TIP 537 implementation: Enable 64-bit indexes in regexp matching

16 files changed, 96 insertions, 102 deletions

diff --git a/doc/RegExp.3 b/doc/RegExp.3
index aa757bc..95716e9 100644
--- a/doc/RegExp.3
+++ b/doc/RegExp.3
@@ -64,7 +64,7 @@ identifies the beginning of the larger string.
 If it is not the same as \fItext\fR, then no
 .QW \fB^\fR
 matches will be allowed.
-.AP int index in
+.AP size_t index in
 Specifies which range is desired:  0 means the range of the entire
 match, 1 or greater means the range that matched a parenthesized
 sub-expression.
@@ -80,14 +80,14 @@ OR-ed combination of the compilation flags \fBTCL_REG_ADVANCED\fR,
 \fBTCL_REG_QUOTE\fR, \fBTCL_REG_NOCASE\fR, \fBTCL_REG_NEWLINE\fR,
 \fBTCL_REG_NLSTOP\fR, \fBTCL_REG_NLANCH\fR, \fBTCL_REG_NOSUB\fR, and
 \fBTCL_REG_CANMATCH\fR. See below for more information.
-.AP int offset in
+.AP size_t offset in
 The character offset into the text where matching should begin.
 The value of the offset has no impact on \fB^\fR matches.  This
 behavior is controlled by \fIeflags\fR.
-.AP int nmatches in
+.AP size_t nmatches in
 The number of matching subexpressions that should be remembered for
 later use.  If this value is 0, then no subexpression match
-information will be computed.  If the value is \-1, then
+information will be computed.  If the value is TCL_INDEX_NONE, then
 all of the matching subexpressions will be remembered.  Any other
 value will be taken as the maximum number of subexpressions to
 remember.
@@ -337,9 +337,9 @@ defined as follows:
 .PP
 .CS
 typedef struct Tcl_RegExpInfo {
-    int \fInsubs\fR;
+    size_t \fInsubs\fR;
     Tcl_RegExpIndices *\fImatches\fR;
-    long \fIextendStart\fR;
+    size_t \fIextendStart\fR;
 } \fBTcl_RegExpInfo\fR;
 .CE
 .PP
@@ -355,8 +355,8 @@ follows:
 .PP
 .CS
 typedef struct Tcl_RegExpIndices {
-    long \fIstart\fR;
-    long \fIend\fR;
+    size_t \fIstart\fR;
+    size_t \fIend\fR;
 } \fBTcl_RegExpIndices\fR;
 .CE
 .PP
diff --git a/doc/Utf.3 b/doc/Utf.3
index 80c86aa..a642d30 100644
--- a/doc/Utf.3
+++ b/doc/Utf.3
@@ -244,7 +244,7 @@ characters.
 \fBTcl_UtfAtIndex\fR returns a pointer to the specified character (not
 byte) \fIindex\fR in the UTF-8 string \fIsrc\fR.  The source string must
 contain at least \fIindex\fR characters.  This is equivalent to calling
-\fBTcl_UtfNext\fR \fIindex\fR times.  If \fIindex\fR is TCL_AUTO_LENGTH,
+\fBTcl_UtfNext\fR \fIindex\fR times.  If \fIindex\fR is TCL_INDEX_NONE,
 the return pointer points to the first character in the source string.
 .PP
 \fBTcl_UtfBackslash\fR is a utility procedure used by several of the Tcl
diff --git a/generic/regcomp.c b/generic/regcomp.c
index 47f06c8..9bf862c 100644
--- a/generic/regcomp.c
+++ b/generic/regcomp.c
@@ -338,7 +338,6 @@ compile(
     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);
 
@@ -2085,8 +2084,8 @@ dump(
     }
 
     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);
+    fprintf(f, "nsub %" TCL_Z_MODIFIER "d, info 0%lo, ntree %d\n",
+	    re->re_nsub, re->re_info, g->ntree);
 
     dumpcolors(&g->cmap, f);
     if (!NULLCNFA(g->search)) {
diff --git a/generic/regex.h b/generic/regex.h
index 81f98a4..47a8103 100644
--- a/generic/regex.h
+++ b/generic/regex.h
@@ -117,8 +117,8 @@ extern "C" {
 /* 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 */
+    size_t re_nsub;		/* number of subexpressions */
 #define	REG_UBACKREF		000001
 #define	REG_ULOOKAHEAD		000002
 #define	REG_UBOUNDS		000004
@@ -133,7 +133,6 @@ typedef struct {
 #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 *' */
@@ -142,8 +141,8 @@ typedef struct {
 
 /* result reporting (may acquire more fields later) */
 typedef struct {
-    long rm_so;		/* start of substring */
-    long rm_eo;		/* end of substring */
+    size_t rm_so;		/* start of substring */
+    size_t rm_eo;		/* end of substring */
 } regmatch_t;
 
 /* supplementary control and reporting */
diff --git a/generic/regexec.c b/generic/regexec.c
index c57f42c..5ad8b17 100644
--- a/generic/regexec.c
+++ b/generic/regexec.c
@@ -187,10 +187,6 @@ exec(
 	FreeVars(v);
 	return REG_INVARG;
     }
-    if (re->re_csize != sizeof(chr)) {
-	FreeVars(v);
-	return REG_MIXED;
-    }
 
     /*
      * Setup.
@@ -889,7 +885,7 @@ cbrdissect(
     MDEBUG(("cbackref n%d %d{%d-%d}\n", t->id, n, min, max));
 
     /* get the backreferenced string */
-    if (v->pmatch[n].rm_so == -1) {
+    if (v->pmatch[n].rm_so == (size_t)-1) {
 	return REG_NOMATCH;
     }
     brstring = v->start + v->pmatch[n].rm_so;
diff --git a/generic/tcl.h b/generic/tcl.h
index 3cf6fad..46561b5 100644
--- a/generic/tcl.h
+++ b/generic/tcl.h
@@ -459,19 +459,18 @@ typedef void (Tcl_ThreadCreateProc) (void *clientData);
  */
 
 typedef struct Tcl_RegExpIndices {
-    long start;			/* Character offset of first character in
+    size_t start;			/* Character offset of first character in
 				 * match. */
-    long end;			/* Character offset of first character after
+    size_t end;			/* Character offset of first character after
 				 * the match. */
 } Tcl_RegExpIndices;
 
 typedef struct Tcl_RegExpInfo {
-    int nsubs;			/* Number of subexpressions in the compiled
+    size_t 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
+    size_t extendStart;		/* The offset at which a subsequent match
 				 * might begin. */
-    long reserved;		/* Reserved for later use. */
 } Tcl_RegExpInfo;
 
 /*
@@ -2171,6 +2170,7 @@ typedef int (Tcl_ArgvGenFuncProc)(void *clientData, Tcl_Interp *interp,
 
 #define TCL_IO_FAILURE	((size_t)-1)
 #define TCL_AUTO_LENGTH	((size_t)-1)
+#define TCL_INDEX_NONE  ((size_t)-1)
 
 /*
  *----------------------------------------------------------------------------
diff --git a/generic/tclCmdMZ.c b/generic/tclCmdMZ.c
index d02a159..e1c2960 100644
--- a/generic/tclCmdMZ.c
+++ b/generic/tclCmdMZ.c
@@ -127,9 +127,8 @@ Tcl_RegexpObjCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    size_t offset;
+    size_t offset, stringLength, matchLength, cflags, eflags;
     int i, indices, match, about, all, doinline, numMatchesSaved;
-    int cflags, eflags, stringLength, matchLength;
     Tcl_RegExp regExpr;
     Tcl_Obj *objPtr, *startIndex = NULL, *resultPtr = NULL;
     Tcl_RegExpInfo info;
@@ -309,7 +308,7 @@ Tcl_RegexpObjCmd(
 
 	if (offset == TCL_INDEX_START) {
 	    eflags = 0;
-	} else if (offset + 1 > (size_t)stringLength + 1) {
+	} else if (offset + 1 > stringLength + 1) {
 	    eflags = TCL_REG_NOTBOL;
 	} else if (Tcl_GetUniChar(objPtr, offset-1) == '\n') {
 	    eflags = 0;
@@ -365,7 +364,7 @@ Tcl_RegexpObjCmd(
 	    Tcl_Obj *newPtr;
 
 	    if (indices) {
-		int start, end;
+		size_t start, end;
 		Tcl_Obj *objs[2];
 
 		/*
@@ -373,7 +372,7 @@ Tcl_RegexpObjCmd(
 		 * area. (Scriptics Bug 4391/SF Bug #219232)
 		 */
 
-		if (i <= info.nsubs && info.matches[i].start >= 0) {
+		if (i <= (int)info.nsubs && info.matches[i].start != TCL_INDEX_NONE) {
 		    start = offset + info.matches[i].start;
 		    end = offset + info.matches[i].end;
 
@@ -382,20 +381,20 @@ Tcl_RegexpObjCmd(
 		     * match instead of the first character after the match.
 		     */
 
-		    if ((size_t)end + 1 >= offset + 1) {
+		    if (end + 1 >= offset + 1) {
 			end--;
 		    }
 		} else {
-		    start = -1;
-		    end = -1;
+		    start = TCL_INDEX_NONE;
+		    end = TCL_INDEX_NONE;
 		}
 
-		objs[0] = Tcl_NewWideIntObj(start);
-		objs[1] = Tcl_NewWideIntObj(end);
+		objs[0] = TclNewWideIntObjFromSize(start);
+		objs[1] = TclNewWideIntObjFromSize(end);
 
 		newPtr = Tcl_NewListObj(2, objs);
 	    } else {
-		if (i <= info.nsubs) {
+		if (i <= (int)info.nsubs) {
 		    newPtr = Tcl_GetRange(objPtr,
 			    offset + info.matches[i].start,
 			    offset + info.matches[i].end - 1);
@@ -445,7 +444,7 @@ Tcl_RegexpObjCmd(
 	    offset++;
 	}
 	all++;
-	if (offset + 1 >= (size_t)stringLength + 1) {
+	if (offset + 1 >= stringLength + 1) {
 	    break;
 	}
     }
@@ -488,9 +487,9 @@ Tcl_RegsubObjCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    int idx, result, cflags, all, numMatches;
-    size_t wlen, wsublen = 0, offset;
-    int start, end, subStart, subEnd, match, command, numParts;
+    int result, cflags, all, match, command, numParts;
+    size_t idx, wlen, wsublen = 0, offset, numMatches;
+    size_t start, end, subStart, subEnd;
     Tcl_RegExp regExpr;
     Tcl_RegExpInfo info;
     Tcl_Obj *resultPtr, *subPtr, *objPtr, *startIndex = NULL;
@@ -513,7 +512,7 @@ Tcl_RegsubObjCmd(
     command = 0;
     resultPtr = NULL;
 
-    for (idx = 1; idx < objc; idx++) {
+    for (idx = 1; idx < (size_t)objc; idx++) {
 	const char *name;
 	int index;
 
@@ -549,7 +548,7 @@ Tcl_RegsubObjCmd(
 	    break;
 	case REGSUB_START: {
 	    size_t temp;
-	    if (++idx >= objc) {
+	    if (++idx >= (size_t)objc) {
 		goto endOfForLoop;
 	    }
 	    if (TclGetIntForIndexM(interp, objv[idx], TCL_INDEX_START, &temp) != TCL_OK) {
@@ -569,7 +568,7 @@ Tcl_RegsubObjCmd(
     }
 
   endOfForLoop:
-    if (objc-idx < 3 || objc-idx > 4) {
+    if ((size_t)objc < idx + 3 || (size_t)objc > idx + 4) {
 	Tcl_WrongNumArgs(interp, 1, objv,
 		"?-option ...? exp string subSpec ?varName?");
     optionError:
@@ -1269,7 +1268,7 @@ Tcl_SplitObjCmd(
 	Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
     } else {
 	const char *element, *p, *splitEnd;
-	int splitLen;
+	size_t splitLen;
 	Tcl_UniChar splitChar = 0;
 
 	/*
@@ -1982,8 +1981,7 @@ StringMapCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    size_t length1, length2;
-    int mapElemc, index;
+    size_t length1, length2,  mapElemc, index;
     int nocase = 0, mapWithDict = 0, copySource = 0;
     Tcl_Obj **mapElemv, *sourceObj, *resultPtr;
     Tcl_UniChar *ustring1, *ustring2, *p, *end;
@@ -2024,8 +2022,8 @@ StringMapCmd(
 	 * sure. This shortens this code quite a bit.
 	 */
 
-	Tcl_DictObjSize(interp, objv[objc-2], &mapElemc);
-	if (mapElemc == 0) {
+	Tcl_DictObjSize(interp, objv[objc-2], &i);
+	if (i == 0) {
 	    /*
 	     * Empty charMap, just return whatever string was given.
 	     */
@@ -2034,7 +2032,7 @@ StringMapCmd(
 	    return TCL_OK;
 	}
 
-	mapElemc *= 2;
+	mapElemc = 2 * i;
 	mapWithDict = 1;
 
 	/*
@@ -2045,15 +2043,17 @@ StringMapCmd(
 	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);
+	for (index=2 ; index<mapElemc ; index+=2) {
+	    Tcl_DictObjNext(&search, mapElemv+index, mapElemv+index+1, &done);
 	}
 	Tcl_DictObjDone(&search);
     } else {
-	if (TclListObjGetElements(interp, objv[objc-2], &mapElemc,
+	int i;
+	if (TclListObjGetElements(interp, objv[objc-2], &i,
 		&mapElemv) != TCL_OK) {
 	    return TCL_ERROR;
 	}
+	mapElemc = i;
 	if (mapElemc == 0) {
 	    /*
 	     * empty charMap, just return whatever string was given.
@@ -3484,9 +3484,9 @@ TclNRSwitchObjCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    int i,j, index, mode, foundmode, splitObjs, numMatchesSaved;
+    int i, index, mode, foundmode, splitObjs, numMatchesSaved;
     int noCase;
-    size_t patternLength;
+    size_t patternLength, j;
     const char *pattern;
     Tcl_Obj *stringObj, *indexVarObj, *matchVarObj;
     Tcl_Obj *const *savedObjv = objv;
@@ -3797,9 +3797,9 @@ TclNRSwitchObjCmd(
 	    if (indexVarObj != NULL) {
 		Tcl_Obj *rangeObjAry[2];
 
-		if (info.matches[j].end > 0) {
-		    rangeObjAry[0] = Tcl_NewWideIntObj(info.matches[j].start);
-		    rangeObjAry[1] = Tcl_NewWideIntObj(info.matches[j].end-1);
+		if (info.matches[j].end + 1 > 1) {
+		    rangeObjAry[0] = TclNewWideIntObjFromSize(info.matches[j].start);
+		    rangeObjAry[1] = TclNewWideIntObjFromSize(info.matches[j].end-1);
 		} else {
 		    rangeObjAry[0] = rangeObjAry[1] = Tcl_NewWideIntObj(-1);
 		}
@@ -3916,7 +3916,7 @@ TclNRSwitchObjCmd(
     }
 
     for (j = i + 1; ; j += 2) {
-	if (j >= objc) {
+	if (j >= (size_t)objc) {
 	    /*
 	     * This shouldn't happen since we've checked that the last body is
 	     * not a continuation...
diff --git a/generic/tclEnv.c b/generic/tclEnv.c
index 8598b9a..49b544d 100644
--- a/generic/tclEnv.c
+++ b/generic/tclEnv.c
@@ -230,7 +230,7 @@ TclSetEnv(
     Tcl_MutexLock(&envMutex);
     index = TclpFindVariable(name, &length);
 
-    if (index == TCL_AUTO_LENGTH) {
+    if (index == TCL_INDEX_NONE) {
 #ifndef USE_PUTENV
 	/*
 	 * We need to handle the case where the environment may be changed
@@ -314,7 +314,7 @@ TclSetEnv(
      * string in the cache.
      */
 
-    if ((index != TCL_AUTO_LENGTH) && (environ[index] == p)) {
+    if ((index != TCL_INDEX_NONE) && (environ[index] == p)) {
 	ReplaceString(oldValue, p);
 #ifdef HAVE_PUTENV_THAT_COPIES
     } else {
diff --git a/generic/tclInt.decls b/generic/tclInt.decls
index 2a66d18..bcdea6c 100644
--- a/generic/tclInt.decls
+++ b/generic/tclInt.decls
@@ -615,8 +615,8 @@ declare 150 {
     int TclRegAbout(Tcl_Interp *interp, Tcl_RegExp re)
 }
 declare 151 {
-    void TclRegExpRangeUniChar(Tcl_RegExp re, size_t index, int *startPtr,
-	    int *endPtr)
+    void TclRegExpRangeUniChar(Tcl_RegExp re, size_t index, size_t *startPtr,
+	    size_t *endPtr)
 }
 declare 152 {
     void TclSetLibraryPath(Tcl_Obj *pathPtr)
diff --git a/generic/tclInt.h b/generic/tclInt.h
index 6f61f1a..33462c04 100644
--- a/generic/tclInt.h
+++ b/generic/tclInt.h
@@ -4128,7 +4128,6 @@ MODULE_SCOPE size_t	TclIndexDecode(int encoded, size_t endValue);
 
 /* Constants used in index value encoding routines. */
 #define TCL_INDEX_END           ((size_t)-2)
-#define TCL_INDEX_NONE          ((size_t)-1) /* Index out of range or END+1 */
 #define TCL_INDEX_START         ((size_t)0)
 
 /*
@@ -4972,7 +4971,7 @@ MODULE_SCOPE Tcl_PackageInitProc Procbodytest_SafeInit;
 #   define TclNewWideIntObjFromSize(value) \
 	Tcl_NewWideIntObj(TclWideIntFromSize(value))
 #else
-#   define TclWideIntFromSize(value)	(value)
+#   define TclWideIntFromSize(value)	((Tcl_WideInt)(value))
 #   define TclNewWideIntObjFromSize Tcl_NewWideIntObj
 #endif
 
diff --git a/generic/tclIntDecls.h b/generic/tclIntDecls.h
index 2d4817b..260ef3e 100644
--- a/generic/tclIntDecls.h
+++ b/generic/tclIntDecls.h
@@ -334,7 +334,7 @@ EXTERN void		TclHandleRelease(TclHandle handle);
 EXTERN int		TclRegAbout(Tcl_Interp *interp, Tcl_RegExp re);
 /* 151 */
 EXTERN void		TclRegExpRangeUniChar(Tcl_RegExp re, size_t index,
-				int *startPtr, int *endPtr);
+				size_t *startPtr, size_t *endPtr);
 /* 152 */
 EXTERN void		TclSetLibraryPath(Tcl_Obj *pathPtr);
 /* 153 */
@@ -739,7 +739,7 @@ typedef struct TclIntStubs {
     TclHandle (*tclHandlePreserve) (TclHandle handle); /* 148 */
     void (*tclHandleRelease) (TclHandle handle); /* 149 */
     int (*tclRegAbout) (Tcl_Interp *interp, Tcl_RegExp re); /* 150 */
-    void (*tclRegExpRangeUniChar) (Tcl_RegExp re, size_t index, int *startPtr, int *endPtr); /* 151 */
+    void (*tclRegExpRangeUniChar) (Tcl_RegExp re, size_t index, size_t *startPtr, size_t *endPtr); /* 151 */
     void (*tclSetLibraryPath) (Tcl_Obj *pathPtr); /* 152 */
     Tcl_Obj * (*tclGetLibraryPath) (void); /* 153 */
     void (*reserved154)(void);
diff --git a/generic/tclRegexp.c b/generic/tclRegexp.c
index 67195bb..5df76c3 100644
--- a/generic/tclRegexp.c
+++ b/generic/tclRegexp.c
@@ -264,7 +264,7 @@ Tcl_RegExpRange(
 
     if (index > regexpPtr->re.re_nsub) {
 	*startPtr = *endPtr = NULL;
-    } else if (regexpPtr->matches[index].rm_so == -1) {
+    } else if (regexpPtr->matches[index].rm_so == (size_t)-1) {
 	*startPtr = *endPtr = NULL;
     } else {
 	if (regexpPtr->objPtr) {
@@ -363,21 +363,21 @@ TclRegExpRangeUniChar(
 				 * passed to Tcl_RegExpExec. */
     size_t 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
+				 * subrange, TCL_INDEX_NONE means the range of the
 				 * rm_extend field. */
-    int *startPtr,		/* Store address of first character in
+    size_t *startPtr,		/* Store address of first character in
 				 * (sub-)range here. */
-    int *endPtr)		/* Store address of character just after last
+    size_t *endPtr)		/* Store address of character just after last
 				 * in (sub-)range here. */
 {
     TclRegexp *regexpPtr = (TclRegexp *) re;
 
-    if ((regexpPtr->flags&REG_EXPECT) && index == TCL_AUTO_LENGTH) {
+    if ((regexpPtr->flags&REG_EXPECT) && (index == TCL_INDEX_NONE)) {
 	*startPtr = regexpPtr->details.rm_extend.rm_so;
 	*endPtr = regexpPtr->details.rm_extend.rm_eo;
-    } else if (index > regexpPtr->re.re_nsub) {
-	*startPtr = -1;
-	*endPtr = -1;
+    } else if (index + 1 > regexpPtr->re.re_nsub + 1) {
+	*startPtr = TCL_INDEX_NONE;
+	*endPtr = TCL_INDEX_NONE;
     } else {
 	*startPtr = regexpPtr->matches[index].rm_so;
 	*endPtr = regexpPtr->matches[index].rm_eo;
@@ -677,7 +677,7 @@ TclRegAbout(
 
     resultObj = Tcl_NewObj();
     Tcl_ListObjAppendElement(NULL, resultObj,
-	    Tcl_NewWideIntObj((Tcl_WideInt) regexpPtr->re.re_nsub));
+	    TclNewWideIntObjFromSize(regexpPtr->re.re_nsub));
 
     /*
      * Now append a list of all the bit-flags set for the RE.
diff --git a/generic/tclStringObj.c b/generic/tclStringObj.c
index a18c13e..f74d13d 100644
--- a/generic/tclStringObj.c
+++ b/generic/tclStringObj.c
@@ -1404,7 +1404,7 @@ AppendUnicodeToUnicodeRep(
     numChars = stringPtr->numChars + appendNumChars;
 
     if (numChars > stringPtr->maxChars) {
-	size_t offset = TCL_AUTO_LENGTH;
+	size_t index = TCL_INDEX_NONE;
 
 	/*
 	 * Protect against case where unicode points into the existing
@@ -1414,7 +1414,7 @@ AppendUnicodeToUnicodeRep(
 
 	if (unicode && unicode >= stringPtr->unicode
 		&& unicode <= stringPtr->unicode + stringPtr->maxChars) {
-	    offset = unicode - stringPtr->unicode;
+	    index = unicode - stringPtr->unicode;
 	}
 
 	GrowUnicodeBuffer(objPtr, numChars);
@@ -1424,8 +1424,8 @@ AppendUnicodeToUnicodeRep(
 	 * Relocate unicode if needed; see above.
 	 */
 
-	if (offset != TCL_AUTO_LENGTH) {
-	    unicode = stringPtr->unicode + offset;
+	if (index != TCL_INDEX_NONE) {
+	    unicode = stringPtr->unicode + index;
 	}
     }
 
diff --git a/generic/tclTest.c b/generic/tclTest.c
index 450d663..1a975e2 100644
--- a/generic/tclTest.c
+++ b/generic/tclTest.c
@@ -345,7 +345,7 @@ static int		TestreturnObjCmd(void *dummy,
 			    Tcl_Interp *interp, int objc,
 			    Tcl_Obj *const objv[]);
 static void		TestregexpXflags(const char *string,
-			    int length, int *cflagsPtr, int *eflagsPtr);
+			    size_t length, int *cflagsPtr, int *eflagsPtr);
 static int		TestsaveresultCmd(void *dummy,
 			    Tcl_Interp *interp, int objc,
 			    Tcl_Obj *const objv[]);
@@ -4007,12 +4007,12 @@ TestregexpObjCmd(
 	if (objc > 2 && (cflags&REG_EXPECT) && indices) {
 	    const char *varName;
 	    const char *value;
-	    int start, end;
+	    size_t 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);
+	    sprintf(resinfo, "%" TCL_LL_MODIFIER "d %" TCL_LL_MODIFIER "d", TclWideIntFromSize(start), TclWideIntFromSize(end-1));
 	    value = Tcl_SetVar2(interp, varName, NULL, resinfo, 0);
 	    if (value == NULL) {
 		Tcl_AppendResult(interp, "couldn't set variable \"",
@@ -4026,7 +4026,7 @@ TestregexpObjCmd(
 
 	    Tcl_RegExpGetInfo(regExpr, &info);
 	    varName = Tcl_GetString(objv[2]);
-	    sprintf(resinfo, "%ld", info.extendStart);
+	    sprintf(resinfo, "%" TCL_LL_MODIFIER "d", TclWideIntFromSize(info.extendStart));
 	    value = Tcl_SetVar2(interp, varName, NULL, resinfo, 0);
 	    if (value == NULL) {
 		Tcl_AppendResult(interp, "couldn't set variable \"",
@@ -4047,7 +4047,7 @@ TestregexpObjCmd(
 
     Tcl_RegExpGetInfo(regExpr, &info);
     for (i = 0; i < objc; i++) {
-	int start, end;
+	size_t start, end;
 	Tcl_Obj *newPtr, *varPtr, *valuePtr;
 
 	varPtr = objv[i];
@@ -4057,9 +4057,9 @@ TestregexpObjCmd(
 
 	    if (ii == TCL_INDEX_NONE) {
 		TclRegExpRangeUniChar(regExpr, ii, &start, &end);
-	    } else if (ii > (size_t)info.nsubs) {
-		start = -1;
-		end = -1;
+	    } else if (ii > info.nsubs) {
+		start = TCL_INDEX_NONE;
+		end = TCL_INDEX_NONE;
 	    } else {
 		start = info.matches[ii].start;
 		end = info.matches[ii].end;
@@ -4070,19 +4070,19 @@ TestregexpObjCmd(
 	     * instead of the first character after the match.
 	     */
 
-	    if (end >= 0) {
+	    if (end != TCL_INDEX_NONE) {
 		end--;
 	    }
 
-	    objs[0] = Tcl_NewWideIntObj(start);
-	    objs[1] = Tcl_NewWideIntObj(end);
+	    objs[0] = TclNewWideIntObjFromSize(start);
+	    objs[1] = TclNewWideIntObjFromSize(end);
 
 	    newPtr = Tcl_NewListObj(2, objs);
 	} else {
 	    if (ii == TCL_INDEX_NONE) {
 		TclRegExpRangeUniChar(regExpr, ii, &start, &end);
 		newPtr = Tcl_GetRange(objPtr, start, end);
-	    } else if (ii > (size_t)info.nsubs) {
+	    } else if (ii > info.nsubs) {
 		newPtr = Tcl_NewObj();
 	    } else {
 		newPtr = Tcl_GetRange(objPtr, info.matches[ii].start,
@@ -4123,11 +4123,12 @@ TestregexpObjCmd(
 static void
 TestregexpXflags(
     const char *string,	/* The string of flags. */
-    int length,			/* The length of the string in bytes. */
+    size_t length,			/* The length of the string in bytes. */
     int *cflagsPtr,		/* compile flags word */
     int *eflagsPtr)		/* exec flags word */
 {
-    int i, cflags, eflags;
+    size_t i;
+    int cflags, eflags;
 
     cflags = *cflagsPtr;
     eflags = *eflagsPtr;
diff --git a/generic/tclUtf.c b/generic/tclUtf.c
index 6a40856..5177fd0 100644
--- a/generic/tclUtf.c
+++ b/generic/tclUtf.c
@@ -1027,7 +1027,7 @@ Tcl_UtfAtIndex(
     size_t len = 0;
 #endif
 
-    if (index != TCL_AUTO_LENGTH) {
+    if (index != TCL_INDEX_NONE) {
 	while (index--) {
 #if TCL_UTF_MAX <= 4
 	    src += (len = TclUtfToUniChar(src, &ch));
diff --git a/generic/tclUtil.c b/generic/tclUtil.c
index 7373085..c9eb8d2 100644
--- a/generic/tclUtil.c
+++ b/generic/tclUtil.c
@@ -2637,10 +2637,10 @@ Tcl_DStringInit(
 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. */
+    const char *bytes,		/* String to append. If length is
+				 * TCL_AUTO_LENGTH then this must be null-terminated. */
     size_t length)			/* Number of bytes from "bytes" to append. If
-				 * -1, then append all of bytes, up to null
+				 * TCL_AUTO_LENGTH, then append all of bytes, up to null
 				 * at end. */
 {
     size_t newSize;
@@ -2664,18 +2664,18 @@ Tcl_DStringAppend(
 	    memcpy(newString, dsPtr->string, dsPtr->length);
 	    dsPtr->string = newString;
 	} else {
-	    size_t offset = TCL_AUTO_LENGTH;
+	    size_t index = TCL_INDEX_NONE;
 
 	    /* See [16896d49fd] */
 	    if (bytes >= dsPtr->string
 		    && bytes <= dsPtr->string + dsPtr->length) {
-		offset = bytes - dsPtr->string;
+		index = bytes - dsPtr->string;
 	    }
 
 	    dsPtr->string = Tcl_Realloc(dsPtr->string, dsPtr->spaceAvl);
 
-	    if (offset != TCL_AUTO_LENGTH) {
-		bytes = dsPtr->string + offset;
+	    if (index != TCL_INDEX_NONE) {
+		bytes = dsPtr->string + index;
 	    }
 	}
     }