blob: 9ae66f3fa5ff290df3b30f4495676bcddcde2653 [file] [log] [blame]
Torok Edwince0c81e2009-08-30 08:24:09 +00001/*-
2 * This code is derived from OpenBSD's libc/regex, original license follows:
3 *
4 * Copyright (c) 1992, 1993, 1994 Henry Spencer.
5 * Copyright (c) 1992, 1993, 1994
6 * The Regents of the University of California. All rights reserved.
7 *
8 * This code is derived from software contributed to Berkeley by
9 * Henry Spencer.
10 *
11 * Redistribution and use in source and binary forms, with or without
12 * modification, are permitted provided that the following conditions
13 * are met:
14 * 1. Redistributions of source code must retain the above copyright
15 * notice, this list of conditions and the following disclaimer.
16 * 2. Redistributions in binary form must reproduce the above copyright
17 * notice, this list of conditions and the following disclaimer in the
18 * documentation and/or other materials provided with the distribution.
19 * 3. Neither the name of the University nor the names of its contributors
20 * may be used to endorse or promote products derived from this software
21 * without specific prior written permission.
22 *
23 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33 * SUCH DAMAGE.
34 *
35 * @(#)regcomp.c 8.5 (Berkeley) 3/20/94
36 */
37
38#include <sys/types.h>
39#include <stdio.h>
40#include <string.h>
41#include <ctype.h>
42#include <limits.h>
43#include <stdlib.h>
44#include "regex_impl.h"
45
46#include "regutils.h"
47#include "regex2.h"
48
49#include "regcclass.h"
50#include "regcname.h"
51
52/*
53 * parse structure, passed up and down to avoid global variables and
54 * other clumsinesses
55 */
56struct parse {
57 char *next; /* next character in RE */
58 char *end; /* end of string (-> NUL normally) */
59 int error; /* has an error been seen? */
60 sop *strip; /* malloced strip */
61 sopno ssize; /* malloced strip size (allocated) */
62 sopno slen; /* malloced strip length (used) */
63 int ncsalloc; /* number of csets allocated */
64 struct re_guts *g;
65# define NPAREN 10 /* we need to remember () 1-9 for back refs */
66 sopno pbegin[NPAREN]; /* -> ( ([0] unused) */
67 sopno pend[NPAREN]; /* -> ) ([0] unused) */
68};
69
70static void p_ere(struct parse *, int);
71static void p_ere_exp(struct parse *);
72static void p_str(struct parse *);
73static void p_bre(struct parse *, int, int);
74static int p_simp_re(struct parse *, int);
75static int p_count(struct parse *);
76static void p_bracket(struct parse *);
77static void p_b_term(struct parse *, cset *);
78static void p_b_cclass(struct parse *, cset *);
79static void p_b_eclass(struct parse *, cset *);
80static char p_b_symbol(struct parse *);
81static char p_b_coll_elem(struct parse *, int);
82static char othercase(int);
83static void bothcases(struct parse *, int);
84static void ordinary(struct parse *, int);
85static void nonnewline(struct parse *);
86static void repeat(struct parse *, sopno, int, int);
87static int seterr(struct parse *, int);
88static cset *allocset(struct parse *);
89static void freeset(struct parse *, cset *);
90static int freezeset(struct parse *, cset *);
91static int firstch(struct parse *, cset *);
92static int nch(struct parse *, cset *);
93static void mcadd(struct parse *, cset *, const char *);
94static void mcinvert(struct parse *, cset *);
95static void mccase(struct parse *, cset *);
96static int isinsets(struct re_guts *, int);
97static int samesets(struct re_guts *, int, int);
98static void categorize(struct parse *, struct re_guts *);
99static sopno dupl(struct parse *, sopno, sopno);
100static void doemit(struct parse *, sop, size_t);
101static void doinsert(struct parse *, sop, size_t, sopno);
102static void dofwd(struct parse *, sopno, sop);
103static void enlarge(struct parse *, sopno);
104static void stripsnug(struct parse *, struct re_guts *);
105static void findmust(struct parse *, struct re_guts *);
106static sopno pluscount(struct parse *, struct re_guts *);
107
108static char nuls[10]; /* place to point scanner in event of error */
109
110/*
111 * macros for use with parse structure
112 * BEWARE: these know that the parse structure is named `p' !!!
113 */
114#define PEEK() (*p->next)
115#define PEEK2() (*(p->next+1))
116#define MORE() (p->next < p->end)
117#define MORE2() (p->next+1 < p->end)
118#define SEE(c) (MORE() && PEEK() == (c))
119#define SEETWO(a, b) (MORE() && MORE2() && PEEK() == (a) && PEEK2() == (b))
120#define EAT(c) ((SEE(c)) ? (NEXT(), 1) : 0)
121#define EATTWO(a, b) ((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
122#define NEXT() (p->next++)
123#define NEXT2() (p->next += 2)
124#define NEXTn(n) (p->next += (n))
125#define GETNEXT() (*p->next++)
126#define SETERROR(e) seterr(p, (e))
127#define REQUIRE(co, e) (void)((co) || SETERROR(e))
128#define MUSTSEE(c, e) (REQUIRE(MORE() && PEEK() == (c), e))
129#define MUSTEAT(c, e) (REQUIRE(MORE() && GETNEXT() == (c), e))
130#define MUSTNOTSEE(c, e) (REQUIRE(!MORE() || PEEK() != (c), e))
131#define EMIT(op, sopnd) doemit(p, (sop)(op), (size_t)(sopnd))
132#define INSERT(op, pos) doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
133#define AHEAD(pos) dofwd(p, pos, HERE()-(pos))
134#define ASTERN(sop, pos) EMIT(sop, HERE()-pos)
135#define HERE() (p->slen)
136#define THERE() (p->slen - 1)
137#define THERETHERE() (p->slen - 2)
138#define DROP(n) (p->slen -= (n))
139
140#ifdef _POSIX2_RE_DUP_MAX
141#define DUPMAX _POSIX2_RE_DUP_MAX
142#else
143#define DUPMAX 255
144#endif
145
146#ifndef NDEBUG
147static int never = 0; /* for use in asserts; shuts lint up */
148#else
149#define never 0 /* some <assert.h>s have bugs too */
150#endif
151
152/*
153 - llvm_regcomp - interface for parser and compilation
154 */
155int /* 0 success, otherwise REG_something */
156llvm_regcomp(llvm_regex_t *preg, const char *pattern, int cflags)
157{
158 struct parse pa;
159 struct re_guts *g;
160 struct parse *p = &pa;
161 int i;
162 size_t len;
163#ifdef REDEBUG
164# define GOODFLAGS(f) (f)
165#else
166# define GOODFLAGS(f) ((f)&~REG_DUMP)
167#endif
168
169 cflags = GOODFLAGS(cflags);
170 if ((cflags&REG_EXTENDED) && (cflags&REG_NOSPEC))
171 return(REG_INVARG);
172
173 if (cflags&REG_PEND) {
174 if (preg->re_endp < pattern)
175 return(REG_INVARG);
176 len = preg->re_endp - pattern;
177 } else
178 len = strlen((const char *)pattern);
179
180 /* do the mallocs early so failure handling is easy */
181 g = (struct re_guts *)malloc(sizeof(struct re_guts) +
182 (NC-1)*sizeof(cat_t));
183 if (g == NULL)
184 return(REG_ESPACE);
185 p->ssize = len/(size_t)2*(size_t)3 + (size_t)1; /* ugh */
186 p->strip = (sop *)calloc(p->ssize, sizeof(sop));
187 p->slen = 0;
188 if (p->strip == NULL) {
189 free((char *)g);
190 return(REG_ESPACE);
191 }
192
193 /* set things up */
194 p->g = g;
195 p->next = (char *)pattern; /* convenience; we do not modify it */
196 p->end = p->next + len;
197 p->error = 0;
198 p->ncsalloc = 0;
199 for (i = 0; i < NPAREN; i++) {
200 p->pbegin[i] = 0;
201 p->pend[i] = 0;
202 }
203 g->csetsize = NC;
204 g->sets = NULL;
205 g->setbits = NULL;
206 g->ncsets = 0;
207 g->cflags = cflags;
208 g->iflags = 0;
209 g->nbol = 0;
210 g->neol = 0;
211 g->must = NULL;
212 g->mlen = 0;
213 g->nsub = 0;
214 g->ncategories = 1; /* category 0 is "everything else" */
215 g->categories = &g->catspace[-(CHAR_MIN)];
216 (void) memset((char *)g->catspace, 0, NC*sizeof(cat_t));
217 g->backrefs = 0;
218
219 /* do it */
220 EMIT(OEND, 0);
221 g->firststate = THERE();
222 if (cflags&REG_EXTENDED)
223 p_ere(p, OUT);
224 else if (cflags&REG_NOSPEC)
225 p_str(p);
226 else
227 p_bre(p, OUT, OUT);
228 EMIT(OEND, 0);
229 g->laststate = THERE();
230
231 /* tidy up loose ends and fill things in */
232 categorize(p, g);
233 stripsnug(p, g);
234 findmust(p, g);
235 g->nplus = pluscount(p, g);
236 g->magic = MAGIC2;
237 preg->re_nsub = g->nsub;
238 preg->re_g = g;
239 preg->re_magic = MAGIC1;
240#ifndef REDEBUG
241 /* not debugging, so can't rely on the assert() in llvm_regexec() */
242 if (g->iflags&REGEX_BAD)
243 SETERROR(REG_ASSERT);
244#endif
245
246 /* win or lose, we're done */
247 if (p->error != 0) /* lose */
248 llvm_regfree(preg);
249 return(p->error);
250}
251
252/*
253 - p_ere - ERE parser top level, concatenation and alternation
254 */
255static void
256p_ere(struct parse *p, int stop) /* character this ERE should end at */
257{
258 char c;
Benjamin Kramer5b7c3562009-09-06 09:29:39 +0000259 sopno prevback;
260 sopno prevfwd;
Torok Edwince0c81e2009-08-30 08:24:09 +0000261 sopno conc;
262 int first = 1; /* is this the first alternative? */
263
264 for (;;) {
265 /* do a bunch of concatenated expressions */
266 conc = HERE();
267 while (MORE() && (c = PEEK()) != '|' && c != stop)
268 p_ere_exp(p);
269 REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
270
271 if (!EAT('|'))
272 break; /* NOTE BREAK OUT */
273
274 if (first) {
275 INSERT(OCH_, conc); /* offset is wrong */
276 prevfwd = conc;
277 prevback = conc;
278 first = 0;
279 }
280 ASTERN(OOR1, prevback);
281 prevback = THERE();
282 AHEAD(prevfwd); /* fix previous offset */
283 prevfwd = HERE();
284 EMIT(OOR2, 0); /* offset is very wrong */
285 }
286
287 if (!first) { /* tail-end fixups */
288 AHEAD(prevfwd);
289 ASTERN(O_CH, prevback);
290 }
291
292 assert(!MORE() || SEE(stop));
293}
294
295/*
296 - p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
297 */
298static void
299p_ere_exp(struct parse *p)
300{
301 char c;
302 sopno pos;
303 int count;
304 int count2;
305 sopno subno;
306 int wascaret = 0;
307
308 assert(MORE()); /* caller should have ensured this */
309 c = GETNEXT();
310
311 pos = HERE();
312 switch (c) {
313 case '(':
314 REQUIRE(MORE(), REG_EPAREN);
315 p->g->nsub++;
316 subno = p->g->nsub;
317 if (subno < NPAREN)
318 p->pbegin[subno] = HERE();
319 EMIT(OLPAREN, subno);
320 if (!SEE(')'))
321 p_ere(p, ')');
322 if (subno < NPAREN) {
323 p->pend[subno] = HERE();
324 assert(p->pend[subno] != 0);
325 }
326 EMIT(ORPAREN, subno);
327 MUSTEAT(')', REG_EPAREN);
328 break;
329#ifndef POSIX_MISTAKE
330 case ')': /* happens only if no current unmatched ( */
331 /*
332 * You may ask, why the ifndef? Because I didn't notice
333 * this until slightly too late for 1003.2, and none of the
334 * other 1003.2 regular-expression reviewers noticed it at
335 * all. So an unmatched ) is legal POSIX, at least until
336 * we can get it fixed.
337 */
338 SETERROR(REG_EPAREN);
339 break;
340#endif
341 case '^':
342 EMIT(OBOL, 0);
343 p->g->iflags |= USEBOL;
344 p->g->nbol++;
345 wascaret = 1;
346 break;
347 case '$':
348 EMIT(OEOL, 0);
349 p->g->iflags |= USEEOL;
350 p->g->neol++;
351 break;
352 case '|':
353 SETERROR(REG_EMPTY);
354 break;
355 case '*':
356 case '+':
357 case '?':
358 SETERROR(REG_BADRPT);
359 break;
360 case '.':
361 if (p->g->cflags&REG_NEWLINE)
362 nonnewline(p);
363 else
364 EMIT(OANY, 0);
365 break;
366 case '[':
367 p_bracket(p);
368 break;
369 case '\\':
370 REQUIRE(MORE(), REG_EESCAPE);
371 c = GETNEXT();
372 ordinary(p, c);
373 break;
374 case '{': /* okay as ordinary except if digit follows */
375 REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
376 /* FALLTHROUGH */
377 default:
378 ordinary(p, c);
379 break;
380 }
381
382 if (!MORE())
383 return;
384 c = PEEK();
385 /* we call { a repetition if followed by a digit */
386 if (!( c == '*' || c == '+' || c == '?' ||
387 (c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
388 return; /* no repetition, we're done */
389 NEXT();
390
391 REQUIRE(!wascaret, REG_BADRPT);
392 switch (c) {
393 case '*': /* implemented as +? */
394 /* this case does not require the (y|) trick, noKLUDGE */
395 INSERT(OPLUS_, pos);
396 ASTERN(O_PLUS, pos);
397 INSERT(OQUEST_, pos);
398 ASTERN(O_QUEST, pos);
399 break;
400 case '+':
401 INSERT(OPLUS_, pos);
402 ASTERN(O_PLUS, pos);
403 break;
404 case '?':
405 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
406 INSERT(OCH_, pos); /* offset slightly wrong */
407 ASTERN(OOR1, pos); /* this one's right */
408 AHEAD(pos); /* fix the OCH_ */
409 EMIT(OOR2, 0); /* offset very wrong... */
410 AHEAD(THERE()); /* ...so fix it */
411 ASTERN(O_CH, THERETHERE());
412 break;
413 case '{':
414 count = p_count(p);
415 if (EAT(',')) {
416 if (isdigit((uch)PEEK())) {
417 count2 = p_count(p);
418 REQUIRE(count <= count2, REG_BADBR);
419 } else /* single number with comma */
420 count2 = INFINITY;
421 } else /* just a single number */
422 count2 = count;
423 repeat(p, pos, count, count2);
424 if (!EAT('}')) { /* error heuristics */
425 while (MORE() && PEEK() != '}')
426 NEXT();
427 REQUIRE(MORE(), REG_EBRACE);
428 SETERROR(REG_BADBR);
429 }
430 break;
431 }
432
433 if (!MORE())
434 return;
435 c = PEEK();
436 if (!( c == '*' || c == '+' || c == '?' ||
437 (c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
438 return;
439 SETERROR(REG_BADRPT);
440}
441
442/*
443 - p_str - string (no metacharacters) "parser"
444 */
445static void
446p_str(struct parse *p)
447{
448 REQUIRE(MORE(), REG_EMPTY);
449 while (MORE())
450 ordinary(p, GETNEXT());
451}
452
453/*
454 - p_bre - BRE parser top level, anchoring and concatenation
455 * Giving end1 as OUT essentially eliminates the end1/end2 check.
456 *
457 * This implementation is a bit of a kludge, in that a trailing $ is first
458 * taken as an ordinary character and then revised to be an anchor. The
459 * only undesirable side effect is that '$' gets included as a character
460 * category in such cases. This is fairly harmless; not worth fixing.
461 * The amount of lookahead needed to avoid this kludge is excessive.
462 */
463static void
464p_bre(struct parse *p,
465 int end1, /* first terminating character */
466 int end2) /* second terminating character */
467{
468 sopno start = HERE();
469 int first = 1; /* first subexpression? */
470 int wasdollar = 0;
471
472 if (EAT('^')) {
473 EMIT(OBOL, 0);
474 p->g->iflags |= USEBOL;
475 p->g->nbol++;
476 }
477 while (MORE() && !SEETWO(end1, end2)) {
478 wasdollar = p_simp_re(p, first);
479 first = 0;
480 }
481 if (wasdollar) { /* oops, that was a trailing anchor */
482 DROP(1);
483 EMIT(OEOL, 0);
484 p->g->iflags |= USEEOL;
485 p->g->neol++;
486 }
487
488 REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
489}
490
491/*
492 - p_simp_re - parse a simple RE, an atom possibly followed by a repetition
493 */
494static int /* was the simple RE an unbackslashed $? */
495p_simp_re(struct parse *p,
496 int starordinary) /* is a leading * an ordinary character? */
497{
498 int c;
499 int count;
500 int count2;
501 sopno pos;
502 int i;
503 sopno subno;
504# define BACKSL (1<<CHAR_BIT)
505
506 pos = HERE(); /* repetion op, if any, covers from here */
507
508 assert(MORE()); /* caller should have ensured this */
509 c = GETNEXT();
510 if (c == '\\') {
511 REQUIRE(MORE(), REG_EESCAPE);
512 c = BACKSL | GETNEXT();
513 }
514 switch (c) {
515 case '.':
516 if (p->g->cflags&REG_NEWLINE)
517 nonnewline(p);
518 else
519 EMIT(OANY, 0);
520 break;
521 case '[':
522 p_bracket(p);
523 break;
524 case BACKSL|'{':
525 SETERROR(REG_BADRPT);
526 break;
527 case BACKSL|'(':
528 p->g->nsub++;
529 subno = p->g->nsub;
530 if (subno < NPAREN)
531 p->pbegin[subno] = HERE();
532 EMIT(OLPAREN, subno);
533 /* the MORE here is an error heuristic */
534 if (MORE() && !SEETWO('\\', ')'))
535 p_bre(p, '\\', ')');
536 if (subno < NPAREN) {
537 p->pend[subno] = HERE();
538 assert(p->pend[subno] != 0);
539 }
540 EMIT(ORPAREN, subno);
541 REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
542 break;
543 case BACKSL|')': /* should not get here -- must be user */
544 case BACKSL|'}':
545 SETERROR(REG_EPAREN);
546 break;
547 case BACKSL|'1':
548 case BACKSL|'2':
549 case BACKSL|'3':
550 case BACKSL|'4':
551 case BACKSL|'5':
552 case BACKSL|'6':
553 case BACKSL|'7':
554 case BACKSL|'8':
555 case BACKSL|'9':
556 i = (c&~BACKSL) - '0';
557 assert(i < NPAREN);
558 if (p->pend[i] != 0) {
559 assert(i <= p->g->nsub);
560 EMIT(OBACK_, i);
561 assert(p->pbegin[i] != 0);
562 assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
563 assert(OP(p->strip[p->pend[i]]) == ORPAREN);
564 (void) dupl(p, p->pbegin[i]+1, p->pend[i]);
565 EMIT(O_BACK, i);
566 } else
567 SETERROR(REG_ESUBREG);
568 p->g->backrefs = 1;
569 break;
570 case '*':
571 REQUIRE(starordinary, REG_BADRPT);
572 /* FALLTHROUGH */
573 default:
574 ordinary(p, (char)c);
575 break;
576 }
577
578 if (EAT('*')) { /* implemented as +? */
579 /* this case does not require the (y|) trick, noKLUDGE */
580 INSERT(OPLUS_, pos);
581 ASTERN(O_PLUS, pos);
582 INSERT(OQUEST_, pos);
583 ASTERN(O_QUEST, pos);
584 } else if (EATTWO('\\', '{')) {
585 count = p_count(p);
586 if (EAT(',')) {
587 if (MORE() && isdigit((uch)PEEK())) {
588 count2 = p_count(p);
589 REQUIRE(count <= count2, REG_BADBR);
590 } else /* single number with comma */
591 count2 = INFINITY;
592 } else /* just a single number */
593 count2 = count;
594 repeat(p, pos, count, count2);
595 if (!EATTWO('\\', '}')) { /* error heuristics */
596 while (MORE() && !SEETWO('\\', '}'))
597 NEXT();
598 REQUIRE(MORE(), REG_EBRACE);
599 SETERROR(REG_BADBR);
600 }
601 } else if (c == '$') /* $ (but not \$) ends it */
602 return(1);
603
604 return(0);
605}
606
607/*
608 - p_count - parse a repetition count
609 */
610static int /* the value */
611p_count(struct parse *p)
612{
613 int count = 0;
614 int ndigits = 0;
615
616 while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
617 count = count*10 + (GETNEXT() - '0');
618 ndigits++;
619 }
620
621 REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
622 return(count);
623}
624
625/*
626 - p_bracket - parse a bracketed character list
627 *
628 * Note a significant property of this code: if the allocset() did SETERROR,
629 * no set operations are done.
630 */
631static void
632p_bracket(struct parse *p)
633{
634 cset *cs;
635 int invert = 0;
636
637 /* Dept of Truly Sickening Special-Case Kludges */
638 if (p->next + 5 < p->end && strncmp(p->next, "[:<:]]", 6) == 0) {
639 EMIT(OBOW, 0);
640 NEXTn(6);
641 return;
642 }
643 if (p->next + 5 < p->end && strncmp(p->next, "[:>:]]", 6) == 0) {
644 EMIT(OEOW, 0);
645 NEXTn(6);
646 return;
647 }
648
649 if ((cs = allocset(p)) == NULL) {
650 /* allocset did set error status in p */
651 return;
652 }
653
654 if (EAT('^'))
655 invert++; /* make note to invert set at end */
656 if (EAT(']'))
657 CHadd(cs, ']');
658 else if (EAT('-'))
659 CHadd(cs, '-');
660 while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
661 p_b_term(p, cs);
662 if (EAT('-'))
663 CHadd(cs, '-');
664 MUSTEAT(']', REG_EBRACK);
665
666 if (p->error != 0) { /* don't mess things up further */
667 freeset(p, cs);
668 return;
669 }
670
671 if (p->g->cflags&REG_ICASE) {
672 int i;
673 int ci;
674
675 for (i = p->g->csetsize - 1; i >= 0; i--)
676 if (CHIN(cs, i) && isalpha(i)) {
677 ci = othercase(i);
678 if (ci != i)
679 CHadd(cs, ci);
680 }
681 if (cs->multis != NULL)
682 mccase(p, cs);
683 }
684 if (invert) {
685 int i;
686
687 for (i = p->g->csetsize - 1; i >= 0; i--)
688 if (CHIN(cs, i))
689 CHsub(cs, i);
690 else
691 CHadd(cs, i);
692 if (p->g->cflags&REG_NEWLINE)
693 CHsub(cs, '\n');
694 if (cs->multis != NULL)
695 mcinvert(p, cs);
696 }
697
698 assert(cs->multis == NULL); /* xxx */
699
700 if (nch(p, cs) == 1) { /* optimize singleton sets */
701 ordinary(p, firstch(p, cs));
702 freeset(p, cs);
703 } else
704 EMIT(OANYOF, freezeset(p, cs));
705}
706
707/*
708 - p_b_term - parse one term of a bracketed character list
709 */
710static void
711p_b_term(struct parse *p, cset *cs)
712{
713 char c;
714 char start, finish;
715 int i;
716
717 /* classify what we've got */
718 switch ((MORE()) ? PEEK() : '\0') {
719 case '[':
720 c = (MORE2()) ? PEEK2() : '\0';
721 break;
722 case '-':
723 SETERROR(REG_ERANGE);
724 return; /* NOTE RETURN */
725 break;
726 default:
727 c = '\0';
728 break;
729 }
730
731 switch (c) {
732 case ':': /* character class */
733 NEXT2();
734 REQUIRE(MORE(), REG_EBRACK);
735 c = PEEK();
736 REQUIRE(c != '-' && c != ']', REG_ECTYPE);
737 p_b_cclass(p, cs);
738 REQUIRE(MORE(), REG_EBRACK);
739 REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
740 break;
741 case '=': /* equivalence class */
742 NEXT2();
743 REQUIRE(MORE(), REG_EBRACK);
744 c = PEEK();
745 REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
746 p_b_eclass(p, cs);
747 REQUIRE(MORE(), REG_EBRACK);
748 REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
749 break;
750 default: /* symbol, ordinary character, or range */
751/* xxx revision needed for multichar stuff */
752 start = p_b_symbol(p);
753 if (SEE('-') && MORE2() && PEEK2() != ']') {
754 /* range */
755 NEXT();
756 if (EAT('-'))
757 finish = '-';
758 else
759 finish = p_b_symbol(p);
760 } else
761 finish = start;
762/* xxx what about signed chars here... */
763 REQUIRE(start <= finish, REG_ERANGE);
764 for (i = start; i <= finish; i++)
765 CHadd(cs, i);
766 break;
767 }
768}
769
770/*
771 - p_b_cclass - parse a character-class name and deal with it
772 */
773static void
774p_b_cclass(struct parse *p, cset *cs)
775{
776 char *sp = p->next;
777 struct cclass *cp;
778 size_t len;
779 const char *u;
780 char c;
781
782 while (MORE() && isalpha(PEEK()))
783 NEXT();
784 len = p->next - sp;
785 for (cp = cclasses; cp->name != NULL; cp++)
786 if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
787 break;
788 if (cp->name == NULL) {
789 /* oops, didn't find it */
790 SETERROR(REG_ECTYPE);
791 return;
792 }
793
794 u = cp->chars;
795 while ((c = *u++) != '\0')
796 CHadd(cs, c);
797 for (u = cp->multis; *u != '\0'; u += strlen(u) + 1)
798 MCadd(p, cs, u);
799}
800
801/*
802 - p_b_eclass - parse an equivalence-class name and deal with it
803 *
804 * This implementation is incomplete. xxx
805 */
806static void
807p_b_eclass(struct parse *p, cset *cs)
808{
809 char c;
810
811 c = p_b_coll_elem(p, '=');
812 CHadd(cs, c);
813}
814
815/*
816 - p_b_symbol - parse a character or [..]ed multicharacter collating symbol
817 */
818static char /* value of symbol */
819p_b_symbol(struct parse *p)
820{
821 char value;
822
823 REQUIRE(MORE(), REG_EBRACK);
824 if (!EATTWO('[', '.'))
825 return(GETNEXT());
826
827 /* collating symbol */
828 value = p_b_coll_elem(p, '.');
829 REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
830 return(value);
831}
832
833/*
834 - p_b_coll_elem - parse a collating-element name and look it up
835 */
836static char /* value of collating element */
837p_b_coll_elem(struct parse *p,
838 int endc) /* name ended by endc,']' */
839{
840 char *sp = p->next;
841 struct cname *cp;
842 int len;
843
844 while (MORE() && !SEETWO(endc, ']'))
845 NEXT();
846 if (!MORE()) {
847 SETERROR(REG_EBRACK);
848 return(0);
849 }
850 len = p->next - sp;
851 for (cp = cnames; cp->name != NULL; cp++)
852 if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
853 return(cp->code); /* known name */
854 if (len == 1)
855 return(*sp); /* single character */
856 SETERROR(REG_ECOLLATE); /* neither */
857 return(0);
858}
859
860/*
861 - othercase - return the case counterpart of an alphabetic
862 */
863static char /* if no counterpart, return ch */
864othercase(int ch)
865{
866 ch = (uch)ch;
867 assert(isalpha(ch));
868 if (isupper(ch))
869 return ((uch)tolower(ch));
870 else if (islower(ch))
871 return ((uch)toupper(ch));
872 else /* peculiar, but could happen */
873 return(ch);
874}
875
876/*
877 - bothcases - emit a dualcase version of a two-case character
878 *
879 * Boy, is this implementation ever a kludge...
880 */
881static void
882bothcases(struct parse *p, int ch)
883{
884 char *oldnext = p->next;
885 char *oldend = p->end;
886 char bracket[3];
887
888 ch = (uch)ch;
889 assert(othercase(ch) != ch); /* p_bracket() would recurse */
890 p->next = bracket;
891 p->end = bracket+2;
892 bracket[0] = ch;
893 bracket[1] = ']';
894 bracket[2] = '\0';
895 p_bracket(p);
896 assert(p->next == bracket+2);
897 p->next = oldnext;
898 p->end = oldend;
899}
900
901/*
902 - ordinary - emit an ordinary character
903 */
904static void
905ordinary(struct parse *p, int ch)
906{
907 cat_t *cap = p->g->categories;
908
909 if ((p->g->cflags&REG_ICASE) && isalpha((uch)ch) && othercase(ch) != ch)
910 bothcases(p, ch);
911 else {
912 EMIT(OCHAR, (uch)ch);
913 if (cap[ch] == 0)
914 cap[ch] = p->g->ncategories++;
915 }
916}
917
918/*
919 - nonnewline - emit REG_NEWLINE version of OANY
920 *
921 * Boy, is this implementation ever a kludge...
922 */
923static void
924nonnewline(struct parse *p)
925{
926 char *oldnext = p->next;
927 char *oldend = p->end;
928 char bracket[4];
929
930 p->next = bracket;
931 p->end = bracket+3;
932 bracket[0] = '^';
933 bracket[1] = '\n';
934 bracket[2] = ']';
935 bracket[3] = '\0';
936 p_bracket(p);
937 assert(p->next == bracket+3);
938 p->next = oldnext;
939 p->end = oldend;
940}
941
942/*
943 - repeat - generate code for a bounded repetition, recursively if needed
944 */
945static void
946repeat(struct parse *p,
947 sopno start, /* operand from here to end of strip */
948 int from, /* repeated from this number */
949 int to) /* to this number of times (maybe INFINITY) */
950{
951 sopno finish = HERE();
952# define N 2
953# define INF 3
954# define REP(f, t) ((f)*8 + (t))
955# define MAP(n) (((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
956 sopno copy;
957
958 if (p->error != 0) /* head off possible runaway recursion */
959 return;
960
961 assert(from <= to);
962
963 switch (REP(MAP(from), MAP(to))) {
964 case REP(0, 0): /* must be user doing this */
965 DROP(finish-start); /* drop the operand */
966 break;
967 case REP(0, 1): /* as x{1,1}? */
968 case REP(0, N): /* as x{1,n}? */
969 case REP(0, INF): /* as x{1,}? */
970 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
971 INSERT(OCH_, start); /* offset is wrong... */
972 repeat(p, start+1, 1, to);
973 ASTERN(OOR1, start);
974 AHEAD(start); /* ... fix it */
975 EMIT(OOR2, 0);
976 AHEAD(THERE());
977 ASTERN(O_CH, THERETHERE());
978 break;
979 case REP(1, 1): /* trivial case */
980 /* done */
981 break;
982 case REP(1, N): /* as x?x{1,n-1} */
983 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
984 INSERT(OCH_, start);
985 ASTERN(OOR1, start);
986 AHEAD(start);
987 EMIT(OOR2, 0); /* offset very wrong... */
988 AHEAD(THERE()); /* ...so fix it */
989 ASTERN(O_CH, THERETHERE());
990 copy = dupl(p, start+1, finish+1);
991 assert(copy == finish+4);
992 repeat(p, copy, 1, to-1);
993 break;
994 case REP(1, INF): /* as x+ */
995 INSERT(OPLUS_, start);
996 ASTERN(O_PLUS, start);
997 break;
998 case REP(N, N): /* as xx{m-1,n-1} */
999 copy = dupl(p, start, finish);
1000 repeat(p, copy, from-1, to-1);
1001 break;
1002 case REP(N, INF): /* as xx{n-1,INF} */
1003 copy = dupl(p, start, finish);
1004 repeat(p, copy, from-1, to);
1005 break;
1006 default: /* "can't happen" */
1007 SETERROR(REG_ASSERT); /* just in case */
1008 break;
1009 }
1010}
1011
1012/*
1013 - seterr - set an error condition
1014 */
1015static int /* useless but makes type checking happy */
1016seterr(struct parse *p, int e)
1017{
1018 if (p->error == 0) /* keep earliest error condition */
1019 p->error = e;
1020 p->next = nuls; /* try to bring things to a halt */
1021 p->end = nuls;
1022 return(0); /* make the return value well-defined */
1023}
1024
1025/*
1026 - allocset - allocate a set of characters for []
1027 */
1028static cset *
1029allocset(struct parse *p)
1030{
1031 int no = p->g->ncsets++;
1032 size_t nc;
1033 size_t nbytes;
1034 cset *cs;
1035 size_t css = (size_t)p->g->csetsize;
1036 int i;
1037
1038 if (no >= p->ncsalloc) { /* need another column of space */
1039 void *ptr;
1040
1041 p->ncsalloc += CHAR_BIT;
1042 nc = p->ncsalloc;
1043 assert(nc % CHAR_BIT == 0);
1044 nbytes = nc / CHAR_BIT * css;
1045
1046 ptr = (cset *)realloc((char *)p->g->sets, nc * sizeof(cset));
1047 if (ptr == NULL)
1048 goto nomem;
1049 p->g->sets = ptr;
1050
1051 ptr = (uch *)realloc((char *)p->g->setbits, nbytes);
1052 if (ptr == NULL)
1053 goto nomem;
1054 p->g->setbits = ptr;
1055
1056 for (i = 0; i < no; i++)
1057 p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
1058
1059 (void) memset((char *)p->g->setbits + (nbytes - css), 0, css);
1060 }
1061 /* XXX should not happen */
1062 if (p->g->sets == NULL || p->g->setbits == NULL)
1063 goto nomem;
1064
1065 cs = &p->g->sets[no];
1066 cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
1067 cs->mask = 1 << ((no) % CHAR_BIT);
1068 cs->hash = 0;
1069 cs->smultis = 0;
1070 cs->multis = NULL;
1071
1072 return(cs);
1073nomem:
1074 free(p->g->sets);
1075 p->g->sets = NULL;
1076 free(p->g->setbits);
1077 p->g->setbits = NULL;
1078
1079 SETERROR(REG_ESPACE);
1080 /* caller's responsibility not to do set ops */
1081 return(NULL);
1082}
1083
1084/*
1085 - freeset - free a now-unused set
1086 */
1087static void
1088freeset(struct parse *p, cset *cs)
1089{
1090 size_t i;
1091 cset *top = &p->g->sets[p->g->ncsets];
1092 size_t css = (size_t)p->g->csetsize;
1093
1094 for (i = 0; i < css; i++)
1095 CHsub(cs, i);
1096 if (cs == top-1) /* recover only the easy case */
1097 p->g->ncsets--;
1098}
1099
1100/*
1101 - freezeset - final processing on a set of characters
1102 *
1103 * The main task here is merging identical sets. This is usually a waste
1104 * of time (although the hash code minimizes the overhead), but can win
1105 * big if REG_ICASE is being used. REG_ICASE, by the way, is why the hash
1106 * is done using addition rather than xor -- all ASCII [aA] sets xor to
1107 * the same value!
1108 */
1109static int /* set number */
1110freezeset(struct parse *p, cset *cs)
1111{
1112 uch h = cs->hash;
1113 size_t i;
1114 cset *top = &p->g->sets[p->g->ncsets];
1115 cset *cs2;
1116 size_t css = (size_t)p->g->csetsize;
1117
1118 /* look for an earlier one which is the same */
1119 for (cs2 = &p->g->sets[0]; cs2 < top; cs2++)
1120 if (cs2->hash == h && cs2 != cs) {
1121 /* maybe */
1122 for (i = 0; i < css; i++)
1123 if (!!CHIN(cs2, i) != !!CHIN(cs, i))
1124 break; /* no */
1125 if (i == css)
1126 break; /* yes */
1127 }
1128
1129 if (cs2 < top) { /* found one */
1130 freeset(p, cs);
1131 cs = cs2;
1132 }
1133
1134 return((int)(cs - p->g->sets));
1135}
1136
1137/*
1138 - firstch - return first character in a set (which must have at least one)
1139 */
1140static int /* character; there is no "none" value */
1141firstch(struct parse *p, cset *cs)
1142{
1143 size_t i;
1144 size_t css = (size_t)p->g->csetsize;
1145
1146 for (i = 0; i < css; i++)
1147 if (CHIN(cs, i))
1148 return((char)i);
1149 assert(never);
1150 return(0); /* arbitrary */
1151}
1152
1153/*
1154 - nch - number of characters in a set
1155 */
1156static int
1157nch(struct parse *p, cset *cs)
1158{
1159 size_t i;
1160 size_t css = (size_t)p->g->csetsize;
1161 int n = 0;
1162
1163 for (i = 0; i < css; i++)
1164 if (CHIN(cs, i))
1165 n++;
1166 return(n);
1167}
1168
1169/*
1170 - mcadd - add a collating element to a cset
1171 */
1172static void
1173mcadd( struct parse *p, cset *cs, const char *cp)
1174{
1175 size_t oldend = cs->smultis;
1176 void *np;
1177
1178 cs->smultis += strlen(cp) + 1;
1179 np = realloc(cs->multis, cs->smultis);
1180 if (np == NULL) {
1181 if (cs->multis)
1182 free(cs->multis);
1183 cs->multis = NULL;
1184 SETERROR(REG_ESPACE);
1185 return;
1186 }
1187 cs->multis = np;
1188
1189 llvm_strlcpy(cs->multis + oldend - 1, cp, cs->smultis - oldend + 1);
1190}
1191
1192/*
1193 - mcinvert - invert the list of collating elements in a cset
1194 *
1195 * This would have to know the set of possibilities. Implementation
1196 * is deferred.
1197 */
1198/* ARGSUSED */
1199static void
1200mcinvert(struct parse *p, cset *cs)
1201{
1202 assert(cs->multis == NULL); /* xxx */
1203}
1204
1205/*
1206 - mccase - add case counterparts of the list of collating elements in a cset
1207 *
1208 * This would have to know the set of possibilities. Implementation
1209 * is deferred.
1210 */
1211/* ARGSUSED */
1212static void
1213mccase(struct parse *p, cset *cs)
1214{
1215 assert(cs->multis == NULL); /* xxx */
1216}
1217
1218/*
1219 - isinsets - is this character in any sets?
1220 */
1221static int /* predicate */
1222isinsets(struct re_guts *g, int c)
1223{
1224 uch *col;
1225 int i;
1226 int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1227 unsigned uc = (uch)c;
1228
1229 for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1230 if (col[uc] != 0)
1231 return(1);
1232 return(0);
1233}
1234
1235/*
1236 - samesets - are these two characters in exactly the same sets?
1237 */
1238static int /* predicate */
1239samesets(struct re_guts *g, int c1, int c2)
1240{
1241 uch *col;
1242 int i;
1243 int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1244 unsigned uc1 = (uch)c1;
1245 unsigned uc2 = (uch)c2;
1246
1247 for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1248 if (col[uc1] != col[uc2])
1249 return(0);
1250 return(1);
1251}
1252
1253/*
1254 - categorize - sort out character categories
1255 */
1256static void
1257categorize(struct parse *p, struct re_guts *g)
1258{
1259 cat_t *cats = g->categories;
1260 int c;
1261 int c2;
1262 cat_t cat;
1263
1264 /* avoid making error situations worse */
1265 if (p->error != 0)
1266 return;
1267
1268 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
1269 if (cats[c] == 0 && isinsets(g, c)) {
1270 cat = g->ncategories++;
1271 cats[c] = cat;
1272 for (c2 = c+1; c2 <= CHAR_MAX; c2++)
1273 if (cats[c2] == 0 && samesets(g, c, c2))
1274 cats[c2] = cat;
1275 }
1276}
1277
1278/*
1279 - dupl - emit a duplicate of a bunch of sops
1280 */
1281static sopno /* start of duplicate */
1282dupl(struct parse *p,
1283 sopno start, /* from here */
1284 sopno finish) /* to this less one */
1285{
1286 sopno ret = HERE();
1287 sopno len = finish - start;
1288
1289 assert(finish >= start);
1290 if (len == 0)
1291 return(ret);
1292 enlarge(p, p->ssize + len); /* this many unexpected additions */
1293 assert(p->ssize >= p->slen + len);
1294 (void) memmove((char *)(p->strip + p->slen),
1295 (char *)(p->strip + start), (size_t)len*sizeof(sop));
1296 p->slen += len;
1297 return(ret);
1298}
1299
1300/*
1301 - doemit - emit a strip operator
1302 *
1303 * It might seem better to implement this as a macro with a function as
1304 * hard-case backup, but it's just too big and messy unless there are
1305 * some changes to the data structures. Maybe later.
1306 */
1307static void
1308doemit(struct parse *p, sop op, size_t opnd)
1309{
1310 /* avoid making error situations worse */
1311 if (p->error != 0)
1312 return;
1313
1314 /* deal with oversize operands ("can't happen", more or less) */
1315 assert(opnd < 1<<OPSHIFT);
1316
1317 /* deal with undersized strip */
1318 if (p->slen >= p->ssize)
1319 enlarge(p, (p->ssize+1) / 2 * 3); /* +50% */
1320 assert(p->slen < p->ssize);
1321
1322 /* finally, it's all reduced to the easy case */
1323 p->strip[p->slen++] = SOP(op, opnd);
1324}
1325
1326/*
1327 - doinsert - insert a sop into the strip
1328 */
1329static void
1330doinsert(struct parse *p, sop op, size_t opnd, sopno pos)
1331{
1332 sopno sn;
1333 sop s;
1334 int i;
1335
1336 /* avoid making error situations worse */
1337 if (p->error != 0)
1338 return;
1339
1340 sn = HERE();
1341 EMIT(op, opnd); /* do checks, ensure space */
1342 assert(HERE() == sn+1);
1343 s = p->strip[sn];
1344
1345 /* adjust paren pointers */
1346 assert(pos > 0);
1347 for (i = 1; i < NPAREN; i++) {
1348 if (p->pbegin[i] >= pos) {
1349 p->pbegin[i]++;
1350 }
1351 if (p->pend[i] >= pos) {
1352 p->pend[i]++;
1353 }
1354 }
1355
1356 memmove((char *)&p->strip[pos+1], (char *)&p->strip[pos],
1357 (HERE()-pos-1)*sizeof(sop));
1358 p->strip[pos] = s;
1359}
1360
1361/*
1362 - dofwd - complete a forward reference
1363 */
1364static void
1365dofwd(struct parse *p, sopno pos, sop value)
1366{
1367 /* avoid making error situations worse */
1368 if (p->error != 0)
1369 return;
1370
1371 assert(value < 1<<OPSHIFT);
1372 p->strip[pos] = OP(p->strip[pos]) | value;
1373}
1374
1375/*
1376 - enlarge - enlarge the strip
1377 */
1378static void
1379enlarge(struct parse *p, sopno size)
1380{
1381 sop *sp;
1382
1383 if (p->ssize >= size)
1384 return;
1385
1386 sp = (sop *)realloc(p->strip, size*sizeof(sop));
1387 if (sp == NULL) {
1388 SETERROR(REG_ESPACE);
1389 return;
1390 }
1391 p->strip = sp;
1392 p->ssize = size;
1393}
1394
1395/*
1396 - stripsnug - compact the strip
1397 */
1398static void
1399stripsnug(struct parse *p, struct re_guts *g)
1400{
1401 g->nstates = p->slen;
1402 g->strip = (sop *)realloc((char *)p->strip, p->slen * sizeof(sop));
1403 if (g->strip == NULL) {
1404 SETERROR(REG_ESPACE);
1405 g->strip = p->strip;
1406 }
1407}
1408
1409/*
1410 - findmust - fill in must and mlen with longest mandatory literal string
1411 *
1412 * This algorithm could do fancy things like analyzing the operands of |
1413 * for common subsequences. Someday. This code is simple and finds most
1414 * of the interesting cases.
1415 *
1416 * Note that must and mlen got initialized during setup.
1417 */
1418static void
1419findmust(struct parse *p, struct re_guts *g)
1420{
1421 sop *scan;
Daniel Dunbar12420d32009-08-30 21:13:58 +00001422 sop *start = 0; /* start initialized in the default case, after that */
1423 sop *newstart = 0; /* newstart was initialized in the OCHAR case */
Torok Edwince0c81e2009-08-30 08:24:09 +00001424 sopno newlen;
1425 sop s;
1426 char *cp;
1427 sopno i;
1428
1429 /* avoid making error situations worse */
1430 if (p->error != 0)
1431 return;
1432
1433 /* find the longest OCHAR sequence in strip */
1434 newlen = 0;
1435 scan = g->strip + 1;
1436 do {
1437 s = *scan++;
1438 switch (OP(s)) {
1439 case OCHAR: /* sequence member */
1440 if (newlen == 0) /* new sequence */
1441 newstart = scan - 1;
1442 newlen++;
1443 break;
1444 case OPLUS_: /* things that don't break one */
1445 case OLPAREN:
1446 case ORPAREN:
1447 break;
1448 case OQUEST_: /* things that must be skipped */
1449 case OCH_:
1450 scan--;
1451 do {
1452 scan += OPND(s);
1453 s = *scan;
1454 /* assert() interferes w debug printouts */
1455 if (OP(s) != O_QUEST && OP(s) != O_CH &&
1456 OP(s) != OOR2) {
1457 g->iflags |= REGEX_BAD;
1458 return;
1459 }
1460 } while (OP(s) != O_QUEST && OP(s) != O_CH);
1461 /* fallthrough */
1462 default: /* things that break a sequence */
1463 if (newlen > g->mlen) { /* ends one */
1464 start = newstart;
1465 g->mlen = newlen;
1466 }
1467 newlen = 0;
1468 break;
1469 }
1470 } while (OP(s) != OEND);
1471
1472 if (g->mlen == 0) /* there isn't one */
1473 return;
1474
1475 /* turn it into a character string */
1476 g->must = malloc((size_t)g->mlen + 1);
1477 if (g->must == NULL) { /* argh; just forget it */
1478 g->mlen = 0;
1479 return;
1480 }
1481 cp = g->must;
1482 scan = start;
1483 for (i = g->mlen; i > 0; i--) {
1484 while (OP(s = *scan++) != OCHAR)
1485 continue;
1486 assert(cp < g->must + g->mlen);
1487 *cp++ = (char)OPND(s);
1488 }
1489 assert(cp == g->must + g->mlen);
1490 *cp++ = '\0'; /* just on general principles */
1491}
1492
1493/*
1494 - pluscount - count + nesting
1495 */
1496static sopno /* nesting depth */
1497pluscount(struct parse *p, struct re_guts *g)
1498{
1499 sop *scan;
1500 sop s;
1501 sopno plusnest = 0;
1502 sopno maxnest = 0;
1503
1504 if (p->error != 0)
1505 return(0); /* there may not be an OEND */
1506
1507 scan = g->strip + 1;
1508 do {
1509 s = *scan++;
1510 switch (OP(s)) {
1511 case OPLUS_:
1512 plusnest++;
1513 break;
1514 case O_PLUS:
1515 if (plusnest > maxnest)
1516 maxnest = plusnest;
1517 plusnest--;
1518 break;
1519 }
1520 } while (OP(s) != OEND);
1521 if (plusnest != 0)
1522 g->iflags |= REGEX_BAD;
1523 return(maxnest);
1524}