xref: /openbsd-src/lib/libc/regex/regcomp.c (revision 46035553bfdd96e63c94e32da0210227ec2e3cf1)
1 /*	$OpenBSD: regcomp.c,v 1.41 2020/12/31 17:24:05 millert Exp $ */
2 /*-
3  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
4  * Copyright (c) 1992, 1993, 1994
5  *	The Regents of the University of California.  All rights reserved.
6  *
7  * This code is derived from software contributed to Berkeley by
8  * Henry Spencer.
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  * 1. Redistributions of source code must retain the above copyright
14  *    notice, this list of conditions and the following disclaimer.
15  * 2. Redistributions in binary form must reproduce the above copyright
16  *    notice, this list of conditions and the following disclaimer in the
17  *    documentation and/or other materials provided with the distribution.
18  * 3. Neither the name of the University nor the names of its contributors
19  *    may be used to endorse or promote products derived from this software
20  *    without specific prior written permission.
21  *
22  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32  * SUCH DAMAGE.
33  *
34  *	@(#)regcomp.c	8.5 (Berkeley) 3/20/94
35  */
36 
37 #include <sys/types.h>
38 #include <stdio.h>
39 #include <string.h>
40 #include <ctype.h>
41 #include <limits.h>
42 #include <stdlib.h>
43 #include <regex.h>
44 
45 #include "utils.h"
46 #include "regex2.h"
47 
48 #include "cclass.h"
49 #include "cname.h"
50 
51 /*
52  * parse structure, passed up and down to avoid global variables and
53  * other clumsinesses
54  */
55 struct parse {
56 	const char *next;	/* next character in RE */
57 	const char *end;	/* end of string (-> NUL normally) */
58 	int error;		/* has an error been seen? */
59 	sop *strip;		/* malloced strip */
60 	sopno ssize;		/* malloced strip size (allocated) */
61 	sopno slen;		/* malloced strip length (used) */
62 	int ncsalloc;		/* number of csets allocated */
63 	struct re_guts *g;
64 #	define	NPAREN	10	/* we need to remember () 1-9 for back refs */
65 	sopno pbegin[NPAREN];	/* -> ( ([0] unused) */
66 	sopno pend[NPAREN];	/* -> ) ([0] unused) */
67 };
68 
69 static void p_ere(struct parse *, int);
70 static void p_ere_exp(struct parse *);
71 static void p_str(struct parse *);
72 static void p_bre(struct parse *, int, int);
73 static int p_simp_re(struct parse *, int);
74 static int p_count(struct parse *);
75 static void p_bracket(struct parse *);
76 static void p_b_term(struct parse *, cset *);
77 static void p_b_cclass(struct parse *, cset *);
78 static void p_b_eclass(struct parse *, cset *);
79 static char p_b_symbol(struct parse *);
80 static char p_b_coll_elem(struct parse *, int);
81 static char othercase(int);
82 static void bothcases(struct parse *, int);
83 static void ordinary(struct parse *, int);
84 static void backslash(struct parse *, int);
85 static void nonnewline(struct parse *);
86 static void repeat(struct parse *, sopno, int, int);
87 static void seterr(struct parse *, int);
88 static cset *allocset(struct parse *);
89 static void freeset(struct parse *, cset *);
90 static int freezeset(struct parse *, cset *);
91 static int firstch(struct parse *, cset *);
92 static int nch(struct parse *, cset *);
93 static int isinsets(struct re_guts *, int);
94 static int samesets(struct re_guts *, int, int);
95 static sopno dupl(struct parse *, sopno, sopno);
96 static void doemit(struct parse *, sop, size_t);
97 static void doinsert(struct parse *, sop, size_t, sopno);
98 static void dofwd(struct parse *, sopno, sop);
99 static int enlarge(struct parse *, sopno);
100 static void stripsnug(struct parse *, struct re_guts *);
101 static void findmust(struct parse *, struct re_guts *);
102 static sopno pluscount(struct parse *, struct re_guts *);
103 
104 static char nuls[10];		/* place to point scanner in event of error */
105 
106 /*
107  * macros for use with parse structure
108  * BEWARE:  these know that the parse structure is named `p' !!!
109  */
110 #define	PEEK()	(*p->next)
111 #define	PEEK2()	(*(p->next+1))
112 #define	MORE()	(p->end - p->next > 0)
113 #define	MORE2()	(p->end - p->next > 1)
114 #define	SEE(c)	(MORE() && PEEK() == (c))
115 #define	SEETWO(a, b)	(MORE2() && PEEK() == (a) && PEEK2() == (b))
116 #define	EAT(c)	((SEE(c)) ? (NEXT(), 1) : 0)
117 #define	EATTWO(a, b)	((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
118 #define	NEXT()	(p->next++)
119 #define	NEXT2()	(p->next += 2)
120 #define	NEXTn(n)	(p->next += (n))
121 #define	GETNEXT()	(*p->next++)
122 #define	SETERROR(e)	seterr(p, (e))
123 #define	REQUIRE(co, e)	do { if (!(co)) SETERROR(e); } while (0)
124 #define	EMIT(op, sopnd)	doemit(p, (sop)(op), (size_t)(sopnd))
125 #define	INSERT(op, pos)	doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
126 #define	AHEAD(pos)		dofwd(p, pos, HERE()-(pos))
127 #define	ASTERN(sop, pos)	EMIT(sop, HERE()-pos)
128 #define	HERE()		(p->slen)
129 #define	THERE()		(p->slen - 1)
130 #define	THERETHERE()	(p->slen - 2)
131 #define	DROP(n)	(p->slen -= (n))
132 
133 #ifndef NDEBUG
134 static int never = 0;		/* for use in asserts; shuts lint up */
135 #else
136 #define	never	0		/* some <assert.h>s have bugs too */
137 #endif
138 
139 /*
140  - regcomp - interface for parser and compilation
141  */
142 int				/* 0 success, otherwise REG_something */
143 regcomp(regex_t *preg, const char *pattern, int cflags)
144 {
145 	struct parse pa;
146 	struct re_guts *g;
147 	struct parse *p = &pa;
148 	int i;
149 	size_t len;
150 #ifdef REDEBUG
151 #	define	GOODFLAGS(f)	(f)
152 #else
153 #	define	GOODFLAGS(f)	((f)&~REG_DUMP)
154 #endif
155 
156 	cflags = GOODFLAGS(cflags);
157 	if ((cflags&REG_EXTENDED) && (cflags&REG_NOSPEC))
158 		return(REG_INVARG);
159 
160 	if (cflags&REG_PEND) {
161 		if (preg->re_endp < pattern)
162 			return(REG_INVARG);
163 		len = preg->re_endp - pattern;
164 	} else
165 		len = strlen((char *)pattern);
166 
167 	/* do the mallocs early so failure handling is easy */
168 	g = malloc(sizeof(struct re_guts));
169 	if (g == NULL)
170 		return(REG_ESPACE);
171 	p->ssize = len/(size_t)2*(size_t)3 + (size_t)1;	/* ugh */
172 	p->strip = reallocarray(NULL, p->ssize, sizeof(sop));
173 	p->slen = 0;
174 	if (p->strip == NULL) {
175 		free(g);
176 		return(REG_ESPACE);
177 	}
178 
179 	/* set things up */
180 	p->g = g;
181 	p->next = pattern;
182 	p->end = p->next + len;
183 	p->error = 0;
184 	p->ncsalloc = 0;
185 	for (i = 0; i < NPAREN; i++) {
186 		p->pbegin[i] = 0;
187 		p->pend[i] = 0;
188 	}
189 	g->csetsize = NC;
190 	g->sets = NULL;
191 	g->setbits = NULL;
192 	g->ncsets = 0;
193 	g->cflags = cflags;
194 	g->iflags = 0;
195 	g->nbol = 0;
196 	g->neol = 0;
197 	g->must = NULL;
198 	g->mlen = 0;
199 	g->nsub = 0;
200 	g->backrefs = 0;
201 
202 	/* do it */
203 	EMIT(OEND, 0);
204 	g->firststate = THERE();
205 	if (cflags&REG_EXTENDED)
206 		p_ere(p, OUT);
207 	else if (cflags&REG_NOSPEC)
208 		p_str(p);
209 	else
210 		p_bre(p, OUT, OUT);
211 	EMIT(OEND, 0);
212 	g->laststate = THERE();
213 
214 	/* tidy up loose ends and fill things in */
215 	stripsnug(p, g);
216 	findmust(p, g);
217 	g->nplus = pluscount(p, g);
218 	g->magic = MAGIC2;
219 	preg->re_nsub = g->nsub;
220 	preg->re_g = g;
221 	preg->re_magic = MAGIC1;
222 #ifndef REDEBUG
223 	/* not debugging, so can't rely on the assert() in regexec() */
224 	if (g->iflags&BAD)
225 		SETERROR(REG_ASSERT);
226 #endif
227 
228 	/* win or lose, we're done */
229 	if (p->error != 0)	/* lose */
230 		regfree(preg);
231 	return(p->error);
232 }
233 
234 /*
235  - p_ere - ERE parser top level, concatenation and alternation
236  */
237 static void
238 p_ere(struct parse *p, int stop)	/* character this ERE should end at */
239 {
240 	char c;
241 	sopno prevback;
242 	sopno prevfwd;
243 	sopno conc;
244 	int first = 1;		/* is this the first alternative? */
245 
246 	for (;;) {
247 		/* do a bunch of concatenated expressions */
248 		conc = HERE();
249 		while (MORE() && (c = PEEK()) != '|' && c != stop)
250 			p_ere_exp(p);
251 		REQUIRE(HERE() != conc, REG_EMPTY);	/* require nonempty */
252 
253 		if (!EAT('|'))
254 			break;		/* NOTE BREAK OUT */
255 
256 		if (first) {
257 			INSERT(OCH_, conc);	/* offset is wrong */
258 			prevfwd = conc;
259 			prevback = conc;
260 			first = 0;
261 		}
262 		ASTERN(OOR1, prevback);
263 		prevback = THERE();
264 		AHEAD(prevfwd);			/* fix previous offset */
265 		prevfwd = HERE();
266 		EMIT(OOR2, 0);			/* offset is very wrong */
267 	}
268 
269 	if (!first) {		/* tail-end fixups */
270 		AHEAD(prevfwd);
271 		ASTERN(O_CH, prevback);
272 	}
273 
274 	assert(!MORE() || SEE(stop));
275 }
276 
277 /*
278  - p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
279  */
280 static void
281 p_ere_exp(struct parse *p)
282 {
283 	char c;
284 	sopno pos;
285 	int count;
286 	int count2;
287 	sopno subno;
288 	int wascaret = 0;
289 
290 	assert(MORE());		/* caller should have ensured this */
291 	c = GETNEXT();
292 
293 	pos = HERE();
294 	switch (c) {
295 	case '(':
296 		REQUIRE(MORE(), REG_EPAREN);
297 		p->g->nsub++;
298 		subno = p->g->nsub;
299 		if (subno < NPAREN)
300 			p->pbegin[subno] = HERE();
301 		EMIT(OLPAREN, subno);
302 		if (!SEE(')'))
303 			p_ere(p, ')');
304 		if (subno < NPAREN) {
305 			p->pend[subno] = HERE();
306 			assert(p->pend[subno] != 0);
307 		}
308 		EMIT(ORPAREN, subno);
309 		REQUIRE(MORE() && GETNEXT() == ')', REG_EPAREN);
310 		break;
311 	case '^':
312 		EMIT(OBOL, 0);
313 		p->g->iflags |= USEBOL;
314 		p->g->nbol++;
315 		wascaret = 1;
316 		break;
317 	case '$':
318 		EMIT(OEOL, 0);
319 		p->g->iflags |= USEEOL;
320 		p->g->neol++;
321 		break;
322 	case '|':
323 		SETERROR(REG_EMPTY);
324 		break;
325 	case '*':
326 	case '+':
327 	case '?':
328 		SETERROR(REG_BADRPT);
329 		break;
330 	case '.':
331 		if (p->g->cflags&REG_NEWLINE)
332 			nonnewline(p);
333 		else
334 			EMIT(OANY, 0);
335 		break;
336 	case '[':
337 		p_bracket(p);
338 		break;
339 	case '\\':
340 		REQUIRE(MORE(), REG_EESCAPE);
341 		c = GETNEXT();
342 		backslash(p, c);
343 		break;
344 	case '{':		/* okay as ordinary except if digit follows */
345 		REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
346 		/* FALLTHROUGH */
347 	default:
348 		if (p->error != 0)
349 			return;
350 		ordinary(p, c);
351 		break;
352 	}
353 
354 	if (!MORE())
355 		return;
356 	c = PEEK();
357 	/* we call { a repetition if followed by a digit */
358 	if (!( c == '*' || c == '+' || c == '?' ||
359 				(c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
360 		return;		/* no repetition, we're done */
361 	NEXT();
362 
363 	REQUIRE(!wascaret, REG_BADRPT);
364 	switch (c) {
365 	case '*':	/* implemented as +? */
366 		/* this case does not require the (y|) trick, noKLUDGE */
367 		INSERT(OPLUS_, pos);
368 		ASTERN(O_PLUS, pos);
369 		INSERT(OQUEST_, pos);
370 		ASTERN(O_QUEST, pos);
371 		break;
372 	case '+':
373 		INSERT(OPLUS_, pos);
374 		ASTERN(O_PLUS, pos);
375 		break;
376 	case '?':
377 		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
378 		INSERT(OCH_, pos);		/* offset slightly wrong */
379 		ASTERN(OOR1, pos);		/* this one's right */
380 		AHEAD(pos);			/* fix the OCH_ */
381 		EMIT(OOR2, 0);			/* offset very wrong... */
382 		AHEAD(THERE());			/* ...so fix it */
383 		ASTERN(O_CH, THERETHERE());
384 		break;
385 	case '{':
386 		count = p_count(p);
387 		if (EAT(',')) {
388 			if (isdigit((uch)PEEK())) {
389 				count2 = p_count(p);
390 				REQUIRE(count <= count2, REG_BADBR);
391 			} else		/* single number with comma */
392 				count2 = INFINITY;
393 		} else		/* just a single number */
394 			count2 = count;
395 		repeat(p, pos, count, count2);
396 		if (!EAT('}')) {	/* error heuristics */
397 			while (MORE() && PEEK() != '}')
398 				NEXT();
399 			REQUIRE(MORE(), REG_EBRACE);
400 			SETERROR(REG_BADBR);
401 		}
402 		break;
403 	}
404 
405 	if (!MORE())
406 		return;
407 	c = PEEK();
408 	if (!( c == '*' || c == '+' || c == '?' ||
409 				(c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
410 		return;
411 	SETERROR(REG_BADRPT);
412 }
413 
414 /*
415  - p_str - string (no metacharacters) "parser"
416  */
417 static void
418 p_str(struct parse *p)
419 {
420 	REQUIRE(MORE(), REG_EMPTY);
421 	while (MORE())
422 		ordinary(p, GETNEXT());
423 }
424 
425 /*
426  - p_bre - BRE parser top level, anchoring and concatenation
427  * Giving end1 as OUT essentially eliminates the end1/end2 check.
428  *
429  * This implementation is a bit of a kludge, in that a trailing $ is first
430  * taken as an ordinary character and then revised to be an anchor.  The
431  * only undesirable side effect is that '$' gets included as a character
432  * category in such cases.  This is fairly harmless; not worth fixing.
433  * The amount of lookahead needed to avoid this kludge is excessive.
434  */
435 static void
436 p_bre(struct parse *p,
437     int end1,		/* first terminating character */
438     int end2)		/* second terminating character */
439 {
440 	sopno start = HERE();
441 	int first = 1;			/* first subexpression? */
442 	int wasdollar = 0;
443 
444 	if (EAT('^')) {
445 		EMIT(OBOL, 0);
446 		p->g->iflags |= USEBOL;
447 		p->g->nbol++;
448 	}
449 	while (MORE() && !SEETWO(end1, end2)) {
450 		wasdollar = p_simp_re(p, first);
451 		first = 0;
452 	}
453 	if (wasdollar) {	/* oops, that was a trailing anchor */
454 		DROP(1);
455 		EMIT(OEOL, 0);
456 		p->g->iflags |= USEEOL;
457 		p->g->neol++;
458 	}
459 
460 	REQUIRE(HERE() != start, REG_EMPTY);	/* require nonempty */
461 }
462 
463 /*
464  - p_simp_re - parse a simple RE, an atom possibly followed by a repetition
465  */
466 static int			/* was the simple RE an unbackslashed $? */
467 p_simp_re(struct parse *p,
468     int starordinary)		/* is a leading * an ordinary character? */
469 {
470 	int c;
471 	int count;
472 	int count2;
473 	sopno pos;
474 	int i;
475 	sopno subno;
476 #	define	BACKSL	(1<<CHAR_BIT)
477 
478 	pos = HERE();		/* repetion op, if any, covers from here */
479 
480 	assert(MORE());		/* caller should have ensured this */
481 	c = GETNEXT();
482 	if (c == '\\') {
483 		REQUIRE(MORE(), REG_EESCAPE);
484 		c = BACKSL | GETNEXT();
485 	}
486 	switch (c) {
487 	case '.':
488 		if (p->g->cflags&REG_NEWLINE)
489 			nonnewline(p);
490 		else
491 			EMIT(OANY, 0);
492 		break;
493 	case '[':
494 		p_bracket(p);
495 		break;
496 	case BACKSL|'<':
497 		EMIT(OBOW, 0);
498 		break;
499 	case BACKSL|'>':
500 		EMIT(OEOW, 0);
501 		break;
502 	case BACKSL|'{':
503 		SETERROR(REG_BADRPT);
504 		break;
505 	case BACKSL|'(':
506 		p->g->nsub++;
507 		subno = p->g->nsub;
508 		if (subno < NPAREN)
509 			p->pbegin[subno] = HERE();
510 		EMIT(OLPAREN, subno);
511 		/* the MORE here is an error heuristic */
512 		if (MORE() && !SEETWO('\\', ')'))
513 			p_bre(p, '\\', ')');
514 		if (subno < NPAREN) {
515 			p->pend[subno] = HERE();
516 			assert(p->pend[subno] != 0);
517 		}
518 		EMIT(ORPAREN, subno);
519 		REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
520 		break;
521 	case BACKSL|')':	/* should not get here -- must be user */
522 	case BACKSL|'}':
523 		SETERROR(REG_EPAREN);
524 		break;
525 	case BACKSL|'1':
526 	case BACKSL|'2':
527 	case BACKSL|'3':
528 	case BACKSL|'4':
529 	case BACKSL|'5':
530 	case BACKSL|'6':
531 	case BACKSL|'7':
532 	case BACKSL|'8':
533 	case BACKSL|'9':
534 		i = (c&~BACKSL) - '0';
535 		assert(i < NPAREN);
536 		if (p->pend[i] != 0) {
537 			assert(i <= p->g->nsub);
538 			EMIT(OBACK_, i);
539 			assert(p->pbegin[i] != 0);
540 			assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
541 			assert(OP(p->strip[p->pend[i]]) == ORPAREN);
542 			(void) dupl(p, p->pbegin[i]+1, p->pend[i]);
543 			EMIT(O_BACK, i);
544 		} else
545 			SETERROR(REG_ESUBREG);
546 		p->g->backrefs = 1;
547 		break;
548 	case '*':
549 		REQUIRE(starordinary, REG_BADRPT);
550 		/* FALLTHROUGH */
551 	default:
552 		if (p->error != 0)
553 			return(0);	/* Definitely not $... */
554 		ordinary(p, (char)c);
555 		break;
556 	}
557 
558 	if (EAT('*')) {		/* implemented as +? */
559 		/* this case does not require the (y|) trick, noKLUDGE */
560 		INSERT(OPLUS_, pos);
561 		ASTERN(O_PLUS, pos);
562 		INSERT(OQUEST_, pos);
563 		ASTERN(O_QUEST, pos);
564 	} else if (EATTWO('\\', '{')) {
565 		count = p_count(p);
566 		if (EAT(',')) {
567 			if (MORE() && isdigit((uch)PEEK())) {
568 				count2 = p_count(p);
569 				REQUIRE(count <= count2, REG_BADBR);
570 			} else		/* single number with comma */
571 				count2 = INFINITY;
572 		} else		/* just a single number */
573 			count2 = count;
574 		repeat(p, pos, count, count2);
575 		if (!EATTWO('\\', '}')) {	/* error heuristics */
576 			while (MORE() && !SEETWO('\\', '}'))
577 				NEXT();
578 			REQUIRE(MORE(), REG_EBRACE);
579 			SETERROR(REG_BADBR);
580 		}
581 	} else if (c == '$')	/* $ (but not \$) ends it */
582 		return(1);
583 
584 	return(0);
585 }
586 
587 /*
588  - p_count - parse a repetition count
589  */
590 static int			/* the value */
591 p_count(struct parse *p)
592 {
593 	int count = 0;
594 	int ndigits = 0;
595 
596 	while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
597 		count = count*10 + (GETNEXT() - '0');
598 		ndigits++;
599 	}
600 
601 	REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
602 	return(count);
603 }
604 
605 /*
606  - p_bracket - parse a bracketed character list
607  *
608  * Note a significant property of this code:  if the allocset() did SETERROR,
609  * no set operations are done.
610  */
611 static void
612 p_bracket(struct parse *p)
613 {
614 	cset *cs;
615 	int invert = 0;
616 
617 	/* Dept of Truly Sickening Special-Case Kludges */
618 	if (p->end - p->next > 5) {
619 		if (strncmp(p->next, "[:<:]]", 6) == 0) {
620 			EMIT(OBOW, 0);
621 			NEXTn(6);
622 			return;
623 		}
624 		if (strncmp(p->next, "[:>:]]", 6) == 0) {
625 			EMIT(OEOW, 0);
626 			NEXTn(6);
627 			return;
628 		}
629 	}
630 
631 	if ((cs = allocset(p)) == NULL) {
632 		/* allocset did set error status in p */
633 		return;
634 	}
635 
636 	if (EAT('^'))
637 		invert++;	/* make note to invert set at end */
638 	if (EAT(']'))
639 		CHadd(cs, ']');
640 	else if (EAT('-'))
641 		CHadd(cs, '-');
642 	while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
643 		p_b_term(p, cs);
644 	if (EAT('-'))
645 		CHadd(cs, '-');
646 	REQUIRE(MORE() && GETNEXT() == ']', REG_EBRACK);
647 
648 	if (p->error != 0) {	/* don't mess things up further */
649 		freeset(p, cs);
650 		return;
651 	}
652 
653 	if (p->g->cflags&REG_ICASE) {
654 		int i;
655 		int ci;
656 
657 		for (i = p->g->csetsize - 1; i >= 0; i--)
658 			if (CHIN(cs, i) && isalpha(i)) {
659 				ci = othercase(i);
660 				if (ci != i)
661 					CHadd(cs, ci);
662 			}
663 	}
664 	if (invert) {
665 		int i;
666 
667 		for (i = p->g->csetsize - 1; i >= 0; i--)
668 			if (CHIN(cs, i))
669 				CHsub(cs, i);
670 			else
671 				CHadd(cs, i);
672 		if (p->g->cflags&REG_NEWLINE)
673 			CHsub(cs, '\n');
674 	}
675 
676 	if (nch(p, cs) == 1) {		/* optimize singleton sets */
677 		ordinary(p, firstch(p, cs));
678 		freeset(p, cs);
679 	} else
680 		EMIT(OANYOF, freezeset(p, cs));
681 }
682 
683 /*
684  - p_b_term - parse one term of a bracketed character list
685  */
686 static void
687 p_b_term(struct parse *p, cset *cs)
688 {
689 	char c;
690 	char start, finish;
691 	int i;
692 
693 	/* classify what we've got */
694 	switch ((MORE()) ? PEEK() : '\0') {
695 	case '[':
696 		c = (MORE2()) ? PEEK2() : '\0';
697 		break;
698 	case '-':
699 		SETERROR(REG_ERANGE);
700 		return;			/* NOTE RETURN */
701 		break;
702 	default:
703 		c = '\0';
704 		break;
705 	}
706 
707 	switch (c) {
708 	case ':':		/* character class */
709 		NEXT2();
710 		REQUIRE(MORE(), REG_EBRACK);
711 		c = PEEK();
712 		REQUIRE(c != '-' && c != ']', REG_ECTYPE);
713 		p_b_cclass(p, cs);
714 		REQUIRE(MORE(), REG_EBRACK);
715 		REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
716 		break;
717 	case '=':		/* equivalence class */
718 		NEXT2();
719 		REQUIRE(MORE(), REG_EBRACK);
720 		c = PEEK();
721 		REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
722 		p_b_eclass(p, cs);
723 		REQUIRE(MORE(), REG_EBRACK);
724 		REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
725 		break;
726 	default:		/* symbol, ordinary character, or range */
727 /* xxx revision needed for multichar stuff */
728 		start = p_b_symbol(p);
729 		if (SEE('-') && MORE2() && PEEK2() != ']') {
730 			/* range */
731 			NEXT();
732 			if (EAT('-'))
733 				finish = '-';
734 			else
735 				finish = p_b_symbol(p);
736 		} else
737 			finish = start;
738 /* xxx what about signed chars here... */
739 		REQUIRE(start <= finish, REG_ERANGE);
740 		for (i = start; i <= finish; i++)
741 			CHadd(cs, i);
742 		break;
743 	}
744 }
745 
746 /*
747  - p_b_cclass - parse a character-class name and deal with it
748  */
749 static void
750 p_b_cclass(struct parse *p, cset *cs)
751 {
752 	const char *sp = p->next;
753 	const struct cclass *cp;
754 	size_t len;
755 	const char *u;
756 	char c;
757 
758 	while (MORE() && isalpha((uch)PEEK()))
759 		NEXT();
760 	len = p->next - sp;
761 	for (cp = cclasses; cp->name != NULL; cp++)
762 		if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
763 			break;
764 	if (cp->name == NULL) {
765 		/* oops, didn't find it */
766 		SETERROR(REG_ECTYPE);
767 		return;
768 	}
769 
770 	u = cp->chars;
771 	while ((c = *u++) != '\0')
772 		CHadd(cs, c);
773 }
774 
775 /*
776  - p_b_eclass - parse an equivalence-class name and deal with it
777  *
778  * This implementation is incomplete. xxx
779  */
780 static void
781 p_b_eclass(struct parse *p, cset *cs)
782 {
783 	char c;
784 
785 	c = p_b_coll_elem(p, '=');
786 	CHadd(cs, c);
787 }
788 
789 /*
790  - p_b_symbol - parse a character or [..]ed multicharacter collating symbol
791  */
792 static char			/* value of symbol */
793 p_b_symbol(struct parse *p)
794 {
795 	char value;
796 
797 	REQUIRE(MORE(), REG_EBRACK);
798 	if (!EATTWO('[', '.'))
799 		return(GETNEXT());
800 
801 	/* collating symbol */
802 	value = p_b_coll_elem(p, '.');
803 	REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
804 	return(value);
805 }
806 
807 /*
808  - p_b_coll_elem - parse a collating-element name and look it up
809  */
810 static char			/* value of collating element */
811 p_b_coll_elem(struct parse *p,
812     int endc)			/* name ended by endc,']' */
813 {
814 	const char *sp = p->next;
815 	const struct cname *cp;
816 	size_t len;
817 
818 	while (MORE() && !SEETWO(endc, ']'))
819 		NEXT();
820 	if (!MORE()) {
821 		SETERROR(REG_EBRACK);
822 		return(0);
823 	}
824 	len = p->next - sp;
825 	for (cp = cnames; cp->name != NULL; cp++)
826 		if (strncmp(cp->name, sp, len) == 0 && strlen(cp->name) == len)
827 			return(cp->code);	/* known name */
828 	if (len == 1)
829 		return(*sp);	/* single character */
830 	SETERROR(REG_ECOLLATE);			/* neither */
831 	return(0);
832 }
833 
834 /*
835  - othercase - return the case counterpart of an alphabetic
836  */
837 static char			/* if no counterpart, return ch */
838 othercase(int ch)
839 {
840 	ch = (uch)ch;
841 	assert(isalpha(ch));
842 	if (isupper(ch))
843 		return ((uch)tolower(ch));
844 	else if (islower(ch))
845 		return ((uch)toupper(ch));
846 	else			/* peculiar, but could happen */
847 		return(ch);
848 }
849 
850 /*
851  - bothcases - emit a dualcase version of a two-case character
852  *
853  * Boy, is this implementation ever a kludge...
854  */
855 static void
856 bothcases(struct parse *p, int ch)
857 {
858 	const char *oldnext = p->next;
859 	const char *oldend = p->end;
860 	char bracket[3];
861 
862 	ch = (uch)ch;
863 	assert(othercase(ch) != ch);	/* p_bracket() would recurse */
864 	p->next = bracket;
865 	p->end = bracket+2;
866 	bracket[0] = ch;
867 	bracket[1] = ']';
868 	bracket[2] = '\0';
869 	p_bracket(p);
870 	assert(p->next == bracket+2);
871 	p->next = oldnext;
872 	p->end = oldend;
873 }
874 
875 /*
876  - ordinary - emit an ordinary character
877  */
878 static void
879 ordinary(struct parse *p, int ch)
880 {
881 	if ((p->g->cflags&REG_ICASE) && isalpha((uch)ch) && othercase(ch) != ch)
882 		bothcases(p, ch);
883 	else
884 		EMIT(OCHAR, (uch)ch);
885 }
886 
887 /*
888  * do something magic with this character, but only if it's extra magic
889  */
890 static void
891 backslash(struct parse *p, int ch)
892 {
893 	switch (ch) {
894 	case '<':
895 		EMIT(OBOW, 0);
896 		break;
897 	case '>':
898 		EMIT(OEOW, 0);
899 		break;
900 	default:
901 		ordinary(p, ch);
902 		break;
903 	}
904 }
905 
906 /*
907  - nonnewline - emit REG_NEWLINE version of OANY
908  *
909  * Boy, is this implementation ever a kludge...
910  */
911 static void
912 nonnewline(struct parse *p)
913 {
914 	const char *oldnext = p->next;
915 	const char *oldend = p->end;
916 	static const char bracket[4] = { '^', '\n', ']', '\0' };
917 
918 	p->next = bracket;
919 	p->end = bracket+3;
920 	p_bracket(p);
921 	assert(p->next == bracket+3);
922 	p->next = oldnext;
923 	p->end = oldend;
924 }
925 
926 /*
927  - repeat - generate code for a bounded repetition, recursively if needed
928  */
929 static void
930 repeat(struct parse *p,
931     sopno start,		/* operand from here to end of strip */
932     int from,			/* repeated from this number */
933     int to)			/* to this number of times (maybe INFINITY) */
934 {
935 	sopno finish = HERE();
936 #	define	N	2
937 #	define	INF	3
938 #	define	REP(f, t)	((f)*8 + (t))
939 #	define	MAP(n)	(((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
940 	sopno copy;
941 
942 	if (p->error != 0)	/* head off possible runaway recursion */
943 		return;
944 
945 	assert(from <= to);
946 
947 	switch (REP(MAP(from), MAP(to))) {
948 	case REP(0, 0):			/* must be user doing this */
949 		DROP(finish-start);	/* drop the operand */
950 		break;
951 	case REP(0, 1):			/* as x{1,1}? */
952 	case REP(0, N):			/* as x{1,n}? */
953 	case REP(0, INF):		/* as x{1,}? */
954 		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
955 		INSERT(OCH_, start);		/* offset is wrong... */
956 		repeat(p, start+1, 1, to);
957 		ASTERN(OOR1, start);
958 		AHEAD(start);			/* ... fix it */
959 		EMIT(OOR2, 0);
960 		AHEAD(THERE());
961 		ASTERN(O_CH, THERETHERE());
962 		break;
963 	case REP(1, 1):			/* trivial case */
964 		/* done */
965 		break;
966 	case REP(1, N):			/* as x?x{1,n-1} */
967 		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
968 		INSERT(OCH_, start);
969 		ASTERN(OOR1, start);
970 		AHEAD(start);
971 		EMIT(OOR2, 0);			/* offset very wrong... */
972 		AHEAD(THERE());			/* ...so fix it */
973 		ASTERN(O_CH, THERETHERE());
974 		copy = dupl(p, start+1, finish+1);
975 		assert(copy == finish+4);
976 		repeat(p, copy, 1, to-1);
977 		break;
978 	case REP(1, INF):		/* as x+ */
979 		INSERT(OPLUS_, start);
980 		ASTERN(O_PLUS, start);
981 		break;
982 	case REP(N, N):			/* as xx{m-1,n-1} */
983 		copy = dupl(p, start, finish);
984 		repeat(p, copy, from-1, to-1);
985 		break;
986 	case REP(N, INF):		/* as xx{n-1,INF} */
987 		copy = dupl(p, start, finish);
988 		repeat(p, copy, from-1, to);
989 		break;
990 	default:			/* "can't happen" */
991 		SETERROR(REG_ASSERT);	/* just in case */
992 		break;
993 	}
994 }
995 
996 /*
997  - seterr - set an error condition
998  */
999 static void
1000 seterr(struct parse *p, int e)
1001 {
1002 	if (p->error == 0)	/* keep earliest error condition */
1003 		p->error = e;
1004 	p->next = nuls;		/* try to bring things to a halt */
1005 	p->end = nuls;
1006 }
1007 
1008 /*
1009  - allocset - allocate a set of characters for []
1010  */
1011 static cset *
1012 allocset(struct parse *p)
1013 {
1014 	int no = p->g->ncsets++;
1015 	size_t nc;
1016 	size_t nbytes;
1017 	cset *cs;
1018 	size_t css = (size_t)p->g->csetsize;
1019 	int i;
1020 
1021 	if (no >= p->ncsalloc) {	/* need another column of space */
1022 		void *ptr;
1023 
1024 		p->ncsalloc += CHAR_BIT;
1025 		nc = p->ncsalloc;
1026 		assert(nc % CHAR_BIT == 0);
1027 
1028 		ptr = reallocarray(p->g->sets, nc, sizeof(cset));
1029 		if (ptr == NULL)
1030 			goto nomem;
1031 		p->g->sets = ptr;
1032 
1033 		ptr = reallocarray(p->g->setbits, nc / CHAR_BIT, css);
1034 		if (ptr == NULL)
1035 			goto nomem;
1036 		nbytes = (nc / CHAR_BIT) * css;
1037 		p->g->setbits = ptr;
1038 
1039 		for (i = 0; i < no; i++)
1040 			p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
1041 
1042 		(void) memset((char *)p->g->setbits + (nbytes - css), 0, css);
1043 	}
1044 	/* XXX should not happen */
1045 	if (p->g->sets == NULL || p->g->setbits == NULL)
1046 		goto nomem;
1047 
1048 	cs = &p->g->sets[no];
1049 	cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
1050 	cs->mask = 1 << ((no) % CHAR_BIT);
1051 	cs->hash = 0;
1052 
1053 	return(cs);
1054 nomem:
1055 	free(p->g->sets);
1056 	p->g->sets = NULL;
1057 	free(p->g->setbits);
1058 	p->g->setbits = NULL;
1059 
1060 	SETERROR(REG_ESPACE);
1061 	/* caller's responsibility not to do set ops */
1062 	return(NULL);
1063 }
1064 
1065 /*
1066  - freeset - free a now-unused set
1067  */
1068 static void
1069 freeset(struct parse *p, cset *cs)
1070 {
1071 	int i;
1072 	cset *top = &p->g->sets[p->g->ncsets];
1073 	size_t css = (size_t)p->g->csetsize;
1074 
1075 	for (i = 0; i < css; i++)
1076 		CHsub(cs, i);
1077 	if (cs == top-1)	/* recover only the easy case */
1078 		p->g->ncsets--;
1079 }
1080 
1081 /*
1082  - freezeset - final processing on a set of characters
1083  *
1084  * The main task here is merging identical sets.  This is usually a waste
1085  * of time (although the hash code minimizes the overhead), but can win
1086  * big if REG_ICASE is being used.  REG_ICASE, by the way, is why the hash
1087  * is done using addition rather than xor -- all ASCII [aA] sets xor to
1088  * the same value!
1089  */
1090 static int			/* set number */
1091 freezeset(struct parse *p, cset *cs)
1092 {
1093 	uch h = cs->hash;
1094 	int i;
1095 	cset *top = &p->g->sets[p->g->ncsets];
1096 	cset *cs2;
1097 	size_t css = (size_t)p->g->csetsize;
1098 
1099 	/* look for an earlier one which is the same */
1100 	for (cs2 = &p->g->sets[0]; cs2 < top; cs2++)
1101 		if (cs2->hash == h && cs2 != cs) {
1102 			/* maybe */
1103 			for (i = 0; i < css; i++)
1104 				if (!!CHIN(cs2, i) != !!CHIN(cs, i))
1105 					break;		/* no */
1106 			if (i == css)
1107 				break;			/* yes */
1108 		}
1109 
1110 	if (cs2 < top) {	/* found one */
1111 		freeset(p, cs);
1112 		cs = cs2;
1113 	}
1114 
1115 	return((int)(cs - p->g->sets));
1116 }
1117 
1118 /*
1119  - firstch - return first character in a set (which must have at least one)
1120  */
1121 static int			/* character; there is no "none" value */
1122 firstch(struct parse *p, cset *cs)
1123 {
1124 	int i;
1125 	size_t css = (size_t)p->g->csetsize;
1126 
1127 	for (i = 0; i < css; i++)
1128 		if (CHIN(cs, i))
1129 			return((char)i);
1130 	assert(never);
1131 	return(0);		/* arbitrary */
1132 }
1133 
1134 /*
1135  - nch - number of characters in a set
1136  */
1137 static int
1138 nch(struct parse *p, cset *cs)
1139 {
1140 	int i;
1141 	size_t css = (size_t)p->g->csetsize;
1142 	int n = 0;
1143 
1144 	for (i = 0; i < css; i++)
1145 		if (CHIN(cs, i))
1146 			n++;
1147 	return(n);
1148 }
1149 
1150 /*
1151  - isinsets - is this character in any sets?
1152  */
1153 static int			/* predicate */
1154 isinsets(struct re_guts *g, int c)
1155 {
1156 	uch *col;
1157 	int i;
1158 	int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1159 	unsigned uc = (uch)c;
1160 
1161 	for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1162 		if (col[uc] != 0)
1163 			return(1);
1164 	return(0);
1165 }
1166 
1167 /*
1168  - samesets - are these two characters in exactly the same sets?
1169  */
1170 static int			/* predicate */
1171 samesets(struct re_guts *g, int c1, int c2)
1172 {
1173 	uch *col;
1174 	int i;
1175 	int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1176 	unsigned uc1 = (uch)c1;
1177 	unsigned uc2 = (uch)c2;
1178 
1179 	for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1180 		if (col[uc1] != col[uc2])
1181 			return(0);
1182 	return(1);
1183 }
1184 
1185 /*
1186  - dupl - emit a duplicate of a bunch of sops
1187  */
1188 static sopno			/* start of duplicate */
1189 dupl(struct parse *p,
1190     sopno start,		/* from here */
1191     sopno finish)		/* to this less one */
1192 {
1193 	sopno ret = HERE();
1194 	sopno len = finish - start;
1195 
1196 	assert(finish >= start);
1197 	if (len == 0)
1198 		return(ret);
1199 	if (!enlarge(p, p->ssize + len)) /* this many unexpected additions */
1200 		return(ret);
1201 	(void) memcpy(p->strip + p->slen, p->strip + start, len * sizeof(sop));
1202 	p->slen += len;
1203 	return(ret);
1204 }
1205 
1206 /*
1207  - doemit - emit a strip operator
1208  *
1209  * It might seem better to implement this as a macro with a function as
1210  * hard-case backup, but it's just too big and messy unless there are
1211  * some changes to the data structures.  Maybe later.
1212  */
1213 static void
1214 doemit(struct parse *p, sop op, size_t opnd)
1215 {
1216 	/* avoid making error situations worse */
1217 	if (p->error != 0)
1218 		return;
1219 
1220 	/* deal with oversize operands ("can't happen", more or less) */
1221 	assert(opnd < 1<<OPSHIFT);
1222 
1223 	/* deal with undersized strip */
1224 	if (p->slen >= p->ssize)
1225 		if (!enlarge(p, (p->ssize+1) / 2 * 3))	/* +50% */
1226 			return;
1227 
1228 	/* finally, it's all reduced to the easy case */
1229 	p->strip[p->slen++] = SOP(op, opnd);
1230 }
1231 
1232 /*
1233  - doinsert - insert a sop into the strip
1234  */
1235 static void
1236 doinsert(struct parse *p, sop op, size_t opnd, sopno pos)
1237 {
1238 	sopno sn;
1239 	sop s;
1240 	int i;
1241 
1242 	/* avoid making error situations worse */
1243 	if (p->error != 0)
1244 		return;
1245 
1246 	sn = HERE();
1247 	EMIT(op, opnd);		/* do checks, ensure space */
1248 	assert(HERE() == sn+1);
1249 	s = p->strip[sn];
1250 
1251 	/* adjust paren pointers */
1252 	assert(pos > 0);
1253 	for (i = 1; i < NPAREN; i++) {
1254 		if (p->pbegin[i] >= pos) {
1255 			p->pbegin[i]++;
1256 		}
1257 		if (p->pend[i] >= pos) {
1258 			p->pend[i]++;
1259 		}
1260 	}
1261 
1262 	memmove((char *)&p->strip[pos+1], (char *)&p->strip[pos],
1263 						(HERE()-pos-1)*sizeof(sop));
1264 	p->strip[pos] = s;
1265 }
1266 
1267 /*
1268  - dofwd - complete a forward reference
1269  */
1270 static void
1271 dofwd(struct parse *p, sopno pos, sop value)
1272 {
1273 	/* avoid making error situations worse */
1274 	if (p->error != 0)
1275 		return;
1276 
1277 	assert(value < 1<<OPSHIFT);
1278 	p->strip[pos] = OP(p->strip[pos]) | value;
1279 }
1280 
1281 /*
1282  - enlarge - enlarge the strip
1283  */
1284 static int
1285 enlarge(struct parse *p, sopno size)
1286 {
1287 	sop *sp;
1288 
1289 	if (p->ssize >= size)
1290 		return 1;
1291 
1292 	sp = reallocarray(p->strip, size, sizeof(sop));
1293 	if (sp == NULL) {
1294 		SETERROR(REG_ESPACE);
1295 		return 0;
1296 	}
1297 	p->strip = sp;
1298 	p->ssize = size;
1299 	return 1;
1300 }
1301 
1302 /*
1303  - stripsnug - compact the strip
1304  */
1305 static void
1306 stripsnug(struct parse *p, struct re_guts *g)
1307 {
1308 	g->nstates = p->slen;
1309 	g->strip = reallocarray(p->strip, p->slen, sizeof(sop));
1310 	if (g->strip == NULL) {
1311 		SETERROR(REG_ESPACE);
1312 		g->strip = p->strip;
1313 	}
1314 }
1315 
1316 /*
1317  - findmust - fill in must and mlen with longest mandatory literal string
1318  *
1319  * This algorithm could do fancy things like analyzing the operands of |
1320  * for common subsequences.  Someday.  This code is simple and finds most
1321  * of the interesting cases.
1322  *
1323  * Note that must and mlen got initialized during setup.
1324  */
1325 static void
1326 findmust(struct parse *p, struct re_guts *g)
1327 {
1328 	sop *scan;
1329 	sop *start;    /* start initialized in the default case, after that */
1330 	sop *newstart; /* newstart was initialized in the OCHAR case */
1331 	sopno newlen;
1332 	sop s;
1333 	char *cp;
1334 	sopno i;
1335 
1336 	/* avoid making error situations worse */
1337 	if (p->error != 0)
1338 		return;
1339 
1340 	/* find the longest OCHAR sequence in strip */
1341 	newlen = 0;
1342 	scan = g->strip + 1;
1343 	do {
1344 		s = *scan++;
1345 		switch (OP(s)) {
1346 		case OCHAR:		/* sequence member */
1347 			if (newlen == 0)		/* new sequence */
1348 				newstart = scan - 1;
1349 			newlen++;
1350 			break;
1351 		case OPLUS_:		/* things that don't break one */
1352 		case OLPAREN:
1353 		case ORPAREN:
1354 			break;
1355 		case OQUEST_:		/* things that must be skipped */
1356 		case OCH_:
1357 			scan--;
1358 			do {
1359 				scan += OPND(s);
1360 				s = *scan;
1361 				/* assert() interferes w debug printouts */
1362 				if (OP(s) != O_QUEST && OP(s) != O_CH &&
1363 							OP(s) != OOR2) {
1364 					g->iflags |= BAD;
1365 					return;
1366 				}
1367 			} while (OP(s) != O_QUEST && OP(s) != O_CH);
1368 			/* fallthrough */
1369 		default:		/* things that break a sequence */
1370 			if (newlen > g->mlen) {		/* ends one */
1371 				start = newstart;
1372 				g->mlen = newlen;
1373 			}
1374 			newlen = 0;
1375 			break;
1376 		}
1377 	} while (OP(s) != OEND);
1378 
1379 	if (g->mlen == 0)		/* there isn't one */
1380 		return;
1381 
1382 	/* turn it into a character string */
1383 	g->must = malloc((size_t)g->mlen + 1);
1384 	if (g->must == NULL) {		/* argh; just forget it */
1385 		g->mlen = 0;
1386 		return;
1387 	}
1388 	cp = g->must;
1389 	scan = start;
1390 	for (i = g->mlen; i > 0; i--) {
1391 		while (OP(s = *scan++) != OCHAR)
1392 			continue;
1393 		assert(cp < g->must + g->mlen);
1394 		*cp++ = (char)OPND(s);
1395 	}
1396 	assert(cp == g->must + g->mlen);
1397 	*cp++ = '\0';		/* just on general principles */
1398 }
1399 
1400 /*
1401  - pluscount - count + nesting
1402  */
1403 static sopno			/* nesting depth */
1404 pluscount(struct parse *p, struct re_guts *g)
1405 {
1406 	sop *scan;
1407 	sop s;
1408 	sopno plusnest = 0;
1409 	sopno maxnest = 0;
1410 
1411 	if (p->error != 0)
1412 		return(0);	/* there may not be an OEND */
1413 
1414 	scan = g->strip + 1;
1415 	do {
1416 		s = *scan++;
1417 		switch (OP(s)) {
1418 		case OPLUS_:
1419 			plusnest++;
1420 			break;
1421 		case O_PLUS:
1422 			if (plusnest > maxnest)
1423 				maxnest = plusnest;
1424 			plusnest--;
1425 			break;
1426 		}
1427 	} while (OP(s) != OEND);
1428 	if (plusnest != 0)
1429 		g->iflags |= BAD;
1430 	return(maxnest);
1431 }
1432