/[pcre]/code/trunk/pcre_dfa_exec.c
ViewVC logotype

Contents of /code/trunk/pcre_dfa_exec.c

Parent Directory Parent Directory | Revision Log Revision Log

Revision 461 - (show annotations)
Mon Oct 5 10:59:35 2009 UTC (10 years, 2 months ago) by ph10
File MIME type: text/plain
File size: 102172 byte(s)
Error occurred while calculating annotation data. 
Tidy up, remove trailing spaces, etc. for 8.00-RC1.
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 (but see
7 below for why this module is different).
8
9 Written by Philip Hazel
10 Copyright (c) 1997-2009 University of Cambridge
11
12 -----------------------------------------------------------------------------
13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions are met:
15
16 * Redistributions of source code must retain the above copyright notice,
17 this list of conditions and the following disclaimer.
18
19 * Redistributions in binary form must reproduce the above copyright
20 notice, this list of conditions and the following disclaimer in the
21 documentation and/or other materials provided with the distribution.
22
23 * Neither the name of the University of Cambridge nor the names of its
24 contributors may be used to endorse or promote products derived from
25 this software without specific prior written permission.
26
27 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37 POSSIBILITY OF SUCH DAMAGE.
38 -----------------------------------------------------------------------------
39 */
40
41
42 /* This module contains the external function pcre_dfa_exec(), which is an
43 alternative matching function that uses a sort of DFA algorithm (not a true
44 FSM). This is NOT Perl- compatible, but it has advantages in certain
45 applications. */
46
47
48 /* NOTE ABOUT PERFORMANCE: A user of this function sent some code that improved
49 the performance of his patterns greatly. I could not use it as it stood, as it
50 was not thread safe, and made assumptions about pattern sizes. Also, it caused
51 test 7 to loop, and test 9 to crash with a segfault.
52
53 The issue is the check for duplicate states, which is done by a simple linear
54 search up the state list. (Grep for "duplicate" below to find the code.) For
55 many patterns, there will never be many states active at one time, so a simple
56 linear search is fine. In patterns that have many active states, it might be a
57 bottleneck. The suggested code used an indexing scheme to remember which states
58 had previously been used for each character, and avoided the linear search when
59 it knew there was no chance of a duplicate. This was implemented when adding
60 states to the state lists.
61
62 I wrote some thread-safe, not-limited code to try something similar at the time
63 of checking for duplicates (instead of when adding states), using index vectors
64 on the stack. It did give a 13% improvement with one specially constructed
65 pattern for certain subject strings, but on other strings and on many of the
66 simpler patterns in the test suite it did worse. The major problem, I think,
67 was the extra time to initialize the index. This had to be done for each call
68 of internal_dfa_exec(). (The supplied patch used a static vector, initialized
69 only once - I suspect this was the cause of the problems with the tests.)
70
71 Overall, I concluded that the gains in some cases did not outweigh the losses
72 in others, so I abandoned this code. */
73
74
75
76 #ifdef HAVE_CONFIG_H
77 #include "config.h"
78 #endif
79
80 #define NLBLOCK md /* Block containing newline information */
81 #define PSSTART start_subject /* Field containing processed string start */
82 #define PSEND end_subject /* Field containing processed string end */
83
84 #include "pcre_internal.h"
85
86
87 /* For use to indent debugging output */
88
89 #define SP " "
90
91
92 /*************************************************
93 * Code parameters and static tables *
94 *************************************************/
95
96 /* These are offsets that are used to turn the OP_TYPESTAR and friends opcodes
97 into others, under special conditions. A gap of 20 between the blocks should be
98 enough. The resulting opcodes don't have to be less than 256 because they are
99 never stored, so we push them well clear of the normal opcodes. */
100
101 #define OP_PROP_EXTRA 300
102 #define OP_EXTUNI_EXTRA 320
103 #define OP_ANYNL_EXTRA 340
104 #define OP_HSPACE_EXTRA 360
105 #define OP_VSPACE_EXTRA 380
106
107
108 /* This table identifies those opcodes that are followed immediately by a
109 character that is to be tested in some way. This makes is possible to
110 centralize the loading of these characters. In the case of Type * etc, the
111 "character" is the opcode for \D, \d, \S, \s, \W, or \w, which will always be a
112 small value. ***NOTE*** If the start of this table is modified, the two tables
113 that follow must also be modified. */
114
115 static const uschar coptable[] = {
116 0, /* End */
117 0, 0, 0, 0, 0, /* \A, \G, \K, \B, \b */
118 0, 0, 0, 0, 0, 0, /* \D, \d, \S, \s, \W, \w */
119 0, 0, 0, /* Any, AllAny, Anybyte */
120 0, 0, 0, /* NOTPROP, PROP, EXTUNI */
121 0, 0, 0, 0, 0, /* \R, \H, \h, \V, \v */
122 0, 0, 0, 0, 0, /* \Z, \z, Opt, ^, $ */
123 1, /* Char */
124 1, /* Charnc */
125 1, /* not */
126 /* Positive single-char repeats */
127 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */
128 3, 3, 3, /* upto, minupto, exact */
129 1, 1, 1, 3, /* *+, ++, ?+, upto+ */
130 /* Negative single-char repeats - only for chars < 256 */
131 1, 1, 1, 1, 1, 1, /* NOT *, *?, +, +?, ?, ?? */
132 3, 3, 3, /* NOT upto, minupto, exact */
133 1, 1, 1, 3, /* NOT *+, ++, ?+, updo+ */
134 /* Positive type repeats */
135 1, 1, 1, 1, 1, 1, /* Type *, *?, +, +?, ?, ?? */
136 3, 3, 3, /* Type upto, minupto, exact */
137 1, 1, 1, 3, /* Type *+, ++, ?+, upto+ */
138 /* Character class & ref repeats */
139 0, 0, 0, 0, 0, 0, /* *, *?, +, +?, ?, ?? */
140 0, 0, /* CRRANGE, CRMINRANGE */
141 0, /* CLASS */
142 0, /* NCLASS */
143 0, /* XCLASS - variable length */
144 0, /* REF */
145 0, /* RECURSE */
146 0, /* CALLOUT */
147 0, /* Alt */
148 0, /* Ket */
149 0, /* KetRmax */
150 0, /* KetRmin */
151 0, /* Assert */
152 0, /* Assert not */
153 0, /* Assert behind */
154 0, /* Assert behind not */
155 0, /* Reverse */
156 0, 0, 0, 0, /* ONCE, BRA, CBRA, COND */
157 0, 0, 0, /* SBRA, SCBRA, SCOND */
158 0, /* CREF */
159 0, /* RREF */
160 0, /* DEF */
161 0, 0, /* BRAZERO, BRAMINZERO */
162 0, 0, 0, 0, /* PRUNE, SKIP, THEN, COMMIT */
163 0, 0, 0 /* FAIL, ACCEPT, SKIPZERO */
164 };
165
166 /* These 2 tables allow for compact code for testing for \D, \d, \S, \s, \W,
167 and \w */
168
169 static const uschar toptable1[] = {
170 0, 0, 0, 0, 0, 0,
171 ctype_digit, ctype_digit,
172 ctype_space, ctype_space,
173 ctype_word, ctype_word,
174 0, 0 /* OP_ANY, OP_ALLANY */
175 };
176
177 static const uschar toptable2[] = {
178 0, 0, 0, 0, 0, 0,
179 ctype_digit, 0,
180 ctype_space, 0,
181 ctype_word, 0,
182 1, 1 /* OP_ANY, OP_ALLANY */
183 };
184
185
186 /* Structure for holding data about a particular state, which is in effect the
187 current data for an active path through the match tree. It must consist
188 entirely of ints because the working vector we are passed, and which we put
189 these structures in, is a vector of ints. */
190
191 typedef struct stateblock {
192 int offset; /* Offset to opcode */
193 int count; /* Count for repeats */
194 int ims; /* ims flag bits */
195 int data; /* Some use extra data */
196 } stateblock;
197
198 #define INTS_PER_STATEBLOCK (sizeof(stateblock)/sizeof(int))
199
200
201 #ifdef DEBUG
202 /*************************************************
203 * Print character string *
204 *************************************************/
205
206 /* Character string printing function for debugging.
207
208 Arguments:
209 p points to string
210 length number of bytes
211 f where to print
212
213 Returns: nothing
214 */
215
216 static void
217 pchars(unsigned char *p, int length, FILE *f)
218 {
219 int c;
220 while (length-- > 0)
221 {
222 if (isprint(c = *(p++)))
223 fprintf(f, "%c", c);
224 else
225 fprintf(f, "\\x%02x", c);
226 }
227 }
228 #endif
229
230
231
232 /*************************************************
233 * Execute a Regular Expression - DFA engine *
234 *************************************************/
235
236 /* This internal function applies a compiled pattern to a subject string,
237 starting at a given point, using a DFA engine. This function is called from the
238 external one, possibly multiple times if the pattern is not anchored. The
239 function calls itself recursively for some kinds of subpattern.
240
241 Arguments:
242 md the match_data block with fixed information
243 this_start_code the opening bracket of this subexpression's code
244 current_subject where we currently are in the subject string
245 start_offset start offset in the subject string
246 offsets vector to contain the matching string offsets
247 offsetcount size of same
248 workspace vector of workspace
249 wscount size of same
250 ims the current ims flags
251 rlevel function call recursion level
252 recursing regex recursive call level
253
254 Returns: > 0 => number of match offset pairs placed in offsets
255 = 0 => offsets overflowed; longest matches are present
256 -1 => failed to match
257 < -1 => some kind of unexpected problem
258
259 The following macros are used for adding states to the two state vectors (one
260 for the current character, one for the following character). */
261
262 #define ADD_ACTIVE(x,y) \
263 if (active_count++ < wscount) \
264 { \
265 next_active_state->offset = (x); \
266 next_active_state->count = (y); \
267 next_active_state->ims = ims; \
268 next_active_state++; \
269 DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
270 } \
271 else return PCRE_ERROR_DFA_WSSIZE
272
273 #define ADD_ACTIVE_DATA(x,y,z) \
274 if (active_count++ < wscount) \
275 { \
276 next_active_state->offset = (x); \
277 next_active_state->count = (y); \
278 next_active_state->ims = ims; \
279 next_active_state->data = (z); \
280 next_active_state++; \
281 DPRINTF(("%.*sADD_ACTIVE_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \
282 } \
283 else return PCRE_ERROR_DFA_WSSIZE
284
285 #define ADD_NEW(x,y) \
286 if (new_count++ < wscount) \
287 { \
288 next_new_state->offset = (x); \
289 next_new_state->count = (y); \
290 next_new_state->ims = ims; \
291 next_new_state++; \
292 DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
293 } \
294 else return PCRE_ERROR_DFA_WSSIZE
295
296 #define ADD_NEW_DATA(x,y,z) \
297 if (new_count++ < wscount) \
298 { \
299 next_new_state->offset = (x); \
300 next_new_state->count = (y); \
301 next_new_state->ims = ims; \
302 next_new_state->data = (z); \
303 next_new_state++; \
304 DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \
305 } \
306 else return PCRE_ERROR_DFA_WSSIZE
307
308 /* And now, here is the code */
309
310 static int
311 internal_dfa_exec(
312 dfa_match_data *md,
313 const uschar *this_start_code,
314 const uschar *current_subject,
315 int start_offset,
316 int *offsets,
317 int offsetcount,
318 int *workspace,
319 int wscount,
320 int ims,
321 int rlevel,
322 int recursing)
323 {
324 stateblock *active_states, *new_states, *temp_states;
325 stateblock *next_active_state, *next_new_state;
326
327 const uschar *ctypes, *lcc, *fcc;
328 const uschar *ptr;
329 const uschar *end_code, *first_op;
330
331 int active_count, new_count, match_count;
332
333 /* Some fields in the md block are frequently referenced, so we load them into
334 independent variables in the hope that this will perform better. */
335
336 const uschar *start_subject = md->start_subject;
337 const uschar *end_subject = md->end_subject;
338 const uschar *start_code = md->start_code;
339
340 #ifdef SUPPORT_UTF8
341 BOOL utf8 = (md->poptions & PCRE_UTF8) != 0;
342 #else
343 BOOL utf8 = FALSE;
344 #endif
345
346 rlevel++;
347 offsetcount &= (-2);
348
349 wscount -= 2;
350 wscount = (wscount - (wscount % (INTS_PER_STATEBLOCK * 2))) /
351 (2 * INTS_PER_STATEBLOCK);
352
353 DPRINTF(("\n%.*s---------------------\n"
354 "%.*sCall to internal_dfa_exec f=%d r=%d\n",
355 rlevel*2-2, SP, rlevel*2-2, SP, rlevel, recursing));
356
357 ctypes = md->tables + ctypes_offset;
358 lcc = md->tables + lcc_offset;
359 fcc = md->tables + fcc_offset;
360
361 match_count = PCRE_ERROR_NOMATCH; /* A negative number */
362
363 active_states = (stateblock *)(workspace + 2);
364 next_new_state = new_states = active_states + wscount;
365 new_count = 0;
366
367 first_op = this_start_code + 1 + LINK_SIZE +
368 ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);
369
370 /* The first thing in any (sub) pattern is a bracket of some sort. Push all
371 the alternative states onto the list, and find out where the end is. This
372 makes is possible to use this function recursively, when we want to stop at a
373 matching internal ket rather than at the end.
374
375 If the first opcode in the first alternative is OP_REVERSE, we are dealing with
376 a backward assertion. In that case, we have to find out the maximum amount to
377 move back, and set up each alternative appropriately. */
378
379 if (*first_op == OP_REVERSE)
380 {
381 int max_back = 0;
382 int gone_back;
383
384 end_code = this_start_code;
385 do
386 {
387 int back = GET(end_code, 2+LINK_SIZE);
388 if (back > max_back) max_back = back;
389 end_code += GET(end_code, 1);
390 }
391 while (*end_code == OP_ALT);
392
393 /* If we can't go back the amount required for the longest lookbehind
394 pattern, go back as far as we can; some alternatives may still be viable. */
395
396 #ifdef SUPPORT_UTF8
397 /* In character mode we have to step back character by character */
398
399 if (utf8)
400 {
401 for (gone_back = 0; gone_back < max_back; gone_back++)
402 {
403 if (current_subject <= start_subject) break;
404 current_subject--;
405 while (current_subject > start_subject &&
406 (*current_subject & 0xc0) == 0x80)
407 current_subject--;
408 }
409 }
410 else
411 #endif
412
413 /* In byte-mode we can do this quickly. */
414
415 {
416 gone_back = (current_subject - max_back < start_subject)?
417 current_subject - start_subject : max_back;
418 current_subject -= gone_back;
419 }
420
421 /* Save the earliest consulted character */
422
423 if (current_subject < md->start_used_ptr)
424 md->start_used_ptr = current_subject;
425
426 /* Now we can process the individual branches. */
427
428 end_code = this_start_code;
429 do
430 {
431 int back = GET(end_code, 2+LINK_SIZE);
432 if (back <= gone_back)
433 {
434 int bstate = end_code - start_code + 2 + 2*LINK_SIZE;
435 ADD_NEW_DATA(-bstate, 0, gone_back - back);
436 }
437 end_code += GET(end_code, 1);
438 }
439 while (*end_code == OP_ALT);
440 }
441
442 /* This is the code for a "normal" subpattern (not a backward assertion). The
443 start of a whole pattern is always one of these. If we are at the top level,
444 we may be asked to restart matching from the same point that we reached for a
445 previous partial match. We still have to scan through the top-level branches to
446 find the end state. */
447
448 else
449 {
450 end_code = this_start_code;
451
452 /* Restarting */
453
454 if (rlevel == 1 && (md->moptions & PCRE_DFA_RESTART) != 0)
455 {
456 do { end_code += GET(end_code, 1); } while (*end_code == OP_ALT);
457 new_count = workspace[1];
458 if (!workspace[0])
459 memcpy(new_states, active_states, new_count * sizeof(stateblock));
460 }
461
462 /* Not restarting */
463
464 else
465 {
466 int length = 1 + LINK_SIZE +
467 ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);
468 do
469 {
470 ADD_NEW(end_code - start_code + length, 0);
471 end_code += GET(end_code, 1);
472 length = 1 + LINK_SIZE;
473 }
474 while (*end_code == OP_ALT);
475 }
476 }
477
478 workspace[0] = 0; /* Bit indicating which vector is current */
479
480 DPRINTF(("%.*sEnd state = %d\n", rlevel*2-2, SP, end_code - start_code));
481
482 /* Loop for scanning the subject */
483
484 ptr = current_subject;
485 for (;;)
486 {
487 int i, j;
488 int clen, dlen;
489 unsigned int c, d;
490 int forced_fail = 0;
491 int reached_end = 0;
492
493 /* Make the new state list into the active state list and empty the
494 new state list. */
495
496 temp_states = active_states;
497 active_states = new_states;
498 new_states = temp_states;
499 active_count = new_count;
500 new_count = 0;
501
502 workspace[0] ^= 1; /* Remember for the restarting feature */
503 workspace[1] = active_count;
504
505 #ifdef DEBUG
506 printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);
507 pchars((uschar *)ptr, strlen((char *)ptr), stdout);
508 printf("\"\n");
509
510 printf("%.*sActive states: ", rlevel*2-2, SP);
511 for (i = 0; i < active_count; i++)
512 printf("%d/%d ", active_states[i].offset, active_states[i].count);
513 printf("\n");
514 #endif
515
516 /* Set the pointers for adding new states */
517
518 next_active_state = active_states + active_count;
519 next_new_state = new_states;
520
521 /* Load the current character from the subject outside the loop, as many
522 different states may want to look at it, and we assume that at least one
523 will. */
524
525 if (ptr < end_subject)
526 {
527 clen = 1; /* Number of bytes in the character */
528 #ifdef SUPPORT_UTF8
529 if (utf8) { GETCHARLEN(c, ptr, clen); } else
530 #endif /* SUPPORT_UTF8 */
531 c = *ptr;
532 }
533 else
534 {
535 clen = 0; /* This indicates the end of the subject */
536 c = NOTACHAR; /* This value should never actually be used */
537 }
538
539 /* Scan up the active states and act on each one. The result of an action
540 may be to add more states to the currently active list (e.g. on hitting a
541 parenthesis) or it may be to put states on the new list, for considering
542 when we move the character pointer on. */
543
544 for (i = 0; i < active_count; i++)
545 {
546 stateblock *current_state = active_states + i;
547 const uschar *code;
548 int state_offset = current_state->offset;
549 int count, codevalue, rrc;
550
551 #ifdef DEBUG
552 printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);
553 if (clen == 0) printf("EOL\n");
554 else if (c > 32 && c < 127) printf("'%c'\n", c);
555 else printf("0x%02x\n", c);
556 #endif
557
558 /* This variable is referred to implicity in the ADD_xxx macros. */
559
560 ims = current_state->ims;
561
562 /* A negative offset is a special case meaning "hold off going to this
563 (negated) state until the number of characters in the data field have
564 been skipped". */
565
566 if (state_offset < 0)
567 {
568 if (current_state->data > 0)
569 {
570 DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));
571 ADD_NEW_DATA(state_offset, current_state->count,
572 current_state->data - 1);
573 continue;
574 }
575 else
576 {
577 current_state->offset = state_offset = -state_offset;
578 }
579 }
580
581 /* Check for a duplicate state with the same count, and skip if found.
582 See the note at the head of this module about the possibility of improving
583 performance here. */
584
585 for (j = 0; j < i; j++)
586 {
587 if (active_states[j].offset == state_offset &&
588 active_states[j].count == current_state->count)
589 {
590 DPRINTF(("%.*sDuplicate state: skipped\n", rlevel*2-2, SP));
591 goto NEXT_ACTIVE_STATE;
592 }
593 }
594
595 /* The state offset is the offset to the opcode */
596
597 code = start_code + state_offset;
598 codevalue = *code;
599
600 /* If this opcode is followed by an inline character, load it. It is
601 tempting to test for the presence of a subject character here, but that
602 is wrong, because sometimes zero repetitions of the subject are
603 permitted.
604
605 We also use this mechanism for opcodes such as OP_TYPEPLUS that take an
606 argument that is not a data character - but is always one byte long. We
607 have to take special action to deal with \P, \p, \H, \h, \V, \v and \X in
608 this case. To keep the other cases fast, convert these ones to new opcodes.
609 */
610
611 if (coptable[codevalue] > 0)
612 {
613 dlen = 1;
614 #ifdef SUPPORT_UTF8
615 if (utf8) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else
616 #endif /* SUPPORT_UTF8 */
617 d = code[coptable[codevalue]];
618 if (codevalue >= OP_TYPESTAR)
619 {
620 switch(d)
621 {
622 case OP_ANYBYTE: return PCRE_ERROR_DFA_UITEM;
623 case OP_NOTPROP:
624 case OP_PROP: codevalue += OP_PROP_EXTRA; break;
625 case OP_ANYNL: codevalue += OP_ANYNL_EXTRA; break;
626 case OP_EXTUNI: codevalue += OP_EXTUNI_EXTRA; break;
627 case OP_NOT_HSPACE:
628 case OP_HSPACE: codevalue += OP_HSPACE_EXTRA; break;
629 case OP_NOT_VSPACE:
630 case OP_VSPACE: codevalue += OP_VSPACE_EXTRA; break;
631 default: break;
632 }
633 }
634 }
635 else
636 {
637 dlen = 0; /* Not strictly necessary, but compilers moan */
638 d = NOTACHAR; /* if these variables are not set. */
639 }
640
641
642 /* Now process the individual opcodes */
643
644 switch (codevalue)
645 {
646
647 /* ========================================================================== */
648 /* Reached a closing bracket. If not at the end of the pattern, carry
649 on with the next opcode. Otherwise, unless we have an empty string and
650 PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the
651 start of the subject, save the match data, shifting up all previous
652 matches so we always have the longest first. */
653
654 case OP_KET:
655 case OP_KETRMIN:
656 case OP_KETRMAX:
657 if (code != end_code)
658 {
659 ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);
660 if (codevalue != OP_KET)
661 {
662 ADD_ACTIVE(state_offset - GET(code, 1), 0);
663 }
664 }
665 else
666 {
667 reached_end++; /* Count branches that reach the end */
668 if (ptr > current_subject ||
669 ((md->moptions & PCRE_NOTEMPTY) == 0 &&
670 ((md->moptions & PCRE_NOTEMPTY_ATSTART) == 0 ||
671 current_subject > start_subject + md->start_offset)))
672 {
673 if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;
674 else if (match_count > 0 && ++match_count * 2 >= offsetcount)
675 match_count = 0;
676 count = ((match_count == 0)? offsetcount : match_count * 2) - 2;
677 if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));
678 if (offsetcount >= 2)
679 {
680 offsets[0] = current_subject - start_subject;
681 offsets[1] = ptr - start_subject;
682 DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,
683 offsets[1] - offsets[0], current_subject));
684 }
685 if ((md->moptions & PCRE_DFA_SHORTEST) != 0)
686 {
687 DPRINTF(("%.*sEnd of internal_dfa_exec %d: returning %d\n"
688 "%.*s---------------------\n\n", rlevel*2-2, SP, rlevel,
689 match_count, rlevel*2-2, SP));
690 return match_count;
691 }
692 }
693 }
694 break;
695
696 /* ========================================================================== */
697 /* These opcodes add to the current list of states without looking
698 at the current character. */
699
700 /*-----------------------------------------------------------------*/
701 case OP_ALT:
702 do { code += GET(code, 1); } while (*code == OP_ALT);
703 ADD_ACTIVE(code - start_code, 0);
704 break;
705
706 /*-----------------------------------------------------------------*/
707 case OP_BRA:
708 case OP_SBRA:
709 do
710 {
711 ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);
712 code += GET(code, 1);
713 }
714 while (*code == OP_ALT);
715 break;
716
717 /*-----------------------------------------------------------------*/
718 case OP_CBRA:
719 case OP_SCBRA:
720 ADD_ACTIVE(code - start_code + 3 + LINK_SIZE, 0);
721 code += GET(code, 1);
722 while (*code == OP_ALT)
723 {
724 ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);
725 code += GET(code, 1);
726 }
727 break;
728
729 /*-----------------------------------------------------------------*/
730 case OP_BRAZERO:
731 case OP_BRAMINZERO:
732 ADD_ACTIVE(state_offset + 1, 0);
733 code += 1 + GET(code, 2);
734 while (*code == OP_ALT) code += GET(code, 1);
735 ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);
736 break;
737
738 /*-----------------------------------------------------------------*/
739 case OP_SKIPZERO:
740 code += 1 + GET(code, 2);
741 while (*code == OP_ALT) code += GET(code, 1);
742 ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);
743 break;
744
745 /*-----------------------------------------------------------------*/
746 case OP_CIRC:
747 if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||
748 ((ims & PCRE_MULTILINE) != 0 &&
749 ptr != end_subject &&
750 WAS_NEWLINE(ptr)))
751 { ADD_ACTIVE(state_offset + 1, 0); }
752 break;
753
754 /*-----------------------------------------------------------------*/
755 case OP_EOD:
756 if (ptr >= end_subject) { ADD_ACTIVE(state_offset + 1, 0); }
757 break;
758
759 /*-----------------------------------------------------------------*/
760 case OP_OPT:
761 ims = code[1];
762 ADD_ACTIVE(state_offset + 2, 0);
763 break;
764
765 /*-----------------------------------------------------------------*/
766 case OP_SOD:
767 if (ptr == start_subject) { ADD_ACTIVE(state_offset + 1, 0); }
768 break;
769
770 /*-----------------------------------------------------------------*/
771 case OP_SOM:
772 if (ptr == start_subject + start_offset) { ADD_ACTIVE(state_offset + 1, 0); }
773 break;
774
775
776 /* ========================================================================== */
777 /* These opcodes inspect the next subject character, and sometimes
778 the previous one as well, but do not have an argument. The variable
779 clen contains the length of the current character and is zero if we are
780 at the end of the subject. */
781
782 /*-----------------------------------------------------------------*/
783 case OP_ANY:
784 if (clen > 0 && !IS_NEWLINE(ptr))
785 { ADD_NEW(state_offset + 1, 0); }
786 break;
787
788 /*-----------------------------------------------------------------*/
789 case OP_ALLANY:
790 if (clen > 0)
791 { ADD_NEW(state_offset + 1, 0); }
792 break;
793
794 /*-----------------------------------------------------------------*/
795 case OP_EODN:
796 if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))
797 { ADD_ACTIVE(state_offset + 1, 0); }
798 break;
799
800 /*-----------------------------------------------------------------*/
801 case OP_DOLL:
802 if ((md->moptions & PCRE_NOTEOL) == 0)
803 {
804 if (clen == 0 ||
805 ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
806 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)
807 ))
808 { ADD_ACTIVE(state_offset + 1, 0); }
809 }
810 else if ((ims & PCRE_MULTILINE) != 0 && IS_NEWLINE(ptr))
811 { ADD_ACTIVE(state_offset + 1, 0); }
812 break;
813
814 /*-----------------------------------------------------------------*/
815
816 case OP_DIGIT:
817 case OP_WHITESPACE:
818 case OP_WORDCHAR:
819 if (clen > 0 && c < 256 &&
820 ((ctypes[c] & toptable1[codevalue]) ^ toptable2[codevalue]) != 0)
821 { ADD_NEW(state_offset + 1, 0); }
822 break;
823
824 /*-----------------------------------------------------------------*/
825 case OP_NOT_DIGIT:
826 case OP_NOT_WHITESPACE:
827 case OP_NOT_WORDCHAR:
828 if (clen > 0 && (c >= 256 ||
829 ((ctypes[c] & toptable1[codevalue]) ^ toptable2[codevalue]) != 0))
830 { ADD_NEW(state_offset + 1, 0); }
831 break;
832
833 /*-----------------------------------------------------------------*/
834 case OP_WORD_BOUNDARY:
835 case OP_NOT_WORD_BOUNDARY:
836 {
837 int left_word, right_word;
838
839 if (ptr > start_subject)
840 {
841 const uschar *temp = ptr - 1;
842 if (temp < md->start_used_ptr) md->start_used_ptr = temp;
843 #ifdef SUPPORT_UTF8
844 if (utf8) BACKCHAR(temp);
845 #endif
846 GETCHARTEST(d, temp);
847 left_word = d < 256 && (ctypes[d] & ctype_word) != 0;
848 }
849 else left_word = 0;
850
851 if (clen > 0)
852 right_word = c < 256 && (ctypes[c] & ctype_word) != 0;
853 else /* This is a fudge to ensure that if this is the */
854 { /* last item in the pattern, we don't count it as */
855 reached_end--; /* reached, thus disabling a partial match. */
856 right_word = 0;
857 }
858
859 if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))
860 { ADD_ACTIVE(state_offset + 1, 0); }
861 }
862 break;
863
864
865 /*-----------------------------------------------------------------*/
866 /* Check the next character by Unicode property. We will get here only
867 if the support is in the binary; otherwise a compile-time error occurs.
868 */
869
870 #ifdef SUPPORT_UCP
871 case OP_PROP:
872 case OP_NOTPROP:
873 if (clen > 0)
874 {
875 BOOL OK;
876 const ucd_record * prop = GET_UCD(c);
877 switch(code[1])
878 {
879 case PT_ANY:
880 OK = TRUE;
881 break;
882
883 case PT_LAMP:
884 OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
885 break;
886
887 case PT_GC:
888 OK = _pcre_ucp_gentype[prop->chartype] == code[2];
889 break;
890
891 case PT_PC:
892 OK = prop->chartype == code[2];
893 break;
894
895 case PT_SC:
896 OK = prop->script == code[2];
897 break;
898
899 /* Should never occur, but keep compilers from grumbling. */
900
901 default:
902 OK = codevalue != OP_PROP;
903 break;
904 }
905
906 if (OK == (codevalue == OP_PROP)) { ADD_NEW(state_offset + 3, 0); }
907 }
908 break;
909 #endif
910
911
912
913 /* ========================================================================== */
914 /* These opcodes likewise inspect the subject character, but have an
915 argument that is not a data character. It is one of these opcodes:
916 OP_ANY, OP_ALLANY, OP_DIGIT, OP_NOT_DIGIT, OP_WHITESPACE, OP_NOT_SPACE,
917 OP_WORDCHAR, OP_NOT_WORDCHAR. The value is loaded into d. */
918
919 case OP_TYPEPLUS:
920 case OP_TYPEMINPLUS:
921 case OP_TYPEPOSPLUS:
922 count = current_state->count; /* Already matched */
923 if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
924 if (clen > 0)
925 {
926 if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
927 (c < 256 &&
928 (d != OP_ANY || !IS_NEWLINE(ptr)) &&
929 ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
930 {
931 if (count > 0 && codevalue == OP_TYPEPOSPLUS)
932 {
933 active_count--; /* Remove non-match possibility */
934 next_active_state--;
935 }
936 count++;
937 ADD_NEW(state_offset, count);
938 }
939 }
940 break;
941
942 /*-----------------------------------------------------------------*/
943 case OP_TYPEQUERY:
944 case OP_TYPEMINQUERY:
945 case OP_TYPEPOSQUERY:
946 ADD_ACTIVE(state_offset + 2, 0);
947 if (clen > 0)
948 {
949 if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
950 (c < 256 &&
951 (d != OP_ANY || !IS_NEWLINE(ptr)) &&
952 ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
953 {
954 if (codevalue == OP_TYPEPOSQUERY)
955 {
956 active_count--; /* Remove non-match possibility */
957 next_active_state--;
958 }
959 ADD_NEW(state_offset + 2, 0);
960 }
961 }
962 break;
963
964 /*-----------------------------------------------------------------*/
965 case OP_TYPESTAR:
966 case OP_TYPEMINSTAR:
967 case OP_TYPEPOSSTAR:
968 ADD_ACTIVE(state_offset + 2, 0);
969 if (clen > 0)
970 {
971 if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
972 (c < 256 &&
973 (d != OP_ANY || !IS_NEWLINE(ptr)) &&
974 ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
975 {
976 if (codevalue == OP_TYPEPOSSTAR)
977 {
978 active_count--; /* Remove non-match possibility */
979 next_active_state--;
980 }
981 ADD_NEW(state_offset, 0);
982 }
983 }
984 break;
985
986 /*-----------------------------------------------------------------*/
987 case OP_TYPEEXACT:
988 count = current_state->count; /* Number already matched */
989 if (clen > 0)
990 {
991 if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
992 (c < 256 &&
993 (d != OP_ANY || !IS_NEWLINE(ptr)) &&
994 ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
995 {
996 if (++count >= GET2(code, 1))
997 { ADD_NEW(state_offset + 4, 0); }
998 else
999 { ADD_NEW(state_offset, count); }
1000 }
1001 }
1002 break;
1003
1004 /*-----------------------------------------------------------------*/
1005 case OP_TYPEUPTO:
1006 case OP_TYPEMINUPTO:
1007 case OP_TYPEPOSUPTO:
1008 ADD_ACTIVE(state_offset + 4, 0);
1009 count = current_state->count; /* Number already matched */
1010 if (clen > 0)
1011 {
1012 if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1013 (c < 256 &&
1014 (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1015 ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
1016 {
1017 if (codevalue == OP_TYPEPOSUPTO)
1018 {
1019 active_count--; /* Remove non-match possibility */
1020 next_active_state--;
1021 }
1022 if (++count >= GET2(code, 1))
1023 { ADD_NEW(state_offset + 4, 0); }
1024 else
1025 { ADD_NEW(state_offset, count); }
1026 }
1027 }
1028 break;
1029
1030 /* ========================================================================== */
1031 /* These are virtual opcodes that are used when something like
1032 OP_TYPEPLUS has OP_PROP, OP_NOTPROP, OP_ANYNL, or OP_EXTUNI as its
1033 argument. It keeps the code above fast for the other cases. The argument
1034 is in the d variable. */
1035
1036 #ifdef SUPPORT_UCP
1037 case OP_PROP_EXTRA + OP_TYPEPLUS:
1038 case OP_PROP_EXTRA + OP_TYPEMINPLUS:
1039 case OP_PROP_EXTRA + OP_TYPEPOSPLUS:
1040 count = current_state->count; /* Already matched */
1041 if (count > 0) { ADD_ACTIVE(state_offset + 4, 0); }
1042 if (clen > 0)
1043 {
1044 BOOL OK;
1045 const ucd_record * prop = GET_UCD(c);
1046 switch(code[2])
1047 {
1048 case PT_ANY:
1049 OK = TRUE;
1050 break;
1051
1052 case PT_LAMP:
1053 OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1054 break;
1055
1056 case PT_GC:
1057 OK = _pcre_ucp_gentype[prop->chartype] == code[3];
1058 break;
1059
1060 case PT_PC:
1061 OK = prop->chartype == code[3];
1062 break;
1063
1064 case PT_SC:
1065 OK = prop->script == code[3];
1066 break;
1067
1068 /* Should never occur, but keep compilers from grumbling. */
1069
1070 default:
1071 OK = codevalue != OP_PROP;
1072 break;
1073 }
1074
1075 if (OK == (d == OP_PROP))
1076 {
1077 if (count > 0 && codevalue == OP_PROP_EXTRA + OP_TYPEPOSPLUS)
1078 {
1079 active_count--; /* Remove non-match possibility */
1080 next_active_state--;
1081 }
1082 count++;
1083 ADD_NEW(state_offset, count);
1084 }
1085 }
1086 break;
1087
1088 /*-----------------------------------------------------------------*/
1089 case OP_EXTUNI_EXTRA + OP_TYPEPLUS:
1090 case OP_EXTUNI_EXTRA + OP_TYPEMINPLUS:
1091 case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:
1092 count = current_state->count; /* Already matched */
1093 if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1094 if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1095 {
1096 const uschar *nptr = ptr + clen;
1097 int ncount = 0;
1098 if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)
1099 {
1100 active_count--; /* Remove non-match possibility */
1101 next_active_state--;
1102 }
1103 while (nptr < end_subject)
1104 {
1105 int nd;
1106 int ndlen = 1;
1107 GETCHARLEN(nd, nptr, ndlen);
1108 if (UCD_CATEGORY(nd) != ucp_M) break;
1109 ncount++;
1110 nptr += ndlen;
1111 }
1112 count++;
1113 ADD_NEW_DATA(-state_offset, count, ncount);
1114 }
1115 break;
1116 #endif
1117
1118 /*-----------------------------------------------------------------*/
1119 case OP_ANYNL_EXTRA + OP_TYPEPLUS:
1120 case OP_ANYNL_EXTRA + OP_TYPEMINPLUS:
1121 case OP_ANYNL_EXTRA + OP_TYPEPOSPLUS:
1122 count = current_state->count; /* Already matched */
1123 if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1124 if (clen > 0)
1125 {
1126 int ncount = 0;
1127 switch (c)
1128 {
1129 case 0x000b:
1130 case 0x000c:
1131 case 0x0085:
1132 case 0x2028:
1133 case 0x2029:
1134 if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1135 goto ANYNL01;
1136
1137 case 0x000d:
1138 if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;
1139 /* Fall through */
1140
1141 ANYNL01:
1142 case 0x000a:
1143 if (count > 0 && codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSPLUS)
1144 {
1145 active_count--; /* Remove non-match possibility */
1146 next_active_state--;
1147 }
1148 count++;
1149 ADD_NEW_DATA(-state_offset, count, ncount);
1150 break;
1151
1152 default:
1153 break;
1154 }
1155 }
1156 break;
1157
1158 /*-----------------------------------------------------------------*/
1159 case OP_VSPACE_EXTRA + OP_TYPEPLUS:
1160 case OP_VSPACE_EXTRA + OP_TYPEMINPLUS:
1161 case OP_VSPACE_EXTRA + OP_TYPEPOSPLUS:
1162 count = current_state->count; /* Already matched */
1163 if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1164 if (clen > 0)
1165 {
1166 BOOL OK;
1167 switch (c)
1168 {
1169 case 0x000a:
1170 case 0x000b:
1171 case 0x000c:
1172 case 0x000d:
1173 case 0x0085:
1174 case 0x2028:
1175 case 0x2029:
1176 OK = TRUE;
1177 break;
1178
1179 default:
1180 OK = FALSE;
1181 break;
1182 }
1183
1184 if (OK == (d == OP_VSPACE))
1185 {
1186 if (count > 0 && codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSPLUS)
1187 {
1188 active_count--; /* Remove non-match possibility */
1189 next_active_state--;
1190 }
1191 count++;
1192 ADD_NEW_DATA(-state_offset, count, 0);
1193 }
1194 }
1195 break;
1196
1197 /*-----------------------------------------------------------------*/
1198 case OP_HSPACE_EXTRA + OP_TYPEPLUS:
1199 case OP_HSPACE_EXTRA + OP_TYPEMINPLUS:
1200 case OP_HSPACE_EXTRA + OP_TYPEPOSPLUS:
1201 count = current_state->count; /* Already matched */
1202 if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1203 if (clen > 0)
1204 {
1205 BOOL OK;
1206 switch (c)
1207 {
1208 case 0x09: /* HT */
1209 case 0x20: /* SPACE */
1210 case 0xa0: /* NBSP */
1211 case 0x1680: /* OGHAM SPACE MARK */
1212 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */
1213 case 0x2000: /* EN QUAD */
1214 case 0x2001: /* EM QUAD */
1215 case 0x2002: /* EN SPACE */
1216 case 0x2003: /* EM SPACE */
1217 case 0x2004: /* THREE-PER-EM SPACE */
1218 case 0x2005: /* FOUR-PER-EM SPACE */
1219 case 0x2006: /* SIX-PER-EM SPACE */
1220 case 0x2007: /* FIGURE SPACE */
1221 case 0x2008: /* PUNCTUATION SPACE */
1222 case 0x2009: /* THIN SPACE */
1223 case 0x200A: /* HAIR SPACE */
1224 case 0x202f: /* NARROW NO-BREAK SPACE */
1225 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */
1226 case 0x3000: /* IDEOGRAPHIC SPACE */
1227 OK = TRUE;
1228 break;
1229
1230 default:
1231 OK = FALSE;
1232 break;
1233 }
1234
1235 if (OK == (d == OP_HSPACE))
1236 {
1237 if (count > 0 && codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSPLUS)
1238 {
1239 active_count--; /* Remove non-match possibility */
1240 next_active_state--;
1241 }
1242 count++;
1243 ADD_NEW_DATA(-state_offset, count, 0);
1244 }
1245 }
1246 break;
1247
1248 /*-----------------------------------------------------------------*/
1249 #ifdef SUPPORT_UCP
1250 case OP_PROP_EXTRA + OP_TYPEQUERY:
1251 case OP_PROP_EXTRA + OP_TYPEMINQUERY:
1252 case OP_PROP_EXTRA + OP_TYPEPOSQUERY:
1253 count = 4;
1254 goto QS1;
1255
1256 case OP_PROP_EXTRA + OP_TYPESTAR:
1257 case OP_PROP_EXTRA + OP_TYPEMINSTAR:
1258 case OP_PROP_EXTRA + OP_TYPEPOSSTAR:
1259 count = 0;
1260
1261 QS1:
1262
1263 ADD_ACTIVE(state_offset + 4, 0);
1264 if (clen > 0)
1265 {
1266 BOOL OK;
1267 const ucd_record * prop = GET_UCD(c);
1268 switch(code[2])
1269 {
1270 case PT_ANY:
1271 OK = TRUE;
1272 break;
1273
1274 case PT_LAMP:
1275 OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1276 break;
1277
1278 case PT_GC:
1279 OK = _pcre_ucp_gentype[prop->chartype] == code[3];
1280 break;
1281
1282 case PT_PC:
1283 OK = prop->chartype == code[3];
1284 break;
1285
1286 case PT_SC:
1287 OK = prop->script == code[3];
1288 break;
1289
1290 /* Should never occur, but keep compilers from grumbling. */
1291
1292 default:
1293 OK = codevalue != OP_PROP;
1294 break;
1295 }
1296
1297 if (OK == (d == OP_PROP))
1298 {
1299 if (codevalue == OP_PROP_EXTRA + OP_TYPEPOSSTAR ||
1300 codevalue == OP_PROP_EXTRA + OP_TYPEPOSQUERY)
1301 {
1302 active_count--; /* Remove non-match possibility */
1303 next_active_state--;
1304 }
1305 ADD_NEW(state_offset + count, 0);
1306 }
1307 }
1308 break;
1309
1310 /*-----------------------------------------------------------------*/
1311 case OP_EXTUNI_EXTRA + OP_TYPEQUERY:
1312 case OP_EXTUNI_EXTRA + OP_TYPEMINQUERY:
1313 case OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY:
1314 count = 2;
1315 goto QS2;
1316
1317 case OP_EXTUNI_EXTRA + OP_TYPESTAR:
1318 case OP_EXTUNI_EXTRA + OP_TYPEMINSTAR:
1319 case OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR:
1320 count = 0;
1321
1322 QS2:
1323
1324 ADD_ACTIVE(state_offset + 2, 0);
1325 if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1326 {
1327 const uschar *nptr = ptr + clen;
1328 int ncount = 0;
1329 if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||
1330 codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)
1331 {
1332 active_count--; /* Remove non-match possibility */
1333 next_active_state--;
1334 }
1335 while (nptr < end_subject)
1336 {
1337 int nd;
1338 int ndlen = 1;
1339 GETCHARLEN(nd, nptr, ndlen);
1340 if (UCD_CATEGORY(nd) != ucp_M) break;
1341 ncount++;
1342 nptr += ndlen;
1343 }
1344 ADD_NEW_DATA(-(state_offset + count), 0, ncount);
1345 }
1346 break;
1347 #endif
1348
1349 /*-----------------------------------------------------------------*/
1350 case OP_ANYNL_EXTRA + OP_TYPEQUERY:
1351 case OP_ANYNL_EXTRA + OP_TYPEMINQUERY:
1352 case OP_ANYNL_EXTRA + OP_TYPEPOSQUERY:
1353 count = 2;
1354 goto QS3;
1355
1356 case OP_ANYNL_EXTRA + OP_TYPESTAR:
1357 case OP_ANYNL_EXTRA + OP_TYPEMINSTAR:
1358 case OP_ANYNL_EXTRA + OP_TYPEPOSSTAR:
1359 count = 0;
1360
1361 QS3:
1362 ADD_ACTIVE(state_offset + 2, 0);
1363 if (clen > 0)
1364 {
1365 int ncount = 0;
1366 switch (c)
1367 {
1368 case 0x000b:
1369 case 0x000c:
1370 case 0x0085:
1371 case 0x2028:
1372 case 0x2029:
1373 if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1374 goto ANYNL02;
1375
1376 case 0x000d:
1377 if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;
1378 /* Fall through */
1379
1380 ANYNL02:
1381 case 0x000a:
1382 if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSSTAR ||
1383 codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSQUERY)
1384 {
1385 active_count--; /* Remove non-match possibility */
1386 next_active_state--;
1387 }
1388 ADD_NEW_DATA(-(state_offset + count), 0, ncount);
1389 break;
1390
1391 default:
1392 break;
1393 }
1394 }
1395 break;
1396
1397 /*-----------------------------------------------------------------*/
1398 case OP_VSPACE_EXTRA + OP_TYPEQUERY:
1399 case OP_VSPACE_EXTRA + OP_TYPEMINQUERY:
1400 case OP_VSPACE_EXTRA + OP_TYPEPOSQUERY:
1401 count = 2;
1402 goto QS4;
1403
1404 case OP_VSPACE_EXTRA + OP_TYPESTAR:
1405 case OP_VSPACE_EXTRA + OP_TYPEMINSTAR:
1406 case OP_VSPACE_EXTRA + OP_TYPEPOSSTAR:
1407 count = 0;
1408
1409 QS4:
1410 ADD_ACTIVE(state_offset + 2, 0);
1411 if (clen > 0)
1412 {
1413 BOOL OK;
1414 switch (c)
1415 {
1416 case 0x000a:
1417 case 0x000b:
1418 case 0x000c:
1419 case 0x000d:
1420 case 0x0085:
1421 case 0x2028:
1422 case 0x2029:
1423 OK = TRUE;
1424 break;
1425
1426 default:
1427 OK = FALSE;
1428 break;
1429 }
1430 if (OK == (d == OP_VSPACE))
1431 {
1432 if (codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSSTAR ||
1433 codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSQUERY)
1434 {
1435 active_count--; /* Remove non-match possibility */
1436 next_active_state--;
1437 }
1438 ADD_NEW_DATA(-(state_offset + count), 0, 0);
1439 }
1440 }
1441 break;
1442
1443 /*-----------------------------------------------------------------*/
1444 case OP_HSPACE_EXTRA + OP_TYPEQUERY:
1445 case OP_HSPACE_EXTRA + OP_TYPEMINQUERY:
1446 case OP_HSPACE_EXTRA + OP_TYPEPOSQUERY:
1447 count = 2;
1448 goto QS5;
1449
1450 case OP_HSPACE_EXTRA + OP_TYPESTAR:
1451 case OP_HSPACE_EXTRA + OP_TYPEMINSTAR:
1452 case OP_HSPACE_EXTRA + OP_TYPEPOSSTAR:
1453 count = 0;
1454
1455 QS5:
1456 ADD_ACTIVE(state_offset + 2, 0);
1457 if (clen > 0)
1458 {
1459 BOOL OK;
1460 switch (c)
1461 {
1462 case 0x09: /* HT */
1463 case 0x20: /* SPACE */
1464 case 0xa0: /* NBSP */
1465 case 0x1680: /* OGHAM SPACE MARK */
1466 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */
1467 case 0x2000: /* EN QUAD */
1468 case 0x2001: /* EM QUAD */
1469 case 0x2002: /* EN SPACE */
1470 case 0x2003: /* EM SPACE */
1471 case 0x2004: /* THREE-PER-EM SPACE */
1472 case 0x2005: /* FOUR-PER-EM SPACE */
1473 case 0x2006: /* SIX-PER-EM SPACE */
1474 case 0x2007: /* FIGURE SPACE */
1475 case 0x2008: /* PUNCTUATION SPACE */
1476 case 0x2009: /* THIN SPACE */
1477 case 0x200A: /* HAIR SPACE */
1478 case 0x202f: /* NARROW NO-BREAK SPACE */
1479 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */
1480 case 0x3000: /* IDEOGRAPHIC SPACE */
1481 OK = TRUE;
1482 break;
1483
1484 default:
1485 OK = FALSE;
1486 break;
1487 }
1488
1489 if (OK == (d == OP_HSPACE))
1490 {
1491 if (codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSSTAR ||
1492 codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSQUERY)
1493 {
1494 active_count--; /* Remove non-match possibility */
1495 next_active_state--;
1496 }
1497 ADD_NEW_DATA(-(state_offset + count), 0, 0);
1498 }
1499 }
1500 break;
1501
1502 /*-----------------------------------------------------------------*/
1503 #ifdef SUPPORT_UCP
1504 case OP_PROP_EXTRA + OP_TYPEEXACT:
1505 case OP_PROP_EXTRA + OP_TYPEUPTO:
1506 case OP_PROP_EXTRA + OP_TYPEMINUPTO:
1507 case OP_PROP_EXTRA + OP_TYPEPOSUPTO:
1508 if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)
1509 { ADD_ACTIVE(state_offset + 6, 0); }
1510 count = current_state->count; /* Number already matched */
1511 if (clen > 0)
1512 {
1513 BOOL OK;
1514 const ucd_record * prop = GET_UCD(c);
1515 switch(code[4])
1516 {
1517 case PT_ANY:
1518 OK = TRUE;
1519 break;
1520
1521 case PT_LAMP:
1522 OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1523 break;
1524
1525 case PT_GC:
1526 OK = _pcre_ucp_gentype[prop->chartype] == code[5];
1527 break;
1528
1529 case PT_PC:
1530 OK = prop->chartype == code[5];
1531 break;
1532
1533 case PT_SC:
1534 OK = prop->script == code[5];
1535 break;
1536
1537 /* Should never occur, but keep compilers from grumbling. */
1538
1539 default:
1540 OK = codevalue != OP_PROP;
1541 break;
1542 }
1543
1544 if (OK == (d == OP_PROP))
1545 {
1546 if (codevalue == OP_PROP_EXTRA + OP_TYPEPOSUPTO)
1547 {
1548 active_count--; /* Remove non-match possibility */
1549 next_active_state--;
1550 }
1551 if (++count >= GET2(code, 1))
1552 { ADD_NEW(state_offset + 6, 0); }
1553 else
1554 { ADD_NEW(state_offset, count); }
1555 }
1556 }
1557 break;
1558
1559 /*-----------------------------------------------------------------*/
1560 case OP_EXTUNI_EXTRA + OP_TYPEEXACT:
1561 case OP_EXTUNI_EXTRA + OP_TYPEUPTO:
1562 case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:
1563 case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:
1564 if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
1565 { ADD_ACTIVE(state_offset + 4, 0); }
1566 count = current_state->count; /* Number already matched */
1567 if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1568 {
1569 const uschar *nptr = ptr + clen;
1570 int ncount = 0;
1571 if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)
1572 {
1573 active_count--; /* Remove non-match possibility */
1574 next_active_state--;
1575 }
1576 while (nptr < end_subject)
1577 {
1578 int nd;
1579 int ndlen = 1;
1580 GETCHARLEN(nd, nptr, ndlen);
1581 if (UCD_CATEGORY(nd) != ucp_M) break;
1582 ncount++;
1583 nptr += ndlen;
1584 }
1585 if (++count >= GET2(code, 1))
1586 { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }
1587 else
1588 { ADD_NEW_DATA(-state_offset, count, ncount); }
1589 }
1590 break;
1591 #endif
1592
1593 /*-----------------------------------------------------------------*/
1594 case OP_ANYNL_EXTRA + OP_TYPEEXACT:
1595 case OP_ANYNL_EXTRA + OP_TYPEUPTO:
1596 case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:
1597 case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:
1598 if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)
1599 { ADD_ACTIVE(state_offset + 4, 0); }
1600 count = current_state->count; /* Number already matched */
1601 if (clen > 0)
1602 {
1603 int ncount = 0;
1604 switch (c)
1605 {
1606 case 0x000b:
1607 case 0x000c:
1608 case 0x0085:
1609 case 0x2028:
1610 case 0x2029:
1611 if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1612 goto ANYNL03;
1613
1614 case 0x000d:
1615 if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;
1616 /* Fall through */
1617
1618 ANYNL03:
1619 case 0x000a:
1620 if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSUPTO)
1621 {
1622 active_count--; /* Remove non-match possibility */
1623 next_active_state--;
1624 }
1625 if (++count >= GET2(code, 1))
1626 { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }
1627 else
1628 { ADD_NEW_DATA(-state_offset, count, ncount); }
1629 break;
1630
1631 default:
1632 break;
1633 }
1634 }
1635 break;
1636
1637 /*-----------------------------------------------------------------*/
1638 case OP_VSPACE_EXTRA + OP_TYPEEXACT:
1639 case OP_VSPACE_EXTRA + OP_TYPEUPTO:
1640 case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:
1641 case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:
1642 if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)
1643 { ADD_ACTIVE(state_offset + 4, 0); }
1644 count = current_state->count; /* Number already matched */
1645 if (clen > 0)
1646 {
1647 BOOL OK;
1648 switch (c)
1649 {
1650 case 0x000a:
1651 case 0x000b:
1652 case 0x000c:
1653 case 0x000d:
1654 case 0x0085:
1655 case 0x2028:
1656 case 0x2029:
1657 OK = TRUE;
1658 break;
1659
1660 default:
1661 OK = FALSE;
1662 }
1663
1664 if (OK == (d == OP_VSPACE))
1665 {
1666 if (codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSUPTO)
1667 {
1668 active_count--; /* Remove non-match possibility */
1669 next_active_state--;
1670 }
1671 if (++count >= GET2(code, 1))
1672 { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }
1673 else
1674 { ADD_NEW_DATA(-state_offset, count, 0); }
1675 }
1676 }
1677 break;
1678
1679 /*-----------------------------------------------------------------*/
1680 case OP_HSPACE_EXTRA + OP_TYPEEXACT:
1681 case OP_HSPACE_EXTRA + OP_TYPEUPTO:
1682 case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:
1683 case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:
1684 if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)
1685 { ADD_ACTIVE(state_offset + 4, 0); }
1686 count = current_state->count; /* Number already matched */
1687 if (clen > 0)
1688 {
1689 BOOL OK;
1690 switch (c)
1691 {
1692 case 0x09: /* HT */
1693 case 0x20: /* SPACE */
1694 case 0xa0: /* NBSP */
1695 case 0x1680: /* OGHAM SPACE MARK */
1696 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */
1697 case 0x2000: /* EN QUAD */
1698 case 0x2001: /* EM QUAD */
1699 case 0x2002: /* EN SPACE */
1700 case 0x2003: /* EM SPACE */
1701 case 0x2004: /* THREE-PER-EM SPACE */
1702 case 0x2005: /* FOUR-PER-EM SPACE */
1703 case 0x2006: /* SIX-PER-EM SPACE */
1704 case 0x2007: /* FIGURE SPACE */
1705 case 0x2008: /* PUNCTUATION SPACE */
1706 case 0x2009: /* THIN SPACE */
1707 case 0x200A: /* HAIR SPACE */
1708 case 0x202f: /* NARROW NO-BREAK SPACE */
1709 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */
1710 case 0x3000: /* IDEOGRAPHIC SPACE */
1711 OK = TRUE;
1712 break;
1713
1714 default:
1715 OK = FALSE;
1716 break;
1717 }
1718
1719 if (OK == (d == OP_HSPACE))
1720 {
1721 if (codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSUPTO)
1722 {
1723 active_count--; /* Remove non-match possibility */
1724 next_active_state--;
1725 }
1726 if (++count >= GET2(code, 1))
1727 { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }
1728 else
1729 { ADD_NEW_DATA(-state_offset, count, 0); }
1730 }
1731 }
1732 break;
1733
1734 /* ========================================================================== */
1735 /* These opcodes are followed by a character that is usually compared
1736 to the current subject character; it is loaded into d. We still get
1737 here even if there is no subject character, because in some cases zero
1738 repetitions are permitted. */
1739
1740 /*-----------------------------------------------------------------*/
1741 case OP_CHAR:
1742 if (clen > 0 && c == d) { ADD_NEW(state_offset + dlen + 1, 0); }
1743 break;
1744
1745 /*-----------------------------------------------------------------*/
1746 case OP_CHARNC:
1747 if (clen == 0) break;
1748
1749 #ifdef SUPPORT_UTF8
1750 if (utf8)
1751 {
1752 if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else
1753 {
1754 unsigned int othercase;
1755 if (c < 128) othercase = fcc[c]; else
1756
1757 /* If we have Unicode property support, we can use it to test the
1758 other case of the character. */
1759
1760 #ifdef SUPPORT_UCP
1761 othercase = UCD_OTHERCASE(c);
1762 #else
1763 othercase = NOTACHAR;
1764 #endif
1765
1766 if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }
1767 }
1768 }
1769 else
1770 #endif /* SUPPORT_UTF8 */
1771
1772 /* Non-UTF-8 mode */
1773 {
1774 if (lcc[c] == lcc[d]) { ADD_NEW(state_offset + 2, 0); }
1775 }
1776 break;
1777
1778
1779 #ifdef SUPPORT_UCP
1780 /*-----------------------------------------------------------------*/
1781 /* This is a tricky one because it can match more than one character.
1782 Find out how many characters to skip, and then set up a negative state
1783 to wait for them to pass before continuing. */
1784
1785 case OP_EXTUNI:
1786 if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1787 {
1788 const uschar *nptr = ptr + clen;
1789 int ncount = 0;
1790 while (nptr < end_subject)
1791 {
1792 int nclen = 1;
1793 GETCHARLEN(c, nptr, nclen);
1794 if (UCD_CATEGORY(c) != ucp_M) break;
1795 ncount++;
1796 nptr += nclen;
1797 }
1798 ADD_NEW_DATA(-(state_offset + 1), 0, ncount);
1799 }
1800 break;
1801 #endif
1802
1803 /*-----------------------------------------------------------------*/
1804 /* This is a tricky like EXTUNI because it too can match more than one
1805 character (when CR is followed by LF). In this case, set up a negative
1806 state to wait for one character to pass before continuing. */
1807
1808 case OP_ANYNL:
1809 if (clen > 0) switch(c)
1810 {
1811 case 0x000b:
1812 case 0x000c:
1813 case 0x0085:
1814 case 0x2028:
1815 case 0x2029:
1816 if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1817
1818 case 0x000a:
1819 ADD_NEW(state_offset + 1, 0);
1820 break;
1821
1822 case 0x000d:
1823 if (ptr + 1 < end_subject && ptr[1] == 0x0a)
1824 {
1825 ADD_NEW_DATA(-(state_offset + 1), 0, 1);
1826 }
1827 else
1828 {
1829 ADD_NEW(state_offset + 1, 0);
1830 }
1831 break;
1832 }
1833 break;
1834
1835 /*-----------------------------------------------------------------*/
1836 case OP_NOT_VSPACE:
1837 if (clen > 0) switch(c)
1838 {
1839 case 0x000a:
1840 case 0x000b:
1841 case 0x000c:
1842 case 0x000d:
1843 case 0x0085:
1844 case 0x2028:
1845 case 0x2029:
1846 break;
1847
1848 default:
1849 ADD_NEW(state_offset + 1, 0);
1850 break;
1851 }
1852 break;
1853
1854 /*-----------------------------------------------------------------*/
1855 case OP_VSPACE:
1856 if (clen > 0) switch(c)
1857 {
1858 case 0x000a:
1859 case 0x000b:
1860 case 0x000c:
1861 case 0x000d:
1862 case 0x0085:
1863 case 0x2028:
1864 case 0x2029:
1865 ADD_NEW(state_offset + 1, 0);
1866 break;
1867
1868 default: break;
1869 }
1870 break;
1871
1872 /*-----------------------------------------------------------------*/
1873 case OP_NOT_HSPACE:
1874 if (clen > 0) switch(c)
1875 {
1876 case 0x09: /* HT */
1877 case 0x20: /* SPACE */
1878 case 0xa0: /* NBSP */
1879 case 0x1680: /* OGHAM SPACE MARK */
1880 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */
1881 case 0x2000: /* EN QUAD */
1882 case 0x2001: /* EM QUAD */
1883 case 0x2002: /* EN SPACE */
1884 case 0x2003: /* EM SPACE */
1885 case 0x2004: /* THREE-PER-EM SPACE */
1886 case 0x2005: /* FOUR-PER-EM SPACE */
1887 case 0x2006: /* SIX-PER-EM SPACE */
1888 case 0x2007: /* FIGURE SPACE */
1889 case 0x2008: /* PUNCTUATION SPACE */
1890 case 0x2009: /* THIN SPACE */
1891 case 0x200A: /* HAIR SPACE */
1892 case 0x202f: /* NARROW NO-BREAK SPACE */
1893 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */
1894 case 0x3000: /* IDEOGRAPHIC SPACE */
1895 break;
1896
1897 default:
1898 ADD_NEW(state_offset + 1, 0);
1899 break;
1900 }
1901 break;
1902
1903 /*-----------------------------------------------------------------*/
1904 case OP_HSPACE:
1905 if (clen > 0) switch(c)
1906 {
1907 case 0x09: /* HT */
1908 case 0x20: /* SPACE */
1909 case 0xa0: /* NBSP */
1910 case 0x1680: /* OGHAM SPACE MARK */
1911 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */
1912 case 0x2000: /* EN QUAD */
1913 case 0x2001: /* EM QUAD */
1914 case 0x2002: /* EN SPACE */
1915 case 0x2003: /* EM SPACE */
1916 case 0x2004: /* THREE-PER-EM SPACE */
1917 case 0x2005: /* FOUR-PER-EM SPACE */
1918 case 0x2006: /* SIX-PER-EM SPACE */
1919 case 0x2007: /* FIGURE SPACE */
1920 case 0x2008: /* PUNCTUATION SPACE */
1921 case 0x2009: /* THIN SPACE */
1922 case 0x200A: /* HAIR SPACE */
1923 case 0x202f: /* NARROW NO-BREAK SPACE */
1924 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */
1925 case 0x3000: /* IDEOGRAPHIC SPACE */
1926 ADD_NEW(state_offset + 1, 0);
1927 break;
1928 }
1929 break;
1930
1931 /*-----------------------------------------------------------------*/
1932 /* Match a negated single character. This is only used for one-byte
1933 characters, that is, we know that d < 256. The character we are
1934 checking (c) can be multibyte. */
1935
1936 case OP_NOT:
1937 if (clen > 0)
1938 {
1939 unsigned int otherd = ((ims & PCRE_CASELESS) != 0)? fcc[d] : d;
1940 if (c != d && c != otherd) { ADD_NEW(state_offset + dlen + 1, 0); }
1941 }
1942 break;
1943
1944 /*-----------------------------------------------------------------*/
1945 case OP_PLUS:
1946 case OP_MINPLUS:
1947 case OP_POSPLUS:
1948 case OP_NOTPLUS:
1949 case OP_NOTMINPLUS:
1950 case OP_NOTPOSPLUS:
1951 count = current_state->count; /* Already matched */
1952 if (count > 0) { ADD_ACTIVE(state_offset + dlen + 1, 0); }
1953 if (clen > 0)
1954 {
1955 unsigned int otherd = NOTACHAR;
1956 if ((ims & PCRE_CASELESS) != 0)
1957 {
1958 #ifdef SUPPORT_UTF8
1959 if (utf8 && d >= 128)
1960 {
1961 #ifdef SUPPORT_UCP
1962 otherd = UCD_OTHERCASE(d);
1963 #endif /* SUPPORT_UCP */
1964 }
1965 else
1966 #endif /* SUPPORT_UTF8 */
1967 otherd = fcc[d];
1968 }
1969 if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
1970 {
1971 if (count > 0 &&
1972 (codevalue == OP_POSPLUS || codevalue == OP_NOTPOSPLUS))
1973 {
1974 active_count--; /* Remove non-match possibility */
1975 next_active_state--;
1976 }
1977 count++;
1978 ADD_NEW(state_offset, count);
1979 }
1980 }
1981 break;
1982
1983 /*-----------------------------------------------------------------*/
1984 case OP_QUERY:
1985 case OP_MINQUERY:
1986 case OP_POSQUERY:
1987 case OP_NOTQUERY:
1988 case OP_NOTMINQUERY:
1989 case OP_NOTPOSQUERY:
1990 ADD_ACTIVE(state_offset + dlen + 1, 0);
1991 if (clen > 0)
1992 {
1993 unsigned int otherd = NOTACHAR;
1994 if ((ims & PCRE_CASELESS) != 0)
1995 {
1996 #ifdef SUPPORT_UTF8
1997 if (utf8 && d >= 128)
1998 {
1999 #ifdef SUPPORT_UCP
2000 otherd = UCD_OTHERCASE(d);
2001 #endif /* SUPPORT_UCP */
2002 }
2003 else
2004 #endif /* SUPPORT_UTF8 */
2005 otherd = fcc[d];
2006 }
2007 if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2008 {
2009 if (codevalue == OP_POSQUERY || codevalue == OP_NOTPOSQUERY)
2010 {
2011 active_count--; /* Remove non-match possibility */
2012 next_active_state--;
2013 }
2014 ADD_NEW(state_offset + dlen + 1, 0);
2015 }
2016 }
2017 break;
2018
2019 /*-----------------------------------------------------------------*/
2020 case OP_STAR:
2021 case OP_MINSTAR:
2022 case OP_POSSTAR:
2023 case OP_NOTSTAR:
2024 case OP_NOTMINSTAR:
2025 case OP_NOTPOSSTAR:
2026 ADD_ACTIVE(state_offset + dlen + 1, 0);
2027 if (clen > 0)
2028 {
2029 unsigned int otherd = NOTACHAR;
2030 if ((ims & PCRE_CASELESS) != 0)
2031 {
2032 #ifdef SUPPORT_UTF8
2033 if (utf8 && d >= 128)
2034 {
2035 #ifdef SUPPORT_UCP
2036 otherd = UCD_OTHERCASE(d);
2037 #endif /* SUPPORT_UCP */
2038 }
2039 else
2040 #endif /* SUPPORT_UTF8 */
2041 otherd = fcc[d];
2042 }
2043 if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2044 {
2045 if (codevalue == OP_POSSTAR || codevalue == OP_NOTPOSSTAR)
2046 {
2047 active_count--; /* Remove non-match possibility */
2048 next_active_state--;
2049 }
2050 ADD_NEW(state_offset, 0);
2051 }
2052 }
2053 break;
2054
2055 /*-----------------------------------------------------------------*/
2056 case OP_EXACT:
2057 case OP_NOTEXACT:
2058 count = current_state->count; /* Number already matched */
2059 if (clen > 0)
2060 {
2061 unsigned int otherd = NOTACHAR;
2062 if ((ims & PCRE_CASELESS) != 0)
2063 {
2064 #ifdef SUPPORT_UTF8
2065 if (utf8 && d >= 128)
2066 {
2067 #ifdef SUPPORT_UCP
2068 otherd = UCD_OTHERCASE(d);
2069 #endif /* SUPPORT_UCP */
2070 }
2071 else
2072 #endif /* SUPPORT_UTF8 */
2073 otherd = fcc[d];
2074 }
2075 if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2076 {
2077 if (++count >= GET2(code, 1))
2078 { ADD_NEW(state_offset + dlen + 3, 0); }
2079 else
2080 { ADD_NEW(state_offset, count); }
2081 }
2082 }
2083 break;
2084
2085 /*-----------------------------------------------------------------*/
2086 case OP_UPTO:
2087 case OP_MINUPTO:
2088 case OP_POSUPTO:
2089 case OP_NOTUPTO:
2090 case OP_NOTMINUPTO:
2091 case OP_NOTPOSUPTO:
2092 ADD_ACTIVE(state_offset + dlen + 3, 0);
2093 count = current_state->count; /* Number already matched */
2094 if (clen > 0)
2095 {
2096 unsigned int otherd = NOTACHAR;
2097 if ((ims & PCRE_CASELESS) != 0)
2098 {
2099 #ifdef SUPPORT_UTF8
2100 if (utf8 && d >= 128)
2101 {
2102 #ifdef SUPPORT_UCP
2103 otherd = UCD_OTHERCASE(d);
2104 #endif /* SUPPORT_UCP */
2105 }
2106 else
2107 #endif /* SUPPORT_UTF8 */
2108 otherd = fcc[d];
2109 }
2110 if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2111 {
2112 if (codevalue == OP_POSUPTO || codevalue == OP_NOTPOSUPTO)
2113 {
2114 active_count--; /* Remove non-match possibility */
2115 next_active_state--;
2116 }
2117 if (++count >= GET2(code, 1))
2118 { ADD_NEW(state_offset + dlen + 3, 0); }
2119 else
2120 { ADD_NEW(state_offset, count); }
2121 }
2122 }
2123 break;
2124
2125
2126 /* ========================================================================== */
2127 /* These are the class-handling opcodes */
2128
2129 case OP_CLASS:
2130 case OP_NCLASS:
2131 case OP_XCLASS:
2132 {
2133 BOOL isinclass = FALSE;
2134 int next_state_offset;
2135 const uschar *ecode;
2136
2137 /* For a simple class, there is always just a 32-byte table, and we
2138 can set isinclass from it. */
2139
2140 if (codevalue != OP_XCLASS)
2141 {
2142 ecode = code + 33;
2143 if (clen > 0)
2144 {
2145 isinclass = (c > 255)? (codevalue == OP_NCLASS) :
2146 ((code[1 + c/8] & (1 << (c&7))) != 0);
2147 }
2148 }
2149
2150 /* An extended class may have a table or a list of single characters,
2151 ranges, or both, and it may be positive or negative. There's a
2152 function that sorts all this out. */
2153
2154 else
2155 {
2156 ecode = code + GET(code, 1);
2157 if (clen > 0) isinclass = _pcre_xclass(c, code + 1 + LINK_SIZE);
2158 }
2159
2160 /* At this point, isinclass is set for all kinds of class, and ecode
2161 points to the byte after the end of the class. If there is a
2162 quantifier, this is where it will be. */
2163
2164 next_state_offset = ecode - start_code;
2165
2166 switch (*ecode)
2167 {
2168 case OP_CRSTAR:
2169 case OP_CRMINSTAR:
2170 ADD_ACTIVE(next_state_offset + 1, 0);
2171 if (isinclass) { ADD_NEW(state_offset, 0); }
2172 break;
2173
2174 case OP_CRPLUS:
2175 case OP_CRMINPLUS:
2176 count = current_state->count; /* Already matched */
2177 if (count > 0) { ADD_ACTIVE(next_state_offset + 1, 0); }
2178 if (isinclass) { count++; ADD_NEW(state_offset, count); }
2179 break;
2180
2181 case OP_CRQUERY:
2182 case OP_CRMINQUERY:
2183 ADD_ACTIVE(next_state_offset + 1, 0);
2184 if (isinclass) { ADD_NEW(next_state_offset + 1, 0); }
2185 break;
2186
2187 case OP_CRRANGE:
2188 case OP_CRMINRANGE:
2189 count = current_state->count; /* Already matched */
2190 if (count >= GET2(ecode, 1))
2191 { ADD_ACTIVE(next_state_offset + 5, 0); }
2192 if (isinclass)
2193 {
2194 int max = GET2(ecode, 3);
2195 if (++count >= max && max != 0) /* Max 0 => no limit */
2196 { ADD_NEW(next_state_offset + 5, 0); }
2197 else
2198 { ADD_NEW(state_offset, count); }
2199 }
2200 break;
2201
2202 default:
2203 if (isinclass) { ADD_NEW(next_state_offset, 0); }
2204 break;
2205 }
2206 }
2207 break;
2208
2209 /* ========================================================================== */
2210 /* These are the opcodes for fancy brackets of various kinds. We have
2211 to use recursion in order to handle them. The "always failing" assertion
2212 (?!) is optimised to OP_FAIL when compiling, so we have to support that,
2213 though the other "backtracking verbs" are not supported. */
2214
2215 case OP_FAIL:
2216 forced_fail++; /* Count FAILs for multiple states */
2217 break;
2218
2219 case OP_ASSERT:
2220 case OP_ASSERT_NOT:
2221 case OP_ASSERTBACK:
2222 case OP_ASSERTBACK_NOT:
2223 {
2224 int rc;
2225 int local_offsets[2];
2226 int local_workspace[1000];
2227 const uschar *endasscode = code + GET(code, 1);
2228
2229 while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2230
2231 rc = internal_dfa_exec(
2232 md, /* static match data */
2233 code, /* this subexpression's code */
2234 ptr, /* where we currently are */
2235 ptr - start_subject, /* start offset */
2236 local_offsets, /* offset vector */
2237 sizeof(local_offsets)/sizeof(int), /* size of same */
2238 local_workspace, /* workspace vector */
2239 sizeof(local_workspace)/sizeof(int), /* size of same */
2240 ims, /* the current ims flags */
2241 rlevel, /* function recursion level */
2242 recursing); /* pass on regex recursion */
2243
2244 if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))
2245 { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }
2246 }
2247 break;
2248
2249 /*-----------------------------------------------------------------*/
2250 case OP_COND:
2251 case OP_SCOND:
2252 {
2253 int local_offsets[1000];
2254 int local_workspace[1000];
2255 int codelink = GET(code, 1);
2256 int condcode;
2257
2258 /* Because of the way auto-callout works during compile, a callout item
2259 is inserted between OP_COND and an assertion condition. This does not
2260 happen for the other conditions. */
2261
2262 if (code[LINK_SIZE+1] == OP_CALLOUT)
2263 {
2264 rrc = 0;
2265 if (pcre_callout != NULL)
2266 {
2267 pcre_callout_block cb;
2268 cb.version = 1; /* Version 1 of the callout block */
2269 cb.callout_number = code[LINK_SIZE+2];
2270 cb.offset_vector = offsets;
2271 cb.subject = (PCRE_SPTR)start_subject;
2272 cb.subject_length = end_subject - start_subject;
2273 cb.start_match = current_subject - start_subject;
2274 cb.current_position = ptr - start_subject;
2275 cb.pattern_position = GET(code, LINK_SIZE + 3);
2276 cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);
2277 cb.capture_top = 1;
2278 cb.capture_last = -1;
2279 cb.callout_data = md->callout_data;
2280 if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc; /* Abandon */
2281 }
2282 if (rrc > 0) break; /* Fail this thread */
2283 code += _pcre_OP_lengths[OP_CALLOUT]; /* Skip callout data */
2284 }
2285
2286 condcode = code[LINK_SIZE+1];
2287
2288 /* Back reference conditions are not supported */
2289
2290 if (condcode == OP_CREF || condcode == OP_NCREF)
2291 return PCRE_ERROR_DFA_UCOND;
2292
2293 /* The DEFINE condition is always false */
2294
2295 if (condcode == OP_DEF)
2296 { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2297
2298 /* The only supported version of OP_RREF is for the value RREF_ANY,
2299 which means "test if in any recursion". We can't test for specifically
2300 recursed groups. */
2301
2302 else if (condcode == OP_RREF || condcode == OP_NRREF)
2303 {
2304 int value = GET2(code, LINK_SIZE+2);
2305 if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
2306 if (recursing > 0)
2307 { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }
2308 else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2309 }
2310
2311 /* Otherwise, the condition is an assertion */
2312
2313 else
2314 {
2315 int rc;
2316 const uschar *asscode = code + LINK_SIZE + 1;
2317 const uschar *endasscode = asscode + GET(asscode, 1);
2318
2319 while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2320
2321 rc = internal_dfa_exec(
2322 md, /* fixed match data */
2323 asscode, /* this subexpression's code */
2324 ptr, /* where we currently are */
2325 ptr - start_subject, /* start offset */
2326 local_offsets, /* offset vector */
2327 sizeof(local_offsets)/sizeof(int), /* size of same */
2328 local_workspace, /* workspace vector */
2329 sizeof(local_workspace)/sizeof(int), /* size of same */
2330 ims, /* the current ims flags */
2331 rlevel, /* function recursion level */
2332 recursing); /* pass on regex recursion */
2333
2334 if ((rc >= 0) ==
2335 (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))
2336 { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }
2337 else
2338 { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2339 }
2340 }
2341 break;
2342
2343 /*-----------------------------------------------------------------*/
2344 case OP_RECURSE:
2345 {
2346 int local_offsets[1000];
2347 int local_workspace[1000];
2348 int rc;
2349
2350 DPRINTF(("%.*sStarting regex recursion %d\n", rlevel*2-2, SP,
2351 recursing + 1));
2352
2353 rc = internal_dfa_exec(
2354 md, /* fixed match data */
2355 start_code + GET(code, 1), /* this subexpression's code */
2356 ptr, /* where we currently are */
2357 ptr - start_subject, /* start offset */
2358 local_offsets, /* offset vector */
2359 sizeof(local_offsets)/sizeof(int), /* size of same */
2360 local_workspace, /* workspace vector */
2361 sizeof(local_workspace)/sizeof(int), /* size of same */
2362 ims, /* the current ims flags */
2363 rlevel, /* function recursion level */
2364 recursing + 1); /* regex recurse level */
2365
2366 DPRINTF(("%.*sReturn from regex recursion %d: rc=%d\n", rlevel*2-2, SP,
2367 recursing + 1, rc));
2368
2369 /* Ran out of internal offsets */
2370
2371 if (rc == 0) return PCRE_ERROR_DFA_RECURSE;
2372
2373 /* For each successful matched substring, set up the next state with a
2374 count of characters to skip before trying it. Note that the count is in
2375 characters, not bytes. */
2376
2377 if (rc > 0)
2378 {
2379 for (rc = rc*2 - 2; rc >= 0; rc -= 2)
2380 {
2381 const uschar *p = start_subject + local_offsets[rc];
2382 const uschar *pp = start_subject + local_offsets[rc+1];
2383 int charcount = local_offsets[rc+1] - local_offsets[rc];
2384 while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;
2385 if (charcount > 0)
2386 {
2387 ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));
2388 }
2389 else
2390 {
2391 ADD_ACTIVE(state_offset + LINK_SIZE + 1, 0);
2392 }
2393 }
2394 }
2395 else if (rc != PCRE_ERROR_NOMATCH) return rc;
2396 }
2397 break;
2398
2399 /*-----------------------------------------------------------------*/
2400 case OP_ONCE:
2401 {
2402 int local_offsets[2];
2403 int local_workspace[1000];
2404
2405 int rc = internal_dfa_exec(
2406 md, /* fixed match data */
2407 code, /* this subexpression's code */
2408 ptr, /* where we currently are */
2409 ptr - start_subject, /* start offset */
2410 local_offsets, /* offset vector */
2411 sizeof(local_offsets)/sizeof(int), /* size of same */
2412 local_workspace, /* workspace vector */
2413 sizeof(local_workspace)/sizeof(int), /* size of same */
2414 ims, /* the current ims flags */
2415 rlevel, /* function recursion level */
2416 recursing); /* pass on regex recursion */
2417
2418 if (rc >= 0)
2419 {
2420 const uschar *end_subpattern = code;
2421 int charcount = local_offsets[1] - local_offsets[0];
2422 int next_state_offset, repeat_state_offset;
2423
2424 do { end_subpattern += GET(end_subpattern, 1); }
2425 while (*end_subpattern == OP_ALT);
2426 next_state_offset = end_subpattern - start_code + LINK_SIZE + 1;
2427
2428 /* If the end of this subpattern is KETRMAX or KETRMIN, we must
2429 arrange for the repeat state also to be added to the relevant list.
2430 Calculate the offset, or set -1 for no repeat. */
2431
2432 repeat_state_offset = (*end_subpattern == OP_KETRMAX ||
2433 *end_subpattern == OP_KETRMIN)?
2434 end_subpattern - start_code - GET(end_subpattern, 1) : -1;
2435
2436 /* If we have matched an empty string, add the next state at the
2437 current character pointer. This is important so that the duplicate
2438 checking kicks in, which is what breaks infinite loops that match an
2439 empty string. */
2440
2441 if (charcount == 0)
2442 {
2443 ADD_ACTIVE(next_state_offset, 0);
2444 }
2445
2446 /* Optimization: if there are no more active states, and there
2447 are no new states yet set up, then skip over the subject string
2448 right here, to save looping. Otherwise, set up the new state to swing
2449 into action when the end of the substring is reached. */
2450
2451 else if (i + 1 >= active_count && new_count == 0)
2452 {
2453 ptr += charcount;
2454 clen = 0;
2455 ADD_NEW(next_state_offset, 0);
2456
2457 /* If we are adding a repeat state at the new character position,
2458 we must fudge things so that it is the only current state.
2459 Otherwise, it might be a duplicate of one we processed before, and
2460 that would cause it to be skipped. */
2461
2462 if (repeat_state_offset >= 0)
2463 {
2464 next_active_state = active_states;
2465 active_count = 0;
2466 i = -1;
2467 ADD_ACTIVE(repeat_state_offset, 0);
2468 }
2469 }
2470 else
2471 {
2472 const uschar *p = start_subject + local_offsets[0];
2473 const uschar *pp = start_subject + local_offsets[1];
2474 while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;
2475 ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2476 if (repeat_state_offset >= 0)
2477 { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }
2478 }
2479
2480 }
2481 else if (rc != PCRE_ERROR_NOMATCH) return rc;
2482 }
2483 break;
2484
2485
2486 /* ========================================================================== */
2487 /* Handle callouts */
2488
2489 case OP_CALLOUT:
2490 rrc = 0;
2491 if (pcre_callout != NULL)
2492 {
2493 pcre_callout_block cb;
2494 cb.version = 1; /* Version 1 of the callout block */
2495 cb.callout_number = code[1];
2496 cb.offset_vector = offsets;
2497 cb.subject = (PCRE_SPTR)start_subject;
2498 cb.subject_length = end_subject - start_subject;
2499 cb.start_match = current_subject - start_subject;
2500 cb.current_position = ptr - start_subject;
2501 cb.pattern_position = GET(code, 2);
2502 cb.next_item_length = GET(code, 2 + LINK_SIZE);
2503 cb.capture_top = 1;
2504 cb.capture_last = -1;
2505 cb.callout_data = md->callout_data;
2506 if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc; /* Abandon */
2507 }
2508 if (rrc == 0)
2509 { ADD_ACTIVE(state_offset + _pcre_OP_lengths[OP_CALLOUT], 0); }
2510 break;
2511
2512
2513 /* ========================================================================== */
2514 default: /* Unsupported opcode */
2515 return PCRE_ERROR_DFA_UITEM;
2516 }
2517
2518 NEXT_ACTIVE_STATE: continue;
2519
2520 } /* End of loop scanning active states */
2521
2522 /* We have finished the processing at the current subject character. If no
2523 new states have been set for the next character, we have found all the
2524 matches that we are going to find. If we are at the top level and partial
2525 matching has been requested, check for appropriate conditions. The "forced_
2526 fail" variable counts the number of (*F) encountered for the character. If it
2527 is equal to the original active_count (saved in workspace[1]) it means that
2528 (*F) was found on every active state. In this case we don't want to give a
2529 partial match. */
2530
2531 if (new_count <= 0)
2532 {
2533 if (rlevel == 1 && /* Top level, and */
2534 reached_end != workspace[1] && /* Not all reached end */
2535 forced_fail != workspace[1] && /* Not all forced fail & */
2536 ( /* either... */
2537 (md->moptions & PCRE_PARTIAL_HARD) != 0 /* Hard partial */
2538 || /* or... */
2539 ((md->moptions & PCRE_PARTIAL_SOFT) != 0 && /* Soft partial and */
2540 match_count < 0) /* no matches */
2541 ) && /* And... */
2542 ptr >= end_subject && /* Reached end of subject */
2543 ptr > current_subject) /* Matched non-empty string */
2544 {
2545 if (offsetcount >= 2)
2546 {
2547 offsets[0] = md->start_used_ptr - start_subject;
2548 offsets[1] = end_subject - start_subject;
2549 }
2550 match_count = PCRE_ERROR_PARTIAL;
2551 }
2552
2553 DPRINTF(("%.*sEnd of internal_dfa_exec %d: returning %d\n"
2554 "%.*s---------------------\n\n", rlevel*2-2, SP, rlevel, match_count,
2555 rlevel*2-2, SP));
2556 break; /* In effect, "return", but see the comment below */
2557 }
2558
2559 /* One or more states are active for the next character. */
2560
2561 ptr += clen; /* Advance to next subject character */
2562 } /* Loop to move along the subject string */
2563
2564 /* Control gets here from "break" a few lines above. We do it this way because
2565 if we use "return" above, we have compiler trouble. Some compilers warn if
2566 there's nothing here because they think the function doesn't return a value. On
2567 the other hand, if we put a dummy statement here, some more clever compilers
2568 complain that it can't be reached. Sigh. */
2569
2570 return match_count;
2571 }
2572
2573
2574
2575
2576 /*************************************************
2577 * Execute a Regular Expression - DFA engine *
2578 *************************************************/
2579
2580 /* This external function applies a compiled re to a subject string using a DFA
2581 engine. This function calls the internal function multiple times if the pattern
2582 is not anchored.
2583
2584 Arguments:
2585 argument_re points to the compiled expression
2586 extra_data points to extra data or is NULL
2587 subject points to the subject string
2588 length length of subject string (may contain binary zeros)
2589 start_offset where to start in the subject string
2590 options option bits
2591 offsets vector of match offsets
2592 offsetcount size of same
2593 workspace workspace vector
2594 wscount size of same
2595
2596 Returns: > 0 => number of match offset pairs placed in offsets
2597 = 0 => offsets overflowed; longest matches are present
2598 -1 => failed to match
2599 < -1 => some kind of unexpected problem
2600 */
2601
2602 PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
2603 pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
2604 const char *subject, int length, int start_offset, int options, int *offsets,
2605 int offsetcount, int *workspace, int wscount)
2606 {
2607 real_pcre *re = (real_pcre *)argument_re;
2608 dfa_match_data match_block;
2609 dfa_match_data *md = &match_block;
2610 BOOL utf8, anchored, startline, firstline;
2611 const uschar *current_subject, *end_subject, *lcc;
2612
2613 pcre_study_data internal_study;
2614 const pcre_study_data *study = NULL;
2615 real_pcre internal_re;
2616
2617 const uschar *req_byte_ptr;
2618 const uschar *start_bits = NULL;
2619 BOOL first_byte_caseless = FALSE;
2620 BOOL req_byte_caseless = FALSE;
2621 int first_byte = -1;
2622 int req_byte = -1;
2623 int req_byte2 = -1;
2624 int newline;
2625
2626 /* Plausibility checks */
2627
2628 if ((options & ~PUBLIC_DFA_EXEC_OPTIONS) != 0) return PCRE_ERROR_BADOPTION;
2629 if (re == NULL || subject == NULL || workspace == NULL ||
2630 (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
2631 if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
2632 if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;
2633
2634 /* We need to find the pointer to any study data before we test for byte
2635 flipping, so we scan the extra_data block first. This may set two fields in the
2636 match block, so we must initialize them beforehand. However, the other fields
2637 in the match block must not be set until after the byte flipping. */
2638
2639 md->tables = re->tables;
2640 md->callout_data = NULL;
2641
2642 if (extra_data != NULL)
2643 {
2644 unsigned int flags = extra_data->flags;
2645 if ((flags & PCRE_EXTRA_STUDY_DATA) != 0)
2646 study = (const pcre_study_data *)extra_data->study_data;
2647 if ((flags & PCRE_EXTRA_MATCH_LIMIT) != 0) return PCRE_ERROR_DFA_UMLIMIT;
2648 if ((flags & PCRE_EXTRA_MATCH_LIMIT_RECURSION) != 0)
2649 return PCRE_ERROR_DFA_UMLIMIT;
2650 if ((flags & PCRE_EXTRA_CALLOUT_DATA) != 0)
2651 md->callout_data = extra_data->callout_data;
2652 if ((flags & PCRE_EXTRA_TABLES) != 0)
2653 md->tables = extra_data->tables;
2654 }
2655
2656 /* Check that the first field in the block is the magic number. If it is not,
2657 test for a regex that was compiled on a host of opposite endianness. If this is
2658 the case, flipped values are put in internal_re and internal_study if there was
2659 study data too. */
2660
2661 if (re->magic_number != MAGIC_NUMBER)
2662 {
2663 re = _pcre_try_flipped(re, &internal_re, study, &internal_study);
2664 if (re == NULL) return PCRE_ERROR_BADMAGIC;
2665 if (study != NULL) study = &internal_study;
2666 }
2667
2668 /* Set some local values */
2669
2670 current_subject = (const unsigned char *)subject + start_offset;
2671 end_subject = (const unsigned char *)subject + length;
2672 req_byte_ptr = current_subject - 1;
2673
2674 #ifdef SUPPORT_UTF8
2675 utf8 = (re->options & PCRE_UTF8) != 0;
2676 #else
2677 utf8 = FALSE;
2678 #endif
2679
2680 anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||
2681 (re->options & PCRE_ANCHORED) != 0;
2682
2683 /* The remaining fixed data for passing around. */
2684
2685 md->start_code = (const uschar *)argument_re +
2686 re->name_table_offset + re->name_count * re->name_entry_size;
2687 md->start_subject = (const unsigned char *)subject;
2688 md->end_subject = end_subject;
2689 md->start_offset = start_offset;
2690 md->moptions = options;
2691 md->poptions = re->options;
2692
2693 /* If the BSR option is not set at match time, copy what was set
2694 at compile time. */
2695
2696 if ((md->moptions & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) == 0)
2697 {
2698 if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
2699 md->moptions |= re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE);
2700 #ifdef BSR_ANYCRLF
2701 else md->moptions |= PCRE_BSR_ANYCRLF;
2702 #endif
2703 }
2704
2705 /* Handle different types of newline. The three bits give eight cases. If
2706 nothing is set at run time, whatever was used at compile time applies. */
2707
2708 switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &
2709 PCRE_NEWLINE_BITS)
2710 {
2711 case 0: newline = NEWLINE; break; /* Compile-time default */
2712 case PCRE_NEWLINE_CR: newline = CHAR_CR; break;
2713 case PCRE_NEWLINE_LF: newline = CHAR_NL; break;
2714 case PCRE_NEWLINE_CR+
2715 PCRE_NEWLINE_LF: newline = (CHAR_CR << 8) | CHAR_NL; break;
2716 case PCRE_NEWLINE_ANY: newline = -1; break;
2717 case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
2718 default: return PCRE_ERROR_BADNEWLINE;
2719 }
2720
2721 if (newline == -2)
2722 {
2723 md->nltype = NLTYPE_ANYCRLF;
2724 }
2725 else if (newline < 0)
2726 {
2727 md->nltype = NLTYPE_ANY;
2728 }
2729 else
2730 {
2731 md->nltype = NLTYPE_FIXED;
2732 if (newline > 255)
2733 {
2734 md->nllen = 2;
2735 md->nl[0] = (newline >> 8) & 255;
2736 md->nl[1] = newline & 255;
2737 }
2738 else
2739 {
2740 md->nllen = 1;
2741 md->nl[0] = newline;
2742 }
2743 }
2744
2745 /* Check a UTF-8 string if required. Unfortunately there's no way of passing
2746 back the character offset. */
2747
2748 #ifdef SUPPORT_UTF8
2749 if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)
2750 {
2751 if (_pcre_valid_utf8((uschar *)subject, length) >= 0)
2752 return PCRE_ERROR_BADUTF8;
2753 if (start_offset > 0 && start_offset < length)
2754 {
2755 int tb = ((uschar *)subject)[start_offset];
2756 if (tb > 127)
2757 {
2758 tb &= 0xc0;
2759 if (tb != 0 && tb != 0xc0) return PCRE_ERROR_BADUTF8_OFFSET;
2760 }
2761 }
2762 }
2763 #endif
2764
2765 /* If the exec call supplied NULL for tables, use the inbuilt ones. This
2766 is a feature that makes it possible to save compiled regex and re-use them
2767 in other programs later. */
2768
2769 if (md->tables == NULL) md->tables = _pcre_default_tables;
2770
2771 /* The lower casing table and the "must be at the start of a line" flag are
2772 used in a loop when finding where to start. */
2773
2774 lcc = md->tables + lcc_offset;
2775 startline = (re->flags & PCRE_STARTLINE) != 0;
2776 firstline = (re->options & PCRE_FIRSTLINE) != 0;
2777
2778 /* Set up the first character to match, if available. The first_byte value is
2779 never set for an anchored regular expression, but the anchoring may be forced
2780 at run time, so we have to test for anchoring. The first char may be unset for
2781 an unanchored pattern, of course. If there's no first char and the pattern was
2782 studied, there may be a bitmap of possible first characters. */
2783
2784 if (!anchored)
2785 {
2786 if ((re->flags & PCRE_FIRSTSET) != 0)
2787 {
2788 first_byte = re->first_byte & 255;
2789 if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)
2790 first_byte = lcc[first_byte];
2791 }
2792 else
2793 {
2794 if (!startline && study != NULL &&
2795 (study->flags & PCRE_STUDY_MAPPED) != 0)
2796 start_bits = study->start_bits;
2797 }
2798 }
2799
2800 /* For anchored or unanchored matches, there may be a "last known required
2801 character" set. */
2802
2803 if ((re->flags & PCRE_REQCHSET) != 0)
2804 {
2805 req_byte = re->req_byte & 255;
2806 req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;
2807 req_byte2 = (md->tables + fcc_offset)[req_byte]; /* case flipped */
2808 }
2809
2810 /* Call the main matching function, looping for a non-anchored regex after a
2811 failed match. If not restarting, perform certain optimizations at the start of
2812 a match. */
2813
2814 for (;;)
2815 {
2816 int rc;
2817
2818 if ((options & PCRE_DFA_RESTART) == 0)
2819 {
2820 const uschar *save_end_subject = end_subject;
2821
2822 /* If firstline is TRUE, the start of the match is constrained to the first
2823 line of a multiline string. Implement this by temporarily adjusting
2824 end_subject so that we stop scanning at a newline. If the match fails at
2825 the newline, later code breaks this loop. */
2826
2827 if (firstline)
2828 {
2829 USPTR t = current_subject;
2830 #ifdef SUPPORT_UTF8
2831 if (utf8)
2832 {
2833 while (t < md->end_subject && !IS_NEWLINE(t))
2834 {
2835 t++;
2836 while (t < end_subject && (*t & 0xc0) == 0x80) t++;
2837 }
2838 }
2839 else
2840 #endif
2841 while (t < md->end_subject && !IS_NEWLINE(t)) t++;
2842 end_subject = t;
2843 }
2844
2845 /* There are some optimizations that avoid running the match if a known
2846 starting point is not found. However, there is an option that disables
2847 these, for testing and for ensuring that all callouts do actually occur. */
2848
2849 if ((options & PCRE_NO_START_OPTIMIZE) == 0)
2850 {
2851 /* Advance to a known first byte. */
2852
2853 if (first_byte >= 0)
2854 {
2855 if (first_byte_caseless)
2856 while (current_subject < end_subject &&
2857 lcc[*current_subject] != first_byte)
2858 current_subject++;
2859 else
2860 while (current_subject < end_subject &&
2861 *current_subject != first_byte)
2862 current_subject++;
2863 }
2864
2865 /* Or to just after a linebreak for a multiline match if possible */
2866
2867 else if (startline)
2868 {
2869 if (current_subject > md->start_subject + start_offset)
2870 {
2871 #ifdef SUPPORT_UTF8
2872 if (utf8)
2873 {
2874 while (current_subject < end_subject &&
2875 !WAS_NEWLINE(current_subject))
2876 {
2877 current_subject++;
2878 while(current_subject < end_subject &&
2879 (*current_subject & 0xc0) == 0x80)
2880 current_subject++;
2881 }
2882 }
2883 else
2884 #endif
2885 while (current_subject < end_subject && !WAS_NEWLINE(current_subject))
2886 current_subject++;
2887
2888 /* If we have just passed a CR and the newline option is ANY or
2889 ANYCRLF, and we are now at a LF, advance the match position by one
2890 more character. */
2891
2892 if (current_subject[-1] == CHAR_CR &&
2893 (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
2894 current_subject < end_subject &&
2895 *current_subject == CHAR_NL)
2896 current_subject++;
2897 }
2898 }
2899
2900 /* Or to a non-unique first char after study */
2901
2902 else if (start_bits != NULL)
2903 {
2904 while (current_subject < end_subject)
2905 {
2906 register unsigned int c = *current_subject;
2907 if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;
2908 else break;
2909 }
2910 }
2911 }
2912
2913 /* Restore fudged end_subject */
2914
2915 end_subject = save_end_subject;
2916
2917 /* The following two optimizations are disabled for partial matching or if
2918 disabling is explicitly requested (and of course, by the test above, this
2919 code is not obeyed when restarting after a partial match). */
2920
2921 if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&
2922 (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)
2923 {
2924 /* If the pattern was studied, a minimum subject length may be set. This
2925 is a lower bound; no actual string of that length may actually match the
2926 pattern. Although the value is, strictly, in characters, we treat it as
2927 bytes to avoid spending too much time in this optimization. */
2928
2929 if (study != NULL && (study->flags & PCRE_STUDY_MINLEN) != 0 &&
2930 end_subject - current_subject < study->minlength)
2931 return PCRE_ERROR_NOMATCH;
2932
2933 /* If req_byte is set, we know that that character must appear in the
2934 subject for the match to succeed. If the first character is set, req_byte
2935 must be later in the subject; otherwise the test starts at the match
2936 point. This optimization can save a huge amount of work in patterns with
2937 nested unlimited repeats that aren't going to match. Writing separate
2938 code for cased/caseless versions makes it go faster, as does using an
2939 autoincrement and backing off on a match.
2940
2941 HOWEVER: when the subject string is very, very long, searching to its end
2942 can take a long time, and give bad performance on quite ordinary
2943 patterns. This showed up when somebody was matching /^C/ on a 32-megabyte
2944 string... so we don't do this when the string is sufficiently long. */
2945
2946 if (req_byte >= 0 && end_subject - current_subject < REQ_BYTE_MAX)
2947 {
2948 register const uschar *p = current_subject + ((first_byte >= 0)? 1 : 0);
2949
2950 /* We don't need to repeat the search if we haven't yet reached the
2951 place we found it at last time. */
2952
2953 if (p > req_byte_ptr)
2954 {
2955 if (req_byte_caseless)
2956 {
2957 while (p < end_subject)
2958 {
2959 register int pp = *p++;
2960 if (pp == req_byte || pp == req_byte2) { p--; break; }
2961 }
2962 }
2963 else
2964 {
2965 while (p < end_subject)
2966 {
2967 if (*p++ == req_byte) { p--; break; }
2968 }
2969 }
2970
2971 /* If we can't find the required character, break the matching loop,
2972 which will cause a return or PCRE_ERROR_NOMATCH. */
2973
2974 if (p >= end_subject) break;
2975
2976 /* If we have found the required character, save the point where we
2977 found it, so that we don't search again next time round the loop if
2978 the start hasn't passed this character yet. */
2979
2980 req_byte_ptr = p;
2981 }
2982 }
2983 }
2984 } /* End of optimizations that are done when not restarting */
2985
2986 /* OK, now we can do the business */
2987
2988 md->start_used_ptr = current_subject;
2989
2990 rc = internal_dfa_exec(
2991 md, /* fixed match data */
2992 md->start_code, /* this subexpression's code */
2993 current_subject, /* where we currently are */
2994 start_offset, /* start offset in subject */
2995 offsets, /* offset vector */
2996 offsetcount, /* size of same */
2997 workspace, /* workspace vector */
2998 wscount, /* size of same */
2999 re->options & (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL), /* ims flags */
3000 0, /* function recurse level */
3001 0); /* regex recurse level */
3002
3003 /* Anything other than "no match" means we are done, always; otherwise, carry
3004 on only if not anchored. */
3005
3006 if (rc != PCRE_ERROR_NOMATCH || anchored) return rc;
3007
3008 /* Advance to the next subject character unless we are at the end of a line
3009 and firstline is set. */
3010
3011 if (firstline && IS_NEWLINE(current_subject)) break;
3012 current_subject++;
3013 if (utf8)
3014 {
3015 while (current_subject < end_subject && (*current_subject & 0xc0) == 0x80)
3016 current_subject++;
3017 }
3018 if (current_subject > end_subject) break;
3019
3020 /* If we have just passed a CR and we are now at a LF, and the pattern does
3021 not contain any explicit matches for \r or \n, and the newline option is CRLF
3022 or ANY or ANYCRLF, advance the match position by one more character. */
3023
3024 if (current_subject[-1] == CHAR_CR &&
3025 current_subject < end_subject &&
3026 *current_subject == CHAR_NL &&
3027 (re->flags & PCRE_HASCRORLF) == 0 &&
3028 (md->nltype == NLTYPE_ANY ||
3029 md->nltype == NLTYPE_ANYCRLF ||
3030 md->nllen == 2))
3031 current_subject++;
3032
3033 } /* "Bumpalong" loop */
3034
3035 return PCRE_ERROR_NOMATCH;
3036 }
3037
3038 /* End of pcre_dfa_exec.c */

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"
  ViewVC Help
Powered by ViewVC 1.1.5