1/*	$NetBSD: engine.c,v 1.7 2011/11/19 17:45:11 tnozaki Exp $ */
2
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 of the University of Toronto.
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 *	@(#)engine.c	8.4 (Berkeley) 3/19/94
36 */
37
38/*
39 * The matching engine and friends.  This file is #included by regexec.c
40 * after suitable #defines of a variety of macros used herein, so that
41 * different state representations can be used without duplicating masses
42 * of code.
43 */
44
45#ifdef SNAMES
46#define	matcher	smatcher
47#define	fast	sfast
48#define	slow	sslow
49#define	dissect	sdissect
50#define	backref	sbackref
51#define	step	sstep
52#define	print	sprint
53#define	at	sat
54#define	match	smat
55#endif
56#ifdef LNAMES
57#define	matcher	lmatcher
58#define	fast	lfast
59#define	slow	lslow
60#define	dissect	ldissect
61#define	backref	lbackref
62#define	step	lstep
63#define	print	lprint
64#define	at	lat
65#define	match	lmat
66#endif
67
68/* another structure passed up and down to avoid zillions of parameters */
69struct match {
70	struct re_guts *g;
71	int eflags;
72	regmatch_t *pmatch;	/* [nsub+1] (0 element unused) */
73	const RCHAR_T *offp;		/* offsets work from here */
74	const RCHAR_T *beginp;		/* start of string -- virtual NUL precedes */
75	const RCHAR_T *endp;		/* end of string -- virtual NUL here */
76	const RCHAR_T *coldp;		/* can be no match starting before here */
77	const RCHAR_T **lastpos;	/* [nplus+1] */
78	STATEVARS;
79	states st;		/* current states */
80	states fresh;		/* states for a fresh start */
81	states tmp;		/* temporary */
82	states empty;		/* empty set of states */
83};
84
85/* ========= begin header generated by ./mkh ========= */
86#ifdef __cplusplus
87extern "C" {
88#endif
89
90/* === engine.c === */
91static int matcher(struct re_guts *g, const RCHAR_T *string, size_t nmatch, regmatch_t pmatch[], int eflags);
92static const RCHAR_T *dissect(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst, sopno stopst);
93static const RCHAR_T *backref(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst, sopno stopst, sopno lev);
94static const RCHAR_T *fast(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst, sopno stopst);
95static const RCHAR_T *slow(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst, sopno stopst);
96static states step(struct re_guts *g, sopno start, sopno stop, states bef, int flag, RCHAR_T ch, states aft);
97#define	BOL	(1)
98#define	EOL	(BOL+1)
99#define	BOLEOL	(BOL+2)
100#define	NOTHING	(BOL+3)
101#define	BOW	(BOL+4)
102#define	EOW	(BOL+5)
103#ifdef REDEBUG
104static void print(struct match *m, char *caption, states st, int ch, FILE *d);
105#endif
106#ifdef REDEBUG
107static void at(struct match *m, char *title, char *start, char *stop, sopno startst, sopno stopst);
108#endif
109#ifdef REDEBUG
110static char *pchar(int ch);
111#endif
112
113#ifdef __cplusplus
114}
115#endif
116/* ========= end header generated by ./mkh ========= */
117
118#ifdef REDEBUG
119#define	SP(t, s, c)	print(m, t, s, c, stdout)
120#define	AT(t, p1, p2, s1, s2)	at(m, t, p1, p2, s1, s2)
121#define	NOTE(str)	{ if (m->eflags&REG_TRACE) printf("=%s\n", (str)); }
122#else
123#define	SP(t, s, c)	/* nothing */
124#define	AT(t, p1, p2, s1, s2)	/* nothing */
125#define	NOTE(s)	/* nothing */
126#endif
127
128/*
129 - matcher - the actual matching engine
130 */
131static int			/* 0 success, REG_NOMATCH failure */
132matcher(struct re_guts *g, const RCHAR_T *string, size_t nmatch,
133    regmatch_t pmatch[], int eflags)
134{
135	const RCHAR_T *endp;
136	size_t i;
137	struct match mv;
138	struct match *m = &mv;
139	const RCHAR_T *dp;
140	const sopno gf = g->firststate+1;	/* +1 for OEND */
141	const sopno gl = g->laststate;
142	const RCHAR_T *start;
143	const RCHAR_T *stop;
144
145	/* simplify the situation where possible */
146	if (g->cflags&REG_NOSUB)
147		nmatch = 0;
148	if (eflags&REG_STARTEND) {
149		start = string + pmatch[0].rm_so;
150		stop = string + pmatch[0].rm_eo;
151	} else {
152		start = string;
153		stop = start + STRLEN(start);
154	}
155	if (stop < start)
156		return(REG_INVARG);
157
158	/* prescreening; this does wonders for this rather slow code */
159	if (g->must != NULL) {
160		for (dp = start; dp < stop; dp++)
161			if (*dp == g->must[0] && (size_t)(stop - dp) >= g->mlen &&
162				MEMCMP(dp, g->must, g->mlen) == 0)
163				break;
164		if (dp == stop)		/* we didn't find g->must */
165			return(REG_NOMATCH);
166	}
167
168	/* match struct setup */
169	m->g = g;
170	m->eflags = eflags;
171	m->pmatch = NULL;
172	m->lastpos = NULL;
173	m->offp = string;
174	m->beginp = start;
175	m->endp = stop;
176	STATESETUP(m, 4);
177	SETUP(m->st);
178	SETUP(m->fresh);
179	SETUP(m->tmp);
180	SETUP(m->empty);
181	CLEAR(m->empty);
182
183	/* this loop does only one repetition except for backrefs */
184	for (;;) {
185		endp = fast(m, start, stop, gf, gl);
186		if (endp == NULL) {		/* a miss */
187			STATETEARDOWN(m);
188			return(REG_NOMATCH);
189		}
190		if (nmatch == 0 && !g->backrefs)
191			break;		/* no further info needed */
192
193		/* where? */
194		assert(m->coldp != NULL);
195		for (;;) {
196			NOTE("finding start");
197			endp = slow(m, m->coldp, stop, gf, gl);
198			if (endp != NULL)
199				break;
200			assert(m->coldp < m->endp);
201			m->coldp++;
202		}
203		if (nmatch == 1 && !g->backrefs)
204			break;		/* no further info needed */
205
206		/* oh my, he wants the subexpressions... */
207		if (m->pmatch == NULL)
208			m->pmatch = (regmatch_t *)malloc((m->g->nsub + 1) *
209							sizeof(regmatch_t));
210		if (m->pmatch == NULL) {
211			STATETEARDOWN(m);
212			return(REG_ESPACE);
213		}
214		for (i = 1; i <= m->g->nsub; i++)
215			m->pmatch[i].rm_so = m->pmatch[i].rm_eo = -1;
216		if (!g->backrefs && !(m->eflags&REG_BACKR)) {
217			NOTE("dissecting");
218			dp = dissect(m, m->coldp, endp, gf, gl);
219		} else {
220			if (g->nplus > 0 && m->lastpos == NULL)
221				m->lastpos = (const RCHAR_T **)malloc((g->nplus+1) *
222							sizeof(const RCHAR_T *));
223			if (g->nplus > 0 && m->lastpos == NULL) {
224				free(m->pmatch);
225				STATETEARDOWN(m);
226				return(REG_ESPACE);
227			}
228			NOTE("backref dissect");
229			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
230		}
231		if (dp != NULL)
232			break;
233
234		/* uh-oh... we couldn't find a subexpression-level match */
235		assert(g->backrefs);	/* must be back references doing it */
236		assert(g->nplus == 0 || m->lastpos != NULL);
237		for (;;) {
238			if (dp != NULL || endp <= m->coldp)
239				break;		/* defeat */
240			NOTE("backoff");
241			endp = slow(m, m->coldp, endp-1, gf, gl);
242			if (endp == NULL)
243				break;		/* defeat */
244			/* try it on a shorter possibility */
245#ifndef NDEBUG
246			for (i = 1; i <= m->g->nsub; i++) {
247				assert(m->pmatch[i].rm_so == -1);
248				assert(m->pmatch[i].rm_eo == -1);
249			}
250#endif
251			NOTE("backoff dissect");
252			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
253		}
254		assert(dp == NULL || dp == endp);
255		if (dp != NULL)		/* found a shorter one */
256			break;
257
258		/* despite initial appearances, there is no match here */
259		NOTE("false alarm");
260		start = m->coldp + 1;	/* recycle starting later */
261		assert(start <= stop);
262	}
263
264	/* fill in the details if requested */
265	if (nmatch > 0) {
266		pmatch[0].rm_so = m->coldp - m->offp;
267		pmatch[0].rm_eo = endp - m->offp;
268	}
269	if (nmatch > 1) {
270		assert(m->pmatch != NULL);
271		for (i = 1; i < nmatch; i++)
272			if (i <= m->g->nsub)
273				pmatch[i] = m->pmatch[i];
274			else {
275				pmatch[i].rm_so = -1;
276				pmatch[i].rm_eo = -1;
277			}
278	}
279
280	if (m->pmatch != NULL)
281		free((char *)m->pmatch);
282	if (m->lastpos != NULL)
283		free((char *)m->lastpos);
284	STATETEARDOWN(m);
285	return(0);
286}
287
288/*
289 - dissect - figure out what matched what, no back references
290 */
291static const RCHAR_T *			/* == stop (success) always */
292dissect(struct match *m, const RCHAR_T *start, const RCHAR_T *stop,
293    sopno startst, sopno stopst)
294{
295	int i;
296	sopno ss;	/* start sop of current subRE */
297	sopno es;	/* end sop of current subRE */
298	const RCHAR_T *sp;	/* start of string matched by it */
299	const RCHAR_T *stp;	/* string matched by it cannot pass here */
300	const RCHAR_T *rest;	/* start of rest of string */
301	const RCHAR_T *tail;	/* string unmatched by rest of RE */
302	sopno ssub;	/* start sop of subsubRE */
303	sopno esub;	/* end sop of subsubRE */
304	const RCHAR_T *ssp;	/* start of string matched by subsubRE */
305	const RCHAR_T *sep;	/* end of string matched by subsubRE */
306	const RCHAR_T *oldssp;	/* previous ssp */
307	const RCHAR_T *dp;
308
309	AT("diss", start, stop, startst, stopst);
310	sp = start;
311	for (ss = startst; ss < stopst; ss = es) {
312		/* identify end of subRE */
313		es = ss;
314		switch (m->g->strip[es]) {
315		case OPLUS_:
316		case OQUEST_:
317			es += m->g->stripdata[es];
318			break;
319		case OCH_:
320			while (m->g->strip[es] != O_CH)
321				es += m->g->stripdata[es];
322			break;
323		}
324		es++;
325
326		/* figure out what it matched */
327		switch (m->g->strip[ss]) {
328		case OEND:
329			assert(nope);
330			break;
331		case OCHAR:
332			sp++;
333			break;
334		case OBOL:
335		case OEOL:
336		case OBOW:
337		case OEOW:
338			break;
339		case OANY:
340		case OANYOF:
341			sp++;
342			break;
343		case OBACK_:
344		case O_BACK:
345			assert(nope);
346			break;
347		/* cases where length of match is hard to find */
348		case OQUEST_:
349			stp = stop;
350			for (;;) {
351				/* how long could this one be? */
352				rest = slow(m, sp, stp, ss, es);
353				assert(rest != NULL);	/* it did match */
354				/* could the rest match the rest? */
355				tail = slow(m, rest, stop, es, stopst);
356				if (tail == stop)
357					break;		/* yes! */
358				/* no -- try a shorter match for this one */
359				stp = rest - 1;
360				assert(stp >= sp);	/* it did work */
361			}
362			ssub = ss + 1;
363			esub = es - 1;
364			/* did innards match? */
365			if (slow(m, sp, rest, ssub, esub) != NULL) {
366				dp = dissect(m, sp, rest, ssub, esub);
367				assert(dp == rest);
368			} else		/* no */
369				assert(sp == rest);
370			sp = rest;
371			break;
372		case OPLUS_:
373			stp = stop;
374			for (;;) {
375				/* how long could this one be? */
376				rest = slow(m, sp, stp, ss, es);
377				assert(rest != NULL);	/* it did match */
378				/* could the rest match the rest? */
379				tail = slow(m, rest, stop, es, stopst);
380				if (tail == stop)
381					break;		/* yes! */
382				/* no -- try a shorter match for this one */
383				stp = rest - 1;
384				assert(stp >= sp);	/* it did work */
385			}
386			ssub = ss + 1;
387			esub = es - 1;
388			ssp = sp;
389			oldssp = ssp;
390			for (;;) {	/* find last match of innards */
391				sep = slow(m, ssp, rest, ssub, esub);
392				if (sep == NULL || sep == ssp)
393					break;	/* failed or matched null */
394				oldssp = ssp;	/* on to next try */
395				ssp = sep;
396			}
397			if (sep == NULL) {
398				/* last successful match */
399				sep = ssp;
400				ssp = oldssp;
401			}
402			assert(sep == rest);	/* must exhaust substring */
403			assert(slow(m, ssp, sep, ssub, esub) == rest);
404			dp = dissect(m, ssp, sep, ssub, esub);
405			assert(dp == sep);
406			sp = rest;
407			break;
408		case OCH_:
409			stp = stop;
410			for (;;) {
411				/* how long could this one be? */
412				rest = slow(m, sp, stp, ss, es);
413				assert(rest != NULL);	/* it did match */
414				/* could the rest match the rest? */
415				tail = slow(m, rest, stop, es, stopst);
416				if (tail == stop)
417					break;		/* yes! */
418				/* no -- try a shorter match for this one */
419				stp = rest - 1;
420				assert(stp >= sp);	/* it did work */
421			}
422			ssub = ss + 1;
423			esub = ss + m->g->stripdata[ss] - 1;
424			assert(m->g->strip[esub] == OOR1);
425			for (;;) {	/* find first matching branch */
426				if (slow(m, sp, rest, ssub, esub) == rest)
427					break;	/* it matched all of it */
428				/* that one missed, try next one */
429				assert(m->g->strip[esub] == OOR1);
430				esub++;
431				assert(m->g->strip[esub] == OOR2);
432				ssub = esub + 1;
433				esub += m->g->stripdata[esub];
434				if (m->g->strip[esub] == OOR2)
435					esub--;
436				else
437					assert(m->g->strip[esub] == O_CH);
438			}
439			dp = dissect(m, sp, rest, ssub, esub);
440			assert(dp == rest);
441			sp = rest;
442			break;
443		case O_PLUS:
444		case O_QUEST:
445		case OOR1:
446		case OOR2:
447		case O_CH:
448			assert(nope);
449			break;
450		case OLPAREN:
451			i = m->g->stripdata[ss];
452			assert(0 < i && i <= m->g->nsub);
453			m->pmatch[i].rm_so = sp - m->offp;
454			break;
455		case ORPAREN:
456			i = m->g->stripdata[ss];
457			assert(0 < i && i <= m->g->nsub);
458			m->pmatch[i].rm_eo = sp - m->offp;
459			break;
460		default:		/* uh oh */
461			assert(nope);
462			break;
463		}
464	}
465
466	assert(sp == stop);
467	return(sp);
468}
469
470/*
471 - backref - figure out what matched what, figuring in back references
472 */
473static const RCHAR_T *			/* == stop (success) or NULL (failure) */
474backref(struct match *m, const RCHAR_T *start, const RCHAR_T *stop,
475    sopno startst, sopno stopst, sopno lev) /* PLUS nesting level */
476{
477	int i;
478	sopno ss;	/* start sop of current subRE */
479	const RCHAR_T *sp;	/* start of string matched by it */
480	sopno ssub;	/* start sop of subsubRE */
481	sopno esub;	/* end sop of subsubRE */
482	const RCHAR_T *ssp;	/* start of string matched by subsubRE */
483	const RCHAR_T *dp;
484	size_t len;
485	int hard;
486	sop s;
487	RCHAR_T d;
488	regoff_t offsave;
489	cset *cs;
490
491	AT("back", start, stop, startst, stopst);
492	sp = start;
493
494	/* get as far as we can with easy stuff */
495	hard = 0;
496	for (ss = startst; !hard && ss < stopst; ss++) {
497		s = m->g->strip[ss];
498		d = m->g->stripdata[ss];
499		switch (s) {
500		case OCHAR:
501			if (sp == stop || *sp++ != d)
502				return(NULL);
503			break;
504		case OANY:
505			if (sp == stop)
506				return(NULL);
507			sp++;
508			break;
509		case OANYOF:
510			cs = &m->g->sets[d];
511			if (sp == stop || !CHIN(cs, *sp++))
512				return(NULL);
513			break;
514		case OBOL:
515			if ( (sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||
516					(sp < m->endp && *(sp-1) == '\n' &&
517						(m->g->cflags&REG_NEWLINE)) )
518				{ /* yes */ }
519			else
520				return(NULL);
521			break;
522		case OEOL:
523			if ( (sp == m->endp && !(m->eflags&REG_NOTEOL)) ||
524					(sp < m->endp && *sp == '\n' &&
525						(m->g->cflags&REG_NEWLINE)) )
526				{ /* yes */ }
527			else
528				return(NULL);
529			break;
530		case OBOW:
531			if (( (sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||
532					(sp < m->endp && *(sp-1) == '\n' &&
533						(m->g->cflags&REG_NEWLINE)) ||
534					(sp > m->beginp &&
535							!ISWORD(*(sp-1))) ) &&
536					(sp < m->endp && ISWORD(*sp)) )
537				{ /* yes */ }
538			else
539				return(NULL);
540			break;
541		case OEOW:
542			if (( (sp == m->endp && !(m->eflags&REG_NOTEOL)) ||
543					(sp < m->endp && *sp == '\n' &&
544						(m->g->cflags&REG_NEWLINE)) ||
545					(sp < m->endp && !ISWORD(*sp)) ) &&
546					(sp > m->beginp && ISWORD(*(sp-1))) )
547				{ /* yes */ }
548			else
549				return(NULL);
550			break;
551		case O_QUEST:
552			break;
553		case OOR1:	/* matches null but needs to skip */
554			ss++;
555			s = m->g->strip[ss];
556			d = m->g->stripdata[ss];
557			do {
558				assert(s == OOR2);
559				ss += d;
560				s = m->g->strip[ss];
561				d = m->g->stripdata[ss];
562			} while (s != O_CH);
563			/* note that the ss++ gets us past the O_CH */
564			break;
565		default:	/* have to make a choice */
566			hard = 1;
567			break;
568		}
569	}
570	if (!hard) {		/* that was it! */
571		if (sp != stop)
572			return(NULL);
573		return(sp);
574	}
575	ss--;			/* adjust for the for's final increment */
576
577	/* the hard stuff */
578	AT("hard", sp, stop, ss, stopst);
579	s = m->g->strip[ss];
580	d = m->g->stripdata[ss];
581	switch (s) {
582	case OBACK_:		/* the vilest depths */
583		i = d;
584		assert(0 < i && i <= m->g->nsub);
585		if (m->pmatch[i].rm_eo == -1)
586			return(NULL);
587		assert(m->pmatch[i].rm_so != -1);
588		len = m->pmatch[i].rm_eo - m->pmatch[i].rm_so;
589		assert(stop - m->beginp >= len);
590		if (sp > stop - len)
591			return(NULL);	/* not enough left to match */
592		ssp = m->offp + m->pmatch[i].rm_so;
593		if (memcmp(sp, ssp, len) != 0)
594			return(NULL);
595		while (m->g->strip[ss] != O_BACK || m->g->stripdata[ss] != i)
596			ss++;
597		return(backref(m, sp+len, stop, ss+1, stopst, lev));
598		break;
599	case OQUEST_:		/* to null or not */
600		dp = backref(m, sp, stop, ss+1, stopst, lev);
601		if (dp != NULL)
602			return(dp);	/* not */
603		return(backref(m, sp, stop, ss+d+1, stopst, lev));
604		break;
605	case OPLUS_:
606		assert(m->lastpos != NULL);
607		assert(lev+1 <= m->g->nplus);
608		m->lastpos[lev+1] = sp;
609		return(backref(m, sp, stop, ss+1, stopst, lev+1));
610		break;
611	case O_PLUS:
612		if (sp == m->lastpos[lev])	/* last pass matched null */
613			return(backref(m, sp, stop, ss+1, stopst, lev-1));
614		/* try another pass */
615		m->lastpos[lev] = sp;
616		dp = backref(m, sp, stop, ss-d+1, stopst, lev);
617		if (dp == NULL)
618			return(backref(m, sp, stop, ss+1, stopst, lev-1));
619		else
620			return(dp);
621		break;
622	case OCH_:		/* find the right one, if any */
623		ssub = ss + 1;
624		esub = ss + d - 1;
625		assert(m->g->strip[esub] == OOR1);
626		for (;;) {	/* find first matching branch */
627			dp = backref(m, sp, stop, ssub, esub, lev);
628			if (dp != NULL)
629				return(dp);
630			/* that one missed, try next one */
631			if (m->g->strip[esub] == O_CH)
632				return(NULL);	/* there is none */
633			esub++;
634			assert(m->g->strip[esub] == OOR2);
635			ssub = esub + 1;
636			esub += m->g->stripdata[esub];
637			if (m->g->strip[esub] == OOR2)
638				esub--;
639			else
640				assert(m->g->strip[esub] == O_CH);
641		}
642		break;
643	case OLPAREN:		/* must undo assignment if rest fails */
644		i = d;
645		assert(0 < i && i <= m->g->nsub);
646		offsave = m->pmatch[i].rm_so;
647		m->pmatch[i].rm_so = sp - m->offp;
648		dp = backref(m, sp, stop, ss+1, stopst, lev);
649		if (dp != NULL)
650			return(dp);
651		m->pmatch[i].rm_so = offsave;
652		return(NULL);
653		break;
654	case ORPAREN:		/* must undo assignment if rest fails */
655		i = d;
656		assert(0 < i && i <= m->g->nsub);
657		offsave = m->pmatch[i].rm_eo;
658		m->pmatch[i].rm_eo = sp - m->offp;
659		dp = backref(m, sp, stop, ss+1, stopst, lev);
660		if (dp != NULL)
661			return(dp);
662		m->pmatch[i].rm_eo = offsave;
663		return(NULL);
664		break;
665	default:		/* uh oh */
666		assert(nope);
667		break;
668	}
669
670	/* "can't happen" */
671	assert(nope);
672	/* NOTREACHED */
673	return NULL;
674}
675
676/*
677 - fast - step through the string at top speed
678 */
679static const RCHAR_T *			/* where tentative match ended, or NULL */
680fast(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst,
681    sopno stopst)
682{
683	states st = m->st;
684	states fresh = m->fresh;
685	states tmp = m->tmp;
686	const RCHAR_T *p = start;
687	RCHAR_T c = (start == m->beginp) ? OUT : *(start-1);
688	RCHAR_T lastc;	/* previous c */
689	int flag;
690	int i;
691	const RCHAR_T *coldp;	/* last p after which no match was underway */
692
693	CLEAR(st);
694	SET1(st, startst);
695	st = step(m->g, startst, stopst, st, NOTHING, OUT, st);
696	ASSIGN(fresh, st);
697	SP("start", st, *p);
698	coldp = NULL;
699	for (;;) {
700		/* next character */
701		lastc = c;
702		c = (p == m->endp) ? OUT : *p;
703		if (EQ(st, fresh))
704			coldp = p;
705
706		/* is there an EOL and/or BOL between lastc and c? */
707		flag = 0;
708		i = 0;
709		if ( (lastc == '\n' && m->g->cflags&REG_NEWLINE) ||
710				(lastc == OUT && !(m->eflags&REG_NOTBOL)) ) {
711			flag = BOL;
712			i = m->g->nbol;
713		}
714		if ( (c == '\n' && m->g->cflags&REG_NEWLINE) ||
715				(c == OUT && !(m->eflags&REG_NOTEOL)) ) {
716			flag = (flag == BOL) ? BOLEOL : EOL;
717			i += m->g->neol;
718		}
719		if (i != 0) {
720			for (; i > 0; i--)
721				st = step(m->g, startst, stopst, st, flag, OUT, st);
722			SP("boleol", st, c);
723		}
724
725		/* how about a word boundary? */
726		if ( (flag == BOL || (lastc != OUT && !ISWORD(lastc))) &&
727					(c != OUT && ISWORD(c)) ) {
728			flag = BOW;
729		}
730		if ( (lastc != OUT && ISWORD(lastc)) &&
731				(flag == EOL || (c != OUT && !ISWORD(c))) ) {
732			flag = EOW;
733		}
734		if (flag == BOW || flag == EOW) {
735			st = step(m->g, startst, stopst, st, flag, OUT, st);
736			SP("boweow", st, c);
737		}
738
739		/* are we done? */
740		if (ISSET(st, stopst) || p == stop)
741			break;		/* NOTE BREAK OUT */
742
743		/* no, we must deal with this character */
744		ASSIGN(tmp, st);
745		ASSIGN(st, fresh);
746		assert(c != OUT);
747		st = step(m->g, startst, stopst, tmp, 0, c, st);
748		SP("aft", st, c);
749		assert(EQ(step(m->g, startst, stopst, st, NOTHING, OUT, st), st));
750		p++;
751	}
752
753	assert(coldp != NULL);
754	m->coldp = coldp;
755	if (ISSET(st, stopst))
756		return(p+1);
757	else
758		return(NULL);
759}
760
761/*
762 - slow - step through the string more deliberately
763 */
764static const RCHAR_T *			/* where it ended */
765slow(struct match *m, const RCHAR_T *start, const RCHAR_T *stop, sopno startst,
766    sopno stopst)
767{
768	states st = m->st;
769	states empty = m->empty;
770	states tmp = m->tmp;
771	const RCHAR_T *p = start;
772	RCHAR_T c = (start == m->beginp) ? OUT : *(start-1);
773	RCHAR_T lastc;	/* previous c */
774	int flag;
775	int i;
776	const RCHAR_T *matchp;	/* last p at which a match ended */
777
778	AT("slow", start, stop, startst, stopst);
779	CLEAR(st);
780	SET1(st, startst);
781	SP("sstart", st, *p);
782	st = step(m->g, startst, stopst, st, NOTHING, OUT, st);
783	matchp = NULL;
784	for (;;) {
785		/* next character */
786		lastc = c;
787		c = (p == m->endp) ? OUT : *p;
788
789		/* is there an EOL and/or BOL between lastc and c? */
790		flag = 0;
791		i = 0;
792		if ( (lastc == '\n' && m->g->cflags&REG_NEWLINE) ||
793				(lastc == OUT && !(m->eflags&REG_NOTBOL)) ) {
794			flag = BOL;
795			i = m->g->nbol;
796		}
797		if ( (c == '\n' && m->g->cflags&REG_NEWLINE) ||
798				(c == OUT && !(m->eflags&REG_NOTEOL)) ) {
799			flag = (flag == BOL) ? BOLEOL : EOL;
800			i += m->g->neol;
801		}
802		if (i != 0) {
803			for (; i > 0; i--)
804				st = step(m->g, startst, stopst, st, flag, OUT, st);
805			SP("sboleol", st, c);
806		}
807
808		/* how about a word boundary? */
809		if ( (flag == BOL || (lastc != OUT && !ISWORD(lastc))) &&
810					(c != OUT && ISWORD(c)) ) {
811			flag = BOW;
812		}
813		if ( (lastc != OUT && ISWORD(lastc)) &&
814				(flag == EOL || (c != OUT && !ISWORD(c))) ) {
815			flag = EOW;
816		}
817		if (flag == BOW || flag == EOW) {
818			st = step(m->g, startst, stopst, st, flag, OUT, st);
819			SP("sboweow", st, c);
820		}
821
822		/* are we done? */
823		if (ISSET(st, stopst))
824			matchp = p;
825		if (EQ(st, empty) || p == stop)
826			break;		/* NOTE BREAK OUT */
827
828		/* no, we must deal with this character */
829		ASSIGN(tmp, st);
830		ASSIGN(st, empty);
831		assert(c != OUT);
832		st = step(m->g, startst, stopst, tmp, 0, c, st);
833		SP("saft", st, c);
834		assert(EQ(step(m->g, startst, stopst, st, NOTHING, OUT, st), st));
835		p++;
836	}
837
838	return(matchp);
839}
840
841
842/*
843 - step - map set of states reachable before char to set reachable after
844 */
845static states
846step(struct re_guts *g,
847    sopno start,			/* start state within strip */
848    sopno stop,			/* state after stop state within strip */
849    states bef,		/* states reachable before */
850    int flag,			/* NONCHAR flag */
851    RCHAR_T ch,			/* character code */
852    states aft)		/* states already known reachable after */
853{
854	cset *cs;
855	sop s;
856	RCHAR_T d;
857	sopno pc;
858	onestate here;		/* note, macros know this name */
859	sopno look;
860	int i;
861
862	for (pc = start, INIT(here, pc); pc != stop; pc++, INC(here)) {
863		s = g->strip[pc];
864		d = g->stripdata[pc];
865		switch (s) {
866		case OEND:
867			assert(pc == stop-1);
868			break;
869		case OCHAR:
870			/* only characters can match */
871			assert(!flag || ch != d);
872			if (ch == d)
873				FWD(aft, bef, 1);
874			break;
875		case OBOL:
876			if (flag == BOL || flag == BOLEOL)
877				FWD(aft, bef, 1);
878			break;
879		case OEOL:
880			if (flag == EOL || flag == BOLEOL)
881				FWD(aft, bef, 1);
882			break;
883		case OBOW:
884			if (flag == BOW)
885				FWD(aft, bef, 1);
886			break;
887		case OEOW:
888			if (flag == EOW)
889				FWD(aft, bef, 1);
890			break;
891		case OANY:
892			if (!flag)
893				FWD(aft, bef, 1);
894			break;
895		case OANYOF:
896			cs = &g->sets[d];
897			if (!flag && CHIN(cs, ch))
898				FWD(aft, bef, 1);
899			break;
900		case OBACK_:		/* ignored here */
901		case O_BACK:
902			FWD(aft, aft, 1);
903			break;
904		case OPLUS_:		/* forward, this is just an empty */
905			FWD(aft, aft, 1);
906			break;
907		case O_PLUS:		/* both forward and back */
908			FWD(aft, aft, 1);
909			i = ISSETBACK(aft, d);
910			BACK(aft, aft, d);
911			if (!i && ISSETBACK(aft, d)) {
912				/* oho, must reconsider loop body */
913				pc -= d + 1;
914				INIT(here, pc);
915			}
916			break;
917		case OQUEST_:		/* two branches, both forward */
918			FWD(aft, aft, 1);
919			FWD(aft, aft, d);
920			break;
921		case O_QUEST:		/* just an empty */
922			FWD(aft, aft, 1);
923			break;
924		case OLPAREN:		/* not significant here */
925		case ORPAREN:
926			FWD(aft, aft, 1);
927			break;
928		case OCH_:		/* mark the first two branches */
929			FWD(aft, aft, 1);
930			assert(OP(g->strip[pc+d]) == OOR2);
931			FWD(aft, aft, d);
932			break;
933		case OOR1:		/* done a branch, find the O_CH */
934			if (ISSTATEIN(aft, here)) {
935				for (look = 1; /**/; look += d) {
936					s = g->strip[pc+look];
937					d = g->stripdata[pc+look];
938					if (s == O_CH)
939						break;
940					assert(s == OOR2);
941				}
942				FWD(aft, aft, look);
943			}
944			break;
945		case OOR2:		/* propagate OCH_'s marking */
946			FWD(aft, aft, 1);
947			if (g->strip[pc+d] != O_CH) {
948				assert(g->strip[pc+d] == OOR2);
949				FWD(aft, aft, d);
950			}
951			break;
952		case O_CH:		/* just empty */
953			FWD(aft, aft, 1);
954			break;
955		default:		/* ooooops... */
956			assert(nope);
957			break;
958		}
959	}
960
961	return(aft);
962}
963
964#ifdef REDEBUG
965/*
966 - print - print a set of states
967 */
968static void
969print(struct match *m, char *caption, states st, int ch, FILE *d)
970{
971	struct re_guts *g = m->g;
972	int i;
973	int first = 1;
974
975	if (!(m->eflags&REG_TRACE))
976		return;
977
978	fprintf(d, "%s", caption);
979	if (ch != '\0')
980		fprintf(d, " %s", pchar(ch));
981	for (i = 0; i < g->nstates; i++)
982		if (ISSET(st, i)) {
983			fprintf(d, "%s%d", (first) ? "\t" : ", ", i);
984			first = 0;
985		}
986	fprintf(d, "\n");
987}
988
989/*
990 - at - print current situation
991 */
992static void
993at(struct match *m, char *title, char *start, char *stop, sopno startst,
994    sopno stopst)
995{
996	if (!(m->eflags&REG_TRACE))
997		return;
998
999	printf("%s %s-", title, pchar(*start));
1000	printf("%s ", pchar(*stop));
1001	printf("%ld-%ld\n", (long)startst, (long)stopst);
1002}
1003
1004#ifndef PCHARDONE
1005#define	PCHARDONE	/* never again */
1006/*
1007 - pchar - make a character printable
1008 *
1009 * Is this identical to regchar() over in debug.c?  Well, yes.  But a
1010 * duplicate here avoids having a debugging-capable regexec.o tied to
1011 * a matching debug.o, and this is convenient.  It all disappears in
1012 * the non-debug compilation anyway, so it doesn't matter much.
1013 */
1014static char *			/* -> representation */
1015pchar(int ch)
1016{
1017	static char pbuf[10];
1018
1019	if (isprint(ch) || ch == ' ')
1020		snprintf(pbuf, sizeof(pbuf), "%c", ch);
1021	else
1022		snprintf(pbuf, sizeof(pbuf), "\\%o", ch);
1023	return(pbuf);
1024}
1025#endif
1026#endif
1027
1028#undef	matcher
1029#undef	fast
1030#undef	slow
1031#undef	dissect
1032#undef	backref
1033#undef	step
1034#undef	print
1035#undef	at
1036#undef	match
1037