2 * Copyright (c) 1992, 1993, 1994 Henry Spencer.
3 * Copyright (c) 1992, 1993, 1994
4 * The Regents of the University of California. All rights reserved.
6 * This code is derived from software contributed to Berkeley by
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 3. All advertising materials mentioning features or use of this software
18 * must display the following acknowledgement:
19 * This product includes software developed by the University of
20 * California, Berkeley and its contributors.
21 * 4. Neither the name of the University nor the names of its contributors
22 * may be used to endorse or promote products derived from this software
23 * without specific prior written permission.
25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
37 * @(#)engine.c 8.5 (Berkeley) 3/20/94
39 * $FreeBSD: src/lib/libc/regex/engine.c,v 1.5.8.1 2000/07/31 06:30:37 dcs Exp $
40 * $DragonFly: src/lib/libcr/regex/Attic/engine.c,v 1.3 2003/11/12 20:21:28 eirikn Exp $
44 * The matching engine and friends. This file is #included by regexec.c
45 * after suitable #defines of a variety of macros used herein, so that
46 * different state representations can be used without duplicating masses
51 #define matcher smatcher
54 #define dissect sdissect
55 #define backref sbackref
62 #define matcher lmatcher
65 #define dissect ldissect
66 #define backref lbackref
73 /* another structure passed up and down to avoid zillions of parameters */
77 regmatch_t *pmatch; /* [nsub+1] (0 element unused) */
78 char *offp; /* offsets work from here */
79 char *beginp; /* start of string -- virtual NUL precedes */
80 char *endp; /* end of string -- virtual NUL here */
81 char *coldp; /* can be no match starting before here */
82 char **lastpos; /* [nplus+1] */
84 states st; /* current states */
85 states fresh; /* states for a fresh start */
86 states tmp; /* temporary */
87 states empty; /* empty set of states */
90 /* ========= begin header generated by ./mkh ========= */
95 /* === engine.c === */
96 static int matcher (struct re_guts *g, char *string, size_t nmatch, regmatch_t pmatch[], int eflags);
97 static char *dissect (struct match *m, char *start, char *stop, sopno startst, sopno stopst);
98 static char *backref (struct match *m, char *start, char *stop, sopno startst, sopno stopst, sopno lev);
99 static char *fast (struct match *m, char *start, char *stop, sopno startst, sopno stopst);
100 static char *slow (struct match *m, char *start, char *stop, sopno startst, sopno stopst);
101 static states step (struct re_guts *g, sopno start, sopno stop, states bef, int ch, states aft);
104 #define BOLEOL (BOL+2)
105 #define NOTHING (BOL+3)
108 #define CODEMAX (BOL+5) /* highest code used */
109 #define NONCHAR(c) ((c) > CHAR_MAX)
110 #define NNONCHAR (CODEMAX-CHAR_MAX)
112 static void print (struct match *m, char *caption, states st, int ch, FILE *d);
115 static void at (struct match *m, char *title, char *start, char *stop, sopno startst, sopno stopst);
118 static char *pchar (int ch);
124 /* ========= end header generated by ./mkh ========= */
127 #define SP(t, s, c) print(m, t, s, c, stdout)
128 #define AT(t, p1, p2, s1, s2) at(m, t, p1, p2, s1, s2)
129 #define NOTE(str) { if (m->eflags®_TRACE) printf("=%s\n", (str)); }
131 #define SP(t, s, c) /* nothing */
132 #define AT(t, p1, p2, s1, s2) /* nothing */
133 #define NOTE(s) /* nothing */
137 - matcher - the actual matching engine
138 == static int matcher(register struct re_guts *g, char *string, \
139 == size_t nmatch, regmatch_t pmatch[], int eflags);
141 static int /* 0 success, REG_NOMATCH failure */
142 matcher(g, string, nmatch, pmatch, eflags)
143 register struct re_guts *g;
152 register struct match *m = &mv;
154 register const sopno gf = g->firststate+1; /* +1 for OEND */
155 register const sopno gl = g->laststate;
158 /* Boyer-Moore algorithms variables */
161 register char *mustfirst;
162 register char *mustlast;
163 register int *matchjump;
164 register int *charjump;
166 /* simplify the situation where possible */
167 if (g->cflags®_NOSUB)
169 if (eflags®_STARTEND) {
170 start = string + pmatch[0].rm_so;
171 stop = string + pmatch[0].rm_eo;
174 stop = start + strlen(start);
179 /* prescreening; this does wonders for this rather slow code */
180 if (g->must != NULL) {
181 if (g->charjump != NULL && g->matchjump != NULL) {
183 mustlast = g->must + g->mlen - 1;
184 charjump = g->charjump;
185 matchjump = g->matchjump;
187 for (dp = start+g->mlen-1; dp < stop;) {
188 /* Fast skip non-matches */
189 while (dp < stop && charjump[*dp])
196 /* We depend on not being used for
197 * for strings of length 1
199 while (*--dp == *--pp && pp != mustfirst);
204 /* Jump to next possible match */
205 mj = matchjump[pp - mustfirst];
207 dp += (cj < mj ? mj : cj);
213 for (dp = start; dp < stop; dp++)
214 if (*dp == g->must[0] &&
215 stop - dp >= g->mlen &&
216 memcmp(dp, g->must, (size_t)g->mlen) == 0)
218 if (dp == stop) /* we didn't find g->must */
223 /* match struct setup */
238 /* Adjust start according to moffset, to speed things up */
240 start = ((dp - g->moffset) < start) ? start : dp - g->moffset;
242 /* this loop does only one repetition except for backrefs */
244 endp = fast(m, start, stop, gf, gl);
245 if (endp == NULL) { /* a miss */
249 if (nmatch == 0 && !g->backrefs)
250 break; /* no further info needed */
253 assert(m->coldp != NULL);
255 NOTE("finding start");
256 endp = slow(m, m->coldp, stop, gf, gl);
259 assert(m->coldp < m->endp);
262 if (nmatch == 1 && !g->backrefs)
263 break; /* no further info needed */
265 /* oh my, he wants the subexpressions... */
266 if (m->pmatch == NULL)
267 m->pmatch = (regmatch_t *)malloc((m->g->nsub + 1) *
269 if (m->pmatch == NULL) {
273 for (i = 1; i <= m->g->nsub; i++)
274 m->pmatch[i].rm_so = m->pmatch[i].rm_eo = -1;
275 if (!g->backrefs && !(m->eflags®_BACKR)) {
277 dp = dissect(m, m->coldp, endp, gf, gl);
279 if (g->nplus > 0 && m->lastpos == NULL)
280 m->lastpos = (char **)malloc((g->nplus+1) *
282 if (g->nplus > 0 && m->lastpos == NULL) {
287 NOTE("backref dissect");
288 dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
293 /* uh-oh... we couldn't find a subexpression-level match */
294 assert(g->backrefs); /* must be back references doing it */
295 assert(g->nplus == 0 || m->lastpos != NULL);
297 if (dp != NULL || endp <= m->coldp)
300 endp = slow(m, m->coldp, endp-1, gf, gl);
303 /* try it on a shorter possibility */
305 for (i = 1; i <= m->g->nsub; i++) {
306 assert(m->pmatch[i].rm_so == -1);
307 assert(m->pmatch[i].rm_eo == -1);
310 NOTE("backoff dissect");
311 dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
313 assert(dp == NULL || dp == endp);
314 if (dp != NULL) /* found a shorter one */
317 /* despite initial appearances, there is no match here */
319 start = m->coldp + 1; /* recycle starting later */
320 assert(start <= stop);
323 /* fill in the details if requested */
325 pmatch[0].rm_so = m->coldp - m->offp;
326 pmatch[0].rm_eo = endp - m->offp;
329 assert(m->pmatch != NULL);
330 for (i = 1; i < nmatch; i++)
332 pmatch[i] = m->pmatch[i];
334 pmatch[i].rm_so = -1;
335 pmatch[i].rm_eo = -1;
339 if (m->pmatch != NULL)
340 free((char *)m->pmatch);
341 if (m->lastpos != NULL)
342 free((char *)m->lastpos);
348 - dissect - figure out what matched what, no back references
349 == static char *dissect(register struct match *m, char *start, \
350 == char *stop, sopno startst, sopno stopst);
352 static char * /* == stop (success) always */
353 dissect(m, start, stop, startst, stopst)
354 register struct match *m;
361 register sopno ss; /* start sop of current subRE */
362 register sopno es; /* end sop of current subRE */
363 register char *sp; /* start of string matched by it */
364 register char *stp; /* string matched by it cannot pass here */
365 register char *rest; /* start of rest of string */
366 register char *tail; /* string unmatched by rest of RE */
367 register sopno ssub; /* start sop of subsubRE */
368 register sopno esub; /* end sop of subsubRE */
369 register char *ssp; /* start of string matched by subsubRE */
370 register char *sep; /* end of string matched by subsubRE */
371 register char *oldssp; /* previous ssp */
374 AT("diss", start, stop, startst, stopst);
376 for (ss = startst; ss < stopst; ss = es) {
377 /* identify end of subRE */
379 switch (OP(m->g->strip[es])) {
382 es += OPND(m->g->strip[es]);
385 while (OP(m->g->strip[es]) != O_CH)
386 es += OPND(m->g->strip[es]);
391 /* figure out what it matched */
392 switch (OP(m->g->strip[ss])) {
412 /* cases where length of match is hard to find */
416 /* how long could this one be? */
417 rest = slow(m, sp, stp, ss, es);
418 assert(rest != NULL); /* it did match */
419 /* could the rest match the rest? */
420 tail = slow(m, rest, stop, es, stopst);
423 /* no -- try a shorter match for this one */
425 assert(stp >= sp); /* it did work */
429 /* did innards match? */
430 if (slow(m, sp, rest, ssub, esub) != NULL) {
431 dp = dissect(m, sp, rest, ssub, esub);
440 /* how long could this one be? */
441 rest = slow(m, sp, stp, ss, es);
442 assert(rest != NULL); /* it did match */
443 /* could the rest match the rest? */
444 tail = slow(m, rest, stop, es, stopst);
447 /* no -- try a shorter match for this one */
449 assert(stp >= sp); /* it did work */
455 for (;;) { /* find last match of innards */
456 sep = slow(m, ssp, rest, ssub, esub);
457 if (sep == NULL || sep == ssp)
458 break; /* failed or matched null */
459 oldssp = ssp; /* on to next try */
463 /* last successful match */
467 assert(sep == rest); /* must exhaust substring */
468 assert(slow(m, ssp, sep, ssub, esub) == rest);
469 dp = dissect(m, ssp, sep, ssub, esub);
476 /* how long could this one be? */
477 rest = slow(m, sp, stp, ss, es);
478 assert(rest != NULL); /* it did match */
479 /* could the rest match the rest? */
480 tail = slow(m, rest, stop, es, stopst);
483 /* no -- try a shorter match for this one */
485 assert(stp >= sp); /* it did work */
488 esub = ss + OPND(m->g->strip[ss]) - 1;
489 assert(OP(m->g->strip[esub]) == OOR1);
490 for (;;) { /* find first matching branch */
491 if (slow(m, sp, rest, ssub, esub) == rest)
492 break; /* it matched all of it */
493 /* that one missed, try next one */
494 assert(OP(m->g->strip[esub]) == OOR1);
496 assert(OP(m->g->strip[esub]) == OOR2);
498 esub += OPND(m->g->strip[esub]);
499 if (OP(m->g->strip[esub]) == OOR2)
502 assert(OP(m->g->strip[esub]) == O_CH);
504 dp = dissect(m, sp, rest, ssub, esub);
516 i = OPND(m->g->strip[ss]);
517 assert(0 < i && i <= m->g->nsub);
518 m->pmatch[i].rm_so = sp - m->offp;
521 i = OPND(m->g->strip[ss]);
522 assert(0 < i && i <= m->g->nsub);
523 m->pmatch[i].rm_eo = sp - m->offp;
536 - backref - figure out what matched what, figuring in back references
537 == static char *backref(register struct match *m, char *start, \
538 == char *stop, sopno startst, sopno stopst, sopno lev);
540 static char * /* == stop (success) or NULL (failure) */
541 backref(m, start, stop, startst, stopst, lev)
542 register struct match *m;
547 sopno lev; /* PLUS nesting level */
550 register sopno ss; /* start sop of current subRE */
551 register char *sp; /* start of string matched by it */
552 register sopno ssub; /* start sop of subsubRE */
553 register sopno esub; /* end sop of subsubRE */
554 register char *ssp; /* start of string matched by subsubRE */
559 register regoff_t offsave;
562 AT("back", start, stop, startst, stopst);
565 /* get as far as we can with easy stuff */
567 for (ss = startst; !hard && ss < stopst; ss++)
568 switch (OP(s = m->g->strip[ss])) {
570 if (sp == stop || *sp++ != (char)OPND(s))
579 cs = &m->g->sets[OPND(s)];
580 if (sp == stop || !CHIN(cs, *sp++))
584 if ( (sp == m->beginp && !(m->eflags®_NOTBOL)) ||
585 (sp < m->endp && *(sp-1) == '\n' &&
586 (m->g->cflags®_NEWLINE)) )
592 if ( (sp == m->endp && !(m->eflags®_NOTEOL)) ||
593 (sp < m->endp && *sp == '\n' &&
594 (m->g->cflags®_NEWLINE)) )
600 if (( (sp == m->beginp && !(m->eflags®_NOTBOL)) ||
601 (sp < m->endp && *(sp-1) == '\n' &&
602 (m->g->cflags®_NEWLINE)) ||
604 !ISWORD(*(sp-1))) ) &&
605 (sp < m->endp && ISWORD(*sp)) )
611 if (( (sp == m->endp && !(m->eflags®_NOTEOL)) ||
612 (sp < m->endp && *sp == '\n' &&
613 (m->g->cflags®_NEWLINE)) ||
614 (sp < m->endp && !ISWORD(*sp)) ) &&
615 (sp > m->beginp && ISWORD(*(sp-1))) )
622 case OOR1: /* matches null but needs to skip */
626 assert(OP(s) == OOR2);
628 } while (OP(s = m->g->strip[ss]) != O_CH);
629 /* note that the ss++ gets us past the O_CH */
631 default: /* have to make a choice */
635 if (!hard) { /* that was it! */
640 ss--; /* adjust for the for's final increment */
643 AT("hard", sp, stop, ss, stopst);
646 case OBACK_: /* the vilest depths */
648 assert(0 < i && i <= m->g->nsub);
649 if (m->pmatch[i].rm_eo == -1)
651 assert(m->pmatch[i].rm_so != -1);
652 len = m->pmatch[i].rm_eo - m->pmatch[i].rm_so;
653 assert(stop - m->beginp >= len);
655 return(NULL); /* not enough left to match */
656 ssp = m->offp + m->pmatch[i].rm_so;
657 if (memcmp(sp, ssp, len) != 0)
659 while (m->g->strip[ss] != SOP(O_BACK, i))
661 return(backref(m, sp+len, stop, ss+1, stopst, lev));
663 case OQUEST_: /* to null or not */
664 dp = backref(m, sp, stop, ss+1, stopst, lev);
666 return(dp); /* not */
667 return(backref(m, sp, stop, ss+OPND(s)+1, stopst, lev));
670 assert(m->lastpos != NULL);
671 assert(lev+1 <= m->g->nplus);
672 m->lastpos[lev+1] = sp;
673 return(backref(m, sp, stop, ss+1, stopst, lev+1));
676 if (sp == m->lastpos[lev]) /* last pass matched null */
677 return(backref(m, sp, stop, ss+1, stopst, lev-1));
678 /* try another pass */
679 m->lastpos[lev] = sp;
680 dp = backref(m, sp, stop, ss-OPND(s)+1, stopst, lev);
682 return(backref(m, sp, stop, ss+1, stopst, lev-1));
686 case OCH_: /* find the right one, if any */
688 esub = ss + OPND(s) - 1;
689 assert(OP(m->g->strip[esub]) == OOR1);
690 for (;;) { /* find first matching branch */
691 dp = backref(m, sp, stop, ssub, esub, lev);
694 /* that one missed, try next one */
695 if (OP(m->g->strip[esub]) == O_CH)
696 return(NULL); /* there is none */
698 assert(OP(m->g->strip[esub]) == OOR2);
700 esub += OPND(m->g->strip[esub]);
701 if (OP(m->g->strip[esub]) == OOR2)
704 assert(OP(m->g->strip[esub]) == O_CH);
707 case OLPAREN: /* must undo assignment if rest fails */
709 assert(0 < i && i <= m->g->nsub);
710 offsave = m->pmatch[i].rm_so;
711 m->pmatch[i].rm_so = sp - m->offp;
712 dp = backref(m, sp, stop, ss+1, stopst, lev);
715 m->pmatch[i].rm_so = offsave;
718 case ORPAREN: /* must undo assignment if rest fails */
720 assert(0 < i && i <= m->g->nsub);
721 offsave = m->pmatch[i].rm_eo;
722 m->pmatch[i].rm_eo = sp - m->offp;
723 dp = backref(m, sp, stop, ss+1, stopst, lev);
726 m->pmatch[i].rm_eo = offsave;
737 return "shut up gcc";
741 - fast - step through the string at top speed
742 == static char *fast(register struct match *m, char *start, \
743 == char *stop, sopno startst, sopno stopst);
745 static char * /* where tentative match ended, or NULL */
746 fast(m, start, stop, startst, stopst)
747 register struct match *m;
753 register states st = m->st;
754 register states fresh = m->fresh;
755 register states tmp = m->tmp;
756 register char *p = start;
757 register int c = (start == m->beginp) ? OUT : *(start-1);
758 register int lastc; /* previous c */
761 register char *coldp; /* last p after which no match was underway */
765 st = step(m->g, startst, stopst, st, NOTHING, st);
772 c = (p == m->endp) ? OUT : *p;
776 /* is there an EOL and/or BOL between lastc and c? */
779 if ( (lastc == '\n' && m->g->cflags®_NEWLINE) ||
780 (lastc == OUT && !(m->eflags®_NOTBOL)) ) {
784 if ( (c == '\n' && m->g->cflags®_NEWLINE) ||
785 (c == OUT && !(m->eflags®_NOTEOL)) ) {
786 flagch = (flagch == BOL) ? BOLEOL : EOL;
791 st = step(m->g, startst, stopst, st, flagch, st);
795 /* how about a word boundary? */
796 if ( (flagch == BOL || (lastc != OUT && !ISWORD(lastc))) &&
797 (c != OUT && ISWORD(c)) ) {
800 if ( (lastc != OUT && ISWORD(lastc)) &&
801 (flagch == EOL || (c != OUT && !ISWORD(c))) ) {
804 if (flagch == BOW || flagch == EOW) {
805 st = step(m->g, startst, stopst, st, flagch, st);
810 if (ISSET(st, stopst) || p == stop)
811 break; /* NOTE BREAK OUT */
813 /* no, we must deal with this character */
817 st = step(m->g, startst, stopst, tmp, c, st);
819 assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st));
823 assert(coldp != NULL);
825 if (ISSET(st, stopst))
832 - slow - step through the string more deliberately
833 == static char *slow(register struct match *m, char *start, \
834 == char *stop, sopno startst, sopno stopst);
836 static char * /* where it ended */
837 slow(m, start, stop, startst, stopst)
838 register struct match *m;
844 register states st = m->st;
845 register states empty = m->empty;
846 register states tmp = m->tmp;
847 register char *p = start;
848 register int c = (start == m->beginp) ? OUT : *(start-1);
849 register int lastc; /* previous c */
852 register char *matchp; /* last p at which a match ended */
854 AT("slow", start, stop, startst, stopst);
857 SP("sstart", st, *p);
858 st = step(m->g, startst, stopst, st, NOTHING, st);
863 c = (p == m->endp) ? OUT : *p;
865 /* is there an EOL and/or BOL between lastc and c? */
868 if ( (lastc == '\n' && m->g->cflags®_NEWLINE) ||
869 (lastc == OUT && !(m->eflags®_NOTBOL)) ) {
873 if ( (c == '\n' && m->g->cflags®_NEWLINE) ||
874 (c == OUT && !(m->eflags®_NOTEOL)) ) {
875 flagch = (flagch == BOL) ? BOLEOL : EOL;
880 st = step(m->g, startst, stopst, st, flagch, st);
881 SP("sboleol", st, c);
884 /* how about a word boundary? */
885 if ( (flagch == BOL || (lastc != OUT && !ISWORD(lastc))) &&
886 (c != OUT && ISWORD(c)) ) {
889 if ( (lastc != OUT && ISWORD(lastc)) &&
890 (flagch == EOL || (c != OUT && !ISWORD(c))) ) {
893 if (flagch == BOW || flagch == EOW) {
894 st = step(m->g, startst, stopst, st, flagch, st);
895 SP("sboweow", st, c);
899 if (ISSET(st, stopst))
901 if (EQ(st, empty) || p == stop)
902 break; /* NOTE BREAK OUT */
904 /* no, we must deal with this character */
908 st = step(m->g, startst, stopst, tmp, c, st);
910 assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st));
919 - step - map set of states reachable before char to set reachable after
920 == static states step(register struct re_guts *g, sopno start, sopno stop, \
921 == register states bef, int ch, register states aft);
922 == #define BOL (OUT+1)
923 == #define EOL (BOL+1)
924 == #define BOLEOL (BOL+2)
925 == #define NOTHING (BOL+3)
926 == #define BOW (BOL+4)
927 == #define EOW (BOL+5)
928 == #define CODEMAX (BOL+5) // highest code used
929 == #define NONCHAR(c) ((c) > CHAR_MAX)
930 == #define NNONCHAR (CODEMAX-CHAR_MAX)
933 step(g, start, stop, bef, ch, aft)
934 register struct re_guts *g;
935 sopno start; /* start state within strip */
936 sopno stop; /* state after stop state within strip */
937 register states bef; /* states reachable before */
938 int ch; /* character or NONCHAR code */
939 register states aft; /* states already known reachable after */
944 register onestate here; /* note, macros know this name */
948 for (pc = start, INIT(here, pc); pc != stop; pc++, INC(here)) {
952 assert(pc == stop-1);
955 /* only characters can match */
956 assert(!NONCHAR(ch) || ch != (char)OPND(s));
957 if (ch == (char)OPND(s))
961 if (ch == BOL || ch == BOLEOL)
965 if (ch == EOL || ch == BOLEOL)
981 cs = &g->sets[OPND(s)];
982 if (!NONCHAR(ch) && CHIN(cs, ch))
985 case OBACK_: /* ignored here */
989 case OPLUS_: /* forward, this is just an empty */
992 case O_PLUS: /* both forward and back */
994 i = ISSETBACK(aft, OPND(s));
995 BACK(aft, aft, OPND(s));
996 if (!i && ISSETBACK(aft, OPND(s))) {
997 /* oho, must reconsider loop body */
1002 case OQUEST_: /* two branches, both forward */
1004 FWD(aft, aft, OPND(s));
1006 case O_QUEST: /* just an empty */
1009 case OLPAREN: /* not significant here */
1013 case OCH_: /* mark the first two branches */
1015 assert(OP(g->strip[pc+OPND(s)]) == OOR2);
1016 FWD(aft, aft, OPND(s));
1018 case OOR1: /* done a branch, find the O_CH */
1019 if (ISSTATEIN(aft, here)) {
1021 OP(s = g->strip[pc+look]) != O_CH;
1023 assert(OP(s) == OOR2);
1024 FWD(aft, aft, look);
1027 case OOR2: /* propagate OCH_'s marking */
1029 if (OP(g->strip[pc+OPND(s)]) != O_CH) {
1030 assert(OP(g->strip[pc+OPND(s)]) == OOR2);
1031 FWD(aft, aft, OPND(s));
1034 case O_CH: /* just empty */
1037 default: /* ooooops... */
1048 - print - print a set of states
1050 == static void print(struct match *m, char *caption, states st, \
1051 == int ch, FILE *d);
1055 print(m, caption, st, ch, d)
1062 register struct re_guts *g = m->g;
1064 register int first = 1;
1066 if (!(m->eflags®_TRACE))
1069 fprintf(d, "%s", caption);
1071 fprintf(d, " %s", pchar(ch));
1072 for (i = 0; i < g->nstates; i++)
1074 fprintf(d, "%s%d", (first) ? "\t" : ", ", i);
1081 - at - print current situation
1083 == static void at(struct match *m, char *title, char *start, char *stop, \
1084 == sopno startst, sopno stopst);
1088 at(m, title, start, stop, startst, stopst)
1096 if (!(m->eflags®_TRACE))
1099 printf("%s %s-", title, pchar(*start));
1100 printf("%s ", pchar(*stop));
1101 printf("%ld-%ld\n", (long)startst, (long)stopst);
1105 #define PCHARDONE /* never again */
1107 - pchar - make a character printable
1109 == static char *pchar(int ch);
1112 * Is this identical to regchar() over in debug.c? Well, yes. But a
1113 * duplicate here avoids having a debugging-capable regexec.o tied to
1114 * a matching debug.o, and this is convenient. It all disappears in
1115 * the non-debug compilation anyway, so it doesn't matter much.
1117 static char * /* -> representation */
1121 static char pbuf[10];
1123 if (isprint((uch)ch) || ch == ' ')
1124 sprintf(pbuf, "%c", ch);
1126 sprintf(pbuf, "\\%o", ch);