientData newClientData;
+
+	if (mPtr->typePtr->cloneProc(interp, mPtr->clientData,
+		&newClientData) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	m2Ptr = (Method *) Tcl_NewMethod(interp, (Tcl_Class) clsPtr,
+		namePtr, mPtr->flags & PUBLIC_METHOD, mPtr->typePtr,
+		newClientData);
+    } else {
+	m2Ptr = (Method *) Tcl_NewMethod(interp, (Tcl_Class) clsPtr,
+		namePtr, mPtr->flags & PUBLIC_METHOD, mPtr->typePtr,
+		mPtr->clientData);
+    }
+    if (m2PtrPtr != NULL) {
+	*m2PtrPtr = m2Ptr;
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_ClassGetMetadata, Tcl_ClassSetMetadata, Tcl_ObjectGetMetadata,
+ * Tcl_ObjectSetMetadata --
+ *
+ *	Metadata management API. The metadata system allows code in extensions
+ *	to attach arbitrary non-NULL pointers to objects and classes without
+ *	the different things that might be interested being able to interfere
+ *	with each other. Apart from non-NULL-ness, these routines attach no
+ *	interpretation to the meaning of the metadata pointers.
+ *
+ *	The Tcl_*GetMetadata routines get the metadata pointer attached that
+ *	has been related with a particular type, or NULL if no metadata
+ *	associated with the given type has been attached.
+ *
+ *	The Tcl_*SetMetadata routines set or delete the metadata pointer that
+ *	is related to a particular type. The value associated with the type is
+ *	deleted (if present; no-op otherwise) if the value is NULL, and
+ *	attached (replacing the previous value, which is deleted if present)
+ *	otherwise. This means it is impossible to attach a NULL value for any
+ *	metadata type.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_ClassGetMetadata(
+    Tcl_Class clazz,
+    const Tcl_ObjectMetadataType *typePtr)
+{
+    Class *clsPtr = (Class *) clazz;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * If there's no metadata store attached, the type in question has
+     * definitely not been attached either!
+     */
+
+    if (clsPtr->metadataPtr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * There is a metadata store, so look in it for the given type.
+     */
+
+    hPtr = Tcl_FindHashEntry(clsPtr->metadataPtr, (char *) typePtr);
+
+    /*
+     * Return the metadata value if we found it, otherwise NULL.
+     */
+
+    if (hPtr == NULL) {
+	return NULL;
+    }
+    return Tcl_GetHashValue(hPtr);
+}
+
+void
+Tcl_ClassSetMetadata(
+    Tcl_Class clazz,
+    const Tcl_ObjectMetadataType *typePtr,
+    ClientData metadata)
+{
+    Class *clsPtr = (Class *) clazz;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    /*
+     * Attach the metadata store if not done already.
+     */
+
+    if (clsPtr->metadataPtr == NULL) {
+	if (metadata == NULL) {
+	    return;
+	}
+	clsPtr->metadataPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(clsPtr->metadataPtr, TCL_ONE_WORD_KEYS);
+    }
+
+    /*
+     * If the metadata is NULL, we're deleting the metadata for the type.
+     */
+
+    if (metadata == NULL) {
+	hPtr = Tcl_FindHashEntry(clsPtr->metadataPtr, (char *) typePtr);
+	if (hPtr != NULL) {
+	    typePtr->deleteProc(Tcl_GetHashValue(hPtr));
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+	return;
+    }
+
+    /*
+     * Otherwise we're attaching the metadata. Note that if there was already
+     * some metadata attached of this type, we delete that first.
+     */
+
+    hPtr = Tcl_CreateHashEntry(clsPtr->metadataPtr, (char *) typePtr, &isNew);
+    if (!isNew) {
+	typePtr->deleteProc(Tcl_GetHashValue(hPtr));
+    }
+    Tcl_SetHashValue(hPtr, metadata);
+}
+
+ClientData
+Tcl_ObjectGetMetadata(
+    Tcl_Object object,
+    const Tcl_ObjectMetadataType *typePtr)
+{
+    Object *oPtr = (Object *) object;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * If there's no metadata store attached, the type in question has
+     * definitely not been attached either!
+     */
+
+    if (oPtr->metadataPtr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * There is a metadata store, so look in it for the given type.
+     */
+
+    hPtr = Tcl_FindHashEntry(oPtr->metadataPtr, (char *) typePtr);
+
+    /*
+     * Return the metadata value if we found it, otherwise NULL.
+     */
+
+    if (hPtr == NULL) {
+	return NULL;
+    }
+    return Tcl_GetHashValue(hPtr);
+}
+
+void
+Tcl_ObjectSetMetadata(
+    Tcl_Object object,
+    const Tcl_ObjectMetadataType *typePtr,
+    ClientData metadata)
+{
+    Object *oPtr = (Object *) object;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    /*
+     * Attach the metadata store if not done already.
+     */
+
+    if (oPtr->metadataPtr == NULL) {
+	if (metadata == NULL) {
+	    return;
+	}
+	oPtr->metadataPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(oPtr->metadataPtr, TCL_ONE_WORD_KEYS);
+    }
+
+    /*
+     * If the metadata is NULL, we're deleting the metadata for the type.
+     */
+
+    if (metadata == NULL) {
+	hPtr = Tcl_FindHashEntry(oPtr->metadataPtr, (char *) typePtr);
+	if (hPtr != NULL) {
+	    typePtr->deleteProc(Tcl_GetHashValue(hPtr));
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+	return;
+    }
+
+    /*
+     * Otherwise we're attaching the metadata. Note that if there was already
+     * some metadata attached of this type, we delete that first.
+     */
+
+    hPtr = Tcl_CreateHashEntry(oPtr->metadataPtr, (char *) typePtr, &isNew);
+    if (!isNew) {
+	typePtr->deleteProc(Tcl_GetHashValue(hPtr));
+    }
+    Tcl_SetHashValue(hPtr, metadata);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * PublicObjectCmd, PrivateObjectCmd, TclOOInvokeObject --
+ *
+ *	Main entry point for object invokations. The Public* and Private*
+ *	wrapper functions (implementations of both object instance commands
+ *	and [my]) are just thin wrappers round the main TclOOObjectCmdCore
+ *	function. Note that the core is function is NRE-aware.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+PublicObjectCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    return Tcl_NRCallObjProc(interp, PublicNRObjectCmd, clientData,objc,objv);
+}
+
+static int
+PublicNRObjectCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    return TclOOObjectCmdCore(clientData, interp, objc, objv, PUBLIC_METHOD,
+	    NULL);
+}
+
+static int
+PrivateObjectCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    return Tcl_NRCallObjProc(interp, PrivateNRObjectCmd,clientData,objc,objv);
+}
+
+static int
+PrivateNRObjectCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    return TclOOObjectCmdCore(clientData, interp, objc, objv, 0, NULL);
+}
+
+int
+TclOOInvokeObject(
+    Tcl_Interp *interp,		/* Interpreter for commands, variables,
+				 * results, error reporting, etc. */
+    Tcl_Object object,		/* The object to invoke. */
+    Tcl_Class startCls,		/* Where in the class chain to start the
+				 * invoke from, or NULL to traverse the whole
+				 * chain including filters. */
+    int publicPrivate,		/* Whether this is an invoke from a public
+				 * context (PUBLIC_METHOD), a private context
+				 * (PRIVATE_METHOD), or a *really* private
+				 * context (any other value; conventionally
+				 * 0). */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* Array of argument objects. It is assumed
+				 * that the name of the method to invoke will
+				 * be at index 1. */
+{
+    switch (publicPrivate) {
+    case PUBLIC_METHOD:
+	return TclOOObjectCmdCore((Object *) object, interp, objc, objv,
+		PUBLIC_METHOD, (Class *) startCls);
+    case PRIVATE_METHOD:
+	return TclOOObjectCmdCore((Object *) object, interp, objc, objv,
+		PRIVATE_METHOD, (Class *) startCls);
+    default:
+	return TclOOObjectCmdCore((Object *) object, interp, objc, objv, 0,
+		(Class *) startCls);
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOObjectCmdCore, FinalizeObjectCall --
+ *
+ *	Main function for object invokations. Does call chain creation,
+ *	management and invokation. The function FinalizeObjectCall exists to
+ *	clean up after the non-recursive processing of TclOOObjectCmdCore.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOObjectCmdCore(
+    Object *oPtr,		/* The object being invoked. */
+    Tcl_Interp *interp,		/* The interpreter containing the object. */
+    int objc,			/* How many arguments are being passed in. */
+    Tcl_Obj *const *objv,	/* The array of arguments. */
+    int flags,			/* Whether this is an invokation through the
+				 * public or the private command interface. */
+    Class *startCls)		/* Where to start in the call chain, or NULL
+				 * if we are to start at the front with
+				 * filters and the object's methods (which is
+				 * the normal case). */
+{
+    CallContext *contextPtr;
+    Tcl_Obj *methodNamePtr;
+    int result;
+
+    /*
+     * If we've no method name, throw this directly into the unknown
+     * processing.
+     */
+
+    if (objc < 2) {
+	flags |= FORCE_UNKNOWN;
+	methodNamePtr = NULL;
+	goto noMapping;
+    }
+
+    /*
+     * Give plugged in code a chance to remap the method name.
+     */
+
+    methodNamePtr = objv[1];
+    if (oPtr->mapMethodNameProc != NULL) {
+	register Class **startClsPtr = &startCls;
+	Tcl_Obj *mappedMethodName = Tcl_DuplicateObj(methodNamePtr);
+
+	result = oPtr->mapMethodNameProc(interp, (Tcl_Object) oPtr,
+		(Tcl_Class *) startClsPtr, mappedMethodName);
+	if (result != TCL_OK) {
+	    TclDecrRefCount(mappedMethodName);
+	    if (result == TCL_BREAK) {
+		goto noMapping;
+	    } else if (result == TCL_ERROR) {
+		Tcl_AddErrorInfo(interp, "\n    (while mapping method name)");
+	    }
+	    return result;
+	}
+
+	/*
+	 * Get the call chain for the remapped name.
+	 */
+
+	Tcl_IncrRefCount(mappedMethodName);
+	contextPtr = TclOOGetCallContext(oPtr, mappedMethodName,
+		flags | (oPtr->flags & FILTER_HANDLING), methodNamePtr);
+	TclDecrRefCount(mappedMethodName);
+	if (contextPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "impossible to invoke method \"%s\": no defined method or"
+		    " unknown method", TclGetString(methodNamePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD_MAPPED",
+		    TclGetString(methodNamePtr), NULL);
+	    return TCL_ERROR;
+	}
+    } else {
+	/*
+	 * Get the call chain.
+	 */
+
+    noMapping:
+	contextPtr = TclOOGetCallContext(oPtr, methodNamePtr,
+		flags | (oPtr->flags & FILTER_HANDLING), NULL);
+	if (contextPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "impossible to invoke method \"%s\": no defined method or"
+		    " unknown method", TclGetString(methodNamePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		    TclGetString(methodNamePtr), NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Check to see if we need to apply magical tricks to start part way
+     * through the call chain.
+     */
+
+    if (startCls != NULL) {
+	for (; contextPtr->index < contextPtr->callPtr->numChain;
+		contextPtr->index++) {
+	    register struct MInvoke *miPtr =
+		    &contextPtr->callPtr->chain[contextPtr->index];
+
+	    if (miPtr->isFilter) {
+		continue;
+	    }
+	    if (miPtr->mPtr->declaringClassPtr == startCls) {
+		break;
+	    }
+	}
+	if (contextPtr->index >= contextPtr->callPtr->numChain) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "no valid method implementation", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		    TclGetString(methodNamePtr), NULL);
+	    TclOODeleteContext(contextPtr);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Invoke the call chain, locking the object structure against deletion
+     * for the duration.
+     */
+
+    TclNRAddCallback(interp, FinalizeObjectCall, contextPtr, NULL,NULL,NULL);
+    return TclOOInvokeContext(contextPtr, interp, objc, objv);
+}
+
+static int
+FinalizeObjectCall(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    /*
+     * Dispose of the call chain, which drops the lock on the object's
+     * structure.
+     */
+
+    TclOODeleteContext(data[0]);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_ObjectContextInvokeNext, TclNRObjectContextInvokeNext, FinalizeNext --
+ *
+ *	Invokes the next stage of the call chain described in an object
+ *	context. This is the core of the implementation of the [next] command.
+ *	Does not do management of the call-frame stack. Available in public
+ *	(standard API) and private (NRE-aware) forms. FinalizeNext is a
+ *	private function used to clean up in the NRE case.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+Tcl_ObjectContextInvokeNext(
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv,
+    int skip)
+{
+    CallContext *contextPtr = (CallContext *) context;
+    int savedIndex = contextPtr->index;
+    int savedSkip = contextPtr->skip;
+    int result;
+
+    if (contextPtr->index+1 >= contextPtr->callPtr->numChain) {
+	/*
+	 * We're at the end of the chain; generate an error message unless the
+	 * interpreter is being torn down, in which case we might be getting
+	 * here because of methods/destructors doing a [next] (or equivalent)
+	 * unexpectedly.
+	 */
+
+	const char *methodType;
+
+	if (Tcl_InterpDeleted(interp)) {
+	    return TCL_OK;
+	}
+
+	if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+	    methodType = "constructor";
+	} else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+	    methodType = "destructor";
+	} else {
+	    methodType = "method";
+	}
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"no next %s implementation", methodType));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "NOTHING_NEXT", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Advance to the next method implementation in the chain in the method
+     * call context while we process the body. However, need to adjust the
+     * argument-skip control because we're guaranteed to have a single prefix
+     * arg (i.e., 'next') and not the variable amount that can happen because
+     * method invokations (i.e., '$obj meth' and 'my meth'), constructors
+     * (i.e., '$cls new' and '$cls create obj') and destructors (no args at
+     * all) come through the same code.
+     */
+
+    contextPtr->index++;
+    contextPtr->skip = skip;
+
+    /*
+     * Invoke the (advanced) method call context in the caller context.
+     */
+
+    result = Tcl_NRCallObjProc(interp, TclOOInvokeContext, contextPtr, objc,
+	    objv);
+
+    /*
+     * Restore the call chain context index as we've finished the inner invoke
+     * and want to operate in the outer context again.
+     */
+
+    contextPtr->index = savedIndex;
+    contextPtr->skip = savedSkip;
+
+    return result;
+}
+
+int
+TclNRObjectContextInvokeNext(
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv,
+    int skip)
+{
+    register CallContext *contextPtr = (CallContext *) context;
+
+    if (contextPtr->index+1 >= contextPtr->callPtr->numChain) {
+	/*
+	 * We're at the end of the chain; generate an error message unless the
+	 * interpreter is being torn down, in which case we might be getting
+	 * here because of methods/destructors doing a [next] (or equivalent)
+	 * unexpectedly.
+	 */
+
+	const char *methodType;
+
+	if (Tcl_InterpDeleted(interp)) {
+	    return TCL_OK;
+	}
+
+	if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+	    methodType = "constructor";
+	} else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+	    methodType = "destructor";
+	} else {
+	    methodType = "method";
+	}
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"no next %s implementation", methodType));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "NOTHING_NEXT", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Advance to the next method implementation in the chain in the method
+     * call context while we process the body. However, need to adjust the
+     * argument-skip control because we're guaranteed to have a single prefix
+     * arg (i.e., 'next') and not the variable amount that can happen because
+     * method invokations (i.e., '$obj meth' and 'my meth'), constructors
+     * (i.e., '$cls new' and '$cls create obj') and destructors (no args at
+     * all) come through the same code.
+     */
+
+    TclNRAddCallback(interp, FinalizeNext, contextPtr,
+	    INT2PTR(contextPtr->index), INT2PTR(contextPtr->skip), NULL);
+    contextPtr->index++;
+    contextPtr->skip = skip;
+
+    /*
+     * Invoke the (advanced) method call context in the caller context.
+     */
+
+    return TclOOInvokeContext(contextPtr, interp, objc, objv);
+}
+
+static int
+FinalizeNext(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+
+    /*
+     * Restore the call chain context index as we've finished the inner invoke
+     * and want to operate in the outer context again.
+     */
+
+    contextPtr->index = PTR2INT(data[1]);
+    contextPtr->skip = PTR2INT(data[2]);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_GetObjectFromObj --
+ *
+ *	Utility function to get an object from a Tcl_Obj containing its name.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Object
+Tcl_GetObjectFromObj(
+    Tcl_Interp *interp,		/* Interpreter in which to locate the object.
+				 * Will have an error message placed in it if
+				 * the name does not refer to an object. */
+    Tcl_Obj *objPtr)		/* The name of the object to look up, which is
+				 * exactly the name of its public command. */
+{
+    Command *cmdPtr = (Command *) Tcl_GetCommandFromObj(interp, objPtr);
+
+    if (cmdPtr == NULL) {
+	goto notAnObject;
+    }
+    if (cmdPtr->objProc != PublicObjectCmd) {
+	cmdPtr = (Command *) TclGetOriginalCommand((Tcl_Command) cmdPtr);
+	if (cmdPtr == NULL || cmdPtr->objProc != PublicObjectCmd) {
+	    goto notAnObject;
+	}
+    }
+    return cmdPtr->objClientData;
+
+  notAnObject:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "%s does not refer to an object", TclGetString(objPtr)));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "OBJECT", TclGetString(objPtr),
+	    NULL);
+    return NULL;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOIsReachable --
+ *
+ *	Utility function that tests whether a class is a subclass (whether
+ *	directly or indirectly) of another class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOIsReachable(
+    Class *targetPtr,
+    Class *startPtr)
+{
+    int i;
+    Class *superPtr;
+
+  tailRecurse:
+    if (startPtr == targetPtr) {
+	return 1;
+    }
+    if (startPtr->superclasses.num == 1 && startPtr->mixins.num == 0) {
+	startPtr = startPtr->superclasses.list[0];
+	goto tailRecurse;
+    }
+    FOREACH(superPtr, startPtr->superclasses) {
+	if (TclOOIsReachable(targetPtr, superPtr)) {
+	    return 1;
+	}
+    }
+    FOREACH(superPtr, startPtr->mixins) {
+	if (TclOOIsReachable(targetPtr, superPtr)) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOObjectName, Tcl_GetObjectName --
+ *
+ *	Utility function that returns the name of the object. Note that this
+ *	simplifies cache management by keeping the code to do it in one place
+ *	and not sprayed all over. The value returned always has a reference
+ *	count of at least one.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclOOObjectName(
+    Tcl_Interp *interp,
+    Object *oPtr)
+{
+    Tcl_Obj *namePtr;
+
+    if (oPtr->cachedNameObj) {
+	return oPtr->cachedNameObj;
+    }
+    namePtr = Tcl_NewObj();
+    Tcl_GetCommandFullName(interp, oPtr->command, namePtr);
+    Tcl_IncrRefCount(namePtr);
+    oPtr->cachedNameObj = namePtr;
+    return namePtr;
+}
+
+Tcl_Obj *
+Tcl_GetObjectName(
+    Tcl_Interp *interp,
+    Tcl_Object object)
+{
+    return TclOOObjectName(interp, (Object *) object);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * assorted trivial 'getter' functions
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Method
+Tcl_ObjectContextMethod(
+    Tcl_ObjectContext context)
+{
+    CallContext *contextPtr = (CallContext *) context;
+    return (Tcl_Method) contextPtr->callPtr->chain[contextPtr->index].mPtr;
+}
+
+int
+Tcl_ObjectContextIsFiltering(
+    Tcl_ObjectContext context)
+{
+    CallContext *contextPtr = (CallContext *) context;
+    return contextPtr->callPtr->chain[contextPtr->index].isFilter;
+}
+
+Tcl_Object
+Tcl_ObjectContextObject(
+    Tcl_ObjectContext context)
+{
+    return (Tcl_Object) ((CallContext *)context)->oPtr;
+}
+
+int
+Tcl_ObjectContextSkippedArgs(
+    Tcl_ObjectContext context)
+{
+    return ((CallContext *)context)->skip;
+}
+
+Tcl_Namespace *
+Tcl_GetObjectNamespace(
+    Tcl_Object object)
+{
+    return ((Object *)object)->namespacePtr;
+}
+
+Tcl_Command
+Tcl_GetObjectCommand(
+    Tcl_Object object)
+{
+    return ((Object *)object)->command;
+}
+
+Tcl_Class
+Tcl_GetObjectAsClass(
+    Tcl_Object object)
+{
+    return (Tcl_Class) ((Object *)object)->classPtr;
+}
+
+int
+Tcl_ObjectDeleted(
+    Tcl_Object object)
+{
+    return Deleted(object) ? 1 : 0;
+}
+
+Tcl_Object
+Tcl_GetClassAsObject(
+    Tcl_Class clazz)
+{
+    return (Tcl_Object) ((Class *)clazz)->thisPtr;
+}
+
+Tcl_ObjectMapMethodNameProc *
+Tcl_ObjectGetMethodNameMapper(
+    Tcl_Object object)
+{
+    return ((Object *) object)->mapMethodNameProc;
+}
+
+void
+Tcl_ObjectSetMethodNameMapper(
+    Tcl_Object object,
+    Tcl_ObjectMapMethodNameProc *mapMethodNameProc)
+{
+    ((Object *) object)->mapMethodNameProc = mapMethodNameProc;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOO.decls b/generic/tclOO.decls
new file mode 100644
index 0000000..265ba88
--- /dev/null
+++ b/generic/tclOO.decls
@@ -0,0 +1,218 @@
+# tclOO.decls --
+#
+#	This file contains the declarations for all supported public functions
+#	that are exported by the TclOO package that is embedded within the Tcl
+#	library via the stubs table.  This file is used to generate the
+#	tclOODecls.h, tclOOIntDecls.h and tclOOStubInit.c files.
+#
+# Copyright (c) 2008-2013 by Donal K. Fellows.
+#
+# See the file "license.terms" for information on usage and redistribution of
+# this file, and for a DISCLAIMER OF ALL WARRANTIES.
+
+library tclOO
+
+######################################################################
+# Public API, exposed for general users of TclOO.
+#
+
+interface tclOO
+hooks tclOOInt
+scspec TCLAPI
+
+declare 0 {
+    Tcl_Object Tcl_CopyObjectInstance(Tcl_Interp *interp,
+	    Tcl_Object sourceObject, const char *targetName,
+	    const char *targetNamespaceName)
+}
+declare 1 {
+    Tcl_Object Tcl_GetClassAsObject(Tcl_Class clazz)
+}
+declare 2 {
+    Tcl_Class Tcl_GetObjectAsClass(Tcl_Object object)
+}
+declare 3 {
+    Tcl_Command Tcl_GetObjectCommand(Tcl_Object object)
+}
+declare 4 {
+    Tcl_Object Tcl_GetObjectFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr)
+}
+declare 5 {
+    Tcl_Namespace *Tcl_GetObjectNamespace(Tcl_Object object)
+}
+declare 6 {
+    Tcl_Class Tcl_MethodDeclarerClass(Tcl_Method method)
+}
+declare 7 {
+    Tcl_Object Tcl_MethodDeclarerObject(Tcl_Method method)
+}
+declare 8 {
+    int Tcl_MethodIsPublic(Tcl_Method method)
+}
+declare 9 {
+    int Tcl_MethodIsType(Tcl_Method method, const Tcl_MethodType *typePtr,
+	    ClientData *clientDataPtr)
+}
+declare 10 {
+    Tcl_Obj *Tcl_MethodName(Tcl_Method method)
+}
+declare 11 {
+    Tcl_Method Tcl_NewInstanceMethod(Tcl_Interp *interp, Tcl_Object object,
+	    Tcl_Obj *nameObj, int isPublic, const Tcl_MethodType *typePtr,
+	    ClientData clientData)
+}
+declare 12 {
+    Tcl_Method Tcl_NewMethod(Tcl_Interp *interp, Tcl_Class cls,
+	    Tcl_Obj *nameObj, int isPublic, const Tcl_MethodType *typePtr,
+	    ClientData clientData)
+}
+declare 13 {
+    Tcl_Object Tcl_NewObjectInstance(Tcl_Interp *interp, Tcl_Class cls,
+	    const char *nameStr, const char *nsNameStr, int objc,
+	    Tcl_Obj *const *objv, int skip)
+}
+declare 14 {
+    int Tcl_ObjectDeleted(Tcl_Object object)
+}
+declare 15 {
+    int Tcl_ObjectContextIsFiltering(Tcl_ObjectContext context)
+}
+declare 16 {
+    Tcl_Method Tcl_ObjectContextMethod(Tcl_ObjectContext context)
+}
+declare 17 {
+    Tcl_Object Tcl_ObjectContextObject(Tcl_ObjectContext context)
+}
+declare 18 {
+    int Tcl_ObjectContextSkippedArgs(Tcl_ObjectContext context)
+}
+declare 19 {
+    ClientData Tcl_ClassGetMetadata(Tcl_Class clazz,
+	    const Tcl_ObjectMetadataType *typePtr)
+}
+declare 20 {
+    void Tcl_ClassSetMetadata(Tcl_Class clazz,
+	    const Tcl_ObjectMetadataType *typePtr, ClientData metadata)
+}
+declare 21 {
+    ClientData Tcl_ObjectGetMetadata(Tcl_Object object,
+	    const Tcl_ObjectMetadataType *typePtr)
+}
+declare 22 {
+    void Tcl_ObjectSetMetadata(Tcl_Object object,
+	    const Tcl_ObjectMetadataType *typePtr, ClientData metadata)
+}
+declare 23 {
+    int Tcl_ObjectContextInvokeNext(Tcl_Interp *interp,
+	    Tcl_ObjectContext context, int objc, Tcl_Obj *const *objv,
+	    int skip)
+}
+declare 24 {
+    Tcl_ObjectMapMethodNameProc *Tcl_ObjectGetMethodNameMapper(
+	    Tcl_Object object)
+}
+declare 25 {
+    void Tcl_ObjectSetMethodNameMapper(Tcl_Object object,
+	    Tcl_ObjectMapMethodNameProc *mapMethodNameProc)
+}
+declare 26 {
+    void Tcl_ClassSetConstructor(Tcl_Interp *interp, Tcl_Class clazz,
+	    Tcl_Method method)
+}
+declare 27 {
+    void Tcl_ClassSetDestructor(Tcl_Interp *interp, Tcl_Class clazz,
+	    Tcl_Method method)
+}
+declare 28 {
+    Tcl_Obj *Tcl_GetObjectName(Tcl_Interp *interp, Tcl_Object object)
+}
+
+######################################################################
+# Private API, exposed to support advanced OO systems that plug in on top of
+# TclOO; not intended for general use and does not have any commitment to
+# long-term support.
+#
+
+interface tclOOInt
+
+declare 0 {
+    Tcl_Object TclOOGetDefineCmdContext(Tcl_Interp *interp)
+}
+declare 1 {
+    Tcl_Method TclOOMakeProcInstanceMethod(Tcl_Interp *interp, Object *oPtr,
+	    int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+	    const Tcl_MethodType *typePtr, ClientData clientData,
+	    Proc **procPtrPtr)
+}
+declare 2 {
+    Tcl_Method TclOOMakeProcMethod(Tcl_Interp *interp, Class *clsPtr,
+	    int flags, Tcl_Obj *nameObj, const char *namePtr,
+	    Tcl_Obj *argsObj, Tcl_Obj *bodyObj, const Tcl_MethodType *typePtr,
+	    ClientData clientData, Proc **procPtrPtr)
+}
+declare 3 {
+    Method *TclOONewProcInstanceMethod(Tcl_Interp *interp, Object *oPtr,
+	    int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+	    ProcedureMethod **pmPtrPtr)
+}
+declare 4 {
+    Method *TclOONewProcMethod(Tcl_Interp *interp, Class *clsPtr,
+	    int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+	    ProcedureMethod **pmPtrPtr)
+}
+declare 5 {
+    int TclOOObjectCmdCore(Object *oPtr, Tcl_Interp *interp, int objc,
+	    Tcl_Obj *const *objv, int publicOnly, Class *startCls)
+}
+declare 6 {
+    int TclOOIsReachable(Class *targetPtr, Class *startPtr)
+}
+declare 7 {
+    Method *TclOONewForwardMethod(Tcl_Interp *interp, Class *clsPtr,
+	    int isPublic, Tcl_Obj *nameObj, Tcl_Obj *prefixObj)
+}
+declare 8 {
+    Method *TclOONewForwardInstanceMethod(Tcl_Interp *interp, Object *oPtr,
+	    int isPublic, Tcl_Obj *nameObj, Tcl_Obj *prefixObj)
+}
+declare 9 {
+    Tcl_Method TclOONewProcInstanceMethodEx(Tcl_Interp *interp,
+	    Tcl_Object oPtr, TclOO_PreCallProc *preCallPtr,
+	    TclOO_PostCallProc *postCallPtr, ProcErrorProc *errProc,
+	    ClientData clientData, Tcl_Obj *nameObj, Tcl_Obj *argsObj,
+	    Tcl_Obj *bodyObj, int flags, void **internalTokenPtr)
+}
+declare 10 {
+    Tcl_Method TclOONewProcMethodEx(Tcl_Interp *interp, Tcl_Class clsPtr,
+	    TclOO_PreCallProc *preCallPtr, TclOO_PostCallProc *postCallPtr,
+	    ProcErrorProc *errProc, ClientData clientData, Tcl_Obj *nameObj,
+	    Tcl_Obj *argsObj, Tcl_Obj *bodyObj, int flags,
+	    void **internalTokenPtr)
+}
+declare 11 {
+    int TclOOInvokeObject(Tcl_Interp *interp, Tcl_Object object,
+	    Tcl_Class startCls, int publicPrivate, int objc,
+	    Tcl_Obj *const *objv)
+}
+declare 12 {
+    void TclOOObjectSetFilters(Object *oPtr, int numFilters,
+	    Tcl_Obj *const *filters)
+}
+declare 13 {
+    void TclOOClassSetFilters(Tcl_Interp *interp, Class *classPtr,
+	    int numFilters, Tcl_Obj *const *filters)
+}
+declare 14 {
+    void TclOOObjectSetMixins(Object *oPtr, int numMixins,
+	    Class *const *mixins)
+}
+declare 15 {
+    void TclOOClassSetMixins(Tcl_Interp *interp, Class *classPtr,
+	    int numMixins, Class *const *mixins)
+}
+
+return
+
+# Local Variables:
+# mode: tcl
+# End:
diff --git a/generic/tclOO.h b/generic/tclOO.h
new file mode 100644
index 0000000..d051e79
--- /dev/null
+++ b/generic/tclOO.h
@@ -0,0 +1,147 @@
+/*
+ * tclOO.h --
+ *
+ *	This file contains the public API definitions and some of the function
+ *	declarations for the object-system (NB: not Tcl_Obj, but ::oo).
+ *
+ * Copyright (c) 2006-2010 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef TCLOO_H_INCLUDED
+#define TCLOO_H_INCLUDED
+
+/*
+ * Be careful when it comes to versioning; need to make sure that the
+ * standalone TclOO version matches. Also make sure that this matches the
+ * version in the files:
+ *
+ * tests/oo.test
+ * tests/ooNext2.test
+ * unix/tclooConfig.sh
+ * win/tclooConfig.sh
+ */
+
+#define TCLOO_VERSION "1.2.0"
+#define TCLOO_PATCHLEVEL TCLOO_VERSION
+
+#include "tcl.h"
+
+/*
+ * For C++ compilers, use extern "C"
+ */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+extern const char *TclOOInitializeStubs(
+	Tcl_Interp *, const char *version);
+#define Tcl_OOInitStubs(interp) \
+    TclOOInitializeStubs((interp), TCLOO_VERSION)
+#ifndef USE_TCL_STUBS
+#   define TclOOInitializeStubs(interp, version) (TCLOO_PATCHLEVEL)
+#endif
+
+/*
+ * These are opaque types.
+ */
+
+typedef struct Tcl_Class_ *Tcl_Class;
+typedef struct Tcl_Method_ *Tcl_Method;
+typedef struct Tcl_Object_ *Tcl_Object;
+typedef struct Tcl_ObjectContext_ *Tcl_ObjectContext;
+
+/*
+ * Public datatypes for callbacks and structures used in the TIP#257 (OO)
+ * implementation. These are used to implement custom types of method calls
+ * and to allow the attachment of arbitrary data to objects and classes.
+ */
+
+typedef int (Tcl_MethodCallProc)(ClientData clientData, Tcl_Interp *interp,
+	Tcl_ObjectContext objectContext, int objc, Tcl_Obj *const *objv);
+typedef void (Tcl_MethodDeleteProc)(ClientData clientData);
+typedef int (Tcl_CloneProc)(Tcl_Interp *interp, ClientData oldClientData,
+	ClientData *newClientData);
+typedef void (Tcl_ObjectMetadataDeleteProc)(ClientData clientData);
+typedef int (Tcl_ObjectMapMethodNameProc)(Tcl_Interp *interp,
+	Tcl_Object object, Tcl_Class *startClsPtr, Tcl_Obj *methodNameObj);
+
+/*
+ * The type of a method implementation. This describes how to call the method
+ * implementation, how to delete it (when the object or class is deleted) and
+ * how to create a clone of it (when the object or class is copied).
+ */
+
+typedef struct {
+    int version;		/* Structure version field. Always to be equal
+				 * to TCL_OO_METHOD_VERSION_CURRENT in
+				 * declarations. */
+    const char *name;		/* Name of this type of method, mostly for
+				 * debugging purposes. */
+    Tcl_MethodCallProc *callProc;
+				/* How to invoke this method. */
+    Tcl_MethodDeleteProc *deleteProc;
+				/* How to delete this method's type-specific
+				 * data, or NULL if the type-specific data
+				 * does not need deleting. */
+    Tcl_CloneProc *cloneProc;	/* How to copy this method's type-specific
+				 * data, or NULL if the type-specific data can
+				 * be copied directly. */
+} Tcl_MethodType;
+
+/*
+ * The correct value for the version field of the Tcl_MethodType structure.
+ * This allows new versions of the structure to be introduced without breaking
+ * binary compatability.
+ */
+
+#define TCL_OO_METHOD_VERSION_CURRENT 1
+
+/*
+ * The type of some object (or class) metadata. This describes how to delete
+ * the metadata (when the object or class is deleted) and how to create a
+ * clone of it (when the object or class is copied).
+ */
+
+typedef struct {
+    int version;		/* Structure version field. Always to be equal
+				 * to TCL_OO_METADATA_VERSION_CURRENT in
+				 * declarations. */
+    const char *name;
+    Tcl_ObjectMetadataDeleteProc *deleteProc;
+				/* How to delete the metadata. This must not
+				 * be NULL. */
+    Tcl_CloneProc *cloneProc;	/* How to copy the metadata, or NULL if the
+				 * type-specific data can be copied
+				 * directly. */
+} Tcl_ObjectMetadataType;
+
+/*
+ * The correct value for the version field of the Tcl_ObjectMetadataType
+ * structure. This allows new versions of the structure to be introduced
+ * without breaking binary compatability.
+ */
+
+#define TCL_OO_METADATA_VERSION_CURRENT 1
+
+/*
+ * Include all the public API, generated from tclOO.decls.
+ */
+
+#include "tclOODecls.h"
+
+#ifdef __cplusplus
+}
+#endif
+#endif
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOBasic.c b/generic/tclOOBasic.c
new file mode 100644
index 0000000..b2c06a7
--- /dev/null
+++ b/generic/tclOOBasic.c
@@ -0,0 +1,1267 @@
+/*
+ * tclOOBasic.c --
+ *
+ *	This file contains implementations of the "simple" commands and
+ *	methods from the object-system core.
+ *
+ * Copyright (c) 2005-2013 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+static inline Tcl_Object *AddConstructionFinalizer(Tcl_Interp *interp);
+static Tcl_NRPostProc	AfterNRDestructor;
+static Tcl_NRPostProc	DecrRefsPostClassConstructor;
+static Tcl_NRPostProc	FinalizeConstruction;
+static Tcl_NRPostProc	FinalizeEval;
+static Tcl_NRPostProc	NextRestoreFrame;
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddCreateCallback, FinalizeConstruction --
+ *
+ *	Special version of TclNRAddCallback that allows the caller to splice
+ *	the object created later on. Always calls FinalizeConstruction, which
+ *	converts the object into its name and stores that in the interpreter
+ *	result. This is shared by all the construction methods (create,
+ *	createWithNamespace, new).
+ *
+ *	Note that this is the only code in this file (or, indeed, the whole of
+ *	TclOO) that uses NRE internals; it is the only code that does
+ *	non-standard poking in the NRE guts.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline Tcl_Object *
+AddConstructionFinalizer(
+    Tcl_Interp *interp)
+{
+    TclNRAddCallback(interp, FinalizeConstruction, NULL, NULL, NULL, NULL);
+    return (Tcl_Object *) &(TOP_CB(interp)->data[0]);
+}
+
+static int
+FinalizeConstruction(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Object *oPtr = data[0];
+
+    if (result != TCL_OK) {
+	return result;
+    }
+    Tcl_SetObjResult(interp, TclOOObjectName(interp, oPtr));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Class_Constructor --
+ *
+ *	Implementation for oo::class constructor.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Class_Constructor(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) Tcl_ObjectContextObject(context);
+    Tcl_Obj **invoke;
+
+    if (objc-1 > Tcl_ObjectContextSkippedArgs(context)) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"?definitionScript?");
+	return TCL_ERROR;
+    } else if (objc == Tcl_ObjectContextSkippedArgs(context)) {
+	return TCL_OK;
+    }
+
+    /*
+     * Delegate to [oo::define] to do the work.
+     */
+
+    invoke = ckalloc(3 * sizeof(Tcl_Obj *));
+    invoke[0] = oPtr->fPtr->defineName;
+    invoke[1] = TclOOObjectName(interp, oPtr);
+    invoke[2] = objv[objc-1];
+
+    /*
+     * Must add references or errors in configuration script will cause
+     * trouble.
+     */
+
+    Tcl_IncrRefCount(invoke[0]);
+    Tcl_IncrRefCount(invoke[1]);
+    Tcl_IncrRefCount(invoke[2]);
+    TclNRAddCallback(interp, DecrRefsPostClassConstructor,
+	    invoke, NULL, NULL, NULL);
+
+    /*
+     * Tricky point: do not want the extra reported level in the Tcl stack
+     * trace, so use TCL_EVAL_NOERR.
+     */
+
+    return TclNREvalObjv(interp, 3, invoke, TCL_EVAL_NOERR, NULL);
+}
+
+static int
+DecrRefsPostClassConstructor(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj **invoke = data[0];
+
+    TclDecrRefCount(invoke[0]);
+    TclDecrRefCount(invoke[1]);
+    TclDecrRefCount(invoke[2]);
+    ckfree(invoke);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Class_Create --
+ *
+ *	Implementation for oo::class->create method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Class_Create(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Object *oPtr = (Object *) Tcl_ObjectContextObject(context);
+    const char *objName;
+    int len;
+
+    /*
+     * Sanity check; should not be possible to invoke this method on a
+     * non-class.
+     */
+
+    if (oPtr->classPtr == NULL) {
+	Tcl_Obj *cmdnameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"object \"%s\" is not a class", TclGetString(cmdnameObj)));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "INSTANTIATE_NONCLASS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check we have the right number of (sensible) arguments.
+     */
+
+    if (objc - Tcl_ObjectContextSkippedArgs(context) < 1) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"objectName ?arg ...?");
+	return TCL_ERROR;
+    }
+    objName = Tcl_GetStringFromObj(
+	    objv[Tcl_ObjectContextSkippedArgs(context)], &len);
+    if (len == 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"object name must not be empty", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "EMPTY_NAME", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the object and return its name.
+     */
+
+    return TclNRNewObjectInstance(interp, (Tcl_Class) oPtr->classPtr,
+	    objName, NULL, objc, objv,
+	    Tcl_ObjectContextSkippedArgs(context)+1,
+	    AddConstructionFinalizer(interp));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Class_CreateNs --
+ *
+ *	Implementation for oo::class->createWithNamespace method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Class_CreateNs(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Object *oPtr = (Object *) Tcl_ObjectContextObject(context);
+    const char *objName, *nsName;
+    int len;
+
+    /*
+     * Sanity check; should not be possible to invoke this method on a
+     * non-class.
+     */
+
+    if (oPtr->classPtr == NULL) {
+	Tcl_Obj *cmdnameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"object \"%s\" is not a class", TclGetString(cmdnameObj)));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "INSTANTIATE_NONCLASS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check we have the right number of (sensible) arguments.
+     */
+
+    if (objc - Tcl_ObjectContextSkippedArgs(context) < 2) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"objectName namespaceName ?arg ...?");
+	return TCL_ERROR;
+    }
+    objName = Tcl_GetStringFromObj(
+	    objv[Tcl_ObjectContextSkippedArgs(context)], &len);
+    if (len == 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"object name must not be empty", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "EMPTY_NAME", NULL);
+	return TCL_ERROR;
+    }
+    nsName = Tcl_GetStringFromObj(
+	    objv[Tcl_ObjectContextSkippedArgs(context)+1], &len);
+    if (len == 0) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"namespace name must not be empty", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "EMPTY_NAME", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the object and return its name.
+     */
+
+    return TclNRNewObjectInstance(interp, (Tcl_Class) oPtr->classPtr,
+	    objName, nsName, objc, objv,
+	    Tcl_ObjectContextSkippedArgs(context)+2,
+	    AddConstructionFinalizer(interp));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Class_New --
+ *
+ *	Implementation for oo::class->new method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Class_New(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Object *oPtr = (Object *) Tcl_ObjectContextObject(context);
+
+    /*
+     * Sanity check; should not be possible to invoke this method on a
+     * non-class.
+     */
+
+    if (oPtr->classPtr == NULL) {
+	Tcl_Obj *cmdnameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"object \"%s\" is not a class", TclGetString(cmdnameObj)));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "INSTANTIATE_NONCLASS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the object and return its name.
+     */
+
+    return TclNRNewObjectInstance(interp, (Tcl_Class) oPtr->classPtr,
+	    NULL, NULL, objc, objv, Tcl_ObjectContextSkippedArgs(context),
+	    AddConstructionFinalizer(interp));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Object_Destroy --
+ *
+ *	Implementation for oo::object->destroy method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Object_Destroy(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Object *oPtr = (Object *) Tcl_ObjectContextObject(context);
+    CallContext *contextPtr;
+
+    if (objc != Tcl_ObjectContextSkippedArgs(context)) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    }
+    if (!(oPtr->flags & DESTRUCTOR_CALLED)) {
+	oPtr->flags |= DESTRUCTOR_CALLED;
+	contextPtr = TclOOGetCallContext(oPtr, NULL, DESTRUCTOR, NULL);
+	if (contextPtr != NULL) {
+	    contextPtr->callPtr->flags |= DESTRUCTOR;
+	    contextPtr->skip = 0;
+	    TclNRAddCallback(interp, AfterNRDestructor, contextPtr,
+		    NULL, NULL, NULL);
+	    TclPushTailcallPoint(interp);
+	    return TclOOInvokeContext(contextPtr, interp, 0, NULL);
+	}
+    }
+    if (oPtr->command) {
+	Tcl_DeleteCommandFromToken(interp, oPtr->command);
+    }
+    return TCL_OK;
+}
+
+static int
+AfterNRDestructor(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+
+    if (contextPtr->oPtr->command) {
+	Tcl_DeleteCommandFromToken(interp, contextPtr->oPtr->command);
+    }
+    TclOODeleteContext(contextPtr);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Object_Eval --
+ *
+ *	Implementation for oo::object->eval method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Object_Eval(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    CallContext *contextPtr = (CallContext *) context;
+    Tcl_Object object = Tcl_ObjectContextObject(context);
+    register const int skip = Tcl_ObjectContextSkippedArgs(context);
+    CallFrame *framePtr, **framePtrPtr = &framePtr;
+    Tcl_Obj *scriptPtr;
+    CmdFrame *invoker;
+
+    if (objc-1 < skip) {
+	Tcl_WrongNumArgs(interp, skip, objv, "arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the object's namespace the current namespace and evaluate the
+     * command(s).
+     */
+
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    Tcl_GetObjectNamespace(object), 0);
+    framePtr->objc = objc;
+    framePtr->objv = objv;	/* Reference counts do not need to be
+				 * incremented here. */
+
+    if (!(contextPtr->callPtr->flags & PUBLIC_METHOD)) {
+	object = NULL;		/* Now just for error mesage printing. */
+    }
+
+    /*
+     * Work out what script we are actually going to evaluate.
+     *
+     * When there's more than one argument, we concatenate them together with
+     * spaces between, then evaluate the result. Tcl_EvalObjEx will delete the
+     * object when it decrements its refcount after eval'ing it.
+     */
+
+    if (objc != skip+1) {
+	scriptPtr = Tcl_ConcatObj(objc-skip, objv+skip);
+	invoker = NULL;
+    } else {
+	scriptPtr = objv[skip];
+	invoker = ((Interp *) interp)->cmdFramePtr;
+    }
+
+    /*
+     * Evaluate the script now, with FinalizeEval to do the processing after
+     * the script completes.
+     */
+
+    TclNRAddCallback(interp, FinalizeEval, object, NULL, NULL, NULL);
+    return TclNREvalObjEx(interp, scriptPtr, 0, invoker, skip);
+}
+
+static int
+FinalizeEval(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    if (result == TCL_ERROR) {
+	Object *oPtr = data[0];
+	const char *namePtr;
+
+	if (oPtr) {
+	    namePtr = TclGetString(TclOOObjectName(interp, oPtr));
+	} else {
+	    namePtr = "my";
+	}
+
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (in \"%s eval\" script line %d)",
+		namePtr, Tcl_GetErrorLine(interp)));
+    }
+
+    /*
+     * Restore the previous "current" namespace.
+     */
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Object_Unknown --
+ *
+ *	Default unknown method handler method (defined in oo::object). This
+ *	just creates a suitable error message.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Object_Unknown(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    CallContext *contextPtr = (CallContext *) context;
+    Object *oPtr = contextPtr->oPtr;
+    const char **methodNames;
+    int numMethodNames, i, skip = Tcl_ObjectContextSkippedArgs(context);
+    Tcl_Obj *errorMsg;
+
+    /*
+     * If no method name, generate an error asking for a method name. (Only by
+     * overriding *this* method can an object handle the absence of a method
+     * name without an error).
+     */
+
+    if (objc < skip+1) {
+	Tcl_WrongNumArgs(interp, skip, objv, "method ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the list of methods that we want to know about.
+     */
+
+    numMethodNames = TclOOGetSortedMethodList(oPtr,
+	    contextPtr->callPtr->flags & PUBLIC_METHOD, &methodNames);
+
+    /*
+     * Special message when there are no visible methods at all.
+     */
+
+    if (numMethodNames == 0) {
+	Tcl_Obj *tmpBuf = TclOOObjectName(interp, oPtr);
+	const char *piece;
+
+	if (contextPtr->callPtr->flags & PUBLIC_METHOD) {
+	    piece = "visible methods";
+	} else {
+	    piece = "methods";
+	}
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"object \"%s\" has no %s", TclGetString(tmpBuf), piece));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[skip]), NULL);
+	return TCL_ERROR;
+    }
+
+    errorMsg = Tcl_ObjPrintf("unknown method \"%s\": must be ",
+	    TclGetString(objv[skip]));
+    for (i=0 ; i<numMethodNames-1 ; i++) {
+	if (i) {
+	    Tcl_AppendToObj(errorMsg, ", ", -1);
+	}
+	Tcl_AppendToObj(errorMsg, methodNames[i], -1);
+    }
+    if (i) {
+	Tcl_AppendToObj(errorMsg, " or ", -1);
+    }
+    Tcl_AppendToObj(errorMsg, methodNames[i], -1);
+    ckfree(methodNames);
+    Tcl_SetObjResult(interp, errorMsg);
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+	    TclGetString(objv[skip]), NULL);
+    return TCL_ERROR;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Object_LinkVar --
+ *
+ *	Implementation of oo::object->variable method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Object_LinkVar(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Object object = Tcl_ObjectContextObject(context);
+    Namespace *savedNsPtr;
+    int i;
+
+    if (objc-Tcl_ObjectContextSkippedArgs(context) < 0) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"?varName ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * A sanity check. Shouldn't ever happen. (This is all that remains of a
+     * more complex check inherited from [global] after we have applied the
+     * fix for [Bug 2903811]; note that the fix involved *removing* code.)
+     */
+
+    if (iPtr->varFramePtr == NULL) {
+	return TCL_OK;
+    }
+
+    for (i=Tcl_ObjectContextSkippedArgs(context) ; i<objc ; i++) {
+	Var *varPtr, *aryPtr;
+	const char *varName = TclGetString(objv[i]);
+
+	/*
+	 * The variable name must not contain a '::' since that's illegal in
+	 * local names.
+	 */
+
+	if (strstr(varName, "::") != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "variable name \"%s\" illegal: must not contain namespace"
+		    " separator", varName));
+	    Tcl_SetErrorCode(interp, "TCL", "UPVAR", "INVERTED", NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Switch to the object's namespace for the duration of this call.
+	 * Like this, the variable is looked up in the namespace of the
+	 * object, and not in the namespace of the caller. Otherwise this
+	 * would only work if the caller was a method of the object itself,
+	 * which might not be true if the method was exported. This is a bit
+	 * of a hack, but the simplest way to do this (pushing a stack frame
+	 * would be horribly expensive by comparison).
+	 */
+
+	savedNsPtr = iPtr->varFramePtr->nsPtr;
+	iPtr->varFramePtr->nsPtr = (Namespace *)
+		Tcl_GetObjectNamespace(object);
+	varPtr = TclObjLookupVar(interp, objv[i], NULL, TCL_NAMESPACE_ONLY,
+		"define", 1, 0, &aryPtr);
+	iPtr->varFramePtr->nsPtr = savedNsPtr;
+
+	if (varPtr == NULL || aryPtr != NULL) {
+	    /*
+	     * Variable cannot be an element in an array. If aryPtr is not
+	     * NULL, it is an element, so throw up an error and return.
+	     */
+
+	    TclVarErrMsg(interp, varName, NULL, "define",
+		    "name refers to an element in an array");
+	    Tcl_SetErrorCode(interp, "TCL", "UPVAR", "LOCAL_ELEMENT", NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Arrange for the lifetime of the variable to be correctly managed.
+	 * This is copied out of Tcl_VariableObjCmd...
+	 */
+
+	if (!TclIsVarNamespaceVar(varPtr)) {
+	    TclSetVarNamespaceVar(varPtr);
+	}
+
+	if (TclPtrMakeUpvar(interp, varPtr, varName, 0, -1) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOO_Object_VarName --
+ *
+ *	Implementation of the oo::object->varname method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOO_Object_VarName(
+    ClientData clientData,	/* Ignored. */
+    Tcl_Interp *interp,		/* Interpreter in which to create the object;
+				 * also used for error reporting. */
+    Tcl_ObjectContext context,	/* The object/call context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* The actual arguments. */
+{
+    Var *varPtr, *aryVar;
+    Tcl_Obj *varNamePtr, *argPtr;
+    const char *arg;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"varName");
+	return TCL_ERROR;
+    }
+    argPtr = objv[objc-1];
+    arg = Tcl_GetString(argPtr);
+
+    /*
+     * Convert the variable name to fully-qualified form if it wasn't already.
+     * This has to be done prior to lookup because we can run into problems
+     * with resolvers otherwise. [Bug 3603695]
+     *
+     * We still need to do the lookup; the variable could be linked to another
+     * variable and we want the target's name.
+     */
+
+    if (arg[0] == ':' && arg[1] == ':') {
+	varNamePtr = argPtr;
+    } else {
+	Tcl_Namespace *namespacePtr =
+		Tcl_GetObjectNamespace(Tcl_ObjectContextObject(context));
+
+	varNamePtr = Tcl_NewStringObj(namespacePtr->fullName, -1);
+	Tcl_AppendToObj(varNamePtr, "::", 2);
+	Tcl_AppendObjToObj(varNamePtr, argPtr);
+    }
+    Tcl_IncrRefCount(varNamePtr);
+    varPtr = TclObjLookupVar(interp, varNamePtr, NULL,
+	    TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG, "refer to", 1, 1, &aryVar);
+    Tcl_DecrRefCount(varNamePtr);
+    if (varPtr == NULL) {
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARIABLE", arg, NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now that we've pinned down what variable we're really talking about
+     * (including traversing variable links), convert back to a name.
+     */
+
+    varNamePtr = Tcl_NewObj();
+    if (aryVar != NULL) {
+	Tcl_HashEntry *hPtr;
+	Tcl_HashSearch search;
+
+	Tcl_GetVariableFullName(interp, (Tcl_Var) aryVar, varNamePtr);
+
+	/*
+	 * WARNING! This code pokes inside the implementation of hash tables!
+	 */
+
+	hPtr = Tcl_FirstHashEntry((Tcl_HashTable *) aryVar->value.tablePtr,
+		&search);
+	while (hPtr != NULL) {
+	    if (varPtr == Tcl_GetHashValue(hPtr)) {
+		Tcl_AppendToObj(varNamePtr, "(", -1);
+		Tcl_AppendObjToObj(varNamePtr, hPtr->key.objPtr);
+		Tcl_AppendToObj(varNamePtr, ")", -1);
+		break;
+	    }
+	    hPtr = Tcl_NextHashEntry(&search);
+	}
+    } else {
+	Tcl_GetVariableFullName(interp, (Tcl_Var) varPtr, varNamePtr);
+    }
+    Tcl_SetObjResult(interp, varNamePtr);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONextObjCmd, TclOONextToObjCmd --
+ *
+ *	Implementation of the [next] and [nextto] commands. Note that these
+ *	commands are only ever to be used inside the body of a procedure-like
+ *	method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOONextObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *framePtr = iPtr->varFramePtr;
+    Tcl_ObjectContext context;
+
+    /*
+     * Start with sanity checks on the calling context to make sure that we
+     * are invoked from a suitable method context. If so, we can safely
+     * retrieve the handle to the object call context.
+     */
+
+    if (framePtr == NULL || !(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s may only be called from inside a method",
+		TclGetString(objv[0])));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	return TCL_ERROR;
+    }
+    context = framePtr->clientData;
+
+    /*
+     * Invoke the (advanced) method call context in the caller context. Note
+     * that this is like [uplevel 1] and not [eval].
+     */
+
+    TclNRAddCallback(interp, NextRestoreFrame, framePtr, NULL,NULL,NULL);
+    iPtr->varFramePtr = framePtr->callerVarPtr;
+    return TclNRObjectContextInvokeNext(interp, context, objc, objv, 1);
+}
+
+int
+TclOONextToObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *framePtr = iPtr->varFramePtr;
+    Class *classPtr;
+    CallContext *contextPtr;
+    int i;
+    Tcl_Object object;
+    const char *methodType;
+
+    /*
+     * Start with sanity checks on the calling context to make sure that we
+     * are invoked from a suitable method context. If so, we can safely
+     * retrieve the handle to the object call context.
+     */
+
+    if (framePtr == NULL || !(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s may only be called from inside a method",
+		TclGetString(objv[0])));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	return TCL_ERROR;
+    }
+    contextPtr = framePtr->clientData;
+
+    /*
+     * Sanity check the arguments; we need the first one to refer to a class.
+     */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "class ?arg...?");
+	return TCL_ERROR;
+    }
+    object = Tcl_GetObjectFromObj(interp, objv[1]);
+    if (object == NULL) {
+	return TCL_ERROR;
+    }
+    classPtr = ((Object *)object)->classPtr;
+    if (classPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" is not a class", TclGetString(objv[1])));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_REQUIRED", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Search for an implementation of a method associated with the current
+     * call on the call chain past the point where we currently are. Do not
+     * allow jumping backwards!
+     */
+
+    for (i=contextPtr->index+1 ; i<contextPtr->callPtr->numChain ; i++) {
+	struct MInvoke *miPtr = contextPtr->callPtr->chain + i;
+
+	if (!miPtr->isFilter && miPtr->mPtr->declaringClassPtr == classPtr) {
+	    /*
+	     * Invoke the (advanced) method call context in the caller
+	     * context. Note that this is like [uplevel 1] and not [eval].
+	     */
+
+	    TclNRAddCallback(interp, NextRestoreFrame, framePtr,
+		    contextPtr, INT2PTR(contextPtr->index), NULL);
+	    contextPtr->index = i-1;
+	    iPtr->varFramePtr = framePtr->callerVarPtr;
+	    return TclNRObjectContextInvokeNext(interp,
+		    (Tcl_ObjectContext) contextPtr, objc, objv, 2);
+	}
+    }
+
+    /*
+     * Generate an appropriate error message, depending on whether the value
+     * is on the chain but unreachable, or not on the chain at all.
+     */
+
+    if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+	methodType = "constructor";
+    } else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+	methodType = "destructor";
+    } else {
+	methodType = "method";
+    }
+
+    for (i=contextPtr->index ; i>=0 ; i--) {
+	struct MInvoke *miPtr = contextPtr->callPtr->chain + i;
+
+	if (!miPtr->isFilter && miPtr->mPtr->declaringClassPtr == classPtr) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "%s implementation by \"%s\" not reachable from here",
+		    methodType, TclGetString(objv[1])));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_NOT_REACHABLE",
+		    NULL);
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "%s has no non-filter implementation by \"%s\"",
+	    methodType, TclGetString(objv[1])));
+    Tcl_SetErrorCode(interp, "TCL", "OO", "CLASS_NOT_THERE", NULL);
+    return TCL_ERROR;
+}
+
+static int
+NextRestoreFrame(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallContext *contextPtr = data[1];
+
+    iPtr->varFramePtr = data[0];
+    if (contextPtr != NULL) {
+	contextPtr->index = PTR2INT(data[2]);
+    }
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOSelfObjCmd --
+ *
+ *	Implementation of the [self] command, which provides introspection of
+ *	the call context.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOSelfObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    static const char *const subcmds[] = {
+	"call", "caller", "class", "filter", "method", "namespace", "next",
+	"object", "target", NULL
+    };
+    enum SelfCmds {
+	SELF_CALL, SELF_CALLER, SELF_CLASS, SELF_FILTER, SELF_METHOD, SELF_NS,
+	SELF_NEXT, SELF_OBJECT, SELF_TARGET
+    };
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *framePtr = iPtr->varFramePtr;
+    CallContext *contextPtr;
+    Tcl_Obj *result[3];
+    int index;
+
+#define CurrentlyInvoked(contextPtr) \
+    ((contextPtr)->callPtr->chain[(contextPtr)->index])
+
+    /*
+     * Start with sanity checks on the calling context and the method context.
+     */
+
+    if (framePtr == NULL || !(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s may only be called from inside a method",
+		TclGetString(objv[0])));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	return TCL_ERROR;
+    }
+
+    contextPtr = framePtr->clientData;
+
+    /*
+     * Now we do "conventional" argument parsing for a while. Note that no
+     * subcommand takes arguments.
+     */
+
+    if (objc > 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "subcommand");
+	return TCL_ERROR;
+    } else if (objc == 1) {
+	index = SELF_OBJECT;
+    } else if (Tcl_GetIndexFromObj(interp, objv[1], subcmds, "subcommand", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum SelfCmds) index) {
+    case SELF_OBJECT:
+	Tcl_SetObjResult(interp, TclOOObjectName(interp, contextPtr->oPtr));
+	return TCL_OK;
+    case SELF_NS:
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		contextPtr->oPtr->namespacePtr->fullName,-1));
+	return TCL_OK;
+    case SELF_CLASS: {
+	Class *clsPtr = CurrentlyInvoked(contextPtr).mPtr->declaringClassPtr;
+
+	if (clsPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "method not defined by a class", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "UNMATCHED_CONTEXT", NULL);
+	    return TCL_ERROR;
+	}
+
+	Tcl_SetObjResult(interp, TclOOObjectName(interp, clsPtr->thisPtr));
+	return TCL_OK;
+    }
+    case SELF_METHOD:
+	if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+	    Tcl_SetObjResult(interp, contextPtr->oPtr->fPtr->constructorName);
+	} else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+	    Tcl_SetObjResult(interp, contextPtr->oPtr->fPtr->destructorName);
+	} else {
+	    Tcl_SetObjResult(interp,
+		    CurrentlyInvoked(contextPtr).mPtr->namePtr);
+	}
+	return TCL_OK;
+    case SELF_FILTER:
+	if (!CurrentlyInvoked(contextPtr).isFilter) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "not inside a filtering context", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "UNMATCHED_CONTEXT", NULL);
+	    return TCL_ERROR;
+	} else {
+	    register struct MInvoke *miPtr = &CurrentlyInvoked(contextPtr);
+	    Object *oPtr;
+	    const char *type;
+
+	    if (miPtr->filterDeclarer != NULL) {
+		oPtr = miPtr->filterDeclarer->thisPtr;
+		type = "class";
+	    } else {
+		oPtr = contextPtr->oPtr;
+		type = "object";
+	    }
+
+	    result[0] = TclOOObjectName(interp, oPtr);
+	    result[1] = Tcl_NewStringObj(type, -1);
+	    result[2] = miPtr->mPtr->namePtr;
+	    Tcl_SetObjResult(interp, Tcl_NewListObj(3, result));
+	    return TCL_OK;
+	}
+    case SELF_CALLER:
+	if ((framePtr->callerVarPtr == NULL) ||
+		!(framePtr->callerVarPtr->isProcCallFrame & FRAME_IS_METHOD)){
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "caller is not an object", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "CONTEXT_REQUIRED", NULL);
+	    return TCL_ERROR;
+	} else {
+	    CallContext *callerPtr = framePtr->callerVarPtr->clientData;
+	    Method *mPtr = callerPtr->callPtr->chain[callerPtr->index].mPtr;
+	    Object *declarerPtr;
+
+	    if (mPtr->declaringClassPtr != NULL) {
+		declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	    } else if (mPtr->declaringObjectPtr != NULL) {
+		declarerPtr = mPtr->declaringObjectPtr;
+	    } else {
+		/*
+		 * This should be unreachable code.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"method without declarer!", -1));
+		return TCL_ERROR;
+	    }
+
+	    result[0] = TclOOObjectName(interp, declarerPtr);
+	    result[1] = TclOOObjectName(interp, callerPtr->oPtr);
+	    if (callerPtr->callPtr->flags & CONSTRUCTOR) {
+		result[2] = declarerPtr->fPtr->constructorName;
+	    } else if (callerPtr->callPtr->flags & DESTRUCTOR) {
+		result[2] = declarerPtr->fPtr->destructorName;
+	    } else {
+		result[2] = mPtr->namePtr;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewListObj(3, result));
+	    return TCL_OK;
+	}
+    case SELF_NEXT:
+	if (contextPtr->index < contextPtr->callPtr->numChain-1) {
+	    Method *mPtr =
+		    contextPtr->callPtr->chain[contextPtr->index+1].mPtr;
+	    Object *declarerPtr;
+
+	    if (mPtr->declaringClassPtr != NULL) {
+		declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	    } else if (mPtr->declaringObjectPtr != NULL) {
+		declarerPtr = mPtr->declaringObjectPtr;
+	    } else {
+		/*
+		 * This should be unreachable code.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"method without declarer!", -1));
+		return TCL_ERROR;
+	    }
+
+	    result[0] = TclOOObjectName(interp, declarerPtr);
+	    if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+		result[1] = declarerPtr->fPtr->constructorName;
+	    } else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+		result[1] = declarerPtr->fPtr->destructorName;
+	    } else {
+		result[1] = mPtr->namePtr;
+	    }
+	    Tcl_SetObjResult(interp, Tcl_NewListObj(2, result));
+	}
+	return TCL_OK;
+    case SELF_TARGET:
+	if (!CurrentlyInvoked(contextPtr).isFilter) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "not inside a filtering context", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "UNMATCHED_CONTEXT", NULL);
+	    return TCL_ERROR;
+	} else {
+	    Method *mPtr;
+	    Object *declarerPtr;
+	    int i;
+
+	    for (i=contextPtr->index ; i<contextPtr->callPtr->numChain ; i++){
+		if (!contextPtr->callPtr->chain[i].isFilter) {
+		    break;
+		}
+	    }
+	    if (i == contextPtr->callPtr->numChain) {
+		Tcl_Panic("filtering call chain without terminal non-filter");
+	    }
+	    mPtr = contextPtr->callPtr->chain[i].mPtr;
+	    if (mPtr->declaringClassPtr != NULL) {
+		declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	    } else if (mPtr->declaringObjectPtr != NULL) {
+		declarerPtr = mPtr->declaringObjectPtr;
+	    } else {
+		/*
+		 * This should be unreachable code.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"method without declarer!", -1));
+		return TCL_ERROR;
+	    }
+	    result[0] = TclOOObjectName(interp, declarerPtr);
+	    result[1] = mPtr->namePtr;
+	    Tcl_SetObjResult(interp, Tcl_NewListObj(2, result));
+	    return TCL_OK;
+	}
+    case SELF_CALL:
+	result[0] = TclOORenderCallChain(interp, contextPtr->callPtr);
+	result[1] = Tcl_NewIntObj(contextPtr->index);
+	Tcl_SetObjResult(interp, Tcl_NewListObj(2, result));
+	return TCL_OK;
+    }
+    return TCL_ERROR;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * CopyObjectCmd --
+ *
+ *	Implementation of the [oo::copy] command, which clones an object (but
+ *	not its namespace). Note that no constructors are called during this
+ *	process.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOCopyObjectCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Object oPtr, o2Ptr;
+
+    if (objc < 2 || objc > 4) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+			 "sourceName ?targetName? ?targetNamespace?");
+	return TCL_ERROR;
+    }
+
+    oPtr = Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create a cloned object of the correct class. Note that constructors are
+     * not called. Also note that we must resolve the object name ourselves
+     * because we do not want to create the object in the current namespace,
+     * but rather in the context of the namespace of the caller of the overall
+     * [oo::define] command.
+     */
+
+    if (objc == 2) {
+	o2Ptr = Tcl_CopyObjectInstance(interp, oPtr, NULL, NULL);
+    } else {
+	const char *name, *namespaceName;
+	Tcl_DString buffer;
+
+	name = TclGetString(objv[2]);
+	Tcl_DStringInit(&buffer);
+	if (name[0] == '\0') {
+	    name = NULL;
+	} else if (name[0]!=':' || name[1]!=':') {
+	    Interp *iPtr = (Interp *) interp;
+
+	    if (iPtr->varFramePtr != NULL) {
+		Tcl_DStringAppend(&buffer,
+			iPtr->varFramePtr->nsPtr->fullName, -1);
+	    }
+	    TclDStringAppendLiteral(&buffer, "::");
+	    Tcl_DStringAppend(&buffer, name, -1);
+	    name = Tcl_DStringValue(&buffer);
+	}
+
+	/*
+	 * Choose a unique namespace name if the user didn't supply one.
+	 */
+
+	namespaceName = NULL;
+	if (objc == 4) {
+	    namespaceName = TclGetString(objv[3]);
+
+	    if (namespaceName[0] == '\0') {
+		namespaceName = NULL;
+	    } else if (Tcl_FindNamespace(interp, namespaceName, NULL,
+		    0) != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"%s refers to an existing namespace", namespaceName));
+		return TCL_ERROR;
+	    }
+	}
+
+	o2Ptr = Tcl_CopyObjectInstance(interp, oPtr, name, namespaceName);
+	Tcl_DStringFree(&buffer);
+    }
+
+    if (o2Ptr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Return the name of the cloned object.
+     */
+
+    Tcl_SetObjResult(interp, TclOOObjectName(interp, (Object *) o2Ptr));
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOCall.c b/generic/tclOOCall.c
new file mode 100644
index 0000000..ac0b94d
--- /dev/null
+++ b/generic/tclOOCall.c
@@ -0,0 +1,1539 @@
+/*
+ * tclOOCall.c --
+ *
+ *	This file contains the method call chain management code for the
+ *	object-system core.
+ *
+ * Copyright (c) 2005-2012 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+/*
+ * Structure containing a CallContext and any other values needed only during
+ * the construction of the CallContext.
+ */
+
+struct ChainBuilder {
+    CallChain *callChainPtr;	/* The call chain being built. */
+    int filterLength;		/* Number of entries in the call chain that
+				 * are due to processing filters and not the
+				 * main call chain. */
+    Object *oPtr;		/* The object that we are building the chain
+				 * for. */
+};
+
+/*
+ * Extra flags used for call chain management.
+ */
+
+#define DEFINITE_PROTECTED 0x100000
+#define DEFINITE_PUBLIC    0x200000
+#define KNOWN_STATE	   (DEFINITE_PROTECTED | DEFINITE_PUBLIC)
+#define SPECIAL		   (CONSTRUCTOR | DESTRUCTOR | FORCE_UNKNOWN)
+#define BUILDING_MIXINS	   0x400000
+#define TRAVERSED_MIXIN	   0x800000
+#define OBJECT_MIXIN	   0x1000000
+#define MIXIN_CONSISTENT(flags) \
+    (((flags) & OBJECT_MIXIN) ||					\
+	!((flags) & BUILDING_MIXINS) == !((flags) & TRAVERSED_MIXIN))
+
+/*
+ * Function declarations for things defined in this file.
+ */
+
+static void		AddClassFiltersToCallContext(Object *const oPtr,
+			    Class *clsPtr, struct ChainBuilder *const cbPtr,
+			    Tcl_HashTable *const doneFilters, int flags);
+static void		AddClassMethodNames(Class *clsPtr, const int flags,
+			    Tcl_HashTable *const namesPtr);
+static inline void	AddMethodToCallChain(Method *const mPtr,
+			    struct ChainBuilder *const cbPtr,
+			    Tcl_HashTable *const doneFilters,
+			    Class *const filterDecl, int flags);
+static inline void	AddSimpleChainToCallContext(Object *const oPtr,
+			    Tcl_Obj *const methodNameObj,
+			    struct ChainBuilder *const cbPtr,
+			    Tcl_HashTable *const doneFilters, int flags,
+			    Class *const filterDecl);
+static void		AddSimpleClassChainToCallContext(Class *classPtr,
+			    Tcl_Obj *const methodNameObj,
+			    struct ChainBuilder *const cbPtr,
+			    Tcl_HashTable *const doneFilters, int flags,
+			    Class *const filterDecl);
+static int		CmpStr(const void *ptr1, const void *ptr2);
+static void		DupMethodNameRep(Tcl_Obj *srcPtr, Tcl_Obj *dstPtr);
+static Tcl_NRPostProc	FinalizeMethodRefs;
+static void		FreeMethodNameRep(Tcl_Obj *objPtr);
+static inline int	IsStillValid(CallChain *callPtr, Object *oPtr,
+			    int flags, int reuseMask);
+static Tcl_NRPostProc	ResetFilterFlags;
+static Tcl_NRPostProc	SetFilterFlags;
+static inline void	StashCallChain(Tcl_Obj *objPtr, CallChain *callPtr);
+
+/*
+ * Object type used to manage type caches attached to method names.
+ */
+
+static const Tcl_ObjType methodNameType = {
+    "TclOO method name",
+    FreeMethodNameRep,
+    DupMethodNameRep,
+    NULL,
+    NULL
+};
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODeleteContext --
+ *
+ *	Destroys a method call-chain context, which should not be in use.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOODeleteContext(
+    CallContext *contextPtr)
+{
+    register Object *oPtr = contextPtr->oPtr;
+
+    TclOODeleteChain(contextPtr->callPtr);
+    if (oPtr != NULL) {
+	TclStackFree(oPtr->fPtr->interp, contextPtr);
+	DelRef(oPtr);
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODeleteChainCache --
+ *
+ *	Destroy the cache of method call-chains.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOODeleteChainCache(
+    Tcl_HashTable *tablePtr)
+{
+    FOREACH_HASH_DECLS;
+    CallChain *callPtr;
+
+    FOREACH_HASH_VALUE(callPtr, tablePtr) {
+	if (callPtr) {
+	    TclOODeleteChain(callPtr);
+	}
+    }
+    Tcl_DeleteHashTable(tablePtr);
+    ckfree(tablePtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODeleteChain --
+ *
+ *	Destroys a method call-chain.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOODeleteChain(
+    CallChain *callPtr)
+{
+    if (callPtr == NULL || callPtr->refCount-- > 1) {
+	return;
+    }
+    if (callPtr->chain != callPtr->staticChain) {
+	ckfree(callPtr->chain);
+    }
+    ckfree(callPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOStashContext --
+ *
+ *	Saves a reference to a method call context in a Tcl_Obj's internal
+ *	representation.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+StashCallChain(
+    Tcl_Obj *objPtr,
+    CallChain *callPtr)
+{
+    callPtr->refCount++;
+    TclGetString(objPtr);
+    TclFreeIntRep(objPtr);
+    objPtr->typePtr = &methodNameType;
+    objPtr->internalRep.twoPtrValue.ptr1 = callPtr;
+}
+
+void
+TclOOStashContext(
+    Tcl_Obj *objPtr,
+    CallContext *contextPtr)
+{
+    StashCallChain(objPtr, contextPtr->callPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * DupMethodNameRep, FreeMethodNameRep --
+ *
+ *	Functions to implement the required parts of the Tcl_Obj guts needed
+ *	for caching of method contexts in Tcl_Objs.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+DupMethodNameRep(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *dstPtr)
+{
+    register CallChain *callPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    dstPtr->typePtr = &methodNameType;
+    dstPtr->internalRep.twoPtrValue.ptr1 = callPtr;
+    callPtr->refCount++;
+}
+
+static void
+FreeMethodNameRep(
+    Tcl_Obj *objPtr)
+{
+    register CallChain *callPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    TclOODeleteChain(callPtr);
+    objPtr->typePtr = NULL;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOInvokeContext --
+ *
+ *	Invokes a single step along a method call-chain context. Note that the
+ *	invokation of a step along the chain can cause further steps along the
+ *	chain to be invoked. Note that this function is written to be as light
+ *	in stack usage as possible.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOInvokeContext(
+    ClientData clientData,	/* The method call context. */
+    Tcl_Interp *interp,		/* Interpreter for error reporting, and many
+				 * other sorts of context handling (e.g.,
+				 * commands, variables) depending on method
+				 * implementation. */
+    int objc,			/* The number of arguments. */
+    Tcl_Obj *const objv[])	/* The arguments as actually seen. */
+{
+    register CallContext *const contextPtr = clientData;
+    Method *const mPtr = contextPtr->callPtr->chain[contextPtr->index].mPtr;
+    const int isFilter =
+	    contextPtr->callPtr->chain[contextPtr->index].isFilter;
+
+    /*
+     * If this is the first step along the chain, we preserve the method
+     * entries in the chain so that they do not get deleted out from under our
+     * feet.
+     */
+
+    if (contextPtr->index == 0) {
+	int i;
+
+	for (i=0 ; i<contextPtr->callPtr->numChain ; i++) {
+	    AddRef(contextPtr->callPtr->chain[i].mPtr);
+	}
+
+	/*
+	 * Ensure that the method name itself is part of the arguments when
+	 * we're doing unknown processing.
+	 */
+
+	if (contextPtr->callPtr->flags & OO_UNKNOWN_METHOD) {
+	    contextPtr->skip--;
+	}
+
+	/*
+	 * Add a callback to ensure that method references are dropped once
+	 * this call is finished.
+	 */
+
+	TclNRAddCallback(interp, FinalizeMethodRefs, contextPtr, NULL, NULL,
+		NULL);
+    }
+
+    /*
+     * Save whether we were in a filter and set up whether we are now.
+     */
+
+    if (contextPtr->oPtr->flags & FILTER_HANDLING) {
+	TclNRAddCallback(interp, SetFilterFlags, contextPtr, NULL,NULL,NULL);
+    } else {
+	TclNRAddCallback(interp, ResetFilterFlags,contextPtr,NULL,NULL,NULL);
+    }
+    if (isFilter || contextPtr->callPtr->flags & FILTER_HANDLING) {
+	contextPtr->oPtr->flags |= FILTER_HANDLING;
+    } else {
+	contextPtr->oPtr->flags &= ~FILTER_HANDLING;
+    }
+
+    /*
+     * Run the method implementation.
+     */
+
+    return mPtr->typePtr->callProc(mPtr->clientData, interp,
+	    (Tcl_ObjectContext) contextPtr, objc, objv);
+}
+
+static int
+SetFilterFlags(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+
+    contextPtr->oPtr->flags |= FILTER_HANDLING;
+    return result;
+}
+
+static int
+ResetFilterFlags(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+
+    contextPtr->oPtr->flags &= ~FILTER_HANDLING;
+    return result;
+}
+
+static int
+FinalizeMethodRefs(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallContext *contextPtr = data[0];
+    int i;
+
+    for (i=0 ; i<contextPtr->callPtr->numChain ; i++) {
+	TclOODelMethodRef(contextPtr->callPtr->chain[i].mPtr);
+    }
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetSortedMethodList, TclOOGetSortedClassMethodList --
+ *
+ *	Discovers the list of method names supported by an object or class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOGetSortedMethodList(
+    Object *oPtr,		/* The object to get the method names for. */
+    int flags,			/* Whether we just want the public method
+				 * names. */
+    const char ***stringsPtr)	/* Where to write a pointer to the array of
+				 * strings to. */
+{
+    Tcl_HashTable names;	/* Tcl_Obj* method name to "wanted in list"
+				 * mapping. */
+    FOREACH_HASH_DECLS;
+    int i;
+    Class *mixinPtr;
+    Tcl_Obj *namePtr;
+    Method *mPtr;
+    int isWantedIn;
+    void *isWanted;
+
+    Tcl_InitObjHashTable(&names);
+
+    /*
+     * Name the bits used in the names table values.
+     */
+#define IN_LIST 1
+#define NO_IMPLEMENTATION 2
+
+    /*
+     * Process method names due to the object.
+     */
+
+    if (oPtr->methodsPtr) {
+	FOREACH_HASH(namePtr, mPtr, oPtr->methodsPtr) {
+	    int isNew;
+
+	    if ((mPtr->flags & PRIVATE_METHOD) && !(flags & PRIVATE_METHOD)) {
+		continue;
+	    }
+	    hPtr = Tcl_CreateHashEntry(&names, (char *) namePtr, &isNew);
+	    if (isNew) {
+		isWantedIn = ((!(flags & PUBLIC_METHOD)
+			|| mPtr->flags & PUBLIC_METHOD) ? IN_LIST : 0);
+		isWantedIn |= (mPtr->typePtr == NULL ? NO_IMPLEMENTATION : 0);
+		Tcl_SetHashValue(hPtr, INT2PTR(isWantedIn));
+	    }
+	}
+    }
+
+    /*
+     * Process method names due to private methods on the object's class.
+     */
+
+    if (flags & PRIVATE_METHOD) {
+	FOREACH_HASH(namePtr, mPtr, &oPtr->selfCls->classMethods) {
+	    if (mPtr->flags & PRIVATE_METHOD) {
+		int isNew;
+
+		hPtr = Tcl_CreateHashEntry(&names, (char *) namePtr, &isNew);
+		if (isNew) {
+		    isWantedIn = IN_LIST;
+		    if (mPtr->typePtr == NULL) {
+			isWantedIn |= NO_IMPLEMENTATION;
+		    }
+		    Tcl_SetHashValue(hPtr, INT2PTR(isWantedIn));
+		} else if (mPtr->typePtr != NULL) {
+		    isWantedIn = PTR2INT(Tcl_GetHashValue(hPtr));
+		    if (isWantedIn & NO_IMPLEMENTATION) {
+			isWantedIn &= ~NO_IMPLEMENTATION;
+			Tcl_SetHashValue(hPtr, INT2PTR(isWantedIn));
+		    }
+		}
+	    }
+	}
+    }
+
+    /*
+     * Process (normal) method names from the class hierarchy and the mixin
+     * hierarchy.
+     */
+
+    AddClassMethodNames(oPtr->selfCls, flags, &names);
+    FOREACH(mixinPtr, oPtr->mixins) {
+	AddClassMethodNames(mixinPtr, flags|TRAVERSED_MIXIN, &names);
+    }
+
+    /*
+     * See how many (visible) method names there are. If none, we do not (and
+     * should not) try to sort the list of them.
+     */
+
+    i = 0;
+    if (names.numEntries != 0) {
+	const char **strings;
+
+	/*
+	 * We need to build the list of methods to sort. We will be using
+	 * qsort() for this, because it is very unlikely that the list will be
+	 * heavily sorted when it is long enough to matter.
+	 */
+
+	strings = ckalloc(sizeof(char *) * names.numEntries);
+	FOREACH_HASH(namePtr, isWanted, &names) {
+	    if (!(flags & PUBLIC_METHOD) || (PTR2INT(isWanted) & IN_LIST)) {
+		if (PTR2INT(isWanted) & NO_IMPLEMENTATION) {
+		    continue;
+		}
+		strings[i++] = TclGetString(namePtr);
+	    }
+	}
+
+	/*
+	 * Note that 'i' may well be less than names.numEntries when we are
+	 * dealing with public method names.
+	 */
+
+	if (i > 0) {
+	    if (i > 1) {
+		qsort((void *) strings, (unsigned) i, sizeof(char *), CmpStr);
+	    }
+	    *stringsPtr = strings;
+	} else {
+	    ckfree(strings);
+	}
+    }
+
+    Tcl_DeleteHashTable(&names);
+    return i;
+}
+
+int
+TclOOGetSortedClassMethodList(
+    Class *clsPtr,		/* The class to get the method names for. */
+    int flags,			/* Whether we just want the public method
+				 * names. */
+    const char ***stringsPtr)	/* Where to write a pointer to the array of
+				 * strings to. */
+{
+    Tcl_HashTable names;	/* Tcl_Obj* method name to "wanted in list"
+				 * mapping. */
+    FOREACH_HASH_DECLS;
+    int i;
+    Tcl_Obj *namePtr;
+    void *isWanted;
+
+    Tcl_InitObjHashTable(&names);
+
+    /*
+     * Process method names from the class hierarchy and the mixin hierarchy.
+     */
+
+    AddClassMethodNames(clsPtr, flags, &names);
+
+    /*
+     * See how many (visible) method names there are. If none, we do not (and
+     * should not) try to sort the list of them.
+     */
+
+    i = 0;
+    if (names.numEntries != 0) {
+	const char **strings;
+
+	/*
+	 * We need to build the list of methods to sort. We will be using
+	 * qsort() for this, because it is very unlikely that the list will be
+	 * heavily sorted when it is long enough to matter.
+	 */
+
+	strings = ckalloc(sizeof(char *) * names.numEntries);
+	FOREACH_HASH(namePtr, isWanted, &names) {
+	    if (!(flags & PUBLIC_METHOD) || (PTR2INT(isWanted) & IN_LIST)) {
+		if (PTR2INT(isWanted) & NO_IMPLEMENTATION) {
+		    continue;
+		}
+		strings[i++] = TclGetString(namePtr);
+	    }
+	}
+
+	/*
+	 * Note that 'i' may well be less than names.numEntries when we are
+	 * dealing with public method names.
+	 */
+
+	if (i > 0) {
+	    if (i > 1) {
+		qsort((void *) strings, (unsigned) i, sizeof(char *), CmpStr);
+	    }
+	    *stringsPtr = strings;
+	} else {
+	    ckfree(strings);
+	}
+    }
+
+    Tcl_DeleteHashTable(&names);
+    return i;
+}
+
+/* Comparator for GetSortedMethodList */
+static int
+CmpStr(
+    const void *ptr1,
+    const void *ptr2)
+{
+    const char **strPtr1 = (const char **) ptr1;
+    const char **strPtr2 = (const char **) ptr2;
+
+    return TclpUtfNcmp2(*strPtr1, *strPtr2, strlen(*strPtr1)+1);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddClassMethodNames --
+ *
+ *	Adds the method names defined by a class (or its superclasses) to the
+ *	collection being built. The collection is built in a hash table to
+ *	ensure that duplicates are excluded. Helper for GetSortedMethodList().
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+AddClassMethodNames(
+    Class *clsPtr,		/* Class to get method names from. */
+    const int flags,		/* Whether we are interested in just the
+				 * public method names. */
+    Tcl_HashTable *const namesPtr)
+				/* Reference to the hash table to put the
+				 * information in. The hash table maps the
+				 * Tcl_Obj * method name to an integral value
+				 * describing whether the method is wanted.
+				 * This ensures that public/private override
+				 * semantics are handled correctly.*/
+{
+    /*
+     * Scope all declarations so that the compiler can stand a good chance of
+     * making the recursive step highly efficient. We also hand-implement the
+     * tail-recursive case using a while loop; C compilers typically cannot do
+     * tail-recursion optimization usefully.
+     */
+
+    if (clsPtr->mixins.num != 0) {
+	Class *mixinPtr;
+	int i;
+
+	/* TODO: Beware of infinite loops! */
+	FOREACH(mixinPtr, clsPtr->mixins) {
+	    AddClassMethodNames(mixinPtr, flags|TRAVERSED_MIXIN, namesPtr);
+	}
+    }
+
+    while (1) {
+	FOREACH_HASH_DECLS;
+	Tcl_Obj *namePtr;
+	Method *mPtr;
+
+	FOREACH_HASH(namePtr, mPtr, &clsPtr->classMethods) {
+	    int isNew;
+
+	    hPtr = Tcl_CreateHashEntry(namesPtr, (char *) namePtr, &isNew);
+	    if (isNew) {
+		int isWanted = (!(flags & PUBLIC_METHOD)
+			|| (mPtr->flags & PUBLIC_METHOD)) ? IN_LIST : 0;
+
+		isWanted |= (mPtr->typePtr == NULL ? NO_IMPLEMENTATION : 0);
+		Tcl_SetHashValue(hPtr, INT2PTR(isWanted));
+	    } else if ((PTR2INT(Tcl_GetHashValue(hPtr)) & NO_IMPLEMENTATION)
+		    && mPtr->typePtr != NULL) {
+		int isWanted = PTR2INT(Tcl_GetHashValue(hPtr));
+
+		isWanted &= ~NO_IMPLEMENTATION;
+		Tcl_SetHashValue(hPtr, INT2PTR(isWanted));
+	    }
+	}
+
+	if (clsPtr->superclasses.num != 1) {
+	    break;
+	}
+	clsPtr = clsPtr->superclasses.list[0];
+    }
+    if (clsPtr->superclasses.num != 0) {
+	Class *superPtr;
+	int i;
+
+	FOREACH(superPtr, clsPtr->superclasses) {
+	    AddClassMethodNames(superPtr, flags, namesPtr);
+	}
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddSimpleChainToCallContext --
+ *
+ *	The core of the call-chain construction engine, this handles calling a
+ *	particular method on a particular object. Note that filters and
+ *	unknown handling are already handled by the logic that uses this
+ *	function.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+AddSimpleChainToCallContext(
+    Object *const oPtr,		/* Object to add call chain entries for. */
+    Tcl_Obj *const methodNameObj,
+				/* Name of method to add the call chain
+				 * entries for. */
+    struct ChainBuilder *const cbPtr,
+				/* Where to add the call chain entries. */
+    Tcl_HashTable *const doneFilters,
+				/* Where to record what call chain entries
+				 * have been processed. */
+    int flags,			/* What sort of call chain are we building. */
+    Class *const filterDecl)	/* The class that declared the filter. If
+				 * NULL, either the filter was declared by the
+				 * object or this isn't a filter. */
+{
+    int i;
+
+    if (!(flags & (KNOWN_STATE | SPECIAL)) && oPtr->methodsPtr) {
+	Tcl_HashEntry *hPtr = Tcl_FindHashEntry(oPtr->methodsPtr,
+		(char *) methodNameObj);
+
+	if (hPtr != NULL) {
+	    Method *mPtr = Tcl_GetHashValue(hPtr);
+
+	    if (flags & PUBLIC_METHOD) {
+		if (!(mPtr->flags & PUBLIC_METHOD)) {
+		    return;
+		} else {
+		    flags |= DEFINITE_PUBLIC;
+		}
+	    } else {
+		flags |= DEFINITE_PROTECTED;
+	    }
+	}
+    }
+    if (!(flags & SPECIAL)) {
+	Tcl_HashEntry *hPtr;
+	Class *mixinPtr;
+
+	FOREACH(mixinPtr, oPtr->mixins) {
+	    AddSimpleClassChainToCallContext(mixinPtr, methodNameObj, cbPtr,
+		    doneFilters, flags|TRAVERSED_MIXIN, filterDecl);
+	}
+	if (oPtr->methodsPtr) {
+	    hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char*) methodNameObj);
+	    if (hPtr != NULL) {
+		AddMethodToCallChain(Tcl_GetHashValue(hPtr), cbPtr,
+			doneFilters, filterDecl, flags);
+	    }
+	}
+    }
+    AddSimpleClassChainToCallContext(oPtr->selfCls, methodNameObj, cbPtr,
+	    doneFilters, flags, filterDecl);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddMethodToCallChain --
+ *
+ *	Utility method that manages the adding of a particular method
+ *	implementation to a call-chain.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+AddMethodToCallChain(
+    Method *const mPtr,		/* Actual method implementation to add to call
+				 * chain (or NULL, a no-op). */
+    struct ChainBuilder *const cbPtr,
+				/* The call chain to add the method
+				 * implementation to. */
+    Tcl_HashTable *const doneFilters,
+				/* Where to record what filters have been
+				 * processed. If NULL, not processing filters.
+				 * Note that this function does not update
+				 * this hashtable. */
+    Class *const filterDecl,	/* The class that declared the filter. If
+				 * NULL, either the filter was declared by the
+				 * object or this isn't a filter. */
+    int flags)			/* Used to check if we're mixin-consistent
+				 * only. Mixin-consistent means that either
+				 * we're looking to add things from a mixin
+				 * and we have passed a mixin, or we're not
+				 * looking to add things from a mixin and have
+				 * not passed a mixin. */
+{
+    register CallChain *callPtr = cbPtr->callChainPtr;
+    int i;
+
+    /*
+     * Return if this is just an entry used to record whether this is a public
+     * method. If so, there's nothing real to call and so nothing to add to
+     * the call chain.
+     *
+     * This is also where we enforce mixin-consistency.
+     */
+
+    if (mPtr == NULL || mPtr->typePtr == NULL || !MIXIN_CONSISTENT(flags)) {
+	return;
+    }
+
+    /*
+     * Enforce real private method handling here. We will skip adding this
+     * method IF
+     *  1) we are not allowing private methods, AND
+     *  2) this is a private method, AND
+     *  3) this is a class method, AND
+     *  4) this method was not declared by the class of the current object.
+     *
+     * This does mean that only classes really handle private methods. This
+     * should be sufficient for [incr Tcl] support though.
+     */
+
+    if (!(callPtr->flags & PRIVATE_METHOD)
+	    && (mPtr->flags & PRIVATE_METHOD)
+	    && (mPtr->declaringClassPtr != NULL)
+	    && (mPtr->declaringClassPtr != cbPtr->oPtr->selfCls)) {
+	return;
+    }
+
+    /*
+     * First test whether the method is already in the call chain. Skip over
+     * any leading filters.
+     */
+
+    for (i=cbPtr->filterLength ; i<callPtr->numChain ; i++) {
+	if (callPtr->chain[i].mPtr == mPtr &&
+		callPtr->chain[i].isFilter == (doneFilters != NULL)) {
+	    /*
+	     * Call chain semantics states that methods come as *late* in the
+	     * call chain as possible. This is done by copying down the
+	     * following methods. Note that this does not change the number of
+	     * method invokations in the call chain; it just rearranges them.
+	     */
+
+	    Class *declCls = callPtr->chain[i].filterDeclarer;
+
+	    for (; i+1<callPtr->numChain ; i++) {
+		callPtr->chain[i] = callPtr->chain[i+1];
+	    }
+	    callPtr->chain[i].mPtr = mPtr;
+	    callPtr->chain[i].isFilter = (doneFilters != NULL);
+	    callPtr->chain[i].filterDeclarer = declCls;
+	    return;
+	}
+    }
+
+    /*
+     * Need to really add the method. This is made a bit more complex by the
+     * fact that we are using some "static" space initially, and only start
+     * realloc-ing if the chain gets long.
+     */
+
+    if (callPtr->numChain == CALL_CHAIN_STATIC_SIZE) {
+	callPtr->chain =
+		ckalloc(sizeof(struct MInvoke) * (callPtr->numChain+1));
+	memcpy(callPtr->chain, callPtr->staticChain,
+		sizeof(struct MInvoke) * callPtr->numChain);
+    } else if (callPtr->numChain > CALL_CHAIN_STATIC_SIZE) {
+	callPtr->chain = ckrealloc(callPtr->chain,
+		sizeof(struct MInvoke) * (callPtr->numChain + 1));
+    }
+    callPtr->chain[i].mPtr = mPtr;
+    callPtr->chain[i].isFilter = (doneFilters != NULL);
+    callPtr->chain[i].filterDeclarer = filterDecl;
+    callPtr->numChain++;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InitCallChain --
+ *	Encoding of the policy of how to set up a call chain. Doesn't populate
+ *	the chain with the method implementation data.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+InitCallChain(
+    CallChain *callPtr,
+    Object *oPtr,
+    int flags)
+{
+    callPtr->flags = flags &
+	    (PUBLIC_METHOD | PRIVATE_METHOD | SPECIAL | FILTER_HANDLING);
+    if (oPtr->flags & USE_CLASS_CACHE) {
+	oPtr = oPtr->selfCls->thisPtr;
+	callPtr->flags |= USE_CLASS_CACHE;
+    }
+    callPtr->epoch = oPtr->fPtr->epoch;
+    callPtr->objectCreationEpoch = oPtr->creationEpoch;
+    callPtr->objectEpoch = oPtr->epoch;
+    callPtr->refCount = 1;
+    callPtr->numChain = 0;
+    callPtr->chain = callPtr->staticChain;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * IsStillValid --
+ *	Calculates whether the given call chain can be used for executing a
+ *	method for the given object. The condition on a chain from a cached
+ *	location being reusable is:
+ *	- Refers to the same object (same creation epoch), and
+ *	- Still across the same class structure (same global epoch), and
+ *	- Still across the same object strucutre (same local epoch), and
+ *	- No public/private/filter magic leakage (same flags, modulo the fact
+ *	  that a public chain will satisfy a non-public call).
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline int
+IsStillValid(
+    CallChain *callPtr,
+    Object *oPtr,
+    int flags,
+    int mask)
+{
+    if ((oPtr->flags & USE_CLASS_CACHE)) {
+	oPtr = oPtr->selfCls->thisPtr;
+	flags |= USE_CLASS_CACHE;
+    }
+    return ((callPtr->objectCreationEpoch == oPtr->creationEpoch)
+	    && (callPtr->epoch == oPtr->fPtr->epoch)
+	    && (callPtr->objectEpoch == oPtr->epoch)
+	    && ((callPtr->flags & mask) == (flags & mask)));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetCallContext --
+ *
+ *	Responsible for constructing the call context, an ordered list of all
+ *	method implementations to be called as part of a method invokation.
+ *	This method is central to the whole operation of the OO system.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+CallContext *
+TclOOGetCallContext(
+    Object *oPtr,		/* The object to get the context for. */
+    Tcl_Obj *methodNameObj,	/* The name of the method to get the context
+				 * for. NULL when getting a constructor or
+				 * destructor chain. */
+    int flags,			/* What sort of context are we looking for.
+				 * Only the bits PUBLIC_METHOD, CONSTRUCTOR,
+				 * PRIVATE_METHOD, DESTRUCTOR and
+				 * FILTER_HANDLING are useful. */
+    Tcl_Obj *cacheInThisObj)	/* What object to cache in, or NULL if it is
+				 * to be in the same object as the
+				 * methodNameObj. */
+{
+    CallContext *contextPtr;
+    CallChain *callPtr;
+    struct ChainBuilder cb;
+    int i, count, doFilters;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashTable doneFilters;
+
+    if (cacheInThisObj == NULL) {
+	cacheInThisObj = methodNameObj;
+    }
+    if (flags&(SPECIAL|FILTER_HANDLING) || (oPtr->flags&FILTER_HANDLING)) {
+	hPtr = NULL;
+	doFilters = 0;
+
+	/*
+	 * Check if we have a cached valid constructor or destructor.
+	 */
+
+	if (flags & CONSTRUCTOR) {
+	    callPtr = oPtr->selfCls->constructorChainPtr;
+	    if ((callPtr != NULL)
+		    && (callPtr->objectEpoch == oPtr->selfCls->thisPtr->epoch)
+		    && (callPtr->epoch == oPtr->fPtr->epoch)) {
+		callPtr->refCount++;
+		goto returnContext;
+	    }
+	} else if (flags & DESTRUCTOR) {
+	    callPtr = oPtr->selfCls->destructorChainPtr;
+	    if ((oPtr->mixins.num == 0) && (callPtr != NULL)
+		    && (callPtr->objectEpoch == oPtr->selfCls->thisPtr->epoch)
+		    && (callPtr->epoch == oPtr->fPtr->epoch)) {
+		callPtr->refCount++;
+		goto returnContext;
+	    }
+	}
+    } else {
+	/*
+	 * Check if we can get the chain out of the Tcl_Obj method name or out
+	 * of the cache. This is made a bit more complex by the fact that
+	 * there are multiple different layers of cache (in the Tcl_Obj, in
+	 * the object, and in the class).
+	 */
+
+	const int reuseMask = ((flags & PUBLIC_METHOD) ? ~0 : ~PUBLIC_METHOD);
+
+	if (cacheInThisObj->typePtr == &methodNameType) {
+	    callPtr = cacheInThisObj->internalRep.twoPtrValue.ptr1;
+	    if (IsStillValid(callPtr, oPtr, flags, reuseMask)) {
+		callPtr->refCount++;
+		goto returnContext;
+	    }
+	    FreeMethodNameRep(cacheInThisObj);
+	}
+
+	if (oPtr->flags & USE_CLASS_CACHE) {
+	    if (oPtr->selfCls->classChainCache != NULL) {
+		hPtr = Tcl_FindHashEntry(oPtr->selfCls->classChainCache,
+			(char *) methodNameObj);
+	    } else {
+		hPtr = NULL;
+	    }
+	} else {
+	    if (oPtr->chainCache != NULL) {
+		hPtr = Tcl_FindHashEntry(oPtr->chainCache,
+			(char *) methodNameObj);
+	    } else {
+		hPtr = NULL;
+	    }
+	}
+
+	if (hPtr != NULL && Tcl_GetHashValue(hPtr) != NULL) {
+	    callPtr = Tcl_GetHashValue(hPtr);
+	    if (IsStillValid(callPtr, oPtr, flags, reuseMask)) {
+		callPtr->refCount++;
+		goto returnContext;
+	    }
+	    Tcl_SetHashValue(hPtr, NULL);
+	    TclOODeleteChain(callPtr);
+	}
+
+	doFilters = 1;
+    }
+
+    callPtr = ckalloc(sizeof(CallChain));
+    InitCallChain(callPtr, oPtr, flags);
+
+    cb.callChainPtr = callPtr;
+    cb.filterLength = 0;
+    cb.oPtr = oPtr;
+
+    /*
+     * If we're working with a forced use of unknown, do that now.
+     */
+
+    if (flags & FORCE_UNKNOWN) {
+	AddSimpleChainToCallContext(oPtr, oPtr->fPtr->unknownMethodNameObj,
+		&cb, NULL, BUILDING_MIXINS, NULL);
+	AddSimpleChainToCallContext(oPtr, oPtr->fPtr->unknownMethodNameObj,
+		&cb, NULL, 0, NULL);
+	callPtr->flags |= OO_UNKNOWN_METHOD;
+	callPtr->epoch = -1;
+	if (callPtr->numChain == 0) {
+	    TclOODeleteChain(callPtr);
+	    return NULL;
+	}
+	goto returnContext;
+    }
+
+    /*
+     * Add all defined filters (if any, and if we're going to be processing
+     * them; they're not processed for constructors, destructors or when we're
+     * in the middle of processing a filter).
+     */
+
+    if (doFilters) {
+	Tcl_Obj *filterObj;
+	Class *mixinPtr;
+
+	doFilters = 1;
+	Tcl_InitObjHashTable(&doneFilters);
+	FOREACH(mixinPtr, oPtr->mixins) {
+	    AddClassFiltersToCallContext(oPtr, mixinPtr, &cb, &doneFilters,
+		    TRAVERSED_MIXIN|BUILDING_MIXINS|OBJECT_MIXIN);
+	    AddClassFiltersToCallContext(oPtr, mixinPtr, &cb, &doneFilters,
+		    OBJECT_MIXIN);
+	}
+	FOREACH(filterObj, oPtr->filters) {
+	    AddSimpleChainToCallContext(oPtr, filterObj, &cb, &doneFilters,
+		    BUILDING_MIXINS, NULL);
+	    AddSimpleChainToCallContext(oPtr, filterObj, &cb, &doneFilters, 0,
+		    NULL);
+	}
+	AddClassFiltersToCallContext(oPtr, oPtr->selfCls, &cb, &doneFilters,
+		BUILDING_MIXINS);
+	AddClassFiltersToCallContext(oPtr, oPtr->selfCls, &cb, &doneFilters,
+		0);
+	Tcl_DeleteHashTable(&doneFilters);
+    }
+    count = cb.filterLength = callPtr->numChain;
+
+    /*
+     * Add the actual method implementations. We have to do this twice to
+     * handle class mixins right.
+     */
+
+    AddSimpleChainToCallContext(oPtr, methodNameObj, &cb, NULL,
+	    flags|BUILDING_MIXINS, NULL);
+    AddSimpleChainToCallContext(oPtr, methodNameObj, &cb, NULL, flags, NULL);
+
+    /*
+     * Check to see if the method has no implementation. If so, we probably
+     * need to add in a call to the unknown method. Otherwise, set up the
+     * cacheing of the method implementation (if relevant).
+     */
+
+    if (count == callPtr->numChain) {
+	/*
+	 * Method does not actually exist. If we're dealing with constructors
+	 * or destructors, this isn't a problem.
+	 */
+
+	if (flags & SPECIAL) {
+	    TclOODeleteChain(callPtr);
+	    return NULL;
+	}
+	AddSimpleChainToCallContext(oPtr, oPtr->fPtr->unknownMethodNameObj,
+		&cb, NULL, BUILDING_MIXINS, NULL);
+	AddSimpleChainToCallContext(oPtr, oPtr->fPtr->unknownMethodNameObj,
+		&cb, NULL, 0, NULL);
+	callPtr->flags |= OO_UNKNOWN_METHOD;
+	callPtr->epoch = -1;
+	if (count == callPtr->numChain) {
+	    TclOODeleteChain(callPtr);
+	    return NULL;
+	}
+    } else if (doFilters) {
+	if (hPtr == NULL) {
+	    if (oPtr->flags & USE_CLASS_CACHE) {
+		if (oPtr->selfCls->classChainCache == NULL) {
+		    oPtr->selfCls->classChainCache =
+			    ckalloc(sizeof(Tcl_HashTable));
+
+		    Tcl_InitObjHashTable(oPtr->selfCls->classChainCache);
+		}
+		hPtr = Tcl_CreateHashEntry(oPtr->selfCls->classChainCache,
+			(char *) methodNameObj, &i);
+	    } else {
+		if (oPtr->chainCache == NULL) {
+		    oPtr->chainCache = ckalloc(sizeof(Tcl_HashTable));
+
+		    Tcl_InitObjHashTable(oPtr->chainCache);
+		}
+		hPtr = Tcl_CreateHashEntry(oPtr->chainCache,
+			(char *) methodNameObj, &i);
+	    }
+	}
+	callPtr->refCount++;
+	Tcl_SetHashValue(hPtr, callPtr);
+	StashCallChain(cacheInThisObj, callPtr);
+    } else if (flags & CONSTRUCTOR) {
+	if (oPtr->selfCls->constructorChainPtr) {
+	    TclOODeleteChain(oPtr->selfCls->constructorChainPtr);
+	}
+	oPtr->selfCls->constructorChainPtr = callPtr;
+	callPtr->refCount++;
+    } else if ((flags & DESTRUCTOR) && oPtr->mixins.num == 0) {
+	if (oPtr->selfCls->destructorChainPtr) {
+	    TclOODeleteChain(oPtr->selfCls->destructorChainPtr);
+	}
+	oPtr->selfCls->destructorChainPtr = callPtr;
+	callPtr->refCount++;
+    }
+
+  returnContext:
+    contextPtr = TclStackAlloc(oPtr->fPtr->interp, sizeof(CallContext));
+    contextPtr->oPtr = oPtr;
+    AddRef(oPtr);
+    contextPtr->callPtr = callPtr;
+    contextPtr->skip = 2;
+    contextPtr->index = 0;
+    return contextPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetStereotypeCallChain --
+ *
+ *	Construct a call-chain for a method that would be used by a
+ *	stereotypical instance of the given class (i.e., where the object has
+ *	no definitions special to itself).
+ *
+ * ----------------------------------------------------------------------
+ */
+
+CallChain *
+TclOOGetStereotypeCallChain(
+    Class *clsPtr,		/* The object to get the context for. */
+    Tcl_Obj *methodNameObj,	/* The name of the method to get the context
+				 * for. NULL when getting a constructor or
+				 * destructor chain. */
+    int flags)			/* What sort of context are we looking for.
+				 * Only the bits PUBLIC_METHOD, CONSTRUCTOR,
+				 * PRIVATE_METHOD, DESTRUCTOR and
+				 * FILTER_HANDLING are useful. */
+{
+    CallChain *callPtr;
+    struct ChainBuilder cb;
+    int i, count;
+    Foundation *fPtr = clsPtr->thisPtr->fPtr;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashTable doneFilters;
+    Object obj;
+
+    /*
+     * Synthesize a temporary stereotypical object so that we can use existing
+     * machinery to produce the stereotypical call chain.
+     */
+
+    memset(&obj, 0, sizeof(Object));
+    obj.fPtr = fPtr;
+    obj.selfCls = clsPtr;
+    obj.refCount = 1;
+    obj.flags = USE_CLASS_CACHE;
+
+    /*
+     * Check if we can get the chain out of the Tcl_Obj method name or out of
+     * the cache. This is made a bit more complex by the fact that there are
+     * multiple different layers of cache (in the Tcl_Obj, in the object, and
+     * in the class).
+     */
+
+    if (clsPtr->classChainCache != NULL) {
+	hPtr = Tcl_FindHashEntry(clsPtr->classChainCache,
+		(char *) methodNameObj);
+	if (hPtr != NULL && Tcl_GetHashValue(hPtr) != NULL) {
+	    const int reuseMask =
+		    ((flags & PUBLIC_METHOD) ? ~0 : ~PUBLIC_METHOD);
+
+	    callPtr = Tcl_GetHashValue(hPtr);
+	    if (IsStillValid(callPtr, &obj, flags, reuseMask)) {
+		callPtr->refCount++;
+		return callPtr;
+	    }
+	    Tcl_SetHashValue(hPtr, NULL);
+	    TclOODeleteChain(callPtr);
+	}
+    } else {
+	hPtr = NULL;
+    }
+
+    callPtr = ckalloc(sizeof(CallChain));
+    memset(callPtr, 0, sizeof(CallChain));
+    callPtr->flags = flags & (PUBLIC_METHOD|PRIVATE_METHOD|FILTER_HANDLING);
+    callPtr->epoch = fPtr->epoch;
+    callPtr->objectCreationEpoch = fPtr->tsdPtr->nsCount;
+    callPtr->objectEpoch = clsPtr->thisPtr->epoch;
+    callPtr->refCount = 1;
+    callPtr->chain = callPtr->staticChain;
+
+    cb.callChainPtr = callPtr;
+    cb.filterLength = 0;
+    cb.oPtr = &obj;
+
+    /*
+     * Add all defined filters (if any, and if we're going to be processing
+     * them; they're not processed for constructors, destructors or when we're
+     * in the middle of processing a filter).
+     */
+
+    Tcl_InitObjHashTable(&doneFilters);
+    AddClassFiltersToCallContext(&obj, clsPtr, &cb, &doneFilters,
+	    BUILDING_MIXINS);
+    AddClassFiltersToCallContext(&obj, clsPtr, &cb, &doneFilters, 0);
+    Tcl_DeleteHashTable(&doneFilters);
+    count = cb.filterLength = callPtr->numChain;
+
+    /*
+     * Add the actual method implementations.
+     */
+
+    AddSimpleChainToCallContext(&obj, methodNameObj, &cb, NULL,
+	    flags|BUILDING_MIXINS, NULL);
+    AddSimpleChainToCallContext(&obj, methodNameObj, &cb, NULL, flags, NULL);
+
+    /*
+     * Check to see if the method has no implementation. If so, we probably
+     * need to add in a call to the unknown method. Otherwise, set up the
+     * cacheing of the method implementation (if relevant).
+     */
+
+    if (count == callPtr->numChain) {
+	AddSimpleChainToCallContext(&obj, fPtr->unknownMethodNameObj, &cb,
+		NULL, BUILDING_MIXINS, NULL);
+	AddSimpleChainToCallContext(&obj, fPtr->unknownMethodNameObj, &cb,
+		NULL, 0, NULL);
+	callPtr->flags |= OO_UNKNOWN_METHOD;
+	callPtr->epoch = -1;
+	if (count == callPtr->numChain) {
+	    TclOODeleteChain(callPtr);
+	    return NULL;
+	}
+    } else {
+	if (hPtr == NULL) {
+	    if (clsPtr->classChainCache == NULL) {
+		clsPtr->classChainCache = ckalloc(sizeof(Tcl_HashTable));
+		Tcl_InitObjHashTable(clsPtr->classChainCache);
+	    }
+	    hPtr = Tcl_CreateHashEntry(clsPtr->classChainCache,
+		    (char *) methodNameObj, &i);
+	}
+	callPtr->refCount++;
+	Tcl_SetHashValue(hPtr, callPtr);
+	StashCallChain(methodNameObj, callPtr);
+    }
+    return callPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddClassFiltersToCallContext --
+ *
+ *	Logic to make extracting all the filters from the class context much
+ *	easier.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+AddClassFiltersToCallContext(
+    Object *const oPtr,		/* Object that the filters operate on. */
+    Class *clsPtr,		/* Class to get the filters from. */
+    struct ChainBuilder *const cbPtr,
+				/* Context to fill with call chain entries. */
+    Tcl_HashTable *const doneFilters,
+				/* Where to record what filters have been
+				 * processed. Keys are objects, values are
+				 * ignored. */
+    int flags)			/* Whether we've gone along a mixin link
+				 * yet. */
+{
+    int i, clearedFlags =
+	    flags & ~(TRAVERSED_MIXIN|OBJECT_MIXIN|BUILDING_MIXINS);
+    Class *superPtr, *mixinPtr;
+    Tcl_Obj *filterObj;
+
+  tailRecurse:
+    if (clsPtr == NULL) {
+	return;
+    }
+
+    /*
+     * Add all the filters defined by classes mixed into the main class
+     * hierarchy.
+     */
+
+    FOREACH(mixinPtr, clsPtr->mixins) {
+	AddClassFiltersToCallContext(oPtr, mixinPtr, cbPtr, doneFilters,
+		flags|TRAVERSED_MIXIN);
+    }
+
+    /*
+     * Add all the class filters from the current class. Note that the filters
+     * are added starting at the object root, as this allows the object to
+     * override how filters work to extend their behaviour.
+     */
+
+    if (MIXIN_CONSISTENT(flags)) {
+	FOREACH(filterObj, clsPtr->filters) {
+	    int isNew;
+
+	    (void) Tcl_CreateHashEntry(doneFilters, (char *) filterObj,
+		    &isNew);
+	    if (isNew) {
+		AddSimpleChainToCallContext(oPtr, filterObj, cbPtr,
+			doneFilters, clearedFlags|BUILDING_MIXINS, clsPtr);
+		AddSimpleChainToCallContext(oPtr, filterObj, cbPtr,
+			doneFilters, clearedFlags, clsPtr);
+	    }
+	}
+    }
+
+    /*
+     * Now process the recursive case. Notice the tail-call optimization.
+     */
+
+    switch (clsPtr->superclasses.num) {
+    case 1:
+	clsPtr = clsPtr->superclasses.list[0];
+	goto tailRecurse;
+    default:
+	FOREACH(superPtr, clsPtr->superclasses) {
+	    AddClassFiltersToCallContext(oPtr, superPtr, cbPtr, doneFilters,
+		    flags);
+	}
+    case 0:
+	return;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AddSimpleClassChainToCallContext --
+ *
+ *	Construct a call-chain from a class hierarchy.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+AddSimpleClassChainToCallContext(
+    Class *classPtr,		/* Class to add the call chain entries for. */
+    Tcl_Obj *const methodNameObj,
+				/* Name of method to add the call chain
+				 * entries for. */
+    struct ChainBuilder *const cbPtr,
+				/* Where to add the call chain entries. */
+    Tcl_HashTable *const doneFilters,
+				/* Where to record what call chain entries
+				 * have been processed. */
+    int flags,			/* What sort of call chain are we building. */
+    Class *const filterDecl)	/* The class that declared the filter. If
+				 * NULL, either the filter was declared by the
+				 * object or this isn't a filter. */
+{
+    int i;
+    Class *superPtr;
+
+    /*
+     * We hard-code the tail-recursive form. It's by far the most common case
+     * *and* it is much more gentle on the stack.
+     *
+     * Note that mixins must be processed before the main class hierarchy.
+     * [Bug 1998221]
+     */
+
+  tailRecurse:
+    FOREACH(superPtr, classPtr->mixins) {
+	AddSimpleClassChainToCallContext(superPtr, methodNameObj, cbPtr,
+		doneFilters, flags|TRAVERSED_MIXIN, filterDecl);
+    }
+
+    if (flags & CONSTRUCTOR) {
+	AddMethodToCallChain(classPtr->constructorPtr, cbPtr, doneFilters,
+		filterDecl, flags);
+
+    } else if (flags & DESTRUCTOR) {
+	AddMethodToCallChain(classPtr->destructorPtr, cbPtr, doneFilters,
+		filterDecl, flags);
+    } else {
+	Tcl_HashEntry *hPtr = Tcl_FindHashEntry(&classPtr->classMethods,
+		(char *) methodNameObj);
+
+	if (hPtr != NULL) {
+	    register Method *mPtr = Tcl_GetHashValue(hPtr);
+
+	    if (!(flags & KNOWN_STATE)) {
+		if (flags & PUBLIC_METHOD) {
+		    if (mPtr->flags & PUBLIC_METHOD) {
+			flags |= DEFINITE_PUBLIC;
+		    } else {
+			return;
+		    }
+		} else {
+		    flags |= DEFINITE_PROTECTED;
+		}
+	    }
+	    AddMethodToCallChain(mPtr, cbPtr, doneFilters, filterDecl, flags);
+	}
+    }
+
+    switch (classPtr->superclasses.num) {
+    case 1:
+	classPtr = classPtr->superclasses.list[0];
+	goto tailRecurse;
+    default:
+	FOREACH(superPtr, classPtr->superclasses) {
+	    AddSimpleClassChainToCallContext(superPtr, methodNameObj, cbPtr,
+		    doneFilters, flags, filterDecl);
+	}
+    case 0:
+	return;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOORenderCallChain --
+ *
+ *	Create a description of a call chain. Used in [info object call],
+ *	[info class call], and [self call].
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclOORenderCallChain(
+    Tcl_Interp *interp,
+    CallChain *callPtr)
+{
+    Tcl_Obj *filterLiteral, *methodLiteral, *objectLiteral;
+    Tcl_Obj *resultObj, *descObjs[4], **objv;
+    Foundation *fPtr = TclOOGetFoundation(interp);
+    int i;
+
+    /*
+     * Allocate the literals (potentially) used in our description.
+     */
+
+    filterLiteral = Tcl_NewStringObj("filter", -1);
+    Tcl_IncrRefCount(filterLiteral);
+    methodLiteral = Tcl_NewStringObj("method", -1);
+    Tcl_IncrRefCount(methodLiteral);
+    objectLiteral = Tcl_NewStringObj("object", -1);
+    Tcl_IncrRefCount(objectLiteral);
+
+    /*
+     * Do the actual construction of the descriptions. They consist of a list
+     * of triples that describe the details of how a method is understood. For
+     * each triple, the first word is the type of invokation ("method" is
+     * normal, "unknown" is special because it adds the method name as an
+     * extra argument when handled by some method types, and "filter" is
+     * special because it's a filter method). The second word is the name of
+     * the method in question (which differs for "unknown" and "filter" types)
+     * and the third word is the full name of the class that declares the
+     * method (or "object" if it is declared on the instance).
+     */
+
+    objv = TclStackAlloc(interp, callPtr->numChain * sizeof(Tcl_Obj *));
+    for (i=0 ; i<callPtr->numChain ; i++) {
+	struct MInvoke *miPtr = &callPtr->chain[i];
+
+	descObjs[0] = miPtr->isFilter
+		? filterLiteral
+		: callPtr->flags & OO_UNKNOWN_METHOD
+			? fPtr->unknownMethodNameObj
+			: methodLiteral;
+	descObjs[1] = callPtr->flags & CONSTRUCTOR
+		? fPtr->constructorName
+		: callPtr->flags & DESTRUCTOR
+			? fPtr->destructorName
+			: miPtr->mPtr->namePtr;
+	descObjs[2] = miPtr->mPtr->declaringClassPtr
+		? Tcl_GetObjectName(interp,
+			(Tcl_Object) miPtr->mPtr->declaringClassPtr->thisPtr)
+		: objectLiteral;
+	descObjs[3] = Tcl_NewStringObj(miPtr->mPtr->typePtr->name, -1);
+
+	objv[i] = Tcl_NewListObj(4, descObjs);
+    }
+
+    /*
+     * Drop the local references to the literals; if they're actually used,
+     * they'll live on the description itself.
+     */
+
+    Tcl_DecrRefCount(filterLiteral);
+    Tcl_DecrRefCount(methodLiteral);
+    Tcl_DecrRefCount(objectLiteral);
+
+    /*
+     * Finish building the description and return it.
+     */
+
+    resultObj = Tcl_NewListObj(callPtr->numChain, objv);
+    TclStackFree(interp, objv);
+    return resultObj;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOODecls.h b/generic/tclOODecls.h
new file mode 100644
index 0000000..9fd62ec
--- /dev/null
+++ b/generic/tclOODecls.h
@@ -0,0 +1,234 @@
+/*
+ * This file is (mostly) automatically generated from tclOO.decls.
+ */
+
+#ifndef _TCLOODECLS
+#define _TCLOODECLS
+
+#ifndef TCLAPI
+#   ifdef BUILD_tcl
+#	define TCLAPI extern DLLEXPORT
+#   else
+#	define TCLAPI extern DLLIMPORT
+#   endif
+#endif
+
+#ifdef USE_TCL_STUBS
+#   undef USE_TCLOO_STUBS
+#   define USE_TCLOO_STUBS
+#endif
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+/* 0 */
+TCLAPI Tcl_Object	Tcl_CopyObjectInstance(Tcl_Interp *interp,
+				Tcl_Object sourceObject,
+				const char *targetName,
+				const char *targetNamespaceName);
+/* 1 */
+TCLAPI Tcl_Object	Tcl_GetClassAsObject(Tcl_Class clazz);
+/* 2 */
+TCLAPI Tcl_Class	Tcl_GetObjectAsClass(Tcl_Object object);
+/* 3 */
+TCLAPI Tcl_Command	Tcl_GetObjectCommand(Tcl_Object object);
+/* 4 */
+TCLAPI Tcl_Object	Tcl_GetObjectFromObj(Tcl_Interp *interp,
+				Tcl_Obj *objPtr);
+/* 5 */
+TCLAPI Tcl_Namespace *	Tcl_GetObjectNamespace(Tcl_Object object);
+/* 6 */
+TCLAPI Tcl_Class	Tcl_MethodDeclarerClass(Tcl_Method method);
+/* 7 */
+TCLAPI Tcl_Object	Tcl_MethodDeclarerObject(Tcl_Method method);
+/* 8 */
+TCLAPI int		Tcl_MethodIsPublic(Tcl_Method method);
+/* 9 */
+TCLAPI int		Tcl_MethodIsType(Tcl_Method method,
+				const Tcl_MethodType *typePtr,
+				ClientData *clientDataPtr);
+/* 10 */
+TCLAPI Tcl_Obj *	Tcl_MethodName(Tcl_Method method);
+/* 11 */
+TCLAPI Tcl_Method	Tcl_NewInstanceMethod(Tcl_Interp *interp,
+				Tcl_Object object, Tcl_Obj *nameObj,
+				int isPublic, const Tcl_MethodType *typePtr,
+				ClientData clientData);
+/* 12 */
+TCLAPI Tcl_Method	Tcl_NewMethod(Tcl_Interp *interp, Tcl_Class cls,
+				Tcl_Obj *nameObj, int isPublic,
+				const Tcl_MethodType *typePtr,
+				ClientData clientData);
+/* 13 */
+TCLAPI Tcl_Object	Tcl_NewObjectInstance(Tcl_Interp *interp,
+				Tcl_Class cls, const char *nameStr,
+				const char *nsNameStr, int objc,
+				Tcl_Obj *const *objv, int skip);
+/* 14 */
+TCLAPI int		Tcl_ObjectDeleted(Tcl_Object object);
+/* 15 */
+TCLAPI int		Tcl_ObjectContextIsFiltering(
+				Tcl_ObjectContext context);
+/* 16 */
+TCLAPI Tcl_Method	Tcl_ObjectContextMethod(Tcl_ObjectContext context);
+/* 17 */
+TCLAPI Tcl_Object	Tcl_ObjectContextObject(Tcl_ObjectContext context);
+/* 18 */
+TCLAPI int		Tcl_ObjectContextSkippedArgs(
+				Tcl_ObjectContext context);
+/* 19 */
+TCLAPI ClientData	Tcl_ClassGetMetadata(Tcl_Class clazz,
+				const Tcl_ObjectMetadataType *typePtr);
+/* 20 */
+TCLAPI void		Tcl_ClassSetMetadata(Tcl_Class clazz,
+				const Tcl_ObjectMetadataType *typePtr,
+				ClientData metadata);
+/* 21 */
+TCLAPI ClientData	Tcl_ObjectGetMetadata(Tcl_Object object,
+				const Tcl_ObjectMetadataType *typePtr);
+/* 22 */
+TCLAPI void		Tcl_ObjectSetMetadata(Tcl_Object object,
+				const Tcl_ObjectMetadataType *typePtr,
+				ClientData metadata);
+/* 23 */
+TCLAPI int		Tcl_ObjectContextInvokeNext(Tcl_Interp *interp,
+				Tcl_ObjectContext context, int objc,
+				Tcl_Obj *const *objv, int skip);
+/* 24 */
+TCLAPI Tcl_ObjectMapMethodNameProc * Tcl_ObjectGetMethodNameMapper(
+				Tcl_Object object);
+/* 25 */
+TCLAPI void		Tcl_ObjectSetMethodNameMapper(Tcl_Object object,
+				Tcl_ObjectMapMethodNameProc *mapMethodNameProc);
+/* 26 */
+TCLAPI void		Tcl_ClassSetConstructor(Tcl_Interp *interp,
+				Tcl_Class clazz, Tcl_Method method);
+/* 27 */
+TCLAPI void		Tcl_ClassSetDestructor(Tcl_Interp *interp,
+				Tcl_Class clazz, Tcl_Method method);
+/* 28 */
+TCLAPI Tcl_Obj *	Tcl_GetObjectName(Tcl_Interp *interp,
+				Tcl_Object object);
+
+typedef struct {
+    const struct TclOOIntStubs *tclOOIntStubs;
+} TclOOStubHooks;
+
+typedef struct TclOOStubs {
+    int magic;
+    const TclOOStubHooks *hooks;
+
+    Tcl_Object (*tcl_CopyObjectInstance) (Tcl_Interp *interp, Tcl_Object sourceObject, const char *targetName, const char *targetNamespaceName); /* 0 */
+    Tcl_Object (*tcl_GetClassAsObject) (Tcl_Class clazz); /* 1 */
+    Tcl_Class (*tcl_GetObjectAsClass) (Tcl_Object object); /* 2 */
+    Tcl_Command (*tcl_GetObjectCommand) (Tcl_Object object); /* 3 */
+    Tcl_Object (*tcl_GetObjectFromObj) (Tcl_Interp *interp, Tcl_Obj *objPtr); /* 4 */
+    Tcl_Namespace * (*tcl_GetObjectNamespace) (Tcl_Object object); /* 5 */
+    Tcl_Class (*tcl_MethodDeclarerClass) (Tcl_Method method); /* 6 */
+    Tcl_Object (*tcl_MethodDeclarerObject) (Tcl_Method method); /* 7 */
+    int (*tcl_MethodIsPublic) (Tcl_Method method); /* 8 */
+    int (*tcl_MethodIsType) (Tcl_Method method, const Tcl_MethodType *typePtr, ClientData *clientDataPtr); /* 9 */
+    Tcl_Obj * (*tcl_MethodName) (Tcl_Method method); /* 10 */
+    Tcl_Method (*tcl_NewInstanceMethod) (Tcl_Interp *interp, Tcl_Object object, Tcl_Obj *nameObj, int isPublic, const Tcl_MethodType *typePtr, ClientData clientData); /* 11 */
+    Tcl_Method (*tcl_NewMethod) (Tcl_Interp *interp, Tcl_Class cls, Tcl_Obj *nameObj, int isPublic, const Tcl_MethodType *typePtr, ClientData clientData); /* 12 */
+    Tcl_Object (*tcl_NewObjectInstance) (Tcl_Interp *interp, Tcl_Class cls, const char *nameStr, const char *nsNameStr, int objc, Tcl_Obj *const *objv, int skip); /* 13 */
+    int (*tcl_ObjectDeleted) (Tcl_Object object); /* 14 */
+    int (*tcl_ObjectContextIsFiltering) (Tcl_ObjectContext context); /* 15 */
+    Tcl_Method (*tcl_ObjectContextMethod) (Tcl_ObjectContext context); /* 16 */
+    Tcl_Object (*tcl_ObjectContextObject) (Tcl_ObjectContext context); /* 17 */
+    int (*tcl_ObjectContextSkippedArgs) (Tcl_ObjectContext context); /* 18 */
+    ClientData (*tcl_ClassGetMetadata) (Tcl_Class clazz, const Tcl_ObjectMetadataType *typePtr); /* 19 */
+    void (*tcl_ClassSetMetadata) (Tcl_Class clazz, const Tcl_ObjectMetadataType *typePtr, ClientData metadata); /* 20 */
+    ClientData (*tcl_ObjectGetMetadata) (Tcl_Object object, const Tcl_ObjectMetadataType *typePtr); /* 21 */
+    void (*tcl_ObjectSetMetadata) (Tcl_Object object, const Tcl_ObjectMetadataType *typePtr, ClientData metadata); /* 22 */
+    int (*tcl_ObjectContextInvokeNext) (Tcl_Interp *interp, Tcl_ObjectContext context, int objc, Tcl_Obj *const *objv, int skip); /* 23 */
+    Tcl_ObjectMapMethodNameProc * (*tcl_ObjectGetMethodNameMapper) (Tcl_Object object); /* 24 */
+    void (*tcl_ObjectSetMethodNameMapper) (Tcl_Object object, Tcl_ObjectMapMethodNameProc *mapMethodNameProc); /* 25 */
+    void (*tcl_ClassSetConstructor) (Tcl_Interp *interp, Tcl_Class clazz, Tcl_Method method); /* 26 */
+    void (*tcl_ClassSetDestructor) (Tcl_Interp *interp, Tcl_Class clazz, Tcl_Method method); /* 27 */
+    Tcl_Obj * (*tcl_GetObjectName) (Tcl_Interp *interp, Tcl_Object object); /* 28 */
+} TclOOStubs;
+
+extern const TclOOStubs *tclOOStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCLOO_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#define Tcl_CopyObjectInstance \
+	(tclOOStubsPtr->tcl_CopyObjectInstance) /* 0 */
+#define Tcl_GetClassAsObject \
+	(tclOOStubsPtr->tcl_GetClassAsObject) /* 1 */
+#define Tcl_GetObjectAsClass \
+	(tclOOStubsPtr->tcl_GetObjectAsClass) /* 2 */
+#define Tcl_GetObjectCommand \
+	(tclOOStubsPtr->tcl_GetObjectCommand) /* 3 */
+#define Tcl_GetObjectFromObj \
+	(tclOOStubsPtr->tcl_GetObjectFromObj) /* 4 */
+#define Tcl_GetObjectNamespace \
+	(tclOOStubsPtr->tcl_GetObjectNamespace) /* 5 */
+#define Tcl_MethodDeclarerClass \
+	(tclOOStubsPtr->tcl_MethodDeclarerClass) /* 6 */
+#define Tcl_MethodDeclarerObject \
+	(tclOOStubsPtr->tcl_MethodDeclarerObject) /* 7 */
+#define Tcl_MethodIsPublic \
+	(tclOOStubsPtr->tcl_MethodIsPublic) /* 8 */
+#define Tcl_MethodIsType \
+	(tclOOStubsPtr->tcl_MethodIsType) /* 9 */
+#define Tcl_MethodName \
+	(tclOOStubsPtr->tcl_MethodName) /* 10 */
+#define Tcl_NewInstanceMethod \
+	(tclOOStubsPtr->tcl_NewInstanceMethod) /* 11 */
+#define Tcl_NewMethod \
+	(tclOOStubsPtr->tcl_NewMethod) /* 12 */
+#define Tcl_NewObjectInstance \
+	(tclOOStubsPtr->tcl_NewObjectInstance) /* 13 */
+#define Tcl_ObjectDeleted \
+	(tclOOStubsPtr->tcl_ObjectDeleted) /* 14 */
+#define Tcl_ObjectContextIsFiltering \
+	(tclOOStubsPtr->tcl_ObjectContextIsFiltering) /* 15 */
+#define Tcl_ObjectContextMethod \
+	(tclOOStubsPtr->tcl_ObjectContextMethod) /* 16 */
+#define Tcl_ObjectContextObject \
+	(tclOOStubsPtr->tcl_ObjectContextObject) /* 17 */
+#define Tcl_ObjectContextSkippedArgs \
+	(tclOOStubsPtr->tcl_ObjectContextSkippedArgs) /* 18 */
+#define Tcl_ClassGetMetadata \
+	(tclOOStubsPtr->tcl_ClassGetMetadata) /* 19 */
+#define Tcl_ClassSetMetadata \
+	(tclOOStubsPtr->tcl_ClassSetMetadata) /* 20 */
+#define Tcl_ObjectGetMetadata \
+	(tclOOStubsPtr->tcl_ObjectGetMetadata) /* 21 */
+#define Tcl_ObjectSetMetadata \
+	(tclOOStubsPtr->tcl_ObjectSetMetadata) /* 22 */
+#define Tcl_ObjectContextInvokeNext \
+	(tclOOStubsPtr->tcl_ObjectContextInvokeNext) /* 23 */
+#define Tcl_ObjectGetMethodNameMapper \
+	(tclOOStubsPtr->tcl_ObjectGetMethodNameMapper) /* 24 */
+#define Tcl_ObjectSetMethodNameMapper \
+	(tclOOStubsPtr->tcl_ObjectSetMethodNameMapper) /* 25 */
+#define Tcl_ClassSetConstructor \
+	(tclOOStubsPtr->tcl_ClassSetConstructor) /* 26 */
+#define Tcl_ClassSetDestructor \
+	(tclOOStubsPtr->tcl_ClassSetDestructor) /* 27 */
+#define Tcl_GetObjectName \
+	(tclOOStubsPtr->tcl_GetObjectName) /* 28 */
+
+#endif /* defined(USE_TCLOO_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#endif /* _TCLOODECLS */
diff --git a/generic/tclOODefineCmds.c b/generic/tclOODefineCmds.c
new file mode 100644
index 0000000..b0bfd9c
--- /dev/null
+++ b/generic/tclOODefineCmds.c
@@ -0,0 +1,2658 @@
+/*
+ * tclOODefineCmds.c --
+ *
+ *	This file contains the implementation of the ::oo::define command,
+ *	part of the object-system core (NB: not Tcl_Obj, but ::oo).
+ *
+ * Copyright (c) 2006-2013 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+/*
+ * The maximum length of fully-qualified object name to use in an errorinfo
+ * message. Longer than this will be curtailed.
+ */
+
+#define OBJNAME_LENGTH_IN_ERRORINFO_LIMIT 30
+
+/*
+ * Some things that make it easier to declare a slot.
+ */
+
+struct DeclaredSlot {
+    const char *name;
+    const Tcl_MethodType getterType;
+    const Tcl_MethodType setterType;
+};
+
+#define SLOT(name,getter,setter)					\
+    {"::oo::" name,							\
+	    {TCL_OO_METHOD_VERSION_CURRENT, "core method: " name " Getter", \
+		    getter, NULL, NULL},				\
+	    {TCL_OO_METHOD_VERSION_CURRENT, "core method: " name " Setter", \
+		    setter, NULL, NULL}}
+
+/*
+ * Forward declarations.
+ */
+
+static inline void	BumpGlobalEpoch(Tcl_Interp *interp, Class *classPtr);
+static Tcl_Command	FindCommand(Tcl_Interp *interp, Tcl_Obj *stringObj,
+			    Tcl_Namespace *const namespacePtr);
+static inline void	GenerateErrorInfo(Tcl_Interp *interp, Object *oPtr,
+			    Tcl_Obj *savedNameObj, const char *typeOfSubject);
+static inline int	MagicDefinitionInvoke(Tcl_Interp *interp,
+			    Tcl_Namespace *nsPtr, int cmdIndex,
+			    int objc, Tcl_Obj *const *objv);
+static inline Class *	GetClassInOuterContext(Tcl_Interp *interp,
+			    Tcl_Obj *className, const char *errMsg);
+static inline int	InitDefineContext(Tcl_Interp *interp,
+			    Tcl_Namespace *namespacePtr, Object *oPtr,
+			    int objc, Tcl_Obj *const objv[]);
+static inline void	RecomputeClassCacheFlag(Object *oPtr);
+static int		RenameDeleteMethod(Tcl_Interp *interp, Object *oPtr,
+			    int useClass, Tcl_Obj *const fromPtr,
+			    Tcl_Obj *const toPtr);
+static int		ClassFilterGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassFilterSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassMixinGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassMixinSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassSuperGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassSuperSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassVarsGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ClassVarsSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjFilterGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjFilterSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjMixinGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjMixinSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjVarsGet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static int		ObjVarsSet(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+
+/*
+ * Now define the slots used in declarations.
+ */
+
+static const struct DeclaredSlot slots[] = {
+    SLOT("define::filter",      ClassFilterGet, ClassFilterSet),
+    SLOT("define::mixin",       ClassMixinGet,  ClassMixinSet),
+    SLOT("define::superclass",  ClassSuperGet,  ClassSuperSet),
+    SLOT("define::variable",    ClassVarsGet,   ClassVarsSet),
+    SLOT("objdefine::filter",   ObjFilterGet,   ObjFilterSet),
+    SLOT("objdefine::mixin",    ObjMixinGet,    ObjMixinSet),
+    SLOT("objdefine::variable", ObjVarsGet,     ObjVarsSet),
+    {NULL, {0, 0, 0, 0, 0}, {0, 0, 0, 0, 0}}
+};
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * BumpGlobalEpoch --
+ *	Utility that ensures that call chains that are invalid will get thrown
+ *	away at an appropriate time. Note that exactly which epoch gets
+ *	advanced will depend on exactly what the class is tangled up in; in
+ *	the worst case, the simplest option is to advance the global epoch,
+ *	causing *everything* to be thrown away on next usage.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+BumpGlobalEpoch(
+    Tcl_Interp *interp,
+    Class *classPtr)
+{
+    if (classPtr != NULL
+	    && classPtr->subclasses.num == 0
+	    && classPtr->instances.num == 0
+	    && classPtr->mixinSubs.num == 0) {
+	/*
+	 * If a class has no subclasses or instances, and is not mixed into
+	 * anything, a change to its structure does not require us to
+	 * invalidate any call chains. Note that we still bump our object's
+	 * epoch if it has any mixins; the relation between a class and its
+	 * representative object is special. But it won't hurt.
+	 */
+
+	if (classPtr->thisPtr->mixins.num > 0) {
+	    classPtr->thisPtr->epoch++;
+	}
+	return;
+    }
+
+    /*
+     * Either there's no class (?!) or we're reconfiguring something that is
+     * in use. Force regeneration of call chains.
+     */
+
+    TclOOGetFoundation(interp)->epoch++;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * RecomputeClassCacheFlag --
+ *	Determine whether the object is prototypical of its class, and hence
+ *	able to use the class's method chain cache.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+RecomputeClassCacheFlag(
+    Object *oPtr)
+{
+    if ((oPtr->methodsPtr == NULL || oPtr->methodsPtr->numEntries == 0)
+	    && (oPtr->mixins.num == 0) && (oPtr->filters.num == 0)) {
+	oPtr->flags |= USE_CLASS_CACHE;
+    } else {
+	oPtr->flags &= ~USE_CLASS_CACHE;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOObjectSetFilters --
+ *	Install a list of filter method names into an object.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOObjectSetFilters(
+    Object *oPtr,
+    int numFilters,
+    Tcl_Obj *const *filters)
+{
+    int i;
+
+    if (oPtr->filters.num) {
+	Tcl_Obj *filterObj;
+
+	FOREACH(filterObj, oPtr->filters) {
+	    Tcl_DecrRefCount(filterObj);
+	}
+    }
+
+    if (numFilters == 0) {
+	/*
+	 * No list of filters was supplied, so we're deleting filters.
+	 */
+
+	ckfree(oPtr->filters.list);
+	oPtr->filters.list = NULL;
+	oPtr->filters.num = 0;
+	RecomputeClassCacheFlag(oPtr);
+    } else {
+	/*
+	 * We've got a list of filters, so we're creating filters.
+	 */
+
+	Tcl_Obj **filtersList;
+	int size = sizeof(Tcl_Obj *) * numFilters;	/* should be size_t */
+
+	if (oPtr->filters.num == 0) {
+	    filtersList = ckalloc(size);
+	} else {
+	    filtersList = ckrealloc(oPtr->filters.list, size);
+	}
+	for (i=0 ; i<numFilters ; i++) {
+	    filtersList[i] = filters[i];
+	    Tcl_IncrRefCount(filters[i]);
+	}
+	oPtr->filters.list = filtersList;
+	oPtr->filters.num = numFilters;
+	oPtr->flags &= ~USE_CLASS_CACHE;
+    }
+    oPtr->epoch++;		/* Only this object can be affected. */
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOClassSetFilters --
+ *	Install a list of filter method names into a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOClassSetFilters(
+    Tcl_Interp *interp,
+    Class *classPtr,
+    int numFilters,
+    Tcl_Obj *const *filters)
+{
+    int i;
+
+    if (classPtr->filters.num) {
+	Tcl_Obj *filterObj;
+
+	FOREACH(filterObj, classPtr->filters) {
+	    Tcl_DecrRefCount(filterObj);
+	}
+    }
+
+    if (numFilters == 0) {
+	/*
+	 * No list of filters was supplied, so we're deleting filters.
+	 */
+
+	ckfree(classPtr->filters.list);
+	classPtr->filters.list = NULL;
+	classPtr->filters.num = 0;
+    } else {
+	/*
+	 * We've got a list of filters, so we're creating filters.
+	 */
+
+	Tcl_Obj **filtersList;
+	int size = sizeof(Tcl_Obj *) * numFilters;	/* should be size_t */
+
+	if (classPtr->filters.num == 0) {
+	    filtersList = ckalloc(size);
+	} else {
+	    filtersList = ckrealloc(classPtr->filters.list, size);
+	}
+	for (i=0 ; i<numFilters ; i++) {
+	    filtersList[i] = filters[i];
+	    Tcl_IncrRefCount(filters[i]);
+	}
+	classPtr->filters.list = filtersList;
+	classPtr->filters.num = numFilters;
+    }
+
+    /*
+     * There may be many objects affected, so bump the global epoch.
+     */
+
+    BumpGlobalEpoch(interp, classPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOObjectSetMixins --
+ *	Install a list of mixin classes into an object.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOObjectSetMixins(
+    Object *oPtr,
+    int numMixins,
+    Class *const *mixins)
+{
+    Class *mixinPtr;
+    int i;
+
+    if (numMixins == 0) {
+	if (oPtr->mixins.num != 0) {
+	    FOREACH(mixinPtr, oPtr->mixins) {
+		if (mixinPtr) {
+		    TclOORemoveFromInstances(oPtr, mixinPtr);
+		}
+	    }
+	    ckfree(oPtr->mixins.list);
+	    oPtr->mixins.num = 0;
+	}
+	RecomputeClassCacheFlag(oPtr);
+    } else {
+	if (oPtr->mixins.num != 0) {
+	    FOREACH(mixinPtr, oPtr->mixins) {
+		if (mixinPtr && mixinPtr != oPtr->selfCls) {
+		    TclOORemoveFromInstances(oPtr, mixinPtr);
+		}
+	    }
+	    oPtr->mixins.list = ckrealloc(oPtr->mixins.list,
+		    sizeof(Class *) * numMixins);
+	} else {
+	    oPtr->mixins.list = ckalloc(sizeof(Class *) * numMixins);
+	    oPtr->flags &= ~USE_CLASS_CACHE;
+	}
+	oPtr->mixins.num = numMixins;
+	memcpy(oPtr->mixins.list, mixins, sizeof(Class *) * numMixins);
+	FOREACH(mixinPtr, oPtr->mixins) {
+	    if (mixinPtr != oPtr->selfCls) {
+		TclOOAddToInstances(oPtr, mixinPtr);
+	    }
+	}
+    }
+    oPtr->epoch++;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOClassSetMixins --
+ *	Install a list of mixin classes into a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOClassSetMixins(
+    Tcl_Interp *interp,
+    Class *classPtr,
+    int numMixins,
+    Class *const *mixins)
+{
+    Class *mixinPtr;
+    int i;
+
+    if (numMixins == 0) {
+	if (classPtr->mixins.num != 0) {
+	    FOREACH(mixinPtr, classPtr->mixins) {
+		TclOORemoveFromMixinSubs(classPtr, mixinPtr);
+	    }
+	    ckfree(classPtr->mixins.list);
+	    classPtr->mixins.num = 0;
+	}
+    } else {
+	if (classPtr->mixins.num != 0) {
+	    FOREACH(mixinPtr, classPtr->mixins) {
+		TclOORemoveFromMixinSubs(classPtr, mixinPtr);
+	    }
+	    classPtr->mixins.list = ckrealloc(classPtr->mixins.list,
+		    sizeof(Class *) * numMixins);
+	} else {
+	    classPtr->mixins.list = ckalloc(sizeof(Class *) * numMixins);
+	}
+	classPtr->mixins.num = numMixins;
+	memcpy(classPtr->mixins.list, mixins, sizeof(Class *) * numMixins);
+	FOREACH(mixinPtr, classPtr->mixins) {
+	    TclOOAddToMixinSubs(classPtr, mixinPtr);
+	}
+    }
+    BumpGlobalEpoch(interp, classPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * RenameDeleteMethod --
+ *	Core of the code to rename and delete methods.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+RenameDeleteMethod(
+    Tcl_Interp *interp,
+    Object *oPtr,
+    int useClass,
+    Tcl_Obj *const fromPtr,
+    Tcl_Obj *const toPtr)
+{
+    Tcl_HashEntry *hPtr, *newHPtr = NULL;
+    Method *mPtr;
+    int isNew;
+
+    if (!useClass) {
+	if (!oPtr->methodsPtr) {
+	noSuchMethod:
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "method %s does not exist", TclGetString(fromPtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		    TclGetString(fromPtr), NULL);
+	    return TCL_ERROR;
+	}
+	hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char *) fromPtr);
+	if (hPtr == NULL) {
+	    goto noSuchMethod;
+	}
+	if (toPtr) {
+	    newHPtr = Tcl_CreateHashEntry(oPtr->methodsPtr, (char *) toPtr,
+		    &isNew);
+	    if (hPtr == newHPtr) {
+	    renameToSelf:
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"cannot rename method to itself", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OO", "RENAME_TO_SELF", NULL);
+		return TCL_ERROR;
+	    } else if (!isNew) {
+	    renameToExisting:
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"method called %s already exists",
+			TclGetString(toPtr)));
+		Tcl_SetErrorCode(interp, "TCL", "OO", "RENAME_OVER", NULL);
+		return TCL_ERROR;
+	    }
+	}
+    } else {
+	hPtr = Tcl_FindHashEntry(&oPtr->classPtr->classMethods,
+		(char *) fromPtr);
+	if (hPtr == NULL) {
+	    goto noSuchMethod;
+	}
+	if (toPtr) {
+	    newHPtr = Tcl_CreateHashEntry(&oPtr->classPtr->classMethods,
+		    (char *) toPtr, &isNew);
+	    if (hPtr == newHPtr) {
+		goto renameToSelf;
+	    } else if (!isNew) {
+		goto renameToExisting;
+	    }
+	}
+    }
+
+    /*
+     * Complete the splicing by changing the method's name.
+     */
+
+    mPtr = Tcl_GetHashValue(hPtr);
+    if (toPtr) {
+	Tcl_IncrRefCount(toPtr);
+	Tcl_DecrRefCount(mPtr->namePtr);
+	mPtr->namePtr = toPtr;
+	Tcl_SetHashValue(newHPtr, mPtr);
+    } else {
+	if (!useClass) {
+	    RecomputeClassCacheFlag(oPtr);
+	}
+	TclOODelMethodRef(mPtr);
+    }
+    Tcl_DeleteHashEntry(hPtr);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOUnknownDefinition --
+ *	Handles what happens when an unknown command is encountered during the
+ *	processing of a definition script. Works by finding a command in the
+ *	operating definition namespace that the requested command is a unique
+ *	prefix of.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOUnknownDefinition(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Namespace *nsPtr = (Namespace *) Tcl_GetCurrentNamespace(interp);
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hPtr;
+    int soughtLen;
+    const char *soughtStr, *matchedStr = NULL;
+
+    if (objc < 2) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"bad call of unknown handler", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_UNKNOWN", NULL);
+	return TCL_ERROR;
+    }
+    if (TclOOGetDefineCmdContext(interp) == NULL) {
+	return TCL_ERROR;
+    }
+
+    soughtStr = TclGetStringFromObj(objv[1], &soughtLen);
+    if (soughtLen == 0) {
+	goto noMatch;
+    }
+    hPtr = Tcl_FirstHashEntry(&nsPtr->cmdTable, &search);
+    while (hPtr != NULL) {
+	const char *nameStr = Tcl_GetHashKey(&nsPtr->cmdTable, hPtr);
+
+	if (strncmp(soughtStr, nameStr, soughtLen) == 0) {
+	    if (matchedStr != NULL) {
+		goto noMatch;
+	    }
+	    matchedStr = nameStr;
+	}
+	hPtr = Tcl_NextHashEntry(&search);
+    }
+
+    if (matchedStr != NULL) {
+	/*
+	 * Got one match, and only one match!
+	 */
+
+	Tcl_Obj **newObjv = TclStackAlloc(interp, sizeof(Tcl_Obj*)*(objc-1));
+	int result;
+
+	newObjv[0] = Tcl_NewStringObj(matchedStr, -1);
+	Tcl_IncrRefCount(newObjv[0]);
+	if (objc > 2) {
+	    memcpy(newObjv+1, objv+2, sizeof(Tcl_Obj *) * (objc-2));
+	}
+	result = Tcl_EvalObjv(interp, objc-1, newObjv, 0);
+	Tcl_DecrRefCount(newObjv[0]);
+	TclStackFree(interp, newObjv);
+	return result;
+    }
+
+  noMatch:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "invalid command name \"%s\"", soughtStr));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "COMMAND", soughtStr, NULL);
+    return TCL_ERROR;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * FindCommand --
+ *	Specialized version of Tcl_FindCommand that handles command prefixes
+ *	and disallows namespace magic.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static Tcl_Command
+FindCommand(
+    Tcl_Interp *interp,
+    Tcl_Obj *stringObj,
+    Tcl_Namespace *const namespacePtr)
+{
+    int length;
+    const char *nameStr, *string = TclGetStringFromObj(stringObj, &length);
+    register Namespace *const nsPtr = (Namespace *) namespacePtr;
+    FOREACH_HASH_DECLS;
+    Tcl_Command cmd, cmd2;
+
+    /*
+     * If someone is playing games, we stop playing right now.
+     */
+
+    if (string[0] == '\0' || strstr(string, "::") != NULL) {
+	return NULL;
+    }
+
+    /*
+     * Do the exact lookup first.
+     */
+
+    cmd = Tcl_FindCommand(interp, string, namespacePtr, TCL_NAMESPACE_ONLY);
+    if (cmd != NULL) {
+	return cmd;
+    }
+
+    /*
+     * Bother, need to perform an approximate match. Iterate across the hash
+     * table of commands in the namespace.
+     */
+
+    FOREACH_HASH(nameStr, cmd2, &nsPtr->cmdTable) {
+	if (strncmp(string, nameStr, length) == 0) {
+	    if (cmd != NULL) {
+		return NULL;
+	    }
+	    cmd = cmd2;
+	}
+    }
+
+    /*
+     * Either we found one thing or we found nothing. Either way, return it.
+     */
+
+    return cmd;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InitDefineContext --
+ *	Does the magic incantations necessary to push the special stack frame
+ *	used when processing object definitions. It is up to the caller to
+ *	dispose of the frame (with TclPopStackFrame) when finished.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline int
+InitDefineContext(
+    Tcl_Interp *interp,
+    Tcl_Namespace *namespacePtr,
+    Object *oPtr,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    CallFrame *framePtr, **framePtrPtr = &framePtr;
+
+    if (namespacePtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"cannot process definitions; support namespace deleted",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    /* framePtrPtr is needed to satisfy GCC 3.3's strict aliasing rules */
+
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    namespacePtr, FRAME_IS_OO_DEFINE);
+    framePtr->clientData = oPtr;
+    framePtr->objc = objc;
+    framePtr->objv = objv;	/* Reference counts do not need to be
+				 * incremented here. */
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetDefineCmdContext --
+ *	Extracts the magic token from the current stack frame, or returns NULL
+ *	(and leaves an error message) otherwise.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Object
+TclOOGetDefineCmdContext(
+    Tcl_Interp *interp)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Object object;
+
+    if ((iPtr->varFramePtr == NULL)
+	    || (iPtr->varFramePtr->isProcCallFrame != FRAME_IS_OO_DEFINE)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"this command may only be called from within the context of"
+		" an ::oo::define or ::oo::objdefine command", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return NULL;
+    }
+    object = iPtr->varFramePtr->clientData;
+    if (Tcl_ObjectDeleted(object)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"this command cannot be called when the object has been"
+		" deleted", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return NULL;
+    }
+    return object;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * GetClassInOuterContext --
+ *	Wrapper round Tcl_GetObjectFromObj to perform the lookup in the
+ *	context that called oo::define (or equivalent). Note that this may
+ *	have to go up multiple levels to get the level that we started doing
+ *	definitions at.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline Class *
+GetClassInOuterContext(
+    Tcl_Interp *interp,
+    Tcl_Obj *className,
+    const char *errMsg)
+{
+    Interp *iPtr = (Interp *) interp;
+    Object *oPtr;
+    CallFrame *savedFramePtr = iPtr->varFramePtr;
+
+    while (iPtr->varFramePtr->isProcCallFrame == FRAME_IS_OO_DEFINE) {
+	if (iPtr->varFramePtr->callerVarPtr == NULL) {
+	    Tcl_Panic("getting outer context when already in global context");
+	}
+	iPtr->varFramePtr = iPtr->varFramePtr->callerVarPtr;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, className);
+    iPtr->varFramePtr = savedFramePtr;
+    if (oPtr == NULL) {
+	return NULL;
+    }
+    if (oPtr->classPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(errMsg, -1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		TclGetString(className), NULL);
+	return NULL;
+    }
+    return oPtr->classPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * GenerateErrorInfo --
+ *	Factored out code to generate part of the error trace messages.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline void
+GenerateErrorInfo(
+    Tcl_Interp *interp,		/* Where to store the error info trace. */
+    Object *oPtr,		/* What object (or class) was being configured
+				 * when the error occurred? */
+    Tcl_Obj *savedNameObj,	/* Name of object saved from before script was
+				 * evaluated, which is needed if the object
+				 * goes away part way through execution. OTOH,
+				 * if the object isn't deleted then its
+				 * current name (post-execution) has to be
+				 * used. This matters, because the object
+				 * could have been renamed... */
+    const char *typeOfSubject)	/* Part of the message, saying whether it was
+				 * an object, class or class-as-object that
+				 * was being configured. */
+{
+    int length;
+    Tcl_Obj *realNameObj = Tcl_ObjectDeleted((Tcl_Object) oPtr)
+	    ? savedNameObj : TclOOObjectName(interp, oPtr);
+    const char *objName = TclGetStringFromObj(realNameObj, &length);
+    int limit = OBJNAME_LENGTH_IN_ERRORINFO_LIMIT;
+    int overflow = (length > limit);
+
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (in definition script for %s \"%.*s%s\" line %d)",
+	    typeOfSubject, (overflow ? limit : length), objName,
+	    (overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * MagicDefinitionInvoke --
+ *	Part of the implementation of the "oo::define" and "oo::objdefine"
+ *	commands that is used to implement the more-than-one-argument case,
+ *	applying ensemble-like tricks with dispatch so that error messages are
+ *	clearer. Doesn't handle the management of the stack frame.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline int
+MagicDefinitionInvoke(
+    Tcl_Interp *interp,
+    Tcl_Namespace *nsPtr,
+    int cmdIndex,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Tcl_Obj *objPtr, *obj2Ptr, **objs;
+    Tcl_Command cmd;
+    int isRoot, dummy, result, offset = cmdIndex + 1;
+
+    /*
+     * More than one argument: fire them through the ensemble processing
+     * engine so that everything appears to be good and proper in error
+     * messages. Note that we cannot just concatenate and send through
+     * Tcl_EvalObjEx, as that doesn't do ensemble processing, and we cannot go
+     * through Tcl_EvalObjv without the extra work to pre-find the command, as
+     * that finds command names in the wrong namespace at the moment. Ugly!
+     */
+
+    isRoot = TclInitRewriteEnsemble(interp, offset, 1, objv);
+
+    /*
+     * Build the list of arguments using a Tcl_Obj as a workspace. See
+     * comments above for why these contortions are necessary.
+     */
+
+    objPtr = Tcl_NewObj();
+    obj2Ptr = Tcl_NewObj();
+    cmd = FindCommand(interp, objv[cmdIndex], nsPtr);
+    if (cmd == NULL) {
+	/* punt this case! */
+	Tcl_AppendObjToObj(obj2Ptr, objv[cmdIndex]);
+    } else {
+	Tcl_GetCommandFullName(interp, cmd, obj2Ptr);
+    }
+    Tcl_ListObjAppendElement(NULL, objPtr, obj2Ptr);
+    /* TODO: overflow? */
+    Tcl_ListObjReplace(NULL, objPtr, 1, 0, objc-offset, objv+offset);
+    Tcl_ListObjGetElements(NULL, objPtr, &dummy, &objs);
+
+    result = Tcl_EvalObjv(interp, objc-cmdIndex, objs, TCL_EVAL_INVOKE);
+    if (isRoot) {
+	TclResetRewriteEnsemble(interp, 1);
+    }
+    Tcl_DecrRefCount(objPtr);
+
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineObjCmd --
+ *	Implementation of the "oo::define" command. Works by effectively doing
+ *	the same as 'namespace eval', but with extra magic applied so that the
+ *	object to be modified is known to the commands in the target
+ *	namespace. Also does ensemble-like tricks with dispatch so that error
+ *	messages are clearer.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Foundation *fPtr = TclOOGetFoundation(interp);
+    Object *oPtr;
+    int result;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (oPtr->classPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"%s does not refer to a class",TclGetString(objv[1])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		TclGetString(objv[1]), NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the oo::define namespace the current namespace and evaluate the
+     * command(s).
+     */
+
+    if (InitDefineContext(interp, fPtr->defineNs, oPtr, objc,objv) != TCL_OK){
+	return TCL_ERROR;
+    }
+
+    AddRef(oPtr);
+    if (objc == 3) {
+	Tcl_Obj *objNameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_IncrRefCount(objNameObj);
+	result = TclEvalObjEx(interp, objv[2], 0,
+		((Interp *)interp)->cmdFramePtr, 2);
+	if (result == TCL_ERROR) {
+	    GenerateErrorInfo(interp, oPtr, objNameObj, "class");
+	}
+	TclDecrRefCount(objNameObj);
+    } else {
+	result = MagicDefinitionInvoke(interp, fPtr->defineNs, 2, objc, objv);
+    }
+    DelRef(oPtr);
+
+    /*
+     * Restore the previous "current" namespace.
+     */
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOObjDefObjCmd --
+ *	Implementation of the "oo::objdefine" command. Works by effectively
+ *	doing the same as 'namespace eval', but with extra magic applied so
+ *	that the object to be modified is known to the commands in the target
+ *	namespace. Also does ensemble-like tricks with dispatch so that error
+ *	messages are clearer.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOOObjDefObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Foundation *fPtr = TclOOGetFoundation(interp);
+    Object *oPtr;
+    int result;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objectName arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make the oo::objdefine namespace the current namespace and evaluate the
+     * command(s).
+     */
+
+    if (InitDefineContext(interp, fPtr->objdefNs, oPtr, objc,objv) != TCL_OK){
+	return TCL_ERROR;
+    }
+
+    AddRef(oPtr);
+    if (objc == 3) {
+	Tcl_Obj *objNameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_IncrRefCount(objNameObj);
+	result = TclEvalObjEx(interp, objv[2], 0,
+		((Interp *)interp)->cmdFramePtr, 2);
+	if (result == TCL_ERROR) {
+	    GenerateErrorInfo(interp, oPtr, objNameObj, "object");
+	}
+	TclDecrRefCount(objNameObj);
+    } else {
+	result = MagicDefinitionInvoke(interp, fPtr->objdefNs, 2, objc, objv);
+    }
+    DelRef(oPtr);
+
+    /*
+     * Restore the previous "current" namespace.
+     */
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineSelfObjCmd --
+ *	Implementation of the "self" subcommand of the "oo::define" command.
+ *	Works by effectively doing the same as 'namespace eval', but with
+ *	extra magic applied so that the object to be modified is known to the
+ *	commands in the target namespace. Also does ensemble-like tricks with
+ *	dispatch so that error messages are clearer.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineSelfObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Foundation *fPtr = TclOOGetFoundation(interp);
+    Object *oPtr;
+    int result;
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (objc < 2) {
+	Tcl_SetObjResult(interp, TclOOObjectName(interp, oPtr));
+	return TCL_OK;
+    }
+
+    /*
+     * Make the oo::objdefine namespace the current namespace and evaluate the
+     * command(s).
+     */
+
+    if (InitDefineContext(interp, fPtr->objdefNs, oPtr, objc,objv) != TCL_OK){
+	return TCL_ERROR;
+    }
+
+    AddRef(oPtr);
+    if (objc == 2) {
+	Tcl_Obj *objNameObj = TclOOObjectName(interp, oPtr);
+
+	Tcl_IncrRefCount(objNameObj);
+	result = TclEvalObjEx(interp, objv[1], 0,
+		((Interp *)interp)->cmdFramePtr, 2);
+	if (result == TCL_ERROR) {
+	    GenerateErrorInfo(interp, oPtr, objNameObj, "class object");
+	}
+	TclDecrRefCount(objNameObj);
+    } else {
+	result = MagicDefinitionInvoke(interp, fPtr->objdefNs, 1, objc, objv);
+    }
+    DelRef(oPtr);
+
+    /*
+     * Restore the previous "current" namespace.
+     */
+
+    TclPopStackFrame(interp);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineObjSelfObjCmd --
+ *	Implementation of the "self" subcommand of the "oo::objdefine"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineObjSelfObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr;
+
+    if (objc != 1) {
+	Tcl_WrongNumArgs(interp, 1, objv, NULL);
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, TclOOObjectName(interp, oPtr));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineClassObjCmd --
+ *	Implementation of the "class" subcommand of the "oo::objdefine"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineClassObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr;
+    Class *clsPtr;
+    Foundation *fPtr = TclOOGetFoundation(interp);
+
+    /*
+     * Parse the context to get the object to operate on.
+     */
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (oPtr->flags & ROOT_OBJECT) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"may not modify the class of the root object class", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+    if (oPtr->flags & ROOT_CLASS) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"may not modify the class of the class of classes", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse the argument to get the class to set the object's class to.
+     */
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassInOuterContext(interp, objv[1],
+	    "the class of an object must be a class");
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Apply semantic checks. In particular, classes and non-classes are not
+     * interchangable (too complicated to do the conversion!) so we must
+     * produce an error if any attempt is made to swap from one to the other.
+     */
+
+    if ((oPtr->classPtr==NULL) == TclOOIsReachable(fPtr->classCls, clsPtr)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"may not change a %sclass object into a %sclass object",
+		(oPtr->classPtr==NULL ? "non-" : ""),
+		(oPtr->classPtr==NULL ? "" : "non-")));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "TRANSMUTATION", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set the object's class.
+     */
+
+    if (oPtr->selfCls != clsPtr) {
+	TclOORemoveFromInstances(oPtr, oPtr->selfCls);
+	oPtr->selfCls = clsPtr;
+	TclOOAddToInstances(oPtr, oPtr->selfCls);
+	if (!(clsPtr->thisPtr->flags & OBJECT_DELETED)) {
+	    oPtr->flags &= ~CLASS_GONE;
+	}
+	if (oPtr->classPtr != NULL) {
+	    BumpGlobalEpoch(interp, oPtr->classPtr);
+	} else {
+	    oPtr->epoch++;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineConstructorObjCmd --
+ *	Implementation of the "constructor" subcommand of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineConstructorObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr;
+    Class *clsPtr;
+    Tcl_Method method;
+    int bodyLength;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arguments body");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Extract and validate the context, which is the class that we wish to
+     * modify.
+     */
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    clsPtr = oPtr->classPtr;
+
+    TclGetStringFromObj(objv[2], &bodyLength);
+    if (bodyLength > 0) {
+	/*
+	 * Create the method structure.
+	 */
+
+	method = (Tcl_Method) TclOONewProcMethod(interp, clsPtr,
+		PUBLIC_METHOD, NULL, objv[1], objv[2], NULL);
+	if (method == NULL) {
+	    return TCL_ERROR;
+	}
+    } else {
+	/*
+	 * Delete the constructor method record and set the field in the
+	 * class record to NULL.
+	 */
+
+	method = NULL;
+    }
+
+    /*
+     * Place the method structure in the class record. Note that we might not
+     * immediately delete the constructor as this might be being done during
+     * execution of the constructor itself.
+     */
+
+    Tcl_ClassSetConstructor(interp, (Tcl_Class) clsPtr, method);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineDeleteMethodObjCmd --
+ *	Implementation of the "deletemethod" subcommand of the "oo::define"
+ *	and "oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineDeleteMethodObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceDeleteMethod = (clientData != NULL);
+    Object *oPtr;
+    int i;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?name ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (!isInstanceDeleteMethod && !oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    for (i=1 ; i<objc ; i++) {
+	/*
+	 * Delete the method structure from the appropriate hash table.
+	 */
+
+	if (RenameDeleteMethod(interp, oPtr, !isInstanceDeleteMethod,
+		objv[i], NULL) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    if (isInstanceDeleteMethod) {
+	oPtr->epoch++;
+    } else {
+	BumpGlobalEpoch(interp, oPtr->classPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineDestructorObjCmd --
+ *	Implementation of the "destructor" subcommand of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineDestructorObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr;
+    Class *clsPtr;
+    Tcl_Method method;
+    int bodyLength;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "body");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    clsPtr = oPtr->classPtr;
+
+    TclGetStringFromObj(objv[1], &bodyLength);
+    if (bodyLength > 0) {
+	/*
+	 * Create the method structure.
+	 */
+
+	method = (Tcl_Method) TclOONewProcMethod(interp, clsPtr,
+		PUBLIC_METHOD, NULL, NULL, objv[1], NULL);
+	if (method == NULL) {
+	    return TCL_ERROR;
+	}
+    } else {
+	/*
+	 * Delete the destructor method record and set the field in the class
+	 * record to NULL.
+	 */
+
+	method = NULL;
+    }
+
+    /*
+     * Place the method structure in the class record. Note that we might not
+     * immediately delete the destructor as this might be being done during
+     * execution of the destructor itself. Also note that setting a
+     * destructor during a destructor is fairly dumb anyway.
+     */
+
+    Tcl_ClassSetDestructor(interp, (Tcl_Class) clsPtr, method);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineExportObjCmd --
+ *	Implementation of the "export" subcommand of the "oo::define" and
+ *	"oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineExportObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceExport = (clientData != NULL);
+    Object *oPtr;
+    Method *mPtr;
+    Tcl_HashEntry *hPtr;
+    Class *clsPtr;
+    int i, isNew, changed = 0;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?name ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    clsPtr = oPtr->classPtr;
+    if (!isInstanceExport && !clsPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    for (i=1 ; i<objc ; i++) {
+	/*
+	 * Exporting is done by adding the PUBLIC_METHOD flag to the method
+	 * record. If there is no such method in this object or class (i.e.
+	 * the method comes from something inherited from or that we're an
+	 * instance of) then we put in a blank record with that flag; such
+	 * records are skipped over by the call chain engine *except* for
+	 * their flags member.
+	 */
+
+	if (isInstanceExport) {
+	    if (!oPtr->methodsPtr) {
+		oPtr->methodsPtr = ckalloc(sizeof(Tcl_HashTable));
+		Tcl_InitObjHashTable(oPtr->methodsPtr);
+		oPtr->flags &= ~USE_CLASS_CACHE;
+	    }
+	    hPtr = Tcl_CreateHashEntry(oPtr->methodsPtr, (char *) objv[i],
+		    &isNew);
+	} else {
+	    hPtr = Tcl_CreateHashEntry(&clsPtr->classMethods, (char*) objv[i],
+		    &isNew);
+	}
+
+	if (isNew) {
+	    mPtr = ckalloc(sizeof(Method));
+	    memset(mPtr, 0, sizeof(Method));
+	    mPtr->refCount = 1;
+	    mPtr->namePtr = objv[i];
+	    Tcl_IncrRefCount(objv[i]);
+	    Tcl_SetHashValue(hPtr, mPtr);
+	} else {
+	    mPtr = Tcl_GetHashValue(hPtr);
+	}
+	if (isNew || !(mPtr->flags & PUBLIC_METHOD)) {
+	    mPtr->flags |= PUBLIC_METHOD;
+	    changed = 1;
+	}
+    }
+
+    /*
+     * Bump the right epoch if we actually changed anything.
+     */
+
+    if (changed) {
+	if (isInstanceExport) {
+	    oPtr->epoch++;
+	} else {
+	    BumpGlobalEpoch(interp, clsPtr);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineForwardObjCmd --
+ *	Implementation of the "forward" subcommand of the "oo::define" and
+ *	"oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineForwardObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceForward = (clientData != NULL);
+    Object *oPtr;
+    Method *mPtr;
+    int isPublic;
+    Tcl_Obj *prefixObj;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name cmdName ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (!isInstanceForward && !oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+    isPublic = Tcl_StringMatch(TclGetString(objv[1]), "[a-z]*")
+	    ? PUBLIC_METHOD : 0;
+
+    /*
+     * Create the method structure.
+     */
+
+    prefixObj = Tcl_NewListObj(objc-2, objv+2);
+    if (isInstanceForward) {
+	mPtr = TclOONewForwardInstanceMethod(interp, oPtr, isPublic, objv[1],
+		prefixObj);
+    } else {
+	mPtr = TclOONewForwardMethod(interp, oPtr->classPtr, isPublic,
+		objv[1], prefixObj);
+    }
+    if (mPtr == NULL) {
+	Tcl_DecrRefCount(prefixObj);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineMethodObjCmd --
+ *	Implementation of the "method" subcommand of the "oo::define" and
+ *	"oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineMethodObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceMethod = (clientData != NULL);
+    Object *oPtr;
+    int isPublic;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name args body");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (!isInstanceMethod && !oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+    isPublic = Tcl_StringMatch(TclGetString(objv[1]), "[a-z]*")
+	    ? PUBLIC_METHOD : 0;
+
+    /*
+     * Create the method by using the right back-end API.
+     */
+
+    if (isInstanceMethod) {
+	if (TclOONewProcInstanceMethod(interp, oPtr, isPublic, objv[1],
+		objv[2], objv[3], NULL) == NULL) {
+	    return TCL_ERROR;
+	}
+    } else {
+	if (TclOONewProcMethod(interp, oPtr->classPtr, isPublic, objv[1],
+		objv[2], objv[3], NULL) == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineMixinObjCmd --
+ *	Implementation of the "mixin" subcommand of the "oo::define" and
+ *	"oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineMixinObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    const int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceMixin = (clientData != NULL);
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Class **mixins;
+    int i;
+
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (!isInstanceMixin && !oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+    mixins = TclStackAlloc(interp, sizeof(Class *) * (objc-1));
+
+    for (i=1 ; i<objc ; i++) {
+	Class *clsPtr = GetClassInOuterContext(interp, objv[i],
+		"may only mix in classes");
+
+	if (clsPtr == NULL) {
+	    goto freeAndError;
+	}
+	if (!isInstanceMixin && TclOOIsReachable(oPtr->classPtr, clsPtr)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "may not mix a class into itself", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "SELF_MIXIN", NULL);
+	    goto freeAndError;
+	}
+	mixins[i-1] = clsPtr;
+    }
+
+    if (isInstanceMixin) {
+	TclOOObjectSetMixins(oPtr, objc-1, mixins);
+    } else {
+	TclOOClassSetMixins(interp, oPtr->classPtr, objc-1, mixins);
+    }
+
+    TclStackFree(interp, mixins);
+    return TCL_OK;
+
+  freeAndError:
+    TclStackFree(interp, mixins);
+    return TCL_ERROR;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineRenameMethodObjCmd --
+ *	Implementation of the "renamemethod" subcommand of the "oo::define"
+ *	and "oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineRenameMethodObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceRenameMethod = (clientData != NULL);
+    Object *oPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "oldName newName");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (!isInstanceRenameMethod && !oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Delete the method entry from the appropriate hash table, and transfer
+     * the thing it points to to its new entry. To do this, we first need to
+     * get the entries from the appropriate hash tables (this can generate a
+     * range of errors...)
+     */
+
+    if (RenameDeleteMethod(interp, oPtr, !isInstanceRenameMethod,
+	    objv[1], objv[2]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (isInstanceRenameMethod) {
+	oPtr->epoch++;
+    } else {
+	BumpGlobalEpoch(interp, oPtr->classPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineUnexportObjCmd --
+ *	Implementation of the "unexport" subcommand of the "oo::define" and
+ *	"oo::objdefine" commands.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineUnexportObjCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    int isInstanceUnexport = (clientData != NULL);
+    Object *oPtr;
+    Method *mPtr;
+    Tcl_HashEntry *hPtr;
+    Class *clsPtr;
+    int i, isNew, changed = 0;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name ?name ...?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    clsPtr = oPtr->classPtr;
+    if (!isInstanceUnexport && !clsPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    for (i=1 ; i<objc ; i++) {
+	/*
+	 * Unexporting is done by removing the PUBLIC_METHOD flag from the
+	 * method record. If there is no such method in this object or class
+	 * (i.e. the method comes from something inherited from or that we're
+	 * an instance of) then we put in a blank record without that flag;
+	 * such records are skipped over by the call chain engine *except* for
+	 * their flags member.
+	 */
+
+	if (isInstanceUnexport) {
+	    if (!oPtr->methodsPtr) {
+		oPtr->methodsPtr = ckalloc(sizeof(Tcl_HashTable));
+		Tcl_InitObjHashTable(oPtr->methodsPtr);
+		oPtr->flags &= ~USE_CLASS_CACHE;
+	    }
+	    hPtr = Tcl_CreateHashEntry(oPtr->methodsPtr, (char *) objv[i],
+		    &isNew);
+	} else {
+	    hPtr = Tcl_CreateHashEntry(&clsPtr->classMethods, (char*) objv[i],
+		    &isNew);
+	}
+
+	if (isNew) {
+	    mPtr = ckalloc(sizeof(Method));
+	    memset(mPtr, 0, sizeof(Method));
+	    mPtr->refCount = 1;
+	    mPtr->namePtr = objv[i];
+	    Tcl_IncrRefCount(objv[i]);
+	    Tcl_SetHashValue(hPtr, mPtr);
+	} else {
+	    mPtr = Tcl_GetHashValue(hPtr);
+	}
+	if (isNew || mPtr->flags & PUBLIC_METHOD) {
+	    mPtr->flags &= ~PUBLIC_METHOD;
+	    changed = 1;
+	}
+    }
+
+    /*
+     * Bump the right epoch if we actually changed anything.
+     */
+
+    if (changed) {
+	if (isInstanceUnexport) {
+	    oPtr->epoch++;
+	} else {
+	    BumpGlobalEpoch(interp, clsPtr);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_ClassSetConstructor, Tcl_ClassSetDestructor --
+ *	How to install a constructor or destructor into a class; API to call
+ *	from C.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+Tcl_ClassSetConstructor(
+    Tcl_Interp *interp,
+    Tcl_Class clazz,
+    Tcl_Method method)
+{
+    Class *clsPtr = (Class *) clazz;
+
+    if (method != (Tcl_Method) clsPtr->constructorPtr) {
+	TclOODelMethodRef(clsPtr->constructorPtr);
+	clsPtr->constructorPtr = (Method *) method;
+
+	/*
+	 * Remember to invalidate the cached constructor chain for this class.
+	 * [Bug 2531577]
+	 */
+
+	if (clsPtr->constructorChainPtr) {
+	    TclOODeleteChain(clsPtr->constructorChainPtr);
+	    clsPtr->constructorChainPtr = NULL;
+	}
+	BumpGlobalEpoch(interp, clsPtr);
+    }
+}
+
+void
+Tcl_ClassSetDestructor(
+    Tcl_Interp *interp,
+    Tcl_Class clazz,
+    Tcl_Method method)
+{
+    Class *clsPtr = (Class *) clazz;
+
+    if (method != (Tcl_Method) clsPtr->destructorPtr) {
+	TclOODelMethodRef(clsPtr->destructorPtr);
+	clsPtr->destructorPtr = (Method *) method;
+	if (clsPtr->destructorChainPtr) {
+	    TclOODeleteChain(clsPtr->destructorChainPtr);
+	    clsPtr->destructorChainPtr = NULL;
+	}
+	BumpGlobalEpoch(interp, clsPtr);
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODefineSlots --
+ *	Create the "::oo::Slot" class and its standard instances. Class
+ *	definition is empty at the stage (added by scripting).
+ *
+ * ----------------------------------------------------------------------
+ */
+
+int
+TclOODefineSlots(
+    Foundation *fPtr)
+{
+    const struct DeclaredSlot *slotInfoPtr;
+    Tcl_Obj *getName = Tcl_NewStringObj("Get", -1);
+    Tcl_Obj *setName = Tcl_NewStringObj("Set", -1);
+    Class *slotCls;
+
+    slotCls = ((Object *) Tcl_NewObjectInstance(fPtr->interp, (Tcl_Class)
+	    fPtr->classCls, "::oo::Slot", NULL, -1, NULL, 0))->classPtr;
+    if (slotCls == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_IncrRefCount(getName);
+    Tcl_IncrRefCount(setName);
+    for (slotInfoPtr = slots ; slotInfoPtr->name ; slotInfoPtr++) {
+	Tcl_Object slotObject = Tcl_NewObjectInstance(fPtr->interp,
+		(Tcl_Class) slotCls, slotInfoPtr->name, NULL,-1,NULL,0);
+
+	if (slotObject == NULL) {
+	    continue;
+	}
+	Tcl_NewInstanceMethod(fPtr->interp, slotObject, getName, 0,
+		&slotInfoPtr->getterType, NULL);
+	Tcl_NewInstanceMethod(fPtr->interp, slotObject, setName, 0,
+		&slotInfoPtr->setterType, NULL);
+    }
+    Tcl_DecrRefCount(getName);
+    Tcl_DecrRefCount(setName);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ClassFilterGet, ClassFilterSet --
+ *	Implementation of the "filter" slot accessors of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ClassFilterGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj, *filterObj;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    }
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(filterObj, oPtr->classPtr->filters) {
+	Tcl_ListObjAppendElement(NULL, resultObj, filterObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ClassFilterSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int filterc;
+    Tcl_Obj **filterv;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"filterList");
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    } else if (Tcl_ListObjGetElements(interp, objv[0], &filterc,
+	    &filterv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    TclOOClassSetFilters(interp, oPtr->classPtr, filterc, filterv);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ClassMixinGet, ClassMixinSet --
+ *	Implementation of the "mixin" slot accessors of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ClassMixinGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj;
+    Class *mixinPtr;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    }
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(mixinPtr, oPtr->classPtr->mixins) {
+	Tcl_ListObjAppendElement(NULL, resultObj,
+		TclOOObjectName(interp, mixinPtr->thisPtr));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+
+}
+
+static int
+ClassMixinSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int mixinc, i;
+    Tcl_Obj **mixinv;
+    Class **mixins;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"mixinList");
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    } else if (Tcl_ListObjGetElements(interp, objv[0], &mixinc,
+	    &mixinv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    mixins = TclStackAlloc(interp, sizeof(Class *) * mixinc);
+
+    for (i=0 ; i<mixinc ; i++) {
+	mixins[i] = GetClassInOuterContext(interp, mixinv[i],
+		"may only mix in classes");
+	if (mixins[i] == NULL) {
+	    goto freeAndError;
+	}
+	if (TclOOIsReachable(oPtr->classPtr, mixins[i])) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "may not mix a class into itself", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "SELF_MIXIN", NULL);
+	    goto freeAndError;
+	}
+    }
+
+    TclOOClassSetMixins(interp, oPtr->classPtr, mixinc, mixins);
+    TclStackFree(interp, mixins);
+    return TCL_OK;
+
+  freeAndError:
+    TclStackFree(interp, mixins);
+    return TCL_ERROR;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ClassSuperGet, ClassSuperSet --
+ *	Implementation of the "superclass" slot accessors of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ClassSuperGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj;
+    Class *superPtr;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    }
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(superPtr, oPtr->classPtr->superclasses) {
+	Tcl_ListObjAppendElement(NULL, resultObj,
+		TclOOObjectName(interp, superPtr->thisPtr));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ClassSuperSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int superc, i, j;
+    Tcl_Obj **superv;
+    Class **superclasses, *superPtr;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"superclassList");
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    } else if (oPtr == oPtr->fPtr->objectCls->thisPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"may not modify the superclass of the root object", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    } else if (Tcl_ListObjGetElements(interp, objv[0], &superc,
+	    &superv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Allocate some working space.
+     */
+
+    superclasses = (Class **) ckalloc(sizeof(Class *) * superc);
+
+    /*
+     * Parse the arguments to get the class to use as superclasses.
+     *
+     * Note that zero classes is special, as it is equivalent to just the
+     * class of objects. [Bug 9d61624b3d]
+     */
+
+    if (superc == 0) {
+	superclasses = ckrealloc(superclasses, sizeof(Class *));
+	superclasses[0] = oPtr->fPtr->objectCls;
+	superc = 1;
+	if (TclOOIsReachable(oPtr->fPtr->classCls, oPtr->classPtr)) {
+	    superclasses[0] = oPtr->fPtr->classCls;
+	}
+    } else {
+	for (i=0 ; i<superc ; i++) {
+	    superclasses[i] = GetClassInOuterContext(interp, superv[i],
+		    "only a class can be a superclass");
+	    if (superclasses[i] == NULL) {
+		goto failedAfterAlloc;
+	    }
+	    for (j=0 ; j<i ; j++) {
+		if (superclasses[j] == superclasses[i]) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "class should only be a direct superclass once",
+			    -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OO", "REPETITIOUS",NULL);
+		    goto failedAfterAlloc;
+		}
+	    }
+	    if (TclOOIsReachable(oPtr->classPtr, superclasses[i])) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"attempt to form circular dependency graph", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OO", "CIRCULARITY", NULL);
+	    failedAfterAlloc:
+		ckfree(superclasses);
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    /*
+     * Install the list of superclasses into the class. Note that this also
+     * involves splicing the class out of the superclasses' subclass list that
+     * it used to be a member of and splicing it into the new superclasses'
+     * subclass list.
+     */
+
+    if (oPtr->classPtr->superclasses.num != 0) {
+	FOREACH(superPtr, oPtr->classPtr->superclasses) {
+	    TclOORemoveFromSubclasses(oPtr->classPtr, superPtr);
+	}
+	ckfree(oPtr->classPtr->superclasses.list);
+    }
+    oPtr->classPtr->superclasses.list = superclasses;
+    oPtr->classPtr->superclasses.num = superc;
+    FOREACH(superPtr, oPtr->classPtr->superclasses) {
+	TclOOAddToSubclasses(oPtr->classPtr, superPtr);
+    }
+    BumpGlobalEpoch(interp, oPtr->classPtr);
+
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ClassVarsGet, ClassVarsSet --
+ *	Implementation of the "variable" slot accessors of the "oo::define"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ClassVarsGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj, *variableObj;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    }
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(variableObj, oPtr->classPtr->variables) {
+	Tcl_ListObjAppendElement(NULL, resultObj, variableObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ClassVarsSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int varc;
+    Tcl_Obj **varv, *variableObj;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"filterList");
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    } else if (!oPtr->classPtr) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"attempt to misuse API", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "MONKEY_BUSINESS", NULL);
+	return TCL_ERROR;
+    } else if (Tcl_ListObjGetElements(interp, objv[0], &varc,
+	    &varv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<varc ; i++) {
+	const char *varName = TclGetString(varv[i]);
+
+	if (strstr(varName, "::") != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "invalid declared variable name \"%s\": must not %s",
+		    varName, "contain namespace separators"));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_DECLVAR", NULL);
+	    return TCL_ERROR;
+	}
+	if (Tcl_StringMatch(varName, "*(*)")) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "invalid declared variable name \"%s\": must not %s",
+		    varName, "refer to an array element"));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_DECLVAR", NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    for (i=0 ; i<varc ; i++) {
+	Tcl_IncrRefCount(varv[i]);
+    }
+    FOREACH(variableObj, oPtr->classPtr->variables) {
+	Tcl_DecrRefCount(variableObj);
+    }
+    if (i != varc) {
+	if (varc == 0) {
+	    ckfree(oPtr->classPtr->variables.list);
+	} else if (i) {
+	    oPtr->classPtr->variables.list = (Tcl_Obj **)
+		    ckrealloc((char *) oPtr->classPtr->variables.list,
+		    sizeof(Tcl_Obj *) * varc);
+	} else {
+	    oPtr->classPtr->variables.list = (Tcl_Obj **)
+		    ckalloc(sizeof(Tcl_Obj *) * varc);
+	}
+    }
+
+    oPtr->classPtr->variables.num = 0;
+    if (varc > 0) {
+	int created, n;
+	Tcl_HashTable uniqueTable;
+
+	Tcl_InitObjHashTable(&uniqueTable);
+	for (i=n=0 ; i<varc ; i++) {
+	    Tcl_CreateHashEntry(&uniqueTable, varv[i], &created);
+	    if (created) {
+		oPtr->classPtr->variables.list[n++] = varv[i];
+	    } else {
+		Tcl_DecrRefCount(varv[i]);
+	    }
+	}
+	oPtr->classPtr->variables.num = n;
+
+	/*
+	 * Shouldn't be necessary, but maintain num/list invariant.
+	 */
+
+	oPtr->classPtr->variables.list = (Tcl_Obj **)
+		ckrealloc((char *) oPtr->classPtr->variables.list,
+		sizeof(Tcl_Obj *) * n);
+	Tcl_DeleteHashTable(&uniqueTable);
+    }
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ObjectFilterGet, ObjectFilterSet --
+ *	Implementation of the "filter" slot accessors of the "oo::objdefine"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ObjFilterGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj, *filterObj;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(filterObj, oPtr->filters) {
+	Tcl_ListObjAppendElement(NULL, resultObj, filterObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ObjFilterSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int filterc;
+    Tcl_Obj **filterv;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"filterList");
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+    if (Tcl_ListObjGetElements(interp, objv[0], &filterc,
+	    &filterv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    TclOOObjectSetFilters(oPtr, filterc, filterv);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ObjectMixinGet, ObjectMixinSet --
+ *	Implementation of the "mixin" slot accessors of the "oo::objdefine"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ObjMixinGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj;
+    Class *mixinPtr;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(mixinPtr, oPtr->mixins) {
+	if (mixinPtr) {
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    TclOOObjectName(interp, mixinPtr->thisPtr));
+	}
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ObjMixinSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int mixinc;
+    Tcl_Obj **mixinv;
+    Class **mixins;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"mixinList");
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+    if (Tcl_ListObjGetElements(interp, objv[0], &mixinc,
+	    &mixinv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    mixins = TclStackAlloc(interp, sizeof(Class *) * mixinc);
+
+    for (i=0 ; i<mixinc ; i++) {
+	mixins[i] = GetClassInOuterContext(interp, mixinv[i],
+		"may only mix in classes");
+	if (mixins[i] == NULL) {
+	    TclStackFree(interp, mixins);
+	    return TCL_ERROR;
+	}
+    }
+
+    TclOOObjectSetMixins(oPtr, mixinc, mixins);
+    TclStackFree(interp, mixins);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ObjectVarsGet, ObjectVarsSet --
+ *	Implementation of the "variable" slot accessors of the "oo::objdefine"
+ *	command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+ObjVarsGet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    Tcl_Obj *resultObj, *variableObj;
+    int i;
+
+    if (Tcl_ObjectContextSkippedArgs(context) != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		NULL);
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(variableObj, oPtr->variables) {
+	Tcl_ListObjAppendElement(NULL, resultObj, variableObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+static int
+ObjVarsSet(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    Object *oPtr = (Object *) TclOOGetDefineCmdContext(interp);
+    int varc, i;
+    Tcl_Obj **varv, *variableObj;
+
+    if (Tcl_ObjectContextSkippedArgs(context)+1 != objc) {
+	Tcl_WrongNumArgs(interp, Tcl_ObjectContextSkippedArgs(context), objv,
+		"variableList");
+	return TCL_ERROR;
+    } else if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    objv += Tcl_ObjectContextSkippedArgs(context);
+    if (Tcl_ListObjGetElements(interp, objv[0], &varc,
+	    &varv) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<varc ; i++) {
+	const char *varName = TclGetString(varv[i]);
+
+	if (strstr(varName, "::") != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "invalid declared variable name \"%s\": must not %s",
+		    varName, "contain namespace separators"));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_DECLVAR", NULL);
+	    return TCL_ERROR;
+	}
+	if (Tcl_StringMatch(varName, "*(*)")) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "invalid declared variable name \"%s\": must not %s",
+		    varName, "refer to an array element"));
+	    Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_DECLVAR", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    for (i=0 ; i<varc ; i++) {
+	Tcl_IncrRefCount(varv[i]);
+    }
+
+    FOREACH(variableObj, oPtr->variables) {
+	Tcl_DecrRefCount(variableObj);
+    }
+    if (i != varc) {
+	if (varc == 0) {
+	    ckfree(oPtr->variables.list);
+	} else if (i) {
+	    oPtr->variables.list = (Tcl_Obj **)
+		    ckrealloc((char *) oPtr->variables.list,
+		    sizeof(Tcl_Obj *) * varc);
+	} else {
+	    oPtr->variables.list = (Tcl_Obj **)
+		    ckalloc(sizeof(Tcl_Obj *) * varc);
+	}
+    }
+    oPtr->variables.num = 0;
+    if (varc > 0) {
+	int created, n;
+	Tcl_HashTable uniqueTable;
+
+	Tcl_InitObjHashTable(&uniqueTable);
+	for (i=n=0 ; i<varc ; i++) {
+	    Tcl_CreateHashEntry(&uniqueTable, varv[i], &created);
+	    if (created) {
+		oPtr->variables.list[n++] = varv[i];
+	    } else {
+		Tcl_DecrRefCount(varv[i]);
+	    }
+	}
+	oPtr->variables.num = n;
+
+	/*
+	 * Shouldn't be necessary, but maintain num/list invariant.
+	 */
+
+	oPtr->variables.list = (Tcl_Obj **)
+		ckrealloc((char *) oPtr->variables.list,
+		sizeof(Tcl_Obj *) * n);
+	Tcl_DeleteHashTable(&uniqueTable);
+    }
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOInfo.c b/generic/tclOOInfo.c
new file mode 100644
index 0000000..76eaef5
--- /dev/null
+++ b/generic/tclOOInfo.c
@@ -0,0 +1,1530 @@
+/*
+ * tclOODefineCmds.c --
+ *
+ *	This file contains the implementation of the ::oo-related [info]
+ *	subcommands.
+ *
+ * Copyright (c) 2006-2011 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+static inline Class *	GetClassFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static Tcl_ObjCmdProc InfoObjectCallCmd;
+static Tcl_ObjCmdProc InfoObjectClassCmd;
+static Tcl_ObjCmdProc InfoObjectDefnCmd;
+static Tcl_ObjCmdProc InfoObjectFiltersCmd;
+static Tcl_ObjCmdProc InfoObjectForwardCmd;
+static Tcl_ObjCmdProc InfoObjectIsACmd;
+static Tcl_ObjCmdProc InfoObjectMethodsCmd;
+static Tcl_ObjCmdProc InfoObjectMethodTypeCmd;
+static Tcl_ObjCmdProc InfoObjectMixinsCmd;
+static Tcl_ObjCmdProc InfoObjectNsCmd;
+static Tcl_ObjCmdProc InfoObjectVarsCmd;
+static Tcl_ObjCmdProc InfoObjectVariablesCmd;
+static Tcl_ObjCmdProc InfoClassCallCmd;
+static Tcl_ObjCmdProc InfoClassConstrCmd;
+static Tcl_ObjCmdProc InfoClassDefnCmd;
+static Tcl_ObjCmdProc InfoClassDestrCmd;
+static Tcl_ObjCmdProc InfoClassFiltersCmd;
+static Tcl_ObjCmdProc InfoClassForwardCmd;
+static Tcl_ObjCmdProc InfoClassInstancesCmd;
+static Tcl_ObjCmdProc InfoClassMethodsCmd;
+static Tcl_ObjCmdProc InfoClassMethodTypeCmd;
+static Tcl_ObjCmdProc InfoClassMixinsCmd;
+static Tcl_ObjCmdProc InfoClassSubsCmd;
+static Tcl_ObjCmdProc InfoClassSupersCmd;
+static Tcl_ObjCmdProc InfoClassVariablesCmd;
+
+/*
+ * List of commands that are used to implement the [info object] subcommands.
+ */
+
+static const EnsembleImplMap infoObjectCmds[] = {
+    {"call",	   InfoObjectCallCmd,	    TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"class",	   InfoObjectClassCmd,	    TclCompileInfoObjectClassCmd, NULL, NULL, 0},
+    {"definition", InfoObjectDefnCmd,	    TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"filters",	   InfoObjectFiltersCmd,    TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"forward",	   InfoObjectForwardCmd,    TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"isa",	   InfoObjectIsACmd,	    TclCompileInfoObjectIsACmd, NULL, NULL, 0},
+    {"methods",	   InfoObjectMethodsCmd,    TclCompileBasicMin1ArgCmd, NULL, NULL, 0},
+    {"methodtype", InfoObjectMethodTypeCmd, TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"mixins",	   InfoObjectMixinsCmd,	    TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"namespace",  InfoObjectNsCmd,	    TclCompileInfoObjectNamespaceCmd, NULL, NULL, 0},
+    {"variables",  InfoObjectVariablesCmd,  TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"vars",	   InfoObjectVarsCmd,	    TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+    {NULL, NULL, NULL, NULL, NULL, 0}
+};
+
+/*
+ * List of commands that are used to implement the [info class] subcommands.
+ */
+
+static const EnsembleImplMap infoClassCmds[] = {
+    {"call",	     InfoClassCallCmd,		TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"constructor",  InfoClassConstrCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"definition",   InfoClassDefnCmd,		TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"destructor",   InfoClassDestrCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"filters",	     InfoClassFiltersCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"forward",	     InfoClassForwardCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"instances",    InfoClassInstancesCmd,	TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+    {"methods",	     InfoClassMethodsCmd,	TclCompileBasicMin1ArgCmd, NULL, NULL, 0},
+    {"methodtype",   InfoClassMethodTypeCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+    {"mixins",	     InfoClassMixinsCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"subclasses",   InfoClassSubsCmd,		TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+    {"superclasses", InfoClassSupersCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {"variables",    InfoClassVariablesCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+    {NULL, NULL, NULL, NULL, NULL, 0}
+};
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOInitInfo --
+ *
+ *	Adjusts the Tcl core [info] command to contain subcommands ("object"
+ *	and "class") for introspection of objects and classes.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOInitInfo(
+    Tcl_Interp *interp)
+{
+    Tcl_Command infoCmd;
+    Tcl_Obj *mapDict;
+
+    /*
+     * Build the ensembles used to implement [info object] and [info class].
+     */
+
+    TclMakeEnsemble(interp, "::oo::InfoObject", infoObjectCmds);
+    TclMakeEnsemble(interp, "::oo::InfoClass", infoClassCmds);
+
+    /*
+     * Install into the master [info] ensemble.
+     */
+
+    infoCmd = Tcl_FindCommand(interp, "info", NULL, TCL_GLOBAL_ONLY);
+    Tcl_GetEnsembleMappingDict(NULL, infoCmd, &mapDict);
+    Tcl_DictObjPut(NULL, mapDict, Tcl_NewStringObj("object", -1),
+	    Tcl_NewStringObj("::oo::InfoObject", -1));
+    Tcl_DictObjPut(NULL, mapDict, Tcl_NewStringObj("class", -1),
+	    Tcl_NewStringObj("::oo::InfoClass", -1));
+    Tcl_SetEnsembleMappingDict(interp, infoCmd, mapDict);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * GetClassFromObj --
+ *
+ *	How to correctly get a class from a Tcl_Obj. Just a wrapper round
+ *	Tcl_GetObjectFromObj, but this is an idiom that was used heavily.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static inline Class *
+GetClassFromObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr)
+{
+    Object *oPtr = (Object *) Tcl_GetObjectFromObj(interp, objPtr);
+
+    if (oPtr == NULL) {
+	return NULL;
+    }
+    if (oPtr->classPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" is not a class", TclGetString(objPtr)));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "CLASS",
+		TclGetString(objPtr), NULL);
+	return NULL;
+    }
+    return oPtr->classPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectClassCmd --
+ *
+ *	Implements [info object class $objName ?$className?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectClassCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+
+    if (objc != 2 && objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName ?className?");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	Tcl_SetObjResult(interp,
+		TclOOObjectName(interp, oPtr->selfCls->thisPtr));
+	return TCL_OK;
+    } else {
+	Class *mixinPtr, *o2clsPtr;
+	int i;
+
+	o2clsPtr = GetClassFromObj(interp, objv[2]);
+	if (o2clsPtr == NULL) {
+	    return TCL_ERROR;
+	}
+
+	FOREACH(mixinPtr, oPtr->mixins) {
+	    if (!mixinPtr) {
+		continue;
+	    }
+	    if (TclOOIsReachable(o2clsPtr, mixinPtr)) {
+		Tcl_SetObjResult(interp, Tcl_NewIntObj(1));
+		return TCL_OK;
+	    }
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(
+		TclOOIsReachable(o2clsPtr, oPtr->selfCls)));
+	return TCL_OK;
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectDefnCmd --
+ *
+ *	Implements [info object definition $objName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectDefnCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    Tcl_HashEntry *hPtr;
+    Proc *procPtr;
+    CompiledLocal *localPtr;
+    Tcl_Obj *resultObjs[2];
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName methodName");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (!oPtr->methodsPtr) {
+	goto unknownMethod;
+    }
+    hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char *) objv[2]);
+    if (hPtr == NULL) {
+    unknownMethod:
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    procPtr = TclOOGetProcFromMethod(Tcl_GetHashValue(hPtr));
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"definition not available for this kind of method", -1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+
+    resultObjs[0] = Tcl_NewObj();
+    for (localPtr=procPtr->firstLocalPtr; localPtr!=NULL;
+	    localPtr=localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)) {
+	    Tcl_Obj *argObj;
+
+	    argObj = Tcl_NewObj();
+	    Tcl_ListObjAppendElement(NULL, argObj,
+		    Tcl_NewStringObj(localPtr->name, -1));
+	    if (localPtr->defValuePtr != NULL) {
+		Tcl_ListObjAppendElement(NULL, argObj, localPtr->defValuePtr);
+	    }
+	    Tcl_ListObjAppendElement(NULL, resultObjs[0], argObj);
+	}
+    }
+    resultObjs[1] = TclOOGetMethodBody(Tcl_GetHashValue(hPtr));
+    Tcl_SetObjResult(interp, Tcl_NewListObj(2, resultObjs));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectFiltersCmd --
+ *
+ *	Implements [info object filters $objName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectFiltersCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int i;
+    Tcl_Obj *filterObj, *resultObj;
+    Object *oPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    resultObj = Tcl_NewObj();
+
+    FOREACH(filterObj, oPtr->filters) {
+	Tcl_ListObjAppendElement(NULL, resultObj, filterObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectForwardCmd --
+ *
+ *	Implements [info object forward $objName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectForwardCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    Tcl_HashEntry *hPtr;
+    Tcl_Obj *prefixObj;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName methodName");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (!oPtr->methodsPtr) {
+	goto unknownMethod;
+    }
+    hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char *) objv[2]);
+    if (hPtr == NULL) {
+    unknownMethod:
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    prefixObj = TclOOGetFwdFromMethod(Tcl_GetHashValue(hPtr));
+    if (prefixObj == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"prefix argument list not available for this kind of method",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, prefixObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectIsACmd --
+ *
+ *	Implements [info object isa $category $objName ...]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectIsACmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const char *const categories[] = {
+	"class", "metaclass", "mixin", "object", "typeof", NULL
+    };
+    enum IsACats {
+	IsClass, IsMetaclass, IsMixin, IsObject, IsType
+    };
+    Object *oPtr, *o2Ptr;
+    int idx, i, result = 0;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "category objName ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], categories, "category", 0,
+	    &idx) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now we know what test we are doing, we can check we've got the right
+     * number of arguments.
+     */
+
+    switch ((enum IsACats) idx) {
+    case IsObject:
+    case IsClass:
+    case IsMetaclass:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "objName");
+	    return TCL_ERROR;
+	}
+	break;
+    case IsMixin:
+    case IsType:
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "objName className");
+	    return TCL_ERROR;
+	}
+	break;
+    }
+
+    /*
+     * Perform the check. Note that we can guarantee that we will not fail
+     * from here on; "failures" result in a false-TCL_OK result.
+     */
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[2]);
+    if (oPtr == NULL) {
+	goto failPrecondition;
+    }
+
+    switch ((enum IsACats) idx) {
+    case IsObject:
+	result = 1;
+	break;
+    case IsClass:
+	result = (oPtr->classPtr != NULL);
+	break;
+    case IsMetaclass:
+	if (oPtr->classPtr != NULL) {
+	    result = TclOOIsReachable(TclOOGetFoundation(interp)->classCls,
+		    oPtr->classPtr);
+	}
+	break;
+    case IsMixin:
+	o2Ptr = (Object *) Tcl_GetObjectFromObj(interp, objv[3]);
+	if (o2Ptr == NULL) {
+	    goto failPrecondition;
+	}
+	if (o2Ptr->classPtr != NULL) {
+	    Class *mixinPtr;
+
+	    FOREACH(mixinPtr, oPtr->mixins) {
+		if (!mixinPtr) {
+		    continue;
+		}
+		if (TclOOIsReachable(o2Ptr->classPtr, mixinPtr)) {
+		    result = 1;
+		    break;
+		}
+	    }
+	}
+	break;
+    case IsType:
+	o2Ptr = (Object *) Tcl_GetObjectFromObj(interp, objv[3]);
+	if (o2Ptr == NULL) {
+	    goto failPrecondition;
+	}
+	if (o2Ptr->classPtr != NULL) {
+	    result = TclOOIsReachable(o2Ptr->classPtr, oPtr->selfCls);
+	}
+	break;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(result));
+    return TCL_OK;
+
+  failPrecondition:
+    Tcl_ResetResult(interp);
+    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(0));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectMethodsCmd --
+ *
+ *	Implements [info object methods $objName ?$option ...?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectMethodsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    int flag = PUBLIC_METHOD, recurse = 0;
+    FOREACH_HASH_DECLS;
+    Tcl_Obj *namePtr, *resultObj;
+    Method *mPtr;
+    static const char *const options[] = {
+	"-all", "-localprivate", "-private", NULL
+    };
+    enum Options {
+	OPT_ALL, OPT_LOCALPRIVATE, OPT_PRIVATE
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName ?-option value ...?");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (objc != 2) {
+	int i, idx;
+
+	for (i=2 ; i<objc ; i++) {
+	    if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		    &idx) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum Options) idx) {
+	    case OPT_ALL:
+		recurse = 1;
+		break;
+	    case OPT_LOCALPRIVATE:
+		flag = PRIVATE_METHOD;
+		break;
+	    case OPT_PRIVATE:
+		flag = 0;
+		break;
+	    }
+	}
+    }
+
+    resultObj = Tcl_NewObj();
+    if (recurse) {
+	const char **names;
+	int i, numNames = TclOOGetSortedMethodList(oPtr, flag, &names);
+
+	for (i=0 ; i<numNames ; i++) {
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(names[i], -1));
+	}
+	if (numNames > 0) {
+	    ckfree(names);
+	}
+    } else if (oPtr->methodsPtr) {
+	FOREACH_HASH(namePtr, mPtr, oPtr->methodsPtr) {
+	    if (mPtr->typePtr != NULL && (mPtr->flags & flag) == flag) {
+		Tcl_ListObjAppendElement(NULL, resultObj, namePtr);
+	    }
+	}
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectMethodTypeCmd --
+ *
+ *	Implements [info object methodtype $objName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectMethodTypeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    Tcl_HashEntry *hPtr;
+    Method *mPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName methodName");
+	return TCL_ERROR;
+    }
+
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (!oPtr->methodsPtr) {
+	goto unknownMethod;
+    }
+    hPtr = Tcl_FindHashEntry(oPtr->methodsPtr, (char *) objv[2]);
+    if (hPtr == NULL) {
+    unknownMethod:
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    mPtr = Tcl_GetHashValue(hPtr);
+    if (mPtr->typePtr == NULL) {
+	/*
+	 * Special entry for visibility control: pretend the method doesnt
+	 * exist.
+	 */
+
+	goto unknownMethod;
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(mPtr->typePtr->name, -1));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectMixinsCmd --
+ *
+ *	Implements [info object mixins $objName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectMixinsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *mixinPtr;
+    Object *oPtr;
+    Tcl_Obj *resultObj;
+    int i;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(mixinPtr, oPtr->mixins) {
+	if (!mixinPtr) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj,
+		TclOOObjectName(interp, mixinPtr->thisPtr));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectNsCmd --
+ *
+ *	Implements [info object namespace $objName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectNsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp,
+	    Tcl_NewStringObj(oPtr->namespacePtr->fullName, -1));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectVariablesCmd --
+ *
+ *	Implements [info object variables $objName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectVariablesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    Tcl_Obj *variableObj, *resultObj;
+    int i;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(variableObj, oPtr->variables) {
+	Tcl_ListObjAppendElement(NULL, resultObj, variableObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectVarsCmd --
+ *
+ *	Implements [info object vars $objName ?$pattern?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectVarsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    const char *pattern = NULL;
+    FOREACH_HASH_DECLS;
+    VarInHash *vihPtr;
+    Tcl_Obj *nameObj, *resultObj;
+
+    if (objc != 2 && objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName ?pattern?");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (objc == 3) {
+	pattern = TclGetString(objv[2]);
+    }
+    resultObj = Tcl_NewObj();
+
+    /*
+     * Extract the information we need from the object's namespace's table of
+     * variables. Note that this involves horrific knowledge of the guts of
+     * tclVar.c, so we can't leverage our hash-iteration macros properly.
+     */
+
+    FOREACH_HASH_VALUE(vihPtr,
+	    &((Namespace *) oPtr->namespacePtr)->varTable.table) {
+	nameObj = vihPtr->entry.key.objPtr;
+
+	if (TclIsVarUndefined(&vihPtr->var)
+		|| !TclIsVarNamespaceVar(&vihPtr->var)) {
+	    continue;
+	}
+	if (pattern != NULL
+		&& !Tcl_StringMatch(TclGetString(nameObj), pattern)) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj, nameObj);
+    }
+
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassConstrCmd --
+ *
+ *	Implements [info class constructor $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassConstrCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Proc *procPtr;
+    CompiledLocal *localPtr;
+    Tcl_Obj *resultObjs[2];
+    Class *clsPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (clsPtr->constructorPtr == NULL) {
+	return TCL_OK;
+    }
+    procPtr = TclOOGetProcFromMethod(clsPtr->constructorPtr);
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"definition not available for this kind of method", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "METHOD_TYPE", NULL);
+	return TCL_ERROR;
+    }
+
+    resultObjs[0] = Tcl_NewObj();
+    for (localPtr=procPtr->firstLocalPtr; localPtr!=NULL;
+	    localPtr=localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)) {
+	    Tcl_Obj *argObj;
+
+	    argObj = Tcl_NewObj();
+	    Tcl_ListObjAppendElement(NULL, argObj,
+		    Tcl_NewStringObj(localPtr->name, -1));
+	    if (localPtr->defValuePtr != NULL) {
+		Tcl_ListObjAppendElement(NULL, argObj, localPtr->defValuePtr);
+	    }
+	    Tcl_ListObjAppendElement(NULL, resultObjs[0], argObj);
+	}
+    }
+    resultObjs[1] = TclOOGetMethodBody(clsPtr->constructorPtr);
+    Tcl_SetObjResult(interp, Tcl_NewListObj(2, resultObjs));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassDefnCmd --
+ *
+ *	Implements [info class definition $clsName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassDefnCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_HashEntry *hPtr;
+    Proc *procPtr;
+    CompiledLocal *localPtr;
+    Tcl_Obj *resultObjs[2];
+    Class *clsPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className methodName");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    hPtr = Tcl_FindHashEntry(&clsPtr->classMethods, (char *) objv[2]);
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    procPtr = TclOOGetProcFromMethod(Tcl_GetHashValue(hPtr));
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"definition not available for this kind of method", -1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+
+    resultObjs[0] = Tcl_NewObj();
+    for (localPtr=procPtr->firstLocalPtr; localPtr!=NULL;
+	    localPtr=localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)) {
+	    Tcl_Obj *argObj;
+
+	    argObj = Tcl_NewObj();
+	    Tcl_ListObjAppendElement(NULL, argObj,
+		    Tcl_NewStringObj(localPtr->name, -1));
+	    if (localPtr->defValuePtr != NULL) {
+		Tcl_ListObjAppendElement(NULL, argObj, localPtr->defValuePtr);
+	    }
+	    Tcl_ListObjAppendElement(NULL, resultObjs[0], argObj);
+	}
+    }
+    resultObjs[1] = TclOOGetMethodBody(Tcl_GetHashValue(hPtr));
+    Tcl_SetObjResult(interp, Tcl_NewListObj(2, resultObjs));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassDestrCmd --
+ *
+ *	Implements [info class destructor $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassDestrCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Proc *procPtr;
+    Class *clsPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (clsPtr->destructorPtr == NULL) {
+	return TCL_OK;
+    }
+    procPtr = TclOOGetProcFromMethod(clsPtr->destructorPtr);
+    if (procPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"definition not available for this kind of method", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "METHOD_TYPE", NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, TclOOGetMethodBody(clsPtr->destructorPtr));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassFiltersCmd --
+ *
+ *	Implements [info class filters $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassFiltersCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int i;
+    Tcl_Obj *filterObj, *resultObj;
+    Class *clsPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(filterObj, clsPtr->filters) {
+	Tcl_ListObjAppendElement(NULL, resultObj, filterObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassForwardCmd --
+ *
+ *	Implements [info class forward $clsName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassForwardCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_HashEntry *hPtr;
+    Tcl_Obj *prefixObj;
+    Class *clsPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className methodName");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    hPtr = Tcl_FindHashEntry(&clsPtr->classMethods, (char *) objv[2]);
+    if (hPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    prefixObj = TclOOGetFwdFromMethod(Tcl_GetHashValue(hPtr));
+    if (prefixObj == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"prefix argument list not available for this kind of method",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_SetObjResult(interp, prefixObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassInstancesCmd --
+ *
+ *	Implements [info class instances $clsName ?$pattern?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassInstancesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    Class *clsPtr;
+    int i;
+    const char *pattern = NULL;
+    Tcl_Obj *resultObj;
+
+    if (objc != 2 && objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className ?pattern?");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (objc == 3) {
+	pattern = TclGetString(objv[2]);
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(oPtr, clsPtr->instances) {
+	Tcl_Obj *tmpObj = TclOOObjectName(interp, oPtr);
+
+	if (pattern && !Tcl_StringMatch(TclGetString(tmpObj), pattern)) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj, tmpObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassMethodsCmd --
+ *
+ *	Implements [info class methods $clsName ?-private?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassMethodsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int flag = PUBLIC_METHOD, recurse = 0;
+    Tcl_Obj *namePtr, *resultObj;
+    Method *mPtr;
+    Class *clsPtr;
+    static const char *const options[] = {
+	"-all", "-localprivate", "-private", NULL
+    };
+    enum Options {
+	OPT_ALL, OPT_LOCALPRIVATE, OPT_PRIVATE
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className ?-option value ...?");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (objc != 2) {
+	int i, idx;
+
+	for (i=2 ; i<objc ; i++) {
+	    if (Tcl_GetIndexFromObj(interp, objv[i], options, "option", 0,
+		    &idx) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum Options) idx) {
+	    case OPT_ALL:
+		recurse = 1;
+		break;
+	    case OPT_LOCALPRIVATE:
+		flag = PRIVATE_METHOD;
+		break;
+	    case OPT_PRIVATE:
+		flag = 0;
+		break;
+	    }
+	}
+    }
+
+    resultObj = Tcl_NewObj();
+    if (recurse) {
+	const char **names;
+	int i, numNames = TclOOGetSortedClassMethodList(clsPtr, flag, &names);
+
+	for (i=0 ; i<numNames ; i++) {
+	    Tcl_ListObjAppendElement(NULL, resultObj,
+		    Tcl_NewStringObj(names[i], -1));
+	}
+	if (numNames > 0) {
+	    ckfree(names);
+	}
+    } else {
+	FOREACH_HASH_DECLS;
+
+	FOREACH_HASH(namePtr, mPtr, &clsPtr->classMethods) {
+	    if (mPtr->typePtr != NULL && (mPtr->flags & flag) == flag) {
+		Tcl_ListObjAppendElement(NULL, resultObj, namePtr);
+	    }
+	}
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassMethodTypeCmd --
+ *
+ *	Implements [info class methodtype $clsName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassMethodTypeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_HashEntry *hPtr;
+    Method *mPtr;
+    Class *clsPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className methodName");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    hPtr = Tcl_FindHashEntry(&clsPtr->classMethods, (char *) objv[2]);
+    if (hPtr == NULL) {
+    unknownMethod:
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown method \"%s\"", TclGetString(objv[2])));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "METHOD",
+		TclGetString(objv[2]), NULL);
+	return TCL_ERROR;
+    }
+    mPtr = Tcl_GetHashValue(hPtr);
+    if (mPtr->typePtr == NULL) {
+	/*
+	 * Special entry for visibility control: pretend the method doesnt
+	 * exist.
+	 */
+
+	goto unknownMethod;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(mPtr->typePtr->name, -1));
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassMixinsCmd --
+ *
+ *	Implements [info class mixins $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassMixinsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *clsPtr, *mixinPtr;
+    Tcl_Obj *resultObj;
+    int i;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(mixinPtr, clsPtr->mixins) {
+	if (!mixinPtr) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj,
+		TclOOObjectName(interp, mixinPtr->thisPtr));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassSubsCmd --
+ *
+ *	Implements [info class subclasses $clsName ?$pattern?]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassSubsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *clsPtr, *subclassPtr;
+    Tcl_Obj *resultObj;
+    int i;
+    const char *pattern = NULL;
+
+    if (objc != 2 && objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className ?pattern?");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (objc == 3) {
+	pattern = TclGetString(objv[2]);
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(subclassPtr, clsPtr->subclasses) {
+	Tcl_Obj *tmpObj = TclOOObjectName(interp, subclassPtr->thisPtr);
+
+	if (pattern && !Tcl_StringMatch(TclGetString(tmpObj), pattern)) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj, tmpObj);
+    }
+    FOREACH(subclassPtr, clsPtr->mixinSubs) {
+	Tcl_Obj *tmpObj = TclOOObjectName(interp, subclassPtr->thisPtr);
+
+	if (pattern && !Tcl_StringMatch(TclGetString(tmpObj), pattern)) {
+	    continue;
+	}
+	Tcl_ListObjAppendElement(NULL, resultObj, tmpObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassSupersCmd --
+ *
+ *	Implements [info class superclasses $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassSupersCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *clsPtr, *superPtr;
+    Tcl_Obj *resultObj;
+    int i;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(superPtr, clsPtr->superclasses) {
+	Tcl_ListObjAppendElement(NULL, resultObj,
+		TclOOObjectName(interp, superPtr->thisPtr));
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassVariablesCmd --
+ *
+ *	Implements [info class variables $clsName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassVariablesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *clsPtr;
+    Tcl_Obj *variableObj, *resultObj;
+    int i;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resultObj = Tcl_NewObj();
+    FOREACH(variableObj, clsPtr->variables) {
+	Tcl_ListObjAppendElement(NULL, resultObj, variableObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoObjectCallCmd --
+ *
+ *	Implements [info object call $objName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoObjectCallCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Object *oPtr;
+    CallContext *contextPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "objName methodName");
+	return TCL_ERROR;
+    }
+    oPtr = (Object *) Tcl_GetObjectFromObj(interp, objv[1]);
+    if (oPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the call context and render its call chain.
+     */
+
+    contextPtr = TclOOGetCallContext(oPtr, objv[2], PUBLIC_METHOD, NULL);
+    if (contextPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"cannot construct any call chain", -1));
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp,
+	    TclOORenderCallChain(interp, contextPtr->callPtr));
+    TclOODeleteContext(contextPtr);
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InfoClassCallCmd --
+ *
+ *	Implements [info class call $clsName $methodName]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InfoClassCallCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Class *clsPtr;
+    CallChain *callPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "className methodName");
+	return TCL_ERROR;
+    }
+    clsPtr = GetClassFromObj(interp, objv[1]);
+    if (clsPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get an render the stereotypical call chain.
+     */
+
+    callPtr = TclOOGetStereotypeCallChain(clsPtr, objv[2], PUBLIC_METHOD);
+    if (callPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"cannot construct any call chain", -1));
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, TclOORenderCallChain(interp, callPtr));
+    TclOODeleteChain(callPtr);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOInt.h b/generic/tclOOInt.h
new file mode 100644
index 0000000..476446d
--- /dev/null
+++ b/generic/tclOOInt.h
@@ -0,0 +1,610 @@
+/*
+ * tclOOInt.h --
+ *
+ *	This file contains the structure definitions and some of the function
+ *	declarations for the object-system (NB: not Tcl_Obj, but ::oo).
+ *
+ * Copyright (c) 2006-2012 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef TCL_OO_INTERNAL_H
+#define TCL_OO_INTERNAL_H 1
+
+#include "tclInt.h"
+#include "tclOO.h"
+
+/*
+ * Hack to make things work with Objective C. Note that ObjC isn't really
+ * supported, but we don't want to to be actively hostile to it. [Bug 2163447]
+ */
+
+#ifdef __OBJC__
+#define Class	TclOOClass
+#define Object	TclOOObject
+#endif /* __OBJC__ */
+
+/*
+ * Forward declarations.
+ */
+
+struct CallChain;
+struct Class;
+struct Foundation;
+struct Object;
+
+/*
+ * The data that needs to be stored per method. This record is used to collect
+ * information about all sorts of methods, including forwards, constructors
+ * and destructors.
+ */
+
+typedef struct Method {
+    const Tcl_MethodType *typePtr;
+				/* The type of method. If NULL, this is a
+				 * special flag record which is just used for
+				 * the setting of the flags field. */
+    int refCount;
+    ClientData clientData;	/* Type-specific data. */
+    Tcl_Obj *namePtr;		/* Name of the method. */
+    struct Object *declaringObjectPtr;
+				/* The object that declares this method, or
+				 * NULL if it was declared by a class. */
+    struct Class *declaringClassPtr;
+				/* The class that declares this method, or
+				 * NULL if it was declared directly on an
+				 * object. */
+    int flags;			/* Assorted flags. Includes whether this
+				 * method is public/exported or not. */
+} Method;
+
+/*
+ * Pre- and post-call callbacks, to allow procedure-like methods to be fine
+ * tuned in their behaviour.
+ */
+
+typedef int (TclOO_PreCallProc)(ClientData clientData, Tcl_Interp *interp,
+	Tcl_ObjectContext context, Tcl_CallFrame *framePtr, int *isFinished);
+typedef int (TclOO_PostCallProc)(ClientData clientData, Tcl_Interp *interp,
+	Tcl_ObjectContext context, Tcl_Namespace *namespacePtr, int result);
+typedef void (TclOO_PmCDDeleteProc)(ClientData clientData);
+typedef ClientData (TclOO_PmCDCloneProc)(ClientData clientData);
+
+/*
+ * Procedure-like methods have the following extra information.
+ */
+
+typedef struct ProcedureMethod {
+    int version;		/* Version of this structure. Currently must
+				 * be 0. */
+    Proc *procPtr;		/* Core of the implementation of the method;
+				 * includes the argument definition and the
+				 * body bytecodes. */
+    int flags;			/* Flags to control features. */
+    int refCount;
+    ClientData clientData;
+    TclOO_PmCDDeleteProc *deleteClientdataProc;
+    TclOO_PmCDCloneProc *cloneClientdataProc;
+    ProcErrorProc *errProc;	/* Replacement error handler. */
+    TclOO_PreCallProc *preCallProc;
+				/* Callback to allow for additional setup
+				 * before the method executes. */
+    TclOO_PostCallProc *postCallProc;
+				/* Callback to allow for additional cleanup
+				 * after the method executes. */
+    GetFrameInfoValueProc *gfivProc;
+				/* Callback to allow for fine tuning of how
+				 * the method reports itself. */
+} ProcedureMethod;
+
+#define TCLOO_PROCEDURE_METHOD_VERSION 0
+
+/*
+ * Flags for use in a ProcedureMethod.
+ *
+ * When the USE_DECLARER_NS flag is set, the method will use the namespace of
+ * the object or class that declared it (or the clone of it, if it was from
+ * such that the implementation of the method came to the particular use)
+ * instead of the namespace of the object on which the method was invoked.
+ * This flag must be distinct from all others that are associated with
+ * methods.
+ */
+
+#define USE_DECLARER_NS		0x80
+
+/*
+ * Forwarded methods have the following extra information.
+ */
+
+typedef struct ForwardMethod {
+    Tcl_Obj *prefixObj;		/* The list of values to use to replace the
+				 * object and method name with. Will be a
+				 * non-empty list. */
+} ForwardMethod;
+
+/*
+ * Helper definitions that declare a "list" array. The two varieties are
+ * either optimized for simplicity (in the case that the whole array is
+ * typically assigned at once) or efficiency (in the case that the array is
+ * expected to be expanded over time). These lists are designed to be iterated
+ * over with the help of the FOREACH macro (see later in this file).
+ *
+ * The "num" field always counts the number of listType_t elements used in the
+ * "list" field. When a "size" field exists, it describes how many elements
+ * are present in the list; when absent, exactly "num" elements are present.
+ */
+
+#define LIST_STATIC(listType_t) \
+    struct { int num; listType_t *list; }
+#define LIST_DYNAMIC(listType_t) \
+    struct { int num, size; listType_t *list; }
+
+/*
+ * Now, the definition of what an object actually is.
+ */
+
+typedef struct Object {
+    struct Foundation *fPtr;	/* The basis for the object system. Putting
+				 * this here allows the avoidance of quite a
+				 * lot of hash lookups on the critical path
+				 * for object invokation and creation. */
+    Tcl_Namespace *namespacePtr;/* This object's tame namespace. */
+    Tcl_Command command;	/* Reference to this object's public
+				 * command. */
+    Tcl_Command myCommand;	/* Reference to this object's internal
+				 * command. */
+    struct Class *selfCls;	/* This object's class. */
+    Tcl_HashTable *methodsPtr;	/* Object-local Tcl_Obj (method name) to
+				 * Method* mapping. */
+    LIST_STATIC(struct Class *) mixins;
+				/* Classes mixed into this object. */
+    LIST_STATIC(Tcl_Obj *) filters;
+				/* List of filter names. */
+    struct Class *classPtr;	/* All classes have this non-NULL; it points
+				 * to the class structure. Everything else has
+				 * this NULL. */
+    int refCount;		/* Number of strong references to this object.
+				 * Note that there may be many more weak
+				 * references; this mechanism is there to
+				 * avoid Tcl_Preserve. */
+    int flags;
+    int creationEpoch;		/* Unique value to make comparisons of objects
+				 * easier. */
+    int epoch;			/* Per-object epoch, incremented when the way
+				 * an object should resolve call chains is
+				 * changed. */
+    Tcl_HashTable *metadataPtr;	/* Mapping from pointers to metadata type to
+				 * the ClientData values that are the values
+				 * of each piece of attached metadata. This
+				 * field starts out as NULL and is only
+				 * allocated if metadata is attached. */
+    Tcl_Obj *cachedNameObj;	/* Cache of the name of the object. */
+    Tcl_HashTable *chainCache;	/* Place to keep unused contexts. This table
+				 * is indexed by method name as Tcl_Obj. */
+    Tcl_ObjectMapMethodNameProc *mapMethodNameProc;
+				/* Function to allow remapping of method
+				 * names. For itcl-ng. */
+    LIST_STATIC(Tcl_Obj *) variables;
+} Object;
+
+#define OBJECT_DELETED	1	/* Flag to say that an object has been
+				 * destroyed. */
+#define DESTRUCTOR_CALLED 2	/* Flag to say that the destructor has been
+				 * called. */
+#define CLASS_GONE	4	/* Indicates that the class of this object has
+				 * been deleted, and so the object should not
+				 * attempt to remove itself from its class. */
+#define ROOT_OBJECT 0x1000	/* Flag to say that this object is the root of
+				 * the class hierarchy and should be treated
+				 * specially during teardown. */
+#define FILTER_HANDLING 0x2000	/* Flag set when the object is processing a
+				 * filter; when set, filters are *not*
+				 * processed on the object, preventing nasty
+				 * recursive filtering problems. */
+#define USE_CLASS_CACHE 0x4000	/* Flag set to say that the object is a pure
+				 * instance of the class, and has had nothing
+				 * added that changes the dispatch chain (i.e.
+				 * no methods, mixins, or filters. */
+#define ROOT_CLASS 0x8000	/* Flag to say that this object is the root
+				 * class of classes, and should be treated
+				 * specially during teardown (and in a few
+				 * other spots). */
+#define FORCE_UNKNOWN 0x10000	/* States that we are *really* looking up the
+				 * unknown method handler at that point. */
+
+/*
+ * And the definition of a class. Note that every class also has an associated
+ * object, through which it is manipulated.
+ */
+
+typedef struct Class {
+    Object *thisPtr;		/* Reference to the object associated with
+				 * this class. */
+    int refCount;		/* Number of strong references to this class.
+				 * Weak references are not counted; the
+				 * purpose of this is to avoid Tcl_Preserve as
+				 * that is quite slow. */
+    int flags;			/* Assorted flags. */
+    LIST_STATIC(struct Class *) superclasses;
+				/* List of superclasses, used for generation
+				 * of method call chains. */
+    LIST_DYNAMIC(struct Class *) subclasses;
+				/* List of subclasses, used to ensure deletion
+				 * of dependent entities happens properly when
+				 * the class itself is deleted. */
+    LIST_DYNAMIC(Object *) instances;
+				/* List of instances, used to ensure deletion
+				 * of dependent entities happens properly when
+				 * the class itself is deleted. */
+    LIST_STATIC(Tcl_Obj *) filters;
+				/* List of filter names, used for generation
+				 * of method call chains. */
+    LIST_STATIC(struct Class *) mixins;
+				/* List of mixin classes, used for generation
+				 * of method call chains. */
+    LIST_DYNAMIC(struct Class *) mixinSubs;
+				/* List of classes that this class is mixed
+				 * into, used to ensure deletion of dependent
+				 * entities happens properly when the class
+				 * itself is deleted. */
+    Tcl_HashTable classMethods;	/* Hash table of all methods. Hash maps from
+				 * the (Tcl_Obj*) method name to the (Method*)
+				 * method record. */
+    Method *constructorPtr;	/* Method record of the class constructor (if
+				 * any). */
+    Method *destructorPtr;	/* Method record of the class destructor (if
+				 * any). */
+    Tcl_HashTable *metadataPtr;	/* Mapping from pointers to metadata type to
+				 * the ClientData values that are the values
+				 * of each piece of attached metadata. This
+				 * field starts out as NULL and is only
+				 * allocated if metadata is attached. */
+    struct CallChain *constructorChainPtr;
+    struct CallChain *destructorChainPtr;
+    Tcl_HashTable *classChainCache;
+				/* Places where call chains are stored. For
+				 * constructors, the class chain is always
+				 * used. For destructors and ordinary methods,
+				 * the class chain is only used when the
+				 * object doesn't override with its own mixins
+				 * (and filters and method implementations for
+				 * when getting method chains). */
+    LIST_STATIC(Tcl_Obj *) variables;
+} Class;
+
+/*
+ * The foundation of the object system within an interpreter contains
+ * references to the key classes and namespaces, together with a few other
+ * useful bits and pieces. Probably ought to eventually go in the Interp
+ * structure itself.
+ */
+
+typedef struct ThreadLocalData {
+    int nsCount;		/* Master epoch counter is used for keeping
+				 * the values used in Tcl_Obj internal
+				 * representations sane. Must be thread-local
+				 * because Tcl_Objs can cross interpreter
+				 * boundaries within a thread (objects don't
+				 * generally cross threads). */
+} ThreadLocalData;
+
+typedef struct Foundation {
+    Tcl_Interp *interp;
+    Class *objectCls;		/* The root of the object system. */
+    Class *classCls;		/* The class of all classes. */
+    Tcl_Namespace *ooNs;	/* Master ::oo namespace. */
+    Tcl_Namespace *defineNs;	/* Namespace containing special commands for
+				 * manipulating objects and classes. The
+				 * "oo::define" command acts as a special kind
+				 * of ensemble for this namespace. */
+    Tcl_Namespace *objdefNs;	/* Namespace containing special commands for
+				 * manipulating objects and classes. The
+				 * "oo::objdefine" command acts as a special
+				 * kind of ensemble for this namespace. */
+    Tcl_Namespace *helpersNs;	/* Namespace containing the commands that are
+				 * only valid when executing inside a
+				 * procedural method. */
+    int epoch;			/* Used to invalidate method chains when the
+				 * class structure changes. */
+    ThreadLocalData *tsdPtr;	/* Counter so we can allocate a unique
+				 * namespace to each object. */
+    Tcl_Obj *unknownMethodNameObj;
+				/* Shared object containing the name of the
+				 * unknown method handler method. */
+    Tcl_Obj *constructorName;	/* Shared object containing the "name" of a
+				 * constructor. */
+    Tcl_Obj *destructorName;	/* Shared object containing the "name" of a
+				 * destructor. */
+    Tcl_Obj *clonedName;	/* Shared object containing the name of a
+				 * "<cloned>" pseudo-constructor. */
+    Tcl_Obj *defineName;	/* Fully qualified name of oo::define. */
+} Foundation;
+
+/*
+ * A call context structure is built when a method is called. They contain the
+ * chain of method implementations that are to be invoked by a particular
+ * call, and the process of calling walks the chain, with the [next] command
+ * proceeding to the next entry in the chain.
+ */
+
+#define CALL_CHAIN_STATIC_SIZE 4
+
+struct MInvoke {
+    Method *mPtr;		/* Reference to the method implementation
+				 * record. */
+    int isFilter;		/* Whether this is a filter invokation. */
+    Class *filterDeclarer;	/* What class decided to add the filter; if
+				 * NULL, it was added by the object. */
+};
+
+typedef struct CallChain {
+    int objectCreationEpoch;	/* The object's creation epoch. Note that the
+				 * object reference is not stored in the call
+				 * chain; it is in the call context. */
+    int objectEpoch;		/* Local (object structure) epoch counter
+				 * snapshot. */
+    int epoch;			/* Global (class structure) epoch counter
+				 * snapshot. */
+    int flags;			/* Assorted flags, see below. */
+    int refCount;		/* Reference count. */
+    int numChain;		/* Size of the call chain. */
+    struct MInvoke *chain;	/* Array of call chain entries. May point to
+				 * staticChain if the number of entries is
+				 * small. */
+    struct MInvoke staticChain[CALL_CHAIN_STATIC_SIZE];
+} CallChain;
+
+typedef struct CallContext {
+    Object *oPtr;		/* The object associated with this call. */
+    int index;			/* Index into the call chain of the currently
+				 * executing method implementation. */
+    int skip;			/* Current number of arguments to skip; can
+				 * vary depending on whether it is a direct
+				 * method call or a continuation via the
+				 * [next] command. */
+    CallChain *callPtr;		/* The actual call chain. */
+} CallContext;
+
+/*
+ * Bits for the 'flags' field of the call chain.
+ */
+
+#define PUBLIC_METHOD     0x01	/* This is a public (exported) method. */
+#define PRIVATE_METHOD    0x02	/* This is a private (class's direct instances
+				 * only) method. */
+#define OO_UNKNOWN_METHOD 0x04	/* This is an unknown method. */
+#define CONSTRUCTOR	  0x08	/* This is a constructor. */
+#define DESTRUCTOR	  0x10	/* This is a destructor. */
+
+/*
+ * Structure containing definition information about basic class methods.
+ */
+
+typedef struct {
+    const char *name;		/* Name of the method in question. */
+    int isPublic;		/* Whether the method is public by default. */
+    Tcl_MethodType definition;	/* How to call the method. */
+} DeclaredClassMethod;
+
+/*
+ *----------------------------------------------------------------
+ * Commands relating to OO support.
+ *----------------------------------------------------------------
+ */
+
+MODULE_SCOPE int	TclOOInit(Tcl_Interp *interp);
+MODULE_SCOPE int	TclOODefineObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOOObjDefObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineConstructorObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineDeleteMethodObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineDestructorObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineExportObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineForwardObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineMethodObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineRenameMethodObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineUnexportObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineClassObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineSelfObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOODefineObjSelfObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOOUnknownDefinition(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOOCopyObjectCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOONextObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOONextToObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOOSelfObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv);
+
+/*
+ * Method implementations (in tclOOBasic.c).
+ */
+
+MODULE_SCOPE int	TclOO_Class_Constructor(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Class_Create(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Class_CreateNs(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Class_New(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Object_Destroy(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Object_Eval(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Object_LinkVar(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Object_Unknown(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+MODULE_SCOPE int	TclOO_Object_VarName(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+
+/*
+ * Private definitions, some of which perhaps ought to be exposed properly or
+ * maybe just put in the internal stubs table.
+ */
+
+MODULE_SCOPE void	TclOOAddToInstances(Object *oPtr, Class *clsPtr);
+MODULE_SCOPE void	TclOOAddToMixinSubs(Class *subPtr, Class *mixinPtr);
+MODULE_SCOPE void	TclOOAddToSubclasses(Class *subPtr, Class *superPtr);
+MODULE_SCOPE int	TclNRNewObjectInstance(Tcl_Interp *interp,
+			    Tcl_Class cls, const char *nameStr,
+			    const char *nsNameStr, int objc,
+			    Tcl_Obj *const *objv, int skip,
+			    Tcl_Object *objectPtr);
+MODULE_SCOPE int	TclOODefineSlots(Foundation *fPtr);
+MODULE_SCOPE void	TclOODeleteChain(CallChain *callPtr);
+MODULE_SCOPE void	TclOODeleteChainCache(Tcl_HashTable *tablePtr);
+MODULE_SCOPE void	TclOODeleteContext(CallContext *contextPtr);
+MODULE_SCOPE void	TclOODelMethodRef(Method *method);
+MODULE_SCOPE CallContext *TclOOGetCallContext(Object *oPtr,
+			    Tcl_Obj *methodNameObj, int flags,
+			    Tcl_Obj *cacheInThisObj);
+MODULE_SCOPE CallChain *TclOOGetStereotypeCallChain(Class *clsPtr,
+			    Tcl_Obj *methodNameObj, int flags);
+MODULE_SCOPE Foundation	*TclOOGetFoundation(Tcl_Interp *interp);
+MODULE_SCOPE Tcl_Obj *	TclOOGetFwdFromMethod(Method *mPtr);
+MODULE_SCOPE Proc *	TclOOGetProcFromMethod(Method *mPtr);
+MODULE_SCOPE Tcl_Obj *	TclOOGetMethodBody(Method *mPtr);
+MODULE_SCOPE int	TclOOGetSortedClassMethodList(Class *clsPtr,
+			    int flags, const char ***stringsPtr);
+MODULE_SCOPE int	TclOOGetSortedMethodList(Object *oPtr, int flags,
+			    const char ***stringsPtr);
+MODULE_SCOPE int	TclOOInit(Tcl_Interp *interp);
+MODULE_SCOPE void	TclOOInitInfo(Tcl_Interp *interp);
+MODULE_SCOPE int	TclOOInvokeContext(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+MODULE_SCOPE int	TclNRObjectContextInvokeNext(Tcl_Interp *interp,
+			    Tcl_ObjectContext context, int objc,
+			    Tcl_Obj *const *objv, int skip);
+MODULE_SCOPE void	TclOONewBasicMethod(Tcl_Interp *interp, Class *clsPtr,
+			    const DeclaredClassMethod *dcm);
+MODULE_SCOPE Tcl_Obj *	TclOOObjectName(Tcl_Interp *interp, Object *oPtr);
+MODULE_SCOPE void	TclOORemoveFromInstances(Object *oPtr, Class *clsPtr);
+MODULE_SCOPE void	TclOORemoveFromMixinSubs(Class *subPtr,
+			    Class *mixinPtr);
+MODULE_SCOPE void	TclOORemoveFromSubclasses(Class *subPtr,
+			    Class *superPtr);
+MODULE_SCOPE Tcl_Obj *	TclOORenderCallChain(Tcl_Interp *interp,
+			    CallChain *callPtr);
+MODULE_SCOPE void	TclOOStashContext(Tcl_Obj *objPtr,
+			    CallContext *contextPtr);
+MODULE_SCOPE void	TclOOSetupVariableResolver(Tcl_Namespace *nsPtr);
+
+/*
+ * Include all the private API, generated from tclOO.decls.
+ */
+
+#include "tclOOIntDecls.h"
+
+/*
+ * A convenience macro for iterating through the lists used in the internal
+ * memory management of objects. This is a bit gnarly because we want to do
+ * the assignment of the picked-out value only when the body test succeeds,
+ * but we cannot rely on the assigned value being useful, forcing us to do
+ * some nasty stuff with the comma operator. The compiler's optimizer should
+ * be able to sort it all out!
+ *
+ * REQUIRES DECLARATION: int i;
+ */
+
+#define FOREACH(var,ary) \
+	for(i=0 ; (i<(ary).num?((var=(ary).list[i]),1):0) ; i++)
+
+/*
+ * Convenience macros for iterating through hash tables. FOREACH_HASH_DECLS
+ * sets up the declarations needed for the main macro, FOREACH_HASH, which
+ * does the actual iteration. FOREACH_HASH_VALUE is a restricted version that
+ * only iterates over values.
+ */
+
+#define FOREACH_HASH_DECLS \
+    Tcl_HashEntry *hPtr;Tcl_HashSearch search
+#define FOREACH_HASH(key,val,tablePtr) \
+    for(hPtr=Tcl_FirstHashEntry((tablePtr),&search); hPtr!=NULL ? \
+	    ((key)=(void *)Tcl_GetHashKey((tablePtr),hPtr),\
+	    (val)=Tcl_GetHashValue(hPtr),1):0; hPtr=Tcl_NextHashEntry(&search))
+#define FOREACH_HASH_VALUE(val,tablePtr) \
+    for(hPtr=Tcl_FirstHashEntry((tablePtr),&search); hPtr!=NULL ? \
+	    ((val)=Tcl_GetHashValue(hPtr),1):0;hPtr=Tcl_NextHashEntry(&search))
+
+/*
+ * Convenience macro for duplicating a list. Needs no external declaration,
+ * but all arguments are used multiple times and so must have no side effects.
+ */
+
+#undef DUPLICATE /* prevent possible conflict with definition in WINAPI nb30.h */
+#define DUPLICATE(target,source,type) \
+    do { \
+	register unsigned len = sizeof(type) * ((target).num=(source).num);\
+	if (len != 0) { \
+	    memcpy(((target).list=(type*)ckalloc(len)), (source).list, len); \
+	} else { \
+	    (target).list = NULL; \
+	} \
+    } while(0)
+
+/*
+ * Alternatives to Tcl_Preserve/Tcl_EventuallyFree/Tcl_Release.
+ */
+
+#define AddRef(ptr) ((ptr)->refCount++)
+#define DelRef(ptr) do {			\
+	if ((ptr)->refCount-- <= 1) {		\
+	    ckfree(ptr);			\
+	}					\
+    } while(0)
+
+#endif /* TCL_OO_INTERNAL_H */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOIntDecls.h b/generic/tclOOIntDecls.h
new file mode 100644
index 0000000..74a8d81
--- /dev/null
+++ b/generic/tclOOIntDecls.h
@@ -0,0 +1,166 @@
+/*
+ * This file is (mostly) automatically generated from tclOO.decls.
+ */
+
+#ifndef _TCLOOINTDECLS
+#define _TCLOOINTDECLS
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+/* 0 */
+TCLAPI Tcl_Object	TclOOGetDefineCmdContext(Tcl_Interp *interp);
+/* 1 */
+TCLAPI Tcl_Method	TclOOMakeProcInstanceMethod(Tcl_Interp *interp,
+				Object *oPtr, int flags, Tcl_Obj *nameObj,
+				Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+				const Tcl_MethodType *typePtr,
+				ClientData clientData, Proc **procPtrPtr);
+/* 2 */
+TCLAPI Tcl_Method	TclOOMakeProcMethod(Tcl_Interp *interp,
+				Class *clsPtr, int flags, Tcl_Obj *nameObj,
+				const char *namePtr, Tcl_Obj *argsObj,
+				Tcl_Obj *bodyObj,
+				const Tcl_MethodType *typePtr,
+				ClientData clientData, Proc **procPtrPtr);
+/* 3 */
+TCLAPI Method *		TclOONewProcInstanceMethod(Tcl_Interp *interp,
+				Object *oPtr, int flags, Tcl_Obj *nameObj,
+				Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+				ProcedureMethod **pmPtrPtr);
+/* 4 */
+TCLAPI Method *		TclOONewProcMethod(Tcl_Interp *interp, Class *clsPtr,
+				int flags, Tcl_Obj *nameObj,
+				Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+				ProcedureMethod **pmPtrPtr);
+/* 5 */
+TCLAPI int		TclOOObjectCmdCore(Object *oPtr, Tcl_Interp *interp,
+				int objc, Tcl_Obj *const *objv,
+				int publicOnly, Class *startCls);
+/* 6 */
+TCLAPI int		TclOOIsReachable(Class *targetPtr, Class *startPtr);
+/* 7 */
+TCLAPI Method *		TclOONewForwardMethod(Tcl_Interp *interp,
+				Class *clsPtr, int isPublic,
+				Tcl_Obj *nameObj, Tcl_Obj *prefixObj);
+/* 8 */
+TCLAPI Method *		TclOONewForwardInstanceMethod(Tcl_Interp *interp,
+				Object *oPtr, int isPublic, Tcl_Obj *nameObj,
+				Tcl_Obj *prefixObj);
+/* 9 */
+TCLAPI Tcl_Method	TclOONewProcInstanceMethodEx(Tcl_Interp *interp,
+				Tcl_Object oPtr,
+				TclOO_PreCallProc *preCallPtr,
+				TclOO_PostCallProc *postCallPtr,
+				ProcErrorProc *errProc,
+				ClientData clientData, Tcl_Obj *nameObj,
+				Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+				int flags, void **internalTokenPtr);
+/* 10 */
+TCLAPI Tcl_Method	TclOONewProcMethodEx(Tcl_Interp *interp,
+				Tcl_Class clsPtr,
+				TclOO_PreCallProc *preCallPtr,
+				TclOO_PostCallProc *postCallPtr,
+				ProcErrorProc *errProc,
+				ClientData clientData, Tcl_Obj *nameObj,
+				Tcl_Obj *argsObj, Tcl_Obj *bodyObj,
+				int flags, void **internalTokenPtr);
+/* 11 */
+TCLAPI int		TclOOInvokeObject(Tcl_Interp *interp,
+				Tcl_Object object, Tcl_Class startCls,
+				int publicPrivate, int objc,
+				Tcl_Obj *const *objv);
+/* 12 */
+TCLAPI void		TclOOObjectSetFilters(Object *oPtr, int numFilters,
+				Tcl_Obj *const *filters);
+/* 13 */
+TCLAPI void		TclOOClassSetFilters(Tcl_Interp *interp,
+				Class *classPtr, int numFilters,
+				Tcl_Obj *const *filters);
+/* 14 */
+TCLAPI void		TclOOObjectSetMixins(Object *oPtr, int numMixins,
+				Class *const *mixins);
+/* 15 */
+TCLAPI void		TclOOClassSetMixins(Tcl_Interp *interp,
+				Class *classPtr, int numMixins,
+				Class *const *mixins);
+
+typedef struct TclOOIntStubs {
+    int magic;
+    void *hooks;
+
+    Tcl_Object (*tclOOGetDefineCmdContext) (Tcl_Interp *interp); /* 0 */
+    Tcl_Method (*tclOOMakeProcInstanceMethod) (Tcl_Interp *interp, Object *oPtr, int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, const Tcl_MethodType *typePtr, ClientData clientData, Proc **procPtrPtr); /* 1 */
+    Tcl_Method (*tclOOMakeProcMethod) (Tcl_Interp *interp, Class *clsPtr, int flags, Tcl_Obj *nameObj, const char *namePtr, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, const Tcl_MethodType *typePtr, ClientData clientData, Proc **procPtrPtr); /* 2 */
+    Method * (*tclOONewProcInstanceMethod) (Tcl_Interp *interp, Object *oPtr, int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, ProcedureMethod **pmPtrPtr); /* 3 */
+    Method * (*tclOONewProcMethod) (Tcl_Interp *interp, Class *clsPtr, int flags, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, ProcedureMethod **pmPtrPtr); /* 4 */
+    int (*tclOOObjectCmdCore) (Object *oPtr, Tcl_Interp *interp, int objc, Tcl_Obj *const *objv, int publicOnly, Class *startCls); /* 5 */
+    int (*tclOOIsReachable) (Class *targetPtr, Class *startPtr); /* 6 */
+    Method * (*tclOONewForwardMethod) (Tcl_Interp *interp, Class *clsPtr, int isPublic, Tcl_Obj *nameObj, Tcl_Obj *prefixObj); /* 7 */
+    Method * (*tclOONewForwardInstanceMethod) (Tcl_Interp *interp, Object *oPtr, int isPublic, Tcl_Obj *nameObj, Tcl_Obj *prefixObj); /* 8 */
+    Tcl_Method (*tclOONewProcInstanceMethodEx) (Tcl_Interp *interp, Tcl_Object oPtr, TclOO_PreCallProc *preCallPtr, TclOO_PostCallProc *postCallPtr, ProcErrorProc *errProc, ClientData clientData, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, int flags, void **internalTokenPtr); /* 9 */
+    Tcl_Method (*tclOONewProcMethodEx) (Tcl_Interp *interp, Tcl_Class clsPtr, TclOO_PreCallProc *preCallPtr, TclOO_PostCallProc *postCallPtr, ProcErrorProc *errProc, ClientData clientData, Tcl_Obj *nameObj, Tcl_Obj *argsObj, Tcl_Obj *bodyObj, int flags, void **internalTokenPtr); /* 10 */
+    int (*tclOOInvokeObject) (Tcl_Interp *interp, Tcl_Object object, Tcl_Class startCls, int publicPrivate, int objc, Tcl_Obj *const *objv); /* 11 */
+    void (*tclOOObjectSetFilters) (Object *oPtr, int numFilters, Tcl_Obj *const *filters); /* 12 */
+    void (*tclOOClassSetFilters) (Tcl_Interp *interp, Class *classPtr, int numFilters, Tcl_Obj *const *filters); /* 13 */
+    void (*tclOOObjectSetMixins) (Object *oPtr, int numMixins, Class *const *mixins); /* 14 */
+    void (*tclOOClassSetMixins) (Tcl_Interp *interp, Class *classPtr, int numMixins, Class *const *mixins); /* 15 */
+} TclOOIntStubs;
+
+extern const TclOOIntStubs *tclOOIntStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCLOO_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#define TclOOGetDefineCmdContext \
+	(tclOOIntStubsPtr->tclOOGetDefineCmdContext) /* 0 */
+#define TclOOMakeProcInstanceMethod \
+	(tclOOIntStubsPtr->tclOOMakeProcInstanceMethod) /* 1 */
+#define TclOOMakeProcMethod \
+	(tclOOIntStubsPtr->tclOOMakeProcMethod) /* 2 */
+#define TclOONewProcInstanceMethod \
+	(tclOOIntStubsPtr->tclOONewProcInstanceMethod) /* 3 */
+#define TclOONewProcMethod \
+	(tclOOIntStubsPtr->tclOONewProcMethod) /* 4 */
+#define TclOOObjectCmdCore \
+	(tclOOIntStubsPtr->tclOOObjectCmdCore) /* 5 */
+#define TclOOIsReachable \
+	(tclOOIntStubsPtr->tclOOIsReachable) /* 6 */
+#define TclOONewForwardMethod \
+	(tclOOIntStubsPtr->tclOONewForwardMethod) /* 7 */
+#define TclOONewForwardInstanceMethod \
+	(tclOOIntStubsPtr->tclOONewForwardInstanceMethod) /* 8 */
+#define TclOONewProcInstanceMethodEx \
+	(tclOOIntStubsPtr->tclOONewProcInstanceMethodEx) /* 9 */
+#define TclOONewProcMethodEx \
+	(tclOOIntStubsPtr->tclOONewProcMethodEx) /* 10 */
+#define TclOOInvokeObject \
+	(tclOOIntStubsPtr->tclOOInvokeObject) /* 11 */
+#define TclOOObjectSetFilters \
+	(tclOOIntStubsPtr->tclOOObjectSetFilters) /* 12 */
+#define TclOOClassSetFilters \
+	(tclOOIntStubsPtr->tclOOClassSetFilters) /* 13 */
+#define TclOOObjectSetMixins \
+	(tclOOIntStubsPtr->tclOOObjectSetMixins) /* 14 */
+#define TclOOClassSetMixins \
+	(tclOOIntStubsPtr->tclOOClassSetMixins) /* 15 */
+
+#endif /* defined(USE_TCLOO_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#endif /* _TCLOOINTDECLS */
diff --git a/generic/tclOOMethod.c b/generic/tclOOMethod.c
new file mode 100644
index 0000000..9c49caa
--- /dev/null
+++ b/generic/tclOOMethod.c
@@ -0,0 +1,1763 @@
+/*
+ * tclOOMethod.c --
+ *
+ *	This file contains code to create and manage methods.
+ *
+ * Copyright (c) 2005-2011 by Donal K. Fellows
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclInt.h"
+#include "tclOOInt.h"
+#include "tclCompile.h"
+
+/*
+ * Structure used to help delay computing names of objects or classes for
+ * [info frame] until needed, making invokation faster in the normal case.
+ */
+
+struct PNI {
+    Tcl_Interp *interp;		/* Interpreter in which to compute the name of
+				 * a method. */
+    Tcl_Method method;		/* Method to compute the name of. */
+};
+
+/*
+ * Structure used to contain all the information needed about a call frame
+ * used in a procedure-like method.
+ */
+
+typedef struct {
+    CallFrame *framePtr;	/* Reference to the call frame itself (it's
+				 * actually allocated on the Tcl stack). */
+    ProcErrorProc *errProc;	/* The error handler for the body. */
+    Tcl_Obj *nameObj;		/* The "name" of the command. */
+    Command cmd;		/* The command structure. Mostly bogus. */
+    ExtraFrameInfo efi;		/* Extra information used for [info frame]. */
+    Command *oldCmdPtr;		/* Saved cmdPtr so that we can be safe after a
+				 * recursive call returns. */
+    struct PNI pni;		/* Specialist information used in the efi
+				 * field for this type of call. */
+} PMFrameData;
+
+/*
+ * Structure used to pass information about variable resolution to the
+ * on-the-ground resolvers used when working with resolved compiled variables.
+ */
+
+typedef struct {
+    Tcl_ResolvedVarInfo info;	/* "Type" information so that the compiled
+				 * variable can be linked to the namespace
+				 * variable at the right time. */
+    Tcl_Obj *variableObj;	/* The name of the variable. */
+    Tcl_Var cachedObjectVar;	/* TODO: When to flush this cache? Can class
+				 * variables be cached? */
+} OOResVarInfo;
+
+/*
+ * Function declarations for things defined in this file.
+ */
+
+static Tcl_Obj **	InitEnsembleRewrite(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const *objv, int toRewrite,
+			    int rewriteLength, Tcl_Obj *const *rewriteObjs,
+			    int *lengthPtr);
+static int		InvokeProcedureMethod(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static Tcl_NRPostProc	FinalizeForwardCall;
+static Tcl_NRPostProc	FinalizePMCall;
+static int		PushMethodCallFrame(Tcl_Interp *interp,
+			    CallContext *contextPtr, ProcedureMethod *pmPtr,
+			    int objc, Tcl_Obj *const *objv,
+			    PMFrameData *fdPtr);
+static void		DeleteProcedureMethodRecord(ProcedureMethod *pmPtr);
+static void		DeleteProcedureMethod(ClientData clientData);
+static int		CloneProcedureMethod(Tcl_Interp *interp,
+			    ClientData clientData, ClientData *newClientData);
+static void		MethodErrorHandler(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj);
+static void		ConstructorErrorHandler(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj);
+static void		DestructorErrorHandler(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj);
+static Tcl_Obj *	RenderDeclarerName(ClientData clientData);
+static int		InvokeForwardMethod(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_ObjectContext context,
+			    int objc, Tcl_Obj *const *objv);
+static void		DeleteForwardMethod(ClientData clientData);
+static int		CloneForwardMethod(Tcl_Interp *interp,
+			    ClientData clientData, ClientData *newClientData);
+static int		ProcedureMethodVarResolver(Tcl_Interp *interp,
+			    const char *varName, Tcl_Namespace *contextNs,
+			    int flags, Tcl_Var *varPtr);
+static int		ProcedureMethodCompiledVarResolver(Tcl_Interp *interp,
+			    const char *varName, int length,
+			    Tcl_Namespace *contextNs,
+			    Tcl_ResolvedVarInfo **rPtrPtr);
+
+/*
+ * The types of methods defined by the core OO system.
+ */
+
+static const Tcl_MethodType procMethodType = {
+    TCL_OO_METHOD_VERSION_CURRENT, "method",
+    InvokeProcedureMethod, DeleteProcedureMethod, CloneProcedureMethod
+};
+static const Tcl_MethodType fwdMethodType = {
+    TCL_OO_METHOD_VERSION_CURRENT, "forward",
+    InvokeForwardMethod, DeleteForwardMethod, CloneForwardMethod
+};
+
+/*
+ * Helper macros (derived from things private to tclVar.c)
+ */
+
+#define TclVarTable(contextNs) \
+    ((Tcl_HashTable *) (&((Namespace *) (contextNs))->varTable))
+#define TclVarHashGetValue(hPtr) \
+    ((Tcl_Var) ((char *)hPtr - TclOffset(VarInHash, entry)))
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_NewInstanceMethod --
+ *
+ *	Attach a method to an object instance.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Method
+Tcl_NewInstanceMethod(
+    Tcl_Interp *interp,		/* Unused? */
+    Tcl_Object object,		/* The object that has the method attached to
+				 * it. */
+    Tcl_Obj *nameObj,		/* The name of the method. May be NULL; if so,
+				 * up to caller to manage storage (e.g., when
+				 * it is a constructor or destructor). */
+    int flags,			/* Whether this is a public method. */
+    const Tcl_MethodType *typePtr,
+				/* The type of method this is, which defines
+				 * how to invoke, delete and clone the
+				 * method. */
+    ClientData clientData)	/* Some data associated with the particular
+				 * method to be created. */
+{
+    register Object *oPtr = (Object *) object;
+    register Method *mPtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (nameObj == NULL) {
+	mPtr = ckalloc(sizeof(Method));
+	mPtr->namePtr = NULL;
+	mPtr->refCount = 1;
+	goto populate;
+    }
+    if (!oPtr->methodsPtr) {
+	oPtr->methodsPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitObjHashTable(oPtr->methodsPtr);
+	oPtr->flags &= ~USE_CLASS_CACHE;
+    }
+    hPtr = Tcl_CreateHashEntry(oPtr->methodsPtr, (char *) nameObj, &isNew);
+    if (isNew) {
+	mPtr = ckalloc(sizeof(Method));
+	mPtr->namePtr = nameObj;
+	mPtr->refCount = 1;
+	Tcl_IncrRefCount(nameObj);
+	Tcl_SetHashValue(hPtr, mPtr);
+    } else {
+	mPtr = Tcl_GetHashValue(hPtr);
+	if (mPtr->typePtr != NULL && mPtr->typePtr->deleteProc != NULL) {
+	    mPtr->typePtr->deleteProc(mPtr->clientData);
+	}
+    }
+
+  populate:
+    mPtr->typePtr = typePtr;
+    mPtr->clientData = clientData;
+    mPtr->flags = 0;
+    mPtr->declaringObjectPtr = oPtr;
+    mPtr->declaringClassPtr = NULL;
+    if (flags) {
+	mPtr->flags |= flags & (PUBLIC_METHOD | PRIVATE_METHOD);
+    }
+    oPtr->epoch++;
+    return (Tcl_Method) mPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * Tcl_NewMethod --
+ *
+ *	Attach a method to a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Method
+Tcl_NewMethod(
+    Tcl_Interp *interp,		/* The interpreter containing the class. */
+    Tcl_Class cls,		/* The class to attach the method to. */
+    Tcl_Obj *nameObj,		/* The name of the object. May be NULL (e.g.,
+				 * for constructors or destructors); if so, up
+				 * to caller to manage storage. */
+    int flags,			/* Whether this is a public method. */
+    const Tcl_MethodType *typePtr,
+				/* The type of method this is, which defines
+				 * how to invoke, delete and clone the
+				 * method. */
+    ClientData clientData)	/* Some data associated with the particular
+				 * method to be created. */
+{
+    register Class *clsPtr = (Class *) cls;
+    register Method *mPtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+
+    if (nameObj == NULL) {
+	mPtr = ckalloc(sizeof(Method));
+	mPtr->namePtr = NULL;
+	mPtr->refCount = 1;
+	goto populate;
+    }
+    hPtr = Tcl_CreateHashEntry(&clsPtr->classMethods, (char *)nameObj,&isNew);
+    if (isNew) {
+	mPtr = ckalloc(sizeof(Method));
+	mPtr->refCount = 1;
+	mPtr->namePtr = nameObj;
+	Tcl_IncrRefCount(nameObj);
+	Tcl_SetHashValue(hPtr, mPtr);
+    } else {
+	mPtr = Tcl_GetHashValue(hPtr);
+	if (mPtr->typePtr != NULL && mPtr->typePtr->deleteProc != NULL) {
+	    mPtr->typePtr->deleteProc(mPtr->clientData);
+	}
+    }
+
+  populate:
+    clsPtr->thisPtr->fPtr->epoch++;
+    mPtr->typePtr = typePtr;
+    mPtr->clientData = clientData;
+    mPtr->flags = 0;
+    mPtr->declaringObjectPtr = NULL;
+    mPtr->declaringClassPtr = clsPtr;
+    if (flags) {
+	mPtr->flags |= flags & (PUBLIC_METHOD | PRIVATE_METHOD);
+    }
+
+    return (Tcl_Method) mPtr;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOODelMethodRef --
+ *
+ *	How to delete a method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOODelMethodRef(
+    Method *mPtr)
+{
+    if ((mPtr != NULL) && (mPtr->refCount-- <= 1)) {
+	if (mPtr->typePtr != NULL && mPtr->typePtr->deleteProc != NULL) {
+	    mPtr->typePtr->deleteProc(mPtr->clientData);
+	}
+	if (mPtr->namePtr != NULL) {
+	    Tcl_DecrRefCount(mPtr->namePtr);
+	}
+
+	ckfree(mPtr);
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONewBasicMethod --
+ *
+ *	Helper that makes it cleaner to create very simple methods during
+ *	basic system initialization. Not suitable for general use.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOONewBasicMethod(
+    Tcl_Interp *interp,
+    Class *clsPtr,		/* Class to attach the method to. */
+    const DeclaredClassMethod *dcm)
+				/* Name of the method, whether it is public,
+				 * and the function to implement it. */
+{
+    Tcl_Obj *namePtr = Tcl_NewStringObj(dcm->name, -1);
+
+    Tcl_IncrRefCount(namePtr);
+    Tcl_NewMethod(interp, (Tcl_Class) clsPtr, namePtr,
+	    (dcm->isPublic ? PUBLIC_METHOD : 0), &dcm->definition, NULL);
+    Tcl_DecrRefCount(namePtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONewProcInstanceMethod --
+ *
+ *	Create a new procedure-like method for an object.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Method *
+TclOONewProcInstanceMethod(
+    Tcl_Interp *interp,		/* The interpreter containing the object. */
+    Object *oPtr,		/* The object to modify. */
+    int flags,			/* Whether this is a public method. */
+    Tcl_Obj *nameObj,		/* The name of the method, which must not be
+				 * NULL. */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which must not be NULL. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which must not be
+				 * NULL. */
+    ProcedureMethod **pmPtrPtr)	/* Place to write pointer to procedure method
+				 * structure to allow for deeper tuning of the
+				 * structure's contents. NULL if caller is not
+				 * interested. */
+{
+    int argsLen;
+    register ProcedureMethod *pmPtr;
+    Tcl_Method method;
+
+    if (Tcl_ListObjLength(interp, argsObj, &argsLen) != TCL_OK) {
+	return NULL;
+    }
+    pmPtr = ckalloc(sizeof(ProcedureMethod));
+    memset(pmPtr, 0, sizeof(ProcedureMethod));
+    pmPtr->version = TCLOO_PROCEDURE_METHOD_VERSION;
+    pmPtr->flags = flags & USE_DECLARER_NS;
+    pmPtr->refCount = 1;
+
+    method = TclOOMakeProcInstanceMethod(interp, oPtr, flags, nameObj,
+	    argsObj, bodyObj, &procMethodType, pmPtr, &pmPtr->procPtr);
+    if (method == NULL) {
+	ckfree(pmPtr);
+    } else if (pmPtrPtr != NULL) {
+	*pmPtrPtr = pmPtr;
+    }
+    return (Method *) method;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONewProcMethod --
+ *
+ *	Create a new procedure-like method for a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Method *
+TclOONewProcMethod(
+    Tcl_Interp *interp,		/* The interpreter containing the class. */
+    Class *clsPtr,		/* The class to modify. */
+    int flags,			/* Whether this is a public method. */
+    Tcl_Obj *nameObj,		/* The name of the method, which may be NULL;
+				 * if so, up to caller to manage storage
+				 * (e.g., because it is a constructor or
+				 * destructor). */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which may be NULL; if so, it is equivalent
+				 * to an empty list. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which must not be
+				 * NULL. */
+    ProcedureMethod **pmPtrPtr)	/* Place to write pointer to procedure method
+				 * structure to allow for deeper tuning of the
+				 * structure's contents. NULL if caller is not
+				 * interested. */
+{
+    int argsLen;		/* -1 => delete argsObj before exit */
+    register ProcedureMethod *pmPtr;
+    const char *procName;
+    Tcl_Method method;
+
+    if (argsObj == NULL) {
+	argsLen = -1;
+	argsObj = Tcl_NewObj();
+	Tcl_IncrRefCount(argsObj);
+	procName = "<destructor>";
+    } else if (Tcl_ListObjLength(interp, argsObj, &argsLen) != TCL_OK) {
+	return NULL;
+    } else {
+	procName = (nameObj==NULL ? "<constructor>" : TclGetString(nameObj));
+    }
+
+    pmPtr = ckalloc(sizeof(ProcedureMethod));
+    memset(pmPtr, 0, sizeof(ProcedureMethod));
+    pmPtr->version = TCLOO_PROCEDURE_METHOD_VERSION;
+    pmPtr->flags = flags & USE_DECLARER_NS;
+    pmPtr->refCount = 1;
+
+    method = TclOOMakeProcMethod(interp, clsPtr, flags, nameObj, procName,
+	    argsObj, bodyObj, &procMethodType, pmPtr, &pmPtr->procPtr);
+
+    if (argsLen == -1) {
+	Tcl_DecrRefCount(argsObj);
+    }
+    if (method == NULL) {
+	ckfree(pmPtr);
+    } else if (pmPtrPtr != NULL) {
+	*pmPtrPtr = pmPtr;
+    }
+
+    return (Method *) method;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOMakeProcInstanceMethod --
+ *
+ *	The guts of the code to make a procedure-like method for an object.
+ *	Split apart so that it is easier for other extensions to reuse (in
+ *	particular, it frees them from having to pry so deeply into Tcl's
+ *	guts).
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Method
+TclOOMakeProcInstanceMethod(
+    Tcl_Interp *interp,		/* The interpreter containing the object. */
+    Object *oPtr,		/* The object to modify. */
+    int flags,			/* Whether this is a public method. */
+    Tcl_Obj *nameObj,		/* The name of the method, which _must not_ be
+				 * NULL. */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which _must not_ be NULL. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which _must not_ be
+				 * NULL. */
+    const Tcl_MethodType *typePtr,
+				/* The type of the method to create. */
+    ClientData clientData,	/* The per-method type-specific data. */
+    Proc **procPtrPtr)		/* A pointer to the variable in which to write
+				 * the procedure record reference. Presumably
+				 * inside the structure indicated by the
+				 * pointer in clientData. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Proc *procPtr;
+
+    if (TclCreateProc(interp, NULL, TclGetString(nameObj), argsObj, bodyObj,
+	    procPtrPtr) != TCL_OK) {
+	return NULL;
+    }
+    procPtr = *procPtrPtr;
+    procPtr->cmdPtr = NULL;
+
+    if (iPtr->cmdFramePtr) {
+	CmdFrame context = *iPtr->cmdFramePtr;
+
+	if (context.type == TCL_LOCATION_BC) {
+	    /*
+	     * Retrieve source information from the bytecode, if possible. If
+	     * the information is retrieved successfully, context.type will be
+	     * TCL_LOCATION_SOURCE and the reference held by
+	     * context.data.eval.path will be counted.
+	     */
+
+	    TclGetSrcInfoForPc(&context);
+	} else if (context.type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * The copy into 'context' up above has created another reference
+	     * to 'context.data.eval.path'; account for it.
+	     */
+
+	    Tcl_IncrRefCount(context.data.eval.path);
+	}
+
+	if (context.type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * We can account for source location within a proc only if the
+	     * proc body was not created by substitution.
+	     * (FIXME: check that this is sane and correct!)
+	     */
+
+	    if (context.line
+		    && (context.nline >= 4) && (context.line[3] >= 0)) {
+		int isNew;
+		CmdFrame *cfPtr = ckalloc(sizeof(CmdFrame));
+		Tcl_HashEntry *hPtr;
+
+		cfPtr->level = -1;
+		cfPtr->type = context.type;
+		cfPtr->line = ckalloc(sizeof(int));
+		cfPtr->line[0] = context.line[3];
+		cfPtr->nline = 1;
+		cfPtr->framePtr = NULL;
+		cfPtr->nextPtr = NULL;
+
+		cfPtr->data.eval.path = context.data.eval.path;
+		Tcl_IncrRefCount(cfPtr->data.eval.path);
+
+		cfPtr->cmd = NULL;
+		cfPtr->len = 0;
+
+		hPtr = Tcl_CreateHashEntry(iPtr->linePBodyPtr,
+			(char *) procPtr, &isNew);
+		Tcl_SetHashValue(hPtr, cfPtr);
+	    }
+
+	    /*
+	     * 'context' is going out of scope; account for the reference that
+	     * it's holding to the path name.
+	     */
+
+	    Tcl_DecrRefCount(context.data.eval.path);
+	    context.data.eval.path = NULL;
+	}
+    }
+
+    return Tcl_NewInstanceMethod(interp, (Tcl_Object) oPtr, nameObj, flags,
+	    typePtr, clientData);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOMakeProcMethod --
+ *
+ *	The guts of the code to make a procedure-like method for a class.
+ *	Split apart so that it is easier for other extensions to reuse (in
+ *	particular, it frees them from having to pry so deeply into Tcl's
+ *	guts).
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Method
+TclOOMakeProcMethod(
+    Tcl_Interp *interp,		/* The interpreter containing the class. */
+    Class *clsPtr,		/* The class to modify. */
+    int flags,			/* Whether this is a public method. */
+    Tcl_Obj *nameObj,		/* The name of the method, which may be NULL;
+				 * if so, up to caller to manage storage
+				 * (e.g., because it is a constructor or
+				 * destructor). */
+    const char *namePtr,	/* The name of the method as a string, which
+				 * _must not_ be NULL. */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which _must not_ be NULL. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which _must not_ be
+				 * NULL. */
+    const Tcl_MethodType *typePtr,
+				/* The type of the method to create. */
+    ClientData clientData,	/* The per-method type-specific data. */
+    Proc **procPtrPtr)		/* A pointer to the variable in which to write
+				 * the procedure record reference. Presumably
+				 * inside the structure indicated by the
+				 * pointer in clientData. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Proc *procPtr;
+
+    if (TclCreateProc(interp, NULL, namePtr, argsObj, bodyObj,
+	    procPtrPtr) != TCL_OK) {
+	return NULL;
+    }
+    procPtr = *procPtrPtr;
+    procPtr->cmdPtr = NULL;
+
+    if (iPtr->cmdFramePtr) {
+	CmdFrame context = *iPtr->cmdFramePtr;
+
+	if (context.type == TCL_LOCATION_BC) {
+	    /*
+	     * Retrieve source information from the bytecode, if possible. If
+	     * the information is retrieved successfully, context.type will be
+	     * TCL_LOCATION_SOURCE and the reference held by
+	     * context.data.eval.path will be counted.
+	     */
+
+	    TclGetSrcInfoForPc(&context);
+	} else if (context.type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * The copy into 'context' up above has created another reference
+	     * to 'context.data.eval.path'; account for it.
+	     */
+
+	    Tcl_IncrRefCount(context.data.eval.path);
+	}
+
+	if (context.type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * We can account for source location within a proc only if the
+	     * proc body was not created by substitution.
+	     * (FIXME: check that this is sane and correct!)
+	     */
+
+	    if (context.line
+		    && (context.nline >= 4) && (context.line[3] >= 0)) {
+		int isNew;
+		CmdFrame *cfPtr = ckalloc(sizeof(CmdFrame));
+		Tcl_HashEntry *hPtr;
+
+		cfPtr->level = -1;
+		cfPtr->type = context.type;
+		cfPtr->line = ckalloc(sizeof(int));
+		cfPtr->line[0] = context.line[3];
+		cfPtr->nline = 1;
+		cfPtr->framePtr = NULL;
+		cfPtr->nextPtr = NULL;
+
+		cfPtr->data.eval.path = context.data.eval.path;
+		Tcl_IncrRefCount(cfPtr->data.eval.path);
+
+		cfPtr->cmd = NULL;
+		cfPtr->len = 0;
+
+		hPtr = Tcl_CreateHashEntry(iPtr->linePBodyPtr,
+			(char *) procPtr, &isNew);
+		Tcl_SetHashValue(hPtr, cfPtr);
+	    }
+
+	    /*
+	     * 'context' is going out of scope; account for the reference that
+	     * it's holding to the path name.
+	     */
+
+	    Tcl_DecrRefCount(context.data.eval.path);
+	    context.data.eval.path = NULL;
+	}
+    }
+
+    return Tcl_NewMethod(interp, (Tcl_Class) clsPtr, nameObj, flags, typePtr,
+	    clientData);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InvokeProcedureMethod, PushMethodCallFrame --
+ *
+ *	How to invoke a procedure-like method.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InvokeProcedureMethod(
+    ClientData clientData,	/* Pointer to some per-method context. */
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,	/* The method calling context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* Arguments as actually seen. */
+{
+    ProcedureMethod *pmPtr = clientData;
+    int result;
+    PMFrameData *fdPtr;		/* Important data that has to have a lifetime
+				 * matched by this function (or rather, by the
+				 * call frame's lifetime). */
+
+    /*
+     * If the interpreter was deleted, we just skip to the next thing in the
+     * chain.
+     */
+
+    if (Tcl_InterpDeleted(interp)) {
+	return TclNRObjectContextInvokeNext(interp, context, objc, objv,
+		Tcl_ObjectContextSkippedArgs(context));
+    }
+
+    /*
+     * Allocate the special frame data.
+     */
+
+    fdPtr = TclStackAlloc(interp, sizeof(PMFrameData));
+
+    /*
+     * Create a call frame for this method.
+     */
+
+    result = PushMethodCallFrame(interp, (CallContext *) context, pmPtr,
+	    objc, objv, fdPtr);
+    if (result != TCL_OK) {
+	TclStackFree(interp, fdPtr);
+	return result;
+    }
+    pmPtr->refCount++;
+
+    /*
+     * Give the pre-call callback a chance to do some setup and, possibly,
+     * veto the call.
+     */
+
+    if (pmPtr->preCallProc != NULL) {
+	int isFinished;
+
+	result = pmPtr->preCallProc(pmPtr->clientData, interp, context,
+		(Tcl_CallFrame *) fdPtr->framePtr, &isFinished);
+	if (isFinished || result != TCL_OK) {
+	    /*
+	     * Restore the old cmdPtr so that a subsequent use of [info frame]
+	     * won't crash on us. [Bug 3001438]
+	     */
+
+	    pmPtr->procPtr->cmdPtr = fdPtr->oldCmdPtr;
+
+	    Tcl_PopCallFrame(interp);
+	    TclStackFree(interp, fdPtr->framePtr);
+	    if (pmPtr->refCount-- <= 1) {
+		DeleteProcedureMethodRecord(pmPtr);
+	    }
+	    TclStackFree(interp, fdPtr);
+	    return result;
+	}
+    }
+
+    /*
+     * Now invoke the body of the method.
+     */
+
+    TclNRAddCallback(interp, FinalizePMCall, pmPtr, context, fdPtr, NULL);
+    return TclNRInterpProcCore(interp, fdPtr->nameObj,
+	    Tcl_ObjectContextSkippedArgs(context), fdPtr->errProc);
+}
+
+static int
+FinalizePMCall(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ProcedureMethod *pmPtr = data[0];
+    Tcl_ObjectContext context = data[1];
+    PMFrameData *fdPtr = data[2];
+
+    /*
+     * Give the post-call callback a chance to do some cleanup. Note that at
+     * this point the call frame itself is invalid; it's already been popped.
+     */
+
+    if (pmPtr->postCallProc) {
+	result = pmPtr->postCallProc(pmPtr->clientData, interp, context,
+		Tcl_GetObjectNamespace(Tcl_ObjectContextObject(context)),
+		result);
+    }
+
+    /*
+     * Restore the old cmdPtr so that a subsequent use of [info frame] won't
+     * crash on us. [Bug 3001438]
+     */
+
+    pmPtr->procPtr->cmdPtr = fdPtr->oldCmdPtr;
+
+    /*
+     * Scrap the special frame data now that we're done with it. Note that we
+     * are inlining DeleteProcedureMethod() here; this location is highly
+     * sensitive when it comes to performance!
+     */
+
+    if (pmPtr->refCount-- <= 1) {
+	DeleteProcedureMethodRecord(pmPtr);
+    }
+    TclStackFree(interp, fdPtr);
+    return result;
+}
+
+static int
+PushMethodCallFrame(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    CallContext *contextPtr,	/* Current method call context. */
+    ProcedureMethod *pmPtr,	/* Information about this procedure-like
+				 * method. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv,	/* Array of arguments. */
+    PMFrameData *fdPtr)		/* Place to store information about the call
+				 * frame. */
+{
+    Namespace *nsPtr = (Namespace *) contextPtr->oPtr->namespacePtr;
+    register int result;
+    const char *namePtr;
+    CallFrame **framePtrPtr = &fdPtr->framePtr;
+
+    /*
+     * Compute basic information on the basis of the type of method it is.
+     */
+
+    if (contextPtr->callPtr->flags & CONSTRUCTOR) {
+	namePtr = "<constructor>";
+	fdPtr->nameObj = contextPtr->oPtr->fPtr->constructorName;
+	fdPtr->errProc = ConstructorErrorHandler;
+    } else if (contextPtr->callPtr->flags & DESTRUCTOR) {
+	namePtr = "<destructor>";
+	fdPtr->nameObj = contextPtr->oPtr->fPtr->destructorName;
+	fdPtr->errProc = DestructorErrorHandler;
+    } else {
+	fdPtr->nameObj = Tcl_MethodName(
+		Tcl_ObjectContextMethod((Tcl_ObjectContext) contextPtr));
+	namePtr = TclGetString(fdPtr->nameObj);
+	fdPtr->errProc = MethodErrorHandler;
+    }
+    if (pmPtr->errProc != NULL) {
+	fdPtr->errProc = pmPtr->errProc;
+    }
+
+    /*
+     * Magic to enable things like [incr Tcl], which wants methods to run in
+     * their class's namespace.
+     */
+
+    if (pmPtr->flags & USE_DECLARER_NS) {
+	register Method *mPtr =
+		contextPtr->callPtr->chain[contextPtr->index].mPtr;
+
+	if (mPtr->declaringClassPtr != NULL) {
+	    nsPtr = (Namespace *)
+		    mPtr->declaringClassPtr->thisPtr->namespacePtr;
+	} else {
+	    nsPtr = (Namespace *) mPtr->declaringObjectPtr->namespacePtr;
+	}
+    }
+
+    /*
+     * Save the old cmdPtr so that when this recursive call returns, we can
+     * restore it. To do otherwise causes crashes in [info frame] after we
+     * return from a recursive call. [Bug 3001438]
+     */
+
+    fdPtr->oldCmdPtr = pmPtr->procPtr->cmdPtr;
+
+    /*
+     * Compile the body. This operation may fail.
+     */
+
+    fdPtr->efi.length = 2;
+    memset(&fdPtr->cmd, 0, sizeof(Command));
+    fdPtr->cmd.nsPtr = nsPtr;
+    fdPtr->cmd.clientData = &fdPtr->efi;
+    pmPtr->procPtr->cmdPtr = &fdPtr->cmd;
+
+    /*
+     * [Bug 2037727] Always call TclProcCompileProc so that we check not only
+     * that we have bytecode, but also that it remains valid. Note that we set
+     * the namespace of the code here directly; this is a hack, but the
+     * alternative is *so* slow...
+     */
+
+    if (pmPtr->procPtr->bodyPtr->typePtr == &tclByteCodeType) {
+	ByteCode *codePtr =
+		pmPtr->procPtr->bodyPtr->internalRep.twoPtrValue.ptr1;
+
+	codePtr->nsPtr = nsPtr;
+    }
+    result = TclProcCompileProc(interp, pmPtr->procPtr,
+	    pmPtr->procPtr->bodyPtr, nsPtr, "body of method", namePtr);
+    if (result != TCL_OK) {
+	goto failureReturn;
+    }
+
+    /*
+     * Make the stack frame and fill it out with information about this call.
+     * This operation may fail.
+     */
+
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    (Tcl_Namespace *) nsPtr, FRAME_IS_PROC|FRAME_IS_METHOD);
+
+    fdPtr->framePtr->clientData = contextPtr;
+    fdPtr->framePtr->objc = objc;
+    fdPtr->framePtr->objv = objv;
+    fdPtr->framePtr->procPtr = pmPtr->procPtr;
+
+    /*
+     * Finish filling out the extra frame info so that [info frame] works.
+     */
+
+    fdPtr->efi.fields[0].name = "method";
+    fdPtr->efi.fields[0].proc = NULL;
+    fdPtr->efi.fields[0].clientData = fdPtr->nameObj;
+    if (pmPtr->gfivProc != NULL) {
+	fdPtr->efi.fields[1].name = "";
+	fdPtr->efi.fields[1].proc = pmPtr->gfivProc;
+	fdPtr->efi.fields[1].clientData = pmPtr;
+    } else {
+	register Tcl_Method method =
+		Tcl_ObjectContextMethod((Tcl_ObjectContext) contextPtr);
+
+	if (Tcl_MethodDeclarerObject(method) != NULL) {
+	    fdPtr->efi.fields[1].name = "object";
+	} else {
+	    fdPtr->efi.fields[1].name = "class";
+	}
+	fdPtr->efi.fields[1].proc = RenderDeclarerName;
+	fdPtr->efi.fields[1].clientData = &fdPtr->pni;
+	fdPtr->pni.interp = interp;
+	fdPtr->pni.method = method;
+    }
+
+    return TCL_OK;
+
+    /*
+     * Restore the old cmdPtr so that a subsequent use of [info frame] won't
+     * crash on us. [Bug 3001438]
+     */
+
+  failureReturn:
+    pmPtr->procPtr->cmdPtr = fdPtr->oldCmdPtr;
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOSetupVariableResolver, etc. --
+ *
+ *	Variable resolution engine used to connect declared variables to local
+ *	variables used in methods. The compiled variable resolver is more
+ *	important, but both are needed as it is possible to have a variable
+ *	that is only referred to in ways that aren't compilable and we can't
+ *	force LVT presence. [TIP #320]
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOOSetupVariableResolver(
+    Tcl_Namespace *nsPtr)
+{
+    Tcl_ResolverInfo info;
+
+    Tcl_GetNamespaceResolvers(nsPtr, &info);
+    if (info.compiledVarResProc == NULL) {
+	Tcl_SetNamespaceResolvers(nsPtr, NULL, ProcedureMethodVarResolver,
+		ProcedureMethodCompiledVarResolver);
+    }
+}
+
+static int
+ProcedureMethodVarResolver(
+    Tcl_Interp *interp,
+    const char *varName,
+    Tcl_Namespace *contextNs,
+    int flags,
+    Tcl_Var *varPtr)
+{
+    int result;
+    Tcl_ResolvedVarInfo *rPtr = NULL;
+
+    result = ProcedureMethodCompiledVarResolver(interp, varName,
+	    strlen(varName), contextNs, &rPtr);
+
+    if (result != TCL_OK) {
+	return result;
+    }
+
+    *varPtr = rPtr->fetchProc(interp, rPtr);
+
+    /*
+     * Must not retain reference to resolved information. [Bug 3105999]
+     */
+
+    if (rPtr != NULL) {
+	rPtr->deleteProc(rPtr);
+    }
+    return (*varPtr? TCL_OK : TCL_CONTINUE);
+}
+
+static Tcl_Var
+ProcedureMethodCompiledVarConnect(
+    Tcl_Interp *interp,
+    Tcl_ResolvedVarInfo *rPtr)
+{
+    OOResVarInfo *infoPtr = (OOResVarInfo *) rPtr;
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *framePtr = iPtr->varFramePtr;
+    CallContext *contextPtr;
+    Tcl_Obj *variableObj;
+    Tcl_HashEntry *hPtr;
+    int i, isNew, cacheIt, varLen, len;
+    const char *match, *varName;
+
+    /*
+     * Check that the variable is being requested in a context that is also a
+     * method call; if not (i.e. we're evaluating in the object's namespace or
+     * in a procedure of that namespace) then we do nothing.
+     */
+
+    if (framePtr == NULL || !(framePtr->isProcCallFrame & FRAME_IS_METHOD)) {
+	return NULL;
+    }
+    contextPtr = framePtr->clientData;
+
+    /*
+     * If we've done the work before (in a comparable context) then reuse that
+     * rather than performing resolution ourselves.
+     */
+
+    if (infoPtr->cachedObjectVar) {
+	return infoPtr->cachedObjectVar;
+    }
+
+    /*
+     * Check if the variable is one we want to resolve at all (i.e. whether it
+     * is in the list provided by the user). If not, we mustn't do anything
+     * either.
+     */
+
+    varName = TclGetStringFromObj(infoPtr->variableObj, &varLen);
+    if (contextPtr->callPtr->chain[contextPtr->index]
+	    .mPtr->declaringClassPtr != NULL) {
+	FOREACH(variableObj, contextPtr->callPtr->chain[contextPtr->index]
+		.mPtr->declaringClassPtr->variables) {
+	    match = TclGetStringFromObj(variableObj, &len);
+	    if ((len == varLen) && !memcmp(match, varName, len)) {
+		cacheIt = 0;
+		goto gotMatch;
+	    }
+	}
+    } else {
+	FOREACH(variableObj, contextPtr->oPtr->variables) {
+	    match = TclGetStringFromObj(variableObj, &len);
+	    if ((len == varLen) && !memcmp(match, varName, len)) {
+		cacheIt = 1;
+		goto gotMatch;
+	    }
+	}
+    }
+    return NULL;
+
+    /*
+     * It is a variable we want to resolve, so resolve it.
+     */
+
+  gotMatch:
+    hPtr = Tcl_CreateHashEntry(TclVarTable(contextPtr->oPtr->namespacePtr),
+	    (char *) variableObj, &isNew);
+    if (isNew) {
+	TclSetVarNamespaceVar((Var *) TclVarHashGetValue(hPtr));
+    }
+    if (cacheIt) {
+	infoPtr->cachedObjectVar = TclVarHashGetValue(hPtr);
+
+	/*
+	 * We must keep a reference to the variable so everything will
+	 * continue to work correctly even if it is unset; being unset does
+	 * not end the life of the variable at this level. [Bug 3185009]
+	 */
+
+	VarHashRefCount(infoPtr->cachedObjectVar)++;
+    }
+    return TclVarHashGetValue(hPtr);
+}
+
+static void
+ProcedureMethodCompiledVarDelete(
+    Tcl_ResolvedVarInfo *rPtr)
+{
+    OOResVarInfo *infoPtr = (OOResVarInfo *) rPtr;
+
+    /*
+     * Release the reference to the variable if we were holding it.
+     */
+
+    if (infoPtr->cachedObjectVar) {
+	VarHashRefCount(infoPtr->cachedObjectVar)--;
+	TclCleanupVar((Var *) infoPtr->cachedObjectVar, NULL);
+    }
+    Tcl_DecrRefCount(infoPtr->variableObj);
+    ckfree(infoPtr);
+}
+
+static int
+ProcedureMethodCompiledVarResolver(
+    Tcl_Interp *interp,
+    const char *varName,
+    int length,
+    Tcl_Namespace *contextNs,
+    Tcl_ResolvedVarInfo **rPtrPtr)
+{
+    OOResVarInfo *infoPtr;
+    Tcl_Obj *variableObj = Tcl_NewStringObj(varName, length);
+
+    /*
+     * Do not create resolvers for cases that contain namespace separators or
+     * which look like array accesses. Both will lead us astray.
+     */
+
+    if (strstr(Tcl_GetString(variableObj), "::") != NULL ||
+	    Tcl_StringMatch(Tcl_GetString(variableObj), "*(*)")) {
+	Tcl_DecrRefCount(variableObj);
+	return TCL_CONTINUE;
+    }
+
+    infoPtr = ckalloc(sizeof(OOResVarInfo));
+    infoPtr->info.fetchProc = ProcedureMethodCompiledVarConnect;
+    infoPtr->info.deleteProc = ProcedureMethodCompiledVarDelete;
+    infoPtr->cachedObjectVar = NULL;
+    infoPtr->variableObj = variableObj;
+    Tcl_IncrRefCount(variableObj);
+    *rPtrPtr = &infoPtr->info;
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * RenderDeclarerName --
+ *
+ *	Returns the name of the entity (object or class) which declared a
+ *	method. Used for producing information for [info frame] in such a way
+ *	that the expensive part of this (generating the object or class name
+ *	itself) isn't done until it is needed.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static Tcl_Obj *
+RenderDeclarerName(
+    ClientData clientData)
+{
+    struct PNI *pni = clientData;
+    Tcl_Object object = Tcl_MethodDeclarerObject(pni->method);
+
+    if (object == NULL) {
+	object = Tcl_GetClassAsObject(Tcl_MethodDeclarerClass(pni->method));
+    }
+    return TclOOObjectName(pni->interp, (Object *) object);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * MethodErrorHandler, ConstructorErrorHandler, DestructorErrorHandler --
+ *
+ *	How to fill in the stack trace correctly upon error in various forms
+ *	of procedure-like methods. LIMIT is how long the inserted strings in
+ *	the error traces should get before being converted to have ellipses,
+ *	and ELLIPSIFY is a macro to do the conversion (with the help of a
+ *	%.*s%s format field). Note that ELLIPSIFY is only safe for use in
+ *	suitable formatting contexts.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+#define LIMIT 60
+#define ELLIPSIFY(str,len) \
+	((len) > LIMIT ? LIMIT : (len)), (str), ((len) > LIMIT ? "..." : "")
+
+static void
+MethodErrorHandler(
+    Tcl_Interp *interp,
+    Tcl_Obj *methodNameObj)
+{
+    int nameLen, objectNameLen;
+    CallContext *contextPtr = ((Interp *) interp)->varFramePtr->clientData;
+    Method *mPtr = contextPtr->callPtr->chain[contextPtr->index].mPtr;
+    const char *objectName, *kindName, *methodName =
+	    TclGetStringFromObj(mPtr->namePtr, &nameLen);
+    Object *declarerPtr;
+
+    if (mPtr->declaringObjectPtr != NULL) {
+	declarerPtr = mPtr->declaringObjectPtr;
+	kindName = "object";
+    } else {
+	if (mPtr->declaringClassPtr == NULL) {
+	    Tcl_Panic("method not declared in class or object");
+	}
+	declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	kindName = "class";
+    }
+
+    objectName = Tcl_GetStringFromObj(TclOOObjectName(interp, declarerPtr),
+	    &objectNameLen);
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (%s \"%.*s%s\" method \"%.*s%s\" line %d)",
+	    kindName, ELLIPSIFY(objectName, objectNameLen),
+	    ELLIPSIFY(methodName, nameLen), Tcl_GetErrorLine(interp)));
+}
+
+static void
+ConstructorErrorHandler(
+    Tcl_Interp *interp,
+    Tcl_Obj *methodNameObj)
+{
+    CallContext *contextPtr = ((Interp *) interp)->varFramePtr->clientData;
+    Method *mPtr = contextPtr->callPtr->chain[contextPtr->index].mPtr;
+    Object *declarerPtr;
+    const char *objectName, *kindName;
+    int objectNameLen;
+
+    if (mPtr->declaringObjectPtr != NULL) {
+	declarerPtr = mPtr->declaringObjectPtr;
+	kindName = "object";
+    } else {
+	if (mPtr->declaringClassPtr == NULL) {
+	    Tcl_Panic("method not declared in class or object");
+	}
+	declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	kindName = "class";
+    }
+
+    objectName = Tcl_GetStringFromObj(TclOOObjectName(interp, declarerPtr),
+	    &objectNameLen);
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (%s \"%.*s%s\" constructor line %d)", kindName,
+	    ELLIPSIFY(objectName, objectNameLen), Tcl_GetErrorLine(interp)));
+}
+
+static void
+DestructorErrorHandler(
+    Tcl_Interp *interp,
+    Tcl_Obj *methodNameObj)
+{
+    CallContext *contextPtr = ((Interp *) interp)->varFramePtr->clientData;
+    Method *mPtr = contextPtr->callPtr->chain[contextPtr->index].mPtr;
+    Object *declarerPtr;
+    const char *objectName, *kindName;
+    int objectNameLen;
+
+    if (mPtr->declaringObjectPtr != NULL) {
+	declarerPtr = mPtr->declaringObjectPtr;
+	kindName = "object";
+    } else {
+	if (mPtr->declaringClassPtr == NULL) {
+	    Tcl_Panic("method not declared in class or object");
+	}
+	declarerPtr = mPtr->declaringClassPtr->thisPtr;
+	kindName = "class";
+    }
+
+    objectName = Tcl_GetStringFromObj(TclOOObjectName(interp, declarerPtr),
+	    &objectNameLen);
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (%s \"%.*s%s\" destructor line %d)", kindName,
+	    ELLIPSIFY(objectName, objectNameLen), Tcl_GetErrorLine(interp)));
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * DeleteProcedureMethod, CloneProcedureMethod --
+ *
+ *	How to delete and clone procedure-like methods.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+DeleteProcedureMethodRecord(
+    ProcedureMethod *pmPtr)
+{
+    TclProcDeleteProc(pmPtr->procPtr);
+    if (pmPtr->deleteClientdataProc) {
+	pmPtr->deleteClientdataProc(pmPtr->clientData);
+    }
+    ckfree(pmPtr);
+}
+
+static void
+DeleteProcedureMethod(
+    ClientData clientData)
+{
+    register ProcedureMethod *pmPtr = clientData;
+
+    if (pmPtr->refCount-- <= 1) {
+	DeleteProcedureMethodRecord(pmPtr);
+    }
+}
+
+static int
+CloneProcedureMethod(
+    Tcl_Interp *interp,
+    ClientData clientData,
+    ClientData *newClientData)
+{
+    ProcedureMethod *pmPtr = clientData;
+    ProcedureMethod *pm2Ptr;
+    Tcl_Obj *bodyObj, *argsObj;
+    CompiledLocal *localPtr;
+
+    /*
+     * Copy the argument list.
+     */
+
+    argsObj = Tcl_NewObj();
+    for (localPtr=pmPtr->procPtr->firstLocalPtr; localPtr!=NULL;
+	    localPtr=localPtr->nextPtr) {
+	if (TclIsVarArgument(localPtr)) {
+	    Tcl_Obj *argObj = Tcl_NewObj();
+
+	    Tcl_ListObjAppendElement(NULL, argObj,
+		    Tcl_NewStringObj(localPtr->name, -1));
+	    if (localPtr->defValuePtr != NULL) {
+		Tcl_ListObjAppendElement(NULL, argObj, localPtr->defValuePtr);
+	    }
+	    Tcl_ListObjAppendElement(NULL, argsObj, argObj);
+	}
+    }
+
+    /*
+     * Must strip the internal representation in order to ensure that any
+     * bound references to instance variables are removed. [Bug 3609693]
+     */
+
+    bodyObj = Tcl_DuplicateObj(pmPtr->procPtr->bodyPtr);
+    TclFreeIntRep(bodyObj);
+
+    /*
+     * Create the actual copy of the method record, manufacturing a new proc
+     * record.
+     */
+
+    pm2Ptr = ckalloc(sizeof(ProcedureMethod));
+    memcpy(pm2Ptr, pmPtr, sizeof(ProcedureMethod));
+    pm2Ptr->refCount = 1;
+    Tcl_IncrRefCount(argsObj);
+    Tcl_IncrRefCount(bodyObj);
+    if (TclCreateProc(interp, NULL, "", argsObj, bodyObj,
+	    &pm2Ptr->procPtr) != TCL_OK) {
+	Tcl_DecrRefCount(argsObj);
+	Tcl_DecrRefCount(bodyObj);
+	ckfree(pm2Ptr);
+	return TCL_ERROR;
+    }
+    Tcl_DecrRefCount(argsObj);
+    Tcl_DecrRefCount(bodyObj);
+
+    if (pmPtr->cloneClientdataProc) {
+	pm2Ptr->clientData = pmPtr->cloneClientdataProc(pmPtr->clientData);
+    }
+    *newClientData = pm2Ptr;
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONewForwardInstanceMethod --
+ *
+ *	Create a forwarded method for an object.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Method *
+TclOONewForwardInstanceMethod(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. */
+    Object *oPtr,		/* The object to attach the method to. */
+    int flags,			/* Whether the method is public or not. */
+    Tcl_Obj *nameObj,		/* The name of the method. */
+    Tcl_Obj *prefixObj)		/* List of arguments that form the command
+				 * prefix to forward to. */
+{
+    int prefixLen;
+    register ForwardMethod *fmPtr;
+
+    if (Tcl_ListObjLength(interp, prefixObj, &prefixLen) != TCL_OK) {
+	return NULL;
+    }
+    if (prefixLen < 1) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"method forward prefix must be non-empty", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_FORWARD", NULL);
+	return NULL;
+    }
+
+    fmPtr = ckalloc(sizeof(ForwardMethod));
+    fmPtr->prefixObj = prefixObj;
+    Tcl_IncrRefCount(prefixObj);
+    return (Method *) Tcl_NewInstanceMethod(interp, (Tcl_Object) oPtr,
+	    nameObj, flags, &fwdMethodType, fmPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOONewForwardMethod --
+ *
+ *	Create a new forwarded method for a class.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Method *
+TclOONewForwardMethod(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. */
+    Class *clsPtr,		/* The class to attach the method to. */
+    int flags,			/* Whether the method is public or not. */
+    Tcl_Obj *nameObj,		/* The name of the method. */
+    Tcl_Obj *prefixObj)		/* List of arguments that form the command
+				 * prefix to forward to. */
+{
+    int prefixLen;
+    register ForwardMethod *fmPtr;
+
+    if (Tcl_ListObjLength(interp, prefixObj, &prefixLen) != TCL_OK) {
+	return NULL;
+    }
+    if (prefixLen < 1) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"method forward prefix must be non-empty", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OO", "BAD_FORWARD", NULL);
+	return NULL;
+    }
+
+    fmPtr = ckalloc(sizeof(ForwardMethod));
+    fmPtr->prefixObj = prefixObj;
+    Tcl_IncrRefCount(prefixObj);
+    return (Method *) Tcl_NewMethod(interp, (Tcl_Class) clsPtr, nameObj,
+	    flags, &fwdMethodType, fmPtr);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InvokeForwardMethod --
+ *
+ *	How to invoke a forwarded method. Works by doing some ensemble-like
+ *	command rearranging and then invokes some other Tcl command.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static int
+InvokeForwardMethod(
+    ClientData clientData,	/* Pointer to some per-method context. */
+    Tcl_Interp *interp,
+    Tcl_ObjectContext context,	/* The method calling context. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* Arguments as actually seen. */
+{
+    CallContext *contextPtr = (CallContext *) context;
+    ForwardMethod *fmPtr = clientData;
+    Tcl_Obj **argObjs, **prefixObjs;
+    int numPrefixes, len, skip = contextPtr->skip;
+
+    /*
+     * Build the real list of arguments to use. Note that we know that the
+     * prefixObj field of the ForwardMethod structure holds a reference to a
+     * non-empty list, so there's a whole class of failures ("not a list") we
+     * can ignore here.
+     */
+
+    Tcl_ListObjGetElements(NULL, fmPtr->prefixObj, &numPrefixes, &prefixObjs);
+    argObjs = InitEnsembleRewrite(interp, objc, objv, skip,
+	    numPrefixes, prefixObjs, &len);
+    Tcl_NRAddCallback(interp, FinalizeForwardCall, argObjs, NULL, NULL, NULL);
+    /*
+     * NOTE: The combination of direct set of iPtr->lookupNsPtr and the use
+     * of the TCL_EVAL_NOERR flag results in an evaluation configuration
+     * very much like TCL_EVAL_INVOKE.
+     */
+    ((Interp *)interp)->lookupNsPtr
+	    = (Namespace *) contextPtr->oPtr->namespacePtr;
+    return TclNREvalObjv(interp, len, argObjs, TCL_EVAL_NOERR, NULL);
+}
+
+static int
+FinalizeForwardCall(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Tcl_Obj **argObjs = data[0];
+
+    TclStackFree(interp, argObjs);
+    return result;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * DeleteForwardMethod, CloneForwardMethod --
+ *
+ *	How to delete and clone forwarded methods.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+DeleteForwardMethod(
+    ClientData clientData)
+{
+    ForwardMethod *fmPtr = clientData;
+
+    Tcl_DecrRefCount(fmPtr->prefixObj);
+    ckfree(fmPtr);
+}
+
+static int
+CloneForwardMethod(
+    Tcl_Interp *interp,
+    ClientData clientData,
+    ClientData *newClientData)
+{
+    ForwardMethod *fmPtr = clientData;
+    ForwardMethod *fm2Ptr = ckalloc(sizeof(ForwardMethod));
+
+    fm2Ptr->prefixObj = fmPtr->prefixObj;
+    Tcl_IncrRefCount(fm2Ptr->prefixObj);
+    *newClientData = fm2Ptr;
+    return TCL_OK;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOOGetProcFromMethod, TclOOGetFwdFromMethod --
+ *
+ *	Utility functions used for procedure-like and forwarding method
+ *	introspection.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Proc *
+TclOOGetProcFromMethod(
+    Method *mPtr)
+{
+    if (mPtr->typePtr == &procMethodType) {
+	ProcedureMethod *pmPtr = mPtr->clientData;
+
+	return pmPtr->procPtr;
+    }
+    return NULL;
+}
+
+Tcl_Obj *
+TclOOGetMethodBody(
+    Method *mPtr)
+{
+    if (mPtr->typePtr == &procMethodType) {
+	ProcedureMethod *pmPtr = mPtr->clientData;
+
+	if (pmPtr->procPtr->bodyPtr->bytes == NULL) {
+	    (void) Tcl_GetString(pmPtr->procPtr->bodyPtr);
+	}
+	return pmPtr->procPtr->bodyPtr;
+    }
+    return NULL;
+}
+
+Tcl_Obj *
+TclOOGetFwdFromMethod(
+    Method *mPtr)
+{
+    if (mPtr->typePtr == &fwdMethodType) {
+	ForwardMethod *fwPtr = mPtr->clientData;
+
+	return fwPtr->prefixObj;
+    }
+    return NULL;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * InitEnsembleRewrite --
+ *
+ *	Utility function that wraps up a lot of the complexity involved in
+ *	doing ensemble-like command forwarding. Here is a picture of memory
+ *	management plan:
+ *
+ *                    <-----------------objc---------------------->
+ *      objv:        |=============|===============================|
+ *                    <-toRewrite->           |
+ *                                             \
+ *                    <-rewriteLength->         \
+ *      rewriteObjs: |=================|         \
+ *                           |                    |
+ *                           V                    V
+ *      argObjs:     |=================|===============================|
+ *                    <------------------*lengthPtr------------------->
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static Tcl_Obj **
+InitEnsembleRewrite(
+    Tcl_Interp *interp,		/* Place to log the rewrite info. */
+    int objc,			/* Number of real arguments. */
+    Tcl_Obj *const *objv,	/* The real arguments. */
+    int toRewrite,		/* Number of real arguments to replace. */
+    int rewriteLength,		/* Number of arguments to insert instead. */
+    Tcl_Obj *const *rewriteObjs,/* Arguments to insert instead. */
+    int *lengthPtr)		/* Where to write the resulting length of the
+				 * array of rewritten arguments. */
+{
+    unsigned len = rewriteLength + objc - toRewrite;
+    Tcl_Obj **argObjs = TclStackAlloc(interp, sizeof(Tcl_Obj *) * len);
+
+    memcpy(argObjs, rewriteObjs, rewriteLength * sizeof(Tcl_Obj *));
+    memcpy(argObjs + rewriteLength, objv + toRewrite,
+	    sizeof(Tcl_Obj *) * (objc - toRewrite));
+
+    /*
+     * Now plumb this into the core ensemble rewrite logging system so that
+     * Tcl_WrongNumArgs() can rewrite its result appropriately. The rules for
+     * how to store the rewrite rules get complex solely because of the case
+     * where an ensemble rewrites itself out of the picture; when that
+     * happens, the quality of the error message rewrite falls drastically
+     * (and unavoidably).
+     */
+
+    if (TclInitRewriteEnsemble(interp, toRewrite, rewriteLength, objv)) {
+	TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL, NULL);
+    }
+    *lengthPtr = len;
+    return argObjs;
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * assorted trivial 'getter' functions
+ *
+ * ----------------------------------------------------------------------
+ */
+
+Tcl_Object
+Tcl_MethodDeclarerObject(
+    Tcl_Method method)
+{
+    return (Tcl_Object) ((Method *) method)->declaringObjectPtr;
+}
+
+Tcl_Class
+Tcl_MethodDeclarerClass(
+    Tcl_Method method)
+{
+    return (Tcl_Class) ((Method *) method)->declaringClassPtr;
+}
+
+Tcl_Obj *
+Tcl_MethodName(
+    Tcl_Method method)
+{
+    return ((Method *) method)->namePtr;
+}
+
+int
+Tcl_MethodIsType(
+    Tcl_Method method,
+    const Tcl_MethodType *typePtr,
+    ClientData *clientDataPtr)
+{
+    Method *mPtr = (Method *) method;
+
+    if (mPtr->typePtr == typePtr) {
+	if (clientDataPtr != NULL) {
+	    *clientDataPtr = mPtr->clientData;
+	}
+	return 1;
+    }
+    return 0;
+}
+
+int
+Tcl_MethodIsPublic(
+    Tcl_Method method)
+{
+    return (((Method *)method)->flags & PUBLIC_METHOD) ? 1 : 0;
+}
+
+/*
+ * Extended method construction for itcl-ng.
+ */
+
+Tcl_Method
+TclOONewProcInstanceMethodEx(
+    Tcl_Interp *interp,		/* The interpreter containing the object. */
+    Tcl_Object oPtr,		/* The object to modify. */
+    TclOO_PreCallProc *preCallPtr,
+    TclOO_PostCallProc *postCallPtr,
+    ProcErrorProc *errProc,
+    ClientData clientData,
+    Tcl_Obj *nameObj,		/* The name of the method, which must not be
+				 * NULL. */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which must not be NULL. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which must not be
+				 * NULL. */
+    int flags,			/* Whether this is a public method. */
+    void **internalTokenPtr)	/* If non-NULL, points to a variable that gets
+				 * the reference to the ProcedureMethod
+				 * structure. */
+{
+    ProcedureMethod *pmPtr;
+    Tcl_Method method = (Tcl_Method) TclOONewProcInstanceMethod(interp,
+	    (Object *) oPtr, flags, nameObj, argsObj, bodyObj, &pmPtr);
+
+    if (method == NULL) {
+	return NULL;
+    }
+    pmPtr->flags = flags & USE_DECLARER_NS;
+    pmPtr->preCallProc = preCallPtr;
+    pmPtr->postCallProc = postCallPtr;
+    pmPtr->errProc = errProc;
+    pmPtr->clientData = clientData;
+    if (internalTokenPtr != NULL) {
+	*internalTokenPtr = pmPtr;
+    }
+    return method;
+}
+
+Tcl_Method
+TclOONewProcMethodEx(
+    Tcl_Interp *interp,		/* The interpreter containing the class. */
+    Tcl_Class clsPtr,		/* The class to modify. */
+    TclOO_PreCallProc *preCallPtr,
+    TclOO_PostCallProc *postCallPtr,
+    ProcErrorProc *errProc,
+    ClientData clientData,
+    Tcl_Obj *nameObj,		/* The name of the method, which may be NULL;
+				 * if so, up to caller to manage storage
+				 * (e.g., because it is a constructor or
+				 * destructor). */
+    Tcl_Obj *argsObj,		/* The formal argument list for the method,
+				 * which may be NULL; if so, it is equivalent
+				 * to an empty list. */
+    Tcl_Obj *bodyObj,		/* The body of the method, which must not be
+				 * NULL. */
+    int flags,			/* Whether this is a public method. */
+    void **internalTokenPtr)	/* If non-NULL, points to a variable that gets
+				 * the reference to the ProcedureMethod
+				 * structure. */
+{
+    ProcedureMethod *pmPtr;
+    Tcl_Method method = (Tcl_Method) TclOONewProcMethod(interp,
+	    (Class *) clsPtr, flags, nameObj, argsObj, bodyObj, &pmPtr);
+
+    if (method == NULL) {
+	return NULL;
+    }
+    pmPtr->flags = flags & USE_DECLARER_NS;
+    pmPtr->preCallProc = preCallPtr;
+    pmPtr->postCallProc = postCallPtr;
+    pmPtr->errProc = errProc;
+    pmPtr->clientData = clientData;
+    if (internalTokenPtr != NULL) {
+	*internalTokenPtr = pmPtr;
+    }
+    return method;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclOOStubInit.c b/generic/tclOOStubInit.c
new file mode 100644
index 0000000..900ab22
--- /dev/null
+++ b/generic/tclOOStubInit.c
@@ -0,0 +1,78 @@
+/*
+ * This file is (mostly) automatically generated from tclOO.decls.
+ * It is compiled and linked in with the tclOO package proper.
+ */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include "tclOOInt.h"
+
+MODULE_SCOPE const TclOOStubs tclOOStubs;
+
+#ifdef __GNUC__
+#pragma GCC dependency "tclOO.decls"
+#endif
+
+/* !BEGIN!: Do not edit below this line. */
+
+static const TclOOIntStubs tclOOIntStubs = {
+    TCL_STUB_MAGIC,
+    0,
+    TclOOGetDefineCmdContext, /* 0 */
+    TclOOMakeProcInstanceMethod, /* 1 */
+    TclOOMakeProcMethod, /* 2 */
+    TclOONewProcInstanceMethod, /* 3 */
+    TclOONewProcMethod, /* 4 */
+    TclOOObjectCmdCore, /* 5 */
+    TclOOIsReachable, /* 6 */
+    TclOONewForwardMethod, /* 7 */
+    TclOONewForwardInstanceMethod, /* 8 */
+    TclOONewProcInstanceMethodEx, /* 9 */
+    TclOONewProcMethodEx, /* 10 */
+    TclOOInvokeObject, /* 11 */
+    TclOOObjectSetFilters, /* 12 */
+    TclOOClassSetFilters, /* 13 */
+    TclOOObjectSetMixins, /* 14 */
+    TclOOClassSetMixins, /* 15 */
+};
+
+static const TclOOStubHooks tclOOStubHooks = {
+    &tclOOIntStubs
+};
+
+const TclOOStubs tclOOStubs = {
+    TCL_STUB_MAGIC,
+    &tclOOStubHooks,
+    Tcl_CopyObjectInstance, /* 0 */
+    Tcl_GetClassAsObject, /* 1 */
+    Tcl_GetObjectAsClass, /* 2 */
+    Tcl_GetObjectCommand, /* 3 */
+    Tcl_GetObjectFromObj, /* 4 */
+    Tcl_GetObjectNamespace, /* 5 */
+    Tcl_MethodDeclarerClass, /* 6 */
+    Tcl_MethodDeclarerObject, /* 7 */
+    Tcl_MethodIsPublic, /* 8 */
+    Tcl_MethodIsType, /* 9 */
+    Tcl_MethodName, /* 10 */
+    Tcl_NewInstanceMethod, /* 11 */
+    Tcl_NewMethod, /* 12 */
+    Tcl_NewObjectInstance, /* 13 */
+    Tcl_ObjectDeleted, /* 14 */
+    Tcl_ObjectContextIsFiltering, /* 15 */
+    Tcl_ObjectContextMethod, /* 16 */
+    Tcl_ObjectContextObject, /* 17 */
+    Tcl_ObjectContextSkippedArgs, /* 18 */
+    Tcl_ClassGetMetadata, /* 19 */
+    Tcl_ClassSetMetadata, /* 20 */
+    Tcl_ObjectGetMetadata, /* 21 */
+    Tcl_ObjectSetMetadata, /* 22 */
+    Tcl_ObjectContextInvokeNext, /* 23 */
+    Tcl_ObjectGetMethodNameMapper, /* 24 */
+    Tcl_ObjectSetMethodNameMapper, /* 25 */
+    Tcl_ClassSetConstructor, /* 26 */
+    Tcl_ClassSetDestructor, /* 27 */
+    Tcl_GetObjectName, /* 28 */
+};
+
+/* !END!: Do not edit above this line. */
diff --git a/generic/tclOOStubLib.c b/generic/tclOOStubLib.c
new file mode 100644
index 0000000..a9fa212
--- /dev/null
+++ b/generic/tclOOStubLib.c
@@ -0,0 +1,71 @@
+/*
+ * ORIGINAL SOURCE: tk/generic/tkStubLib.c, version 1.9 2004/03/17
+ */
+
+#include "tclOOInt.h"
+
+MODULE_SCOPE const TclOOStubs *tclOOStubsPtr;
+MODULE_SCOPE const TclOOIntStubs *tclOOIntStubsPtr;
+
+const TclOOStubs *tclOOStubsPtr = NULL;
+const TclOOIntStubs *tclOOIntStubsPtr = NULL;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclOOInitializeStubs --
+ *	Load the tclOO package, initialize stub table pointer. Do not call
+ *	this function directly, use Tcl_OOInitStubs() macro instead.
+ *
+ * Results:
+ *	The actual version of the package that satisfies the request, or NULL
+ *	to indicate that an error occurred.
+ *
+ * Side effects:
+ *	Sets the stub table pointers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef TclOOInitializeStubs
+
+MODULE_SCOPE const char *
+TclOOInitializeStubs(
+    Tcl_Interp *interp,
+    const char *version)
+{
+    int exact = 0;
+    const char *packageName = "TclOO";
+    const char *errMsg = NULL;
+    TclOOStubs *stubsPtr = NULL;
+    const char *actualVersion = tclStubsPtr->tcl_PkgRequireEx(interp,
+	    packageName, version, exact, &stubsPtr);
+
+    if (actualVersion == NULL) {
+	return NULL;
+    }
+    if (stubsPtr == NULL) {
+	errMsg = "missing stub table pointer";
+    } else {
+	tclOOStubsPtr = stubsPtr;
+	if (stubsPtr->hooks) {
+	    tclOOIntStubsPtr = stubsPtr->hooks->tclOOIntStubs;
+	} else {
+	    tclOOIntStubsPtr = NULL;
+	}
+	return actualVersion;
+    }
+    tclStubsPtr->tcl_ResetResult(interp);
+    tclStubsPtr->tcl_AppendResult(interp, "Error loading ", packageName,
+	    " (requested version ", version, ", actual version ",
+	    actualVersion, "): ", errMsg, NULL);
+    return NULL;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclObj.c b/generic/tclObj.c
new file mode 100644
index 0000000..f4b81f2
--- /dev/null
+++ b/generic/tclObj.c
@@ -0,0 +1,4508 @@
+/*
+ * tclObj.c --
+ *
+ *	This file contains Tcl object-related functions that are used by many
+ *	Tcl commands.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1999 by Scriptics Corporation.
+ * Copyright (c) 2001 by ActiveState Corporation.
+ * Copyright (c) 2005 by Kevin B. Kenny.  All rights reserved.
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+#include <math.h>
+
+/*
+ * Table of all object types.
+ */
+
+static Tcl_HashTable typeTable;
+static int typeTableInitialized = 0;	/* 0 means not yet initialized. */
+TCL_DECLARE_MUTEX(tableMutex)
+
+/*
+ * Head of the list of free Tcl_Obj structs we maintain.
+ */
+
+Tcl_Obj *tclFreeObjList = NULL;
+
+/*
+ * The object allocator is single threaded. This mutex is referenced by the
+ * TclNewObj macro, however, so must be visible.
+ */
+
+#ifdef TCL_THREADS
+MODULE_SCOPE Tcl_Mutex tclObjMutex;
+Tcl_Mutex tclObjMutex;
+#endif
+
+/*
+ * Pointer to a heap-allocated string of length zero that the Tcl core uses as
+ * the value of an empty string representation for an object. This value is
+ * shared by all new objects allocated by Tcl_NewObj.
+ */
+
+char tclEmptyString = '\0';
+
+#if defined(TCL_MEM_DEBUG) && defined(TCL_THREADS)
+/*
+ * Structure for tracking the source file and line number where a given
+ * Tcl_Obj was allocated.  We also track the pointer to the Tcl_Obj itself,
+ * for sanity checking purposes.
+ */
+
+typedef struct ObjData {
+    Tcl_Obj *objPtr;		/* The pointer to the allocated Tcl_Obj. */
+    const char *file;		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line;			/* Line number in the source file; used for
+				 * debugging. */
+} ObjData;
+#endif /* TCL_MEM_DEBUG && TCL_THREADS */
+
+/*
+ * All static variables used in this file are collected into a single instance
+ * of the following structure.  For multi-threaded implementations, there is
+ * one instance of this structure for each thread.
+ *
+ * Notice that different structures with the same name appear in other files.
+ * The structure defined below is used in this file only.
+ */
+
+typedef struct {
+    Tcl_HashTable *lineCLPtr;   /* This table remembers for each Tcl_Obj
+                                 * generated by a call to the function
+                                 * TclSubstTokens() from a literal text
+                                 * where bs+nl sequences occured in it, if
+                                 * any. I.e. this table keeps track of
+                                 * invisible and stripped continuation lines.
+                                 * Its keys are Tcl_Obj pointers, the values
+                                 * are ContLineLoc pointers. See the file
+                                 * tclCompile.h for the definition of this
+                                 * structure, and for references to all
+                                 * related places in the core. */
+#if defined(TCL_MEM_DEBUG) && defined(TCL_THREADS)
+    Tcl_HashTable *objThreadMap;/* Thread local table that is used to check
+                                 * that a Tcl_Obj was not allocated by some
+                                 * other thread. */
+#endif /* TCL_MEM_DEBUG && TCL_THREADS */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+static void             TclThreadFinalizeContLines(ClientData clientData);
+static ThreadSpecificData *TclGetContLineTable(void);
+
+/*
+ * Nested Tcl_Obj deletion management support
+ *
+ * All context references used in the object freeing code are pointers to this
+ * structure; every thread will have its own structure instance. The purpose
+ * of this structure is to allow deeply nested collections of Tcl_Objs to be
+ * freed without taking a vast depth of C stack (which could cause all sorts
+ * of breakage.)
+ */
+
+typedef struct PendingObjData {
+    int deletionCount;		/* Count of the number of invokations of
+				 * TclFreeObj() are on the stack (at least
+				 * conceptually; many are actually expanded
+				 * macros). */
+    Tcl_Obj *deletionStack;	/* Stack of objects that have had TclFreeObj()
+				 * invoked upon them but which can't be
+				 * deleted yet because they are in a nested
+				 * invokation of TclFreeObj(). By postponing
+				 * this way, we limit the maximum overall C
+				 * stack depth when deleting a complex object.
+				 * The down-side is that we alter the overall
+				 * behaviour by altering the order in which
+				 * objects are deleted, and we change the
+				 * order in which the string rep and the
+				 * internal rep of an object are deleted. Note
+				 * that code which assumes the previous
+				 * behaviour in either of these respects is
+				 * unsafe anyway; it was never documented as
+				 * to exactly what would happen in these
+				 * cases, and the overall contract of a
+				 * user-level Tcl_DecrRefCount() is still
+				 * preserved (assuming that a particular T_DRC
+				 * would delete an object is not very
+				 * safe). */
+} PendingObjData;
+
+/*
+ * These are separated out so that some semantic content is attached
+ * to them.
+ */
+#define ObjDeletionLock(contextPtr)	((contextPtr)->deletionCount++)
+#define ObjDeletionUnlock(contextPtr)	((contextPtr)->deletionCount--)
+#define ObjDeletePending(contextPtr)	((contextPtr)->deletionCount > 0)
+#define ObjOnStack(contextPtr)		((contextPtr)->deletionStack != NULL)
+#define PushObjToDelete(contextPtr,objPtr) \
+    /* The string rep is already invalidated so we can use the bytes value \
+     * for our pointer chain: push onto the head of the stack. */       \
+    (objPtr)->bytes = (char *) ((contextPtr)->deletionStack);           \
+    (contextPtr)->deletionStack = (objPtr)
+#define PopObjToDelete(contextPtr,objPtrVar) \
+    (objPtrVar) = (contextPtr)->deletionStack;                          \
+    (contextPtr)->deletionStack = (Tcl_Obj *) (objPtrVar)->bytes
+
+/*
+ * Macro to set up the local reference to the deletion context.
+ */
+#ifndef TCL_THREADS
+static PendingObjData pendingObjData;
+#define ObjInitDeletionContext(contextPtr) \
+    PendingObjData *const contextPtr = &pendingObjData
+#elif HAVE_FAST_TSD
+static __thread PendingObjData pendingObjData;
+#define ObjInitDeletionContext(contextPtr) \
+    PendingObjData *const contextPtr = &pendingObjData
+#else
+static Tcl_ThreadDataKey pendingObjDataKey;
+#define ObjInitDeletionContext(contextPtr) \
+    PendingObjData *const contextPtr =                                  \
+	    Tcl_GetThreadData(&pendingObjDataKey, sizeof(PendingObjData))
+#endif
+
+/*
+ * Macros to pack/unpack a bignum's fields in a Tcl_Obj internal rep
+ */
+
+#define PACK_BIGNUM(bignum, objPtr) \
+    if ((bignum).used > 0x7fff) {                                       \
+	mp_int *temp = (void *) ckalloc((unsigned) sizeof(mp_int));     \
+	*temp = bignum;                                                 \
+	(objPtr)->internalRep.twoPtrValue.ptr1 = temp;                 \
+	(objPtr)->internalRep.twoPtrValue.ptr2 = INT2PTR(-1); \
+    } else {                                                            \
+	if ((bignum).alloc > 0x7fff) {                                  \
+	    mp_shrink(&(bignum));                                       \
+	}                                                               \
+	(objPtr)->internalRep.twoPtrValue.ptr1 = (void *) (bignum).dp; \
+	(objPtr)->internalRep.twoPtrValue.ptr2 = INT2PTR( ((bignum).sign << 30) \
+		| ((bignum).alloc << 15) | ((bignum).used));            \
+    }
+
+#define UNPACK_BIGNUM(objPtr, bignum) \
+    if ((objPtr)->internalRep.twoPtrValue.ptr2 == INT2PTR(-1)) { \
+	(bignum) = *((mp_int *) ((objPtr)->internalRep.twoPtrValue.ptr1)); \
+    } else {                                                            \
+	(bignum).dp = (objPtr)->internalRep.twoPtrValue.ptr1;          \
+	(bignum).sign = PTR2INT((objPtr)->internalRep.twoPtrValue.ptr2) >> 30; \
+	(bignum).alloc =                                                \
+		(PTR2INT((objPtr)->internalRep.twoPtrValue.ptr2) >> 15) & 0x7fff; \
+	(bignum).used = PTR2INT((objPtr)->internalRep.twoPtrValue.ptr2) & 0x7fff; \
+    }
+
+/*
+ * Prototypes for functions defined later in this file:
+ */
+
+static int		ParseBoolean(Tcl_Obj *objPtr);
+static int		SetDoubleFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static int		SetIntFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		UpdateStringOfDouble(Tcl_Obj *objPtr);
+static void		UpdateStringOfInt(Tcl_Obj *objPtr);
+#ifndef TCL_WIDE_INT_IS_LONG
+static void		UpdateStringOfWideInt(Tcl_Obj *objPtr);
+static int		SetWideIntFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+#endif
+static void		FreeBignum(Tcl_Obj *objPtr);
+static void		DupBignum(Tcl_Obj *objPtr, Tcl_Obj *copyPtr);
+static void		UpdateStringOfBignum(Tcl_Obj *objPtr);
+static int		GetBignumFromObj(Tcl_Interp *interp, Tcl_Obj *objPtr,
+			    int copy, mp_int *bignumValue);
+
+/*
+ * Prototypes for the array hash key methods.
+ */
+
+static Tcl_HashEntry *	AllocObjEntry(Tcl_HashTable *tablePtr, void *keyPtr);
+
+/*
+ * Prototypes for the CommandName object type.
+ */
+
+static void		DupCmdNameInternalRep(Tcl_Obj *objPtr,
+			    Tcl_Obj *copyPtr);
+static void		FreeCmdNameInternalRep(Tcl_Obj *objPtr);
+static int		SetCmdNameFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+
+/*
+ * The structures below defines the Tcl object types defined in this file by
+ * means of functions that can be invoked by generic object code. See also
+ * tclStringObj.c, tclListObj.c, tclByteCode.c for other type manager
+ * implementations.
+ */
+
+static const Tcl_ObjType oldBooleanType = {
+    "boolean",			/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    TclSetBooleanFromAny		/* setFromAnyProc */
+};
+const Tcl_ObjType tclBooleanType = {
+    "booleanString",		/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    TclSetBooleanFromAny		/* setFromAnyProc */
+};
+const Tcl_ObjType tclDoubleType = {
+    "double",			/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    UpdateStringOfDouble,	/* updateStringProc */
+    SetDoubleFromAny		/* setFromAnyProc */
+};
+const Tcl_ObjType tclIntType = {
+    "int",			/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    UpdateStringOfInt,		/* updateStringProc */
+    SetIntFromAny		/* setFromAnyProc */
+};
+#ifndef TCL_WIDE_INT_IS_LONG
+const Tcl_ObjType tclWideIntType = {
+    "wideInt",			/* name */
+    NULL,			/* freeIntRepProc */
+    NULL,			/* dupIntRepProc */
+    UpdateStringOfWideInt,	/* updateStringProc */
+    SetWideIntFromAny		/* setFromAnyProc */
+};
+#endif
+const Tcl_ObjType tclBignumType = {
+    "bignum",			/* name */
+    FreeBignum,			/* freeIntRepProc */
+    DupBignum,			/* dupIntRepProc */
+    UpdateStringOfBignum,	/* updateStringProc */
+    NULL			/* setFromAnyProc */
+};
+
+/*
+ * The structure below defines the Tcl obj hash key type.
+ */
+
+const Tcl_HashKeyType tclObjHashKeyType = {
+    TCL_HASH_KEY_TYPE_VERSION,	/* version */
+    0,				/* flags */
+    TclHashObjKey,		/* hashKeyProc */
+    TclCompareObjKeys,		/* compareKeysProc */
+    AllocObjEntry,		/* allocEntryProc */
+    TclFreeObjEntry		/* freeEntryProc */
+};
+
+/*
+ * The structure below defines the command name Tcl object type by means of
+ * functions that can be invoked by generic object code. Objects of this type
+ * cache the Command pointer that results from looking up command names in the
+ * command hashtable. Such objects appear as the zeroth ("command name")
+ * argument in a Tcl command.
+ *
+ * NOTE: the ResolvedCmdName that gets cached is stored in the
+ * twoPtrValue.ptr1 field, and the twoPtrValue.ptr2 field is unused. You might
+ * think you could use the simpler otherValuePtr field to store the single
+ * ResolvedCmdName pointer, but DO NOT DO THIS. It seems that some extensions
+ * use the second internal pointer field of the twoPtrValue field for their
+ * own purposes.
+ *
+ * TRICKY POINT! Some extensions update this structure! (Notably, these
+ * include TclBlend and TCom). This is highly ill-advised on their part, but
+ * does allow them to delete a command when references to it are gone, which
+ * is fragile but useful given their somewhat-OO style. Because of this, this
+ * structure MUST NOT be const so that the C compiler puts the data in
+ * writable memory. [Bug 2558422] [Bug 07d13d99b0a9]
+ * TODO: Provide a better API for those extensions so that they can coexist...
+ */
+
+Tcl_ObjType tclCmdNameType = {
+    "cmdName",			/* name */
+    FreeCmdNameInternalRep,	/* freeIntRepProc */
+    DupCmdNameInternalRep,	/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    SetCmdNameFromAny		/* setFromAnyProc */
+};
+
+/*
+ * Structure containing a cached pointer to a command that is the result of
+ * resolving the command's name in some namespace. It is the internal
+ * representation for a cmdName object. It contains the pointer along with
+ * some information that is used to check the pointer's validity.
+ */
+
+typedef struct ResolvedCmdName {
+    Command *cmdPtr;		/* A cached Command pointer. */
+    Namespace *refNsPtr;	/* Points to the namespace containing the
+				 * reference (not the namespace that contains
+				 * the referenced command). NULL if the name
+				 * is fully qualified.*/
+    size_t refNsId;		/* refNsPtr's unique namespace id. Used to
+				 * verify that refNsPtr is still valid (e.g.,
+				 * it's possible that the cmd's containing
+				 * namespace was deleted and a new one created
+				 * at the same address). */
+    size_t refNsCmdEpoch;	/* Value of the referencing namespace's
+				 * cmdRefEpoch when the pointer was cached.
+				 * Before using the cached pointer, we check
+				 * if the namespace's epoch was incremented;
+				 * if so, this cached pointer is invalid. */
+    size_t cmdEpoch;		/* Value of the command's cmdEpoch when this
+				 * pointer was cached. Before using the cached
+				 * pointer, we check if the cmd's epoch was
+				 * incremented; if so, the cmd was renamed,
+				 * deleted, hidden, or exposed, and so the
+				 * pointer is invalid. */
+    size_t refCount;		/* Reference count: 1 for each cmdName object
+				 * that has a pointer to this ResolvedCmdName
+				 * structure as its internal rep. This
+				 * structure can be freed when refCount
+				 * becomes zero. */
+} ResolvedCmdName;
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclInitObjectSubsystem --
+ *
+ *	This function is invoked to perform once-only initialization of the
+ *	type table. It also registers the object types defined in this file.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Initializes the table of defined object types "typeTable" with builtin
+ *	object types defined in this file.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+TclInitObjSubsystem(void)
+{
+    Tcl_MutexLock(&tableMutex);
+    typeTableInitialized = 1;
+    Tcl_InitHashTable(&typeTable, TCL_STRING_KEYS);
+    Tcl_MutexUnlock(&tableMutex);
+
+    Tcl_RegisterObjType(&tclByteArrayType);
+    Tcl_RegisterObjType(&tclDoubleType);
+    Tcl_RegisterObjType(&tclEndOffsetType);
+    Tcl_RegisterObjType(&tclIntType);
+    Tcl_RegisterObjType(&tclStringType);
+    Tcl_RegisterObjType(&tclListType);
+    Tcl_RegisterObjType(&tclDictType);
+    Tcl_RegisterObjType(&tclByteCodeType);
+    Tcl_RegisterObjType(&tclCmdNameType);
+    Tcl_RegisterObjType(&tclRegexpType);
+    Tcl_RegisterObjType(&tclProcBodyType);
+
+    /* For backward compatibility only ... */
+    Tcl_RegisterObjType(&oldBooleanType);
+#ifndef TCL_WIDE_INT_IS_LONG
+    Tcl_RegisterObjType(&tclWideIntType);
+#endif
+
+#ifdef TCL_COMPILE_STATS
+    Tcl_MutexLock(&tclObjMutex);
+    tclObjsAlloced = 0;
+    tclObjsFreed = 0;
+    {
+	int i;
+
+	for (i=0 ; i<TCL_MAX_SHARED_OBJ_STATS ; i++) {
+	    tclObjsShared[i] = 0;
+	}
+    }
+    Tcl_MutexUnlock(&tclObjMutex);
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadObjects --
+ *
+ *	This function is called by Tcl_FinalizeThread to clean up thread
+ *	specific Tcl_Obj information.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadObjects(void)
+{
+#if defined(TCL_MEM_DEBUG) && defined(TCL_THREADS)
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch hSearch;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    Tcl_HashTable *tablePtr = tsdPtr->objThreadMap;
+
+    if (tablePtr != NULL) {
+	for (hPtr = Tcl_FirstHashEntry(tablePtr, &hSearch);
+		hPtr != NULL; hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    ObjData *objData = Tcl_GetHashValue(hPtr);
+
+	    if (objData != NULL) {
+		ckfree(objData);
+	    }
+	}
+
+	Tcl_DeleteHashTable(tablePtr);
+	ckfree(tablePtr);
+	tsdPtr->objThreadMap = NULL;
+    }
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeObjects --
+ *
+ *	This function is called by Tcl_Finalize to clean up all registered
+ *	Tcl_ObjType's and to reset the tclFreeObjList.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeObjects(void)
+{
+    Tcl_MutexLock(&tableMutex);
+    if (typeTableInitialized) {
+	Tcl_DeleteHashTable(&typeTable);
+	typeTableInitialized = 0;
+    }
+    Tcl_MutexUnlock(&tableMutex);
+
+    /*
+     * All we do here is reset the head pointer of the linked list of free
+     * Tcl_Obj's to NULL; the memory finalization will take care of releasing
+     * memory for us.
+     */
+    Tcl_MutexLock(&tclObjMutex);
+    tclFreeObjList = NULL;
+    Tcl_MutexUnlock(&tclObjMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetContLineTable --
+ *
+ *	This procedure is a helper which returns the thread-specific
+ *	hash-table used to track continuation line information associated with
+ *	Tcl_Obj*, and the objThreadMap, etc.
+ *
+ * Results:
+ *	A reference to the thread-data.
+ *
+ * Side effects:
+ *	May allocate memory for the thread-data.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+static ThreadSpecificData *
+TclGetContLineTable(void)
+{
+    /*
+     * Initialize the hashtable tracking invisible continuation lines.  For
+     * the release we use a thread exit handler to ensure that this is done
+     * before TSD blocks are made invalid. The TclFinalizeObjects() which
+     * would be the natural place for this is invoked afterwards, meaning that
+     * we try to operate on a data structure already gone.
+     */
+
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (!tsdPtr->lineCLPtr) {
+	tsdPtr->lineCLPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_InitHashTable(tsdPtr->lineCLPtr, TCL_ONE_WORD_KEYS);
+	Tcl_CreateThreadExitHandler(TclThreadFinalizeContLines,NULL);
+    }
+    return tsdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclContinuationsEnter --
+ *
+ *	This procedure is a helper which saves the continuation line
+ *	information associated with a Tcl_Obj*.
+ *
+ * Results:
+ *	A reference to the newly created continuation line location table.
+ *
+ * Side effects:
+ *	Allocates memory for the table of continuation line locations.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+ContLineLoc *
+TclContinuationsEnter(
+    Tcl_Obj *objPtr,
+    int num,
+    int *loc)
+{
+    int newEntry;
+    ThreadSpecificData *tsdPtr = TclGetContLineTable();
+    Tcl_HashEntry *hPtr =
+	    Tcl_CreateHashEntry(tsdPtr->lineCLPtr, objPtr, &newEntry);
+    ContLineLoc *clLocPtr = ckalloc(sizeof(ContLineLoc) + num*sizeof(int));
+
+    if (!newEntry) {
+	/*
+	 * We're entering ContLineLoc data for the same value more than one
+	 * time. Taking care not to leak the old entry.
+	 *
+	 * This can happen when literals in a proc body are shared. See for
+	 * example test info-30.19 where the action (code) for all branches of
+	 * the switch command is identical, mapping them all to the same
+	 * literal. An interesting result of this is that the number and
+	 * locations (offset) of invisible continuation lines in the literal
+	 * are the same for all occurences.
+	 *
+	 * Note that while reusing the existing entry is possible it requires
+	 * the same actions as for a new entry because we have to copy the
+	 * incoming num/loc data even so. Because we are called from
+	 * TclContinuationsEnterDerived for this case, which modified the
+	 * stored locations (Rebased to the proper relative offset). Just
+	 * returning the stored entry would rebase them a second time, or
+	 * more, hosing the data. It is easier to simply replace, as we are
+	 * doing.
+	 */
+
+	ckfree(Tcl_GetHashValue(hPtr));
+    }
+
+    clLocPtr->num = num;
+    memcpy(&clLocPtr->loc, loc, num*sizeof(int));
+    clLocPtr->loc[num] = CLL_END;       /* Sentinel */
+    Tcl_SetHashValue(hPtr, clLocPtr);
+
+    return clLocPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclContinuationsEnterDerived --
+ *
+ *	This procedure is a helper which computes the continuation line
+ *	information associated with a Tcl_Obj* cut from the middle of a
+ *	script.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory for the table of continuation line locations.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclContinuationsEnterDerived(
+    Tcl_Obj *objPtr,
+    int start,
+    int *clNext)
+{
+    int length, end, num;
+    int *wordCLLast = clNext;
+
+    /*
+     * We have to handle invisible continuations lines here as well, despite
+     * the code we have in TclSubstTokens (TST) for that. Why ?  Nesting. If
+     * our script is the sole argument to an 'eval' command, for example, the
+     * scriptCLLocPtr we are using was generated by a previous call to TST,
+     * and while the words we have here may contain continuation lines they
+     * are invisible already, and the inner call to TST had no bs+nl sequences
+     * to trigger its code.
+     *
+     * Luckily for us, the table we have to create here for the current word
+     * has to be a slice of the table currently in use, with the locations
+     * suitably modified to be relative to the start of the word instead of
+     * relative to the script.
+     *
+     * That is what we are doing now. Determine the slice we need, and if not
+     * empty, wrap it into a new table, and save the result into our
+     * thread-global hashtable, as usual.
+     */
+
+    /*
+     * First compute the range of the word within the script. (Is there a
+     * better way which doesn't shimmer?)
+     */
+
+    TclGetStringFromObj(objPtr, &length);
+    end = start + length;       /* First char after the word */
+
+    /*
+     * Then compute the table slice covering the range of the word.
+     */
+
+    while (*wordCLLast >= 0 && *wordCLLast < end) {
+	wordCLLast++;
+    }
+
+    /*
+     * And generate the table from the slice, if it was not empty.
+     */
+
+    num = wordCLLast - clNext;
+    if (num) {
+	int i;
+	ContLineLoc *clLocPtr = TclContinuationsEnter(objPtr, num, clNext);
+
+	/*
+	 * Re-base the locations.
+	 */
+
+	for (i=0 ; i<num ; i++) {
+	    clLocPtr->loc[i] -= start;
+
+	    /*
+	     * Continuation lines coming before the string and affecting us
+	     * should not happen, due to the proper maintenance of clNext
+	     * during compilation.
+	     */
+
+	    if (clLocPtr->loc[i] < 0) {
+		Tcl_Panic("Derived ICL data for object using offsets from before the script");
+	    }
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclContinuationsCopy --
+ *
+ *	This procedure is a helper which copies the continuation line
+ *	information associated with a Tcl_Obj* to another Tcl_Obj*. It is
+ *	assumed that both contain the same string/script. Use this when a
+ *	script is duplicated because it was shared.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory for the table of continuation line locations.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+void
+TclContinuationsCopy(
+    Tcl_Obj *objPtr,
+    Tcl_Obj *originObjPtr)
+{
+    ThreadSpecificData *tsdPtr = TclGetContLineTable();
+    Tcl_HashEntry *hPtr =
+            Tcl_FindHashEntry(tsdPtr->lineCLPtr, originObjPtr);
+
+    if (hPtr) {
+	ContLineLoc *clLocPtr = Tcl_GetHashValue(hPtr);
+
+	TclContinuationsEnter(objPtr, clLocPtr->num, clLocPtr->loc);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclContinuationsGet --
+ *
+ *	This procedure is a helper which retrieves the continuation line
+ *	information associated with a Tcl_Obj*, if it has any.
+ *
+ * Results:
+ *	A reference to the continuation line location table, or NULL if the
+ *	Tcl_Obj* has no such information associated with it.
+ *
+ * Side effects:
+ *	None.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+ContLineLoc *
+TclContinuationsGet(
+    Tcl_Obj *objPtr)
+{
+    ThreadSpecificData *tsdPtr = TclGetContLineTable();
+    Tcl_HashEntry *hPtr =
+            Tcl_FindHashEntry(tsdPtr->lineCLPtr, objPtr);
+
+    if (!hPtr) {
+        return NULL;
+    }
+    return Tcl_GetHashValue(hPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadFinalizeContLines --
+ *
+ *	This procedure is a helper which releases all continuation line
+ *	information currently known. It is run as a thread exit handler.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Releases memory.
+ *
+ * TIP #280
+ *----------------------------------------------------------------------
+ */
+
+static void
+TclThreadFinalizeContLines(
+    ClientData clientData)
+{
+    /*
+     * Release the hashtable tracking invisible continuation lines.
+     */
+
+    ThreadSpecificData *tsdPtr = TclGetContLineTable();
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch hSearch;
+
+    for (hPtr = Tcl_FirstHashEntry(tsdPtr->lineCLPtr, &hSearch);
+	    hPtr != NULL; hPtr = Tcl_NextHashEntry(&hSearch)) {
+	ckfree(Tcl_GetHashValue(hPtr));
+	Tcl_DeleteHashEntry(hPtr);
+    }
+    Tcl_DeleteHashTable(tsdPtr->lineCLPtr);
+    ckfree(tsdPtr->lineCLPtr);
+    tsdPtr->lineCLPtr = NULL;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_RegisterObjType --
+ *
+ *	This function is called to register a new Tcl object type in the table
+ *	of all object types supported by Tcl.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The type is registered in the Tcl type table. If there was already a
+ *	type with the same name as in typePtr, it is replaced with the new
+ *	type.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+Tcl_RegisterObjType(
+    const Tcl_ObjType *typePtr)	/* Information about object type; storage must
+				 * be statically allocated (must live
+				 * forever). */
+{
+    int isNew;
+
+    Tcl_MutexLock(&tableMutex);
+    Tcl_SetHashValue(
+	    Tcl_CreateHashEntry(&typeTable, typePtr->name, &isNew), typePtr);
+    Tcl_MutexUnlock(&tableMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendAllObjTypes --
+ *
+ *	This function appends onto the argument object the name of each object
+ *	type as a list element. This includes the builtin object types (e.g.
+ *	int, list) as well as those added using Tcl_NewObj. These names can be
+ *	used, for example, with Tcl_GetObjType to get pointers to the
+ *	corresponding Tcl_ObjType structures.
+ *
+ * Results:
+ *	The return value is normally TCL_OK; in this case the object
+ *	referenced by objPtr has each type name appended to it. If an error
+ *	occurs, TCL_ERROR is returned and the interpreter's result holds an
+ *	error message.
+ *
+ * Side effects:
+ *	If necessary, the object referenced by objPtr is converted into a list
+ *	object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AppendAllObjTypes(
+    Tcl_Interp *interp,		/* Interpreter used for error reporting. */
+    Tcl_Obj *objPtr)		/* Points to the Tcl object onto which the
+				 * name of each registered type is appended as
+				 * a list element. */
+{
+    register Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+    int numElems;
+
+    /*
+     * Get the test for a valid list out of the way first.
+     */
+
+    if (TclListObjLength(interp, objPtr, &numElems) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Type names are NUL-terminated, not counted strings. This code relies on
+     * that.
+     */
+
+    Tcl_MutexLock(&tableMutex);
+    for (hPtr = Tcl_FirstHashEntry(&typeTable, &search);
+	    hPtr != NULL; hPtr = Tcl_NextHashEntry(&search)) {
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewStringObj(Tcl_GetHashKey(&typeTable, hPtr), -1));
+    }
+    Tcl_MutexUnlock(&tableMutex);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetObjType --
+ *
+ *	This function looks up an object type by name.
+ *
+ * Results:
+ *	If an object type with name matching "typeName" is found, a pointer to
+ *	its Tcl_ObjType structure is returned; otherwise, NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const Tcl_ObjType *
+Tcl_GetObjType(
+    const char *typeName)	/* Name of Tcl object type to look up. */
+{
+    register Tcl_HashEntry *hPtr;
+    const Tcl_ObjType *typePtr = NULL;
+
+    Tcl_MutexLock(&tableMutex);
+    hPtr = Tcl_FindHashEntry(&typeTable, typeName);
+    if (hPtr != NULL) {
+	typePtr = Tcl_GetHashValue(hPtr);
+    }
+    Tcl_MutexUnlock(&tableMutex);
+    return typePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConvertToType --
+ *
+ *	Convert the Tcl object "objPtr" to have type "typePtr" if possible.
+ *
+ * Results:
+ *	The return value is TCL_OK on success and TCL_ERROR on failure. If
+ *	TCL_ERROR is returned, then the interpreter's result contains an error
+ *	message unless "interp" is NULL. Passing a NULL "interp" allows this
+ *	function to be used as a test whether the conversion could be done
+ *	(and in fact was done).
+ *
+ * Side effects:
+ *	Any internal representation for the old type is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ConvertToType(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr,		/* The object to convert. */
+    const Tcl_ObjType *typePtr)	/* The target type. */
+{
+    if (objPtr->typePtr == typePtr) {
+	return TCL_OK;
+    }
+
+    /*
+     * Use the target type's Tcl_SetFromAnyProc to set "objPtr"s internal form
+     * as appropriate for the target type. This frees the old internal
+     * representation.
+     */
+
+    if (typePtr->setFromAnyProc == NULL) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "can't convert value to type %s", typePtr->name));
+	    Tcl_SetErrorCode(interp, "TCL", "API_ABUSE", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    return typePtr->setFromAnyProc(interp, objPtr);
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * TclDbDumpActiveObjects --
+ *
+ *	This function is called to dump all of the active Tcl_Obj structs this
+ *	allocator knows about.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+TclDbDumpActiveObjects(
+    FILE *outFile)
+{
+#if defined(TCL_MEM_DEBUG) && defined(TCL_THREADS)
+    Tcl_HashSearch hSearch;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashTable *tablePtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    tablePtr = tsdPtr->objThreadMap;
+
+    if (tablePtr != NULL) {
+	fprintf(outFile, "total objects: %d\n", tablePtr->numEntries);
+	for (hPtr = Tcl_FirstHashEntry(tablePtr, &hSearch); hPtr != NULL;
+		hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    ObjData *objData = Tcl_GetHashValue(hPtr);
+
+	    if (objData != NULL) {
+		fprintf(outFile,
+			"key = 0x%p, objPtr = 0x%p, file = %s, line = %d\n",
+			Tcl_GetHashKey(tablePtr, hPtr), objData->objPtr,
+			objData->file, objData->line);
+	    } else {
+		fprintf(outFile, "key = 0x%p\n",
+			Tcl_GetHashKey(tablePtr, hPtr));
+	    }
+	}
+    }
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDbInitNewObj --
+ *
+ *	Called via the TclNewObj or TclDbNewObj macros when TCL_MEM_DEBUG is
+ *	enabled. This function will initialize the members of a Tcl_Obj
+ *	struct. Initilization would be done inline via the TclNewObj macro
+ *	when compiling without TCL_MEM_DEBUG.
+ *
+ * Results:
+ *	The Tcl_Obj struct members are initialized.
+ *
+ * Side effects:
+ *	None.
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+void
+TclDbInitNewObj(
+    register Tcl_Obj *objPtr,
+    register const char *file,	/* The name of the source file calling this
+				 * function; used for debugging. */
+    register int line)		/* Line number in the source file; used for
+				 * debugging. */
+{
+    objPtr->refCount = 0;
+    objPtr->bytes = &tclEmptyString;
+    objPtr->length = 0;
+    objPtr->typePtr = NULL;
+
+#ifdef TCL_THREADS
+    /*
+     * Add entry to a thread local map used to check if a Tcl_Obj was
+     * allocated by the currently executing thread.
+     */
+
+    if (!TclInExit()) {
+	Tcl_HashEntry *hPtr;
+	Tcl_HashTable *tablePtr;
+	int isNew;
+	ObjData *objData;
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+	if (tsdPtr->objThreadMap == NULL) {
+	    tsdPtr->objThreadMap = ckalloc(sizeof(Tcl_HashTable));
+	    Tcl_InitHashTable(tsdPtr->objThreadMap, TCL_ONE_WORD_KEYS);
+	}
+	tablePtr = tsdPtr->objThreadMap;
+	hPtr = Tcl_CreateHashEntry(tablePtr, objPtr, &isNew);
+	if (!isNew) {
+	    Tcl_Panic("expected to create new entry for object map");
+	}
+
+	/*
+	 * Record the debugging information.
+	 */
+
+	objData = ckalloc(sizeof(ObjData));
+	objData->objPtr = objPtr;
+	objData->file = file;
+	objData->line = line;
+	Tcl_SetHashValue(hPtr, objData);
+    }
+#endif /* TCL_THREADS */
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates new Tcl objects that denote
+ *	the empty string. These objects have a NULL object type and NULL
+ *	string representation byte pointer. Type managers call this routine to
+ *	allocate new objects that they further initialize.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewObj.
+ *
+ * Results:
+ *	The result is a newly allocated object that represents the empty
+ *	string. The new object's typePtr is set NULL and its ref count is set
+ *	to 0.
+ *
+ * Side effects:
+ *	If compiling with TCL_COMPILE_STATS, this function increments the
+ *	global count of allocated objects (tclObjsAlloced).
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewObj
+
+Tcl_Obj *
+Tcl_NewObj(void)
+{
+    return Tcl_DbNewObj("unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewObj(void)
+{
+    register Tcl_Obj *objPtr;
+
+    /*
+     * Use the macro defined in tclInt.h - it will use the correct allocator.
+     */
+
+    TclNewObj(objPtr);
+    return objPtr;
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new Tcl objects that denote the
+ *	empty string. It is the same as the Tcl_NewObj function above except
+ *	that it calls Tcl_DbCkalloc directly with the file name and line
+ *	number from its caller. This simplifies debugging since then the
+ *	[memory active] command will report the correct file name and line
+ *	number when reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewObj.
+ *
+ * Results:
+ *	The result is a newly allocated that represents the empty string. The
+ *	new object's typePtr is set NULL and its ref count is set to 0.
+ *
+ * Side effects:
+ *	If compiling with TCL_COMPILE_STATS, this function increments the
+ *	global count of allocated objects (tclObjsAlloced).
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewObj(
+    register const char *file,	/* The name of the source file calling this
+				 * function; used for debugging. */
+    register int line)		/* Line number in the source file; used for
+				 * debugging. */
+{
+    register Tcl_Obj *objPtr;
+
+    /*
+     * Use the macro defined in tclInt.h - it will use the correct allocator.
+     */
+
+    TclDbNewObj(objPtr, file, line);
+    return objPtr;
+}
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewObj(
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewObj();
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclAllocateFreeObjects --
+ *
+ *	Function to allocate a number of free Tcl_Objs. This is done using a
+ *	single ckalloc to reduce the overhead for Tcl_Obj allocation.
+ *
+ *	Assumes mutex is held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	tclFreeObjList, the head of the list of free Tcl_Objs, is set to the
+ *	first of a number of free Tcl_Obj's linked together by their
+ *	internalRep.twoPtrValue.ptr1's.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#define OBJS_TO_ALLOC_EACH_TIME 100
+
+void
+TclAllocateFreeObjects(void)
+{
+    size_t bytesToAlloc = (OBJS_TO_ALLOC_EACH_TIME * sizeof(Tcl_Obj));
+    char *basePtr;
+    register Tcl_Obj *prevPtr, *objPtr;
+    register int i;
+
+    /*
+     * This has been noted by Purify to be a potential leak. The problem is
+     * that Tcl, when not TCL_MEM_DEBUG compiled, keeps around all allocated
+     * Tcl_Obj's, pointed to by tclFreeObjList, when freed instead of actually
+     * freeing the memory. TclFinalizeObjects() does not ckfree() this memory,
+     * but leaves it to Tcl's memory subsystem finalization to release it.
+     * Purify apparently can't figure that out, and fires a false alarm.
+     */
+
+    basePtr = ckalloc(bytesToAlloc);
+
+    prevPtr = NULL;
+    objPtr = (Tcl_Obj *) basePtr;
+    for (i = 0; i < OBJS_TO_ALLOC_EACH_TIME; i++) {
+	objPtr->internalRep.twoPtrValue.ptr1 = prevPtr;
+	prevPtr = objPtr;
+	objPtr++;
+    }
+    tclFreeObjList = prevPtr;
+}
+#undef OBJS_TO_ALLOC_EACH_TIME
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreeObj --
+ *
+ *	This function frees the memory associated with the argument object.
+ *	It is called by the tcl.h macro Tcl_DecrRefCount when an object's ref
+ *	count is zero. It is only "public" since it must be callable by that
+ *	macro wherever the macro is used. It should not be directly called by
+ *	clients.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deallocates the storage for the object's Tcl_Obj structure after
+ *	deallocating the string representation and calling the type-specific
+ *	Tcl_FreeInternalRepProc to deallocate the object's internal
+ *	representation. If compiling with TCL_COMPILE_STATS, this function
+ *	increments the global count of freed objects (tclObjsFreed).
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+void
+TclFreeObj(
+    register Tcl_Obj *objPtr)	/* The object to be freed. */
+{
+    register const Tcl_ObjType *typePtr = objPtr->typePtr;
+
+    /*
+     * This macro declares a variable, so must come here...
+     */
+
+    ObjInitDeletionContext(context);
+
+# ifdef TCL_THREADS
+    /*
+     * Check to make sure that the Tcl_Obj was allocated by the current
+     * thread. Don't do this check when shutting down since thread local
+     * storage can be finalized before the last Tcl_Obj is freed.
+     */
+
+    if (!TclInExit()) {
+	Tcl_HashTable *tablePtr;
+	Tcl_HashEntry *hPtr;
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+	tablePtr = tsdPtr->objThreadMap;
+	if (!tablePtr) {
+	    Tcl_Panic("TclFreeObj: object table not initialized");
+	}
+	hPtr = Tcl_FindHashEntry(tablePtr, (char *) objPtr);
+	if (hPtr) {
+	    /*
+	     * As the Tcl_Obj is going to be deleted we remove the entry.
+	     */
+
+	    ObjData *objData = Tcl_GetHashValue(hPtr);
+
+	    if (objData != NULL) {
+		ckfree(objData);
+	    }
+
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+    }
+# endif
+
+    /*
+     * Check for a double free of the same value.  This is slightly tricky
+     * because it is customary to free a Tcl_Obj when its refcount falls
+     * either from 1 to 0, or from 0 to -1.  Falling from -1 to -2, though,
+     * and so on, is always a sign of a botch in the caller.
+     */
+    if (objPtr->refCount < -1) {
+	Tcl_Panic("Reference count for %p was negative", objPtr);
+    }
+    /*
+     * Now, in case we just approved drop from 1 to 0 as acceptable, make
+     * sure we do not accept a second free when falling from 0 to -1.
+     * Skip that possibility so any double free will trigger the panic.
+     */
+    objPtr->refCount = -1;
+
+    /*
+     * Invalidate the string rep first so we can use the bytes value for our
+     * pointer chain, and signal an obj deletion (as opposed to shimmering)
+     * with 'length == -1'.
+     */
+
+    TclInvalidateStringRep(objPtr);
+    objPtr->length = -1;
+
+    if (ObjDeletePending(context)) {
+	PushObjToDelete(context, objPtr);
+    } else {
+	TCL_DTRACE_OBJ_FREE(objPtr);
+	if ((typePtr != NULL) && (typePtr->freeIntRepProc != NULL)) {
+	    ObjDeletionLock(context);
+	    typePtr->freeIntRepProc(objPtr);
+	    ObjDeletionUnlock(context);
+	}
+
+	Tcl_MutexLock(&tclObjMutex);
+	ckfree(objPtr);
+	Tcl_MutexUnlock(&tclObjMutex);
+	TclIncrObjsFreed();
+	ObjDeletionLock(context);
+	while (ObjOnStack(context)) {
+	    Tcl_Obj *objToFree;
+
+	    PopObjToDelete(context, objToFree);
+	    TCL_DTRACE_OBJ_FREE(objToFree);
+	    TclFreeIntRep(objToFree);
+
+	    Tcl_MutexLock(&tclObjMutex);
+	    ckfree(objToFree);
+	    Tcl_MutexUnlock(&tclObjMutex);
+	    TclIncrObjsFreed();
+	}
+	ObjDeletionUnlock(context);
+    }
+
+    /*
+     * We cannot use TclGetContinuationTable() here, because that may
+     * re-initialize the thread-data for calls coming after the finalization.
+     * We have to access it using the low-level call and then check for
+     * validity. This function can be called after TclFinalizeThreadData() has
+     * already killed the thread-global data structures. Performing
+     * TCL_TSD_INIT will leave us with an un-initialized memory block upon
+     * which we crash (if we where to access the uninitialized hashtable).
+     */
+
+    {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+        Tcl_HashEntry *hPtr;
+
+	if (tsdPtr->lineCLPtr) {
+            hPtr = Tcl_FindHashEntry(tsdPtr->lineCLPtr, objPtr);
+	    if (hPtr) {
+		ckfree(Tcl_GetHashValue(hPtr));
+		Tcl_DeleteHashEntry(hPtr);
+	    }
+	}
+    }
+}
+#else /* TCL_MEM_DEBUG */
+
+void
+TclFreeObj(
+    register Tcl_Obj *objPtr)	/* The object to be freed. */
+{
+    /*
+     * Invalidate the string rep first so we can use the bytes value for our
+     * pointer chain, and signal an obj deletion (as opposed to shimmering)
+     * with 'length == -1'.
+     */
+
+    TclInvalidateStringRep(objPtr);
+    objPtr->length = -1;
+
+    if (!objPtr->typePtr || !objPtr->typePtr->freeIntRepProc) {
+	/*
+	 * objPtr can be freed safely, as it will not attempt to free any
+	 * other objects: it will not cause recursive calls to this function.
+	 */
+
+	TCL_DTRACE_OBJ_FREE(objPtr);
+	TclFreeObjStorage(objPtr);
+	TclIncrObjsFreed();
+    } else {
+	/*
+	 * This macro declares a variable, so must come here...
+	 */
+
+	ObjInitDeletionContext(context);
+
+	if (ObjDeletePending(context)) {
+	    PushObjToDelete(context, objPtr);
+	} else {
+	    /*
+	     * Note that the contents of the while loop assume that the string
+	     * rep has already been freed and we don't want to do anything
+	     * fancy with adding to the queue inside ourselves. Must take care
+	     * to unstack the object first since freeing the internal rep can
+	     * add further objects to the stack. The code assumes that it is
+	     * the first thing in a block; all current usages in the core
+	     * satisfy this.
+	     */
+
+	    TCL_DTRACE_OBJ_FREE(objPtr);
+	    ObjDeletionLock(context);
+	    objPtr->typePtr->freeIntRepProc(objPtr);
+	    ObjDeletionUnlock(context);
+
+	    TclFreeObjStorage(objPtr);
+	    TclIncrObjsFreed();
+	    ObjDeletionLock(context);
+	    while (ObjOnStack(context)) {
+		Tcl_Obj *objToFree;
+
+		PopObjToDelete(context, objToFree);
+		TCL_DTRACE_OBJ_FREE(objToFree);
+		if ((objToFree->typePtr != NULL)
+			&& (objToFree->typePtr->freeIntRepProc != NULL)) {
+		    objToFree->typePtr->freeIntRepProc(objToFree);
+		}
+		TclFreeObjStorage(objToFree);
+		TclIncrObjsFreed();
+	    }
+	    ObjDeletionUnlock(context);
+	}
+    }
+
+    /*
+     * We cannot use TclGetContinuationTable() here, because that may
+     * re-initialize the thread-data for calls coming after the finalization.
+     * We have to access it using the low-level call and then check for
+     * validity. This function can be called after TclFinalizeThreadData() has
+     * already killed the thread-global data structures. Performing
+     * TCL_TSD_INIT will leave us with an un-initialized memory block upon
+     * which we crash (if we where to access the uninitialized hashtable).
+     */
+
+    {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+        Tcl_HashEntry *hPtr;
+
+	if (tsdPtr->lineCLPtr) {
+            hPtr = Tcl_FindHashEntry(tsdPtr->lineCLPtr, objPtr);
+	    if (hPtr) {
+		ckfree(Tcl_GetHashValue(hPtr));
+		Tcl_DeleteHashEntry(hPtr);
+	    }
+	}
+    }
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjBeingDeleted --
+ *
+ *	This function returns 1 when the Tcl_Obj is being deleted. It is
+ *	provided for the rare cases where the reason for the loss of an
+ *	internal rep might be relevant. [FR 1512138]
+ *
+ * Results:
+ *	1 if being deleted, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjBeingDeleted(
+    Tcl_Obj *objPtr)
+{
+    return (objPtr->length == -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DuplicateObj --
+ *
+ *	Create and return a new object that is a duplicate of the argument
+ *	object.
+ *
+ * Results:
+ *	The return value is a pointer to a newly created Tcl_Obj. This object
+ *	has reference count 0 and the same type, if any, as the source object
+ *	objPtr. Also:
+ *	  1) If the source object has a valid string rep, we copy it;
+ *	     otherwise, the duplicate's string rep is set NULL to mark it
+ *	     invalid.
+ *	  2) If the source object has an internal representation (i.e. its
+ *	     typePtr is non-NULL), the new object's internal rep is set to a
+ *	     copy; otherwise the new internal rep is marked invalid.
+ *
+ * Side effects:
+ *	What constitutes "copying" the internal representation depends on the
+ *	type. For example, if the argument object is a list, the element
+ *	objects it points to will not actually be copied but will be shared
+ *	with the duplicate list. That is, the ref counts of the element
+ *	objects will be incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#define SetDuplicateObj(dupPtr, objPtr)					\
+    {									\
+	const Tcl_ObjType *typePtr = (objPtr)->typePtr;			\
+	const char *bytes = (objPtr)->bytes;				\
+	if (bytes) {							\
+	    TclInitStringRep((dupPtr), bytes, (objPtr)->length);	\
+	} else {							\
+	    (dupPtr)->bytes = NULL;					\
+	}								\
+	if (typePtr) {							\
+	    if (typePtr->dupIntRepProc) {				\
+		typePtr->dupIntRepProc((objPtr), (dupPtr));		\
+	    } else {							\
+		(dupPtr)->internalRep = (objPtr)->internalRep;		\
+		(dupPtr)->typePtr = typePtr;				\
+	    }								\
+	}								\
+    }
+
+Tcl_Obj *
+Tcl_DuplicateObj(
+    Tcl_Obj *objPtr)		/* The object to duplicate. */
+{
+    Tcl_Obj *dupPtr;
+
+    TclNewObj(dupPtr);
+    SetDuplicateObj(dupPtr, objPtr);
+    return dupPtr;
+}
+
+void
+TclSetDuplicateObj(
+    Tcl_Obj *dupPtr,
+    Tcl_Obj *objPtr)
+{
+    if (Tcl_IsShared(dupPtr)) {
+	Tcl_Panic("%s called with shared object", "TclSetDuplicateObj");
+    }
+    TclInvalidateStringRep(dupPtr);
+    TclFreeIntRep(dupPtr);
+    SetDuplicateObj(dupPtr, objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetString --
+ *
+ *	Returns the string representation byte array pointer for an object.
+ *
+ * Results:
+ *	Returns a pointer to the string representation of objPtr. The byte
+ *	array referenced by the returned pointer must not be modified by the
+ *	caller. Furthermore, the caller must copy the bytes if they need to
+ *	retain them since the object's string rep can change as a result of
+ *	other operations.
+ *
+ * Side effects:
+ *	May call the object's updateStringProc to update the string
+ *	representation from the internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_GetString(
+    register Tcl_Obj *objPtr)	/* Object whose string rep byte pointer should
+				 * be returned. */
+{
+    if (objPtr->bytes != NULL) {
+	return objPtr->bytes;
+    }
+
+    /*
+     * Note we do not check for objPtr->typePtr == NULL.  An invariant of
+     * a properly maintained Tcl_Obj is that at least  one of objPtr->bytes
+     * and objPtr->typePtr must not be NULL.  If broken extensions fail to
+     * maintain that invariant, we can crash here.
+     */
+
+    if (objPtr->typePtr->updateStringProc == NULL) {
+	/*
+	 * Those Tcl_ObjTypes which choose not to define an updateStringProc
+	 * must be written in such a way that (objPtr->bytes) never becomes
+	 * NULL.  This panic was added in Tcl 8.1.
+	 */
+
+	Tcl_Panic("UpdateStringProc should not be invoked for type %s",
+		objPtr->typePtr->name);
+    }
+    objPtr->typePtr->updateStringProc(objPtr);
+    if (objPtr->bytes == NULL || objPtr->length < 0
+	    || objPtr->bytes[objPtr->length] != '\0') {
+	Tcl_Panic("UpdateStringProc for type '%s' "
+		"failed to create a valid string rep", objPtr->typePtr->name);
+    }
+    return objPtr->bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetStringFromObj --
+ *
+ *	Returns the string representation's byte array pointer and length for
+ *	an object.
+ *
+ * Results:
+ *	Returns a pointer to the string representation of objPtr. If lengthPtr
+ *	isn't NULL, the length of the string representation is stored at
+ *	*lengthPtr. The byte array referenced by the returned pointer must not
+ *	be modified by the caller. Furthermore, the caller must copy the bytes
+ *	if they need to retain them since the object's string rep can change
+ *	as a result of other operations.
+ *
+ * Side effects:
+ *	May call the object's updateStringProc to update the string
+ *	representation from the internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_GetStringFromObj(
+    register Tcl_Obj *objPtr,	/* Object whose string rep byte pointer should
+				 * be returned. */
+    register int *lengthPtr)	/* If non-NULL, the location where the string
+				 * rep's byte array length should * be stored.
+				 * If NULL, no length is stored. */
+{
+    (void) TclGetString(objPtr);
+
+    if (lengthPtr != NULL) {
+	*lengthPtr = objPtr->length;
+    }
+    return objPtr->bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InvalidateStringRep --
+ *
+ *	This function is called to invalidate an object's string
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Deallocates the storage for any old string representation, then sets
+ *	the string representation NULL to mark it invalid.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_InvalidateStringRep(
+    register Tcl_Obj *objPtr)	/* Object whose string rep byte pointer should
+				 * be freed. */
+{
+    TclInvalidateStringRep(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewBooleanObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates a new Tcl_Obj and
+ *	initializes it from the argument boolean value. A nonzero "boolValue"
+ *	is coerced to 1.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewLongObj.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_NewBooleanObj
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_NewBooleanObj(
+    register int boolValue)	/* Boolean used to initialize new object. */
+{
+    return Tcl_DbNewLongObj(boolValue!=0, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewBooleanObj(
+    register int boolValue)	/* Boolean used to initialize new object. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclNewLongObj(objPtr, boolValue!=0);
+    return objPtr;
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewBooleanObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new boolean objects. It is the
+ *	same as the Tcl_NewBooleanObj function above except that it calls
+ *	Tcl_DbCkalloc directly with the file name and line number from its
+ *	caller. This simplifies debugging since then the [memory active]
+ *	command will report the correct file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewBooleanObj.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_DbNewBooleanObj
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewBooleanObj(
+    register int boolValue,	/* Boolean used to initialize new object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    objPtr->bytes = NULL;
+
+    objPtr->internalRep.longValue = (boolValue != 0);
+    objPtr->typePtr = &tclIntType;
+    return objPtr;
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewBooleanObj(
+    register int boolValue,	/* Boolean used to initialize new object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewBooleanObj(boolValue);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetBooleanObj --
+ *
+ *	Modify an object to be a boolean object and to have the specified
+ *	boolean value. A nonzero "boolValue" is coerced to 1.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep, if any, is freed. Also, any old internal
+ *	rep is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_SetBooleanObj
+void
+Tcl_SetBooleanObj(
+    register Tcl_Obj *objPtr,	/* Object whose internal rep to init. */
+    register int boolValue)	/* Boolean used to set object's value. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetBooleanObj");
+    }
+
+    TclSetLongObj(objPtr, boolValue!=0);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetBooleanFromObj --
+ *
+ *	Attempt to return a boolean from the Tcl object "objPtr". This
+ *	includes conversion from any of Tcl's numeric types.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	The intrep of *objPtr may be changed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetBooleanFromObj(
+    Tcl_Interp *interp,         /* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr,	/* The object from which to get boolean. */
+    register int *boolPtr)	/* Place to store resulting boolean. */
+{
+    do {
+	if (objPtr->typePtr == &tclIntType) {
+	    *boolPtr = (objPtr->internalRep.longValue != 0);
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclBooleanType) {
+	    *boolPtr = (int) objPtr->internalRep.longValue;
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclDoubleType) {
+	    /*
+	     * Caution: Don't be tempted to check directly for the "double"
+	     * Tcl_ObjType and then compare the intrep to 0.0. This isn't
+	     * reliable because a "double" Tcl_ObjType can hold the NaN value.
+	     * Use the API Tcl_GetDoubleFromObj, which does the checking and
+	     * sets the proper error message for us.
+	     */
+
+	    double d;
+
+	    if (Tcl_GetDoubleFromObj(interp, objPtr, &d) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    *boolPtr = (d != 0.0);
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclBignumType) {
+	    *boolPtr = 1;
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    *boolPtr = (objPtr->internalRep.wideValue != 0);
+	    return TCL_OK;
+	}
+#endif
+    } while ((ParseBoolean(objPtr) == TCL_OK) || (TCL_OK ==
+	    TclParseNumber(interp, objPtr, "boolean value", NULL,-1,NULL,0)));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetBooleanFromAny --
+ *
+ *	Attempt to generate a boolean internal form for the Tcl object
+ *	"objPtr".
+ *
+ * Results:
+ *	The return value is a standard Tcl result. If an error occurs during
+ *	conversion, an error message is left in the interpreter's result
+ *	unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If no error occurs, an integer 1 or 0 is stored as "objPtr"s internal
+ *	representation and the type of "objPtr" is set to boolean.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSetBooleanFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    /*
+     * For some "pure" numeric Tcl_ObjTypes (no string rep), we can determine
+     * whether a boolean conversion is possible without generating the string
+     * rep.
+     */
+
+    if (objPtr->bytes == NULL) {
+	if (objPtr->typePtr == &tclIntType) {
+	    switch (objPtr->internalRep.longValue) {
+	    case 0L: case 1L:
+		return TCL_OK;
+	    }
+	    goto badBoolean;
+	}
+
+	if (objPtr->typePtr == &tclBignumType) {
+	    goto badBoolean;
+	}
+
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    goto badBoolean;
+	}
+#endif
+
+	if (objPtr->typePtr == &tclDoubleType) {
+	    goto badBoolean;
+	}
+    }
+
+    if (ParseBoolean(objPtr) == TCL_OK) {
+	return TCL_OK;
+    }
+
+  badBoolean:
+    if (interp != NULL) {
+	int length;
+	const char *str = TclGetStringFromObj(objPtr, &length);
+	Tcl_Obj *msg;
+
+	TclNewLiteralStringObj(msg, "expected boolean value but got \"");
+	Tcl_AppendLimitedToObj(msg, str, length, 50, "");
+	Tcl_AppendToObj(msg, "\"", -1);
+	Tcl_SetObjResult(interp, msg);
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "BOOLEAN", NULL);
+    }
+    return TCL_ERROR;
+}
+
+static int
+ParseBoolean(
+    register Tcl_Obj *objPtr)	/* The object to parse/convert. */
+{
+    int newBool;
+    char lowerCase[6];
+    const char *str = TclGetString(objPtr);
+    size_t i, length = objPtr->length;
+
+    if ((length == 0) || (length > 5)) {
+	/*
+         * Longest valid boolean string rep. is "false".
+         */
+
+	return TCL_ERROR;
+    }
+
+    switch (str[0]) {
+    case '0':
+	if (length == 1) {
+	    newBool = 0;
+	    goto numericBoolean;
+	}
+	return TCL_ERROR;
+    case '1':
+	if (length == 1) {
+	    newBool = 1;
+	    goto numericBoolean;
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Force to lower case for case-insensitive detection. Filter out known
+     * invalid characters at the same time.
+     */
+
+    for (i=0; i < length; i++) {
+	char c = str[i];
+
+	switch (c) {
+	case 'A': case 'E': case 'F': case 'L': case 'N':
+	case 'O': case 'R': case 'S': case 'T': case 'U': case 'Y':
+	    lowerCase[i] = c + (char) ('a' - 'A');
+	    break;
+	case 'a': case 'e': case 'f': case 'l': case 'n':
+	case 'o': case 'r': case 's': case 't': case 'u': case 'y':
+	    lowerCase[i] = c;
+	    break;
+	default:
+	    return TCL_ERROR;
+	}
+    }
+    lowerCase[length] = 0;
+    switch (lowerCase[0]) {
+    case 'y':
+	/*
+	 * Checking the 'y' is redundant, but makes the code clearer.
+	 */
+	if (strncmp(lowerCase, "yes", length) == 0) {
+	    newBool = 1;
+	    goto goodBoolean;
+	}
+	return TCL_ERROR;
+    case 'n':
+	if (strncmp(lowerCase, "no", length) == 0) {
+	    newBool = 0;
+	    goto goodBoolean;
+	}
+	return TCL_ERROR;
+    case 't':
+	if (strncmp(lowerCase, "true", length) == 0) {
+	    newBool = 1;
+	    goto goodBoolean;
+	}
+	return TCL_ERROR;
+    case 'f':
+	if (strncmp(lowerCase, "false", length) == 0) {
+	    newBool = 0;
+	    goto goodBoolean;
+	}
+	return TCL_ERROR;
+    case 'o':
+	if (length < 2) {
+	    return TCL_ERROR;
+	}
+	if (strncmp(lowerCase, "on", length) == 0) {
+	    newBool = 1;
+	    goto goodBoolean;
+	} else if (strncmp(lowerCase, "off", length) == 0) {
+	    newBool = 0;
+	    goto goodBoolean;
+	}
+	return TCL_ERROR;
+    default:
+	return TCL_ERROR;
+    }
+
+    /*
+     * Free the old internalRep before setting the new one. We do this as late
+     * as possible to allow the conversion code, in particular
+     * Tcl_GetStringFromObj, to use that old internalRep.
+     */
+
+  goodBoolean:
+    TclFreeIntRep(objPtr);
+    objPtr->internalRep.longValue = newBool;
+    objPtr->typePtr = &tclBooleanType;
+    return TCL_OK;
+
+  numericBoolean:
+    TclFreeIntRep(objPtr);
+    objPtr->internalRep.longValue = newBool;
+    objPtr->typePtr = &tclIntType;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewDoubleObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates a new double object and
+ *	initializes it from the argument double value.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewDoubleObj.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an
+ *	invalid string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewDoubleObj
+
+Tcl_Obj *
+Tcl_NewDoubleObj(
+    register double dblValue)	/* Double used to initialize the object. */
+{
+    return Tcl_DbNewDoubleObj(dblValue, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewDoubleObj(
+    register double dblValue)	/* Double used to initialize the object. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclNewDoubleObj(objPtr, dblValue);
+    return objPtr;
+}
+#endif /* if TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewDoubleObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new double objects. It is the
+ *	same as the Tcl_NewDoubleObj function above except that it calls
+ *	Tcl_DbCkalloc directly with the file name and line number from its
+ *	caller. This simplifies debugging since then the [memory active]
+ *	command will report the correct file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewDoubleObj.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewDoubleObj(
+    register double dblValue,	/* Double used to initialize the object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    objPtr->bytes = NULL;
+
+    objPtr->internalRep.doubleValue = dblValue;
+    objPtr->typePtr = &tclDoubleType;
+    return objPtr;
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewDoubleObj(
+    register double dblValue,	/* Double used to initialize the object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewDoubleObj(dblValue);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetDoubleObj --
+ *
+ *	Modify an object to be a double object and to have the specified
+ *	double value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep, if any, is freed. Also, any old internal
+ *	rep is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetDoubleObj(
+    register Tcl_Obj *objPtr,	/* Object whose internal rep to init. */
+    register double dblValue)	/* Double used to set the object's value. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetDoubleObj");
+    }
+
+    TclSetDoubleObj(objPtr, dblValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetDoubleFromObj --
+ *
+ *	Attempt to return a double from the Tcl object "objPtr". If the object
+ *	is not already a double, an attempt will be made to convert it to one.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If the object is not already a double, the conversion will free any
+ *	old internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetDoubleFromObj(
+    Tcl_Interp *interp,         /* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr,	/* The object from which to get a double. */
+    register double *dblPtr)	/* Place to store resulting double. */
+{
+    do {
+	if (objPtr->typePtr == &tclDoubleType) {
+	    if (TclIsNaN(objPtr->internalRep.doubleValue)) {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "floating point value is Not a Number", -1));
+                    Tcl_SetErrorCode(interp, "TCL", "VALUE", "DOUBLE", "NAN",
+                            NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    *dblPtr = (double) objPtr->internalRep.doubleValue;
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclIntType) {
+	    *dblPtr = objPtr->internalRep.longValue;
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclBignumType) {
+	    mp_int big;
+
+	    UNPACK_BIGNUM(objPtr, big);
+	    *dblPtr = TclBignumToDouble(&big);
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    *dblPtr = (double) objPtr->internalRep.wideValue;
+	    return TCL_OK;
+	}
+#endif
+    } while (SetDoubleFromAny(interp, objPtr) == TCL_OK);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetDoubleFromAny --
+ *
+ *	Attempt to generate an double-precision floating point internal form
+ *	for the Tcl object "objPtr".
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If no error occurs, a double is stored as "objPtr"s internal
+ *	representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetDoubleFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    return TclParseNumber(interp, objPtr, "floating-point number", NULL, -1,
+	    NULL, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfDouble --
+ *
+ *	Update the string representation for a double-precision floating point
+ *	object. This must obey the current tcl_precision value for
+ *	double-to-string conversions. Note: This function does not free an
+ *	existing old string rep so storage will be lost if this has not
+ *	already been done.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	double-to-string conversion.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfDouble(
+    register Tcl_Obj *objPtr)	/* Double obj with string rep to update. */
+{
+    char buffer[TCL_DOUBLE_SPACE];
+    register int len;
+
+    Tcl_PrintDouble(NULL, objPtr->internalRep.doubleValue, buffer);
+    len = strlen(buffer);
+
+    objPtr->bytes = ckalloc(len + 1);
+    memcpy(objPtr->bytes, buffer, (unsigned) len + 1);
+    objPtr->length = len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewIntObj --
+ *
+ *	If a client is compiled with TCL_MEM_DEBUG defined, calls to
+ *	Tcl_NewIntObj to create a new integer object end up calling the
+ *	debugging function Tcl_DbNewLongObj instead.
+ *
+ *	Otherwise, if the client is compiled without TCL_MEM_DEBUG defined,
+ *	calls to Tcl_NewIntObj result in a call to one of the two
+ *	Tcl_NewIntObj implementations below. We provide two implementations so
+ *	that the Tcl core can be compiled to do memory debugging of the core
+ *	even if a client does not request it for itself.
+ *
+ *	Integer and long integer objects share the same "integer" type
+ *	implementation. We store all integers as longs and Tcl_GetIntFromObj
+ *	checks whether the current value of the long can be represented by an
+ *	int.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_NewIntObj
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_NewIntObj(
+    register int intValue)	/* Int used to initialize the new object. */
+{
+    return Tcl_DbNewLongObj((long)intValue, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewIntObj(
+    register int intValue)	/* Int used to initialize the new object. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclNewLongObj(objPtr, intValue);
+    return objPtr;
+}
+#endif /* if TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetIntObj --
+ *
+ *	Modify an object to be an integer and to have the specified integer
+ *	value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep, if any, is freed. Also, any old internal
+ *	rep is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_SetIntObj
+void
+Tcl_SetIntObj(
+    register Tcl_Obj *objPtr,	/* Object whose internal rep to init. */
+    register int intValue)	/* Integer used to set object's value. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetIntObj");
+    }
+
+    TclSetLongObj(objPtr, intValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetIntFromObj --
+ *
+ *	Attempt to return an int from the Tcl object "objPtr". If the object
+ *	is not already an int, an attempt will be made to convert it to one.
+ *
+ *	Integer and long integer objects share the same "integer" type
+ *	implementation. We store all integers as longs and Tcl_GetIntFromObj
+ *	checks whether the current value of the long can be represented by an
+ *	int.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion or if the long integer held by the object can not be
+ *	represented by an int, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If the object is not already an int, the conversion will free any old
+ *	internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetIntFromObj(
+    Tcl_Interp *interp,         /* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr,	/* The object from which to get a int. */
+    register int *intPtr)	/* Place to store resulting int. */
+{
+#if (LONG_MAX == INT_MAX)
+    return TclGetLongFromObj(interp, objPtr, (long *) intPtr);
+#else
+    long l;
+
+    if (TclGetLongFromObj(interp, objPtr, &l) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if ((ULONG_MAX > UINT_MAX) && ((l > UINT_MAX) || (l < -(long)UINT_MAX))) {
+	if (interp != NULL) {
+	    const char *s =
+		    "integer value too large to represent as non-long integer";
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(s, -1));
+	    Tcl_SetErrorCode(interp, "ARITH", "IOVERFLOW", s, NULL);
+	}
+	return TCL_ERROR;
+    }
+    *intPtr = (int) l;
+    return TCL_OK;
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetIntFromAny --
+ *
+ *	Attempts to force the internal representation for a Tcl object to
+ *	tclIntType, specifically.
+ *
+ * Results:
+ *	The return value is a standard object Tcl result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetIntFromAny(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    Tcl_Obj *objPtr)		/* Pointer to the object to convert */
+{
+    long l;
+
+    return TclGetLongFromObj(interp, objPtr, &l);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfInt --
+ *
+ *	Update the string representation for an integer object. Note: This
+ *	function does not free an existing old string rep so storage will be
+ *	lost if this has not already been done.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	int-to-string conversion.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfInt(
+    register Tcl_Obj *objPtr)	/* Int object whose string rep to update. */
+{
+    char buffer[TCL_INTEGER_SPACE];
+    register int len;
+
+    len = TclFormatInt(buffer, objPtr->internalRep.longValue);
+
+    objPtr->bytes = ckalloc(len + 1);
+    memcpy(objPtr->bytes, buffer, (unsigned) len + 1);
+    objPtr->length = len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewLongObj --
+ *
+ *	If a client is compiled with TCL_MEM_DEBUG defined, calls to
+ *	Tcl_NewLongObj to create a new long integer object end up calling the
+ *	debugging function Tcl_DbNewLongObj instead.
+ *
+ *	Otherwise, if the client is compiled without TCL_MEM_DEBUG defined,
+ *	calls to Tcl_NewLongObj result in a call to one of the two
+ *	Tcl_NewLongObj implementations below. We provide two implementations
+ *	so that the Tcl core can be compiled to do memory debugging of the
+ *	core even if a client does not request it for itself.
+ *
+ *	Integer and long integer objects share the same "integer" type
+ *	implementation. We store all integers as longs and Tcl_GetIntFromObj
+ *	checks whether the current value of the long can be represented by an
+ *	int.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewLongObj
+
+Tcl_Obj *
+Tcl_NewLongObj(
+    register long longValue)	/* Long integer used to initialize the
+				 * new object. */
+{
+    return Tcl_DbNewLongObj(longValue, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewLongObj(
+    register long longValue)	/* Long integer used to initialize the
+				 * new object. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclNewLongObj(objPtr, longValue);
+    return objPtr;
+}
+#endif /* if TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewLongObj --
+ *
+ *	If a client is compiled with TCL_MEM_DEBUG defined, calls to
+ *	Tcl_NewIntObj and Tcl_NewLongObj to create new integer or long integer
+ *	objects end up calling the debugging function Tcl_DbNewLongObj
+ *	instead. We provide two implementations of Tcl_DbNewLongObj so that
+ *	whether the Tcl core is compiled to do memory debugging of the core is
+ *	independent of whether a client requests debugging for itself.
+ *
+ *	When the core is compiled with TCL_MEM_DEBUG defined, Tcl_DbNewLongObj
+ *	calls Tcl_DbCkalloc directly with the file name and line number from
+ *	its caller. This simplifies debugging since then the [memory active]
+ *	command will report the caller's file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	Otherwise, when the core is compiled without TCL_MEM_DEBUG defined,
+ *	this function just returns the result of calling Tcl_NewLongObj.
+ *
+ * Results:
+ *	The newly created long integer object is returned. This object will
+ *	have an invalid string representation. The returned object has ref
+ *	count 0.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewLongObj(
+    register long longValue,	/* Long integer used to initialize the new
+				 * object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    objPtr->bytes = NULL;
+
+    objPtr->internalRep.longValue = longValue;
+    objPtr->typePtr = &tclIntType;
+    return objPtr;
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewLongObj(
+    register long longValue,	/* Long integer used to initialize the new
+				 * object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewLongObj(longValue);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetLongObj --
+ *
+ *	Modify an object to be an integer object and to have the specified
+ *	long integer value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep, if any, is freed. Also, any old internal
+ *	rep is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetLongObj(
+    register Tcl_Obj *objPtr,	/* Object whose internal rep to init. */
+    register long longValue)	/* Long integer used to initialize the
+				 * object's value. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetLongObj");
+    }
+
+    TclSetLongObj(objPtr, longValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetLongFromObj --
+ *
+ *	Attempt to return an long integer from the Tcl object "objPtr". If the
+ *	object is not already an int object, an attempt will be made to
+ *	convert it to one.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If the object is not already an int object, the conversion will free
+ *	any old internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetLongFromObj(
+    Tcl_Interp *interp,         /* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr,	/* The object from which to get a long. */
+    register long *longPtr)	/* Place to store resulting long. */
+{
+    do {
+	if (objPtr->typePtr == &tclIntType) {
+	    *longPtr = objPtr->internalRep.longValue;
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    /*
+	     * We return any integer in the range -ULONG_MAX to ULONG_MAX
+	     * converted to a long, ignoring overflow. The rule preserves
+	     * existing semantics for conversion of integers on input, but
+	     * avoids inadvertent demotion of wide integers to 32-bit ones in
+	     * the internal rep.
+	     */
+
+	    Tcl_WideInt w = objPtr->internalRep.wideValue;
+
+	    if (w >= -(Tcl_WideInt)(ULONG_MAX)
+		    && w <= (Tcl_WideInt)(ULONG_MAX)) {
+		*longPtr = Tcl_WideAsLong(w);
+		return TCL_OK;
+	    }
+	    goto tooLarge;
+	}
+#endif
+	if (objPtr->typePtr == &tclDoubleType) {
+	    if (interp != NULL) {
+                Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "expected integer but got \"%s\"",
+                        TclGetString(objPtr)));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "INTEGER", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	if (objPtr->typePtr == &tclBignumType) {
+	    /*
+	     * Must check for those bignum values that can fit in a long, even
+	     * when auto-narrowing is enabled. Only those values in the signed
+	     * long range get auto-narrowed to tclIntType, while all the
+	     * values in the unsigned long range will fit in a long.
+	     */
+
+	    mp_int big;
+
+	    UNPACK_BIGNUM(objPtr, big);
+	    if ((size_t) big.used <= (CHAR_BIT * sizeof(long) + DIGIT_BIT - 1)
+		    / DIGIT_BIT) {
+		unsigned long value = 0, numBytes = sizeof(long);
+		long scratch;
+		unsigned char *bytes = (unsigned char *) &scratch;
+
+		if (mp_to_unsigned_bin_n(&big, bytes, &numBytes) == MP_OKAY) {
+		    while (numBytes-- > 0) {
+			value = (value << CHAR_BIT) | *bytes++;
+		    }
+		    if (big.sign) {
+			*longPtr = - (long) value;
+		    } else {
+			*longPtr = (long) value;
+		    }
+		    return TCL_OK;
+		}
+	    }
+#ifndef TCL_WIDE_INT_IS_LONG
+	tooLarge:
+#endif
+	    if (interp != NULL) {
+		const char *s = "integer value too large to represent";
+		Tcl_Obj *msg = Tcl_NewStringObj(s, -1);
+
+		Tcl_SetObjResult(interp, msg);
+		Tcl_SetErrorCode(interp, "ARITH", "IOVERFLOW", s, NULL);
+	    }
+	    return TCL_ERROR;
+	}
+    } while (TclParseNumber(interp, objPtr, "integer", NULL, -1, NULL,
+	    TCL_PARSE_INTEGER_ONLY)==TCL_OK);
+    return TCL_ERROR;
+}
+#ifndef TCL_WIDE_INT_IS_LONG
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfWideInt --
+ *
+ *	Update the string representation for a wide integer object. Note: this
+ *	function does not free an existing old string rep so storage will be
+ *	lost if this has not already been done.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to a valid string that results from the
+ *	wideInt-to-string conversion.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfWideInt(
+    register Tcl_Obj *objPtr)	/* Int object whose string rep to update. */
+{
+    char buffer[TCL_INTEGER_SPACE+2];
+    register unsigned len;
+    register Tcl_WideInt wideVal = objPtr->internalRep.wideValue;
+
+    /*
+     * Note that sprintf will generate a compiler warning under Mingw claiming
+     * %I64 is an unknown format specifier. Just ignore this warning. We can't
+     * use %L as the format specifier since that gets printed as a 32 bit
+     * value.
+     */
+
+    sprintf(buffer, "%" TCL_LL_MODIFIER "d", wideVal);
+    len = strlen(buffer);
+    objPtr->bytes = ckalloc(len + 1);
+    memcpy(objPtr->bytes, buffer, len + 1);
+    objPtr->length = len;
+}
+#endif /* !TCL_WIDE_INT_IS_LONG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewWideIntObj --
+ *
+ *	If a client is compiled with TCL_MEM_DEBUG defined, calls to
+ *	Tcl_NewWideIntObj to create a new 64-bit integer object end up calling
+ *	the debugging function Tcl_DbNewWideIntObj instead.
+ *
+ *	Otherwise, if the client is compiled without TCL_MEM_DEBUG defined,
+ *	calls to Tcl_NewWideIntObj result in a call to one of the two
+ *	Tcl_NewWideIntObj implementations below. We provide two
+ *	implementations so that the Tcl core can be compiled to do memory
+ *	debugging of the core even if a client does not request it for itself.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have an invalid
+ *	string representation. The returned object has ref count 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewWideIntObj
+
+Tcl_Obj *
+Tcl_NewWideIntObj(
+    register Tcl_WideInt wideValue)
+				/* Wide integer used to initialize the new
+				 * object. */
+{
+    return Tcl_DbNewWideIntObj(wideValue, "unknown", 0);
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_NewWideIntObj(
+    register Tcl_WideInt wideValue)
+				/* Wide integer used to initialize the new
+				 * object. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclNewObj(objPtr);
+    Tcl_SetWideIntObj(objPtr, wideValue);
+    return objPtr;
+}
+#endif /* if TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewWideIntObj --
+ *
+ *	If a client is compiled with TCL_MEM_DEBUG defined, calls to
+ *	Tcl_NewWideIntObj to create new wide integer end up calling the
+ *	debugging function Tcl_DbNewWideIntObj instead. We provide two
+ *	implementations of Tcl_DbNewWideIntObj so that whether the Tcl core is
+ *	compiled to do memory debugging of the core is independent of whether
+ *	a client requests debugging for itself.
+ *
+ *	When the core is compiled with TCL_MEM_DEBUG defined,
+ *	Tcl_DbNewWideIntObj calls Tcl_DbCkalloc directly with the file name
+ *	and line number from its caller. This simplifies debugging since then
+ *	the checkmem command will report the caller's file name and line
+ *	number when reporting objects that haven't been freed.
+ *
+ *	Otherwise, when the core is compiled without TCL_MEM_DEBUG defined,
+ *	this function just returns the result of calling Tcl_NewWideIntObj.
+ *
+ * Results:
+ *	The newly created wide integer object is returned. This object will
+ *	have an invalid string representation. The returned object has ref
+ *	count 0.
+ *
+ * Side effects:
+ *	Allocates memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+
+Tcl_Obj *
+Tcl_DbNewWideIntObj(
+    register Tcl_WideInt wideValue,
+				/* Wide integer used to initialize the new
+				 * object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    register Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    Tcl_SetWideIntObj(objPtr, wideValue);
+    return objPtr;
+}
+
+#else /* if not TCL_MEM_DEBUG */
+
+Tcl_Obj *
+Tcl_DbNewWideIntObj(
+    register Tcl_WideInt wideValue,
+				/* Long integer used to initialize the new
+				 * object. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewWideIntObj(wideValue);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetWideIntObj --
+ *
+ *	Modify an object to be a wide integer object and to have the specified
+ *	wide integer value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old string rep, if any, is freed. Also, any old internal
+ *	rep is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetWideIntObj(
+    register Tcl_Obj *objPtr,	/* Object w. internal rep to init. */
+    register Tcl_WideInt wideValue)
+				/* Wide integer used to initialize the
+				 * object's value. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetWideIntObj");
+    }
+
+    if ((wideValue >= (Tcl_WideInt) LONG_MIN)
+	    && (wideValue <= (Tcl_WideInt) LONG_MAX)) {
+	TclSetLongObj(objPtr, (long) wideValue);
+    } else {
+#ifndef TCL_WIDE_INT_IS_LONG
+	TclSetWideIntObj(objPtr, wideValue);
+#else
+	mp_int big;
+
+	TclBNInitBignumFromWideInt(&big, wideValue);
+	Tcl_SetBignumObj(objPtr, &big);
+#endif
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetWideIntFromObj --
+ *
+ *	Attempt to return a wide integer from the Tcl object "objPtr". If the
+ *	object is not already a wide int object, an attempt will be made to
+ *	convert it to one.
+ *
+ * Results:
+ *	The return value is a standard Tcl object result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If the object is not already an int object, the conversion will free
+ *	any old internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetWideIntFromObj(
+    Tcl_Interp *interp,         /* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr,	/* Object from which to get a wide int. */
+    register Tcl_WideInt *wideIntPtr)
+				/* Place to store resulting long. */
+{
+    do {
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    *wideIntPtr = objPtr->internalRep.wideValue;
+	    return TCL_OK;
+	}
+#endif
+	if (objPtr->typePtr == &tclIntType) {
+	    *wideIntPtr = (Tcl_WideInt) objPtr->internalRep.longValue;
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclDoubleType) {
+	    if (interp != NULL) {
+                Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "expected integer but got \"%s\"",
+                        TclGetString(objPtr)));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "INTEGER", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	if (objPtr->typePtr == &tclBignumType) {
+	    /*
+	     * Must check for those bignum values that can fit in a
+	     * Tcl_WideInt, even when auto-narrowing is enabled.
+	     */
+
+	    mp_int big;
+
+	    UNPACK_BIGNUM(objPtr, big);
+	    if ((size_t) big.used <= (CHAR_BIT * sizeof(Tcl_WideInt)
+		     + DIGIT_BIT - 1) / DIGIT_BIT) {
+		Tcl_WideUInt value = 0;
+		unsigned long numBytes = sizeof(Tcl_WideInt);
+		Tcl_WideInt scratch;
+		unsigned char *bytes = (unsigned char *) &scratch;
+
+		if (mp_to_unsigned_bin_n(&big, bytes, &numBytes) == MP_OKAY) {
+		    while (numBytes-- > 0) {
+			value = (value << CHAR_BIT) | *bytes++;
+		    }
+		    if (big.sign) {
+			*wideIntPtr = - (Tcl_WideInt) value;
+		    } else {
+			*wideIntPtr = (Tcl_WideInt) value;
+		    }
+		    return TCL_OK;
+		}
+	    }
+	    if (interp != NULL) {
+		const char *s = "integer value too large to represent";
+		Tcl_Obj *msg = Tcl_NewStringObj(s, -1);
+
+		Tcl_SetObjResult(interp, msg);
+		Tcl_SetErrorCode(interp, "ARITH", "IOVERFLOW", s, NULL);
+	    }
+	    return TCL_ERROR;
+	}
+    } while (TclParseNumber(interp, objPtr, "integer", NULL, -1, NULL,
+	    TCL_PARSE_INTEGER_ONLY)==TCL_OK);
+    return TCL_ERROR;
+}
+#ifndef TCL_WIDE_INT_IS_LONG
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetWideIntFromAny --
+ *
+ *	Attempts to force the internal representation for a Tcl object to
+ *	tclWideIntType, specifically.
+ *
+ * Results:
+ *	The return value is a standard object Tcl result. If an error occurs
+ *	during conversion, an error message is left in the interpreter's
+ *	result unless "interp" is NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetWideIntFromAny(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    Tcl_Obj *objPtr)		/* Pointer to the object to convert */
+{
+    Tcl_WideInt w;
+    return Tcl_GetWideIntFromObj(interp, objPtr, &w);
+}
+#endif /* !TCL_WIDE_INT_IS_LONG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeBignum --
+ *
+ *	This function frees the internal rep of a bignum.
+ *
+ * Results:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeBignum(
+    Tcl_Obj *objPtr)
+{
+    mp_int toFree;		/* Bignum to free */
+
+    UNPACK_BIGNUM(objPtr, toFree);
+    mp_clear(&toFree);
+    if (PTR2INT(objPtr->internalRep.twoPtrValue.ptr2) < 0) {
+	ckfree(objPtr->internalRep.twoPtrValue.ptr1);
+    }
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupBignum --
+ *
+ *	This function duplicates the internal rep of a bignum.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The destination object receies a copy of the source object
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupBignum(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *copyPtr)
+{
+    mp_int bignumVal;
+    mp_int bignumCopy;
+
+    copyPtr->typePtr = &tclBignumType;
+    UNPACK_BIGNUM(srcPtr, bignumVal);
+    if (mp_init_copy(&bignumCopy, &bignumVal) != MP_OKAY) {
+	Tcl_Panic("initialization failure in DupBignum");
+    }
+    PACK_BIGNUM(bignumCopy, copyPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfBignum --
+ *
+ *	This function updates the string representation of a bignum object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string is set to whatever results from the bignum-
+ *	to-string conversion.
+ *
+ * The object's existing string representation is NOT freed; memory will leak
+ * if the string rep is still valid at the time this function is called.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfBignum(
+    Tcl_Obj *objPtr)
+{
+    mp_int bignumVal;
+    int size;
+    int status;
+    char *stringVal;
+
+    UNPACK_BIGNUM(objPtr, bignumVal);
+    status = mp_radix_size(&bignumVal, 10, &size);
+    if (status != MP_OKAY) {
+	Tcl_Panic("radix size failure in UpdateStringOfBignum");
+    }
+    if (size < 2) {
+	/*
+	 * mp_radix_size() returns < 2 when more than INT_MAX bytes would be
+	 * needed to hold the string rep (because mp_radix_size ignores
+	 * integer overflow issues).
+	 *
+	 * Note that so long as we enforce our bignums to the size that fits
+	 * in a packed bignum, this branch will never be taken.
+	 */
+
+	Tcl_Panic("UpdateStringOfBignum: string length limit exceeded");
+    }
+    stringVal = ckalloc(size);
+    status = mp_toradix_n(&bignumVal, stringVal, 10, size);
+    if (status != MP_OKAY) {
+	Tcl_Panic("conversion failure in UpdateStringOfBignum");
+    }
+    objPtr->bytes = stringVal;
+    objPtr->length = size - 1;	/* size includes a trailing NUL byte. */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewBignumObj --
+ *
+ *	Creates an initializes a bignum object.
+ *
+ * Results:
+ *	Returns the newly created object.
+ *
+ * Side effects:
+ *	The bignum value is cleared, since ownership has transferred to Tcl.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewBignumObj
+
+Tcl_Obj *
+Tcl_NewBignumObj(
+    mp_int *bignumValue)
+{
+    return Tcl_DbNewBignumObj(bignumValue, "unknown", 0);
+}
+#else
+Tcl_Obj *
+Tcl_NewBignumObj(
+    mp_int *bignumValue)
+{
+    Tcl_Obj *objPtr;
+
+    TclNewObj(objPtr);
+    Tcl_SetBignumObj(objPtr, bignumValue);
+    return objPtr;
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewBignumObj --
+ *
+ *	This function is normally called when debugging: that is, when
+ *	TCL_MEM_DEBUG is defined. It constructs a bignum object, recording the
+ *	creation point so that [memory active] can report it.
+ *
+ * Results:
+ *	Returns the newly created object.
+ *
+ * Side effects:
+ *	The bignum value is cleared, since ownership has transferred to Tcl.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+Tcl_Obj *
+Tcl_DbNewBignumObj(
+    mp_int *bignumValue,
+    const char *file,
+    int line)
+{
+    Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    Tcl_SetBignumObj(objPtr, bignumValue);
+    return objPtr;
+}
+#else
+Tcl_Obj *
+Tcl_DbNewBignumObj(
+    mp_int *bignumValue,
+    const char *file,
+    int line)
+{
+    return Tcl_NewBignumObj(bignumValue);
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetBignumFromObj --
+ *
+ *	This function retrieves a 'bignum' value from a Tcl object, converting
+ *	the object if necessary. Either copies or transfers the mp_int value
+ *	depending on the copy flag value passed in.
+ *
+ * Results:
+ *	Returns TCL_OK if the conversion is successful, TCL_ERROR otherwise.
+ *
+ * Side effects:
+ *	A copy of bignum is stored in *bignumValue, which is expected to be
+ *	uninitialized or cleared. If conversion fails, and the 'interp'
+ *	argument is not NULL, an error message is stored in the interpreter
+ *	result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetBignumFromObj(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting */
+    Tcl_Obj *objPtr,		/* Object to read */
+    int copy,			/* Whether to copy the returned bignum value */
+    mp_int *bignumValue)	/* Returned bignum value. */
+{
+    do {
+	if (objPtr->typePtr == &tclBignumType) {
+	    if (copy || Tcl_IsShared(objPtr)) {
+		mp_int temp;
+
+		UNPACK_BIGNUM(objPtr, temp);
+		mp_init_copy(bignumValue, &temp);
+	    } else {
+		UNPACK_BIGNUM(objPtr, *bignumValue);
+		objPtr->internalRep.twoPtrValue.ptr1 = NULL;
+		objPtr->internalRep.twoPtrValue.ptr2 = NULL;
+		objPtr->typePtr = NULL;
+		if (objPtr->bytes == NULL) {
+		    TclInitStringRep(objPtr, &tclEmptyString, 0);
+		}
+	    }
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclIntType) {
+	    TclBNInitBignumFromLong(bignumValue, objPtr->internalRep.longValue);
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    TclBNInitBignumFromWideInt(bignumValue,
+		    objPtr->internalRep.wideValue);
+	    return TCL_OK;
+	}
+#endif
+	if (objPtr->typePtr == &tclDoubleType) {
+	    if (interp != NULL) {
+                Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "expected integer but got \"%s\"",
+                        TclGetString(objPtr)));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "INTEGER", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+    } while (TclParseNumber(interp, objPtr, "integer", NULL, -1, NULL,
+	    TCL_PARSE_INTEGER_ONLY)==TCL_OK);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetBignumFromObj --
+ *
+ *	This function retrieves a 'bignum' value from a Tcl object, converting
+ *	the object if necessary.
+ *
+ * Results:
+ *	Returns TCL_OK if the conversion is successful, TCL_ERROR otherwise.
+ *
+ * Side effects:
+ *	A copy of bignum is stored in *bignumValue, which is expected to be
+ *	uninitialized or cleared. If conversion fails, an the 'interp'
+ *	argument is not NULL, an error message is stored in the interpreter
+ *	result.
+ *
+ *	It is expected that the caller will NOT have invoked mp_init on the
+ *	bignum value before passing it in. Tcl will initialize the mp_int as
+ *	it sets the value. The value is a copy of the value in objPtr, so it
+ *	becomes the responsibility of the caller to call mp_clear on it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetBignumFromObj(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting */
+    Tcl_Obj *objPtr,		/* Object to read */
+    mp_int *bignumValue)	/* Returned bignum value. */
+{
+    return GetBignumFromObj(interp, objPtr, 1, bignumValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TakeBignumFromObj --
+ *
+ *	This function retrieves a 'bignum' value from a Tcl object, converting
+ *	the object if necessary.
+ *
+ * Results:
+ *	Returns TCL_OK if the conversion is successful, TCL_ERROR otherwise.
+ *
+ * Side effects:
+ *	A copy of bignum is stored in *bignumValue, which is expected to be
+ *	uninitialized or cleared. If conversion fails, an the 'interp'
+ *	argument is not NULL, an error message is stored in the interpreter
+ *	result.
+ *
+ *	It is expected that the caller will NOT have invoked mp_init on the
+ *	bignum value before passing it in. Tcl will initialize the mp_int as
+ *	it sets the value. The value is transferred from the internals of
+ *	objPtr to the caller, passing responsibility of the caller to call
+ *	mp_clear on it. The objPtr is cleared to hold an empty value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_TakeBignumFromObj(
+    Tcl_Interp *interp,		/* Tcl interpreter for error reporting */
+    Tcl_Obj *objPtr,		/* Object to read */
+    mp_int *bignumValue)	/* Returned bignum value. */
+{
+    return GetBignumFromObj(interp, objPtr, 0, bignumValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetBignumObj --
+ *
+ *	This function sets the value of a Tcl_Obj to a large integer.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Object value is stored. The bignum value is cleared, since ownership
+ *	has transferred to Tcl.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetBignumObj(
+    Tcl_Obj *objPtr,		/* Object to set */
+    mp_int *bignumValue)	/* Value to store */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetBignumObj");
+    }
+    if ((size_t) bignumValue->used
+	    <= (CHAR_BIT * sizeof(long) + DIGIT_BIT - 1) / DIGIT_BIT) {
+	unsigned long value = 0, numBytes = sizeof(long);
+	long scratch;
+	unsigned char *bytes = (unsigned char *) &scratch;
+
+	if (mp_to_unsigned_bin_n(bignumValue, bytes, &numBytes) != MP_OKAY) {
+	    goto tooLargeForLong;
+	}
+	while (numBytes-- > 0) {
+	    value = (value << CHAR_BIT) | *bytes++;
+	}
+	if (value > (((~(unsigned long)0) >> 1) + bignumValue->sign)) {
+	    goto tooLargeForLong;
+	}
+	if (bignumValue->sign) {
+	    TclSetLongObj(objPtr, -(long)value);
+	} else {
+	    TclSetLongObj(objPtr, (long)value);
+	}
+	mp_clear(bignumValue);
+	return;
+    }
+  tooLargeForLong:
+#ifndef TCL_WIDE_INT_IS_LONG
+    if ((size_t) bignumValue->used
+	    <= (CHAR_BIT * sizeof(Tcl_WideInt) + DIGIT_BIT - 1) / DIGIT_BIT) {
+	Tcl_WideUInt value = 0;
+	unsigned long numBytes = sizeof(Tcl_WideInt);
+	Tcl_WideInt scratch;
+	unsigned char *bytes = (unsigned char *)&scratch;
+
+	if (mp_to_unsigned_bin_n(bignumValue, bytes, &numBytes) != MP_OKAY) {
+	    goto tooLargeForWide;
+	}
+	while (numBytes-- > 0) {
+	    value = (value << CHAR_BIT) | *bytes++;
+	}
+	if (value > (((~(Tcl_WideUInt)0) >> 1) + bignumValue->sign)) {
+	    goto tooLargeForWide;
+	}
+	if (bignumValue->sign) {
+	    TclSetWideIntObj(objPtr, -(Tcl_WideInt)value);
+	} else {
+	    TclSetWideIntObj(objPtr, (Tcl_WideInt)value);
+	}
+	mp_clear(bignumValue);
+	return;
+    }
+  tooLargeForWide:
+#endif
+    TclInvalidateStringRep(objPtr);
+    TclFreeIntRep(objPtr);
+    TclSetBignumIntRep(objPtr, bignumValue);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetBignumIntRep --
+ *
+ *	Install a bignum into the internal representation of an object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Object internal representation is updated and object type is set. The
+ *	bignum value is cleared, since ownership has transferred to the
+ *	object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetBignumIntRep(
+    Tcl_Obj *objPtr,
+    mp_int *bignumValue)
+{
+    objPtr->typePtr = &tclBignumType;
+    PACK_BIGNUM(*bignumValue, objPtr);
+
+    /*
+     * Clear the mp_int value.
+     *
+     * Don't call mp_clear() because it would free the digit array we just
+     * packed into the Tcl_Obj.
+     */
+
+    bignumValue->dp = NULL;
+    bignumValue->alloc = bignumValue->used = 0;
+    bignumValue->sign = MP_NEG;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetNumberFromObj --
+ *
+ *      Extracts a number (of any possible numeric type) from an object.
+ *
+ * Results:
+ *      Whether the extraction worked. The type is stored in the variable
+ *      referred to by the typePtr argument, and a pointer to the
+ *      representation is stored in the variable referred to by the
+ *      clientDataPtr.
+ *
+ * Side effects:
+ *      Can allocate thread-specific data for handling the copy-out space for
+ *      bignums; this space is shared within a thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetNumberFromObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    ClientData *clientDataPtr,
+    int *typePtr)
+{
+    do {
+	if (objPtr->typePtr == &tclDoubleType) {
+	    if (TclIsNaN(objPtr->internalRep.doubleValue)) {
+		*typePtr = TCL_NUMBER_NAN;
+	    } else {
+		*typePtr = TCL_NUMBER_DOUBLE;
+	    }
+	    *clientDataPtr = &objPtr->internalRep.doubleValue;
+	    return TCL_OK;
+	}
+	if (objPtr->typePtr == &tclIntType) {
+	    *typePtr = TCL_NUMBER_LONG;
+	    *clientDataPtr = &objPtr->internalRep.longValue;
+	    return TCL_OK;
+	}
+#ifndef TCL_WIDE_INT_IS_LONG
+	if (objPtr->typePtr == &tclWideIntType) {
+	    *typePtr = TCL_NUMBER_WIDE;
+	    *clientDataPtr = &objPtr->internalRep.wideValue;
+	    return TCL_OK;
+	}
+#endif
+	if (objPtr->typePtr == &tclBignumType) {
+	    static Tcl_ThreadDataKey bignumKey;
+	    mp_int *bigPtr = Tcl_GetThreadData(&bignumKey,
+		    (int) sizeof(mp_int));
+
+	    UNPACK_BIGNUM(objPtr, *bigPtr);
+	    *typePtr = TCL_NUMBER_BIG;
+	    *clientDataPtr = bigPtr;
+	    return TCL_OK;
+	}
+    } while (TCL_OK ==
+	    TclParseNumber(interp, objPtr, "number", NULL, -1, NULL, 0));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbIncrRefCount --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. This checks to see whether or not the memory
+ *	has been freed before incrementing the ref count.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just increments the
+ *	reference count of the object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's ref count is incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DbIncrRefCount(
+    register Tcl_Obj *objPtr,	/* The object we are registering a reference
+				 * to. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+#ifdef TCL_MEM_DEBUG
+    if (objPtr->refCount == 0x61616161) {
+	fprintf(stderr, "file = %s, line = %d\n", file, line);
+	fflush(stderr);
+	Tcl_Panic("incrementing refCount of previously disposed object");
+    }
+
+# ifdef TCL_THREADS
+    /*
+     * Check to make sure that the Tcl_Obj was allocated by the current
+     * thread. Don't do this check when shutting down since thread local
+     * storage can be finalized before the last Tcl_Obj is freed.
+     */
+
+    if (!TclInExit()) {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+	Tcl_HashTable *tablePtr = tsdPtr->objThreadMap;
+	Tcl_HashEntry *hPtr;
+
+	if (!tablePtr) {
+	    Tcl_Panic("object table not initialized");
+	}
+	hPtr = Tcl_FindHashEntry(tablePtr, objPtr);
+	if (!hPtr) {
+	    Tcl_Panic("Trying to %s of Tcl_Obj allocated in another thread",
+                    "incr ref count");
+	}
+    }
+# endif /* TCL_THREADS */
+#endif /* TCL_MEM_DEBUG */
+    ++(objPtr)->refCount;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbDecrRefCount --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. This checks to see whether or not the memory
+ *	has been freed before decrementing the ref count.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just decrements the
+ *	reference count of the object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's ref count is incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DbDecrRefCount(
+    register Tcl_Obj *objPtr,	/* The object we are releasing a reference
+				 * to. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+#ifdef TCL_MEM_DEBUG
+    if (objPtr->refCount == 0x61616161) {
+	fprintf(stderr, "file = %s, line = %d\n", file, line);
+	fflush(stderr);
+	Tcl_Panic("decrementing refCount of previously disposed object");
+    }
+
+# ifdef TCL_THREADS
+    /*
+     * Check to make sure that the Tcl_Obj was allocated by the current
+     * thread. Don't do this check when shutting down since thread local
+     * storage can be finalized before the last Tcl_Obj is freed.
+     */
+
+    if (!TclInExit()) {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+	Tcl_HashTable *tablePtr = tsdPtr->objThreadMap;
+	Tcl_HashEntry *hPtr;
+
+	if (!tablePtr) {
+	    Tcl_Panic("object table not initialized");
+	}
+	hPtr = Tcl_FindHashEntry(tablePtr, objPtr);
+	if (!hPtr) {
+	    Tcl_Panic("Trying to %s of Tcl_Obj allocated in another thread",
+                    "decr ref count");
+	}
+    }
+# endif /* TCL_THREADS */
+#endif /* TCL_MEM_DEBUG */
+
+    if (objPtr->refCount-- <= 1) {
+	TclFreeObj(objPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbIsShared --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It tests whether the object has a ref count
+ *	greater than one.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just tests if the
+ *	object has a ref count greater than one.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_DbIsShared(
+    register Tcl_Obj *objPtr,	/* The object to test for being shared. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+#ifdef TCL_MEM_DEBUG
+    if (objPtr->refCount == 0x61616161) {
+	fprintf(stderr, "file = %s, line = %d\n", file, line);
+	fflush(stderr);
+	Tcl_Panic("checking whether previously disposed object is shared");
+    }
+
+# ifdef TCL_THREADS
+    /*
+     * Check to make sure that the Tcl_Obj was allocated by the current
+     * thread. Don't do this check when shutting down since thread local
+     * storage can be finalized before the last Tcl_Obj is freed.
+     */
+
+    if (!TclInExit()) {
+	ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+	Tcl_HashTable *tablePtr = tsdPtr->objThreadMap;
+	Tcl_HashEntry *hPtr;
+
+	if (!tablePtr) {
+	    Tcl_Panic("object table not initialized");
+	}
+	hPtr = Tcl_FindHashEntry(tablePtr, objPtr);
+	if (!hPtr) {
+	    Tcl_Panic("Trying to %s of Tcl_Obj allocated in another thread",
+                    "check shared status");
+	}
+    }
+# endif /* TCL_THREADS */
+#endif /* TCL_MEM_DEBUG */
+
+#ifdef TCL_COMPILE_STATS
+    Tcl_MutexLock(&tclObjMutex);
+    if ((objPtr)->refCount <= 1) {
+	tclObjsShared[1]++;
+    } else if ((objPtr)->refCount < TCL_MAX_SHARED_OBJ_STATS) {
+	tclObjsShared[(objPtr)->refCount]++;
+    } else {
+	tclObjsShared[0]++;
+    }
+    Tcl_MutexUnlock(&tclObjMutex);
+#endif /* TCL_COMPILE_STATS */
+
+    return ((objPtr)->refCount > 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitObjHashTable --
+ *
+ *	Given storage for a hash table, set up the fields to prepare the hash
+ *	table for use, the keys are Tcl_Obj *.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	TablePtr is now ready to be passed to Tcl_FindHashEntry and
+ *	Tcl_CreateHashEntry.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_InitObjHashTable(
+    register Tcl_HashTable *tablePtr)
+				/* Pointer to table record, which is supplied
+				 * by the caller. */
+{
+    Tcl_InitCustomHashTable(tablePtr, TCL_CUSTOM_PTR_KEYS,
+	    &tclObjHashKeyType);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AllocObjEntry --
+ *
+ *	Allocate space for a Tcl_HashEntry containing the Tcl_Obj * key.
+ *
+ * Results:
+ *	The return value is a pointer to the created entry.
+ *
+ * Side effects:
+ *	Increments the reference count on the object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashEntry *
+AllocObjEntry(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key to store in the hash table entry. */
+{
+    Tcl_Obj *objPtr = keyPtr;
+    Tcl_HashEntry *hPtr = ckalloc(sizeof(Tcl_HashEntry));
+
+    hPtr->key.objPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);
+    hPtr->clientData = NULL;
+
+    return hPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCompareObjKeys --
+ *
+ *	Compares two Tcl_Obj * keys.
+ *
+ * Results:
+ *	The return value is 0 if they are different and 1 if they are the
+ *	same.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCompareObjKeys(
+    void *keyPtr,		/* New key to compare. */
+    Tcl_HashEntry *hPtr)	/* Existing key to compare. */
+{
+    Tcl_Obj *objPtr1 = keyPtr;
+    Tcl_Obj *objPtr2 = (Tcl_Obj *) hPtr->key.oneWordValue;
+    register const char *p1, *p2;
+    register size_t l1, l2;
+
+    /*
+     * If the object pointers are the same then they match.
+     * OPT: this comparison was moved to the caller
+
+       if (objPtr1 == objPtr2) return 1;
+    */
+
+    /*
+     * Don't use Tcl_GetStringFromObj as it would prevent l1 and l2 being
+     * in a register.
+     */
+
+    p1 = TclGetString(objPtr1);
+    l1 = objPtr1->length;
+    p2 = TclGetString(objPtr2);
+    l2 = objPtr2->length;
+
+    /*
+     * Only compare if the string representations are of the same length.
+     */
+
+    if (l1 == l2) {
+	for (;; p1++, p2++, l1--) {
+	    if (*p1 != *p2) {
+		break;
+	    }
+	    if (l1 == 0) {
+		return 1;
+	    }
+	}
+    }
+
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreeObjEntry --
+ *
+ *	Frees space for a Tcl_HashEntry containing the Tcl_Obj * key.
+ *
+ * Results:
+ *	The return value is a pointer to the created entry.
+ *
+ * Side effects:
+ *	Decrements the reference count of the object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFreeObjEntry(
+    Tcl_HashEntry *hPtr)	/* Hash entry to free. */
+{
+    Tcl_Obj *objPtr = (Tcl_Obj *) hPtr->key.oneWordValue;
+
+    Tcl_DecrRefCount(objPtr);
+    ckfree(hPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclHashObjKey --
+ *
+ *	Compute a one-word summary of the string representation of the
+ *	Tcl_Obj, which can be used to generate a hash index.
+ *
+ * Results:
+ *	The return value is a one-word summary of the information in the
+ *	string representation of the Tcl_Obj.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+TCL_HASH_TYPE
+TclHashObjKey(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key from which to compute hash value. */
+{
+    Tcl_Obj *objPtr = keyPtr;
+    int length;
+    const char *string = TclGetStringFromObj(objPtr, &length);
+    unsigned int result = 0;
+
+    /*
+     * I tried a zillion different hash functions and asked many other people
+     * for advice. Many people had their own favorite functions, all
+     * different, but no-one had much idea why they were good ones. I chose
+     * the one below (multiply by 9 and add new character) because of the
+     * following reasons:
+     *
+     * 1. Multiplying by 10 is perfect for keys that are decimal strings, and
+     *    multiplying by 9 is just about as good.
+     * 2. Times-9 is (shift-left-3) plus (old). This means that each
+     *    character's bits hang around in the low-order bits of the hash value
+     *    for ever, plus they spread fairly rapidly up to the high-order bits
+     *    to fill out the hash value. This seems works well both for decimal
+     *    and non-decimal strings.
+     *
+     * Note that this function is very weak against malicious strings; it's
+     * very easy to generate multiple keys that have the same hashcode. On the
+     * other hand, that hardly ever actually occurs and this function *is*
+     * very cheap, even by comparison with industry-standard hashes like FNV.
+     * If real strength of hash is required though, use a custom hash based on
+     * Bob Jenkins's lookup3(), but be aware that it's significantly slower.
+     * Tcl does not use that level of strength because it typically does not
+     * need it (and some of the aspects of that strength are genuinely
+     * unnecessary given the rest of Tcl's hash machinery, and the fact that
+     * we do not either transfer hashes to another machine, use them as a true
+     * substitute for equality, or attempt to minimize work in rebuilding the
+     * hash table).
+     *
+     * See also HashStringKey in tclHash.c.
+     * See also HashString in tclLiteral.c.
+     *
+     * See [tcl-Feature Request #2958832]
+     */
+
+    if (length > 0) {
+	result = UCHAR(*string);
+	while (--length) {
+	    result += (result << 3) + UCHAR(*++string);
+	}
+    }
+    return (TCL_HASH_TYPE) result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCommandFromObj --
+ *
+ *	Returns the command specified by the name in a Tcl_Obj.
+ *
+ * Results:
+ *	Returns a token for the command if it is found. Otherwise, if it can't
+ *	be found or there is an error, returns NULL.
+ *
+ * Side effects:
+ *	May update the internal representation for the object, caching the
+ *	command reference so that the next time this function is called with
+ *	the same object, the command can be found quickly.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Command
+Tcl_GetCommandFromObj(
+    Tcl_Interp *interp,		/* The interpreter in which to resolve the
+				 * command and to report errors. */
+    register Tcl_Obj *objPtr)	/* The object containing the command's name.
+				 * If the name starts with "::", will be
+				 * looked up in global namespace. Else, looked
+				 * up first in the current namespace, then in
+				 * global namespace. */
+{
+    register ResolvedCmdName *resPtr;
+
+    /*
+     * Get the internal representation, converting to a command type if
+     * needed. The internal representation is a ResolvedCmdName that points to
+     * the actual command.
+     *
+     * Check the context namespace and the namespace epoch of the resolved
+     * symbol to make sure that it is fresh. Note that we verify that the
+     * namespace id of the context namespace is the same as the one we cached;
+     * this insures that the namespace wasn't deleted and a new one created at
+     * the same address with the same command epoch. Note that fully qualified
+     * names have a NULL refNsPtr, these checks needn't be made.
+     *
+     * Check also that the command's epoch is up to date, and that the command
+     * is not deleted.
+     *
+     * If any check fails, then force another conversion to the command type,
+     * to discard the old rep and create a new one.
+     */
+
+    resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    if (objPtr->typePtr == &tclCmdNameType) {
+        register Command *cmdPtr = resPtr->cmdPtr;
+
+        if ((cmdPtr->cmdEpoch == resPtr->cmdEpoch)
+                && (interp == cmdPtr->nsPtr->interp)
+                && !(cmdPtr->nsPtr->flags & NS_DYING)) {
+            register Namespace *refNsPtr = (Namespace *)
+                    TclGetCurrentNamespace(interp);
+
+            if ((resPtr->refNsPtr == NULL)
+                || ((refNsPtr == resPtr->refNsPtr)
+                    && (resPtr->refNsId == refNsPtr->nsId)
+                    && (resPtr->refNsCmdEpoch == refNsPtr->cmdRefEpoch))) {
+                return (Tcl_Command) cmdPtr;
+            }
+        }
+    }
+
+    /*
+     * OK, must create a new internal representation (or fail) as any cache we
+     * had is invalid one way or another.
+     */
+
+    /* See [07d13d99b0a9] why we cannot call SetCmdNameFromAny() directly here. */
+    if (tclCmdNameType.setFromAnyProc(interp, objPtr) != TCL_OK) {
+        return NULL;
+    }
+    resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    return (Tcl_Command) (resPtr ? resPtr->cmdPtr : NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetCmdNameObj --
+ *
+ *	Modify an object to be an CmdName object that refers to the argument
+ *	Command structure.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's old internal rep is freed. It's string rep is not
+ *	changed. The refcount in the Command structure is incremented to keep
+ *	it from being freed if the command is later deleted until
+ *	TclNRExecuteByteCode has a chance to recognize that it was deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+SetCmdNameObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    Command *cmdPtr,
+    ResolvedCmdName *resPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    ResolvedCmdName *fillPtr;
+    const char *name = TclGetString(objPtr);
+
+    if (resPtr) {
+	fillPtr = resPtr;
+    } else {
+	fillPtr = ckalloc(sizeof(ResolvedCmdName));
+	fillPtr->refCount = 1;
+    }
+
+    fillPtr->cmdPtr = cmdPtr;
+    cmdPtr->refCount++;
+    fillPtr->cmdEpoch = cmdPtr->cmdEpoch;
+
+    /* NOTE: relying on NULL termination here. */
+    if ((name[0] == ':') && (name[1] == ':')) {
+	/*
+	 * Fully qualified names always resolve to same thing. No need
+	 * to record resolution context information.
+	 */
+
+	fillPtr->refNsPtr = NULL;
+	fillPtr->refNsId = 0;		/* Will not be read */
+	fillPtr->refNsCmdEpoch = 0;	/* Will not be read */
+    } else {
+	/*
+	 * Record current state of current namespace as the resolution
+	 * context of this command name lookup.
+	 */
+	Namespace *currNsPtr = iPtr->varFramePtr->nsPtr;
+
+	fillPtr->refNsPtr = currNsPtr;
+	fillPtr->refNsId = currNsPtr->nsId;
+	fillPtr->refNsCmdEpoch = currNsPtr->cmdRefEpoch;
+    }
+
+    if (resPtr == NULL) {
+	TclFreeIntRep(objPtr);
+
+	objPtr->internalRep.twoPtrValue.ptr1 = fillPtr;
+	objPtr->internalRep.twoPtrValue.ptr2 = NULL;
+	objPtr->typePtr = &tclCmdNameType;
+    }
+}
+
+void
+TclSetCmdNameObj(
+    Tcl_Interp *interp,		/* Points to interpreter containing command
+				 * that should be cached in objPtr. */
+    register Tcl_Obj *objPtr,	/* Points to Tcl object to be changed to a
+				 * CmdName object. */
+    Command *cmdPtr)		/* Points to Command structure that the
+				 * CmdName object should refer to. */
+{
+    register ResolvedCmdName *resPtr;
+
+    if (objPtr->typePtr == &tclCmdNameType) {
+	resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+	if (resPtr != NULL && resPtr->cmdPtr == cmdPtr) {
+	    return;
+	}
+    }
+
+    SetCmdNameObj(interp, objPtr, cmdPtr, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeCmdNameInternalRep --
+ *
+ *	Frees the resources associated with a cmdName object's internal
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Decrements the ref count of any cached ResolvedCmdName structure
+ *	pointed to by the cmdName's internal representation. If this is the
+ *	last use of the ResolvedCmdName, it is freed. This in turn decrements
+ *	the ref count of the Command structure pointed to by the
+ *	ResolvedSymbol, which may free the Command structure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeCmdNameInternalRep(
+    register Tcl_Obj *objPtr)	/* CmdName object with internal
+				 * representation to free. */
+{
+    register ResolvedCmdName *resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+	/*
+	 * Decrement the reference count of the ResolvedCmdName structure. If
+	 * there are no more uses, free the ResolvedCmdName structure.
+	 */
+
+	if (resPtr->refCount-- <= 1) {
+	    /*
+	     * Now free the cached command, unless it is still in its hash
+	     * table or if there are other references to it from other cmdName
+	     * objects.
+	     */
+
+	    Command *cmdPtr = resPtr->cmdPtr;
+
+	    TclCleanupCommandMacro(cmdPtr);
+	    ckfree(resPtr);
+	}
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupCmdNameInternalRep --
+ *
+ *	Initialize the internal representation of an cmdName Tcl_Obj to a copy
+ *	of the internal representation of an existing cmdName object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	"copyPtr"s internal rep is set to point to the ResolvedCmdName
+ *	structure corresponding to "srcPtr"s internal rep. Increments the ref
+ *	count of the ResolvedCmdName structure pointed to by the cmdName's
+ *	internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupCmdNameInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    register Tcl_Obj *copyPtr)	/* Object with internal rep to set. */
+{
+    register ResolvedCmdName *resPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    copyPtr->internalRep.twoPtrValue.ptr1 = resPtr;
+    copyPtr->internalRep.twoPtrValue.ptr2 = NULL;
+	resPtr->refCount++;
+    copyPtr->typePtr = &tclCmdNameType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetCmdNameFromAny --
+ *
+ *	Generate an cmdName internal form for the Tcl object "objPtr".
+ *
+ * Results:
+ *	The return value is a standard Tcl result. The conversion always
+ *	succeeds and TCL_OK is returned.
+ *
+ * Side effects:
+ *	A pointer to a ResolvedCmdName structure that holds a cached pointer
+ *	to the command with a name that matches objPtr's string rep is stored
+ *	as objPtr's internal representation. This ResolvedCmdName pointer will
+ *	be NULL if no matching command was found. The ref count of the cached
+ *	Command's structure (if any) is also incremented.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetCmdNameFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    const char *name;
+    register Command *cmdPtr;
+    register ResolvedCmdName *resPtr;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the Command structure, if any, that describes the command called
+     * "name". Build a ResolvedCmdName that holds a cached pointer to this
+     * Command, and bump the reference count in the referenced Command
+     * structure. A Command structure will not be deleted as long as it is
+     * referenced from a CmdName object.
+     */
+
+    name = TclGetString(objPtr);
+    cmdPtr = (Command *)
+	    Tcl_FindCommand(interp, name, /*ns*/ NULL, /*flags*/ 0);
+
+    /*
+     * Stop shimmering and caching nothing when we found nothing.  Just
+     * report the failure to find the command as an error.
+     */
+
+    if (cmdPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    resPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    if ((objPtr->typePtr == &tclCmdNameType) && (resPtr->refCount == 1)) {
+	/*
+	 * Re-use existing ResolvedCmdName struct when possible.
+	 * Cleanup the old fields that need it.
+	 */
+
+	Command *oldCmdPtr = resPtr->cmdPtr;
+
+	if (oldCmdPtr->refCount-- <= 1) {
+	    TclCleanupCommandMacro(oldCmdPtr);
+	}
+    } else {
+	resPtr = NULL;
+    }
+
+    SetCmdNameObj(interp, objPtr, cmdPtr, resPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RepresentationCmd --
+ *
+ *	Implementation of the "tcl::unsupported::representation" command.
+ *
+ * Results:
+ *	Reports the current representation (Tcl_Obj type) of its argument.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RepresentationCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_Obj *descObj;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "value");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Value is a bignum with a refcount of 14, object pointer at 0x12345678,
+     * internal representation 0x45671234:0x98765432, string representation
+     * "1872361827361287"
+     */
+
+    descObj = Tcl_ObjPrintf("value is a %s with a refcount of %d,"
+	    " object pointer at %p",
+	    objv[1]->typePtr ? objv[1]->typePtr->name : "pure string",
+	    objv[1]->refCount, objv[1]);
+
+    if (objv[1]->typePtr) {
+	if (objv[1]->typePtr == &tclDoubleType) {
+	    Tcl_AppendPrintfToObj(descObj, ", internal representation %g",
+		    objv[1]->internalRep.doubleValue);
+	} else {
+	    Tcl_AppendPrintfToObj(descObj, ", internal representation %p:%p",
+		    (void *) objv[1]->internalRep.twoPtrValue.ptr1,
+		    (void *) objv[1]->internalRep.twoPtrValue.ptr2);
+	}
+    }
+
+    if (objv[1]->bytes) {
+        Tcl_AppendToObj(descObj, ", string representation \"", -1);
+	Tcl_AppendLimitedToObj(descObj, objv[1]->bytes, objv[1]->length,
+                16, "...");
+	Tcl_AppendToObj(descObj, "\"", -1);
+    } else {
+	Tcl_AppendToObj(descObj, ", no string representation", -1);
+    }
+
+    Tcl_SetObjResult(interp, descObj);
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclOptimize.c b/generic/tclOptimize.c
new file mode 100644
index 0000000..8267a7d
--- /dev/null
+++ b/generic/tclOptimize.c
@@ -0,0 +1,444 @@
+/*
+ * tclOptimize.c --
+ *
+ *	This file contains the bytecode optimizer.
+ *
+ * Copyright (c) 2013 by Donal Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+#include <assert.h>
+
+/*
+ * Forward declarations.
+ */
+
+static void		AdvanceJumps(CompileEnv *envPtr);
+static void		ConvertZeroEffectToNOP(CompileEnv *envPtr);
+static void		LocateTargetAddresses(CompileEnv *envPtr,
+			    Tcl_HashTable *tablePtr);
+static void		TrimUnreachable(CompileEnv *envPtr);
+
+/*
+ * Helper macros.
+ */
+
+#define DefineTargetAddress(tablePtr, address) \
+    ((void) Tcl_CreateHashEntry((tablePtr), (void *) (address), &isNew))
+#define IsTargetAddress(tablePtr, address) \
+    (Tcl_FindHashEntry((tablePtr), (void *) (address)) != NULL)
+#define AddrLength(address) \
+    (tclInstructionTable[*(unsigned char *)(address)].numBytes)
+#define InstLength(instruction) \
+    (tclInstructionTable[(unsigned char)(instruction)].numBytes)
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * LocateTargetAddresses --
+ *
+ *	Populate a hash table with places that we need to be careful around
+ *	because they're the targets of various kinds of jumps and other
+ *	non-local behavior.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+LocateTargetAddresses(
+    CompileEnv *envPtr,
+    Tcl_HashTable *tablePtr)
+{
+    unsigned char *currentInstPtr, *targetInstPtr;
+    int isNew, i;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch hSearch;
+
+    Tcl_InitHashTable(tablePtr, TCL_ONE_WORD_KEYS);
+
+    /*
+     * The starts of commands represent target addresses.
+     */
+
+    for (i=0 ; i<envPtr->numCommands ; i++) {
+	DefineTargetAddress(tablePtr,
+		envPtr->codeStart + envPtr->cmdMapPtr[i].codeOffset);
+    }
+
+    /*
+     * Find places where we should be careful about replacing instructions
+     * because they are the targets of various types of jumps.
+     */
+
+    for (currentInstPtr = envPtr->codeStart ;
+	    currentInstPtr < envPtr->codeNext ;
+	    currentInstPtr += AddrLength(currentInstPtr)) {
+	switch (*currentInstPtr) {
+	case INST_JUMP1:
+	case INST_JUMP_TRUE1:
+	case INST_JUMP_FALSE1:
+	    targetInstPtr = currentInstPtr+TclGetInt1AtPtr(currentInstPtr+1);
+	    goto storeTarget;
+	case INST_JUMP4:
+	case INST_JUMP_TRUE4:
+	case INST_JUMP_FALSE4:
+	case INST_START_CMD:
+	    targetInstPtr = currentInstPtr+TclGetInt4AtPtr(currentInstPtr+1);
+	    goto storeTarget;
+	case INST_BEGIN_CATCH4:
+	    targetInstPtr = envPtr->codeStart + envPtr->exceptArrayPtr[
+		    TclGetUInt4AtPtr(currentInstPtr+1)].codeOffset;
+	storeTarget:
+	    DefineTargetAddress(tablePtr, targetInstPtr);
+	    break;
+	case INST_JUMP_TABLE:
+	    hPtr = Tcl_FirstHashEntry(
+		    &JUMPTABLEINFO(envPtr, currentInstPtr+1)->hashTable,
+		    &hSearch);
+	    for (; hPtr ; hPtr = Tcl_NextHashEntry(&hSearch)) {
+		targetInstPtr = currentInstPtr +
+			PTR2INT(Tcl_GetHashValue(hPtr));
+		DefineTargetAddress(tablePtr, targetInstPtr);
+	    }
+	    break;
+	case INST_RETURN_CODE_BRANCH:
+	    for (i=TCL_ERROR ; i<TCL_CONTINUE+1 ; i++) {
+		DefineTargetAddress(tablePtr, currentInstPtr + 2*i - 1);
+	    }
+	    break;
+	}
+    }
+
+    /*
+     * Add a marker *after* the last bytecode instruction. WARNING: points to
+     * one past the end!
+     */
+
+    DefineTargetAddress(tablePtr, currentInstPtr);
+
+    /*
+     * Enter in the targets of exception ranges.
+     */
+
+    for (i=0 ; i<envPtr->exceptArrayNext ; i++) {
+	ExceptionRange *rangePtr = &envPtr->exceptArrayPtr[i];
+
+	if (rangePtr->type == CATCH_EXCEPTION_RANGE) {
+	    targetInstPtr = envPtr->codeStart + rangePtr->catchOffset;
+	    DefineTargetAddress(tablePtr, targetInstPtr);
+	} else {
+	    targetInstPtr = envPtr->codeStart + rangePtr->breakOffset;
+	    DefineTargetAddress(tablePtr, targetInstPtr);
+	    if (rangePtr->continueOffset >= 0) {
+		targetInstPtr = envPtr->codeStart + rangePtr->continueOffset;
+		DefineTargetAddress(tablePtr, targetInstPtr);
+	    }
+	}
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TrimUnreachable --
+ *
+ *	Converts code that provably can't be executed into NOPs and reduces
+ *	the overall reported length of the bytecode where that is possible.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+TrimUnreachable(
+    CompileEnv *envPtr)
+{
+    unsigned char *currentInstPtr;
+    Tcl_HashTable targets;
+
+    LocateTargetAddresses(envPtr, &targets);
+
+    for (currentInstPtr = envPtr->codeStart ;
+	    currentInstPtr < envPtr->codeNext-1 ;
+	    currentInstPtr += AddrLength(currentInstPtr)) {
+	int clear = 0;
+
+	if (*currentInstPtr != INST_DONE) {
+	    continue;
+	}
+
+	while (!IsTargetAddress(&targets, currentInstPtr + 1 + clear)) {
+	    clear += AddrLength(currentInstPtr + 1 + clear);
+	}
+	if (currentInstPtr + 1 + clear == envPtr->codeNext) {
+	    envPtr->codeNext -= clear;
+	} else {
+	    while (clear --> 0) {
+		*(currentInstPtr + 1 + clear) = INST_NOP;
+	    }
+	}
+    }
+
+    Tcl_DeleteHashTable(&targets);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * ConvertZeroEffectToNOP --
+ *
+ *	Replace PUSH/POP sequences (when non-hazardous) with NOPs. Also
+ *	replace PUSH empty/STR_CONCAT and TRY_CVT_NUMERIC (when followed by an
+ *	operation that guarantees the check for arithmeticity) and eliminate
+ *	LNOT when we can invert the following JUMP condition.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+ConvertZeroEffectToNOP(
+    CompileEnv *envPtr)
+{
+    unsigned char *currentInstPtr;
+    int size;
+    Tcl_HashTable targets;
+
+    LocateTargetAddresses(envPtr, &targets);
+    for (currentInstPtr = envPtr->codeStart ;
+	    currentInstPtr < envPtr->codeNext ; currentInstPtr += size) {
+	int blank = 0, i, nextInst;
+
+	size = AddrLength(currentInstPtr);
+	while ((currentInstPtr + size < envPtr->codeNext)
+		&& *(currentInstPtr+size) == INST_NOP) {
+	    if (IsTargetAddress(&targets, currentInstPtr + size)) {
+		break;
+	    }
+	    size += InstLength(INST_NOP);
+	}
+	if (IsTargetAddress(&targets, currentInstPtr + size)) {
+	    continue;
+	}
+	nextInst = *(currentInstPtr + size);
+	switch (*currentInstPtr) {
+	case INST_PUSH1:
+	    if (nextInst == INST_POP) {
+		blank = size + InstLength(nextInst);
+	    } else if (nextInst == INST_STR_CONCAT1
+		    && TclGetUInt1AtPtr(currentInstPtr + size + 1) == 2) {
+		Tcl_Obj *litPtr = TclFetchLiteral(envPtr,
+			TclGetUInt1AtPtr(currentInstPtr + 1));
+		int numBytes;
+
+		(void) TclGetStringFromObj(litPtr, &numBytes);
+		if (numBytes == 0) {
+		    blank = size + InstLength(nextInst);
+		}
+	    }
+	    break;
+	case INST_PUSH4:
+	    if (nextInst == INST_POP) {
+		blank = size + 1;
+	    } else if (nextInst == INST_STR_CONCAT1
+		    && TclGetUInt1AtPtr(currentInstPtr + size + 1) == 2) {
+		Tcl_Obj *litPtr = TclFetchLiteral(envPtr,
+			TclGetUInt4AtPtr(currentInstPtr + 1));
+		int numBytes;
+
+		(void) TclGetStringFromObj(litPtr, &numBytes);
+		if (numBytes == 0) {
+		    blank = size + InstLength(nextInst);
+		}
+	    }
+	    break;
+
+	case INST_LNOT:
+	    switch (nextInst) {
+	    case INST_JUMP_TRUE1:
+		blank = size;
+		*(currentInstPtr + size) = INST_JUMP_FALSE1;
+		break;
+	    case INST_JUMP_FALSE1:
+		blank = size;
+		*(currentInstPtr + size) = INST_JUMP_TRUE1;
+		break;
+	    case INST_JUMP_TRUE4:
+		blank = size;
+		*(currentInstPtr + size) = INST_JUMP_FALSE4;
+		break;
+	    case INST_JUMP_FALSE4:
+		blank = size;
+		*(currentInstPtr + size) = INST_JUMP_TRUE4;
+		break;
+	    }
+	    break;
+
+	case INST_TRY_CVT_TO_NUMERIC:
+	    switch (nextInst) {
+	    case INST_JUMP_TRUE1:
+	    case INST_JUMP_TRUE4:
+	    case INST_JUMP_FALSE1:
+	    case INST_JUMP_FALSE4:
+	    case INST_INCR_SCALAR1:
+	    case INST_INCR_ARRAY1:
+	    case INST_INCR_ARRAY_STK:
+	    case INST_INCR_SCALAR_STK:
+	    case INST_INCR_STK:
+	    case INST_LOR:
+	    case INST_LAND:
+	    case INST_EQ:
+	    case INST_NEQ:
+	    case INST_LT:
+	    case INST_LE:
+	    case INST_GT:
+	    case INST_GE:
+	    case INST_MOD:
+	    case INST_LSHIFT:
+	    case INST_RSHIFT:
+	    case INST_BITOR:
+	    case INST_BITXOR:
+	    case INST_BITAND:
+	    case INST_EXPON:
+	    case INST_ADD:
+	    case INST_SUB:
+	    case INST_DIV:
+	    case INST_MULT:
+	    case INST_LNOT:
+	    case INST_BITNOT:
+	    case INST_UMINUS:
+	    case INST_UPLUS:
+	    case INST_TRY_CVT_TO_NUMERIC:
+		blank = size;
+		break;
+	    }
+	    break;
+	}
+
+	if (blank > 0) {
+	    for (i=0 ; i<blank ; i++) {
+		*(currentInstPtr + i) = INST_NOP;
+	    }
+	    size = blank;
+	}
+    }
+    Tcl_DeleteHashTable(&targets);
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * AdvanceJumps --
+ *
+ *	Advance jumps past NOPs and chained JUMPs. After this runs, the only
+ *	JUMPs that jump to a NOP or a JUMP will be length-1 ones that run out
+ *	of room in their opcode to be targeted to where they really belong.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+static void
+AdvanceJumps(
+    CompileEnv *envPtr)
+{
+    unsigned char *currentInstPtr;
+    Tcl_HashTable jumps;
+
+    for (currentInstPtr = envPtr->codeStart ;
+	    currentInstPtr < envPtr->codeNext-1 ;
+	    currentInstPtr += AddrLength(currentInstPtr)) {
+	int offset, delta, isNew;
+
+	switch (*currentInstPtr) {
+	case INST_JUMP1:
+	case INST_JUMP_TRUE1:
+	case INST_JUMP_FALSE1:
+	    offset = TclGetInt1AtPtr(currentInstPtr + 1);
+	    Tcl_InitHashTable(&jumps, TCL_ONE_WORD_KEYS);
+	    for (delta=0 ; offset+delta != 0 ;) {
+		if (offset + delta < -128 || offset + delta > 127) {
+		    break;
+		}
+		Tcl_CreateHashEntry(&jumps, INT2PTR(offset), &isNew);
+		if (!isNew) {
+		    offset = TclGetInt1AtPtr(currentInstPtr + 1);
+		    break;
+		}
+		offset += delta;
+		switch (*(currentInstPtr + offset)) {
+		case INST_NOP:
+		    delta = InstLength(INST_NOP);
+		    continue;
+		case INST_JUMP1:
+		    delta = TclGetInt1AtPtr(currentInstPtr + offset + 1);
+		    continue;
+		case INST_JUMP4:
+		    delta = TclGetInt4AtPtr(currentInstPtr + offset + 1);
+		    continue;
+		}
+		break;
+	    }
+	    Tcl_DeleteHashTable(&jumps);
+	    TclStoreInt1AtPtr(offset, currentInstPtr + 1);
+	    continue;
+
+	case INST_JUMP4:
+	case INST_JUMP_TRUE4:
+	case INST_JUMP_FALSE4:
+	    Tcl_InitHashTable(&jumps, TCL_ONE_WORD_KEYS);
+	    Tcl_CreateHashEntry(&jumps, INT2PTR(0), &isNew);
+	    for (offset = TclGetInt4AtPtr(currentInstPtr + 1); offset!=0 ;) {
+		Tcl_CreateHashEntry(&jumps, INT2PTR(offset), &isNew);
+		if (!isNew) {
+		    offset = TclGetInt4AtPtr(currentInstPtr + 1);
+		    break;
+		}
+		switch (*(currentInstPtr + offset)) {
+		case INST_NOP:
+		    offset += InstLength(INST_NOP);
+		    continue;
+		case INST_JUMP1:
+		    offset += TclGetInt1AtPtr(currentInstPtr + offset + 1);
+		    continue;
+		case INST_JUMP4:
+		    offset += TclGetInt4AtPtr(currentInstPtr + offset + 1);
+		    continue;
+		}
+		break;
+	    }
+	    Tcl_DeleteHashTable(&jumps);
+	    TclStoreInt4AtPtr(offset, currentInstPtr + 1);
+	    continue;
+	}
+    }
+}
+
+/*
+ * ----------------------------------------------------------------------
+ *
+ * TclOptimizeBytecode --
+ *
+ *	A very simple peephole optimizer for bytecode.
+ *
+ * ----------------------------------------------------------------------
+ */
+
+void
+TclOptimizeBytecode(
+    void *envPtr)
+{
+    ConvertZeroEffectToNOP(envPtr);
+    AdvanceJumps(envPtr);
+    TrimUnreachable(envPtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * End:
+ */
diff --git a/generic/tclPanic.c b/generic/tclPanic.c
new file mode 100644
index 0000000..b03ad41
--- /dev/null
+++ b/generic/tclPanic.c
@@ -0,0 +1,170 @@
+/*
+ * tclPanic.c --
+ *
+ *	Source code for the "Tcl_Panic" library procedure for Tcl; individual
+ *	applications will probably call Tcl_SetPanicProc() to set an
+ *	application-specific panic procedure.
+ *
+ * Copyright (c) 1988-1993 The Regents of the University of California.
+ * Copyright (c) 1994 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#if defined(_WIN32) || defined(__CYGWIN__)
+    MODULE_SCOPE TCL_NORETURN void tclWinDebugPanic(const char *format, ...);
+#endif
+
+/*
+ * The panicProc variable contains a pointer to an application specific panic
+ * procedure.
+ */
+
+#if defined(__CYGWIN__)
+static TCL_NORETURN Tcl_PanicProc *panicProc = tclWinDebugPanic;
+#else
+static TCL_NORETURN1 Tcl_PanicProc *panicProc = NULL;
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetPanicProc --
+ *
+ *	Replace the default panic behavior with the specified function.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Sets the panicProc variable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetPanicProc(
+    TCL_NORETURN1 Tcl_PanicProc *proc)
+{
+#if defined(_WIN32)
+    /* tclWinDebugPanic only installs if there is no panicProc yet. */
+    if ((proc != tclWinDebugPanic) || (panicProc == NULL))
+#elif defined(__CYGWIN__)
+    if (proc == NULL)
+	panicProc = tclWinDebugPanic;
+    else
+#endif
+    panicProc = proc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PanicVA --
+ *
+ *	Print an error message and kill the process.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The process dies, entering the debugger if possible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_PanicVA(
+    const char *format,		/* Format string, suitable for passing to
+				 * fprintf. */
+    va_list argList)		/* Variable argument list. */
+{
+    char *arg1, *arg2, *arg3;	/* Additional arguments (variable in number)
+				 * to pass to fprintf. */
+    char *arg4, *arg5, *arg6, *arg7, *arg8;
+
+    arg1 = va_arg(argList, char *);
+    arg2 = va_arg(argList, char *);
+    arg3 = va_arg(argList, char *);
+    arg4 = va_arg(argList, char *);
+    arg5 = va_arg(argList, char *);
+    arg6 = va_arg(argList, char *);
+    arg7 = va_arg(argList, char *);
+    arg8 = va_arg(argList, char *);
+
+    if (panicProc != NULL) {
+	panicProc(format, arg1, arg2, arg3, arg4, arg5, arg6, arg7, arg8);
+#ifdef _WIN32
+    } else if (IsDebuggerPresent()) {
+	tclWinDebugPanic(format, arg1, arg2, arg3, arg4, arg5, arg6, arg7, arg8);
+#endif
+    } else {
+	fprintf(stderr, format, arg1, arg2, arg3, arg4, arg5, arg6, arg7,
+		arg8);
+	fprintf(stderr, "\n");
+	fflush(stderr);
+#if defined(_WIN32) || defined(__CYGWIN__)
+#   if defined(__GNUC__)
+	__builtin_trap();
+#   elif defined(_WIN64)
+	__debugbreak();
+#   elif defined(_MSC_VER) && defined (_M_IX86)
+	_asm {int 3}
+#   else
+	DebugBreak();
+#   endif
+#endif
+#if defined(_WIN32)
+	ExitProcess(1);
+#else
+	abort();
+#endif
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Panic --
+ *
+ *	Print an error message and kill the process.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The process dies, entering the debugger if possible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+/* ARGSUSED */
+
+/*
+ * The following comment is here so that Coverity's static analizer knows that
+ * a Tcl_Panic() call can never return and avoids lots of false positives.
+ */
+
+/* coverity[+kill] */
+void
+Tcl_Panic(
+    const char *format,
+    ...)
+{
+    va_list argList;
+
+    va_start(argList, format);
+    Tcl_PanicVA(format, argList);
+    va_end (argList);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclParse.c b/generic/tclParse.c
new file mode 100644
index 0000000..a2227f7
--- /dev/null
+++ b/generic/tclParse.c
@@ -0,0 +1,2513 @@
+/*
+ * tclParse.c --
+ *
+ *	This file contains functions that parse Tcl scripts. They do so in a
+ *	general-purpose fashion that can be used for many different purposes,
+ *	including compilation, direct execution, code analysis, etc.
+ *
+ * Copyright (c) 1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 Ajuba Solutions.
+ * Contributions from Don Porter, NIST, 2002. (not subject to US copyright)
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclParse.h"
+#include <assert.h>
+
+/*
+ * The following table provides parsing information about each possible 8-bit
+ * character. The table is designed to be referenced with either signed or
+ * unsigned characters, so it has 384 entries. The first 128 entries
+ * correspond to negative character values, the next 256 correspond to
+ * positive character values. The last 128 entries are identical to the first
+ * 128. The table is always indexed with a 128-byte offset (the 128th entry
+ * corresponds to a character value of 0).
+ *
+ * The macro CHAR_TYPE is used to index into the table and return information
+ * about its character argument. The following return values are defined.
+ *
+ * TYPE_NORMAL -	All characters that don't have special significance to
+ *			the Tcl parser.
+ * TYPE_SPACE -		The character is a whitespace character other than
+ *			newline.
+ * TYPE_COMMAND_END -	Character is newline or semicolon.
+ * TYPE_SUBS -		Character begins a substitution or has other special
+ *			meaning in ParseTokens: backslash, dollar sign, or
+ *			open bracket.
+ * TYPE_QUOTE -		Character is a double quote.
+ * TYPE_CLOSE_PAREN -	Character is a right parenthesis.
+ * TYPE_CLOSE_BRACK -	Character is a right square bracket.
+ * TYPE_BRACE -		Character is a curly brace (either left or right).
+ */
+
+const char tclCharTypeTable[] = {
+    /*
+     * Negative character values, from -128 to -1:
+     */
+
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+
+    /*
+     * Positive character values, from 0-127:
+     */
+
+    TYPE_SUBS,        TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_SPACE,       TYPE_COMMAND_END, TYPE_SPACE,
+    TYPE_SPACE,       TYPE_SPACE,       TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_SPACE,       TYPE_NORMAL,      TYPE_QUOTE,       TYPE_NORMAL,
+    TYPE_SUBS,        TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_CLOSE_PAREN, TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_COMMAND_END,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_SUBS,
+    TYPE_SUBS,        TYPE_CLOSE_BRACK, TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_BRACE,
+    TYPE_NORMAL,      TYPE_BRACE,       TYPE_NORMAL,      TYPE_NORMAL,
+
+    /*
+     * Large unsigned character values, from 128-255:
+     */
+
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+    TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,      TYPE_NORMAL,
+};
+
+/*
+ * Prototypes for local functions defined in this file:
+ */
+
+static inline int	CommandComplete(const char *script, int numBytes);
+static int		ParseComment(const char *src, int numBytes,
+			    Tcl_Parse *parsePtr);
+static int		ParseTokens(const char *src, int numBytes, int mask,
+			    int flags, Tcl_Parse *parsePtr);
+static int		ParseWhiteSpace(const char *src, int numBytes,
+			    int *incompletePtr, char *typePtr);
+static int		ParseAllWhiteSpace(const char *src, int numBytes,
+			    int *incompletePtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclParseInit --
+ *
+ *	Initialize the fields of a Tcl_Parse struct.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The Tcl_Parse struct pointed to by parsePtr gets initialized.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclParseInit(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting */
+    const char *start,		/* Start of string to be parsed. */
+    int numBytes,		/* Total number of bytes in string. If < 0,
+				 * the script consists of all bytes up to the
+				 * first null character. */
+    Tcl_Parse *parsePtr)	/* Points to struct to initialize */
+{
+    parsePtr->numWords = 0;
+    parsePtr->tokenPtr = parsePtr->staticTokens;
+    parsePtr->numTokens = 0;
+    parsePtr->tokensAvailable = NUM_STATIC_TOKENS;
+    parsePtr->string = start;
+    parsePtr->end = start + numBytes;
+    parsePtr->term = parsePtr->end;
+    parsePtr->interp = interp;
+    parsePtr->incomplete = 0;
+    parsePtr->errorType = TCL_PARSE_SUCCESS;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseCommand --
+ *
+ *	Given a string, this function parses the first Tcl command in the
+ *	string and returns information about the structure of the command.
+ *
+ * Results:
+ *	The return value is TCL_OK if the command was parsed successfully and
+ *	TCL_ERROR otherwise. If an error occurs and interp isn't NULL then an
+ *	error message is left in its result. On a successful return, parsePtr
+ *	is filled in with information about the command that was parsed.
+ *
+ * Side effects:
+ *	If there is insufficient space in parsePtr to hold all the information
+ *	about the command, then additional space is malloc-ed. If the function
+ *	returns TCL_OK then the caller must eventually invoke Tcl_FreeParse to
+ *	release any additional space that was allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseCommand(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting; if
+				 * NULL, then no error message is provided. */
+    const char *start,		/* First character of string containing one or
+				 * more Tcl commands. */
+    register int numBytes,	/* Total number of bytes in string. If < 0,
+				 * the script consists of all bytes up to the
+				 * first null character. */
+    int nested,			/* Non-zero means this is a nested command:
+				 * close bracket should be considered a
+				 * command terminator. If zero, then close
+				 * bracket has no special meaning. */
+    register Tcl_Parse *parsePtr)
+				/* Structure to fill in with information about
+				 * the parsed command; any previous
+				 * information in the structure is ignored. */
+{
+    register const char *src;	/* Points to current character in the
+				 * command. */
+    char type;			/* Result returned by CHAR_TYPE(*src). */
+    Tcl_Token *tokenPtr;	/* Pointer to token being filled in. */
+    int wordIndex;		/* Index of word token for current word. */
+    int terminators;		/* CHAR_TYPE bits that indicate the end of a
+				 * command. */
+    const char *termPtr;	/* Set by Tcl_ParseBraces/QuotedString to
+				 * point to char after terminating one. */
+    int scanned;
+
+    if ((start == NULL) && (numBytes != 0)) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "can't parse a NULL pointer", -1));
+	}
+	return TCL_ERROR;
+    }
+    if (numBytes < 0) {
+	numBytes = strlen(start);
+    }
+    TclParseInit(interp, start, numBytes, parsePtr);
+    parsePtr->commentStart = NULL;
+    parsePtr->commentSize = 0;
+    parsePtr->commandStart = NULL;
+    parsePtr->commandSize = 0;
+    if (nested != 0) {
+	terminators = TYPE_COMMAND_END | TYPE_CLOSE_BRACK;
+    } else {
+	terminators = TYPE_COMMAND_END;
+    }
+
+    /*
+     * Parse any leading space and comments before the first word of the
+     * command.
+     */
+
+    scanned = ParseComment(start, numBytes, parsePtr);
+    src = (start + scanned);
+    numBytes -= scanned;
+    if (numBytes == 0) {
+	if (nested) {
+	    parsePtr->incomplete = nested;
+	}
+    }
+
+    /*
+     * The following loop parses the words of the command, one word in each
+     * iteration through the loop.
+     */
+
+    parsePtr->commandStart = src;
+    type = CHAR_TYPE(*src);
+    scanned = 1;	/* Can't have missing whitepsace before first word. */
+    while (1) {
+	int expandWord = 0;
+
+	/* Are we at command termination? */
+
+	if ((numBytes == 0) || (type & terminators) != 0) {
+	    parsePtr->term = src;
+	    parsePtr->commandSize = src + (numBytes != 0)
+		    - parsePtr->commandStart;
+	    return TCL_OK;
+	}
+
+	/* Are we missing white space after previous word? */
+
+	if (scanned == 0) {
+	    if (src[-1] == '"') {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "extra characters after close-quote", -1));
+		}
+		parsePtr->errorType = TCL_PARSE_QUOTE_EXTRA;
+	    } else {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "extra characters after close-brace", -1));
+		}
+		parsePtr->errorType = TCL_PARSE_BRACE_EXTRA;
+	    }
+	    parsePtr->term = src;
+	error:
+	    Tcl_FreeParse(parsePtr);
+	    parsePtr->commandSize = parsePtr->end - parsePtr->commandStart;
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Create the token for the word.
+	 */
+
+	TclGrowParseTokenArray(parsePtr, 1);
+	wordIndex = parsePtr->numTokens;
+	tokenPtr = &parsePtr->tokenPtr[wordIndex];
+	tokenPtr->type = TCL_TOKEN_WORD;
+
+	tokenPtr->start = src;
+	parsePtr->numTokens++;
+	parsePtr->numWords++;
+
+	/*
+	 * At this point the word can have one of four forms: something
+	 * enclosed in quotes, something enclosed in braces, and expanding
+	 * word, or an unquoted word (anything else).
+	 */
+
+    parseWord:
+	if (*src == '"') {
+	    if (Tcl_ParseQuotedString(interp, src, numBytes, parsePtr, 1,
+		    &termPtr) != TCL_OK) {
+		goto error;
+	    }
+	    src = termPtr;
+	    numBytes = parsePtr->end - src;
+	} else if (*src == '{') {
+	    int expIdx = wordIndex + 1;
+	    Tcl_Token *expPtr;
+
+	    if (Tcl_ParseBraces(interp, src, numBytes, parsePtr, 1,
+		    &termPtr) != TCL_OK) {
+		goto error;
+	    }
+	    src = termPtr;
+	    numBytes = parsePtr->end - src;
+
+	    /*
+	     * Check whether the braces contained the word expansion prefix
+	     * {*}
+	     */
+
+	    expPtr = &parsePtr->tokenPtr[expIdx];
+	    if ((0 == expandWord)
+		    /* Haven't seen prefix already */
+		    && (1 == parsePtr->numTokens - expIdx)
+		    /* Only one token */
+		    && (((1 == (size_t) expPtr->size)
+			    /* Same length as prefix */
+			    && (expPtr->start[0] == '*')))
+			    /* Is the prefix */
+		    && (numBytes > 0) && (0 == ParseWhiteSpace(termPtr,
+			    numBytes, &parsePtr->incomplete, &type))
+		    && (type != TYPE_COMMAND_END)
+		    /* Non-whitespace follows */) {
+		expandWord = 1;
+		parsePtr->numTokens--;
+		goto parseWord;
+	    }
+	} else {
+	    /*
+	     * This is an unquoted word. Call ParseTokens and let it do all of
+	     * the work.
+	     */
+
+	    if (ParseTokens(src, numBytes, TYPE_SPACE|terminators,
+		    TCL_SUBST_ALL, parsePtr) != TCL_OK) {
+		goto error;
+	    }
+	    src = parsePtr->term;
+	    numBytes = parsePtr->end - src;
+	}
+
+	/*
+	 * Finish filling in the token for the word and check for the special
+	 * case of a word consisting of a single range of literal text.
+	 */
+
+	tokenPtr = &parsePtr->tokenPtr[wordIndex];
+	tokenPtr->size = src - tokenPtr->start;
+	tokenPtr->numComponents = parsePtr->numTokens - (wordIndex + 1);
+	if (expandWord) {
+	    int i, isLiteral = 1;
+
+	    /*
+	     * When a command includes a word that is an expanded literal; for
+	     * example, {*}{1 2 3}, the parser performs that expansion
+	     * immediately, generating several TCL_TOKEN_SIMPLE_WORDs instead
+	     * of a single TCL_TOKEN_EXPAND_WORD that the Tcl_ParseCommand()
+	     * caller might have to expand. This notably makes it simpler for
+	     * those callers that wish to track line endings, such as those
+	     * that implement key parts of TIP 280.
+	     *
+	     * First check whether the thing to be expanded is a literal,
+	     * in the sense of being composed entirely of TCL_TOKEN_TEXT
+	     * tokens.
+	     */
+
+	    for (i = 1; i <= tokenPtr->numComponents; i++) {
+		if (tokenPtr[i].type != TCL_TOKEN_TEXT) {
+		    isLiteral = 0;
+		    break;
+		}
+	    }
+
+	    if (isLiteral) {
+		int elemCount = 0, code = TCL_OK, literal = 1;
+		const char *nextElem, *listEnd, *elemStart;
+
+		/*
+		 * The word to be expanded is a literal, so determine the
+		 * boundaries of the literal string to be treated as a list
+		 * and expanded. That literal string starts at
+		 * tokenPtr[1].start, and includes all bytes up to, but not
+		 * including (tokenPtr[tokenPtr->numComponents].start +
+		 * tokenPtr[tokenPtr->numComponents].size)
+		 */
+
+		listEnd = (tokenPtr[tokenPtr->numComponents].start +
+			tokenPtr[tokenPtr->numComponents].size);
+		nextElem = tokenPtr[1].start;
+
+		/*
+		 * Step through the literal string, parsing and counting list
+		 * elements.
+		 */
+
+		while (nextElem < listEnd) {
+		    int size;
+
+		    code = TclFindElement(NULL, nextElem, listEnd - nextElem,
+			    &elemStart, &nextElem, &size, &literal);
+		    if ((code != TCL_OK) || !literal) {
+			break;
+		    }
+		    if (elemStart < listEnd) {
+			elemCount++;
+		    }
+		}
+
+		if ((code != TCL_OK) || !literal) {
+		    /*
+		     * Some list element could not be parsed, or is not
+		     * present as a literal substring of the script.  The
+		     * compiler cannot handle list elements that get generated
+		     * by a call to TclCopyAndCollapse(). Defer  the
+		     * handling of  this to  compile/eval time, where  code is
+		     * already  in place to  report the  "attempt to  expand a
+		     * non-list" error or expand lists that require
+		     * substitution.
+		     */
+
+		    tokenPtr->type = TCL_TOKEN_EXPAND_WORD;
+		} else if (elemCount == 0) {
+		    /*
+		     * We are expanding a literal empty list. This means that
+		     * the expanding word completely disappears, leaving no
+		     * word generated this pass through the loop. Adjust
+		     * accounting appropriately.
+		     */
+
+		    parsePtr->numWords--;
+		    parsePtr->numTokens = wordIndex;
+		} else {
+		    /*
+		     * Recalculate the number of Tcl_Tokens needed to store
+		     * tokens representing the expanded list.
+		     */
+
+		    const char *listStart;
+		    int growthNeeded = wordIndex + 2*elemCount
+			    - parsePtr->numTokens;
+
+		    parsePtr->numWords += elemCount - 1;
+		    if (growthNeeded > 0) {
+			TclGrowParseTokenArray(parsePtr, growthNeeded);
+			tokenPtr = &parsePtr->tokenPtr[wordIndex];
+		    }
+		    parsePtr->numTokens = wordIndex + 2*elemCount;
+
+		    /*
+		     * Generate a TCL_TOKEN_SIMPLE_WORD token sequence for
+		     * each element of the literal list we are expanding in
+		     * place. Take care with the start and size fields of each
+		     * token so they point to the right literal characters in
+		     * the original script to represent the right expanded
+		     * word value.
+		     */
+
+		    listStart = nextElem = tokenPtr[1].start;
+		    while (nextElem < listEnd) {
+			int quoted;
+
+			tokenPtr->type = TCL_TOKEN_SIMPLE_WORD;
+			tokenPtr->numComponents = 1;
+
+			tokenPtr++;
+			tokenPtr->type = TCL_TOKEN_TEXT;
+			tokenPtr->numComponents = 0;
+			TclFindElement(NULL, nextElem, listEnd - nextElem,
+				&(tokenPtr->start), &nextElem,
+				&(tokenPtr->size), NULL);
+
+			quoted = (tokenPtr->start[-1] == '{'
+				|| tokenPtr->start[-1] == '"')
+				&& tokenPtr->start > listStart;
+			tokenPtr[-1].start = tokenPtr->start - quoted;
+			tokenPtr[-1].size = tokenPtr->start + tokenPtr->size
+				- tokenPtr[-1].start + quoted;
+
+			tokenPtr++;
+		    }
+		}
+	    } else {
+		/*
+		 * The word to be expanded is not a literal, so defer
+		 * expansion to compile/eval time by marking with a
+		 * TCL_TOKEN_EXPAND_WORD token.
+		 */
+
+		tokenPtr->type = TCL_TOKEN_EXPAND_WORD;
+	    }
+	} else if ((tokenPtr->numComponents == 1)
+		&& (tokenPtr[1].type == TCL_TOKEN_TEXT)) {
+	    tokenPtr->type = TCL_TOKEN_SIMPLE_WORD;
+	}
+
+	/* Parse the whitespace between words. */
+
+	scanned = ParseWhiteSpace(src,numBytes, &parsePtr->incomplete, &type);
+	src += scanned;
+	numBytes -= scanned;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclIsSpaceProc --
+ *
+ *	Report whether byte is in the set of whitespace characters used by
+ *	Tcl to separate words in scripts or elements in lists.
+ *
+ * Results:
+ *	Returns 1, if byte is in the set, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclIsSpaceProc(
+    char byte)
+{
+    return CHAR_TYPE(byte) & (TYPE_SPACE) || byte == '\n';
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclIsBareword--
+ *
+ *	Report whether byte is one that can be part of a "bareword".
+ *	This concept is named in expression parsing, where it determines
+ *	what can be a legal function name, but is the same definition used
+ *	in determining what variable names can be parsed as variable
+ *	substitutions without the benefit of enclosing braces.  The set of
+ *	ASCII chars that are accepted are the numeric chars ('0'-'9'),
+ *	the alphabetic chars ('a'-'z', 'A'-'Z')	and underscore ('_').
+ *
+ * Results:
+ *	Returns 1, if byte is in the accepted set of chars, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclIsBareword(
+    char byte)
+{
+    if (byte < '0' || byte > 'z') {
+	return 0;
+    }
+    if (byte <= '9' || byte >= 'a') {
+	return 1;
+    }
+    if (byte == '_') {
+	return 1;
+    }
+    if (byte < 'A' || byte > 'Z') {
+	return 0;
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseWhiteSpace --
+ *
+ *	Scans up to numBytes bytes starting at src, consuming white space
+ *	between words as defined by Tcl's parsing rules.
+ *
+ * Results:
+ *	Returns the number of bytes recognized as white space. Records at
+ *	parsePtr, information about the parse. Records at typePtr the
+ *	character type of the non-whitespace character that terminated the
+ *	scan.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseWhiteSpace(
+    const char *src,		/* First character to parse. */
+    register int numBytes,	/* Max number of bytes to scan. */
+    int *incompletePtr,		/* Set this boolean memory to true if parsing
+				 * indicates an incomplete command. */
+    char *typePtr)		/* Points to location to store character type
+				 * of character that ends run of whitespace */
+{
+    register char type = TYPE_NORMAL;
+    register const char *p = src;
+
+    while (1) {
+	while (numBytes && ((type = CHAR_TYPE(*p)) & TYPE_SPACE)) {
+	    numBytes--;
+	    p++;
+	}
+	if (numBytes && (type & TYPE_SUBS)) {
+	    if (*p != '\\') {
+		break;
+	    }
+	    if (--numBytes == 0) {
+		break;
+	    }
+	    if (p[1] != '\n') {
+		break;
+	    }
+	    p += 2;
+	    if (--numBytes == 0) {
+		*incompletePtr = 1;
+		break;
+	    }
+	    continue;
+	}
+	break;
+    }
+    *typePtr = type;
+    return (p - src);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclParseAllWhiteSpace --
+ *
+ *	Scans up to numBytes bytes starting at src, consuming all white space
+ *	including the command-terminating newline characters.
+ *
+ * Results:
+ *	Returns the number of bytes recognized as white space.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseAllWhiteSpace(
+    const char *src,		/* First character to parse. */
+    int numBytes,		/* Max number of byes to scan */
+    int *incompletePtr)		/* Set true if parse is incomplete. */
+{
+    char type;
+    const char *p = src;
+
+    do {
+	int scanned = ParseWhiteSpace(p, numBytes, incompletePtr, &type);
+
+	p += scanned;
+	numBytes -= scanned;
+    } while (numBytes && (*p == '\n') && (p++, --numBytes));
+    return (p-src);
+}
+
+int
+TclParseAllWhiteSpace(
+    const char *src,		/* First character to parse. */
+    int numBytes)		/* Max number of byes to scan */
+{
+    int dummy;
+    return ParseAllWhiteSpace(src, numBytes, &dummy);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclParseHex --
+ *
+ *	Scans a hexadecimal number as a Tcl_UniChar value (e.g., for parsing
+ *	\x and \u escape sequences). At most numBytes bytes are scanned.
+ *
+ * Results:
+ *	The numeric value is stored in *resultPtr. Returns the number of bytes
+ *	consumed.
+ *
+ * Notes:
+ *	Relies on the following properties of the ASCII character set, with
+ *	which UTF-8 is compatible:
+ *
+ *	The digits '0' .. '9' and the letters 'A' .. 'Z' and 'a' .. 'z' occupy
+ *	consecutive code points, and '0' < 'A' < 'a'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclParseHex(
+    const char *src,		/* First character to parse. */
+    int numBytes,		/* Max number of byes to scan */
+    int *resultPtr)	/* Points to storage provided by caller where
+				 * the character resulting from the
+				 * conversion is to be written. */
+{
+    int result = 0;
+    register const char *p = src;
+
+    while (numBytes--) {
+	unsigned char digit = UCHAR(*p);
+
+	if (!isxdigit(digit) || (result > 0x10fff)) {
+	    break;
+	}
+
+	p++;
+	result <<= 4;
+
+	if (digit >= 'a') {
+	    result |= (10 + digit - 'a');
+	} else if (digit >= 'A') {
+	    result |= (10 + digit - 'A');
+	} else {
+	    result |= (digit - '0');
+	}
+    }
+
+    *resultPtr = result;
+    return (p - src);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclParseBackslash --
+ *
+ *	Scans up to numBytes bytes starting at src, consuming a backslash
+ *	sequence as defined by Tcl's parsing rules.
+ *
+ * Results:
+ *	Records at readPtr the number of bytes making up the backslash
+ *	sequence. Records at dst the UTF-8 encoded equivalent of that
+ *	backslash sequence. Returns the number of bytes written to dst, at
+ *	most TCL_UTF_MAX. Either readPtr or dst may be NULL, if the results
+ *	are not needed, but the return value is the same either way.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclParseBackslash(
+    const char *src,		/* Points to the backslash character of a a
+				 * backslash sequence. */
+    int numBytes,		/* Max number of bytes to scan. */
+    int *readPtr,		/* NULL, or points to storage where the number
+				 * of bytes scanned should be written. */
+    char *dst)			/* NULL, or points to buffer where the UTF-8
+				 * encoding of the backslash sequence is to be
+				 * written. At most TCL_UTF_MAX bytes will be
+				 * written there. */
+{
+    register const char *p = src+1;
+    Tcl_UniChar unichar = 0;
+    int result;
+    int count;
+    char buf[TCL_UTF_MAX];
+
+    if (numBytes == 0) {
+	if (readPtr != NULL) {
+	    *readPtr = 0;
+	}
+	return 0;
+    }
+
+    if (dst == NULL) {
+	dst = buf;
+    }
+
+    if (numBytes == 1) {
+	/*
+	 * Can only scan the backslash, so return it.
+	 */
+
+	result = '\\';
+	count = 1;
+	goto done;
+    }
+
+    count = 2;
+    switch (*p) {
+	/*
+	 * Note: in the conversions below, use absolute values (e.g., 0xa)
+	 * rather than symbolic values (e.g. \n) that get converted by the
+	 * compiler. It's possible that compilers on some platforms will do
+	 * the symbolic conversions differently, which could result in
+	 * non-portable Tcl scripts.
+	 */
+
+    case 'a':
+	result = 0x7;
+	break;
+    case 'b':
+	result = 0x8;
+	break;
+    case 'f':
+	result = 0xc;
+	break;
+    case 'n':
+	result = 0xa;
+	break;
+    case 'r':
+	result = 0xd;
+	break;
+    case 't':
+	result = 0x9;
+	break;
+    case 'v':
+	result = 0xb;
+	break;
+    case 'x':
+	count += TclParseHex(p+1, (numBytes > 3) ? 2 : numBytes-2, &result);
+	if (count == 2) {
+	    /*
+	     * No hexadigits -> This is just "x".
+	     */
+
+	    result = 'x';
+	} else {
+	    /*
+	     * Keep only the last byte (2 hex digits).
+	     */
+	    result = (unsigned char) result;
+	}
+	break;
+    case 'u':
+	count += TclParseHex(p+1, (numBytes > 5) ? 4 : numBytes-2, &result);
+	if (count == 2) {
+	    /*
+	     * No hexadigits -> This is just "u".
+	     */
+	    result = 'u';
+	}
+	break;
+    case 'U':
+	count += TclParseHex(p+1, (numBytes > 9) ? 8 : numBytes-2, &result);
+	if (count == 2) {
+	    /*
+	     * No hexadigits -> This is just "U".
+	     */
+	    result = 'U';
+	}
+	break;
+    case '\n':
+	count--;
+	do {
+	    p++;
+	    count++;
+	} while ((count < numBytes) && ((*p == ' ') || (*p == '\t')));
+	result = ' ';
+	break;
+    case 0:
+	result = '\\';
+	count = 1;
+	break;
+    default:
+	/*
+	 * Check for an octal number \oo?o?
+	 */
+
+	if (isdigit(UCHAR(*p)) && (UCHAR(*p) < '8')) {	/* INTL: digit */
+	    result = *p - '0';
+	    p++;
+	    if ((numBytes == 2) || !isdigit(UCHAR(*p))	/* INTL: digit */
+		    || (UCHAR(*p) >= '8')) {
+		break;
+	    }
+	    count = 3;
+	    result = (result << 3) + (*p - '0');
+	    p++;
+	    if ((numBytes == 3) || !isdigit(UCHAR(*p))	/* INTL: digit */
+		    || (UCHAR(*p) >= '8') || (result >= 0x20)) {
+		break;
+	    }
+	    count = 4;
+	    result = UCHAR((result << 3) + (*p - '0'));
+	    break;
+	}
+
+	/*
+	 * We have to convert here in case the user has put a backslash in
+	 * front of a multi-byte utf-8 character. While this means nothing
+	 * special, we shouldn't break up a correct utf-8 character. [Bug
+	 * #217987] test subst-3.2
+	 */
+
+	if (Tcl_UtfCharComplete(p, numBytes - 1)) {
+	    count = TclUtfToUniChar(p, &unichar) + 1;	/* +1 for '\' */
+	} else {
+	    char utfBytes[TCL_UTF_MAX];
+
+	    memcpy(utfBytes, p, (size_t) (numBytes - 1));
+	    utfBytes[numBytes - 1] = '\0';
+	    count = TclUtfToUniChar(utfBytes, &unichar) + 1;
+	}
+	result = unichar;
+	break;
+    }
+
+  done:
+    if (readPtr != NULL) {
+	*readPtr = count;
+    }
+    return Tcl_UniCharToUtf(result, dst);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseComment --
+ *
+ *	Scans up to numBytes bytes starting at src, consuming a Tcl comment as
+ *	defined by Tcl's parsing rules.
+ *
+ * Results:
+ *	Records in parsePtr information about the parse. Returns the number of
+ *	bytes consumed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseComment(
+    const char *src,		/* First character to parse. */
+    register int numBytes,	/* Max number of bytes to scan. */
+    Tcl_Parse *parsePtr)	/* Information about parse in progress.
+				 * Updated if parsing indicates an incomplete
+				 * command. */
+{
+    register const char *p = src;
+    int incomplete = parsePtr->incomplete;
+
+    while (numBytes) {
+	int scanned = ParseAllWhiteSpace(p, numBytes, &incomplete);
+	p += scanned;
+	numBytes -= scanned;
+
+	if ((numBytes == 0) || (*p != '#')) {
+	    break;
+	}
+	if (parsePtr->commentStart == NULL) {
+	    parsePtr->commentStart = p;
+	}
+
+	p++;
+	numBytes--;
+	while (numBytes) {
+	    if (*p == '\n') {
+		p++;
+		numBytes--;
+		break;
+	    }
+	    if (*p == '\\') {
+		p++;
+		numBytes--;
+		if (numBytes == 0) {
+		    break;
+		}
+	    }
+	    incomplete = (*p == '\n');
+	    p++;
+	    numBytes--;
+	}
+	parsePtr->commentSize = p - parsePtr->commentStart;
+    }
+    parsePtr->incomplete = incomplete;
+    return (p - src);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseTokens --
+ *
+ *	This function forms the heart of the Tcl parser. It parses one or more
+ *	tokens from a string, up to a termination point specified by the
+ *	caller. This function is used to parse unquoted command words (those
+ *	not in quotes or braces), words in quotes, and array indices for
+ *	variables. No more than numBytes bytes will be scanned.
+ *
+ * Results:
+ *	Tokens are added to parsePtr and parsePtr->term is filled in with the
+ *	address of the character that terminated the parse (the first one
+ *	whose CHAR_TYPE matched mask or the character at parsePtr->end). The
+ *	return value is TCL_OK if the parse completed successfully and
+ *	TCL_ERROR otherwise. If a parse error occurs and parsePtr->interp is
+ *	not NULL, then an error message is left in the interpreter's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ParseTokens(
+    register const char *src,	/* First character to parse. */
+    register int numBytes,	/* Max number of bytes to scan. */
+    int mask,			/* Specifies when to stop parsing. The parse
+				 * stops at the first unquoted character whose
+				 * CHAR_TYPE contains any of the bits in
+				 * mask. */
+    int flags,			/* OR-ed bits indicating what substitutions to
+				 * perform: TCL_SUBST_COMMANDS,
+				 * TCL_SUBST_VARIABLES, and
+				 * TCL_SUBST_BACKSLASHES */
+    Tcl_Parse *parsePtr)	/* Information about parse in progress.
+				 * Updated with additional tokens and
+				 * termination information. */
+{
+    char type;
+    int originalTokens;
+    int noSubstCmds = !(flags & TCL_SUBST_COMMANDS);
+    int noSubstVars = !(flags & TCL_SUBST_VARIABLES);
+    int noSubstBS = !(flags & TCL_SUBST_BACKSLASHES);
+    Tcl_Token *tokenPtr;
+
+    /*
+     * Each iteration through the following loop adds one token of type
+     * TCL_TOKEN_TEXT, TCL_TOKEN_BS, TCL_TOKEN_COMMAND, or TCL_TOKEN_VARIABLE
+     * to parsePtr. For TCL_TOKEN_VARIABLE tokens, additional tokens are added
+     * for the parsed variable name.
+     */
+
+    originalTokens = parsePtr->numTokens;
+    while (numBytes && !((type = CHAR_TYPE(*src)) & mask)) {
+	TclGrowParseTokenArray(parsePtr, 1);
+	tokenPtr = &parsePtr->tokenPtr[parsePtr->numTokens];
+	tokenPtr->start = src;
+	tokenPtr->numComponents = 0;
+
+	if ((type & TYPE_SUBS) == 0) {
+	    /*
+	     * This is a simple range of characters. Scan to find the end of
+	     * the range.
+	     */
+
+	    while ((++src, --numBytes)
+		    && !(CHAR_TYPE(*src) & (mask | TYPE_SUBS))) {
+		/* empty loop */
+	    }
+	    tokenPtr->type = TCL_TOKEN_TEXT;
+	    tokenPtr->size = src - tokenPtr->start;
+	    parsePtr->numTokens++;
+	} else if (*src == '$') {
+	    int varToken;
+
+	    if (noSubstVars) {
+		tokenPtr->type = TCL_TOKEN_TEXT;
+		tokenPtr->size = 1;
+		parsePtr->numTokens++;
+		src++;
+		numBytes--;
+		continue;
+	    }
+
+	    /*
+	     * This is a variable reference. Call Tcl_ParseVarName to do all
+	     * the dirty work of parsing the name.
+	     */
+
+	    varToken = parsePtr->numTokens;
+	    if (Tcl_ParseVarName(parsePtr->interp, src, numBytes, parsePtr,
+		    1) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    src += parsePtr->tokenPtr[varToken].size;
+	    numBytes -= parsePtr->tokenPtr[varToken].size;
+	} else if (*src == '[') {
+	    Tcl_Parse *nestedPtr;
+
+	    if (noSubstCmds) {
+		tokenPtr->type = TCL_TOKEN_TEXT;
+		tokenPtr->size = 1;
+		parsePtr->numTokens++;
+		src++;
+		numBytes--;
+		continue;
+	    }
+
+	    /*
+	     * Command substitution. Call Tcl_ParseCommand recursively (and
+	     * repeatedly) to parse the nested command(s), then throw away the
+	     * parse information.
+	     */
+
+	    src++;
+	    numBytes--;
+	    nestedPtr = TclStackAlloc(parsePtr->interp, sizeof(Tcl_Parse));
+	    while (1) {
+		const char *curEnd;
+
+		if (Tcl_ParseCommand(parsePtr->interp, src, numBytes, 1,
+			nestedPtr) != TCL_OK) {
+		    parsePtr->errorType = nestedPtr->errorType;
+		    parsePtr->term = nestedPtr->term;
+		    parsePtr->incomplete = nestedPtr->incomplete;
+		    TclStackFree(parsePtr->interp, nestedPtr);
+		    return TCL_ERROR;
+		}
+		curEnd = src + numBytes;
+		src = nestedPtr->commandStart + nestedPtr->commandSize;
+		numBytes = curEnd - src;
+		Tcl_FreeParse(nestedPtr);
+
+		/*
+		 * Check for the closing ']' that ends the command
+		 * substitution. It must have been the last character of the
+		 * parsed command.
+		 */
+
+		if ((nestedPtr->term < parsePtr->end)
+			&& (*(nestedPtr->term) == ']')
+			&& !(nestedPtr->incomplete)) {
+		    break;
+		}
+		if (numBytes == 0) {
+		    if (parsePtr->interp != NULL) {
+			Tcl_SetObjResult(parsePtr->interp, Tcl_NewStringObj(
+				"missing close-bracket", -1));
+		    }
+		    parsePtr->errorType = TCL_PARSE_MISSING_BRACKET;
+		    parsePtr->term = tokenPtr->start;
+		    parsePtr->incomplete = 1;
+		    TclStackFree(parsePtr->interp, nestedPtr);
+		    return TCL_ERROR;
+		}
+	    }
+	    TclStackFree(parsePtr->interp, nestedPtr);
+	    tokenPtr->type = TCL_TOKEN_COMMAND;
+	    tokenPtr->size = src - tokenPtr->start;
+	    parsePtr->numTokens++;
+	} else if (*src == '\\') {
+	    if (noSubstBS) {
+		tokenPtr->type = TCL_TOKEN_TEXT;
+		tokenPtr->size = 1;
+		parsePtr->numTokens++;
+		src++;
+		numBytes--;
+		continue;
+	    }
+
+	    /*
+	     * Backslash substitution.
+	     */
+
+	    TclParseBackslash(src, numBytes, &tokenPtr->size, NULL);
+
+	    if (tokenPtr->size == 1) {
+		/*
+		 * Just a backslash, due to end of string.
+		 */
+
+		tokenPtr->type = TCL_TOKEN_TEXT;
+		parsePtr->numTokens++;
+		src++;
+		numBytes--;
+		continue;
+	    }
+
+	    if (src[1] == '\n') {
+		if (numBytes == 2) {
+		    parsePtr->incomplete = 1;
+		}
+
+		/*
+		 * Note: backslash-newline is special in that it is treated
+		 * the same as a space character would be. This means that it
+		 * could terminate the token.
+		 */
+
+		if (mask & TYPE_SPACE) {
+		    if (parsePtr->numTokens == originalTokens) {
+			goto finishToken;
+		    }
+		    break;
+		}
+	    }
+
+	    tokenPtr->type = TCL_TOKEN_BS;
+	    parsePtr->numTokens++;
+	    src += tokenPtr->size;
+	    numBytes -= tokenPtr->size;
+	} else if (*src == 0) {
+	    tokenPtr->type = TCL_TOKEN_TEXT;
+	    tokenPtr->size = 1;
+	    parsePtr->numTokens++;
+	    src++;
+	    numBytes--;
+	} else {
+	    Tcl_Panic("ParseTokens encountered unknown character");
+	}
+    }
+    if (parsePtr->numTokens == originalTokens) {
+	/*
+	 * There was nothing in this range of text. Add an empty token for the
+	 * empty range, so that there is always at least one token added.
+	 */
+
+	TclGrowParseTokenArray(parsePtr, 1);
+	tokenPtr = &parsePtr->tokenPtr[parsePtr->numTokens];
+	tokenPtr->start = src;
+	tokenPtr->numComponents = 0;
+
+    finishToken:
+	tokenPtr->type = TCL_TOKEN_TEXT;
+	tokenPtr->size = 0;
+	parsePtr->numTokens++;
+    }
+    parsePtr->term = src;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FreeParse --
+ *
+ *	This function is invoked to free any dynamic storage that may have
+ *	been allocated by a previous call to Tcl_ParseCommand.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there is any dynamically allocated memory in *parsePtr, it is
+ *	freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_FreeParse(
+    Tcl_Parse *parsePtr)	/* Structure that was filled in by a previous
+				 * call to Tcl_ParseCommand. */
+{
+    if (parsePtr->tokenPtr != parsePtr->staticTokens) {
+	ckfree(parsePtr->tokenPtr);
+	parsePtr->tokenPtr = parsePtr->staticTokens;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseVarName --
+ *
+ *	Given a string starting with a $ sign, parse off a variable name and
+ *	return information about the parse. No more than numBytes bytes will
+ *	be scanned.
+ *
+ * Results:
+ *	The return value is TCL_OK if the command was parsed successfully and
+ *	TCL_ERROR otherwise. If an error occurs and interp isn't NULL then an
+ *	error message is left in its result. On a successful return, tokenPtr
+ *	and numTokens fields of parsePtr are filled in with information about
+ *	the variable name that was parsed. The "size" field of the first new
+ *	token gives the total number of bytes in the variable name. Other
+ *	fields in parsePtr are undefined.
+ *
+ * Side effects:
+ *	If there is insufficient space in parsePtr to hold all the information
+ *	about the command, then additional space is malloc-ed. If the function
+ *	returns TCL_OK then the caller must eventually invoke Tcl_FreeParse to
+ *	release any additional space that was allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseVarName(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting; if
+				 * NULL, then no error message is provided. */
+    const char *start,		/* Start of variable substitution string.
+				 * First character must be "$". */
+    register int numBytes,	/* Total number of bytes in string. If < 0,
+				 * the string consists of all bytes up to the
+				 * first null character. */
+    Tcl_Parse *parsePtr,	/* Structure to fill in with information about
+				 * the variable name. */
+    int append)			/* Non-zero means append tokens to existing
+				 * information in parsePtr; zero means ignore
+				 * existing tokens in parsePtr and
+				 * reinitialize it. */
+{
+    Tcl_Token *tokenPtr;
+    register const char *src;
+    int varIndex;
+    unsigned array;
+
+    if ((numBytes == 0) || (start == NULL)) {
+	return TCL_ERROR;
+    }
+    if (numBytes < 0) {
+	numBytes = strlen(start);
+    }
+
+    if (!append) {
+	TclParseInit(interp, start, numBytes, parsePtr);
+    }
+
+    /*
+     * Generate one token for the variable, an additional token for the name,
+     * plus any number of additional tokens for the index, if there is one.
+     */
+
+    src = start;
+    TclGrowParseTokenArray(parsePtr, 2);
+    tokenPtr = &parsePtr->tokenPtr[parsePtr->numTokens];
+    tokenPtr->type = TCL_TOKEN_VARIABLE;
+    tokenPtr->start = src;
+    varIndex = parsePtr->numTokens;
+    parsePtr->numTokens++;
+    tokenPtr++;
+    src++;
+    numBytes--;
+    if (numBytes == 0) {
+	goto justADollarSign;
+    }
+    tokenPtr->type = TCL_TOKEN_TEXT;
+    tokenPtr->start = src;
+    tokenPtr->numComponents = 0;
+
+    /*
+     * The name of the variable can have three forms:
+     * 1. The $ sign is followed by an open curly brace. Then the variable
+     *	  name is everything up to the next close curly brace, and the
+     *	  variable is a scalar variable.
+     * 2. The $ sign is not followed by an open curly brace. Then the variable
+     *	  name is everything up to the next character that isn't a letter,
+     *	  digit, or underscore. :: sequences are also considered part of the
+     *	  variable name, in order to support namespaces. If the following
+     *	  character is an open parenthesis, then the information between
+     *	  parentheses is the array element name.
+     * 3. The $ sign is followed by something that isn't a letter, digit, or
+     *	  underscore: in this case, there is no variable name and the token is
+     *	  just "$".
+     */
+
+    if (*src == '{') {
+	src++;
+	numBytes--;
+	tokenPtr->type = TCL_TOKEN_TEXT;
+	tokenPtr->start = src;
+	tokenPtr->numComponents = 0;
+
+	while (numBytes && (*src != '}')) {
+	    numBytes--;
+	    src++;
+	}
+	if (numBytes == 0) {
+	    if (parsePtr->interp != NULL) {
+		Tcl_SetObjResult(parsePtr->interp, Tcl_NewStringObj(
+			"missing close-brace for variable name", -1));
+	    }
+	    parsePtr->errorType = TCL_PARSE_MISSING_VAR_BRACE;
+	    parsePtr->term = tokenPtr->start-1;
+	    parsePtr->incomplete = 1;
+	    goto error;
+	}
+	tokenPtr->size = src - tokenPtr->start;
+	tokenPtr[-1].size = src - tokenPtr[-1].start;
+	parsePtr->numTokens++;
+	src++;
+    } else {
+	tokenPtr->type = TCL_TOKEN_TEXT;
+	tokenPtr->start = src;
+	tokenPtr->numComponents = 0;
+
+	while (numBytes) {
+	    if (TclIsBareword(*src)) {
+		src += 1;
+		numBytes -= 1;
+		continue;
+	    }
+	    if ((src[0] == ':') && (numBytes != 1) && (src[1] == ':')) {
+		src += 2;
+		numBytes -= 2;
+		while (numBytes && (*src == ':')) {
+		    src++;
+		    numBytes--;
+		}
+		continue;
+	    }
+	    break;
+	}
+
+	/*
+	 * Support for empty array names here.
+	 */
+
+	array = (numBytes && (*src == '('));
+	tokenPtr->size = src - tokenPtr->start;
+	if ((tokenPtr->size == 0) && !array) {
+	    goto justADollarSign;
+	}
+	parsePtr->numTokens++;
+	if (array) {
+	    /*
+	     * This is a reference to an array element. Call ParseTokens
+	     * recursively to parse the element name, since it could contain
+	     * any number of substitutions.
+	     */
+
+	    if (TCL_OK != ParseTokens(src+1, numBytes-1, TYPE_CLOSE_PAREN,
+		    TCL_SUBST_ALL, parsePtr)) {
+		goto error;
+	    }
+	    if ((parsePtr->term == src+numBytes) || (*parsePtr->term != ')')){
+		if (parsePtr->interp != NULL) {
+		    Tcl_SetObjResult(parsePtr->interp, Tcl_NewStringObj(
+			    "missing )", -1));
+		}
+		parsePtr->errorType = TCL_PARSE_MISSING_PAREN;
+		parsePtr->term = src;
+		parsePtr->incomplete = 1;
+		goto error;
+	    }
+	    src = parsePtr->term + 1;
+	}
+    }
+    tokenPtr = &parsePtr->tokenPtr[varIndex];
+    tokenPtr->size = src - tokenPtr->start;
+    tokenPtr->numComponents = parsePtr->numTokens - (varIndex + 1);
+    return TCL_OK;
+
+    /*
+     * The dollar sign isn't followed by a variable name. Replace the
+     * TCL_TOKEN_VARIABLE token with a TCL_TOKEN_TEXT token for the dollar
+     * sign.
+     */
+
+  justADollarSign:
+    tokenPtr = &parsePtr->tokenPtr[varIndex];
+    tokenPtr->type = TCL_TOKEN_TEXT;
+    tokenPtr->size = 1;
+    tokenPtr->numComponents = 0;
+    return TCL_OK;
+
+  error:
+    Tcl_FreeParse(parsePtr);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseVar --
+ *
+ *	Given a string starting with a $ sign, parse off a variable name and
+ *	return its value.
+ *
+ * Results:
+ *	The return value is the contents of the variable given by the leading
+ *	characters of string. If termPtr isn't NULL, *termPtr gets filled in
+ *	with the address of the character just after the last one in the
+ *	variable specifier. If the variable doesn't exist, then the return
+ *	value is NULL and an error message will be left in interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_ParseVar(
+    Tcl_Interp *interp,		/* Context for looking up variable. */
+    register const char *start,	/* Start of variable substitution. First
+				 * character must be "$". */
+    const char **termPtr)	/* If non-NULL, points to word to fill in with
+				 * character just after last one in the
+				 * variable specifier. */
+{
+    register Tcl_Obj *objPtr;
+    int code;
+    Tcl_Parse *parsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse));
+
+    if (Tcl_ParseVarName(interp, start, -1, parsePtr, 0) != TCL_OK) {
+	TclStackFree(interp, parsePtr);
+	return NULL;
+    }
+
+    if (termPtr != NULL) {
+	*termPtr = start + parsePtr->tokenPtr->size;
+    }
+    if (parsePtr->numTokens == 1) {
+	/*
+	 * There isn't a variable name after all: the $ is just a $.
+	 */
+
+	TclStackFree(interp, parsePtr);
+	return "$";
+    }
+
+    code = TclSubstTokens(interp, parsePtr->tokenPtr, parsePtr->numTokens,
+	    NULL, 1, NULL, NULL);
+    Tcl_FreeParse(parsePtr);
+    TclStackFree(interp, parsePtr);
+    if (code != TCL_OK) {
+	return NULL;
+    }
+    objPtr = Tcl_GetObjResult(interp);
+
+    /*
+     * At this point we should have an object containing the value of a
+     * variable. Just return the string from that object.
+     *
+     * Since TclSubstTokens above returned TCL_OK, we know that objPtr
+     * is shared.  It is in both the interp result and the value of the
+     * variable.  Returning the string relies on that to be true.
+     */
+
+    assert( Tcl_IsShared(objPtr) );
+
+    Tcl_ResetResult(interp);
+    return TclGetString(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseBraces --
+ *
+ *	Given a string in braces such as a Tcl command argument or a string
+ *	value in a Tcl expression, this function parses the string and returns
+ *	information about the parse. No more than numBytes bytes will be
+ *	scanned.
+ *
+ * Results:
+ *	The return value is TCL_OK if the string was parsed successfully and
+ *	TCL_ERROR otherwise. If an error occurs and interp isn't NULL then an
+ *	error message is left in its result. On a successful return, tokenPtr
+ *	and numTokens fields of parsePtr are filled in with information about
+ *	the string that was parsed. Other fields in parsePtr are undefined.
+ *	termPtr is set to point to the character just after the last one in
+ *	the braced string.
+ *
+ * Side effects:
+ *	If there is insufficient space in parsePtr to hold all the information
+ *	about the command, then additional space is malloc-ed. If the function
+ *	returns TCL_OK then the caller must eventually invoke Tcl_FreeParse to
+ *	release any additional space that was allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseBraces(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting; if
+				 * NULL, then no error message is provided. */
+    const char *start,		/* Start of string enclosed in braces. The
+				 * first character must be {'. */
+    register int numBytes,	/* Total number of bytes in string. If < 0,
+				 * the string consists of all bytes up to the
+				 * first null character. */
+    register Tcl_Parse *parsePtr,
+				/* Structure to fill in with information about
+				 * the string. */
+    int append,			/* Non-zero means append tokens to existing
+				 * information in parsePtr; zero means ignore
+				 * existing tokens in parsePtr and
+				 * reinitialize it. */
+    const char **termPtr)	/* If non-NULL, points to word in which to
+				 * store a pointer to the character just after
+				 * the terminating '}' if the parse was
+				 * successful. */
+{
+    Tcl_Token *tokenPtr;
+    register const char *src;
+    int startIndex, level, length;
+
+    if ((numBytes == 0) || (start == NULL)) {
+	return TCL_ERROR;
+    }
+    if (numBytes < 0) {
+	numBytes = strlen(start);
+    }
+
+    if (!append) {
+	TclParseInit(interp, start, numBytes, parsePtr);
+    }
+
+    src = start;
+    startIndex = parsePtr->numTokens;
+
+    TclGrowParseTokenArray(parsePtr, 1);
+    tokenPtr = &parsePtr->tokenPtr[startIndex];
+    tokenPtr->type = TCL_TOKEN_TEXT;
+    tokenPtr->start = src+1;
+    tokenPtr->numComponents = 0;
+    level = 1;
+    while (1) {
+	while (++src, --numBytes) {
+	    if (CHAR_TYPE(*src) != TYPE_NORMAL) {
+		break;
+	    }
+	}
+	if (numBytes == 0) {
+	    goto missingBraceError;
+	}
+
+	switch (*src) {
+	case '{':
+	    level++;
+	    break;
+	case '}':
+	    if (--level == 0) {
+		/*
+		 * Decide if we need to finish emitting a partially-finished
+		 * token. There are 3 cases:
+		 *     {abc \newline xyz} or {xyz}
+		 *		- finish emitting "xyz" token
+		 *     {abc \newline}
+		 *		- don't emit token after \newline
+		 *     {}	- finish emitting zero-sized token
+		 *
+		 * The last case ensures that there is a token (even if empty)
+		 * that describes the braced string.
+		 */
+
+		if ((src != tokenPtr->start)
+			|| (parsePtr->numTokens == startIndex)) {
+		    tokenPtr->size = (src - tokenPtr->start);
+		    parsePtr->numTokens++;
+		}
+		if (termPtr != NULL) {
+		    *termPtr = src+1;
+		}
+		return TCL_OK;
+	    }
+	    break;
+	case '\\':
+	    TclParseBackslash(src, numBytes, &length, NULL);
+	    if ((length > 1) && (src[1] == '\n')) {
+		/*
+		 * A backslash-newline sequence must be collapsed, even inside
+		 * braces, so we have to split the word into multiple tokens
+		 * so that the backslash-newline can be represented
+		 * explicitly.
+		 */
+
+		if (numBytes == 2) {
+		    parsePtr->incomplete = 1;
+		}
+		tokenPtr->size = (src - tokenPtr->start);
+		if (tokenPtr->size != 0) {
+		    parsePtr->numTokens++;
+		}
+		TclGrowParseTokenArray(parsePtr, 2);
+		tokenPtr = &parsePtr->tokenPtr[parsePtr->numTokens];
+		tokenPtr->type = TCL_TOKEN_BS;
+		tokenPtr->start = src;
+		tokenPtr->size = length;
+		tokenPtr->numComponents = 0;
+		parsePtr->numTokens++;
+
+		src += length - 1;
+		numBytes -= length - 1;
+		tokenPtr++;
+		tokenPtr->type = TCL_TOKEN_TEXT;
+		tokenPtr->start = src + 1;
+		tokenPtr->numComponents = 0;
+	    } else {
+		src += length - 1;
+		numBytes -= length - 1;
+	    }
+	    break;
+	}
+    }
+
+  missingBraceError:
+    parsePtr->errorType = TCL_PARSE_MISSING_BRACE;
+    parsePtr->term = start;
+    parsePtr->incomplete = 1;
+    if (parsePtr->interp == NULL) {
+	/*
+	 * Skip straight to the exit code since we have no interpreter to put
+	 * error message in.
+	 */
+
+	goto error;
+    }
+
+    Tcl_SetObjResult(parsePtr->interp, Tcl_NewStringObj(
+	    "missing close-brace", -1));
+
+    /*
+     * Guess if the problem is due to comments by searching the source string
+     * for a possible open brace within the context of a comment. Since we
+     * aren't performing a full Tcl parse, just look for an open brace
+     * preceded by a '<whitespace>#' on the same line.
+     */
+
+    {
+	register int openBrace = 0;
+
+	while (--src > start) {
+	    switch (*src) {
+	    case '{':
+		openBrace = 1;
+		break;
+	    case '\n':
+		openBrace = 0;
+		break;
+	    case '#' :
+		if (openBrace && TclIsSpaceProc(src[-1])) {
+		    Tcl_AppendToObj(Tcl_GetObjResult(parsePtr->interp),
+			    ": possible unbalanced brace in comment", -1);
+		    goto error;
+		}
+		break;
+	    }
+	}
+    }
+
+  error:
+    Tcl_FreeParse(parsePtr);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ParseQuotedString --
+ *
+ *	Given a double-quoted string such as a quoted Tcl command argument or
+ *	a quoted value in a Tcl expression, this function parses the string
+ *	and returns information about the parse. No more than numBytes bytes
+ *	will be scanned.
+ *
+ * Results:
+ *	The return value is TCL_OK if the string was parsed successfully and
+ *	TCL_ERROR otherwise. If an error occurs and interp isn't NULL then an
+ *	error message is left in its result. On a successful return, tokenPtr
+ *	and numTokens fields of parsePtr are filled in with information about
+ *	the string that was parsed. Other fields in parsePtr are undefined.
+ *	termPtr is set to point to the character just after the quoted
+ *	string's terminating close-quote.
+ *
+ * Side effects:
+ *	If there is insufficient space in parsePtr to hold all the information
+ *	about the command, then additional space is malloc-ed. If the function
+ *	returns TCL_OK then the caller must eventually invoke Tcl_FreeParse to
+ *	release any additional space that was allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ParseQuotedString(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting; if
+				 * NULL, then no error message is provided. */
+    const char *start,		/* Start of the quoted string. The first
+				 * character must be '"'. */
+    register int numBytes,	/* Total number of bytes in string. If < 0,
+				 * the string consists of all bytes up to the
+				 * first null character. */
+    register Tcl_Parse *parsePtr,
+				/* Structure to fill in with information about
+				 * the string. */
+    int append,			/* Non-zero means append tokens to existing
+				 * information in parsePtr; zero means ignore
+				 * existing tokens in parsePtr and
+				 * reinitialize it. */
+    const char **termPtr)	/* If non-NULL, points to word in which to
+				 * store a pointer to the character just after
+				 * the quoted string's terminating close-quote
+				 * if the parse succeeds. */
+{
+    if ((numBytes == 0) || (start == NULL)) {
+	return TCL_ERROR;
+    }
+    if (numBytes < 0) {
+	numBytes = strlen(start);
+    }
+
+    if (!append) {
+	TclParseInit(interp, start, numBytes, parsePtr);
+    }
+
+    if (TCL_OK != ParseTokens(start+1, numBytes-1, TYPE_QUOTE, TCL_SUBST_ALL,
+	    parsePtr)) {
+	goto error;
+    }
+    if (*parsePtr->term != '"') {
+	if (parsePtr->interp != NULL) {
+	    Tcl_SetObjResult(parsePtr->interp, Tcl_NewStringObj(
+		    "missing \"", -1));
+	}
+	parsePtr->errorType = TCL_PARSE_MISSING_QUOTE;
+	parsePtr->term = start;
+	parsePtr->incomplete = 1;
+	goto error;
+    }
+    if (termPtr != NULL) {
+	*termPtr = (parsePtr->term + 1);
+    }
+    return TCL_OK;
+
+  error:
+    Tcl_FreeParse(parsePtr);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSubstParse --
+ *
+ *	Token parser used by the [subst] command. Parses the string made up of
+ *	'numBytes' bytes starting at 'bytes'. Parsing is controlled by the
+ *	flags argument to provide support for the -nobackslashes, -nocommands,
+ *	and -novariables options, as represented by the flag values
+ *	TCL_SUBST_BACKSLASHES, TCL_SUBST_COMMANDS, TCL_SUBST_VARIABLES.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The Tcl_Parse struct '*parsePtr' is filled with parse results.
+ *	The caller is expected to eventually call Tcl_FreeParse() to properly
+ *	cleanup the value written there.
+ *
+ *	If a parse error occurs, the Tcl_InterpState value '*statePtr' is
+ *	filled with the state created by that error. When *statePtr is written
+ *	to, the caller is expected to make the required calls to either
+ *	Tcl_RestoreInterpState() or Tcl_DiscardInterpState() to dispose of the
+ *	value written there.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSubstParse(
+    Tcl_Interp *interp,
+    const char *bytes,
+    int numBytes,
+    int flags,
+    Tcl_Parse *parsePtr,
+    Tcl_InterpState *statePtr)
+{
+    int length = numBytes;
+    const char *p = bytes;
+
+    TclParseInit(interp, p, length, parsePtr);
+
+    /*
+     * First parse the string rep of objPtr, as if it were enclosed as a
+     * "-quoted word in a normal Tcl command. Honor flags that selectively
+     * inhibit types of substitution.
+     */
+
+    if (TCL_OK != ParseTokens(p, length, /* mask */ 0, flags, parsePtr)) {
+	/*
+	 * There was a parse error. Save the interpreter state for possible
+	 * error reporting later.
+	 */
+
+	*statePtr = Tcl_SaveInterpState(interp, TCL_ERROR);
+
+	/*
+	 * We need to re-parse to get the portion of the string we can [subst]
+	 * before the parse error. Sadly, all the Tcl_Token's created by the
+	 * first parse attempt are gone, freed according to the public spec
+	 * for the Tcl_Parse* routines. The only clue we have is parse.term,
+	 * which points to either the unmatched opener, or to characters that
+	 * follow a close brace or close quote.
+	 *
+	 * Call ParseTokens again, working on the string up to parse.term.
+	 * Keep repeating until we get a good parse on a prefix.
+	 */
+
+	do {
+	    parsePtr->numTokens = 0;
+	    parsePtr->tokensAvailable = NUM_STATIC_TOKENS;
+	    parsePtr->end = parsePtr->term;
+	    parsePtr->incomplete = 0;
+	    parsePtr->errorType = TCL_PARSE_SUCCESS;
+	} while (TCL_OK !=
+		ParseTokens(p, parsePtr->end - p, 0, flags, parsePtr));
+
+	/*
+	 * The good parse will have to be followed by {, (, or [.
+	 */
+
+	switch (*(parsePtr->term)) {
+	case '{':
+	    /*
+	     * Parse error was a missing } in a ${varname} variable
+	     * substitution at the toplevel. We will subst everything up to
+	     * that broken variable substitution before reporting the parse
+	     * error. Substituting the leftover '$' will have no side-effects,
+	     * so the current token stream is fine.
+	     */
+	    break;
+
+	case '(':
+	    /*
+	     * Parse error was during the parsing of the index part of an
+	     * array variable substitution at the toplevel.
+	     */
+
+	    if (*(parsePtr->term - 1) == '$') {
+		/*
+		 * Special case where removing the array index left us with
+		 * just a dollar sign (array variable with name the empty
+		 * string as its name), instead of with a scalar variable
+		 * reference.
+		 *
+		 * As in the previous case, existing token stream is OK.
+		 */
+	    } else {
+		/*
+		 * The current parse includes a successful parse of a scalar
+		 * variable substitution where there should have been an array
+		 * variable substitution. We remove that mistaken part of the
+		 * parse before moving on. A scalar variable substitution is
+		 * two tokens.
+		 */
+
+		Tcl_Token *varTokenPtr =
+			parsePtr->tokenPtr + parsePtr->numTokens - 2;
+
+		if (varTokenPtr->type != TCL_TOKEN_VARIABLE) {
+		    Tcl_Panic("TclSubstParse: programming error");
+		}
+		if (varTokenPtr[1].type != TCL_TOKEN_TEXT) {
+		    Tcl_Panic("TclSubstParse: programming error");
+		}
+		parsePtr->numTokens -= 2;
+	    }
+	    break;
+	case '[':
+	    /*
+	     * Parse error occurred during parsing of a toplevel command
+	     * substitution.
+	     */
+
+	    parsePtr->end = p + length;
+	    p = parsePtr->term + 1;
+	    length = parsePtr->end - p;
+	    if (length == 0) {
+		/*
+		 * No commands, just an unmatched [. As in previous cases,
+		 * existing token stream is OK.
+		 */
+	    } else {
+		/*
+		 * We want to add the parsing of as many commands as we can
+		 * within that substitution until we reach the actual parse
+		 * error. We'll do additional parsing to determine what length
+		 * to claim for the final TCL_TOKEN_COMMAND token.
+		 */
+
+		Tcl_Token *tokenPtr;
+		const char *lastTerm = parsePtr->term;
+		Tcl_Parse *nestedPtr =
+			TclStackAlloc(interp, sizeof(Tcl_Parse));
+
+		while (TCL_OK ==
+			Tcl_ParseCommand(NULL, p, length, 0, nestedPtr)) {
+		    Tcl_FreeParse(nestedPtr);
+		    p = nestedPtr->term + (nestedPtr->term < nestedPtr->end);
+		    length = nestedPtr->end - p;
+		    if ((length == 0) && (nestedPtr->term == nestedPtr->end)) {
+			/*
+			 * If we run out of string, blame the missing close
+			 * bracket on the last command, and do not evaluate it
+			 * during substitution.
+			 */
+
+			break;
+		    }
+		    lastTerm = nestedPtr->term;
+		}
+		TclStackFree(interp, nestedPtr);
+
+		if (lastTerm == parsePtr->term) {
+		    /*
+		     * Parse error in first command. No commands to subst, add
+		     * no more tokens.
+		     */
+		    break;
+		}
+
+		/*
+		 * Create a command substitution token for whatever commands
+		 * got parsed.
+		 */
+
+		TclGrowParseTokenArray(parsePtr, 1);
+		tokenPtr = &(parsePtr->tokenPtr[parsePtr->numTokens]);
+		tokenPtr->start = parsePtr->term;
+		tokenPtr->numComponents = 0;
+		tokenPtr->type = TCL_TOKEN_COMMAND;
+		tokenPtr->size = lastTerm - tokenPtr->start + 1;
+		parsePtr->numTokens++;
+	    }
+	    break;
+
+	default:
+	    Tcl_Panic("bad parse in TclSubstParse: %c", p[length]);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSubstTokens --
+ *
+ *	Accepts an array of count Tcl_Token's, and creates a result value in
+ *	the interp from concatenating the results of performing Tcl
+ *	substitution on each Tcl_Token. Substitution is interrupted if any
+ *	non-TCL_OK completion code arises.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code. The result in
+ *	interp is the substituted value, or an error message if TCL_ERROR is
+ *	returned. If tokensLeftPtr is not NULL, then it points to an int where
+ *	the number of tokens remaining to be processed is written.
+ *
+ * Side effects:
+ *	Can be anything, depending on the types of substitution done.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclSubstTokens(
+    Tcl_Interp *interp,		/* Interpreter in which to lookup variables,
+				 * execute nested commands, and report
+				 * errors. */
+    Tcl_Token *tokenPtr,	/* Pointer to first in an array of tokens to
+				 * evaluate and concatenate. */
+    int count,			/* Number of tokens to consider at tokenPtr.
+				 * Must be at least 1. */
+    int *tokensLeftPtr,		/* If not NULL, points to memory where an
+				 * integer representing the number of tokens
+				 * left to be substituted will be written */
+    int line,			/* The line the script starts on. */
+    int *clNextOuter,		/* Information about an outer context for */
+    const char *outerScript)	/* continuation line data. This is set by
+				 * EvalEx() to properly handle [...]-nested
+				 * commands. The 'outerScript' refers to the
+				 * most-outer script containing the embedded
+				 * command, which is refered to by 'script'.
+				 * The 'clNextOuter' refers to the current
+				 * entry in the table of continuation lines in
+				 * this "master script", and the character
+				 * offsets are relative to the 'outerScript'
+				 * as well.
+				 *
+				 * If outerScript == script, then this call is
+				 * for words in the outer-most script or
+				 * command. See Tcl_EvalEx and TclEvalObjEx
+				 * for the places generating arguments for
+				 * which this is true. */
+{
+    Tcl_Obj *result;
+    int code = TCL_OK;
+#define NUM_STATIC_POS 20
+    int isLiteral, maxNumCL, numCL, i, adjust;
+    int *clPosition = NULL;
+    Interp *iPtr = (Interp *) interp;
+    int inFile = iPtr->evalFlags & TCL_EVAL_FILE;
+
+    /*
+     * Each pass through this loop will substitute one token, and its
+     * components, if any. The only thing tricky here is that we go to some
+     * effort to pass Tcl_Obj's through untouched, to avoid string copying and
+     * Tcl_Obj creation if possible, to aid performance and limit shimmering.
+     *
+     * Further optimization opportunities might be to check for the equivalent
+     * of Tcl_SetObjResult(interp, Tcl_GetObjResult(interp)) and omit them.
+     */
+
+    /*
+     * For the handling of continuation lines in literals we first check if
+     * this is actually a literal. For if not we can forego the additional
+     * processing. Otherwise we pre-allocate a small table to store the
+     * locations of all continuation lines we find in this literal, if any.
+     * The table is extended if needed.
+     */
+
+    numCL = 0;
+    maxNumCL = 0;
+    isLiteral = 1;
+    for (i=0 ; i < count; i++) {
+	if ((tokenPtr[i].type != TCL_TOKEN_TEXT)
+		&& (tokenPtr[i].type != TCL_TOKEN_BS)) {
+	    isLiteral = 0;
+	    break;
+	}
+    }
+
+    if (isLiteral) {
+	maxNumCL = NUM_STATIC_POS;
+	clPosition = ckalloc(maxNumCL * sizeof(int));
+    }
+
+    adjust = 0;
+    result = NULL;
+    for (; count>0 && code==TCL_OK ; count--, tokenPtr++) {
+	Tcl_Obj *appendObj = NULL;
+	const char *append = NULL;
+	int appendByteLength = 0;
+	char utfCharBytes[TCL_UTF_MAX];
+
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_TEXT:
+	    append = tokenPtr->start;
+	    appendByteLength = tokenPtr->size;
+	    break;
+
+	case TCL_TOKEN_BS:
+	    appendByteLength = TclParseBackslash(tokenPtr->start,
+		    tokenPtr->size, NULL, utfCharBytes);
+	    append = utfCharBytes;
+
+	    /*
+	     * If the backslash sequence we found is in a literal, and
+	     * represented a continuation line, we compute and store its
+	     * location (as char offset to the beginning of the _result_
+	     * script). We may have to extend the table of locations.
+	     *
+	     * Note that the continuation line information is relevant even if
+	     * the word we are processing is not a literal, as it can affect
+	     * nested commands. See the branch for TCL_TOKEN_COMMAND below,
+	     * where the adjustment we are tracking here is taken into
+	     * account. The good thing is that we do not need a table of
+	     * everything, just the number of lines we have to add as
+	     * correction.
+	     */
+
+	    if ((appendByteLength == 1) && (utfCharBytes[0] == ' ')
+		    && (tokenPtr->start[1] == '\n')) {
+		if (isLiteral) {
+		    int clPos;
+
+		    if (result == 0) {
+			clPos = 0;
+		    } else {
+			TclGetStringFromObj(result, &clPos);
+		    }
+
+		    if (numCL >= maxNumCL) {
+			maxNumCL *= 2;
+			clPosition = ckrealloc(clPosition,
+				maxNumCL * sizeof(int));
+		    }
+		    clPosition[numCL] = clPos;
+		    numCL++;
+		}
+		adjust++;
+	    }
+	    break;
+
+	case TCL_TOKEN_COMMAND: {
+	    /* TIP #280: Transfer line information to nested command */
+	    iPtr->numLevels++;
+	    code = TclInterpReady(interp);
+	    if (code == TCL_OK) {
+		/*
+		 * Test cases: info-30.{6,8,9}
+		 */
+
+		int theline;
+
+		TclAdvanceContinuations(&line, &clNextOuter,
+			tokenPtr->start - outerScript);
+		theline = line + adjust;
+		code = TclEvalEx(interp, tokenPtr->start+1, tokenPtr->size-2,
+			0, theline, clNextOuter, outerScript);
+
+		TclAdvanceLines(&line, tokenPtr->start+1,
+			tokenPtr->start + tokenPtr->size - 1);
+
+		/*
+		 * Restore flag reset by nested eval for future bracketed
+		 * commands and their cmdframe setup
+		 */
+
+		if (inFile) {
+		    iPtr->evalFlags |= TCL_EVAL_FILE;
+		}
+	    }
+	    iPtr->numLevels--;
+	    TclResetCancellation(interp, 0);
+	    appendObj = Tcl_GetObjResult(interp);
+	    break;
+	}
+
+	case TCL_TOKEN_VARIABLE: {
+	    Tcl_Obj *arrayIndex = NULL;
+	    Tcl_Obj *varName = NULL;
+
+	    if (tokenPtr->numComponents > 1) {
+		/*
+		 * Subst the index part of an array variable reference.
+		 */
+
+		code = TclSubstTokens(interp, tokenPtr+2,
+			tokenPtr->numComponents - 1, NULL, line, NULL, NULL);
+		arrayIndex = Tcl_GetObjResult(interp);
+		Tcl_IncrRefCount(arrayIndex);
+	    }
+
+	    if (code == TCL_OK) {
+		varName = Tcl_NewStringObj(tokenPtr[1].start,
+			tokenPtr[1].size);
+		appendObj = Tcl_ObjGetVar2(interp, varName, arrayIndex,
+			TCL_LEAVE_ERR_MSG);
+		Tcl_DecrRefCount(varName);
+		if (appendObj == NULL) {
+		    code = TCL_ERROR;
+		}
+	    }
+
+	    switch (code) {
+	    case TCL_OK:	/* Got value */
+	    case TCL_ERROR:	/* Already have error message */
+	    case TCL_BREAK:	/* Will not substitute anyway */
+	    case TCL_CONTINUE:	/* Will not substitute anyway */
+		break;
+	    default:
+		/*
+		 * All other return codes, we will subst the result from the
+		 * code-throwing evaluation.
+		 */
+
+		appendObj = Tcl_GetObjResult(interp);
+	    }
+
+	    if (arrayIndex != NULL) {
+		Tcl_DecrRefCount(arrayIndex);
+	    }
+	    count -= tokenPtr->numComponents;
+	    tokenPtr += tokenPtr->numComponents;
+	    break;
+	}
+
+	default:
+	    Tcl_Panic("unexpected token type in TclSubstTokens: %d",
+		    tokenPtr->type);
+	}
+
+	if ((code == TCL_BREAK) || (code == TCL_CONTINUE)) {
+	    /*
+	     * Inhibit substitution.
+	     */
+	    continue;
+	}
+
+	if (result == NULL) {
+	    /*
+	     * First pass through. If we have a Tcl_Obj, just use it. If not,
+	     * create one from our string.
+	     */
+
+	    if (appendObj != NULL) {
+		result = appendObj;
+	    } else {
+		result = Tcl_NewStringObj(append, appendByteLength);
+	    }
+	    Tcl_IncrRefCount(result);
+	} else {
+	    /*
+	     * Subsequent passes. Append to result.
+	     */
+
+	    if (Tcl_IsShared(result)) {
+		Tcl_DecrRefCount(result);
+		result = Tcl_DuplicateObj(result);
+		Tcl_IncrRefCount(result);
+	    }
+	    if (appendObj != NULL) {
+		Tcl_AppendObjToObj(result, appendObj);
+	    } else {
+		Tcl_AppendToObj(result, append, appendByteLength);
+	    }
+	}
+    }
+
+    if (code != TCL_ERROR) {		/* Keep error message in result! */
+	if (result != NULL) {
+	    Tcl_SetObjResult(interp, result);
+
+	    /*
+	     * If the code found continuation lines (which implies that this
+	     * word is a literal), then we store the accumulated table of
+	     * locations in the thread-global data structure for the bytecode
+	     * compiler to find later, assuming that the literal is a script
+	     * which will be compiled.
+	     */
+
+	    if (numCL) {
+		TclContinuationsEnter(result, numCL, clPosition);
+	    }
+
+	    /*
+	     * Release the temp table we used to collect the locations of
+	     * continuation lines, if any.
+	     */
+
+	    if (maxNumCL) {
+		ckfree(clPosition);
+	    }
+	} else {
+	    Tcl_ResetResult(interp);
+	}
+    }
+    if (tokensLeftPtr != NULL) {
+	*tokensLeftPtr = count;
+    }
+    if (result != NULL) {
+	Tcl_DecrRefCount(result);
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CommandComplete --
+ *
+ *	This function is shared by TclCommandComplete and
+ *	Tcl_ObjCommandComplete; it does all the real work of seeing whether a
+ *	script is complete
+ *
+ * Results:
+ *	1 is returned if the script is complete, 0 if there are open
+ *	delimiters such as " or (. 1 is also returned if there is a parse
+ *	error in the script other than unmatched delimiters.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+CommandComplete(
+    const char *script,		/* Script to check. */
+    int numBytes)		/* Number of bytes in script. */
+{
+    Tcl_Parse parse;
+    const char *p, *end;
+    int result;
+
+    p = script;
+    end = p + numBytes;
+    while (Tcl_ParseCommand(NULL, p, end - p, 0, &parse) == TCL_OK) {
+	p = parse.commandStart + parse.commandSize;
+	if (p >= end) {
+	    break;
+	}
+	Tcl_FreeParse(&parse);
+    }
+    if (parse.incomplete) {
+	result = 0;
+    } else {
+	result = 1;
+    }
+    Tcl_FreeParse(&parse);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CommandComplete --
+ *
+ *	Given a partial or complete Tcl script, this function determines
+ *	whether the script is complete in the sense of having matched braces
+ *	and quotes and brackets.
+ *
+ * Results:
+ *	1 is returned if the script is complete, 0 otherwise. 1 is also
+ *	returned if there is a parse error in the script other than unmatched
+ *	delimiters.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_CommandComplete(
+    const char *script)		/* Script to check. */
+{
+    return CommandComplete(script, (int) strlen(script));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjCommandComplete --
+ *
+ *	Given a partial or complete Tcl command in a Tcl object, this function
+ *	determines whether the command is complete in the sense of having
+ *	matched braces and quotes and brackets.
+ *
+ * Results:
+ *	1 is returned if the command is complete, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjCommandComplete(
+    Tcl_Obj *objPtr)		/* Points to object holding script to
+				 * check. */
+{
+    int length;
+    const char *script = TclGetStringFromObj(objPtr, &length);
+
+    return CommandComplete(script, length);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclParse.h b/generic/tclParse.h
new file mode 100644
index 0000000..20c609c
--- /dev/null
+++ b/generic/tclParse.h
@@ -0,0 +1,17 @@
+/*
+ * Minimal set of shared macro definitions and declarations so that multiple
+ * source files can make use of the parsing table in tclParse.c
+ */
+
+#define TYPE_NORMAL		0
+#define TYPE_SPACE		0x1
+#define TYPE_COMMAND_END	0x2
+#define TYPE_SUBS		0x4
+#define TYPE_QUOTE		0x8
+#define TYPE_CLOSE_PAREN	0x10
+#define TYPE_CLOSE_BRACK	0x20
+#define TYPE_BRACE		0x40
+
+#define CHAR_TYPE(c) (tclCharTypeTable+128)[(int)(c)]
+
+MODULE_SCOPE const char tclCharTypeTable[];
diff --git a/generic/tclPathObj.c b/generic/tclPathObj.c
new file mode 100644
index 0000000..49d62dc
--- /dev/null
+++ b/generic/tclPathObj.c
@@ -0,0 +1,2708 @@
+/*
+ * tclPathObj.c --
+ *
+ *	This file contains the implementation of Tcl's "path" object type used
+ *	to represent and manipulate a general (virtual) filesystem entity in
+ *	an efficient manner.
+ *
+ * Copyright (c) 2003 Vince Darley.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclFileSystem.h"
+#include <assert.h>
+
+/*
+ * Prototypes for functions defined later in this file.
+ */
+
+static Tcl_Obj *	AppendPath(Tcl_Obj *head, Tcl_Obj *tail);
+static void		DupFsPathInternalRep(Tcl_Obj *srcPtr,
+			    Tcl_Obj *copyPtr);
+static void		FreeFsPathInternalRep(Tcl_Obj *pathPtr);
+static void		UpdateStringOfFsPath(Tcl_Obj *pathPtr);
+static int		SetFsPathFromAny(Tcl_Interp *interp, Tcl_Obj *pathPtr);
+static int		FindSplitPos(const char *path, int separator);
+static int		IsSeparatorOrNull(int ch);
+static Tcl_Obj *	GetExtension(Tcl_Obj *pathPtr);
+static int		MakePathFromNormalized(Tcl_Interp *interp,
+			    Tcl_Obj *pathPtr);
+
+/*
+ * Define the 'path' object type, which Tcl uses to represent file paths
+ * internally.
+ */
+
+static const Tcl_ObjType tclFsPathType = {
+    "path",				/* name */
+    FreeFsPathInternalRep,		/* freeIntRepProc */
+    DupFsPathInternalRep,		/* dupIntRepProc */
+    UpdateStringOfFsPath,		/* updateStringProc */
+    SetFsPathFromAny			/* setFromAnyProc */
+};
+
+/*
+ * struct FsPath --
+ *
+ * Internal representation of a Tcl_Obj of "path" type. This can be used to
+ * represent relative or absolute paths, and has certain optimisations when
+ * used to represent paths which are already normalized and absolute.
+ *
+ * Note that both 'translatedPathPtr' and 'normPathPtr' can be a circular
+ * reference to the container Tcl_Obj of this FsPath.
+ *
+ * There are two cases, with the first being the most common:
+ *
+ * (i) flags == 0, => Ordinary path.
+ *
+ * translatedPathPtr contains the translated path (which may be a circular
+ * reference to the object itself). If it is NULL then the path is pure
+ * normalized (and the normPathPtr will be a circular reference). cwdPtr is
+ * null for an absolute path, and non-null for a relative path (unless the cwd
+ * has never been set, in which case the cwdPtr may also be null for a
+ * relative path).
+ *
+ * (ii) flags != 0, => Special path, see TclNewFSPathObj
+ *
+ * Now, this is a path like 'file join $dir $tail' where, cwdPtr is the $dir
+ * and normPathPtr is the $tail.
+ *
+ */
+
+typedef struct FsPath {
+    Tcl_Obj *translatedPathPtr; /* Name without any ~user sequences. If this
+				 * is NULL, then this is a pure normalized,
+				 * absolute path object, in which the parent
+				 * Tcl_Obj's string rep is already both
+				 * translated and normalized. */
+    Tcl_Obj *normPathPtr;	/* Normalized absolute path, without ., .. or
+				 * ~user sequences. If the Tcl_Obj containing
+				 * this FsPath is already normalized, this may
+				 * be a circular reference back to the
+				 * container. If that is NOT the case, we have
+				 * a refCount on the object. */
+    Tcl_Obj *cwdPtr;		/* If null, path is absolute, else this points
+				 * to the cwd object used for this path. We
+				 * have a refCount on the object. */
+    int flags;			/* Flags to describe interpretation - see
+				 * below. */
+    ClientData nativePathPtr;	/* Native representation of this path, which
+				 * is filesystem dependent. */
+    int filesystemEpoch;	/* Used to ensure the path representation was
+				 * generated during the correct filesystem
+				 * epoch. The epoch changes when
+				 * filesystem-mounts are changed. */
+    const Tcl_Filesystem *fsPtr;/* The Tcl_Filesystem that claims this path */
+} FsPath;
+
+/*
+ * Flag values for FsPath->flags.
+ */
+
+#define TCLPATH_APPENDED 1
+#define TCLPATH_NEEDNORM 4
+
+/*
+ * Define some macros to give us convenient access to path-object specific
+ * fields.
+ */
+
+#define PATHOBJ(pathPtr) ((FsPath *) (pathPtr)->internalRep.twoPtrValue.ptr1)
+#define SETPATHOBJ(pathPtr,fsPathPtr) \
+	((pathPtr)->internalRep.twoPtrValue.ptr1 = (void *) (fsPathPtr))
+#define PATHFLAGS(pathPtr) (PATHOBJ(pathPtr)->flags)
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFSNormalizeAbsolutePath --
+ *
+ *	Takes an absolute path specification and computes a 'normalized' path
+ *	from it.
+ *
+ *	A normalized path is one which has all '../', './' removed. Also it is
+ *	one which is in the 'standard' format for the native platform. On
+ *	Unix, this means the path must be free of symbolic links/aliases, and
+ *	on Windows it means we want the long form, with that long form's
+ *	case-dependence (which gives us a unique, case-dependent path).
+ *
+ *	The behaviour of this function if passed a non-absolute path is NOT
+ *	defined.
+ *
+ *	pathPtr may have a refCount of zero, or may be a shared object.
+ *
+ * Results:
+ *	The result is returned in a Tcl_Obj with a refCount of 1, which is
+ *	therefore owned by the caller. It must be freed (with
+ *	Tcl_DecrRefCount) by the caller when no longer needed.
+ *
+ * Side effects:
+ *	None (beyond the memory allocation for the result).
+ *
+ * Special note:
+ *	This code was originally based on code from Matt Newman and
+ *	Jean-Claude Wippler, but has since been totally rewritten by Vince
+ *	Darley to deal with symbolic links.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclFSNormalizeAbsolutePath(
+    Tcl_Interp *interp,		/* Interpreter to use */
+    Tcl_Obj *pathPtr)		/* Absolute path to normalize */
+{
+    const char *dirSep, *oldDirSep;
+    int first = 1;		/* Set to zero once we've passed the first
+				 * directory separator - we can't use '..' to
+				 * remove the volume in a path. */
+    Tcl_Obj *retVal = NULL;
+    dirSep = TclGetString(pathPtr);
+
+    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+	if (   (dirSep[0] == '/' || dirSep[0] == '\\')
+	    && (dirSep[1] == '/' || dirSep[1] == '\\')
+	    && (dirSep[2] == '?')
+	    && (dirSep[3] == '/' || dirSep[3] == '\\')) {
+	    /* NT extended path */
+	    dirSep += 4;
+
+	    if (   (dirSep[0] == 'U' || dirSep[0] == 'u')
+		&& (dirSep[1] == 'N' || dirSep[1] == 'n')
+		&& (dirSep[2] == 'C' || dirSep[2] == 'c')
+		&& (dirSep[3] == '/' || dirSep[3] == '\\')) {
+		/* NT extended UNC path */
+		dirSep += 4;
+	    }
+	}
+	if (dirSep[0] != 0 && dirSep[1] == ':' &&
+		(dirSep[2] == '/' || dirSep[2] == '\\')) {
+	    /* Do nothing */
+	} else if ((dirSep[0] == '/' || dirSep[0] == '\\')
+		&& (dirSep[1] == '/' || dirSep[1] == '\\')) {
+	    /*
+	     * UNC style path, where we must skip over the first separator,
+	     * since the first two segments are actually inseparable.
+	     */
+
+	    dirSep += 2;
+	    dirSep += FindSplitPos(dirSep, '/');
+	    if (*dirSep != 0) {
+		dirSep++;
+	    }
+	}
+    }
+
+    /*
+     * Scan forward from one directory separator to the next, checking for
+     * '..' and '.' sequences which must be handled specially. In particular
+     * handling of '..' can be complicated if the directory before is a link,
+     * since we will have to expand the link to be able to back up one level.
+     */
+
+    while (*dirSep != 0) {
+	oldDirSep = dirSep;
+	if (!first) {
+	    dirSep++;
+	}
+	dirSep += FindSplitPos(dirSep, '/');
+	if (dirSep[0] == 0 || dirSep[1] == 0) {
+	    if (retVal != NULL) {
+		Tcl_AppendToObj(retVal, oldDirSep, dirSep - oldDirSep);
+	    }
+	    break;
+	}
+	if (dirSep[1] == '.') {
+	    if (retVal != NULL) {
+		Tcl_AppendToObj(retVal, oldDirSep, dirSep - oldDirSep);
+		oldDirSep = dirSep;
+	    }
+	again:
+	    if (IsSeparatorOrNull(dirSep[2])) {
+		/*
+		 * Need to skip '.' in the path.
+		 */
+		int curLen;
+
+		if (retVal == NULL) {
+		    const char *path = TclGetString(pathPtr);
+		    retVal = Tcl_NewStringObj(path, dirSep - path);
+		    Tcl_IncrRefCount(retVal);
+		}
+		TclGetStringFromObj(retVal, &curLen);
+		if (curLen == 0) {
+		    Tcl_AppendToObj(retVal, dirSep, 1);
+		}
+		dirSep += 2;
+		oldDirSep = dirSep;
+		if (dirSep[0] != 0 && dirSep[1] == '.') {
+		    goto again;
+		}
+		continue;
+	    }
+	    if (dirSep[2] == '.' && IsSeparatorOrNull(dirSep[3])) {
+		Tcl_Obj *linkObj;
+		int curLen;
+		char *linkStr;
+
+		/*
+		 * Have '..' so need to skip previous directory.
+		 */
+
+		if (retVal == NULL) {
+		    const char *path = TclGetString(pathPtr);
+
+		    retVal = Tcl_NewStringObj(path, dirSep - path);
+		    Tcl_IncrRefCount(retVal);
+		}
+		TclGetStringFromObj(retVal, &curLen);
+		if (curLen == 0) {
+		    Tcl_AppendToObj(retVal, dirSep, 1);
+		}
+		if (!first || (tclPlatform == TCL_PLATFORM_UNIX)) {
+		    linkObj = Tcl_FSLink(retVal, NULL, 0);
+
+		    /* Safety check in case driver caused sharing */
+		    if (Tcl_IsShared(retVal)) {
+			TclDecrRefCount(retVal);
+			retVal = Tcl_DuplicateObj(retVal);
+			Tcl_IncrRefCount(retVal);
+		    }
+
+		    if (linkObj != NULL) {
+			/*
+			 * Got a link. Need to check if the link is relative
+			 * or absolute, for those platforms where relative
+			 * links exist.
+			 */
+
+			if (tclPlatform != TCL_PLATFORM_WINDOWS
+				&& Tcl_FSGetPathType(linkObj)
+					== TCL_PATH_RELATIVE) {
+			    /*
+			     * We need to follow this link which is relative
+			     * to retVal's directory. This means concatenating
+			     * the link onto the directory of the path so far.
+			     */
+
+			    const char *path =
+				    TclGetStringFromObj(retVal, &curLen);
+
+			    while (--curLen >= 0) {
+				if (IsSeparatorOrNull(path[curLen])) {
+				    break;
+				}
+			    }
+
+			    /*
+			     * We want the trailing slash.
+			     */
+
+			    Tcl_SetObjLength(retVal, curLen+1);
+			    Tcl_AppendObjToObj(retVal, linkObj);
+			    TclDecrRefCount(linkObj);
+			    linkStr = TclGetStringFromObj(retVal, &curLen);
+			} else {
+			    /*
+			     * Absolute link.
+			     */
+
+			    TclDecrRefCount(retVal);
+			    if (Tcl_IsShared(linkObj)) {
+				retVal = Tcl_DuplicateObj(linkObj);
+				TclDecrRefCount(linkObj);
+			    } else {
+				retVal = linkObj;
+			    }
+			    linkStr = TclGetStringFromObj(retVal, &curLen);
+
+			    /*
+			     * Convert to forward-slashes on windows.
+			     */
+
+			    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+				int i;
+
+				for (i = 0; i < curLen; i++) {
+				    if (linkStr[i] == '\\') {
+					linkStr[i] = '/';
+				    }
+				}
+			    }
+			}
+		    } else {
+			linkStr = TclGetStringFromObj(retVal, &curLen);
+		    }
+
+		    /*
+		     * Either way, we now remove the last path element (but
+		     * not the first character of the path).
+		     */
+
+		    while (--curLen >= 0) {
+			if (IsSeparatorOrNull(linkStr[curLen])) {
+			    if (curLen) {
+				Tcl_SetObjLength(retVal, curLen);
+			    } else {
+				Tcl_SetObjLength(retVal, 1);
+			    }
+			    break;
+			}
+		    }
+		}
+		dirSep += 3;
+		oldDirSep = dirSep;
+
+		if ((curLen == 0) && (dirSep[0] != 0)) {
+		    Tcl_SetObjLength(retVal, 0);
+		}
+
+		if (dirSep[0] != 0 && dirSep[1] == '.') {
+		    goto again;
+		}
+		continue;
+	    }
+	}
+	first = 0;
+	if (retVal != NULL) {
+	    Tcl_AppendToObj(retVal, oldDirSep, dirSep - oldDirSep);
+	}
+    }
+
+    /*
+     * If we didn't make any changes, just use the input path.
+     */
+
+    if (retVal == NULL) {
+	retVal = pathPtr;
+	Tcl_IncrRefCount(retVal);
+
+	if (Tcl_IsShared(retVal)) {
+	    /*
+	     * Unfortunately, the platform-specific normalization code which
+	     * will be called below has no way of dealing with the case where
+	     * an object is shared. It is expecting to modify an object in
+	     * place. So, we must duplicate this here to ensure an object with
+	     * a single ref-count.
+	     *
+	     * If that changes in the future (e.g. the normalize proc is given
+	     * one object and is able to return a different one), then we
+	     * could remove this code.
+	     */
+
+	    TclDecrRefCount(retVal);
+	    retVal = Tcl_DuplicateObj(pathPtr);
+	    Tcl_IncrRefCount(retVal);
+	}
+    }
+
+    /*
+     * Ensure a windows drive like C:/ has a trailing separator.
+     */
+
+    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+	int len;
+	const char *path = TclGetStringFromObj(retVal, &len);
+
+	if (len == 2 && path[0] != 0 && path[1] == ':') {
+	    if (Tcl_IsShared(retVal)) {
+		TclDecrRefCount(retVal);
+		retVal = Tcl_DuplicateObj(retVal);
+		Tcl_IncrRefCount(retVal);
+	    }
+	    Tcl_AppendToObj(retVal, "/", 1);
+	}
+    }
+
+    /*
+     * Now we have an absolute path, with no '..', '.' sequences, but it still
+     * may not be in 'unique' form, depending on the platform. For instance,
+     * Unix is case-sensitive, so the path is ok. Windows is case-insensitive,
+     * and also has the weird 'longname/shortname' thing (e.g. C:/Program
+     * Files/ and C:/Progra~1/ are equivalent).
+     *
+     * Virtual file systems which may be registered may have other criteria
+     * for normalizing a path.
+     */
+
+    TclFSNormalizeToUniquePath(interp, retVal, 0);
+
+    /*
+     * Since we know it is a normalized path, we can actually convert this
+     * object into an FsPath for greater efficiency
+     */
+
+    MakePathFromNormalized(interp, retVal);
+
+    /*
+     * This has a refCount of 1 for the caller, unlike many Tcl_Obj APIs.
+     */
+
+    return retVal;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FSGetPathType --
+ *
+ *	Determines whether a given path is relative to the current directory,
+ *	relative to the current volume, or absolute.
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE, TCL_PATH_RELATIVE, or
+ *	TCL_PATH_VOLUME_RELATIVE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+Tcl_FSGetPathType(
+    Tcl_Obj *pathPtr)
+{
+    return TclFSGetPathType(pathPtr, NULL, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFSGetPathType --
+ *
+ *	Determines whether a given path is relative to the current directory,
+ *	relative to the current volume, or absolute. If the caller wishes to
+ *	know which filesystem claimed the path (in the case for which the path
+ *	is absolute), then a reference to a filesystem pointer can be passed
+ *	in (but passing NULL is acceptable).
+ *
+ * Results:
+ *	Returns one of TCL_PATH_ABSOLUTE, TCL_PATH_RELATIVE, or
+ *	TCL_PATH_VOLUME_RELATIVE. The filesystem reference will be set if and
+ *	only if it is non-NULL and the function's return value is
+ *	TCL_PATH_ABSOLUTE.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_PathType
+TclFSGetPathType(
+    Tcl_Obj *pathPtr,
+    const Tcl_Filesystem **filesystemPtrPtr,
+    int *driveNameLengthPtr)
+{
+    FsPath *fsPathPtr;
+
+    if (Tcl_FSConvertToPathType(NULL, pathPtr) != TCL_OK) {
+	return TclGetPathType(pathPtr, filesystemPtrPtr, driveNameLengthPtr,
+		NULL);
+    }
+
+    fsPathPtr = PATHOBJ(pathPtr);
+    if (fsPathPtr->cwdPtr == NULL) {
+	return TclGetPathType(pathPtr, filesystemPtrPtr, driveNameLengthPtr,
+		NULL);
+    }
+
+    if (PATHFLAGS(pathPtr) == 0) {
+	/* The path is not absolute... */
+#ifdef _WIN32
+	/* ... on Windows we must make another call to determine whether
+	 * it's relative or volumerelative [Bug 2571597]. */
+	return TclGetPathType(pathPtr, filesystemPtrPtr, driveNameLengthPtr,
+		NULL);
+#else
+	/* On other systems, quickly deduce !absolute -> relative */
+	return TCL_PATH_RELATIVE;
+#endif
+    }
+    return TclFSGetPathType(fsPathPtr->cwdPtr, filesystemPtrPtr,
+	    driveNameLengthPtr);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclPathPart
+ *
+ *	This function calculates the requested part of the given path, which
+ *	can be:
+ *
+ *	- the directory above ('file dirname')
+ *	- the tail            ('file tail')
+ *	- the extension       ('file extension')
+ *	- the root            ('file root')
+ *
+ *	The 'portion' parameter dictates which of these to calculate. There
+ *	are a number of special cases both to be more efficient, and because
+ *	the behaviour when given a path with only a single element is defined
+ *	to require the expansion of that single element, where possible.
+ *
+ *	Should look into integrating 'FileBasename' in tclFCmd.c into this
+ *	function.
+ *
+ * Results:
+ *	NULL if an error occurred, otherwise a Tcl_Obj owned by the caller
+ *	(i.e. most likely with refCount 1).
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPathPart(
+    Tcl_Interp *interp,		/* Used for error reporting */
+    Tcl_Obj *pathPtr,		/* Path to take dirname of */
+    Tcl_PathPart portion)	/* Requested portion of name */
+{
+    if (pathPtr->typePtr == &tclFsPathType) {
+	FsPath *fsPathPtr = PATHOBJ(pathPtr);
+
+	if (PATHFLAGS(pathPtr) != 0) {
+	    switch (portion) {
+	    case TCL_PATH_DIRNAME: {
+		/*
+		 * Check if the joined-on bit has any directory delimiters in
+		 * it. If so, the 'dirname' would be a joining of the main
+		 * part with the dirname of the joined-on bit. We could handle
+		 * that special case here, but we don't, and instead just use
+		 * the standardPath code.
+		 */
+
+		int numBytes;
+		const char *rest =
+			TclGetStringFromObj(fsPathPtr->normPathPtr, &numBytes);
+
+		if (strchr(rest, '/') != NULL) {
+		    goto standardPath;
+		}
+		/*
+		 * If the joined-on bit is empty, then [file dirname] is
+		 * documented to return all but the last non-empty element
+		 * of the path, so we need to split apart the main part to
+		 * get the right answer.  We could do that here, but it's
+		 * simpler to fall back to the standardPath code.
+		 * [Bug 2710920]
+		 */
+		if (numBytes == 0) {
+		    goto standardPath;
+		}
+		if (tclPlatform == TCL_PLATFORM_WINDOWS
+			&& strchr(rest, '\\') != NULL) {
+		    goto standardPath;
+		}
+
+		/*
+		 * The joined-on path is simple, so we can just return here.
+		 */
+
+		Tcl_IncrRefCount(fsPathPtr->cwdPtr);
+		return fsPathPtr->cwdPtr;
+	    }
+	    case TCL_PATH_TAIL: {
+		/*
+		 * Check if the joined-on bit has any directory delimiters in
+		 * it. If so, the 'tail' would be only the part following the
+		 * last delimiter. We could handle that special case here, but
+		 * we don't, and instead just use the standardPath code.
+		 */
+
+		int numBytes;
+		const char *rest =
+			TclGetStringFromObj(fsPathPtr->normPathPtr, &numBytes);
+
+		if (strchr(rest, '/') != NULL) {
+		    goto standardPath;
+		}
+		/*
+		 * If the joined-on bit is empty, then [file tail] is
+		 * documented to return the last non-empty element
+		 * of the path, so we need to split off the last element
+		 * of the main part to get the right answer.  We could do
+		 * that here, but it's simpler to fall back to the
+		 * standardPath code.  [Bug 2710920]
+		 */
+		if (numBytes == 0) {
+		    goto standardPath;
+		}
+		if (tclPlatform == TCL_PLATFORM_WINDOWS
+			&& strchr(rest, '\\') != NULL) {
+		    goto standardPath;
+		}
+		Tcl_IncrRefCount(fsPathPtr->normPathPtr);
+		return fsPathPtr->normPathPtr;
+	    }
+	    case TCL_PATH_EXTENSION:
+		return GetExtension(fsPathPtr->normPathPtr);
+	    case TCL_PATH_ROOT: {
+		const char *fileName, *extension;
+		int length;
+
+		fileName = TclGetStringFromObj(fsPathPtr->normPathPtr,
+			&length);
+		extension = TclGetExtension(fileName);
+		if (extension == NULL) {
+		    /*
+		     * There is no extension so the root is the same as the
+		     * path we were given.
+		     */
+
+		    Tcl_IncrRefCount(pathPtr);
+		    return pathPtr;
+		} else {
+		    /*
+		     * Need to return the whole path with the extension
+		     * suffix removed.  Do that by joining our "head" to
+		     * our "tail" with the extension suffix removed from
+		     * the tail.
+		     */
+
+		    Tcl_Obj *resultPtr =
+			    TclNewFSPathObj(fsPathPtr->cwdPtr, fileName,
+			    (int)(length - strlen(extension)));
+
+		    Tcl_IncrRefCount(resultPtr);
+		    return resultPtr;
+		}
+	    }
+	    default:
+		/* We should never get here */
+		Tcl_Panic("Bad portion to TclPathPart");
+		/* For less clever compilers */
+		return NULL;
+	    }
+	} else if (fsPathPtr->cwdPtr != NULL) {
+	    /* Relative path */
+	    goto standardPath;
+	} else {
+	    /* Absolute path */
+	    goto standardPath;
+	}
+    } else {
+	int splitElements;
+	Tcl_Obj *splitPtr, *resultPtr;
+
+    standardPath:
+	resultPtr = NULL;
+	if (portion == TCL_PATH_EXTENSION) {
+	    return GetExtension(pathPtr);
+	} else if (portion == TCL_PATH_ROOT) {
+	    int length;
+	    const char *fileName, *extension;
+
+	    fileName = TclGetStringFromObj(pathPtr, &length);
+	    extension = TclGetExtension(fileName);
+	    if (extension == NULL) {
+		Tcl_IncrRefCount(pathPtr);
+		return pathPtr;
+	    } else {
+		Tcl_Obj *root = Tcl_NewStringObj(fileName,
+			(int) (length - strlen(extension)));
+
+		Tcl_IncrRefCount(root);
+		return root;
+	    }
+	}
+
+	/*
+	 * The behaviour we want here is slightly different to the standard
+	 * Tcl_FSSplitPath in the handling of home directories;
+	 * Tcl_FSSplitPath preserves the "~" while this code computes the
+	 * actual full path name, if we had just a single component.
+	 */
+
+	splitPtr = Tcl_FSSplitPath(pathPtr, &splitElements);
+	Tcl_IncrRefCount(splitPtr);
+	if (splitElements == 1  &&  TclGetString(pathPtr)[0] == '~') {
+	    Tcl_Obj *norm;
+
+	    TclDecrRefCount(splitPtr);
+	    norm = Tcl_FSGetNormalizedPath(interp, pathPtr);
+	    if (norm == NULL) {
+		return NULL;
+	    }
+	    splitPtr = Tcl_FSSplitPath(norm, &splitElements);
+	    Tcl_IncrRefCount(splitPtr);
+	}
+	if (portion == TCL_PATH_TAIL) {
+	    /*
+	     * Return the last component, unless it is the only component, and
+	     * it is the root of an absolute path.
+	     */
+
+	    if ((splitElements > 0) && ((splitElements > 1) ||
+		    (Tcl_FSGetPathType(pathPtr) == TCL_PATH_RELATIVE))) {
+		Tcl_ListObjIndex(NULL, splitPtr, splitElements-1, &resultPtr);
+	    } else {
+		resultPtr = Tcl_NewObj();
+	    }
+	} else {
+	    /*
+	     * Return all but the last component. If there is only one
+	     * component, return it if the path was non-relative, otherwise
+	     * return the current directory.
+	     */
+
+	    if (splitElements > 1) {
+		resultPtr = Tcl_FSJoinPath(splitPtr, splitElements - 1);
+	    } else if (splitElements == 0 ||
+		    (Tcl_FSGetPathType(pathPtr) == TCL_PATH_RELATIVE)) {
+		TclNewLiteralStringObj(resultPtr, ".");
+	    } else {
+		Tcl_ListObjIndex(NULL, splitPtr, 0, &resultPtr);
+	    }
+	}
+	Tcl_IncrRefCount(resultPtr);
+	TclDecrRefCount(splitPtr);
+	return resultPtr;
+    }
+}
+
+/*
+ * Simple helper function
+ */
+
+static Tcl_Obj *
+GetExtension(
+    Tcl_Obj *pathPtr)
+{
+    const char *tail, *extension;
+    Tcl_Obj *ret;
+
+    tail = TclGetString(pathPtr);
+    extension = TclGetExtension(tail);
+    if (extension == NULL) {
+	ret = Tcl_NewObj();
+    } else {
+	ret = Tcl_NewStringObj(extension, -1);
+    }
+    Tcl_IncrRefCount(ret);
+    return ret;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSJoinPath --
+ *
+ *	This function takes the given Tcl_Obj, which should be a valid list,
+ *	and returns the path object given by considering the first 'elements'
+ *	elements as valid path segments (each path segment may be a complete
+ *	path, a partial path or just a single possible directory or file
+ *	name). If any path segment is actually an absolute path, then all
+ *	prior path segments are discarded.
+ *
+ *	If elements < 0, we use the entire list that was given.
+ *
+ *	It is possible that the returned object is actually an element of the
+ *	given list, so the caller should be careful to store a refCount to it
+ *	before freeing the list.
+ *
+ * Results:
+ *	Returns object with refCount of zero, (or if non-zero, it has
+ *	references elsewhere in Tcl). Either way, the caller must increment
+ *	its refCount before use. Note that in the case where the caller has
+ *	asked to join zero elements of the list, the return value will be an
+ *	empty-string Tcl_Obj.
+ *
+ *	If the given listObj was invalid, then the calling routine has a bug,
+ *	and this function will just return NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSJoinPath(
+    Tcl_Obj *listObj,		/* Path elements to join, may have a zero
+				 * reference count. */
+    int elements)		/* Number of elements to use (-1 = all) */
+{
+    Tcl_Obj *copy, *res;
+    int objc;
+    Tcl_Obj **objv;
+
+    if (Tcl_ListObjLength(NULL, listObj, &objc) != TCL_OK) {
+	return NULL;
+    }
+
+    elements = ((elements >= 0) && (elements <= objc)) ? elements : objc;
+    copy = TclListObjCopy(NULL, listObj);
+    Tcl_ListObjGetElements(NULL, listObj, &objc, &objv);
+    res = TclJoinPath(elements, objv);
+    Tcl_DecrRefCount(copy);
+    return res;
+}
+
+Tcl_Obj *
+TclJoinPath(
+    int elements,
+    Tcl_Obj * const objv[])
+{
+    Tcl_Obj *res = NULL;
+    int i;
+    const Tcl_Filesystem *fsPtr = NULL;
+
+    assert ( elements >= 0 );
+
+    if (elements == 0) {
+	return Tcl_NewObj();
+    }
+
+    assert ( elements > 0 );
+
+    if (elements == 2) {
+	Tcl_Obj *elt = objv[0];
+
+	/*
+	 * This is a special case where we can be much more efficient, where
+	 * we are joining a single relative path onto an object that is
+	 * already of path type. The 'TclNewFSPathObj' call below creates an
+	 * object which can be normalized more efficiently. Currently we only
+	 * use the special case when we have exactly two elements, but we
+	 * could expand that in the future.
+	 *
+	 * Bugfix [a47641a0]. TclNewFSPathObj requires first argument
+	 * to be an absolute path. Added a check for that elt is absolute.
+	 */
+
+	if ((elt->typePtr == &tclFsPathType)
+		&& !((elt->bytes != NULL) && (elt->bytes[0] == '\0'))
+                && TclGetPathType(elt, NULL, NULL, NULL) == TCL_PATH_ABSOLUTE) {
+            Tcl_Obj *tailObj = objv[1];
+	    Tcl_PathType type = TclGetPathType(tailObj, NULL, NULL, NULL);
+
+	    if (type == TCL_PATH_RELATIVE) {
+		const char *str;
+		int len;
+
+		str = TclGetStringFromObj(tailObj, &len);
+		if (len == 0) {
+		    /*
+		     * This happens if we try to handle the root volume '/'.
+		     * There's no need to return a special path object, when
+		     * the base itself is just fine!
+		     */
+
+		    return elt;
+		}
+
+		/*
+		 * If it doesn't begin with '.' and is a unix path or it a
+		 * windows path without backslashes, then we can be very
+		 * efficient here. (In fact even a windows path with
+		 * backslashes can be joined efficiently, but the path object
+		 * would not have forward slashes only, and this would
+		 * therefore contradict our 'file join' documentation).
+		 */
+
+		if (str[0] != '.' && ((tclPlatform != TCL_PLATFORM_WINDOWS)
+			|| (strchr(str, '\\') == NULL))) {
+		    /*
+		     * Finally, on Windows, 'file join' is defined to convert
+		     * all backslashes to forward slashes, so the base part
+		     * cannot have backslashes either.
+		     */
+
+		    if ((tclPlatform != TCL_PLATFORM_WINDOWS)
+			    || (strchr(Tcl_GetString(elt), '\\') == NULL)) {
+
+			if (PATHFLAGS(elt)) {
+			    return TclNewFSPathObj(elt, str, len);
+			}
+			if (TCL_PATH_ABSOLUTE != Tcl_FSGetPathType(elt)) {
+			    return TclNewFSPathObj(elt, str, len);
+			}
+			(void) Tcl_FSGetNormalizedPath(NULL, elt);
+			if (elt == PATHOBJ(elt)->normPathPtr) {
+			    return TclNewFSPathObj(elt, str, len);
+			}
+		    }
+		}
+
+		/*
+		 * Otherwise we don't have an easy join, and we must let the
+		 * more general code below handle things.
+		 */
+	    } else if (tclPlatform == TCL_PLATFORM_UNIX) {
+		return tailObj;
+	    } else {
+		const char *str = TclGetString(tailObj);
+
+		if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+		    if (strchr(str, '\\') == NULL) {
+			return tailObj;
+		    }
+		}
+	    }
+	}
+    }
+
+    assert ( res == NULL );
+
+    for (i = 0; i < elements; i++) {
+	int driveNameLength, strEltLen, length;
+	Tcl_PathType type;
+	char *strElt, *ptr;
+	Tcl_Obj *driveName = NULL;
+	Tcl_Obj *elt = objv[i];
+
+	strElt = TclGetStringFromObj(elt, &strEltLen);
+	driveNameLength = 0;
+	type = TclGetPathType(elt, &fsPtr, &driveNameLength, &driveName);
+	if (type != TCL_PATH_RELATIVE) {
+	    /*
+	     * Zero out the current result.
+	     */
+
+	    if (res != NULL) {
+		TclDecrRefCount(res);
+	    }
+
+	    if (driveName != NULL) {
+		/*
+		 * We've been given a separate drive-name object, because the
+		 * prefix in 'elt' is not in a suitable format for us (e.g. it
+		 * may contain irrelevant multiple separators, like
+		 * C://///foo).
+		 */
+
+		res = Tcl_DuplicateObj(driveName);
+		TclDecrRefCount(driveName);
+
+		/*
+		 * Do not set driveName to NULL, because we will check its
+		 * value below (but we won't access the contents, since those
+		 * have been cleaned-up).
+		 */
+	    } else {
+		res = Tcl_NewStringObj(strElt, driveNameLength);
+	    }
+	    strElt += driveNameLength;
+	} else if (driveName != NULL) {
+	    Tcl_DecrRefCount(driveName);
+	}
+
+	/*
+	 * Optimisation block: if this is the last element to be examined, and
+	 * it is absolute or the only element, and the drive-prefix was ok (if
+	 * there is one), it might be that the path is already in a suitable
+	 * form to be returned. Then we can short-cut the rest of this
+	 * function.
+	 */
+
+	if ((driveName == NULL) && (i == (elements - 1))
+		&& (type != TCL_PATH_RELATIVE || res == NULL)) {
+	    /*
+	     * It's the last path segment. Perform a quick check if the path
+	     * is already in a suitable form.
+	     */
+
+	    if (tclPlatform == TCL_PLATFORM_WINDOWS) {
+		if (strchr(strElt, '\\') != NULL) {
+		    goto noQuickReturn;
+		}
+	    }
+	    ptr = strElt;
+	    /* [Bug f34cf83dd0] */
+	    if (driveNameLength > 0) {
+		if (ptr[0] == '/' && ptr[-1] == '/') {
+		    goto noQuickReturn;
+		}
+	    }
+	    while (*ptr != '\0') {
+		if (*ptr == '/' && (ptr[1] == '/' || ptr[1] == '\0')) {
+		    /*
+		     * We have a repeated file separator, which means the path
+		     * is not in normalized form
+		     */
+
+		    goto noQuickReturn;
+		}
+		ptr++;
+	    }
+	    if (res != NULL) {
+		TclDecrRefCount(res);
+	    }
+
+	    /*
+	     * This element is just what we want to return already; no further
+	     * manipulation is requred.
+	     */
+
+	    return elt;
+	}
+
+	/*
+	 * The path element was not of a suitable form to be returned as is.
+	 * We need to perform a more complex operation here.
+	 */
+
+    noQuickReturn:
+	if (res == NULL) {
+	    res = Tcl_NewObj();
+	}
+	ptr = TclGetStringFromObj(res, &length);
+
+	/*
+	 * Strip off any './' before a tilde, unless this is the beginning of
+	 * the path.
+	 */
+
+	if (length > 0 && strEltLen > 0 && (strElt[0] == '.') &&
+		(strElt[1] == '/') && (strElt[2] == '~')) {
+	    strElt += 2;
+	}
+
+	/*
+	 * A NULL value for fsPtr at this stage basically means we're trying
+	 * to join a relative path onto something which is also relative (or
+	 * empty). There's nothing particularly wrong with that.
+	 */
+
+	if (*strElt == '\0') {
+	    continue;
+	}
+
+	if (fsPtr == &tclNativeFilesystem || fsPtr == NULL) {
+	    TclpNativeJoinPath(res, strElt);
+	} else {
+	    char separator = '/';
+	    int needsSep = 0;
+
+	    if (fsPtr->filesystemSeparatorProc != NULL) {
+		Tcl_Obj *sep = fsPtr->filesystemSeparatorProc(res);
+
+		if (sep != NULL) {
+		    separator = TclGetString(sep)[0];
+		    TclDecrRefCount(sep);
+		}
+		/* Safety check in case the VFS driver caused sharing */
+		if (Tcl_IsShared(res)) {
+		    TclDecrRefCount(res);
+		    res = Tcl_DuplicateObj(res);
+		    Tcl_IncrRefCount(res);
+		}
+	    }
+
+	    if (length > 0 && ptr[length -1] != '/') {
+		Tcl_AppendToObj(res, &separator, 1);
+		TclGetStringFromObj(res, &length);
+	    }
+	    Tcl_SetObjLength(res, length + (int) strlen(strElt));
+
+	    ptr = TclGetString(res) + length;
+	    for (; *strElt != '\0'; strElt++) {
+		if (*strElt == separator) {
+		    while (strElt[1] == separator) {
+			strElt++;
+		    }
+		    if (strElt[1] != '\0') {
+			if (needsSep) {
+			    *ptr++ = separator;
+			}
+		    }
+		} else {
+		    *ptr++ = *strElt;
+		    needsSep = 1;
+		}
+	    }
+	    length = ptr - TclGetString(res);
+	    Tcl_SetObjLength(res, length);
+	}
+    }
+    assert ( res != NULL );
+    return res;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSConvertToPathType --
+ *
+ *	This function tries to convert the given Tcl_Obj to a valid Tcl path
+ *	type, taking account of the fact that the cwd may have changed even if
+ *	this object is already supposedly of the correct type.
+ *
+ *	The filename may begin with "~" (to indicate current user's home
+ *	directory) or "~<user>" (to indicate any user's home directory).
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	The old representation may be freed, and new memory allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSConvertToPathType(
+    Tcl_Interp *interp,		/* Interpreter in which to store error message
+				 * (if necessary). */
+    Tcl_Obj *pathPtr)		/* Object to convert to a valid, current path
+				 * type. */
+{
+    /*
+     * While it is bad practice to examine an object's type directly, this is
+     * actually the best thing to do here. The reason is that if we are
+     * converting this object to FsPath type for the first time, we don't need
+     * to worry whether the 'cwd' has changed. On the other hand, if this
+     * object is already of FsPath type, and is a relative path, we do have to
+     * worry about the cwd. If the cwd has changed, we must recompute the
+     * path.
+     */
+
+    if (pathPtr->typePtr == &tclFsPathType) {
+	if (TclFSEpochOk(PATHOBJ(pathPtr)->filesystemEpoch)) {
+	    return TCL_OK;
+	}
+
+	if (pathPtr->bytes == NULL) {
+	    UpdateStringOfFsPath(pathPtr);
+	}
+	FreeFsPathInternalRep(pathPtr);
+    }
+
+    return SetFsPathFromAny(interp, pathPtr);
+
+    /*
+     * We used to have more complex code here:
+     *
+     * FsPath *fsPathPtr = PATHOBJ(pathPtr);
+     * if (fsPathPtr->cwdPtr == NULL || PATHFLAGS(pathPtr) != 0) {
+     *     return TCL_OK;
+     * } else {
+     *     if (TclFSCwdPointerEquals(&fsPathPtr->cwdPtr)) {
+     *         return TCL_OK;
+     *     } else {
+     *         if (pathPtr->bytes == NULL) {
+     *             UpdateStringOfFsPath(pathPtr);
+     *         }
+     *         FreeFsPathInternalRep(pathPtr);
+     *         return Tcl_ConvertToType(interp, pathPtr, &tclFsPathType);
+     *     }
+     * }
+     *
+     * But we no longer believe this is necessary.
+     */
+}
+
+/*
+ * Helper function for normalization.
+ */
+
+static int
+IsSeparatorOrNull(
+    int ch)
+{
+    if (ch == 0) {
+	return 1;
+    }
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	return (ch == '/' ? 1 : 0);
+    case TCL_PLATFORM_WINDOWS:
+	return ((ch == '/' || ch == '\\') ? 1 : 0);
+    }
+    return 0;
+}
+
+/*
+ * Helper function for SetFsPathFromAny. Returns position of first directory
+ * delimiter in the path. If no separator is found, then returns the position
+ * of the end of the string.
+ */
+
+static int
+FindSplitPos(
+    const char *path,
+    int separator)
+{
+    int count = 0;
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	while (path[count] != 0) {
+	    if (path[count] == separator) {
+		return count;
+	    }
+	    count++;
+	}
+	break;
+
+    case TCL_PLATFORM_WINDOWS:
+	while (path[count] != 0) {
+	    if (path[count] == separator || path[count] == '\\') {
+		return count;
+	    }
+	    count++;
+	}
+	break;
+    }
+    return count;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclNewFSPathObj --
+ *
+ *	Creates a path object whose string representation is '[file join
+ *	dirPtr addStrRep]', but does so in a way that allows for more
+ *	efficient creation and caching of normalized paths, and more efficient
+ *	'file dirname', 'file tail', etc.
+ *
+ * Assumptions:
+ *	'dirPtr' must be an absolute path. 'len' may not be zero.
+ *
+ * Results:
+ *	The new Tcl object, with refCount zero.
+ *
+ * Side effects:
+ *	Memory is allocated. 'dirPtr' gets an additional refCount.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclNewFSPathObj(
+    Tcl_Obj *dirPtr,
+    const char *addStrRep,
+    int len)
+{
+    FsPath *fsPathPtr;
+    Tcl_Obj *pathPtr;
+    const char *p;
+    int state = 0, count = 0;
+
+    /* [Bug 2806250] - this is only a partial solution of the problem.
+     * The PATHFLAGS != 0 representation assumes in many places that
+     * the "tail" part stored in the normPathPtr field is itself a
+     * relative path.  Strings that begin with "~" are not relative paths,
+     * so we must prevent their storage in the normPathPtr field.
+     *
+     * More generally we ought to be testing "addStrRep" for any value
+     * that is not a relative path, but in an unconstrained VFS world
+     * that could be just about anything, and testing could be expensive.
+     * Since this routine plays a big role in [glob], anything that slows
+     * it down would be unwelcome.  For now, continue the risk of further
+     * bugs when some Tcl_Filesystem uses otherwise relative path strings
+     * as absolute path strings.  Sensible Tcl_Filesystems will avoid
+     * that by mounting on path prefixes like foo:// which cannot be the
+     * name of a file or directory read from a native [glob] operation.
+     */
+    if (addStrRep[0] == '~') {
+	Tcl_Obj *tail = Tcl_NewStringObj(addStrRep, len);
+
+	pathPtr = AppendPath(dirPtr, tail);
+	Tcl_DecrRefCount(tail);
+	return pathPtr;
+    }
+
+    pathPtr = Tcl_NewObj();
+    fsPathPtr = ckalloc(sizeof(FsPath));
+
+    /*
+     * Set up the path.
+     */
+
+    fsPathPtr->translatedPathPtr = NULL;
+    fsPathPtr->normPathPtr = Tcl_NewStringObj(addStrRep, len);
+    Tcl_IncrRefCount(fsPathPtr->normPathPtr);
+    fsPathPtr->cwdPtr = dirPtr;
+    Tcl_IncrRefCount(dirPtr);
+    fsPathPtr->nativePathPtr = NULL;
+    fsPathPtr->fsPtr = NULL;
+    fsPathPtr->filesystemEpoch = 0;
+
+    SETPATHOBJ(pathPtr, fsPathPtr);
+    PATHFLAGS(pathPtr) = TCLPATH_APPENDED;
+    pathPtr->typePtr = &tclFsPathType;
+    pathPtr->bytes = NULL;
+    pathPtr->length = 0;
+
+    /*
+     * Look for path components made up of only "."
+     * This is overly conservative analysis to keep simple. It may mark some
+     * things as needing more aggressive normalization that don't actually
+     * need it. No harm done.
+     */
+    for (p = addStrRep; len > 0; p++, len--) {
+	switch (state) {
+	case 0:		/* So far only "." since last dirsep or start */
+	    switch (*p) {
+	    case '.':
+		count++;
+		break;
+	    case '/':
+	    case '\\':
+	    case ':':
+		if (count) {
+		    PATHFLAGS(pathPtr) |= TCLPATH_NEEDNORM;
+		    len = 0;
+		}
+		break;
+	    default:
+		count = 0;
+		state = 1;
+	    }
+	case 1:		/* Scanning for next dirsep */
+	    switch (*p) {
+	    case '/':
+	    case '\\':
+	    case ':':
+		state = 0;
+		break;
+	    }
+	}
+    }
+    if (len == 0 && count) {
+	PATHFLAGS(pathPtr) |= TCLPATH_NEEDNORM;
+    }
+
+    return pathPtr;
+}
+
+static Tcl_Obj *
+AppendPath(
+    Tcl_Obj *head,
+    Tcl_Obj *tail)
+{
+    int numBytes;
+    const char *bytes;
+    Tcl_Obj *copy = Tcl_DuplicateObj(head);
+
+    /*
+     * This is likely buggy when dealing with virtual filesystem drivers
+     * that use some character other than "/" as a path separator.  I know
+     * of no evidence that such a foolish thing exists.  This solution was
+     * chosen so that "JoinPath" operations that pass through either path
+     * intrep produce the same results; that is, bugward compatibility.  If
+     * we need to fix that bug here, it needs fixing in TclJoinPath() too.
+     */
+    bytes = TclGetStringFromObj(tail, &numBytes);
+    if (numBytes == 0) {
+	Tcl_AppendToObj(copy, "/", 1);
+    } else {
+	TclpNativeJoinPath(copy, bytes);
+    }
+    return copy;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFSMakePathRelative --
+ *
+ *	Only for internal use.
+ *
+ *	Takes a path and a directory, where we _assume_ both path and
+ *	directory are absolute, normalized and that the path lies inside the
+ *	directory. Returns a Tcl_Obj representing filename of the path
+ *	relative to the directory.
+ *
+ * Results:
+ *	NULL on error, otherwise a valid object, typically with refCount of
+ *	zero, which it is assumed the caller will increment.
+ *
+ * Side effects:
+ *	The old representation may be freed, and new memory allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclFSMakePathRelative(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *pathPtr,		/* The path we have. */
+    Tcl_Obj *cwdPtr)		/* Make it relative to this. */
+{
+    int cwdLen, len;
+    const char *tempStr;
+
+    if (pathPtr->typePtr == &tclFsPathType) {
+	FsPath *fsPathPtr = PATHOBJ(pathPtr);
+
+	if (PATHFLAGS(pathPtr) != 0 && fsPathPtr->cwdPtr == cwdPtr) {
+	    return fsPathPtr->normPathPtr;
+	}
+    }
+
+    /*
+     * We know the cwd is a normalised object which does not end in a
+     * directory delimiter, unless the cwd is the name of a volume, in which
+     * case it will end in a delimiter! We handle this situation here. A
+     * better test than the '!= sep' might be to simply check if 'cwd' is a
+     * root volume.
+     *
+     * Note that if we get this wrong, we will strip off either too much or
+     * too little below, leading to wrong answers returned by glob.
+     */
+
+    tempStr = TclGetStringFromObj(cwdPtr, &cwdLen);
+
+    /*
+     * Should we perhaps use 'Tcl_FSPathSeparator'? But then what about the
+     * Windows special case? Perhaps we should just check if cwd is a root
+     * volume.
+     */
+
+    switch (tclPlatform) {
+    case TCL_PLATFORM_UNIX:
+	if (tempStr[cwdLen-1] != '/') {
+	    cwdLen++;
+	}
+	break;
+    case TCL_PLATFORM_WINDOWS:
+	if (tempStr[cwdLen-1] != '/' && tempStr[cwdLen-1] != '\\') {
+	    cwdLen++;
+	}
+	break;
+    }
+    tempStr = TclGetStringFromObj(pathPtr, &len);
+
+    return Tcl_NewStringObj(tempStr + cwdLen, len - cwdLen);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * MakePathFromNormalized --
+ *
+ *	Like SetFsPathFromAny, but assumes the given object is an absolute
+ *	normalized path. Only for internal use.
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	The old representation may be freed, and new memory allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+MakePathFromNormalized(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *pathPtr)		/* The object to convert. */
+{
+    FsPath *fsPathPtr;
+
+    if (pathPtr->typePtr == &tclFsPathType) {
+	return TCL_OK;
+    }
+
+    /*
+     * Free old representation
+     */
+
+    if (pathPtr->typePtr != NULL) {
+	if (pathPtr->bytes == NULL) {
+	    if (pathPtr->typePtr->updateStringProc == NULL) {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "can't find object string representation", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "PATH", "WTF",
+			    NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    pathPtr->typePtr->updateStringProc(pathPtr);
+	}
+	TclFreeIntRep(pathPtr);
+    }
+
+    fsPathPtr = ckalloc(sizeof(FsPath));
+
+    /*
+     * It's a pure normalized absolute path.
+     */
+
+    fsPathPtr->translatedPathPtr = NULL;
+
+    /*
+     * Circular reference by design.
+     */
+
+    fsPathPtr->normPathPtr = pathPtr;
+    fsPathPtr->cwdPtr = NULL;
+    fsPathPtr->nativePathPtr = NULL;
+    fsPathPtr->fsPtr = NULL;
+    /* Remember the epoch under which we decided pathPtr was normalized */
+    fsPathPtr->filesystemEpoch = TclFSEpoch();
+
+    SETPATHOBJ(pathPtr, fsPathPtr);
+    PATHFLAGS(pathPtr) = 0;
+    pathPtr->typePtr = &tclFsPathType;
+
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSNewNativePath --
+ *
+ *	This function performs the something like the reverse of the usual
+ *	obj->path->nativerep conversions. If some code retrieves a path in
+ *	native form (from, e.g. readlink or a native dialog), and that path is
+ *	to be used at the Tcl level, then calling this function is an
+ *	efficient way of creating the appropriate path object type.
+ *
+ *	Any memory which is allocated for 'clientData' should be retained
+ *	until clientData is passed to the filesystem's freeInternalRepProc
+ *	when it can be freed. The built in platform-specific filesystems use
+ *	'ckalloc' to allocate clientData, and ckfree to free it.
+ *
+ * Results:
+ *	NULL or a valid path object pointer, with refCount zero.
+ *
+ * Side effects:
+ *	New memory may be allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSNewNativePath(
+    const Tcl_Filesystem *fromFilesystem,
+    ClientData clientData)
+{
+    Tcl_Obj *pathPtr = NULL;
+    FsPath *fsPathPtr;
+
+
+    if (fromFilesystem->internalToNormalizedProc != NULL) {
+	pathPtr = (*fromFilesystem->internalToNormalizedProc)(clientData);
+    }
+    if (pathPtr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * Free old representation; shouldn't normally be any, but best to be
+     * safe.
+     */
+
+    if (pathPtr->typePtr != NULL) {
+	if (pathPtr->bytes == NULL) {
+	    if (pathPtr->typePtr->updateStringProc == NULL) {
+		return NULL;
+	    }
+	    pathPtr->typePtr->updateStringProc(pathPtr);
+	}
+	TclFreeIntRep(pathPtr);
+    }
+
+    fsPathPtr = ckalloc(sizeof(FsPath));
+
+    fsPathPtr->translatedPathPtr = NULL;
+
+    /*
+     * Circular reference, by design.
+     */
+
+    fsPathPtr->normPathPtr = pathPtr;
+    fsPathPtr->cwdPtr = NULL;
+    fsPathPtr->nativePathPtr = clientData;
+    fsPathPtr->fsPtr = fromFilesystem;
+    fsPathPtr->filesystemEpoch = TclFSEpoch();
+
+    SETPATHOBJ(pathPtr, fsPathPtr);
+    PATHFLAGS(pathPtr) = 0;
+    pathPtr->typePtr = &tclFsPathType;
+
+    return pathPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetTranslatedPath --
+ *
+ *	This function attempts to extract the translated path from the given
+ *	Tcl_Obj. If the translation succeeds (i.e. the object is a valid
+ *	path), then it is returned. Otherwise NULL will be returned, and an
+ *	error message may be left in the interpreter (if it is non-NULL)
+ *
+ * Results:
+ *	NULL or a valid Tcl_Obj pointer.
+ *
+ * Side effects:
+ *	Only those of 'Tcl_FSConvertToPathType'
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSGetTranslatedPath(
+    Tcl_Interp *interp,
+    Tcl_Obj *pathPtr)
+{
+    Tcl_Obj *retObj = NULL;
+    FsPath *srcFsPathPtr;
+
+    if (Tcl_FSConvertToPathType(interp, pathPtr) != TCL_OK) {
+	return NULL;
+    }
+    srcFsPathPtr = PATHOBJ(pathPtr);
+    if (srcFsPathPtr->translatedPathPtr == NULL) {
+	if (PATHFLAGS(pathPtr) != 0) {
+	    /*
+	     * We lack a translated path result, but we have a directory
+	     * (cwdPtr) and a tail (normPathPtr), and if we join the
+	     * translated version of cwdPtr to normPathPtr, we'll get the
+	     * translated result we need, and can store it for future use.
+	     */
+
+	    Tcl_Obj *translatedCwdPtr = Tcl_FSGetTranslatedPath(interp,
+		    srcFsPathPtr->cwdPtr);
+	    if (translatedCwdPtr == NULL) {
+		return NULL;
+	    }
+
+	    retObj = Tcl_FSJoinToPath(translatedCwdPtr, 1,
+		    &srcFsPathPtr->normPathPtr);
+	    srcFsPathPtr->translatedPathPtr = retObj;
+	    if (translatedCwdPtr->typePtr == &tclFsPathType) {
+		srcFsPathPtr->filesystemEpoch
+			= PATHOBJ(translatedCwdPtr)->filesystemEpoch;
+	    } else {
+		srcFsPathPtr->filesystemEpoch = 0;
+	    }
+	    Tcl_IncrRefCount(retObj);
+	    Tcl_DecrRefCount(translatedCwdPtr);
+	} else {
+	    /*
+	     * It is a pure absolute, normalized path object. This is
+	     * something like being a 'pure list'. The object's string,
+	     * translatedPath and normalizedPath are all identical.
+	     */
+
+	    retObj = srcFsPathPtr->normPathPtr;
+	}
+    } else {
+	/*
+	 * It is an ordinary path object.
+	 */
+
+	retObj = srcFsPathPtr->translatedPathPtr;
+    }
+
+    if (retObj != NULL) {
+	Tcl_IncrRefCount(retObj);
+    }
+    return retObj;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetTranslatedStringPath --
+ *
+ *	This function attempts to extract the translated path from the given
+ *	Tcl_Obj. If the translation succeeds (i.e. the object is a valid
+ *	path), then the path is returned. Otherwise NULL will be returned, and
+ *	an error message may be left in the interpreter (if it is non-NULL)
+ *
+ * Results:
+ *	NULL or a valid string.
+ *
+ * Side effects:
+ *	Only those of 'Tcl_FSConvertToPathType'
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_FSGetTranslatedStringPath(
+    Tcl_Interp *interp,
+    Tcl_Obj *pathPtr)
+{
+    Tcl_Obj *transPtr = Tcl_FSGetTranslatedPath(interp, pathPtr);
+
+    if (transPtr != NULL) {
+	int len;
+	const char *orig = TclGetStringFromObj(transPtr, &len);
+	char *result = ckalloc(len+1);
+
+	memcpy(result, orig, (size_t) len+1);
+	TclDecrRefCount(transPtr);
+	return result;
+    }
+
+    return NULL;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetNormalizedPath --
+ *
+ *	This important function attempts to extract from the given Tcl_Obj a
+ *	unique normalised path representation, whose string value can be used
+ *	as a unique identifier for the file.
+ *
+ * Results:
+ *	NULL or a valid path object pointer.
+ *
+ * Side effects:
+ *	New memory may be allocated. The Tcl 'errno' may be modified in the
+ *	process of trying to examine various path possibilities.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_FSGetNormalizedPath(
+    Tcl_Interp *interp,
+    Tcl_Obj *pathPtr)
+{
+    FsPath *fsPathPtr;
+
+    if (Tcl_FSConvertToPathType(interp, pathPtr) != TCL_OK) {
+	return NULL;
+    }
+    fsPathPtr = PATHOBJ(pathPtr);
+
+    if (PATHFLAGS(pathPtr) != 0) {
+	/*
+	 * This is a special path object which is the result of something like
+	 * 'file join'
+	 */
+
+	Tcl_Obj *dir, *copy;
+	int tailLen, cwdLen, pathType;
+
+	pathType = Tcl_FSGetPathType(fsPathPtr->cwdPtr);
+	dir = Tcl_FSGetNormalizedPath(interp, fsPathPtr->cwdPtr);
+	if (dir == NULL) {
+	    return NULL;
+	}
+	/* TODO: Figure out why this is needed. */
+	if (pathPtr->bytes == NULL) {
+	    UpdateStringOfFsPath(pathPtr);
+	}
+
+	TclGetStringFromObj(fsPathPtr->normPathPtr, &tailLen);
+	if (tailLen) {
+	    copy = AppendPath(dir, fsPathPtr->normPathPtr);
+	} else {
+	    copy = Tcl_DuplicateObj(dir);
+	}
+	Tcl_IncrRefCount(dir);
+	Tcl_IncrRefCount(copy);
+
+	/*
+	 * We now own a reference on both 'dir' and 'copy'
+	 */
+
+	(void) TclGetStringFromObj(dir, &cwdLen);
+	cwdLen += (Tcl_GetString(copy)[cwdLen] == '/');
+
+	/* Normalize the combined string. */
+
+	if (PATHFLAGS(pathPtr) & TCLPATH_NEEDNORM) {
+	    /*
+	     * If the "tail" part has components (like /../) that cause the
+	     * combined path to need more complete normalizing, call on the
+	     * more powerful routine to accomplish that so we avoid [Bug
+	     * 2385549] ...
+	     */
+
+	    Tcl_Obj *newCopy = TclFSNormalizeAbsolutePath(interp, copy);
+
+	    Tcl_DecrRefCount(copy);
+	    copy = newCopy;
+	} else {
+	    /*
+	     * ... but in most cases where we join a trouble free tail to a
+	     * normalized head, we can more efficiently normalize the combined
+	     * path by passing over only the unnormalized tail portion. When
+	     * this is sufficient, prior developers claim this should be much
+	     * faster. We use 'cwdLen-1' so that we are already pointing at
+	     * the dir-separator that we know about. The normalization code
+	     * will actually start off directly after that separator.
+	     */
+
+	    TclFSNormalizeToUniquePath(interp, copy, cwdLen-1);
+	}
+
+	/* Now we need to construct the new path object. */
+
+	if (pathType == TCL_PATH_RELATIVE) {
+	    Tcl_Obj *origDir = fsPathPtr->cwdPtr;
+
+	    /*
+	     * NOTE: here we are (dangerously?) assuming that origDir points
+	     * to a Tcl_Obj with Tcl_ObjType == &tclFsPathType. The
+	     *     pathType = Tcl_FSGetPathType(fsPathPtr->cwdPtr);
+	     * above that set the pathType value should have established that,
+	     * but it's far less clear on what basis we know there's been no
+	     * shimmering since then.
+	     */
+
+	    FsPath *origDirFsPathPtr = PATHOBJ(origDir);
+
+	    fsPathPtr->cwdPtr = origDirFsPathPtr->cwdPtr;
+	    Tcl_IncrRefCount(fsPathPtr->cwdPtr);
+
+	    TclDecrRefCount(fsPathPtr->normPathPtr);
+	    fsPathPtr->normPathPtr = copy;
+
+	    /*
+	     * That's our reference to copy used.
+	     */
+
+	    TclDecrRefCount(dir);
+	    TclDecrRefCount(origDir);
+	} else {
+	    TclDecrRefCount(fsPathPtr->cwdPtr);
+	    fsPathPtr->cwdPtr = NULL;
+	    TclDecrRefCount(fsPathPtr->normPathPtr);
+	    fsPathPtr->normPathPtr = copy;
+
+	    /*
+	     * That's our reference to copy used.
+	     */
+
+	    TclDecrRefCount(dir);
+	}
+	PATHFLAGS(pathPtr) = 0;
+    }
+
+    /*
+     * Ensure cwd hasn't changed.
+     */
+
+    if (fsPathPtr->cwdPtr != NULL) {
+	if (!TclFSCwdPointerEquals(&fsPathPtr->cwdPtr)) {
+	    if (pathPtr->bytes == NULL) {
+		UpdateStringOfFsPath(pathPtr);
+	    }
+	    FreeFsPathInternalRep(pathPtr);
+	    if (SetFsPathFromAny(interp, pathPtr) != TCL_OK) {
+		return NULL;
+	    }
+	    fsPathPtr = PATHOBJ(pathPtr);
+	} else if (fsPathPtr->normPathPtr == NULL) {
+	    int cwdLen;
+	    Tcl_Obj *copy;
+
+	    copy = AppendPath(fsPathPtr->cwdPtr, pathPtr);
+
+	    (void) TclGetStringFromObj(fsPathPtr->cwdPtr, &cwdLen);
+	    cwdLen += (Tcl_GetString(copy)[cwdLen] == '/');
+
+	    /*
+	     * Normalize the combined string, but only starting after the end
+	     * of the previously normalized 'dir'. This should be much faster!
+	     */
+
+	    TclFSNormalizeToUniquePath(interp, copy, cwdLen-1);
+	    fsPathPtr->normPathPtr = copy;
+	    Tcl_IncrRefCount(fsPathPtr->normPathPtr);
+	}
+    }
+    if (fsPathPtr->normPathPtr == NULL) {
+	Tcl_Obj *useThisCwd = NULL;
+	int pureNormalized = 1;
+
+	/*
+	 * Since normPathPtr is NULL, but this is a valid path object, we know
+	 * that the translatedPathPtr cannot be NULL.
+	 */
+
+	Tcl_Obj *absolutePath = fsPathPtr->translatedPathPtr;
+	const char *path = TclGetString(absolutePath);
+
+	Tcl_IncrRefCount(absolutePath);
+
+	/*
+	 * We have to be a little bit careful here to avoid infinite loops
+	 * we're asking Tcl_FSGetPathType to return the path's type, but that
+	 * call can actually result in a lot of other filesystem action, which
+	 * might loop back through here.
+	 */
+
+	if (path[0] == '\0') {
+	    /*
+	     * Special handling for the empty string value. This one is very
+	     * weird with [file normalize {}] => {}. (The reasoning supporting
+	     * this is unknown to DGP, but he fears changing it.) Attempt here
+	     * to keep the expectations of other parts of Tcl_Filesystem code
+	     * about state of the FsPath fields satisfied.
+	     *
+	     * In particular, capture the cwd value and save so it can be
+	     * stored in the cwdPtr field below.
+	     */
+
+	    useThisCwd = Tcl_FSGetCwd(interp);
+	} else {
+	    /*
+	     * We don't ask for the type of 'pathPtr' here, because that is
+	     * not correct for our purposes when we have a path like '~'. Tcl
+	     * has a bit of a contradiction in that '~' paths are defined as
+	     * 'absolute', but in reality can be just about anything,
+	     * depending on how env(HOME) is set.
+	     */
+
+	    Tcl_PathType type = Tcl_FSGetPathType(absolutePath);
+
+	    if (type == TCL_PATH_RELATIVE) {
+		useThisCwd = Tcl_FSGetCwd(interp);
+
+		if (useThisCwd == NULL) {
+		    return NULL;
+		}
+
+		pureNormalized = 0;
+		Tcl_DecrRefCount(absolutePath);
+		absolutePath = Tcl_FSJoinToPath(useThisCwd, 1, &absolutePath);
+		Tcl_IncrRefCount(absolutePath);
+
+		/*
+		 * We have a refCount on the cwd.
+		 */
+#ifdef _WIN32
+	    } else if (type == TCL_PATH_VOLUME_RELATIVE) {
+		/*
+		 * Only Windows has volume-relative paths.
+		 */
+
+		Tcl_DecrRefCount(absolutePath);
+		absolutePath = TclWinVolumeRelativeNormalize(interp,
+			path, &useThisCwd);
+		if (absolutePath == NULL) {
+		    return NULL;
+		}
+		pureNormalized = 0;
+#endif /* _WIN32 */
+	    }
+	}
+
+	/*
+	 * Already has refCount incremented.
+	 */
+
+	fsPathPtr->normPathPtr = TclFSNormalizeAbsolutePath(interp,
+		absolutePath);
+
+	/*
+	 * Check if path is pure normalized (this can only be the case if it
+	 * is an absolute path).
+	 */
+
+	if (pureNormalized) {
+	    int normPathLen, pathLen;
+	    const char *normPath;
+
+	    path = TclGetStringFromObj(pathPtr, &pathLen);
+	    normPath = TclGetStringFromObj(fsPathPtr->normPathPtr, &normPathLen);
+	    if ((pathLen == normPathLen) && !memcmp(path, normPath, pathLen)) {
+		/*
+		 * The path was already normalized. Get rid of the duplicate.
+		 */
+
+		TclDecrRefCount(fsPathPtr->normPathPtr);
+
+		/*
+		 * We do *not* increment the refCount for this circular
+		 * reference.
+		 */
+
+		fsPathPtr->normPathPtr = pathPtr;
+	    }
+	}
+	if (useThisCwd != NULL) {
+	    /*
+	     * We just need to free an object we allocated above for relative
+	     * paths (this was returned by Tcl_FSJoinToPath above), and then
+	     * of course store the cwd.
+	     */
+
+	    fsPathPtr->cwdPtr = useThisCwd;
+	}
+	TclDecrRefCount(absolutePath);
+    }
+
+    return fsPathPtr->normPathPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSGetInternalRep --
+ *
+ *	Extract the internal representation of a given path object, in the
+ *	given filesystem. If the path object belongs to a different
+ *	filesystem, we return NULL.
+ *
+ *	If the internal representation is currently NULL, we attempt to
+ *	generate it, by calling the filesystem's
+ *	'Tcl_FSCreateInternalRepProc'.
+ *
+ * Results:
+ *	NULL or a valid internal representation.
+ *
+ * Side effects:
+ *	An attempt may be made to convert the object.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_FSGetInternalRep(
+    Tcl_Obj *pathPtr,
+    const Tcl_Filesystem *fsPtr)
+{
+    FsPath *srcFsPathPtr;
+
+    if (Tcl_FSConvertToPathType(NULL, pathPtr) != TCL_OK) {
+	return NULL;
+    }
+    srcFsPathPtr = PATHOBJ(pathPtr);
+
+    /*
+     * We will only return the native representation for the caller's
+     * filesystem. Otherwise we will simply return NULL. This means that there
+     * must be a unique bi-directional mapping between paths and filesystems,
+     * and that this mapping will not allow 'remapped' files -- files which
+     * are in one filesystem but mapped into another. Another way of putting
+     * this is that 'stacked' filesystems are not allowed. We recognise that
+     * this is a potentially useful feature for the future.
+     *
+     * Even something simple like a 'pass through' filesystem which logs all
+     * activity and passes the calls onto the native system would be nice, but
+     * not easily achievable with the current implementation.
+     */
+
+    if (srcFsPathPtr->fsPtr == NULL) {
+	/*
+	 * This only usually happens in wrappers like TclpStat which create a
+	 * string object and pass it to TclpObjStat. Code which calls the
+	 * Tcl_FS.. functions should always have a filesystem already set.
+	 * Whether this code path is legal or not depends on whether we decide
+	 * to allow external code to call the native filesystem directly. It
+	 * is at least safer to allow this sub-optimal routing.
+	 */
+
+	Tcl_FSGetFileSystemForPath(pathPtr);
+
+	/*
+	 * If we fail through here, then the path is probably not a valid path
+	 * in the filesystsem, and is most likely to be a use of the empty
+	 * path "" via a direct call to one of the objectified interfaces
+	 * (e.g. from the Tcl testsuite).
+	 */
+
+	srcFsPathPtr = PATHOBJ(pathPtr);
+	if (srcFsPathPtr->fsPtr == NULL) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * There is still one possibility we should consider; if the file belongs
+     * to a different filesystem, perhaps it is actually linked through to a
+     * file in our own filesystem which we do care about. The way we can check
+     * for this is we ask what filesystem this path belongs to.
+     */
+
+    if (fsPtr != srcFsPathPtr->fsPtr) {
+	const Tcl_Filesystem *actualFs = Tcl_FSGetFileSystemForPath(pathPtr);
+
+	if (actualFs == fsPtr) {
+	    return Tcl_FSGetInternalRep(pathPtr, fsPtr);
+	}
+	return NULL;
+    }
+
+    if (srcFsPathPtr->nativePathPtr == NULL) {
+	Tcl_FSCreateInternalRepProc *proc;
+	char *nativePathPtr;
+
+	proc = srcFsPathPtr->fsPtr->createInternalRepProc;
+	if (proc == NULL) {
+	    return NULL;
+	}
+
+	nativePathPtr = proc(pathPtr);
+	srcFsPathPtr = PATHOBJ(pathPtr);
+	srcFsPathPtr->nativePathPtr = nativePathPtr;
+    }
+
+    return srcFsPathPtr->nativePathPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFSEnsureEpochOk --
+ *
+ *	This will ensure the pathPtr is up to date and can be converted into a
+ *	"path" type, and that we are able to generate a complete normalized
+ *	path which is used to determine the filesystem match.
+ *
+ * Results:
+ *	Standard Tcl return code.
+ *
+ * Side effects:
+ *	An attempt may be made to convert the object.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclFSEnsureEpochOk(
+    Tcl_Obj *pathPtr,
+    const Tcl_Filesystem **fsPtrPtr)
+{
+    FsPath *srcFsPathPtr;
+
+    if (pathPtr->typePtr != &tclFsPathType) {
+	return TCL_OK;
+    }
+
+    srcFsPathPtr = PATHOBJ(pathPtr);
+
+    /*
+     * Check if the filesystem has changed in some way since this object's
+     * internal representation was calculated.
+     */
+
+    if (!TclFSEpochOk(srcFsPathPtr->filesystemEpoch)) {
+	/*
+	 * We have to discard the stale representation and recalculate it.
+	 */
+
+	if (pathPtr->bytes == NULL) {
+	    UpdateStringOfFsPath(pathPtr);
+	}
+	FreeFsPathInternalRep(pathPtr);
+	if (SetFsPathFromAny(NULL, pathPtr) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	srcFsPathPtr = PATHOBJ(pathPtr);
+    }
+
+    /*
+     * Check whether the object is already assigned to a fs.
+     */
+
+    if (srcFsPathPtr->fsPtr != NULL) {
+	*fsPtrPtr = srcFsPathPtr->fsPtr;
+    }
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclFSSetPathDetails --
+ *
+ *	???
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	???
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclFSSetPathDetails(
+    Tcl_Obj *pathPtr,
+    const Tcl_Filesystem *fsPtr,
+    ClientData clientData)
+{
+    FsPath *srcFsPathPtr;
+
+    /*
+     * Make sure pathPtr is of the correct type.
+     */
+
+    if (pathPtr->typePtr != &tclFsPathType) {
+	if (SetFsPathFromAny(NULL, pathPtr) != TCL_OK) {
+	    return;
+	}
+    }
+
+    srcFsPathPtr = PATHOBJ(pathPtr);
+    srcFsPathPtr->fsPtr = fsPtr;
+    srcFsPathPtr->nativePathPtr = clientData;
+    srcFsPathPtr->filesystemEpoch = TclFSEpoch();
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_FSEqualPaths --
+ *
+ *	This function tests whether the two paths given are equal path
+ *	objects. If either or both is NULL, 0 is always returned.
+ *
+ * Results:
+ *	1 or 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_FSEqualPaths(
+    Tcl_Obj *firstPtr,
+    Tcl_Obj *secondPtr)
+{
+    const char *firstStr, *secondStr;
+    int firstLen, secondLen, tempErrno;
+
+    if (firstPtr == secondPtr) {
+	return 1;
+    }
+
+    if (firstPtr == NULL || secondPtr == NULL) {
+	return 0;
+    }
+    firstStr = TclGetStringFromObj(firstPtr, &firstLen);
+    secondStr = TclGetStringFromObj(secondPtr, &secondLen);
+    if ((firstLen == secondLen) && !memcmp(firstStr, secondStr, firstLen)) {
+	return 1;
+    }
+
+    /*
+     * Try the most thorough, correct method of comparing fully normalized
+     * paths.
+     */
+
+    tempErrno = Tcl_GetErrno();
+    firstPtr = Tcl_FSGetNormalizedPath(NULL, firstPtr);
+    secondPtr = Tcl_FSGetNormalizedPath(NULL, secondPtr);
+    Tcl_SetErrno(tempErrno);
+
+    if (firstPtr == NULL || secondPtr == NULL) {
+	return 0;
+    }
+
+    firstStr = TclGetStringFromObj(firstPtr, &firstLen);
+    secondStr = TclGetStringFromObj(secondPtr, &secondLen);
+    return ((firstLen == secondLen) && !memcmp(firstStr, secondStr, firstLen));
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * SetFsPathFromAny --
+ *
+ *	This function tries to convert the given Tcl_Obj to a valid Tcl path
+ *	type.
+ *
+ *	The filename may begin with "~" (to indicate current user's home
+ *	directory) or "~<user>" (to indicate any user's home directory).
+ *
+ * Results:
+ *	Standard Tcl error code.
+ *
+ * Side effects:
+ *	The old representation may be freed, and new memory allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static int
+SetFsPathFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *pathPtr)		/* The object to convert. */
+{
+    int len;
+    FsPath *fsPathPtr;
+    Tcl_Obj *transPtr;
+    char *name;
+
+    if (pathPtr->typePtr == &tclFsPathType) {
+	return TCL_OK;
+    }
+
+    /*
+     * First step is to translate the filename. This is similar to
+     * Tcl_TranslateFilename, but shouldn't convert everything to windows
+     * backslashes on that platform. The current implementation of this piece
+     * is a slightly optimised version of the various Tilde/Split/Join stuff
+     * to avoid multiple split/join operations.
+     *
+     * We remove any trailing directory separator.
+     *
+     * However, the split/join routines are quite complex, and one has to make
+     * sure not to break anything on Unix or Win (fCmd.test, fileName.test and
+     * cmdAH.test exercise most of the code).
+     */
+
+    name = TclGetStringFromObj(pathPtr, &len);
+
+    /*
+     * Handle tilde substitutions, if needed.
+     */
+
+    if (name[0] == '~') {
+	Tcl_DString temp;
+	int split;
+	char separator = '/';
+
+	split = FindSplitPos(name, separator);
+	if (split != len) {
+	    /*
+	     * We have multiple pieces '~user/foo/bar...'
+	     */
+
+	    name[split] = '\0';
+	}
+
+	/*
+	 * Do some tilde substitution.
+	 */
+
+	if (name[1] == '\0') {
+	    /*
+	     * We have just '~'
+	     */
+
+	    const char *dir;
+	    Tcl_DString dirString;
+
+	    if (split != len) {
+		name[split] = separator;
+	    }
+
+	    dir = TclGetEnv("HOME", &dirString);
+	    if (dir == NULL) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "couldn't find HOME environment variable to"
+			    " expand path", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "PATH",
+			    "HOMELESS", NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    Tcl_DStringInit(&temp);
+	    Tcl_JoinPath(1, &dir, &temp);
+	    Tcl_DStringFree(&dirString);
+	} else {
+	    /*
+	     * We have a user name '~user'
+	     */
+
+	    Tcl_DStringInit(&temp);
+	    if (TclpGetUserHome(name+1, &temp) == NULL) {
+		if (interp != NULL) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "user \"%s\" doesn't exist", name+1));
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "PATH", "NOUSER",
+			    NULL);
+		}
+		Tcl_DStringFree(&temp);
+		if (split != len) {
+		    name[split] = separator;
+		}
+		return TCL_ERROR;
+	    }
+	    if (split != len) {
+		name[split] = separator;
+	    }
+	}
+
+	transPtr = TclDStringToObj(&temp);
+
+	if (split != len) {
+	    /*
+	     * Join up the tilde substitution with the rest.
+	     */
+
+	    if (name[split+1] == separator) {
+		/*
+		 * Somewhat tricky case like ~//foo/bar. Make use of
+		 * Split/Join machinery to get it right. Assumes all paths
+		 * beginning with ~ are part of the native filesystem.
+		 */
+
+		int objc;
+		Tcl_Obj **objv;
+		Tcl_Obj *parts = TclpNativeSplitPath(pathPtr, NULL);
+
+		Tcl_ListObjGetElements(NULL, parts, &objc, &objv);
+
+		/*
+		 * Skip '~'. It's replaced by its expansion.
+		 */
+
+		objc--; objv++;
+		while (objc--) {
+		    TclpNativeJoinPath(transPtr, Tcl_GetString(*objv++));
+		}
+		TclDecrRefCount(parts);
+	    } else {
+		Tcl_Obj *pair[2];
+
+		pair[0] = transPtr;
+		pair[1] = Tcl_NewStringObj(name+split+1, -1);
+		transPtr = TclJoinPath(2, pair);
+		Tcl_DecrRefCount(pair[0]);
+		Tcl_DecrRefCount(pair[1]);
+	    }
+	}
+    } else {
+	transPtr = TclJoinPath(1, &pathPtr);
+    }
+
+    /*
+     * Now we have a translated filename in 'transPtr'. This will have forward
+     * slashes on Windows, and will not contain any ~user sequences.
+     */
+
+    fsPathPtr = ckalloc(sizeof(FsPath));
+
+    fsPathPtr->translatedPathPtr = transPtr;
+    if (transPtr != pathPtr) {
+	Tcl_IncrRefCount(fsPathPtr->translatedPathPtr);
+	/* Redo translation when $env(HOME) changes */
+	fsPathPtr->filesystemEpoch = TclFSEpoch();
+    } else {
+	fsPathPtr->filesystemEpoch = 0;
+    }
+    fsPathPtr->normPathPtr = NULL;
+    fsPathPtr->cwdPtr = NULL;
+    fsPathPtr->nativePathPtr = NULL;
+    fsPathPtr->fsPtr = NULL;
+
+    /*
+     * Free old representation before installing our new one.
+     */
+
+    TclFreeIntRep(pathPtr);
+    SETPATHOBJ(pathPtr, fsPathPtr);
+    PATHFLAGS(pathPtr) = 0;
+    pathPtr->typePtr = &tclFsPathType;
+    return TCL_OK;
+}
+
+static void
+FreeFsPathInternalRep(
+    Tcl_Obj *pathPtr)		/* Path object with internal rep to free. */
+{
+    FsPath *fsPathPtr = PATHOBJ(pathPtr);
+
+    if (fsPathPtr->translatedPathPtr != NULL) {
+	if (fsPathPtr->translatedPathPtr != pathPtr) {
+	    TclDecrRefCount(fsPathPtr->translatedPathPtr);
+	}
+    }
+    if (fsPathPtr->normPathPtr != NULL) {
+	if (fsPathPtr->normPathPtr != pathPtr) {
+	    TclDecrRefCount(fsPathPtr->normPathPtr);
+	}
+	fsPathPtr->normPathPtr = NULL;
+    }
+    if (fsPathPtr->cwdPtr != NULL) {
+	TclDecrRefCount(fsPathPtr->cwdPtr);
+    }
+    if (fsPathPtr->nativePathPtr != NULL && fsPathPtr->fsPtr != NULL) {
+	Tcl_FSFreeInternalRepProc *freeProc =
+		fsPathPtr->fsPtr->freeInternalRepProc;
+
+	if (freeProc != NULL) {
+	    freeProc(fsPathPtr->nativePathPtr);
+	    fsPathPtr->nativePathPtr = NULL;
+	}
+    }
+
+    ckfree(fsPathPtr);
+    pathPtr->typePtr = NULL;
+}
+
+static void
+DupFsPathInternalRep(
+    Tcl_Obj *srcPtr,		/* Path obj with internal rep to copy. */
+    Tcl_Obj *copyPtr)		/* Path obj with internal rep to set. */
+{
+    FsPath *srcFsPathPtr = PATHOBJ(srcPtr);
+    FsPath *copyFsPathPtr = ckalloc(sizeof(FsPath));
+
+    SETPATHOBJ(copyPtr, copyFsPathPtr);
+
+    if (srcFsPathPtr->translatedPathPtr == srcPtr) {
+	/* Cycle in src -> make cycle in copy. */
+	copyFsPathPtr->translatedPathPtr = copyPtr;
+    } else {
+	copyFsPathPtr->translatedPathPtr = srcFsPathPtr->translatedPathPtr;
+	if (copyFsPathPtr->translatedPathPtr != NULL) {
+	    Tcl_IncrRefCount(copyFsPathPtr->translatedPathPtr);
+	}
+    }
+
+    if (srcFsPathPtr->normPathPtr == srcPtr) {
+	/* Cycle in src -> make cycle in copy. */
+	copyFsPathPtr->normPathPtr = copyPtr;
+    } else {
+	copyFsPathPtr->normPathPtr = srcFsPathPtr->normPathPtr;
+	if (copyFsPathPtr->normPathPtr != NULL) {
+	    Tcl_IncrRefCount(copyFsPathPtr->normPathPtr);
+	}
+    }
+
+    copyFsPathPtr->cwdPtr = srcFsPathPtr->cwdPtr;
+    if (copyFsPathPtr->cwdPtr != NULL) {
+	Tcl_IncrRefCount(copyFsPathPtr->cwdPtr);
+    }
+
+    copyFsPathPtr->flags = srcFsPathPtr->flags;
+
+    if (srcFsPathPtr->fsPtr != NULL
+	    && srcFsPathPtr->nativePathPtr != NULL) {
+	Tcl_FSDupInternalRepProc *dupProc =
+		srcFsPathPtr->fsPtr->dupInternalRepProc;
+
+	if (dupProc != NULL) {
+	    copyFsPathPtr->nativePathPtr =
+		    dupProc(srcFsPathPtr->nativePathPtr);
+	} else {
+	    copyFsPathPtr->nativePathPtr = NULL;
+	}
+    } else {
+	copyFsPathPtr->nativePathPtr = NULL;
+    }
+    copyFsPathPtr->fsPtr = srcFsPathPtr->fsPtr;
+    copyFsPathPtr->filesystemEpoch = srcFsPathPtr->filesystemEpoch;
+
+    copyPtr->typePtr = &tclFsPathType;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * UpdateStringOfFsPath --
+ *
+ *	Gives an object a valid string rep.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory may be allocated.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfFsPath(
+    register Tcl_Obj *pathPtr)	/* path obj with string rep to update. */
+{
+    FsPath *fsPathPtr = PATHOBJ(pathPtr);
+    int cwdLen;
+    Tcl_Obj *copy;
+
+    if (PATHFLAGS(pathPtr) == 0 || fsPathPtr->cwdPtr == NULL) {
+	Tcl_Panic("Called UpdateStringOfFsPath with invalid object");
+    }
+
+    copy = AppendPath(fsPathPtr->cwdPtr, fsPathPtr->normPathPtr);
+
+    pathPtr->bytes = TclGetStringFromObj(copy, &cwdLen);
+    pathPtr->length = cwdLen;
+    copy->bytes = &tclEmptyString;
+    copy->length = 0;
+    TclDecrRefCount(copy);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclNativePathInFilesystem --
+ *
+ *	Any path object is acceptable to the native filesystem, by default (we
+ *	will throw errors when illegal paths are actually tried to be used).
+ *
+ *	However, this behavior means the native filesystem must be the last
+ *	filesystem in the lookup list (otherwise it will claim all files
+ *	belong to it, and other filesystems will never get a look in).
+ *
+ * Results:
+ *	TCL_OK, to indicate 'yes', -1 to indicate no.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclNativePathInFilesystem(
+    Tcl_Obj *pathPtr,
+    ClientData *clientDataPtr)
+{
+    /*
+     * A special case is required to handle the empty path "". This is a valid
+     * path (i.e. the user should be able to do 'file exists ""' without
+     * throwing an error), but equally the path doesn't exist. Those are the
+     * semantics of Tcl (at present anyway), so we have to abide by them here.
+     */
+
+    if (pathPtr->typePtr == &tclFsPathType) {
+	if (pathPtr->bytes != NULL && pathPtr->bytes[0] == '\0') {
+	    /*
+	     * We reject the empty path "".
+	     */
+
+	    return -1;
+	}
+
+	/*
+	 * Otherwise there is no way this path can be empty.
+	 */
+    } else {
+	/*
+	 * It is somewhat unusual to reach this code path without the object
+	 * being of tclFsPathType. However, we do our best to deal with the
+	 * situation.
+	 */
+
+	int len;
+
+	(void) TclGetStringFromObj(pathPtr, &len);
+	if (len == 0) {
+	    /*
+	     * We reject the empty path "".
+	     */
+
+	    return -1;
+	}
+    }
+
+    /*
+     * Path is of correct type, or is of non-zero length, so we accept it.
+     */
+
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclPipe.c b/generic/tclPipe.c
new file mode 100644
index 0000000..b679ec4
--- /dev/null
+++ b/generic/tclPipe.c
@@ -0,0 +1,1141 @@
+/*
+ * tclPipe.c --
+ *
+ *	This file contains the generic portion of the command channel driver
+ *	as well as various utility routines used in managing subprocesses.
+ *
+ * Copyright (c) 1997 by Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * A linked list of the following structures is used to keep track of child
+ * processes that have been detached but haven't exited yet, so we can make
+ * sure that they're properly "reaped" (officially waited for) and don't lie
+ * around as zombies cluttering the system.
+ */
+
+typedef struct Detached {
+    Tcl_Pid pid;		/* Id of process that's been detached but
+				 * isn't known to have exited. */
+    struct Detached *nextPtr;	/* Next in list of all detached processes. */
+} Detached;
+
+static Detached *detList = NULL;/* List of all detached proceses. */
+TCL_DECLARE_MUTEX(pipeMutex)	/* Guard access to detList. */
+
+/*
+ * Declarations for local functions defined in this file:
+ */
+
+static TclFile		FileForRedirect(Tcl_Interp *interp, const char *spec,
+			    int atOk, const char *arg, const char *nextArg,
+			    int flags, int *skipPtr, int *closePtr,
+			    int *releasePtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FileForRedirect --
+ *
+ *	This function does much of the work of parsing redirection operators.
+ *	It handles "@" if specified and allowed, and a file name, and opens
+ *	the file if necessary.
+ *
+ * Results:
+ *	The return value is the descriptor number for the file. If an error
+ *	occurs then NULL is returned and an error message is left in the
+ *	interp's result. Several arguments are side-effected; see the argument
+ *	list below for details.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static TclFile
+FileForRedirect(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    const char *spec,		/* Points to character just after redirection
+				 * character. */
+    int atOK,			/* Non-zero means that '@' notation can be
+				 * used to specify a channel, zero means that
+				 * it isn't. */
+    const char *arg,		/* Pointer to entire argument containing spec:
+				 * used for error reporting. */
+    const char *nextArg,	/* Next argument in argc/argv array, if needed
+				 * for file name or channel name. May be
+				 * NULL. */
+    int flags,			/* Flags to use for opening file or to specify
+				 * mode for channel. */
+    int *skipPtr,		/* Filled with 1 if redirection target was in
+				 * spec, 2 if it was in nextArg. */
+    int *closePtr,		/* Filled with one if the caller should close
+				 * the file when done with it, zero
+				 * otherwise. */
+    int *releasePtr)
+{
+    int writing = (flags & O_WRONLY);
+    Tcl_Channel chan;
+    TclFile file;
+
+    *skipPtr = 1;
+    if ((atOK != 0) && (*spec == '@')) {
+	spec++;
+	if (*spec == '\0') {
+	    spec = nextArg;
+	    if (spec == NULL) {
+		goto badLastArg;
+	    }
+	    *skipPtr = 2;
+	}
+	chan = Tcl_GetChannel(interp, spec, NULL);
+	if (chan == (Tcl_Channel) NULL) {
+	    return NULL;
+	}
+	file = TclpMakeFile(chan, writing ? TCL_WRITABLE : TCL_READABLE);
+	if (file == NULL) {
+	    Tcl_Obj *msg;
+
+	    Tcl_GetChannelError(chan, &msg);
+	    if (msg) {
+		Tcl_SetObjResult(interp, msg);
+	    } else {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"channel \"%s\" wasn't opened for %s",
+			Tcl_GetChannelName(chan),
+			((writing) ? "writing" : "reading")));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+			"BADCHAN", NULL);
+	    }
+	    return NULL;
+	}
+	*releasePtr = 1;
+	if (writing) {
+	    /*
+	     * Be sure to flush output to the file, so that anything written
+	     * by the child appears after stuff we've already written.
+	     */
+
+	    Tcl_Flush(chan);
+	}
+    } else {
+	const char *name;
+	Tcl_DString nameString;
+
+	if (*spec == '\0') {
+	    spec = nextArg;
+	    if (spec == NULL) {
+		goto badLastArg;
+	    }
+	    *skipPtr = 2;
+	}
+	name = Tcl_TranslateFileName(interp, spec, &nameString);
+	if (name == NULL) {
+	    return NULL;
+	}
+	file = TclpOpenFile(name, flags);
+	Tcl_DStringFree(&nameString);
+	if (file == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "couldn't %s file \"%s\": %s",
+		    (writing ? "write" : "read"), spec,
+		    Tcl_PosixError(interp)));
+	    return NULL;
+	}
+	*closePtr = 1;
+    }
+    return file;
+
+  badLastArg:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "can't specify \"%s\" as last word in command", arg));
+    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC", "SYNTAX", NULL);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DetachPids --
+ *
+ *	This function is called to indicate that one or more child processes
+ *	have been placed in background and will never be waited for; they
+ *	should eventually be reaped by Tcl_ReapDetachedProcs.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DetachPids(
+    int numPids,		/* Number of pids to detach: gives size of
+				 * array pointed to by pidPtr. */
+    Tcl_Pid *pidPtr)		/* Array of pids to detach. */
+{
+    register Detached *detPtr;
+    int i;
+
+    Tcl_MutexLock(&pipeMutex);
+    for (i = 0; i < numPids; i++) {
+	detPtr = ckalloc(sizeof(Detached));
+	detPtr->pid = pidPtr[i];
+	detPtr->nextPtr = detList;
+	detList = detPtr;
+    }
+    Tcl_MutexUnlock(&pipeMutex);
+
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ReapDetachedProcs --
+ *
+ *	This function checks to see if any detached processes have exited and,
+ *	if so, it "reaps" them by officially waiting on them. It should be
+ *	called "occasionally" to make sure that all detached processes are
+ *	eventually reaped.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Processes are waited on, so that they can be reaped by the system.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ReapDetachedProcs(void)
+{
+    register Detached *detPtr;
+    Detached *nextPtr, *prevPtr;
+    int status;
+    Tcl_Pid pid;
+
+    Tcl_MutexLock(&pipeMutex);
+    for (detPtr = detList, prevPtr = NULL; detPtr != NULL; ) {
+	pid = Tcl_WaitPid(detPtr->pid, &status, WNOHANG);
+	if ((pid == 0) || ((pid == (Tcl_Pid) -1) && (errno != ECHILD))) {
+	    prevPtr = detPtr;
+	    detPtr = detPtr->nextPtr;
+	    continue;
+	}
+	nextPtr = detPtr->nextPtr;
+	if (prevPtr == NULL) {
+	    detList = detPtr->nextPtr;
+	} else {
+	    prevPtr->nextPtr = detPtr->nextPtr;
+	}
+	ckfree(detPtr);
+	detPtr = nextPtr;
+    }
+    Tcl_MutexUnlock(&pipeMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCleanupChildren --
+ *
+ *	This is a utility function used to wait for child processes to exit,
+ *	record information about abnormal exits, and then collect any stderr
+ *	output generated by them.
+ *
+ * Results:
+ *	The return value is a standard Tcl result. If anything at weird
+ *	happened with the child processes, TCL_ERROR is returned and a message
+ *	is left in the interp's result.
+ *
+ * Side effects:
+ *	If the last character of the interp's result is a newline, then it is
+ *	removed unless keepNewline is non-zero. File errorId gets closed, and
+ *	pidPtr is freed back to the storage allocator.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCleanupChildren(
+    Tcl_Interp *interp,		/* Used for error messages. */
+    int numPids,		/* Number of entries in pidPtr array. */
+    Tcl_Pid *pidPtr,		/* Array of process ids of children. */
+    Tcl_Channel errorChan)	/* Channel for file containing stderr output
+				 * from pipeline. NULL means there isn't any
+				 * stderr output. */
+{
+    int result = TCL_OK;
+    int i, abnormalExit, anyErrorInfo;
+    Tcl_Pid pid;
+    int waitStatus;
+    const char *msg;
+    unsigned long resolvedPid;
+
+    abnormalExit = 0;
+    for (i = 0; i < numPids; i++) {
+	/*
+	 * We need to get the resolved pid before we wait on it as the windows
+	 * implementation of Tcl_WaitPid deletes the information such that any
+	 * following calls to TclpGetPid fail.
+	 */
+
+	resolvedPid = TclpGetPid(pidPtr[i]);
+	pid = Tcl_WaitPid(pidPtr[i], &waitStatus, 0);
+	if (pid == (Tcl_Pid) -1) {
+	    result = TCL_ERROR;
+	    if (interp != NULL) {
+		msg = Tcl_PosixError(interp);
+		if (errno == ECHILD) {
+		    /*
+		     * This changeup in message suggested by Mark Diekhans to
+		     * remind people that ECHILD errors can occur on some
+		     * systems if SIGCHLD isn't in its default state.
+		     */
+
+		    msg =
+			"child process lost (is SIGCHLD ignored or trapped?)";
+		}
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error waiting for process to exit: %s", msg));
+	    }
+	    continue;
+	}
+
+	/*
+	 * Create error messages for unusual process exits. An extra newline
+	 * gets appended to each error message, but it gets removed below (in
+	 * the same fashion that an extra newline in the command's output is
+	 * removed).
+	 */
+
+	if (!WIFEXITED(waitStatus) || (WEXITSTATUS(waitStatus) != 0)) {
+	    char msg1[TCL_INTEGER_SPACE], msg2[TCL_INTEGER_SPACE];
+
+	    result = TCL_ERROR;
+	    sprintf(msg1, "%lu", resolvedPid);
+	    if (WIFEXITED(waitStatus)) {
+		if (interp != NULL) {
+		    sprintf(msg2, "%u", WEXITSTATUS(waitStatus));
+		    Tcl_SetErrorCode(interp, "CHILDSTATUS", msg1, msg2, NULL);
+		}
+		abnormalExit = 1;
+	    } else if (interp != NULL) {
+		const char *p;
+
+		if (WIFSIGNALED(waitStatus)) {
+		    p = Tcl_SignalMsg(WTERMSIG(waitStatus));
+		    Tcl_SetErrorCode(interp, "CHILDKILLED", msg1,
+			    Tcl_SignalId(WTERMSIG(waitStatus)), p, NULL);
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "child killed: %s\n", p));
+		} else if (WIFSTOPPED(waitStatus)) {
+		    p = Tcl_SignalMsg(WSTOPSIG(waitStatus));
+		    Tcl_SetErrorCode(interp, "CHILDSUSP", msg1,
+			    Tcl_SignalId(WSTOPSIG(waitStatus)), p, NULL);
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "child suspended: %s\n", p));
+		} else {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "child wait status didn't make sense\n", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+			    "ODDWAITRESULT", msg1, NULL);
+		}
+	    }
+	}
+    }
+
+    /*
+     * Read the standard error file. If there's anything there, then return an
+     * error and add the file's contents to the result string.
+     */
+
+    anyErrorInfo = 0;
+    if (errorChan != NULL) {
+	/*
+	 * Make sure we start at the beginning of the file.
+	 */
+
+	if (interp != NULL) {
+	    int count;
+	    Tcl_Obj *objPtr;
+
+	    Tcl_Seek(errorChan, (Tcl_WideInt)0, SEEK_SET);
+	    objPtr = Tcl_NewObj();
+	    count = Tcl_ReadChars(errorChan, objPtr, -1, 0);
+	    if (count < 0) {
+		result = TCL_ERROR;
+		Tcl_DecrRefCount(objPtr);
+		Tcl_ResetResult(interp);
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"error reading stderr output file: %s",
+			Tcl_PosixError(interp)));
+	    } else if (count > 0) {
+		anyErrorInfo = 1;
+		Tcl_SetObjResult(interp, objPtr);
+		result = TCL_ERROR;
+	    } else {
+		Tcl_DecrRefCount(objPtr);
+	    }
+	}
+	Tcl_Close(NULL, errorChan);
+    }
+
+    /*
+     * If a child exited abnormally but didn't output any error information at
+     * all, generate an error message here.
+     */
+
+    if ((abnormalExit != 0) && (anyErrorInfo == 0) && (interp != NULL)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"child process exited abnormally", -1));
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreatePipeline --
+ *
+ *	Given an argc/argv array, instantiate a pipeline of processes as
+ *	described by the argv.
+ *
+ *	This function is unofficially exported for use by BLT.
+ *
+ * Results:
+ *	The return value is a count of the number of new processes created, or
+ *	-1 if an error occurred while creating the pipeline. *pidArrayPtr is
+ *	filled in with the address of a dynamically allocated array giving the
+ *	ids of all of the processes. It is up to the caller to free this array
+ *	when it isn't needed anymore. If inPipePtr is non-NULL, *inPipePtr is
+ *	filled in with the file id for the input pipe for the pipeline (if
+ *	any): the caller must eventually close this file. If outPipePtr isn't
+ *	NULL, then *outPipePtr is filled in with the file id for the output
+ *	pipe from the pipeline: the caller must close this file. If errFilePtr
+ *	isn't NULL, then *errFilePtr is filled with a file id that may be used
+ *	to read error output after the pipeline completes.
+ *
+ * Side effects:
+ *	Processes and pipes are created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCreatePipeline(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    int argc,			/* Number of entries in argv. */
+    const char **argv,		/* Array of strings describing commands in
+				 * pipeline plus I/O redirection with <, <<,
+				 * >, etc. Argv[argc] must be NULL. */
+    Tcl_Pid **pidArrayPtr,	/* Word at *pidArrayPtr gets filled in with
+				 * address of array of pids for processes in
+				 * pipeline (first pid is first process in
+				 * pipeline). */
+    TclFile *inPipePtr,		/* If non-NULL, input to the pipeline comes
+				 * from a pipe (unless overridden by
+				 * redirection in the command). The file id
+				 * with which to write to this pipe is stored
+				 * at *inPipePtr. NULL means command specified
+				 * its own input source. */
+    TclFile *outPipePtr,	/* If non-NULL, output to the pipeline goes to
+				 * a pipe, unless overriden by redirection in
+				 * the command. The file id with which to read
+				 * frome this pipe is stored at *outPipePtr.
+				 * NULL means command specified its own output
+				 * sink. */
+    TclFile *errFilePtr)	/* If non-NULL, all stderr output from the
+				 * pipeline will go to a temporary file
+				 * created here, and a descriptor to read the
+				 * file will be left at *errFilePtr. The file
+				 * will be removed already, so closing this
+				 * descriptor will be the end of the file. If
+				 * this is NULL, then all stderr output goes
+				 * to our stderr. If the pipeline specifies
+				 * redirection then the file will still be
+				 * created but it will never get any data. */
+{
+    Tcl_Pid *pidPtr = NULL;	/* Points to malloc-ed array holding all the
+				 * pids of child processes. */
+    int numPids;		/* Actual number of processes that exist at
+				 * *pidPtr right now. */
+    int cmdCount;		/* Count of number of distinct commands found
+				 * in argc/argv. */
+    const char *inputLiteral = NULL;
+				/* If non-null, then this points to a string
+				 * containing input data (specified via <<) to
+				 * be piped to the first process in the
+				 * pipeline. */
+    TclFile inputFile = NULL;	/* If != NULL, gives file to use as input for
+				 * first process in pipeline (specified via <
+				 * or <@). */
+    int inputClose = 0;		/* If non-zero, then inputFile should be
+				 * closed when cleaning up. */
+    int inputRelease = 0;
+    TclFile outputFile = NULL;	/* Writable file for output from last command
+				 * in pipeline (could be file or pipe). NULL
+				 * means use stdout. */
+    int outputClose = 0;	/* If non-zero, then outputFile should be
+				 * closed when cleaning up. */
+    int outputRelease = 0;
+    TclFile errorFile = NULL;	/* Writable file for error output from all
+				 * commands in pipeline. NULL means use
+				 * stderr. */
+    int errorClose = 0;		/* If non-zero, then errorFile should be
+				 * closed when cleaning up. */
+    int errorRelease = 0;
+    const char *p;
+    const char *nextArg;
+    int skip, lastBar, lastArg, i, j, atOK, flags, needCmd, errorToOutput = 0;
+    Tcl_DString execBuffer;
+    TclFile pipeIn;
+    TclFile curInFile, curOutFile, curErrFile;
+    Tcl_Channel channel;
+
+    if (inPipePtr != NULL) {
+	*inPipePtr = NULL;
+    }
+    if (outPipePtr != NULL) {
+	*outPipePtr = NULL;
+    }
+    if (errFilePtr != NULL) {
+	*errFilePtr = NULL;
+    }
+
+    Tcl_DStringInit(&execBuffer);
+
+    pipeIn = NULL;
+    curInFile = NULL;
+    curOutFile = NULL;
+    numPids = 0;
+
+    /*
+     * First, scan through all the arguments to figure out the structure of
+     * the pipeline. Process all of the input and output redirection arguments
+     * and remove them from the argument list in the pipeline. Count the
+     * number of distinct processes (it's the number of "|" arguments plus
+     * one) but don't remove the "|" arguments because they'll be used in the
+     * second pass to seperate the individual child processes. Cannot start
+     * the child processes in this pass because the redirection symbols may
+     * appear anywhere in the command line - e.g., the '<' that specifies the
+     * input to the entire pipe may appear at the very end of the argument
+     * list.
+     */
+
+    lastBar = -1;
+    cmdCount = 1;
+    needCmd = 1;
+    for (i = 0; i < argc; i++) {
+	errorToOutput = 0;
+	skip = 0;
+	p = argv[i];
+	switch (*p++) {
+	case '|':
+	    if (*p == '&') {
+		p++;
+	    }
+	    if (*p == '\0') {
+		if ((i == (lastBar + 1)) || (i == (argc - 1))) {
+		    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			    "illegal use of | or |& in command", -1));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+			    "PIPESYNTAX", NULL);
+		    goto error;
+		}
+	    }
+	    lastBar = i;
+	    cmdCount++;
+	    needCmd = 1;
+	    break;
+
+	case '<':
+	    if (inputClose != 0) {
+		inputClose = 0;
+		TclpCloseFile(inputFile);
+	    }
+	    if (inputRelease != 0) {
+		inputRelease = 0;
+		TclpReleaseFile(inputFile);
+	    }
+	    if (*p == '<') {
+		inputFile = NULL;
+		inputLiteral = p + 1;
+		skip = 1;
+		if (*inputLiteral == '\0') {
+		    inputLiteral = ((i + 1) == argc) ? NULL : argv[i + 1];
+		    if (inputLiteral == NULL) {
+			Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+				"can't specify \"%s\" as last word in command",
+				argv[i]));
+			Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+				"PIPESYNTAX", NULL);
+			goto error;
+		    }
+		    skip = 2;
+		}
+	    } else {
+		nextArg = ((i + 1) == argc) ? NULL : argv[i + 1];
+		inputLiteral = NULL;
+		inputFile = FileForRedirect(interp, p, 1, argv[i], nextArg,
+			O_RDONLY, &skip, &inputClose, &inputRelease);
+		if (inputFile == NULL) {
+		    goto error;
+		}
+	    }
+	    break;
+
+	case '>':
+	    atOK = 1;
+	    flags = O_WRONLY | O_CREAT | O_TRUNC;
+	    if (*p == '>') {
+		p++;
+		atOK = 0;
+
+		/*
+		 * Note that the O_APPEND flag only has an effect on POSIX
+		 * platforms. On Windows, we just have to carry on regardless.
+		 */
+
+		flags = O_WRONLY | O_CREAT | O_APPEND;
+	    }
+	    if (*p == '&') {
+		if (errorClose != 0) {
+		    errorClose = 0;
+		    TclpCloseFile(errorFile);
+		}
+		errorToOutput = 1;
+		p++;
+	    }
+
+	    /*
+	     * Close the old output file, but only if the error file is not
+	     * also using it.
+	     */
+
+	    if (outputClose != 0) {
+		outputClose = 0;
+		if (errorFile == outputFile) {
+		    errorClose = 1;
+		} else {
+		    TclpCloseFile(outputFile);
+		}
+	    }
+	    if (outputRelease != 0) {
+		outputRelease = 0;
+		if (errorFile == outputFile) {
+		    errorRelease = 1;
+		} else {
+		    TclpReleaseFile(outputFile);
+		}
+	    }
+	    nextArg = ((i + 1) == argc) ? NULL : argv[i + 1];
+	    outputFile = FileForRedirect(interp, p, atOK, argv[i], nextArg,
+		    flags, &skip, &outputClose, &outputRelease);
+	    if (outputFile == NULL) {
+		goto error;
+	    }
+	    if (errorToOutput) {
+		if (errorClose != 0) {
+		    errorClose = 0;
+		    TclpCloseFile(errorFile);
+		}
+		if (errorRelease != 0) {
+		    errorRelease = 0;
+		    TclpReleaseFile(errorFile);
+		}
+		errorFile = outputFile;
+	    }
+	    break;
+
+	case '2':
+	    if (*p != '>') {
+		break;
+	    }
+	    p++;
+	    atOK = 1;
+	    flags = O_WRONLY | O_CREAT | O_TRUNC;
+	    if (*p == '>') {
+		p++;
+		atOK = 0;
+
+		/*
+		 * Note that the O_APPEND flag only has an effect on POSIX
+		 * platforms. On Windows, we just have to carry on regardless.
+		 */
+
+		flags = O_WRONLY | O_CREAT | O_APPEND;
+	    }
+	    if (errorClose != 0) {
+		errorClose = 0;
+		TclpCloseFile(errorFile);
+	    }
+	    if (errorRelease != 0) {
+		errorRelease = 0;
+		TclpReleaseFile(errorFile);
+	    }
+	    if (atOK && p[0] == '@' && p[1] == '1' && p[2] == '\0') {
+		/*
+		 * Special case handling of 2>@1 to redirect stderr to the
+		 * exec/open output pipe as well. This is meant for the end of
+		 * the command string, otherwise use |& between commands.
+		 */
+
+		if (i != argc-1) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "must specify \"%s\" as last word in command",
+			    argv[i]));
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+			    "PIPESYNTAX", NULL);
+		    goto error;
+		}
+		errorFile = outputFile;
+		errorToOutput = 2;
+		skip = 1;
+	    } else {
+		nextArg = ((i + 1) == argc) ? NULL : argv[i + 1];
+		errorFile = FileForRedirect(interp, p, atOK, argv[i],
+			nextArg, flags, &skip, &errorClose, &errorRelease);
+		if (errorFile == NULL) {
+		    goto error;
+		}
+	    }
+	    break;
+
+	default:
+	    /*
+	     * Got a command word, not a redirection.
+	     */
+
+	    needCmd = 0;
+	    break;
+	}
+
+	if (skip != 0) {
+	    for (j = i + skip; j < argc; j++) {
+		argv[j - skip] = argv[j];
+	    }
+	    argc -= skip;
+	    i -= 1;
+	}
+    }
+
+    if (needCmd) {
+	/*
+	 * We had a bar followed only by redirections.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"illegal use of | or |& in command", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC", "PIPESYNTAX",
+		NULL);
+	goto error;
+    }
+
+    if (inputFile == NULL) {
+	if (inputLiteral != NULL) {
+	    /*
+	     * The input for the first process is immediate data coming from
+	     * Tcl. Create a temporary file for it and put the data into the
+	     * file.
+	     */
+
+	    inputFile = TclpCreateTempFile(inputLiteral);
+	    if (inputFile == NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"couldn't create input file for command: %s",
+			Tcl_PosixError(interp)));
+		goto error;
+	    }
+	    inputClose = 1;
+	} else if (inPipePtr != NULL) {
+	    /*
+	     * The input for the first process in the pipeline is to come from
+	     * a pipe that can be written from by the caller.
+	     */
+
+	    if (TclpCreatePipe(&inputFile, inPipePtr) == 0) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"couldn't create input pipe for command: %s",
+			Tcl_PosixError(interp)));
+		goto error;
+	    }
+	    inputClose = 1;
+	} else {
+	    /*
+	     * The input for the first process comes from stdin.
+	     */
+
+	    channel = Tcl_GetStdChannel(TCL_STDIN);
+	    if (channel != NULL) {
+		inputFile = TclpMakeFile(channel, TCL_READABLE);
+		if (inputFile != NULL) {
+		    inputRelease = 1;
+		}
+	    }
+	}
+    }
+
+    if (outputFile == NULL) {
+	if (outPipePtr != NULL) {
+	    /*
+	     * Output from the last process in the pipeline is to go to a pipe
+	     * that can be read by the caller.
+	     */
+
+	    if (TclpCreatePipe(outPipePtr, &outputFile) == 0) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"couldn't create output pipe for command: %s",
+			Tcl_PosixError(interp)));
+		goto error;
+	    }
+	    outputClose = 1;
+	} else {
+	    /*
+	     * The output for the last process goes to stdout.
+	     */
+
+	    channel = Tcl_GetStdChannel(TCL_STDOUT);
+	    if (channel) {
+		outputFile = TclpMakeFile(channel, TCL_WRITABLE);
+		if (outputFile != NULL) {
+		    outputRelease = 1;
+		}
+	    }
+	}
+    }
+
+    if (errorFile == NULL) {
+	if (errorToOutput == 2) {
+	    /*
+	     * Handle 2>@1 special case at end of cmd line.
+	     */
+
+	    errorFile = outputFile;
+	} else if (errFilePtr != NULL) {
+	    /*
+	     * Set up the standard error output sink for the pipeline, if
+	     * requested. Use a temporary file which is opened, then deleted.
+	     * Could potentially just use pipe, but if it filled up it could
+	     * cause the pipeline to deadlock: we'd be waiting for processes
+	     * to complete before reading stderr, and processes couldn't
+	     * complete because stderr was backed up.
+	     */
+
+	    errorFile = TclpCreateTempFile(NULL);
+	    if (errorFile == NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"couldn't create error file for command: %s",
+			Tcl_PosixError(interp)));
+		goto error;
+	    }
+	    *errFilePtr = errorFile;
+	} else {
+	    /*
+	     * Errors from the pipeline go to stderr.
+	     */
+
+	    channel = Tcl_GetStdChannel(TCL_STDERR);
+	    if (channel) {
+		errorFile = TclpMakeFile(channel, TCL_WRITABLE);
+		if (errorFile != NULL) {
+		    errorRelease = 1;
+		}
+	    }
+	}
+    }
+
+    /*
+     * Scan through the argc array, creating a process for each group of
+     * arguments between the "|" characters.
+     */
+
+    Tcl_ReapDetachedProcs();
+    pidPtr = ckalloc(cmdCount * sizeof(Tcl_Pid));
+
+    curInFile = inputFile;
+
+    for (i = 0; i < argc; i = lastArg + 1) {
+	int result, joinThisError;
+	Tcl_Pid pid;
+	const char *oldName;
+
+	/*
+	 * Convert the program name into native form.
+	 */
+
+	if (Tcl_TranslateFileName(interp, argv[i], &execBuffer) == NULL) {
+	    goto error;
+	}
+
+	/*
+	 * Find the end of the current segment of the pipeline.
+	 */
+
+	joinThisError = 0;
+	for (lastArg = i; lastArg < argc; lastArg++) {
+	    if (argv[lastArg][0] != '|') {
+		continue;
+	    }
+	    if (argv[lastArg][1] == '\0') {
+		break;
+	    }
+	    if ((argv[lastArg][1] == '&') && (argv[lastArg][2] == '\0')) {
+		joinThisError = 1;
+		break;
+	    }
+	}
+
+	/*
+	 * If this is the last segment, use the specified outputFile.
+	 * Otherwise create an intermediate pipe. pipeIn will become the
+	 * curInFile for the next segment of the pipe.
+	 */
+
+	if (lastArg == argc) {
+	    curOutFile = outputFile;
+	} else {
+	    argv[lastArg] = NULL;
+	    if (TclpCreatePipe(&pipeIn, &curOutFile) == 0) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"couldn't create pipe: %s", Tcl_PosixError(interp)));
+		goto error;
+	    }
+	}
+
+	if (joinThisError != 0) {
+	    curErrFile = curOutFile;
+	} else {
+	    curErrFile = errorFile;
+	}
+
+	/*
+	 * Restore argv[i], since a caller wouldn't expect the contents of
+	 * argv to be modified.
+	 */
+
+	oldName = argv[i];
+	argv[i] = Tcl_DStringValue(&execBuffer);
+	result = TclpCreateProcess(interp, lastArg - i, argv + i,
+		curInFile, curOutFile, curErrFile, &pid);
+	argv[i] = oldName;
+	if (result != TCL_OK) {
+	    goto error;
+	}
+	Tcl_DStringFree(&execBuffer);
+
+	pidPtr[numPids] = pid;
+	numPids++;
+
+	/*
+	 * Close off our copies of file descriptors that were set up for this
+	 * child, then set up the input for the next child.
+	 */
+
+	if ((curInFile != NULL) && (curInFile != inputFile)) {
+	    TclpCloseFile(curInFile);
+	}
+	curInFile = pipeIn;
+	pipeIn = NULL;
+
+	if ((curOutFile != NULL) && (curOutFile != outputFile)) {
+	    TclpCloseFile(curOutFile);
+	}
+	curOutFile = NULL;
+    }
+
+    *pidArrayPtr = pidPtr;
+
+    /*
+     * All done. Cleanup open files lying around and then return.
+     */
+
+  cleanup:
+    Tcl_DStringFree(&execBuffer);
+
+    if (inputClose) {
+	TclpCloseFile(inputFile);
+    } else if (inputRelease) {
+	TclpReleaseFile(inputFile);
+    }
+    if (outputClose) {
+	TclpCloseFile(outputFile);
+    } else if (outputRelease) {
+	TclpReleaseFile(outputFile);
+    }
+    if (errorClose) {
+	TclpCloseFile(errorFile);
+    } else if (errorRelease) {
+	TclpReleaseFile(errorFile);
+    }
+    return numPids;
+
+    /*
+     * An error occurred. There could have been extra files open, such as
+     * pipes between children. Clean them all up. Detach any child processes
+     * that have been created.
+     */
+
+  error:
+    if (pipeIn != NULL) {
+	TclpCloseFile(pipeIn);
+    }
+    if ((curOutFile != NULL) && (curOutFile != outputFile)) {
+	TclpCloseFile(curOutFile);
+    }
+    if ((curInFile != NULL) && (curInFile != inputFile)) {
+	TclpCloseFile(curInFile);
+    }
+    if ((inPipePtr != NULL) && (*inPipePtr != NULL)) {
+	TclpCloseFile(*inPipePtr);
+	*inPipePtr = NULL;
+    }
+    if ((outPipePtr != NULL) && (*outPipePtr != NULL)) {
+	TclpCloseFile(*outPipePtr);
+	*outPipePtr = NULL;
+    }
+    if ((errFilePtr != NULL) && (*errFilePtr != NULL)) {
+	TclpCloseFile(*errFilePtr);
+	*errFilePtr = NULL;
+    }
+    if (pidPtr != NULL) {
+	for (i = 0; i < numPids; i++) {
+	    if (pidPtr[i] != (Tcl_Pid) -1) {
+		Tcl_DetachPids(1, &pidPtr[i]);
+	    }
+	}
+	ckfree(pidPtr);
+    }
+    numPids = -1;
+    goto cleanup;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_OpenCommandChannel --
+ *
+ *	Opens an I/O channel to one or more subprocesses specified by argc and
+ *	argv. The flags argument determines the disposition of the stdio
+ *	handles. If the TCL_STDIN flag is set then the standard input for the
+ *	first subprocess will be tied to the channel: writing to the channel
+ *	will provide input to the subprocess. If TCL_STDIN is not set, then
+ *	standard input for the first subprocess will be the same as this
+ *	application's standard input. If TCL_STDOUT is set then standard
+ *	output from the last subprocess can be read from the channel;
+ *	otherwise it goes to this application's standard output. If TCL_STDERR
+ *	is set, standard error output for all subprocesses is returned to the
+ *	channel and results in an error when the channel is closed; otherwise
+ *	it goes to this application's standard error. If TCL_ENFORCE_MODE is
+ *	not set, then argc and argv can redirect the stdio handles to override
+ *	TCL_STDIN, TCL_STDOUT, and TCL_STDERR; if it is set, then it is an
+ *	error for argc and argv to override stdio channels for which
+ *	TCL_STDIN, TCL_STDOUT, and TCL_STDERR have been set.
+ *
+ * Results:
+ *	A new command channel, or NULL on failure with an error message left
+ *	in interp.
+ *
+ * Side effects:
+ *	Creates processes, opens pipes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Channel
+Tcl_OpenCommandChannel(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. Can NOT be
+				 * NULL. */
+    int argc,			/* How many arguments. */
+    const char **argv,		/* Array of arguments for command pipe. */
+    int flags)			/* Or'ed combination of TCL_STDIN, TCL_STDOUT,
+				 * TCL_STDERR, and TCL_ENFORCE_MODE. */
+{
+    TclFile *inPipePtr, *outPipePtr, *errFilePtr;
+    TclFile inPipe, outPipe, errFile;
+    int numPids;
+    Tcl_Pid *pidPtr;
+    Tcl_Channel channel;
+
+    inPipe = outPipe = errFile = NULL;
+
+    inPipePtr = (flags & TCL_STDIN) ? &inPipe : NULL;
+    outPipePtr = (flags & TCL_STDOUT) ? &outPipe : NULL;
+    errFilePtr = (flags & TCL_STDERR) ? &errFile : NULL;
+
+    numPids = TclCreatePipeline(interp, argc, argv, &pidPtr, inPipePtr,
+	    outPipePtr, errFilePtr);
+
+    if (numPids < 0) {
+	goto error;
+    }
+
+    /*
+     * Verify that the pipes that were created satisfy the readable/writable
+     * constraints.
+     */
+
+    if (flags & TCL_ENFORCE_MODE) {
+	if ((flags & TCL_STDOUT) && (outPipe == NULL)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "can't read output from command:"
+		    " standard output was redirected", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+		    "BADREDIRECT", NULL);
+	    goto error;
+	}
+	if ((flags & TCL_STDIN) && (inPipe == NULL)) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "can't write input to command:"
+		    " standard input was redirected", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC",
+		    "BADREDIRECT", NULL);
+	    goto error;
+	}
+    }
+
+    channel = TclpCreateCommandChannel(outPipe, inPipe, errFile,
+	    numPids, pidPtr);
+
+    if (channel == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"pipe for command could not be created", -1));
+	Tcl_SetErrorCode(interp, "TCL", "OPERATION", "EXEC", "NOPIPE", NULL);
+	goto error;
+    }
+    return channel;
+
+  error:
+    if (numPids > 0) {
+	Tcl_DetachPids(numPids, pidPtr);
+	ckfree(pidPtr);
+    }
+    if (inPipe != NULL) {
+	TclpCloseFile(inPipe);
+    }
+    if (outPipe != NULL) {
+	TclpCloseFile(outPipe);
+    }
+    if (errFile != NULL) {
+	TclpCloseFile(errFile);
+    }
+    return NULL;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclPkg.c b/generic/tclPkg.c
new file mode 100644
index 0000000..eb4dc9b
--- /dev/null
+++ b/generic/tclPkg.c
@@ -0,0 +1,2043 @@
+/*
+ * tclPkg.c --
+ *
+ *	This file implements package and version control for Tcl via the
+ *	"package" command and a few C APIs.
+ *
+ * Copyright (c) 1996 Sun Microsystems, Inc.
+ * Copyright (c) 2006 Andreas Kupries <andreas_kupries@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ *
+ * TIP #268.
+ * Heavily rewritten to handle the extend version numbers, and extended
+ * package requirements.
+ */
+
+#include "tclInt.h"
+
+MODULE_SCOPE char *tclEmptyStringRep;
+
+char *tclEmptyStringRep = &tclEmptyString;
+
+/*
+ * Each invocation of the "package ifneeded" command creates a structure of
+ * the following type, which is used to load the package into the interpreter
+ * if it is requested with a "package require" command.
+ */
+
+typedef struct PkgAvail {
+    char *version;		/* Version string; malloc'ed. */
+    char *script;		/* Script to invoke to provide this version of
+				 * the package. Malloc'ed and protected by
+				 * Tcl_Preserve and Tcl_Release. */
+    char *pkgIndex;		/* Full file name of pkgIndex file */
+    struct PkgAvail *nextPtr;	/* Next in list of available versions of the
+				 * same package. */
+} PkgAvail;
+
+typedef struct PkgName {
+    struct PkgName *nextPtr;	/* Next in list of package names being initialized. */
+    char name[1];
+} PkgName;
+
+typedef struct PkgFiles {
+    PkgName *names;		/* Package names being initialized. Must be first field*/
+    Tcl_HashTable table;	/* Table which contains files for each package */
+} PkgFiles;
+
+
+/*
+ * For each package that is known in any way to an interpreter, there is one
+ * record of the following type. These records are stored in the
+ * "packageTable" hash table in the interpreter, keyed by package name such as
+ * "Tk" (no version number).
+ */
+
+typedef struct Package {
+    char *version;		/* Version that has been supplied in this
+				 * interpreter via "package provide"
+				 * (malloc'ed). NULL means the package doesn't
+				 * exist in this interpreter yet. */
+    PkgAvail *availPtr;		/* First in list of all available versions of
+				 * this package. */
+    const void *clientData;	/* Client data. */
+} Package;
+
+/*
+ * Prototypes for functions defined in this file:
+ */
+
+static int		CheckVersionAndConvert(Tcl_Interp *interp,
+			    const char *string, char **internal, int *stable);
+static int		CompareVersions(char *v1i, char *v2i,
+			    int *isMajorPtr);
+static int		CheckRequirement(Tcl_Interp *interp,
+			    const char *string);
+static int		CheckAllRequirements(Tcl_Interp *interp, int reqc,
+			    Tcl_Obj *const reqv[]);
+static int		RequirementSatisfied(char *havei, const char *req);
+static int		SomeRequirementSatisfied(char *havei, int reqc,
+			    Tcl_Obj *const reqv[]);
+static void		AddRequirementsToResult(Tcl_Interp *interp, int reqc,
+			    Tcl_Obj *const reqv[]);
+static void		AddRequirementsToDString(Tcl_DString *dstring,
+			    int reqc, Tcl_Obj *const reqv[]);
+static Package *	FindPackage(Tcl_Interp *interp, const char *name);
+static const char *	PkgRequireCore(Tcl_Interp *interp, const char *name,
+			    int reqc, Tcl_Obj *const reqv[],
+			    void *clientDataPtr);
+
+/*
+ * Helper macros.
+ */
+
+#define DupBlock(v,s,len) \
+    ((v) = ckalloc(len), memcpy((v),(s),(len)))
+#define DupString(v,s) \
+    do { \
+	size_t local__len = strlen(s) + 1; \
+	DupBlock((v),(s),local__len); \
+    } while (0)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PkgProvide / Tcl_PkgProvideEx --
+ *
+ *	This function is invoked to declare that a particular version of a
+ *	particular package is now present in an interpreter. There must not be
+ *	any other version of this package already provided in the interpreter.
+ *
+ * Results:
+ *	Normally returns TCL_OK; if there is already another version of the
+ *	package loaded then TCL_ERROR is returned and an error message is left
+ *	in the interp's result.
+ *
+ * Side effects:
+ *	The interpreter remembers that this package is available, so that no
+ *	other version of the package may be provided for the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_PkgProvide
+int
+Tcl_PkgProvide(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of package. */
+    const char *version)	/* Version string for package. */
+{
+    return Tcl_PkgProvideEx(interp, name, version, NULL);
+}
+
+int
+Tcl_PkgProvideEx(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of package. */
+    const char *version,	/* Version string for package. */
+    const void *clientData)	/* clientdata for this package (normally used
+				 * for C callback function table) */
+{
+    Package *pkgPtr;
+    char *pvi, *vi;
+    int res;
+
+    pkgPtr = FindPackage(interp, name);
+    if (pkgPtr->version == NULL) {
+	DupString(pkgPtr->version, version);
+	pkgPtr->clientData = clientData;
+	return TCL_OK;
+    }
+
+    if (CheckVersionAndConvert(interp, pkgPtr->version, &pvi,
+	    NULL) != TCL_OK) {
+	return TCL_ERROR;
+    } else if (CheckVersionAndConvert(interp, version, &vi, NULL) != TCL_OK) {
+	ckfree(pvi);
+	return TCL_ERROR;
+    }
+
+    res = CompareVersions(pvi, vi, NULL);
+    ckfree(pvi);
+    ckfree(vi);
+
+    if (res == 0) {
+	if (clientData != NULL) {
+	    pkgPtr->clientData = clientData;
+	}
+	return TCL_OK;
+    }
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "conflicting versions provided for package \"%s\": %s, then %s",
+	    name, pkgPtr->version, version));
+    Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "VERSIONCONFLICT", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PkgRequire / Tcl_PkgRequireEx / Tcl_PkgRequireProc --
+ *
+ *	This function is called by code that depends on a particular version
+ *	of a particular package. If the package is not already provided in the
+ *	interpreter, this function invokes a Tcl script to provide it. If the
+ *	package is already provided, this function makes sure that the
+ *	caller's needs don't conflict with the version that is present.
+ *
+ * Results:
+ *	If successful, returns the version string for the currently provided
+ *	version of the package, which may be different from the "version"
+ *	argument. If the caller's requirements cannot be met (e.g. the version
+ *	requested conflicts with a currently provided version, or the required
+ *	version cannot be found, or the script to provide the required version
+ *	generates an error), NULL is returned and an error message is left in
+ *	the interp's result.
+ *
+ * Side effects:
+ *	The script from some previous "package ifneeded" command may be
+ *	invoked to provide the package.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void PkgFilesCleanupProc(ClientData clientData,
+    			    Tcl_Interp *interp)
+{
+    PkgFiles *pkgFiles = (PkgFiles *) clientData;
+    Tcl_HashSearch search;
+    Tcl_HashEntry *entry;
+
+    while (pkgFiles->names) {
+	PkgName *name = pkgFiles->names;
+	pkgFiles->names = name->nextPtr;
+	ckfree(name);
+    }
+    entry = Tcl_FirstHashEntry(&pkgFiles->table, &search);
+    while (entry) {
+	Tcl_Obj *obj = (Tcl_Obj *)Tcl_GetHashValue(entry);
+	Tcl_DecrRefCount(obj);
+	entry = Tcl_NextHashEntry(&search);
+    }
+    Tcl_DeleteHashTable(&pkgFiles->table);
+    ckfree(pkgFiles);
+    return;
+}
+
+void *TclInitPkgFiles(Tcl_Interp *interp)
+{
+    /* If assocdata "tclPkgFiles" doesn't exist yet, create it */
+    PkgFiles *pkgFiles = Tcl_GetAssocData(interp, "tclPkgFiles", NULL);
+    if (!pkgFiles) {
+	pkgFiles = ckalloc(sizeof(PkgFiles));
+	pkgFiles->names = NULL;
+	Tcl_InitHashTable(&pkgFiles->table, TCL_STRING_KEYS);
+	Tcl_SetAssocData(interp, "tclPkgFiles", PkgFilesCleanupProc, pkgFiles);
+    }
+    return pkgFiles;
+}
+
+void TclPkgFileSeen(Tcl_Interp *interp, const char *fileName)
+{
+    PkgFiles *pkgFiles = (PkgFiles *) Tcl_GetAssocData(interp, "tclPkgFiles", NULL);
+    if (pkgFiles && pkgFiles->names) {
+	const char *name = pkgFiles->names->name;
+	Tcl_HashTable *table = &pkgFiles->table;
+	int new;
+	Tcl_HashEntry *entry = Tcl_CreateHashEntry(table, name, &new);
+	Tcl_Obj *list;
+
+	if (new) {
+	    list = Tcl_NewObj();
+	    Tcl_SetHashValue(entry, list);
+	    Tcl_IncrRefCount(list);
+	} else {
+	    list = Tcl_GetHashValue(entry);
+	}
+	Tcl_ListObjAppendElement(interp, list, Tcl_NewStringObj(fileName, -1));
+    }
+}
+
+#undef Tcl_PkgRequire
+const char *
+Tcl_PkgRequire(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    const char *version,	/* Version string for desired version; NULL
+				 * means use the latest version available. */
+    int exact)			/* Non-zero means that only the particular
+				 * version given is acceptable. Zero means use
+				 * the latest compatible version. */
+{
+    return Tcl_PkgRequireEx(interp, name, version, exact, NULL);
+}
+
+const char *
+Tcl_PkgRequireEx(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    const char *version,	/* Version string for desired version; NULL
+				 * means use the latest version available. */
+    int exact,			/* Non-zero means that only the particular
+				 * version given is acceptable. Zero means use
+				 * the latest compatible version. */
+    void *clientDataPtr)	/* Used to return the client data for this
+				 * package. If it is NULL then the client data
+				 * is not returned. This is unchanged if this
+				 * call fails for any reason. */
+{
+    Tcl_Obj *ov;
+    const char *result = NULL;
+
+    /*
+     * If an attempt is being made to load this into a standalone executable
+     * on a platform where backlinking is not supported then this must be a
+     * shared version of Tcl (Otherwise the load would have failed). Detect
+     * this situation by checking that this library has been correctly
+     * initialised. If it has not been then return immediately as nothing will
+     * work.
+     */
+
+    if (tclEmptyStringRep == NULL) {
+	/*
+	 * OK, so what's going on here?
+	 *
+	 * First, what are we doing? We are performing a check on behalf of
+	 * one particular caller, Tcl_InitStubs(). When a package is stub-
+	 * enabled, it is statically linked to libtclstub.a, which contains a
+	 * copy of Tcl_InitStubs(). When a stub-enabled package is loaded, its
+	 * *_Init() function is supposed to call Tcl_InitStubs() before
+	 * calling any other functions in the Tcl library. The first Tcl
+	 * function called by Tcl_InitStubs() through the stub table is
+	 * Tcl_PkgRequireEx(), so this code right here is the first code that
+	 * is part of the original Tcl library in the executable that gets
+	 * executed on behalf of a newly loaded stub-enabled package.
+	 *
+	 * One easy error for the developer/builder of a stub-enabled package
+	 * to make is to forget to define USE_TCL_STUBS when compiling the
+	 * package. When that happens, the package will contain symbols that
+	 * are references to the Tcl library, rather than function pointers
+	 * referencing the stub table. On platforms that lack backlinking,
+	 * those unresolved references may cause the loading of the package to
+	 * also load a second copy of the Tcl library, leading to all kinds of
+	 * trouble. We would like to catch that error and report a useful
+	 * message back to the user. That's what we're doing.
+	 *
+	 * Second, how does this work? If we reach this point, then the global
+	 * variable tclEmptyStringRep has the value NULL. Compare that with
+	 * the definition of tclEmptyStringRep near the top of this file.
+	 * It clearly should not have the value NULL; it
+	 * should point to the char tclEmptyString. If we see it having the
+	 * value NULL, then somehow we are seeing a Tcl library that isn't
+	 * completely initialized, and that's an indicator for the error
+	 * condition described above. (Further explanation is welcome.)
+	 *
+	 * Third, so what do we do about it? This situation indicates the
+	 * package we just loaded wasn't properly compiled to be stub-enabled,
+	 * yet it thinks it is stub-enabled (it called Tcl_InitStubs()). We
+	 * want to report that the package just loaded is broken, so we want
+	 * to place an error message in the interpreter result and return NULL
+	 * to indicate failure to Tcl_InitStubs() so that it will also fail.
+	 * (Further explanation why we don't want to Tcl_Panic() is welcome.
+	 * After all, two Tcl libraries can't be a good thing!)
+	 *
+	 * Trouble is that's going to be tricky. We're now using a Tcl library
+	 * that's not fully initialized. Functions in it may not work
+	 * reliably, so be very careful about adding any other calls here
+	 * without checking how they behave when initialization is incomplete.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"Cannot load package \"%s\" in standalone executable:"
+		" This package is not compiled with stub support", name));
+	Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "UNSTUBBED", NULL);
+	return NULL;
+    }
+
+    /*
+     * Translate between old and new API, and defer to the new function.
+     */
+
+    if (version == NULL) {
+	result = PkgRequireCore(interp, name, 0, NULL, clientDataPtr);
+    } else {
+	if (exact && TCL_OK
+		!= CheckVersionAndConvert(interp, version, NULL, NULL)) {
+	    return NULL;
+	}
+	ov = Tcl_NewStringObj(version, -1);
+	if (exact) {
+	    Tcl_AppendStringsToObj(ov, "-", version, NULL);
+	}
+	Tcl_IncrRefCount(ov);
+	result = PkgRequireCore(interp, name, 1, &ov, clientDataPtr);
+	TclDecrRefCount(ov);
+    }
+
+    return result;
+}
+
+int
+Tcl_PkgRequireProc(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    int reqc,			/* Requirements constraining the desired
+				 * version. */
+    Tcl_Obj *const reqv[],	/* 0 means to use the latest version
+				 * available. */
+    void *clientDataPtr)
+{
+    const char *result =
+	    PkgRequireCore(interp, name, reqc, reqv, clientDataPtr);
+
+    if (result == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(result, -1));
+    return TCL_OK;
+}
+
+static const char *
+PkgRequireCore(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    int reqc,			/* Requirements constraining the desired
+				 * version. */
+    Tcl_Obj *const reqv[],	/* 0 means to use the latest version
+				 * available. */
+    void *clientDataPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    Package *pkgPtr;
+    PkgAvail *availPtr, *bestPtr, *bestStablePtr;
+    char *availVersion, *bestVersion;
+				/* Internal rep. of versions */
+    int availStable, code, satisfies, pass;
+    char *script, *pkgVersionI;
+    Tcl_DString command;
+
+    if (TCL_OK != CheckAllRequirements(interp, reqc, reqv)) {
+	return NULL;
+    }
+
+    /*
+     * It can take up to three passes to find the package: one pass to run the
+     * "package unknown" script, one to run the "package ifneeded" script for
+     * a specific version, and a final pass to lookup the package loaded by
+     * the "package ifneeded" script.
+     */
+
+    for (pass=1 ;; pass++) {
+	pkgPtr = FindPackage(interp, name);
+	if (pkgPtr->version != NULL) {
+	    break;
+	}
+
+	/*
+	 * Check whether we're already attempting to load some version of this
+	 * package (circular dependency detection).
+	 */
+
+	if (pkgPtr->clientData != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "circular package dependency:"
+		    " attempt to provide %s %s requires %s",
+		    name, (char *) pkgPtr->clientData, name));
+	    AddRequirementsToResult(interp, reqc, reqv);
+	    Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "CIRCULARITY", NULL);
+	    return NULL;
+	}
+
+	/*
+	 * The package isn't yet present. Search the list of available
+	 * versions and invoke the script for the best available version. We
+	 * are actually locating the best, and the best stable version. One of
+	 * them is then chosen based on the selection mode.
+	 */
+
+	bestPtr = NULL;
+	bestStablePtr = NULL;
+	bestVersion = NULL;
+
+	for (availPtr = pkgPtr->availPtr; availPtr != NULL;
+		availPtr = availPtr->nextPtr) {
+	    if (CheckVersionAndConvert(interp, availPtr->version,
+		    &availVersion, &availStable) != TCL_OK) {
+		/*
+		 * The provided version number has invalid syntax. This
+		 * should not happen. This should have been caught by the
+		 * 'package ifneeded' registering the package.
+		 */
+
+		continue;
+	    }
+
+	    if (bestPtr != NULL) {
+		int res = CompareVersions(availVersion, bestVersion, NULL);
+
+		/*
+		 * Note: Use internal reps!
+		 */
+
+		if (res <= 0) {
+		    /*
+		     * The version of the package sought is not as good as the
+		     * currently selected version. Ignore it.
+		     */
+
+		    ckfree(availVersion);
+		    availVersion = NULL;
+		    continue;
+		}
+	    }
+
+	    /*
+	     * We have found a version which is better than our max.
+	     */
+
+	    if (reqc > 0) {
+		/* Check satisfaction of requirements. */
+
+		satisfies = SomeRequirementSatisfied(availVersion, reqc, reqv);
+		if (!satisfies) {
+		    ckfree(availVersion);
+		    availVersion = NULL;
+		    continue;
+		}
+	    }
+
+	    bestPtr = availPtr;
+
+	    if (bestVersion != NULL) {
+		ckfree(bestVersion);
+	    }
+	    bestVersion = availVersion;
+
+	    /*
+	     * If this new best version is stable then it also has to be
+	     * better than the max stable version found so far.
+	     */
+
+	    if (availStable) {
+		bestStablePtr = availPtr;
+	    }
+	}
+
+	if (bestVersion != NULL) {
+	    ckfree(bestVersion);
+	}
+
+	/*
+	 * Now choose a version among the two best. For 'latest' we simply
+	 * take (actually keep) the best. For 'stable' we take the best
+	 * stable, if there is any, or the best if there is nothing stable.
+	 */
+
+	if ((iPtr->packagePrefer == PKG_PREFER_STABLE)
+		&& (bestStablePtr != NULL)) {
+	    bestPtr = bestStablePtr;
+	}
+
+	if (bestPtr != NULL) {
+	    /*
+	     * We found an ifneeded script for the package. Be careful while
+	     * executing it: this could cause reentrancy, so (a) protect the
+	     * script itself from deletion and (b) don't assume that bestPtr
+	     * will still exist when the script completes.
+	     */
+
+	    char *versionToProvide = bestPtr->version;
+	    PkgFiles *pkgFiles;
+	    PkgName *pkgName;
+	    script = bestPtr->script;
+
+	    pkgPtr->clientData = versionToProvide;
+	    Tcl_Preserve(versionToProvide);
+	    Tcl_Preserve(script);
+	    pkgFiles = TclInitPkgFiles(interp);
+	    /* Push "ifneeded" package name in "tclPkgFiles" assocdata. */
+	    pkgName = ckalloc(sizeof(PkgName) + strlen(name));
+	    pkgName->nextPtr = pkgFiles->names;
+	    strcpy(pkgName->name, name);
+	    pkgFiles->names = pkgName;
+	    if (bestPtr->pkgIndex) {
+		TclPkgFileSeen(interp, bestPtr->pkgIndex);
+	    }
+	    code = Tcl_EvalEx(interp, script, -1, TCL_EVAL_GLOBAL);
+	    /* Pop the "ifneeded" package name from "tclPkgFiles" assocdata*/
+	    pkgFiles->names = pkgName->nextPtr;
+	    ckfree(pkgName);
+	    Tcl_Release(script);
+
+	    pkgPtr = FindPackage(interp, name);
+	    if (code == TCL_OK) {
+		Tcl_ResetResult(interp);
+		if (pkgPtr->version == NULL) {
+		    code = TCL_ERROR;
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "attempt to provide package %s %s failed:"
+			    " no version of package %s provided",
+			    name, versionToProvide, name));
+		    Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "UNPROVIDED",
+			    NULL);
+		} else {
+		    char *pvi, *vi;
+
+		    if (CheckVersionAndConvert(interp, pkgPtr->version, &pvi,
+			    NULL) != TCL_OK) {
+			code = TCL_ERROR;
+		    } else if (CheckVersionAndConvert(interp,
+			    versionToProvide, &vi, NULL) != TCL_OK) {
+			ckfree(pvi);
+			code = TCL_ERROR;
+		    } else {
+			int res = CompareVersions(pvi, vi, NULL);
+
+			ckfree(pvi);
+			ckfree(vi);
+			if (res != 0) {
+			    code = TCL_ERROR;
+			    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+				    "attempt to provide package %s %s failed:"
+				    " package %s %s provided instead",
+				    name, versionToProvide,
+				    name, pkgPtr->version));
+			    Tcl_SetErrorCode(interp, "TCL", "PACKAGE",
+				    "WRONGPROVIDE", NULL);
+			}
+		    }
+		}
+	    } else if (code != TCL_ERROR) {
+		Tcl_Obj *codePtr = Tcl_NewIntObj(code);
+
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"attempt to provide package %s %s failed:"
+			" bad return code: %s",
+			name, versionToProvide, TclGetString(codePtr)));
+		Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "BADRESULT", NULL);
+		TclDecrRefCount(codePtr);
+		code = TCL_ERROR;
+	    }
+
+	    if (code == TCL_ERROR) {
+		Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+			"\n    (\"package ifneeded %s %s\" script)",
+			name, versionToProvide));
+	    }
+	    Tcl_Release(versionToProvide);
+
+	    if (code != TCL_OK) {
+		/*
+		 * Take a non-TCL_OK code from the script as an indication the
+		 * package wasn't loaded properly, so the package system
+		 * should not remember an improper load.
+		 *
+		 * This is consistent with our returning NULL. If we're not
+		 * willing to tell our caller we got a particular version, we
+		 * shouldn't store that version for telling future callers
+		 * either.
+		 */
+
+		if (pkgPtr->version != NULL) {
+		    ckfree(pkgPtr->version);
+		    pkgPtr->version = NULL;
+		}
+		pkgPtr->clientData = NULL;
+		return NULL;
+	    }
+
+	    break;
+	}
+
+	/*
+	 * The package is not in the database. If there is a "package unknown"
+	 * command, invoke it (but only on the first pass; after that, we
+	 * should not get here in the first place).
+	 */
+
+	if (pass > 1) {
+	    break;
+	}
+
+	script = ((Interp *) interp)->packageUnknown;
+	if (script != NULL) {
+	    Tcl_DStringInit(&command);
+	    Tcl_DStringAppend(&command, script, -1);
+	    Tcl_DStringAppendElement(&command, name);
+	    AddRequirementsToDString(&command, reqc, reqv);
+
+	    code = Tcl_EvalEx(interp, Tcl_DStringValue(&command),
+		    Tcl_DStringLength(&command), TCL_EVAL_GLOBAL);
+	    Tcl_DStringFree(&command);
+
+	    if ((code != TCL_OK) && (code != TCL_ERROR)) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"bad return code: %d", code));
+		Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "BADRESULT", NULL);
+		code = TCL_ERROR;
+	    }
+	    if (code == TCL_ERROR) {
+		Tcl_AddErrorInfo(interp,
+			"\n    (\"package unknown\" script)");
+		return NULL;
+	    }
+	    Tcl_ResetResult(interp);
+	}
+    }
+
+    if (pkgPtr->version == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't find package %s", name));
+	Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "UNFOUND", NULL);
+	AddRequirementsToResult(interp, reqc, reqv);
+	return NULL;
+    }
+
+    /*
+     * At this point we know that the package is present. Make sure that the
+     * provided version meets the current requirements.
+     */
+
+    if (reqc != 0) {
+	CheckVersionAndConvert(interp, pkgPtr->version, &pkgVersionI, NULL);
+	satisfies = SomeRequirementSatisfied(pkgVersionI, reqc, reqv);
+
+	ckfree(pkgVersionI);
+
+	if (!satisfies) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "version conflict for package \"%s\": have %s, need",
+		    name, pkgPtr->version));
+	    Tcl_SetErrorCode(interp, "TCL", "PACKAGE", "VERSIONCONFLICT",
+		    NULL);
+	    AddRequirementsToResult(interp, reqc, reqv);
+	    return NULL;
+	}
+    }
+
+    if (clientDataPtr) {
+	const void **ptr = (const void **) clientDataPtr;
+
+	*ptr = pkgPtr->clientData;
+    }
+    return pkgPtr->version;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PkgPresent / Tcl_PkgPresentEx --
+ *
+ *	Checks to see whether the specified package is present. If it is not
+ *	then no additional action is taken.
+ *
+ * Results:
+ *	If successful, returns the version string for the currently provided
+ *	version of the package, which may be different from the "version"
+ *	argument. If the caller's requirements cannot be met (e.g. the version
+ *	requested conflicts with a currently provided version), NULL is
+ *	returned and an error message is left in interp->result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_PkgPresent
+const char *
+Tcl_PkgPresent(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    const char *version,	/* Version string for desired version; NULL
+				 * means use the latest version available. */
+    int exact)			/* Non-zero means that only the particular
+				 * version given is acceptable. Zero means use
+				 * the latest compatible version. */
+{
+    return Tcl_PkgPresentEx(interp, name, version, exact, NULL);
+}
+
+const char *
+Tcl_PkgPresentEx(
+    Tcl_Interp *interp,		/* Interpreter in which package is now
+				 * available. */
+    const char *name,		/* Name of desired package. */
+    const char *version,	/* Version string for desired version; NULL
+				 * means use the latest version available. */
+    int exact,			/* Non-zero means that only the particular
+				 * version given is acceptable. Zero means use
+				 * the latest compatible version. */
+    void *clientDataPtr)	/* Used to return the client data for this
+				 * package. If it is NULL then the client data
+				 * is not returned. This is unchanged if this
+				 * call fails for any reason. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hPtr;
+    Package *pkgPtr;
+
+    hPtr = Tcl_FindHashEntry(&iPtr->packageTable, name);
+    if (hPtr) {
+	pkgPtr = Tcl_GetHashValue(hPtr);
+	if (pkgPtr->version != NULL) {
+	    /*
+	     * At this point we know that the package is present. Make sure
+	     * that the provided version meets the current requirement by
+	     * calling Tcl_PkgRequireEx() to check for us.
+	     */
+
+	    const char *foundVersion = Tcl_PkgRequireEx(interp, name, version,
+		    exact, clientDataPtr);
+
+	    if (foundVersion == NULL) {
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PACKAGE", name,
+			NULL);
+	    }
+	    return foundVersion;
+	}
+    }
+
+    if (version != NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"package %s %s is not present", name, version));
+    } else {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"package %s is not present", name));
+    }
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "PACKAGE", name, NULL);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PackageObjCmd --
+ *
+ *	This function is invoked to process the "package" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_PackageObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    static const char *const pkgOptions[] = {
+	"files",  "forget",  "ifneeded", "names",   "prefer",
+	"present", "provide", "require",  "unknown", "vcompare",
+	"versions", "vsatisfies", NULL
+    };
+    enum pkgOptions {
+	PKG_FILES,  PKG_FORGET,  PKG_IFNEEDED, PKG_NAMES,   PKG_PREFER,
+	PKG_PRESENT, PKG_PROVIDE, PKG_REQUIRE,  PKG_UNKNOWN, PKG_VCOMPARE,
+	PKG_VERSIONS, PKG_VSATISFIES
+    };
+    Interp *iPtr = (Interp *) interp;
+    int optionIndex, exact, i, satisfies;
+    PkgAvail *availPtr, *prevPtr;
+    Package *pkgPtr;
+    Tcl_HashEntry *hPtr;
+    Tcl_HashSearch search;
+    Tcl_HashTable *tablePtr;
+    const char *version;
+    const char *argv2, *argv3, *argv4;
+    char *iva = NULL, *ivb = NULL;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], pkgOptions, "option", 0,
+	    &optionIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch ((enum pkgOptions) optionIndex) {
+    case PKG_FILES: {
+	PkgFiles *pkgFiles;
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "package");
+	    return TCL_ERROR;
+	}
+	pkgFiles = (PkgFiles *) Tcl_GetAssocData(interp, "tclPkgFiles", NULL);
+	if (pkgFiles) {
+	    Tcl_HashEntry *entry = Tcl_FindHashEntry(&pkgFiles->table, Tcl_GetString(objv[2]));
+	    if (entry) {
+		Tcl_SetObjResult(interp, (Tcl_Obj *)Tcl_GetHashValue(entry));
+	    }
+	}
+	break;
+    }
+    case PKG_FORGET: {
+	const char *keyString;
+	PkgFiles *pkgFiles = (PkgFiles *) Tcl_GetAssocData(interp, "tclPkgFiles", NULL);
+
+	for (i = 2; i < objc; i++) {
+	    keyString = TclGetString(objv[i]);
+	    if (pkgFiles) {
+		hPtr = Tcl_FindHashEntry(&pkgFiles->table, keyString);
+		if (hPtr) {
+		    Tcl_Obj *obj = Tcl_GetHashValue(hPtr);
+		    Tcl_DeleteHashEntry(hPtr);
+		    Tcl_DecrRefCount(obj);
+		}
+	    }
+
+	    hPtr = Tcl_FindHashEntry(&iPtr->packageTable, keyString);
+	    if (hPtr == NULL) {
+		continue;
+	    }
+	    pkgPtr = Tcl_GetHashValue(hPtr);
+	    Tcl_DeleteHashEntry(hPtr);
+	    if (pkgPtr->version != NULL) {
+		ckfree(pkgPtr->version);
+	    }
+	    while (pkgPtr->availPtr != NULL) {
+		availPtr = pkgPtr->availPtr;
+		pkgPtr->availPtr = availPtr->nextPtr;
+		Tcl_EventuallyFree(availPtr->version, TCL_DYNAMIC);
+		Tcl_EventuallyFree(availPtr->script, TCL_DYNAMIC);
+		if (availPtr->pkgIndex) {
+		    Tcl_EventuallyFree(availPtr->pkgIndex, TCL_DYNAMIC);
+		}
+		ckfree(availPtr);
+	    }
+	    ckfree(pkgPtr);
+	}
+	break;
+    }
+    case PKG_IFNEEDED: {
+	int length, res;
+	char *argv3i, *avi;
+
+	if ((objc != 4) && (objc != 5)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "package version ?script?");
+	    return TCL_ERROR;
+	}
+	argv3 = TclGetString(objv[3]);
+	if (CheckVersionAndConvert(interp, argv3, &argv3i, NULL) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	argv2 = TclGetString(objv[2]);
+	if (objc == 4) {
+	    hPtr = Tcl_FindHashEntry(&iPtr->packageTable, argv2);
+	    if (hPtr == NULL) {
+		ckfree(argv3i);
+		return TCL_OK;
+	    }
+	    pkgPtr = Tcl_GetHashValue(hPtr);
+	} else {
+	    pkgPtr = FindPackage(interp, argv2);
+	}
+	argv3 = TclGetStringFromObj(objv[3], &length);
+
+	for (availPtr = pkgPtr->availPtr, prevPtr = NULL; availPtr != NULL;
+		prevPtr = availPtr, availPtr = availPtr->nextPtr) {
+	    if (CheckVersionAndConvert(interp, availPtr->version, &avi,
+		    NULL) != TCL_OK) {
+		ckfree(argv3i);
+		return TCL_ERROR;
+	    }
+
+	    res = CompareVersions(avi, argv3i, NULL);
+	    ckfree(avi);
+
+	    if (res == 0){
+		if (objc == 4) {
+		    ckfree(argv3i);
+		    Tcl_SetObjResult(interp,
+			    Tcl_NewStringObj(availPtr->script, -1));
+		    return TCL_OK;
+		}
+		Tcl_EventuallyFree(availPtr->script, TCL_DYNAMIC);
+		if (availPtr->pkgIndex) {
+		    Tcl_EventuallyFree(availPtr->pkgIndex, TCL_DYNAMIC);
+		}
+		break;
+	    }
+	}
+	ckfree(argv3i);
+
+	if (objc == 4) {
+	    return TCL_OK;
+	}
+	if (availPtr == NULL) {
+	    availPtr = ckalloc(sizeof(PkgAvail));
+	    availPtr->pkgIndex = 0;
+	    DupBlock(availPtr->version, argv3, (unsigned) length + 1);
+
+	    if (prevPtr == NULL) {
+		availPtr->nextPtr = pkgPtr->availPtr;
+		pkgPtr->availPtr = availPtr;
+	    } else {
+		availPtr->nextPtr = prevPtr->nextPtr;
+		prevPtr->nextPtr = availPtr;
+	    }
+	}
+	if (iPtr->scriptFile) {
+	    argv4 = TclGetStringFromObj(iPtr->scriptFile, &length);
+	    DupBlock(availPtr->pkgIndex, argv4, (unsigned) length + 1);
+	}
+	argv4 = TclGetStringFromObj(objv[4], &length);
+	DupBlock(availPtr->script, argv4, (unsigned) length + 1);
+	break;
+    }
+    case PKG_NAMES:
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	} else {
+	    Tcl_Obj *resultObj;
+
+	    resultObj = Tcl_NewObj();
+	    tablePtr = &iPtr->packageTable;
+	    for (hPtr = Tcl_FirstHashEntry(tablePtr, &search); hPtr != NULL;
+		    hPtr = Tcl_NextHashEntry(&search)) {
+		pkgPtr = Tcl_GetHashValue(hPtr);
+		if ((pkgPtr->version != NULL) || (pkgPtr->availPtr != NULL)) {
+		    Tcl_ListObjAppendElement(NULL,resultObj, Tcl_NewStringObj(
+			    Tcl_GetHashKey(tablePtr, hPtr), -1));
+		}
+	    }
+	    Tcl_SetObjResult(interp, resultObj);
+	}
+	break;
+    case PKG_PRESENT: {
+	const char *name;
+
+	if (objc < 3) {
+	    goto require;
+	}
+	argv2 = TclGetString(objv[2]);
+	if ((argv2[0] == '-') && (strcmp(argv2, "-exact") == 0)) {
+	    if (objc != 5) {
+		goto requireSyntax;
+	    }
+	    exact = 1;
+	    name = TclGetString(objv[3]);
+	} else {
+	    exact = 0;
+	    name = argv2;
+	}
+
+	hPtr = Tcl_FindHashEntry(&iPtr->packageTable, name);
+	if (hPtr != NULL) {
+	    pkgPtr = Tcl_GetHashValue(hPtr);
+	    if (pkgPtr->version != NULL) {
+		goto require;
+	    }
+	}
+
+	version = NULL;
+	if (exact) {
+	    version = TclGetString(objv[4]);
+	    if (CheckVersionAndConvert(interp, version, NULL,
+		    NULL) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	} else {
+	    if (CheckAllRequirements(interp, objc-3, objv+3) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if ((objc > 3) && (CheckVersionAndConvert(interp,
+		    TclGetString(objv[3]), NULL, NULL) == TCL_OK)) {
+		version = TclGetString(objv[3]);
+	    }
+	}
+	Tcl_PkgPresentEx(interp, name, version, exact, NULL);
+	return TCL_ERROR;
+	break;
+    }
+    case PKG_PROVIDE:
+	if ((objc != 3) && (objc != 4)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "package ?version?");
+	    return TCL_ERROR;
+	}
+	argv2 = TclGetString(objv[2]);
+	if (objc == 3) {
+	    hPtr = Tcl_FindHashEntry(&iPtr->packageTable, argv2);
+	    if (hPtr != NULL) {
+		pkgPtr = Tcl_GetHashValue(hPtr);
+		if (pkgPtr->version != NULL) {
+		    Tcl_SetObjResult(interp,
+			    Tcl_NewStringObj(pkgPtr->version, -1));
+		}
+	    }
+	    return TCL_OK;
+	}
+	argv3 = TclGetString(objv[3]);
+	if (CheckVersionAndConvert(interp, argv3, NULL, NULL) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	return Tcl_PkgProvideEx(interp, argv2, argv3, NULL);
+    case PKG_REQUIRE:
+    require:
+	if (objc < 3) {
+	requireSyntax:
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "?-exact? package ?requirement ...?");
+	    return TCL_ERROR;
+	}
+
+	version = NULL;
+
+	argv2 = TclGetString(objv[2]);
+	if ((argv2[0] == '-') && (strcmp(argv2, "-exact") == 0)) {
+	    Tcl_Obj *ov;
+	    int res;
+
+	    if (objc != 5) {
+		goto requireSyntax;
+	    }
+
+	    version = TclGetString(objv[4]);
+	    if (CheckVersionAndConvert(interp, version, NULL,
+		    NULL) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+
+	    /*
+	     * Create a new-style requirement for the exact version.
+	     */
+
+	    ov = Tcl_NewStringObj(version, -1);
+	    Tcl_AppendStringsToObj(ov, "-", version, NULL);
+	    version = NULL;
+	    argv3 = TclGetString(objv[3]);
+
+	    Tcl_IncrRefCount(ov);
+	    res = Tcl_PkgRequireProc(interp, argv3, 1, &ov, NULL);
+	    TclDecrRefCount(ov);
+	    return res;
+	} else {
+	    if (CheckAllRequirements(interp, objc-3, objv+3) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+
+	    return Tcl_PkgRequireProc(interp, argv2, objc-3, objv+3, NULL);
+	}
+	break;
+    case PKG_UNKNOWN: {
+	int length;
+
+	if (objc == 2) {
+	    if (iPtr->packageUnknown != NULL) {
+		Tcl_SetObjResult(interp,
+			Tcl_NewStringObj(iPtr->packageUnknown, -1));
+	    }
+	} else if (objc == 3) {
+	    if (iPtr->packageUnknown != NULL) {
+		ckfree(iPtr->packageUnknown);
+	    }
+	    argv2 = TclGetStringFromObj(objv[2], &length);
+	    if (argv2[0] == 0) {
+		iPtr->packageUnknown = NULL;
+	    } else {
+		DupBlock(iPtr->packageUnknown, argv2, (unsigned) length+1);
+	    }
+	} else {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?command?");
+	    return TCL_ERROR;
+	}
+	break;
+    }
+    case PKG_PREFER: {
+	static const char *const pkgPreferOptions[] = {
+	    "latest", "stable", NULL
+	};
+
+	/*
+	 * See tclInt.h for the enum, just before Interp.
+	 */
+
+	if (objc > 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?latest|stable?");
+	    return TCL_ERROR;
+	} else if (objc == 3) {
+	    /*
+	     * Seting the value.
+	     */
+
+	    int newPref;
+
+	    if (Tcl_GetIndexFromObj(interp, objv[2], pkgPreferOptions,
+		    "preference", 0, &newPref) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+
+	    if (newPref < iPtr->packagePrefer) {
+		iPtr->packagePrefer = newPref;
+	    }
+	}
+
+	/*
+	 * Always return current value.
+	 */
+
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj(pkgPreferOptions[iPtr->packagePrefer], -1));
+	break;
+    }
+    case PKG_VCOMPARE:
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "version1 version2");
+	    return TCL_ERROR;
+	}
+	argv3 = TclGetString(objv[3]);
+	argv2 = TclGetString(objv[2]);
+	if (CheckVersionAndConvert(interp, argv2, &iva, NULL) != TCL_OK ||
+		CheckVersionAndConvert(interp, argv3, &ivb, NULL) != TCL_OK) {
+	    if (iva != NULL) {
+		ckfree(iva);
+	    }
+
+	    /*
+	     * ivb cannot be set in this branch.
+	     */
+
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Comparison is done on the internal representation.
+	 */
+
+	Tcl_SetObjResult(interp,
+		Tcl_NewIntObj(CompareVersions(iva, ivb, NULL)));
+	ckfree(iva);
+	ckfree(ivb);
+	break;
+    case PKG_VERSIONS:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "package");
+	    return TCL_ERROR;
+	} else {
+	    Tcl_Obj *resultObj = Tcl_NewObj();
+
+	    argv2 = TclGetString(objv[2]);
+	    hPtr = Tcl_FindHashEntry(&iPtr->packageTable, argv2);
+	    if (hPtr != NULL) {
+		pkgPtr = Tcl_GetHashValue(hPtr);
+		for (availPtr = pkgPtr->availPtr; availPtr != NULL;
+			availPtr = availPtr->nextPtr) {
+		    Tcl_ListObjAppendElement(NULL, resultObj,
+			    Tcl_NewStringObj(availPtr->version, -1));
+		}
+	    }
+	    Tcl_SetObjResult(interp, resultObj);
+	}
+	break;
+    case PKG_VSATISFIES: {
+	char *argv2i = NULL;
+
+	if (objc < 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "version ?requirement ...?");
+	    return TCL_ERROR;
+	}
+
+	argv2 = TclGetString(objv[2]);
+	if (CheckVersionAndConvert(interp, argv2, &argv2i, NULL) != TCL_OK) {
+	    return TCL_ERROR;
+	} else if (CheckAllRequirements(interp, objc-3, objv+3) != TCL_OK) {
+	    ckfree(argv2i);
+	    return TCL_ERROR;
+	}
+
+	satisfies = SomeRequirementSatisfied(argv2i, objc-3, objv+3);
+	ckfree(argv2i);
+
+	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(satisfies));
+	break;
+    }
+    default:
+	Tcl_Panic("Tcl_PackageObjCmd: bad option index to pkgOptions");
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FindPackage --
+ *
+ *	This function finds the Package record for a particular package in a
+ *	particular interpreter, creating a record if one doesn't already
+ *	exist.
+ *
+ * Results:
+ *	The return value is a pointer to the Package record for the package.
+ *
+ * Side effects:
+ *	A new Package record may be created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Package *
+FindPackage(
+    Tcl_Interp *interp,		/* Interpreter to use for package lookup. */
+    const char *name)		/* Name of package to fine. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+    Package *pkgPtr;
+
+    hPtr = Tcl_CreateHashEntry(&iPtr->packageTable, name, &isNew);
+    if (isNew) {
+	pkgPtr = ckalloc(sizeof(Package));
+	pkgPtr->version = NULL;
+	pkgPtr->availPtr = NULL;
+	pkgPtr->clientData = NULL;
+	Tcl_SetHashValue(hPtr, pkgPtr);
+    } else {
+	pkgPtr = Tcl_GetHashValue(hPtr);
+    }
+    return pkgPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreePackageInfo --
+ *
+ *	This function is called during interpreter deletion to free all of the
+ *	package-related information for the interpreter.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFreePackageInfo(
+    Interp *iPtr)		/* Interpreter that is being deleted. */
+{
+    Package *pkgPtr;
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hPtr;
+    PkgAvail *availPtr;
+
+    for (hPtr = Tcl_FirstHashEntry(&iPtr->packageTable, &search);
+	    hPtr != NULL; hPtr = Tcl_NextHashEntry(&search)) {
+	pkgPtr = Tcl_GetHashValue(hPtr);
+	if (pkgPtr->version != NULL) {
+	    ckfree(pkgPtr->version);
+	}
+	while (pkgPtr->availPtr != NULL) {
+	    availPtr = pkgPtr->availPtr;
+	    pkgPtr->availPtr = availPtr->nextPtr;
+	    Tcl_EventuallyFree(availPtr->version, TCL_DYNAMIC);
+	    Tcl_EventuallyFree(availPtr->script, TCL_DYNAMIC);
+	    if (availPtr->pkgIndex) {
+		Tcl_EventuallyFree(availPtr->pkgIndex, TCL_DYNAMIC);
+	    }
+	    ckfree(availPtr);
+	}
+	ckfree(pkgPtr);
+    }
+    Tcl_DeleteHashTable(&iPtr->packageTable);
+    if (iPtr->packageUnknown != NULL) {
+	ckfree(iPtr->packageUnknown);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckVersionAndConvert --
+ *
+ *	This function checks to see whether a version number has valid syntax.
+ *	It also generates a semi-internal representation (string rep of a list
+ *	of numbers).
+ *
+ * Results:
+ *	If string is a properly formed version number the TCL_OK is returned.
+ *	Otherwise TCL_ERROR is returned and an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckVersionAndConvert(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    const char *string,		/* Supposedly a version number, which is
+				 * groups of decimal digits separated by
+				 * dots. */
+    char **internal,		/* Internal normalized representation */
+    int *stable)		/* Flag: Version is (un)stable. */
+{
+    const char *p = string;
+    char prevChar;
+    int hasunstable = 0;
+    /*
+     * 4* assuming that each char is a separator (a,b become ' -x ').
+     * 4+ to have spce for an additional -2 at the end
+     */
+    char *ibuf = ckalloc(4 + 4*strlen(string));
+    char *ip = ibuf;
+
+    /*
+     * Basic rules
+     * (1) First character has to be a digit.
+     * (2) All other characters have to be a digit or '.'
+     * (3) Two '.'s may not follow each other.
+     *
+     * TIP 268, Modified rules
+     * (1) s.a.
+     * (2) All other characters have to be a digit, 'a', 'b', or '.'
+     * (3) s.a.
+     * (4) Only one of 'a' or 'b' may occur.
+     * (5) Neither 'a', nor 'b' may occur before or after a '.'
+     */
+
+    if (!isdigit(UCHAR(*p))) {				/* INTL: digit */
+	goto error;
+    }
+
+    *ip++ = *p;
+
+    for (prevChar = *p, p++; *p != 0; p++) {
+	if (!isdigit(UCHAR(*p)) &&			/* INTL: digit */
+		((*p!='.' && *p!='a' && *p!='b') ||
+		((hasunstable && (*p=='a' || *p=='b')) ||
+		((prevChar=='a' || prevChar=='b' || prevChar=='.')
+			&& (*p=='.')) ||
+		((*p=='a' || *p=='b' || *p=='.') && prevChar=='.')))) {
+	    goto error;
+	}
+
+	if (*p == 'a' || *p == 'b') {
+	    hasunstable = 1;
+	}
+
+	/*
+	 * Translation to the internal rep. Regular version chars are copied
+	 * as is. The separators are translated to numerics. The new separator
+	 * for all parts is space.
+	 */
+
+	if (*p == '.') {
+	    *ip++ = ' ';
+	    *ip++ = '0';
+	    *ip++ = ' ';
+	} else if (*p == 'a') {
+	    *ip++ = ' ';
+	    *ip++ = '-';
+	    *ip++ = '2';
+	    *ip++ = ' ';
+	} else if (*p == 'b') {
+	    *ip++ = ' ';
+	    *ip++ = '-';
+	    *ip++ = '1';
+	    *ip++ = ' ';
+	} else {
+	    *ip++ = *p;
+	}
+
+	prevChar = *p;
+    }
+    if (prevChar!='.' && prevChar!='a' && prevChar!='b') {
+	*ip = '\0';
+	if (internal != NULL) {
+	    *internal = ibuf;
+	} else {
+	    ckfree(ibuf);
+	}
+	if (stable != NULL) {
+	    *stable = !hasunstable;
+	}
+	return TCL_OK;
+    }
+
+  error:
+    ckfree(ibuf);
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "expected version number but got \"%s\"", string));
+    Tcl_SetErrorCode(interp, "TCL", "VALUE", "VERSION", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CompareVersions --
+ *
+ *	This function compares two version numbers (in internal rep).
+ *
+ * Results:
+ *	The return value is -1 if v1 is less than v2, 0 if the two version
+ *	numbers are the same, and 1 if v1 is greater than v2. If *satPtr is
+ *	non-NULL, the word it points to is filled in with 1 if v2 >= v1 and
+ *	both numbers have the same major number or 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CompareVersions(
+    char *v1, char *v2,		/* Versions strings, of form 2.1.3 (any number
+				 * of version numbers). */
+    int *isMajorPtr)		/* If non-null, the word pointed to is filled
+				 * in with a 0/1 value. 1 means that the
+				 * difference occured in the first element. */
+{
+    int thisIsMajor, res, flip;
+    char *s1, *e1, *s2, *e2, o1, o2;
+
+    /*
+     * Each iteration of the following loop processes one number from each
+     * string, terminated by a " " (space). If those numbers don't match then
+     * the comparison is over; otherwise, we loop back for the next number.
+     *
+     * TIP 268.
+     * This is identical the function 'ComparePkgVersion', but using the new
+     * space separator as used by the internal rep of version numbers. The
+     * special separators 'a' and 'b' have already been dealt with in
+     * 'CheckVersionAndConvert', they were translated into numbers as well.
+     * This keeps the comparison sane. Otherwise we would have to compare
+     * numerics, the separators, and also deal with the special case of
+     * end-of-string compared to separators. The semi-list rep we get here is
+     * much easier to handle, as it is still regular.
+     *
+     * Rewritten to not compute a numeric value for the extracted version
+     * number, but do string comparison. Skip any leading zeros for that to
+     * work. This change breaks through the 32bit-limit on version numbers.
+     */
+
+    thisIsMajor = 1;
+    s1 = v1;
+    s2 = v2;
+
+    while (1) {
+	/*
+	 * Parse one decimal number from the front of each string. Skip
+	 * leading zeros. Terminate found number for upcoming string-wise
+	 * comparison, if needed.
+	 */
+
+	while ((*s1 != 0) && (*s1 == '0')) {
+	    s1++;
+	}
+	while ((*s2 != 0) && (*s2 == '0')) {
+	    s2++;
+	}
+
+	/*
+	 * s1, s2 now point to the beginnings of the numbers to compare. Test
+	 * for their signs first, as shortcut to the result (different signs),
+	 * or determines if result has to be flipped (both negative). If there
+	 * is no shortcut we have to insert terminators later to limit the
+	 * strcmp.
+	 */
+
+	if ((*s1 == '-') && (*s2 != '-')) {
+	    /* s1 < 0, s2 >= 0 => s1 < s2 */
+	    res = -1;
+	    break;
+	}
+	if ((*s1 != '-') && (*s2 == '-')) {
+	    /* s1 >= 0, s2 < 0 => s1 > s2 */
+	    res = 1;
+	    break;
+	}
+
+	if ((*s1 == '-') && (*s2 == '-')) {
+	    /* a < b => -a > -b, etc. */
+	    s1++;
+	    s2++;
+	    flip = 1;
+	} else {
+	    flip = 0;
+	}
+
+	/*
+	 * The string comparison is needed, so now we determine where the
+	 * numbers end.
+	 */
+
+	e1 = s1;
+	while ((*e1 != 0) && (*e1 != ' ')) {
+	    e1++;
+	}
+	e2 = s2;
+	while ((*e2 != 0) && (*e2 != ' ')) {
+	    e2++;
+	}
+
+	/*
+	 * s1 .. e1 and s2 .. e2 now bracket the numbers to compare. Insert
+	 * terminators, compare, and restore actual contents. First however
+	 * another shortcut. Compare lengths. Shorter string is smaller
+	 * number! Thus we strcmp only strings of identical length.
+	 */
+
+	if ((e1-s1) < (e2-s2)) {
+	    res = -1;
+	} else if ((e2-s2) < (e1-s1)) {
+	    res = 1;
+	} else {
+	    o1 = *e1;
+	    *e1 = '\0';
+	    o2 = *e2;
+	    *e2 = '\0';
+
+	    res = strcmp(s1, s2);
+	    res = (res < 0) ? -1 : (res ? 1 : 0);
+
+	    *e1 = o1;
+	    *e2 = o2;
+	}
+
+	/*
+	 * Stop comparing segments when a difference has been found. Here we
+	 * may have to flip the result to account for signs.
+	 */
+
+	if (res != 0) {
+	    if (flip) {
+		res = -res;
+	    }
+	    break;
+	}
+
+	/*
+	 * Go on to the next version number if the current numbers match.
+	 * However stop processing if the end of both numbers has been
+	 * reached.
+	 */
+
+	s1 = e1;
+	s2 = e2;
+
+	if (*s1 != 0) {
+	    s1++;
+	} else if (*s2 == 0) {
+	    /*
+	     * s1, s2 both at the end => identical
+	     */
+
+	    res = 0;
+	    break;
+	}
+	if (*s2 != 0) {
+	    s2++;
+	}
+	thisIsMajor = 0;
+    }
+
+    if (isMajorPtr != NULL) {
+	*isMajorPtr = thisIsMajor;
+    }
+
+    return res;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckAllRequirements --
+ *
+ *	This function checks to see whether all requirements in a set have
+ *	valid syntax.
+ *
+ * Results:
+ *	TCL_OK is returned if all requirements are valid. Otherwise TCL_ERROR
+ *	is returned and an error message is left in the interp's result.
+ *
+ * Side effects:
+ *	May modify the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckAllRequirements(
+    Tcl_Interp *interp,
+    int reqc,			/* Requirements to check. */
+    Tcl_Obj *const reqv[])
+{
+    int i;
+
+    for (i = 0; i < reqc; i++) {
+	if ((CheckRequirement(interp, TclGetString(reqv[i])) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckRequirement --
+ *
+ *	This function checks to see whether a requirement has valid syntax.
+ *
+ * Results:
+ *	If string is a properly formed requirement then TCL_OK is returned.
+ *	Otherwise TCL_ERROR is returned and an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckRequirement(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    const char *string)		/* Supposedly a requirement. */
+{
+    /*
+     * Syntax of requirement = version
+     *			     = version-version
+     *			     = version-
+     */
+
+    char *dash = NULL, *buf;
+
+    dash = strchr(string, '-');
+    if (dash == NULL) {
+	/*
+	 * No dash found, has to be a simple version.
+	 */
+
+	return CheckVersionAndConvert(interp, string, NULL, NULL);
+    }
+
+    if (strchr(dash+1, '-') != NULL) {
+	/*
+	 * More dashes found after the first. This is wrong.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"expected versionMin-versionMax but got \"%s\"", string));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "VERSIONRANGE", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Exactly one dash is present. Copy the string, split at the location of
+     * dash and check that both parts are versions. Note that the max part can
+     * be empty. Also note that the string allocated with strdup() must be
+     * freed with free() and not ckfree().
+     */
+
+    DupString(buf, string);
+    dash = buf + (dash - string);
+    *dash = '\0';		/* buf now <=> min part */
+    dash++;			/* dash now <=> max part */
+
+    if ((CheckVersionAndConvert(interp, buf, NULL, NULL) != TCL_OK) ||
+	    ((*dash != '\0') &&
+	    (CheckVersionAndConvert(interp, dash, NULL, NULL) != TCL_OK))) {
+	ckfree(buf);
+	return TCL_ERROR;
+    }
+
+    ckfree(buf);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AddRequirementsToResult --
+ *
+ *	This function accumulates requirements in the interpreter result.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interpreter result is extended.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AddRequirementsToResult(
+    Tcl_Interp *interp,
+    int reqc,			/* Requirements constraining the desired
+				 * version. */
+    Tcl_Obj *const reqv[])	/* 0 means to use the latest version
+				 * available. */
+{
+    Tcl_Obj *result = Tcl_GetObjResult(interp);
+    int i, length;
+
+    for (i = 0; i < reqc; i++) {
+	const char *v = TclGetStringFromObj(reqv[i], &length);
+
+	if ((length & 0x1) && (v[length/2] == '-')
+		&& (strncmp(v, v+((length+1)/2), length/2) == 0)) {
+	    Tcl_AppendPrintfToObj(result, " exactly %s", v+((length+1)/2));
+	} else {
+	    Tcl_AppendPrintfToObj(result, " %s", v);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AddRequirementsToDString --
+ *
+ *	This function accumulates requirements in a DString.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The DString argument is extended.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AddRequirementsToDString(
+    Tcl_DString *dsPtr,
+    int reqc,			/* Requirements constraining the desired
+				 * version. */
+    Tcl_Obj *const reqv[])	/* 0 means to use the latest version
+				 * available. */
+{
+    int i;
+
+    if (reqc > 0) {
+	for (i = 0; i < reqc; i++) {
+	    TclDStringAppendLiteral(dsPtr, " ");
+	    TclDStringAppendObj(dsPtr, reqv[i]);
+	}
+    } else {
+	TclDStringAppendLiteral(dsPtr, " 0-");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SomeRequirementSatisfied --
+ *
+ *	This function checks to see whether a version satisfies at least one
+ *	of a set of requirements.
+ *
+ * Results:
+ *	If the requirements are satisfied 1 is returned. Otherwise 0 is
+ *	returned. The function assumes that all pieces have valid syntax. And
+ *	is allowed to make that assumption.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SomeRequirementSatisfied(
+    char *availVersionI,	/* Candidate version to check against the
+				 * requirements. */
+    int reqc,			/* Requirements constraining the desired
+				 * version. */
+    Tcl_Obj *const reqv[])	/* 0 means to use the latest version
+				 * available. */
+{
+    int i;
+
+    for (i = 0; i < reqc; i++) {
+	if (RequirementSatisfied(availVersionI, TclGetString(reqv[i]))) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RequirementSatisfied --
+ *
+ *	This function checks to see whether a version satisfies a requirement.
+ *
+ * Results:
+ *	If the requirement is satisfied 1 is returned. Otherwise 0 is
+ *	returned. The function assumes that all pieces have valid syntax, and
+ *	is allowed to make that assumption.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+RequirementSatisfied(
+    char *havei,		/* Version string, of candidate package we
+				 * have. */
+    const char *req)		/* Requirement string the candidate has to
+				 * satisfy. */
+{
+    /*
+     * The have candidate is already in internal rep.
+     */
+
+    int satisfied, res;
+    char *dash = NULL, *buf, *min, *max;
+
+    dash = strchr(req, '-');
+    if (dash == NULL) {
+	/*
+	 * No dash found, is a simple version, fallback to regular check. The
+	 * 'CheckVersionAndConvert' cannot fail. We pad the requirement with
+	 * 'a0', i.e '-2' before doing the comparison to properly accept
+	 * unstables as well.
+	 */
+
+	char *reqi = NULL;
+	int thisIsMajor;
+
+	CheckVersionAndConvert(NULL, req, &reqi, NULL);
+	strcat(reqi, " -2");
+	res = CompareVersions(havei, reqi, &thisIsMajor);
+	satisfied = (res == 0) || ((res == 1) && !thisIsMajor);
+	ckfree(reqi);
+	return satisfied;
+    }
+
+    /*
+     * Exactly one dash is present (Assumption of valid syntax). Copy the req,
+     * split at the location of dash and check that both parts are versions.
+     * Note that the max part can be empty.
+     */
+
+    DupString(buf, req);
+    dash = buf + (dash - req);
+    *dash = '\0';		/* buf now <=> min part */
+    dash++;			/* dash now <=> max part */
+
+    if (*dash == '\0') {
+	/*
+	 * We have a min, but no max. For the comparison we generate the
+	 * internal rep, padded with 'a0' i.e. '-2'.
+	 */
+
+	CheckVersionAndConvert(NULL, buf, &min, NULL);
+	strcat(min, " -2");
+	satisfied = (CompareVersions(havei, min, NULL) >= 0);
+	ckfree(min);
+	ckfree(buf);
+	return satisfied;
+    }
+
+    /*
+     * We have both min and max, and generate their internal reps. When
+     * identical we compare as is, otherwise we pad with 'a0' to ove the range
+     * a bit.
+     */
+
+    CheckVersionAndConvert(NULL, buf, &min, NULL);
+    CheckVersionAndConvert(NULL, dash, &max, NULL);
+
+    if (CompareVersions(min, max, NULL) == 0) {
+	satisfied = (CompareVersions(min, havei, NULL) == 0);
+    } else {
+	strcat(min, " -2");
+	strcat(max, " -2");
+	satisfied = ((CompareVersions(min, havei, NULL) <= 0) &&
+		(CompareVersions(havei, max, NULL) < 0));
+    }
+
+    ckfree(min);
+    ckfree(max);
+    ckfree(buf);
+    return satisfied;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PkgInitStubsCheck --
+ *
+ *	This is a replacement routine for Tcl_InitStubs() that is called
+ *	from code where -DUSE_TCL_STUBS has not been enabled.
+ *
+ * Results:
+ *	Returns the version of a conforming stubs table, or NULL, if
+ *	the table version doesn't satisfy the requested requirements,
+ *	according to historical practice.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_PkgInitStubsCheck(
+    Tcl_Interp *interp,
+    const char * version,
+    int exact)
+{
+    const char *actualVersion = Tcl_PkgPresent(interp, "Tcl", version, 0);
+
+    if ((exact&1) && actualVersion) {
+	const char *p = version;
+	int count = 0;
+
+	while (*p) {
+	    count += !isdigit(UCHAR(*p++));
+	}
+	if (count == 1) {
+	    if (0 != strncmp(version, actualVersion, strlen(version))) {
+		/* Construct error message */
+		Tcl_PkgPresent(interp, "Tcl", version, 1);
+		return NULL;
+	    }
+	} else {
+	    return Tcl_PkgPresent(interp, "Tcl", version, 1);
+	}
+    }
+    return actualVersion;
+}
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclPkgConfig.c b/generic/tclPkgConfig.c
new file mode 100644
index 0000000..466d535
--- /dev/null
+++ b/generic/tclPkgConfig.c
@@ -0,0 +1,135 @@
+/*
+ * tclPkgConfig.c --
+ *
+ *	This file contains the configuration information to embed into the tcl
+ *	binary library.
+ *
+ * Copyright (c) 2002 Andreas Kupries <andreas_kupries@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/* Note, the definitions in this module are influenced by the following C
+ * preprocessor macros:
+ *
+ * OSCMa  = shortcut for "old style configuration macro activates"
+ * NSCMdt = shortcut for "new style configuration macro declares that"
+ *
+ * - TCL_THREADS		OSCMa compilation as threaded core.
+ * - TCL_MEM_DEBUG		OSCMa memory debugging.
+ * - TCL_COMPILE_DEBUG		OSCMa debugging of bytecode compiler.
+ * - TCL_COMPILE_STATS		OSCMa bytecode compiler statistics.
+ *
+ * - TCL_CFG_DO64BIT		NSCMdt tcl is compiled for a 64bit system.
+ * - NDEBUG		NSCMdt tcl is compiled with symbol info off.
+ * - TCL_CFG_OPTIMIZED		NSCMdt tcl is compiled with cc optimizations on
+ * - TCL_CFG_PROFILED		NSCMdt tcl is compiled with profiling info.
+ *
+ * - CFG_RUNTIME_*		Paths to various stuff at runtime.
+ * - CFG_INSTALL_*		Paths to various stuff at installation time.
+ *
+ * - TCL_CFGVAL_ENCODING	string containing the encoding used for the
+ *				configuration values.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Use C preprocessor statements to define the various values for the embedded
+ * configuration information.
+ */
+
+#ifdef TCL_THREADS
+#  define  CFG_THREADED		"1"
+#else
+#  define  CFG_THREADED		"0"
+#endif
+
+#ifdef TCL_MEM_DEBUG
+#  define CFG_MEMDEBUG		"1"
+#else
+#  define CFG_MEMDEBUG		"0"
+#endif
+
+#ifdef TCL_COMPILE_DEBUG
+#  define CFG_COMPILE_DEBUG	"1"
+#else
+#  define CFG_COMPILE_DEBUG	"0"
+#endif
+
+#ifdef TCL_COMPILE_STATS
+#  define CFG_COMPILE_STATS	"1"
+#else
+#  define CFG_COMPILE_STATS	"0"
+#endif
+
+#ifdef TCL_CFG_DO64BIT
+#  define CFG_64		"1"
+#else
+#  define CFG_64		"0"
+#endif
+
+#ifndef NDEBUG
+#  define CFG_DEBUG		"1"
+#else
+#  define CFG_DEBUG		"0"
+#endif
+
+#ifdef TCL_CFG_OPTIMIZED
+#  define CFG_OPTIMIZED		"1"
+#else
+#  define CFG_OPTIMIZED		"0"
+#endif
+
+#ifdef TCL_CFG_PROFILED
+#  define CFG_PROFILED		"1"
+#else
+#  define CFG_PROFILED		"0"
+#endif
+
+static Tcl_Config const cfg[] = {
+    {"debug",			CFG_DEBUG},
+    {"threaded",		CFG_THREADED},
+    {"profiled",		CFG_PROFILED},
+    {"64bit",			CFG_64},
+    {"optimized",		CFG_OPTIMIZED},
+    {"mem_debug",		CFG_MEMDEBUG},
+    {"compile_debug",		CFG_COMPILE_DEBUG},
+    {"compile_stats",		CFG_COMPILE_STATS},
+
+    /* Runtime paths to various stuff */
+
+    {"libdir,runtime",		CFG_RUNTIME_LIBDIR},
+    {"bindir,runtime",		CFG_RUNTIME_BINDIR},
+    {"scriptdir,runtime",	CFG_RUNTIME_SCRDIR},
+    {"includedir,runtime",	CFG_RUNTIME_INCDIR},
+    {"docdir,runtime",		CFG_RUNTIME_DOCDIR},
+
+    /* Installation paths to various stuff */
+
+    {"libdir,install",		CFG_INSTALL_LIBDIR},
+    {"bindir,install",		CFG_INSTALL_BINDIR},
+    {"scriptdir,install",	CFG_INSTALL_SCRDIR},
+    {"includedir,install",	CFG_INSTALL_INCDIR},
+    {"docdir,install",		CFG_INSTALL_DOCDIR},
+
+    /* Last entry, closes the array */
+    {NULL, NULL}
+};
+
+void
+TclInitEmbeddedConfigurationInformation(
+    Tcl_Interp *interp)		/* Interpreter the configuration command is
+				 * registered in. */
+{
+    Tcl_RegisterConfig(interp, "tcl", cfg, TCL_CFGVAL_ENCODING);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclPlatDecls.h b/generic/tclPlatDecls.h
new file mode 100644
index 0000000..abc8ee8
--- /dev/null
+++ b/generic/tclPlatDecls.h
@@ -0,0 +1,122 @@
+/*
+ * tclPlatDecls.h --
+ *
+ *	Declarations of platform specific Tcl APIs.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * All rights reserved.
+ */
+
+#ifndef _TCLPLATDECLS
+#define _TCLPLATDECLS
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/*
+ * WARNING: This file is automatically generated by the tools/genStubs.tcl
+ * script.  Any modifications to the function declarations below should be made
+ * in the generic/tcl.decls script.
+ */
+
+/*
+ * TCHAR is needed here for win32, so if it is not defined yet do it here.
+ * This way, we don't need to include <tchar.h> just for one define.
+ */
+#if (defined(_WIN32) || defined(__CYGWIN__)) && !defined(_TCHAR_DEFINED)
+#   if defined(_UNICODE)
+	typedef wchar_t TCHAR;
+#   else
+	typedef char TCHAR;
+#   endif
+#   define _TCHAR_DEFINED
+#endif
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+/* 0 */
+EXTERN TCHAR *		Tcl_WinUtfToTChar(const char *str, int len,
+				Tcl_DString *dsPtr);
+/* 1 */
+EXTERN char *		Tcl_WinTCharToUtf(const TCHAR *str, int len,
+				Tcl_DString *dsPtr);
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+/* 0 */
+EXTERN int		Tcl_MacOSXOpenBundleResources(Tcl_Interp *interp,
+				const char *bundleName, int hasResourceFile,
+				int maxPathLen, char *libraryPath);
+/* 1 */
+EXTERN int		Tcl_MacOSXOpenVersionedBundleResources(
+				Tcl_Interp *interp, const char *bundleName,
+				const char *bundleVersion,
+				int hasResourceFile, int maxPathLen,
+				char *libraryPath);
+#endif /* MACOSX */
+
+typedef struct TclPlatStubs {
+    int magic;
+    void *hooks;
+
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+    TCHAR * (*tcl_WinUtfToTChar) (const char *str, int len, Tcl_DString *dsPtr); /* 0 */
+    char * (*tcl_WinTCharToUtf) (const TCHAR *str, int len, Tcl_DString *dsPtr); /* 1 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    int (*tcl_MacOSXOpenBundleResources) (Tcl_Interp *interp, const char *bundleName, int hasResourceFile, int maxPathLen, char *libraryPath); /* 0 */
+    int (*tcl_MacOSXOpenVersionedBundleResources) (Tcl_Interp *interp, const char *bundleName, const char *bundleVersion, int hasResourceFile, int maxPathLen, char *libraryPath); /* 1 */
+#endif /* MACOSX */
+} TclPlatStubs;
+
+extern const TclPlatStubs *tclPlatStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCL_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+#define Tcl_WinUtfToTChar \
+	(tclPlatStubsPtr->tcl_WinUtfToTChar) /* 0 */
+#define Tcl_WinTCharToUtf \
+	(tclPlatStubsPtr->tcl_WinTCharToUtf) /* 1 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+#define Tcl_MacOSXOpenBundleResources \
+	(tclPlatStubsPtr->tcl_MacOSXOpenBundleResources) /* 0 */
+#define Tcl_MacOSXOpenVersionedBundleResources \
+	(tclPlatStubsPtr->tcl_MacOSXOpenVersionedBundleResources) /* 1 */
+#endif /* MACOSX */
+
+#endif /* defined(USE_TCL_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#undef TCL_STORAGE_CLASS
+#define TCL_STORAGE_CLASS DLLIMPORT
+
+#endif /* _TCLPLATDECLS */
+
+
diff --git a/generic/tclPort.h b/generic/tclPort.h
new file mode 100644
index 0000000..12a60db
--- /dev/null
+++ b/generic/tclPort.h
@@ -0,0 +1,43 @@
+/*
+ * tclPort.h --
+ *
+ *	This header file handles porting issues that occur because
+ *	of differences between systems.  It reads in platform specific
+ *	portability files.
+ *
+ * Copyright (c) 1994-1995 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLPORT
+#define _TCLPORT
+
+#ifdef HAVE_TCL_CONFIG_H
+#include "tclConfig.h"
+#endif
+#if defined(_WIN32)
+#   include "tclWinPort.h"
+#else
+#   include "tclUnixPort.h"
+#endif
+#include "tcl.h"
+
+#if !defined(LLONG_MIN)
+#   ifdef TCL_WIDE_INT_IS_LONG
+#      define LLONG_MIN LONG_MIN
+#   else
+#      ifdef LLONG_BIT
+#         define LLONG_MIN ((Tcl_WideInt)(Tcl_LongAsWide(1)<<(LLONG_BIT-1)))
+#      else
+/* Assume we're on a system with a 64-bit 'long long' type */
+#         define LLONG_MIN ((Tcl_WideInt)(Tcl_LongAsWide(1)<<63))
+#      endif
+#   endif
+/* Assume that if LLONG_MIN is undefined, then so is LLONG_MAX */
+#   define LLONG_MAX (~LLONG_MIN)
+#endif
+
+
+#endif /* _TCLPORT */
diff --git a/generic/tclPosixStr.c b/generic/tclPosixStr.c
new file mode 100644
index 0000000..411eb27
--- /dev/null
+++ b/generic/tclPosixStr.c
@@ -0,0 +1,1211 @@
+/*
+ * tclPosixStr.c --
+ *
+ *	This file contains procedures that generate strings corresponding to
+ *	various POSIX-related codes, such as errno and signals.
+ *
+ * Copyright (c) 1991-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1996 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ErrnoId --
+ *
+ *	Return a textual identifier for the current errno value.
+ *
+ * Results:
+ *	This procedure returns a machine-readable textual identifier that
+ *	corresponds to the current errno value (e.g. "EPERM"). The identifier
+ *	is the same as the #define name in errno.h.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_ErrnoId(void)
+{
+    switch (errno) {
+#if defined(E2BIG) && (!defined(EOVERFLOW) || (E2BIG != EOVERFLOW))
+    case E2BIG: return "E2BIG";
+#endif
+#ifdef EACCES
+    case EACCES: return "EACCES";
+#endif
+#ifdef EADDRINUSE
+    case EADDRINUSE: return "EADDRINUSE";
+#endif
+#ifdef EADDRNOTAVAIL
+    case EADDRNOTAVAIL: return "EADDRNOTAVAIL";
+#endif
+#ifdef EADV
+    case EADV: return "EADV";
+#endif
+#ifdef EAFNOSUPPORT
+    case EAFNOSUPPORT: return "EAFNOSUPPORT";
+#endif
+#ifdef EAGAIN
+    case EAGAIN: return "EAGAIN";
+#endif
+#ifdef EALIGN
+    case EALIGN: return "EALIGN";
+#endif
+#if defined(EALREADY) && (!defined(EBUSY) || (EALREADY != EBUSY))
+    case EALREADY: return "EALREADY";
+#endif
+#ifdef EBADE
+    case EBADE: return "EBADE";
+#endif
+#ifdef EBADF
+    case EBADF: return "EBADF";
+#endif
+#ifdef EBADFD
+    case EBADFD: return "EBADFD";
+#endif
+#ifdef EBADMSG
+    case EBADMSG: return "EBADMSG";
+#endif
+#ifdef ECANCELED
+    case ECANCELED: return "ECANCELED";
+#endif
+#ifdef EBADR
+    case EBADR: return "EBADR";
+#endif
+#ifdef EBADRPC
+    case EBADRPC: return "EBADRPC";
+#endif
+#ifdef EBADRQC
+    case EBADRQC: return "EBADRQC";
+#endif
+#ifdef EBADSLT
+    case EBADSLT: return "EBADSLT";
+#endif
+#ifdef EBFONT
+    case EBFONT: return "EBFONT";
+#endif
+#ifdef EBUSY
+    case EBUSY: return "EBUSY";
+#endif
+#ifdef ECHILD
+    case ECHILD: return "ECHILD";
+#endif
+#ifdef ECHRNG
+    case ECHRNG: return "ECHRNG";
+#endif
+#ifdef ECOMM
+    case ECOMM: return "ECOMM";
+#endif
+#ifdef ECONNABORTED
+    case ECONNABORTED: return "ECONNABORTED";
+#endif
+#ifdef ECONNREFUSED
+    case ECONNREFUSED: return "ECONNREFUSED";
+#endif
+#ifdef ECONNRESET
+    case ECONNRESET: return "ECONNRESET";
+#endif
+#if defined(EDEADLK) && (!defined(EWOULDBLOCK) || (EDEADLK != EWOULDBLOCK))
+    case EDEADLK: return "EDEADLK";
+#endif
+#if defined(EDEADLOCK) && (!defined(EDEADLK) || (EDEADLOCK != EDEADLK))
+    case EDEADLOCK: return "EDEADLOCK";
+#endif
+#ifdef EDESTADDRREQ
+    case EDESTADDRREQ: return "EDESTADDRREQ";
+#endif
+#ifdef EDIRTY
+    case EDIRTY: return "EDIRTY";
+#endif
+#ifdef EDOM
+    case EDOM: return "EDOM";
+#endif
+#ifdef EDOTDOT
+    case EDOTDOT: return "EDOTDOT";
+#endif
+#ifdef EDQUOT
+    case EDQUOT: return "EDQUOT";
+#endif
+#ifdef EDUPPKG
+    case EDUPPKG: return "EDUPPKG";
+#endif
+#ifdef EEXIST
+    case EEXIST: return "EEXIST";
+#endif
+#ifdef EFAULT
+    case EFAULT: return "EFAULT";
+#endif
+#ifdef EFBIG
+    case EFBIG: return "EFBIG";
+#endif
+#ifdef EHOSTDOWN
+    case EHOSTDOWN: return "EHOSTDOWN";
+#endif
+#ifdef EHOSTUNREACH
+    case EHOSTUNREACH: return "EHOSTUNREACH";
+#endif
+#if defined(EIDRM) && (!defined(EINPROGRESS) || (EIDRM != EINPROGRESS))
+    case EIDRM: return "EIDRM";
+#endif
+#ifdef EINIT
+    case EINIT: return "EINIT";
+#endif
+#ifdef EINPROGRESS
+    case EINPROGRESS: return "EINPROGRESS";
+#endif
+#ifdef EINTR
+    case EINTR: return "EINTR";
+#endif
+#ifdef EINVAL
+    case EINVAL: return "EINVAL";
+#endif
+#ifdef EIO
+    case EIO: return "EIO";
+#endif
+#ifdef EISCONN
+    case EISCONN: return "EISCONN";
+#endif
+#ifdef EISDIR
+    case EISDIR: return "EISDIR";
+#endif
+#ifdef EISNAME
+    case EISNAM: return "EISNAM";
+#endif
+#ifdef ELBIN
+    case ELBIN: return "ELBIN";
+#endif
+#ifdef EL2HLT
+    case EL2HLT: return "EL2HLT";
+#endif
+#ifdef EL2NSYNC
+    case EL2NSYNC: return "EL2NSYNC";
+#endif
+#ifdef EL3HLT
+    case EL3HLT: return "EL3HLT";
+#endif
+#ifdef EL3RST
+    case EL3RST: return "EL3RST";
+#endif
+#ifdef ELIBACC
+    case ELIBACC: return "ELIBACC";
+#endif
+#ifdef ELIBBAD
+    case ELIBBAD: return "ELIBBAD";
+#endif
+#ifdef ELIBEXEC
+    case ELIBEXEC: return "ELIBEXEC";
+#endif
+#if defined(ELIBMAX) && (!defined(ECANCELED) || (ELIBMAX != ECANCELED))
+    case ELIBMAX: return "ELIBMAX";
+#endif
+#ifdef ELIBSCN
+    case ELIBSCN: return "ELIBSCN";
+#endif
+#ifdef ELNRNG
+    case ELNRNG: return "ELNRNG";
+#endif
+#if defined(ELOOP) && (!defined(ENOENT) || (ELOOP != ENOENT))
+    case ELOOP: return "ELOOP";
+#endif
+#ifdef EMFILE
+    case EMFILE: return "EMFILE";
+#endif
+#ifdef EMLINK
+    case EMLINK: return "EMLINK";
+#endif
+#ifdef EMSGSIZE
+    case EMSGSIZE: return "EMSGSIZE";
+#endif
+#ifdef EMULTIHOP
+    case EMULTIHOP: return "EMULTIHOP";
+#endif
+#ifdef ENAMETOOLONG
+    case ENAMETOOLONG: return "ENAMETOOLONG";
+#endif
+#ifdef ENAVAIL
+    case ENAVAIL: return "ENAVAIL";
+#endif
+#ifdef ENET
+    case ENET: return "ENET";
+#endif
+#ifdef ENETDOWN
+    case ENETDOWN: return "ENETDOWN";
+#endif
+#ifdef ENETRESET
+    case ENETRESET: return "ENETRESET";
+#endif
+#ifdef ENETUNREACH
+    case ENETUNREACH: return "ENETUNREACH";
+#endif
+#ifdef ENFILE
+    case ENFILE: return "ENFILE";
+#endif
+#ifdef ENOANO
+    case ENOANO: return "ENOANO";
+#endif
+#if defined(ENOBUFS) && (!defined(ENOSR) || (ENOBUFS != ENOSR))
+    case ENOBUFS: return "ENOBUFS";
+#endif
+#ifdef ENOCSI
+    case ENOCSI: return "ENOCSI";
+#endif
+#if defined(ENODATA) && (!defined(ECONNREFUSED) || (ENODATA != ECONNREFUSED))
+    case ENODATA: return "ENODATA";
+#endif
+#ifdef ENODEV
+    case ENODEV: return "ENODEV";
+#endif
+#ifdef ENOENT
+    case ENOENT: return "ENOENT";
+#endif
+#ifdef ENOEXEC
+    case ENOEXEC: return "ENOEXEC";
+#endif
+#ifdef ENOLCK
+    case ENOLCK: return "ENOLCK";
+#endif
+#ifdef ENOLINK
+    case ENOLINK: return "ENOLINK";
+#endif
+#ifdef ENOMEM
+    case ENOMEM: return "ENOMEM";
+#endif
+#ifdef ENOMSG
+    case ENOMSG: return "ENOMSG";
+#endif
+#ifdef ENONET
+    case ENONET: return "ENONET";
+#endif
+#ifdef ENOPKG
+    case ENOPKG: return "ENOPKG";
+#endif
+#ifdef ENOPROTOOPT
+    case ENOPROTOOPT: return "ENOPROTOOPT";
+#endif
+#ifdef ENOSPC
+    case ENOSPC: return "ENOSPC";
+#endif
+#if defined(ENOSR) && (!defined(ENAMETOOLONG) || (ENAMETOOLONG != ENOSR))
+    case ENOSR: return "ENOSR";
+#endif
+#if defined(ENOSTR) && (!defined(ENOTTY) || (ENOTTY != ENOSTR))
+    case ENOSTR: return "ENOSTR";
+#endif
+#ifdef ENOSYM
+    case ENOSYM: return "ENOSYM";
+#endif
+#ifdef ENOSYS
+    case ENOSYS: return "ENOSYS";
+#endif
+#ifdef ENOTBLK
+    case ENOTBLK: return "ENOTBLK";
+#endif
+#ifdef ENOTCONN
+    case ENOTCONN: return "ENOTCONN";
+#endif
+#ifdef ENOTRECOVERABLE
+    case ENOTRECOVERABLE: return "ENOTRECOVERABLE";
+#endif
+#ifdef ENOTDIR
+    case ENOTDIR: return "ENOTDIR";
+#endif
+#if defined(ENOTEMPTY) && (!defined(EEXIST) || (ENOTEMPTY != EEXIST))
+    case ENOTEMPTY: return "ENOTEMPTY";
+#endif
+#ifdef ENOTNAM
+    case ENOTNAM: return "ENOTNAM";
+#endif
+#ifdef ENOTSOCK
+    case ENOTSOCK: return "ENOTSOCK";
+#endif
+#ifdef ENOTSUP
+    case ENOTSUP: return "ENOTSUP";
+#endif
+#ifdef ENOTTY
+    case ENOTTY: return "ENOTTY";
+#endif
+#ifdef ENOTUNIQ
+    case ENOTUNIQ: return "ENOTUNIQ";
+#endif
+#ifdef ENXIO
+    case ENXIO: return "ENXIO";
+#endif
+#if defined(EOPNOTSUPP) &&  (!defined(ENOTSUP) || (ENOTSUP != EOPNOTSUPP))
+    case EOPNOTSUPP: return "EOPNOTSUPP";
+#endif
+#ifdef EOTHER
+    case EOTHER: return "EOTHER";
+#endif
+#if defined(EOVERFLOW) && (!defined(EFBIG) || (EOVERFLOW != EFBIG)) && (!defined(EINVAL) || (EOVERFLOW != EINVAL))
+    case EOVERFLOW: return "EOVERFLOW";
+#endif
+#ifdef EOWNERDEAD
+    case EOWNERDEAD: return "EOWNERDEAD";
+#endif
+#ifdef EPERM
+    case EPERM: return "EPERM";
+#endif
+#if defined(EPFNOSUPPORT) && (!defined(ENOLCK) || (ENOLCK != EPFNOSUPPORT))
+    case EPFNOSUPPORT: return "EPFNOSUPPORT";
+#endif
+#ifdef EPIPE
+    case EPIPE: return "EPIPE";
+#endif
+#ifdef EPROCLIM
+    case EPROCLIM: return "EPROCLIM";
+#endif
+#ifdef EPROCUNAVAIL
+    case EPROCUNAVAIL: return "EPROCUNAVAIL";
+#endif
+#ifdef EPROGMISMATCH
+    case EPROGMISMATCH: return "EPROGMISMATCH";
+#endif
+#ifdef EPROGUNAVAIL
+    case EPROGUNAVAIL: return "EPROGUNAVAIL";
+#endif
+#ifdef EPROTO
+    case EPROTO: return "EPROTO";
+#endif
+#ifdef EPROTONOSUPPORT
+    case EPROTONOSUPPORT: return "EPROTONOSUPPORT";
+#endif
+#ifdef EPROTOTYPE
+    case EPROTOTYPE: return "EPROTOTYPE";
+#endif
+#ifdef ERANGE
+    case ERANGE: return "ERANGE";
+#endif
+#if defined(EREFUSED) && (!defined(ECONNREFUSED) || (EREFUSED != ECONNREFUSED))
+    case EREFUSED: return "EREFUSED";
+#endif
+#ifdef EREMCHG
+    case EREMCHG: return "EREMCHG";
+#endif
+#ifdef EREMDEV
+    case EREMDEV: return "EREMDEV";
+#endif
+#ifdef EREMOTE
+    case EREMOTE: return "EREMOTE";
+#endif
+#ifdef EREMOTEIO
+    case EREMOTEIO: return "EREMOTEIO";
+#endif
+#ifdef EREMOTERELEASE
+    case EREMOTERELEASE: return "EREMOTERELEASE";
+#endif
+#ifdef EROFS
+    case EROFS: return "EROFS";
+#endif
+#ifdef ERPCMISMATCH
+    case ERPCMISMATCH: return "ERPCMISMATCH";
+#endif
+#ifdef ERREMOTE
+    case ERREMOTE: return "ERREMOTE";
+#endif
+#ifdef ESHUTDOWN
+    case ESHUTDOWN: return "ESHUTDOWN";
+#endif
+#ifdef ESOCKTNOSUPPORT
+    case ESOCKTNOSUPPORT: return "ESOCKTNOSUPPORT";
+#endif
+#ifdef ESPIPE
+    case ESPIPE: return "ESPIPE";
+#endif
+#ifdef ESRCH
+    case ESRCH: return "ESRCH";
+#endif
+#ifdef ESRMNT
+    case ESRMNT: return "ESRMNT";
+#endif
+#ifdef ESTALE
+    case ESTALE: return "ESTALE";
+#endif
+#ifdef ESUCCESS
+    case ESUCCESS: return "ESUCCESS";
+#endif
+#if defined(ETIME) && (!defined(ELOOP) || (ETIME != ELOOP))
+    case ETIME: return "ETIME";
+#endif
+#if defined(ETIMEDOUT) && (!defined(ENOSTR) || (ETIMEDOUT != ENOSTR))
+    case ETIMEDOUT: return "ETIMEDOUT";
+#endif
+#ifdef ETOOMANYREFS
+    case ETOOMANYREFS: return "ETOOMANYREFS";
+#endif
+#ifdef ETXTBSY
+    case ETXTBSY: return "ETXTBSY";
+#endif
+#ifdef EUCLEAN
+    case EUCLEAN: return "EUCLEAN";
+#endif
+#ifdef EUNATCH
+    case EUNATCH: return "EUNATCH";
+#endif
+#ifdef EUSERS
+    case EUSERS: return "EUSERS";
+#endif
+#ifdef EVERSION
+    case EVERSION: return "EVERSION";
+#endif
+#if defined(EWOULDBLOCK) && (!defined(EAGAIN) || (EWOULDBLOCK != EAGAIN))
+    case EWOULDBLOCK: return "EWOULDBLOCK";
+#endif
+#ifdef EXDEV
+    case EXDEV: return "EXDEV";
+#endif
+#ifdef EXFULL
+    case EXFULL: return "EXFULL";
+#endif
+    }
+    return "unknown error";
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ErrnoMsg --
+ *
+ *	Return a human-readable message corresponding to a given errno value.
+ *
+ * Results:
+ *	The return value is the standard POSIX error message for errno.  This
+ *	procedure is used instead of strerror because strerror returns
+ *	slightly different values on different machines (e.g. different
+ *	capitalizations), which cause problems for things such as regression
+ *	tests. This procedure provides messages for most standard errors, then
+ *	it calls strerror for things it doesn't understand.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_ErrnoMsg(
+     int err)			/* Error number (such as in errno variable). */
+{
+    switch (err) {
+#if defined(E2BIG) && (!defined(EOVERFLOW) || (E2BIG != EOVERFLOW))
+    case E2BIG: return "argument list too long";
+#endif
+#ifdef EACCES
+    case EACCES: return "permission denied";
+#endif
+#ifdef EADDRINUSE
+    case EADDRINUSE: return "address already in use";
+#endif
+#ifdef EADDRNOTAVAIL
+    case EADDRNOTAVAIL: return "cannot assign requested address";
+#endif
+#ifdef EADV
+    case EADV: return "advertise error";
+#endif
+#ifdef EAFNOSUPPORT
+    case EAFNOSUPPORT: return "address family not supported by protocol";
+#endif
+#ifdef EAGAIN
+    case EAGAIN: return "resource temporarily unavailable";
+#endif
+#ifdef EALIGN
+    case EALIGN: return "EALIGN";
+#endif
+#if defined(EALREADY) && (!defined(EBUSY) || (EALREADY != EBUSY))
+    case EALREADY: return "operation already in progress";
+#endif
+#ifdef EBADE
+    case EBADE: return "bad exchange descriptor";
+#endif
+#ifdef EBADF
+    case EBADF: return "bad file number";
+#endif
+#ifdef EBADFD
+    case EBADFD: return "file descriptor in bad state";
+#endif
+#ifdef EBADMSG
+    case EBADMSG: return "not a data message";
+#endif
+#ifdef ECANCELED
+    case ECANCELED: return "operation canceled";
+#endif
+#ifdef EBADR
+    case EBADR: return "bad request descriptor";
+#endif
+#ifdef EBADRPC
+    case EBADRPC: return "RPC structure is bad";
+#endif
+#ifdef EBADRQC
+    case EBADRQC: return "bad request code";
+#endif
+#ifdef EBADSLT
+    case EBADSLT: return "invalid slot";
+#endif
+#ifdef EBFONT
+    case EBFONT: return "bad font file format";
+#endif
+#ifdef EBUSY
+    case EBUSY: return "file busy";
+#endif
+#ifdef ECHILD
+    case ECHILD: return "no children";
+#endif
+#ifdef ECHRNG
+    case ECHRNG: return "channel number out of range";
+#endif
+#ifdef ECOMM
+    case ECOMM: return "communication error on send";
+#endif
+#ifdef ECONNABORTED
+    case ECONNABORTED: return "software caused connection abort";
+#endif
+#ifdef ECONNREFUSED
+    case ECONNREFUSED: return "connection refused";
+#endif
+#ifdef ECONNRESET
+    case ECONNRESET: return "connection reset by peer";
+#endif
+#if defined(EDEADLK) && (!defined(EWOULDBLOCK) || (EDEADLK != EWOULDBLOCK))
+    case EDEADLK: return "resource deadlock avoided";
+#endif
+#if defined(EDEADLOCK) && (!defined(EDEADLK) || (EDEADLOCK != EDEADLK))
+    case EDEADLOCK: return "resource deadlock avoided";
+#endif
+#ifdef EDESTADDRREQ
+    case EDESTADDRREQ: return "destination address required";
+#endif
+#ifdef EDIRTY
+    case EDIRTY: return "mounting a dirty fs w/o force";
+#endif
+#ifdef EDOM
+    case EDOM: return "math argument out of range";
+#endif
+#ifdef EDOTDOT
+    case EDOTDOT: return "cross mount point";
+#endif
+#ifdef EDQUOT
+    case EDQUOT: return "disk quota exceeded";
+#endif
+#ifdef EDUPPKG
+    case EDUPPKG: return "duplicate package name";
+#endif
+#ifdef EEXIST
+    case EEXIST: return "file already exists";
+#endif
+#ifdef EFAULT
+    case EFAULT: return "bad address in system call argument";
+#endif
+#ifdef EFBIG
+    case EFBIG: return "file too large";
+#endif
+#ifdef EHOSTDOWN
+    case EHOSTDOWN: return "host is down";
+#endif
+#ifdef EHOSTUNREACH
+    case EHOSTUNREACH: return "host is unreachable";
+#endif
+#if defined(EIDRM) && (!defined(EINPROGRESS) || (EIDRM != EINPROGRESS))
+    case EIDRM: return "identifier removed";
+#endif
+#ifdef EINIT
+    case EINIT: return "initialization error";
+#endif
+#ifdef EINPROGRESS
+    case EINPROGRESS: return "operation now in progress";
+#endif
+#ifdef EINTR
+    case EINTR: return "interrupted system call";
+#endif
+#ifdef EINVAL
+    case EINVAL: return "invalid argument";
+#endif
+#ifdef EIO
+    case EIO: return "I/O error";
+#endif
+#ifdef EISCONN
+    case EISCONN: return "socket is already connected";
+#endif
+#ifdef EISDIR
+    case EISDIR: return "illegal operation on a directory";
+#endif
+#ifdef EISNAME
+    case EISNAM: return "is a name file";
+#endif
+#ifdef ELBIN
+    case ELBIN: return "ELBIN";
+#endif
+#ifdef EL2HLT
+    case EL2HLT: return "level 2 halted";
+#endif
+#ifdef EL2NSYNC
+    case EL2NSYNC: return "level 2 not synchronized";
+#endif
+#ifdef EL3HLT
+    case EL3HLT: return "level 3 halted";
+#endif
+#ifdef EL3RST
+    case EL3RST: return "level 3 reset";
+#endif
+#ifdef ELIBACC
+    case ELIBACC: return "cannot access a needed shared library";
+#endif
+#ifdef ELIBBAD
+    case ELIBBAD: return "accessing a corrupted shared library";
+#endif
+#ifdef ELIBEXEC
+    case ELIBEXEC: return "cannot exec a shared library directly";
+#endif
+#if defined(ELIBMAX) && (!defined(ECANCELED) || (ELIBMAX != ECANCELED))
+    case ELIBMAX: return
+	    "attempting to link in more shared libraries than system limit";
+#endif
+#ifdef ELIBSCN
+    case ELIBSCN: return ".lib section in a.out corrupted";
+#endif
+#ifdef ELNRNG
+    case ELNRNG: return "link number out of range";
+#endif
+#if defined(ELOOP) && (!defined(ENOENT) || (ELOOP != ENOENT))
+    case ELOOP: return "too many levels of symbolic links";
+#endif
+#ifdef EMFILE
+    case EMFILE: return "too many open files";
+#endif
+#ifdef EMLINK
+    case EMLINK: return "too many links";
+#endif
+#ifdef EMSGSIZE
+    case EMSGSIZE: return "message too long";
+#endif
+#ifdef EMULTIHOP
+    case EMULTIHOP: return "multihop attempted";
+#endif
+#ifdef ENAMETOOLONG
+    case ENAMETOOLONG: return "file name too long";
+#endif
+#ifdef ENAVAIL
+    case ENAVAIL: return "not available";
+#endif
+#ifdef ENET
+    case ENET: return "ENET";
+#endif
+#ifdef ENETDOWN
+    case ENETDOWN: return "network is down";
+#endif
+#ifdef ENETRESET
+    case ENETRESET: return "network dropped connection on reset";
+#endif
+#ifdef ENETUNREACH
+    case ENETUNREACH: return "network is unreachable";
+#endif
+#ifdef ENFILE
+    case ENFILE: return "file table overflow";
+#endif
+#ifdef ENOANO
+    case ENOANO: return "anode table overflow";
+#endif
+#if defined(ENOBUFS) && (!defined(ENOSR) || (ENOBUFS != ENOSR))
+    case ENOBUFS: return "no buffer space available";
+#endif
+#ifdef ENOCSI
+    case ENOCSI: return "no CSI structure available";
+#endif
+#if defined(ENODATA) && (!defined(ECONNREFUSED) || (ENODATA != ECONNREFUSED))
+    case ENODATA: return "no data available";
+#endif
+#ifdef ENODEV
+    case ENODEV: return "no such device";
+#endif
+#ifdef ENOENT
+    case ENOENT: return "no such file or directory";
+#endif
+#ifdef ENOEXEC
+    case ENOEXEC: return "exec format error";
+#endif
+#ifdef ENOLCK
+    case ENOLCK: return "no locks available";
+#endif
+#ifdef ENOLINK
+    case ENOLINK: return "link has been severed";
+#endif
+#ifdef ENOMEM
+    case ENOMEM: return "not enough memory";
+#endif
+#ifdef ENOMSG
+    case ENOMSG: return "no message of desired type";
+#endif
+#ifdef ENONET
+    case ENONET: return "machine is not on the network";
+#endif
+#ifdef ENOPKG
+    case ENOPKG: return "package not installed";
+#endif
+#ifdef ENOPROTOOPT
+    case ENOPROTOOPT: return "bad protocol option";
+#endif
+#ifdef ENOSPC
+    case ENOSPC: return "no space left on device";
+#endif
+#if defined(ENOSR) && (!defined(ENAMETOOLONG) || (ENAMETOOLONG != ENOSR))
+    case ENOSR: return "out of stream resources";
+#endif
+#if defined(ENOSTR) && (!defined(ENOTTY) || (ENOTTY != ENOSTR))
+    case ENOSTR: return "not a stream device";
+#endif
+#ifdef ENOSYM
+    case ENOSYM: return "unresolved symbol name";
+#endif
+#ifdef ENOSYS
+    case ENOSYS: return "function not implemented";
+#endif
+#ifdef ENOTBLK
+    case ENOTBLK: return "block device required";
+#endif
+#ifdef ENOTCONN
+    case ENOTCONN: return "socket is not connected";
+#endif
+#ifdef ENOTRECOVERABLE
+    case ENOTRECOVERABLE: return "state not recoverable";
+#endif
+#ifdef ENOTDIR
+    case ENOTDIR: return "not a directory";
+#endif
+#if defined(ENOTEMPTY) && (!defined(EEXIST) || (ENOTEMPTY != EEXIST))
+    case ENOTEMPTY: return "directory not empty";
+#endif
+#ifdef ENOTNAM
+    case ENOTNAM: return "not a name file";
+#endif
+#ifdef ENOTSOCK
+    case ENOTSOCK: return "socket operation on non-socket";
+#endif
+#ifdef ENOTSUP
+    case ENOTSUP: return "operation not supported";
+#endif
+#ifdef ENOTTY
+    case ENOTTY: return "inappropriate device for ioctl";
+#endif
+#ifdef ENOTUNIQ
+    case ENOTUNIQ: return "name not unique on network";
+#endif
+#ifdef ENXIO
+    case ENXIO: return "no such device or address";
+#endif
+#if defined(EOPNOTSUPP) &&  (!defined(ENOTSUP) || (ENOTSUP != EOPNOTSUPP))
+    case EOPNOTSUPP: return "operation not supported on socket";
+#endif
+#ifdef EOTHER
+    case EOTHER: return "other error";
+#endif
+#if defined(EOVERFLOW) && (!defined(EFBIG) || (EOVERFLOW != EFBIG)) && (!defined(EINVAL) || (EOVERFLOW != EINVAL))
+    case EOVERFLOW: return "file too big";
+#endif
+#ifdef EOWNERDEAD
+    case EOWNERDEAD: return "owner died";
+#endif
+#ifdef EPERM
+    case EPERM: return "not owner";
+#endif
+#if defined(EPFNOSUPPORT) && (!defined(ENOLCK) || (ENOLCK != EPFNOSUPPORT))
+    case EPFNOSUPPORT: return "protocol family not supported";
+#endif
+#ifdef EPIPE
+    case EPIPE: return "broken pipe";
+#endif
+#ifdef EPROCLIM
+    case EPROCLIM: return "too many processes";
+#endif
+#ifdef EPROCUNAVAIL
+    case EPROCUNAVAIL: return "bad procedure for program";
+#endif
+#ifdef EPROGMISMATCH
+    case EPROGMISMATCH: return "program version wrong";
+#endif
+#ifdef EPROGUNAVAIL
+    case EPROGUNAVAIL: return "RPC program not available";
+#endif
+#ifdef EPROTO
+    case EPROTO: return "protocol error";
+#endif
+#ifdef EPROTONOSUPPORT
+    case EPROTONOSUPPORT: return "protocol not supported";
+#endif
+#ifdef EPROTOTYPE
+    case EPROTOTYPE: return "protocol wrong type for socket";
+#endif
+#ifdef ERANGE
+    case ERANGE: return "math result unrepresentable";
+#endif
+#if defined(EREFUSED) && (!defined(ECONNREFUSED) || (EREFUSED != ECONNREFUSED))
+    case EREFUSED: return "EREFUSED";
+#endif
+#ifdef EREMCHG
+    case EREMCHG: return "remote address changed";
+#endif
+#ifdef EREMDEV
+    case EREMDEV: return "remote device";
+#endif
+#ifdef EREMOTE
+    case EREMOTE: return "pathname hit remote file system";
+#endif
+#ifdef EREMOTEIO
+    case EREMOTEIO: return "remote i/o error";
+#endif
+#ifdef EREMOTERELEASE
+    case EREMOTERELEASE: return "EREMOTERELEASE";
+#endif
+#ifdef EROFS
+    case EROFS: return "read-only file system";
+#endif
+#ifdef ERPCMISMATCH
+    case ERPCMISMATCH: return "RPC version is wrong";
+#endif
+#ifdef ERREMOTE
+    case ERREMOTE: return "object is remote";
+#endif
+#ifdef ESHUTDOWN
+    case ESHUTDOWN: return "cannot send after socket shutdown";
+#endif
+#ifdef ESOCKTNOSUPPORT
+    case ESOCKTNOSUPPORT: return "socket type not supported";
+#endif
+#ifdef ESPIPE
+    case ESPIPE: return "invalid seek";
+#endif
+#ifdef ESRCH
+    case ESRCH: return "no such process";
+#endif
+#ifdef ESRMNT
+    case ESRMNT: return "srmount error";
+#endif
+#ifdef ESTALE
+    case ESTALE: return "stale remote file handle";
+#endif
+#ifdef ESUCCESS
+    case ESUCCESS: return "Error 0";
+#endif
+#if defined(ETIME) && (!defined(ELOOP) || (ETIME != ELOOP))
+    case ETIME: return "timer expired";
+#endif
+#if defined(ETIMEDOUT) && (!defined(ENOSTR) || (ETIMEDOUT != ENOSTR))
+    case ETIMEDOUT: return "connection timed out";
+#endif
+#ifdef ETOOMANYREFS
+    case ETOOMANYREFS: return "too many references: cannot splice";
+#endif
+#ifdef ETXTBSY
+    case ETXTBSY: return "text file or pseudo-device busy";
+#endif
+#ifdef EUCLEAN
+    case EUCLEAN: return "structure needs cleaning";
+#endif
+#ifdef EUNATCH
+    case EUNATCH: return "protocol driver not attached";
+#endif
+#ifdef EUSERS
+    case EUSERS: return "too many users";
+#endif
+#ifdef EVERSION
+    case EVERSION: return "version mismatch";
+#endif
+#if defined(EWOULDBLOCK) && (!defined(EAGAIN) || (EWOULDBLOCK != EAGAIN))
+    case EWOULDBLOCK: return "operation would block";
+#endif
+#ifdef EXDEV
+    case EXDEV: return "cross-domain link";
+#endif
+#ifdef EXFULL
+    case EXFULL: return "message tables full";
+#endif
+    default:
+#ifdef NO_STRERROR
+	return "unknown POSIX error";
+#else
+	return strerror(err);
+#endif
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SignalId --
+ *
+ *	Return a textual identifier for a signal number.
+ *
+ * Results:
+ *	This procedure returns a machine-readable textual identifier that
+ *	corresponds to sig.  The identifier is the same as the #define name in
+ *	signal.h.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_SignalId(
+     int sig)			/* Number of signal. */
+{
+    switch (sig) {
+#ifdef SIGABRT
+    case SIGABRT: return "SIGABRT";
+#endif
+#ifdef SIGALRM
+    case SIGALRM: return "SIGALRM";
+#endif
+#ifdef SIGBUS
+    case SIGBUS: return "SIGBUS";
+#endif
+#ifdef SIGCHLD
+    case SIGCHLD: return "SIGCHLD";
+#endif
+#if defined(SIGCLD) && (!defined(SIGCHLD) || (SIGCLD != SIGCHLD))
+    case SIGCLD: return "SIGCLD";
+#endif
+#ifdef SIGCONT
+    case SIGCONT: return "SIGCONT";
+#endif
+#if defined(SIGEMT) && (!defined(SIGXCPU) || (SIGEMT != SIGXCPU))
+    case SIGEMT: return "SIGEMT";
+#endif
+#ifdef SIGFPE
+    case SIGFPE: return "SIGFPE";
+#endif
+#ifdef SIGHUP
+    case SIGHUP: return "SIGHUP";
+#endif
+#ifdef SIGILL
+    case SIGILL: return "SIGILL";
+#endif
+#ifdef SIGINT
+    case SIGINT: return "SIGINT";
+#endif
+#ifdef SIGIO
+    case SIGIO: return "SIGIO";
+#endif
+#if defined(SIGIOT) && (!defined(SIGABRT) || (SIGIOT != SIGABRT))
+    case SIGIOT: return "SIGIOT";
+#endif
+#ifdef SIGKILL
+    case SIGKILL: return "SIGKILL";
+#endif
+#if defined(SIGLOST) && (!defined(SIGIOT) || (SIGLOST != SIGIOT)) && (!defined(SIGURG) || (SIGLOST != SIGURG)) && (!defined(SIGPROF) || (SIGLOST != SIGPROF)) && (!defined(SIGIO) || (SIGLOST != SIGIO))
+    case SIGLOST: return "SIGLOST";
+#endif
+#ifdef SIGPIPE
+    case SIGPIPE: return "SIGPIPE";
+#endif
+#if defined(SIGPOLL) && (!defined(SIGIO) || (SIGPOLL != SIGIO))
+    case SIGPOLL: return "SIGPOLL";
+#endif
+#ifdef SIGPROF
+    case SIGPROF: return "SIGPROF";
+#endif
+#if defined(SIGPWR) && (!defined(SIGXFSZ) || (SIGPWR != SIGXFSZ)) && (!defined(SIGLOST) || (SIGPWR != SIGLOST))
+    case SIGPWR: return "SIGPWR";
+#endif
+#ifdef SIGQUIT
+    case SIGQUIT: return "SIGQUIT";
+#endif
+#if defined(SIGSEGV) && (!defined(SIGBUS) || (SIGSEGV != SIGBUS))
+    case SIGSEGV: return "SIGSEGV";
+#endif
+#ifdef SIGSTOP
+    case SIGSTOP: return "SIGSTOP";
+#endif
+#ifdef SIGSYS
+    case SIGSYS: return "SIGSYS";
+#endif
+#ifdef SIGTERM
+    case SIGTERM: return "SIGTERM";
+#endif
+#ifdef SIGTRAP
+    case SIGTRAP: return "SIGTRAP";
+#endif
+#ifdef SIGTSTP
+    case SIGTSTP: return "SIGTSTP";
+#endif
+#ifdef SIGTTIN
+    case SIGTTIN: return "SIGTTIN";
+#endif
+#ifdef SIGTTOU
+    case SIGTTOU: return "SIGTTOU";
+#endif
+#if defined(SIGURG) && (!defined(SIGIO) || (SIGURG != SIGIO))
+    case SIGURG: return "SIGURG";
+#endif
+#if defined(SIGUSR1) && (!defined(SIGIO) || (SIGUSR1 != SIGIO))
+    case SIGUSR1: return "SIGUSR1";
+#endif
+#if defined(SIGUSR2) && (!defined(SIGURG) || (SIGUSR2 != SIGURG))
+    case SIGUSR2: return "SIGUSR2";
+#endif
+#ifdef SIGVTALRM
+    case SIGVTALRM: return "SIGVTALRM";
+#endif
+#ifdef SIGWINCH
+    case SIGWINCH: return "SIGWINCH";
+#endif
+#ifdef SIGXCPU
+    case SIGXCPU: return "SIGXCPU";
+#endif
+#ifdef SIGXFSZ
+    case SIGXFSZ: return "SIGXFSZ";
+#endif
+#if defined(SIGINFO) && (!defined(SIGPWR) || (SIGINFO != SIGPWR))
+    case SIGINFO: return "SIGINFO";
+#endif
+    }
+    return "unknown signal";
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SignalMsg --
+ *
+ *	Return a human-readable message describing a signal.
+ *
+ * Results:
+ *	This procedure returns a string describing sig that should make sense
+ *	to a human.  It may not be easy for a machine to parse.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_SignalMsg(
+     int sig)			/* Number of signal. */
+{
+    switch (sig) {
+#ifdef SIGABRT
+    case SIGABRT: return "SIGABRT";
+#endif
+#ifdef SIGALRM
+    case SIGALRM: return "alarm clock";
+#endif
+#ifdef SIGBUS
+    case SIGBUS: return "bus error";
+#endif
+#ifdef SIGCHLD
+    case SIGCHLD: return "child status changed";
+#endif
+#if defined(SIGCLD) && (!defined(SIGCHLD) || (SIGCLD != SIGCHLD))
+    case SIGCLD: return "child status changed";
+#endif
+#ifdef SIGCONT
+    case SIGCONT: return "continue after stop";
+#endif
+#if defined(SIGEMT) && (!defined(SIGXCPU) || (SIGEMT != SIGXCPU))
+    case SIGEMT: return "EMT instruction";
+#endif
+#ifdef SIGFPE
+    case SIGFPE: return "floating-point exception";
+#endif
+#ifdef SIGHUP
+    case SIGHUP: return "hangup";
+#endif
+#ifdef SIGILL
+    case SIGILL: return "illegal instruction";
+#endif
+#ifdef SIGINT
+    case SIGINT: return "interrupt";
+#endif
+#ifdef SIGIO
+    case SIGIO: return "input/output possible on file";
+#endif
+#if defined(SIGIOT) && (!defined(SIGABRT) || (SIGABRT != SIGIOT))
+    case SIGIOT: return "IOT instruction";
+#endif
+#ifdef SIGKILL
+    case SIGKILL: return "kill signal";
+#endif
+#if defined(SIGLOST) && (!defined(SIGIOT) || (SIGLOST != SIGIOT)) && (!defined(SIGURG) || (SIGLOST != SIGURG)) && (!defined(SIGPROF) || (SIGLOST != SIGPROF)) && (!defined(SIGIO) || (SIGLOST != SIGIO))
+    case SIGLOST: return "resource lost";
+#endif
+#ifdef SIGPIPE
+    case SIGPIPE: return "write on pipe with no readers";
+#endif
+#if defined(SIGPOLL) && (!defined(SIGIO) || (SIGPOLL != SIGIO))
+    case SIGPOLL: return "input/output possible on file";
+#endif
+#ifdef SIGPROF
+    case SIGPROF: return "profiling alarm";
+#endif
+#if defined(SIGPWR) && (!defined(SIGXFSZ) || (SIGPWR != SIGXFSZ)) && (!defined(SIGLOST) || (SIGPWR != SIGLOST))
+    case SIGPWR: return "power-fail restart";
+#endif
+#ifdef SIGQUIT
+    case SIGQUIT: return "quit signal";
+#endif
+#if defined(SIGSEGV) && (!defined(SIGBUS) || (SIGSEGV != SIGBUS))
+    case SIGSEGV: return "segmentation violation";
+#endif
+#ifdef SIGSTOP
+    case SIGSTOP: return "stop";
+#endif
+#ifdef SIGSYS
+    case SIGSYS: return "bad argument to system call";
+#endif
+#ifdef SIGTERM
+    case SIGTERM: return "software termination signal";
+#endif
+#ifdef SIGTRAP
+    case SIGTRAP: return "trace trap";
+#endif
+#ifdef SIGTSTP
+    case SIGTSTP: return "stop signal from tty";
+#endif
+#ifdef SIGTTIN
+    case SIGTTIN: return "background tty read";
+#endif
+#ifdef SIGTTOU
+    case SIGTTOU: return "background tty write";
+#endif
+#if defined(SIGURG) && (!defined(SIGIO) || (SIGURG != SIGIO))
+    case SIGURG: return "urgent I/O condition";
+#endif
+#if defined(SIGUSR1) && (!defined(SIGIO) || (SIGUSR1 != SIGIO))
+    case SIGUSR1: return "user-defined signal 1";
+#endif
+#if defined(SIGUSR2) && (!defined(SIGURG) || (SIGUSR2 != SIGURG))
+    case SIGUSR2: return "user-defined signal 2";
+#endif
+#ifdef SIGVTALRM
+    case SIGVTALRM: return "virtual time alarm";
+#endif
+#ifdef SIGWINCH
+    case SIGWINCH: return "window changed";
+#endif
+#ifdef SIGXCPU
+    case SIGXCPU: return "exceeded CPU time limit";
+#endif
+#ifdef SIGXFSZ
+    case SIGXFSZ: return "exceeded file size limit";
+#endif
+#if defined(SIGINFO) && (!defined(SIGPWR) || (SIGINFO != SIGPWR))
+    case SIGINFO: return "information request";
+#endif
+    }
+    return "unknown signal";
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclPreserve.c b/generic/tclPreserve.c
new file mode 100644
index 0000000..5c6097f
--- /dev/null
+++ b/generic/tclPreserve.c
@@ -0,0 +1,473 @@
+/*
+ * tclPreserve.c --
+ *
+ *	This file contains a collection of functions that are used to make
+ *	sure that widget records and other data structures aren't reallocated
+ *	when there are nested functions that depend on their existence.
+ *
+ * Copyright (c) 1991-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * The following data structure is used to keep track of all the Tcl_Preserve
+ * calls that are still in effect. It grows as needed to accommodate any
+ * number of calls in effect.
+ */
+
+typedef struct {
+    ClientData clientData;	/* Address of preserved block. */
+    size_t refCount;		/* Number of Tcl_Preserve calls in effect for
+				 * block. */
+    int mustFree;		/* Non-zero means Tcl_EventuallyFree was
+				 * called while a Tcl_Preserve call was in
+				 * effect, so the structure must be freed when
+				 * refCount becomes zero. */
+    Tcl_FreeProc *freeProc;	/* Function to call to free. */
+} Reference;
+
+/*
+ * Global data structures used to hold the list of preserved data references.
+ * These variables are protected by "preserveMutex".
+ */
+
+static Reference *refArray = NULL;	/* First in array of references. */
+static int spaceAvl = 0;	/* Total number of structures available at
+				 * *firstRefPtr. */
+static int inUse = 0;		/* Count of structures currently in use in
+				 * refArray. */
+TCL_DECLARE_MUTEX(preserveMutex)/* To protect the above statics */
+
+#define INITIAL_SIZE	2	/* Initial number of reference slots to make */
+
+/*
+ * The following data structure is used to keep track of whether an arbitrary
+ * block of memory has been deleted. This is used by the TclHandle code to
+ * avoid the more time-expensive algorithm of Tcl_Preserve(). This mechanism
+ * is mainly used when we have lots of references to a few big, expensive
+ * objects that we don't want to live any longer than necessary.
+ */
+
+typedef struct HandleStruct {
+    void *ptr;			/* Pointer to the memory block being tracked.
+				 * This field will become NULL when the memory
+				 * block is deleted. This field must be the
+				 * first in the structure. */
+#ifdef TCL_MEM_DEBUG
+    void *ptr2;			/* Backup copy of the above pointer used to
+				 * ensure that the contents of the handle are
+				 * not changed by anyone else. */
+#endif
+    size_t refCount;		/* Number of TclHandlePreserve() calls in
+				 * effect on this handle. */
+} HandleStruct;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizePreserve --
+ *
+ *	Called during exit processing to clean up the reference array.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees the storage of the reference array.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+void
+TclFinalizePreserve(void)
+{
+    Tcl_MutexLock(&preserveMutex);
+    if (spaceAvl != 0) {
+	ckfree(refArray);
+	refArray = NULL;
+	inUse = 0;
+	spaceAvl = 0;
+    }
+    Tcl_MutexUnlock(&preserveMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Preserve --
+ *
+ *	This function is used by a function to declare its interest in a
+ *	particular block of memory, so that the block will not be reallocated
+ *	until a matching call to Tcl_Release has been made.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Information is retained so that the block of memory will not be freed
+ *	until at least the matching call to Tcl_Release.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_Preserve(
+    ClientData clientData)	/* Pointer to malloc'ed block of memory. */
+{
+    Reference *refPtr;
+    int i;
+
+    /*
+     * See if there is already a reference for this pointer. If so, just
+     * increment its reference count.
+     */
+
+    Tcl_MutexLock(&preserveMutex);
+    for (i=0, refPtr=refArray ; i<inUse ; i++, refPtr++) {
+	if (refPtr->clientData == clientData) {
+	    refPtr->refCount++;
+	    Tcl_MutexUnlock(&preserveMutex);
+	    return;
+	}
+    }
+
+    /*
+     * Make a reference array if it doesn't already exist, or make it bigger
+     * if it is full.
+     */
+
+    if (inUse == spaceAvl) {
+	spaceAvl = spaceAvl ? 2*spaceAvl : INITIAL_SIZE;
+	refArray = ckrealloc(refArray, spaceAvl * sizeof(Reference));
+    }
+
+    /*
+     * Make a new entry for the new reference.
+     */
+
+    refPtr = &refArray[inUse];
+    refPtr->clientData = clientData;
+    refPtr->refCount = 1;
+    refPtr->mustFree = 0;
+    refPtr->freeProc = 0;
+    inUse += 1;
+    Tcl_MutexUnlock(&preserveMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Release --
+ *
+ *	This function is called to cancel a previous call to Tcl_Preserve,
+ *	thereby allowing a block of memory to be freed (if no one else cares
+ *	about it).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If Tcl_EventuallyFree has been called for clientData, and if no other
+ *	call to Tcl_Preserve is still in effect, the block of memory is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_Release(
+    ClientData clientData)	/* Pointer to malloc'ed block of memory. */
+{
+    Reference *refPtr;
+    int i;
+
+    Tcl_MutexLock(&preserveMutex);
+    for (i=0, refPtr=refArray ; i<inUse ; i++, refPtr++) {
+	int mustFree;
+	Tcl_FreeProc *freeProc;
+
+	if (refPtr->clientData != clientData) {
+	    continue;
+	}
+
+	if (refPtr->refCount-- > 1) {
+	    Tcl_MutexUnlock(&preserveMutex);
+	    return;
+	}
+
+	/*
+	 * Must remove information from the slot before calling freeProc to
+	 * avoid reentrancy problems if the freeProc calls Tcl_Preserve on the
+	 * same clientData. Copy down the last reference in the array to
+	 * overwrite the current slot.
+	 */
+
+	freeProc = refPtr->freeProc;
+	mustFree = refPtr->mustFree;
+	inUse--;
+	if (i < inUse) {
+	    refArray[i] = refArray[inUse];
+	}
+
+	/*
+	 * Now committed to disposing the data. But first, we've patched up
+	 * all the global data structures so we should release the mutex now.
+	 * Only then should we dabble around with potentially-slow memory
+	 * managers...
+	 */
+
+	Tcl_MutexUnlock(&preserveMutex);
+	if (mustFree) {
+	    if (freeProc == TCL_DYNAMIC) {
+		ckfree(clientData);
+	    } else {
+		freeProc(clientData);
+	    }
+	}
+	return;
+    }
+    Tcl_MutexUnlock(&preserveMutex);
+
+    /*
+     * Reference not found. This is a bug in the caller.
+     */
+
+    Tcl_Panic("Tcl_Release couldn't find reference for %p", clientData);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_EventuallyFree --
+ *
+ *	Free up a block of memory, unless a call to Tcl_Preserve is in effect
+ *	for that block. In this case, defer the free until all calls to
+ *	Tcl_Preserve have been undone by matching calls to Tcl_Release.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Ptr may be released by calling free().
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_EventuallyFree(
+    ClientData clientData,	/* Pointer to malloc'ed block of memory. */
+    Tcl_FreeProc *freeProc)	/* Function to actually do free. */
+{
+    Reference *refPtr;
+    int i;
+
+    /*
+     * See if there is a reference for this pointer. If so, set its "mustFree"
+     * flag (the flag had better not be set already!).
+     */
+
+    Tcl_MutexLock(&preserveMutex);
+    for (i = 0, refPtr = refArray; i < inUse; i++, refPtr++) {
+	if (refPtr->clientData != clientData) {
+	    continue;
+	}
+	if (refPtr->mustFree) {
+	    Tcl_Panic("Tcl_EventuallyFree called twice for %p", clientData);
+	}
+	refPtr->mustFree = 1;
+	refPtr->freeProc = freeProc;
+	Tcl_MutexUnlock(&preserveMutex);
+	return;
+    }
+    Tcl_MutexUnlock(&preserveMutex);
+
+    /*
+     * No reference for this block.  Free it now.
+     */
+
+    if (freeProc == TCL_DYNAMIC) {
+	ckfree(clientData);
+    } else {
+	freeProc(clientData);
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclHandleCreate --
+ *
+ *	Allocate a handle that contains enough information to determine if an
+ *	arbitrary malloc'd block has been deleted. This is used to avoid the
+ *	more time-expensive algorithm of Tcl_Preserve().
+ *
+ * Results:
+ *	The return value is a TclHandle that refers to the given malloc'd
+ *	block. Doubly dereferencing the returned handle will give back the
+ *	pointer to the block, or will give NULL if the block has been deleted.
+ *
+ * Side effects:
+ *	The caller must keep track of this handle (generally by storing it in
+ *	a field in the malloc'd block) and call TclHandleFree() on this handle
+ *	when the block is deleted. Everything else that wishes to keep track
+ *	of whether the malloc'd block has been deleted should use calls to
+ *	TclHandlePreserve() and TclHandleRelease() on the associated handle.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+TclHandle
+TclHandleCreate(
+    void *ptr)			/* Pointer to an arbitrary block of memory to
+				 * be tracked for deletion. Must not be
+				 * NULL. */
+{
+    HandleStruct *handlePtr = ckalloc(sizeof(HandleStruct));
+
+    handlePtr->ptr = ptr;
+#ifdef TCL_MEM_DEBUG
+    handlePtr->ptr2 = ptr;
+#endif
+    handlePtr->refCount = 0;
+    return (TclHandle) handlePtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclHandleFree --
+ *
+ *	Called when the arbitrary malloc'd block associated with the handle is
+ *	being deleted. Modifies the handle so that doubly dereferencing it
+ *	will give NULL. This informs any user of the handle that the block of
+ *	memory formerly referenced by the handle has been freed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If nothing is referring to the handle, the handle will be reclaimed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclHandleFree(
+    TclHandle handle)		/* Previously created handle associated with a
+				 * malloc'd block that is being deleted. The
+				 * handle is modified so that doubly
+				 * dereferencing it will give NULL. */
+{
+    HandleStruct *handlePtr;
+
+    handlePtr = (HandleStruct *) handle;
+#ifdef TCL_MEM_DEBUG
+    if (handlePtr->refCount == 0x61616161) {
+	Tcl_Panic("using previously disposed TclHandle %p", handlePtr);
+    }
+    if (handlePtr->ptr2 != handlePtr->ptr) {
+	Tcl_Panic("someone has changed the block referenced by the handle %p\nfrom %p to %p",
+		handlePtr, handlePtr->ptr2, handlePtr->ptr);
+    }
+#endif
+    handlePtr->ptr = NULL;
+    if (handlePtr->refCount == 0) {
+	ckfree(handlePtr);
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclHandlePreserve --
+ *
+ *	Declare an interest in the arbitrary malloc'd block associated with
+ *	the handle.
+ *
+ * Results:
+ *	The return value is the handle argument, with its ref count
+ *	incremented.
+ *
+ * Side effects:
+ *	For each call to TclHandlePreserve(), there should be a matching call
+ *	to TclHandleRelease() when the caller is no longer interested in the
+ *	malloc'd block associated with the handle.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+TclHandle
+TclHandlePreserve(
+    TclHandle handle)		/* Declare an interest in the block of memory
+				 * referenced by this handle. */
+{
+    HandleStruct *handlePtr;
+
+    handlePtr = (HandleStruct *) handle;
+#ifdef TCL_MEM_DEBUG
+    if (handlePtr->refCount == 0x61616161) {
+	Tcl_Panic("using previously disposed TclHandle %p", handlePtr);
+    }
+    if ((handlePtr->ptr != NULL) && (handlePtr->ptr != handlePtr->ptr2)) {
+	Tcl_Panic("someone has changed the block referenced by the handle %p\nfrom %p to %p",
+		handlePtr, handlePtr->ptr2, handlePtr->ptr);
+    }
+#endif
+    handlePtr->refCount++;
+
+    return handle;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclHandleRelease --
+ *
+ *	This function is called to release an interest in the malloc'd block
+ *	associated with the handle.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The ref count of the handle is decremented. If the malloc'd block has
+ *	been freed and if no one is using the handle any more, the handle will
+ *	be reclaimed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclHandleRelease(
+    TclHandle handle)		/* Unregister interest in the block of memory
+				 * referenced by this handle. */
+{
+    HandleStruct *handlePtr;
+
+    handlePtr = (HandleStruct *) handle;
+#ifdef TCL_MEM_DEBUG
+    if (handlePtr->refCount == 0x61616161) {
+	Tcl_Panic("using previously disposed TclHandle %p", handlePtr);
+    }
+    if ((handlePtr->ptr != NULL) && (handlePtr->ptr != handlePtr->ptr2)) {
+	Tcl_Panic("someone has changed the block referenced by the handle %p\nfrom %p to %p",
+		handlePtr, handlePtr->ptr2, handlePtr->ptr);
+    }
+#endif
+    if ((handlePtr->refCount-- <= 1) && (handlePtr->ptr == NULL)) {
+	ckfree(handlePtr);
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclProc.c b/generic/tclProc.c
new file mode 100644
index 0000000..96bdcf3
--- /dev/null
+++ b/generic/tclProc.c
@@ -0,0 +1,2793 @@
+/*
+ * tclProc.c --
+ *
+ *	This file contains routines that implement Tcl procedures, including
+ *	the "proc" and "uplevel" commands.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2004-2006 Miguel Sofer
+ * Copyright (c) 2007 Daniel A. Steffen <das@users.sourceforge.net>
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclCompile.h"
+
+/*
+ * Variables that are part of the [apply] command implementation and which
+ * have to be passed to the other side of the NRE call.
+ */
+
+typedef struct {
+    Command cmd;
+    ExtraFrameInfo efi;
+} ApplyExtraData;
+
+/*
+ * Prototypes for static functions in this file
+ */
+
+static void		DupLambdaInternalRep(Tcl_Obj *objPtr,
+			    Tcl_Obj *copyPtr);
+static void		FreeLambdaInternalRep(Tcl_Obj *objPtr);
+static int		InitArgsAndLocals(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj, int skip);
+static void		InitResolvedLocals(Tcl_Interp *interp,
+			    ByteCode *codePtr, Var *defPtr,
+			    Namespace *nsPtr);
+static void		InitLocalCache(Proc *procPtr);
+static void		ProcBodyDup(Tcl_Obj *srcPtr, Tcl_Obj *dupPtr);
+static void		ProcBodyFree(Tcl_Obj *objPtr);
+static int		ProcWrongNumArgs(Tcl_Interp *interp, int skip);
+static void		MakeProcError(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj);
+static void		MakeLambdaError(Tcl_Interp *interp,
+			    Tcl_Obj *procNameObj);
+static int		SetLambdaFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+
+static Tcl_NRPostProc ApplyNR2;
+static Tcl_NRPostProc InterpProcNR2;
+static Tcl_NRPostProc Uplevel_Callback;
+
+/*
+ * The ProcBodyObjType type
+ */
+
+const Tcl_ObjType tclProcBodyType = {
+    "procbody",			/* name for this type */
+    ProcBodyFree,		/* FreeInternalRep function */
+    ProcBodyDup,		/* DupInternalRep function */
+    NULL,			/* UpdateString function; Tcl_GetString and
+				 * Tcl_GetStringFromObj should panic
+				 * instead. */
+    NULL			/* SetFromAny function; Tcl_ConvertToType
+				 * should panic instead. */
+};
+
+/*
+ * The [upvar]/[uplevel] level reference type. Uses the longValue field
+ * to remember the integer value of a parsed #<integer> format.
+ *
+ * Uses the default behaviour throughout, and never disposes of the string
+ * rep; it's just a cache type.
+ */
+
+static const Tcl_ObjType levelReferenceType = {
+    "levelReference",
+    NULL, NULL, NULL, NULL
+};
+
+/*
+ * The type of lambdas. Note that every lambda will *always* have a string
+ * representation.
+ *
+ * Internally, ptr1 is a pointer to a Proc instance that is not bound to a
+ * command name, and ptr2 is a pointer to the namespace that the Proc instance
+ * will execute within. IF YOU CHANGE THIS, CHECK IN tclDisassemble.c TOO.
+ */
+
+const Tcl_ObjType tclLambdaType = {
+    "lambdaExpr",		/* name */
+    FreeLambdaInternalRep,	/* freeIntRepProc */
+    DupLambdaInternalRep,	/* dupIntRepProc */
+    NULL,			/* updateStringProc */
+    SetLambdaFromAny		/* setFromAnyProc */
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ProcObjCmd --
+ *
+ *	This object-based function is invoked to process the "proc" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	A new procedure gets created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ProcObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    Proc *procPtr;
+    const char *fullName;
+    const char *procName, *procArgs, *procBody;
+    Namespace *nsPtr, *altNsPtr, *cxtNsPtr;
+    Tcl_Command cmd;
+    Tcl_DString ds;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name args body");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Determine the namespace where the procedure should reside. Unless the
+     * command name includes namespace qualifiers, this will be the current
+     * namespace.
+     */
+
+    fullName = TclGetString(objv[1]);
+    TclGetNamespaceForQualName(interp, fullName, NULL, 0,
+	    &nsPtr, &altNsPtr, &cxtNsPtr, &procName);
+
+    if (nsPtr == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create procedure \"%s\": unknown namespace",
+		fullName));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", NULL);
+	return TCL_ERROR;
+    }
+    if (procName == NULL) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create procedure \"%s\": bad procedure name",
+		fullName));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", NULL);
+	return TCL_ERROR;
+    }
+    if ((nsPtr != iPtr->globalNsPtr)
+	    && (procName != NULL) && (procName[0] == ':')) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't create procedure \"%s\" in non-global namespace with"
+		" name starting with \":\"", procName));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMMAND", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create the data structure to represent the procedure.
+     */
+
+    if (TclCreateProc(interp, nsPtr, procName, objv[2], objv[3],
+	    &procPtr) != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (creating proc \"");
+	Tcl_AddErrorInfo(interp, procName);
+	Tcl_AddErrorInfo(interp, "\")");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Now create a command for the procedure. This will initially be in the
+     * current namespace unless the procedure's name included namespace
+     * qualifiers. To create the new command in the right namespace, we
+     * generate a fully qualified name for it.
+     */
+
+    Tcl_DStringInit(&ds);
+    if (nsPtr != iPtr->globalNsPtr) {
+	Tcl_DStringAppend(&ds, nsPtr->fullName, -1);
+	TclDStringAppendLiteral(&ds, "::");
+    }
+    Tcl_DStringAppend(&ds, procName, -1);
+
+    cmd = Tcl_NRCreateCommand(interp, Tcl_DStringValue(&ds), TclObjInterpProc,
+	    TclNRInterpProc, procPtr, TclProcDeleteProc);
+    Tcl_DStringFree(&ds);
+
+    /*
+     * Now initialize the new procedure's cmdPtr field. This will be used
+     * later when the procedure is called to determine what namespace the
+     * procedure will run in. This will be different than the current
+     * namespace if the proc was renamed into a different namespace.
+     */
+
+    procPtr->cmdPtr = (Command *) cmd;
+
+    /*
+     * TIP #280: Remember the line the procedure body is starting on. In a
+     * bytecode context we ask the engine to provide us with the necessary
+     * information. This is for the initialization of the byte code compiler
+     * when the body is used for the first time.
+     *
+     * This code is nearly identical to the #280 code in SetLambdaFromAny, see
+     * this file. The differences are the different index of the body in the
+     * line array of the context, and the lambda code requires some special
+     * processing. Find a way to factor the common elements into a single
+     * function.
+     */
+
+    if (iPtr->cmdFramePtr) {
+	CmdFrame *contextPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+
+	*contextPtr = *iPtr->cmdFramePtr;
+	if (contextPtr->type == TCL_LOCATION_BC) {
+	    /*
+	     * Retrieve source information from the bytecode, if possible. If
+	     * the information is retrieved successfully, context.type will be
+	     * TCL_LOCATION_SOURCE and the reference held by
+	     * context.data.eval.path will be counted.
+	     */
+
+	    TclGetSrcInfoForPc(contextPtr);
+	} else if (contextPtr->type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * The copy into 'context' up above has created another reference
+	     * to 'context.data.eval.path'; account for it.
+	     */
+
+	    Tcl_IncrRefCount(contextPtr->data.eval.path);
+	}
+
+	if (contextPtr->type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * We can account for source location within a proc only if the
+	     * proc body was not created by substitution.
+	     */
+
+	    if (contextPtr->line
+		    && (contextPtr->nline >= 4) && (contextPtr->line[3] >= 0)) {
+		int isNew;
+		Tcl_HashEntry *hePtr;
+		CmdFrame *cfPtr = ckalloc(sizeof(CmdFrame));
+
+		cfPtr->level = -1;
+		cfPtr->type = contextPtr->type;
+		cfPtr->line = ckalloc(sizeof(int));
+		cfPtr->line[0] = contextPtr->line[3];
+		cfPtr->nline = 1;
+		cfPtr->framePtr = NULL;
+		cfPtr->nextPtr = NULL;
+
+		cfPtr->data.eval.path = contextPtr->data.eval.path;
+		Tcl_IncrRefCount(cfPtr->data.eval.path);
+
+		cfPtr->cmd = NULL;
+		cfPtr->len = 0;
+
+		hePtr = Tcl_CreateHashEntry(iPtr->linePBodyPtr,
+			procPtr, &isNew);
+		if (!isNew) {
+		    /*
+		     * Get the old command frame and release it. See also
+		     * TclProcCleanupProc in this file. Currently it seems as
+		     * if only the procbodytest::proc command of the testsuite
+		     * is able to trigger this situation.
+		     */
+
+		    CmdFrame *cfOldPtr = Tcl_GetHashValue(hePtr);
+
+		    if (cfOldPtr->type == TCL_LOCATION_SOURCE) {
+			Tcl_DecrRefCount(cfOldPtr->data.eval.path);
+			cfOldPtr->data.eval.path = NULL;
+		    }
+		    ckfree(cfOldPtr->line);
+		    cfOldPtr->line = NULL;
+		    ckfree(cfOldPtr);
+		}
+		Tcl_SetHashValue(hePtr, cfPtr);
+	    }
+
+	    /*
+	     * 'contextPtr' is going out of scope; account for the reference
+	     * that it's holding to the path name.
+	     */
+
+	    Tcl_DecrRefCount(contextPtr->data.eval.path);
+	    contextPtr->data.eval.path = NULL;
+	}
+	TclStackFree(interp, contextPtr);
+    }
+
+    /*
+     * Optimize for no-op procs: if the body is not precompiled (like a TclPro
+     * procbody), and the argument list is just "args" and the body is empty,
+     * define a compileProc to compile a no-op.
+     *
+     * Notes:
+     *	 - cannot be done for any argument list without having different
+     *	   compiled/not-compiled behaviour in the "wrong argument #" case, or
+     *	   making this code much more complicated. In any case, it doesn't
+     *	   seem to make a lot of sense to verify the number of arguments we
+     *	   are about to ignore ...
+     *	 - could be enhanced to handle also non-empty bodies that contain only
+     *	   comments; however, parsing the body will slow down the compilation
+     *	   of all procs whose argument list is just _args_
+     */
+
+    if (objv[3]->typePtr == &tclProcBodyType) {
+	goto done;
+    }
+
+    procArgs = TclGetString(objv[2]);
+
+    while (*procArgs == ' ') {
+	procArgs++;
+    }
+
+    if ((procArgs[0] == 'a') && (strncmp(procArgs, "args", 4) == 0)) {
+	int numBytes;
+
+	procArgs +=4;
+	while (*procArgs != '\0') {
+	    if (*procArgs != ' ') {
+		goto done;
+	    }
+	    procArgs++;
+	}
+
+	/*
+	 * The argument list is just "args"; check the body
+	 */
+
+	procBody = TclGetStringFromObj(objv[3], &numBytes);
+	if (TclParseAllWhiteSpace(procBody, numBytes) < numBytes) {
+	    goto done;
+	}
+
+	/*
+	 * The body is just spaces: link the compileProc
+	 */
+
+	((Command *) cmd)->compileProc = TclCompileNoOp;
+    }
+
+  done:
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCreateProc --
+ *
+ *	Creates the data associated with a Tcl procedure definition. This
+ *	function knows how to handle two types of body objects: strings and
+ *	procbody. Strings are the traditional (and common) value for bodies,
+ *	procbody are values created by extensions that have loaded a
+ *	previously compiled script.
+ *
+ * Results:
+ *	Returns TCL_OK on success, along with a pointer to a Tcl procedure
+ *	definition in procPtrPtr where the cmdPtr field is not initialised.
+ *	This definition should be freed by calling TclProcCleanupProc() when
+ *	it is no longer needed. Returns TCL_ERROR if anything goes wrong.
+ *
+ * Side effects:
+ *	If anything goes wrong, this function returns an error message in the
+ *	interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCreateProc(
+    Tcl_Interp *interp,		/* Interpreter containing proc. */
+    Namespace *nsPtr,		/* Namespace containing this proc. */
+    const char *procName,	/* Unqualified name of this proc. */
+    Tcl_Obj *argsPtr,		/* Description of arguments. */
+    Tcl_Obj *bodyPtr,		/* Command body. */
+    Proc **procPtrPtr)		/* Returns: pointer to proc data. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char **argArray = NULL;
+
+    register Proc *procPtr;
+    int i, length, result, numArgs;
+    const char *args, *bytes, *p;
+    register CompiledLocal *localPtr = NULL;
+    Tcl_Obj *defPtr;
+    int precompiled = 0;
+
+    if (bodyPtr->typePtr == &tclProcBodyType) {
+	/*
+	 * Because the body is a TclProProcBody, the actual body is already
+	 * compiled, and it is not shared with anyone else, so it's OK not to
+	 * unshare it (as a matter of fact, it is bad to unshare it, because
+	 * there may be no source code).
+	 *
+	 * We don't create and initialize a Proc structure for the procedure;
+	 * rather, we use what is in the body object. We increment the ref
+	 * count of the Proc struct since the command (soon to be created)
+	 * will be holding a reference to it.
+	 */
+
+	procPtr = bodyPtr->internalRep.twoPtrValue.ptr1;
+	procPtr->iPtr = iPtr;
+	procPtr->refCount++;
+	precompiled = 1;
+    } else {
+	/*
+	 * If the procedure's body object is shared because its string value
+	 * is identical to, e.g., the body of another procedure, we must
+	 * create a private copy for this procedure to use. Such sharing of
+	 * procedure bodies is rare but can cause problems. A procedure body
+	 * is compiled in a context that includes the number of "slots"
+	 * allocated by the compiler for local variables. There is a local
+	 * variable slot for each formal parameter (the
+	 * "procPtr->numCompiledLocals = numArgs" assignment below). This
+	 * means that the same code can not be shared by two procedures that
+	 * have a different number of arguments, even if their bodies are
+	 * identical. Note that we don't use Tcl_DuplicateObj since we would
+	 * not want any bytecode internal representation.
+	 */
+
+	if (Tcl_IsShared(bodyPtr)) {
+	    Tcl_Obj *sharedBodyPtr = bodyPtr;
+
+	    bytes = TclGetStringFromObj(bodyPtr, &length);
+	    bodyPtr = Tcl_NewStringObj(bytes, length);
+
+	    /*
+	     * TIP #280.
+	     * Ensure that the continuation line data for the original body is
+	     * not lost and applies to the new body as well.
+	     */
+
+	    TclContinuationsCopy(bodyPtr, sharedBodyPtr);
+	}
+
+	/*
+	 * Create and initialize a Proc structure for the procedure. We
+	 * increment the ref count of the procedure's body object since there
+	 * will be a reference to it in the Proc structure.
+	 */
+
+	Tcl_IncrRefCount(bodyPtr);
+
+	procPtr = ckalloc(sizeof(Proc));
+	procPtr->iPtr = iPtr;
+	procPtr->refCount = 1;
+	procPtr->bodyPtr = bodyPtr;
+	procPtr->numArgs = 0;	/* Actual argument count is set below. */
+	procPtr->numCompiledLocals = 0;
+	procPtr->firstLocalPtr = NULL;
+	procPtr->lastLocalPtr = NULL;
+    }
+
+    /*
+     * Break up the argument list into argument specifiers, then process each
+     * argument specifier. If the body is precompiled, processing is limited
+     * to checking that the parsed argument is consistent with the one stored
+     * in the Proc.
+     *
+     * THIS FAILS IF THE ARG LIST OBJECT'S STRING REP CONTAINS NULS.
+     */
+
+    args = TclGetStringFromObj(argsPtr, &length);
+    result = Tcl_SplitList(interp, args, &numArgs, &argArray);
+    if (result != TCL_OK) {
+	goto procError;
+    }
+
+    if (precompiled) {
+	if (numArgs > procPtr->numArgs) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "procedure \"%s\": arg list contains %d entries, "
+		    "precompiled header expects %d", procName, numArgs,
+		    procPtr->numArgs));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+		    "BYTECODELIES", NULL);
+	    goto procError;
+	}
+	localPtr = procPtr->firstLocalPtr;
+    } else {
+	procPtr->numArgs = numArgs;
+	procPtr->numCompiledLocals = numArgs;
+    }
+
+    for (i = 0; i < numArgs; i++) {
+	int fieldCount, nameLength;
+	size_t valueLength;
+	const char **fieldValues;
+
+	/*
+	 * Now divide the specifier up into name and default.
+	 */
+
+	result = Tcl_SplitList(interp, argArray[i], &fieldCount,
+		&fieldValues);
+	if (result != TCL_OK) {
+	    goto procError;
+	}
+	if (fieldCount > 2) {
+	    ckfree(fieldValues);
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "too many fields in argument specifier \"%s\"",
+		    argArray[i]));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+		    "FORMALARGUMENTFORMAT", NULL);
+	    goto procError;
+	}
+	if ((fieldCount == 0) || (*fieldValues[0] == 0)) {
+	    ckfree(fieldValues);
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "argument with no name", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+		    "FORMALARGUMENTFORMAT", NULL);
+	    goto procError;
+	}
+
+	nameLength = strlen(fieldValues[0]);
+	if (fieldCount == 2) {
+	    valueLength = strlen(fieldValues[1]);
+	} else {
+	    valueLength = 0;
+	}
+
+	/*
+	 * Check that the formal parameter name is a scalar.
+	 */
+
+	p = fieldValues[0];
+	while (*p != '\0') {
+	    if (*p == '(') {
+		const char *q = p;
+		do {
+		    q++;
+		} while (*q != '\0');
+		q--;
+		if (*q == ')') {	/* We have an array element. */
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "formal parameter \"%s\" is an array element",
+			    fieldValues[0]));
+		    ckfree(fieldValues);
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+			    "FORMALARGUMENTFORMAT", NULL);
+		    goto procError;
+		}
+	    } else if ((*p == ':') && (*(p+1) == ':')) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"formal parameter \"%s\" is not a simple name",
+			fieldValues[0]));
+		ckfree(fieldValues);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+			"FORMALARGUMENTFORMAT", NULL);
+		goto procError;
+	    }
+	    p++;
+	}
+
+	if (precompiled) {
+	    /*
+	     * Compare the parsed argument with the stored one. Note that the
+	     * only flag value that makes sense at this point is VAR_ARGUMENT
+	     * (its value was kept the same as pre VarReform to simplify
+	     * tbcload's processing of older byetcodes).
+	     *
+	     * The only other flag vlaue that is important to retrieve from
+	     * precompiled procs is VAR_TEMPORARY (also unchanged). It is
+	     * needed later when retrieving the variable names.
+	     */
+
+	    if ((localPtr->nameLength != nameLength)
+		    || (strcmp(localPtr->name, fieldValues[0]))
+		    || (localPtr->frameIndex != i)
+		    || !(localPtr->flags & VAR_ARGUMENT)
+		    || (localPtr->defValuePtr == NULL && fieldCount == 2)
+		    || (localPtr->defValuePtr != NULL && fieldCount != 2)) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"procedure \"%s\": formal parameter %d is "
+			"inconsistent with precompiled body", procName, i));
+		ckfree(fieldValues);
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+			"BYTECODELIES", NULL);
+		goto procError;
+	    }
+
+	    /*
+	     * Compare the default value if any.
+	     */
+
+	    if (localPtr->defValuePtr != NULL) {
+		const char *tmpPtr = TclGetString(localPtr->defValuePtr);
+		size_t tmpLength = localPtr->defValuePtr->length;
+
+		if ((valueLength != tmpLength) ||
+			strncmp(fieldValues[1], tmpPtr, tmpLength)) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "procedure \"%s\": formal parameter \"%s\" has "
+			    "default value inconsistent with precompiled body",
+			    procName, fieldValues[0]));
+		    ckfree(fieldValues);
+		    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+			    "BYTECODELIES", NULL);
+		    goto procError;
+		}
+	    }
+	    if ((i == numArgs - 1)
+		    && (localPtr->nameLength == 4)
+		    && (localPtr->name[0] == 'a')
+		    && (strcmp(localPtr->name, "args") == 0)) {
+		localPtr->flags |= VAR_IS_ARGS;
+	    }
+
+	    localPtr = localPtr->nextPtr;
+	} else {
+	    /*
+	     * Allocate an entry in the runtime procedure frame's array of
+	     * local variables for the argument.
+	     */
+
+	    localPtr = ckalloc(TclOffset(CompiledLocal, name) + nameLength+1);
+	    if (procPtr->firstLocalPtr == NULL) {
+		procPtr->firstLocalPtr = procPtr->lastLocalPtr = localPtr;
+	    } else {
+		procPtr->lastLocalPtr->nextPtr = localPtr;
+		procPtr->lastLocalPtr = localPtr;
+	    }
+	    localPtr->nextPtr = NULL;
+	    localPtr->nameLength = nameLength;
+	    localPtr->frameIndex = i;
+	    localPtr->flags = VAR_ARGUMENT;
+	    localPtr->resolveInfo = NULL;
+
+	    if (fieldCount == 2) {
+		localPtr->defValuePtr =
+			Tcl_NewStringObj(fieldValues[1], valueLength);
+		Tcl_IncrRefCount(localPtr->defValuePtr);
+	    } else {
+		localPtr->defValuePtr = NULL;
+	    }
+	    memcpy(localPtr->name, fieldValues[0], nameLength + 1);
+	    if ((i == numArgs - 1)
+		    && (localPtr->nameLength == 4)
+		    && (localPtr->name[0] == 'a')
+		    && (strcmp(localPtr->name, "args") == 0)) {
+		localPtr->flags |= VAR_IS_ARGS;
+	    }
+	}
+
+	ckfree(fieldValues);
+    }
+
+    *procPtrPtr = procPtr;
+    ckfree(argArray);
+    return TCL_OK;
+
+  procError:
+    if (precompiled) {
+	procPtr->refCount--;
+    } else {
+	Tcl_DecrRefCount(bodyPtr);
+	while (procPtr->firstLocalPtr != NULL) {
+	    localPtr = procPtr->firstLocalPtr;
+	    procPtr->firstLocalPtr = localPtr->nextPtr;
+
+	    defPtr = localPtr->defValuePtr;
+	    if (defPtr != NULL) {
+		Tcl_DecrRefCount(defPtr);
+	    }
+
+	    ckfree(localPtr);
+	}
+	ckfree(procPtr);
+    }
+    if (argArray != NULL) {
+	ckfree(argArray);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetFrame --
+ *
+ *	Given a description of a procedure frame, such as the first argument
+ *	to an "uplevel" or "upvar" command, locate the call frame for the
+ *	appropriate level of procedure.
+ *
+ * Results:
+ *	The return value is -1 if an error occurred in finding the frame (in
+ *	this case an error message is left in the interp's result). 1 is
+ *	returned if string was either a number or a number preceded by "#" and
+ *	it specified a valid frame. 0 is returned if string isn't one of the
+ *	two things above (in this case, the lookup acts as if string were
+ *	"1"). The variable pointed to by framePtrPtr is filled in with the
+ *	address of the desired frame (unless an error occurs, in which case it
+ *	isn't modified).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetFrame(
+    Tcl_Interp *interp,		/* Interpreter in which to find frame. */
+    const char *name,		/* String describing frame. */
+    CallFrame **framePtrPtr)	/* Store pointer to frame here (or NULL if
+				 * global frame indicated). */
+{
+    register Interp *iPtr = (Interp *) interp;
+    int curLevel, level, result;
+    CallFrame *framePtr;
+
+    /*
+     * Parse string to figure out which level number to go to.
+     */
+
+    result = 1;
+    curLevel = iPtr->varFramePtr->level;
+    if (*name== '#') {
+	if (Tcl_GetInt(interp, name+1, &level) != TCL_OK || level < 0) {
+	    goto levelError;
+	}
+    } else if (isdigit(UCHAR(*name))) { /* INTL: digit */
+	if (Tcl_GetInt(interp, name, &level) != TCL_OK) {
+	    goto levelError;
+	}
+	level = curLevel - level;
+    } else {
+	level = curLevel - 1;
+	result = 0;
+    }
+
+    /*
+     * Figure out which frame to use, and return it to the caller.
+     */
+
+    for (framePtr = iPtr->varFramePtr; framePtr != NULL;
+	    framePtr = framePtr->callerVarPtr) {
+	if (framePtr->level == level) {
+	    break;
+	}
+    }
+    if (framePtr == NULL) {
+	goto levelError;
+    }
+
+    *framePtrPtr = framePtr;
+    return result;
+
+  levelError:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf("bad level \"%s\"", name));
+    Tcl_SetErrorCode(interp, "TCL", "VALUE", "STACKLEVEL", NULL);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjGetFrame --
+ *
+ *	Given a description of a procedure frame, such as the first argument
+ *	to an "uplevel" or "upvar" command, locate the call frame for the
+ *	appropriate level of procedure.
+ *
+ * Results:
+ *	The return value is -1 if an error occurred in finding the frame (in
+ *	this case an error message is left in the interp's result). 1 is
+ *	returned if objPtr was either an int or an int preceded by "#" and
+ *	it specified a valid frame. 0 is returned if objPtr isn't one of the
+ *	two things above (in this case, the lookup acts as if objPtr were
+ *	"1"). The variable pointed to by framePtrPtr is filled in with the
+ *	address of the desired frame (unless an error occurs, in which case it
+ *	isn't modified).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjGetFrame(
+    Tcl_Interp *interp,		/* Interpreter in which to find frame. */
+    Tcl_Obj *objPtr,		/* Object describing frame. */
+    CallFrame **framePtrPtr)	/* Store pointer to frame here (or NULL if
+				 * global frame indicated). */
+{
+    register Interp *iPtr = (Interp *) interp;
+    int curLevel, level, result;
+    const char *name = NULL;
+
+    /*
+     * Parse object to figure out which level number to go to.
+     */
+
+    result = 0;
+    curLevel = iPtr->varFramePtr->level;
+
+    /*
+     * Check for integer first, since that has potential to spare us
+     * a generation of a stringrep.
+     */
+
+    if (objPtr == NULL) {
+	/* Do nothing */
+    } else if (TCL_OK == Tcl_GetIntFromObj(NULL, objPtr, &level)
+	    && (level >= 0)) {
+	level = curLevel - level;
+	result = 1;
+    } else if (objPtr->typePtr == &levelReferenceType) {
+	level = (int) objPtr->internalRep.longValue;
+	result = 1;
+    } else {
+	name = TclGetString(objPtr);
+	if (name[0] == '#') {
+	    if (TCL_OK == Tcl_GetInt(NULL, name+1, &level) && level >= 0) {
+		TclFreeIntRep(objPtr);
+		objPtr->typePtr = &levelReferenceType;
+		objPtr->internalRep.longValue = level;
+		result = 1;
+	    } else {
+		result = -1;
+	    }
+	} else if (isdigit(UCHAR(name[0]))) { /* INTL: digit */
+	    /*
+	     * If this were an integer, we'd have succeeded already.
+	     * Docs say we have to treat this as a 'bad level'  error.
+	     */
+	    result = -1;
+	}
+    }
+
+    if (result == 0) {
+	level = curLevel - 1;
+	name = "1";
+    }
+    if (result != -1) {
+	if (level >= 0) {
+	    CallFrame *framePtr;
+	    for (framePtr = iPtr->varFramePtr; framePtr != NULL;
+		    framePtr = framePtr->callerVarPtr) {
+		if (framePtr->level == level) {
+		    *framePtrPtr = framePtr;
+		    return result;
+		}
+	    }
+	}
+	if (name == NULL) {
+	    name = TclGetString(objPtr);
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf("bad level \"%s\"", name));
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "LEVEL", name, NULL);
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UplevelObjCmd --
+ *
+ *	This object function is invoked to process the "uplevel" Tcl command.
+ *	See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+Uplevel_Callback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    CallFrame *savedVarFramePtr = data[0];
+
+    if (result == TCL_ERROR) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (\"uplevel\" body line %d)", Tcl_GetErrorLine(interp)));
+    }
+
+    /*
+     * Restore the variable frame, and return.
+     */
+
+    ((Interp *)interp)->varFramePtr = savedVarFramePtr;
+    return result;
+}
+
+	/* ARGSUSED */
+int
+Tcl_UplevelObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRUplevelObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRUplevelObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+
+    register Interp *iPtr = (Interp *) interp;
+    CmdFrame *invoker = NULL;
+    int word = 0;
+    int result;
+    CallFrame *savedVarFramePtr, *framePtr;
+    Tcl_Obj *objPtr;
+
+    if (objc < 2) {
+    uplevelSyntax:
+	Tcl_WrongNumArgs(interp, 1, objv, "?level? command ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the level to use for executing the command.
+     */
+
+    result = TclObjGetFrame(interp, objv[1], &framePtr);
+    if (result == -1) {
+	return TCL_ERROR;
+    }
+    objc -= result + 1;
+    if (objc == 0) {
+	goto uplevelSyntax;
+    }
+    objv += result + 1;
+
+    /*
+     * Modify the interpreter state to execute in the given frame.
+     */
+
+    savedVarFramePtr = iPtr->varFramePtr;
+    iPtr->varFramePtr = framePtr;
+
+    /*
+     * Execute the residual arguments as a command.
+     */
+
+    if (objc == 1) {
+	/*
+	 * TIP #280. Make actual argument location available to eval'd script
+	 */
+
+	TclArgumentGet(interp, objv[0], &invoker, &word);
+	objPtr = objv[0];
+
+    } else {
+	/*
+	 * More than one argument: concatenate them together with spaces
+	 * between, then evaluate the result. Tcl_EvalObjEx will delete the
+	 * object when it decrements its refcount after eval'ing it.
+	 */
+
+	objPtr = Tcl_ConcatObj(objc, objv);
+    }
+
+    TclNRAddCallback(interp, Uplevel_Callback, savedVarFramePtr, NULL, NULL,
+	    NULL);
+    return TclNREvalObjEx(interp, objPtr, 0, invoker, word);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFindProc --
+ *
+ *	Given the name of a procedure, return a pointer to the record
+ *	describing the procedure. The procedure will be looked up using the
+ *	usual rules: first in the current namespace and then in the global
+ *	namespace.
+ *
+ * Results:
+ *	NULL is returned if the name doesn't correspond to any procedure.
+ *	Otherwise, the return value is a pointer to the procedure's record. If
+ *	the name is found but refers to an imported command that points to a
+ *	"real" procedure defined in another namespace, a pointer to that
+ *	"real" procedure's structure is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Proc *
+TclFindProc(
+    Interp *iPtr,		/* Interpreter in which to look. */
+    const char *procName)	/* Name of desired procedure. */
+{
+    Tcl_Command cmd;
+    Command *cmdPtr;
+
+    cmd = Tcl_FindCommand((Tcl_Interp *) iPtr, procName, NULL, /*flags*/ 0);
+    if (cmd == (Tcl_Command) NULL) {
+	return NULL;
+    }
+    cmdPtr = (Command *) cmd;
+
+    return TclIsProc(cmdPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclIsProc --
+ *
+ *	Tells whether a command is a Tcl procedure or not.
+ *
+ * Results:
+ *	If the given command is actually a Tcl procedure, the return value is
+ *	the address of the record describing the procedure. Otherwise the
+ *	return value is 0.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Proc *
+TclIsProc(
+    Command *cmdPtr)		/* Command to test. */
+{
+    Tcl_Command origCmd = TclGetOriginalCommand((Tcl_Command) cmdPtr);
+
+    if (origCmd != NULL) {
+	cmdPtr = (Command *) origCmd;
+    }
+    if (cmdPtr->deleteProc == TclProcDeleteProc) {
+	return cmdPtr->objClientData;
+    }
+    return NULL;
+}
+
+static int
+ProcWrongNumArgs(
+    Tcl_Interp *interp,
+    int skip)
+{
+    CallFrame *framePtr = ((Interp *)interp)->varFramePtr;
+    register Proc *procPtr = framePtr->procPtr;
+    register Var *defPtr;
+    int localCt = procPtr->numCompiledLocals, numArgs, i;
+    Tcl_Obj **desiredObjs;
+    const char *final = NULL;
+
+    /*
+     * Build up desired argument list for Tcl_WrongNumArgs
+     */
+
+    numArgs = framePtr->procPtr->numArgs;
+    desiredObjs = TclStackAlloc(interp,
+	    (int) sizeof(Tcl_Obj *) * (numArgs+1));
+
+    if (framePtr->isProcCallFrame & FRAME_IS_LAMBDA) {
+	desiredObjs[0] = Tcl_NewStringObj("lambdaExpr", -1);
+    } else {
+#ifdef AVOID_HACKS_FOR_ITCL
+	desiredObjs[0] = framePtr->objv[skip-1];
+#else
+	desiredObjs[0] = Tcl_NewListObj(1, framePtr->objv + skip - 1);
+#endif /* AVOID_HACKS_FOR_ITCL */
+    }
+    Tcl_IncrRefCount(desiredObjs[0]);
+
+    defPtr = (Var *) (&framePtr->localCachePtr->varName0 + localCt);
+    for (i=1 ; i<=numArgs ; i++, defPtr++) {
+	Tcl_Obj *argObj;
+	Tcl_Obj *namePtr = localName(framePtr, i-1);
+
+	if (defPtr->value.objPtr != NULL) {
+	    TclNewObj(argObj);
+	    Tcl_AppendStringsToObj(argObj, "?", TclGetString(namePtr), "?", NULL);
+	} else if (defPtr->flags & VAR_IS_ARGS) {
+	    numArgs--;
+	    final = "?arg ...?";
+	    break;
+	} else {
+	    argObj = namePtr;
+	    Tcl_IncrRefCount(namePtr);
+	}
+	desiredObjs[i] = argObj;
+    }
+
+    Tcl_ResetResult(interp);
+    Tcl_WrongNumArgs(interp, numArgs+1, desiredObjs, final);
+
+    for (i=0 ; i<=numArgs ; i++) {
+	Tcl_DecrRefCount(desiredObjs[i]);
+    }
+    TclStackFree(interp, desiredObjs);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitCompiledLocals --
+ *
+ *	This routine is invoked in order to initialize the compiled locals
+ *	table for a new call frame.
+ *
+ *	DEPRECATED: functionality has been inlined elsewhere; this function
+ *	remains to insure binary compatibility with Itcl.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May invoke various name resolvers in order to determine which
+ *	variables are being referenced at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitCompiledLocals(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    CallFrame *framePtr,	/* Call frame to initialize. */
+    Namespace *nsPtr)		/* Pointer to current namespace. */
+{
+    Var *varPtr = framePtr->compiledLocals;
+    Tcl_Obj *bodyPtr;
+    ByteCode *codePtr;
+
+    bodyPtr = framePtr->procPtr->bodyPtr;
+    if (bodyPtr->typePtr != &tclByteCodeType) {
+	Tcl_Panic("body object for proc attached to frame is not a byte code type");
+    }
+    codePtr = bodyPtr->internalRep.twoPtrValue.ptr1;
+
+    if (framePtr->numCompiledLocals) {
+	if (!codePtr->localCachePtr) {
+	    InitLocalCache(framePtr->procPtr) ;
+	}
+	framePtr->localCachePtr = codePtr->localCachePtr;
+	framePtr->localCachePtr->refCount++;
+    }
+
+    InitResolvedLocals(interp, codePtr, varPtr, nsPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InitResolvedLocals --
+ *
+ *	This routine is invoked in order to initialize the compiled locals
+ *	table for a new call frame.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May invoke various name resolvers in order to determine which
+ *	variables are being referenced at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+InitResolvedLocals(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    ByteCode *codePtr,
+    Var *varPtr,
+    Namespace *nsPtr)		/* Pointer to current namespace. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int haveResolvers = (nsPtr->compiledVarResProc || iPtr->resolverPtr);
+    CompiledLocal *firstLocalPtr, *localPtr;
+    int varNum;
+    Tcl_ResolvedVarInfo *resVarInfo;
+
+    /*
+     * Find the localPtr corresponding to varPtr
+     */
+
+    varNum = varPtr - iPtr->framePtr->compiledLocals;
+    localPtr = iPtr->framePtr->procPtr->firstLocalPtr;
+    while (varNum--) {
+	localPtr = localPtr->nextPtr;
+    }
+
+    if (!(haveResolvers && (codePtr->flags & TCL_BYTECODE_RESOLVE_VARS))) {
+	goto doInitResolvedLocals;
+    }
+
+    /*
+     * This is the first run after a recompile, or else the resolver epoch
+     * has changed: update the resolver cache.
+     */
+
+    firstLocalPtr = localPtr;
+    for (; localPtr != NULL; localPtr = localPtr->nextPtr) {
+	if (localPtr->resolveInfo) {
+	    if (localPtr->resolveInfo->deleteProc) {
+		localPtr->resolveInfo->deleteProc(localPtr->resolveInfo);
+	    } else {
+		ckfree(localPtr->resolveInfo);
+	    }
+	    localPtr->resolveInfo = NULL;
+	}
+	localPtr->flags &= ~VAR_RESOLVED;
+
+	if (haveResolvers &&
+		!(localPtr->flags & (VAR_ARGUMENT|VAR_TEMPORARY))) {
+	    ResolverScheme *resPtr = iPtr->resolverPtr;
+	    Tcl_ResolvedVarInfo *vinfo;
+	    int result;
+
+	    if (nsPtr->compiledVarResProc) {
+		result = nsPtr->compiledVarResProc(nsPtr->interp,
+			localPtr->name, localPtr->nameLength,
+			(Tcl_Namespace *) nsPtr, &vinfo);
+	    } else {
+		result = TCL_CONTINUE;
+	    }
+
+	    while ((result == TCL_CONTINUE) && resPtr) {
+		if (resPtr->compiledVarResProc) {
+		    result = resPtr->compiledVarResProc(nsPtr->interp,
+			    localPtr->name, localPtr->nameLength,
+			    (Tcl_Namespace *) nsPtr, &vinfo);
+		}
+		resPtr = resPtr->nextPtr;
+	    }
+	    if (result == TCL_OK) {
+		localPtr->resolveInfo = vinfo;
+		localPtr->flags |= VAR_RESOLVED;
+	    }
+	}
+    }
+    localPtr = firstLocalPtr;
+    codePtr->flags &= ~TCL_BYTECODE_RESOLVE_VARS;
+
+    /*
+     * Initialize the array of local variables stored in the call frame. Some
+     * variables may have special resolution rules. In that case, we call
+     * their "resolver" procs to get our hands on the variable, and we make
+     * the compiled local a link to the real variable.
+     */
+
+  doInitResolvedLocals:
+    for (; localPtr != NULL; varPtr++, localPtr = localPtr->nextPtr) {
+	varPtr->flags = 0;
+	varPtr->value.objPtr = NULL;
+
+	/*
+	 * Now invoke the resolvers to determine the exact variables that
+	 * should be used.
+	 */
+
+	resVarInfo = localPtr->resolveInfo;
+	if (resVarInfo && resVarInfo->fetchProc) {
+	    register Var *resolvedVarPtr = (Var *)
+		    resVarInfo->fetchProc(interp, resVarInfo);
+
+	    if (resolvedVarPtr) {
+		if (TclIsVarInHash(resolvedVarPtr)) {
+		    VarHashRefCount(resolvedVarPtr)++;
+		}
+		varPtr->flags = VAR_LINK;
+		varPtr->value.linkPtr = resolvedVarPtr;
+	    }
+	}
+    }
+}
+
+void
+TclFreeLocalCache(
+    Tcl_Interp *interp,
+    LocalCache *localCachePtr)
+{
+    int i;
+    Tcl_Obj **namePtrPtr = &localCachePtr->varName0;
+
+    for (i = 0; i < localCachePtr->numVars; i++, namePtrPtr++) {
+	register Tcl_Obj *objPtr = *namePtrPtr;
+
+	if (objPtr) {
+	    /* TclReleaseLiteral calls Tcl_DecrRefCount for us */
+	    TclReleaseLiteral(interp, objPtr);
+	}
+    }
+    ckfree(localCachePtr);
+}
+
+static void
+InitLocalCache(
+    Proc *procPtr)
+{
+    Interp *iPtr = procPtr->iPtr;
+    ByteCode *codePtr = procPtr->bodyPtr->internalRep.twoPtrValue.ptr1;
+    int localCt = procPtr->numCompiledLocals;
+    int numArgs = procPtr->numArgs, i = 0;
+
+    Tcl_Obj **namePtr;
+    Var *varPtr;
+    LocalCache *localCachePtr;
+    CompiledLocal *localPtr;
+    int new;
+
+    /*
+     * Cache the names and initial values of local variables; store the
+     * cache in both the framePtr for this execution and in the codePtr
+     * for future calls.
+     */
+
+    localCachePtr = ckalloc(sizeof(LocalCache)
+	    + (localCt - 1) * sizeof(Tcl_Obj *)
+	    + numArgs * sizeof(Var));
+
+    namePtr = &localCachePtr->varName0;
+    varPtr = (Var *) (namePtr + localCt);
+    localPtr = procPtr->firstLocalPtr;
+    while (localPtr) {
+	if (TclIsVarTemporary(localPtr)) {
+	    *namePtr = NULL;
+	} else {
+	    *namePtr = TclCreateLiteral(iPtr, localPtr->name,
+		    localPtr->nameLength, /* hash */ (unsigned int) -1,
+		    &new, /* nsPtr */ NULL, 0, NULL);
+	    Tcl_IncrRefCount(*namePtr);
+	}
+
+	if (i < numArgs) {
+	    varPtr->flags = (localPtr->flags & VAR_IS_ARGS);
+	    varPtr->value.objPtr = localPtr->defValuePtr;
+	    varPtr++;
+	    i++;
+	}
+	namePtr++;
+	localPtr = localPtr->nextPtr;
+    }
+    codePtr->localCachePtr = localCachePtr;
+    localCachePtr->refCount = 1;
+    localCachePtr->numVars = localCt;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InitArgsAndLocals --
+ *
+ *	This routine is invoked in order to initialize the arguments and other
+ *	compiled locals table for a new call frame.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Allocates memory on the stack for the compiled local variables, the
+ *	caller is responsible for freeing them. Initialises all variables. May
+ *	invoke various name resolvers in order to determine which variables
+ *	are being referenced at runtime.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+InitArgsAndLocals(
+    register Tcl_Interp *interp,/* Interpreter in which procedure was
+				 * invoked. */
+    Tcl_Obj *procNameObj,	/* Procedure name for error reporting. */
+    int skip)			/* Number of initial arguments to be skipped,
+				 * i.e., words in the "command name". */
+{
+    CallFrame *framePtr = ((Interp *)interp)->varFramePtr;
+    register Proc *procPtr = framePtr->procPtr;
+    ByteCode *codePtr = procPtr->bodyPtr->internalRep.twoPtrValue.ptr1;
+    register Var *varPtr, *defPtr;
+    int localCt = procPtr->numCompiledLocals, numArgs, argCt, i, imax;
+    Tcl_Obj *const *argObjs;
+
+    /*
+     * Make sure that the local cache of variable names and initial values has
+     * been initialised properly .
+     */
+
+    if (localCt) {
+	if (!codePtr->localCachePtr) {
+	    InitLocalCache(procPtr) ;
+	}
+	framePtr->localCachePtr = codePtr->localCachePtr;
+	framePtr->localCachePtr->refCount++;
+	defPtr = (Var *) (&framePtr->localCachePtr->varName0 + localCt);
+    } else {
+	defPtr = NULL;
+    }
+
+    /*
+     * Create the "compiledLocals" array. Make sure it is large enough to hold
+     * all the procedure's compiled local variables, including its formal
+     * parameters.
+     */
+
+    varPtr = TclStackAlloc(interp, (int)(localCt * sizeof(Var)));
+    framePtr->compiledLocals = varPtr;
+    framePtr->numCompiledLocals = localCt;
+
+    /*
+     * Match and assign the call's actual parameters to the procedure's formal
+     * arguments. The formal arguments are described by the first numArgs
+     * entries in both the Proc structure's local variable list and the call
+     * frame's local variable array.
+     */
+
+    numArgs = procPtr->numArgs;
+    argCt = framePtr->objc - skip;	/* Set it to the number of args to the
+					 * procedure. */
+    argObjs = framePtr->objv + skip;
+    if (numArgs == 0) {
+	if (argCt) {
+	    goto incorrectArgs;
+	} else {
+	    goto correctArgs;
+	}
+    }
+    imax = ((argCt < numArgs-1) ? argCt : numArgs-1);
+    for (i = 0; i < imax; i++, varPtr++, defPtr ? defPtr++ : defPtr) {
+	/*
+	 * "Normal" arguments; last formal is special, depends on it being
+	 * 'args'.
+	 */
+
+	Tcl_Obj *objPtr = argObjs[i];
+
+	varPtr->flags = 0;
+	varPtr->value.objPtr = objPtr;
+	Tcl_IncrRefCount(objPtr);	/* Local var is a reference. */
+    }
+    for (; i < numArgs-1; i++, varPtr++, defPtr ? defPtr++ : defPtr) {
+	/*
+	 * This loop is entered if argCt < (numArgs-1). Set default values;
+	 * last formal is special.
+	 */
+
+	Tcl_Obj *objPtr = defPtr ? defPtr->value.objPtr : NULL;
+
+	if (!objPtr) {
+	    goto incorrectArgs;
+	}
+	varPtr->flags = 0;
+	varPtr->value.objPtr = objPtr;
+	Tcl_IncrRefCount(objPtr);	/* Local var reference. */
+    }
+
+    /*
+     * When we get here, the last formal argument remains to be defined:
+     * defPtr and varPtr point to the last argument to be initialized.
+     */
+
+    varPtr->flags = 0;
+    if (defPtr && defPtr->flags & VAR_IS_ARGS) {
+	Tcl_Obj *listPtr = Tcl_NewListObj(argCt-i, argObjs+i);
+
+	varPtr->value.objPtr = listPtr;
+	Tcl_IncrRefCount(listPtr);	/* Local var is a reference. */
+    } else if (argCt == numArgs) {
+	Tcl_Obj *objPtr = argObjs[i];
+
+	varPtr->value.objPtr = objPtr;
+	Tcl_IncrRefCount(objPtr);	/* Local var is a reference. */
+    } else if ((argCt < numArgs) && defPtr && defPtr->value.objPtr) {
+	Tcl_Obj *objPtr = defPtr->value.objPtr;
+
+	varPtr->value.objPtr = objPtr;
+	Tcl_IncrRefCount(objPtr);	/* Local var is a reference. */
+    } else {
+	goto incorrectArgs;
+    }
+    varPtr++;
+
+    /*
+     * Initialise and resolve the remaining compiledLocals. In the absence of
+     * resolvers, they are undefined local vars: (flags=0, value=NULL).
+     */
+
+  correctArgs:
+    if (numArgs < localCt) {
+	if (!framePtr->nsPtr->compiledVarResProc
+		&& !((Interp *)interp)->resolverPtr) {
+	    memset(varPtr, 0, (localCt - numArgs)*sizeof(Var));
+	} else {
+	    InitResolvedLocals(interp, codePtr, varPtr, framePtr->nsPtr);
+	}
+    }
+
+    return TCL_OK;
+
+    /*
+     * Initialise all compiled locals to avoid problems at DeleteLocalVars.
+     */
+
+  incorrectArgs:
+    if ((skip != 1) &&
+	    TclInitRewriteEnsemble(interp, skip-1, 0, framePtr->objv)) {
+	TclNRAddCallback(interp, TclClearRootEnsemble, NULL, NULL, NULL, NULL);
+    }
+    memset(varPtr, 0,
+	    ((framePtr->compiledLocals + localCt)-varPtr) * sizeof(Var));
+    return ProcWrongNumArgs(interp, skip);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPushProcCallFrame --
+ *
+ *	Compiles a proc body if necessary, then pushes a CallFrame suitable
+ *	for executing it.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	The proc's body may be recompiled. A CallFrame is pushed, it will have
+ *	to be popped by the caller.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPushProcCallFrame(
+    ClientData clientData,	/* Record describing procedure to be
+				 * interpreted. */
+    register Tcl_Interp *interp,/* Interpreter in which procedure was
+				 * invoked. */
+    int objc,			/* Count of number of arguments to this
+				 * procedure. */
+    Tcl_Obj *const objv[],	/* Argument value objects. */
+    int isLambda)		/* 1 if this is a call by ApplyObjCmd: it
+				 * needs special rules for error msg */
+{
+    Proc *procPtr = clientData;
+    Namespace *nsPtr = procPtr->cmdPtr->nsPtr;
+    CallFrame *framePtr, **framePtrPtr;
+    int result;
+    ByteCode *codePtr;
+
+    /*
+     * If necessary (i.e. if we haven't got a suitable compilation already
+     * cached) compile the procedure's body. The compiler will allocate frame
+     * slots for the procedure's non-argument local variables. Note that
+     * compiling the body might increase procPtr->numCompiledLocals if new
+     * local variables are found while compiling.
+     */
+
+    if (procPtr->bodyPtr->typePtr == &tclByteCodeType) {
+	Interp *iPtr = (Interp *) interp;
+
+	/*
+	 * When we've got bytecode, this is the check for validity. That is,
+	 * the bytecode must be for the right interpreter (no cross-leaks!),
+	 * the code must be from the current epoch (so subcommand compilation
+	 * is up-to-date), the namespace must match (so variable handling
+	 * is right) and the resolverEpoch must match (so that new shadowed
+	 * commands and/or resolver changes are considered).
+	 */
+
+	codePtr = procPtr->bodyPtr->internalRep.twoPtrValue.ptr1;
+	if (((Interp *) *codePtr->interpHandle != iPtr)
+		|| (codePtr->compileEpoch != iPtr->compileEpoch)
+		|| (codePtr->nsPtr != nsPtr)
+		|| (codePtr->nsEpoch != nsPtr->resolverEpoch)) {
+	    goto doCompilation;
+	}
+    } else {
+    doCompilation:
+	result = TclProcCompileProc(interp, procPtr, procPtr->bodyPtr, nsPtr,
+		(isLambda ? "body of lambda term" : "body of proc"),
+		TclGetString(objv[isLambda]));
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+
+    /*
+     * Set up and push a new call frame for the new procedure invocation.
+     * This call frame will execute in the proc's namespace, which might be
+     * different than the current namespace. The proc's namespace is that of
+     * its command, which can change if the command is renamed from one
+     * namespace to another.
+     */
+
+    framePtrPtr = &framePtr;
+    (void) TclPushStackFrame(interp, (Tcl_CallFrame **) framePtrPtr,
+	    (Tcl_Namespace *) nsPtr,
+	    (isLambda? (FRAME_IS_PROC|FRAME_IS_LAMBDA) : FRAME_IS_PROC));
+
+    framePtr->objc = objc;
+    framePtr->objv = objv;
+    framePtr->procPtr = procPtr;
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjInterpProc --
+ *
+ *	When a Tcl procedure gets invoked during bytecode evaluation, this
+ *	object-based routine gets invoked to interpret the procedure.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	Depends on the commands in the procedure.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjInterpProc(
+    ClientData clientData,	/* Record describing procedure to be
+				 * interpreted. */
+    register Tcl_Interp *interp,/* Interpreter in which procedure was
+				 * invoked. */
+    int objc,			/* Count of number of arguments to this
+				 * procedure. */
+    Tcl_Obj *const objv[])	/* Argument value objects. */
+{
+    /*
+     * Not used much in the core; external interface for iTcl
+     */
+
+    return Tcl_NRCallObjProc(interp, TclNRInterpProc, clientData, objc, objv);
+}
+
+int
+TclNRInterpProc(
+    ClientData clientData,	/* Record describing procedure to be
+				 * interpreted. */
+    register Tcl_Interp *interp,/* Interpreter in which procedure was
+				 * invoked. */
+    int objc,			/* Count of number of arguments to this
+				 * procedure. */
+    Tcl_Obj *const objv[])	/* Argument value objects. */
+{
+    int result = TclPushProcCallFrame(clientData, interp, objc, objv,
+	    /*isLambda*/ 0);
+
+    if (result != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return TclNRInterpProcCore(interp, objv[0], 1, &MakeProcError);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNRInterpProcCore --
+ *
+ *	When a Tcl procedure, lambda term or anything else that works like a
+ *	procedure gets invoked during bytecode evaluation, this object-based
+ *	routine gets invoked to interpret the body.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	Nearly anything; depends on the commands in the procedure body.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNRInterpProcCore(
+    register Tcl_Interp *interp,/* Interpreter in which procedure was
+				 * invoked. */
+    Tcl_Obj *procNameObj,	/* Procedure name for error reporting. */
+    int skip,			/* Number of initial arguments to be skipped,
+				 * i.e., words in the "command name". */
+    ProcErrorProc *errorProc)	/* How to convert results from the script into
+				 * results of the overall procedure. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register Proc *procPtr = iPtr->varFramePtr->procPtr;
+    int result;
+    CallFrame *freePtr;
+    ByteCode *codePtr;
+
+    result = InitArgsAndLocals(interp, procNameObj, skip);
+    if (result != TCL_OK) {
+	freePtr = iPtr->framePtr;
+	Tcl_PopCallFrame(interp);	/* Pop but do not free. */
+	TclStackFree(interp, freePtr->compiledLocals);
+					/* Free compiledLocals. */
+	TclStackFree(interp, freePtr);	/* Free CallFrame. */
+	return TCL_ERROR;
+    }
+
+#if defined(TCL_COMPILE_DEBUG)
+    if (tclTraceExec >= 1) {
+	register CallFrame *framePtr = iPtr->varFramePtr;
+	register int i;
+
+	if (framePtr->isProcCallFrame & FRAME_IS_LAMBDA) {
+	    fprintf(stdout, "Calling lambda ");
+	} else {
+	    fprintf(stdout, "Calling proc ");
+	}
+	for (i = 0; i < framePtr->objc; i++) {
+	    TclPrintObject(stdout, framePtr->objv[i], 15);
+	    fprintf(stdout, " ");
+	}
+	fprintf(stdout, "\n");
+	fflush(stdout);
+    }
+#endif /*TCL_COMPILE_DEBUG*/
+
+#ifdef USE_DTRACE
+    if (TCL_DTRACE_PROC_ARGS_ENABLED()) {
+	int l = iPtr->varFramePtr->isProcCallFrame & FRAME_IS_LAMBDA ? 1 : 0;
+	const char *a[10];
+	int i;
+
+	for (i = 0 ; i < 10 ; i++) {
+	    a[i] = (l < iPtr->varFramePtr->objc ?
+		    TclGetString(iPtr->varFramePtr->objv[l]) : NULL);
+	    l++;
+	}
+	TCL_DTRACE_PROC_ARGS(a[0], a[1], a[2], a[3], a[4], a[5], a[6], a[7],
+		a[8], a[9]);
+    }
+    if (TCL_DTRACE_PROC_INFO_ENABLED() && iPtr->cmdFramePtr) {
+	Tcl_Obj *info = TclInfoFrame(interp, iPtr->cmdFramePtr);
+	const char *a[6]; int i[2];
+
+	TclDTraceInfo(info, a, i);
+	TCL_DTRACE_PROC_INFO(a[0], a[1], a[2], a[3], i[0], i[1], a[4], a[5]);
+	TclDecrRefCount(info);
+    }
+    if (TCL_DTRACE_PROC_ENTRY_ENABLED()) {
+	int l = iPtr->varFramePtr->isProcCallFrame & FRAME_IS_LAMBDA ? 1 : 0;
+
+	TCL_DTRACE_PROC_ENTRY(l < iPtr->varFramePtr->objc ?
+		TclGetString(iPtr->varFramePtr->objv[l]) : NULL,
+		iPtr->varFramePtr->objc - l - 1,
+		(Tcl_Obj **)(iPtr->varFramePtr->objv + l + 1));
+    }
+    if (TCL_DTRACE_PROC_ENTRY_ENABLED()) {
+	int l = iPtr->varFramePtr->isProcCallFrame & FRAME_IS_LAMBDA ? 1 : 0;
+
+	TCL_DTRACE_PROC_ENTRY(l < iPtr->varFramePtr->objc ?
+		TclGetString(iPtr->varFramePtr->objv[l]) : NULL,
+		iPtr->varFramePtr->objc - l - 1,
+		(Tcl_Obj **)(iPtr->varFramePtr->objv + l + 1));
+    }
+#endif /* USE_DTRACE */
+
+    /*
+     * Invoke the commands in the procedure's body.
+     */
+
+    procPtr->refCount++;
+    codePtr = procPtr->bodyPtr->internalRep.twoPtrValue.ptr1;
+
+    TclNRAddCallback(interp, InterpProcNR2, procNameObj, errorProc,
+	    NULL, NULL);
+    return TclNRExecuteByteCode(interp, codePtr);
+}
+
+static int
+InterpProcNR2(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Proc *procPtr = iPtr->varFramePtr->procPtr;
+    CallFrame *freePtr;
+    Tcl_Obj *procNameObj = data[0];
+    ProcErrorProc *errorProc = (ProcErrorProc *)data[1];
+
+    if (TCL_DTRACE_PROC_RETURN_ENABLED()) {
+	int l = iPtr->varFramePtr->isProcCallFrame & FRAME_IS_LAMBDA ? 1 : 0;
+
+	TCL_DTRACE_PROC_RETURN(l < iPtr->varFramePtr->objc ?
+		TclGetString(iPtr->varFramePtr->objv[l]) : NULL, result);
+    }
+    if (procPtr->refCount-- <= 1) {
+	TclProcCleanupProc(procPtr);
+    }
+
+    /*
+     * Free the stack-allocated compiled locals and CallFrame. It is important
+     * to pop the call frame without freeing it first: the compiledLocals
+     * cannot be freed before the frame is popped, as the local variables must
+     * be deleted. But the compiledLocals must be freed first, as they were
+     * allocated later on the stack.
+     */
+
+    if (result != TCL_OK) {
+	goto process;
+    }
+
+    done:
+    if (TCL_DTRACE_PROC_RESULT_ENABLED()) {
+	int l = iPtr->varFramePtr->isProcCallFrame & FRAME_IS_LAMBDA ? 1 : 0;
+	Tcl_Obj *r = Tcl_GetObjResult(interp);
+
+	TCL_DTRACE_PROC_RESULT(l < iPtr->varFramePtr->objc ?
+		TclGetString(iPtr->varFramePtr->objv[l]) : NULL, result,
+		TclGetString(r), r);
+    }
+
+    freePtr = iPtr->framePtr;
+    Tcl_PopCallFrame(interp);		/* Pop but do not free. */
+    TclStackFree(interp, freePtr->compiledLocals);
+					/* Free compiledLocals. */
+    TclStackFree(interp, freePtr);	/* Free CallFrame. */
+    return result;
+
+    /*
+     * Process any non-TCL_OK result code.
+     */
+
+    process:
+    switch (result) {
+    case TCL_RETURN:
+	/*
+	 * If it is a 'return', do the TIP#90 processing now.
+	 */
+
+	result = TclUpdateReturnInfo((Interp *) interp);
+	break;
+
+    case TCL_CONTINUE:
+    case TCL_BREAK:
+	/*
+	 * It's an error to get to this point from a 'break' or 'continue', so
+	 * transform to an error now.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"invoked \"%s\" outside of a loop",
+		((result == TCL_BREAK) ? "break" : "continue")));
+	Tcl_SetErrorCode(interp, "TCL", "RESULT", "UNEXPECTED", NULL);
+	result = TCL_ERROR;
+
+	/*
+	 * Fall through to the TCL_ERROR handling code.
+	 */
+
+    case TCL_ERROR:
+	/*
+	 * Now it _must_ be an error, so we need to log it as such. This means
+	 * filling out the error trace. Luckily, we just hand this off to the
+	 * function handed to us as an argument.
+	 */
+
+	errorProc(interp, procNameObj);
+    }
+    goto done;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclProcCompileProc --
+ *
+ *	Called just before a procedure is executed to compile the body to byte
+ *	codes. If the type of the body is not "byte code" or if the compile
+ *	conditions have changed (namespace context, epoch counters, etc.) then
+ *	the body is recompiled. Otherwise, this function does nothing.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May change the internal representation of the body object to compiled
+ *	code.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclProcCompileProc(
+    Tcl_Interp *interp,		/* Interpreter containing procedure. */
+    Proc *procPtr,		/* Data associated with procedure. */
+    Tcl_Obj *bodyPtr,		/* Body of proc. (Usually procPtr->bodyPtr,
+				 * but could be any code fragment compiled in
+				 * the context of this procedure.) */
+    Namespace *nsPtr,		/* Namespace containing procedure. */
+    const char *description,	/* string describing this body of code. */
+    const char *procName)	/* Name of this procedure. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_CallFrame *framePtr;
+    ByteCode *codePtr = bodyPtr->internalRep.twoPtrValue.ptr1;
+
+    /*
+     * If necessary, compile the procedure's body. The compiler will allocate
+     * frame slots for the procedure's non-argument local variables. If the
+     * ByteCode already exists, make sure it hasn't been invalidated by
+     * someone redefining a core command (this might make the compiled code
+     * wrong). Also, if the code was compiled in/for a different interpreter,
+     * we recompile it. Note that compiling the body might increase
+     * procPtr->numCompiledLocals if new local variables are found while
+     * compiling.
+     *
+     * Precompiled procedure bodies, however, are immutable and therefore they
+     * are not recompiled, even if things have changed.
+     */
+
+    if (bodyPtr->typePtr == &tclByteCodeType) {
+	if (((Interp *) *codePtr->interpHandle == iPtr)
+		&& (codePtr->compileEpoch == iPtr->compileEpoch)
+		&& (codePtr->nsPtr == nsPtr)
+		&& (codePtr->nsEpoch == nsPtr->resolverEpoch)) {
+	    return TCL_OK;
+	}
+
+	if (codePtr->flags & TCL_BYTECODE_PRECOMPILED) {
+	    if ((Interp *) *codePtr->interpHandle != iPtr) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"a precompiled script jumped interps", -1));
+		Tcl_SetErrorCode(interp, "TCL", "OPERATION", "PROC",
+			"CROSSINTERPBYTECODE", NULL);
+		return TCL_ERROR;
+	    }
+	    codePtr->compileEpoch = iPtr->compileEpoch;
+	    codePtr->nsPtr = nsPtr;
+	} else {
+	    TclFreeIntRep(bodyPtr);
+	}
+    }
+
+    if (bodyPtr->typePtr != &tclByteCodeType) {
+	Tcl_HashEntry *hePtr;
+
+#ifdef TCL_COMPILE_DEBUG
+	if (tclTraceCompile >= 1) {
+	    /*
+	     * Display a line summarizing the top level command we are about
+	     * to compile.
+	     */
+
+	    Tcl_Obj *message;
+
+	    TclNewLiteralStringObj(message, "Compiling ");
+	    Tcl_IncrRefCount(message);
+	    Tcl_AppendStringsToObj(message, description, " \"", NULL);
+	    Tcl_AppendLimitedToObj(message, procName, -1, 50, NULL);
+	    fprintf(stdout, "%s\"\n", TclGetString(message));
+	    Tcl_DecrRefCount(message);
+	}
+#endif
+
+	/*
+	 * Plug the current procPtr into the interpreter and coerce the code
+	 * body to byte codes. The interpreter needs to know which proc it's
+	 * compiling so that it can access its list of compiled locals.
+	 *
+	 * TRICKY NOTE: Be careful to push a call frame with the proper
+	 *   namespace context, so that the byte codes are compiled in the
+	 *   appropriate class context.
+	 */
+
+	iPtr->compiledProcPtr = procPtr;
+
+	if (procPtr->numCompiledLocals > procPtr->numArgs) {
+	    CompiledLocal *clPtr = procPtr->firstLocalPtr;
+	    CompiledLocal *lastPtr = NULL;
+	    int i, numArgs = procPtr->numArgs;
+
+	    for (i = 0; i < numArgs; i++) {
+		lastPtr = clPtr;
+		clPtr = clPtr->nextPtr;
+	    }
+
+	    if (lastPtr) {
+		lastPtr->nextPtr = NULL;
+	    } else {
+		procPtr->firstLocalPtr = NULL;
+	    }
+	    procPtr->lastLocalPtr = lastPtr;
+	    while (clPtr) {
+		CompiledLocal *toFree = clPtr;
+
+		clPtr = clPtr->nextPtr;
+		if (toFree->resolveInfo) {
+		    if (toFree->resolveInfo->deleteProc) {
+			toFree->resolveInfo->deleteProc(toFree->resolveInfo);
+		    } else {
+			ckfree(toFree->resolveInfo);
+		    }
+		}
+		ckfree(toFree);
+	    }
+	    procPtr->numCompiledLocals = procPtr->numArgs;
+	}
+
+	(void) TclPushStackFrame(interp, &framePtr, (Tcl_Namespace *) nsPtr,
+		/* isProcCallFrame */ 0);
+
+	/*
+	 * TIP #280: We get the invoking context from the cmdFrame which
+	 * was saved by 'Tcl_ProcObjCmd' (using linePBodyPtr).
+	 */
+
+	hePtr = Tcl_FindHashEntry(iPtr->linePBodyPtr, (char *) procPtr);
+
+	/*
+	 * Constructed saved frame has body as word 0. See Tcl_ProcObjCmd.
+	 */
+
+	iPtr->invokeWord = 0;
+	iPtr->invokeCmdFramePtr = (hePtr ? Tcl_GetHashValue(hePtr) : NULL);
+	TclSetByteCodeFromAny(interp, bodyPtr, NULL, NULL);
+	iPtr->invokeCmdFramePtr = NULL;
+	TclPopStackFrame(interp);
+    } else if (codePtr->nsEpoch != nsPtr->resolverEpoch) {
+	/*
+	 * The resolver epoch has changed, but we only need to invalidate the
+	 * resolver cache.
+	 */
+
+	codePtr->nsEpoch = nsPtr->resolverEpoch;
+	codePtr->flags |= TCL_BYTECODE_RESOLVE_VARS;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeProcError --
+ *
+ *	Function called by TclObjInterpProc to create the stack information
+ *	upon an error from a procedure.
+ *
+ * Results:
+ *	The interpreter's error info trace is set to a value that supplements
+ *	the error code.
+ *
+ * Side effects:
+ *	none.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MakeProcError(
+    Tcl_Interp *interp,		/* The interpreter in which the procedure was
+				 * called. */
+    Tcl_Obj *procNameObj)	/* Name of the procedure. Used for error
+				 * messages and trace information. */
+{
+    int overflow, limit = 60, nameLen;
+    const char *procName = TclGetStringFromObj(procNameObj, &nameLen);
+
+    overflow = (nameLen > limit);
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (procedure \"%.*s%s\" line %d)",
+	    (overflow ? limit : nameLen), procName,
+	    (overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclProcDeleteProc --
+ *
+ *	This function is invoked just before a command procedure is removed
+ *	from an interpreter. Its job is to release all the resources allocated
+ *	to the procedure.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory gets freed, unless the procedure is actively being executed.
+ *	In this case the cleanup is delayed until the last call to the current
+ *	procedure completes.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclProcDeleteProc(
+    ClientData clientData)	/* Procedure to be deleted. */
+{
+    Proc *procPtr = clientData;
+
+    if (procPtr->refCount-- <= 1) {
+	TclProcCleanupProc(procPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclProcCleanupProc --
+ *
+ *	This function does all the real work of freeing up a Proc structure.
+ *	It's called only when the structure's reference count becomes zero.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory gets freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclProcCleanupProc(
+    register Proc *procPtr)	/* Procedure to be deleted. */
+{
+    register CompiledLocal *localPtr;
+    Tcl_Obj *bodyPtr = procPtr->bodyPtr;
+    Tcl_Obj *defPtr;
+    Tcl_ResolvedVarInfo *resVarInfo;
+    Tcl_HashEntry *hePtr = NULL;
+    CmdFrame *cfPtr = NULL;
+    Interp *iPtr = procPtr->iPtr;
+
+    if (bodyPtr != NULL) {
+	Tcl_DecrRefCount(bodyPtr);
+    }
+    for (localPtr = procPtr->firstLocalPtr; localPtr != NULL; ) {
+	CompiledLocal *nextPtr = localPtr->nextPtr;
+
+	resVarInfo = localPtr->resolveInfo;
+	if (resVarInfo) {
+	    if (resVarInfo->deleteProc) {
+		resVarInfo->deleteProc(resVarInfo);
+	    } else {
+		ckfree(resVarInfo);
+	    }
+	}
+
+	if (localPtr->defValuePtr != NULL) {
+	    defPtr = localPtr->defValuePtr;
+	    Tcl_DecrRefCount(defPtr);
+	}
+	ckfree(localPtr);
+	localPtr = nextPtr;
+    }
+    ckfree(procPtr);
+
+    /*
+     * TIP #280: Release the location data associated with this Proc
+     * structure, if any. The interpreter may not exist (For example for
+     * procbody structures created by tbcload.
+     */
+
+    if (iPtr == NULL) {
+	return;
+    }
+
+    hePtr = Tcl_FindHashEntry(iPtr->linePBodyPtr, (char *) procPtr);
+    if (!hePtr) {
+	return;
+    }
+
+    cfPtr = Tcl_GetHashValue(hePtr);
+
+    if (cfPtr) {
+	if (cfPtr->type == TCL_LOCATION_SOURCE) {
+	    Tcl_DecrRefCount(cfPtr->data.eval.path);
+	    cfPtr->data.eval.path = NULL;
+	}
+	ckfree(cfPtr->line);
+	cfPtr->line = NULL;
+	ckfree(cfPtr);
+    }
+    Tcl_DeleteHashEntry(hePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclUpdateReturnInfo --
+ *
+ *	This function is called when procedures return, and at other points
+ *	where the TCL_RETURN code is used. It examines the returnLevel and
+ *	returnCode to determine the real return status.
+ *
+ * Results:
+ *	The return value is the true completion code to use for the procedure
+ *	or script, instead of TCL_RETURN.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclUpdateReturnInfo(
+    Interp *iPtr)		/* Interpreter for which TCL_RETURN exception
+				 * is being processed. */
+{
+    int code = TCL_RETURN;
+
+    iPtr->returnLevel--;
+    if (iPtr->returnLevel < 0) {
+	Tcl_Panic("TclUpdateReturnInfo: negative return level");
+    }
+    if (iPtr->returnLevel == 0) {
+	/*
+	 * Now we've reached the level to return the requested -code.
+	 * Since iPtr->returnLevel and iPtr->returnCode have completed
+	 * their task, we now reset them to default values so that any
+	 * bare "return TCL_RETURN" that may follow will work [Bug 2152286].
+	 */
+
+	code = iPtr->returnCode;
+	iPtr->returnLevel = 1;
+	iPtr->returnCode = TCL_OK;
+	if (code == TCL_ERROR) {
+	    iPtr->flags |= ERR_LEGACY_COPY;
+	}
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetObjInterpProc --
+ *
+ *	Returns a pointer to the TclObjInterpProc function; this is different
+ *	from the value obtained from the TclObjInterpProc reference on systems
+ *	like Windows where import and export versions of a function exported
+ *	by a DLL exist.
+ *
+ * Results:
+ *	Returns the internal address of the TclObjInterpProc function.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+TclObjCmdProcType
+TclGetObjInterpProc(void)
+{
+    return (TclObjCmdProcType) TclObjInterpProc;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNewProcBodyObj --
+ *
+ *	Creates a new object, of type "procbody", whose internal
+ *	representation is the given Proc struct. The newly created object's
+ *	reference count is 0.
+ *
+ * Results:
+ *	Returns a pointer to a newly allocated Tcl_Obj, NULL on error.
+ *
+ * Side effects:
+ *	The reference count in the ByteCode attached to the Proc is bumped up
+ *	by one, since the internal rep stores a pointer to it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclNewProcBodyObj(
+    Proc *procPtr)		/* the Proc struct to store as the internal
+				 * representation. */
+{
+    Tcl_Obj *objPtr;
+
+    if (!procPtr) {
+	return NULL;
+    }
+
+    TclNewObj(objPtr);
+    if (objPtr) {
+	objPtr->typePtr = &tclProcBodyType;
+	objPtr->internalRep.twoPtrValue.ptr1 = procPtr;
+
+	procPtr->refCount++;
+    }
+
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ProcBodyDup --
+ *
+ *	Tcl_ObjType's Dup function for the proc body object. Bumps the
+ *	reference count on the Proc stored in the internal representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Sets up the object in dupPtr to be a duplicate of the one in srcPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ProcBodyDup(
+    Tcl_Obj *srcPtr,		/* Object to copy. */
+    Tcl_Obj *dupPtr)		/* Target object for the duplication. */
+{
+    Proc *procPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    dupPtr->typePtr = &tclProcBodyType;
+    dupPtr->internalRep.twoPtrValue.ptr1 = procPtr;
+    procPtr->refCount++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ProcBodyFree --
+ *
+ *	Tcl_ObjType's Free function for the proc body object. The reference
+ *	count on its Proc struct is decreased by 1; if the count reaches 0,
+ *	the proc is freed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the reference count on the Proc struct reaches 0, the struct is
+ *	freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ProcBodyFree(
+    Tcl_Obj *objPtr)		/* The object to clean up. */
+{
+    Proc *procPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    if (procPtr->refCount-- <= 1) {
+	TclProcCleanupProc(procPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupLambdaInternalRep, FreeLambdaInternalRep, SetLambdaFromAny --
+ *
+ *	How to manage the internal representations of lambda term objects.
+ *	Syntactically they look like a two- or three-element list, where the
+ *	first element is the formal arguments, the second is the the body, and
+ *	the (optional) third is the namespace to execute the lambda term
+ *	within (the global namespace is assumed if it is absent).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupLambdaInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    register Tcl_Obj *copyPtr)	/* Object with internal rep to set. */
+{
+    Proc *procPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+    Tcl_Obj *nsObjPtr = srcPtr->internalRep.twoPtrValue.ptr2;
+
+    copyPtr->internalRep.twoPtrValue.ptr1 = procPtr;
+    copyPtr->internalRep.twoPtrValue.ptr2 = nsObjPtr;
+
+    procPtr->refCount++;
+    Tcl_IncrRefCount(nsObjPtr);
+    copyPtr->typePtr = &tclLambdaType;
+}
+
+static void
+FreeLambdaInternalRep(
+    register Tcl_Obj *objPtr)	/* CmdName object with internal representation
+				 * to free. */
+{
+    Proc *procPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    Tcl_Obj *nsObjPtr = objPtr->internalRep.twoPtrValue.ptr2;
+
+    if (procPtr->refCount-- == 1) {
+	TclProcCleanupProc(procPtr);
+    }
+    TclDecrRefCount(nsObjPtr);
+    objPtr->typePtr = NULL;
+}
+
+static int
+SetLambdaFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    register Tcl_Obj *objPtr)	/* The object to convert. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *name;
+    Tcl_Obj *argsPtr, *bodyPtr, *nsObjPtr, **objv;
+    int isNew, objc, result;
+    CmdFrame *cfPtr = NULL;
+    Proc *procPtr;
+
+    if (interp == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert objPtr to list type first; if it cannot be converted, or if its
+     * length is not 2, then it cannot be converted to tclLambdaType.
+     */
+
+    result = TclListObjGetElements(NULL, objPtr, &objc, &objv);
+    if ((result != TCL_OK) || ((objc != 2) && (objc != 3))) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"can't interpret \"%s\" as a lambda expression",
+		Tcl_GetString(objPtr)));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "LAMBDA", NULL);
+	return TCL_ERROR;
+    }
+
+    argsPtr = objv[0];
+    bodyPtr = objv[1];
+
+    /*
+     * Create and initialize the Proc struct. The cmdPtr field is set to NULL
+     * to signal that this is an anonymous function.
+     */
+
+    name = TclGetString(objPtr);
+
+    if (TclCreateProc(interp, /*ignored nsPtr*/ NULL, name, argsPtr, bodyPtr,
+	    &procPtr) != TCL_OK) {
+	Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+		"\n    (parsing lambda expression \"%s\")", name));
+	return TCL_ERROR;
+    }
+
+    /*
+     * CAREFUL: TclCreateProc returns refCount==1! [Bug 1578454]
+     * procPtr->refCount = 1;
+     */
+
+    procPtr->cmdPtr = NULL;
+
+    /*
+     * TIP #280: Remember the line the apply body is starting on. In a Byte
+     * code context we ask the engine to provide us with the necessary
+     * information. This is for the initialization of the byte code compiler
+     * when the body is used for the first time.
+     *
+     * NOTE: The body is the second word in the 'objPtr'. Its location,
+     * accessible through 'context.line[1]' (see below) is therefore only the
+     * first approximation of the actual line the body is on. We have to use
+     * the string rep of the 'objPtr' to determine the exact line. This is
+     * available already through 'name'. Use 'TclListLines', see 'switch'
+     * (tclCmdMZ.c).
+     *
+     * This code is nearly identical to the #280 code in Tcl_ProcObjCmd, see
+     * this file. The differences are the different index of the body in the
+     * line array of the context, and the special processing mentioned in the
+     * previous paragraph to track into the list. Find a way to factor the
+     * common elements into a single function.
+     */
+
+    if (iPtr->cmdFramePtr) {
+	CmdFrame *contextPtr = TclStackAlloc(interp, sizeof(CmdFrame));
+
+	*contextPtr = *iPtr->cmdFramePtr;
+	if (contextPtr->type == TCL_LOCATION_BC) {
+	    /*
+	     * Retrieve the source context from the bytecode. This call
+	     * accounts for the reference to the source file, if any, held in
+	     * 'context.data.eval.path'.
+	     */
+
+	    TclGetSrcInfoForPc(contextPtr);
+	} else if (contextPtr->type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * We created a new reference to the source file path name when we
+	     * created 'context' above. Account for the reference.
+	     */
+
+	    Tcl_IncrRefCount(contextPtr->data.eval.path);
+
+	}
+
+	if (contextPtr->type == TCL_LOCATION_SOURCE) {
+	    /*
+	     * We can record source location within a lambda only if the body
+	     * was not created by substitution.
+	     */
+
+	    if (contextPtr->line
+		    && (contextPtr->nline >= 2) && (contextPtr->line[1] >= 0)) {
+		int buf[2];
+
+		/*
+		 * Move from approximation (line of list cmd word) to actual
+		 * location (line of 2nd list element).
+		 */
+
+		cfPtr = ckalloc(sizeof(CmdFrame));
+		TclListLines(objPtr, contextPtr->line[1], 2, buf, NULL);
+
+		cfPtr->level = -1;
+		cfPtr->type = contextPtr->type;
+		cfPtr->line = ckalloc(sizeof(int));
+		cfPtr->line[0] = buf[1];
+		cfPtr->nline = 1;
+		cfPtr->framePtr = NULL;
+		cfPtr->nextPtr = NULL;
+
+		cfPtr->data.eval.path = contextPtr->data.eval.path;
+		Tcl_IncrRefCount(cfPtr->data.eval.path);
+
+		cfPtr->cmd = NULL;
+		cfPtr->len = 0;
+	    }
+
+	    /*
+	     * 'contextPtr' is going out of scope. Release the reference that
+	     * it's holding to the source file path
+	     */
+
+	    Tcl_DecrRefCount(contextPtr->data.eval.path);
+	}
+	TclStackFree(interp, contextPtr);
+    }
+    Tcl_SetHashValue(Tcl_CreateHashEntry(iPtr->linePBodyPtr, procPtr,
+	    &isNew), cfPtr);
+
+    /*
+     * Set the namespace for this lambda: given by objv[2] understood as a
+     * global reference, or else global per default.
+     */
+
+    if (objc == 2) {
+	TclNewLiteralStringObj(nsObjPtr, "::");
+    } else {
+	const char *nsName = TclGetString(objv[2]);
+
+	if ((*nsName != ':') || (*(nsName+1) != ':')) {
+	    TclNewLiteralStringObj(nsObjPtr, "::");
+	    Tcl_AppendObjToObj(nsObjPtr, objv[2]);
+	} else {
+	    nsObjPtr = objv[2];
+	}
+    }
+
+    Tcl_IncrRefCount(nsObjPtr);
+
+    /*
+     * Free the list internalrep of objPtr - this will free argsPtr, but
+     * bodyPtr retains a reference from the Proc structure. Then finish the
+     * conversion to tclLambdaType.
+     */
+
+    TclFreeIntRep(objPtr);
+
+    objPtr->internalRep.twoPtrValue.ptr1 = procPtr;
+    objPtr->internalRep.twoPtrValue.ptr2 = nsObjPtr;
+    objPtr->typePtr = &tclLambdaType;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ApplyObjCmd --
+ *
+ *	This object-based function is invoked to process the "apply" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	Depends on the content of the lambda term (i.e., objv[1]).
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ApplyObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return Tcl_NRCallObjProc(interp, TclNRApplyObjCmd, dummy, objc, objv);
+}
+
+int
+TclNRApplyObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Proc *procPtr = NULL;
+    Tcl_Obj *lambdaPtr, *nsObjPtr;
+    int result;
+    Tcl_Namespace *nsPtr;
+    ApplyExtraData *extraPtr;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "lambdaExpr ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set lambdaPtr, convert it to tclLambdaType in the current interp if
+     * necessary.
+     */
+
+    lambdaPtr = objv[1];
+    if (lambdaPtr->typePtr == &tclLambdaType) {
+	procPtr = lambdaPtr->internalRep.twoPtrValue.ptr1;
+    }
+
+    if ((procPtr == NULL) || (procPtr->iPtr != iPtr)) {
+	result = SetLambdaFromAny(interp, lambdaPtr);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	procPtr = lambdaPtr->internalRep.twoPtrValue.ptr1;
+    }
+
+    /*
+     * Find the namespace where this lambda should run, and push a call frame
+     * for that namespace. Note that TclObjInterpProc() will pop it.
+     */
+
+    nsObjPtr = lambdaPtr->internalRep.twoPtrValue.ptr2;
+    result = TclGetNamespaceFromObj(interp, nsObjPtr, &nsPtr);
+    if (result != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    extraPtr = TclStackAlloc(interp, sizeof(ApplyExtraData));
+    memset(&extraPtr->cmd, 0, sizeof(Command));
+    procPtr->cmdPtr = &extraPtr->cmd;
+    extraPtr->cmd.nsPtr = (Namespace *) nsPtr;
+
+    /*
+     * TIP#280 (semi-)HACK!
+     *
+     * Using cmd.clientData to tell [info frame] how to render the lambdaPtr.
+     * The InfoFrameCmd will detect this case by testing cmd.hPtr for NULL.
+     * This condition holds here because of the memset() above, and nowhere
+     * else (in the core). Regular commands always have a valid hPtr, and
+     * lambda's never.
+     */
+
+    extraPtr->efi.length = 1;
+    extraPtr->efi.fields[0].name = "lambda";
+    extraPtr->efi.fields[0].proc = NULL;
+    extraPtr->efi.fields[0].clientData = lambdaPtr;
+    extraPtr->cmd.clientData = &extraPtr->efi;
+
+    result = TclPushProcCallFrame(procPtr, interp, objc, objv, 1);
+    if (result == TCL_OK) {
+	TclNRAddCallback(interp, ApplyNR2, extraPtr, NULL, NULL, NULL);
+	result = TclNRInterpProcCore(interp, objv[1], 2, &MakeLambdaError);
+    }
+    return result;
+}
+
+static int
+ApplyNR2(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    ApplyExtraData *extraPtr = data[0];
+
+    TclStackFree(interp, extraPtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeLambdaError --
+ *
+ *	Function called by TclObjInterpProc to create the stack information
+ *	upon an error from a lambda term.
+ *
+ * Results:
+ *	The interpreter's error info trace is set to a value that supplements
+ *	the error code.
+ *
+ * Side effects:
+ *	none.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MakeLambdaError(
+    Tcl_Interp *interp,		/* The interpreter in which the procedure was
+				 * called. */
+    Tcl_Obj *procNameObj)	/* Name of the procedure. Used for error
+				 * messages and trace information. */
+{
+    int overflow, limit = 60, nameLen;
+    const char *procName = TclGetStringFromObj(procNameObj, &nameLen);
+
+    overflow = (nameLen > limit);
+    Tcl_AppendObjToErrorInfo(interp, Tcl_ObjPrintf(
+	    "\n    (lambda term \"%.*s%s\" line %d)",
+	    (overflow ? limit : nameLen), procName,
+	    (overflow ? "..." : ""), Tcl_GetErrorLine(interp)));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetCmdFrameForProcedure --
+ *
+ *	How to get the CmdFrame information for a procedure.
+ *
+ * Results:
+ *	A pointer to the CmdFrame (only guaranteed to be valid until the next
+ *	Tcl command is processed or the interpreter's state is otherwise
+ *	modified) or a NULL if the information is not available.
+ *
+ * Side effects:
+ *	none.
+ *
+ *----------------------------------------------------------------------
+ */
+
+CmdFrame *
+TclGetCmdFrameForProcedure(
+    Proc *procPtr)		/* The procedure whose cmd-frame is to be
+				 * looked up. */
+{
+    Tcl_HashEntry *hePtr;
+
+    if (procPtr == NULL || procPtr->iPtr == NULL) {
+	return NULL;
+    }
+    hePtr = Tcl_FindHashEntry(procPtr->iPtr->linePBodyPtr, procPtr);
+    if (hePtr == NULL) {
+	return NULL;
+    }
+    return (CmdFrame *) Tcl_GetHashValue(hePtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclRegexp.c b/generic/tclRegexp.c
new file mode 100644
index 0000000..5f8dc20
--- /dev/null
+++ b/generic/tclRegexp.c
@@ -0,0 +1,1081 @@
+/*
+ * tclRegexp.c --
+ *
+ *	This file contains the public interfaces to the Tcl regular expression
+ *	mechanism.
+ *
+ * Copyright (c) 1998 by Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclRegexp.h"
+
+/*
+ *----------------------------------------------------------------------
+ * The routines in this file use Henry Spencer's regular expression package
+ * contained in the following additional source files:
+ *
+ *	regc_color.c	regc_cvec.c	regc_lex.c
+ *	regc_nfa.c	regcomp.c	regcustom.h
+ *	rege_dfa.c	regerror.c	regerrs.h
+ *	regex.h		regexec.c	regfree.c
+ *	regfronts.c	regguts.h
+ *
+ * Copyright (c) 1998 Henry Spencer.  All rights reserved.
+ *
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results. The author
+ * thanks all of them.
+ *
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ *
+ * I'd appreciate being given credit for this package in the documentation of
+ * software which uses it, but that is not a requirement.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * *** NOTE: this code has been altered slightly for use in Tcl: ***
+ * *** 1. Names have been changed, e.g. from re_comp to		 ***
+ * ***    TclRegComp, to avoid clashes with other		 ***
+ * ***    regexp implementations used by applications.		 ***
+ */
+
+/*
+ * Thread local storage used to maintain a per-thread cache of compiled
+ * regular expressions.
+ */
+
+#define NUM_REGEXPS 30
+
+typedef struct {
+    int initialized;		/* Set to 1 when the module is initialized. */
+    char *patterns[NUM_REGEXPS];/* Strings corresponding to compiled regular
+				 * expression patterns. NULL means that this
+				 * slot isn't used. Malloc-ed. */
+    int patLengths[NUM_REGEXPS];/* Number of non-null characters in
+				 * corresponding entry in patterns. -1 means
+				 * entry isn't used. */
+    struct TclRegexp *regexps[NUM_REGEXPS];
+				/* Compiled forms of above strings. Also
+				 * malloc-ed, or NULL if not in use yet. */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * Declarations for functions used only in this file.
+ */
+
+static TclRegexp *	CompileRegexp(Tcl_Interp *interp, const char *pattern,
+			    int length, int flags);
+static void		DupRegexpInternalRep(Tcl_Obj *srcPtr,
+			    Tcl_Obj *copyPtr);
+static void		FinalizeRegexp(ClientData clientData);
+static void		FreeRegexp(TclRegexp *regexpPtr);
+static void		FreeRegexpInternalRep(Tcl_Obj *objPtr);
+static int		RegExpExecUniChar(Tcl_Interp *interp, Tcl_RegExp re,
+			    const Tcl_UniChar *uniString, int numChars,
+			    int nmatches, int flags);
+static int		SetRegexpFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+
+/*
+ * The regular expression Tcl object type. This serves as a cache of the
+ * compiled form of the regular expression.
+ */
+
+const Tcl_ObjType tclRegexpType = {
+    "regexp",				/* name */
+    FreeRegexpInternalRep,		/* freeIntRepProc */
+    DupRegexpInternalRep,		/* dupIntRepProc */
+    NULL,				/* updateStringProc */
+    SetRegexpFromAny			/* setFromAnyProc */
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpCompile --
+ *
+ *	Compile a regular expression into a form suitable for fast matching.
+ *	This function is DEPRECATED in favor of the object version of the
+ *	command.
+ *
+ * Results:
+ *	The return value is a pointer to the compiled form of string, suitable
+ *	for passing to Tcl_RegExpExec. This compiled form is only valid up
+ *	until the next call to this function, so don't keep these around for a
+ *	long time! If an error occurred while compiling the pattern, then NULL
+ *	is returned and an error message is left in the interp's result.
+ *
+ * Side effects:
+ *	Updates the cache of compiled regexps.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_RegExp
+Tcl_RegExpCompile(
+    Tcl_Interp *interp,		/* For use in error reporting and to access
+				 * the interp regexp cache. */
+    const char *pattern)	/* String for which to produce compiled
+				 * regular expression. */
+{
+    return (Tcl_RegExp) CompileRegexp(interp, pattern, (int) strlen(pattern),
+	    REG_ADVANCED);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpExec --
+ *
+ *	Execute the regular expression matcher using a compiled form of a
+ *	regular expression and save information about any match that is found.
+ *
+ * Results:
+ *	If an error occurs during the matching operation then -1 is returned
+ *	and the interp's result contains an error message. Otherwise the
+ *	return value is 1 if a matching range is found and 0 if there is no
+ *	matching range.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegExpExec(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    Tcl_RegExp re,		/* Compiled regular expression; must have been
+				 * returned by previous call to
+				 * Tcl_GetRegExpFromObj. */
+    const char *text,		/* Text against which to match re. */
+    const char *start)		/* If text is part of a larger string, this
+				 * identifies beginning of larger string, so
+				 * that "^" won't match. */
+{
+    int flags, result, numChars;
+    TclRegexp *regexp = (TclRegexp *) re;
+    Tcl_DString ds;
+    const Tcl_UniChar *ustr;
+
+    /*
+     * If the starting point is offset from the beginning of the buffer, then
+     * we need to tell the regexp engine not to match "^".
+     */
+
+    if (text > start) {
+	flags = REG_NOTBOL;
+    } else {
+	flags = 0;
+    }
+
+    /*
+     * Remember the string for use by Tcl_RegExpRange().
+     */
+
+    regexp->string = text;
+    regexp->objPtr = NULL;
+
+    /*
+     * Convert the string to Unicode and perform the match.
+     */
+
+    Tcl_DStringInit(&ds);
+    ustr = Tcl_UtfToUniCharDString(text, -1, &ds);
+    numChars = Tcl_DStringLength(&ds) / sizeof(Tcl_UniChar);
+    result = RegExpExecUniChar(interp, re, ustr, numChars, -1 /* nmatches */,
+	    flags);
+    Tcl_DStringFree(&ds);
+
+    return result;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_RegExpRange --
+ *
+ *	Returns pointers describing the range of a regular expression match,
+ *	or one of the subranges within the match.
+ *
+ * Results:
+ *	The variables at *startPtr and *endPtr are modified to hold the
+ *	addresses of the endpoints of the range given by index. If the
+ *	specified range doesn't exist then NULLs are returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+Tcl_RegExpRange(
+    Tcl_RegExp re,		/* Compiled regular expression that has been
+				 * passed to Tcl_RegExpExec. */
+    int index,			/* 0 means give the range of the entire match,
+				 * > 0 means give the range of a matching
+				 * subrange. */
+    const char **startPtr,	/* Store address of first character in
+				 * (sub-)range here. */
+    const char **endPtr)	/* Store address of character just after last
+				 * in (sub-)range here. */
+{
+    TclRegexp *regexpPtr = (TclRegexp *) re;
+    const char *string;
+
+    if ((size_t) index > regexpPtr->re.re_nsub) {
+	*startPtr = *endPtr = NULL;
+    } else if (regexpPtr->matches[index].rm_so < 0) {
+	*startPtr = *endPtr = NULL;
+    } else {
+	if (regexpPtr->objPtr) {
+	    string = TclGetString(regexpPtr->objPtr);
+	} else {
+	    string = regexpPtr->string;
+	}
+	*startPtr = Tcl_UtfAtIndex(string, regexpPtr->matches[index].rm_so);
+	*endPtr = Tcl_UtfAtIndex(string, regexpPtr->matches[index].rm_eo);
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * RegExpExecUniChar --
+ *
+ *	Execute the regular expression matcher using a compiled form of a
+ *	regular expression and save information about any match that is found.
+ *
+ * Results:
+ *	If an error occurs during the matching operation then -1 is returned
+ *	and an error message is left in interp's result. Otherwise the return
+ *	value is 1 if a matching range was found or 0 if there was no matching
+ *	range.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+RegExpExecUniChar(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    Tcl_RegExp re,		/* Compiled regular expression; returned by a
+				 * previous call to Tcl_GetRegExpFromObj */
+    const Tcl_UniChar *wString,	/* String against which to match re. */
+    int numChars,		/* Length of Tcl_UniChar string (must be
+				 * >=0). */
+    int nmatches,		/* How many subexpression matches (counting
+				 * the whole match as subexpression 0) are of
+				 * interest. -1 means "don't know". */
+    int flags)			/* Regular expression flags. */
+{
+    int status;
+    TclRegexp *regexpPtr = (TclRegexp *) re;
+    size_t last = regexpPtr->re.re_nsub + 1;
+    size_t nm = last;
+
+    if (nmatches >= 0 && (size_t) nmatches < nm) {
+	nm = (size_t) nmatches;
+    }
+
+    status = TclReExec(&regexpPtr->re, wString, (size_t) numChars,
+	    &regexpPtr->details, nm, regexpPtr->matches, flags);
+
+    /*
+     * Check for errors.
+     */
+
+    if (status != REG_OKAY) {
+	if (status == REG_NOMATCH) {
+	    return 0;
+	}
+	if (interp != NULL) {
+	    TclRegError(interp, "error while matching regular expression: ",
+		    status);
+	}
+	return -1;
+    }
+    return 1;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclRegExpRangeUniChar --
+ *
+ *	Returns pointers describing the range of a regular expression match,
+ *	or one of the subranges within the match, or the hypothetical range
+ *	represented by the rm_extend field of the rm_detail_t.
+ *
+ * Results:
+ *	The variables at *startPtr and *endPtr are modified to hold the
+ *	offsets of the endpoints of the range given by index. If the specified
+ *	range doesn't exist then -1s are supplied.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+TclRegExpRangeUniChar(
+    Tcl_RegExp re,		/* Compiled regular expression that has been
+				 * passed to Tcl_RegExpExec. */
+    int index,			/* 0 means give the range of the entire match,
+				 * > 0 means give the range of a matching
+				 * subrange, -1 means the range of the
+				 * rm_extend field. */
+    int *startPtr,		/* Store address of first character in
+				 * (sub-)range here. */
+    int *endPtr)		/* Store address of character just after last
+				 * in (sub-)range here. */
+{
+    TclRegexp *regexpPtr = (TclRegexp *) re;
+
+    if ((regexpPtr->flags&REG_EXPECT) && index == -1) {
+	*startPtr = regexpPtr->details.rm_extend.rm_so;
+	*endPtr = regexpPtr->details.rm_extend.rm_eo;
+    } else if ((size_t) index > regexpPtr->re.re_nsub) {
+	*startPtr = -1;
+	*endPtr = -1;
+    } else {
+	*startPtr = regexpPtr->matches[index].rm_so;
+	*endPtr = regexpPtr->matches[index].rm_eo;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpMatch --
+ *
+ *	See if a string matches a regular expression.
+ *
+ * Results:
+ *	If an error occurs during the matching operation then -1 is returned
+ *	and the interp's result contains an error message. Otherwise the
+ *	return value is 1 if "text" matches "pattern" and 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegExpMatch(
+    Tcl_Interp *interp,		/* Used for error reporting. May be NULL. */
+    const char *text,		/* Text to search for pattern matches. */
+    const char *pattern)	/* Regular expression to match against text. */
+{
+    Tcl_RegExp re = Tcl_RegExpCompile(interp, pattern);
+
+    if (re == NULL) {
+	return -1;
+    }
+    return Tcl_RegExpExec(interp, re, text, text);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpExecObj --
+ *
+ *	Execute a precompiled regexp against the given object.
+ *
+ * Results:
+ *	If an error occurs during the matching operation then -1 is returned
+ *	and the interp's result contains an error message. Otherwise the
+ *	return value is 1 if "string" matches "pattern" and 0 otherwise.
+ *
+ * Side effects:
+ *	Converts the object to a Unicode object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegExpExecObj(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. */
+    Tcl_RegExp re,		/* Compiled regular expression; must have been
+				 * returned by previous call to
+				 * Tcl_GetRegExpFromObj. */
+    Tcl_Obj *textObj,		/* Text against which to match re. */
+    int offset,			/* Character index that marks where matching
+				 * should begin. */
+    int nmatches,		/* How many subexpression matches (counting
+				 * the whole match as subexpression 0) are of
+				 * interest. -1 means all of them. */
+    int flags)			/* Regular expression execution flags. */
+{
+    TclRegexp *regexpPtr = (TclRegexp *) re;
+    Tcl_UniChar *udata;
+    int length;
+    int reflags = regexpPtr->flags;
+#define TCL_REG_GLOBOK_FLAGS \
+	(TCL_REG_ADVANCED | TCL_REG_NOSUB | TCL_REG_NOCASE)
+
+    /*
+     * Take advantage of the equivalent glob pattern, if one exists.
+     * This is possible based only on the right mix of incoming flags (0)
+     * and regexp compile flags.
+     */
+    if ((offset == 0) && (nmatches == 0) && (flags == 0)
+	    && !(reflags & ~TCL_REG_GLOBOK_FLAGS)
+	    && (regexpPtr->globObjPtr != NULL)) {
+	int nocase = (reflags & TCL_REG_NOCASE) ? TCL_MATCH_NOCASE : 0;
+
+	/*
+	 * Pass to TclStringMatchObj for obj-specific handling.
+	 * XXX: Currently doesn't take advantage of exact-ness that
+	 * XXX: TclReToGlob tells us about
+	 */
+
+	return TclStringMatchObj(textObj, regexpPtr->globObjPtr, nocase);
+    }
+
+    /*
+     * Save the target object so we can extract strings from it later.
+     */
+
+    regexpPtr->string = NULL;
+    regexpPtr->objPtr = textObj;
+
+    udata = Tcl_GetUnicodeFromObj(textObj, &length);
+
+    if (offset > length) {
+	offset = length;
+    }
+    udata += offset;
+    length -= offset;
+
+    return RegExpExecUniChar(interp, re, udata, length, nmatches, flags);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpMatchObj --
+ *
+ *	See if an object matches a regular expression.
+ *
+ * Results:
+ *	If an error occurs during the matching operation then -1 is returned
+ *	and the interp's result contains an error message. Otherwise the
+ *	return value is 1 if "text" matches "pattern" and 0 otherwise.
+ *
+ * Side effects:
+ *	Changes the internal rep of the pattern and string objects.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RegExpMatchObj(
+    Tcl_Interp *interp,		/* Used for error reporting. May be NULL. */
+    Tcl_Obj *textObj,		/* Object containing the String to search. */
+    Tcl_Obj *patternObj)	/* Regular expression to match against
+				 * string. */
+{
+    Tcl_RegExp re;
+
+    /*
+     * For performance reasons, first try compiling the RE without support for
+     * subexpressions. On failure, try again without TCL_REG_NOSUB in case the
+     * RE has backreferences in it. Closely related to [Bug 1366683]. If this
+     * still fails, an error message will be left in the interpreter.
+     */
+
+    if (!(re = Tcl_GetRegExpFromObj(interp, patternObj,
+	    TCL_REG_ADVANCED | TCL_REG_NOSUB))
+     && !(re = Tcl_GetRegExpFromObj(interp, patternObj, TCL_REG_ADVANCED))) {
+	return -1;
+    }
+    return Tcl_RegExpExecObj(interp, re, textObj, 0 /* offset */,
+	    0 /* nmatches */, 0 /* flags */);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RegExpGetInfo --
+ *
+ *	Retrieve information about the current match.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_RegExpGetInfo(
+    Tcl_RegExp regexp,		/* Pattern from which to get subexpressions. */
+    Tcl_RegExpInfo *infoPtr)	/* Match information is stored here. */
+{
+    TclRegexp *regexpPtr = (TclRegexp *) regexp;
+
+    infoPtr->nsubs = regexpPtr->re.re_nsub;
+    infoPtr->matches = (Tcl_RegExpIndices *) regexpPtr->matches;
+    infoPtr->extendStart = regexpPtr->details.rm_extend.rm_so;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetRegExpFromObj --
+ *
+ *	Compile a regular expression into a form suitable for fast matching.
+ *	This function caches the result in a Tcl_Obj.
+ *
+ * Results:
+ *	The return value is a pointer to the compiled form of string, suitable
+ *	for passing to Tcl_RegExpExec. If an error occurred while compiling
+ *	the pattern, then NULL is returned and an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	Updates the native rep of the Tcl_Obj.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_RegExp
+Tcl_GetRegExpFromObj(
+    Tcl_Interp *interp,		/* For use in error reporting, and to access
+				 * the interp regexp cache. */
+    Tcl_Obj *objPtr,		/* Object whose string rep contains regular
+				 * expression pattern. Internal rep will be
+				 * changed to compiled form of this regular
+				 * expression. */
+    int flags)			/* Regular expression compilation flags. */
+{
+    int length;
+    TclRegexp *regexpPtr;
+    const char *pattern;
+
+    /*
+     * This is OK because we only actually interpret this value properly as a
+     * TclRegexp* when the type is tclRegexpType.
+     */
+
+    regexpPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    if ((objPtr->typePtr != &tclRegexpType) || (regexpPtr->flags != flags)) {
+	pattern = TclGetStringFromObj(objPtr, &length);
+
+	regexpPtr = CompileRegexp(interp, pattern, length, flags);
+	if (regexpPtr == NULL) {
+	    return NULL;
+	}
+
+	/*
+	 * Add a reference to the regexp so it will persist even if it is
+	 * pushed out of the current thread's regexp cache. This reference
+	 * will be removed when the object's internal rep is freed.
+	 */
+
+	regexpPtr->refCount++;
+
+	/*
+	 * Free the old representation and set our type.
+	 */
+
+	TclFreeIntRep(objPtr);
+	objPtr->internalRep.twoPtrValue.ptr1 = regexpPtr;
+	objPtr->typePtr = &tclRegexpType;
+    }
+    return (Tcl_RegExp) regexpPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRegAbout --
+ *
+ *	Return information about a compiled regular expression.
+ *
+ * Results:
+ *	The return value is -1 for failure, 0 for success, although at the
+ *	moment there's nothing that could fail. On success, a list is left in
+ *	the interp's result: first element is the subexpression count, second
+ *	is a list of re_info bit names.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclRegAbout(
+    Tcl_Interp *interp,		/* For use in variable assignment. */
+    Tcl_RegExp re)		/* The compiled regular expression. */
+{
+    TclRegexp *regexpPtr = (TclRegexp *) re;
+    struct infoname {
+	int bit;
+	const char *text;
+    };
+    static const struct infoname infonames[] = {
+	{REG_UBACKREF,		"REG_UBACKREF"},
+	{REG_ULOOKAHEAD,	"REG_ULOOKAHEAD"},
+	{REG_UBOUNDS,		"REG_UBOUNDS"},
+	{REG_UBRACES,		"REG_UBRACES"},
+	{REG_UBSALNUM,		"REG_UBSALNUM"},
+	{REG_UPBOTCH,		"REG_UPBOTCH"},
+	{REG_UBBS,		"REG_UBBS"},
+	{REG_UNONPOSIX,		"REG_UNONPOSIX"},
+	{REG_UUNSPEC,		"REG_UUNSPEC"},
+	{REG_UUNPORT,		"REG_UUNPORT"},
+	{REG_ULOCALE,		"REG_ULOCALE"},
+	{REG_UEMPTYMATCH,	"REG_UEMPTYMATCH"},
+	{REG_UIMPOSSIBLE,	"REG_UIMPOSSIBLE"},
+	{REG_USHORTEST,		"REG_USHORTEST"},
+	{0,			NULL}
+    };
+    const struct infoname *inf;
+    Tcl_Obj *infoObj, *resultObj;
+
+    /*
+     * The reset here guarantees that the interpreter result is empty and
+     * unshared. This means that we can use Tcl_ListObjAppendElement on the
+     * result object quite safely.
+     */
+
+    Tcl_ResetResult(interp);
+
+    /*
+     * Assume that there will never be more than INT_MAX subexpressions. This
+     * is a pretty reasonable assumption; the RE engine doesn't scale _that_
+     * well and Tcl has other limits that constrain things as well...
+     */
+
+    resultObj = Tcl_NewObj();
+    Tcl_ListObjAppendElement(NULL, resultObj,
+	    Tcl_NewWideIntObj((Tcl_WideInt) regexpPtr->re.re_nsub));
+
+    /*
+     * Now append a list of all the bit-flags set for the RE.
+     */
+
+    TclNewObj(infoObj);
+    for (inf=infonames ; inf->bit != 0 ; inf++) {
+	if (regexpPtr->re.re_info & inf->bit) {
+	    Tcl_ListObjAppendElement(NULL, infoObj,
+		    Tcl_NewStringObj(inf->text, -1));
+	}
+    }
+    Tcl_ListObjAppendElement(NULL, resultObj, infoObj);
+    Tcl_SetObjResult(interp, resultObj);
+
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRegError --
+ *
+ *	Generate an error message based on the regexp status code.
+ *
+ * Results:
+ *	Places an error in the interpreter.
+ *
+ * Side effects:
+ *	Sets errorCode as well.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclRegError(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. */
+    const char *msg,		/* Message to prepend to error. */
+    int status)			/* Status code to report. */
+{
+    char buf[100];		/* ample in practice */
+    char cbuf[TCL_INTEGER_SPACE];
+    size_t n;
+    const char *p;
+
+    Tcl_ResetResult(interp);
+    n = TclReError(status, NULL, buf, sizeof(buf));
+    p = (n > sizeof(buf)) ? "..." : "";
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf("%s%s%s", msg, buf, p));
+
+    sprintf(cbuf, "%d", status);
+    (void) TclReError(REG_ITOA, NULL, cbuf, sizeof(cbuf));
+    Tcl_SetErrorCode(interp, "REGEXP", cbuf, buf, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeRegexpInternalRep --
+ *
+ *	Deallocate the storage associated with a regexp object's internal
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees the compiled regular expression.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeRegexpInternalRep(
+    Tcl_Obj *objPtr)		/* Regexp object with internal rep to free. */
+{
+    TclRegexp *regexpRepPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    /*
+     * If this is the last reference to the regexp, free it.
+     */
+
+    if (regexpRepPtr->refCount-- <= 1) {
+	FreeRegexp(regexpRepPtr);
+    }
+    objPtr->typePtr = NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupRegexpInternalRep --
+ *
+ *	We copy the reference to the compiled regexp and bump its reference
+ *	count.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Increments the reference count of the regexp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupRegexpInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. */
+    Tcl_Obj *copyPtr)		/* Object with internal rep to set. */
+{
+    TclRegexp *regexpPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    regexpPtr->refCount++;
+    copyPtr->internalRep.twoPtrValue.ptr1 = srcPtr->internalRep.twoPtrValue.ptr1;
+    copyPtr->typePtr = &tclRegexpType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetRegexpFromAny --
+ *
+ *	Attempt to generate a compiled regular expression for the Tcl object
+ *	"objPtr".
+ *
+ * Results:
+ *	The return value is TCL_OK or TCL_ERROR. If an error occurs during
+ *	conversion, an error message is left in the interpreter's result
+ *	unless "interp" is NULL.
+ *
+ * Side effects:
+ *	If no error occurs, a regular expression is stored as "objPtr"s
+ *	internal representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetRegexpFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr)		/* The object to convert. */
+{
+    if (Tcl_GetRegExpFromObj(interp, objPtr, REG_ADVANCED) == NULL) {
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * CompileRegexp --
+ *
+ *	Attempt to compile the given regexp pattern. If the compiled regular
+ *	expression can be found in the per-thread cache, it will be used
+ *	instead of compiling a new copy.
+ *
+ * Results:
+ *	The return value is a pointer to a newly allocated TclRegexp that
+ *	represents the compiled pattern, or NULL if the pattern could not be
+ *	compiled. If NULL is returned, an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	The thread-local regexp cache is updated and a new TclRegexp may be
+ *	allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static TclRegexp *
+CompileRegexp(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    const char *string,		/* The regexp to compile (UTF-8). */
+    int length,			/* The length of the string in bytes. */
+    int flags)			/* Compilation flags. */
+{
+    TclRegexp *regexpPtr;
+    const Tcl_UniChar *uniString;
+    int numChars, status, i, exact;
+    Tcl_DString stringBuf;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (!tsdPtr->initialized) {
+	tsdPtr->initialized = 1;
+	Tcl_CreateThreadExitHandler(FinalizeRegexp, NULL);
+    }
+
+    /*
+     * This routine maintains a second-level regular expression cache in
+     * addition to the per-object regexp cache. The per-thread cache is needed
+     * to handle the case where for various reasons the object is lost between
+     * invocations of the regexp command, but the literal pattern is the same.
+     */
+
+    /*
+     * Check the per-thread compiled regexp cache. We can only reuse a regexp
+     * if it has the same pattern and the same flags.
+     */
+
+    for (i = 0; (i < NUM_REGEXPS) && (tsdPtr->patterns[i] != NULL); i++) {
+	if ((length == tsdPtr->patLengths[i])
+		&& (tsdPtr->regexps[i]->flags == flags)
+		&& (strcmp(string, tsdPtr->patterns[i]) == 0)) {
+	    /*
+	     * Move the matched pattern to the first slot in the cache and
+	     * shift the other patterns down one position.
+	     */
+
+	    if (i != 0) {
+		int j;
+		char *cachedString;
+
+		cachedString = tsdPtr->patterns[i];
+		regexpPtr = tsdPtr->regexps[i];
+		for (j = i-1; j >= 0; j--) {
+		    tsdPtr->patterns[j+1] = tsdPtr->patterns[j];
+		    tsdPtr->patLengths[j+1] = tsdPtr->patLengths[j];
+		    tsdPtr->regexps[j+1] = tsdPtr->regexps[j];
+		}
+		tsdPtr->patterns[0] = cachedString;
+		tsdPtr->patLengths[0] = length;
+		tsdPtr->regexps[0] = regexpPtr;
+	    }
+	    return tsdPtr->regexps[0];
+	}
+    }
+
+    /*
+     * This is a new expression, so compile it and add it to the cache.
+     */
+
+    regexpPtr = ckalloc(sizeof(TclRegexp));
+    regexpPtr->objPtr = NULL;
+    regexpPtr->string = NULL;
+    regexpPtr->details.rm_extend.rm_so = -1;
+    regexpPtr->details.rm_extend.rm_eo = -1;
+
+    /*
+     * Get the up-to-date string representation and map to unicode.
+     */
+
+    Tcl_DStringInit(&stringBuf);
+    uniString = Tcl_UtfToUniCharDString(string, length, &stringBuf);
+    numChars = Tcl_DStringLength(&stringBuf) / sizeof(Tcl_UniChar);
+
+    /*
+     * Compile the string and check for errors.
+     */
+
+    regexpPtr->flags = flags;
+    status = TclReComp(&regexpPtr->re, uniString, (size_t) numChars, flags);
+    Tcl_DStringFree(&stringBuf);
+
+    if (status != REG_OKAY) {
+	/*
+	 * Clean up and report errors in the interpreter, if possible.
+	 */
+
+	ckfree(regexpPtr);
+	if (interp) {
+	    TclRegError(interp,
+		    "couldn't compile regular expression pattern: ", status);
+	}
+	return NULL;
+    }
+
+    /*
+     * Convert RE to a glob pattern equivalent, if any, and cache it.  If this
+     * is not possible, then globObjPtr will be NULL.  This is used by
+     * Tcl_RegExpExecObj to optionally do a fast match (avoids RE engine).
+     */
+
+    if (TclReToGlob(NULL, string, length, &stringBuf, &exact,
+	    NULL) == TCL_OK) {
+	regexpPtr->globObjPtr = TclDStringToObj(&stringBuf);
+	Tcl_IncrRefCount(regexpPtr->globObjPtr);
+    } else {
+	regexpPtr->globObjPtr = NULL;
+    }
+
+    /*
+     * Allocate enough space for all of the subexpressions, plus one extra for
+     * the entire pattern.
+     */
+
+    regexpPtr->matches =
+	    ckalloc(sizeof(regmatch_t) * (regexpPtr->re.re_nsub + 1));
+
+    /*
+     * Initialize the refcount to one initially, since it is in the cache.
+     */
+
+    regexpPtr->refCount = 1;
+
+    /*
+     * Free the last regexp, if necessary, and make room at the head of the
+     * list for the new regexp.
+     */
+
+    if (tsdPtr->patterns[NUM_REGEXPS-1] != NULL) {
+	TclRegexp *oldRegexpPtr = tsdPtr->regexps[NUM_REGEXPS-1];
+
+	if (oldRegexpPtr->refCount-- <= 1) {
+	    FreeRegexp(oldRegexpPtr);
+	}
+	ckfree(tsdPtr->patterns[NUM_REGEXPS-1]);
+    }
+    for (i = NUM_REGEXPS - 2; i >= 0; i--) {
+	tsdPtr->patterns[i+1] = tsdPtr->patterns[i];
+	tsdPtr->patLengths[i+1] = tsdPtr->patLengths[i];
+	tsdPtr->regexps[i+1] = tsdPtr->regexps[i];
+    }
+    tsdPtr->patterns[0] = ckalloc(length + 1);
+    memcpy(tsdPtr->patterns[0], string, (unsigned) length + 1);
+    tsdPtr->patLengths[0] = length;
+    tsdPtr->regexps[0] = regexpPtr;
+
+    return regexpPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeRegexp --
+ *
+ *	Release the storage associated with a TclRegexp.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeRegexp(
+    TclRegexp *regexpPtr)	/* Compiled regular expression to free. */
+{
+    TclReFree(&regexpPtr->re);
+    if (regexpPtr->globObjPtr) {
+	TclDecrRefCount(regexpPtr->globObjPtr);
+    }
+    if (regexpPtr->matches) {
+	ckfree(regexpPtr->matches);
+    }
+    ckfree(regexpPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FinalizeRegexp --
+ *
+ *	Release the storage associated with the per-thread regexp cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FinalizeRegexp(
+    ClientData clientData)	/* Not used. */
+{
+    int i;
+    TclRegexp *regexpPtr;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    for (i = 0; (i < NUM_REGEXPS) && (tsdPtr->patterns[i] != NULL); i++) {
+	regexpPtr = tsdPtr->regexps[i];
+	if (regexpPtr->refCount-- <= 1) {
+	    FreeRegexp(regexpPtr);
+	}
+	ckfree(tsdPtr->patterns[i]);
+	tsdPtr->patterns[i] = NULL;
+    }
+
+    /*
+     * We may find ourselves reinitialized if another finalization routine
+     * invokes regexps.
+     */
+
+    tsdPtr->initialized = 0;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclRegexp.h b/generic/tclRegexp.h
new file mode 100644
index 0000000..eac0aaa
--- /dev/null
+++ b/generic/tclRegexp.h
@@ -0,0 +1,52 @@
+/*
+ * tclRegexp.h --
+ *
+ *	This file contains definitions used internally by Henry Spencer's
+ *	regular expression code.
+ *
+ * Copyright (c) 1998 by Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLREGEXP
+#define _TCLREGEXP
+
+#include "regex.h"
+
+/*
+ * The TclRegexp structure encapsulates a compiled regex_t, the flags that
+ * were used to compile it, and an array of pointers that are used to indicate
+ * subexpressions after a call to Tcl_RegExpExec. Note that the string and
+ * objPtr are mutually exclusive. These values are needed by Tcl_RegExpRange
+ * in order to return pointers into the original string.
+ */
+
+typedef struct TclRegexp {
+    int flags;			/* Regexp compile flags. */
+    regex_t re;			/* Compiled re, includes number of
+				 * subexpressions. */
+    const char *string;		/* Last string passed to Tcl_RegExpExec. */
+    Tcl_Obj *objPtr;		/* Last object passed to Tcl_RegExpExecObj. */
+    Tcl_Obj *globObjPtr;	/* Glob pattern rep of RE or NULL if none. */
+    regmatch_t *matches;	/* Array of indices into the Tcl_UniChar
+				 * representation of the last string matched
+				 * with this regexp to indicate the location
+				 * of subexpressions. */
+    rm_detail_t details;	/* Detailed information on match (currently
+				 * used only for REG_EXPECT). */
+    unsigned int refCount;	/* Count of number of references to this
+				 * compiled regexp. */
+} TclRegexp;
+
+#endif /* _TCLREGEXP */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclResolve.c b/generic/tclResolve.c
new file mode 100644
index 0000000..974737e
--- /dev/null
+++ b/generic/tclResolve.c
@@ -0,0 +1,424 @@
+/*
+ * tclResolve.c --
+ *
+ *	Contains hooks for customized command/variable name resolution
+ *	schemes. These hooks allow extensions like [incr Tcl] to add their own
+ *	name resolution rules to the Tcl language. Rules can be applied to a
+ *	particular namespace, to the interpreter as a whole, or both.
+ *
+ * Copyright (c) 1998 Lucent Technologies, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Declarations for functions local to this file:
+ */
+
+static void		BumpCmdRefEpochs(Namespace *nsPtr);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AddInterpResolvers --
+ *
+ *	Adds a set of command/variable resolution functions to an interpreter.
+ *	These functions are consulted when commands are resolved in
+ *	Tcl_FindCommand, and when variables are resolved in TclLookupVar and
+ *	LookupCompiledLocal. Each namespace may also have its own set of
+ *	resolution functions which take precedence over those for the
+ *	interpreter.
+ *
+ *	When a name is resolved, it is handled as follows. First, the name is
+ *	passed to the resolution functions for the namespace. If not resolved,
+ *	the name is passed to each of the resolution functions added to the
+ *	interpreter. Finally, if still not resolved, the name is handled using
+ *	the default Tcl rules for name resolution.
+ *
+ * Results:
+ *	Returns pointers to the current name resolution functions in the
+ *	cmdProcPtr, varProcPtr and compiledVarProcPtr arguments.
+ *
+ * Side effects:
+ *	If a compiledVarProc is specified, this function bumps the
+ *	compileEpoch for the interpreter, forcing all code to be recompiled.
+ *	If a cmdProc is specified, this function bumps the cmdRefEpoch in all
+ *	namespaces, forcing commands to be resolved again using the new rules.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AddInterpResolvers(
+    Tcl_Interp *interp,		/* Interpreter whose name resolution rules are
+				 * being modified. */
+    const char *name,		/* Name of this resolution scheme. */
+    Tcl_ResolveCmdProc *cmdProc,/* New function for command resolution. */
+    Tcl_ResolveVarProc *varProc,/* Function for variable resolution at
+				 * runtime. */
+    Tcl_ResolveCompiledVarProc *compiledVarProc)
+				/* Function for variable resolution at compile
+				 * time. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ResolverScheme *resPtr;
+    unsigned len;
+
+    /*
+     * Since we're adding a new name resolution scheme, we must force all code
+     * to be recompiled to use the new scheme. If there are new compiled
+     * variable resolution rules, bump the compiler epoch to invalidate
+     * compiled code. If there are new command resolution rules, bump the
+     * cmdRefEpoch in all namespaces.
+     */
+
+    if (compiledVarProc) {
+	iPtr->compileEpoch++;
+    }
+    if (cmdProc) {
+	BumpCmdRefEpochs(iPtr->globalNsPtr);
+    }
+
+    /*
+     * Look for an existing scheme with the given name. If found, then replace
+     * its rules.
+     */
+
+    for (resPtr=iPtr->resolverPtr ; resPtr!=NULL ; resPtr=resPtr->nextPtr) {
+	if (*name == *resPtr->name && strcmp(name, resPtr->name) == 0) {
+	    resPtr->cmdResProc = cmdProc;
+	    resPtr->varResProc = varProc;
+	    resPtr->compiledVarResProc = compiledVarProc;
+	    return;
+	}
+    }
+
+    /*
+     * Otherwise, this is a new scheme. Add it to the FRONT of the linked
+     * list, so that it overrides existing schemes.
+     */
+
+    resPtr = ckalloc(sizeof(ResolverScheme));
+    len = strlen(name) + 1;
+    resPtr->name = ckalloc(len);
+    memcpy(resPtr->name, name, len);
+    resPtr->cmdResProc = cmdProc;
+    resPtr->varResProc = varProc;
+    resPtr->compiledVarResProc = compiledVarProc;
+    resPtr->nextPtr = iPtr->resolverPtr;
+    iPtr->resolverPtr = resPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetInterpResolvers --
+ *
+ *	Looks for a set of command/variable resolution functions with the
+ *	given name in an interpreter. These functions are registered by
+ *	calling Tcl_AddInterpResolvers.
+ *
+ * Results:
+ *	If the name is recognized, this function returns non-zero, along with
+ *	pointers to the name resolution functions in the Tcl_ResolverInfo
+ *	structure. If the name is not recognized, this function returns zero.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetInterpResolvers(
+    Tcl_Interp *interp,		/* Interpreter whose name resolution rules are
+				 * being queried. */
+    const char *name,		/* Look for a scheme with this name. */
+    Tcl_ResolverInfo *resInfoPtr)
+				/* Returns pointers to the functions, if
+				 * found */
+{
+    Interp *iPtr = (Interp *) interp;
+    ResolverScheme *resPtr;
+
+    /*
+     * Look for an existing scheme with the given name. If found, then return
+     * pointers to its functions.
+     */
+
+    for (resPtr=iPtr->resolverPtr ; resPtr!=NULL ; resPtr=resPtr->nextPtr) {
+	if (*name == *resPtr->name && strcmp(name, resPtr->name) == 0) {
+	    resInfoPtr->cmdResProc = resPtr->cmdResProc;
+	    resInfoPtr->varResProc = resPtr->varResProc;
+	    resInfoPtr->compiledVarResProc = resPtr->compiledVarResProc;
+	    return 1;
+	}
+    }
+
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RemoveInterpResolvers --
+ *
+ *	Removes a set of command/variable resolution functions previously
+ *	added by Tcl_AddInterpResolvers. The next time a command/variable name
+ *	is resolved, these functions won't be consulted.
+ *
+ * Results:
+ *	Returns non-zero if the name was recognized and the resolution scheme
+ *	was deleted. Returns zero otherwise.
+ *
+ * Side effects:
+ *	If a scheme with a compiledVarProc was deleted, this function bumps
+ *	the compileEpoch for the interpreter, forcing all code to be
+ *	recompiled. If a scheme with a cmdProc was deleted, this function
+ *	bumps the cmdRefEpoch in all namespaces, forcing commands to be
+ *	resolved again using the new rules.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RemoveInterpResolvers(
+    Tcl_Interp *interp,		/* Interpreter whose name resolution rules are
+				 * being modified. */
+    const char *name)		/* Name of the scheme to be removed. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ResolverScheme **prevPtrPtr, *resPtr;
+
+    /*
+     * Look for an existing scheme with the given name.
+     */
+
+    prevPtrPtr = &iPtr->resolverPtr;
+    for (resPtr=iPtr->resolverPtr ; resPtr!=NULL ; resPtr=resPtr->nextPtr) {
+	if (*name == *resPtr->name && strcmp(name, resPtr->name) == 0) {
+	    break;
+	}
+	prevPtrPtr = &resPtr->nextPtr;
+    }
+
+    /*
+     * If we found the scheme, delete it.
+     */
+
+    if (resPtr) {
+	/*
+	 * If we're deleting a scheme with compiled variable resolution rules,
+	 * bump the compiler epoch to invalidate compiled code. If we're
+	 * deleting a scheme with command resolution rules, bump the
+	 * cmdRefEpoch in all namespaces.
+	 */
+
+	if (resPtr->compiledVarResProc) {
+	    iPtr->compileEpoch++;
+	}
+	if (resPtr->cmdResProc) {
+	    BumpCmdRefEpochs(iPtr->globalNsPtr);
+	}
+
+	*prevPtrPtr = resPtr->nextPtr;
+	ckfree(resPtr->name);
+	ckfree(resPtr);
+
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BumpCmdRefEpochs --
+ *
+ *	This function is used to bump the cmdRefEpoch counters in the
+ *	specified namespace and all of its child namespaces. It is used
+ *	whenever name resolution schemes are added/removed from an
+ *	interpreter, to invalidate all command references.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Bumps the cmdRefEpoch in the specified namespace and its children,
+ *	recursively.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+BumpCmdRefEpochs(
+    Namespace *nsPtr)		/* Namespace being modified. */
+{
+    Tcl_HashEntry *entry;
+    Tcl_HashSearch search;
+
+    nsPtr->cmdRefEpoch++;
+
+#ifndef BREAK_NAMESPACE_COMPAT
+    for (entry = Tcl_FirstHashEntry(&nsPtr->childTable, &search);
+	    entry != NULL; entry = Tcl_NextHashEntry(&search)) {
+	Namespace *childNsPtr = Tcl_GetHashValue(entry);
+
+	BumpCmdRefEpochs(childNsPtr);
+    }
+#else
+    if (nsPtr->childTablePtr != NULL) {
+	for (entry = Tcl_FirstHashEntry(nsPtr->childTablePtr, &search);
+		entry != NULL; entry = Tcl_NextHashEntry(&search)) {
+	    Namespace *childNsPtr = Tcl_GetHashValue(entry);
+
+	    BumpCmdRefEpochs(childNsPtr);
+	}
+    }
+#endif
+    TclInvalidateNsPath(nsPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetNamespaceResolvers --
+ *
+ *	Sets the command/variable resolution functions for a namespace,
+ *	thereby changing the way that command/variable names are interpreted.
+ *	This allows extension writers to support different name resolution
+ *	schemes, such as those for object-oriented packages.
+ *
+ *	Command resolution is handled by a function of the following type:
+ *
+ *	  typedef int (Tcl_ResolveCmdProc)(Tcl_Interp *interp,
+ *		  const char *name, Tcl_Namespace *context,
+ *		  int flags, Tcl_Command *rPtr);
+ *
+ *	Whenever a command is executed or Tcl_FindCommand is invoked within
+ *	the namespace, this function is called to resolve the command name. If
+ *	this function is able to resolve the name, it should return the status
+ *	code TCL_OK, along with the corresponding Tcl_Command in the rPtr
+ *	argument. Otherwise, the function can return TCL_CONTINUE, and the
+ *	command will be treated under the usual name resolution rules. Or, it
+ *	can return TCL_ERROR, and the command will be considered invalid.
+ *
+ *	Variable resolution is handled by two functions. The first is called
+ *	whenever a variable needs to be resolved at compile time:
+ *
+ *	  typedef int (Tcl_ResolveCompiledVarProc)(Tcl_Interp *interp,
+ *		  const char *name, Tcl_Namespace *context,
+ *		  Tcl_ResolvedVarInfo *rPtr);
+ *
+ *	If this function is able to resolve the name, it should return the
+ *	status code TCL_OK, along with variable resolution info in the rPtr
+ *	argument; this info will be used to set up compiled locals in the call
+ *	frame at runtime. The function may also return TCL_CONTINUE, and the
+ *	variable will be treated under the usual name resolution rules. Or, it
+ *	can return TCL_ERROR, and the variable will be considered invalid.
+ *
+ *	Another function is used whenever a variable needs to be resolved at
+ *	runtime but it is not recognized as a compiled local. (For example,
+ *	the variable may be requested via Tcl_FindNamespaceVar.) This function
+ *	has the following type:
+ *
+ *	  typedef int (Tcl_ResolveVarProc)(Tcl_Interp *interp,
+ *		  const char *name, Tcl_Namespace *context,
+ *		  int flags, Tcl_Var *rPtr);
+ *
+ *	This function is quite similar to the compile-time version. It returns
+ *	the same status codes, but if variable resolution succeeds, this
+ *	function returns a Tcl_Var directly via the rPtr argument.
+ *
+ * Results:
+ *	Nothing.
+ *
+ * Side effects:
+ *	Bumps the command epoch counter for the namespace, invalidating all
+ *	command references in that namespace. Also bumps the resolver epoch
+ *	counter for the namespace, forcing all code in the namespace to be
+ *	recompiled.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetNamespaceResolvers(
+    Tcl_Namespace *namespacePtr,/* Namespace whose resolution rules are being
+				 * modified. */
+    Tcl_ResolveCmdProc *cmdProc,/* Function for command resolution */
+    Tcl_ResolveVarProc *varProc,/* Function for variable resolution at
+				 * run-time */
+    Tcl_ResolveCompiledVarProc *compiledVarProc)
+				/* Function for variable resolution at compile
+				 * time. */
+{
+    Namespace *nsPtr = (Namespace *) namespacePtr;
+
+    /*
+     * Plug in the new command resolver, and bump the epoch counters so that
+     * all code will have to be recompiled and all commands will have to be
+     * resolved again using the new policy.
+     */
+
+    nsPtr->cmdResProc = cmdProc;
+    nsPtr->varResProc = varProc;
+    nsPtr->compiledVarResProc = compiledVarProc;
+
+    nsPtr->cmdRefEpoch++;
+    nsPtr->resolverEpoch++;
+    TclInvalidateNsPath(nsPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetNamespaceResolvers --
+ *
+ *	Returns the current command/variable resolution functions for a
+ *	namespace. By default, these functions are NULL. New functions can be
+ *	installed by calling Tcl_SetNamespaceResolvers, to provide new name
+ *	resolution rules.
+ *
+ * Results:
+ *	Returns non-zero if any name resolution functions have been assigned
+ *	to this namespace; also returns pointers to the functions in the
+ *	Tcl_ResolverInfo structure. Returns zero otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetNamespaceResolvers(
+    Tcl_Namespace *namespacePtr,/* Namespace whose resolution rules are being
+				 * modified. */
+    Tcl_ResolverInfo *resInfoPtr)
+				/* Returns: pointers for all name resolution
+				 * functions assigned to this namespace. */
+{
+    Namespace *nsPtr = (Namespace *) namespacePtr;
+
+    resInfoPtr->cmdResProc = nsPtr->cmdResProc;
+    resInfoPtr->varResProc = nsPtr->varResProc;
+    resInfoPtr->compiledVarResProc = nsPtr->compiledVarResProc;
+
+    if (nsPtr->cmdResProc != NULL || nsPtr->varResProc != NULL ||
+	    nsPtr->compiledVarResProc != NULL) {
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclResult.c b/generic/tclResult.c
new file mode 100644
index 0000000..57a6de5
--- /dev/null
+++ b/generic/tclResult.c
@@ -0,0 +1,1784 @@
+/*
+ * tclResult.c --
+ *
+ *	This file contains code to manage the interpreter result.
+ *
+ * Copyright (c) 1997 by Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Indices of the standard return options dictionary keys.
+ */
+
+enum returnKeys {
+    KEY_CODE,	KEY_ERRORCODE,	KEY_ERRORINFO,	KEY_ERRORLINE,
+    KEY_LEVEL,	KEY_OPTIONS,	KEY_ERRORSTACK,	KEY_LAST
+};
+
+/*
+ * Function prototypes for local functions in this file:
+ */
+
+static Tcl_Obj **	GetKeys(void);
+static void		ReleaseKeys(ClientData clientData);
+static void		ResetObjResult(Interp *iPtr);
+#ifndef TCL_NO_DEPRECATED
+static void		SetupAppendBuffer(Interp *iPtr, int newSpace);
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * This structure is used to take a snapshot of the interpreter state in
+ * Tcl_SaveInterpState. You can snapshot the state, execute a command, and
+ * then back up to the result or the error that was previously in progress.
+ */
+
+typedef struct {
+    int status;			/* return code status */
+    int flags;			/* Each remaining field saves the */
+    int returnLevel;		/* corresponding field of the Interp */
+    int returnCode;		/* struct. These fields taken together are */
+    Tcl_Obj *errorInfo;		/* the "state" of the interp. */
+    Tcl_Obj *errorCode;
+    Tcl_Obj *returnOpts;
+    Tcl_Obj *objResult;
+    Tcl_Obj *errorStack;
+    int resetErrorStack;
+} InterpState;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SaveInterpState --
+ *
+ *	Fills a token with a snapshot of the current state of the interpreter.
+ *	The snapshot can be restored at any point by TclRestoreInterpState.
+ *
+ *	The token returned must be eventally passed to one of the routines
+ *	TclRestoreInterpState or TclDiscardInterpState, or there will be a
+ *	memory leak.
+ *
+ * Results:
+ *	Returns a token representing the interp state.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_InterpState
+Tcl_SaveInterpState(
+    Tcl_Interp *interp,		/* Interpreter's state to be saved */
+    int status)			/* status code for current operation */
+{
+    Interp *iPtr = (Interp *) interp;
+    InterpState *statePtr = ckalloc(sizeof(InterpState));
+
+    statePtr->status = status;
+    statePtr->flags = iPtr->flags & ERR_ALREADY_LOGGED;
+    statePtr->returnLevel = iPtr->returnLevel;
+    statePtr->returnCode = iPtr->returnCode;
+    statePtr->errorInfo = iPtr->errorInfo;
+    statePtr->errorStack = iPtr->errorStack;
+    statePtr->resetErrorStack = iPtr->resetErrorStack;
+    if (statePtr->errorInfo) {
+	Tcl_IncrRefCount(statePtr->errorInfo);
+    }
+    statePtr->errorCode = iPtr->errorCode;
+    if (statePtr->errorCode) {
+	Tcl_IncrRefCount(statePtr->errorCode);
+    }
+    statePtr->returnOpts = iPtr->returnOpts;
+    if (statePtr->returnOpts) {
+	Tcl_IncrRefCount(statePtr->returnOpts);
+    }
+    if (statePtr->errorStack) {
+	Tcl_IncrRefCount(statePtr->errorStack);
+    }
+    statePtr->objResult = Tcl_GetObjResult(interp);
+    Tcl_IncrRefCount(statePtr->objResult);
+    return (Tcl_InterpState) statePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RestoreInterpState --
+ *
+ *	Accepts an interp and a token previously returned by
+ *	Tcl_SaveInterpState. Restore the state of the interp to what it was at
+ *	the time of the Tcl_SaveInterpState call.
+ *
+ * Results:
+ *	Returns the status value originally passed in to Tcl_SaveInterpState.
+ *
+ * Side effects:
+ *	Restores the interp state and frees memory held by token.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_RestoreInterpState(
+    Tcl_Interp *interp,		/* Interpreter's state to be restored. */
+    Tcl_InterpState state)	/* Saved interpreter state. */
+{
+    Interp *iPtr = (Interp *) interp;
+    InterpState *statePtr = (InterpState *) state;
+    int status = statePtr->status;
+
+    iPtr->flags &= ~ERR_ALREADY_LOGGED;
+    iPtr->flags |= (statePtr->flags & ERR_ALREADY_LOGGED);
+
+    iPtr->returnLevel = statePtr->returnLevel;
+    iPtr->returnCode = statePtr->returnCode;
+    iPtr->resetErrorStack = statePtr->resetErrorStack;
+    if (iPtr->errorInfo) {
+	Tcl_DecrRefCount(iPtr->errorInfo);
+    }
+    iPtr->errorInfo = statePtr->errorInfo;
+    if (iPtr->errorInfo) {
+	Tcl_IncrRefCount(iPtr->errorInfo);
+    }
+    if (iPtr->errorCode) {
+	Tcl_DecrRefCount(iPtr->errorCode);
+    }
+    iPtr->errorCode = statePtr->errorCode;
+    if (iPtr->errorCode) {
+	Tcl_IncrRefCount(iPtr->errorCode);
+    }
+    if (iPtr->errorStack) {
+	Tcl_DecrRefCount(iPtr->errorStack);
+    }
+    iPtr->errorStack = statePtr->errorStack;
+    if (iPtr->errorStack) {
+	Tcl_IncrRefCount(iPtr->errorStack);
+    }
+    if (iPtr->returnOpts) {
+	Tcl_DecrRefCount(iPtr->returnOpts);
+    }
+    iPtr->returnOpts = statePtr->returnOpts;
+    if (iPtr->returnOpts) {
+	Tcl_IncrRefCount(iPtr->returnOpts);
+    }
+    Tcl_SetObjResult(interp, statePtr->objResult);
+    Tcl_DiscardInterpState(state);
+    return status;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DiscardInterpState --
+ *
+ *	Accepts a token previously returned by Tcl_SaveInterpState. Frees the
+ *	memory it uses.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DiscardInterpState(
+    Tcl_InterpState state)	/* saved interpreter state */
+{
+    InterpState *statePtr = (InterpState *) state;
+
+    if (statePtr->errorInfo) {
+	Tcl_DecrRefCount(statePtr->errorInfo);
+    }
+    if (statePtr->errorCode) {
+	Tcl_DecrRefCount(statePtr->errorCode);
+    }
+    if (statePtr->returnOpts) {
+	Tcl_DecrRefCount(statePtr->returnOpts);
+    }
+    if (statePtr->errorStack) {
+	Tcl_DecrRefCount(statePtr->errorStack);
+    }
+    Tcl_DecrRefCount(statePtr->objResult);
+    ckfree(statePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SaveResult --
+ *
+ *	Takes a snapshot of the current result state of the interpreter. The
+ *	snapshot can be restored at any point by Tcl_RestoreResult. Note that
+ *	this routine does not preserve the errorCode, errorInfo, or flags
+ *	fields so it should not be used if an error is in progress.
+ *
+ *	Once a snapshot is saved, it must be restored by calling
+ *	Tcl_RestoreResult, or discarded by calling Tcl_DiscardResult.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Resets the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_SaveResult
+void
+Tcl_SaveResult(
+    Tcl_Interp *interp,		/* Interpreter to save. */
+    Tcl_SavedResult *statePtr)	/* Pointer to state structure. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    /*
+     * Move the result object into the save state. Note that we don't need to
+     * change its refcount because we're moving it, not adding a new
+     * reference. Put an empty object into the interpreter.
+     */
+
+    statePtr->objResultPtr = iPtr->objResultPtr;
+    iPtr->objResultPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(iPtr->objResultPtr);
+
+    /*
+     * Save the string result.
+     */
+
+    statePtr->freeProc = iPtr->freeProc;
+    if (iPtr->result == iPtr->resultSpace) {
+	/*
+	 * Copy the static string data out of the interp buffer.
+	 */
+
+	statePtr->result = statePtr->resultSpace;
+	strcpy(statePtr->result, iPtr->result);
+	statePtr->appendResult = NULL;
+    } else if (iPtr->result == iPtr->appendResult) {
+	/*
+	 * Move the append buffer out of the interp.
+	 */
+
+	statePtr->appendResult = iPtr->appendResult;
+	statePtr->appendAvl = iPtr->appendAvl;
+	statePtr->appendUsed = iPtr->appendUsed;
+	statePtr->result = statePtr->appendResult;
+	iPtr->appendResult = NULL;
+	iPtr->appendAvl = 0;
+	iPtr->appendUsed = 0;
+    } else {
+	/*
+	 * Move the dynamic or static string out of the interpreter.
+	 */
+
+	statePtr->result = iPtr->result;
+	statePtr->appendResult = NULL;
+    }
+
+    iPtr->result = iPtr->resultSpace;
+    iPtr->resultSpace[0] = 0;
+    iPtr->freeProc = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_RestoreResult --
+ *
+ *	Restores the state of the interpreter to a snapshot taken by
+ *	Tcl_SaveResult. After this call, the token for the interpreter state
+ *	is no longer valid.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Restores the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_RestoreResult
+void
+Tcl_RestoreResult(
+    Tcl_Interp *interp,		/* Interpreter being restored. */
+    Tcl_SavedResult *statePtr)	/* State returned by Tcl_SaveResult. */
+{
+    Interp *iPtr = (Interp *) interp;
+
+    Tcl_ResetResult(interp);
+
+    /*
+     * Restore the string result.
+     */
+
+    iPtr->freeProc = statePtr->freeProc;
+    if (statePtr->result == statePtr->resultSpace) {
+	/*
+	 * Copy the static string data into the interp buffer.
+	 */
+
+	iPtr->result = iPtr->resultSpace;
+	strcpy(iPtr->result, statePtr->result);
+    } else if (statePtr->result == statePtr->appendResult) {
+	/*
+	 * Move the append buffer back into the interp.
+	 */
+
+	if (iPtr->appendResult != NULL) {
+	    ckfree(iPtr->appendResult);
+	}
+
+	iPtr->appendResult = statePtr->appendResult;
+	iPtr->appendAvl = statePtr->appendAvl;
+	iPtr->appendUsed = statePtr->appendUsed;
+	iPtr->result = iPtr->appendResult;
+    } else {
+	/*
+	 * Move the dynamic or static string back into the interpreter.
+	 */
+
+	iPtr->result = statePtr->result;
+    }
+
+    /*
+     * Restore the object result.
+     */
+
+    Tcl_DecrRefCount(iPtr->objResultPtr);
+    iPtr->objResultPtr = statePtr->objResultPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DiscardResult --
+ *
+ *	Frees the memory associated with an interpreter snapshot taken by
+ *	Tcl_SaveResult. If the snapshot is not restored, this function must be
+ *	called to discard it, or the memory will be lost.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_DiscardResult
+void
+Tcl_DiscardResult(
+    Tcl_SavedResult *statePtr)	/* State returned by Tcl_SaveResult. */
+{
+    TclDecrRefCount(statePtr->objResultPtr);
+
+    if (statePtr->result == statePtr->appendResult) {
+	ckfree(statePtr->appendResult);
+    } else if (statePtr->freeProc == TCL_DYNAMIC) {
+        ckfree(statePtr->result);
+    } else if (statePtr->freeProc) {
+        statePtr->freeProc(statePtr->result);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetResult --
+ *
+ *	Arrange for "result" to be the Tcl return value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	interp->result is left pointing either to "result" or to a copy of it.
+ *	Also, the object result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetResult(
+    Tcl_Interp *interp,		/* Interpreter with which to associate the
+				 * return value. */
+    register char *result,	/* Value to be returned. If NULL, the result
+				 * is set to an empty string. */
+    Tcl_FreeProc *freeProc)	/* Gives information about the string:
+				 * TCL_STATIC, TCL_VOLATILE, or the address of
+				 * a Tcl_FreeProc such as free. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register Tcl_FreeProc *oldFreeProc = iPtr->freeProc;
+    char *oldResult = iPtr->result;
+
+    if (result == NULL) {
+	iPtr->resultSpace[0] = 0;
+	iPtr->result = iPtr->resultSpace;
+	iPtr->freeProc = 0;
+    } else if (freeProc == TCL_VOLATILE) {
+	int length = strlen(result);
+
+	if (length > TCL_RESULT_SIZE) {
+	    iPtr->result = ckalloc(length + 1);
+	    iPtr->freeProc = TCL_DYNAMIC;
+	} else {
+	    iPtr->result = iPtr->resultSpace;
+	    iPtr->freeProc = 0;
+	}
+	memcpy(iPtr->result, result, (unsigned) length+1);
+    } else {
+	iPtr->result = (char *) result;
+	iPtr->freeProc = freeProc;
+    }
+
+    /*
+     * If the old result was dynamically-allocated, free it up. Do it here,
+     * rather than at the beginning, in case the new result value was part of
+     * the old result value.
+     */
+
+    if (oldFreeProc != 0) {
+	if (oldFreeProc == TCL_DYNAMIC) {
+	    ckfree(oldResult);
+	} else {
+	    oldFreeProc(oldResult);
+	}
+    }
+
+    /*
+     * Reset the object result since we just set the string result.
+     */
+
+    ResetObjResult(iPtr);
+}
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetStringResult --
+ *
+ *	Returns an interpreter's result value as a string.
+ *
+ * Results:
+ *	The interpreter's result as a string.
+ *
+ * Side effects:
+ *	If the string result is empty, the object result is moved to the
+ *	string result, then the object result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetStringResult(
+    register Tcl_Interp *interp)/* Interpreter whose result to return. */
+{
+    Interp *iPtr = (Interp *) interp;
+#ifdef TCL_NO_DEPRECATED
+    return Tcl_GetString(iPtr->objResultPtr);
+#else
+    /*
+     * If the string result is empty, move the object result to the string
+     * result, then reset the object result.
+     */
+
+    if (*(iPtr->result) == 0) {
+	Tcl_SetResult(interp, TclGetString(Tcl_GetObjResult(interp)),
+		TCL_VOLATILE);
+    }
+    return iPtr->result;
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetObjResult --
+ *
+ *	Arrange for objPtr to be an interpreter's result value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	interp->objResultPtr is left pointing to the object referenced by
+ *	objPtr. The object's reference count is incremented since there is now
+ *	a new reference to it. The reference count for any old objResultPtr
+ *	value is decremented. Also, the string result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetObjResult(
+    Tcl_Interp *interp,		/* Interpreter with which to associate the
+				 * return object value. */
+    register Tcl_Obj *objPtr)	/* Tcl object to be returned. If NULL, the obj
+				 * result is made an empty string object. */
+{
+    register Interp *iPtr = (Interp *) interp;
+    register Tcl_Obj *oldObjResult = iPtr->objResultPtr;
+
+    iPtr->objResultPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);	/* since interp result is a reference */
+
+    /*
+     * We wait until the end to release the old object result, in case we are
+     * setting the result to itself.
+     */
+
+    TclDecrRefCount(oldObjResult);
+
+#ifndef TCL_NO_DEPRECATED
+    /*
+     * Reset the string result since we just set the result object.
+     */
+
+    if (iPtr->freeProc != NULL) {
+	if (iPtr->freeProc == TCL_DYNAMIC) {
+	    ckfree(iPtr->result);
+	} else {
+	    iPtr->freeProc(iPtr->result);
+	}
+	iPtr->freeProc = 0;
+    }
+    iPtr->result = iPtr->resultSpace;
+    iPtr->resultSpace[0] = 0;
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetObjResult --
+ *
+ *	Returns an interpreter's result value as a Tcl object. The object's
+ *	reference count is not modified; the caller must do that if it needs
+ *	to hold on to a long-term reference to it.
+ *
+ * Results:
+ *	The interpreter's result as an object.
+ *
+ * Side effects:
+ *	If the interpreter has a non-empty string result, the result object is
+ *	either empty or stale because some function set interp->result
+ *	directly. If so, the string result is moved to the result object then
+ *	the string result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetObjResult(
+    Tcl_Interp *interp)		/* Interpreter whose result to return. */
+{
+    register Interp *iPtr = (Interp *) interp;
+#ifndef TCL_NO_DEPRECATED
+    Tcl_Obj *objResultPtr;
+    int length;
+
+    /*
+     * If the string result is non-empty, move the string result to the object
+     * result, then reset the string result.
+     */
+
+    if (iPtr->result[0] != 0) {
+	ResetObjResult(iPtr);
+
+	objResultPtr = iPtr->objResultPtr;
+	length = strlen(iPtr->result);
+	TclInitStringRep(objResultPtr, iPtr->result, length);
+
+	if (iPtr->freeProc != NULL) {
+	    if (iPtr->freeProc == TCL_DYNAMIC) {
+		ckfree(iPtr->result);
+	    } else {
+		iPtr->freeProc(iPtr->result);
+	    }
+	    iPtr->freeProc = 0;
+	}
+	iPtr->result = iPtr->resultSpace;
+	iPtr->result[0] = 0;
+    }
+#endif /* !TCL_NO_DEPRECATED */
+    return iPtr->objResultPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendResultVA --
+ *
+ *	Append a variable number of strings onto the interpreter's result.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The result of the interpreter given by the first argument is extended
+ *	by the strings in the va_list (up to a terminating NULL argument).
+ *
+ *	If the string result is non-empty, the object result forced to be a
+ *	duplicate of it first. There will be a string result afterwards.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendResultVA(
+    Tcl_Interp *interp,		/* Interpreter with which to associate the
+				 * return value. */
+    va_list argList)		/* Variable argument list. */
+{
+    Tcl_Obj *objPtr = Tcl_GetObjResult(interp);
+
+    if (Tcl_IsShared(objPtr)) {
+	objPtr = Tcl_DuplicateObj(objPtr);
+    }
+    Tcl_AppendStringsToObjVA(objPtr, argList);
+    Tcl_SetObjResult(interp, objPtr);
+
+    /*
+     * Strictly we should call Tcl_GetStringResult(interp) here to make sure
+     * that interp->result is correct according to the old contract, but that
+     * makes the performance of much code (e.g. in Tk) absolutely awful. So we
+     * leave it out; code that really wants interp->result can just insert the
+     * calls to Tcl_GetStringResult() itself. [Patch 1041072 discussion]
+     */
+
+#ifdef USE_INTERP_RESULT
+    /*
+     * Ensure that the interp->result is legal so old Tcl 7.* code still
+     * works. There's still embarrasingly much of it about...
+     */
+
+    (void) Tcl_GetStringResult(interp);
+#endif /* USE_INTERP_RESULT */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendResult --
+ *
+ *	Append a variable number of strings onto the interpreter's result.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The result of the interpreter given by the first argument is extended
+ *	by the strings given by the second and following arguments (up to a
+ *	terminating NULL argument).
+ *
+ *	If the string result is non-empty, the object result forced to be a
+ *	duplicate of it first. There will be a string result afterwards.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendResult(
+    Tcl_Interp *interp, ...)
+{
+    va_list argList;
+
+    va_start(argList, interp);
+    Tcl_AppendResultVA(interp, argList);
+    va_end(argList);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendElement --
+ *
+ *	Convert a string to a valid Tcl list element and append it to the
+ *	result (which is ostensibly a list).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The result in the interpreter given by the first argument is extended
+ *	with a list element converted from string. A separator space is added
+ *	before the converted list element unless the current result is empty,
+ *	contains the single character "{", or ends in " {".
+ *
+ *	If the string result is empty, the object result is moved to the
+ *	string result, then the object result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendElement(
+    Tcl_Interp *interp,		/* Interpreter whose result is to be
+				 * extended. */
+    const char *element)	/* String to convert to list element and add
+				 * to result. */
+{
+    Interp *iPtr = (Interp *) interp;
+#ifdef TCL_NO_DEPRECATED
+    Tcl_Obj *elementPtr = Tcl_NewStringObj(element, -1);
+    Tcl_Obj *listPtr = Tcl_NewListObj(1, &elementPtr);
+    const char *bytes;
+
+    if (Tcl_IsShared(iPtr->objResultPtr)) {
+	Tcl_SetObjResult(interp, Tcl_DuplicateObj(iPtr->objResultPtr));
+    }
+    bytes = TclGetString(iPtr->objResultPtr);
+    if (TclNeedSpace(bytes, bytes+iPtr->objResultPtr->length)) {
+	Tcl_AppendToObj(iPtr->objResultPtr, " ", 1);
+    }
+    Tcl_AppendObjToObj(iPtr->objResultPtr, listPtr);
+    Tcl_DecrRefCount(listPtr);
+#else
+    char *dst;
+    int size;
+    int flags;
+
+    /*
+     * If the string result is empty, move the object result to the string
+     * result, then reset the object result.
+     */
+
+    (void) Tcl_GetStringResult(interp);
+
+    /*
+     * See how much space is needed, and grow the append buffer if needed to
+     * accommodate the list element.
+     */
+
+    size = Tcl_ScanElement(element, &flags) + 1;
+    if ((iPtr->result != iPtr->appendResult)
+	    || (iPtr->appendResult[iPtr->appendUsed] != 0)
+	    || ((size + iPtr->appendUsed) >= iPtr->appendAvl)) {
+	SetupAppendBuffer(iPtr, size+iPtr->appendUsed);
+    }
+
+    /*
+     * Convert the string into a list element and copy it to the buffer that's
+     * forming, with a space separator if needed.
+     */
+
+    dst = iPtr->appendResult + iPtr->appendUsed;
+    if (TclNeedSpace(iPtr->appendResult, dst)) {
+	iPtr->appendUsed++;
+	*dst = ' ';
+	dst++;
+
+	/*
+	 * If we need a space to separate this element from preceding stuff,
+	 * then this element will not lead a list, and need not have it's
+	 * leading '#' quoted.
+	 */
+
+	flags |= TCL_DONT_QUOTE_HASH;
+    }
+    iPtr->appendUsed += Tcl_ConvertElement(element, dst, flags);
+#endif /* !TCL_NO_DEPRECATED */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetupAppendBuffer --
+ *
+ *	This function makes sure that there is an append buffer properly
+ *	initialized, if necessary, from the interpreter's result, and that it
+ *	has at least enough room to accommodate newSpace new bytes of
+ *	information.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+static void
+SetupAppendBuffer(
+    Interp *iPtr,		/* Interpreter whose result is being set up. */
+    int newSpace)		/* Make sure that at least this many bytes of
+				 * new information may be added. */
+{
+    int totalSpace;
+
+    /*
+     * Make the append buffer larger, if that's necessary, then copy the
+     * result into the append buffer and make the append buffer the official
+     * Tcl result.
+     */
+
+    if (iPtr->result != iPtr->appendResult) {
+	/*
+	 * If an oversized buffer was used recently, then free it up so we go
+	 * back to a smaller buffer. This avoids tying up memory forever after
+	 * a large operation.
+	 */
+
+	if (iPtr->appendAvl > 500) {
+	    ckfree(iPtr->appendResult);
+	    iPtr->appendResult = NULL;
+	    iPtr->appendAvl = 0;
+	}
+	iPtr->appendUsed = strlen(iPtr->result);
+    } else if (iPtr->result[iPtr->appendUsed] != 0) {
+	/*
+	 * Most likely someone has modified a result created by
+	 * Tcl_AppendResult et al. so that it has a different size. Just
+	 * recompute the size.
+	 */
+
+	iPtr->appendUsed = strlen(iPtr->result);
+    }
+
+    totalSpace = newSpace + iPtr->appendUsed;
+    if (totalSpace >= iPtr->appendAvl) {
+	char *new;
+
+	if (totalSpace < 100) {
+	    totalSpace = 200;
+	} else {
+	    totalSpace *= 2;
+	}
+	new = ckalloc(totalSpace);
+	strcpy(new, iPtr->result);
+	if (iPtr->appendResult != NULL) {
+	    ckfree(iPtr->appendResult);
+	}
+	iPtr->appendResult = new;
+	iPtr->appendAvl = totalSpace;
+    } else if (iPtr->result != iPtr->appendResult) {
+	strcpy(iPtr->appendResult, iPtr->result);
+    }
+
+    Tcl_FreeResult((Tcl_Interp *) iPtr);
+    iPtr->result = iPtr->appendResult;
+}
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FreeResult --
+ *
+ *	This function frees up the memory associated with an interpreter's
+ *	string result. It also resets the interpreter's result object.
+ *	Tcl_FreeResult is most commonly used when a function is about to
+ *	replace one result value with another.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees the memory associated with interp's string result and sets
+ *	interp->freeProc to zero, but does not change interp->result or clear
+ *	error state. Resets interp's result object to an unshared empty
+ *	object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_FreeResult(
+    register Tcl_Interp *interp)/* Interpreter for which to free result. */
+{
+    register Interp *iPtr = (Interp *) interp;
+
+#ifndef TCL_NO_DEPRECATED
+    if (iPtr->freeProc != NULL) {
+	if (iPtr->freeProc == TCL_DYNAMIC) {
+	    ckfree(iPtr->result);
+	} else {
+	    iPtr->freeProc(iPtr->result);
+	}
+	iPtr->freeProc = 0;
+    }
+
+#endif /* !TCL_NO_DEPRECATED */
+    ResetObjResult(iPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ResetResult --
+ *
+ *	This function resets both the interpreter's string and object results.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	It resets the result object to an unshared empty object. It then
+ *	restores the interpreter's string result area to its default
+ *	initialized state, freeing up any memory that may have been allocated.
+ *	It also clears any error information for the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ResetResult(
+    register Tcl_Interp *interp)/* Interpreter for which to clear result. */
+{
+    register Interp *iPtr = (Interp *) interp;
+
+    ResetObjResult(iPtr);
+#ifndef TCL_NO_DEPRECATED
+    if (iPtr->freeProc != NULL) {
+	if (iPtr->freeProc == TCL_DYNAMIC) {
+	    ckfree(iPtr->result);
+	} else {
+	    iPtr->freeProc(iPtr->result);
+	}
+	iPtr->freeProc = 0;
+    }
+    iPtr->result = iPtr->resultSpace;
+    iPtr->resultSpace[0] = 0;
+#endif /* !TCL_NO_DEPRECATED */
+    if (iPtr->errorCode) {
+	/* Legacy support */
+	if (iPtr->flags & ERR_LEGACY_COPY) {
+	    Tcl_ObjSetVar2(interp, iPtr->ecVar, NULL,
+		    iPtr->errorCode, TCL_GLOBAL_ONLY);
+	}
+	Tcl_DecrRefCount(iPtr->errorCode);
+	iPtr->errorCode = NULL;
+    }
+    if (iPtr->errorInfo) {
+	/* Legacy support */
+	if (iPtr->flags & ERR_LEGACY_COPY) {
+	    Tcl_ObjSetVar2(interp, iPtr->eiVar, NULL,
+		    iPtr->errorInfo, TCL_GLOBAL_ONLY);
+	}
+	Tcl_DecrRefCount(iPtr->errorInfo);
+	iPtr->errorInfo = NULL;
+    }
+    iPtr->resetErrorStack = 1;
+    iPtr->returnLevel = 1;
+    iPtr->returnCode = TCL_OK;
+    if (iPtr->returnOpts) {
+	Tcl_DecrRefCount(iPtr->returnOpts);
+	iPtr->returnOpts = NULL;
+    }
+    iPtr->flags &= ~(ERR_ALREADY_LOGGED | ERR_LEGACY_COPY);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResetObjResult --
+ *
+ *	Function used to reset an interpreter's Tcl result object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Resets the interpreter's result object to an unshared empty string
+ *	object with ref count one. It does not clear any error information in
+ *	the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ResetObjResult(
+    register Interp *iPtr)	/* Points to the interpreter whose result
+				 * object should be reset. */
+{
+    register Tcl_Obj *objResultPtr = iPtr->objResultPtr;
+
+    if (Tcl_IsShared(objResultPtr)) {
+	TclDecrRefCount(objResultPtr);
+	TclNewObj(objResultPtr);
+	Tcl_IncrRefCount(objResultPtr);
+	iPtr->objResultPtr = objResultPtr;
+    } else {
+	if (objResultPtr->bytes != &tclEmptyString) {
+	    if (objResultPtr->bytes) {
+		ckfree(objResultPtr->bytes);
+	    }
+	    objResultPtr->bytes = &tclEmptyString;
+	    objResultPtr->length = 0;
+	}
+	TclFreeIntRep(objResultPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetErrorCodeVA --
+ *
+ *	This function is called to record machine-readable information about
+ *	an error that is about to be returned.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The errorCode field of the interp is modified to hold all of the
+ *	arguments to this function, in a list form with each argument becoming
+ *	one element of the list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetErrorCodeVA(
+    Tcl_Interp *interp,		/* Interpreter in which to set errorCode */
+    va_list argList)		/* Variable argument list. */
+{
+    Tcl_Obj *errorObj = Tcl_NewObj();
+
+    /*
+     * Scan through the arguments one at a time, appending them to the
+     * errorCode field as list elements.
+     */
+
+    while (1) {
+	char *elem = va_arg(argList, char *);
+
+	if (elem == NULL) {
+	    break;
+	}
+	Tcl_ListObjAppendElement(NULL, errorObj, Tcl_NewStringObj(elem, -1));
+    }
+    Tcl_SetObjErrorCode(interp, errorObj);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetErrorCode --
+ *
+ *	This function is called to record machine-readable information about
+ *	an error that is about to be returned.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The errorCode field of the interp is modified to hold all of the
+ *	arguments to this function, in a list form with each argument becoming
+ *	one element of the list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetErrorCode(
+    Tcl_Interp *interp, ...)
+{
+    va_list argList;
+
+    /*
+     * Scan through the arguments one at a time, appending them to the
+     * errorCode field as list elements.
+     */
+
+    va_start(argList, interp);
+    Tcl_SetErrorCodeVA(interp, argList);
+    va_end(argList);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetObjErrorCode --
+ *
+ *	This function is called to record machine-readable information about
+ *	an error that is about to be returned. The caller should build a list
+ *	object up and pass it to this routine.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The errorCode field of the interp is set to the new value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetObjErrorCode(
+    Tcl_Interp *interp,
+    Tcl_Obj *errorObjPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    if (iPtr->errorCode) {
+	Tcl_DecrRefCount(iPtr->errorCode);
+    }
+    iPtr->errorCode = errorObjPtr;
+    Tcl_IncrRefCount(iPtr->errorCode);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetErrorLine --
+ *
+ *      Returns the line number associated with the current error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_GetErrorLine
+int
+Tcl_GetErrorLine(
+    Tcl_Interp *interp)
+{
+    return ((Interp *) interp)->errorLine;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetErrorLine --
+ *
+ *      Sets the line number associated with the current error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_SetErrorLine
+void
+Tcl_SetErrorLine(
+    Tcl_Interp *interp,
+    int value)
+{
+    ((Interp *) interp)->errorLine = value;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetKeys --
+ *
+ *	Returns a Tcl_Obj * array of the standard keys used in the return
+ *	options dictionary.
+ *
+ *	Broadly sharing one copy of these key values helps with both memory
+ *	efficiency and dictionary lookup times.
+ *
+ * Results:
+ *	A Tcl_Obj * array.
+ *
+ * Side effects:
+ *	First time called in a thread, creates the keys (allocating memory)
+ *	and arranges for their cleanup at thread exit.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Obj **
+GetKeys(void)
+{
+    static Tcl_ThreadDataKey returnKeysKey;
+    Tcl_Obj **keys = Tcl_GetThreadData(&returnKeysKey,
+	    (int) (KEY_LAST * sizeof(Tcl_Obj *)));
+
+    if (keys[0] == NULL) {
+	/*
+	 * First call in this thread, create the keys...
+	 */
+
+	int i;
+
+	TclNewLiteralStringObj(keys[KEY_CODE],	    "-code");
+	TclNewLiteralStringObj(keys[KEY_ERRORCODE], "-errorcode");
+	TclNewLiteralStringObj(keys[KEY_ERRORINFO], "-errorinfo");
+	TclNewLiteralStringObj(keys[KEY_ERRORLINE], "-errorline");
+	TclNewLiteralStringObj(keys[KEY_ERRORSTACK],"-errorstack");
+	TclNewLiteralStringObj(keys[KEY_LEVEL],	    "-level");
+	TclNewLiteralStringObj(keys[KEY_OPTIONS],   "-options");
+
+	for (i = KEY_CODE; i < KEY_LAST; i++) {
+	    Tcl_IncrRefCount(keys[i]);
+	}
+
+	/*
+	 * ... and arrange for their clenaup.
+	 */
+
+	Tcl_CreateThreadExitHandler(ReleaseKeys, keys);
+    }
+    return keys;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReleaseKeys --
+ *
+ *	Called as a thread exit handler to cleanup return options dictionary
+ *	keys.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReleaseKeys(
+    ClientData clientData)
+{
+    Tcl_Obj **keys = clientData;
+    int i;
+
+    for (i = KEY_CODE; i < KEY_LAST; i++) {
+	Tcl_DecrRefCount(keys[i]);
+	keys[i] = NULL;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclProcessReturn --
+ *
+ *	Does the work of the [return] command based on the code, level, and
+ *	returnOpts arguments. Note that the code argument must agree with the
+ *	-code entry in returnOpts and the level argument must agree with the
+ *	-level entry in returnOpts, as is the case for values returned from
+ *	TclMergeReturnOptions.
+ *
+ * Results:
+ *	Returns the return code the [return] command should return.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclProcessReturn(
+    Tcl_Interp *interp,
+    int code,
+    int level,
+    Tcl_Obj *returnOpts)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *valuePtr;
+    Tcl_Obj **keys = GetKeys();
+
+    /*
+     * Store the merged return options.
+     */
+
+    if (iPtr->returnOpts != returnOpts) {
+	if (iPtr->returnOpts) {
+	    Tcl_DecrRefCount(iPtr->returnOpts);
+	}
+	iPtr->returnOpts = returnOpts;
+	Tcl_IncrRefCount(iPtr->returnOpts);
+    }
+
+    if (code == TCL_ERROR) {
+	if (iPtr->errorInfo) {
+	    Tcl_DecrRefCount(iPtr->errorInfo);
+	    iPtr->errorInfo = NULL;
+	}
+	Tcl_DictObjGet(NULL, iPtr->returnOpts, keys[KEY_ERRORINFO],
+                &valuePtr);
+	if (valuePtr != NULL) {
+	    (void) TclGetString(valuePtr);
+	    if (valuePtr->length) {
+		iPtr->errorInfo = valuePtr;
+		Tcl_IncrRefCount(iPtr->errorInfo);
+		iPtr->flags |= ERR_ALREADY_LOGGED;
+	    }
+	}
+	Tcl_DictObjGet(NULL, iPtr->returnOpts, keys[KEY_ERRORSTACK],
+                &valuePtr);
+	if (valuePtr != NULL) {
+            int len, valueObjc;
+            Tcl_Obj **valueObjv;
+
+            if (Tcl_IsShared(iPtr->errorStack)) {
+                Tcl_Obj *newObj;
+
+                newObj = Tcl_DuplicateObj(iPtr->errorStack);
+                Tcl_DecrRefCount(iPtr->errorStack);
+                Tcl_IncrRefCount(newObj);
+                iPtr->errorStack = newObj;
+            }
+
+            /*
+             * List extraction done after duplication to avoid moving the rug
+             * if someone does [return -errorstack [info errorstack]]
+             */
+
+            if (Tcl_ListObjGetElements(interp, valuePtr, &valueObjc,
+                    &valueObjv) == TCL_ERROR) {
+                return TCL_ERROR;
+            }
+            iPtr->resetErrorStack = 0;
+            Tcl_ListObjLength(interp, iPtr->errorStack, &len);
+
+            /*
+             * Reset while keeping the list intrep as much as possible.
+             */
+
+            Tcl_ListObjReplace(interp, iPtr->errorStack, 0, len, valueObjc,
+                    valueObjv);
+ 	}
+	Tcl_DictObjGet(NULL, iPtr->returnOpts, keys[KEY_ERRORCODE],
+                &valuePtr);
+	if (valuePtr != NULL) {
+	    Tcl_SetObjErrorCode(interp, valuePtr);
+	} else {
+	    Tcl_SetErrorCode(interp, "NONE", NULL);
+	}
+
+	Tcl_DictObjGet(NULL, iPtr->returnOpts, keys[KEY_ERRORLINE],
+                &valuePtr);
+	if (valuePtr != NULL) {
+	    TclGetIntFromObj(NULL, valuePtr, &iPtr->errorLine);
+	}
+    }
+    if (level != 0) {
+	iPtr->returnLevel = level;
+	iPtr->returnCode = code;
+	return TCL_RETURN;
+    }
+    if (code == TCL_ERROR) {
+	iPtr->flags |= ERR_LEGACY_COPY;
+    }
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclMergeReturnOptions --
+ *
+ *	Parses, checks, and stores the options to the [return] command.
+ *
+ * Results:
+ *	Returns TCL_ERROR if any of the option values are invalid. Otherwise,
+ *	returns TCL_OK, and writes the returnOpts, code, and level values to
+ *	the pointers provided.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclMergeReturnOptions(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[],	/* Argument objects. */
+    Tcl_Obj **optionsPtrPtr,	/* If not NULL, points to space for a (Tcl_Obj
+				 * *) where the pointer to the merged return
+				 * options dictionary should be written. */
+    int *codePtr,		/* If not NULL, points to space where the
+				 * -code value should be written. */
+    int *levelPtr)		/* If not NULL, points to space where the
+				 * -level value should be written. */
+{
+    int code = TCL_OK;
+    int level = 1;
+    Tcl_Obj *valuePtr;
+    Tcl_Obj *returnOpts = Tcl_NewObj();
+    Tcl_Obj **keys = GetKeys();
+
+    for (;  objc > 1;  objv += 2, objc -= 2) {
+	const char *opt = TclGetString(objv[0]);
+	const char *compare = TclGetString(keys[KEY_OPTIONS]);
+
+	if ((objv[0]->length == keys[KEY_OPTIONS]->length)
+		&& (memcmp(opt, compare, objv[0]->length) == 0)) {
+	    Tcl_DictSearch search;
+	    int done = 0;
+	    Tcl_Obj *keyPtr;
+	    Tcl_Obj *dict = objv[1];
+
+	nestedOptions:
+	    if (TCL_ERROR == Tcl_DictObjFirst(NULL, dict, &search,
+		    &keyPtr, &valuePtr, &done)) {
+		/*
+		 * Value is not a legal dictionary.
+		 */
+
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                        "bad %s value: expected dictionary but got \"%s\"",
+                        compare, TclGetString(objv[1])));
+		Tcl_SetErrorCode(interp, "TCL", "RESULT", "ILLEGAL_OPTIONS",
+			NULL);
+		goto error;
+	    }
+
+	    while (!done) {
+		Tcl_DictObjPut(NULL, returnOpts, keyPtr, valuePtr);
+		Tcl_DictObjNext(&search, &keyPtr, &valuePtr, &done);
+	    }
+
+	    Tcl_DictObjGet(NULL, returnOpts, keys[KEY_OPTIONS], &valuePtr);
+	    if (valuePtr != NULL) {
+		dict = valuePtr;
+		Tcl_DictObjRemove(NULL, returnOpts, keys[KEY_OPTIONS]);
+		goto nestedOptions;
+	    }
+
+	} else {
+	    Tcl_DictObjPut(NULL, returnOpts, objv[0], objv[1]);
+	}
+    }
+
+    /*
+     * Check for bogus -code value.
+     */
+
+    Tcl_DictObjGet(NULL, returnOpts, keys[KEY_CODE], &valuePtr);
+    if (valuePtr != NULL) {
+	if (TclGetCompletionCodeFromObj(interp, valuePtr,
+                &code) == TCL_ERROR) {
+	    goto error;
+	}
+	Tcl_DictObjRemove(NULL, returnOpts, keys[KEY_CODE]);
+    }
+
+    /*
+     * Check for bogus -level value.
+     */
+
+    Tcl_DictObjGet(NULL, returnOpts, keys[KEY_LEVEL], &valuePtr);
+    if (valuePtr != NULL) {
+	if ((TCL_ERROR == TclGetIntFromObj(NULL, valuePtr, &level))
+		|| (level < 0)) {
+	    /*
+	     * Value is not a legal level.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "bad -level value: expected non-negative integer but got"
+                    " \"%s\"", TclGetString(valuePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "RESULT", "ILLEGAL_LEVEL", NULL);
+	    goto error;
+	}
+	Tcl_DictObjRemove(NULL, returnOpts, keys[KEY_LEVEL]);
+    }
+
+    /*
+     * Check for bogus -errorcode value.
+     */
+
+    Tcl_DictObjGet(NULL, returnOpts, keys[KEY_ERRORCODE], &valuePtr);
+    if (valuePtr != NULL) {
+	int length;
+
+	if (TCL_ERROR == Tcl_ListObjLength(NULL, valuePtr, &length )) {
+	    /*
+	     * Value is not a list, which is illegal for -errorcode.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "bad -errorcode value: expected a list but got \"%s\"",
+                    TclGetString(valuePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "RESULT", "ILLEGAL_ERRORCODE",
+		    NULL);
+	    goto error;
+	}
+    }
+
+    /*
+     * Check for bogus -errorstack value.
+     */
+
+    Tcl_DictObjGet(NULL, returnOpts, keys[KEY_ERRORSTACK], &valuePtr);
+    if (valuePtr != NULL) {
+	int length;
+
+	if (TCL_ERROR == Tcl_ListObjLength(NULL, valuePtr, &length )) {
+	    /*
+	     * Value is not a list, which is illegal for -errorstack.
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "bad -errorstack value: expected a list but got \"%s\"",
+                    TclGetString(valuePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "RESULT", "NONLIST_ERRORSTACK",
+                    NULL);
+	    goto error;
+	}
+        if (length % 2) {
+            /*
+             * Errorstack must always be an even-sized list
+             */
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "forbidden odd-sized list for -errorstack: \"%s\"",
+		    TclGetString(valuePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "RESULT",
+                    "ODDSIZEDLIST_ERRORSTACK", NULL);
+	    goto error;
+        }
+    }
+
+    /*
+     * Convert [return -code return -level X] to [return -code ok -level X+1]
+     */
+
+    if (code == TCL_RETURN) {
+	level++;
+	code = TCL_OK;
+    }
+
+    if (codePtr != NULL) {
+	*codePtr = code;
+    }
+    if (levelPtr != NULL) {
+	*levelPtr = level;
+    }
+
+    if (optionsPtrPtr == NULL) {
+	/*
+	 * Not passing back the options (?!), so clean them up.
+	 */
+
+	Tcl_DecrRefCount(returnOpts);
+    } else {
+	*optionsPtrPtr = returnOpts;
+    }
+    return TCL_OK;
+
+  error:
+    Tcl_DecrRefCount(returnOpts);
+    return TCL_ERROR;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_GetReturnOptions --
+ *
+ *	Packs up the interp state into a dictionary of return options.
+ *
+ * Results:
+ *	A dictionary of return options.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetReturnOptions(
+    Tcl_Interp *interp,
+    int result)
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *options;
+    Tcl_Obj **keys = GetKeys();
+
+    if (iPtr->returnOpts) {
+	options = Tcl_DuplicateObj(iPtr->returnOpts);
+    } else {
+	options = Tcl_NewObj();
+    }
+
+    if (result == TCL_RETURN) {
+	Tcl_DictObjPut(NULL, options, keys[KEY_CODE],
+		Tcl_NewIntObj(iPtr->returnCode));
+	Tcl_DictObjPut(NULL, options, keys[KEY_LEVEL],
+		Tcl_NewIntObj(iPtr->returnLevel));
+    } else {
+	Tcl_DictObjPut(NULL, options, keys[KEY_CODE],
+		Tcl_NewIntObj(result));
+	Tcl_DictObjPut(NULL, options, keys[KEY_LEVEL],
+		Tcl_NewIntObj(0));
+    }
+
+    if (result == TCL_ERROR) {
+	Tcl_AddErrorInfo(interp, "");
+        Tcl_DictObjPut(NULL, options, keys[KEY_ERRORSTACK], iPtr->errorStack);
+    }
+    if (iPtr->errorCode) {
+	Tcl_DictObjPut(NULL, options, keys[KEY_ERRORCODE], iPtr->errorCode);
+    }
+    if (iPtr->errorInfo) {
+	Tcl_DictObjPut(NULL, options, keys[KEY_ERRORINFO], iPtr->errorInfo);
+	Tcl_DictObjPut(NULL, options, keys[KEY_ERRORLINE],
+		Tcl_NewIntObj(iPtr->errorLine));
+    }
+    return options;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * TclNoErrorStack --
+ *
+ *	Removes the -errorstack entry from an options dict to avoid reference
+ *	cycles.
+ *
+ * Results:
+ *	The (unshared) argument options dict, modified in -place.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclNoErrorStack(
+    Tcl_Interp *interp,
+    Tcl_Obj *options)
+{
+    Tcl_Obj **keys = GetKeys();
+
+    Tcl_DictObjRemove(interp, options, keys[KEY_ERRORSTACK]);
+    return options;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_SetReturnOptions --
+ *
+ *	Accepts an interp and a dictionary of return options, and sets the
+ *	return options of the interp to match the dictionary.
+ *
+ * Results:
+ *	A standard status code. Usually TCL_OK, but TCL_ERROR if an invalid
+ *	option value was found in the dictionary. If a -level value of 0 is in
+ *	the dictionary, then the -code value in the dictionary will be
+ *	returned (TCL_OK default).
+ *
+ * Side effects:
+ *	Sets the state of the interp.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+int
+Tcl_SetReturnOptions(
+    Tcl_Interp *interp,
+    Tcl_Obj *options)
+{
+    int objc, level, code;
+    Tcl_Obj **objv, *mergedOpts;
+
+    Tcl_IncrRefCount(options);
+    if (TCL_ERROR == TclListObjGetElements(interp, options, &objc, &objv)
+	    || (objc % 2)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                "expected dict but got \"%s\"", TclGetString(options)));
+	Tcl_SetErrorCode(interp, "TCL", "RESULT", "ILLEGAL_OPTIONS", NULL);
+	code = TCL_ERROR;
+    } else if (TCL_ERROR == TclMergeReturnOptions(interp, objc, objv,
+	    &mergedOpts, &code, &level)) {
+	code = TCL_ERROR;
+    } else {
+	code = TclProcessReturn(interp, code, level, mergedOpts);
+    }
+
+    Tcl_DecrRefCount(options);
+    return code;
+}
+
+/*
+ *-------------------------------------------------------------------------
+ *
+ * Tcl_TransferResult --
+ *
+ *	Copy the result (and error information) from one interp to another.
+ *	Used when one interp has caused another interp to evaluate a script
+ *	and then wants to transfer the results back to itself.
+ *
+ *	This routine copies the string reps of the result and error
+ *	information. It does not simply increment the refcounts of the result
+ *	and error information objects themselves. It is not legal to exchange
+ *	objects between interps, because an object may be kept alive by one
+ *	interp, but have an internal rep that is only valid while some other
+ *	interp is alive.
+ *
+ * Results:
+ *	The target interp's result is set to a copy of the source interp's
+ *	result. The source's errorInfo field may be transferred to the
+ *	target's errorInfo field, and the source's errorCode field may be
+ *	transferred to the target's errorCode field.
+ *
+ * Side effects:
+ *	None.
+ *
+ *-------------------------------------------------------------------------
+ */
+
+void
+Tcl_TransferResult(
+    Tcl_Interp *sourceInterp,	/* Interp whose result and error information
+				 * should be moved to the target interp.
+				 * After moving result, this interp's result
+				 * is reset. */
+    int result,			/* TCL_OK if just the result should be copied,
+				 * TCL_ERROR if both the result and error
+				 * information should be copied. */
+    Tcl_Interp *targetInterp)	/* Interp where result and error information
+				 * should be stored. If source and target are
+				 * the same, nothing is done. */
+{
+    Interp *tiPtr = (Interp *) targetInterp;
+    Interp *siPtr = (Interp *) sourceInterp;
+
+    if (sourceInterp == targetInterp) {
+	return;
+    }
+
+    if (result == TCL_OK && siPtr->returnOpts == NULL) {
+	/*
+	 * Special optimization for the common case of normal command return
+	 * code and no explicit return options.
+	 */
+
+	if (tiPtr->returnOpts) {
+	    Tcl_DecrRefCount(tiPtr->returnOpts);
+	    tiPtr->returnOpts = NULL;
+	}
+    } else {
+	Tcl_SetReturnOptions(targetInterp,
+		Tcl_GetReturnOptions(sourceInterp, result));
+	tiPtr->flags &= ~(ERR_ALREADY_LOGGED);
+    }
+    Tcl_SetObjResult(targetInterp, Tcl_GetObjResult(sourceInterp));
+    Tcl_ResetResult(sourceInterp);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclScan.c b/generic/tclScan.c
new file mode 100644
index 0000000..7a6a8a2
--- /dev/null
+++ b/generic/tclScan.c
@@ -0,0 +1,1079 @@
+/*
+ * tclScan.c --
+ *
+ *	This file contains the implementation of the "scan" command.
+ *
+ * Copyright (c) 1998 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Flag values used by Tcl_ScanObjCmd.
+ */
+
+#define SCAN_NOSKIP	0x1		/* Don't skip blanks. */
+#define SCAN_SUPPRESS	0x2		/* Suppress assignment. */
+#define SCAN_UNSIGNED	0x4		/* Read an unsigned value. */
+#define SCAN_WIDTH	0x8		/* A width value was supplied. */
+
+#define SCAN_LONGER	0x400		/* Asked for a wide value. */
+#define SCAN_BIG	0x800		/* Asked for a bignum value. */
+
+/*
+ * The following structure contains the information associated with a
+ * character set.
+ */
+
+typedef struct CharSet {
+    int exclude;		/* 1 if this is an exclusion set. */
+    int nchars;
+    Tcl_UniChar *chars;
+    int nranges;
+    struct Range {
+	Tcl_UniChar start;
+	Tcl_UniChar end;
+    } *ranges;
+} CharSet;
+
+/*
+ * Declarations for functions used only in this file.
+ */
+
+static const char *	BuildCharSet(CharSet *cset, const char *format);
+static int		CharInSet(CharSet *cset, int ch);
+static void		ReleaseCharSet(CharSet *cset);
+static int		ValidateFormat(Tcl_Interp *interp, const char *format,
+			    int numVars, int *totalVars);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BuildCharSet --
+ *
+ *	This function examines a character set format specification and builds
+ *	a CharSet containing the individual characters and character ranges
+ *	specified.
+ *
+ * Results:
+ *	Returns the next format position.
+ *
+ * Side effects:
+ *	Initializes the charset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static const char *
+BuildCharSet(
+    CharSet *cset,
+    const char *format)		/* Points to first char of set. */
+{
+    Tcl_UniChar ch, start;
+    int offset, nranges;
+    const char *end;
+
+    memset(cset, 0, sizeof(CharSet));
+
+    offset = TclUtfToUniChar(format, &ch);
+    if (ch == '^') {
+	cset->exclude = 1;
+	format += offset;
+	offset = TclUtfToUniChar(format, &ch);
+    }
+    end = format + offset;
+
+    /*
+     * Find the close bracket so we can overallocate the set.
+     */
+
+    if (ch == ']') {
+	end += TclUtfToUniChar(end, &ch);
+    }
+    nranges = 0;
+    while (ch != ']') {
+	if (ch == '-') {
+	    nranges++;
+	}
+	end += TclUtfToUniChar(end, &ch);
+    }
+
+    cset->chars = ckalloc(sizeof(Tcl_UniChar) * (end - format - 1));
+    if (nranges > 0) {
+	cset->ranges = ckalloc(sizeof(struct Range) * nranges);
+    } else {
+	cset->ranges = NULL;
+    }
+
+    /*
+     * Now build the character set.
+     */
+
+    cset->nchars = cset->nranges = 0;
+    format += TclUtfToUniChar(format, &ch);
+    start = ch;
+    if (ch == ']' || ch == '-') {
+	cset->chars[cset->nchars++] = ch;
+	format += TclUtfToUniChar(format, &ch);
+    }
+    while (ch != ']') {
+	if (*format == '-') {
+	    /*
+	     * This may be the first character of a range, so don't add it
+	     * yet.
+	     */
+
+	    start = ch;
+	} else if (ch == '-') {
+	    /*
+	     * Check to see if this is the last character in the set, in which
+	     * case it is not a range and we should add the previous character
+	     * as well as the dash.
+	     */
+
+	    if (*format == ']') {
+		cset->chars[cset->nchars++] = start;
+		cset->chars[cset->nchars++] = ch;
+	    } else {
+		format += TclUtfToUniChar(format, &ch);
+
+		/*
+		 * Check to see if the range is in reverse order.
+		 */
+
+		if (start < ch) {
+		    cset->ranges[cset->nranges].start = start;
+		    cset->ranges[cset->nranges].end = ch;
+		} else {
+		    cset->ranges[cset->nranges].start = ch;
+		    cset->ranges[cset->nranges].end = start;
+		}
+		cset->nranges++;
+	    }
+	} else {
+	    cset->chars[cset->nchars++] = ch;
+	}
+	format += TclUtfToUniChar(format, &ch);
+    }
+    return format;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CharInSet --
+ *
+ *	Check to see if a character matches the given set.
+ *
+ * Results:
+ *	Returns non-zero if the character matches the given set.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CharInSet(
+    CharSet *cset,
+    int c)			/* Character to test, passed as int because of
+				 * non-ANSI prototypes. */
+{
+    Tcl_UniChar ch = (Tcl_UniChar) c;
+    int i, match = 0;
+
+    for (i = 0; i < cset->nchars; i++) {
+	if (cset->chars[i] == ch) {
+	    match = 1;
+	    break;
+	}
+    }
+    if (!match) {
+	for (i = 0; i < cset->nranges; i++) {
+	    if ((cset->ranges[i].start <= ch) && (ch <= cset->ranges[i].end)) {
+		match = 1;
+		break;
+	    }
+	}
+    }
+    return (cset->exclude ? !match : match);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ReleaseCharSet --
+ *
+ *	Free the storage associated with a character set.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ReleaseCharSet(
+    CharSet *cset)
+{
+    ckfree(cset->chars);
+    if (cset->ranges) {
+	ckfree(cset->ranges);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ValidateFormat --
+ *
+ *	Parse the format string and verify that it is properly formed and that
+ *	there are exactly enough variables on the command line.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May place an error in the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ValidateFormat(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    const char *format,		/* The format string. */
+    int numVars,		/* The number of variables passed to the scan
+				 * command. */
+    int *totalSubs)		/* The number of variables that will be
+				 * required. */
+{
+    int gotXpg, gotSequential, value, i, flags;
+    char *end;
+    Tcl_UniChar ch = 0;
+    int objIndex, xpgSize, nspace = numVars;
+    int *nassign = TclStackAlloc(interp, nspace * sizeof(int));
+    char buf[TCL_UTF_MAX+1];
+    Tcl_Obj *errorMsg;		/* Place to build an error messages. Note that
+				 * these are messy operations because we do
+				 * not want to use the formatting engine;
+				 * we're inside there! */
+
+    /*
+     * Initialize an array that records the number of times a variable is
+     * assigned to by the format string. We use this to detect if a variable
+     * is multiply assigned or left unassigned.
+     */
+
+    for (i = 0; i < nspace; i++) {
+	nassign[i] = 0;
+    }
+
+    xpgSize = objIndex = gotXpg = gotSequential = 0;
+
+    while (*format != '\0') {
+	format += TclUtfToUniChar(format, &ch);
+
+	flags = 0;
+
+	if (ch != '%') {
+	    continue;
+	}
+	format += TclUtfToUniChar(format, &ch);
+	if (ch == '%') {
+	    continue;
+	}
+	if (ch == '*') {
+	    flags |= SCAN_SUPPRESS;
+	    format += TclUtfToUniChar(format, &ch);
+	    goto xpgCheckDone;
+	}
+
+	if ((ch < 0x80) && isdigit(UCHAR(ch))) {	/* INTL: "C" locale. */
+	    /*
+	     * Check for an XPG3-style %n$ specification. Note: there must
+	     * not be a mixture of XPG3 specs and non-XPG3 specs in the same
+	     * format string.
+	     */
+
+	    value = strtoul(format-1, &end, 10);	/* INTL: "C" locale. */
+	    if (*end != '$') {
+		goto notXpg;
+	    }
+	    format = end+1;
+	    format += TclUtfToUniChar(format, &ch);
+	    gotXpg = 1;
+	    if (gotSequential) {
+		goto mixedXPG;
+	    }
+	    objIndex = value - 1;
+	    if ((objIndex < 0) || (numVars && (objIndex >= numVars))) {
+		goto badIndex;
+	    } else if (numVars == 0) {
+		/*
+		 * In the case where no vars are specified, the user can
+		 * specify %9999$ legally, so we have to consider special
+		 * rules for growing the assign array. 'value' is guaranteed
+		 * to be > 0.
+		 */
+		xpgSize = (xpgSize > value) ? xpgSize : value;
+	    }
+	    goto xpgCheckDone;
+	}
+
+    notXpg:
+	gotSequential = 1;
+	if (gotXpg) {
+	mixedXPG:
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "cannot mix \"%\" and \"%n$\" conversion specifiers",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "FORMAT", "MIXEDSPECTYPES", NULL);
+	    goto error;
+	}
+
+    xpgCheckDone:
+	/*
+	 * Parse any width specifier.
+	 */
+
+	if ((ch < 0x80) && isdigit(UCHAR(ch))) {	/* INTL: "C" locale. */
+	    value = strtoul(format-1, (char **) &format, 10);	/* INTL: "C" locale. */
+	    flags |= SCAN_WIDTH;
+	    format += TclUtfToUniChar(format, &ch);
+	}
+
+	/*
+	 * Handle any size specifier.
+	 */
+
+	switch (ch) {
+	case 'l':
+	    if (*format == 'l') {
+		flags |= SCAN_BIG;
+		format += 1;
+		format += TclUtfToUniChar(format, &ch);
+		break;
+	    }
+	case 'L':
+	    flags |= SCAN_LONGER;
+	case 'h':
+	    format += TclUtfToUniChar(format, &ch);
+	}
+
+	if (!(flags & SCAN_SUPPRESS) && numVars && (objIndex >= numVars)) {
+	    goto badIndex;
+	}
+
+	/*
+	 * Handle the various field types.
+	 */
+
+	switch (ch) {
+	case 'c':
+	    if (flags & SCAN_WIDTH) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"field width may not be specified in %c conversion",
+			-1));
+		Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BADWIDTH", NULL);
+		goto error;
+	    }
+	    /*
+	     * Fall through!
+	     */
+	case 'n':
+	case 's':
+	    if (flags & (SCAN_LONGER|SCAN_BIG)) {
+	    invalidFieldSize:
+		buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
+		errorMsg = Tcl_NewStringObj(
+			"field size modifier may not be specified in %", -1);
+		Tcl_AppendToObj(errorMsg, buf, -1);
+		Tcl_AppendToObj(errorMsg, " conversion", -1);
+		Tcl_SetObjResult(interp, errorMsg);
+		Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BADSIZE", NULL);
+		goto error;
+	    }
+	    /*
+	     * Fall through!
+	     */
+	case 'd':
+	case 'e':
+	case 'E':
+	case 'f':
+	case 'g':
+	case 'G':
+	case 'i':
+	case 'o':
+	case 'x':
+	case 'X':
+	case 'b':
+	    break;
+	case 'u':
+	    if (flags & SCAN_BIG) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"unsigned bignum scans are invalid", -1));
+		Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BADUNSIGNED",NULL);
+		goto error;
+	    }
+	    break;
+	    /*
+	     * Bracket terms need special checking
+	     */
+	case '[':
+	    if (flags & (SCAN_LONGER|SCAN_BIG)) {
+		goto invalidFieldSize;
+	    }
+	    if (*format == '\0') {
+		goto badSet;
+	    }
+	    format += TclUtfToUniChar(format, &ch);
+	    if (ch == '^') {
+		if (*format == '\0') {
+		    goto badSet;
+		}
+		format += TclUtfToUniChar(format, &ch);
+	    }
+	    if (ch == ']') {
+		if (*format == '\0') {
+		    goto badSet;
+		}
+		format += TclUtfToUniChar(format, &ch);
+	    }
+	    while (ch != ']') {
+		if (*format == '\0') {
+		    goto badSet;
+		}
+		format += TclUtfToUniChar(format, &ch);
+	    }
+	    break;
+	badSet:
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "unmatched [ in format string", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BRACKET", NULL);
+	    goto error;
+	default:
+	    buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
+	    errorMsg = Tcl_NewStringObj(
+		    "bad scan conversion character \"", -1);
+	    Tcl_AppendToObj(errorMsg, buf, -1);
+	    Tcl_AppendToObj(errorMsg, "\"", -1);
+	    Tcl_SetObjResult(interp, errorMsg);
+	    Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BADTYPE", NULL);
+	    goto error;
+	}
+	if (!(flags & SCAN_SUPPRESS)) {
+	    if (objIndex >= nspace) {
+		/*
+		 * Expand the nassign buffer. If we are using XPG specifiers,
+		 * make sure that we grow to a large enough size. xpgSize is
+		 * guaranteed to be at least one larger than objIndex.
+		 */
+
+		value = nspace;
+		if (xpgSize) {
+		    nspace = xpgSize;
+		} else {
+		    nspace += 16;	/* formerly STATIC_LIST_SIZE */
+		}
+		nassign = TclStackRealloc(interp, nassign,
+			nspace * sizeof(int));
+		for (i = value; i < nspace; i++) {
+		    nassign[i] = 0;
+		}
+	    }
+	    nassign[objIndex]++;
+	    objIndex++;
+	}
+    }
+
+    /*
+     * Verify that all of the variable were assigned exactly once.
+     */
+
+    if (numVars == 0) {
+	if (xpgSize) {
+	    numVars = xpgSize;
+	} else {
+	    numVars = objIndex;
+	}
+    }
+    if (totalSubs) {
+	*totalSubs = numVars;
+    }
+    for (i = 0; i < numVars; i++) {
+	if (nassign[i] > 1) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "variable is assigned by multiple \"%n$\" conversion specifiers",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "FORMAT", "POLYASSIGNED", NULL);
+	    goto error;
+	} else if (!xpgSize && (nassign[i] == 0)) {
+	    /*
+	     * If the space is empty, and xpgSize is 0 (means XPG wasn't used,
+	     * and/or numVars != 0), then too many vars were given
+	     */
+
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "variable is not assigned by any conversion specifiers",
+		    -1));
+	    Tcl_SetErrorCode(interp, "TCL", "FORMAT", "UNASSIGNED", NULL);
+	    goto error;
+	}
+    }
+
+    TclStackFree(interp, nassign);
+    return TCL_OK;
+
+  badIndex:
+    if (gotXpg) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"\"%n$\" argument index out of range", -1));
+	Tcl_SetErrorCode(interp, "TCL", "FORMAT", "INDEXRANGE", NULL);
+    } else {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"different numbers of variable names and field specifiers",
+		-1));
+	Tcl_SetErrorCode(interp, "TCL", "FORMAT", "FIELDVARMISMATCH", NULL);
+    }
+
+  error:
+    TclStackFree(interp, nassign);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ScanObjCmd --
+ *
+ *	This function is invoked to process the "scan" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_ScanObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *format;
+    int numVars, nconversions, totalVars = -1;
+    int objIndex, offset, i, result, code;
+    long value;
+    const char *string, *end, *baseString;
+    char op = 0;
+    int width, underflow = 0;
+    Tcl_WideInt wideValue;
+    Tcl_UniChar ch, sch;
+    Tcl_Obj **objs = NULL, *objPtr = NULL;
+    int flags;
+    char buf[513];		/* Temporary buffer to hold scanned number
+				 * strings before they are passed to
+				 * strtoul. */
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"string format ?varName ...?");
+	return TCL_ERROR;
+    }
+
+    format = Tcl_GetString(objv[2]);
+    numVars = objc-3;
+
+    /*
+     * Check for errors in the format string.
+     */
+
+    if (ValidateFormat(interp, format, numVars, &totalVars) == TCL_ERROR) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Allocate space for the result objects.
+     */
+
+    if (totalVars > 0) {
+	objs = ckalloc(sizeof(Tcl_Obj *) * totalVars);
+	for (i = 0; i < totalVars; i++) {
+	    objs[i] = NULL;
+	}
+    }
+
+    string = Tcl_GetString(objv[1]);
+    baseString = string;
+
+    /*
+     * Iterate over the format string filling in the result objects until we
+     * reach the end of input, the end of the format string, or there is a
+     * mismatch.
+     */
+
+    objIndex = 0;
+    nconversions = 0;
+    while (*format != '\0') {
+	int parseFlag = TCL_PARSE_NO_WHITESPACE;
+	format += TclUtfToUniChar(format, &ch);
+
+	flags = 0;
+
+	/*
+	 * If we see whitespace in the format, skip whitespace in the string.
+	 */
+
+	if (Tcl_UniCharIsSpace(ch)) {
+	    offset = TclUtfToUniChar(string, &sch);
+	    while (Tcl_UniCharIsSpace(sch)) {
+		if (*string == '\0') {
+		    goto done;
+		}
+		string += offset;
+		offset = TclUtfToUniChar(string, &sch);
+	    }
+	    continue;
+	}
+
+	if (ch != '%') {
+	literal:
+	    if (*string == '\0') {
+		underflow = 1;
+		goto done;
+	    }
+	    string += TclUtfToUniChar(string, &sch);
+	    if (ch != sch) {
+		goto done;
+	    }
+	    continue;
+	}
+
+	format += TclUtfToUniChar(format, &ch);
+	if (ch == '%') {
+	    goto literal;
+	}
+
+	/*
+	 * Check for assignment suppression ('*') or an XPG3-style assignment
+	 * ('%n$').
+	 */
+
+	if (ch == '*') {
+	    flags |= SCAN_SUPPRESS;
+	    format += TclUtfToUniChar(format, &ch);
+	} else if ((ch < 0x80) && isdigit(UCHAR(ch))) {	/* INTL: "C" locale. */
+	    char *formatEnd;
+	    value = strtoul(format-1, &formatEnd, 10);/* INTL: "C" locale. */
+	    if (*formatEnd == '$') {
+		format = formatEnd+1;
+		format += TclUtfToUniChar(format, &ch);
+		objIndex = (int) value - 1;
+	    }
+	}
+
+	/*
+	 * Parse any width specifier.
+	 */
+
+	if ((ch < 0x80) && isdigit(UCHAR(ch))) {	/* INTL: "C" locale. */
+	    width = (int) strtoul(format-1, (char **) &format, 10);/* INTL: "C" locale. */
+	    format += TclUtfToUniChar(format, &ch);
+	} else {
+	    width = 0;
+	}
+
+	/*
+	 * Handle any size specifier.
+	 */
+
+	switch (ch) {
+	case 'l':
+	    if (*format == 'l') {
+		flags |= SCAN_BIG;
+		format += 1;
+		format += TclUtfToUniChar(format, &ch);
+		break;
+	    }
+	case 'L':
+	    flags |= SCAN_LONGER;
+	    /*
+	     * Fall through so we skip to the next character.
+	     */
+	case 'h':
+	    format += TclUtfToUniChar(format, &ch);
+	}
+
+	/*
+	 * Handle the various field types.
+	 */
+
+	switch (ch) {
+	case 'n':
+	    if (!(flags & SCAN_SUPPRESS)) {
+		objPtr = Tcl_NewIntObj(string - baseString);
+		Tcl_IncrRefCount(objPtr);
+		CLANG_ASSERT(objs);
+		objs[objIndex++] = objPtr;
+	    }
+	    nconversions++;
+	    continue;
+
+	case 'd':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_DECIMAL_ONLY;
+	    break;
+	case 'i':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_SCAN_PREFIXES;
+	    break;
+	case 'o':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_OCTAL_ONLY | TCL_PARSE_SCAN_PREFIXES;
+	    break;
+	case 'x':
+	case 'X':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_HEXADECIMAL_ONLY;
+	    break;
+	case 'b':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_BINARY_ONLY;
+	    break;
+	case 'u':
+	    op = 'i';
+	    parseFlag |= TCL_PARSE_DECIMAL_ONLY;
+	    flags |= SCAN_UNSIGNED;
+	    break;
+
+	case 'f':
+	case 'e':
+	case 'E':
+	case 'g':
+	case 'G':
+	    op = 'f';
+	    break;
+
+	case 's':
+	    op = 's';
+	    break;
+
+	case 'c':
+	    op = 'c';
+	    flags |= SCAN_NOSKIP;
+	    break;
+	case '[':
+	    op = '[';
+	    flags |= SCAN_NOSKIP;
+	    break;
+	}
+
+	/*
+	 * At this point, we will need additional characters from the string
+	 * to proceed.
+	 */
+
+	if (*string == '\0') {
+	    underflow = 1;
+	    goto done;
+	}
+
+	/*
+	 * Skip any leading whitespace at the beginning of a field unless the
+	 * format suppresses this behavior.
+	 */
+
+	if (!(flags & SCAN_NOSKIP)) {
+	    while (*string != '\0') {
+		offset = TclUtfToUniChar(string, &sch);
+		if (!Tcl_UniCharIsSpace(sch)) {
+		    break;
+		}
+		string += offset;
+	    }
+	    if (*string == '\0') {
+		underflow = 1;
+		goto done;
+	    }
+	}
+
+	/*
+	 * Perform the requested scanning operation.
+	 */
+
+	switch (op) {
+	case 's':
+	    /*
+	     * Scan a string up to width characters or whitespace.
+	     */
+
+	    if (width == 0) {
+		width = ~0;
+	    }
+	    end = string;
+	    while (*end != '\0') {
+		offset = TclUtfToUniChar(end, &sch);
+		if (Tcl_UniCharIsSpace(sch)) {
+		    break;
+		}
+		end += offset;
+		if (--width == 0) {
+		    break;
+		}
+	    }
+	    if (!(flags & SCAN_SUPPRESS)) {
+		objPtr = Tcl_NewStringObj(string, end-string);
+		Tcl_IncrRefCount(objPtr);
+		CLANG_ASSERT(objs);
+		objs[objIndex++] = objPtr;
+	    }
+	    string = end;
+	    break;
+
+	case '[': {
+	    CharSet cset;
+
+	    if (width == 0) {
+		width = ~0;
+	    }
+	    end = string;
+
+	    format = BuildCharSet(&cset, format);
+	    while (*end != '\0') {
+		offset = TclUtfToUniChar(end, &sch);
+		if (!CharInSet(&cset, (int)sch)) {
+		    break;
+		}
+		end += offset;
+		if (--width == 0) {
+		    break;
+		}
+	    }
+	    ReleaseCharSet(&cset);
+
+	    if (string == end) {
+		/*
+		 * Nothing matched the range, stop processing.
+		 */
+		goto done;
+	    }
+	    if (!(flags & SCAN_SUPPRESS)) {
+		objPtr = Tcl_NewStringObj(string, end-string);
+		Tcl_IncrRefCount(objPtr);
+		objs[objIndex++] = objPtr;
+	    }
+	    string = end;
+
+	    break;
+	}
+	case 'c':
+	    /*
+	     * Scan a single Unicode character.
+	     */
+
+	    string += TclUtfToUniChar(string, &sch);
+	    if (!(flags & SCAN_SUPPRESS)) {
+		objPtr = Tcl_NewIntObj((int)sch);
+		Tcl_IncrRefCount(objPtr);
+		CLANG_ASSERT(objs);
+		objs[objIndex++] = objPtr;
+	    }
+	    break;
+
+	case 'i':
+	    /*
+	     * Scan an unsigned or signed integer.
+	     */
+	    objPtr = Tcl_NewLongObj(0);
+	    Tcl_IncrRefCount(objPtr);
+	    if (width == 0) {
+		width = ~0;
+	    }
+	    if (TCL_OK != TclParseNumber(NULL, objPtr, NULL, string, width,
+		    &end, TCL_PARSE_INTEGER_ONLY | parseFlag)) {
+		Tcl_DecrRefCount(objPtr);
+		if (width < 0) {
+		    if (*end == '\0') {
+			underflow = 1;
+		    }
+		} else {
+		    if (end == string + width) {
+			underflow = 1;
+		    }
+		}
+		goto done;
+	    }
+	    string = end;
+	    if (flags & SCAN_SUPPRESS) {
+		Tcl_DecrRefCount(objPtr);
+		break;
+	    }
+	    if (flags & SCAN_LONGER) {
+		if (Tcl_GetWideIntFromObj(NULL, objPtr, &wideValue) != TCL_OK) {
+		    wideValue = ~(Tcl_WideUInt)0 >> 1;	/* WIDE_MAX */
+		    if (TclGetString(objPtr)[0] == '-') {
+			wideValue++;	/* WIDE_MAX + 1 = WIDE_MIN */
+		    }
+		}
+		if ((flags & SCAN_UNSIGNED) && (wideValue < 0)) {
+		    sprintf(buf, "%" TCL_LL_MODIFIER "u",
+			    (Tcl_WideUInt)wideValue);
+		    Tcl_SetStringObj(objPtr, buf, -1);
+		} else {
+		    Tcl_SetWideIntObj(objPtr, wideValue);
+		}
+	    } else if (!(flags & SCAN_BIG)) {
+		if (TclGetLongFromObj(NULL, objPtr, &value) != TCL_OK) {
+		    if (TclGetString(objPtr)[0] == '-') {
+			value = LONG_MIN;
+		    } else {
+			value = LONG_MAX;
+		    }
+		}
+		if ((flags & SCAN_UNSIGNED) && (value < 0)) {
+		    sprintf(buf, "%lu", value);	/* INTL: ISO digit */
+		    Tcl_SetStringObj(objPtr, buf, -1);
+		} else {
+		    Tcl_SetLongObj(objPtr, value);
+		}
+	    }
+	    objs[objIndex++] = objPtr;
+	    break;
+
+	case 'f':
+	    /*
+	     * Scan a floating point number
+	     */
+
+	    objPtr = Tcl_NewDoubleObj(0.0);
+	    Tcl_IncrRefCount(objPtr);
+	    if (width == 0) {
+		width = ~0;
+	    }
+	    if (TCL_OK != TclParseNumber(NULL, objPtr, NULL, string, width,
+		    &end, TCL_PARSE_DECIMAL_ONLY | TCL_PARSE_NO_WHITESPACE)) {
+		Tcl_DecrRefCount(objPtr);
+		if (width < 0) {
+		    if (*end == '\0') {
+			underflow = 1;
+		    }
+		} else {
+		    if (end == string + width) {
+			underflow = 1;
+		    }
+		}
+		goto done;
+	    } else if (flags & SCAN_SUPPRESS) {
+		Tcl_DecrRefCount(objPtr);
+		string = end;
+	    } else {
+		double dvalue;
+		if (Tcl_GetDoubleFromObj(NULL, objPtr, &dvalue) != TCL_OK) {
+#ifdef ACCEPT_NAN
+		    if (objPtr->typePtr == &tclDoubleType) {
+			dvalue = objPtr->internalRep.doubleValue;
+		    } else
+#endif
+		    {
+			Tcl_DecrRefCount(objPtr);
+			goto done;
+		    }
+		}
+		Tcl_SetDoubleObj(objPtr, dvalue);
+		CLANG_ASSERT(objs);
+		objs[objIndex++] = objPtr;
+		string = end;
+	    }
+	}
+	nconversions++;
+    }
+
+  done:
+    result = 0;
+    code = TCL_OK;
+
+    if (numVars) {
+	/*
+	 * In this case, variables were specified (classic scan).
+	 */
+
+	for (i = 0; i < totalVars; i++) {
+	    if (objs[i] == NULL) {
+		continue;
+	    }
+	    result++;
+
+	    /*
+	     * In case of multiple errors in setting variables, just report
+	     * the first one.
+	     */
+
+	    if (Tcl_ObjSetVar2(interp, objv[i+3], NULL, objs[i],
+		    (code == TCL_OK) ? TCL_LEAVE_ERR_MSG : 0) == NULL) {
+		code = TCL_ERROR;
+	    }
+	    Tcl_DecrRefCount(objs[i]);
+	}
+    } else {
+	/*
+	 * Here no vars were specified, we want a list returned (inline scan)
+	 */
+
+	objPtr = Tcl_NewObj();
+	for (i = 0; i < totalVars; i++) {
+	    if (objs[i] != NULL) {
+		Tcl_ListObjAppendElement(NULL, objPtr, objs[i]);
+		Tcl_DecrRefCount(objs[i]);
+	    } else {
+		/*
+		 * More %-specifiers than matching chars, so we just spit out
+		 * empty strings for these.
+		 */
+
+		Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewObj());
+	    }
+	}
+    }
+    if (objs != NULL) {
+	ckfree(objs);
+    }
+    if (code == TCL_OK) {
+	if (underflow && (nconversions == 0)) {
+	    if (numVars) {
+		objPtr = Tcl_NewIntObj(-1);
+	    } else {
+		if (objPtr) {
+		    Tcl_SetListObj(objPtr, 0, NULL);
+		} else {
+		    objPtr = Tcl_NewObj();
+		}
+	    }
+	} else if (numVars) {
+	    objPtr = Tcl_NewIntObj(result);
+	}
+	Tcl_SetObjResult(interp, objPtr);
+    }
+    return code;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStrToD.c b/generic/tclStrToD.c
new file mode 100644
index 0000000..539a92c
--- /dev/null
+++ b/generic/tclStrToD.c
@@ -0,0 +1,5070 @@
+/*
+ * tclStrToD.c --
+ *
+ *	This file contains a collection of procedures for managing conversions
+ *	to/from floating-point in Tcl. They include TclParseNumber, which
+ *	parses numbers from strings; TclDoubleDigits, which formats numbers
+ *	into strings of digits, and procedures for interconversion among
+ *	'double' and 'mp_int' types.
+ *
+ * Copyright (c) 2005 by Kevin B. Kenny. All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+#include <math.h>
+
+/*
+ * This code supports (at least hypothetically), IBM, Cray, VAX and IEEE-754
+ * floating point; of these, only IEEE-754 can represent NaN. IEEE-754 can be
+ * uniquely determined by radix and by the widths of significand and exponent.
+ */
+
+#if (FLT_RADIX == 2) && (DBL_MANT_DIG == 53) && (DBL_MAX_EXP == 1024)
+#   define IEEE_FLOATING_POINT
+#endif
+
+/*
+ * Rounding controls. (Thanks a lot, Intel!)
+ */
+
+#ifdef __i386
+/*
+ * gcc on x86 needs access to rounding controls, because of a questionable
+ * feature where it retains intermediate results as IEEE 'long double' values
+ * somewhat unpredictably. It is tempting to include fpu_control.h, but that
+ * file exists only on Linux; it is missing on Cygwin and MinGW. Most gcc-isms
+ * and ix86-isms are factored out here.
+ */
+
+#if defined(__GNUC__)
+typedef unsigned int	fpu_control_t __attribute__ ((__mode__ (__HI__)));
+
+#define _FPU_GETCW(cw)	__asm__ __volatile__ ("fnstcw %0" : "=m" (*&cw))
+#define _FPU_SETCW(cw)	__asm__ __volatile__ ("fldcw %0" : : "m" (*&cw))
+#   define FPU_IEEE_ROUNDING	0x027f
+#   define ADJUST_FPU_CONTROL_WORD
+#define TCL_IEEE_DOUBLE_ROUNDING \
+    fpu_control_t roundTo53Bits = FPU_IEEE_ROUNDING;	\
+    fpu_control_t oldRoundingMode;			\
+    _FPU_GETCW(oldRoundingMode);			\
+    _FPU_SETCW(roundTo53Bits)
+#define TCL_DEFAULT_DOUBLE_ROUNDING \
+    _FPU_SETCW(oldRoundingMode)
+
+/*
+ * Sun ProC needs sunmath for rounding control on x86 like gcc above.
+ */
+#elif defined(__sun)
+#include <sunmath.h>
+#define TCL_IEEE_DOUBLE_ROUNDING \
+    ieee_flags("set","precision","double",NULL)
+#define TCL_DEFAULT_DOUBLE_ROUNDING \
+    ieee_flags("clear","precision",NULL,NULL)
+
+/*
+ * Other platforms are assumed to always operate in full IEEE mode, so we make
+ * the macros to go in and out of that mode do nothing.
+ */
+
+#else /* !__GNUC__ && !__sun */
+#define TCL_IEEE_DOUBLE_ROUNDING	((void) 0)
+#define TCL_DEFAULT_DOUBLE_ROUNDING	((void) 0)
+#endif
+#else /* !__i386 */
+#define TCL_IEEE_DOUBLE_ROUNDING	((void) 0)
+#define TCL_DEFAULT_DOUBLE_ROUNDING	((void) 0)
+#endif
+
+/*
+ * MIPS floating-point units need special settings in control registers to use
+ * gradual underflow as we expect.  This fix is for the MIPSpro compiler.
+ */
+
+#if defined(__sgi) && defined(_COMPILER_VERSION)
+#include <sys/fpu.h>
+#endif
+
+/*
+ * HP's PA_RISC architecture uses 7ff4000000000000 to represent a quiet NaN.
+ * Everyone else uses 7ff8000000000000. (Why, HP, why?)
+ */
+
+#ifdef __hppa
+#   define NAN_START	0x7ff4
+#   define NAN_MASK	(((Tcl_WideUInt) 1) << 50)
+#else
+#   define NAN_START	0x7ff8
+#   define NAN_MASK	(((Tcl_WideUInt) 1) << 51)
+#endif
+
+/*
+ * Constants used by this file (most of which are only ever calculated at
+ * runtime).
+ */
+
+/* Magic constants */
+
+#define LOG10_2 0.3010299956639812
+#define TWO_OVER_3LOG10 0.28952965460216784
+#define LOG10_3HALVES_PLUS_FUDGE 0.1760912590558
+
+/*
+ * Definitions of the parts of an IEEE754-format floating point number.
+ */
+
+#define SIGN_BIT 	0x80000000
+				/* Mask for the sign bit in the first word of
+				 * a double. */
+#define EXP_MASK	0x7ff00000
+				/* Mask for the exponent field in the first
+				 * word of a double. */
+#define EXP_SHIFT	20	/* Shift count to make the exponent an
+				 * integer. */
+#define HIDDEN_BIT	(((Tcl_WideUInt) 0x00100000) << 32)
+				/* Hidden 1 bit for the significand. */
+#define HI_ORDER_SIG_MASK 0x000fffff
+				/* Mask for the high-order part of the
+				 * significand in the first word of a
+				 * double. */
+#define SIG_MASK	(((Tcl_WideUInt) HI_ORDER_SIG_MASK << 32) \
+			| 0xffffffff)
+				/* Mask for the 52-bit significand. */
+#define FP_PRECISION	53	/* Number of bits of significand plus the
+				 * hidden bit. */
+#define EXPONENT_BIAS	0x3ff	/* Bias of the exponent 0. */
+
+/*
+ * Derived quantities.
+ */
+
+#define TEN_PMAX	22	/* floor(FP_PRECISION*log(2)/log(5)) */
+#define QUICK_MAX	14	/* floor((FP_PRECISION-1)*log(2)/log(10))-1 */
+#define BLETCH		0x10	/* Highest power of two that is greater than
+				 * DBL_MAX_10_EXP, divided by 16. */
+#define DIGIT_GROUP	8	/* floor(DIGIT_BIT*log(2)/log(10)) */
+
+/*
+ * Union used to dismantle floating point numbers.
+ */
+
+typedef union Double {
+    struct {
+#ifdef WORDS_BIGENDIAN
+	int word0;
+	int word1;
+#else
+	int word1;
+	int word0;
+#endif
+    } w;
+    double d;
+    Tcl_WideUInt q;
+} Double;
+
+static int maxpow10_wide;	/* The powers of ten that can be represented
+				 * exactly as wide integers. */
+static Tcl_WideUInt *pow10_wide;
+#define MAXPOW	22
+static double pow10vals[MAXPOW+1];
+				/* The powers of ten that can be represented
+				 * exactly as IEEE754 doubles. */
+static int mmaxpow;		/* Largest power of ten that can be
+				 * represented exactly in a 'double'. */
+static int log10_DIGIT_MAX;	/* The number of decimal digits that fit in an
+				 * mp_digit. */
+static int log2FLT_RADIX;	/* Logarithm of the floating point radix. */
+static int mantBits;		/* Number of bits in a double's significand */
+static mp_int pow5[9];		/* Table of powers of 5**(2**n), up to
+				 * 5**256 */
+static double tiny = 0.0;	/* The smallest representable double. */
+static int maxDigits;		/* The maximum number of digits to the left of
+				 * the decimal point of a double. */
+static int minDigits;		/* The maximum number of digits to the right
+				 * of the decimal point in a double. */
+static const double pow_10_2_n[] = {	/* Inexact higher powers of ten. */
+    1.0,
+    100.0,
+    10000.0,
+    1.0e+8,
+    1.0e+16,
+    1.0e+32,
+    1.0e+64,
+    1.0e+128,
+    1.0e+256
+};
+
+static int n770_fp;		/* Flag is 1 on Nokia N770 floating point.
+				 * Nokia's floating point has the words
+				 * reversed: if big-endian is 7654 3210,
+				 * and little-endian is       0123 4567,
+				 * then Nokia's FP is         4567 0123;
+				 * little-endian within the 32-bit words but
+				 * big-endian between them. */
+
+/*
+ * Table of powers of 5 that are small enough to fit in an mp_digit.
+ */
+
+static const mp_digit dpow5[13] = {
+               1,              5,             25,            125,
+             625,           3125,          15625,          78125,
+          390625,        1953125,        9765625,       48828125,
+       244140625
+};
+
+/*
+ * Table of powers: pow5_13[n] = 5**(13*2**(n+1))
+ */
+
+static mp_int pow5_13[5];	/* Table of powers: 5**13, 5**26, 5**52,
+				 * 5**104, 5**208 */
+static const double tens[] = {
+    1e00, 1e01, 1e02, 1e03, 1e04, 1e05, 1e06, 1e07, 1e08, 1e09,
+    1e10, 1e11, 1e12, 1e13, 1e14, 1e15, 1e16, 1e17, 1e18, 1e19,
+    1e20, 1e21, 1e22
+};
+
+static const int itens [] = {
+    1,
+    10,
+    100,
+    1000,
+    10000,
+    100000,
+    1000000,
+    10000000,
+    100000000
+};
+
+static const double bigtens[] = {
+    1e016, 1e032, 1e064, 1e128, 1e256
+};
+#define N_BIGTENS 5
+
+static const int log2pow5[27] = {
+    01,  3,  5,  7, 10, 12, 14, 17, 19, 21,
+    24, 26, 28, 31, 33, 35, 38, 40, 42, 45,
+    47, 49, 52, 54, 56, 59, 61
+};
+#define N_LOG2POW5 27
+
+static const Tcl_WideUInt wuipow5[27] = {
+    (Tcl_WideUInt) 1,		/* 5**0 */
+    (Tcl_WideUInt) 5,
+    (Tcl_WideUInt) 25,
+    (Tcl_WideUInt) 125,
+    (Tcl_WideUInt) 625,
+    (Tcl_WideUInt) 3125,	/* 5**5 */
+    (Tcl_WideUInt) 3125*5,
+    (Tcl_WideUInt) 3125*25,
+    (Tcl_WideUInt) 3125*125,
+    (Tcl_WideUInt) 3125*625,
+    (Tcl_WideUInt) 3125*3125,	/* 5**10 */
+    (Tcl_WideUInt) 3125*3125*5,
+    (Tcl_WideUInt) 3125*3125*25,
+    (Tcl_WideUInt) 3125*3125*125,
+    (Tcl_WideUInt) 3125*3125*625,
+    (Tcl_WideUInt) 3125*3125*3125, /* 5**15 */
+    (Tcl_WideUInt) 3125*3125*3125*5,
+    (Tcl_WideUInt) 3125*3125*3125*25,
+    (Tcl_WideUInt) 3125*3125*3125*125,
+    (Tcl_WideUInt) 3125*3125*3125*625,
+    (Tcl_WideUInt) 3125*3125*3125*3125,	/* 5**20 */
+    (Tcl_WideUInt) 3125*3125*3125*3125*5,
+    (Tcl_WideUInt) 3125*3125*3125*3125*25,
+    (Tcl_WideUInt) 3125*3125*3125*3125*125,
+    (Tcl_WideUInt) 3125*3125*3125*3125*625,
+    (Tcl_WideUInt) 3125*3125*3125*3125*3125,  /* 5**25 */
+    (Tcl_WideUInt) 3125*3125*3125*3125*3125*5 /* 5**26 */
+};
+
+/*
+ * Static functions defined in this file.
+ */
+
+static int		AccumulateDecimalDigit(unsigned, int,
+			    Tcl_WideUInt *, mp_int *, int);
+static double		MakeHighPrecisionDouble(int signum,
+			    mp_int *significand, int nSigDigs, int exponent);
+static double		MakeLowPrecisionDouble(int signum,
+			    Tcl_WideUInt significand, int nSigDigs,
+			    int exponent);
+#ifdef IEEE_FLOATING_POINT
+static double		MakeNaN(int signum, Tcl_WideUInt tag);
+#endif
+static double		RefineApproximation(double approx,
+			    mp_int *exactSignificand, int exponent);
+static void		MulPow5(mp_int *, unsigned, mp_int *);
+static int 		NormalizeRightward(Tcl_WideUInt *);
+static int		RequiredPrecision(Tcl_WideUInt);
+static void		DoubleToExpAndSig(double, Tcl_WideUInt *, int *,
+			    int *);
+static void		TakeAbsoluteValue(Double *, int *);
+static char *		FormatInfAndNaN(Double *, int *, char **);
+static char *		FormatZero(int *, char **);
+static int		ApproximateLog10(Tcl_WideUInt, int, int);
+static int		BetterLog10(double, int, int *);
+static void		ComputeScale(int, int, int *, int *, int *, int *);
+static void		SetPrecisionLimits(int, int, int *, int *, int *,
+			    int *);
+static char *		BumpUp(char *, char *, int *);
+static int		AdjustRange(double *, int);
+static char *		ShorteningQuickFormat(double, int, int, double,
+			    char *, int *);
+static char *		StrictQuickFormat(double, int, int, double,
+			    char *, int *);
+static char *		QuickConversion(double, int, int, int, int, int, int,
+			    int *, char **);
+static void		CastOutPowersOf2(int *, int *, int *);
+static char *		ShorteningInt64Conversion(Double *, int, Tcl_WideUInt,
+			    int, int, int, int, int, int, int, int, int,
+			    int, int, int *, char **);
+static char *		StrictInt64Conversion(Double *, int, Tcl_WideUInt,
+			    int, int, int, int, int, int,
+			    int, int, int *, char **);
+static int		ShouldBankerRoundUpPowD(mp_int *, int, int);
+static int		ShouldBankerRoundUpToNextPowD(mp_int *, mp_int *,
+			    int, int, int, mp_int *);
+static char *		ShorteningBignumConversionPowD(Double *dPtr,
+			    int convType, Tcl_WideUInt bw, int b2, int b5,
+			    int m2plus, int m2minus, int m5,
+			    int sd, int k, int len,
+			    int ilim, int ilim1, int *decpt,
+			    char **endPtr);
+static char *		StrictBignumConversionPowD(Double *dPtr, int convType,
+			    Tcl_WideUInt bw, int b2, int b5,
+			    int sd, int k, int len,
+			    int ilim, int ilim1, int *decpt,
+			    char **endPtr);
+static int		ShouldBankerRoundUp(mp_int *, mp_int *, int);
+static int		ShouldBankerRoundUpToNext(mp_int *, mp_int *,
+			    mp_int *, int, int, mp_int *);
+static char *		ShorteningBignumConversion(Double *dPtr, int convType,
+			    Tcl_WideUInt bw, int b2,
+			    int m2plus, int m2minus,
+			    int s2, int s5, int k, int len,
+			    int ilim, int ilim1, int *decpt,
+			    char **endPtr);
+static char *		StrictBignumConversion(Double *dPtr, int convType,
+			    Tcl_WideUInt bw, int b2,
+			    int s2, int s5, int k, int len,
+			    int ilim, int ilim1, int *decpt,
+			    char **endPtr);
+static double		BignumToBiasedFrExp(const mp_int *big, int *machexp);
+static double		Pow10TimesFrExp(int exponent, double fraction,
+			    int *machexp);
+static double		SafeLdExp(double fraction, int exponent);
+#ifdef IEEE_FLOATING_POINT
+static Tcl_WideUInt	Nokia770Twiddle(Tcl_WideUInt w);
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclParseNumber --
+ *
+ *	Scans bytes, interpreted as characters in Tcl's internal encoding, and
+ *	parses the longest prefix that is the string representation of a
+ *	number in a format recognized by Tcl.
+ *
+ *	The arguments bytes, numBytes, and objPtr are the inputs which
+ *	determine the string to be parsed. If bytes is non-NULL, it points to
+ *	the first byte to be scanned. If bytes is NULL, then objPtr must be
+ *	non-NULL, and the string representation of objPtr will be scanned
+ *	(generated first, if necessary). The numBytes argument determines the
+ *	number of bytes to be scanned. If numBytes is negative, the first NUL
+ *	byte encountered will terminate the scan. If numBytes is non-negative,
+ *	then no more than numBytes bytes will be scanned.
+ *
+ *	The argument flags is an input that controls the numeric formats
+ *	recognized by the parser. The flag bits are:
+ *
+ *	- TCL_PARSE_INTEGER_ONLY:	accept only integer values; reject
+ *		strings that denote floating point values (or accept only the
+ *		leading portion of them that are integer values).
+ *	- TCL_PARSE_SCAN_PREFIXES:	ignore the prefixes 0b and 0o that are
+ *		not part of the [scan] command's vocabulary. Use only in
+ *		combination with TCL_PARSE_INTEGER_ONLY.
+ *	- TCL_PARSE_BINARY_ONLY:	parse only in the binary format, whether
+ *		or not a prefix is present that would lead to binary parsing.
+ *		Use only in combination with TCL_PARSE_INTEGER_ONLY.
+ *	- TCL_PARSE_OCTAL_ONLY:		parse only in the octal format, whether
+ *		or not a prefix is present that would lead to octal parsing.
+ *		Use only in combination with TCL_PARSE_INTEGER_ONLY.
+ *	- TCL_PARSE_HEXADECIMAL_ONLY:	parse only in the hexadecimal format,
+ *		whether or not a prefix is present that would lead to
+ *		hexadecimal parsing. Use only in combination with
+ *		TCL_PARSE_INTEGER_ONLY.
+ *	- TCL_PARSE_DECIMAL_ONLY:	parse only in the decimal format, no
+ *		matter whether a 0 prefix would normally force a different
+ *		base.
+ *	- TCL_PARSE_NO_WHITESPACE:	reject any leading/trailing whitespace
+ *
+ *	The arguments interp and expected are inputs that control error
+ *	message generation. If interp is NULL, no error message will be
+ *	generated. If interp is non-NULL, then expected must also be non-NULL.
+ *	When TCL_ERROR is returned, an error message will be left in the
+ *	result of interp, and the expected argument will appear in the error
+ *	message as the thing TclParseNumber expected, but failed to find in
+ *	the string.
+ *
+ *	The arguments objPtr and endPtrPtr as well as the return code are the
+ *	outputs.
+ *
+ *	When the parser cannot find any prefix of the string that matches a
+ *	format it is looking for, TCL_ERROR is returned and an error message
+ *	may be generated and returned as described above. The contents of
+ *	objPtr will not be changed. If endPtrPtr is non-NULL, a pointer to the
+ *	character in the string that terminated the scan will be written to
+ *	*endPtrPtr.
+ *
+ *	When the parser determines that the entire string matches a format it
+ *	is looking for, TCL_OK is returned, and if objPtr is non-NULL, then
+ *	the internal rep and Tcl_ObjType of objPtr are set to the "canonical"
+ *	numeric value that matches the scanned string. If endPtrPtr is not
+ *	NULL, a pointer to the end of the string will be written to *endPtrPtr
+ *	(that is, either bytes+numBytes or a pointer to a terminating NUL
+ *	byte).
+ *
+ *	When the parser determines that a partial string matches a format it
+ *	is looking for, the value of endPtrPtr determines what happens:
+ *
+ *	- If endPtrPtr is NULL, then TCL_ERROR is returned, with error message
+ *		generation as above.
+ *
+ *	- If endPtrPtr is non-NULL, then TCL_OK is returned and objPtr
+ *		internals are set as above. Also, a pointer to the first
+ *		character following the parsed numeric string is written to
+ *		*endPtrPtr.
+ *
+ *	In some cases where the string being scanned is the string rep of
+ *	objPtr, this routine can leave objPtr in an inconsistent state where
+ *	its string rep and its internal rep do not agree. In these cases the
+ *	internal rep will be in agreement with only some substring of the
+ *	string rep. This might happen if the caller passes in a non-NULL bytes
+ *	value that points somewhere into the string rep. It might happen if
+ *	the caller passes in a numBytes value that limits the scan to only a
+ *	prefix of the string rep. Or it might happen if a non-NULL value of
+ *	endPtrPtr permits a TCL_OK return from only a partial string match. It
+ *	is the responsibility of the caller to detect and correct such
+ *	inconsistencies when they can and do arise.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	The string representaton of objPtr may be generated.
+ *
+ *	The internal representation and Tcl_ObjType of objPtr may be changed.
+ *	This may involve allocation and/or freeing of memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclParseNumber(
+    Tcl_Interp *interp,		/* Used for error reporting. May be NULL. */
+    Tcl_Obj *objPtr,		/* Object to receive the internal rep. */
+    const char *expected,	/* Description of the type of number the
+				 * caller expects to be able to parse
+				 * ("integer", "boolean value", etc.). */
+    const char *bytes,		/* Pointer to the start of the string to
+				 * scan. */
+    int numBytes,		/* Maximum number of bytes to scan, see
+				 * above. */
+    const char **endPtrPtr,	/* Place to store pointer to the character
+				 * that terminated the scan. */
+    int flags)			/* Flags governing the parse. */
+{
+    enum State {
+	INITIAL, SIGNUM, ZERO, ZERO_X,
+	ZERO_O, ZERO_B, ZERO_D, BINARY,
+	HEXADECIMAL, OCTAL, BAD_OCTAL, DECIMAL,
+	LEADING_RADIX_POINT, FRACTION,
+	EXPONENT_START, EXPONENT_SIGNUM, EXPONENT,
+	sI, sIN, sINF, sINFI, sINFIN, sINFINI, sINFINIT, sINFINITY
+#ifdef IEEE_FLOATING_POINT
+	, sN, sNA, sNAN, sNANPAREN, sNANHEX, sNANFINISH
+#endif
+    } state = INITIAL;
+    enum State acceptState = INITIAL;
+
+    int signum = 0;		/* Sign of the number being parsed. */
+    Tcl_WideUInt significandWide = 0;
+				/* Significand of the number being parsed (if
+				 * no overflow). */
+    mp_int significandBig;	/* Significand of the number being parsed (if
+				 * it overflows significandWide). */
+    int significandOverflow = 0;/* Flag==1 iff significandBig is used. */
+    Tcl_WideUInt octalSignificandWide = 0;
+				/* Significand of an octal number; needed
+				 * because we don't know whether a number with
+				 * a leading zero is octal or decimal until
+				 * we've scanned forward to a '.' or 'e'. */
+    mp_int octalSignificandBig;	/* Significand of octal number once
+				 * octalSignificandWide overflows. */
+    int octalSignificandOverflow = 0;
+				/* Flag==1 if octalSignificandBig is used. */
+    int numSigDigs = 0;		/* Number of significant digits in the decimal
+				 * significand. */
+    int numTrailZeros = 0;	/* Number of trailing zeroes at the current
+				 * point in the parse. */
+    int numDigitsAfterDp = 0;	/* Number of digits scanned after the decimal
+				 * point. */
+    int exponentSignum = 0;	/* Signum of the exponent of a floating point
+				 * number. */
+    long exponent = 0;		/* Exponent of a floating point number. */
+    const char *p;		/* Pointer to next character to scan. */
+    size_t len;			/* Number of characters remaining after p. */
+    const char *acceptPoint;	/* Pointer to position after last character in
+				 * an acceptable number. */
+    size_t acceptLen;		/* Number of characters following that
+				 * point. */
+    int status = TCL_OK;	/* Status to return to caller. */
+    char d = 0;			/* Last hexadecimal digit scanned; initialized
+				 * to avoid a compiler warning. */
+    int shift = 0;		/* Amount to shift when accumulating binary */
+    int explicitOctal = 0;
+
+#define ALL_BITS	(~(Tcl_WideUInt)0)
+#define MOST_BITS	(ALL_BITS >> 1)
+
+    /*
+     * Initialize bytes to start of the object's string rep if the caller
+     * didn't pass anything else.
+     */
+
+    if (bytes == NULL) {
+	if (interp == NULL && endPtrPtr == NULL) {
+	    if (objPtr->typePtr == &tclDictType) {
+		/* A dict can never be a (single) number */
+		return TCL_ERROR;
+	    }
+	    if (objPtr->typePtr == &tclListType) {
+		int length;
+		/* A list can only be a (single) number if its length == 1 */
+		TclListObjLength(NULL, objPtr, &length);
+		if (length != 1) {
+		    return TCL_ERROR;
+		}
+	    }
+	}
+	bytes = TclGetString(objPtr);
+    }
+
+    p = bytes;
+    len = numBytes;
+    acceptPoint = p;
+    acceptLen = len;
+    while (1) {
+	char c = len ? *p : '\0';
+	switch (state) {
+
+	case INITIAL:
+	    /*
+	     * Initial state. Acceptable characters are +, -, digits, period,
+	     * I, N, and whitespace.
+	     */
+
+	    if (TclIsSpaceProc(c)) {
+		if (flags & TCL_PARSE_NO_WHITESPACE) {
+		    goto endgame;
+		}
+		break;
+	    } else if (c == '+') {
+		state = SIGNUM;
+		break;
+	    } else if (c == '-') {
+		signum = 1;
+		state = SIGNUM;
+		break;
+	    }
+	    /* FALLTHROUGH */
+
+	case SIGNUM:
+	    /*
+	     * Scanned a leading + or -. Acceptable characters are digits,
+	     * period, I, and N.
+	     */
+
+	    if (c == '0') {
+		if (flags & TCL_PARSE_DECIMAL_ONLY) {
+		    state = DECIMAL;
+		} else {
+		    state = ZERO;
+		}
+		break;
+	    } else if (flags & TCL_PARSE_HEXADECIMAL_ONLY) {
+		goto zerox;
+	    } else if (flags & TCL_PARSE_BINARY_ONLY) {
+		goto zerob;
+	    } else if (flags & TCL_PARSE_OCTAL_ONLY) {
+		goto zeroo;
+	    } else if (isdigit(UCHAR(c))) {
+		significandWide = c - '0';
+		numSigDigs = 1;
+		state = DECIMAL;
+		break;
+	    } else if (flags & TCL_PARSE_INTEGER_ONLY) {
+		goto endgame;
+	    } else if (c == '.') {
+		state = LEADING_RADIX_POINT;
+		break;
+	    } else if (c == 'I' || c == 'i') {
+		state = sI;
+		break;
+#ifdef IEEE_FLOATING_POINT
+	    } else if (c == 'N' || c == 'n') {
+		state = sN;
+		break;
+#endif
+	    }
+	    goto endgame;
+
+	case ZERO:
+	    /*
+	     * Scanned a leading zero (perhaps with a + or -). Acceptable
+	     * inputs are digits, period, X, b, and E. If 8 or 9 is
+	     * encountered, the number can't be octal. This state and the
+	     * OCTAL state differ only in whether they recognize 'X' and 'b'.
+	     */
+
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (c == 'x' || c == 'X') {
+		if (flags & (TCL_PARSE_OCTAL_ONLY|TCL_PARSE_BINARY_ONLY)) {
+		    goto endgame;
+		}
+		state = ZERO_X;
+		break;
+	    }
+	    if (flags & TCL_PARSE_HEXADECIMAL_ONLY) {
+		goto zerox;
+	    }
+	    if (flags & TCL_PARSE_SCAN_PREFIXES) {
+		goto zeroo;
+	    }
+	    if (c == 'b' || c == 'B') {
+		if (flags & TCL_PARSE_OCTAL_ONLY) {
+		    goto endgame;
+		}
+		state = ZERO_B;
+		break;
+	    }
+	    if (flags & TCL_PARSE_BINARY_ONLY) {
+		goto zerob;
+	    }
+	    if (c == 'o' || c == 'O') {
+		explicitOctal = 1;
+		state = ZERO_O;
+		break;
+	    }
+	    if (c == 'd' || c == 'D') {
+		state = ZERO_D;
+		break;
+	    }
+#ifdef TCL_NO_DEPRECATED
+	    goto decimal;
+#endif
+	    /* FALLTHROUGH */
+
+	case OCTAL:
+	    /*
+	     * Scanned an optional + or -, followed by a string of octal
+	     * digits. Acceptable inputs are more digits, period, or E. If 8
+	     * or 9 is encountered, commit to floating point.
+	     */
+
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    /* FALLTHROUGH */
+	case ZERO_O:
+	zeroo:
+	    if (c == '0') {
+		numTrailZeros++;
+		state = OCTAL;
+		break;
+	    } else if (c >= '1' && c <= '7') {
+		if (objPtr != NULL) {
+		    shift = 3 * (numTrailZeros + 1);
+		    significandOverflow = AccumulateDecimalDigit(
+			    (unsigned)(c-'0'), numTrailZeros,
+			    &significandWide, &significandBig,
+			    significandOverflow);
+
+		    if (!octalSignificandOverflow) {
+			/*
+			 * Shifting by more bits than are in the value being
+			 * shifted is at least de facto nonportable. Check for
+			 * too large shifts first.
+			 */
+
+			if ((octalSignificandWide != 0)
+				&& (((size_t)shift >=
+					CHAR_BIT*sizeof(Tcl_WideUInt))
+				|| (octalSignificandWide >
+					(~(Tcl_WideUInt)0 >> shift)))) {
+			    octalSignificandOverflow = 1;
+			    TclBNInitBignumFromWideUInt(&octalSignificandBig,
+				    octalSignificandWide);
+			}
+		    }
+		    if (!octalSignificandOverflow) {
+			octalSignificandWide =
+				(octalSignificandWide << shift) + (c - '0');
+		    } else {
+			mp_mul_2d(&octalSignificandBig, shift,
+				&octalSignificandBig);
+			mp_add_d(&octalSignificandBig, (mp_digit)(c - '0'),
+				&octalSignificandBig);
+		    }
+		}
+		if (numSigDigs != 0) {
+		    numSigDigs += numTrailZeros+1;
+		} else {
+		    numSigDigs = 1;
+		}
+		numTrailZeros = 0;
+		state = OCTAL;
+		break;
+	    }
+	    /* FALLTHROUGH */
+
+	case BAD_OCTAL:
+	    if (explicitOctal) {
+		/*
+		 * No forgiveness for bad digits in explicitly octal numbers.
+		 */
+
+		goto endgame;
+	    }
+	    if (flags & TCL_PARSE_INTEGER_ONLY) {
+		/*
+		 * No seeking floating point when parsing only integer.
+		 */
+
+		goto endgame;
+	    }
+#ifndef TCL_NO_DEPRECATED
+
+	    /*
+	     * Scanned a number with a leading zero that contains an 8, 9,
+	     * radix point or E. This is an invalid octal number, but might
+	     * still be floating point.
+	     */
+
+	    if (c == '0') {
+		numTrailZeros++;
+		state = BAD_OCTAL;
+		break;
+	    } else if (isdigit(UCHAR(c))) {
+		if (objPtr != NULL) {
+		    significandOverflow = AccumulateDecimalDigit(
+			    (unsigned)(c-'0'), numTrailZeros,
+			    &significandWide, &significandBig,
+			    significandOverflow);
+		}
+		if (numSigDigs != 0) {
+		    numSigDigs += (numTrailZeros + 1);
+		} else {
+		    numSigDigs = 1;
+		}
+		numTrailZeros = 0;
+		state = BAD_OCTAL;
+		break;
+	    } else if (c == '.') {
+		state = FRACTION;
+		break;
+	    } else if (c == 'E' || c == 'e') {
+		state = EXPONENT_START;
+		break;
+	    }
+#endif
+	    goto endgame;
+
+	    /*
+	     * Scanned 0x. If state is HEXADECIMAL, scanned at least one
+	     * character following the 0x. The only acceptable inputs are
+	     * hexadecimal digits.
+	     */
+
+	case HEXADECIMAL:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    /* FALLTHROUGH */
+
+	case ZERO_X:
+	zerox:
+	    if (c == '0') {
+		numTrailZeros++;
+		state = HEXADECIMAL;
+		break;
+	    } else if (isdigit(UCHAR(c))) {
+		d = (c-'0');
+	    } else if (c >= 'A' && c <= 'F') {
+		d = (c-'A'+10);
+	    } else if (c >= 'a' && c <= 'f') {
+		d = (c-'a'+10);
+	    } else {
+		goto endgame;
+	    }
+	    if (objPtr != NULL) {
+		shift = 4 * (numTrailZeros + 1);
+		if (!significandOverflow) {
+		    /*
+		     * Shifting by more bits than are in the value being
+		     * shifted is at least de facto nonportable. Check for too
+		     * large shifts first.
+		     */
+
+		    if (significandWide != 0 &&
+			    ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideUInt) ||
+			    significandWide > (~(Tcl_WideUInt)0 >> shift))) {
+			significandOverflow = 1;
+			TclBNInitBignumFromWideUInt(&significandBig,
+				significandWide);
+		    }
+		}
+		if (!significandOverflow) {
+		    significandWide = (significandWide << shift) + d;
+		} else {
+		    mp_mul_2d(&significandBig, shift, &significandBig);
+		    mp_add_d(&significandBig, (mp_digit) d, &significandBig);
+		}
+	    }
+	    numTrailZeros = 0;
+	    state = HEXADECIMAL;
+	    break;
+
+	case BINARY:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	case ZERO_B:
+	zerob:
+	    if (c == '0') {
+		numTrailZeros++;
+		state = BINARY;
+		break;
+	    } else if (c != '1') {
+		goto endgame;
+	    }
+	    if (objPtr != NULL) {
+		shift = numTrailZeros + 1;
+		if (!significandOverflow) {
+		    /*
+		     * Shifting by more bits than are in the value being
+		     * shifted is at least de facto nonportable. Check for too
+		     * large shifts first.
+		     */
+
+		    if (significandWide != 0 &&
+			    ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideUInt) ||
+			    significandWide > (~(Tcl_WideUInt)0 >> shift))) {
+			significandOverflow = 1;
+			TclBNInitBignumFromWideUInt(&significandBig,
+				significandWide);
+		    }
+		}
+		if (!significandOverflow) {
+		    significandWide = (significandWide << shift) + 1;
+		} else {
+		    mp_mul_2d(&significandBig, shift, &significandBig);
+		    mp_add_d(&significandBig, (mp_digit) 1, &significandBig);
+		}
+	    }
+	    numTrailZeros = 0;
+	    state = BINARY;
+	    break;
+
+	case ZERO_D:
+	    if (c == '0') {
+		numTrailZeros++;
+	    } else if ( ! isdigit(UCHAR(c))) {
+		goto endgame;
+	    }
+	    state = DECIMAL;
+	    flags |= TCL_PARSE_INTEGER_ONLY;
+	    /* FALLTHROUGH */
+
+	case DECIMAL:
+	    /*
+	     * Scanned an optional + or - followed by a string of decimal
+	     * digits.
+	     */
+
+#ifdef TCL_NO_DEPRECATED
+	decimal:
+#endif
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (c == '0') {
+		numTrailZeros++;
+		state = DECIMAL;
+		break;
+	    } else if (isdigit(UCHAR(c))) {
+		if (objPtr != NULL) {
+		    significandOverflow = AccumulateDecimalDigit(
+			    (unsigned)(c - '0'), numTrailZeros,
+			    &significandWide, &significandBig,
+			    significandOverflow);
+		}
+		numSigDigs += numTrailZeros+1;
+		numTrailZeros = 0;
+		state = DECIMAL;
+		break;
+	    } else if (flags & TCL_PARSE_INTEGER_ONLY) {
+		goto endgame;
+	    } else if (c == '.') {
+		state = FRACTION;
+		break;
+	    } else if (c == 'E' || c == 'e') {
+		state = EXPONENT_START;
+		break;
+	    }
+	    goto endgame;
+
+	    /*
+	     * Found a decimal point. If no digits have yet been scanned, E is
+	     * not allowed; otherwise, it introduces the exponent. If at least
+	     * one digit has been found, we have a possible complete number.
+	     */
+
+	case FRACTION:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (c == 'E' || c=='e') {
+		state = EXPONENT_START;
+		break;
+	    }
+	    /* FALLTHROUGH */
+
+	case LEADING_RADIX_POINT:
+	    if (c == '0') {
+		numDigitsAfterDp++;
+		numTrailZeros++;
+		state = FRACTION;
+		break;
+	    } else if (isdigit(UCHAR(c))) {
+		numDigitsAfterDp++;
+		if (objPtr != NULL) {
+		    significandOverflow = AccumulateDecimalDigit(
+			    (unsigned)(c-'0'), numTrailZeros,
+			    &significandWide, &significandBig,
+			    significandOverflow);
+		}
+		if (numSigDigs != 0) {
+		    numSigDigs += numTrailZeros+1;
+		} else {
+		    numSigDigs = 1;
+		}
+		numTrailZeros = 0;
+		state = FRACTION;
+		break;
+	    }
+	    goto endgame;
+
+	case EXPONENT_START:
+	    /*
+	     * Scanned the E at the start of an exponent. Make sure a legal
+	     * character follows before using the C library strtol routine,
+	     * which allows whitespace.
+	     */
+
+	    if (c == '+') {
+		state = EXPONENT_SIGNUM;
+		break;
+	    } else if (c == '-') {
+		exponentSignum = 1;
+		state = EXPONENT_SIGNUM;
+		break;
+	    }
+	    /* FALLTHROUGH */
+
+	case EXPONENT_SIGNUM:
+	    /*
+	     * Found the E at the start of the exponent, followed by a sign
+	     * character.
+	     */
+
+	    if (isdigit(UCHAR(c))) {
+		exponent = c - '0';
+		state = EXPONENT;
+		break;
+	    }
+	    goto endgame;
+
+	case EXPONENT:
+	    /*
+	     * Found an exponent with at least one digit. Accumulate it,
+	     * making sure to hard-pin it to LONG_MAX on overflow.
+	     */
+
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (isdigit(UCHAR(c))) {
+		if (exponent < (LONG_MAX - 9) / 10) {
+		    exponent = 10 * exponent + (c - '0');
+		} else {
+		    exponent = LONG_MAX;
+		}
+		state = EXPONENT;
+		break;
+	    }
+	    goto endgame;
+
+	    /*
+	     * Parse out INFINITY by simply spelling it out. INF is accepted
+	     * as an abbreviation; other prefices are not.
+	     */
+
+	case sI:
+	    if (c == 'n' || c == 'N') {
+		state = sIN;
+		break;
+	    }
+	    goto endgame;
+	case sIN:
+	    if (c == 'f' || c == 'F') {
+		state = sINF;
+		break;
+	    }
+	    goto endgame;
+	case sINF:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (c == 'i' || c == 'I') {
+		state = sINFI;
+		break;
+	    }
+	    goto endgame;
+	case sINFI:
+	    if (c == 'n' || c == 'N') {
+		state = sINFIN;
+		break;
+	    }
+	    goto endgame;
+	case sINFIN:
+	    if (c == 'i' || c == 'I') {
+		state = sINFINI;
+		break;
+	    }
+	    goto endgame;
+	case sINFINI:
+	    if (c == 't' || c == 'T') {
+		state = sINFINIT;
+		break;
+	    }
+	    goto endgame;
+	case sINFINIT:
+	    if (c == 'y' || c == 'Y') {
+		state = sINFINITY;
+		break;
+	    }
+	    goto endgame;
+
+	    /*
+	     * Parse NaN's.
+	     */
+#ifdef IEEE_FLOATING_POINT
+	case sN:
+	    if (c == 'a' || c == 'A') {
+		state = sNA;
+		break;
+	    }
+	    goto endgame;
+	case sNA:
+	    if (c == 'n' || c == 'N') {
+		state = sNAN;
+		break;
+	    }
+	    goto endgame;
+	case sNAN:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    if (c == '(') {
+		state = sNANPAREN;
+		break;
+	    }
+	    goto endgame;
+
+	    /*
+	     * Parse NaN(hexdigits)
+	     */
+	case sNANHEX:
+	    if (c == ')') {
+		state = sNANFINISH;
+		break;
+	    }
+	    /* FALLTHROUGH */
+	case sNANPAREN:
+	    if (TclIsSpaceProc(c)) {
+		break;
+	    }
+	    if (numSigDigs < 13) {
+		if (c >= '0' && c <= '9') {
+		    d = c - '0';
+		} else if (c >= 'a' && c <= 'f') {
+		    d = 10 + c - 'a';
+		} else if (c >= 'A' && c <= 'F') {
+		    d = 10 + c - 'A';
+		} else {
+		    goto endgame;
+		}
+		numSigDigs++;
+		significandWide = (significandWide << 4) + d;
+		state = sNANHEX;
+		break;
+	    }
+	    goto endgame;
+	case sNANFINISH:
+#endif
+
+	case sINFINITY:
+	    acceptState = state;
+	    acceptPoint = p;
+	    acceptLen = len;
+	    goto endgame;
+	}
+	p++;
+	len--;
+    }
+
+  endgame:
+    if (acceptState == INITIAL) {
+	/*
+	 * No numeric string at all found.
+	 */
+
+	status = TCL_ERROR;
+	if (endPtrPtr != NULL) {
+	    *endPtrPtr = p;
+	}
+    } else {
+	/*
+	 * Back up to the last accepting state in the lexer.
+	 */
+
+	p = acceptPoint;
+	len = acceptLen;
+	if (!(flags & TCL_PARSE_NO_WHITESPACE)) {
+	    /*
+	     * Accept trailing whitespace.
+	     */
+
+	    while (len != 0 && TclIsSpaceProc(*p)) {
+		p++;
+		len--;
+	    }
+	}
+	if (endPtrPtr == NULL) {
+	    if ((len != 0) && ((numBytes > 0) || (*p != '\0'))) {
+		status = TCL_ERROR;
+	    }
+	} else {
+	    *endPtrPtr = p;
+	}
+    }
+
+    /*
+     * Generate and store the appropriate internal rep.
+     */
+
+    if (status == TCL_OK && objPtr != NULL) {
+	TclFreeIntRep(objPtr);
+	switch (acceptState) {
+	case SIGNUM:
+	case BAD_OCTAL:
+	case ZERO_X:
+	case ZERO_O:
+	case ZERO_B:
+	case ZERO_D:
+	case LEADING_RADIX_POINT:
+	case EXPONENT_START:
+	case EXPONENT_SIGNUM:
+	case sI:
+	case sIN:
+	case sINFI:
+	case sINFIN:
+	case sINFINI:
+	case sINFINIT:
+#ifdef IEEE_FLOATING_POINT
+	case sN:
+	case sNA:
+	case sNANPAREN:
+	case sNANHEX:
+#endif
+	    Tcl_Panic("TclParseNumber: bad acceptState %d parsing '%s'",
+		    acceptState, bytes);
+	case BINARY:
+	    shift = numTrailZeros;
+	    if (!significandOverflow && significandWide != 0 &&
+		    ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideUInt) ||
+		    significandWide > (MOST_BITS + signum) >> shift)) {
+		significandOverflow = 1;
+		TclBNInitBignumFromWideUInt(&significandBig, significandWide);
+	    }
+	    if (shift) {
+		if (!significandOverflow) {
+		    significandWide <<= shift;
+		} else {
+		    mp_mul_2d(&significandBig, shift, &significandBig);
+		}
+	    }
+	    goto returnInteger;
+
+	case HEXADECIMAL:
+	    /*
+	     * Returning a hex integer. Final scaling step.
+	     */
+
+	    shift = 4 * numTrailZeros;
+	    if (!significandOverflow && significandWide !=0 &&
+		    ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideUInt) ||
+		    significandWide > (MOST_BITS + signum) >> shift)) {
+		significandOverflow = 1;
+		TclBNInitBignumFromWideUInt(&significandBig, significandWide);
+	    }
+	    if (shift) {
+		if (!significandOverflow) {
+		    significandWide <<= shift;
+		} else {
+		    mp_mul_2d(&significandBig, shift, &significandBig);
+		}
+	    }
+	    goto returnInteger;
+
+	case OCTAL:
+	    /*
+	     * Returning an octal integer. Final scaling step.
+	     */
+
+	    shift = 3 * numTrailZeros;
+	    if (!octalSignificandOverflow && octalSignificandWide != 0 &&
+		    ((size_t)shift >= CHAR_BIT*sizeof(Tcl_WideUInt) ||
+		    octalSignificandWide > (MOST_BITS + signum) >> shift)) {
+		octalSignificandOverflow = 1;
+		TclBNInitBignumFromWideUInt(&octalSignificandBig,
+			octalSignificandWide);
+	    }
+	    if (shift) {
+		if (!octalSignificandOverflow) {
+		    octalSignificandWide <<= shift;
+		} else {
+		    mp_mul_2d(&octalSignificandBig, shift,
+			    &octalSignificandBig);
+		}
+	    }
+	    if (!octalSignificandOverflow) {
+		if (octalSignificandWide >
+			(Tcl_WideUInt)(((~(unsigned long)0) >> 1) + signum)) {
+#ifndef TCL_WIDE_INT_IS_LONG
+		    if (octalSignificandWide <= (MOST_BITS + signum)) {
+			objPtr->typePtr = &tclWideIntType;
+			if (signum) {
+			    objPtr->internalRep.wideValue =
+				    - (Tcl_WideInt) octalSignificandWide;
+			} else {
+			    objPtr->internalRep.wideValue =
+				    (Tcl_WideInt) octalSignificandWide;
+			}
+			break;
+		    }
+#endif
+		    TclBNInitBignumFromWideUInt(&octalSignificandBig,
+			    octalSignificandWide);
+		    octalSignificandOverflow = 1;
+		} else {
+		    objPtr->typePtr = &tclIntType;
+		    if (signum) {
+			objPtr->internalRep.longValue =
+				- (long) octalSignificandWide;
+		    } else {
+			objPtr->internalRep.longValue =
+				(long) octalSignificandWide;
+		    }
+		}
+	    }
+	    if (octalSignificandOverflow) {
+		if (signum) {
+		    mp_neg(&octalSignificandBig, &octalSignificandBig);
+		}
+		TclSetBignumIntRep(objPtr, &octalSignificandBig);
+	    }
+	    break;
+
+	case ZERO:
+	case DECIMAL:
+	    significandOverflow = AccumulateDecimalDigit(0, numTrailZeros-1,
+		    &significandWide, &significandBig, significandOverflow);
+	    if (!significandOverflow && (significandWide > MOST_BITS+signum)){
+		significandOverflow = 1;
+		TclBNInitBignumFromWideUInt(&significandBig, significandWide);
+	    }
+	returnInteger:
+	    if (!significandOverflow) {
+		if (significandWide >
+			(Tcl_WideUInt)(((~(unsigned long)0) >> 1) + signum)) {
+#ifndef TCL_WIDE_INT_IS_LONG
+		    if (significandWide <= MOST_BITS+signum) {
+			objPtr->typePtr = &tclWideIntType;
+			if (signum) {
+			    objPtr->internalRep.wideValue =
+				    - (Tcl_WideInt) significandWide;
+			} else {
+			    objPtr->internalRep.wideValue =
+				    (Tcl_WideInt) significandWide;
+			}
+			break;
+		    }
+#endif
+		    TclBNInitBignumFromWideUInt(&significandBig,
+			    significandWide);
+		    significandOverflow = 1;
+		} else {
+		    objPtr->typePtr = &tclIntType;
+		    if (signum) {
+			objPtr->internalRep.longValue =
+				- (long) significandWide;
+		    } else {
+			objPtr->internalRep.longValue =
+				(long) significandWide;
+		    }
+		}
+	    }
+	    if (significandOverflow) {
+		if (signum) {
+		    mp_neg(&significandBig, &significandBig);
+		}
+		TclSetBignumIntRep(objPtr, &significandBig);
+	    }
+	    break;
+
+	case FRACTION:
+	case EXPONENT:
+
+	    /*
+	     * Here, we're parsing a floating-point number. 'significandWide'
+	     * or 'significandBig' contains the exact significand, according
+	     * to whether 'significandOverflow' is set. The desired floating
+	     * point value is significand * 10**k, where
+	     * k = numTrailZeros+exponent-numDigitsAfterDp.
+	     */
+
+	    objPtr->typePtr = &tclDoubleType;
+	    if (exponentSignum) {
+		exponent = -exponent;
+	    }
+	    if (!significandOverflow) {
+		objPtr->internalRep.doubleValue = MakeLowPrecisionDouble(
+			signum, significandWide, numSigDigs,
+			numTrailZeros + exponent - numDigitsAfterDp);
+	    } else {
+		objPtr->internalRep.doubleValue = MakeHighPrecisionDouble(
+			signum, &significandBig, numSigDigs,
+			numTrailZeros + exponent - numDigitsAfterDp);
+	    }
+	    break;
+
+	case sINF:
+	case sINFINITY:
+	    if (signum) {
+		objPtr->internalRep.doubleValue = -HUGE_VAL;
+	    } else {
+		objPtr->internalRep.doubleValue = HUGE_VAL;
+	    }
+	    objPtr->typePtr = &tclDoubleType;
+	    break;
+
+#ifdef IEEE_FLOATING_POINT
+	case sNAN:
+	case sNANFINISH:
+	    objPtr->internalRep.doubleValue = MakeNaN(signum,significandWide);
+	    objPtr->typePtr = &tclDoubleType;
+	    break;
+#endif
+	case INITIAL:
+	    /* This case only to silence compiler warning. */
+	    Tcl_Panic("TclParseNumber: state INITIAL can't happen here");
+	}
+    }
+
+    /*
+     * Format an error message when an invalid number is encountered.
+     */
+
+    if (status != TCL_OK) {
+	if (interp != NULL) {
+	    Tcl_Obj *msg = Tcl_ObjPrintf("expected %s but got \"",
+		    expected);
+
+	    Tcl_AppendLimitedToObj(msg, bytes, numBytes, 50, "");
+	    Tcl_AppendToObj(msg, "\"", -1);
+	    if (state == BAD_OCTAL) {
+		Tcl_AppendToObj(msg, " (looks like invalid octal number)", -1);
+	    }
+	    Tcl_SetObjResult(interp, msg);
+	    Tcl_SetErrorCode(interp, "TCL", "VALUE", "NUMBER", NULL);
+	}
+    }
+
+    /*
+     * Free memory.
+     */
+
+    if (octalSignificandOverflow) {
+	mp_clear(&octalSignificandBig);
+    }
+    if (significandOverflow) {
+	mp_clear(&significandBig);
+    }
+    return status;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AccumulateDecimalDigit --
+ *
+ *	Consume a decimal digit in a number being scanned.
+ *
+ * Results:
+ *	Returns 1 if the number has overflowed to a bignum, 0 if it still fits
+ *	in a wide integer.
+ *
+ * Side effects:
+ *	Updates either the wide or bignum representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AccumulateDecimalDigit(
+    unsigned digit,		/* Digit being scanned. */
+    int numZeros,		/* Count of zero digits preceding the digit
+				 * being scanned. */
+    Tcl_WideUInt *wideRepPtr,	/* Representation of the partial number as a
+				 * wide integer. */
+    mp_int *bignumRepPtr,	/* Representation of the partial number as a
+				 * bignum. */
+    int bignumFlag)		/* Flag == 1 if the number overflowed previous
+				 * to this digit. */
+{
+    int i, n;
+    Tcl_WideUInt w;
+
+    /*
+     * Try wide multiplication first.
+     */
+
+    if (!bignumFlag) {
+	w = *wideRepPtr;
+	if (w == 0) {
+	    /*
+	     * There's no need to multiply if the multiplicand is zero.
+	     */
+
+	    *wideRepPtr = digit;
+	    return 0;
+	} else if (numZeros >= maxpow10_wide
+		|| w > ((~(Tcl_WideUInt)0)-digit)/pow10_wide[numZeros+1]) {
+	    /*
+	     * Wide multiplication will overflow.  Expand the number to a
+	     * bignum and fall through into the bignum case.
+	     */
+
+	    TclBNInitBignumFromWideUInt(bignumRepPtr, w);
+	} else {
+	    /*
+	     * Wide multiplication.
+	     */
+
+	    *wideRepPtr = w * pow10_wide[numZeros+1] + digit;
+	    return 0;
+	}
+    }
+
+    /*
+     * Bignum multiplication.
+     */
+
+    if (numZeros < log10_DIGIT_MAX) {
+	/*
+	 * Up to about 8 zeros - single digit multiplication.
+	 */
+
+	mp_mul_d(bignumRepPtr, (mp_digit) pow10_wide[numZeros+1],
+		bignumRepPtr);
+	mp_add_d(bignumRepPtr, (mp_digit) digit, bignumRepPtr);
+    } else {
+	/*
+	 * More than single digit multiplication. Multiply by the appropriate
+	 * small powers of 5, and then shift. Large strings of zeroes are
+	 * eaten 256 at a time; this is less efficient than it could be, but
+	 * seems implausible. We presume that DIGIT_BIT is at least 27. The
+	 * first multiplication, by up to 10**7, is done with a one-DIGIT
+	 * multiply (this presumes that DIGIT_BIT >= 24).
+	 */
+
+	n = numZeros + 1;
+	mp_mul_d(bignumRepPtr, (mp_digit) pow10_wide[n&0x7], bignumRepPtr);
+	for (i=3; i<=7; ++i) {
+	    if (n & (1 << i)) {
+		mp_mul(bignumRepPtr, pow5+i, bignumRepPtr);
+	    }
+	}
+	while (n >= 256) {
+	    mp_mul(bignumRepPtr, pow5+8, bignumRepPtr);
+	    n -= 256;
+	}
+	mp_mul_2d(bignumRepPtr, (int)(numZeros+1)&~0x7, bignumRepPtr);
+	mp_add_d(bignumRepPtr, (mp_digit) digit, bignumRepPtr);
+    }
+
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeLowPrecisionDouble --
+ *
+ *	Makes the double precision number, signum*significand*10**exponent.
+ *
+ * Results:
+ *	Returns the constructed number.
+ *
+ *	Common cases, where there are few enough digits that the number can be
+ *	represented with at most roundoff, are handled specially here. If the
+ *	number requires more than one rounded operation to compute, the code
+ *	promotes the significand to a bignum and calls MakeHighPrecisionDouble
+ *	to do it instead.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+MakeLowPrecisionDouble(
+    int signum,			/* 1 if the number is negative, 0 otherwise */
+    Tcl_WideUInt significand,	/* Significand of the number. */
+    int numSigDigs,		/* Number of digits in the significand. */
+    int exponent)		/* Power of ten. */
+{
+    double retval;		/* Value of the number. */
+    mp_int significandBig;	/* Significand expressed as a bignum. */
+
+    /*
+     * With gcc on x86, the floating point rounding mode is double-extended.
+     * This causes the result of double-precision calculations to be rounded
+     * twice: once to the precision of double-extended and then again to the
+     * precision of double. Double-rounding introduces gratuitous errors of 1
+     * ulp, so we need to change rounding mode to 53-bits.
+     */
+
+    TCL_IEEE_DOUBLE_ROUNDING;
+
+    /*
+     * Test for the easy cases.
+     */
+
+    if (numSigDigs <= QUICK_MAX) {
+	if (exponent >= 0) {
+	    if (exponent <= mmaxpow) {
+		/*
+		 * The significand is an exact integer, and so is
+		 * 10**exponent. The product will be correct to within 1/2 ulp
+		 * without special handling.
+		 */
+
+		retval = (double)
+			((Tcl_WideInt)significand * pow10vals[exponent]);
+		goto returnValue;
+	    } else {
+		int diff = QUICK_MAX - numSigDigs;
+
+		if (exponent-diff <= mmaxpow) {
+		    /*
+		     * 10**exponent is not an exact integer, but
+		     * 10**(exponent-diff) is exact, and so is
+		     * significand*10**diff, so we can still compute the value
+		     * with only one roundoff.
+		     */
+
+		    volatile double factor = (double)
+			    ((Tcl_WideInt)significand * pow10vals[diff]);
+		    retval = factor * pow10vals[exponent-diff];
+		    goto returnValue;
+		}
+	    }
+	} else {
+	    if (exponent >= -mmaxpow) {
+		/*
+		 * 10**-exponent is an exact integer, and so is the
+		 * significand. Compute the result by one division, again with
+		 * only one rounding.
+		 */
+
+		retval = (double)
+			((Tcl_WideInt)significand / pow10vals[-exponent]);
+		goto returnValue;
+	    }
+	}
+    }
+
+    /*
+     * All the easy cases have failed. Promote ths significand to bignum and
+     * call MakeHighPrecisionDouble to do it the hard way.
+     */
+
+    TclBNInitBignumFromWideUInt(&significandBig, significand);
+    retval = MakeHighPrecisionDouble(0, &significandBig, numSigDigs,
+	    exponent);
+    mp_clear(&significandBig);
+
+    /*
+     * Come here to return the computed value.
+     */
+
+  returnValue:
+    if (signum) {
+	retval = -retval;
+    }
+
+    /*
+     * On gcc on x86, restore the floating point mode word.
+     */
+
+    TCL_DEFAULT_DOUBLE_ROUNDING;
+
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeHighPrecisionDouble --
+ *
+ *	Makes the double precision number, signum*significand*10**exponent.
+ *
+ * Results:
+ *	Returns the constructed number.
+ *
+ *	MakeHighPrecisionDouble is used when arbitrary-precision arithmetic is
+ *	needed to ensure correct rounding. It begins by calculating a
+ *	low-precision approximation to the desired number, and then refines
+ *	the answer in high precision.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+MakeHighPrecisionDouble(
+    int signum,			/* 1=negative, 0=nonnegative. */
+    mp_int *significand,	/* Exact significand of the number. */
+    int numSigDigs,		/* Number of significant digits. */
+    int exponent)		/* Power of 10 by which to multiply. */
+{
+    double retval;
+    int machexp;		/* Machine exponent of a power of 10. */
+
+    /*
+     * With gcc on x86, the floating point rounding mode is double-extended.
+     * This causes the result of double-precision calculations to be rounded
+     * twice: once to the precision of double-extended and then again to the
+     * precision of double. Double-rounding introduces gratuitous errors of 1
+     * ulp, so we need to change rounding mode to 53-bits.
+     */
+
+    TCL_IEEE_DOUBLE_ROUNDING;
+
+    /*
+     * Quick checks for over/underflow.
+     */
+
+    if (numSigDigs+exponent-1 > maxDigits) {
+	retval = HUGE_VAL;
+	goto returnValue;
+    }
+    if (numSigDigs+exponent-1 < minDigits) {
+	retval = 0;
+	goto returnValue;
+    }
+
+    /*
+     * Develop a first approximation to the significand. It is tempting simply
+     * to force bignum to double, but that will overflow on input numbers like
+     * 1.[string repeat 0 1000]1; while this is a not terribly likely
+     * scenario, we still have to deal with it. Use fraction and exponent
+     * instead. Once we have the significand, multiply by 10**exponent. Test
+     * for overflow. Convert back to a double, and test for underflow.
+     */
+
+    retval = BignumToBiasedFrExp(significand, &machexp);
+    retval = Pow10TimesFrExp(exponent, retval, &machexp);
+    if (machexp > DBL_MAX_EXP*log2FLT_RADIX) {
+	retval = HUGE_VAL;
+	goto returnValue;
+    }
+    retval = SafeLdExp(retval, machexp);
+	if (tiny == 0.0) {
+	    tiny = SafeLdExp(1.0, DBL_MIN_EXP * log2FLT_RADIX - mantBits);
+	}
+    if (retval < tiny) {
+	retval = tiny;
+    }
+
+    /*
+     * Refine the result twice. (The second refinement should be necessary
+     * only if the best approximation is a power of 2 minus 1/2 ulp).
+     */
+
+    retval = RefineApproximation(retval, significand, exponent);
+    retval = RefineApproximation(retval, significand, exponent);
+
+    /*
+     * Come here to return the computed value.
+     */
+
+  returnValue:
+    if (signum) {
+	retval = -retval;
+    }
+
+    /*
+     * On gcc on x86, restore the floating point mode word.
+     */
+
+    TCL_DEFAULT_DOUBLE_ROUNDING;
+
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MakeNaN --
+ *
+ *	Makes a "Not a Number" given a set of bits to put in the tag bits
+ *
+ *	Note that a signalling NaN is never returned.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef IEEE_FLOATING_POINT
+static double
+MakeNaN(
+    int signum,			/* Sign bit (1=negative, 0=nonnegative. */
+    Tcl_WideUInt tags)		/* Tag bits to put in the NaN. */
+{
+    union {
+	Tcl_WideUInt iv;
+	double dv;
+    } theNaN;
+
+    theNaN.iv = tags;
+    theNaN.iv &= (((Tcl_WideUInt) 1) << 51) - 1;
+    if (signum) {
+	theNaN.iv |= ((Tcl_WideUInt) (0x8000 | NAN_START)) << 48;
+    } else {
+	theNaN.iv |= ((Tcl_WideUInt) NAN_START) << 48;
+    }
+    if (n770_fp) {
+	theNaN.iv = Nokia770Twiddle(theNaN.iv);
+    }
+    return theNaN.dv;
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RefineApproximation --
+ *
+ *	Given a poor approximation to a floating point number, returns a
+ *	better one. (The better approximation is correct to within 1 ulp, and
+ *	is entirely correct if the poor approximation is correct to 1 ulp.)
+ *
+ * Results:
+ *	Returns the improved result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+RefineApproximation(
+    double approxResult,	/* Approximate result of conversion. */
+    mp_int *exactSignificand,	/* Integer significand. */
+    int exponent)		/* Power of 10 to multiply by significand. */
+{
+    int M2, M5;			/* Powers of 2 and of 5 needed to put the
+				 * decimal and binary numbers over a common
+				 * denominator. */
+    double significand;		/* Sigificand of the binary number. */
+    int binExponent;		/* Exponent of the binary number. */
+    int msb;			/* Most significant bit position of an
+				 * intermediate result. */
+    int nDigits;		/* Number of mp_digit's in an intermediate
+				 * result. */
+    mp_int twoMv;		/* Approx binary value expressed as an exact
+				 * integer scaled by the multiplier 2M. */
+    mp_int twoMd;		/* Exact decimal value expressed as an exact
+				 * integer scaled by the multiplier 2M. */
+    int scale;			/* Scale factor for M. */
+    int multiplier;		/* Power of two to scale M. */
+    double num, den;		/* Numerator and denominator of the correction
+				 * term. */
+    double quot;		/* Correction term. */
+    double minincr;		/* Lower bound on the absolute value of the
+				 * correction term. */
+    int roundToEven = 0;	/* Flag == TRUE if we need to invoke
+				 * "round to even" functionality */
+    double rteSignificand;	/* Significand of the round-to-even result */
+    int rteExponent;		/* Exponent of the round-to-even result */
+    Tcl_WideInt rteSigWide;	/* Wide integer version of the significand
+				 * for testing evenness */
+    int i;
+
+    /*
+     * The first approximation is always low. If we find that it's HUGE_VAL,
+     * we're done.
+     */
+
+    if (approxResult == HUGE_VAL) {
+	return approxResult;
+    }
+
+    /*
+     * Find a common denominator for the decimal and binary fractions. The
+     * common denominator will be 2**M2 + 5**M5.
+     */
+
+    significand = frexp(approxResult, &binExponent);
+    i = mantBits - binExponent;
+    if (i < 0) {
+	M2 = 0;
+    } else {
+	M2 = i;
+    }
+    if (exponent > 0) {
+	M5 = 0;
+    } else {
+	M5 = -exponent;
+	if (M5 - 1 > M2) {
+	    M2 = M5 - 1;
+	}
+    }
+
+    /*
+     * The floating point number is significand*2**binExponent. Compute the
+     * large integer significand*2**(binExponent+M2+1). The 2**-1 bit of the
+     * significand (the most significant) corresponds to the
+     * 2**(binExponent+M2 + 1) bit of 2*M2*v. Allocate enough digits to hold
+     * that quantity, then convert the significand to a large integer, scaled
+     * appropriately. Then multiply by the appropriate power of 5.
+     */
+
+    msb = binExponent + M2;	/* 1008 */
+    nDigits = msb / DIGIT_BIT + 1;
+    mp_init_size(&twoMv, nDigits);
+    i = (msb % DIGIT_BIT + 1);
+    twoMv.used = nDigits;
+    significand *= SafeLdExp(1.0, i);
+    while (--nDigits >= 0) {
+	twoMv.dp[nDigits] = (mp_digit) significand;
+	significand -= (mp_digit) significand;
+	significand = SafeLdExp(significand, DIGIT_BIT);
+    }
+    for (i = 0; i <= 8; ++i) {
+	if (M5 & (1 << i)) {
+	    mp_mul(&twoMv, pow5+i, &twoMv);
+	}
+    }
+
+    /*
+     * Collect the decimal significand as a high precision integer. The least
+     * significant bit corresponds to bit M2+exponent+1 so it will need to be
+     * shifted left by that many bits after being multiplied by
+     * 5**(M5+exponent).
+     */
+
+    mp_init_copy(&twoMd, exactSignificand);
+    for (i=0; i<=8; ++i) {
+	if ((M5 + exponent) & (1 << i)) {
+	    mp_mul(&twoMd, pow5+i, &twoMd);
+	}
+    }
+    mp_mul_2d(&twoMd, M2+exponent+1, &twoMd);
+    mp_sub(&twoMd, &twoMv, &twoMd);
+
+    /*
+     * The result, 2Mv-2Md, needs to be divided by 2M to yield a correction
+     * term. Because 2M may well overflow a double, we need to scale the
+     * denominator by a factor of 2**binExponent-mantBits.
+     */
+
+    scale = binExponent - mantBits - 1;
+
+    mp_set(&twoMv, 1);
+    for (i=0; i<=8; ++i) {
+	if (M5 & (1 << i)) {
+	    mp_mul(&twoMv, pow5+i, &twoMv);
+	}
+    }
+    multiplier = M2 + scale + 1;
+    if (multiplier > 0) {
+	mp_mul_2d(&twoMv, multiplier, &twoMv);
+    } else if (multiplier < 0) {
+	mp_div_2d(&twoMv, -multiplier, &twoMv, NULL);
+    }
+
+    switch (mp_cmp_mag(&twoMd, &twoMv)) {
+    case MP_LT:
+	/*
+	 * If the result is less than unity, the error is less than 1/2 unit in
+	 * the last place, so there's no correction to make.
+	 */
+	mp_clear(&twoMd);
+	mp_clear(&twoMv);
+	return approxResult;
+    case MP_EQ:
+	/*
+	 * If the result is exactly unity, we need to round to even.
+	 */
+	roundToEven = 1;
+	break;
+    case MP_GT:
+	break;
+    }
+
+    if (roundToEven) {
+	rteSignificand = frexp(approxResult, &rteExponent);
+	rteSigWide = (Tcl_WideInt) ldexp(rteSignificand, FP_PRECISION);
+	if ((rteSigWide & 1) == 0) {
+	    mp_clear(&twoMd);
+	    mp_clear(&twoMv);
+	    return approxResult;
+	}
+    }
+
+    /*
+     * Convert the numerator and denominator of the corrector term accurately
+     * to floating point numbers.
+     */
+
+    num = TclBignumToDouble(&twoMd);
+    den = TclBignumToDouble(&twoMv);
+
+    quot = SafeLdExp(num/den, scale);
+    minincr = SafeLdExp(1.0, binExponent-mantBits);
+
+    if (quot<0. && quot>-minincr) {
+	quot = -minincr;
+    } else if (quot>0. && quot<minincr) {
+	quot = minincr;
+    }
+
+    mp_clear(&twoMd);
+    mp_clear(&twoMv);
+
+    return approxResult + quot;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MultPow5 --
+ *
+ *	Multiply a bignum by a power of 5.
+ *
+ * Side effects:
+ *	Stores base*5**n in result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+MulPow5(
+    mp_int *base, 		/* Number to multiply. */
+    unsigned n,			/* Power of 5 to multiply by. */
+    mp_int *result)		/* Place to store the result. */
+{
+    mp_int *p = base;
+    int n13 = n / 13;
+    int r = n % 13;
+
+    if (r != 0) {
+	mp_mul_d(p, dpow5[r], result);
+	p = result;
+    }
+    r = 0;
+    while (n13 != 0) {
+	if (n13 & 1) {
+	    mp_mul(p, pow5_13+r, result);
+	    p = result;
+	}
+	n13 >>= 1;
+	++r;
+    }
+    if (p != result) {
+	mp_copy(p, result);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NormalizeRightward --
+ *
+ *	Shifts a number rightward until it is odd (that is, until the least
+ *	significant bit is nonzero.
+ *
+ * Results:
+ *	Returns the number of bit positions by which the number was shifted.
+ *
+ * Side effects:
+ *	Shifts the number in place; *wPtr is replaced by the shifted number.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+NormalizeRightward(
+    Tcl_WideUInt *wPtr)		/* INOUT: Number to shift. */
+{
+    int rv = 0;
+    Tcl_WideUInt w = *wPtr;
+
+    if (!(w & (Tcl_WideUInt) 0xffffffff)) {
+	w >>= 32; rv += 32;
+    }
+    if (!(w & (Tcl_WideUInt) 0xffff)) {
+	w >>= 16; rv += 16;
+    }
+    if (!(w & (Tcl_WideUInt) 0xff)) {
+	w >>= 8; rv += 8;
+    }
+    if (!(w & (Tcl_WideUInt) 0xf)) {
+	w >>= 4; rv += 4;
+    }
+    if (!(w & 0x3)) {
+	w >>= 2; rv += 2;
+    }
+    if (!(w & 0x1)) {
+	w >>= 1; ++rv;
+    }
+    *wPtr = w;
+    return rv;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RequiredPrecision --
+ *
+ *	Determines the number of bits needed to hold an intger.
+ *
+ * Results:
+ *	Returns the position of the most significant bit (0 - 63).  Returns 0
+ *	if the number is zero.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+RequiredPrecision(
+    Tcl_WideUInt w)		/* Number to interrogate. */
+{
+    int rv;
+    unsigned long wi;
+
+    if (w & ((Tcl_WideUInt) 0xffffffff << 32)) {
+	wi = (unsigned long) (w >> 32); rv = 32;
+    } else {
+	wi = (unsigned long) w; rv = 0;
+    }
+    if (wi & 0xffff0000) {
+	wi >>= 16; rv += 16;
+    }
+    if (wi & 0xff00) {
+	wi >>= 8; rv += 8;
+    }
+    if (wi & 0xf0) {
+	wi >>= 4; rv += 4;
+    }
+    if (wi & 0xc) {
+	wi >>= 2; rv += 2;
+    }
+    if (wi & 0x2) {
+	wi >>= 1; ++rv;
+    }
+    if (wi & 0x1) {
+	++rv;
+    }
+    return rv;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DoubleToExpAndSig --
+ *
+ *	Separates a 'double' into exponent and significand.
+ *
+ * Side effects:
+ *	Stores the significand in '*significand' and the exponent in '*expon'
+ *	so that dv == significand * 2.0**expon, and significand is odd.  Also
+ *	stores the position of the leftmost 1-bit in 'significand' in 'bits'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+DoubleToExpAndSig(
+    double dv,			/* Number to convert. */
+    Tcl_WideUInt *significand,	/* OUTPUT: Significand of the number. */
+    int *expon,			/* OUTPUT: Exponent to multiply the number
+				 * by. */
+    int *bits)			/* OUTPUT: Number of significant bits. */
+{
+    Double d;			/* Number being converted. */
+    Tcl_WideUInt z;		/* Significand under construction. */
+    int de;			/* Exponent of the number. */
+    int k;			/* Bit count. */
+
+    d.d = dv;
+
+    /*
+     * Extract exponent and significand.
+     */
+
+    de = (d.w.word0 & EXP_MASK) >> EXP_SHIFT;
+    z = d.q & SIG_MASK;
+    if (de != 0) {
+	z |= HIDDEN_BIT;
+	k = NormalizeRightward(&z);
+	*bits = FP_PRECISION - k;
+	*expon = k + (de - EXPONENT_BIAS) - (FP_PRECISION-1);
+    } else {
+	k = NormalizeRightward(&z);
+	*expon = k + (de - EXPONENT_BIAS) - (FP_PRECISION-1) + 1;
+	*bits = RequiredPrecision(z);
+    }
+    *significand = z;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TakeAbsoluteValue --
+ *
+ *	Takes the absolute value of a 'double' including 0, Inf and NaN
+ *
+ * Side effects:
+ *	The 'double' in *d is replaced with its absolute value. The signum is
+ *	stored in 'sign': 1 for negative, 0 for nonnegative.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+TakeAbsoluteValue(
+    Double *d,			/* Number to replace with absolute value. */
+    int *sign)			/* Place to put the signum. */
+{
+    if (d->w.word0 & SIGN_BIT) {
+	*sign = 1;
+	d->w.word0 &= ~SIGN_BIT;
+    } else {
+	*sign = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FormatInfAndNaN --
+ *
+ *	Bailout for formatting infinities and Not-A-Number.
+ *
+ * Results:
+ *	Returns one of the strings 'Infinity' and 'NaN'.  The string returned
+ *	must be freed by the caller using 'ckfree'.
+ *
+ * Side effects:
+ *	Stores 9999 in *decpt, and sets '*endPtr' to designate the terminating
+ *	NUL byte of the string if 'endPtr' is not NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+FormatInfAndNaN(
+    Double *d,			/* Exceptional number to format. */
+    int *decpt,			/* Decimal point to set to a bogus value. */
+    char **endPtr)		/* Pointer to the end of the formatted data */
+{
+    char *retval;
+
+    *decpt = 9999;
+    if (!(d->w.word1) && !(d->w.word0 & HI_ORDER_SIG_MASK)) {
+	retval = ckalloc(9);
+	strcpy(retval, "Infinity");
+	if (endPtr) {
+	    *endPtr = retval + 8;
+	}
+    } else {
+	retval = ckalloc(4);
+	strcpy(retval, "NaN");
+	if (endPtr) {
+	    *endPtr = retval + 3;
+	}
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FormatZero --
+ *
+ *	Bailout to format a zero floating-point number.
+ *
+ * Results:
+ *	Returns the constant string "0"
+ *
+ * Side effects:
+ *	Stores 1 in '*decpt' and puts a pointer to the NUL byte terminating
+ *	the string in '*endPtr' if 'endPtr' is not NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+FormatZero(
+    int *decpt,			/* Location of the decimal point. */
+    char **endPtr)		/* Pointer to the end of the formatted data */
+{
+    char *retval = ckalloc(2);
+
+    strcpy(retval, "0");
+    if (endPtr) {
+	*endPtr = retval+1;
+    }
+    *decpt = 0;
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ApproximateLog10 --
+ *
+ *	Computes a two-term Taylor series approximation to the common log of a
+ *	number, and computes the number's binary log.
+ *
+ * Results:
+ *	Return an approximation to floor(log10(bw*2**be)) that is either exact
+ *	or 1 too high.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ApproximateLog10(
+    Tcl_WideUInt bw,		/* Integer significand of the number. */
+    int be,			/* Power of two to scale bw. */
+    int bbits)			/* Number of bits of precision in bw. */
+{
+    int i;			/* Log base 2 of the number. */
+    int k;			/* Floor(Log base 10 of the number) */
+    double ds;			/* Mantissa of the number. */
+    Double d2;
+
+    /*
+     * Compute i and d2 such that d = d2*2**i, and 1 < d2 < 2.
+     * Compute an approximation to log10(d),
+     *   log10(d) ~ log10(2) * i + log10(1.5)
+     *            + (significand-1.5)/(1.5 * log(10))
+     */
+
+    d2.q = bw << (FP_PRECISION - bbits) & SIG_MASK;
+    d2.w.word0 |= (EXPONENT_BIAS) << EXP_SHIFT;
+    i = be + bbits - 1;
+    ds = (d2.d - 1.5) * TWO_OVER_3LOG10
+	    + LOG10_3HALVES_PLUS_FUDGE + LOG10_2 * i;
+    k = (int) ds;
+    if (k > ds) {
+	--k;
+    }
+    return k;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BetterLog10 --
+ *
+ *	Improves the result of ApproximateLog10 for numbers in the range
+ *	1 .. 10**(TEN_PMAX)-1
+ *
+ * Side effects:
+ *	Sets k_check to 0 if the new result is known to be exact, and to 1 if
+ *	it may still be one too high.
+ *
+ * Results:
+ *	Returns the improved approximation to log10(d).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+BetterLog10(
+    double d,			/* Original number to format. */
+    int k,			/* Characteristic(Log base 10) of the
+				 * number. */
+    int *k_check)		/* Flag == 1 if k is inexact. */
+{
+    /*
+     * Performance hack. If k is in the range 0..TEN_PMAX, then we can use a
+     * powers-of-ten table to check it.
+     */
+
+    if (k >= 0 && k <= TEN_PMAX) {
+	if (d < tens[k]) {
+	    k--;
+	}
+	*k_check = 0;
+    } else {
+	*k_check = 1;
+    }
+    return k;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ComputeScale --
+ *
+ *	Prepares to format a floating-point number as decimal.
+ *
+ * Parameters:
+ *	floor(log10*x) is k (or possibly k-1).  floor(log2(x) is i.  The
+ *	significand of x requires bbits bits to represent.
+ *
+ * Results:
+ *	Determines integers b2, b5, s2, s5 so that sig*2**b2*5**b5/2**s2*2**s5
+ *	exactly represents the value of the x/10**k. This value will lie in
+ *	the range [1 .. 10), and allows for computing successive digits by
+ *	multiplying sig%10 by 10.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+ComputeScale(
+    int be,			/* Exponent part of number: d = bw * 2**be. */
+    int k,			/* Characteristic of log10(number). */
+    int *b2,			/* OUTPUT: Power of 2 in the numerator. */
+    int *b5,			/* OUTPUT: Power of 5 in the numerator. */
+    int *s2,			/* OUTPUT: Power of 2 in the denominator. */
+    int *s5)			/* OUTPUT: Power of 5 in the denominator. */
+{
+    /*
+     * Scale numerator and denominator powers of 2 so that the input binary
+     * number is the ratio of integers.
+     */
+
+    if (be <= 0) {
+	*b2 = 0;
+	*s2 = -be;
+    } else {
+	*b2 = be;
+	*s2 = 0;
+    }
+
+    /*
+     * Scale numerator and denominator so that the output decimal number is
+     * the ratio of integers.
+     */
+
+    if (k >= 0) {
+	*b5 = 0;
+	*s5 = k;
+	*s2 += k;
+    } else {
+	*b2 -= k;
+	*b5 = -k;
+	*s5 = 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetPrecisionLimits --
+ *
+ *	Determines how many digits of significance should be computed (and,
+ *	hence, how much memory need be allocated) for formatting a floating
+ *	point number.
+ *
+ * Given that 'k' is floor(log10(x)):
+ * if 'shortest' format is used, there will be at most 18 digits in the
+ * result.
+ * if 'F' format is used, there will be at most 'ndigits' + k + 1 digits
+ * if 'E' format is used, there will be exactly 'ndigits' digits.
+ *
+ * Side effects:
+ *	Adjusts '*ndigitsPtr' to have a valid value. Stores the maximum memory
+ *	allocation needed in *iPtr.  Sets '*iLimPtr' to the limiting number of
+ *	digits to convert if k has been guessed correctly, and '*iLim1Ptr' to
+ *	the limiting number of digits to convert if k has been guessed to be
+ *	one too high.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+SetPrecisionLimits(
+    int convType,		/* Type of conversion: TCL_DD_SHORTEST,
+				 * TCL_DD_STEELE0, TCL_DD_E_FMT,
+				 * TCL_DD_F_FMT. */
+    int k,			/* Floor(log10(number to convert)) */
+    int *ndigitsPtr,		/* IN/OUT: Number of digits requested (will be
+				 *         adjusted if needed). */
+    int *iPtr,			/* OUT: Maximum number of digits to return. */
+    int *iLimPtr,		/* OUT: Number of digits of significance if
+				 *      the bignum method is used.*/
+    int *iLim1Ptr)		/* OUT: Number of digits of significance if
+				 *      the quick method is used. */
+{
+    switch (convType) {
+    case TCL_DD_SHORTEST0:
+    case TCL_DD_STEELE0:
+	*iLimPtr = *iLim1Ptr = -1;
+	*iPtr = 18;
+	*ndigitsPtr = 0;
+	break;
+    case TCL_DD_E_FORMAT:
+	if (*ndigitsPtr <= 0) {
+	    *ndigitsPtr = 1;
+	}
+	*iLimPtr = *iLim1Ptr = *iPtr = *ndigitsPtr;
+	break;
+    case TCL_DD_F_FORMAT:
+	*iPtr = *ndigitsPtr + k + 1;
+	*iLimPtr = *iPtr;
+	*iLim1Ptr = *iPtr - 1;
+	if (*iPtr <= 0) {
+	    *iPtr = 1;
+	}
+	break;
+    default:
+	*iPtr = -1;
+	*iLimPtr = -1;
+	*iLim1Ptr = -1;
+	Tcl_Panic("impossible conversion type in TclDoubleDigits");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BumpUp --
+ *
+ *	Increases a string of digits ending in a series of nines to designate
+ *	the next higher number.  xxxxb9999... -> xxxx(b+1)0000...
+ *
+ * Results:
+ *	Returns a pointer to the end of the adjusted string.
+ *
+ * Side effects:
+ *	In the case that the string consists solely of '999999', sets it to
+ *	"1" and moves the decimal point (*kPtr) one place to the right.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+BumpUp(
+    char *s,		    	/* Cursor pointing one past the end of the
+				 * string. */
+    char *retval,		/* Start of the string of digits. */
+    int *kPtr)			/* Position of the decimal point. */
+{
+    while (*--s == '9') {
+	if (s == retval) {
+	    ++(*kPtr);
+	    *s = '1';
+	    return s+1;
+	}
+    }
+    ++*s;
+    ++s;
+    return s;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AdjustRange --
+ *
+ *	Rescales a 'double' in preparation for formatting it using the 'quick'
+ *	double-to-string method.
+ *
+ * Results:
+ *	Returns the precision that has been lost in the prescaling as a count
+ *	of units in the least significant place.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+AdjustRange(
+    double *dPtr,		/* INOUT: Number to adjust. */
+    int k)			/* IN: floor(log10(d)) */
+{
+    int ieps;			/* Number of roundoff errors that have
+				 * accumulated. */
+    double d = *dPtr;		/* Number to adjust. */
+    double ds;
+    int i, j, j1;
+
+    ieps = 2;
+
+    if (k > 0) {
+	/*
+	 * The number must be reduced to bring it into range.
+	 */
+
+	ds = tens[k & 0xf];
+	j = k >> 4;
+	if (j & BLETCH) {
+	    j &= (BLETCH-1);
+	    d /= bigtens[N_BIGTENS - 1];
+	    ieps++;
+	}
+	i = 0;
+	for (; j != 0; j>>=1) {
+	    if (j & 1) {
+		ds *= bigtens[i];
+		++ieps;
+	    }
+	    ++i;
+	}
+	d /= ds;
+    } else if ((j1 = -k) != 0) {
+	/*
+	 * The number must be increased to bring it into range.
+	 */
+
+	d *= tens[j1 & 0xf];
+	i = 0;
+	for (j = j1>>4; j; j>>=1) {
+	    if (j & 1) {
+		ieps++;
+		d *= bigtens[i];
+	    }
+	    ++i;
+	}
+    }
+
+    *dPtr = d;
+    return ieps;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShorteningQuickFormat --
+ *
+ *	Returns a 'quick' format of a double precision number to a string of
+ *	digits, preferring a shorter string of digits if the shorter string is
+ *	still within 1/2 ulp of the number.
+ *
+ * Results:
+ *	Returns the string of digits. Returns NULL if the 'quick' method fails
+ *	and the bignum method must be used.
+ *
+ * Side effects:
+ *	Stores the position of the decimal point at '*kPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+ShorteningQuickFormat(
+    double d,			/* Number to convert. */
+    int k,			/* floor(log10(d)) */
+    int ilim,			/* Number of significant digits to return. */
+    double eps,			/* Estimated roundoff error. */
+    char *retval,		/* Buffer to receive the digit string. */
+    int *kPtr)			/* Pointer to stash the position of the
+				 * decimal point. */
+{
+    char *s = retval;		/* Cursor in the return value. */
+    int digit;			/* Current digit. */
+    int i;
+
+    eps = 0.5 / tens[ilim-1] - eps;
+    i = 0;
+    for (;;) {
+	/*
+	 * Convert a digit.
+	 */
+
+	digit = (int) d;
+	d -= digit;
+	*s++ = '0' + digit;
+
+	/*
+	 * Truncate the conversion if the string of digits is within 1/2 ulp
+	 * of the actual value.
+	 */
+
+	if (d < eps) {
+	    *kPtr = k;
+	    return s;
+	}
+	if ((1. - d) < eps) {
+	    *kPtr = k;
+	    return BumpUp(s, retval, kPtr);
+	}
+
+	/*
+	 * Bail out if the conversion fails to converge to a sufficiently
+	 * precise value.
+	 */
+
+	if (++i >= ilim) {
+	    return NULL;
+	}
+
+	/*
+	 * Bring the next digit to the integer part.
+	 */
+
+	eps *= 10;
+	d *= 10.0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StrictQuickFormat --
+ *
+ *	Convert a double precision number of a string of a precise number of
+ *	digits, using the 'quick' double precision method.
+ *
+ * Results:
+ *	Returns the digit string, or NULL if the bignum method must be used to
+ *	do the formatting.
+ *
+ * Side effects:
+ *	Stores the position of the decimal point in '*kPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+StrictQuickFormat(
+    double d,			/* Number to convert. */
+    int k,			/* floor(log10(d)) */
+    int ilim,			/* Number of significant digits to return. */
+    double eps,			/* Estimated roundoff error. */
+    char *retval,		/* Start of the digit string. */
+    int *kPtr)			/* Pointer to stash the position of the
+				 * decimal point. */
+{
+    char *s = retval;		/* Cursor in the return value. */
+    int digit;			/* Current digit of the answer. */
+    int i;
+
+    eps *= tens[ilim-1];
+    i = 1;
+    for (;;) {
+	/*
+	 * Extract a digit.
+	 */
+
+	digit = (int) d;
+	d -= digit;
+	if (d == 0.0) {
+	    ilim = i;
+	}
+	*s++ = '0' + digit;
+
+	/*
+	 * When the given digit count is reached, handle trailing strings of 0
+	 * and 9.
+	 */
+
+	if (i == ilim) {
+	    if (d > 0.5 + eps) {
+		*kPtr = k;
+		return BumpUp(s, retval, kPtr);
+	    } else if (d < 0.5 - eps) {
+		while (*--s == '0') {
+		    /* do nothing */
+		}
+		s++;
+		*kPtr = k;
+		return s;
+	    } else {
+		return NULL;
+	    }
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	++i;
+	d *= 10.0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * QuickConversion --
+ *
+ *	Converts a floating point number the 'quick' way, when only a limited
+ *	number of digits is required and floating point arithmetic can
+ *	therefore be used for the intermediate results.
+ *
+ * Results:
+ *	Returns the converted string, or NULL if the bignum method must be
+ *	used.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+QuickConversion(
+    double e,			/* Number to format. */
+    int k,			/* floor(log10(d)), approximately. */
+    int k_check,		/* 0 if k is exact, 1 if it may be too high */
+    int flags,			/* Flags passed to dtoa:
+				 *    TCL_DD_SHORTEN_FLAG */
+    int len,			/* Length of the return value. */
+    int ilim,			/* Number of digits to store. */
+    int ilim1,			/* Number of digits to store if we misguessed
+				 * k. */
+    int *decpt,			/* OUTPUT: Location of the decimal point. */
+    char **endPtr)		/* OUTPUT: Pointer to the terminal null
+				 * byte. */
+{
+    int ieps;			/* Number of 1-ulp roundoff errors that have
+				 * accumulated in the calculation. */
+    Double eps;			/* Estimated roundoff error. */
+    char *retval;		/* Returned string. */
+    char *end;			/* Pointer to the terminal null byte in the
+				 * returned string. */
+    volatile double d;		/* Workaround for a bug in mingw gcc 3.4.5 */
+
+    /*
+     * Bring d into the range [1 .. 10).
+     */
+
+    ieps = AdjustRange(&e, k);
+    d = e;
+
+    /*
+     * If the guessed value of k didn't get d into range, adjust it by one. If
+     * that leaves us outside the range in which quick format is accurate,
+     * bail out.
+     */
+
+    if (k_check && d < 1. && ilim > 0) {
+	if (ilim1 < 0) {
+	    return NULL;
+	}
+	ilim = ilim1;
+	--k;
+	d *= 10.0;
+	++ieps;
+    }
+
+    /*
+     * Compute estimated roundoff error.
+     */
+
+    eps.d = ieps * d + 7.;
+    eps.w.word0 -= (FP_PRECISION-1) << EXP_SHIFT;
+
+    /*
+     * Handle the peculiar case where the result has no significant digits.
+     */
+
+    retval = ckalloc(len + 1);
+    if (ilim == 0) {
+	d -= 5.;
+	if (d > eps.d) {
+	    *retval = '1';
+	    *decpt = k;
+	    return retval;
+	} else if (d < -eps.d) {
+	    *decpt = k;
+	    return retval;
+	} else {
+	    ckfree(retval);
+	    return NULL;
+	}
+    }
+
+    /*
+     * Format the digit string.
+     */
+
+    if (flags & TCL_DD_SHORTEN_FLAG) {
+	end = ShorteningQuickFormat(d, k, ilim, eps.d, retval, decpt);
+    } else {
+	end = StrictQuickFormat(d, k, ilim, eps.d, retval, decpt);
+    }
+    if (end == NULL) {
+	ckfree(retval);
+	return NULL;
+    }
+    *end = '\0';
+    if (endPtr != NULL) {
+	*endPtr = end;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CastOutPowersOf2 --
+ *
+ *	Adjust the factors 'b2', 'm2', and 's2' to cast out common powers of 2
+ *	from numerator and denominator in preparation for the 'bignum' method
+ *	of floating point conversion.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+CastOutPowersOf2(
+    int *b2,			/* Power of 2 to multiply the significand. */
+    int *m2,			/* Power of 2 to multiply 1/2 ulp. */
+    int *s2)			/* Power of 2 to multiply the common
+				 * denominator. */
+{
+    int i;
+
+    if (*m2 > 0 && *s2 > 0) {	/* Find the smallest power of 2 in the
+				 * numerator. */
+	if (*m2 < *s2) {	/* Find the lowest common denominator. */
+	    i = *m2;
+	} else {
+	    i = *s2;
+	}
+	*b2 -= i;		/* Reduce to lowest terms. */
+	*m2 -= i;
+	*s2 -= i;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShorteningInt64Conversion --
+ *
+ *	Converts a double-precision number to the shortest string of digits
+ *	that reconverts exactly to the given number, or to 'ilim' digits if
+ *	that will yield a shorter result. The numerator and denominator in
+ *	David Gay's conversion algorithm are known to fit in Tcl_WideUInt,
+ *	giving considerably faster arithmetic than mp_int's.
+ *
+ * Results:
+ *	Returns the string of significant decimal digits, in newly allocated
+ *	memory
+ *
+ * Side effects:
+ *	Stores the location of the decimal point in '*decpt' and the location
+ *	of the terminal null byte in '*endPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+ShorteningInt64Conversion(
+    Double *dPtr,		/* Original number to convert. */
+    int convType,		/* Type of conversion (shortest, Steele,
+				 * E format, F format). */
+    Tcl_WideUInt bw,		/* Integer significand. */
+    int b2, int b5,		/* Scale factor for the significand in the
+				 * numerator. */
+    int m2plus, int m2minus, int m5,
+				/* Scale factors for 1/2 ulp in the numerator
+				 * (will be different if bw == 1. */
+    int s2, int s5,		/* Scale factors for the denominator. */
+    int k,			/* Number of output digits before the decimal
+				 * point. */
+    int len,			/* Number of digits to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Position of the terminal '\0' at
+				 *	   the end of the returned string. */
+{
+    char *retval = ckalloc(len + 1);
+				/* Output buffer. */
+    Tcl_WideUInt b = (bw * wuipow5[b5]) << b2;
+				/* Numerator of the fraction being
+				 * converted. */
+    Tcl_WideUInt S = wuipow5[s5] << s2;
+				/* Denominator of the fraction being
+				 * converted. */
+    Tcl_WideUInt mplus, mminus;	/* Ranges for testing whether the result is
+				 * within roundoff of being exact. */
+    int digit;			/* Current output digit. */
+    char *s = retval;		/* Cursor in the output buffer. */
+    int i;			/* Current position in the output buffer. */
+
+    /*
+     * Adjust if the logarithm was guessed wrong.
+     */
+
+    if (b < S) {
+	b = 10 * b;
+	++m2plus; ++m2minus; ++m5;
+	ilim = ilim1;
+	--k;
+    }
+
+    /*
+     * Compute roundoff ranges.
+     */
+
+    mplus = wuipow5[m5] << m2plus;
+    mminus = wuipow5[m5] << m2minus;
+
+    /*
+     * Loop through the digits.
+     */
+
+    i = 1;
+    for (;;) {
+	digit = (int)(b / S);
+	if (digit > 10) {
+	    Tcl_Panic("wrong digit!");
+	}
+	b = b % S;
+
+	/*
+	 * Does the current digit put us on the low side of the exact value
+	 * but within within roundoff of being exact?
+	 */
+
+	if (b < mplus || (b == mplus
+		&& convType != TCL_DD_STEELE0 && (dPtr->w.word1 & 1) == 0)) {
+	    /*
+	     * Make sure we shouldn't be rounding *up* instead, in case the
+	     * next number above is closer.
+	     */
+
+	    if (2 * b > S || (2 * b == S && (digit & 1) != 0)) {
+		++digit;
+		if (digit == 10) {
+		    *s++ = '9';
+		    s = BumpUp(s, retval, &k);
+		    break;
+		}
+	    }
+
+	    /*
+	     * Stash the current digit.
+	     */
+
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Does one plus the current digit put us within roundoff of the
+	 * number?
+	 */
+
+	if (b > S - mminus || (b == S - mminus
+		&& convType != TCL_DD_STEELE0 && (dPtr->w.word1 & 1) == 0)) {
+	    if (digit == 9) {
+		*s++ = '9';
+		s = BumpUp(s, retval, &k);
+		break;
+	    }
+	    ++digit;
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Have we converted all the requested digits?
+	 */
+
+	*s++ = '0' + digit;
+	if (i == ilim) {
+	    if (2*b > S || (2*b == S && (digit & 1) != 0)) {
+		s = BumpUp(s, retval, &k);
+	    }
+	    break;
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	b = 10 * b;
+	mplus = 10 * mplus;
+	mminus = 10 * mminus;
+	++i;
+    }
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StrictInt64Conversion --
+ *
+ *	Converts a double-precision number to a fixed-length string of 'ilim'
+ *	digits that reconverts exactly to the given number.  ('ilim' should be
+ *	replaced with 'ilim1' in the case where log10(d) has been
+ *	overestimated).  The numerator and denominator in David Gay's
+ *	conversion algorithm are known to fit in Tcl_WideUInt, giving
+ *	considerably faster arithmetic than mp_int's.
+ *
+ * Results:
+ *	Returns the string of significant decimal digits, in newly allocated
+ *	memory
+ *
+ * Side effects:
+ *	Stores the location of the decimal point in '*decpt' and the location
+ *	of the terminal null byte in '*endPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+StrictInt64Conversion(
+    Double *dPtr,		/* Original number to convert. */
+    int convType,		/* Type of conversion (shortest, Steele,
+				 * E format, F format). */
+    Tcl_WideUInt bw,		/* Integer significand. */
+    int b2, int b5,		/* Scale factor for the significand in the
+				 * numerator. */
+    int s2, int s5,		/* Scale factors for the denominator. */
+    int k,			/* Number of output digits before the decimal
+				 * point. */
+    int len,			/* Number of digits to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Position of the terminal '\0' at
+				 *	   the end of the returned string. */
+{
+    char *retval = ckalloc(len + 1);
+				/* Output buffer. */
+    Tcl_WideUInt b = (bw * wuipow5[b5]) << b2;
+				/* Numerator of the fraction being
+				 * converted. */
+    Tcl_WideUInt S = wuipow5[s5] << s2;
+				/* Denominator of the fraction being
+				 * converted. */
+    int digit;			/* Current output digit. */
+    char *s = retval;		/* Cursor in the output buffer. */
+    int i;			/* Current position in the output buffer. */
+
+    /*
+     * Adjust if the logarithm was guessed wrong.
+     */
+
+    if (b < S) {
+	b = 10 * b;
+	ilim = ilim1;
+	--k;
+    }
+
+    /*
+     * Loop through the digits.
+     */
+
+    i = 1;
+    for (;;) {
+	digit = (int)(b / S);
+	if (digit > 10) {
+	    Tcl_Panic("wrong digit!");
+	}
+	b = b % S;
+
+	/*
+	 * Have we converted all the requested digits?
+	 */
+
+	*s++ = '0' + digit;
+	if (i == ilim) {
+	    if (2*b > S || (2*b == S && (digit & 1) != 0)) {
+		s = BumpUp(s, retval, &k);
+	    } else {
+		while (*--s == '0') {
+		    /* do nothing */
+		}
+		++s;
+	    }
+	    break;
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	b = 10 * b;
+	++i;
+    }
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShouldBankerRoundUpPowD --
+ *
+ *	Test whether bankers' rounding should round a digit up. Assumption is
+ *	made that the denominator of the fraction being tested is a power of
+ *	2**DIGIT_BIT.
+ *
+ * Results:
+ *	Returns 1 iff the fraction is more than 1/2, or if the fraction is
+ *	exactly 1/2 and the digit is odd.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ShouldBankerRoundUpPowD(
+    mp_int *b,			/* Numerator of the fraction. */
+    int sd,			/* Denominator is 2**(sd*DIGIT_BIT). */
+    int isodd)			/* 1 if the digit is odd, 0 if even. */
+{
+    int i;
+    static const mp_digit topbit = 1 << (DIGIT_BIT - 1);
+
+    if (b->used < sd || (b->dp[sd-1] & topbit) == 0) {
+	return 0;
+    }
+    if (b->dp[sd-1] != topbit) {
+	return 1;
+    }
+    for (i = sd-2; i >= 0; --i) {
+	if (b->dp[i] != 0) {
+	    return 1;
+	}
+    }
+    return isodd;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShouldBankerRoundUpToNextPowD --
+ *
+ *	Tests whether bankers' rounding will round down in the "denominator is
+ *	a power of 2**MP_DIGIT" case.
+ *
+ * Results:
+ *	Returns 1 if the rounding will be performed - which increases the
+ *	digit by one - and 0 otherwise.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ShouldBankerRoundUpToNextPowD(
+    mp_int *b,			/* Numerator of the fraction. */
+    mp_int *m,			/* Numerator of the rounding tolerance. */
+    int sd,			/* Common denominator is 2**(sd*DIGIT_BIT). */
+    int convType,		/* Conversion type: STEELE defeats
+				 * round-to-even (not sure why one wants to do
+				 * this; I copied it from Gay). FIXME */
+    int isodd,			/* 1 if the integer significand is odd. */
+    mp_int *temp)		/* Work area for the calculation. */
+{
+    int i;
+
+    /*
+     * Compare B and S-m - which is the same as comparing B+m and S - which we
+     * do by computing b+m and doing a bitwhack compare against
+     * 2**(DIGIT_BIT*sd)
+     */
+
+    mp_add(b, m, temp);
+    if (temp->used <= sd) {	/* Too few digits to be > s */
+	return 0;
+    }
+    if (temp->used > sd+1 || temp->dp[sd] > 1) {
+				/* >= 2s */
+	return 1;
+    }
+    for (i = sd-1; i >= 0; --i) {
+				/* Check for ==s */
+	if (temp->dp[i] != 0) {	/* > s */
+	    return 1;
+	}
+    }
+    if (convType == TCL_DD_STEELE0) {
+				/* Biased rounding. */
+	return 0;
+    }
+    return isodd;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShorteningBignumConversionPowD --
+ *
+ *	Converts a double-precision number to the shortest string of digits
+ *	that reconverts exactly to the given number, or to 'ilim' digits if
+ *	that will yield a shorter result. The denominator in David Gay's
+ *	conversion algorithm is known to be a power of 2**DIGIT_BIT, and hence
+ *	the division in the main loop may be replaced by a digit shift and
+ *	mask.
+ *
+ * Results:
+ *	Returns the string of significant decimal digits, in newly allocated
+ *	memory
+ *
+ * Side effects:
+ *	Stores the location of the decimal point in '*decpt' and the location
+ *	of the terminal null byte in '*endPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+ShorteningBignumConversionPowD(
+    Double *dPtr,		/* Original number to convert. */
+    int convType,		/* Type of conversion (shortest, Steele,
+				 * E format, F format). */
+    Tcl_WideUInt bw,		/* Integer significand. */
+    int b2, int b5,		/* Scale factor for the significand in the
+				 * numerator. */
+    int m2plus, int m2minus, int m5,
+				/* Scale factors for 1/2 ulp in the numerator
+				 * (will be different if bw == 1). */
+    int sd,			/* Scale factor for the denominator. */
+    int k,			/* Number of output digits before the decimal
+				 * point. */
+    int len,			/* Number of digits to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Position of the terminal '\0' at
+				 *	   the end of the returned string. */
+{
+    char *retval = ckalloc(len + 1);
+				/* Output buffer. */
+    mp_int b;			/* Numerator of the fraction being
+				 * converted. */
+    mp_int mplus, mminus;	/* Bounds for roundoff. */
+    mp_digit digit;		/* Current output digit. */
+    char *s = retval;		/* Cursor in the output buffer. */
+    int i;			/* Index in the output buffer. */
+    mp_int temp;
+    int r1;
+
+    /*
+     * b = bw * 2**b2 * 5**b5
+     * mminus = 5**m5
+     */
+
+    TclBNInitBignumFromWideUInt(&b, bw);
+    mp_init_set_int(&mminus, 1);
+    MulPow5(&b, b5, &b);
+    mp_mul_2d(&b, b2, &b);
+
+    /*
+     * Adjust if the logarithm was guessed wrong.
+     */
+
+    if (b.used <= sd) {
+	mp_mul_d(&b, 10, &b);
+	++m2plus; ++m2minus; ++m5;
+	ilim = ilim1;
+	--k;
+    }
+
+    /*
+     * mminus = 5**m5 * 2**m2minus
+     * mplus = 5**m5 * 2**m2plus
+     */
+
+    mp_mul_2d(&mminus, m2minus, &mminus);
+    MulPow5(&mminus, m5, &mminus);
+    if (m2plus > m2minus) {
+	mp_init_copy(&mplus, &mminus);
+	mp_mul_2d(&mplus, m2plus-m2minus, &mplus);
+    }
+    mp_init(&temp);
+
+    /*
+     * Loop through the digits. Do division and mod by s == 2**(sd*DIGIT_BIT)
+     * by mp_digit extraction.
+     */
+
+    i = 0;
+    for (;;) {
+	if (b.used <= sd) {
+	    digit = 0;
+	} else {
+	    digit = b.dp[sd];
+	    if (b.used > sd+1 || digit >= 10) {
+		Tcl_Panic("wrong digit!");
+	    }
+	    --b.used; mp_clamp(&b);
+	}
+
+	/*
+	 * Does the current digit put us on the low side of the exact value
+	 * but within within roundoff of being exact?
+	 */
+
+	r1 = mp_cmp_mag(&b, (m2plus > m2minus)? &mplus : &mminus);
+	if (r1 == MP_LT || (r1 == MP_EQ
+		&& convType != TCL_DD_STEELE0 && (dPtr->w.word1 & 1) == 0)) {
+	    /*
+	     * Make sure we shouldn't be rounding *up* instead, in case the
+	     * next number above is closer.
+	     */
+
+	    if (ShouldBankerRoundUpPowD(&b, sd, digit&1)) {
+		++digit;
+		if (digit == 10) {
+		    *s++ = '9';
+		    s = BumpUp(s, retval, &k);
+		    break;
+		}
+	    }
+
+	    /*
+	     * Stash the last digit.
+	     */
+
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Does one plus the current digit put us within roundoff of the
+	 * number?
+	 */
+
+	if (ShouldBankerRoundUpToNextPowD(&b, &mminus, sd, convType,
+		dPtr->w.word1 & 1, &temp)) {
+	    if (digit == 9) {
+		*s++ = '9';
+		s = BumpUp(s, retval, &k);
+		break;
+	    }
+	    ++digit;
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Have we converted all the requested digits?
+	 */
+
+	*s++ = '0' + digit;
+	if (i == ilim) {
+	    if (ShouldBankerRoundUpPowD(&b, sd, digit&1)) {
+		s = BumpUp(s, retval, &k);
+	    }
+	    break;
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	mp_mul_d(&b, 10, &b);
+	mp_mul_d(&mminus, 10, &mminus);
+	if (m2plus > m2minus) {
+	    mp_mul_2d(&mminus, m2plus-m2minus, &mplus);
+	}
+	++i;
+    }
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    if (m2plus > m2minus) {
+	mp_clear(&mplus);
+    }
+    mp_clear_multi(&b, &mminus, &temp, NULL);
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StrictBignumConversionPowD --
+ *
+ *	Converts a double-precision number to a fixed-lengt string of 'ilim'
+ *	digits (or 'ilim1' if log10(d) has been overestimated).  The
+ *	denominator in David Gay's conversion algorithm is known to be a power
+ *	of 2**DIGIT_BIT, and hence the division in the main loop may be
+ *	replaced by a digit shift and mask.
+ *
+ * Results:
+ *	Returns the string of significant decimal digits, in newly allocated
+ *	memory.
+ *
+ * Side effects:
+ *	Stores the location of the decimal point in '*decpt' and the location
+ *	of the terminal null byte in '*endPtr'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+StrictBignumConversionPowD(
+    Double *dPtr,		/* Original number to convert. */
+    int convType,		/* Type of conversion (shortest, Steele,
+				 * E format, F format). */
+    Tcl_WideUInt bw,		/* Integer significand. */
+    int b2, int b5,		/* Scale factor for the significand in the
+				 * numerator. */
+    int sd,			/* Scale factor for the denominator. */
+    int k,			/* Number of output digits before the decimal
+				 * point. */
+    int len,			/* Number of digits to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Position of the terminal '\0' at
+				 *	   the end of the returned string. */
+{
+    char *retval = ckalloc(len + 1);
+				/* Output buffer. */
+    mp_int b;			/* Numerator of the fraction being
+				 * converted. */
+    mp_digit digit;		/* Current output digit. */
+    char *s = retval;		/* Cursor in the output buffer. */
+    int i;			/* Index in the output buffer. */
+    mp_int temp;
+
+    /*
+     * b = bw * 2**b2 * 5**b5
+     */
+
+    TclBNInitBignumFromWideUInt(&b, bw);
+    MulPow5(&b, b5, &b);
+    mp_mul_2d(&b, b2, &b);
+
+    /*
+     * Adjust if the logarithm was guessed wrong.
+     */
+
+    if (b.used <= sd) {
+	mp_mul_d(&b, 10, &b);
+	ilim = ilim1;
+	--k;
+    }
+    mp_init(&temp);
+
+    /*
+     * Loop through the digits. Do division and mod by s == 2**(sd*DIGIT_BIT)
+     * by mp_digit extraction.
+     */
+
+    i = 1;
+    for (;;) {
+	if (b.used <= sd) {
+	    digit = 0;
+	} else {
+	    digit = b.dp[sd];
+	    if (b.used > sd+1 || digit >= 10) {
+		Tcl_Panic("wrong digit!");
+	    }
+	    --b.used;
+	    mp_clamp(&b);
+	}
+
+	/*
+	 * Have we converted all the requested digits?
+	 */
+
+	*s++ = '0' + digit;
+	if (i == ilim) {
+	    if (ShouldBankerRoundUpPowD(&b, sd, digit&1)) {
+		s = BumpUp(s, retval, &k);
+	    }
+	    while (*--s == '0') {
+		/* do nothing */
+	    }
+	    ++s;
+	    break;
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	mp_mul_d(&b, 10, &b);
+	++i;
+    }
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    mp_clear_multi(&b, &temp, NULL);
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShouldBankerRoundUp --
+ *
+ *	Tests whether a digit should be rounded up or down when finishing
+ *	bignum-based floating point conversion.
+ *
+ * Results:
+ *	Returns 1 if the number needs to be rounded up, 0 otherwise.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ShouldBankerRoundUp(
+    mp_int *twor,		/* 2x the remainder from thd division that
+				 * produced the last digit. */
+    mp_int *S,			/* Denominator. */
+    int isodd)			/* Flag == 1 if the last digit is odd. */
+{
+    int r = mp_cmp_mag(twor, S);
+
+    switch (r) {
+    case MP_LT:
+	return 0;
+    case MP_EQ:
+	return isodd;
+    case MP_GT:
+	return 1;
+    }
+    Tcl_Panic("in ShouldBankerRoundUp, trichotomy fails!");
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShouldBankerRoundUpToNext --
+ *
+ *	Tests whether the remainder is great enough to force rounding to the
+ *	next higher digit.
+ *
+ * Results:
+ *	Returns 1 if the number should be rounded up, 0 otherwise.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ShouldBankerRoundUpToNext(
+    mp_int *b,			/* Remainder from the division that produced
+				 * the last digit. */
+    mp_int *m,			/* Numerator of the rounding tolerance. */
+    mp_int *S,			/* Denominator. */
+    int convType,		/* Conversion type: STEELE0 defeats
+				 * round-to-even. (Not sure why one would want
+				 * this; I coped it from Gay). FIXME */
+    int isodd,			/* 1 if the integer significand is odd. */
+    mp_int *temp)		/* Work area needed for the calculation. */
+{
+    int r;
+
+    /*
+     * Compare b and S-m: this is the same as comparing B+m and S.
+     */
+
+    mp_add(b, m, temp);
+    r = mp_cmp_mag(temp, S);
+    switch(r) {
+    case MP_LT:
+	return 0;
+    case MP_EQ:
+	if (convType == TCL_DD_STEELE0) {
+	    return 0;
+	} else {
+	    return isodd;
+	}
+    case MP_GT:
+	return 1;
+    }
+    Tcl_Panic("in ShouldBankerRoundUpToNext, trichotomy fails!");
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ShorteningBignumConversion --
+ *
+ *	Convert a floating point number to a variable-length digit string
+ *	using the multiprecision method.
+ *
+ * Results:
+ *	Returns the string of digits.
+ *
+ * Side effects:
+ *	Stores the position of the decimal point in *decpt.  Stores a pointer
+ *	to the end of the number in *endPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+ShorteningBignumConversion(
+    Double *dPtr,		/* Original number being converted. */
+    int convType,		/* Conversion type. */
+    Tcl_WideUInt bw,		/* Integer significand and exponent. */
+    int b2,			/* Scale factor for the significand. */
+    int m2plus, int m2minus,	/* Scale factors for 1/2 ulp in numerator. */
+    int s2, int s5,		/* Scale factors for denominator. */
+    int k,			/* Guessed position of the decimal point. */
+    int len,			/* Size of the digit buffer to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Pointer to the end of the number */
+{
+    char *retval = ckalloc(len+1);
+				/* Buffer of digits to return. */
+    char *s = retval;		/* Cursor in the return value. */
+    mp_int b;			/* Numerator of the result. */
+    mp_int mminus;		/* 1/2 ulp below the result. */
+    mp_int mplus;		/* 1/2 ulp above the result. */
+    mp_int S;			/* Denominator of the result. */
+    mp_int dig;			/* Current digit of the result. */
+    int digit;			/* Current digit of the result. */
+    mp_int temp;		/* Work area. */
+    int minit = 1;		/* Fudge factor for when we misguess k. */
+    int i;
+    int r1;
+
+    /*
+     * b = bw * 2**b2 * 5**b5
+     * S = 2**s2 * 5*s5
+     */
+
+    TclBNInitBignumFromWideUInt(&b, bw);
+    mp_mul_2d(&b, b2, &b);
+    mp_init_set_int(&S, 1);
+    MulPow5(&S, s5, &S); mp_mul_2d(&S, s2, &S);
+
+    /*
+     * Handle the case where we guess the position of the decimal point wrong.
+     */
+
+    if (mp_cmp_mag(&b, &S) == MP_LT) {
+	mp_mul_d(&b, 10, &b);
+	minit = 10;
+	ilim =ilim1;
+	--k;
+    }
+
+    /*
+     * mminus = 2**m2minus * 5**m5
+     */
+
+    mp_init_set_int(&mminus, minit);
+    mp_mul_2d(&mminus, m2minus, &mminus);
+    if (m2plus > m2minus) {
+	mp_init_copy(&mplus, &mminus);
+	mp_mul_2d(&mplus, m2plus-m2minus, &mplus);
+    }
+    mp_init(&temp);
+
+    /*
+     * Loop through the digits.
+     */
+
+    mp_init(&dig);
+    i = 1;
+    for (;;) {
+	mp_div(&b, &S, &dig, &b);
+	if (dig.used > 1 || dig.dp[0] >= 10) {
+	    Tcl_Panic("wrong digit!");
+	}
+	digit = dig.dp[0];
+
+	/*
+	 * Does the current digit leave us with a remainder small enough to
+	 * round to it?
+	 */
+
+	r1 = mp_cmp_mag(&b, (m2plus > m2minus)? &mplus : &mminus);
+	if (r1 == MP_LT || (r1 == MP_EQ
+		&& convType != TCL_DD_STEELE0 && (dPtr->w.word1 & 1) == 0)) {
+	    mp_mul_2d(&b, 1, &b);
+	    if (ShouldBankerRoundUp(&b, &S, digit&1)) {
+		++digit;
+		if (digit == 10) {
+		    *s++ = '9';
+		    s = BumpUp(s, retval, &k);
+		    break;
+		}
+	    }
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Does the current digit leave us with a remainder large enough to
+	 * commit to rounding up to the next higher digit?
+	 */
+
+	if (ShouldBankerRoundUpToNext(&b, &mminus, &S, convType,
+		dPtr->w.word1 & 1, &temp)) {
+	    ++digit;
+	    if (digit == 10) {
+		*s++ = '9';
+		s = BumpUp(s, retval, &k);
+		break;
+	    }
+	    *s++ = '0' + digit;
+	    break;
+	}
+
+	/*
+	 * Have we converted all the requested digits?
+	 */
+
+	*s++ = '0' + digit;
+	if (i == ilim) {
+	    mp_mul_2d(&b, 1, &b);
+	    if (ShouldBankerRoundUp(&b, &S, digit&1)) {
+		s = BumpUp(s, retval, &k);
+	    }
+	    break;
+	}
+
+	/*
+	 * Advance to the next digit.
+	 */
+
+	if (s5 > 0) {
+	    /*
+	     * Can possibly shorten the denominator.
+	     */
+
+	    mp_mul_2d(&b, 1, &b);
+	    mp_mul_2d(&mminus, 1, &mminus);
+	    if (m2plus > m2minus) {
+		mp_mul_2d(&mplus, 1, &mplus);
+	    }
+	    mp_div_d(&S, 5, &S, NULL);
+	    --s5;
+
+	    /*
+	     * IDEA: It might possibly be a win to fall back to int64_t
+	     *       arithmetic here if S < 2**64/10. But it's a win only for
+	     *       a fairly narrow range of magnitudes so perhaps not worth
+	     *       bothering.  We already know that we shorten the
+	     *       denominator by at least 1 mp_digit, perhaps 2, as we do
+	     *       the conversion for 17 digits of significance.
+	     * Possible savings:
+	     * 10**26   1 trip through loop before fallback possible
+	     * 10**27   1 trip
+	     * 10**28   2 trips
+	     * 10**29   3 trips
+	     * 10**30   4 trips
+	     * 10**31   5 trips
+	     * 10**32   6 trips
+	     * 10**33   7 trips
+	     * 10**34   8 trips
+	     * 10**35   9 trips
+	     * 10**36  10 trips
+	     * 10**37  11 trips
+	     * 10**38  12 trips
+	     * 10**39  13 trips
+	     * 10**40  14 trips
+	     * 10**41  15 trips
+	     * 10**42  16 trips
+	     * thereafter no gain.
+	     */
+	} else {
+	    mp_mul_d(&b, 10, &b);
+	    mp_mul_d(&mminus, 10, &mminus);
+	    if (m2plus > m2minus) {
+		mp_mul_2d(&mplus, 10, &mplus);
+	    }
+	}
+
+	++i;
+    }
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    if (m2plus > m2minus) {
+	mp_clear(&mplus);
+    }
+    mp_clear_multi(&b, &mminus, &temp, &dig, &S, NULL);
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StrictBignumConversion --
+ *
+ *	Convert a floating point number to a fixed-length digit string using
+ *	the multiprecision method.
+ *
+ * Results:
+ *	Returns the string of digits.
+ *
+ * Side effects:
+ *	Stores the position of the decimal point in *decpt.  Stores a pointer
+ *	to the end of the number in *endPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline char *
+StrictBignumConversion(
+    Double *dPtr,		/* Original number being converted. */
+    int convType,		/* Conversion type. */
+    Tcl_WideUInt bw,		/* Integer significand and exponent. */
+    int b2,			/* Scale factor for the significand. */
+    int s2, int s5,		/* Scale factors for denominator. */
+    int k,			/* Guessed position of the decimal point. */
+    int len,			/* Size of the digit buffer to allocate. */
+    int ilim,			/* Number of digits to convert if b >= s */
+    int ilim1,			/* Number of digits to convert if b < s */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    char **endPtr)		/* OUTPUT: Pointer to the end of the number */
+{
+    char *retval = ckalloc(len+1);
+				/* Buffer of digits to return. */
+    char *s = retval;		/* Cursor in the return value. */
+    mp_int b;			/* Numerator of the result. */
+    mp_int S;			/* Denominator of the result. */
+    mp_int dig;			/* Current digit of the result. */
+    int digit;			/* Current digit of the result. */
+    mp_int temp;		/* Work area. */
+    int g;			/* Size of the current digit ground. */
+    int i, j;
+
+    /*
+     * b = bw * 2**b2 * 5**b5
+     * S = 2**s2 * 5*s5
+     */
+
+    mp_init_multi(&temp, &dig, NULL);
+    TclBNInitBignumFromWideUInt(&b, bw);
+    mp_mul_2d(&b, b2, &b);
+    mp_init_set_int(&S, 1);
+    MulPow5(&S, s5, &S); mp_mul_2d(&S, s2, &S);
+
+    /*
+     * Handle the case where we guess the position of the decimal point wrong.
+     */
+
+    if (mp_cmp_mag(&b, &S) == MP_LT) {
+	mp_mul_d(&b, 10, &b);
+	ilim =ilim1;
+	--k;
+    }
+
+    /*
+     * Convert the leading digit.
+     */
+
+    i = 0;
+    mp_div(&b, &S, &dig, &b);
+    if (dig.used > 1 || dig.dp[0] >= 10) {
+	Tcl_Panic("wrong digit!");
+    }
+    digit = dig.dp[0];
+
+    /*
+     * Is a single digit all that was requested?
+     */
+
+    *s++ = '0' + digit;
+    if (++i >= ilim) {
+	mp_mul_2d(&b, 1, &b);
+	if (ShouldBankerRoundUp(&b, &S, digit&1)) {
+	    s = BumpUp(s, retval, &k);
+	}
+    } else {
+	for (;;) {
+	    /*
+	     * Shift by a group of digits.
+	     */
+
+	    g = ilim - i;
+	    if (g > DIGIT_GROUP) {
+		g = DIGIT_GROUP;
+	    }
+	    if (s5 >= g) {
+		mp_div_d(&S, dpow5[g], &S, NULL);
+		s5 -= g;
+	    } else if (s5 > 0) {
+		mp_div_d(&S, dpow5[s5], &S, NULL);
+		mp_mul_d(&b, dpow5[g - s5], &b);
+		s5 = 0;
+	    } else {
+		mp_mul_d(&b, dpow5[g], &b);
+	    }
+	    mp_mul_2d(&b, g, &b);
+
+	    /*
+	     * As with the shortening bignum conversion, it's possible at this
+	     * point that we will have reduced the denominator to less than
+	     * 2**64/10, at which point it would be possible to fall back to
+	     * to int64_t arithmetic. But the potential payoff is tremendously
+	     * less - unless we're working in F format - because we know that
+	     * three groups of digits will always suffice for %#.17e, the
+	     * longest format that doesn't introduce empty precision.
+	     *
+	     * Extract the next group of digits.
+	     */
+
+	    mp_div(&b, &S, &dig, &b);
+	    if (dig.used > 1) {
+		Tcl_Panic("wrong digit!");
+	    }
+	    digit = dig.dp[0];
+	    for (j = g-1; j >= 0; --j) {
+		int t = itens[j];
+
+		*s++ = digit / t + '0';
+		digit %= t;
+	    }
+	    i += g;
+
+	    /*
+	     * Have we converted all the requested digits?
+	     */
+
+	    if (i == ilim) {
+		mp_mul_2d(&b, 1, &b);
+		if (ShouldBankerRoundUp(&b, &S, digit&1)) {
+		    s = BumpUp(s, retval, &k);
+		}
+		break;
+	    }
+	}
+    }
+    while (*--s == '0') {
+	/* do nothing */
+    }
+    ++s;
+
+    /*
+     * Endgame - store the location of the decimal point and the end of the
+     * string.
+     */
+
+    mp_clear_multi(&b, &S, &temp, &dig, NULL);
+    *s = '\0';
+    *decpt = k;
+    if (endPtr) {
+	*endPtr = s;
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDoubleDigits --
+ *
+ *	Core of Tcl's conversion of double-precision floating point numbers to
+ *	decimal.
+ *
+ * Results:
+ *	Returns a newly-allocated string of digits.
+ *
+ * Side effects:
+ *	Sets *decpt to the index of the character in the string before the
+ *	place that the decimal point should go. If 'endPtr' is not NULL, sets
+ *	endPtr to point to the terminating '\0' byte of the string. Sets *sign
+ *	to 1 if a minus sign should be printed with the number, or 0 if a plus
+ *	sign (or no sign) should appear.
+ *
+ * This function is a service routine that produces the string of digits for
+ * floating-point-to-decimal conversion. It can do a number of things
+ * according to the 'flags' argument. Valid values for 'flags' include:
+ *	TCL_DD_SHORTEST - This is the default for floating point conversion if
+ *		::tcl_precision is 0. It constructs the shortest string of
+ *		digits that will reconvert to the given number when scanned.
+ *		For floating point numbers that are exactly between two
+ *		decimal numbers, it resolves using the 'round to even' rule.
+ *		With this value, the 'ndigits' parameter is ignored.
+ *	TCL_DD_STEELE - This value is not recommended and may be removed in
+ *		the future. It follows the conversion algorithm outlined in
+ *		"How to Print Floating-Point Numbers Accurately" by Guy
+ *		L. Steele, Jr. and Jon L. White [Proc. ACM SIGPLAN '90,
+ *		pp. 112-126]. This rule has the effect of rendering 1e23 as
+ *		9.9999999999999999e22 - which is a 'better' approximation in
+ *		the sense that it will reconvert correctly even if a
+ *		subsequent input conversion is 'round up' or 'round down'
+ *		rather than 'round to nearest', but is surprising otherwise.
+ *	TCL_DD_E_FORMAT - This value is used to prepare numbers for %e format
+ *		conversion (or for default floating->string if tcl_precision
+ *		is not 0). It constructs a string of at most 'ndigits' digits,
+ *		choosing the one that is closest to the given number (and
+ *		resolving ties with 'round to even').  It is allowed to return
+ *		fewer than 'ndigits' if the number converts exactly; if the
+ *		TCL_DD_E_FORMAT|TCL_DD_SHORTEN_FLAG is supplied instead, it
+ *		also returns fewer digits if the shorter string will still
+ *		reconvert without loss to the given input number. In any case,
+ *		strings of trailing zeroes are suppressed.
+ *	TCL_DD_F_FORMAT - This value is used to prepare numbers for %f format
+ *		conversion. It requests that conversion proceed until
+ *		'ndigits' digits after the decimal point have been converted.
+ *		It is possible for this format to result in a zero-length
+ *		string if the number is sufficiently small. Again, it is
+ *		permissible for TCL_DD_F_FORMAT to return fewer digits for a
+ *		number that converts exactly, and changing the argument to
+ *		TCL_DD_F_FORMAT|TCL_DD_SHORTEN_FLAG will allow the routine
+ *		also to return fewer digits if the shorter string will still
+ *		reconvert without loss to the given input number. Strings of
+ *		trailing zeroes are suppressed.
+ *
+ *	To any of these flags may be OR'ed TCL_DD_NO_QUICK; this flag requires
+ *	all calculations to be done in exact arithmetic. Normally, E and F
+ *	format with fewer than about 14 digits will be done with a quick
+ *	floating point approximation and fall back on the exact arithmetic
+ *	only if the input number is close enough to the midpoint between two
+ *	decimal strings that more precision is needed to resolve which string
+ *	is correct.
+ *
+ * The value stored in the 'decpt' argument on return may be negative
+ * (indicating that the decimal point falls to the left of the string) or
+ * greater than the length of the string. In addition, the value -9999 is used
+ * as a sentinel to indicate that the string is one of the special values
+ * "Infinity" and "NaN", and that no decimal point should be inserted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclDoubleDigits(
+    double dv,			/* Number to convert. */
+    int ndigits,		/* Number of digits requested. */
+    int flags,			/* Conversion flags. */
+    int *decpt,			/* OUTPUT: Position of the decimal point. */
+    int *sign,			/* OUTPUT: 1 if the result is negative. */
+    char **endPtr)		/* OUTPUT: If not NULL, receives a pointer to
+				 *	   one character beyond the end of the
+				 *	   returned string. */
+{
+    int convType = (flags & TCL_DD_CONVERSION_TYPE_MASK);
+				/* Type of conversion being performed:
+				 * TCL_DD_SHORTEST0, TCL_DD_STEELE0,
+				 * TCL_DD_E_FORMAT, or TCL_DD_F_FORMAT. */
+    Double d;			/* Union for deconstructing doubles. */
+    Tcl_WideUInt bw;		/* Integer significand. */
+    int be;			/* Power of 2 by which b must be multiplied */
+    int bbits;			/* Number of bits needed to represent b. */
+    int denorm;			/* Flag == 1 iff the input number was
+				 * denormalized. */
+    int k;			/* Estimate of floor(log10(d)). */
+    int k_check;		/* Flag == 1 if d is near enough to a power of
+				 * ten that k must be checked. */
+    int b2, b5, s2, s5;		/* Powers of 2 and 5 in the numerator and
+				 * denominator of intermediate results. */
+    int ilim = -1, ilim1 = -1;	/* Number of digits to convert, and number to
+				 * convert if log10(d) has been
+				 * overestimated. */
+    char *retval;		/* Return value from this function. */
+    int i = -1;
+
+    /*
+     * Put the input number into a union for bit-whacking.
+     */
+
+    d.d = dv;
+
+    /*
+     * Handle the cases of negative numbers (by taking the absolute value:
+     * this includes -Inf and -NaN!), infinity, Not a Number, and zero.
+     */
+
+    TakeAbsoluteValue(&d, sign);
+    if ((d.w.word0 & EXP_MASK) == EXP_MASK) {
+	return FormatInfAndNaN(&d, decpt, endPtr);
+    }
+    if (d.d == 0.0) {
+	return FormatZero(decpt, endPtr);
+    }
+
+    /*
+     * Unpack the floating point into a wide integer and an exponent.
+     * Determine the number of bits that the big integer requires, and compute
+     * a quick approximation (which may be one too high) of ceil(log10(d.d)).
+     */
+
+    denorm = ((d.w.word0 & EXP_MASK) == 0);
+    DoubleToExpAndSig(d.d, &bw, &be, &bbits);
+    k = ApproximateLog10(bw, be, bbits);
+    k = BetterLog10(d.d, k, &k_check);
+
+    /* At this point, we have:
+     *	  d is the number to convert.
+     *    bw are significand and exponent: d == bw*2**be,
+     *    bbits is the length of bw: 2**bbits-1 <= bw < 2**bbits
+     *	  k is either ceil(log10(d)) or ceil(log10(d))+1. k_check is 0 if we
+     *      know that k is exactly ceil(log10(d)) and 1 if we need to check.
+     *    We want a rational number
+     *      r = b * 10**(1-k) = bw * 2**b2 * 5**b5 / (2**s2 / 5**s5),
+     *    with b2, b5, s2, s5 >= 0.  Note that the most significant decimal
+     *    digit is floor(r) and that successive digits can be obtained by
+     *    setting r <- 10*floor(r) (or b <= 10 * (b % S)).  Find appropriate
+     *    b2, b5, s2, s5.
+     */
+
+    ComputeScale(be, k, &b2, &b5, &s2, &s5);
+
+    /*
+     * Correct an incorrect caller-supplied 'ndigits'.  Also determine:
+     *	i = The maximum number of decimal digits that will be returned in the
+     *      formatted string.  This is k + 1 + ndigits for F format, 18 for
+     *      shortest and Steele, and ndigits for E format.
+     *  ilim = The number of significant digits to convert if k has been
+     *         guessed correctly. This is -1 for shortest and Steele (which
+     *         stop when all significance has been lost), 'ndigits' for E
+     *         format, and 'k + 1 + ndigits' for F format.
+     *  ilim1 = The minimum number of significant digits to convert if k has
+     *	        been guessed 1 too high. This, too, is -1 for shortest and
+     *	        Steele, and 'ndigits' for E format, but it's 'ndigits-1' for F
+     *	        format.
+     */
+
+    SetPrecisionLimits(convType, k, &ndigits, &i, &ilim, &ilim1);
+
+    /*
+     * Try to do low-precision conversion in floating point rather than
+     * resorting to expensive multiprecision arithmetic.
+     */
+
+    if (ilim >= 0 && ilim <= QUICK_MAX && !(flags & TCL_DD_NO_QUICK)) {
+	retval = QuickConversion(d.d, k, k_check, flags, i, ilim, ilim1,
+		decpt, endPtr);
+	if (retval != NULL) {
+	    return retval;
+	}
+    }
+
+    /*
+     * For shortening conversions, determine the upper and lower bounds for
+     * the remainder at which we can stop.
+     *   m+ = (2**m2plus * 5**m5) / (2**s2 * 5**s5) is the limit on the high
+     *        side, and
+     *   m- = (2**m2minus * 5**m5) / (2**s2 * 5**s5) is the limit on the low
+     *        side.
+     * We may need to increase s2 to put m2plus, m2minus, b2 over a common
+     * denominator.
+     */
+
+    if (flags & TCL_DD_SHORTEN_FLAG) {
+	int m2minus = b2;
+	int m2plus;
+	int m5 = b5;
+	int len = i;
+
+	/*
+	 * Find the quantity i so that (2**i*5**b5)/(2**s2*5**s5) is 1/2 unit
+	 * in the least significant place of the floating point number.
+	 */
+
+	if (denorm) {
+	    i = be + EXPONENT_BIAS + (FP_PRECISION-1);
+	} else {
+	    i = 1 + FP_PRECISION - bbits;
+	}
+	b2 += i;
+	s2 += i;
+
+	/*
+	 * Reduce the fractions to lowest terms, since the above calculation
+	 * may have left excess powers of 2 in numerator and denominator.
+	 */
+
+	CastOutPowersOf2(&b2, &m2minus, &s2);
+
+	/*
+	 * In the special case where bw==1, the nearest floating point number
+	 * to it on the low side is 1/4 ulp below it. Adjust accordingly.
+	 */
+
+	m2plus = m2minus;
+	if (!denorm && bw == 1) {
+	    ++b2;
+	    ++s2;
+	    ++m2plus;
+	}
+
+	if (s5+1 < N_LOG2POW5 && s2+1 + log2pow5[s5+1] <= 64) {
+	    /*
+	     * If 10*2**s2*5**s5 == 2**(s2+1)+5**(s5+1) fits in a 64-bit word,
+	     * then all our intermediate calculations can be done using exact
+	     * 64-bit arithmetic with no need for expensive multiprecision
+	     * operations. (This will be true for all numbers in the range
+	     * [1.0e-3 .. 1.0e+24]).
+	     */
+
+	    return ShorteningInt64Conversion(&d, convType, bw, b2, b5, m2plus,
+		    m2minus, m5, s2, s5, k, len, ilim, ilim1, decpt, endPtr);
+	} else if (s5 == 0) {
+	    /*
+	     * The denominator is a power of 2, so we can replace division by
+	     * digit shifts. First we round up s2 to a multiple of DIGIT_BIT,
+	     * and adjust m2 and b2 accordingly. Then we launch into a version
+	     * of the comparison that's specialized for the 'power of mp_digit
+	     * in the denominator' case.
+	     */
+
+	    if (s2 % DIGIT_BIT != 0) {
+		int delta = DIGIT_BIT - (s2 % DIGIT_BIT);
+
+		b2 += delta;
+		m2plus += delta;
+		m2minus += delta;
+		s2 += delta;
+	    }
+	    return ShorteningBignumConversionPowD(&d, convType, bw, b2, b5,
+		    m2plus, m2minus, m5, s2/DIGIT_BIT, k, len, ilim, ilim1,
+		    decpt, endPtr);
+	} else {
+	    /*
+	     * Alas, there's no helpful special case; use full-up bignum
+	     * arithmetic for the conversion.
+	     */
+
+	    return ShorteningBignumConversion(&d, convType, bw, b2, m2plus,
+		    m2minus, s2, s5, k, len, ilim, ilim1, decpt, endPtr);
+	}
+    } else {
+	/*
+	 * Non-shortening conversion.
+	 */
+
+	int len = i;
+
+	/*
+	 * Reduce numerator and denominator to lowest terms.
+	 */
+
+	if (b2 >= s2 && s2 > 0) {
+	    b2 -= s2; s2 = 0;
+	} else if (s2 >= b2 && b2 > 0) {
+	    s2 -= b2; b2 = 0;
+	}
+
+	if (s5+1 < N_LOG2POW5 && s2+1 + log2pow5[s5+1] <= 64) {
+	    /*
+	     * If 10*2**s2*5**s5 == 2**(s2+1)+5**(s5+1) fits in a 64-bit word,
+	     * then all our intermediate calculations can be done using exact
+	     * 64-bit arithmetic with no need for expensive multiprecision
+	     * operations.
+	     */
+
+	    return StrictInt64Conversion(&d, convType, bw, b2, b5, s2, s5, k,
+		    len, ilim, ilim1, decpt, endPtr);
+	} else if (s5 == 0) {
+	    /*
+	     * The denominator is a power of 2, so we can replace division by
+	     * digit shifts. First we round up s2 to a multiple of DIGIT_BIT,
+	     * and adjust m2 and b2 accordingly. Then we launch into a version
+	     * of the comparison that's specialized for the 'power of mp_digit
+	     * in the denominator' case.
+	     */
+
+	    if (s2 % DIGIT_BIT != 0) {
+		int delta = DIGIT_BIT - (s2 % DIGIT_BIT);
+
+		b2 += delta;
+		s2 += delta;
+	    }
+	    return StrictBignumConversionPowD(&d, convType, bw, b2, b5,
+		    s2/DIGIT_BIT, k, len, ilim, ilim1, decpt, endPtr);
+	} else {
+	    /*
+	     * There are no helpful special cases, but at least we know in
+	     * advance how many digits we will convert. We can run the
+	     * conversion in steps of DIGIT_GROUP digits, so as to have many
+	     * fewer mp_int divisions.
+	     */
+
+	    return StrictBignumConversion(&d, convType, bw, b2, s2, s5, k,
+		    len, ilim, ilim1, decpt, endPtr);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitDoubleConversion --
+ *
+ *	Initializes constants that are needed for conversions to and from
+ *	'double'
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The log base 2 of the floating point radix, the number of bits in a
+ *	double mantissa, and a table of the powers of five and ten are
+ *	computed and stored.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitDoubleConversion(void)
+{
+    int i;
+    int x;
+    Tcl_WideUInt u;
+    double d;
+#ifdef IEEE_FLOATING_POINT
+    union {
+	double dv;
+	Tcl_WideUInt iv;
+    } bitwhack;
+#endif
+#if defined(__sgi) && defined(_COMPILER_VERSION)
+    union fpc_csr mipsCR;
+
+    mipsCR.fc_word = get_fpc_csr();
+    mipsCR.fc_struct.flush = 0;
+    set_fpc_csr(mipsCR.fc_word);
+#endif
+
+    /*
+     * Initialize table of powers of 10 expressed as wide integers.
+     */
+
+    maxpow10_wide = (int)
+	    floor(sizeof(Tcl_WideUInt) * CHAR_BIT * log(2.) / log(10.));
+    pow10_wide = ckalloc((maxpow10_wide + 1) * sizeof(Tcl_WideUInt));
+    u = 1;
+    for (i = 0; i < maxpow10_wide; ++i) {
+	pow10_wide[i] = u;
+	u *= 10;
+    }
+    pow10_wide[i] = u;
+
+    /*
+     * Determine how many bits of precision a double has, and how many decimal
+     * digits that represents.
+     */
+
+    if (frexp((double) FLT_RADIX, &log2FLT_RADIX) != 0.5) {
+	Tcl_Panic("This code doesn't work on a decimal machine!");
+    }
+    log2FLT_RADIX--;
+    mantBits = DBL_MANT_DIG * log2FLT_RADIX;
+    d = 1.0;
+
+    /*
+     * Initialize a table of powers of ten that can be exactly represented in
+     * a double.
+     */
+
+    x = (int) (DBL_MANT_DIG * log((double) FLT_RADIX) / log(5.0));
+    if (x < MAXPOW) {
+	mmaxpow = x;
+    } else {
+	mmaxpow = MAXPOW;
+    }
+    for (i=0 ; i<=mmaxpow ; ++i) {
+	pow10vals[i] = d;
+	d *= 10.0;
+    }
+
+    /*
+     * Initialize a table of large powers of five.
+     */
+
+    for (i=0; i<9; ++i) {
+	mp_init(pow5 + i);
+    }
+    mp_set(pow5, 5);
+    for (i=0; i<8; ++i) {
+	mp_sqr(pow5+i, pow5+i+1);
+    }
+    mp_init_set_int(pow5_13, 1220703125);
+    for (i = 1; i < 5; ++i) {
+	mp_init(pow5_13 + i);
+	mp_sqr(pow5_13 + i - 1, pow5_13 + i);
+    }
+
+    /*
+     * Determine the number of decimal digits to the left and right of the
+     * decimal point in the largest and smallest double, the smallest double
+     * that differs from zero, and the number of mp_digits needed to represent
+     * the significand of a double.
+     */
+
+    maxDigits = (int) ((DBL_MAX_EXP * log((double) FLT_RADIX)
+	    + 0.5 * log(10.)) / log(10.));
+    minDigits = (int) floor((DBL_MIN_EXP - DBL_MANT_DIG)
+	    * log((double) FLT_RADIX) / log(10.));
+    log10_DIGIT_MAX = (int) floor(DIGIT_BIT * log(2.) / log(10.));
+
+    /*
+     * Nokia 770's software-emulated floating point is "middle endian": the
+     * bytes within a 32-bit word are little-endian (like the native
+     * integers), but the two words of a 'double' are presented most
+     * significant word first.
+     */
+
+#ifdef IEEE_FLOATING_POINT
+    bitwhack.dv = 1.000000238418579;
+				/* 3ff0 0000 4000 0000 */
+    if ((bitwhack.iv >> 32) == 0x3ff00000) {
+	n770_fp = 0;
+    } else if ((bitwhack.iv & 0xffffffff) == 0x3ff00000) {
+	n770_fp = 1;
+    } else {
+	Tcl_Panic("unknown floating point word order on this machine");
+    }
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeDoubleConversion --
+ *
+ *	Cleans up this file on exit.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Memory allocated by TclInitDoubleConversion is freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeDoubleConversion(void)
+{
+    int i;
+
+    ckfree(pow10_wide);
+    for (i=0; i<9; ++i) {
+	mp_clear(pow5 + i);
+    }
+    for (i=0; i < 5; ++i) {
+	mp_clear(pow5_13 + i);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitBignumFromDouble --
+ *
+ *	Extracts the integer part of a double and converts it to an arbitrary
+ *	precision integer.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Initializes the bignum supplied, and stores the converted number in
+ *	it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_InitBignumFromDouble(
+    Tcl_Interp *interp,		/* For error message. */
+    double d,			/* Number to convert. */
+    mp_int *b)			/* Place to store the result. */
+{
+    double fract;
+    int expt;
+
+    /*
+     * Infinite values can't convert to bignum.
+     */
+
+    if (TclIsInfinite(d)) {
+	if (interp != NULL) {
+	    const char *s = "integer value too large to represent";
+
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(s, -1));
+	    Tcl_SetErrorCode(interp, "ARITH", "IOVERFLOW", s, NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    fract = frexp(d,&expt);
+    if (expt <= 0) {
+	mp_init(b);
+	mp_zero(b);
+    } else {
+	Tcl_WideInt w = (Tcl_WideInt) ldexp(fract, mantBits);
+	int shift = expt - mantBits;
+
+	TclBNInitBignumFromWideInt(b, w);
+	if (shift < 0) {
+	    mp_div_2d(b, -shift, b, NULL);
+	} else if (shift > 0) {
+	    mp_mul_2d(b, shift, b);
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBignumToDouble --
+ *
+ *	Convert an arbitrary-precision integer to a native floating point
+ *	number.
+ *
+ * Results:
+ *	Returns the converted number. Sets errno to ERANGE if the number is
+ *	too large to convert.
+ *
+ *----------------------------------------------------------------------
+ */
+
+double
+TclBignumToDouble(
+    const mp_int *a)			/* Integer to convert. */
+{
+    mp_int b;
+    int bits, shift, i, lsb;
+    double r;
+
+
+    /*
+     * We need a 'mantBits'-bit significand.  Determine what shift will
+     * give us that.
+     */
+
+    bits = mp_count_bits(a);
+    if (bits > DBL_MAX_EXP*log2FLT_RADIX) {
+	errno = ERANGE;
+	if (a->sign == MP_ZPOS) {
+	    return HUGE_VAL;
+	} else {
+	    return -HUGE_VAL;
+	}
+    }
+    shift = mantBits - bits;
+
+    /*
+     * If shift > 0, shift the significand left by the requisite number of
+     * bits.  If shift == 0, the significand is already exactly 'mantBits'
+     * in length.  If shift < 0, we will need to shift the significand right
+     * by the requisite number of bits, and round it. If the '1-shift'
+     * least significant bits are 0, but the 'shift'th bit is nonzero,
+     * then the significand lies exactly between two values and must be
+     * 'rounded to even'.
+     */
+
+    mp_init(&b);
+    if (shift == 0) {
+	mp_copy(a, &b);
+    } else if (shift > 0) {
+	mp_mul_2d(a, shift, &b);
+    } else if (shift < 0) {
+	lsb = mp_cnt_lsb(a);
+	if (lsb == -1-shift) {
+
+	    /*
+	     * Round to even
+	     */
+
+	    mp_div_2d(a, -shift, &b, NULL);
+	    if (mp_isodd(&b)) {
+		if (b.sign == MP_ZPOS) {
+		    mp_add_d(&b, 1, &b);
+		} else {
+		    mp_sub_d(&b, 1, &b);
+		}
+	    }
+	} else {
+
+	    /*
+	     * Ordinary rounding
+	     */
+
+	    mp_div_2d(a, -1-shift, &b, NULL);
+	    if (b.sign == MP_ZPOS) {
+		mp_add_d(&b, 1, &b);
+	    } else {
+		mp_sub_d(&b, 1, &b);
+	    }
+	    mp_div_2d(&b, 1, &b, NULL);
+	}
+    }
+
+    /*
+     * Accumulate the result, one mp_digit at a time.
+     */
+
+    r = 0.0;
+    for (i=b.used-1 ; i>=0 ; --i) {
+	r = ldexp(r, DIGIT_BIT) + b.dp[i];
+    }
+    mp_clear(&b);
+
+    /*
+     * Scale the result to the correct number of bits.
+     */
+
+    r = ldexp(r, bits - mantBits);
+
+    /*
+     * Return the result with the appropriate sign.
+     */
+
+    if (a->sign == MP_ZPOS) {
+	return r;
+    } else {
+	return -r;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCeil --
+ *
+ *	Computes the smallest floating point number that is at least the
+ *	mp_int argument.
+ *
+ * Results:
+ *	Returns the floating point number.
+ *
+ *----------------------------------------------------------------------
+ */
+
+double
+TclCeil(
+    const mp_int *a)			/* Integer to convert. */
+{
+    double r = 0.0;
+    mp_int b;
+
+    mp_init(&b);
+    if (mp_cmp_d(a, 0) == MP_LT) {
+	mp_neg(a, &b);
+	r = -TclFloor(&b);
+    } else {
+	int bits = mp_count_bits(a);
+
+	if (bits > DBL_MAX_EXP*log2FLT_RADIX) {
+	    r = HUGE_VAL;
+	} else {
+	    int i, exact = 1, shift = mantBits - bits;
+
+	    if (shift > 0) {
+		mp_mul_2d(a, shift, &b);
+	    } else if (shift < 0) {
+		mp_int d;
+		mp_init(&d);
+		mp_div_2d(a, -shift, &b, &d);
+		exact = mp_iszero(&d);
+		mp_clear(&d);
+	    } else {
+		mp_copy(a, &b);
+	    }
+	    if (!exact) {
+		mp_add_d(&b, 1, &b);
+	    }
+	    for (i=b.used-1 ; i>=0 ; --i) {
+		r = ldexp(r, DIGIT_BIT) + b.dp[i];
+	    }
+	    r = ldexp(r, bits - mantBits);
+	}
+    }
+    mp_clear(&b);
+    return r;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFloor --
+ *
+ *	Computes the largest floating point number less than or equal to the
+ *	mp_int argument.
+ *
+ * Results:
+ *	Returns the floating point value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+double
+TclFloor(
+    const mp_int *a)			/* Integer to convert. */
+{
+    double r = 0.0;
+    mp_int b;
+
+    mp_init(&b);
+    if (mp_cmp_d(a, 0) == MP_LT) {
+	mp_neg(a, &b);
+	r = -TclCeil(&b);
+    } else {
+	int bits = mp_count_bits(a);
+
+	if (bits > DBL_MAX_EXP*log2FLT_RADIX) {
+	    r = DBL_MAX;
+	} else {
+	    int i, shift = mantBits - bits;
+
+	    if (shift > 0) {
+		mp_mul_2d(a, shift, &b);
+	    } else if (shift < 0) {
+		mp_div_2d(a, -shift, &b, NULL);
+	    } else {
+		mp_copy(a, &b);
+	    }
+	    for (i=b.used-1 ; i>=0 ; --i) {
+		r = ldexp(r, DIGIT_BIT) + b.dp[i];
+	    }
+	    r = ldexp(r, bits - mantBits);
+	}
+    }
+    mp_clear(&b);
+    return r;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * BignumToBiasedFrExp --
+ *
+ *	Convert an arbitrary-precision integer to a native floating point
+ *	number in the range [0.5,1) times a power of two. NOTE: Intentionally
+ *	converts to a number that's a few ulp too small, so that
+ *	RefineApproximation will not overflow near the high end of the
+ *	machine's arithmetic range.
+ *
+ * Results:
+ *	Returns the converted number.
+ *
+ * Side effects:
+ *	Stores the exponent of two in 'machexp'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+BignumToBiasedFrExp(
+    const mp_int *a,		/* Integer to convert. */
+    int *machexp)		/* Power of two. */
+{
+    mp_int b;
+    int bits;
+    int shift;
+    int i;
+    double r;
+
+    /*
+     * Determine how many bits we need, and extract that many from the input.
+     * Round to nearest unit in the last place.
+     */
+
+    bits = mp_count_bits(a);
+    shift = mantBits - 2 - bits;
+    mp_init(&b);
+    if (shift > 0) {
+	mp_mul_2d(a, shift, &b);
+    } else if (shift < 0) {
+	mp_div_2d(a, -shift, &b, NULL);
+    } else {
+	mp_copy(a, &b);
+    }
+
+    /*
+     * Accumulate the result, one mp_digit at a time.
+     */
+
+    r = 0.0;
+    for (i=b.used-1; i>=0; --i) {
+	r = ldexp(r, DIGIT_BIT) + b.dp[i];
+    }
+    mp_clear(&b);
+
+    /*
+     * Return the result with the appropriate sign.
+     */
+
+    *machexp = bits - mantBits + 2;
+    return ((a->sign == MP_ZPOS) ? r : -r);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Pow10TimesFrExp --
+ *
+ *	Multiply a power of ten by a number expressed as fraction and
+ *	exponent.
+ *
+ * Results:
+ *	Returns the significand of the result.
+ *
+ * Side effects:
+ *	Overwrites the 'machexp' parameter with the exponent of the result.
+ *
+ * Assumes that 'exponent' is such that 10**exponent would be a double, even
+ * though 'fraction*10**(machexp+exponent)' might overflow.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+Pow10TimesFrExp(
+    int exponent,		/* Power of 10 to multiply by. */
+    double fraction,		/* Significand of multiplicand. */
+    int *machexp)		/* On input, exponent of multiplicand. On
+				 * output, exponent of result. */
+{
+    int i, j;
+    int expt = *machexp;
+    double retval = fraction;
+
+    if (exponent > 0) {
+	/*
+	 * Multiply by 10**exponent.
+	 */
+
+	retval = frexp(retval * pow10vals[exponent&0xf], &j);
+	expt += j;
+	for (i=4; i<9; ++i) {
+	    if (exponent & (1<<i)) {
+		retval = frexp(retval * pow_10_2_n[i], &j);
+		expt += j;
+	    }
+	}
+    } else if (exponent < 0) {
+	/*
+	 * Divide by 10**-exponent.
+	 */
+
+	retval = frexp(retval / pow10vals[(-exponent) & 0xf], &j);
+	expt += j;
+	for (i=4; i<9; ++i) {
+	    if ((-exponent) & (1<<i)) {
+		retval = frexp(retval / pow_10_2_n[i], &j);
+		expt += j;
+	    }
+	}
+    }
+
+    *machexp = expt;
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SafeLdExp --
+ *
+ *	Do an 'ldexp' operation, but handle denormals gracefully.
+ *
+ * Results:
+ *	Returns the appropriately scaled value.
+ *
+ *	On some platforms, 'ldexp' fails when presented with a number too
+ *	small to represent as a normalized double. This routine does 'ldexp'
+ *	in two steps for those numbers, to return correctly denormalized
+ *	values.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static double
+SafeLdExp(
+    double fract,
+    int expt)
+{
+    int minexpt = DBL_MIN_EXP * log2FLT_RADIX;
+    volatile double a, b, retval;
+
+    if (expt < minexpt) {
+	a = ldexp(fract, expt - mantBits - minexpt);
+	b = ldexp(1.0, mantBits + minexpt);
+	retval = a * b;
+    } else {
+	retval = ldexp(fract, expt);
+    }
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFormatNaN --
+ *
+ *	Makes the string representation of a "Not a Number"
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores the string representation in the supplied buffer, which must be
+ *	at least TCL_DOUBLE_SPACE characters.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFormatNaN(
+    double value,		/* The Not-a-Number to format. */
+    char *buffer)		/* String representation. */
+{
+#ifndef IEEE_FLOATING_POINT
+    strcpy(buffer, "NaN");
+    return;
+#else
+    union {
+	double dv;
+	Tcl_WideUInt iv;
+    } bitwhack;
+
+    bitwhack.dv = value;
+    if (n770_fp) {
+	bitwhack.iv = Nokia770Twiddle(bitwhack.iv);
+    }
+    if (bitwhack.iv & ((Tcl_WideUInt) 1 << 63)) {
+	bitwhack.iv &= ~ ((Tcl_WideUInt) 1 << 63);
+	*buffer++ = '-';
+    }
+    *buffer++ = 'N';
+    *buffer++ = 'a';
+    *buffer++ = 'N';
+    bitwhack.iv &= (((Tcl_WideUInt) 1) << 51) - 1;
+    if (bitwhack.iv != 0) {
+	sprintf(buffer, "(%" TCL_LL_MODIFIER "x)", bitwhack.iv);
+    } else {
+	*buffer = '\0';
+    }
+#endif /* IEEE_FLOATING_POINT */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Nokia770Twiddle --
+ *
+ *	Transpose the two words of a number for Nokia 770 floating point
+ *	handling.
+ *
+ *----------------------------------------------------------------------
+ */
+#ifdef IEEE_FLOATING_POINT
+static Tcl_WideUInt
+Nokia770Twiddle(
+    Tcl_WideUInt w)		/* Number to transpose. */
+{
+    return (((w >> 32) & 0xffffffff) | (w << 32));
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNokia770Doubles --
+ *
+ *	Transpose the two words of a number for Nokia 770 floating point
+ *	handling.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNokia770Doubles(void)
+{
+    return n770_fp;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStringObj.c b/generic/tclStringObj.c
new file mode 100644
index 0000000..209f982
--- /dev/null
+++ b/generic/tclStringObj.c
@@ -0,0 +1,3846 @@
+/*
+ * tclStringObj.c --
+ *
+ *	This file contains functions that implement string operations on Tcl
+ *	objects. Some string operations work with UTF strings and others
+ *	require Unicode format. Functions that require knowledge of the width
+ *	of each character, such as indexing, operate on Unicode data.
+ *
+ *	A Unicode string is an internationalized string. Conceptually, a
+ *	Unicode string is an array of 16-bit quantities organized as a
+ *	sequence of properly formed UTF-8 characters. There is a one-to-one
+ *	map between Unicode and UTF characters. Because Unicode characters
+ *	have a fixed width, operations such as indexing operate on Unicode
+ *	data. The String object is optimized for the case where each UTF char
+ *	in a string is only one byte. In this case, we store the value of
+ *	numChars, but we don't store the Unicode data (unless Tcl_GetUnicode
+ *	is explicitly called).
+ *
+ *	The String object type stores one or both formats. The default
+ *	behavior is to store UTF. Once Unicode is calculated by a function, it
+ *	is stored in the internal rep for future access (without an additional
+ *	O(n) cost).
+ *
+ *	To allow many appends to be done to an object without constantly
+ *	reallocating the space for the string or Unicode representation, we
+ *	allocate double the space for the string or Unicode and use the
+ *	internal representation to keep track of how much space is used vs.
+ *	allocated.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+#include "tclStringRep.h"
+
+/*
+ * Prototypes for functions defined later in this file:
+ */
+
+static void		AppendPrintfToObjVA(Tcl_Obj *objPtr,
+			    const char *format, va_list argList);
+static void		AppendUnicodeToUnicodeRep(Tcl_Obj *objPtr,
+			    const Tcl_UniChar *unicode, int appendNumChars);
+static void		AppendUnicodeToUtfRep(Tcl_Obj *objPtr,
+			    const Tcl_UniChar *unicode, int numChars);
+static void		AppendUtfToUnicodeRep(Tcl_Obj *objPtr,
+			    const char *bytes, int numBytes);
+static void		AppendUtfToUtfRep(Tcl_Obj *objPtr,
+			    const char *bytes, int numBytes);
+static void		DupStringInternalRep(Tcl_Obj *objPtr,
+			    Tcl_Obj *copyPtr);
+static int		ExtendStringRepWithUnicode(Tcl_Obj *objPtr,
+			    const Tcl_UniChar *unicode, int numChars);
+static void		ExtendUnicodeRepWithString(Tcl_Obj *objPtr,
+			    const char *bytes, int numBytes,
+			    int numAppendChars);
+static void		FillUnicodeRep(Tcl_Obj *objPtr);
+static void		FreeStringInternalRep(Tcl_Obj *objPtr);
+static void		GrowStringBuffer(Tcl_Obj *objPtr, int needed, int flag);
+static void		GrowUnicodeBuffer(Tcl_Obj *objPtr, int needed);
+static int		SetStringFromAny(Tcl_Interp *interp, Tcl_Obj *objPtr);
+static void		SetUnicodeObj(Tcl_Obj *objPtr,
+			    const Tcl_UniChar *unicode, int numChars);
+static int		UnicodeLength(const Tcl_UniChar *unicode);
+static void		UpdateStringOfString(Tcl_Obj *objPtr);
+
+/*
+ * The structure below defines the string Tcl object type by means of
+ * functions that can be invoked by generic object code.
+ */
+
+const Tcl_ObjType tclStringType = {
+    "string",			/* name */
+    FreeStringInternalRep,	/* freeIntRepPro */
+    DupStringInternalRep,	/* dupIntRepProc */
+    UpdateStringOfString,	/* updateStringProc */
+    SetStringFromAny		/* setFromAnyProc */
+};
+
+/*
+ * TCL STRING GROWTH ALGORITHM
+ *
+ * When growing strings (during an append, for example), the following growth
+ * algorithm is used:
+ *
+ *   Attempt to allocate 2 * (originalLength + appendLength)
+ *   On failure:
+ *	attempt to allocate originalLength + 2*appendLength + TCL_MIN_GROWTH
+ *
+ * This algorithm allows very good performance, as it rapidly increases the
+ * memory allocated for a given string, which minimizes the number of
+ * reallocations that must be performed. However, using only the doubling
+ * algorithm can lead to a significant waste of memory. In particular, it may
+ * fail even when there is sufficient memory available to complete the append
+ * request (but there is not 2*totalLength memory available). So when the
+ * doubling fails (because there is not enough memory available), the
+ * algorithm requests a smaller amount of memory, which is still enough to
+ * cover the request, but which hopefully will be less than the total
+ * available memory.
+ *
+ * The addition of TCL_MIN_GROWTH allows for efficient handling of very
+ * small appends. Without this extra slush factor, a sequence of several small
+ * appends would cause several memory allocations. As long as
+ * TCL_MIN_GROWTH is a reasonable size, we can avoid that behavior.
+ *
+ * The growth algorithm can be tuned by adjusting the following parameters:
+ *
+ * TCL_MIN_GROWTH		Additional space, in bytes, to allocate when
+ *				the double allocation has failed. Default is
+ *				1024 (1 kilobyte).  See tclInt.h.
+ */
+
+#ifndef TCL_MIN_UNICHAR_GROWTH
+#define TCL_MIN_UNICHAR_GROWTH	TCL_MIN_GROWTH/sizeof(Tcl_UniChar)
+#endif
+
+static void
+GrowStringBuffer(
+    Tcl_Obj *objPtr,
+    int needed,
+    int flag)
+{
+    /*
+     * Pre-conditions:
+     *	objPtr->typePtr == &tclStringType
+     *	needed > stringPtr->allocated
+     *	flag || objPtr->bytes != NULL
+     */
+
+    String *stringPtr = GET_STRING(objPtr);
+    char *ptr = NULL;
+    int attempt;
+
+    if (objPtr->bytes == &tclEmptyString) {
+	objPtr->bytes = NULL;
+    }
+    if (flag == 0 || stringPtr->allocated > 0) {
+	attempt = 2 * needed;
+	if (attempt >= 0) {
+	    ptr = attemptckrealloc(objPtr->bytes, attempt + 1);
+	}
+	if (ptr == NULL) {
+	    /*
+	     * Take care computing the amount of modest growth to avoid
+	     * overflow into invalid argument values for attempt.
+	     */
+
+	    unsigned int limit = INT_MAX - needed;
+	    unsigned int extra = needed - objPtr->length + TCL_MIN_GROWTH;
+	    int growth = (int) ((extra > limit) ? limit : extra);
+
+	    attempt = needed + growth;
+	    ptr = attemptckrealloc(objPtr->bytes, attempt + 1);
+	}
+    }
+    if (ptr == NULL) {
+	/*
+	 * First allocation - just big enough; or last chance fallback.
+	 */
+
+	attempt = needed;
+	ptr = ckrealloc(objPtr->bytes, attempt + 1);
+    }
+    objPtr->bytes = ptr;
+    stringPtr->allocated = attempt;
+}
+
+static void
+GrowUnicodeBuffer(
+    Tcl_Obj *objPtr,
+    int needed)
+{
+    /*
+     * Pre-conditions:
+     *	objPtr->typePtr == &tclStringType
+     *	needed > stringPtr->maxChars
+     *	needed < STRING_MAXCHARS
+     */
+
+    String *ptr = NULL, *stringPtr = GET_STRING(objPtr);
+    int attempt;
+
+    if (stringPtr->maxChars > 0) {
+	/*
+	 * Subsequent appends - apply the growth algorithm.
+	 */
+
+	attempt = 2 * needed;
+	if (attempt >= 0 && attempt <= STRING_MAXCHARS) {
+	    ptr = stringAttemptRealloc(stringPtr, attempt);
+	}
+	if (ptr == NULL) {
+	    /*
+	     * Take care computing the amount of modest growth to avoid
+	     * overflow into invalid argument values for attempt.
+	     */
+
+	    unsigned int limit = STRING_MAXCHARS - needed;
+	    unsigned int extra = needed - stringPtr->numChars
+		    + TCL_MIN_UNICHAR_GROWTH;
+	    int growth = (int) ((extra > limit) ? limit : extra);
+
+	    attempt = needed + growth;
+	    ptr = stringAttemptRealloc(stringPtr, attempt);
+	}
+    }
+    if (ptr == NULL) {
+	/*
+	 * First allocation - just big enough; or last chance fallback.
+	 */
+
+	attempt = needed;
+	ptr = stringRealloc(stringPtr, attempt);
+    }
+    stringPtr = ptr;
+    stringPtr->maxChars = attempt;
+    SET_STRING(objPtr, stringPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_NewStringObj --
+ *
+ *	This function is normally called when not debugging: i.e., when
+ *	TCL_MEM_DEBUG is not defined. It creates a new string object and
+ *	initializes it from the byte pointer and length arguments.
+ *
+ *	When TCL_MEM_DEBUG is defined, this function just returns the result
+ *	of calling the debugging version Tcl_DbNewStringObj.
+ *
+ * Results:
+ *	A newly created string object is returned that has ref count zero.
+ *
+ * Side effects:
+ *	The new object's internal string representation will be set to a copy
+ *	of the length bytes starting at "bytes". If "length" is negative, use
+ *	bytes up to the first NUL byte; i.e., assume "bytes" points to a
+ *	C-style NUL-terminated string. The object's type is set to NULL. An
+ *	extra NUL is added to the end of the new object's byte array.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+#undef Tcl_NewStringObj
+Tcl_Obj *
+Tcl_NewStringObj(
+    const char *bytes,		/* Points to the first of the length bytes
+				 * used to initialize the new object. */
+    int length)			/* The number of bytes to copy from "bytes"
+				 * when initializing the new object. If
+				 * negative, use bytes up to the first NUL
+				 * byte. */
+{
+    return Tcl_DbNewStringObj(bytes, length, "unknown", 0);
+}
+#else /* if not TCL_MEM_DEBUG */
+Tcl_Obj *
+Tcl_NewStringObj(
+    const char *bytes,		/* Points to the first of the length bytes
+				 * used to initialize the new object. */
+    int length)			/* The number of bytes to copy from "bytes"
+				 * when initializing the new object. If
+				 * negative, use bytes up to the first NUL
+				 * byte. */
+{
+    Tcl_Obj *objPtr;
+
+    if (length < 0) {
+	length = (bytes? strlen(bytes) : 0);
+    }
+    TclNewStringObj(objPtr, bytes, length);
+    return objPtr;
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DbNewStringObj --
+ *
+ *	This function is normally called when debugging: i.e., when
+ *	TCL_MEM_DEBUG is defined. It creates new string objects. It is the
+ *	same as the Tcl_NewStringObj function above except that it calls
+ *	Tcl_DbCkalloc directly with the file name and line number from its
+ *	caller. This simplifies debugging since then the [memory active]
+ *	command will report the correct file name and line number when
+ *	reporting objects that haven't been freed.
+ *
+ *	When TCL_MEM_DEBUG is not defined, this function just returns the
+ *	result of calling Tcl_NewStringObj.
+ *
+ * Results:
+ *	A newly created string object is returned that has ref count zero.
+ *
+ * Side effects:
+ *	The new object's internal string representation will be set to a copy
+ *	of the length bytes starting at "bytes". If "length" is negative, use
+ *	bytes up to the first NUL byte; i.e., assume "bytes" points to a
+ *	C-style NUL-terminated string. The object's type is set to NULL. An
+ *	extra NUL is added to the end of the new object's byte array.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_MEM_DEBUG
+Tcl_Obj *
+Tcl_DbNewStringObj(
+    const char *bytes,		/* Points to the first of the length bytes
+				 * used to initialize the new object. */
+    int length,			/* The number of bytes to copy from "bytes"
+				 * when initializing the new object. If
+				 * negative, use bytes up to the first NUL
+				 * byte. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    Tcl_Obj *objPtr;
+
+    if (length < 0) {
+	length = (bytes? strlen(bytes) : 0);
+    }
+    TclDbNewObj(objPtr, file, line);
+    TclInitStringRep(objPtr, bytes, length);
+    return objPtr;
+}
+#else /* if not TCL_MEM_DEBUG */
+Tcl_Obj *
+Tcl_DbNewStringObj(
+    const char *bytes,		/* Points to the first of the length bytes
+				 * used to initialize the new object. */
+    int length,			/* The number of bytes to copy from "bytes"
+				 * when initializing the new object. If
+				 * negative, use bytes up to the first NUL
+				 * byte. */
+    const char *file,		/* The name of the source file calling this
+				 * function; used for debugging. */
+    int line)			/* Line number in the source file; used for
+				 * debugging. */
+{
+    return Tcl_NewStringObj(bytes, length);
+}
+#endif /* TCL_MEM_DEBUG */
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_NewUnicodeObj --
+ *
+ *	This function is creates a new String object and initializes it from
+ *	the given Unicode String. If the Utf String is the same size as the
+ *	Unicode string, don't duplicate the data.
+ *
+ * Results:
+ *	The newly created object is returned. This object will have no initial
+ *	string representation. The returned object has a ref count of 0.
+ *
+ * Side effects:
+ *	Memory allocated for new object and copy of Unicode argument.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_NewUnicodeObj(
+    const Tcl_UniChar *unicode,	/* The unicode string used to initialize the
+				 * new object. */
+    int numChars)		/* Number of characters in the unicode
+				 * string. */
+{
+    Tcl_Obj *objPtr;
+
+    TclNewObj(objPtr);
+    SetUnicodeObj(objPtr, unicode, numChars);
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetCharLength --
+ *
+ *	Get the length of the Unicode string from the Tcl object.
+ *
+ * Results:
+ *	Pointer to unicode string representing the unicode object.
+ *
+ * Side effects:
+ *	Frees old internal rep. Allocates memory for new "String" internal
+ *	rep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GetCharLength(
+    Tcl_Obj *objPtr)		/* The String object to get the num chars
+				 * of. */
+{
+    String *stringPtr;
+    int numChars;
+
+    /*
+     * Quick, no-shimmer return for short string reps.
+     */
+
+    if ((objPtr->bytes) && (objPtr->length < 2)) {
+	/* 0 bytes -> 0 chars; 1 byte -> 1 char */
+	return objPtr->length;
+    }
+
+    /*
+     * Optimize the case where we're really dealing with a bytearray object
+     * without string representation; we don't need to convert to a string to
+     * perform the get-length operation.
+     */
+
+    if (TclIsPureByteArray(objPtr)) {
+	int length;
+
+	(void) Tcl_GetByteArrayFromObj(objPtr, &length);
+	return length;
+    }
+
+    /*
+     * OK, need to work with the object as a string.
+     */
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+    numChars = stringPtr->numChars;
+
+    /*
+     * If numChars is unknown, compute it.
+     */
+
+    if (numChars == -1) {
+	TclNumUtfChars(numChars, objPtr->bytes, objPtr->length);
+	stringPtr->numChars = numChars;
+    }
+    return numChars;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetUniChar --
+ *
+ *	Get the index'th Unicode character from the String object. The index
+ *	is assumed to be in the appropriate range.
+ *
+ * Results:
+ *	Returns the index'th Unicode character in the Object.
+ *
+ * Side effects:
+ *	Fills unichar with the index'th Unicode character.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar
+Tcl_GetUniChar(
+    Tcl_Obj *objPtr,		/* The object to get the Unicode charater
+				 * from. */
+    int index)			/* Get the index'th Unicode character. */
+{
+    String *stringPtr;
+
+    /*
+     * Optimize the case where we're really dealing with a bytearray object
+     * without string representation; we don't need to convert to a string to
+     * perform the indexing operation.
+     */
+
+    if (TclIsPureByteArray(objPtr)) {
+	unsigned char *bytes = Tcl_GetByteArrayFromObj(objPtr, NULL);
+
+	return (Tcl_UniChar) bytes[index];
+    }
+
+    /*
+     * OK, need to work with the object as a string.
+     */
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (stringPtr->hasUnicode == 0) {
+	/*
+	 * If numChars is unknown, compute it.
+	 */
+
+	if (stringPtr->numChars == -1) {
+	    TclNumUtfChars(stringPtr->numChars, objPtr->bytes, objPtr->length);
+	}
+	if (stringPtr->numChars == objPtr->length) {
+	    return (Tcl_UniChar) objPtr->bytes[index];
+	}
+	FillUnicodeRep(objPtr);
+	stringPtr = GET_STRING(objPtr);
+    }
+    return stringPtr->unicode[index];
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetUnicode --
+ *
+ *	Get the Unicode form of the String object. If the object is not
+ *	already a String object, it will be converted to one. If the String
+ *	object does not have a Unicode rep, then one is create from the UTF
+ *	string format.
+ *
+ * Results:
+ *	Returns a pointer to the object's internal Unicode string.
+ *
+ * Side effects:
+ *	Converts the object to have the String internal rep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar *
+Tcl_GetUnicode(
+    Tcl_Obj *objPtr)		/* The object to find the unicode string
+				 * for. */
+{
+    return Tcl_GetUnicodeFromObj(objPtr, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetUnicodeFromObj --
+ *
+ *	Get the Unicode form of the String object with length. If the object
+ *	is not already a String object, it will be converted to one. If the
+ *	String object does not have a Unicode rep, then one is create from the
+ *	UTF string format.
+ *
+ * Results:
+ *	Returns a pointer to the object's internal Unicode string.
+ *
+ * Side effects:
+ *	Converts the object to have the String internal rep.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar *
+Tcl_GetUnicodeFromObj(
+    Tcl_Obj *objPtr,		/* The object to find the unicode string
+				 * for. */
+    int *lengthPtr)		/* If non-NULL, the location where the string
+				 * rep's unichar length should be stored. If
+				 * NULL, no length is stored. */
+{
+    String *stringPtr;
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (stringPtr->hasUnicode == 0) {
+	FillUnicodeRep(objPtr);
+	stringPtr = GET_STRING(objPtr);
+    }
+
+    if (lengthPtr != NULL) {
+	*lengthPtr = stringPtr->numChars;
+    }
+    return stringPtr->unicode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetRange --
+ *
+ *	Create a Tcl Object that contains the chars between first and last of
+ *	the object indicated by "objPtr". If the object is not already a
+ *	String object, convert it to one. The first and last indices are
+ *	assumed to be in the appropriate range.
+ *
+ * Results:
+ *	Returns a new Tcl Object of the String type.
+ *
+ * Side effects:
+ *	Changes the internal rep of "objPtr" to the String type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetRange(
+    Tcl_Obj *objPtr,		/* The Tcl object to find the range of. */
+    int first,			/* First index of the range. */
+    int last)			/* Last index of the range. */
+{
+    Tcl_Obj *newObjPtr;		/* The Tcl object to find the range of. */
+    String *stringPtr;
+
+    /*
+     * Optimize the case where we're really dealing with a bytearray object
+     * without string representation; we don't need to convert to a string to
+     * perform the substring operation.
+     */
+
+    if (TclIsPureByteArray(objPtr)) {
+	unsigned char *bytes = Tcl_GetByteArrayFromObj(objPtr, NULL);
+
+	return Tcl_NewByteArrayObj(bytes+first, last-first+1);
+    }
+
+    /*
+     * OK, need to work with the object as a string.
+     */
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (stringPtr->hasUnicode == 0) {
+	/*
+	 * If numChars is unknown, compute it.
+	 */
+
+	if (stringPtr->numChars == -1) {
+	    TclNumUtfChars(stringPtr->numChars, objPtr->bytes, objPtr->length);
+	}
+	if (stringPtr->numChars == objPtr->length) {
+	    newObjPtr = Tcl_NewStringObj(objPtr->bytes + first, last-first+1);
+
+	    /*
+	     * Since we know the char length of the result, store it.
+	     */
+
+	    SetStringFromAny(NULL, newObjPtr);
+	    stringPtr = GET_STRING(newObjPtr);
+	    stringPtr->numChars = newObjPtr->length;
+	    return newObjPtr;
+	}
+	FillUnicodeRep(objPtr);
+	stringPtr = GET_STRING(objPtr);
+    }
+
+    return Tcl_NewUnicodeObj(stringPtr->unicode + first, last-first+1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetStringObj --
+ *
+ *	Modify an object to hold a string that is a copy of the bytes
+ *	indicated by the byte pointer and length arguments.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string representation will be set to a copy of the
+ *	"length" bytes starting at "bytes". If "length" is negative, use bytes
+ *	up to the first NUL byte; i.e., assume "bytes" points to a C-style
+ *	NUL-terminated string. The object's old string and internal
+ *	representations are freed and the object's type is set NULL.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetStringObj(
+    Tcl_Obj *objPtr,		/* Object whose internal rep to init. */
+    const char *bytes,		/* Points to the first of the length bytes
+				 * used to initialize the object. */
+    int length)			/* The number of bytes to copy from "bytes"
+				 * when initializing the object. If negative,
+				 * use bytes up to the first NUL byte.*/
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetStringObj");
+    }
+
+    /*
+     * Set the type to NULL and free any internal rep for the old type.
+     */
+
+    TclFreeIntRep(objPtr);
+
+    /*
+     * Free any old string rep, then set the string rep to a copy of the
+     * length bytes starting at "bytes".
+     */
+
+    TclInvalidateStringRep(objPtr);
+    if (length < 0) {
+	length = (bytes? strlen(bytes) : 0);
+    }
+    TclInitStringRep(objPtr, bytes, length);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetObjLength --
+ *
+ *	This function changes the length of the string representation of an
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the size of objPtr's string representation is greater than length,
+ *	then it is reduced to length and a new terminating null byte is stored
+ *	in the strength. If the length of the string representation is greater
+ *	than length, the storage space is reallocated to the given length; a
+ *	null byte is stored at the end, but other bytes past the end of the
+ *	original string representation are undefined. The object's internal
+ *	representation is changed to "expendable string".
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_SetObjLength(
+    Tcl_Obj *objPtr,		/* Pointer to object. This object must not
+				 * currently be shared. */
+    int length)			/* Number of bytes desired for string
+				 * representation of object, not including
+				 * terminating null byte. */
+{
+    String *stringPtr;
+
+    if (length < 0) {
+	/*
+	 * Setting to a negative length is nonsense. This is probably the
+	 * result of overflowing the signed integer range.
+	 */
+
+	Tcl_Panic("Tcl_SetObjLength: negative length requested: "
+		"%d (integer overflow?)", length);
+    }
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetObjLength");
+    }
+
+    if (objPtr->bytes && objPtr->length == length) {
+	return;
+    }
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (objPtr->bytes != NULL) {
+	/*
+	 * Change length of an existing string rep.
+	 */
+	if (length > stringPtr->allocated) {
+	    /*
+	     * Need to enlarge the buffer.
+	     */
+	    if (objPtr->bytes == &tclEmptyString) {
+		objPtr->bytes = ckalloc(length + 1);
+	    } else {
+		objPtr->bytes = ckrealloc(objPtr->bytes, length + 1);
+	    }
+	    stringPtr->allocated = length;
+	}
+
+	objPtr->length = length;
+	objPtr->bytes[length] = 0;
+
+	/*
+	 * Invalidate the unicode data.
+	 */
+
+	stringPtr->numChars = -1;
+	stringPtr->hasUnicode = 0;
+    } else {
+	/*
+	 * Changing length of pure unicode string.
+	 */
+
+	stringCheckLimits(length);
+	if (length > stringPtr->maxChars) {
+	    stringPtr = stringRealloc(stringPtr, length);
+	    SET_STRING(objPtr, stringPtr);
+	    stringPtr->maxChars = length;
+	}
+
+	/*
+	 * Mark the new end of the unicode string
+	 */
+
+	stringPtr->numChars = length;
+	stringPtr->unicode[length] = 0;
+	stringPtr->hasUnicode = 1;
+
+	/*
+	 * Can only get here when objPtr->bytes == NULL. No need to invalidate
+	 * the string rep.
+	 */
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AttemptSetObjLength --
+ *
+ *	This function changes the length of the string representation of an
+ *	object. It uses the attempt* (non-panic'ing) memory allocators.
+ *
+ * Results:
+ *	1 if the requested memory was allocated, 0 otherwise.
+ *
+ * Side effects:
+ *	If the size of objPtr's string representation is greater than length,
+ *	then it is reduced to length and a new terminating null byte is stored
+ *	in the strength. If the length of the string representation is greater
+ *	than length, the storage space is reallocated to the given length; a
+ *	null byte is stored at the end, but other bytes past the end of the
+ *	original string representation are undefined. The object's internal
+ *	representation is changed to "expendable string".
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AttemptSetObjLength(
+    Tcl_Obj *objPtr,		/* Pointer to object. This object must not
+				 * currently be shared. */
+    int length)			/* Number of bytes desired for string
+				 * representation of object, not including
+				 * terminating null byte. */
+{
+    String *stringPtr;
+
+    if (length < 0) {
+	/*
+	 * Setting to a negative length is nonsense. This is probably the
+	 * result of overflowing the signed integer range.
+	 */
+
+	return 0;
+    }
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_AttemptSetObjLength");
+    }
+    if (objPtr->bytes && objPtr->length == length) {
+	return 1;
+    }
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (objPtr->bytes != NULL) {
+	/*
+	 * Change length of an existing string rep.
+	 */
+	if (length > stringPtr->allocated) {
+	    /*
+	     * Need to enlarge the buffer.
+	     */
+
+	    char *newBytes;
+
+	    if (objPtr->bytes == &tclEmptyString) {
+		newBytes = attemptckalloc(length + 1);
+	    } else {
+		newBytes = attemptckrealloc(objPtr->bytes, length + 1);
+	    }
+	    if (newBytes == NULL) {
+		return 0;
+	    }
+	    objPtr->bytes = newBytes;
+	    stringPtr->allocated = length;
+	}
+
+	objPtr->length = length;
+	objPtr->bytes[length] = 0;
+
+	/*
+	 * Invalidate the unicode data.
+	 */
+
+	stringPtr->numChars = -1;
+	stringPtr->hasUnicode = 0;
+    } else {
+	/*
+	 * Changing length of pure unicode string.
+	 */
+
+	if (length > STRING_MAXCHARS) {
+	    return 0;
+	}
+	if (length > stringPtr->maxChars) {
+	    stringPtr = stringAttemptRealloc(stringPtr, length);
+	    if (stringPtr == NULL) {
+		return 0;
+	    }
+	    SET_STRING(objPtr, stringPtr);
+	    stringPtr->maxChars = length;
+	}
+
+	/*
+	 * Mark the new end of the unicode string.
+	 */
+
+	stringPtr->unicode[length] = 0;
+	stringPtr->numChars = length;
+	stringPtr->hasUnicode = 1;
+
+	/*
+	 * Can only get here when objPtr->bytes == NULL. No need to invalidate
+	 * the string rep.
+	 */
+    }
+    return 1;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_SetUnicodeObj --
+ *
+ *	Modify an object to hold the Unicode string indicated by "unicode".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory allocated for new "String" internal rep.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+Tcl_SetUnicodeObj(
+    Tcl_Obj *objPtr,		/* The object to set the string of. */
+    const Tcl_UniChar *unicode,	/* The unicode string used to initialize the
+				 * object. */
+    int numChars)		/* Number of characters in the unicode
+				 * string. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_SetUnicodeObj");
+    }
+    TclFreeIntRep(objPtr);
+    SetUnicodeObj(objPtr, unicode, numChars);
+}
+
+static int
+UnicodeLength(
+    const Tcl_UniChar *unicode)
+{
+    int numChars = 0;
+
+    if (unicode) {
+	while (numChars >= 0 && unicode[numChars] != 0) {
+	    numChars++;
+	}
+    }
+    stringCheckLimits(numChars);
+    return numChars;
+}
+
+static void
+SetUnicodeObj(
+    Tcl_Obj *objPtr,		/* The object to set the string of. */
+    const Tcl_UniChar *unicode,	/* The unicode string used to initialize the
+				 * object. */
+    int numChars)		/* Number of characters in the unicode
+				 * string. */
+{
+    String *stringPtr;
+
+    if (numChars < 0) {
+	numChars = UnicodeLength(unicode);
+    }
+
+    /*
+     * Allocate enough space for the String structure + Unicode string.
+     */
+
+    stringCheckLimits(numChars);
+    stringPtr = stringAlloc(numChars);
+    SET_STRING(objPtr, stringPtr);
+    objPtr->typePtr = &tclStringType;
+
+    stringPtr->maxChars = numChars;
+    memcpy(stringPtr->unicode, unicode, numChars * sizeof(Tcl_UniChar));
+    stringPtr->unicode[numChars] = 0;
+    stringPtr->numChars = numChars;
+    stringPtr->hasUnicode = 1;
+
+    TclInvalidateStringRep(objPtr);
+    stringPtr->allocated = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendLimitedToObj --
+ *
+ *	This function appends a limited number of bytes from a sequence of
+ *	bytes to an object, marking any limitation with an ellipsis.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The bytes at *bytes are appended to the string representation of
+ *	objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendLimitedToObj(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const char *bytes,		/* Points to the bytes to append to the
+				 * object. */
+    int length,			/* The number of bytes available to be
+				 * appended from "bytes". If < 0, then all
+				 * bytes up to a NUL byte are available. */
+    int limit,			/* The maximum number of bytes to append to
+				 * the object. */
+    const char *ellipsis)	/* Ellipsis marker string, appended to the
+				 * object to indicate not all available bytes
+				 * at "bytes" were appended. */
+{
+    String *stringPtr;
+    int toCopy = 0;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_AppendLimitedToObj");
+    }
+
+    if (length < 0) {
+	length = (bytes ? strlen(bytes) : 0);
+    }
+    if (length == 0) {
+	return;
+    }
+
+    if (length <= limit) {
+	toCopy = length;
+    } else {
+	if (ellipsis == NULL) {
+	    ellipsis = "...";
+	}
+	toCopy = (bytes == NULL) ? limit
+		: Tcl_UtfPrev(bytes+limit+1-strlen(ellipsis), bytes) - bytes;
+    }
+
+    /*
+     * If objPtr has a valid Unicode rep, then append the Unicode conversion
+     * of "bytes" to the objPtr's Unicode rep, otherwise append "bytes" to
+     * objPtr's string rep.
+     */
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (stringPtr->hasUnicode && stringPtr->numChars > 0) {
+	AppendUtfToUnicodeRep(objPtr, bytes, toCopy);
+    } else {
+	AppendUtfToUtfRep(objPtr, bytes, toCopy);
+    }
+
+    if (length <= limit) {
+	return;
+    }
+
+    stringPtr = GET_STRING(objPtr);
+    if (stringPtr->hasUnicode && stringPtr->numChars > 0) {
+	AppendUtfToUnicodeRep(objPtr, ellipsis, strlen(ellipsis));
+    } else {
+	AppendUtfToUtfRep(objPtr, ellipsis, strlen(ellipsis));
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendToObj --
+ *
+ *	This function appends a sequence of bytes to an object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The bytes at *bytes are appended to the string representation of
+ *	objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendToObj(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const char *bytes,		/* Points to the bytes to append to the
+				 * object. */
+    int length)			/* The number of bytes to append from "bytes".
+				 * If < 0, then append all bytes up to NUL
+				 * byte. */
+{
+    Tcl_AppendLimitedToObj(objPtr, bytes, length, INT_MAX, NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendUnicodeToObj --
+ *
+ *	This function appends a Unicode string to an object in the most
+ *	efficient manner possible. Length must be >= 0.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Invalidates the string rep and creates a new Unicode string.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendUnicodeToObj(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const Tcl_UniChar *unicode,	/* The unicode string to append to the
+				 * object. */
+    int length)			/* Number of chars in "unicode". */
+{
+    String *stringPtr;
+
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_AppendUnicodeToObj");
+    }
+
+    if (length == 0) {
+	return;
+    }
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    /*
+     * If objPtr has a valid Unicode rep, then append the "unicode" to the
+     * objPtr's Unicode rep, otherwise the UTF conversion of "unicode" to
+     * objPtr's string rep.
+     */
+
+    if (stringPtr->hasUnicode) {
+	AppendUnicodeToUnicodeRep(objPtr, unicode, length);
+    } else {
+	AppendUnicodeToUtfRep(objPtr, unicode, length);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendObjToObj --
+ *
+ *	This function appends the string rep of one object to another.
+ *	"objPtr" cannot be a shared object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The string rep of appendObjPtr is appended to the string
+ *	representation of objPtr.
+ *	IMPORTANT: This routine does not and MUST NOT shimmer appendObjPtr.
+ *	Callers are counting on that.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendObjToObj(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    Tcl_Obj *appendObjPtr)	/* Object to append. */
+{
+    String *stringPtr;
+    int length, numChars, appendNumChars = -1;
+    const char *bytes;
+
+    /*
+     * Special case: second object is standard-empty is fast case. We know
+     * that appending nothing to anything leaves that starting anything...
+     */
+
+    if (appendObjPtr->bytes == &tclEmptyString) {
+	return;
+    }
+
+    /*
+     * Handle append of one bytearray object to another as a special case.
+     * Note that we only do this when the objects don't have string reps; if
+     * it did, then appending the byte arrays together could well lose
+     * information; this is a special-case optimization only.
+     */
+
+    if ((TclIsPureByteArray(objPtr) || objPtr->bytes == &tclEmptyString)
+	    && TclIsPureByteArray(appendObjPtr)) {
+
+	/*
+	 * You might expect the code here to be
+	 *
+	 *  bytes = Tcl_GetByteArrayFromObj(appendObjPtr, &length);
+	 *  TclAppendBytesToByteArray(objPtr, bytes, length);
+	 *
+	 * and essentially all of the time that would be fine.  However,
+	 * it would run into trouble in the case where objPtr and
+	 * appendObjPtr point to the same thing.  That may never be a
+	 * good idea.  It seems to violate Copy On Write, and we don't
+	 * have any tests for the situation, since making any Tcl commands
+	 * that call Tcl_AppendObjToObj() do that appears impossible
+	 * (They honor Copy On Write!).  For the sake of extensions that
+	 * go off into that realm, though, here's a more complex approach
+	 * that can handle all the cases.
+	 */
+
+	/* Get lengths */
+	int lengthSrc;
+
+	(void) Tcl_GetByteArrayFromObj(objPtr, &length);
+	(void) Tcl_GetByteArrayFromObj(appendObjPtr, &lengthSrc);
+
+	/* Grow buffer enough for the append */
+	TclAppendBytesToByteArray(objPtr, NULL, lengthSrc);
+
+	/* Reset objPtr back to the original value */
+	Tcl_SetByteArrayLength(objPtr, length);
+
+	/*
+	 * Now do the append knowing that buffer growth cannot cause
+	 * any trouble.
+	 */
+
+	TclAppendBytesToByteArray(objPtr,
+		Tcl_GetByteArrayFromObj(appendObjPtr, NULL), lengthSrc);
+	return;
+    }
+
+    /*
+     * Must append as strings.
+     */
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    /*
+     * If objPtr has a valid Unicode rep, then get a Unicode string from
+     * appendObjPtr and append it.
+     */
+
+    if (stringPtr->hasUnicode) {
+	/*
+	 * If appendObjPtr is not of the "String" type, don't convert it.
+	 */
+
+	if (appendObjPtr->typePtr == &tclStringType) {
+	    Tcl_UniChar *unicode =
+		    Tcl_GetUnicodeFromObj(appendObjPtr, &numChars);
+
+	    AppendUnicodeToUnicodeRep(objPtr, unicode, numChars);
+	} else {
+	    bytes = TclGetStringFromObj(appendObjPtr, &length);
+	    AppendUtfToUnicodeRep(objPtr, bytes, length);
+	}
+	return;
+    }
+
+    /*
+     * Append to objPtr's UTF string rep. If we know the number of characters
+     * in both objects before appending, then set the combined number of
+     * characters in the final (appended-to) object.
+     */
+
+    bytes = TclGetStringFromObj(appendObjPtr, &length);
+
+    numChars = stringPtr->numChars;
+    if ((numChars >= 0) && (appendObjPtr->typePtr == &tclStringType)) {
+	String *appendStringPtr = GET_STRING(appendObjPtr);
+	appendNumChars = appendStringPtr->numChars;
+    }
+
+    AppendUtfToUtfRep(objPtr, bytes, length);
+
+    if (numChars >= 0 && appendNumChars >= 0) {
+	stringPtr->numChars = numChars + appendNumChars;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AppendUnicodeToUnicodeRep --
+ *
+ *	This function appends the contents of "unicode" to the Unicode rep of
+ *	"objPtr". objPtr must already have a valid Unicode rep.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	objPtr's internal rep is reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AppendUnicodeToUnicodeRep(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const Tcl_UniChar *unicode,	/* String to append. */
+    int appendNumChars)		/* Number of chars of "unicode" to append. */
+{
+    String *stringPtr;
+    int numChars;
+
+    if (appendNumChars < 0) {
+	appendNumChars = UnicodeLength(unicode);
+    }
+    if (appendNumChars == 0) {
+	return;
+    }
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    /*
+     * If not enough space has been allocated for the unicode rep, reallocate
+     * the internal rep object with additional space. First try to double the
+     * required allocation; if that fails, try a more modest increase. See the
+     * "TCL STRING GROWTH ALGORITHM" comment at the top of this file for an
+     * explanation of this growth algorithm.
+     */
+
+    numChars = stringPtr->numChars + appendNumChars;
+    stringCheckLimits(numChars);
+
+    if (numChars > stringPtr->maxChars) {
+	int offset = -1;
+
+	/*
+	 * Protect against case where unicode points into the existing
+	 * stringPtr->unicode array. Force it to follow any relocations due to
+	 * the reallocs below.
+	 */
+
+	if (unicode && unicode >= stringPtr->unicode
+		&& unicode <= stringPtr->unicode + stringPtr->maxChars) {
+	    offset = unicode - stringPtr->unicode;
+	}
+
+	GrowUnicodeBuffer(objPtr, numChars);
+	stringPtr = GET_STRING(objPtr);
+
+	/*
+	 * Relocate unicode if needed; see above.
+	 */
+
+	if (offset >= 0) {
+	    unicode = stringPtr->unicode + offset;
+	}
+    }
+
+    /*
+     * Copy the new string onto the end of the old string, then add the
+     * trailing null.
+     */
+
+    if (unicode) {
+	memmove(stringPtr->unicode + stringPtr->numChars, unicode,
+		appendNumChars * sizeof(Tcl_UniChar));
+    }
+    stringPtr->unicode[numChars] = 0;
+    stringPtr->numChars = numChars;
+    stringPtr->allocated = 0;
+
+    TclInvalidateStringRep(objPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AppendUnicodeToUtfRep --
+ *
+ *	This function converts the contents of "unicode" to UTF and appends
+ *	the UTF to the string rep of "objPtr".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	objPtr's internal rep is reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AppendUnicodeToUtfRep(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const Tcl_UniChar *unicode,	/* String to convert to UTF. */
+    int numChars)		/* Number of chars of "unicode" to convert. */
+{
+    String *stringPtr = GET_STRING(objPtr);
+
+    numChars = ExtendStringRepWithUnicode(objPtr, unicode, numChars);
+
+    if (stringPtr->numChars != -1) {
+	stringPtr->numChars += numChars;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AppendUtfToUnicodeRep --
+ *
+ *	This function converts the contents of "bytes" to Unicode and appends
+ *	the Unicode to the Unicode rep of "objPtr". objPtr must already have a
+ *	valid Unicode rep. numBytes must be non-negative.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	objPtr's internal rep is reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AppendUtfToUnicodeRep(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const char *bytes,		/* String to convert to Unicode. */
+    int numBytes)		/* Number of bytes of "bytes" to convert. */
+{
+    String *stringPtr;
+
+    if (numBytes == 0) {
+	return;
+    }
+
+    ExtendUnicodeRepWithString(objPtr, bytes, numBytes, -1);
+    TclInvalidateStringRep(objPtr);
+    stringPtr = GET_STRING(objPtr);
+    stringPtr->allocated = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AppendUtfToUtfRep --
+ *
+ *	This function appends "numBytes" bytes of "bytes" to the UTF string
+ *	rep of "objPtr". objPtr must already have a valid String rep.
+ *	numBytes must be non-negative.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	objPtr's internal rep is reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AppendUtfToUtfRep(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    const char *bytes,		/* String to append. */
+    int numBytes)		/* Number of bytes of "bytes" to append. */
+{
+    String *stringPtr;
+    int newLength, oldLength;
+
+    if (numBytes == 0) {
+	return;
+    }
+
+    /*
+     * Copy the new string onto the end of the old string, then add the
+     * trailing null.
+     */
+
+    if (objPtr->bytes == NULL) {
+	objPtr->length = 0;
+    }
+    oldLength = objPtr->length;
+    newLength = numBytes + oldLength;
+    if (newLength < 0) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+
+    stringPtr = GET_STRING(objPtr);
+    if (newLength > stringPtr->allocated) {
+	int offset = -1;
+
+	/*
+	 * Protect against case where unicode points into the existing
+	 * stringPtr->unicode array. Force it to follow any relocations due to
+	 * the reallocs below.
+	 */
+
+	if (bytes && bytes >= objPtr->bytes
+		&& bytes <= objPtr->bytes + objPtr->length) {
+	    offset = bytes - objPtr->bytes;
+	}
+
+	/*
+	 * TODO: consider passing flag=1: no overalloc on first append. This
+	 * would make test stringObj-8.1 fail.
+	 */
+
+	GrowStringBuffer(objPtr, newLength, 0);
+
+	/*
+	 * Relocate bytes if needed; see above.
+	 */
+
+	if (offset >= 0) {
+	    bytes = objPtr->bytes + offset;
+	}
+    }
+
+    /*
+     * Invalidate the unicode data.
+     */
+
+    stringPtr->numChars = -1;
+    stringPtr->hasUnicode = 0;
+
+    if (bytes) {
+	memmove(objPtr->bytes + oldLength, bytes, numBytes);
+    }
+    objPtr->bytes[newLength] = 0;
+    objPtr->length = newLength;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendStringsToObjVA --
+ *
+ *	This function appends one or more null-terminated strings to an
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The contents of all the string arguments are appended to the string
+ *	representation of objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendStringsToObjVA(
+    Tcl_Obj *objPtr,		/* Points to the object to append to. */
+    va_list argList)		/* Variable argument list. */
+{
+    if (Tcl_IsShared(objPtr)) {
+	Tcl_Panic("%s called with shared object", "Tcl_AppendStringsToObj");
+    }
+
+    while (1) {
+	const char *bytes = va_arg(argList, char *);
+
+	if (bytes == NULL) {
+	    break;
+	}
+	Tcl_AppendToObj(objPtr, bytes, -1);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendStringsToObj --
+ *
+ *	This function appends one or more null-terminated strings to an
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The contents of all the string arguments are appended to the string
+ *	representation of objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendStringsToObj(
+    Tcl_Obj *objPtr,
+    ...)
+{
+    va_list argList;
+
+    va_start(argList, objPtr);
+    Tcl_AppendStringsToObjVA(objPtr, argList);
+    va_end(argList);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendFormatToObj --
+ *
+ *	This function appends a list of Tcl_Obj's to a Tcl_Obj according to
+ *	the formatting instructions embedded in the format string. The
+ *	formatting instructions are inspired by sprintf(). Returns TCL_OK when
+ *	successful. If there's an error in the arguments, TCL_ERROR is
+ *	returned, and an error message is written to the interp, if non-NULL.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_AppendFormatToObj(
+    Tcl_Interp *interp,
+    Tcl_Obj *appendObj,
+    const char *format,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    const char *span = format, *msg, *errCode;
+    int numBytes = 0, objIndex = 0, gotXpg = 0, gotSequential = 0;
+    int originalLength, limit;
+    static const char *mixedXPG =
+	    "cannot mix \"%\" and \"%n$\" conversion specifiers";
+    static const char *const badIndex[2] = {
+	"not enough arguments for all format specifiers",
+	"\"%n$\" argument index out of range"
+    };
+    static const char *overflow = "max size for a Tcl value exceeded";
+
+    if (Tcl_IsShared(appendObj)) {
+	Tcl_Panic("%s called with shared object", "Tcl_AppendFormatToObj");
+    }
+    TclGetStringFromObj(appendObj, &originalLength);
+    limit = INT_MAX - originalLength;
+
+    /*
+     * Format string is NUL-terminated.
+     */
+
+    while (*format != '\0') {
+	char *end;
+	int gotMinus = 0, gotHash = 0, gotZero = 0, gotSpace = 0, gotPlus = 0;
+	int width, gotPrecision, precision, sawFlag, useShort = 0, useBig = 0;
+#ifndef TCL_WIDE_INT_IS_LONG
+	int useWide = 0;
+#endif
+	int newXpg, numChars, allocSegment = 0, segmentLimit, segmentNumBytes;
+	Tcl_Obj *segment;
+	Tcl_UniChar ch = 0;
+	int step = TclUtfToUniChar(format, &ch);
+
+	format += step;
+	if (ch != '%') {
+	    numBytes += step;
+	    continue;
+	}
+	if (numBytes) {
+	    if (numBytes > limit) {
+		msg = overflow;
+		errCode = "OVERFLOW";
+		goto errorMsg;
+	    }
+	    Tcl_AppendToObj(appendObj, span, numBytes);
+	    limit -= numBytes;
+	    numBytes = 0;
+	}
+
+	/*
+	 * Saw a % : process the format specifier.
+	 *
+	 * Step 0. Handle special case of escaped format marker (i.e., %%).
+	 */
+
+	step = TclUtfToUniChar(format, &ch);
+	if (ch == '%') {
+	    span = format;
+	    numBytes = step;
+	    format += step;
+	    continue;
+	}
+
+	/*
+	 * Step 1. XPG3 position specifier
+	 */
+
+	newXpg = 0;
+	if (isdigit(UCHAR(ch))) {
+	    int position = strtoul(format, &end, 10);
+
+	    if (*end == '$') {
+		newXpg = 1;
+		objIndex = position - 1;
+		format = end + 1;
+		step = TclUtfToUniChar(format, &ch);
+	    }
+	}
+	if (newXpg) {
+	    if (gotSequential) {
+		msg = mixedXPG;
+		errCode = "MIXEDSPECTYPES";
+		goto errorMsg;
+	    }
+	    gotXpg = 1;
+	} else {
+	    if (gotXpg) {
+		msg = mixedXPG;
+		errCode = "MIXEDSPECTYPES";
+		goto errorMsg;
+	    }
+	    gotSequential = 1;
+	}
+	if ((objIndex < 0) || (objIndex >= objc)) {
+	    msg = badIndex[gotXpg];
+	    errCode = gotXpg ? "INDEXRANGE" : "FIELDVARMISMATCH";
+	    goto errorMsg;
+	}
+
+	/*
+	 * Step 2. Set of flags.
+	 */
+
+	sawFlag = 1;
+	do {
+	    switch (ch) {
+	    case '-':
+		gotMinus = 1;
+		break;
+	    case '#':
+		gotHash = 1;
+		break;
+	    case '0':
+		gotZero = 1;
+		break;
+	    case ' ':
+		gotSpace = 1;
+		break;
+	    case '+':
+		gotPlus = 1;
+		break;
+	    default:
+		sawFlag = 0;
+	    }
+	    if (sawFlag) {
+		format += step;
+		step = TclUtfToUniChar(format, &ch);
+	    }
+	} while (sawFlag);
+
+	/*
+	 * Step 3. Minimum field width.
+	 */
+
+	width = 0;
+	if (isdigit(UCHAR(ch))) {
+	    width = strtoul(format, &end, 10);
+	    format = end;
+	    step = TclUtfToUniChar(format, &ch);
+	} else if (ch == '*') {
+	    if (objIndex >= objc - 1) {
+		msg = badIndex[gotXpg];
+		errCode = gotXpg ? "INDEXRANGE" : "FIELDVARMISMATCH";
+		goto errorMsg;
+	    }
+	    if (TclGetIntFromObj(interp, objv[objIndex], &width) != TCL_OK) {
+		goto error;
+	    }
+	    if (width < 0) {
+		width = -width;
+		gotMinus = 1;
+	    }
+	    objIndex++;
+	    format += step;
+	    step = TclUtfToUniChar(format, &ch);
+	}
+	if (width > limit) {
+	    msg = overflow;
+	    errCode = "OVERFLOW";
+	    goto errorMsg;
+	}
+
+	/*
+	 * Step 4. Precision.
+	 */
+
+	gotPrecision = precision = 0;
+	if (ch == '.') {
+	    gotPrecision = 1;
+	    format += step;
+	    step = TclUtfToUniChar(format, &ch);
+	}
+	if (isdigit(UCHAR(ch))) {
+	    precision = strtoul(format, &end, 10);
+	    format = end;
+	    step = TclUtfToUniChar(format, &ch);
+	} else if (ch == '*') {
+	    if (objIndex >= objc - 1) {
+		msg = badIndex[gotXpg];
+		errCode = gotXpg ? "INDEXRANGE" : "FIELDVARMISMATCH";
+		goto errorMsg;
+	    }
+	    if (TclGetIntFromObj(interp, objv[objIndex], &precision)
+		    != TCL_OK) {
+		goto error;
+	    }
+
+	    /*
+	     * TODO: Check this truncation logic.
+	     */
+
+	    if (precision < 0) {
+		precision = 0;
+	    }
+	    objIndex++;
+	    format += step;
+	    step = TclUtfToUniChar(format, &ch);
+	}
+
+	/*
+	 * Step 5. Length modifier.
+	 */
+
+	if (ch == 'h') {
+	    useShort = 1;
+	    format += step;
+	    step = TclUtfToUniChar(format, &ch);
+	} else if (ch == 'l') {
+	    format += step;
+	    step = TclUtfToUniChar(format, &ch);
+	    if (ch == 'l') {
+		useBig = 1;
+		format += step;
+		step = TclUtfToUniChar(format, &ch);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    } else {
+		useWide = 1;
+#endif
+	    }
+	} else if (ch == 'I') {
+	    if ((format[1] == '6') && (format[2] == '4')) {
+		format += (step + 2);
+		step = Tcl_UtfToUniChar(format, &ch);
+#ifndef TCL_WIDE_INT_IS_LONG
+		useWide = 1;
+#endif
+	    } else if ((format[1] == '3') && (format[2] == '2')) {
+		format += (step + 2);
+		step = Tcl_UtfToUniChar(format, &ch);
+	    } else {
+		format += step;
+		step = Tcl_UtfToUniChar(format, &ch);
+	    }
+	} else if ((ch == 't') || (ch == 'z')) {
+	    format += step;
+	    step = Tcl_UtfToUniChar(format, &ch);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    if (sizeof(size_t) > sizeof(int)) {
+		useWide = 1;
+	    }
+#endif
+	} else if ((ch == 'q') ||(ch == 'j')) {
+	    format += step;
+	    step = Tcl_UtfToUniChar(format, &ch);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    useWide = 1;
+#endif
+	}
+
+	format += step;
+	span = format;
+
+	/*
+	 * Step 6. The actual conversion character.
+	 */
+
+	segment = objv[objIndex];
+	numChars = -1;
+	if (ch == 'i') {
+	    ch = 'd';
+	}
+	switch (ch) {
+	case '\0':
+	    msg = "format string ended in middle of field specifier";
+	    errCode = "INCOMPLETE";
+	    goto errorMsg;
+	case 's':
+	    if (gotPrecision) {
+		numChars = Tcl_GetCharLength(segment);
+		if (precision < numChars) {
+		    segment = Tcl_GetRange(segment, 0, precision - 1);
+		    numChars = precision;
+		    Tcl_IncrRefCount(segment);
+		    allocSegment = 1;
+		}
+	    }
+	    break;
+	case 'c': {
+	    char buf[TCL_UTF_MAX];
+	    int code, length;
+
+	    if (TclGetIntFromObj(interp, segment, &code) != TCL_OK) {
+		goto error;
+	    }
+	    length = Tcl_UniCharToUtf(code, buf);
+	    segment = Tcl_NewStringObj(buf, length);
+	    Tcl_IncrRefCount(segment);
+	    allocSegment = 1;
+	    break;
+	}
+
+	case 'u':
+	    if (useBig) {
+		msg = "unsigned bignum format is invalid";
+		errCode = "BADUNSIGNED";
+		goto errorMsg;
+	    }
+	case 'd':
+	case 'o':
+	case 'p':
+	case 'x':
+	case 'X':
+	case 'b': {
+	    short s = 0;	/* Silence compiler warning; only defined and
+				 * used when useShort is true. */
+	    long l;
+	    Tcl_WideInt w;
+	    mp_int big;
+	    int toAppend, isNegative = 0;
+
+#ifndef TCL_WIDE_INT_IS_LONG
+	    if (ch == 'p') {
+		useWide = 1;
+	    }
+#endif
+	    if (useBig) {
+		if (Tcl_GetBignumFromObj(interp, segment, &big) != TCL_OK) {
+		    goto error;
+		}
+		isNegative = (mp_cmp_d(&big, 0) == MP_LT);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    } else if (useWide) {
+		if (Tcl_GetWideIntFromObj(NULL, segment, &w) != TCL_OK) {
+		    Tcl_Obj *objPtr;
+
+		    if (Tcl_GetBignumFromObj(interp,segment,&big) != TCL_OK) {
+			goto error;
+		    }
+		    mp_mod_2d(&big, (int) CHAR_BIT*sizeof(Tcl_WideInt), &big);
+		    objPtr = Tcl_NewBignumObj(&big);
+		    Tcl_IncrRefCount(objPtr);
+		    Tcl_GetWideIntFromObj(NULL, objPtr, &w);
+		    Tcl_DecrRefCount(objPtr);
+		}
+		isNegative = (w < (Tcl_WideInt) 0);
+#endif
+	    } else if (TclGetLongFromObj(NULL, segment, &l) != TCL_OK) {
+		if (Tcl_GetWideIntFromObj(NULL, segment, &w) != TCL_OK) {
+		    Tcl_Obj *objPtr;
+
+		    if (Tcl_GetBignumFromObj(interp,segment,&big) != TCL_OK) {
+			goto error;
+		    }
+		    mp_mod_2d(&big, (int) CHAR_BIT * sizeof(long), &big);
+		    objPtr = Tcl_NewBignumObj(&big);
+		    Tcl_IncrRefCount(objPtr);
+		    TclGetLongFromObj(NULL, objPtr, &l);
+		    Tcl_DecrRefCount(objPtr);
+		} else {
+		    l = Tcl_WideAsLong(w);
+		}
+		if (useShort) {
+		    s = (short) l;
+		    isNegative = (s < (short) 0);
+		} else {
+		    isNegative = (l < (long) 0);
+		}
+	    } else if (useShort) {
+		s = (short) l;
+		isNegative = (s < (short) 0);
+	    } else {
+		isNegative = (l < (long) 0);
+	    }
+
+	    segment = Tcl_NewObj();
+	    allocSegment = 1;
+	    segmentLimit = INT_MAX;
+	    Tcl_IncrRefCount(segment);
+
+	    if ((isNegative || gotPlus || gotSpace) && (useBig || ch=='d')) {
+		Tcl_AppendToObj(segment,
+			(isNegative ? "-" : gotPlus ? "+" : " "), 1);
+		segmentLimit -= 1;
+	    }
+
+	    if (gotHash || (ch == 'p')) {
+		switch (ch) {
+		case 'o':
+		    Tcl_AppendToObj(segment, "0", 1);
+		    segmentLimit -= 1;
+		    precision--;
+		    break;
+		case 'X':
+		    Tcl_AppendToObj(segment, "0X", 2);
+		    segmentLimit -= 2;
+		    break;
+		case 'p':
+		case 'x':
+		    Tcl_AppendToObj(segment, "0x", 2);
+		    segmentLimit -= 2;
+		    break;
+		case 'b':
+		    Tcl_AppendToObj(segment, "0b", 2);
+		    segmentLimit -= 2;
+		    break;
+		case 'd':
+		    Tcl_AppendToObj(segment, "0d", 2);
+		    segmentLimit -= 2;
+		    break;
+		}
+	    }
+
+	    switch (ch) {
+	    case 'd': {
+		int length;
+		Tcl_Obj *pure;
+		const char *bytes;
+
+		if (useShort) {
+		    pure = Tcl_NewIntObj((int) s);
+#ifndef TCL_WIDE_INT_IS_LONG
+		} else if (useWide) {
+		    pure = Tcl_NewWideIntObj(w);
+#endif
+		} else if (useBig) {
+		    pure = Tcl_NewBignumObj(&big);
+		} else {
+		    pure = Tcl_NewLongObj(l);
+		}
+		Tcl_IncrRefCount(pure);
+		bytes = TclGetStringFromObj(pure, &length);
+
+		/*
+		 * Already did the sign above.
+		 */
+
+		if (*bytes == '-') {
+		    length--;
+		    bytes++;
+		}
+		toAppend = length;
+
+		/*
+		 * Canonical decimal string reps for integers are composed
+		 * entirely of one-byte encoded characters, so "length" is the
+		 * number of chars.
+		 */
+
+		if (gotPrecision) {
+		    if (length < precision) {
+			segmentLimit -= precision - length;
+		    }
+		    while (length < precision) {
+			Tcl_AppendToObj(segment, "0", 1);
+			length++;
+		    }
+		    gotZero = 0;
+		}
+		if (gotZero) {
+		    length += Tcl_GetCharLength(segment);
+		    if (length < width) {
+			segmentLimit -= width - length;
+		    }
+		    while (length < width) {
+			Tcl_AppendToObj(segment, "0", 1);
+			length++;
+		    }
+		}
+		if (toAppend > segmentLimit) {
+		    msg = overflow;
+		    errCode = "OVERFLOW";
+		    goto errorMsg;
+		}
+		Tcl_AppendToObj(segment, bytes, toAppend);
+		Tcl_DecrRefCount(pure);
+		break;
+	    }
+
+	    case 'u':
+	    case 'o':
+	    case 'p':
+	    case 'x':
+	    case 'X':
+	    case 'b': {
+		Tcl_WideUInt bits = (Tcl_WideUInt) 0;
+		Tcl_WideInt numDigits = (Tcl_WideInt) 0;
+		int length, numBits = 4, base = 16, index = 0, shift = 0;
+		Tcl_Obj *pure;
+		char *bytes;
+
+		if (ch == 'u') {
+		    base = 10;
+		} else if (ch == 'o') {
+		    base = 8;
+		    numBits = 3;
+		} else if (ch == 'b') {
+		    base = 2;
+		    numBits = 1;
+		}
+		if (useShort) {
+		    unsigned short us = (unsigned short) s;
+
+		    bits = (Tcl_WideUInt) us;
+		    while (us) {
+			numDigits++;
+			us /= base;
+		    }
+#ifndef TCL_WIDE_INT_IS_LONG
+		} else if (useWide) {
+		    Tcl_WideUInt uw = (Tcl_WideUInt) w;
+
+		    bits = uw;
+		    while (uw) {
+			numDigits++;
+			uw /= base;
+		    }
+#endif
+		} else if (useBig && big.used) {
+		    int leftover = (big.used * DIGIT_BIT) % numBits;
+		    mp_digit mask = (~(mp_digit)0) << (DIGIT_BIT-leftover);
+
+		    numDigits = 1 +
+			    (((Tcl_WideInt) big.used * DIGIT_BIT) / numBits);
+		    while ((mask & big.dp[big.used-1]) == 0) {
+			numDigits--;
+			mask >>= numBits;
+		    }
+		    if (numDigits > INT_MAX) {
+			msg = overflow;
+			errCode = "OVERFLOW";
+			goto errorMsg;
+		    }
+		} else if (!useBig) {
+		    unsigned long ul = (unsigned long) l;
+
+		    bits = (Tcl_WideUInt) ul;
+		    while (ul) {
+			numDigits++;
+			ul /= base;
+		    }
+		}
+
+		/*
+		 * Need to be sure zero becomes "0", not "".
+		 */
+
+		if ((numDigits == 0) && !((ch == 'o') && gotHash)) {
+		    numDigits = 1;
+		}
+		pure = Tcl_NewObj();
+		Tcl_SetObjLength(pure, (int) numDigits);
+		bytes = TclGetString(pure);
+		toAppend = length = (int) numDigits;
+		while (numDigits--) {
+		    int digitOffset;
+
+		    if (useBig && big.used) {
+			if (index < big.used && (size_t) shift <
+				CHAR_BIT*sizeof(Tcl_WideUInt) - DIGIT_BIT) {
+			    bits |= ((Tcl_WideUInt) big.dp[index++]) << shift;
+			    shift += DIGIT_BIT;
+			}
+			shift -= numBits;
+		    }
+		    digitOffset = (int) (bits % base);
+		    if (digitOffset > 9) {
+			if (ch == 'X') {
+			    bytes[numDigits] = 'A' + digitOffset - 10;
+			} else {
+			    bytes[numDigits] = 'a' + digitOffset - 10;
+			}
+		    } else {
+			bytes[numDigits] = '0' + digitOffset;
+		    }
+		    bits /= base;
+		}
+		if (useBig) {
+		    mp_clear(&big);
+		}
+		if (gotPrecision) {
+		    if (length < precision) {
+			segmentLimit -= precision - length;
+		    }
+		    while (length < precision) {
+			Tcl_AppendToObj(segment, "0", 1);
+			length++;
+		    }
+		    gotZero = 0;
+		}
+		if (gotZero) {
+		    length += Tcl_GetCharLength(segment);
+		    if (length < width) {
+			segmentLimit -= width - length;
+		    }
+		    while (length < width) {
+			Tcl_AppendToObj(segment, "0", 1);
+			length++;
+		    }
+		}
+		if (toAppend > segmentLimit) {
+		    msg = overflow;
+		    errCode = "OVERFLOW";
+		    goto errorMsg;
+		}
+		Tcl_AppendObjToObj(segment, pure);
+		Tcl_DecrRefCount(pure);
+		break;
+	    }
+
+	    }
+	    break;
+	}
+
+	case 'e':
+	case 'E':
+	case 'f':
+	case 'g':
+	case 'G': {
+#define MAX_FLOAT_SIZE 320
+	    char spec[2*TCL_INTEGER_SPACE + 9], *p = spec;
+	    double d;
+	    int length = MAX_FLOAT_SIZE;
+	    char *bytes;
+
+	    if (Tcl_GetDoubleFromObj(interp, segment, &d) != TCL_OK) {
+		/* TODO: Figure out ACCEPT_NAN here */
+		goto error;
+	    }
+	    *p++ = '%';
+	    if (gotMinus) {
+		*p++ = '-';
+	    }
+	    if (gotHash) {
+		*p++ = '#';
+	    }
+	    if (gotZero) {
+		*p++ = '0';
+	    }
+	    if (gotSpace) {
+		*p++ = ' ';
+	    }
+	    if (gotPlus) {
+		*p++ = '+';
+	    }
+	    if (width) {
+		p += sprintf(p, "%d", width);
+		if (width > length) {
+		    length = width;
+		}
+	    }
+	    if (gotPrecision) {
+		*p++ = '.';
+		p += sprintf(p, "%d", precision);
+		if (precision > INT_MAX - length) {
+		    msg = overflow;
+		    errCode = "OVERFLOW";
+		    goto errorMsg;
+		}
+		length += precision;
+	    }
+
+	    /*
+	     * Don't pass length modifiers!
+	     */
+
+	    *p++ = (char) ch;
+	    *p = '\0';
+
+	    segment = Tcl_NewObj();
+	    allocSegment = 1;
+	    if (!Tcl_AttemptSetObjLength(segment, length)) {
+		msg = overflow;
+		errCode = "OVERFLOW";
+		goto errorMsg;
+	    }
+	    bytes = TclGetString(segment);
+	    if (!Tcl_AttemptSetObjLength(segment, sprintf(bytes, spec, d))) {
+		msg = overflow;
+		errCode = "OVERFLOW";
+		goto errorMsg;
+	    }
+	    break;
+	}
+	default:
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp,
+			Tcl_ObjPrintf("bad field specifier \"%c\"", ch));
+		Tcl_SetErrorCode(interp, "TCL", "FORMAT", "BADTYPE", NULL);
+	    }
+	    goto error;
+	}
+
+	if (width>0 && numChars<0) {
+	    numChars = Tcl_GetCharLength(segment);
+	}
+	if (!gotMinus && width>0) {
+	    if (numChars < width) {
+		limit -= width - numChars;
+	    }
+	    while (numChars < width) {
+		Tcl_AppendToObj(appendObj, (gotZero ? "0" : " "), 1);
+		numChars++;
+	    }
+	}
+
+	TclGetStringFromObj(segment, &segmentNumBytes);
+	if (segmentNumBytes > limit) {
+	    if (allocSegment) {
+		Tcl_DecrRefCount(segment);
+	    }
+	    msg = overflow;
+	    errCode = "OVERFLOW";
+	    goto errorMsg;
+	}
+	Tcl_AppendObjToObj(appendObj, segment);
+	limit -= segmentNumBytes;
+	if (allocSegment) {
+	    Tcl_DecrRefCount(segment);
+	}
+	if (width > 0) {
+	    if (numChars < width) {
+		limit -= width-numChars;
+	    }
+	    while (numChars < width) {
+		Tcl_AppendToObj(appendObj, (gotZero ? "0" : " "), 1);
+		numChars++;
+	    }
+	}
+
+	objIndex += gotSequential;
+    }
+    if (numBytes) {
+	if (numBytes > limit) {
+	    msg = overflow;
+	    errCode = "OVERFLOW";
+	    goto errorMsg;
+	}
+	Tcl_AppendToObj(appendObj, span, numBytes);
+	limit -= numBytes;
+	numBytes = 0;
+    }
+
+    return TCL_OK;
+
+  errorMsg:
+    if (interp != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(msg, -1));
+	Tcl_SetErrorCode(interp, "TCL", "FORMAT", errCode, NULL);
+    }
+  error:
+    Tcl_SetObjLength(appendObj, originalLength);
+    return TCL_ERROR;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_Format--
+ *
+ * Results:
+ *	A refcount zero Tcl_Obj.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_Format(
+    Tcl_Interp *interp,
+    const char *format,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int result;
+    Tcl_Obj *objPtr = Tcl_NewObj();
+
+    result = Tcl_AppendFormatToObj(interp, objPtr, format, objc, objv);
+    if (result != TCL_OK) {
+	Tcl_DecrRefCount(objPtr);
+	return NULL;
+    }
+    return objPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * AppendPrintfToObjVA --
+ *
+ * Results:
+ *
+ * Side effects:
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+AppendPrintfToObjVA(
+    Tcl_Obj *objPtr,
+    const char *format,
+    va_list argList)
+{
+    int code, objc;
+    Tcl_Obj **objv, *list = Tcl_NewObj();
+    const char *p;
+
+    p = format;
+    Tcl_IncrRefCount(list);
+    while (*p != '\0') {
+	int size = 0, seekingConversion = 1, gotPrecision = 0;
+	int lastNum = -1;
+
+	if (*p++ != '%') {
+	    continue;
+	}
+	if (*p == '%') {
+	    p++;
+	    continue;
+	}
+	do {
+	    switch (*p) {
+	    case '\0':
+		seekingConversion = 0;
+		break;
+	    case 's': {
+		const char *q, *end, *bytes = va_arg(argList, char *);
+		seekingConversion = 0;
+
+		/*
+		 * The buffer to copy characters from starts at bytes and ends
+		 * at either the first NUL byte, or after lastNum bytes, when
+		 * caller has indicated a limit.
+		 */
+
+		end = bytes;
+		while ((!gotPrecision || lastNum--) && (*end != '\0')) {
+		    end++;
+		}
+
+		/*
+		 * Within that buffer, we trim both ends if needed so that we
+		 * copy only whole characters, and avoid copying any partial
+		 * multi-byte characters.
+		 */
+
+		q = Tcl_UtfPrev(end, bytes);
+		if (!Tcl_UtfCharComplete(q, (int)(end - q))) {
+		    end = q;
+		}
+
+		q = bytes + TCL_UTF_MAX;
+		while ((bytes < end) && (bytes < q)
+			&& ((*bytes & 0xC0) == 0x80)) {
+		    bytes++;
+		}
+
+		Tcl_ListObjAppendElement(NULL, list,
+			Tcl_NewStringObj(bytes , (int)(end - bytes)));
+
+		break;
+	    }
+	    case 'c':
+	    case 'i':
+	    case 'u':
+	    case 'd':
+	    case 'o':
+	    case 'p':
+	    case 'x':
+	    case 'X':
+		seekingConversion = 0;
+		switch (size) {
+		case -1:
+		case 0:
+		    Tcl_ListObjAppendElement(NULL, list, Tcl_NewLongObj(
+			    (long) va_arg(argList, int)));
+		    break;
+		case 1:
+		    Tcl_ListObjAppendElement(NULL, list, Tcl_NewLongObj(
+			    va_arg(argList, long)));
+		    break;
+		case 2:
+		    Tcl_ListObjAppendElement(NULL, list, Tcl_NewWideIntObj(
+			    va_arg(argList, Tcl_WideInt)));
+		    break;
+		}
+		break;
+	    case 'e':
+	    case 'E':
+	    case 'f':
+	    case 'g':
+	    case 'G':
+		Tcl_ListObjAppendElement(NULL, list, Tcl_NewDoubleObj(
+			va_arg(argList, double)));
+		seekingConversion = 0;
+		break;
+	    case '*':
+		lastNum = (int) va_arg(argList, int);
+		Tcl_ListObjAppendElement(NULL, list, Tcl_NewIntObj(lastNum));
+		p++;
+		break;
+	    case '0': case '1': case '2': case '3': case '4':
+	    case '5': case '6': case '7': case '8': case '9': {
+		char *end;
+
+		lastNum = (int) strtoul(p, &end, 10);
+		p = end;
+		break;
+	    }
+	    case '.':
+		gotPrecision = 1;
+		p++;
+		break;
+	    /* TODO: support for bignum arguments */
+	    case 'l':
+		++size;
+		p++;
+		break;
+	    case 't':
+	    case 'z':
+		if (sizeof(size_t) == sizeof(Tcl_WideInt)) {
+		    size = 2;
+		}
+		p++;
+		break;
+	    case 'j':
+	    case 'q':
+		size = 2;
+		p++;
+		break;
+	    case 'I':
+		if (p[1]=='6' && p[2]=='4') {
+		    p += 2;
+		    size = 2;
+		} else if (p[1]=='3' && p[2]=='2') {
+		    p += 2;
+		} else if (sizeof(size_t) == sizeof(Tcl_WideInt)) {
+		    size = 2;
+		}
+		p++;
+		break;
+	    case 'h':
+		size = -1;
+	    default:
+		p++;
+	    }
+	} while (seekingConversion);
+    }
+    TclListObjGetElements(NULL, list, &objc, &objv);
+    code = Tcl_AppendFormatToObj(NULL, objPtr, format, objc, objv);
+    if (code != TCL_OK) {
+	Tcl_AppendPrintfToObj(objPtr,
+		"Unable to format \"%s\" with supplied arguments: %s",
+		format, Tcl_GetString(list));
+    }
+    Tcl_DecrRefCount(list);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_AppendPrintfToObj --
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+void
+Tcl_AppendPrintfToObj(
+    Tcl_Obj *objPtr,
+    const char *format,
+    ...)
+{
+    va_list argList;
+
+    va_start(argList, format);
+    AppendPrintfToObjVA(objPtr, format, argList);
+    va_end(argList);
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_ObjPrintf --
+ *
+ * Results:
+ *	A refcount zero Tcl_Obj.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ObjPrintf(
+    const char *format,
+    ...)
+{
+    va_list argList;
+    Tcl_Obj *objPtr = Tcl_NewObj();
+
+    va_start(argList, format);
+    AppendPrintfToObjVA(objPtr, format, argList);
+    va_end(argList);
+    return objPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclGetStringStorage --
+ *
+ *	Returns the string storage space of a Tcl_Obj.
+ *
+ * Results:
+ *	The pointer value objPtr->bytes is returned and the number of bytes
+ *	allocated there is written to *sizePtr (if known).
+ *
+ * Side effects:
+ *	May set objPtr->bytes.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+char *
+TclGetStringStorage(
+    Tcl_Obj *objPtr,
+    unsigned int *sizePtr)
+{
+    String *stringPtr;
+
+    if (objPtr->typePtr != &tclStringType || objPtr->bytes == NULL) {
+	return TclGetStringFromObj(objPtr, (int *)sizePtr);
+    }
+
+    stringPtr = GET_STRING(objPtr);
+    *sizePtr = stringPtr->allocated;
+    return objPtr->bytes;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclStringRepeat --
+ *
+ *	Performs the [string repeat] function.
+ *
+ * Results:
+ * 	A standard Tcl result.
+ *
+ * Side effects:
+ * 	Writes to *objPtrPtr the address of Tcl_Obj that is concatenation
+ * 	of count copies of the value in objPtr.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclStringRepeat(
+    Tcl_Interp *interp,
+    Tcl_Obj *objPtr,
+    int count,
+    Tcl_Obj **objPtrPtr)
+{
+    Tcl_Obj *objResultPtr;
+    int length = 0, unichar = 0, done = 1;
+    int binary = TclIsPureByteArray(objPtr);
+
+    /* assert (count >= 2) */
+
+    /*
+     * Analyze to determine what representation result should be.
+     * GOALS:	Avoid shimmering & string rep generation.
+     * 		Produce pure bytearray when possible.
+     * 		Error on overflow.
+     */
+
+    if (!binary) {
+	if (objPtr->typePtr == &tclStringType) {
+	    String *stringPtr = GET_STRING(objPtr);
+	    if (stringPtr->hasUnicode) {
+		unichar = 1;
+	    }
+	}
+    }
+
+    if (binary) {
+	/* Result will be pure byte array. Pre-size it */
+	Tcl_GetByteArrayFromObj(objPtr, &length);
+    } else if (unichar) {
+	/* Result will be pure Tcl_UniChar array. Pre-size it. */
+	Tcl_GetUnicodeFromObj(objPtr, &length);
+    } else {
+	/* Result will be concat of string reps. Pre-size it. */
+	Tcl_GetStringFromObj(objPtr, &length);
+    }
+
+    if (length == 0) {
+	/* Any repeats of empty is empty. */
+	*objPtrPtr = objPtr;
+	return TCL_OK;
+    }
+
+    if (count > INT_MAX/length) {
+	if (interp) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "max size for a Tcl value (%d bytes) exceeded", INT_MAX));
+	    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    if (binary) {
+	/* Efficiently produce a pure byte array result */
+	objResultPtr = Tcl_IsShared(objPtr) ? Tcl_DuplicateObj(objPtr)
+		: objPtr;
+
+	Tcl_SetByteArrayLength(objResultPtr, count*length); /* PANIC? */
+	Tcl_SetByteArrayLength(objResultPtr, length);
+	while (count - done > done) {
+	    Tcl_AppendObjToObj(objResultPtr, objResultPtr);
+	    done *= 2;
+	}
+	TclAppendBytesToByteArray(objResultPtr,
+		Tcl_GetByteArrayFromObj(objResultPtr, NULL),
+		(count - done) * length);
+    } else if (unichar) {
+	/* Efficiently produce a pure Tcl_UniChar array result */
+	if (Tcl_IsShared(objPtr)) {
+	    objResultPtr = Tcl_NewUnicodeObj(Tcl_GetUnicode(objPtr), length);
+	} else {
+	    TclInvalidateStringRep(objPtr);
+	    objResultPtr = objPtr;
+	}
+
+        if (0 == Tcl_AttemptSetObjLength(objResultPtr, count*length)) {
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"string size overflow: unable to alloc %"
+			TCL_LL_MODIFIER "d bytes",
+			(Tcl_WideUInt)STRING_SIZE(count*length)));
+		Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjLength(objResultPtr, length);
+	while (count - done > done) {
+	    Tcl_AppendObjToObj(objResultPtr, objResultPtr);
+	    done *= 2;
+	}
+	Tcl_AppendUnicodeToObj(objResultPtr, Tcl_GetUnicode(objResultPtr),
+		(count - done) * length);
+    } else {
+	/* Efficiently concatenate string reps */
+	if (Tcl_IsShared(objPtr)) {
+	    objResultPtr = Tcl_NewStringObj(Tcl_GetString(objPtr), length);
+	} else {
+	    TclFreeIntRep(objPtr);
+	    objResultPtr = objPtr;
+	}
+        if (0 == Tcl_AttemptSetObjLength(objResultPtr, count*length)) {
+	    if (interp) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"string size overflow: unable to alloc %u bytes",
+			count*length));
+		Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjLength(objResultPtr, length);
+	while (count - done > done) {
+	    Tcl_AppendObjToObj(objResultPtr, objResultPtr);
+	    done *= 2;
+	}
+	Tcl_AppendToObj(objResultPtr, Tcl_GetString(objResultPtr),
+		(count - done) * length);
+    }
+    *objPtrPtr = objResultPtr;
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclStringCatObjv --
+ *
+ *	Performs the [string cat] function.
+ *
+ * Results:
+ * 	A standard Tcl result.
+ *
+ * Side effects:
+ * 	Writes to *objPtrPtr the address of Tcl_Obj that is concatenation
+ * 	of all objc values in objv.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclStringCatObjv(
+    Tcl_Interp *interp,
+    int inPlace,
+    int objc,
+    Tcl_Obj * const objv[],
+    Tcl_Obj **objPtrPtr)
+{
+    Tcl_Obj *objResultPtr, * const *ov;
+    int oc, length = 0, binary = 1;
+    int allowUniChar = 1, requestUniChar = 0;
+    int first = objc - 1;	/* Index of first value possibly not empty */
+    int last = 0;		/* Index of last value possibly not empty */
+
+    /* assert ( objc >= 0 ) */
+
+    if (objc <= 1) {
+	/* Only one or no objects; return first or empty */
+	*objPtrPtr = objc ? objv[0] : Tcl_NewObj();
+	return TCL_OK;
+    }
+
+    /* assert ( objc >= 2 ) */
+
+    /*
+     * Analyze to determine what representation result should be.
+     * GOALS:	Avoid shimmering & string rep generation.
+     * 		Produce pure bytearray when possible.
+     * 		Error on overflow.
+     */
+
+    ov = objv, oc = objc;
+    do {
+	Tcl_Obj *objPtr = *ov++;
+
+	if (objPtr->bytes) {
+	    /* Value has a string rep. */
+	    if (objPtr->length) {
+		/*
+		 * Non-empty string rep. Not a pure bytearray, so we
+		 * won't create a pure bytearray
+		 */
+	 	binary = 0;
+		if ((objPtr->typePtr) && (objPtr->typePtr != &tclStringType)) {
+		    /* Prevent shimmer of non-string types. */
+		    allowUniChar = 0;
+		}
+	    }
+	} else {
+	    /* assert (objPtr->typePtr != NULL) -- stork! */
+	    if (TclIsPureByteArray(objPtr)) {
+		allowUniChar = 0;
+	    } else {
+		binary = 0;
+		if (objPtr->typePtr == &tclStringType) {
+		    /* Have a pure Unicode value; ask to preserve it */
+		    requestUniChar = 1;
+		} else {
+		    /* Have another type; prevent shimmer */
+		    allowUniChar = 0;
+		}
+	    }
+	}
+    } while (--oc && (binary || allowUniChar));
+
+    if (binary) {
+	/* Result will be pure byte array. Pre-size it */
+	ov = objv; oc = objc;
+	do {
+	    Tcl_Obj *objPtr = *ov++;
+
+	    if (objPtr->bytes == NULL) {
+		int numBytes;
+
+		Tcl_GetByteArrayFromObj(objPtr, &numBytes); /* PANIC? */
+		if (numBytes) {
+		    last = objc - oc;
+		    if (length == 0) {
+			first = last;
+		    } else if (numBytes > INT_MAX - length) {
+			goto overflow;
+		    }
+		    length += numBytes;
+		}
+	    }
+	} while (--oc);
+    } else if (allowUniChar && requestUniChar) {
+	/* Result will be pure Tcl_UniChar array. Pre-size it. */
+	ov = objv; oc = objc;
+	do {
+	    Tcl_Obj *objPtr = *ov++;
+
+	    if ((objPtr->bytes == NULL) || (objPtr->length)) {
+		int numChars;
+
+		Tcl_GetUnicodeFromObj(objPtr, &numChars); /* PANIC? */
+		if (numChars) {
+		    last = objc - oc;
+		    if (length == 0) {
+			first = last;
+		    } else if (numChars > INT_MAX - length) {
+			goto overflow;
+		    }
+		    length += numChars;
+		}
+	    }
+	} while (--oc);
+    } else {
+	/* Result will be concat of string reps. Pre-size it. */
+	ov = objv; oc = objc;
+	do {
+	    Tcl_Obj *pendingPtr = NULL;
+
+	    /*
+	     * Loop until a possibly non-empty value is reached.
+	     * Keep string rep generation pending when possible.
+	     */
+
+	    do {
+		/* assert ( pendingPtr == NULL ) */
+		/* assert ( length == 0 ) */
+
+		Tcl_Obj *objPtr = *ov++;
+
+		if (objPtr->bytes == NULL) {
+		    /* No string rep; Take the chance we can avoid making it */
+		    pendingPtr = objPtr;
+		} else {
+		    Tcl_GetStringFromObj(objPtr, &length); /* PANIC? */
+		}
+	    } while (--oc && (length == 0) && (pendingPtr == NULL));
+
+	    /*
+ 	     * Either we found a possibly non-empty value, and we
+ 	     * remember this index as the first and last such value so
+ 	     * far seen, or (oc == 0) and all values are known empty,
+ 	     * so first = last = objc - 1 signals the right quick return.
+ 	     */
+
+	    first = last = objc - oc - 1;
+
+	    if (oc && (length == 0)) {
+		int numBytes;
+
+		/* assert ( pendingPtr != NULL ) */
+
+		/*
+		 * There's a pending value followed by more values.
+		 * Loop over remaining values generating strings until
+		 * a non-empty value is found, or the pending value gets
+		 * its string generated.
+		 */
+
+		do {
+		    Tcl_Obj *objPtr = *ov++;
+		    Tcl_GetStringFromObj(objPtr, &numBytes); /* PANIC? */
+		} while (--oc && numBytes == 0 && pendingPtr->bytes == NULL);
+
+		if (numBytes) {
+		    last = objc -oc -1;
+		}
+		if (oc || numBytes) {
+		    Tcl_GetStringFromObj(pendingPtr, &length);
+		}
+		if (length == 0) {
+		    if (numBytes) {
+			first = last;
+		    }
+		} else if (numBytes > INT_MAX - length) {
+		    goto overflow;
+		}
+		length += numBytes;
+	    }
+	} while (oc && (length == 0));
+
+	while (oc) {
+	    int numBytes;
+	    Tcl_Obj *objPtr = *ov++;
+
+	    /* assert ( length > 0 && pendingPtr == NULL )  */
+
+	    Tcl_GetStringFromObj(objPtr, &numBytes); /* PANIC? */
+	    if (numBytes) {
+		last = objc - oc;
+		if (numBytes > INT_MAX - length) {
+		    goto overflow;
+		}
+		length += numBytes;
+	    }
+	    --oc;
+	}
+    }
+
+    if (last <= first /*|| length == 0 */) {
+	/* Only one non-empty value or zero length; return first */
+	/* NOTE: (length == 0) implies (last <= first) */
+	*objPtrPtr = objv[first];
+	return TCL_OK;
+    }
+
+    objv += first; objc = (last - first + 1);
+
+    if (binary) {
+	/* Efficiently produce a pure byte array result */
+	unsigned char *dst;
+
+	/*
+	 * Broken interface! Byte array value routines offer no way
+	 * to handle failure to allocate enough space. Following
+	 * stanza may panic.
+	 */
+	if (inPlace && !Tcl_IsShared(*objv)) {
+	    int start;
+
+	    objResultPtr = *objv++; objc--;
+	    Tcl_GetByteArrayFromObj(objResultPtr, &start);
+	    dst = Tcl_SetByteArrayLength(objResultPtr, length) + start;
+	} else {
+	    objResultPtr = Tcl_NewByteArrayObj(NULL, length);
+	    dst = Tcl_SetByteArrayLength(objResultPtr, length);
+	}
+	while (objc--) {
+	    Tcl_Obj *objPtr = *objv++;
+
+	    if (objPtr->bytes == NULL) {
+		int more;
+		unsigned char *src = Tcl_GetByteArrayFromObj(objPtr, &more);
+		memcpy(dst, src, (size_t) more);
+		dst += more;
+	    }
+	}
+    } else if (allowUniChar && requestUniChar) {
+	/* Efficiently produce a pure Tcl_UniChar array result */
+	Tcl_UniChar *dst;
+
+	if (inPlace && !Tcl_IsShared(*objv)) {
+	    int start;
+
+	    objResultPtr = *objv++; objc--;
+
+	    /* Ugly interface! Force resize of the unicode array. */
+	    Tcl_GetUnicodeFromObj(objResultPtr, &start);
+	    Tcl_InvalidateStringRep(objResultPtr);
+	    if (0 == Tcl_AttemptSetObjLength(objResultPtr, length)) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    	"concatenation failed: unable to alloc %"
+			TCL_LL_MODIFIER "d bytes",
+			(Tcl_WideUInt)STRING_SIZE(length)));
+		    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    dst = Tcl_GetUnicode(objResultPtr) + start;
+	} else {
+	    Tcl_UniChar ch = 0;
+
+	    /* Ugly interface! No scheme to init array size. */
+	    objResultPtr = Tcl_NewUnicodeObj(&ch, 0);	/* PANIC? */
+	    if (0 == Tcl_AttemptSetObjLength(objResultPtr, length)) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    	"concatenation failed: unable to alloc %"
+			TCL_LL_MODIFIER "d bytes",
+			(Tcl_WideUInt)STRING_SIZE(length)));
+		    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    dst = Tcl_GetUnicode(objResultPtr);
+	}
+	while (objc--) {
+	    Tcl_Obj *objPtr = *objv++;
+
+	    if ((objPtr->bytes == NULL) || (objPtr->length)) {
+		int more;
+		Tcl_UniChar *src = Tcl_GetUnicodeFromObj(objPtr, &more);
+		memcpy(dst, src, more * sizeof(Tcl_UniChar));
+		dst += more;
+	    }
+	}
+    } else {
+	/* Efficiently concatenate string reps */
+	char *dst;
+
+	if (inPlace && !Tcl_IsShared(*objv)) {
+	    int start;
+
+	    objResultPtr = *objv++; objc--;
+
+	    Tcl_GetStringFromObj(objResultPtr, &start);
+	    if (0 == Tcl_AttemptSetObjLength(objResultPtr, length)) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    	"concatenation failed: unable to alloc %u bytes",
+			length));
+		    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    dst = Tcl_GetString(objResultPtr) + start;
+
+	    /* assert ( length > start ) */
+	    TclFreeIntRep(objResultPtr);
+	} else {
+	    objResultPtr = Tcl_NewObj();	/* PANIC? */
+	    if (0 == Tcl_AttemptSetObjLength(objResultPtr, length)) {
+		if (interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    	"concatenation failed: unable to alloc %u bytes",
+			length));
+		    Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    dst = Tcl_GetString(objResultPtr);
+	}
+	while (objc--) {
+	    Tcl_Obj *objPtr = *objv++;
+
+	    if ((objPtr->bytes == NULL) || (objPtr->length)) {
+		int more;
+		char *src = Tcl_GetStringFromObj(objPtr, &more);
+		memcpy(dst, src, (size_t) more);
+		dst += more;
+	    }
+	}
+    }
+    *objPtrPtr = objResultPtr;
+    return TCL_OK;
+
+  overflow:
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "max size for a Tcl value (%d bytes) exceeded", INT_MAX));
+	Tcl_SetErrorCode(interp, "TCL", "MEMORY", NULL);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclStringFind --
+ *
+ *	Implements the [string first] operation.
+ *
+ * Results:
+ *	If needle is found as a substring of haystack, the index of the
+ *	first instance of such a find is returned.  If needle is not present
+ *	as a substring of haystack, -1 is returned.
+ *
+ * Side effects:
+ *	needle and haystack may have their Tcl_ObjType changed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclStringFind(
+    Tcl_Obj *needle,
+    Tcl_Obj *haystack,
+    int start)
+{
+    int lh, ln = Tcl_GetCharLength(needle);
+
+    if (ln == 0) {
+	/*
+	 * 	We don't find empty substrings.  Bizarre!
+	 *
+	 * 	TODO: When we one day make this a true substring
+	 * 	finder, change this to "return 0"
+	 */
+	return -1;
+    }
+
+    if (TclIsPureByteArray(needle) && TclIsPureByteArray(haystack)) {
+	unsigned char *end, *try, *bh;
+	unsigned char *bn = Tcl_GetByteArrayFromObj(needle, &ln);
+
+	bh = Tcl_GetByteArrayFromObj(haystack, &lh);
+	end = bh + lh;
+
+	try = bh + start;
+	while (try + ln <= end) {
+	    try = memchr(try, bn[0], end - try);
+
+	    if (try == NULL) {
+		return -1;
+	    }
+	    if (0 == memcmp(try+1, bn+1, ln-1)) {
+		return (try - bh);
+	    }
+	    try++;
+	}
+	return -1;
+    }
+
+    lh = Tcl_GetCharLength(haystack);
+    if (haystack->bytes && (lh == haystack->length)) {
+	/* haystack is all single-byte chars */
+
+	if (needle->bytes && (ln == needle->length)) {
+	    /* needle is also all single-byte chars */
+	    char *found = strstr(haystack->bytes + start, needle->bytes);
+
+	    if (found) {
+		return (found - haystack->bytes);
+	    } else {
+		return -1;
+	    }
+	} else {
+	    /*
+	     * Cannot find substring with a multi-byte char inside
+	     * a string with no multi-byte chars.
+	     */
+	    return -1;
+	}
+    } else {
+	Tcl_UniChar *try, *end, *uh;
+	Tcl_UniChar *un = Tcl_GetUnicodeFromObj(needle, &ln);
+
+	uh = Tcl_GetUnicodeFromObj(haystack, &lh);
+	end = uh + lh;
+
+	try = uh + start;
+	while (try + ln <= end) {
+	    if ((*try == *un)
+		    && (0 == memcmp(try+1, un+1, (ln-1)*sizeof(Tcl_UniChar)))) {
+		return (try - uh);
+	    }
+	    try++;
+	}
+	return -1;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclStringLast --
+ *
+ *	Implements the [string last] operation.
+ *
+ * Results:
+ *	If needle is found as a substring of haystack, the index of the
+ *	last instance of such a find is returned.  If needle is not present
+ *	as a substring of haystack, -1 is returned.
+ *
+ * Side effects:
+ *	needle and haystack may have their Tcl_ObjType changed.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclStringLast(
+    Tcl_Obj *needle,
+    Tcl_Obj *haystack,
+    int last)
+{
+    int lh, ln = Tcl_GetCharLength(needle);
+
+    if (ln == 0) {
+	/*
+	 * 	We don't find empty substrings.  Bizarre!
+	 *
+	 * 	TODO: When we one day make this a true substring
+	 * 	finder, change this to "return 0"
+	 */
+	return -1;
+    }
+
+    if (ln > last + 1) {
+	return -1;
+    }
+
+    if (TclIsPureByteArray(needle) && TclIsPureByteArray(haystack)) {
+	unsigned char *try, *bh;
+	unsigned char *bn = Tcl_GetByteArrayFromObj(needle, &ln);
+
+	bh = Tcl_GetByteArrayFromObj(haystack, &lh);
+
+	if (last + 1 > lh) {
+	    last = lh - 1;
+	}
+	try = bh + last + 1 - ln;
+	while (try >= bh) {
+	    if ((*try == bn[0])
+		    && (0 == memcmp(try+1, bn+1, ln-1))) {
+		return (try - bh);
+	    }
+	    try--;
+	}
+	return -1;
+    }
+
+    lh = Tcl_GetCharLength(haystack);
+    if (last + 1 > lh) {
+	last = lh - 1;
+    }
+    if (haystack->bytes && (lh == haystack->length)) {
+	/* haystack is all single-byte chars */
+
+	if (needle->bytes && (ln == needle->length)) {
+	    /* needle is also all single-byte chars */
+
+	    char *try = haystack->bytes + last + 1 - ln;
+	    while (try >= haystack->bytes) {
+		if ((*try == needle->bytes[0])
+			&& (0 == memcmp(try+1, needle->bytes + 1, ln - 1))) {
+		    return (try - haystack->bytes);
+		}
+		try--;
+	    }
+	    return -1;
+	} else {
+	    /*
+	     * Cannot find substring with a multi-byte char inside
+	     * a string with no multi-byte chars.
+	     */
+	    return -1;
+	}
+    } else {
+	Tcl_UniChar *try, *uh;
+	Tcl_UniChar *un = Tcl_GetUnicodeFromObj(needle, &ln);
+
+	uh = Tcl_GetUnicodeFromObj(haystack, &lh);
+
+	try = uh + last + 1 - ln;
+	while (try >= uh) {
+	    if ((*try == un[0])
+		    && (0 == memcmp(try+1, un+1, (ln-1)*sizeof(Tcl_UniChar)))) {
+		return (try - uh);
+	    }
+	    try--;
+	}
+	return -1;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclStringObjReverse --
+ *
+ *	Implements the [string reverse] operation.
+ *
+ * Results:
+ *	An unshared Tcl value which is the [string reverse] of the argument
+ *	supplied. When sharing rules permit, the returned value might be the
+ *	argument with modifications done in place.
+ *
+ * Side effects:
+ *	May allocate a new Tcl_Obj.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+ReverseBytes(
+    unsigned char *to,		/* Copy bytes into here... */
+    unsigned char *from,	/* ...from here... */
+    int count)		/* Until this many are copied, */
+				/* reversing as you go. */
+{
+    unsigned char *src = from + count;
+    if (to == from) {
+	/* Reversing in place */
+	while (--src > to) {
+	    unsigned char c = *src;
+	    *src = *to;
+	    *to++ = c;
+	}
+    }  else {
+	while (--src >= from) {
+	    *to++ = *src;
+	}
+    }
+}
+
+Tcl_Obj *
+TclStringObjReverse(
+    Tcl_Obj *objPtr)
+{
+    String *stringPtr;
+    Tcl_UniChar ch = 0;
+
+    if (TclIsPureByteArray(objPtr)) {
+	int numBytes;
+	unsigned char *from = Tcl_GetByteArrayFromObj(objPtr, &numBytes);
+
+	if (Tcl_IsShared(objPtr)) {
+	    objPtr = Tcl_NewByteArrayObj(NULL, numBytes);
+	}
+	ReverseBytes(Tcl_GetByteArrayFromObj(objPtr, NULL), from, numBytes);
+	return objPtr;
+    }
+
+    SetStringFromAny(NULL, objPtr);
+    stringPtr = GET_STRING(objPtr);
+
+    if (stringPtr->hasUnicode) {
+	Tcl_UniChar *from = Tcl_GetUnicode(objPtr);
+	Tcl_UniChar *src = from + stringPtr->numChars;
+
+	if (Tcl_IsShared(objPtr)) {
+	    Tcl_UniChar *to;
+
+	    /*
+	     * Create a non-empty, pure unicode value, so we can coax
+	     * Tcl_SetObjLength into growing the unicode rep buffer.
+	     */
+
+	    ch = 0;
+	    objPtr = Tcl_NewUnicodeObj(&ch, 1);
+	    Tcl_SetObjLength(objPtr, stringPtr->numChars);
+	    to = Tcl_GetUnicode(objPtr);
+	    while (--src >= from) {
+		*to++ = *src;
+	    }
+	} else {
+	    /* Reversing in place */
+	    while (--src > from) {
+		ch = *src;
+		*src = *from;
+		*from++ = ch;
+	    }
+	}
+    }
+
+    if (objPtr->bytes) {
+	int numChars = stringPtr->numChars;
+	int numBytes = objPtr->length;
+	char *to, *from = objPtr->bytes;
+
+	if (Tcl_IsShared(objPtr)) {
+	    objPtr = Tcl_NewObj();
+	    Tcl_SetObjLength(objPtr, numBytes);
+	}
+	to = objPtr->bytes;
+
+	if (numChars < numBytes) {
+	    /*
+	     * Either numChars == -1 and we don't know how many chars are
+	     * represented by objPtr->bytes and we need Pass 1 just in case,
+	     * or numChars >= 0 and we know we have fewer chars than bytes,
+	     * so we know there's a multibyte character needing Pass 1.
+	     *
+	     * Pass 1. Reverse the bytes of each multi-byte character.
+	     */
+	    int charCount = 0;
+	    int bytesLeft = numBytes;
+
+	    while (bytesLeft) {
+		/*
+		 * NOTE: We know that the from buffer is NUL-terminated.
+		 * It's part of the contract for objPtr->bytes values.
+		 * Thus, we can skip calling Tcl_UtfCharComplete() here.
+		 */
+		int bytesInChar = TclUtfToUniChar(from, &ch);
+
+		ReverseBytes((unsigned char *)to, (unsigned char *)from,
+			bytesInChar);
+		to += bytesInChar;
+		from += bytesInChar;
+		bytesLeft -= bytesInChar;
+		charCount++;
+	    }
+
+	    from = to = objPtr->bytes;
+	    stringPtr->numChars = charCount;
+	}
+	/* Pass 2. Reverse all the bytes. */
+	ReverseBytes((unsigned char *)to, (unsigned char *)from, numBytes);
+    }
+
+    return objPtr;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * FillUnicodeRep --
+ *
+ *	Populate the Unicode internal rep with the Unicode form of its string
+ *	rep. The object must alread have a "String" internal rep.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Reallocates the String internal rep.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+static void
+FillUnicodeRep(
+    Tcl_Obj *objPtr)		/* The object in which to fill the unicode
+				 * rep. */
+{
+    String *stringPtr = GET_STRING(objPtr);
+
+    ExtendUnicodeRepWithString(objPtr, objPtr->bytes, objPtr->length,
+	    stringPtr->numChars);
+}
+
+static void
+ExtendUnicodeRepWithString(
+    Tcl_Obj *objPtr,
+    const char *bytes,
+    int numBytes,
+    int numAppendChars)
+{
+    String *stringPtr = GET_STRING(objPtr);
+    int needed, numOrigChars = 0;
+    Tcl_UniChar *dst;
+
+    if (stringPtr->hasUnicode) {
+	numOrigChars = stringPtr->numChars;
+    }
+    if (numAppendChars == -1) {
+	TclNumUtfChars(numAppendChars, bytes, numBytes);
+    }
+    needed = numOrigChars + numAppendChars;
+    stringCheckLimits(needed);
+
+    if (needed > stringPtr->maxChars) {
+	GrowUnicodeBuffer(objPtr, needed);
+	stringPtr = GET_STRING(objPtr);
+    }
+
+    stringPtr->hasUnicode = 1;
+    if (bytes) {
+	stringPtr->numChars = needed;
+    } else {
+	numAppendChars = 0;
+    }
+    for (dst=stringPtr->unicode + numOrigChars; numAppendChars-- > 0; dst++) {
+	bytes += TclUtfToUniChar(bytes, dst);
+    }
+    *dst = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DupStringInternalRep --
+ *
+ *	Initialize the internal representation of a new Tcl_Obj to a copy of
+ *	the internal representation of an existing string object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	copyPtr's internal rep is set to a copy of srcPtr's internal
+ *	representation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DupStringInternalRep(
+    Tcl_Obj *srcPtr,		/* Object with internal rep to copy. Must have
+				 * an internal rep of type "String". */
+    Tcl_Obj *copyPtr)		/* Object with internal rep to set. Must not
+				 * currently have an internal rep.*/
+{
+    String *srcStringPtr = GET_STRING(srcPtr);
+    String *copyStringPtr = NULL;
+
+    if (srcStringPtr->numChars == -1) {
+	/*
+	 * The String struct in the source value holds zero useful data. Don't
+	 * bother copying it. Don't even bother allocating space in which to
+	 * copy it. Just let the copy be untyped.
+	 */
+
+	return;
+    }
+
+    if (srcStringPtr->hasUnicode) {
+	int copyMaxChars;
+
+	if (srcStringPtr->maxChars / 2 >= srcStringPtr->numChars) {
+	    copyMaxChars = 2 * srcStringPtr->numChars;
+	} else {
+	    copyMaxChars = srcStringPtr->maxChars;
+	}
+	copyStringPtr = stringAttemptAlloc(copyMaxChars);
+	if (copyStringPtr == NULL) {
+	    copyMaxChars = srcStringPtr->numChars;
+	    copyStringPtr = stringAlloc(copyMaxChars);
+	}
+	copyStringPtr->maxChars = copyMaxChars;
+	memcpy(copyStringPtr->unicode, srcStringPtr->unicode,
+		srcStringPtr->numChars * sizeof(Tcl_UniChar));
+	copyStringPtr->unicode[srcStringPtr->numChars] = 0;
+    } else {
+	copyStringPtr = stringAlloc(0);
+	copyStringPtr->maxChars = 0;
+	copyStringPtr->unicode[0] = 0;
+    }
+    copyStringPtr->hasUnicode = srcStringPtr->hasUnicode;
+    copyStringPtr->numChars = srcStringPtr->numChars;
+
+    /*
+     * Tricky point: the string value was copied by generic object management
+     * code, so it doesn't contain any extra bytes that might exist in the
+     * source object.
+     */
+
+    copyStringPtr->allocated = copyPtr->bytes ? copyPtr->length : 0;
+
+    SET_STRING(copyPtr, copyStringPtr);
+    copyPtr->typePtr = &tclStringType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetStringFromAny --
+ *
+ *	Create an internal representation of type "String" for an object.
+ *
+ * Results:
+ *	This operation always succeeds and returns TCL_OK.
+ *
+ * Side effects:
+ *	Any old internal reputation for objPtr is freed and the internal
+ *	representation is set to "String".
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetStringFromAny(
+    Tcl_Interp *interp,		/* Used for error reporting if not NULL. */
+    Tcl_Obj *objPtr)		/* The object to convert. */
+{
+    if (objPtr->typePtr != &tclStringType) {
+	String *stringPtr = stringAlloc(0);
+
+	/*
+	 * Convert whatever we have into an untyped value. Just A String.
+	 */
+
+	(void) TclGetString(objPtr);
+	TclFreeIntRep(objPtr);
+
+	/*
+	 * Create a basic String intrep that just points to the UTF-8 string
+	 * already in place at objPtr->bytes.
+	 */
+
+	stringPtr->numChars = -1;
+	stringPtr->allocated = objPtr->length;
+	stringPtr->maxChars = 0;
+	stringPtr->hasUnicode = 0;
+	SET_STRING(objPtr, stringPtr);
+	objPtr->typePtr = &tclStringType;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfString --
+ *
+ *	Update the string representation for an object whose internal
+ *	representation is "String".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The object's string may be set by converting its Unicode represention
+ *	to UTF format.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfString(
+    Tcl_Obj *objPtr)		/* Object with string rep to update. */
+{
+    String *stringPtr = GET_STRING(objPtr);
+
+    /*
+     * This routine is only called when we need to generate the
+     * string rep objPtr->bytes because it does not exist -- it is NULL.
+     * In that circumstance, any lingering claim about the size of
+     * memory pointed to by that NULL pointer is clearly bogus, and
+     * needs a reset.
+     */
+
+    stringPtr->allocated = 0;
+
+    if (stringPtr->numChars == 0) {
+	TclInitStringRep(objPtr, &tclEmptyString, 0);
+    } else {
+	(void) ExtendStringRepWithUnicode(objPtr, stringPtr->unicode,
+		stringPtr->numChars);
+    }
+}
+
+static int
+ExtendStringRepWithUnicode(
+    Tcl_Obj *objPtr,
+    const Tcl_UniChar *unicode,
+    int numChars)
+{
+    /*
+     * Pre-condition: this is the "string" Tcl_ObjType.
+     */
+
+    int i, origLength, size = 0;
+    char *dst;
+    String *stringPtr = GET_STRING(objPtr);
+
+    if (numChars < 0) {
+	numChars = UnicodeLength(unicode);
+    }
+
+    if (numChars == 0) {
+	return 0;
+    }
+
+    if (objPtr->bytes == NULL) {
+	objPtr->length = 0;
+    }
+    size = origLength = objPtr->length;
+
+    /*
+     * Quick cheap check in case we have more than enough room.
+     */
+
+    if (numChars <= (INT_MAX - size)/TCL_UTF_MAX
+	    && stringPtr->allocated >= size + numChars * TCL_UTF_MAX) {
+	goto copyBytes;
+    }
+
+    for (i = 0; i < numChars && size >= 0; i++) {
+	size += TclUtfCount(unicode[i]);
+    }
+    if (size < 0) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+
+    /*
+     * Grow space if needed.
+     */
+
+    if (size > stringPtr->allocated) {
+	GrowStringBuffer(objPtr, size, 1);
+    }
+
+  copyBytes:
+    dst = objPtr->bytes + origLength;
+    for (i = 0; i < numChars; i++) {
+	dst += Tcl_UniCharToUtf((int) unicode[i], dst);
+    }
+    *dst = '\0';
+    objPtr->length = dst - objPtr->bytes;
+    return numChars;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeStringInternalRep --
+ *
+ *	Deallocate the storage associated with a String data object's internal
+ *	representation.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeStringInternalRep(
+    Tcl_Obj *objPtr)		/* Object with internal rep to free. */
+{
+    ckfree(GET_STRING(objPtr));
+    objPtr->typePtr = NULL;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStringRep.h b/generic/tclStringRep.h
new file mode 100644
index 0000000..1ef1957
--- /dev/null
+++ b/generic/tclStringRep.h
@@ -0,0 +1,97 @@
+/*
+ * tclStringRep.h --
+ *
+ *	This file contains the definition of the Unicode string internal
+ *	representation and macros to access it.
+ *
+ *	A Unicode string is an internationalized string. Conceptually, a
+ *	Unicode string is an array of 16-bit quantities organized as a
+ *	sequence of properly formed UTF-8 characters. There is a one-to-one
+ *	map between Unicode and UTF characters. Because Unicode characters
+ *	have a fixed width, operations such as indexing operate on Unicode
+ *	data. The String object is optimized for the case where each UTF char
+ *	in a string is only one byte. In this case, we store the value of
+ *	numChars, but we don't store the Unicode data (unless Tcl_GetUnicode
+ *	is explicitly called).
+ *
+ *	The String object type stores one or both formats. The default
+ *	behavior is to store UTF. Once Unicode is calculated by a function, it
+ *	is stored in the internal rep for future access (without an additional
+ *	O(n) cost).
+ *
+ *	To allow many appends to be done to an object without constantly
+ *	reallocating the space for the string or Unicode representation, we
+ *	allocate double the space for the string or Unicode and use the
+ *	internal representation to keep track of how much space is used vs.
+ *	allocated.
+ *
+ * Copyright (c) 1995-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+/*
+ * The following structure is the internal rep for a String object. It keeps
+ * track of how much memory has been used and how much has been allocated for
+ * the Unicode and UTF string to enable growing and shrinking of the UTF and
+ * Unicode reps of the String object with fewer mallocs. To optimize string
+ * length and indexing operations, this structure also stores the number of
+ * characters (same of UTF and Unicode!) once that value has been computed.
+ *
+ * Under normal configurations, what Tcl calls "Unicode" is actually UTF-16
+ * restricted to the Basic Multilingual Plane (i.e. U+00000 to U+0FFFF). This
+ * can be officially modified by altering the definition of Tcl_UniChar in
+ * tcl.h, but do not do that unless you are sure what you're doing!
+ */
+
+typedef struct {
+    int numChars;		/* The number of chars in the string. -1 means
+				 * this value has not been calculated. >= 0
+				 * means that there is a valid Unicode rep, or
+				 * that the number of UTF bytes == the number
+				 * of chars. */
+    int allocated;		/* The amount of space actually allocated for
+				 * the UTF string (minus 1 byte for the
+				 * termination char). */
+    int maxChars;		/* Max number of chars that can fit in the
+				 * space allocated for the unicode array. */
+    int hasUnicode;		/* Boolean determining whether the string has
+				 * a Unicode representation. */
+    Tcl_UniChar unicode[1];	/* The array of Unicode chars. The actual size
+				 * of this field depends on the 'maxChars'
+				 * field above. */
+} String;
+
+#define STRING_MAXCHARS \
+    (int)(((size_t)UINT_MAX - sizeof(String))/sizeof(Tcl_UniChar))
+#define STRING_SIZE(numChars) \
+    (sizeof(String) + ((numChars) * sizeof(Tcl_UniChar)))
+#define stringCheckLimits(numChars) \
+    do {								\
+	if ((numChars) < 0 || (numChars) > STRING_MAXCHARS) {		\
+	    Tcl_Panic("max length for a Tcl unicode value (%d chars) exceeded", \
+		      (int)STRING_MAXCHARS);					\
+	}								\
+    } while (0)
+#define stringAttemptAlloc(numChars) \
+    (String *) attemptckalloc(STRING_SIZE(numChars))
+#define stringAlloc(numChars) \
+    (String *) ckalloc(STRING_SIZE(numChars))
+#define stringRealloc(ptr, numChars) \
+    (String *) ckrealloc((ptr), STRING_SIZE(numChars))
+#define stringAttemptRealloc(ptr, numChars) \
+    (String *) attemptckrealloc((ptr), STRING_SIZE(numChars))
+#define GET_STRING(objPtr) \
+    ((String *) (objPtr)->internalRep.twoPtrValue.ptr1)
+#define SET_STRING(objPtr, stringPtr) \
+    ((objPtr)->internalRep.twoPtrValue.ptr1 = (void *) (stringPtr))
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStringTrim.h b/generic/tclStringTrim.h
new file mode 100644
index 0000000..030e4ec
--- /dev/null
+++ b/generic/tclStringTrim.h
@@ -0,0 +1,43 @@
+/*
+ * tclStringTrim.h --
+ *
+ *	This file contains the definition of what characters are to be trimmed
+ *	from a string by [string trim] by default. It's only needed by Tcl's
+ *	implementation; it does not form a public or private API at all.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 Scriptics Corporation.
+ * Copyright (c) 2002 ActiveState Corporation.
+ * Copyright (c) 2003-2013 Donal K. Fellows.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef TCL_STRING_TRIM_H
+#define TCL_STRING_TRIM_H
+
+/*
+ * Default set of characters to trim in [string trim] and friends. This is a
+ * UTF-8 literal string containing all Unicode space characters. [TIP #413]
+ */
+
+MODULE_SCOPE const char tclDefaultTrimSet[];
+
+/*
+ * The whitespace trimming set used when [concat]enating. This is a subset of
+ * the above, and deliberately so.
+ */
+
+#define CONCAT_TRIM_SET " \f\v\r\t\n"
+
+#endif /* TCL_STRING_TRIM_H */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStubInit.c b/generic/tclStubInit.c
new file mode 100644
index 0000000..8cc21aa
--- /dev/null
+++ b/generic/tclStubInit.c
@@ -0,0 +1,1531 @@
+/*
+ * tclStubInit.c --
+ *
+ *	This file contains the initializers for the Tcl stub vectors.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+
+#ifdef __GNUC__
+#pragma GCC dependency "tcl.decls"
+#pragma GCC dependency "tclInt.decls"
+#pragma GCC dependency "tclTomMath.decls"
+#endif
+
+/*
+ * Remove macros that will interfere with the definitions below.
+ */
+
+#undef Tcl_Alloc
+#undef Tcl_Free
+#undef Tcl_Realloc
+#undef Tcl_NewBooleanObj
+#undef Tcl_NewByteArrayObj
+#undef Tcl_NewDoubleObj
+#undef Tcl_NewIntObj
+#undef Tcl_NewListObj
+#undef Tcl_NewLongObj
+#undef Tcl_NewObj
+#undef Tcl_NewStringObj
+#undef Tcl_DumpActiveMemory
+#undef Tcl_ValidateAllMemory
+#undef Tcl_FindHashEntry
+#undef Tcl_CreateHashEntry
+#undef Tcl_Panic
+#undef Tcl_FindExecutable
+#undef TclpGetPid
+#undef TclSockMinimumBuffers
+#undef Tcl_SetIntObj
+#undef TclpInetNtoa
+#undef TclWinGetServByName
+#undef TclWinGetSockOpt
+#undef TclWinSetSockOpt
+#undef TclWinNToHS
+
+/* See bug 510001: TclSockMinimumBuffers needs plat imp */
+#if defined(_WIN64) || defined(TCL_NO_DEPRECATED)
+#   define TclSockMinimumBuffersOld 0
+#else
+#define TclSockMinimumBuffersOld sockMinimumBuffersOld
+static int TclSockMinimumBuffersOld(int sock, int size)
+{
+    return TclSockMinimumBuffers(INT2PTR(sock), size);
+}
+#endif
+
+#if defined(TCL_NO_DEPRECATED)
+#   define TclSetStartupScriptPath 0
+#   define TclGetStartupScriptPath 0
+#   define TclSetStartupScriptFileName 0
+#   define TclGetStartupScriptFileName 0
+#   define TclpInetNtoa 0
+#   define TclWinGetServByName 0
+#   define TclWinGetSockOpt 0
+#   define TclWinSetSockOpt 0
+#   define TclWinNToHS 0
+#else
+#define TclSetStartupScriptPath setStartupScriptPath
+static void TclSetStartupScriptPath(Tcl_Obj *path)
+{
+    Tcl_SetStartupScript(path, NULL);
+}
+#define TclGetStartupScriptPath getStartupScriptPath
+static Tcl_Obj *TclGetStartupScriptPath(void)
+{
+    return Tcl_GetStartupScript(NULL);
+}
+#define TclSetStartupScriptFileName setStartupScriptFileName
+static void TclSetStartupScriptFileName(
+    const char *fileName)
+{
+    Tcl_SetStartupScript(Tcl_NewStringObj(fileName,-1), NULL);
+}
+#define TclGetStartupScriptFileName getStartupScriptFileName
+static const char *TclGetStartupScriptFileName(void)
+{
+    Tcl_Obj *path = Tcl_GetStartupScript(NULL);
+    if (path == NULL) {
+	return NULL;
+    }
+    return Tcl_GetString(path);
+}
+
+#if defined(_WIN32) || defined(__CYGWIN__)
+#undef TclWinNToHS
+#define TclWinNToHS winNToHS
+static unsigned short TclWinNToHS(unsigned short ns) {
+	return ntohs(ns);
+}
+#endif
+#endif /* TCL_NO_DEPRECATED */
+
+#ifdef _WIN32
+#   define TclUnixWaitForFile 0
+#   define TclUnixCopyFile 0
+#   define TclUnixOpenTemporaryFile 0
+#   define TclpReaddir 0
+#   define TclpIsAtty 0
+#elif defined(__CYGWIN__)
+#   define TclpIsAtty TclPlatIsAtty
+#   define TclWinSetInterfaces (void (*) (int)) doNothing
+#   define TclWinAddProcess (void (*) (void *, unsigned int)) doNothing
+#   define TclWinFlushDirtyChannels doNothing
+#   define TclWinResetInterfaces doNothing
+
+static Tcl_Encoding winTCharEncoding;
+
+static int
+TclpIsAtty(int fd)
+{
+    return isatty(fd);
+}
+
+#define TclWinGetPlatformId winGetPlatformId
+static int
+TclWinGetPlatformId()
+{
+    /* Don't bother to determine the real platform on cygwin,
+     * because VER_PLATFORM_WIN32_NT is the only supported platform */
+    return 2; /* VER_PLATFORM_WIN32_NT */;
+}
+
+void *TclWinGetTclInstance()
+{
+    void *hInstance = NULL;
+    GetModuleHandleExW(GET_MODULE_HANDLE_EX_FLAG_FROM_ADDRESS,
+	    (const char *)&winTCharEncoding, &hInstance);
+    return hInstance;
+}
+
+#ifndef TCL_NO_DEPRECATED
+#define TclWinSetSockOpt winSetSockOpt
+static int
+TclWinSetSockOpt(SOCKET s, int level, int optname,
+	    const char *optval, int optlen)
+{
+    return setsockopt((int) s, level, optname, optval, optlen);
+}
+
+#define TclWinGetSockOpt winGetSockOpt
+static int
+TclWinGetSockOpt(SOCKET s, int level, int optname,
+	    char *optval, int *optlen)
+{
+    return getsockopt((int) s, level, optname, optval, optlen);
+}
+
+#define TclWinGetServByName winGetServByName
+static struct servent *
+TclWinGetServByName(const char *name, const char *proto)
+{
+    return getservbyname(name, proto);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+#define TclWinNoBackslash winNoBackslash
+static char *
+TclWinNoBackslash(char *path)
+{
+    char *p;
+
+    for (p = path; *p != '\0'; p++) {
+	if (*p == '\\') {
+	    *p = '/';
+	}
+    }
+    return path;
+}
+
+int
+TclpGetPid(Tcl_Pid pid)
+{
+    return (int) (size_t) pid;
+}
+
+static void
+doNothing(void)
+{
+    /* dummy implementation, no need to do anything */
+}
+
+char *
+Tcl_WinUtfToTChar(
+    const char *string,
+    int len,
+    Tcl_DString *dsPtr)
+{
+    if (!winTCharEncoding) {
+	winTCharEncoding = Tcl_GetEncoding(0, "unicode");
+    }
+    return Tcl_UtfToExternalDString(winTCharEncoding,
+	    string, len, dsPtr);
+}
+
+char *
+Tcl_WinTCharToUtf(
+    const char *string,
+    int len,
+    Tcl_DString *dsPtr)
+{
+    if (!winTCharEncoding) {
+	winTCharEncoding = Tcl_GetEncoding(0, "unicode");
+    }
+    return Tcl_ExternalToUtfDString(winTCharEncoding,
+	    string, len, dsPtr);
+}
+
+#if defined(TCL_WIDE_INT_IS_LONG)
+/* On Cygwin64, long is 64-bit while on Win64 long is 32-bit. Therefore
+ * we have to make sure that all stub entries on Cygwin64 follow the Win64
+ * signature. Tcl 9 must find a better solution, but that cannot be done
+ * without introducing a binary incompatibility.
+ */
+#define Tcl_DbNewLongObj ((Tcl_Obj*(*)(long,const char*,int))dbNewLongObj)
+static Tcl_Obj *dbNewLongObj(
+    int intValue,
+    const char *file,
+    int line
+) {
+#ifdef TCL_MEM_DEBUG
+    register Tcl_Obj *objPtr;
+
+    TclDbNewObj(objPtr, file, line);
+    objPtr->bytes = NULL;
+
+    objPtr->internalRep.longValue = (long) intValue;
+    objPtr->typePtr = &tclIntType;
+    return objPtr;
+#else
+    return Tcl_NewIntObj(intValue);
+#endif
+}
+#define Tcl_GetLongFromObj (int(*)(Tcl_Interp*,Tcl_Obj*,long*))Tcl_GetIntFromObj
+#define Tcl_NewLongObj (Tcl_Obj*(*)(long))Tcl_NewIntObj
+#define Tcl_SetLongObj (void(*)(Tcl_Obj*,long))Tcl_SetIntObj
+static int exprInt(Tcl_Interp *interp, const char *expr, int *ptr){
+    long longValue;
+    int result = Tcl_ExprLong(interp, expr, &longValue);
+    if (result == TCL_OK) {
+	    if ((longValue >= -(long)(UINT_MAX))
+		    && (longValue <= (long)(UINT_MAX))) {
+	    *ptr = (int)longValue;
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "integer value too large to represent as non-long integer", -1));
+	    result = TCL_ERROR;
+	}
+    }
+    return result;
+}
+#define Tcl_ExprLong (int(*)(Tcl_Interp*,const char*,long*))exprInt
+static int exprIntObj(Tcl_Interp *interp, Tcl_Obj*expr, int *ptr){
+    long longValue;
+    int result = Tcl_ExprLongObj(interp, expr, &longValue);
+    if (result == TCL_OK) {
+	    if ((longValue >= -(long)(UINT_MAX))
+		    && (longValue <= (long)(UINT_MAX))) {
+	    *ptr = (int)longValue;
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "integer value too large to represent as non-long integer", -1));
+	    result = TCL_ERROR;
+	}
+    }
+    return result;
+}
+#define Tcl_ExprLongObj (int(*)(Tcl_Interp*,Tcl_Obj*,long*))exprIntObj
+static int uniCharNcmp(const Tcl_UniChar *ucs, const Tcl_UniChar *uct, unsigned int n){
+   return Tcl_UniCharNcmp(ucs, uct, (unsigned long)n);
+}
+#define Tcl_UniCharNcmp (int(*)(const Tcl_UniChar*,const Tcl_UniChar*,unsigned long))uniCharNcmp
+static int utfNcmp(const char *s1, const char *s2, unsigned int n){
+   return Tcl_UtfNcmp(s1, s2, (unsigned long)n);
+}
+#define Tcl_UtfNcmp (int(*)(const char*,const char*,unsigned long))utfNcmp
+static int utfNcasecmp(const char *s1, const char *s2, unsigned int n){
+   return Tcl_UtfNcasecmp(s1, s2, (unsigned long)n);
+}
+#define Tcl_UtfNcasecmp (int(*)(const char*,const char*,unsigned long))utfNcasecmp
+static int uniCharNcasecmp(const Tcl_UniChar *ucs, const Tcl_UniChar *uct, unsigned int n){
+   return Tcl_UniCharNcasecmp(ucs, uct, (unsigned long)n);
+}
+#define Tcl_UniCharNcasecmp (int(*)(const Tcl_UniChar*,const Tcl_UniChar*,unsigned long))uniCharNcasecmp
+static int formatInt(char *buffer, int n){
+   return TclFormatInt(buffer, (long)n);
+}
+#define TclFormatInt (int(*)(char *, long))formatInt
+
+#endif /* TCL_WIDE_INT_IS_LONG */
+
+#endif /* __CYGWIN__ */
+
+#ifdef TCL_NO_DEPRECATED
+#   define Tcl_SeekOld 0
+#   define Tcl_TellOld 0
+#   undef Tcl_SetBooleanObj
+#   define Tcl_SetBooleanObj 0
+#   undef Tcl_PkgPresent
+#   define Tcl_PkgPresent 0
+#   undef Tcl_PkgProvide
+#   define Tcl_PkgProvide 0
+#   undef Tcl_PkgRequire
+#   define Tcl_PkgRequire 0
+#   undef Tcl_GetIndexFromObj
+#   define Tcl_GetIndexFromObj 0
+#   define Tcl_NewBooleanObj 0
+#   undef Tcl_DbNewBooleanObj
+#   define Tcl_DbNewBooleanObj 0
+#   undef Tcl_SetBooleanObj
+#   define Tcl_SetBooleanObj 0
+#   undef Tcl_SetVar
+#   define Tcl_SetVar 0
+#   undef Tcl_UnsetVar
+#   define Tcl_UnsetVar 0
+#   undef Tcl_GetVar
+#   define Tcl_GetVar 0
+#   undef Tcl_TraceVar
+#   define Tcl_TraceVar 0
+#   undef Tcl_UntraceVar
+#   define Tcl_UntraceVar 0
+#   undef Tcl_VarTraceInfo
+#   define Tcl_VarTraceInfo 0
+#   undef Tcl_UpVar
+#   define Tcl_UpVar 0
+#   undef Tcl_AddErrorInfo
+#   define Tcl_AddErrorInfo 0
+#   undef Tcl_AddObjErrorInfo
+#   define Tcl_AddObjErrorInfo 0
+#   undef Tcl_Eval
+#   define Tcl_Eval 0
+#   undef Tcl_GlobalEval
+#   define Tcl_GlobalEval 0
+#   undef Tcl_SaveResult
+#   define Tcl_SaveResult 0
+#   undef Tcl_RestoreResult
+#   define Tcl_RestoreResult 0
+#   undef Tcl_DiscardResult
+#   define Tcl_DiscardResult 0
+#   undef Tcl_SetResult
+#   define Tcl_SetResult 0
+#   undef Tcl_EvalObj
+#   define Tcl_EvalObj 0
+#   undef Tcl_GlobalEvalObj
+#   define Tcl_GlobalEvalObj 0
+#   define TclBackgroundException 0
+#   undef TclpReaddir
+#   define TclpReaddir 0
+#   undef TclpGetDate
+#   define TclpGetDate 0
+#   undef TclpLocaltime
+#   define TclpLocaltime 0
+#   undef TclpGmtime
+#   define TclpGmtime 0
+#   define TclpLocaltime_unix 0
+#   define TclpGmtime_unix 0
+#else /* TCL_NO_DEPRECATED */
+#   define Tcl_SeekOld seekOld
+#   define Tcl_TellOld tellOld
+#   define TclBackgroundException Tcl_BackgroundException
+#   define TclpLocaltime_unix TclpLocaltime
+#   define TclpGmtime_unix TclpGmtime
+
+static int
+seekOld(
+    Tcl_Channel chan,		/* The channel on which to seek. */
+    int offset,			/* Offset to seek to. */
+    int mode)			/* Relative to which location to seek? */
+{
+    Tcl_WideInt wOffset, wResult;
+
+    wOffset = Tcl_LongAsWide((long) offset);
+    wResult = Tcl_Seek(chan, wOffset, mode);
+    return (int) Tcl_WideAsLong(wResult);
+}
+
+static int
+tellOld(
+    Tcl_Channel chan)		/* The channel to return pos for. */
+{
+    Tcl_WideInt wResult = Tcl_Tell(chan);
+
+    return (int) Tcl_WideAsLong(wResult);
+}
+#endif /* !TCL_NO_DEPRECATED */
+
+/*
+ * WARNING: The contents of this file is automatically generated by the
+ * tools/genStubs.tcl script. Any modifications to the function declarations
+ * below should be made in the generic/tcl.decls script.
+ */
+
+MODULE_SCOPE const TclStubs tclStubs;
+MODULE_SCOPE const TclTomMathStubs tclTomMathStubs;
+
+/* !BEGIN!: Do not edit below this line. */
+
+static const TclIntStubs tclIntStubs = {
+    TCL_STUB_MAGIC,
+    0,
+    0, /* 0 */
+    0, /* 1 */
+    0, /* 2 */
+    TclAllocateFreeObjects, /* 3 */
+    0, /* 4 */
+    TclCleanupChildren, /* 5 */
+    TclCleanupCommand, /* 6 */
+    TclCopyAndCollapse, /* 7 */
+    TclCopyChannelOld, /* 8 */
+    TclCreatePipeline, /* 9 */
+    TclCreateProc, /* 10 */
+    TclDeleteCompiledLocalVars, /* 11 */
+    TclDeleteVars, /* 12 */
+    0, /* 13 */
+    TclDumpMemoryInfo, /* 14 */
+    0, /* 15 */
+    TclExprFloatError, /* 16 */
+    0, /* 17 */
+    0, /* 18 */
+    0, /* 19 */
+    0, /* 20 */
+    0, /* 21 */
+    TclFindElement, /* 22 */
+    TclFindProc, /* 23 */
+    TclFormatInt, /* 24 */
+    TclFreePackageInfo, /* 25 */
+    0, /* 26 */
+    0, /* 27 */
+    TclpGetDefaultStdChannel, /* 28 */
+    0, /* 29 */
+    0, /* 30 */
+    TclGetExtension, /* 31 */
+    TclGetFrame, /* 32 */
+    0, /* 33 */
+    TclGetIntForIndex, /* 34 */
+    0, /* 35 */
+    0, /* 36 */
+    TclGetLoadedPackages, /* 37 */
+    TclGetNamespaceForQualName, /* 38 */
+    TclGetObjInterpProc, /* 39 */
+    TclGetOpenMode, /* 40 */
+    TclGetOriginalCommand, /* 41 */
+    TclpGetUserHome, /* 42 */
+    0, /* 43 */
+    TclGuessPackageName, /* 44 */
+    TclHideUnsafeCommands, /* 45 */
+    TclInExit, /* 46 */
+    0, /* 47 */
+    0, /* 48 */
+    0, /* 49 */
+    TclInitCompiledLocals, /* 50 */
+    TclInterpInit, /* 51 */
+    0, /* 52 */
+    TclInvokeObjectCommand, /* 53 */
+    TclInvokeStringCommand, /* 54 */
+    TclIsProc, /* 55 */
+    0, /* 56 */
+    0, /* 57 */
+    TclLookupVar, /* 58 */
+    0, /* 59 */
+    TclNeedSpace, /* 60 */
+    TclNewProcBodyObj, /* 61 */
+    TclObjCommandComplete, /* 62 */
+    TclObjInterpProc, /* 63 */
+    TclObjInvoke, /* 64 */
+    0, /* 65 */
+    0, /* 66 */
+    0, /* 67 */
+    0, /* 68 */
+    TclpAlloc, /* 69 */
+    0, /* 70 */
+    0, /* 71 */
+    0, /* 72 */
+    0, /* 73 */
+    TclpFree, /* 74 */
+    TclpGetClicks, /* 75 */
+    TclpGetSeconds, /* 76 */
+    TclpGetTime, /* 77 */
+    0, /* 78 */
+    0, /* 79 */
+    0, /* 80 */
+    TclpRealloc, /* 81 */
+    0, /* 82 */
+    0, /* 83 */
+    0, /* 84 */
+    0, /* 85 */
+    0, /* 86 */
+    0, /* 87 */
+    TclPrecTraceProc, /* 88 */
+    TclPreventAliasLoop, /* 89 */
+    0, /* 90 */
+    TclProcCleanupProc, /* 91 */
+    TclProcCompileProc, /* 92 */
+    TclProcDeleteProc, /* 93 */
+    0, /* 94 */
+    0, /* 95 */
+    TclRenameCommand, /* 96 */
+    TclResetShadowedCmdRefs, /* 97 */
+    TclServiceIdle, /* 98 */
+    0, /* 99 */
+    0, /* 100 */
+    TclSetPreInitScript, /* 101 */
+    TclSetupEnv, /* 102 */
+    TclSockGetPort, /* 103 */
+    TclSockMinimumBuffersOld, /* 104 */
+    0, /* 105 */
+    0, /* 106 */
+    0, /* 107 */
+    TclTeardownNamespace, /* 108 */
+    TclUpdateReturnInfo, /* 109 */
+    TclSockMinimumBuffers, /* 110 */
+    Tcl_AddInterpResolvers, /* 111 */
+    Tcl_AppendExportList, /* 112 */
+    Tcl_CreateNamespace, /* 113 */
+    Tcl_DeleteNamespace, /* 114 */
+    Tcl_Export, /* 115 */
+    Tcl_FindCommand, /* 116 */
+    Tcl_FindNamespace, /* 117 */
+    Tcl_GetInterpResolvers, /* 118 */
+    Tcl_GetNamespaceResolvers, /* 119 */
+    Tcl_FindNamespaceVar, /* 120 */
+    Tcl_ForgetImport, /* 121 */
+    Tcl_GetCommandFromObj, /* 122 */
+    Tcl_GetCommandFullName, /* 123 */
+    Tcl_GetCurrentNamespace, /* 124 */
+    Tcl_GetGlobalNamespace, /* 125 */
+    Tcl_GetVariableFullName, /* 126 */
+    Tcl_Import, /* 127 */
+    Tcl_PopCallFrame, /* 128 */
+    Tcl_PushCallFrame, /* 129 */
+    Tcl_RemoveInterpResolvers, /* 130 */
+    Tcl_SetNamespaceResolvers, /* 131 */
+    TclpHasSockets, /* 132 */
+    TclpGetDate, /* 133 */
+    0, /* 134 */
+    0, /* 135 */
+    0, /* 136 */
+    0, /* 137 */
+    TclGetEnv, /* 138 */
+    0, /* 139 */
+    0, /* 140 */
+    TclpGetCwd, /* 141 */
+    TclSetByteCodeFromAny, /* 142 */
+    TclAddLiteralObj, /* 143 */
+    TclHideLiteral, /* 144 */
+    TclGetAuxDataType, /* 145 */
+    TclHandleCreate, /* 146 */
+    TclHandleFree, /* 147 */
+    TclHandlePreserve, /* 148 */
+    TclHandleRelease, /* 149 */
+    TclRegAbout, /* 150 */
+    TclRegExpRangeUniChar, /* 151 */
+    TclSetLibraryPath, /* 152 */
+    TclGetLibraryPath, /* 153 */
+    0, /* 154 */
+    0, /* 155 */
+    TclRegError, /* 156 */
+    TclVarTraceExists, /* 157 */
+    TclSetStartupScriptFileName, /* 158 */
+    TclGetStartupScriptFileName, /* 159 */
+    0, /* 160 */
+    TclChannelTransform, /* 161 */
+    TclChannelEventScriptInvoker, /* 162 */
+    TclGetInstructionTable, /* 163 */
+    TclExpandCodeArray, /* 164 */
+    TclpSetInitialEncodings, /* 165 */
+    TclListObjSetElement, /* 166 */
+    TclSetStartupScriptPath, /* 167 */
+    TclGetStartupScriptPath, /* 168 */
+    TclpUtfNcmp2, /* 169 */
+    TclCheckInterpTraces, /* 170 */
+    TclCheckExecutionTraces, /* 171 */
+    TclInThreadExit, /* 172 */
+    TclUniCharMatch, /* 173 */
+    0, /* 174 */
+    TclCallVarTraces, /* 175 */
+    TclCleanupVar, /* 176 */
+    TclVarErrMsg, /* 177 */
+    Tcl_SetStartupScript, /* 178 */
+    Tcl_GetStartupScript, /* 179 */
+    0, /* 180 */
+    0, /* 181 */
+    TclpLocaltime, /* 182 */
+    TclpGmtime, /* 183 */
+    0, /* 184 */
+    0, /* 185 */
+    0, /* 186 */
+    0, /* 187 */
+    0, /* 188 */
+    0, /* 189 */
+    0, /* 190 */
+    0, /* 191 */
+    0, /* 192 */
+    0, /* 193 */
+    0, /* 194 */
+    0, /* 195 */
+    0, /* 196 */
+    0, /* 197 */
+    TclObjGetFrame, /* 198 */
+    0, /* 199 */
+    TclpObjRemoveDirectory, /* 200 */
+    TclpObjCopyDirectory, /* 201 */
+    TclpObjCreateDirectory, /* 202 */
+    TclpObjDeleteFile, /* 203 */
+    TclpObjCopyFile, /* 204 */
+    TclpObjRenameFile, /* 205 */
+    TclpObjStat, /* 206 */
+    TclpObjAccess, /* 207 */
+    TclpOpenFileChannel, /* 208 */
+    0, /* 209 */
+    0, /* 210 */
+    0, /* 211 */
+    TclpFindExecutable, /* 212 */
+    TclGetObjNameOfExecutable, /* 213 */
+    TclSetObjNameOfExecutable, /* 214 */
+    TclStackAlloc, /* 215 */
+    TclStackFree, /* 216 */
+    TclPushStackFrame, /* 217 */
+    TclPopStackFrame, /* 218 */
+    0, /* 219 */
+    0, /* 220 */
+    0, /* 221 */
+    0, /* 222 */
+    0, /* 223 */
+    TclGetPlatform, /* 224 */
+    TclTraceDictPath, /* 225 */
+    TclObjBeingDeleted, /* 226 */
+    TclSetNsPath, /* 227 */
+    0, /* 228 */
+    TclPtrMakeUpvar, /* 229 */
+    TclObjLookupVar, /* 230 */
+    TclGetNamespaceFromObj, /* 231 */
+    TclEvalObjEx, /* 232 */
+    TclGetSrcInfoForPc, /* 233 */
+    TclVarHashCreateVar, /* 234 */
+    TclInitVarHashTable, /* 235 */
+    TclBackgroundException, /* 236 */
+    TclResetCancellation, /* 237 */
+    TclNRInterpProc, /* 238 */
+    TclNRInterpProcCore, /* 239 */
+    TclNRRunCallbacks, /* 240 */
+    TclNREvalObjEx, /* 241 */
+    TclNREvalObjv, /* 242 */
+    TclDbDumpActiveObjects, /* 243 */
+    TclGetNamespaceChildTable, /* 244 */
+    TclGetNamespaceCommandTable, /* 245 */
+    TclInitRewriteEnsemble, /* 246 */
+    TclResetRewriteEnsemble, /* 247 */
+    TclCopyChannel, /* 248 */
+    TclDoubleDigits, /* 249 */
+    TclSetSlaveCancelFlags, /* 250 */
+    TclRegisterLiteral, /* 251 */
+    TclPtrGetVar, /* 252 */
+    TclPtrSetVar, /* 253 */
+    TclPtrIncrObjVar, /* 254 */
+    TclPtrObjMakeUpvar, /* 255 */
+    TclPtrUnsetVar, /* 256 */
+};
+
+static const TclIntPlatStubs tclIntPlatStubs = {
+    TCL_STUB_MAGIC,
+    0,
+#if !defined(_WIN32) && !defined(__CYGWIN__) && !defined(MAC_OSX_TCL) /* UNIX */
+    TclGetAndDetachPids, /* 0 */
+    TclpCloseFile, /* 1 */
+    TclpCreateCommandChannel, /* 2 */
+    TclpCreatePipe, /* 3 */
+    TclpCreateProcess, /* 4 */
+    0, /* 5 */
+    TclpMakeFile, /* 6 */
+    TclpOpenFile, /* 7 */
+    TclUnixWaitForFile, /* 8 */
+    TclpCreateTempFile, /* 9 */
+    TclpReaddir, /* 10 */
+    TclpLocaltime_unix, /* 11 */
+    TclpGmtime_unix, /* 12 */
+    TclpInetNtoa, /* 13 */
+    TclUnixCopyFile, /* 14 */
+    0, /* 15 */
+    0, /* 16 */
+    0, /* 17 */
+    0, /* 18 */
+    0, /* 19 */
+    0, /* 20 */
+    0, /* 21 */
+    0, /* 22 */
+    0, /* 23 */
+    0, /* 24 */
+    0, /* 25 */
+    0, /* 26 */
+    0, /* 27 */
+    0, /* 28 */
+    TclWinCPUID, /* 29 */
+    TclUnixOpenTemporaryFile, /* 30 */
+#endif /* UNIX */
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+    TclWinConvertError, /* 0 */
+    TclWinConvertWSAError, /* 1 */
+    TclWinGetServByName, /* 2 */
+    TclWinGetSockOpt, /* 3 */
+    TclWinGetTclInstance, /* 4 */
+    TclUnixWaitForFile, /* 5 */
+    TclWinNToHS, /* 6 */
+    TclWinSetSockOpt, /* 7 */
+    TclpGetPid, /* 8 */
+    TclWinGetPlatformId, /* 9 */
+    TclpReaddir, /* 10 */
+    TclGetAndDetachPids, /* 11 */
+    TclpCloseFile, /* 12 */
+    TclpCreateCommandChannel, /* 13 */
+    TclpCreatePipe, /* 14 */
+    TclpCreateProcess, /* 15 */
+    TclpIsAtty, /* 16 */
+    TclUnixCopyFile, /* 17 */
+    TclpMakeFile, /* 18 */
+    TclpOpenFile, /* 19 */
+    TclWinAddProcess, /* 20 */
+    TclpInetNtoa, /* 21 */
+    TclpCreateTempFile, /* 22 */
+    0, /* 23 */
+    TclWinNoBackslash, /* 24 */
+    0, /* 25 */
+    TclWinSetInterfaces, /* 26 */
+    TclWinFlushDirtyChannels, /* 27 */
+    TclWinResetInterfaces, /* 28 */
+    TclWinCPUID, /* 29 */
+    TclUnixOpenTemporaryFile, /* 30 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    TclGetAndDetachPids, /* 0 */
+    TclpCloseFile, /* 1 */
+    TclpCreateCommandChannel, /* 2 */
+    TclpCreatePipe, /* 3 */
+    TclpCreateProcess, /* 4 */
+    0, /* 5 */
+    TclpMakeFile, /* 6 */
+    TclpOpenFile, /* 7 */
+    TclUnixWaitForFile, /* 8 */
+    TclpCreateTempFile, /* 9 */
+    TclpReaddir, /* 10 */
+    TclpLocaltime_unix, /* 11 */
+    TclpGmtime_unix, /* 12 */
+    TclpInetNtoa, /* 13 */
+    TclUnixCopyFile, /* 14 */
+    TclMacOSXGetFileAttribute, /* 15 */
+    TclMacOSXSetFileAttribute, /* 16 */
+    TclMacOSXCopyFileAttributes, /* 17 */
+    TclMacOSXMatchType, /* 18 */
+    TclMacOSXNotifierAddRunLoopMode, /* 19 */
+    0, /* 20 */
+    0, /* 21 */
+    0, /* 22 */
+    0, /* 23 */
+    0, /* 24 */
+    0, /* 25 */
+    0, /* 26 */
+    0, /* 27 */
+    0, /* 28 */
+    TclWinCPUID, /* 29 */
+    TclUnixOpenTemporaryFile, /* 30 */
+#endif /* MACOSX */
+};
+
+static const TclPlatStubs tclPlatStubs = {
+    TCL_STUB_MAGIC,
+    0,
+#if defined(_WIN32) || defined(__CYGWIN__) /* WIN */
+    Tcl_WinUtfToTChar, /* 0 */
+    Tcl_WinTCharToUtf, /* 1 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    Tcl_MacOSXOpenBundleResources, /* 0 */
+    Tcl_MacOSXOpenVersionedBundleResources, /* 1 */
+#endif /* MACOSX */
+};
+
+const TclTomMathStubs tclTomMathStubs = {
+    TCL_STUB_MAGIC,
+    0,
+    TclBN_epoch, /* 0 */
+    TclBN_revision, /* 1 */
+    TclBN_mp_add, /* 2 */
+    TclBN_mp_add_d, /* 3 */
+    TclBN_mp_and, /* 4 */
+    TclBN_mp_clamp, /* 5 */
+    TclBN_mp_clear, /* 6 */
+    TclBN_mp_clear_multi, /* 7 */
+    TclBN_mp_cmp, /* 8 */
+    TclBN_mp_cmp_d, /* 9 */
+    TclBN_mp_cmp_mag, /* 10 */
+    TclBN_mp_copy, /* 11 */
+    TclBN_mp_count_bits, /* 12 */
+    TclBN_mp_div, /* 13 */
+    TclBN_mp_div_d, /* 14 */
+    TclBN_mp_div_2, /* 15 */
+    TclBN_mp_div_2d, /* 16 */
+    TclBN_mp_div_3, /* 17 */
+    TclBN_mp_exch, /* 18 */
+    TclBN_mp_expt_d, /* 19 */
+    TclBN_mp_grow, /* 20 */
+    TclBN_mp_init, /* 21 */
+    TclBN_mp_init_copy, /* 22 */
+    TclBN_mp_init_multi, /* 23 */
+    TclBN_mp_init_set, /* 24 */
+    TclBN_mp_init_size, /* 25 */
+    TclBN_mp_lshd, /* 26 */
+    TclBN_mp_mod, /* 27 */
+    TclBN_mp_mod_2d, /* 28 */
+    TclBN_mp_mul, /* 29 */
+    TclBN_mp_mul_d, /* 30 */
+    TclBN_mp_mul_2, /* 31 */
+    TclBN_mp_mul_2d, /* 32 */
+    TclBN_mp_neg, /* 33 */
+    TclBN_mp_or, /* 34 */
+    TclBN_mp_radix_size, /* 35 */
+    TclBN_mp_read_radix, /* 36 */
+    TclBN_mp_rshd, /* 37 */
+    TclBN_mp_shrink, /* 38 */
+    TclBN_mp_set, /* 39 */
+    TclBN_mp_sqr, /* 40 */
+    TclBN_mp_sqrt, /* 41 */
+    TclBN_mp_sub, /* 42 */
+    TclBN_mp_sub_d, /* 43 */
+    TclBN_mp_to_unsigned_bin, /* 44 */
+    TclBN_mp_to_unsigned_bin_n, /* 45 */
+    TclBN_mp_toradix_n, /* 46 */
+    TclBN_mp_unsigned_bin_size, /* 47 */
+    TclBN_mp_xor, /* 48 */
+    TclBN_mp_zero, /* 49 */
+    TclBN_reverse, /* 50 */
+    TclBN_fast_s_mp_mul_digs, /* 51 */
+    TclBN_fast_s_mp_sqr, /* 52 */
+    TclBN_mp_karatsuba_mul, /* 53 */
+    TclBN_mp_karatsuba_sqr, /* 54 */
+    TclBN_mp_toom_mul, /* 55 */
+    TclBN_mp_toom_sqr, /* 56 */
+    TclBN_s_mp_add, /* 57 */
+    TclBN_s_mp_mul_digs, /* 58 */
+    TclBN_s_mp_sqr, /* 59 */
+    TclBN_s_mp_sub, /* 60 */
+    TclBN_mp_init_set_int, /* 61 */
+    TclBN_mp_set_int, /* 62 */
+    TclBN_mp_cnt_lsb, /* 63 */
+    TclBNInitBignumFromLong, /* 64 */
+    TclBNInitBignumFromWideInt, /* 65 */
+    TclBNInitBignumFromWideUInt, /* 66 */
+    TclBN_mp_expt_d_ex, /* 67 */
+};
+
+static const TclStubHooks tclStubHooks = {
+    &tclPlatStubs,
+    &tclIntStubs,
+    &tclIntPlatStubs
+};
+
+const TclStubs tclStubs = {
+    TCL_STUB_MAGIC,
+    &tclStubHooks,
+    Tcl_PkgProvideEx, /* 0 */
+    Tcl_PkgRequireEx, /* 1 */
+    Tcl_Panic, /* 2 */
+    Tcl_Alloc, /* 3 */
+    Tcl_Free, /* 4 */
+    Tcl_Realloc, /* 5 */
+    Tcl_DbCkalloc, /* 6 */
+    Tcl_DbCkfree, /* 7 */
+    Tcl_DbCkrealloc, /* 8 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    Tcl_CreateFileHandler, /* 9 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    0, /* 9 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    Tcl_CreateFileHandler, /* 9 */
+#endif /* MACOSX */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    Tcl_DeleteFileHandler, /* 10 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    0, /* 10 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    Tcl_DeleteFileHandler, /* 10 */
+#endif /* MACOSX */
+    Tcl_SetTimer, /* 11 */
+    Tcl_Sleep, /* 12 */
+    Tcl_WaitForEvent, /* 13 */
+    Tcl_AppendAllObjTypes, /* 14 */
+    Tcl_AppendStringsToObj, /* 15 */
+    Tcl_AppendToObj, /* 16 */
+    Tcl_ConcatObj, /* 17 */
+    Tcl_ConvertToType, /* 18 */
+    Tcl_DbDecrRefCount, /* 19 */
+    Tcl_DbIncrRefCount, /* 20 */
+    Tcl_DbIsShared, /* 21 */
+    Tcl_DbNewBooleanObj, /* 22 */
+    Tcl_DbNewByteArrayObj, /* 23 */
+    Tcl_DbNewDoubleObj, /* 24 */
+    Tcl_DbNewListObj, /* 25 */
+    Tcl_DbNewLongObj, /* 26 */
+    Tcl_DbNewObj, /* 27 */
+    Tcl_DbNewStringObj, /* 28 */
+    Tcl_DuplicateObj, /* 29 */
+    TclFreeObj, /* 30 */
+    Tcl_GetBoolean, /* 31 */
+    Tcl_GetBooleanFromObj, /* 32 */
+    Tcl_GetByteArrayFromObj, /* 33 */
+    Tcl_GetDouble, /* 34 */
+    Tcl_GetDoubleFromObj, /* 35 */
+    Tcl_GetIndexFromObj, /* 36 */
+    Tcl_GetInt, /* 37 */
+    Tcl_GetIntFromObj, /* 38 */
+    Tcl_GetLongFromObj, /* 39 */
+    Tcl_GetObjType, /* 40 */
+    Tcl_GetStringFromObj, /* 41 */
+    Tcl_InvalidateStringRep, /* 42 */
+    Tcl_ListObjAppendList, /* 43 */
+    Tcl_ListObjAppendElement, /* 44 */
+    Tcl_ListObjGetElements, /* 45 */
+    Tcl_ListObjIndex, /* 46 */
+    Tcl_ListObjLength, /* 47 */
+    Tcl_ListObjReplace, /* 48 */
+    Tcl_NewBooleanObj, /* 49 */
+    Tcl_NewByteArrayObj, /* 50 */
+    Tcl_NewDoubleObj, /* 51 */
+    Tcl_NewIntObj, /* 52 */
+    Tcl_NewListObj, /* 53 */
+    Tcl_NewLongObj, /* 54 */
+    Tcl_NewObj, /* 55 */
+    Tcl_NewStringObj, /* 56 */
+    Tcl_SetBooleanObj, /* 57 */
+    Tcl_SetByteArrayLength, /* 58 */
+    Tcl_SetByteArrayObj, /* 59 */
+    Tcl_SetDoubleObj, /* 60 */
+    Tcl_SetIntObj, /* 61 */
+    Tcl_SetListObj, /* 62 */
+    Tcl_SetLongObj, /* 63 */
+    Tcl_SetObjLength, /* 64 */
+    Tcl_SetStringObj, /* 65 */
+    Tcl_AddErrorInfo, /* 66 */
+    Tcl_AddObjErrorInfo, /* 67 */
+    Tcl_AllowExceptions, /* 68 */
+    Tcl_AppendElement, /* 69 */
+    Tcl_AppendResult, /* 70 */
+    Tcl_AsyncCreate, /* 71 */
+    Tcl_AsyncDelete, /* 72 */
+    Tcl_AsyncInvoke, /* 73 */
+    Tcl_AsyncMark, /* 74 */
+    Tcl_AsyncReady, /* 75 */
+    Tcl_BackgroundError, /* 76 */
+    Tcl_Backslash, /* 77 */
+    Tcl_BadChannelOption, /* 78 */
+    Tcl_CallWhenDeleted, /* 79 */
+    Tcl_CancelIdleCall, /* 80 */
+    Tcl_Close, /* 81 */
+    Tcl_CommandComplete, /* 82 */
+    Tcl_Concat, /* 83 */
+    Tcl_ConvertElement, /* 84 */
+    Tcl_ConvertCountedElement, /* 85 */
+    Tcl_CreateAlias, /* 86 */
+    Tcl_CreateAliasObj, /* 87 */
+    Tcl_CreateChannel, /* 88 */
+    Tcl_CreateChannelHandler, /* 89 */
+    Tcl_CreateCloseHandler, /* 90 */
+    Tcl_CreateCommand, /* 91 */
+    Tcl_CreateEventSource, /* 92 */
+    Tcl_CreateExitHandler, /* 93 */
+    Tcl_CreateInterp, /* 94 */
+    Tcl_CreateMathFunc, /* 95 */
+    Tcl_CreateObjCommand, /* 96 */
+    Tcl_CreateSlave, /* 97 */
+    Tcl_CreateTimerHandler, /* 98 */
+    Tcl_CreateTrace, /* 99 */
+    Tcl_DeleteAssocData, /* 100 */
+    Tcl_DeleteChannelHandler, /* 101 */
+    Tcl_DeleteCloseHandler, /* 102 */
+    Tcl_DeleteCommand, /* 103 */
+    Tcl_DeleteCommandFromToken, /* 104 */
+    Tcl_DeleteEvents, /* 105 */
+    Tcl_DeleteEventSource, /* 106 */
+    Tcl_DeleteExitHandler, /* 107 */
+    Tcl_DeleteHashEntry, /* 108 */
+    Tcl_DeleteHashTable, /* 109 */
+    Tcl_DeleteInterp, /* 110 */
+    Tcl_DetachPids, /* 111 */
+    Tcl_DeleteTimerHandler, /* 112 */
+    Tcl_DeleteTrace, /* 113 */
+    Tcl_DontCallWhenDeleted, /* 114 */
+    Tcl_DoOneEvent, /* 115 */
+    Tcl_DoWhenIdle, /* 116 */
+    Tcl_DStringAppend, /* 117 */
+    Tcl_DStringAppendElement, /* 118 */
+    Tcl_DStringEndSublist, /* 119 */
+    Tcl_DStringFree, /* 120 */
+    Tcl_DStringGetResult, /* 121 */
+    Tcl_DStringInit, /* 122 */
+    Tcl_DStringResult, /* 123 */
+    Tcl_DStringSetLength, /* 124 */
+    Tcl_DStringStartSublist, /* 125 */
+    Tcl_Eof, /* 126 */
+    Tcl_ErrnoId, /* 127 */
+    Tcl_ErrnoMsg, /* 128 */
+    Tcl_Eval, /* 129 */
+    Tcl_EvalFile, /* 130 */
+    Tcl_EvalObj, /* 131 */
+    Tcl_EventuallyFree, /* 132 */
+    Tcl_Exit, /* 133 */
+    Tcl_ExposeCommand, /* 134 */
+    Tcl_ExprBoolean, /* 135 */
+    Tcl_ExprBooleanObj, /* 136 */
+    Tcl_ExprDouble, /* 137 */
+    Tcl_ExprDoubleObj, /* 138 */
+    Tcl_ExprLong, /* 139 */
+    Tcl_ExprLongObj, /* 140 */
+    Tcl_ExprObj, /* 141 */
+    Tcl_ExprString, /* 142 */
+    Tcl_Finalize, /* 143 */
+    Tcl_FindExecutable, /* 144 */
+    Tcl_FirstHashEntry, /* 145 */
+    Tcl_Flush, /* 146 */
+    Tcl_FreeResult, /* 147 */
+    Tcl_GetAlias, /* 148 */
+    Tcl_GetAliasObj, /* 149 */
+    Tcl_GetAssocData, /* 150 */
+    Tcl_GetChannel, /* 151 */
+    Tcl_GetChannelBufferSize, /* 152 */
+    Tcl_GetChannelHandle, /* 153 */
+    Tcl_GetChannelInstanceData, /* 154 */
+    Tcl_GetChannelMode, /* 155 */
+    Tcl_GetChannelName, /* 156 */
+    Tcl_GetChannelOption, /* 157 */
+    Tcl_GetChannelType, /* 158 */
+    Tcl_GetCommandInfo, /* 159 */
+    Tcl_GetCommandName, /* 160 */
+    Tcl_GetErrno, /* 161 */
+    Tcl_GetHostName, /* 162 */
+    Tcl_GetInterpPath, /* 163 */
+    Tcl_GetMaster, /* 164 */
+    Tcl_GetNameOfExecutable, /* 165 */
+    Tcl_GetObjResult, /* 166 */
+#if !defined(_WIN32) && !defined(MAC_OSX_TCL) /* UNIX */
+    Tcl_GetOpenFile, /* 167 */
+#endif /* UNIX */
+#if defined(_WIN32) /* WIN */
+    0, /* 167 */
+#endif /* WIN */
+#ifdef MAC_OSX_TCL /* MACOSX */
+    Tcl_GetOpenFile, /* 167 */
+#endif /* MACOSX */
+    Tcl_GetPathType, /* 168 */
+    Tcl_Gets, /* 169 */
+    Tcl_GetsObj, /* 170 */
+    Tcl_GetServiceMode, /* 171 */
+    Tcl_GetSlave, /* 172 */
+    Tcl_GetStdChannel, /* 173 */
+    Tcl_GetStringResult, /* 174 */
+    Tcl_GetVar, /* 175 */
+    Tcl_GetVar2, /* 176 */
+    Tcl_GlobalEval, /* 177 */
+    Tcl_GlobalEvalObj, /* 178 */
+    Tcl_HideCommand, /* 179 */
+    Tcl_Init, /* 180 */
+    Tcl_InitHashTable, /* 181 */
+    Tcl_InputBlocked, /* 182 */
+    Tcl_InputBuffered, /* 183 */
+    Tcl_InterpDeleted, /* 184 */
+    Tcl_IsSafe, /* 185 */
+    Tcl_JoinPath, /* 186 */
+    Tcl_LinkVar, /* 187 */
+    0, /* 188 */
+    Tcl_MakeFileChannel, /* 189 */
+    Tcl_MakeSafe, /* 190 */
+    Tcl_MakeTcpClientChannel, /* 191 */
+    Tcl_Merge, /* 192 */
+    Tcl_NextHashEntry, /* 193 */
+    Tcl_NotifyChannel, /* 194 */
+    Tcl_ObjGetVar2, /* 195 */
+    Tcl_ObjSetVar2, /* 196 */
+    Tcl_OpenCommandChannel, /* 197 */
+    Tcl_OpenFileChannel, /* 198 */
+    Tcl_OpenTcpClient, /* 199 */
+    Tcl_OpenTcpServer, /* 200 */
+    Tcl_Preserve, /* 201 */
+    Tcl_PrintDouble, /* 202 */
+    Tcl_PutEnv, /* 203 */
+    Tcl_PosixError, /* 204 */
+    Tcl_QueueEvent, /* 205 */
+    Tcl_Read, /* 206 */
+    Tcl_ReapDetachedProcs, /* 207 */
+    Tcl_RecordAndEval, /* 208 */
+    Tcl_RecordAndEvalObj, /* 209 */
+    Tcl_RegisterChannel, /* 210 */
+    Tcl_RegisterObjType, /* 211 */
+    Tcl_RegExpCompile, /* 212 */
+    Tcl_RegExpExec, /* 213 */
+    Tcl_RegExpMatch, /* 214 */
+    Tcl_RegExpRange, /* 215 */
+    Tcl_Release, /* 216 */
+    Tcl_ResetResult, /* 217 */
+    Tcl_ScanElement, /* 218 */
+    Tcl_ScanCountedElement, /* 219 */
+    Tcl_SeekOld, /* 220 */
+    Tcl_ServiceAll, /* 221 */
+    Tcl_ServiceEvent, /* 222 */
+    Tcl_SetAssocData, /* 223 */
+    Tcl_SetChannelBufferSize, /* 224 */
+    Tcl_SetChannelOption, /* 225 */
+    Tcl_SetCommandInfo, /* 226 */
+    Tcl_SetErrno, /* 227 */
+    Tcl_SetErrorCode, /* 228 */
+    Tcl_SetMaxBlockTime, /* 229 */
+    Tcl_SetPanicProc, /* 230 */
+    Tcl_SetRecursionLimit, /* 231 */
+    Tcl_SetResult, /* 232 */
+    Tcl_SetServiceMode, /* 233 */
+    Tcl_SetObjErrorCode, /* 234 */
+    Tcl_SetObjResult, /* 235 */
+    Tcl_SetStdChannel, /* 236 */
+    Tcl_SetVar, /* 237 */
+    Tcl_SetVar2, /* 238 */
+    Tcl_SignalId, /* 239 */
+    Tcl_SignalMsg, /* 240 */
+    Tcl_SourceRCFile, /* 241 */
+    Tcl_SplitList, /* 242 */
+    Tcl_SplitPath, /* 243 */
+    Tcl_StaticPackage, /* 244 */
+    Tcl_StringMatch, /* 245 */
+    Tcl_TellOld, /* 246 */
+    Tcl_TraceVar, /* 247 */
+    Tcl_TraceVar2, /* 248 */
+    Tcl_TranslateFileName, /* 249 */
+    Tcl_Ungets, /* 250 */
+    Tcl_UnlinkVar, /* 251 */
+    Tcl_UnregisterChannel, /* 252 */
+    Tcl_UnsetVar, /* 253 */
+    Tcl_UnsetVar2, /* 254 */
+    Tcl_UntraceVar, /* 255 */
+    Tcl_UntraceVar2, /* 256 */
+    Tcl_UpdateLinkedVar, /* 257 */
+    Tcl_UpVar, /* 258 */
+    Tcl_UpVar2, /* 259 */
+    Tcl_VarEval, /* 260 */
+    Tcl_VarTraceInfo, /* 261 */
+    Tcl_VarTraceInfo2, /* 262 */
+    Tcl_Write, /* 263 */
+    Tcl_WrongNumArgs, /* 264 */
+    Tcl_DumpActiveMemory, /* 265 */
+    Tcl_ValidateAllMemory, /* 266 */
+    Tcl_AppendResultVA, /* 267 */
+    Tcl_AppendStringsToObjVA, /* 268 */
+    Tcl_HashStats, /* 269 */
+    Tcl_ParseVar, /* 270 */
+    Tcl_PkgPresent, /* 271 */
+    Tcl_PkgPresentEx, /* 272 */
+    Tcl_PkgProvide, /* 273 */
+    Tcl_PkgRequire, /* 274 */
+    Tcl_SetErrorCodeVA, /* 275 */
+    Tcl_VarEvalVA, /* 276 */
+    Tcl_WaitPid, /* 277 */
+    Tcl_PanicVA, /* 278 */
+    Tcl_GetVersion, /* 279 */
+    Tcl_InitMemory, /* 280 */
+    Tcl_StackChannel, /* 281 */
+    Tcl_UnstackChannel, /* 282 */
+    Tcl_GetStackedChannel, /* 283 */
+    Tcl_SetMainLoop, /* 284 */
+    0, /* 285 */
+    Tcl_AppendObjToObj, /* 286 */
+    Tcl_CreateEncoding, /* 287 */
+    Tcl_CreateThreadExitHandler, /* 288 */
+    Tcl_DeleteThreadExitHandler, /* 289 */
+    Tcl_DiscardResult, /* 290 */
+    Tcl_EvalEx, /* 291 */
+    Tcl_EvalObjv, /* 292 */
+    Tcl_EvalObjEx, /* 293 */
+    Tcl_ExitThread, /* 294 */
+    Tcl_ExternalToUtf, /* 295 */
+    Tcl_ExternalToUtfDString, /* 296 */
+    Tcl_FinalizeThread, /* 297 */
+    Tcl_FinalizeNotifier, /* 298 */
+    Tcl_FreeEncoding, /* 299 */
+    Tcl_GetCurrentThread, /* 300 */
+    Tcl_GetEncoding, /* 301 */
+    Tcl_GetEncodingName, /* 302 */
+    Tcl_GetEncodingNames, /* 303 */
+    Tcl_GetIndexFromObjStruct, /* 304 */
+    Tcl_GetThreadData, /* 305 */
+    Tcl_GetVar2Ex, /* 306 */
+    Tcl_InitNotifier, /* 307 */
+    Tcl_MutexLock, /* 308 */
+    Tcl_MutexUnlock, /* 309 */
+    Tcl_ConditionNotify, /* 310 */
+    Tcl_ConditionWait, /* 311 */
+    Tcl_NumUtfChars, /* 312 */
+    Tcl_ReadChars, /* 313 */
+    Tcl_RestoreResult, /* 314 */
+    Tcl_SaveResult, /* 315 */
+    Tcl_SetSystemEncoding, /* 316 */
+    Tcl_SetVar2Ex, /* 317 */
+    Tcl_ThreadAlert, /* 318 */
+    Tcl_ThreadQueueEvent, /* 319 */
+    Tcl_UniCharAtIndex, /* 320 */
+    Tcl_UniCharToLower, /* 321 */
+    Tcl_UniCharToTitle, /* 322 */
+    Tcl_UniCharToUpper, /* 323 */
+    Tcl_UniCharToUtf, /* 324 */
+    Tcl_UtfAtIndex, /* 325 */
+    Tcl_UtfCharComplete, /* 326 */
+    Tcl_UtfBackslash, /* 327 */
+    Tcl_UtfFindFirst, /* 328 */
+    Tcl_UtfFindLast, /* 329 */
+    Tcl_UtfNext, /* 330 */
+    Tcl_UtfPrev, /* 331 */
+    Tcl_UtfToExternal, /* 332 */
+    Tcl_UtfToExternalDString, /* 333 */
+    Tcl_UtfToLower, /* 334 */
+    Tcl_UtfToTitle, /* 335 */
+    Tcl_UtfToUniChar, /* 336 */
+    Tcl_UtfToUpper, /* 337 */
+    Tcl_WriteChars, /* 338 */
+    Tcl_WriteObj, /* 339 */
+    Tcl_GetString, /* 340 */
+    Tcl_GetDefaultEncodingDir, /* 341 */
+    Tcl_SetDefaultEncodingDir, /* 342 */
+    Tcl_AlertNotifier, /* 343 */
+    Tcl_ServiceModeHook, /* 344 */
+    Tcl_UniCharIsAlnum, /* 345 */
+    Tcl_UniCharIsAlpha, /* 346 */
+    Tcl_UniCharIsDigit, /* 347 */
+    Tcl_UniCharIsLower, /* 348 */
+    Tcl_UniCharIsSpace, /* 349 */
+    Tcl_UniCharIsUpper, /* 350 */
+    Tcl_UniCharIsWordChar, /* 351 */
+    Tcl_UniCharLen, /* 352 */
+    Tcl_UniCharNcmp, /* 353 */
+    Tcl_UniCharToUtfDString, /* 354 */
+    Tcl_UtfToUniCharDString, /* 355 */
+    Tcl_GetRegExpFromObj, /* 356 */
+    Tcl_EvalTokens, /* 357 */
+    Tcl_FreeParse, /* 358 */
+    Tcl_LogCommandInfo, /* 359 */
+    Tcl_ParseBraces, /* 360 */
+    Tcl_ParseCommand, /* 361 */
+    Tcl_ParseExpr, /* 362 */
+    Tcl_ParseQuotedString, /* 363 */
+    Tcl_ParseVarName, /* 364 */
+    Tcl_GetCwd, /* 365 */
+    Tcl_Chdir, /* 366 */
+    Tcl_Access, /* 367 */
+    Tcl_Stat, /* 368 */
+    Tcl_UtfNcmp, /* 369 */
+    Tcl_UtfNcasecmp, /* 370 */
+    Tcl_StringCaseMatch, /* 371 */
+    Tcl_UniCharIsControl, /* 372 */
+    Tcl_UniCharIsGraph, /* 373 */
+    Tcl_UniCharIsPrint, /* 374 */
+    Tcl_UniCharIsPunct, /* 375 */
+    Tcl_RegExpExecObj, /* 376 */
+    Tcl_RegExpGetInfo, /* 377 */
+    Tcl_NewUnicodeObj, /* 378 */
+    Tcl_SetUnicodeObj, /* 379 */
+    Tcl_GetCharLength, /* 380 */
+    Tcl_GetUniChar, /* 381 */
+    Tcl_GetUnicode, /* 382 */
+    Tcl_GetRange, /* 383 */
+    Tcl_AppendUnicodeToObj, /* 384 */
+    Tcl_RegExpMatchObj, /* 385 */
+    Tcl_SetNotifier, /* 386 */
+    Tcl_GetAllocMutex, /* 387 */
+    Tcl_GetChannelNames, /* 388 */
+    Tcl_GetChannelNamesEx, /* 389 */
+    Tcl_ProcObjCmd, /* 390 */
+    Tcl_ConditionFinalize, /* 391 */
+    Tcl_MutexFinalize, /* 392 */
+    Tcl_CreateThread, /* 393 */
+    Tcl_ReadRaw, /* 394 */
+    Tcl_WriteRaw, /* 395 */
+    Tcl_GetTopChannel, /* 396 */
+    Tcl_ChannelBuffered, /* 397 */
+    Tcl_ChannelName, /* 398 */
+    Tcl_ChannelVersion, /* 399 */
+    Tcl_ChannelBlockModeProc, /* 400 */
+    Tcl_ChannelCloseProc, /* 401 */
+    Tcl_ChannelClose2Proc, /* 402 */
+    Tcl_ChannelInputProc, /* 403 */
+    Tcl_ChannelOutputProc, /* 404 */
+    Tcl_ChannelSeekProc, /* 405 */
+    Tcl_ChannelSetOptionProc, /* 406 */
+    Tcl_ChannelGetOptionProc, /* 407 */
+    Tcl_ChannelWatchProc, /* 408 */
+    Tcl_ChannelGetHandleProc, /* 409 */
+    Tcl_ChannelFlushProc, /* 410 */
+    Tcl_ChannelHandlerProc, /* 411 */
+    Tcl_JoinThread, /* 412 */
+    Tcl_IsChannelShared, /* 413 */
+    Tcl_IsChannelRegistered, /* 414 */
+    Tcl_CutChannel, /* 415 */
+    Tcl_SpliceChannel, /* 416 */
+    Tcl_ClearChannelHandlers, /* 417 */
+    Tcl_IsChannelExisting, /* 418 */
+    Tcl_UniCharNcasecmp, /* 419 */
+    Tcl_UniCharCaseMatch, /* 420 */
+    Tcl_FindHashEntry, /* 421 */
+    Tcl_CreateHashEntry, /* 422 */
+    Tcl_InitCustomHashTable, /* 423 */
+    Tcl_InitObjHashTable, /* 424 */
+    Tcl_CommandTraceInfo, /* 425 */
+    Tcl_TraceCommand, /* 426 */
+    Tcl_UntraceCommand, /* 427 */
+    Tcl_AttemptAlloc, /* 428 */
+    Tcl_AttemptDbCkalloc, /* 429 */
+    Tcl_AttemptRealloc, /* 430 */
+    Tcl_AttemptDbCkrealloc, /* 431 */
+    Tcl_AttemptSetObjLength, /* 432 */
+    Tcl_GetChannelThread, /* 433 */
+    Tcl_GetUnicodeFromObj, /* 434 */
+    Tcl_GetMathFuncInfo, /* 435 */
+    Tcl_ListMathFuncs, /* 436 */
+    Tcl_SubstObj, /* 437 */
+    Tcl_DetachChannel, /* 438 */
+    Tcl_IsStandardChannel, /* 439 */
+    Tcl_FSCopyFile, /* 440 */
+    Tcl_FSCopyDirectory, /* 441 */
+    Tcl_FSCreateDirectory, /* 442 */
+    Tcl_FSDeleteFile, /* 443 */
+    Tcl_FSLoadFile, /* 444 */
+    Tcl_FSMatchInDirectory, /* 445 */
+    Tcl_FSLink, /* 446 */
+    Tcl_FSRemoveDirectory, /* 447 */
+    Tcl_FSRenameFile, /* 448 */
+    Tcl_FSLstat, /* 449 */
+    Tcl_FSUtime, /* 450 */
+    Tcl_FSFileAttrsGet, /* 451 */
+    Tcl_FSFileAttrsSet, /* 452 */
+    Tcl_FSFileAttrStrings, /* 453 */
+    Tcl_FSStat, /* 454 */
+    Tcl_FSAccess, /* 455 */
+    Tcl_FSOpenFileChannel, /* 456 */
+    Tcl_FSGetCwd, /* 457 */
+    Tcl_FSChdir, /* 458 */
+    Tcl_FSConvertToPathType, /* 459 */
+    Tcl_FSJoinPath, /* 460 */
+    Tcl_FSSplitPath, /* 461 */
+    Tcl_FSEqualPaths, /* 462 */
+    Tcl_FSGetNormalizedPath, /* 463 */
+    Tcl_FSJoinToPath, /* 464 */
+    Tcl_FSGetInternalRep, /* 465 */
+    Tcl_FSGetTranslatedPath, /* 466 */
+    Tcl_FSEvalFile, /* 467 */
+    Tcl_FSNewNativePath, /* 468 */
+    Tcl_FSGetNativePath, /* 469 */
+    Tcl_FSFileSystemInfo, /* 470 */
+    Tcl_FSPathSeparator, /* 471 */
+    Tcl_FSListVolumes, /* 472 */
+    Tcl_FSRegister, /* 473 */
+    Tcl_FSUnregister, /* 474 */
+    Tcl_FSData, /* 475 */
+    Tcl_FSGetTranslatedStringPath, /* 476 */
+    Tcl_FSGetFileSystemForPath, /* 477 */
+    Tcl_FSGetPathType, /* 478 */
+    Tcl_OutputBuffered, /* 479 */
+    Tcl_FSMountsChanged, /* 480 */
+    Tcl_EvalTokensStandard, /* 481 */
+    Tcl_GetTime, /* 482 */
+    Tcl_CreateObjTrace, /* 483 */
+    Tcl_GetCommandInfoFromToken, /* 484 */
+    Tcl_SetCommandInfoFromToken, /* 485 */
+    Tcl_DbNewWideIntObj, /* 486 */
+    Tcl_GetWideIntFromObj, /* 487 */
+    Tcl_NewWideIntObj, /* 488 */
+    Tcl_SetWideIntObj, /* 489 */
+    Tcl_AllocStatBuf, /* 490 */
+    Tcl_Seek, /* 491 */
+    Tcl_Tell, /* 492 */
+    Tcl_ChannelWideSeekProc, /* 493 */
+    Tcl_DictObjPut, /* 494 */
+    Tcl_DictObjGet, /* 495 */
+    Tcl_DictObjRemove, /* 496 */
+    Tcl_DictObjSize, /* 497 */
+    Tcl_DictObjFirst, /* 498 */
+    Tcl_DictObjNext, /* 499 */
+    Tcl_DictObjDone, /* 500 */
+    Tcl_DictObjPutKeyList, /* 501 */
+    Tcl_DictObjRemoveKeyList, /* 502 */
+    Tcl_NewDictObj, /* 503 */
+    Tcl_DbNewDictObj, /* 504 */
+    Tcl_RegisterConfig, /* 505 */
+    Tcl_CreateNamespace, /* 506 */
+    Tcl_DeleteNamespace, /* 507 */
+    Tcl_AppendExportList, /* 508 */
+    Tcl_Export, /* 509 */
+    Tcl_Import, /* 510 */
+    Tcl_ForgetImport, /* 511 */
+    Tcl_GetCurrentNamespace, /* 512 */
+    Tcl_GetGlobalNamespace, /* 513 */
+    Tcl_FindNamespace, /* 514 */
+    Tcl_FindCommand, /* 515 */
+    Tcl_GetCommandFromObj, /* 516 */
+    Tcl_GetCommandFullName, /* 517 */
+    Tcl_FSEvalFileEx, /* 518 */
+    Tcl_SetExitProc, /* 519 */
+    Tcl_LimitAddHandler, /* 520 */
+    Tcl_LimitRemoveHandler, /* 521 */
+    Tcl_LimitReady, /* 522 */
+    Tcl_LimitCheck, /* 523 */
+    Tcl_LimitExceeded, /* 524 */
+    Tcl_LimitSetCommands, /* 525 */
+    Tcl_LimitSetTime, /* 526 */
+    Tcl_LimitSetGranularity, /* 527 */
+    Tcl_LimitTypeEnabled, /* 528 */
+    Tcl_LimitTypeExceeded, /* 529 */
+    Tcl_LimitTypeSet, /* 530 */
+    Tcl_LimitTypeReset, /* 531 */
+    Tcl_LimitGetCommands, /* 532 */
+    Tcl_LimitGetTime, /* 533 */
+    Tcl_LimitGetGranularity, /* 534 */
+    Tcl_SaveInterpState, /* 535 */
+    Tcl_RestoreInterpState, /* 536 */
+    Tcl_DiscardInterpState, /* 537 */
+    Tcl_SetReturnOptions, /* 538 */
+    Tcl_GetReturnOptions, /* 539 */
+    Tcl_IsEnsemble, /* 540 */
+    Tcl_CreateEnsemble, /* 541 */
+    Tcl_FindEnsemble, /* 542 */
+    Tcl_SetEnsembleSubcommandList, /* 543 */
+    Tcl_SetEnsembleMappingDict, /* 544 */
+    Tcl_SetEnsembleUnknownHandler, /* 545 */
+    Tcl_SetEnsembleFlags, /* 546 */
+    Tcl_GetEnsembleSubcommandList, /* 547 */
+    Tcl_GetEnsembleMappingDict, /* 548 */
+    Tcl_GetEnsembleUnknownHandler, /* 549 */
+    Tcl_GetEnsembleFlags, /* 550 */
+    Tcl_GetEnsembleNamespace, /* 551 */
+    Tcl_SetTimeProc, /* 552 */
+    Tcl_QueryTimeProc, /* 553 */
+    Tcl_ChannelThreadActionProc, /* 554 */
+    Tcl_NewBignumObj, /* 555 */
+    Tcl_DbNewBignumObj, /* 556 */
+    Tcl_SetBignumObj, /* 557 */
+    Tcl_GetBignumFromObj, /* 558 */
+    Tcl_TakeBignumFromObj, /* 559 */
+    Tcl_TruncateChannel, /* 560 */
+    Tcl_ChannelTruncateProc, /* 561 */
+    Tcl_SetChannelErrorInterp, /* 562 */
+    Tcl_GetChannelErrorInterp, /* 563 */
+    Tcl_SetChannelError, /* 564 */
+    Tcl_GetChannelError, /* 565 */
+    Tcl_InitBignumFromDouble, /* 566 */
+    Tcl_GetNamespaceUnknownHandler, /* 567 */
+    Tcl_SetNamespaceUnknownHandler, /* 568 */
+    Tcl_GetEncodingFromObj, /* 569 */
+    Tcl_GetEncodingSearchPath, /* 570 */
+    Tcl_SetEncodingSearchPath, /* 571 */
+    Tcl_GetEncodingNameFromEnvironment, /* 572 */
+    Tcl_PkgRequireProc, /* 573 */
+    Tcl_AppendObjToErrorInfo, /* 574 */
+    Tcl_AppendLimitedToObj, /* 575 */
+    Tcl_Format, /* 576 */
+    Tcl_AppendFormatToObj, /* 577 */
+    Tcl_ObjPrintf, /* 578 */
+    Tcl_AppendPrintfToObj, /* 579 */
+    Tcl_CancelEval, /* 580 */
+    Tcl_Canceled, /* 581 */
+    Tcl_CreatePipe, /* 582 */
+    Tcl_NRCreateCommand, /* 583 */
+    Tcl_NREvalObj, /* 584 */
+    Tcl_NREvalObjv, /* 585 */
+    Tcl_NRCmdSwap, /* 586 */
+    Tcl_NRAddCallback, /* 587 */
+    Tcl_NRCallObjProc, /* 588 */
+    Tcl_GetFSDeviceFromStat, /* 589 */
+    Tcl_GetFSInodeFromStat, /* 590 */
+    Tcl_GetModeFromStat, /* 591 */
+    Tcl_GetLinkCountFromStat, /* 592 */
+    Tcl_GetUserIdFromStat, /* 593 */
+    Tcl_GetGroupIdFromStat, /* 594 */
+    Tcl_GetDeviceTypeFromStat, /* 595 */
+    Tcl_GetAccessTimeFromStat, /* 596 */
+    Tcl_GetModificationTimeFromStat, /* 597 */
+    Tcl_GetChangeTimeFromStat, /* 598 */
+    Tcl_GetSizeFromStat, /* 599 */
+    Tcl_GetBlocksFromStat, /* 600 */
+    Tcl_GetBlockSizeFromStat, /* 601 */
+    Tcl_SetEnsembleParameterList, /* 602 */
+    Tcl_GetEnsembleParameterList, /* 603 */
+    Tcl_ParseArgsObjv, /* 604 */
+    Tcl_GetErrorLine, /* 605 */
+    Tcl_SetErrorLine, /* 606 */
+    Tcl_TransferResult, /* 607 */
+    Tcl_InterpActive, /* 608 */
+    Tcl_BackgroundException, /* 609 */
+    Tcl_ZlibDeflate, /* 610 */
+    Tcl_ZlibInflate, /* 611 */
+    Tcl_ZlibCRC32, /* 612 */
+    Tcl_ZlibAdler32, /* 613 */
+    Tcl_ZlibStreamInit, /* 614 */
+    Tcl_ZlibStreamGetCommandName, /* 615 */
+    Tcl_ZlibStreamEof, /* 616 */
+    Tcl_ZlibStreamChecksum, /* 617 */
+    Tcl_ZlibStreamPut, /* 618 */
+    Tcl_ZlibStreamGet, /* 619 */
+    Tcl_ZlibStreamClose, /* 620 */
+    Tcl_ZlibStreamReset, /* 621 */
+    Tcl_SetStartupScript, /* 622 */
+    Tcl_GetStartupScript, /* 623 */
+    Tcl_CloseEx, /* 624 */
+    Tcl_NRExprObj, /* 625 */
+    Tcl_NRSubstObj, /* 626 */
+    Tcl_LoadFile, /* 627 */
+    Tcl_FindSymbol, /* 628 */
+    Tcl_FSUnloadFile, /* 629 */
+    Tcl_ZlibStreamSetCompressionDictionary, /* 630 */
+    Tcl_OpenTcpServerEx, /* 631 */
+};
+
+/* !END!: Do not edit above this line. */
diff --git a/generic/tclStubLib.c b/generic/tclStubLib.c
new file mode 100644
index 0000000..5261591
--- /dev/null
+++ b/generic/tclStubLib.c
@@ -0,0 +1,129 @@
+/*
+ * tclStubLib.c --
+ *
+ *	Stub object that will be statically linked into extensions that want
+ *	to access Tcl.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 1998 Paul Duffin.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+MODULE_SCOPE const TclStubs *tclStubsPtr;
+MODULE_SCOPE const TclPlatStubs *tclPlatStubsPtr;
+MODULE_SCOPE const TclIntStubs *tclIntStubsPtr;
+MODULE_SCOPE const TclIntPlatStubs *tclIntPlatStubsPtr;
+
+const TclStubs *tclStubsPtr = NULL;
+const TclPlatStubs *tclPlatStubsPtr = NULL;
+const TclIntStubs *tclIntStubsPtr = NULL;
+const TclIntPlatStubs *tclIntPlatStubsPtr = NULL;
+
+/*
+ * Use our own ISDIGIT to avoid linking to libc on windows
+ */
+
+#define ISDIGIT(c) (((unsigned)((c)-'0')) <= 9)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_InitStubs --
+ *
+ *	Tries to initialise the stub table pointers and ensures that the
+ *	correct version of Tcl is loaded.
+ *
+ * Results:
+ *	The actual version of Tcl that satisfies the request, or NULL to
+ *	indicate that an error occurred.
+ *
+ * Side effects:
+ *	Sets the stub table pointers.
+ *
+ *----------------------------------------------------------------------
+ */
+#undef Tcl_InitStubs
+MODULE_SCOPE const char *
+Tcl_InitStubs(
+    Tcl_Interp *interp,
+    const char *version,
+    int exact,
+    int magic)
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *actualVersion = NULL;
+    ClientData pkgData = NULL;
+    const TclStubs *stubsPtr = iPtr->stubTable;
+
+    /*
+     * We can't optimize this check by caching tclStubsPtr because that
+     * prevents apps from being able to load/unload Tcl dynamically multiple
+     * times. [Bug 615304]
+     */
+
+    if (!stubsPtr || (stubsPtr->magic != (((exact&0xff00) >= 0x900) ? magic : TCL_STUB_MAGIC))) {
+	iPtr->result = (char *)"interpreter uses an incompatible stubs mechanism";
+	iPtr->freeProc = 0;
+	return NULL;
+    }
+
+    actualVersion = stubsPtr->tcl_PkgRequireEx(interp, "Tcl", version, 0, &pkgData);
+    if (actualVersion == NULL) {
+	return NULL;
+    }
+    if (exact&1) {
+	const char *p = version;
+	int count = 0;
+
+	while (*p) {
+	    count += !ISDIGIT(*p++);
+	}
+	if (count == 1) {
+	    const char *q = actualVersion;
+
+	    p = version;
+	    while (*p && (*p == *q)) {
+		p++; q++;
+	    }
+	    if (*p || ISDIGIT(*q)) {
+		/* Construct error message */
+		stubsPtr->tcl_PkgRequireEx(interp, "Tcl", version, 1, NULL);
+		return NULL;
+	    }
+	} else {
+	    actualVersion = stubsPtr->tcl_PkgRequireEx(interp, "Tcl", version, 1, NULL);
+	    if (actualVersion == NULL) {
+		return NULL;
+	    }
+	}
+    }
+    if (((exact&0xff00) < 0x900)) {
+	/* We are running Tcl 8.x */
+	stubsPtr = (TclStubs *)pkgData;
+    }
+    tclStubsPtr = stubsPtr;
+
+    if (stubsPtr->hooks) {
+	tclPlatStubsPtr = stubsPtr->hooks->tclPlatStubs;
+	tclIntStubsPtr = stubsPtr->hooks->tclIntStubs;
+	tclIntPlatStubsPtr = stubsPtr->hooks->tclIntPlatStubs;
+    } else {
+	tclPlatStubsPtr = NULL;
+	tclIntStubsPtr = NULL;
+	tclIntPlatStubsPtr = NULL;
+    }
+
+    return actualVersion;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclStubLibTbl.c b/generic/tclStubLibTbl.c
new file mode 100644
index 0000000..0391502
--- /dev/null
+++ b/generic/tclStubLibTbl.c
@@ -0,0 +1,58 @@
+/*
+ * tclStubLibTbl.c --
+ *
+ *	Stub object that will be statically linked into extensions that want
+ *	to access Tcl.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 1998 Paul Duffin.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitStubTable --
+ *
+ *	Initialize the stub table, using the structure pointed at
+ *	by the "version" argument.
+ *
+ * Results:
+ *	Outputs the value of the "version" argument.
+ *
+ * Side effects:
+ *	Sets the stub table pointers.
+ *
+ *----------------------------------------------------------------------
+ */
+MODULE_SCOPE const char *
+TclInitStubTable(
+	const char *version) /* points to the version field of a
+	                        TclStubInfoType structure variable. */
+{
+    tclStubsPtr = ((const TclStubInfoType *) version)->stubs;
+
+    if (tclStubsPtr->hooks) {
+	tclPlatStubsPtr = tclStubsPtr->hooks->tclPlatStubs;
+	tclIntStubsPtr = tclStubsPtr->hooks->tclIntStubs;
+	tclIntPlatStubsPtr = tclStubsPtr->hooks->tclIntPlatStubs;
+    } else {
+	tclPlatStubsPtr = NULL;
+	tclIntStubsPtr = NULL;
+	tclIntPlatStubsPtr = NULL;
+    }
+
+    return version;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclTest.c b/generic/tclTest.c
new file mode 100644
index 0000000..ebd90ae
--- /dev/null
+++ b/generic/tclTest.c
@@ -0,0 +1,7701 @@
+/*
+ * tclTest.c --
+ *
+ *	This file contains C command functions for a bunch of additional Tcl
+ *	commands that are used for testing out Tcl's C interfaces. These
+ *	commands are not normally included in Tcl applications; they're only
+ *	used for testing.
+ *
+ * Copyright (c) 1993-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 Ajuba Solutions.
+ * Copyright (c) 2003 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#undef STATIC_BUILD
+#ifndef USE_TCL_STUBS
+#   define USE_TCL_STUBS
+#endif
+#include "tclInt.h"
+#include "tclOO.h"
+#include <math.h>
+
+/*
+ * Required for Testregexp*Cmd
+ */
+#include "tclRegexp.h"
+
+/*
+ * Required for TestlocaleCmd
+ */
+#include <locale.h>
+
+/*
+ * Required for the TestChannelCmd and TestChannelEventCmd
+ */
+#include "tclIO.h"
+
+/*
+ * Declare external functions used in Windows tests.
+ */
+
+DLLEXPORT int		Tcltest_Init(Tcl_Interp *interp);
+DLLEXPORT int		Tcltest_SafeInit(Tcl_Interp *interp);
+
+/*
+ * Dynamic string shared by TestdcallCmd and DelCallbackProc; used to collect
+ * the results of the various deletion callbacks.
+ */
+
+static Tcl_DString delString;
+static Tcl_Interp *delInterp;
+
+/*
+ * One of the following structures exists for each asynchronous handler
+ * created by the "testasync" command".
+ */
+
+typedef struct TestAsyncHandler {
+    int id;			/* Identifier for this handler. */
+    Tcl_AsyncHandler handler;	/* Tcl's token for the handler. */
+    char *command;		/* Command to invoke when the handler is
+				 * invoked. */
+    struct TestAsyncHandler *nextPtr;
+				/* Next is list of handlers. */
+} TestAsyncHandler;
+
+/*
+ * Start of the socket driver state structure to acces field testFlags
+ */
+
+typedef struct TcpState TcpState;
+
+struct TcpState {
+    Tcl_Channel channel;	/* Channel associated with this socket. */
+    int testFlags;              /* bit field for tests. Is set by testsocket
+                                 * test procedure */
+};
+
+TCL_DECLARE_MUTEX(asyncTestMutex)
+
+static TestAsyncHandler *firstHandler = NULL;
+
+/*
+ * The dynamic string below is used by the "testdstring" command to test the
+ * dynamic string facilities.
+ */
+
+static Tcl_DString dstring;
+
+/*
+ * The command trace below is used by the "testcmdtraceCmd" command to test
+ * the command tracing facilities.
+ */
+
+static Tcl_Trace cmdTrace;
+
+/*
+ * One of the following structures exists for each command created by
+ * TestdelCmd:
+ */
+
+typedef struct {
+    Tcl_Interp *interp;		/* Interpreter in which command exists. */
+    char *deleteCmd;		/* Script to execute when command is deleted.
+				 * Malloc'ed. */
+} DelCmd;
+
+/*
+ * The following is used to keep track of an encoding that invokes a Tcl
+ * command.
+ */
+
+typedef struct {
+    Tcl_Interp *interp;
+    char *toUtfCmd;
+    char *fromUtfCmd;
+} TclEncoding;
+
+/*
+ * The counter below is used to determine if the TestsaveresultFree routine
+ * was called for a result.
+ */
+
+static int freeCount;
+
+/*
+ * Boolean flag used by the "testsetmainloop" and "testexitmainloop" commands.
+ */
+
+static int exitMainLoop = 0;
+
+/*
+ * Event structure used in testing the event queue management procedures.
+ */
+
+typedef struct {
+    Tcl_Event header;		/* Header common to all events */
+    Tcl_Interp *interp;		/* Interpreter that will handle the event */
+    Tcl_Obj *command;		/* Command to evaluate when the event occurs */
+    Tcl_Obj *tag;		/* Tag for this event used to delete it */
+} TestEvent;
+
+/*
+ * Simple detach/attach facility for testchannel cut|splice. Allow testing of
+ * channel transfer in core testsuite.
+ */
+
+typedef struct TestChannel {
+    Tcl_Channel chan;		/* Detached channel */
+    struct TestChannel *nextPtr;/* Next in detached channel pool */
+} TestChannel;
+
+static TestChannel *firstDetached;
+
+/*
+ * Forward declarations for procedures defined later in this file:
+ */
+
+static int		AsyncHandlerProc(ClientData clientData,
+			    Tcl_Interp *interp, int code);
+#ifdef TCL_THREADS
+static Tcl_ThreadCreateType AsyncThreadProc(ClientData);
+#endif
+static void		CleanupTestSetassocdataTests(
+			    ClientData clientData, Tcl_Interp *interp);
+static void		CmdDelProc1(ClientData clientData);
+static void		CmdDelProc2(ClientData clientData);
+static int		CmdProc1(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		CmdProc2(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static void		CmdTraceDeleteProc(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int level, char *command, Tcl_CmdProc *cmdProc,
+			    ClientData cmdClientData, int argc,
+			    const char *argv[]);
+static void		CmdTraceProc(ClientData clientData,
+			    Tcl_Interp *interp, int level, char *command,
+			    Tcl_CmdProc *cmdProc, ClientData cmdClientData,
+			    int argc, const char *argv[]);
+static int		CreatedCommandProc(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int argc, const char **argv);
+static int		CreatedCommandProc2(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int argc, const char **argv);
+static void		DelCallbackProc(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		DelCmdProc(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static void		DelDeleteProc(ClientData clientData);
+static void		EncodingFreeProc(ClientData clientData);
+static int		EncodingToUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst,
+			    int dstLen, int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static int		EncodingFromUtfProc(ClientData clientData,
+			    const char *src, int srcLen, int flags,
+			    Tcl_EncodingState *statePtr, char *dst,
+			    int dstLen, int *srcReadPtr, int *dstWrotePtr,
+			    int *dstCharsPtr);
+static void		ExitProcEven(ClientData clientData);
+static void		ExitProcOdd(ClientData clientData);
+static int		GetTimesObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		MainLoop(void);
+static int		NoopCmd(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		NoopObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		ObjTraceProc(ClientData clientData,
+			    Tcl_Interp *interp, int level, const char *command,
+			    Tcl_Command commandToken, int objc,
+			    Tcl_Obj *const objv[]);
+static void		ObjTraceDeleteProc(ClientData clientData);
+static void		PrintParse(Tcl_Interp *interp, Tcl_Parse *parsePtr);
+static void		SpecialFree(char *blockPtr);
+static int		StaticInitProc(Tcl_Interp *interp);
+static int		TestasyncCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestbytestringObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestcmdinfoCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestcmdtokenCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestcmdtraceCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestconcatobjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestcreatecommandCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestdcallCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestdelCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestdelassocdataCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestdoubledigitsObjCmd(ClientData dummy,
+					       Tcl_Interp* interp,
+					       int objc, Tcl_Obj* const objv[]);
+static int		TestdstringCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestencodingObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestevalexObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestevalobjvObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TesteventObjCmd(ClientData unused,
+			    Tcl_Interp *interp, int argc,
+			    Tcl_Obj *const objv[]);
+static int		TesteventProc(Tcl_Event *event, int flags);
+static int		TesteventDeleteProc(Tcl_Event *event,
+			    ClientData clientData);
+static int		TestexithandlerCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestexprlongCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestexprlongobjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestexprdoubleCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestexprdoubleobjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestexprparserObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestexprstringCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestfileCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]);
+static int		TestfilelinkCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]);
+static int		TestfeventCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestgetassocdataCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestgetintCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestgetplatformCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestgetvarfullnameCmd(
+			    ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestinterpdeleteCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestlinkCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestlocaleCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#ifndef TCL_NO_DEPRECATED
+static int		TestMathFunc(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_Value *args,
+			    Tcl_Value *resultPtr);
+static int		TestMathFunc2(ClientData clientData,
+			    Tcl_Interp *interp, Tcl_Value *args,
+			    Tcl_Value *resultPtr);
+#endif /* TCL_NO_DEPRECATED */
+static int		TestmainthreadCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestsetmainloopCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestexitmainloopCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestpanicCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestparseargsCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestparserObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestparsevarObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestparsevarnameObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestpreferstableObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestprintObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestregexpObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestreturnObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		TestregexpXflags(const char *string,
+			    int length, int *cflagsPtr, int *eflagsPtr);
+static int		TestsaveresultCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		TestsaveresultFree(char *blockPtr);
+static int		TestsetassocdataCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestsetCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		Testset2Cmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestseterrorcodeCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestsetobjerrorcodeCmd(
+			    ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestsetplatformCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TeststaticpkgCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TesttranslatefilenameCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestupvarCmd(ClientData dummy,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestWrongNumArgsObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestGetIndexFromObjStructObjCmd(
+			    ClientData clientData, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestChannelCmd(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestChannelEventCmd(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestSocketCmd(ClientData clientData,
+			    Tcl_Interp *interp, int argc, const char **argv);
+static int		TestFilesystemObjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestSimpleFilesystemObjCmd(
+			    ClientData dummy, Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static void		TestReport(const char *cmd, Tcl_Obj *arg1,
+			    Tcl_Obj *arg2);
+static Tcl_Obj *	TestReportGetNativePath(Tcl_Obj *pathPtr);
+static Tcl_FSStatProc TestReportStat;
+static Tcl_FSAccessProc TestReportAccess;
+static Tcl_FSOpenFileChannelProc TestReportOpenFileChannel;
+static Tcl_FSMatchInDirectoryProc TestReportMatchInDirectory;
+static Tcl_FSChdirProc TestReportChdir;
+static Tcl_FSLstatProc TestReportLstat;
+static Tcl_FSCopyFileProc TestReportCopyFile;
+static Tcl_FSDeleteFileProc TestReportDeleteFile;
+static Tcl_FSRenameFileProc TestReportRenameFile;
+static Tcl_FSCreateDirectoryProc TestReportCreateDirectory;
+static Tcl_FSCopyDirectoryProc TestReportCopyDirectory;
+static Tcl_FSRemoveDirectoryProc TestReportRemoveDirectory;
+static int TestReportLoadFile(Tcl_Interp *interp, Tcl_Obj *pathPtr,
+	Tcl_LoadHandle *handlePtr, Tcl_FSUnloadFileProc **unloadProcPtr);
+static Tcl_FSLinkProc TestReportLink;
+static Tcl_FSFileAttrStringsProc TestReportFileAttrStrings;
+static Tcl_FSFileAttrsGetProc TestReportFileAttrsGet;
+static Tcl_FSFileAttrsSetProc TestReportFileAttrsSet;
+static Tcl_FSUtimeProc TestReportUtime;
+static Tcl_FSNormalizePathProc TestReportNormalizePath;
+static Tcl_FSPathInFilesystemProc TestReportInFilesystem;
+static Tcl_FSFreeInternalRepProc TestReportFreeInternalRep;
+static Tcl_FSDupInternalRepProc TestReportDupInternalRep;
+
+static Tcl_FSStatProc SimpleStat;
+static Tcl_FSAccessProc SimpleAccess;
+static Tcl_FSOpenFileChannelProc SimpleOpenFileChannel;
+static Tcl_FSListVolumesProc SimpleListVolumes;
+static Tcl_FSPathInFilesystemProc SimplePathInFilesystem;
+static Tcl_Obj *	SimpleRedirect(Tcl_Obj *pathPtr);
+static Tcl_FSMatchInDirectoryProc SimpleMatchInDirectory;
+static int		TestNumUtfCharsCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestHashSystemHashCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+
+static Tcl_NRPostProc	NREUnwind_callback;
+static int		TestNREUnwind(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestNRELevels(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestInterpResolverCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+#if defined(HAVE_CPUID) || defined(_WIN32)
+static int		TestcpuidCmd(ClientData dummy,
+			    Tcl_Interp* interp, int objc,
+			    Tcl_Obj *const objv[]);
+#endif
+
+static const Tcl_Filesystem testReportingFilesystem = {
+    "reporting",
+    sizeof(Tcl_Filesystem),
+    TCL_FILESYSTEM_VERSION_1,
+    TestReportInFilesystem, /* path in */
+    TestReportDupInternalRep,
+    TestReportFreeInternalRep,
+    NULL, /* native to norm */
+    NULL, /* convert to native */
+    TestReportNormalizePath,
+    NULL, /* path type */
+    NULL, /* separator */
+    TestReportStat,
+    TestReportAccess,
+    TestReportOpenFileChannel,
+    TestReportMatchInDirectory,
+    TestReportUtime,
+    TestReportLink,
+    NULL /* list volumes */,
+    TestReportFileAttrStrings,
+    TestReportFileAttrsGet,
+    TestReportFileAttrsSet,
+    TestReportCreateDirectory,
+    TestReportRemoveDirectory,
+    TestReportDeleteFile,
+    TestReportCopyFile,
+    TestReportRenameFile,
+    TestReportCopyDirectory,
+    TestReportLstat,
+    (Tcl_FSLoadFileProc *) TestReportLoadFile,
+    NULL /* cwd */,
+    TestReportChdir
+};
+
+static const Tcl_Filesystem simpleFilesystem = {
+    "simple",
+    sizeof(Tcl_Filesystem),
+    TCL_FILESYSTEM_VERSION_1,
+    SimplePathInFilesystem,
+    NULL,
+    NULL,
+    /* No internal to normalized, since we don't create any
+     * pure 'internal' Tcl_Obj path representations */
+    NULL,
+    /* No create native rep function, since we don't use it
+     * or 'Tcl_FSNewNativePath' */
+    NULL,
+    /* Normalize path isn't needed - we assume paths only have
+     * one representation */
+    NULL,
+    NULL,
+    NULL,
+    SimpleStat,
+    SimpleAccess,
+    SimpleOpenFileChannel,
+    SimpleMatchInDirectory,
+    NULL,
+    /* We choose not to support symbolic links inside our vfs's */
+    NULL,
+    SimpleListVolumes,
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    /* No copy file - fallback will occur at Tcl level */
+    NULL,
+    /* No rename file - fallback will occur at Tcl level */
+    NULL,
+    /* No copy directory - fallback will occur at Tcl level */
+    NULL,
+    /* Use stat for lstat */
+    NULL,
+    /* No load - fallback on core implementation */
+    NULL,
+    /* We don't need a getcwd or chdir - fallback on Tcl's versions */
+    NULL,
+    NULL
+};
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcltest_Init --
+ *
+ *	This procedure performs application-specific initialization. Most
+ *	applications, especially those that incorporate additional packages,
+ *	will have their own version of this procedure.
+ *
+ * Results:
+ *	Returns a standard Tcl completion code, and leaves an error message in
+ *	the interp's result if an error occurs.
+ *
+ * Side effects:
+ *	Depends on the startup script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcltest_Init(
+    Tcl_Interp *interp)		/* Interpreter for application. */
+{
+#ifndef TCL_NO_DEPRECATED
+    Tcl_ValueType t3ArgTypes[2];
+#endif /* TCL_NO_DEPRECATED */
+
+    Tcl_Obj *listPtr;
+    Tcl_Obj **objv;
+    int objc, index;
+    static const char *const specialOptions[] = {
+	"-appinitprocerror", "-appinitprocdeleteinterp",
+	"-appinitprocclosestderr", "-appinitprocsetrcfile", NULL
+    };
+
+    if (Tcl_InitStubs(interp, "8.5-", 0) == NULL) {
+	return TCL_ERROR;
+    }
+    if (Tcl_TomMath_InitStubs(interp, "8.5-") == NULL) {
+	return TCL_ERROR;
+    }
+    if (Tcl_OOInitStubs(interp) == NULL) {
+	return TCL_ERROR;
+    }
+    /* TIP #268: Full patchlevel instead of just major.minor */
+
+    if (Tcl_PkgProvideEx(interp, "Tcltest", TCL_PATCH_LEVEL, NULL) == TCL_ERROR) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create additional commands and math functions for testing Tcl.
+     */
+
+    Tcl_CreateObjCommand(interp, "gettimes", GetTimesObjCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "noop", NoopCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "noop", NoopObjCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testbytestring", TestbytestringObjCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testwrongnumargs", TestWrongNumArgsObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testfilesystem", TestFilesystemObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testsimplefilesystem", TestSimpleFilesystemObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testgetindexfromobjstruct",
+	    TestGetIndexFromObjStructObjCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testasync", TestasyncCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testchannel", TestChannelCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testchannelevent", TestChannelEventCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testcmdtoken", TestcmdtokenCmd, NULL,
+	    NULL);
+    Tcl_CreateCommand(interp, "testcmdinfo", TestcmdinfoCmd, NULL,
+	    NULL);
+    Tcl_CreateCommand(interp, "testcmdtrace", TestcmdtraceCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testconcatobj", TestconcatobjCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testcreatecommand", TestcreatecommandCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testdcall", TestdcallCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testdel", TestdelCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testdelassocdata", TestdelassocdataCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testdoubledigits", TestdoubledigitsObjCmd,
+			 NULL, NULL);
+    Tcl_DStringInit(&dstring);
+    Tcl_CreateCommand(interp, "testdstring", TestdstringCmd, NULL,
+	    NULL);
+    Tcl_CreateObjCommand(interp, "testencoding", TestencodingObjCmd, NULL,
+	    NULL);
+    Tcl_CreateObjCommand(interp, "testevalex", TestevalexObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testevalobjv", TestevalobjvObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testevent", TesteventObjCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testexithandler", TestexithandlerCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testexprlong", TestexprlongCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testexprlongobj", TestexprlongobjCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testexprdouble", TestexprdoubleCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testexprdoubleobj", TestexprdoubleobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testexprparser", TestexprparserObjCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testexprstring", TestexprstringCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testfevent", TestfeventCmd, NULL,
+	    NULL);
+    Tcl_CreateObjCommand(interp, "testfilelink", TestfilelinkCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testfile", TestfileCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testhashsystemhash",
+	    TestHashSystemHashCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testgetassocdata", TestgetassocdataCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testgetint", TestgetintCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testgetplatform", TestgetplatformCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testgetvarfullname",
+	    TestgetvarfullnameCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testinterpdelete", TestinterpdeleteCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testlink", TestlinkCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testlocale", TestlocaleCmd, NULL,
+	    NULL);
+    Tcl_CreateCommand(interp, "testpanic", TestpanicCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testparseargs", TestparseargsCmd,NULL,NULL);
+    Tcl_CreateObjCommand(interp, "testparser", TestparserObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testparsevar", TestparsevarObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testparsevarname", TestparsevarnameObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testpreferstable", TestpreferstableObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testprint", TestprintObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testregexp", TestregexpObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testreturn", TestreturnObjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testsaveresult", TestsaveresultCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testsetassocdata", TestsetassocdataCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testsetnoerr", TestsetCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testseterr", TestsetCmd,
+	    (ClientData) TCL_LEAVE_ERR_MSG, NULL);
+    Tcl_CreateCommand(interp, "testset2", Testset2Cmd,
+	    (ClientData) TCL_LEAVE_ERR_MSG, NULL);
+    Tcl_CreateCommand(interp, "testseterrorcode", TestseterrorcodeCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testsetobjerrorcode",
+	    TestsetobjerrorcodeCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testnumutfchars",
+	    TestNumUtfCharsCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testsetplatform", TestsetplatformCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testsocket", TestSocketCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "teststaticpkg", TeststaticpkgCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testtranslatefilename",
+	    TesttranslatefilenameCmd, NULL, NULL);
+    Tcl_CreateCommand(interp, "testupvar", TestupvarCmd, NULL, NULL);
+#ifndef TCL_NO_DEPRECATED
+    Tcl_CreateMathFunc(interp, "T1", 0, NULL, TestMathFunc, (ClientData) 123);
+    Tcl_CreateMathFunc(interp, "T2", 0, NULL, TestMathFunc, (ClientData) 345);
+#endif /* TCL_NO_DEPRECATED */
+    Tcl_CreateCommand(interp, "testmainthread", TestmainthreadCmd, NULL,
+	    NULL);
+    Tcl_CreateCommand(interp, "testsetmainloop", TestsetmainloopCmd,
+	    NULL, NULL);
+    Tcl_CreateCommand(interp, "testexitmainloop", TestexitmainloopCmd,
+	    NULL, NULL);
+#if defined(HAVE_CPUID) || defined(_WIN32)
+    Tcl_CreateObjCommand(interp, "testcpuid", TestcpuidCmd,
+	    (ClientData) 0, NULL);
+#endif
+#ifndef TCL_NO_DEPRECATED
+    t3ArgTypes[0] = TCL_EITHER;
+    t3ArgTypes[1] = TCL_EITHER;
+    Tcl_CreateMathFunc(interp, "T3", 2, t3ArgTypes, TestMathFunc2,
+	    NULL);
+#endif /* TCL_NO_DEPRECATED */
+
+    Tcl_CreateObjCommand(interp, "testnreunwind", TestNREUnwind,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testnrelevels", TestNRELevels,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testinterpresolver", TestInterpResolverCmd,
+	    NULL, NULL);
+
+    if (TclObjTest_Init(interp) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Procbodytest_Init(interp) != TCL_OK) {
+	return TCL_ERROR;
+    }
+#ifdef TCL_THREADS
+    if (TclThread_Init(interp) != TCL_OK) {
+	return TCL_ERROR;
+    }
+#endif
+
+    /*
+     * Check for special options used in ../tests/main.test
+     */
+
+    listPtr = Tcl_GetVar2Ex(interp, "argv", NULL, TCL_GLOBAL_ONLY);
+    if (listPtr != NULL) {
+	if (Tcl_ListObjGetElements(interp, listPtr, &objc, &objv) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (objc && (Tcl_GetIndexFromObj(NULL, objv[0], specialOptions, NULL,
+		TCL_EXACT, &index) == TCL_OK)) {
+	    switch (index) {
+	    case 0:
+		return TCL_ERROR;
+	    case 1:
+		Tcl_DeleteInterp(interp);
+		return TCL_ERROR;
+	    case 2: {
+		int mode;
+		Tcl_UnregisterChannel(interp,
+			Tcl_GetChannel(interp, "stderr", &mode));
+		return TCL_ERROR;
+	    }
+	    case 3:
+		if (objc-1) {
+		    Tcl_SetVar2Ex(interp, "tcl_rcFileName", NULL, objv[1],
+			    TCL_GLOBAL_ONLY);
+		}
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    /*
+     * And finally add any platform specific test commands.
+     */
+
+    return TclplatformtestInit(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcltest_SafeInit --
+ *
+ *	This procedure performs application-specific initialization. Most
+ *	applications, especially those that incorporate additional packages,
+ *	will have their own version of this procedure.
+ *
+ * Results:
+ *	Returns a standard Tcl completion code, and leaves an error message in
+ *	the interp's result if an error occurs.
+ *
+ * Side effects:
+ *	Depends on the startup script.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcltest_SafeInit(
+    Tcl_Interp *interp)		/* Interpreter for application. */
+{
+    if (Tcl_InitStubs(interp, "8.5-", 0) == NULL) {
+	return TCL_ERROR;
+    }
+    return Procbodytest_SafeInit(interp);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestasyncCmd --
+ *
+ *	This procedure implements the "testasync" command.  It is used
+ *	to test the asynchronous handler facilities of Tcl.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates, deletes, and invokes handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestasyncCmd(
+    ClientData dummy,			/* Not used. */
+    Tcl_Interp *interp,			/* Current interpreter. */
+    int argc,				/* Number of arguments. */
+    const char **argv)			/* Argument strings. */
+{
+    TestAsyncHandler *asyncPtr, *prevPtr;
+    int id, code;
+    static int nextId = 1;
+
+    if (argc < 2) {
+	wrongNumArgs:
+	Tcl_AppendResult(interp, "wrong # args", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	asyncPtr = ckalloc(sizeof(TestAsyncHandler));
+	asyncPtr->command = ckalloc(strlen(argv[2]) + 1);
+	strcpy(asyncPtr->command, argv[2]);
+        Tcl_MutexLock(&asyncTestMutex);
+	asyncPtr->id = nextId;
+	nextId++;
+	asyncPtr->handler = Tcl_AsyncCreate(AsyncHandlerProc,
+                                            INT2PTR(asyncPtr->id));
+	asyncPtr->nextPtr = firstHandler;
+	firstHandler = asyncPtr;
+        Tcl_MutexUnlock(&asyncTestMutex);
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(asyncPtr->id));
+    } else if (strcmp(argv[1], "delete") == 0) {
+	if (argc == 2) {
+            Tcl_MutexLock(&asyncTestMutex);
+	    while (firstHandler != NULL) {
+		asyncPtr = firstHandler;
+		firstHandler = asyncPtr->nextPtr;
+		Tcl_AsyncDelete(asyncPtr->handler);
+		ckfree(asyncPtr->command);
+		ckfree(asyncPtr);
+	    }
+            Tcl_MutexUnlock(&asyncTestMutex);
+	    return TCL_OK;
+	}
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_GetInt(interp, argv[2], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+        Tcl_MutexLock(&asyncTestMutex);
+	for (prevPtr = NULL, asyncPtr = firstHandler; asyncPtr != NULL;
+		prevPtr = asyncPtr, asyncPtr = asyncPtr->nextPtr) {
+	    if (asyncPtr->id != id) {
+		continue;
+	    }
+	    if (prevPtr == NULL) {
+		firstHandler = asyncPtr->nextPtr;
+	    } else {
+		prevPtr->nextPtr = asyncPtr->nextPtr;
+	    }
+	    Tcl_AsyncDelete(asyncPtr->handler);
+	    ckfree(asyncPtr->command);
+	    ckfree(asyncPtr);
+	    break;
+	}
+        Tcl_MutexUnlock(&asyncTestMutex);
+    } else if (strcmp(argv[1], "mark") == 0) {
+	if (argc != 5) {
+	    goto wrongNumArgs;
+	}
+	if ((Tcl_GetInt(interp, argv[2], &id) != TCL_OK)
+		|| (Tcl_GetInt(interp, argv[4], &code) != TCL_OK)) {
+	    return TCL_ERROR;
+	}
+	Tcl_MutexLock(&asyncTestMutex);
+	for (asyncPtr = firstHandler; asyncPtr != NULL;
+		asyncPtr = asyncPtr->nextPtr) {
+	    if (asyncPtr->id == id) {
+		Tcl_AsyncMark(asyncPtr->handler);
+		break;
+	    }
+	}
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(argv[3], -1));
+	Tcl_MutexUnlock(&asyncTestMutex);
+	return code;
+#ifdef TCL_THREADS
+    } else if (strcmp(argv[1], "marklater") == 0) {
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_GetInt(interp, argv[2], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+        Tcl_MutexLock(&asyncTestMutex);
+	for (asyncPtr = firstHandler; asyncPtr != NULL;
+		asyncPtr = asyncPtr->nextPtr) {
+	    if (asyncPtr->id == id) {
+		Tcl_ThreadId threadID;
+		if (Tcl_CreateThread(&threadID, AsyncThreadProc,
+			INT2PTR(id), TCL_THREAD_STACK_DEFAULT,
+			TCL_THREAD_NOFLAGS) != TCL_OK) {
+		    Tcl_AppendResult(interp, "can't create thread", NULL);
+		    Tcl_MutexUnlock(&asyncTestMutex);
+		    return TCL_ERROR;
+		}
+		break;
+	    }
+	}
+        Tcl_MutexUnlock(&asyncTestMutex);
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create, delete, int, mark, or marklater", NULL);
+	return TCL_ERROR;
+#else /* !TCL_THREADS */
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create, delete, int, or mark", NULL);
+	return TCL_ERROR;
+#endif
+    }
+    return TCL_OK;
+}
+
+static int
+AsyncHandlerProc(
+    ClientData clientData,	/* If of TestAsyncHandler structure.
+                                 * in global list. */
+    Tcl_Interp *interp,		/* Interpreter in which command was
+				 * executed, or NULL. */
+    int code)			/* Current return code from command. */
+{
+    TestAsyncHandler *asyncPtr;
+    int id = PTR2INT(clientData);
+    const char *listArgv[4], *cmd;
+    char string[TCL_INTEGER_SPACE];
+
+    Tcl_MutexLock(&asyncTestMutex);
+    for (asyncPtr = firstHandler; asyncPtr != NULL;
+         asyncPtr = asyncPtr->nextPtr) {
+        if (asyncPtr->id == id) break;
+    }
+    Tcl_MutexUnlock(&asyncTestMutex);
+
+    if (!asyncPtr) {
+        /* Woops - this one was deleted between the AsyncMark and now */
+        return TCL_OK;
+    }
+
+    TclFormatInt(string, code);
+    listArgv[0] = asyncPtr->command;
+    listArgv[1] = Tcl_GetString(Tcl_GetObjResult(interp));
+    listArgv[2] = string;
+    listArgv[3] = NULL;
+    cmd = Tcl_Merge(3, listArgv);
+    if (interp != NULL) {
+	code = Tcl_EvalEx(interp, cmd, -1, 0);
+    } else {
+	/*
+	 * this should not happen, but by definition of how async handlers are
+	 * invoked, it's possible.  Better error checking is needed here.
+	 */
+    }
+    ckfree(cmd);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AsyncThreadProc --
+ *
+ *	Delivers an asynchronous event to a handler in another thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Invokes Tcl_AsyncMark on the handler
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifdef TCL_THREADS
+static Tcl_ThreadCreateType
+AsyncThreadProc(
+    ClientData clientData)	/* Parameter is the id of a
+				 * TestAsyncHandler, defined above. */
+{
+    TestAsyncHandler *asyncPtr;
+    int id = PTR2INT(clientData);
+
+    Tcl_Sleep(1);
+    Tcl_MutexLock(&asyncTestMutex);
+    for (asyncPtr = firstHandler; asyncPtr != NULL;
+         asyncPtr = asyncPtr->nextPtr) {
+        if (asyncPtr->id == id) {
+            Tcl_AsyncMark(asyncPtr->handler);
+            break;
+        }
+    }
+    Tcl_MutexUnlock(&asyncTestMutex);
+    Tcl_ExitThread(TCL_OK);
+    TCL_THREAD_CREATE_RETURN;
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestcmdinfoCmd --
+ *
+ *	This procedure implements the "testcmdinfo" command.  It is used to
+ *	test Tcl_GetCommandInfo, Tcl_SetCommandInfo, and command creation and
+ *	deletion.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes various commands and modifies their data.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestcmdinfoCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_CmdInfo info;
+
+    if (argc != 3) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option cmdName\"", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	Tcl_CreateCommand(interp, argv[2], CmdProc1, (ClientData) "original",
+		CmdDelProc1);
+    } else if (strcmp(argv[1], "delete") == 0) {
+	Tcl_DStringInit(&delString);
+	Tcl_DeleteCommand(interp, argv[2]);
+	Tcl_DStringResult(interp, &delString);
+    } else if (strcmp(argv[1], "get") == 0) {
+	if (Tcl_GetCommandInfo(interp, argv[2], &info) ==0) {
+	    Tcl_AppendResult(interp, "??", NULL);
+	    return TCL_OK;
+	}
+	if (info.proc == CmdProc1) {
+	    Tcl_AppendResult(interp, "CmdProc1", " ",
+		    (char *) info.clientData, NULL);
+	} else if (info.proc == CmdProc2) {
+	    Tcl_AppendResult(interp, "CmdProc2", " ",
+		    (char *) info.clientData, NULL);
+	} else {
+	    Tcl_AppendResult(interp, "unknown", NULL);
+	}
+	if (info.deleteProc == CmdDelProc1) {
+	    Tcl_AppendResult(interp, " CmdDelProc1", " ",
+		    (char *) info.deleteData, NULL);
+	} else if (info.deleteProc == CmdDelProc2) {
+	    Tcl_AppendResult(interp, " CmdDelProc2", " ",
+		    (char *) info.deleteData, NULL);
+	} else {
+	    Tcl_AppendResult(interp, " unknown", NULL);
+	}
+	Tcl_AppendResult(interp, " ", info.namespacePtr->fullName, NULL);
+	if (info.isNativeObjectProc) {
+	    Tcl_AppendResult(interp, " nativeObjectProc", NULL);
+	} else {
+	    Tcl_AppendResult(interp, " stringProc", NULL);
+	}
+    } else if (strcmp(argv[1], "modify") == 0) {
+	info.proc = CmdProc2;
+	info.clientData = (ClientData) "new_command_data";
+	info.objProc = NULL;
+	info.objClientData = NULL;
+	info.deleteProc = CmdDelProc2;
+	info.deleteData = (ClientData) "new_delete_data";
+	if (Tcl_SetCommandInfo(interp, argv[2], &info) == 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(0));
+	} else {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(1));
+	}
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create, delete, get, or modify", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+	/*ARGSUSED*/
+static int
+CmdProc1(
+    ClientData clientData,	/* String to return. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_AppendResult(interp, "CmdProc1 ", (char *) clientData, NULL);
+    return TCL_OK;
+}
+
+	/*ARGSUSED*/
+static int
+CmdProc2(
+    ClientData clientData,	/* String to return. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_AppendResult(interp, "CmdProc2 ", (char *) clientData, NULL);
+    return TCL_OK;
+}
+
+static void
+CmdDelProc1(
+    ClientData clientData)	/* String to save. */
+{
+    Tcl_DStringInit(&delString);
+    Tcl_DStringAppend(&delString, "CmdDelProc1 ", -1);
+    Tcl_DStringAppend(&delString, (char *) clientData, -1);
+}
+
+static void
+CmdDelProc2(
+    ClientData clientData)	/* String to save. */
+{
+    Tcl_DStringInit(&delString);
+    Tcl_DStringAppend(&delString, "CmdDelProc2 ", -1);
+    Tcl_DStringAppend(&delString, (char *) clientData, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestcmdtokenCmd --
+ *
+ *	This procedure implements the "testcmdtoken" command. It is used to
+ *	test Tcl_Command tokens and procedures such as Tcl_GetCommandFullName.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes various commands and modifies their data.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestcmdtokenCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_Command token;
+    int *l;
+    char buf[30];
+
+    if (argc != 3) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option arg\"", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	token = Tcl_CreateCommand(interp, argv[2], CmdProc1,
+		(ClientData) "original", NULL);
+	sprintf(buf, "%p", (void *)token);
+	Tcl_AppendResult(interp, buf, NULL);
+    } else if (strcmp(argv[1], "name") == 0) {
+	Tcl_Obj *objPtr;
+
+	if (sscanf(argv[2], "%p", &l) != 1) {
+	    Tcl_AppendResult(interp, "bad command token \"", argv[2],
+		    "\"", NULL);
+	    return TCL_ERROR;
+	}
+
+	objPtr = Tcl_NewObj();
+	Tcl_GetCommandFullName(interp, (Tcl_Command) l, objPtr);
+
+	Tcl_AppendElement(interp,
+		Tcl_GetCommandName(interp, (Tcl_Command) l));
+	Tcl_AppendElement(interp, Tcl_GetString(objPtr));
+	Tcl_DecrRefCount(objPtr);
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create or name", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestcmdtraceCmd --
+ *
+ *	This procedure implements the "testcmdtrace" command. It is used
+ *	to test Tcl_CreateTrace and Tcl_DeleteTrace.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes a command trace, and tests the invocation of
+ *	a procedure by the command trace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestcmdtraceCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_DString buffer;
+    int result;
+
+    if (argc != 3) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option script\"", NULL);
+	return TCL_ERROR;
+    }
+
+    if (strcmp(argv[1], "tracetest") == 0) {
+	Tcl_DStringInit(&buffer);
+	cmdTrace = Tcl_CreateTrace(interp, 50000, CmdTraceProc, &buffer);
+	result = Tcl_EvalEx(interp, argv[2], -1, 0);
+	if (result == TCL_OK) {
+	    Tcl_ResetResult(interp);
+	    Tcl_AppendResult(interp, Tcl_DStringValue(&buffer), NULL);
+	}
+	Tcl_DeleteTrace(interp, cmdTrace);
+	Tcl_DStringFree(&buffer);
+    } else if (strcmp(argv[1], "deletetest") == 0) {
+	/*
+	 * Create a command trace then eval a script to check whether it is
+	 * called. Note that this trace procedure removes itself as a further
+	 * check of the robustness of the trace proc calling code in
+	 * TclNRExecuteByteCode.
+	 */
+
+	cmdTrace = Tcl_CreateTrace(interp, 50000, CmdTraceDeleteProc, NULL);
+	Tcl_EvalEx(interp, argv[2], -1, 0);
+    } else if (strcmp(argv[1], "leveltest") == 0) {
+	Interp *iPtr = (Interp *) interp;
+	Tcl_DStringInit(&buffer);
+	cmdTrace = Tcl_CreateTrace(interp, iPtr->numLevels + 4, CmdTraceProc,
+		&buffer);
+	result = Tcl_EvalEx(interp, argv[2], -1, 0);
+	if (result == TCL_OK) {
+	    Tcl_ResetResult(interp);
+	    Tcl_AppendResult(interp, Tcl_DStringValue(&buffer), NULL);
+	}
+	Tcl_DeleteTrace(interp, cmdTrace);
+	Tcl_DStringFree(&buffer);
+    } else if (strcmp(argv[1], "resulttest") == 0) {
+	/* Create an object-based trace, then eval a script. This is used
+	 * to test return codes other than TCL_OK from the trace engine.
+	 */
+
+	static int deleteCalled;
+
+	deleteCalled = 0;
+	cmdTrace = Tcl_CreateObjTrace(interp, 50000,
+		TCL_ALLOW_INLINE_COMPILATION, ObjTraceProc,
+		(ClientData) &deleteCalled, ObjTraceDeleteProc);
+	result = Tcl_EvalEx(interp, argv[2], -1, 0);
+	Tcl_DeleteTrace(interp, cmdTrace);
+	if (!deleteCalled) {
+	    Tcl_AppendResult(interp, "Delete wasn't called", NULL);
+	    return TCL_ERROR;
+	} else {
+	    return result;
+	}
+    } else if (strcmp(argv[1], "doubletest") == 0) {
+	Tcl_Trace t1, t2;
+
+	Tcl_DStringInit(&buffer);
+	t1 = Tcl_CreateTrace(interp, 1, CmdTraceProc, &buffer);
+	t2 = Tcl_CreateTrace(interp, 50000, CmdTraceProc, &buffer);
+	result = Tcl_EvalEx(interp, argv[2], -1, 0);
+	if (result == TCL_OK) {
+	    Tcl_ResetResult(interp);
+	    Tcl_AppendResult(interp, Tcl_DStringValue(&buffer), NULL);
+	}
+	Tcl_DeleteTrace(interp, t2);
+	Tcl_DeleteTrace(interp, t1);
+	Tcl_DStringFree(&buffer);
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be tracetest, deletetest, doubletest or resulttest", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+static void
+CmdTraceProc(
+    ClientData clientData,	/* Pointer to buffer in which the
+				 * command and arguments are appended.
+				 * Accumulates test result. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int level,			/* Current trace level. */
+    char *command,		/* The command being traced (after
+				 * substitutions). */
+    Tcl_CmdProc *cmdProc,	/* Points to command's command procedure. */
+    ClientData cmdClientData,	/* Client data associated with command
+				 * procedure. */
+    int argc,			/* Number of arguments. */
+    const char *argv[])		/* Argument strings. */
+{
+    Tcl_DString *bufPtr = (Tcl_DString *) clientData;
+    int i;
+
+    Tcl_DStringAppendElement(bufPtr, command);
+
+    Tcl_DStringStartSublist(bufPtr);
+    for (i = 0;  i < argc;  i++) {
+	Tcl_DStringAppendElement(bufPtr, argv[i]);
+    }
+    Tcl_DStringEndSublist(bufPtr);
+}
+
+static void
+CmdTraceDeleteProc(
+    ClientData clientData,	/* Unused. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int level,			/* Current trace level. */
+    char *command,		/* The command being traced (after
+				 * substitutions). */
+    Tcl_CmdProc *cmdProc,	/* Points to command's command procedure. */
+    ClientData cmdClientData,	/* Client data associated with command
+				 * procedure. */
+    int argc,			/* Number of arguments. */
+    const char *argv[])		/* Argument strings. */
+{
+    /*
+     * Remove ourselves to test whether calling Tcl_DeleteTrace within a trace
+     * callback causes the for loop in TclNRExecuteByteCode that calls traces to
+     * reference freed memory.
+     */
+
+    Tcl_DeleteTrace(interp, cmdTrace);
+}
+
+static int
+ObjTraceProc(
+    ClientData clientData,	/* unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int level,			/* Execution level */
+    const char *command,	/* Command being executed */
+    Tcl_Command token,		/* Command information */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const objv[])	/* Parameter list */
+{
+    const char *word = Tcl_GetString(objv[0]);
+
+    if (!strcmp(word, "Error")) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(command, -1));
+	return TCL_ERROR;
+    } else if (!strcmp(word, "Break")) {
+	return TCL_BREAK;
+    } else if (!strcmp(word, "Continue")) {
+	return TCL_CONTINUE;
+    } else if (!strcmp(word, "Return")) {
+	return TCL_RETURN;
+    } else if (!strcmp(word, "OtherStatus")) {
+	return 6;
+    } else {
+	return TCL_OK;
+    }
+}
+
+static void
+ObjTraceDeleteProc(
+    ClientData clientData)
+{
+    int *intPtr = (int *) clientData;
+    *intPtr = 1;		/* Record that the trace was deleted */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestcreatecommandCmd --
+ *
+ *	This procedure implements the "testcreatecommand" command. It is used
+ *	to test that the Tcl_CreateCommand creates a new command in the
+ *	namespace specified as part of its name, if any. It also checks that
+ *	the namespace code ignore single ":"s in the middle or end of a
+ *	command name.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes two commands ("test_ns_basic::createdcommand"
+ *	and "value:at:").
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestcreatecommandCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option\"", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	Tcl_CreateCommand(interp, "test_ns_basic::createdcommand",
+		CreatedCommandProc, NULL, NULL);
+    } else if (strcmp(argv[1], "delete") == 0) {
+	Tcl_DeleteCommand(interp, "test_ns_basic::createdcommand");
+    } else if (strcmp(argv[1], "create2") == 0) {
+	Tcl_CreateCommand(interp, "value:at:",
+		CreatedCommandProc2, NULL, NULL);
+    } else if (strcmp(argv[1], "delete2") == 0) {
+	Tcl_DeleteCommand(interp, "value:at:");
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create, delete, create2, or delete2", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+static int
+CreatedCommandProc(
+    ClientData clientData,	/* String to return. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_CmdInfo info;
+    int found;
+
+    found = Tcl_GetCommandInfo(interp, "test_ns_basic::createdcommand",
+	    &info);
+    if (!found) {
+	Tcl_AppendResult(interp, "CreatedCommandProc could not get command info for test_ns_basic::createdcommand",
+		NULL);
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "CreatedCommandProc in ",
+	    info.namespacePtr->fullName, NULL);
+    return TCL_OK;
+}
+
+static int
+CreatedCommandProc2(
+    ClientData clientData,	/* String to return. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_CmdInfo info;
+    int found;
+
+    found = Tcl_GetCommandInfo(interp, "value:at:", &info);
+    if (!found) {
+	Tcl_AppendResult(interp, "CreatedCommandProc2 could not get command info for test_ns_basic::createdcommand",
+		NULL);
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "CreatedCommandProc2 in ",
+	    info.namespacePtr->fullName, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestdcallCmd --
+ *
+ *	This procedure implements the "testdcall" command.  It is used
+ *	to test Tcl_CallWhenDeleted.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes interpreters.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestdcallCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int i, id;
+
+    delInterp = Tcl_CreateInterp();
+    Tcl_DStringInit(&delString);
+    for (i = 1; i < argc; i++) {
+	if (Tcl_GetInt(interp, argv[i], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (id < 0) {
+	    Tcl_DontCallWhenDeleted(delInterp, DelCallbackProc,
+		    (ClientData) INT2PTR(-id));
+	} else {
+	    Tcl_CallWhenDeleted(delInterp, DelCallbackProc,
+		    (ClientData) INT2PTR(id));
+	}
+    }
+    Tcl_DeleteInterp(delInterp);
+    Tcl_DStringResult(interp, &delString);
+    return TCL_OK;
+}
+
+/*
+ * The deletion callback used by TestdcallCmd:
+ */
+
+static void
+DelCallbackProc(
+    ClientData clientData,	/* Numerical value to append to delString. */
+    Tcl_Interp *interp)		/* Interpreter being deleted. */
+{
+    int id = PTR2INT(clientData);
+    char buffer[TCL_INTEGER_SPACE];
+
+    TclFormatInt(buffer, id);
+    Tcl_DStringAppendElement(&delString, buffer);
+    if (interp != delInterp) {
+	Tcl_DStringAppendElement(&delString, "bogus interpreter argument!");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestdelCmd --
+ *
+ *	This procedure implements the "testdel" command.  It is used
+ *	to test calling of command deletion callbacks.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates a command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestdelCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    DelCmd *dPtr;
+    Tcl_Interp *slave;
+
+    if (argc != 4) {
+	Tcl_AppendResult(interp, "wrong # args", NULL);
+	return TCL_ERROR;
+    }
+
+    slave = Tcl_GetSlave(interp, argv[1]);
+    if (slave == NULL) {
+	return TCL_ERROR;
+    }
+
+    dPtr = ckalloc(sizeof(DelCmd));
+    dPtr->interp = interp;
+    dPtr->deleteCmd = ckalloc(strlen(argv[3]) + 1);
+    strcpy(dPtr->deleteCmd, argv[3]);
+
+    Tcl_CreateCommand(slave, argv[2], DelCmdProc, (ClientData) dPtr,
+	    DelDeleteProc);
+    return TCL_OK;
+}
+
+static int
+DelCmdProc(
+    ClientData clientData,	/* String result to return. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    DelCmd *dPtr = (DelCmd *) clientData;
+
+    Tcl_AppendResult(interp, dPtr->deleteCmd, NULL);
+    ckfree(dPtr->deleteCmd);
+    ckfree(dPtr);
+    return TCL_OK;
+}
+
+static void
+DelDeleteProc(
+    ClientData clientData)	/* String command to evaluate. */
+{
+    DelCmd *dPtr = clientData;
+
+    Tcl_EvalEx(dPtr->interp, dPtr->deleteCmd, -1, 0);
+    Tcl_ResetResult(dPtr->interp);
+    ckfree(dPtr->deleteCmd);
+    ckfree(dPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestdelassocdataCmd --
+ *
+ *	This procedure implements the "testdelassocdata" command. It is used
+ *	to test Tcl_DeleteAssocData.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Deletes an association between a key and associated data from an
+ *	interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestdelassocdataCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" data_key\"", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_DeleteAssocData(interp, argv[1]);
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TestdoubledigitsCmd --
+ *
+ *	This procedure implements the 'testdoubledigits' command. It is
+ *	used to test the low-level floating-point formatting primitives
+ *	in Tcl.
+ *
+ * Usage:
+ *	testdoubledigits fpval ndigits type ?shorten"
+ *
+ * Parameters:
+ *	fpval - Floating-point value to format.
+ *	ndigits - Digit count to request from Tcl_DoubleDigits
+ *	type - One of 'shortest', 'Steele', 'e', 'f'
+ *	shorten - Indicates that the 'shorten' flag should be passed in.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+TestdoubledigitsObjCmd(ClientData unused,
+				/* NULL */
+		       Tcl_Interp* interp,
+				/* Tcl interpreter */
+		       int objc,
+				/* Parameter count */
+		       Tcl_Obj* const objv[])
+				/* Parameter vector */
+{
+    static const char* options[] = {
+	"shortest",
+	"Steele",
+	"e",
+	"f",
+	NULL
+    };
+    static const int types[] = {
+	TCL_DD_SHORTEST,
+	TCL_DD_STEELE,
+	TCL_DD_E_FORMAT,
+	TCL_DD_F_FORMAT
+    };
+
+    const Tcl_ObjType* doubleType;
+    double d;
+    int status;
+    int ndigits;
+    int type;
+    int decpt;
+    int signum;
+    char* str;
+    char* endPtr;
+    Tcl_Obj* strObj;
+    Tcl_Obj* retval;
+
+    if (objc < 4 || objc > 5) {
+	Tcl_WrongNumArgs(interp, 1, objv, "fpval ndigits type ?shorten?");
+	return TCL_ERROR;
+    }
+    status = Tcl_GetDoubleFromObj(interp, objv[1], &d);
+    if (status != TCL_OK) {
+	doubleType = Tcl_GetObjType("double");
+	if (objv[1]->typePtr == doubleType
+	    || TclIsNaN(objv[1]->internalRep.doubleValue)) {
+	    status = TCL_OK;
+	    memcpy(&d, &(objv[1]->internalRep.doubleValue), sizeof(double));
+	}
+    }
+    if (status != TCL_OK
+	|| Tcl_GetIntFromObj(interp, objv[2], &ndigits) != TCL_OK
+	|| Tcl_GetIndexFromObj(interp, objv[3], options, "conversion type",
+			       TCL_EXACT, &type) != TCL_OK) {
+	fprintf(stderr, "bad value? %g\n", d);
+	return TCL_ERROR;
+    }
+    type = types[type];
+    if (objc > 4) {
+	if (strcmp(Tcl_GetString(objv[4]), "shorten")) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("bad flag", -1));
+	    return TCL_ERROR;
+	}
+	type |= TCL_DD_SHORTEN_FLAG;
+    }
+    str = TclDoubleDigits(d, ndigits, type, &decpt, &signum, &endPtr);
+    strObj = Tcl_NewStringObj(str, endPtr-str);
+    ckfree(str);
+    retval = Tcl_NewListObj(1, &strObj);
+    Tcl_ListObjAppendElement(NULL, retval, Tcl_NewIntObj(decpt));
+    strObj = Tcl_NewStringObj(signum ? "-" : "+", 1);
+    Tcl_ListObjAppendElement(NULL, retval, strObj);
+    Tcl_SetObjResult(interp, retval);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestdstringCmd --
+ *
+ *	This procedure implements the "testdstring" command.  It is used
+ *	to test the dynamic string facilities of Tcl.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates, deletes, and invokes handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestdstringCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int count;
+
+    if (argc < 2) {
+	wrongNumArgs:
+	Tcl_AppendResult(interp, "wrong # args", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "append") == 0) {
+	if (argc != 4) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_GetInt(interp, argv[3], &count) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_DStringAppend(&dstring, argv[2], count);
+    } else if (strcmp(argv[1], "element") == 0) {
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	Tcl_DStringAppendElement(&dstring, argv[2]);
+    } else if (strcmp(argv[1], "end") == 0) {
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_DStringEndSublist(&dstring);
+    } else if (strcmp(argv[1], "free") == 0) {
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_DStringFree(&dstring);
+    } else if (strcmp(argv[1], "get") == 0) {
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_SetResult(interp, Tcl_DStringValue(&dstring), TCL_VOLATILE);
+    } else if (strcmp(argv[1], "gresult") == 0) {
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (strcmp(argv[2], "staticsmall") == 0) {
+	    Tcl_AppendResult(interp, "short", NULL);
+	} else if (strcmp(argv[2], "staticlarge") == 0) {
+	    Tcl_AppendResult(interp, "first0 first1 first2 first3 first4 first5 first6 first7 first8 first9\nsecond0 second1 second2 second3 second4 second5 second6 second7 second8 second9\nthird0 third1 third2 third3 third4 third5 third6 third7 third8 third9\nfourth0 fourth1 fourth2 fourth3 fourth4 fourth5 fourth6 fourth7 fourth8 fourth9\nfifth0 fifth1 fifth2 fifth3 fifth4 fifth5 fifth6 fifth7 fifth8 fifth9\nsixth0 sixth1 sixth2 sixth3 sixth4 sixth5 sixth6 sixth7 sixth8 sixth9\nseventh0 seventh1 seventh2 seventh3 seventh4 seventh5 seventh6 seventh7 seventh8 seventh9\n", NULL);
+	} else if (strcmp(argv[2], "free") == 0) {
+	    char *s = ckalloc(100);
+	    strcpy(s, "This is a malloc-ed string");
+	    Tcl_SetResult(interp, s, TCL_DYNAMIC);
+	} else if (strcmp(argv[2], "special") == 0) {
+	    char *s = (char*)ckalloc(100) + 16;
+	    strcpy(s, "This is a specially-allocated string");
+	    Tcl_SetResult(interp, s, SpecialFree);
+	} else {
+	    Tcl_AppendResult(interp, "bad gresult option \"", argv[2],
+		    "\": must be staticsmall, staticlarge, free, or special",
+		    NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_DStringGetResult(interp, &dstring);
+    } else if (strcmp(argv[1], "length") == 0) {
+
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(Tcl_DStringLength(&dstring)));
+    } else if (strcmp(argv[1], "result") == 0) {
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_DStringResult(interp, &dstring);
+    } else if (strcmp(argv[1], "trunc") == 0) {
+	if (argc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_GetInt(interp, argv[2], &count) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_DStringSetLength(&dstring, count);
+    } else if (strcmp(argv[1], "start") == 0) {
+	if (argc != 2) {
+	    goto wrongNumArgs;
+	}
+	Tcl_DStringStartSublist(&dstring);
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be append, element, end, free, get, length, "
+		"result, trunc, or start", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ * The procedure below is used as a special freeProc to test how well
+ * Tcl_DStringGetResult handles freeProc's other than free.
+ */
+
+static void SpecialFree(blockPtr)
+    char *blockPtr;			/* Block to free. */
+{
+    ckfree(blockPtr - 16);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestencodingCmd --
+ *
+ *	This procedure implements the "testencoding" command.  It is used
+ *	to test the encoding package.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Load encodings.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestencodingObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Encoding encoding;
+    int index, length;
+    const char *string;
+    TclEncoding *encodingPtr;
+    static const char *const optionStrings[] = {
+	"create",	"delete",	NULL
+    };
+    enum options {
+	ENC_CREATE,	ENC_DELETE
+    };
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], optionStrings, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum options) index) {
+    case ENC_CREATE: {
+	Tcl_EncodingType type;
+
+	if (objc != 5) {
+	    return TCL_ERROR;
+	}
+	encodingPtr = ckalloc(sizeof(TclEncoding));
+	encodingPtr->interp = interp;
+
+	string = Tcl_GetStringFromObj(objv[3], &length);
+	encodingPtr->toUtfCmd = ckalloc(length + 1);
+	memcpy(encodingPtr->toUtfCmd, string, (unsigned) length + 1);
+
+	string = Tcl_GetStringFromObj(objv[4], &length);
+	encodingPtr->fromUtfCmd = ckalloc(length + 1);
+	memcpy(encodingPtr->fromUtfCmd, string, (unsigned) (length + 1));
+
+	string = Tcl_GetStringFromObj(objv[2], &length);
+
+	type.encodingName = string;
+	type.toUtfProc = EncodingToUtfProc;
+	type.fromUtfProc = EncodingFromUtfProc;
+	type.freeProc = EncodingFreeProc;
+	type.clientData = (ClientData) encodingPtr;
+	type.nullSize = 1;
+
+	Tcl_CreateEncoding(&type);
+	break;
+    }
+    case ENC_DELETE:
+	if (objc != 3) {
+	    return TCL_ERROR;
+	}
+	encoding = Tcl_GetEncoding(NULL, Tcl_GetString(objv[2]));
+	Tcl_FreeEncoding(encoding);
+	Tcl_FreeEncoding(encoding);
+	break;
+    }
+    return TCL_OK;
+}
+
+static int
+EncodingToUtfProc(
+    ClientData clientData,	/* TclEncoding structure. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Current state. */
+    char *dst,			/* Output buffer. */
+    int dstLen,			/* The maximum length of output buffer. */
+    int *srcReadPtr,		/* Filled with number of bytes read. */
+    int *dstWrotePtr,		/* Filled with number of bytes stored. */
+    int *dstCharsPtr)		/* Filled with number of chars stored. */
+{
+    int len;
+    TclEncoding *encodingPtr;
+
+    encodingPtr = (TclEncoding *) clientData;
+    Tcl_EvalEx(encodingPtr->interp, encodingPtr->toUtfCmd, -1, TCL_EVAL_GLOBAL);
+
+    len = strlen(Tcl_GetStringResult(encodingPtr->interp));
+    if (len > dstLen) {
+	len = dstLen;
+    }
+    memcpy(dst, Tcl_GetStringResult(encodingPtr->interp), (unsigned) len);
+    Tcl_ResetResult(encodingPtr->interp);
+
+    *srcReadPtr = srcLen;
+    *dstWrotePtr = len;
+    *dstCharsPtr = len;
+    return TCL_OK;
+}
+
+static int
+EncodingFromUtfProc(
+    ClientData clientData,	/* TclEncoding structure. */
+    const char *src,		/* Source string in specified encoding. */
+    int srcLen,			/* Source string length in bytes. */
+    int flags,			/* Conversion control flags. */
+    Tcl_EncodingState *statePtr,/* Current state. */
+    char *dst,			/* Output buffer. */
+    int dstLen,			/* The maximum length of output buffer. */
+    int *srcReadPtr,		/* Filled with number of bytes read. */
+    int *dstWrotePtr,		/* Filled with number of bytes stored. */
+    int *dstCharsPtr)		/* Filled with number of chars stored. */
+{
+    int len;
+    TclEncoding *encodingPtr;
+
+    encodingPtr = (TclEncoding *) clientData;
+    Tcl_EvalEx(encodingPtr->interp, encodingPtr->fromUtfCmd, -1, TCL_EVAL_GLOBAL);
+
+    len = strlen(Tcl_GetStringResult(encodingPtr->interp));
+    if (len > dstLen) {
+	len = dstLen;
+    }
+    memcpy(dst, Tcl_GetStringResult(encodingPtr->interp), (unsigned) len);
+    Tcl_ResetResult(encodingPtr->interp);
+
+    *srcReadPtr = srcLen;
+    *dstWrotePtr = len;
+    *dstCharsPtr = len;
+    return TCL_OK;
+}
+
+static void
+EncodingFreeProc(
+    ClientData clientData)	/* ClientData associated with type. */
+{
+    TclEncoding *encodingPtr = clientData;
+
+    ckfree(encodingPtr->toUtfCmd);
+    ckfree(encodingPtr->fromUtfCmd);
+    ckfree(encodingPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestevalexObjCmd --
+ *
+ *	This procedure implements the "testevalex" command.  It is
+ *	used to test Tcl_EvalEx.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestevalexObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int length, flags;
+    const char *script;
+
+    flags = 0;
+    if (objc == 3) {
+	const char *global = Tcl_GetStringFromObj(objv[2], &length);
+	if (strcmp(global, "global") != 0) {
+	    Tcl_AppendResult(interp, "bad value \"", global,
+		    "\": must be global", NULL);
+	    return TCL_ERROR;
+	}
+	flags = TCL_EVAL_GLOBAL;
+    } else if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "script ?global?");
+	return TCL_ERROR;
+    }
+
+    script = Tcl_GetStringFromObj(objv[1], &length);
+    return Tcl_EvalEx(interp, script, length, flags);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestevalobjvObjCmd --
+ *
+ *	This procedure implements the "testevalobjv" command.  It is
+ *	used to test Tcl_EvalObjv.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestevalobjvObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int evalGlobal;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "global word ?word ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIntFromObj(interp, objv[1], &evalGlobal) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return Tcl_EvalObjv(interp, objc-2, objv+2,
+	    (evalGlobal) ? TCL_EVAL_GLOBAL : 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TesteventObjCmd --
+ *
+ *	This procedure implements a 'testevent' command.  The command
+ *	is used to test event queue management.
+ *
+ * The command takes two forms:
+ *	- testevent queue name position script
+ *		Queues an event at the given position in the queue, and
+ *		associates a given name with it (the same name may be
+ *		associated with multiple events). When the event comes
+ *		to the head of the queue, executes the given script at
+ *		global level in the current interp. The position may be
+ *		one of 'head', 'tail' or 'mark'.
+ *	- testevent delete name
+ *		Deletes any events associated with the given name from
+ *		the queue.
+ *
+ * Return value:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Manipulates the event queue as directed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TesteventObjCmd(
+    ClientData unused,		/* Not used */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const objv[])	/* Parameter vector */
+{
+    static const char *const subcommands[] = { /* Possible subcommands */
+	"queue", "delete", NULL
+    };
+    int subCmdIndex;		/* Index of the chosen subcommand */
+    static const char *const positions[] = { /* Possible queue positions */
+	"head", "tail", "mark", NULL
+    };
+    int posIndex;		/* Index of the chosen position */
+    static const Tcl_QueuePosition posNum[] = {
+				/* Interpretation of the chosen position */
+	TCL_QUEUE_HEAD,
+	TCL_QUEUE_TAIL,
+	TCL_QUEUE_MARK
+    };
+    TestEvent *ev;		/* Event to be queued */
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "subcommand ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], subcommands, "subcommand",
+	    TCL_EXACT, &subCmdIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch (subCmdIndex) {
+    case 0:			/* queue */
+	if (objc != 5) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "name position script");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[3], positions,
+		"position specifier", TCL_EXACT, &posIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	ev = ckalloc(sizeof(TestEvent));
+	ev->header.proc = TesteventProc;
+	ev->header.nextPtr = NULL;
+	ev->interp = interp;
+	ev->command = objv[4];
+	Tcl_IncrRefCount(ev->command);
+	ev->tag = objv[2];
+	Tcl_IncrRefCount(ev->tag);
+	Tcl_QueueEvent((Tcl_Event *) ev, posNum[posIndex]);
+	break;
+
+    case 1:			/* delete */
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "name");
+	    return TCL_ERROR;
+	}
+	Tcl_DeleteEvents(TesteventDeleteProc, objv[2]);
+	break;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TesteventProc --
+ *
+ *	Delivers a test event to the Tcl interpreter as part of event
+ *	queue testing.
+ *
+ * Results:
+ *	Returns 1 if the event has been serviced, 0 otherwise.
+ *
+ * Side effects:
+ *	Evaluates the event's callback script, so has whatever side effects
+ *	the callback has.  The return value of the callback script becomes the
+ *	return value of this function.  If the callback script reports an
+ *	error, it is reported as a background error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TesteventProc(
+    Tcl_Event *event,		/* Event to deliver */
+    int flags)			/* Current flags for Tcl_ServiceEvent */
+{
+    TestEvent *ev = (TestEvent *) event;
+    Tcl_Interp *interp = ev->interp;
+    Tcl_Obj *command = ev->command;
+    int result = Tcl_EvalObjEx(interp, command,
+	    TCL_EVAL_GLOBAL | TCL_EVAL_DIRECT);
+    int retval;
+
+    if (result != TCL_OK) {
+	Tcl_AddErrorInfo(interp,
+		"    (command bound to \"testevent\" callback)");
+	Tcl_BackgroundError(interp);
+	return 1;		/* Avoid looping on errors */
+    }
+    if (Tcl_GetBooleanFromObj(interp, Tcl_GetObjResult(interp),
+	    &retval) != TCL_OK) {
+	Tcl_AddErrorInfo(interp,
+		"    (return value from \"testevent\" callback)");
+	Tcl_BackgroundError(interp);
+	return 1;
+    }
+    if (retval) {
+	Tcl_DecrRefCount(ev->tag);
+	Tcl_DecrRefCount(ev->command);
+    }
+
+    return retval;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TesteventDeleteProc --
+ *
+ *	Removes some set of events from the queue.
+ *
+ * This procedure is used as part of testing event queue management.
+ *
+ * Results:
+ *	Returns 1 if a given event should be deleted, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TesteventDeleteProc(
+    Tcl_Event *event,		/* Event to examine */
+    ClientData clientData)	/* Tcl_Obj containing the name of the event(s)
+				 * to remove */
+{
+    TestEvent *ev;		/* Event to examine */
+    const char *evNameStr;
+    Tcl_Obj *targetName;	/* Name of the event(s) to delete */
+    const char *targetNameStr;
+
+    if (event->proc != TesteventProc) {
+	return 0;
+    }
+    targetName = (Tcl_Obj *) clientData;
+    targetNameStr = (char *) Tcl_GetString(targetName);
+    ev = (TestEvent *) event;
+    evNameStr = Tcl_GetString(ev->tag);
+    if (strcmp(evNameStr, targetNameStr) == 0) {
+	Tcl_DecrRefCount(ev->tag);
+	Tcl_DecrRefCount(ev->command);
+	return 1;
+    } else {
+	return 0;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexithandlerCmd --
+ *
+ *	This procedure implements the "testexithandler" command. It is
+ *	used to test Tcl_CreateExitHandler and Tcl_DeleteExitHandler.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexithandlerCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int value;
+
+    if (argc != 3) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" create|delete value\"", NULL);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetInt(interp, argv[2], &value) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	Tcl_CreateExitHandler((value & 1) ? ExitProcOdd : ExitProcEven,
+		(ClientData) INT2PTR(value));
+    } else if (strcmp(argv[1], "delete") == 0) {
+	Tcl_DeleteExitHandler((value & 1) ? ExitProcOdd : ExitProcEven,
+		(ClientData) INT2PTR(value));
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": must be create or delete", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+static void
+ExitProcOdd(
+    ClientData clientData)	/* Integer value to print. */
+{
+    char buf[16 + TCL_INTEGER_SPACE];
+    size_t len;
+
+    sprintf(buf, "odd %d\n", PTR2INT(clientData));
+    len = strlen(buf);
+    if (len != (size_t) write(1, buf, len)) {
+	Tcl_Panic("ExitProcOdd: unable to write to stdout");
+    }
+}
+
+static void
+ExitProcEven(
+    ClientData clientData)	/* Integer value to print. */
+{
+    char buf[16 + TCL_INTEGER_SPACE];
+    size_t len;
+
+    sprintf(buf, "even %d\n", PTR2INT(clientData));
+    len = strlen(buf);
+    if (len != (size_t) write(1, buf, len)) {
+	Tcl_Panic("ExitProcEven: unable to write to stdout");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprlongCmd --
+ *
+ *	This procedure verifies that Tcl_ExprLong does not modify the
+ *	interpreter result if there is no error.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprlongCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    long exprResult;
+    char buf[4 + TCL_INTEGER_SPACE];
+    int result;
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" expression\"", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "This is a result", NULL);
+    result = Tcl_ExprLong(interp, argv[1], &exprResult);
+    if (result != TCL_OK) {
+	return result;
+    }
+    sprintf(buf, ": %ld", exprResult);
+    Tcl_AppendResult(interp, buf, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprlongobjCmd --
+ *
+ *	This procedure verifies that Tcl_ExprLongObj does not modify the
+ *	interpreter result if there is no error.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprlongobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* Argument objects. */
+{
+    long exprResult;
+    char buf[4 + TCL_INTEGER_SPACE];
+    int result;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "expression");
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "This is a result", NULL);
+    result = Tcl_ExprLongObj(interp, objv[1], &exprResult);
+    if (result != TCL_OK) {
+	return result;
+    }
+    sprintf(buf, ": %ld", exprResult);
+    Tcl_AppendResult(interp, buf, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprdoubleCmd --
+ *
+ *	This procedure verifies that Tcl_ExprDouble does not modify the
+ *	interpreter result if there is no error.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprdoubleCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    double exprResult;
+    char buf[4 + TCL_DOUBLE_SPACE];
+    int result;
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" expression\"", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "This is a result", NULL);
+    result = Tcl_ExprDouble(interp, argv[1], &exprResult);
+    if (result != TCL_OK) {
+	return result;
+    }
+    strcpy(buf, ": ");
+    Tcl_PrintDouble(interp, exprResult, buf+2);
+    Tcl_AppendResult(interp, buf, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprdoubleobjCmd --
+ *
+ *	This procedure verifies that Tcl_ExprLongObj does not modify the
+ *	interpreter result if there is no error.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprdoubleobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const *objv)	/* Argument objects. */
+{
+    double exprResult;
+    char buf[4 + TCL_DOUBLE_SPACE];
+    int result;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "expression");
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, "This is a result", NULL);
+    result = Tcl_ExprDoubleObj(interp, objv[1], &exprResult);
+    if (result != TCL_OK) {
+	return result;
+    }
+    strcpy(buf, ": ");
+    Tcl_PrintDouble(interp, exprResult, buf+2);
+    Tcl_AppendResult(interp, buf, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprstringCmd --
+ *
+ *	This procedure tests the basic operation of Tcl_ExprString.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprstringCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" expression\"", NULL);
+	return TCL_ERROR;
+    }
+    return Tcl_ExprString(interp, argv[1]);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestfilelinkCmd --
+ *
+ *	This procedure implements the "testfilelink" command.  It is used to
+ *	test the effects of creating and manipulating filesystem links in Tcl.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	May create a link on disk.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestfilelinkCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    Tcl_Obj *contents;
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "source ?target?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_FSConvertToPathType(interp, objv[1]) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	/* Create link from source to target */
+	contents = Tcl_FSLink(objv[1], objv[2],
+		TCL_CREATE_SYMBOLIC_LINK|TCL_CREATE_HARD_LINK);
+	if (contents == NULL) {
+	    Tcl_AppendResult(interp, "could not create link from \"",
+		    Tcl_GetString(objv[1]), "\" to \"",
+		    Tcl_GetString(objv[2]), "\": ",
+		    Tcl_PosixError(interp), NULL);
+	    return TCL_ERROR;
+	}
+    } else {
+	/* Read link */
+	contents = Tcl_FSLink(objv[1], NULL, 0);
+	if (contents == NULL) {
+	    Tcl_AppendResult(interp, "could not read link \"",
+		    Tcl_GetString(objv[1]), "\": ",
+		    Tcl_PosixError(interp), NULL);
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_SetObjResult(interp, contents);
+    if (objc == 2) {
+	/*
+	 * If we are creating a link, this will actually just
+	 * be objv[3], and we don't own it
+	 */
+	Tcl_DecrRefCount(contents);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestgetassocdataCmd --
+ *
+ *	This procedure implements the "testgetassocdata" command. It is
+ *	used to test Tcl_GetAssocData.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestgetassocdataCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    char *res;
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" data_key\"", NULL);
+	return TCL_ERROR;
+    }
+    res = (char *) Tcl_GetAssocData(interp, argv[1], NULL);
+    if (res != NULL) {
+	Tcl_AppendResult(interp, res, NULL);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestgetplatformCmd --
+ *
+ *	This procedure implements the "testgetplatform" command. It is
+ *	used to retrievel the value of the tclPlatform global variable.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestgetplatformCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    static const char *const platformStrings[] = { "unix", "mac", "windows" };
+    TclPlatformType *platform;
+
+    platform = TclGetPlatform();
+
+    if (argc != 1) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_AppendResult(interp, platformStrings[*platform], NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestinterpdeleteCmd --
+ *
+ *	This procedure tests the code in tclInterp.c that deals with
+ *	interpreter deletion. It deletes a user-specified interpreter
+ *	from the hierarchy, and subsequent code checks integrity.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Deletes one or more interpreters.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestinterpdeleteCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_Interp *slaveToDelete;
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" path\"", NULL);
+	return TCL_ERROR;
+    }
+    slaveToDelete = Tcl_GetSlave(interp, argv[1]);
+    if (slaveToDelete == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_DeleteInterp(slaveToDelete);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestlinkCmd --
+ *
+ *	This procedure implements the "testlink" command.  It is used
+ *	to test Tcl_LinkVar and related library procedures.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes various variable links, plus returns
+ *	values of the linked variables.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestlinkCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    static int intVar = 43;
+    static int boolVar = 4;
+    static double realVar = 1.23;
+    static Tcl_WideInt wideVar = Tcl_LongAsWide(79);
+    static char *stringVar = NULL;
+    static char charVar = '@';
+    static unsigned char ucharVar = 130;
+    static short shortVar = 3000;
+    static unsigned short ushortVar = 60000;
+    static unsigned int uintVar = 0xbeeffeed;
+    static long longVar = 123456789L;
+    static unsigned long ulongVar = 3456789012UL;
+    static float floatVar = 4.5;
+    static Tcl_WideUInt uwideVar = (Tcl_WideUInt) Tcl_LongAsWide(123);
+    static int created = 0;
+    char buffer[2*TCL_DOUBLE_SPACE];
+    int writable, flag;
+    Tcl_Obj *tmp;
+
+    if (argc < 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option ?arg arg arg arg arg arg arg arg arg arg arg arg"
+		" arg arg?\"", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "create") == 0) {
+	if (argc != 16) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"",
+		argv[0], " ", argv[1],
+		" intRO realRO boolRO stringRO wideRO charRO ucharRO shortRO"
+		" ushortRO uintRO longRO ulongRO floatRO uwideRO\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (created) {
+	    Tcl_UnlinkVar(interp, "int");
+	    Tcl_UnlinkVar(interp, "real");
+	    Tcl_UnlinkVar(interp, "bool");
+	    Tcl_UnlinkVar(interp, "string");
+	    Tcl_UnlinkVar(interp, "wide");
+	    Tcl_UnlinkVar(interp, "char");
+	    Tcl_UnlinkVar(interp, "uchar");
+	    Tcl_UnlinkVar(interp, "short");
+	    Tcl_UnlinkVar(interp, "ushort");
+	    Tcl_UnlinkVar(interp, "uint");
+	    Tcl_UnlinkVar(interp, "long");
+	    Tcl_UnlinkVar(interp, "ulong");
+	    Tcl_UnlinkVar(interp, "float");
+	    Tcl_UnlinkVar(interp, "uwide");
+	}
+	created = 1;
+	if (Tcl_GetBoolean(interp, argv[2], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "int", (char *) &intVar,
+		TCL_LINK_INT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[3], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "real", (char *) &realVar,
+		TCL_LINK_DOUBLE | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[4], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "bool", (char *) &boolVar,
+		TCL_LINK_BOOLEAN | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[5], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "string", (char *) &stringVar,
+		TCL_LINK_STRING | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[6], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "wide", (char *) &wideVar,
+			TCL_LINK_WIDE_INT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[7], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "char", (char *) &charVar,
+		TCL_LINK_CHAR | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[8], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "uchar", (char *) &ucharVar,
+		TCL_LINK_UCHAR | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[9], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "short", (char *) &shortVar,
+		TCL_LINK_SHORT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[10], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "ushort", (char *) &ushortVar,
+		TCL_LINK_USHORT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[11], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "uint", (char *) &uintVar,
+		TCL_LINK_UINT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[12], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "long", (char *) &longVar,
+		TCL_LINK_LONG | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[13], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "ulong", (char *) &ulongVar,
+		TCL_LINK_ULONG | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[14], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "float", (char *) &floatVar,
+		TCL_LINK_FLOAT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBoolean(interp, argv[15], &writable) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	flag = (writable != 0) ? 0 : TCL_LINK_READ_ONLY;
+	if (Tcl_LinkVar(interp, "uwide", (char *) &uwideVar,
+		TCL_LINK_WIDE_UINT | flag) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+    } else if (strcmp(argv[1], "delete") == 0) {
+	Tcl_UnlinkVar(interp, "int");
+	Tcl_UnlinkVar(interp, "real");
+	Tcl_UnlinkVar(interp, "bool");
+	Tcl_UnlinkVar(interp, "string");
+	Tcl_UnlinkVar(interp, "wide");
+	Tcl_UnlinkVar(interp, "char");
+	Tcl_UnlinkVar(interp, "uchar");
+	Tcl_UnlinkVar(interp, "short");
+	Tcl_UnlinkVar(interp, "ushort");
+	Tcl_UnlinkVar(interp, "uint");
+	Tcl_UnlinkVar(interp, "long");
+	Tcl_UnlinkVar(interp, "ulong");
+	Tcl_UnlinkVar(interp, "float");
+	Tcl_UnlinkVar(interp, "uwide");
+	created = 0;
+    } else if (strcmp(argv[1], "get") == 0) {
+	TclFormatInt(buffer, intVar);
+	Tcl_AppendElement(interp, buffer);
+	Tcl_PrintDouble(NULL, realVar, buffer);
+	Tcl_AppendElement(interp, buffer);
+	TclFormatInt(buffer, boolVar);
+	Tcl_AppendElement(interp, buffer);
+	Tcl_AppendElement(interp, (stringVar == NULL) ? "-" : stringVar);
+	/*
+	 * Wide ints only have an object-based interface.
+	 */
+	tmp = Tcl_NewWideIntObj(wideVar);
+	Tcl_AppendElement(interp, Tcl_GetString(tmp));
+	Tcl_DecrRefCount(tmp);
+	TclFormatInt(buffer, (int) charVar);
+	Tcl_AppendElement(interp, buffer);
+	TclFormatInt(buffer, (int) ucharVar);
+	Tcl_AppendElement(interp, buffer);
+	TclFormatInt(buffer, (int) shortVar);
+	Tcl_AppendElement(interp, buffer);
+	TclFormatInt(buffer, (int) ushortVar);
+	Tcl_AppendElement(interp, buffer);
+	TclFormatInt(buffer, (int) uintVar);
+	Tcl_AppendElement(interp, buffer);
+	tmp = Tcl_NewLongObj(longVar);
+	Tcl_AppendElement(interp, Tcl_GetString(tmp));
+	Tcl_DecrRefCount(tmp);
+	tmp = Tcl_NewLongObj((long)ulongVar);
+	Tcl_AppendElement(interp, Tcl_GetString(tmp));
+	Tcl_DecrRefCount(tmp);
+	Tcl_PrintDouble(NULL, (double)floatVar, buffer);
+	Tcl_AppendElement(interp, buffer);
+	tmp = Tcl_NewWideIntObj((Tcl_WideInt)uwideVar);
+	Tcl_AppendElement(interp, Tcl_GetString(tmp));
+	Tcl_DecrRefCount(tmp);
+    } else if (strcmp(argv[1], "set") == 0) {
+	int v;
+
+	if (argc != 16) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"",
+		    argv[0], " ", argv[1],
+		    " intValue realValue boolValue stringValue wideValue"
+		    " charValue ucharValue shortValue ushortValue uintValue"
+		    " longValue ulongValue floatValue uwideValue\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (argv[2][0] != 0) {
+	    if (Tcl_GetInt(interp, argv[2], &intVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	if (argv[3][0] != 0) {
+	    if (Tcl_GetDouble(interp, argv[3], &realVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	if (argv[4][0] != 0) {
+	    if (Tcl_GetInt(interp, argv[4], &boolVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	if (argv[5][0] != 0) {
+	    if (stringVar != NULL) {
+		ckfree(stringVar);
+	    }
+	    if (strcmp(argv[5], "-") == 0) {
+		stringVar = NULL;
+	    } else {
+		stringVar = ckalloc(strlen(argv[5]) + 1);
+		strcpy(stringVar, argv[5]);
+	    }
+	}
+	if (argv[6][0] != 0) {
+	    tmp = Tcl_NewStringObj(argv[6], -1);
+	    if (Tcl_GetWideIntFromObj(interp, tmp, &wideVar) != TCL_OK) {
+		Tcl_DecrRefCount(tmp);
+		return TCL_ERROR;
+	    }
+	    Tcl_DecrRefCount(tmp);
+	}
+	if (argv[7][0]) {
+	    if (Tcl_GetInt(interp, argv[7], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    charVar = (char) v;
+	}
+	if (argv[8][0]) {
+	    if (Tcl_GetInt(interp, argv[8], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ucharVar = (unsigned char) v;
+	}
+	if (argv[9][0]) {
+	    if (Tcl_GetInt(interp, argv[9], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    shortVar = (short) v;
+	}
+	if (argv[10][0]) {
+	    if (Tcl_GetInt(interp, argv[10], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ushortVar = (unsigned short) v;
+	}
+	if (argv[11][0]) {
+	    if (Tcl_GetInt(interp, argv[11], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    uintVar = (unsigned int) v;
+	}
+	if (argv[12][0]) {
+	    if (Tcl_GetInt(interp, argv[12], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    longVar = (long) v;
+	}
+	if (argv[13][0]) {
+	    if (Tcl_GetInt(interp, argv[13], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ulongVar = (unsigned long) v;
+	}
+	if (argv[14][0]) {
+	    double d;
+	    if (Tcl_GetDouble(interp, argv[14], &d) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    floatVar = (float) d;
+	}
+	if (argv[15][0]) {
+	    Tcl_WideInt w;
+	    tmp = Tcl_NewStringObj(argv[15], -1);
+	    if (Tcl_GetWideIntFromObj(interp, tmp, &w) != TCL_OK) {
+		Tcl_DecrRefCount(tmp);
+		return TCL_ERROR;
+	    }
+	    Tcl_DecrRefCount(tmp);
+	    uwideVar = (Tcl_WideUInt) w;
+	}
+    } else if (strcmp(argv[1], "update") == 0) {
+	int v;
+
+	if (argc != 16) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"",
+		    argv[0], " ", argv[1],
+		    " intValue realValue boolValue stringValue wideValue"
+		    " charValue ucharValue shortValue ushortValue uintValue"
+		    " longValue ulongValue floatValue uwideValue\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (argv[2][0] != 0) {
+	    if (Tcl_GetInt(interp, argv[2], &intVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    Tcl_UpdateLinkedVar(interp, "int");
+	}
+	if (argv[3][0] != 0) {
+	    if (Tcl_GetDouble(interp, argv[3], &realVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    Tcl_UpdateLinkedVar(interp, "real");
+	}
+	if (argv[4][0] != 0) {
+	    if (Tcl_GetInt(interp, argv[4], &boolVar) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    Tcl_UpdateLinkedVar(interp, "bool");
+	}
+	if (argv[5][0] != 0) {
+	    if (stringVar != NULL) {
+		ckfree(stringVar);
+	    }
+	    if (strcmp(argv[5], "-") == 0) {
+		stringVar = NULL;
+	    } else {
+		stringVar = ckalloc(strlen(argv[5]) + 1);
+		strcpy(stringVar, argv[5]);
+	    }
+	    Tcl_UpdateLinkedVar(interp, "string");
+	}
+	if (argv[6][0] != 0) {
+	    tmp = Tcl_NewStringObj(argv[6], -1);
+	    if (Tcl_GetWideIntFromObj(interp, tmp, &wideVar) != TCL_OK) {
+		Tcl_DecrRefCount(tmp);
+		return TCL_ERROR;
+	    }
+	    Tcl_DecrRefCount(tmp);
+	    Tcl_UpdateLinkedVar(interp, "wide");
+	}
+	if (argv[7][0]) {
+	    if (Tcl_GetInt(interp, argv[7], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    charVar = (char) v;
+	    Tcl_UpdateLinkedVar(interp, "char");
+	}
+	if (argv[8][0]) {
+	    if (Tcl_GetInt(interp, argv[8], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ucharVar = (unsigned char) v;
+	    Tcl_UpdateLinkedVar(interp, "uchar");
+	}
+	if (argv[9][0]) {
+	    if (Tcl_GetInt(interp, argv[9], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    shortVar = (short) v;
+	    Tcl_UpdateLinkedVar(interp, "short");
+	}
+	if (argv[10][0]) {
+	    if (Tcl_GetInt(interp, argv[10], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ushortVar = (unsigned short) v;
+	    Tcl_UpdateLinkedVar(interp, "ushort");
+	}
+	if (argv[11][0]) {
+	    if (Tcl_GetInt(interp, argv[11], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    uintVar = (unsigned int) v;
+	    Tcl_UpdateLinkedVar(interp, "uint");
+	}
+	if (argv[12][0]) {
+	    if (Tcl_GetInt(interp, argv[12], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    longVar = (long) v;
+	    Tcl_UpdateLinkedVar(interp, "long");
+	}
+	if (argv[13][0]) {
+	    if (Tcl_GetInt(interp, argv[13], &v) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    ulongVar = (unsigned long) v;
+	    Tcl_UpdateLinkedVar(interp, "ulong");
+	}
+	if (argv[14][0]) {
+	    double d;
+	    if (Tcl_GetDouble(interp, argv[14], &d) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    floatVar = (float) d;
+	    Tcl_UpdateLinkedVar(interp, "float");
+	}
+	if (argv[15][0]) {
+	    Tcl_WideInt w;
+	    tmp = Tcl_NewStringObj(argv[15], -1);
+	    if (Tcl_GetWideIntFromObj(interp, tmp, &w) != TCL_OK) {
+		Tcl_DecrRefCount(tmp);
+		return TCL_ERROR;
+	    }
+	    Tcl_DecrRefCount(tmp);
+	    uwideVar = (Tcl_WideUInt) w;
+	    Tcl_UpdateLinkedVar(interp, "uwide");
+	}
+    } else {
+	Tcl_AppendResult(interp, "bad option \"", argv[1],
+		"\": should be create, delete, get, set, or update", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestlocaleCmd --
+ *
+ *	This procedure implements the "testlocale" command.  It is used
+ *	to test the effects of setting different locales in Tcl.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Modifies the current C locale.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestlocaleCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    int index;
+    const char *locale;
+
+    static const char *const optionStrings[] = {
+	"ctype", "numeric", "time", "collate", "monetary",
+	"all",	NULL
+    };
+    static const int lcTypes[] = {
+	LC_CTYPE, LC_NUMERIC, LC_TIME, LC_COLLATE, LC_MONETARY,
+	LC_ALL
+    };
+
+    /*
+     * LC_CTYPE, etc. correspond to the indices for the strings.
+     */
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "category ?locale?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], optionStrings, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (objc == 3) {
+	locale = Tcl_GetString(objv[2]);
+    } else {
+	locale = NULL;
+    }
+    locale = setlocale(lcTypes[index], locale);
+    if (locale) {
+	Tcl_SetStringObj(Tcl_GetObjResult(interp), locale, -1);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestMathFunc --
+ *
+ *	This is a user-defined math procedure to test out math procedures
+ *	with no arguments.
+ *
+ * Results:
+ *	A normal Tcl completion code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+#ifndef TCL_NO_DEPRECATED
+static int
+TestMathFunc(
+    ClientData clientData,	/* Integer value to return. */
+    Tcl_Interp *interp,		/* Not used. */
+    Tcl_Value *args,		/* Not used. */
+    Tcl_Value *resultPtr)	/* Where to store result. */
+{
+    resultPtr->type = TCL_INT;
+    resultPtr->intValue = PTR2INT(clientData);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestMathFunc2 --
+ *
+ *	This is a user-defined math procedure to test out math procedures
+ *	that do have arguments, in this case 2.
+ *
+ * Results:
+ *	A normal Tcl completion code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestMathFunc2(
+    ClientData clientData,	/* Integer value to return. */
+    Tcl_Interp *interp,		/* Used to report errors. */
+    Tcl_Value *args,		/* Points to an array of two Tcl_Value structs
+				 * for the two arguments. */
+    Tcl_Value *resultPtr)	/* Where to store the result. */
+{
+    int result = TCL_OK;
+
+    /*
+     * Return the maximum of the two arguments with the correct type.
+     */
+
+    if (args[0].type == TCL_INT) {
+	int i0 = args[0].intValue;
+
+	if (args[1].type == TCL_INT) {
+	    int i1 = args[1].intValue;
+
+	    resultPtr->type = TCL_INT;
+	    resultPtr->intValue = ((i0 > i1)? i0 : i1);
+	} else if (args[1].type == TCL_DOUBLE) {
+	    double d0 = i0;
+	    double d1 = args[1].doubleValue;
+
+	    resultPtr->type = TCL_DOUBLE;
+	    resultPtr->doubleValue = ((d0 > d1)? d0 : d1);
+	} else if (args[1].type == TCL_WIDE_INT) {
+	    Tcl_WideInt w0 = Tcl_LongAsWide(i0);
+	    Tcl_WideInt w1 = args[1].wideValue;
+
+	    resultPtr->type = TCL_WIDE_INT;
+	    resultPtr->wideValue = ((w0 > w1)? w0 : w1);
+	} else {
+	    Tcl_AppendResult(interp, "T3: wrong type for arg 2", NULL);
+	    result = TCL_ERROR;
+	}
+    } else if (args[0].type == TCL_DOUBLE) {
+	double d0 = args[0].doubleValue;
+
+	if (args[1].type == TCL_INT) {
+	    double d1 = args[1].intValue;
+
+	    resultPtr->type = TCL_DOUBLE;
+	    resultPtr->doubleValue = ((d0 > d1)? d0 : d1);
+	} else if (args[1].type == TCL_DOUBLE) {
+	    double d1 = args[1].doubleValue;
+
+	    resultPtr->type = TCL_DOUBLE;
+	    resultPtr->doubleValue = ((d0 > d1)? d0 : d1);
+	} else if (args[1].type == TCL_WIDE_INT) {
+	    double d1 = Tcl_WideAsDouble(args[1].wideValue);
+
+	    resultPtr->type = TCL_DOUBLE;
+	    resultPtr->doubleValue = ((d0 > d1)? d0 : d1);
+	} else {
+	    Tcl_AppendResult(interp, "T3: wrong type for arg 2", NULL);
+	    result = TCL_ERROR;
+	}
+    } else if (args[0].type == TCL_WIDE_INT) {
+	Tcl_WideInt w0 = args[0].wideValue;
+
+	if (args[1].type == TCL_INT) {
+	    Tcl_WideInt w1 = Tcl_LongAsWide(args[1].intValue);
+
+	    resultPtr->type = TCL_WIDE_INT;
+	    resultPtr->wideValue = ((w0 > w1)? w0 : w1);
+	} else if (args[1].type == TCL_DOUBLE) {
+	    double d0 = Tcl_WideAsDouble(w0);
+	    double d1 = args[1].doubleValue;
+
+	    resultPtr->type = TCL_DOUBLE;
+	    resultPtr->doubleValue = ((d0 > d1)? d0 : d1);
+	} else if (args[1].type == TCL_WIDE_INT) {
+	    Tcl_WideInt w1 = args[1].wideValue;
+
+	    resultPtr->type = TCL_WIDE_INT;
+	    resultPtr->wideValue = ((w0 > w1)? w0 : w1);
+	} else {
+	    Tcl_AppendResult(interp, "T3: wrong type for arg 2", NULL);
+	    result = TCL_ERROR;
+	}
+    } else {
+	Tcl_AppendResult(interp, "T3: wrong type for arg 1", NULL);
+	result = TCL_ERROR;
+    }
+    return result;
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CleanupTestSetassocdataTests --
+ *
+ *	This function is called when an interpreter is deleted to clean
+ *	up any data left over from running the testsetassocdata command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Releases storage.
+ *
+ *----------------------------------------------------------------------
+ */
+	/* ARGSUSED */
+static void
+CleanupTestSetassocdataTests(
+    ClientData clientData,	/* Data to be released. */
+    Tcl_Interp *interp)		/* Interpreter being deleted. */
+{
+    ckfree(clientData);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestparserObjCmd --
+ *
+ *	This procedure implements the "testparser" command.  It is
+ *	used for testing the new Tcl script parser in Tcl 8.1.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestparserObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    const char *script;
+    int length, dummy;
+    Tcl_Parse parse;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "script length");
+	return TCL_ERROR;
+    }
+    script = Tcl_GetStringFromObj(objv[1], &dummy);
+    if (Tcl_GetIntFromObj(interp, objv[2], &length)) {
+	return TCL_ERROR;
+    }
+    if (length == 0) {
+	length = dummy;
+    }
+    if (Tcl_ParseCommand(interp, script, length, 0, &parse) != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (remainder of script: \"");
+	Tcl_AddErrorInfo(interp, parse.term);
+	Tcl_AddErrorInfo(interp, "\")");
+	return TCL_ERROR;
+    }
+
+    /*
+     * The parse completed successfully.  Just print out the contents
+     * of the parse structure into the interpreter's result.
+     */
+
+    PrintParse(interp, &parse);
+    Tcl_FreeParse(&parse);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexprparserObjCmd --
+ *
+ *	This procedure implements the "testexprparser" command.  It is
+ *	used for testing the new Tcl expression parser in Tcl 8.1.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexprparserObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    const char *script;
+    int length, dummy;
+    Tcl_Parse parse;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "expr length");
+	return TCL_ERROR;
+    }
+    script = Tcl_GetStringFromObj(objv[1], &dummy);
+    if (Tcl_GetIntFromObj(interp, objv[2], &length)) {
+	return TCL_ERROR;
+    }
+    if (length == 0) {
+	length = dummy;
+    }
+    parse.commentStart = NULL;
+    parse.commentSize = 0;
+    parse.commandStart = NULL;
+    parse.commandSize = 0;
+    if (Tcl_ParseExpr(interp, script, length, &parse) != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (remainder of expr: \"");
+	Tcl_AddErrorInfo(interp, parse.term);
+	Tcl_AddErrorInfo(interp, "\")");
+	return TCL_ERROR;
+    }
+
+    /*
+     * The parse completed successfully.  Just print out the contents
+     * of the parse structure into the interpreter's result.
+     */
+
+    PrintParse(interp, &parse);
+    Tcl_FreeParse(&parse);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PrintParse --
+ *
+ *	This procedure prints out the contents of a Tcl_Parse structure
+ *	in the result of an interpreter.
+ *
+ * Results:
+ *	Interp's result is set to a prettily formatted version of the
+ *	contents of parsePtr.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PrintParse(
+    Tcl_Interp *interp,		/* Interpreter whose result is to be set to
+				 * the contents of a parse structure. */
+    Tcl_Parse *parsePtr)	/* Parse structure to print out. */
+{
+    Tcl_Obj *objPtr;
+    const char *typeString;
+    Tcl_Token *tokenPtr;
+    int i;
+
+    objPtr = Tcl_GetObjResult(interp);
+    if (parsePtr->commentSize > 0) {
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewStringObj(parsePtr->commentStart,
+			parsePtr->commentSize));
+    } else {
+	Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewStringObj("-", 1));
+    }
+    Tcl_ListObjAppendElement(NULL, objPtr,
+	    Tcl_NewStringObj(parsePtr->commandStart, parsePtr->commandSize));
+    Tcl_ListObjAppendElement(NULL, objPtr,
+	    Tcl_NewIntObj(parsePtr->numWords));
+    for (i = 0; i < parsePtr->numTokens; i++) {
+	tokenPtr = &parsePtr->tokenPtr[i];
+	switch (tokenPtr->type) {
+	case TCL_TOKEN_EXPAND_WORD:
+	    typeString = "expand";
+	    break;
+	case TCL_TOKEN_WORD:
+	    typeString = "word";
+	    break;
+	case TCL_TOKEN_SIMPLE_WORD:
+	    typeString = "simple";
+	    break;
+	case TCL_TOKEN_TEXT:
+	    typeString = "text";
+	    break;
+	case TCL_TOKEN_BS:
+	    typeString = "backslash";
+	    break;
+	case TCL_TOKEN_COMMAND:
+	    typeString = "command";
+	    break;
+	case TCL_TOKEN_VARIABLE:
+	    typeString = "variable";
+	    break;
+	case TCL_TOKEN_SUB_EXPR:
+	    typeString = "subexpr";
+	    break;
+	case TCL_TOKEN_OPERATOR:
+	    typeString = "operator";
+	    break;
+	default:
+	    typeString = "??";
+	    break;
+	}
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewStringObj(typeString, -1));
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewStringObj(tokenPtr->start, tokenPtr->size));
+	Tcl_ListObjAppendElement(NULL, objPtr,
+		Tcl_NewIntObj(tokenPtr->numComponents));
+    }
+    Tcl_ListObjAppendElement(NULL, objPtr,
+	    Tcl_NewStringObj(parsePtr->commandStart + parsePtr->commandSize,
+	    -1));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestparsevarObjCmd --
+ *
+ *	This procedure implements the "testparsevar" command.  It is
+ *	used for testing Tcl_ParseVar.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestparsevarObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    const char *value, *name, *termPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName");
+	return TCL_ERROR;
+    }
+    name = Tcl_GetString(objv[1]);
+    value = Tcl_ParseVar(interp, name, &termPtr);
+    if (value == NULL) {
+	return TCL_ERROR;
+    }
+
+    Tcl_AppendElement(interp, value);
+    Tcl_AppendElement(interp, termPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestparsevarnameObjCmd --
+ *
+ *	This procedure implements the "testparsevarname" command.  It is
+ *	used for testing the new Tcl script parser in Tcl 8.1.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestparsevarnameObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    const char *script;
+    int append, length, dummy;
+    Tcl_Parse parse;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "script length append");
+	return TCL_ERROR;
+    }
+    script = Tcl_GetStringFromObj(objv[1], &dummy);
+    if (Tcl_GetIntFromObj(interp, objv[2], &length)) {
+	return TCL_ERROR;
+    }
+    if (length == 0) {
+	length = dummy;
+    }
+    if (Tcl_GetIntFromObj(interp, objv[3], &append)) {
+	return TCL_ERROR;
+    }
+    if (Tcl_ParseVarName(interp, script, length, &parse, append) != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (remainder of script: \"");
+	Tcl_AddErrorInfo(interp, parse.term);
+	Tcl_AddErrorInfo(interp, "\")");
+	return TCL_ERROR;
+    }
+
+    /*
+     * The parse completed successfully.  Just print out the contents
+     * of the parse structure into the interpreter's result.
+     */
+
+    parse.commentSize = 0;
+    parse.commandStart = script + parse.tokenPtr->size;
+    parse.commandSize = 0;
+    PrintParse(interp, &parse);
+    Tcl_FreeParse(&parse);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestpreferstableObjCmd --
+ *
+ *	This procedure implements the "testpreferstable" command.  It is
+ *	used for being able to test the "package" command even when the
+ *  environment variable TCL_PKG_PREFER_LATEST is set in your environment.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestpreferstableObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    iPtr->packagePrefer = PKG_PREFER_STABLE;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestprintObjCmd --
+ *
+ *	This procedure implements the "testprint" command.  It is
+ *	used for being able to test the Tcl_ObjPrintf() function.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestprintObjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    Tcl_WideInt argv1 = 0;
+    size_t argv2;
+
+    if (objc < 2 || objc > 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "format wideint");
+    }
+
+    if (objc > 1) {
+	Tcl_GetWideIntFromObj(interp, objv[2], &argv1);
+    }
+    argv2 = (size_t)argv1;
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(Tcl_GetString(objv[1]), argv1, argv2, argv2));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestregexpObjCmd --
+ *
+ *	This procedure implements the "testregexp" command. It is used to give
+ *	a direct interface for regexp flags. It's identical to
+ *	Tcl_RegexpObjCmd except for the -xflags option, and the consequences
+ *	thereof (including the REG_EXPECT kludge).
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestregexpObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int i, ii, indices, stringLength, match, about;
+    int hasxflags, cflags, eflags;
+    Tcl_RegExp regExpr;
+    const char *string;
+    Tcl_Obj *objPtr;
+    Tcl_RegExpInfo info;
+    static const char *const options[] = {
+	"-indices",	"-nocase",	"-about",	"-expanded",
+	"-line",	"-linestop",	"-lineanchor",
+	"-xflags",
+	"--",		NULL
+    };
+    enum options {
+	REGEXP_INDICES, REGEXP_NOCASE,	REGEXP_ABOUT,	REGEXP_EXPANDED,
+	REGEXP_MULTI,	REGEXP_NOCROSS,	REGEXP_NEWL,
+	REGEXP_XFLAGS,
+	REGEXP_LAST
+    };
+
+    indices = 0;
+    about = 0;
+    cflags = REG_ADVANCED;
+    eflags = 0;
+    hasxflags = 0;
+
+    for (i = 1; i < objc; i++) {
+	const char *name;
+	int index;
+
+	name = Tcl_GetString(objv[i]);
+	if (name[0] != '-') {
+	    break;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[i], options, "switch", TCL_EXACT,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	switch ((enum options) index) {
+	case REGEXP_INDICES:
+	    indices = 1;
+	    break;
+	case REGEXP_NOCASE:
+	    cflags |= REG_ICASE;
+	    break;
+	case REGEXP_ABOUT:
+	    about = 1;
+	    break;
+	case REGEXP_EXPANDED:
+	    cflags |= REG_EXPANDED;
+	    break;
+	case REGEXP_MULTI:
+	    cflags |= REG_NEWLINE;
+	    break;
+	case REGEXP_NOCROSS:
+	    cflags |= REG_NLSTOP;
+	    break;
+	case REGEXP_NEWL:
+	    cflags |= REG_NLANCH;
+	    break;
+	case REGEXP_XFLAGS:
+	    hasxflags = 1;
+	    break;
+	case REGEXP_LAST:
+	    i++;
+	    goto endOfForLoop;
+	}
+    }
+
+  endOfForLoop:
+    if (objc - i < hasxflags + 2 - about) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?-switch ...? exp string ?matchVar? ?subMatchVar ...?");
+	return TCL_ERROR;
+    }
+    objc -= i;
+    objv += i;
+
+    if (hasxflags) {
+	string = Tcl_GetStringFromObj(objv[0], &stringLength);
+	TestregexpXflags(string, stringLength, &cflags, &eflags);
+	objc--;
+	objv++;
+    }
+
+    regExpr = Tcl_GetRegExpFromObj(interp, objv[0], cflags);
+    if (regExpr == NULL) {
+	return TCL_ERROR;
+    }
+
+    if (about) {
+	if (TclRegAbout(interp, regExpr) < 0) {
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+
+    objPtr = objv[1];
+    match = Tcl_RegExpExecObj(interp, regExpr, objPtr, 0 /* offset */,
+	    objc-2 /* nmatches */, eflags);
+
+    if (match < 0) {
+	return TCL_ERROR;
+    }
+    if (match == 0) {
+	/*
+	 * Set the interpreter's object result to an integer object w/
+	 * value 0.
+	 */
+
+	Tcl_SetIntObj(Tcl_GetObjResult(interp), 0);
+	if (objc > 2 && (cflags&REG_EXPECT) && indices) {
+	    const char *varName;
+	    const char *value;
+	    int start, end;
+	    char resinfo[TCL_INTEGER_SPACE * 2];
+
+	    varName = Tcl_GetString(objv[2]);
+	    TclRegExpRangeUniChar(regExpr, -1, &start, &end);
+	    sprintf(resinfo, "%d %d", start, end-1);
+	    value = Tcl_SetVar2(interp, varName, NULL, resinfo, 0);
+	    if (value == NULL) {
+		Tcl_AppendResult(interp, "couldn't set variable \"",
+			varName, "\"", NULL);
+		return TCL_ERROR;
+	    }
+	} else if (cflags & TCL_REG_CANMATCH) {
+	    const char *varName;
+	    const char *value;
+	    char resinfo[TCL_INTEGER_SPACE * 2];
+
+	    Tcl_RegExpGetInfo(regExpr, &info);
+	    varName = Tcl_GetString(objv[2]);
+	    sprintf(resinfo, "%ld", info.extendStart);
+	    value = Tcl_SetVar2(interp, varName, NULL, resinfo, 0);
+	    if (value == NULL) {
+		Tcl_AppendResult(interp, "couldn't set variable \"",
+			varName, "\"", NULL);
+		return TCL_ERROR;
+	    }
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * If additional variable names have been specified, return
+     * index information in those variables.
+     */
+
+    objc -= 2;
+    objv += 2;
+
+    Tcl_RegExpGetInfo(regExpr, &info);
+    for (i = 0; i < objc; i++) {
+	int start, end;
+	Tcl_Obj *newPtr, *varPtr, *valuePtr;
+
+	varPtr = objv[i];
+	ii = ((cflags&REG_EXPECT) && i == objc-1) ? -1 : i;
+	if (indices) {
+	    Tcl_Obj *objs[2];
+
+	    if (ii == -1) {
+		TclRegExpRangeUniChar(regExpr, ii, &start, &end);
+	    } else if (ii > info.nsubs) {
+		start = -1;
+		end = -1;
+	    } else {
+		start = info.matches[ii].start;
+		end = info.matches[ii].end;
+	    }
+
+	    /*
+	     * Adjust index so it refers to the last character in the match
+	     * instead of the first character after the match.
+	     */
+
+	    if (end >= 0) {
+		end--;
+	    }
+
+	    objs[0] = Tcl_NewLongObj(start);
+	    objs[1] = Tcl_NewLongObj(end);
+
+	    newPtr = Tcl_NewListObj(2, objs);
+	} else {
+	    if (ii == -1) {
+		TclRegExpRangeUniChar(regExpr, ii, &start, &end);
+		newPtr = Tcl_GetRange(objPtr, start, end);
+	    } else if (ii > info.nsubs) {
+		newPtr = Tcl_NewObj();
+	    } else {
+		newPtr = Tcl_GetRange(objPtr, info.matches[ii].start,
+			info.matches[ii].end - 1);
+	    }
+	}
+	valuePtr = Tcl_ObjSetVar2(interp, varPtr, NULL, newPtr, TCL_LEAVE_ERR_MSG);
+	if (valuePtr == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Set the interpreter's object result to an integer object w/ value 1.
+     */
+
+    Tcl_SetIntObj(Tcl_GetObjResult(interp), 1);
+    return TCL_OK;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TestregexpXflags --
+ *
+ *	Parse a string of extended regexp flag letters, for testing.
+ *
+ * Results:
+ *	No return value (you're on your own for errors here).
+ *
+ * Side effects:
+ *	Modifies *cflagsPtr, a regcomp flags word, and *eflagsPtr, a
+ *	regexec flags word, as appropriate.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TestregexpXflags(
+    const char *string,	/* The string of flags. */
+    int length,			/* The length of the string in bytes. */
+    int *cflagsPtr,		/* compile flags word */
+    int *eflagsPtr)		/* exec flags word */
+{
+    int i, cflags, eflags;
+
+    cflags = *cflagsPtr;
+    eflags = *eflagsPtr;
+    for (i = 0; i < length; i++) {
+	switch (string[i]) {
+	case 'a':
+	    cflags |= REG_ADVF;
+	    break;
+	case 'b':
+	    cflags &= ~REG_ADVANCED;
+	    break;
+	case 'c':
+	    cflags |= TCL_REG_CANMATCH;
+	    break;
+	case 'e':
+	    cflags &= ~REG_ADVANCED;
+	    cflags |= REG_EXTENDED;
+	    break;
+	case 'q':
+	    cflags &= ~REG_ADVANCED;
+	    cflags |= REG_QUOTE;
+	    break;
+	case 'o':			/* o for opaque */
+	    cflags |= REG_NOSUB;
+	    break;
+	case 's':			/* s for start */
+	    cflags |= REG_BOSONLY;
+	    break;
+	case '+':
+	    cflags |= REG_FAKE;
+	    break;
+	case ',':
+	    cflags |= REG_PROGRESS;
+	    break;
+	case '.':
+	    cflags |= REG_DUMP;
+	    break;
+	case ':':
+	    eflags |= REG_MTRACE;
+	    break;
+	case ';':
+	    eflags |= REG_FTRACE;
+	    break;
+	case '^':
+	    eflags |= REG_NOTBOL;
+	    break;
+	case '$':
+	    eflags |= REG_NOTEOL;
+	    break;
+	case 't':
+	    cflags |= REG_EXPECT;
+	    break;
+	case '%':
+	    eflags |= REG_SMALL;
+	    break;
+	}
+    }
+
+    *cflagsPtr = cflags;
+    *eflagsPtr = eflags;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestreturnObjCmd --
+ *
+ *	This procedure implements the "testreturn" command. It is
+ *	used to verify that a
+ *		return TCL_RETURN;
+ *	has same behavior as
+ *		return Tcl_SetReturnOptions(interp, Tcl_NewObj());
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestreturnObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    return TCL_RETURN;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsetassocdataCmd --
+ *
+ *	This procedure implements the "testsetassocdata" command. It is used
+ *	to test Tcl_SetAssocData.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Modifies or creates an association between a key and associated
+ *	data for this interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestsetassocdataCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    char *buf, *oldData;
+    Tcl_InterpDeleteProc *procPtr;
+
+    if (argc != 3) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" data_key data_item\"", NULL);
+	return TCL_ERROR;
+    }
+
+    buf = ckalloc(strlen(argv[2]) + 1);
+    strcpy(buf, argv[2]);
+
+    /*
+     * If we previously associated a malloced value with the variable,
+     * free it before associating a new value.
+     */
+
+    oldData = (char *) Tcl_GetAssocData(interp, argv[1], &procPtr);
+    if ((oldData != NULL) && (procPtr == CleanupTestSetassocdataTests)) {
+	ckfree(oldData);
+    }
+
+    Tcl_SetAssocData(interp, argv[1], CleanupTestSetassocdataTests,
+	(ClientData) buf);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsetplatformCmd --
+ *
+ *	This procedure implements the "testsetplatform" command. It is
+ *	used to change the tclPlatform global variable so all file
+ *	name conversions can be tested on a single platform.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Sets the tclPlatform global variable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestsetplatformCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    size_t length;
+    TclPlatformType *platform;
+
+    platform = TclGetPlatform();
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"", argv[0],
+		" platform\"", NULL);
+	return TCL_ERROR;
+    }
+
+    length = strlen(argv[1]);
+    if (strncmp(argv[1], "unix", length) == 0) {
+	*platform = TCL_PLATFORM_UNIX;
+    } else if (strncmp(argv[1], "windows", length) == 0) {
+	*platform = TCL_PLATFORM_WINDOWS;
+    } else {
+	Tcl_AppendResult(interp, "unsupported platform: should be one of "
+		"unix, or windows", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TeststaticpkgCmd --
+ *
+ *	This procedure implements the "teststaticpkg" command.
+ *	It is used to test the procedure Tcl_StaticPackage.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	When the packge given by argv[1] is loaded into an interpeter,
+ *	variable "x" in that interpreter is set to "loaded".
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TeststaticpkgCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int safe, loaded;
+
+    if (argc != 4) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"",
+		argv[0], " pkgName safe loaded\"", NULL);
+	return TCL_ERROR;
+    }
+    if (Tcl_GetInt(interp, argv[2], &safe) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetInt(interp, argv[3], &loaded) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    tclStubsPtr->tcl_StaticPackage((loaded) ? interp : NULL, argv[1],
+	    StaticInitProc, (safe) ? StaticInitProc : NULL);
+    return TCL_OK;
+}
+
+static int
+StaticInitProc(
+    Tcl_Interp *interp)		/* Interpreter in which package is supposedly
+				 * being loaded. */
+{
+    Tcl_SetVar2(interp, "x", NULL, "loaded", TCL_GLOBAL_ONLY);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TesttranslatefilenameCmd --
+ *
+ *	This procedure implements the "testtranslatefilename" command.
+ *	It is used to test the Tcl_TranslateFileName command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TesttranslatefilenameCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_DString buffer;
+    const char *result;
+
+    if (argc != 2) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"",
+		argv[0], " path\"", NULL);
+	return TCL_ERROR;
+    }
+    result = Tcl_TranslateFileName(interp, argv[1], &buffer);
+    if (result == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_AppendResult(interp, result, NULL);
+    Tcl_DStringFree(&buffer);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestupvarCmd --
+ *
+ *	This procedure implements the "testupvar2" command.  It is used
+ *	to test Tcl_UpVar and Tcl_UpVar2.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates or modifies an "upvar" reference.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestupvarCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int flags = 0;
+
+    if ((argc != 5) && (argc != 6)) {
+	Tcl_AppendResult(interp, "wrong # arguments: should be \"",
+		argv[0], " level name ?name2? dest global\"", NULL);
+	return TCL_ERROR;
+    }
+
+    if (argc == 5) {
+	if (strcmp(argv[4], "global") == 0) {
+	    flags = TCL_GLOBAL_ONLY;
+	} else if (strcmp(argv[4], "namespace") == 0) {
+	    flags = TCL_NAMESPACE_ONLY;
+	}
+	return Tcl_UpVar2(interp, argv[1], argv[2], NULL, argv[3], flags);
+    } else {
+	if (strcmp(argv[5], "global") == 0) {
+	    flags = TCL_GLOBAL_ONLY;
+	} else if (strcmp(argv[5], "namespace") == 0) {
+	    flags = TCL_NAMESPACE_ONLY;
+	}
+	return Tcl_UpVar2(interp, argv[1], argv[2],
+		(argv[3][0] == 0) ? NULL : argv[3], argv[4],
+		flags);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestseterrorcodeCmd --
+ *
+ *	This procedure implements the "testseterrorcodeCmd".  This tests up to
+ *	five elements passed to the Tcl_SetErrorCode command.
+ *
+ * Results:
+ *	A standard Tcl result. Always returns TCL_ERROR so that
+ *	the error code can be tested.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestseterrorcodeCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    if (argc > 6) {
+	Tcl_AppendResult(interp, "too many args", NULL);
+	return TCL_ERROR;
+    }
+    switch (argc) {
+    case 1:
+	Tcl_SetErrorCode(interp, "NONE", NULL);
+	break;
+    case 2:
+	Tcl_SetErrorCode(interp, argv[1], NULL);
+	break;
+    case 3:
+	Tcl_SetErrorCode(interp, argv[1], argv[2], NULL);
+	break;
+    case 4:
+	Tcl_SetErrorCode(interp, argv[1], argv[2], argv[3], NULL);
+	break;
+    case 5:
+	Tcl_SetErrorCode(interp, argv[1], argv[2], argv[3], argv[4], NULL);
+	break;
+    case 6:
+	Tcl_SetErrorCode(interp, argv[1], argv[2], argv[3], argv[4],
+		argv[5], NULL);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsetobjerrorcodeCmd --
+ *
+ *	This procedure implements the "testsetobjerrorcodeCmd".
+ *	This tests the Tcl_SetObjErrorCode function.
+ *
+ * Results:
+ *	A standard Tcl result. Always returns TCL_ERROR so that
+ *	the error code can be tested.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestsetobjerrorcodeCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    Tcl_SetObjErrorCode(interp, Tcl_ConcatObj(objc - 1, objv + 1));
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestfeventCmd --
+ *
+ *	This procedure implements the "testfevent" command.  It is
+ *	used for testing the "fileevent" command.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates and deletes interpreters.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestfeventCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    static Tcl_Interp *interp2 = NULL;
+    int code;
+    Tcl_Channel chan;
+
+    if (argc < 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" option ?arg ...?", NULL);
+	return TCL_ERROR;
+    }
+    if (strcmp(argv[1], "cmd") == 0) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " cmd script", NULL);
+	    return TCL_ERROR;
+	}
+	if (interp2 != NULL) {
+	    code = Tcl_EvalEx(interp2, argv[2], -1, TCL_EVAL_GLOBAL);
+	    Tcl_SetObjResult(interp, Tcl_GetObjResult(interp2));
+	    return code;
+	} else {
+	    Tcl_AppendResult(interp,
+		    "called \"testfevent code\" before \"testfevent create\"",
+		    NULL);
+	    return TCL_ERROR;
+	}
+    } else if (strcmp(argv[1], "create") == 0) {
+	if (interp2 != NULL) {
+	    Tcl_DeleteInterp(interp2);
+	}
+	interp2 = Tcl_CreateInterp();
+	return Tcl_Init(interp2);
+    } else if (strcmp(argv[1], "delete") == 0) {
+	if (interp2 != NULL) {
+	    Tcl_DeleteInterp(interp2);
+	}
+	interp2 = NULL;
+    } else if (strcmp(argv[1], "share") == 0) {
+	if (interp2 != NULL) {
+	    chan = Tcl_GetChannel(interp, argv[2], NULL);
+	    if (chan == (Tcl_Channel) NULL) {
+		return TCL_ERROR;
+	    }
+	    Tcl_RegisterChannel(interp2, chan);
+	}
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestpanicCmd --
+ *
+ *	Calls the panic routine.
+ *
+ * Results:
+ *	Always returns TCL_OK.
+ *
+ * Side effects:
+ *	May exit application.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestpanicCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    char *argString;
+
+    /*
+     *  Put the arguments into a var args structure
+     *  Append all of the arguments together separated by spaces
+     */
+
+    argString = Tcl_Merge(argc-1, argv+1);
+    Tcl_Panic("%s", argString);
+    ckfree(argString);
+
+    return TCL_OK;
+}
+
+static int
+TestfileCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    Tcl_Obj *const argv[])	/* The argument objects. */
+{
+    int force, i, j, result;
+    Tcl_Obj *error = NULL;
+    const char *subcmd;
+
+    if (argc < 3) {
+	return TCL_ERROR;
+    }
+
+    force = 0;
+    i = 2;
+    if (strcmp(Tcl_GetString(argv[2]), "-force") == 0) {
+	force = 1;
+	i = 3;
+    }
+
+    if (argc - i > 2) {
+	return TCL_ERROR;
+    }
+
+    for (j = i; j < argc; j++) {
+	if (Tcl_FSGetNormalizedPath(interp, argv[j]) == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    subcmd = Tcl_GetString(argv[1]);
+
+    if (strcmp(subcmd, "mv") == 0) {
+	result = TclpObjRenameFile(argv[i], argv[i + 1]);
+    } else if (strcmp(subcmd, "cp") == 0) {
+	result = TclpObjCopyFile(argv[i], argv[i + 1]);
+    } else if (strcmp(subcmd, "rm") == 0) {
+	result = TclpObjDeleteFile(argv[i]);
+    } else if (strcmp(subcmd, "mkdir") == 0) {
+	result = TclpObjCreateDirectory(argv[i]);
+    } else if (strcmp(subcmd, "cpdir") == 0) {
+	result = TclpObjCopyDirectory(argv[i], argv[i + 1], &error);
+    } else if (strcmp(subcmd, "rmdir") == 0) {
+	result = TclpObjRemoveDirectory(argv[i], force, &error);
+    } else {
+	result = TCL_ERROR;
+	goto end;
+    }
+
+    if (result != TCL_OK) {
+	if (error != NULL) {
+	    if (Tcl_GetString(error)[0] != '\0') {
+		Tcl_AppendResult(interp, Tcl_GetString(error), " ", NULL);
+	    }
+	    Tcl_DecrRefCount(error);
+	}
+	Tcl_AppendResult(interp, Tcl_ErrnoId(), NULL);
+    }
+
+  end:
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestgetvarfullnameCmd --
+ *
+ *	Implements the "testgetvarfullname" cmd that is used when testing
+ *	the Tcl_GetVariableFullName procedure.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestgetvarfullnameCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    const char *name, *arg;
+    int flags = 0;
+    Tcl_Namespace *namespacePtr;
+    Tcl_CallFrame *framePtr;
+    Tcl_Var variable;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "name scope");
+	return TCL_ERROR;
+    }
+
+    name = Tcl_GetString(objv[1]);
+
+    arg = Tcl_GetString(objv[2]);
+    if (strcmp(arg, "global") == 0) {
+	flags = TCL_GLOBAL_ONLY;
+    } else if (strcmp(arg, "namespace") == 0) {
+	flags = TCL_NAMESPACE_ONLY;
+    }
+
+    /*
+     * This command, like any other created with Tcl_Create[Obj]Command, runs
+     * in the global namespace. As a "namespace-aware" command that needs to
+     * run in a particular namespace, it must activate that namespace itself.
+     */
+
+    if (flags == TCL_NAMESPACE_ONLY) {
+	namespacePtr = Tcl_FindNamespace(interp, "::test_ns_var", NULL,
+		TCL_LEAVE_ERR_MSG);
+	if (namespacePtr == NULL) {
+	    return TCL_ERROR;
+	}
+	(void) TclPushStackFrame(interp, &framePtr, namespacePtr,
+		/*isProcCallFrame*/ 0);
+    }
+
+    variable = Tcl_FindNamespaceVar(interp, name, NULL,
+	    (flags | TCL_LEAVE_ERR_MSG));
+
+    if (flags == TCL_NAMESPACE_ONLY) {
+	TclPopStackFrame(interp);
+    }
+    if (variable == (Tcl_Var) NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_GetVariableFullName(interp, variable, Tcl_GetObjResult(interp));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetTimesObjCmd --
+ *
+ *	This procedure implements the "gettimes" command.  It is used for
+ *	computing the time needed for various basic operations such as reading
+ *	variables, allocating memory, sprintf, converting variables, etc.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Allocates and frees memory, sets a variable "a" in the interpreter.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetTimesObjCmd(
+    ClientData unused,		/* Unused. */
+    Tcl_Interp *interp,		/* The current interpreter. */
+    int notused1,			/* Number of arguments. */
+    Tcl_Obj *const notused2[])	/* The argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int i, n;
+    double timePer;
+    Tcl_Time start, stop;
+    Tcl_Obj *objPtr, **objv;
+    const char *s;
+    char newString[TCL_INTEGER_SPACE];
+
+    /* alloc & free 100000 times */
+    fprintf(stderr, "alloc & free 100000 6 word items\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	objPtr = ckalloc(sizeof(Tcl_Obj));
+	ckfree(objPtr);
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per alloc+free\n", timePer/100000);
+
+    /* alloc 5000 times */
+    fprintf(stderr, "alloc 5000 6 word items\n");
+    objv = ckalloc(5000 * sizeof(Tcl_Obj *));
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 5000;  i++) {
+	objv[i] = ckalloc(sizeof(Tcl_Obj));
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per alloc\n", timePer/5000);
+
+    /* free 5000 times */
+    fprintf(stderr, "free 5000 6 word items\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 5000;  i++) {
+	ckfree(objv[i]);
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per free\n", timePer/5000);
+
+    /* Tcl_NewObj 5000 times */
+    fprintf(stderr, "Tcl_NewObj 5000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 5000;  i++) {
+	objv[i] = Tcl_NewObj();
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_NewObj\n", timePer/5000);
+
+    /* Tcl_DecrRefCount 5000 times */
+    fprintf(stderr, "Tcl_DecrRefCount 5000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 5000;  i++) {
+	objPtr = objv[i];
+	Tcl_DecrRefCount(objPtr);
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_DecrRefCount\n", timePer/5000);
+    ckfree(objv);
+
+    /* TclGetString 100000 times */
+    fprintf(stderr, "TclGetStringFromObj of \"12345\" 100000 times\n");
+    objPtr = Tcl_NewStringObj("12345", -1);
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	(void) TclGetString(objPtr);
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per TclGetStringFromObj of \"12345\"\n",
+	    timePer/100000);
+
+    /* Tcl_GetIntFromObj 100000 times */
+    fprintf(stderr, "Tcl_GetIntFromObj of \"12345\" 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	if (Tcl_GetIntFromObj(interp, objPtr, &n) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_GetIntFromObj of \"12345\"\n",
+	    timePer/100000);
+    Tcl_DecrRefCount(objPtr);
+
+    /* Tcl_GetInt 100000 times */
+    fprintf(stderr, "Tcl_GetInt of \"12345\" 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	if (Tcl_GetInt(interp, "12345", &n) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_GetInt of \"12345\"\n",
+	    timePer/100000);
+
+    /* sprintf 100000 times */
+    fprintf(stderr, "sprintf of 12345 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	sprintf(newString, "%d", 12345);
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per sprintf of 12345\n",
+	    timePer/100000);
+
+    /* hashtable lookup 100000 times */
+    fprintf(stderr, "hashtable lookup of \"gettimes\" 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	(void) Tcl_FindHashEntry(&iPtr->globalNsPtr->cmdTable, "gettimes");
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per hashtable lookup of \"gettimes\"\n",
+	    timePer/100000);
+
+    /* Tcl_SetVar 100000 times */
+    fprintf(stderr, "Tcl_SetVar2 of \"12345\" 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	s = Tcl_SetVar2(interp, "a", NULL, "12345", TCL_LEAVE_ERR_MSG);
+	if (s == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_SetVar of a to \"12345\"\n",
+	    timePer/100000);
+
+    /* Tcl_GetVar 100000 times */
+    fprintf(stderr, "Tcl_GetVar of a==\"12345\" 100000 times\n");
+    Tcl_GetTime(&start);
+    for (i = 0;  i < 100000;  i++) {
+	s = Tcl_GetVar2(interp, "a", NULL, TCL_LEAVE_ERR_MSG);
+	if (s == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+    Tcl_GetTime(&stop);
+    timePer = (stop.sec - start.sec)*1000000 + (stop.usec - start.usec);
+    fprintf(stderr, "   %.3f usec per Tcl_GetVar of a==\"12345\"\n",
+	    timePer/100000);
+
+    Tcl_ResetResult(interp);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NoopCmd --
+ *
+ *	This procedure is just used to time the overhead involved in
+ *	parsing and invoking a command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NoopCmd(
+    ClientData unused,		/* Unused. */
+    Tcl_Interp *interp,		/* The current interpreter. */
+    int argc,			/* The number of arguments. */
+    const char **argv)		/* The argument strings. */
+{
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * NoopObjCmd --
+ *
+ *	This object-based procedure is just used to time the overhead
+ *	involved in parsing and invoking a command.
+ *
+ * Results:
+ *	Returns the TCL_OK result code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+NoopObjCmd(
+    ClientData unused,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestbytestringObjCmd --
+ *
+ *	This object-based procedure constructs a string which can
+ *	possibly contain invalid UTF-8 bytes.
+ *
+ * Results:
+ *	Returns the TCL_OK result code.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestbytestringObjCmd(
+    ClientData unused,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    int n;
+    const char *p;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "bytearray");
+	return TCL_ERROR;
+    }
+    p = (const char *)Tcl_GetByteArrayFromObj(objv[1], &n);
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(p, n));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsetCmd --
+ *
+ *	Implements the "testset{err,noerr}" cmds that are used when testing
+ *	Tcl_Set/GetVar C Api with/without TCL_LEAVE_ERR_MSG flag
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *     Variables may be set.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestsetCmd(
+    ClientData data,		/* Additional flags for Get/SetVar2. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int flags = PTR2INT(data);
+    const char *value;
+
+    if (argc == 2) {
+	Tcl_AppendResult(interp, "before get", NULL);
+	value = Tcl_GetVar2(interp, argv[1], NULL, flags);
+	if (value == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_AppendElement(interp, value);
+	return TCL_OK;
+    } else if (argc == 3) {
+	Tcl_AppendResult(interp, "before set", NULL);
+	value = Tcl_SetVar2(interp, argv[1], NULL, argv[2], flags);
+	if (value == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_AppendElement(interp, value);
+	return TCL_OK;
+    } else {
+	Tcl_AppendResult(interp, "wrong # args: should be \"",
+		argv[0], " varName ?newValue?\"", NULL);
+	return TCL_ERROR;
+    }
+}
+static int
+Testset2Cmd(
+    ClientData data,		/* Additional flags for Get/SetVar2. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    int flags = PTR2INT(data);
+    const char *value;
+
+    if (argc == 3) {
+	Tcl_AppendResult(interp, "before get", NULL);
+	value = Tcl_GetVar2(interp, argv[1], argv[2], flags);
+	if (value == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_AppendElement(interp, value);
+	return TCL_OK;
+    } else if (argc == 4) {
+	Tcl_AppendResult(interp, "before set", NULL);
+	value = Tcl_SetVar2(interp, argv[1], argv[2], argv[3], flags);
+	if (value == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_AppendElement(interp, value);
+	return TCL_OK;
+    } else {
+	Tcl_AppendResult(interp, "wrong # args: should be \"",
+		argv[0], " varName elemName ?newValue?\"", NULL);
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsaveresultCmd --
+ *
+ *	Implements the "testsaveresult" cmd that is used when testing the
+ *	Tcl_SaveResult, Tcl_RestoreResult, and Tcl_DiscardResult interfaces.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestsaveresultCmd(
+    ClientData dummy,		/* Not used. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* The argument objects. */
+{
+    Interp* iPtr = (Interp*) interp;
+    int discard, result, index;
+    Tcl_SavedResult state;
+    Tcl_Obj *objPtr;
+    static const char *const optionStrings[] = {
+	"append", "dynamic", "free", "object", "small", NULL
+    };
+    enum options {
+	RESULT_APPEND, RESULT_DYNAMIC, RESULT_FREE, RESULT_OBJECT, RESULT_SMALL
+    };
+
+    /*
+     * Parse arguments
+     */
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "type script discard");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], optionStrings, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetBooleanFromObj(interp, objv[3], &discard) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    freeCount = 0;
+    objPtr = NULL;		/* Lint. */
+    switch ((enum options) index) {
+    case RESULT_SMALL:
+	Tcl_AppendResult(interp, "small result", NULL);
+	break;
+    case RESULT_APPEND:
+	Tcl_AppendResult(interp, "append result", NULL);
+	break;
+    case RESULT_FREE: {
+	char *buf = ckalloc(200);
+
+	strcpy(buf, "free result");
+	Tcl_SetResult(interp, buf, TCL_DYNAMIC);
+	break;
+    }
+    case RESULT_DYNAMIC:
+	Tcl_SetResult(interp, (char *)"dynamic result", TestsaveresultFree);
+	break;
+    case RESULT_OBJECT:
+	objPtr = Tcl_NewStringObj("object result", -1);
+	Tcl_SetObjResult(interp, objPtr);
+	break;
+    }
+
+    Tcl_SaveResult(interp, &state);
+
+    if (((enum options) index) == RESULT_OBJECT) {
+	result = Tcl_EvalObjEx(interp, objv[2], 0);
+    } else {
+	result = Tcl_EvalEx(interp, Tcl_GetString(objv[2]), -1, 0);
+    }
+
+    if (discard) {
+	Tcl_DiscardResult(&state);
+    } else {
+	Tcl_RestoreResult(interp, &state);
+	result = TCL_OK;
+    }
+
+    switch ((enum options) index) {
+    case RESULT_DYNAMIC: {
+	int presentOrFreed = (iPtr->freeProc == TestsaveresultFree) ^ freeCount;
+
+	Tcl_AppendElement(interp, presentOrFreed ? "presentOrFreed" : "missingOrLeak");
+	break;
+    }
+    case RESULT_OBJECT:
+	Tcl_AppendElement(interp, Tcl_GetObjResult(interp) == objPtr
+		? "same" : "different");
+	break;
+    default:
+	break;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsaveresultFree --
+ *
+ *	Special purpose freeProc used by TestsaveresultCmd.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Increments the freeCount.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TestsaveresultFree(
+    char *blockPtr)
+{
+    freeCount++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestmainthreadCmd  --
+ *
+ *	Implements the "testmainthread" cmd that is used to test the
+ *	'Tcl_GetCurrentThread' API.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestmainthreadCmd(
+    ClientData dummy,		/* Not used. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    if (argc == 1) {
+	Tcl_Obj *idObj = Tcl_NewLongObj((long)(size_t)Tcl_GetCurrentThread());
+
+	Tcl_SetObjResult(interp, idObj);
+	return TCL_OK;
+    } else {
+	Tcl_AppendResult(interp, "wrong # args", NULL);
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MainLoop --
+ *
+ *	A main loop set by TestsetmainloopCmd below.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Event handlers could do anything.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MainLoop(void)
+{
+    while (!exitMainLoop) {
+	Tcl_DoOneEvent(0);
+    }
+    fprintf(stdout,"Exit MainLoop\n");
+    fflush(stdout);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestsetmainloopCmd  --
+ *
+ *	Implements the "testsetmainloop" cmd that is used to test the
+ *	'Tcl_SetMainLoop' API.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestsetmainloopCmd(
+    ClientData dummy,		/* Not used. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+  exitMainLoop = 0;
+  Tcl_SetMainLoop(MainLoop);
+  return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestexitmainloopCmd  --
+ *
+ *	Implements the "testexitmainloop" cmd that is used to test the
+ *	'Tcl_SetMainLoop' API.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestexitmainloopCmd(
+    ClientData dummy,		/* Not used. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+  exitMainLoop = 1;
+  return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestChannelCmd --
+ *
+ *	Implements the Tcl "testchannel" debugging command and its
+ *	subcommands. This is part of the testing environment.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestChannelCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter for result. */
+    int argc,			/* Count of additional args. */
+    const char **argv)		/* Additional arg strings. */
+{
+    const char *cmdName;	/* Sub command. */
+    Tcl_HashTable *hTblPtr;	/* Hash table of channels. */
+    Tcl_HashSearch hSearch;	/* Search variable. */
+    Tcl_HashEntry *hPtr;	/* Search variable. */
+    Channel *chanPtr;		/* The actual channel. */
+    ChannelState *statePtr;	/* state info for channel */
+    Tcl_Channel chan;		/* The opaque type. */
+    size_t len;			/* Length of subcommand string. */
+    int IOQueued;		/* How much IO is queued inside channel? */
+    char buf[TCL_INTEGER_SPACE];/* For sprintf. */
+    int mode;			/* rw mode of the channel */
+
+    if (argc < 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" subcommand ?additional args..?\"", NULL);
+	return TCL_ERROR;
+    }
+    cmdName = argv[1];
+    len = strlen(cmdName);
+
+    chanPtr = NULL;
+
+    if (argc > 2) {
+	if ((cmdName[0] == 's') && (strncmp(cmdName, "splice", len) == 0)) {
+	    /* For splice access the pool of detached channels.
+	     * Locate channel, remove from the list.
+	     */
+
+	    TestChannel **nextPtrPtr, *curPtr;
+
+	    chan = (Tcl_Channel) NULL;
+	    for (nextPtrPtr = &firstDetached, curPtr = firstDetached;
+		 curPtr != NULL;
+		 nextPtrPtr = &(curPtr->nextPtr), curPtr = curPtr->nextPtr) {
+
+		if (strcmp(argv[2], Tcl_GetChannelName(curPtr->chan)) == 0) {
+		    *nextPtrPtr = curPtr->nextPtr;
+		    curPtr->nextPtr = NULL;
+		    chan = curPtr->chan;
+		    ckfree(curPtr);
+		    break;
+		}
+	    }
+	} else {
+	    chan = Tcl_GetChannel(interp, argv[2], &mode);
+	}
+	if (chan == (Tcl_Channel) NULL) {
+	    return TCL_ERROR;
+	}
+	chanPtr		= (Channel *) chan;
+	statePtr	= chanPtr->state;
+	chanPtr		= statePtr->topChanPtr;
+	chan		= (Tcl_Channel) chanPtr;
+    } else {
+	/* lint */
+	statePtr	= NULL;
+	chan		= NULL;
+    }
+
+    if ((cmdName[0] == 's') && (strncmp(cmdName, "setchannelerror", len) == 0)) {
+
+	Tcl_Obj *msg = Tcl_NewStringObj(argv[3],-1);
+
+	Tcl_IncrRefCount(msg);
+	Tcl_SetChannelError(chan, msg);
+	Tcl_DecrRefCount(msg);
+
+	Tcl_GetChannelError(chan, &msg);
+	Tcl_SetObjResult(interp, msg);
+	Tcl_DecrRefCount(msg);
+	return TCL_OK;
+    }
+    if ((cmdName[0] == 's') && (strncmp(cmdName, "setchannelerrorinterp", len) == 0)) {
+
+	Tcl_Obj *msg = Tcl_NewStringObj(argv[3],-1);
+
+	Tcl_IncrRefCount(msg);
+	Tcl_SetChannelErrorInterp(interp, msg);
+	Tcl_DecrRefCount(msg);
+
+	Tcl_GetChannelErrorInterp(interp, &msg);
+	Tcl_SetObjResult(interp, msg);
+	Tcl_DecrRefCount(msg);
+	return TCL_OK;
+    }
+
+    /*
+     * "cut" is actually more a simplified detach facility as provided by the
+     * Thread package. Without the safeguards of a regular command (no
+     * checking that the command is truly cut'able, no mutexes for
+     * thread-safety). Its complementary command is "splice", see below.
+     */
+
+    if ((cmdName[0] == 'c') && (strncmp(cmdName, "cut", len) == 0)) {
+	TestChannel *det;
+
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " cut channelName\"", NULL);
+	    return TCL_ERROR;
+	}
+
+	Tcl_RegisterChannel(NULL, chan); /* prevent closing */
+	Tcl_UnregisterChannel(interp, chan);
+
+	Tcl_CutChannel(chan);
+
+	/* Remember the channel in the pool of detached channels */
+
+	det = ckalloc(sizeof(TestChannel));
+	det->chan     = chan;
+	det->nextPtr  = firstDetached;
+	firstDetached = det;
+
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'c') &&
+	    (strncmp(cmdName, "clearchannelhandlers", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " clearchannelhandlers channelName\"", NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_ClearChannelHandlers(chan);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'i') && (strncmp(cmdName, "info", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " info channelName\"", NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_AppendElement(interp, argv[2]);
+	Tcl_AppendElement(interp, Tcl_ChannelName(chanPtr->typePtr));
+	if (statePtr->flags & TCL_READABLE) {
+	    Tcl_AppendElement(interp, "read");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	if (statePtr->flags & TCL_WRITABLE) {
+	    Tcl_AppendElement(interp, "write");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	if (statePtr->flags & CHANNEL_NONBLOCKING) {
+	    Tcl_AppendElement(interp, "nonblocking");
+	} else {
+	    Tcl_AppendElement(interp, "blocking");
+	}
+	if (statePtr->flags & CHANNEL_LINEBUFFERED) {
+	    Tcl_AppendElement(interp, "line");
+	} else if (statePtr->flags & CHANNEL_UNBUFFERED) {
+	    Tcl_AppendElement(interp, "none");
+	} else {
+	    Tcl_AppendElement(interp, "full");
+	}
+	if (statePtr->flags & BG_FLUSH_SCHEDULED) {
+	    Tcl_AppendElement(interp, "async_flush");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	if (statePtr->flags & CHANNEL_EOF) {
+	    Tcl_AppendElement(interp, "eof");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	if (statePtr->flags & CHANNEL_BLOCKED) {
+	    Tcl_AppendElement(interp, "blocked");
+	} else {
+	    Tcl_AppendElement(interp, "unblocked");
+	}
+	if (statePtr->inputTranslation == TCL_TRANSLATE_AUTO) {
+	    Tcl_AppendElement(interp, "auto");
+	    if (statePtr->flags & INPUT_SAW_CR) {
+		Tcl_AppendElement(interp, "saw_cr");
+	    } else {
+		Tcl_AppendElement(interp, "");
+	    }
+	} else if (statePtr->inputTranslation == TCL_TRANSLATE_LF) {
+	    Tcl_AppendElement(interp, "lf");
+	    Tcl_AppendElement(interp, "");
+	} else if (statePtr->inputTranslation == TCL_TRANSLATE_CR) {
+	    Tcl_AppendElement(interp, "cr");
+	    Tcl_AppendElement(interp, "");
+	} else if (statePtr->inputTranslation == TCL_TRANSLATE_CRLF) {
+	    Tcl_AppendElement(interp, "crlf");
+	    if (statePtr->flags & INPUT_SAW_CR) {
+		Tcl_AppendElement(interp, "queued_cr");
+	    } else {
+		Tcl_AppendElement(interp, "");
+	    }
+	}
+	if (statePtr->outputTranslation == TCL_TRANSLATE_AUTO) {
+	    Tcl_AppendElement(interp, "auto");
+	} else if (statePtr->outputTranslation == TCL_TRANSLATE_LF) {
+	    Tcl_AppendElement(interp, "lf");
+	} else if (statePtr->outputTranslation == TCL_TRANSLATE_CR) {
+	    Tcl_AppendElement(interp, "cr");
+	} else if (statePtr->outputTranslation == TCL_TRANSLATE_CRLF) {
+	    Tcl_AppendElement(interp, "crlf");
+	}
+	IOQueued = Tcl_InputBuffered(chan);
+	TclFormatInt(buf, IOQueued);
+	Tcl_AppendElement(interp, buf);
+
+	IOQueued = Tcl_OutputBuffered(chan);
+	TclFormatInt(buf, IOQueued);
+	Tcl_AppendElement(interp, buf);
+
+	TclFormatInt(buf, (int)Tcl_Tell(chan));
+	Tcl_AppendElement(interp, buf);
+
+	TclFormatInt(buf, statePtr->refCount);
+	Tcl_AppendElement(interp, buf);
+
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'i') &&
+	    (strncmp(cmdName, "inputbuffered", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+	IOQueued = Tcl_InputBuffered(chan);
+	TclFormatInt(buf, IOQueued);
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'i') && (strncmp(cmdName, "isshared", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	TclFormatInt(buf, Tcl_IsChannelShared(chan));
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'i') && (strncmp(cmdName, "isstandard", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	TclFormatInt(buf, Tcl_IsStandardChannel(chan));
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'm') && (strncmp(cmdName, "mode", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	if (statePtr->flags & TCL_READABLE) {
+	    Tcl_AppendElement(interp, "read");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	if (statePtr->flags & TCL_WRITABLE) {
+	    Tcl_AppendElement(interp, "write");
+	} else {
+	    Tcl_AppendElement(interp, "");
+	}
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'm') && (strncmp(cmdName, "mthread", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	TclFormatInt(buf, (size_t) Tcl_GetChannelThread(chan));
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'n') && (strncmp(cmdName, "name", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_AppendResult(interp, statePtr->channelName, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'o') && (strncmp(cmdName, "open", len) == 0)) {
+	hTblPtr = (Tcl_HashTable *) Tcl_GetAssocData(interp, "tclIO", NULL);
+	if (hTblPtr == NULL) {
+	    return TCL_OK;
+	}
+	for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch);
+	     hPtr != NULL;
+	     hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    Tcl_AppendElement(interp, Tcl_GetHashKey(hTblPtr, hPtr));
+	}
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'o') &&
+	    (strncmp(cmdName, "outputbuffered", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	IOQueued = Tcl_OutputBuffered(chan);
+	TclFormatInt(buf, IOQueued);
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'q') &&
+	    (strncmp(cmdName, "queuedcr", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	Tcl_AppendResult(interp,
+		(statePtr->flags & INPUT_SAW_CR) ? "1" : "0", NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'r') && (strncmp(cmdName, "readable", len) == 0)) {
+	hTblPtr = (Tcl_HashTable *) Tcl_GetAssocData(interp, "tclIO", NULL);
+	if (hTblPtr == NULL) {
+	    return TCL_OK;
+	}
+	for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch);
+	     hPtr != NULL;
+	     hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    chanPtr  = (Channel *) Tcl_GetHashValue(hPtr);
+	    statePtr = chanPtr->state;
+	    if (statePtr->flags & TCL_READABLE) {
+		Tcl_AppendElement(interp, Tcl_GetHashKey(hTblPtr, hPtr));
+	    }
+	}
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'r') && (strncmp(cmdName, "refcount", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	TclFormatInt(buf, statePtr->refCount);
+	Tcl_AppendResult(interp, buf, NULL);
+	return TCL_OK;
+    }
+
+    /*
+     * "splice" is actually more a simplified attach facility as provided by
+     * the Thread package. Without the safeguards of a regular command (no
+     * checking that the command is truly cut'able, no mutexes for
+     * thread-safety). Its complementary command is "cut", see above.
+     */
+
+    if ((cmdName[0] == 's') && (strncmp(cmdName, "splice", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+
+	Tcl_SpliceChannel(chan);
+
+	Tcl_RegisterChannel(interp, chan);
+	Tcl_UnregisterChannel(NULL, chan);
+
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 't') && (strncmp(cmdName, "type", len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "channel name required", NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_AppendResult(interp, Tcl_ChannelName(chanPtr->typePtr), NULL);
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 'w') && (strncmp(cmdName, "writable", len) == 0)) {
+	hTblPtr = (Tcl_HashTable *) Tcl_GetAssocData(interp, "tclIO", NULL);
+	if (hTblPtr == NULL) {
+	    return TCL_OK;
+	}
+	for (hPtr = Tcl_FirstHashEntry(hTblPtr, &hSearch);
+		hPtr != NULL; hPtr = Tcl_NextHashEntry(&hSearch)) {
+	    chanPtr = (Channel *) Tcl_GetHashValue(hPtr);
+	    statePtr = chanPtr->state;
+	    if (statePtr->flags & TCL_WRITABLE) {
+		Tcl_AppendElement(interp, Tcl_GetHashKey(hTblPtr, hPtr));
+	    }
+	}
+	return TCL_OK;
+    }
+
+    if ((cmdName[0] == 't') && (strncmp(cmdName, "transform", len) == 0)) {
+	/*
+	 * Syntax: transform channel -command command
+	 */
+
+	if (argc != 5) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " transform channelId -command cmd\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (strcmp(argv[3], "-command") != 0) {
+	    Tcl_AppendResult(interp, "bad argument \"", argv[3],
+		    "\": should be \"-command\"", NULL);
+	    return TCL_ERROR;
+	}
+
+	return TclChannelTransform(interp, chan,
+		Tcl_NewStringObj(argv[4], -1));
+    }
+
+    if ((cmdName[0] == 'u') && (strncmp(cmdName, "unstack", len) == 0)) {
+	/*
+	 * Syntax: unstack channel
+	 */
+
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " unstack channel\"", NULL);
+	    return TCL_ERROR;
+	}
+	return Tcl_UnstackChannel(interp, chan);
+    }
+
+    Tcl_AppendResult(interp, "bad option \"", cmdName, "\": should be "
+	    "cut, clearchannelhandlers, info, isshared, mode, open, "
+	    "readable, splice, writable, transform, unstack", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestChannelEventCmd --
+ *
+ *	This procedure implements the "testchannelevent" command. It is used
+ *	to test the Tcl channel event mechanism.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Creates, deletes and returns channel event handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestChannelEventCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_Obj *resultListPtr;
+    Channel *chanPtr;
+    ChannelState *statePtr;	/* state info for channel */
+    EventScriptRecord *esPtr, *prevEsPtr, *nextEsPtr;
+    const char *cmd;
+    int index, i, mask, len;
+
+    if ((argc < 3) || (argc > 5)) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" channelName cmd ?arg1? ?arg2?\"", NULL);
+	return TCL_ERROR;
+    }
+    chanPtr = (Channel *) Tcl_GetChannel(interp, argv[1], NULL);
+    if (chanPtr == NULL) {
+	return TCL_ERROR;
+    }
+    statePtr = chanPtr->state;
+
+    cmd = argv[2];
+    len = strlen(cmd);
+    if ((cmd[0] == 'a') && (strncmp(cmd, "add", (unsigned) len) == 0)) {
+	if (argc != 5) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " channelName add eventSpec script\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (strcmp(argv[3], "readable") == 0) {
+	    mask = TCL_READABLE;
+	} else if (strcmp(argv[3], "writable") == 0) {
+	    mask = TCL_WRITABLE;
+	} else if (strcmp(argv[3], "none") == 0) {
+	    mask = 0;
+	} else {
+	    Tcl_AppendResult(interp, "bad event name \"", argv[3],
+		    "\": must be readable, writable, or none", NULL);
+	    return TCL_ERROR;
+	}
+
+	esPtr = ckalloc(sizeof(EventScriptRecord));
+	esPtr->nextPtr = statePtr->scriptRecordPtr;
+	statePtr->scriptRecordPtr = esPtr;
+
+	esPtr->chanPtr = chanPtr;
+	esPtr->interp = interp;
+	esPtr->mask = mask;
+	esPtr->scriptPtr = Tcl_NewStringObj(argv[4], -1);
+	Tcl_IncrRefCount(esPtr->scriptPtr);
+
+	Tcl_CreateChannelHandler((Tcl_Channel) chanPtr, mask,
+		TclChannelEventScriptInvoker, (ClientData) esPtr);
+
+	return TCL_OK;
+    }
+
+    if ((cmd[0] == 'd') && (strncmp(cmd, "delete", (unsigned) len) == 0)) {
+	if (argc != 4) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " channelName delete index\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetInt(interp, argv[3], &index) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (index < 0) {
+	    Tcl_AppendResult(interp, "bad event index: ", argv[3],
+		    ": must be nonnegative", NULL);
+	    return TCL_ERROR;
+	}
+	for (i = 0, esPtr = statePtr->scriptRecordPtr;
+	     (i < index) && (esPtr != NULL);
+	     i++, esPtr = esPtr->nextPtr) {
+	    /* Empty loop body. */
+	}
+	if (esPtr == NULL) {
+	    Tcl_AppendResult(interp, "bad event index ", argv[3],
+		    ": out of range", NULL);
+	    return TCL_ERROR;
+	}
+	if (esPtr == statePtr->scriptRecordPtr) {
+	    statePtr->scriptRecordPtr = esPtr->nextPtr;
+	} else {
+	    for (prevEsPtr = statePtr->scriptRecordPtr;
+		 (prevEsPtr != NULL) &&
+		     (prevEsPtr->nextPtr != esPtr);
+		 prevEsPtr = prevEsPtr->nextPtr) {
+		/* Empty loop body. */
+	    }
+	    if (prevEsPtr == NULL) {
+		Tcl_Panic("TestChannelEventCmd: damaged event script list");
+	    }
+	    prevEsPtr->nextPtr = esPtr->nextPtr;
+	}
+	Tcl_DeleteChannelHandler((Tcl_Channel) chanPtr,
+		TclChannelEventScriptInvoker, (ClientData) esPtr);
+	Tcl_DecrRefCount(esPtr->scriptPtr);
+	ckfree(esPtr);
+
+	return TCL_OK;
+    }
+
+    if ((cmd[0] == 'l') && (strncmp(cmd, "list", (unsigned) len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " channelName list\"", NULL);
+	    return TCL_ERROR;
+	}
+	resultListPtr = Tcl_GetObjResult(interp);
+	for (esPtr = statePtr->scriptRecordPtr;
+	     esPtr != NULL;
+	     esPtr = esPtr->nextPtr) {
+	    if (esPtr->mask) {
+		Tcl_ListObjAppendElement(interp, resultListPtr, Tcl_NewStringObj(
+		    (esPtr->mask == TCL_READABLE) ? "readable" : "writable", -1));
+	    } else {
+		Tcl_ListObjAppendElement(interp, resultListPtr,
+			Tcl_NewStringObj("none", -1));
+	    }
+	    Tcl_ListObjAppendElement(interp, resultListPtr, esPtr->scriptPtr);
+	}
+	Tcl_SetObjResult(interp, resultListPtr);
+	return TCL_OK;
+    }
+
+    if ((cmd[0] == 'r') && (strncmp(cmd, "removeall", (unsigned) len) == 0)) {
+	if (argc != 3) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " channelName removeall\"", NULL);
+	    return TCL_ERROR;
+	}
+	for (esPtr = statePtr->scriptRecordPtr;
+	     esPtr != NULL;
+	     esPtr = nextEsPtr) {
+	    nextEsPtr = esPtr->nextPtr;
+	    Tcl_DeleteChannelHandler((Tcl_Channel) chanPtr,
+		    TclChannelEventScriptInvoker, (ClientData) esPtr);
+	    Tcl_DecrRefCount(esPtr->scriptPtr);
+	    ckfree(esPtr);
+	}
+	statePtr->scriptRecordPtr = NULL;
+	return TCL_OK;
+    }
+
+    if	((cmd[0] == 's') && (strncmp(cmd, "set", (unsigned) len) == 0)) {
+	if (argc != 5) {
+	    Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		    " channelName delete index event\"", NULL);
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetInt(interp, argv[3], &index) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (index < 0) {
+	    Tcl_AppendResult(interp, "bad event index: ", argv[3],
+		    ": must be nonnegative", NULL);
+	    return TCL_ERROR;
+	}
+	for (i = 0, esPtr = statePtr->scriptRecordPtr;
+	     (i < index) && (esPtr != NULL);
+	     i++, esPtr = esPtr->nextPtr) {
+	    /* Empty loop body. */
+	}
+	if (esPtr == NULL) {
+	    Tcl_AppendResult(interp, "bad event index ", argv[3],
+		    ": out of range", NULL);
+	    return TCL_ERROR;
+	}
+
+	if (strcmp(argv[4], "readable") == 0) {
+	    mask = TCL_READABLE;
+	} else if (strcmp(argv[4], "writable") == 0) {
+	    mask = TCL_WRITABLE;
+	} else if (strcmp(argv[4], "none") == 0) {
+	    mask = 0;
+	} else {
+	    Tcl_AppendResult(interp, "bad event name \"", argv[4],
+		    "\": must be readable, writable, or none", NULL);
+	    return TCL_ERROR;
+	}
+	esPtr->mask = mask;
+	Tcl_CreateChannelHandler((Tcl_Channel) chanPtr, mask,
+		TclChannelEventScriptInvoker, (ClientData) esPtr);
+	return TCL_OK;
+    }
+    Tcl_AppendResult(interp, "bad command ", cmd, ", must be one of "
+	    "add, delete, list, set, or removeall", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestSocketCmd --
+ *
+ *	Implements the Tcl "testsocket" debugging command and its
+ *	subcommands. This is part of the testing environment.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+TestSocketCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter for result. */
+    int argc,			/* Count of additional args. */
+    const char **argv)		/* Additional arg strings. */
+{
+    const char *cmdName;	/* Sub command. */
+    size_t len;			/* Length of subcommand string. */
+
+    if (argc < 2) {
+	Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+		" subcommand ?additional args..?\"", NULL);
+	return TCL_ERROR;
+    }
+    cmdName = argv[1];
+    len = strlen(cmdName);
+
+    if ((cmdName[0] == 't') && (strncmp(cmdName, "testflags", len) == 0)) {
+        Tcl_Channel hChannel;
+        int modePtr;
+        TcpState *statePtr;
+        /* Set test value in the socket driver
+         */
+        /* Check for argument "channel name"
+         */
+        if (argc < 4) {
+            Tcl_AppendResult(interp, "wrong # args: should be \"", argv[0],
+                    " testflags channel flags\"", NULL);
+            return TCL_ERROR;
+        }
+        hChannel = Tcl_GetChannel(interp, argv[2], &modePtr);
+        if ( NULL == hChannel ) {
+            Tcl_AppendResult(interp, "unknown channel:", argv[2], NULL);
+            return TCL_ERROR;
+        }
+        statePtr = (TcpState *)Tcl_GetChannelInstanceData(hChannel);
+        if ( NULL == statePtr) {
+            Tcl_AppendResult(interp, "No channel instance data:", argv[2],
+                    NULL);
+            return TCL_ERROR;
+        }
+        statePtr->testFlags = atoi(argv[3]);
+        return TCL_OK;
+    }
+
+    Tcl_AppendResult(interp, "bad option \"", cmdName, "\": should be "
+	    "testflags", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestWrongNumArgsObjCmd --
+ *
+ *	Test the Tcl_WrongNumArgs function.
+ *
+ * Results:
+ *	Standard Tcl result.
+ *
+ * Side effects:
+ *	Sets interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestWrongNumArgsObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int i, length;
+    const char *msg;
+
+    if (objc < 3) {
+	/*
+	 * Don't use Tcl_WrongNumArgs here, as that is the function
+	 * we want to test!
+	 */
+	Tcl_AppendResult(interp, "insufficient arguments", NULL);
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIntFromObj(interp, objv[1], &i) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    msg = Tcl_GetStringFromObj(objv[2], &length);
+    if (length == 0) {
+	msg = NULL;
+    }
+
+    if (i > objc - 3) {
+	/*
+	 * Asked for more arguments than were given.
+	 */
+	Tcl_AppendResult(interp, "insufficient arguments", NULL);
+	return TCL_ERROR;
+    }
+
+    Tcl_WrongNumArgs(interp, i, &(objv[3]), msg);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestGetIndexFromObjStructObjCmd --
+ *
+ *	Test the Tcl_GetIndexFromObjStruct function.
+ *
+ * Results:
+ *	Standard Tcl result.
+ *
+ * Side effects:
+ *	Sets interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestGetIndexFromObjStructObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *const ary[] = {
+	"a", "b", "c", "d", "e", "f", NULL, NULL
+    };
+    int idx,target;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "argument targetvalue");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObjStruct(interp, objv[1], ary, 2*sizeof(char *),
+	    "dummy", 0, &idx) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIntFromObj(interp, objv[2], &target) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (idx != target) {
+	char buffer[64];
+	sprintf(buffer, "%d", idx);
+	Tcl_AppendResult(interp, "index value comparison failed: got ",
+		buffer, NULL);
+	sprintf(buffer, "%d", target);
+	Tcl_AppendResult(interp, " when ", buffer, " expected", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_WrongNumArgs(interp, 3, objv, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestFilesystemObjCmd --
+ *
+ *	This procedure implements the "testfilesystem" command. It is used to
+ *	test Tcl_FSRegister, Tcl_FSUnregister, and can be used to test that
+ *	the pluggable filesystem works.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Inserts or removes a filesystem from Tcl's stack.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestFilesystemObjCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int res, boolVal;
+    const char *msg;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "boolean");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetBooleanFromObj(interp, objv[1], &boolVal) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (boolVal) {
+	res = Tcl_FSRegister((ClientData)interp, &testReportingFilesystem);
+	msg = (res == TCL_OK) ? "registered" : "failed";
+    } else {
+	res = Tcl_FSUnregister(&testReportingFilesystem);
+	msg = (res == TCL_OK) ? "unregistered" : "failed";
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(msg , -1));
+    return res;
+}
+
+static int
+TestReportInFilesystem(
+    Tcl_Obj *pathPtr,
+    ClientData *clientDataPtr)
+{
+    static Tcl_Obj *lastPathPtr = NULL;
+    Tcl_Obj *newPathPtr;
+
+    if (pathPtr == lastPathPtr) {
+	/* Reject all files second time around */
+	return -1;
+    }
+
+    /* Try to claim all files first time around */
+
+    newPathPtr = Tcl_DuplicateObj(pathPtr);
+    lastPathPtr = newPathPtr;
+    Tcl_IncrRefCount(newPathPtr);
+    if (Tcl_FSGetFileSystemForPath(newPathPtr) == NULL) {
+	/* Nothing claimed it. Therefore we don't either */
+	Tcl_DecrRefCount(newPathPtr);
+	lastPathPtr = NULL;
+	return -1;
+    }
+    lastPathPtr = NULL;
+    *clientDataPtr = (ClientData) newPathPtr;
+    return TCL_OK;
+}
+
+/*
+ * Simple helper function to extract the native vfs representation of a path
+ * object, or NULL if no such representation exists.
+ */
+
+static Tcl_Obj *
+TestReportGetNativePath(
+    Tcl_Obj *pathPtr)
+{
+    return (Tcl_Obj*) Tcl_FSGetInternalRep(pathPtr, &testReportingFilesystem);
+}
+
+static void
+TestReportFreeInternalRep(
+    ClientData clientData)
+{
+    Tcl_Obj *nativeRep = (Tcl_Obj *) clientData;
+
+    if (nativeRep != NULL) {
+	/* Free the path */
+	Tcl_DecrRefCount(nativeRep);
+    }
+}
+
+static ClientData
+TestReportDupInternalRep(
+    ClientData clientData)
+{
+    Tcl_Obj *original = (Tcl_Obj *) clientData;
+
+    Tcl_IncrRefCount(original);
+    return clientData;
+}
+
+static void
+TestReport(
+    const char *cmd,
+    Tcl_Obj *path,
+    Tcl_Obj *arg2)
+{
+    Tcl_Interp *interp = (Tcl_Interp *) Tcl_FSData(&testReportingFilesystem);
+
+    if (interp == NULL) {
+	/* This is bad, but not much we can do about it */
+    } else {
+	/*
+	 * No idea why I decided to program this up using the old string-based
+	 * API, but there you go. We should convert it to objects.
+	 */
+
+	Tcl_Obj *savedResult;
+	Tcl_DString ds;
+
+	Tcl_DStringInit(&ds);
+	Tcl_DStringAppend(&ds, "lappend filesystemReport ", -1);
+	Tcl_DStringStartSublist(&ds);
+	Tcl_DStringAppendElement(&ds, cmd);
+	if (path != NULL) {
+	    Tcl_DStringAppendElement(&ds, Tcl_GetString(path));
+	}
+	if (arg2 != NULL) {
+	    Tcl_DStringAppendElement(&ds, Tcl_GetString(arg2));
+	}
+	Tcl_DStringEndSublist(&ds);
+	savedResult = Tcl_GetObjResult(interp);
+	Tcl_IncrRefCount(savedResult);
+	Tcl_SetObjResult(interp, Tcl_NewObj());
+	Tcl_EvalEx(interp, Tcl_DStringValue(&ds), -1, 0);
+	Tcl_DStringFree(&ds);
+	Tcl_ResetResult(interp);
+	Tcl_SetObjResult(interp, savedResult);
+	Tcl_DecrRefCount(savedResult);
+    }
+}
+
+static int
+TestReportStat(
+    Tcl_Obj *path,		/* Path of file to stat (in current CP). */
+    Tcl_StatBuf *buf)		/* Filled with results of stat call. */
+{
+    TestReport("stat", path, NULL);
+    return Tcl_FSStat(TestReportGetNativePath(path), buf);
+}
+
+static int
+TestReportLstat(
+    Tcl_Obj *path,		/* Path of file to stat (in current CP). */
+    Tcl_StatBuf *buf)		/* Filled with results of stat call. */
+{
+    TestReport("lstat", path, NULL);
+    return Tcl_FSLstat(TestReportGetNativePath(path), buf);
+}
+
+static int
+TestReportAccess(
+    Tcl_Obj *path,		/* Path of file to access (in current CP). */
+    int mode)			/* Permission setting. */
+{
+    TestReport("access", path, NULL);
+    return Tcl_FSAccess(TestReportGetNativePath(path), mode);
+}
+
+static Tcl_Channel
+TestReportOpenFileChannel(
+    Tcl_Interp *interp,		/* Interpreter for error reporting; can be
+				 * NULL. */
+    Tcl_Obj *fileName,		/* Name of file to open. */
+    int mode,			/* POSIX open mode. */
+    int permissions)		/* If the open involves creating a file, with
+				 * what modes to create it? */
+{
+    TestReport("open", fileName, NULL);
+    return TclpOpenFileChannel(interp, TestReportGetNativePath(fileName),
+	    mode, permissions);
+}
+
+static int
+TestReportMatchInDirectory(
+    Tcl_Interp *interp,		/* Interpreter for error messages. */
+    Tcl_Obj *resultPtr,		/* Object to lappend results. */
+    Tcl_Obj *dirPtr,		/* Contains path to directory to search. */
+    const char *pattern,	/* Pattern to match against. */
+    Tcl_GlobTypeData *types)	/* Object containing list of acceptable types.
+				 * May be NULL. */
+{
+    if (types != NULL && types->type & TCL_GLOB_TYPE_MOUNT) {
+	TestReport("matchmounts", dirPtr, NULL);
+	return TCL_OK;
+    } else {
+	TestReport("matchindirectory", dirPtr, NULL);
+	return Tcl_FSMatchInDirectory(interp, resultPtr,
+		TestReportGetNativePath(dirPtr), pattern, types);
+    }
+}
+
+static int
+TestReportChdir(
+    Tcl_Obj *dirName)
+{
+    TestReport("chdir", dirName, NULL);
+    return Tcl_FSChdir(TestReportGetNativePath(dirName));
+}
+
+static int
+TestReportLoadFile(
+    Tcl_Interp *interp,		/* Used for error reporting. */
+    Tcl_Obj *fileName,		/* Name of the file containing the desired
+				 * code. */
+    Tcl_LoadHandle *handlePtr,	/* Filled with token for dynamically loaded
+				 * file which will be passed back to
+				 * (*unloadProcPtr)() to unload the file. */
+    Tcl_FSUnloadFileProc **unloadProcPtr)
+				/* Filled with address of Tcl_FSUnloadFileProc
+				 * function which should be used for
+				 * this file. */
+{
+    TestReport("loadfile", fileName, NULL);
+    return Tcl_FSLoadFile(interp, TestReportGetNativePath(fileName), NULL,
+	    NULL, NULL, NULL, handlePtr, unloadProcPtr);
+}
+
+static Tcl_Obj *
+TestReportLink(
+    Tcl_Obj *path,		/* Path of file to readlink or link */
+    Tcl_Obj *to,		/* Path of file to link to, or NULL */
+    int linkType)
+{
+    TestReport("link", path, to);
+    return Tcl_FSLink(TestReportGetNativePath(path), to, linkType);
+}
+
+static int
+TestReportRenameFile(
+    Tcl_Obj *src,		/* Pathname of file or dir to be renamed
+				 * (UTF-8). */
+    Tcl_Obj *dst)		/* New pathname of file or directory
+				 * (UTF-8). */
+{
+    TestReport("renamefile", src, dst);
+    return Tcl_FSRenameFile(TestReportGetNativePath(src),
+	    TestReportGetNativePath(dst));
+}
+
+static int
+TestReportCopyFile(
+    Tcl_Obj *src,		/* Pathname of file to be copied (UTF-8). */
+    Tcl_Obj *dst)		/* Pathname of file to copy to (UTF-8). */
+{
+    TestReport("copyfile", src, dst);
+    return Tcl_FSCopyFile(TestReportGetNativePath(src),
+	    TestReportGetNativePath(dst));
+}
+
+static int
+TestReportDeleteFile(
+    Tcl_Obj *path)		/* Pathname of file to be removed (UTF-8). */
+{
+    TestReport("deletefile", path, NULL);
+    return Tcl_FSDeleteFile(TestReportGetNativePath(path));
+}
+
+static int
+TestReportCreateDirectory(
+    Tcl_Obj *path)		/* Pathname of directory to create (UTF-8). */
+{
+    TestReport("createdirectory", path, NULL);
+    return Tcl_FSCreateDirectory(TestReportGetNativePath(path));
+}
+
+static int
+TestReportCopyDirectory(
+    Tcl_Obj *src,		/* Pathname of directory to be copied
+				 * (UTF-8). */
+    Tcl_Obj *dst,		/* Pathname of target directory (UTF-8). */
+    Tcl_Obj **errorPtr)		/* If non-NULL, to be filled with UTF-8 name
+				 * of file causing error. */
+{
+    TestReport("copydirectory", src, dst);
+    return Tcl_FSCopyDirectory(TestReportGetNativePath(src),
+	    TestReportGetNativePath(dst), errorPtr);
+}
+
+static int
+TestReportRemoveDirectory(
+    Tcl_Obj *path,		/* Pathname of directory to be removed
+				 * (UTF-8). */
+    int recursive,		/* If non-zero, removes directories that
+				 * are nonempty.  Otherwise, will only remove
+				 * empty directories. */
+    Tcl_Obj **errorPtr)		/* If non-NULL, to be filled with UTF-8 name
+				 * of file causing error. */
+{
+    TestReport("removedirectory", path, NULL);
+    return Tcl_FSRemoveDirectory(TestReportGetNativePath(path), recursive,
+	    errorPtr);
+}
+
+static const char *const *
+TestReportFileAttrStrings(
+    Tcl_Obj *fileName,
+    Tcl_Obj **objPtrRef)
+{
+    TestReport("fileattributestrings", fileName, NULL);
+    return Tcl_FSFileAttrStrings(TestReportGetNativePath(fileName), objPtrRef);
+}
+
+static int
+TestReportFileAttrsGet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *fileName,		/* filename we are operating on. */
+    Tcl_Obj **objPtrRef)	/* for output. */
+{
+    TestReport("fileattributesget", fileName, NULL);
+    return Tcl_FSFileAttrsGet(interp, index,
+	    TestReportGetNativePath(fileName), objPtrRef);
+}
+
+static int
+TestReportFileAttrsSet(
+    Tcl_Interp *interp,		/* The interpreter for error reporting. */
+    int index,			/* index of the attribute command. */
+    Tcl_Obj *fileName,		/* filename we are operating on. */
+    Tcl_Obj *objPtr)		/* for input. */
+{
+    TestReport("fileattributesset", fileName, objPtr);
+    return Tcl_FSFileAttrsSet(interp, index,
+	    TestReportGetNativePath(fileName), objPtr);
+}
+
+static int
+TestReportUtime(
+    Tcl_Obj *fileName,
+    struct utimbuf *tval)
+{
+    TestReport("utime", fileName, NULL);
+    return Tcl_FSUtime(TestReportGetNativePath(fileName), tval);
+}
+
+static int
+TestReportNormalizePath(
+    Tcl_Interp *interp,
+    Tcl_Obj *pathPtr,
+    int nextCheckpoint)
+{
+    TestReport("normalizepath", pathPtr, NULL);
+    return nextCheckpoint;
+}
+
+static int
+SimplePathInFilesystem(
+    Tcl_Obj *pathPtr,
+    ClientData *clientDataPtr)
+{
+    const char *str = Tcl_GetString(pathPtr);
+
+    if (strncmp(str, "simplefs:/", 10)) {
+	return -1;
+    }
+    return TCL_OK;
+}
+
+/*
+ * This is a slightly 'hacky' filesystem which is used just to test a few
+ * important features of the vfs code: (1) that you can load a shared library
+ * from a vfs, (2) that when copying files from one fs to another, the 'mtime'
+ * is preserved. (3) that recursive cross-filesystem directory copies have the
+ * correct behaviour with/without -force.
+ *
+ * It treats any file in 'simplefs:/' as a file, which it routes to the
+ * current directory. The real file it uses is whatever follows the trailing
+ * '/' (e.g. 'foo' in 'simplefs:/foo'), and that file exists or not according
+ * to what is in the native pwd.
+ *
+ * Please do not consider this filesystem a model of how things are to be
+ * done. It is quite the opposite!  But, it does allow us to test some
+ * important features.
+ */
+
+static int
+TestSimpleFilesystemObjCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int res, boolVal;
+    const char *msg;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "boolean");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetBooleanFromObj(interp, objv[1], &boolVal) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (boolVal) {
+	res = Tcl_FSRegister((ClientData)interp, &simpleFilesystem);
+	msg = (res == TCL_OK) ? "registered" : "failed";
+    } else {
+	res = Tcl_FSUnregister(&simpleFilesystem);
+	msg = (res == TCL_OK) ? "unregistered" : "failed";
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(msg , -1));
+    return res;
+}
+
+/*
+ * Treats a file name 'simplefs:/foo' by using the file 'foo' in the current
+ * (native) directory.
+ */
+
+static Tcl_Obj *
+SimpleRedirect(
+    Tcl_Obj *pathPtr)		/* Name of file to copy. */
+{
+    int len;
+    const char *str;
+    Tcl_Obj *origPtr;
+
+    /*
+     * We assume the same name in the current directory is ok.
+     */
+
+    str = Tcl_GetStringFromObj(pathPtr, &len);
+    if (len < 10 || strncmp(str, "simplefs:/", 10)) {
+	/* Probably shouldn't ever reach here */
+	Tcl_IncrRefCount(pathPtr);
+	return pathPtr;
+    }
+    origPtr = Tcl_NewStringObj(str+10,-1);
+    Tcl_IncrRefCount(origPtr);
+    return origPtr;
+}
+
+static int
+SimpleMatchInDirectory(
+    Tcl_Interp *interp,		/* Interpreter for error
+				 * messages. */
+    Tcl_Obj *resultPtr,		/* Object to lappend results. */
+    Tcl_Obj *dirPtr,		/* Contains path to directory to search. */
+    const char *pattern,	/* Pattern to match against. */
+    Tcl_GlobTypeData *types)	/* Object containing list of acceptable types.
+				 * May be NULL. */
+{
+    int res;
+    Tcl_Obj *origPtr;
+    Tcl_Obj *resPtr;
+
+    /* We only provide a new volume, therefore no mounts at all */
+    if (types != NULL && types->type & TCL_GLOB_TYPE_MOUNT) {
+	return TCL_OK;
+    }
+
+    /*
+     * We assume the same name in the current directory is ok.
+     */
+    resPtr = Tcl_NewObj();
+    Tcl_IncrRefCount(resPtr);
+    origPtr = SimpleRedirect(dirPtr);
+    res = Tcl_FSMatchInDirectory(interp, resPtr, origPtr, pattern, types);
+    if (res == TCL_OK) {
+	int gLength, j;
+	Tcl_ListObjLength(NULL, resPtr, &gLength);
+	for (j = 0; j < gLength; j++) {
+	    Tcl_Obj *gElt, *nElt;
+	    Tcl_ListObjIndex(NULL, resPtr, j, &gElt);
+	    nElt = Tcl_NewStringObj("simplefs:/",10);
+	    Tcl_AppendObjToObj(nElt, gElt);
+	    Tcl_ListObjAppendElement(NULL, resultPtr, nElt);
+	}
+    }
+    Tcl_DecrRefCount(origPtr);
+    Tcl_DecrRefCount(resPtr);
+    return res;
+}
+
+static Tcl_Channel
+SimpleOpenFileChannel(
+    Tcl_Interp *interp,		/* Interpreter for error reporting; can be
+				 * NULL. */
+    Tcl_Obj *pathPtr,		/* Name of file to open. */
+    int mode,			/* POSIX open mode. */
+    int permissions)		/* If the open involves creating a file, with
+				 * what modes to create it? */
+{
+    Tcl_Obj *tempPtr;
+    Tcl_Channel chan;
+
+    if ((mode != 0) && !(mode & O_RDONLY)) {
+	Tcl_AppendResult(interp, "read-only", NULL);
+	return NULL;
+    }
+
+    tempPtr = SimpleRedirect(pathPtr);
+    chan = Tcl_FSOpenFileChannel(interp, tempPtr, "r", permissions);
+    Tcl_DecrRefCount(tempPtr);
+    return chan;
+}
+
+static int
+SimpleAccess(
+    Tcl_Obj *pathPtr,		/* Path of file to access (in current CP). */
+    int mode)			/* Permission setting. */
+{
+    Tcl_Obj *tempPtr = SimpleRedirect(pathPtr);
+    int res = Tcl_FSAccess(tempPtr, mode);
+
+    Tcl_DecrRefCount(tempPtr);
+    return res;
+}
+
+static int
+SimpleStat(
+    Tcl_Obj *pathPtr,		/* Path of file to stat (in current CP). */
+    Tcl_StatBuf *bufPtr)	/* Filled with results of stat call. */
+{
+    Tcl_Obj *tempPtr = SimpleRedirect(pathPtr);
+    int res = Tcl_FSStat(tempPtr, bufPtr);
+
+    Tcl_DecrRefCount(tempPtr);
+    return res;
+}
+
+static Tcl_Obj *
+SimpleListVolumes(void)
+{
+    /* Add one new volume */
+    Tcl_Obj *retVal;
+
+    retVal = Tcl_NewStringObj("simplefs:/", -1);
+    Tcl_IncrRefCount(retVal);
+    return retVal;
+}
+
+/*
+ * Used to check correct string-length determining in Tcl_NumUtfChars
+ */
+
+static int
+TestNumUtfCharsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    if (objc > 1) {
+	int len = -1;
+
+	if (objc > 2) {
+	    (void) Tcl_GetIntFromObj(interp, objv[2], &len);
+	}
+	len = Tcl_NumUtfChars(Tcl_GetString(objv[1]), len);
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(len));
+    }
+    return TCL_OK;
+}
+
+#if defined(HAVE_CPUID) || defined(_WIN32)
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestcpuidCmd --
+ *
+ *	Retrieves CPU ID information.
+ *
+ * Usage:
+ *	testwincpuid <eax>
+ *
+ * Parameters:
+ *	eax - The value to pass in the EAX register to a CPUID instruction.
+ *
+ * Results:
+ *	Returns a four-element list containing the values from the EAX, EBX,
+ *	ECX and EDX registers returned from the CPUID instruction.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestcpuidCmd(
+    ClientData dummy,
+    Tcl_Interp* interp,		/* Tcl interpreter */
+    int objc,			/* Parameter count */
+    Tcl_Obj *const * objv)	/* Parameter vector */
+{
+    int status, index, i;
+    int regs[4];
+    Tcl_Obj *regsObjs[4];
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "eax");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIntFromObj(interp, objv[1], &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    status = TclWinCPUID(index, regs);
+    if (status != TCL_OK) {
+	Tcl_SetObjResult(interp,
+		Tcl_NewStringObj("operation not available", -1));
+	return status;
+    }
+    for (i=0 ; i<4 ; ++i) {
+	regsObjs[i] = Tcl_NewIntObj(regs[i]);
+    }
+    Tcl_SetObjResult(interp, Tcl_NewListObj(4, regsObjs));
+    return TCL_OK;
+}
+#endif
+
+/*
+ * Used to do basic checks of the TCL_HASH_KEY_SYSTEM_HASH flag
+ */
+
+static int
+TestHashSystemHashCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const Tcl_HashKeyType hkType = {
+	TCL_HASH_KEY_TYPE_VERSION, TCL_HASH_KEY_SYSTEM_HASH,
+	NULL, NULL, NULL, NULL
+    };
+    Tcl_HashTable hash;
+    Tcl_HashEntry *hPtr;
+    int i, isNew, limit = 100;
+
+    if (objc>1 && Tcl_GetIntFromObj(interp, objv[1], &limit)!=TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    Tcl_InitCustomHashTable(&hash, TCL_CUSTOM_TYPE_KEYS, &hkType);
+
+    if (hash.numEntries != 0) {
+	Tcl_AppendResult(interp, "non-zero initial size", NULL);
+	Tcl_DeleteHashTable(&hash);
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<limit ; i++) {
+	hPtr = Tcl_CreateHashEntry(&hash, INT2PTR(i), &isNew);
+	if (!isNew) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(i));
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp)," creation problem",-1);
+	    Tcl_DeleteHashTable(&hash);
+	    return TCL_ERROR;
+	}
+	Tcl_SetHashValue(hPtr, INT2PTR(i+42));
+    }
+
+    if (hash.numEntries != limit) {
+	Tcl_AppendResult(interp, "unexpected maximal size", NULL);
+	Tcl_DeleteHashTable(&hash);
+	return TCL_ERROR;
+    }
+
+    for (i=0 ; i<limit ; i++) {
+	hPtr = Tcl_FindHashEntry(&hash, (char *) INT2PTR(i));
+	if (hPtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(i));
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp)," lookup problem",-1);
+	    Tcl_DeleteHashTable(&hash);
+	    return TCL_ERROR;
+	}
+	if (PTR2INT(Tcl_GetHashValue(hPtr)) != i+42) {
+	    Tcl_SetObjResult(interp, Tcl_NewIntObj(i));
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp)," value problem",-1);
+	    Tcl_DeleteHashTable(&hash);
+	    return TCL_ERROR;
+	}
+	Tcl_DeleteHashEntry(hPtr);
+    }
+
+    if (hash.numEntries != 0) {
+	Tcl_AppendResult(interp, "non-zero final size", NULL);
+	Tcl_DeleteHashTable(&hash);
+	return TCL_ERROR;
+    }
+
+    Tcl_DeleteHashTable(&hash);
+    Tcl_AppendResult(interp, "OK", NULL);
+    return TCL_OK;
+}
+
+/*
+ * Used for testing Tcl_GetInt which is no longer used directly by the
+ * core very much.
+ */
+static int
+TestgetintCmd(
+    ClientData dummy,
+    Tcl_Interp *interp,
+    int argc,
+    const char **argv)
+{
+    if (argc < 2) {
+	Tcl_AppendResult(interp, "wrong # args", NULL);
+	return TCL_ERROR;
+    } else {
+	int val, i, total=0;
+
+	for (i=1 ; i<argc ; i++) {
+	    if (Tcl_GetInt(interp, argv[i], &val) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    total += val;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(total));
+	return TCL_OK;
+    }
+}
+
+static int
+NREUnwind_callback(
+    ClientData data[],
+    Tcl_Interp *interp,
+    int result)
+{
+    int none;
+
+    if (data[0] == INT2PTR(-1)) {
+        Tcl_NRAddCallback(interp, NREUnwind_callback, &none, INT2PTR(-1),
+                INT2PTR(-1), NULL);
+    } else if (data[1] == INT2PTR(-1)) {
+        Tcl_NRAddCallback(interp, NREUnwind_callback, data[0], &none,
+                INT2PTR(-1), NULL);
+    } else if (data[2] == INT2PTR(-1)) {
+        Tcl_NRAddCallback(interp, NREUnwind_callback, data[0], data[1],
+                &none, NULL);
+    } else {
+        Tcl_Obj *idata[3];
+        idata[0] = Tcl_NewIntObj((int) ((char *) data[1] - (char *) data[0]));
+        idata[1] = Tcl_NewIntObj((int) ((char *) data[2] - (char *) data[0]));
+        idata[2] = Tcl_NewIntObj((int) ((char *) &none   - (char *) data[0]));
+        Tcl_SetObjResult(interp, Tcl_NewListObj(3, idata));
+    }
+    return TCL_OK;
+}
+
+static int
+TestNREUnwind(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    /*
+     * Insure that callbacks effectively run at the proper level during the
+     * unwinding of the NRE stack.
+     */
+
+    Tcl_NRAddCallback(interp, NREUnwind_callback, INT2PTR(-1), INT2PTR(-1),
+            INT2PTR(-1), NULL);
+    return TCL_OK;
+}
+
+
+static int
+TestNRELevels(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    static ptrdiff_t *refDepth = NULL;
+    ptrdiff_t depth;
+    Tcl_Obj *levels[6];
+    int i = 0;
+    NRE_callback *cbPtr = iPtr->execEnvPtr->callbackPtr;
+
+    if (refDepth == NULL) {
+	refDepth = &depth;
+    }
+
+    depth = (refDepth - &depth);
+
+    levels[0] = Tcl_NewIntObj(depth);
+    levels[1] = Tcl_NewIntObj(iPtr->numLevels);
+    levels[2] = Tcl_NewIntObj(iPtr->cmdFramePtr->level);
+    levels[3] = Tcl_NewIntObj(iPtr->varFramePtr->level);
+    levels[4] = Tcl_NewIntObj(iPtr->execEnvPtr->execStackPtr->tosPtr
+	    - iPtr->execEnvPtr->execStackPtr->stackWords);
+
+    while (cbPtr) {
+	i++;
+	cbPtr = cbPtr->nextPtr;
+    }
+    levels[5] = Tcl_NewIntObj(i);
+
+    Tcl_SetObjResult(interp, Tcl_NewListObj(6, levels));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestconcatobjCmd --
+ *
+ *	This procedure implements the "testconcatobj" command. It is used
+ *	to test that Tcl_ConcatObj does indeed return a fresh Tcl_Obj in all
+ *	cases and thet it never corrupts its arguments. In other words, that
+ *	[Bug 1447328] was fixed properly.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestconcatobjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int argc,			/* Number of arguments. */
+    const char **argv)		/* Argument strings. */
+{
+    Tcl_Obj *list1Ptr, *list2Ptr, *emptyPtr, *concatPtr, *tmpPtr;
+    int result = TCL_OK, len;
+    Tcl_Obj *objv[3];
+
+    /*
+     * Set the start of the error message as obj result; it will be cleared at
+     * the end if no errors were found.
+     */
+
+    Tcl_SetObjResult(interp,
+	    Tcl_NewStringObj("Tcl_ConcatObj is unsafe:", -1));
+
+    emptyPtr = Tcl_NewObj();
+
+    list1Ptr = Tcl_NewStringObj("foo bar sum", -1);
+    Tcl_ListObjLength(NULL, list1Ptr, &len);
+    if (list1Ptr->bytes != NULL) {
+	ckfree(list1Ptr->bytes);
+	list1Ptr->bytes = NULL;
+    }
+
+    list2Ptr = Tcl_NewStringObj("eeny meeny", -1);
+    Tcl_ListObjLength(NULL, list2Ptr, &len);
+    if (list2Ptr->bytes != NULL) {
+	ckfree(list2Ptr->bytes);
+	list2Ptr->bytes = NULL;
+    }
+
+    /*
+     * Verify that concat'ing a list obj with one or more empty strings does
+     * return a fresh Tcl_Obj (see also [Bug 2055782]).
+     */
+
+    tmpPtr = Tcl_DuplicateObj(list1Ptr);
+
+    objv[0] = tmpPtr;
+    objv[1] = emptyPtr;
+    concatPtr = Tcl_ConcatObj(2, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (a) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (a) concatObj is not a new obj ",
+		NULL);
+	switch (tmpPtr->refCount) {
+	case 0:
+	    Tcl_AppendResult(interp, "(no new refCount)", NULL);
+	    break;
+	case 1:
+	    Tcl_AppendResult(interp, "(refCount added)", NULL);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(more than one refCount added!)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[0] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    Tcl_IncrRefCount(tmpPtr);
+    concatPtr = Tcl_ConcatObj(2, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (b) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (b) concatObj is not a new obj ",
+		NULL);
+	switch (tmpPtr->refCount) {
+	case 0:
+	    Tcl_AppendResult(interp, "(refCount removed?)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	    break;
+	case 1:
+	    Tcl_AppendResult(interp, "(no new refCount)", NULL);
+	    break;
+	case 2:
+	    Tcl_AppendResult(interp, "(refCount added)", NULL);
+	    Tcl_DecrRefCount(tmpPtr);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(more than one refCount added!)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[0] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    objv[0] = emptyPtr;
+    objv[1] = tmpPtr;
+    objv[2] = emptyPtr;
+    concatPtr = Tcl_ConcatObj(3, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (c) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (c) concatObj is not a new obj ",
+		NULL);
+	switch (tmpPtr->refCount) {
+	case 0:
+	    Tcl_AppendResult(interp, "(no new refCount)", NULL);
+	    break;
+	case 1:
+	    Tcl_AppendResult(interp, "(refCount added)", NULL);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(more than one refCount added!)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[1] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    Tcl_IncrRefCount(tmpPtr);
+    concatPtr = Tcl_ConcatObj(3, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (d) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (d) concatObj is not a new obj ",
+		NULL);
+	switch (tmpPtr->refCount) {
+	case 0:
+	    Tcl_AppendResult(interp, "(refCount removed?)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	    break;
+	case 1:
+	    Tcl_AppendResult(interp, "(no new refCount)", NULL);
+	    break;
+	case 2:
+	    Tcl_AppendResult(interp, "(refCount added)", NULL);
+	    Tcl_DecrRefCount(tmpPtr);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(more than one refCount added!)", NULL);
+	    Tcl_Panic("extremely unsafe behaviour by Tcl_ConcatObj()");
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[1] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    /*
+     * Verify that an unshared list is not corrupted when concat'ing things to
+     * it.
+     */
+
+    objv[0] = tmpPtr;
+    objv[1] = list2Ptr;
+    concatPtr = Tcl_ConcatObj(2, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (e) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	int len;
+
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (e) concatObj is not a new obj ",
+		NULL);
+
+	(void) Tcl_ListObjLength(NULL, concatPtr, &len);
+	switch (tmpPtr->refCount) {
+	case 3:
+	    Tcl_AppendResult(interp, "(failed to concat)", NULL);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(corrupted input!)", NULL);
+	}
+	if (Tcl_IsShared(tmpPtr)) {
+	    Tcl_DecrRefCount(tmpPtr);
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[0] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    objv[0] = tmpPtr;
+    objv[1] = list2Ptr;
+    Tcl_IncrRefCount(tmpPtr);
+    concatPtr = Tcl_ConcatObj(2, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (f) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	int len;
+
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (f) concatObj is not a new obj ",
+		NULL);
+
+	(void) Tcl_ListObjLength(NULL, concatPtr, &len);
+	switch (tmpPtr->refCount) {
+	case 3:
+	    Tcl_AppendResult(interp, "(failed to concat)", NULL);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(corrupted input!)", NULL);
+	}
+	if (Tcl_IsShared(tmpPtr)) {
+	    Tcl_DecrRefCount(tmpPtr);
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[0] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    objv[0] = tmpPtr;
+    objv[1] = list2Ptr;
+    Tcl_IncrRefCount(tmpPtr);
+    Tcl_IncrRefCount(tmpPtr);
+    concatPtr = Tcl_ConcatObj(2, objv);
+    if (concatPtr->refCount != 0) {
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp,
+		"\n\t* (g) concatObj does not have refCount 0", NULL);
+    }
+    if (concatPtr == tmpPtr) {
+	int len;
+
+	result = TCL_ERROR;
+	Tcl_AppendResult(interp, "\n\t* (g) concatObj is not a new obj ",
+		NULL);
+
+	(void) Tcl_ListObjLength(NULL, concatPtr, &len);
+	switch (tmpPtr->refCount) {
+	case 3:
+	    Tcl_AppendResult(interp, "(failed to concat)", NULL);
+	    break;
+	default:
+	    Tcl_AppendResult(interp, "(corrupted input!)", NULL);
+	}
+	Tcl_DecrRefCount(tmpPtr);
+	if (Tcl_IsShared(tmpPtr)) {
+	    Tcl_DecrRefCount(tmpPtr);
+	}
+	tmpPtr = Tcl_DuplicateObj(list1Ptr);
+	objv[0] = tmpPtr;
+    }
+    Tcl_DecrRefCount(concatPtr);
+
+    /*
+     * Clean everything up. Note that we don't actually know how many
+     * references there are to tmpPtr here; in the no-error case, it should be
+     * five... [Bug 2895367]
+     */
+
+    Tcl_DecrRefCount(list1Ptr);
+    Tcl_DecrRefCount(list2Ptr);
+    Tcl_DecrRefCount(emptyPtr);
+    while (tmpPtr->refCount > 1) {
+	Tcl_DecrRefCount(tmpPtr);
+    }
+    Tcl_DecrRefCount(tmpPtr);
+
+    if (result == TCL_OK) {
+	Tcl_ResetResult(interp);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestparseargsCmd --
+ *
+ *	This procedure implements the "testparseargs" command. It is used to
+ *	test that Tcl_ParseArgsObjv does indeed return the right number of
+ *	arguments. In other words, that [Bug 3413857] was fixed properly.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestparseargsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Arguments. */
+{
+    static int foo = 0;
+    int count = objc;
+    Tcl_Obj **remObjv, *result[3];
+    Tcl_ArgvInfo argTable[] = {
+        {TCL_ARGV_CONSTANT, "-bool", INT2PTR(1), &foo, "booltest", NULL},
+        TCL_ARGV_AUTO_REST, TCL_ARGV_AUTO_HELP, TCL_ARGV_TABLE_END
+    };
+
+    foo = 0;
+    if (Tcl_ParseArgsObjv(interp, argTable, &count, objv, &remObjv)!=TCL_OK) {
+        return TCL_ERROR;
+    }
+    result[0] = Tcl_NewIntObj(foo);
+    result[1] = Tcl_NewIntObj(count);
+    result[2] = Tcl_NewListObj(count, remObjv);
+    Tcl_SetObjResult(interp, Tcl_NewListObj(3, result));
+    ckfree(remObjv);
+    return TCL_OK;
+}
+
+/**
+ * Test harness for command and variable resolvers.
+ */
+
+static int
+InterpCmdResolver(
+    Tcl_Interp *interp,
+    const char *name,
+    Tcl_Namespace *context,
+    int flags,
+    Tcl_Command *rPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *varFramePtr = iPtr->varFramePtr;
+    Proc *procPtr = (varFramePtr->isProcCallFrame & FRAME_IS_PROC) ?
+            varFramePtr->procPtr : NULL;
+    Namespace *callerNsPtr = varFramePtr->nsPtr;
+    Tcl_Command resolvedCmdPtr = NULL;
+
+    /*
+     * Just do something special on a cmd literal "z" in two cases:
+     *  A)  when the caller is a proc "x", and the proc is either in "::" or in "::ns2".
+     *  B) the caller's namespace is "ctx1" or "ctx2"
+     */
+    if ( (name[0] == 'z') && (name[1] == '\0') ) {
+        Namespace *ns2NsPtr = (Namespace *) Tcl_FindNamespace(interp, "::ns2", NULL, 0);
+
+        if (procPtr != NULL
+            && ((procPtr->cmdPtr->nsPtr == iPtr->globalNsPtr)
+                || (ns2NsPtr != NULL && procPtr->cmdPtr->nsPtr == ns2NsPtr)
+                )
+            ) {
+            /*
+             * Case A)
+             *
+             *    - The context, in which this resolver becomes active, is
+             *      determined by the name of the caller proc, which has to be
+             *      named "x".
+             *
+             *    - To determine the name of the caller proc, the proc is taken
+             *      from the topmost stack frame.
+             *
+             *    - Note that the context is NOT provided during byte-code
+             *      compilation (e.g. in TclProcCompileProc)
+             *
+             *   When these conditions hold, this function resolves the
+             *   passed-in cmd literal into a cmd "y", which is taken from the
+             *   the global namespace (for simplicity).
+             */
+
+            const char *callingCmdName =
+                Tcl_GetCommandName(interp, (Tcl_Command) procPtr->cmdPtr);
+
+            if ( callingCmdName[0] == 'x' && callingCmdName[1] == '\0' ) {
+                resolvedCmdPtr = Tcl_FindCommand(interp, "y", NULL, TCL_GLOBAL_ONLY);
+            }
+        } else if (callerNsPtr != NULL) {
+            /*
+             * Case B)
+             *
+             *    - The context, in which this resolver becomes active, is
+             *      determined by the name of the parent namespace, which has
+             *      to be named "ctx1" or "ctx2".
+             *
+             *    - To determine the name of the parent namesace, it is taken
+             *      from the 2nd highest stack frame.
+             *
+             *    - Note that the context can be provided during byte-code
+             *      compilation (e.g. in TclProcCompileProc)
+             *
+             *   When these conditions hold, this function resolves the
+             *   passed-in cmd literal into a cmd "y" or "Y" depending on the
+             *   context. The resolved procs are taken from the the global
+             *   namespace (for simplicity).
+             */
+
+            CallFrame *parentFramePtr = varFramePtr->callerPtr;
+            const char *context = parentFramePtr != NULL ? parentFramePtr->nsPtr->name : "(NULL)";
+
+            if (strcmp(context, "ctx1") == 0 && (name[0] == 'z') && (name[1] == '\0')) {
+                resolvedCmdPtr = Tcl_FindCommand(interp, "y", NULL, TCL_GLOBAL_ONLY);
+                /* fprintf(stderr, "... y ==> %p\n", resolvedCmdPtr);*/
+
+            } else if (strcmp(context, "ctx2") == 0 && (name[0] == 'z') && (name[1] == '\0')) {
+                resolvedCmdPtr = Tcl_FindCommand(interp, "Y", NULL, TCL_GLOBAL_ONLY);
+                /*fprintf(stderr, "... Y ==> %p\n", resolvedCmdPtr);*/
+            }
+        }
+
+        if (resolvedCmdPtr != NULL) {
+            *rPtr = resolvedCmdPtr;
+            return TCL_OK;
+        }
+    }
+    return TCL_CONTINUE;
+}
+
+static int
+InterpVarResolver(
+    Tcl_Interp *interp,
+    const char *name,
+    Tcl_Namespace *context,
+    int flags,
+    Tcl_Var *rPtr)
+{
+    /*
+     * Don't resolve the variable; use standard rules.
+     */
+
+    return TCL_CONTINUE;
+}
+
+typedef struct MyResolvedVarInfo {
+    Tcl_ResolvedVarInfo vInfo;  /* This must be the first element. */
+    Tcl_Var var;
+    Tcl_Obj *nameObj;
+} MyResolvedVarInfo;
+
+static inline void
+HashVarFree(
+    Tcl_Var var)
+{
+    if (VarHashRefCount(var) < 2) {
+        ckfree(var);
+    } else {
+        VarHashRefCount(var)--;
+    }
+}
+
+static void
+MyCompiledVarFree(
+    Tcl_ResolvedVarInfo *vInfoPtr)
+{
+    MyResolvedVarInfo *resVarInfo = (MyResolvedVarInfo *) vInfoPtr;
+
+    Tcl_DecrRefCount(resVarInfo->nameObj);
+    if (resVarInfo->var) {
+        HashVarFree(resVarInfo->var);
+    }
+    ckfree(vInfoPtr);
+}
+
+#define TclVarHashGetValue(hPtr) \
+    ((Var *) ((char *)hPtr - TclOffset(VarInHash, entry)))
+
+static Tcl_Var
+MyCompiledVarFetch(
+    Tcl_Interp *interp,
+    Tcl_ResolvedVarInfo *vinfoPtr)
+{
+    MyResolvedVarInfo *resVarInfo = (MyResolvedVarInfo *) vinfoPtr;
+    Tcl_Var var = resVarInfo->var;
+    int isNewVar;
+    Interp *iPtr = (Interp *) interp;
+    Tcl_HashEntry *hPtr;
+
+    if (var != NULL) {
+        if (!(((Var *) var)->flags & VAR_DEAD_HASH)) {
+            /*
+             * The cached variable is valid, return it.
+             */
+
+            return var;
+        }
+
+        /*
+         * The variable is not valid anymore. Clean it up.
+         */
+
+        HashVarFree(var);
+    }
+
+    hPtr = Tcl_CreateHashEntry((Tcl_HashTable *) &iPtr->globalNsPtr->varTable,
+            (char *) resVarInfo->nameObj, &isNewVar);
+    if (hPtr) {
+        var = (Tcl_Var) TclVarHashGetValue(hPtr);
+    } else {
+        var = NULL;
+    }
+    resVarInfo->var = var;
+
+    /*
+     * Increment the reference counter to avoid ckfree() of the variable in
+     * Tcl's FreeVarEntry(); for cleanup, we provide our own HashVarFree();
+     */
+
+    VarHashRefCount(var)++;
+    return var;
+}
+
+static int
+InterpCompiledVarResolver(
+    Tcl_Interp *interp,
+    const char *name,
+    int length,
+    Tcl_Namespace *context,
+    Tcl_ResolvedVarInfo **rPtr)
+{
+    if (*name == 'T') {
+ 	MyResolvedVarInfo *resVarInfo = ckalloc(sizeof(MyResolvedVarInfo));
+
+ 	resVarInfo->vInfo.fetchProc = MyCompiledVarFetch;
+ 	resVarInfo->vInfo.deleteProc = MyCompiledVarFree;
+ 	resVarInfo->var = NULL;
+ 	resVarInfo->nameObj = Tcl_NewStringObj(name, -1);
+ 	Tcl_IncrRefCount(resVarInfo->nameObj);
+ 	*rPtr = &resVarInfo->vInfo;
+ 	return TCL_OK;
+    }
+    return TCL_CONTINUE;
+}
+
+static int
+TestInterpResolverCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const char *const table[] = {
+        "down", "up", NULL
+    };
+    int idx;
+#define RESOLVER_KEY "testInterpResolver"
+
+    if ((objc < 2) || (objc > 3)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "up|down ?interp?");
+	return TCL_ERROR;
+    }
+    if (objc == 3) {
+	interp = Tcl_GetSlave(interp, Tcl_GetString(objv[2]));
+	if (interp == NULL) {
+	    Tcl_AppendResult(interp, "provided interpreter not found", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], table, "operation", TCL_EXACT,
+            &idx) != TCL_OK) {
+        return TCL_ERROR;
+    }
+    switch (idx) {
+    case 1: /* up */
+        Tcl_AddInterpResolvers(interp, RESOLVER_KEY, InterpCmdResolver,
+                InterpVarResolver, InterpCompiledVarResolver);
+        break;
+    case 0: /*down*/
+        if (!Tcl_RemoveInterpResolvers(interp, RESOLVER_KEY)) {
+            Tcl_AppendResult(interp, "could not remove the resolver scheme",
+                    NULL);
+            return TCL_ERROR;
+        }
+    }
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclTestObj.c b/generic/tclTestObj.c
new file mode 100644
index 0000000..5627608
--- /dev/null
+++ b/generic/tclTestObj.c
@@ -0,0 +1,1526 @@
+/*
+ * tclTestObj.c --
+ *
+ *	This file contains C command functions for the additional Tcl commands
+ *	that are used for testing implementations of the Tcl object types.
+ *	These commands are not normally included in Tcl applications; they're
+ *	only used for testing.
+ *
+ * Copyright (c) 1995-1998 Sun Microsystems, Inc.
+ * Copyright (c) 1999 by Scriptics Corporation.
+ * Copyright (c) 2005 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef USE_TCL_STUBS
+#   define USE_TCL_STUBS
+#endif
+#include "tclInt.h"
+#include "tommath.h"
+#include "tclStringRep.h"
+
+
+/*
+ * Forward declarations for functions defined later in this file:
+ */
+
+static int		CheckIfVarUnset(Tcl_Interp *interp, Tcl_Obj **varPtr, int varIndex);
+static int		GetVariableIndex(Tcl_Interp *interp,
+			    const char *string, int *indexPtr);
+static void		SetVarToObj(Tcl_Obj **varPtr, int varIndex, Tcl_Obj *objPtr);
+static int		TestbignumobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestbooleanobjCmd(ClientData dummy,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		TestdoubleobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestindexobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestintobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int 		TestlistobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TestobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+static int		TeststringobjCmd(ClientData dummy, Tcl_Interp *interp,
+			    int objc, Tcl_Obj *const objv[]);
+
+#define VARPTR_KEY "TCLOBJTEST_VARPTR"
+#define NUMBER_OF_OBJECT_VARS 20
+
+static void VarPtrDeleteProc(ClientData clientData, Tcl_Interp *interp)
+{
+    register int i;
+    Tcl_Obj **varPtr = (Tcl_Obj **) clientData;
+    for (i = 0;  i < NUMBER_OF_OBJECT_VARS;  i++) {
+	if (varPtr[i]) Tcl_DecrRefCount(varPtr[i]);
+    }
+    Tcl_DeleteAssocData(interp, VARPTR_KEY);
+    ckfree(varPtr);
+}
+
+static Tcl_Obj **GetVarPtr(Tcl_Interp *interp)
+{
+    Tcl_InterpDeleteProc *proc;
+
+    return (Tcl_Obj **) Tcl_GetAssocData(interp, VARPTR_KEY, &proc);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjTest_Init --
+ *
+ *	This function creates additional commands that are used to test the
+ *	Tcl object support.
+ *
+ * Results:
+ *	Returns a standard Tcl completion code, and leaves an error
+ *	message in the interp's result if an error occurs.
+ *
+ * Side effects:
+ *	Creates and registers several new testing commands.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjTest_Init(
+    Tcl_Interp *interp)
+{
+    register int i;
+    /*
+     * An array of Tcl_Obj pointers used in the commands that operate on or get
+     * the values of Tcl object-valued variables. varPtr[i] is the i-th variable's
+     * Tcl_Obj *.
+     */
+    Tcl_Obj **varPtr;
+
+    varPtr = (Tcl_Obj **) ckalloc(NUMBER_OF_OBJECT_VARS *sizeof(varPtr[0]));
+    if (!varPtr) {
+	return TCL_ERROR;
+    }
+    Tcl_SetAssocData(interp, VARPTR_KEY, VarPtrDeleteProc, varPtr);
+    for (i = 0;  i < NUMBER_OF_OBJECT_VARS;  i++) {
+	varPtr[i] = NULL;
+    }
+
+    Tcl_CreateObjCommand(interp, "testbignumobj", TestbignumobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testbooleanobj", TestbooleanobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testdoubleobj", TestdoubleobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testintobj", TestintobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testindexobj", TestindexobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testlistobj", TestlistobjCmd,
+	    NULL, NULL);
+    Tcl_CreateObjCommand(interp, "testobj", TestobjCmd, NULL, NULL);
+    Tcl_CreateObjCommand(interp, "teststringobj", TeststringobjCmd,
+	    NULL, NULL);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestbignumobjCmd --
+ *
+ *	This function implmenets the "testbignumobj" command.  It is used
+ *	to exercise the bignum Tcl object type implementation.
+ *
+ * Results:
+ *	Returns a standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees bignum objects; converts objects to have bignum
+ *	type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestbignumobjCmd(
+    ClientData clientData,	/* unused */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Argument count */
+    Tcl_Obj *const objv[])	/* Argument vector */
+{
+    const char *const subcmds[] = {
+	"set", "get", "mult10", "div10", "iseven", "radixsize", NULL
+    };
+    enum options {
+	BIGNUM_SET, BIGNUM_GET, BIGNUM_MULT10, BIGNUM_DIV10, BIGNUM_ISEVEN,
+	BIGNUM_RADIXSIZE
+    };
+    int index, varIndex;
+    const char *string;
+    mp_int bignumValue, newValue;
+    Tcl_Obj **varPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], subcmds, "option", 0,
+	    &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    string = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, string, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    varPtr = GetVarPtr(interp);
+
+    switch (index) {
+    case BIGNUM_SET:
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "var value");
+	    return TCL_ERROR;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (mp_init(&bignumValue) != MP_OKAY) {
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("error in mp_init", -1));
+	    return TCL_ERROR;
+	}
+	if (mp_read_radix(&bignumValue, string, 10) != MP_OKAY) {
+	    mp_clear(&bignumValue);
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("error in mp_read_radix", -1));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * If the object currently bound to the variable with index varIndex
+	 * has ref count 1 (i.e. the object is unshared) we can modify that
+	 * object directly.  Otherwise, if RC>1 (i.e. the object is shared),
+	 * we must create a new object to modify/set and decrement the old
+	 * formerly-shared object's ref count. This is "copy on write".
+	 */
+
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetBignumObj(varPtr[varIndex], &bignumValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewBignumObj(&bignumValue));
+	}
+	break;
+
+    case BIGNUM_GET:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	break;
+
+    case BIGNUM_MULT10:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBignumFromObj(interp, varPtr[varIndex],
+		&bignumValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (mp_init(&newValue) != MP_OKAY
+		|| (mp_mul_d(&bignumValue, 10, &newValue) != MP_OKAY)) {
+	    mp_clear(&bignumValue);
+	    mp_clear(&newValue);
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("error in mp_mul_d", -1));
+	    return TCL_ERROR;
+	}
+	mp_clear(&bignumValue);
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetBignumObj(varPtr[varIndex], &newValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewBignumObj(&newValue));
+	}
+	break;
+
+    case BIGNUM_DIV10:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBignumFromObj(interp, varPtr[varIndex],
+		&bignumValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (mp_init(&newValue) != MP_OKAY
+		|| (mp_div_d(&bignumValue, 10, &newValue, NULL) != MP_OKAY)) {
+	    mp_clear(&bignumValue);
+	    mp_clear(&newValue);
+	    Tcl_SetObjResult(interp,
+		    Tcl_NewStringObj("error in mp_div_d", -1));
+	    return TCL_ERROR;
+	}
+	mp_clear(&bignumValue);
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetBignumObj(varPtr[varIndex], &newValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewBignumObj(&newValue));
+	}
+	break;
+
+    case BIGNUM_ISEVEN:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBignumFromObj(interp, varPtr[varIndex],
+		&bignumValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], mp_iseven(&bignumValue));
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(mp_iseven(&bignumValue)));
+	}
+	mp_clear(&bignumValue);
+	break;
+
+    case BIGNUM_RADIXSIZE:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBignumFromObj(interp, varPtr[varIndex],
+		&bignumValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (mp_radix_size(&bignumValue, 10, &index) != MP_OKAY) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], index);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(index));
+	}
+	mp_clear(&bignumValue);
+	break;
+    }
+
+    Tcl_SetObjResult(interp, varPtr[varIndex]);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestbooleanobjCmd --
+ *
+ *	This function implements the "testbooleanobj" command.  It is used to
+ *	test the boolean Tcl object type implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees boolean objects, and also converts objects to
+ *	have boolean type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestbooleanobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int varIndex, boolValue;
+    const char *index, *subCmd;
+    Tcl_Obj **varPtr;
+
+    if (objc < 3) {
+	wrongNumArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    index = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    varPtr = GetVarPtr(interp);
+
+    subCmd = Tcl_GetString(objv[1]);
+    if (strcmp(subCmd, "set") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_GetBooleanFromObj(interp, objv[3], &boolValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * If the object currently bound to the variable with index varIndex
+	 * has ref count 1 (i.e. the object is unshared) we can modify that
+	 * object directly. Otherwise, if RC>1 (i.e. the object is shared),
+	 * we must create a new object to modify/set and decrement the old
+	 * formerly-shared object's ref count. This is "copy on write".
+	 */
+
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetBooleanObj(varPtr[varIndex], boolValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewBooleanObj(boolValue));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "get") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "not") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetBooleanFromObj(interp, varPtr[varIndex],
+				  &boolValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetBooleanObj(varPtr[varIndex], !boolValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewBooleanObj(!boolValue));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"bad option \"", Tcl_GetString(objv[1]),
+		"\": must be set, get, or not", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestdoubleobjCmd --
+ *
+ *	This function implements the "testdoubleobj" command.  It is used to
+ *	test the double-precision floating point Tcl object type
+ *	implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees double objects, and also converts objects to
+ *	have double type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestdoubleobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int varIndex;
+    double doubleValue;
+    const char *index, *subCmd, *string;
+    Tcl_Obj **varPtr;
+
+    if (objc < 3) {
+	wrongNumArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    varPtr = GetVarPtr(interp);
+
+    index = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    subCmd = Tcl_GetString(objv[1]);
+    if (strcmp(subCmd, "set") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (Tcl_GetDouble(interp, string, &doubleValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * If the object currently bound to the variable with index varIndex
+	 * has ref count 1 (i.e. the object is unshared) we can modify that
+	 * object directly. Otherwise, if RC>1 (i.e. the object is shared), we
+	 * must create a new object to modify/set and decrement the old
+	 * formerly-shared object's ref count. This is "copy on write".
+	 */
+
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetDoubleObj(varPtr[varIndex], doubleValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewDoubleObj(doubleValue));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "get") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "mult10") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetDoubleFromObj(interp, varPtr[varIndex],
+				 &doubleValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetDoubleObj(varPtr[varIndex], doubleValue * 10.0);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewDoubleObj(doubleValue * 10.0));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "div10") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetDoubleFromObj(interp, varPtr[varIndex],
+		&doubleValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetDoubleObj(varPtr[varIndex], doubleValue / 10.0);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewDoubleObj(doubleValue / 10.0));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"bad option \"", Tcl_GetString(objv[1]),
+		"\": must be set, get, mult10, or div10", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestindexobjCmd --
+ *
+ *	This function implements the "testindexobj" command. It is used to
+ *	test the index Tcl object type implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees int objects, and also converts objects to
+ *	have int type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestindexobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int allowAbbrev, index, index2, setError, i, result;
+    const char **argv;
+    static const char *const tablePtr[] = {"a", "b", "check", NULL};
+    /*
+     * Keep this structure declaration in sync with tclIndexObj.c
+     */
+    struct IndexRep {
+	void *tablePtr;		/* Pointer to the table of strings. */
+	int offset;		/* Offset between table entries. */
+	int index;		/* Selected index into table. */
+    };
+    struct IndexRep *indexRep;
+
+    if ((objc == 3) && (strcmp(Tcl_GetString(objv[1]),
+	    "check") == 0)) {
+	/*
+	 * This code checks to be sure that the results of Tcl_GetIndexFromObj
+	 * are properly cached in the object and returned on subsequent
+	 * lookups.
+	 */
+
+	if (Tcl_GetIntFromObj(interp, objv[2], &index2) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	Tcl_GetIndexFromObj(NULL, objv[1], tablePtr, "token", 0, &index);
+	indexRep = objv[1]->internalRep.twoPtrValue.ptr1;
+	indexRep->index = index2;
+	result = Tcl_GetIndexFromObj(NULL, objv[1],
+		tablePtr, "token", 0, &index);
+	if (result == TCL_OK) {
+	    Tcl_SetIntObj(Tcl_GetObjResult(interp), index);
+	}
+	return result;
+    }
+
+    if (objc < 5) {
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), "wrong # args", -1);
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetBooleanFromObj(interp, objv[1], &setError) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetBooleanFromObj(interp, objv[2], &allowAbbrev) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    argv = ckalloc((objc-3) * sizeof(char *));
+    for (i = 4; i < objc; i++) {
+	argv[i-4] = Tcl_GetString(objv[i]);
+    }
+    argv[objc-4] = NULL;
+
+    result = Tcl_GetIndexFromObj((setError? interp : NULL), objv[3],
+	    argv, "token", INDEX_TEMP_TABLE|(allowAbbrev? 0 : TCL_EXACT),
+	    &index);
+    ckfree(argv);
+    if (result == TCL_OK) {
+	Tcl_SetIntObj(Tcl_GetObjResult(interp), index);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestintobjCmd --
+ *
+ *	This function implements the "testintobj" command. It is used to
+ *	test the int Tcl object type implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees int objects, and also converts objects to
+ *	have int type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestintobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int intValue, varIndex, i;
+    long longValue;
+    const char *index, *subCmd, *string;
+    Tcl_Obj **varPtr;
+
+    if (objc < 3) {
+	wrongNumArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    varPtr = GetVarPtr(interp);
+    index = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    subCmd = Tcl_GetString(objv[1]);
+    if (strcmp(subCmd, "set") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (Tcl_GetInt(interp, string, &i) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	intValue = i;
+
+	/*
+	 * If the object currently bound to the variable with index varIndex
+	 * has ref count 1 (i.e. the object is unshared) we can modify that
+	 * object directly. Otherwise, if RC>1 (i.e. the object is shared), we
+	 * must create a new object to modify/set and decrement the old
+	 * formerly-shared object's ref count. This is "copy on write".
+	 */
+
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], intValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(intValue));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "set2") == 0) { /* doesn't set result */
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (Tcl_GetInt(interp, string, &i) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	intValue = i;
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], intValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(intValue));
+	}
+    } else if (strcmp(subCmd, "setlong") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (Tcl_GetInt(interp, string, &i) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	intValue = i;
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetLongObj(varPtr[varIndex], intValue);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewLongObj(intValue));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "setmaxlong") == 0) {
+	long maxLong = LONG_MAX;
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetLongObj(varPtr[varIndex], maxLong);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewLongObj(maxLong));
+	}
+    } else if (strcmp(subCmd, "ismaxlong") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetLongFromObj(interp, varPtr[varIndex], &longValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_AppendToObj(Tcl_GetObjResult(interp),
+		((longValue == LONG_MAX)? "1" : "0"), -1);
+    } else if (strcmp(subCmd, "get") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "get2") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	string = Tcl_GetString(varPtr[varIndex]);
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), string, -1);
+    } else if (strcmp(subCmd, "inttoobigtest") == 0) {
+	/*
+	 * If long ints have more bits than ints on this platform, verify that
+	 * Tcl_GetIntFromObj returns an error if the long int held in an
+	 * integer object's internal representation is too large to fit in an
+	 * int.
+	 */
+
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+#if (INT_MAX == LONG_MAX)   /* int is same size as long int */
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), "1", -1);
+#else
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetLongObj(varPtr[varIndex], LONG_MAX);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewLongObj(LONG_MAX));
+	}
+	if (Tcl_GetIntFromObj(interp, varPtr[varIndex], &i) != TCL_OK) {
+	    Tcl_ResetResult(interp);
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp), "1", -1);
+	    return TCL_OK;
+	}
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), "0", -1);
+#endif
+    } else if (strcmp(subCmd, "mult10") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIntFromObj(interp, varPtr[varIndex],
+		&intValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], intValue * 10);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(intValue * 10));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "div10") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIntFromObj(interp, varPtr[varIndex],
+		&intValue) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (!Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetIntObj(varPtr[varIndex], intValue / 10);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewIntObj(intValue / 10));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"bad option \"", Tcl_GetString(objv[1]),
+		"\": must be set, get, get2, mult10, or div10", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *-----------------------------------------------------------------------------
+ *
+ * TestlistobjCmd --
+ *
+ *	This function implements the 'testlistobj' command. It is used to
+ *	test a few possible corner cases in list object manipulation from
+ *	C code that cannot occur at the Tcl level.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates, manipulates and frees list objects.
+ *
+ *-----------------------------------------------------------------------------
+ */
+
+static int
+TestlistobjCmd(
+    ClientData clientData,	/* Not used */
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int objc,			/* Number of arguments */
+    Tcl_Obj *const objv[])	/* Argument objects */
+{
+    /* Subcommands supported by this command */
+    const char* subcommands[] = {
+	"set",
+	"get",
+	"replace"
+    };
+    enum listobjCmdIndex {
+	LISTOBJ_SET,
+	LISTOBJ_GET,
+	LISTOBJ_REPLACE
+    };
+
+    const char* index;		/* Argument giving the variable number */
+    int varIndex;		/* Variable number converted to binary */
+    int cmdIndex;		/* Ordinal number of the subcommand */
+    int first;			/* First index in the list */
+    int count;			/* Count of elements in a list */
+    Tcl_Obj **varPtr;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg...?");
+	return TCL_ERROR;
+    }
+    varPtr = GetVarPtr(interp);
+    index = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], subcommands, "command",
+			    0, &cmdIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch(cmdIndex) {
+    case LISTOBJ_SET:
+	if ((varPtr[varIndex] != NULL) && !Tcl_IsShared(varPtr[varIndex])) {
+	    Tcl_SetListObj(varPtr[varIndex], objc-3, objv+3);
+	} else {
+	    SetVarToObj(varPtr, varIndex, Tcl_NewListObj(objc-3, objv+3));
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+	break;
+
+    case LISTOBJ_GET:
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "varIndex");
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+	break;
+
+    case LISTOBJ_REPLACE:
+	if (objc < 5) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+			     "varIndex start count ?element...?");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIntFromObj(interp, objv[3], &first) != TCL_OK
+	    || Tcl_GetIntFromObj(interp, objv[4], &count) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (Tcl_IsShared(varPtr[varIndex])) {
+	    SetVarToObj(varPtr, varIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	}
+	Tcl_ResetResult(interp);
+	return Tcl_ListObjReplace(interp, varPtr[varIndex], first, count,
+				  objc-5, objv+5);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TestobjCmd --
+ *
+ *	This function implements the "testobj" command. It is used to test
+ *	the type-independent portions of the Tcl object type implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees objects.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TestobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int varIndex, destIndex, i;
+    const char *index, *subCmd, *string;
+    const Tcl_ObjType *targetType;
+    Tcl_Obj **varPtr;
+
+    if (objc < 2) {
+	wrongNumArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    varPtr = GetVarPtr(interp);
+    subCmd = Tcl_GetString(objv[1]);
+    if (strcmp(subCmd, "assign") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (GetVariableIndex(interp, string, &destIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	SetVarToObj(varPtr, destIndex, varPtr[varIndex]);
+	Tcl_SetObjResult(interp, varPtr[destIndex]);
+    } else if (strcmp(subCmd, "bug3598580") == 0) {
+	Tcl_Obj *listObjPtr, *elemObjPtr;
+	if (objc != 2) {
+	    goto wrongNumArgs;
+	}
+	elemObjPtr = Tcl_NewIntObj(123);
+	listObjPtr = Tcl_NewListObj(1, &elemObjPtr);
+	/* Replace the single list element through itself, nonsense but legal. */
+	Tcl_ListObjReplace(interp, listObjPtr, 0, 1, 1, &elemObjPtr);
+	Tcl_SetObjResult(interp, listObjPtr);
+	return TCL_OK;
+    } else if (strcmp(subCmd, "convert") == 0) {
+	const char *typeName;
+
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	typeName = Tcl_GetString(objv[3]);
+	if ((targetType = Tcl_GetObjType(typeName)) == NULL) {
+	    Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		    "no type ", typeName, " found", NULL);
+	    return TCL_ERROR;
+	}
+	if (Tcl_ConvertToType(interp, varPtr[varIndex], targetType)
+		!= TCL_OK) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "duplicate") == 0) {
+	if (objc != 4) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	string = Tcl_GetString(objv[3]);
+	if (GetVariableIndex(interp, string, &destIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	SetVarToObj(varPtr, destIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	Tcl_SetObjResult(interp, varPtr[destIndex]);
+    } else if (strcmp(subCmd, "freeallvars") == 0) {
+	if (objc != 2) {
+	    goto wrongNumArgs;
+	}
+	for (i = 0;  i < NUMBER_OF_OBJECT_VARS;  i++) {
+	    if (varPtr[i] != NULL) {
+		Tcl_DecrRefCount(varPtr[i]);
+		varPtr[i] = NULL;
+	    }
+	}
+    } else if (strcmp(subCmd, "invalidateStringRep") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_InvalidateStringRep(varPtr[varIndex]);
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "newobj") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	SetVarToObj(varPtr, varIndex, Tcl_NewObj());
+	Tcl_SetObjResult(interp, varPtr[varIndex]);
+    } else if (strcmp(subCmd, "objtype") == 0) {
+	const char *typeName;
+
+	/*
+	 * Return an object containing the name of the argument's type of
+	 * internal rep. If none exists, return "none".
+	 */
+
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	if (objv[2]->typePtr == NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj("none", -1));
+	} else {
+	    typeName = objv[2]->typePtr->name;
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(typeName, -1));
+	}
+    } else if (strcmp(subCmd, "refcount") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj(varPtr[varIndex]->refCount));
+    } else if (strcmp(subCmd, "type") == 0) {
+	if (objc != 3) {
+	    goto wrongNumArgs;
+	}
+	index = Tcl_GetString(objv[2]);
+	if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+	    return TCL_ERROR;
+	}
+	if (varPtr[varIndex]->typePtr == NULL) { /* a string! */
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp), "string", -1);
+	} else {
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp),
+		    varPtr[varIndex]->typePtr->name, -1);
+	}
+    } else if (strcmp(subCmd, "types") == 0) {
+	if (objc != 2) {
+	    goto wrongNumArgs;
+	}
+	if (Tcl_AppendAllObjTypes(interp,
+		Tcl_GetObjResult(interp)) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    } else {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"bad option \"", Tcl_GetString(objv[1]),
+		"\": must be assign, convert, duplicate, freeallvars, "
+		"newobj, objcount, objtype, refcount, type, or types", NULL);
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TeststringobjCmd --
+ *
+ *	This function implements the "teststringobj" command. It is used to
+ *	test the string Tcl object type implementation.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	Creates and frees string objects, and also converts objects to
+ *	have string type.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TeststringobjCmd(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_UniChar *unicode;
+    int varIndex, option, i, length;
+#define MAX_STRINGS 11
+    const char *index, *string, *strings[MAX_STRINGS+1];
+    String *strPtr;
+    Tcl_Obj **varPtr;
+    static const char *const options[] = {
+	"append", "appendstrings", "get", "get2", "length", "length2",
+	"set", "set2", "setlength", "maxchars", "getunicode",
+	"appendself", "appendself2", NULL
+    };
+
+    if (objc < 3) {
+	wrongNumArgs:
+	Tcl_WrongNumArgs(interp, 1, objv, "option arg ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    varPtr = GetVarPtr(interp);
+    index = Tcl_GetString(objv[2]);
+    if (GetVariableIndex(interp, index, &varIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], options, "option", 0, &option)
+	    != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch (option) {
+	case 0:				/* append */
+	    if (objc != 5) {
+		goto wrongNumArgs;
+	    }
+	    if (Tcl_GetIntFromObj(interp, objv[4], &length) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (varPtr[varIndex] == NULL) {
+		SetVarToObj(varPtr, varIndex, Tcl_NewObj());
+	    }
+
+	    /*
+	     * If the object bound to variable "varIndex" is shared, we must
+	     * "copy on write" and append to a copy of the object.
+	     */
+
+	    if (Tcl_IsShared(varPtr[varIndex])) {
+		SetVarToObj(varPtr, varIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	    }
+	    string = Tcl_GetString(objv[3]);
+	    Tcl_AppendToObj(varPtr[varIndex], string, length);
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+	case 1:				/* appendstrings */
+	    if (objc > (MAX_STRINGS+3)) {
+		goto wrongNumArgs;
+	    }
+	    if (varPtr[varIndex] == NULL) {
+		SetVarToObj(varPtr, varIndex, Tcl_NewObj());
+	    }
+
+	    /*
+	     * If the object bound to variable "varIndex" is shared, we must
+	     * "copy on write" and append to a copy of the object.
+	     */
+
+	    if (Tcl_IsShared(varPtr[varIndex])) {
+		SetVarToObj(varPtr, varIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	    }
+	    for (i = 3;  i < objc;  i++) {
+		strings[i-3] = Tcl_GetString(objv[i]);
+	    }
+	    for ( ; i < 12 + 3; i++) {
+		strings[i - 3] = NULL;
+	    }
+	    Tcl_AppendStringsToObj(varPtr[varIndex], strings[0], strings[1],
+		    strings[2], strings[3], strings[4], strings[5],
+		    strings[6], strings[7], strings[8], strings[9],
+		    strings[10], strings[11]);
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+	case 2:				/* get */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    if (CheckIfVarUnset(interp, varPtr,varIndex)) {
+		return TCL_ERROR;
+	    }
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+	case 3:				/* get2 */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    if (CheckIfVarUnset(interp, varPtr, varIndex)) {
+		return TCL_ERROR;
+	    }
+	    string = Tcl_GetString(varPtr[varIndex]);
+	    Tcl_AppendToObj(Tcl_GetObjResult(interp), string, -1);
+	    break;
+	case 4:				/* length */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    Tcl_SetIntObj(Tcl_GetObjResult(interp), (varPtr[varIndex] != NULL)
+		    ? varPtr[varIndex]->length : -1);
+	    break;
+	case 5:				/* length2 */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    if (varPtr[varIndex] != NULL) {
+		Tcl_ConvertToType(NULL, varPtr[varIndex],
+			Tcl_GetObjType("string"));
+		strPtr = varPtr[varIndex]->internalRep.twoPtrValue.ptr1;
+		length = (int) strPtr->allocated;
+	    } else {
+		length = -1;
+	    }
+	    Tcl_SetIntObj(Tcl_GetObjResult(interp), length);
+	    break;
+	case 6:				/* set */
+	    if (objc != 4) {
+		goto wrongNumArgs;
+	    }
+
+	    /*
+	     * If the object currently bound to the variable with index
+	     * varIndex has ref count 1 (i.e. the object is unshared) we can
+	     * modify that object directly. Otherwise, if RC>1 (i.e. the
+	     * object is shared), we must create a new object to modify/set
+	     * and decrement the old formerly-shared object's ref count. This
+	     * is "copy on write".
+	     */
+
+	    string = Tcl_GetStringFromObj(objv[3], &length);
+	    if ((varPtr[varIndex] != NULL)
+		    && !Tcl_IsShared(varPtr[varIndex])) {
+		Tcl_SetStringObj(varPtr[varIndex], string, length);
+	    } else {
+		SetVarToObj(varPtr, varIndex, Tcl_NewStringObj(string, length));
+	    }
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+	case 7:				/* set2 */
+	    if (objc != 4) {
+		goto wrongNumArgs;
+	    }
+	    SetVarToObj(varPtr, varIndex, objv[3]);
+	    break;
+	case 8:				/* setlength */
+	    if (objc != 4) {
+		goto wrongNumArgs;
+	    }
+	    if (Tcl_GetIntFromObj(interp, objv[3], &length) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (varPtr[varIndex] != NULL) {
+		Tcl_SetObjLength(varPtr[varIndex], length);
+	    }
+	    break;
+	case 9:				/* maxchars */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    if (varPtr[varIndex] != NULL) {
+		Tcl_ConvertToType(NULL, varPtr[varIndex],
+			Tcl_GetObjType("string"));
+		strPtr = varPtr[varIndex]->internalRep.twoPtrValue.ptr1;
+		length = strPtr->maxChars;
+	    } else {
+		length = -1;
+	    }
+	    Tcl_SetIntObj(Tcl_GetObjResult(interp), length);
+	    break;
+	case 10:			/* getunicode */
+	    if (objc != 3) {
+		goto wrongNumArgs;
+	    }
+	    Tcl_GetUnicodeFromObj(varPtr[varIndex], NULL);
+	    break;
+	case 11:			/* appendself */
+	    if (objc != 4) {
+		goto wrongNumArgs;
+	    }
+	    if (varPtr[varIndex] == NULL) {
+		SetVarToObj(varPtr, varIndex, Tcl_NewObj());
+	    }
+
+	    /*
+	     * If the object bound to variable "varIndex" is shared, we must
+	     * "copy on write" and append to a copy of the object.
+	     */
+
+	    if (Tcl_IsShared(varPtr[varIndex])) {
+		SetVarToObj(varPtr, varIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	    }
+
+	    string = Tcl_GetStringFromObj(varPtr[varIndex], &length);
+
+	    if (Tcl_GetIntFromObj(interp, objv[3], &i) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if ((i < 0) || (i > length)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"index value out of range", -1));
+		return TCL_ERROR;
+	    }
+
+	    Tcl_AppendToObj(varPtr[varIndex], string + i, length - i);
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+	case 12:			/* appendself2 */
+	    if (objc != 4) {
+		goto wrongNumArgs;
+	    }
+	    if (varPtr[varIndex] == NULL) {
+		SetVarToObj(varPtr, varIndex, Tcl_NewObj());
+	    }
+
+	    /*
+	     * If the object bound to variable "varIndex" is shared, we must
+	     * "copy on write" and append to a copy of the object.
+	     */
+
+	    if (Tcl_IsShared(varPtr[varIndex])) {
+		SetVarToObj(varPtr, varIndex, Tcl_DuplicateObj(varPtr[varIndex]));
+	    }
+
+	    unicode = Tcl_GetUnicodeFromObj(varPtr[varIndex], &length);
+
+	    if (Tcl_GetIntFromObj(interp, objv[3], &i) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if ((i < 0) || (i > length)) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"index value out of range", -1));
+		return TCL_ERROR;
+	    }
+
+	    Tcl_AppendUnicodeToObj(varPtr[varIndex], unicode + i, length - i);
+	    Tcl_SetObjResult(interp, varPtr[varIndex]);
+	    break;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetVarToObj --
+ *
+ *	Utility routine to assign a Tcl_Obj* to a test variable. The
+ *	Tcl_Obj* can be NULL.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	This routine handles ref counting details for assignment: i.e. the old
+ *	value's ref count must be decremented (if not NULL) and the new one
+ *	incremented (also if not NULL).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+SetVarToObj(
+    Tcl_Obj **varPtr,
+    int varIndex,		/* Designates the assignment variable. */
+    Tcl_Obj *objPtr)		/* Points to object to assign to var. */
+{
+    if (varPtr[varIndex] != NULL) {
+	Tcl_DecrRefCount(varPtr[varIndex]);
+    }
+    varPtr[varIndex] = objPtr;
+    if (objPtr != NULL) {
+	Tcl_IncrRefCount(objPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetVariableIndex --
+ *
+ *	Utility routine to get a test variable index from the command line.
+ *
+ * Results:
+ *	A standard Tcl object result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetVariableIndex(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. */
+    const char *string,		/* String containing a variable index
+				 * specified as a nonnegative number less than
+				 * NUMBER_OF_OBJECT_VARS. */
+    int *indexPtr)		/* Place to store converted result. */
+{
+    int index;
+
+    if (Tcl_GetInt(interp, string, &index) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (index < 0 || index >= NUMBER_OF_OBJECT_VARS) {
+	Tcl_ResetResult(interp);
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), "bad variable index", -1);
+	return TCL_ERROR;
+    }
+
+    *indexPtr = index;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CheckIfVarUnset --
+ *
+ *	Utility function that checks whether a test variable is readable:
+ *	i.e., that varPtr[varIndex] is non-NULL.
+ *
+ * Results:
+ *	1 if the test variable is unset (NULL); 0 otherwise.
+ *
+ * Side effects:
+ *	Sets the interpreter result to an error message if the variable is
+ *	unset (NULL).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CheckIfVarUnset(
+    Tcl_Interp *interp,		/* Interpreter for error reporting. */
+    Tcl_Obj ** varPtr,
+    int varIndex)		/* Index of the test variable to check. */
+{
+    if (varPtr[varIndex] == NULL) {
+	char buf[32 + TCL_INTEGER_SPACE];
+
+	sprintf(buf, "variable %d is unset (NULL)", varIndex);
+	Tcl_ResetResult(interp);
+	Tcl_AppendToObj(Tcl_GetObjResult(interp), buf, -1);
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclTestProcBodyObj.c b/generic/tclTestProcBodyObj.c
new file mode 100644
index 0000000..4d32c5a
--- /dev/null
+++ b/generic/tclTestProcBodyObj.c
@@ -0,0 +1,309 @@
+/*
+ * tclTestProcBodyObj.c --
+ *
+ *	Implements the "procbodytest" package, which contains commands to test
+ *	creation of Tcl procedures whose body argument is a Tcl_Obj of type
+ *	"procbody" rather than a string.
+ *
+ * Copyright (c) 1998 by Scriptics Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef USE_TCL_STUBS
+#   define USE_TCL_STUBS
+#endif
+#include "tclInt.h"
+
+/*
+ * name and version of this package
+ */
+
+static const char packageName[] = "procbodytest";
+static const char packageVersion[] = "1.0";
+
+/*
+ * Name of the commands exported by this package
+ */
+
+static const char procCommand[] = "proc";
+
+/*
+ * this struct describes an entry in the table of command names and command
+ * procs
+ */
+
+typedef struct CmdTable {
+    const char *cmdName;		/* command name */
+    Tcl_ObjCmdProc *proc;	/* command proc */
+    int exportIt;		/* if 1, export the command */
+} CmdTable;
+
+/*
+ * Declarations for functions defined in this file.
+ */
+
+static int	ProcBodyTestProcObjCmd(ClientData dummy,
+			Tcl_Interp *interp, int objc, Tcl_Obj *const objv[]);
+static int	ProcBodyTestInitInternal(Tcl_Interp *interp, int isSafe);
+static int	RegisterCommand(Tcl_Interp* interp,
+			const char *namespace, const CmdTable *cmdTablePtr);
+
+/*
+ * List of commands to create when the package is loaded; must go after the
+ * declarations of the enable command procedure.
+ */
+
+static const CmdTable commands[] = {
+    { procCommand,	ProcBodyTestProcObjCmd,	1 },
+    { 0, 0, 0 }
+};
+
+static const CmdTable safeCommands[] = {
+    { procCommand,	ProcBodyTestProcObjCmd,	1 },
+    { 0, 0, 0 }
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Procbodytest_Init --
+ *
+ *	This function initializes the "procbodytest" package.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Procbodytest_Init(
+    Tcl_Interp *interp)		/* the Tcl interpreter for which the package
+				 * is initialized */
+{
+    return ProcBodyTestInitInternal(interp, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Procbodytest_SafeInit --
+ *
+ *	This function initializes the "procbodytest" package.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Procbodytest_SafeInit(
+    Tcl_Interp *interp)		/* the Tcl interpreter for which the package
+				 * is initialized */
+{
+    return ProcBodyTestInitInternal(interp, 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RegisterCommand --
+ *
+ *	This function registers a command in the context of the given
+ *	namespace.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+RegisterCommand(
+    Tcl_Interp* interp,		/* the Tcl interpreter for which the operation
+				 * is performed */
+    const char *namespace,		/* the namespace in which the command is
+				 * registered */
+    const CmdTable *cmdTablePtr)/* the command to register */
+{
+    char buf[128];
+
+    if (cmdTablePtr->exportIt) {
+	sprintf(buf, "namespace eval %s { namespace export %s }",
+		namespace, cmdTablePtr->cmdName);
+	if (Tcl_EvalEx(interp, buf, -1, 0) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    sprintf(buf, "%s::%s", namespace, cmdTablePtr->cmdName);
+    Tcl_CreateObjCommand(interp, buf, cmdTablePtr->proc, 0, 0);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ProcBodyTestInitInternal --
+ *
+ *  This function initializes the Loader package.
+ *  The isSafe flag is 1 if the interpreter is safe, 0 otherwise.
+ *
+ * Results:
+ *  A standard Tcl result.
+ *
+ * Side effects:
+ *  None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ProcBodyTestInitInternal(
+    Tcl_Interp *interp,		/* the Tcl interpreter for which the package
+				 * is initialized */
+    int isSafe)			/* 1 if this is a safe interpreter */
+{
+    const CmdTable *cmdTablePtr;
+
+    cmdTablePtr = (isSafe) ? &safeCommands[0] : &commands[0];
+    for ( ; cmdTablePtr->cmdName ; cmdTablePtr++) {
+	if (RegisterCommand(interp, packageName, cmdTablePtr) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+
+    return Tcl_PkgProvide(interp, packageName, packageVersion);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ProcBodyTestProcObjCmd --
+ *
+ *  Implements the "procbodytest::proc" command. Here is the command
+ *  description:
+ *	procbodytest::proc newName argList bodyName
+ *  Looks up a procedure called $bodyName and, if the procedure exists,
+ *  constructs a Tcl_Obj of type "procbody" and calls Tcl_ProcObjCmd.
+ *  Arguments:
+ *    newName		the name of the procedure to be created
+ *    argList		the argument list for the procedure
+ *    bodyName		the name of an existing procedure from which the
+ *			body is to be copied.
+ *  This command can be used to trigger the branches in Tcl_ProcObjCmd that
+ *  construct a proc from a "procbody", for example:
+ *	proc a {x} {return $x}
+ *	a 123
+ *	procbodytest::proc b {x} a
+ *  Note the call to "a 123", which is necessary so that the Proc pointer
+ *  for "a" is filled in by the internal compiler; this is a hack.
+ *
+ * Results:
+ *  Returns a standard Tcl code.
+ *
+ * Side effects:
+ *  A new procedure is created.
+ *  Leaves an error message in the interp's result on error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ProcBodyTestProcObjCmd(
+    ClientData dummy,		/* context; not used */
+    Tcl_Interp *interp,		/* the current interpreter */
+    int objc,			/* argument count */
+    Tcl_Obj *const objv[])	/* arguments */
+{
+    const char *fullName;
+    Tcl_Command procCmd;
+    Command *cmdPtr;
+    Proc *procPtr = NULL;
+    Tcl_Obj *bodyObjPtr;
+    Tcl_Obj *myobjv[5];
+    int result;
+
+    if (objc != 4) {
+	Tcl_WrongNumArgs(interp, 1, objv, "newName argsList bodyName");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Find the Command pointer to this procedure
+     */
+
+    fullName = Tcl_GetString(objv[3]);
+    procCmd = Tcl_FindCommand(interp, fullName, NULL, TCL_LEAVE_ERR_MSG);
+    if (procCmd == NULL) {
+	return TCL_ERROR;
+    }
+
+    cmdPtr = (Command *) procCmd;
+
+    /*
+     * check that this is a procedure and not a builtin command:
+     * If a procedure, cmdPtr->objClientData is TclIsProc(cmdPtr).
+     */
+
+    if (cmdPtr->objClientData != TclIsProc(cmdPtr)) {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"command \"", fullName, "\" is not a Tcl procedure", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * it is a Tcl procedure: the client data is the Proc structure
+     */
+
+    procPtr = (Proc *) cmdPtr->objClientData;
+    if (procPtr == NULL) {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp), "procedure \"",
+		fullName, "\" does not have a Proc struct!", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * create a new object, initialize our argument vector, call into Tcl
+     */
+
+    bodyObjPtr = TclNewProcBodyObj(procPtr);
+    if (bodyObjPtr == NULL) {
+	Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
+		"failed to create a procbody object for procedure \"",
+		fullName, "\"", NULL);
+	return TCL_ERROR;
+    }
+    Tcl_IncrRefCount(bodyObjPtr);
+
+    myobjv[0] = objv[0];
+    myobjv[1] = objv[1];
+    myobjv[2] = objv[2];
+    myobjv[3] = bodyObjPtr;
+    myobjv[4] = NULL;
+
+    result = Tcl_ProcObjCmd(NULL, interp, objc, myobjv);
+    Tcl_DecrRefCount(bodyObjPtr);
+
+    return result;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclThread.c b/generic/tclThread.c
new file mode 100644
index 0000000..198fa6a
--- /dev/null
+++ b/generic/tclThread.c
@@ -0,0 +1,538 @@
+/*
+ * tclThread.c --
+ *
+ *	This file implements Platform independent thread operations. Most of
+ *	the real work is done in the platform dependent files.
+ *
+ * Copyright (c) 1998 by Sun Microsystems, Inc.
+ * Copyright (c) 2008 by George Peter Staplin
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * There are three classes of synchronization objects: mutexes, thread data
+ * keys, and condition variables. The following are used to record the memory
+ * used for these objects so they can be finalized.
+ *
+ * These statics are guarded by the mutex in the caller of
+ * TclRememberThreadData, e.g., TclpThreadDataKeyInit
+ */
+
+typedef struct {
+    int num;		/* Number of objects remembered */
+    int max;		/* Max size of the array */
+    void **list;	/* List of pointers */
+} SyncObjRecord;
+
+static SyncObjRecord keyRecord = {0, 0, NULL};
+static SyncObjRecord mutexRecord = {0, 0, NULL};
+static SyncObjRecord condRecord = {0, 0, NULL};
+
+/*
+ * Prototypes of functions used only in this file.
+ */
+
+static void		ForgetSyncObject(void *objPtr, SyncObjRecord *recPtr);
+static void		RememberSyncObject(void *objPtr,
+			    SyncObjRecord *recPtr);
+
+/*
+ * Several functions are #defined to nothing in tcl.h if TCL_THREADS is not
+ * specified. Here we undo that so the functions are defined in the stubs
+ * table.
+ */
+
+#ifndef TCL_THREADS
+#undef Tcl_MutexLock
+#undef Tcl_MutexUnlock
+#undef Tcl_MutexFinalize
+#undef Tcl_ConditionNotify
+#undef Tcl_ConditionWait
+#undef Tcl_ConditionFinalize
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetThreadData --
+ *
+ *	This function allocates and initializes a chunk of thread local
+ *	storage.
+ *
+ * Results:
+ *	A thread-specific pointer to the data structure.
+ *
+ * Side effects:
+ *	Will allocate memory the first time this thread calls for this chunk
+ *	of storage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void *
+Tcl_GetThreadData(
+    Tcl_ThreadDataKey *keyPtr,	/* Identifier for the data chunk */
+    int size)			/* Size of storage block */
+{
+    void *result;
+#ifdef TCL_THREADS
+    /*
+     * Initialize the key for this thread.
+     */
+
+    result = TclThreadStorageKeyGet(keyPtr);
+
+    if (result == NULL) {
+	result = ckalloc(size);
+	memset(result, 0, (size_t) size);
+	TclThreadStorageKeySet(keyPtr, result);
+    }
+#else /* TCL_THREADS */
+    if (*keyPtr == NULL) {
+	result = ckalloc(size);
+	memset(result, 0, (size_t)size);
+	*keyPtr = result;
+	RememberSyncObject(keyPtr, &keyRecord);
+    } else {
+	result = *keyPtr;
+    }
+#endif /* TCL_THREADS */
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadDataKeyGet --
+ *
+ *	This function returns a pointer to a block of thread local storage.
+ *
+ * Results:
+ *	A thread-specific pointer to the data structure, or NULL if the memory
+ *	has not been assigned to this key for this thread.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void *
+TclThreadDataKeyGet(
+    Tcl_ThreadDataKey *keyPtr)	/* Identifier for the data chunk. */
+
+{
+#ifdef TCL_THREADS
+    return TclThreadStorageKeyGet(keyPtr);
+#else /* TCL_THREADS */
+    return *keyPtr;
+#endif /* TCL_THREADS */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * RememberSyncObject
+ *
+ *	Keep a list of (mutexes/condition variable/data key) used during
+ *	finalization.
+ *
+ *	Assume master lock is held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Add to the appropriate list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+RememberSyncObject(
+    void *objPtr,		/* Pointer to sync object */
+    SyncObjRecord *recPtr)	/* Record of sync objects */
+{
+    void **newList;
+    int i, j;
+
+
+    /*
+     * Reuse any free slot in the list.
+     */
+
+    for (i=0 ; i < recPtr->num ; ++i) {
+	if (recPtr->list[i] == NULL) {
+	    recPtr->list[i] = objPtr;
+	    return;
+	}
+    }
+
+    /*
+     * Grow the list of pointers if necessary, copying only non-NULL
+     * pointers to the new list.
+     */
+
+    if (recPtr->num >= recPtr->max) {
+	recPtr->max += 8;
+	newList = ckalloc(recPtr->max * sizeof(void *));
+	for (i=0,j=0 ; i<recPtr->num ; i++) {
+	    if (recPtr->list[i] != NULL) {
+		newList[j++] = recPtr->list[i];
+	    }
+	}
+	if (recPtr->list != NULL) {
+	    ckfree(recPtr->list);
+	}
+	recPtr->list = newList;
+	recPtr->num = j;
+    }
+
+    recPtr->list[recPtr->num] = objPtr;
+    recPtr->num++;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ForgetSyncObject
+ *
+ *	Remove a single object from the list.
+ *	Assume master lock is held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Remove from the appropriate list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ForgetSyncObject(
+    void *objPtr,		/* Pointer to sync object */
+    SyncObjRecord *recPtr)	/* Record of sync objects */
+{
+    int i;
+
+    for (i=0 ; i<recPtr->num ; i++) {
+	if (objPtr == recPtr->list[i]) {
+	    recPtr->list[i] = NULL;
+	    return;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRememberMutex
+ *
+ *	Keep a list of mutexes used during finalization.
+ *	Assume master lock is held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Add to the mutex list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclRememberMutex(
+    Tcl_Mutex *mutexPtr)
+{
+    RememberSyncObject(mutexPtr, &mutexRecord);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_MutexFinalize --
+ *
+ *	Finalize a single mutex and remove it from the list of remembered
+ *	objects.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Remove the mutex from the list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_MutexFinalize(
+    Tcl_Mutex *mutexPtr)
+{
+#ifdef TCL_THREADS
+    TclpFinalizeMutex(mutexPtr);
+#endif
+    TclpMasterLock();
+    ForgetSyncObject(mutexPtr, &mutexRecord);
+    TclpMasterUnlock();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRememberCondition
+ *
+ *	Keep a list of condition variables used during finalization.
+ *	Assume master lock is held.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Add to the condition variable list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclRememberCondition(
+    Tcl_Condition *condPtr)
+{
+    RememberSyncObject(condPtr, &condRecord);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConditionFinalize --
+ *
+ *	Finalize a single condition variable and remove it from the list of
+ *	remembered objects.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Remove the condition variable from the list.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ConditionFinalize(
+    Tcl_Condition *condPtr)
+{
+#ifdef TCL_THREADS
+    TclpFinalizeCondition(condPtr);
+#endif
+    TclpMasterLock();
+    ForgetSyncObject(condPtr, &condRecord);
+    TclpMasterUnlock();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadData --
+ *
+ *	This function cleans up the thread-local storage. Secondary, it cleans
+ *	thread alloc cache.
+ *	This is called once for each thread before thread exits.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees up all thread local storage.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadData(int quick)
+{
+    TclFinalizeThreadDataThread();
+#if defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+    if (!quick) {
+	/*
+	 * Quick exit principle makes it useless to terminate allocators
+	 */
+	TclFinalizeThreadAllocThread();
+    }
+#endif
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeSynchronization --
+ *
+ *	This function cleans up all synchronization objects: mutexes,
+ *	condition variables, and thread-local storage.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Frees up the memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeSynchronization(void)
+{
+    int i;
+    void *blockPtr;
+    Tcl_ThreadDataKey *keyPtr;
+#ifdef TCL_THREADS
+    Tcl_Mutex *mutexPtr;
+    Tcl_Condition *condPtr;
+
+    TclpMasterLock();
+#endif
+
+    /*
+     * If we're running unthreaded, the TSD blocks are simply stored inside
+     * their thread data keys. Free them here.
+     */
+
+    if (keyRecord.list != NULL) {
+	for (i=0 ; i<keyRecord.num ; i++) {
+	    keyPtr = (Tcl_ThreadDataKey *) keyRecord.list[i];
+	    blockPtr = *keyPtr;
+	    ckfree(blockPtr);
+	}
+	ckfree(keyRecord.list);
+	keyRecord.list = NULL;
+    }
+    keyRecord.max = 0;
+    keyRecord.num = 0;
+
+#ifdef TCL_THREADS
+    /*
+     * Call thread storage master cleanup.
+     */
+
+    TclFinalizeThreadStorage();
+
+    for (i=0 ; i<mutexRecord.num ; i++) {
+	mutexPtr = (Tcl_Mutex *)mutexRecord.list[i];
+	if (mutexPtr != NULL) {
+	    TclpFinalizeMutex(mutexPtr);
+	}
+    }
+    if (mutexRecord.list != NULL) {
+	ckfree(mutexRecord.list);
+	mutexRecord.list = NULL;
+    }
+    mutexRecord.max = 0;
+    mutexRecord.num = 0;
+
+    for (i=0 ; i<condRecord.num ; i++) {
+	condPtr = (Tcl_Condition *) condRecord.list[i];
+	if (condPtr != NULL) {
+	    TclpFinalizeCondition(condPtr);
+	}
+    }
+    if (condRecord.list != NULL) {
+	ckfree(condRecord.list);
+	condRecord.list = NULL;
+    }
+    condRecord.max = 0;
+    condRecord.num = 0;
+
+    TclpMasterUnlock();
+#endif /* TCL_THREADS */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ExitThread --
+ *
+ *	This function is called to terminate the current thread. This should
+ *	be used by extensions that create threads with additional interpreters
+ *	in them.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	All thread exit handlers are invoked, then the thread dies.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ExitThread(
+    int status)
+{
+    Tcl_FinalizeThread();
+#ifdef TCL_THREADS
+    TclpThreadExit(status);
+#endif
+}
+
+#ifndef TCL_THREADS
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConditionWait, et al. --
+ *
+ *	These noop functions are provided so the stub table does not have to
+ *	be conditionalized for threads. The real implementations of these
+ *	functions live in the platform specific files.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#undef Tcl_ConditionWait
+void
+Tcl_ConditionWait(
+    Tcl_Condition *condPtr,	/* Really (pthread_cond_t **) */
+    Tcl_Mutex *mutexPtr,	/* Really (pthread_mutex_t **) */
+    const Tcl_Time *timePtr) /* Timeout on waiting period */
+{
+}
+
+#undef Tcl_ConditionNotify
+void
+Tcl_ConditionNotify(
+    Tcl_Condition *condPtr)
+{
+}
+
+#undef Tcl_MutexLock
+void
+Tcl_MutexLock(
+    Tcl_Mutex *mutexPtr)
+{
+}
+
+#undef Tcl_MutexUnlock
+void
+Tcl_MutexUnlock(
+    Tcl_Mutex *mutexPtr)
+{
+}
+#endif /* !TCL_THREADS */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclThreadAlloc.c b/generic/tclThreadAlloc.c
new file mode 100644
index 0000000..8077de4
--- /dev/null
+++ b/generic/tclThreadAlloc.c
@@ -0,0 +1,1210 @@
+/*
+ * tclThreadAlloc.c --
+ *
+ *	This is a very fast storage allocator for used with threads (designed
+ *	avoid lock contention). The basic strategy is to allocate memory in
+ *	fixed size blocks from block caches.
+ *
+ * The Initial Developer of the Original Code is America Online, Inc.
+ * Portions created by AOL are Copyright (C) 1999 America Online, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#if defined(TCL_THREADS) && defined(USE_THREAD_ALLOC)
+
+/*
+ * If range checking is enabled, an additional byte will be allocated to store
+ * the magic number at the end of the requested memory.
+ */
+
+#ifndef RCHECK
+#ifdef  NDEBUG
+#define RCHECK		0
+#else
+#define RCHECK		1
+#endif
+#endif
+
+/*
+ * The following define the number of Tcl_Obj's to allocate/move at a time and
+ * the high water mark to prune a per-thread cache. On a 32 bit system,
+ * sizeof(Tcl_Obj) = 24 so 800 * 24 = ~16k.
+ */
+
+#define NOBJALLOC	800
+
+/* Actual definition moved to tclInt.h */
+#define NOBJHIGH	ALLOC_NOBJHIGH
+
+/*
+ * The following union stores accounting information for each block including
+ * two small magic numbers and a bucket number when in use or a next pointer
+ * when free. The original requested size (not including the Block overhead)
+ * is also maintained.
+ */
+
+typedef union Block {
+    struct {
+	union {
+	    union Block *next;		/* Next in free list. */
+	    struct {
+		unsigned char magic1;	/* First magic number. */
+		unsigned char bucket;	/* Bucket block allocated from. */
+		unsigned char unused;	/* Padding. */
+		unsigned char magic2;	/* Second magic number. */
+	    } s;
+	} u;
+	size_t reqSize;			/* Requested allocation size. */
+    } b;
+    unsigned char padding[TCL_ALLOCALIGN];
+} Block;
+#define nextBlock	b.u.next
+#define sourceBucket	b.u.s.bucket
+#define magicNum1	b.u.s.magic1
+#define magicNum2	b.u.s.magic2
+#define MAGIC		0xEF
+#define blockReqSize	b.reqSize
+
+/*
+ * The following defines the minimum and and maximum block sizes and the number
+ * of buckets in the bucket cache.
+ */
+
+#define MINALLOC	((sizeof(Block) + 8 + (TCL_ALLOCALIGN-1)) & ~(TCL_ALLOCALIGN-1))
+#define NBUCKETS	(11 - (MINALLOC >> 5))
+#define MAXALLOC	(MINALLOC << (NBUCKETS - 1))
+
+/*
+ * The following structure defines a bucket of blocks with various accounting
+ * and statistics information.
+ */
+
+typedef struct Bucket {
+    Block *firstPtr;		/* First block available */
+    Block *lastPtr;		/* End of block list */
+    long numFree;		/* Number of blocks available */
+
+    /* All fields below for accounting only */
+
+    long numRemoves;		/* Number of removes from bucket */
+    long numInserts;		/* Number of inserts into bucket */
+    long numWaits;		/* Number of waits to acquire a lock */
+    long numLocks;		/* Number of locks acquired */
+    long totalAssigned;		/* Total space assigned to bucket */
+} Bucket;
+
+/*
+ * The following structure defines a cache of buckets and objs, of which there
+ * will be (at most) one per thread. Any changes need to be reflected in the
+ * struct AllocCache defined in tclInt.h, possibly also in the initialisation
+ * code in Tcl_CreateInterp().
+ */
+
+typedef struct Cache {
+    struct Cache *nextPtr;	/* Linked list of cache entries */
+    Tcl_ThreadId owner;		/* Which thread's cache is this? */
+    Tcl_Obj *firstObjPtr;	/* List of free objects for thread */
+    int numObjects;		/* Number of objects for thread */
+    Tcl_Obj *lastPtr;		/* Last object in this cache */
+    int totalAssigned;		/* Total space assigned to thread */
+    Bucket buckets[NBUCKETS];	/* The buckets for this thread */
+} Cache;
+
+/*
+ * The following array specifies various per-bucket limits and locks. The
+ * values are statically initialized to avoid calculating them repeatedly.
+ */
+
+static struct {
+    size_t blockSize;		/* Bucket blocksize. */
+    int maxBlocks;		/* Max blocks before move to share. */
+    int numMove;		/* Num blocks to move to share. */
+    Tcl_Mutex *lockPtr;		/* Share bucket lock. */
+} bucketInfo[NBUCKETS];
+
+/*
+ * Static functions defined in this file.
+ */
+
+static Cache *	GetCache(void);
+static void	LockBucket(Cache *cachePtr, int bucket);
+static void	UnlockBucket(Cache *cachePtr, int bucket);
+static void	PutBlocks(Cache *cachePtr, int bucket, int numMove);
+static int	GetBlocks(Cache *cachePtr, int bucket);
+static Block *	Ptr2Block(char *ptr);
+static char *	Block2Ptr(Block *blockPtr, int bucket, unsigned int reqSize);
+static void	MoveObjs(Cache *fromPtr, Cache *toPtr, int numMove);
+static void	PutObjs(Cache *fromPtr, int numMove);
+
+/*
+ * Local variables defined in this file and initialized at startup.
+ */
+
+static Tcl_Mutex *listLockPtr;
+static Tcl_Mutex *objLockPtr;
+static Cache sharedCache;
+static Cache *sharedPtr = &sharedCache;
+static Cache *firstCachePtr = &sharedCache;
+
+#if defined(HAVE_FAST_TSD)
+static __thread Cache *tcachePtr;
+
+# define GETCACHE(cachePtr)			\
+    do {					\
+	if (!tcachePtr) {			\
+	    tcachePtr = GetCache();		\
+	}					\
+	(cachePtr) = tcachePtr;			\
+    } while (0)
+#else
+# define GETCACHE(cachePtr)			\
+    do {					\
+	(cachePtr) = TclpGetAllocCache();	\
+	if ((cachePtr) == NULL) {		\
+	    (cachePtr) = GetCache();		\
+	}					\
+    } while (0)
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetCache ---
+ *
+ *	Gets per-thread memory cache, allocating it if necessary.
+ *
+ * Results:
+ *	Pointer to cache.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Cache *
+GetCache(void)
+{
+    Cache *cachePtr;
+
+    /*
+     * Check for first-time initialization.
+     */
+
+    if (listLockPtr == NULL) {
+	Tcl_Mutex *initLockPtr;
+
+	initLockPtr = Tcl_GetAllocMutex();
+	Tcl_MutexLock(initLockPtr);
+	if (listLockPtr == NULL) {
+	    TclInitThreadAlloc();
+	}
+	Tcl_MutexUnlock(initLockPtr);
+    }
+
+    /*
+     * Get this thread's cache, allocating if necessary.
+     */
+
+    cachePtr = TclpGetAllocCache();
+    if (cachePtr == NULL) {
+	cachePtr = TclpSysAlloc(sizeof(Cache), 0);
+	if (cachePtr == NULL) {
+	    Tcl_Panic("alloc: could not allocate new cache");
+	}
+        memset(cachePtr, 0, sizeof(Cache));
+	Tcl_MutexLock(listLockPtr);
+	cachePtr->nextPtr = firstCachePtr;
+	firstCachePtr = cachePtr;
+	Tcl_MutexUnlock(listLockPtr);
+	cachePtr->owner = Tcl_GetCurrentThread();
+	TclpSetAllocCache(cachePtr);
+    }
+    return cachePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFreeAllocCache --
+ *
+ *	Flush and delete a cache, removing from list of caches.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFreeAllocCache(
+    void *arg)
+{
+    Cache *cachePtr = arg;
+    Cache **nextPtrPtr;
+    register unsigned int bucket;
+
+    /*
+     * Flush blocks.
+     */
+
+    for (bucket = 0; bucket < NBUCKETS; ++bucket) {
+	if (cachePtr->buckets[bucket].numFree > 0) {
+	    PutBlocks(cachePtr, bucket, cachePtr->buckets[bucket].numFree);
+	}
+    }
+
+    /*
+     * Flush objs.
+     */
+
+    if (cachePtr->numObjects > 0) {
+	PutObjs(cachePtr, cachePtr->numObjects);
+    }
+
+    /*
+     * Remove from pool list.
+     */
+
+    Tcl_MutexLock(listLockPtr);
+    nextPtrPtr = &firstCachePtr;
+    while (*nextPtrPtr != cachePtr) {
+	nextPtrPtr = &(*nextPtrPtr)->nextPtr;
+    }
+    *nextPtrPtr = cachePtr->nextPtr;
+    cachePtr->nextPtr = NULL;
+    Tcl_MutexUnlock(listLockPtr);
+    TclpSysFree(cachePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpAlloc --
+ *
+ *	Allocate memory.
+ *
+ * Results:
+ *	Pointer to memory just beyond Block pointer.
+ *
+ * Side effects:
+ *	May allocate more blocks for a bucket.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpAlloc(
+    unsigned int reqSize)
+{
+    Cache *cachePtr;
+    Block *blockPtr;
+    register int bucket;
+    size_t size;
+
+#ifndef __LP64__
+    if (sizeof(int) >= sizeof(size_t)) {
+	/* An unsigned int overflow can also be a size_t overflow */
+	const size_t zero = 0;
+	const size_t max = ~zero;
+
+	if (((size_t) reqSize) > max - sizeof(Block) - RCHECK) {
+	    /* Requested allocation exceeds memory */
+	    return NULL;
+	}
+    }
+#endif
+
+    GETCACHE(cachePtr);
+
+    /*
+     * Increment the requested size to include room for the Block structure.
+     * Call TclpSysAlloc() directly if the required amount is greater than the
+     * largest block, otherwise pop the smallest block large enough,
+     * allocating more blocks if necessary.
+     */
+
+    blockPtr = NULL;
+    size = reqSize + sizeof(Block);
+#if RCHECK
+    size++;
+#endif
+    if (size > MAXALLOC) {
+	bucket = NBUCKETS;
+	blockPtr = TclpSysAlloc(size, 0);
+	if (blockPtr != NULL) {
+	    cachePtr->totalAssigned += reqSize;
+	}
+    } else {
+	bucket = 0;
+	while (bucketInfo[bucket].blockSize < size) {
+	    bucket++;
+	}
+	if (cachePtr->buckets[bucket].numFree || GetBlocks(cachePtr, bucket)) {
+	    blockPtr = cachePtr->buckets[bucket].firstPtr;
+	    cachePtr->buckets[bucket].firstPtr = blockPtr->nextBlock;
+	    cachePtr->buckets[bucket].numFree--;
+	    cachePtr->buckets[bucket].numRemoves++;
+	    cachePtr->buckets[bucket].totalAssigned += reqSize;
+	}
+    }
+    if (blockPtr == NULL) {
+	return NULL;
+    }
+    return Block2Ptr(blockPtr, bucket, reqSize);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpFree --
+ *
+ *	Return blocks to the thread block cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May move blocks to shared cache.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclpFree(
+    char *ptr)
+{
+    Cache *cachePtr;
+    Block *blockPtr;
+    int bucket;
+
+    if (ptr == NULL) {
+	return;
+    }
+
+    GETCACHE(cachePtr);
+
+    /*
+     * Get the block back from the user pointer and call system free directly
+     * for large blocks. Otherwise, push the block back on the bucket and move
+     * blocks to the shared cache if there are now too many free.
+     */
+
+    blockPtr = Ptr2Block(ptr);
+    bucket = blockPtr->sourceBucket;
+    if (bucket == NBUCKETS) {
+	cachePtr->totalAssigned -= blockPtr->blockReqSize;
+	TclpSysFree(blockPtr);
+	return;
+    }
+
+    cachePtr->buckets[bucket].totalAssigned -= blockPtr->blockReqSize;
+    blockPtr->nextBlock = cachePtr->buckets[bucket].firstPtr;
+    cachePtr->buckets[bucket].firstPtr = blockPtr;
+    if (cachePtr->buckets[bucket].numFree == 0) {
+	cachePtr->buckets[bucket].lastPtr = blockPtr;
+    }
+    cachePtr->buckets[bucket].numFree++;
+    cachePtr->buckets[bucket].numInserts++;
+
+    if (cachePtr != sharedPtr &&
+	    cachePtr->buckets[bucket].numFree > bucketInfo[bucket].maxBlocks) {
+	PutBlocks(cachePtr, bucket, bucketInfo[bucket].numMove);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpRealloc --
+ *
+ *	Re-allocate memory to a larger or smaller size.
+ *
+ * Results:
+ *	Pointer to memory just beyond Block pointer.
+ *
+ * Side effects:
+ *	Previous memory, if any, may be freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclpRealloc(
+    char *ptr,
+    unsigned int reqSize)
+{
+    Cache *cachePtr;
+    Block *blockPtr;
+    void *newPtr;
+    size_t size, min;
+    int bucket;
+
+    if (ptr == NULL) {
+	return TclpAlloc(reqSize);
+    }
+
+#ifndef __LP64__
+    if (sizeof(int) >= sizeof(size_t)) {
+	/* An unsigned int overflow can also be a size_t overflow */
+	const size_t zero = 0;
+	const size_t max = ~zero;
+
+	if (((size_t) reqSize) > max - sizeof(Block) - RCHECK) {
+	    /* Requested allocation exceeds memory */
+	    return NULL;
+	}
+    }
+#endif
+
+    GETCACHE(cachePtr);
+
+    /*
+     * If the block is not a system block and fits in place, simply return the
+     * existing pointer. Otherwise, if the block is a system block and the new
+     * size would also require a system block, call TclpSysRealloc() directly.
+     */
+
+    blockPtr = Ptr2Block(ptr);
+    size = reqSize + sizeof(Block);
+#if RCHECK
+    size++;
+#endif
+    bucket = blockPtr->sourceBucket;
+    if (bucket != NBUCKETS) {
+	if (bucket > 0) {
+	    min = bucketInfo[bucket-1].blockSize;
+	} else {
+	    min = 0;
+	}
+	if (size > min && size <= bucketInfo[bucket].blockSize) {
+	    cachePtr->buckets[bucket].totalAssigned -= blockPtr->blockReqSize;
+	    cachePtr->buckets[bucket].totalAssigned += reqSize;
+	    return Block2Ptr(blockPtr, bucket, reqSize);
+	}
+    } else if (size > MAXALLOC) {
+	cachePtr->totalAssigned -= blockPtr->blockReqSize;
+	cachePtr->totalAssigned += reqSize;
+	blockPtr = TclpSysRealloc(blockPtr, size);
+	if (blockPtr == NULL) {
+	    return NULL;
+	}
+	return Block2Ptr(blockPtr, NBUCKETS, reqSize);
+    }
+
+    /*
+     * Finally, perform an expensive malloc/copy/free.
+     */
+
+    newPtr = TclpAlloc(reqSize);
+    if (newPtr != NULL) {
+	if (reqSize > blockPtr->blockReqSize) {
+	    reqSize = blockPtr->blockReqSize;
+	}
+	memcpy(newPtr, ptr, reqSize);
+	TclpFree(ptr);
+    }
+    return newPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadAllocObj --
+ *
+ *	Allocate a Tcl_Obj from the per-thread cache.
+ *
+ * Results:
+ *	Pointer to uninitialized Tcl_Obj.
+ *
+ * Side effects:
+ *	May move Tcl_Obj's from shared cached or allocate new Tcl_Obj's if
+ *	list is empty.
+ *
+ * Note:
+ *	If this code is updated, the changes need to be reflected in the macro
+ *	TclAllocObjStorageEx() defined in tclInt.h
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclThreadAllocObj(void)
+{
+    register Cache *cachePtr;
+    register Tcl_Obj *objPtr;
+
+    GETCACHE(cachePtr);
+
+    /*
+     * Get this thread's obj list structure and move or allocate new objs if
+     * necessary.
+     */
+
+    if (cachePtr->numObjects == 0) {
+	register int numMove;
+
+	Tcl_MutexLock(objLockPtr);
+	numMove = sharedPtr->numObjects;
+	if (numMove > 0) {
+	    if (numMove > NOBJALLOC) {
+		numMove = NOBJALLOC;
+	    }
+	    MoveObjs(sharedPtr, cachePtr, numMove);
+	}
+	Tcl_MutexUnlock(objLockPtr);
+	if (cachePtr->numObjects == 0) {
+	    Tcl_Obj *newObjsPtr;
+
+	    cachePtr->numObjects = numMove = NOBJALLOC;
+	    newObjsPtr = TclpSysAlloc(sizeof(Tcl_Obj) * numMove, 0);
+	    if (newObjsPtr == NULL) {
+		Tcl_Panic("alloc: could not allocate %d new objects", numMove);
+	    }
+	    cachePtr->lastPtr = newObjsPtr + numMove - 1;
+	    objPtr = cachePtr->firstObjPtr;	/* NULL */
+	    while (--numMove >= 0) {
+		newObjsPtr[numMove].internalRep.twoPtrValue.ptr1 = objPtr;
+		objPtr = newObjsPtr + numMove;
+	    }
+	    cachePtr->firstObjPtr = newObjsPtr;
+	}
+    }
+
+    /*
+     * Pop the first object.
+     */
+
+    objPtr = cachePtr->firstObjPtr;
+    cachePtr->firstObjPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    cachePtr->numObjects--;
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadFreeObj --
+ *
+ *	Return a free Tcl_Obj to the per-thread cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May move free Tcl_Obj's to shared list upon hitting high water mark.
+ *
+ * Note:
+ *	If this code is updated, the changes need to be reflected in the macro
+ *	TclAllocObjStorageEx() defined in tclInt.h
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclThreadFreeObj(
+    Tcl_Obj *objPtr)
+{
+    Cache *cachePtr;
+
+    GETCACHE(cachePtr);
+
+    /*
+     * Get this thread's list and push on the free Tcl_Obj.
+     */
+
+    objPtr->internalRep.twoPtrValue.ptr1 = cachePtr->firstObjPtr;
+    cachePtr->firstObjPtr = objPtr;
+    if (cachePtr->numObjects == 0) {
+	cachePtr->lastPtr = objPtr;
+    }
+    cachePtr->numObjects++;
+
+    /*
+     * If the number of free objects has exceeded the high water mark, move
+     * some blocks to the shared list.
+     */
+
+    if (cachePtr->numObjects > NOBJHIGH) {
+	PutObjs(cachePtr, NOBJALLOC);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetMemoryInfo --
+ *
+ *	Return a list-of-lists of memory stats.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	List appended to given dstring.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetMemoryInfo(
+    Tcl_DString *dsPtr)
+{
+    Cache *cachePtr;
+    char buf[200];
+    unsigned int n;
+
+    Tcl_MutexLock(listLockPtr);
+    cachePtr = firstCachePtr;
+    while (cachePtr != NULL) {
+	Tcl_DStringStartSublist(dsPtr);
+	if (cachePtr == sharedPtr) {
+	    Tcl_DStringAppendElement(dsPtr, "shared");
+	} else {
+	    sprintf(buf, "thread%p", cachePtr->owner);
+	    Tcl_DStringAppendElement(dsPtr, buf);
+	}
+	for (n = 0; n < NBUCKETS; ++n) {
+	    sprintf(buf, "%lu %ld %ld %ld %ld %ld %ld",
+		    (unsigned long) bucketInfo[n].blockSize,
+		    cachePtr->buckets[n].numFree,
+		    cachePtr->buckets[n].numRemoves,
+		    cachePtr->buckets[n].numInserts,
+		    cachePtr->buckets[n].totalAssigned,
+		    cachePtr->buckets[n].numLocks,
+		    cachePtr->buckets[n].numWaits);
+	    Tcl_DStringAppendElement(dsPtr, buf);
+	}
+	Tcl_DStringEndSublist(dsPtr);
+	cachePtr = cachePtr->nextPtr;
+    }
+    Tcl_MutexUnlock(listLockPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * MoveObjs --
+ *
+ *	Move Tcl_Obj's between caches.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+MoveObjs(
+    Cache *fromPtr,
+    Cache *toPtr,
+    int numMove)
+{
+    register Tcl_Obj *objPtr = fromPtr->firstObjPtr;
+    Tcl_Obj *fromFirstObjPtr = objPtr;
+
+    toPtr->numObjects += numMove;
+    fromPtr->numObjects -= numMove;
+
+    /*
+     * Find the last object to be moved; set the next one (the first one not
+     * to be moved) as the first object in the 'from' cache.
+     */
+
+    while (--numMove) {
+	objPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    }
+    fromPtr->firstObjPtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    /*
+     * Move all objects as a block - they are already linked to each other, we
+     * just have to update the first and last.
+     */
+
+    toPtr->lastPtr = objPtr;
+    objPtr->internalRep.twoPtrValue.ptr1 = toPtr->firstObjPtr; /* NULL */
+    toPtr->firstObjPtr = fromFirstObjPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PutObjs --
+ *
+ *	Move Tcl_Obj's from thread cache to shared cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PutObjs(
+    Cache *fromPtr,
+    int numMove)
+{
+    int keep = fromPtr->numObjects - numMove;
+    Tcl_Obj *firstPtr, *lastPtr = NULL;
+
+    fromPtr->numObjects = keep;
+    firstPtr = fromPtr->firstObjPtr;
+    if (keep == 0) {
+	fromPtr->firstObjPtr = NULL;
+    } else {
+	do {
+	    lastPtr = firstPtr;
+	    firstPtr = firstPtr->internalRep.twoPtrValue.ptr1;
+	} while (--keep > 0);
+	lastPtr->internalRep.twoPtrValue.ptr1 = NULL;
+    }
+
+    /*
+     * Move all objects as a block - they are already linked to each other, we
+     * just have to update the first and last.
+     */
+
+    Tcl_MutexLock(objLockPtr);
+    fromPtr->lastPtr->internalRep.twoPtrValue.ptr1 = sharedPtr->firstObjPtr;
+    sharedPtr->firstObjPtr = firstPtr;
+    if (sharedPtr->numObjects == 0) {
+	sharedPtr->lastPtr = fromPtr->lastPtr;
+    }
+    sharedPtr->numObjects += numMove;
+    Tcl_MutexUnlock(objLockPtr);
+
+    fromPtr->lastPtr = lastPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Block2Ptr, Ptr2Block --
+ *
+ *	Convert between internal blocks and user pointers.
+ *
+ * Results:
+ *	User pointer or internal block.
+ *
+ * Side effects:
+ *	Invalid blocks will abort the server.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static char *
+Block2Ptr(
+    Block *blockPtr,
+    int bucket,
+    unsigned int reqSize)
+{
+    register void *ptr;
+
+    blockPtr->magicNum1 = blockPtr->magicNum2 = MAGIC;
+    blockPtr->sourceBucket = bucket;
+    blockPtr->blockReqSize = reqSize;
+    ptr = ((void *) (blockPtr + 1));
+#if RCHECK
+    ((unsigned char *)(ptr))[reqSize] = MAGIC;
+#endif
+    return (char *) ptr;
+}
+
+static Block *
+Ptr2Block(
+    char *ptr)
+{
+    register Block *blockPtr;
+
+    blockPtr = (((Block *) ptr) - 1);
+    if (blockPtr->magicNum1 != MAGIC || blockPtr->magicNum2 != MAGIC) {
+	Tcl_Panic("alloc: invalid block: %p: %x %x",
+		blockPtr, blockPtr->magicNum1, blockPtr->magicNum2);
+    }
+#if RCHECK
+    if (((unsigned char *) ptr)[blockPtr->blockReqSize] != MAGIC) {
+	Tcl_Panic("alloc: invalid block: %p: %x %x %x",
+		blockPtr, blockPtr->magicNum1, blockPtr->magicNum2,
+		((unsigned char *) ptr)[blockPtr->blockReqSize]);
+    }
+#endif
+    return blockPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * LockBucket, UnlockBucket --
+ *
+ *	Set/unset the lock to access a bucket in the shared cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Lock activity and contention are monitored globally and on a per-cache
+ *	basis.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+LockBucket(
+    Cache *cachePtr,
+    int bucket)
+{
+    Tcl_MutexLock(bucketInfo[bucket].lockPtr);
+    cachePtr->buckets[bucket].numLocks++;
+    sharedPtr->buckets[bucket].numLocks++;
+}
+
+static void
+UnlockBucket(
+    Cache *cachePtr,
+    int bucket)
+{
+    Tcl_MutexUnlock(bucketInfo[bucket].lockPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * PutBlocks --
+ *
+ *	Return unused blocks to the shared cache.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+PutBlocks(
+    Cache *cachePtr,
+    int bucket,
+    int numMove)
+{
+    /*
+     * We have numFree.  Want to shed numMove. So compute how many
+     * Blocks to keep.
+     */
+
+    int keep = cachePtr->buckets[bucket].numFree - numMove;
+    Block *lastPtr = NULL, *firstPtr;
+
+    cachePtr->buckets[bucket].numFree = keep;
+    firstPtr = cachePtr->buckets[bucket].firstPtr;
+    if (keep == 0) {
+	cachePtr->buckets[bucket].firstPtr = NULL;
+    } else {
+	do {
+	    lastPtr = firstPtr;
+	    firstPtr = firstPtr->nextBlock;
+	} while (--keep > 0);
+	lastPtr->nextBlock = NULL;
+    }
+
+    /*
+     * Aquire the lock and place the list of blocks at the front of the shared
+     * cache bucket.
+     */
+
+    LockBucket(cachePtr, bucket);
+    cachePtr->buckets[bucket].lastPtr->nextBlock
+	    = sharedPtr->buckets[bucket].firstPtr;
+    sharedPtr->buckets[bucket].firstPtr = firstPtr;
+    if (sharedPtr->buckets[bucket].numFree == 0) {
+	sharedPtr->buckets[bucket].lastPtr
+		= cachePtr->buckets[bucket].lastPtr;
+    }
+    sharedPtr->buckets[bucket].numFree += numMove;
+    UnlockBucket(cachePtr, bucket);
+
+    cachePtr->buckets[bucket].lastPtr = lastPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetBlocks --
+ *
+ *	Get more blocks for a bucket.
+ *
+ * Results:
+ *	1 if blocks where allocated, 0 otherwise.
+ *
+ * Side effects:
+ *	Cache may be filled with available blocks.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+GetBlocks(
+    Cache *cachePtr,
+    int bucket)
+{
+    register Block *blockPtr;
+    register int n;
+
+    /*
+     * First, atttempt to move blocks from the shared cache. Note the
+     * potentially dirty read of numFree before acquiring the lock which is a
+     * slight performance enhancement. The value is verified after the lock is
+     * actually acquired.
+     */
+
+    if (cachePtr != sharedPtr && sharedPtr->buckets[bucket].numFree > 0) {
+	LockBucket(cachePtr, bucket);
+	if (sharedPtr->buckets[bucket].numFree > 0) {
+
+	    /*
+	     * Either move the entire list or walk the list to find the last
+	     * block to move.
+	     */
+
+	    n = bucketInfo[bucket].numMove;
+	    if (n >= sharedPtr->buckets[bucket].numFree) {
+		cachePtr->buckets[bucket].firstPtr =
+			sharedPtr->buckets[bucket].firstPtr;
+		cachePtr->buckets[bucket].lastPtr =
+			sharedPtr->buckets[bucket].lastPtr;
+		cachePtr->buckets[bucket].numFree =
+			sharedPtr->buckets[bucket].numFree;
+		sharedPtr->buckets[bucket].firstPtr = NULL;
+		sharedPtr->buckets[bucket].numFree = 0;
+	    } else {
+		blockPtr = sharedPtr->buckets[bucket].firstPtr;
+		cachePtr->buckets[bucket].firstPtr = blockPtr;
+		sharedPtr->buckets[bucket].numFree -= n;
+		cachePtr->buckets[bucket].numFree = n;
+		while (--n > 0) {
+		    blockPtr = blockPtr->nextBlock;
+		}
+		sharedPtr->buckets[bucket].firstPtr = blockPtr->nextBlock;
+		cachePtr->buckets[bucket].lastPtr = blockPtr;
+		blockPtr->nextBlock = NULL;
+	    }
+	}
+	UnlockBucket(cachePtr, bucket);
+    }
+
+    if (cachePtr->buckets[bucket].numFree == 0) {
+	register size_t size;
+
+	/*
+	 * If no blocks could be moved from shared, first look for a larger
+	 * block in this cache to split up.
+	 */
+
+	blockPtr = NULL;
+	n = NBUCKETS;
+	size = 0; /* lint */
+	while (--n > bucket) {
+	    if (cachePtr->buckets[n].numFree > 0) {
+		size = bucketInfo[n].blockSize;
+		blockPtr = cachePtr->buckets[n].firstPtr;
+		cachePtr->buckets[n].firstPtr = blockPtr->nextBlock;
+		cachePtr->buckets[n].numFree--;
+		break;
+	    }
+	}
+
+	/*
+	 * Otherwise, allocate a big new block directly.
+	 */
+
+	if (blockPtr == NULL) {
+	    size = MAXALLOC;
+	    blockPtr = TclpSysAlloc(size, 0);
+	    if (blockPtr == NULL) {
+		return 0;
+	    }
+	}
+
+	/*
+	 * Split the larger block into smaller blocks for this bucket.
+	 */
+
+	n = size / bucketInfo[bucket].blockSize;
+	cachePtr->buckets[bucket].numFree = n;
+	cachePtr->buckets[bucket].firstPtr = blockPtr;
+	while (--n > 0) {
+	    blockPtr->nextBlock = (Block *)
+		((char *) blockPtr + bucketInfo[bucket].blockSize);
+	    blockPtr = blockPtr->nextBlock;
+	}
+	cachePtr->buckets[bucket].lastPtr = blockPtr;
+	blockPtr->nextBlock = NULL;
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitThreadAlloc --
+ *
+ *	Initializes the allocator cache-maintenance structures.
+ *      It is done early and protected during the TclInitSubsystems().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitThreadAlloc(void)
+{
+    unsigned int i;
+
+    listLockPtr = TclpNewAllocMutex();
+    objLockPtr = TclpNewAllocMutex();
+    for (i = 0; i < NBUCKETS; ++i) {
+	bucketInfo[i].blockSize = MINALLOC << i;
+	bucketInfo[i].maxBlocks = 1 << (NBUCKETS - 1 - i);
+	bucketInfo[i].numMove = i < NBUCKETS - 1 ?
+		1 << (NBUCKETS - 2 - i) : 1;
+	bucketInfo[i].lockPtr = TclpNewAllocMutex();
+    }
+    TclpInitAllocCache();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadAlloc --
+ *
+ *	This procedure is used to destroy all private resources used in this
+ *	file.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadAlloc(void)
+{
+    unsigned int i;
+
+    for (i = 0; i < NBUCKETS; ++i) {
+	TclpFreeAllocMutex(bucketInfo[i].lockPtr);
+	bucketInfo[i].lockPtr = NULL;
+    }
+
+    TclpFreeAllocMutex(objLockPtr);
+    objLockPtr = NULL;
+
+    TclpFreeAllocMutex(listLockPtr);
+    listLockPtr = NULL;
+
+    TclpFreeAllocCache(NULL);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadAllocThread --
+ *
+ *	This procedure is used to destroy single thread private resources
+ *	defined in this file. Called either during Tcl_FinalizeThread() or
+ *	Tcl_Finalize().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadAllocThread(void)
+{
+    Cache *cachePtr = TclpGetAllocCache();
+    if (cachePtr != NULL) {
+	TclpFreeAllocCache(cachePtr);
+    }
+}
+
+#else /* !(TCL_THREADS && USE_THREAD_ALLOC) */
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetMemoryInfo --
+ *
+ *	Return a list-of-lists of memory stats.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	List appended to given dstring.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetMemoryInfo(
+    Tcl_DString *dsPtr)
+{
+    Tcl_Panic("Tcl_GetMemoryInfo called when threaded memory allocator not in use");
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadAlloc --
+ *
+ *	This procedure is used to destroy all private resources used in this
+ *	file.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadAlloc(void)
+{
+    Tcl_Panic("TclFinalizeThreadAlloc called when threaded memory allocator not in use");
+}
+#endif /* TCL_THREADS && USE_THREAD_ALLOC */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclThreadJoin.c b/generic/tclThreadJoin.c
new file mode 100644
index 0000000..5c70a62
--- /dev/null
+++ b/generic/tclThreadJoin.c
@@ -0,0 +1,316 @@
+/*
+ * tclThreadJoin.c --
+ *
+ *	This file implements a platform independent emulation layer for the
+ *	handling of joinable threads. The Windows platform uses this code to
+ *	provide the functionality of joining threads.  This code is currently
+ *	not necessary on Unix.
+ *
+ * Copyright (c) 2000 by Scriptics Corporation
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+#ifdef _WIN32
+
+/*
+ * The information about each joinable thread is remembered in a structure as
+ * defined below.
+ */
+
+typedef struct JoinableThread {
+    Tcl_ThreadId  id;		/* The id of the joinable thread. */
+    int result;			/* A place for the result after the demise of
+				 * the thread. */
+    int done;			/* Boolean flag. Initialized to 0 and set to 1
+				 * after the exit of the thread. This allows a
+				 * thread requesting a join to detect when
+				 * waiting is not necessary. */
+    int waitedUpon;		/* Boolean flag. Initialized to 0 and set to 1
+				 * by the thread waiting for this one via
+				 * Tcl_JoinThread.  Used to lock any other
+				 * thread trying to wait on this one. */
+    Tcl_Mutex threadMutex;	/* The mutex used to serialize access to this
+				 * structure. */
+    Tcl_Condition cond;		/* This is the condition a thread has to wait
+				 * upon to get notified of the end of the
+				 * described thread. It is signaled indirectly
+				 * by Tcl_ExitThread. */
+    struct JoinableThread *nextThreadPtr;
+				/* Reference to the next thread in the list of
+				 * joinable threads. */
+} JoinableThread;
+
+/*
+ * The following variable is used to maintain the global list of all joinable
+ * threads. Usage by a thread is allowed only if the thread acquired the
+ * 'joinMutex'.
+ */
+
+TCL_DECLARE_MUTEX(joinMutex)
+
+static JoinableThread *firstThreadPtr;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclJoinThread --
+ *
+ *	This procedure waits for the exit of the thread with the specified id
+ *	and returns its result.
+ *
+ * Results:
+ *	A standard tcl result signaling the overall success/failure of the
+ *	operation and an integer result delivered by the thread which was
+ *	waited upon.
+ *
+ * Side effects:
+ *	Deallocates the memory allocated by TclRememberJoinableThread.
+ *	Removes the data associated to the thread waited upon from the list of
+ *	joinable threads.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclJoinThread(
+    Tcl_ThreadId id,		/* The id of the thread to wait upon. */
+    int *result)		/* Reference to a location for the result of
+				 * the thread we are waiting upon. */
+{
+    JoinableThread *threadPtr;
+
+    /*
+     * Steps done here:
+     * i.    Acquire the joinMutex and search for the thread.
+     * ii.   Error out if it could not be found.
+     * iii.  If found, switch from exclusive access to the list to exclusive
+     *	     access to the thread structure.
+     * iv.   Error out if some other is already waiting.
+     * v.    Skip the waiting part of the thread is already done.
+     * vi.   Wait for the thread to exit, mark it as waited upon too.
+     * vii.  Get the result form the structure,
+     * viii. switch to exclusive access of the list,
+     * ix.   remove the structure from the list,
+     * x.    then switch back to exclusive access to the structure
+     * xi.   and delete it.
+     */
+
+    Tcl_MutexLock(&joinMutex);
+
+    threadPtr = firstThreadPtr;
+    while (threadPtr!=NULL && threadPtr->id!=id) {
+	threadPtr = threadPtr->nextThreadPtr;
+    }
+
+    if (threadPtr == NULL) {
+	/*
+	 * Thread not found. Either not joinable, or already waited upon and
+	 * exited. Whatever, an error is in order.
+	 */
+
+	Tcl_MutexUnlock(&joinMutex);
+	return TCL_ERROR;
+    }
+
+    /*
+     * [1] If we don't lock the structure before giving up exclusive access to
+     * the list some other thread just completing its wait on the same thread
+     * can delete the structure from under us, leaving us with a dangling
+     * pointer.
+     */
+
+    Tcl_MutexLock(&threadPtr->threadMutex);
+    Tcl_MutexUnlock(&joinMutex);
+
+    /*
+     * [2] Now that we have the structure mutex any other thread that just
+     * tries to delete structure will wait at location [3] until we are done
+     * with the structure. And in that case we are done with it rather quickly
+     * as 'waitedUpon' will be set and we will have to error out.
+     */
+
+    if (threadPtr->waitedUpon) {
+	Tcl_MutexUnlock(&threadPtr->threadMutex);
+	return TCL_ERROR;
+    }
+
+    /*
+     * We are waiting now, let other threads recognize this.
+     */
+
+    threadPtr->waitedUpon = 1;
+
+    while (!threadPtr->done) {
+	Tcl_ConditionWait(&threadPtr->cond, &threadPtr->threadMutex, NULL);
+    }
+
+    /*
+     * We have to release the structure before trying to access the list again
+     * or we can run into deadlock with a thread at [1] (see above) because of
+     * us holding the structure and the other holding the list.  There is no
+     * problem with dangling pointers here as 'waitedUpon == 1' is still valid
+     * and any other thread will error out and not come to this place. IOW,
+     * the fact that we are here also means that no other thread came here
+     * before us and is able to delete the structure.
+     */
+
+    Tcl_MutexUnlock(&threadPtr->threadMutex);
+    Tcl_MutexLock(&joinMutex);
+
+    /*
+     * We have to search the list again as its structure may (may, almost
+     * certainly) have changed while we were waiting. Especially now is the
+     * time to compute the predecessor in the list. Any earlier result can be
+     * dangling by now.
+     */
+
+    if (firstThreadPtr == threadPtr) {
+	firstThreadPtr = threadPtr->nextThreadPtr;
+    } else {
+	JoinableThread *prevThreadPtr = firstThreadPtr;
+
+	while (prevThreadPtr->nextThreadPtr != threadPtr) {
+	    prevThreadPtr = prevThreadPtr->nextThreadPtr;
+	}
+	prevThreadPtr->nextThreadPtr = threadPtr->nextThreadPtr;
+    }
+
+    Tcl_MutexUnlock(&joinMutex);
+
+    /*
+     * [3] Now that the structure is not part of the list anymore no other
+     * thread can acquire its mutex from now on. But it is possible that
+     * another thread is still holding the mutex though, see location [2].  So
+     * we have to acquire the mutex one more time to wait for that thread to
+     * finish. We can (and have to) release the mutex immediately.
+     */
+
+    Tcl_MutexLock(&threadPtr->threadMutex);
+    Tcl_MutexUnlock(&threadPtr->threadMutex);
+
+    /*
+     * Copy the result to us, finalize the synchronisation objects, then free
+     * the structure and return.
+     */
+
+    *result = threadPtr->result;
+
+    Tcl_ConditionFinalize(&threadPtr->cond);
+    Tcl_MutexFinalize(&threadPtr->threadMutex);
+    ckfree(threadPtr);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclRememberJoinableThread --
+ *
+ *	This procedure remebers a thread as joinable. Only a call to
+ *	TclJoinThread will remove the structre created (and initialized) here.
+ *	IOW, not waiting upon a joinable thread will cause memory leaks.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates memory, adds it to the global list of all joinable threads.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclRememberJoinableThread(
+    Tcl_ThreadId id)		/* The thread to remember as joinable */
+{
+    JoinableThread *threadPtr;
+
+    threadPtr = ckalloc(sizeof(JoinableThread));
+    threadPtr->id = id;
+    threadPtr->done = 0;
+    threadPtr->waitedUpon = 0;
+    threadPtr->threadMutex = (Tcl_Mutex) NULL;
+    threadPtr->cond = (Tcl_Condition) NULL;
+
+    Tcl_MutexLock(&joinMutex);
+
+    threadPtr->nextThreadPtr = firstThreadPtr;
+    firstThreadPtr = threadPtr;
+
+    Tcl_MutexUnlock(&joinMutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSignalExitThread --
+ *
+ *	This procedure signals that the specified thread is done with its
+ *	work. If the thread is joinable this signal is propagated to the
+ *	thread waiting upon it.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Modifies the associated structure to hold the result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSignalExitThread(
+    Tcl_ThreadId id,		/* Id of the thread signaling its exit. */
+    int result)			/* The result from the thread. */
+{
+    JoinableThread *threadPtr;
+
+    Tcl_MutexLock(&joinMutex);
+
+    threadPtr = firstThreadPtr;
+    while ((threadPtr != NULL) && (threadPtr->id != id)) {
+	threadPtr = threadPtr->nextThreadPtr;
+    }
+
+    if (threadPtr == NULL) {
+	/*
+	 * Thread not found. Not joinable. No problem, nothing to do.
+	 */
+
+	Tcl_MutexUnlock(&joinMutex);
+	return;
+    }
+
+    /*
+     * Switch over the exclusive access from the list to the structure, then
+     * store the result, set the flag and notify the waiting thread, provided
+     * that it exists. The order of lock/unlock ensures that a thread entering
+     * 'TclJoinThread' will not interfere with us.
+     */
+
+    Tcl_MutexLock(&threadPtr->threadMutex);
+    Tcl_MutexUnlock(&joinMutex);
+
+    threadPtr->done = 1;
+    threadPtr->result = result;
+
+    if (threadPtr->waitedUpon) {
+	Tcl_ConditionNotify(&threadPtr->cond);
+    }
+
+    Tcl_MutexUnlock(&threadPtr->threadMutex);
+}
+#endif /* _WIN32 */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclThreadStorage.c b/generic/tclThreadStorage.c
new file mode 100644
index 0000000..755a461
--- /dev/null
+++ b/generic/tclThreadStorage.c
@@ -0,0 +1,373 @@
+/*
+ * tclThreadStorage.c --
+ *
+ *	This file implements platform independent thread storage operations to
+ *	work around system limits on the number of thread-specific variables.
+ *
+ * Copyright (c) 2003-2004 by Joe Mistachkin
+ * Copyright (c) 2008 by George Peter Staplin
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+#ifdef TCL_THREADS
+#include <signal.h>
+
+/*
+ * IMPLEMENTATION NOTES:
+ *
+ * The primary idea is that we create one platform-specific TSD slot, and use
+ * it for storing a table pointer. Each Tcl_ThreadDataKey has an offset into
+ * the table of TSD values. We don't use more than 1 platform-specific TSD
+ * slot, because there is a hard limit on the number of TSD slots. Valid key
+ * offsets are greater than 0; 0 is for the initialized Tcl_ThreadDataKey.
+ */
+
+/*
+ * The master collection of information about TSDs. This is shared across the
+ * whole process, and includes the mutex used to protect it.
+ */
+
+static struct TSDMaster {
+    void *key;			/* Key into the system TSD structure. The
+				 * collection of Tcl TSD values for a
+				 * particular thread will hang off the
+				 * back-end of this. */
+    sig_atomic_t counter;	/* The number of different Tcl TSDs used
+				 * across *all* threads. This is a strictly
+				 * increasing value. */
+    Tcl_Mutex mutex;		/* Protection for the rest of this structure,
+				 * which holds per-process data. */
+} tsdMaster = { NULL, 0, NULL };
+
+/*
+ * The type of the data held per thread in a system TSD.
+ */
+
+typedef struct TSDTable {
+    ClientData *tablePtr;	/* The table of Tcl TSDs. */
+    sig_atomic_t allocated;	/* The size of the table in the current
+				 * thread. */
+} TSDTable;
+
+/*
+ * The actual type of Tcl_ThreadDataKey.
+ */
+
+typedef union TSDUnion {
+    volatile sig_atomic_t offset;
+				/* The type is really an offset into the
+				 * thread-local table of TSDs, which is this
+				 * field. */
+    void *ptr;			/* For alignment purposes only. Not actually
+				 * accessed through this. */
+} TSDUnion;
+
+/*
+ * Forward declarations of functions in this file.
+ */
+
+static TSDTable *	TSDTableCreate(void);
+static void		TSDTableDelete(TSDTable *tsdTablePtr);
+static void		TSDTableGrow(TSDTable *tsdTablePtr,
+			    sig_atomic_t atLeast);
+
+/*
+ * Allocator and deallocator for a TSDTable structure.
+ */
+
+static TSDTable *
+TSDTableCreate(void)
+{
+    TSDTable *tsdTablePtr;
+    sig_atomic_t i;
+
+    tsdTablePtr = TclpSysAlloc(sizeof(TSDTable), 0);
+    if (tsdTablePtr == NULL) {
+	Tcl_Panic("unable to allocate TSDTable");
+    }
+
+    tsdTablePtr->allocated = 8;
+    tsdTablePtr->tablePtr =
+	    TclpSysAlloc(sizeof(void *) * tsdTablePtr->allocated, 0);
+    if (tsdTablePtr->tablePtr == NULL) {
+	Tcl_Panic("unable to allocate TSDTable");
+    }
+
+    for (i = 0; i < tsdTablePtr->allocated; ++i) {
+	tsdTablePtr->tablePtr[i] = NULL;
+    }
+
+    return tsdTablePtr;
+}
+
+static void
+TSDTableDelete(
+    TSDTable *tsdTablePtr)
+{
+    sig_atomic_t i;
+
+    for (i=0 ; i<tsdTablePtr->allocated ; i++) {
+	if (tsdTablePtr->tablePtr[i] != NULL) {
+	    /*
+	     * These values were allocated in Tcl_GetThreadData in tclThread.c
+	     * and must now be deallocated or they will leak.
+	     */
+
+	    ckfree(tsdTablePtr->tablePtr[i]);
+	}
+    }
+
+    TclpSysFree(tsdTablePtr->tablePtr);
+    TclpSysFree(tsdTablePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TSDTableGrow --
+ *
+ *	This procedure makes the passed TSDTable grow to fit the atLeast
+ *	value.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The table is enlarged.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TSDTableGrow(
+    TSDTable *tsdTablePtr,
+    sig_atomic_t atLeast)
+{
+    sig_atomic_t newAllocated = tsdTablePtr->allocated * 2;
+    ClientData *newTablePtr;
+    sig_atomic_t i;
+
+    if (newAllocated <= atLeast) {
+	newAllocated = atLeast + 10;
+    }
+
+    newTablePtr = TclpSysRealloc(tsdTablePtr->tablePtr,
+	    sizeof(ClientData) * newAllocated);
+    if (newTablePtr == NULL) {
+	Tcl_Panic("unable to reallocate TSDTable");
+    }
+
+    for (i = tsdTablePtr->allocated; i < newAllocated; ++i) {
+	newTablePtr[i] = NULL;
+    }
+
+    tsdTablePtr->allocated = newAllocated;
+    tsdTablePtr->tablePtr = newTablePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadStorageKeyGet --
+ *
+ *    This procedure gets the value associated with the passed key.
+ *
+ * Results:
+ *	A pointer value associated with the Tcl_ThreadDataKey or NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void *
+TclThreadStorageKeyGet(
+    Tcl_ThreadDataKey *dataKeyPtr)
+{
+    TSDTable *tsdTablePtr = TclpThreadGetMasterTSD(tsdMaster.key);
+    ClientData resultPtr = NULL;
+    TSDUnion *keyPtr = (TSDUnion *) dataKeyPtr;
+    sig_atomic_t offset = keyPtr->offset;
+
+    if ((tsdTablePtr != NULL) && (offset > 0)
+	    && (offset < tsdTablePtr->allocated)) {
+	resultPtr = tsdTablePtr->tablePtr[offset];
+    }
+    return resultPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThreadStorageKeySet --
+ *
+ *	This procedure set an association of value with the key passed. The
+ *	associated value may be retrieved with TclThreadDataKeyGet().
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The thread-specific table may be created or reallocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclThreadStorageKeySet(
+    Tcl_ThreadDataKey *dataKeyPtr,
+    void *value)
+{
+    TSDTable *tsdTablePtr = TclpThreadGetMasterTSD(tsdMaster.key);
+    TSDUnion *keyPtr = (TSDUnion *) dataKeyPtr;
+
+    if (tsdTablePtr == NULL) {
+	tsdTablePtr = TSDTableCreate();
+	TclpThreadSetMasterTSD(tsdMaster.key, tsdTablePtr);
+    }
+
+    /*
+     * Get the lock while we check if this TSD is new or not. Note that this
+     * is the only place where Tcl_ThreadDataKey values are set. We use a
+     * double-checked lock to try to avoid having to grab this lock a lot,
+     * since it is on quite a few critical paths and will only get set once in
+     * each location.
+     */
+
+    if (keyPtr->offset == 0) {
+	Tcl_MutexLock(&tsdMaster.mutex);
+	if (keyPtr->offset == 0) {
+	    /*
+	     * The Tcl_ThreadDataKey hasn't been used yet. Make a new one.
+	     */
+
+	    keyPtr->offset = ++tsdMaster.counter;
+	}
+	Tcl_MutexUnlock(&tsdMaster.mutex);
+    }
+
+    /*
+     * Check if this is the first time this Tcl_ThreadDataKey has been used
+     * with the current thread. Note that we don't need to hold a lock when
+     * doing this, as we are *definitely* the only point accessing this
+     * tsdTablePtr right now; it's thread-local.
+     */
+
+    if (keyPtr->offset >= tsdTablePtr->allocated) {
+	TSDTableGrow(tsdTablePtr, keyPtr->offset);
+    }
+
+    /*
+     * Set the value in the Tcl thread-local variable.
+     */
+
+    tsdTablePtr->tablePtr[keyPtr->offset] = value;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadDataThread --
+ *
+ *	This procedure finalizes the data for a single thread.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The TSDTable is deleted/freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadDataThread(void)
+{
+    TSDTable *tsdTablePtr = TclpThreadGetMasterTSD(tsdMaster.key);
+
+    if (tsdTablePtr != NULL) {
+	TSDTableDelete(tsdTablePtr);
+	TclpThreadSetMasterTSD(tsdMaster.key, NULL);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitializeThreadStorage --
+ *
+ *	This procedure initializes the TSD subsystem with per-platform code.
+ *	This should be called before any Tcl threads are created.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocates a system TSD.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclInitThreadStorage(void)
+{
+    tsdMaster.key = TclpThreadCreateKey();
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFinalizeThreadStorage --
+ *
+ *	This procedure cleans up the thread storage data key for all threads.
+ *	IMPORTANT: All Tcl threads must be finalized before calling this!
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Releases the thread data key.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFinalizeThreadStorage(void)
+{
+    TclpThreadDeleteKey(tsdMaster.key);
+    tsdMaster.key = NULL;
+}
+
+#else /* !TCL_THREADS */
+/*
+ * Stub functions for non-threaded builds
+ */
+
+void
+TclInitThreadStorage(void)
+{
+}
+
+void
+TclFinalizeThreadDataThread(void)
+{
+}
+
+void
+TclFinalizeThreadStorage(void)
+{
+}
+#endif /* TCL_THREADS */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclThreadTest.c b/generic/tclThreadTest.c
new file mode 100644
index 0000000..9c5fecb
--- /dev/null
+++ b/generic/tclThreadTest.c
@@ -0,0 +1,1211 @@
+/*
+ * tclThreadTest.c --
+ *
+ *	This file implements the testthread command. Eventually this should be
+ *	tclThreadCmd.c
+ *	Some of this code is based on work done by Richard Hipp on behalf of
+ *	Conservation Through Innovation, Limited, with their permission.
+ *
+ * Copyright (c) 1998 by Sun Microsystems, Inc.
+ * Copyright (c) 2006-2008 by Joe Mistachkin.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef USE_TCL_STUBS
+#   define USE_TCL_STUBS
+#endif
+#include "tclInt.h"
+
+#ifdef TCL_THREADS
+/*
+ * Each thread has an single instance of the following structure. There is one
+ * instance of this structure per thread even if that thread contains multiple
+ * interpreters. The interpreter identified by this structure is the main
+ * interpreter for the thread.
+ *
+ * The main interpreter is the one that will process any messages received by
+ * a thread. Any thread can send messages but only the main interpreter can
+ * receive them.
+ */
+
+typedef struct ThreadSpecificData {
+    Tcl_ThreadId threadId;	/* Tcl ID for this thread */
+    Tcl_Interp *interp;		/* Main interpreter for this thread */
+    int flags;			/* See the TP_ defines below... */
+    struct ThreadSpecificData *nextPtr;
+				/* List for "thread names" */
+    struct ThreadSpecificData *prevPtr;
+				/* List for "thread names" */
+} ThreadSpecificData;
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * This list is used to list all threads that have interpreters. This is
+ * protected by threadMutex.
+ */
+
+static ThreadSpecificData *threadList = NULL;
+
+/*
+ * The following bit-values are legal for the "flags" field of the
+ * ThreadSpecificData structure.
+ */
+
+#define TP_Dying		0x001 /* This thread is being canceled */
+
+/*
+ * An instance of the following structure contains all information that is
+ * passed into a new thread when the thread is created using either the
+ * "thread create" Tcl command or the ThreadCreate() C function.
+ */
+
+typedef struct ThreadCtrl {
+    const char *script;		/* The Tcl command this thread should
+				 * execute */
+    int flags;			/* Initial value of the "flags" field in the
+				 * ThreadSpecificData structure for the new
+				 * thread. Might contain TP_Detached or
+				 * TP_TclThread. */
+    Tcl_Condition condWait;	/* This condition variable is used to
+				 * synchronize the parent and child threads.
+				 * The child won't run until it acquires
+				 * threadMutex, and the parent function won't
+				 * complete until signaled on this condition
+				 * variable. */
+} ThreadCtrl;
+
+/*
+ * This is the event used to send scripts to other threads.
+ */
+
+typedef struct ThreadEvent {
+    Tcl_Event event;		/* Must be first */
+    char *script;		/* The script to execute. */
+    struct ThreadEventResult *resultPtr;
+				/* To communicate the result. This is NULL if
+				 * we don't care about it. */
+} ThreadEvent;
+
+typedef struct ThreadEventResult {
+    Tcl_Condition done;		/* Signaled when the script completes */
+    int code;			/* Return value of Tcl_Eval */
+    char *result;		/* Result from the script */
+    char *errorInfo;		/* Copy of errorInfo variable */
+    char *errorCode;		/* Copy of errorCode variable */
+    Tcl_ThreadId srcThreadId;	/* Id of sending thread, in case it dies */
+    Tcl_ThreadId dstThreadId;	/* Id of target thread, in case it dies */
+    struct ThreadEvent *eventPtr;	/* Back pointer */
+    struct ThreadEventResult *nextPtr;	/* List for cleanup */
+    struct ThreadEventResult *prevPtr;
+
+} ThreadEventResult;
+
+static ThreadEventResult *resultList;
+
+/*
+ * This is for simple error handling when a thread script exits badly.
+ */
+
+static Tcl_ThreadId mainThreadId;
+static Tcl_ThreadId errorThreadId;
+static char *errorProcString;
+
+/*
+ * Access to the list of threads and to the thread send results is guarded by
+ * this mutex.
+ */
+
+TCL_DECLARE_MUTEX(threadMutex)
+
+static int		ThreadObjCmd(ClientData clientData,
+			    Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static int		ThreadCreate(Tcl_Interp *interp, const char *script,
+			    int joinable);
+static int		ThreadList(Tcl_Interp *interp);
+static int		ThreadSend(Tcl_Interp *interp, Tcl_ThreadId id,
+			    const char *script, int wait);
+static int		ThreadCancel(Tcl_Interp *interp, Tcl_ThreadId id,
+			    const char *result, int flags);
+
+static Tcl_ThreadCreateType	NewTestThread(ClientData clientData);
+static void		ListRemove(ThreadSpecificData *tsdPtr);
+static void		ListUpdateInner(ThreadSpecificData *tsdPtr);
+static int		ThreadEventProc(Tcl_Event *evPtr, int mask);
+static void		ThreadErrorProc(Tcl_Interp *interp);
+static void		ThreadFreeProc(ClientData clientData);
+static int		ThreadDeleteEvent(Tcl_Event *eventPtr,
+			    ClientData clientData);
+static void		ThreadExitProc(ClientData clientData);
+extern int		Tcltest_Init(Tcl_Interp *interp);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclThread_Init --
+ *
+ *	Initialize the test thread command.
+ *
+ * Results:
+ *	TCL_OK if the package was properly initialized.
+ *
+ * Side effects:
+ *	Add the "testthread" command to the interp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclThread_Init(
+    Tcl_Interp *interp)		/* The current Tcl interpreter */
+{
+    /*
+     * If the main thread Id has not been set, do it now.
+     */
+
+    Tcl_MutexLock(&threadMutex);
+    if (mainThreadId == 0) {
+	mainThreadId = Tcl_GetCurrentThread();
+    }
+    Tcl_MutexUnlock(&threadMutex);
+
+    Tcl_CreateObjCommand(interp, "testthread", ThreadObjCmd, NULL, NULL);
+    return TCL_OK;
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ThreadObjCmd --
+ *
+ *	This procedure is invoked to process the "testthread" Tcl command. See
+ *	the user documentation for details on what it does.
+ *
+ *	thread cancel ?-unwind? id ?result?
+ *	thread create ?-joinable? ?script?
+ *	thread send ?-async? id script
+ *	thread event
+ *	thread exit
+ *	thread id ?-main?
+ *	thread names
+ *	thread wait
+ *	thread errorproc proc
+ *	thread join id
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ThreadObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    int option;
+    static const char *const threadOptions[] = {
+	"cancel", "create", "event", "exit", "id",
+	"join", "names", "send", "wait", "errorproc",
+	NULL
+    };
+    enum options {
+	THREAD_CANCEL, THREAD_CREATE, THREAD_EVENT, THREAD_EXIT,
+	THREAD_ID, THREAD_JOIN, THREAD_NAMES, THREAD_SEND,
+	THREAD_WAIT, THREAD_ERRORPROC
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], threadOptions, "option", 0,
+	    &option) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make sure the initial thread is on the list before doing anything.
+     */
+
+    if (tsdPtr->interp == NULL) {
+	Tcl_MutexLock(&threadMutex);
+	tsdPtr->interp = interp;
+	ListUpdateInner(tsdPtr);
+	Tcl_CreateThreadExitHandler(ThreadExitProc, NULL);
+	Tcl_MutexUnlock(&threadMutex);
+    }
+
+    switch ((enum options)option) {
+    case THREAD_CANCEL: {
+	long id;
+	const char *result;
+	int flags, arg;
+
+	if ((objc < 3) || (objc > 5)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?-unwind? id ?result?");
+	    return TCL_ERROR;
+	}
+	flags = 0;
+	arg = 2;
+	if ((objc == 4) || (objc == 5)) {
+	    if (strcmp("-unwind", Tcl_GetString(objv[arg])) == 0) {
+		flags = TCL_CANCEL_UNWIND;
+		arg++;
+	    }
+	}
+	if (Tcl_GetLongFromObj(interp, objv[arg], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	arg++;
+	if (arg < objc) {
+	    result = Tcl_GetString(objv[arg]);
+	} else {
+	    result = NULL;
+	}
+	return ThreadCancel(interp, (Tcl_ThreadId) (size_t) id, result, flags);
+    }
+    case THREAD_CREATE: {
+	const char *script;
+	int joinable, len;
+
+	if (objc == 2) {
+	    /*
+	     * Neither joinable nor special script
+	     */
+
+	    joinable = 0;
+	    script = "testthread wait";		/* Just enter event loop */
+	} else if (objc == 3) {
+	    /*
+	     * Possibly -joinable, then no special script, no joinable, then
+	     * its a script.
+	     */
+
+	    script = Tcl_GetStringFromObj(objv[2], &len);
+
+	    if ((len > 1) && (script[0] == '-') && (script[1] == 'j') &&
+		    (0 == strncmp(script, "-joinable", (size_t) len))) {
+		joinable = 1;
+		script = "testthread wait";	/* Just enter event loop */
+	    } else {
+		/*
+		 * Remember the script
+		 */
+
+		joinable = 0;
+	    }
+	} else if (objc == 4) {
+	    /*
+	     * Definitely a script available, but is the flag -joinable?
+	     */
+
+	    script = Tcl_GetStringFromObj(objv[2], &len);
+	    joinable = ((len > 1) && (script[0] == '-') && (script[1] == 'j')
+		    && (0 == strncmp(script, "-joinable", (size_t) len)));
+	    script = Tcl_GetString(objv[3]);
+	} else {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?-joinable? ?script?");
+	    return TCL_ERROR;
+	}
+	return ThreadCreate(interp, script, joinable);
+    }
+    case THREAD_EXIT:
+	if (objc > 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	ListRemove(NULL);
+	Tcl_ExitThread(0);
+	return TCL_OK;
+    case THREAD_ID:
+	if (objc == 2 || objc == 3) {
+	    Tcl_Obj *idObj;
+
+	    /*
+	     * Check if they want the main thread id or the current thread id.
+	     */
+
+	    if (objc == 2) {
+		idObj = Tcl_NewWideIntObj((Tcl_WideInt)(size_t)Tcl_GetCurrentThread());
+	    } else if (objc == 3
+		    && strcmp("-main", Tcl_GetString(objv[2])) == 0) {
+		Tcl_MutexLock(&threadMutex);
+		idObj = Tcl_NewWideIntObj((Tcl_WideInt)(size_t)mainThreadId);
+		Tcl_MutexUnlock(&threadMutex);
+	    } else {
+		Tcl_WrongNumArgs(interp, 2, objv, NULL);
+		return TCL_ERROR;
+	    }
+
+	    Tcl_SetObjResult(interp, idObj);
+	    return TCL_OK;
+	} else {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+    case THREAD_JOIN: {
+	Tcl_WideInt id;
+	int result, status;
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "id");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetWideIntFromObj(interp, objv[2], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	result = Tcl_JoinThread((Tcl_ThreadId)(size_t)id, &status);
+	if (result == TCL_OK) {
+	    Tcl_SetIntObj(Tcl_GetObjResult(interp), status);
+	} else {
+	    char buf[20];
+
+	    sprintf(buf, "%" TCL_LL_MODIFIER "d", id);
+	    Tcl_AppendResult(interp, "cannot join thread ", buf, NULL);
+	}
+	return result;
+    }
+    case THREAD_NAMES:
+	if (objc > 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return ThreadList(interp);
+    case THREAD_SEND: {
+	Tcl_WideInt id;
+	const char *script;
+	int wait, arg;
+
+	if ((objc != 4) && (objc != 5)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?-async? id script");
+	    return TCL_ERROR;
+	}
+	if (objc == 5) {
+	    if (strcmp("-async", Tcl_GetString(objv[2])) != 0) {
+		Tcl_WrongNumArgs(interp, 2, objv, "?-async? id script");
+		return TCL_ERROR;
+	    }
+	    wait = 0;
+	    arg = 3;
+	} else {
+	    wait = 1;
+	    arg = 2;
+	}
+	if (Tcl_GetWideIntFromObj(interp, objv[arg], &id) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	arg++;
+	script = Tcl_GetString(objv[arg]);
+	return ThreadSend(interp, (Tcl_ThreadId)(size_t)id, script, wait);
+    }
+    case THREAD_EVENT: {
+	if (objc > 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(
+		Tcl_DoOneEvent(TCL_ALL_EVENTS | TCL_DONT_WAIT)));
+	return TCL_OK;
+    }
+    case THREAD_ERRORPROC: {
+	/*
+	 * Arrange for this proc to handle thread death errors.
+	 */
+
+	const char *proc;
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "proc");
+	    return TCL_ERROR;
+	}
+	Tcl_MutexLock(&threadMutex);
+	errorThreadId = Tcl_GetCurrentThread();
+	if (errorProcString) {
+	    ckfree(errorProcString);
+	}
+	proc = Tcl_GetString(objv[2]);
+	errorProcString = ckalloc(strlen(proc) + 1);
+	strcpy(errorProcString, proc);
+	Tcl_MutexUnlock(&threadMutex);
+	return TCL_OK;
+    }
+    case THREAD_WAIT:
+	if (objc > 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "");
+	    return TCL_ERROR;
+	}
+	while (1) {
+	    /*
+	     * If the script has been unwound, bail out immediately. This does
+	     * not follow the recommended guidelines for how extensions should
+	     * handle the script cancellation functionality because this is
+	     * not a "normal" extension. Most extensions do not have a command
+	     * that simply enters an infinite Tcl event loop. Normal
+	     * extensions should not specify the TCL_CANCEL_UNWIND when
+	     * calling Tcl_Canceled to check if the command has been canceled.
+	     */
+
+	    if (Tcl_Canceled(interp,
+		    TCL_LEAVE_ERR_MSG | TCL_CANCEL_UNWIND) == TCL_ERROR) {
+		break;
+	    }
+	    (void) Tcl_DoOneEvent(TCL_ALL_EVENTS);
+	}
+
+	/*
+	 * If we get to this point, we have been canceled by another thread,
+	 * which is considered to be an "error".
+	 */
+
+	ThreadErrorProc(interp);
+	return TCL_OK;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ThreadCreate --
+ *
+ *	This procedure is invoked to create a thread containing an interp to
+ *	run a script. This returns after the thread has started executing.
+ *
+ * Results:
+ *	A standard Tcl result, which is the thread ID.
+ *
+ * Side effects:
+ *	Create a thread.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ThreadCreate(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    const char *script,		/* Script to execute */
+    int joinable)		/* Flag, joinable thread or not */
+{
+    ThreadCtrl ctrl;
+    Tcl_ThreadId id;
+
+    ctrl.script = script;
+    ctrl.condWait = NULL;
+    ctrl.flags = 0;
+
+    joinable = joinable ? TCL_THREAD_JOINABLE : TCL_THREAD_NOFLAGS;
+
+    Tcl_MutexLock(&threadMutex);
+    if (Tcl_CreateThread(&id, NewTestThread, (ClientData) &ctrl,
+	    TCL_THREAD_STACK_DEFAULT, joinable) != TCL_OK) {
+	Tcl_MutexUnlock(&threadMutex);
+	Tcl_AppendResult(interp, "can't create a new thread", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Wait for the thread to start because it is using something on our stack!
+     */
+
+    Tcl_ConditionWait(&ctrl.condWait, &threadMutex, NULL);
+    Tcl_MutexUnlock(&threadMutex);
+    Tcl_ConditionFinalize(&ctrl.condWait);
+    Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt)(size_t)id));
+    return TCL_OK;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * NewTestThread --
+ *
+ *	This routine is the "main()" for a new thread whose task is to execute
+ *	a single Tcl script. The argument to this function is a pointer to a
+ *	structure that contains the text of the TCL script to be executed.
+ *
+ *	Space to hold the script field of the ThreadControl structure passed
+ *	in as the only argument was obtained from malloc() and must be freed
+ *	by this function before it exits. Space to hold the ThreadControl
+ *	structure itself is released by the calling function, and the two
+ *	condition variables in the ThreadControl structure are destroyed by
+ *	the calling function. The calling function will destroy the
+ *	ThreadControl structure and the condition variable as soon as
+ *	ctrlPtr->condWait is signaled, so this routine must make copies of any
+ *	data it might need after that point.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	A Tcl script is executed in a new thread.
+ *
+ *------------------------------------------------------------------------
+ */
+
+Tcl_ThreadCreateType
+NewTestThread(
+    ClientData clientData)
+{
+    ThreadCtrl *ctrlPtr = clientData;
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    int result;
+    char *threadEvalScript;
+
+    /*
+     * Initialize the interpreter. This should be more general.
+     */
+
+    tsdPtr->interp = Tcl_CreateInterp();
+    result = Tcl_Init(tsdPtr->interp);
+    if (result != TCL_OK) {
+	ThreadErrorProc(tsdPtr->interp);
+    }
+
+    /*
+     * This is part of the test facility. Initialize _ALL_ test commands for
+     * use by the new thread.
+     */
+
+    result = Tcltest_Init(tsdPtr->interp);
+    if (result != TCL_OK) {
+	ThreadErrorProc(tsdPtr->interp);
+    }
+
+    /*
+     * Update the list of threads.
+     */
+
+    Tcl_MutexLock(&threadMutex);
+    ListUpdateInner(tsdPtr);
+
+    /*
+     * We need to keep a pointer to the alloc'ed mem of the script we are
+     * eval'ing, for the case that we exit during evaluation
+     */
+
+    threadEvalScript = ckalloc(strlen(ctrlPtr->script) + 1);
+    strcpy(threadEvalScript, ctrlPtr->script);
+
+    Tcl_CreateThreadExitHandler(ThreadExitProc, threadEvalScript);
+
+    /*
+     * Notify the parent we are alive.
+     */
+
+    Tcl_ConditionNotify(&ctrlPtr->condWait);
+    Tcl_MutexUnlock(&threadMutex);
+
+    /*
+     * Run the script.
+     */
+
+    Tcl_Preserve(tsdPtr->interp);
+    result = Tcl_EvalEx(tsdPtr->interp, threadEvalScript, -1, 0);
+    if (result != TCL_OK) {
+	ThreadErrorProc(tsdPtr->interp);
+    }
+
+    /*
+     * Clean up.
+     */
+
+    Tcl_DeleteInterp(tsdPtr->interp);
+    Tcl_Release(tsdPtr->interp);
+    ListRemove(tsdPtr);
+    Tcl_ExitThread(result);
+
+    TCL_THREAD_CREATE_RETURN;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadErrorProc --
+ *
+ *	Send a message to the thread willing to hear about errors.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Send an event.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static void
+ThreadErrorProc(
+    Tcl_Interp *interp)		/* Interp that failed */
+{
+    Tcl_Channel errChannel;
+    const char *errorInfo, *argv[3];
+    char *script;
+    char buf[TCL_DOUBLE_SPACE+1];
+
+    sprintf(buf, "%p", Tcl_GetCurrentThread());
+
+    errorInfo = Tcl_GetVar2(interp, "errorInfo", NULL, TCL_GLOBAL_ONLY);
+    if (errorProcString == NULL) {
+	errChannel = Tcl_GetStdChannel(TCL_STDERR);
+	Tcl_WriteChars(errChannel, "Error from thread ", -1);
+	Tcl_WriteChars(errChannel, buf, -1);
+	Tcl_WriteChars(errChannel, "\n", 1);
+	Tcl_WriteChars(errChannel, errorInfo, -1);
+	Tcl_WriteChars(errChannel, "\n", 1);
+    } else {
+	argv[0] = errorProcString;
+	argv[1] = buf;
+	argv[2] = errorInfo;
+	script = Tcl_Merge(3, argv);
+	ThreadSend(interp, errorThreadId, script, 0);
+	ckfree(script);
+    }
+}
+
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ListUpdateInner --
+ *
+ *	Add the thread local storage to the list. This assumes the caller has
+ *	obtained the mutex.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Add the thread local storage to its list.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static void
+ListUpdateInner(
+    ThreadSpecificData *tsdPtr)
+{
+    if (tsdPtr == NULL) {
+	tsdPtr = TCL_TSD_INIT(&dataKey);
+    }
+    tsdPtr->threadId = Tcl_GetCurrentThread();
+    tsdPtr->nextPtr = threadList;
+    if (threadList) {
+	threadList->prevPtr = tsdPtr;
+    }
+    tsdPtr->prevPtr = NULL;
+    threadList = tsdPtr;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ListRemove --
+ *
+ *	Remove the thread local storage from its list. This grabs the mutex to
+ *	protect the list.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Remove the thread local storage from its list.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static void
+ListRemove(
+    ThreadSpecificData *tsdPtr)
+{
+    if (tsdPtr == NULL) {
+	tsdPtr = TCL_TSD_INIT(&dataKey);
+    }
+    Tcl_MutexLock(&threadMutex);
+    if (tsdPtr->prevPtr) {
+	tsdPtr->prevPtr->nextPtr = tsdPtr->nextPtr;
+    } else {
+	threadList = tsdPtr->nextPtr;
+    }
+    if (tsdPtr->nextPtr) {
+	tsdPtr->nextPtr->prevPtr = tsdPtr->prevPtr;
+    }
+    tsdPtr->nextPtr = tsdPtr->prevPtr = 0;
+    tsdPtr->interp = NULL;
+    Tcl_MutexUnlock(&threadMutex);
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadList --
+ *
+ *    Return a list of threads running Tcl interpreters.
+ *
+ * Results:
+ *    A standard Tcl result.
+ *
+ * Side effects:
+ *    None.
+ *
+ *------------------------------------------------------------------------
+ */
+static int
+ThreadList(
+    Tcl_Interp *interp)
+{
+    ThreadSpecificData *tsdPtr;
+    Tcl_Obj *listPtr;
+
+    listPtr = Tcl_NewListObj(0, NULL);
+    Tcl_MutexLock(&threadMutex);
+    for (tsdPtr = threadList ; tsdPtr ; tsdPtr = tsdPtr->nextPtr) {
+	Tcl_ListObjAppendElement(interp, listPtr,
+		Tcl_NewWideIntObj((Tcl_WideInt)(size_t)tsdPtr->threadId));
+    }
+    Tcl_MutexUnlock(&threadMutex);
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadSend --
+ *
+ *    Send a script to another thread.
+ *
+ * Results:
+ *    A standard Tcl result.
+ *
+ * Side effects:
+ *    None.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static int
+ThreadSend(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    Tcl_ThreadId id,		/* Thread Id of other interpreter. */
+    const char *script,		/* The script to evaluate. */
+    int wait)			/* If 1, we block for the result. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ThreadEvent *threadEventPtr;
+    ThreadEventResult *resultPtr;
+    int found, code;
+    Tcl_ThreadId threadId = (Tcl_ThreadId) id;
+
+    /*
+     * Verify the thread exists.
+     */
+
+    Tcl_MutexLock(&threadMutex);
+    found = 0;
+    for (tsdPtr = threadList ; tsdPtr ; tsdPtr = tsdPtr->nextPtr) {
+	if (tsdPtr->threadId == threadId) {
+	    found = 1;
+	    break;
+	}
+    }
+    if (!found) {
+	Tcl_MutexUnlock(&threadMutex);
+	Tcl_AppendResult(interp, "invalid thread id", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Short circut sends to ourself. Ought to do something with -async, like
+     * run in an idle handler.
+     */
+
+    if (threadId == Tcl_GetCurrentThread()) {
+	Tcl_MutexUnlock(&threadMutex);
+	return Tcl_EvalEx(interp, script,-1,TCL_EVAL_GLOBAL);
+    }
+
+    /*
+     * Create the event for its event queue.
+     */
+
+    threadEventPtr = ckalloc(sizeof(ThreadEvent));
+    threadEventPtr->script = ckalloc(strlen(script) + 1);
+    strcpy(threadEventPtr->script, script);
+    if (!wait) {
+	resultPtr = threadEventPtr->resultPtr = NULL;
+    } else {
+	resultPtr = ckalloc(sizeof(ThreadEventResult));
+	threadEventPtr->resultPtr = resultPtr;
+
+	/*
+	 * Initialize the result fields.
+	 */
+
+	resultPtr->done = NULL;
+	resultPtr->code = 0;
+	resultPtr->result = NULL;
+	resultPtr->errorInfo = NULL;
+	resultPtr->errorCode = NULL;
+
+	/*
+	 * Maintain the cleanup list.
+	 */
+
+	resultPtr->srcThreadId = Tcl_GetCurrentThread();
+	resultPtr->dstThreadId = threadId;
+	resultPtr->eventPtr = threadEventPtr;
+	resultPtr->nextPtr = resultList;
+	if (resultList) {
+	    resultList->prevPtr = resultPtr;
+	}
+	resultPtr->prevPtr = NULL;
+	resultList = resultPtr;
+    }
+
+    /*
+     * Queue the event and poke the other thread's notifier.
+     */
+
+    threadEventPtr->event.proc = ThreadEventProc;
+    Tcl_ThreadQueueEvent(threadId, (Tcl_Event *) threadEventPtr,
+	    TCL_QUEUE_TAIL);
+    Tcl_ThreadAlert(threadId);
+
+    if (!wait) {
+	Tcl_MutexUnlock(&threadMutex);
+	return TCL_OK;
+    }
+
+    /*
+     * Block on the results and then get them.
+     */
+
+    Tcl_ResetResult(interp);
+    while (resultPtr->result == NULL) {
+	Tcl_ConditionWait(&resultPtr->done, &threadMutex, NULL);
+    }
+
+    /*
+     * Unlink result from the result list.
+     */
+
+    if (resultPtr->prevPtr) {
+	resultPtr->prevPtr->nextPtr = resultPtr->nextPtr;
+    } else {
+	resultList = resultPtr->nextPtr;
+    }
+    if (resultPtr->nextPtr) {
+	resultPtr->nextPtr->prevPtr = resultPtr->prevPtr;
+    }
+    resultPtr->eventPtr = NULL;
+    resultPtr->nextPtr = NULL;
+    resultPtr->prevPtr = NULL;
+
+    Tcl_MutexUnlock(&threadMutex);
+
+    if (resultPtr->code != TCL_OK) {
+	if (resultPtr->errorCode) {
+	    Tcl_SetErrorCode(interp, resultPtr->errorCode, NULL);
+	    ckfree(resultPtr->errorCode);
+	}
+	if (resultPtr->errorInfo) {
+	    Tcl_AddErrorInfo(interp, resultPtr->errorInfo);
+	    ckfree(resultPtr->errorInfo);
+	}
+    }
+    Tcl_AppendResult(interp, resultPtr->result, NULL);
+    Tcl_ConditionFinalize(&resultPtr->done);
+    code = resultPtr->code;
+
+    ckfree(resultPtr->result);
+    ckfree(resultPtr);
+
+    return code;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadCancel --
+ *
+ *    Cancels a script in another thread.
+ *
+ * Results:
+ *    A standard Tcl result.
+ *
+ * Side effects:
+ *    None.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static int
+ThreadCancel(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    Tcl_ThreadId id,		/* Thread Id of other interpreter. */
+    const char *result,		/* The result or NULL for default. */
+    int flags)			/* Flags for Tcl_CancelEval. */
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    int found;
+    Tcl_ThreadId threadId = (Tcl_ThreadId) id;
+
+    /*
+     * Verify the thread exists.
+     */
+
+    Tcl_MutexLock(&threadMutex);
+    found = 0;
+    for (tsdPtr = threadList ; tsdPtr ; tsdPtr = tsdPtr->nextPtr) {
+	if (tsdPtr->threadId == threadId) {
+	    found = 1;
+	    break;
+	}
+    }
+    if (!found) {
+	Tcl_MutexUnlock(&threadMutex);
+	Tcl_AppendResult(interp, "invalid thread id", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Since Tcl_CancelEval can be safely called from any thread,
+     * we do it now.
+     */
+
+    Tcl_MutexUnlock(&threadMutex);
+    Tcl_ResetResult(interp);
+    return Tcl_CancelEval(tsdPtr->interp,
+    	(result != NULL) ? Tcl_NewStringObj(result, -1) : NULL, 0, flags);
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadEventProc --
+ *
+ *    Handle the event in the target thread.
+ *
+ * Results:
+ *    Returns 1 to indicate that the event was processed.
+ *
+ * Side effects:
+ *    Fills out the ThreadEventResult struct.
+ *
+ *------------------------------------------------------------------------
+ */
+
+static int
+ThreadEventProc(
+    Tcl_Event *evPtr,		/* Really ThreadEvent */
+    int mask)
+{
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+    ThreadEvent *threadEventPtr = (ThreadEvent *) evPtr;
+    ThreadEventResult *resultPtr = threadEventPtr->resultPtr;
+    Tcl_Interp *interp = tsdPtr->interp;
+    int code;
+    const char *result, *errorCode, *errorInfo;
+
+    if (interp == NULL) {
+	code = TCL_ERROR;
+	result = "no target interp!";
+	errorCode = "THREAD";
+	errorInfo = "";
+    } else {
+	Tcl_Preserve(interp);
+	Tcl_ResetResult(interp);
+	Tcl_CreateThreadExitHandler(ThreadFreeProc, threadEventPtr->script);
+	code = Tcl_EvalEx(interp, threadEventPtr->script,-1,TCL_EVAL_GLOBAL);
+	Tcl_DeleteThreadExitHandler(ThreadFreeProc, threadEventPtr->script);
+	if (code != TCL_OK) {
+	    errorCode = Tcl_GetVar2(interp, "errorCode", NULL, TCL_GLOBAL_ONLY);
+	    errorInfo = Tcl_GetVar2(interp, "errorInfo", NULL, TCL_GLOBAL_ONLY);
+	} else {
+	    errorCode = errorInfo = NULL;
+	}
+	result = Tcl_GetStringResult(interp);
+    }
+    ckfree(threadEventPtr->script);
+    if (resultPtr) {
+	Tcl_MutexLock(&threadMutex);
+	resultPtr->code = code;
+	resultPtr->result = ckalloc(strlen(result) + 1);
+	strcpy(resultPtr->result, result);
+	if (errorCode != NULL) {
+	    resultPtr->errorCode = ckalloc(strlen(errorCode) + 1);
+	    strcpy(resultPtr->errorCode, errorCode);
+	}
+	if (errorInfo != NULL) {
+	    resultPtr->errorInfo = ckalloc(strlen(errorInfo) + 1);
+	    strcpy(resultPtr->errorInfo, errorInfo);
+	}
+	Tcl_ConditionNotify(&resultPtr->done);
+	Tcl_MutexUnlock(&threadMutex);
+    }
+    if (interp != NULL) {
+	Tcl_Release(interp);
+    }
+    return 1;
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadFreeProc --
+ *
+ *    This is called from when we are exiting and memory needs
+ *    to be freed.
+ *
+ * Results:
+ *    None.
+ *
+ * Side effects:
+ *	Clears up mem specified in ClientData
+ *
+ *------------------------------------------------------------------------
+ */
+
+     /* ARGSUSED */
+static void
+ThreadFreeProc(
+    ClientData clientData)
+{
+    if (clientData) {
+	ckfree(clientData);
+    }
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadDeleteEvent --
+ *
+ *    This is called from the ThreadExitProc to delete memory related
+ *    to events that we put on the queue.
+ *
+ * Results:
+ *    1 it was our event and we want it removed, 0 otherwise.
+ *
+ * Side effects:
+ *	It cleans up our events in the event queue for this thread.
+ *
+ *------------------------------------------------------------------------
+ */
+
+     /* ARGSUSED */
+static int
+ThreadDeleteEvent(
+    Tcl_Event *eventPtr,	/* Really ThreadEvent */
+    ClientData clientData)	/* dummy */
+{
+    if (eventPtr->proc == ThreadEventProc) {
+	ckfree(((ThreadEvent *) eventPtr)->script);
+	return 1;
+    }
+
+    /*
+     * If it was NULL, we were in the middle of servicing the event and it
+     * should be removed
+     */
+
+    return (eventPtr->proc == NULL);
+}
+
+/*
+ *------------------------------------------------------------------------
+ *
+ * ThreadExitProc --
+ *
+ *    This is called when the thread exits.
+ *
+ * Results:
+ *    None.
+ *
+ * Side effects:
+ *	It unblocks anyone that is waiting on a send to this thread. It cleans
+ *	up any events in the event queue for this thread.
+ *
+ *------------------------------------------------------------------------
+ */
+
+     /* ARGSUSED */
+static void
+ThreadExitProc(
+    ClientData clientData)
+{
+    char *threadEvalScript = clientData;
+    ThreadEventResult *resultPtr, *nextPtr;
+    Tcl_ThreadId self = Tcl_GetCurrentThread();
+    ThreadSpecificData *tsdPtr = TCL_TSD_INIT(&dataKey);
+
+    if (tsdPtr->interp != NULL) {
+	ListRemove(tsdPtr);
+    }
+
+    Tcl_MutexLock(&threadMutex);
+
+    if (threadEvalScript) {
+	ckfree(threadEvalScript);
+	threadEvalScript = NULL;
+    }
+    Tcl_DeleteEvents((Tcl_EventDeleteProc *) ThreadDeleteEvent, NULL);
+
+    for (resultPtr = resultList ; resultPtr ; resultPtr = nextPtr) {
+	nextPtr = resultPtr->nextPtr;
+	if (resultPtr->srcThreadId == self) {
+	    /*
+	     * We are going away. By freeing up the result we signal to the
+	     * other thread we don't care about the result.
+	     */
+
+	    if (resultPtr->prevPtr) {
+		resultPtr->prevPtr->nextPtr = resultPtr->nextPtr;
+	    } else {
+		resultList = resultPtr->nextPtr;
+	    }
+	    if (resultPtr->nextPtr) {
+		resultPtr->nextPtr->prevPtr = resultPtr->prevPtr;
+	    }
+	    resultPtr->nextPtr = resultPtr->prevPtr = 0;
+	    resultPtr->eventPtr->resultPtr = NULL;
+	    ckfree(resultPtr);
+	} else if (resultPtr->dstThreadId == self) {
+	    /*
+	     * Dang. The target is going away. Unblock the caller. The result
+	     * string must be dynamically allocated because the main thread is
+	     * going to call free on it.
+	     */
+
+	    const char *msg = "target thread died";
+
+	    resultPtr->result = ckalloc(strlen(msg) + 1);
+	    strcpy(resultPtr->result, msg);
+	    resultPtr->code = TCL_ERROR;
+	    Tcl_ConditionNotify(&resultPtr->done);
+	}
+    }
+    Tcl_MutexUnlock(&threadMutex);
+}
+#endif /* TCL_THREADS */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclTimer.c b/generic/tclTimer.c
new file mode 100644
index 0000000..3467305
--- /dev/null
+++ b/generic/tclTimer.c
@@ -0,0 +1,1299 @@
+/*
+ * tclTimer.c --
+ *
+ *	This file provides timer event management facilities for Tcl,
+ *	including the "after" command.
+ *
+ * Copyright (c) 1997 by Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * For each timer callback that's pending there is one record of the following
+ * type. The normal handlers (created by Tcl_CreateTimerHandler) are chained
+ * together in a list sorted by time (earliest event first).
+ */
+
+typedef struct TimerHandler {
+    Tcl_Time time;		/* When timer is to fire. */
+    Tcl_TimerProc *proc;	/* Function to call. */
+    ClientData clientData;	/* Argument to pass to proc. */
+    Tcl_TimerToken token;	/* Identifies handler so it can be deleted. */
+    struct TimerHandler *nextPtr;
+				/* Next event in queue, or NULL for end of
+				 * queue. */
+} TimerHandler;
+
+/*
+ * The data structure below is used by the "after" command to remember the
+ * command to be executed later. All of the pending "after" commands for an
+ * interpreter are linked together in a list.
+ */
+
+typedef struct AfterInfo {
+    struct AfterAssocData *assocPtr;
+				/* Pointer to the "tclAfter" assocData for the
+				 * interp in which command will be
+				 * executed. */
+    Tcl_Obj *commandPtr;	/* Command to execute. */
+    int id;			/* Integer identifier for command; used to
+				 * cancel it. */
+    Tcl_TimerToken token;	/* Used to cancel the "after" command. NULL
+				 * means that the command is run as an idle
+				 * handler rather than as a timer handler.
+				 * NULL means this is an "after idle" handler
+				 * rather than a timer handler. */
+    struct AfterInfo *nextPtr;	/* Next in list of all "after" commands for
+				 * this interpreter. */
+} AfterInfo;
+
+/*
+ * One of the following structures is associated with each interpreter for
+ * which an "after" command has ever been invoked. A pointer to this structure
+ * is stored in the AssocData for the "tclAfter" key.
+ */
+
+typedef struct AfterAssocData {
+    Tcl_Interp *interp;		/* The interpreter for which this data is
+				 * registered. */
+    AfterInfo *firstAfterPtr;	/* First in list of all "after" commands still
+				 * pending for this interpreter, or NULL if
+				 * none. */
+} AfterAssocData;
+
+/*
+ * There is one of the following structures for each of the handlers declared
+ * in a call to Tcl_DoWhenIdle. All of the currently-active handlers are
+ * linked together into a list.
+ */
+
+typedef struct IdleHandler {
+    Tcl_IdleProc *proc;		/* Function to call. */
+    ClientData clientData;	/* Value to pass to proc. */
+    int generation;		/* Used to distinguish older handlers from
+				 * recently-created ones. */
+    struct IdleHandler *nextPtr;/* Next in list of active handlers. */
+} IdleHandler;
+
+/*
+ * The timer and idle queues are per-thread because they are associated with
+ * the notifier, which is also per-thread.
+ *
+ * All static variables used in this file are collected into a single instance
+ * of the following structure. For multi-threaded implementations, there is
+ * one instance of this structure for each thread.
+ *
+ * Notice that different structures with the same name appear in other files.
+ * The structure defined below is used in this file only.
+ */
+
+typedef struct {
+    TimerHandler *firstTimerHandlerPtr;	/* First event in queue. */
+    int lastTimerId;		/* Timer identifier of most recently created
+				 * timer. */
+    int timerPending;		/* 1 if a timer event is in the queue. */
+    IdleHandler *idleList;	/* First in list of all idle handlers. */
+    IdleHandler *lastIdlePtr;	/* Last in list (or NULL for empty list). */
+    int idleGeneration;		/* Used to fill in the "generation" fields of
+				 * IdleHandler structures. Increments each
+				 * time Tcl_DoOneEvent starts calling idle
+				 * handlers, so that all old handlers can be
+				 * called without calling any of the new ones
+				 * created by old ones. */
+    int afterId;		/* For unique identifiers of after events. */
+} ThreadSpecificData;
+
+static Tcl_ThreadDataKey dataKey;
+
+/*
+ * Helper macros for working with times. TCL_TIME_BEFORE encodes how to write
+ * the ordering relation on (normalized) times, and TCL_TIME_DIFF_MS computes
+ * the number of milliseconds difference between two times. Both macros use
+ * both of their arguments multiple times, so make sure they are cheap and
+ * side-effect free. The "prototypes" for these macros are:
+ *
+ * static int	TCL_TIME_BEFORE(Tcl_Time t1, Tcl_Time t2);
+ * static long	TCL_TIME_DIFF_MS(Tcl_Time t1, Tcl_Time t2);
+ */
+
+#define TCL_TIME_BEFORE(t1, t2) \
+    (((t1).sec<(t2).sec) || ((t1).sec==(t2).sec && (t1).usec<(t2).usec))
+
+#define TCL_TIME_DIFF_MS(t1, t2) \
+    (1000*((Tcl_WideInt)(t1).sec - (Tcl_WideInt)(t2).sec) + \
+	    ((long)(t1).usec - (long)(t2).usec)/1000)
+
+#define TCL_TIME_DIFF_MS_CEILING(t1, t2) \
+    (1000*((Tcl_WideInt)(t1).sec - (Tcl_WideInt)(t2).sec) + \
+	    ((long)(t1).usec - (long)(t2).usec + 999)/1000)
+
+/*
+ * Sleeps under that number of milliseconds don't get double-checked
+ * and are done in exactly one Tcl_Sleep(). This to limit gettimeofday()s.
+ */
+
+#define SLEEP_OFFLOAD_GETTIMEOFDAY 20
+
+/*
+ * The maximum number of milliseconds for each Tcl_Sleep call in AfterDelay.
+ * This is used to limit the maximum lag between interp limit and script
+ * cancellation checks.
+ */
+
+#define TCL_TIME_MAXIMUM_SLICE 500
+
+/*
+ * Prototypes for functions referenced only in this file:
+ */
+
+static void		AfterCleanupProc(ClientData clientData,
+			    Tcl_Interp *interp);
+static int		AfterDelay(Tcl_Interp *interp, Tcl_WideInt ms);
+static void		AfterProc(ClientData clientData);
+static void		FreeAfterPtr(AfterInfo *afterPtr);
+static AfterInfo *	GetAfterEvent(AfterAssocData *assocPtr,
+			    Tcl_Obj *commandPtr);
+static ThreadSpecificData *InitTimer(void);
+static void		TimerExitProc(ClientData clientData);
+static int		TimerHandlerEventProc(Tcl_Event *evPtr, int flags);
+static void		TimerCheckProc(ClientData clientData, int flags);
+static void		TimerSetupProc(ClientData clientData, int flags);
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InitTimer --
+ *
+ *	This function initializes the timer module.
+ *
+ * Results:
+ *	A pointer to the thread specific data.
+ *
+ * Side effects:
+ *	Registers the idle and timer event sources.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ThreadSpecificData *
+InitTimer(void)
+{
+    ThreadSpecificData *tsdPtr = TclThreadDataKeyGet(&dataKey);
+
+    if (tsdPtr == NULL) {
+	tsdPtr = TCL_TSD_INIT(&dataKey);
+	Tcl_CreateEventSource(TimerSetupProc, TimerCheckProc, NULL);
+	Tcl_CreateThreadExitHandler(TimerExitProc, NULL);
+    }
+    return tsdPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerExitProc --
+ *
+ *	This function is call at exit or unload time to remove the timer and
+ *	idle event sources.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Removes the timer and idle event sources and remaining events.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerExitProc(
+    ClientData clientData)	/* Not used. */
+{
+    ThreadSpecificData *tsdPtr = TclThreadDataKeyGet(&dataKey);
+
+    Tcl_DeleteEventSource(TimerSetupProc, TimerCheckProc, NULL);
+    if (tsdPtr != NULL) {
+	register TimerHandler *timerHandlerPtr;
+
+	timerHandlerPtr = tsdPtr->firstTimerHandlerPtr;
+	while (timerHandlerPtr != NULL) {
+	    tsdPtr->firstTimerHandlerPtr = timerHandlerPtr->nextPtr;
+	    ckfree(timerHandlerPtr);
+	    timerHandlerPtr = tsdPtr->firstTimerHandlerPtr;
+	}
+    }
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_CreateTimerHandler --
+ *
+ *	Arrange for a given function to be invoked at a particular time in the
+ *	future.
+ *
+ * Results:
+ *	The return value is a token for the timer event, which may be used to
+ *	delete the event before it fires.
+ *
+ * Side effects:
+ *	When milliseconds have elapsed, proc will be invoked exactly once.
+ *
+ *--------------------------------------------------------------
+ */
+
+Tcl_TimerToken
+Tcl_CreateTimerHandler(
+    int milliseconds,		/* How many milliseconds to wait before
+				 * invoking proc. */
+    Tcl_TimerProc *proc,	/* Function to invoke. */
+    ClientData clientData)	/* Arbitrary data to pass to proc. */
+{
+    Tcl_Time time;
+
+    /*
+     * Compute when the event should fire.
+     */
+
+    Tcl_GetTime(&time);
+    time.sec += milliseconds/1000;
+    time.usec += (milliseconds%1000)*1000;
+    if (time.usec >= 1000000) {
+	time.usec -= 1000000;
+	time.sec += 1;
+    }
+    return TclCreateAbsoluteTimerHandler(&time, proc, clientData);
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * TclCreateAbsoluteTimerHandler --
+ *
+ *	Arrange for a given function to be invoked at a particular time in the
+ *	future.
+ *
+ * Results:
+ *	The return value is a token for the timer event, which may be used to
+ *	delete the event before it fires.
+ *
+ * Side effects:
+ *	When the time in timePtr has been reached, proc will be invoked
+ *	exactly once.
+ *
+ *--------------------------------------------------------------
+ */
+
+Tcl_TimerToken
+TclCreateAbsoluteTimerHandler(
+    Tcl_Time *timePtr,
+    Tcl_TimerProc *proc,
+    ClientData clientData)
+{
+    register TimerHandler *timerHandlerPtr, *tPtr2, *prevPtr;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    timerHandlerPtr = ckalloc(sizeof(TimerHandler));
+
+    /*
+     * Fill in fields for the event.
+     */
+
+    memcpy(&timerHandlerPtr->time, timePtr, sizeof(Tcl_Time));
+    timerHandlerPtr->proc = proc;
+    timerHandlerPtr->clientData = clientData;
+    tsdPtr->lastTimerId++;
+    timerHandlerPtr->token = (Tcl_TimerToken) INT2PTR(tsdPtr->lastTimerId);
+
+    /*
+     * Add the event to the queue in the correct position
+     * (ordered by event firing time).
+     */
+
+    for (tPtr2 = tsdPtr->firstTimerHandlerPtr, prevPtr = NULL; tPtr2 != NULL;
+	    prevPtr = tPtr2, tPtr2 = tPtr2->nextPtr) {
+	if (TCL_TIME_BEFORE(timerHandlerPtr->time, tPtr2->time)) {
+	    break;
+	}
+    }
+    timerHandlerPtr->nextPtr = tPtr2;
+    if (prevPtr == NULL) {
+	tsdPtr->firstTimerHandlerPtr = timerHandlerPtr;
+    } else {
+	prevPtr->nextPtr = timerHandlerPtr;
+    }
+
+    TimerSetupProc(NULL, TCL_ALL_EVENTS);
+
+    return timerHandlerPtr->token;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_DeleteTimerHandler --
+ *
+ *	Delete a previously-registered timer handler.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Destroy the timer callback identified by TimerToken, so that its
+ *	associated function will not be called. If the callback has already
+ *	fired, or if the given token doesn't exist, then nothing happens.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteTimerHandler(
+    Tcl_TimerToken token)	/* Result previously returned by
+				 * Tcl_DeleteTimerHandler. */
+{
+    register TimerHandler *timerHandlerPtr, *prevPtr;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    if (token == NULL) {
+	return;
+    }
+
+    for (timerHandlerPtr = tsdPtr->firstTimerHandlerPtr, prevPtr = NULL;
+	    timerHandlerPtr != NULL; prevPtr = timerHandlerPtr,
+	    timerHandlerPtr = timerHandlerPtr->nextPtr) {
+	if (timerHandlerPtr->token != token) {
+	    continue;
+	}
+	if (prevPtr == NULL) {
+	    tsdPtr->firstTimerHandlerPtr = timerHandlerPtr->nextPtr;
+	} else {
+	    prevPtr->nextPtr = timerHandlerPtr->nextPtr;
+	}
+	ckfree(timerHandlerPtr);
+	return;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerSetupProc --
+ *
+ *	This function is called by Tcl_DoOneEvent to setup the timer event
+ *	source for before blocking. This routine checks both the idle and
+ *	after timer lists.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May update the maximum notifier block time.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerSetupProc(
+    ClientData data,		/* Not used. */
+    int flags)			/* Event flags as passed to Tcl_DoOneEvent. */
+{
+    Tcl_Time blockTime;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    if (((flags & TCL_IDLE_EVENTS) && tsdPtr->idleList)
+	    || ((flags & TCL_TIMER_EVENTS) && tsdPtr->timerPending)) {
+	/*
+	 * There is an idle handler or a pending timer event, so just poll.
+	 */
+
+	blockTime.sec = 0;
+	blockTime.usec = 0;
+    } else if ((flags & TCL_TIMER_EVENTS) && tsdPtr->firstTimerHandlerPtr) {
+	/*
+	 * Compute the timeout for the next timer on the list.
+	 */
+
+	Tcl_GetTime(&blockTime);
+	blockTime.sec = tsdPtr->firstTimerHandlerPtr->time.sec - blockTime.sec;
+	blockTime.usec = tsdPtr->firstTimerHandlerPtr->time.usec -
+		blockTime.usec;
+	if (blockTime.usec < 0) {
+	    blockTime.sec -= 1;
+	    blockTime.usec += 1000000;
+	}
+	if (blockTime.sec < 0) {
+	    blockTime.sec = 0;
+	    blockTime.usec = 0;
+	}
+    } else {
+	return;
+    }
+
+    Tcl_SetMaxBlockTime(&blockTime);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerCheckProc --
+ *
+ *	This function is called by Tcl_DoOneEvent to check the timer event
+ *	source for events. This routine checks both the idle and after timer
+ *	lists.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May queue an event and update the maximum notifier block time.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+TimerCheckProc(
+    ClientData data,		/* Not used. */
+    int flags)			/* Event flags as passed to Tcl_DoOneEvent. */
+{
+    Tcl_Event *timerEvPtr;
+    Tcl_Time blockTime;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    if ((flags & TCL_TIMER_EVENTS) && tsdPtr->firstTimerHandlerPtr) {
+	/*
+	 * Compute the timeout for the next timer on the list.
+	 */
+
+	Tcl_GetTime(&blockTime);
+	blockTime.sec = tsdPtr->firstTimerHandlerPtr->time.sec - blockTime.sec;
+	blockTime.usec = tsdPtr->firstTimerHandlerPtr->time.usec -
+		blockTime.usec;
+	if (blockTime.usec < 0) {
+	    blockTime.sec -= 1;
+	    blockTime.usec += 1000000;
+	}
+	if (blockTime.sec < 0) {
+	    blockTime.sec = 0;
+	    blockTime.usec = 0;
+	}
+
+	/*
+	 * If the first timer has expired, stick an event on the queue.
+	 */
+
+	if (blockTime.sec == 0 && blockTime.usec == 0 &&
+		!tsdPtr->timerPending) {
+	    tsdPtr->timerPending = 1;
+	    timerEvPtr = ckalloc(sizeof(Tcl_Event));
+	    timerEvPtr->proc = TimerHandlerEventProc;
+	    Tcl_QueueEvent(timerEvPtr, TCL_QUEUE_TAIL);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TimerHandlerEventProc --
+ *
+ *	This function is called by Tcl_ServiceEvent when a timer event reaches
+ *	the front of the event queue. This function handles the event by
+ *	invoking the callbacks for all timers that are ready.
+ *
+ * Results:
+ *	Returns 1 if the event was handled, meaning it should be removed from
+ *	the queue. Returns 0 if the event was not handled, meaning it should
+ *	stay on the queue. The only time the event isn't handled is if the
+ *	TCL_TIMER_EVENTS flag bit isn't set.
+ *
+ * Side effects:
+ *	Whatever the timer handler callback functions do.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TimerHandlerEventProc(
+    Tcl_Event *evPtr,		/* Event to service. */
+    int flags)			/* Flags that indicate what events to handle,
+				 * such as TCL_FILE_EVENTS. */
+{
+    TimerHandler *timerHandlerPtr, **nextPtrPtr;
+    Tcl_Time time;
+    int currentTimerId;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    /*
+     * Do nothing if timers aren't enabled. This leaves the event on the
+     * queue, so we will get to it as soon as ServiceEvents() is called with
+     * timers enabled.
+     */
+
+    if (!(flags & TCL_TIMER_EVENTS)) {
+	return 0;
+    }
+
+    /*
+     * The code below is trickier than it may look, for the following reasons:
+     *
+     * 1. New handlers can get added to the list while the current one is
+     *	  being processed. If new ones get added, we don't want to process
+     *	  them during this pass through the list to avoid starving other event
+     *	  sources. This is implemented using the token number in the handler:
+     *	  new handlers will have a newer token than any of the ones currently
+     *	  on the list.
+     * 2. The handler can call Tcl_DoOneEvent, so we have to remove the
+     *	  handler from the list before calling it. Otherwise an infinite loop
+     *	  could result.
+     * 3. Tcl_DeleteTimerHandler can be called to remove an element from the
+     *	  list while a handler is executing, so the list could change
+     *	  structure during the call.
+     * 4. Because we only fetch the current time before entering the loop, the
+     *	  only way a new timer will even be considered runnable is if its
+     *	  expiration time is within the same millisecond as the current time.
+     *	  This is fairly likely on Windows, since it has a course granularity
+     *	  clock. Since timers are placed on the queue in time order with the
+     *	  most recently created handler appearing after earlier ones with the
+     *	  same expiration time, we don't have to worry about newer generation
+     *	  timers appearing before later ones.
+     */
+
+    tsdPtr->timerPending = 0;
+    currentTimerId = tsdPtr->lastTimerId;
+    Tcl_GetTime(&time);
+    while (1) {
+	nextPtrPtr = &tsdPtr->firstTimerHandlerPtr;
+	timerHandlerPtr = tsdPtr->firstTimerHandlerPtr;
+	if (timerHandlerPtr == NULL) {
+	    break;
+	}
+
+	if (TCL_TIME_BEFORE(time, timerHandlerPtr->time)) {
+	    break;
+	}
+
+	/*
+	 * Bail out if the next timer is of a newer generation.
+	 */
+
+	if ((currentTimerId - PTR2INT(timerHandlerPtr->token)) < 0) {
+	    break;
+	}
+
+	/*
+	 * Remove the handler from the queue before invoking it, to avoid
+	 * potential reentrancy problems.
+	 */
+
+	*nextPtrPtr = timerHandlerPtr->nextPtr;
+	timerHandlerPtr->proc(timerHandlerPtr->clientData);
+	ckfree(timerHandlerPtr);
+    }
+    TimerSetupProc(NULL, TCL_TIMER_EVENTS);
+    return 1;
+}
+
+/*
+ *--------------------------------------------------------------
+ *
+ * Tcl_DoWhenIdle --
+ *
+ *	Arrange for proc to be invoked the next time the system is idle (i.e.,
+ *	just before the next time that Tcl_DoOneEvent would have to wait for
+ *	something to happen).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Proc will eventually be called, with clientData as argument. See the
+ *	manual entry for details.
+ *
+ *--------------------------------------------------------------
+ */
+
+void
+Tcl_DoWhenIdle(
+    Tcl_IdleProc *proc,		/* Function to invoke. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    register IdleHandler *idlePtr;
+    Tcl_Time blockTime;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    idlePtr = ckalloc(sizeof(IdleHandler));
+    idlePtr->proc = proc;
+    idlePtr->clientData = clientData;
+    idlePtr->generation = tsdPtr->idleGeneration;
+    idlePtr->nextPtr = NULL;
+    if (tsdPtr->lastIdlePtr == NULL) {
+	tsdPtr->idleList = idlePtr;
+    } else {
+	tsdPtr->lastIdlePtr->nextPtr = idlePtr;
+    }
+    tsdPtr->lastIdlePtr = idlePtr;
+
+    blockTime.sec = 0;
+    blockTime.usec = 0;
+    Tcl_SetMaxBlockTime(&blockTime);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CancelIdleCall --
+ *
+ *	If there are any when-idle calls requested to a given function with
+ *	given clientData, cancel all of them.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the proc/clientData combination were on the when-idle list, they
+ *	are removed so that they will never be called.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_CancelIdleCall(
+    Tcl_IdleProc *proc,		/* Function that was previously registered. */
+    ClientData clientData)	/* Arbitrary value to pass to proc. */
+{
+    register IdleHandler *idlePtr, *prevPtr;
+    IdleHandler *nextPtr;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    for (prevPtr = NULL, idlePtr = tsdPtr->idleList; idlePtr != NULL;
+	    prevPtr = idlePtr, idlePtr = idlePtr->nextPtr) {
+	while ((idlePtr->proc == proc)
+		&& (idlePtr->clientData == clientData)) {
+	    nextPtr = idlePtr->nextPtr;
+	    ckfree(idlePtr);
+	    idlePtr = nextPtr;
+	    if (prevPtr == NULL) {
+		tsdPtr->idleList = idlePtr;
+	    } else {
+		prevPtr->nextPtr = idlePtr;
+	    }
+	    if (idlePtr == NULL) {
+		tsdPtr->lastIdlePtr = prevPtr;
+		return;
+	    }
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclServiceIdle --
+ *
+ *	This function is invoked by the notifier when it becomes idle. It will
+ *	invoke all idle handlers that are present at the time the call is
+ *	invoked, but not those added during idle processing.
+ *
+ * Results:
+ *	The return value is 1 if TclServiceIdle found something to do,
+ *	otherwise return value is 0.
+ *
+ * Side effects:
+ *	Invokes all pending idle handlers.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclServiceIdle(void)
+{
+    IdleHandler *idlePtr;
+    int oldGeneration;
+    Tcl_Time blockTime;
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    if (tsdPtr->idleList == NULL) {
+	return 0;
+    }
+
+    oldGeneration = tsdPtr->idleGeneration;
+    tsdPtr->idleGeneration++;
+
+    /*
+     * The code below is trickier than it may look, for the following reasons:
+     *
+     * 1. New handlers can get added to the list while the current one is
+     *	  being processed. If new ones get added, we don't want to process
+     *	  them during this pass through the list (want to check for other work
+     *	  to do first). This is implemented using the generation number in the
+     *	  handler: new handlers will have a different generation than any of
+     *	  the ones currently on the list.
+     * 2. The handler can call Tcl_DoOneEvent, so we have to remove the
+     *	  handler from the list before calling it. Otherwise an infinite loop
+     *	  could result.
+     * 3. Tcl_CancelIdleCall can be called to remove an element from the list
+     *	  while a handler is executing, so the list could change structure
+     *	  during the call.
+     */
+
+    for (idlePtr = tsdPtr->idleList;
+	    ((idlePtr != NULL)
+		    && ((oldGeneration - idlePtr->generation) >= 0));
+	    idlePtr = tsdPtr->idleList) {
+	tsdPtr->idleList = idlePtr->nextPtr;
+	if (tsdPtr->idleList == NULL) {
+	    tsdPtr->lastIdlePtr = NULL;
+	}
+	idlePtr->proc(idlePtr->clientData);
+	ckfree(idlePtr);
+    }
+    if (tsdPtr->idleList) {
+	blockTime.sec = 0;
+	blockTime.usec = 0;
+	Tcl_SetMaxBlockTime(&blockTime);
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AfterObjCmd --
+ *
+ *	This function is invoked to process the "after" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_AfterObjCmd(
+    ClientData clientData,	/* Unused */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_WideInt ms = 0;		/* Number of milliseconds to wait */
+    Tcl_Time wakeup;
+    AfterInfo *afterPtr;
+    AfterAssocData *assocPtr;
+    int length;
+    int index;
+    static const char *const afterSubCmds[] = {
+	"cancel", "idle", "info", NULL
+    };
+    enum afterSubCmds {AFTER_CANCEL, AFTER_IDLE, AFTER_INFO};
+    ThreadSpecificData *tsdPtr = InitTimer();
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Create the "after" information associated for this interpreter, if it
+     * doesn't already exist.
+     */
+
+    assocPtr = Tcl_GetAssocData(interp, "tclAfter", NULL);
+    if (assocPtr == NULL) {
+	assocPtr = ckalloc(sizeof(AfterAssocData));
+	assocPtr->interp = interp;
+	assocPtr->firstAfterPtr = NULL;
+	Tcl_SetAssocData(interp, "tclAfter", AfterCleanupProc, assocPtr);
+    }
+
+    /*
+     * First lets see if the command was passed a number as the first argument.
+     */
+
+    if (objv[1]->typePtr == &tclIntType
+#ifndef TCL_WIDE_INT_IS_LONG
+	    || objv[1]->typePtr == &tclWideIntType
+#endif
+	    || objv[1]->typePtr == &tclBignumType
+	    || (Tcl_GetIndexFromObj(NULL, objv[1], afterSubCmds, "", 0,
+		    &index) != TCL_OK)) {
+	index = -1;
+	if (Tcl_GetWideIntFromObj(NULL, objv[1], &ms) != TCL_OK) {
+            const char *arg = Tcl_GetString(objv[1]);
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "bad argument \"%s\": must be"
+                    " cancel, idle, info, or an integer", arg));
+            Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "INDEX", "argument",
+                    arg, NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * At this point, either index = -1 and ms contains the number of ms
+     * to wait, or else index is the index of a subcommand.
+     */
+
+    switch (index) {
+    case -1: {
+	if (ms < 0) {
+	    ms = 0;
+	}
+	if (objc == 2) {
+	    return AfterDelay(interp, ms);
+	}
+	afterPtr = ckalloc(sizeof(AfterInfo));
+	afterPtr->assocPtr = assocPtr;
+	if (objc == 3) {
+	    afterPtr->commandPtr = objv[2];
+	} else {
+	    afterPtr->commandPtr = Tcl_ConcatObj(objc-2, objv+2);
+	}
+	Tcl_IncrRefCount(afterPtr->commandPtr);
+
+	/*
+	 * The variable below is used to generate unique identifiers for after
+	 * commands. This id can wrap around, which can potentially cause
+	 * problems. However, there are not likely to be problems in practice,
+	 * because after commands can only be requested to about a month in
+	 * the future, and wrap-around is unlikely to occur in less than about
+	 * 1-10 years. Thus it's unlikely that any old ids will still be
+	 * around when wrap-around occurs.
+	 */
+
+	afterPtr->id = tsdPtr->afterId;
+	tsdPtr->afterId += 1;
+	Tcl_GetTime(&wakeup);
+	wakeup.sec += (long)(ms / 1000);
+	wakeup.usec += ((long)(ms % 1000)) * 1000;
+	if (wakeup.usec > 1000000) {
+	    wakeup.sec++;
+	    wakeup.usec -= 1000000;
+	}
+	afterPtr->token = TclCreateAbsoluteTimerHandler(&wakeup,
+		AfterProc, afterPtr);
+	afterPtr->nextPtr = assocPtr->firstAfterPtr;
+	assocPtr->firstAfterPtr = afterPtr;
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf("after#%d", afterPtr->id));
+	return TCL_OK;
+    }
+    case AFTER_CANCEL: {
+	Tcl_Obj *commandPtr;
+	const char *command, *tempCommand;
+	int tempLength;
+
+	if (objc < 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "id|command");
+	    return TCL_ERROR;
+	}
+	if (objc == 3) {
+	    commandPtr = objv[2];
+	} else {
+	    commandPtr = Tcl_ConcatObj(objc-2, objv+2);;
+	}
+	command = TclGetStringFromObj(commandPtr, &length);
+	for (afterPtr = assocPtr->firstAfterPtr;  afterPtr != NULL;
+		afterPtr = afterPtr->nextPtr) {
+	    tempCommand = TclGetStringFromObj(afterPtr->commandPtr,
+		    &tempLength);
+	    if ((length == tempLength)
+		    && !memcmp(command, tempCommand, (unsigned) length)) {
+		break;
+	    }
+	}
+	if (afterPtr == NULL) {
+	    afterPtr = GetAfterEvent(assocPtr, commandPtr);
+	}
+	if (objc != 3) {
+	    Tcl_DecrRefCount(commandPtr);
+	}
+	if (afterPtr != NULL) {
+	    if (afterPtr->token != NULL) {
+		Tcl_DeleteTimerHandler(afterPtr->token);
+	    } else {
+		Tcl_CancelIdleCall(AfterProc, afterPtr);
+	    }
+	    FreeAfterPtr(afterPtr);
+	}
+	break;
+    }
+    case AFTER_IDLE:
+	if (objc < 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "script ?script ...?");
+	    return TCL_ERROR;
+	}
+	afterPtr = ckalloc(sizeof(AfterInfo));
+	afterPtr->assocPtr = assocPtr;
+	if (objc == 3) {
+	    afterPtr->commandPtr = objv[2];
+	} else {
+	    afterPtr->commandPtr = Tcl_ConcatObj(objc-2, objv+2);
+	}
+	Tcl_IncrRefCount(afterPtr->commandPtr);
+	afterPtr->id = tsdPtr->afterId;
+	tsdPtr->afterId += 1;
+	afterPtr->token = NULL;
+	afterPtr->nextPtr = assocPtr->firstAfterPtr;
+	assocPtr->firstAfterPtr = afterPtr;
+	Tcl_DoWhenIdle(AfterProc, afterPtr);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf("after#%d", afterPtr->id));
+	break;
+    case AFTER_INFO:
+	if (objc == 2) {
+            Tcl_Obj *resultObj = Tcl_NewObj();
+
+	    for (afterPtr = assocPtr->firstAfterPtr; afterPtr != NULL;
+		    afterPtr = afterPtr->nextPtr) {
+		if (assocPtr->interp == interp) {
+                    Tcl_ListObjAppendElement(NULL, resultObj, Tcl_ObjPrintf(
+                            "after#%d", afterPtr->id));
+		}
+	    }
+            Tcl_SetObjResult(interp, resultObj);
+	    return TCL_OK;
+	}
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?id?");
+	    return TCL_ERROR;
+	}
+	afterPtr = GetAfterEvent(assocPtr, objv[2]);
+	if (afterPtr == NULL) {
+            const char *eventStr = TclGetString(objv[2]);
+
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+                    "event \"%s\" doesn't exist", eventStr));
+            Tcl_SetErrorCode(interp, "TCL","LOOKUP","EVENT", eventStr, NULL);
+	    return TCL_ERROR;
+	} else {
+            Tcl_Obj *resultListPtr = Tcl_NewObj();
+
+            Tcl_ListObjAppendElement(interp, resultListPtr,
+                    afterPtr->commandPtr);
+            Tcl_ListObjAppendElement(interp, resultListPtr, Tcl_NewStringObj(
+		    (afterPtr->token == NULL) ? "idle" : "timer", -1));
+            Tcl_SetObjResult(interp, resultListPtr);
+        }
+	break;
+    default:
+	Tcl_Panic("Tcl_AfterObjCmd: bad subcommand index to afterSubCmds");
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AfterDelay --
+ *
+ *	Implements the blocking delay behaviour of [after $time]. Tricky
+ *	because it has to take into account any time limit that has been set.
+ *
+ * Results:
+ *	Standard Tcl result code (with error set if an error occurred due to a
+ *	time limit being exceeded or being canceled).
+ *
+ * Side effects:
+ *	May adjust the time limit granularity marker.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+AfterDelay(
+    Tcl_Interp *interp,
+    Tcl_WideInt ms)
+{
+    Interp *iPtr = (Interp *) interp;
+
+    Tcl_Time endTime, now;
+    Tcl_WideInt diff;
+
+    Tcl_GetTime(&now);
+    endTime = now;
+    endTime.sec += (long)(ms/1000);
+    endTime.usec += ((int)(ms%1000))*1000;
+    if (endTime.usec >= 1000000) {
+	endTime.sec++;
+	endTime.usec -= 1000000;
+    }
+
+    do {
+	if (Tcl_AsyncReady()) {
+	    if (Tcl_AsyncInvoke(interp, TCL_OK) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	if (Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	if (iPtr->limit.timeEvent != NULL
+		&& TCL_TIME_BEFORE(iPtr->limit.time, now)) {
+	    iPtr->limit.granularityTicker = 0;
+	    if (Tcl_LimitCheck(interp) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	if (iPtr->limit.timeEvent == NULL
+		|| TCL_TIME_BEFORE(endTime, iPtr->limit.time)) {
+	    diff = TCL_TIME_DIFF_MS_CEILING(endTime, now);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    if (diff > LONG_MAX) {
+		diff = LONG_MAX;
+	    }
+#endif
+	    if (diff > TCL_TIME_MAXIMUM_SLICE) {
+		diff = TCL_TIME_MAXIMUM_SLICE;
+	    }
+            if (diff == 0 && TCL_TIME_BEFORE(now, endTime)) {
+                diff = 1;
+            }
+	    if (diff > 0) {
+		Tcl_Sleep((long) diff);
+                if (diff < SLEEP_OFFLOAD_GETTIMEOFDAY) {
+                    break;
+                }
+	    } else {
+                break;
+            }
+	} else {
+	    diff = TCL_TIME_DIFF_MS(iPtr->limit.time, now);
+#ifndef TCL_WIDE_INT_IS_LONG
+	    if (diff > LONG_MAX) {
+		diff = LONG_MAX;
+	    }
+#endif
+	    if (diff > TCL_TIME_MAXIMUM_SLICE) {
+		diff = TCL_TIME_MAXIMUM_SLICE;
+	    }
+	    if (diff > 0) {
+		Tcl_Sleep((long) diff);
+	    }
+	    if (Tcl_AsyncReady()) {
+		if (Tcl_AsyncInvoke(interp, TCL_OK) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+	    }
+	    if (Tcl_Canceled(interp, TCL_LEAVE_ERR_MSG) == TCL_ERROR) {
+		return TCL_ERROR;
+	    }
+	    if (Tcl_LimitCheck(interp) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+        Tcl_GetTime(&now);
+    } while (TCL_TIME_BEFORE(now, endTime));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetAfterEvent --
+ *
+ *	This function parses an "after" id such as "after#4" and returns a
+ *	pointer to the AfterInfo structure.
+ *
+ * Results:
+ *	The return value is either a pointer to an AfterInfo structure, if one
+ *	is found that corresponds to "cmdString" and is for interp, or NULL if
+ *	no corresponding after event can be found.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static AfterInfo *
+GetAfterEvent(
+    AfterAssocData *assocPtr,	/* Points to "after"-related information for
+				 * this interpreter. */
+    Tcl_Obj *commandPtr)
+{
+    const char *cmdString;	/* Textual identifier for after event, such as
+				 * "after#6". */
+    AfterInfo *afterPtr;
+    int id;
+    char *end;
+
+    cmdString = TclGetString(commandPtr);
+    if (strncmp(cmdString, "after#", 6) != 0) {
+	return NULL;
+    }
+    cmdString += 6;
+    id = strtoul(cmdString, &end, 10);
+    if ((end == cmdString) || (*end != 0)) {
+	return NULL;
+    }
+    for (afterPtr = assocPtr->firstAfterPtr; afterPtr != NULL;
+	    afterPtr = afterPtr->nextPtr) {
+	if (afterPtr->id == id) {
+	    return afterPtr;
+	}
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AfterProc --
+ *
+ *	Timer callback to execute commands registered with the "after"
+ *	command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Executes whatever command was specified. If the command returns an
+ *	error, then the command "bgerror" is invoked to process the error; if
+ *	bgerror fails then information about the error is output on stderr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AfterProc(
+    ClientData clientData)	/* Describes command to execute. */
+{
+    AfterInfo *afterPtr = clientData;
+    AfterAssocData *assocPtr = afterPtr->assocPtr;
+    AfterInfo *prevPtr;
+    int result;
+    Tcl_Interp *interp;
+
+    /*
+     * First remove the callback from our list of callbacks; otherwise someone
+     * could delete the callback while it's being executed, which could cause
+     * a core dump.
+     */
+
+    if (assocPtr->firstAfterPtr == afterPtr) {
+	assocPtr->firstAfterPtr = afterPtr->nextPtr;
+    } else {
+	for (prevPtr = assocPtr->firstAfterPtr; prevPtr->nextPtr != afterPtr;
+		prevPtr = prevPtr->nextPtr) {
+	    /* Empty loop body. */
+	}
+	prevPtr->nextPtr = afterPtr->nextPtr;
+    }
+
+    /*
+     * Execute the callback.
+     */
+
+    interp = assocPtr->interp;
+    Tcl_Preserve(interp);
+    result = Tcl_EvalObjEx(interp, afterPtr->commandPtr, TCL_EVAL_GLOBAL);
+    if (result != TCL_OK) {
+	Tcl_AddErrorInfo(interp, "\n    (\"after\" script)");
+	Tcl_BackgroundException(interp, result);
+    }
+    Tcl_Release(interp);
+
+    /*
+     * Free the memory for the callback.
+     */
+
+    Tcl_DecrRefCount(afterPtr->commandPtr);
+    ckfree(afterPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeAfterPtr --
+ *
+ *	This function removes an "after" command from the list of those that
+ *	are pending and frees its resources. This function does *not* cancel
+ *	the timer handler; if that's needed, the caller must do it.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The memory associated with afterPtr is released.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeAfterPtr(
+    AfterInfo *afterPtr)		/* Command to be deleted. */
+{
+    AfterInfo *prevPtr;
+    AfterAssocData *assocPtr = afterPtr->assocPtr;
+
+    if (assocPtr->firstAfterPtr == afterPtr) {
+	assocPtr->firstAfterPtr = afterPtr->nextPtr;
+    } else {
+	for (prevPtr = assocPtr->firstAfterPtr; prevPtr->nextPtr != afterPtr;
+		prevPtr = prevPtr->nextPtr) {
+	    /* Empty loop body. */
+	}
+	prevPtr->nextPtr = afterPtr->nextPtr;
+    }
+    Tcl_DecrRefCount(afterPtr->commandPtr);
+    ckfree(afterPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AfterCleanupProc --
+ *
+ *	This function is invoked whenever an interpreter is deleted
+ *	to cleanup the AssocData for "tclAfter".
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	After commands are removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static void
+AfterCleanupProc(
+    ClientData clientData,	/* Points to AfterAssocData for the
+				 * interpreter. */
+    Tcl_Interp *interp)		/* Interpreter that is being deleted. */
+{
+    AfterAssocData *assocPtr = clientData;
+    AfterInfo *afterPtr;
+
+    while (assocPtr->firstAfterPtr != NULL) {
+	afterPtr = assocPtr->firstAfterPtr;
+	assocPtr->firstAfterPtr = afterPtr->nextPtr;
+	if (afterPtr->token != NULL) {
+	    Tcl_DeleteTimerHandler(afterPtr->token);
+	} else {
+	    Tcl_CancelIdleCall(AfterProc, afterPtr);
+	}
+	Tcl_DecrRefCount(afterPtr->commandPtr);
+	ckfree(afterPtr);
+    }
+    ckfree(assocPtr);
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/generic/tclTomMath.decls b/generic/tclTomMath.decls
new file mode 100644
index 0000000..74ccefc
--- /dev/null
+++ b/generic/tclTomMath.decls
@@ -0,0 +1,243 @@
+# tclTomMath.decls --
+#
+#	This file contains the declarations for the functions in 'libtommath'
+#	that are contained within the Tcl library.  This file is used to
+#	generate the 'tclTomMathDecls.h' and 'tclStubInit.c' files.
+#
+# If you edit this file, advance the revision number (and the epoch
+# if the new stubs are not backward compatible) in tclTomMathDecls.h
+#
+# Copyright (c) 2005 by Kevin B. Kenny.  All rights reserved.
+#
+# See the file "license.terms" for information on usage and redistribution
+# of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+
+library tcl
+
+# Define the unsupported generic interfaces.
+
+interface tclTomMath
+# hooks {tclTomMathInt}
+scspec EXTERN
+
+# Declare each of the functions in the Tcl tommath interface
+
+declare 0 {
+    int TclBN_epoch(void)
+}
+declare 1 {
+    int TclBN_revision(void)
+}
+
+declare 2 {
+    int TclBN_mp_add(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 3 {
+    int TclBN_mp_add_d(mp_int *a, mp_digit b, mp_int *c)
+}
+declare 4 {
+    int TclBN_mp_and(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 5 {
+    void TclBN_mp_clamp(mp_int *a)
+}
+declare 6 {
+    void TclBN_mp_clear(mp_int *a)
+}
+declare 7 {
+    void TclBN_mp_clear_multi(mp_int *a, ...)
+}
+declare 8 {
+    int TclBN_mp_cmp(const mp_int *a, const mp_int *b)
+}
+declare 9 {
+    int TclBN_mp_cmp_d(const mp_int *a, mp_digit b)
+}
+declare 10 {
+    int TclBN_mp_cmp_mag(const mp_int *a, const mp_int *b)
+}
+declare 11 {
+    int TclBN_mp_copy(const mp_int *a, mp_int *b)
+}
+declare 12 {
+    int TclBN_mp_count_bits(const mp_int *a)
+}
+declare 13 {
+    int TclBN_mp_div(mp_int *a, mp_int *b, mp_int *q, mp_int *r)
+}
+declare 14 {
+    int TclBN_mp_div_d(mp_int *a, mp_digit b, mp_int *q, mp_digit *r)
+}
+declare 15 {
+    int TclBN_mp_div_2(mp_int *a, mp_int *q)
+}
+declare 16 {
+    int TclBN_mp_div_2d(const mp_int *a, int b, mp_int *q, mp_int *r)
+}
+declare 17 {
+    int TclBN_mp_div_3(mp_int *a, mp_int *q, mp_digit *r)
+}
+declare 18 {
+    void TclBN_mp_exch(mp_int *a, mp_int *b)
+}
+declare 19 {
+    int TclBN_mp_expt_d(mp_int *a, mp_digit b, mp_int *c)
+}
+declare 20 {
+    int TclBN_mp_grow(mp_int *a, int size)
+}
+declare 21 {
+    int TclBN_mp_init(mp_int *a)
+}
+declare 22 {
+    int TclBN_mp_init_copy(mp_int *a, const mp_int *b)
+}
+declare 23 {
+    int TclBN_mp_init_multi(mp_int *a, ...)
+}
+declare 24 {
+    int TclBN_mp_init_set(mp_int *a, mp_digit b)
+}
+declare 25 {
+    int TclBN_mp_init_size(mp_int *a, int size)
+}
+declare 26 {
+    int TclBN_mp_lshd(mp_int *a, int shift)
+}
+declare 27 {
+    int TclBN_mp_mod(mp_int *a, mp_int *b, mp_int *r)
+}
+declare 28 {
+    int TclBN_mp_mod_2d(const mp_int *a, int b, mp_int *r)
+}
+declare 29 {
+    int TclBN_mp_mul(mp_int *a, mp_int *b, mp_int *p)
+}
+declare 30 {
+    int TclBN_mp_mul_d(mp_int *a, mp_digit b, mp_int *p)
+}
+declare 31 {
+    int TclBN_mp_mul_2(mp_int *a, mp_int *p)
+}
+declare 32 {
+    int TclBN_mp_mul_2d(const mp_int *a, int d, mp_int *p)
+}
+declare 33 {
+    int TclBN_mp_neg(const mp_int *a, mp_int *b)
+}
+declare 34 {
+    int TclBN_mp_or(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 35 {
+    int TclBN_mp_radix_size(const mp_int *a, int radix, int *size)
+}
+declare 36 {
+    int TclBN_mp_read_radix(mp_int *a, const char *str, int radix)
+}
+declare 37 {
+    void TclBN_mp_rshd(mp_int *a, int shift)
+}
+declare 38 {
+    int TclBN_mp_shrink(mp_int *a)
+}
+declare 39 {
+    void TclBN_mp_set(mp_int *a, mp_digit b)
+}
+declare 40 {
+    int TclBN_mp_sqr(mp_int *a, mp_int *b)
+}
+declare 41 {
+    int TclBN_mp_sqrt(mp_int *a, mp_int *b)
+}
+declare 42 {
+    int TclBN_mp_sub(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 43 {
+    int TclBN_mp_sub_d(mp_int *a, mp_digit b, mp_int *c)
+}
+declare 44 {
+    int TclBN_mp_to_unsigned_bin(mp_int *a, unsigned char *b)
+}
+declare 45 {
+    int TclBN_mp_to_unsigned_bin_n(mp_int *a, unsigned char *b,
+	    unsigned long *outlen)
+}
+declare 46 {
+    int TclBN_mp_toradix_n(mp_int *a, char *str, int radix, int maxlen)
+}
+declare 47 {
+    int TclBN_mp_unsigned_bin_size(mp_int *a)
+}
+declare 48 {
+    int TclBN_mp_xor(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 49 {
+    void TclBN_mp_zero(mp_int *a)
+}
+
+# internal routines to libtommath - should not be called but must be
+# exported to accommodate the "tommath" extension
+
+declare 50 {
+    void TclBN_reverse(unsigned char *s, int len)
+}
+declare 51 {
+    int TclBN_fast_s_mp_mul_digs(mp_int *a, mp_int *b, mp_int *c, int digs)
+}
+declare 52 {
+    int TclBN_fast_s_mp_sqr(mp_int *a, mp_int *b)
+}
+declare 53 {
+    int TclBN_mp_karatsuba_mul(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 54 {
+    int TclBN_mp_karatsuba_sqr(mp_int *a, mp_int *b)
+}
+declare 55 {
+    int TclBN_mp_toom_mul(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 56 {
+    int TclBN_mp_toom_sqr(mp_int *a, mp_int *b)
+}
+declare 57 {
+    int TclBN_s_mp_add(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 58 {
+    int TclBN_s_mp_mul_digs(mp_int *a, mp_int *b, mp_int *c, int digs)
+}
+declare 59 {
+    int TclBN_s_mp_sqr(mp_int *a, mp_int *b)
+}
+declare 60 {
+    int TclBN_s_mp_sub(mp_int *a, mp_int *b, mp_int *c)
+}
+declare 61 {
+    int TclBN_mp_init_set_int(mp_int *a, unsigned long i)
+}
+declare 62 {
+    int TclBN_mp_set_int(mp_int *a, unsigned long i)
+}
+declare 63 {
+    int TclBN_mp_cnt_lsb(const mp_int *a)
+}
+
+# Formerly internal API to allow initialisation of bignums without knowing the
+# typedefs of how a bignum works internally.
+declare 64 {
+    void TclBNInitBignumFromLong(mp_int *bignum, long initVal)
+}
+declare 65 {
+    void TclBNInitBignumFromWideInt(mp_int *bignum, Tcl_WideInt initVal)
+}
+declare 66 {
+    void TclBNInitBignumFromWideUInt(mp_int *bignum, Tcl_WideUInt initVal)
+}
+
+# Added in libtommath 1.0
+declare 67 {
+    int TclBN_mp_expt_d_ex(mp_int *a, mp_digit b, mp_int *c, int fast)
+}
+
+# Local Variables:
+# mode: tcl
+# End:
diff --git a/generic/tclTomMath.h b/generic/tclTomMath.h
new file mode 100644
index 0000000..87fe756
--- /dev/null
+++ b/generic/tclTomMath.h
@@ -0,0 +1,792 @@
+/* LibTomMath, multiple-precision integer library -- Tom St Denis
+ *
+ * LibTomMath is a library that provides multiple-precision
+ * integer arithmetic as well as number theoretic functionality.
+ *
+ * The library was designed directly after the MPI library by
+ * Michael Fromberger but has been written from scratch with
+ * additional optimizations in place.
+ *
+ * The library is free for all purposes without any express
+ * guarantee it works.
+ *
+ * Tom St Denis, tstdenis82@gmail.com, http://math.libtomcrypt.com
+ */
+#ifndef BN_H_
+#define BN_H_
+
+#include "tclTomMathDecls.h"
+#ifndef MODULE_SCOPE
+#define MODULE_SCOPE extern
+#endif
+
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* detect 64-bit mode if possible */
+#if defined(NEVER) /* 128-bit ints fail in too many places */
+   #if !(defined(MP_32BIT) || defined(MP_16BIT) || defined(MP_8BIT))
+      #define MP_64BIT
+   #endif
+#endif
+
+/* some default configurations.
+ *
+ * A "mp_digit" must be able to hold DIGIT_BIT + 1 bits
+ * A "mp_word" must be able to hold 2*DIGIT_BIT + 1 bits
+ *
+ * At the very least a mp_digit must be able to hold 7 bits
+ * [any size beyond that is ok provided it doesn't overflow the data type]
+ */
+#ifdef MP_8BIT
+#ifndef MP_DIGIT_DECLARED
+   typedef uint8_t              mp_digit;
+#define MP_DIGIT_DECLARED
+#endif
+   typedef uint16_t             mp_word;
+#define MP_SIZEOF_MP_DIGIT      1
+#ifdef DIGIT_BIT
+#error You must not define DIGIT_BIT when using MP_8BIT
+#endif
+#elif defined(MP_16BIT)
+#ifndef MP_DIGIT_DECLARED
+   typedef uint16_t             mp_digit;
+#define MP_DIGIT_DECLARED
+#endif
+   typedef uint32_t             mp_word;
+#define MP_SIZEOF_MP_DIGIT      2
+#ifdef DIGIT_BIT
+#error You must not define DIGIT_BIT when using MP_16BIT
+#endif
+#elif defined(MP_64BIT)
+   /* for GCC only on supported platforms */
+#ifndef MP_DIGIT_DECLARED
+   typedef uint64_t mp_digit;
+#define MP_DIGIT_DECLARED
+#endif
+#if defined(_WIN32)
+   typedef unsigned __int128    mp_word;
+#elif defined(__GNUC__)
+   typedef unsigned long        mp_word __attribute__ ((mode(TI)));
+#else
+   /* it seems you have a problem
+    * but we assume you can somewhere define your own uint128_t */
+   typedef uint128_t            mp_word;
+#endif
+
+   #define DIGIT_BIT            60
+#else
+   /* this is the default case, 28-bit digits */
+
+   /* this is to make porting into LibTomCrypt easier :-) */
+#ifndef MP_DIGIT_DECLARED
+   typedef uint32_t             mp_digit;
+#define MP_DIGIT_DECLARED
+#endif
+   typedef uint64_t             mp_word;
+
+#ifdef MP_31BIT
+   /* this is an extension that uses 31-bit digits */
+   #define DIGIT_BIT            31
+#else
+   /* default case is 28-bit digits, defines MP_28BIT as a handy macro to test */
+   #define DIGIT_BIT            28
+   #define MP_28BIT
+#endif
+#endif
+
+/* otherwise the bits per digit is calculated automatically from the size of a mp_digit */
+#ifndef DIGIT_BIT
+   #define DIGIT_BIT     (((CHAR_BIT * MP_SIZEOF_MP_DIGIT) - 1))  /* bits per digit */
+   typedef uint_least32_t mp_min_u32;
+#else
+   typedef mp_digit mp_min_u32;
+#endif
+
+/* use arc4random on platforms that support it */
+#if defined(__FreeBSD__) || defined(__OpenBSD__) || defined(__NetBSD__) || defined(__DragonFly__)
+    #define MP_GEN_RANDOM()    arc4random()
+    #define MP_GEN_RANDOM_MAX  0xffffffff
+#endif
+
+/* use rand() as fall-back if there's no better rand function */
+#ifndef MP_GEN_RANDOM
+    #define MP_GEN_RANDOM()    rand()
+    #define MP_GEN_RANDOM_MAX  RAND_MAX
+#endif
+
+#define MP_DIGIT_BIT     DIGIT_BIT
+#define MP_MASK          ((((mp_digit)1)<<((mp_digit)DIGIT_BIT))-((mp_digit)1))
+#define MP_DIGIT_MAX     MP_MASK
+
+/* equalities */
+#define MP_LT        -1   /* less than */
+#define MP_EQ         0   /* equal to */
+#define MP_GT         1   /* greater than */
+
+#define MP_ZPOS       0   /* positive integer */
+#define MP_NEG        1   /* negative */
+
+#define MP_OKAY       0   /* ok result */
+#define MP_MEM        -2  /* out of mem */
+#define MP_VAL        -3  /* invalid input */
+#define MP_RANGE      MP_VAL
+
+#define MP_YES        1   /* yes response */
+#define MP_NO         0   /* no response */
+
+/* Primality generation flags */
+#define LTM_PRIME_BBS      0x0001 /* BBS style prime */
+#define LTM_PRIME_SAFE     0x0002 /* Safe prime (p-1)/2 == prime */
+#define LTM_PRIME_2MSB_ON  0x0008 /* force 2nd MSB to 1 */
+
+typedef int           mp_err;
+
+/* you'll have to tune these... */
+#if defined(BUILD_tcl) || !defined(_WIN32)
+MODULE_SCOPE int KARATSUBA_MUL_CUTOFF,
+           KARATSUBA_SQR_CUTOFF,
+           TOOM_MUL_CUTOFF,
+           TOOM_SQR_CUTOFF;
+#endif
+
+/* define this to use lower memory usage routines (exptmods mostly) */
+/* #define MP_LOW_MEM */
+
+/* default precision */
+#ifndef MP_PREC
+   #ifndef MP_LOW_MEM
+      #define MP_PREC                 32     /* default digits of precision */
+   #else
+      #define MP_PREC                 8      /* default digits of precision */
+   #endif
+#endif
+
+/* size of comba arrays, should be at least 2 * 2**(BITS_PER_WORD - BITS_PER_DIGIT*2) */
+#define MP_WARRAY               (1 << (((sizeof(mp_word) * CHAR_BIT) - (2 * DIGIT_BIT)) + 1))
+
+/* the infamous mp_int structure */
+#ifndef MP_INT_DECLARED
+#define MP_INT_DECLARED
+typedef struct mp_int mp_int;
+#endif
+struct mp_int {
+    int used, alloc, sign;
+    mp_digit *dp;
+};
+
+/* callback for mp_prime_random, should fill dst with random bytes and return how many read [upto len] */
+typedef int ltm_prime_callback(unsigned char *dst, int len, void *dat);
+
+
+#define USED(m)    ((m)->used)
+#define DIGIT(m,k) ((m)->dp[(k)])
+#define SIGN(m)    ((m)->sign)
+
+/* error code to char* string */
+const char *mp_error_to_string(int code);
+
+/* ---> init and deinit bignum functions <--- */
+/* init a bignum */
+/*
+int mp_init(mp_int *a);
+*/
+
+/* free a bignum */
+/*
+void mp_clear(mp_int *a);
+*/
+
+/* init a null terminated series of arguments */
+/*
+int mp_init_multi(mp_int *mp, ...);
+*/
+
+/* clear a null terminated series of arguments */
+/*
+void mp_clear_multi(mp_int *mp, ...);
+*/
+
+/* exchange two ints */
+/*
+void mp_exch(mp_int *a, mp_int *b);
+*/
+
+/* shrink ram required for a bignum */
+/*
+int mp_shrink(mp_int *a);
+*/
+
+/* grow an int to a given size */
+/*
+int mp_grow(mp_int *a, int size);
+*/
+
+/* init to a given number of digits */
+/*
+int mp_init_size(mp_int *a, int size);
+*/
+
+/* ---> Basic Manipulations <--- */
+#define mp_iszero(a) (((a)->used == 0) ? MP_YES : MP_NO)
+#define mp_iseven(a) ((((a)->used == 0) || (((a)->dp[0] & 1u) == 0u)) ? MP_YES : MP_NO)
+#define mp_isodd(a)  ((((a)->used > 0) && (((a)->dp[0] & 1u) == 1u)) ? MP_YES : MP_NO)
+#define mp_isneg(a)  (((a)->sign != MP_ZPOS) ? MP_YES : MP_NO)
+
+/* set to zero */
+/*
+void mp_zero(mp_int *a);
+*/
+
+/* set to a digit */
+/*
+void mp_set(mp_int *a, mp_digit b);
+*/
+
+/* set a 32-bit const */
+/*
+int mp_set_int(mp_int *a, unsigned long b);
+*/
+
+/* set a platform dependent unsigned long value */
+/*
+int mp_set_long(mp_int *a, unsigned long b);
+*/
+
+/* set a platform dependent unsigned long long value */
+/*
+int mp_set_long_long(mp_int *a, unsigned long long b);
+*/
+
+/* get a 32-bit value */
+unsigned long mp_get_int(mp_int * a);
+
+/* get a platform dependent unsigned long value */
+unsigned long mp_get_long(mp_int * a);
+
+/* get a platform dependent unsigned long long value */
+unsigned long long mp_get_long_long(mp_int * a);
+
+/* initialize and set a digit */
+/*
+int mp_init_set (mp_int * a, mp_digit b);
+*/
+
+/* initialize and set 32-bit value */
+/*
+int mp_init_set_int (mp_int * a, unsigned long b);
+*/
+
+/* copy, b = a */
+/*
+int mp_copy(const mp_int *a, mp_int *b);
+*/
+
+/* inits and copies, a = b */
+/*
+int mp_init_copy(mp_int *a, const mp_int *b);
+*/
+
+/* trim unused digits */
+/*
+void mp_clamp(mp_int *a);
+*/
+
+/* import binary data */
+/*
+int mp_import(mp_int* rop, size_t count, int order, size_t size, int endian, size_t nails, const void* op);
+*/
+
+/* export binary data */
+/*
+int mp_export(void* rop, size_t* countp, int order, size_t size, int endian, size_t nails, mp_int* op);
+*/
+
+/* ---> digit manipulation <--- */
+
+/* right shift by "b" digits */
+/*
+void mp_rshd(mp_int *a, int b);
+*/
+
+/* left shift by "b" digits */
+/*
+int mp_lshd(mp_int *a, int b);
+*/
+
+/* c = a / 2**b, implemented as c = a >> b */
+/*
+int mp_div_2d(const mp_int *a, int b, mp_int *c, mp_int *d);
+*/
+
+/* b = a/2 */
+/*
+int mp_div_2(mp_int *a, mp_int *b);
+*/
+
+/* c = a * 2**b, implemented as c = a << b */
+/*
+int mp_mul_2d(const mp_int *a, int b, mp_int *c);
+*/
+
+/* b = a*2 */
+/*
+int mp_mul_2(mp_int *a, mp_int *b);
+*/
+
+/* c = a mod 2**b */
+/*
+int mp_mod_2d(const mp_int *a, int b, mp_int *c);
+*/
+
+/* computes a = 2**b */
+/*
+int mp_2expt(mp_int *a, int b);
+*/
+
+/* Counts the number of lsbs which are zero before the first zero bit */
+/*
+int mp_cnt_lsb(const mp_int *a);
+*/
+
+/* I Love Earth! */
+
+/* makes a pseudo-random int of a given size */
+/*
+int mp_rand(mp_int *a, int digits);
+*/
+
+/* ---> binary operations <--- */
+/* c = a XOR b  */
+/*
+int mp_xor(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = a OR b */
+/*
+int mp_or(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = a AND b */
+/*
+int mp_and(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* ---> Basic arithmetic <--- */
+
+/* b = -a */
+/*
+int mp_neg(const mp_int *a, mp_int *b);
+*/
+
+/* b = |a| */
+/*
+int mp_abs(mp_int *a, mp_int *b);
+*/
+
+/* compare a to b */
+/*
+int mp_cmp(const mp_int *a, const mp_int *b);
+*/
+
+/* compare |a| to |b| */
+/*
+int mp_cmp_mag(const mp_int *a, const mp_int *b);
+*/
+
+/* c = a + b */
+/*
+int mp_add(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = a - b */
+/*
+int mp_sub(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = a * b */
+/*
+int mp_mul(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* b = a*a  */
+/*
+int mp_sqr(mp_int *a, mp_int *b);
+*/
+
+/* a/b => cb + d == a */
+/*
+int mp_div(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
+*/
+
+/* c = a mod b, 0 <= c < b  */
+/*
+int mp_mod(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* ---> single digit functions <--- */
+
+/* compare against a single digit */
+/*
+int mp_cmp_d(const mp_int *a, mp_digit b);
+*/
+
+/* c = a + b */
+/*
+int mp_add_d(mp_int *a, mp_digit b, mp_int *c);
+*/
+
+/* c = a - b */
+/*
+int mp_sub_d(mp_int *a, mp_digit b, mp_int *c);
+*/
+
+/* c = a * b */
+/*
+int mp_mul_d(mp_int *a, mp_digit b, mp_int *c);
+*/
+
+/* a/b => cb + d == a */
+/*
+int mp_div_d(mp_int *a, mp_digit b, mp_int *c, mp_digit *d);
+*/
+
+/* a/3 => 3c + d == a */
+/*
+int mp_div_3(mp_int *a, mp_int *c, mp_digit *d);
+*/
+
+/* c = a**b */
+/*
+int mp_expt_d(mp_int *a, mp_digit b, mp_int *c);
+*/
+/*
+int mp_expt_d_ex (mp_int * a, mp_digit b, mp_int * c, int fast);
+*/
+
+/* c = a mod b, 0 <= c < b  */
+/*
+int mp_mod_d(mp_int *a, mp_digit b, mp_digit *c);
+*/
+
+/* ---> number theory <--- */
+
+/* d = a + b (mod c) */
+/*
+int mp_addmod(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
+*/
+
+/* d = a - b (mod c) */
+/*
+int mp_submod(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
+*/
+
+/* d = a * b (mod c) */
+/*
+int mp_mulmod(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
+*/
+
+/* c = a * a (mod b) */
+/*
+int mp_sqrmod(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = 1/a (mod b) */
+/*
+int mp_invmod(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* c = (a, b) */
+/*
+int mp_gcd(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* produces value such that U1*a + U2*b = U3 */
+/*
+int mp_exteuclid(mp_int *a, mp_int *b, mp_int *U1, mp_int *U2, mp_int *U3);
+*/
+
+/* c = [a, b] or (a*b)/(a, b) */
+/*
+int mp_lcm(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* finds one of the b'th root of a, such that |c|**b <= |a|
+ *
+ * returns error if a < 0 and b is even
+ */
+/*
+int mp_n_root(mp_int *a, mp_digit b, mp_int *c);
+*/
+/*
+int mp_n_root_ex (mp_int * a, mp_digit b, mp_int * c, int fast);
+*/
+
+/* special sqrt algo */
+/*
+int mp_sqrt(mp_int *arg, mp_int *ret);
+*/
+
+/* special sqrt (mod prime) */
+/*
+int mp_sqrtmod_prime(mp_int *arg, mp_int *prime, mp_int *ret);
+*/
+
+/* is number a square? */
+/*
+int mp_is_square(mp_int *arg, int *ret);
+*/
+
+/* computes the jacobi c = (a | n) (or Legendre if b is prime)  */
+/*
+int mp_jacobi(mp_int *a, mp_int *n, int *c);
+*/
+
+/* used to setup the Barrett reduction for a given modulus b */
+/*
+int mp_reduce_setup(mp_int *a, mp_int *b);
+*/
+
+/* Barrett Reduction, computes a (mod b) with a precomputed value c
+ *
+ * Assumes that 0 < a <= b*b, note if 0 > a > -(b*b) then you can merely
+ * compute the reduction as -1 * mp_reduce(mp_abs(a)) [pseudo code].
+ */
+/*
+int mp_reduce(mp_int *a, mp_int *b, mp_int *c);
+*/
+
+/* setups the montgomery reduction */
+/*
+int mp_montgomery_setup(mp_int *a, mp_digit *mp);
+*/
+
+/* computes a = B**n mod b without division or multiplication useful for
+ * normalizing numbers in a Montgomery system.
+ */
+/*
+int mp_montgomery_calc_normalization(mp_int *a, mp_int *b);
+*/
+
+/* computes x/R == x (mod N) via Montgomery Reduction */
+/*
+int mp_montgomery_reduce(mp_int *a, mp_int *m, mp_digit mp);
+*/
+
+/* returns 1 if a is a valid DR modulus */
+/*
+int mp_dr_is_modulus(mp_int *a);
+*/
+
+/* sets the value of "d" required for mp_dr_reduce */
+/*
+void mp_dr_setup(mp_int *a, mp_digit *d);
+*/
+
+/* reduces a modulo b using the Diminished Radix method */
+/*
+int mp_dr_reduce(mp_int *a, mp_int *b, mp_digit mp);
+*/
+
+/* returns true if a can be reduced with mp_reduce_2k */
+/*
+int mp_reduce_is_2k(mp_int *a);
+*/
+
+/* determines k value for 2k reduction */
+/*
+int mp_reduce_2k_setup(mp_int *a, mp_digit *d);
+*/
+
+/* reduces a modulo b where b is of the form 2**p - k [0 <= a] */
+/*
+int mp_reduce_2k(mp_int *a, mp_int *n, mp_digit d);
+*/
+
+/* returns true if a can be reduced with mp_reduce_2k_l */
+/*
+int mp_reduce_is_2k_l(mp_int *a);
+*/
+
+/* determines k value for 2k reduction */
+/*
+int mp_reduce_2k_setup_l(mp_int *a, mp_int *d);
+*/
+
+/* reduces a modulo b where b is of the form 2**p - k [0 <= a] */
+/*
+int mp_reduce_2k_l(mp_int *a, mp_int *n, mp_int *d);
+*/
+
+/* d = a**b (mod c) */
+/*
+int mp_exptmod(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
+*/
+
+/* ---> Primes <--- */
+
+/* number of primes */
+#ifdef MP_8BIT
+#  define PRIME_SIZE      31
+#else
+#  define PRIME_SIZE      256
+#endif
+
+/* table of first PRIME_SIZE primes */
+#if defined(BUILD_tcl) || !defined(_WIN32)
+MODULE_SCOPE const mp_digit ltm_prime_tab[PRIME_SIZE];
+#endif
+
+/* result=1 if a is divisible by one of the first PRIME_SIZE primes */
+/*
+int mp_prime_is_divisible(mp_int *a, int *result);
+*/
+
+/* performs one Fermat test of "a" using base "b".
+ * Sets result to 0 if composite or 1 if probable prime
+ */
+/*
+int mp_prime_fermat(mp_int *a, mp_int *b, int *result);
+*/
+
+/* performs one Miller-Rabin test of "a" using base "b".
+ * Sets result to 0 if composite or 1 if probable prime
+ */
+/*
+int mp_prime_miller_rabin(mp_int *a, mp_int *b, int *result);
+*/
+
+/* This gives [for a given bit size] the number of trials required
+ * such that Miller-Rabin gives a prob of failure lower than 2^-96
+ */
+/*
+int mp_prime_rabin_miller_trials(int size);
+*/
+
+/* performs t rounds of Miller-Rabin on "a" using the first
+ * t prime bases.  Also performs an initial sieve of trial
+ * division.  Determines if "a" is prime with probability
+ * of error no more than (1/4)**t.
+ *
+ * Sets result to 1 if probably prime, 0 otherwise
+ */
+/*
+int mp_prime_is_prime(mp_int *a, int t, int *result);
+*/
+
+/* finds the next prime after the number "a" using "t" trials
+ * of Miller-Rabin.
+ *
+ * bbs_style = 1 means the prime must be congruent to 3 mod 4
+ */
+/*
+int mp_prime_next_prime(mp_int *a, int t, int bbs_style);
+*/
+
+/* makes a truly random prime of a given size (bytes),
+ * call with bbs = 1 if you want it to be congruent to 3 mod 4
+ *
+ * You have to supply a callback which fills in a buffer with random bytes.  "dat" is a parameter you can
+ * have passed to the callback (e.g. a state or something).  This function doesn't use "dat" itself
+ * so it can be NULL
+ *
+ * The prime generated will be larger than 2^(8*size).
+ */
+#define mp_prime_random(a, t, size, bbs, cb, dat) mp_prime_random_ex(a, t, ((size) * 8) + 1, (bbs==1)?LTM_PRIME_BBS:0, cb, dat)
+
+/* makes a truly random prime of a given size (bits),
+ *
+ * Flags are as follows:
+ *
+ *   LTM_PRIME_BBS      - make prime congruent to 3 mod 4
+ *   LTM_PRIME_SAFE     - make sure (p-1)/2 is prime as well (implies LTM_PRIME_BBS)
+ *   LTM_PRIME_2MSB_ON  - make the 2nd highest bit one
+ *
+ * You have to supply a callback which fills in a buffer with random bytes.  "dat" is a parameter you can
+ * have passed to the callback (e.g. a state or something).  This function doesn't use "dat" itself
+ * so it can be NULL
+ *
+ */
+/*
+int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback cb, void *dat);
+*/
+
+/* ---> radix conversion <--- */
+/*
+int mp_count_bits(const mp_int *a);
+*/
+
+/*
+int mp_unsigned_bin_size(mp_int *a);
+*/
+/*
+int mp_read_unsigned_bin(mp_int *a, const unsigned char *b, int c);
+*/
+/*
+int mp_to_unsigned_bin(mp_int *a, unsigned char *b);
+*/
+/*
+int mp_to_unsigned_bin_n (mp_int * a, unsigned char *b, unsigned long *outlen);
+*/
+
+/*
+int mp_signed_bin_size(mp_int *a);
+*/
+/*
+int mp_read_signed_bin(mp_int *a, const unsigned char *b, int c);
+*/
+/*
+int mp_to_signed_bin(mp_int *a,  unsigned char *b);
+*/
+/*
+int mp_to_signed_bin_n (mp_int * a, unsigned char *b, unsigned long *outlen);
+*/
+
+/*
+int mp_read_radix(mp_int *a, const char *str, int radix);
+*/
+/*
+int mp_toradix(mp_int *a, char *str, int radix);
+*/
+/*
+int mp_toradix_n(mp_int * a, char *str, int radix, int maxlen);
+*/
+/*
+int mp_radix_size(const mp_int *a, int radix, int *size);
+*/
+
+#ifndef LTM_NO_FILE
+/*
+int mp_fread(mp_int *a, int radix, FILE *stream);
+*/
+/*
+int mp_fwrite(mp_int *a, int radix, FILE *stream);
+*/
+#endif
+
+#define mp_read_raw(mp, str, len) mp_read_signed_bin((mp), (str), (len))
+#define mp_raw_size(mp)           mp_signed_bin_size(mp)
+#define mp_toraw(mp, str)         mp_to_signed_bin((mp), (str))
+#define mp_read_mag(mp, str, len) mp_read_unsigned_bin((mp), (str), (len))
+#define mp_mag_size(mp)           mp_unsigned_bin_size(mp)
+#define mp_tomag(mp, str)         mp_to_unsigned_bin((mp), (str))
+
+#define mp_tobinary(M, S)  mp_toradix((M), (S), 2)
+#define mp_tooctal(M, S)   mp_toradix((M), (S), 8)
+#define mp_todecimal(M, S) mp_toradix((M), (S), 10)
+#define mp_tohex(M, S)     mp_toradix((M), (S), 16)
+
+#ifdef __cplusplus
+   }
+#endif
+
+#endif
+
+
+/* ref:         tag: v1.0.1, master */
+/* git commit:  5953f62e42b24af93748b1ee5e1d062e242c2546 */
+/* commit time: 2017-08-29 22:27:36 +0200 */
+
diff --git a/generic/tclTomMathDecls.h b/generic/tclTomMathDecls.h
new file mode 100644
index 0000000..209c486
--- /dev/null
+++ b/generic/tclTomMathDecls.h
@@ -0,0 +1,526 @@
+/*
+ *----------------------------------------------------------------------
+ *
+ * tclTomMathDecls.h --
+ *
+ *	This file contains the declarations for the 'libtommath'
+ *	functions that are exported by the Tcl library.
+ *
+ * Copyright (c) 2005 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution
+ * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#ifndef _TCLTOMMATHDECLS
+#define _TCLTOMMATHDECLS
+
+#include "tcl.h"
+
+/*
+ * Define the version of the Stubs table that's exported for tommath
+ */
+
+#define TCLTOMMATH_EPOCH 0
+#define TCLTOMMATH_REVISION 0
+
+#define Tcl_TomMath_InitStubs(interp,version) \
+    (TclTomMathInitializeStubs((interp),(version),\
+                               TCLTOMMATH_EPOCH,TCLTOMMATH_REVISION))
+
+/* Define custom memory allocation for libtommath */
+
+/* MODULE_SCOPE void* TclBNAlloc( size_t ); */
+#define TclBNAlloc(s) ((void*)ckalloc((size_t)(s)))
+/* MODULE_SCOPE void* TclBNRealloc( void*, size_t ); */
+#define TclBNRealloc(x,s) ((void*)ckrealloc((char*)(x),(size_t)(s)))
+/* MODULE_SCOPE void  TclBNFree( void* ); */
+#define TclBNFree(x) (ckfree((char*)(x)))
+/* MODULE_SCOPE void* TclBNCalloc( size_t, size_t ); */
+/* unused - no macro */
+
+#define XMALLOC(x) TclBNAlloc(x)
+#define XFREE(x) TclBNFree(x)
+#define XREALLOC(x,n) TclBNRealloc(x,n)
+#define XCALLOC(n,x) TclBNCalloc(n,x)
+
+/* Rename the global symbols in libtommath to avoid linkage conflicts */
+
+#define KARATSUBA_MUL_CUTOFF TclBNKaratsubaMulCutoff
+#define KARATSUBA_SQR_CUTOFF TclBNKaratsubaSqrCutoff
+#define TOOM_MUL_CUTOFF TclBNToomMulCutoff
+#define TOOM_SQR_CUTOFF TclBNToomSqrCutoff
+
+#define bn_reverse TclBN_reverse
+#define fast_s_mp_mul_digs TclBN_fast_s_mp_mul_digs
+#define fast_s_mp_sqr TclBN_fast_s_mp_sqr
+#define mp_add TclBN_mp_add
+#define mp_add_d TclBN_mp_add_d
+#define mp_and TclBN_mp_and
+#define mp_clamp TclBN_mp_clamp
+#define mp_clear TclBN_mp_clear
+#define mp_clear_multi TclBN_mp_clear_multi
+#define mp_cmp TclBN_mp_cmp
+#define mp_cmp_d TclBN_mp_cmp_d
+#define mp_cmp_mag TclBN_mp_cmp_mag
+#define mp_cnt_lsb TclBN_mp_cnt_lsb
+#define mp_copy TclBN_mp_copy
+#define mp_count_bits TclBN_mp_count_bits
+#define mp_div TclBN_mp_div
+#define mp_div_2 TclBN_mp_div_2
+#define mp_div_2d TclBN_mp_div_2d
+#define mp_div_3 TclBN_mp_div_3
+#define mp_div_d TclBN_mp_div_d
+#define mp_exch TclBN_mp_exch
+#define mp_expt_d TclBN_mp_expt_d
+#define mp_expt_d_ex TclBN_mp_expt_d_ex
+#define mp_grow TclBN_mp_grow
+#define mp_init TclBN_mp_init
+#define mp_init_copy TclBN_mp_init_copy
+#define mp_init_multi TclBN_mp_init_multi
+#define mp_init_set TclBN_mp_init_set
+#define mp_init_set_int TclBN_mp_init_set_int
+#define mp_init_size TclBN_mp_init_size
+#define mp_karatsuba_mul TclBN_mp_karatsuba_mul
+#define mp_karatsuba_sqr TclBN_mp_karatsuba_sqr
+#define mp_lshd TclBN_mp_lshd
+#define mp_mod TclBN_mp_mod
+#define mp_mod_2d TclBN_mp_mod_2d
+#define mp_mul TclBN_mp_mul
+#define mp_mul_2 TclBN_mp_mul_2
+#define mp_mul_2d TclBN_mp_mul_2d
+#define mp_mul_d TclBN_mp_mul_d
+#define mp_neg TclBN_mp_neg
+#define mp_or TclBN_mp_or
+#define mp_radix_size TclBN_mp_radix_size
+#define mp_read_radix TclBN_mp_read_radix
+#define mp_rshd TclBN_mp_rshd
+#define mp_s_rmap TclBNMpSRmap
+#define mp_set TclBN_mp_set
+#define mp_set_int TclBN_mp_set_int
+#define mp_shrink TclBN_mp_shrink
+#define mp_sqr TclBN_mp_sqr
+#define mp_sqrt TclBN_mp_sqrt
+#define mp_sub TclBN_mp_sub
+#define mp_sub_d TclBN_mp_sub_d
+#define mp_to_unsigned_bin TclBN_mp_to_unsigned_bin
+#define mp_to_unsigned_bin_n TclBN_mp_to_unsigned_bin_n
+#define mp_toom_mul TclBN_mp_toom_mul
+#define mp_toom_sqr TclBN_mp_toom_sqr
+#define mp_toradix_n TclBN_mp_toradix_n
+#define mp_unsigned_bin_size TclBN_mp_unsigned_bin_size
+#define mp_xor TclBN_mp_xor
+#define mp_zero TclBN_mp_zero
+#define s_mp_add TclBN_s_mp_add
+#define s_mp_mul_digs TclBN_s_mp_mul_digs
+#define s_mp_sqr TclBN_s_mp_sqr
+#define s_mp_sub TclBN_s_mp_sub
+
+#undef TCL_STORAGE_CLASS
+#ifdef BUILD_tcl
+#   define TCL_STORAGE_CLASS DLLEXPORT
+#else
+#   ifdef USE_TCL_STUBS
+#      define TCL_STORAGE_CLASS
+#   else
+#      define TCL_STORAGE_CLASS DLLIMPORT
+#   endif
+#endif
+
+/*
+ * WARNING: This file is automatically generated by the tools/genStubs.tcl
+ * script.  Any modifications to the function declarations below should be made
+ * in the generic/tclInt.decls script.
+ */
+
+/* !BEGIN!: Do not edit below this line. */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+ * Exported function declarations:
+ */
+
+/* 0 */
+EXTERN int		TclBN_epoch(void);
+/* 1 */
+EXTERN int		TclBN_revision(void);
+/* 2 */
+EXTERN int		TclBN_mp_add(mp_int *a, mp_int *b, mp_int *c);
+/* 3 */
+EXTERN int		TclBN_mp_add_d(mp_int *a, mp_digit b, mp_int *c);
+/* 4 */
+EXTERN int		TclBN_mp_and(mp_int *a, mp_int *b, mp_int *c);
+/* 5 */
+EXTERN void		TclBN_mp_clamp(mp_int *a);
+/* 6 */
+EXTERN void		TclBN_mp_clear(mp_int *a);
+/* 7 */
+EXTERN void		TclBN_mp_clear_multi(mp_int *a, ...);
+/* 8 */
+EXTERN int		TclBN_mp_cmp(const mp_int *a, const mp_int *b);
+/* 9 */
+EXTERN int		TclBN_mp_cmp_d(const mp_int *a, mp_digit b);
+/* 10 */
+EXTERN int		TclBN_mp_cmp_mag(const mp_int *a, const mp_int *b);
+/* 11 */
+EXTERN int		TclBN_mp_copy(const mp_int *a, mp_int *b);
+/* 12 */
+EXTERN int		TclBN_mp_count_bits(const mp_int *a);
+/* 13 */
+EXTERN int		TclBN_mp_div(mp_int *a, mp_int *b, mp_int *q,
+				mp_int *r);
+/* 14 */
+EXTERN int		TclBN_mp_div_d(mp_int *a, mp_digit b, mp_int *q,
+				mp_digit *r);
+/* 15 */
+EXTERN int		TclBN_mp_div_2(mp_int *a, mp_int *q);
+/* 16 */
+EXTERN int		TclBN_mp_div_2d(const mp_int *a, int b, mp_int *q,
+				mp_int *r);
+/* 17 */
+EXTERN int		TclBN_mp_div_3(mp_int *a, mp_int *q, mp_digit *r);
+/* 18 */
+EXTERN void		TclBN_mp_exch(mp_int *a, mp_int *b);
+/* 19 */
+EXTERN int		TclBN_mp_expt_d(mp_int *a, mp_digit b, mp_int *c);
+/* 20 */
+EXTERN int		TclBN_mp_grow(mp_int *a, int size);
+/* 21 */
+EXTERN int		TclBN_mp_init(mp_int *a);
+/* 22 */
+EXTERN int		TclBN_mp_init_copy(mp_int *a, const mp_int *b);
+/* 23 */
+EXTERN int		TclBN_mp_init_multi(mp_int *a, ...);
+/* 24 */
+EXTERN int		TclBN_mp_init_set(mp_int *a, mp_digit b);
+/* 25 */
+EXTERN int		TclBN_mp_init_size(mp_int *a, int size);
+/* 26 */
+EXTERN int		TclBN_mp_lshd(mp_int *a, int shift);
+/* 27 */
+EXTERN int		TclBN_mp_mod(mp_int *a, mp_int *b, mp_int *r);
+/* 28 */
+EXTERN int		TclBN_mp_mod_2d(const mp_int *a, int b, mp_int *r);
+/* 29 */
+EXTERN int		TclBN_mp_mul(mp_int *a, mp_int *b, mp_int *p);
+/* 30 */
+EXTERN int		TclBN_mp_mul_d(mp_int *a, mp_digit b, mp_int *p);
+/* 31 */
+EXTERN int		TclBN_mp_mul_2(mp_int *a, mp_int *p);
+/* 32 */
+EXTERN int		TclBN_mp_mul_2d(const mp_int *a, int d, mp_int *p);
+/* 33 */
+EXTERN int		TclBN_mp_neg(const mp_int *a, mp_int *b);
+/* 34 */
+EXTERN int		TclBN_mp_or(mp_int *a, mp_int *b, mp_int *c);
+/* 35 */
+EXTERN int		TclBN_mp_radix_size(const mp_int *a, int radix,
+				int *size);
+/* 36 */
+EXTERN int		TclBN_mp_read_radix(mp_int *a, const char *str,
+				int radix);
+/* 37 */
+EXTERN void		TclBN_mp_rshd(mp_int *a, int shift);
+/* 38 */
+EXTERN int		TclBN_mp_shrink(mp_int *a);
+/* 39 */
+EXTERN void		TclBN_mp_set(mp_int *a, mp_digit b);
+/* 40 */
+EXTERN int		TclBN_mp_sqr(mp_int *a, mp_int *b);
+/* 41 */
+EXTERN int		TclBN_mp_sqrt(mp_int *a, mp_int *b);
+/* 42 */
+EXTERN int		TclBN_mp_sub(mp_int *a, mp_int *b, mp_int *c);
+/* 43 */
+EXTERN int		TclBN_mp_sub_d(mp_int *a, mp_digit b, mp_int *c);
+/* 44 */
+EXTERN int		TclBN_mp_to_unsigned_bin(mp_int *a, unsigned char *b);
+/* 45 */
+EXTERN int		TclBN_mp_to_unsigned_bin_n(mp_int *a,
+				unsigned char *b, unsigned long *outlen);
+/* 46 */
+EXTERN int		TclBN_mp_toradix_n(mp_int *a, char *str, int radix,
+				int maxlen);
+/* 47 */
+EXTERN int		TclBN_mp_unsigned_bin_size(mp_int *a);
+/* 48 */
+EXTERN int		TclBN_mp_xor(mp_int *a, mp_int *b, mp_int *c);
+/* 49 */
+EXTERN void		TclBN_mp_zero(mp_int *a);
+/* 50 */
+EXTERN void		TclBN_reverse(unsigned char *s, int len);
+/* 51 */
+EXTERN int		TclBN_fast_s_mp_mul_digs(mp_int *a, mp_int *b,
+				mp_int *c, int digs);
+/* 52 */
+EXTERN int		TclBN_fast_s_mp_sqr(mp_int *a, mp_int *b);
+/* 53 */
+EXTERN int		TclBN_mp_karatsuba_mul(mp_int *a, mp_int *b,
+				mp_int *c);
+/* 54 */
+EXTERN int		TclBN_mp_karatsuba_sqr(mp_int *a, mp_int *b);
+/* 55 */
+EXTERN int		TclBN_mp_toom_mul(mp_int *a, mp_int *b, mp_int *c);
+/* 56 */
+EXTERN int		TclBN_mp_toom_sqr(mp_int *a, mp_int *b);
+/* 57 */
+EXTERN int		TclBN_s_mp_add(mp_int *a, mp_int *b, mp_int *c);
+/* 58 */
+EXTERN int		TclBN_s_mp_mul_digs(mp_int *a, mp_int *b, mp_int *c,
+				int digs);
+/* 59 */
+EXTERN int		TclBN_s_mp_sqr(mp_int *a, mp_int *b);
+/* 60 */
+EXTERN int		TclBN_s_mp_sub(mp_int *a, mp_int *b, mp_int *c);
+/* 61 */
+EXTERN int		TclBN_mp_init_set_int(mp_int *a, unsigned long i);
+/* 62 */
+EXTERN int		TclBN_mp_set_int(mp_int *a, unsigned long i);
+/* 63 */
+EXTERN int		TclBN_mp_cnt_lsb(const mp_int *a);
+/* 64 */
+EXTERN void		TclBNInitBignumFromLong(mp_int *bignum, long initVal);
+/* 65 */
+EXTERN void		TclBNInitBignumFromWideInt(mp_int *bignum,
+				Tcl_WideInt initVal);
+/* 66 */
+EXTERN void		TclBNInitBignumFromWideUInt(mp_int *bignum,
+				Tcl_WideUInt initVal);
+/* 67 */
+EXTERN int		TclBN_mp_expt_d_ex(mp_int *a, mp_digit b, mp_int *c,
+				int fast);
+
+typedef struct TclTomMathStubs {
+    int magic;
+    void *hooks;
+
+    int (*tclBN_epoch) (void); /* 0 */
+    int (*tclBN_revision) (void); /* 1 */
+    int (*tclBN_mp_add) (mp_int *a, mp_int *b, mp_int *c); /* 2 */
+    int (*tclBN_mp_add_d) (mp_int *a, mp_digit b, mp_int *c); /* 3 */
+    int (*tclBN_mp_and) (mp_int *a, mp_int *b, mp_int *c); /* 4 */
+    void (*tclBN_mp_clamp) (mp_int *a); /* 5 */
+    void (*tclBN_mp_clear) (mp_int *a); /* 6 */
+    void (*tclBN_mp_clear_multi) (mp_int *a, ...); /* 7 */
+    int (*tclBN_mp_cmp) (const mp_int *a, const mp_int *b); /* 8 */
+    int (*tclBN_mp_cmp_d) (const mp_int *a, mp_digit b); /* 9 */
+    int (*tclBN_mp_cmp_mag) (const mp_int *a, const mp_int *b); /* 10 */
+    int (*tclBN_mp_copy) (const mp_int *a, mp_int *b); /* 11 */
+    int (*tclBN_mp_count_bits) (const mp_int *a); /* 12 */
+    int (*tclBN_mp_div) (mp_int *a, mp_int *b, mp_int *q, mp_int *r); /* 13 */
+    int (*tclBN_mp_div_d) (mp_int *a, mp_digit b, mp_int *q, mp_digit *r); /* 14 */
+    int (*tclBN_mp_div_2) (mp_int *a, mp_int *q); /* 15 */
+    int (*tclBN_mp_div_2d) (const mp_int *a, int b, mp_int *q, mp_int *r); /* 16 */
+    int (*tclBN_mp_div_3) (mp_int *a, mp_int *q, mp_digit *r); /* 17 */
+    void (*tclBN_mp_exch) (mp_int *a, mp_int *b); /* 18 */
+    int (*tclBN_mp_expt_d) (mp_int *a, mp_digit b, mp_int *c); /* 19 */
+    int (*tclBN_mp_grow) (mp_int *a, int size); /* 20 */
+    int (*tclBN_mp_init) (mp_int *a); /* 21 */
+    int (*tclBN_mp_init_copy) (mp_int *a, const mp_int *b); /* 22 */
+    int (*tclBN_mp_init_multi) (mp_int *a, ...); /* 23 */
+    int (*tclBN_mp_init_set) (mp_int *a, mp_digit b); /* 24 */
+    int (*tclBN_mp_init_size) (mp_int *a, int size); /* 25 */
+    int (*tclBN_mp_lshd) (mp_int *a, int shift); /* 26 */
+    int (*tclBN_mp_mod) (mp_int *a, mp_int *b, mp_int *r); /* 27 */
+    int (*tclBN_mp_mod_2d) (const mp_int *a, int b, mp_int *r); /* 28 */
+    int (*tclBN_mp_mul) (mp_int *a, mp_int *b, mp_int *p); /* 29 */
+    int (*tclBN_mp_mul_d) (mp_int *a, mp_digit b, mp_int *p); /* 30 */
+    int (*tclBN_mp_mul_2) (mp_int *a, mp_int *p); /* 31 */
+    int (*tclBN_mp_mul_2d) (const mp_int *a, int d, mp_int *p); /* 32 */
+    int (*tclBN_mp_neg) (const mp_int *a, mp_int *b); /* 33 */
+    int (*tclBN_mp_or) (mp_int *a, mp_int *b, mp_int *c); /* 34 */
+    int (*tclBN_mp_radix_size) (const mp_int *a, int radix, int *size); /* 35 */
+    int (*tclBN_mp_read_radix) (mp_int *a, const char *str, int radix); /* 36 */
+    void (*tclBN_mp_rshd) (mp_int *a, int shift); /* 37 */
+    int (*tclBN_mp_shrink) (mp_int *a); /* 38 */
+    void (*tclBN_mp_set) (mp_int *a, mp_digit b); /* 39 */
+    int (*tclBN_mp_sqr) (mp_int *a, mp_int *b); /* 40 */
+    int (*tclBN_mp_sqrt) (mp_int *a, mp_int *b); /* 41 */
+    int (*tclBN_mp_sub) (mp_int *a, mp_int *b, mp_int *c); /* 42 */
+    int (*tclBN_mp_sub_d) (mp_int *a, mp_digit b, mp_int *c); /* 43 */
+    int (*tclBN_mp_to_unsigned_bin) (mp_int *a, unsigned char *b); /* 44 */
+    int (*tclBN_mp_to_unsigned_bin_n) (mp_int *a, unsigned char *b, unsigned long *outlen); /* 45 */
+    int (*tclBN_mp_toradix_n) (mp_int *a, char *str, int radix, int maxlen); /* 46 */
+    int (*tclBN_mp_unsigned_bin_size) (mp_int *a); /* 47 */
+    int (*tclBN_mp_xor) (mp_int *a, mp_int *b, mp_int *c); /* 48 */
+    void (*tclBN_mp_zero) (mp_int *a); /* 49 */
+    void (*tclBN_reverse) (unsigned char *s, int len); /* 50 */
+    int (*tclBN_fast_s_mp_mul_digs) (mp_int *a, mp_int *b, mp_int *c, int digs); /* 51 */
+    int (*tclBN_fast_s_mp_sqr) (mp_int *a, mp_int *b); /* 52 */
+    int (*tclBN_mp_karatsuba_mul) (mp_int *a, mp_int *b, mp_int *c); /* 53 */
+    int (*tclBN_mp_karatsuba_sqr) (mp_int *a, mp_int *b); /* 54 */
+    int (*tclBN_mp_toom_mul) (mp_int *a, mp_int *b, mp_int *c); /* 55 */
+    int (*tclBN_mp_toom_sqr) (mp_int *a, mp_int *b); /* 56 */
+    int (*tclBN_s_mp_add) (mp_int *a, mp_int *b, mp_int *c); /* 57 */
+    int (*tclBN_s_mp_mul_digs) (mp_int *a, mp_int *b, mp_int *c, int digs); /* 58 */
+    int (*tclBN_s_mp_sqr) (mp_int *a, mp_int *b); /* 59 */
+    int (*tclBN_s_mp_sub) (mp_int *a, mp_int *b, mp_int *c); /* 60 */
+    int (*tclBN_mp_init_set_int) (mp_int *a, unsigned long i); /* 61 */
+    int (*tclBN_mp_set_int) (mp_int *a, unsigned long i); /* 62 */
+    int (*tclBN_mp_cnt_lsb) (const mp_int *a); /* 63 */
+    void (*tclBNInitBignumFromLong) (mp_int *bignum, long initVal); /* 64 */
+    void (*tclBNInitBignumFromWideInt) (mp_int *bignum, Tcl_WideInt initVal); /* 65 */
+    void (*tclBNInitBignumFromWideUInt) (mp_int *bignum, Tcl_WideUInt initVal); /* 66 */
+    int (*tclBN_mp_expt_d_ex) (mp_int *a, mp_digit b, mp_int *c, int fast); /* 67 */
+} TclTomMathStubs;
+
+extern const TclTomMathStubs *tclTomMathStubsPtr;
+
+#ifdef __cplusplus
+}
+#endif
+
+#if defined(USE_TCL_STUBS)
+
+/*
+ * Inline function declarations:
+ */
+
+#define TclBN_epoch \
+	(tclTomMathStubsPtr->tclBN_epoch) /* 0 */
+#define TclBN_revision \
+	(tclTomMathStubsPtr->tclBN_revision) /* 1 */
+#define TclBN_mp_add \
+	(tclTomMathStubsPtr->tclBN_mp_add) /* 2 */
+#define TclBN_mp_add_d \
+	(tclTomMathStubsPtr->tclBN_mp_add_d) /* 3 */
+#define TclBN_mp_and \
+	(tclTomMathStubsPtr->tclBN_mp_and) /* 4 */
+#define TclBN_mp_clamp \
+	(tclTomMathStubsPtr->tclBN_mp_clamp) /* 5 */
+#define TclBN_mp_clear \
+	(tclTomMathStubsPtr->tclBN_mp_clear) /* 6 */
+#define TclBN_mp_clear_multi \
+	(tclTomMathStubsPtr->tclBN_mp_clear_multi) /* 7 */
+#define TclBN_mp_cmp \
+	(tclTomMathStubsPtr->tclBN_mp_cmp) /* 8 */
+#define TclBN_mp_cmp_d \
+	(tclTomMathStubsPtr->tclBN_mp_cmp_d) /* 9 */
+#define TclBN_mp_cmp_mag \
+	(tclTomMathStubsPtr->tclBN_mp_cmp_mag) /* 10 */
+#define TclBN_mp_copy \
+	(tclTomMathStubsPtr->tclBN_mp_copy) /* 11 */
+#define TclBN_mp_count_bits \
+	(tclTomMathStubsPtr->tclBN_mp_count_bits) /* 12 */
+#define TclBN_mp_div \
+	(tclTomMathStubsPtr->tclBN_mp_div) /* 13 */
+#define TclBN_mp_div_d \
+	(tclTomMathStubsPtr->tclBN_mp_div_d) /* 14 */
+#define TclBN_mp_div_2 \
+	(tclTomMathStubsPtr->tclBN_mp_div_2) /* 15 */
+#define TclBN_mp_div_2d \
+	(tclTomMathStubsPtr->tclBN_mp_div_2d) /* 16 */
+#define TclBN_mp_div_3 \
+	(tclTomMathStubsPtr->tclBN_mp_div_3) /* 17 */
+#define TclBN_mp_exch \
+	(tclTomMathStubsPtr->tclBN_mp_exch) /* 18 */
+#define TclBN_mp_expt_d \
+	(tclTomMathStubsPtr->tclBN_mp_expt_d) /* 19 */
+#define TclBN_mp_grow \
+	(tclTomMathStubsPtr->tclBN_mp_grow) /* 20 */
+#define TclBN_mp_init \
+	(tclTomMathStubsPtr->tclBN_mp_init) /* 21 */
+#define TclBN_mp_init_copy \
+	(tclTomMathStubsPtr->tclBN_mp_init_copy) /* 22 */
+#define TclBN_mp_init_multi \
+	(tclTomMathStubsPtr->tclBN_mp_init_multi) /* 23 */
+#define TclBN_mp_init_set \
+	(tclTomMathStubsPtr->tclBN_mp_init_set) /* 24 */
+#define TclBN_mp_init_size \
+	(tclTomMathStubsPtr->tclBN_mp_init_size) /* 25 */
+#define TclBN_mp_lshd \
+	(tclTomMathStubsPtr->tclBN_mp_lshd) /* 26 */
+#define TclBN_mp_mod \
+	(tclTomMathStubsPtr->tclBN_mp_mod) /* 27 */
+#define TclBN_mp_mod_2d \
+	(tclTomMathStubsPtr->tclBN_mp_mod_2d) /* 28 */
+#define TclBN_mp_mul \
+	(tclTomMathStubsPtr->tclBN_mp_mul) /* 29 */
+#define TclBN_mp_mul_d \
+	(tclTomMathStubsPtr->tclBN_mp_mul_d) /* 30 */
+#define TclBN_mp_mul_2 \
+	(tclTomMathStubsPtr->tclBN_mp_mul_2) /* 31 */
+#define TclBN_mp_mul_2d \
+	(tclTomMathStubsPtr->tclBN_mp_mul_2d) /* 32 */
+#define TclBN_mp_neg \
+	(tclTomMathStubsPtr->tclBN_mp_neg) /* 33 */
+#define TclBN_mp_or \
+	(tclTomMathStubsPtr->tclBN_mp_or) /* 34 */
+#define TclBN_mp_radix_size \
+	(tclTomMathStubsPtr->tclBN_mp_radix_size) /* 35 */
+#define TclBN_mp_read_radix \
+	(tclTomMathStubsPtr->tclBN_mp_read_radix) /* 36 */
+#define TclBN_mp_rshd \
+	(tclTomMathStubsPtr->tclBN_mp_rshd) /* 37 */
+#define TclBN_mp_shrink \
+	(tclTomMathStubsPtr->tclBN_mp_shrink) /* 38 */
+#define TclBN_mp_set \
+	(tclTomMathStubsPtr->tclBN_mp_set) /* 39 */
+#define TclBN_mp_sqr \
+	(tclTomMathStubsPtr->tclBN_mp_sqr) /* 40 */
+#define TclBN_mp_sqrt \
+	(tclTomMathStubsPtr->tclBN_mp_sqrt) /* 41 */
+#define TclBN_mp_sub \
+	(tclTomMathStubsPtr->tclBN_mp_sub) /* 42 */
+#define TclBN_mp_sub_d \
+	(tclTomMathStubsPtr->tclBN_mp_sub_d) /* 43 */
+#define TclBN_mp_to_unsigned_bin \
+	(tclTomMathStubsPtr->tclBN_mp_to_unsigned_bin) /* 44 */
+#define TclBN_mp_to_unsigned_bin_n \
+	(tclTomMathStubsPtr->tclBN_mp_to_unsigned_bin_n) /* 45 */
+#define TclBN_mp_toradix_n \
+	(tclTomMathStubsPtr->tclBN_mp_toradix_n) /* 46 */
+#define TclBN_mp_unsigned_bin_size \
+	(tclTomMathStubsPtr->tclBN_mp_unsigned_bin_size) /* 47 */
+#define TclBN_mp_xor \
+	(tclTomMathStubsPtr->tclBN_mp_xor) /* 48 */
+#define TclBN_mp_zero \
+	(tclTomMathStubsPtr->tclBN_mp_zero) /* 49 */
+#define TclBN_reverse \
+	(tclTomMathStubsPtr->tclBN_reverse) /* 50 */
+#define TclBN_fast_s_mp_mul_digs \
+	(tclTomMathStubsPtr->tclBN_fast_s_mp_mul_digs) /* 51 */
+#define TclBN_fast_s_mp_sqr \
+	(tclTomMathStubsPtr->tclBN_fast_s_mp_sqr) /* 52 */
+#define TclBN_mp_karatsuba_mul \
+	(tclTomMathStubsPtr->tclBN_mp_karatsuba_mul) /* 53 */
+#define TclBN_mp_karatsuba_sqr \
+	(tclTomMathStubsPtr->tclBN_mp_karatsuba_sqr) /* 54 */
+#define TclBN_mp_toom_mul \
+	(tclTomMathStubsPtr->tclBN_mp_toom_mul) /* 55 */
+#define TclBN_mp_toom_sqr \
+	(tclTomMathStubsPtr->tclBN_mp_toom_sqr) /* 56 */
+#define TclBN_s_mp_add \
+	(tclTomMathStubsPtr->tclBN_s_mp_add) /* 57 */
+#define TclBN_s_mp_mul_digs \
+	(tclTomMathStubsPtr->tclBN_s_mp_mul_digs) /* 58 */
+#define TclBN_s_mp_sqr \
+	(tclTomMathStubsPtr->tclBN_s_mp_sqr) /* 59 */
+#define TclBN_s_mp_sub \
+	(tclTomMathStubsPtr->tclBN_s_mp_sub) /* 60 */
+#define TclBN_mp_init_set_int \
+	(tclTomMathStubsPtr->tclBN_mp_init_set_int) /* 61 */
+#define TclBN_mp_set_int \
+	(tclTomMathStubsPtr->tclBN_mp_set_int) /* 62 */
+#define TclBN_mp_cnt_lsb \
+	(tclTomMathStubsPtr->tclBN_mp_cnt_lsb) /* 63 */
+#define TclBNInitBignumFromLong \
+	(tclTomMathStubsPtr->tclBNInitBignumFromLong) /* 64 */
+#define TclBNInitBignumFromWideInt \
+	(tclTomMathStubsPtr->tclBNInitBignumFromWideInt) /* 65 */
+#define TclBNInitBignumFromWideUInt \
+	(tclTomMathStubsPtr->tclBNInitBignumFromWideUInt) /* 66 */
+#define TclBN_mp_expt_d_ex \
+	(tclTomMathStubsPtr->tclBN_mp_expt_d_ex) /* 67 */
+
+#endif /* defined(USE_TCL_STUBS) */
+
+/* !END!: Do not edit above this line. */
+
+#undef TCL_STORAGE_CLASS
+#define TCL_STORAGE_CLASS DLLIMPORT
+
+#endif /* _TCLINTDECLS */
diff --git a/generic/tclTomMathInt.h b/generic/tclTomMathInt.h
new file mode 100644
index 0000000..831f13f
--- /dev/null
+++ b/generic/tclTomMathInt.h
@@ -0,0 +1,3 @@
+#include "tclInt.h"
+#include "tclTomMath.h"
+#include "tommath_class.h"
diff --git a/generic/tclTomMathInterface.c b/generic/tclTomMathInterface.c
new file mode 100644
index 0000000..48db8c3
--- /dev/null
+++ b/generic/tclTomMathInterface.c
@@ -0,0 +1,310 @@
+/*
+ *----------------------------------------------------------------------
+ *
+ * tclTomMathInterface.c --
+ *
+ *	This file contains procedures that are used as a 'glue' layer between
+ *	Tcl and libtommath.
+ *
+ * Copyright (c) 2005 by Kevin B. Kenny.  All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tommath.h"
+
+MODULE_SCOPE const TclTomMathStubs tclTomMathStubs;
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTommath_Init --
+ *
+ *	Initializes the TclTomMath 'package', which exists as a
+ *	placeholder so that the package data can be used to hold
+ *	a stub table pointer.
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * Side effects:
+ *	Installs the stub table for tommath.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclTommath_Init(
+    Tcl_Interp *interp)		/* Tcl interpreter */
+{
+    /* TIP #268: Full patchlevel instead of just major.minor */
+
+    if (Tcl_PkgProvideEx(interp, "tcl::tommath", TCL_PATCH_LEVEL,
+	    &tclTomMathStubs) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBN_epoch --
+ *
+ *	Return the epoch number of the TclTomMath stubs table
+ *
+ * Results:
+ *	Returns an arbitrary integer that does not decrease with
+ *	release.  Stubs tables with different epochs are incompatible.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclBN_epoch(void)
+{
+    return TCLTOMMATH_EPOCH;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBN_revision --
+ *
+ *	Returns the revision level of the TclTomMath stubs table
+ *
+ * Results:
+ *	Returns an arbitrary integer that increases with revisions.
+ *	If a client requires a given epoch and revision, any Stubs table
+ *	with the same epoch and an equal or higher revision satisfies
+ *	the request.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclBN_revision(void)
+{
+    return TCLTOMMATH_REVISION;
+}
+#if 0
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNAlloc --
+ *
+ *	Allocate memory for libtommath.
+ *
+ * Results:
+ *	Returns a pointer to the allocated block.
+ *
+ * This procedure is a wrapper around Tcl_Alloc, needed because of a
+ * mismatched type signature between Tcl_Alloc and malloc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+extern void *
+TclBNAlloc(
+    size_t x)
+{
+    return (void *) ckalloc((unsigned int) x);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNRealloc --
+ *
+ *	Change the size of an allocated block of memory in libtommath
+ *
+ * Results:
+ *	Returns a pointer to the allocated block.
+ *
+ * This procedure is a wrapper around Tcl_Realloc, needed because of a
+ * mismatched type signature between Tcl_Realloc and realloc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void *
+TclBNRealloc(
+    void *p,
+    size_t s)
+{
+    return (void *) ckrealloc((char *) p, (unsigned int) s);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNFree --
+ *
+ *	Free allocated memory in libtommath.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is freed.
+ *
+ * This function is simply a wrapper around Tcl_Free, needed in libtommath
+ * because of a type mismatch between free and Tcl_Free.
+ *
+ *----------------------------------------------------------------------
+ */
+
+extern void
+TclBNFree(
+    void *p)
+{
+    ckree((char *) p);
+}
+#endif
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNInitBignumFromLong --
+ *
+ *	Allocate and initialize a 'bignum' from a native 'long'.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The 'bignum' is constructed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+extern void
+TclBNInitBignumFromLong(
+    mp_int *a,
+    long initVal)
+{
+    int status;
+    unsigned long v;
+    mp_digit *p;
+
+    /*
+     * Allocate enough memory to hold the largest possible long
+     */
+
+    status = mp_init_size(a,
+	    (CHAR_BIT * sizeof(long) + DIGIT_BIT - 1) / DIGIT_BIT);
+    if (status != MP_OKAY) {
+	Tcl_Panic("initialization failure in TclBNInitBignumFromLong");
+    }
+
+    /*
+     * Convert arg to sign and magnitude.
+     */
+
+    if (initVal < 0) {
+	a->sign = MP_NEG;
+	v = -initVal;
+    } else {
+	a->sign = MP_ZPOS;
+	v = initVal;
+    }
+
+    /*
+     * Store the magnitude in the bignum.
+     */
+
+    p = a->dp;
+    while (v) {
+	*p++ = (mp_digit) (v & MP_MASK);
+	v >>= MP_DIGIT_BIT;
+    }
+    a->used = p - a->dp;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNInitBignumFromWideInt --
+ *
+ *	Allocate and initialize a 'bignum' from a Tcl_WideInt
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The 'bignum' is constructed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+extern void
+TclBNInitBignumFromWideInt(
+    mp_int *a,			/* Bignum to initialize */
+    Tcl_WideInt v)		/* Initial value */
+{
+    if (v < (Tcl_WideInt)0) {
+	TclBNInitBignumFromWideUInt(a, (Tcl_WideUInt)(-v));
+	mp_neg(a, a);
+    } else {
+	TclBNInitBignumFromWideUInt(a, (Tcl_WideUInt)v);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclBNInitBignumFromWideUInt --
+ *
+ *	Allocate and initialize a 'bignum' from a Tcl_WideUInt
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The 'bignum' is constructed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+extern void
+TclBNInitBignumFromWideUInt(
+    mp_int *a,			/* Bignum to initialize */
+    Tcl_WideUInt v)		/* Initial value */
+{
+    int status;
+    mp_digit *p;
+
+    /*
+     * Allocate enough memory to hold the largest possible Tcl_WideUInt.
+     */
+
+    status = mp_init_size(a,
+	    (CHAR_BIT * sizeof(Tcl_WideUInt) + DIGIT_BIT - 1) / DIGIT_BIT);
+    if (status != MP_OKAY) {
+	Tcl_Panic("initialization failure in TclBNInitBignumFromWideUInt");
+    }
+
+    a->sign = MP_ZPOS;
+
+    /*
+     * Store the magnitude in the bignum.
+     */
+
+    p = a->dp;
+    while (v) {
+	*p++ = (mp_digit) (v & MP_MASK);
+	v >>= MP_DIGIT_BIT;
+    }
+    a->used = p - a->dp;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclTomMathStubLib.c b/generic/tclTomMathStubLib.c
new file mode 100644
index 0000000..324f2a3
--- /dev/null
+++ b/generic/tclTomMathStubLib.c
@@ -0,0 +1,79 @@
+/*
+ * tclTomMathStubLib.c --
+ *
+ *	Stub object that will be statically linked into extensions that want
+ *	to access Tcl.
+ *
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 1998 Paul Duffin.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+MODULE_SCOPE const TclTomMathStubs *tclTomMathStubsPtr;
+
+const TclTomMathStubs *tclTomMathStubsPtr = NULL;
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTomMathInitStubs --
+ *
+ *	Initializes the Stubs table for Tcl's subset of libtommath
+ *
+ * Results:
+ *	Returns a standard Tcl result.
+ *
+ * This procedure should not be called directly, but rather through
+ * the TclTomMath_InitStubs macro, to insure that the Stubs table
+ * matches the header files used in compilation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+MODULE_SCOPE const char *
+TclTomMathInitializeStubs(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    const char *version,	/* Tcl version needed */
+    int epoch,			/* Stubs table epoch from the header files */
+    int revision)		/* Stubs table revision number from the
+				 * header files */
+{
+    int exact = 0;
+    const char *packageName = "tcl::tommath";
+    const char *errMsg = NULL;
+    TclTomMathStubs *stubsPtr = NULL;
+    const char *actualVersion = tclStubsPtr->tcl_PkgRequireEx(interp,
+	    packageName, version, exact, &stubsPtr);
+
+    if (actualVersion == NULL) {
+	return NULL;
+    }
+    if (stubsPtr == NULL) {
+	errMsg = "missing stub table pointer";
+    } else if(stubsPtr->tclBN_epoch() != epoch) {
+	errMsg = "epoch number mismatch";
+    } else if(stubsPtr->tclBN_revision() != revision) {
+	errMsg = "requires a later revision";
+    } else {
+	tclTomMathStubsPtr = stubsPtr;
+	return actualVersion;
+    }
+    tclStubsPtr->tcl_ResetResult(interp);
+    tclStubsPtr->tcl_AppendResult(interp, "Error loading ", packageName,
+	    " (requested version ", version, ", actual version ",
+	    actualVersion, "): ", errMsg, NULL);
+    return NULL;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclTrace.c b/generic/tclTrace.c
new file mode 100644
index 0000000..f86f472
--- /dev/null
+++ b/generic/tclTrace.c
@@ -0,0 +1,3277 @@
+/*
+ * tclTrace.c --
+ *
+ *	This file contains code to handle most trace management.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-2000 Scriptics Corporation.
+ * Copyright (c) 2002 ActiveState Corporation.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Structures used to hold information about variable traces:
+ */
+
+typedef struct {
+    int flags;			/* Operations for which Tcl command is to be
+				 * invoked. */
+    size_t length;		/* Number of non-NUL chars. in command. */
+    char command[1];		/* Space for Tcl command to invoke. Actual
+				 * size will be as large as necessary to hold
+				 * command. This field must be the last in the
+				 * structure, so that it can be larger than 1
+				 * byte. */
+} TraceVarInfo;
+
+typedef struct {
+    VarTrace traceInfo;
+    TraceVarInfo traceCmdInfo;
+} CombinedTraceVarInfo;
+
+/*
+ * Structure used to hold information about command traces:
+ */
+
+typedef struct {
+    int flags;			/* Operations for which Tcl command is to be
+				 * invoked. */
+    size_t length;		/* Number of non-NUL chars. in command. */
+    Tcl_Trace stepTrace;	/* Used for execution traces, when tracing
+				 * inside the given command */
+    int startLevel;		/* Used for bookkeeping with step execution
+				 * traces, store the level at which the step
+				 * trace was invoked */
+    char *startCmd;		/* Used for bookkeeping with step execution
+				 * traces, store the command name which
+				 * invoked step trace */
+    int curFlags;		/* Trace flags for the current command */
+    int curCode;		/* Return code for the current command */
+    size_t refCount;		/* Used to ensure this structure is not
+				 * deleted too early. Keeps track of how many
+				 * pieces of code have a pointer to this
+				 * structure. */
+    char command[1];		/* Space for Tcl command to invoke. Actual
+				 * size will be as large as necessary to hold
+				 * command. This field must be the last in the
+				 * structure, so that it can be larger than 1
+				 * byte. */
+} TraceCommandInfo;
+
+/*
+ * Used by command execution traces. Note that we assume in the code that
+ * TCL_TRACE_ENTER_DURING_EXEC == 4 * TCL_TRACE_ENTER_EXEC and that
+ * TCL_TRACE_LEAVE_DURING_EXEC == 4 * TCL_TRACE_LEAVE_EXEC.
+ *
+ * TCL_TRACE_ENTER_DURING_EXEC  - Trace each command inside the command
+ *				  currently being traced, before execution.
+ * TCL_TRACE_LEAVE_DURING_EXEC  - Trace each command inside the command
+ *				  currently being traced, after execution.
+ * TCL_TRACE_ANY_EXEC		- OR'd combination of all EXEC flags.
+ * TCL_TRACE_EXEC_IN_PROGRESS   - The callback function on this trace is
+ *				  currently executing. Therefore we don't let
+ *				  further traces execute.
+ * TCL_TRACE_EXEC_DIRECT	- This execution trace is triggered directly
+ *				  by the command being traced, not because of
+ *				  an internal trace.
+ * The flags 'TCL_TRACE_DESTROYED' and 'TCL_INTERP_DESTROYED' may also be used
+ * in command execution traces.
+ */
+
+#define TCL_TRACE_ENTER_DURING_EXEC	4
+#define TCL_TRACE_LEAVE_DURING_EXEC	8
+#define TCL_TRACE_ANY_EXEC		15
+#define TCL_TRACE_EXEC_IN_PROGRESS	0x10
+#define TCL_TRACE_EXEC_DIRECT		0x20
+
+/*
+ * Forward declarations for functions defined in this file:
+ */
+
+typedef int (Tcl_TraceTypeObjCmd)(Tcl_Interp *interp, int optionIndex,
+	int objc, Tcl_Obj *const objv[]);
+
+static Tcl_TraceTypeObjCmd TraceVariableObjCmd;
+static Tcl_TraceTypeObjCmd TraceCommandObjCmd;
+static Tcl_TraceTypeObjCmd TraceExecutionObjCmd;
+
+/*
+ * Each subcommand has a number of 'types' to which it can apply. Currently
+ * 'execution', 'command' and 'variable' are the only types supported. These
+ * three arrays MUST be kept in sync! In the future we may provide an API to
+ * add to the list of supported trace types.
+ */
+
+static const char *const traceTypeOptions[] = {
+    "execution", "command", "variable", NULL
+};
+static Tcl_TraceTypeObjCmd *const traceSubCmds[] = {
+    TraceExecutionObjCmd,
+    TraceCommandObjCmd,
+    TraceVariableObjCmd
+};
+
+/*
+ * Declarations for local functions to this file:
+ */
+
+static int		CallTraceFunction(Tcl_Interp *interp, Trace *tracePtr,
+			    Command *cmdPtr, const char *command, int numChars,
+			    int objc, Tcl_Obj *const objv[]);
+static char *		TraceVarProc(ClientData clientData, Tcl_Interp *interp,
+			    const char *name1, const char *name2, int flags);
+static void		TraceCommandProc(ClientData clientData,
+			    Tcl_Interp *interp, const char *oldName,
+			    const char *newName, int flags);
+static Tcl_CmdObjTraceProc TraceExecutionProc;
+static int		StringTraceProc(ClientData clientData,
+			    Tcl_Interp *interp, int level,
+			    const char *command, Tcl_Command commandInfo,
+			    int objc, Tcl_Obj *const objv[]);
+static void		StringTraceDeleteProc(ClientData clientData);
+static void		DisposeTraceResult(int flags, char *result);
+static int		TraceVarEx(Tcl_Interp *interp, const char *part1,
+			    const char *part2, register VarTrace *tracePtr);
+
+/*
+ * The following structure holds the client data for string-based
+ * trace procs
+ */
+
+typedef struct {
+    ClientData clientData;	/* Client data from Tcl_CreateTrace */
+    Tcl_CmdTraceProc *proc;	/* Trace function from Tcl_CreateTrace */
+} StringTraceData;
+
+/*
+ * Convenience macros for iterating over the list of traces. Note that each of
+ * these *must* be treated as a command, and *must* have a block following it.
+ */
+
+#define FOREACH_VAR_TRACE(interp, name, clientData) \
+    (clientData) = NULL; \
+    while (((clientData) = Tcl_VarTraceInfo2((interp), (name), NULL, \
+	    0, TraceVarProc, (clientData))) != NULL)
+
+#define FOREACH_COMMAND_TRACE(interp, name, clientData) \
+    (clientData) = NULL; \
+    while ((clientData = Tcl_CommandTraceInfo(interp, name, 0, \
+	    TraceCommandProc, clientData)) != NULL)
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TraceObjCmd --
+ *
+ *	This function is invoked to process the "trace" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ *	Standard syntax as of Tcl 8.4 is:
+ *	    trace {add|info|remove} {command|variable} name ops cmd
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_TraceObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int optionIndex;
+    const char *name;
+    const char *flagOps, *p;
+    /* Main sub commands to 'trace' */
+    static const char *const traceOptions[] = {
+	"add", "info", "remove",
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+	"variable", "vdelete", "vinfo",
+#endif
+	NULL
+    };
+    /* 'OLD' options are pre-Tcl-8.4 style */
+    enum traceOptions {
+	TRACE_ADD, TRACE_INFO, TRACE_REMOVE,
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+	TRACE_OLD_VARIABLE, TRACE_OLD_VDELETE, TRACE_OLD_VINFO
+#endif
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option ?arg ...?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], traceOptions, "option", 0,
+	    &optionIndex) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch ((enum traceOptions) optionIndex) {
+    case TRACE_ADD:
+    case TRACE_REMOVE: {
+	/*
+	 * All sub commands of trace add/remove must take at least one more
+	 * argument. Beyond that we let the subcommand itself control the
+	 * argument structure.
+	 */
+
+	int typeIndex;
+
+	if (objc < 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "type ?arg ...?");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[2], traceTypeOptions, "option",
+		0, &typeIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	return traceSubCmds[typeIndex](interp, optionIndex, objc, objv);
+    }
+    case TRACE_INFO: {
+	/*
+	 * All sub commands of trace info must take exactly two more arguments
+	 * which name the type of thing being traced and the name of the thing
+	 * being traced.
+	 */
+
+	int typeIndex;
+	if (objc < 3) {
+	    /*
+	     * Delegate other complaints to the type-specific code which can
+	     * give a better error message.
+	     */
+
+	    Tcl_WrongNumArgs(interp, 2, objv, "type name");
+	    return TCL_ERROR;
+	}
+	if (Tcl_GetIndexFromObj(interp, objv[2], traceTypeOptions, "option",
+		0, &typeIndex) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	return traceSubCmds[typeIndex](interp, optionIndex, objc, objv);
+	break;
+    }
+
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+    case TRACE_OLD_VARIABLE:
+    case TRACE_OLD_VDELETE: {
+	Tcl_Obj *copyObjv[6];
+	Tcl_Obj *opsList;
+	int code, numFlags;
+
+	if (objc != 5) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "name ops command");
+	    return TCL_ERROR;
+	}
+
+	opsList = Tcl_NewObj();
+	Tcl_IncrRefCount(opsList);
+	flagOps = TclGetStringFromObj(objv[3], &numFlags);
+	if (numFlags == 0) {
+	    Tcl_DecrRefCount(opsList);
+	    goto badVarOps;
+	}
+	for (p = flagOps; *p != 0; p++) {
+	    Tcl_Obj *opObj;
+
+	    if (*p == 'r') {
+		TclNewLiteralStringObj(opObj, "read");
+	    } else if (*p == 'w') {
+		TclNewLiteralStringObj(opObj, "write");
+	    } else if (*p == 'u') {
+		TclNewLiteralStringObj(opObj, "unset");
+	    } else if (*p == 'a') {
+		TclNewLiteralStringObj(opObj, "array");
+	    } else {
+		Tcl_DecrRefCount(opsList);
+		goto badVarOps;
+	    }
+	    Tcl_ListObjAppendElement(NULL, opsList, opObj);
+	}
+	copyObjv[0] = NULL;
+	memcpy(copyObjv+1, objv, objc*sizeof(Tcl_Obj *));
+	copyObjv[4] = opsList;
+	if (optionIndex == TRACE_OLD_VARIABLE) {
+	    code = traceSubCmds[2](interp, TRACE_ADD, objc+1, copyObjv);
+	} else {
+	    code = traceSubCmds[2](interp, TRACE_REMOVE, objc+1, copyObjv);
+	}
+	Tcl_DecrRefCount(opsList);
+	return code;
+    }
+    case TRACE_OLD_VINFO: {
+	ClientData clientData;
+	char ops[5];
+	Tcl_Obj *resultListPtr, *pairObjPtr, *elemObjPtr;
+
+	if (objc != 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "name");
+	    return TCL_ERROR;
+	}
+	resultListPtr = Tcl_NewObj();
+	name = Tcl_GetString(objv[2]);
+	FOREACH_VAR_TRACE(interp, name, clientData) {
+	    TraceVarInfo *tvarPtr = clientData;
+	    char *q = ops;
+
+	    pairObjPtr = Tcl_NewListObj(0, NULL);
+	    if (tvarPtr->flags & TCL_TRACE_READS) {
+		*q = 'r';
+		q++;
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_WRITES) {
+		*q = 'w';
+		q++;
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_UNSETS) {
+		*q = 'u';
+		q++;
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_ARRAY) {
+		*q = 'a';
+		q++;
+	    }
+	    *q = '\0';
+
+	    /*
+	     * Build a pair (2-item list) with the ops string as the first obj
+	     * element and the tvarPtr->command string as the second obj
+	     * element. Append the pair (as an element) to the end of the
+	     * result object list.
+	     */
+
+	    elemObjPtr = Tcl_NewStringObj(ops, -1);
+	    Tcl_ListObjAppendElement(NULL, pairObjPtr, elemObjPtr);
+	    elemObjPtr = Tcl_NewStringObj(tvarPtr->command, -1);
+	    Tcl_ListObjAppendElement(NULL, pairObjPtr, elemObjPtr);
+	    Tcl_ListObjAppendElement(interp, resultListPtr, pairObjPtr);
+	}
+	Tcl_SetObjResult(interp, resultListPtr);
+	break;
+    }
+#endif /* TCL_REMOVE_OBSOLETE_TRACES */
+    }
+    return TCL_OK;
+
+  badVarOps:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "bad operations \"%s\": should be one or more of rwua",
+	    flagOps));
+    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRACE", "BADOPS", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceExecutionObjCmd --
+ *
+ *	Helper function for Tcl_TraceObjCmd; implements the [trace
+ *	{add|remove|info} execution ...] subcommands. See the user
+ *	documentation for details on what these do.
+ *
+ * Results:
+ *	Standard Tcl result.
+ *
+ * Side effects:
+ *	Depends on the operation (add, remove, or info) being performed; may
+ *	add or remove command traces on a command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TraceExecutionObjCmd(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int optionIndex,		/* Add, info or remove */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int commandLength, index;
+    const char *name, *command;
+    size_t length;
+    enum traceOptions {
+	TRACE_ADD, TRACE_INFO, TRACE_REMOVE
+    };
+    static const char *const opStrings[] = {
+	"enter", "leave", "enterstep", "leavestep", NULL
+    };
+    enum operations {
+	TRACE_EXEC_ENTER, TRACE_EXEC_LEAVE,
+	TRACE_EXEC_ENTER_STEP, TRACE_EXEC_LEAVE_STEP
+    };
+
+    switch ((enum traceOptions) optionIndex) {
+    case TRACE_ADD:
+    case TRACE_REMOVE: {
+	int flags = 0;
+	int i, listLen, result;
+	Tcl_Obj **elemPtrs;
+
+	if (objc != 6) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name opList command");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Make sure the ops argument is a list object; get its length and a
+	 * pointer to its array of element pointers.
+	 */
+
+	result = Tcl_ListObjGetElements(interp, objv[4], &listLen, &elemPtrs);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	if (listLen == 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "bad operation list \"\": must be one or more of"
+		    " enter, leave, enterstep, or leavestep", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRACE", "NOOPS",
+		    NULL);
+	    return TCL_ERROR;
+	}
+	for (i = 0; i < listLen; i++) {
+	    if (Tcl_GetIndexFromObj(interp, elemPtrs[i], opStrings,
+		    "operation", TCL_EXACT, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum operations) index) {
+	    case TRACE_EXEC_ENTER:
+		flags |= TCL_TRACE_ENTER_EXEC;
+		break;
+	    case TRACE_EXEC_LEAVE:
+		flags |= TCL_TRACE_LEAVE_EXEC;
+		break;
+	    case TRACE_EXEC_ENTER_STEP:
+		flags |= TCL_TRACE_ENTER_DURING_EXEC;
+		break;
+	    case TRACE_EXEC_LEAVE_STEP:
+		flags |= TCL_TRACE_LEAVE_DURING_EXEC;
+		break;
+	    }
+	}
+	command = TclGetStringFromObj(objv[5], &commandLength);
+	length = (size_t) commandLength;
+	if ((enum traceOptions) optionIndex == TRACE_ADD) {
+	    TraceCommandInfo *tcmdPtr = ckalloc(
+		    TclOffset(TraceCommandInfo, command) + 1 + length);
+
+	    tcmdPtr->flags = flags;
+	    tcmdPtr->stepTrace = NULL;
+	    tcmdPtr->startLevel = 0;
+	    tcmdPtr->startCmd = NULL;
+	    tcmdPtr->length = length;
+	    tcmdPtr->refCount = 1;
+	    flags |= TCL_TRACE_DELETE;
+	    if (flags & (TCL_TRACE_ENTER_DURING_EXEC |
+		    TCL_TRACE_LEAVE_DURING_EXEC)) {
+		flags |= (TCL_TRACE_ENTER_EXEC | TCL_TRACE_LEAVE_EXEC);
+	    }
+	    memcpy(tcmdPtr->command, command, length+1);
+	    name = Tcl_GetString(objv[3]);
+	    if (Tcl_TraceCommand(interp, name, flags, TraceCommandProc,
+		    tcmdPtr) != TCL_OK) {
+		ckfree(tcmdPtr);
+		return TCL_ERROR;
+	    }
+	} else {
+	    /*
+	     * Search through all of our traces on this command to see if
+	     * there's one with the given command. If so, then delete the
+	     * first one that matches.
+	     */
+
+	    ClientData clientData;
+
+	    /*
+	     * First ensure the name given is valid.
+	     */
+
+	    name = Tcl_GetString(objv[3]);
+	    if (Tcl_FindCommand(interp,name,NULL,TCL_LEAVE_ERR_MSG) == NULL) {
+		return TCL_ERROR;
+	    }
+
+	    FOREACH_COMMAND_TRACE(interp, name, clientData) {
+		TraceCommandInfo *tcmdPtr = clientData;
+
+		/*
+		 * In checking the 'flags' field we must remove any extraneous
+		 * flags which may have been temporarily added by various
+		 * pieces of the trace mechanism.
+		 */
+
+		if ((tcmdPtr->length == length)
+			&& ((tcmdPtr->flags & (TCL_TRACE_ANY_EXEC |
+				TCL_TRACE_RENAME | TCL_TRACE_DELETE)) == flags)
+			&& (strncmp(command, tcmdPtr->command,
+				(size_t) length) == 0)) {
+		    flags |= TCL_TRACE_DELETE;
+		    if (flags & (TCL_TRACE_ENTER_DURING_EXEC |
+			    TCL_TRACE_LEAVE_DURING_EXEC)) {
+			flags |= (TCL_TRACE_ENTER_EXEC | TCL_TRACE_LEAVE_EXEC);
+		    }
+		    Tcl_UntraceCommand(interp, name, flags,
+			    TraceCommandProc, clientData);
+		    if (tcmdPtr->stepTrace != NULL) {
+			/*
+			 * We need to remove the interpreter-wide trace which
+			 * we created to allow 'step' traces.
+			 */
+
+			Tcl_DeleteTrace(interp, tcmdPtr->stepTrace);
+			tcmdPtr->stepTrace = NULL;
+			if (tcmdPtr->startCmd != NULL) {
+			    ckfree(tcmdPtr->startCmd);
+			}
+		    }
+		    if (tcmdPtr->flags & TCL_TRACE_EXEC_IN_PROGRESS) {
+			/*
+			 * Postpone deletion.
+			 */
+
+			tcmdPtr->flags = 0;
+		    }
+		    if (tcmdPtr->refCount-- <= 1) {
+			ckfree(tcmdPtr);
+		    }
+		    break;
+		}
+	    }
+	}
+	break;
+    }
+    case TRACE_INFO: {
+	ClientData clientData;
+	Tcl_Obj *resultListPtr;
+
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name");
+	    return TCL_ERROR;
+	}
+
+	name = Tcl_GetString(objv[3]);
+
+	/*
+	 * First ensure the name given is valid.
+	 */
+
+	if (Tcl_FindCommand(interp, name, NULL, TCL_LEAVE_ERR_MSG) == NULL) {
+	    return TCL_ERROR;
+	}
+
+	resultListPtr = Tcl_NewListObj(0, NULL);
+	FOREACH_COMMAND_TRACE(interp, name, clientData) {
+	    int numOps = 0;
+	    Tcl_Obj *opObj, *eachTraceObjPtr, *elemObjPtr;
+	    TraceCommandInfo *tcmdPtr = clientData;
+
+	    /*
+	     * Build a list with the ops list as the first obj element and the
+	     * tcmdPtr->command string as the second obj element. Append this
+	     * list (as an element) to the end of the result object list.
+	     */
+
+	    elemObjPtr = Tcl_NewListObj(0, NULL);
+	    Tcl_IncrRefCount(elemObjPtr);
+	    if (tcmdPtr->flags & TCL_TRACE_ENTER_EXEC) {
+		TclNewLiteralStringObj(opObj, "enter");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    if (tcmdPtr->flags & TCL_TRACE_LEAVE_EXEC) {
+		TclNewLiteralStringObj(opObj, "leave");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    if (tcmdPtr->flags & TCL_TRACE_ENTER_DURING_EXEC) {
+		TclNewLiteralStringObj(opObj, "enterstep");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    if (tcmdPtr->flags & TCL_TRACE_LEAVE_DURING_EXEC) {
+		TclNewLiteralStringObj(opObj, "leavestep");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    Tcl_ListObjLength(NULL, elemObjPtr, &numOps);
+	    if (0 == numOps) {
+		Tcl_DecrRefCount(elemObjPtr);
+		continue;
+	    }
+	    eachTraceObjPtr = Tcl_NewListObj(0, NULL);
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr, elemObjPtr);
+	    Tcl_DecrRefCount(elemObjPtr);
+	    elemObjPtr = NULL;
+
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr,
+		    Tcl_NewStringObj(tcmdPtr->command, -1));
+	    Tcl_ListObjAppendElement(interp, resultListPtr, eachTraceObjPtr);
+	}
+	Tcl_SetObjResult(interp, resultListPtr);
+	break;
+    }
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceCommandObjCmd --
+ *
+ *	Helper function for Tcl_TraceObjCmd; implements the [trace
+ *	{add|info|remove} command ...] subcommands. See the user documentation
+ *	for details on what these do.
+ *
+ * Results:
+ *	Standard Tcl result.
+ *
+ * Side effects:
+ *	Depends on the operation (add, remove, or info) being performed; may
+ *	add or remove command traces on a command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TraceCommandObjCmd(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int optionIndex,		/* Add, info or remove */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int commandLength, index;
+    const char *name, *command;
+    size_t length;
+    enum traceOptions { TRACE_ADD, TRACE_INFO, TRACE_REMOVE };
+    static const char *const opStrings[] = { "delete", "rename", NULL };
+    enum operations { TRACE_CMD_DELETE, TRACE_CMD_RENAME };
+
+    switch ((enum traceOptions) optionIndex) {
+    case TRACE_ADD:
+    case TRACE_REMOVE: {
+	int flags = 0;
+	int i, listLen, result;
+	Tcl_Obj **elemPtrs;
+
+	if (objc != 6) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name opList command");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Make sure the ops argument is a list object; get its length and a
+	 * pointer to its array of element pointers.
+	 */
+
+	result = Tcl_ListObjGetElements(interp, objv[4], &listLen, &elemPtrs);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	if (listLen == 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "bad operation list \"\": must be one or more of"
+		    " delete or rename", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRACE", "NOOPS",
+		    NULL);
+	    return TCL_ERROR;
+	}
+
+	for (i = 0; i < listLen; i++) {
+	    if (Tcl_GetIndexFromObj(interp, elemPtrs[i], opStrings,
+		    "operation", TCL_EXACT, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum operations) index) {
+	    case TRACE_CMD_RENAME:
+		flags |= TCL_TRACE_RENAME;
+		break;
+	    case TRACE_CMD_DELETE:
+		flags |= TCL_TRACE_DELETE;
+		break;
+	    }
+	}
+
+	command = TclGetStringFromObj(objv[5], &commandLength);
+	length = (size_t) commandLength;
+	if ((enum traceOptions) optionIndex == TRACE_ADD) {
+	    TraceCommandInfo *tcmdPtr = ckalloc(
+		    TclOffset(TraceCommandInfo, command) + 1 + length);
+
+	    tcmdPtr->flags = flags;
+	    tcmdPtr->stepTrace = NULL;
+	    tcmdPtr->startLevel = 0;
+	    tcmdPtr->startCmd = NULL;
+	    tcmdPtr->length = length;
+	    tcmdPtr->refCount = 1;
+	    flags |= TCL_TRACE_DELETE;
+	    memcpy(tcmdPtr->command, command, length+1);
+	    name = Tcl_GetString(objv[3]);
+	    if (Tcl_TraceCommand(interp, name, flags, TraceCommandProc,
+		    tcmdPtr) != TCL_OK) {
+		ckfree(tcmdPtr);
+		return TCL_ERROR;
+	    }
+	} else {
+	    /*
+	     * Search through all of our traces on this command to see if
+	     * there's one with the given command. If so, then delete the
+	     * first one that matches.
+	     */
+
+	    ClientData clientData;
+
+	    /*
+	     * First ensure the name given is valid.
+	     */
+
+	    name = Tcl_GetString(objv[3]);
+	    if (Tcl_FindCommand(interp,name,NULL,TCL_LEAVE_ERR_MSG) == NULL) {
+		return TCL_ERROR;
+	    }
+
+	    FOREACH_COMMAND_TRACE(interp, name, clientData) {
+		TraceCommandInfo *tcmdPtr = clientData;
+
+		if ((tcmdPtr->length == length) && (tcmdPtr->flags == flags)
+			&& (strncmp(command, tcmdPtr->command,
+				(size_t) length) == 0)) {
+		    Tcl_UntraceCommand(interp, name, flags | TCL_TRACE_DELETE,
+			    TraceCommandProc, clientData);
+		    tcmdPtr->flags |= TCL_TRACE_DESTROYED;
+		    if (tcmdPtr->refCount-- <= 1) {
+			ckfree(tcmdPtr);
+		    }
+		    break;
+		}
+	    }
+	}
+	break;
+    }
+    case TRACE_INFO: {
+	ClientData clientData;
+	Tcl_Obj *resultListPtr;
+
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * First ensure the name given is valid.
+	 */
+
+	name = Tcl_GetString(objv[3]);
+	if (Tcl_FindCommand(interp, name, NULL, TCL_LEAVE_ERR_MSG) == NULL) {
+	    return TCL_ERROR;
+	}
+
+	resultListPtr = Tcl_NewListObj(0, NULL);
+	FOREACH_COMMAND_TRACE(interp, name, clientData) {
+	    int numOps = 0;
+	    Tcl_Obj *opObj, *eachTraceObjPtr, *elemObjPtr;
+	    TraceCommandInfo *tcmdPtr = clientData;
+
+	    /*
+	     * Build a list with the ops list as the first obj element and the
+	     * tcmdPtr->command string as the second obj element. Append this
+	     * list (as an element) to the end of the result object list.
+	     */
+
+	    elemObjPtr = Tcl_NewListObj(0, NULL);
+	    Tcl_IncrRefCount(elemObjPtr);
+	    if (tcmdPtr->flags & TCL_TRACE_RENAME) {
+		TclNewLiteralStringObj(opObj, "rename");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    if (tcmdPtr->flags & TCL_TRACE_DELETE) {
+		TclNewLiteralStringObj(opObj, "delete");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObj);
+	    }
+	    Tcl_ListObjLength(NULL, elemObjPtr, &numOps);
+	    if (0 == numOps) {
+		Tcl_DecrRefCount(elemObjPtr);
+		continue;
+	    }
+	    eachTraceObjPtr = Tcl_NewListObj(0, NULL);
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr, elemObjPtr);
+	    Tcl_DecrRefCount(elemObjPtr);
+
+	    elemObjPtr = Tcl_NewStringObj(tcmdPtr->command, -1);
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr, elemObjPtr);
+	    Tcl_ListObjAppendElement(interp, resultListPtr, eachTraceObjPtr);
+	}
+	Tcl_SetObjResult(interp, resultListPtr);
+	break;
+    }
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceVariableObjCmd --
+ *
+ *	Helper function for Tcl_TraceObjCmd; implements the [trace
+ *	{add|info|remove} variable ...] subcommands. See the user
+ *	documentation for details on what these do.
+ *
+ * Results:
+ *	Standard Tcl result.
+ *
+ * Side effects:
+ *	Depends on the operation (add, remove, or info) being performed; may
+ *	add or remove variable traces on a variable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TraceVariableObjCmd(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int optionIndex,		/* Add, info or remove */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    int commandLength, index;
+    const char *name, *command;
+    size_t length;
+    ClientData clientData;
+    enum traceOptions { TRACE_ADD, TRACE_INFO, TRACE_REMOVE };
+    static const char *const opStrings[] = {
+	"array", "read", "unset", "write", NULL
+    };
+    enum operations {
+	TRACE_VAR_ARRAY, TRACE_VAR_READ, TRACE_VAR_UNSET, TRACE_VAR_WRITE
+    };
+
+    switch ((enum traceOptions) optionIndex) {
+    case TRACE_ADD:
+    case TRACE_REMOVE: {
+	int flags = 0;
+	int i, listLen, result;
+	Tcl_Obj **elemPtrs;
+
+	if (objc != 6) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name opList command");
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Make sure the ops argument is a list object; get its length and a
+	 * pointer to its array of element pointers.
+	 */
+
+	result = Tcl_ListObjGetElements(interp, objv[4], &listLen, &elemPtrs);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	if (listLen == 0) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "bad operation list \"\": must be one or more of"
+		    " array, read, unset, or write", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "OPERATION", "TRACE", "NOOPS",
+		    NULL);
+	    return TCL_ERROR;
+	}
+	for (i = 0; i < listLen ; i++) {
+	    if (Tcl_GetIndexFromObj(interp, elemPtrs[i], opStrings,
+		    "operation", TCL_EXACT, &index) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch ((enum operations) index) {
+	    case TRACE_VAR_ARRAY:
+		flags |= TCL_TRACE_ARRAY;
+		break;
+	    case TRACE_VAR_READ:
+		flags |= TCL_TRACE_READS;
+		break;
+	    case TRACE_VAR_UNSET:
+		flags |= TCL_TRACE_UNSETS;
+		break;
+	    case TRACE_VAR_WRITE:
+		flags |= TCL_TRACE_WRITES;
+		break;
+	    }
+	}
+	command = TclGetStringFromObj(objv[5], &commandLength);
+	length = (size_t) commandLength;
+	if ((enum traceOptions) optionIndex == TRACE_ADD) {
+	    CombinedTraceVarInfo *ctvarPtr = ckalloc(
+		    TclOffset(CombinedTraceVarInfo, traceCmdInfo.command)
+		    + 1 + length);
+
+	    ctvarPtr->traceCmdInfo.flags = flags;
+	    if (objv[0] == NULL) {
+		ctvarPtr->traceCmdInfo.flags |= TCL_TRACE_OLD_STYLE;
+	    }
+	    ctvarPtr->traceCmdInfo.length = length;
+	    flags |= TCL_TRACE_UNSETS | TCL_TRACE_RESULT_OBJECT;
+	    memcpy(ctvarPtr->traceCmdInfo.command, command, length+1);
+	    ctvarPtr->traceInfo.traceProc = TraceVarProc;
+	    ctvarPtr->traceInfo.clientData = &ctvarPtr->traceCmdInfo;
+	    ctvarPtr->traceInfo.flags = flags;
+	    name = Tcl_GetString(objv[3]);
+	    if (TraceVarEx(interp, name, NULL, (VarTrace *) ctvarPtr)
+		    != TCL_OK) {
+		ckfree(ctvarPtr);
+		return TCL_ERROR;
+	    }
+	} else {
+	    /*
+	     * Search through all of our traces on this variable to see if
+	     * there's one with the given command. If so, then delete the
+	     * first one that matches.
+	     */
+
+	    name = Tcl_GetString(objv[3]);
+	    FOREACH_VAR_TRACE(interp, name, clientData) {
+		TraceVarInfo *tvarPtr = clientData;
+
+		if ((tvarPtr->length == length)
+			&& ((tvarPtr->flags & ~TCL_TRACE_OLD_STYLE)==flags)
+			&& (strncmp(command, tvarPtr->command,
+				(size_t) length) == 0)) {
+		    Tcl_UntraceVar2(interp, name, NULL,
+			    flags | TCL_TRACE_UNSETS | TCL_TRACE_RESULT_OBJECT,
+			    TraceVarProc, clientData);
+		    break;
+		}
+	    }
+	}
+	break;
+    }
+    case TRACE_INFO: {
+	Tcl_Obj *resultListPtr;
+
+	if (objc != 4) {
+	    Tcl_WrongNumArgs(interp, 3, objv, "name");
+	    return TCL_ERROR;
+	}
+
+	resultListPtr = Tcl_NewObj();
+	name = Tcl_GetString(objv[3]);
+	FOREACH_VAR_TRACE(interp, name, clientData) {
+	    Tcl_Obj *opObjPtr, *eachTraceObjPtr, *elemObjPtr;
+	    TraceVarInfo *tvarPtr = clientData;
+
+	    /*
+	     * Build a list with the ops list as the first obj element and the
+	     * tcmdPtr->command string as the second obj element. Append this
+	     * list (as an element) to the end of the result object list.
+	     */
+
+	    elemObjPtr = Tcl_NewListObj(0, NULL);
+	    if (tvarPtr->flags & TCL_TRACE_ARRAY) {
+		TclNewLiteralStringObj(opObjPtr, "array");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObjPtr);
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_READS) {
+		TclNewLiteralStringObj(opObjPtr, "read");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObjPtr);
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_WRITES) {
+		TclNewLiteralStringObj(opObjPtr, "write");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObjPtr);
+	    }
+	    if (tvarPtr->flags & TCL_TRACE_UNSETS) {
+		TclNewLiteralStringObj(opObjPtr, "unset");
+		Tcl_ListObjAppendElement(NULL, elemObjPtr, opObjPtr);
+	    }
+	    eachTraceObjPtr = Tcl_NewListObj(0, NULL);
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr, elemObjPtr);
+
+	    elemObjPtr = Tcl_NewStringObj(tvarPtr->command, -1);
+	    Tcl_ListObjAppendElement(NULL, eachTraceObjPtr, elemObjPtr);
+	    Tcl_ListObjAppendElement(interp, resultListPtr,
+		    eachTraceObjPtr);
+	}
+	Tcl_SetObjResult(interp, resultListPtr);
+	break;
+    }
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CommandTraceInfo --
+ *
+ *	Return the clientData value associated with a trace on a command.
+ *	This function can also be used to step through all of the traces on a
+ *	particular command that have the same trace function.
+ *
+ * Results:
+ *	The return value is the clientData value associated with a trace on
+ *	the given command. Information will only be returned for a trace with
+ *	proc as trace function. If the clientData argument is NULL then the
+ *	first such trace is returned; otherwise, the next relevant one after
+ *	the one given by clientData will be returned. If the command doesn't
+ *	exist then an error message is left in the interpreter and NULL is
+ *	returned. Also, if there are no (more) traces for the given command,
+ *	NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_CommandTraceInfo(
+    Tcl_Interp *interp,		/* Interpreter containing command. */
+    const char *cmdName,	/* Name of command. */
+    int flags,			/* OR-ed combo or TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY (can be 0). */
+    Tcl_CommandTraceProc *proc,	/* Function assocated with trace. */
+    ClientData prevClientData)	/* If non-NULL, gives last value returned by
+				 * this function, so this call will return the
+				 * next trace after that one. If NULL, this
+				 * call will return the first trace. */
+{
+    Command *cmdPtr;
+    register CommandTrace *tracePtr;
+
+    cmdPtr = (Command *) Tcl_FindCommand(interp, cmdName, NULL,
+	    TCL_LEAVE_ERR_MSG);
+    if (cmdPtr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * Find the relevant trace, if any, and return its clientData.
+     */
+
+    tracePtr = cmdPtr->tracePtr;
+    if (prevClientData != NULL) {
+	for (; tracePtr!=NULL ; tracePtr=tracePtr->nextPtr) {
+	    if ((tracePtr->clientData == prevClientData)
+		    && (tracePtr->traceProc == proc)) {
+		tracePtr = tracePtr->nextPtr;
+		break;
+	    }
+	}
+    }
+    for (; tracePtr!=NULL ; tracePtr=tracePtr->nextPtr) {
+	if (tracePtr->traceProc == proc) {
+	    return tracePtr->clientData;
+	}
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TraceCommand --
+ *
+ *	Arrange for rename/deletes to a command to cause a function to be
+ *	invoked, which can monitor the operations.
+ *
+ *	Also optionally arrange for execution of that command to cause a
+ *	function to be invoked.
+ *
+ * Results:
+ *	A standard Tcl return value.
+ *
+ * Side effects:
+ *	A trace is set up on the command given by cmdName, such that future
+ *	changes to the command will be intermediated by proc. See the manual
+ *	entry for complete details on the calling sequence for proc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_TraceCommand(
+    Tcl_Interp *interp,		/* Interpreter in which command is to be
+				 * traced. */
+    const char *cmdName,	/* Name of command. */
+    int flags,			/* OR-ed collection of bits, including any of
+				 * TCL_TRACE_RENAME, TCL_TRACE_DELETE, and any
+				 * of the TRACE_*_EXEC flags */
+    Tcl_CommandTraceProc *proc,	/* Function to call when specified ops are
+				 * invoked upon cmdName. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    Command *cmdPtr;
+    register CommandTrace *tracePtr;
+
+    cmdPtr = (Command *) Tcl_FindCommand(interp, cmdName, NULL,
+	    TCL_LEAVE_ERR_MSG);
+    if (cmdPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Set up trace information.
+     */
+
+    tracePtr = ckalloc(sizeof(CommandTrace));
+    tracePtr->traceProc = proc;
+    tracePtr->clientData = clientData;
+    tracePtr->flags = flags &
+	    (TCL_TRACE_RENAME | TCL_TRACE_DELETE | TCL_TRACE_ANY_EXEC);
+    tracePtr->nextPtr = cmdPtr->tracePtr;
+    tracePtr->refCount = 1;
+    cmdPtr->tracePtr = tracePtr;
+    if (tracePtr->flags & TCL_TRACE_ANY_EXEC) {
+	/*
+	 * Bug 3484621: up the interp's epoch if this is a BC'ed command
+	 */
+
+	if ((cmdPtr->compileProc != NULL) && !(cmdPtr->flags & CMD_HAS_EXEC_TRACES)){
+	    Interp *iPtr = (Interp *) interp;
+	    iPtr->compileEpoch++;
+	}
+	cmdPtr->flags |= CMD_HAS_EXEC_TRACES;
+    }
+
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UntraceCommand --
+ *
+ *	Remove a previously-created trace for a command.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there exists a trace for the command given by cmdName with the
+ *	given flags, proc, and clientData, then that trace is removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_UntraceCommand(
+    Tcl_Interp *interp,		/* Interpreter containing command. */
+    const char *cmdName,	/* Name of command. */
+    int flags,			/* OR-ed collection of bits, including any of
+				 * TCL_TRACE_RENAME, TCL_TRACE_DELETE, and any
+				 * of the TRACE_*_EXEC flags */
+    Tcl_CommandTraceProc *proc,	/* Function assocated with trace. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    register CommandTrace *tracePtr;
+    CommandTrace *prevPtr;
+    Command *cmdPtr;
+    Interp *iPtr = (Interp *) interp;
+    ActiveCommandTrace *activePtr;
+    int hasExecTraces = 0;
+
+    cmdPtr = (Command *) Tcl_FindCommand(interp, cmdName, NULL,
+	    TCL_LEAVE_ERR_MSG);
+    if (cmdPtr == NULL) {
+	return;
+    }
+
+    flags &= (TCL_TRACE_RENAME | TCL_TRACE_DELETE | TCL_TRACE_ANY_EXEC);
+
+    for (tracePtr = cmdPtr->tracePtr, prevPtr = NULL; ;
+	    prevPtr = tracePtr, tracePtr = tracePtr->nextPtr) {
+	if (tracePtr == NULL) {
+	    return;
+	}
+	if ((tracePtr->traceProc == proc)
+		&& ((tracePtr->flags & (TCL_TRACE_RENAME | TCL_TRACE_DELETE |
+			TCL_TRACE_ANY_EXEC)) == flags)
+		&& (tracePtr->clientData == clientData)) {
+	    if (tracePtr->flags & TCL_TRACE_ANY_EXEC) {
+		hasExecTraces = 1;
+	    }
+	    break;
+	}
+    }
+
+    /*
+     * The code below makes it possible to delete traces while traces are
+     * active: it makes sure that the deleted trace won't be processed by
+     * CallCommandTraces.
+     */
+
+    for (activePtr = iPtr->activeCmdTracePtr;  activePtr != NULL;
+	    activePtr = activePtr->nextPtr) {
+	if (activePtr->nextTracePtr == tracePtr) {
+	    if (activePtr->reverseScan) {
+		activePtr->nextTracePtr = prevPtr;
+	    } else {
+		activePtr->nextTracePtr = tracePtr->nextPtr;
+	    }
+	}
+    }
+    if (prevPtr == NULL) {
+	cmdPtr->tracePtr = tracePtr->nextPtr;
+    } else {
+	prevPtr->nextPtr = tracePtr->nextPtr;
+    }
+    tracePtr->flags = 0;
+
+    if (tracePtr->refCount-- <= 1) {
+	ckfree(tracePtr);
+    }
+
+    if (hasExecTraces) {
+	for (tracePtr = cmdPtr->tracePtr, prevPtr = NULL; tracePtr != NULL ;
+		prevPtr = tracePtr, tracePtr = tracePtr->nextPtr) {
+	    if (tracePtr->flags & TCL_TRACE_ANY_EXEC) {
+		return;
+	    }
+	}
+
+	/*
+	 * None of the remaining traces on this command are execution traces.
+	 * We therefore remove this flag:
+	 */
+
+	cmdPtr->flags &= ~CMD_HAS_EXEC_TRACES;
+
+        /*
+	 * Bug 3484621: up the interp's epoch if this is a BC'ed command
+	 */
+
+	if (cmdPtr->compileProc != NULL) {
+	    Interp *iPtr = (Interp *) interp;
+	    iPtr->compileEpoch++;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceCommandProc --
+ *
+ *	This function is called to handle command changes that have been
+ *	traced using the "trace" command, when using the 'rename' or 'delete'
+ *	options.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Depends on the command associated with the trace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static void
+TraceCommandProc(
+    ClientData clientData,	/* Information about the command trace. */
+    Tcl_Interp *interp,		/* Interpreter containing command. */
+    const char *oldName,	/* Name of command being changed. */
+    const char *newName,	/* New name of command. Empty string or NULL
+				 * means command is being deleted (renamed to
+				 * ""). */
+    int flags)			/* OR-ed bits giving operation and other
+				 * information. */
+{
+    TraceCommandInfo *tcmdPtr = clientData;
+    int code;
+    Tcl_DString cmd;
+
+    tcmdPtr->refCount++;
+
+    if ((tcmdPtr->flags & flags) && !Tcl_InterpDeleted(interp)
+	    && !Tcl_LimitExceeded(interp)) {
+	/*
+	 * Generate a command to execute by appending list elements for the
+	 * old and new command name and the operation.
+	 */
+
+	Tcl_DStringInit(&cmd);
+	Tcl_DStringAppend(&cmd, tcmdPtr->command, (int) tcmdPtr->length);
+	Tcl_DStringAppendElement(&cmd, oldName);
+	Tcl_DStringAppendElement(&cmd, (newName ? newName : ""));
+	if (flags & TCL_TRACE_RENAME) {
+	    TclDStringAppendLiteral(&cmd, " rename");
+	} else if (flags & TCL_TRACE_DELETE) {
+	    TclDStringAppendLiteral(&cmd, " delete");
+	}
+
+	/*
+	 * Execute the command. We discard any object result the command
+	 * returns.
+	 *
+	 * Add the TCL_TRACE_DESTROYED flag to tcmdPtr to indicate to other
+	 * areas that this will be destroyed by us, otherwise a double-free
+	 * might occur depending on what the eval does.
+	 */
+
+	if (flags & TCL_TRACE_DESTROYED) {
+	    tcmdPtr->flags |= TCL_TRACE_DESTROYED;
+	}
+	code = Tcl_EvalEx(interp, Tcl_DStringValue(&cmd),
+		Tcl_DStringLength(&cmd), 0);
+	if (code != TCL_OK) {
+	    /* We ignore errors in these traced commands */
+	    /*** QUESTION: Use Tcl_BackgroundException(interp, code); instead? ***/
+	}
+	Tcl_DStringFree(&cmd);
+    }
+
+    /*
+     * We delete when the trace was destroyed or if this is a delete trace,
+     * because command deletes are unconditional, so the trace must go away.
+     */
+
+    if (flags & (TCL_TRACE_DESTROYED | TCL_TRACE_DELETE)) {
+	int untraceFlags = tcmdPtr->flags;
+	Tcl_InterpState state;
+
+	if (tcmdPtr->stepTrace != NULL) {
+	    Tcl_DeleteTrace(interp, tcmdPtr->stepTrace);
+	    tcmdPtr->stepTrace = NULL;
+	    if (tcmdPtr->startCmd != NULL) {
+		ckfree(tcmdPtr->startCmd);
+	    }
+	}
+	if (tcmdPtr->flags & TCL_TRACE_EXEC_IN_PROGRESS) {
+	    /*
+	     * Postpone deletion, until exec trace returns.
+	     */
+
+	    tcmdPtr->flags = 0;
+	}
+
+	/*
+	 * We need to construct the same flags for Tcl_UntraceCommand as were
+	 * passed to Tcl_TraceCommand. Reproduce the processing of [trace add
+	 * execution/command]. Be careful to keep this code in sync with that.
+	 */
+
+	if (untraceFlags & TCL_TRACE_ANY_EXEC) {
+	    untraceFlags |= TCL_TRACE_DELETE;
+	    if (untraceFlags & (TCL_TRACE_ENTER_DURING_EXEC
+		    | TCL_TRACE_LEAVE_DURING_EXEC)) {
+		untraceFlags |= (TCL_TRACE_ENTER_EXEC | TCL_TRACE_LEAVE_EXEC);
+	    }
+	} else if (untraceFlags & TCL_TRACE_RENAME) {
+	    untraceFlags |= TCL_TRACE_DELETE;
+	}
+
+	/*
+	 * Remove the trace since TCL_TRACE_DESTROYED tells us to, or the
+	 * command we're tracing has just gone away. Then decrement the
+	 * clientData refCount that was set up by trace creation.
+	 *
+	 * Note that we save the (return) state of the interpreter to prevent
+	 * bizarre error messages.
+	 */
+
+	state = Tcl_SaveInterpState(interp, TCL_OK);
+	Tcl_UntraceCommand(interp, oldName, untraceFlags,
+		TraceCommandProc, clientData);
+	Tcl_RestoreInterpState(interp, state);
+	tcmdPtr->refCount--;
+    }
+    if (tcmdPtr->refCount-- <= 1) {
+	ckfree(tcmdPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCheckExecutionTraces --
+ *
+ *	Checks on all current command execution traces, and invokes functions
+ *	which have been registered. This function can be used by other code
+ *	which performs execution to unify the tracing system, so that
+ *	execution traces will function for that other code.
+ *
+ *	For instance extensions like [incr Tcl] which use their own execution
+ *	technique can make use of Tcl's tracing.
+ *
+ *	This function is called by 'TclEvalObjvInternal'
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR, etc.
+ *
+ * Side effects:
+ *	Those side effects made by any trace functions called.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCheckExecutionTraces(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    const char *command,	/* Pointer to beginning of the current command
+				 * string. */
+    int numChars,		/* The number of characters in 'command' which
+				 * are part of the command string. */
+    Command *cmdPtr,		/* Points to command's Command struct. */
+    int code,			/* The current result code. */
+    int traceFlags,		/* Current tracing situation. */
+    int objc,			/* Number of arguments for the command. */
+    Tcl_Obj *const objv[])	/* Pointers to Tcl_Obj of each argument. */
+{
+    Interp *iPtr = (Interp *) interp;
+    CommandTrace *tracePtr, *lastTracePtr;
+    ActiveCommandTrace active;
+    int curLevel;
+    int traceCode = TCL_OK;
+    Tcl_InterpState state = NULL;
+
+    if (cmdPtr->tracePtr == NULL) {
+	return traceCode;
+    }
+
+    curLevel = iPtr->varFramePtr->level;
+
+    active.nextPtr = iPtr->activeCmdTracePtr;
+    iPtr->activeCmdTracePtr = &active;
+
+    active.cmdPtr = cmdPtr;
+    lastTracePtr = NULL;
+    for (tracePtr = cmdPtr->tracePtr;
+	    (traceCode == TCL_OK) && (tracePtr != NULL);
+	    tracePtr = active.nextTracePtr) {
+	if (traceFlags & TCL_TRACE_LEAVE_EXEC) {
+	    /*
+	     * Execute the trace command in order of creation for "leave".
+	     */
+
+	    active.reverseScan = 1;
+	    active.nextTracePtr = NULL;
+	    tracePtr = cmdPtr->tracePtr;
+	    while (tracePtr->nextPtr != lastTracePtr) {
+		active.nextTracePtr = tracePtr;
+		tracePtr = tracePtr->nextPtr;
+	    }
+	} else {
+	    active.reverseScan = 0;
+	    active.nextTracePtr = tracePtr->nextPtr;
+	}
+	if (tracePtr->traceProc == TraceCommandProc) {
+	    TraceCommandInfo *tcmdPtr = tracePtr->clientData;
+
+	    if (tcmdPtr->flags != 0) {
+		tcmdPtr->curFlags = traceFlags | TCL_TRACE_EXEC_DIRECT;
+		tcmdPtr->curCode  = code;
+		tcmdPtr->refCount++;
+		if (state == NULL) {
+		    state = Tcl_SaveInterpState(interp, code);
+		}
+		traceCode = TraceExecutionProc(tcmdPtr, interp, curLevel,
+			command, (Tcl_Command) cmdPtr, objc, objv);
+		if (tcmdPtr->refCount-- <= 1) {
+		    ckfree(tcmdPtr);
+		}
+	    }
+	}
+	if (active.nextTracePtr) {
+	    lastTracePtr = active.nextTracePtr->nextPtr;
+	}
+    }
+    iPtr->activeCmdTracePtr = active.nextPtr;
+    if (state) {
+	if (traceCode == TCL_OK) {
+	    (void) Tcl_RestoreInterpState(interp, state);
+	} else {
+	    Tcl_DiscardInterpState(state);
+	}
+    }
+
+    return traceCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCheckInterpTraces --
+ *
+ *	Checks on all current traces, and invokes functions which have been
+ *	registered. This function can be used by other code which performs
+ *	execution to unify the tracing system. For instance extensions like
+ *	[incr Tcl] which use their own execution technique can make use of
+ *	Tcl's tracing.
+ *
+ *	This function is called by 'TclEvalObjvInternal'
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR, etc.
+ *
+ * Side effects:
+ *	Those side effects made by any trace functions called.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCheckInterpTraces(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    const char *command,	/* Pointer to beginning of the current command
+				 * string. */
+    int numChars,		/* The number of characters in 'command' which
+				 * are part of the command string. */
+    Command *cmdPtr,		/* Points to command's Command struct. */
+    int code,			/* The current result code. */
+    int traceFlags,		/* Current tracing situation. */
+    int objc,			/* Number of arguments for the command. */
+    Tcl_Obj *const objv[])	/* Pointers to Tcl_Obj of each argument. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Trace *tracePtr, *lastTracePtr;
+    ActiveInterpTrace active;
+    int curLevel;
+    int traceCode = TCL_OK;
+    Tcl_InterpState state = NULL;
+
+    if ((iPtr->tracePtr == NULL)
+	    || (iPtr->flags & INTERP_TRACE_IN_PROGRESS)) {
+	return(traceCode);
+    }
+
+    curLevel = iPtr->numLevels;
+
+    active.nextPtr = iPtr->activeInterpTracePtr;
+    iPtr->activeInterpTracePtr = &active;
+
+    lastTracePtr = NULL;
+    for (tracePtr = iPtr->tracePtr;
+	    (traceCode == TCL_OK) && (tracePtr != NULL);
+	    tracePtr = active.nextTracePtr) {
+	if (traceFlags & TCL_TRACE_ENTER_EXEC) {
+	    /*
+	     * Execute the trace command in reverse order of creation for
+	     * "enterstep" operation. The order is changed for "enterstep"
+	     * instead of for "leavestep" as was done in
+	     * TclCheckExecutionTraces because for step traces,
+	     * Tcl_CreateObjTrace creates one more linked list of traces which
+	     * results in one more reversal of trace invocation.
+	     */
+
+	    active.reverseScan = 1;
+	    active.nextTracePtr = NULL;
+	    tracePtr = iPtr->tracePtr;
+	    while (tracePtr->nextPtr != lastTracePtr) {
+		active.nextTracePtr = tracePtr;
+		tracePtr = tracePtr->nextPtr;
+	    }
+	    if (active.nextTracePtr) {
+		lastTracePtr = active.nextTracePtr->nextPtr;
+	    }
+	} else {
+	    active.reverseScan = 0;
+	    active.nextTracePtr = tracePtr->nextPtr;
+	}
+
+	if (tracePtr->level > 0 && curLevel > tracePtr->level) {
+	    continue;
+	}
+
+	if (!(tracePtr->flags & TCL_TRACE_EXEC_IN_PROGRESS)) {
+	    /*
+	     * The proc invoked might delete the traced command which which
+	     * might try to free tracePtr. We want to use tracePtr until the
+	     * end of this if section, so we use Tcl_Preserve() and
+	     * Tcl_Release() to be sure it is not freed while we still need
+	     * it.
+	     */
+
+	    Tcl_Preserve(tracePtr);
+	    tracePtr->flags |= TCL_TRACE_EXEC_IN_PROGRESS;
+	    if (state == NULL) {
+		state = Tcl_SaveInterpState(interp, code);
+	    }
+
+	    if (tracePtr->flags &
+		    (TCL_TRACE_ENTER_EXEC | TCL_TRACE_LEAVE_EXEC)) {
+		/*
+		 * New style trace.
+		 */
+
+		if (tracePtr->flags & traceFlags) {
+		    if (tracePtr->proc == TraceExecutionProc) {
+			TraceCommandInfo *tcmdPtr = tracePtr->clientData;
+
+			tcmdPtr->curFlags = traceFlags;
+			tcmdPtr->curCode = code;
+		    }
+		    traceCode = tracePtr->proc(tracePtr->clientData, interp,
+			    curLevel, command, (Tcl_Command) cmdPtr, objc,
+			    objv);
+		}
+	    } else {
+		/*
+		 * Old-style trace.
+		 */
+
+		if (traceFlags & TCL_TRACE_ENTER_EXEC) {
+		    /*
+		     * Old-style interpreter-wide traces only trigger before
+		     * the command is executed.
+		     */
+
+		    traceCode = CallTraceFunction(interp, tracePtr, cmdPtr,
+			    command, numChars, objc, objv);
+		}
+	    }
+	    tracePtr->flags &= ~TCL_TRACE_EXEC_IN_PROGRESS;
+	    Tcl_Release(tracePtr);
+	}
+    }
+    iPtr->activeInterpTracePtr = active.nextPtr;
+    if (state) {
+	if (traceCode == TCL_OK) {
+	    Tcl_RestoreInterpState(interp, state);
+	} else {
+	    Tcl_DiscardInterpState(state);
+	}
+    }
+
+    return traceCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CallTraceFunction --
+ *
+ *	Invokes a trace function registered with an interpreter. These
+ *	functions trace command execution. Currently this trace function is
+ *	called with the address of the string-based Tcl_CmdProc for the
+ *	command, not the Tcl_ObjCmdProc.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Those side effects made by the trace function.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+CallTraceFunction(
+    Tcl_Interp *interp,		/* The current interpreter. */
+    register Trace *tracePtr,	/* Describes the trace function to call. */
+    Command *cmdPtr,		/* Points to command's Command struct. */
+    const char *command,	/* Points to the first character of the
+				 * command's source before substitutions. */
+    int numChars,		/* The number of characters in the command's
+				 * source. */
+    register int objc,		/* Number of arguments for the command. */
+    Tcl_Obj *const objv[])	/* Pointers to Tcl_Obj of each argument. */
+{
+    Interp *iPtr = (Interp *) interp;
+    char *commandCopy;
+    int traceCode;
+
+    /*
+     * Copy the command characters into a new string.
+     */
+
+    commandCopy = TclStackAlloc(interp, (unsigned) numChars + 1);
+    memcpy(commandCopy, command, (size_t) numChars);
+    commandCopy[numChars] = '\0';
+
+    /*
+     * Call the trace function then free allocated storage.
+     */
+
+    traceCode = tracePtr->proc(tracePtr->clientData, (Tcl_Interp *) iPtr,
+	    iPtr->numLevels, commandCopy, (Tcl_Command) cmdPtr, objc, objv);
+
+    TclStackFree(interp, commandCopy);
+    return traceCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * CommandObjTraceDeleted --
+ *
+ *	Ensure the trace is correctly deleted by decrementing its refCount and
+ *	only deleting if no other references exist.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	May release memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+CommandObjTraceDeleted(
+    ClientData clientData)
+{
+    TraceCommandInfo *tcmdPtr = clientData;
+
+    if (tcmdPtr->refCount-- <= 1) {
+	ckfree(tcmdPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceExecutionProc --
+ *
+ *	This function is invoked whenever code relevant to a 'trace execution'
+ *	command is executed. It is called in one of two ways in Tcl's core:
+ *
+ *	(i) by the TclCheckExecutionTraces, when an execution trace has been
+ *	triggered.
+ *	(ii) by TclCheckInterpTraces, when a prior execution trace has created
+ *	a trace of the internals of a procedure, passing in this function as
+ *	the one to be called.
+ *
+ * Results:
+ *	The return value is a standard Tcl completion code such as TCL_OK or
+ *	TCL_ERROR, etc.
+ *
+ * Side effects:
+ *	May invoke an arbitrary Tcl procedure, and may create or delete an
+ *	interpreter-wide trace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TraceExecutionProc(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int level,
+    const char *command,
+    Tcl_Command cmdInfo,
+    int objc,
+    struct Tcl_Obj *const objv[])
+{
+    int call = 0;
+    Interp *iPtr = (Interp *) interp;
+    TraceCommandInfo *tcmdPtr = clientData;
+    int flags = tcmdPtr->curFlags;
+    int code = tcmdPtr->curCode;
+    int traceCode = TCL_OK;
+
+    if (tcmdPtr->flags & TCL_TRACE_EXEC_IN_PROGRESS) {
+	/*
+	 * Inside any kind of execution trace callback, we do not allow any
+	 * further execution trace callbacks to be called for the same trace.
+	 */
+
+	return traceCode;
+    }
+
+    if (!Tcl_InterpDeleted(interp) && !Tcl_LimitExceeded(interp)) {
+	/*
+	 * Check whether the current call is going to eval arbitrary Tcl code
+	 * with a generated trace, or whether we are only going to setup
+	 * interpreter-wide traces to implement the 'step' traces. This latter
+	 * situation can happen if we create a command trace without either
+	 * before or after operations, but with either of the step operations.
+	 */
+
+	if (flags & TCL_TRACE_EXEC_DIRECT) {
+	    call = flags & tcmdPtr->flags &
+		    (TCL_TRACE_ENTER_EXEC | TCL_TRACE_LEAVE_EXEC);
+	} else {
+	    call = 1;
+	}
+
+	/*
+	 * First, if we have returned back to the level at which we created an
+	 * interpreter trace for enterstep and/or leavestep execution traces,
+	 * we remove it here.
+	 */
+
+	if ((flags & TCL_TRACE_LEAVE_EXEC) && (tcmdPtr->stepTrace != NULL)
+		&& (level == tcmdPtr->startLevel)
+		&& (strcmp(command, tcmdPtr->startCmd) == 0)) {
+	    Tcl_DeleteTrace(interp, tcmdPtr->stepTrace);
+	    tcmdPtr->stepTrace = NULL;
+	    if (tcmdPtr->startCmd != NULL) {
+		ckfree(tcmdPtr->startCmd);
+	    }
+	}
+
+	/*
+	 * Second, create the tcl callback, if required.
+	 */
+
+	if (call) {
+	    Tcl_DString cmd, sub;
+	    int i, saveInterpFlags;
+
+	    Tcl_DStringInit(&cmd);
+	    Tcl_DStringAppend(&cmd, tcmdPtr->command, (int)tcmdPtr->length);
+
+	    /*
+	     * Append command with arguments.
+	     */
+
+	    Tcl_DStringInit(&sub);
+	    for (i = 0; i < objc; i++) {
+		Tcl_DStringAppendElement(&sub, Tcl_GetString(objv[i]));
+	    }
+	    Tcl_DStringAppendElement(&cmd, Tcl_DStringValue(&sub));
+	    Tcl_DStringFree(&sub);
+
+	    if (flags & TCL_TRACE_ENTER_EXEC) {
+		/*
+		 * Append trace operation.
+		 */
+
+		if (flags & TCL_TRACE_EXEC_DIRECT) {
+		    Tcl_DStringAppendElement(&cmd, "enter");
+		} else {
+		    Tcl_DStringAppendElement(&cmd, "enterstep");
+		}
+	    } else if (flags & TCL_TRACE_LEAVE_EXEC) {
+		Tcl_Obj *resultCode;
+		const char *resultCodeStr;
+
+		/*
+		 * Append result code.
+		 */
+
+		resultCode = Tcl_NewIntObj(code);
+		resultCodeStr = Tcl_GetString(resultCode);
+		Tcl_DStringAppendElement(&cmd, resultCodeStr);
+		Tcl_DecrRefCount(resultCode);
+
+		/*
+		 * Append result string.
+		 */
+
+		Tcl_DStringAppendElement(&cmd, Tcl_GetStringResult(interp));
+
+		/*
+		 * Append trace operation.
+		 */
+
+		if (flags & TCL_TRACE_EXEC_DIRECT) {
+		    Tcl_DStringAppendElement(&cmd, "leave");
+		} else {
+		    Tcl_DStringAppendElement(&cmd, "leavestep");
+		}
+	    } else {
+		Tcl_Panic("TraceExecutionProc: bad flag combination");
+	    }
+
+	    /*
+	     * Execute the command. We discard any object result the command
+	     * returns.
+	     */
+
+	    saveInterpFlags = iPtr->flags;
+	    iPtr->flags    |= INTERP_TRACE_IN_PROGRESS;
+	    tcmdPtr->flags |= TCL_TRACE_EXEC_IN_PROGRESS;
+	    tcmdPtr->refCount++;
+
+	    /*
+	     * This line can have quite arbitrary side-effects, including
+	     * deleting the trace, the command being traced, or even the
+	     * interpreter.
+	     */
+
+	    traceCode = Tcl_EvalEx(interp, Tcl_DStringValue(&cmd),
+		    Tcl_DStringLength(&cmd), 0);
+	    tcmdPtr->flags &= ~TCL_TRACE_EXEC_IN_PROGRESS;
+
+	    /*
+	     * Restore the interp tracing flag to prevent cmd traces from
+	     * affecting interp traces.
+	     */
+
+	    iPtr->flags = saveInterpFlags;
+	    if (tcmdPtr->flags == 0) {
+		flags |= TCL_TRACE_DESTROYED;
+	    }
+	    Tcl_DStringFree(&cmd);
+	}
+
+	/*
+	 * Third, if there are any step execution traces for this proc, we
+	 * register an interpreter trace to invoke enterstep and/or leavestep
+	 * traces. We also need to save the current stack level and the proc
+	 * string in startLevel and startCmd so that we can delete this
+	 * interpreter trace when it reaches the end of this proc.
+	 */
+
+	if ((flags & TCL_TRACE_ENTER_EXEC) && (tcmdPtr->stepTrace == NULL)
+		&& (tcmdPtr->flags & (TCL_TRACE_ENTER_DURING_EXEC |
+			TCL_TRACE_LEAVE_DURING_EXEC))) {
+	    register unsigned len = strlen(command) + 1;
+
+	    tcmdPtr->startLevel = level;
+	    tcmdPtr->startCmd = ckalloc(len);
+	    memcpy(tcmdPtr->startCmd, command, len);
+	    tcmdPtr->refCount++;
+	    tcmdPtr->stepTrace = Tcl_CreateObjTrace(interp, 0,
+		   (tcmdPtr->flags & TCL_TRACE_ANY_EXEC) >> 2,
+		   TraceExecutionProc, tcmdPtr, CommandObjTraceDeleted);
+	}
+    }
+    if (flags & TCL_TRACE_DESTROYED) {
+	if (tcmdPtr->stepTrace != NULL) {
+	    Tcl_DeleteTrace(interp, tcmdPtr->stepTrace);
+	    tcmdPtr->stepTrace = NULL;
+	    if (tcmdPtr->startCmd != NULL) {
+		ckfree(tcmdPtr->startCmd);
+	    }
+	}
+    }
+    if (call) {
+	if (tcmdPtr->refCount-- <= 1) {
+	    ckfree(tcmdPtr);
+	}
+    }
+    return traceCode;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceVarProc --
+ *
+ *	This function is called to handle variable accesses that have been
+ *	traced using the "trace" command.
+ *
+ * Results:
+ *	Normally returns NULL. If the trace command returns an error, then
+ *	this function returns an error string.
+ *
+ * Side effects:
+ *	Depends on the command associated with the trace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static char *
+TraceVarProc(
+    ClientData clientData,	/* Information about the variable trace. */
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *name1,		/* Name of variable or array. */
+    const char *name2,		/* Name of element within array; NULL means
+				 * scalar variable is being referenced. */
+    int flags)			/* OR-ed bits giving operation and other
+				 * information. */
+{
+    TraceVarInfo *tvarPtr = clientData;
+    char *result;
+    int code, destroy = 0;
+    Tcl_DString cmd;
+    int rewind = ((Interp *)interp)->execEnvPtr->rewind;
+
+    /*
+     * We might call Tcl_Eval() below, and that might evaluate [trace vdelete]
+     * which might try to free tvarPtr. We want to use tvarPtr until the end
+     * of this function, so we use Tcl_Preserve() and Tcl_Release() to be sure
+     * it is not freed while we still need it.
+     */
+
+    result = NULL;
+    if ((tvarPtr->flags & flags) && !Tcl_InterpDeleted(interp)
+	    && !Tcl_LimitExceeded(interp)) {
+	if (tvarPtr->length != (size_t) 0) {
+	    /*
+	     * Generate a command to execute by appending list elements for
+	     * the two variable names and the operation.
+	     */
+
+	    Tcl_DStringInit(&cmd);
+	    Tcl_DStringAppend(&cmd, tvarPtr->command, (int) tvarPtr->length);
+	    Tcl_DStringAppendElement(&cmd, name1);
+	    Tcl_DStringAppendElement(&cmd, (name2 ? name2 : ""));
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+	    if (tvarPtr->flags & TCL_TRACE_OLD_STYLE) {
+		if (flags & TCL_TRACE_ARRAY) {
+		    TclDStringAppendLiteral(&cmd, " a");
+		} else if (flags & TCL_TRACE_READS) {
+		    TclDStringAppendLiteral(&cmd, " r");
+		} else if (flags & TCL_TRACE_WRITES) {
+		    TclDStringAppendLiteral(&cmd, " w");
+		} else if (flags & TCL_TRACE_UNSETS) {
+		    TclDStringAppendLiteral(&cmd, " u");
+		}
+	    } else {
+#endif
+		if (flags & TCL_TRACE_ARRAY) {
+		    TclDStringAppendLiteral(&cmd, " array");
+		} else if (flags & TCL_TRACE_READS) {
+		    TclDStringAppendLiteral(&cmd, " read");
+		} else if (flags & TCL_TRACE_WRITES) {
+		    TclDStringAppendLiteral(&cmd, " write");
+		} else if (flags & TCL_TRACE_UNSETS) {
+		    TclDStringAppendLiteral(&cmd, " unset");
+		}
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+	    }
+#endif
+
+	    /*
+	     * Execute the command. We discard any object result the command
+	     * returns.
+	     *
+	     * Add the TCL_TRACE_DESTROYED flag to tvarPtr to indicate to
+	     * other areas that this will be destroyed by us, otherwise a
+	     * double-free might occur depending on what the eval does.
+	     */
+
+	    if ((flags & TCL_TRACE_DESTROYED)
+		    && !(tvarPtr->flags & TCL_TRACE_DESTROYED)) {
+		destroy = 1;
+		tvarPtr->flags |= TCL_TRACE_DESTROYED;
+	    }
+
+	    /*
+	     * Make sure that unset traces are rune even if the execEnv is
+	     * rewinding (coroutine deletion, [Bug 2093947]
+	     */
+
+	    if (rewind && (flags & TCL_TRACE_UNSETS)) {
+		((Interp *)interp)->execEnvPtr->rewind = 0;
+	    }
+	    code = Tcl_EvalEx(interp, Tcl_DStringValue(&cmd),
+		    Tcl_DStringLength(&cmd), 0);
+	    if (rewind) {
+		((Interp *)interp)->execEnvPtr->rewind = rewind;
+	    }
+	    if (code != TCL_OK) {		/* copy error msg to result */
+		Tcl_Obj *errMsgObj = Tcl_GetObjResult(interp);
+
+		Tcl_IncrRefCount(errMsgObj);
+		result = (char *) errMsgObj;
+	    }
+	    Tcl_DStringFree(&cmd);
+	}
+    }
+    if (destroy && result != NULL) {
+	register Tcl_Obj *errMsgObj = (Tcl_Obj *) result;
+
+	Tcl_DecrRefCount(errMsgObj);
+	result = NULL;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateObjTrace --
+ *
+ *	Arrange for a function to be called to trace command execution.
+ *
+ * Results:
+ *	The return value is a token for the trace, which may be passed to
+ *	Tcl_DeleteTrace to eliminate the trace.
+ *
+ * Side effects:
+ *	From now on, proc will be called just before a command function is
+ *	called to execute a Tcl command. Calls to proc will have the following
+ *	form:
+ *
+ *	void proc(ClientData	 clientData,
+ *		  Tcl_Interp *	 interp,
+ *		  int		 level,
+ *		  const char *	 command,
+ *		  Tcl_Command	 commandInfo,
+ *		  int		 objc,
+ *		  Tcl_Obj *const objv[]);
+ *
+ *	The 'clientData' and 'interp' arguments to 'proc' will be the same as
+ *	the arguments to Tcl_CreateObjTrace. The 'level' argument gives the
+ *	nesting depth of command interpretation within the interpreter. The
+ *	'command' argument is the ASCII text of the command being evaluated -
+ *	before any substitutions are performed. The 'commandInfo' argument
+ *	gives a handle to the command procedure that will be evaluated. The
+ *	'objc' and 'objv' parameters give the parameter vector that will be
+ *	passed to the command procedure. Proc does not return a value.
+ *
+ *	It is permissible for 'proc' to call Tcl_SetCommandTokenInfo to change
+ *	the command procedure or client data for the command being evaluated,
+ *	and these changes will take effect with the current evaluation.
+ *
+ *	The 'level' argument specifies the maximum nesting level of calls to
+ *	be traced. If the execution depth of the interpreter exceeds 'level',
+ *	the trace callback is not executed.
+ *
+ *	The 'flags' argument is either zero or the value,
+ *	TCL_ALLOW_INLINE_COMPILATION. If the TCL_ALLOW_INLINE_COMPILATION flag
+ *	is not present, the bytecode compiler will not generate inline code
+ *	for Tcl's built-in commands. This behavior will have a significant
+ *	impact on performance, but will ensure that all command evaluations
+ *	are traced. If the TCL_ALLOW_INLINE_COMPILATION flag is present, the
+ *	bytecode compiler will have its normal behavior of compiling in-line
+ *	code for some of Tcl's built-in commands. In this case, the tracing
+ *	will be imprecise - in-line code will not be traced - but run-time
+ *	performance will be improved. The latter behavior is desired for many
+ *	applications such as profiling of run time.
+ *
+ *	When the trace is deleted, the 'delProc' function will be invoked,
+ *	passing it the original client data.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Trace
+Tcl_CreateObjTrace(
+    Tcl_Interp *interp,		/* Tcl interpreter */
+    int level,			/* Maximum nesting level */
+    int flags,			/* Flags, see above */
+    Tcl_CmdObjTraceProc *proc,	/* Trace callback */
+    ClientData clientData,	/* Client data for the callback */
+    Tcl_CmdObjTraceDeleteProc *delProc)
+				/* Function to call when trace is deleted */
+{
+    register Trace *tracePtr;
+    register Interp *iPtr = (Interp *) interp;
+
+    /*
+     * Test if this trace allows inline compilation of commands.
+     */
+
+    if (!(flags & TCL_ALLOW_INLINE_COMPILATION)) {
+	if (iPtr->tracesForbiddingInline == 0) {
+	    /*
+	     * When the first trace forbidding inline compilation is created,
+	     * invalidate existing compiled code for this interpreter and
+	     * arrange (by setting the DONT_COMPILE_CMDS_INLINE flag) that
+	     * when compiling new code, no commands will be compiled inline
+	     * (i.e., into an inline sequence of instructions). We do this
+	     * because commands that were compiled inline will never result in
+	     * a command trace being called.
+	     */
+
+	    iPtr->compileEpoch++;
+	    iPtr->flags |= DONT_COMPILE_CMDS_INLINE;
+	}
+	iPtr->tracesForbiddingInline++;
+    }
+
+    tracePtr = ckalloc(sizeof(Trace));
+    tracePtr->level = level;
+    tracePtr->proc = proc;
+    tracePtr->clientData = clientData;
+    tracePtr->delProc = delProc;
+    tracePtr->nextPtr = iPtr->tracePtr;
+    tracePtr->flags = flags;
+    iPtr->tracePtr = tracePtr;
+
+    return (Tcl_Trace) tracePtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_CreateTrace --
+ *
+ *	Arrange for a function to be called to trace command execution.
+ *
+ * Results:
+ *	The return value is a token for the trace, which may be passed to
+ *	Tcl_DeleteTrace to eliminate the trace.
+ *
+ * Side effects:
+ *	From now on, proc will be called just before a command procedure is
+ *	called to execute a Tcl command. Calls to proc will have the following
+ *	form:
+ *
+ *	void
+ *	proc(clientData, interp, level, command, cmdProc, cmdClientData,
+ *		argc, argv)
+ *	    ClientData clientData;
+ *	    Tcl_Interp *interp;
+ *	    int level;
+ *	    char *command;
+ *	    int (*cmdProc)();
+ *	    ClientData cmdClientData;
+ *	    int argc;
+ *	    char **argv;
+ *	{
+ *	}
+ *
+ *	The clientData and interp arguments to proc will be the same as the
+ *	corresponding arguments to this function. Level gives the nesting
+ *	level of command interpretation for this interpreter (0 corresponds to
+ *	top level). Command gives the ASCII text of the raw command, cmdProc
+ *	and cmdClientData give the function that will be called to process the
+ *	command and the ClientData value it will receive, and argc and argv
+ *	give the arguments to the command, after any argument parsing and
+ *	substitution. Proc does not return a value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Trace
+Tcl_CreateTrace(
+    Tcl_Interp *interp,		/* Interpreter in which to create trace. */
+    int level,			/* Only call proc for commands at nesting
+				 * level<=argument level (1=>top level). */
+    Tcl_CmdTraceProc *proc,	/* Function to call before executing each
+				 * command. */
+    ClientData clientData)	/* Arbitrary value word to pass to proc. */
+{
+    StringTraceData *data = ckalloc(sizeof(StringTraceData));
+
+    data->clientData = clientData;
+    data->proc = proc;
+    return Tcl_CreateObjTrace(interp, level, 0, StringTraceProc,
+	    data, StringTraceDeleteProc);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTraceProc --
+ *
+ *	Invoke a string-based trace function from an object-based callback.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Whatever the string-based trace function does.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+StringTraceProc(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int level,
+    const char *command,
+    Tcl_Command commandInfo,
+    int objc,
+    Tcl_Obj *const *objv)
+{
+    StringTraceData *data = clientData;
+    Command *cmdPtr = (Command *) commandInfo;
+    const char **argv;		/* Args to pass to string trace proc */
+    int i;
+
+    /*
+     * This is a bit messy because we have to emulate the old trace interface,
+     * which uses strings for everything.
+     */
+
+    argv = (const char **) TclStackAlloc(interp,
+	    (unsigned) ((objc + 1) * sizeof(const char *)));
+    for (i = 0; i < objc; i++) {
+	argv[i] = Tcl_GetString(objv[i]);
+    }
+    argv[objc] = 0;
+
+    /*
+     * Invoke the command function. Note that we cast away const-ness on two
+     * parameters for compatibility with legacy code; the code MUST NOT modify
+     * either command or argv.
+     */
+
+    data->proc(data->clientData, interp, level, (char *) command,
+	    cmdPtr->proc, cmdPtr->clientData, objc, argv);
+    TclStackFree(interp, (void *) argv);
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * StringTraceDeleteProc --
+ *
+ *	Clean up memory when a string-based trace is deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Allocated memory is returned to the system.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+StringTraceDeleteProc(
+    ClientData clientData)
+{
+    ckfree(clientData);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DeleteTrace --
+ *
+ *	Remove a trace.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	From now on there will be no more calls to the function given in
+ *	trace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DeleteTrace(
+    Tcl_Interp *interp,		/* Interpreter that contains trace. */
+    Tcl_Trace trace)		/* Token for trace (returned previously by
+				 * Tcl_CreateTrace). */
+{
+    Interp *iPtr = (Interp *) interp;
+    Trace *prevPtr, *tracePtr = (Trace *) trace;
+    register Trace **tracePtr2 = &iPtr->tracePtr;
+    ActiveInterpTrace *activePtr;
+
+    /*
+     * Locate the trace entry in the interpreter's trace list, and remove it
+     * from the list.
+     */
+
+    prevPtr = NULL;
+    while (*tracePtr2 != NULL && *tracePtr2 != tracePtr) {
+	prevPtr = *tracePtr2;
+	tracePtr2 = &prevPtr->nextPtr;
+    }
+    if (*tracePtr2 == NULL) {
+	return;
+    }
+    *tracePtr2 = (*tracePtr2)->nextPtr;
+
+    /*
+     * The code below makes it possible to delete traces while traces are
+     * active: it makes sure that the deleted trace won't be processed by
+     * TclCheckInterpTraces.
+     */
+
+    for (activePtr = iPtr->activeInterpTracePtr;  activePtr != NULL;
+	    activePtr = activePtr->nextPtr) {
+	if (activePtr->nextTracePtr == tracePtr) {
+	    if (activePtr->reverseScan) {
+		activePtr->nextTracePtr = prevPtr;
+	    } else {
+		activePtr->nextTracePtr = tracePtr->nextPtr;
+	    }
+	}
+    }
+
+    /*
+     * If the trace forbids bytecode compilation, change the interpreter's
+     * state. If bytecode compilation is now permitted, flag the fact and
+     * advance the compilation epoch so that procs will be recompiled to take
+     * advantage of it.
+     */
+
+    if (!(tracePtr->flags & TCL_ALLOW_INLINE_COMPILATION)) {
+	iPtr->tracesForbiddingInline--;
+	if (iPtr->tracesForbiddingInline == 0) {
+	    iPtr->flags &= ~DONT_COMPILE_CMDS_INLINE;
+	    iPtr->compileEpoch++;
+	}
+    }
+
+    /*
+     * Execute any delete callback.
+     */
+
+    if (tracePtr->delProc != NULL) {
+	tracePtr->delProc(tracePtr->clientData);
+    }
+
+    /*
+     * Delete the trace object.
+     */
+
+    Tcl_EventuallyFree((char *) tracePtr, TCL_DYNAMIC);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTraceVarExists --
+ *
+ *	This is called from info exists. We need to trigger read and/or array
+ *	traces because they may end up creating a variable that doesn't
+ *	currently exist.
+ *
+ * Results:
+ *	A pointer to the Var structure, or NULL.
+ *
+ * Side effects:
+ *	May fill in error messages in the interp.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Var *
+TclVarTraceExists(
+    Tcl_Interp *interp,		/* The interpreter */
+    const char *varName)	/* The variable name */
+{
+    Var *varPtr, *arrayPtr;
+
+    /*
+     * The choice of "create" flag values is delicate here, and matches the
+     * semantics of GetVar. Things are still not perfect, however, because if
+     * you do "info exists x" you get a varPtr and therefore trigger traces.
+     * However, if you do "info exists x(i)", then you only get a varPtr if x
+     * is already known to be an array. Otherwise you get NULL, and no trace
+     * is triggered. This matches Tcl 7.6 semantics.
+     */
+
+    varPtr = TclLookupVar(interp, varName, NULL, 0, "access",
+	    /*createPart1*/ 0, /*createPart2*/ 1, &arrayPtr);
+
+    if (varPtr == NULL) {
+	return NULL;
+    }
+
+    if ((varPtr->flags & VAR_TRACED_READ)
+	    || (arrayPtr && (arrayPtr->flags & VAR_TRACED_READ))) {
+	TclCallVarTraces((Interp *) interp, arrayPtr, varPtr, varName, NULL,
+		TCL_TRACE_READS, /* leaveErrMsg */ 0);
+    }
+
+    /*
+     * If the variable doesn't exist anymore and no-one's using it, then free
+     * up the relevant structures and hash table entries.
+     */
+
+    if (TclIsVarUndefined(varPtr)) {
+	TclCleanupVar(varPtr, arrayPtr);
+	return NULL;
+    }
+
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCallVarTraces --
+ *
+ *	This function is invoked to find and invoke relevant trace functions
+ *	associated with a particular operation on a variable. This function
+ *	invokes traces both on the variable and on its containing array (where
+ *	relevant).
+ *
+ * Results:
+ *	Returns TCL_OK to indicate normal operation. Returns TCL_ERROR if
+ *	invocation of a trace function indicated an error. When TCL_ERROR is
+ *	returned and leaveErrMsg is true, then the errorInfo field of iPtr has
+ *	information about the error placed in it.
+ *
+ * Side effects:
+ *	Almost anything can happen, depending on trace; this function itself
+ *	doesn't have any side effects.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjCallVarTraces(
+    Interp *iPtr,		/* Interpreter containing variable. */
+    register Var *arrayPtr,	/* Pointer to array variable that contains the
+				 * variable, or NULL if the variable isn't an
+				 * element of an array. */
+    Var *varPtr,		/* Variable whose traces are to be invoked. */
+    Tcl_Obj *part1Ptr,
+    Tcl_Obj *part2Ptr,		/* Variable's two-part name. */
+    int flags,			/* Flags passed to trace functions: indicates
+				 * what's happening to variable, plus maybe
+				 * TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY */
+    int leaveErrMsg,		/* If true, and one of the traces indicates an
+				 * error, then leave an error message and
+				 * stack trace information in *iPTr. */
+    int index)			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+{
+    const char *part1, *part2;
+
+    if (!part1Ptr) {
+	part1Ptr = localName(iPtr->varFramePtr, index);
+    }
+    if (!part1Ptr) {
+	Tcl_Panic("Cannot trace a variable with no name");
+    }
+    part1 = TclGetString(part1Ptr);
+    part2 = part2Ptr? TclGetString(part2Ptr) : NULL;
+
+    return TclCallVarTraces(iPtr, arrayPtr, varPtr, part1, part2, flags,
+	    leaveErrMsg);
+}
+
+int
+TclCallVarTraces(
+    Interp *iPtr,		/* Interpreter containing variable. */
+    register Var *arrayPtr,	/* Pointer to array variable that contains the
+				 * variable, or NULL if the variable isn't an
+				 * element of an array. */
+    Var *varPtr,		/* Variable whose traces are to be invoked. */
+    const char *part1,
+    const char *part2,		/* Variable's two-part name. */
+    int flags,			/* Flags passed to trace functions: indicates
+				 * what's happening to variable, plus maybe
+				 * TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY */
+    int leaveErrMsg)		/* If true, and one of the traces indicates an
+				 * error, then leave an error message and
+				 * stack trace information in *iPTr. */
+{
+    register VarTrace *tracePtr;
+    ActiveVarTrace active;
+    char *result;
+    const char *openParen, *p;
+    Tcl_DString nameCopy;
+    int copiedName;
+    int code = TCL_OK;
+    int disposeFlags = 0;
+    Tcl_InterpState state = NULL;
+    Tcl_HashEntry *hPtr;
+    int traceflags = flags & VAR_ALL_TRACES;
+
+    /*
+     * If there are already similar trace functions active for the variable,
+     * don't call them again.
+     */
+
+    if (TclIsVarTraceActive(varPtr)) {
+	return code;
+    }
+    TclSetVarTraceActive(varPtr);
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)++;
+    }
+    if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	VarHashRefCount(arrayPtr)++;
+    }
+
+    /*
+     * If the variable name hasn't been parsed into array name and element, do
+     * it here. If there really is an array element, make a copy of the
+     * original name so that NULLs can be inserted into it to separate the
+     * names (can't modify the name string in place, because the string might
+     * get used by the callbacks we invoke).
+     */
+
+    copiedName = 0;
+    if (part2 == NULL) {
+	for (p = part1; *p ; p++) {
+	    if (*p == '(') {
+		openParen = p;
+		do {
+		    p++;
+		} while (*p != '\0');
+		p--;
+		if (*p == ')') {
+		    int offset = (openParen - part1);
+		    char *newPart1;
+
+		    Tcl_DStringInit(&nameCopy);
+		    Tcl_DStringAppend(&nameCopy, part1, p-part1);
+		    newPart1 = Tcl_DStringValue(&nameCopy);
+		    newPart1[offset] = 0;
+		    part1 = newPart1;
+		    part2 = newPart1 + offset + 1;
+		    copiedName = 1;
+		}
+		break;
+	    }
+	}
+    }
+
+    /*
+     * Ignore any caller-provided TCL_INTERP_DESTROYED flag.  Only we can
+     * set it correctly.
+     */
+
+    flags &= ~TCL_INTERP_DESTROYED;
+
+    /*
+     * Invoke traces on the array containing the variable, if relevant.
+     */
+
+    result = NULL;
+    active.nextPtr = iPtr->activeVarTracePtr;
+    iPtr->activeVarTracePtr = &active;
+    Tcl_Preserve(iPtr);
+    if (arrayPtr && !TclIsVarTraceActive(arrayPtr)
+	    && (arrayPtr->flags & traceflags)) {
+	hPtr = Tcl_FindHashEntry(&iPtr->varTraces, (char *) arrayPtr);
+	active.varPtr = arrayPtr;
+	for (tracePtr = Tcl_GetHashValue(hPtr);
+		tracePtr != NULL; tracePtr = active.nextTracePtr) {
+	    active.nextTracePtr = tracePtr->nextPtr;
+	    if (!(tracePtr->flags & flags)) {
+		continue;
+	    }
+	    Tcl_Preserve(tracePtr);
+	    if (state == NULL) {
+		state = Tcl_SaveInterpState((Tcl_Interp *) iPtr, code);
+	    }
+	    if (Tcl_InterpDeleted((Tcl_Interp *) iPtr)) {
+		flags |= TCL_INTERP_DESTROYED;
+	    }
+	    result = tracePtr->traceProc(tracePtr->clientData,
+		    (Tcl_Interp *) iPtr, part1, part2, flags);
+	    if (result != NULL) {
+		if (flags & TCL_TRACE_UNSETS) {
+		    /*
+		     * Ignore errors in unset traces.
+		     */
+
+		    DisposeTraceResult(tracePtr->flags, result);
+		} else {
+		    disposeFlags = tracePtr->flags;
+		    code = TCL_ERROR;
+		}
+	    }
+	    Tcl_Release(tracePtr);
+	    if (code == TCL_ERROR) {
+		goto done;
+	    }
+	}
+    }
+
+    /*
+     * Invoke traces on the variable itself.
+     */
+
+    if (flags & TCL_TRACE_UNSETS) {
+	flags |= TCL_TRACE_DESTROYED;
+    }
+    active.varPtr = varPtr;
+    if (varPtr->flags & traceflags) {
+	hPtr = Tcl_FindHashEntry(&iPtr->varTraces, (char *) varPtr);
+	for (tracePtr = Tcl_GetHashValue(hPtr);
+		tracePtr != NULL; tracePtr = active.nextTracePtr) {
+	    active.nextTracePtr = tracePtr->nextPtr;
+	    if (!(tracePtr->flags & flags)) {
+		continue;
+	    }
+	    Tcl_Preserve(tracePtr);
+	    if (state == NULL) {
+		state = Tcl_SaveInterpState((Tcl_Interp *) iPtr, code);
+	    }
+	    if (Tcl_InterpDeleted((Tcl_Interp *) iPtr)) {
+		flags |= TCL_INTERP_DESTROYED;
+	    }
+	    result = tracePtr->traceProc(tracePtr->clientData,
+		    (Tcl_Interp *) iPtr, part1, part2, flags);
+	    if (result != NULL) {
+		if (flags & TCL_TRACE_UNSETS) {
+		    /*
+		     * Ignore errors in unset traces.
+		     */
+
+		    DisposeTraceResult(tracePtr->flags, result);
+		} else {
+		    disposeFlags = tracePtr->flags;
+		    code = TCL_ERROR;
+		}
+	    }
+	    Tcl_Release(tracePtr);
+	    if (code == TCL_ERROR) {
+		goto done;
+	    }
+	}
+    }
+
+    /*
+     * Restore the variable's flags, remove the record of our active traces,
+     * and then return.
+     */
+
+  done:
+    if (code == TCL_ERROR) {
+	if (leaveErrMsg) {
+	    const char *verb = "";
+	    const char *type = "";
+
+	    switch (flags&(TCL_TRACE_READS|TCL_TRACE_WRITES|TCL_TRACE_ARRAY)) {
+	    case TCL_TRACE_READS:
+		verb = "read";
+		type = verb;
+		break;
+	    case TCL_TRACE_WRITES:
+		verb = "set";
+		type = "write";
+		break;
+	    case TCL_TRACE_ARRAY:
+		verb = "trace array";
+		type = "array";
+		break;
+	    }
+
+	    if (disposeFlags & TCL_TRACE_RESULT_OBJECT) {
+		Tcl_SetObjResult((Tcl_Interp *)iPtr, (Tcl_Obj *) result);
+	    } else {
+		Tcl_SetObjResult((Tcl_Interp *)iPtr,
+			Tcl_NewStringObj(result, -1));
+	    }
+	    Tcl_AddErrorInfo((Tcl_Interp *)iPtr, "");
+
+	    Tcl_AppendObjToErrorInfo((Tcl_Interp *)iPtr, Tcl_ObjPrintf(
+		    "\n    (%s trace on \"%s%s%s%s\")", type, part1,
+		    (part2 ? "(" : ""), (part2 ? part2 : ""),
+		    (part2 ? ")" : "") ));
+	    if (disposeFlags & TCL_TRACE_RESULT_OBJECT) {
+		TclVarErrMsg((Tcl_Interp *) iPtr, part1, part2, verb,
+			Tcl_GetString((Tcl_Obj *) result));
+	    } else {
+		TclVarErrMsg((Tcl_Interp *) iPtr, part1, part2, verb, result);
+	    }
+	    iPtr->flags &= ~(ERR_ALREADY_LOGGED);
+	    Tcl_DiscardInterpState(state);
+	} else {
+	    Tcl_RestoreInterpState((Tcl_Interp *) iPtr, state);
+	}
+	DisposeTraceResult(disposeFlags,result);
+    } else if (state) {
+	if (code == TCL_OK) {
+	    code = Tcl_RestoreInterpState((Tcl_Interp *) iPtr, state);
+	} else {
+	    Tcl_DiscardInterpState(state);
+	}
+    }
+
+    if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	VarHashRefCount(arrayPtr)--;
+    }
+    if (copiedName) {
+	Tcl_DStringFree(&nameCopy);
+    }
+    TclClearVarTraceActive(varPtr);
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)--;
+    }
+    iPtr->activeVarTracePtr = active.nextPtr;
+    Tcl_Release(iPtr);
+    return code;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DisposeTraceResult--
+ *
+ *	This function is called to dispose of the result returned from a trace
+ *	function. The disposal method appropriate to the type of result is
+ *	determined by flags.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The memory allocated for the trace result may be freed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DisposeTraceResult(
+    int flags,			/* Indicates type of result to determine
+				 * proper disposal method. */
+    char *result)		/* The result returned from a trace function
+				 * to be disposed. */
+{
+    if (flags & TCL_TRACE_RESULT_DYNAMIC) {
+	ckfree(result);
+    } else if (flags & TCL_TRACE_RESULT_OBJECT) {
+	Tcl_DecrRefCount((Tcl_Obj *) result);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UntraceVar --
+ *
+ *	Remove a previously-created trace for a variable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there exists a trace for the variable given by varName with the
+ *	given flags, proc, and clientData, then that trace is removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_UntraceVar
+void
+Tcl_UntraceVar(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *varName,	/* Name of variable; may end with "(index)" to
+				 * signify an array reference. */
+    int flags,			/* OR-ed collection of bits describing current
+				 * trace, including any of TCL_TRACE_READS,
+				 * TCL_TRACE_WRITES, TCL_TRACE_UNSETS,
+				 * TCL_GLOBAL_ONLY and TCL_NAMESPACE_ONLY. */
+    Tcl_VarTraceProc *proc,	/* Function assocated with trace. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    Tcl_UntraceVar2(interp, varName, NULL, flags, proc, clientData);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UntraceVar2 --
+ *
+ *	Remove a previously-created trace for a variable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If there exists a trace for the variable given by part1 and part2 with
+ *	the given flags, proc, and clientData, then that trace is removed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_UntraceVar2(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *part1,		/* Name of variable or array. */
+    const char *part2,		/* Name of element within array; NULL means
+				 * trace applies to scalar variable or array
+				 * as-a-whole. */
+    int flags,			/* OR-ed collection of bits describing current
+				 * trace, including any of TCL_TRACE_READS,
+				 * TCL_TRACE_WRITES, TCL_TRACE_UNSETS,
+				 * TCL_GLOBAL_ONLY, and TCL_NAMESPACE_ONLY. */
+    Tcl_VarTraceProc *proc,	/* Function assocated with trace. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    register VarTrace *tracePtr;
+    VarTrace *prevPtr, *nextPtr;
+    Var *varPtr, *arrayPtr;
+    Interp *iPtr = (Interp *) interp;
+    ActiveVarTrace *activePtr;
+    int flagMask, allFlags = 0;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * Set up a mask to mask out the parts of the flags that we are not
+     * interested in now.
+     */
+
+    flagMask = TCL_GLOBAL_ONLY | TCL_NAMESPACE_ONLY;
+    varPtr = TclLookupVar(interp, part1, part2, flags & flagMask, /*msg*/ NULL,
+	    /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+    if (varPtr == NULL || !(varPtr->flags & VAR_ALL_TRACES & flags)) {
+	return;
+    }
+
+    /*
+     * Set up a mask to mask out the parts of the flags that we are not
+     * interested in now.
+     */
+
+    flagMask = TCL_TRACE_READS | TCL_TRACE_WRITES | TCL_TRACE_UNSETS |
+	  TCL_TRACE_ARRAY | TCL_TRACE_RESULT_DYNAMIC | TCL_TRACE_RESULT_OBJECT;
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+    flagMask |= TCL_TRACE_OLD_STYLE;
+#endif
+    flags &= flagMask;
+
+    hPtr = Tcl_FindHashEntry(&iPtr->varTraces, (char *) varPtr);
+    for (tracePtr = Tcl_GetHashValue(hPtr), prevPtr = NULL; ;
+	    prevPtr = tracePtr, tracePtr = tracePtr->nextPtr) {
+	if (tracePtr == NULL) {
+	    goto updateFlags;
+	}
+	if ((tracePtr->traceProc == proc) && (tracePtr->flags == flags)
+		&& (tracePtr->clientData == clientData)) {
+	    break;
+	}
+	allFlags |= tracePtr->flags;
+    }
+
+    /*
+     * The code below makes it possible to delete traces while traces are
+     * active: it makes sure that the deleted trace won't be processed by
+     * TclCallVarTraces.
+     *
+     * Caveat (Bug 3062331): When an unset trace handler on a variable
+     * tries to delete a different unset trace handler on the same variable,
+     * the results may be surprising.  When variable unset traces fire, the
+     * traced variable is already gone.  So the TclLookupVar() call above
+     * will not find that variable, and not finding it will never reach here
+     * to perform the deletion.  This means callers of Tcl_UntraceVar*()
+     * attempting to delete unset traces from within the handler of another
+     * unset trace have to account for the possibility that their call to
+     * Tcl_UntraceVar*() is a no-op.
+     */
+
+    for (activePtr = iPtr->activeVarTracePtr;  activePtr != NULL;
+	    activePtr = activePtr->nextPtr) {
+	if (activePtr->nextTracePtr == tracePtr) {
+	    activePtr->nextTracePtr = tracePtr->nextPtr;
+	}
+    }
+    nextPtr = tracePtr->nextPtr;
+    if (prevPtr == NULL) {
+	if (nextPtr) {
+	    Tcl_SetHashValue(hPtr, nextPtr);
+	} else {
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+    } else {
+	prevPtr->nextPtr = nextPtr;
+    }
+    tracePtr->nextPtr = NULL;
+    Tcl_EventuallyFree(tracePtr, TCL_DYNAMIC);
+
+    for (tracePtr = nextPtr; tracePtr != NULL;
+	    tracePtr = tracePtr->nextPtr) {
+	allFlags |= tracePtr->flags;
+    }
+
+  updateFlags:
+    varPtr->flags &= ~VAR_ALL_TRACES;
+    if (allFlags & VAR_ALL_TRACES) {
+	varPtr->flags |= (allFlags & VAR_ALL_TRACES);
+    } else if (TclIsVarUndefined(varPtr)) {
+	/*
+	 * If this is the last trace on the variable, and the variable is
+	 * unset and unused, then free up the variable.
+	 */
+
+	TclCleanupVar(varPtr, NULL);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_VarTraceInfo --
+ *
+ *	Return the clientData value associated with a trace on a variable.
+ *	This function can also be used to step through all of the traces on a
+ *	particular variable that have the same trace function.
+ *
+ * Results:
+ *	The return value is the clientData value associated with a trace on
+ *	the given variable. Information will only be returned for a trace with
+ *	proc as trace function. If the clientData argument is NULL then the
+ *	first such trace is returned; otherwise, the next relevant one after
+ *	the one given by clientData will be returned. If the variable doesn't
+ *	exist, or if there are no (more) traces for it, then NULL is returned.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_VarTraceInfo
+ClientData
+Tcl_VarTraceInfo(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *varName,	/* Name of variable; may end with "(index)" to
+				 * signify an array reference. */
+    int flags,			/* OR-ed combo or TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY (can be 0). */
+    Tcl_VarTraceProc *proc,	/* Function assocated with trace. */
+    ClientData prevClientData)	/* If non-NULL, gives last value returned by
+				 * this function, so this call will return the
+				 * next trace after that one. If NULL, this
+				 * call will return the first trace. */
+{
+    return Tcl_VarTraceInfo2(interp, varName, NULL, flags, proc,
+	    prevClientData);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_VarTraceInfo2 --
+ *
+ *	Same as Tcl_VarTraceInfo, except takes name in two pieces instead of
+ *	one.
+ *
+ * Results:
+ *	Same as Tcl_VarTraceInfo.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+ClientData
+Tcl_VarTraceInfo2(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *part1,		/* Name of variable or array. */
+    const char *part2,		/* Name of element within array; NULL means
+				 * trace applies to scalar variable or array
+				 * as-a-whole. */
+    int flags,			/* OR-ed combination of TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY. */
+    Tcl_VarTraceProc *proc,	/* Function assocated with trace. */
+    ClientData prevClientData)	/* If non-NULL, gives last value returned by
+				 * this function, so this call will return the
+				 * next trace after that one. If NULL, this
+				 * call will return the first trace. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+    Tcl_HashEntry *hPtr;
+
+    varPtr = TclLookupVar(interp, part1, part2,
+	    flags & (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY), /*msg*/ NULL,
+	    /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+    if (varPtr == NULL) {
+	return NULL;
+    }
+
+    /*
+     * Find the relevant trace, if any, and return its clientData.
+     */
+
+    hPtr = Tcl_FindHashEntry(&iPtr->varTraces, (char *) varPtr);
+
+    if (hPtr) {
+	register VarTrace *tracePtr = Tcl_GetHashValue(hPtr);
+
+	if (prevClientData != NULL) {
+	    for (; tracePtr != NULL; tracePtr = tracePtr->nextPtr) {
+		if ((tracePtr->clientData == prevClientData)
+			&& (tracePtr->traceProc == proc)) {
+		    tracePtr = tracePtr->nextPtr;
+		    break;
+		}
+	    }
+	}
+	for (; tracePtr != NULL ; tracePtr = tracePtr->nextPtr) {
+	    if (tracePtr->traceProc == proc) {
+		return tracePtr->clientData;
+	    }
+	}
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TraceVar --
+ *
+ *	Arrange for reads and/or writes to a variable to cause a function to
+ *	be invoked, which can monitor the operations and/or change their
+ *	actions.
+ *
+ * Results:
+ *	A standard Tcl return value.
+ *
+ * Side effects:
+ *	A trace is set up on the variable given by varName, such that future
+ *	references to the variable will be intermediated by proc. See the
+ *	manual entry for complete details on the calling sequence for proc.
+ *     The variable's flags are updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_TraceVar
+int
+Tcl_TraceVar(
+    Tcl_Interp *interp,		/* Interpreter in which variable is to be
+				 * traced. */
+    const char *varName,	/* Name of variable; may end with "(index)" to
+				 * signify an array reference. */
+    int flags,			/* OR-ed collection of bits, including any of
+				 * TCL_TRACE_READS, TCL_TRACE_WRITES,
+				 * TCL_TRACE_UNSETS, TCL_GLOBAL_ONLY, and
+				 * TCL_NAMESPACE_ONLY. */
+    Tcl_VarTraceProc *proc,	/* Function to call when specified ops are
+				 * invoked upon varName. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    return Tcl_TraceVar2(interp, varName, NULL, flags, proc, clientData);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_TraceVar2 --
+ *
+ *	Arrange for reads and/or writes to a variable to cause a function to
+ *	be invoked, which can monitor the operations and/or change their
+ *	actions.
+ *
+ * Results:
+ *	A standard Tcl return value.
+ *
+ * Side effects:
+ *	A trace is set up on the variable given by part1 and part2, such that
+ *	future references to the variable will be intermediated by proc. See
+ *	the manual entry for complete details on the calling sequence for
+ *	proc. The variable's flags are updated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_TraceVar2(
+    Tcl_Interp *interp,		/* Interpreter in which variable is to be
+				 * traced. */
+    const char *part1,		/* Name of scalar variable or array. */
+    const char *part2,		/* Name of element within array; NULL means
+				 * trace applies to scalar variable or array
+				 * as-a-whole. */
+    int flags,			/* OR-ed collection of bits, including any of
+				 * TCL_TRACE_READS, TCL_TRACE_WRITES,
+				 * TCL_TRACE_UNSETS, TCL_GLOBAL_ONLY, and
+				 * TCL_NAMESPACE_ONLY. */
+    Tcl_VarTraceProc *proc,	/* Function to call when specified ops are
+				 * invoked upon varName. */
+    ClientData clientData)	/* Arbitrary argument to pass to proc. */
+{
+    register VarTrace *tracePtr;
+    int result;
+
+    tracePtr = ckalloc(sizeof(VarTrace));
+    tracePtr->traceProc = proc;
+    tracePtr->clientData = clientData;
+    tracePtr->flags = flags;
+
+    result = TraceVarEx(interp, part1, part2, tracePtr);
+
+    if (result != TCL_OK) {
+	ckfree(tracePtr);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TraceVarEx --
+ *
+ *	Arrange for reads and/or writes to a variable to cause a function to
+ *	be invoked, which can monitor the operations and/or change their
+ *	actions.
+ *
+ * Results:
+ *	A standard Tcl return value.
+ *
+ * Side effects:
+ *	A trace is set up on the variable given by part1 and part2, such that
+ *	future references to the variable will be intermediated by the
+ *	traceProc listed in tracePtr. See the manual entry for complete
+ *	details on the calling sequence for proc.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+TraceVarEx(
+    Tcl_Interp *interp,		/* Interpreter in which variable is to be
+				 * traced. */
+    const char *part1,		/* Name of scalar variable or array. */
+    const char *part2,		/* Name of element within array; NULL means
+				 * trace applies to scalar variable or array
+				 * as-a-whole. */
+    register VarTrace *tracePtr)/* Structure containing flags, traceProc and
+				 * clientData fields. Others should be left
+				 * blank. Will be ckfree()d (eventually) if
+				 * this function returns TCL_OK, and up to
+				 * caller to free if this function returns
+				 * TCL_ERROR. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+    int flagMask, isNew;
+    Tcl_HashEntry *hPtr;
+
+    /*
+     * We strip 'flags' down to just the parts which are relevant to
+     * TclLookupVar, to avoid conflicts between trace flags and internal
+     * namespace flags such as 'TCL_FIND_ONLY_NS'. This can now occur since we
+     * have trace flags with values 0x1000 and higher.
+     */
+
+    flagMask = TCL_GLOBAL_ONLY | TCL_NAMESPACE_ONLY;
+    varPtr = TclLookupVar(interp, part1, part2,
+	    (tracePtr->flags & flagMask) | TCL_LEAVE_ERR_MSG,
+	    "trace", /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check for a nonsense flag combination. Note that this is a Tcl_Panic()
+     * because there should be no code path that ever sets both flags.
+     */
+
+    if ((tracePtr->flags & TCL_TRACE_RESULT_DYNAMIC)
+	    && (tracePtr->flags & TCL_TRACE_RESULT_OBJECT)) {
+	Tcl_Panic("bad result flag combination");
+    }
+
+    /*
+     * Set up trace information.
+     */
+
+    flagMask = TCL_TRACE_READS | TCL_TRACE_WRITES | TCL_TRACE_UNSETS |
+	  TCL_TRACE_ARRAY | TCL_TRACE_RESULT_DYNAMIC | TCL_TRACE_RESULT_OBJECT;
+#ifndef TCL_REMOVE_OBSOLETE_TRACES
+    flagMask |= TCL_TRACE_OLD_STYLE;
+#endif
+    tracePtr->flags = tracePtr->flags & flagMask;
+
+    hPtr = Tcl_CreateHashEntry(&iPtr->varTraces, varPtr, &isNew);
+    if (isNew) {
+	tracePtr->nextPtr = NULL;
+    } else {
+	tracePtr->nextPtr = Tcl_GetHashValue(hPtr);
+    }
+    Tcl_SetHashValue(hPtr, tracePtr);
+
+    /*
+     * Mark the variable as traced so we know to call them.
+     */
+
+    varPtr->flags |= (tracePtr->flags & VAR_ALL_TRACES);
+
+    return TCL_OK;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclUniData.c b/generic/tclUniData.c
new file mode 100644
index 0000000..9f05230
--- /dev/null
+++ b/generic/tclUniData.c
@@ -0,0 +1,1632 @@
+/*
+ * tclUniData.c --
+ *
+ *	Declarations of Unicode character information tables.  This file is
+ *	automatically generated by the tools/uniParse.tcl script.  Do not
+ *	modify this file by hand.
+ *
+ * Copyright (c) 1998 by Scriptics Corporation.
+ * All rights reserved.
+ */
+
+/*
+ * A 16-bit Unicode character is split into two parts in order to index
+ * into the following tables.  The lower OFFSET_BITS comprise an offset
+ * into a page of characters.  The upper bits comprise the page number.
+ */
+
+#define OFFSET_BITS 5
+
+/*
+ * The pageMap is indexed by page number and returns an alternate page number
+ * that identifies a unique page of characters.  Many Unicode characters map
+ * to the same alternate page number.
+ */
+
+static const unsigned short pageMap[] = {
+    0, 32, 64, 96, 0, 128, 160, 192, 224, 256, 288, 320, 352, 384, 416,
+    448, 224, 480, 512, 544, 576, 608, 640, 672, 704, 704, 736, 768, 800,
+    832, 864, 896, 928, 960, 992, 224, 1024, 224, 1056, 224, 224, 1088,
+    1120, 1152, 1184, 1216, 1248, 1280, 1312, 1344, 1376, 1408, 1344, 1344,
+    1440, 1472, 1504, 1536, 1568, 1344, 1344, 1600, 1632, 1664, 1696, 1728,
+    1760, 1792, 1824, 1856, 1888, 1920, 1952, 1984, 2016, 2048, 2080, 2112,
+    2144, 2176, 2208, 2240, 2272, 2304, 2336, 2368, 2400, 2432, 2464, 2496,
+    2528, 2560, 2592, 2624, 2656, 2688, 2720, 2752, 2784, 2816, 2848, 2880,
+    2912, 2944, 2976, 3008, 3040, 3072, 3104, 3136, 3168, 3200, 3232, 3264,
+    3296, 1824, 3328, 3360, 3392, 1824, 3424, 3456, 3488, 3520, 3552, 3584,
+    3616, 1824, 1344, 3648, 3680, 3712, 3744, 3776, 3808, 3840, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 3872, 1344, 3904, 3936,
+    3968, 1344, 4000, 1344, 4032, 4064, 4096, 4128, 4128, 4160, 4192, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 4224, 4256, 1344, 1344, 4288, 4320, 4352,
+    4384, 4416, 1344, 4448, 4480, 4512, 4544, 1344, 4576, 4608, 4640, 4672,
+    1344, 4704, 4736, 4768, 4800, 4832, 1344, 4864, 4896, 4928, 4960, 1344,
+    4992, 5024, 5056, 5088, 1824, 1824, 5120, 5152, 5184, 5216, 5248, 5280,
+    1344, 5312, 1344, 5344, 5376, 5408, 5440, 1824, 5472, 5504, 5536, 5568,
+    5600, 5632, 5664, 5600, 704, 5696, 224, 224, 224, 224, 5728, 224, 224,
+    224, 5760, 5792, 5824, 5856, 5888, 5920, 5952, 5984, 6016, 6048, 6080,
+    6112, 6144, 6176, 6208, 6240, 6272, 6304, 6336, 6368, 6400, 6432, 6464,
+    6496, 6528, 6528, 6528, 6528, 6528, 6528, 6528, 6528, 6560, 6592, 4928,
+    6624, 6656, 6688, 6720, 6752, 4928, 6784, 6816, 6848, 6880, 6912, 6944,
+    6976, 4928, 4928, 4928, 4928, 4928, 7008, 7040, 7072, 4928, 4928, 4928,
+    7104, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 7136, 7168, 4928, 7200,
+    7232, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 6528, 6528, 6528,
+    6528, 7264, 6528, 7296, 7328, 6528, 6528, 6528, 6528, 6528, 6528, 6528,
+    6528, 4928, 7360, 7392, 7424, 7456, 7488, 7520, 7552, 7584, 7616, 7648,
+    7680, 224, 224, 224, 7712, 7744, 7776, 1344, 7808, 7840, 7872, 7872,
+    704, 7904, 7936, 7968, 1824, 8000, 4928, 4928, 8032, 4928, 4928, 4928,
+    4928, 4928, 4928, 8064, 8096, 8128, 8160, 3232, 1344, 8192, 4192, 1344,
+    8224, 8256, 8288, 1344, 1344, 8320, 8352, 4928, 8384, 8416, 8448, 8480,
+    4928, 8448, 8512, 4928, 8416, 4928, 4928, 4928, 4928, 4928, 4928, 4928,
+    4928, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 4704, 4928, 4928, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1792, 8544, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 8576, 4928, 8608, 5408, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 8640, 8672, 224, 8704, 8736, 1344, 1344, 8768, 8800, 8832, 224,
+    8864, 8896, 8928, 1824, 8960, 8992, 9024, 1344, 9056, 9088, 9120, 9152,
+    9184, 1632, 9216, 9248, 9280, 1952, 9312, 9344, 9376, 1344, 9408, 9440,
+    9472, 1344, 9504, 9536, 9568, 9600, 9632, 9664, 9696, 9728, 9728, 1344,
+    9760, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 9792, 9824, 9856, 9888, 9888, 9888, 9888, 9888, 9888, 9888,
+    9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888,
+    9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888,
+    9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888,
+    9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888,
+    9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9888, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920, 9920,
+    9920, 9920, 9920, 9920, 9920, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 9952, 1344, 1344, 9984, 1824, 10016, 10048,
+    10080, 1344, 1344, 10112, 10144, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 10176, 10208, 1344, 10240, 1344, 10272, 10304,
+    10336, 10368, 10400, 10432, 1344, 1344, 1344, 10464, 10496, 64, 10528,
+    10560, 10592, 4736, 10624, 10656
+#if TCL_UTF_MAX > 3
+    ,10688, 10720, 10752, 1824, 1344, 1344, 1344, 8352, 10784, 10816, 10848,
+    10880, 10912, 10944, 10976, 11008, 1824, 1824, 1824, 1824, 9280, 1344,
+    11040, 11072, 1344, 11104, 11136, 11168, 11200, 1344, 11232, 1824,
+    11264, 11296, 11328, 1344, 11360, 11392, 11424, 11456, 1344, 11488,
+    1344, 11520, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 7840, 4704, 10272, 1824, 1824, 1824, 1824,
+    11552, 11584, 11616, 11648, 4736, 11680, 1824, 11712, 11744, 11776,
+    1824, 1824, 1344, 11808, 11840, 6848, 11872, 11904, 11936, 11968, 12000,
+    1824, 12032, 12064, 1344, 12096, 12128, 12160, 12192, 12224, 1824,
+    1824, 1344, 1344, 12256, 1824, 12288, 12320, 12352, 12384, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 12416, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 12448,
+    12480, 12512, 12544, 5248, 12576, 12608, 12640, 12672, 12704, 12736,
+    12768, 5248, 12800, 12832, 12864, 12896, 12928, 1824, 1824, 12960,
+    12992, 13024, 13056, 13088, 2368, 13120, 13152, 1824, 1824, 1824, 1824,
+    1344, 13184, 13216, 1824, 1344, 13248, 13280, 1824, 1824, 1824, 1824,
+    1824, 1344, 13312, 13344, 1824, 1344, 13376, 13408, 13440, 1344, 13472,
+    13504, 1824, 13536, 13568, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 13600, 13632, 13664, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 13696, 13728, 13760, 1344, 13792, 13824, 1344,
+    13856, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 13888, 13920,
+    13952, 13984, 14016, 14048, 1824, 1824, 14080, 14112, 14144, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    9984, 1824, 1824, 1824, 10848, 10848, 10848, 14176, 1344, 1344, 1344,
+    1344, 1344, 1344, 14208, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 14240, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 14272, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 13856, 4736, 14304, 1824, 1824, 10208,
+    14336, 1344, 14368, 14400, 14432, 14464, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1344, 1344,
+    14496, 14528, 14560, 1824, 1824, 14592, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 14624, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 14656, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 4736, 1824, 1824, 10208, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 9856, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 14688, 14720,
+    14752, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    4928, 4928, 4928, 4928, 4928, 4928, 4928, 8064, 4928, 14784, 4928,
+    14816, 14848, 14880, 4928, 14912, 4928, 4928, 14944, 1824, 1824, 1824,
+    1824, 1824, 4928, 4928, 14976, 15008, 1824, 1824, 1824, 1824, 15040,
+    15072, 15104, 15136, 15168, 15200, 15232, 15264, 15296, 15328, 15360,
+    15392, 15424, 15040, 15072, 15456, 15136, 15488, 15520, 15552, 15264,
+    15584, 15616, 15648, 15680, 15712, 15744, 15776, 15808, 15840, 15872,
+    15904, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928,
+    4928, 4928, 4928, 4928, 4928, 4928, 704, 15936, 704, 15968, 16000,
+    16032, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 16064, 16096, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1344, 1344, 1344, 1344, 1344, 1344, 16128, 1824, 16160, 16192,
+    16224, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 16256, 16288, 16320, 16352, 16384, 16416, 1824, 16448,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 4928, 16480, 4928,
+    4928, 8032, 16512, 16544, 8064, 16576, 16608, 4928, 16480, 4928, 16640,
+    1824, 16672, 16704, 16736, 16768, 16800, 1824, 1824, 1824, 1824, 4928,
+    4928, 4928, 4928, 4928, 4928, 4928, 16832, 4928, 4928, 4928, 4928,
+    4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928, 4928,
+    4928, 4928, 4928, 4928, 4928, 4928, 16864, 16896, 4928, 4928, 4928,
+    8032, 4928, 4928, 16864, 1824, 16480, 4928, 16928, 4928, 16960, 16992,
+    1824, 1824, 16480, 8416, 17024, 17056, 17088, 1824, 17120, 6784, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 7840, 1824, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    17152, 1344, 1344, 1344, 1344, 1344, 1344, 11360, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 17184, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 17216, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824, 1824,
+    1824, 1824, 1824, 1824, 1824, 1824, 1344, 1344, 1344, 1344, 1344, 1344,
+    1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 1344, 11360
+#endif /* TCL_UTF_MAX > 3 */
+};
+
+/*
+ * The groupMap is indexed by combining the alternate page number with
+ * the page offset and returns a group number that identifies a unique
+ * set of character attributes.
+ */
+
+static const unsigned char groupMap[] = {
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 3, 3, 4, 3, 3, 3, 5, 6, 3, 7, 3, 8,
+    3, 3, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 3, 7, 7, 7, 3, 3, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 5, 3, 6, 11, 12, 11, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 5, 7, 6, 7, 1, 2, 3, 4, 4, 4, 4, 14, 3, 11, 14, 15, 16,
+    7, 17, 14, 11, 14, 7, 18, 18, 11, 19, 3, 3, 11, 18, 15, 20, 18, 18,
+    18, 3, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 7, 10, 10, 10, 10, 10, 10, 10, 21,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 7, 13, 13, 13, 13, 13, 13, 13, 22, 23, 24,
+    23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 25, 26, 23, 24, 23,
+    24, 23, 24, 21, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 21, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 27,
+    23, 24, 23, 24, 23, 24, 28, 29, 30, 23, 24, 23, 24, 31, 23, 24, 32,
+    32, 23, 24, 21, 33, 34, 35, 23, 24, 32, 36, 37, 38, 39, 23, 24, 40,
+    21, 38, 41, 42, 43, 23, 24, 23, 24, 23, 24, 44, 23, 24, 44, 21, 21,
+    23, 24, 44, 23, 24, 45, 45, 23, 24, 23, 24, 46, 23, 24, 21, 15, 23,
+    24, 21, 47, 15, 15, 15, 15, 48, 49, 50, 48, 49, 50, 48, 49, 50, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 51, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    21, 48, 49, 50, 23, 24, 52, 53, 23, 24, 23, 24, 23, 24, 23, 24, 54,
+    21, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 21, 21, 21, 21, 21, 21, 55, 23, 24, 56, 57, 58, 58, 23, 24,
+    59, 60, 61, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 62, 63, 64, 65,
+    66, 21, 67, 67, 21, 68, 21, 69, 70, 21, 21, 21, 67, 71, 21, 72, 21,
+    73, 74, 21, 75, 76, 74, 77, 78, 21, 21, 76, 21, 79, 80, 21, 21, 81,
+    21, 21, 21, 21, 21, 21, 21, 82, 21, 21, 83, 21, 21, 83, 21, 21, 21,
+    84, 83, 85, 86, 86, 87, 21, 21, 21, 21, 21, 88, 21, 15, 21, 21, 21,
+    21, 21, 21, 21, 21, 89, 90, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 91, 91, 91, 11, 11, 11, 11, 91, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 91, 91, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
+    11, 11, 11, 11, 91, 91, 91, 91, 91, 11, 11, 11, 11, 11, 11, 11, 91,
+    11, 91, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
+    11, 11, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 93, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 23, 24, 23,
+    24, 91, 11, 23, 24, 0, 0, 91, 42, 42, 42, 3, 94, 0, 0, 0, 0, 11, 11,
+    95, 3, 96, 96, 96, 0, 97, 0, 98, 98, 21, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 0, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 99, 100, 100, 100, 21, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 101, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 102, 103, 103, 104, 105, 106, 107, 107, 107, 108, 109, 110,
+    23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 111, 112, 113, 114, 115, 116, 7, 23, 24,
+    117, 23, 24, 21, 54, 54, 54, 118, 118, 118, 118, 118, 118, 118, 118,
+    118, 118, 118, 118, 118, 118, 118, 118, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 112, 112, 112, 112, 112, 112, 112, 112, 112,
+    112, 112, 112, 112, 112, 112, 112, 23, 24, 14, 92, 92, 92, 92, 92,
+    119, 119, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 120, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 121, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 0, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    0, 0, 91, 3, 3, 3, 3, 3, 3, 0, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 123, 21, 0, 3, 8, 0, 0, 14, 14, 4, 0, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 8, 92, 3, 92, 92, 3, 92, 92, 3, 92, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0,
+    15, 15, 15, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 17, 17, 17, 17,
+    17, 17, 7, 7, 7, 3, 3, 4, 3, 3, 14, 14, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 3, 17, 0, 3, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 91, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 3, 3, 3, 15, 15, 92,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 3, 15, 92, 92, 92, 92, 92, 92, 92, 17, 14, 92, 92, 92, 92, 92,
+    92, 91, 91, 92, 92, 14, 92, 92, 92, 92, 15, 15, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 15, 15, 15, 14, 14, 15, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
+    3, 3, 0, 17, 15, 92, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 15, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 91, 91, 14, 3, 3, 3, 91, 0, 0, 0, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 92, 92, 92, 92, 91, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 91, 92, 92, 92, 91, 92, 92, 92, 92, 92, 0, 0, 3, 3, 3, 3, 3,
+    3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92,
+    92, 92, 0, 0, 3, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 17, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 124, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92,
+    124, 92, 15, 124, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92, 124, 124,
+    124, 124, 92, 124, 124, 15, 92, 92, 92, 92, 92, 92, 92, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 92, 92, 3, 3, 9, 9, 9, 9, 9, 9, 9, 9, 9,
+    9, 3, 91, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 124, 124, 0, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 0, 0, 0,
+    15, 15, 15, 15, 0, 0, 92, 15, 124, 124, 124, 92, 92, 92, 92, 0, 0,
+    124, 124, 0, 0, 124, 124, 92, 15, 0, 0, 0, 0, 0, 0, 0, 0, 124, 0, 0,
+    0, 0, 15, 15, 0, 15, 15, 15, 92, 92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 15, 15, 4, 4, 18, 18, 18, 18, 18, 18, 14, 4, 15, 3, 0, 0, 0,
+    92, 92, 124, 0, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 15, 15, 0, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 0, 15, 15,
+    0, 15, 15, 0, 0, 92, 0, 124, 124, 124, 92, 92, 0, 0, 0, 0, 92, 92,
+    0, 0, 92, 92, 92, 0, 0, 0, 92, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15,
+    0, 15, 0, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 92, 92, 15,
+    15, 15, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 92, 124, 0, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 0, 15, 15, 15, 15, 15, 0, 0,
+    92, 15, 124, 124, 124, 92, 92, 92, 92, 92, 0, 92, 92, 124, 0, 124,
+    124, 92, 0, 0, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15,
+    15, 92, 92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 4, 0, 0, 0, 0, 0,
+    0, 0, 15, 92, 92, 92, 92, 92, 92, 0, 92, 124, 124, 0, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 0, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15,
+    15, 15, 15, 15, 15, 0, 15, 15, 0, 15, 15, 15, 15, 15, 0, 0, 92, 15,
+    124, 92, 124, 92, 92, 92, 92, 0, 0, 124, 124, 0, 0, 124, 124, 92, 0,
+    0, 0, 0, 0, 0, 0, 0, 92, 124, 0, 0, 0, 0, 15, 15, 0, 15, 15, 15, 92,
+    92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 14, 15, 18, 18, 18, 18, 18,
+    18, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 15, 0, 15, 15, 15, 15, 15, 15,
+    0, 0, 0, 15, 15, 15, 0, 15, 15, 15, 15, 0, 0, 0, 15, 15, 0, 15, 0,
+    15, 15, 0, 0, 0, 15, 15, 0, 0, 0, 15, 15, 15, 0, 0, 0, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 124, 124, 92, 124,
+    124, 0, 0, 0, 124, 124, 124, 0, 124, 124, 124, 92, 0, 0, 15, 0, 0,
+    0, 0, 0, 0, 124, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 18, 18, 18, 14, 14, 14, 14, 14, 14, 4, 14, 0,
+    0, 0, 0, 0, 92, 124, 124, 124, 0, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 15, 92, 92, 92, 124,
+    124, 124, 124, 0, 92, 92, 92, 0, 92, 92, 92, 92, 0, 0, 0, 0, 0, 0,
+    0, 92, 92, 0, 15, 15, 15, 0, 0, 0, 0, 0, 15, 15, 92, 92, 0, 0, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18,
+    18, 18, 14, 15, 92, 124, 124, 0, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 0, 0, 92, 15, 124, 92, 124,
+    124, 124, 124, 124, 0, 92, 124, 124, 0, 124, 124, 92, 92, 0, 0, 0,
+    0, 0, 0, 0, 124, 124, 0, 0, 0, 0, 0, 0, 0, 15, 0, 15, 15, 92, 92, 0,
+    0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 92, 92, 124, 124, 0, 15, 15, 15, 15, 15, 15, 15, 15,
+    0, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92, 15, 124, 124, 124,
+    92, 92, 92, 92, 0, 124, 124, 124, 0, 124, 124, 124, 92, 15, 14, 0,
+    0, 0, 0, 15, 15, 15, 124, 18, 18, 18, 18, 18, 18, 18, 15, 15, 15, 92,
+    92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 14, 15, 15, 15, 15, 15, 15, 0, 0, 124, 124, 0, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    0, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 92, 0, 0, 0, 0, 124,
+    124, 124, 92, 92, 92, 0, 92, 0, 124, 124, 124, 124, 124, 124, 124,
+    124, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 124, 124,
+    3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 92, 15, 15, 92, 92, 92, 92, 92, 92, 92,
+    0, 0, 0, 0, 4, 15, 15, 15, 15, 15, 15, 91, 92, 92, 92, 92, 92, 92,
+    92, 92, 3, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 3, 0, 0, 0, 0, 0, 15, 15,
+    0, 15, 0, 0, 15, 15, 0, 15, 0, 0, 15, 0, 0, 0, 0, 0, 0, 15, 15, 15,
+    15, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 0, 15, 0, 15, 0,
+    0, 15, 15, 0, 15, 15, 15, 15, 92, 15, 15, 92, 92, 92, 92, 92, 92, 0,
+    92, 92, 15, 0, 0, 15, 15, 15, 15, 15, 0, 91, 0, 92, 92, 92, 92, 92,
+    92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 15, 15, 15, 15, 15, 14,
+    14, 14, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 14, 3, 14, 14,
+    14, 92, 92, 14, 14, 14, 14, 14, 14, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 14, 92, 14, 92, 14, 92, 5, 6, 5,
+    6, 124, 124, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 124, 92, 92,
+    92, 92, 92, 3, 92, 92, 15, 15, 15, 15, 15, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 0, 14, 14, 14, 14, 14, 14, 14, 14,
+    92, 14, 14, 14, 14, 14, 14, 0, 14, 14, 3, 3, 3, 3, 3, 14, 14, 14, 14,
+    3, 3, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124,
+    124, 92, 92, 92, 92, 124, 92, 92, 92, 92, 92, 92, 124, 92, 92, 124,
+    124, 92, 92, 15, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 3, 3, 3, 3, 3, 15,
+    15, 15, 15, 15, 15, 124, 124, 92, 92, 15, 15, 15, 15, 92, 92, 92, 15,
+    124, 124, 124, 15, 15, 124, 124, 124, 124, 124, 124, 124, 15, 15, 15,
+    92, 92, 92, 92, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 124, 124, 92, 92, 124, 124, 124, 124, 124, 124, 92, 15, 124, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 124, 124, 124, 92, 14, 14, 125, 125, 125,
+    125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125,
+    125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125, 125,
+    125, 125, 125, 125, 125, 125, 125, 0, 125, 0, 0, 0, 0, 0, 125, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 3, 91, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 0, 15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15,
+    15, 15, 15, 0, 15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 0, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 92, 92, 92, 3, 3, 3, 3, 3, 3,
+    3, 3, 3, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0,
+    0, 0, 0, 0, 0, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126,
+    126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126,
+    126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126, 126,
+    126, 126, 126, 126, 126, 126, 126, 126, 126, 104, 104, 104, 104, 104,
+    104, 0, 0, 110, 110, 110, 110, 110, 110, 0, 0, 8, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 3, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 2, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 5,
+    6, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 3, 3, 3, 127,
+    127, 127, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15,
+    92, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92, 92, 3, 3, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15,
+    0, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92,
+    124, 92, 92, 92, 92, 92, 92, 92, 124, 124, 124, 124, 124, 124, 124,
+    124, 92, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 3, 3,
+    3, 91, 3, 3, 3, 4, 15, 92, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0,
+    0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0,
+    0, 3, 3, 3, 3, 3, 3, 8, 3, 3, 3, 3, 92, 92, 92, 17, 0, 9, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 15, 15, 15, 91, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 92, 92, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 15, 0, 0, 0, 0, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 92, 92, 92, 124, 124, 124, 124, 92,
+    92, 124, 124, 124, 0, 0, 0, 0, 124, 124, 92, 124, 124, 124, 124, 124,
+    124, 92, 92, 92, 0, 0, 0, 0, 14, 0, 0, 0, 3, 3, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 15,
+    15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 18, 0, 0, 0,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 92, 92, 124, 124, 92, 0, 0, 3, 3, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 124, 92, 124, 92, 92, 92, 92, 92, 92, 92, 0, 92, 124, 92, 124,
+    124, 92, 92, 92, 92, 92, 92, 92, 92, 124, 124, 124, 124, 124, 124,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 0, 0, 92, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0,
+    0, 0, 0, 3, 3, 3, 3, 3, 3, 3, 91, 3, 3, 3, 3, 3, 3, 0, 0, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 119, 0, 92, 92, 92, 92,
+    124, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 124, 92,
+    92, 92, 92, 92, 124, 92, 124, 124, 124, 124, 124, 92, 124, 124, 15,
+    15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3,
+    3, 3, 3, 3, 3, 3, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0,
+    92, 92, 124, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124,
+    92, 92, 92, 92, 124, 124, 92, 92, 124, 92, 92, 92, 15, 15, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 124, 92, 92, 124, 124, 124, 92, 124, 92, 92, 92, 124, 124, 0, 0,
+    0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 15, 15, 15, 15, 124, 124, 124, 124, 124,
+    124, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92, 124, 124, 92, 92, 0,
+    0, 0, 3, 3, 3, 3, 3, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 15, 15,
+    15, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 91, 91, 91, 91, 91, 91, 3, 3, 128, 129, 130, 131, 131,
+    132, 133, 134, 135, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0,
+    92, 92, 92, 3, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    124, 92, 92, 92, 92, 92, 92, 92, 15, 15, 15, 15, 92, 15, 15, 15, 15,
+    124, 124, 92, 15, 15, 124, 92, 92, 0, 0, 0, 0, 0, 0, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    91, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 91, 136, 21,
+    21, 21, 137, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 91, 91,
+    91, 91, 91, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 0, 92, 92, 92, 92,
+    92, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 23, 24, 21, 21, 21, 21, 21, 138, 21, 21, 139, 21, 140,
+    140, 140, 140, 140, 140, 140, 140, 141, 141, 141, 141, 141, 141, 141,
+    141, 140, 140, 140, 140, 140, 140, 0, 0, 141, 141, 141, 141, 141, 141,
+    0, 0, 140, 140, 140, 140, 140, 140, 140, 140, 141, 141, 141, 141, 141,
+    141, 141, 141, 140, 140, 140, 140, 140, 140, 140, 140, 141, 141, 141,
+    141, 141, 141, 141, 141, 140, 140, 140, 140, 140, 140, 0, 0, 141, 141,
+    141, 141, 141, 141, 0, 0, 21, 140, 21, 140, 21, 140, 21, 140, 0, 141,
+    0, 141, 0, 141, 0, 141, 140, 140, 140, 140, 140, 140, 140, 140, 141,
+    141, 141, 141, 141, 141, 141, 141, 142, 142, 143, 143, 143, 143, 144,
+    144, 145, 145, 146, 146, 147, 147, 0, 0, 140, 140, 140, 140, 140, 140,
+    140, 140, 148, 148, 148, 148, 148, 148, 148, 148, 140, 140, 140, 140,
+    140, 140, 140, 140, 148, 148, 148, 148, 148, 148, 148, 148, 140, 140,
+    140, 140, 140, 140, 140, 140, 148, 148, 148, 148, 148, 148, 148, 148,
+    140, 140, 21, 149, 21, 0, 21, 21, 141, 141, 150, 150, 151, 11, 152,
+    11, 11, 11, 21, 149, 21, 0, 21, 21, 153, 153, 153, 153, 151, 11, 11,
+    11, 140, 140, 21, 21, 0, 0, 21, 21, 141, 141, 154, 154, 0, 11, 11,
+    11, 140, 140, 21, 21, 21, 113, 21, 21, 141, 141, 155, 155, 117, 11,
+    11, 11, 0, 0, 21, 149, 21, 0, 21, 21, 156, 156, 157, 157, 151, 11,
+    11, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 17, 17, 17, 17, 17, 8, 8, 8,
+    8, 8, 8, 3, 3, 16, 20, 5, 16, 16, 20, 5, 16, 3, 3, 3, 3, 3, 3, 3, 3,
+    158, 159, 17, 17, 17, 17, 17, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 16, 20,
+    3, 3, 3, 3, 12, 12, 3, 3, 3, 7, 5, 6, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
+    3, 7, 3, 12, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 17, 17, 17, 17, 17, 0,
+    17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 18, 91, 0, 0, 18, 18, 18, 18,
+    18, 18, 7, 7, 7, 5, 6, 91, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    7, 7, 7, 5, 6, 0, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4,
+    4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    119, 119, 119, 119, 92, 119, 119, 119, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14,
+    14, 107, 14, 14, 14, 14, 107, 14, 14, 21, 107, 107, 107, 21, 21, 107,
+    107, 107, 21, 14, 107, 14, 14, 7, 107, 107, 107, 107, 107, 14, 14,
+    14, 14, 14, 14, 107, 14, 160, 14, 107, 14, 161, 162, 107, 107, 14,
+    21, 107, 107, 163, 107, 21, 15, 15, 15, 15, 21, 14, 14, 21, 21, 107,
+    107, 7, 7, 7, 7, 7, 107, 21, 21, 21, 21, 14, 7, 14, 14, 164, 14, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 165, 165,
+    165, 165, 165, 165, 165, 165, 165, 165, 165, 165, 165, 165, 165, 165,
+    166, 166, 166, 166, 166, 166, 166, 166, 166, 166, 166, 166, 166, 166,
+    166, 166, 127, 127, 127, 23, 24, 127, 127, 127, 127, 18, 14, 14, 0,
+    0, 0, 0, 7, 7, 7, 7, 7, 14, 14, 14, 14, 14, 7, 7, 14, 14, 14, 14, 7,
+    14, 14, 7, 14, 14, 7, 14, 14, 14, 14, 14, 14, 14, 7, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 7, 14, 14, 7, 14, 7, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 14, 14, 14, 14, 14,
+    14, 14, 14, 5, 6, 5, 6, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 7, 14, 14, 14, 14, 14, 14, 14,
+    5, 6, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 7, 7, 7, 7, 7, 7, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 167, 167, 167, 167, 167, 167, 167, 167,
+    167, 167, 167, 167, 167, 167, 167, 167, 167, 167, 167, 167, 167, 167,
+    167, 167, 167, 167, 168, 168, 168, 168, 168, 168, 168, 168, 168, 168,
+    168, 168, 168, 168, 168, 168, 168, 168, 168, 168, 168, 168, 168, 168,
+    168, 168, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 7, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 7, 7, 7, 7, 7, 7, 7, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 7, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 7, 7, 7, 7, 7, 5, 6, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5,
+    6, 5, 6, 5, 6, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 5, 6, 5, 6, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 5, 6, 7, 7, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
+    7, 7, 7, 7, 14, 14, 7, 7, 7, 7, 7, 7, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122, 122,
+    122, 122, 122, 122, 122, 122, 122, 122, 122, 0, 123, 123, 123, 123,
+    123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123, 123,
+    123, 0, 23, 24, 169, 170, 171, 172, 173, 23, 24, 23, 24, 23, 24, 174,
+    175, 176, 177, 21, 23, 24, 21, 23, 24, 21, 21, 21, 21, 21, 91, 91,
+    178, 178, 23, 24, 23, 24, 21, 14, 14, 14, 14, 14, 14, 23, 24, 23, 24,
+    92, 92, 92, 23, 24, 0, 0, 0, 0, 0, 3, 3, 3, 3, 18, 3, 3, 179, 179,
+    179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179,
+    179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, 179,
+    179, 179, 179, 179, 179, 179, 179, 179, 0, 179, 0, 0, 0, 0, 0, 179,
+    0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 91, 3, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15,
+    15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15,
+    0, 3, 3, 16, 20, 16, 20, 3, 3, 3, 16, 20, 3, 16, 20, 3, 3, 3, 3, 3,
+    3, 3, 3, 3, 8, 3, 3, 8, 3, 16, 20, 3, 3, 16, 20, 5, 6, 5, 6, 5, 6,
+    5, 6, 3, 3, 3, 3, 3, 91, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 8, 8, 3, 3,
+    3, 3, 8, 3, 5, 3, 3, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 0, 0, 0, 0, 2, 3, 3, 3, 14, 91, 15, 127, 5, 6, 5, 6, 5,
+    6, 5, 6, 5, 6, 14, 14, 5, 6, 5, 6, 5, 6, 5, 6, 8, 5, 6, 6, 14, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 92, 92, 92, 92, 124, 124, 8,
+    91, 91, 91, 91, 91, 14, 14, 127, 127, 127, 91, 15, 3, 14, 14, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 0, 92, 92, 11, 11, 91, 91, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 3, 91, 91, 91, 15, 0, 0, 0, 0, 0, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 0, 14, 14, 18, 18, 18, 18, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 14, 14,
+    14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 18, 18, 18,
+    18, 18, 18, 18, 18, 14, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 91, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 91, 3, 3, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 15, 15, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 15, 92, 119, 119, 119,
+    3, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 3, 91, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 23, 24, 91, 91, 92, 92, 15, 15, 15, 15, 15, 15, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 127, 92, 92, 3, 3, 3, 3, 3,
+    3, 0, 0, 0, 0, 0, 0, 0, 0, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
+    11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 91, 91, 91, 91,
+    91, 91, 91, 91, 91, 11, 11, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 23, 24, 21, 21, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24,
+    23, 24, 91, 21, 21, 21, 21, 21, 21, 21, 21, 23, 24, 23, 24, 180, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 91, 11, 11, 23, 24, 181, 21, 15,
+    23, 24, 23, 24, 21, 21, 23, 24, 23, 24, 23, 24, 23, 24, 23, 24, 23,
+    24, 23, 24, 23, 24, 23, 24, 23, 24, 182, 183, 184, 185, 182, 0, 186,
+    187, 188, 189, 23, 24, 23, 24, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 91, 91,
+    21, 15, 15, 15, 15, 15, 15, 15, 92, 15, 15, 15, 92, 15, 15, 15, 15,
+    92, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 124, 124, 92, 92, 124, 14, 14, 14, 14,
+    0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 14, 14, 4, 14, 0, 0, 0, 0, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 124, 124, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124, 124, 124,
+    124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 92,
+    92, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0,
+    0, 0, 0, 0, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 15, 15, 15, 15, 15, 15, 3, 3, 3, 15, 3, 15, 0,
+    0, 15, 15, 15, 15, 15, 15, 92, 92, 92, 92, 92, 92, 92, 92, 3, 3, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 124,
+    124, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 124, 124, 92, 92, 92, 92, 124,
+    124, 92, 124, 124, 124, 124, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
+    0, 91, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 3, 3, 15, 15, 15,
+    15, 15, 92, 91, 15, 15, 15, 15, 15, 15, 15, 15, 15, 9, 9, 9, 9, 9,
+    9, 9, 9, 9, 9, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 92, 92, 92, 92, 92, 92, 124, 124, 92, 92, 124, 124, 92, 92, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 92, 15, 15, 15, 15, 15, 15, 15,
+    15, 92, 124, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 3, 3, 3, 3,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 91,
+    15, 15, 15, 15, 15, 15, 14, 14, 14, 15, 124, 92, 124, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 15, 92,
+    92, 92, 15, 15, 92, 92, 15, 15, 15, 15, 15, 92, 92, 15, 92, 15, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    15, 15, 91, 3, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124,
+    92, 92, 124, 124, 3, 3, 15, 91, 91, 124, 92, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 0, 0, 15,
+    15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 0, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 190, 21, 21, 21, 21, 21,
+    21, 21, 11, 91, 91, 91, 91, 21, 21, 21, 21, 21, 21, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191,
+    191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191,
+    191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191, 191,
+    191, 191, 191, 191, 191, 191, 191, 191, 191, 15, 15, 15, 124, 124,
+    92, 124, 124, 92, 124, 124, 3, 124, 92, 0, 0, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 192, 192, 192,
+    192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192,
+    192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192, 192,
+    192, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193,
+    193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193, 193,
+    193, 193, 193, 193, 193, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 21,
+    21, 21, 21, 21, 21, 21, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 21, 21,
+    21, 21, 21, 0, 0, 0, 0, 0, 15, 92, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 7, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15,
+    15, 15, 15, 15, 0, 15, 0, 15, 15, 0, 15, 15, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
+    11, 11, 11, 11, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 6, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 4,
+    14, 0, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 3, 3, 3, 3, 3, 3, 3, 5, 6, 3, 0, 0, 0, 0, 0, 0, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 3, 8, 8, 12, 12, 5,
+    6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 5, 6, 3, 3, 5, 6, 3, 3, 3, 3,
+    12, 12, 12, 3, 3, 3, 0, 3, 3, 3, 3, 8, 5, 6, 5, 6, 5, 6, 3, 3, 3, 7,
+    8, 7, 7, 7, 0, 3, 4, 3, 3, 0, 0, 0, 0, 15, 15, 15, 15, 15, 0, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 17, 0, 3, 3, 3, 4, 3, 3, 3, 5, 6, 3, 7, 3, 8, 3,
+    3, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 3, 3, 7, 7, 7, 3, 11, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 5, 7, 6, 7, 5, 6, 3, 5, 6, 3, 3, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 91, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 91, 91, 0, 0, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15,
+    15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 15, 0, 0, 0, 4,
+    4, 7, 11, 14, 4, 4, 0, 14, 7, 7, 7, 7, 14, 14, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 17, 17, 17, 14, 14, 0, 0
+#if TCL_UTF_MAX > 3
+    ,15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 0, 3, 3, 3, 0, 0, 0, 0, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 127, 18,
+    18, 18, 18, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 18, 18, 14, 14, 14, 0, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 0, 0, 0, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 92, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 18, 18, 18, 18,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 127, 15, 15, 15, 15, 15, 15,
+    15, 15, 127, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92, 92, 92, 92, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    3, 15, 15, 15, 15, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 3, 127,
+    127, 127, 127, 127, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 194, 194, 194, 194,
+    194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194,
+    194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194,
+    194, 194, 194, 194, 194, 194, 194, 194, 195, 195, 195, 195, 195, 195,
+    195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195,
+    195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195,
+    195, 195, 195, 195, 195, 195, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 194,
+    194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194,
+    194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194, 194,
+    194, 194, 194, 194, 194, 194, 194, 0, 0, 0, 0, 195, 195, 195, 195,
+    195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195,
+    195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195, 195,
+    195, 195, 195, 195, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 0, 0, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15,
+    0, 0, 0, 15, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 3, 18, 18, 18, 18, 18,
+    18, 18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 14, 14, 18, 18, 18, 18, 18, 18,
+    18, 0, 0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 0, 0, 0,
+    0, 0, 18, 18, 18, 18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 18, 18, 18, 18, 18, 18,
+    0, 0, 0, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 3, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 18, 18, 15, 15, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 15, 92, 92, 92, 0, 92, 92,
+    0, 0, 0, 0, 0, 92, 92, 92, 92, 15, 15, 15, 15, 0, 15, 15, 15, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 92, 92, 92, 0, 0, 0,
+    0, 92, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3,
+    3, 3, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 18, 18, 3, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 18, 18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 14, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 92, 0, 0, 0, 0, 18, 18, 18,
+    18, 18, 3, 3, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 0, 0, 0, 3, 3, 3, 3, 3, 3, 3, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 18, 18,
+    18, 18, 18, 18, 18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18,
+    18, 18, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 97,
+    97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+    97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+    97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 102, 102, 102, 102, 102, 102, 102,
+    102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102,
+    102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102,
+    102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102, 102,
+    102, 102, 0, 0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 124, 92, 124, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 3, 3, 3,
+    3, 3, 3, 3, 0, 0, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124, 124, 124, 92, 92, 92,
+    92, 124, 124, 92, 92, 3, 3, 17, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 92, 92, 92, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 92, 92, 92, 92, 124, 92, 92, 92, 92, 92, 92, 92, 92, 0, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 92, 3, 3, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124, 124, 124,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 124, 124, 15, 15, 15, 15, 3, 3,
+    3, 3, 3, 92, 92, 92, 3, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 15, 3,
+    15, 3, 3, 3, 0, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 124, 124, 124, 92, 92, 92, 124, 124,
+    92, 124, 92, 92, 3, 3, 3, 3, 3, 3, 92, 0, 15, 15, 15, 15, 15, 15, 15,
+    0, 15, 0, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 3, 0,
+    0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 124, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92, 0, 0, 0, 0, 0, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 92, 92, 124, 124, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 0, 15, 15, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 92, 124, 124, 124, 124, 0, 0, 124,
+    124, 0, 0, 124, 124, 124, 0, 0, 15, 0, 0, 0, 0, 0, 0, 124, 0, 0, 0,
+    0, 0, 15, 15, 15, 15, 15, 124, 124, 0, 0, 92, 92, 92, 92, 92, 92, 92,
+    0, 0, 0, 92, 92, 92, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 124, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92, 124, 124, 92,
+    92, 92, 124, 92, 15, 15, 15, 15, 3, 3, 3, 3, 3, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 0, 3, 0, 3, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 124, 124, 124, 92, 92, 92, 92, 92, 92, 124,
+    92, 124, 124, 124, 124, 92, 92, 124, 92, 92, 15, 15, 3, 15, 0, 0, 0,
+    0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 124, 124, 124,
+    92, 92, 92, 92, 0, 0, 124, 124, 124, 124, 92, 92, 124, 92, 92, 3, 3,
+    3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 15,
+    15, 15, 15, 92, 92, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 124, 124, 124, 92, 92, 92, 92, 92, 92, 92, 92,
+    124, 124, 92, 124, 92, 92, 3, 3, 3, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 3,
+    3, 3, 3, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 92, 124, 92,
+    124, 124, 92, 92, 92, 92, 92, 92, 124, 92, 0, 0, 0, 0, 0, 0, 0, 0,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    0, 0, 92, 92, 92, 124, 124, 92, 92, 92, 92, 124, 92, 92, 92, 92, 92,
+    0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 18, 18, 3, 3, 3, 14, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+    10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 92, 92, 92, 92, 92, 92, 124, 124, 92, 92, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 92, 92, 92, 92, 92, 92, 124, 15, 92, 92, 92, 92, 3,
+    3, 3, 3, 3, 3, 3, 3, 92, 0, 0, 0, 0, 0, 0, 0, 0, 15, 92, 92, 92, 92,
+    92, 92, 124, 124, 92, 92, 92, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0,
+    15, 15, 15, 15, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    124, 92, 92, 3, 3, 3, 0, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 124, 92, 92, 92, 92, 92, 92, 92, 0, 92,
+    92, 92, 92, 92, 92, 124, 92, 15, 3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 3, 3, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 0, 124, 92, 92, 92, 92, 92, 92, 92, 124, 92, 92, 124, 92, 92, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 0, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 92, 92, 92, 92, 92, 92, 0, 0, 0, 92, 0, 92, 92, 0, 92,
+    92, 92, 92, 92, 92, 92, 15, 92, 0, 0, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 0, 0, 127, 127, 127, 127, 127, 127, 127,
+    127, 127, 127, 127, 127, 127, 127, 127, 0, 3, 3, 3, 3, 3, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 92, 92, 92, 92, 92, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    92, 92, 92, 92, 92, 92, 92, 3, 3, 3, 3, 3, 14, 14, 14, 14, 91, 91,
+    91, 91, 3, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 0, 18, 18, 18, 18, 18, 18, 18, 0, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0,
+    0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15,
+    15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 124, 124, 124, 124,
+    124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124,
+    124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124,
+    124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124, 124,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 92, 92, 92, 91,
+    91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 0, 0, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0,
+    0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 14, 92, 92, 3, 17,
+    17, 17, 17, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 0, 0, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 124, 124, 92, 92, 92, 14, 14, 14,
+    124, 124, 124, 124, 124, 124, 17, 17, 17, 17, 17, 17, 17, 17, 92, 92,
+    92, 92, 92, 92, 92, 92, 14, 14, 92, 92, 92, 92, 92, 92, 92, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 92, 92, 92, 92, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 92, 92, 92, 14, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 21, 21, 21, 21, 21, 21,
+    21, 0, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 107, 0, 107, 107, 0, 0, 107,
+    0, 0, 107, 107, 0, 0, 107, 107, 107, 107, 0, 107, 107, 107, 107, 107,
+    107, 107, 107, 21, 21, 21, 21, 0, 21, 0, 21, 21, 21, 21, 21, 21, 21,
+    0, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 107, 107, 0, 107, 107, 107, 107, 0, 0, 107, 107, 107, 107,
+    107, 107, 107, 107, 0, 107, 107, 107, 107, 107, 107, 107, 0, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 107, 107, 0, 107, 107, 107, 107, 0, 107,
+    107, 107, 107, 107, 0, 107, 0, 0, 0, 107, 107, 107, 107, 107, 107,
+    107, 0, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 107, 107, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 21, 21, 21, 21, 21, 21, 0, 0, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 7, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 7, 21, 21, 21, 21, 21, 21, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 7, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 7, 21, 21,
+    21, 21, 21, 21, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    7, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 7, 21, 21, 21, 21, 21, 21, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 7, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 7, 21, 21, 21, 21, 21, 21, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107, 107,
+    107, 107, 107, 107, 107, 107, 7, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+    21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 7,
+    21, 21, 21, 21, 21, 21, 107, 21, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9,
+    9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 14, 14, 14, 14, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 92, 92, 92, 14, 14, 14, 14, 14, 14, 14, 14, 92, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 92, 14, 14, 3,
+    3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 92, 92,
+    92, 92, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 92, 92, 92, 92,
+    92, 92, 92, 0, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+    92, 92, 92, 92, 0, 0, 92, 92, 92, 92, 92, 92, 92, 0, 92, 92, 0, 92,
+    92, 92, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 15, 15, 15, 15, 15, 0, 0, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 92, 92, 92, 92, 92, 92, 92, 0, 0, 0, 0, 0, 0, 0, 0, 0, 196, 196,
+    196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196,
+    196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196, 196,
+    196, 196, 196, 196, 197, 197, 197, 197, 197, 197, 197, 197, 197, 197,
+    197, 197, 197, 197, 197, 197, 197, 197, 197, 197, 197, 197, 197, 197,
+    197, 197, 197, 197, 197, 197, 197, 197, 197, 197, 92, 92, 92, 92, 92,
+    92, 92, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0, 0, 3,
+    3, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15,
+    15, 0, 15, 0, 0, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0,
+    15, 15, 15, 15, 0, 15, 0, 15, 0, 0, 0, 0, 0, 0, 15, 0, 0, 0, 0, 15,
+    0, 15, 0, 15, 0, 15, 15, 15, 0, 15, 15, 0, 15, 0, 0, 15, 0, 15, 0,
+    15, 0, 15, 0, 15, 0, 15, 15, 0, 15, 0, 0, 15, 15, 15, 15, 0, 15, 15,
+    15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 0, 15, 15, 15, 15, 0, 15, 0,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 15, 15,
+    15, 0, 15, 15, 15, 15, 15, 0, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 7, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 18, 18, 18, 18, 18, 18, 18, 18,
+    18, 18, 18, 18, 18, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0,
+    0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 14,
+    14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 11, 11, 11, 11, 11,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0,
+    0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0,
+    0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 14, 14,
+    14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 0, 0, 0, 0, 0, 0, 0, 0, 14, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+    14, 14, 14, 14, 14, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    15, 15, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 15, 15, 15, 15, 15,
+    15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0
+#endif /* TCL_UTF_MAX > 3 */
+};
+
+/*
+ * Each group represents a unique set of character attributes.  The attributes
+ * are encoded into a 32-bit value as follows:
+ *
+ * Bits 0-4	Character category: see the constants listed below.
+ *
+ * Bits 5-7	Case delta type: 000 = identity
+ *				 010 = add delta for lower
+ *				 011 = add delta for lower, add 1 for title
+ *				 100 = subtract delta for title/upper
+ *				 101 = sub delta for upper, sub 1 for title
+ *				 110 = sub delta for upper, add delta for lower
+ *
+ * Bits 8-31	Case delta: delta for case conversions.  This should be the
+ *			    highest field so we can easily sign extend.
+ */
+
+static const int groups[] = {
+    0, 15, 12, 25, 27, 21, 22, 26, 20, 9, 8257, 28, 19, 8322, 29,
+    5, 23, 16, 11, -190078, 24, 2, -30846, 321, 386, -50879, 59522,
+    -30911, 76930, -49790, 53825, 52801, 52545, 20289, 51777, 52033,
+    53057, -24702, 54081, 53569, -41598, 54593, -33150, 54849, 55873,
+    55617, 56129, -14206, 609, 451, 674, 20354, -24767, -14271, -33215,
+    2763585, -41663, 2762817, -2768510, -49855, 17729, 18241, -2760318,
+    -2759550, -2760062, 53890, 52866, 52610, 51842, 52098, -10833534,
+    -10832510, 53122, -10823550, -10830718, 53634, 54146, -2750078,
+    -10829950, -2751614, 54658, 54914, -2745982, 55938, -10824062,
+    17794, 55682, 18306, 56194, -10818686, -10817918, 4, 6, -21370,
+    29761, 9793, 9537, 16449, 16193, 9858, 9602, 8066, 16514, 16258,
+    2113, 16002, 14722, 1, 12162, 13954, 2178, 22146, 20610, -1662,
+    29826, -15295, 24706, -1727, 20545, 7, 3905, 3970, 12353, 12418,
+    8, 1859649, 9949249, 10, 1601154, 1600898, 1598594, 1598082, 1598338,
+    1596546, 1582466, -9027966, -9044862, -976254, 15234, -1949375,
+    -1918, -1983, -18814, -21886, -25470, -32638, -28542, -32126,
+    -1981, -2174, -18879, -2237, 1844610, -21951, -25535, -28607,
+    -32703, -32191, 13, 14, -1924287, -2145983, -2115007, 7233, 7298,
+    4170, 4234, 6749, 6813, -2750143, -976319, -2746047, 2763650,
+    2762882, -2759615, -2751679, -2760383, -2760127, -2768575, 1859714,
+    -9044927, -10823615, -10830783, -10833599, -10832575, -10830015,
+    -10817983, -10824127, -10818751, 237633, 237698, 9949314, 18,
+    17, 10305, 10370, 8769, 8834
+};
+
+#if TCL_UTF_MAX > 3
+#   define UNICODE_OUT_OF_RANGE(ch) (((ch) & 0x1fffff) >= 0x2fa20)
+#else
+#   define UNICODE_OUT_OF_RANGE(ch) (((ch) & 0x1f0000) != 0)
+#endif
+
+/*
+ * The following constants are used to determine the category of a
+ * Unicode character.
+ */
+
+enum {
+    UNASSIGNED,
+    UPPERCASE_LETTER,
+    LOWERCASE_LETTER,
+    TITLECASE_LETTER,
+    MODIFIER_LETTER,
+    OTHER_LETTER,
+    NON_SPACING_MARK,
+    ENCLOSING_MARK,
+    COMBINING_SPACING_MARK,
+    DECIMAL_DIGIT_NUMBER,
+    LETTER_NUMBER,
+    OTHER_NUMBER,
+    SPACE_SEPARATOR,
+    LINE_SEPARATOR,
+    PARAGRAPH_SEPARATOR,
+    CONTROL,
+    FORMAT,
+    PRIVATE_USE,
+    SURROGATE,
+    CONNECTOR_PUNCTUATION,
+    DASH_PUNCTUATION,
+    OPEN_PUNCTUATION,
+    CLOSE_PUNCTUATION,
+    INITIAL_QUOTE_PUNCTUATION,
+    FINAL_QUOTE_PUNCTUATION,
+    OTHER_PUNCTUATION,
+    MATH_SYMBOL,
+    CURRENCY_SYMBOL,
+    MODIFIER_SYMBOL,
+    OTHER_SYMBOL
+};
+
+/*
+ * The following macros extract the fields of the character info.  The
+ * GetDelta() macro is complicated because we can't rely on the C compiler
+ * to do sign extension on right shifts.
+ */
+
+#define GetCaseType(info) (((info) & 0xe0) >> 5)
+#define GetCategory(ch) (GetUniCharInfo(ch) & 0x1f)
+#define GetDelta(info) ((info) >> 8)
+
+/*
+ * This macro extracts the information about a character from the
+ * Unicode character tables.
+ */
+
+#if TCL_UTF_MAX > 3
+#   define GetUniCharInfo(ch) (groups[groupMap[pageMap[((ch) & 0x1fffff) >> OFFSET_BITS] | ((ch) & ((1 << OFFSET_BITS)-1))]])
+#else
+#   define GetUniCharInfo(ch) (groups[groupMap[pageMap[((ch) & 0xffff) >> OFFSET_BITS] | ((ch) & ((1 << OFFSET_BITS)-1))]])
+#endif
diff --git a/generic/tclUtf.c b/generic/tclUtf.c
new file mode 100644
index 0000000..25cc2d1
--- /dev/null
+++ b/generic/tclUtf.c
@@ -0,0 +1,2071 @@
+/*
+ * tclUtf.c --
+ *
+ *	Routines for manipulating UTF-8 strings.
+ *
+ * Copyright (c) 1997-1998 Sun Microsystems, Inc.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+
+/*
+ * Include the static character classification tables and macros.
+ */
+
+#include "tclUniData.c"
+
+/*
+ * The following macros are used for fast character category tests. The x_BITS
+ * values are shifted right by the category value to determine whether the
+ * given category is included in the set.
+ */
+
+#define ALPHA_BITS ((1 << UPPERCASE_LETTER) | (1 << LOWERCASE_LETTER) \
+	| (1 << TITLECASE_LETTER) | (1 << MODIFIER_LETTER) | (1<<OTHER_LETTER))
+
+#define CONTROL_BITS ((1 << CONTROL) | (1 << FORMAT) | (1 << PRIVATE_USE))
+
+#define DIGIT_BITS (1 << DECIMAL_DIGIT_NUMBER)
+
+#define SPACE_BITS ((1 << SPACE_SEPARATOR) | (1 << LINE_SEPARATOR) \
+	| (1 << PARAGRAPH_SEPARATOR))
+
+#define WORD_BITS (ALPHA_BITS | DIGIT_BITS | (1 << CONNECTOR_PUNCTUATION))
+
+#define PUNCT_BITS ((1 << CONNECTOR_PUNCTUATION) | \
+	(1 << DASH_PUNCTUATION) | (1 << OPEN_PUNCTUATION) | \
+	(1 << CLOSE_PUNCTUATION) | (1 << INITIAL_QUOTE_PUNCTUATION) | \
+	(1 << FINAL_QUOTE_PUNCTUATION) | (1 << OTHER_PUNCTUATION))
+
+#define GRAPH_BITS (WORD_BITS | PUNCT_BITS | \
+	(1 << NON_SPACING_MARK) | (1 << ENCLOSING_MARK) | \
+	(1 << COMBINING_SPACING_MARK) | (1 << LETTER_NUMBER) | \
+	(1 << OTHER_NUMBER) | \
+	(1 << MATH_SYMBOL) | (1 << CURRENCY_SYMBOL) | \
+	(1 << MODIFIER_SYMBOL) | (1 << OTHER_SYMBOL))
+
+/*
+ * Unicode characters less than this value are represented by themselves in
+ * UTF-8 strings.
+ */
+
+#define UNICODE_SELF	0x80
+
+/*
+ * The following structures are used when mapping between Unicode (UCS-2) and
+ * UTF-8.
+ */
+
+static const unsigned char totalBytes[256] = {
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+    2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+    3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,
+#if TCL_UTF_MAX > 3
+    4,4,4,4,4,4,4,4,
+#else
+    1,1,1,1,1,1,1,1,
+#endif
+    1,1,1,1,1,1,1,1
+};
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * TclUtfCount --
+ *
+ *	Find the number of bytes in the Utf character "ch".
+ *
+ * Results:
+ *	The return values is the number of bytes in the Utf character "ch".
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+TclUtfCount(
+    int ch)			/* The Tcl_UniChar whose size is returned. */
+{
+    if ((unsigned)(ch - 1) < (UNICODE_SELF - 1)) {
+	return 1;
+    }
+    if (ch <= 0x7FF) {
+	return 2;
+    }
+#if TCL_UTF_MAX > 3
+    if (((unsigned)(ch - 0x10000) <= 0xFFFFF)) {
+	return 4;
+    }
+#endif
+    return 3;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UniCharToUtf --
+ *
+ *	Store the given Tcl_UniChar as a sequence of UTF-8 bytes in the
+ *	provided buffer. Equivalent to Plan 9 runetochar().
+ *
+ * Results:
+ *	The return values is the number of bytes in the buffer that were
+ *	consumed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharToUtf(
+    int ch,			/* The Tcl_UniChar to be stored in the
+				 * buffer. */
+    char *buf)			/* Buffer in which the UTF-8 representation of
+				 * the Tcl_UniChar is stored. Buffer must be
+				 * large enough to hold the UTF-8 character
+				 * (at most TCL_UTF_MAX bytes). */
+{
+    if ((unsigned)(ch - 1) < (UNICODE_SELF - 1)) {
+	buf[0] = (char) ch;
+	return 1;
+    }
+    if (ch >= 0) {
+	if (ch <= 0x7FF) {
+	    buf[1] = (char) ((ch | 0x80) & 0xBF);
+	    buf[0] = (char) ((ch >> 6) | 0xC0);
+	    return 2;
+	}
+	if (ch <= 0xFFFF) {
+#if TCL_UTF_MAX == 4
+	    if ((ch & 0xF800) == 0xD800) {
+		if (ch & 0x0400) {
+		    /* Low surrogate */
+		    buf[3] = (char) ((ch | 0x80) & 0xBF);
+		    buf[2] |= (char) (((ch >> 6) | 0x80) & 0x8F);
+		    return 4;
+		} else {
+		    /* High surrogate */
+		    ch += 0x40;
+		    buf[2] = (char) (((ch << 4) | 0x80) & 0xB0);
+		    buf[1] = (char) (((ch >> 2) | 0x80) & 0xBF);
+		    buf[0] = (char) (((ch >> 8) | 0xF0) & 0xF7);
+		    return 0;
+		}
+	    }
+#endif
+	    goto three;
+	}
+
+#if TCL_UTF_MAX > 3
+	if (ch <= 0x10FFFF) {
+	    buf[3] = (char) ((ch | 0x80) & 0xBF);
+	    buf[2] = (char) (((ch >> 6) | 0x80) & 0xBF);
+	    buf[1] = (char) (((ch >> 12) | 0x80) & 0xBF);
+	    buf[0] = (char) ((ch >> 18) | 0xF0);
+	    return 4;
+	}
+#endif
+    }
+
+    ch = 0xFFFD;
+three:
+    buf[2] = (char) ((ch | 0x80) & 0xBF);
+    buf[1] = (char) (((ch >> 6) | 0x80) & 0xBF);
+    buf[0] = (char) ((ch >> 12) | 0xE0);
+    return 3;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UniCharToUtfDString --
+ *
+ *	Convert the given Unicode string to UTF-8.
+ *
+ * Results:
+ *	The return value is a pointer to the UTF-8 representation of the
+ *	Unicode string. Storage for the return value is appended to the end of
+ *	dsPtr.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+char *
+Tcl_UniCharToUtfDString(
+    const Tcl_UniChar *uniStr,	/* Unicode string to convert to UTF-8. */
+    int uniLength,		/* Length of Unicode string in Tcl_UniChars
+				 * (must be >= 0). */
+    Tcl_DString *dsPtr)		/* UTF-8 representation of string is appended
+				 * to this previously initialized DString. */
+{
+    const Tcl_UniChar *w, *wEnd;
+    char *p, *string;
+    int oldLength;
+
+    /*
+     * UTF-8 string length in bytes will be <= Unicode string length *
+     * TCL_UTF_MAX.
+     */
+
+    oldLength = Tcl_DStringLength(dsPtr);
+    Tcl_DStringSetLength(dsPtr, (oldLength + uniLength + 1) * TCL_UTF_MAX);
+    string = Tcl_DStringValue(dsPtr) + oldLength;
+
+    p = string;
+    wEnd = uniStr + uniLength;
+    for (w = uniStr; w < wEnd; ) {
+	p += Tcl_UniCharToUtf(*w, p);
+	w++;
+    }
+    Tcl_DStringSetLength(dsPtr, oldLength + (p - string));
+
+    return string;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfToUniChar --
+ *
+ *	Extract the Tcl_UniChar represented by the UTF-8 string. Bad UTF-8
+ *	sequences are converted to valid Tcl_UniChars and processing
+ *	continues. Equivalent to Plan 9 chartorune().
+ *
+ *	The caller must ensure that the source buffer is long enough that this
+ *	routine does not run off the end and dereference non-existent memory
+ *	looking for trail bytes. If the source buffer is known to be '\0'
+ *	terminated, this cannot happen. Otherwise, the caller should call
+ *	Tcl_UtfCharComplete() before calling this routine to ensure that
+ *	enough bytes remain in the string.
+ *
+ *	If TCL_UTF_MAX == 4, special handling of Surrogate pairs is done:
+ *	For any UTF-8 string containing a character outside of the BMP, the
+ *	first call to this function will fill *chPtr with the high surrogate
+ *	and generate a return value of 0. Calling Tcl_UtfToUniChar again
+ *	will produce the low surrogate and a return value of 4. Because *chPtr
+ *	is used to remember whether the high surrogate is already produced, it
+ *	is recommended to initialize the variable it points to as 0 before
+ *	the first call to Tcl_UtfToUniChar is done.
+ *
+ * Results:
+ *	*chPtr is filled with the Tcl_UniChar, and the return value is the
+ *	number of bytes from the UTF-8 string that were consumed.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfToUniChar(
+    register const char *src,	/* The UTF-8 string. */
+    register Tcl_UniChar *chPtr)/* Filled with the Tcl_UniChar represented by
+				 * the UTF-8 string. */
+{
+    register int byte;
+
+    /*
+     * Unroll 1 to 3 (or 4) byte UTF-8 sequences.
+     */
+
+    byte = *((unsigned char *) src);
+    if (byte < 0xC0) {
+	/*
+	 * Handles properly formed UTF-8 characters between 0x01 and 0x7F.
+	 * Also treats \0 and naked trail bytes 0x80 to 0xBF as valid
+	 * characters representing themselves.
+	 */
+
+	*chPtr = (Tcl_UniChar) byte;
+	return 1;
+    } else if (byte < 0xE0) {
+	if ((src[1] & 0xC0) == 0x80) {
+	    /*
+	     * Two-byte-character lead-byte followed by a trail-byte.
+	     */
+
+	    *chPtr = (Tcl_UniChar) (((byte & 0x1F) << 6) | (src[1] & 0x3F));
+	    if ((unsigned)(*chPtr - 1) >= (UNICODE_SELF - 1)) {
+		return 2;
+	    }
+	}
+
+	/*
+	 * A two-byte-character lead-byte not followed by trail-byte
+	 * represents itself.
+	 */
+    } else if (byte < 0xF0) {
+	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80)) {
+	    /*
+	     * Three-byte-character lead byte followed by two trail bytes.
+	     */
+
+	    *chPtr = (Tcl_UniChar) (((byte & 0x0F) << 12)
+		    | ((src[1] & 0x3F) << 6) | (src[2] & 0x3F));
+	    if (*chPtr > 0x7FF) {
+		return 3;
+	    }
+	}
+
+	/*
+	 * A three-byte-character lead-byte not followed by two trail-bytes
+	 * represents itself.
+	 */
+    }
+#if TCL_UTF_MAX > 3
+    else if (byte < 0xF8) {
+	if (((src[1] & 0xC0) == 0x80) && ((src[2] & 0xC0) == 0x80) && ((src[3] & 0xC0) == 0x80)) {
+	    /*
+	     * Four-byte-character lead byte followed by three trail bytes.
+	     */
+#if TCL_UTF_MAX == 4
+	    Tcl_UniChar surrogate;
+
+	    byte = (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
+		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F)) - 0x10000;
+	    surrogate = (Tcl_UniChar) (0xD800 + (byte >> 10));
+	    if (byte & 0x100000) {
+		/* out of range, < 0x10000 or > 0x10ffff */
+	    } else if (*chPtr != surrogate) {
+		/* produce high surrogate, but don't advance source pointer */
+		*chPtr = surrogate;
+		return 0;
+	    } else {
+		/* produce low surrogate, and advance source pointer */
+		*chPtr = (Tcl_UniChar) (0xDC00 | (byte & 0x3FF));
+		return 4;
+	    }
+#else
+	    *chPtr = (Tcl_UniChar) (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
+		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F));
+	    if ((unsigned)(*chPtr - 0x10000) <= 0xFFFFF) {
+		return 4;
+	    }
+#endif
+	}
+
+	/*
+	 * A four-byte-character lead-byte not followed by two trail-bytes
+	 * represents itself.
+	 */
+    }
+#endif
+
+    *chPtr = (Tcl_UniChar) byte;
+    return 1;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfToUniCharDString --
+ *
+ *	Convert the UTF-8 string to Unicode.
+ *
+ * Results:
+ *	The return value is a pointer to the Unicode representation of the
+ *	UTF-8 string. Storage for the return value is appended to the end of
+ *	dsPtr. The Unicode string is terminated with a Unicode NULL character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_UniChar *
+Tcl_UtfToUniCharDString(
+    const char *src,		/* UTF-8 string to convert to Unicode. */
+    int length,			/* Length of UTF-8 string in bytes, or -1 for
+				 * strlen(). */
+    Tcl_DString *dsPtr)		/* Unicode representation of string is
+				 * appended to this previously initialized
+				 * DString. */
+{
+    Tcl_UniChar ch, *w, *wString;
+    const char *p, *end;
+    int oldLength;
+
+    if (length < 0) {
+	length = strlen(src);
+    }
+
+    /*
+     * Unicode string length in Tcl_UniChars will be <= UTF-8 string length in
+     * bytes.
+     */
+
+    oldLength = Tcl_DStringLength(dsPtr);
+/* TODO: fix overreach! */
+    Tcl_DStringSetLength(dsPtr,
+	    (int) ((oldLength + length + 1) * sizeof(Tcl_UniChar)));
+    wString = (Tcl_UniChar *) (Tcl_DStringValue(dsPtr) + oldLength);
+
+    w = wString;
+    end = src + length;
+    for (p = src; p < end; ) {
+	p += TclUtfToUniChar(p, &ch);
+	*w++ = ch;
+    }
+    *w = '\0';
+    Tcl_DStringSetLength(dsPtr,
+	    (oldLength + ((char *) w - (char *) wString)));
+
+    return wString;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfCharComplete --
+ *
+ *	Determine if the UTF-8 string of the given length is long enough to be
+ *	decoded by Tcl_UtfToUniChar(). This does not ensure that the UTF-8
+ *	string is properly formed. Equivalent to Plan 9 fullrune().
+ *
+ * Results:
+ *	The return value is 0 if the string is not long enough, non-zero
+ *	otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfCharComplete(
+    const char *src,		/* String to check if first few bytes contain
+				 * a complete UTF-8 character. */
+    int length)			/* Length of above string in bytes. */
+{
+    return length >= totalBytes[(unsigned char)*src];
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_NumUtfChars --
+ *
+ *	Returns the number of characters (not bytes) in the UTF-8 string, not
+ *	including the terminating NULL byte. This is equivalent to Plan 9
+ *	utflen() and utfnlen().
+ *
+ * Results:
+ *	As above.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_NumUtfChars(
+    register const char *src,	/* The UTF-8 string to measure. */
+    int length)			/* The length of the string in bytes, or -1
+				 * for strlen(string). */
+{
+    Tcl_UniChar ch = 0;
+    register int i = 0;
+
+    /*
+     * The separate implementations are faster.
+     *
+     * Since this is a time-sensitive function, we also do the check for the
+     * single-byte char case specially.
+     */
+
+    if (length < 0) {
+	while (*src != '\0') {
+	    src += TclUtfToUniChar(src, &ch);
+	    i++;
+	}
+	if (i < 0) i = INT_MAX; /* Bug [2738427] */
+    } else {
+	register const char *endPtr = src + length - TCL_UTF_MAX;
+
+	while (src < endPtr) {
+	    src += TclUtfToUniChar(src, &ch);
+	    i++;
+	}
+	endPtr += TCL_UTF_MAX;
+	while ((src < endPtr) && Tcl_UtfCharComplete(src, endPtr - src)) {
+	    src += TclUtfToUniChar(src, &ch);
+	    i++;
+	}
+	if (src < endPtr) {
+	    i += endPtr - src;
+	}
+    }
+    return i;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfFindFirst --
+ *
+ *	Returns a pointer to the first occurance of the given Tcl_UniChar in
+ *	the NULL-terminated UTF-8 string. The NULL terminator is considered
+ *	part of the UTF-8 string. Equivalent to Plan 9 utfrune().
+ *
+ * Results:
+ *	As above. If the Tcl_UniChar does not exist in the given string, the
+ *	return value is NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_UtfFindFirst(
+    const char *src,		/* The UTF-8 string to be searched. */
+    int ch)			/* The Tcl_UniChar to search for. */
+{
+    int len;
+    Tcl_UniChar find = 0;
+
+    while (1) {
+	len = TclUtfToUniChar(src, &find);
+	if (find == ch) {
+	    return src;
+	}
+	if (*src == '\0') {
+	    return NULL;
+	}
+	src += len;
+    }
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfFindLast --
+ *
+ *	Returns a pointer to the last occurance of the given Tcl_UniChar in
+ *	the NULL-terminated UTF-8 string. The NULL terminator is considered
+ *	part of the UTF-8 string. Equivalent to Plan 9 utfrrune().
+ *
+ * Results:
+ *	As above. If the Tcl_UniChar does not exist in the given string, the
+ *	return value is NULL.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_UtfFindLast(
+    const char *src,		/* The UTF-8 string to be searched. */
+    int ch)			/* The Tcl_UniChar to search for. */
+{
+    int len;
+    Tcl_UniChar find = 0;
+    const char *last;
+
+    last = NULL;
+    while (1) {
+	len = TclUtfToUniChar(src, &find);
+	if (find == ch) {
+	    last = src;
+	}
+	if (*src == '\0') {
+	    break;
+	}
+	src += len;
+    }
+    return last;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfNext --
+ *
+ *	Given a pointer to some current location in a UTF-8 string, move
+ *	forward one character. The caller must ensure that they are not asking
+ *	for the next character after the last character in the string.
+ *
+ * Results:
+ *	The return value is the pointer to the next character in the UTF-8
+ *	string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_UtfNext(
+    const char *src)		/* The current location in the string. */
+{
+    Tcl_UniChar ch = 0;
+    int len = TclUtfToUniChar(src, &ch);
+
+#if TCL_UTF_MAX == 4
+    if (len == 0) {
+      len = TclUtfToUniChar(src, &ch);
+    }
+#endif
+    return src + len;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfPrev --
+ *
+ *	Given a pointer to some current location in a UTF-8 string, move
+ *	backwards one character. This works correctly when the pointer is in
+ *	the middle of a UTF-8 character.
+ *
+ * Results:
+ *	The return value is a pointer to the previous character in the UTF-8
+ *	string. If the current location was already at the beginning of the
+ *	string, the return value will also be a pointer to the beginning of
+ *	the string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_UtfPrev(
+    const char *src,		/* The current location in the string. */
+    const char *start)		/* Pointer to the beginning of the string, to
+				 * avoid going backwards too far. */
+{
+    const char *look;
+    int i, byte;
+
+    look = --src;
+    for (i = 0; i < TCL_UTF_MAX; i++) {
+	if (look < start) {
+	    if (src < start) {
+		src = start;
+	    }
+	    break;
+	}
+	byte = *((unsigned char *) look);
+	if (byte < 0x80) {
+	    break;
+	}
+	if (byte >= 0xC0) {
+	    return look;
+	}
+	look--;
+    }
+    return src;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UniCharAtIndex --
+ *
+ *	Returns the Unicode character represented at the specified character
+ *	(not byte) position in the UTF-8 string.
+ *
+ * Results:
+ *	As above.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+Tcl_UniChar
+Tcl_UniCharAtIndex(
+    register const char *src,	/* The UTF-8 string to dereference. */
+    register int index)		/* The position of the desired character. */
+{
+    Tcl_UniChar ch = 0;
+
+    while (index >= 0) {
+	index--;
+	src += TclUtfToUniChar(src, &ch);
+    }
+    return ch;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfAtIndex --
+ *
+ *	Returns a pointer to the specified character (not byte) position in
+ *	the UTF-8 string.
+ *
+ * Results:
+ *	As above.
+ *
+ * Side effects:
+ *	None.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+const char *
+Tcl_UtfAtIndex(
+    register const char *src,	/* The UTF-8 string. */
+    register int index)		/* The position of the desired character. */
+{
+    Tcl_UniChar ch = 0;
+
+    while (index > 0) {
+	index--;
+	src += TclUtfToUniChar(src, &ch);
+    }
+    return src;
+}
+
+/*
+ *---------------------------------------------------------------------------
+ *
+ * Tcl_UtfBackslash --
+ *
+ *	Figure out how to handle a backslash sequence.
+ *
+ * Results:
+ *	Stores the bytes represented by the backslash sequence in dst and
+ *	returns the number of bytes written to dst. At most TCL_UTF_MAX bytes
+ *	are written to dst; dst must have been large enough to accept those
+ *	bytes. If readPtr isn't NULL then it is filled in with a count of the
+ *	number of bytes in the backslash sequence.
+ *
+ * Side effects:
+ *	The maximum number of bytes it takes to represent a Unicode character
+ *	in UTF-8 is guaranteed to be less than the number of bytes used to
+ *	express the backslash sequence that represents that Unicode character.
+ *	If the target buffer into which the caller is going to store the bytes
+ *	that represent the Unicode character is at least as large as the
+ *	source buffer from which the backslashed sequence was extracted, no
+ *	buffer overruns should occur.
+ *
+ *---------------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfBackslash(
+    const char *src,		/* Points to the backslash character of a
+				 * backslash sequence. */
+    int *readPtr,		/* Fill in with number of characters read from
+				 * src, unless NULL. */
+    char *dst)			/* Filled with the bytes represented by the
+				 * backslash sequence. */
+{
+#define LINE_LENGTH 128
+    int numRead;
+    int result;
+
+    result = TclParseBackslash(src, LINE_LENGTH, &numRead, dst);
+    if (numRead == LINE_LENGTH) {
+	/*
+	 * We ate a whole line. Pay the price of a strlen()
+	 */
+
+	result = TclParseBackslash(src, (int)strlen(src), &numRead, dst);
+    }
+    if (readPtr != NULL) {
+	*readPtr = numRead;
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfToUpper --
+ *
+ *	Convert lowercase characters to uppercase characters in a UTF string
+ *	in place. The conversion may shrink the UTF string.
+ *
+ * Results:
+ *	Returns the number of bytes in the resulting string excluding the
+ *	trailing null.
+ *
+ * Side effects:
+ *	Writes a terminating null after the last converted character.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfToUpper(
+    char *str)			/* String to convert in place. */
+{
+    Tcl_UniChar ch = 0, upChar;
+    char *src, *dst;
+    int bytes;
+
+    /*
+     * Iterate over the string until we hit the terminating null.
+     */
+
+    src = dst = str;
+    while (*src) {
+	bytes = TclUtfToUniChar(src, &ch);
+	upChar = Tcl_UniCharToUpper(ch);
+
+	/*
+	 * To keep badly formed Utf strings from getting inflated by the
+	 * conversion (thereby causing a segfault), only copy the upper case
+	 * char to dst if its size is <= the original char.
+	 */
+
+	if (bytes < TclUtfCount(upChar)) {
+	    memcpy(dst, src, (size_t) bytes);
+	    dst += bytes;
+	} else {
+	    dst += Tcl_UniCharToUtf(upChar, dst);
+	}
+	src += bytes;
+    }
+    *dst = '\0';
+    return (dst - str);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfToLower --
+ *
+ *	Convert uppercase characters to lowercase characters in a UTF string
+ *	in place. The conversion may shrink the UTF string.
+ *
+ * Results:
+ *	Returns the number of bytes in the resulting string excluding the
+ *	trailing null.
+ *
+ * Side effects:
+ *	Writes a terminating null after the last converted character.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfToLower(
+    char *str)			/* String to convert in place. */
+{
+    Tcl_UniChar ch = 0, lowChar;
+    char *src, *dst;
+    int bytes;
+
+    /*
+     * Iterate over the string until we hit the terminating null.
+     */
+
+    src = dst = str;
+    while (*src) {
+	bytes = TclUtfToUniChar(src, &ch);
+	lowChar = Tcl_UniCharToLower(ch);
+
+	/*
+	 * To keep badly formed Utf strings from getting inflated by the
+	 * conversion (thereby causing a segfault), only copy the lower case
+	 * char to dst if its size is <= the original char.
+	 */
+
+	if (bytes < TclUtfCount(lowChar)) {
+	    memcpy(dst, src, (size_t) bytes);
+	    dst += bytes;
+	} else {
+	    dst += Tcl_UniCharToUtf(lowChar, dst);
+	}
+	src += bytes;
+    }
+    *dst = '\0';
+    return (dst - str);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfToTitle --
+ *
+ *	Changes the first character of a UTF string to title case or uppercase
+ *	and the rest of the string to lowercase. The conversion happens in
+ *	place and may shrink the UTF string.
+ *
+ * Results:
+ *	Returns the number of bytes in the resulting string excluding the
+ *	trailing null.
+ *
+ * Side effects:
+ *	Writes a terminating null after the last converted character.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfToTitle(
+    char *str)			/* String to convert in place. */
+{
+    Tcl_UniChar ch = 0, titleChar, lowChar;
+    char *src, *dst;
+    int bytes;
+
+    /*
+     * Capitalize the first character and then lowercase the rest of the
+     * characters until we get to a null.
+     */
+
+    src = dst = str;
+
+    if (*src) {
+	bytes = TclUtfToUniChar(src, &ch);
+	titleChar = Tcl_UniCharToTitle(ch);
+
+	if (bytes < TclUtfCount(titleChar)) {
+	    memcpy(dst, src, (size_t) bytes);
+	    dst += bytes;
+	} else {
+	    dst += Tcl_UniCharToUtf(titleChar, dst);
+	}
+	src += bytes;
+    }
+    while (*src) {
+	bytes = TclUtfToUniChar(src, &ch);
+	lowChar = Tcl_UniCharToLower(ch);
+
+	if (bytes < TclUtfCount(lowChar)) {
+	    memcpy(dst, src, (size_t) bytes);
+	    dst += bytes;
+	} else {
+	    dst += Tcl_UniCharToUtf(lowChar, dst);
+	}
+	src += bytes;
+    }
+    *dst = '\0';
+    return (dst - str);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpUtfNcmp2 --
+ *
+ *	Compare at most numBytes bytes of utf-8 strings cs and ct. Both cs and
+ *	ct are assumed to be at least numBytes bytes long.
+ *
+ * Results:
+ *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclpUtfNcmp2(
+    const char *cs,		/* UTF string to compare to ct. */
+    const char *ct,		/* UTF string cs is compared to. */
+    unsigned long numBytes)	/* Number of *bytes* to compare. */
+{
+    /*
+     * We can't simply call 'memcmp(cs, ct, numBytes);' because we need to
+     * check for Tcl's \xC0\x80 non-utf-8 null encoding. Otherwise utf-8 lexes
+     * fine in the strcmp manner.
+     */
+
+    register int result = 0;
+
+    for ( ; numBytes != 0; numBytes--, cs++, ct++) {
+	if (*cs != *ct) {
+	    result = UCHAR(*cs) - UCHAR(*ct);
+	    break;
+	}
+    }
+    if (numBytes && ((UCHAR(*cs) == 0xC0) || (UCHAR(*ct) == 0xC0))) {
+	unsigned char c1, c2;
+
+	c1 = ((UCHAR(*cs) == 0xC0) && (UCHAR(cs[1]) == 0x80)) ? 0 : UCHAR(*cs);
+	c2 = ((UCHAR(*ct) == 0xC0) && (UCHAR(ct[1]) == 0x80)) ? 0 : UCHAR(*ct);
+	result = (c1 - c2);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfNcmp --
+ *
+ *	Compare at most numChars UTF chars of string cs to string ct. Both cs
+ *	and ct are assumed to be at least numChars UTF chars long.
+ *
+ * Results:
+ *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfNcmp(
+    const char *cs,		/* UTF string to compare to ct. */
+    const char *ct,		/* UTF string cs is compared to. */
+    unsigned long numChars)	/* Number of UTF chars to compare. */
+{
+    Tcl_UniChar ch1 = 0, ch2 = 0;
+
+    /*
+     * Cannot use 'memcmp(cs, ct, n);' as byte representation of \u0000 (the
+     * pair of bytes 0xC0,0x80) is larger than byte representation of \u0001
+     * (the byte 0x01.)
+     */
+
+    while (numChars-- > 0) {
+	/*
+	 * n must be interpreted as chars, not bytes. This should be called
+	 * only when both strings are of at least n chars long (no need for \0
+	 * check)
+	 */
+
+	cs += TclUtfToUniChar(cs, &ch1);
+	ct += TclUtfToUniChar(ct, &ch2);
+	if (ch1 != ch2) {
+	    return (ch1 - ch2);
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfNcasecmp --
+ *
+ *	Compare at most numChars UTF chars of string cs to string ct case
+ *	insensitive. Both cs and ct are assumed to be at least numChars UTF
+ *	chars long.
+ *
+ * Results:
+ *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UtfNcasecmp(
+    const char *cs,		/* UTF string to compare to ct. */
+    const char *ct,		/* UTF string cs is compared to. */
+    unsigned long numChars)	/* Number of UTF chars to compare. */
+{
+    Tcl_UniChar ch1 = 0, ch2 = 0;
+    while (numChars-- > 0) {
+	/*
+	 * n must be interpreted as chars, not bytes.
+	 * This should be called only when both strings are of
+	 * at least n chars long (no need for \0 check)
+	 */
+	cs += TclUtfToUniChar(cs, &ch1);
+	ct += TclUtfToUniChar(ct, &ch2);
+	if (ch1 != ch2) {
+	    ch1 = Tcl_UniCharToLower(ch1);
+	    ch2 = Tcl_UniCharToLower(ch2);
+	    if (ch1 != ch2) {
+		return (ch1 - ch2);
+	    }
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UtfNcasecmp --
+ *
+ *	Compare UTF chars of string cs to string ct case insensitively.
+ *	Replacement for strcasecmp in Tcl core, in places where UTF-8 should
+ *	be handled.
+ *
+ * Results:
+ *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclUtfCasecmp(
+    const char *cs,		/* UTF string to compare to ct. */
+    const char *ct)		/* UTF string cs is compared to. */
+{
+    while (*cs && *ct) {
+	Tcl_UniChar ch1, ch2;
+
+	cs += TclUtfToUniChar(cs, &ch1);
+	ct += TclUtfToUniChar(ct, &ch2);
+	if (ch1 != ch2) {
+	    ch1 = Tcl_UniCharToLower(ch1);
+	    ch2 = Tcl_UniCharToLower(ch2);
+	    if (ch1 != ch2) {
+		return ch1 - ch2;
+	    }
+	}
+    }
+    return UCHAR(*cs) - UCHAR(*ct);
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharToUpper --
+ *
+ *	Compute the uppercase equivalent of the given Unicode character.
+ *
+ * Results:
+ *	Returns the uppercase Unicode character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar
+Tcl_UniCharToUpper(
+    int ch)			/* Unicode character to convert. */
+{
+    int info = GetUniCharInfo(ch);
+
+    if (GetCaseType(info) & 0x04) {
+	ch -= GetDelta(info);
+    }
+    return (Tcl_UniChar) ch;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharToLower --
+ *
+ *	Compute the lowercase equivalent of the given Unicode character.
+ *
+ * Results:
+ *	Returns the lowercase Unicode character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar
+Tcl_UniCharToLower(
+    int ch)			/* Unicode character to convert. */
+{
+    int info = GetUniCharInfo(ch);
+
+    if (GetCaseType(info) & 0x02) {
+	ch += GetDelta(info);
+    }
+    return (Tcl_UniChar) ch;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharToTitle --
+ *
+ *	Compute the titlecase equivalent of the given Unicode character.
+ *
+ * Results:
+ *	Returns the titlecase Unicode character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_UniChar
+Tcl_UniCharToTitle(
+    int ch)			/* Unicode character to convert. */
+{
+    int info = GetUniCharInfo(ch);
+    int mode = GetCaseType(info);
+
+    if (mode & 0x1) {
+	/*
+	 * Subtract or add one depending on the original case.
+	 */
+
+	ch += ((mode & 0x4) ? -1 : 1);
+    } else if (mode == 0x4) {
+	ch -= GetDelta(info);
+    }
+    return (Tcl_UniChar) ch;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharLen --
+ *
+ *	Find the length of a UniChar string. The str input must be null
+ *	terminated.
+ *
+ * Results:
+ *	Returns the length of str in UniChars (not bytes).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharLen(
+    const Tcl_UniChar *uniStr)	/* Unicode string to find length of. */
+{
+    int len = 0;
+
+    while (*uniStr != '\0') {
+	len++;
+	uniStr++;
+    }
+    return len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharNcmp --
+ *
+ *	Compare at most numChars unichars of string ucs to string uct.
+ *	Both ucs and uct are assumed to be at least numChars unichars long.
+ *
+ * Results:
+ *	Return <0 if ucs < uct, 0 if ucs == uct, or >0 if ucs > uct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharNcmp(
+    const Tcl_UniChar *ucs,	/* Unicode string to compare to uct. */
+    const Tcl_UniChar *uct,	/* Unicode string ucs is compared to. */
+    unsigned long numChars)	/* Number of unichars to compare. */
+{
+#ifdef WORDS_BIGENDIAN
+    /*
+     * We are definitely on a big-endian machine; memcmp() is safe
+     */
+
+    return memcmp(ucs, uct, numChars*sizeof(Tcl_UniChar));
+
+#else /* !WORDS_BIGENDIAN */
+    /*
+     * We can't simply call memcmp() because that is not lexically correct.
+     */
+
+    for ( ; numChars != 0; ucs++, uct++, numChars--) {
+	if (*ucs != *uct) {
+	    return (*ucs - *uct);
+	}
+    }
+    return 0;
+#endif /* WORDS_BIGENDIAN */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharNcasecmp --
+ *
+ *	Compare at most numChars unichars of string ucs to string uct case
+ *	insensitive. Both ucs and uct are assumed to be at least numChars
+ *	unichars long.
+ *
+ * Results:
+ *	Return <0 if ucs < uct, 0 if ucs == uct, or >0 if ucs > uct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharNcasecmp(
+    const Tcl_UniChar *ucs,	/* Unicode string to compare to uct. */
+    const Tcl_UniChar *uct,	/* Unicode string ucs is compared to. */
+    unsigned long numChars)	/* Number of unichars to compare. */
+{
+    for ( ; numChars != 0; numChars--, ucs++, uct++) {
+	if (*ucs != *uct) {
+	    Tcl_UniChar lcs = Tcl_UniCharToLower(*ucs);
+	    Tcl_UniChar lct = Tcl_UniCharToLower(*uct);
+
+	    if (lcs != lct) {
+		return (lcs - lct);
+	    }
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsAlnum --
+ *
+ *	Test if a character is an alphanumeric Unicode character.
+ *
+ * Results:
+ *	Returns 1 if character is alphanumeric.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsAlnum(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return (((ALPHA_BITS | DIGIT_BITS) >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsAlpha --
+ *
+ *	Test if a character is an alphabetic Unicode character.
+ *
+ * Results:
+ *	Returns 1 if character is alphabetic.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsAlpha(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return ((ALPHA_BITS >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsControl --
+ *
+ *	Test if a character is a Unicode control character.
+ *
+ * Results:
+ *	Returns non-zero if character is a control.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsControl(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	ch &= 0x1FFFFF;
+	if ((ch == 0xE0001) || ((ch >= 0xE0020) && (ch <= 0xE007f))) {
+	    return 1;
+	}
+	if ((ch >= 0xF0000) && ((ch & 0xFFFF) <= 0xFFFD)) {
+	    return 1;
+	}
+	return 0;
+    }
+#endif
+    return ((CONTROL_BITS >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsDigit --
+ *
+ *	Test if a character is a numeric Unicode character.
+ *
+ * Results:
+ *	Returns non-zero if character is a digit.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsDigit(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return (GetCategory(ch) == DECIMAL_DIGIT_NUMBER);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsGraph --
+ *
+ *	Test if a character is any Unicode print character except space.
+ *
+ * Results:
+ *	Returns non-zero if character is printable, but not space.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsGraph(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	ch &= 0x1FFFFF;
+	return (ch >= 0xE0100) && (ch <= 0xE01EF);
+    }
+#endif
+    return ((GRAPH_BITS >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsLower --
+ *
+ *	Test if a character is a lowercase Unicode character.
+ *
+ * Results:
+ *	Returns non-zero if character is lowercase.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsLower(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return (GetCategory(ch) == LOWERCASE_LETTER);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsPrint --
+ *
+ *	Test if a character is a Unicode print character.
+ *
+ * Results:
+ *	Returns non-zero if character is printable.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsPrint(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	ch &= 0x1FFFFF;
+	return (ch >= 0xE0100) && (ch <= 0xE01EF);
+    }
+#endif
+    return (((GRAPH_BITS|SPACE_BITS) >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsPunct --
+ *
+ *	Test if a character is a Unicode punctuation character.
+ *
+ * Results:
+ *	Returns non-zero if character is punct.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsPunct(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return ((PUNCT_BITS >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsSpace --
+ *
+ *	Test if a character is a whitespace Unicode character.
+ *
+ * Results:
+ *	Returns non-zero if character is a space.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsSpace(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    /* Ignore upper 11 bits. */
+    ch &= 0x1FFFFF;
+#else
+    /* Ignore upper 16 bits. */
+    ch &= 0xFFFF;
+#endif
+
+    /*
+     * If the character is within the first 127 characters, just use the
+     * standard C function, otherwise consult the Unicode table.
+     */
+
+    if (ch < 0x80) {
+	return TclIsSpaceProc((char) ch);
+#if TCL_UTF_MAX > 3
+    } else if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+#endif
+    } else if (ch == 0x0085 || ch == 0x180E || ch == 0x200B
+	    || ch == 0x202F || ch == 0x2060 || ch == 0xFEFF) {
+	return 1;
+    } else {
+	return ((SPACE_BITS >> GetCategory(ch)) & 1);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsUpper --
+ *
+ *	Test if a character is a uppercase Unicode character.
+ *
+ * Results:
+ *	Returns non-zero if character is uppercase.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsUpper(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return (GetCategory(ch) == UPPERCASE_LETTER);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharIsWordChar --
+ *
+ *	Test if a character is alphanumeric or a connector punctuation mark.
+ *
+ * Results:
+ *	Returns 1 if character is a word character.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharIsWordChar(
+    int ch)			/* Unicode character to test. */
+{
+#if TCL_UTF_MAX > 3
+    if (UNICODE_OUT_OF_RANGE(ch)) {
+	return 0;
+    }
+#endif
+    return ((WORD_BITS >> GetCategory(ch)) & 1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UniCharCaseMatch --
+ *
+ *	See if a particular Unicode string matches a particular pattern.
+ *	Allows case insensitivity. This is the Unicode equivalent of the char*
+ *	Tcl_StringCaseMatch. The UniChar strings must be NULL-terminated.
+ *	This has no provision for counted UniChar strings, thus should not be
+ *	used where NULLs are expected in the UniChar string. Use
+ *	TclUniCharMatch where possible.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UniCharCaseMatch(
+    const Tcl_UniChar *uniStr,	/* Unicode String. */
+    const Tcl_UniChar *uniPattern,
+				/* Pattern, which may contain special
+				 * characters. */
+    int nocase)			/* 0 for case sensitive, 1 for insensitive */
+{
+    Tcl_UniChar ch1 = 0, p;
+
+    while (1) {
+	p = *uniPattern;
+
+	/*
+	 * See if we're at the end of both the pattern and the string. If so,
+	 * we succeeded. If we're at the end of the pattern but not at the end
+	 * of the string, we failed.
+	 */
+
+	if (p == 0) {
+	    return (*uniStr == 0);
+	}
+	if ((*uniStr == 0) && (p != '*')) {
+	    return 0;
+	}
+
+	/*
+	 * Check for a "*" as the next pattern character. It matches any
+	 * substring. We handle this by skipping all the characters up to the
+	 * next matching one in the pattern, and then calling ourselves
+	 * recursively for each postfix of string, until either we match or we
+	 * reach the end of the string.
+	 */
+
+	if (p == '*') {
+	    /*
+	     * Skip all successive *'s in the pattern
+	     */
+
+	    while (*(++uniPattern) == '*') {
+		/* empty body */
+	    }
+	    p = *uniPattern;
+	    if (p == 0) {
+		return 1;
+	    }
+	    if (nocase) {
+		p = Tcl_UniCharToLower(p);
+	    }
+	    while (1) {
+		/*
+		 * Optimization for matching - cruise through the string
+		 * quickly if the next char in the pattern isn't a special
+		 * character
+		 */
+
+		if ((p != '[') && (p != '?') && (p != '\\')) {
+		    if (nocase) {
+			while (*uniStr && (p != *uniStr)
+				&& (p != Tcl_UniCharToLower(*uniStr))) {
+			    uniStr++;
+			}
+		    } else {
+			while (*uniStr && (p != *uniStr)) {
+			    uniStr++;
+			}
+		    }
+		}
+		if (Tcl_UniCharCaseMatch(uniStr, uniPattern, nocase)) {
+		    return 1;
+		}
+		if (*uniStr == 0) {
+		    return 0;
+		}
+		uniStr++;
+	    }
+	}
+
+	/*
+	 * Check for a "?" as the next pattern character. It matches any
+	 * single character.
+	 */
+
+	if (p == '?') {
+	    uniPattern++;
+	    uniStr++;
+	    continue;
+	}
+
+	/*
+	 * Check for a "[" as the next pattern character. It is followed by a
+	 * list of characters that are acceptable, or by a range (two
+	 * characters separated by "-").
+	 */
+
+	if (p == '[') {
+	    Tcl_UniChar startChar, endChar;
+
+	    uniPattern++;
+	    ch1 = (nocase ? Tcl_UniCharToLower(*uniStr) : *uniStr);
+	    uniStr++;
+	    while (1) {
+		if ((*uniPattern == ']') || (*uniPattern == 0)) {
+		    return 0;
+		}
+		startChar = (nocase ? Tcl_UniCharToLower(*uniPattern)
+			: *uniPattern);
+		uniPattern++;
+		if (*uniPattern == '-') {
+		    uniPattern++;
+		    if (*uniPattern == 0) {
+			return 0;
+		    }
+		    endChar = (nocase ? Tcl_UniCharToLower(*uniPattern)
+			    : *uniPattern);
+		    uniPattern++;
+		    if (((startChar <= ch1) && (ch1 <= endChar))
+			    || ((endChar <= ch1) && (ch1 <= startChar))) {
+			/*
+			 * Matches ranges of form [a-z] or [z-a].
+			 */
+			break;
+		    }
+		} else if (startChar == ch1) {
+		    break;
+		}
+	    }
+	    while (*uniPattern != ']') {
+		if (*uniPattern == 0) {
+		    uniPattern--;
+		    break;
+		}
+		uniPattern++;
+	    }
+	    uniPattern++;
+	    continue;
+	}
+
+	/*
+	 * If the next pattern character is '\', just strip off the '\' so we
+	 * do exact matching on the character that follows.
+	 */
+
+	if (p == '\\') {
+	    if (*(++uniPattern) == '\0') {
+		return 0;
+	    }
+	}
+
+	/*
+	 * There's no special character. Just make sure that the next bytes of
+	 * each string match.
+	 */
+
+	if (nocase) {
+	    if (Tcl_UniCharToLower(*uniStr) !=
+		    Tcl_UniCharToLower(*uniPattern)) {
+		return 0;
+	    }
+	} else if (*uniStr != *uniPattern) {
+	    return 0;
+	}
+	uniStr++;
+	uniPattern++;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclUniCharMatch --
+ *
+ *	See if a particular Unicode string matches a particular pattern.
+ *	Allows case insensitivity. This is the Unicode equivalent of the char*
+ *	Tcl_StringCaseMatch. This variant of Tcl_UniCharCaseMatch uses counted
+ *	Strings, so embedded NULLs are allowed.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclUniCharMatch(
+    const Tcl_UniChar *string,	/* Unicode String. */
+    int strLen,			/* Length of String */
+    const Tcl_UniChar *pattern,	/* Pattern, which may contain special
+				 * characters. */
+    int ptnLen,			/* Length of Pattern */
+    int nocase)			/* 0 for case sensitive, 1 for insensitive */
+{
+    const Tcl_UniChar *stringEnd, *patternEnd;
+    Tcl_UniChar p;
+
+    stringEnd = string + strLen;
+    patternEnd = pattern + ptnLen;
+
+    while (1) {
+	/*
+	 * See if we're at the end of both the pattern and the string. If so,
+	 * we succeeded. If we're at the end of the pattern but not at the end
+	 * of the string, we failed.
+	 */
+
+	if (pattern == patternEnd) {
+	    return (string == stringEnd);
+	}
+	p = *pattern;
+	if ((string == stringEnd) && (p != '*')) {
+	    return 0;
+	}
+
+	/*
+	 * Check for a "*" as the next pattern character. It matches any
+	 * substring. We handle this by skipping all the characters up to the
+	 * next matching one in the pattern, and then calling ourselves
+	 * recursively for each postfix of string, until either we match or we
+	 * reach the end of the string.
+	 */
+
+	if (p == '*') {
+	    /*
+	     * Skip all successive *'s in the pattern.
+	     */
+
+	    while (*(++pattern) == '*') {
+		/* empty body */
+	    }
+	    if (pattern == patternEnd) {
+		return 1;
+	    }
+	    p = *pattern;
+	    if (nocase) {
+		p = Tcl_UniCharToLower(p);
+	    }
+	    while (1) {
+		/*
+		 * Optimization for matching - cruise through the string
+		 * quickly if the next char in the pattern isn't a special
+		 * character.
+		 */
+
+		if ((p != '[') && (p != '?') && (p != '\\')) {
+		    if (nocase) {
+			while ((string < stringEnd) && (p != *string)
+				&& (p != Tcl_UniCharToLower(*string))) {
+			    string++;
+			}
+		    } else {
+			while ((string < stringEnd) && (p != *string)) {
+			    string++;
+			}
+		    }
+		}
+		if (TclUniCharMatch(string, stringEnd - string,
+			pattern, patternEnd - pattern, nocase)) {
+		    return 1;
+		}
+		if (string == stringEnd) {
+		    return 0;
+		}
+		string++;
+	    }
+	}
+
+	/*
+	 * Check for a "?" as the next pattern character. It matches any
+	 * single character.
+	 */
+
+	if (p == '?') {
+	    pattern++;
+	    string++;
+	    continue;
+	}
+
+	/*
+	 * Check for a "[" as the next pattern character. It is followed by a
+	 * list of characters that are acceptable, or by a range (two
+	 * characters separated by "-").
+	 */
+
+	if (p == '[') {
+	    Tcl_UniChar ch1, startChar, endChar;
+
+	    pattern++;
+	    ch1 = (nocase ? Tcl_UniCharToLower(*string) : *string);
+	    string++;
+	    while (1) {
+		if ((*pattern == ']') || (pattern == patternEnd)) {
+		    return 0;
+		}
+		startChar = (nocase ? Tcl_UniCharToLower(*pattern) : *pattern);
+		pattern++;
+		if (*pattern == '-') {
+		    pattern++;
+		    if (pattern == patternEnd) {
+			return 0;
+		    }
+		    endChar = (nocase ? Tcl_UniCharToLower(*pattern)
+			    : *pattern);
+		    pattern++;
+		    if (((startChar <= ch1) && (ch1 <= endChar))
+			    || ((endChar <= ch1) && (ch1 <= startChar))) {
+			/*
+			 * Matches ranges of form [a-z] or [z-a].
+			 */
+			break;
+		    }
+		} else if (startChar == ch1) {
+		    break;
+		}
+	    }
+	    while (*pattern != ']') {
+		if (pattern == patternEnd) {
+		    pattern--;
+		    break;
+		}
+		pattern++;
+	    }
+	    pattern++;
+	    continue;
+	}
+
+	/*
+	 * If the next pattern character is '\', just strip off the '\' so we
+	 * do exact matching on the character that follows.
+	 */
+
+	if (p == '\\') {
+	    if (++pattern == patternEnd) {
+		return 0;
+	    }
+	}
+
+	/*
+	 * There's no special character. Just make sure that the next bytes of
+	 * each string match.
+	 */
+
+	if (nocase) {
+	    if (Tcl_UniCharToLower(*string) != Tcl_UniCharToLower(*pattern)) {
+		return 0;
+	    }
+	} else if (*string != *pattern) {
+	    return 0;
+	}
+	string++;
+	pattern++;
+    }
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclUtil.c b/generic/tclUtil.c
new file mode 100644
index 0000000..608cd15
--- /dev/null
+++ b/generic/tclUtil.c
@@ -0,0 +1,4493 @@
+/*
+ * tclUtil.c --
+ *
+ *	This file contains utility functions that are used by many Tcl
+ *	commands.
+ *
+ * Copyright (c) 1987-1993 The Regents of the University of California.
+ * Copyright (c) 1994-1998 Sun Microsystems, Inc.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclParse.h"
+#include "tclStringTrim.h"
+#include <math.h>
+
+/*
+ * The absolute pathname of the executable in which this Tcl library is
+ * running.
+ */
+
+static ProcessGlobalValue executableName = {
+    0, 0, NULL, NULL, NULL, NULL, NULL
+};
+
+/*
+ * The following values are used in the flags arguments of Tcl*Scan*Element
+ * and Tcl*Convert*Element.  The values TCL_DONT_USE_BRACES and
+ * TCL_DONT_QUOTE_HASH are defined in tcl.h, like so:
+ *
+#define TCL_DONT_USE_BRACES     1
+#define TCL_DONT_QUOTE_HASH     8
+ *
+ * Those are public flag bits which callers of the public routines
+ * Tcl_Convert*Element() can use to indicate:
+ *
+ * TCL_DONT_USE_BRACES -	1 means the caller is insisting that brace
+ *				quoting not be used when converting the list
+ *				element.
+ * TCL_DONT_QUOTE_HASH -	1 means the caller insists that a leading hash
+ *				character ('#') should *not* be quoted. This
+ *				is appropriate when the caller can guarantee
+ *				the element is not the first element of a
+ *				list, so [eval] cannot mis-parse the element
+ *				as a comment.
+ *
+ * The remaining values which can be carried by the flags of these routines
+ * are for internal use only.  Make sure they do not overlap with the public
+ * values above.
+ *
+ * The Tcl*Scan*Element() routines make a determination which of 4 modes of
+ * conversion is most appropriate for Tcl*Convert*Element() to perform, and
+ * sets two bits of the flags value to indicate the mode selected.
+ *
+ * CONVERT_NONE		The element needs no quoting. Its literal string is
+ *			suitable as is.
+ * CONVERT_BRACE	The conversion should be enclosing the literal string
+ *			in braces.
+ * CONVERT_ESCAPE	The conversion should be using backslashes to escape
+ *			any characters in the string that require it.
+ * CONVERT_MASK		A mask value used to extract the conversion mode from
+ *			the flags argument.
+ *			Also indicates a strange conversion mode where all
+ *			special characters are escaped with backslashes
+ *			*except for braces*. This is a strange and unnecessary
+ *			case, but it's part of the historical way in which
+ *			lists have been formatted in Tcl. To experiment with
+ *			removing this case, set the value of COMPAT to 0.
+ *
+ * One last flag value is used only by callers of TclScanElement(). The flag
+ * value produced by a call to Tcl*Scan*Element() will never leave this bit
+ * set.
+ *
+ * CONVERT_ANY		The caller of TclScanElement() declares it can make no
+ *			promise about what public flags will be passed to the
+ *			matching call of TclConvertElement(). As such,
+ *			TclScanElement() has to determine the worst case
+ *			destination buffer length over all possibilities, and
+ *			in other cases this means an overestimate of the
+ *			required size.
+ *
+ * For more details, see the comments on the Tcl*Scan*Element and
+ * Tcl*Convert*Element routines.
+ */
+
+#define COMPAT 1
+#define CONVERT_NONE	0
+#define CONVERT_BRACE	2
+#define CONVERT_ESCAPE	4
+#define CONVERT_MASK	(CONVERT_BRACE | CONVERT_ESCAPE)
+#define CONVERT_ANY	16
+
+/*
+ * The following key is used by Tcl_PrintDouble and TclPrecTraceProc to
+ * access the precision to be used for double formatting.
+ */
+
+static Tcl_ThreadDataKey precisionKey;
+
+/*
+ * Prototypes for functions defined later in this file.
+ */
+
+static void		ClearHash(Tcl_HashTable *tablePtr);
+static void		FreeProcessGlobalValue(ClientData clientData);
+static void		FreeThreadHash(ClientData clientData);
+static Tcl_HashTable *	GetThreadHash(Tcl_ThreadDataKey *keyPtr);
+static int		SetEndOffsetFromAny(Tcl_Interp *interp,
+			    Tcl_Obj *objPtr);
+static void		UpdateStringOfEndOffset(Tcl_Obj *objPtr);
+static int		FindElement(Tcl_Interp *interp, const char *string,
+			    int stringLength, const char *typeStr,
+			    const char *typeCode, const char **elementPtr,
+			    const char **nextPtr, int *sizePtr,
+			    int *literalPtr);
+/*
+ * The following is the Tcl object type definition for an object that
+ * represents a list index in the form, "end-offset". It is used as a
+ * performance optimization in TclGetIntForIndex. The internal rep is an
+ * integer, so no memory management is required for it.
+ */
+
+const Tcl_ObjType tclEndOffsetType = {
+    "end-offset",			/* name */
+    NULL,				/* freeIntRepProc */
+    NULL,				/* dupIntRepProc */
+    UpdateStringOfEndOffset,		/* updateStringProc */
+    SetEndOffsetFromAny
+};
+
+/*
+ *	*	STRING REPRESENTATION OF LISTS	*	*	*
+ *
+ * The next several routines implement the conversions of strings to and from
+ * Tcl lists. To understand their operation, the rules of parsing and
+ * generating the string representation of lists must be known.  Here we
+ * describe them in one place.
+ *
+ * A list is made up of zero or more elements. Any string is a list if it is
+ * made up of alternating substrings of element-separating ASCII whitespace
+ * and properly formatted elements.
+ *
+ * The ASCII characters which can make up the whitespace between list elements
+ * are:
+ *
+ *	\u0009	\t	TAB
+ *	\u000A	\n	NEWLINE
+ *	\u000B	\v	VERTICAL TAB
+ *	\u000C	\f	FORM FEED
+ * 	\u000D	\r	CARRIAGE RETURN
+ *	\u0020		SPACE
+ *
+ * NOTE: differences between this and other places where Tcl defines a role
+ * for "whitespace".
+ *
+ *	* Unlike command parsing, here NEWLINE is just another whitespace
+ *	  character; its role as a command terminator in a script has no
+ *	  importance here.
+ *
+ *	* Unlike command parsing, the BACKSLASH NEWLINE sequence is not
+ *	  considered to be a whitespace character.
+ *
+ *	* Other Unicode whitespace characters (recognized by [string is space]
+ *	  or Tcl_UniCharIsSpace()) do not play any role as element separators
+ *	  in Tcl lists.
+ *
+ *	* The NUL byte ought not appear, as it is not in strings properly
+ *	  encoded for Tcl, but if it is present, it is not treated as
+ *	  separating whitespace, or a string terminator. It is just another
+ *	  character in a list element.
+ *
+ * The interpretation of a formatted substring as a list element follows rules
+ * similar to the parsing of the words of a command in a Tcl script. Backslash
+ * substitution plays a key role, and is defined exactly as it is in command
+ * parsing. The same routine, TclParseBackslash() is used in both command
+ * parsing and list parsing.
+ *
+ * NOTE: This means that if and when backslash substitution rules ever change
+ * for command parsing, the interpretation of strings as lists also changes.
+ *
+ * Backslash substitution replaces an "escape sequence" of one or more
+ * characters starting with
+ *		\u005c	\	BACKSLASH
+ * with a single character. The one character escape sequence case happens only
+ * when BACKSLASH is the last character in the string. In all other cases, the
+ * escape sequence is at least two characters long.
+ *
+ * The formatted substrings are interpreted as element values according to the
+ * following cases:
+ *
+ * * If the first character of a formatted substring is
+ *		\u007b	{	OPEN BRACE
+ *   then the end of the substring is the matching
+ *		\u007d	}	CLOSE BRACE
+ *   character, where matching is determined by counting nesting levels, and
+ *   not including any brace characters that are contained within a backslash
+ *   escape sequence in the nesting count. Having found the matching brace,
+ *   all characters between the braces are the string value of the element.
+ *   If no matching close brace is found before the end of the string, the
+ *   string is not a Tcl list. If the character following the close brace is
+ *   not an element separating whitespace character, or the end of the string,
+ *   then the string is not a Tcl list.
+ *
+ *   NOTE: this differs from a brace-quoted word in the parsing of a Tcl
+ *   command only in its treatment of the backslash-newline sequence. In a
+ *   list element, the literal characters in the backslash-newline sequence
+ *   become part of the element value. In a script word, conversion to a
+ *   single SPACE character is done.
+ *
+ *   NOTE: Most list element values can be represented by a formatted
+ *   substring using brace quoting. The exceptions are any element value that
+ *   includes an unbalanced brace not in a backslash escape sequence, and any
+ *   value that ends with a backslash not itself in a backslash escape
+ *   sequence.
+ *
+ * * If the first character of a formatted substring is
+ *		\u0022	"	QUOTE
+ *   then the end of the substring is the next QUOTE character, not counting
+ *   any QUOTE characters that are contained within a backslash escape
+ *   sequence. If no next QUOTE is found before the end of the string, the
+ *   string is not a Tcl list. If the character following the closing QUOTE is
+ *   not an element separating whitespace character, or the end of the string,
+ *   then the string is not a Tcl list. Having found the limits of the
+ *   substring, the element value is produced by performing backslash
+ *   substitution on the character sequence between the open and close QUOTEs.
+ *
+ *   NOTE: Any element value can be represented by this style of formatting,
+ *   given suitable choice of backslash escape sequences.
+ *
+ * * All other formatted substrings are terminated by the next element
+ *   separating whitespace character in the string.  Having found the limits
+ *   of the substring, the element value is produced by performing backslash
+ *   substitution on it.
+ *
+ *   NOTE: Any element value can be represented by this style of formatting,
+ *   given suitable choice of backslash escape sequences, with one exception.
+ *   The empty string cannot be represented as a list element without the use
+ *   of either braces or quotes to delimit it.
+ *
+ * This collection of parsing rules is implemented in the routine
+ * FindElement().
+ *
+ * In order to produce lists that can be parsed by these rules, we need the
+ * ability to distinguish between characters that are part of a list element
+ * value from characters providing syntax that define the structure of the
+ * list. This means that our code that generates lists must at a minimum be
+ * able to produce escape sequences for the 10 characters identified above
+ * that have significance to a list parser.
+ *
+ *	*	*	CANONICAL LISTS	*	*	*	*	*
+ *
+ * In addition to the basic rules for parsing strings into Tcl lists, there
+ * are additional properties to be met by the set of list values that are
+ * generated by Tcl.  Such list values are often said to be in "canonical
+ * form":
+ *
+ * * When any canonical list is evaluated as a Tcl script, it is a script of
+ *   either zero commands (an empty list) or exactly one command. The command
+ *   word is exactly the first element of the list, and each argument word is
+ *   exactly one of the following elements of the list. This means that any
+ *   characters that have special meaning during script evaluation need
+ *   special treatment when canonical lists are produced:
+ *
+ *	* Whitespace between elements may not include NEWLINE.
+ *	* The command terminating character,
+ *		\u003b	;	SEMICOLON
+ *	  must be BRACEd, QUOTEd, or escaped so that it does not terminate the
+ * 	  command prematurely.
+ *	* Any of the characters that begin substitutions in scripts,
+ *		\u0024	$	DOLLAR
+ *		\u005b	[	OPEN BRACKET
+ *		\u005c	\	BACKSLASH
+ *	  need to be BRACEd or escaped.
+ *	* In any list where the first character of the first element is
+ *		\u0023	#	HASH
+ *	  that HASH character must be BRACEd, QUOTEd, or escaped so that it
+ *	  does not convert the command into a comment.
+ *	* Any list element that contains the character sequence BACKSLASH
+ *	  NEWLINE cannot be formatted with BRACEs. The BACKSLASH character
+ *	  must be represented by an escape sequence, and unless QUOTEs are
+ *	  used, the NEWLINE must be as well.
+ *
+ * * It is also guaranteed that one can use a canonical list as a building
+ *   block of a larger script within command substitution, as in this example:
+ *	set script "puts \[[list $cmd $arg]]"; eval $script
+ *   To support this usage, any appearance of the character
+ *		\u005d	]	CLOSE BRACKET
+ *   in a list element must be BRACEd, QUOTEd, or escaped.
+ *
+ * * Finally it is guaranteed that enclosing a canonical list in braces
+ *   produces a new value that is also a canonical list.  This new list has
+ *   length 1, and its only element is the original canonical list.  This same
+ *   guarantee also makes it possible to construct scripts where an argument
+ *   word is given a list value by enclosing the canonical form of that list
+ *   in braces:
+ *	set script "puts {[list $one $two $three]}"; eval $script
+ *   This sort of coding was once fairly common, though it's become more
+ *   idiomatic to see the following instead:
+ *	set script [list puts [list $one $two $three]]; eval $script
+ *   In order to support this guarantee, every canonical list must have
+ *   balance when counting those braces that are not in escape sequences.
+ *
+ * Within these constraints, the canonical list generation routines
+ * TclScanElement() and TclConvertElement() attempt to generate the string for
+ * any list that is easiest to read. When an element value is itself
+ * acceptable as the formatted substring, it is usually used (CONVERT_NONE).
+ * When some quoting or escaping is required, use of BRACEs (CONVERT_BRACE) is
+ * usually preferred over the use of escape sequences (CONVERT_ESCAPE). There
+ * are some exceptions to both of these preferences for reasons of code
+ * simplicity, efficiency, and continuation of historical habits. Canonical
+ * lists never use the QUOTE formatting to delimit their elements because that
+ * form of quoting does not nest, which makes construction of nested lists far
+ * too much trouble.  Canonical lists always use only a single SPACE character
+ * for element-separating whitespace.
+ *
+ *	*	*	FUTURE CONSIDERATIONS	*	*	*
+ *
+ * When a list element requires quoting or escaping due to a CLOSE BRACKET
+ * character or an internal QUOTE character, a strange formatting mode is
+ * recommended. For example, if the value "a{b]c}d" is converted by the usual
+ * modes:
+ *
+ *	CONVERT_BRACE:	a{b]c}d		=> {a{b]c}d}
+ *	CONVERT_ESCAPE:	a{b]c}d		=> a\{b\]c\}d
+ *
+ * we get perfectly usable formatted list elements. However, this is not what
+ * Tcl releases have been producing. Instead, we have:
+ *
+ *	CONVERT_MASK:	a{b]c}d		=> a{b\]c}d
+ *
+ * where the CLOSE BRACKET is escaped, but the BRACEs are not. The same effect
+ * can be seen replacing ] with " in this example. There does not appear to be
+ * any functional or aesthetic purpose for this strange additional mode. The
+ * sole purpose I can see for preserving it is to keep generating the same
+ * formatted lists programmers have become accustomed to, and perhaps written
+ * tests to expect. That is, compatibility only. The additional code
+ * complexity required to support this mode is significant. The lines of code
+ * supporting it are delimited in the routines below with #if COMPAT
+ * directives. This makes it easy to experiment with eliminating this
+ * formatting mode simply with "#define COMPAT 0" above. I believe this is
+ * worth considering.
+ *
+ * Another consideration is the treatment of QUOTE characters in list
+ * elements. TclConvertElement() must have the ability to produce the escape
+ * sequence \" so that when a list element begins with a QUOTE we do not
+ * confuse that first character with a QUOTE used as list syntax to define
+ * list structure. However, that is the only place where QUOTE characters need
+ * quoting. In this way, handling QUOTE could really be much more like the way
+ * we handle HASH which also needs quoting and escaping only in particular
+ * situations. Following up this could increase the set of list elements that
+ * can use the CONVERT_NONE formatting mode.
+ *
+ * More speculative is that the demands of canonical list form require brace
+ * balance for the list as a whole, while the current implementation achieves
+ * this by establishing brace balance for every element.
+ *
+ * Finally, a reminder that the rules for parsing and formatting lists are
+ * closely tied together with the rules for parsing and evaluating scripts,
+ * and will need to evolve in sync.
+ */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclMaxListLength --
+ *
+ *	Given 'bytes' pointing to 'numBytes' bytes, scan through them and
+ *	count the number of whitespace runs that could be list element
+ *	separators. If 'numBytes' is -1, scan to the terminating '\0'. Not a
+ *	full list parser. Typically used to get a quick and dirty overestimate
+ *	of length size in order to allocate space for an actual list parser to
+ *	operate with.
+ *
+ * Results:
+ *	Returns the largest number of list elements that could possibly be in
+ *	this string, interpreted as a Tcl list. If 'endPtr' is not NULL,
+ *	writes a pointer to the end of the string scanned there.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclMaxListLength(
+    const char *bytes,
+    int numBytes,
+    const char **endPtr)
+{
+    int count = 0;
+
+    if ((numBytes == 0) || ((numBytes == -1) && (*bytes == '\0'))) {
+	/* Empty string case - quick exit */
+	goto done;
+    }
+
+    /*
+     * No list element before leading white space.
+     */
+
+    count += 1 - TclIsSpaceProc(*bytes);
+
+    /*
+     * Count white space runs as potential element separators.
+     */
+
+    while (numBytes) {
+	if ((numBytes == -1) && (*bytes == '\0')) {
+	    break;
+	}
+	if (TclIsSpaceProc(*bytes)) {
+	    /*
+	     * Space run started; bump count.
+	     */
+
+	    count++;
+	    do {
+		bytes++;
+		numBytes -= (numBytes != -1);
+	    } while (numBytes && TclIsSpaceProc(*bytes));
+	    if ((numBytes == 0) || ((numBytes == -1) && (*bytes == '\0'))) {
+		break;
+	    }
+
+	    /*
+	     * (*bytes) is non-space; return to counting state.
+	     */
+	}
+	bytes++;
+	numBytes -= (numBytes != -1);
+    }
+
+    /*
+     * No list element following trailing white space.
+     */
+
+    count -= TclIsSpaceProc(bytes[-1]);
+
+  done:
+    if (endPtr) {
+	*endPtr = bytes;
+    }
+    return count;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFindElement --
+ *
+ *	Given a pointer into a Tcl list, locate the first (or next) element in
+ *	the list.
+ *
+ * Results:
+ *	The return value is normally TCL_OK, which means that the element was
+ *	successfully located. If TCL_ERROR is returned it means that list
+ *	didn't have proper list structure; the interp's result contains a more
+ *	detailed error message.
+ *
+ *	If TCL_OK is returned, then *elementPtr will be set to point to the
+ *	first element of list, and *nextPtr will be set to point to the
+ *	character just after any white space following the last character
+ *	that's part of the element. If this is the last argument in the list,
+ *	then *nextPtr will point just after the last character in the list
+ *	(i.e., at the character at list+listLength). If sizePtr is non-NULL,
+ *	*sizePtr is filled in with the number of bytes in the element. If the
+ *	element is in braces, then *elementPtr will point to the character
+ *	after the opening brace and *sizePtr will not include either of the
+ *	braces. If there isn't an element in the list, *sizePtr will be zero,
+ *	and both *elementPtr and *nextPtr will point just after the last
+ *	character in the list. If literalPtr is non-NULL, *literalPtr is set
+ *	to a boolean value indicating whether the substring returned as the
+ *	values of **elementPtr and *sizePtr is the literal value of a list
+ *	element. If not, a call to TclCopyAndCollapse() is needed to produce
+ *	the actual value of the list element. Note: this function does NOT
+ *	collapse backslash sequences, but uses *literalPtr to tell callers
+ *	when it is required for them to do so.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFindElement(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, then no error message is left after
+				 * errors. */
+    const char *list,		/* Points to the first byte of a string
+				 * containing a Tcl list with zero or more
+				 * elements (possibly in braces). */
+    int listLength,		/* Number of bytes in the list's string. */
+    const char **elementPtr,	/* Where to put address of first significant
+				 * character in first element of list. */
+    const char **nextPtr,	/* Fill in with location of character just
+				 * after all white space following end of
+				 * argument (next arg or end of list). */
+    int *sizePtr,		/* If non-zero, fill in with size of
+				 * element. */
+    int *literalPtr)		/* If non-zero, fill in with non-zero/zero to
+				 * indicate that the substring of *sizePtr
+				 * bytes starting at **elementPtr is/is not
+				 * the literal list element and therefore
+				 * does not/does require a call to
+				 * TclCopyAndCollapse() by the caller. */
+{
+    return FindElement(interp, list, listLength, "list", "LIST", elementPtr,
+	    nextPtr, sizePtr, literalPtr);
+}
+
+int
+TclFindDictElement(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, then no error message is left after
+				 * errors. */
+    const char *dict,		/* Points to the first byte of a string
+				 * containing a Tcl dictionary with zero or
+				 * more keys and values (possibly in
+				 * braces). */
+    int dictLength,		/* Number of bytes in the dict's string. */
+    const char **elementPtr,	/* Where to put address of first significant
+				 * character in the first element (i.e., key
+				 * or value) of dict. */
+    const char **nextPtr,	/* Fill in with location of character just
+				 * after all white space following end of
+				 * element (next arg or end of list). */
+    int *sizePtr,		/* If non-zero, fill in with size of
+				 * element. */
+    int *literalPtr)		/* If non-zero, fill in with non-zero/zero to
+				 * indicate that the substring of *sizePtr
+				 * bytes starting at **elementPtr is/is not
+				 * the literal key or value and therefore
+				 * does not/does require a call to
+				 * TclCopyAndCollapse() by the caller. */
+{
+    return FindElement(interp, dict, dictLength, "dict", "DICTIONARY",
+	    elementPtr, nextPtr, sizePtr, literalPtr);
+}
+
+static int
+FindElement(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, then no error message is left after
+				 * errors. */
+    const char *string,		/* Points to the first byte of a string
+				 * containing a Tcl list or dictionary with
+				 * zero or more elements (possibly in
+				 * braces). */
+    int stringLength,		/* Number of bytes in the string. */
+    const char *typeStr,	/* The name of the type of thing we are
+				 * parsing, for error messages. */
+    const char *typeCode,	/* The type code for thing we are parsing, for
+				 * error messages. */
+    const char **elementPtr,	/* Where to put address of first significant
+				 * character in first element. */
+    const char **nextPtr,	/* Fill in with location of character just
+				 * after all white space following end of
+				 * argument (next arg or end of list/dict). */
+    int *sizePtr,		/* If non-zero, fill in with size of
+				 * element. */
+    int *literalPtr)		/* If non-zero, fill in with non-zero/zero to
+				 * indicate that the substring of *sizePtr
+				 * bytes starting at **elementPtr is/is not
+				 * the literal list/dict element and therefore
+				 * does not/does require a call to
+				 * TclCopyAndCollapse() by the caller. */
+{
+    const char *p = string;
+    const char *elemStart;	/* Points to first byte of first element. */
+    const char *limit;		/* Points just after list/dict's last byte. */
+    int openBraces = 0;		/* Brace nesting level during parse. */
+    int inQuotes = 0;
+    int size = 0;		/* lint. */
+    int numChars;
+    int literal = 1;
+    const char *p2;
+
+    /*
+     * Skim off leading white space and check for an opening brace or quote.
+     * We treat embedded NULLs in the list/dict as bytes belonging to a list
+     * element (or dictionary key or value).
+     */
+
+    limit = (string + stringLength);
+    while ((p < limit) && (TclIsSpaceProc(*p))) {
+	p++;
+    }
+    if (p == limit) {		/* no element found */
+	elemStart = limit;
+	goto done;
+    }
+
+    if (*p == '{') {
+	openBraces = 1;
+	p++;
+    } else if (*p == '"') {
+	inQuotes = 1;
+	p++;
+    }
+    elemStart = p;
+
+    /*
+     * Find element's end (a space, close brace, or the end of the string).
+     */
+
+    while (p < limit) {
+	switch (*p) {
+	    /*
+	     * Open brace: don't treat specially unless the element is in
+	     * braces. In this case, keep a nesting count.
+	     */
+
+	case '{':
+	    if (openBraces != 0) {
+		openBraces++;
+	    }
+	    break;
+
+	    /*
+	     * Close brace: if element is in braces, keep nesting count and
+	     * quit when the last close brace is seen.
+	     */
+
+	case '}':
+	    if (openBraces > 1) {
+		openBraces--;
+	    } else if (openBraces == 1) {
+		size = (p - elemStart);
+		p++;
+		if ((p >= limit) || TclIsSpaceProc(*p)) {
+		    goto done;
+		}
+
+		/*
+		 * Garbage after the closing brace; return an error.
+		 */
+
+		if (interp != NULL) {
+		    p2 = p;
+		    while ((p2 < limit) && (!TclIsSpaceProc(*p2))
+			    && (p2 < p+20)) {
+			p2++;
+		    }
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "%s element in braces followed by \"%.*s\" "
+			    "instead of space", typeStr, (int) (p2-p), p));
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", typeCode, "JUNK",
+			    NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    break;
+
+	    /*
+	     * Backslash: skip over everything up to the end of the backslash
+	     * sequence.
+	     */
+
+	case '\\':
+	    if (openBraces == 0) {
+		/*
+		 * A backslash sequence not within a brace quoted element
+		 * means the value of the element is different from the
+		 * substring we are parsing. A call to TclCopyAndCollapse() is
+		 * needed to produce the element value. Inform the caller.
+		 */
+
+		literal = 0;
+	    }
+	    TclParseBackslash(p, limit - p, &numChars, NULL);
+	    p += (numChars - 1);
+	    break;
+
+	    /*
+	     * Space: ignore if element is in braces or quotes; otherwise
+	     * terminate element.
+	     */
+
+	case ' ':
+	case '\f':
+	case '\n':
+	case '\r':
+	case '\t':
+	case '\v':
+	    if ((openBraces == 0) && !inQuotes) {
+		size = (p - elemStart);
+		goto done;
+	    }
+	    break;
+
+	    /*
+	     * Double-quote: if element is in quotes then terminate it.
+	     */
+
+	case '"':
+	    if (inQuotes) {
+		size = (p - elemStart);
+		p++;
+		if ((p >= limit) || TclIsSpaceProc(*p)) {
+		    goto done;
+		}
+
+		/*
+		 * Garbage after the closing quote; return an error.
+		 */
+
+		if (interp != NULL) {
+		    p2 = p;
+		    while ((p2 < limit) && (!TclIsSpaceProc(*p2))
+			    && (p2 < p+20)) {
+			p2++;
+		    }
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "%s element in quotes followed by \"%.*s\" "
+			    "instead of space", typeStr, (int) (p2-p), p));
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", typeCode, "JUNK",
+			    NULL);
+		}
+		return TCL_ERROR;
+	    }
+	    break;
+	}
+	p++;
+    }
+
+    /*
+     * End of list/dict: terminate element.
+     */
+
+    if (p == limit) {
+	if (openBraces != 0) {
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"unmatched open brace in %s", typeStr));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", typeCode, "BRACE",
+			NULL);
+	    }
+	    return TCL_ERROR;
+	} else if (inQuotes) {
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"unmatched open quote in %s", typeStr));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", typeCode, "QUOTE",
+			NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	size = (p - elemStart);
+    }
+
+  done:
+    while ((p < limit) && (TclIsSpaceProc(*p))) {
+	p++;
+    }
+    *elementPtr = elemStart;
+    *nextPtr = p;
+    if (sizePtr != 0) {
+	*sizePtr = size;
+    }
+    if (literalPtr != 0) {
+	*literalPtr = literal;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCopyAndCollapse --
+ *
+ *	Copy a string and substitute all backslash escape sequences
+ *
+ * Results:
+ *	Count bytes get copied from src to dst. Along the way, backslash
+ *	sequences are substituted in the copy. After scanning count bytes from
+ *	src, a null character is placed at the end of dst. Returns the number
+ *	of bytes that got written to dst.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCopyAndCollapse(
+    int count,			/* Number of byte to copy from src. */
+    const char *src,		/* Copy from here... */
+    char *dst)			/* ... to here. */
+{
+    int newCount = 0;
+
+    while (count > 0) {
+	char c = *src;
+
+	if (c == '\\') {
+	    int numRead;
+	    int backslashCount = TclParseBackslash(src, count, &numRead, dst);
+
+	    dst += backslashCount;
+	    newCount += backslashCount;
+	    src += numRead;
+	    count -= numRead;
+	} else {
+	    *dst = c;
+	    dst++;
+	    newCount++;
+	    src++;
+	    count--;
+	}
+    }
+    *dst = 0;
+    return newCount;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SplitList --
+ *
+ *	Splits a list up into its constituent fields.
+ *
+ * Results
+ *	The return value is normally TCL_OK, which means that the list was
+ *	successfully split up. If TCL_ERROR is returned, it means that "list"
+ *	didn't have proper list structure; the interp's result will contain a
+ *	more detailed error message.
+ *
+ *	*argvPtr will be filled in with the address of an array whose elements
+ *	point to the elements of list, in order. *argcPtr will get filled in
+ *	with the number of valid elements in the array. A single block of
+ *	memory is dynamically allocated to hold both the argv array and a copy
+ *	of the list (with backslashes and braces removed in the standard way).
+ *	The caller must eventually free this memory by calling free() on
+ *	*argvPtr. Note: *argvPtr and *argcPtr are only modified if the
+ *	function returns normally.
+ *
+ * Side effects:
+ *	Memory is allocated.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_SplitList(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, no error message is left. */
+    const char *list,		/* Pointer to string with list structure. */
+    int *argcPtr,		/* Pointer to location to fill in with the
+				 * number of elements in the list. */
+    const char ***argvPtr)	/* Pointer to place to store pointer to array
+				 * of pointers to list elements. */
+{
+    const char **argv, *end, *element;
+    char *p;
+    int length, size, i, result, elSize;
+
+    /*
+     * Allocate enough space to work in. A (const char *) for each (possible)
+     * list element plus one more for terminating NULL, plus as many bytes as
+     * in the original string value, plus one more for a terminating '\0'.
+     * Space used to hold element separating white space in the original
+     * string gets re-purposed to hold '\0' characters in the argv array.
+     */
+
+    size = TclMaxListLength(list, -1, &end) + 1;
+    length = end - list;
+    argv = ckalloc((size * sizeof(char *)) + length + 1);
+
+    for (i = 0, p = ((char *) argv) + size*sizeof(char *);
+	    *list != 0;  i++) {
+	const char *prevList = list;
+	int literal;
+
+	result = TclFindElement(interp, list, length, &element, &list,
+		&elSize, &literal);
+	length -= (list - prevList);
+	if (result != TCL_OK) {
+	    ckfree(argv);
+	    return result;
+	}
+	if (*element == 0) {
+	    break;
+	}
+	if (i >= size) {
+	    ckfree(argv);
+	    if (interp != NULL) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"internal error in Tcl_SplitList", -1));
+		Tcl_SetErrorCode(interp, "TCL", "INTERNAL", "Tcl_SplitList",
+			NULL);
+	    }
+	    return TCL_ERROR;
+	}
+	argv[i] = p;
+	if (literal) {
+	    memcpy(p, element, (size_t) elSize);
+	    p += elSize;
+	    *p = 0;
+	    p++;
+	} else {
+	    p += 1 + TclCopyAndCollapse(elSize, element, p);
+	}
+    }
+
+    argv[i] = NULL;
+    *argvPtr = argv;
+    *argcPtr = i;
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ScanElement --
+ *
+ *	This function is a companion function to Tcl_ConvertElement. It scans
+ *	a string to see what needs to be done to it (e.g. add backslashes or
+ *	enclosing braces) to make the string into a valid Tcl list element.
+ *
+ * Results:
+ *	The return value is an overestimate of the number of bytes that will
+ *	be needed by Tcl_ConvertElement to produce a valid list element from
+ *	src. The word at *flagPtr is filled in with a value needed by
+ *	Tcl_ConvertElement when doing the actual conversion.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ScanElement(
+    register const char *src,	/* String to convert to list element. */
+    register int *flagPtr)	/* Where to store information to guide
+				 * Tcl_ConvertCountedElement. */
+{
+    return Tcl_ScanCountedElement(src, -1, flagPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ScanCountedElement --
+ *
+ *	This function is a companion function to Tcl_ConvertCountedElement. It
+ *	scans a string to see what needs to be done to it (e.g. add
+ *	backslashes or enclosing braces) to make the string into a valid Tcl
+ *	list element. If length is -1, then the string is scanned from src up
+ *	to the first null byte.
+ *
+ * Results:
+ *	The return value is an overestimate of the number of bytes that will
+ *	be needed by Tcl_ConvertCountedElement to produce a valid list element
+ *	from src. The word at *flagPtr is filled in with a value needed by
+ *	Tcl_ConvertCountedElement when doing the actual conversion.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ScanCountedElement(
+    const char *src,		/* String to convert to Tcl list element. */
+    int length,			/* Number of bytes in src, or -1. */
+    int *flagPtr)		/* Where to store information to guide
+				 * Tcl_ConvertElement. */
+{
+    int flags = CONVERT_ANY;
+    int numBytes = TclScanElement(src, length, &flags);
+
+    *flagPtr = flags;
+    return numBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclScanElement --
+ *
+ *	This function is a companion function to TclConvertElement. It scans a
+ *	string to see what needs to be done to it (e.g. add backslashes or
+ *	enclosing braces) to make the string into a valid Tcl list element. If
+ *	length is -1, then the string is scanned from src up to the first null
+ *	byte. A NULL value for src is treated as an empty string. The incoming
+ *	value of *flagPtr is a report from the caller what additional flags it
+ *	will pass to TclConvertElement().
+ *
+ * Results:
+ *	The recommended formatting mode for the element is determined and a
+ *	value is written to *flagPtr indicating that recommendation. This
+ *	recommendation is combined with the incoming flag values in *flagPtr
+ *	set by the caller to determine how many bytes will be needed by
+ *	TclConvertElement() in which to write the formatted element following
+ *	the recommendation modified by the flag values. This number of bytes
+ *	is the return value of the routine.  In some situations it may be an
+ *	overestimate, but so long as the caller passes the same flags to
+ *	TclConvertElement(), it will be large enough.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclScanElement(
+    const char *src,		/* String to convert to Tcl list element. */
+    int length,			/* Number of bytes in src, or -1. */
+    int *flagPtr)		/* Where to store information to guide
+				 * Tcl_ConvertElement. */
+{
+    const char *p = src;
+    int nestingLevel = 0;	/* Brace nesting count */
+    int forbidNone = 0;		/* Do not permit CONVERT_NONE mode. Something
+				 * needs protection or escape. */
+    int requireEscape = 0;	/* Force use of CONVERT_ESCAPE mode.  For some
+				 * reason bare or brace-quoted form fails. */
+    int extra = 0;		/* Count of number of extra bytes needed for
+				 * formatted element, assuming we use escape
+				 * sequences in formatting. */
+    int bytesNeeded;		/* Buffer length computed to complete the
+				 * element formatting in the selected mode. */
+#if COMPAT
+    int preferEscape = 0;	/* Use preferences to track whether to use */
+    int preferBrace = 0;	/* CONVERT_MASK mode. */
+    int braceCount = 0;		/* Count of all braces '{' '}' seen. */
+#endif /* COMPAT */
+
+    if ((p == NULL) || (length == 0) || ((*p == '\0') && (length == -1))) {
+	/*
+	 * Empty string element must be brace quoted.
+	 */
+
+	*flagPtr = CONVERT_BRACE;
+	return 2;
+    }
+
+    if ((*p == '{') || (*p == '"')) {
+	/*
+	 * Must escape or protect so leading character of value is not
+	 * misinterpreted as list element delimiting syntax.
+	 */
+
+	forbidNone = 1;
+#if COMPAT
+	preferBrace = 1;
+#endif /* COMPAT */
+    }
+
+    while (length) {
+      if (CHAR_TYPE(*p) != TYPE_NORMAL) {
+	switch (*p) {
+	case '{':	/* TYPE_BRACE */
+#if COMPAT
+	    braceCount++;
+#endif /* COMPAT */
+	    extra++;				/* Escape '{' => '\{' */
+	    nestingLevel++;
+	    break;
+	case '}':	/* TYPE_BRACE */
+#if COMPAT
+	    braceCount++;
+#endif /* COMPAT */
+	    extra++;				/* Escape '}' => '\}' */
+	    nestingLevel--;
+	    if (nestingLevel < 0) {
+		/*
+		 * Unbalanced braces!  Cannot format with brace quoting.
+		 */
+
+		requireEscape = 1;
+	    }
+	    break;
+	case ']':	/* TYPE_CLOSE_BRACK */
+	case '"':	/* TYPE_SPACE */
+#if COMPAT
+	    forbidNone = 1;
+	    extra++;		/* Escapes all just prepend a backslash */
+	    preferEscape = 1;
+	    break;
+#else
+	    /* FLOW THROUGH */
+#endif /* COMPAT */
+	case '[':	/* TYPE_SUBS */
+	case '$':	/* TYPE_SUBS */
+	case ';':	/* TYPE_COMMAND_END */
+	case ' ':	/* TYPE_SPACE */
+	case '\f':	/* TYPE_SPACE */
+	case '\n':	/* TYPE_COMMAND_END */
+	case '\r':	/* TYPE_SPACE */
+	case '\t':	/* TYPE_SPACE */
+	case '\v':	/* TYPE_SPACE */
+	    forbidNone = 1;
+	    extra++;		/* Escape sequences all one byte longer. */
+#if COMPAT
+	    preferBrace = 1;
+#endif /* COMPAT */
+	    break;
+	case '\\':	/* TYPE_SUBS */
+	    extra++;				/* Escape '\' => '\\' */
+	    if ((length == 1) || ((length == -1) && (p[1] == '\0'))) {
+		/*
+		 * Final backslash. Cannot format with brace quoting.
+		 */
+
+		requireEscape = 1;
+		break;
+	    }
+	    if (p[1] == '\n') {
+		extra++;	/* Escape newline => '\n', one byte longer */
+
+		/*
+		 * Backslash newline sequence.  Brace quoting not permitted.
+		 */
+
+		requireEscape = 1;
+		length -= (length > 0);
+		p++;
+		break;
+	    }
+	    if ((p[1] == '{') || (p[1] == '}') || (p[1] == '\\')) {
+		extra++;	/* Escape sequences all one byte longer. */
+		length -= (length > 0);
+		p++;
+	    }
+	    forbidNone = 1;
+#if COMPAT
+	    preferBrace = 1;
+#endif /* COMPAT */
+	    break;
+	case '\0':	/* TYPE_SUBS */
+	    if (length == -1) {
+		goto endOfString;
+	    }
+	    /* TODO: Panic on improper encoding? */
+	    break;
+	}
+      }
+	length -= (length > 0);
+	p++;
+    }
+
+  endOfString:
+    if (nestingLevel != 0) {
+	/*
+	 * Unbalanced braces!  Cannot format with brace quoting.
+	 */
+
+	requireEscape = 1;
+    }
+
+    /*
+     * We need at least as many bytes as are in the element value...
+     */
+
+    bytesNeeded = p - src;
+
+    if (requireEscape) {
+	/*
+	 * We must use escape sequences.  Add all the extra bytes needed to
+	 * have room to create them.
+	 */
+
+	bytesNeeded += extra;
+
+	/*
+	 * Make room to escape leading #, if needed.
+	 */
+
+	if ((*src == '#') && !(*flagPtr & TCL_DONT_QUOTE_HASH)) {
+	    bytesNeeded++;
+	}
+	*flagPtr = CONVERT_ESCAPE;
+	goto overflowCheck;
+    }
+    if (*flagPtr & CONVERT_ANY) {
+	/*
+	 * The caller has not let us know what flags it will pass to
+	 * TclConvertElement() so compute the max size we might need for any
+	 * possible choice.  Normally the formatting using escape sequences is
+	 * the longer one, and a minimum "extra" value of 2 makes sure we
+	 * don't request too small a buffer in those edge cases where that's
+	 * not true.
+	 */
+
+	if (extra < 2) {
+	    extra = 2;
+	}
+	*flagPtr &= ~CONVERT_ANY;
+	*flagPtr |= TCL_DONT_USE_BRACES;
+    }
+    if (forbidNone) {
+	/*
+	 * We must request some form of quoting of escaping...
+	 */
+
+#if COMPAT
+	if (preferEscape && !preferBrace) {
+	    /*
+	     * If we are quoting solely due to ] or internal " characters use
+	     * the CONVERT_MASK mode where we escape all special characters
+	     * except for braces. "extra" counted space needed to escape
+	     * braces too, so substract "braceCount" to get our actual needs.
+	     */
+
+	    bytesNeeded += (extra - braceCount);
+	    /* Make room to escape leading #, if needed. */
+	    if ((*src == '#') && !(*flagPtr & TCL_DONT_QUOTE_HASH)) {
+		bytesNeeded++;
+	    }
+
+	    /*
+	     * If the caller reports it will direct TclConvertElement() to
+	     * use full escapes on the element, add back the bytes needed to
+	     * escape the braces.
+	     */
+
+	    if (*flagPtr & TCL_DONT_USE_BRACES) {
+		bytesNeeded += braceCount;
+	    }
+	    *flagPtr = CONVERT_MASK;
+	    goto overflowCheck;
+	}
+#endif /* COMPAT */
+	if (*flagPtr & TCL_DONT_USE_BRACES) {
+	    /*
+	     * If the caller reports it will direct TclConvertElement() to
+	     * use escapes, add the extra bytes needed to have room for them.
+	     */
+
+	    bytesNeeded += extra;
+
+	    /*
+	     * Make room to escape leading #, if needed.
+	     */
+
+	    if ((*src == '#') && !(*flagPtr & TCL_DONT_QUOTE_HASH)) {
+		bytesNeeded++;
+	    }
+	} else {
+	    /*
+	     * Add 2 bytes for room for the enclosing braces.
+	     */
+
+	    bytesNeeded += 2;
+	}
+	*flagPtr = CONVERT_BRACE;
+	goto overflowCheck;
+    }
+
+    /*
+     * So far, no need to quote or escape anything.
+     */
+
+    if ((*src == '#') && !(*flagPtr & TCL_DONT_QUOTE_HASH)) {
+	/*
+	 * If we need to quote a leading #, make room to enclose in braces.
+	 */
+
+	bytesNeeded += 2;
+    }
+    *flagPtr = CONVERT_NONE;
+
+  overflowCheck:
+    if (bytesNeeded < 0) {
+	Tcl_Panic("TclScanElement: string length overflow");
+    }
+    return bytesNeeded;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConvertElement --
+ *
+ *	This is a companion function to Tcl_ScanElement. Given the information
+ *	produced by Tcl_ScanElement, this function converts a string to a list
+ *	element equal to that string.
+ *
+ * Results:
+ *	Information is copied to *dst in the form of a list element identical
+ *	to src (i.e. if Tcl_SplitList is applied to dst it will produce a
+ *	string identical to src). The return value is a count of the number of
+ *	characters copied (not including the terminating NULL character).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ConvertElement(
+    register const char *src,	/* Source information for list element. */
+    register char *dst,		/* Place to put list-ified element. */
+    register int flags)		/* Flags produced by Tcl_ScanElement. */
+{
+    return Tcl_ConvertCountedElement(src, -1, dst, flags);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConvertCountedElement --
+ *
+ *	This is a companion function to Tcl_ScanCountedElement. Given the
+ *	information produced by Tcl_ScanCountedElement, this function converts
+ *	a string to a list element equal to that string.
+ *
+ * Results:
+ *	Information is copied to *dst in the form of a list element identical
+ *	to src (i.e. if Tcl_SplitList is applied to dst it will produce a
+ *	string identical to src). The return value is a count of the number of
+ *	characters copied (not including the terminating NULL character).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ConvertCountedElement(
+    register const char *src,	/* Source information for list element. */
+    int length,			/* Number of bytes in src, or -1. */
+    char *dst,			/* Place to put list-ified element. */
+    int flags)			/* Flags produced by Tcl_ScanElement. */
+{
+    int numBytes = TclConvertElement(src, length, dst, flags);
+    dst[numBytes] = '\0';
+    return numBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclConvertElement --
+ *
+ *	This is a companion function to TclScanElement. Given the information
+ *	produced by TclScanElement, this function converts a string to a list
+ *	element equal to that string.
+ *
+ * Results:
+ *	Information is copied to *dst in the form of a list element identical
+ *	to src (i.e. if Tcl_SplitList is applied to dst it will produce a
+ *	string identical to src). The return value is a count of the number of
+ *	characters copied (not including the terminating NULL character).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclConvertElement(
+    register const char *src,	/* Source information for list element. */
+    int length,			/* Number of bytes in src, or -1. */
+    char *dst,			/* Place to put list-ified element. */
+    int flags)			/* Flags produced by Tcl_ScanElement. */
+{
+    int conversion = flags & CONVERT_MASK;
+    char *p = dst;
+
+    /*
+     * Let the caller demand we use escape sequences rather than braces.
+     */
+
+    if ((flags & TCL_DONT_USE_BRACES) && (conversion & CONVERT_BRACE)) {
+	conversion = CONVERT_ESCAPE;
+    }
+
+    /*
+     * No matter what the caller demands, empty string must be braced!
+     */
+
+    if ((src == NULL) || (length == 0) || (*src == '\0' && length == -1)) {
+	src = &tclEmptyString;
+	length = 0;
+	conversion = CONVERT_BRACE;
+    }
+
+    /*
+     * Escape leading hash as needed and requested.
+     */
+
+    if ((*src == '#') && !(flags & TCL_DONT_QUOTE_HASH)) {
+	if (conversion == CONVERT_ESCAPE) {
+	    p[0] = '\\';
+	    p[1] = '#';
+	    p += 2;
+	    src++;
+	    length -= (length > 0);
+	} else {
+	    conversion = CONVERT_BRACE;
+	}
+    }
+
+    /*
+     * No escape or quoting needed.  Copy the literal string value.
+     */
+
+    if (conversion == CONVERT_NONE) {
+	if (length == -1) {
+	    /* TODO: INT_MAX overflow? */
+	    while (*src) {
+		*p++ = *src++;
+	    }
+	    return p - dst;
+	} else {
+	    memcpy(dst, src, length);
+	    return length;
+	}
+    }
+
+    /*
+     * Formatted string is original string enclosed in braces.
+     */
+
+    if (conversion == CONVERT_BRACE) {
+	*p = '{';
+	p++;
+	if (length == -1) {
+	    /* TODO: INT_MAX overflow? */
+	    while (*src) {
+		*p++ = *src++;
+	    }
+	} else {
+	    memcpy(p, src, length);
+	    p += length;
+	}
+	*p = '}';
+	p++;
+	return p - dst;
+    }
+
+    /* conversion == CONVERT_ESCAPE or CONVERT_MASK */
+
+    /*
+     * Formatted string is original string converted to escape sequences.
+     */
+
+    for ( ; length; src++, length -= (length > 0)) {
+	switch (*src) {
+	case ']':
+	case '[':
+	case '$':
+	case ';':
+	case ' ':
+	case '\\':
+	case '"':
+	    *p = '\\';
+	    p++;
+	    break;
+	case '{':
+	case '}':
+#if COMPAT
+	    if (conversion == CONVERT_ESCAPE)
+#endif /* COMPAT */
+	    {
+		*p = '\\';
+		p++;
+	    }
+	    break;
+	case '\f':
+	    *p = '\\';
+	    p++;
+	    *p = 'f';
+	    p++;
+	    continue;
+	case '\n':
+	    *p = '\\';
+	    p++;
+	    *p = 'n';
+	    p++;
+	    continue;
+	case '\r':
+	    *p = '\\';
+	    p++;
+	    *p = 'r';
+	    p++;
+	    continue;
+	case '\t':
+	    *p = '\\';
+	    p++;
+	    *p = 't';
+	    p++;
+	    continue;
+	case '\v':
+	    *p = '\\';
+	    p++;
+	    *p = 'v';
+	    p++;
+	    continue;
+	case '\0':
+	    if (length == -1) {
+		return p - dst;
+	    }
+
+	    /*
+	     * If we reach this point, there's an embedded NULL in the string
+	     * range being processed, which should not happen when the
+	     * encoding rules for Tcl strings are properly followed.  If the
+	     * day ever comes when we stop tolerating such things, this is
+	     * where to put the Tcl_Panic().
+	     */
+
+	    break;
+	}
+	*p = *src;
+	p++;
+    }
+    return p - dst;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Merge --
+ *
+ *	Given a collection of strings, merge them together into a single
+ *	string that has proper Tcl list structured (i.e. Tcl_SplitList may be
+ *	used to retrieve strings equal to the original elements, and Tcl_Eval
+ *	will parse the string back into its original elements).
+ *
+ * Results:
+ *	The return value is the address of a dynamically-allocated string
+ *	containing the merged list.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_Merge(
+    int argc,			/* How many strings to merge. */
+    const char *const *argv)	/* Array of string values. */
+{
+#define LOCAL_SIZE 20
+    int localFlags[LOCAL_SIZE], *flagPtr = NULL;
+    int i, bytesNeeded = 0;
+    char *result, *dst;
+    const int maxFlags = UINT_MAX / sizeof(int);
+
+    /*
+     * Handle empty list case first, so logic of the general case can be
+     * simpler.
+     */
+
+    if (argc == 0) {
+	result = ckalloc(1);
+	result[0] = '\0';
+	return result;
+    }
+
+    /*
+     * Pass 1: estimate space, gather flags.
+     */
+
+    if (argc <= LOCAL_SIZE) {
+	flagPtr = localFlags;
+    } else if (argc > maxFlags) {
+	/*
+	 * We cannot allocate a large enough flag array to format this list in
+	 * one pass.  We could imagine converting this routine to a multi-pass
+	 * implementation, but for sizeof(int) == 4, the limit is a max of
+	 * 2^30 list elements and since each element is at least one byte
+	 * formatted, and requires one byte space between it and the next one,
+	 * that a minimum space requirement of 2^31 bytes, which is already
+	 * INT_MAX. If we tried to format a list of > maxFlags elements, we're
+	 * just going to overflow the size limits on the formatted string
+	 * anyway, so just issue that same panic early.
+	 */
+
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    } else {
+	flagPtr = ckalloc(argc * sizeof(int));
+    }
+    for (i = 0; i < argc; i++) {
+	flagPtr[i] = ( i ? TCL_DONT_QUOTE_HASH : 0 );
+	bytesNeeded += TclScanElement(argv[i], -1, &flagPtr[i]);
+	if (bytesNeeded < 0) {
+	    Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+	}
+    }
+    if (bytesNeeded > INT_MAX - argc + 1) {
+	Tcl_Panic("max size for a Tcl value (%d bytes) exceeded", INT_MAX);
+    }
+    bytesNeeded += argc;
+
+    /*
+     * Pass two: copy into the result area.
+     */
+
+    result = ckalloc(bytesNeeded);
+    dst = result;
+    for (i = 0; i < argc; i++) {
+	flagPtr[i] |= ( i ? TCL_DONT_QUOTE_HASH : 0 );
+	dst += TclConvertElement(argv[i], -1, dst, flagPtr[i]);
+	*dst = ' ';
+	dst++;
+    }
+    dst[-1] = 0;
+
+    if (flagPtr != localFlags) {
+	ckfree(flagPtr);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Backslash --
+ *
+ *	Figure out how to handle a backslash sequence.
+ *
+ * Results:
+ *	The return value is the character that should be substituted in place
+ *	of the backslash sequence that starts at src. If readPtr isn't NULL
+ *	then it is filled in with a count of the number of characters in the
+ *	backslash sequence.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char
+Tcl_Backslash(
+    const char *src,		/* Points to the backslash character of a
+				 * backslash sequence. */
+    int *readPtr)		/* Fill in with number of characters read from
+				 * src, unless NULL. */
+{
+    char buf[TCL_UTF_MAX];
+    Tcl_UniChar ch = 0;
+
+    Tcl_UtfBackslash(src, readPtr, buf);
+    TclUtfToUniChar(buf, &ch);
+    return (char) ch;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTrimRight --
+ *
+ *	Takes two counted strings in the Tcl encoding which must both be null
+ *	terminated. Conceptually trims from the right side of the first string
+ *	all characters found in the second string.
+ *
+ * Results:
+ *	The number of bytes to be removed from the end of the string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclTrimRight(
+    const char *bytes,		/* String to be trimmed... */
+    int numBytes,		/* ...and its length in bytes */
+    const char *trim,		/* String of trim characters... */
+    int numTrim)		/* ...and its length in bytes */
+{
+    const char *p = bytes + numBytes;
+    int pInc;
+
+    if ((bytes[numBytes] != '\0') || (trim[numTrim] != '\0')) {
+	Tcl_Panic("TclTrimRight works only on null-terminated strings");
+    }
+
+    /*
+     * Empty strings -> nothing to do.
+     */
+
+    if ((numBytes == 0) || (numTrim == 0)) {
+	return 0;
+    }
+
+    /*
+     * Outer loop: iterate over string to be trimmed.
+     */
+
+    do {
+	Tcl_UniChar ch1;
+	const char *q = trim;
+	int bytesLeft = numTrim;
+
+	p = Tcl_UtfPrev(p, bytes);
+ 	pInc = TclUtfToUniChar(p, &ch1);
+
+	/*
+	 * Inner loop: scan trim string for match to current character.
+	 */
+
+	do {
+	    Tcl_UniChar ch2;
+	    int qInc = TclUtfToUniChar(q, &ch2);
+
+	    if (ch1 == ch2) {
+		break;
+	    }
+
+	    q += qInc;
+	    bytesLeft -= qInc;
+	} while (bytesLeft);
+
+	if (bytesLeft == 0) {
+	    /*
+	     * No match; trim task done; *p is last non-trimmed char.
+	     */
+
+	    p += pInc;
+	    break;
+	}
+    } while (p > bytes);
+
+    return numBytes - (p - bytes);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclTrimLeft --
+ *
+ *	Takes two counted strings in the Tcl encoding which must both be null
+ *	terminated. Conceptually trims from the left side of the first string
+ *	all characters found in the second string.
+ *
+ * Results:
+ *	The number of bytes to be removed from the start of the string.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclTrimLeft(
+    const char *bytes,		/* String to be trimmed... */
+    int numBytes,		/* ...and its length in bytes */
+    const char *trim,		/* String of trim characters... */
+    int numTrim)		/* ...and its length in bytes */
+{
+    const char *p = bytes;
+
+    if ((bytes[numBytes] != '\0') || (trim[numTrim] != '\0')) {
+	Tcl_Panic("TclTrimLeft works only on null-terminated strings");
+    }
+
+    /*
+     * Empty strings -> nothing to do.
+     */
+
+    if ((numBytes == 0) || (numTrim == 0)) {
+	return 0;
+    }
+
+    /*
+     * Outer loop: iterate over string to be trimmed.
+     */
+
+    do {
+	Tcl_UniChar ch1;
+	int pInc = TclUtfToUniChar(p, &ch1);
+	const char *q = trim;
+	int bytesLeft = numTrim;
+
+	/*
+	 * Inner loop: scan trim string for match to current character.
+	 */
+
+	do {
+	    Tcl_UniChar ch2;
+	    int qInc = TclUtfToUniChar(q, &ch2);
+
+	    if (ch1 == ch2) {
+		break;
+	    }
+
+	    q += qInc;
+	    bytesLeft -= qInc;
+	} while (bytesLeft);
+
+	if (bytesLeft == 0) {
+	    /*
+	     * No match; trim task done; *p is first non-trimmed char.
+	     */
+
+	    break;
+	}
+
+	p += pInc;
+	numBytes -= pInc;
+    } while (numBytes);
+
+    return p - bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_Concat --
+ *
+ *	Concatenate a set of strings into a single large string.
+ *
+ * Results:
+ *	The return value is dynamically-allocated string containing a
+ *	concatenation of all the strings in argv, with spaces between the
+ *	original argv elements.
+ *
+ * Side effects:
+ *	Memory is allocated for the result; the caller is responsible for
+ *	freeing the memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+/* The whitespace characters trimmed during [concat] operations */
+#define CONCAT_WS_SIZE (int) (sizeof(CONCAT_TRIM_SET "") - 1)
+
+char *
+Tcl_Concat(
+    int argc,			/* Number of strings to concatenate. */
+    const char *const *argv)	/* Array of strings to concatenate. */
+{
+    int i, needSpace = 0, bytesNeeded = 0;
+    char *result, *p;
+
+    /*
+     * Dispose of the empty result corner case first to simplify later code.
+     */
+
+    if (argc == 0) {
+	result = (char *) ckalloc(1);
+	result[0] = '\0';
+	return result;
+    }
+
+    /*
+     * First allocate the result buffer at the size required.
+     */
+
+    for (i = 0;  i < argc;  i++) {
+	bytesNeeded += strlen(argv[i]);
+	if (bytesNeeded < 0) {
+	    Tcl_Panic("Tcl_Concat: max size of Tcl value exceeded");
+	}
+    }
+    if (bytesNeeded + argc - 1 < 0) {
+	/*
+	 * Panic test could be tighter, but not going to bother for this
+	 * legacy routine.
+	 */
+
+	Tcl_Panic("Tcl_Concat: max size of Tcl value exceeded");
+    }
+
+    /*
+     * All element bytes + (argc - 1) spaces + 1 terminating NULL.
+     */
+
+    result = ckalloc((unsigned) (bytesNeeded + argc));
+
+    for (p = result, i = 0;  i < argc;  i++) {
+	int trim, elemLength;
+	const char *element;
+
+	element = argv[i];
+	elemLength = strlen(argv[i]);
+
+	/*
+	 * Trim away the leading whitespace.
+	 */
+
+	trim = TclTrimLeft(element, elemLength, CONCAT_TRIM_SET,
+		CONCAT_WS_SIZE);
+	element += trim;
+	elemLength -= trim;
+
+	/*
+	 * Trim away the trailing whitespace. Do not permit trimming to expose
+	 * a final backslash character.
+	 */
+
+	trim = TclTrimRight(element, elemLength, CONCAT_TRIM_SET,
+		CONCAT_WS_SIZE);
+	trim -= trim && (element[elemLength - trim - 1] == '\\');
+	elemLength -= trim;
+
+	/*
+	 * If we're left with empty element after trimming, do nothing.
+	 */
+
+	if (elemLength == 0) {
+	    continue;
+	}
+
+	/*
+	 * Append to the result with space if needed.
+	 */
+
+	if (needSpace) {
+	    *p++ = ' ';
+	}
+	memcpy(p, element, (size_t) elemLength);
+	p += elemLength;
+	needSpace = 1;
+    }
+    *p = '\0';
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ConcatObj --
+ *
+ *	Concatenate the strings from a set of objects into a single string
+ *	object with spaces between the original strings.
+ *
+ * Results:
+ *	The return value is a new string object containing a concatenation of
+ *	the strings in objv. Its ref count is zero.
+ *
+ * Side effects:
+ *	A new object is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ConcatObj(
+    int objc,			/* Number of objects to concatenate. */
+    Tcl_Obj *const objv[])	/* Array of objects to concatenate. */
+{
+    int i, elemLength, needSpace = 0, bytesNeeded = 0;
+    const char *element;
+    Tcl_Obj *objPtr, *resPtr;
+
+    /*
+     * Check first to see if all the items are of list type or empty. If so,
+     * we will concat them together as lists, and return a list object. This
+     * is only valid when the lists are in canonical form.
+     */
+
+    for (i = 0;  i < objc;  i++) {
+	int length;
+
+	objPtr = objv[i];
+	if (TclListObjIsCanonical(objPtr)) {
+	    continue;
+	}
+	TclGetStringFromObj(objPtr, &length);
+	if (length > 0) {
+	    break;
+	}
+    }
+    if (i == objc) {
+	resPtr = NULL;
+	for (i = 0;  i < objc;  i++) {
+	    objPtr = objv[i];
+	    if (objPtr->bytes && objPtr->length == 0) {
+		continue;
+	    }
+	    if (resPtr) {
+		if (TCL_OK != Tcl_ListObjAppendList(NULL, resPtr, objPtr)) {
+		    /* Abandon ship! */
+		    Tcl_DecrRefCount(resPtr);
+		    goto slow;
+		}
+	    } else {
+		resPtr = TclListObjCopy(NULL, objPtr);
+	    }
+	}
+	if (!resPtr) {
+	    resPtr = Tcl_NewObj();
+	}
+	return resPtr;
+    }
+
+  slow:
+    /*
+     * Something cannot be determined to be safe, so build the concatenation
+     * the slow way, using the string representations.
+     *
+     * First try to pre-allocate the size required.
+     */
+
+    for (i = 0;  i < objc;  i++) {
+	element = TclGetStringFromObj(objv[i], &elemLength);
+	bytesNeeded += elemLength;
+	if (bytesNeeded < 0) {
+	    break;
+	}
+    }
+
+    /*
+     * Does not matter if this fails, will simply try later to build up the
+     * string with each Append reallocating as needed with the usual string
+     * append algorithm.  When that fails it will report the error.
+     */
+
+    TclNewObj(resPtr);
+    (void) Tcl_AttemptSetObjLength(resPtr, bytesNeeded + objc - 1);
+    Tcl_SetObjLength(resPtr, 0);
+
+    for (i = 0;  i < objc;  i++) {
+	int trim;
+
+	element = TclGetStringFromObj(objv[i], &elemLength);
+
+	/*
+	 * Trim away the leading whitespace.
+	 */
+
+	trim = TclTrimLeft(element, elemLength, CONCAT_TRIM_SET,
+		CONCAT_WS_SIZE);
+	element += trim;
+	elemLength -= trim;
+
+	/*
+	 * Trim away the trailing whitespace. Do not permit trimming to expose
+	 * a final backslash character.
+	 */
+
+	trim = TclTrimRight(element, elemLength, CONCAT_TRIM_SET,
+		CONCAT_WS_SIZE);
+	trim -= trim && (element[elemLength - trim - 1] == '\\');
+	elemLength -= trim;
+
+	/*
+	 * If we're left with empty element after trimming, do nothing.
+	 */
+
+	if (elemLength == 0) {
+	    continue;
+	}
+
+	/*
+	 * Append to the result with space if needed.
+	 */
+
+	if (needSpace) {
+	    Tcl_AppendToObj(resPtr, " ", 1);
+	}
+	Tcl_AppendToObj(resPtr, element, elemLength);
+	needSpace = 1;
+    }
+    return resPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_StringMatch --
+ *
+ *	See if a particular string matches a particular pattern.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_StringMatch(
+    const char *str,		/* String. */
+    const char *pattern)	/* Pattern, which may contain special
+				 * characters. */
+{
+    return Tcl_StringCaseMatch(str, pattern, 0);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_StringCaseMatch --
+ *
+ *	See if a particular string matches a particular pattern. Allows case
+ *	insensitivity.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_StringCaseMatch(
+    const char *str,		/* String. */
+    const char *pattern,	/* Pattern, which may contain special
+				 * characters. */
+    int nocase)			/* 0 for case sensitive, 1 for insensitive */
+{
+    int p, charLen;
+    const char *pstart = pattern;
+    Tcl_UniChar ch1, ch2;
+
+    while (1) {
+	p = *pattern;
+
+	/*
+	 * See if we're at the end of both the pattern and the string. If so,
+	 * we succeeded. If we're at the end of the pattern but not at the end
+	 * of the string, we failed.
+	 */
+
+	if (p == '\0') {
+	    return (*str == '\0');
+	}
+	if ((*str == '\0') && (p != '*')) {
+	    return 0;
+	}
+
+	/*
+	 * Check for a "*" as the next pattern character. It matches any
+	 * substring. We handle this by calling ourselves recursively for each
+	 * postfix of string, until either we match or we reach the end of the
+	 * string.
+	 */
+
+	if (p == '*') {
+	    /*
+	     * Skip all successive *'s in the pattern
+	     */
+
+	    while (*(++pattern) == '*') {}
+	    p = *pattern;
+	    if (p == '\0') {
+		return 1;
+	    }
+
+	    /*
+	     * This is a special case optimization for single-byte utf.
+	     */
+
+	    if (UCHAR(*pattern) < 0x80) {
+		ch2 = (Tcl_UniChar)
+			(nocase ? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
+	    } else {
+		Tcl_UtfToUniChar(pattern, &ch2);
+		if (nocase) {
+		    ch2 = Tcl_UniCharToLower(ch2);
+		}
+	    }
+
+	    while (1) {
+		/*
+		 * Optimization for matching - cruise through the string
+		 * quickly if the next char in the pattern isn't a special
+		 * character
+		 */
+
+		if ((p != '[') && (p != '?') && (p != '\\')) {
+		    if (nocase) {
+			while (*str) {
+			    charLen = TclUtfToUniChar(str, &ch1);
+			    if (ch2==ch1 || ch2==Tcl_UniCharToLower(ch1)) {
+				break;
+			    }
+			    str += charLen;
+			}
+		    } else {
+			/*
+			 * There's no point in trying to make this code
+			 * shorter, as the number of bytes you want to compare
+			 * each time is non-constant.
+			 */
+
+			while (*str) {
+			    charLen = TclUtfToUniChar(str, &ch1);
+			    if (ch2 == ch1) {
+				break;
+			    }
+			    str += charLen;
+			}
+		    }
+		}
+		if (Tcl_StringCaseMatch(str, pattern, nocase)) {
+		    return 1;
+		}
+		if (*str == '\0') {
+		    return 0;
+		}
+		str += TclUtfToUniChar(str, &ch1);
+	    }
+	}
+
+	/*
+	 * Check for a "?" as the next pattern character. It matches any
+	 * single character.
+	 */
+
+	if (p == '?') {
+	    pattern++;
+	    str += TclUtfToUniChar(str, &ch1);
+	    continue;
+	}
+
+	/*
+	 * Check for a "[" as the next pattern character. It is followed by a
+	 * list of characters that are acceptable, or by a range (two
+	 * characters separated by "-").
+	 */
+
+	if (p == '[') {
+	    Tcl_UniChar startChar, endChar;
+
+	    pattern++;
+	    if (UCHAR(*str) < 0x80) {
+		ch1 = (Tcl_UniChar)
+			(nocase ? tolower(UCHAR(*str)) : UCHAR(*str));
+		str++;
+	    } else {
+		str += Tcl_UtfToUniChar(str, &ch1);
+		if (nocase) {
+		    ch1 = Tcl_UniCharToLower(ch1);
+		}
+	    }
+	    while (1) {
+		if ((*pattern == ']') || (*pattern == '\0')) {
+		    return 0;
+		}
+		if (UCHAR(*pattern) < 0x80) {
+		    startChar = (Tcl_UniChar) (nocase
+			    ? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
+		    pattern++;
+		} else {
+		    pattern += Tcl_UtfToUniChar(pattern, &startChar);
+		    if (nocase) {
+			startChar = Tcl_UniCharToLower(startChar);
+		    }
+		}
+		if (*pattern == '-') {
+		    pattern++;
+		    if (*pattern == '\0') {
+			return 0;
+		    }
+		    if (UCHAR(*pattern) < 0x80) {
+			endChar = (Tcl_UniChar) (nocase
+				? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
+			pattern++;
+		    } else {
+			pattern += Tcl_UtfToUniChar(pattern, &endChar);
+			if (nocase) {
+			    endChar = Tcl_UniCharToLower(endChar);
+			}
+		    }
+		    if (((startChar <= ch1) && (ch1 <= endChar))
+			    || ((endChar <= ch1) && (ch1 <= startChar))) {
+			/*
+			 * Matches ranges of form [a-z] or [z-a].
+			 */
+
+			break;
+		    }
+		} else if (startChar == ch1) {
+		    break;
+		}
+	    }
+	    while (*pattern != ']') {
+		if (*pattern == '\0') {
+		    pattern = Tcl_UtfPrev(pattern, pstart);
+		    break;
+		}
+		pattern++;
+	    }
+	    pattern++;
+	    continue;
+	}
+
+	/*
+	 * If the next pattern character is '\', just strip off the '\' so we
+	 * do exact matching on the character that follows.
+	 */
+
+	if (p == '\\') {
+	    pattern++;
+	    if (*pattern == '\0') {
+		return 0;
+	    }
+	}
+
+	/*
+	 * There's no special character. Just make sure that the next bytes of
+	 * each string match.
+	 */
+
+	str += TclUtfToUniChar(str, &ch1);
+	pattern += TclUtfToUniChar(pattern, &ch2);
+	if (nocase) {
+	    if (Tcl_UniCharToLower(ch1) != Tcl_UniCharToLower(ch2)) {
+		return 0;
+	    }
+	} else if (ch1 != ch2) {
+	    return 0;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclByteArrayMatch --
+ *
+ *	See if a particular string matches a particular pattern.  Does not
+ *	allow for case insensitivity.
+ *	Parallels tclUtf.c:TclUniCharMatch, adjusted for char* and sans nocase.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclByteArrayMatch(
+    const unsigned char *string,/* String. */
+    int strLen,			/* Length of String */
+    const unsigned char *pattern,
+				/* Pattern, which may contain special
+				 * characters. */
+    int ptnLen,			/* Length of Pattern */
+    int flags)
+{
+    const unsigned char *stringEnd, *patternEnd;
+    unsigned char p;
+
+    stringEnd = string + strLen;
+    patternEnd = pattern + ptnLen;
+
+    while (1) {
+	/*
+	 * See if we're at the end of both the pattern and the string. If so,
+	 * we succeeded. If we're at the end of the pattern but not at the end
+	 * of the string, we failed.
+	 */
+
+	if (pattern == patternEnd) {
+	    return (string == stringEnd);
+	}
+	p = *pattern;
+	if ((string == stringEnd) && (p != '*')) {
+	    return 0;
+	}
+
+	/*
+	 * Check for a "*" as the next pattern character. It matches any
+	 * substring. We handle this by skipping all the characters up to the
+	 * next matching one in the pattern, and then calling ourselves
+	 * recursively for each postfix of string, until either we match or we
+	 * reach the end of the string.
+	 */
+
+	if (p == '*') {
+	    /*
+	     * Skip all successive *'s in the pattern.
+	     */
+
+	    while ((++pattern < patternEnd) && (*pattern == '*')) {
+		/* empty body */
+	    }
+	    if (pattern == patternEnd) {
+		return 1;
+	    }
+	    p = *pattern;
+	    while (1) {
+		/*
+		 * Optimization for matching - cruise through the string
+		 * quickly if the next char in the pattern isn't a special
+		 * character.
+		 */
+
+		if ((p != '[') && (p != '?') && (p != '\\')) {
+		    while ((string < stringEnd) && (p != *string)) {
+			string++;
+		    }
+		}
+		if (TclByteArrayMatch(string, stringEnd - string,
+				pattern, patternEnd - pattern, 0)) {
+		    return 1;
+		}
+		if (string == stringEnd) {
+		    return 0;
+		}
+		string++;
+	    }
+	}
+
+	/*
+	 * Check for a "?" as the next pattern character. It matches any
+	 * single character.
+	 */
+
+	if (p == '?') {
+	    pattern++;
+	    string++;
+	    continue;
+	}
+
+	/*
+	 * Check for a "[" as the next pattern character. It is followed by a
+	 * list of characters that are acceptable, or by a range (two
+	 * characters separated by "-").
+	 */
+
+	if (p == '[') {
+	    unsigned char ch1, startChar, endChar;
+
+	    pattern++;
+	    ch1 = *string;
+	    string++;
+	    while (1) {
+		if ((*pattern == ']') || (pattern == patternEnd)) {
+		    return 0;
+		}
+		startChar = *pattern;
+		pattern++;
+		if (*pattern == '-') {
+		    pattern++;
+		    if (pattern == patternEnd) {
+			return 0;
+		    }
+		    endChar = *pattern;
+		    pattern++;
+		    if (((startChar <= ch1) && (ch1 <= endChar))
+			    || ((endChar <= ch1) && (ch1 <= startChar))) {
+			/*
+			 * Matches ranges of form [a-z] or [z-a].
+			 */
+
+			break;
+		    }
+		} else if (startChar == ch1) {
+		    break;
+		}
+	    }
+	    while (*pattern != ']') {
+		if (pattern == patternEnd) {
+		    pattern--;
+		    break;
+		}
+		pattern++;
+	    }
+	    pattern++;
+	    continue;
+	}
+
+	/*
+	 * If the next pattern character is '\', just strip off the '\' so we
+	 * do exact matching on the character that follows.
+	 */
+
+	if (p == '\\') {
+	    if (++pattern == patternEnd) {
+		return 0;
+	    }
+	}
+
+	/*
+	 * There's no special character. Just make sure that the next bytes of
+	 * each string match.
+	 */
+
+	if (*string != *pattern) {
+	    return 0;
+	}
+	string++;
+	pattern++;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclStringMatchObj --
+ *
+ *	See if a particular string matches a particular pattern. Allows case
+ *	insensitivity. This is the generic multi-type handler for the various
+ *	matching algorithms.
+ *
+ * Results:
+ *	The return value is 1 if string matches pattern, and 0 otherwise. The
+ *	matching operation permits the following special characters in the
+ *	pattern: *?\[] (see the manual entry for details on what these mean).
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclStringMatchObj(
+    Tcl_Obj *strObj,		/* string object. */
+    Tcl_Obj *ptnObj,		/* pattern object. */
+    int flags)			/* Only TCL_MATCH_NOCASE should be passed, or
+				 * 0. */
+{
+    int match, length, plen;
+
+    /*
+     * Promote based on the type of incoming object.
+     * XXX: Currently doesn't take advantage of exact-ness that
+     * XXX: TclReToGlob tells us about
+    trivial = nocase ? 0 : TclMatchIsTrivial(TclGetString(ptnObj));
+     */
+
+    if ((strObj->typePtr == &tclStringType) || (strObj->typePtr == NULL)) {
+	Tcl_UniChar *udata, *uptn;
+
+	udata = Tcl_GetUnicodeFromObj(strObj, &length);
+	uptn  = Tcl_GetUnicodeFromObj(ptnObj, &plen);
+	match = TclUniCharMatch(udata, length, uptn, plen, flags);
+    } else if (TclIsPureByteArray(strObj) && !flags) {
+	unsigned char *data, *ptn;
+
+	data = Tcl_GetByteArrayFromObj(strObj, &length);
+	ptn  = Tcl_GetByteArrayFromObj(ptnObj, &plen);
+	match = TclByteArrayMatch(data, length, ptn, plen, 0);
+    } else {
+	match = Tcl_StringCaseMatch(TclGetString(strObj),
+		TclGetString(ptnObj), flags);
+    }
+    return match;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringInit --
+ *
+ *	Initializes a dynamic string, discarding any previous contents of the
+ *	string (Tcl_DStringFree should have been called already if the dynamic
+ *	string was previously in use).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The dynamic string is initialized to be empty.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringInit(
+    Tcl_DString *dsPtr)		/* Pointer to structure for dynamic string. */
+{
+    dsPtr->string = dsPtr->staticSpace;
+    dsPtr->length = 0;
+    dsPtr->spaceAvl = TCL_DSTRING_STATIC_SIZE;
+    dsPtr->staticSpace[0] = '\0';
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringAppend --
+ *
+ *	Append more bytes to the current value of a dynamic string.
+ *
+ * Results:
+ *	The return value is a pointer to the dynamic string's new value.
+ *
+ * Side effects:
+ *	Length bytes from "bytes" (or all of "bytes" if length is less than
+ *	zero) are added to the current value of the string. Memory gets
+ *	reallocated if needed to accomodate the string's new size.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_DStringAppend(
+    Tcl_DString *dsPtr,		/* Structure describing dynamic string. */
+    const char *bytes,		/* String to append. If length is -1 then this
+				 * must be null-terminated. */
+    int length)			/* Number of bytes from "bytes" to append. If
+				 * < 0, then append all of bytes, up to null
+				 * at end. */
+{
+    int newSize;
+
+    if (length < 0) {
+	length = strlen(bytes);
+    }
+    newSize = length + dsPtr->length;
+
+    /*
+     * Allocate a larger buffer for the string if the current one isn't large
+     * enough. Allocate extra space in the new buffer so that there will be
+     * room to grow before we have to allocate again.
+     */
+
+    if (newSize >= dsPtr->spaceAvl) {
+	dsPtr->spaceAvl = newSize * 2;
+	if (dsPtr->string == dsPtr->staticSpace) {
+	    char *newString = ckalloc(dsPtr->spaceAvl);
+
+	    memcpy(newString, dsPtr->string, (size_t) dsPtr->length);
+	    dsPtr->string = newString;
+	} else {
+	    int offset = -1;
+
+	    /* See [16896d49fd] */
+	    if (bytes >= dsPtr->string
+		    && bytes <= dsPtr->string + dsPtr->length) {
+		offset = bytes - dsPtr->string;
+	    }
+
+	    dsPtr->string = ckrealloc(dsPtr->string, dsPtr->spaceAvl);
+
+	    if (offset >= 0) {
+		bytes = dsPtr->string + offset;
+	    }
+	}
+    }
+
+    /*
+     * Copy the new string into the buffer at the end of the old one.
+     */
+
+    memcpy(dsPtr->string + dsPtr->length, bytes, length);
+    dsPtr->length += length;
+    dsPtr->string[dsPtr->length] = '\0';
+    return dsPtr->string;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDStringAppendObj, TclDStringAppendDString --
+ *
+ *	Simple wrappers round Tcl_DStringAppend that make it easier to append
+ *	from particular sources of strings.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+TclDStringAppendObj(
+    Tcl_DString *dsPtr,
+    Tcl_Obj *objPtr)
+{
+    int length;
+    char *bytes = TclGetStringFromObj(objPtr, &length);
+
+    return Tcl_DStringAppend(dsPtr, bytes, length);
+}
+
+char *
+TclDStringAppendDString(
+    Tcl_DString *dsPtr,
+    Tcl_DString *toAppendPtr)
+{
+    return Tcl_DStringAppend(dsPtr, Tcl_DStringValue(toAppendPtr),
+	    Tcl_DStringLength(toAppendPtr));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringAppendElement --
+ *
+ *	Append a list element to the current value of a dynamic string.
+ *
+ * Results:
+ *	The return value is a pointer to the dynamic string's new value.
+ *
+ * Side effects:
+ *	String is reformatted as a list element and added to the current value
+ *	of the string. Memory gets reallocated if needed to accomodate the
+ *	string's new size.
+ *
+ *----------------------------------------------------------------------
+ */
+
+char *
+Tcl_DStringAppendElement(
+    Tcl_DString *dsPtr,		/* Structure describing dynamic string. */
+    const char *element)	/* String to append. Must be
+				 * null-terminated. */
+{
+    char *dst = dsPtr->string + dsPtr->length;
+    int needSpace = TclNeedSpace(dsPtr->string, dst);
+    int flags = needSpace ? TCL_DONT_QUOTE_HASH : 0;
+    int newSize = dsPtr->length + needSpace
+	    + TclScanElement(element, -1, &flags);
+
+    /*
+     * Allocate a larger buffer for the string if the current one isn't large
+     * enough. Allocate extra space in the new buffer so that there will be
+     * room to grow before we have to allocate again. SPECIAL NOTE: must use
+     * memcpy, not strcpy, to copy the string to a larger buffer, since there
+     * may be embedded NULLs in the string in some cases.
+     */
+
+    if (newSize >= dsPtr->spaceAvl) {
+	dsPtr->spaceAvl = newSize * 2;
+	if (dsPtr->string == dsPtr->staticSpace) {
+	    char *newString = ckalloc(dsPtr->spaceAvl);
+
+	    memcpy(newString, dsPtr->string, (size_t) dsPtr->length);
+	    dsPtr->string = newString;
+	} else {
+	    int offset = -1;
+
+	    /* See [16896d49fd] */
+	    if (element >= dsPtr->string
+		    && element <= dsPtr->string + dsPtr->length) {
+		offset = element - dsPtr->string;
+	    }
+
+	    dsPtr->string = ckrealloc(dsPtr->string, dsPtr->spaceAvl);
+
+	    if (offset >= 0) {
+		element = dsPtr->string + offset;
+	    }
+	}
+	dst = dsPtr->string + dsPtr->length;
+    }
+
+    /*
+     * Convert the new string to a list element and copy it into the buffer at
+     * the end, with a space, if needed.
+     */
+
+    if (needSpace) {
+	*dst = ' ';
+	dst++;
+	dsPtr->length++;
+
+	/*
+	 * If we need a space to separate this element from preceding stuff,
+	 * then this element will not lead a list, and need not have it's
+	 * leading '#' quoted.
+	 */
+
+	flags |= TCL_DONT_QUOTE_HASH;
+    }
+    dsPtr->length += TclConvertElement(element, -1, dst, flags);
+    dsPtr->string[dsPtr->length] = '\0';
+    return dsPtr->string;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringSetLength --
+ *
+ *	Change the length of a dynamic string. This can cause the string to
+ *	either grow or shrink, depending on the value of length.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The length of dsPtr is changed to length and a null byte is stored at
+ *	that position in the string. If length is larger than the space
+ *	allocated for dsPtr, then a panic occurs.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringSetLength(
+    Tcl_DString *dsPtr,		/* Structure describing dynamic string. */
+    int length)			/* New length for dynamic string. */
+{
+    int newsize;
+
+    if (length < 0) {
+	length = 0;
+    }
+    if (length >= dsPtr->spaceAvl) {
+	/*
+	 * There are two interesting cases here. In the first case, the user
+	 * may be trying to allocate a large buffer of a specific size. It
+	 * would be wasteful to overallocate that buffer, so we just allocate
+	 * enough for the requested size plus the trailing null byte. In the
+	 * second case, we are growing the buffer incrementally, so we need
+	 * behavior similar to Tcl_DStringAppend. The requested length will
+	 * usually be a small delta above the current spaceAvl, so we'll end
+	 * up doubling the old size. This won't grow the buffer quite as
+	 * quickly, but it should be close enough.
+	 */
+
+	newsize = dsPtr->spaceAvl * 2;
+	if (length < newsize) {
+	    dsPtr->spaceAvl = newsize;
+	} else {
+	    dsPtr->spaceAvl = length + 1;
+	}
+	if (dsPtr->string == dsPtr->staticSpace) {
+	    char *newString = ckalloc(dsPtr->spaceAvl);
+
+	    memcpy(newString, dsPtr->string, (size_t) dsPtr->length);
+	    dsPtr->string = newString;
+	} else {
+	    dsPtr->string = ckrealloc(dsPtr->string, dsPtr->spaceAvl);
+	}
+    }
+    dsPtr->length = length;
+    dsPtr->string[length] = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringFree --
+ *
+ *	Frees up any memory allocated for the dynamic string and reinitializes
+ *	the string to an empty state.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The previous contents of the dynamic string are lost, and the new
+ *	value is an empty string.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringFree(
+    Tcl_DString *dsPtr)		/* Structure describing dynamic string. */
+{
+    if (dsPtr->string != dsPtr->staticSpace) {
+	ckfree(dsPtr->string);
+    }
+    dsPtr->string = dsPtr->staticSpace;
+    dsPtr->length = 0;
+    dsPtr->spaceAvl = TCL_DSTRING_STATIC_SIZE;
+    dsPtr->staticSpace[0] = '\0';
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringResult --
+ *
+ *	This function moves the value of a dynamic string into an interpreter
+ *	as its string result. Afterwards, the dynamic string is reset to an
+ *	empty string.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The string is "moved" to interp's result, and any existing string
+ *	result for interp is freed. dsPtr is reinitialized to an empty string.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringResult(
+    Tcl_Interp *interp,		/* Interpreter whose result is to be reset. */
+    Tcl_DString *dsPtr)		/* Dynamic string that is to become the
+				 * result of interp. */
+{
+    Tcl_SetObjResult(interp, TclDStringToObj(dsPtr));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringGetResult --
+ *
+ *	This function moves an interpreter's result into a dynamic string.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interpreter's string result is cleared, and the previous contents
+ *	of dsPtr are freed.
+ *
+ *	If the string result is empty, the object result is moved to the
+ *	string result, then the object result is reset.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringGetResult(
+    Tcl_Interp *interp,		/* Interpreter whose result is to be reset. */
+    Tcl_DString *dsPtr)		/* Dynamic string that is to become the result
+				 * of interp. */
+{
+#ifdef TCL_NO_DEPRECATED
+    Tcl_Obj *obj = Tcl_GetObjResult(interp);
+    const char *bytes = TclGetString(obj);
+
+    Tcl_DStringFree(dsPtr);
+    Tcl_DStringAppend(dsPtr, bytes, obj->length);
+    Tcl_ResetResult(interp);
+#else
+    Interp *iPtr = (Interp *) interp;
+
+    if (dsPtr->string != dsPtr->staticSpace) {
+	ckfree(dsPtr->string);
+    }
+
+    /*
+     * Do more efficient transfer when we know the result is a Tcl_Obj. When
+     * there's no string result, we only have to deal with two cases:
+     *
+     *  1. When the string rep is the empty string, when we don't copy but
+     *     instead use the staticSpace in the DString to hold an empty string.
+
+     *  2. When the string rep is not there or there's a real string rep, when
+     *     we use Tcl_GetString to fetch (or generate) the string rep - which
+     *     we know to have been allocated with ckalloc() - and use it to
+     *     populate the DString space. Then, we free the internal rep. and set
+     *     the object's string representation back to the canonical empty
+     *     string.
+     */
+
+    if (!iPtr->result[0] && iPtr->objResultPtr
+	    && !Tcl_IsShared(iPtr->objResultPtr)) {
+	if (iPtr->objResultPtr->bytes == &tclEmptyString) {
+	    dsPtr->string = dsPtr->staticSpace;
+	    dsPtr->string[0] = 0;
+	    dsPtr->length = 0;
+	    dsPtr->spaceAvl = TCL_DSTRING_STATIC_SIZE;
+	} else {
+	    dsPtr->string = TclGetString(iPtr->objResultPtr);
+	    dsPtr->length = iPtr->objResultPtr->length;
+	    dsPtr->spaceAvl = dsPtr->length + 1;
+	    TclFreeIntRep(iPtr->objResultPtr);
+	    iPtr->objResultPtr->bytes = &tclEmptyString;
+	    iPtr->objResultPtr->length = 0;
+	}
+	return;
+    }
+
+    /*
+     * If the string result is empty, move the object result to the string
+     * result, then reset the object result.
+     */
+
+    (void) Tcl_GetStringResult(interp);
+
+    dsPtr->length = strlen(iPtr->result);
+    if (iPtr->freeProc != NULL) {
+	if (iPtr->freeProc == TCL_DYNAMIC) {
+	    dsPtr->string = iPtr->result;
+	    dsPtr->spaceAvl = dsPtr->length+1;
+	} else {
+	    dsPtr->string = ckalloc(dsPtr->length+1);
+	    memcpy(dsPtr->string, iPtr->result, (unsigned) dsPtr->length+1);
+	    iPtr->freeProc(iPtr->result);
+	}
+	dsPtr->spaceAvl = dsPtr->length+1;
+	iPtr->freeProc = NULL;
+    } else {
+	if (dsPtr->length < TCL_DSTRING_STATIC_SIZE) {
+	    dsPtr->string = dsPtr->staticSpace;
+	    dsPtr->spaceAvl = TCL_DSTRING_STATIC_SIZE;
+	} else {
+	    dsPtr->string = ckalloc(dsPtr->length+1);
+	    dsPtr->spaceAvl = dsPtr->length + 1;
+	}
+	memcpy(dsPtr->string, iPtr->result, (unsigned) dsPtr->length+1);
+    }
+
+    iPtr->result = iPtr->resultSpace;
+    iPtr->resultSpace[0] = 0;
+#endif /* !TCL_NO_DEPRECATED */
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDStringToObj --
+ *
+ *	This function moves a dynamic string's contents to a new Tcl_Obj. Be
+ *	aware that this function does *not* check that the encoding of the
+ *	contents of the dynamic string is correct; this is the caller's
+ *	responsibility to enforce.
+ *
+ * Results:
+ *	The newly-allocated untyped (i.e., typePtr==NULL) Tcl_Obj with a
+ *	reference count of zero.
+ *
+ * Side effects:
+ *	The string is "moved" to the object. dsPtr is reinitialized to an
+ *	empty string; it does not need to be Tcl_DStringFree'd after this if
+ *	not used further.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclDStringToObj(
+    Tcl_DString *dsPtr)
+{
+    Tcl_Obj *result;
+
+    if (dsPtr->string == dsPtr->staticSpace) {
+	if (dsPtr->length == 0) {
+	    TclNewObj(result);
+	} else {
+	    /*
+	     * Static buffer, so must copy.
+	     */
+
+	    TclNewStringObj(result, dsPtr->string, dsPtr->length);
+	}
+    } else {
+	/*
+	 * Dynamic buffer, so transfer ownership and reset.
+	 */
+
+	TclNewObj(result);
+	result->bytes = dsPtr->string;
+	result->length = dsPtr->length;
+    }
+
+    /*
+     * Re-establish the DString as empty with no buffer allocated.
+     */
+
+    dsPtr->string = dsPtr->staticSpace;
+    dsPtr->spaceAvl = TCL_DSTRING_STATIC_SIZE;
+    dsPtr->length = 0;
+    dsPtr->staticSpace[0] = '\0';
+
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringStartSublist --
+ *
+ *	This function adds the necessary information to a dynamic string
+ *	(e.g. " {") to start a sublist. Future element appends will be in the
+ *	sublist rather than the main list.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Characters get added to the dynamic string.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringStartSublist(
+    Tcl_DString *dsPtr)		/* Dynamic string. */
+{
+    if (TclNeedSpace(dsPtr->string, dsPtr->string + dsPtr->length)) {
+	TclDStringAppendLiteral(dsPtr, " {");
+    } else {
+	TclDStringAppendLiteral(dsPtr, "{");
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_DStringEndSublist --
+ *
+ *	This function adds the necessary characters to a dynamic string to end
+ *	a sublist (e.g. "}"). Future element appends will be in the enclosing
+ *	(sub)list rather than the current sublist.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_DStringEndSublist(
+    Tcl_DString *dsPtr)		/* Dynamic string. */
+{
+    TclDStringAppendLiteral(dsPtr, "}");
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_PrintDouble --
+ *
+ *	Given a floating-point value, this function converts it to an ASCII
+ *	string using.
+ *
+ * Results:
+ *	The ASCII equivalent of "value" is written at "dst". It is written
+ *	using the current precision, and it is guaranteed to contain a decimal
+ *	point or exponent, so that it looks like a floating-point value and
+ *	not an integer.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_PrintDouble(
+    Tcl_Interp *interp,		/* Interpreter whose tcl_precision variable
+				 * used to be used to control printing. It's
+				 * ignored now. */
+    double value,		/* Value to print as string. */
+    char *dst)			/* Where to store converted value; must have
+				 * at least TCL_DOUBLE_SPACE characters. */
+{
+    char *p, c;
+    int exponent;
+    int signum;
+    char *digits;
+    char *end;
+    int *precisionPtr = Tcl_GetThreadData(&precisionKey, sizeof(int));
+
+    /*
+     * Handle NaN.
+     */
+
+    if (TclIsNaN(value)) {
+	TclFormatNaN(value, dst);
+	return;
+    }
+
+    /*
+     * Handle infinities.
+     */
+
+    if (TclIsInfinite(value)) {
+	/*
+	 * Remember to copy the terminating NUL too.
+	 */
+
+	if (value < 0) {
+	    memcpy(dst, "-Inf", 5);
+	} else {
+	    memcpy(dst, "Inf", 4);
+	}
+	return;
+    }
+
+    /*
+     * Ordinary (normal and denormal) values.
+     */
+
+    if (*precisionPtr == 0) {
+	digits = TclDoubleDigits(value, -1, TCL_DD_SHORTEST,
+		&exponent, &signum, &end);
+    } else {
+	/*
+	 * There are at least two possible interpretations for tcl_precision.
+	 *
+	 * The first is, "choose the decimal representation having
+	 * $tcl_precision digits of significance that is nearest to the given
+	 * number, breaking ties by rounding to even, and then trimming
+	 * trailing zeros." This gives the greatest possible precision in the
+	 * decimal string, but offers the anomaly that [expr 0.1] will be
+	 * "0.10000000000000001".
+	 *
+	 * The second is "choose the decimal representation having at most
+	 * $tcl_precision digits of significance that is nearest to the given
+	 * number. If no such representation converts exactly to the given
+	 * number, choose the one that is closest, breaking ties by rounding
+	 * to even. If more than one such representation converts exactly to
+	 * the given number, choose the shortest, breaking ties in favour of
+	 * the nearest, breaking remaining ties in favour of the one ending in
+	 * an even digit."
+	 *
+	 * Tcl 8.4 implements the first of these, which gives rise to
+	 * anomalies in formatting:
+	 *
+	 *	% expr 0.1
+	 *	0.10000000000000001
+	 *	% expr 0.01
+	 *	0.01
+	 *	% expr 1e-7
+	 *	9.9999999999999995e-08
+	 *
+	 * For human readability, it appears better to choose the second rule,
+	 * and let [expr 0.1] return 0.1. But for 8.4 compatibility, we prefer
+	 * the first (the recommended zero value for tcl_precision avoids the
+	 * problem entirely).
+	 *
+	 * Uncomment TCL_DD_SHORTEN_FLAG in the next call to prefer the method
+	 * that allows floating point values to be shortened if it can be done
+	 * without loss of precision.
+	 */
+
+	digits = TclDoubleDigits(value, *precisionPtr,
+		TCL_DD_E_FORMAT /* | TCL_DD_SHORTEN_FLAG */,
+		&exponent, &signum, &end);
+    }
+    if (signum) {
+	*dst++ = '-';
+    }
+    p = digits;
+    if (exponent < -4 || exponent > 16) {
+	/*
+	 * E format for numbers < 1e-3 or >= 1e17.
+	 */
+
+	*dst++ = *p++;
+	c = *p;
+	if (c != '\0') {
+	    *dst++ = '.';
+	    while (c != '\0') {
+		*dst++ = c;
+		c = *++p;
+	    }
+	}
+
+	/*
+	 * Tcl 8.4 appears to format with at least a two-digit exponent;
+	 * preserve that behaviour when tcl_precision != 0
+	 */
+
+	if (*precisionPtr == 0) {
+	    sprintf(dst, "e%+d", exponent);
+	} else {
+	    sprintf(dst, "e%+03d", exponent);
+	}
+    } else {
+	/*
+	 * F format for others.
+	 */
+
+	if (exponent < 0) {
+	    *dst++ = '0';
+	}
+	c = *p;
+	while (exponent-- >= 0) {
+	    if (c != '\0') {
+		*dst++ = c;
+		c = *++p;
+	    } else {
+		*dst++ = '0';
+	    }
+	}
+	*dst++ = '.';
+	if (c == '\0') {
+	    *dst++ = '0';
+	} else {
+	    while (++exponent < -1) {
+		*dst++ = '0';
+	    }
+	    while (c != '\0') {
+		*dst++ = c;
+		c = *++p;
+	    }
+	}
+	*dst++ = '\0';
+    }
+    ckfree(digits);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPrecTraceProc --
+ *
+ *	This function is invoked whenever the variable "tcl_precision" is
+ *	written.
+ *
+ * Results:
+ *	Returns NULL if all went well, or an error message if the new value
+ *	for the variable doesn't make sense.
+ *
+ * Side effects:
+ *	If the new value doesn't make sense then this function undoes the
+ *	effect of the variable modification. Otherwise it modifies the format
+ *	string that's used by Tcl_PrintDouble.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+char *
+TclPrecTraceProc(
+    ClientData clientData,	/* Not used. */
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const char *name1,		/* Name of variable. */
+    const char *name2,		/* Second part of variable name. */
+    int flags)			/* Information about what happened. */
+{
+    Tcl_Obj *value;
+    int prec;
+    int *precisionPtr = Tcl_GetThreadData(&precisionKey, sizeof(int));
+
+    /*
+     * If the variable is unset, then recreate the trace.
+     */
+
+    if (flags & TCL_TRACE_UNSETS) {
+	if ((flags & TCL_TRACE_DESTROYED) && !Tcl_InterpDeleted(interp)) {
+	    Tcl_TraceVar2(interp, name1, name2,
+		    TCL_GLOBAL_ONLY|TCL_TRACE_READS|TCL_TRACE_WRITES
+		    |TCL_TRACE_UNSETS, TclPrecTraceProc, clientData);
+	}
+	return NULL;
+    }
+
+    /*
+     * When the variable is read, reset its value from our shared value. This
+     * is needed in case the variable was modified in some other interpreter
+     * so that this interpreter's value is out of date.
+     */
+
+
+    if (flags & TCL_TRACE_READS) {
+	Tcl_SetVar2Ex(interp, name1, name2, Tcl_NewIntObj(*precisionPtr),
+		flags & TCL_GLOBAL_ONLY);
+	return NULL;
+    }
+
+    /*
+     * The variable is being written. Check the new value and disallow it if
+     * it isn't reasonable or if this is a safe interpreter (we don't want
+     * safe interpreters messing up the precision of other interpreters).
+     */
+
+    if (Tcl_IsSafe(interp)) {
+	return (char *) "can't modify precision from a safe interpreter";
+    }
+    value = Tcl_GetVar2Ex(interp, name1, name2, flags & TCL_GLOBAL_ONLY);
+    if (value == NULL
+	    || Tcl_GetIntFromObj(NULL, value, &prec) != TCL_OK
+	    || prec < 0 || prec > TCL_MAX_PREC) {
+	return (char *) "improper value for precision";
+    }
+    *precisionPtr = prec;
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclNeedSpace --
+ *
+ *	This function checks to see whether it is appropriate to add a space
+ *	before appending a new list element to an existing string.
+ *
+ * Results:
+ *	The return value is 1 if a space is appropriate, 0 otherwise.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclNeedSpace(
+    const char *start,		/* First character in string. */
+    const char *end)		/* End of string (place where space will be
+				 * added, if appropriate). */
+{
+    /*
+     * A space is needed unless either:
+     * (a) we're at the start of the string, or
+     */
+
+    if (end == start) {
+	return 0;
+    }
+
+    /*
+     * (b) we're at the start of a nested list-element, quoted with an open
+     *	   curly brace; we can be nested arbitrarily deep, so long as the
+     *	   first curly brace starts an element, so backtrack over open curly
+     *	   braces that are trailing characters of the string; and
+     */
+
+    end = Tcl_UtfPrev(end, start);
+    while (*end == '{') {
+	if (end == start) {
+	    return 0;
+	}
+	end = Tcl_UtfPrev(end, start);
+    }
+
+    /*
+     * (c) the trailing character of the string is already a list-element
+     *	   separator (according to TclFindElement); that is, one of these
+     *	   characters:
+     *		\u0009	\t	TAB
+     *		\u000A	\n	NEWLINE
+     *		\u000B	\v	VERTICAL TAB
+     *		\u000C	\f	FORM FEED
+     *		\u000D	\r	CARRIAGE RETURN
+     *		\u0020		SPACE
+     *	   with the condition that the penultimate character is not a
+     *	   backslash.
+     */
+
+    if (*end > 0x20) {
+	/*
+	 * Performance tweak. All ASCII spaces are <= 0x20. So get a quick
+	 * answer for most characters before comparing against all spaces in
+	 * the switch below.
+	 *
+	 * NOTE: Remove this if other Unicode spaces ever get accepted as
+	 * list-element separators.
+	 */
+
+	return 1;
+    }
+    switch (*end) {
+    case ' ':
+    case '\t':
+    case '\n':
+    case '\r':
+    case '\v':
+    case '\f':
+	if ((end == start) || (end[-1] != '\\')) {
+	    return 0;
+	}
+    }
+    return 1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFormatInt --
+ *
+ *	This procedure formats an integer into a sequence of decimal digit
+ *	characters in a buffer. If the integer is negative, a minus sign is
+ *	inserted at the start of the buffer. A null character is inserted at
+ *	the end of the formatted characters. It is the caller's responsibility
+ *	to ensure that enough storage is available. This procedure has the
+ *	effect of sprintf(buffer, "%ld", n) but is faster as proven in
+ *	benchmarks.  This is key to UpdateStringOfInt, which is a common path
+ *	for a lot of code (e.g. int-indexed arrays).
+ *
+ * Results:
+ *	An integer representing the number of characters formatted, not
+ *	including the terminating \0.
+ *
+ * Side effects:
+ *	The formatted characters are written into the storage pointer to by
+ *	the "buffer" argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclFormatInt(
+    char *buffer,		/* Points to the storage into which the
+				 * formatted characters are written. */
+    long n)			/* The integer to format. */
+{
+    long intVal;
+    int i;
+    int numFormatted, j;
+    const char *digits = "0123456789";
+
+    /*
+     * Check first whether "n" is zero.
+     */
+
+    if (n == 0) {
+	buffer[0] = '0';
+	buffer[1] = 0;
+	return 1;
+    }
+
+    /*
+     * Check whether "n" is the maximum negative value. This is -2^(m-1) for
+     * an m-bit word, and has no positive equivalent; negating it produces the
+     * same value.
+     */
+
+    intVal = -n;			/* [Bug 3390638] Workaround for*/
+    if (n == -n || intVal == n) {	/* broken compiler optimizers. */
+	return sprintf(buffer, "%ld", n);
+    }
+
+    /*
+     * Generate the characters of the result backwards in the buffer.
+     */
+
+    intVal = (n < 0? -n : n);
+    i = 0;
+    buffer[0] = '\0';
+    do {
+	i++;
+	buffer[i] = digits[intVal % 10];
+	intVal = intVal/10;
+    } while (intVal > 0);
+    if (n < 0) {
+	i++;
+	buffer[i] = '-';
+    }
+    numFormatted = i;
+
+    /*
+     * Now reverse the characters.
+     */
+
+    for (j = 0;  j < i;  j++, i--) {
+	char tmp = buffer[i];
+
+	buffer[i] = buffer[j];
+	buffer[j] = tmp;
+    }
+    return numFormatted;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetIntForIndex --
+ *
+ *	This function returns an integer corresponding to the list index held
+ *	in a Tcl object. The Tcl object's value is expected to be in the
+ *	format integer([+-]integer)? or the format end([+-]integer)?.
+ *
+ * Results:
+ *	The return value is normally TCL_OK, which means that the index was
+ *	successfully stored into the location referenced by "indexPtr". If the
+ *	Tcl object referenced by "objPtr" has the value "end", the value
+ *	stored is "endValue". If "objPtr"s values is not of one of the
+ *	expected formats, TCL_ERROR is returned and, if "interp" is non-NULL,
+ *	an error message is left in the interpreter's result object.
+ *
+ * Side effects:
+ *	The object referenced by "objPtr" might be converted to an integer,
+ *	wide integer, or end-based-index object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclGetIntForIndex(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, then no error message is left after
+				 * errors. */
+    Tcl_Obj *objPtr,		/* Points to an object containing either "end"
+				 * or an integer. */
+    int endValue,		/* The value to be stored at "indexPtr" if
+				 * "objPtr" holds "end". */
+    int *indexPtr)		/* Location filled in with an integer
+				 * representing an index. */
+{
+    size_t length;
+    char *opPtr;
+    const char *bytes;
+
+    if (TclGetIntFromObj(NULL, objPtr, indexPtr) == TCL_OK) {
+	return TCL_OK;
+    }
+
+    if (SetEndOffsetFromAny(NULL, objPtr) == TCL_OK) {
+	/*
+	 * If the object is already an offset from the end of the list, or can
+	 * be converted to one, use it.
+	 */
+
+	*indexPtr = endValue + objPtr->internalRep.longValue;
+	return TCL_OK;
+    }
+
+    bytes = TclGetString(objPtr);
+    length = objPtr->length;
+
+    /*
+     * Leading whitespace is acceptable in an index.
+     */
+
+    while (length && TclIsSpaceProc(*bytes)) {
+	bytes++;
+	length--;
+    }
+
+    if (TclParseNumber(NULL, NULL, NULL, bytes, length, (const char **)&opPtr,
+	    TCL_PARSE_INTEGER_ONLY | TCL_PARSE_NO_WHITESPACE) == TCL_OK) {
+	int code, first, second;
+	char savedOp = *opPtr;
+
+	if ((savedOp != '+') && (savedOp != '-')) {
+	    goto parseError;
+	}
+	if (TclIsSpaceProc(opPtr[1])) {
+	    goto parseError;
+	}
+	*opPtr = '\0';
+	code = Tcl_GetInt(interp, bytes, &first);
+	*opPtr = savedOp;
+	if (code == TCL_ERROR) {
+	    goto parseError;
+	}
+	if (TCL_ERROR == Tcl_GetInt(interp, opPtr+1, &second)) {
+	    goto parseError;
+	}
+	if (savedOp == '+') {
+	    *indexPtr = first + second;
+	} else {
+	    *indexPtr = first - second;
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * Report a parse error.
+     */
+
+  parseError:
+    if (interp != NULL) {
+	bytes = TclGetString(objPtr);
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad index \"%s\": must be integer?[+-]integer? or"
+		" end?[+-]integer?", bytes));
+	if (!strncmp(bytes, "end-", 4)) {
+	    bytes += 4;
+	}
+	TclCheckBadOctal(interp, bytes);
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "INDEX", NULL);
+    }
+
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UpdateStringOfEndOffset --
+ *
+ *	Update the string rep of a Tcl object holding an "end-offset"
+ *	expression.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores a valid string in the object's string rep.
+ *
+ * This function does NOT free any earlier string rep. If it is called on an
+ * object that already has a valid string rep, it will leak memory.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UpdateStringOfEndOffset(
+    register Tcl_Obj *objPtr)
+{
+    char buffer[TCL_INTEGER_SPACE + 5];
+    register int len = 3;
+
+    memcpy(buffer, "end", 4);
+    if (objPtr->internalRep.longValue != 0) {
+	buffer[len++] = '-';
+	len += TclFormatInt(buffer+len, -(objPtr->internalRep.longValue));
+    }
+    objPtr->bytes = ckalloc((unsigned) len+1);
+    memcpy(objPtr->bytes, buffer, (unsigned) len+1);
+    objPtr->length = len;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * SetEndOffsetFromAny --
+ *
+ *	Look for a string of the form "end[+-]offset" and convert it to an
+ *	internal representation holding the offset.
+ *
+ * Results:
+ *	Returns TCL_OK if ok, TCL_ERROR if the string was badly formed.
+ *
+ * Side effects:
+ *	If interp is not NULL, stores an error message in the interpreter
+ *	result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+SetEndOffsetFromAny(
+    Tcl_Interp *interp,		/* Tcl interpreter or NULL */
+    Tcl_Obj *objPtr)		/* Pointer to the object to parse */
+{
+    int offset;			/* Offset in the "end-offset" expression */
+    register const char *bytes;	/* String rep of the object */
+    int length;			/* Length of the object's string rep */
+
+    /*
+     * If it's already the right type, we're fine.
+     */
+
+    if (objPtr->typePtr == &tclEndOffsetType) {
+	return TCL_OK;
+    }
+
+    /*
+     * Check for a string rep of the right form.
+     */
+
+    bytes = TclGetStringFromObj(objPtr, &length);
+    if ((*bytes != 'e') || (strncmp(bytes, "end",
+	    (size_t)((length > 3) ? 3 : length)) != 0)) {
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad index \"%s\": must be end?[+-]integer?", bytes));
+	    Tcl_SetErrorCode(interp, "TCL", "VALUE", "INDEX", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * Convert the string rep.
+     */
+
+    if (length <= 3) {
+	offset = 0;
+    } else if ((length > 4) && ((bytes[3] == '-') || (bytes[3] == '+'))) {
+	/*
+	 * This is our limited string expression evaluator. Pass everything
+	 * after "end-" to Tcl_GetInt, then reverse for offset.
+	 */
+
+	if (TclIsSpaceProc(bytes[4])) {
+	    goto badIndexFormat;
+	}
+	if (Tcl_GetInt(interp, bytes+4, &offset) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (bytes[3] == '-') {
+	    offset = -offset;
+	}
+    } else {
+	/*
+	 * Conversion failed. Report the error.
+	 */
+
+    badIndexFormat:
+	if (interp != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "bad index \"%s\": must be end?[+-]integer?", bytes));
+	    Tcl_SetErrorCode(interp, "TCL", "VALUE", "INDEX", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    /*
+     * The conversion succeeded. Free the old internal rep and set the new
+     * one.
+     */
+
+    TclFreeIntRep(objPtr);
+    objPtr->internalRep.longValue = offset;
+    objPtr->typePtr = &tclEndOffsetType;
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCheckBadOctal --
+ *
+ *	This function checks for a bad octal value and appends a meaningful
+ *	error to the interp's result.
+ *
+ * Results:
+ *	1 if the argument was a bad octal, else 0.
+ *
+ * Side effects:
+ *	The interpreter's result is modified.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclCheckBadOctal(
+    Tcl_Interp *interp,		/* Interpreter to use for error reporting. If
+				 * NULL, then no error message is left after
+				 * errors. */
+    const char *value)		/* String to check. */
+{
+    register const char *p = value;
+
+    /*
+     * A frequent mistake is invalid octal values due to an unwanted leading
+     * zero. Try to generate a meaningful error message.
+     */
+
+    while (TclIsSpaceProc(*p)) {
+	p++;
+    }
+    if (*p == '+' || *p == '-') {
+	p++;
+    }
+    if (*p == '0') {
+	if ((p[1] == 'o') || p[1] == 'O') {
+	    p += 2;
+	}
+	while (isdigit(UCHAR(*p))) {	/* INTL: digit. */
+	    p++;
+	}
+	while (TclIsSpaceProc(*p)) {
+	    p++;
+	}
+	if (*p == '\0') {
+	    /*
+	     * Reached end of string.
+	     */
+
+	    if (interp != NULL) {
+		/*
+		 * Don't reset the result here because we want this result to
+		 * be added to an existing error message as extra info.
+		 */
+
+		Tcl_AppendToObj(Tcl_GetObjResult(interp),
+			" (looks like invalid octal number)", -1);
+	    }
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ClearHash --
+ *
+ *	Remove all the entries in the hash table *tablePtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ClearHash(
+    Tcl_HashTable *tablePtr)
+{
+    Tcl_HashSearch search;
+    Tcl_HashEntry *hPtr;
+
+    for (hPtr = Tcl_FirstHashEntry(tablePtr, &search); hPtr != NULL;
+	    hPtr = Tcl_NextHashEntry(&search)) {
+	Tcl_Obj *objPtr = Tcl_GetHashValue(hPtr);
+
+	Tcl_DecrRefCount(objPtr);
+	Tcl_DeleteHashEntry(hPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GetThreadHash --
+ *
+ *	Get a thread-specific (Tcl_HashTable *) associated with a thread data
+ *	key.
+ *
+ * Results:
+ *	The Tcl_HashTable * corresponding to *keyPtr.
+ *
+ * Side effects:
+ *	The first call on a keyPtr in each thread creates a new Tcl_HashTable,
+ *	and registers a thread exit handler to dispose of it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_HashTable *
+GetThreadHash(
+    Tcl_ThreadDataKey *keyPtr)
+{
+    Tcl_HashTable **tablePtrPtr =
+	    Tcl_GetThreadData(keyPtr, sizeof(Tcl_HashTable *));
+
+    if (NULL == *tablePtrPtr) {
+	*tablePtrPtr = ckalloc(sizeof(Tcl_HashTable));
+	Tcl_CreateThreadExitHandler(FreeThreadHash, *tablePtrPtr);
+	Tcl_InitHashTable(*tablePtrPtr, TCL_ONE_WORD_KEYS);
+    }
+    return *tablePtrPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeThreadHash --
+ *
+ *	Thread exit handler used by GetThreadHash to dispose of a thread hash
+ *	table.
+ *
+ * Side effects:
+ *	Frees a Tcl_HashTable.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeThreadHash(
+    ClientData clientData)
+{
+    Tcl_HashTable *tablePtr = clientData;
+
+    ClearHash(tablePtr);
+    Tcl_DeleteHashTable(tablePtr);
+    ckfree(tablePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * FreeProcessGlobalValue --
+ *
+ *	Exit handler used by Tcl(Set|Get)ProcessGlobalValue to cleanup a
+ *	ProcessGlobalValue at exit.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+FreeProcessGlobalValue(
+    ClientData clientData)
+{
+    ProcessGlobalValue *pgvPtr = clientData;
+
+    pgvPtr->epoch++;
+    pgvPtr->numBytes = 0;
+    ckfree(pgvPtr->value);
+    pgvPtr->value = NULL;
+    if (pgvPtr->encoding) {
+	Tcl_FreeEncoding(pgvPtr->encoding);
+	pgvPtr->encoding = NULL;
+    }
+    Tcl_MutexFinalize(&pgvPtr->mutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetProcessGlobalValue --
+ *
+ *	Utility routine to set a global value shared by all threads in the
+ *	process while keeping a thread-local copy as well.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetProcessGlobalValue(
+    ProcessGlobalValue *pgvPtr,
+    Tcl_Obj *newValue,
+    Tcl_Encoding encoding)
+{
+    const char *bytes;
+    Tcl_HashTable *cacheMap;
+    Tcl_HashEntry *hPtr;
+    int dummy;
+
+    Tcl_MutexLock(&pgvPtr->mutex);
+
+    /*
+     * Fill the global string value.
+     */
+
+    pgvPtr->epoch++;
+    if (NULL != pgvPtr->value) {
+	ckfree(pgvPtr->value);
+    } else {
+	Tcl_CreateExitHandler(FreeProcessGlobalValue, pgvPtr);
+    }
+    bytes = TclGetString(newValue);
+    pgvPtr->numBytes = newValue->length;
+    pgvPtr->value = ckalloc(pgvPtr->numBytes + 1);
+    memcpy(pgvPtr->value, bytes, pgvPtr->numBytes + 1);
+    if (pgvPtr->encoding) {
+	Tcl_FreeEncoding(pgvPtr->encoding);
+    }
+    pgvPtr->encoding = encoding;
+
+    /*
+     * Fill the local thread copy directly with the Tcl_Obj value to avoid
+     * loss of the intrep. Increment newValue refCount early to handle case
+     * where we set a PGV to itself.
+     */
+
+    Tcl_IncrRefCount(newValue);
+    cacheMap = GetThreadHash(&pgvPtr->key);
+    ClearHash(cacheMap);
+    hPtr = Tcl_CreateHashEntry(cacheMap, (void *)(pgvPtr->epoch), &dummy);
+    Tcl_SetHashValue(hPtr, newValue);
+    Tcl_MutexUnlock(&pgvPtr->mutex);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetProcessGlobalValue --
+ *
+ *	Retrieve a global value shared among all threads of the process,
+ *	preferring a thread-local copy as long as it remains valid.
+ *
+ * Results:
+ *	Returns a (Tcl_Obj *) that holds a copy of the global value.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetProcessGlobalValue(
+    ProcessGlobalValue *pgvPtr)
+{
+    Tcl_Obj *value = NULL;
+    Tcl_HashTable *cacheMap;
+    Tcl_HashEntry *hPtr;
+    size_t epoch = pgvPtr->epoch;
+
+    if (pgvPtr->encoding) {
+	Tcl_Encoding current = Tcl_GetEncoding(NULL, NULL);
+
+	if (pgvPtr->encoding != current) {
+	    /*
+	     * The system encoding has changed since the master string value
+	     * was saved. Convert the master value to be based on the new
+	     * system encoding.
+	     */
+
+	    Tcl_DString native, newValue;
+
+	    Tcl_MutexLock(&pgvPtr->mutex);
+	    epoch = ++pgvPtr->epoch;
+	    Tcl_UtfToExternalDString(pgvPtr->encoding, pgvPtr->value,
+		    pgvPtr->numBytes, &native);
+	    Tcl_ExternalToUtfDString(current, Tcl_DStringValue(&native),
+	    Tcl_DStringLength(&native), &newValue);
+	    Tcl_DStringFree(&native);
+	    ckfree(pgvPtr->value);
+	    pgvPtr->value = ckalloc(Tcl_DStringLength(&newValue) + 1);
+	    memcpy(pgvPtr->value, Tcl_DStringValue(&newValue),
+		    (size_t) Tcl_DStringLength(&newValue) + 1);
+	    Tcl_DStringFree(&newValue);
+	    Tcl_FreeEncoding(pgvPtr->encoding);
+	    pgvPtr->encoding = current;
+	    Tcl_MutexUnlock(&pgvPtr->mutex);
+	} else {
+	    Tcl_FreeEncoding(current);
+	}
+    }
+    cacheMap = GetThreadHash(&pgvPtr->key);
+    hPtr = Tcl_FindHashEntry(cacheMap, (void *) (epoch));
+    if (NULL == hPtr) {
+	int dummy;
+
+	/*
+	 * No cache for the current epoch - must be a new one.
+	 *
+	 * First, clear the cacheMap, as anything in it must refer to some
+	 * expired epoch.
+	 */
+
+	ClearHash(cacheMap);
+
+	/*
+	 * If no thread has set the shared value, call the initializer.
+	 */
+
+	Tcl_MutexLock(&pgvPtr->mutex);
+	if ((NULL == pgvPtr->value) && (pgvPtr->proc)) {
+	    pgvPtr->epoch++;
+	    pgvPtr->proc(&pgvPtr->value,&pgvPtr->numBytes,&pgvPtr->encoding);
+	    if (pgvPtr->value == NULL) {
+		Tcl_Panic("PGV Initializer did not initialize");
+	    }
+	    Tcl_CreateExitHandler(FreeProcessGlobalValue, pgvPtr);
+	}
+
+	/*
+	 * Store a copy of the shared value in our epoch-indexed cache.
+	 */
+
+	value = Tcl_NewStringObj(pgvPtr->value, pgvPtr->numBytes);
+	hPtr = Tcl_CreateHashEntry(cacheMap,
+		(void *)(pgvPtr->epoch), &dummy);
+	Tcl_MutexUnlock(&pgvPtr->mutex);
+	Tcl_SetHashValue(hPtr, value);
+	Tcl_IncrRefCount(value);
+    }
+    return Tcl_GetHashValue(hPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclSetObjNameOfExecutable --
+ *
+ *	This function stores the absolute pathname of the executable file
+ *	(normally as computed by TclpFindExecutable).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores the executable name.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclSetObjNameOfExecutable(
+    Tcl_Obj *name,
+    Tcl_Encoding encoding)
+{
+    TclSetProcessGlobalValue(&executableName, name, encoding);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetObjNameOfExecutable --
+ *
+ *	This function retrieves the absolute pathname of the application in
+ *	which the Tcl library is running, usually as previously stored by
+ *	TclpFindExecutable(). This function call is the C API equivalent to
+ *	the "info nameofexecutable" command.
+ *
+ * Results:
+ *	A pointer to an "fsPath" Tcl_Obj, or to an empty Tcl_Obj if the
+ *	pathname of the application is unknown.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclGetObjNameOfExecutable(void)
+{
+    return TclGetProcessGlobalValue(&executableName);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetNameOfExecutable --
+ *
+ *	This function retrieves the absolute pathname of the application in
+ *	which the Tcl library is running, and returns it in string form.
+ *
+ *	The returned string belongs to Tcl and should be copied if the caller
+ *	plans to keep it, to guard against it becoming invalid.
+ *
+ * Results:
+ *	A pointer to the internal string or NULL if the internal full path
+ *	name has not been computed or unknown.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetNameOfExecutable(void)
+{
+    Tcl_Obj *obj = TclGetObjNameOfExecutable();
+    const char *bytes = TclGetString(obj);
+
+    if (obj->length == 0) {
+	return NULL;
+    }
+    return bytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclpGetTime --
+ *
+ *	Deprecated synonym for Tcl_GetTime. This function is provided for the
+ *	benefit of extensions written before Tcl_GetTime was exported from the
+ *	library.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Stores current time in the buffer designated by "timePtr"
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclpGetTime(
+    Tcl_Time *timePtr)
+{
+    Tcl_GetTime(timePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclGetPlatform --
+ *
+ *	This is a kludge that allows the test library to get access the
+ *	internal tclPlatform variable.
+ *
+ * Results:
+ *	Returns a pointer to the tclPlatform variable.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+TclPlatformType *
+TclGetPlatform(void)
+{
+    return &tclPlatform;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclReToGlob --
+ *
+ *	Attempt to convert a regular expression to an equivalent glob pattern.
+ *
+ * Results:
+ *	Returns TCL_OK on success, TCL_ERROR on failure. If interp is not
+ *	NULL, an error message is placed in the result. On success, the
+ *	DString will contain an exact equivalent glob pattern. The caller is
+ *	responsible for calling Tcl_DStringFree on success. If exactPtr is not
+ *	NULL, it will be 1 if an exact match qualifies.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclReToGlob(
+    Tcl_Interp *interp,
+    const char *reStr,
+    int reStrLen,
+    Tcl_DString *dsPtr,
+    int *exactPtr,
+    int *quantifiersFoundPtr)
+{
+    int anchorLeft, anchorRight, lastIsStar, numStars;
+    char *dsStr, *dsStrStart;
+    const char *msg, *p, *strEnd, *code;
+
+    strEnd = reStr + reStrLen;
+    Tcl_DStringInit(dsPtr);
+    if (quantifiersFoundPtr != NULL) {
+	*quantifiersFoundPtr = 0;
+    }
+
+    /*
+     * "***=xxx" == "*xxx*", watch for glob-sensitive chars.
+     */
+
+    if ((reStrLen >= 4) && (memcmp("***=", reStr, 4) == 0)) {
+	/*
+	 * At most, the glob pattern has length 2*reStrLen + 2 to backslash
+	 * escape every character and have * at each end.
+	 */
+
+	Tcl_DStringSetLength(dsPtr, reStrLen + 2);
+	dsStr = dsStrStart = Tcl_DStringValue(dsPtr);
+	*dsStr++ = '*';
+	for (p = reStr + 4; p < strEnd; p++) {
+	    switch (*p) {
+	    case '\\': case '*': case '[': case ']': case '?':
+		/* Only add \ where necessary for glob */
+		*dsStr++ = '\\';
+		/* fall through */
+	    default:
+		*dsStr++ = *p;
+		break;
+	    }
+	}
+	*dsStr++ = '*';
+	Tcl_DStringSetLength(dsPtr, dsStr - dsStrStart);
+	if (exactPtr) {
+	    *exactPtr = 0;
+	}
+	return TCL_OK;
+    }
+
+    /*
+     * At most, the glob pattern has length reStrLen + 2 to account for
+     * possible * at each end.
+     */
+
+    Tcl_DStringSetLength(dsPtr, reStrLen + 2);
+    dsStr = dsStrStart = Tcl_DStringValue(dsPtr);
+
+    /*
+     * Check for anchored REs (ie ^foo$), so we can use string equal if
+     * possible. Do not alter the start of str so we can free it correctly.
+     *
+     * Keep track of the last char being an unescaped star to prevent multiple
+     * instances.  Simpler than checking that the last star may be escaped.
+     */
+
+    msg = NULL;
+    code = NULL;
+    p = reStr;
+    anchorRight = 0;
+    lastIsStar = 0;
+    numStars = 0;
+
+    if (*p == '^') {
+	anchorLeft = 1;
+	p++;
+    } else {
+	anchorLeft = 0;
+	*dsStr++ = '*';
+	lastIsStar = 1;
+    }
+
+    for ( ; p < strEnd; p++) {
+	switch (*p) {
+	case '\\':
+	    p++;
+	    switch (*p) {
+	    case 'a':
+		*dsStr++ = '\a';
+		break;
+	    case 'b':
+		*dsStr++ = '\b';
+		break;
+	    case 'f':
+		*dsStr++ = '\f';
+		break;
+	    case 'n':
+		*dsStr++ = '\n';
+		break;
+	    case 'r':
+		*dsStr++ = '\r';
+		break;
+	    case 't':
+		*dsStr++ = '\t';
+		break;
+	    case 'v':
+		*dsStr++ = '\v';
+		break;
+	    case 'B': case '\\':
+		*dsStr++ = '\\';
+		*dsStr++ = '\\';
+		anchorLeft = 0; /* prevent exact match */
+		break;
+	    case '*': case '[': case ']': case '?':
+		/* Only add \ where necessary for glob */
+		*dsStr++ = '\\';
+		anchorLeft = 0; /* prevent exact match */
+		/* fall through */
+	    case '{': case '}': case '(': case ')': case '+':
+	    case '.': case '|': case '^': case '$':
+		*dsStr++ = *p;
+		break;
+	    default:
+		msg = "invalid escape sequence";
+		code = "BADESCAPE";
+		goto invalidGlob;
+	    }
+	    break;
+	case '.':
+	    if (quantifiersFoundPtr != NULL) {
+		*quantifiersFoundPtr = 1;
+	    }
+	    anchorLeft = 0; /* prevent exact match */
+	    if (p+1 < strEnd) {
+		if (p[1] == '*') {
+		    p++;
+		    if (!lastIsStar) {
+			*dsStr++ = '*';
+			lastIsStar = 1;
+			numStars++;
+		    }
+		    continue;
+		} else if (p[1] == '+') {
+		    p++;
+		    *dsStr++ = '?';
+		    *dsStr++ = '*';
+		    lastIsStar = 1;
+		    numStars++;
+		    continue;
+		}
+	    }
+	    *dsStr++ = '?';
+	    break;
+	case '$':
+	    if (p+1 != strEnd) {
+		msg = "$ not anchor";
+		code = "NONANCHOR";
+		goto invalidGlob;
+	    }
+	    anchorRight = 1;
+	    break;
+	case '*': case '+': case '?': case '|': case '^':
+	case '{': case '}': case '(': case ')': case '[': case ']':
+	    msg = "unhandled RE special char";
+	    code = "UNHANDLED";
+	    goto invalidGlob;
+	default:
+	    *dsStr++ = *p;
+	    break;
+	}
+	lastIsStar = 0;
+    }
+    if (numStars > 1) {
+	/*
+	 * Heuristic: if >1 non-anchoring *, the risk is large that glob
+	 * matching is slower than the RE engine, so report invalid.
+	 */
+
+	msg = "excessive recursive glob backtrack potential";
+	code = "OVERCOMPLEX";
+	goto invalidGlob;
+    }
+
+    if (!anchorRight && !lastIsStar) {
+	*dsStr++ = '*';
+    }
+    Tcl_DStringSetLength(dsPtr, dsStr - dsStrStart);
+
+    if (exactPtr) {
+	*exactPtr = (anchorLeft && anchorRight);
+    }
+
+    return TCL_OK;
+
+  invalidGlob:
+    if (interp != NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(msg, -1));
+	Tcl_SetErrorCode(interp, "TCL", "RE2GLOB", code, NULL);
+    }
+    Tcl_DStringFree(dsPtr);
+    return TCL_ERROR;
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclVar.c b/generic/tclVar.c
new file mode 100644
index 0000000..7c8bb73
--- /dev/null
+++ b/generic/tclVar.c
@@ -0,0 +1,6319 @@
+/*
+ * tclVar.c --
+ *
+ *	This file contains routines that implement Tcl variables (both scalars
+ *	and arrays).
+ *
+ *	The implementation of arrays is modelled after an initial
+ *	implementation by Mark Diekhans and Karl Lehenbauer.
+ *
+ * Copyright (c) 1987-1994 The Regents of the University of California.
+ * Copyright (c) 1994-1997 Sun Microsystems, Inc.
+ * Copyright (c) 1998-1999 by Scriptics Corporation.
+ * Copyright (c) 2001 by Kevin B. Kenny. All rights reserved.
+ * Copyright (c) 2007 Miguel Sofer
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#include "tclOOInt.h"
+
+/*
+ * Prototypes for the variable hash key methods.
+ */
+
+static Tcl_HashEntry *	AllocVarEntry(Tcl_HashTable *tablePtr, void *keyPtr);
+static void		FreeVarEntry(Tcl_HashEntry *hPtr);
+static int		CompareVarKeys(void *keyPtr, Tcl_HashEntry *hPtr);
+
+static const Tcl_HashKeyType tclVarHashKeyType = {
+    TCL_HASH_KEY_TYPE_VERSION,	/* version */
+    0,				/* flags */
+    TclHashObjKey,		/* hashKeyProc */
+    CompareVarKeys,		/* compareKeysProc */
+    AllocVarEntry,		/* allocEntryProc */
+    FreeVarEntry		/* freeEntryProc */
+};
+
+static inline Var *	VarHashCreateVar(TclVarHashTable *tablePtr,
+			    Tcl_Obj *key, int *newPtr);
+static inline Var *	VarHashFirstVar(TclVarHashTable *tablePtr,
+			    Tcl_HashSearch *searchPtr);
+static inline Var *	VarHashNextVar(Tcl_HashSearch *searchPtr);
+static inline void	CleanupVar(Var *varPtr, Var *arrayPtr);
+
+#define VarHashGetValue(hPtr) \
+    ((Var *) ((char *)hPtr - TclOffset(VarInHash, entry)))
+
+/*
+ * NOTE: VarHashCreateVar increments the recount of its key argument.
+ * All callers that will call Tcl_DecrRefCount on that argument must
+ * call Tcl_IncrRefCount on it before passing it in.  This requirement
+ * can bubble up to callers of callers .... etc.
+ */
+
+static inline Var *
+VarHashCreateVar(
+    TclVarHashTable *tablePtr,
+    Tcl_Obj *key,
+    int *newPtr)
+{
+    Tcl_HashEntry *hPtr = Tcl_CreateHashEntry(&tablePtr->table,
+	    key, newPtr);
+
+    if (hPtr) {
+	return VarHashGetValue(hPtr);
+    } else {
+	return NULL;
+    }
+}
+
+#define VarHashFindVar(tablePtr, key) \
+    VarHashCreateVar((tablePtr), (key), NULL)
+
+#define VarHashInvalidateEntry(varPtr) \
+    ((varPtr)->flags |= VAR_DEAD_HASH)
+
+#define VarHashDeleteEntry(varPtr) \
+    Tcl_DeleteHashEntry(&(((VarInHash *) varPtr)->entry))
+
+#define VarHashFirstEntry(tablePtr, searchPtr) \
+    Tcl_FirstHashEntry(&(tablePtr)->table, (searchPtr))
+
+#define VarHashNextEntry(searchPtr) \
+    Tcl_NextHashEntry((searchPtr))
+
+static inline Var *
+VarHashFirstVar(
+    TclVarHashTable *tablePtr,
+    Tcl_HashSearch *searchPtr)
+{
+    Tcl_HashEntry *hPtr = VarHashFirstEntry(tablePtr, searchPtr);
+
+    if (hPtr) {
+	return VarHashGetValue(hPtr);
+    } else {
+	return NULL;
+    }
+}
+
+static inline Var *
+VarHashNextVar(
+    Tcl_HashSearch *searchPtr)
+{
+    Tcl_HashEntry *hPtr = VarHashNextEntry(searchPtr);
+
+    if (hPtr) {
+	return VarHashGetValue(hPtr);
+    } else {
+	return NULL;
+    }
+}
+
+#define VarHashGetKey(varPtr) \
+    (((VarInHash *)(varPtr))->entry.key.objPtr)
+
+#define VarHashDeleteTable(tablePtr) \
+    Tcl_DeleteHashTable(&(tablePtr)->table)
+
+/*
+ * The strings below are used to indicate what went wrong when a variable
+ * access is denied.
+ */
+
+static const char *noSuchVar =		"no such variable";
+static const char *isArray =		"variable is array";
+static const char *needArray =		"variable isn't array";
+static const char *noSuchElement =	"no such element in array";
+static const char *danglingElement =
+	"upvar refers to element in deleted array";
+static const char *danglingVar =
+	"upvar refers to variable in deleted namespace";
+static const char *badNamespace =	"parent namespace doesn't exist";
+static const char *missingName =	"missing variable name";
+static const char *isArrayElement =
+	"name refers to an element in an array";
+
+/*
+ * A test to see if we are in a call frame that has local variables. This is
+ * true if we are inside a procedure body.
+ */
+
+#define HasLocalVars(framePtr) ((framePtr)->isProcCallFrame & FRAME_IS_PROC)
+
+/*
+ * The following structure describes an enumerative search in progress on an
+ * array variable; this are invoked with options to the "array" command.
+ */
+
+typedef struct ArraySearch {
+    Tcl_Obj *name;		/* Name of this search */
+    int id;			/* Integer id used to distinguish among
+				 * multiple concurrent searches for the same
+				 * array. */
+    struct Var *varPtr;		/* Pointer to array variable that's being
+				 * searched. */
+    Tcl_HashSearch search;	/* Info kept by the hash module about progress
+				 * through the array. */
+    Tcl_HashEntry *nextEntry;	/* Non-null means this is the next element to
+				 * be enumerated (it's leftover from the
+				 * Tcl_FirstHashEntry call or from an "array
+				 * anymore" command). NULL means must call
+				 * Tcl_NextHashEntry to get value to
+				 * return. */
+    struct ArraySearch *nextPtr;/* Next in list of all active searches for
+				 * this variable, or NULL if this is the last
+				 * one. */
+} ArraySearch;
+
+/*
+ * Forward references to functions defined later in this file:
+ */
+
+static void		AppendLocals(Tcl_Interp *interp, Tcl_Obj *listPtr,
+			    Tcl_Obj *patternPtr, int includeLinks);
+static void		DeleteSearches(Interp *iPtr, Var *arrayVarPtr);
+static void		DeleteArray(Interp *iPtr, Tcl_Obj *arrayNamePtr,
+			    Var *varPtr, int flags, int index);
+static Tcl_Var		ObjFindNamespaceVar(Tcl_Interp *interp,
+			    Tcl_Obj *namePtr, Tcl_Namespace *contextNsPtr,
+			    int flags);
+static int		ObjMakeUpvar(Tcl_Interp *interp,
+			    CallFrame *framePtr, Tcl_Obj *otherP1Ptr,
+			    const char *otherP2, const int otherFlags,
+			    Tcl_Obj *myNamePtr, int myFlags, int index);
+static ArraySearch *	ParseSearchId(Tcl_Interp *interp, const Var *varPtr,
+			    Tcl_Obj *varNamePtr, Tcl_Obj *handleObj);
+static void		UnsetVarStruct(Var *varPtr, Var *arrayPtr,
+			    Interp *iPtr, Tcl_Obj *part1Ptr,
+			    Tcl_Obj *part2Ptr, int flags, int index);
+static Var *		VerifyArray(Tcl_Interp *interp, Tcl_Obj *varNameObj);
+
+/*
+ * Functions defined in this file that may be exported in the future for use
+ * by the bytecode compiler and engine or to the public interface.
+ */
+
+MODULE_SCOPE Var *	TclLookupSimpleVar(Tcl_Interp *interp,
+			    Tcl_Obj *varNamePtr, int flags, const int create,
+			    const char **errMsgPtr, int *indexPtr);
+
+static Tcl_DupInternalRepProc	DupLocalVarName;
+static Tcl_FreeInternalRepProc	FreeLocalVarName;
+
+static Tcl_FreeInternalRepProc	FreeParsedVarName;
+static Tcl_DupInternalRepProc	DupParsedVarName;
+
+/*
+ * Types of Tcl_Objs used to cache variable lookups.
+ *
+ * localVarName - INTERNALREP DEFINITION:
+ *   twoPtrValue.ptr1:   pointer to name obj in varFramePtr->localCache
+ *			  or NULL if it is this same obj
+ *   twoPtrValue.ptr2: index into locals table
+ *
+ * parsedVarName - INTERNALREP DEFINITION:
+ *   twoPtrValue.ptr1:	pointer to the array name Tcl_Obj, or NULL if it is a
+ *			scalar variable
+ *   twoPtrValue.ptr2:	pointer to the element name string (owned by this
+ *			Tcl_Obj), or NULL if it is a scalar variable
+ */
+
+static const Tcl_ObjType localVarNameType = {
+    "localVarName",
+    FreeLocalVarName, DupLocalVarName, NULL, NULL
+};
+
+static const Tcl_ObjType tclParsedVarNameType = {
+    "parsedVarName",
+    FreeParsedVarName, DupParsedVarName, NULL, NULL
+};
+
+
+Var *
+TclVarHashCreateVar(
+    TclVarHashTable *tablePtr,
+    const char *key,
+    int *newPtr)
+{
+    Tcl_Obj *keyPtr;
+    Var *varPtr;
+
+    keyPtr = Tcl_NewStringObj(key, -1);
+    Tcl_IncrRefCount(keyPtr);
+    varPtr = VarHashCreateVar(tablePtr, keyPtr, newPtr);
+    Tcl_DecrRefCount(keyPtr);
+
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclCleanupVar --
+ *
+ *	This function is called when it looks like it may be OK to free up a
+ *	variable's storage. If the variable is in a hashtable, its Var
+ *	structure and hash table entry will be freed along with those of its
+ *	containing array, if any. This function is called, for example, when
+ *	a trace on a variable deletes a variable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the variable (or its containing array) really is dead and in a
+ *	hashtable, then its Var structure, and possibly its hash table entry,
+ *	is freed up.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline void
+CleanupVar(
+    Var *varPtr,		/* Pointer to variable that may be a candidate
+				 * for being expunged. */
+    Var *arrayPtr)		/* Array that contains the variable, or NULL
+				 * if this variable isn't an array element. */
+{
+    if (TclIsVarUndefined(varPtr) && TclIsVarInHash(varPtr)
+	    && !TclIsVarTraced(varPtr)
+	    && (VarHashRefCount(varPtr) == !TclIsVarDeadHash(varPtr))) {
+	if (VarHashRefCount(varPtr) == 0) {
+	    ckfree(varPtr);
+	} else {
+	    VarHashDeleteEntry(varPtr);
+	}
+    }
+    if (arrayPtr != NULL && TclIsVarUndefined(arrayPtr) &&
+	    TclIsVarInHash(arrayPtr) && !TclIsVarTraced(arrayPtr) &&
+	    (VarHashRefCount(arrayPtr) == !TclIsVarDeadHash(arrayPtr))) {
+	if (VarHashRefCount(arrayPtr) == 0) {
+	    ckfree(arrayPtr);
+	} else {
+	    VarHashDeleteEntry(arrayPtr);
+	}
+    }
+}
+
+void
+TclCleanupVar(
+    Var *varPtr,		/* Pointer to variable that may be a candidate
+				 * for being expunged. */
+    Var *arrayPtr)		/* Array that contains the variable, or NULL
+				 * if this variable isn't an array element. */
+{
+    CleanupVar(varPtr, arrayPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLookupVar --
+ *
+ *	This function is used to locate a variable given its name(s). It has
+ *	been mostly superseded by TclObjLookupVar, it is now only used by the
+ *	trace code. It is kept in tcl8.5 mainly because it is in the internal
+ *	stubs table, so that some extension may be calling it.
+ *
+ * Results:
+ *	The return value is a pointer to the variable structure indicated by
+ *	part1 and part2, or NULL if the variable couldn't be found. If the
+ *	variable is found, *arrayPtrPtr is filled in with the address of the
+ *	variable structure for the array that contains the variable (or NULL
+ *	if the variable is a scalar). If the variable can't be found and
+ *	either createPart1 or createPart2 are 1, a new as-yet-undefined
+ *	(VAR_UNDEFINED) variable structure is created, entered into a hash
+ *	table, and returned.
+ *
+ *	If the variable isn't found and creation wasn't specified, or some
+ *	other error occurs, NULL is returned and an error message is left in
+ *	the interp's result if TCL_LEAVE_ERR_MSG is set in flags.
+ *
+ *	Note: it's possible for the variable returned to be VAR_UNDEFINED even
+ *	if createPart1 or createPart2 are 1 (these only cause the hash table
+ *	entry or array to be created). For example, the variable might be a
+ *	global that has been unset but is still referenced by a procedure, or
+ *	a variable that has been unset but it only being kept in existence (if
+ *	VAR_UNDEFINED) by a trace.
+ *
+ * Side effects:
+ *	New hashtable entries may be created if createPart1 or createPart2
+ *	are 1.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Var *
+TclLookupVar(
+    Tcl_Interp *interp,		/* Interpreter to use for lookup. */
+    const char *part1,		/* If part2 isn't NULL, this is the name of an
+				 * array. Otherwise, this is a full variable
+				 * name that could include a parenthesized
+				 * array element. */
+    const char *part2,		/* Name of element within array, or NULL. */
+    int flags,			/* Only TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * and TCL_LEAVE_ERR_MSG bits matter. */
+    const char *msg,		/* Verb to use in error messages, e.g. "read"
+				 * or "set". Only needed if TCL_LEAVE_ERR_MSG
+				 * is set in flags. */
+    int createPart1,		/* If 1, create hash table entry for part 1 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    int createPart2,		/* If 1, create hash table entry for part 2 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    Var **arrayPtrPtr)		/* If the name refers to an element of an
+				 * array, *arrayPtrPtr gets filled in with
+				 * address of array variable. Otherwise this
+				 * is set to NULL. */
+{
+    Var *varPtr;
+    Tcl_Obj *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    if (createPart1) {
+	Tcl_IncrRefCount(part1Ptr);
+    }
+
+    varPtr = TclObjLookupVar(interp, part1Ptr, part2, flags, msg,
+	    createPart1, createPart2, arrayPtrPtr);
+
+    TclDecrRefCount(part1Ptr);
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjLookupVar, TclObjLookupVarEx --
+ *
+ *	This function is used by virtually all of the variable code to locate
+ *	a variable given its name(s). The parsing into array/element
+ *	components and (if possible) the lookup results are cached in
+ *	part1Ptr, which is converted to one of the varNameTypes.
+ *
+ * Results:
+ *	The return value is a pointer to the variable structure indicated by
+ *	part1Ptr and part2, or NULL if the variable couldn't be found. If *
+ *	the variable is found, *arrayPtrPtr is filled with the address of the
+ *	variable structure for the array that contains the variable (or NULL
+ *	if the variable is a scalar). If the variable can't be found and
+ *	either createPart1 or createPart2 are 1, a new as-yet-undefined
+ *	(VAR_UNDEFINED) variable structure is created, entered into a hash
+ *	table, and returned.
+ *
+ *	If the variable isn't found and creation wasn't specified, or some
+ *	other error occurs, NULL is returned and an error message is left in
+ *	the interp's result if TCL_LEAVE_ERR_MSG is set in flags.
+ *
+ *	Note: it's possible for the variable returned to be VAR_UNDEFINED even
+ *	if createPart1 or createPart2 are 1 (these only cause the hash table
+ *	entry or array to be created). For example, the variable might be a
+ *	global that has been unset but is still referenced by a procedure, or
+ *	a variable that has been unset but it only being kept in existence (if
+ *	VAR_UNDEFINED) by a trace.
+ *
+ * Side effects:
+ *	New hashtable entries may be created if createPart1 or createPart2
+ *	are 1. The object part1Ptr is converted to one of localVarNameType,
+ *	tclNsVarNameType or tclParsedVarNameType and caches as much of the
+ *	lookup as it can.
+ *	When createPart1 is 1, callers must IncrRefCount part1Ptr if they
+ *	plan to DecrRefCount it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Var *
+TclObjLookupVar(
+    Tcl_Interp *interp,		/* Interpreter to use for lookup. */
+    register Tcl_Obj *part1Ptr,	/* If part2 isn't NULL, this is the name of an
+				 * array. Otherwise, this is a full variable
+				 * name that could include a parenthesized
+				 * array element. */
+    const char *part2,		/* Name of element within array, or NULL. */
+    int flags,			/* Only TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * and TCL_LEAVE_ERR_MSG bits matter. */
+    const char *msg,		/* Verb to use in error messages, e.g. "read"
+				 * or "set". Only needed if TCL_LEAVE_ERR_MSG
+				 * is set in flags. */
+    const int createPart1,	/* If 1, create hash table entry for part 1 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    const int createPart2,	/* If 1, create hash table entry for part 2 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    Var **arrayPtrPtr)		/* If the name refers to an element of an
+				 * array, *arrayPtrPtr gets filled in with
+				 * address of array variable. Otherwise this
+				 * is set to NULL. */
+{
+    Tcl_Obj *part2Ptr = NULL;
+    Var *resPtr;
+
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+	if (createPart2) {
+	    Tcl_IncrRefCount(part2Ptr);
+	}
+    }
+
+    resPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr,
+	    flags, msg, createPart1, createPart2, arrayPtrPtr);
+
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+
+    return resPtr;
+}
+
+/*
+ *	When createPart1 is 1, callers must IncrRefCount part1Ptr if they
+ *	plan to DecrRefCount it.
+ *	When createPart2 is 1, callers must IncrRefCount part2Ptr if they
+ *	plan to DecrRefCount it.
+ */
+Var *
+TclObjLookupVarEx(
+    Tcl_Interp *interp,		/* Interpreter to use for lookup. */
+    Tcl_Obj *part1Ptr,		/* If part2Ptr isn't NULL, this is the name of
+				 * an array. Otherwise, this is a full
+				 * variable name that could include a
+				 * parenthesized array element. */
+    Tcl_Obj *part2Ptr,		/* Name of element within array, or NULL. */
+    int flags,			/* Only TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * and TCL_LEAVE_ERR_MSG bits matter. */
+    const char *msg,		/* Verb to use in error messages, e.g. "read"
+				 * or "set". Only needed if TCL_LEAVE_ERR_MSG
+				 * is set in flags. */
+    const int createPart1,	/* If 1, create hash table entry for part 1 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    const int createPart2,	/* If 1, create hash table entry for part 2 of
+				 * name, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    Var **arrayPtrPtr)		/* If the name refers to an element of an
+				 * array, *arrayPtrPtr gets filled in with
+				 * address of array variable. Otherwise this
+				 * is set to NULL. */
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *varFramePtr = iPtr->varFramePtr;
+    register Var *varPtr;	/* Points to the variable's in-frame Var
+				 * structure. */
+    const char *errMsg = NULL;
+    int index, parsed = 0;
+    const Tcl_ObjType *typePtr = part1Ptr->typePtr;
+
+    *arrayPtrPtr = NULL;
+
+    if (typePtr == &localVarNameType) {
+	int localIndex;
+
+    localVarNameTypeHandling:
+	localIndex = PTR2INT(part1Ptr->internalRep.twoPtrValue.ptr2);
+	if (HasLocalVars(varFramePtr)
+		&& !(flags & (TCL_GLOBAL_ONLY | TCL_NAMESPACE_ONLY))
+		&& (localIndex < varFramePtr->numCompiledLocals)) {
+	    /*
+	     * Use the cached index if the names coincide.
+	     */
+
+	    Tcl_Obj *namePtr = part1Ptr->internalRep.twoPtrValue.ptr1;
+	    Tcl_Obj *checkNamePtr = localName(varFramePtr, localIndex);
+
+	    if ((!namePtr && (checkNamePtr == part1Ptr)) ||
+		    (namePtr && (checkNamePtr == namePtr))) {
+		varPtr = (Var *) &(varFramePtr->compiledLocals[localIndex]);
+		goto donePart1;
+	    }
+	}
+	goto doneParsing;
+    }
+
+    /*
+     * If part1Ptr is a tclParsedVarNameType, separate it into the pre-parsed
+     * parts.
+     */
+
+    if (typePtr == &tclParsedVarNameType) {
+	if (part1Ptr->internalRep.twoPtrValue.ptr1 != NULL) {
+	    if (part2Ptr != NULL) {
+		/*
+		 * ERROR: part1Ptr is already an array element, cannot specify
+		 * a part2.
+		 */
+
+		if (flags & TCL_LEAVE_ERR_MSG) {
+		    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, msg,
+			    noSuchVar, -1);
+		    Tcl_SetErrorCode(interp, "TCL", "VALUE", "VARNAME", NULL);
+		}
+		return NULL;
+	    }
+	    part2Ptr = part1Ptr->internalRep.twoPtrValue.ptr2;
+	    part1Ptr = part1Ptr->internalRep.twoPtrValue.ptr1;
+	    typePtr = part1Ptr->typePtr;
+	    if (typePtr == &localVarNameType) {
+		goto localVarNameTypeHandling;
+	    }
+	}
+	parsed = 1;
+    }
+
+    if (!parsed) {
+
+	/*
+	 * part1Ptr is possibly an unparsed array element.
+	 */
+
+	int len;
+	const char *part1 = TclGetStringFromObj(part1Ptr, &len);
+
+	if (len > 1 && (part1[len - 1] == ')')) {
+
+	  const char *part2 = strchr(part1, '(');
+
+	  if (part2) {
+	    Tcl_Obj *arrayPtr;
+
+		if (part2Ptr != NULL) {
+		    if (flags & TCL_LEAVE_ERR_MSG) {
+			TclObjVarErrMsg(interp, part1Ptr, part2Ptr, msg,
+				needArray, -1);
+			Tcl_SetErrorCode(interp, "TCL", "VALUE", "VARNAME",
+				NULL);
+		    }
+		    return NULL;
+		}
+
+	    arrayPtr = Tcl_NewStringObj(part1, (part2 - part1));
+	    part2Ptr = Tcl_NewStringObj(part2 + 1, len - (part2 - part1) - 2);
+
+	    TclFreeIntRep(part1Ptr);
+
+	    Tcl_IncrRefCount(arrayPtr);
+	    part1Ptr->internalRep.twoPtrValue.ptr1 = arrayPtr;
+	    Tcl_IncrRefCount(part2Ptr);
+	    part1Ptr->internalRep.twoPtrValue.ptr2 = part2Ptr;
+	    part1Ptr->typePtr = &tclParsedVarNameType;
+
+	    part1Ptr = arrayPtr;
+	  }
+	}
+    }
+
+  doneParsing:
+    /*
+     * part1Ptr is not an array element; look it up, and convert it to one of
+     * the cached types if possible.
+     */
+
+    varPtr = TclLookupSimpleVar(interp, part1Ptr, flags, createPart1,
+	    &errMsg, &index);
+    if (varPtr == NULL) {
+	if ((errMsg != NULL) && (flags & TCL_LEAVE_ERR_MSG)) {
+	    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, msg, errMsg, -1);
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+		    TclGetString(part1Ptr), NULL);
+	}
+	return NULL;
+    }
+
+    /*
+     * Cache the newly found variable if possible.
+     */
+
+    TclFreeIntRep(part1Ptr);
+    if (index >= 0) {
+	/*
+	 * An indexed local variable.
+	 */
+	Tcl_Obj *cachedNamePtr = localName(varFramePtr, index);
+
+	part1Ptr->typePtr = &localVarNameType;
+	if (part1Ptr != cachedNamePtr) {
+	    part1Ptr->internalRep.twoPtrValue.ptr1 = cachedNamePtr;
+	    Tcl_IncrRefCount(cachedNamePtr);
+	    if (cachedNamePtr->typePtr != &localVarNameType
+		    || cachedNamePtr->internalRep.twoPtrValue.ptr1 != NULL) {
+	        TclFreeIntRep(cachedNamePtr);
+	    }
+	} else {
+	    part1Ptr->internalRep.twoPtrValue.ptr1 = NULL;
+	}
+	part1Ptr->internalRep.twoPtrValue.ptr2 = INT2PTR(index);
+    } else {
+	/*
+	 * At least mark part1Ptr as already parsed.
+	 */
+
+	part1Ptr->typePtr = &tclParsedVarNameType;
+	part1Ptr->internalRep.twoPtrValue.ptr1 = NULL;
+	part1Ptr->internalRep.twoPtrValue.ptr2 = NULL;
+    }
+
+  donePart1:
+    while (TclIsVarLink(varPtr)) {
+	varPtr = varPtr->value.linkPtr;
+    }
+
+    if (part2Ptr != NULL) {
+	/*
+	 * Array element sought: look it up.
+	 */
+
+	*arrayPtrPtr = varPtr;
+	varPtr = TclLookupArrayElement(interp, part1Ptr, part2Ptr, flags, msg,
+		createPart1, createPart2, varPtr, -1);
+    }
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLookupSimpleVar --
+ *
+ *	This function is used by to locate a simple variable (i.e., not an
+ *	array element) given its name.
+ *
+ * Results:
+ *	The return value is a pointer to the variable structure indicated by
+ *	varName, or NULL if the variable couldn't be found. If the variable
+ *	can't be found and create is 1, a new as-yet-undefined (VAR_UNDEFINED)
+ *	variable structure is created, entered into a hash table, and
+ *	returned.
+ *
+ *	If the current CallFrame corresponds to a proc and the variable found
+ *	is one of the compiledLocals, its index is placed in *indexPtr.
+ *	Otherwise, *indexPtr will be set to (according to the needs of
+ *	TclObjLookupVar):
+ *	    -1 a global reference
+ *	    -2 a reference to a namespace variable
+ *	    -3 a non-cachable reference, i.e., one of:
+ *		. non-indexed local var
+ *		. a reference of unknown origin;
+ *		. resolution by a namespace or interp resolver
+ *
+ *	If the variable isn't found and creation wasn't specified, or some
+ *	other error occurs, NULL is returned and the corresponding error
+ *	message is left in *errMsgPtr.
+ *
+ *	Note: it's possible for the variable returned to be VAR_UNDEFINED even
+ *	if create is 1 (this only causes the hash table entry to be created).
+ *	For example, the variable might be a global that has been unset but is
+ *	still referenced by a procedure, or a variable that has been unset but
+ *	it only being kept in existence (if VAR_UNDEFINED) by a trace.
+ *
+ * Side effects:
+ *	A new hashtable entry may be created if create is 1.
+ *	Callers must Incr varNamePtr if they plan to Decr it if create is 1.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Var *
+TclLookupSimpleVar(
+    Tcl_Interp *interp,		/* Interpreter to use for lookup. */
+    Tcl_Obj *varNamePtr,	/* This is a simple variable name that could
+				 * represent a scalar or an array. */
+    int flags,			/* Only TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_AVOID_RESOLVERS and TCL_LEAVE_ERR_MSG
+				 * bits matter. */
+    const int create,		/* If 1, create hash table entry for varname,
+				 * if it doesn't already exist. If 0, return
+				 * error if it doesn't exist. */
+    const char **errMsgPtr,
+    int *indexPtr)
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *varFramePtr = iPtr->varFramePtr;
+				/* Points to the procedure call frame whose
+				 * variables are currently in use. Same as the
+				 * current procedure's frame, if any, unless
+				 * an "uplevel" is executing. */
+    TclVarHashTable *tablePtr;	/* Points to the hashtable, if any, in which
+				 * to look up the variable. */
+    Tcl_Var var;		/* Used to search for global names. */
+    Var *varPtr;		/* Points to the Var structure returned for
+				 * the variable. */
+    Namespace *varNsPtr, *cxtNsPtr, *dummy1Ptr, *dummy2Ptr;
+    ResolverScheme *resPtr;
+    int isNew, i, result, varLen;
+    const char *varName = TclGetStringFromObj(varNamePtr, &varLen);
+
+    varPtr = NULL;
+    varNsPtr = NULL;		/* Set non-NULL if a nonlocal variable. */
+    *indexPtr = -3;
+
+    if (flags & TCL_GLOBAL_ONLY) {
+	cxtNsPtr = iPtr->globalNsPtr;
+    } else {
+	cxtNsPtr = iPtr->varFramePtr->nsPtr;
+    }
+
+    /*
+     * If this namespace has a variable resolver, then give it first crack at
+     * the variable resolution. It may return a Tcl_Var value, it may signal
+     * to continue onward, or it may signal an error.
+     */
+
+    if ((cxtNsPtr->varResProc != NULL || iPtr->resolverPtr != NULL)
+	    && !(flags & TCL_AVOID_RESOLVERS)) {
+	resPtr = iPtr->resolverPtr;
+	if (cxtNsPtr->varResProc) {
+	    result = cxtNsPtr->varResProc(interp, varName,
+		    (Tcl_Namespace *) cxtNsPtr, flags, &var);
+	} else {
+	    result = TCL_CONTINUE;
+	}
+
+	while (result == TCL_CONTINUE && resPtr) {
+	    if (resPtr->varResProc) {
+		result = resPtr->varResProc(interp, varName,
+			(Tcl_Namespace *) cxtNsPtr, flags, &var);
+	    }
+	    resPtr = resPtr->nextPtr;
+	}
+
+	if (result == TCL_OK) {
+	    return (Var *) var;
+	} else if (result != TCL_CONTINUE) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * Look up varName. Look it up as either a namespace variable or as a
+     * local variable in a procedure call frame (varFramePtr). Interpret
+     * varName as a namespace variable if:
+     *    1) so requested by a TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY flag,
+     *    2) there is no active frame (we're at the global :: scope),
+     *    3) the active frame was pushed to define the namespace context for a
+     *	     "namespace eval" or "namespace inscope" command,
+     *    4) the name has namespace qualifiers ("::"s).
+     * Otherwise, if varName is a local variable, search first in the frame's
+     * array of compiler-allocated local variables, then in its hashtable for
+     * runtime-created local variables.
+     *
+     * If create and the variable isn't found, create the variable and, if
+     * necessary, create varFramePtr's local var hashtable.
+     */
+
+    if (((flags & (TCL_GLOBAL_ONLY | TCL_NAMESPACE_ONLY)) != 0)
+	    || !HasLocalVars(varFramePtr)
+	    || (strstr(varName, "::") != NULL)) {
+	const char *tail;
+	int lookGlobal = (flags & TCL_GLOBAL_ONLY)
+		|| (cxtNsPtr == iPtr->globalNsPtr)
+		|| ((*varName == ':') && (*(varName+1) == ':'));
+
+	if (lookGlobal) {
+	    *indexPtr = -1;
+	    flags = (flags | TCL_GLOBAL_ONLY) & ~TCL_NAMESPACE_ONLY;
+	} else {
+	    if (flags & TCL_AVOID_RESOLVERS) {
+		flags = (flags | TCL_NAMESPACE_ONLY);
+	    }
+	    if (flags & TCL_NAMESPACE_ONLY) {
+		*indexPtr = -2;
+	    }
+	}
+
+	/*
+	 * Don't pass TCL_LEAVE_ERR_MSG, we may yet create the variable, or
+	 * otherwise generate our own error!
+	 */
+
+	varPtr = (Var *) ObjFindNamespaceVar(interp, varNamePtr,
+		(Tcl_Namespace *) cxtNsPtr,
+		(flags | TCL_AVOID_RESOLVERS) & ~TCL_LEAVE_ERR_MSG);
+	if (varPtr == NULL) {
+	    Tcl_Obj *tailPtr;
+
+	    if (create) {	/* Var wasn't found so create it. */
+		TclGetNamespaceForQualName(interp, varName, cxtNsPtr,
+			flags, &varNsPtr, &dummy1Ptr, &dummy2Ptr, &tail);
+		if (varNsPtr == NULL) {
+		    *errMsgPtr = badNamespace;
+		    return NULL;
+		} else if (tail == NULL) {
+		    *errMsgPtr = missingName;
+		    return NULL;
+		}
+		if (tail != varName) {
+		    tailPtr = Tcl_NewStringObj(tail, -1);
+		} else {
+		    tailPtr = varNamePtr;
+		}
+		varPtr = VarHashCreateVar(&varNsPtr->varTable, tailPtr,
+			&isNew);
+		if (lookGlobal) {
+		    /*
+		     * The variable was created starting from the global
+		     * namespace: a global reference is returned even if it
+		     * wasn't explicitly requested.
+		     */
+
+		    *indexPtr = -1;
+		} else {
+		    *indexPtr = -2;
+		}
+	    } else {		/* Var wasn't found and not to create it. */
+		*errMsgPtr = noSuchVar;
+		return NULL;
+	    }
+	}
+    } else {			/* Local var: look in frame varFramePtr. */
+	int localLen, localCt = varFramePtr->numCompiledLocals;
+	Tcl_Obj **objPtrPtr = &varFramePtr->localCachePtr->varName0;
+	const char *localNameStr;
+
+	for (i=0 ; i<localCt ; i++, objPtrPtr++) {
+	    register Tcl_Obj *objPtr = *objPtrPtr;
+
+	    if (objPtr) {
+		localNameStr = TclGetStringFromObj(objPtr, &localLen);
+
+		if ((varLen == localLen) && (varName[0] == localNameStr[0])
+			&& !memcmp(varName, localNameStr, varLen)) {
+		    *indexPtr = i;
+		    return (Var *) &varFramePtr->compiledLocals[i];
+		}
+	    }
+	}
+	tablePtr = varFramePtr->varTablePtr;
+	if (create) {
+	    if (tablePtr == NULL) {
+		tablePtr = ckalloc(sizeof(TclVarHashTable));
+		TclInitVarHashTable(tablePtr, NULL);
+		varFramePtr->varTablePtr = tablePtr;
+	    }
+	    varPtr = VarHashCreateVar(tablePtr, varNamePtr, &isNew);
+	} else {
+	    varPtr = NULL;
+	    if (tablePtr != NULL) {
+		varPtr = VarHashFindVar(tablePtr, varNamePtr);
+	    }
+	    if (varPtr == NULL) {
+		*errMsgPtr = noSuchVar;
+	    }
+	}
+    }
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclLookupArrayElement --
+ *
+ *	This function is used to locate a variable which is in an array's
+ *	hashtable given a pointer to the array's Var structure and the
+ *	element's name.
+ *
+ * Results:
+ *	The return value is a pointer to the variable structure , or NULL if
+ *	the variable couldn't be found.
+ *
+ *	If arrayPtr points to a variable that isn't an array and createPart1
+ *	is 1, the corresponding variable will be converted to an array.
+ *	Otherwise, NULL is returned and an error message is left in the
+ *	interp's result if TCL_LEAVE_ERR_MSG is set in flags.
+ *
+ *	If the variable is not found and createPart2 is 1, the variable is
+ *	created. Otherwise, NULL is returned and an error message is left in
+ *	the interp's result if TCL_LEAVE_ERR_MSG is set in flags.
+ *
+ *	Note: it's possible for the variable returned to be VAR_UNDEFINED even
+ *	if createPart1 or createPart2 are 1 (these only cause the hash table
+ *	entry or array to be created). For example, the variable might be a
+ *	global that has been unset but is still referenced by a procedure, or
+ *	a variable that has been unset but it only being kept in existence (if
+ *	VAR_UNDEFINED) by a trace.
+ *
+ * Side effects:
+ *	The variable at arrayPtr may be converted to be an array if
+ *	createPart1 is 1. A new hashtable entry may be created if createPart2
+ *	is 1.
+ *	When createElem is 1, callers must incr elNamePtr if they plan
+ *	to decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Var *
+TclLookupArrayElement(
+    Tcl_Interp *interp,		/* Interpreter to use for lookup. */
+    Tcl_Obj *arrayNamePtr,	/* This is the name of the array, or NULL if
+				 * index>= 0. */
+    Tcl_Obj *elNamePtr,		/* Name of element within array. */
+    const int flags,		/* Only TCL_LEAVE_ERR_MSG bit matters. */
+    const char *msg,		/* Verb to use in error messages, e.g. "read"
+				 * or "set". Only needed if TCL_LEAVE_ERR_MSG
+				 * is set in flags. */
+    const int createArray,	/* If 1, transform arrayName to be an array if
+				 * it isn't one yet and the transformation is
+				 * possible. If 0, return error if it isn't
+				 * already an array. */
+    const int createElem,	/* If 1, create hash table entry for the
+				 * element, if it doesn't already exist. If 0,
+				 * return error if it doesn't exist. */
+    Var *arrayPtr,		/* Pointer to the array's Var structure. */
+    int index)			/* If >=0, the index of the local array. */
+{
+    int isNew;
+    Var *varPtr;
+    TclVarHashTable *tablePtr;
+    Namespace *nsPtr;
+
+    /*
+     * We're dealing with an array element. Make sure the variable is an array
+     * and look up the element (create the element if desired).
+     */
+
+    if (TclIsVarUndefined(arrayPtr) && !TclIsVarArrayElement(arrayPtr)) {
+	if (!createArray) {
+	    if (flags & TCL_LEAVE_ERR_MSG) {
+		TclObjVarErrMsg(interp, arrayNamePtr, elNamePtr, msg,
+			noSuchVar, index);
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+			arrayNamePtr?TclGetString(arrayNamePtr):NULL, NULL);
+	    }
+	    return NULL;
+	}
+
+	/*
+	 * Make sure we are not resurrecting a namespace variable from a
+	 * deleted namespace!
+	 */
+
+	if (TclIsVarDeadHash(arrayPtr)) {
+	    if (flags & TCL_LEAVE_ERR_MSG) {
+		TclObjVarErrMsg(interp, arrayNamePtr, elNamePtr, msg,
+			danglingVar, index);
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+			arrayNamePtr?TclGetString(arrayNamePtr):NULL, NULL);
+	    }
+	    return NULL;
+	}
+
+	TclSetVarArray(arrayPtr);
+	tablePtr = ckalloc(sizeof(TclVarHashTable));
+	arrayPtr->value.tablePtr = tablePtr;
+
+	if (TclIsVarInHash(arrayPtr) && TclGetVarNsPtr(arrayPtr)) {
+	    nsPtr = TclGetVarNsPtr(arrayPtr);
+	} else {
+	    nsPtr = NULL;
+	}
+	TclInitVarHashTable(arrayPtr->value.tablePtr, nsPtr);
+    } else if (!TclIsVarArray(arrayPtr)) {
+	if (flags & TCL_LEAVE_ERR_MSG) {
+	    TclObjVarErrMsg(interp, arrayNamePtr, elNamePtr, msg, needArray,
+		    index);
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+		    arrayNamePtr?TclGetString(arrayNamePtr):NULL, NULL);
+	}
+	return NULL;
+    }
+
+    if (createElem) {
+	varPtr = VarHashCreateVar(arrayPtr->value.tablePtr, elNamePtr,
+		&isNew);
+	if (isNew) {
+	    if (arrayPtr->flags & VAR_SEARCH_ACTIVE) {
+		DeleteSearches((Interp *) interp, arrayPtr);
+	    }
+	    TclSetVarArrayElement(varPtr);
+	}
+    } else {
+	varPtr = VarHashFindVar(arrayPtr->value.tablePtr, elNamePtr);
+	if (varPtr == NULL) {
+	    if (flags & TCL_LEAVE_ERR_MSG) {
+		TclObjVarErrMsg(interp, arrayNamePtr, elNamePtr, msg,
+			noSuchElement, index);
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ELEMENT",
+			TclGetString(elNamePtr), NULL);
+	    }
+	}
+    }
+    return varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetVar --
+ *
+ *	Return the value of a Tcl variable as a string.
+ *
+ * Results:
+ *	The return value points to the current value of varName as a string.
+ *	If the variable is not defined or can't be read because of a clash in
+ *	array usage then a NULL pointer is returned and an error message is
+ *	left in the interp's result if the TCL_LEAVE_ERR_MSG flag is set.
+ *	Note: the return value is only valid up until the next change to the
+ *	variable; if you depend on the value lasting longer than that, then
+ *	make yourself a private copy.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_GetVar
+const char *
+Tcl_GetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    const char *varName,	/* Name of a variable in interp. */
+    int flags)			/* OR-ed combination of TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY or TCL_LEAVE_ERR_MSG
+				 * bits. */
+{
+    Tcl_Obj *varNamePtr = Tcl_NewStringObj(varName, -1);
+    Tcl_Obj *resultPtr = Tcl_ObjGetVar2(interp, varNamePtr, NULL, flags);
+
+    TclDecrRefCount(varNamePtr);
+
+    if (resultPtr == NULL) {
+	return NULL;
+    }
+    return TclGetString(resultPtr);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetVar2 --
+ *
+ *	Return the value of a Tcl variable as a string, given a two-part name
+ *	consisting of array name and element within array.
+ *
+ * Results:
+ *	The return value points to the current value of the variable given by
+ *	part1 and part2 as a string. If the specified variable doesn't exist,
+ *	or if there is a clash in array usage, then NULL is returned and a
+ *	message will be left in the interp's result if the TCL_LEAVE_ERR_MSG
+ *	flag is set. Note: the return value is only valid up until the next
+ *	change to the variable; if you depend on the value lasting longer than
+ *	that, then make yourself a private copy.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_GetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    const char *part1,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    const char *part2,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    int flags)			/* OR-ed combination of TCL_GLOBAL_ONLY,
+				 * TCL_NAMESPACE_ONLY and TCL_LEAVE_ERR_MSG *
+				 * bits. */
+{
+    Tcl_Obj *resultPtr;
+    Tcl_Obj *part2Ptr = NULL, *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+	Tcl_IncrRefCount(part2Ptr);
+    }
+
+    resultPtr = Tcl_ObjGetVar2(interp, part1Ptr, part2Ptr, flags);
+
+    Tcl_DecrRefCount(part1Ptr);
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+    if (resultPtr == NULL) {
+	return NULL;
+    }
+    return TclGetString(resultPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetVar2Ex --
+ *
+ *	Return the value of a Tcl variable as a Tcl object, given a two-part
+ *	name consisting of array name and element within array.
+ *
+ * Results:
+ *	The return value points to the current object value of the variable
+ *	given by part1Ptr and part2Ptr. If the specified variable doesn't
+ *	exist, or if there is a clash in array usage, then NULL is returned
+ *	and a message will be left in the interpreter's result if the
+ *	TCL_LEAVE_ERR_MSG flag is set.
+ *
+ * Side effects:
+ *	The ref count for the returned object is _not_ incremented to reflect
+ *	the returned reference; if you want to keep a reference to the object
+ *	you must increment its ref count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_GetVar2Ex(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    const char *part1,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    const char *part2,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    int flags)			/* OR-ed combination of TCL_GLOBAL_ONLY, and
+				 * TCL_LEAVE_ERR_MSG bits. */
+{
+    Tcl_Obj *resPtr, *part2Ptr = NULL, *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+	Tcl_IncrRefCount(part2Ptr);
+    }
+
+    resPtr = Tcl_ObjGetVar2(interp, part1Ptr, part2Ptr, flags);
+
+    Tcl_DecrRefCount(part1Ptr);
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+
+    return resPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ObjGetVar2 --
+ *
+ *	Return the value of a Tcl variable as a Tcl object, given a two-part
+ *	name consisting of array name and element within array.
+ *
+ * Results:
+ *	The return value points to the current object value of the variable
+ *	given by part1Ptr and part2Ptr. If the specified variable doesn't
+ *	exist, or if there is a clash in array usage, then NULL is returned
+ *	and a message will be left in the interpreter's result if the
+ *	TCL_LEAVE_ERR_MSG flag is set.
+ *
+ * Side effects:
+ *	The ref count for the returned object is _not_ incremented to reflect
+ *	the returned reference; if you want to keep a reference to the object
+ *	you must increment its ref count yourself.
+ *
+ *	Callers must incr part2Ptr if they plan to decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ObjGetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    register Tcl_Obj *part1Ptr,	/* Points to an object holding the name of an
+				 * array (if part2 is non-NULL) or the name of
+				 * a variable. */
+    register Tcl_Obj *part2Ptr,	/* If non-null, points to an object holding
+				 * the name of an element in the array
+				 * part1Ptr. */
+    int flags)			/* OR-ed combination of TCL_GLOBAL_ONLY and
+				 * TCL_LEAVE_ERR_MSG bits. */
+{
+    Var *varPtr, *arrayPtr;
+
+    /*
+     * Filter to pass through only the flags this interface supports.
+     */
+
+    flags &= (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG);
+    varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr, flags, "read",
+	    /*createPart1*/ 0, /*createPart2*/ 1, &arrayPtr);
+    if (varPtr == NULL) {
+	return NULL;
+    }
+
+    return TclPtrGetVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrGetVar --
+ *
+ *	Return the value of a Tcl variable as a Tcl object, given the pointers
+ *	to the variable's (and possibly containing array's) VAR structure.
+ *
+ * Results:
+ *	The return value points to the current object value of the variable
+ *	given by varPtr. If the specified variable doesn't exist, or if there
+ *	is a clash in array usage, then NULL is returned and a message will be
+ *	left in the interpreter's result if the TCL_LEAVE_ERR_MSG flag is set.
+ *
+ * Side effects:
+ *	The ref count for the returned object is _not_ incremented to reflect
+ *	the returned reference; if you want to keep a reference to the object
+ *	you must increment its ref count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrGetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    Tcl_Var varPtr,		/* The variable to be read.*/
+    Tcl_Var arrayPtr,		/* NULL for scalar variables, pointer to the
+				 * containing array otherwise. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    const int flags)		/* OR-ed combination of TCL_GLOBAL_ONLY, and
+				 * TCL_LEAVE_ERR_MSG bits. */
+{
+    if (varPtr == NULL) {
+	Tcl_Panic("varPtr must not be NULL");
+    }
+    if (part1Ptr == NULL) {
+	Tcl_Panic("part1Ptr must not be NULL");
+    }
+    return TclPtrGetVarIdx(interp, (Var *) varPtr, (Var *) arrayPtr,
+	    part1Ptr, part2Ptr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrGetVarIdx --
+ *
+ *	Return the value of a Tcl variable as a Tcl object, given the pointers
+ *	to the variable's (and possibly containing array's) VAR structure.
+ *
+ * Results:
+ *	The return value points to the current object value of the variable
+ *	given by varPtr. If the specified variable doesn't exist, or if there
+ *	is a clash in array usage, then NULL is returned and a message will be
+ *	left in the interpreter's result if the TCL_LEAVE_ERR_MSG flag is set.
+ *
+ * Side effects:
+ *	The ref count for the returned object is _not_ incremented to reflect
+ *	the returned reference; if you want to keep a reference to the object
+ *	you must increment its ref count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrGetVarIdx(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    register Var *varPtr,	/* The variable to be read.*/
+    Var *arrayPtr,		/* NULL for scalar variables, pointer to the
+				 * containing array otherwise. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    const int flags,		/* OR-ed combination of TCL_GLOBAL_ONLY, and
+				 * TCL_LEAVE_ERR_MSG bits. */
+    int index)			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *msg;
+
+    /*
+     * Invoke any read traces that have been set for the variable.
+     */
+
+    if ((varPtr->flags & VAR_TRACED_READ)
+	    || (arrayPtr && (arrayPtr->flags & VAR_TRACED_READ))) {
+	if (TCL_ERROR == TclObjCallVarTraces(iPtr, arrayPtr, varPtr,
+		part1Ptr, part2Ptr,
+		(flags & (TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY))
+		| TCL_TRACE_READS, (flags & TCL_LEAVE_ERR_MSG), index)) {
+	    goto errorReturn;
+	}
+    }
+
+    /*
+     * Return the element if it's an existing scalar variable.
+     */
+
+    if (TclIsVarScalar(varPtr) && !TclIsVarUndefined(varPtr)) {
+	return varPtr->value.objPtr;
+    }
+
+    if (flags & TCL_LEAVE_ERR_MSG) {
+	if (TclIsVarUndefined(varPtr) && arrayPtr
+		&& !TclIsVarUndefined(arrayPtr)) {
+	    msg = noSuchElement;
+	} else if (TclIsVarArray(varPtr)) {
+	    msg = isArray;
+	} else {
+	    msg = noSuchVar;
+	}
+	TclObjVarErrMsg(interp, part1Ptr, part2Ptr, "read", msg, index);
+    }
+
+    /*
+     * An error. If the variable doesn't exist anymore and no-one's using it,
+     * then free up the relevant structures and hash table entries.
+     */
+
+  errorReturn:
+    Tcl_SetErrorCode(interp, "TCL", "READ", "VARNAME", NULL);
+    if (TclIsVarUndefined(varPtr)) {
+	TclCleanupVar(varPtr, arrayPtr);
+    }
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetObjCmd --
+ *
+ *	This function is invoked to process the "set" Tcl command. See the
+ *	user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result value.
+ *
+ * Side effects:
+ *	A variable's value may be changed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_SetObjCmd(
+    ClientData dummy,		/* Not used. */
+    register Tcl_Interp *interp,/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *varValueObj;
+
+    if (objc == 2) {
+	varValueObj = Tcl_ObjGetVar2(interp, objv[1], NULL,TCL_LEAVE_ERR_MSG);
+	if (varValueObj == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varValueObj);
+	return TCL_OK;
+    } else if (objc == 3) {
+	varValueObj = Tcl_ObjSetVar2(interp, objv[1], NULL, objv[2],
+		TCL_LEAVE_ERR_MSG);
+	if (varValueObj == NULL) {
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, varValueObj);
+	return TCL_OK;
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName ?newValue?");
+	return TCL_ERROR;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetVar --
+ *
+ *	Change the value of a variable.
+ *
+ * Results:
+ *	Returns a pointer to the malloc'ed string which is the character
+ *	representation of the variable's new value. The caller must not modify
+ *	this string. If the write operation was disallowed then NULL is
+ *	returned; if the TCL_LEAVE_ERR_MSG flag is set, then an explanatory
+ *	message will be left in the interp's result. Note that the returned
+ *	string may not be the same as newValue; this is because variable
+ *	traces may modify the variable's value.
+ *
+ * Side effects:
+ *	If varName is defined as a local or global variable in interp, its
+ *	value is changed to newValue. If varName isn't currently defined, then
+ *	a new global variable by that name is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_SetVar
+const char *
+Tcl_SetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    const char *varName,	/* Name of a variable in interp. */
+    const char *newValue,	/* New value for varName. */
+    int flags)			/* Various flags that tell how to set value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Tcl_Obj *varValuePtr = Tcl_SetVar2Ex(interp, varName, NULL,
+	    Tcl_NewStringObj(newValue, -1), flags);
+
+    if (varValuePtr == NULL) {
+	return NULL;
+    }
+    return TclGetString(varValuePtr);
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetVar2 --
+ *
+ *	Given a two-part variable name, which may refer either to a scalar
+ *	variable or an element of an array, change the value of the variable.
+ *	If the named scalar or array or element doesn't exist then create one.
+ *
+ * Results:
+ *	Returns a pointer to the malloc'ed string which is the character
+ *	representation of the variable's new value. The caller must not modify
+ *	this string. If the write operation was disallowed because an array
+ *	was expected but not found (or vice versa), then NULL is returned; if
+ *	the TCL_LEAVE_ERR_MSG flag is set, then an explanatory message will be
+ *	left in the interp's result. Note that the returned string may not be
+ *	the same as newValue; this is because variable traces may modify the
+ *	variable's value.
+ *
+ * Side effects:
+ *	The value of the given variable is set. If either the array or the
+ *	entry didn't exist then a new one is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+const char *
+Tcl_SetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    const char *part1,		/* If part2 is NULL, this is name of scalar
+				 * variable. Otherwise it is the name of an
+				 * array. */
+    const char *part2,		/* Name of an element within an array, or
+				 * NULL. */
+    const char *newValue,	/* New value for variable. */
+    int flags)			/* Various flags that tell how to set value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT, or
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Tcl_Obj *varValuePtr = Tcl_SetVar2Ex(interp, part1, part2,
+	    Tcl_NewStringObj(newValue, -1), flags);
+
+    if (varValuePtr == NULL) {
+	return NULL;
+    }
+    return TclGetString(varValuePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_SetVar2Ex --
+ *
+ *	Given a two-part variable name, which may refer either to a scalar
+ *	variable or an element of an array, change the value of the variable
+ *	to a new Tcl object value. If the named scalar or array or element
+ *	doesn't exist then create one.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the write operation was disallowed because an array was
+ *	expected but not found (or vice versa), then NULL is returned; if the
+ *	TCL_LEAVE_ERR_MSG flag is set, then an explanatory message will be
+ *	left in the interpreter's result. Note that the returned object may
+ *	not be the same one referenced by newValuePtr; this is because
+ *	variable traces may modify the variable's value.
+ *
+ * Side effects:
+ *	The value of the given variable is set. If either the array or the
+ *	entry didn't exist then a new variable is created.
+ *
+ *	The reference count is decremented for any old value of the variable
+ *	and incremented for its new value. If the new value for the variable
+ *	is not the same one referenced by newValuePtr (perhaps as a result of
+ *	a variable trace), then newValuePtr's ref count is left unchanged by
+ *	Tcl_SetVar2Ex. newValuePtr's ref count is also left unchanged if we
+ *	are appending it as a string value: that is, if "flags" includes
+ *	TCL_APPEND_VALUE but not TCL_LIST_ELEMENT.
+ *
+ *	The reference count for the returned object is _not_ incremented: if
+ *	you want to keep a reference to the object you must increment its ref
+ *	count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_SetVar2Ex(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be found. */
+    const char *part1,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    const char *part2,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    Tcl_Obj *newValuePtr,	/* New value for variable. */
+    int flags)			/* Various flags that tell how to set value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT or
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Tcl_Obj *resPtr, *part2Ptr = NULL, *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    Tcl_IncrRefCount(part1Ptr);
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+	Tcl_IncrRefCount(part2Ptr);
+    }
+
+    resPtr = Tcl_ObjSetVar2(interp, part1Ptr, part2Ptr, newValuePtr, flags);
+
+    Tcl_DecrRefCount(part1Ptr);
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+
+    return resPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ObjSetVar2 --
+ *
+ *	This function is the same as Tcl_SetVar2Ex above, except the variable
+ *	names are passed in Tcl object instead of strings.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the write operation was disallowed because an array was
+ *	expected but not found (or vice versa), then NULL is returned; if the
+ *	TCL_LEAVE_ERR_MSG flag is set, then an explanatory message will be
+ *	left in the interpreter's result. Note that the returned object may
+ *	not be the same one referenced by newValuePtr; this is because
+ *	variable traces may modify the variable's value.
+ *
+ * Side effects:
+ *	The value of the given variable is set. If either the array or the
+ *	entry didn't exist then a new variable is created.
+ *	Callers must Incr part1Ptr if they plan to Decr it.
+ *	Callers must Incr part2Ptr if they plan to Decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ObjSetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be found. */
+    register Tcl_Obj *part1Ptr,	/* Points to an object holding the name of an
+				 * array (if part2 is non-NULL) or the name of
+				 * a variable. */
+    register Tcl_Obj *part2Ptr,	/* If non-NULL, points to an object holding
+				 * the name of an element in the array
+				 * part1Ptr. */
+    Tcl_Obj *newValuePtr,	/* New value for variable. */
+    int flags)			/* Various flags that tell how to set value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT, or
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Var *varPtr, *arrayPtr;
+
+    /*
+     * Filter to pass through only the flags this interface supports.
+     */
+
+    flags &= (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG
+	    |TCL_APPEND_VALUE|TCL_LIST_ELEMENT);
+    varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr, flags, "set",
+	    /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+    if (varPtr == NULL) {
+	if (newValuePtr->refCount == 0) {
+	    Tcl_DecrRefCount(newValuePtr);
+	}
+	return NULL;
+    }
+
+    return TclPtrSetVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    newValuePtr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrSetVar --
+ *
+ *	This function is the same as Tcl_SetVar2Ex above, except that it
+ *	requires pointers to the variable's Var structs in addition to the
+ *	variable names.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the write operation was disallowed because an array was
+ *	expected but not found (or vice versa), then NULL is returned; if the
+ *	TCL_LEAVE_ERR_MSG flag is set, then an explanatory message will be
+ *	left in the interpreter's result. Note that the returned object may
+ *	not be the same one referenced by newValuePtr; this is because
+ *	variable traces may modify the variable's value.
+ *
+ * Side effects:
+ *	The value of the given variable is set. If either the array or the
+ *	entry didn't exist then a new variable is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrSetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    Tcl_Var varPtr,		/* Reference to the variable to set. */
+    Tcl_Var arrayPtr,		/* Reference to the array containing the
+				 * variable, or NULL if the variable is a
+				 * scalar. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    Tcl_Obj *newValuePtr,	/* New value for variable. */
+    const int flags)		/* OR-ed combination of TCL_GLOBAL_ONLY, and
+				 * TCL_LEAVE_ERR_MSG bits. */
+{
+    if (varPtr == NULL) {
+	Tcl_Panic("varPtr must not be NULL");
+    }
+    if (part1Ptr == NULL) {
+	Tcl_Panic("part1Ptr must not be NULL");
+    }
+    if (newValuePtr == NULL) {
+	Tcl_Panic("newValuePtr must not be NULL");
+    }
+    return TclPtrSetVarIdx(interp, (Var *) varPtr, (Var *) arrayPtr,
+	    part1Ptr, part2Ptr, newValuePtr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrSetVarIdx --
+ *
+ *	This function is the same as Tcl_SetVar2Ex above, except that it
+ *	requires pointers to the variable's Var structs in addition to the
+ *	variable names.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the write operation was disallowed because an array was
+ *	expected but not found (or vice versa), then NULL is returned; if the
+ *	TCL_LEAVE_ERR_MSG flag is set, then an explanatory message will be
+ *	left in the interpreter's result. Note that the returned object may
+ *	not be the same one referenced by newValuePtr; this is because
+ *	variable traces may modify the variable's value.
+ *
+ * Side effects:
+ *	The value of the given variable is set. If either the array or the
+ *	entry didn't exist then a new variable is created.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrSetVarIdx(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be looked up. */
+    register Var *varPtr,	/* Reference to the variable to set. */
+    Var *arrayPtr,		/* Reference to the array containing the
+				 * variable, or NULL if the variable is a
+				 * scalar. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. NULL if the 'index'
+				 * parameter is >= 0 */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    Tcl_Obj *newValuePtr,	/* New value for variable. */
+    const int flags,		/* OR-ed combination of TCL_GLOBAL_ONLY, and
+				 * TCL_LEAVE_ERR_MSG bits. */
+    int index)			/* Index of local var where part1 is to be
+				 * found. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *oldValuePtr;
+    Tcl_Obj *resultPtr = NULL;
+    int result;
+    int cleanupOnEarlyError = (newValuePtr->refCount == 0);
+
+    /*
+     * If the variable is in a hashtable and its hPtr field is NULL, then we
+     * may have an upvar to an array element where the array was deleted or an
+     * upvar to a namespace variable whose namespace was deleted. Generate an
+     * error (allowing the variable to be reset would screw up our storage
+     * allocation and is meaningless anyway).
+     */
+
+    if (TclIsVarDeadHash(varPtr)) {
+	if (flags & TCL_LEAVE_ERR_MSG) {
+	    if (TclIsVarArrayElement(varPtr)) {
+		TclObjVarErrMsg(interp, part1Ptr, part2Ptr, "set",
+			danglingElement, index);
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ELEMENT", NULL);
+	    } else {
+		TclObjVarErrMsg(interp, part1Ptr, part2Ptr, "set",
+			danglingVar, index);
+		Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME", NULL);
+	    }
+	}
+	goto earlyError;
+    }
+
+    /*
+     * It's an error to try to set an array variable itself.
+     */
+
+    if (TclIsVarArray(varPtr)) {
+	if (flags & TCL_LEAVE_ERR_MSG) {
+	    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, "set", isArray,index);
+	    Tcl_SetErrorCode(interp, "TCL", "WRITE", "ARRAY", NULL);
+	}
+	goto earlyError;
+    }
+
+    /*
+     * Invoke any read traces that have been set for the variable if it is
+     * requested. This was done for INST_LAPPEND_* but that was inconsistent
+     * with the non-bc instruction, and would cause failures trying to
+     * lappend to any non-existing ::env var, which is inconsistent with
+     * documented behavior. [Bug #3057639].
+     */
+
+    if ((flags & TCL_TRACE_READS) && ((varPtr->flags & VAR_TRACED_READ)
+	    || (arrayPtr && (arrayPtr->flags & VAR_TRACED_READ)))) {
+	if (TCL_ERROR == TclObjCallVarTraces(iPtr, arrayPtr, varPtr,
+		part1Ptr, part2Ptr,
+		TCL_TRACE_READS, (flags & TCL_LEAVE_ERR_MSG), index)) {
+	    goto earlyError;
+	}
+    }
+
+    /*
+     * Set the variable's new value. If appending, append the new value to the
+     * variable, either as a list element or as a string. Also, if appending,
+     * then if the variable's old value is unshared we can modify it directly,
+     * otherwise we must create a new copy to modify: this is "copy on write".
+     */
+
+    oldValuePtr = varPtr->value.objPtr;
+    if (flags & TCL_LIST_ELEMENT && !(flags & TCL_APPEND_VALUE)) {
+	varPtr->value.objPtr = NULL;
+    }
+    if (flags & (TCL_APPEND_VALUE|TCL_LIST_ELEMENT)) {
+	if (flags & TCL_LIST_ELEMENT) {		/* Append list element. */
+	    if (oldValuePtr == NULL) {
+		TclNewObj(oldValuePtr);
+		varPtr->value.objPtr = oldValuePtr;
+		Tcl_IncrRefCount(oldValuePtr);	/* Since var is referenced. */
+	    } else if (Tcl_IsShared(oldValuePtr)) {
+		varPtr->value.objPtr = Tcl_DuplicateObj(oldValuePtr);
+		TclDecrRefCount(oldValuePtr);
+		oldValuePtr = varPtr->value.objPtr;
+		Tcl_IncrRefCount(oldValuePtr);	/* Since var is referenced. */
+	    }
+	    result = Tcl_ListObjAppendElement(interp, oldValuePtr,
+		    newValuePtr);
+	    if (result != TCL_OK) {
+		goto earlyError;
+	    }
+	} else {				/* Append string. */
+	    /*
+	     * We append newValuePtr's bytes but don't change its ref count.
+	     */
+
+	    if (oldValuePtr == NULL) {
+		varPtr->value.objPtr = newValuePtr;
+		Tcl_IncrRefCount(newValuePtr);
+	    } else {
+		if (Tcl_IsShared(oldValuePtr)) {	/* Append to copy. */
+		    varPtr->value.objPtr = Tcl_DuplicateObj(oldValuePtr);
+
+		    TclContinuationsCopy(varPtr->value.objPtr, oldValuePtr);
+
+		    TclDecrRefCount(oldValuePtr);
+		    oldValuePtr = varPtr->value.objPtr;
+		    Tcl_IncrRefCount(oldValuePtr);	/* Since var is ref */
+		}
+		Tcl_AppendObjToObj(oldValuePtr, newValuePtr);
+		if (newValuePtr->refCount == 0) {
+		    Tcl_DecrRefCount(newValuePtr);
+		}
+	    }
+	}
+    } else if (newValuePtr != oldValuePtr) {
+	/*
+	 * In this case we are replacing the value, so we don't need to do
+	 * more than swap the objects.
+	 */
+
+	varPtr->value.objPtr = newValuePtr;
+	Tcl_IncrRefCount(newValuePtr);		/* Var is another ref. */
+	if (oldValuePtr != NULL) {
+	    TclDecrRefCount(oldValuePtr);	/* Discard old value. */
+	}
+    }
+
+    /*
+     * Invoke any write traces for the variable.
+     */
+
+    if ((varPtr->flags & VAR_TRACED_WRITE)
+	    || (arrayPtr && (arrayPtr->flags & VAR_TRACED_WRITE))) {
+	if (TCL_ERROR == TclObjCallVarTraces(iPtr, arrayPtr, varPtr, part1Ptr,
+		part2Ptr, (flags & (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY))
+		| TCL_TRACE_WRITES, (flags & TCL_LEAVE_ERR_MSG), index)) {
+	    goto cleanup;
+	}
+    }
+
+    /*
+     * Return the variable's value unless the variable was changed in some
+     * gross way by a trace (e.g. it was unset and then recreated as an
+     * array).
+     */
+
+    if (TclIsVarScalar(varPtr) && !TclIsVarUndefined(varPtr)) {
+	return varPtr->value.objPtr;
+    }
+
+    /*
+     * A trace changed the value in some gross way. Return an empty string
+     * object.
+     */
+
+    resultPtr = iPtr->emptyObjPtr;
+
+    /*
+     * If the variable doesn't exist anymore and no-one's using it, then free
+     * up the relevant structures and hash table entries.
+     */
+
+  cleanup:
+    if (resultPtr == NULL) {
+	Tcl_SetErrorCode(interp, "TCL", "WRITE", "VARNAME", NULL);
+    }
+    if (TclIsVarUndefined(varPtr)) {
+	TclCleanupVar(varPtr, arrayPtr);
+    }
+    return resultPtr;
+
+  earlyError:
+    if (cleanupOnEarlyError) {
+	Tcl_DecrRefCount(newValuePtr);
+    }
+    goto cleanup;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclIncrObjVar2 --
+ *
+ *	Given a two-part variable name, which may refer either to a scalar
+ *	variable or an element of an array, increment the Tcl object value of
+ *	the variable by a specified Tcl_Obj increment value.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the specified variable doesn't exist, or there is a clash
+ *	in array usage, or an error occurs while executing variable traces,
+ *	then NULL is returned and a message will be left in the interpreter's
+ *	result.
+ *
+ * Side effects:
+ *	The value of the given variable is incremented by the specified
+ *	amount. If either the array or the entry didn't exist then a new
+ *	variable is created. The ref count for the returned object is _not_
+ *	incremented to reflect the returned reference; if you want to keep a
+ *	reference to the object you must increment its ref count yourself.
+ *	Callers must Incr part1Ptr if they plan to Decr it.
+ *	Callers must Incr part2Ptr if they plan to Decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclIncrObjVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be found. */
+    Tcl_Obj *part1Ptr,		/* Points to an object holding the name of an
+				 * array (if part2 is non-NULL) or the name of
+				 * a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-null, points to an object holding
+				 * the name of an element in the array
+				 * part1Ptr. */
+    Tcl_Obj *incrPtr,		/* Amount to be added to variable. */
+    int flags)			/* Various flags that tell how to incr value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Var *varPtr, *arrayPtr;
+
+    varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr, flags, "read",
+	    1, 1, &arrayPtr);
+    if (varPtr == NULL) {
+	Tcl_AddErrorInfo(interp,
+		"\n    (reading value of variable to increment)");
+	return NULL;
+    }
+    return TclPtrIncrObjVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    incrPtr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrIncrObjVar --
+ *
+ *	Given the pointers to a variable and possible containing array,
+ *	increment the Tcl object value of the variable by a Tcl_Obj increment.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the specified variable doesn't exist, or there is a clash
+ *	in array usage, or an error occurs while executing variable traces,
+ *	then NULL is returned and a message will be left in the interpreter's
+ *	result.
+ *
+ * Side effects:
+ *	The value of the given variable is incremented by the specified
+ *	amount. If either the array or the entry didn't exist then a new
+ *	variable is created. The ref count for the returned object is _not_
+ *	incremented to reflect the returned reference; if you want to keep a
+ *	reference to the object you must increment its ref count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrIncrObjVar(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be found. */
+    Tcl_Var varPtr,		/* Reference to the variable to set. */
+    Tcl_Var arrayPtr,		/* Reference to the array containing the
+				 * variable, or NULL if the variable is a
+				 * scalar. */
+    Tcl_Obj *part1Ptr,		/* Points to an object holding the name of an
+				 * array (if part2 is non-NULL) or the name of
+				 * a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-null, points to an object holding
+				 * the name of an element in the array
+				 * part1Ptr. */
+    Tcl_Obj *incrPtr,		/* Increment value. */
+/* TODO: Which of these flag values really make sense? */
+    const int flags)		/* Various flags that tell how to incr value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    if (varPtr == NULL) {
+	Tcl_Panic("varPtr must not be NULL");
+    }
+    if (part1Ptr == NULL) {
+	Tcl_Panic("part1Ptr must not be NULL");
+    }
+    return TclPtrIncrObjVarIdx(interp, (Var *) varPtr, (Var *) arrayPtr,
+	    part1Ptr, part2Ptr, incrPtr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrIncrObjVarIdx --
+ *
+ *	Given the pointers to a variable and possible containing array,
+ *	increment the Tcl object value of the variable by a Tcl_Obj increment.
+ *
+ * Results:
+ *	Returns a pointer to the Tcl_Obj holding the new value of the
+ *	variable. If the specified variable doesn't exist, or there is a clash
+ *	in array usage, or an error occurs while executing variable traces,
+ *	then NULL is returned and a message will be left in the interpreter's
+ *	result.
+ *
+ * Side effects:
+ *	The value of the given variable is incremented by the specified
+ *	amount. If either the array or the entry didn't exist then a new
+ *	variable is created. The ref count for the returned object is _not_
+ *	incremented to reflect the returned reference; if you want to keep a
+ *	reference to the object you must increment its ref count yourself.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+TclPtrIncrObjVarIdx(
+    Tcl_Interp *interp,		/* Command interpreter in which variable is to
+				 * be found. */
+    Var *varPtr,		/* Reference to the variable to set. */
+    Var *arrayPtr,		/* Reference to the array containing the
+				 * variable, or NULL if the variable is a
+				 * scalar. */
+    Tcl_Obj *part1Ptr,		/* Points to an object holding the name of an
+				 * array (if part2 is non-NULL) or the name of
+				 * a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-null, points to an object holding
+				 * the name of an element in the array
+				 * part1Ptr. */
+    Tcl_Obj *incrPtr,		/* Increment value. */
+/* TODO: Which of these flag values really make sense? */
+    const int flags,		/* Various flags that tell how to incr value:
+				 * any of TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_APPEND_VALUE, TCL_LIST_ELEMENT,
+				 * TCL_LEAVE_ERR_MSG. */
+    int index)			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+{
+    register Tcl_Obj *varValuePtr;
+
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)++;
+    }
+    varValuePtr = TclPtrGetVarIdx(interp, varPtr, arrayPtr, part1Ptr,
+	    part2Ptr, flags, index);
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)--;
+    }
+    if (varValuePtr == NULL) {
+	varValuePtr = Tcl_NewIntObj(0);
+    }
+    if (Tcl_IsShared(varValuePtr)) {
+	/* Copy on write */
+	varValuePtr = Tcl_DuplicateObj(varValuePtr);
+
+	if (TCL_OK == TclIncrObj(interp, varValuePtr, incrPtr)) {
+	    return TclPtrSetVarIdx(interp, varPtr, arrayPtr, part1Ptr,
+		    part2Ptr, varValuePtr, flags, index);
+	} else {
+	    Tcl_DecrRefCount(varValuePtr);
+	    return NULL;
+	}
+    } else {
+	/* Unshared - can Incr in place */
+	if (TCL_OK == TclIncrObj(interp, varValuePtr, incrPtr)) {
+
+	    /*
+	     * This seems dumb to write the incremeted value into the var
+	     * after we just adjusted the value in place, but the spec for
+	     * [incr] requires that write traces fire, and making this call
+	     * is the way to make that happen.
+	     */
+
+	    return TclPtrSetVarIdx(interp, varPtr, arrayPtr, part1Ptr,
+		    part2Ptr, varValuePtr, flags, index);
+	} else {
+	    return NULL;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnsetVar --
+ *
+ *	Delete a variable, so that it may not be accessed anymore.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable was successfully deleted, TCL_ERROR if
+ *	the variable can't be unset. In the event of an error, if the
+ *	TCL_LEAVE_ERR_MSG flag is set then an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	If varName is defined as a local or global variable in interp, it is
+ *	deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_UnsetVar
+int
+Tcl_UnsetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    const char *varName,	/* Name of a variable in interp. May be either
+				 * a scalar name or an array name or an
+				 * element in an array. */
+    int flags)			/* OR-ed combination of any of
+				 * TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY or
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    int result;
+    Tcl_Obj *varNamePtr;
+
+    varNamePtr = Tcl_NewStringObj(varName, -1);
+    Tcl_IncrRefCount(varNamePtr);
+
+    /*
+     * Filter to pass through only the flags this interface supports.
+     */
+
+    flags &= (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG);
+    result = TclObjUnsetVar2(interp, varNamePtr, NULL, flags);
+
+    Tcl_DecrRefCount(varNamePtr);
+    return result;
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnsetVar2 --
+ *
+ *	Delete a variable, given a 2-part name.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable was successfully deleted, TCL_ERROR if
+ *	the variable can't be unset. In the event of an error, if the
+ *	TCL_LEAVE_ERR_MSG flag is set then an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	If part1 and part2 indicate a local or global variable in interp, it
+ *	is deleted. If part1 is an array name and part2 is NULL, then the
+ *	whole array is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UnsetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    const char *part1,		/* Name of variable or array. */
+    const char *part2,		/* Name of element within array or NULL. */
+    int flags)			/* OR-ed combination of any of
+				 * TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    int result;
+    Tcl_Obj *part2Ptr = NULL, *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+    }
+
+    /*
+     * Filter to pass through only the flags this interface supports.
+     */
+
+    flags &= (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY|TCL_LEAVE_ERR_MSG);
+    result = TclObjUnsetVar2(interp, part1Ptr, part2Ptr, flags);
+
+    Tcl_DecrRefCount(part1Ptr);
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjUnsetVar2 --
+ *
+ *	Delete a variable, given a 2-object name.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable was successfully deleted, TCL_ERROR if
+ *	the variable can't be unset. In the event of an error, if the
+ *	TCL_LEAVE_ERR_MSG flag is set then an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	If part1ptr and part2Ptr indicate a local or global variable in
+ *	interp, it is deleted. If part1Ptr is an array name and part2Ptr is
+ *	NULL, then the whole array is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclObjUnsetVar2(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    Tcl_Obj *part1Ptr,		/* Name of variable or array. */
+    Tcl_Obj *part2Ptr,		/* Name of element within array or NULL. */
+    int flags)			/* OR-ed combination of any of
+				 * TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    Var *varPtr, *arrayPtr;
+
+    varPtr = TclObjLookupVarEx(interp, part1Ptr, part2Ptr, flags, "unset",
+	    /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    return TclPtrUnsetVarIdx(interp, varPtr, arrayPtr, part1Ptr, part2Ptr,
+	    flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrUnsetVar --
+ *
+ *	Delete a variable, given the pointers to the variable's (and possibly
+ *	containing array's) VAR structure.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable was successfully deleted, TCL_ERROR if
+ *	the variable can't be unset. In the event of an error, if the
+ *	TCL_LEAVE_ERR_MSG flag is set then an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	If varPtr and arrayPtr indicate a local or global variable in interp,
+ *	it is deleted. If varPtr is an array reference and part2Ptr is NULL,
+ *	then the whole array is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPtrUnsetVar(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    Tcl_Var varPtr,		/* The variable to be unset. */
+    Tcl_Var arrayPtr,		/* NULL for scalar variables, pointer to the
+				 * containing array otherwise. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    const int flags)		/* OR-ed combination of any of
+				 * TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_LEAVE_ERR_MSG. */
+{
+    if (varPtr == NULL) {
+	Tcl_Panic("varPtr must not be NULL");
+    }
+    if (part1Ptr == NULL) {
+	Tcl_Panic("part1Ptr must not be NULL");
+    }
+    return TclPtrUnsetVarIdx(interp, (Var *) varPtr, (Var *) arrayPtr,
+	    part1Ptr, part2Ptr, flags, -1);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrUnsetVarIdx --
+ *
+ *	Delete a variable, given the pointers to the variable's (and possibly
+ *	containing array's) VAR structure.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable was successfully deleted, TCL_ERROR if
+ *	the variable can't be unset. In the event of an error, if the
+ *	TCL_LEAVE_ERR_MSG flag is set then an error message is left in the
+ *	interp's result.
+ *
+ * Side effects:
+ *	If varPtr and arrayPtr indicate a local or global variable in interp,
+ *	it is deleted. If varPtr is an array reference and part2Ptr is NULL,
+ *	then the whole array is deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPtrUnsetVarIdx(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    register Var *varPtr,	/* The variable to be unset. */
+    Var *arrayPtr,		/* NULL for scalar variables, pointer to the
+				 * containing array otherwise. */
+    Tcl_Obj *part1Ptr,		/* Name of an array (if part2 is non-NULL) or
+				 * the name of a variable. */
+    Tcl_Obj *part2Ptr,		/* If non-NULL, gives the name of an element
+				 * in the array part1. */
+    const int flags,		/* OR-ed combination of any of
+				 * TCL_GLOBAL_ONLY, TCL_NAMESPACE_ONLY,
+				 * TCL_LEAVE_ERR_MSG. */
+    int index)			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+{
+    Interp *iPtr = (Interp *) interp;
+    int result = (TclIsVarUndefined(varPtr)? TCL_ERROR : TCL_OK);
+
+    /*
+     * Keep the variable alive until we're done with it. We used to
+     * increase/decrease the refCount for each operation, making it hard to
+     * find [Bug 735335] - caused by unsetting the variable whose value was
+     * the variable's name.
+     */
+
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)++;
+    }
+
+    UnsetVarStruct(varPtr, arrayPtr, iPtr, part1Ptr, part2Ptr, flags, index);
+
+    /*
+     * It's an error to unset an undefined variable.
+     */
+
+    if (result != TCL_OK) {
+	if (flags & TCL_LEAVE_ERR_MSG) {
+	    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, "unset",
+		    ((arrayPtr == NULL) ? noSuchVar : noSuchElement), index);
+	    Tcl_SetErrorCode(interp, "TCL", "UNSET", "VARNAME", NULL);
+	}
+    }
+
+    /*
+     * Finally, if the variable is truly not in use then free up its Var
+     * structure and remove it from its hash table, if any. The ref count of
+     * its value object, if any, was decremented above.
+     */
+
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)--;
+	CleanupVar(varPtr, arrayPtr);
+    }
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * UnsetVarStruct --
+ *
+ *	Unset and delete a variable. This does the internal work for
+ *	TclObjUnsetVar2 and TclDeleteNamespaceVars, which call here for each
+ *	variable to be unset and deleted.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	If the arguments indicate a local or global variable in iPtr, it is
+ *	unset and deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+UnsetVarStruct(
+    Var *varPtr,
+    Var *arrayPtr,
+    Interp *iPtr,
+    Tcl_Obj *part1Ptr,
+    Tcl_Obj *part2Ptr,
+    int flags,
+    int index)
+{
+    Var dummyVar;
+    int traced = TclIsVarTraced(varPtr)
+	    || (arrayPtr && (arrayPtr->flags & VAR_TRACED_UNSET));
+
+    if (arrayPtr && (arrayPtr->flags & VAR_SEARCH_ACTIVE)) {
+	DeleteSearches(iPtr, arrayPtr);
+    } else if (varPtr->flags & VAR_SEARCH_ACTIVE) {
+	DeleteSearches(iPtr, varPtr);
+    }
+
+    /*
+     * The code below is tricky, because of the possibility that a trace
+     * function might try to access a variable being deleted. To handle this
+     * situation gracefully, do things in three steps:
+     * 1. Copy the contents of the variable to a dummy variable structure, and
+     *    mark the original Var structure as undefined.
+     * 2. Invoke traces and clean up the variable, using the dummy copy.
+     * 3. If at the end of this the original variable is still undefined and
+     *    has no outstanding references, then delete it (but it could have
+     *    gotten recreated by a trace).
+     */
+
+    dummyVar = *varPtr;
+    dummyVar.flags &= ~VAR_ALL_HASH;
+    TclSetVarUndefined(varPtr);
+
+    /*
+     * Call trace functions for the variable being deleted. Then delete its
+     * traces. Be sure to abort any other traces for the variable that are
+     * still pending. Special tricks:
+     * 1. We need to increment varPtr's refCount around this: TclCallVarTraces
+     *    will use dummyVar so it won't increment varPtr's refCount itself.
+     * 2. Turn off the VAR_TRACE_ACTIVE flag in dummyVar: we want to call
+     *    unset traces even if other traces are pending.
+     */
+
+    if (traced) {
+	VarTrace *tracePtr = NULL;
+	Tcl_HashEntry *tPtr;
+
+	if (TclIsVarTraced(&dummyVar)) {
+	    /*
+	     * Transfer any existing traces on var, IF there are unset traces.
+	     * Otherwise just delete them.
+	     */
+
+	    int isNew;
+
+	    tPtr = Tcl_FindHashEntry(&iPtr->varTraces, varPtr);
+	    tracePtr = Tcl_GetHashValue(tPtr);
+	    varPtr->flags &= ~VAR_ALL_TRACES;
+	    Tcl_DeleteHashEntry(tPtr);
+	    if (dummyVar.flags & VAR_TRACED_UNSET) {
+		tPtr = Tcl_CreateHashEntry(&iPtr->varTraces,
+			&dummyVar, &isNew);
+		Tcl_SetHashValue(tPtr, tracePtr);
+	    }
+	}
+
+	if ((dummyVar.flags & VAR_TRACED_UNSET)
+		|| (arrayPtr && (arrayPtr->flags & VAR_TRACED_UNSET))) {
+	    dummyVar.flags &= ~VAR_TRACE_ACTIVE;
+	    TclObjCallVarTraces(iPtr, arrayPtr, &dummyVar, part1Ptr, part2Ptr,
+		    (flags & (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY))
+			    | TCL_TRACE_UNSETS,
+		    /* leaveErrMsg */ 0, index);
+
+	    /*
+	     * The traces that we just called may have triggered a change in
+	     * the set of traces. If so, reload the traces to manipulate.
+	     */
+
+	    tracePtr = NULL;
+	    if (TclIsVarTraced(&dummyVar)) {
+		tPtr = Tcl_FindHashEntry(&iPtr->varTraces, &dummyVar);
+		if (tPtr) {
+		    tracePtr = Tcl_GetHashValue(tPtr);
+		    Tcl_DeleteHashEntry(tPtr);
+		}
+	    }
+	}
+
+	if (tracePtr) {
+	    ActiveVarTrace *activePtr;
+
+	    while (tracePtr) {
+		VarTrace *prevPtr = tracePtr;
+
+		tracePtr = tracePtr->nextPtr;
+		prevPtr->nextPtr = NULL;
+		Tcl_EventuallyFree(prevPtr, TCL_DYNAMIC);
+	    }
+	    for (activePtr = iPtr->activeVarTracePtr;  activePtr != NULL;
+		    activePtr = activePtr->nextPtr) {
+		if (activePtr->varPtr == varPtr) {
+		    activePtr->nextTracePtr = NULL;
+		}
+	    }
+	    dummyVar.flags &= ~VAR_ALL_TRACES;
+	}
+    }
+
+    if (TclIsVarScalar(&dummyVar) && (dummyVar.value.objPtr != NULL)) {
+	/*
+	 * Decrement the ref count of the var's value.
+	 */
+
+	Tcl_Obj *objPtr = dummyVar.value.objPtr;
+
+	TclDecrRefCount(objPtr);
+    } else if (TclIsVarArray(&dummyVar)) {
+	/*
+	 * If the variable is an array, delete all of its elements. This must
+	 * be done after calling and deleting the traces on the array, above
+	 * (that's the way traces are defined). If the array name is not
+	 * present and is required for a trace on some element, it will be
+	 * computed at DeleteArray.
+	 */
+
+	DeleteArray(iPtr, part1Ptr, (Var *) &dummyVar, (flags
+		& (TCL_GLOBAL_ONLY|TCL_NAMESPACE_ONLY)) | TCL_TRACE_UNSETS,
+		index);
+    } else if (TclIsVarLink(&dummyVar)) {
+	/*
+	 * For global/upvar variables referenced in procedures, decrement the
+	 * reference count on the variable referred to, and free the
+	 * referenced variable if it's no longer needed.
+	 */
+
+	Var *linkPtr = dummyVar.value.linkPtr;
+
+	if (TclIsVarInHash(linkPtr)) {
+	    VarHashRefCount(linkPtr)--;
+	    CleanupVar(linkPtr, NULL);
+	}
+    }
+
+    /*
+     * If the variable was a namespace variable, decrement its reference
+     * count.
+     */
+
+    TclClearVarNamespaceVar(varPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UnsetObjCmd --
+ *
+ *	This object-based function is invoked to process the "unset" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_UnsetObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    register int i, flags = TCL_LEAVE_ERR_MSG;
+    register const char *name;
+
+    if (objc == 1) {
+	/*
+	 * Do nothing if no arguments supplied, so as to match command
+	 * documentation.
+	 */
+
+	return TCL_OK;
+    }
+
+    /*
+     * Simple, restrictive argument parsing. The only options are -- and
+     * -nocomplain (which must come first and be given exactly to be an
+     * option).
+     */
+
+    i = 1;
+    name = TclGetString(objv[i]);
+    if (name[0] == '-') {
+	if (strcmp("-nocomplain", name) == 0) {
+	    i++;
+	    if (i == objc) {
+		return TCL_OK;
+	    }
+	    flags = 0;
+	    name = TclGetString(objv[i]);
+	}
+	if (strcmp("--", name) == 0) {
+	    i++;
+	}
+    }
+
+    for (; i < objc; i++) {
+	if ((TclObjUnsetVar2(interp, objv[i], NULL, flags) != TCL_OK)
+		&& (flags == TCL_LEAVE_ERR_MSG)) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_AppendObjCmd --
+ *
+ *	This object-based function is invoked to process the "append" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	A variable's value may be changed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_AppendObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Var *varPtr, *arrayPtr;
+    register Tcl_Obj *varValuePtr = NULL;
+				/* Initialized to avoid compiler warning. */
+    int i;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName ?value ...?");
+	return TCL_ERROR;
+    }
+
+    if (objc == 2) {
+	varValuePtr = Tcl_ObjGetVar2(interp, objv[1], NULL,TCL_LEAVE_ERR_MSG);
+	if (varValuePtr == NULL) {
+	    return TCL_ERROR;
+	}
+    } else {
+	varPtr = TclObjLookupVarEx(interp, objv[1], NULL, TCL_LEAVE_ERR_MSG,
+		"set", /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+	if (varPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	for (i=2 ; i<objc ; i++) {
+	    /*
+	     * Note that we do not need to increase the refCount of the Var
+	     * pointers: should a trace delete the variable, the return value
+	     * of TclPtrSetVarIdx will be NULL or emptyObjPtr, and we will not
+	     * access the variable again.
+	     */
+
+	    varValuePtr = TclPtrSetVarIdx(interp, varPtr, arrayPtr, objv[1],
+		    NULL, objv[i], TCL_APPEND_VALUE|TCL_LEAVE_ERR_MSG, -1);
+	    if ((varValuePtr == NULL) ||
+		    (varValuePtr == ((Interp *) interp)->emptyObjPtr)) {
+		return TCL_ERROR;
+	    }
+	}
+    }
+    Tcl_SetObjResult(interp, varValuePtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_LappendObjCmd --
+ *
+ *	This object-based function is invoked to process the "lappend" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	A variable's value may be changed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_LappendObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Tcl_Obj *varValuePtr, *newValuePtr;
+    int numElems, createdNewObj;
+    Var *varPtr, *arrayPtr;
+    int result;
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "varName ?value ...?");
+	return TCL_ERROR;
+    }
+    if (objc == 2) {
+	newValuePtr = Tcl_ObjGetVar2(interp, objv[1], NULL, 0);
+	if (newValuePtr == NULL) {
+	    /*
+	     * The variable doesn't exist yet. Just create it with an empty
+	     * initial value.
+	     */
+
+	    TclNewObj(varValuePtr);
+	    newValuePtr = Tcl_ObjSetVar2(interp, objv[1], NULL, varValuePtr,
+		    TCL_LEAVE_ERR_MSG);
+	    if (newValuePtr == NULL) {
+		return TCL_ERROR;
+	    }
+	} else {
+	    result = TclListObjLength(interp, newValuePtr, &numElems);
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+    } else {
+	/*
+	 * We have arguments to append. We used to call Tcl_SetVar2 to append
+	 * each argument one at a time to ensure that traces were run for each
+	 * append step. We now append the arguments all at once because it's
+	 * faster. Note that a read trace and a write trace for the variable
+	 * will now each only be called once. Also, if the variable's old
+	 * value is unshared we modify it directly, otherwise we create a new
+	 * copy to modify: this is "copy on write".
+	 */
+
+	createdNewObj = 0;
+
+	/*
+	 * Protect the variable pointers around the TclPtrGetVarIdx call
+	 * to insure that they remain valid even if the variable was undefined
+	 * and unused.
+	 */
+
+	varPtr = TclObjLookupVarEx(interp, objv[1], NULL, TCL_LEAVE_ERR_MSG,
+		"set", /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+	if (varPtr == NULL) {
+	    return TCL_ERROR;
+	}
+	if (TclIsVarInHash(varPtr)) {
+	    VarHashRefCount(varPtr)++;
+	}
+	if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	    VarHashRefCount(arrayPtr)++;
+	}
+	varValuePtr = TclPtrGetVarIdx(interp, varPtr, arrayPtr, objv[1], NULL,
+		TCL_LEAVE_ERR_MSG, -1);
+	if (TclIsVarInHash(varPtr)) {
+	    VarHashRefCount(varPtr)--;
+	}
+	if (arrayPtr && TclIsVarInHash(arrayPtr)) {
+	    VarHashRefCount(arrayPtr)--;
+	}
+
+	if (varValuePtr == NULL) {
+	    /*
+	     * We couldn't read the old value: either the var doesn't yet
+	     * exist or it's an array element. If it's new, we will try to
+	     * create it with Tcl_ObjSetVar2 below.
+	     */
+
+	    TclNewObj(varValuePtr);
+	    createdNewObj = 1;
+	} else if (Tcl_IsShared(varValuePtr)) {
+	    varValuePtr = Tcl_DuplicateObj(varValuePtr);
+	    createdNewObj = 1;
+	}
+
+	result = TclListObjLength(interp, varValuePtr, &numElems);
+	if (result == TCL_OK) {
+	    result = Tcl_ListObjReplace(interp, varValuePtr, numElems, 0,
+		    (objc-2), (objv+2));
+	}
+	if (result != TCL_OK) {
+	    if (createdNewObj) {
+		TclDecrRefCount(varValuePtr); /* Free unneeded obj. */
+	    }
+	    return result;
+	}
+
+	/*
+	 * Now store the list object back into the variable. If there is an
+	 * error setting the new value, decrement its ref count if it was new
+	 * and we didn't create the variable.
+	 */
+
+	newValuePtr = TclPtrSetVarIdx(interp, varPtr, arrayPtr, objv[1], NULL,
+		varValuePtr, TCL_LEAVE_ERR_MSG, -1);
+	if (newValuePtr == NULL) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Set the interpreter's object result to refer to the variable's value
+     * object.
+     */
+
+    Tcl_SetObjResult(interp, newValuePtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclArraySet --
+ *
+ *	Set the elements of an array. If there are no elements to set, create
+ *	an empty array. This routine is used by the Tcl_ArrayObjCmd and by the
+ *	TclSetupEnv routine.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	A variable will be created if one does not already exist.
+ *	Callers must Incr arrayNameObj if they pland to Decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclArraySet(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Obj *arrayNameObj,	/* The array name. */
+    Tcl_Obj *arrayElemObj)	/* The array elements list or dict. If this is
+				 * NULL, create an empty array. */
+{
+    Var *varPtr, *arrayPtr;
+    int result, i;
+
+    varPtr = TclObjLookupVarEx(interp, arrayNameObj, NULL,
+	    /*flags*/ TCL_LEAVE_ERR_MSG, /*msg*/ "set", /*createPart1*/ 1,
+	    /*createPart2*/ 1, &arrayPtr);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+    if (arrayPtr) {
+	CleanupVar(varPtr, arrayPtr);
+	TclObjVarErrMsg(interp, arrayNameObj, NULL, "set", needArray, -1);
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+		TclGetString(arrayNameObj), NULL);
+	return TCL_ERROR;
+    }
+
+    if (arrayElemObj == NULL) {
+	goto ensureArray;
+    }
+
+    /*
+     * Install the contents of the dictionary or list into the array.
+     */
+
+    if (arrayElemObj->typePtr == &tclDictType) {
+	Tcl_Obj *keyPtr, *valuePtr;
+	Tcl_DictSearch search;
+	int done;
+
+	if (Tcl_DictObjSize(interp, arrayElemObj, &done) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (done == 0) {
+	    /*
+	     * Empty, so we'll just force the array to be properly existing
+	     * instead.
+	     */
+
+	    goto ensureArray;
+	}
+
+	/*
+	 * Don't need to look at result of Tcl_DictObjFirst as we've just
+	 * successfully used a dictionary operation on the same object.
+	 */
+
+	for (Tcl_DictObjFirst(interp, arrayElemObj, &search,
+		&keyPtr, &valuePtr, &done) ; !done ;
+		Tcl_DictObjNext(&search, &keyPtr, &valuePtr, &done)) {
+	    /*
+	     * At this point, it would be nice if the key was directly usable
+	     * by the array. This isn't the case though.
+	     */
+
+	    Var *elemVarPtr = TclLookupArrayElement(interp, arrayNameObj,
+		    keyPtr, TCL_LEAVE_ERR_MSG, "set", 1, 1, varPtr, -1);
+
+	    if ((elemVarPtr == NULL) ||
+		    (TclPtrSetVarIdx(interp, elemVarPtr, varPtr, arrayNameObj,
+		    keyPtr, valuePtr, TCL_LEAVE_ERR_MSG, -1) == NULL)) {
+		Tcl_DictObjDone(&search);
+		return TCL_ERROR;
+	    }
+	}
+	return TCL_OK;
+    } else {
+	/*
+	 * Not a dictionary, so assume (and convert to, for backward-
+	 * -compatibility reasons) a list.
+	 */
+
+	int elemLen;
+	Tcl_Obj **elemPtrs, *copyListObj;
+
+	result = TclListObjGetElements(interp, arrayElemObj,
+		&elemLen, &elemPtrs);
+	if (result != TCL_OK) {
+	    return result;
+	}
+	if (elemLen & 1) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "list must have an even number of elements", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ARGUMENT", "FORMAT", NULL);
+	    return TCL_ERROR;
+	}
+	if (elemLen == 0) {
+	    goto ensureArray;
+	}
+
+	/*
+	 * We needn't worry about traces invalidating arrayPtr: should that be
+	 * the case, TclPtrSetVarIdx will return NULL so that we break out of
+	 * the loop and return an error.
+	 */
+
+	copyListObj = TclListObjCopy(NULL, arrayElemObj);
+	for (i=0 ; i<elemLen ; i+=2) {
+	    Var *elemVarPtr = TclLookupArrayElement(interp, arrayNameObj,
+		    elemPtrs[i], TCL_LEAVE_ERR_MSG, "set", 1, 1, varPtr, -1);
+
+	    if ((elemVarPtr == NULL) ||
+		    (TclPtrSetVarIdx(interp, elemVarPtr, varPtr, arrayNameObj,
+		    elemPtrs[i],elemPtrs[i+1],TCL_LEAVE_ERR_MSG,-1) == NULL)){
+		result = TCL_ERROR;
+		break;
+	    }
+	}
+	Tcl_DecrRefCount(copyListObj);
+	return result;
+    }
+
+    /*
+     * The list is empty make sure we have an array, or create one if
+     * necessary.
+     */
+
+  ensureArray:
+    if (varPtr != NULL) {
+	if (TclIsVarArray(varPtr)) {
+	    /*
+	     * Already an array, done.
+	     */
+
+	    return TCL_OK;
+	}
+	if (TclIsVarArrayElement(varPtr) || !TclIsVarUndefined(varPtr)) {
+	    /*
+	     * Either an array element, or a scalar: lose!
+	     */
+
+	    TclObjVarErrMsg(interp, arrayNameObj, NULL, "array set",
+		    needArray, -1);
+	    Tcl_SetErrorCode(interp, "TCL", "WRITE", "ARRAY", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    TclSetVarArray(varPtr);
+    varPtr->value.tablePtr = ckalloc(sizeof(TclVarHashTable));
+    TclInitVarHashTable(varPtr->value.tablePtr, TclGetVarNsPtr(varPtr));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayStartSearchCmd --
+ *
+ *	This object-based function is invoked to process the "array
+ *	startsearch" Tcl command. See the user documentation for details on
+ *	what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+
+static Var *
+VerifyArray(
+    Tcl_Interp *interp,
+    Tcl_Obj *varNameObj)
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *varName = TclGetString(varNameObj);
+    Var *arrayPtr;
+
+    /*
+     * Locate the array variable.
+     */
+
+    Var *varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces.
+     */
+
+    if ((varPtr == NULL) || !TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" isn't an array", varName));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ARRAY", varName, NULL);
+	return NULL;
+    }
+
+    return varPtr;
+}
+
+static int
+ArrayStartSearchCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr;
+    Tcl_HashEntry *hPtr;
+    int isNew;
+    ArraySearch *searchPtr;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName");
+	return TCL_ERROR;
+    }
+
+    varPtr = VerifyArray(interp, objv[1]);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Make a new array search with a free name.
+     */
+
+    searchPtr = ckalloc(sizeof(ArraySearch));
+    hPtr = Tcl_CreateHashEntry(&iPtr->varSearches, varPtr, &isNew);
+    if (isNew) {
+	searchPtr->id = 1;
+	varPtr->flags |= VAR_SEARCH_ACTIVE;
+	searchPtr->nextPtr = NULL;
+    } else {
+	searchPtr->id = ((ArraySearch *) Tcl_GetHashValue(hPtr))->id + 1;
+	searchPtr->nextPtr = Tcl_GetHashValue(hPtr);
+    }
+    searchPtr->varPtr = varPtr;
+    searchPtr->nextEntry = VarHashFirstEntry(varPtr->value.tablePtr,
+	    &searchPtr->search);
+    Tcl_SetHashValue(hPtr, searchPtr);
+    searchPtr->name = Tcl_ObjPrintf("s-%d-%s", searchPtr->id, TclGetString(objv[1]));
+    Tcl_IncrRefCount(searchPtr->name);
+    Tcl_SetObjResult(interp, searchPtr->name);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayAnyMoreCmd --
+ *
+ *	This object-based function is invoked to process the "array anymore"
+ *	Tcl command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayAnyMoreCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr;
+    Tcl_Obj *varNameObj, *searchObj;
+    int gotValue;
+    ArraySearch *searchPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName searchId");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+    searchObj = objv[2];
+
+    varPtr = VerifyArray(interp, varNameObj);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the search.
+     */
+
+    searchPtr = ParseSearchId(interp, varPtr, varNameObj, searchObj);
+    if (searchPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Scan forward to find if there are any further elements in the array
+     * that are defined.
+     */
+
+    while (1) {
+	if (searchPtr->nextEntry != NULL) {
+	    varPtr = VarHashGetValue(searchPtr->nextEntry);
+	    if (!TclIsVarUndefined(varPtr)) {
+		gotValue = 1;
+		break;
+	    }
+	}
+	searchPtr->nextEntry = Tcl_NextHashEntry(&searchPtr->search);
+	if (searchPtr->nextEntry == NULL) {
+	    gotValue = 0;
+	    break;
+	}
+    }
+    Tcl_SetObjResult(interp, iPtr->execEnvPtr->constants[gotValue]);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayNextElementCmd --
+ *
+ *	This object-based function is invoked to process the "array
+ *	nextelement" Tcl command. See the user documentation for details on
+ *	what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayNextElementCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Var *varPtr;
+    Tcl_Obj *varNameObj, *searchObj;
+    ArraySearch *searchPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName searchId");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+    searchObj = objv[2];
+
+    varPtr = VerifyArray(interp, varNameObj);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the search.
+     */
+
+    searchPtr = ParseSearchId(interp, varPtr, varNameObj, searchObj);
+    if (searchPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the next element from the search, or the empty string on
+     * exhaustion. Note that the [array anymore] command may well have already
+     * pulled a value from the hash enumeration, so we have to check the cache
+     * there first.
+     */
+
+    while (1) {
+	Tcl_HashEntry *hPtr = searchPtr->nextEntry;
+
+	if (hPtr == NULL) {
+	    hPtr = Tcl_NextHashEntry(&searchPtr->search);
+	    if (hPtr == NULL) {
+		return TCL_OK;
+	    }
+	} else {
+	    searchPtr->nextEntry = NULL;
+	}
+	varPtr = VarHashGetValue(hPtr);
+	if (!TclIsVarUndefined(varPtr)) {
+	    Tcl_SetObjResult(interp, VarHashGetKey(varPtr));
+	    return TCL_OK;
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayDoneSearchCmd --
+ *
+ *	This object-based function is invoked to process the "array
+ *	donesearch" Tcl command. See the user documentation for details on
+ *	what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayDoneSearchCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr;
+    Tcl_HashEntry *hPtr;
+    Tcl_Obj *varNameObj, *searchObj;
+    ArraySearch *searchPtr, *prevPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName searchId");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+    searchObj = objv[2];
+
+    varPtr = VerifyArray(interp, varNameObj);
+    if (varPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get the search.
+     */
+
+    searchPtr = ParseSearchId(interp, varPtr, varNameObj, searchObj);
+    if (searchPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Unhook the search from the list of searches associated with the
+     * variable.
+     */
+
+    hPtr = Tcl_FindHashEntry(&iPtr->varSearches, varPtr);
+    if (searchPtr == Tcl_GetHashValue(hPtr)) {
+	if (searchPtr->nextPtr) {
+	    Tcl_SetHashValue(hPtr, searchPtr->nextPtr);
+	} else {
+	    varPtr->flags &= ~VAR_SEARCH_ACTIVE;
+	    Tcl_DeleteHashEntry(hPtr);
+	}
+    } else {
+	for (prevPtr=Tcl_GetHashValue(hPtr) ;; prevPtr=prevPtr->nextPtr) {
+	    if (prevPtr->nextPtr == searchPtr) {
+		prevPtr->nextPtr = searchPtr->nextPtr;
+		break;
+	    }
+	}
+    }
+    Tcl_DecrRefCount(searchPtr->name);
+    ckfree(searchPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayExistsCmd --
+ *
+ *	This object-based function is invoked to process the "array exists"
+ *	Tcl command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayExistsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+    Tcl_Obj *arrayNameObj;
+    int notArray;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName");
+	return TCL_ERROR;
+    }
+    arrayNameObj = objv[1];
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, arrayNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, arrayNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Check whether we've actually got an array variable.
+     */
+
+    notArray = ((varPtr == NULL) || !TclIsVarArray(varPtr)
+	    || TclIsVarUndefined(varPtr));
+    Tcl_SetObjResult(interp, iPtr->execEnvPtr->constants[!notArray]);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayGetCmd --
+ *
+ *	This object-based function is invoked to process the "array get" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayGetCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr, *varPtr2;
+    Tcl_Obj *varNameObj, *nameObj, *valueObj, *nameLstObj, *tmpResObj;
+    Tcl_Obj **nameObjPtr, *patternObj;
+    Tcl_HashSearch search;
+    const char *pattern;
+    int i, count, result;
+
+    switch (objc) {
+    case 2:
+	varNameObj = objv[1];
+	patternObj = NULL;
+	break;
+    case 3:
+	varNameObj = objv[1];
+	patternObj = objv[2];
+	break;
+    default:
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName ?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces. If not an array, it's an empty result.
+     */
+
+    if ((varPtr == NULL) || !TclIsVarArray(varPtr)
+	    || TclIsVarUndefined(varPtr)) {
+	return TCL_OK;
+    }
+
+    pattern = (patternObj ? TclGetString(patternObj) : NULL);
+
+    /*
+     * Store the array names in a new object.
+     */
+
+    TclNewObj(nameLstObj);
+    Tcl_IncrRefCount(nameLstObj);
+    if ((patternObj != NULL) && TclMatchIsTrivial(pattern)) {
+	varPtr2 = VarHashFindVar(varPtr->value.tablePtr, patternObj);
+	if (varPtr2 == NULL) {
+	    goto searchDone;
+	}
+	if (TclIsVarUndefined(varPtr2)) {
+	    goto searchDone;
+	}
+	result = Tcl_ListObjAppendElement(interp, nameLstObj,
+		VarHashGetKey(varPtr2));
+	if (result != TCL_OK) {
+	    TclDecrRefCount(nameLstObj);
+	    return result;
+	}
+	goto searchDone;
+    }
+
+    for (varPtr2 = VarHashFirstVar(varPtr->value.tablePtr, &search);
+	    varPtr2; varPtr2 = VarHashNextVar(&search)) {
+	if (TclIsVarUndefined(varPtr2)) {
+	    continue;
+	}
+	nameObj = VarHashGetKey(varPtr2);
+	if (patternObj && !Tcl_StringMatch(TclGetString(nameObj), pattern)) {
+	    continue;		/* Element name doesn't match pattern. */
+	}
+
+	result = Tcl_ListObjAppendElement(interp, nameLstObj, nameObj);
+	if (result != TCL_OK) {
+	    TclDecrRefCount(nameLstObj);
+	    return result;
+	}
+    }
+
+    /*
+     * Make sure the Var structure of the array is not removed by a trace
+     * while we're working.
+     */
+
+  searchDone:
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)++;
+    }
+
+    /*
+     * Get the array values corresponding to each element name.
+     */
+
+    TclNewObj(tmpResObj);
+    result = Tcl_ListObjGetElements(interp, nameLstObj, &count, &nameObjPtr);
+    if (result != TCL_OK) {
+	goto errorInArrayGet;
+    }
+
+    for (i=0 ; i<count ; i++) {
+	nameObj = *nameObjPtr++;
+	valueObj = Tcl_ObjGetVar2(interp, varNameObj, nameObj,
+		TCL_LEAVE_ERR_MSG);
+	if (valueObj == NULL) {
+	    /*
+	     * Some trace played a trick on us; we need to diagnose to adapt
+	     * our behaviour: was the array element unset, or did the
+	     * modification modify the complete array?
+	     */
+
+	    if (TclIsVarArray(varPtr)) {
+		/*
+		 * The array itself looks OK, the variable was undefined:
+		 * forget it.
+		 */
+
+		continue;
+	    }
+	    result = TCL_ERROR;
+	    goto errorInArrayGet;
+	}
+	result = Tcl_DictObjPut(interp, tmpResObj, nameObj, valueObj);
+	if (result != TCL_OK) {
+	    goto errorInArrayGet;
+	}
+    }
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)--;
+    }
+    Tcl_SetObjResult(interp, tmpResObj);
+    TclDecrRefCount(nameLstObj);
+    return TCL_OK;
+
+  errorInArrayGet:
+    if (TclIsVarInHash(varPtr)) {
+	VarHashRefCount(varPtr)--;
+    }
+    TclDecrRefCount(nameLstObj);
+    TclDecrRefCount(tmpResObj);	/* Free unneeded temp result. */
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayNamesCmd --
+ *
+ *	This object-based function is invoked to process the "array names" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayNamesCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const char *const options[] = {
+	"-exact", "-glob", "-regexp", NULL
+    };
+    enum options { OPT_EXACT, OPT_GLOB, OPT_REGEXP };
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr, *varPtr2;
+    Tcl_Obj *varNameObj, *nameObj, *resultObj, *patternObj;
+    Tcl_HashSearch search;
+    const char *pattern = NULL;
+    int mode = OPT_GLOB;
+
+    if ((objc < 2) || (objc > 4)) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName ?mode? ?pattern?");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+    patternObj = (objc > 2 ? objv[objc-1] : NULL);
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Finish parsing the arguments.
+     */
+
+    if ((objc == 4) && Tcl_GetIndexFromObj(interp, objv[2], options, "option",
+	    0, &mode) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces. If not an array, the result is empty.
+     */
+
+    if ((varPtr == NULL) || !TclIsVarArray(varPtr)
+	    || TclIsVarUndefined(varPtr)) {
+	return TCL_OK;
+    }
+
+    /*
+     * Check for the trivial cases where we can use a direct lookup.
+     */
+
+    TclNewObj(resultObj);
+    if (patternObj) {
+	pattern = TclGetString(patternObj);
+    }
+    if ((mode==OPT_GLOB && patternObj && TclMatchIsTrivial(pattern))
+	    || (mode==OPT_EXACT)) {
+	varPtr2 = VarHashFindVar(varPtr->value.tablePtr, patternObj);
+	if ((varPtr2 != NULL) && !TclIsVarUndefined(varPtr2)) {
+	    /*
+	     * This can't fail; lappending to an empty object always works.
+	     */
+
+	    Tcl_ListObjAppendElement(NULL, resultObj, VarHashGetKey(varPtr2));
+	}
+	Tcl_SetObjResult(interp, resultObj);
+	return TCL_OK;
+    }
+
+    /*
+     * Must scan the array to select the elements.
+     */
+
+    for (varPtr2=VarHashFirstVar(varPtr->value.tablePtr, &search);
+	    varPtr2!=NULL ; varPtr2=VarHashNextVar(&search)) {
+	if (TclIsVarUndefined(varPtr2)) {
+	    continue;
+	}
+	nameObj = VarHashGetKey(varPtr2);
+	if (patternObj) {
+	    const char *name = TclGetString(nameObj);
+	    int matched = 0;
+
+	    switch ((enum options) mode) {
+	    case OPT_EXACT:
+		Tcl_Panic("exact matching shouldn't get here");
+	    case OPT_GLOB:
+		matched = Tcl_StringMatch(name, pattern);
+		break;
+	    case OPT_REGEXP:
+		matched = Tcl_RegExpMatchObj(interp, nameObj, patternObj);
+		if (matched < 0) {
+		    TclDecrRefCount(resultObj);
+		    return TCL_ERROR;
+		}
+		break;
+	    }
+	    if (matched == 0) {
+		continue;
+	    }
+	}
+
+	Tcl_ListObjAppendElement(NULL, resultObj, nameObj);
+    }
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclFindArrayPtrElements --
+ *
+ *	Fill out a hash table (which *must* use Tcl_Obj* keys) with an entry
+ *	for each existing element of the given array. The provided hash table
+ *	is assumed to be initially empty.
+ *
+ * Result:
+ *	none
+ *
+ * Side effects:
+ *	The keys of the array gain an extra reference. The supplied hash table
+ *	has elements added to it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclFindArrayPtrElements(
+    Var *arrayPtr,
+    Tcl_HashTable *tablePtr)
+{
+    Var *varPtr;
+    Tcl_HashSearch search;
+
+    if ((arrayPtr == NULL) || !TclIsVarArray(arrayPtr)
+	    || TclIsVarUndefined(arrayPtr)) {
+	return;
+    }
+
+    for (varPtr=VarHashFirstVar(arrayPtr->value.tablePtr, &search);
+	    varPtr!=NULL ; varPtr=VarHashNextVar(&search)) {
+	Tcl_HashEntry *hPtr;
+	Tcl_Obj *nameObj;
+	int dummy;
+
+	if (TclIsVarUndefined(varPtr)) {
+	    continue;
+	}
+	nameObj = VarHashGetKey(varPtr);
+	hPtr = Tcl_CreateHashEntry(tablePtr, (char *) nameObj, &dummy);
+	Tcl_SetHashValue(hPtr, nameObj);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArraySetCmd --
+ *
+ *	This object-based function is invoked to process the "array set" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArraySetCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+
+    if (objc != 3) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName list");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, objv[1], NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, objv[1], NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    return TclArraySet(interp, objv[1], objv[2]);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArraySizeCmd --
+ *
+ *	This object-based function is invoked to process the "array size" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArraySizeCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+    Tcl_Obj *varNameObj;
+    Tcl_HashSearch search;
+    Var *varPtr2;
+    int size = 0;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces. We can only iterate over the array if it exists...
+     */
+
+    if (varPtr && TclIsVarArray(varPtr) && !TclIsVarUndefined(varPtr)) {
+	/*
+	 * Must iterate in order to get chance to check for present but
+	 * "undefined" entries.
+	 */
+
+	for (varPtr2=VarHashFirstVar(varPtr->value.tablePtr, &search);
+		varPtr2!=NULL ; varPtr2=VarHashNextVar(&search)) {
+	    if (!TclIsVarUndefined(varPtr2)) {
+		size++;
+	    }
+	}
+    }
+
+    Tcl_SetObjResult(interp, Tcl_NewIntObj(size));
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayStatsCmd --
+ *
+ *	This object-based function is invoked to process the "array
+ *	statistics" Tcl command. See the user documentation for details on
+ *	what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayStatsCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr;
+    Tcl_Obj *varNameObj;
+    char *stats;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName");
+	return TCL_ERROR;
+    }
+    varNameObj = objv[1];
+
+    /*
+     * Locate the array variable.
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces.
+     */
+
+    if ((varPtr == NULL) || !TclIsVarArray(varPtr)
+	    || TclIsVarUndefined(varPtr)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"\"%s\" isn't an array", TclGetString(varNameObj)));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ARRAY",
+		TclGetString(varNameObj), NULL);
+	return TCL_ERROR;
+    }
+
+    stats = Tcl_HashStats((Tcl_HashTable *) varPtr->value.tablePtr);
+    if (stats == NULL) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"error reading array statistics", -1));
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(stats, -1));
+    ckfree(stats);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ArrayUnsetCmd --
+ *
+ *	This object-based function is invoked to process the "array unset" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl result object.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+static int
+ArrayUnsetCmd(
+    ClientData clientData,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr, *arrayPtr, *varPtr2, *protectedVarPtr;
+    Tcl_Obj *varNameObj, *patternObj, *nameObj;
+    Tcl_HashSearch search;
+    const char *pattern;
+    const int unsetFlags = 0;	/* Should this be TCL_LEAVE_ERR_MSG? */
+
+    switch (objc) {
+    case 2:
+	varNameObj = objv[1];
+	patternObj = NULL;
+	break;
+    case 3:
+	varNameObj = objv[1];
+	patternObj = objv[2];
+	break;
+    default:
+	Tcl_WrongNumArgs(interp, 1, objv, "arrayName ?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Locate the array variable
+     */
+
+    varPtr = TclObjLookupVarEx(interp, varNameObj, NULL, /*flags*/ 0,
+	    /*msg*/ 0, /*createPart1*/ 0, /*createPart2*/ 0, &arrayPtr);
+
+    /*
+     * Special array trace used to keep the env array in sync for array names,
+     * array get, etc.
+     */
+
+    if (varPtr && (varPtr->flags & VAR_TRACED_ARRAY)
+	    && (TclIsVarArray(varPtr) || TclIsVarUndefined(varPtr))) {
+	if (TclObjCallVarTraces(iPtr, arrayPtr, varPtr, varNameObj, NULL,
+		(TCL_LEAVE_ERR_MSG|TCL_NAMESPACE_ONLY|TCL_GLOBAL_ONLY|
+		TCL_TRACE_ARRAY), /* leaveErrMsg */ 1, -1) == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+    }
+
+    /*
+     * Verify that it is indeed an array variable. This test comes after the
+     * traces - the variable may actually become an array as an effect of said
+     * traces.
+     */
+
+    if ((varPtr == NULL) || !TclIsVarArray(varPtr)
+	    || TclIsVarUndefined(varPtr)) {
+	return TCL_OK;
+    }
+
+    if (!patternObj) {
+	/*
+	 * When no pattern is given, just unset the whole array.
+	 */
+
+	return TclObjUnsetVar2(interp, varNameObj, NULL, 0);
+    }
+
+    /*
+     * With a trivial pattern, we can just unset.
+     */
+
+    pattern = TclGetString(patternObj);
+    if (TclMatchIsTrivial(pattern)) {
+	varPtr2 = VarHashFindVar(varPtr->value.tablePtr, patternObj);
+	if (!varPtr2 || TclIsVarUndefined(varPtr2)) {
+	    return TCL_OK;
+	}
+	return TclPtrUnsetVarIdx(interp, varPtr2, varPtr, varNameObj,
+		patternObj, unsetFlags, -1);
+    }
+
+    /*
+     * Non-trivial case (well, deeply tricky really). We peek inside the hash
+     * iterator in order to allow us to guarantee that the following element
+     * in the array will not be scrubbed until we have dealt with it. This
+     * stops the overall iterator from ending up pointing into deallocated
+     * memory. [Bug 2939073]
+     */
+
+    protectedVarPtr = NULL;
+    for (varPtr2=VarHashFirstVar(varPtr->value.tablePtr, &search);
+	    varPtr2!=NULL ; varPtr2=VarHashNextVar(&search)) {
+	/*
+	 * Drop the extra ref immediately. We don't need to free it at this
+	 * point though; we'll be unsetting it if necessary soon.
+	 */
+
+	if (varPtr2 == protectedVarPtr) {
+	    VarHashRefCount(varPtr2)--;
+	}
+
+	/*
+	 * Guard the next (peeked) item in the search chain by incrementing
+	 * its refcount. This guarantees that the hash table iterator won't be
+	 * dangling on the next time through the loop.
+	 */
+
+	if (search.nextEntryPtr != NULL) {
+	    protectedVarPtr = VarHashGetValue(search.nextEntryPtr);
+	    VarHashRefCount(protectedVarPtr)++;
+	} else {
+	    protectedVarPtr = NULL;
+	}
+
+	/*
+	 * If the variable is undefined, clean it out as it has been hit by
+	 * something else (i.e., an unset trace).
+	 */
+
+	if (TclIsVarUndefined(varPtr2)) {
+	    CleanupVar(varPtr2, varPtr);
+	    continue;
+	}
+
+	nameObj = VarHashGetKey(varPtr2);
+	if (Tcl_StringMatch(TclGetString(nameObj), pattern)
+		&& TclPtrUnsetVarIdx(interp, varPtr2, varPtr, varNameObj,
+			nameObj, unsetFlags, -1) != TCL_OK) {
+	    /*
+	     * If we incremented a refcount, we must decrement it here as we
+	     * will not be coming back properly due to the error.
+	     */
+
+	    if (protectedVarPtr) {
+		VarHashRefCount(protectedVarPtr)--;
+		CleanupVar(protectedVarPtr, varPtr);
+	    }
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInitArrayCmd --
+ *
+ *	This creates the ensemble for the "array" command.
+ *
+ * Results:
+ *	The handle for the created ensemble.
+ *
+ * Side effects:
+ *	Creates a command in the global namespace.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+Tcl_Command
+TclInitArrayCmd(
+    Tcl_Interp *interp)		/* Current interpreter. */
+{
+    static const EnsembleImplMap arrayImplMap[] = {
+	{"anymore",	ArrayAnyMoreCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"donesearch",	ArrayDoneSearchCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"exists",	ArrayExistsCmd,		TclCompileArrayExistsCmd, NULL, NULL, 0},
+	{"get",		ArrayGetCmd,		TclCompileBasic1Or2ArgCmd, NULL, NULL, 0},
+	{"names",	ArrayNamesCmd,		TclCompileBasic1To3ArgCmd, NULL, NULL, 0},
+	{"nextelement",	ArrayNextElementCmd,	TclCompileBasic2ArgCmd, NULL, NULL, 0},
+	{"set",		ArraySetCmd,		TclCompileArraySetCmd, NULL, NULL, 0},
+	{"size",	ArraySizeCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"startsearch",	ArrayStartSearchCmd,	TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"statistics",	ArrayStatsCmd,		TclCompileBasic1ArgCmd, NULL, NULL, 0},
+	{"unset",	ArrayUnsetCmd,		TclCompileArrayUnsetCmd, NULL, NULL, 0},
+	{NULL, NULL, NULL, NULL, NULL, 0}
+    };
+
+    return TclMakeEnsemble(interp, "array", arrayImplMap);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ObjMakeUpvar --
+ *
+ *	This function does all of the work of the "global" and "upvar"
+ *	commands.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs then an error
+ *	message is left in interp.
+ *
+ * Side effects:
+ *	The variable given by myName is linked to the variable in framePtr
+ *	given by otherP1 and otherP2, so that references to myName are
+ *	redirected to the other variable like a symbolic link.
+ *	Callers must Incr myNamePtr if they plan to Decr it.
+ *	Callers must Incr otherP1Ptr if they plan to Decr it.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ObjMakeUpvar(
+    Tcl_Interp *interp,		/* Interpreter containing variables. Used for
+				 * error messages, too. */
+    CallFrame *framePtr,	/* Call frame containing "other" variable.
+				 * NULL means use global :: context. */
+    Tcl_Obj *otherP1Ptr,
+    const char *otherP2,	/* Two-part name of variable in framePtr. */
+    const int otherFlags,	/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of "other" variable. */
+    Tcl_Obj *myNamePtr,		/* Name of variable which will refer to
+				 * otherP1/otherP2. Must be a scalar. */
+    int myFlags,		/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of myName. */
+    int index)			/* If the variable to be linked is an indexed
+				 * scalar, this is its index. Otherwise, -1 */
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *otherPtr, *arrayPtr;
+    CallFrame *varFramePtr;
+
+    /*
+     * Find "other" in "framePtr". If not looking up other in just the current
+     * namespace, temporarily replace the current var frame pointer in the
+     * interpreter in order to use TclObjLookupVar.
+     */
+
+    if (framePtr == NULL) {
+	framePtr = iPtr->rootFramePtr;
+    }
+
+    varFramePtr = iPtr->varFramePtr;
+    if (!(otherFlags & TCL_NAMESPACE_ONLY)) {
+	iPtr->varFramePtr = framePtr;
+    }
+    otherPtr = TclObjLookupVar(interp, otherP1Ptr, otherP2,
+	    (otherFlags | TCL_LEAVE_ERR_MSG), "access",
+	    /*createPart1*/ 1, /*createPart2*/ 1, &arrayPtr);
+    if (!(otherFlags & TCL_NAMESPACE_ONLY)) {
+	iPtr->varFramePtr = varFramePtr;
+    }
+    if (otherPtr == NULL) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Check that we are not trying to create a namespace var linked to a
+     * local variable in a procedure. If we allowed this, the local
+     * variable in the shorter-lived procedure frame could go away leaving
+     * the namespace var's reference invalid.
+     */
+
+    if (index < 0) {
+	if (!(arrayPtr != NULL
+		     ? (TclIsVarInHash(arrayPtr) && TclGetVarNsPtr(arrayPtr))
+		     : (TclIsVarInHash(otherPtr) && TclGetVarNsPtr(otherPtr)))
+		&& ((myFlags & (TCL_GLOBAL_ONLY | TCL_NAMESPACE_ONLY))
+			|| (varFramePtr == NULL)
+			|| !HasLocalVars(varFramePtr)
+			|| (strstr(TclGetString(myNamePtr), "::") != NULL))) {
+	    Tcl_SetObjResult((Tcl_Interp *) iPtr, Tcl_ObjPrintf(
+		    "bad variable name \"%s\": can't create namespace "
+		    "variable that refers to procedure variable",
+		    TclGetString(myNamePtr)));
+	    Tcl_SetErrorCode(interp, "TCL", "UPVAR", "INVERTED", NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    return TclPtrObjMakeUpvarIdx(interp, otherPtr, myNamePtr, myFlags, index);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclPtrMakeUpvar --
+ *
+ *	This procedure does all of the work of the "global" and "upvar"
+ *	commands.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs then an error
+ *	message is left in interp.
+ *
+ * Side effects:
+ *	The variable given by myName is linked to the variable in framePtr
+ *	given by otherP1 and otherP2, so that references to myName are
+ *	redirected to the other variable like a symbolic link.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclPtrMakeUpvar(
+    Tcl_Interp *interp,		/* Interpreter containing variables. Used for
+				 * error messages, too. */
+    Var *otherPtr,		/* Pointer to the variable being linked-to. */
+    const char *myName,		/* Name of variable which will refer to
+				 * otherP1/otherP2. Must be a scalar. */
+    int myFlags,		/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of myName. */
+    int index)			/* If the variable to be linked is an indexed
+				 * scalar, this is its index. Otherwise, -1 */
+{
+    Tcl_Obj *myNamePtr = NULL;
+    int result;
+
+    if (myName) {
+	myNamePtr = Tcl_NewStringObj(myName, -1);
+	Tcl_IncrRefCount(myNamePtr);
+    }
+    result = TclPtrObjMakeUpvarIdx(interp, otherPtr, myNamePtr, myFlags,
+	    index);
+    if (myNamePtr) {
+	Tcl_DecrRefCount(myNamePtr);
+    }
+    return result;
+}
+
+int
+TclPtrObjMakeUpvar(
+    Tcl_Interp *interp,		/* Interpreter containing variables. Used for
+				 * error messages, too. */
+    Tcl_Var otherPtr,		/* Pointer to the variable being linked-to. */
+    Tcl_Obj *myNamePtr,		/* Name of variable which will refer to
+				 * otherP1/otherP2. Must be a scalar. */
+    int myFlags)		/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of myName. */
+{
+    return TclPtrObjMakeUpvarIdx(interp, (Var *) otherPtr, myNamePtr, myFlags,
+	    -1);
+}
+
+/* Callers must Incr myNamePtr if they plan to Decr it. */
+
+int
+TclPtrObjMakeUpvarIdx(
+    Tcl_Interp *interp,		/* Interpreter containing variables. Used for
+				 * error messages, too. */
+    Var *otherPtr,		/* Pointer to the variable being linked-to. */
+    Tcl_Obj *myNamePtr,		/* Name of variable which will refer to
+				 * otherP1/otherP2. Must be a scalar. */
+    int myFlags,		/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of myName. */
+    int index)			/* If the variable to be linked is an indexed
+				 * scalar, this is its index. Otherwise, -1 */
+{
+    Interp *iPtr = (Interp *) interp;
+    CallFrame *varFramePtr = iPtr->varFramePtr;
+    const char *errMsg, *p, *myName;
+    Var *varPtr;
+
+    if (index >= 0) {
+	if (!HasLocalVars(varFramePtr)) {
+	    Tcl_Panic("ObjMakeUpvar called with an index outside from a proc");
+	}
+	varPtr = (Var *) &(varFramePtr->compiledLocals[index]);
+	myNamePtr = localName(iPtr->varFramePtr, index);
+	myName = myNamePtr? TclGetString(myNamePtr) : NULL;
+    } else {
+	/*
+	 * Do not permit the new variable to look like an array reference, as
+	 * it will not be reachable in that case [Bug 600812, TIP 184]. The
+	 * "definition" of what "looks like an array reference" is consistent
+	 * (and must remain consistent) with the code in TclObjLookupVar().
+	 */
+
+	myName = TclGetString(myNamePtr);
+	p = strstr(myName, "(");
+	if (p != NULL) {
+	    p += strlen(p)-1;
+	    if (*p == ')') {
+		/*
+		 * myName looks like an array reference.
+		 */
+
+		Tcl_SetObjResult((Tcl_Interp *) iPtr, Tcl_ObjPrintf(
+			"bad variable name \"%s\": can't create a scalar "
+			"variable that looks like an array element", myName));
+		Tcl_SetErrorCode(interp, "TCL", "UPVAR", "LOCAL_ELEMENT",
+			NULL);
+		return TCL_ERROR;
+	    }
+	}
+
+	/*
+	 * Lookup and eventually create the new variable. Set the flag bit
+	 * TCL_AVOID_RESOLVERS to indicate the special resolution rules for
+	 * upvar purposes:
+	 *   - Bug #696893 - variable is either proc-local or in the current
+	 *     namespace; never follow the second (global) resolution path.
+	 *   - Bug #631741 - do not use special namespace or interp resolvers.
+	 */
+
+	varPtr = TclLookupSimpleVar(interp, myNamePtr,
+		myFlags|TCL_AVOID_RESOLVERS, /* create */ 1, &errMsg, &index);
+	if (varPtr == NULL) {
+	    TclObjVarErrMsg(interp, myNamePtr, NULL, "create", errMsg, -1);
+	    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARNAME",
+		    TclGetString(myNamePtr), NULL);
+	    return TCL_ERROR;
+	}
+    }
+
+    if (varPtr == otherPtr) {
+	Tcl_SetObjResult((Tcl_Interp *) iPtr, Tcl_NewStringObj(
+		"can't upvar from variable to itself", -1));
+	Tcl_SetErrorCode(interp, "TCL", "UPVAR", "SELF", NULL);
+	return TCL_ERROR;
+    }
+
+    if (TclIsVarTraced(varPtr)) {
+	Tcl_SetObjResult((Tcl_Interp *) iPtr, Tcl_ObjPrintf(
+		"variable \"%s\" has traces: can't use for upvar", myName));
+	Tcl_SetErrorCode(interp, "TCL", "UPVAR", "TRACED", NULL);
+	return TCL_ERROR;
+    } else if (!TclIsVarUndefined(varPtr)) {
+	Var *linkPtr;
+
+	/*
+	 * The variable already existed. Make sure this variable "varPtr"
+	 * isn't the same as "otherPtr" (avoid circular links). Also, if it's
+	 * not an upvar then it's an error. If it is an upvar, then just
+	 * disconnect it from the thing it currently refers to.
+	 */
+
+	if (!TclIsVarLink(varPtr)) {
+	    Tcl_SetObjResult((Tcl_Interp *) iPtr, Tcl_ObjPrintf(
+		    "variable \"%s\" already exists", myName));
+	    Tcl_SetErrorCode(interp, "TCL", "UPVAR", "EXISTS", NULL);
+	    return TCL_ERROR;
+	}
+
+	linkPtr = varPtr->value.linkPtr;
+	if (linkPtr == otherPtr) {
+	    return TCL_OK;
+	}
+	if (TclIsVarInHash(linkPtr)) {
+	    VarHashRefCount(linkPtr)--;
+	    if (TclIsVarUndefined(linkPtr)) {
+		CleanupVar(linkPtr, NULL);
+	    }
+	}
+    }
+    TclSetVarLink(varPtr);
+    varPtr->value.linkPtr = otherPtr;
+    if (TclIsVarInHash(otherPtr)) {
+	VarHashRefCount(otherPtr)++;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UpVar --
+ *
+ *	This function links one variable to another, just like the "upvar"
+ *	command.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs then an error
+ *	message is left in the interp's result.
+ *
+ * Side effects:
+ *	The variable in frameName whose name is given by varName becomes
+ *	accessible under the name localNameStr, so that references to
+ *	localNameStr are redirected to the other variable like a symbolic
+ *	link.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#ifndef TCL_NO_DEPRECATED
+#undef Tcl_UpVar
+int
+Tcl_UpVar(
+    Tcl_Interp *interp,		/* Command interpreter in which varName is to
+				 * be looked up. */
+    const char *frameName,	/* Name of the frame containing the source
+				 * variable, such as "1" or "#0". */
+    const char *varName,	/* Name of a variable in interp to link to.
+				 * May be either a scalar name or an element
+				 * in an array. */
+    const char *localNameStr,	/* Name of link variable. */
+    int flags)			/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of localNameStr. */
+{
+    int result;
+    CallFrame *framePtr;
+    Tcl_Obj *varNamePtr, *localNamePtr;
+
+    if (TclGetFrame(interp, frameName, &framePtr) == -1) {
+	return TCL_ERROR;
+    }
+
+    varNamePtr = Tcl_NewStringObj(varName, -1);
+    Tcl_IncrRefCount(varNamePtr);
+    localNamePtr = Tcl_NewStringObj(localNameStr, -1);
+    Tcl_IncrRefCount(localNamePtr);
+
+    result = ObjMakeUpvar(interp, framePtr, varNamePtr, NULL, 0,
+	    localNamePtr, flags, -1);
+    Tcl_DecrRefCount(varNamePtr);
+    Tcl_DecrRefCount(localNamePtr);
+    return result;
+}
+#endif /* TCL_NO_DEPRECATED */
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UpVar2 --
+ *
+ *	This function links one variable to another, just like the "upvar"
+ *	command.
+ *
+ * Results:
+ *	A standard Tcl completion code. If an error occurs then an error
+ *	message is left in the interp's result.
+ *
+ * Side effects:
+ *	The variable in frameName whose name is given by part1 and part2
+ *	becomes accessible under the name localNameStr, so that references to
+ *	localNameStr are redirected to the other variable like a symbolic
+ *	link.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_UpVar2(
+    Tcl_Interp *interp,		/* Interpreter containing variables. Used for
+				 * error messages too. */
+    const char *frameName,	/* Name of the frame containing the source
+				 * variable, such as "1" or "#0". */
+    const char *part1,
+    const char *part2,		/* Two parts of source variable name to link
+				 * to. */
+    const char *localNameStr,	/* Name of link variable. */
+    int flags)			/* 0, TCL_GLOBAL_ONLY or TCL_NAMESPACE_ONLY:
+				 * indicates scope of localNameStr. */
+{
+    int result;
+    CallFrame *framePtr;
+    Tcl_Obj *part1Ptr, *localNamePtr;
+
+    if (TclGetFrame(interp, frameName, &framePtr) == -1) {
+	return TCL_ERROR;
+    }
+
+    part1Ptr = Tcl_NewStringObj(part1, -1);
+    Tcl_IncrRefCount(part1Ptr);
+    localNamePtr = Tcl_NewStringObj(localNameStr, -1);
+    Tcl_IncrRefCount(localNamePtr);
+
+    result = ObjMakeUpvar(interp, framePtr, part1Ptr, part2, 0,
+	    localNamePtr, flags, -1);
+    Tcl_DecrRefCount(part1Ptr);
+    Tcl_DecrRefCount(localNamePtr);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GetVariableFullName --
+ *
+ *	Given a Tcl_Var token returned by Tcl_FindNamespaceVar, this function
+ *	appends to an object the namespace variable's full name, qualified by
+ *	a sequence of parent namespace names.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The variable's fully-qualified name is appended to the string
+ *	representation of objPtr.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_GetVariableFullName(
+    Tcl_Interp *interp,		/* Interpreter containing the variable. */
+    Tcl_Var variable,		/* Token for the variable returned by a
+				 * previous call to Tcl_FindNamespaceVar. */
+    Tcl_Obj *objPtr)		/* Points to the object onto which the
+				 * variable's full name is appended. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register Var *varPtr = (Var *) variable;
+    Tcl_Obj *namePtr;
+    Namespace *nsPtr;
+
+    if (!varPtr || TclIsVarArrayElement(varPtr)) {
+	return;
+    }
+
+    /*
+     * Add the full name of the containing namespace (if any), followed by the
+     * "::" separator, then the variable name.
+     */
+
+    nsPtr = TclGetVarNsPtr(varPtr);
+    if (nsPtr) {
+	Tcl_AppendToObj(objPtr, nsPtr->fullName, -1);
+	if (nsPtr != iPtr->globalNsPtr) {
+	    Tcl_AppendToObj(objPtr, "::", 2);
+	}
+    }
+    if (TclIsVarInHash(varPtr)) {
+	if (!TclIsVarDeadHash(varPtr)) {
+	    namePtr = VarHashGetKey(varPtr);
+	    Tcl_AppendObjToObj(objPtr, namePtr);
+	}
+    } else if (iPtr->varFramePtr->procPtr) {
+	int index = varPtr - iPtr->varFramePtr->compiledLocals;
+
+	if (index >= 0 && index < iPtr->varFramePtr->numCompiledLocals) {
+	    namePtr = localName(iPtr->varFramePtr, index);
+	    Tcl_AppendObjToObj(objPtr, namePtr);
+	}
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_GlobalObjCmd --
+ *
+ *	This object-based function is invoked to process the "global" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_GlobalObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    register Tcl_Obj *objPtr, *tailPtr;
+    const char *varName;
+    register const char *tail;
+    int result, i;
+
+    /*
+     * If we are not executing inside a Tcl procedure, just return.
+     */
+
+    if (!HasLocalVars(iPtr->varFramePtr)) {
+	return TCL_OK;
+    }
+
+    for (i=1 ; i<objc ; i++) {
+	/*
+	 * Make a local variable linked to its counterpart in the global ::
+	 * namespace.
+	 */
+
+	objPtr = objv[i];
+	varName = TclGetString(objPtr);
+
+	/*
+	 * The variable name might have a scope qualifier, but the name for
+	 * the local "link" variable must be the simple name at the tail.
+	 */
+
+	for (tail=varName ; *tail!='\0' ; tail++) {
+	    /* empty body */
+	}
+	while ((tail > varName) && ((*tail != ':') || (*(tail-1) != ':'))) {
+	    tail--;
+	}
+	if ((*tail == ':') && (tail > varName)) {
+	    tail++;
+	}
+
+	if (tail == varName) {
+	    tailPtr = objPtr;
+	} else {
+	    tailPtr = Tcl_NewStringObj(tail, -1);
+	    Tcl_IncrRefCount(tailPtr);
+	}
+
+	/*
+	 * Link to the variable "varName" in the global :: namespace.
+	 */
+
+	result = ObjMakeUpvar(interp, NULL, objPtr, NULL,
+		TCL_GLOBAL_ONLY, /*myName*/ tailPtr, /*myFlags*/ 0, -1);
+
+	if (tail != varName) {
+	    Tcl_DecrRefCount(tailPtr);
+	}
+
+	if (result != TCL_OK) {
+	    return result;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_VariableObjCmd --
+ *
+ *	Invoked to implement the "variable" command that creates one or more
+ *	global variables. Handles the following syntax:
+ *
+ *	    variable ?name value...? name ?value?
+ *
+ *	One or more variables can be created. The variables are initialized
+ *	with the specified values. The value for the last variable is
+ *	optional.
+ *
+ *	If the variable does not exist, it is created and given the optional
+ *	value. If it already exists, it is simply set to the optional value.
+ *	Normally, "name" is an unqualified name, so it is created in the
+ *	current namespace. If it includes namespace qualifiers, it can be
+ *	created in another namespace.
+ *
+ *	If the variable command is executed inside a Tcl procedure, it creates
+ *	a local variable linked to the newly-created namespace variable.
+ *
+ * Results:
+ *	Returns TCL_OK if the variable is found or created. Returns TCL_ERROR
+ *	if anything goes wrong.
+ *
+ * Side effects:
+ *	If anything goes wrong, this function returns an error message as the
+ *	result in the interpreter's result object.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_VariableObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *varName, *tail, *cp;
+    Var *varPtr, *arrayPtr;
+    Tcl_Obj *varValuePtr;
+    int i, result;
+    Tcl_Obj *varNamePtr, *tailPtr;
+
+    for (i=1 ; i<objc ; i+=2) {
+	/*
+	 * Look up each variable in the current namespace context, creating it
+	 * if necessary.
+	 */
+
+	varNamePtr = objv[i];
+	varName = TclGetString(varNamePtr);
+	varPtr = TclObjLookupVarEx(interp, varNamePtr, NULL,
+		(TCL_NAMESPACE_ONLY | TCL_LEAVE_ERR_MSG), "define",
+		/*createPart1*/ 1, /*createPart2*/ 0, &arrayPtr);
+
+	if (arrayPtr != NULL) {
+	    /*
+	     * Variable cannot be an element in an array. If arrayPtr is
+	     * non-NULL, it is, so throw up an error and return.
+	     */
+
+	    TclObjVarErrMsg(interp, varNamePtr, NULL, "define",
+		    isArrayElement, -1);
+	    Tcl_SetErrorCode(interp, "TCL", "UPVAR", "LOCAL_ELEMENT", NULL);
+	    return TCL_ERROR;
+	}
+
+	if (varPtr == NULL) {
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Mark the variable as a namespace variable and increment its
+	 * reference count so that it will persist until its namespace is
+	 * destroyed or until the variable is unset.
+	 */
+
+	TclSetVarNamespaceVar(varPtr);
+
+	/*
+	 * If a value was specified, set the variable to that value.
+	 * Otherwise, if the variable is new, leave it undefined. (If the
+	 * variable already exists and no value was specified, leave its value
+	 * unchanged; just create the local link if we're in a Tcl procedure).
+	 */
+
+	if (i+1 < objc) {	/* A value was specified. */
+	    varValuePtr = TclPtrSetVarIdx(interp, varPtr, arrayPtr,
+		    varNamePtr, NULL, objv[i+1],
+		    (TCL_NAMESPACE_ONLY | TCL_LEAVE_ERR_MSG), -1);
+	    if (varValuePtr == NULL) {
+		return TCL_ERROR;
+	    }
+	}
+
+	/*
+	 * If we are executing inside a Tcl procedure, create a local variable
+	 * linked to the new namespace variable "varName".
+	 */
+
+	if (HasLocalVars(iPtr->varFramePtr)) {
+	    /*
+	     * varName might have a scope qualifier, but the name for the
+	     * local "link" variable must be the simple name at the tail.
+	     *
+	     * Locate tail in one pass: drop any prefix after two *or more*
+	     * consecutive ":" characters).
+	     */
+
+	    for (tail=cp=varName ; *cp!='\0' ;) {
+		if (*cp++ == ':') {
+		    while (*cp == ':') {
+			tail = ++cp;
+		    }
+		}
+	    }
+
+	    /*
+	     * Create a local link "tail" to the variable "varName" in the
+	     * current namespace.
+	     */
+
+	    if (tail == varName) {
+		tailPtr = varNamePtr;
+	    } else {
+		tailPtr = Tcl_NewStringObj(tail, -1);
+		Tcl_IncrRefCount(tailPtr);
+	    }
+
+	    result = ObjMakeUpvar(interp, NULL, varNamePtr, /*otherP2*/ NULL,
+		    /*otherFlags*/ TCL_NAMESPACE_ONLY,
+		    /*myName*/ tailPtr, /*myFlags*/ 0, -1);
+
+	    if (tail != varName) {
+		Tcl_DecrRefCount(tailPtr);
+	    }
+
+	    if (result != TCL_OK) {
+		return result;
+	    }
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_UpvarObjCmd --
+ *
+ *	This object-based function is invoked to process the "upvar" Tcl
+ *	command. See the user documentation for details on what it does.
+ *
+ * Results:
+ *	A standard Tcl object result value.
+ *
+ * Side effects:
+ *	See the user documentation.
+ *
+ *----------------------------------------------------------------------
+ */
+
+	/* ARGSUSED */
+int
+Tcl_UpvarObjCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    CallFrame *framePtr;
+    int result, hasLevel;
+    Tcl_Obj *levelObj;
+
+    if (objc < 3) {
+	Tcl_WrongNumArgs(interp, 1, objv,
+		"?level? otherVar localVar ?otherVar localVar ...?");
+	return TCL_ERROR;
+    }
+
+    if (objc & 1) {
+	/*
+	 * Even number of arguments, so use the default level of "1" by
+	 * passing NULL to TclObjGetFrame.
+	 */
+
+	levelObj = NULL;
+	hasLevel = 0;
+    } else {
+	/*
+	 * Odd number of arguments, so objv[1] must contain the level.
+	 */
+
+	levelObj = objv[1];
+	hasLevel = 1;
+    }
+
+    /*
+     * Find the call frame containing each of the "other variables" to be
+     * linked to.
+     */
+
+    result = TclObjGetFrame(interp, levelObj, &framePtr);
+    if (result == -1) {
+	return TCL_ERROR;
+    }
+    if ((result == 0) && hasLevel) {
+	/*
+	 * Synthesize an error message since TclObjGetFrame doesn't do this
+	 * for this particular case.
+	 */
+
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"bad level \"%s\"", TclGetString(levelObj)));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "LEVEL",
+		TclGetString(levelObj), NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * We've now finished with parsing levels; skip to the variable names.
+     */
+
+    objc -= hasLevel + 1;
+    objv += hasLevel + 1;
+
+    /*
+     * Iterate over each (other variable, local variable) pair. Divide the
+     * other variable name into two parts, then call MakeUpvar to do all the
+     * work of linking it to the local variable.
+     */
+
+    for (; objc>0 ; objc-=2, objv+=2) {
+	result = ObjMakeUpvar(interp, framePtr, /* othervarName */ objv[0],
+		NULL, 0, /* myVarName */ objv[1], /*flags*/ 0, -1);
+	if (result != TCL_OK) {
+	    return TCL_ERROR;
+	}
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ParseSearchId --
+ *
+ *	This function translates from a tcl object to a pointer to an active
+ *	array search (if there is one that matches the string).
+ *
+ * Results:
+ *	The return value is a pointer to the array search indicated by string,
+ *	or NULL if there isn't one. If NULL is returned, the interp's result
+ *	contains an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static ArraySearch *
+ParseSearchId(
+    Tcl_Interp *interp,		/* Interpreter containing variable. */
+    const Var *varPtr,		/* Array variable search is for. */
+    Tcl_Obj *varNamePtr,	/* Name of array variable that search is
+				 * supposed to be for. */
+    Tcl_Obj *handleObj)		/* Object containing id of search. Must have
+				 * form "search-num-var" where "num" is a
+				 * decimal number and "var" is a variable
+				 * name. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ArraySearch *searchPtr;
+    const char *handle = TclGetString(handleObj);
+    char *end;
+
+    if (varPtr->flags & VAR_SEARCH_ACTIVE) {
+	Tcl_HashEntry *hPtr =
+		Tcl_FindHashEntry(&iPtr->varSearches, varPtr);
+
+	/* First look for same (Tcl_Obj *) */
+	for (searchPtr = Tcl_GetHashValue(hPtr); searchPtr != NULL;
+		searchPtr = searchPtr->nextPtr) {
+	    if (searchPtr->name == handleObj) {
+		return searchPtr;
+	    }
+	}
+	/* Fallback: do string compares. */
+	for (searchPtr = Tcl_GetHashValue(hPtr); searchPtr != NULL;
+		searchPtr = searchPtr->nextPtr) {
+	    if (strcmp(TclGetString(searchPtr->name), handle) == 0) {
+		return searchPtr;
+	    }
+	}
+    }
+    if ((handle[0] != 's') || (handle[1] != '-')
+	    || (strtoul(handle + 2, &end, 10), end == (handle + 2))
+	    || (*end != '-')) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"illegal search identifier \"%s\"", handle));
+    } else if (strcmp(end + 1, TclGetString(varNamePtr)) != 0) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"search identifier \"%s\" isn't for variable \"%s\"",
+		handle, TclGetString(varNamePtr)));
+    } else {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"couldn't find search \"%s\"", handle));
+    }
+    Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "ARRAYSEARCH", handle, NULL);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteSearches --
+ *
+ *	This function is called to free up all of the searches associated
+ *	with an array variable.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Memory is released to the storage allocator.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteSearches(
+    Interp *iPtr,
+    register Var *arrayVarPtr)	/* Variable whose searches are to be
+				 * deleted. */
+{
+    ArraySearch *searchPtr, *nextPtr;
+    Tcl_HashEntry *sPtr;
+
+    if (arrayVarPtr->flags & VAR_SEARCH_ACTIVE) {
+	sPtr = Tcl_FindHashEntry(&iPtr->varSearches, arrayVarPtr);
+	for (searchPtr = Tcl_GetHashValue(sPtr); searchPtr != NULL;
+		searchPtr = nextPtr) {
+	    nextPtr = searchPtr->nextPtr;
+	    Tcl_DecrRefCount(searchPtr->name);
+	    ckfree(searchPtr);
+	}
+	arrayVarPtr->flags &= ~VAR_SEARCH_ACTIVE;
+	Tcl_DeleteHashEntry(sPtr);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteNamespaceVars --
+ *
+ *	This function is called to recycle all the storage space associated
+ *	with a namespace's table of variables.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Variables are deleted and trace functions are invoked, if any are
+ *	declared.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDeleteNamespaceVars(
+    Namespace *nsPtr)
+{
+    TclVarHashTable *tablePtr = &nsPtr->varTable;
+    Tcl_Interp *interp = nsPtr->interp;
+    Interp *iPtr = (Interp *)interp;
+    Tcl_HashSearch search;
+    int flags = 0;
+    Var *varPtr;
+
+    /*
+     * Determine what flags to pass to the trace callback functions.
+     */
+
+    if (nsPtr == iPtr->globalNsPtr) {
+	flags = TCL_GLOBAL_ONLY;
+    } else if (nsPtr == (Namespace *) TclGetCurrentNamespace(interp)) {
+	flags = TCL_NAMESPACE_ONLY;
+    }
+
+    for (varPtr = VarHashFirstVar(tablePtr, &search);  varPtr != NULL;
+	    varPtr = VarHashFirstVar(tablePtr, &search)) {
+	Tcl_Obj *objPtr = Tcl_NewObj();
+	VarHashRefCount(varPtr)++;	/* Make sure we get to remove from
+					 * hash. */
+	Tcl_GetVariableFullName(interp, (Tcl_Var) varPtr, objPtr);
+	UnsetVarStruct(varPtr, NULL, iPtr, /* part1 */ objPtr,
+		NULL, flags, -1);
+
+	/*
+	 * We just unset the variable. However, an unset trace might
+	 * have re-set it, or might have re-established traces on it.
+	 * This namespace and its vartable are going away unconditionally,
+	 * so we cannot let such things linger. That would be a leak.
+	 *
+	 * First we destroy all traces. ...
+	 */
+
+	if (TclIsVarTraced(varPtr)) {
+	    Tcl_HashEntry *tPtr = Tcl_FindHashEntry(&iPtr->varTraces, varPtr);
+	    VarTrace *tracePtr = Tcl_GetHashValue(tPtr);
+	    ActiveVarTrace *activePtr;
+
+	    while (tracePtr) {
+		VarTrace *prevPtr = tracePtr;
+
+		tracePtr = tracePtr->nextPtr;
+		prevPtr->nextPtr = NULL;
+		Tcl_EventuallyFree(prevPtr, TCL_DYNAMIC);
+	    }
+	    Tcl_DeleteHashEntry(tPtr);
+	    varPtr->flags &= ~VAR_ALL_TRACES;
+	    for (activePtr = iPtr->activeVarTracePtr; activePtr != NULL;
+		    activePtr = activePtr->nextPtr) {
+		if (activePtr->varPtr == varPtr) {
+		    activePtr->nextTracePtr = NULL;
+		}
+	    }
+	}
+
+	/*
+	 * ...and then, if the variable still holds a value, we unset it
+	 * again. This time with no traces left, we're sure it goes away.
+	 */
+
+	if (!TclIsVarUndefined(varPtr)) {
+	    UnsetVarStruct(varPtr, NULL, iPtr, /* part1 */ objPtr,
+		    NULL, flags, -1);
+	}
+	Tcl_DecrRefCount(objPtr); /* free no longer needed obj */
+	VarHashRefCount(varPtr)--;
+	VarHashDeleteEntry(varPtr);
+    }
+    VarHashDeleteTable(tablePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteVars --
+ *
+ *	This function is called to recycle all the storage space associated
+ *	with a table of variables. For this function to work correctly, it
+ *	must not be possible for any of the variables in the table to be
+ *	accessed from Tcl commands (e.g. from trace functions).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Variables are deleted and trace functions are invoked, if any are
+ *	declared.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDeleteVars(
+    Interp *iPtr,		/* Interpreter to which variables belong. */
+    TclVarHashTable *tablePtr)	/* Hash table containing variables to
+				 * delete. */
+{
+    Tcl_Interp *interp = (Tcl_Interp *) iPtr;
+    Tcl_HashSearch search;
+    register Var *varPtr;
+    int flags;
+    Namespace *currNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+
+    /*
+     * Determine what flags to pass to the trace callback functions.
+     */
+
+    flags = TCL_TRACE_UNSETS;
+    if (tablePtr == &iPtr->globalNsPtr->varTable) {
+	flags |= TCL_GLOBAL_ONLY;
+    } else if (tablePtr == &currNsPtr->varTable) {
+	flags |= TCL_NAMESPACE_ONLY;
+    }
+
+    for (varPtr = VarHashFirstVar(tablePtr, &search); varPtr != NULL;
+	 varPtr = VarHashFirstVar(tablePtr, &search)) {
+	UnsetVarStruct(varPtr, NULL, iPtr, VarHashGetKey(varPtr), NULL, flags,
+		-1);
+	VarHashDeleteEntry(varPtr);
+    }
+    VarHashDeleteTable(tablePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclDeleteCompiledLocalVars --
+ *
+ *	This function is called to recycle storage space associated with the
+ *	compiler-allocated array of local variables in a procedure call frame.
+ *	This function resembles TclDeleteVars above except that each variable
+ *	is stored in a call frame and not a hash table. For this function to
+ *	work correctly, it must not be possible for any of the variable in the
+ *	table to be accessed from Tcl commands (e.g. from trace functions).
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Variables are deleted and trace functions are invoked, if any are
+ *	declared.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclDeleteCompiledLocalVars(
+    Interp *iPtr,		/* Interpreter to which variables belong. */
+    CallFrame *framePtr)	/* Procedure call frame containing compiler-
+				 * assigned local variables to delete. */
+{
+    register Var *varPtr;
+    int numLocals, i;
+    Tcl_Obj **namePtrPtr;
+
+    numLocals = framePtr->numCompiledLocals;
+    varPtr = framePtr->compiledLocals;
+    namePtrPtr = &localName(framePtr, 0);
+    for (i=0 ; i<numLocals ; i++, namePtrPtr++, varPtr++) {
+	UnsetVarStruct(varPtr, NULL, iPtr, *namePtrPtr, NULL,
+		TCL_TRACE_UNSETS, i);
+    }
+    framePtr->numCompiledLocals = 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * DeleteArray --
+ *
+ *	This function is called to free up everything in an array variable.
+ *	It's the caller's responsibility to make sure that the array is no
+ *	longer accessible before this function is called.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	All storage associated with varPtr's array elements is deleted
+ *	(including the array's hash table). Deletion trace functions for
+ *	array elements are invoked, then deleted. Any pending traces for array
+ *	elements are also deleted.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+DeleteArray(
+    Interp *iPtr,		/* Interpreter containing array. */
+    Tcl_Obj *arrayNamePtr,	/* Name of array (used for trace callbacks),
+				 * or NULL if it is to be computed on
+				 * demand. */
+    Var *varPtr,		/* Pointer to variable structure. */
+    int flags,			/* Flags to pass to TclCallVarTraces:
+				 * TCL_TRACE_UNSETS and sometimes
+				 * TCL_NAMESPACE_ONLY or TCL_GLOBAL_ONLY. */
+    int index)
+{
+    Tcl_HashSearch search;
+    Tcl_HashEntry *tPtr;
+    register Var *elPtr;
+    ActiveVarTrace *activePtr;
+    Tcl_Obj *objPtr;
+    VarTrace *tracePtr;
+
+    if (varPtr->flags & VAR_SEARCH_ACTIVE) {
+	DeleteSearches(iPtr, varPtr);
+    }
+    for (elPtr = VarHashFirstVar(varPtr->value.tablePtr, &search);
+	    elPtr != NULL; elPtr = VarHashNextVar(&search)) {
+	if (TclIsVarScalar(elPtr) && (elPtr->value.objPtr != NULL)) {
+	    objPtr = elPtr->value.objPtr;
+	    TclDecrRefCount(objPtr);
+	    elPtr->value.objPtr = NULL;
+	}
+
+	/*
+	 * Lie about the validity of the hashtable entry. In this way the
+	 * variables will be deleted by VarHashDeleteTable.
+	 */
+
+	VarHashInvalidateEntry(elPtr);
+	if (TclIsVarTraced(elPtr)) {
+	    /*
+	     * Compute the array name if it was not supplied.
+	     */
+
+	    if (elPtr->flags & VAR_TRACED_UNSET) {
+		Tcl_Obj *elNamePtr = VarHashGetKey(elPtr);
+
+		elPtr->flags &= ~VAR_TRACE_ACTIVE;
+		TclObjCallVarTraces(iPtr, NULL, elPtr, arrayNamePtr,
+			elNamePtr, flags,/* leaveErrMsg */ 0, index);
+	    }
+	    tPtr = Tcl_FindHashEntry(&iPtr->varTraces, elPtr);
+	    tracePtr = Tcl_GetHashValue(tPtr);
+	    while (tracePtr) {
+		VarTrace *prevPtr = tracePtr;
+
+		tracePtr = tracePtr->nextPtr;
+		prevPtr->nextPtr = NULL;
+		Tcl_EventuallyFree(prevPtr, TCL_DYNAMIC);
+	    }
+	    Tcl_DeleteHashEntry(tPtr);
+	    elPtr->flags &= ~VAR_ALL_TRACES;
+	    for (activePtr = iPtr->activeVarTracePtr; activePtr != NULL;
+		    activePtr = activePtr->nextPtr) {
+		if (activePtr->varPtr == elPtr) {
+		    activePtr->nextTracePtr = NULL;
+		}
+	    }
+	}
+	TclSetVarUndefined(elPtr);
+
+	/*
+	 * Even though array elements are not supposed to be namespace
+	 * variables, some combinations of [upvar] and [variable] may create
+	 * such beasts - see [Bug 604239]. This is necessary to avoid leaking
+	 * the corresponding Var struct, and is otherwise harmless.
+	 */
+
+	TclClearVarNamespaceVar(elPtr);
+    }
+    VarHashDeleteTable(varPtr->value.tablePtr);
+    ckfree(varPtr->value.tablePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclObjVarErrMsg --
+ *
+ *	Generate a reasonable error message describing why a variable
+ *	operation failed.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	The interp's result is set to hold a message identifying the variable
+ *	given by part1 and part2 and describing why the variable operation
+ *	failed.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+TclVarErrMsg(
+    Tcl_Interp *interp,		/* Interpreter in which to record message. */
+    const char *part1,
+    const char *part2,		/* Variable's two-part name. */
+    const char *operation,	/* String describing operation that failed,
+				 * e.g. "read", "set", or "unset". */
+    const char *reason)		/* String describing why operation failed. */
+{
+    Tcl_Obj *part2Ptr = NULL, *part1Ptr = Tcl_NewStringObj(part1, -1);
+
+    if (part2) {
+	part2Ptr = Tcl_NewStringObj(part2, -1);
+    }
+
+    TclObjVarErrMsg(interp, part1Ptr, part2Ptr, operation, reason, -1);
+
+    Tcl_DecrRefCount(part1Ptr);
+    if (part2Ptr) {
+	Tcl_DecrRefCount(part2Ptr);
+    }
+}
+
+void
+TclObjVarErrMsg(
+    Tcl_Interp *interp,		/* Interpreter in which to record message. */
+    Tcl_Obj *part1Ptr,		/* (may be NULL, if index >= 0) */
+    Tcl_Obj *part2Ptr,		/* Variable's two-part name. */
+    const char *operation,	/* String describing operation that failed,
+				 * e.g. "read", "set", or "unset". */
+    const char *reason,		/* String describing why operation failed. */
+    int index)			/* Index into the local variable table of the
+				 * variable, or -1. Only used when part1Ptr is
+				 * NULL. */
+{
+    if (!part1Ptr) {
+	if (index == -1) {
+	    Tcl_Panic("invalid part1Ptr and invalid index together");
+	}
+	part1Ptr = localName(((Interp *)interp)->varFramePtr, index);
+    }
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf("can't %s \"%s%s%s%s\": %s",
+	    operation, TclGetString(part1Ptr), (part2Ptr ? "(" : ""),
+	    (part2Ptr ? TclGetString(part2Ptr) : ""), (part2Ptr ? ")" : ""),
+	    reason));
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Internal functions for variable name object types --
+ *
+ *----------------------------------------------------------------------
+ */
+
+/*
+ * localVarName -
+ *
+ * INTERNALREP DEFINITION:
+ *   twoPtrValue.ptr1:   pointer to name obj in varFramePtr->localCache
+ *			  or NULL if it is this same obj
+ *   twoPtrValue.ptr2: index into locals table
+ */
+
+static void
+FreeLocalVarName(
+    Tcl_Obj *objPtr)
+{
+    Tcl_Obj *namePtr = objPtr->internalRep.twoPtrValue.ptr1;
+
+    if (namePtr) {
+	Tcl_DecrRefCount(namePtr);
+    }
+    objPtr->typePtr = NULL;
+}
+
+static void
+DupLocalVarName(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *dupPtr)
+{
+    Tcl_Obj *namePtr = srcPtr->internalRep.twoPtrValue.ptr1;
+
+    if (!namePtr) {
+	namePtr = srcPtr;
+    }
+    dupPtr->internalRep.twoPtrValue.ptr1 = namePtr;
+    Tcl_IncrRefCount(namePtr);
+
+    dupPtr->internalRep.twoPtrValue.ptr2 =
+	    srcPtr->internalRep.twoPtrValue.ptr2;
+    dupPtr->typePtr = &localVarNameType;
+}
+
+/*
+ * parsedVarName -
+ *
+ * INTERNALREP DEFINITION:
+ *   twoPtrValue.ptr1 = pointer to the array name Tcl_Obj (NULL if scalar)
+ *   twoPtrValue.ptr2 = pointer to the element name string (owned by this
+ *			Tcl_Obj), or NULL if it is a scalar variable
+ */
+
+static void
+FreeParsedVarName(
+    Tcl_Obj *objPtr)
+{
+    register Tcl_Obj *arrayPtr = objPtr->internalRep.twoPtrValue.ptr1;
+    register Tcl_Obj *elem = objPtr->internalRep.twoPtrValue.ptr2;
+
+    if (arrayPtr != NULL) {
+	TclDecrRefCount(arrayPtr);
+	TclDecrRefCount(elem);
+    }
+    objPtr->typePtr = NULL;
+}
+
+static void
+DupParsedVarName(
+    Tcl_Obj *srcPtr,
+    Tcl_Obj *dupPtr)
+{
+    register Tcl_Obj *arrayPtr = srcPtr->internalRep.twoPtrValue.ptr1;
+    register Tcl_Obj *elem = srcPtr->internalRep.twoPtrValue.ptr2;
+
+    if (arrayPtr != NULL) {
+	Tcl_IncrRefCount(arrayPtr);
+	Tcl_IncrRefCount(elem);
+    }
+
+    dupPtr->internalRep.twoPtrValue.ptr1 = arrayPtr;
+    dupPtr->internalRep.twoPtrValue.ptr2 = elem;
+    dupPtr->typePtr = &tclParsedVarNameType;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_FindNamespaceVar -- MOVED OVER from tclNamesp.c
+ *
+ *	Searches for a namespace variable, a variable not local to a
+ *	procedure. The variable can be either a scalar or an array, but may
+ *	not be an element of an array.
+ *
+ * Results:
+ *	Returns a token for the variable if it is found. Otherwise, if it
+ *	can't be found or there is an error, returns NULL and leaves an error
+ *	message in the interpreter's result object if "flags" contains
+ *	TCL_LEAVE_ERR_MSG.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Var
+Tcl_FindNamespaceVar(
+    Tcl_Interp *interp,		/* The interpreter in which to find the
+				 * variable. */
+    const char *name,		/* Variable's name. If it starts with "::",
+				 * will be looked up in global namespace.
+				 * Else, looked up first in contextNsPtr
+				 * (current namespace if contextNsPtr is
+				 * NULL), then in global namespace. */
+    Tcl_Namespace *contextNsPtr,/* Ignored if TCL_GLOBAL_ONLY flag set.
+				 * Otherwise, points to namespace in which to
+				 * resolve name. If NULL, look up name in the
+				 * current namespace. */
+    int flags)			/* An OR'd combination of:
+				 * TCL_AVOID_RESOLVERS, TCL_GLOBAL_ONLY (look
+				 * up name only in global namespace),
+				 * TCL_NAMESPACE_ONLY (look up only in
+				 * contextNsPtr, or the current namespace if
+				 * contextNsPtr is NULL), and
+				 * TCL_LEAVE_ERR_MSG. If both TCL_GLOBAL_ONLY
+				 * and TCL_NAMESPACE_ONLY are given,
+				 * TCL_GLOBAL_ONLY is ignored. */
+{
+    Tcl_Obj *namePtr = Tcl_NewStringObj(name, -1);
+    Tcl_Var var;
+
+    var = ObjFindNamespaceVar(interp, namePtr, contextNsPtr, flags);
+    Tcl_DecrRefCount(namePtr);
+    return var;
+}
+
+static Tcl_Var
+ObjFindNamespaceVar(
+    Tcl_Interp *interp,		/* The interpreter in which to find the
+				 * variable. */
+    Tcl_Obj *namePtr,		/* Variable's name. If it starts with "::",
+				 * will be looked up in global namespace.
+				 * Else, looked up first in contextNsPtr
+				 * (current namespace if contextNsPtr is
+				 * NULL), then in global namespace. */
+    Tcl_Namespace *contextNsPtr,/* Ignored if TCL_GLOBAL_ONLY flag set.
+				 * Otherwise, points to namespace in which to
+				 * resolve name. If NULL, look up name in the
+				 * current namespace. */
+    int flags)			/* An OR'd combination of:
+				 * TCL_AVOID_RESOLVERS, TCL_GLOBAL_ONLY (look
+				 * up name only in global namespace),
+				 * TCL_NAMESPACE_ONLY (look up only in
+				 * contextNsPtr, or the current namespace if
+				 * contextNsPtr is NULL), and
+				 * TCL_LEAVE_ERR_MSG. If both TCL_GLOBAL_ONLY
+				 * and TCL_NAMESPACE_ONLY are given,
+				 * TCL_GLOBAL_ONLY is ignored. */
+{
+    Interp *iPtr = (Interp *) interp;
+    ResolverScheme *resPtr;
+    Namespace *nsPtr[2], *cxtNsPtr;
+    const char *simpleName;
+    Var *varPtr;
+    register int search;
+    int result;
+    Tcl_Var var;
+    Tcl_Obj *simpleNamePtr;
+    const char *name = TclGetString(namePtr);
+
+    /*
+     * If this namespace has a variable resolver, then give it first crack at
+     * the variable resolution. It may return a Tcl_Var value, it may signal
+     * to continue onward, or it may signal an error.
+     */
+
+    if ((flags & TCL_GLOBAL_ONLY) != 0) {
+	cxtNsPtr = (Namespace *) TclGetGlobalNamespace(interp);
+    } else if (contextNsPtr != NULL) {
+	cxtNsPtr = (Namespace *) contextNsPtr;
+    } else {
+	cxtNsPtr = (Namespace *) TclGetCurrentNamespace(interp);
+    }
+
+    if (!(flags & TCL_AVOID_RESOLVERS) &&
+	    (cxtNsPtr->varResProc != NULL || iPtr->resolverPtr != NULL)) {
+	resPtr = iPtr->resolverPtr;
+
+	if (cxtNsPtr->varResProc) {
+	    result = cxtNsPtr->varResProc(interp, name,
+		    (Tcl_Namespace *) cxtNsPtr, flags, &var);
+	} else {
+	    result = TCL_CONTINUE;
+	}
+
+	while (result == TCL_CONTINUE && resPtr) {
+	    if (resPtr->varResProc) {
+		result = resPtr->varResProc(interp, name,
+			(Tcl_Namespace *) cxtNsPtr, flags, &var);
+	    }
+	    resPtr = resPtr->nextPtr;
+	}
+
+	if (result == TCL_OK) {
+	    return var;
+	} else if (result != TCL_CONTINUE) {
+	    return NULL;
+	}
+    }
+
+    /*
+     * Find the namespace(s) that contain the variable.
+     */
+
+    TclGetNamespaceForQualName(interp, name, (Namespace *) contextNsPtr,
+	    flags, &nsPtr[0], &nsPtr[1], &cxtNsPtr, &simpleName);
+
+    /*
+     * Look for the variable in the variable table of its namespace. Be sure
+     * to check both possible search paths: from the specified namespace
+     * context and from the global namespace.
+     */
+
+    varPtr = NULL;
+    if (simpleName != name) {
+	simpleNamePtr = Tcl_NewStringObj(simpleName, -1);
+    } else {
+	simpleNamePtr = namePtr;
+    }
+
+    for (search = 0;  (search < 2) && (varPtr == NULL);  search++) {
+	if ((nsPtr[search] != NULL) && (simpleName != NULL)) {
+	    varPtr = VarHashFindVar(&nsPtr[search]->varTable, simpleNamePtr);
+	}
+    }
+    if (simpleName != name) {
+	Tcl_DecrRefCount(simpleNamePtr);
+    }
+    if ((varPtr == NULL) && (flags & TCL_LEAVE_ERR_MSG)) {
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		"unknown variable \"%s\"", name));
+	Tcl_SetErrorCode(interp, "TCL", "LOOKUP", "VARIABLE", name, NULL);
+    }
+    return (Tcl_Var) varPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoVarsCmd -- (moved over from tclCmdIL.c)
+ *
+ *	Called to implement the "info vars" command that returns the list of
+ *	variables in the interpreter that match an optional pattern. The
+ *	pattern, if any, consists of an optional sequence of namespace names
+ *	separated by "::" qualifiers, which is followed by a glob-style
+ *	pattern that restricts which variables are returned. Handles the
+ *	following syntax:
+ *
+ *	    info vars ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInfoVarsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    const char *varName, *pattern, *simplePattern;
+    Tcl_HashSearch search;
+    Var *varPtr;
+    Namespace *nsPtr;
+    Namespace *globalNsPtr = (Namespace *) Tcl_GetGlobalNamespace(interp);
+    Namespace *currNsPtr = (Namespace *) Tcl_GetCurrentNamespace(interp);
+    Tcl_Obj *listPtr, *elemObjPtr, *varNamePtr;
+    int specificNsInPattern = 0;/* Init. to avoid compiler warning. */
+    Tcl_Obj *simplePatternPtr = NULL;
+
+    /*
+     * Get the pattern and find the "effective namespace" in which to list
+     * variables. We only use this effective namespace if there's no active
+     * Tcl procedure frame.
+     */
+
+    if (objc == 1) {
+	simplePattern = NULL;
+	nsPtr = currNsPtr;
+	specificNsInPattern = 0;
+    } else if (objc == 2) {
+	/*
+	 * From the pattern, get the effective namespace and the simple
+	 * pattern (no namespace qualifiers or ::'s) at the end. If an error
+	 * was found while parsing the pattern, return it. Otherwise, if the
+	 * namespace wasn't found, just leave nsPtr NULL: we will return an
+	 * empty list since no variables there can be found.
+	 */
+
+	Namespace *dummy1NsPtr, *dummy2NsPtr;
+
+	pattern = TclGetString(objv[1]);
+	TclGetNamespaceForQualName(interp, pattern, NULL, /*flags*/ 0,
+		&nsPtr, &dummy1NsPtr, &dummy2NsPtr, &simplePattern);
+
+	if (nsPtr != NULL) {	/* We successfully found the pattern's ns. */
+	    specificNsInPattern = (strcmp(simplePattern, pattern) != 0);
+	    if (simplePattern == pattern) {
+		simplePatternPtr = objv[1];
+	    } else {
+		simplePatternPtr = Tcl_NewStringObj(simplePattern, -1);
+	    }
+	    Tcl_IncrRefCount(simplePatternPtr);
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * If the namespace specified in the pattern wasn't found, just return.
+     */
+
+    if (nsPtr == NULL) {
+	return TCL_OK;
+    }
+
+    listPtr = Tcl_NewListObj(0, NULL);
+
+    if (!HasLocalVars(iPtr->varFramePtr) || specificNsInPattern) {
+	/*
+	 * There is no frame pointer, the frame pointer was pushed only to
+	 * activate a namespace, or we are in a procedure call frame but a
+	 * specific namespace was specified. Create a list containing only the
+	 * variables in the effective namespace's variable table.
+	 */
+
+	if (simplePattern && TclMatchIsTrivial(simplePattern)) {
+	    /*
+	     * If we can just do hash lookups, that simplifies things a lot.
+	     */
+
+	    varPtr = VarHashFindVar(&nsPtr->varTable, simplePatternPtr);
+	    if (varPtr) {
+		if (!TclIsVarUndefined(varPtr)
+			|| TclIsVarNamespaceVar(varPtr)) {
+		    if (specificNsInPattern) {
+			elemObjPtr = Tcl_NewObj();
+			Tcl_GetVariableFullName(interp, (Tcl_Var) varPtr,
+				elemObjPtr);
+		    } else {
+			elemObjPtr = VarHashGetKey(varPtr);
+		    }
+		    Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		}
+	    } else if ((nsPtr != globalNsPtr) && !specificNsInPattern) {
+		varPtr = VarHashFindVar(&globalNsPtr->varTable,
+			simplePatternPtr);
+		if (varPtr) {
+		    if (!TclIsVarUndefined(varPtr)
+			    || TclIsVarNamespaceVar(varPtr)) {
+			Tcl_ListObjAppendElement(interp, listPtr,
+				VarHashGetKey(varPtr));
+		    }
+		}
+	    }
+	} else {
+	    /*
+	     * Have to scan the tables of variables.
+	     */
+
+	    varPtr = VarHashFirstVar(&nsPtr->varTable, &search);
+	    while (varPtr) {
+		if (!TclIsVarUndefined(varPtr)
+			|| TclIsVarNamespaceVar(varPtr)) {
+		    varNamePtr = VarHashGetKey(varPtr);
+		    varName = TclGetString(varNamePtr);
+		    if ((simplePattern == NULL)
+			    || Tcl_StringMatch(varName, simplePattern)) {
+			if (specificNsInPattern) {
+			    elemObjPtr = Tcl_NewObj();
+			    Tcl_GetVariableFullName(interp, (Tcl_Var) varPtr,
+				    elemObjPtr);
+			} else {
+			    elemObjPtr = varNamePtr;
+			}
+			Tcl_ListObjAppendElement(interp, listPtr, elemObjPtr);
+		    }
+		}
+		varPtr = VarHashNextVar(&search);
+	    }
+
+	    /*
+	     * If the effective namespace isn't the global :: namespace, and a
+	     * specific namespace wasn't requested in the pattern (i.e., the
+	     * pattern only specifies variable names), then add in all global
+	     * :: variables that match the simple pattern. Of course, add in
+	     * only those variables that aren't hidden by a variable in the
+	     * effective namespace.
+	     */
+
+	    if ((nsPtr != globalNsPtr) && !specificNsInPattern) {
+		varPtr = VarHashFirstVar(&globalNsPtr->varTable,&search);
+		while (varPtr) {
+		    if (!TclIsVarUndefined(varPtr)
+			    || TclIsVarNamespaceVar(varPtr)) {
+			varNamePtr = VarHashGetKey(varPtr);
+			varName = TclGetString(varNamePtr);
+			if ((simplePattern == NULL)
+				|| Tcl_StringMatch(varName, simplePattern)) {
+			    if (VarHashFindVar(&nsPtr->varTable,
+				    varNamePtr) == NULL) {
+				Tcl_ListObjAppendElement(interp, listPtr,
+					varNamePtr);
+			    }
+			}
+		    }
+		    varPtr = VarHashNextVar(&search);
+		}
+	    }
+	}
+    } else if (iPtr->varFramePtr->procPtr != NULL) {
+	AppendLocals(interp, listPtr, simplePatternPtr, 1);
+    }
+
+    if (simplePatternPtr) {
+	Tcl_DecrRefCount(simplePatternPtr);
+    }
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * InfoGlobalsCmd -- (moved over from tclCmdIL.c)
+ *
+ *	Called to implement the "info globals" command that returns the list
+ *	of global variables matching an optional pattern. Handles the
+ *	following syntax:
+ *
+ *	    info globals ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInfoGlobalsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    const char *varName, *pattern;
+    Namespace *globalNsPtr = (Namespace *) Tcl_GetGlobalNamespace(interp);
+    Tcl_HashSearch search;
+    Var *varPtr;
+    Tcl_Obj *listPtr, *varNamePtr, *patternPtr;
+
+    if (objc == 1) {
+	pattern = NULL;
+    } else if (objc == 2) {
+	pattern = TclGetString(objv[1]);
+
+	/*
+	 * Strip leading global-namespace qualifiers. [Bug 1057461]
+	 */
+
+	if (pattern[0] == ':' && pattern[1] == ':') {
+	    while (*pattern == ':') {
+		pattern++;
+	    }
+	}
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Scan through the global :: namespace's variable table and create a list
+     * of all global variables that match the pattern.
+     */
+
+    listPtr = Tcl_NewListObj(0, NULL);
+    if (pattern != NULL && TclMatchIsTrivial(pattern)) {
+	if (pattern == TclGetString(objv[1])) {
+	    patternPtr = objv[1];
+	} else {
+	    patternPtr = Tcl_NewStringObj(pattern, -1);
+	}
+	Tcl_IncrRefCount(patternPtr);
+
+	varPtr = VarHashFindVar(&globalNsPtr->varTable, patternPtr);
+	if (varPtr) {
+	    if (!TclIsVarUndefined(varPtr)) {
+		Tcl_ListObjAppendElement(interp, listPtr,
+			VarHashGetKey(varPtr));
+	    }
+	}
+	Tcl_DecrRefCount(patternPtr);
+    } else {
+	for (varPtr = VarHashFirstVar(&globalNsPtr->varTable, &search);
+		varPtr != NULL;
+		varPtr = VarHashNextVar(&search)) {
+	    if (TclIsVarUndefined(varPtr)) {
+		continue;
+	    }
+	    varNamePtr = VarHashGetKey(varPtr);
+	    varName = TclGetString(varNamePtr);
+	    if ((pattern == NULL) || Tcl_StringMatch(varName, pattern)) {
+		Tcl_ListObjAppendElement(interp, listPtr, varNamePtr);
+	    }
+	}
+    }
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * TclInfoLocalsCmd -- (moved over from tclCmdIl.c)
+ *
+ *	Called to implement the "info locals" command to return a list of
+ *	local variables that match an optional pattern. Handles the following
+ *	syntax:
+ *
+ *	    info locals ?pattern?
+ *
+ * Results:
+ *	Returns TCL_OK if successful and TCL_ERROR if there is an error.
+ *
+ * Side effects:
+ *	Returns a result in the interpreter's result object. If there is an
+ *	error, the result is an error message.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+TclInfoLocalsCmd(
+    ClientData dummy,		/* Not used. */
+    Tcl_Interp *interp,		/* Current interpreter. */
+    int objc,			/* Number of arguments. */
+    Tcl_Obj *const objv[])	/* Argument objects. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Tcl_Obj *patternPtr, *listPtr;
+
+    if (objc == 1) {
+	patternPtr = NULL;
+    } else if (objc == 2) {
+	patternPtr = objv[1];
+    } else {
+	Tcl_WrongNumArgs(interp, 1, objv, "?pattern?");
+	return TCL_ERROR;
+    }
+
+    if (!HasLocalVars(iPtr->varFramePtr)) {
+	return TCL_OK;
+    }
+
+    /*
+     * Return a list containing names of first the compiled locals (i.e. the
+     * ones stored in the call frame), then the variables in the local hash
+     * table (if one exists).
+     */
+
+    listPtr = Tcl_NewListObj(0, NULL);
+    AppendLocals(interp, listPtr, patternPtr, 0);
+    Tcl_SetObjResult(interp, listPtr);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * AppendLocals --
+ *
+ *	Append the local variables for the current frame to the specified list
+ *	object.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+AppendLocals(
+    Tcl_Interp *interp,		/* Current interpreter. */
+    Tcl_Obj *listPtr,		/* List object to append names to. */
+    Tcl_Obj *patternPtr,	/* Pattern to match against. */
+    int includeLinks)		/* 1 if upvars should be included, else 0. */
+{
+    Interp *iPtr = (Interp *) interp;
+    Var *varPtr;
+    int i, localVarCt, added;
+    Tcl_Obj **varNamePtr, *objNamePtr;
+    const char *varName;
+    TclVarHashTable *localVarTablePtr;
+    Tcl_HashSearch search;
+    Tcl_HashTable addedTable;
+    const char *pattern = patternPtr? TclGetString(patternPtr) : NULL;
+
+    localVarCt = iPtr->varFramePtr->numCompiledLocals;
+    varPtr = iPtr->varFramePtr->compiledLocals;
+    localVarTablePtr = iPtr->varFramePtr->varTablePtr;
+    varNamePtr = &iPtr->varFramePtr->localCachePtr->varName0;
+    if (includeLinks) {
+	Tcl_InitObjHashTable(&addedTable);
+    }
+
+    for (i = 0; i < localVarCt; i++, varNamePtr++) {
+	/*
+	 * Skip nameless (temporary) variables and undefined variables.
+	 */
+
+	if (*varNamePtr && !TclIsVarUndefined(varPtr)
+		&& (includeLinks || !TclIsVarLink(varPtr))) {
+	    varName = TclGetString(*varNamePtr);
+	    if ((pattern == NULL) || Tcl_StringMatch(varName, pattern)) {
+		Tcl_ListObjAppendElement(interp, listPtr, *varNamePtr);
+		if (includeLinks) {
+		    Tcl_CreateHashEntry(&addedTable, *varNamePtr, &added);
+		}
+	    }
+	}
+	varPtr++;
+    }
+
+    /*
+     * Do nothing if no local variables.
+     */
+
+    if (localVarTablePtr == NULL) {
+	goto objectVars;
+    }
+
+    /*
+     * Check for the simple and fast case.
+     */
+
+    if ((pattern != NULL) && TclMatchIsTrivial(pattern)) {
+	varPtr = VarHashFindVar(localVarTablePtr, patternPtr);
+	if (varPtr != NULL) {
+	    if (!TclIsVarUndefined(varPtr)
+		    && (includeLinks || !TclIsVarLink(varPtr))) {
+		Tcl_ListObjAppendElement(interp, listPtr,
+			VarHashGetKey(varPtr));
+		if (includeLinks) {
+		    Tcl_CreateHashEntry(&addedTable, VarHashGetKey(varPtr),
+			    &added);
+		}
+	    }
+	}
+	goto objectVars;
+    }
+
+    /*
+     * Scan over and process all local variables.
+     */
+
+    for (varPtr = VarHashFirstVar(localVarTablePtr, &search);
+	    varPtr != NULL;
+	    varPtr = VarHashNextVar(&search)) {
+	if (!TclIsVarUndefined(varPtr)
+		&& (includeLinks || !TclIsVarLink(varPtr))) {
+	    objNamePtr = VarHashGetKey(varPtr);
+	    varName = TclGetString(objNamePtr);
+	    if ((pattern == NULL) || Tcl_StringMatch(varName, pattern)) {
+		Tcl_ListObjAppendElement(interp, listPtr, objNamePtr);
+		if (includeLinks) {
+		    Tcl_CreateHashEntry(&addedTable, objNamePtr, &added);
+		}
+	    }
+	}
+    }
+
+  objectVars:
+    if (!includeLinks) {
+	return;
+    }
+
+    if (iPtr->varFramePtr->isProcCallFrame & FRAME_IS_METHOD) {
+	CallContext *contextPtr = iPtr->varFramePtr->clientData;
+	Method *mPtr = contextPtr->callPtr->chain[contextPtr->index].mPtr;
+
+	if (mPtr->declaringObjectPtr) {
+	    FOREACH(objNamePtr, mPtr->declaringObjectPtr->variables) {
+		Tcl_CreateHashEntry(&addedTable, objNamePtr, &added);
+		if (added && (!pattern ||
+			Tcl_StringMatch(TclGetString(objNamePtr), pattern))) {
+		    Tcl_ListObjAppendElement(interp, listPtr, objNamePtr);
+		}
+	    }
+	} else {
+	    FOREACH(objNamePtr, mPtr->declaringClassPtr->variables) {
+		Tcl_CreateHashEntry(&addedTable, objNamePtr, &added);
+		if (added && (!pattern ||
+			Tcl_StringMatch(TclGetString(objNamePtr), pattern))) {
+		    Tcl_ListObjAppendElement(interp, listPtr, objNamePtr);
+		}
+	    }
+	}
+    }
+    Tcl_DeleteHashTable(&addedTable);
+}
+
+/*
+ * Hash table implementation - first, just copy and adapt the obj key stuff
+ */
+
+void
+TclInitVarHashTable(
+    TclVarHashTable *tablePtr,
+    Namespace *nsPtr)
+{
+    Tcl_InitCustomHashTable(&tablePtr->table,
+	    TCL_CUSTOM_TYPE_KEYS, &tclVarHashKeyType);
+    tablePtr->nsPtr = nsPtr;
+}
+
+static Tcl_HashEntry *
+AllocVarEntry(
+    Tcl_HashTable *tablePtr,	/* Hash table. */
+    void *keyPtr)		/* Key to store in the hash table entry. */
+{
+    Tcl_Obj *objPtr = keyPtr;
+    Tcl_HashEntry *hPtr;
+    Var *varPtr;
+
+    varPtr = ckalloc(sizeof(VarInHash));
+    varPtr->flags = VAR_IN_HASHTABLE;
+    varPtr->value.objPtr = NULL;
+    VarHashRefCount(varPtr) = 1;
+
+    hPtr = &(((VarInHash *) varPtr)->entry);
+    Tcl_SetHashValue(hPtr, varPtr);
+    hPtr->key.objPtr = objPtr;
+    Tcl_IncrRefCount(objPtr);
+
+    return hPtr;
+}
+
+static void
+FreeVarEntry(
+    Tcl_HashEntry *hPtr)
+{
+    Var *varPtr = VarHashGetValue(hPtr);
+    Tcl_Obj *objPtr = hPtr->key.objPtr;
+
+    if (TclIsVarUndefined(varPtr) && !TclIsVarTraced(varPtr)
+	    && (VarHashRefCount(varPtr) == 1)) {
+	ckfree(varPtr);
+    } else {
+	VarHashInvalidateEntry(varPtr);
+	TclSetVarUndefined(varPtr);
+	VarHashRefCount(varPtr)--;
+    }
+    Tcl_DecrRefCount(objPtr);
+}
+
+static int
+CompareVarKeys(
+    void *keyPtr,		/* New key to compare. */
+    Tcl_HashEntry *hPtr)	/* Existing key to compare. */
+{
+    Tcl_Obj *objPtr1 = keyPtr;
+    Tcl_Obj *objPtr2 = hPtr->key.objPtr;
+    register const char *p1, *p2;
+    register int l1, l2;
+
+    /*
+     * If the object pointers are the same then they match.
+     * OPT: this comparison was moved to the caller
+
+       if (objPtr1 == objPtr2) return 1;
+    */
+
+    /*
+     * Don't use Tcl_GetStringFromObj as it would prevent l1 and l2 being in a
+     * register.
+     */
+
+    p1 = TclGetString(objPtr1);
+    l1 = objPtr1->length;
+    p2 = TclGetString(objPtr2);
+    l2 = objPtr2->length;
+
+    /*
+     * Only compare string representations of the same length.
+     */
+
+    return ((l1 == l2) && !memcmp(p1, p2, l1));
+}
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tclZlib.c b/generic/tclZlib.c
new file mode 100644
index 0000000..33eebd1
--- /dev/null
+++ b/generic/tclZlib.c
@@ -0,0 +1,4058 @@
+/*
+ * tclZlib.c --
+ *
+ *	This file provides the interface to the Zlib library.
+ *
+ * Copyright (C) 2004-2005 Pascal Scheffers <pascal@scheffers.net>
+ * Copyright (C) 2005 Unitas Software B.V.
+ * Copyright (c) 2008-2012 Donal K. Fellows
+ *
+ * Parts written by Jean-Claude Wippler, as part of Tclkit, placed in the
+ * public domain March 2003.
+ *
+ * See the file "license.terms" for information on usage and redistribution of
+ * this file, and for a DISCLAIMER OF ALL WARRANTIES.
+ */
+
+#include "tclInt.h"
+#ifdef HAVE_ZLIB
+#include <zlib.h>
+#include "tclIO.h"
+
+/*
+ * The version of the zlib "package" that this implements. Note that this
+ * thoroughly supersedes the versions included with tclkit, which are "1.1",
+ * so this is at least "2.0" (there's no general *commitment* to have the same
+ * interface, even if that is mostly true).
+ */
+
+#define TCL_ZLIB_VERSION	"2.0.1"
+
+/*
+ * Magic flags used with wbits fields to indicate that we're handling the gzip
+ * format or automatic detection of format. Putting it here is slightly less
+ * gross!
+ */
+
+#define WBITS_RAW		(-MAX_WBITS)
+#define WBITS_ZLIB		(MAX_WBITS)
+#define WBITS_GZIP		(MAX_WBITS | 16)
+#define WBITS_AUTODETECT	(MAX_WBITS | 32)
+
+/*
+ * Structure used for handling gzip headers that are generated from a
+ * dictionary. It comprises the header structure itself plus some working
+ * space that it is very convenient to have attached.
+ */
+
+#define MAX_COMMENT_LEN		256
+
+typedef struct {
+    gz_header header;
+    char nativeFilenameBuf[MAXPATHLEN];
+    char nativeCommentBuf[MAX_COMMENT_LEN];
+} GzipHeader;
+
+/*
+ * Structure used for the Tcl_ZlibStream* commands and [zlib stream ...]
+ */
+
+typedef struct {
+    Tcl_Interp *interp;
+    z_stream stream;		/* The interface to the zlib library. */
+    int streamEnd;		/* If we've got to end-of-stream. */
+    Tcl_Obj *inData, *outData;	/* Input / output buffers (lists) */
+    Tcl_Obj *currentInput;	/* Pointer to what is currently being
+				 * inflated. */
+    int outPos;
+    int mode;			/* Either TCL_ZLIB_STREAM_DEFLATE or
+				 * TCL_ZLIB_STREAM_INFLATE. */
+    int format;			/* Flags from the TCL_ZLIB_FORMAT_* */
+    int level;			/* Default 5, 0-9 */
+    int flush;			/* Stores the flush param for deferred the
+				 * decompression. */
+    int wbits;			/* The encoded compression mode, so we can
+				 * restart the stream if necessary. */
+    Tcl_Command cmd;		/* Token for the associated Tcl command. */
+    Tcl_Obj *compDictObj;	/* Byte-array object containing compression
+				 * dictionary (not dictObj!) to use if
+				 * necessary. */
+    int flags;			/* Miscellaneous flag bits. */
+    GzipHeader *gzHeaderPtr;	/* If we've allocated a gzip header
+				 * structure. */
+} ZlibStreamHandle;
+
+#define DICT_TO_SET	0x1	/* If we need to set a compression dictionary
+				 * in the low-level engine at the next
+				 * opportunity. */
+
+/*
+ * Macros to make it clearer in some of the twiddlier accesses what is
+ * happening.
+ */
+
+#define IsRawStream(zshPtr)	((zshPtr)->format == TCL_ZLIB_FORMAT_RAW)
+#define HaveDictToSet(zshPtr)	((zshPtr)->flags & DICT_TO_SET)
+#define DictWasSet(zshPtr)	((zshPtr)->flags |= ~DICT_TO_SET)
+
+/*
+ * Structure used for stacked channel compression and decompression.
+ */
+
+typedef struct {
+    Tcl_Channel chan;		/* Reference to the channel itself. */
+    Tcl_Channel parent;		/* The underlying source and sink of bytes. */
+    int flags;			/* General flag bits, see below... */
+    int mode;			/* Either the value TCL_ZLIB_STREAM_DEFLATE
+				 * for compression on output, or
+				 * TCL_ZLIB_STREAM_INFLATE for decompression
+				 * on input. */
+    int format;			/* What format of data is going on the wire.
+				 * Needed so that the correct [fconfigure]
+				 * options can be enabled. */
+    int readAheadLimit;		/* The maximum number of bytes to read from
+				 * the underlying stream in one go. */
+    z_stream inStream;		/* Structure used by zlib for decompression of
+				 * input. */
+    z_stream outStream;		/* Structure used by zlib for compression of
+				 * output. */
+    char *inBuffer, *outBuffer;	/* Working buffers. */
+    int inAllocated, outAllocated;
+				/* Sizes of working buffers. */
+    GzipHeader inHeader;	/* Header read from input stream, when
+				 * decompressing a gzip stream. */
+    GzipHeader outHeader;	/* Header to write to an output stream, when
+				 * compressing a gzip stream. */
+    Tcl_TimerToken timer;	/* Timer used for keeping events fresh. */
+    Tcl_DString decompressed;	/* Buffer for decompression results. */
+    Tcl_Obj *compDictObj;	/* Byte-array object containing compression
+				 * dictionary (not dictObj!) to use if
+				 * necessary. */
+} ZlibChannelData;
+
+/*
+ * Value bits for the flags field. Definitions are:
+ *	ASYNC -		Whether this is an asynchronous channel.
+ *	IN_HEADER -	Whether the inHeader field has been registered with
+ *			the input compressor.
+ *	OUT_HEADER -	Whether the outputHeader field has been registered
+ *			with the output decompressor.
+ */
+
+#define ASYNC			0x1
+#define IN_HEADER		0x2
+#define OUT_HEADER		0x4
+
+/*
+ * Size of buffers allocated by default, and the range it can be set to.  The
+ * same sorts of values apply to streams, except with different limits (they
+ * permit byte-level activity). Channels always use bytes unless told to use
+ * larger buffers.
+ */
+
+#define DEFAULT_BUFFER_SIZE	4096
+#define MIN_NONSTREAM_BUFFER_SIZE 16
+#define MAX_BUFFER_SIZE		65536
+
+/*
+ * Prototypes for private procedures defined later in this file:
+ */
+
+static Tcl_CmdDeleteProc	ZlibStreamCmdDelete;
+static Tcl_DriverBlockModeProc	ZlibTransformBlockMode;
+static Tcl_DriverCloseProc	ZlibTransformClose;
+static Tcl_DriverGetHandleProc	ZlibTransformGetHandle;
+static Tcl_DriverGetOptionProc	ZlibTransformGetOption;
+static Tcl_DriverHandlerProc	ZlibTransformEventHandler;
+static Tcl_DriverInputProc	ZlibTransformInput;
+static Tcl_DriverOutputProc	ZlibTransformOutput;
+static Tcl_DriverSetOptionProc	ZlibTransformSetOption;
+static Tcl_DriverWatchProc	ZlibTransformWatch;
+static Tcl_ObjCmdProc		ZlibCmd;
+static Tcl_ObjCmdProc		ZlibStreamCmd;
+static Tcl_ObjCmdProc		ZlibStreamAddCmd;
+static Tcl_ObjCmdProc		ZlibStreamHeaderCmd;
+static Tcl_ObjCmdProc		ZlibStreamPutCmd;
+
+static void		ConvertError(Tcl_Interp *interp, int code,
+			    uLong adler);
+static Tcl_Obj *	ConvertErrorToList(int code, uLong adler);
+static inline int	Deflate(z_streamp strm, void *bufferPtr,
+			    int bufferSize, int flush, int *writtenPtr);
+static void		ExtractHeader(gz_header *headerPtr, Tcl_Obj *dictObj);
+static int		GenerateHeader(Tcl_Interp *interp, Tcl_Obj *dictObj,
+			    GzipHeader *headerPtr, int *extraSizePtr);
+static int		ZlibPushSubcmd(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static inline int	ResultCopy(ZlibChannelData *cd, char *buf,
+			    int toRead);
+static int		ResultGenerate(ZlibChannelData *cd, int n, int flush,
+			    int *errorCodePtr);
+static Tcl_Channel	ZlibStackChannelTransform(Tcl_Interp *interp,
+			    int mode, int format, int level, int limit,
+			    Tcl_Channel channel, Tcl_Obj *gzipHeaderDictPtr,
+			    Tcl_Obj *compDictObj);
+static void		ZlibStreamCleanup(ZlibStreamHandle *zshPtr);
+static int		ZlibStreamSubcmd(Tcl_Interp *interp, int objc,
+			    Tcl_Obj *const objv[]);
+static inline void	ZlibTransformEventTimerKill(ZlibChannelData *cd);
+static void		ZlibTransformTimerRun(ClientData clientData);
+
+/*
+ * Type of zlib-based compressing and decompressing channels.
+ */
+
+static const Tcl_ChannelType zlibChannelType = {
+    "zlib",
+    TCL_CHANNEL_VERSION_3,
+    ZlibTransformClose,
+    ZlibTransformInput,
+    ZlibTransformOutput,
+    NULL,			/* seekProc */
+    ZlibTransformSetOption,
+    ZlibTransformGetOption,
+    ZlibTransformWatch,
+    ZlibTransformGetHandle,
+    NULL,			/* close2Proc */
+    ZlibTransformBlockMode,
+    NULL,			/* flushProc */
+    ZlibTransformEventHandler,
+    NULL,			/* wideSeekProc */
+    NULL,
+    NULL
+};
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ConvertError --
+ *
+ *	Utility function for converting a zlib error into a Tcl error.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Updates the interpreter result and errorcode.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ConvertError(
+    Tcl_Interp *interp,		/* Interpreter to store the error in. May be
+				 * NULL, in which case nothing happens. */
+    int code,			/* The zlib error code. */
+    uLong adler)		/* The checksum expected (for Z_NEED_DICT) */
+{
+    const char *codeStr, *codeStr2 = NULL;
+    char codeStrBuf[TCL_INTEGER_SPACE];
+
+    if (interp == NULL) {
+	return;
+    }
+
+    switch (code) {
+	/*
+	 * Firstly, the case that is *different* because it's really coming
+	 * from the OS and is just being reported via zlib. It should be
+	 * really uncommon because Tcl handles all I/O rather than delegating
+	 * it to zlib, but proving it can't happen is hard.
+	 */
+
+    case Z_ERRNO:
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(Tcl_PosixError(interp),-1));
+	return;
+
+	/*
+	 * Normal errors/conditions, some of which have additional detail and
+	 * some which don't. (This is not defined by array lookup because zlib
+	 * error codes are sometimes negative.)
+	 */
+
+    case Z_STREAM_ERROR:
+	codeStr = "STREAM";
+	break;
+    case Z_DATA_ERROR:
+	codeStr = "DATA";
+	break;
+    case Z_MEM_ERROR:
+	codeStr = "MEM";
+	break;
+    case Z_BUF_ERROR:
+	codeStr = "BUF";
+	break;
+    case Z_VERSION_ERROR:
+	codeStr = "VERSION";
+	break;
+    case Z_NEED_DICT:
+	codeStr = "NEED_DICT";
+	codeStr2 = codeStrBuf;
+	sprintf(codeStrBuf, "%lu", adler);
+	break;
+
+	/*
+	 * These should _not_ happen! This function is for dealing with error
+	 * cases, not non-errors!
+	 */
+
+    case Z_OK:
+	Tcl_Panic("unexpected zlib result in error handler: Z_OK");
+    case Z_STREAM_END:
+	Tcl_Panic("unexpected zlib result in error handler: Z_STREAM_END");
+
+	/*
+	 * Anything else is bad news; it's unexpected. Convert to generic
+	 * error.
+	 */
+
+    default:
+	codeStr = "UNKNOWN";
+	codeStr2 = codeStrBuf;
+	sprintf(codeStrBuf, "%d", code);
+	break;
+    }
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(zError(code), -1));
+
+    /*
+     * Tricky point! We might pass NULL twice here (and will when the error
+     * type is known).
+     */
+
+    Tcl_SetErrorCode(interp, "TCL", "ZLIB", codeStr, codeStr2, NULL);
+}
+
+static Tcl_Obj *
+ConvertErrorToList(
+    int code,			/* The zlib error code. */
+    uLong adler)		/* The checksum expected (for Z_NEED_DICT) */
+{
+    Tcl_Obj *objv[4];
+
+    TclNewLiteralStringObj(objv[0], "TCL");
+    TclNewLiteralStringObj(objv[1], "ZLIB");
+    switch (code) {
+    case Z_STREAM_ERROR:
+	TclNewLiteralStringObj(objv[2], "STREAM");
+	return Tcl_NewListObj(3, objv);
+    case Z_DATA_ERROR:
+	TclNewLiteralStringObj(objv[2], "DATA");
+	return Tcl_NewListObj(3, objv);
+    case Z_MEM_ERROR:
+	TclNewLiteralStringObj(objv[2], "MEM");
+	return Tcl_NewListObj(3, objv);
+    case Z_BUF_ERROR:
+	TclNewLiteralStringObj(objv[2], "BUF");
+	return Tcl_NewListObj(3, objv);
+    case Z_VERSION_ERROR:
+	TclNewLiteralStringObj(objv[2], "VERSION");
+	return Tcl_NewListObj(3, objv);
+    case Z_ERRNO:
+	TclNewLiteralStringObj(objv[2], "POSIX");
+	objv[3] = Tcl_NewStringObj(Tcl_ErrnoId(), -1);
+	return Tcl_NewListObj(4, objv);
+    case Z_NEED_DICT:
+	TclNewLiteralStringObj(objv[2], "NEED_DICT");
+	objv[3] = Tcl_NewWideIntObj((Tcl_WideInt) adler);
+	return Tcl_NewListObj(4, objv);
+
+	/*
+	 * These should _not_ happen! This function is for dealing with error
+	 * cases, not non-errors!
+	 */
+
+    case Z_OK:
+	Tcl_Panic("unexpected zlib result in error handler: Z_OK");
+    case Z_STREAM_END:
+	Tcl_Panic("unexpected zlib result in error handler: Z_STREAM_END");
+
+	/*
+	 * Catch-all. Should be unreachable because all cases are already
+	 * listed above.
+	 */
+
+    default:
+	TclNewLiteralStringObj(objv[2], "UNKNOWN");
+	TclNewLongObj(objv[3], code);
+	return Tcl_NewListObj(4, objv);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * GenerateHeader --
+ *
+ *	Function for creating a gzip header from the contents of a dictionary
+ *	(as described in the documentation). GetValue is a helper function.
+ *
+ * Results:
+ *	A Tcl result code.
+ *
+ * Side effects:
+ *	Updates the fields of the given gz_header structure. Adds amount of
+ *	extra space required for the header to the variable referenced by the
+ *	extraSizePtr argument.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+GetValue(
+    Tcl_Interp *interp,
+    Tcl_Obj *dictObj,
+    const char *nameStr,
+    Tcl_Obj **valuePtrPtr)
+{
+    Tcl_Obj *name = Tcl_NewStringObj(nameStr, -1);
+    int result = Tcl_DictObjGet(interp, dictObj, name, valuePtrPtr);
+
+    TclDecrRefCount(name);
+    return result;
+}
+
+static int
+GenerateHeader(
+    Tcl_Interp *interp,		/* Where to put error messages. */
+    Tcl_Obj *dictObj,		/* The dictionary whose contents are to be
+				 * parsed. */
+    GzipHeader *headerPtr,	/* Where to store the parsed-out values. */
+    int *extraSizePtr)		/* Variable to add the length of header
+				 * strings (filename, comment) to. */
+{
+    Tcl_Obj *value;
+    int len, result = TCL_ERROR;
+    const char *valueStr;
+    Tcl_Encoding latin1enc;
+    static const char *const types[] = {
+	"binary", "text"
+    };
+
+    /*
+     * RFC 1952 says that header strings are in ISO 8859-1 (LATIN-1).
+     */
+
+    latin1enc = Tcl_GetEncoding(NULL, "iso8859-1");
+    if (latin1enc == NULL) {
+	Tcl_Panic("no latin-1 encoding");
+    }
+
+    if (GetValue(interp, dictObj, "comment", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL) {
+	valueStr = TclGetStringFromObj(value, &len);
+	Tcl_UtfToExternal(NULL, latin1enc, valueStr, len, 0, NULL,
+		headerPtr->nativeCommentBuf, MAX_COMMENT_LEN-1, NULL, &len,
+		NULL);
+	headerPtr->nativeCommentBuf[len] = '\0';
+	headerPtr->header.comment = (Bytef *) headerPtr->nativeCommentBuf;
+	if (extraSizePtr != NULL) {
+	    *extraSizePtr += len;
+	}
+    }
+
+    if (GetValue(interp, dictObj, "crc", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL &&
+	    Tcl_GetBooleanFromObj(interp, value, &headerPtr->header.hcrc)) {
+	goto error;
+    }
+
+    if (GetValue(interp, dictObj, "filename", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL) {
+	valueStr = TclGetStringFromObj(value, &len);
+	Tcl_UtfToExternal(NULL, latin1enc, valueStr, len, 0, NULL,
+		headerPtr->nativeFilenameBuf, MAXPATHLEN-1, NULL, &len, NULL);
+	headerPtr->nativeFilenameBuf[len] = '\0';
+	headerPtr->header.name = (Bytef *) headerPtr->nativeFilenameBuf;
+	if (extraSizePtr != NULL) {
+	    *extraSizePtr += len;
+	}
+    }
+
+    if (GetValue(interp, dictObj, "os", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL && Tcl_GetIntFromObj(interp, value,
+	    &headerPtr->header.os) != TCL_OK) {
+	goto error;
+    }
+
+    /*
+     * Ignore the 'size' field, since that is controlled by the size of the
+     * input data.
+     */
+
+    if (GetValue(interp, dictObj, "time", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL && Tcl_GetLongFromObj(interp, value,
+	    (long *) &headerPtr->header.time) != TCL_OK) {
+	goto error;
+    }
+
+    if (GetValue(interp, dictObj, "type", &value) != TCL_OK) {
+	goto error;
+    } else if (value != NULL && Tcl_GetIndexFromObj(interp, value, types,
+	    "type", TCL_EXACT, &headerPtr->header.text) != TCL_OK) {
+	goto error;
+    }
+
+    result = TCL_OK;
+  error:
+    Tcl_FreeEncoding(latin1enc);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ExtractHeader --
+ *
+ *	Take the values out of a gzip header and store them in a dictionary.
+ *	SetValue is a helper macro.
+ *
+ * Results:
+ *	None.
+ *
+ * Side effects:
+ *	Updates the dictionary, which must be writable (i.e. refCount < 2).
+ *
+ *----------------------------------------------------------------------
+ */
+
+#define SetValue(dictObj, key, value) \
+	Tcl_DictObjPut(NULL, (dictObj), Tcl_NewStringObj((key), -1), (value))
+
+static void
+ExtractHeader(
+    gz_header *headerPtr,	/* The gzip header to extract from. */
+    Tcl_Obj *dictObj)		/* The dictionary to store in. */
+{
+    Tcl_Encoding latin1enc = NULL;
+    Tcl_DString tmp;
+
+    if (headerPtr->comment != Z_NULL) {
+	if (latin1enc == NULL) {
+	    /*
+	     * RFC 1952 says that header strings are in ISO 8859-1 (LATIN-1).
+	     */
+
+	    latin1enc = Tcl_GetEncoding(NULL, "iso8859-1");
+	    if (latin1enc == NULL) {
+		Tcl_Panic("no latin-1 encoding");
+	    }
+	}
+
+	Tcl_ExternalToUtfDString(latin1enc, (char *) headerPtr->comment, -1,
+		&tmp);
+	SetValue(dictObj, "comment", TclDStringToObj(&tmp));
+    }
+    SetValue(dictObj, "crc", Tcl_NewBooleanObj(headerPtr->hcrc));
+    if (headerPtr->name != Z_NULL) {
+	if (latin1enc == NULL) {
+	    /*
+	     * RFC 1952 says that header strings are in ISO 8859-1 (LATIN-1).
+	     */
+
+	    latin1enc = Tcl_GetEncoding(NULL, "iso8859-1");
+	    if (latin1enc == NULL) {
+		Tcl_Panic("no latin-1 encoding");
+	    }
+	}
+
+	Tcl_ExternalToUtfDString(latin1enc, (char *) headerPtr->name, -1,
+		&tmp);
+	SetValue(dictObj, "filename", TclDStringToObj(&tmp));
+    }
+    if (headerPtr->os != 255) {
+	SetValue(dictObj, "os", Tcl_NewIntObj(headerPtr->os));
+    }
+    if (headerPtr->time != 0 /* magic - no time */) {
+	SetValue(dictObj, "time", Tcl_NewLongObj((long) headerPtr->time));
+    }
+    if (headerPtr->text != Z_UNKNOWN) {
+	SetValue(dictObj, "type",
+		Tcl_NewStringObj(headerPtr->text ? "text" : "binary", -1));
+    }
+
+    if (latin1enc != NULL) {
+	Tcl_FreeEncoding(latin1enc);
+    }
+}
+
+/*
+ * Disentangle the worst of how the zlib API is used.
+ */
+
+static int
+SetInflateDictionary(
+    z_streamp strm,
+    Tcl_Obj *compDictObj)
+{
+    if (compDictObj != NULL) {
+	int length;
+	unsigned char *bytes = Tcl_GetByteArrayFromObj(compDictObj, &length);
+
+	return inflateSetDictionary(strm, bytes, (unsigned) length);
+    }
+    return Z_OK;
+}
+
+static int
+SetDeflateDictionary(
+    z_streamp strm,
+    Tcl_Obj *compDictObj)
+{
+    if (compDictObj != NULL) {
+	int length;
+	unsigned char *bytes = Tcl_GetByteArrayFromObj(compDictObj, &length);
+
+	return deflateSetDictionary(strm, bytes, (unsigned) length);
+    }
+    return Z_OK;
+}
+
+static inline int
+Deflate(
+    z_streamp strm,
+    void *bufferPtr,
+    int bufferSize,
+    int flush,
+    int *writtenPtr)
+{
+    int e;
+
+    strm->next_out = (Bytef *) bufferPtr;
+    strm->avail_out = (unsigned) bufferSize;
+    e = deflate(strm, flush);
+    if (writtenPtr != NULL) {
+	*writtenPtr = bufferSize - strm->avail_out;
+    }
+    return e;
+}
+
+static inline void
+AppendByteArray(
+    Tcl_Obj *listObj,
+    void *buffer,
+    int size)
+{
+    if (size > 0) {
+	Tcl_Obj *baObj = Tcl_NewByteArrayObj((unsigned char *) buffer, size);
+
+	Tcl_ListObjAppendElement(NULL, listObj, baObj);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamInit --
+ *
+ *	This command initializes a (de)compression context/handle for
+ *	(de)compressing data in chunks.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	The variable pointed to by zshandlePtr is initialised and memory
+ *	allocated for internal state. Additionally, if interp is not null, a
+ *	Tcl command is created and its name placed in the interp result obj.
+ *
+ * Note:
+ *	At least one of interp and zshandlePtr should be non-NULL or the
+ *	reference to the stream will be completely lost.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamInit(
+    Tcl_Interp *interp,
+    int mode,			/* Either TCL_ZLIB_STREAM_INFLATE or
+				 * TCL_ZLIB_STREAM_DEFLATE. */
+    int format,			/* Flags from the TCL_ZLIB_FORMAT_* set. */
+    int level,			/* 0-9 or TCL_ZLIB_COMPRESS_DEFAULT. */
+    Tcl_Obj *dictObj,		/* Dictionary containing headers for gzip. */
+    Tcl_ZlibStream *zshandlePtr)
+{
+    int wbits = 0;
+    int e;
+    ZlibStreamHandle *zshPtr = NULL;
+    Tcl_DString cmdname;
+    GzipHeader *gzHeaderPtr = NULL;
+
+    switch (mode) {
+    case TCL_ZLIB_STREAM_DEFLATE:
+	/*
+	 * Compressed format is specified by the wbits parameter. See zlib.h
+	 * for details.
+	 */
+
+	switch (format) {
+	case TCL_ZLIB_FORMAT_RAW:
+	    wbits = WBITS_RAW;
+	    break;
+	case TCL_ZLIB_FORMAT_GZIP:
+	    wbits = WBITS_GZIP;
+	    if (dictObj) {
+		gzHeaderPtr = ckalloc(sizeof(GzipHeader));
+		memset(gzHeaderPtr, 0, sizeof(GzipHeader));
+		if (GenerateHeader(interp, dictObj, gzHeaderPtr,
+			NULL) != TCL_OK) {
+		    ckfree(gzHeaderPtr);
+		    return TCL_ERROR;
+		}
+	    }
+	    break;
+	case TCL_ZLIB_FORMAT_ZLIB:
+	    wbits = WBITS_ZLIB;
+	    break;
+	default:
+	    Tcl_Panic("incorrect zlib data format, must be "
+		    "TCL_ZLIB_FORMAT_ZLIB, TCL_ZLIB_FORMAT_GZIP or "
+		    "TCL_ZLIB_FORMAT_RAW");
+	}
+	if (level < -1 || level > 9) {
+	    Tcl_Panic("compression level should be between 0 (no compression)"
+		    " and 9 (best compression) or -1 for default compression "
+		    "level");
+	}
+	break;
+    case TCL_ZLIB_STREAM_INFLATE:
+	/*
+	 * wbits are the same as DEFLATE, but FORMAT_AUTO is valid too.
+	 */
+
+	switch (format) {
+	case TCL_ZLIB_FORMAT_RAW:
+	    wbits = WBITS_RAW;
+	    break;
+	case TCL_ZLIB_FORMAT_GZIP:
+	    wbits = WBITS_GZIP;
+	    gzHeaderPtr = ckalloc(sizeof(GzipHeader));
+	    memset(gzHeaderPtr, 0, sizeof(GzipHeader));
+	    gzHeaderPtr->header.name = (Bytef *)
+		    gzHeaderPtr->nativeFilenameBuf;
+	    gzHeaderPtr->header.name_max = MAXPATHLEN - 1;
+	    gzHeaderPtr->header.comment = (Bytef *)
+		    gzHeaderPtr->nativeCommentBuf;
+	    gzHeaderPtr->header.name_max = MAX_COMMENT_LEN - 1;
+	    break;
+	case TCL_ZLIB_FORMAT_ZLIB:
+	    wbits = WBITS_ZLIB;
+	    break;
+	case TCL_ZLIB_FORMAT_AUTO:
+	    wbits = WBITS_AUTODETECT;
+	    break;
+	default:
+	    Tcl_Panic("incorrect zlib data format, must be "
+		    "TCL_ZLIB_FORMAT_ZLIB, TCL_ZLIB_FORMAT_GZIP, "
+		    "TCL_ZLIB_FORMAT_RAW or TCL_ZLIB_FORMAT_AUTO");
+	}
+	break;
+    default:
+	Tcl_Panic("bad mode, must be TCL_ZLIB_STREAM_DEFLATE or"
+		" TCL_ZLIB_STREAM_INFLATE");
+    }
+
+    zshPtr = ckalloc(sizeof(ZlibStreamHandle));
+    zshPtr->interp = interp;
+    zshPtr->mode = mode;
+    zshPtr->format = format;
+    zshPtr->level = level;
+    zshPtr->wbits = wbits;
+    zshPtr->currentInput = NULL;
+    zshPtr->streamEnd = 0;
+    zshPtr->compDictObj = NULL;
+    zshPtr->flags = 0;
+    zshPtr->gzHeaderPtr = gzHeaderPtr;
+    memset(&zshPtr->stream, 0, sizeof(z_stream));
+    zshPtr->stream.adler = 1;
+
+    /*
+     * No output buffer available yet
+     */
+
+    if (mode == TCL_ZLIB_STREAM_DEFLATE) {
+	e = deflateInit2(&zshPtr->stream, level, Z_DEFLATED, wbits,
+		MAX_MEM_LEVEL, Z_DEFAULT_STRATEGY);
+	if (e == Z_OK && zshPtr->gzHeaderPtr) {
+	    e = deflateSetHeader(&zshPtr->stream,
+		    &zshPtr->gzHeaderPtr->header);
+	}
+    } else {
+	e = inflateInit2(&zshPtr->stream, wbits);
+	if (e == Z_OK && zshPtr->gzHeaderPtr) {
+	    e = inflateGetHeader(&zshPtr->stream,
+		    &zshPtr->gzHeaderPtr->header);
+	}
+    }
+
+    if (e != Z_OK) {
+	ConvertError(interp, e, zshPtr->stream.adler);
+	goto error;
+    }
+
+    /*
+     * I could do all this in C, but this is easier.
+     */
+
+    if (interp != NULL) {
+	if (Tcl_EvalEx(interp, "::incr ::tcl::zlib::cmdcounter", -1, 0) != TCL_OK) {
+	    goto error;
+	}
+	Tcl_DStringInit(&cmdname);
+	TclDStringAppendLiteral(&cmdname, "::tcl::zlib::streamcmd_");
+	TclDStringAppendObj(&cmdname, Tcl_GetObjResult(interp));
+	if (Tcl_FindCommand(interp, Tcl_DStringValue(&cmdname),
+		NULL, 0) != NULL) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "BUG: Stream command name already exists", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "BUG", "EXISTING_CMD", NULL);
+	    Tcl_DStringFree(&cmdname);
+	    goto error;
+	}
+	Tcl_ResetResult(interp);
+
+	/*
+	 * Create the command.
+	 */
+
+	zshPtr->cmd = Tcl_CreateObjCommand(interp, Tcl_DStringValue(&cmdname),
+		ZlibStreamCmd, zshPtr, ZlibStreamCmdDelete);
+	Tcl_DStringFree(&cmdname);
+	if (zshPtr->cmd == NULL) {
+	    goto error;
+	}
+    } else {
+	zshPtr->cmd = NULL;
+    }
+
+    /*
+     * Prepare the buffers for use.
+     */
+
+    zshPtr->inData = Tcl_NewListObj(0, NULL);
+    Tcl_IncrRefCount(zshPtr->inData);
+    zshPtr->outData = Tcl_NewListObj(0, NULL);
+    Tcl_IncrRefCount(zshPtr->outData);
+
+    zshPtr->outPos = 0;
+
+    /*
+     * Now set the variable pointed to by *zshandlePtr to the pointer to the
+     * zsh struct.
+     */
+
+    if (zshandlePtr) {
+	*zshandlePtr = (Tcl_ZlibStream) zshPtr;
+    }
+
+    return TCL_OK;
+
+  error:
+    if (zshPtr->compDictObj) {
+	Tcl_DecrRefCount(zshPtr->compDictObj);
+    }
+    if (zshPtr->gzHeaderPtr) {
+	ckfree(zshPtr->gzHeaderPtr);
+    }
+    ckfree(zshPtr);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibStreamCmdDelete --
+ *
+ *	This is the delete command which Tcl invokes when a zlibstream command
+ *	is deleted from the interpreter (on stream close, usually).
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Invalidates the zlib stream handle as obtained from Tcl_ZlibStreamInit
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ZlibStreamCmdDelete(
+    ClientData cd)
+{
+    ZlibStreamHandle *zshPtr = cd;
+
+    zshPtr->cmd = NULL;
+    ZlibStreamCleanup(zshPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamClose --
+ *
+ *	This procedure must be called after (de)compression is done to ensure
+ *	memory is freed and the command is deleted from the interpreter (if
+ *	any).
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Invalidates the zlib stream handle as obtained from Tcl_ZlibStreamInit
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamClose(
+    Tcl_ZlibStream zshandle)	/* As obtained from Tcl_ZlibStreamInit. */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+
+    /*
+     * If the interp is set, deleting the command will trigger
+     * ZlibStreamCleanup in ZlibStreamCmdDelete. If no interp is set, call
+     * ZlibStreamCleanup directly.
+     */
+
+    if (zshPtr->interp && zshPtr->cmd) {
+	Tcl_DeleteCommandFromToken(zshPtr->interp, zshPtr->cmd);
+    } else {
+	ZlibStreamCleanup(zshPtr);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibStreamCleanup --
+ *
+ *	This procedure is called by either Tcl_ZlibStreamClose or
+ *	ZlibStreamCmdDelete to cleanup the stream context.
+ *
+ * Results:
+ *	None
+ *
+ * Side effects:
+ *	Invalidates the zlib stream handle.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+ZlibStreamCleanup(
+    ZlibStreamHandle *zshPtr)
+{
+    if (!zshPtr->streamEnd) {
+	if (zshPtr->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	    deflateEnd(&zshPtr->stream);
+	} else {
+	    inflateEnd(&zshPtr->stream);
+	}
+    }
+
+    if (zshPtr->inData) {
+	Tcl_DecrRefCount(zshPtr->inData);
+    }
+    if (zshPtr->outData) {
+	Tcl_DecrRefCount(zshPtr->outData);
+    }
+    if (zshPtr->currentInput) {
+	Tcl_DecrRefCount(zshPtr->currentInput);
+    }
+    if (zshPtr->compDictObj) {
+	Tcl_DecrRefCount(zshPtr->compDictObj);
+    }
+    if (zshPtr->gzHeaderPtr) {
+	ckfree(zshPtr->gzHeaderPtr);
+    }
+
+    ckfree(zshPtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamReset --
+ *
+ *	This procedure will reinitialize an existing stream handle.
+ *
+ * Results:
+ *	A standard Tcl result.
+ *
+ * Side effects:
+ *	Any data left in the (de)compression buffer is lost.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamReset(
+    Tcl_ZlibStream zshandle)	/* As obtained from Tcl_ZlibStreamInit */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+    int e;
+
+    if (!zshPtr->streamEnd) {
+	if (zshPtr->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	    deflateEnd(&zshPtr->stream);
+	} else {
+	    inflateEnd(&zshPtr->stream);
+	}
+    }
+    Tcl_SetByteArrayLength(zshPtr->inData, 0);
+    Tcl_SetByteArrayLength(zshPtr->outData, 0);
+    if (zshPtr->currentInput) {
+	Tcl_DecrRefCount(zshPtr->currentInput);
+	zshPtr->currentInput = NULL;
+    }
+
+    zshPtr->outPos = 0;
+    zshPtr->streamEnd = 0;
+    memset(&zshPtr->stream, 0, sizeof(z_stream));
+
+    /*
+     * No output buffer available yet.
+     */
+
+    if (zshPtr->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	e = deflateInit2(&zshPtr->stream, zshPtr->level, Z_DEFLATED,
+		zshPtr->wbits, MAX_MEM_LEVEL, Z_DEFAULT_STRATEGY);
+	if (e == Z_OK && HaveDictToSet(zshPtr)) {
+	    e = SetDeflateDictionary(&zshPtr->stream, zshPtr->compDictObj);
+	    if (e == Z_OK) {
+		DictWasSet(zshPtr);
+	    }
+	}
+    } else {
+	e = inflateInit2(&zshPtr->stream, zshPtr->wbits);
+	if (IsRawStream(zshPtr) && HaveDictToSet(zshPtr) && e == Z_OK) {
+	    e = SetInflateDictionary(&zshPtr->stream, zshPtr->compDictObj);
+	    if (e == Z_OK) {
+		DictWasSet(zshPtr);
+	    }
+	}
+    }
+
+    if (e != Z_OK) {
+	ConvertError(zshPtr->interp, e, zshPtr->stream.adler);
+	/* TODO:cleanup */
+	return TCL_ERROR;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamGetCommandName --
+ *
+ *	This procedure will return the command name associated with the
+ *	stream.
+ *
+ * Results:
+ *	A Tcl_Obj with the name of the Tcl command or NULL if no command is
+ *	associated with the stream.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+Tcl_Obj *
+Tcl_ZlibStreamGetCommandName(
+    Tcl_ZlibStream zshandle)	/* As obtained from Tcl_ZlibStreamInit */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+    Tcl_Obj *objPtr;
+
+    if (!zshPtr->interp) {
+	return NULL;
+    }
+
+    TclNewObj(objPtr);
+    Tcl_GetCommandFullName(zshPtr->interp, zshPtr->cmd, objPtr);
+    return objPtr;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamEof --
+ *
+ *	This procedure This function returns 0 or 1 depending on the state of
+ *	the (de)compressor. For decompression, eof is reached when the entire
+ *	compressed stream has been decompressed. For compression, eof is
+ *	reached when the stream has been flushed with TCL_ZLIB_FINALIZE.
+ *
+ * Results:
+ *	Integer.
+ *
+ * Side effects:
+ *	None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamEof(
+    Tcl_ZlibStream zshandle)	/* As obtained from Tcl_ZlibStreamInit */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+
+    return zshPtr->streamEnd;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamChecksum --
+ *
+ *	Return the checksum of the uncompressed data seen so far by the
+ *	stream.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamChecksum(
+    Tcl_ZlibStream zshandle)	/* As obtained from Tcl_ZlibStreamInit */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+
+    return zshPtr->stream.adler;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamSetCompressionDictionary --
+ *
+ *	Sets the compression dictionary for a stream. This will be used as
+ *	appropriate for the next compression or decompression action performed
+ *	on the stream.
+ *
+ *----------------------------------------------------------------------
+ */
+
+void
+Tcl_ZlibStreamSetCompressionDictionary(
+    Tcl_ZlibStream zshandle,
+    Tcl_Obj *compressionDictionaryObj)
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+
+    if (compressionDictionaryObj != NULL) {
+	if (Tcl_IsShared(compressionDictionaryObj)) {
+	    compressionDictionaryObj =
+		    Tcl_DuplicateObj(compressionDictionaryObj);
+	}
+	Tcl_IncrRefCount(compressionDictionaryObj);
+	zshPtr->flags |= DICT_TO_SET;
+    } else {
+	zshPtr->flags &= ~DICT_TO_SET;
+    }
+    if (zshPtr->compDictObj != NULL) {
+	Tcl_DecrRefCount(zshPtr->compDictObj);
+    }
+    zshPtr->compDictObj = compressionDictionaryObj;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamPut --
+ *
+ *	Add data to the stream for compression or decompression from a
+ *	bytearray Tcl_Obj.
+ *
+ *----------------------------------------------------------------------
+ */
+
+#define BUFFER_SIZE_LIMIT	0xFFFF
+
+int
+Tcl_ZlibStreamPut(
+    Tcl_ZlibStream zshandle,	/* As obtained from Tcl_ZlibStreamInit */
+    Tcl_Obj *data,		/* Data to compress/decompress */
+    int flush)			/* TCL_ZLIB_NO_FLUSH, TCL_ZLIB_FLUSH,
+				 * TCL_ZLIB_FULLFLUSH, or TCL_ZLIB_FINALIZE */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+    char *dataTmp = NULL;
+    int e, size, outSize, toStore;
+
+    if (zshPtr->streamEnd) {
+	if (zshPtr->interp) {
+	    Tcl_SetObjResult(zshPtr->interp, Tcl_NewStringObj(
+		    "already past compressed stream end", -1));
+	    Tcl_SetErrorCode(zshPtr->interp, "TCL", "ZIP", "CLOSED", NULL);
+	}
+	return TCL_ERROR;
+    }
+
+    if (zshPtr->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	zshPtr->stream.next_in = Tcl_GetByteArrayFromObj(data, &size);
+	zshPtr->stream.avail_in = size;
+
+	/*
+	 * Must not do a zero-length compress unless finalizing. [Bug 25842c161]
+	 */
+
+	if (size == 0 && flush != Z_FINISH) {
+	    return TCL_OK;
+	}
+
+	if (HaveDictToSet(zshPtr)) {
+	    e = SetDeflateDictionary(&zshPtr->stream, zshPtr->compDictObj);
+	    if (e != Z_OK) {
+		ConvertError(zshPtr->interp, e, zshPtr->stream.adler);
+		return TCL_ERROR;
+	    }
+	    DictWasSet(zshPtr);
+	}
+
+	/*
+	 * deflateBound() doesn't seem to take various header sizes into
+	 * account, so we add 100 extra bytes. However, we can also loop
+	 * around again so we also set an upper bound on the output buffer
+	 * size.
+	 */
+
+	outSize = deflateBound(&zshPtr->stream, size) + 100;
+	if (outSize > BUFFER_SIZE_LIMIT) {
+	    outSize = BUFFER_SIZE_LIMIT;
+	}
+	dataTmp = ckalloc(outSize);
+
+	while (1) {
+	    e = Deflate(&zshPtr->stream, dataTmp, outSize, flush, &toStore);
+
+	    /*
+	     * Test if we've filled the buffer up and have to ask deflate() to
+	     * give us some more. Note that the condition for needing to
+	     * repeat a buffer transfer when the result is Z_OK is whether
+	     * there is no more space in the buffer we provided; the zlib
+	     * library does not necessarily return a different code in that
+	     * case. [Bug b26e38a3e4] [Tk Bug 10f2e7872b]
+	     */
+
+	    if ((e != Z_BUF_ERROR) && (e != Z_OK || toStore < outSize)) {
+		if ((e == Z_OK) || (flush == Z_FINISH && e == Z_STREAM_END)) {
+		    break;
+		}
+		ConvertError(zshPtr->interp, e, zshPtr->stream.adler);
+		return TCL_ERROR;
+	    }
+
+	    /*
+	     * Output buffer too small to hold the data being generated or we
+	     * are doing the end-of-stream flush (which can spit out masses of
+	     * data). This means we need to put a new buffer into place after
+	     * saving the old generated data to the outData list.
+	     */
+
+	    AppendByteArray(zshPtr->outData, dataTmp, outSize);
+
+	    if (outSize < BUFFER_SIZE_LIMIT) {
+		outSize = BUFFER_SIZE_LIMIT;
+		/* There may be *lots* of data left to output... */
+		dataTmp = ckrealloc(dataTmp, outSize);
+	    }
+	}
+
+	/*
+	 * And append the final data block to the outData list.
+	 */
+
+	AppendByteArray(zshPtr->outData, dataTmp, toStore);
+	ckfree(dataTmp);
+    } else {
+	/*
+	 * This is easy. Just append to the inData list.
+	 */
+
+	Tcl_ListObjAppendElement(NULL, zshPtr->inData, data);
+
+	/*
+	 * and we'll need the flush parameter for the Inflate call.
+	 */
+
+	zshPtr->flush = flush;
+    }
+
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibStreamGet --
+ *
+ *	Retrieve data (now compressed or decompressed) from the stream into a
+ *	bytearray Tcl_Obj.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibStreamGet(
+    Tcl_ZlibStream zshandle,	/* As obtained from Tcl_ZlibStreamInit */
+    Tcl_Obj *data,		/* A place to append the data. */
+    int count)			/* Number of bytes to grab as a maximum, you
+				 * may get less! */
+{
+    ZlibStreamHandle *zshPtr = (ZlibStreamHandle *) zshandle;
+    int e, i, listLen, itemLen, dataPos = 0;
+    Tcl_Obj *itemObj;
+    unsigned char *dataPtr, *itemPtr;
+    int existing;
+
+    /*
+     * Getting beyond the of stream, just return empty string.
+     */
+
+    if (zshPtr->streamEnd) {
+	return TCL_OK;
+    }
+
+    (void) Tcl_GetByteArrayFromObj(data, &existing);
+
+    if (zshPtr->mode == TCL_ZLIB_STREAM_INFLATE) {
+	if (count == -1) {
+	    /*
+	     * The only safe thing to do is restict to 65k. We might cause a
+	     * panic for out of memory if we just kept growing the buffer.
+	     */
+
+	    count = MAX_BUFFER_SIZE;
+	}
+
+	/*
+	 * Prepare the place to store the data.
+	 */
+
+	dataPtr = Tcl_SetByteArrayLength(data, existing+count);
+	dataPtr += existing;
+
+	zshPtr->stream.next_out = dataPtr;
+	zshPtr->stream.avail_out = count;
+	if (zshPtr->stream.avail_in == 0) {
+	    /*
+	     * zlib will probably need more data to decompress.
+	     */
+
+	    if (zshPtr->currentInput) {
+		Tcl_DecrRefCount(zshPtr->currentInput);
+		zshPtr->currentInput = NULL;
+	    }
+	    Tcl_ListObjLength(NULL, zshPtr->inData, &listLen);
+	    if (listLen > 0) {
+		/*
+		 * There is more input available, get it from the list and
+		 * give it to zlib. At this point, the data must not be shared
+		 * since we require the bytearray representation to not vanish
+		 * under our feet. [Bug 3081008]
+		 */
+
+		Tcl_ListObjIndex(NULL, zshPtr->inData, 0, &itemObj);
+		if (Tcl_IsShared(itemObj)) {
+		    itemObj = Tcl_DuplicateObj(itemObj);
+		}
+		itemPtr = Tcl_GetByteArrayFromObj(itemObj, &itemLen);
+		Tcl_IncrRefCount(itemObj);
+		zshPtr->currentInput = itemObj;
+		zshPtr->stream.next_in = itemPtr;
+		zshPtr->stream.avail_in = itemLen;
+
+		/*
+		 * And remove it from the list
+		 */
+
+		Tcl_ListObjReplace(NULL, zshPtr->inData, 0, 1, 0, NULL);
+	    }
+	}
+
+	/*
+	 * When dealing with a raw stream, we set the dictionary here, once.
+	 * (You can't do it in response to getting Z_NEED_DATA as raw streams
+	 * don't ever issue that.)
+	 */
+
+	if (IsRawStream(zshPtr) && HaveDictToSet(zshPtr)) {
+	    e = SetInflateDictionary(&zshPtr->stream, zshPtr->compDictObj);
+	    if (e != Z_OK) {
+		ConvertError(zshPtr->interp, e, zshPtr->stream.adler);
+		return TCL_ERROR;
+	    }
+	    DictWasSet(zshPtr);
+	}
+	e = inflate(&zshPtr->stream, zshPtr->flush);
+	if (e == Z_NEED_DICT && HaveDictToSet(zshPtr)) {
+	    e = SetInflateDictionary(&zshPtr->stream, zshPtr->compDictObj);
+	    if (e == Z_OK) {
+		DictWasSet(zshPtr);
+		e = inflate(&zshPtr->stream, zshPtr->flush);
+	    }
+	};
+	Tcl_ListObjLength(NULL, zshPtr->inData, &listLen);
+
+	while ((zshPtr->stream.avail_out > 0)
+		&& (e == Z_OK || e == Z_BUF_ERROR) && (listLen > 0)) {
+	    /*
+	     * State: We have not satisfied the request yet and there may be
+	     * more to inflate.
+	     */
+
+	    if (zshPtr->stream.avail_in > 0) {
+		if (zshPtr->interp) {
+		    Tcl_SetObjResult(zshPtr->interp, Tcl_NewStringObj(
+			    "unexpected zlib internal state during"
+			    " decompression", -1));
+		    Tcl_SetErrorCode(zshPtr->interp, "TCL", "ZIP", "STATE",
+			    NULL);
+		}
+		Tcl_SetByteArrayLength(data, existing);
+		return TCL_ERROR;
+	    }
+
+	    if (zshPtr->currentInput) {
+		Tcl_DecrRefCount(zshPtr->currentInput);
+		zshPtr->currentInput = 0;
+	    }
+
+	    /*
+	     * Get the next block of data to go to inflate. At this point, the
+	     * data must not be shared since we require the bytearray
+	     * representation to not vanish under our feet. [Bug 3081008]
+	     */
+
+	    Tcl_ListObjIndex(zshPtr->interp, zshPtr->inData, 0, &itemObj);
+	    if (Tcl_IsShared(itemObj)) {
+		itemObj = Tcl_DuplicateObj(itemObj);
+	    }
+	    itemPtr = Tcl_GetByteArrayFromObj(itemObj, &itemLen);
+	    Tcl_IncrRefCount(itemObj);
+	    zshPtr->currentInput = itemObj;
+	    zshPtr->stream.next_in = itemPtr;
+	    zshPtr->stream.avail_in = itemLen;
+
+	    /*
+	     * Remove it from the list.
+	     */
+
+	    Tcl_ListObjReplace(NULL, zshPtr->inData, 0, 1, 0, NULL);
+	    listLen--;
+
+	    /*
+	     * And call inflate again.
+	     */
+
+	    do {
+		e = inflate(&zshPtr->stream, zshPtr->flush);
+		if (e != Z_NEED_DICT || !HaveDictToSet(zshPtr)) {
+		    break;
+		}
+		e = SetInflateDictionary(&zshPtr->stream,zshPtr->compDictObj);
+		DictWasSet(zshPtr);
+	    } while (e == Z_OK);
+	}
+	if (zshPtr->stream.avail_out > 0) {
+	    Tcl_SetByteArrayLength(data,
+		    existing + count - zshPtr->stream.avail_out);
+	}
+	if (!(e==Z_OK || e==Z_STREAM_END || e==Z_BUF_ERROR)) {
+	    Tcl_SetByteArrayLength(data, existing);
+	    ConvertError(zshPtr->interp, e, zshPtr->stream.adler);
+	    return TCL_ERROR;
+	}
+	if (e == Z_STREAM_END) {
+	    zshPtr->streamEnd = 1;
+	    if (zshPtr->currentInput) {
+		Tcl_DecrRefCount(zshPtr->currentInput);
+		zshPtr->currentInput = 0;
+	    }
+	    inflateEnd(&zshPtr->stream);
+	}
+    } else {
+	Tcl_ListObjLength(NULL, zshPtr->outData, &listLen);
+	if (count == -1) {
+	    count = 0;
+	    for (i=0; i<listLen; i++) {
+		Tcl_ListObjIndex(NULL, zshPtr->outData, i, &itemObj);
+		itemPtr = Tcl_GetByteArrayFromObj(itemObj, &itemLen);
+		if (i == 0) {
+		    count += itemLen - zshPtr->outPos;
+		} else {
+		    count += itemLen;
+		}
+	    }
+	}
+
+	/*
+	 * Prepare the place to store the data.
+	 */
+
+	dataPtr = Tcl_SetByteArrayLength(data, existing + count);
+	dataPtr += existing;
+
+	while ((count > dataPos) &&
+		(Tcl_ListObjLength(NULL, zshPtr->outData, &listLen) == TCL_OK)
+		&& (listLen > 0)) {
+	    /*
+	     * Get the next chunk off our list of chunks and grab the data out
+	     * of it.
+	     */
+
+	    Tcl_ListObjIndex(NULL, zshPtr->outData, 0, &itemObj);
+	    itemPtr = Tcl_GetByteArrayFromObj(itemObj, &itemLen);
+	    if (itemLen-zshPtr->outPos >= count-dataPos) {
+		unsigned len = count - dataPos;
+
+		memcpy(dataPtr + dataPos, itemPtr + zshPtr->outPos, len);
+		zshPtr->outPos += len;
+		dataPos += len;
+		if (zshPtr->outPos == itemLen) {
+		    zshPtr->outPos = 0;
+		}
+	    } else {
+		unsigned len = itemLen - zshPtr->outPos;
+
+		memcpy(dataPtr + dataPos, itemPtr + zshPtr->outPos, len);
+		dataPos += len;
+		zshPtr->outPos = 0;
+	    }
+	    if (zshPtr->outPos == 0) {
+		Tcl_ListObjReplace(NULL, zshPtr->outData, 0, 1, 0, NULL);
+		listLen--;
+	    }
+	}
+	Tcl_SetByteArrayLength(data, existing + dataPos);
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibDeflate --
+ *
+ *	Compress the contents of Tcl_Obj *data with compression level in
+ *	output format, producing the compressed data in the interpreter
+ *	result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibDeflate(
+    Tcl_Interp *interp,
+    int format,
+    Tcl_Obj *data,
+    int level,
+    Tcl_Obj *gzipHeaderDictObj)
+{
+    int wbits = 0, inLen = 0, e = 0, extraSize = 0;
+    Byte *inData = NULL;
+    z_stream stream;
+    GzipHeader header;
+    gz_header *headerPtr = NULL;
+    Tcl_Obj *obj;
+
+    if (!interp) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Compressed format is specified by the wbits parameter. See zlib.h for
+     * details.
+     */
+
+    if (format == TCL_ZLIB_FORMAT_RAW) {
+	wbits = WBITS_RAW;
+    } else if (format == TCL_ZLIB_FORMAT_GZIP) {
+	wbits = WBITS_GZIP;
+
+	/*
+	 * Need to allocate extra space for the gzip header and footer. The
+	 * amount of space is (a bit less than) 32 bytes, plus a byte for each
+	 * byte of string that we add. Note that over-allocation is not a
+	 * problem. [Bug 2419061]
+	 */
+
+	extraSize = 32;
+	if (gzipHeaderDictObj) {
+	    headerPtr = &header.header;
+	    memset(headerPtr, 0, sizeof(gz_header));
+	    if (GenerateHeader(interp, gzipHeaderDictObj, &header,
+		    &extraSize) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+    } else if (format == TCL_ZLIB_FORMAT_ZLIB) {
+	wbits = WBITS_ZLIB;
+    } else {
+	Tcl_Panic("incorrect zlib data format, must be TCL_ZLIB_FORMAT_ZLIB, "
+		"TCL_ZLIB_FORMAT_GZIP or TCL_ZLIB_FORMAT_ZLIB");
+    }
+
+    if (level < -1 || level > 9) {
+	Tcl_Panic("compression level should be between 0 (uncompressed) and "
+		"9 (best compression) or -1 for default compression level");
+    }
+
+    /*
+     * Allocate some space to store the output.
+     */
+
+    TclNewObj(obj);
+
+    /*
+     * Obtain the pointer to the byte array, we'll pass this pointer straight
+     * to the deflate command.
+     */
+
+    inData = Tcl_GetByteArrayFromObj(data, &inLen);
+    memset(&stream, 0, sizeof(z_stream));
+    stream.avail_in = (uInt) inLen;
+    stream.next_in = inData;
+
+    /*
+     * No output buffer available yet, will alloc after deflateInit2.
+     */
+
+    e = deflateInit2(&stream, level, Z_DEFLATED, wbits, MAX_MEM_LEVEL,
+	    Z_DEFAULT_STRATEGY);
+    if (e != Z_OK) {
+	goto error;
+    }
+
+    if (headerPtr != NULL) {
+	e = deflateSetHeader(&stream, headerPtr);
+	if (e != Z_OK) {
+	    goto error;
+	}
+    }
+
+    /*
+     * Allocate the output buffer from the value of deflateBound(). This is
+     * probably too much space. Before returning to the caller, we will reduce
+     * it back to the actual compressed size.
+     */
+
+    stream.avail_out = deflateBound(&stream, inLen) + extraSize;
+    stream.next_out = Tcl_SetByteArrayLength(obj, stream.avail_out);
+
+    /*
+     * Perform the compression, Z_FINISH means do it in one go.
+     */
+
+    e = deflate(&stream, Z_FINISH);
+
+    if (e != Z_STREAM_END) {
+	e = deflateEnd(&stream);
+
+	/*
+	 * deflateEnd() returns Z_OK when there are bytes left to compress, at
+	 * this point we consider that an error, although we could continue by
+	 * allocating more memory and calling deflate() again.
+	 */
+
+	if (e == Z_OK) {
+	    e = Z_BUF_ERROR;
+	}
+    } else {
+	e = deflateEnd(&stream);
+    }
+
+    if (e != Z_OK) {
+	goto error;
+    }
+
+    /*
+     * Reduce the bytearray length to the actual data length produced by
+     * deflate.
+     */
+
+    Tcl_SetByteArrayLength(obj, stream.total_out);
+    Tcl_SetObjResult(interp, obj);
+    return TCL_OK;
+
+  error:
+    ConvertError(interp, e, stream.adler);
+    TclDecrRefCount(obj);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibInflate --
+ *
+ *	Decompress data in an object into the interpreter result.
+ *
+ *----------------------------------------------------------------------
+ */
+
+int
+Tcl_ZlibInflate(
+    Tcl_Interp *interp,
+    int format,
+    Tcl_Obj *data,
+    int bufferSize,
+    Tcl_Obj *gzipHeaderDictObj)
+{
+    int wbits = 0, inLen = 0, e = 0, newBufferSize;
+    Byte *inData = NULL, *outData = NULL, *newOutData = NULL;
+    z_stream stream;
+    gz_header header, *headerPtr = NULL;
+    Tcl_Obj *obj;
+    char *nameBuf = NULL, *commentBuf = NULL;
+
+    if (!interp) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Compressed format is specified by the wbits parameter. See zlib.h for
+     * details.
+     */
+
+    switch (format) {
+    case TCL_ZLIB_FORMAT_RAW:
+	wbits = WBITS_RAW;
+	gzipHeaderDictObj = NULL;
+	break;
+    case TCL_ZLIB_FORMAT_ZLIB:
+	wbits = WBITS_ZLIB;
+	gzipHeaderDictObj = NULL;
+	break;
+    case TCL_ZLIB_FORMAT_GZIP:
+	wbits = WBITS_GZIP;
+	break;
+    case TCL_ZLIB_FORMAT_AUTO:
+	wbits = WBITS_AUTODETECT;
+	break;
+    default:
+	Tcl_Panic("incorrect zlib data format, must be TCL_ZLIB_FORMAT_ZLIB, "
+		"TCL_ZLIB_FORMAT_GZIP, TCL_ZLIB_FORMAT_RAW or "
+		"TCL_ZLIB_FORMAT_AUTO");
+    }
+
+    if (gzipHeaderDictObj) {
+	headerPtr = &header;
+	memset(headerPtr, 0, sizeof(gz_header));
+	nameBuf = ckalloc(MAXPATHLEN);
+	header.name = (Bytef *) nameBuf;
+	header.name_max = MAXPATHLEN - 1;
+	commentBuf = ckalloc(MAX_COMMENT_LEN);
+	header.comment = (Bytef *) commentBuf;
+	header.comm_max = MAX_COMMENT_LEN - 1;
+    }
+
+    inData = Tcl_GetByteArrayFromObj(data, &inLen);
+    if (bufferSize < 1) {
+	/*
+	 * Start with a buffer (up to) 3 times the size of the input data.
+	 */
+
+	if (inLen < 32*1024*1024) {
+	    bufferSize = 3*inLen;
+	} else if (inLen < 256*1024*1024) {
+	    bufferSize = 2*inLen;
+	} else {
+	    bufferSize = inLen;
+	}
+    }
+
+    TclNewObj(obj);
+    outData = Tcl_SetByteArrayLength(obj, bufferSize);
+    memset(&stream, 0, sizeof(z_stream));
+    stream.avail_in = (uInt) inLen+1;	/* +1 because zlib can "over-request"
+					 * input (but ignore it!) */
+    stream.next_in = inData;
+    stream.avail_out = bufferSize;
+    stream.next_out = outData;
+
+    /*
+     * Initialize zlib for decompression.
+     */
+
+    e = inflateInit2(&stream, wbits);
+    if (e != Z_OK) {
+	goto error;
+    }
+    if (headerPtr) {
+	e = inflateGetHeader(&stream, headerPtr);
+	if (e != Z_OK) {
+	    inflateEnd(&stream);
+	    goto error;
+	}
+    }
+
+    /*
+     * Start the decompression cycle.
+     */
+
+    while (1) {
+	e = inflate(&stream, Z_FINISH);
+	if (e != Z_BUF_ERROR) {
+	    break;
+	}
+
+	/*
+	 * Not enough room in the output buffer. Increase it by five times the
+	 * bytes still in the input buffer. (Because 3 times didn't do the
+	 * trick before, 5 times is what we do next.) Further optimization
+	 * should be done by the user, specify the decompressed size!
+	 */
+
+	if ((stream.avail_in == 0) && (stream.avail_out > 0)) {
+	    e = Z_STREAM_ERROR;
+	    break;
+	}
+	newBufferSize = bufferSize + 5 * stream.avail_in;
+	if (newBufferSize == bufferSize) {
+	    newBufferSize = bufferSize+1000;
+	}
+	newOutData = Tcl_SetByteArrayLength(obj, newBufferSize);
+
+	/*
+	 * Set next out to the same offset in the new location.
+	 */
+
+	stream.next_out = newOutData + stream.total_out;
+
+	/*
+	 * And increase avail_out with the number of new bytes allocated.
+	 */
+
+	stream.avail_out += newBufferSize - bufferSize;
+	outData = newOutData;
+	bufferSize = newBufferSize;
+    }
+
+    if (e != Z_STREAM_END) {
+	inflateEnd(&stream);
+	goto error;
+    }
+
+    e = inflateEnd(&stream);
+    if (e != Z_OK) {
+	goto error;
+    }
+
+    /*
+     * Reduce the BA length to the actual data length produced by deflate.
+     */
+
+    Tcl_SetByteArrayLength(obj, stream.total_out);
+    if (headerPtr != NULL) {
+	ExtractHeader(&header, gzipHeaderDictObj);
+	SetValue(gzipHeaderDictObj, "size",
+		Tcl_NewLongObj((long) stream.total_out));
+	ckfree(nameBuf);
+	ckfree(commentBuf);
+    }
+    Tcl_SetObjResult(interp, obj);
+    return TCL_OK;
+
+  error:
+    TclDecrRefCount(obj);
+    ConvertError(interp, e, stream.adler);
+    if (nameBuf) {
+	ckfree(nameBuf);
+    }
+    if (commentBuf) {
+	ckfree(commentBuf);
+    }
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * Tcl_ZlibCRC32, Tcl_ZlibAdler32 --
+ *
+ *	Access to the checksumming engines.
+ *
+ *----------------------------------------------------------------------
+ */
+
+unsigned int
+Tcl_ZlibCRC32(
+    unsigned int crc,
+    const unsigned char *buf,
+    int len)
+{
+    /* Nothing much to do, just wrap the crc32(). */
+    return crc32(crc, (Bytef *) buf, (unsigned) len);
+}
+
+unsigned int
+Tcl_ZlibAdler32(
+    unsigned int adler,
+    const unsigned char *buf,
+    int len)
+{
+    return adler32(adler, (Bytef *) buf, (unsigned) len);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibCmd --
+ *
+ *	Implementation of the [zlib] command.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibCmd(
+    ClientData notUsed,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    int command, dlen, i, option, level = -1;
+    unsigned start, buffersize = 0;
+    Byte *data;
+    Tcl_Obj *headerDictObj;
+    const char *extraInfoStr = NULL;
+    static const char *const commands[] = {
+	"adler32", "compress", "crc32", "decompress", "deflate", "gunzip",
+	"gzip", "inflate", "push", "stream",
+	NULL
+    };
+    enum zlibCommands {
+	CMD_ADLER, CMD_COMPRESS, CMD_CRC, CMD_DECOMPRESS, CMD_DEFLATE,
+	CMD_GUNZIP, CMD_GZIP, CMD_INFLATE, CMD_PUSH, CMD_STREAM
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "command arg ?...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[1], commands, "command", 0,
+	    &command) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum zlibCommands) command) {
+    case CMD_ADLER:			/* adler32 str ?startvalue?
+					 * -> checksum */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?startValue?");
+	    return TCL_ERROR;
+	}
+	if (objc>3 && Tcl_GetIntFromObj(interp, objv[3],
+		(int *) &start) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (objc < 4) {
+	    start = Tcl_ZlibAdler32(0, NULL, 0);
+	}
+	data = Tcl_GetByteArrayFromObj(objv[2], &dlen);
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt)
+		(uLong) Tcl_ZlibAdler32(start, data, dlen)));
+	return TCL_OK;
+    case CMD_CRC:			/* crc32 str ?startvalue?
+					 * -> checksum */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?startValue?");
+	    return TCL_ERROR;
+	}
+	if (objc>3 && Tcl_GetIntFromObj(interp, objv[3],
+		(int *) &start) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (objc < 4) {
+	    start = Tcl_ZlibCRC32(0, NULL, 0);
+	}
+	data = Tcl_GetByteArrayFromObj(objv[2], &dlen);
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt)
+		(uLong) Tcl_ZlibCRC32(start, data, dlen)));
+	return TCL_OK;
+    case CMD_DEFLATE:			/* deflate data ?level?
+					 * -> rawCompressedData */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?level?");
+	    return TCL_ERROR;
+	}
+	if (objc > 3) {
+	    if (Tcl_GetIntFromObj(interp, objv[3], &level) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (level < 0 || level > 9) {
+		goto badLevel;
+	    }
+	}
+	return Tcl_ZlibDeflate(interp, TCL_ZLIB_FORMAT_RAW, objv[2], level,
+		NULL);
+    case CMD_COMPRESS:			/* compress data ?level?
+					 * -> zlibCompressedData */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?level?");
+	    return TCL_ERROR;
+	}
+	if (objc > 3) {
+	    if (Tcl_GetIntFromObj(interp, objv[3], &level) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (level < 0 || level > 9) {
+		goto badLevel;
+	    }
+	}
+	return Tcl_ZlibDeflate(interp, TCL_ZLIB_FORMAT_ZLIB, objv[2], level,
+		NULL);
+    case CMD_GZIP:			/* gzip data ?level?
+					 * -> gzippedCompressedData */
+	headerDictObj = NULL;
+
+	/*
+	 * Legacy argument format support.
+	 */
+
+	if (objc == 4
+		&& Tcl_GetIntFromObj(interp, objv[3], &level) == TCL_OK) {
+	    if (level < 0 || level > 9) {
+		extraInfoStr = "\n    (in -level option)";
+		goto badLevel;
+	    }
+	    return Tcl_ZlibDeflate(interp, TCL_ZLIB_FORMAT_GZIP, objv[2],
+		    level, NULL);
+	}
+
+	if (objc < 3 || objc > 7 || ((objc & 1) == 0)) {
+	    Tcl_WrongNumArgs(interp, 2, objv,
+		    "data ?-level level? ?-header header?");
+	    return TCL_ERROR;
+	}
+	for (i=3 ; i<objc ; i+=2) {
+	    static const char *const gzipopts[] = {
+		"-header", "-level", NULL
+	    };
+
+	    if (Tcl_GetIndexFromObj(interp, objv[i], gzipopts, "option", 0,
+		    &option) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch (option) {
+	    case 0:
+		headerDictObj = objv[i+1];
+		break;
+	    case 1:
+		if (Tcl_GetIntFromObj(interp, objv[i+1],
+			&level) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (level < 0 || level > 9) {
+		    extraInfoStr = "\n    (in -level option)";
+		    goto badLevel;
+		}
+		break;
+	    }
+	}
+	return Tcl_ZlibDeflate(interp, TCL_ZLIB_FORMAT_GZIP, objv[2], level,
+		headerDictObj);
+    case CMD_INFLATE:			/* inflate rawcomprdata ?bufferSize?
+					 *	-> decompressedData */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?bufferSize?");
+	    return TCL_ERROR;
+	}
+	if (objc > 3) {
+	    if (Tcl_GetIntFromObj(interp, objv[3],
+		    (int *) &buffersize) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (buffersize < MIN_NONSTREAM_BUFFER_SIZE
+		    || buffersize > MAX_BUFFER_SIZE) {
+		goto badBuffer;
+	    }
+	}
+	return Tcl_ZlibInflate(interp, TCL_ZLIB_FORMAT_RAW, objv[2],
+		buffersize, NULL);
+    case CMD_DECOMPRESS:		/* decompress zlibcomprdata \
+					 *    ?bufferSize?
+					 *	-> decompressedData */
+	if (objc < 3 || objc > 4) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?bufferSize?");
+	    return TCL_ERROR;
+	}
+	if (objc > 3) {
+	    if (Tcl_GetIntFromObj(interp, objv[3],
+		    (int *) &buffersize) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (buffersize < MIN_NONSTREAM_BUFFER_SIZE
+		    || buffersize > MAX_BUFFER_SIZE) {
+		goto badBuffer;
+	    }
+	}
+	return Tcl_ZlibInflate(interp, TCL_ZLIB_FORMAT_ZLIB, objv[2],
+		buffersize, NULL);
+    case CMD_GUNZIP: {			/* gunzip gzippeddata ?bufferSize?
+					 *	-> decompressedData */
+	Tcl_Obj *headerVarObj;
+
+	if (objc < 3 || objc > 5 || ((objc & 1) == 0)) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "data ?-headerVar varName?");
+	    return TCL_ERROR;
+	}
+	headerDictObj = headerVarObj = NULL;
+	for (i=3 ; i<objc ; i+=2) {
+	    static const char *const gunzipopts[] = {
+		"-buffersize", "-headerVar", NULL
+	    };
+
+	    if (Tcl_GetIndexFromObj(interp, objv[i], gunzipopts, "option", 0,
+		    &option) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    switch (option) {
+	    case 0:
+		if (Tcl_GetIntFromObj(interp, objv[i+1],
+			(int *) &buffersize) != TCL_OK) {
+		    return TCL_ERROR;
+		}
+		if (buffersize < MIN_NONSTREAM_BUFFER_SIZE
+			|| buffersize > MAX_BUFFER_SIZE) {
+		    goto badBuffer;
+		}
+		break;
+	    case 1:
+		headerVarObj = objv[i+1];
+		headerDictObj = Tcl_NewObj();
+		break;
+	    }
+	}
+	if (Tcl_ZlibInflate(interp, TCL_ZLIB_FORMAT_GZIP, objv[2],
+		buffersize, headerDictObj) != TCL_OK) {
+	    if (headerDictObj) {
+		TclDecrRefCount(headerDictObj);
+	    }
+	    return TCL_ERROR;
+	}
+	if (headerVarObj != NULL && Tcl_ObjSetVar2(interp, headerVarObj, NULL,
+		headerDictObj, TCL_LEAVE_ERR_MSG) == NULL) {
+	    return TCL_ERROR;
+	}
+	return TCL_OK;
+    }
+    case CMD_STREAM:			/* stream deflate/inflate/...gunzip \
+					 *    ?options...?
+					 *	-> handleCmd */
+	return ZlibStreamSubcmd(interp, objc, objv);
+    case CMD_PUSH:			/* push mode channel options...
+					 *	-> channel */
+	return ZlibPushSubcmd(interp, objc, objv);
+    };
+
+    return TCL_ERROR;
+
+  badLevel:
+    Tcl_SetObjResult(interp, Tcl_NewStringObj("level must be 0 to 9", -1));
+    Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMPRESSIONLEVEL", NULL);
+    if (extraInfoStr) {
+	Tcl_AddErrorInfo(interp, extraInfoStr);
+    }
+    return TCL_ERROR;
+  badBuffer:
+    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+	    "buffer size must be %d to %d",
+	    MIN_NONSTREAM_BUFFER_SIZE, MAX_BUFFER_SIZE));
+    Tcl_SetErrorCode(interp, "TCL", "VALUE", "BUFFERSIZE", NULL);
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibStreamSubcmd --
+ *
+ *	Implementation of the [zlib stream] subcommand.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibStreamSubcmd(
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const char *const stream_formats[] = {
+	"compress", "decompress", "deflate", "gunzip", "gzip", "inflate",
+	NULL
+    };
+    enum zlibFormats {
+	FMT_COMPRESS, FMT_DECOMPRESS, FMT_DEFLATE, FMT_GUNZIP, FMT_GZIP,
+	FMT_INFLATE
+    };
+    int i, format, mode = 0, option, level;
+    enum objIndices {
+	OPT_COMPRESSION_DICTIONARY = 0,
+	OPT_GZIP_HEADER = 1,
+	OPT_COMPRESSION_LEVEL = 2,
+	OPT_END = -1
+    };
+    Tcl_Obj *obj[3] = { NULL, NULL, NULL };
+#define compDictObj	obj[OPT_COMPRESSION_DICTIONARY]
+#define gzipHeaderObj	obj[OPT_GZIP_HEADER]
+#define levelObj	obj[OPT_COMPRESSION_LEVEL]
+    typedef struct {
+	const char *name;
+	enum objIndices offset;
+    } OptDescriptor;
+    static const OptDescriptor compressionOpts[] = {
+	{ "-dictionary", OPT_COMPRESSION_DICTIONARY },
+	{ "-level",	 OPT_COMPRESSION_LEVEL },
+	{ NULL, OPT_END }
+    };
+    static const OptDescriptor gzipOpts[] = {
+	{ "-header",	 OPT_GZIP_HEADER },
+	{ "-level",	 OPT_COMPRESSION_LEVEL },
+	{ NULL, OPT_END }
+    };
+    static const OptDescriptor expansionOpts[] = {
+	{ "-dictionary", OPT_COMPRESSION_DICTIONARY },
+	{ NULL, OPT_END }
+    };
+    static const OptDescriptor gunzipOpts[] = {
+	{ NULL, OPT_END }
+    };
+    const OptDescriptor *desc = NULL;
+    Tcl_ZlibStream zh;
+
+    if (objc < 3 || !(objc & 1)) {
+	Tcl_WrongNumArgs(interp, 2, objv, "mode ?-option value...?");
+	return TCL_ERROR;
+    }
+    if (Tcl_GetIndexFromObj(interp, objv[2], stream_formats, "mode", 0,
+	    &format) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * The format determines the compression mode and the options that may be
+     * specified.
+     */
+
+    switch ((enum zlibFormats) format) {
+    case FMT_DEFLATE:
+	desc = compressionOpts;
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_RAW;
+	break;
+    case FMT_INFLATE:
+	desc = expansionOpts;
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_RAW;
+	break;
+    case FMT_COMPRESS:
+	desc = compressionOpts;
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_ZLIB;
+	break;
+    case FMT_DECOMPRESS:
+	desc = expansionOpts;
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_ZLIB;
+	break;
+    case FMT_GZIP:
+	desc = gzipOpts;
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_GZIP;
+	break;
+    case FMT_GUNZIP:
+	desc = gunzipOpts;
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_GZIP;
+	break;
+    default:
+	Tcl_Panic("should be unreachable");
+    }
+
+    /*
+     * Parse the options.
+     */
+
+    for (i=3 ; i<objc ; i+=2) {
+	if (Tcl_GetIndexFromObjStruct(interp, objv[i], desc,
+		sizeof(OptDescriptor), "option", 0, &option) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	obj[desc[option].offset] = objv[i+1];
+    }
+
+    /*
+     * If a compression level was given, parse it (integral: 0..9). Otherwise
+     * use the default.
+     */
+
+    if (levelObj == NULL) {
+	level = Z_DEFAULT_COMPRESSION;
+    } else if (Tcl_GetIntFromObj(interp, levelObj, &level) != TCL_OK) {
+	return TCL_ERROR;
+    } else if (level < 0 || level > 9) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("level must be 0 to 9",-1));
+	Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMPRESSIONLEVEL", NULL);
+	Tcl_AddErrorInfo(interp, "\n    (in -level option)");
+	return TCL_ERROR;
+    }
+
+    /*
+     * Construct the stream now we know its configuration.
+     */
+
+    if (Tcl_ZlibStreamInit(interp, mode, format, level, gzipHeaderObj,
+	    &zh) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    if (compDictObj != NULL) {
+	Tcl_ZlibStreamSetCompressionDictionary(zh, compDictObj);
+    }
+    Tcl_SetObjResult(interp, Tcl_ZlibStreamGetCommandName(zh));
+    return TCL_OK;
+#undef compDictObj
+#undef gzipHeaderObj
+#undef levelObj
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibPushSubcmd --
+ *
+ *	Implementation of the [zlib push] subcommand.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibPushSubcmd(
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    static const char *const stream_formats[] = {
+	"compress", "decompress", "deflate", "gunzip", "gzip", "inflate",
+	NULL
+    };
+    enum zlibFormats {
+	FMT_COMPRESS, FMT_DECOMPRESS, FMT_DEFLATE, FMT_GUNZIP, FMT_GZIP,
+	FMT_INFLATE
+    };
+    Tcl_Channel chan;
+    int chanMode, format, mode = 0, level, i, option;
+    static const char *const pushCompressOptions[] = {
+	"-dictionary", "-header", "-level", NULL
+    };
+    static const char *const pushDecompressOptions[] = {
+	"-dictionary", "-header", "-level", "-limit", NULL
+    };
+    const char *const *pushOptions = pushDecompressOptions;
+    enum pushOptions {poDictionary, poHeader, poLevel, poLimit};
+    Tcl_Obj *headerObj = NULL, *compDictObj = NULL;
+    int limit = 1, dummy;
+
+    if (objc < 4) {
+	Tcl_WrongNumArgs(interp, 2, objv, "mode channel ?options...?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[2], stream_formats, "mode", 0,
+	    &format) != TCL_OK) {
+	return TCL_ERROR;
+    }
+    switch ((enum zlibFormats) format) {
+    case FMT_DEFLATE:
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_RAW;
+	pushOptions = pushCompressOptions;
+	break;
+    case FMT_INFLATE:
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_RAW;
+	break;
+    case FMT_COMPRESS:
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_ZLIB;
+	pushOptions = pushCompressOptions;
+	break;
+    case FMT_DECOMPRESS:
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_ZLIB;
+	break;
+    case FMT_GZIP:
+	mode = TCL_ZLIB_STREAM_DEFLATE;
+	format = TCL_ZLIB_FORMAT_GZIP;
+	pushOptions = pushCompressOptions;
+	break;
+    case FMT_GUNZIP:
+	mode = TCL_ZLIB_STREAM_INFLATE;
+	format = TCL_ZLIB_FORMAT_GZIP;
+	break;
+    default:
+	Tcl_Panic("should be unreachable");
+    }
+
+    if (TclGetChannelFromObj(interp, objv[3], &chan, &chanMode, 0) != TCL_OK){
+	return TCL_ERROR;
+    }
+
+    /*
+     * Sanity checks.
+     */
+
+    if (mode == TCL_ZLIB_STREAM_DEFLATE && !(chanMode & TCL_WRITABLE)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"compression may only be applied to writable channels", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ZIP", "UNWRITABLE", NULL);
+	return TCL_ERROR;
+    }
+    if (mode == TCL_ZLIB_STREAM_INFLATE && !(chanMode & TCL_READABLE)) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"decompression may only be applied to readable channels",-1));
+	Tcl_SetErrorCode(interp, "TCL", "ZIP", "UNREADABLE", NULL);
+	return TCL_ERROR;
+    }
+
+    /*
+     * Parse options.
+     */
+
+    level = Z_DEFAULT_COMPRESSION;
+    for (i=4 ; i<objc ; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], pushOptions, "option", 0,
+		&option) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+	if (++i > objc-1) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "value missing for %s option", pushOptions[option]));
+	    Tcl_SetErrorCode(interp, "TCL", "ZIP", "NOVAL", NULL);
+	    return TCL_ERROR;
+	}
+	switch ((enum pushOptions) option) {
+	case poHeader:
+	    headerObj = objv[i];
+	    if (Tcl_DictObjSize(interp, headerObj, &dummy) != TCL_OK) {
+		goto genericOptionError;
+	    }
+	    break;
+	case poLevel:
+	    if (Tcl_GetIntFromObj(interp, objv[i], (int*) &level) != TCL_OK) {
+		goto genericOptionError;
+	    }
+	    if (level < 0 || level > 9) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"level must be 0 to 9", -1));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "COMPRESSIONLEVEL",
+			NULL);
+		goto genericOptionError;
+	    }
+	    break;
+	case poLimit:
+	    if (Tcl_GetIntFromObj(interp, objv[i], (int*) &limit) != TCL_OK) {
+		goto genericOptionError;
+	    }
+	    if (limit < 1 || limit > MAX_BUFFER_SIZE) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"read ahead limit must be 1 to %d",
+			MAX_BUFFER_SIZE));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "BUFFERSIZE", NULL);
+		goto genericOptionError;
+	    }
+	    break;
+	case poDictionary:
+	    if (format == TCL_ZLIB_FORMAT_GZIP) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"a compression dictionary may not be set in the "
+			"gzip format", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ZIP", "BADOPT", NULL);
+		goto genericOptionError;
+	    }
+	    compDictObj = objv[i];
+	    break;
+	}
+    }
+
+    if (ZlibStackChannelTransform(interp, mode, format, level, limit, chan,
+	    headerObj, compDictObj) == NULL) {
+	return TCL_ERROR;
+    }
+    Tcl_SetObjResult(interp, objv[3]);
+    return TCL_OK;
+
+  genericOptionError:
+    Tcl_AddErrorInfo(interp, "\n    (in ");
+    Tcl_AddErrorInfo(interp, pushOptions[option]);
+    Tcl_AddErrorInfo(interp, " option)");
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibStreamCmd --
+ *
+ *	Implementation of the commands returned by [zlib stream].
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibStreamCmd(
+    ClientData cd,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_ZlibStream zstream = cd;
+    int command, count, code;
+    Tcl_Obj *obj;
+    static const char *const cmds[] = {
+	"add", "checksum", "close", "eof", "finalize", "flush",
+	"fullflush", "get", "header", "put", "reset",
+	NULL
+    };
+    enum zlibStreamCommands {
+	zs_add, zs_checksum, zs_close, zs_eof, zs_finalize, zs_flush,
+	zs_fullflush, zs_get, zs_header, zs_put, zs_reset
+    };
+
+    if (objc < 2) {
+	Tcl_WrongNumArgs(interp, 1, objv, "option data ?...?");
+	return TCL_ERROR;
+    }
+
+    if (Tcl_GetIndexFromObj(interp, objv[1], cmds, "option", 0,
+	    &command) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    switch ((enum zlibStreamCommands) command) {
+    case zs_add:		/* $strm add ?$flushopt? $data */
+	return ZlibStreamAddCmd(zstream, interp, objc, objv);
+    case zs_header:		/* $strm header */
+	return ZlibStreamHeaderCmd(zstream, interp, objc, objv);
+    case zs_put:		/* $strm put ?$flushopt? $data */
+	return ZlibStreamPutCmd(zstream, interp, objc, objv);
+
+    case zs_get:		/* $strm get ?count? */
+	if (objc > 3) {
+	    Tcl_WrongNumArgs(interp, 2, objv, "?count?");
+	    return TCL_ERROR;
+	}
+
+	count = -1;
+	if (objc >= 3) {
+	    if (Tcl_GetIntFromObj(interp, objv[2], &count) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	}
+	TclNewObj(obj);
+	code = Tcl_ZlibStreamGet(zstream, obj, count);
+	if (code == TCL_OK) {
+	    Tcl_SetObjResult(interp, obj);
+	} else {
+	    TclDecrRefCount(obj);
+	}
+	return code;
+    case zs_flush:		/* $strm flush */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	TclNewObj(obj);
+	Tcl_IncrRefCount(obj);
+	code = Tcl_ZlibStreamPut(zstream, obj, Z_SYNC_FLUSH);
+	TclDecrRefCount(obj);
+	return code;
+    case zs_fullflush:		/* $strm fullflush */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	TclNewObj(obj);
+	Tcl_IncrRefCount(obj);
+	code = Tcl_ZlibStreamPut(zstream, obj, Z_FULL_FLUSH);
+	TclDecrRefCount(obj);
+	return code;
+    case zs_finalize:		/* $strm finalize */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * The flush commands slightly abuse the empty result obj as input
+	 * data.
+	 */
+
+	TclNewObj(obj);
+	Tcl_IncrRefCount(obj);
+	code = Tcl_ZlibStreamPut(zstream, obj, Z_FINISH);
+	TclDecrRefCount(obj);
+	return code;
+    case zs_close:		/* $strm close */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return Tcl_ZlibStreamClose(zstream);
+    case zs_eof:		/* $strm eof */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewIntObj(Tcl_ZlibStreamEof(zstream)));
+	return TCL_OK;
+    case zs_checksum:		/* $strm checksum */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	Tcl_SetObjResult(interp, Tcl_NewWideIntObj((Tcl_WideInt)
+		(uLong) Tcl_ZlibStreamChecksum(zstream)));
+	return TCL_OK;
+    case zs_reset:		/* $strm reset */
+	if (objc != 2) {
+	    Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	    return TCL_ERROR;
+	}
+	return Tcl_ZlibStreamReset(zstream);
+    }
+
+    return TCL_OK;
+}
+
+static int
+ZlibStreamAddCmd(
+    ClientData cd,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_ZlibStream zstream = cd;
+    int index, code, buffersize = -1, flush = -1, i;
+    Tcl_Obj *obj, *compDictObj = NULL;
+    static const char *const add_options[] = {
+	"-buffer", "-dictionary", "-finalize", "-flush", "-fullflush", NULL
+    };
+    enum addOptions {
+	ao_buffer, ao_dictionary, ao_finalize, ao_flush, ao_fullflush
+    };
+
+    for (i=2; i<objc-1; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], add_options, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	switch ((enum addOptions) index) {
+	case ao_flush: /* -flush */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_SYNC_FLUSH;
+	    }
+	    break;
+	case ao_fullflush: /* -fullflush */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_FULL_FLUSH;
+	    }
+	    break;
+	case ao_finalize: /* -finalize */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_FINISH;
+	    }
+	    break;
+	case ao_buffer: /* -buffer */
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-buffer\" option must be followed by integer "
+			"decompression buffersize", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ZIP", "NOVAL", NULL);
+		return TCL_ERROR;
+	    }
+	    if (Tcl_GetIntFromObj(interp, objv[++i], &buffersize) != TCL_OK) {
+		return TCL_ERROR;
+	    }
+	    if (buffersize < 1 || buffersize > MAX_BUFFER_SIZE) {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"buffer size must be 1 to %d",
+			MAX_BUFFER_SIZE));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "BUFFERSIZE", NULL);
+		return TCL_ERROR;
+	    }
+	    break;
+	case ao_dictionary:
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-dictionary\" option must be followed by"
+			" compression dictionary bytes", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ZIP", "NOVAL", NULL);
+		return TCL_ERROR;
+	    }
+	    compDictObj = objv[++i];
+	    break;
+	}
+
+	if (flush == -2) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "\"-flush\", \"-fullflush\" and \"-finalize\" options"
+		    " are mutually exclusive", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ZIP", "EXCLUSIVE", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    if (flush == -1) {
+	flush = 0;
+    }
+
+    /*
+     * Set the compression dictionary if requested.
+     */
+
+    if (compDictObj != NULL) {
+	int len;
+
+	(void) Tcl_GetByteArrayFromObj(compDictObj, &len);
+	if (len == 0) {
+	    compDictObj = NULL;
+	}
+	Tcl_ZlibStreamSetCompressionDictionary(zstream, compDictObj);
+    }
+
+    /*
+     * Send the data to the stream core, along with any flushing directive.
+     */
+
+    if (Tcl_ZlibStreamPut(zstream, objv[objc-1], flush) != TCL_OK) {
+	return TCL_ERROR;
+    }
+
+    /*
+     * Get such data out as we can (up to the requested length).
+     */
+
+    TclNewObj(obj);
+    code = Tcl_ZlibStreamGet(zstream, obj, buffersize);
+    if (code == TCL_OK) {
+	Tcl_SetObjResult(interp, obj);
+    } else {
+	TclDecrRefCount(obj);
+    }
+    return code;
+}
+
+static int
+ZlibStreamPutCmd(
+    ClientData cd,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    Tcl_ZlibStream zstream = cd;
+    int index, flush = -1, i;
+    Tcl_Obj *compDictObj = NULL;
+    static const char *const put_options[] = {
+	"-dictionary", "-finalize", "-flush", "-fullflush", NULL
+    };
+    enum putOptions {
+	po_dictionary, po_finalize, po_flush, po_fullflush
+    };
+
+    for (i=2; i<objc-1; i++) {
+	if (Tcl_GetIndexFromObj(interp, objv[i], put_options, "option", 0,
+		&index) != TCL_OK) {
+	    return TCL_ERROR;
+	}
+
+	switch ((enum putOptions) index) {
+	case po_flush: /* -flush */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_SYNC_FLUSH;
+	    }
+	    break;
+	case po_fullflush: /* -fullflush */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_FULL_FLUSH;
+	    }
+	    break;
+	case po_finalize: /* -finalize */
+	    if (flush > -1) {
+		flush = -2;
+	    } else {
+		flush = Z_FINISH;
+	    }
+	    break;
+	case po_dictionary:
+	    if (i == objc-2) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"\"-dictionary\" option must be followed by"
+			" compression dictionary bytes", -1));
+		Tcl_SetErrorCode(interp, "TCL", "ZIP", "NOVAL", NULL);
+		return TCL_ERROR;
+	    }
+	    compDictObj = objv[++i];
+	    break;
+	}
+	if (flush == -2) {
+	    Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		    "\"-flush\", \"-fullflush\" and \"-finalize\" options"
+		    " are mutually exclusive", -1));
+	    Tcl_SetErrorCode(interp, "TCL", "ZIP", "EXCLUSIVE", NULL);
+	    return TCL_ERROR;
+	}
+    }
+    if (flush == -1) {
+	flush = 0;
+    }
+
+    /*
+     * Set the compression dictionary if requested.
+     */
+
+    if (compDictObj != NULL) {
+	int len;
+
+	(void) Tcl_GetByteArrayFromObj(compDictObj, &len);
+	if (len == 0) {
+	    compDictObj = NULL;
+	}
+	Tcl_ZlibStreamSetCompressionDictionary(zstream, compDictObj);
+    }
+
+    /*
+     * Send the data to the stream core, along with any flushing directive.
+     */
+
+    return Tcl_ZlibStreamPut(zstream, objv[objc-1], flush);
+}
+
+static int
+ZlibStreamHeaderCmd(
+    ClientData cd,
+    Tcl_Interp *interp,
+    int objc,
+    Tcl_Obj *const objv[])
+{
+    ZlibStreamHandle *zshPtr = cd;
+    Tcl_Obj *resultObj;
+
+    if (objc != 2) {
+	Tcl_WrongNumArgs(interp, 2, objv, NULL);
+	return TCL_ERROR;
+    } else if (zshPtr->mode != TCL_ZLIB_STREAM_INFLATE
+	    || zshPtr->format != TCL_ZLIB_FORMAT_GZIP) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj(
+		"only gunzip streams can produce header information", -1));
+	Tcl_SetErrorCode(interp, "TCL", "ZIP", "BADOP", NULL);
+	return TCL_ERROR;
+    }
+
+    TclNewObj(resultObj);
+    ExtractHeader(&zshPtr->gzHeaderPtr->header, resultObj);
+    Tcl_SetObjResult(interp, resultObj);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *	Set of functions to support channel stacking.
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformClose --
+ *
+ *	How to shut down a stacked compressing/decompressing transform.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformClose(
+    ClientData instanceData,
+    Tcl_Interp *interp)
+{
+    ZlibChannelData *cd = instanceData;
+    int e, written, result = TCL_OK;
+
+    /*
+     * Delete the support timer.
+     */
+
+    ZlibTransformEventTimerKill(cd);
+
+    /*
+     * Flush any data waiting to be compressed.
+     */
+
+    if (cd->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	cd->outStream.avail_in = 0;
+	do {
+	    e = Deflate(&cd->outStream, cd->outBuffer, cd->outAllocated,
+		    Z_FINISH, &written);
+
+	    /*
+	     * Can't be sure that deflate() won't declare the buffer to be
+	     * full (with Z_BUF_ERROR) so handle that case.
+	     */
+
+	    if (e == Z_BUF_ERROR) {
+		e = Z_OK;
+		written = cd->outAllocated;
+	    }
+	    if (e != Z_OK && e != Z_STREAM_END) {
+		/* TODO: is this the right way to do errors on close? */
+		if (!TclInThreadExit()) {
+		    ConvertError(interp, e, cd->outStream.adler);
+		}
+		result = TCL_ERROR;
+		break;
+	    }
+	    if (written && Tcl_WriteRaw(cd->parent, cd->outBuffer, written) < 0) {
+		/* TODO: is this the right way to do errors on close?
+		 * Note: when close is called from FinalizeIOSubsystem then
+		 * interp may be NULL */
+		if (!TclInThreadExit() && interp) {
+		    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			    "error while finalizing file: %s",
+			    Tcl_PosixError(interp)));
+		}
+		result = TCL_ERROR;
+		break;
+	    }
+	} while (e != Z_STREAM_END);
+	(void) deflateEnd(&cd->outStream);
+    } else {
+	(void) inflateEnd(&cd->inStream);
+    }
+
+    /*
+     * Release all memory.
+     */
+
+    if (cd->compDictObj) {
+	Tcl_DecrRefCount(cd->compDictObj);
+	cd->compDictObj = NULL;
+    }
+    Tcl_DStringFree(&cd->decompressed);
+
+    if (cd->inBuffer) {
+	ckfree(cd->inBuffer);
+	cd->inBuffer = NULL;
+    }
+    if (cd->outBuffer) {
+	ckfree(cd->outBuffer);
+	cd->outBuffer = NULL;
+    }
+    ckfree(cd);
+    return result;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformInput --
+ *
+ *	Reader filter that does decompression.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformInput(
+    ClientData instanceData,
+    char *buf,
+    int toRead,
+    int *errorCodePtr)
+{
+    ZlibChannelData *cd = instanceData;
+    Tcl_DriverInputProc *inProc =
+	    Tcl_ChannelInputProc(Tcl_GetChannelType(cd->parent));
+    int readBytes, gotBytes, copied;
+
+    if (cd->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	return inProc(Tcl_GetChannelInstanceData(cd->parent), buf, toRead,
+		errorCodePtr);
+    }
+
+    gotBytes = 0;
+    while (toRead > 0) {
+	/*
+	 * Loop until the request is satisfied (or no data available from
+	 * below, possibly EOF).
+	 */
+
+	copied = ResultCopy(cd, buf, toRead);
+	toRead -= copied;
+	buf += copied;
+	gotBytes += copied;
+
+	if (toRead == 0) {
+	    return gotBytes;
+	}
+
+	/*
+	 * The buffer is exhausted, but the caller wants even more. We now
+	 * have to go to the underlying channel, get more bytes and then
+	 * transform them for delivery. We may not get what we want (full EOF
+	 * or temporarily out of data).
+	 *
+	 * Length (cd->decompressed) == 0, toRead > 0 here.
+	 *
+	 * The zlib transform allows us to read at most one character from the
+	 * underlying channel to properly identify Z_STREAM_END without
+	 * reading over the border.
+	 */
+
+	readBytes = Tcl_ReadRaw(cd->parent, cd->inBuffer, cd->readAheadLimit);
+
+	/*
+	 * Three cases here:
+	 *  1.	Got some data from the underlying channel (readBytes > 0) so
+	 *	it should be fed through the decompression engine.
+	 *  2.	Got an error (readBytes < 0) which we should report up except
+	 *	for the case where we can convert it to a short read.
+	 *  3.	Got an end-of-data from EOF or blocking (readBytes == 0). If
+	 *	it is EOF, try flushing the data out of the decompressor.
+	 */
+
+	if (readBytes < 0) {
+
+	    /* See ReflectInput() in tclIORTrans.c */
+	    if (Tcl_InputBlocked(cd->parent) && (gotBytes > 0)) {
+		return gotBytes;
+	    }
+
+	    *errorCodePtr = Tcl_GetErrno();
+	    return -1;
+	}
+	if (readBytes == 0) {
+	    /*
+	     * Eof in parent.
+	     *
+	     * Now this is a bit different. The partial data waiting is
+	     * converted and returned.
+	     */
+
+	    if (ResultGenerate(cd, 0, Z_SYNC_FLUSH, errorCodePtr) != TCL_OK) {
+		return -1;
+	    }
+
+	    if (Tcl_DStringLength(&cd->decompressed) == 0) {
+		/*
+		 * The drain delivered nothing. Time to deliver what we've
+		 * got.
+		 */
+
+		return gotBytes;
+	    }
+	} else /* readBytes > 0 */ {
+	    /*
+	     * Transform the read chunk, which was not empty. Anything we get
+	     * back is a transformation result to be put into our buffers, and
+	     * the next iteration will put it into the result.
+	     */
+
+	    if (ResultGenerate(cd, readBytes, Z_NO_FLUSH,
+		    errorCodePtr) != TCL_OK) {
+		return -1;
+	    }
+	}
+    }
+    return gotBytes;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformOutput --
+ *
+ *	Writer filter that does compression.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformOutput(
+    ClientData instanceData,
+    const char *buf,
+    int toWrite,
+    int *errorCodePtr)
+{
+    ZlibChannelData *cd = instanceData;
+    Tcl_DriverOutputProc *outProc =
+	    Tcl_ChannelOutputProc(Tcl_GetChannelType(cd->parent));
+    int e, produced;
+    Tcl_Obj *errObj;
+
+    if (cd->mode == TCL_ZLIB_STREAM_INFLATE) {
+	return outProc(Tcl_GetChannelInstanceData(cd->parent), buf, toWrite,
+		errorCodePtr);
+    }
+
+    /*
+     * No zero-length writes. Flushes must be explicit.
+     */
+
+    if (toWrite == 0) {
+	return 0;
+    }
+
+    cd->outStream.next_in = (Bytef *) buf;
+    cd->outStream.avail_in = toWrite;
+    while (cd->outStream.avail_in > 0) {
+	e = Deflate(&cd->outStream, cd->outBuffer, cd->outAllocated,
+		Z_NO_FLUSH, &produced);
+	if (e != Z_OK || produced == 0) {
+	    break;
+	}
+
+	if (Tcl_WriteRaw(cd->parent, cd->outBuffer, produced) < 0) {
+	    *errorCodePtr = Tcl_GetErrno();
+	    return -1;
+	}
+    }
+
+    if (e == Z_OK) {
+	return toWrite - cd->outStream.avail_in;
+    }
+
+    errObj = Tcl_NewListObj(0, NULL);
+    Tcl_ListObjAppendElement(NULL, errObj, Tcl_NewStringObj("-errorcode",-1));
+    Tcl_ListObjAppendElement(NULL, errObj,
+	    ConvertErrorToList(e, cd->outStream.adler));
+    Tcl_ListObjAppendElement(NULL, errObj,
+	    Tcl_NewStringObj(cd->outStream.msg, -1));
+    Tcl_SetChannelError(cd->parent, errObj);
+    *errorCodePtr = EINVAL;
+    return -1;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformFlush --
+ *
+ *	How to perform a flush of a compressing transform.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformFlush(
+    Tcl_Interp *interp,
+    ZlibChannelData *cd,
+    int flushType)
+{
+    int e, len;
+
+    cd->outStream.avail_in = 0;
+    do {
+	/*
+	 * Get the bytes to go out of the compression engine.
+	 */
+
+	e = Deflate(&cd->outStream, cd->outBuffer, cd->outAllocated,
+		flushType, &len);
+	if (e != Z_OK && e != Z_BUF_ERROR) {
+	    ConvertError(interp, e, cd->outStream.adler);
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * Write the bytes we've received to the next layer.
+	 */
+
+	if (len > 0 && Tcl_WriteRaw(cd->parent, cd->outBuffer, len) < 0) {
+	    Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+		    "problem flushing channel: %s",
+		    Tcl_PosixError(interp)));
+	    return TCL_ERROR;
+	}
+
+	/*
+	 * If we get to this point, either we're in the Z_OK or the
+	 * Z_BUF_ERROR state. In the former case, we're done. In the latter
+	 * case, it's because there's more bytes to go than would fit in the
+	 * buffer we provided, and we need to go round again to get some more.
+	 *
+	 * We also stop the loop if we would have done a zero-length write.
+	 * Those can cause problems at the OS level.
+	 */
+    } while (len > 0 && e == Z_BUF_ERROR);
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformSetOption --
+ *
+ *	Writing side of [fconfigure] on our channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformSetOption(			/* not used */
+    ClientData instanceData,
+    Tcl_Interp *interp,
+    const char *optionName,
+    const char *value)
+{
+    ZlibChannelData *cd = instanceData;
+    Tcl_DriverSetOptionProc *setOptionProc =
+	    Tcl_ChannelSetOptionProc(Tcl_GetChannelType(cd->parent));
+    static const char *compressChanOptions = "dictionary flush";
+    static const char *gzipChanOptions = "flush";
+    static const char *decompressChanOptions = "dictionary limit";
+    static const char *gunzipChanOptions = "flush limit";
+    int haveFlushOpt = (cd->mode == TCL_ZLIB_STREAM_DEFLATE);
+
+    if (optionName && (strcmp(optionName, "-dictionary") == 0)
+	    && (cd->format != TCL_ZLIB_FORMAT_GZIP)) {
+	Tcl_Obj *compDictObj;
+	int code;
+
+	TclNewStringObj(compDictObj, value, strlen(value));
+	Tcl_IncrRefCount(compDictObj);
+	(void) Tcl_GetByteArrayFromObj(compDictObj, NULL);
+	if (cd->compDictObj) {
+	    TclDecrRefCount(cd->compDictObj);
+	}
+	cd->compDictObj = compDictObj;
+	code = Z_OK;
+	if (cd->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	    code = SetDeflateDictionary(&cd->outStream, compDictObj);
+	    if (code != Z_OK) {
+		ConvertError(interp, code, cd->outStream.adler);
+		return TCL_ERROR;
+	    }
+	} else if (cd->format == TCL_ZLIB_FORMAT_RAW) {
+	    code = SetInflateDictionary(&cd->inStream, compDictObj);
+	    if (code != Z_OK) {
+		ConvertError(interp, code, cd->inStream.adler);
+		return TCL_ERROR;
+	    }
+	}
+	return TCL_OK;
+    }
+
+    if (haveFlushOpt) {
+	if (optionName && strcmp(optionName, "-flush") == 0) {
+	    int flushType;
+
+	    if (value[0] == 'f' && strcmp(value, "full") == 0) {
+		flushType = Z_FULL_FLUSH;
+	    } else if (value[0] == 's' && strcmp(value, "sync") == 0) {
+		flushType = Z_SYNC_FLUSH;
+	    } else {
+		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
+			"unknown -flush type \"%s\": must be full or sync",
+			value));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "FLUSH", NULL);
+		return TCL_ERROR;
+	    }
+
+	    /*
+	     * Try to actually do the flush now.
+	     */
+
+	    return ZlibTransformFlush(interp, cd, flushType);
+	}
+    } else {
+	if (optionName && strcmp(optionName, "-limit") == 0) {
+	    int newLimit;
+
+	    if (Tcl_GetInt(interp, value, &newLimit) != TCL_OK) {
+		return TCL_ERROR;
+	    } else if (newLimit < 1 || newLimit > MAX_BUFFER_SIZE) {
+		Tcl_SetObjResult(interp, Tcl_NewStringObj(
+			"-limit must be between 1 and 65536", -1));
+		Tcl_SetErrorCode(interp, "TCL", "VALUE", "READLIMIT", NULL);
+		return TCL_ERROR;
+	    }
+	}
+    }
+
+    if (setOptionProc == NULL) {
+	if (cd->format == TCL_ZLIB_FORMAT_GZIP) {
+	    return Tcl_BadChannelOption(interp, optionName,
+		    (cd->mode == TCL_ZLIB_STREAM_DEFLATE)
+		    ? gzipChanOptions : gunzipChanOptions);
+	} else {
+	    return Tcl_BadChannelOption(interp, optionName,
+		    (cd->mode == TCL_ZLIB_STREAM_DEFLATE)
+		    ? compressChanOptions : decompressChanOptions);
+	}
+    }
+
+    /*
+     * Pass all unknown options down, to deeper transforms and/or the base
+     * channel.
+     */
+
+    return setOptionProc(Tcl_GetChannelInstanceData(cd->parent), interp,
+	    optionName, value);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformGetOption --
+ *
+ *	Reading side of [fconfigure] on our channel.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformGetOption(
+    ClientData instanceData,
+    Tcl_Interp *interp,
+    const char *optionName,
+    Tcl_DString *dsPtr)
+{
+    ZlibChannelData *cd = instanceData;
+    Tcl_DriverGetOptionProc *getOptionProc =
+	    Tcl_ChannelGetOptionProc(Tcl_GetChannelType(cd->parent));
+    static const char *compressChanOptions = "checksum dictionary";
+    static const char *gzipChanOptions = "checksum";
+    static const char *decompressChanOptions = "checksum dictionary limit";
+    static const char *gunzipChanOptions = "checksum header limit";
+
+    /*
+     * The "crc" option reports the current CRC (calculated with the Adler32
+     * or CRC32 algorithm according to the format) given the data that has
+     * been processed so far.
+     */
+
+    if (optionName == NULL || strcmp(optionName, "-checksum") == 0) {
+	uLong crc;
+	char buf[12];
+
+	if (cd->mode == TCL_ZLIB_STREAM_DEFLATE) {
+	    crc = cd->outStream.adler;
+	} else {
+	    crc = cd->inStream.adler;
+	}
+
+	sprintf(buf, "%lu", crc);
+	if (optionName == NULL) {
+	    Tcl_DStringAppendElement(dsPtr, "-checksum");
+	    Tcl_DStringAppendElement(dsPtr, buf);
+	} else {
+	    Tcl_DStringAppend(dsPtr, buf, -1);
+	    return TCL_OK;
+	}
+    }
+
+    if ((cd->format != TCL_ZLIB_FORMAT_GZIP) &&
+	    (optionName == NULL || strcmp(optionName, "-dictionary") == 0)) {
+	/*
+	 * Embedded NUL bytes are ok; they'll be C080-encoded.
+	 */
+
+	if (optionName == NULL) {
+	    Tcl_DStringAppendElement(dsPtr, "-dictionary");
+	    if (cd->compDictObj) {
+		Tcl_DStringAppendElement(dsPtr,
+			Tcl_GetString(cd->compDictObj));
+	    } else {
+		Tcl_DStringAppendElement(dsPtr, "");
+	    }
+	} else {
+	    if (cd->compDictObj) {
+		int len;
+		const char *str = TclGetStringFromObj(cd->compDictObj, &len);
+
+		Tcl_DStringAppend(dsPtr, str, len);
+	    }
+	    return TCL_OK;
+	}
+    }
+
+    /*
+     * The "header" option, which is only valid on inflating gzip channels,
+     * reports the header that has been read from the start of the stream.
+     */
+
+    if ((cd->flags & IN_HEADER) && ((optionName == NULL) ||
+	    (strcmp(optionName, "-header") == 0))) {
+	Tcl_Obj *tmpObj = Tcl_NewObj();
+
+	ExtractHeader(&cd->inHeader.header, tmpObj);
+	if (optionName == NULL) {
+	    Tcl_DStringAppendElement(dsPtr, "-header");
+	    Tcl_DStringAppendElement(dsPtr, Tcl_GetString(tmpObj));
+	    Tcl_DecrRefCount(tmpObj);
+	} else {
+	    TclDStringAppendObj(dsPtr, tmpObj);
+	    Tcl_DecrRefCount(tmpObj);
+	    return TCL_OK;
+	}
+    }
+
+    /*
+     * Now we do the standard processing of the stream we wrapped.
+     */
+
+    if (getOptionProc) {
+	return getOptionProc(Tcl_GetChannelInstanceData(cd->parent),
+		interp, optionName, dsPtr);
+    }
+    if (optionName == NULL) {
+	return TCL_OK;
+    }
+    if (cd->format == TCL_ZLIB_FORMAT_GZIP) {
+	return Tcl_BadChannelOption(interp, optionName,
+		(cd->mode == TCL_ZLIB_STREAM_DEFLATE)
+		? gzipChanOptions : gunzipChanOptions);
+    } else {
+	return Tcl_BadChannelOption(interp, optionName,
+		(cd->mode == TCL_ZLIB_STREAM_DEFLATE)
+		? compressChanOptions : decompressChanOptions);
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformWatch, ZlibTransformEventHandler --
+ *
+ *	If we have data pending, trigger a readable event after a short time
+ *	(in order to allow a real event to catch up).
+ *
+ *----------------------------------------------------------------------
+ */
+
+static void
+ZlibTransformWatch(
+    ClientData instanceData,
+    int mask)
+{
+    ZlibChannelData *cd = instanceData;
+    Tcl_DriverWatchProc *watchProc;
+
+    /*
+     * This code is based on the code in tclIORTrans.c
+     */
+
+    watchProc = Tcl_ChannelWatchProc(Tcl_GetChannelType(cd->parent));
+    watchProc(Tcl_GetChannelInstanceData(cd->parent), mask);
+
+    if (!(mask & TCL_READABLE) || Tcl_DStringLength(&cd->decompressed) == 0) {
+	ZlibTransformEventTimerKill(cd);
+    } else if (cd->timer == NULL) {
+	cd->timer = Tcl_CreateTimerHandler(SYNTHETIC_EVENT_TIME,
+		ZlibTransformTimerRun, cd);
+    }
+}
+
+static int
+ZlibTransformEventHandler(
+    ClientData instanceData,
+    int interestMask)
+{
+    ZlibChannelData *cd = instanceData;
+
+    ZlibTransformEventTimerKill(cd);
+    return interestMask;
+}
+
+static inline void
+ZlibTransformEventTimerKill(
+    ZlibChannelData *cd)
+{
+    if (cd->timer != NULL) {
+	Tcl_DeleteTimerHandler(cd->timer);
+	cd->timer = NULL;
+    }
+}
+
+static void
+ZlibTransformTimerRun(
+    ClientData clientData)
+{
+    ZlibChannelData *cd = clientData;
+
+    cd->timer = NULL;
+    Tcl_NotifyChannel(cd->chan, TCL_READABLE);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformGetHandle --
+ *
+ *	Anything that needs the OS handle is told to get it from what we are
+ *	stacked on top of.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformGetHandle(
+    ClientData instanceData,
+    int direction,
+    ClientData *handlePtr)
+{
+    ZlibChannelData *cd = instanceData;
+
+    return Tcl_GetChannelHandle(cd->parent, direction, handlePtr);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibTransformBlockMode --
+ *
+ *	We need to keep track of the blocking mode; it changes our behavior.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ZlibTransformBlockMode(
+    ClientData instanceData,
+    int mode)
+{
+    ZlibChannelData *cd = instanceData;
+
+    if (mode == TCL_MODE_NONBLOCKING) {
+	cd->flags |= ASYNC;
+    } else {
+	cd->flags &= ~ASYNC;
+    }
+    return TCL_OK;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ZlibStackChannelTransform --
+ *
+ *	Stacks either compression or decompression onto a channel.
+ *
+ * Results:
+ *	The stacked channel, or NULL if there was an error.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static Tcl_Channel
+ZlibStackChannelTransform(
+    Tcl_Interp *interp,		/* Where to write error messages. */
+    int mode,			/* Whether this is a compressing transform
+				 * (TCL_ZLIB_STREAM_DEFLATE) or a
+				 * decompressing transform
+				 * (TCL_ZLIB_STREAM_INFLATE). Note that
+				 * compressing transforms require that the
+				 * channel is writable, and decompressing
+				 * transforms require that the channel is
+				 * readable. */
+    int format,			/* One of the TCL_ZLIB_FORMAT_* values that
+				 * indicates what compressed format to allow.
+				 * TCL_ZLIB_FORMAT_AUTO is only supported for
+				 * decompressing transforms. */
+    int level,			/* What compression level to use. Ignored for
+				 * decompressing transforms. */
+    int limit,			/* The limit on the number of bytes to read
+				 * ahead; always at least 1. */
+    Tcl_Channel channel,	/* The channel to attach to. */
+    Tcl_Obj *gzipHeaderDictPtr,	/* A description of header to use, or NULL to
+				 * use a default. Ignored if not compressing
+				 * to produce gzip-format data. */
+    Tcl_Obj *compDictObj)	/* Byte-array object containing compression
+				 * dictionary (not dictObj!) to use if
+				 * necessary. */
+{
+    ZlibChannelData *cd = ckalloc(sizeof(ZlibChannelData));
+    Tcl_Channel chan;
+    int wbits = 0;
+
+    if (mode != TCL_ZLIB_STREAM_DEFLATE && mode != TCL_ZLIB_STREAM_INFLATE) {
+	Tcl_Panic("unknown mode: %d", mode);
+    }
+
+    memset(cd, 0, sizeof(ZlibChannelData));
+    cd->mode = mode;
+    cd->format = format;
+    cd->readAheadLimit = limit;
+
+    if (format == TCL_ZLIB_FORMAT_GZIP || format == TCL_ZLIB_FORMAT_AUTO) {
+	if (mode == TCL_ZLIB_STREAM_DEFLATE) {
+	    if (gzipHeaderDictPtr) {
+		cd->flags |= OUT_HEADER;
+		if (GenerateHeader(interp, gzipHeaderDictPtr, &cd->outHeader,
+			NULL) != TCL_OK) {
+		    goto error;
+		}
+	    }
+	} else {
+	    cd->flags |= IN_HEADER;
+	    cd->inHeader.header.name = (Bytef *)
+		    &cd->inHeader.nativeFilenameBuf;
+	    cd->inHeader.header.name_max = MAXPATHLEN - 1;
+	    cd->inHeader.header.comment = (Bytef *)
+		    &cd->inHeader.nativeCommentBuf;
+	    cd->inHeader.header.comm_max = MAX_COMMENT_LEN - 1;
+	}
+    }
+
+    if (compDictObj != NULL) {
+	cd->compDictObj = Tcl_DuplicateObj(compDictObj);
+	Tcl_IncrRefCount(cd->compDictObj);
+	Tcl_GetByteArrayFromObj(cd->compDictObj, NULL);
+    }
+
+    if (format == TCL_ZLIB_FORMAT_RAW) {
+	wbits = WBITS_RAW;
+    } else if (format == TCL_ZLIB_FORMAT_ZLIB) {
+	wbits = WBITS_ZLIB;
+    } else if (format == TCL_ZLIB_FORMAT_GZIP) {
+	wbits = WBITS_GZIP;
+    } else if (format == TCL_ZLIB_FORMAT_AUTO) {
+	wbits = WBITS_AUTODETECT;
+    } else {
+	Tcl_Panic("bad format: %d", format);
+    }
+
+    /*
+     * Initialize input inflater or the output deflater.
+     */
+
+    if (mode == TCL_ZLIB_STREAM_INFLATE) {
+	if (inflateInit2(&cd->inStream, wbits) != Z_OK) {
+	    goto error;
+	}
+	cd->inAllocated = DEFAULT_BUFFER_SIZE;
+	cd->inBuffer = ckalloc(cd->inAllocated);
+	if (cd->flags & IN_HEADER) {
+	    if (inflateGetHeader(&cd->inStream, &cd->inHeader.header) != Z_OK) {
+		goto error;
+	    }
+	}
+	if (cd->format == TCL_ZLIB_FORMAT_RAW && cd->compDictObj) {
+	    if (SetInflateDictionary(&cd->inStream, cd->compDictObj) != Z_OK) {
+		goto error;
+	    }
+	}
+    } else {
+	if (deflateInit2(&cd->outStream, level, Z_DEFLATED, wbits,
+		MAX_MEM_LEVEL, Z_DEFAULT_STRATEGY) != Z_OK) {
+	    goto error;
+	}
+	cd->outAllocated = DEFAULT_BUFFER_SIZE;
+	cd->outBuffer = ckalloc(cd->outAllocated);
+	if (cd->flags & OUT_HEADER) {
+	    if (deflateSetHeader(&cd->outStream, &cd->outHeader.header) != Z_OK) {
+		goto error;
+	    }
+	}
+	if (cd->compDictObj) {
+	    if (SetDeflateDictionary(&cd->outStream, cd->compDictObj) != Z_OK) {
+		goto error;
+	    }
+	}
+    }
+
+    Tcl_DStringInit(&cd->decompressed);
+
+    chan = Tcl_StackChannel(interp, &zlibChannelType, cd,
+	    Tcl_GetChannelMode(channel), channel);
+    if (chan == NULL) {
+	goto error;
+    }
+    cd->chan = chan;
+    cd->parent = Tcl_GetStackedChannel(chan);
+    Tcl_SetObjResult(interp, Tcl_NewStringObj(Tcl_GetChannelName(chan), -1));
+    return chan;
+
+  error:
+    if (cd->inBuffer) {
+	ckfree(cd->inBuffer);
+	inflateEnd(&cd->inStream);
+    }
+    if (cd->outBuffer) {
+	ckfree(cd->outBuffer);
+	deflateEnd(&cd->outStream);
+    }
+    if (cd->compDictObj) {
+	Tcl_DecrRefCount(cd->compDictObj);
+    }
+    ckfree(cd);
+    return NULL;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultCopy --
+ *
+ *	Copies the requested number of bytes from the buffer into the
+ *	specified array and removes them from the buffer afterward. Copies
+ *	less if there is not enough data in the buffer.
+ *
+ * Side effects:
+ *	See above.
+ *
+ * Result:
+ *	The number of actually copied bytes, possibly less than 'toRead'.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static inline int
+ResultCopy(
+    ZlibChannelData *cd,	/* The location of the buffer to read from. */
+    char *buf,			/* The buffer to copy into */
+    int toRead)			/* Number of requested bytes */
+{
+    int have = Tcl_DStringLength(&cd->decompressed);
+
+    if (have == 0) {
+	/*
+	 * Nothing to copy in the case of an empty buffer.
+	 */
+
+	return 0;
+    } else if (have > toRead) {
+	/*
+	 * The internal buffer contains more than requested. Copy the
+	 * requested subset to the caller, shift the remaining bytes down, and
+	 * truncate.
+	 */
+
+	char *src = Tcl_DStringValue(&cd->decompressed);
+
+	memcpy(buf, src, toRead);
+	memmove(src, src + toRead, have - toRead);
+
+	Tcl_DStringSetLength(&cd->decompressed, have - toRead);
+	return toRead;
+    } else /* have <= toRead */ {
+	/*
+	 * There is just or not enough in the buffer to fully satisfy the
+	 * caller, so take everything as best effort.
+	 */
+
+	memcpy(buf, Tcl_DStringValue(&cd->decompressed), have);
+	TclDStringClear(&cd->decompressed);
+	return have;
+    }
+}
+
+/*
+ *----------------------------------------------------------------------
+ *
+ * ResultGenerate --
+ *
+ *	Extract uncompressed bytes from the compression engine and store them
+ *	in our working buffer.
+ *
+ * Result:
+ *	TCL_OK/TCL_ERROR (with *errorCodePtr updated with reason).
+ *
+ * Side effects:
+ *	See above.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static int
+ResultGenerate(
+    ZlibChannelData *cd,
+    int n,
+    int flush,
+    int *errorCodePtr)
+{
+#define MAXBUF	1024
+    unsigned char buf[MAXBUF];
+    int e, written;
+    Tcl_Obj *errObj;
+
+    cd->inStream.next_in = (Bytef *) cd->inBuffer;
+    cd->inStream.avail_in = n;
+
+    while (1) {
+	cd->inStream.next_out = (Bytef *) buf;
+	cd->inStream.avail_out = MAXBUF;
+
+	e = inflate(&cd->inStream, flush);
+	if (e == Z_NEED_DICT && cd->compDictObj) {
+	    e = SetInflateDictionary(&cd->inStream, cd->compDictObj);
+	    if (e == Z_OK) {
+		/*
+		 * A repetition of Z_NEED_DICT is just an error.
+		 */
+
+		cd->inStream.next_out = (Bytef *) buf;
+		cd->inStream.avail_out = MAXBUF;
+		e = inflate(&cd->inStream, flush);
+	    }
+	}
+
+	/*
+	 * avail_out is now the left over space in the output.  Therefore
+	 * "MAXBUF - avail_out" is the amount of bytes generated.
+	 */
+
+	written = MAXBUF - cd->inStream.avail_out;
+	if (written) {
+	    Tcl_DStringAppend(&cd->decompressed, (char *) buf, written);
+	}
+
+	/*
+	 * The cases where we're definitely done.
+	 */
+
+	if (((flush == Z_SYNC_FLUSH) && (e == Z_BUF_ERROR))
+		|| (e == Z_STREAM_END)
+		|| (e == Z_OK && cd->inStream.avail_out == 0)) {
+	    return TCL_OK;
+	}
+
+	/*
+	 * Z_BUF_ERROR can be ignored as per http://www.zlib.net/zlib_how.html
+	 *
+	 * Just indicates that the zlib couldn't consume input/produce output,
+	 * and is fixed by supplying more input.
+	 *
+	 * Otherwise, we've got errors and need to report to higher-up.
+	 */
+
+	if ((e != Z_OK) && (e != Z_BUF_ERROR)) {
+	    goto handleError;
+	}
+
+	/*
+	 * Check if the inflate stopped early.
+	 */
+
+	if (cd->inStream.avail_in <= 0 && flush != Z_SYNC_FLUSH) {
+	    return TCL_OK;
+	}
+    }
+
+  handleError:
+    errObj = Tcl_NewListObj(0, NULL);
+    Tcl_ListObjAppendElement(NULL, errObj, Tcl_NewStringObj("-errorcode",-1));
+    Tcl_ListObjAppendElement(NULL, errObj,
+	    ConvertErrorToList(e, cd->inStream.adler));
+    Tcl_ListObjAppendElement(NULL, errObj,
+	    Tcl_NewStringObj(cd->inStream.msg, -1));
+    Tcl_SetChannelError(cd->parent, errObj);
+    *errorCodePtr = EINVAL;
+    return TCL_ERROR;
+}
+
+/*
+ *----------------------------------------------------------------------
+ *	Finally, the TclZlibInit function. Used to install the zlib API.
+ *----------------------------------------------------------------------
+ */
+
+int
+TclZlibInit(
+    Tcl_Interp *interp)
+{
+    Tcl_Config cfg[2];
+
+    /*
+     * This does two things. It creates a counter used in the creation of
+     * stream commands, and it creates the namespace that will contain those
+     * commands.
+     */
+
+    Tcl_EvalEx(interp, "namespace eval ::tcl::zlib {variable cmdcounter 0}", -1, 0);
+
+    /*
+     * Create the public scripted interface to this file's functionality.
+     */
+
+    Tcl_CreateObjCommand(interp, "zlib", ZlibCmd, 0, 0);
+
+    /*
+     * Store the underlying configuration information.
+     *
+     * TODO: Describe whether we're using the system version of the library or
+     * a compatibility version built into Tcl?
+     */
+
+    cfg[0].key = "zlibVersion";
+    cfg[0].value = zlibVersion();
+    cfg[1].key = NULL;
+    Tcl_RegisterConfig(interp, "zlib", cfg, "iso8859-1");
+
+    /*
+     * Formally provide the package as a Tcl built-in.
+     */
+
+    return Tcl_PkgProvide(interp, "zlib", TCL_ZLIB_VERSION);
+}
+
+/*
+ *----------------------------------------------------------------------
+ *	Stubs used when a suitable zlib installation was not found during
+ *	configure.
+ *----------------------------------------------------------------------
+ */
+
+#else /* !HAVE_ZLIB */
+int
+Tcl_ZlibStreamInit(
+    Tcl_Interp *interp,
+    int mode,
+    int format,
+    int level,
+    Tcl_Obj *dictObj,
+    Tcl_ZlibStream *zshandle)
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("unimplemented", -1));
+	Tcl_SetErrorCode(interp, "TCL", "UNIMPLEMENTED", NULL);
+    }
+    return TCL_ERROR;
+}
+
+int
+Tcl_ZlibStreamClose(
+    Tcl_ZlibStream zshandle)
+{
+    return TCL_OK;
+}
+
+int
+Tcl_ZlibStreamReset(
+    Tcl_ZlibStream zshandle)
+{
+    return TCL_OK;
+}
+
+Tcl_Obj *
+Tcl_ZlibStreamGetCommandName(
+    Tcl_ZlibStream zshandle)
+{
+    return NULL;
+}
+
+int
+Tcl_ZlibStreamEof(
+    Tcl_ZlibStream zshandle)
+{
+    return 1;
+}
+
+int
+Tcl_ZlibStreamChecksum(
+    Tcl_ZlibStream zshandle)
+{
+    return 0;
+}
+
+int
+Tcl_ZlibStreamPut(
+    Tcl_ZlibStream zshandle,
+    Tcl_Obj *data,
+    int flush)
+{
+    return TCL_OK;
+}
+
+int
+Tcl_ZlibStreamGet(
+    Tcl_ZlibStream zshandle,
+    Tcl_Obj *data,
+    int count)
+{
+    return TCL_OK;
+}
+
+int
+Tcl_ZlibDeflate(
+    Tcl_Interp *interp,
+    int format,
+    Tcl_Obj *data,
+    int level,
+    Tcl_Obj *gzipHeaderDictObj)
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("unimplemented", -1));
+	Tcl_SetErrorCode(interp, "TCL", "UNIMPLEMENTED", NULL);
+    }
+    return TCL_ERROR;
+}
+
+int
+Tcl_ZlibInflate(
+    Tcl_Interp *interp,
+    int format,
+    Tcl_Obj *data,
+    int bufferSize,
+    Tcl_Obj *gzipHeaderDictObj)
+{
+    if (interp) {
+	Tcl_SetObjResult(interp, Tcl_NewStringObj("unimplemented", -1));
+	Tcl_SetErrorCode(interp, "TCL", "UNIMPLEMENTED", NULL);
+    }
+    return TCL_ERROR;
+}
+
+unsigned int
+Tcl_ZlibCRC32(
+    unsigned int crc,
+    const char *buf,
+    int len)
+{
+    return 0;
+}
+
+unsigned int
+Tcl_ZlibAdler32(
+    unsigned int adler,
+    const char *buf,
+    int len)
+{
+    return 0;
+}
+
+void
+Tcl_ZlibStreamSetCompressionDictionary(
+    Tcl_ZlibStream zshandle,
+    Tcl_Obj *compressionDictionaryObj)
+{
+    /* Do nothing. */
+}
+#endif /* HAVE_ZLIB */
+
+/*
+ * Local Variables:
+ * mode: c
+ * c-basic-offset: 4
+ * fill-column: 78
+ * End:
+ */
diff --git a/generic/tommath.h b/generic/tommath.h
new file mode 100644
index 0000000..028a84d
--- /dev/null
+++ b/generic/tommath.h
@@ -0,0 +1 @@
+#include "tclTomMathInt.h"