1 /*
2 * regcomp and regexec -- regsub and regerror are elsewhere
3 *
4 * Copyright (c) 1986 by University of Toronto.
5 * Written by Henry Spencer. Not derived from licensed software.
6 *
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:
10 *
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.
14 *
15 * 2. The origin of this software must not be misrepresented, either
16 * by explicit claim or by omission.
17 *
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.
30 *
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.
34 */
35
36 #include <sys/cdefs.h>
37 #if defined(LIBC_SCCS) && !defined(lint)
38 __RCSID("$NetBSD: regexp.c,v 1.19 2016/01/26 16:05:18 christos Exp $");
39 #endif /* LIBC_SCCS and not lint */
40
41 #include <ctype.h>
42 #include <regexp.h>
43 #include <stdio.h>
44 #include <stdlib.h>
45 #include <string.h>
46 #include "regmagic.h"
47
48 /*
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:
52 *
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
57 *
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.
66 */
67
68 /*
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:
82 */
83
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. */
102
103 /*
104 * Opcode notes:
105 *
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.
113 *
114 * BACK Normal "next" pointers all implicitly point forward; BACK
115 * exists to make loop structures possible.
116 *
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.
121 *
122 * OPEN,CLOSE ...are numbered at compile time.
123 */
124
125 /*
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.)
131 *
132 * Using two bytes for the "next" pointer is vast overkill for most things,
133 * but allows patterns to get big without disasters.
134 */
135 #define OP(p) (*(p))
136 #define NEXT(p) (((*((p)+1)&0377)<<8) + (*((p)+2)&0377))
137 #define OPERAND(p) ((p) + 3)
138
139 /*
140 * See regmagic.h for one further detail of program structure.
141 */
142
143
144 /*
145 * Utility definitions.
146 */
147 #ifndef CHARBITS
148 #define UCHARAT(p) ((int)*(unsigned char *)(p))
149 #else
150 #define UCHARAT(p) ((int)*(p)&CHARBITS)
151 #endif
152
153 #define FAIL(m) { regerror(m); return(NULL); }
154 #define ISMULT(c) ((c) == '*' || (c) == '+' || (c) == '?')
155
156 /*
157 * Flags to be passed up and down.
158 */
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. */
163
164 /*
165 * Global work variables for regcomp().
166 */
167 static char *regparse; /* Input-scan pointer. */
168 static int regnpar; /* () count. */
169 static char regdummy;
170 static char *regcode; /* Code-emit pointer; ®dummy = don't. */
171 static long regsize; /* Code size. */
172
173 /*
174 * Forward declarations for regcomp()'s friends.
175 */
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
192
193 /*
194 - regcomp - compile a regular expression into internal code
195 *
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.)
204 *
205 * Beware that the optimization-preparation code in here knows about some
206 * of the structure of the compiled regexp.
207 */
208 regexp *
__compat_regcomp(expn)209 __compat_regcomp(expn)
210 const char *expn;
211 {
212 regexp *r;
213 char *scan;
214 char *longest;
215 int len;
216 int flags;
217
218 if (expn == NULL)
219 FAIL("NULL argument");
220
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 = ®dummy;
230 regc(MAGIC);
231 if (reg(0, &flags) == NULL)
232 return(NULL);
233
234 /* Small enough for pointer-storage convention? */
235 if (regsize >= 32767L) /* Probably could be 65535L. */
236 FAIL("regexp too big");
237
238 /* Allocate space. */
239 r = malloc(sizeof(regexp) + (unsigned)regsize);
240 if (r == NULL)
241 FAIL("out of space");
242
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 free(r);
251 return(NULL);
252 }
253
254 /* Dig out information for optimizations. */
255 r->regstart = '\0'; /* Worst-case defaults. */
256 r->reganch = 0;
257 r->regmust = NULL;
258 r->regmlen = 0;
259 scan = r->program+1; /* First BRANCH. */
260 if (OP(regnext(scan)) == END) { /* Only one top-level choice. */
261 scan = OPERAND(scan);
262
263 /* Starting-point info. */
264 if (OP(scan) == EXACTLY)
265 r->regstart = *OPERAND(scan);
266 else if (OP(scan) == BOL)
267 r->reganch++;
268
269 /*
270 * If there's something expensive in the r.e., find the
271 * longest literal string that must appear and make it the
272 * regmust. Resolve ties in favor of later strings, since
273 * the regstart check works with the beginning of the r.e.
274 * and avoiding duplication strengthens checking. Not a
275 * strong reason, but sufficient in the absence of others.
276 */
277 if (flags&SPSTART) {
278 longest = NULL;
279 len = 0;
280 for (; scan != NULL; scan = regnext(scan))
281 if (OP(scan) == EXACTLY && (int) strlen(OPERAND(scan)) >= len) {
282 longest = OPERAND(scan);
283 len = strlen(OPERAND(scan));
284 }
285 r->regmust = longest;
286 r->regmlen = len;
287 }
288 }
289
290 return(r);
291 }
292
293 /*
294 - reg - regular expression, i.e. main body or parenthesized thing
295 *
296 * Caller must absorb opening parenthesis.
297 *
298 * Combining parenthesis handling with the base level of regular expression
299 * is a trifle forced, but the need to tie the tails of the branches to what
300 * follows makes it hard to avoid.
301 */
302 static char *
reg(paren,flagp)303 reg(paren, flagp)
304 int paren; /* Parenthesized? */
305 int *flagp;
306 {
307 char *ret;
308 char *br;
309 char *ender;
310 int parno = 0;
311 int flags;
312
313 *flagp = HASWIDTH; /* Tentatively. */
314
315 /* Make an OPEN node, if parenthesized. */
316 if (paren) {
317 if (regnpar >= NSUBEXP)
318 FAIL("too many ()");
319 parno = regnpar;
320 regnpar++;
321 ret = regnode(OPEN+parno);
322 } else
323 ret = NULL;
324
325 /* Pick up the branches, linking them together. */
326 br = regbranch(&flags);
327 if (br == NULL)
328 return(NULL);
329 if (ret != NULL)
330 regtail(ret, br); /* OPEN -> first. */
331 else
332 ret = br;
333 if (!(flags&HASWIDTH))
334 *flagp &= ~HASWIDTH;
335 *flagp |= flags&SPSTART;
336 while (*regparse == '|' || *regparse == '\n') {
337 regparse++;
338 br = regbranch(&flags);
339 if (br == NULL)
340 return(NULL);
341 regtail(ret, br); /* BRANCH -> BRANCH. */
342 if (!(flags&HASWIDTH))
343 *flagp &= ~HASWIDTH;
344 *flagp |= flags&SPSTART;
345 }
346
347 /* Make a closing node, and hook it on the end. */
348 ender = regnode((paren) ? CLOSE+parno : END);
349 regtail(ret, ender);
350
351 /* Hook the tails of the branches to the closing node. */
352 for (br = ret; br != NULL; br = regnext(br))
353 regoptail(br, ender);
354
355 /* Check for proper termination. */
356 if (paren && *regparse++ != ')') {
357 FAIL("unmatched ()");
358 } else if (!paren && *regparse != '\0') {
359 if (*regparse == ')') {
360 FAIL("unmatched ()");
361 } else
362 FAIL("junk on end"); /* "Can't happen". */
363 /* NOTREACHED */
364 }
365
366 return(ret);
367 }
368
369 /*
370 - regbranch - one alternative of an | operator
371 *
372 * Implements the concatenation operator.
373 */
374 static char *
regbranch(flagp)375 regbranch(flagp)
376 int *flagp;
377 {
378 char *ret;
379 char *chain;
380 char *latest;
381 int flags;
382
383 *flagp = WORST; /* Tentatively. */
384
385 ret = regnode(BRANCH);
386 chain = NULL;
387 while (*regparse != '\0' && *regparse != ')' &&
388 *regparse != '\n' && *regparse != '|') {
389 latest = regpiece(&flags);
390 if (latest == NULL)
391 return(NULL);
392 *flagp |= flags&HASWIDTH;
393 if (chain == NULL) /* First piece. */
394 *flagp |= flags&SPSTART;
395 else
396 regtail(chain, latest);
397 chain = latest;
398 }
399 if (chain == NULL) /* Loop ran zero times. */
400 (void) regnode(NOTHING);
401
402 return(ret);
403 }
404
405 /*
406 - regpiece - something followed by possible [*+?]
407 *
408 * Note that the branching code sequences used for ? and the general cases
409 * of * and + are somewhat optimized: they use the same NOTHING node as
410 * both the endmarker for their branch list and the body of the last branch.
411 * It might seem that this node could be dispensed with entirely, but the
412 * endmarker role is not redundant.
413 */
414 static char *
regpiece(flagp)415 regpiece(flagp)
416 int *flagp;
417 {
418 char *ret;
419 char op;
420 char *next;
421 int flags;
422
423 ret = regatom(&flags);
424 if (ret == NULL)
425 return(NULL);
426
427 op = *regparse;
428 if (!ISMULT(op)) {
429 *flagp = flags;
430 return(ret);
431 }
432
433 if (!(flags&HASWIDTH) && op != '?')
434 FAIL("*+ operand could be empty");
435 *flagp = (op != '+') ? (WORST|SPSTART) : (WORST|HASWIDTH);
436
437 if (op == '*' && (flags&SIMPLE))
438 reginsert(STAR, ret);
439 else if (op == '*') {
440 /* Emit x* as (x&|), where & means "self". */
441 reginsert(BRANCH, ret); /* Either x */
442 regoptail(ret, regnode(BACK)); /* and loop */
443 regoptail(ret, ret); /* back */
444 regtail(ret, regnode(BRANCH)); /* or */
445 regtail(ret, regnode(NOTHING)); /* null. */
446 } else if (op == '+' && (flags&SIMPLE))
447 reginsert(PLUS, ret);
448 else if (op == '+') {
449 /* Emit x+ as x(&|), where & means "self". */
450 next = regnode(BRANCH); /* Either */
451 regtail(ret, next);
452 regtail(regnode(BACK), ret); /* loop back */
453 regtail(next, regnode(BRANCH)); /* or */
454 regtail(ret, regnode(NOTHING)); /* null. */
455 } else if (op == '?') {
456 /* Emit x? as (x|) */
457 reginsert(BRANCH, ret); /* Either x */
458 regtail(ret, regnode(BRANCH)); /* or */
459 next = regnode(NOTHING); /* null. */
460 regtail(ret, next);
461 regoptail(ret, next);
462 }
463 regparse++;
464 if (ISMULT(*regparse))
465 FAIL("nested *?+");
466
467 return(ret);
468 }
469
470 /*
471 - regatom - the lowest level
472 *
473 * Optimization: gobbles an entire sequence of ordinary characters so that
474 * it can turn them into a single node, which is smaller to store and
475 * faster to run. Backslashed characters are exceptions, each becoming a
476 * separate node; the code is simpler that way and it's not worth fixing.
477 */
478 static char *
regatom(flagp)479 regatom(flagp)
480 int *flagp;
481 {
482 char *ret;
483 int flags;
484
485 *flagp = WORST; /* Tentatively. */
486
487 switch (*regparse++) {
488 /* FIXME: these chars only have meaning at beg/end of pat? */
489 case '^':
490 ret = regnode(BOL);
491 break;
492 case '$':
493 ret = regnode(EOL);
494 break;
495 case '.':
496 ret = regnode(ANY);
497 *flagp |= HASWIDTH|SIMPLE;
498 break;
499 case '[': {
500 int class;
501 int classend;
502
503 if (*regparse == '^') { /* Complement of range. */
504 ret = regnode(ANYBUT);
505 regparse++;
506 } else
507 ret = regnode(ANYOF);
508 if (*regparse == ']' || *regparse == '-')
509 regc(*regparse++);
510 while (*regparse != '\0' && *regparse != ']') {
511 if (*regparse == '-') {
512 regparse++;
513 if (*regparse == ']' || *regparse == '\0')
514 regc('-');
515 else {
516 class = UCHARAT(regparse-2)+1;
517 classend = UCHARAT(regparse);
518 if (class > classend+1)
519 FAIL("invalid [] range");
520 for (; class <= classend; class++)
521 regc(class);
522 regparse++;
523 }
524 } else
525 regc(*regparse++);
526 }
527 regc('\0');
528 if (*regparse != ']')
529 FAIL("unmatched []");
530 regparse++;
531 *flagp |= HASWIDTH|SIMPLE;
532 }
533 break;
534 case '(':
535 ret = reg(1, &flags);
536 if (ret == NULL)
537 return(NULL);
538 *flagp |= flags&(HASWIDTH|SPSTART);
539 break;
540 case '\0':
541 case '|':
542 case '\n':
543 case ')':
544 FAIL("internal urp"); /* Supposed to be caught earlier. */
545 case '?':
546 case '+':
547 case '*':
548 FAIL("?+* follows nothing");
549 case '\\':
550 switch (*regparse++) {
551 case '\0':
552 FAIL("trailing \\");
553 case '<':
554 ret = regnode(WORDA);
555 break;
556 case '>':
557 ret = regnode(WORDZ);
558 break;
559 /* FIXME: Someday handle \1, \2, ... */
560 default:
561 /* Handle general quoted chars in exact-match routine */
562 goto de_fault;
563 }
564 break;
565 de_fault:
566 /*FALLTHROUGH*/
567 default:
568 /*
569 * Encode a string of characters to be matched exactly.
570 *
571 * This is a bit tricky due to quoted chars and due to
572 * '*', '+', and '?' taking the SINGLE char previous
573 * as their operand.
574 *
575 * On entry, the char at regparse[-1] is going to go
576 * into the string, no matter what it is. (It could be
577 * following a \ if we are entered from the '\' case.)
578 *
579 * Basic idea is to pick up a good char in ch and
580 * examine the next char. If it's *+? then we twiddle.
581 * If it's \ then we frozzle. If it's other magic char
582 * we push ch and terminate the string. If none of the
583 * above, we push ch on the string and go around again.
584 *
585 * regprev is used to remember where "the current char"
586 * starts in the string, if due to a *+? we need to back
587 * up and put the current char in a separate, 1-char, string.
588 * When regprev is NULL, ch is the only char in the
589 * string; this is used in *+? handling, and in setting
590 * flags |= SIMPLE at the end.
591 */
592 {
593 char *regprev;
594 char ch;
595
596 regparse--; /* Look at cur char */
597 ret = regnode(EXACTLY);
598 for ( regprev = 0 ; ; ) {
599 ch = *regparse++; /* Get current char */
600 switch (*regparse) { /* look at next one */
601
602 default:
603 regc(ch); /* Add cur to string */
604 break;
605
606 case '.': case '[': case '(':
607 case ')': case '|': case '\n':
608 case '$': case '^':
609 case '\0':
610 /* FIXME, $ and ^ should not always be magic */
611 magic:
612 regc(ch); /* dump cur char */
613 goto done; /* and we are done */
614
615 case '?': case '+': case '*':
616 if (!regprev) /* If just ch in str, */
617 goto magic; /* use it */
618 /* End mult-char string one early */
619 regparse = regprev; /* Back up parse */
620 goto done;
621
622 case '\\':
623 regc(ch); /* Cur char OK */
624 switch (regparse[1]){ /* Look after \ */
625 case '\0':
626 case '<':
627 case '>':
628 /* FIXME: Someday handle \1, \2, ... */
629 goto done; /* Not quoted */
630 default:
631 /* Backup point is \, scan * point is after it. */
632 regprev = regparse;
633 regparse++;
634 continue; /* NOT break; */
635 }
636 }
637 regprev = regparse; /* Set backup point */
638 }
639 done:
640 regc('\0');
641 *flagp |= HASWIDTH;
642 if (!regprev) /* One char? */
643 *flagp |= SIMPLE;
644 }
645 break;
646 }
647
648 return(ret);
649 }
650
651 /*
652 - regnode - emit a node
653 */
654 static char * /* Location. */
regnode(op)655 regnode(op)
656 int op;
657 {
658 char *ret;
659 char *ptr;
660
661 ret = regcode;
662 if (ret == ®dummy) {
663 regsize += 3;
664 return(ret);
665 }
666
667 ptr = ret;
668 *ptr++ = op;
669 *ptr++ = '\0'; /* Null "next" pointer. */
670 *ptr++ = '\0';
671 regcode = ptr;
672
673 return(ret);
674 }
675
676 /*
677 - regc - emit (if appropriate) a byte of code
678 */
679 static void
regc(b)680 regc(b)
681 int b;
682 {
683 if (regcode != ®dummy)
684 *regcode++ = b;
685 else
686 regsize++;
687 }
688
689 /*
690 - reginsert - insert an operator in front of already-emitted operand
691 *
692 * Means relocating the operand.
693 */
694 static void
reginsert(op,opnd)695 reginsert(op, opnd)
696 int op;
697 char *opnd;
698 {
699 char *src;
700 char *dst;
701 char *place;
702
703 if (regcode == ®dummy) {
704 regsize += 3;
705 return;
706 }
707
708 src = regcode;
709 regcode += 3;
710 dst = regcode;
711 while (src > opnd)
712 *--dst = *--src;
713
714 place = opnd; /* Op node, where operand used to be. */
715 *place++ = op;
716 *place++ = '\0';
717 *place++ = '\0';
718 }
719
720 /*
721 - regtail - set the next-pointer at the end of a node chain
722 */
723 static void
regtail(p,val)724 regtail(p, val)
725 char *p;
726 char *val;
727 {
728 char *scan;
729 char *temp;
730 int offset;
731
732 if (p == ®dummy)
733 return;
734
735 /* Find last node. */
736 scan = p;
737 for (;;) {
738 temp = regnext(scan);
739 if (temp == NULL)
740 break;
741 scan = temp;
742 }
743
744 if (OP(scan) == BACK)
745 offset = scan - val;
746 else
747 offset = val - scan;
748 *(scan+1) = ((unsigned int)offset>>8)&0377;
749 *(scan+2) = offset&0377;
750 }
751
752 /*
753 - regoptail - regtail on operand of first argument; nop if operandless
754 */
755 static void
regoptail(p,val)756 regoptail(p, val)
757 char *p;
758 char *val;
759 {
760 /* "Operandless" and "op != BRANCH" are synonymous in practice. */
761 if (p == NULL || p == ®dummy || OP(p) != BRANCH)
762 return;
763 regtail(OPERAND(p), val);
764 }
765
766 /*
767 * regexec and friends
768 */
769
770 /*
771 * Global work variables for regexec().
772 */
773 static char *reginput; /* String-input pointer. */
774 static char *regbol; /* Beginning of input, for ^ check. */
775 static char **regstartp; /* Pointer to startp array. */
776 static char **regendp; /* Ditto for endp. */
777
778 /*
779 * Forwards.
780 */
781 STATIC int regtry __P((const regexp *, const char *));
782 STATIC int regmatch __P((char *));
783 STATIC int regrepeat __P((char *));
784
785 #ifdef DEBUG
786 int regnarrate = 0;
787 void regdump __P((regexp *));
788 STATIC char *regprop __P((char *));
789 #endif
790
791 /*
792 - regexec - match a regexp against a string
793 */
794 int
__compat_regexec(prog,string)795 __compat_regexec(prog, string)
796 const regexp *prog;
797 const char *string;
798 {
799 char *s;
800
801 /* Be paranoid... */
802 if (prog == NULL || string == NULL) {
803 regerror("NULL parameter");
804 return(0);
805 }
806
807 /* Check validity of program. */
808 if (UCHARAT(prog->program) != MAGIC) {
809 regerror("corrupted program");
810 return(0);
811 }
812
813 /* If there is a "must appear" string, look for it. */
814 if (prog->regmust != NULL) {
815 /* LINTED const castaway */
816 s = (char *)string;
817 while ((s = strchr(s, prog->regmust[0])) != NULL) {
818 if (strncmp(s, prog->regmust,
819 (size_t)prog->regmlen) == 0)
820 break; /* Found it. */
821 s++;
822 }
823 if (s == NULL) /* Not present. */
824 return(0);
825 }
826
827 /* Mark beginning of line for ^ . */
828 /* LINTED const castaway */
829 regbol = (char *)string;
830
831 /* Simplest case: anchored match need be tried only once. */
832 if (prog->reganch)
833 return(regtry(prog, string));
834
835 /* Messy cases: unanchored match. */
836 /* LINTED const castaway */
837 s = (char *)string;
838 if (prog->regstart != '\0')
839 /* We know what char it must start with. */
840 while ((s = strchr(s, prog->regstart)) != NULL) {
841 if (regtry(prog, s))
842 return(1);
843 s++;
844 }
845 else
846 /* We don't -- general case. */
847 do {
848 if (regtry(prog, s))
849 return(1);
850 } while (*s++ != '\0');
851
852 /* Failure. */
853 return(0);
854 }
855
856 /*
857 - regtry - try match at specific point
858 */
859 static int /* 0 failure, 1 success */
regtry(prog,string)860 regtry(prog, string)
861 const regexp *prog;
862 const char *string;
863 {
864 int i;
865 char **sp;
866 char **ep;
867
868 /* LINTED const castaway */
869 reginput = (char *)string; /* XXX */
870 regstartp = (char **)prog->startp; /* XXX */
871 regendp = (char **)prog->endp; /* XXX */
872
873 sp = (char **)prog->startp; /* XXX */
874 ep = (char **)prog->endp; /* XXX */
875 for (i = NSUBEXP; i > 0; i--) {
876 *sp++ = NULL;
877 *ep++ = NULL;
878 }
879 if (regmatch((char *)prog->program + 1)) { /* XXX */
880 /* LINTED const castaway */
881 ((regexp *)prog)->startp[0] = (char *)string; /* XXX */
882 /* LINTED const castaway */
883 ((regexp *)prog)->endp[0] = reginput; /* XXX */
884 return(1);
885 } else
886 return(0);
887 }
888
889 /*
890 - regmatch - main matching routine
891 *
892 * Conceptually the strategy is simple: check to see whether the current
893 * node matches, call self recursively to see whether the rest matches,
894 * and then act accordingly. In practice we make some effort to avoid
895 * recursion, in particular by going through "ordinary" nodes (that don't
896 * need to know whether the rest of the match failed) by a loop instead of
897 * by recursion.
898 */
899 static int /* 0 failure, 1 success */
regmatch(prog)900 regmatch(prog)
901 char *prog;
902 {
903 char *scan; /* Current node. */
904 char *next; /* Next node. */
905
906 scan = prog;
907 #ifdef DEBUG
908 if (scan != NULL && regnarrate)
909 fprintf(stderr, "%s(\n", regprop(scan));
910 #endif
911 while (scan != NULL) {
912 #ifdef DEBUG
913 if (regnarrate)
914 fprintf(stderr, "%s...\n", regprop(scan));
915 #endif
916 next = regnext(scan);
917
918 switch (OP(scan)) {
919 case BOL:
920 if (reginput != regbol)
921 return(0);
922 break;
923 case EOL:
924 if (*reginput != '\0')
925 return(0);
926 break;
927 case WORDA:
928 /* Must be looking at a letter, digit, or _ */
929 if ((!isalnum(UCHARAT(reginput))) && *reginput != '_')
930 return(0);
931 /* Prev must be BOL or nonword */
932 if (reginput > regbol &&
933 (isalnum(UCHARAT(reginput - 1))
934 || reginput[-1] == '_'))
935 return(0);
936 break;
937 case WORDZ:
938 /* Must be looking at non letter, digit, or _ */
939 if (isalnum(UCHARAT(reginput)) || *reginput == '_')
940 return(0);
941 /* We don't care what the previous char was */
942 break;
943 case ANY:
944 if (*reginput == '\0')
945 return(0);
946 reginput++;
947 break;
948 case EXACTLY: {
949 int len;
950 char *opnd;
951
952 opnd = OPERAND(scan);
953 /* Inline the first character, for speed. */
954 if (*opnd != *reginput)
955 return(0);
956 len = strlen(opnd);
957 if (len > 1 && strncmp(opnd, reginput,
958 (size_t)len) != 0)
959 return(0);
960 reginput += len;
961 }
962 break;
963 case ANYOF:
964 if (*reginput == '\0' || strchr(OPERAND(scan), *reginput) == NULL)
965 return(0);
966 reginput++;
967 break;
968 case ANYBUT:
969 if (*reginput == '\0' || strchr(OPERAND(scan), *reginput) != NULL)
970 return(0);
971 reginput++;
972 break;
973 case NOTHING:
974 break;
975 case BACK:
976 break;
977 case OPEN+1:
978 case OPEN+2:
979 case OPEN+3:
980 case OPEN+4:
981 case OPEN+5:
982 case OPEN+6:
983 case OPEN+7:
984 case OPEN+8:
985 case OPEN+9: {
986 int no;
987 char *save;
988
989 no = OP(scan) - OPEN;
990 save = reginput;
991
992 if (regmatch(next)) {
993 /*
994 * Don't set startp if some later
995 * invocation of the same parentheses
996 * already has.
997 */
998 if (regstartp[no] == NULL)
999 regstartp[no] = save;
1000 return(1);
1001 } else
1002 return(0);
1003 }
1004 case CLOSE+1:
1005 case CLOSE+2:
1006 case CLOSE+3:
1007 case CLOSE+4:
1008 case CLOSE+5:
1009 case CLOSE+6:
1010 case CLOSE+7:
1011 case CLOSE+8:
1012 case CLOSE+9: {
1013 int no;
1014 char *save;
1015
1016 no = OP(scan) - CLOSE;
1017 save = reginput;
1018
1019 if (regmatch(next)) {
1020 /*
1021 * Don't set endp if some later
1022 * invocation of the same parentheses
1023 * already has.
1024 */
1025 if (regendp[no] == NULL)
1026 regendp[no] = save;
1027 return(1);
1028 } else
1029 return(0);
1030 }
1031 case BRANCH: {
1032 char *save;
1033
1034 if (OP(next) != BRANCH) /* No choice. */
1035 next = OPERAND(scan); /* Avoid recursion. */
1036 else {
1037 do {
1038 save = reginput;
1039 if (regmatch(OPERAND(scan)))
1040 return(1);
1041 reginput = save;
1042 scan = regnext(scan);
1043 } while (scan != NULL && OP(scan) == BRANCH);
1044 return(0);
1045 /* NOTREACHED */
1046 }
1047 }
1048 break;
1049 case STAR:
1050 case PLUS: {
1051 char nextch;
1052 int no;
1053 char *save;
1054 int min;
1055
1056 /*
1057 * Lookahead to avoid useless match attempts
1058 * when we know what character comes next.
1059 */
1060 nextch = '\0';
1061 if (OP(next) == EXACTLY)
1062 nextch = *OPERAND(next);
1063 min = (OP(scan) == STAR) ? 0 : 1;
1064 save = reginput;
1065 no = regrepeat(OPERAND(scan));
1066 while (no >= min) {
1067 /* If it could work, try it. */
1068 if (nextch == '\0' || *reginput == nextch)
1069 if (regmatch(next))
1070 return(1);
1071 /* Couldn't or didn't -- back up. */
1072 no--;
1073 reginput = save + no;
1074 }
1075 return(0);
1076 }
1077 case END:
1078 return(1); /* Success! */
1079 default:
1080 regerror("memory corruption");
1081 return(0);
1082 }
1083
1084 scan = next;
1085 }
1086
1087 /*
1088 * We get here only if there's trouble -- normally "case END" is
1089 * the terminating point.
1090 */
1091 regerror("corrupted pointers");
1092 return(0);
1093 }
1094
1095 /*
1096 - regrepeat - repeatedly match something simple, report how many
1097 */
1098 static int
regrepeat(p)1099 regrepeat(p)
1100 char *p;
1101 {
1102 int count = 0;
1103 char *scan;
1104 char *opnd;
1105
1106 scan = reginput;
1107 opnd = OPERAND(p);
1108 switch (OP(p)) {
1109 case ANY:
1110 count = strlen(scan);
1111 scan += count;
1112 break;
1113 case EXACTLY:
1114 while (*opnd == *scan) {
1115 count++;
1116 scan++;
1117 }
1118 break;
1119 case ANYOF:
1120 while (*scan != '\0' && strchr(opnd, *scan) != NULL) {
1121 count++;
1122 scan++;
1123 }
1124 break;
1125 case ANYBUT:
1126 while (*scan != '\0' && strchr(opnd, *scan) == NULL) {
1127 count++;
1128 scan++;
1129 }
1130 break;
1131 default: /* Oh dear. Called inappropriately. */
1132 regerror("internal foulup");
1133 count = 0; /* Best compromise. */
1134 break;
1135 }
1136 reginput = scan;
1137
1138 return(count);
1139 }
1140
1141 /*
1142 - regnext - dig the "next" pointer out of a node
1143 */
1144 static char *
regnext(p)1145 regnext(p)
1146 char *p;
1147 {
1148 int offset;
1149
1150 if (p == ®dummy)
1151 return(NULL);
1152
1153 offset = NEXT(p);
1154 if (offset == 0)
1155 return(NULL);
1156
1157 if (OP(p) == BACK)
1158 return(p-offset);
1159 else
1160 return(p+offset);
1161 }
1162
1163 #ifdef DEBUG
1164
1165 /*
1166 - regdump - dump a regexp onto stdout in vaguely comprehensible form
1167 */
1168 void
regdump(r)1169 regdump(r)
1170 regexp *r;
1171 {
1172 char *s;
1173 char op = EXACTLY; /* Arbitrary non-END op. */
1174 char *next;
1175
1176
1177 s = r->program + 1;
1178 while (op != END) { /* While that wasn't END last time... */
1179 op = OP(s);
1180 printf("%2td%s", s-r->program, regprop(s)); /* Where, what. */
1181 next = regnext(s);
1182 if (next == NULL) /* Next ptr. */
1183 printf("(0)");
1184 else
1185 printf("(%td)", (s-r->program)+(next-s));
1186 s += 3;
1187 if (op == ANYOF || op == ANYBUT || op == EXACTLY) {
1188 /* Literal string, where present. */
1189 while (*s != '\0') {
1190 putchar(*s);
1191 s++;
1192 }
1193 s++;
1194 }
1195 putchar('\n');
1196 }
1197
1198 /* Header fields of interest. */
1199 if (r->regstart != '\0')
1200 printf("start `%c' ", r->regstart);
1201 if (r->reganch)
1202 printf("anchored ");
1203 if (r->regmust != NULL)
1204 printf("must have \"%s\"", r->regmust);
1205 printf("\n");
1206 }
1207
1208 /*
1209 - regprop - printable representation of opcode
1210 */
1211 static char *
regprop(op)1212 regprop(op)
1213 char *op;
1214 {
1215 char *p;
1216 static char buf[50];
1217
1218 (void)strncpy(buf, ":", sizeof(buf) - 1);
1219
1220 switch (OP(op)) {
1221 case BOL:
1222 p = "BOL";
1223 break;
1224 case EOL:
1225 p = "EOL";
1226 break;
1227 case ANY:
1228 p = "ANY";
1229 break;
1230 case ANYOF:
1231 p = "ANYOF";
1232 break;
1233 case ANYBUT:
1234 p = "ANYBUT";
1235 break;
1236 case BRANCH:
1237 p = "BRANCH";
1238 break;
1239 case EXACTLY:
1240 p = "EXACTLY";
1241 break;
1242 case NOTHING:
1243 p = "NOTHING";
1244 break;
1245 case BACK:
1246 p = "BACK";
1247 break;
1248 case END:
1249 p = "END";
1250 break;
1251 case OPEN+1:
1252 case OPEN+2:
1253 case OPEN+3:
1254 case OPEN+4:
1255 case OPEN+5:
1256 case OPEN+6:
1257 case OPEN+7:
1258 case OPEN+8:
1259 case OPEN+9:
1260 (void)snprintf(buf+strlen(buf), sizeof(buf) - strlen(buf),
1261 "OPEN%d", OP(op)-OPEN);
1262 p = NULL;
1263 break;
1264 case CLOSE+1:
1265 case CLOSE+2:
1266 case CLOSE+3:
1267 case CLOSE+4:
1268 case CLOSE+5:
1269 case CLOSE+6:
1270 case CLOSE+7:
1271 case CLOSE+8:
1272 case CLOSE+9:
1273 (void)snprintf(buf+strlen(buf), sizeof(buf) - strlen(buf),
1274 "CLOSE%d", OP(op)-CLOSE);
1275 p = NULL;
1276 break;
1277 case STAR:
1278 p = "STAR";
1279 break;
1280 case PLUS:
1281 p = "PLUS";
1282 break;
1283 case WORDA:
1284 p = "WORDA";
1285 break;
1286 case WORDZ:
1287 p = "WORDZ";
1288 break;
1289 default:
1290 p = NULL;
1291 regerror("corrupted opcode");
1292 break;
1293 }
1294 if (p != NULL)
1295 (void)strncat(buf, p, sizeof(buf) - strlen(buf) - 1);
1296 return(buf);
1297 }
1298 #endif
1299
1300 /*
1301 * The following is provided for those people who do not have strcspn() in
1302 * their C libraries. They should get off their butts and do something
1303 * about it; at least one public-domain implementation of those (highly
1304 * useful) string routines has been published on Usenet.
1305 */
1306 #ifdef STRCSPN
1307 /*
1308 * strcspn - find length of initial segment of s1 consisting entirely
1309 * of characters not from s2
1310 */
1311
1312 static int
strcspn(s1,s2)1313 strcspn(s1, s2)
1314 char *s1;
1315 char *s2;
1316 {
1317 char *scan1;
1318 char *scan2;
1319 int count;
1320
1321 count = 0;
1322 for (scan1 = s1; *scan1 != '\0'; scan1++) {
1323 for (scan2 = s2; *scan2 != '\0';) /* ++ moved down. */
1324 if (*scan1 == *scan2++)
1325 return(count);
1326 count++;
1327 }
1328 return(count);
1329 }
1330 #endif
1331