Sync usage with man page.
[netbsd-mini2440.git] / lib / libcompat / regexp / regexp.c
blob174f1d2d869a528938954b499cc10f68fa3d1b27
1 /*
2 * regcomp and regexec -- regsub and regerror are elsewhere
4 * Copyright (c) 1986 by University of Toronto.
5 * Written by Henry Spencer. Not derived from licensed software.
7 * Permission is granted to anyone to use this software for any
8 * purpose on any computer system, and to redistribute it freely,
9 * subject to the following restrictions:
11 * 1. The author is not responsible for the consequences of use of
12 * this software, no matter how awful, even if they arise
13 * from defects in it.
15 * 2. The origin of this software must not be misrepresented, either
16 * by explicit claim or by omission.
18 * 3. Altered versions must be plainly marked as such, and must not
19 * be misrepresented as being the original software.
20 *** THIS IS AN ALTERED VERSION. It was altered by John Gilmore,
21 *** hoptoad!gnu, on 27 Dec 1986, to add \n as an alternative to |
22 *** to assist in implementing egrep.
23 *** THIS IS AN ALTERED VERSION. It was altered by John Gilmore,
24 *** hoptoad!gnu, on 27 Dec 1986, to add \< and \> for word-matching
25 *** as in BSD grep and ex.
26 *** THIS IS AN ALTERED VERSION. It was altered by John Gilmore,
27 *** hoptoad!gnu, on 28 Dec 1986, to optimize characters quoted with \.
28 *** THIS IS AN ALTERED VERSION. It was altered by James A. Woods,
29 *** ames!jaw, on 19 June 1987, to quash a regcomp() redundancy.
31 * Beware that some of this code is subtly aware of the way operator
32 * precedence is structured in regular expressions. Serious changes in
33 * regular-expression syntax might require a total rethink.
36 #include <sys/cdefs.h>
37 #if defined(LIBC_SCCS) && !defined(lint)
38 __RCSID("$NetBSD: regexp.c,v 1.17 2006/04/08 22:05:36 christos Exp $");
39 #endif /* LIBC_SCCS and not lint */
41 #include <ctype.h>
42 #include <regexp.h>
43 #include <stdio.h>
44 #include <stdlib.h>
45 #include <string.h>
46 #include "regmagic.h"
49 * The "internal use only" fields in regexp.h are present to pass info from
50 * compile to execute that permits the execute phase to run lots faster on
51 * simple cases. They are:
53 * regstart char that must begin a match; '\0' if none obvious
54 * reganch is the match anchored (at beginning-of-line only)?
55 * regmust string (pointer into program) that match must include, or NULL
56 * regmlen length of regmust string
58 * Regstart and reganch permit very fast decisions on suitable starting points
59 * for a match, cutting down the work a lot. Regmust permits fast rejection
60 * of lines that cannot possibly match. The regmust tests are costly enough
61 * that regcomp() supplies a regmust only if the r.e. contains something
62 * potentially expensive (at present, the only such thing detected is * or +
63 * at the start of the r.e., which can involve a lot of backup). Regmlen is
64 * supplied because the test in regexec() needs it and regcomp() is computing
65 * it anyway.
69 * Structure for regexp "program". This is essentially a linear encoding
70 * of a nondeterministic finite-state machine (aka syntax charts or
71 * "railroad normal form" in parsing technology). Each node is an opcode
72 * plus a "next" pointer, possibly plus an operand. "Next" pointers of
73 * all nodes except BRANCH implement concatenation; a "next" pointer with
74 * a BRANCH on both ends of it is connecting two alternatives. (Here we
75 * have one of the subtle syntax dependencies: an individual BRANCH (as
76 * opposed to a collection of them) is never concatenated with anything
77 * because of operator precedence.) The operand of some types of node is
78 * a literal string; for others, it is a node leading into a sub-FSM. In
79 * particular, the operand of a BRANCH node is the first node of the branch.
80 * (NB this is *not* a tree structure: the tail of the branch connects
81 * to the thing following the set of BRANCHes.) The opcodes are:
84 /* definition number opnd? meaning */
85 #define END 0 /* no End of program. */
86 #define BOL 1 /* no Match "" at beginning of line. */
87 #define EOL 2 /* no Match "" at end of line. */
88 #define ANY 3 /* no Match any one character. */
89 #define ANYOF 4 /* str Match any character in this string. */
90 #define ANYBUT 5 /* str Match any character not in this string. */
91 #define BRANCH 6 /* node Match this alternative, or the next... */
92 #define BACK 7 /* no Match "", "next" ptr points backward. */
93 #define EXACTLY 8 /* str Match this string. */
94 #define NOTHING 9 /* no Match empty string. */
95 #define STAR 10 /* node Match this (simple) thing 0 or more times. */
96 #define PLUS 11 /* node Match this (simple) thing 1 or more times. */
97 #define WORDA 12 /* no Match "" at wordchar, where prev is nonword */
98 #define WORDZ 13 /* no Match "" at nonwordchar, where prev is word */
99 #define OPEN 20 /* no Mark this point in input as start of #n. */
100 /* OPEN+1 is number 1, etc. */
101 #define CLOSE 30 /* no Analogous to OPEN. */
104 * Opcode notes:
106 * BRANCH The set of branches constituting a single choice are hooked
107 * together with their "next" pointers, since precedence prevents
108 * anything being concatenated to any individual branch. The
109 * "next" pointer of the last BRANCH in a choice points to the
110 * thing following the whole choice. This is also where the
111 * final "next" pointer of each individual branch points; each
112 * branch starts with the operand node of a BRANCH node.
114 * BACK Normal "next" pointers all implicitly point forward; BACK
115 * exists to make loop structures possible.
117 * STAR,PLUS '?', and complex '*' and '+', are implemented as circular
118 * BRANCH structures using BACK. Simple cases (one character
119 * per match) are implemented with STAR and PLUS for speed
120 * and to minimize recursive plunges.
122 * OPEN,CLOSE ...are numbered at compile time.
126 * A node is one char of opcode followed by two chars of "next" pointer.
127 * "Next" pointers are stored as two 8-bit pieces, high order first. The
128 * value is a positive offset from the opcode of the node containing it.
129 * An operand, if any, simply follows the node. (Note that much of the
130 * code generation knows about this implicit relationship.)
132 * Using two bytes for the "next" pointer is vast overkill for most things,
133 * but allows patterns to get big without disasters.
135 #define OP(p) (*(p))
136 #define NEXT(p) (((*((p)+1)&0377)<<8) + (*((p)+2)&0377))
137 #define OPERAND(p) ((p) + 3)
140 * See regmagic.h for one further detail of program structure.
145 * Utility definitions.
147 #ifndef CHARBITS
148 #define UCHARAT(p) ((int)*(unsigned char *)(p))
149 #else
150 #define UCHARAT(p) ((int)*(p)&CHARBITS)
151 #endif
153 #define FAIL(m) { regerror(m); return(NULL); }
154 #define ISMULT(c) ((c) == '*' || (c) == '+' || (c) == '?')
157 * Flags to be passed up and down.
159 #define HASWIDTH 01 /* Known never to match null string. */
160 #define SIMPLE 02 /* Simple enough to be STAR/PLUS operand. */
161 #define SPSTART 04 /* Starts with * or +. */
162 #define WORST 0 /* Worst case. */
165 * Global work variables for regcomp().
167 static char *regparse; /* Input-scan pointer. */
168 static int regnpar; /* () count. */
169 static char regdummy;
170 static char *regcode; /* Code-emit pointer; &regdummy = don't. */
171 static long regsize; /* Code size. */
174 * Forward declarations for regcomp()'s friends.
176 #ifndef STATIC
177 #define STATIC static
178 #endif
179 STATIC char *reg __P((int, int *));
180 STATIC char *regbranch __P((int *));
181 STATIC char *regpiece __P((int *));
182 STATIC char *regatom __P((int *));
183 STATIC char *regnode __P((int));
184 STATIC char *regnext __P((char *));
185 STATIC void regc __P((int));
186 STATIC void reginsert __P((int, char *));
187 STATIC void regtail __P((char *, char *));
188 STATIC void regoptail __P((char *, char *));
189 #ifdef STRCSPN
190 STATIC int strcspn __P((char *, char *));
191 #endif
194 - regcomp - compile a regular expression into internal code
196 * We can't allocate space until we know how big the compiled form will be,
197 * but we can't compile it (and thus know how big it is) until we've got a
198 * place to put the code. So we cheat: we compile it twice, once with code
199 * generation turned off and size counting turned on, and once "for real".
200 * This also means that we don't allocate space until we are sure that the
201 * thing really will compile successfully, and we never have to move the
202 * code and thus invalidate pointers into it. (Note that it has to be in
203 * one piece because free() must be able to free it all.)
205 * Beware that the optimization-preparation code in here knows about some
206 * of the structure of the compiled regexp.
208 regexp *
209 __compat_regcomp(expn)
210 const char *expn;
212 regexp *r;
213 char *scan;
214 char *longest;
215 int len;
216 int flags;
218 if (expn == NULL)
219 FAIL("NULL argument");
221 /* First pass: determine size, legality. */
222 #ifdef notdef
223 if (expn[0] == '.' && expn[1] == '*') expn += 2; /* aid grep */
224 #endif
225 /* LINTED const castaway */
226 regparse = (char *)expn;
227 regnpar = 1;
228 regsize = 0L;
229 regcode = &regdummy;
230 regc(MAGIC);
231 if (reg(0, &flags) == NULL)
232 return(NULL);
234 /* Small enough for pointer-storage convention? */
235 if (regsize >= 32767L) /* Probably could be 65535L. */
236 FAIL("regexp too big");
238 /* Allocate space. */
239 r = (regexp *)malloc(sizeof(regexp) + (unsigned)regsize);
240 if (r == NULL)
241 FAIL("out of space");
243 /* Second pass: emit code. */
244 /* LINTED const castaway */
245 regparse = (char *)expn;
246 regnpar = 1;
247 regcode = r->program;
248 regc(MAGIC);
249 if (reg(0, &flags) == NULL)
250 return(NULL);
252 /* Dig out information for optimizations. */
253 r->regstart = '\0'; /* Worst-case defaults. */
254 r->reganch = 0;
255 r->regmust = NULL;
256 r->regmlen = 0;
257 scan = r->program+1; /* First BRANCH. */
258 if (OP(regnext(scan)) == END) { /* Only one top-level choice. */
259 scan = OPERAND(scan);
261 /* Starting-point info. */
262 if (OP(scan) == EXACTLY)
263 r->regstart = *OPERAND(scan);
264 else if (OP(scan) == BOL)
265 r->reganch++;
268 * If there's something expensive in the r.e., find the
269 * longest literal string that must appear and make it the
270 * regmust. Resolve ties in favor of later strings, since
271 * the regstart check works with the beginning of the r.e.
272 * and avoiding duplication strengthens checking. Not a
273 * strong reason, but sufficient in the absence of others.
275 if (flags&SPSTART) {
276 longest = NULL;
277 len = 0;
278 for (; scan != NULL; scan = regnext(scan))
279 if (OP(scan) == EXACTLY && (int) strlen(OPERAND(scan)) >= len) {
280 longest = OPERAND(scan);
281 len = strlen(OPERAND(scan));
283 r->regmust = longest;
284 r->regmlen = len;
288 return(r);
292 - reg - regular expression, i.e. main body or parenthesized thing
294 * Caller must absorb opening parenthesis.
296 * Combining parenthesis handling with the base level of regular expression
297 * is a trifle forced, but the need to tie the tails of the branches to what
298 * follows makes it hard to avoid.
300 static char *
301 reg(paren, flagp)
302 int paren; /* Parenthesized? */
303 int *flagp;
305 char *ret;
306 char *br;
307 char *ender;
308 int parno = 0;
309 int flags;
311 *flagp = HASWIDTH; /* Tentatively. */
313 /* Make an OPEN node, if parenthesized. */
314 if (paren) {
315 if (regnpar >= NSUBEXP)
316 FAIL("too many ()");
317 parno = regnpar;
318 regnpar++;
319 ret = regnode(OPEN+parno);
320 } else
321 ret = NULL;
323 /* Pick up the branches, linking them together. */
324 br = regbranch(&flags);
325 if (br == NULL)
326 return(NULL);
327 if (ret != NULL)
328 regtail(ret, br); /* OPEN -> first. */
329 else
330 ret = br;
331 if (!(flags&HASWIDTH))
332 *flagp &= ~HASWIDTH;
333 *flagp |= flags&SPSTART;
334 while (*regparse == '|' || *regparse == '\n') {
335 regparse++;
336 br = regbranch(&flags);
337 if (br == NULL)
338 return(NULL);
339 regtail(ret, br); /* BRANCH -> BRANCH. */
340 if (!(flags&HASWIDTH))
341 *flagp &= ~HASWIDTH;
342 *flagp |= flags&SPSTART;
345 /* Make a closing node, and hook it on the end. */
346 ender = regnode((paren) ? CLOSE+parno : END);
347 regtail(ret, ender);
349 /* Hook the tails of the branches to the closing node. */
350 for (br = ret; br != NULL; br = regnext(br))
351 regoptail(br, ender);
353 /* Check for proper termination. */
354 if (paren && *regparse++ != ')') {
355 FAIL("unmatched ()");
356 } else if (!paren && *regparse != '\0') {
357 if (*regparse == ')') {
358 FAIL("unmatched ()");
359 } else
360 FAIL("junk on end"); /* "Can't happen". */
361 /* NOTREACHED */
364 return(ret);
368 - regbranch - one alternative of an | operator
370 * Implements the concatenation operator.
372 static char *
373 regbranch(flagp)
374 int *flagp;
376 char *ret;
377 char *chain;
378 char *latest;
379 int flags;
381 *flagp = WORST; /* Tentatively. */
383 ret = regnode(BRANCH);
384 chain = NULL;
385 while (*regparse != '\0' && *regparse != ')' &&
386 *regparse != '\n' && *regparse != '|') {
387 latest = regpiece(&flags);
388 if (latest == NULL)
389 return(NULL);
390 *flagp |= flags&HASWIDTH;
391 if (chain == NULL) /* First piece. */
392 *flagp |= flags&SPSTART;
393 else
394 regtail(chain, latest);
395 chain = latest;
397 if (chain == NULL) /* Loop ran zero times. */
398 (void) regnode(NOTHING);
400 return(ret);
404 - regpiece - something followed by possible [*+?]
406 * Note that the branching code sequences used for ? and the general cases
407 * of * and + are somewhat optimized: they use the same NOTHING node as
408 * both the endmarker for their branch list and the body of the last branch.
409 * It might seem that this node could be dispensed with entirely, but the
410 * endmarker role is not redundant.
412 static char *
413 regpiece(flagp)
414 int *flagp;
416 char *ret;
417 char op;
418 char *next;
419 int flags;
421 ret = regatom(&flags);
422 if (ret == NULL)
423 return(NULL);
425 op = *regparse;
426 if (!ISMULT(op)) {
427 *flagp = flags;
428 return(ret);
431 if (!(flags&HASWIDTH) && op != '?')
432 FAIL("*+ operand could be empty");
433 *flagp = (op != '+') ? (WORST|SPSTART) : (WORST|HASWIDTH);
435 if (op == '*' && (flags&SIMPLE))
436 reginsert(STAR, ret);
437 else if (op == '*') {
438 /* Emit x* as (x&|), where & means "self". */
439 reginsert(BRANCH, ret); /* Either x */
440 regoptail(ret, regnode(BACK)); /* and loop */
441 regoptail(ret, ret); /* back */
442 regtail(ret, regnode(BRANCH)); /* or */
443 regtail(ret, regnode(NOTHING)); /* null. */
444 } else if (op == '+' && (flags&SIMPLE))
445 reginsert(PLUS, ret);
446 else if (op == '+') {
447 /* Emit x+ as x(&|), where & means "self". */
448 next = regnode(BRANCH); /* Either */
449 regtail(ret, next);
450 regtail(regnode(BACK), ret); /* loop back */
451 regtail(next, regnode(BRANCH)); /* or */
452 regtail(ret, regnode(NOTHING)); /* null. */
453 } else if (op == '?') {
454 /* Emit x? as (x|) */
455 reginsert(BRANCH, ret); /* Either x */
456 regtail(ret, regnode(BRANCH)); /* or */
457 next = regnode(NOTHING); /* null. */
458 regtail(ret, next);
459 regoptail(ret, next);
461 regparse++;
462 if (ISMULT(*regparse))
463 FAIL("nested *?+");
465 return(ret);
469 - regatom - the lowest level
471 * Optimization: gobbles an entire sequence of ordinary characters so that
472 * it can turn them into a single node, which is smaller to store and
473 * faster to run. Backslashed characters are exceptions, each becoming a
474 * separate node; the code is simpler that way and it's not worth fixing.
476 static char *
477 regatom(flagp)
478 int *flagp;
480 char *ret;
481 int flags;
483 *flagp = WORST; /* Tentatively. */
485 switch (*regparse++) {
486 /* FIXME: these chars only have meaning at beg/end of pat? */
487 case '^':
488 ret = regnode(BOL);
489 break;
490 case '$':
491 ret = regnode(EOL);
492 break;
493 case '.':
494 ret = regnode(ANY);
495 *flagp |= HASWIDTH|SIMPLE;
496 break;
497 case '[': {
498 int class;
499 int classend;
501 if (*regparse == '^') { /* Complement of range. */
502 ret = regnode(ANYBUT);
503 regparse++;
504 } else
505 ret = regnode(ANYOF);
506 if (*regparse == ']' || *regparse == '-')
507 regc(*regparse++);
508 while (*regparse != '\0' && *regparse != ']') {
509 if (*regparse == '-') {
510 regparse++;
511 if (*regparse == ']' || *regparse == '\0')
512 regc('-');
513 else {
514 class = UCHARAT(regparse-2)+1;
515 classend = UCHARAT(regparse);
516 if (class > classend+1)
517 FAIL("invalid [] range");
518 for (; class <= classend; class++)
519 regc(class);
520 regparse++;
522 } else
523 regc(*regparse++);
525 regc('\0');
526 if (*regparse != ']')
527 FAIL("unmatched []");
528 regparse++;
529 *flagp |= HASWIDTH|SIMPLE;
531 break;
532 case '(':
533 ret = reg(1, &flags);
534 if (ret == NULL)
535 return(NULL);
536 *flagp |= flags&(HASWIDTH|SPSTART);
537 break;
538 case '\0':
539 case '|':
540 case '\n':
541 case ')':
542 FAIL("internal urp"); /* Supposed to be caught earlier. */
543 case '?':
544 case '+':
545 case '*':
546 FAIL("?+* follows nothing");
547 case '\\':
548 switch (*regparse++) {
549 case '\0':
550 FAIL("trailing \\");
551 case '<':
552 ret = regnode(WORDA);
553 break;
554 case '>':
555 ret = regnode(WORDZ);
556 break;
557 /* FIXME: Someday handle \1, \2, ... */
558 default:
559 /* Handle general quoted chars in exact-match routine */
560 goto de_fault;
562 break;
563 de_fault:
564 /*FALLTHROUGH*/
565 default:
567 * Encode a string of characters to be matched exactly.
569 * This is a bit tricky due to quoted chars and due to
570 * '*', '+', and '?' taking the SINGLE char previous
571 * as their operand.
573 * On entry, the char at regparse[-1] is going to go
574 * into the string, no matter what it is. (It could be
575 * following a \ if we are entered from the '\' case.)
577 * Basic idea is to pick up a good char in ch and
578 * examine the next char. If it's *+? then we twiddle.
579 * If it's \ then we frozzle. If it's other magic char
580 * we push ch and terminate the string. If none of the
581 * above, we push ch on the string and go around again.
583 * regprev is used to remember where "the current char"
584 * starts in the string, if due to a *+? we need to back
585 * up and put the current char in a separate, 1-char, string.
586 * When regprev is NULL, ch is the only char in the
587 * string; this is used in *+? handling, and in setting
588 * flags |= SIMPLE at the end.
591 char *regprev;
592 char ch;
594 regparse--; /* Look at cur char */
595 ret = regnode(EXACTLY);
596 for ( regprev = 0 ; ; ) {
597 ch = *regparse++; /* Get current char */
598 switch (*regparse) { /* look at next one */
600 default:
601 regc(ch); /* Add cur to string */
602 break;
604 case '.': case '[': case '(':
605 case ')': case '|': case '\n':
606 case '$': case '^':
607 case '\0':
608 /* FIXME, $ and ^ should not always be magic */
609 magic:
610 regc(ch); /* dump cur char */
611 goto done; /* and we are done */
613 case '?': case '+': case '*':
614 if (!regprev) /* If just ch in str, */
615 goto magic; /* use it */
616 /* End mult-char string one early */
617 regparse = regprev; /* Back up parse */
618 goto done;
620 case '\\':
621 regc(ch); /* Cur char OK */
622 switch (regparse[1]){ /* Look after \ */
623 case '\0':
624 case '<':
625 case '>':
626 /* FIXME: Someday handle \1, \2, ... */
627 goto done; /* Not quoted */
628 default:
629 /* Backup point is \, scan * point is after it. */
630 regprev = regparse;
631 regparse++;
632 continue; /* NOT break; */
635 regprev = regparse; /* Set backup point */
637 done:
638 regc('\0');
639 *flagp |= HASWIDTH;
640 if (!regprev) /* One char? */
641 *flagp |= SIMPLE;
643 break;
646 return(ret);
650 - regnode - emit a node
652 static char * /* Location. */
653 regnode(op)
654 int op;
656 char *ret;
657 char *ptr;
659 ret = regcode;
660 if (ret == &regdummy) {
661 regsize += 3;
662 return(ret);
665 ptr = ret;
666 *ptr++ = op;
667 *ptr++ = '\0'; /* Null "next" pointer. */
668 *ptr++ = '\0';
669 regcode = ptr;
671 return(ret);
675 - regc - emit (if appropriate) a byte of code
677 static void
678 regc(b)
679 int b;
681 if (regcode != &regdummy)
682 *regcode++ = b;
683 else
684 regsize++;
688 - reginsert - insert an operator in front of already-emitted operand
690 * Means relocating the operand.
692 static void
693 reginsert(op, opnd)
694 int op;
695 char *opnd;
697 char *src;
698 char *dst;
699 char *place;
701 if (regcode == &regdummy) {
702 regsize += 3;
703 return;
706 src = regcode;
707 regcode += 3;
708 dst = regcode;
709 while (src > opnd)
710 *--dst = *--src;
712 place = opnd; /* Op node, where operand used to be. */
713 *place++ = op;
714 *place++ = '\0';
715 *place++ = '\0';
719 - regtail - set the next-pointer at the end of a node chain
721 static void
722 regtail(p, val)
723 char *p;
724 char *val;
726 char *scan;
727 char *temp;
728 int offset;
730 if (p == &regdummy)
731 return;
733 /* Find last node. */
734 scan = p;
735 for (;;) {
736 temp = regnext(scan);
737 if (temp == NULL)
738 break;
739 scan = temp;
742 if (OP(scan) == BACK)
743 offset = scan - val;
744 else
745 offset = val - scan;
746 *(scan+1) = ((unsigned int)offset>>8)&0377;
747 *(scan+2) = offset&0377;
751 - regoptail - regtail on operand of first argument; nop if operandless
753 static void
754 regoptail(p, val)
755 char *p;
756 char *val;
758 /* "Operandless" and "op != BRANCH" are synonymous in practice. */
759 if (p == NULL || p == &regdummy || OP(p) != BRANCH)
760 return;
761 regtail(OPERAND(p), val);
765 * regexec and friends
769 * Global work variables for regexec().
771 static char *reginput; /* String-input pointer. */
772 static char *regbol; /* Beginning of input, for ^ check. */
773 static char **regstartp; /* Pointer to startp array. */
774 static char **regendp; /* Ditto for endp. */
777 * Forwards.
779 STATIC int regtry __P((const regexp *, const char *));
780 STATIC int regmatch __P((char *));
781 STATIC int regrepeat __P((char *));
783 #ifdef DEBUG
784 int regnarrate = 0;
785 void regdump __P((regexp *));
786 STATIC char *regprop __P((char *));
787 #endif
790 - regexec - match a regexp against a string
793 __compat_regexec(prog, string)
794 const regexp *prog;
795 const char *string;
797 char *s;
799 /* Be paranoid... */
800 if (prog == NULL || string == NULL) {
801 regerror("NULL parameter");
802 return(0);
805 /* Check validity of program. */
806 if (UCHARAT(prog->program) != MAGIC) {
807 regerror("corrupted program");
808 return(0);
811 /* If there is a "must appear" string, look for it. */
812 if (prog->regmust != NULL) {
813 /* LINTED const castaway */
814 s = (char *)string;
815 while ((s = strchr(s, prog->regmust[0])) != NULL) {
816 if (strncmp(s, prog->regmust,
817 (size_t)prog->regmlen) == 0)
818 break; /* Found it. */
819 s++;
821 if (s == NULL) /* Not present. */
822 return(0);
825 /* Mark beginning of line for ^ . */
826 /* LINTED const castaway */
827 regbol = (char *)string;
829 /* Simplest case: anchored match need be tried only once. */
830 if (prog->reganch)
831 return(regtry(prog, string));
833 /* Messy cases: unanchored match. */
834 /* LINTED const castaway */
835 s = (char *)string;
836 if (prog->regstart != '\0')
837 /* We know what char it must start with. */
838 while ((s = strchr(s, prog->regstart)) != NULL) {
839 if (regtry(prog, s))
840 return(1);
841 s++;
843 else
844 /* We don't -- general case. */
845 do {
846 if (regtry(prog, s))
847 return(1);
848 } while (*s++ != '\0');
850 /* Failure. */
851 return(0);
855 - regtry - try match at specific point
857 static int /* 0 failure, 1 success */
858 regtry(prog, string)
859 const regexp *prog;
860 const char *string;
862 int i;
863 char **sp;
864 char **ep;
866 /* LINTED const castaway */
867 reginput = (char *)string; /* XXX */
868 regstartp = (char **)prog->startp; /* XXX */
869 regendp = (char **)prog->endp; /* XXX */
871 sp = (char **)prog->startp; /* XXX */
872 ep = (char **)prog->endp; /* XXX */
873 for (i = NSUBEXP; i > 0; i--) {
874 *sp++ = NULL;
875 *ep++ = NULL;
877 if (regmatch((char *)prog->program + 1)) { /* XXX */
878 /* LINTED const castaway */
879 ((regexp *)prog)->startp[0] = (char *)string; /* XXX */
880 /* LINTED const castaway */
881 ((regexp *)prog)->endp[0] = reginput; /* XXX */
882 return(1);
883 } else
884 return(0);
888 - regmatch - main matching routine
890 * Conceptually the strategy is simple: check to see whether the current
891 * node matches, call self recursively to see whether the rest matches,
892 * and then act accordingly. In practice we make some effort to avoid
893 * recursion, in particular by going through "ordinary" nodes (that don't
894 * need to know whether the rest of the match failed) by a loop instead of
895 * by recursion.
897 static int /* 0 failure, 1 success */
898 regmatch(prog)
899 char *prog;
901 char *scan; /* Current node. */
902 char *next; /* Next node. */
904 scan = prog;
905 #ifdef DEBUG
906 if (scan != NULL && regnarrate)
907 fprintf(stderr, "%s(\n", regprop(scan));
908 #endif
909 while (scan != NULL) {
910 #ifdef DEBUG
911 if (regnarrate)
912 fprintf(stderr, "%s...\n", regprop(scan));
913 #endif
914 next = regnext(scan);
916 switch (OP(scan)) {
917 case BOL:
918 if (reginput != regbol)
919 return(0);
920 break;
921 case EOL:
922 if (*reginput != '\0')
923 return(0);
924 break;
925 case WORDA:
926 /* Must be looking at a letter, digit, or _ */
927 if ((!isalnum(UCHARAT(reginput))) && *reginput != '_')
928 return(0);
929 /* Prev must be BOL or nonword */
930 if (reginput > regbol &&
931 (isalnum(UCHARAT(reginput - 1))
932 || reginput[-1] == '_'))
933 return(0);
934 break;
935 case WORDZ:
936 /* Must be looking at non letter, digit, or _ */
937 if (isalnum(UCHARAT(reginput)) || *reginput == '_')
938 return(0);
939 /* We don't care what the previous char was */
940 break;
941 case ANY:
942 if (*reginput == '\0')
943 return(0);
944 reginput++;
945 break;
946 case EXACTLY: {
947 int len;
948 char *opnd;
950 opnd = OPERAND(scan);
951 /* Inline the first character, for speed. */
952 if (*opnd != *reginput)
953 return(0);
954 len = strlen(opnd);
955 if (len > 1 && strncmp(opnd, reginput,
956 (size_t)len) != 0)
957 return(0);
958 reginput += len;
960 break;
961 case ANYOF:
962 if (*reginput == '\0' || strchr(OPERAND(scan), *reginput) == NULL)
963 return(0);
964 reginput++;
965 break;
966 case ANYBUT:
967 if (*reginput == '\0' || strchr(OPERAND(scan), *reginput) != NULL)
968 return(0);
969 reginput++;
970 break;
971 case NOTHING:
972 break;
973 case BACK:
974 break;
975 case OPEN+1:
976 case OPEN+2:
977 case OPEN+3:
978 case OPEN+4:
979 case OPEN+5:
980 case OPEN+6:
981 case OPEN+7:
982 case OPEN+8:
983 case OPEN+9: {
984 int no;
985 char *save;
987 no = OP(scan) - OPEN;
988 save = reginput;
990 if (regmatch(next)) {
992 * Don't set startp if some later
993 * invocation of the same parentheses
994 * already has.
996 if (regstartp[no] == NULL)
997 regstartp[no] = save;
998 return(1);
999 } else
1000 return(0);
1002 case CLOSE+1:
1003 case CLOSE+2:
1004 case CLOSE+3:
1005 case CLOSE+4:
1006 case CLOSE+5:
1007 case CLOSE+6:
1008 case CLOSE+7:
1009 case CLOSE+8:
1010 case CLOSE+9: {
1011 int no;
1012 char *save;
1014 no = OP(scan) - CLOSE;
1015 save = reginput;
1017 if (regmatch(next)) {
1019 * Don't set endp if some later
1020 * invocation of the same parentheses
1021 * already has.
1023 if (regendp[no] == NULL)
1024 regendp[no] = save;
1025 return(1);
1026 } else
1027 return(0);
1029 case BRANCH: {
1030 char *save;
1032 if (OP(next) != BRANCH) /* No choice. */
1033 next = OPERAND(scan); /* Avoid recursion. */
1034 else {
1035 do {
1036 save = reginput;
1037 if (regmatch(OPERAND(scan)))
1038 return(1);
1039 reginput = save;
1040 scan = regnext(scan);
1041 } while (scan != NULL && OP(scan) == BRANCH);
1042 return(0);
1043 /* NOTREACHED */
1046 break;
1047 case STAR:
1048 case PLUS: {
1049 char nextch;
1050 int no;
1051 char *save;
1052 int min;
1055 * Lookahead to avoid useless match attempts
1056 * when we know what character comes next.
1058 nextch = '\0';
1059 if (OP(next) == EXACTLY)
1060 nextch = *OPERAND(next);
1061 min = (OP(scan) == STAR) ? 0 : 1;
1062 save = reginput;
1063 no = regrepeat(OPERAND(scan));
1064 while (no >= min) {
1065 /* If it could work, try it. */
1066 if (nextch == '\0' || *reginput == nextch)
1067 if (regmatch(next))
1068 return(1);
1069 /* Couldn't or didn't -- back up. */
1070 no--;
1071 reginput = save + no;
1073 return(0);
1075 case END:
1076 return(1); /* Success! */
1077 default:
1078 regerror("memory corruption");
1079 return(0);
1082 scan = next;
1086 * We get here only if there's trouble -- normally "case END" is
1087 * the terminating point.
1089 regerror("corrupted pointers");
1090 return(0);
1094 - regrepeat - repeatedly match something simple, report how many
1096 static int
1097 regrepeat(p)
1098 char *p;
1100 int count = 0;
1101 char *scan;
1102 char *opnd;
1104 scan = reginput;
1105 opnd = OPERAND(p);
1106 switch (OP(p)) {
1107 case ANY:
1108 count = strlen(scan);
1109 scan += count;
1110 break;
1111 case EXACTLY:
1112 while (*opnd == *scan) {
1113 count++;
1114 scan++;
1116 break;
1117 case ANYOF:
1118 while (*scan != '\0' && strchr(opnd, *scan) != NULL) {
1119 count++;
1120 scan++;
1122 break;
1123 case ANYBUT:
1124 while (*scan != '\0' && strchr(opnd, *scan) == NULL) {
1125 count++;
1126 scan++;
1128 break;
1129 default: /* Oh dear. Called inappropriately. */
1130 regerror("internal foulup");
1131 count = 0; /* Best compromise. */
1132 break;
1134 reginput = scan;
1136 return(count);
1140 - regnext - dig the "next" pointer out of a node
1142 static char *
1143 regnext(p)
1144 char *p;
1146 int offset;
1148 if (p == &regdummy)
1149 return(NULL);
1151 offset = NEXT(p);
1152 if (offset == 0)
1153 return(NULL);
1155 if (OP(p) == BACK)
1156 return(p-offset);
1157 else
1158 return(p+offset);
1161 #ifdef DEBUG
1164 - regdump - dump a regexp onto stdout in vaguely comprehensible form
1166 void
1167 regdump(r)
1168 regexp *r;
1170 char *s;
1171 char op = EXACTLY; /* Arbitrary non-END op. */
1172 char *next;
1175 s = r->program + 1;
1176 while (op != END) { /* While that wasn't END last time... */
1177 op = OP(s);
1178 printf("%2td%s", s-r->program, regprop(s)); /* Where, what. */
1179 next = regnext(s);
1180 if (next == NULL) /* Next ptr. */
1181 printf("(0)");
1182 else
1183 printf("(%td)", (s-r->program)+(next-s));
1184 s += 3;
1185 if (op == ANYOF || op == ANYBUT || op == EXACTLY) {
1186 /* Literal string, where present. */
1187 while (*s != '\0') {
1188 putchar(*s);
1189 s++;
1191 s++;
1193 putchar('\n');
1196 /* Header fields of interest. */
1197 if (r->regstart != '\0')
1198 printf("start `%c' ", r->regstart);
1199 if (r->reganch)
1200 printf("anchored ");
1201 if (r->regmust != NULL)
1202 printf("must have \"%s\"", r->regmust);
1203 printf("\n");
1207 - regprop - printable representation of opcode
1209 static char *
1210 regprop(op)
1211 char *op;
1213 char *p;
1214 static char buf[50];
1216 (void)strncpy(buf, ":", sizeof(buf) - 1);
1218 switch (OP(op)) {
1219 case BOL:
1220 p = "BOL";
1221 break;
1222 case EOL:
1223 p = "EOL";
1224 break;
1225 case ANY:
1226 p = "ANY";
1227 break;
1228 case ANYOF:
1229 p = "ANYOF";
1230 break;
1231 case ANYBUT:
1232 p = "ANYBUT";
1233 break;
1234 case BRANCH:
1235 p = "BRANCH";
1236 break;
1237 case EXACTLY:
1238 p = "EXACTLY";
1239 break;
1240 case NOTHING:
1241 p = "NOTHING";
1242 break;
1243 case BACK:
1244 p = "BACK";
1245 break;
1246 case END:
1247 p = "END";
1248 break;
1249 case OPEN+1:
1250 case OPEN+2:
1251 case OPEN+3:
1252 case OPEN+4:
1253 case OPEN+5:
1254 case OPEN+6:
1255 case OPEN+7:
1256 case OPEN+8:
1257 case OPEN+9:
1258 (void)snprintf(buf+strlen(buf), sizeof(buf) - strlen(buf),
1259 "OPEN%d", OP(op)-OPEN);
1260 p = NULL;
1261 break;
1262 case CLOSE+1:
1263 case CLOSE+2:
1264 case CLOSE+3:
1265 case CLOSE+4:
1266 case CLOSE+5:
1267 case CLOSE+6:
1268 case CLOSE+7:
1269 case CLOSE+8:
1270 case CLOSE+9:
1271 (void)snprintf(buf+strlen(buf), sizeof(buf) - strlen(buf),
1272 "CLOSE%d", OP(op)-CLOSE);
1273 p = NULL;
1274 break;
1275 case STAR:
1276 p = "STAR";
1277 break;
1278 case PLUS:
1279 p = "PLUS";
1280 break;
1281 case WORDA:
1282 p = "WORDA";
1283 break;
1284 case WORDZ:
1285 p = "WORDZ";
1286 break;
1287 default:
1288 p = NULL;
1289 regerror("corrupted opcode");
1290 break;
1292 if (p != NULL)
1293 (void)strncat(buf, p, sizeof(buf) - strlen(buf) - 1);
1294 return(buf);
1296 #endif
1299 * The following is provided for those people who do not have strcspn() in
1300 * their C libraries. They should get off their butts and do something
1301 * about it; at least one public-domain implementation of those (highly
1302 * useful) string routines has been published on Usenet.
1304 #ifdef STRCSPN
1306 * strcspn - find length of initial segment of s1 consisting entirely
1307 * of characters not from s2
1310 static int
1311 strcspn(s1, s2)
1312 char *s1;
1313 char *s2;
1315 char *scan1;
1316 char *scan2;
1317 int count;
1319 count = 0;
1320 for (scan1 = s1; *scan1 != '\0'; scan1++) {
1321 for (scan2 = s2; *scan2 != '\0';) /* ++ moved down. */
1322 if (*scan1 == *scan2++)
1323 return(count);
1324 count++;
1326 return(count);
1328 #endif