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 * @(#)regcomp.c 8.5 (Berkeley) 3/20/94
39 * $FreeBSD: src/lib/libc/regex/regcomp.c,v 1.13.2.1 2000/07/31 06:30:37 dcs Exp $
42 #if defined(LIBC_SCCS) && !defined(lint)
43 static char sccsid[] = "@(#)regcomp.c 8.5 (Berkeley) 3/20/94";
44 #endif /* LIBC_SCCS and not lint */
46 #include <sys/types.h>
63 * parse structure, passed up and down to avoid global variables and
67 char *next; /* next character in RE */
68 char *end; /* end of string (-> NUL normally) */
69 int error; /* has an error been seen? */
70 sop *strip; /* malloced strip */
71 sopno ssize; /* malloced strip size (allocated) */
72 sopno slen; /* malloced strip length (used) */
73 int ncsalloc; /* number of csets allocated */
75 # define NPAREN 10 /* we need to remember () 1-9 for back refs */
76 sopno pbegin[NPAREN]; /* -> ( ([0] unused) */
77 sopno pend[NPAREN]; /* -> ) ([0] unused) */
80 /* ========= begin header generated by ./mkh ========= */
85 /* === regcomp.c === */
86 static void p_ere __P((struct parse *p, int stop));
87 static void p_ere_exp __P((struct parse *p));
88 static void p_str __P((struct parse *p));
89 static void p_bre __P((struct parse *p, int end1, int end2));
90 static int p_simp_re __P((struct parse *p, int starordinary));
91 static int p_count __P((struct parse *p));
92 static void p_bracket __P((struct parse *p));
93 static void p_b_term __P((struct parse *p, cset *cs));
94 static void p_b_cclass __P((struct parse *p, cset *cs));
95 static void p_b_eclass __P((struct parse *p, cset *cs));
96 static char p_b_symbol __P((struct parse *p));
97 static char p_b_coll_elem __P((struct parse *p, int endc));
98 static char othercase __P((int ch));
99 static void bothcases __P((struct parse *p, int ch));
100 static void ordinary __P((struct parse *p, int ch));
101 static void nonnewline __P((struct parse *p));
102 static void repeat __P((struct parse *p, sopno start, int from, int to));
103 static int seterr __P((struct parse *p, int e));
104 static cset *allocset __P((struct parse *p));
105 static void freeset __P((struct parse *p, cset *cs));
106 static int freezeset __P((struct parse *p, cset *cs));
107 static int firstch __P((struct parse *p, cset *cs));
108 static int nch __P((struct parse *p, cset *cs));
109 static void mcadd __P((struct parse *p, cset *cs, char *cp));
111 static void mcsub __P((cset *cs, char *cp));
112 static int mcin __P((cset *cs, char *cp));
113 static char *mcfind __P((cset *cs, char *cp));
115 static void mcinvert __P((struct parse *p, cset *cs));
116 static void mccase __P((struct parse *p, cset *cs));
117 static int isinsets __P((struct re_guts *g, int c));
118 static int samesets __P((struct re_guts *g, int c1, int c2));
119 static void categorize __P((struct parse *p, struct re_guts *g));
120 static sopno dupl __P((struct parse *p, sopno start, sopno finish));
121 static void doemit __P((struct parse *p, sop op, size_t opnd));
122 static void doinsert __P((struct parse *p, sop op, size_t opnd, sopno pos));
123 static void dofwd __P((struct parse *p, sopno pos, sop value));
124 static void enlarge __P((struct parse *p, sopno size));
125 static void stripsnug __P((struct parse *p, struct re_guts *g));
126 static void findmust __P((struct parse *p, struct re_guts *g));
127 static int altoffset __P((sop *scan, int offset, int mccs));
128 static void computejumps __P((struct parse *p, struct re_guts *g));
129 static void computematchjumps __P((struct parse *p, struct re_guts *g));
130 static sopno pluscount __P((struct parse *p, struct re_guts *g));
135 /* ========= end header generated by ./mkh ========= */
137 static char nuls[10]; /* place to point scanner in event of error */
140 * macros for use with parse structure
141 * BEWARE: these know that the parse structure is named `p' !!!
143 #define PEEK() (*p->next)
144 #define PEEK2() (*(p->next+1))
145 #define MORE() (p->next < p->end)
146 #define MORE2() (p->next+1 < p->end)
147 #define SEE(c) (MORE() && PEEK() == (c))
148 #define SEETWO(a, b) (MORE() && MORE2() && PEEK() == (a) && PEEK2() == (b))
149 #define EAT(c) ((SEE(c)) ? (NEXT(), 1) : 0)
150 #define EATTWO(a, b) ((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
151 #define NEXT() (p->next++)
152 #define NEXT2() (p->next += 2)
153 #define NEXTn(n) (p->next += (n))
154 #define GETNEXT() (*p->next++)
155 #define SETERROR(e) seterr(p, (e))
156 #define REQUIRE(co, e) ((co) || SETERROR(e))
157 #define MUSTSEE(c, e) (REQUIRE(MORE() && PEEK() == (c), e))
158 #define MUSTEAT(c, e) (REQUIRE(MORE() && GETNEXT() == (c), e))
159 #define MUSTNOTSEE(c, e) (REQUIRE(!MORE() || PEEK() != (c), e))
160 #define EMIT(op, sopnd) doemit(p, (sop)(op), (size_t)(sopnd))
161 #define INSERT(op, pos) doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
162 #define AHEAD(pos) dofwd(p, pos, HERE()-(pos))
163 #define ASTERN(sop, pos) EMIT(sop, HERE()-pos)
164 #define HERE() (p->slen)
165 #define THERE() (p->slen - 1)
166 #define THERETHERE() (p->slen - 2)
167 #define DROP(n) (p->slen -= (n))
170 static int never = 0; /* for use in asserts; shuts lint up */
172 #define never 0 /* some <assert.h>s have bugs too */
175 /* Macro used by computejump()/computematchjump() */
176 #define MIN(a,b) ((a)<(b)?(a):(b))
179 - regcomp - interface for parser and compilation
180 = extern int regcomp(regex_t *, const char *, int);
181 = #define REG_BASIC 0000
182 = #define REG_EXTENDED 0001
183 = #define REG_ICASE 0002
184 = #define REG_NOSUB 0004
185 = #define REG_NEWLINE 0010
186 = #define REG_NOSPEC 0020
187 = #define REG_PEND 0040
188 = #define REG_DUMP 0200
190 int /* 0 success, otherwise REG_something */
191 regcomp(preg, pattern, cflags)
197 register struct re_guts *g;
198 register struct parse *p = &pa;
202 # define GOODFLAGS(f) (f)
204 # define GOODFLAGS(f) ((f)&~REG_DUMP)
207 cflags = GOODFLAGS(cflags);
208 if ((cflags®_EXTENDED) && (cflags®_NOSPEC))
211 if (cflags®_PEND) {
212 if (preg->re_endp < pattern)
214 len = preg->re_endp - pattern;
216 len = strlen((char *)pattern);
218 /* do the mallocs early so failure handling is easy */
219 g = (struct re_guts *)malloc(sizeof(struct re_guts) +
220 (NC-1)*sizeof(cat_t));
223 p->ssize = len/(size_t)2*(size_t)3 + (size_t)1; /* ugh */
224 p->strip = (sop *)malloc(p->ssize * sizeof(sop));
226 if (p->strip == NULL) {
233 p->next = (char *)pattern; /* convenience; we do not modify it */
234 p->end = p->next + len;
237 for (i = 0; i < NPAREN; i++) {
255 g->ncategories = 1; /* category 0 is "everything else" */
256 g->categories = &g->catspace[-(CHAR_MIN)];
257 (void) memset((char *)g->catspace, 0, NC*sizeof(cat_t));
262 g->firststate = THERE();
263 if (cflags®_EXTENDED)
265 else if (cflags®_NOSPEC)
270 g->laststate = THERE();
272 /* tidy up loose ends and fill things in */
276 /* only use Boyer-Moore algorithm if the pattern is bigger
277 * than three characters
281 computematchjumps(p, g);
282 if(g->matchjump == NULL && g->charjump != NULL) {
287 g->nplus = pluscount(p, g);
289 preg->re_nsub = g->nsub;
291 preg->re_magic = MAGIC1;
293 /* not debugging, so can't rely on the assert() in regexec() */
295 SETERROR(REG_ASSERT);
298 /* win or lose, we're done */
299 if (p->error != 0) /* lose */
305 - p_ere - ERE parser top level, concatenation and alternation
306 == static void p_ere(register struct parse *p, int stop);
310 register struct parse *p;
311 int stop; /* character this ERE should end at */
314 register sopno prevback;
315 register sopno prevfwd;
317 register int first = 1; /* is this the first alternative? */
320 /* do a bunch of concatenated expressions */
322 while (MORE() && (c = PEEK()) != '|' && c != stop)
324 (void)REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
327 break; /* NOTE BREAK OUT */
330 INSERT(OCH_, conc); /* offset is wrong */
335 ASTERN(OOR1, prevback);
337 AHEAD(prevfwd); /* fix previous offset */
339 EMIT(OOR2, 0); /* offset is very wrong */
342 if (!first) { /* tail-end fixups */
344 ASTERN(O_CH, prevback);
347 assert(!MORE() || SEE(stop));
351 - p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
352 == static void p_ere_exp(register struct parse *p);
356 register struct parse *p;
362 register sopno subno;
365 assert(MORE()); /* caller should have ensured this */
371 (void)REQUIRE(MORE(), REG_EPAREN);
375 p->pbegin[subno] = HERE();
376 EMIT(OLPAREN, subno);
379 if (subno < NPAREN) {
380 p->pend[subno] = HERE();
381 assert(p->pend[subno] != 0);
383 EMIT(ORPAREN, subno);
384 (void)MUSTEAT(')', REG_EPAREN);
386 #ifndef POSIX_MISTAKE
387 case ')': /* happens only if no current unmatched ( */
389 * You may ask, why the ifndef? Because I didn't notice
390 * this until slightly too late for 1003.2, and none of the
391 * other 1003.2 regular-expression reviewers noticed it at
392 * all. So an unmatched ) is legal POSIX, at least until
393 * we can get it fixed.
395 SETERROR(REG_EPAREN);
400 p->g->iflags |= USEBOL;
406 p->g->iflags |= USEEOL;
415 SETERROR(REG_BADRPT);
418 if (p->g->cflags®_NEWLINE)
427 (void)REQUIRE(MORE(), REG_EESCAPE);
431 case '{': /* okay as ordinary except if digit follows */
432 (void)REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
442 /* we call { a repetition if followed by a digit */
443 if (!( c == '*' || c == '+' || c == '?' ||
444 (c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
445 return; /* no repetition, we're done */
448 (void)REQUIRE(!wascaret, REG_BADRPT);
450 case '*': /* implemented as +? */
451 /* this case does not require the (y|) trick, noKLUDGE */
454 INSERT(OQUEST_, pos);
455 ASTERN(O_QUEST, pos);
462 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
463 INSERT(OCH_, pos); /* offset slightly wrong */
464 ASTERN(OOR1, pos); /* this one's right */
465 AHEAD(pos); /* fix the OCH_ */
466 EMIT(OOR2, 0); /* offset very wrong... */
467 AHEAD(THERE()); /* ...so fix it */
468 ASTERN(O_CH, THERETHERE());
473 if (isdigit((uch)PEEK())) {
475 (void)REQUIRE(count <= count2, REG_BADBR);
476 } else /* single number with comma */
478 } else /* just a single number */
480 repeat(p, pos, count, count2);
481 if (!EAT('}')) { /* error heuristics */
482 while (MORE() && PEEK() != '}')
484 (void)REQUIRE(MORE(), REG_EBRACE);
493 if (!( c == '*' || c == '+' || c == '?' ||
494 (c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
496 SETERROR(REG_BADRPT);
500 - p_str - string (no metacharacters) "parser"
501 == static void p_str(register struct parse *p);
505 register struct parse *p;
507 (void)REQUIRE(MORE(), REG_EMPTY);
509 ordinary(p, GETNEXT());
513 - p_bre - BRE parser top level, anchoring and concatenation
514 == static void p_bre(register struct parse *p, register int end1, \
515 == register int end2);
516 * Giving end1 as OUT essentially eliminates the end1/end2 check.
518 * This implementation is a bit of a kludge, in that a trailing $ is first
519 * taken as an ordinary character and then revised to be an anchor. The
520 * only undesirable side effect is that '$' gets included as a character
521 * category in such cases. This is fairly harmless; not worth fixing.
522 * The amount of lookahead needed to avoid this kludge is excessive.
526 register struct parse *p;
527 register int end1; /* first terminating character */
528 register int end2; /* second terminating character */
530 register sopno start = HERE();
531 register int first = 1; /* first subexpression? */
532 register int wasdollar = 0;
536 p->g->iflags |= USEBOL;
539 while (MORE() && !SEETWO(end1, end2)) {
540 wasdollar = p_simp_re(p, first);
543 if (wasdollar) { /* oops, that was a trailing anchor */
546 p->g->iflags |= USEEOL;
550 (void)REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
554 - p_simp_re - parse a simple RE, an atom possibly followed by a repetition
555 == static int p_simp_re(register struct parse *p, int starordinary);
557 static int /* was the simple RE an unbackslashed $? */
558 p_simp_re(p, starordinary)
559 register struct parse *p;
560 int starordinary; /* is a leading * an ordinary character? */
567 register sopno subno;
568 # define BACKSL (1<<CHAR_BIT)
570 pos = HERE(); /* repetion op, if any, covers from here */
572 assert(MORE()); /* caller should have ensured this */
575 (void)REQUIRE(MORE(), REG_EESCAPE);
576 c = BACKSL | GETNEXT();
580 if (p->g->cflags®_NEWLINE)
589 SETERROR(REG_BADRPT);
595 p->pbegin[subno] = HERE();
596 EMIT(OLPAREN, subno);
597 /* the MORE here is an error heuristic */
598 if (MORE() && !SEETWO('\\', ')'))
600 if (subno < NPAREN) {
601 p->pend[subno] = HERE();
602 assert(p->pend[subno] != 0);
604 EMIT(ORPAREN, subno);
605 (void)REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
607 case BACKSL|')': /* should not get here -- must be user */
609 SETERROR(REG_EPAREN);
620 i = (c&~BACKSL) - '0';
622 if (p->pend[i] != 0) {
623 assert(i <= p->g->nsub);
625 assert(p->pbegin[i] != 0);
626 assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
627 assert(OP(p->strip[p->pend[i]]) == ORPAREN);
628 (void) dupl(p, p->pbegin[i]+1, p->pend[i]);
631 SETERROR(REG_ESUBREG);
635 (void)REQUIRE(starordinary, REG_BADRPT);
638 ordinary(p, (char)c);
642 if (EAT('*')) { /* implemented as +? */
643 /* this case does not require the (y|) trick, noKLUDGE */
646 INSERT(OQUEST_, pos);
647 ASTERN(O_QUEST, pos);
648 } else if (EATTWO('\\', '{')) {
651 if (MORE() && isdigit((uch)PEEK())) {
653 (void)REQUIRE(count <= count2, REG_BADBR);
654 } else /* single number with comma */
656 } else /* just a single number */
658 repeat(p, pos, count, count2);
659 if (!EATTWO('\\', '}')) { /* error heuristics */
660 while (MORE() && !SEETWO('\\', '}'))
662 (void)REQUIRE(MORE(), REG_EBRACE);
665 } else if (c == '$') /* $ (but not \$) ends it */
672 - p_count - parse a repetition count
673 == static int p_count(register struct parse *p);
675 static int /* the value */
677 register struct parse *p;
679 register int count = 0;
680 register int ndigits = 0;
682 while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
683 count = count*10 + (GETNEXT() - '0');
687 (void)REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
692 - p_bracket - parse a bracketed character list
693 == static void p_bracket(register struct parse *p);
695 * Note a significant property of this code: if the allocset() did SETERROR,
696 * no set operations are done.
700 register struct parse *p;
702 register cset *cs = allocset(p);
703 register int invert = 0;
705 /* Dept of Truly Sickening Special-Case Kludges */
706 if (p->next + 5 < p->end && strncmp(p->next, "[:<:]]", 6) == 0) {
711 if (p->next + 5 < p->end && strncmp(p->next, "[:>:]]", 6) == 0) {
718 invert++; /* make note to invert set at end */
723 while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
727 (void)MUSTEAT(']', REG_EBRACK);
729 if (p->error != 0) /* don't mess things up further */
732 if (p->g->cflags®_ICASE) {
736 for (i = p->g->csetsize - 1; i >= 0; i--)
737 if (CHIN(cs, i) && isalpha(i)) {
742 if (cs->multis != NULL)
748 for (i = p->g->csetsize - 1; i >= 0; i--)
753 if (p->g->cflags®_NEWLINE)
755 if (cs->multis != NULL)
759 assert(cs->multis == NULL); /* xxx */
761 if (nch(p, cs) == 1) { /* optimize singleton sets */
762 ordinary(p, firstch(p, cs));
765 EMIT(OANYOF, freezeset(p, cs));
769 - p_b_term - parse one term of a bracketed character list
770 == static void p_b_term(register struct parse *p, register cset *cs);
774 register struct parse *p;
778 register char start, finish;
781 /* classify what we've got */
782 switch ((MORE()) ? PEEK() : '\0') {
784 c = (MORE2()) ? PEEK2() : '\0';
787 SETERROR(REG_ERANGE);
788 return; /* NOTE RETURN */
796 case ':': /* character class */
798 (void)REQUIRE(MORE(), REG_EBRACK);
800 (void)REQUIRE(c != '-' && c != ']', REG_ECTYPE);
802 (void)REQUIRE(MORE(), REG_EBRACK);
803 (void)REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
805 case '=': /* equivalence class */
807 (void)REQUIRE(MORE(), REG_EBRACK);
809 (void)REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
811 (void)REQUIRE(MORE(), REG_EBRACK);
812 (void)REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
814 default: /* symbol, ordinary character, or range */
815 /* xxx revision needed for multichar stuff */
816 start = p_b_symbol(p);
817 if (SEE('-') && MORE2() && PEEK2() != ']') {
823 finish = p_b_symbol(p);
829 if (__collate_load_error) {
830 (void)REQUIRE((uch)start <= (uch)finish, REG_ERANGE);
831 for (i = (uch)start; i <= (uch)finish; i++)
834 (void)REQUIRE(__collate_range_cmp(start, finish) <= 0, REG_ERANGE);
835 for (i = CHAR_MIN; i <= CHAR_MAX; i++) {
836 if ( __collate_range_cmp(start, i) <= 0
837 && __collate_range_cmp(i, finish) <= 0
848 - p_b_cclass - parse a character-class name and deal with it
849 == static void p_b_cclass(register struct parse *p, register cset *cs);
853 register struct parse *p;
857 register char *sp = p->next;
858 register struct cclass *cp;
861 while (MORE() && isalpha((uch)PEEK()))
864 for (cp = cclasses; cp->name != NULL; cp++)
865 if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
867 if (cp->name == NULL) {
868 /* oops, didn't find it */
869 SETERROR(REG_ECTYPE);
875 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
880 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
885 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
890 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
895 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
900 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
905 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
910 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
915 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
920 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
925 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
930 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
931 if (isxdigit((uch)c))
936 for (u = cp->multis; *u != '\0'; u += strlen(u) + 1)
942 - p_b_eclass - parse an equivalence-class name and deal with it
943 == static void p_b_eclass(register struct parse *p, register cset *cs);
945 * This implementation is incomplete. xxx
949 register struct parse *p;
954 c = p_b_coll_elem(p, '=');
959 - p_b_symbol - parse a character or [..]ed multicharacter collating symbol
960 == static char p_b_symbol(register struct parse *p);
962 static char /* value of symbol */
964 register struct parse *p;
968 (void)REQUIRE(MORE(), REG_EBRACK);
969 if (!EATTWO('[', '.'))
972 /* collating symbol */
973 value = p_b_coll_elem(p, '.');
974 (void)REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
979 - p_b_coll_elem - parse a collating-element name and look it up
980 == static char p_b_coll_elem(register struct parse *p, int endc);
982 static char /* value of collating element */
983 p_b_coll_elem(p, endc)
984 register struct parse *p;
985 int endc; /* name ended by endc,']' */
987 register char *sp = p->next;
988 register struct cname *cp;
991 while (MORE() && !SEETWO(endc, ']'))
994 SETERROR(REG_EBRACK);
998 for (cp = cnames; cp->name != NULL; cp++)
999 if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
1000 return(cp->code); /* known name */
1002 return(*sp); /* single character */
1003 SETERROR(REG_ECOLLATE); /* neither */
1008 - othercase - return the case counterpart of an alphabetic
1009 == static char othercase(int ch);
1011 static char /* if no counterpart, return ch */
1016 assert(isalpha(ch));
1018 return(tolower(ch));
1019 else if (islower(ch))
1020 return(toupper(ch));
1021 else /* peculiar, but could happen */
1026 - bothcases - emit a dualcase version of a two-case character
1027 == static void bothcases(register struct parse *p, int ch);
1029 * Boy, is this implementation ever a kludge...
1033 register struct parse *p;
1036 register char *oldnext = p->next;
1037 register char *oldend = p->end;
1041 assert(othercase(ch) != ch); /* p_bracket() would recurse */
1048 assert(p->next == bracket+2);
1054 - ordinary - emit an ordinary character
1055 == static void ordinary(register struct parse *p, register int ch);
1059 register struct parse *p;
1062 register cat_t *cap = p->g->categories;
1064 if ((p->g->cflags®_ICASE) && isalpha((uch)ch) && othercase(ch) != ch)
1067 EMIT(OCHAR, (uch)ch);
1069 cap[ch] = p->g->ncategories++;
1074 - nonnewline - emit REG_NEWLINE version of OANY
1075 == static void nonnewline(register struct parse *p);
1077 * Boy, is this implementation ever a kludge...
1081 register struct parse *p;
1083 register char *oldnext = p->next;
1084 register char *oldend = p->end;
1094 assert(p->next == bracket+3);
1100 - repeat - generate code for a bounded repetition, recursively if needed
1101 == static void repeat(register struct parse *p, sopno start, int from, int to);
1104 repeat(p, start, from, to)
1105 register struct parse *p;
1106 sopno start; /* operand from here to end of strip */
1107 int from; /* repeated from this number */
1108 int to; /* to this number of times (maybe INFINITY) */
1110 register sopno finish = HERE();
1113 # define REP(f, t) ((f)*8 + (t))
1114 # define MAP(n) (((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
1115 register sopno copy;
1117 if (p->error != 0) /* head off possible runaway recursion */
1122 switch (REP(MAP(from), MAP(to))) {
1123 case REP(0, 0): /* must be user doing this */
1124 DROP(finish-start); /* drop the operand */
1126 case REP(0, 1): /* as x{1,1}? */
1127 case REP(0, N): /* as x{1,n}? */
1128 case REP(0, INF): /* as x{1,}? */
1129 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
1130 INSERT(OCH_, start); /* offset is wrong... */
1131 repeat(p, start+1, 1, to);
1132 ASTERN(OOR1, start);
1133 AHEAD(start); /* ... fix it */
1136 ASTERN(O_CH, THERETHERE());
1138 case REP(1, 1): /* trivial case */
1141 case REP(1, N): /* as x?x{1,n-1} */
1142 /* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
1143 INSERT(OCH_, start);
1144 ASTERN(OOR1, start);
1146 EMIT(OOR2, 0); /* offset very wrong... */
1147 AHEAD(THERE()); /* ...so fix it */
1148 ASTERN(O_CH, THERETHERE());
1149 copy = dupl(p, start+1, finish+1);
1150 assert(copy == finish+4);
1151 repeat(p, copy, 1, to-1);
1153 case REP(1, INF): /* as x+ */
1154 INSERT(OPLUS_, start);
1155 ASTERN(O_PLUS, start);
1157 case REP(N, N): /* as xx{m-1,n-1} */
1158 copy = dupl(p, start, finish);
1159 repeat(p, copy, from-1, to-1);
1161 case REP(N, INF): /* as xx{n-1,INF} */
1162 copy = dupl(p, start, finish);
1163 repeat(p, copy, from-1, to);
1165 default: /* "can't happen" */
1166 SETERROR(REG_ASSERT); /* just in case */
1172 - seterr - set an error condition
1173 == static int seterr(register struct parse *p, int e);
1175 static int /* useless but makes type checking happy */
1177 register struct parse *p;
1180 if (p->error == 0) /* keep earliest error condition */
1182 p->next = nuls; /* try to bring things to a halt */
1184 return(0); /* make the return value well-defined */
1188 - allocset - allocate a set of characters for []
1189 == static cset *allocset(register struct parse *p);
1193 register struct parse *p;
1195 register int no = p->g->ncsets++;
1197 register size_t nbytes;
1199 register size_t css = (size_t)p->g->csetsize;
1202 if (no >= p->ncsalloc) { /* need another column of space */
1203 p->ncsalloc += CHAR_BIT;
1205 assert(nc % CHAR_BIT == 0);
1206 nbytes = nc / CHAR_BIT * css;
1207 if (p->g->sets == NULL)
1208 p->g->sets = (cset *)malloc(nc * sizeof(cset));
1210 p->g->sets = (cset *)reallocf((char *)p->g->sets,
1212 if (p->g->setbits == NULL)
1213 p->g->setbits = (uch *)malloc(nbytes);
1215 p->g->setbits = (uch *)reallocf((char *)p->g->setbits,
1217 /* xxx this isn't right if setbits is now NULL */
1218 for (i = 0; i < no; i++)
1219 p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
1221 if (p->g->sets != NULL && p->g->setbits != NULL)
1222 (void) memset((char *)p->g->setbits + (nbytes - css),
1226 SETERROR(REG_ESPACE);
1227 /* caller's responsibility not to do set ops */
1231 assert(p->g->sets != NULL); /* xxx */
1232 cs = &p->g->sets[no];
1233 cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
1234 cs->mask = 1 << ((no) % CHAR_BIT);
1243 - freeset - free a now-unused set
1244 == static void freeset(register struct parse *p, register cset *cs);
1248 register struct parse *p;
1252 register cset *top = &p->g->sets[p->g->ncsets];
1253 register size_t css = (size_t)p->g->csetsize;
1255 for (i = 0; i < css; i++)
1257 if (cs == top-1) /* recover only the easy case */
1262 - freezeset - final processing on a set of characters
1263 == static int freezeset(register struct parse *p, register cset *cs);
1265 * The main task here is merging identical sets. This is usually a waste
1266 * of time (although the hash code minimizes the overhead), but can win
1267 * big if REG_ICASE is being used. REG_ICASE, by the way, is why the hash
1268 * is done using addition rather than xor -- all ASCII [aA] sets xor to
1271 static int /* set number */
1273 register struct parse *p;
1276 register short h = cs->hash;
1278 register cset *top = &p->g->sets[p->g->ncsets];
1280 register size_t css = (size_t)p->g->csetsize;
1282 /* look for an earlier one which is the same */
1283 for (cs2 = &p->g->sets[0]; cs2 < top; cs2++)
1284 if (cs2->hash == h && cs2 != cs) {
1286 for (i = 0; i < css; i++)
1287 if (!!CHIN(cs2, i) != !!CHIN(cs, i))
1293 if (cs2 < top) { /* found one */
1298 return((int)(cs - p->g->sets));
1302 - firstch - return first character in a set (which must have at least one)
1303 == static int firstch(register struct parse *p, register cset *cs);
1305 static int /* character; there is no "none" value */
1307 register struct parse *p;
1311 register size_t css = (size_t)p->g->csetsize;
1313 for (i = 0; i < css; i++)
1317 return(0); /* arbitrary */
1321 - nch - number of characters in a set
1322 == static int nch(register struct parse *p, register cset *cs);
1326 register struct parse *p;
1330 register size_t css = (size_t)p->g->csetsize;
1333 for (i = 0; i < css; i++)
1340 - mcadd - add a collating element to a cset
1341 == static void mcadd(register struct parse *p, register cset *cs, \
1342 == register char *cp);
1346 register struct parse *p;
1350 register size_t oldend = cs->smultis;
1352 cs->smultis += strlen(cp) + 1;
1353 if (cs->multis == NULL)
1354 cs->multis = malloc(cs->smultis);
1356 cs->multis = reallocf(cs->multis, cs->smultis);
1357 if (cs->multis == NULL) {
1358 SETERROR(REG_ESPACE);
1362 (void) strcpy(cs->multis + oldend - 1, cp);
1363 cs->multis[cs->smultis - 1] = '\0';
1368 - mcsub - subtract a collating element from a cset
1369 == static void mcsub(register cset *cs, register char *cp);
1376 register char *fp = mcfind(cs, cp);
1377 register size_t len = strlen(fp);
1380 (void) memmove(fp, fp + len + 1,
1381 cs->smultis - (fp + len + 1 - cs->multis));
1384 if (cs->smultis == 0) {
1390 cs->multis = reallocf(cs->multis, cs->smultis);
1391 assert(cs->multis != NULL);
1395 - mcin - is a collating element in a cset?
1396 == static int mcin(register cset *cs, register char *cp);
1403 return(mcfind(cs, cp) != NULL);
1407 - mcfind - find a collating element in a cset
1408 == static char *mcfind(register cset *cs, register char *cp);
1417 if (cs->multis == NULL)
1419 for (p = cs->multis; *p != '\0'; p += strlen(p) + 1)
1420 if (strcmp(cp, p) == 0)
1427 - mcinvert - invert the list of collating elements in a cset
1428 == static void mcinvert(register struct parse *p, register cset *cs);
1430 * This would have to know the set of possibilities. Implementation
1435 register struct parse *p;
1438 assert(cs->multis == NULL); /* xxx */
1442 - mccase - add case counterparts of the list of collating elements in a cset
1443 == static void mccase(register struct parse *p, register cset *cs);
1445 * This would have to know the set of possibilities. Implementation
1450 register struct parse *p;
1453 assert(cs->multis == NULL); /* xxx */
1457 - isinsets - is this character in any sets?
1458 == static int isinsets(register struct re_guts *g, int c);
1460 static int /* predicate */
1462 register struct re_guts *g;
1467 register int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1468 register unsigned uc = (uch)c;
1470 for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1477 - samesets - are these two characters in exactly the same sets?
1478 == static int samesets(register struct re_guts *g, int c1, int c2);
1480 static int /* predicate */
1482 register struct re_guts *g;
1488 register int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
1489 register unsigned uc1 = (uch)c1;
1490 register unsigned uc2 = (uch)c2;
1492 for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
1493 if (col[uc1] != col[uc2])
1499 - categorize - sort out character categories
1500 == static void categorize(struct parse *p, register struct re_guts *g);
1505 register struct re_guts *g;
1507 register cat_t *cats = g->categories;
1512 /* avoid making error situations worse */
1516 for (c = CHAR_MIN; c <= CHAR_MAX; c++)
1517 if (cats[c] == 0 && isinsets(g, c)) {
1518 cat = g->ncategories++;
1520 for (c2 = c+1; c2 <= CHAR_MAX; c2++)
1521 if (cats[c2] == 0 && samesets(g, c, c2))
1527 - dupl - emit a duplicate of a bunch of sops
1528 == static sopno dupl(register struct parse *p, sopno start, sopno finish);
1530 static sopno /* start of duplicate */
1531 dupl(p, start, finish)
1532 register struct parse *p;
1533 sopno start; /* from here */
1534 sopno finish; /* to this less one */
1536 register sopno ret = HERE();
1537 register sopno len = finish - start;
1539 assert(finish >= start);
1542 enlarge(p, p->ssize + len); /* this many unexpected additions */
1543 assert(p->ssize >= p->slen + len);
1544 (void) memcpy((char *)(p->strip + p->slen),
1545 (char *)(p->strip + start), (size_t)len*sizeof(sop));
1551 - doemit - emit a strip operator
1552 == static void doemit(register struct parse *p, sop op, size_t opnd);
1554 * It might seem better to implement this as a macro with a function as
1555 * hard-case backup, but it's just too big and messy unless there are
1556 * some changes to the data structures. Maybe later.
1560 register struct parse *p;
1564 /* avoid making error situations worse */
1568 /* deal with oversize operands ("can't happen", more or less) */
1569 assert(opnd < 1<<OPSHIFT);
1571 /* deal with undersized strip */
1572 if (p->slen >= p->ssize)
1573 enlarge(p, (p->ssize+1) / 2 * 3); /* +50% */
1574 assert(p->slen < p->ssize);
1576 /* finally, it's all reduced to the easy case */
1577 p->strip[p->slen++] = SOP(op, opnd);
1581 - doinsert - insert a sop into the strip
1582 == static void doinsert(register struct parse *p, sop op, size_t opnd, sopno pos);
1585 doinsert(p, op, opnd, pos)
1586 register struct parse *p;
1595 /* avoid making error situations worse */
1600 EMIT(op, opnd); /* do checks, ensure space */
1601 assert(HERE() == sn+1);
1604 /* adjust paren pointers */
1606 for (i = 1; i < NPAREN; i++) {
1607 if (p->pbegin[i] >= pos) {
1610 if (p->pend[i] >= pos) {
1615 memmove((char *)&p->strip[pos+1], (char *)&p->strip[pos],
1616 (HERE()-pos-1)*sizeof(sop));
1621 - dofwd - complete a forward reference
1622 == static void dofwd(register struct parse *p, sopno pos, sop value);
1625 dofwd(p, pos, value)
1626 register struct parse *p;
1630 /* avoid making error situations worse */
1634 assert(value < 1<<OPSHIFT);
1635 p->strip[pos] = OP(p->strip[pos]) | value;
1639 - enlarge - enlarge the strip
1640 == static void enlarge(register struct parse *p, sopno size);
1644 register struct parse *p;
1645 register sopno size;
1649 if (p->ssize >= size)
1652 sp = (sop *)realloc(p->strip, size*sizeof(sop));
1654 SETERROR(REG_ESPACE);
1662 - stripsnug - compact the strip
1663 == static void stripsnug(register struct parse *p, register struct re_guts *g);
1667 register struct parse *p;
1668 register struct re_guts *g;
1670 g->nstates = p->slen;
1671 g->strip = (sop *)realloc((char *)p->strip, p->slen * sizeof(sop));
1672 if (g->strip == NULL) {
1673 SETERROR(REG_ESPACE);
1674 g->strip = p->strip;
1679 - findmust - fill in must and mlen with longest mandatory literal string
1680 == static void findmust(register struct parse *p, register struct re_guts *g);
1682 * This algorithm could do fancy things like analyzing the operands of |
1683 * for common subsequences. Someday. This code is simple and finds most
1684 * of the interesting cases.
1686 * Note that must and mlen got initialized during setup.
1691 register struct re_guts *g;
1695 register sop *newstart;
1696 register sopno newlen;
1703 /* avoid making error situations worse */
1707 /* Find out if we can handle OANYOF or not */
1709 for (cs = 0; cs < g->ncsets; cs++)
1710 if (g->sets[cs].multis != NULL)
1713 /* find the longest OCHAR sequence in strip */
1717 scan = g->strip + 1;
1721 case OCHAR: /* sequence member */
1722 if (newlen == 0) /* new sequence */
1723 newstart = scan - 1;
1726 case OPLUS_: /* things that don't break one */
1730 case OQUEST_: /* things that must be skipped */
1732 offset = altoffset(scan, offset, mccs);
1737 /* assert() interferes w debug printouts */
1738 if (OP(s) != O_QUEST && OP(s) != O_CH &&
1743 } while (OP(s) != O_QUEST && OP(s) != O_CH);
1745 case OBOW: /* things that break a sequence */
1752 if (newlen > g->mlen) { /* ends one */
1756 g->moffset += offset;
1759 g->moffset = offset;
1767 if (newlen > g->mlen) { /* ends one */
1771 g->moffset += offset;
1774 g->moffset = offset;
1783 case OANYOF: /* may or may not invalidate offset */
1784 /* First, everything as OANY */
1785 if (newlen > g->mlen) { /* ends one */
1789 g->moffset += offset;
1792 g->moffset = offset;
1800 /* And, now, if we found out we can't deal with
1801 * it, make offset = -1.
1807 /* Anything here makes it impossible or too hard
1808 * to calculate the offset -- so we give up;
1809 * save the last known good offset, in case the
1810 * must sequence doesn't occur later.
1812 if (newlen > g->mlen) { /* ends one */
1816 g->moffset += offset;
1818 g->moffset = offset;
1824 } while (OP(s) != OEND);
1826 if (g->mlen == 0) { /* there isn't one */
1831 /* turn it into a character string */
1832 g->must = malloc((size_t)g->mlen + 1);
1833 if (g->must == NULL) { /* argh; just forget it */
1840 for (i = g->mlen; i > 0; i--) {
1841 while (OP(s = *scan++) != OCHAR)
1843 assert(cp < g->must + g->mlen);
1844 *cp++ = (char)OPND(s);
1846 assert(cp == g->must + g->mlen);
1847 *cp++ = '\0'; /* just on general principles */
1851 - altoffset - choose biggest offset among multiple choices
1852 == static int altoffset(sop *scan, int offset, int mccs);
1854 * Compute, recursively if necessary, the largest offset among multiple
1858 altoffset(scan, offset, mccs)
1867 /* If we gave up already on offsets, return */
1874 while (OP(s) != O_QUEST && OP(s) != O_CH) {
1883 try = altoffset(scan, try, mccs);
1890 if (OP(s) != O_QUEST && OP(s) != O_CH &&
1893 } while (OP(s) != O_QUEST && OP(s) != O_CH);
1894 /* We must skip to the next position, or we'll
1895 * leave altoffset() too early.
1923 return largest+offset;
1927 - computejumps - compute char jumps for BM scan
1928 == static void computejumps(register struct parse *p, register struct re_guts *g);
1930 * This algorithm assumes g->must exists and is has size greater than
1931 * zero. It's based on the algorithm found on Computer Algorithms by
1934 * A char jump is the number of characters one needs to jump based on
1935 * the value of the character from the text that was mismatched.
1945 /* Avoid making errors worse */
1949 g->charjump = (int*) malloc((NC + 1) * sizeof(int));
1950 if (g->charjump == NULL) /* Not a fatal error */
1952 /* Adjust for signed chars, if necessary */
1953 g->charjump = &g->charjump[-(CHAR_MIN)];
1955 /* If the character does not exist in the pattern, the jump
1956 * is equal to the number of characters in the pattern.
1958 for (ch = CHAR_MIN; ch < (CHAR_MAX + 1); ch++)
1959 g->charjump[ch] = g->mlen;
1961 /* If the character does exist, compute the jump that would
1962 * take us to the last character in the pattern equal to it
1963 * (notice that we match right to left, so that last character
1964 * is the first one that would be matched).
1966 for (mindex = 0; mindex < g->mlen; mindex++)
1967 g->charjump[g->must[mindex]] = g->mlen - mindex - 1;
1971 - computematchjumps - compute match jumps for BM scan
1972 == static void computematchjumps(register struct parse *p, register struct re_guts *g);
1974 * This algorithm assumes g->must exists and is has size greater than
1975 * zero. It's based on the algorithm found on Computer Algorithms by
1978 * A match jump is the number of characters one needs to advance based
1979 * on the already-matched suffix.
1980 * Notice that all values here are minus (g->mlen-1), because of the way
1981 * the search algorithm works.
1984 computematchjumps(p, g)
1988 int mindex; /* General "must" iterator */
1989 int suffix; /* Keeps track of matching suffix */
1990 int ssuffix; /* Keeps track of suffixes' suffix */
1991 int* pmatches; /* pmatches[k] points to the next i
1992 * such that i+1...mlen is a substring
1993 * of k+1...k+mlen-i-1
1996 /* Avoid making errors worse */
2000 pmatches = (int*) malloc(g->mlen * sizeof(unsigned int));
2001 if (pmatches == NULL) {
2002 g->matchjump = NULL;
2006 g->matchjump = (int*) malloc(g->mlen * sizeof(unsigned int));
2007 if (g->matchjump == NULL) /* Not a fatal error */
2010 /* Set maximum possible jump for each character in the pattern */
2011 for (mindex = 0; mindex < g->mlen; mindex++)
2012 g->matchjump[mindex] = 2*g->mlen - mindex - 1;
2014 /* Compute pmatches[] */
2015 for (mindex = g->mlen - 1, suffix = g->mlen; mindex >= 0;
2016 mindex--, suffix--) {
2017 pmatches[mindex] = suffix;
2019 /* If a mismatch is found, interrupting the substring,
2020 * compute the matchjump for that position. If no
2021 * mismatch is found, then a text substring mismatched
2022 * against the suffix will also mismatch against the
2025 while (suffix < g->mlen
2026 && g->must[mindex] != g->must[suffix]) {
2027 g->matchjump[suffix] = MIN(g->matchjump[suffix],
2028 g->mlen - mindex - 1);
2029 suffix = pmatches[suffix];
2033 /* Compute the matchjump up to the last substring found to jump
2034 * to the beginning of the largest must pattern prefix matching
2037 for (mindex = 0; mindex <= suffix; mindex++)
2038 g->matchjump[mindex] = MIN(g->matchjump[mindex],
2039 g->mlen + suffix - mindex);
2041 ssuffix = pmatches[suffix];
2042 while (suffix < g->mlen) {
2043 while (suffix <= ssuffix && suffix < g->mlen) {
2044 g->matchjump[suffix] = MIN(g->matchjump[suffix],
2045 g->mlen + ssuffix - suffix);
2048 ssuffix = pmatches[ssuffix];
2055 - pluscount - count + nesting
2056 == static sopno pluscount(register struct parse *p, register struct re_guts *g);
2058 static sopno /* nesting depth */
2061 register struct re_guts *g;
2065 register sopno plusnest = 0;
2066 register sopno maxnest = 0;
2069 return(0); /* there may not be an OEND */
2071 scan = g->strip + 1;
2079 if (plusnest > maxnest)
2084 } while (OP(s) != OEND);