summaryrefslogtreecommitdiffstats
path: root/generic/regcomp.c
diff options
context:
space:
mode:
Diffstat (limited to 'generic/regcomp.c')
-rw-r--r--generic/regcomp.c107
1 files changed, 64 insertions, 43 deletions
diff --git a/generic/regcomp.c b/generic/regcomp.c
index 983cd7a..2bc0744 100644
--- a/generic/regcomp.c
+++ b/generic/regcomp.c
@@ -2,7 +2,7 @@
* re_*comp and friends - compile REs
* This file #includes several others (see the bottom).
*
- * Copyright © 1998, 1999 Henry Spencer. All rights reserved.
+ * 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
@@ -56,9 +56,10 @@ 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 *sub_re(struct vars *, int, int, struct state *, struct state *);
+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 *);
@@ -78,10 +79,13 @@ 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 chr lexdigits(struct vars *, int, int, int);
static int brenext(struct vars *, pchr);
static void skip(struct vars *);
-static chr newline(void);
+static chr newline(NOPARMS);
+#ifdef REG_DEBUG
+static const chr *ch(NOPARMS);
+#endif
static chr chrnamed(struct vars *, const chr *, const chr *, pchr);
/* === regc_color.c === */
static void initcm(struct vars *, struct colormap *);
@@ -205,11 +209,11 @@ struct vars {
int cflags; /* copy of compile flags */
int lasttype; /* type of previous token */
int nexttype; /* type of next token */
- int nextvalue; /* value (if any) 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 */
- int nsubs; /* length of 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 */
@@ -243,7 +247,6 @@ struct vars {
#define EMPTYARC(x, y) newarc(v->nfa, EMPTY, 0, x, y)
/* token type codes, some also used as NFA arc types */
-#undef DIGIT /* prevent conflict with libtommath */
#define EMPTY 'n' /* no token present */
#define EOS 'e' /* end of string */
#define PLAIN 'p' /* ordinary character */
@@ -287,7 +290,8 @@ compile(
{
AllocVars(v);
struct guts *g;
- int i, j;
+ int i;
+ size_t j;
FILE *debug = (flags&REG_PROGRESS) ? stdout : NULL;
#define CNOERR() { if (ISERR()) return freev(v, v->err); }
@@ -340,13 +344,13 @@ compile(
re->re_info = 0; /* bits get set during parse */
re->re_csize = sizeof(chr);
re->re_guts = NULL;
- re->re_fns = (void*)(&functions);
+ re->re_fns = VS(&functions);
/*
* More complex setup, malloced things.
*/
- re->re_guts = (void*)(MALLOC(sizeof(struct guts)));
+ re->re_guts = VS(MALLOC(sizeof(struct guts)));
if (re->re_guts == NULL) {
return freev(v, REG_ESPACE);
}
@@ -394,6 +398,7 @@ compile(
dumpnfa(v->nfa, debug);
dumpst(v->tree, debug, 1);
}
+ optst(v, v->tree);
v->ntree = numst(v->tree, 1);
markst(v->tree);
cleanst(v);
@@ -432,7 +437,7 @@ compile(
* Can sacrifice main NFA now, so use it as work area.
*/
- (void) optimize(v->nfa, debug);
+ (DISCARD) optimize(v->nfa, debug);
CNOERR();
makesearch(v, v->nfa);
CNOERR();
@@ -475,10 +480,10 @@ moresubs(
int wanted) /* want enough room for this one */
{
struct subre **p;
- int n;
+ size_t n;
- assert(wanted > 0 && wanted >= v->nsubs);
- n = wanted * 3 / 2 + 1;
+ 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) {
@@ -497,7 +502,7 @@ moresubs(
*p = NULL;
}
assert(v->nsubs == n);
- assert(wanted < v->nsubs);
+ assert((size_t)wanted < v->nsubs);
}
/*
@@ -511,7 +516,7 @@ freev(
struct vars *v,
int err)
{
- int ret;
+ register int ret;
if (v->re != NULL) {
rfree(v->re);
@@ -663,7 +668,7 @@ parse(
assert(stopper == ')' || stopper == EOS);
- branches = sub_re(v, '|', LONGER, init, final);
+ branches = subre(v, '|', LONGER, init, final);
NOERRN();
branch = branches;
firstbranch = 1;
@@ -673,7 +678,7 @@ parse(
* Need a place to hang the branch.
*/
- branch->right = sub_re(v, '|', LONGER, init, final);
+ branch->right = subre(v, '|', LONGER, init, final);
NOERRN();
branch = branch->right;
}
@@ -744,7 +749,7 @@ parsebranch(
lp = left;
seencontent = 0;
- t = sub_re(v, '=', 0, left, right); /* op '=' is tentative */
+ t = subre(v, '=', 0, left, right); /* op '=' is tentative */
NOERRN();
while (!SEE('|') && !SEE(stopper) && !SEE(EOS)) {
if (seencontent) { /* implicit concat operator */
@@ -808,7 +813,7 @@ parseqatom(
atom = NULL;
assert(lp->nouts == 0); /* must string new code */
assert(rp->nins == 0); /* between lp and rp */
- subno = 0;
+ subno = 0; /* just to shut lint up */
/*
* An atom or constraint...
@@ -921,7 +926,7 @@ parseqatom(
*/
NOTE(REG_UPBOTCH);
- /* FALLTHRU */
+ /* fallthrough into case PLAIN */
case PLAIN:
onechr(v, v->nextvalue, lp, rp);
okcolors(v->nfa, v->cm);
@@ -952,10 +957,10 @@ parseqatom(
if (cap) {
v->nsubexp++;
subno = v->nsubexp;
- if (subno >= v->nsubs) {
+ if ((size_t)subno >= v->nsubs) {
moresubs(v, subno);
}
- assert(subno < v->nsubs);
+ assert((size_t)subno < v->nsubs);
} else {
atomtype = PLAIN; /* something that's not '(' */
}
@@ -977,7 +982,7 @@ parseqatom(
NOERR();
if (cap) {
v->subs[subno] = atom;
- t = sub_re(v, '(', atom->flags|CAP, lp, rp);
+ t = subre(v, '(', atom->flags|CAP, lp, rp);
NOERR();
t->subno = subno;
t->left = atom;
@@ -995,7 +1000,7 @@ parseqatom(
INSIST(v->subs[v->nextvalue] != NULL, REG_ESUBREG);
NOERR();
assert(v->nextvalue > 0);
- atom = sub_re(v, 'b', BACKR, lp, rp);
+ atom = subre(v, 'b', BACKR, lp, rp);
NOERR();
subno = v->nextvalue;
atom->subno = subno;
@@ -1110,7 +1115,7 @@ parseqatom(
*/
if (atom == NULL) {
- atom = sub_re(v, '=', 0, lp, rp);
+ atom = subre(v, '=', 0, lp, rp);
NOERR();
}
@@ -1147,7 +1152,7 @@ parseqatom(
* Break remaining subRE into x{...} and what follows.
*/
- t = sub_re(v, '.', COMBINE(qprefer, atom->flags), lp, rp);
+ t = subre(v, '.', COMBINE(qprefer, atom->flags), lp, rp);
NOERR();
t->left = atom;
atomp = &t->left;
@@ -1161,7 +1166,7 @@ parseqatom(
*/
assert(top->op == '=' && top->left == NULL && top->right == NULL);
- top->left = sub_re(v, '=', top->flags, top->begin, lp);
+ top->left = subre(v, '=', top->flags, top->begin, lp);
NOERR();
top->op = '.';
top->right = t;
@@ -1230,9 +1235,9 @@ parseqatom(
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 = sub_re(v, '.', f, s, atom->end); /* prefix and atom */
+ t = subre(v, '.', f, s, atom->end); /* prefix and atom */
NOERR();
- t->left = sub_re(v, '=', PREF(f), s, atom->begin);
+ t->left = subre(v, '=', PREF(f), s, atom->begin);
NOERR();
t->right = atom;
*atomp = t;
@@ -1247,7 +1252,7 @@ parseqatom(
dupnfa(v->nfa, atom->begin, atom->end, s, s2);
repeat(v, s, s2, m, n);
f = COMBINE(qprefer, atom->flags);
- t = sub_re(v, '*', f, s, s2);
+ t = subre(v, '*', f, s, s2);
NOERR();
t->min = (short) m;
t->max = (short) n;
@@ -1265,7 +1270,7 @@ parseqatom(
t->right = parsebranch(v, stopper, type, s2, rp, 1);
} else {
EMPTYARC(s2, rp);
- t->right = sub_re(v, '=', 0, s2, rp);
+ t->right = subre(v, '=', 0, s2, rp);
}
NOERR();
assert(SEE('|') || SEE(stopper) || SEE(EOS));
@@ -1717,12 +1722,12 @@ wordchrs(
}
/*
- - sub_re - allocate a subre
- ^ static struct subre *sub_re(struct vars *, int, int, struct state *,
+ - subre - allocate a subre
+ ^ static struct subre *subre(struct vars *, int, int, struct state *,
^ struct state *);
*/
static struct subre *
-sub_re(
+subre(
struct vars *v,
int op,
int flags,
@@ -1810,6 +1815,25 @@ freesrnode(
}
/*
+ - 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);
*/
@@ -1899,10 +1923,10 @@ nfatree(
assert(t != NULL && t->begin != NULL);
if (t->left != NULL) {
- (void) nfatree(v, t->left, f);
+ (DISCARD) nfatree(v, t->left, f);
}
if (t->right != NULL) {
- (void) nfatree(v, t->right, f);
+ (DISCARD) nfatree(v, t->right, f);
}
return nfanode(v, t, f);
@@ -2080,9 +2104,6 @@ dump(
}
fprintf(f, "\n");
dumpst(g->tree, f, 0);
-#else
- (void)re;
- (void)f;
#endif
}
@@ -2146,7 +2167,7 @@ stdump(
fprintf(f, "}");
}
if (nfapresent) {
- fprintf(f, " %d-%d", t->begin->no, t->end->no);
+ 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)));
@@ -2185,9 +2206,9 @@ stid(
return "unable";
}
if (t->id != 0) {
- snprintf(buf, bufsize, "%d", t->id);
+ sprintf(buf, "%d", t->id);
} else {
- snprintf(buf, bufsize, "%p", t);
+ sprintf(buf, "%p", t);
}
return buf;
}
generated by cgit v0.12 at 2025-12-20 13:00:27 (GMT)