/[pcre]/code/branches/pcre16/pcre_study.c
ViewVC logotype

Contents of /code/branches/pcre16/pcre_study.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 688 - (show annotations) (download)
Fri Sep 9 09:35:48 2011 UTC (2 years, 10 months ago) by ph10
Original Path: code/trunk/pcre_study.c
File MIME type: text/plain
File size: 38764 byte(s)
Tidy up interface from pcre_study() to JIT, and ignore a minimum length of 0.

1 /*************************************************
2 * Perl-Compatible Regular Expressions *
3 *************************************************/
4
5 /* PCRE is a library of functions to support regular expressions whose syntax
6 and semantics are as close as possible to those of the Perl 5 language.
7
8 Written by Philip Hazel
9 Copyright (c) 1997-2010 University of Cambridge
10
11 -----------------------------------------------------------------------------
12 Redistribution and use in source and binary forms, with or without
13 modification, are permitted provided that the following conditions are met:
14
15 * Redistributions of source code must retain the above copyright notice,
16 this list of conditions and the following disclaimer.
17
18 * Redistributions in binary form must reproduce the above copyright
19 notice, this list of conditions and the following disclaimer in the
20 documentation and/or other materials provided with the distribution.
21
22 * Neither the name of the University of Cambridge nor the names of its
23 contributors may be used to endorse or promote products derived from
24 this software without specific prior written permission.
25
26 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
27 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
30 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36 POSSIBILITY OF SUCH DAMAGE.
37 -----------------------------------------------------------------------------
38 */
39
40
41 /* This module contains the external function pcre_study(), along with local
42 supporting functions. */
43
44
45 #ifdef HAVE_CONFIG_H
46 #include "config.h"
47 #endif
48
49 #include "pcre_internal.h"
50
51 #define SET_BIT(c) start_bits[c/8] |= (1 << (c&7))
52
53 /* Returns from set_start_bits() */
54
55 enum { SSB_FAIL, SSB_DONE, SSB_CONTINUE, SSB_UNKNOWN };
56
57
58
59 /*************************************************
60 * Find the minimum subject length for a group *
61 *************************************************/
62
63 /* Scan a parenthesized group and compute the minimum length of subject that
64 is needed to match it. This is a lower bound; it does not mean there is a
65 string of that length that matches. In UTF8 mode, the result is in characters
66 rather than bytes.
67
68 Arguments:
69 code pointer to start of group (the bracket)
70 startcode pointer to start of the whole pattern
71 options the compiling options
72 had_accept pointer to flag for (*ACCEPT) encountered
73 int RECURSE depth
74
75 Returns: the minimum length
76 -1 if \C was encountered
77 -2 internal error (missing capturing bracket)
78 -3 internal error (opcode not listed)
79 */
80
81 static int
82 find_minlength(const uschar *code, const uschar *startcode, int options,
83 BOOL *had_accept_ptr, int recurse_depth)
84 {
85 int length = -1;
86 BOOL utf8 = (options & PCRE_UTF8) != 0;
87 BOOL had_recurse = FALSE;
88 register int branchlength = 0;
89 register uschar *cc = (uschar *)code + 1 + LINK_SIZE;
90
91 if (*code == OP_CBRA || *code == OP_SCBRA ||
92 *code == OP_CBRAPOS || *code == OP_SCBRAPOS) cc += 2;
93
94 /* Scan along the opcodes for this branch. If we get to the end of the
95 branch, check the length against that of the other branches. */
96
97 for (;;)
98 {
99 int d, min;
100 uschar *cs, *ce;
101 register int op = *cc;
102
103 switch (op)
104 {
105 case OP_COND:
106 case OP_SCOND:
107
108 /* If there is only one branch in a condition, the implied branch has zero
109 length, so we don't add anything. This covers the DEFINE "condition"
110 automatically. */
111
112 cs = cc + GET(cc, 1);
113 if (*cs != OP_ALT)
114 {
115 cc = cs + 1 + LINK_SIZE;
116 break;
117 }
118
119 /* Otherwise we can fall through and treat it the same as any other
120 subpattern. */
121
122 case OP_CBRA:
123 case OP_SCBRA:
124 case OP_BRA:
125 case OP_SBRA:
126 case OP_CBRAPOS:
127 case OP_SCBRAPOS:
128 case OP_BRAPOS:
129 case OP_SBRAPOS:
130 case OP_ONCE:
131 d = find_minlength(cc, startcode, options, had_accept_ptr, recurse_depth);
132 if (d < 0) return d;
133 branchlength += d;
134 if (*had_accept_ptr) return branchlength;
135 do cc += GET(cc, 1); while (*cc == OP_ALT);
136 cc += 1 + LINK_SIZE;
137 break;
138
139 /* Reached end of a branch; if it's a ket it is the end of a nested
140 call. If it's ALT it is an alternation in a nested call. If it is END it's
141 the end of the outer call. All can be handled by the same code. If it is
142 ACCEPT, it is essentially the same as END, but we set a flag so that
143 counting stops. */
144
145 case OP_ACCEPT:
146 case OP_ASSERT_ACCEPT:
147 *had_accept_ptr = TRUE;
148 /* Fall through */
149 case OP_ALT:
150 case OP_KET:
151 case OP_KETRMAX:
152 case OP_KETRMIN:
153 case OP_KETRPOS:
154 case OP_END:
155 if (length < 0 || (!had_recurse && branchlength < length))
156 length = branchlength;
157 if (op != OP_ALT) return length;
158 cc += 1 + LINK_SIZE;
159 branchlength = 0;
160 had_recurse = FALSE;
161 break;
162
163 /* Skip over assertive subpatterns */
164
165 case OP_ASSERT:
166 case OP_ASSERT_NOT:
167 case OP_ASSERTBACK:
168 case OP_ASSERTBACK_NOT:
169 do cc += GET(cc, 1); while (*cc == OP_ALT);
170 /* Fall through */
171
172 /* Skip over things that don't match chars */
173
174 case OP_REVERSE:
175 case OP_CREF:
176 case OP_NCREF:
177 case OP_RREF:
178 case OP_NRREF:
179 case OP_DEF:
180 case OP_CALLOUT:
181 case OP_SOD:
182 case OP_SOM:
183 case OP_EOD:
184 case OP_EODN:
185 case OP_CIRC:
186 case OP_CIRCM:
187 case OP_DOLL:
188 case OP_DOLLM:
189 case OP_NOT_WORD_BOUNDARY:
190 case OP_WORD_BOUNDARY:
191 cc += _pcre_OP_lengths[*cc];
192 break;
193
194 /* Skip over a subpattern that has a {0} or {0,x} quantifier */
195
196 case OP_BRAZERO:
197 case OP_BRAMINZERO:
198 case OP_BRAPOSZERO:
199 case OP_SKIPZERO:
200 cc += _pcre_OP_lengths[*cc];
201 do cc += GET(cc, 1); while (*cc == OP_ALT);
202 cc += 1 + LINK_SIZE;
203 break;
204
205 /* Handle literal characters and + repetitions */
206
207 case OP_CHAR:
208 case OP_CHARI:
209 case OP_NOT:
210 case OP_NOTI:
211 case OP_PLUS:
212 case OP_PLUSI:
213 case OP_MINPLUS:
214 case OP_MINPLUSI:
215 case OP_POSPLUS:
216 case OP_POSPLUSI:
217 case OP_NOTPLUS:
218 case OP_NOTPLUSI:
219 case OP_NOTMINPLUS:
220 case OP_NOTMINPLUSI:
221 case OP_NOTPOSPLUS:
222 case OP_NOTPOSPLUSI:
223 branchlength++;
224 cc += 2;
225 #ifdef SUPPORT_UTF8
226 if (utf8 && cc[-1] >= 0xc0) cc += _pcre_utf8_table4[cc[-1] & 0x3f];
227 #endif
228 break;
229
230 case OP_TYPEPLUS:
231 case OP_TYPEMINPLUS:
232 case OP_TYPEPOSPLUS:
233 branchlength++;
234 cc += (cc[1] == OP_PROP || cc[1] == OP_NOTPROP)? 4 : 2;
235 break;
236
237 /* Handle exact repetitions. The count is already in characters, but we
238 need to skip over a multibyte character in UTF8 mode. */
239
240 case OP_EXACT:
241 case OP_EXACTI:
242 case OP_NOTEXACT:
243 case OP_NOTEXACTI:
244 branchlength += GET2(cc,1);
245 cc += 4;
246 #ifdef SUPPORT_UTF8
247 if (utf8 && cc[-1] >= 0xc0) cc += _pcre_utf8_table4[cc[-1] & 0x3f];
248 #endif
249 break;
250
251 case OP_TYPEEXACT:
252 branchlength += GET2(cc,1);
253 cc += (cc[3] == OP_PROP || cc[3] == OP_NOTPROP)? 6 : 4;
254 break;
255
256 /* Handle single-char non-literal matchers */
257
258 case OP_PROP:
259 case OP_NOTPROP:
260 cc += 2;
261 /* Fall through */
262
263 case OP_NOT_DIGIT:
264 case OP_DIGIT:
265 case OP_NOT_WHITESPACE:
266 case OP_WHITESPACE:
267 case OP_NOT_WORDCHAR:
268 case OP_WORDCHAR:
269 case OP_ANY:
270 case OP_ALLANY:
271 case OP_EXTUNI:
272 case OP_HSPACE:
273 case OP_NOT_HSPACE:
274 case OP_VSPACE:
275 case OP_NOT_VSPACE:
276 branchlength++;
277 cc++;
278 break;
279
280 /* "Any newline" might match two characters, but it also might match just
281 one. */
282
283 case OP_ANYNL:
284 branchlength += 1;
285 cc++;
286 break;
287
288 /* The single-byte matcher means we can't proceed in UTF-8 mode */
289
290 case OP_ANYBYTE:
291 #ifdef SUPPORT_UTF8
292 if (utf8) return -1;
293 #endif
294 branchlength++;
295 cc++;
296 break;
297
298 /* For repeated character types, we have to test for \p and \P, which have
299 an extra two bytes of parameters. */
300
301 case OP_TYPESTAR:
302 case OP_TYPEMINSTAR:
303 case OP_TYPEQUERY:
304 case OP_TYPEMINQUERY:
305 case OP_TYPEPOSSTAR:
306 case OP_TYPEPOSQUERY:
307 if (cc[1] == OP_PROP || cc[1] == OP_NOTPROP) cc += 2;
308 cc += _pcre_OP_lengths[op];
309 break;
310
311 case OP_TYPEUPTO:
312 case OP_TYPEMINUPTO:
313 case OP_TYPEPOSUPTO:
314 if (cc[3] == OP_PROP || cc[3] == OP_NOTPROP) cc += 2;
315 cc += _pcre_OP_lengths[op];
316 break;
317
318 /* Check a class for variable quantification */
319
320 #ifdef SUPPORT_UTF8
321 case OP_XCLASS:
322 cc += GET(cc, 1) - 33;
323 /* Fall through */
324 #endif
325
326 case OP_CLASS:
327 case OP_NCLASS:
328 cc += 33;
329
330 switch (*cc)
331 {
332 case OP_CRPLUS:
333 case OP_CRMINPLUS:
334 branchlength++;
335 /* Fall through */
336
337 case OP_CRSTAR:
338 case OP_CRMINSTAR:
339 case OP_CRQUERY:
340 case OP_CRMINQUERY:
341 cc++;
342 break;
343
344 case OP_CRRANGE:
345 case OP_CRMINRANGE:
346 branchlength += GET2(cc,1);
347 cc += 5;
348 break;
349
350 default:
351 branchlength++;
352 break;
353 }
354 break;
355
356 /* Backreferences and subroutine calls are treated in the same way: we find
357 the minimum length for the subpattern. A recursion, however, causes an
358 a flag to be set that causes the length of this branch to be ignored. The
359 logic is that a recursion can only make sense if there is another
360 alternation that stops the recursing. That will provide the minimum length
361 (when no recursion happens). A backreference within the group that it is
362 referencing behaves in the same way.
363
364 If PCRE_JAVASCRIPT_COMPAT is set, a backreference to an unset bracket
365 matches an empty string (by default it causes a matching failure), so in
366 that case we must set the minimum length to zero. */
367
368 case OP_REF:
369 case OP_REFI:
370 if ((options & PCRE_JAVASCRIPT_COMPAT) == 0)
371 {
372 ce = cs = (uschar *)_pcre_find_bracket(startcode, utf8, GET2(cc, 1));
373 if (cs == NULL) return -2;
374 do ce += GET(ce, 1); while (*ce == OP_ALT);
375 if (cc > cs && cc < ce)
376 {
377 d = 0;
378 had_recurse = TRUE;
379 }
380 else
381 {
382 d = find_minlength(cs, startcode, options, had_accept_ptr,
383 recurse_depth);
384 *had_accept_ptr = FALSE;
385 }
386 }
387 else d = 0;
388 cc += 3;
389
390 /* Handle repeated back references */
391
392 switch (*cc)
393 {
394 case OP_CRSTAR:
395 case OP_CRMINSTAR:
396 case OP_CRQUERY:
397 case OP_CRMINQUERY:
398 min = 0;
399 cc++;
400 break;
401
402 case OP_CRPLUS:
403 case OP_CRMINPLUS:
404 min = 1;
405 cc++;
406 break;
407
408 case OP_CRRANGE:
409 case OP_CRMINRANGE:
410 min = GET2(cc, 1);
411 cc += 5;
412 break;
413
414 default:
415 min = 1;
416 break;
417 }
418
419 branchlength += min * d;
420 break;
421
422 /* We can easily detect direct recursion, but not mutual recursion. This is
423 caught by a recursion depth count. */
424
425 case OP_RECURSE:
426 cs = ce = (uschar *)startcode + GET(cc, 1);
427 if (cs == NULL) return -2;
428 do ce += GET(ce, 1); while (*ce == OP_ALT);
429 if ((cc > cs && cc < ce) || recurse_depth > 10)
430 had_recurse = TRUE;
431 else
432 {
433 branchlength += find_minlength(cs, startcode, options, had_accept_ptr,
434 recurse_depth + 1);
435 *had_accept_ptr = FALSE;
436 }
437 cc += 1 + LINK_SIZE;
438 break;
439
440 /* Anything else does not or need not match a character. We can get the
441 item's length from the table, but for those that can match zero occurrences
442 of a character, we must take special action for UTF-8 characters. As it
443 happens, the "NOT" versions of these opcodes are used at present only for
444 ASCII characters, so they could be omitted from this list. However, in
445 future that may change, so we include them here so as not to leave a
446 gotcha for a future maintainer. */
447
448 case OP_UPTO:
449 case OP_UPTOI:
450 case OP_NOTUPTO:
451 case OP_NOTUPTOI:
452 case OP_MINUPTO:
453 case OP_MINUPTOI:
454 case OP_NOTMINUPTO:
455 case OP_NOTMINUPTOI:
456 case OP_POSUPTO:
457 case OP_POSUPTOI:
458 case OP_NOTPOSUPTO:
459 case OP_NOTPOSUPTOI:
460
461 case OP_STAR:
462 case OP_STARI:
463 case OP_NOTSTAR:
464 case OP_NOTSTARI:
465 case OP_MINSTAR:
466 case OP_MINSTARI:
467 case OP_NOTMINSTAR:
468 case OP_NOTMINSTARI:
469 case OP_POSSTAR:
470 case OP_POSSTARI:
471 case OP_NOTPOSSTAR:
472 case OP_NOTPOSSTARI:
473
474 case OP_QUERY:
475 case OP_QUERYI:
476 case OP_NOTQUERY:
477 case OP_NOTQUERYI:
478 case OP_MINQUERY:
479 case OP_MINQUERYI:
480 case OP_NOTMINQUERY:
481 case OP_NOTMINQUERYI:
482 case OP_POSQUERY:
483 case OP_POSQUERYI:
484 case OP_NOTPOSQUERY:
485 case OP_NOTPOSQUERYI:
486
487 cc += _pcre_OP_lengths[op];
488 #ifdef SUPPORT_UTF8
489 if (utf8 && cc[-1] >= 0xc0) cc += _pcre_utf8_table4[cc[-1] & 0x3f];
490 #endif
491 break;
492
493 /* Skip these, but we need to add in the name length. */
494
495 case OP_MARK:
496 case OP_PRUNE_ARG:
497 case OP_SKIP_ARG:
498 cc += _pcre_OP_lengths[op] + cc[1];
499 break;
500
501 case OP_THEN_ARG:
502 cc += _pcre_OP_lengths[op] + cc[1+LINK_SIZE];
503 break;
504
505 /* The remaining opcodes are just skipped over. */
506
507 case OP_CLOSE:
508 case OP_COMMIT:
509 case OP_FAIL:
510 case OP_PRUNE:
511 case OP_SET_SOM:
512 case OP_SKIP:
513 case OP_THEN:
514 cc += _pcre_OP_lengths[op];
515 break;
516
517 /* This should not occur: we list all opcodes explicitly so that when
518 new ones get added they are properly considered. */
519
520 default:
521 return -3;
522 }
523 }
524 /* Control never gets here */
525 }
526
527
528
529 /*************************************************
530 * Set a bit and maybe its alternate case *
531 *************************************************/
532
533 /* Given a character, set its first byte's bit in the table, and also the
534 corresponding bit for the other version of a letter if we are caseless. In
535 UTF-8 mode, for characters greater than 127, we can only do the caseless thing
536 when Unicode property support is available.
537
538 Arguments:
539 start_bits points to the bit map
540 p points to the character
541 caseless the caseless flag
542 cd the block with char table pointers
543 utf8 TRUE for UTF-8 mode
544
545 Returns: pointer after the character
546 */
547
548 static const uschar *
549 set_table_bit(uschar *start_bits, const uschar *p, BOOL caseless,
550 compile_data *cd, BOOL utf8)
551 {
552 unsigned int c = *p;
553
554 SET_BIT(c);
555
556 #ifdef SUPPORT_UTF8
557 if (utf8 && c > 127)
558 {
559 GETCHARINC(c, p);
560 #ifdef SUPPORT_UCP
561 if (caseless)
562 {
563 uschar buff[8];
564 c = UCD_OTHERCASE(c);
565 (void)_pcre_ord2utf8(c, buff);
566 SET_BIT(buff[0]);
567 }
568 #endif
569 return p;
570 }
571 #endif
572
573 /* Not UTF-8 mode, or character is less than 127. */
574
575 if (caseless && (cd->ctypes[c] & ctype_letter) != 0) SET_BIT(cd->fcc[c]);
576 return p + 1;
577 }
578
579
580
581 /*************************************************
582 * Set bits for a positive character type *
583 *************************************************/
584
585 /* This function sets starting bits for a character type. In UTF-8 mode, we can
586 only do a direct setting for bytes less than 128, as otherwise there can be
587 confusion with bytes in the middle of UTF-8 characters. In a "traditional"
588 environment, the tables will only recognize ASCII characters anyway, but in at
589 least one Windows environment, some higher bytes bits were set in the tables.
590 So we deal with that case by considering the UTF-8 encoding.
591
592 Arguments:
593 start_bits the starting bitmap
594 cbit type the type of character wanted
595 table_limit 32 for non-UTF-8; 16 for UTF-8
596 cd the block with char table pointers
597
598 Returns: nothing
599 */
600
601 static void
602 set_type_bits(uschar *start_bits, int cbit_type, int table_limit,
603 compile_data *cd)
604 {
605 register int c;
606 for (c = 0; c < table_limit; c++) start_bits[c] |= cd->cbits[c+cbit_type];
607 if (table_limit == 32) return;
608 for (c = 128; c < 256; c++)
609 {
610 if ((cd->cbits[c/8] & (1 << (c&7))) != 0)
611 {
612 uschar buff[8];
613 (void)_pcre_ord2utf8(c, buff);
614 SET_BIT(buff[0]);
615 }
616 }
617 }
618
619
620 /*************************************************
621 * Set bits for a negative character type *
622 *************************************************/
623
624 /* This function sets starting bits for a negative character type such as \D.
625 In UTF-8 mode, we can only do a direct setting for bytes less than 128, as
626 otherwise there can be confusion with bytes in the middle of UTF-8 characters.
627 Unlike in the positive case, where we can set appropriate starting bits for
628 specific high-valued UTF-8 characters, in this case we have to set the bits for
629 all high-valued characters. The lowest is 0xc2, but we overkill by starting at
630 0xc0 (192) for simplicity.
631
632 Arguments:
633 start_bits the starting bitmap
634 cbit type the type of character wanted
635 table_limit 32 for non-UTF-8; 16 for UTF-8
636 cd the block with char table pointers
637
638 Returns: nothing
639 */
640
641 static void
642 set_nottype_bits(uschar *start_bits, int cbit_type, int table_limit,
643 compile_data *cd)
644 {
645 register int c;
646 for (c = 0; c < table_limit; c++) start_bits[c] |= ~cd->cbits[c+cbit_type];
647 if (table_limit != 32) for (c = 24; c < 32; c++) start_bits[c] = 0xff;
648 }
649
650
651
652 /*************************************************
653 * Create bitmap of starting bytes *
654 *************************************************/
655
656 /* This function scans a compiled unanchored expression recursively and
657 attempts to build a bitmap of the set of possible starting bytes. As time goes
658 by, we may be able to get more clever at doing this. The SSB_CONTINUE return is
659 useful for parenthesized groups in patterns such as (a*)b where the group
660 provides some optional starting bytes but scanning must continue at the outer
661 level to find at least one mandatory byte. At the outermost level, this
662 function fails unless the result is SSB_DONE.
663
664 Arguments:
665 code points to an expression
666 start_bits points to a 32-byte table, initialized to 0
667 utf8 TRUE if in UTF-8 mode
668 cd the block with char table pointers
669
670 Returns: SSB_FAIL => Failed to find any starting bytes
671 SSB_DONE => Found mandatory starting bytes
672 SSB_CONTINUE => Found optional starting bytes
673 SSB_UNKNOWN => Hit an unrecognized opcode
674 */
675
676 static int
677 set_start_bits(const uschar *code, uschar *start_bits, BOOL utf8,
678 compile_data *cd)
679 {
680 register int c;
681 int yield = SSB_DONE;
682 int table_limit = utf8? 16:32;
683
684 #if 0
685 /* ========================================================================= */
686 /* The following comment and code was inserted in January 1999. In May 2006,
687 when it was observed to cause compiler warnings about unused values, I took it
688 out again. If anybody is still using OS/2, they will have to put it back
689 manually. */
690
691 /* This next statement and the later reference to dummy are here in order to
692 trick the optimizer of the IBM C compiler for OS/2 into generating correct
693 code. Apparently IBM isn't going to fix the problem, and we would rather not
694 disable optimization (in this module it actually makes a big difference, and
695 the pcre module can use all the optimization it can get). */
696
697 volatile int dummy;
698 /* ========================================================================= */
699 #endif
700
701 do
702 {
703 BOOL try_next = TRUE;
704 const uschar *tcode = code + 1 + LINK_SIZE;
705
706 if (*code == OP_CBRA || *code == OP_SCBRA ||
707 *code == OP_CBRAPOS || *code == OP_SCBRAPOS) tcode += 2;
708
709 while (try_next) /* Loop for items in this branch */
710 {
711 int rc;
712
713 switch(*tcode)
714 {
715 /* If we reach something we don't understand, it means a new opcode has
716 been created that hasn't been added to this code. Hopefully this problem
717 will be discovered during testing. */
718
719 default:
720 return SSB_UNKNOWN;
721
722 /* Fail for a valid opcode that implies no starting bits. */
723
724 case OP_ACCEPT:
725 case OP_ASSERT_ACCEPT:
726 case OP_ALLANY:
727 case OP_ANY:
728 case OP_ANYBYTE:
729 case OP_CIRC:
730 case OP_CIRCM:
731 case OP_CLOSE:
732 case OP_COMMIT:
733 case OP_COND:
734 case OP_CREF:
735 case OP_DEF:
736 case OP_DOLL:
737 case OP_DOLLM:
738 case OP_END:
739 case OP_EOD:
740 case OP_EODN:
741 case OP_EXTUNI:
742 case OP_FAIL:
743 case OP_MARK:
744 case OP_NCREF:
745 case OP_NOT:
746 case OP_NOTEXACT:
747 case OP_NOTEXACTI:
748 case OP_NOTI:
749 case OP_NOTMINPLUS:
750 case OP_NOTMINPLUSI:
751 case OP_NOTMINQUERY:
752 case OP_NOTMINQUERYI:
753 case OP_NOTMINSTAR:
754 case OP_NOTMINSTARI:
755 case OP_NOTMINUPTO:
756 case OP_NOTMINUPTOI:
757 case OP_NOTPLUS:
758 case OP_NOTPLUSI:
759 case OP_NOTPOSPLUS:
760 case OP_NOTPOSPLUSI:
761 case OP_NOTPOSQUERY:
762 case OP_NOTPOSQUERYI:
763 case OP_NOTPOSSTAR:
764 case OP_NOTPOSSTARI:
765 case OP_NOTPOSUPTO:
766 case OP_NOTPOSUPTOI:
767 case OP_NOTPROP:
768 case OP_NOTQUERY:
769 case OP_NOTQUERYI:
770 case OP_NOTSTAR:
771 case OP_NOTSTARI:
772 case OP_NOTUPTO:
773 case OP_NOTUPTOI:
774 case OP_NOT_HSPACE:
775 case OP_NOT_VSPACE:
776 case OP_NRREF:
777 case OP_PROP:
778 case OP_PRUNE:
779 case OP_PRUNE_ARG:
780 case OP_RECURSE:
781 case OP_REF:
782 case OP_REFI:
783 case OP_REVERSE:
784 case OP_RREF:
785 case OP_SCOND:
786 case OP_SET_SOM:
787 case OP_SKIP:
788 case OP_SKIP_ARG:
789 case OP_SOD:
790 case OP_SOM:
791 case OP_THEN:
792 case OP_THEN_ARG:
793 case OP_XCLASS:
794 return SSB_FAIL;
795
796 /* We can ignore word boundary tests. */
797
798 case OP_WORD_BOUNDARY:
799 case OP_NOT_WORD_BOUNDARY:
800 tcode++;
801 break;
802
803 /* If we hit a bracket or a positive lookahead assertion, recurse to set
804 bits from within the subpattern. If it can't find anything, we have to
805 give up. If it finds some mandatory character(s), we are done for this
806 branch. Otherwise, carry on scanning after the subpattern. */
807
808 case OP_BRA:
809 case OP_SBRA:
810 case OP_CBRA:
811 case OP_SCBRA:
812 case OP_BRAPOS:
813 case OP_SBRAPOS:
814 case OP_CBRAPOS:
815 case OP_SCBRAPOS:
816 case OP_ONCE:
817 case OP_ASSERT:
818 rc = set_start_bits(tcode, start_bits, utf8, cd);
819 if (rc == SSB_FAIL || rc == SSB_UNKNOWN) return rc;
820 if (rc == SSB_DONE) try_next = FALSE; else
821 {
822 do tcode += GET(tcode, 1); while (*tcode == OP_ALT);
823 tcode += 1 + LINK_SIZE;
824 }
825 break;
826
827 /* If we hit ALT or KET, it means we haven't found anything mandatory in
828 this branch, though we might have found something optional. For ALT, we
829 continue with the next alternative, but we have to arrange that the final
830 result from subpattern is SSB_CONTINUE rather than SSB_DONE. For KET,
831 return SSB_CONTINUE: if this is the top level, that indicates failure,
832 but after a nested subpattern, it causes scanning to continue. */
833
834 case OP_ALT:
835 yield = SSB_CONTINUE;
836 try_next = FALSE;
837 break;
838
839 case OP_KET:
840 case OP_KETRMAX:
841 case OP_KETRMIN:
842 case OP_KETRPOS:
843 return SSB_CONTINUE;
844
845 /* Skip over callout */
846
847 case OP_CALLOUT:
848 tcode += 2 + 2*LINK_SIZE;
849 break;
850
851 /* Skip over lookbehind and negative lookahead assertions */
852
853 case OP_ASSERT_NOT:
854 case OP_ASSERTBACK:
855 case OP_ASSERTBACK_NOT:
856 do tcode += GET(tcode, 1); while (*tcode == OP_ALT);
857 tcode += 1 + LINK_SIZE;
858 break;
859
860 /* BRAZERO does the bracket, but carries on. */
861
862 case OP_BRAZERO:
863 case OP_BRAMINZERO:
864 case OP_BRAPOSZERO:
865 rc = set_start_bits(++tcode, start_bits, utf8, cd);
866 if (rc == SSB_FAIL || rc == SSB_UNKNOWN) return rc;
867 /* =========================================================================
868 See the comment at the head of this function concerning the next line,
869 which was an old fudge for the benefit of OS/2.
870 dummy = 1;
871 ========================================================================= */
872 do tcode += GET(tcode,1); while (*tcode == OP_ALT);
873 tcode += 1 + LINK_SIZE;
874 break;
875
876 /* SKIPZERO skips the bracket. */
877
878 case OP_SKIPZERO:
879 tcode++;
880 do tcode += GET(tcode,1); while (*tcode == OP_ALT);
881 tcode += 1 + LINK_SIZE;
882 break;
883
884 /* Single-char * or ? sets the bit and tries the next item */
885
886 case OP_STAR:
887 case OP_MINSTAR:
888 case OP_POSSTAR:
889 case OP_QUERY:
890 case OP_MINQUERY:
891 case OP_POSQUERY:
892 tcode = set_table_bit(start_bits, tcode + 1, FALSE, cd, utf8);
893 break;
894
895 case OP_STARI:
896 case OP_MINSTARI:
897 case OP_POSSTARI:
898 case OP_QUERYI:
899 case OP_MINQUERYI:
900 case OP_POSQUERYI:
901 tcode = set_table_bit(start_bits, tcode + 1, TRUE, cd, utf8);
902 break;
903
904 /* Single-char upto sets the bit and tries the next */
905
906 case OP_UPTO:
907 case OP_MINUPTO:
908 case OP_POSUPTO:
909 tcode = set_table_bit(start_bits, tcode + 3, FALSE, cd, utf8);
910 break;
911
912 case OP_UPTOI:
913 case OP_MINUPTOI:
914 case OP_POSUPTOI:
915 tcode = set_table_bit(start_bits, tcode + 3, TRUE, cd, utf8);
916 break;
917
918 /* At least one single char sets the bit and stops */
919
920 case OP_EXACT:
921 tcode += 2;
922 /* Fall through */
923 case OP_CHAR:
924 case OP_PLUS:
925 case OP_MINPLUS:
926 case OP_POSPLUS:
927 (void)set_table_bit(start_bits, tcode + 1, FALSE, cd, utf8);
928 try_next = FALSE;
929 break;
930
931 case OP_EXACTI:
932 tcode += 2;
933 /* Fall through */
934 case OP_CHARI:
935 case OP_PLUSI:
936 case OP_MINPLUSI:
937 case OP_POSPLUSI:
938 (void)set_table_bit(start_bits, tcode + 1, TRUE, cd, utf8);
939 try_next = FALSE;
940 break;
941
942 /* Special spacing and line-terminating items. These recognize specific
943 lists of characters. The difference between VSPACE and ANYNL is that the
944 latter can match the two-character CRLF sequence, but that is not
945 relevant for finding the first character, so their code here is
946 identical. */
947
948 case OP_HSPACE:
949 SET_BIT(0x09);
950 SET_BIT(0x20);
951 if (utf8)
952 {
953 SET_BIT(0xC2); /* For U+00A0 */
954 SET_BIT(0xE1); /* For U+1680, U+180E */
955 SET_BIT(0xE2); /* For U+2000 - U+200A, U+202F, U+205F */
956 SET_BIT(0xE3); /* For U+3000 */
957 }
958 else SET_BIT(0xA0);
959 try_next = FALSE;
960 break;
961
962 case OP_ANYNL:
963 case OP_VSPACE:
964 SET_BIT(0x0A);
965 SET_BIT(0x0B);
966 SET_BIT(0x0C);
967 SET_BIT(0x0D);
968 if (utf8)
969 {
970 SET_BIT(0xC2); /* For U+0085 */
971 SET_BIT(0xE2); /* For U+2028, U+2029 */
972 }
973 else SET_BIT(0x85);
974 try_next = FALSE;
975 break;
976
977 /* Single character types set the bits and stop. Note that if PCRE_UCP
978 is set, we do not see these op codes because \d etc are converted to
979 properties. Therefore, these apply in the case when only characters less
980 than 256 are recognized to match the types. */
981
982 case OP_NOT_DIGIT:
983 set_nottype_bits(start_bits, cbit_digit, table_limit, cd);
984 try_next = FALSE;
985 break;
986
987 case OP_DIGIT:
988 set_type_bits(start_bits, cbit_digit, table_limit, cd);
989 try_next = FALSE;
990 break;
991
992 /* The cbit_space table has vertical tab as whitespace; we have to
993 ensure it is set as not whitespace. */
994
995 case OP_NOT_WHITESPACE:
996 set_nottype_bits(start_bits, cbit_space, table_limit, cd);
997 start_bits[1] |= 0x08;
998 try_next = FALSE;
999 break;
1000
1001 /* The cbit_space table has vertical tab as whitespace; we have to
1002 not set it from the table. */
1003
1004 case OP_WHITESPACE:
1005 c = start_bits[1]; /* Save in case it was already set */
1006 set_type_bits(start_bits, cbit_space, table_limit, cd);
1007 start_bits[1] = (start_bits[1] & ~0x08) | c;
1008 try_next = FALSE;
1009 break;
1010
1011 case OP_NOT_WORDCHAR:
1012 set_nottype_bits(start_bits, cbit_word, table_limit, cd);
1013 try_next = FALSE;
1014 break;
1015
1016 case OP_WORDCHAR:
1017 set_type_bits(start_bits, cbit_word, table_limit, cd);
1018 try_next = FALSE;
1019 break;
1020
1021 /* One or more character type fudges the pointer and restarts, knowing
1022 it will hit a single character type and stop there. */
1023
1024 case OP_TYPEPLUS:
1025 case OP_TYPEMINPLUS:
1026 case OP_TYPEPOSPLUS:
1027 tcode++;
1028 break;
1029
1030 case OP_TYPEEXACT:
1031 tcode += 3;
1032 break;
1033
1034 /* Zero or more repeats of character types set the bits and then
1035 try again. */
1036
1037 case OP_TYPEUPTO:
1038 case OP_TYPEMINUPTO:
1039 case OP_TYPEPOSUPTO:
1040 tcode += 2; /* Fall through */
1041
1042 case OP_TYPESTAR:
1043 case OP_TYPEMINSTAR:
1044 case OP_TYPEPOSSTAR:
1045 case OP_TYPEQUERY:
1046 case OP_TYPEMINQUERY:
1047 case OP_TYPEPOSQUERY:
1048 switch(tcode[1])
1049 {
1050 default:
1051 case OP_ANY:
1052 case OP_ALLANY:
1053 return SSB_FAIL;
1054
1055 case OP_HSPACE:
1056 SET_BIT(0x09);
1057 SET_BIT(0x20);
1058 if (utf8)
1059 {
1060 SET_BIT(0xC2); /* For U+00A0 */
1061 SET_BIT(0xE1); /* For U+1680, U+180E */
1062 SET_BIT(0xE2); /* For U+2000 - U+200A, U+202F, U+205F */
1063 SET_BIT(0xE3); /* For U+3000 */
1064 }
1065 else SET_BIT(0xA0);
1066 break;
1067
1068 case OP_ANYNL:
1069 case OP_VSPACE:
1070 SET_BIT(0x0A);
1071 SET_BIT(0x0B);
1072 SET_BIT(0x0C);
1073 SET_BIT(0x0D);
1074 if (utf8)
1075 {
1076 SET_BIT(0xC2); /* For U+0085 */
1077 SET_BIT(0xE2); /* For U+2028, U+2029 */
1078 }
1079 else SET_BIT(0x85);
1080 break;
1081
1082 case OP_NOT_DIGIT:
1083 set_nottype_bits(start_bits, cbit_digit, table_limit, cd);
1084 break;
1085
1086 case OP_DIGIT:
1087 set_type_bits(start_bits, cbit_digit, table_limit, cd);
1088 break;
1089
1090 /* The cbit_space table has vertical tab as whitespace; we have to
1091 ensure it gets set as not whitespace. */
1092
1093 case OP_NOT_WHITESPACE:
1094 set_nottype_bits(start_bits, cbit_space, table_limit, cd);
1095 start_bits[1] |= 0x08;
1096 break;
1097
1098 /* The cbit_space table has vertical tab as whitespace; we have to
1099 avoid setting it. */
1100
1101 case OP_WHITESPACE:
1102 c = start_bits[1]; /* Save in case it was already set */
1103 set_type_bits(start_bits, cbit_space, table_limit, cd);
1104 start_bits[1] = (start_bits[1] & ~0x08) | c;
1105 break;
1106
1107 case OP_NOT_WORDCHAR:
1108 set_nottype_bits(start_bits, cbit_word, table_limit, cd);
1109 break;
1110
1111 case OP_WORDCHAR:
1112 set_type_bits(start_bits, cbit_word, table_limit, cd);
1113 break;
1114 }
1115
1116 tcode += 2;
1117 break;
1118
1119 /* Character class where all the information is in a bit map: set the
1120 bits and either carry on or not, according to the repeat count. If it was
1121 a negative class, and we are operating with UTF-8 characters, any byte
1122 with a value >= 0xc4 is a potentially valid starter because it starts a
1123 character with a value > 255. */
1124
1125 case OP_NCLASS:
1126 #ifdef SUPPORT_UTF8
1127 if (utf8)
1128 {
1129 start_bits[24] |= 0xf0; /* Bits for 0xc4 - 0xc8 */
1130 memset(start_bits+25, 0xff, 7); /* Bits for 0xc9 - 0xff */
1131 }
1132 #endif
1133 /* Fall through */
1134
1135 case OP_CLASS:
1136 {
1137 tcode++;
1138
1139 /* In UTF-8 mode, the bits in a bit map correspond to character
1140 values, not to byte values. However, the bit map we are constructing is
1141 for byte values. So we have to do a conversion for characters whose
1142 value is > 127. In fact, there are only two possible starting bytes for
1143 characters in the range 128 - 255. */
1144
1145 #ifdef SUPPORT_UTF8
1146 if (utf8)
1147 {
1148 for (c = 0; c < 16; c++) start_bits[c] |= tcode[c];
1149 for (c = 128; c < 256; c++)
1150 {
1151 if ((tcode[c/8] && (1 << (c&7))) != 0)
1152 {
1153 int d = (c >> 6) | 0xc0; /* Set bit for this starter */
1154 start_bits[d/8] |= (1 << (d&7)); /* and then skip on to the */
1155 c = (c & 0xc0) + 0x40 - 1; /* next relevant character. */
1156 }
1157 }
1158 }
1159
1160 /* In non-UTF-8 mode, the two bit maps are completely compatible. */
1161
1162 else
1163 #endif
1164 {
1165 for (c = 0; c < 32; c++) start_bits[c] |= tcode[c];
1166 }
1167
1168 /* Advance past the bit map, and act on what follows. For a zero
1169 minimum repeat, continue; otherwise stop processing. */
1170
1171 tcode += 32;
1172 switch (*tcode)
1173 {
1174 case OP_CRSTAR:
1175 case OP_CRMINSTAR:
1176 case OP_CRQUERY:
1177 case OP_CRMINQUERY:
1178 tcode++;
1179 break;
1180
1181 case OP_CRRANGE:
1182 case OP_CRMINRANGE:
1183 if (((tcode[1] << 8) + tcode[2]) == 0) tcode += 5;
1184 else try_next = FALSE;
1185 break;
1186
1187 default:
1188 try_next = FALSE;
1189 break;
1190 }
1191 }
1192 break; /* End of bitmap class handling */
1193
1194 } /* End of switch */
1195 } /* End of try_next loop */
1196
1197 code += GET(code, 1); /* Advance to next branch */
1198 }
1199 while (*code == OP_ALT);
1200 return yield;
1201 }
1202
1203
1204
1205
1206
1207 /*************************************************
1208 * Study a compiled expression *
1209 *************************************************/
1210
1211 /* This function is handed a compiled expression that it must study to produce
1212 information that will speed up the matching. It returns a pcre_extra block
1213 which then gets handed back to pcre_exec().
1214
1215 Arguments:
1216 re points to the compiled expression
1217 options contains option bits
1218 errorptr points to where to place error messages;
1219 set NULL unless error
1220
1221 Returns: pointer to a pcre_extra block, with study_data filled in and the
1222 appropriate flags set;
1223 NULL on error or if no optimization possible
1224 */
1225
1226 PCRE_EXP_DEFN pcre_extra * PCRE_CALL_CONVENTION
1227 pcre_study(const pcre *external_re, int options, const char **errorptr)
1228 {
1229 int min;
1230 BOOL bits_set = FALSE;
1231 BOOL had_accept = FALSE;
1232 uschar start_bits[32];
1233 pcre_extra *extra = NULL;
1234 pcre_study_data *study;
1235 const uschar *tables;
1236 uschar *code;
1237 compile_data compile_block;
1238 const real_pcre *re = (const real_pcre *)external_re;
1239
1240 *errorptr = NULL;
1241
1242 if (re == NULL || re->magic_number != MAGIC_NUMBER)
1243 {
1244 *errorptr = "argument is not a compiled regular expression";
1245 return NULL;
1246 }
1247
1248 if ((options & ~PUBLIC_STUDY_OPTIONS) != 0)
1249 {
1250 *errorptr = "unknown or incorrect option bit(s) set";
1251 return NULL;
1252 }
1253
1254 code = (uschar *)re + re->name_table_offset +
1255 (re->name_count * re->name_entry_size);
1256
1257 /* For an anchored pattern, or an unanchored pattern that has a first char, or
1258 a multiline pattern that matches only at "line starts", there is no point in
1259 seeking a list of starting bytes. */
1260
1261 if ((re->options & PCRE_ANCHORED) == 0 &&
1262 (re->flags & (PCRE_FIRSTSET|PCRE_STARTLINE)) == 0)
1263 {
1264 int rc;
1265
1266 /* Set the character tables in the block that is passed around */
1267
1268 tables = re->tables;
1269 if (tables == NULL)
1270 (void)pcre_fullinfo(external_re, NULL, PCRE_INFO_DEFAULT_TABLES,
1271 (void *)(&tables));
1272
1273 compile_block.lcc = tables + lcc_offset;
1274 compile_block.fcc = tables + fcc_offset;
1275 compile_block.cbits = tables + cbits_offset;
1276 compile_block.ctypes = tables + ctypes_offset;
1277
1278 /* See if we can find a fixed set of initial characters for the pattern. */
1279
1280 memset(start_bits, 0, 32 * sizeof(uschar));
1281 rc = set_start_bits(code, start_bits, (re->options & PCRE_UTF8) != 0,
1282 &compile_block);
1283 bits_set = rc == SSB_DONE;
1284 if (rc == SSB_UNKNOWN)
1285 {
1286 *errorptr = "internal error: opcode not recognized";
1287 return NULL;
1288 }
1289 }
1290
1291 /* Find the minimum length of subject string. */
1292
1293 switch(min = find_minlength(code, code, re->options, &had_accept, 0))
1294 {
1295 case -2: *errorptr = "internal error: missing capturing bracket"; return NULL;
1296 case -3: *errorptr = "internal error: opcode not recognized"; return NULL;
1297 default: break;
1298 }
1299
1300 /* If a set of starting bytes has been identified, or if the minimum length is
1301 greater than zero, or if JIT optimization has been requested, get a pcre_extra
1302 block and a pcre_study_data block. The study data is put in the latter, which
1303 is pointed to by the former, which may also get additional data set later by
1304 the calling program. At the moment, the size of pcre_study_data is fixed. We
1305 nevertheless save it in a field for returning via the pcre_fullinfo() function
1306 so that if it becomes variable in the future, we don't have to change that
1307 code. */
1308
1309 if (bits_set || min > 0
1310 #ifdef SUPPORT_JIT
1311 || (options & PCRE_STUDY_JIT_COMPILE) != 0
1312 #endif
1313 )
1314 {
1315 extra = (pcre_extra *)(pcre_malloc)
1316 (sizeof(pcre_extra) + sizeof(pcre_study_data));
1317 if (extra == NULL)
1318 {
1319 *errorptr = "failed to get memory";
1320 return NULL;
1321 }
1322
1323 study = (pcre_study_data *)((char *)extra + sizeof(pcre_extra));
1324 extra->flags = PCRE_EXTRA_STUDY_DATA;
1325 extra->study_data = study;
1326
1327 study->size = sizeof(pcre_study_data);
1328 study->flags = 0;
1329
1330 if (bits_set)
1331 {
1332 study->flags |= PCRE_STUDY_MAPPED;
1333 memcpy(study->start_bits, start_bits, sizeof(start_bits));
1334 }
1335
1336 if (min > 0)
1337 {
1338 study->flags |= PCRE_STUDY_MINLEN;
1339 study->minlength = min;
1340 }
1341
1342 /* If JIT support was compiled and requested, attempt the JIT compilation.
1343 If no starting bytes were found, and the minimum length is zero, and JIT
1344 compilation fails, no flags will be set, so abandon the extra block and
1345 return NULL. */
1346
1347 #ifdef SUPPORT_JIT
1348 extra->executable_jit = NULL;
1349 if ((options & PCRE_STUDY_JIT_COMPILE) != 0) _pcre_jit_compile(re, extra);
1350 if (study->flags == 0)
1351 {
1352 pcre_free_study(extra);
1353 extra = NULL;
1354 }
1355 #endif
1356 }
1357
1358 return extra;
1359 }
1360
1361
1362 /*************************************************
1363 * Free the study data *
1364 *************************************************/
1365
1366 /* This function frees the memory that was obtained by pcre_study().
1367
1368 Argument: a pointer to the pcre_extra block
1369 Returns: nothing
1370 */
1371
1372 PCRE_EXP_DEFN void
1373 pcre_free_study(pcre_extra *extra)
1374 {
1375 #ifdef SUPPORT_JIT
1376 if ((extra->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0 &&
1377 extra->executable_jit != NULL)
1378 _pcre_jit_free(extra->executable_jit);
1379 #endif
1380 pcre_free(extra);
1381 }
1382
1383 /* End of pcre_study.c */

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12