5 * "One Ring to rule them all, One Ring to find them..."
8 /* This file contains functions for executing a regular expression. See
9 * also regcomp.c which funnily enough, contains functions for compiling
10 * a regular expression.
12 * This file is also copied at build time to ext/re/re_exec.c, where
13 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
14 * This causes the main functions to be compiled under new names and with
15 * debugging support added, which makes "use re 'debug'" work.
18 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
19 * confused with the original package (see point 3 below). Thanks, Henry!
22 /* Additional note: this code is very heavily munged from Henry's version
23 * in places. In some spots I've traded clarity for efficiency, so don't
24 * blame Henry for some of the lack of readability.
27 /* The names of the functions have been changed from regcomp and
28 * regexec to pregcomp and pregexec in order to avoid conflicts
29 * with the POSIX routines of the same names.
32 #ifdef PERL_EXT_RE_BUILD
37 * pregcomp and pregexec -- regsub and regerror are not used in perl
39 * Copyright (c) 1986 by University of Toronto.
40 * Written by Henry Spencer. Not derived from licensed software.
42 * Permission is granted to anyone to use this software for any
43 * purpose on any computer system, and to redistribute it freely,
44 * subject to the following restrictions:
46 * 1. The author is not responsible for the consequences of use of
47 * this software, no matter how awful, even if they arise
50 * 2. The origin of this software must not be misrepresented, either
51 * by explicit claim or by omission.
53 * 3. Altered versions must be plainly marked as such, and must not
54 * be misrepresented as being the original software.
56 **** Alterations to Henry's code are...
58 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
59 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007 by Larry Wall and others
61 **** You may distribute under the terms of either the GNU General Public
62 **** License or the Artistic License, as specified in the README file.
64 * Beware that some of this code is subtly aware of the way operator
65 * precedence is structured in regular expressions. Serious changes in
66 * regular-expression syntax might require a total rethink.
69 #define PERL_IN_REGEXEC_C
72 #ifdef PERL_IN_XSUB_RE
78 #define RF_tainted 1 /* tainted information used? */
79 #define RF_warned 2 /* warned about big count? */
81 #define RF_utf8 8 /* Pattern contains multibyte chars? */
83 #define UTF ((PL_reg_flags & RF_utf8) != 0)
85 #define RS_init 1 /* eval environment created */
86 #define RS_set 2 /* replsv value is set */
92 #define REGINCLASS(prog,p,c) (ANYOF_FLAGS(p) ? reginclass(prog,p,c,0,0) : ANYOF_BITMAP_TEST(p,*(c)))
98 #define CHR_SVLEN(sv) (do_utf8 ? sv_len_utf8(sv) : SvCUR(sv))
99 #define CHR_DIST(a,b) (PL_reg_match_utf8 ? utf8_distance(a,b) : a - b)
101 #define HOPc(pos,off) \
102 (char *)(PL_reg_match_utf8 \
103 ? reghop3((U8*)pos, off, (U8*)(off >= 0 ? PL_regeol : PL_bostr)) \
105 #define HOPBACKc(pos, off) \
106 (char*)(PL_reg_match_utf8\
107 ? reghopmaybe3((U8*)pos, -off, (U8*)PL_bostr) \
108 : (pos - off >= PL_bostr) \
112 #define HOP3(pos,off,lim) (PL_reg_match_utf8 ? reghop3((U8*)(pos), off, (U8*)(lim)) : (U8*)(pos + off))
113 #define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim))
115 #define LOAD_UTF8_CHARCLASS(class,str) STMT_START { \
116 if (!CAT2(PL_utf8_,class)) { bool ok; ENTER; save_re_context(); ok=CAT2(is_utf8_,class)((const U8*)str); assert(ok); LEAVE; } } STMT_END
117 #define LOAD_UTF8_CHARCLASS_ALNUM() LOAD_UTF8_CHARCLASS(alnum,"a")
118 #define LOAD_UTF8_CHARCLASS_DIGIT() LOAD_UTF8_CHARCLASS(digit,"0")
119 #define LOAD_UTF8_CHARCLASS_SPACE() LOAD_UTF8_CHARCLASS(space," ")
120 #define LOAD_UTF8_CHARCLASS_MARK() LOAD_UTF8_CHARCLASS(mark, "\xcd\x86")
122 /* TODO: Combine JUMPABLE and HAS_TEXT to cache OP(rn) */
124 /* for use after a quantifier and before an EXACT-like node -- japhy */
125 /* it would be nice to rework regcomp.sym to generate this stuff. sigh */
126 #define JUMPABLE(rn) ( \
128 (OP(rn) == CLOSE && (!cur_eval || cur_eval->u.eval.close_paren != ARG(rn))) || \
130 OP(rn) == SUSPEND || OP(rn) == IFMATCH || \
131 OP(rn) == PLUS || OP(rn) == MINMOD || \
132 OP(rn) == KEEPS || (PL_regkind[OP(rn)] == VERB) || \
133 (PL_regkind[OP(rn)] == CURLY && ARG1(rn) > 0) \
135 #define IS_EXACT(rn) (PL_regkind[OP(rn)] == EXACT)
137 #define HAS_TEXT(rn) ( IS_EXACT(rn) || PL_regkind[OP(rn)] == REF )
140 /* Currently these are only used when PL_regkind[OP(rn)] == EXACT so
141 we don't need this definition. */
142 #define IS_TEXT(rn) ( OP(rn)==EXACT || OP(rn)==REF || OP(rn)==NREF )
143 #define IS_TEXTF(rn) ( OP(rn)==EXACTF || OP(rn)==REFF || OP(rn)==NREFF )
144 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL || OP(rn)==REFFL || OP(rn)==NREFFL )
147 /* ... so we use this as its faster. */
148 #define IS_TEXT(rn) ( OP(rn)==EXACT )
149 #define IS_TEXTF(rn) ( OP(rn)==EXACTF )
150 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL )
155 Search for mandatory following text node; for lookahead, the text must
156 follow but for lookbehind (rn->flags != 0) we skip to the next step.
158 #define FIND_NEXT_IMPT(rn) STMT_START { \
159 while (JUMPABLE(rn)) { \
160 const OPCODE type = OP(rn); \
161 if (type == SUSPEND || PL_regkind[type] == CURLY) \
162 rn = NEXTOPER(NEXTOPER(rn)); \
163 else if (type == PLUS) \
165 else if (type == IFMATCH) \
166 rn = (rn->flags == 0) ? NEXTOPER(NEXTOPER(rn)) : rn + ARG(rn); \
167 else rn += NEXT_OFF(rn); \
172 static void restore_pos(pTHX_ void *arg);
175 S_regcppush(pTHX_ I32 parenfloor)
178 const int retval = PL_savestack_ix;
179 #define REGCP_PAREN_ELEMS 4
180 const int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS;
182 GET_RE_DEBUG_FLAGS_DECL;
184 if (paren_elems_to_push < 0)
185 Perl_croak(aTHX_ "panic: paren_elems_to_push < 0");
187 #define REGCP_OTHER_ELEMS 7
188 SSGROW(paren_elems_to_push + REGCP_OTHER_ELEMS);
190 for (p = PL_regsize; p > parenfloor; p--) {
191 /* REGCP_PARENS_ELEMS are pushed per pairs of parentheses. */
192 SSPUSHINT(PL_regoffs[p].end);
193 SSPUSHINT(PL_regoffs[p].start);
194 SSPUSHPTR(PL_reg_start_tmp[p]);
196 DEBUG_BUFFERS_r(PerlIO_printf(Perl_debug_log,
197 " saving \\%"UVuf" %"IVdf"(%"IVdf")..%"IVdf"\n",
198 (UV)p, (IV)PL_regoffs[p].start,
199 (IV)(PL_reg_start_tmp[p] - PL_bostr),
200 (IV)PL_regoffs[p].end
203 /* REGCP_OTHER_ELEMS are pushed in any case, parentheses or no. */
204 SSPUSHPTR(PL_regoffs);
205 SSPUSHINT(PL_regsize);
206 SSPUSHINT(*PL_reglastparen);
207 SSPUSHINT(*PL_reglastcloseparen);
208 SSPUSHPTR(PL_reginput);
209 #define REGCP_FRAME_ELEMS 2
210 /* REGCP_FRAME_ELEMS are part of the REGCP_OTHER_ELEMS and
211 * are needed for the regexp context stack bookkeeping. */
212 SSPUSHINT(paren_elems_to_push + REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS);
213 SSPUSHINT(SAVEt_REGCONTEXT); /* Magic cookie. */
218 /* These are needed since we do not localize EVAL nodes: */
219 #define REGCP_SET(cp) \
221 PerlIO_printf(Perl_debug_log, \
222 " Setting an EVAL scope, savestack=%"IVdf"\n", \
223 (IV)PL_savestack_ix)); \
226 #define REGCP_UNWIND(cp) \
228 if (cp != PL_savestack_ix) \
229 PerlIO_printf(Perl_debug_log, \
230 " Clearing an EVAL scope, savestack=%"IVdf"..%"IVdf"\n", \
231 (IV)(cp), (IV)PL_savestack_ix)); \
235 S_regcppop(pTHX_ const regexp *rex)
241 GET_RE_DEBUG_FLAGS_DECL;
243 /* Pop REGCP_OTHER_ELEMS before the parentheses loop starts. */
245 assert(i == SAVEt_REGCONTEXT); /* Check that the magic cookie is there. */
246 i = SSPOPINT; /* Parentheses elements to pop. */
247 input = (char *) SSPOPPTR;
248 *PL_reglastcloseparen = SSPOPINT;
249 *PL_reglastparen = SSPOPINT;
250 PL_regsize = SSPOPINT;
251 PL_regoffs=(regexp_paren_pair *) SSPOPPTR;
254 /* Now restore the parentheses context. */
255 for (i -= (REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS);
256 i > 0; i -= REGCP_PAREN_ELEMS) {
258 U32 paren = (U32)SSPOPINT;
259 PL_reg_start_tmp[paren] = (char *) SSPOPPTR;
260 PL_regoffs[paren].start = SSPOPINT;
262 if (paren <= *PL_reglastparen)
263 PL_regoffs[paren].end = tmps;
265 PerlIO_printf(Perl_debug_log,
266 " restoring \\%"UVuf" to %"IVdf"(%"IVdf")..%"IVdf"%s\n",
267 (UV)paren, (IV)PL_regoffs[paren].start,
268 (IV)(PL_reg_start_tmp[paren] - PL_bostr),
269 (IV)PL_regoffs[paren].end,
270 (paren > *PL_reglastparen ? "(no)" : ""));
274 if (*PL_reglastparen + 1 <= rex->nparens) {
275 PerlIO_printf(Perl_debug_log,
276 " restoring \\%"IVdf"..\\%"IVdf" to undef\n",
277 (IV)(*PL_reglastparen + 1), (IV)rex->nparens);
281 /* It would seem that the similar code in regtry()
282 * already takes care of this, and in fact it is in
283 * a better location to since this code can #if 0-ed out
284 * but the code in regtry() is needed or otherwise tests
285 * requiring null fields (pat.t#187 and split.t#{13,14}
286 * (as of patchlevel 7877) will fail. Then again,
287 * this code seems to be necessary or otherwise
288 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
289 * --jhi updated by dapm */
290 for (i = *PL_reglastparen + 1; i <= rex->nparens; i++) {
292 PL_regoffs[i].start = -1;
293 PL_regoffs[i].end = -1;
299 #define regcpblow(cp) LEAVE_SCOPE(cp) /* Ignores regcppush()ed data. */
302 * pregexec and friends
305 #ifndef PERL_IN_XSUB_RE
307 - pregexec - match a regexp against a string
310 Perl_pregexec(pTHX_ REGEXP * const prog, char* stringarg, register char *strend,
311 char *strbeg, I32 minend, SV *screamer, U32 nosave)
312 /* strend: pointer to null at end of string */
313 /* strbeg: real beginning of string */
314 /* minend: end of match must be >=minend after stringarg. */
315 /* nosave: For optimizations. */
318 regexec_flags(prog, stringarg, strend, strbeg, minend, screamer, NULL,
319 nosave ? 0 : REXEC_COPY_STR);
324 * Need to implement the following flags for reg_anch:
326 * USE_INTUIT_NOML - Useful to call re_intuit_start() first
328 * INTUIT_AUTORITATIVE_NOML - Can trust a positive answer
329 * INTUIT_AUTORITATIVE_ML
330 * INTUIT_ONCE_NOML - Intuit can match in one location only.
333 * Another flag for this function: SECOND_TIME (so that float substrs
334 * with giant delta may be not rechecked).
337 /* Assumptions: if ANCH_GPOS, then strpos is anchored. XXXX Check GPOS logic */
339 /* If SCREAM, then SvPVX_const(sv) should be compatible with strpos and strend.
340 Otherwise, only SvCUR(sv) is used to get strbeg. */
342 /* XXXX We assume that strpos is strbeg unless sv. */
344 /* XXXX Some places assume that there is a fixed substring.
345 An update may be needed if optimizer marks as "INTUITable"
346 RExen without fixed substrings. Similarly, it is assumed that
347 lengths of all the strings are no more than minlen, thus they
348 cannot come from lookahead.
349 (Or minlen should take into account lookahead.)
350 NOTE: Some of this comment is not correct. minlen does now take account
351 of lookahead/behind. Further research is required. -- demerphq
355 /* A failure to find a constant substring means that there is no need to make
356 an expensive call to REx engine, thus we celebrate a failure. Similarly,
357 finding a substring too deep into the string means that less calls to
358 regtry() should be needed.
360 REx compiler's optimizer found 4 possible hints:
361 a) Anchored substring;
363 c) Whether we are anchored (beginning-of-line or \G);
364 d) First node (of those at offset 0) which may distingush positions;
365 We use a)b)d) and multiline-part of c), and try to find a position in the
366 string which does not contradict any of them.
369 /* Most of decisions we do here should have been done at compile time.
370 The nodes of the REx which we used for the search should have been
371 deleted from the finite automaton. */
374 Perl_re_intuit_start(pTHX_ REGEXP * const prog, SV *sv, char *strpos,
375 char *strend, const U32 flags, re_scream_pos_data *data)
378 register I32 start_shift = 0;
379 /* Should be nonnegative! */
380 register I32 end_shift = 0;
385 const bool do_utf8 = (sv && SvUTF8(sv)) ? 1 : 0; /* if no sv we have to assume bytes */
387 register char *other_last = NULL; /* other substr checked before this */
388 char *check_at = NULL; /* check substr found at this pos */
389 const I32 multiline = prog->extflags & RXf_PMf_MULTILINE;
390 RXi_GET_DECL(prog,progi);
392 const char * const i_strpos = strpos;
395 GET_RE_DEBUG_FLAGS_DECL;
397 RX_MATCH_UTF8_set(prog,do_utf8);
399 if (prog->extflags & RXf_UTF8) {
400 PL_reg_flags |= RF_utf8;
403 debug_start_match(prog, do_utf8, strpos, strend,
404 sv ? "Guessing start of match in sv for"
405 : "Guessing start of match in string for");
408 /* CHR_DIST() would be more correct here but it makes things slow. */
409 if (prog->minlen > strend - strpos) {
410 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
411 "String too short... [re_intuit_start]\n"));
415 strbeg = (sv && SvPOK(sv)) ? strend - SvCUR(sv) : strpos;
418 if (!prog->check_utf8 && prog->check_substr)
419 to_utf8_substr(prog);
420 check = prog->check_utf8;
422 if (!prog->check_substr && prog->check_utf8)
423 to_byte_substr(prog);
424 check = prog->check_substr;
426 if (check == &PL_sv_undef) {
427 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
428 "Non-utf8 string cannot match utf8 check string\n"));
431 if (prog->extflags & RXf_ANCH) { /* Match at beg-of-str or after \n */
432 ml_anch = !( (prog->extflags & RXf_ANCH_SINGLE)
433 || ( (prog->extflags & RXf_ANCH_BOL)
434 && !multiline ) ); /* Check after \n? */
437 if ( !(prog->extflags & RXf_ANCH_GPOS) /* Checked by the caller */
438 && !(prog->intflags & PREGf_IMPLICIT) /* not a real BOL */
439 /* SvCUR is not set on references: SvRV and SvPVX_const overlap */
441 && (strpos != strbeg)) {
442 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not at start...\n"));
445 if (prog->check_offset_min == prog->check_offset_max &&
446 !(prog->extflags & RXf_CANY_SEEN)) {
447 /* Substring at constant offset from beg-of-str... */
450 s = HOP3c(strpos, prog->check_offset_min, strend);
453 slen = SvCUR(check); /* >= 1 */
455 if ( strend - s > slen || strend - s < slen - 1
456 || (strend - s == slen && strend[-1] != '\n')) {
457 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String too long...\n"));
460 /* Now should match s[0..slen-2] */
462 if (slen && (*SvPVX_const(check) != *s
464 && memNE(SvPVX_const(check), s, slen)))) {
466 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String not equal...\n"));
470 else if (*SvPVX_const(check) != *s
471 || ((slen = SvCUR(check)) > 1
472 && memNE(SvPVX_const(check), s, slen)))
475 goto success_at_start;
478 /* Match is anchored, but substr is not anchored wrt beg-of-str. */
480 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
481 end_shift = prog->check_end_shift;
484 const I32 end = prog->check_offset_max + CHR_SVLEN(check)
485 - (SvTAIL(check) != 0);
486 const I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end;
488 if (end_shift < eshift)
492 else { /* Can match at random position */
495 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
496 end_shift = prog->check_end_shift;
498 /* end shift should be non negative here */
501 #ifdef QDEBUGGING /* 7/99: reports of failure (with the older version) */
503 Perl_croak(aTHX_ "panic: end_shift: %"IVdf" pattern:\n%s\n ",
504 (IV)end_shift, prog->precomp);
508 /* Find a possible match in the region s..strend by looking for
509 the "check" substring in the region corrected by start/end_shift. */
512 I32 srch_start_shift = start_shift;
513 I32 srch_end_shift = end_shift;
514 if (srch_start_shift < 0 && strbeg - s > srch_start_shift) {
515 srch_end_shift -= ((strbeg - s) - srch_start_shift);
516 srch_start_shift = strbeg - s;
518 DEBUG_OPTIMISE_MORE_r({
519 PerlIO_printf(Perl_debug_log, "Check offset min: %"IVdf" Start shift: %"IVdf" End shift %"IVdf" Real End Shift: %"IVdf"\n",
520 (IV)prog->check_offset_min,
521 (IV)srch_start_shift,
523 (IV)prog->check_end_shift);
526 if (flags & REXEC_SCREAM) {
527 I32 p = -1; /* Internal iterator of scream. */
528 I32 * const pp = data ? data->scream_pos : &p;
530 if (PL_screamfirst[BmRARE(check)] >= 0
531 || ( BmRARE(check) == '\n'
532 && (BmPREVIOUS(check) == SvCUR(check) - 1)
534 s = screaminstr(sv, check,
535 srch_start_shift + (s - strbeg), srch_end_shift, pp, 0);
538 /* we may be pointing at the wrong string */
539 if (s && RX_MATCH_COPIED(prog))
540 s = strbeg + (s - SvPVX_const(sv));
542 *data->scream_olds = s;
547 if (prog->extflags & RXf_CANY_SEEN) {
548 start_point= (U8*)(s + srch_start_shift);
549 end_point= (U8*)(strend - srch_end_shift);
551 start_point= HOP3(s, srch_start_shift, srch_start_shift < 0 ? strbeg : strend);
552 end_point= HOP3(strend, -srch_end_shift, strbeg);
554 DEBUG_OPTIMISE_MORE_r({
555 PerlIO_printf(Perl_debug_log, "fbm_instr len=%d str=<%.*s>\n",
556 (int)(end_point - start_point),
557 (int)(end_point - start_point) > 20 ? 20 : (int)(end_point - start_point),
561 s = fbm_instr( start_point, end_point,
562 check, multiline ? FBMrf_MULTILINE : 0);
565 /* Update the count-of-usability, remove useless subpatterns,
569 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
570 SvPVX_const(check), RE_SV_DUMPLEN(check), 30);
571 PerlIO_printf(Perl_debug_log, "%s %s substr %s%s%s",
572 (s ? "Found" : "Did not find"),
573 (check == (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)
574 ? "anchored" : "floating"),
577 (s ? " at offset " : "...\n") );
582 /* Finish the diagnostic message */
583 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%ld...\n", (long)(s - i_strpos)) );
585 /* XXX dmq: first branch is for positive lookbehind...
586 Our check string is offset from the beginning of the pattern.
587 So we need to do any stclass tests offset forward from that
596 /* Got a candidate. Check MBOL anchoring, and the *other* substr.
597 Start with the other substr.
598 XXXX no SCREAM optimization yet - and a very coarse implementation
599 XXXX /ttx+/ results in anchored="ttx", floating="x". floating will
600 *always* match. Probably should be marked during compile...
601 Probably it is right to do no SCREAM here...
604 if (do_utf8 ? (prog->float_utf8 && prog->anchored_utf8)
605 : (prog->float_substr && prog->anchored_substr))
607 /* Take into account the "other" substring. */
608 /* XXXX May be hopelessly wrong for UTF... */
611 if (check == (do_utf8 ? prog->float_utf8 : prog->float_substr)) {
614 char * const last = HOP3c(s, -start_shift, strbeg);
616 char * const saved_s = s;
619 t = s - prog->check_offset_max;
620 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
622 || ((t = (char*)reghopmaybe3((U8*)s, -(prog->check_offset_max), (U8*)strpos))
627 t = HOP3c(t, prog->anchored_offset, strend);
628 if (t < other_last) /* These positions already checked */
630 last2 = last1 = HOP3c(strend, -prog->minlen, strbeg);
633 /* XXXX It is not documented what units *_offsets are in.
634 We assume bytes, but this is clearly wrong.
635 Meaning this code needs to be carefully reviewed for errors.
639 /* On end-of-str: see comment below. */
640 must = do_utf8 ? prog->anchored_utf8 : prog->anchored_substr;
641 if (must == &PL_sv_undef) {
643 DEBUG_r(must = prog->anchored_utf8); /* for debug */
648 HOP3(HOP3(last1, prog->anchored_offset, strend)
649 + SvCUR(must), -(SvTAIL(must)!=0), strbeg),
651 multiline ? FBMrf_MULTILINE : 0
654 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
655 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
656 PerlIO_printf(Perl_debug_log, "%s anchored substr %s%s",
657 (s ? "Found" : "Contradicts"),
658 quoted, RE_SV_TAIL(must));
663 if (last1 >= last2) {
664 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
665 ", giving up...\n"));
668 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
669 ", trying floating at offset %ld...\n",
670 (long)(HOP3c(saved_s, 1, strend) - i_strpos)));
671 other_last = HOP3c(last1, prog->anchored_offset+1, strend);
672 s = HOP3c(last, 1, strend);
676 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
677 (long)(s - i_strpos)));
678 t = HOP3c(s, -prog->anchored_offset, strbeg);
679 other_last = HOP3c(s, 1, strend);
687 else { /* Take into account the floating substring. */
689 char * const saved_s = s;
692 t = HOP3c(s, -start_shift, strbeg);
694 HOP3c(strend, -prog->minlen + prog->float_min_offset, strbeg);
695 if (CHR_DIST((U8*)last, (U8*)t) > prog->float_max_offset)
696 last = HOP3c(t, prog->float_max_offset, strend);
697 s = HOP3c(t, prog->float_min_offset, strend);
700 /* XXXX It is not documented what units *_offsets are in. Assume bytes. */
701 must = do_utf8 ? prog->float_utf8 : prog->float_substr;
702 /* fbm_instr() takes into account exact value of end-of-str
703 if the check is SvTAIL(ed). Since false positives are OK,
704 and end-of-str is not later than strend we are OK. */
705 if (must == &PL_sv_undef) {
707 DEBUG_r(must = prog->float_utf8); /* for debug message */
710 s = fbm_instr((unsigned char*)s,
711 (unsigned char*)last + SvCUR(must)
713 must, multiline ? FBMrf_MULTILINE : 0);
715 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
716 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
717 PerlIO_printf(Perl_debug_log, "%s floating substr %s%s",
718 (s ? "Found" : "Contradicts"),
719 quoted, RE_SV_TAIL(must));
723 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
724 ", giving up...\n"));
727 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
728 ", trying anchored starting at offset %ld...\n",
729 (long)(saved_s + 1 - i_strpos)));
731 s = HOP3c(t, 1, strend);
735 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
736 (long)(s - i_strpos)));
737 other_last = s; /* Fix this later. --Hugo */
747 t= (char*)HOP3( s, -prog->check_offset_max, (prog->check_offset_max<0) ? strend : strpos);
749 DEBUG_OPTIMISE_MORE_r(
750 PerlIO_printf(Perl_debug_log,
751 "Check offset min:%"IVdf" max:%"IVdf" S:%"IVdf" t:%"IVdf" D:%"IVdf" end:%"IVdf"\n",
752 (IV)prog->check_offset_min,
753 (IV)prog->check_offset_max,
761 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
763 || ((t = (char*)reghopmaybe3((U8*)s, -prog->check_offset_max, (U8*) ((prog->check_offset_max<0) ? strend : strpos)))
766 /* Fixed substring is found far enough so that the match
767 cannot start at strpos. */
769 if (ml_anch && t[-1] != '\n') {
770 /* Eventually fbm_*() should handle this, but often
771 anchored_offset is not 0, so this check will not be wasted. */
772 /* XXXX In the code below we prefer to look for "^" even in
773 presence of anchored substrings. And we search even
774 beyond the found float position. These pessimizations
775 are historical artefacts only. */
777 while (t < strend - prog->minlen) {
779 if (t < check_at - prog->check_offset_min) {
780 if (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) {
781 /* Since we moved from the found position,
782 we definitely contradict the found anchored
783 substr. Due to the above check we do not
784 contradict "check" substr.
785 Thus we can arrive here only if check substr
786 is float. Redo checking for "other"=="fixed".
789 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld, rescanning for anchored from offset %ld...\n",
790 PL_colors[0], PL_colors[1], (long)(strpos - i_strpos), (long)(strpos - i_strpos + prog->anchored_offset)));
791 goto do_other_anchored;
793 /* We don't contradict the found floating substring. */
794 /* XXXX Why not check for STCLASS? */
796 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld...\n",
797 PL_colors[0], PL_colors[1], (long)(s - i_strpos)));
800 /* Position contradicts check-string */
801 /* XXXX probably better to look for check-string
802 than for "\n", so one should lower the limit for t? */
803 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m, restarting lookup for check-string at offset %ld...\n",
804 PL_colors[0], PL_colors[1], (long)(t + 1 - i_strpos)));
805 other_last = strpos = s = t + 1;
810 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Did not find /%s^%s/m...\n",
811 PL_colors[0], PL_colors[1]));
815 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Starting position does not contradict /%s^%s/m...\n",
816 PL_colors[0], PL_colors[1]));
820 ++BmUSEFUL(do_utf8 ? prog->check_utf8 : prog->check_substr); /* hooray/5 */
823 /* The found string does not prohibit matching at strpos,
824 - no optimization of calling REx engine can be performed,
825 unless it was an MBOL and we are not after MBOL,
826 or a future STCLASS check will fail this. */
828 /* Even in this situation we may use MBOL flag if strpos is offset
829 wrt the start of the string. */
830 if (ml_anch && sv && !SvROK(sv) /* See prev comment on SvROK */
831 && (strpos != strbeg) && strpos[-1] != '\n'
832 /* May be due to an implicit anchor of m{.*foo} */
833 && !(prog->intflags & PREGf_IMPLICIT))
838 DEBUG_EXECUTE_r( if (ml_anch)
839 PerlIO_printf(Perl_debug_log, "Position at offset %ld does not contradict /%s^%s/m...\n",
840 (long)(strpos - i_strpos), PL_colors[0], PL_colors[1]);
843 if (!(prog->intflags & PREGf_NAUGHTY) /* XXXX If strpos moved? */
845 prog->check_utf8 /* Could be deleted already */
846 && --BmUSEFUL(prog->check_utf8) < 0
847 && (prog->check_utf8 == prog->float_utf8)
849 prog->check_substr /* Could be deleted already */
850 && --BmUSEFUL(prog->check_substr) < 0
851 && (prog->check_substr == prog->float_substr)
854 /* If flags & SOMETHING - do not do it many times on the same match */
855 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "... Disabling check substring...\n"));
856 SvREFCNT_dec(do_utf8 ? prog->check_utf8 : prog->check_substr);
857 if (do_utf8 ? prog->check_substr : prog->check_utf8)
858 SvREFCNT_dec(do_utf8 ? prog->check_substr : prog->check_utf8);
859 prog->check_substr = prog->check_utf8 = NULL; /* disable */
860 prog->float_substr = prog->float_utf8 = NULL; /* clear */
861 check = NULL; /* abort */
863 /* XXXX This is a remnant of the old implementation. It
864 looks wasteful, since now INTUIT can use many
866 prog->extflags &= ~RXf_USE_INTUIT;
873 /* XXXX BmUSEFUL already changed, maybe multiple change is meaningful... */
874 /* trie stclasses are too expensive to use here, we are better off to
875 leave it to regmatch itself */
876 if (progi->regstclass && PL_regkind[OP(progi->regstclass)]!=TRIE) {
877 /* minlen == 0 is possible if regstclass is \b or \B,
878 and the fixed substr is ''$.
879 Since minlen is already taken into account, s+1 is before strend;
880 accidentally, minlen >= 1 guaranties no false positives at s + 1
881 even for \b or \B. But (minlen? 1 : 0) below assumes that
882 regstclass does not come from lookahead... */
883 /* If regstclass takes bytelength more than 1: If charlength==1, OK.
884 This leaves EXACTF only, which is dealt with in find_byclass(). */
885 const U8* const str = (U8*)STRING(progi->regstclass);
886 const int cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
887 ? CHR_DIST(str+STR_LEN(progi->regstclass), str)
890 if (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
891 endpos= HOP3c(s, (prog->minlen ? cl_l : 0), strend);
892 else if (prog->float_substr || prog->float_utf8)
893 endpos= HOP3c(HOP3c(check_at, -start_shift, strbeg), cl_l, strend);
897 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "start_shift: %"IVdf" check_at: %"IVdf" s: %"IVdf" endpos: %"IVdf"\n",
898 (IV)start_shift, (IV)(check_at - strbeg), (IV)(s - strbeg), (IV)(endpos - strbeg)));
901 s = find_byclass(prog, progi->regstclass, s, endpos, NULL);
904 const char *what = NULL;
906 if (endpos == strend) {
907 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
908 "Could not match STCLASS...\n") );
911 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
912 "This position contradicts STCLASS...\n") );
913 if ((prog->extflags & RXf_ANCH) && !ml_anch)
915 /* Contradict one of substrings */
916 if (prog->anchored_substr || prog->anchored_utf8) {
917 if ((do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) == check) {
918 DEBUG_EXECUTE_r( what = "anchored" );
920 s = HOP3c(t, 1, strend);
921 if (s + start_shift + end_shift > strend) {
922 /* XXXX Should be taken into account earlier? */
923 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
924 "Could not match STCLASS...\n") );
929 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
930 "Looking for %s substr starting at offset %ld...\n",
931 what, (long)(s + start_shift - i_strpos)) );
934 /* Have both, check_string is floating */
935 if (t + start_shift >= check_at) /* Contradicts floating=check */
936 goto retry_floating_check;
937 /* Recheck anchored substring, but not floating... */
941 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
942 "Looking for anchored substr starting at offset %ld...\n",
943 (long)(other_last - i_strpos)) );
944 goto do_other_anchored;
946 /* Another way we could have checked stclass at the
947 current position only: */
952 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
953 "Looking for /%s^%s/m starting at offset %ld...\n",
954 PL_colors[0], PL_colors[1], (long)(t - i_strpos)) );
957 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr)) /* Could have been deleted */
959 /* Check is floating subtring. */
960 retry_floating_check:
961 t = check_at - start_shift;
962 DEBUG_EXECUTE_r( what = "floating" );
963 goto hop_and_restart;
966 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
967 "By STCLASS: moving %ld --> %ld\n",
968 (long)(t - i_strpos), (long)(s - i_strpos))
972 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
973 "Does not contradict STCLASS...\n");
978 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s%s:%s match at offset %ld\n",
979 PL_colors[4], (check ? "Guessed" : "Giving up"),
980 PL_colors[5], (long)(s - i_strpos)) );
983 fail_finish: /* Substring not found */
984 if (prog->check_substr || prog->check_utf8) /* could be removed already */
985 BmUSEFUL(do_utf8 ? prog->check_utf8 : prog->check_substr) += 5; /* hooray */
987 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch rejected by optimizer%s\n",
988 PL_colors[4], PL_colors[5]));
992 #define DECL_TRIE_TYPE(scan) \
993 const enum { trie_plain, trie_utf8, trie_utf8_fold, trie_latin_utf8_fold } \
994 trie_type = (scan->flags != EXACT) \
995 ? (do_utf8 ? trie_utf8_fold : (UTF ? trie_latin_utf8_fold : trie_plain)) \
996 : (do_utf8 ? trie_utf8 : trie_plain)
998 #define REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc, uscan, len, \
999 uvc, charid, foldlen, foldbuf, uniflags) STMT_START { \
1000 switch (trie_type) { \
1001 case trie_utf8_fold: \
1002 if ( foldlen>0 ) { \
1003 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1008 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1009 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1010 foldlen -= UNISKIP( uvc ); \
1011 uscan = foldbuf + UNISKIP( uvc ); \
1014 case trie_latin_utf8_fold: \
1015 if ( foldlen>0 ) { \
1016 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1022 uvc = to_uni_fold( *(U8*)uc, foldbuf, &foldlen ); \
1023 foldlen -= UNISKIP( uvc ); \
1024 uscan = foldbuf + UNISKIP( uvc ); \
1028 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1036 charid = trie->charmap[ uvc ]; \
1040 if (widecharmap) { \
1041 SV** const svpp = hv_fetch(widecharmap, \
1042 (char*)&uvc, sizeof(UV), 0); \
1044 charid = (U16)SvIV(*svpp); \
1049 #define REXEC_FBC_EXACTISH_CHECK(CoNd) \
1051 char *my_strend= (char *)strend; \
1054 !ibcmp_utf8(s, &my_strend, 0, do_utf8, \
1055 m, NULL, ln, (bool)UTF)) \
1056 && (!reginfo || regtry(reginfo, &s)) ) \
1059 U8 foldbuf[UTF8_MAXBYTES_CASE+1]; \
1060 uvchr_to_utf8(tmpbuf, c); \
1061 f = to_utf8_fold(tmpbuf, foldbuf, &foldlen); \
1063 && (f == c1 || f == c2) \
1065 !ibcmp_utf8(s, &my_strend, 0, do_utf8,\
1066 m, NULL, ln, (bool)UTF)) \
1067 && (!reginfo || regtry(reginfo, &s)) ) \
1073 #define REXEC_FBC_EXACTISH_SCAN(CoNd) \
1077 && (ln == 1 || !(OP(c) == EXACTF \
1079 : ibcmp_locale(s, m, ln))) \
1080 && (!reginfo || regtry(reginfo, &s)) ) \
1086 #define REXEC_FBC_UTF8_SCAN(CoDe) \
1088 while (s + (uskip = UTF8SKIP(s)) <= strend) { \
1094 #define REXEC_FBC_SCAN(CoDe) \
1096 while (s < strend) { \
1102 #define REXEC_FBC_UTF8_CLASS_SCAN(CoNd) \
1103 REXEC_FBC_UTF8_SCAN( \
1105 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1114 #define REXEC_FBC_CLASS_SCAN(CoNd) \
1117 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1126 #define REXEC_FBC_TRYIT \
1127 if ((!reginfo || regtry(reginfo, &s))) \
1130 #define REXEC_FBC_CSCAN(CoNdUtF8,CoNd) \
1132 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1135 REXEC_FBC_CLASS_SCAN(CoNd); \
1139 #define REXEC_FBC_CSCAN_PRELOAD(UtFpReLoAd,CoNdUtF8,CoNd) \
1142 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1145 REXEC_FBC_CLASS_SCAN(CoNd); \
1149 #define REXEC_FBC_CSCAN_TAINT(CoNdUtF8,CoNd) \
1150 PL_reg_flags |= RF_tainted; \
1152 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1155 REXEC_FBC_CLASS_SCAN(CoNd); \
1159 #define DUMP_EXEC_POS(li,s,doutf8) \
1160 dump_exec_pos(li,s,(PL_regeol),(PL_bostr),(PL_reg_starttry),doutf8)
1162 /* We know what class REx starts with. Try to find this position... */
1163 /* if reginfo is NULL, its a dryrun */
1164 /* annoyingly all the vars in this routine have different names from their counterparts
1165 in regmatch. /grrr */
1168 S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s,
1169 const char *strend, regmatch_info *reginfo)
1172 const I32 doevery = (prog->intflags & PREGf_SKIP) == 0;
1176 register STRLEN uskip;
1180 register I32 tmp = 1; /* Scratch variable? */
1181 register const bool do_utf8 = PL_reg_match_utf8;
1182 RXi_GET_DECL(prog,progi);
1184 /* We know what class it must start with. */
1188 REXEC_FBC_UTF8_CLASS_SCAN((ANYOF_FLAGS(c) & ANYOF_UNICODE) ||
1189 !UTF8_IS_INVARIANT((U8)s[0]) ?
1190 reginclass(prog, c, (U8*)s, 0, do_utf8) :
1191 REGINCLASS(prog, c, (U8*)s));
1194 while (s < strend) {
1197 if (REGINCLASS(prog, c, (U8*)s) ||
1198 (ANYOF_FOLD_SHARP_S(c, s, strend) &&
1199 /* The assignment of 2 is intentional:
1200 * for the folded sharp s, the skip is 2. */
1201 (skip = SHARP_S_SKIP))) {
1202 if (tmp && (!reginfo || regtry(reginfo, &s)))
1215 if (tmp && (!reginfo || regtry(reginfo, &s)))
1223 ln = STR_LEN(c); /* length to match in octets/bytes */
1224 lnc = (I32) ln; /* length to match in characters */
1226 STRLEN ulen1, ulen2;
1228 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
1229 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
1230 /* used by commented-out code below */
1231 /*const U32 uniflags = UTF8_ALLOW_DEFAULT;*/
1233 /* XXX: Since the node will be case folded at compile
1234 time this logic is a little odd, although im not
1235 sure that its actually wrong. --dmq */
1237 c1 = to_utf8_lower((U8*)m, tmpbuf1, &ulen1);
1238 c2 = to_utf8_upper((U8*)m, tmpbuf2, &ulen2);
1240 /* XXX: This is kinda strange. to_utf8_XYZ returns the
1241 codepoint of the first character in the converted
1242 form, yet originally we did the extra step.
1243 No tests fail by commenting this code out however
1244 so Ive left it out. -- dmq.
1246 c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXBYTES_CASE,
1248 c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXBYTES_CASE,
1253 while (sm < ((U8 *) m + ln)) {
1268 c2 = PL_fold_locale[c1];
1270 e = HOP3c(strend, -((I32)lnc), s);
1272 if (!reginfo && e < s)
1273 e = s; /* Due to minlen logic of intuit() */
1275 /* The idea in the EXACTF* cases is to first find the
1276 * first character of the EXACTF* node and then, if
1277 * necessary, case-insensitively compare the full
1278 * text of the node. The c1 and c2 are the first
1279 * characters (though in Unicode it gets a bit
1280 * more complicated because there are more cases
1281 * than just upper and lower: one needs to use
1282 * the so-called folding case for case-insensitive
1283 * matching (called "loose matching" in Unicode).
1284 * ibcmp_utf8() will do just that. */
1286 if (do_utf8 || UTF) {
1288 U8 tmpbuf [UTF8_MAXBYTES+1];
1291 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1293 /* Upper and lower of 1st char are equal -
1294 * probably not a "letter". */
1297 c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
1302 REXEC_FBC_EXACTISH_CHECK(c == c1);
1308 c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
1314 /* Handle some of the three Greek sigmas cases.
1315 * Note that not all the possible combinations
1316 * are handled here: some of them are handled
1317 * by the standard folding rules, and some of
1318 * them (the character class or ANYOF cases)
1319 * are handled during compiletime in
1320 * regexec.c:S_regclass(). */
1321 if (c == (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA ||
1322 c == (UV)UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA)
1323 c = (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA;
1325 REXEC_FBC_EXACTISH_CHECK(c == c1 || c == c2);
1330 /* Neither pattern nor string are UTF8 */
1332 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1);
1334 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1 || *(U8*)s == c2);
1338 PL_reg_flags |= RF_tainted;
1345 U8 * const r = reghop3((U8*)s, -1, (U8*)PL_bostr);
1346 tmp = utf8n_to_uvchr(r, UTF8SKIP(r), 0, UTF8_ALLOW_DEFAULT);
1348 tmp = ((OP(c) == BOUND ?
1349 isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
1350 LOAD_UTF8_CHARCLASS_ALNUM();
1351 REXEC_FBC_UTF8_SCAN(
1352 if (tmp == !(OP(c) == BOUND ?
1353 (bool)swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
1354 isALNUM_LC_utf8((U8*)s)))
1362 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n';
1363 tmp = ((OP(c) == BOUND ? isALNUM(tmp) : isALNUM_LC(tmp)) != 0);
1366 !(OP(c) == BOUND ? isALNUM(*s) : isALNUM_LC(*s))) {
1372 if ((!prog->minlen && tmp) && (!reginfo || regtry(reginfo, &s)))
1376 PL_reg_flags |= RF_tainted;
1383 U8 * const r = reghop3((U8*)s, -1, (U8*)PL_bostr);
1384 tmp = utf8n_to_uvchr(r, UTF8SKIP(r), 0, UTF8_ALLOW_DEFAULT);
1386 tmp = ((OP(c) == NBOUND ?
1387 isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
1388 LOAD_UTF8_CHARCLASS_ALNUM();
1389 REXEC_FBC_UTF8_SCAN(
1390 if (tmp == !(OP(c) == NBOUND ?
1391 (bool)swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
1392 isALNUM_LC_utf8((U8*)s)))
1394 else REXEC_FBC_TRYIT;
1398 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n';
1399 tmp = ((OP(c) == NBOUND ?
1400 isALNUM(tmp) : isALNUM_LC(tmp)) != 0);
1403 !(OP(c) == NBOUND ? isALNUM(*s) : isALNUM_LC(*s)))
1405 else REXEC_FBC_TRYIT;
1408 if ((!prog->minlen && !tmp) && (!reginfo || regtry(reginfo, &s)))
1412 REXEC_FBC_CSCAN_PRELOAD(
1413 LOAD_UTF8_CHARCLASS_ALNUM(),
1414 swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8),
1418 REXEC_FBC_CSCAN_TAINT(
1419 isALNUM_LC_utf8((U8*)s),
1423 REXEC_FBC_CSCAN_PRELOAD(
1424 LOAD_UTF8_CHARCLASS_ALNUM(),
1425 !swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8),
1429 REXEC_FBC_CSCAN_TAINT(
1430 !isALNUM_LC_utf8((U8*)s),
1434 REXEC_FBC_CSCAN_PRELOAD(
1435 LOAD_UTF8_CHARCLASS_SPACE(),
1436 *s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8),
1440 REXEC_FBC_CSCAN_TAINT(
1441 *s == ' ' || isSPACE_LC_utf8((U8*)s),
1445 REXEC_FBC_CSCAN_PRELOAD(
1446 LOAD_UTF8_CHARCLASS_SPACE(),
1447 !(*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8)),
1451 REXEC_FBC_CSCAN_TAINT(
1452 !(*s == ' ' || isSPACE_LC_utf8((U8*)s)),
1456 REXEC_FBC_CSCAN_PRELOAD(
1457 LOAD_UTF8_CHARCLASS_DIGIT(),
1458 swash_fetch(PL_utf8_digit,(U8*)s, do_utf8),
1462 REXEC_FBC_CSCAN_TAINT(
1463 isDIGIT_LC_utf8((U8*)s),
1467 REXEC_FBC_CSCAN_PRELOAD(
1468 LOAD_UTF8_CHARCLASS_DIGIT(),
1469 !swash_fetch(PL_utf8_digit,(U8*)s, do_utf8),
1473 REXEC_FBC_CSCAN_TAINT(
1474 !isDIGIT_LC_utf8((U8*)s),
1480 is_LNBREAK_latin1(s)
1490 !is_VERTWS_latin1(s)
1495 is_HORIZWS_latin1(s)
1499 !is_HORIZWS_utf8(s),
1500 !is_HORIZWS_latin1(s)
1506 /* what trie are we using right now */
1508 = (reg_ac_data*)progi->data->data[ ARG( c ) ];
1510 = (reg_trie_data*)progi->data->data[ aho->trie ];
1511 HV *widecharmap = (HV*) progi->data->data[ aho->trie + 1 ];
1513 const char *last_start = strend - trie->minlen;
1515 const char *real_start = s;
1517 STRLEN maxlen = trie->maxlen;
1519 U8 **points; /* map of where we were in the input string
1520 when reading a given char. For ASCII this
1521 is unnecessary overhead as the relationship
1522 is always 1:1, but for Unicode, especially
1523 case folded Unicode this is not true. */
1524 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1528 GET_RE_DEBUG_FLAGS_DECL;
1530 /* We can't just allocate points here. We need to wrap it in
1531 * an SV so it gets freed properly if there is a croak while
1532 * running the match */
1535 sv_points=newSV(maxlen * sizeof(U8 *));
1536 SvCUR_set(sv_points,
1537 maxlen * sizeof(U8 *));
1538 SvPOK_on(sv_points);
1539 sv_2mortal(sv_points);
1540 points=(U8**)SvPV_nolen(sv_points );
1541 if ( trie_type != trie_utf8_fold
1542 && (trie->bitmap || OP(c)==AHOCORASICKC) )
1545 bitmap=(U8*)trie->bitmap;
1547 bitmap=(U8*)ANYOF_BITMAP(c);
1549 /* this is the Aho-Corasick algorithm modified a touch
1550 to include special handling for long "unknown char"
1551 sequences. The basic idea being that we use AC as long
1552 as we are dealing with a possible matching char, when
1553 we encounter an unknown char (and we have not encountered
1554 an accepting state) we scan forward until we find a legal
1556 AC matching is basically that of trie matching, except
1557 that when we encounter a failing transition, we fall back
1558 to the current states "fail state", and try the current char
1559 again, a process we repeat until we reach the root state,
1560 state 1, or a legal transition. If we fail on the root state
1561 then we can either terminate if we have reached an accepting
1562 state previously, or restart the entire process from the beginning
1566 while (s <= last_start) {
1567 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1575 U8 *uscan = (U8*)NULL;
1576 U8 *leftmost = NULL;
1578 U32 accepted_word= 0;
1582 while ( state && uc <= (U8*)strend ) {
1584 U32 word = aho->states[ state ].wordnum;
1588 DEBUG_TRIE_EXECUTE_r(
1589 if ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1590 dump_exec_pos( (char *)uc, c, strend, real_start,
1591 (char *)uc, do_utf8 );
1592 PerlIO_printf( Perl_debug_log,
1593 " Scanning for legal start char...\n");
1596 while ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1601 if (uc >(U8*)last_start) break;
1605 U8 *lpos= points[ (pointpos - trie->wordlen[word-1] ) % maxlen ];
1606 if (!leftmost || lpos < leftmost) {
1607 DEBUG_r(accepted_word=word);
1613 points[pointpos++ % maxlen]= uc;
1614 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
1615 uscan, len, uvc, charid, foldlen,
1617 DEBUG_TRIE_EXECUTE_r({
1618 dump_exec_pos( (char *)uc, c, strend, real_start,
1620 PerlIO_printf(Perl_debug_log,
1621 " Charid:%3u CP:%4"UVxf" ",
1627 word = aho->states[ state ].wordnum;
1629 base = aho->states[ state ].trans.base;
1631 DEBUG_TRIE_EXECUTE_r({
1633 dump_exec_pos( (char *)uc, c, strend, real_start,
1635 PerlIO_printf( Perl_debug_log,
1636 "%sState: %4"UVxf", word=%"UVxf,
1637 failed ? " Fail transition to " : "",
1638 (UV)state, (UV)word);
1643 (base + charid > trie->uniquecharcount )
1644 && (base + charid - 1 - trie->uniquecharcount
1646 && trie->trans[base + charid - 1 -
1647 trie->uniquecharcount].check == state
1648 && (tmp=trie->trans[base + charid - 1 -
1649 trie->uniquecharcount ].next))
1651 DEBUG_TRIE_EXECUTE_r(
1652 PerlIO_printf( Perl_debug_log," - legal\n"));
1657 DEBUG_TRIE_EXECUTE_r(
1658 PerlIO_printf( Perl_debug_log," - fail\n"));
1660 state = aho->fail[state];
1664 /* we must be accepting here */
1665 DEBUG_TRIE_EXECUTE_r(
1666 PerlIO_printf( Perl_debug_log," - accepting\n"));
1675 if (!state) state = 1;
1678 if ( aho->states[ state ].wordnum ) {
1679 U8 *lpos = points[ (pointpos - trie->wordlen[aho->states[ state ].wordnum-1]) % maxlen ];
1680 if (!leftmost || lpos < leftmost) {
1681 DEBUG_r(accepted_word=aho->states[ state ].wordnum);
1686 s = (char*)leftmost;
1687 DEBUG_TRIE_EXECUTE_r({
1689 Perl_debug_log,"Matches word #%"UVxf" at position %"IVdf". Trying full pattern...\n",
1690 (UV)accepted_word, (IV)(s - real_start)
1693 if (!reginfo || regtry(reginfo, &s)) {
1699 DEBUG_TRIE_EXECUTE_r({
1700 PerlIO_printf( Perl_debug_log,"Pattern failed. Looking for new start point...\n");
1703 DEBUG_TRIE_EXECUTE_r(
1704 PerlIO_printf( Perl_debug_log,"No match.\n"));
1713 Perl_croak(aTHX_ "panic: unknown regstclass %d", (int)OP(c));
1722 S_swap_match_buff (pTHX_ regexp *prog) {
1723 regexp_paren_pair *t;
1726 /* We have to be careful. If the previous successful match
1727 was from this regex we don't want a subsequent paritally
1728 successful match to clobber the old results.
1729 So when we detect this possibility we add a swap buffer
1730 to the re, and switch the buffer each match. If we fail
1731 we switch it back, otherwise we leave it swapped.
1733 Newxz(prog->swap, (prog->nparens + 1), regexp_paren_pair);
1736 prog->swap = prog->offs;
1742 - regexec_flags - match a regexp against a string
1745 Perl_regexec_flags(pTHX_ REGEXP * const prog, char *stringarg, register char *strend,
1746 char *strbeg, I32 minend, SV *sv, void *data, U32 flags)
1747 /* strend: pointer to null at end of string */
1748 /* strbeg: real beginning of string */
1749 /* minend: end of match must be >=minend after stringarg. */
1750 /* data: May be used for some additional optimizations.
1751 Currently its only used, with a U32 cast, for transmitting
1752 the ganch offset when doing a /g match. This will change */
1753 /* nosave: For optimizations. */
1756 /*register*/ char *s;
1757 register regnode *c;
1758 /*register*/ char *startpos = stringarg;
1759 I32 minlen; /* must match at least this many chars */
1760 I32 dontbother = 0; /* how many characters not to try at end */
1761 I32 end_shift = 0; /* Same for the end. */ /* CC */
1762 I32 scream_pos = -1; /* Internal iterator of scream. */
1763 char *scream_olds = NULL;
1764 const bool do_utf8 = (bool)DO_UTF8(sv);
1766 RXi_GET_DECL(prog,progi);
1767 regmatch_info reginfo; /* create some info to pass to regtry etc */
1768 bool swap_on_fail = 0;
1770 GET_RE_DEBUG_FLAGS_DECL;
1772 PERL_UNUSED_ARG(data);
1774 /* Be paranoid... */
1775 if (prog == NULL || startpos == NULL) {
1776 Perl_croak(aTHX_ "NULL regexp parameter");
1780 multiline = prog->extflags & RXf_PMf_MULTILINE;
1781 reginfo.prog = prog;
1783 RX_MATCH_UTF8_set(prog, do_utf8);
1785 debug_start_match(prog, do_utf8, startpos, strend,
1789 minlen = prog->minlen;
1791 if (strend - startpos < (minlen+(prog->check_offset_min<0?prog->check_offset_min:0))) {
1792 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
1793 "String too short [regexec_flags]...\n"));
1798 /* Check validity of program. */
1799 if (UCHARAT(progi->program) != REG_MAGIC) {
1800 Perl_croak(aTHX_ "corrupted regexp program");
1804 PL_reg_eval_set = 0;
1807 if (prog->extflags & RXf_UTF8)
1808 PL_reg_flags |= RF_utf8;
1810 /* Mark beginning of line for ^ and lookbehind. */
1811 reginfo.bol = startpos; /* XXX not used ??? */
1815 /* Mark end of line for $ (and such) */
1818 /* see how far we have to get to not match where we matched before */
1819 reginfo.till = startpos+minend;
1821 /* If there is a "must appear" string, look for it. */
1824 if (prog->extflags & RXf_GPOS_SEEN) { /* Need to set reginfo->ganch */
1827 if (flags & REXEC_IGNOREPOS) /* Means: check only at start */
1828 reginfo.ganch = startpos + prog->gofs;
1829 else if (sv && SvTYPE(sv) >= SVt_PVMG
1831 && (mg = mg_find(sv, PERL_MAGIC_regex_global))
1832 && mg->mg_len >= 0) {
1833 reginfo.ganch = strbeg + mg->mg_len; /* Defined pos() */
1834 if (prog->extflags & RXf_ANCH_GPOS) {
1835 if (s > reginfo.ganch)
1837 s = reginfo.ganch - prog->gofs;
1841 reginfo.ganch = strbeg + PTR2UV(data);
1842 } else /* pos() not defined */
1843 reginfo.ganch = strbeg;
1845 if (PL_curpm && (PM_GETRE(PL_curpm) == prog)) {
1847 swap_match_buff(prog); /* do we need a save destructor here for
1850 if (!(flags & REXEC_CHECKED) && (prog->check_substr != NULL || prog->check_utf8 != NULL)) {
1851 re_scream_pos_data d;
1853 d.scream_olds = &scream_olds;
1854 d.scream_pos = &scream_pos;
1855 s = re_intuit_start(prog, sv, s, strend, flags, &d);
1857 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not present...\n"));
1858 goto phooey; /* not present */
1864 /* Simplest case: anchored match need be tried only once. */
1865 /* [unless only anchor is BOL and multiline is set] */
1866 if (prog->extflags & (RXf_ANCH & ~RXf_ANCH_GPOS)) {
1867 if (s == startpos && regtry(®info, &startpos))
1869 else if (multiline || (prog->intflags & PREGf_IMPLICIT)
1870 || (prog->extflags & RXf_ANCH_MBOL)) /* XXXX SBOL? */
1875 dontbother = minlen - 1;
1876 end = HOP3c(strend, -dontbother, strbeg) - 1;
1877 /* for multiline we only have to try after newlines */
1878 if (prog->check_substr || prog->check_utf8) {
1882 if (regtry(®info, &s))
1887 if (prog->extflags & RXf_USE_INTUIT) {
1888 s = re_intuit_start(prog, sv, s + 1, strend, flags, NULL);
1899 if (*s++ == '\n') { /* don't need PL_utf8skip here */
1900 if (regtry(®info, &s))
1907 } else if (RXf_GPOS_CHECK == (prog->extflags & RXf_GPOS_CHECK))
1909 /* the warning about reginfo.ganch being used without intialization
1910 is bogus -- we set it above, when prog->extflags & RXf_GPOS_SEEN
1911 and we only enter this block when the same bit is set. */
1912 char *tmp_s = reginfo.ganch - prog->gofs;
1913 if (regtry(®info, &tmp_s))
1918 /* Messy cases: unanchored match. */
1919 if ((prog->anchored_substr || prog->anchored_utf8) && prog->intflags & PREGf_SKIP) {
1920 /* we have /x+whatever/ */
1921 /* it must be a one character string (XXXX Except UTF?) */
1926 if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr))
1927 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1928 ch = SvPVX_const(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0];
1933 DEBUG_EXECUTE_r( did_match = 1 );
1934 if (regtry(®info, &s)) goto got_it;
1936 while (s < strend && *s == ch)
1944 DEBUG_EXECUTE_r( did_match = 1 );
1945 if (regtry(®info, &s)) goto got_it;
1947 while (s < strend && *s == ch)
1952 DEBUG_EXECUTE_r(if (!did_match)
1953 PerlIO_printf(Perl_debug_log,
1954 "Did not find anchored character...\n")
1957 else if (prog->anchored_substr != NULL
1958 || prog->anchored_utf8 != NULL
1959 || ((prog->float_substr != NULL || prog->float_utf8 != NULL)
1960 && prog->float_max_offset < strend - s)) {
1965 char *last1; /* Last position checked before */
1969 if (prog->anchored_substr || prog->anchored_utf8) {
1970 if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr))
1971 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1972 must = do_utf8 ? prog->anchored_utf8 : prog->anchored_substr;
1973 back_max = back_min = prog->anchored_offset;
1975 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr))
1976 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1977 must = do_utf8 ? prog->float_utf8 : prog->float_substr;
1978 back_max = prog->float_max_offset;
1979 back_min = prog->float_min_offset;
1983 if (must == &PL_sv_undef)
1984 /* could not downgrade utf8 check substring, so must fail */
1990 last = HOP3c(strend, /* Cannot start after this */
1991 -(I32)(CHR_SVLEN(must)
1992 - (SvTAIL(must) != 0) + back_min), strbeg);
1995 last1 = HOPc(s, -1);
1997 last1 = s - 1; /* bogus */
1999 /* XXXX check_substr already used to find "s", can optimize if
2000 check_substr==must. */
2002 dontbother = end_shift;
2003 strend = HOPc(strend, -dontbother);
2004 while ( (s <= last) &&
2005 ((flags & REXEC_SCREAM)
2006 ? (s = screaminstr(sv, must, HOP3c(s, back_min, (back_min<0 ? strbeg : strend)) - strbeg,
2007 end_shift, &scream_pos, 0))
2008 : (s = fbm_instr((unsigned char*)HOP3(s, back_min, (back_min<0 ? strbeg : strend)),
2009 (unsigned char*)strend, must,
2010 multiline ? FBMrf_MULTILINE : 0))) ) {
2011 /* we may be pointing at the wrong string */
2012 if ((flags & REXEC_SCREAM) && RX_MATCH_COPIED(prog))
2013 s = strbeg + (s - SvPVX_const(sv));
2014 DEBUG_EXECUTE_r( did_match = 1 );
2015 if (HOPc(s, -back_max) > last1) {
2016 last1 = HOPc(s, -back_min);
2017 s = HOPc(s, -back_max);
2020 char * const t = (last1 >= PL_bostr) ? HOPc(last1, 1) : last1 + 1;
2022 last1 = HOPc(s, -back_min);
2026 while (s <= last1) {
2027 if (regtry(®info, &s))
2033 while (s <= last1) {
2034 if (regtry(®info, &s))
2040 DEBUG_EXECUTE_r(if (!did_match) {
2041 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
2042 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
2043 PerlIO_printf(Perl_debug_log, "Did not find %s substr %s%s...\n",
2044 ((must == prog->anchored_substr || must == prog->anchored_utf8)
2045 ? "anchored" : "floating"),
2046 quoted, RE_SV_TAIL(must));
2050 else if ( (c = progi->regstclass) ) {
2052 const OPCODE op = OP(progi->regstclass);
2053 /* don't bother with what can't match */
2054 if (PL_regkind[op] != EXACT && op != CANY && PL_regkind[op] != TRIE)
2055 strend = HOPc(strend, -(minlen - 1));
2058 SV * const prop = sv_newmortal();
2059 regprop(prog, prop, c);
2061 RE_PV_QUOTED_DECL(quoted,do_utf8,PERL_DEBUG_PAD_ZERO(1),
2063 PerlIO_printf(Perl_debug_log,
2064 "Matching stclass %.*s against %s (%d chars)\n",
2065 (int)SvCUR(prop), SvPVX_const(prop),
2066 quoted, (int)(strend - s));
2069 if (find_byclass(prog, c, s, strend, ®info))
2071 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass... [regexec_flags]\n"));
2075 if (prog->float_substr != NULL || prog->float_utf8 != NULL) {
2080 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr))
2081 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
2082 float_real = do_utf8 ? prog->float_utf8 : prog->float_substr;
2084 if (flags & REXEC_SCREAM) {
2085 last = screaminstr(sv, float_real, s - strbeg,
2086 end_shift, &scream_pos, 1); /* last one */
2088 last = scream_olds; /* Only one occurrence. */
2089 /* we may be pointing at the wrong string */
2090 else if (RX_MATCH_COPIED(prog))
2091 s = strbeg + (s - SvPVX_const(sv));
2095 const char * const little = SvPV_const(float_real, len);
2097 if (SvTAIL(float_real)) {
2098 if (memEQ(strend - len + 1, little, len - 1))
2099 last = strend - len + 1;
2100 else if (!multiline)
2101 last = memEQ(strend - len, little, len)
2102 ? strend - len : NULL;
2108 last = rninstr(s, strend, little, little + len);
2110 last = strend; /* matching "$" */
2115 PerlIO_printf(Perl_debug_log,
2116 "%sCan't trim the tail, match fails (should not happen)%s\n",
2117 PL_colors[4], PL_colors[5]));
2118 goto phooey; /* Should not happen! */
2120 dontbother = strend - last + prog->float_min_offset;
2122 if (minlen && (dontbother < minlen))
2123 dontbother = minlen - 1;
2124 strend -= dontbother; /* this one's always in bytes! */
2125 /* We don't know much -- general case. */
2128 if (regtry(®info, &s))
2137 if (regtry(®info, &s))
2139 } while (s++ < strend);
2147 RX_MATCH_TAINTED_set(prog, PL_reg_flags & RF_tainted);
2149 if (PL_reg_eval_set)
2150 restore_pos(aTHX_ prog);
2151 if (prog->paren_names)
2152 (void)hv_iterinit(prog->paren_names);
2154 /* make sure $`, $&, $', and $digit will work later */
2155 if ( !(flags & REXEC_NOT_FIRST) ) {
2156 RX_MATCH_COPY_FREE(prog);
2157 if (flags & REXEC_COPY_STR) {
2158 const I32 i = PL_regeol - startpos + (stringarg - strbeg);
2159 #ifdef PERL_OLD_COPY_ON_WRITE
2161 || (SvFLAGS(sv) & CAN_COW_MASK) == CAN_COW_FLAGS)) {
2163 PerlIO_printf(Perl_debug_log,
2164 "Copy on write: regexp capture, type %d\n",
2167 prog->saved_copy = sv_setsv_cow(prog->saved_copy, sv);
2168 prog->subbeg = (char *)SvPVX_const(prog->saved_copy);
2169 assert (SvPOKp(prog->saved_copy));
2173 RX_MATCH_COPIED_on(prog);
2174 s = savepvn(strbeg, i);
2180 prog->subbeg = strbeg;
2181 prog->sublen = PL_regeol - strbeg; /* strend may have been modified */
2188 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch failed%s\n",
2189 PL_colors[4], PL_colors[5]));
2190 if (PL_reg_eval_set)
2191 restore_pos(aTHX_ prog);
2193 /* we failed :-( roll it back */
2194 swap_match_buff(prog);
2201 - regtry - try match at specific point
2203 STATIC I32 /* 0 failure, 1 success */
2204 S_regtry(pTHX_ regmatch_info *reginfo, char **startpos)
2208 regexp *prog = reginfo->prog;
2209 RXi_GET_DECL(prog,progi);
2210 GET_RE_DEBUG_FLAGS_DECL;
2211 reginfo->cutpoint=NULL;
2213 if ((prog->extflags & RXf_EVAL_SEEN) && !PL_reg_eval_set) {
2216 PL_reg_eval_set = RS_init;
2217 DEBUG_EXECUTE_r(DEBUG_s(
2218 PerlIO_printf(Perl_debug_log, " setting stack tmpbase at %"IVdf"\n",
2219 (IV)(PL_stack_sp - PL_stack_base));
2222 cxstack[cxstack_ix].blk_oldsp = PL_stack_sp - PL_stack_base;
2223 /* Otherwise OP_NEXTSTATE will free whatever on stack now. */
2225 /* Apparently this is not needed, judging by wantarray. */
2226 /* SAVEI8(cxstack[cxstack_ix].blk_gimme);
2227 cxstack[cxstack_ix].blk_gimme = G_SCALAR; */
2230 /* Make $_ available to executed code. */
2231 if (reginfo->sv != DEFSV) {
2233 DEFSV = reginfo->sv;
2236 if (!(SvTYPE(reginfo->sv) >= SVt_PVMG && SvMAGIC(reginfo->sv)
2237 && (mg = mg_find(reginfo->sv, PERL_MAGIC_regex_global)))) {
2238 /* prepare for quick setting of pos */
2239 #ifdef PERL_OLD_COPY_ON_WRITE
2240 if (SvIsCOW(reginfo->sv))
2241 sv_force_normal_flags(reginfo->sv, 0);
2243 mg = sv_magicext(reginfo->sv, NULL, PERL_MAGIC_regex_global,
2244 &PL_vtbl_mglob, NULL, 0);
2248 PL_reg_oldpos = mg->mg_len;
2249 SAVEDESTRUCTOR_X(restore_pos, prog);
2251 if (!PL_reg_curpm) {
2252 Newxz(PL_reg_curpm, 1, PMOP);
2255 SV* const repointer = newSViv(0);
2256 /* so we know which PL_regex_padav element is PL_reg_curpm */
2257 SvFLAGS(repointer) |= SVf_BREAK;
2258 av_push(PL_regex_padav,repointer);
2259 PL_reg_curpm->op_pmoffset = av_len(PL_regex_padav);
2260 PL_regex_pad = AvARRAY(PL_regex_padav);
2264 PM_SETRE(PL_reg_curpm, prog);
2265 PL_reg_oldcurpm = PL_curpm;
2266 PL_curpm = PL_reg_curpm;
2267 if (RX_MATCH_COPIED(prog)) {
2268 /* Here is a serious problem: we cannot rewrite subbeg,
2269 since it may be needed if this match fails. Thus
2270 $` inside (?{}) could fail... */
2271 PL_reg_oldsaved = prog->subbeg;
2272 PL_reg_oldsavedlen = prog->sublen;
2273 #ifdef PERL_OLD_COPY_ON_WRITE
2274 PL_nrs = prog->saved_copy;
2276 RX_MATCH_COPIED_off(prog);
2279 PL_reg_oldsaved = NULL;
2280 prog->subbeg = PL_bostr;
2281 prog->sublen = PL_regeol - PL_bostr; /* strend may have been modified */
2283 DEBUG_EXECUTE_r(PL_reg_starttry = *startpos);
2284 prog->offs[0].start = *startpos - PL_bostr;
2285 PL_reginput = *startpos;
2286 PL_reglastparen = &prog->lastparen;
2287 PL_reglastcloseparen = &prog->lastcloseparen;
2288 prog->lastparen = 0;
2289 prog->lastcloseparen = 0;
2291 PL_regoffs = prog->offs;
2292 if (PL_reg_start_tmpl <= prog->nparens) {
2293 PL_reg_start_tmpl = prog->nparens*3/2 + 3;
2294 if(PL_reg_start_tmp)
2295 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2297 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2300 /* XXXX What this code is doing here?!!! There should be no need
2301 to do this again and again, PL_reglastparen should take care of
2304 /* Tests pat.t#187 and split.t#{13,14} seem to depend on this code.
2305 * Actually, the code in regcppop() (which Ilya may be meaning by
2306 * PL_reglastparen), is not needed at all by the test suite
2307 * (op/regexp, op/pat, op/split), but that code is needed otherwise
2308 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
2309 * Meanwhile, this code *is* needed for the
2310 * above-mentioned test suite tests to succeed. The common theme
2311 * on those tests seems to be returning null fields from matches.
2312 * --jhi updated by dapm */
2314 if (prog->nparens) {
2315 regexp_paren_pair *pp = PL_regoffs;
2317 for (i = prog->nparens; i > (I32)*PL_reglastparen; i--) {
2325 if (regmatch(reginfo, progi->program + 1)) {
2326 PL_regoffs[0].end = PL_reginput - PL_bostr;
2329 if (reginfo->cutpoint)
2330 *startpos= reginfo->cutpoint;
2331 REGCP_UNWIND(lastcp);
2336 #define sayYES goto yes
2337 #define sayNO goto no
2338 #define sayNO_SILENT goto no_silent
2340 /* we dont use STMT_START/END here because it leads to
2341 "unreachable code" warnings, which are bogus, but distracting. */
2342 #define CACHEsayNO \
2343 if (ST.cache_mask) \
2344 PL_reg_poscache[ST.cache_offset] |= ST.cache_mask; \
2347 /* this is used to determine how far from the left messages like
2348 'failed...' are printed. It should be set such that messages
2349 are inline with the regop output that created them.
2351 #define REPORT_CODE_OFF 32
2354 /* Make sure there is a test for this +1 options in re_tests */
2355 #define TRIE_INITAL_ACCEPT_BUFFLEN 4;
2357 #define CHRTEST_UNINIT -1001 /* c1/c2 haven't been calculated yet */
2358 #define CHRTEST_VOID -1000 /* the c1/c2 "next char" test should be skipped */
2360 #define SLAB_FIRST(s) (&(s)->states[0])
2361 #define SLAB_LAST(s) (&(s)->states[PERL_REGMATCH_SLAB_SLOTS-1])
2363 /* grab a new slab and return the first slot in it */
2365 STATIC regmatch_state *
2368 #if PERL_VERSION < 9 && !defined(PERL_CORE)
2371 regmatch_slab *s = PL_regmatch_slab->next;
2373 Newx(s, 1, regmatch_slab);
2374 s->prev = PL_regmatch_slab;
2376 PL_regmatch_slab->next = s;
2378 PL_regmatch_slab = s;
2379 return SLAB_FIRST(s);
2383 /* push a new state then goto it */
2385 #define PUSH_STATE_GOTO(state, node) \
2387 st->resume_state = state; \
2390 /* push a new state with success backtracking, then goto it */
2392 #define PUSH_YES_STATE_GOTO(state, node) \
2394 st->resume_state = state; \
2395 goto push_yes_state;
2401 regmatch() - main matching routine
2403 This is basically one big switch statement in a loop. We execute an op,
2404 set 'next' to point the next op, and continue. If we come to a point which
2405 we may need to backtrack to on failure such as (A|B|C), we push a
2406 backtrack state onto the backtrack stack. On failure, we pop the top
2407 state, and re-enter the loop at the state indicated. If there are no more
2408 states to pop, we return failure.
2410 Sometimes we also need to backtrack on success; for example /A+/, where
2411 after successfully matching one A, we need to go back and try to
2412 match another one; similarly for lookahead assertions: if the assertion
2413 completes successfully, we backtrack to the state just before the assertion
2414 and then carry on. In these cases, the pushed state is marked as
2415 'backtrack on success too'. This marking is in fact done by a chain of
2416 pointers, each pointing to the previous 'yes' state. On success, we pop to
2417 the nearest yes state, discarding any intermediate failure-only states.
2418 Sometimes a yes state is pushed just to force some cleanup code to be
2419 called at the end of a successful match or submatch; e.g. (??{$re}) uses
2420 it to free the inner regex.
2422 Note that failure backtracking rewinds the cursor position, while
2423 success backtracking leaves it alone.
2425 A pattern is complete when the END op is executed, while a subpattern
2426 such as (?=foo) is complete when the SUCCESS op is executed. Both of these
2427 ops trigger the "pop to last yes state if any, otherwise return true"
2430 A common convention in this function is to use A and B to refer to the two
2431 subpatterns (or to the first nodes thereof) in patterns like /A*B/: so A is
2432 the subpattern to be matched possibly multiple times, while B is the entire
2433 rest of the pattern. Variable and state names reflect this convention.
2435 The states in the main switch are the union of ops and failure/success of
2436 substates associated with with that op. For example, IFMATCH is the op
2437 that does lookahead assertions /(?=A)B/ and so the IFMATCH state means
2438 'execute IFMATCH'; while IFMATCH_A is a state saying that we have just
2439 successfully matched A and IFMATCH_A_fail is a state saying that we have
2440 just failed to match A. Resume states always come in pairs. The backtrack
2441 state we push is marked as 'IFMATCH_A', but when that is popped, we resume
2442 at IFMATCH_A or IFMATCH_A_fail, depending on whether we are backtracking
2443 on success or failure.
2445 The struct that holds a backtracking state is actually a big union, with
2446 one variant for each major type of op. The variable st points to the
2447 top-most backtrack struct. To make the code clearer, within each
2448 block of code we #define ST to alias the relevant union.
2450 Here's a concrete example of a (vastly oversimplified) IFMATCH
2456 #define ST st->u.ifmatch
2458 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2459 ST.foo = ...; // some state we wish to save
2461 // push a yes backtrack state with a resume value of
2462 // IFMATCH_A/IFMATCH_A_fail, then continue execution at the
2464 PUSH_YES_STATE_GOTO(IFMATCH_A, A);
2467 case IFMATCH_A: // we have successfully executed A; now continue with B
2469 bar = ST.foo; // do something with the preserved value
2472 case IFMATCH_A_fail: // A failed, so the assertion failed
2473 ...; // do some housekeeping, then ...
2474 sayNO; // propagate the failure
2481 For any old-timers reading this who are familiar with the old recursive
2482 approach, the code above is equivalent to:
2484 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2493 ...; // do some housekeeping, then ...
2494 sayNO; // propagate the failure
2497 The topmost backtrack state, pointed to by st, is usually free. If you
2498 want to claim it, populate any ST.foo fields in it with values you wish to
2499 save, then do one of
2501 PUSH_STATE_GOTO(resume_state, node);
2502 PUSH_YES_STATE_GOTO(resume_state, node);
2504 which sets that backtrack state's resume value to 'resume_state', pushes a
2505 new free entry to the top of the backtrack stack, then goes to 'node'.
2506 On backtracking, the free slot is popped, and the saved state becomes the
2507 new free state. An ST.foo field in this new top state can be temporarily
2508 accessed to retrieve values, but once the main loop is re-entered, it
2509 becomes available for reuse.
2511 Note that the depth of the backtrack stack constantly increases during the
2512 left-to-right execution of the pattern, rather than going up and down with
2513 the pattern nesting. For example the stack is at its maximum at Z at the
2514 end of the pattern, rather than at X in the following:
2516 /(((X)+)+)+....(Y)+....Z/
2518 The only exceptions to this are lookahead/behind assertions and the cut,
2519 (?>A), which pop all the backtrack states associated with A before
2522 Bascktrack state structs are allocated in slabs of about 4K in size.
2523 PL_regmatch_state and st always point to the currently active state,
2524 and PL_regmatch_slab points to the slab currently containing
2525 PL_regmatch_state. The first time regmatch() is called, the first slab is
2526 allocated, and is never freed until interpreter destruction. When the slab
2527 is full, a new one is allocated and chained to the end. At exit from
2528 regmatch(), slabs allocated since entry are freed.
2533 #define DEBUG_STATE_pp(pp) \
2535 DUMP_EXEC_POS(locinput, scan, do_utf8); \
2536 PerlIO_printf(Perl_debug_log, \
2537 " %*s"pp" %s%s%s%s%s\n", \
2539 PL_reg_name[st->resume_state], \
2540 ((st==yes_state||st==mark_state) ? "[" : ""), \
2541 ((st==yes_state) ? "Y" : ""), \
2542 ((st==mark_state) ? "M" : ""), \
2543 ((st==yes_state||st==mark_state) ? "]" : "") \
2548 #define REG_NODE_NUM(x) ((x) ? (int)((x)-prog) : -1)
2553 S_debug_start_match(pTHX_ const regexp *prog, const bool do_utf8,
2554 const char *start, const char *end, const char *blurb)
2556 const bool utf8_pat= prog->extflags & RXf_UTF8 ? 1 : 0;
2560 RE_PV_QUOTED_DECL(s0, utf8_pat, PERL_DEBUG_PAD_ZERO(0),
2561 prog->precomp, prog->prelen, 60);
2563 RE_PV_QUOTED_DECL(s1, do_utf8, PERL_DEBUG_PAD_ZERO(1),
2564 start, end - start, 60);
2566 PerlIO_printf(Perl_debug_log,
2567 "%s%s REx%s %s against %s\n",
2568 PL_colors[4], blurb, PL_colors[5], s0, s1);
2570 if (do_utf8||utf8_pat)
2571 PerlIO_printf(Perl_debug_log, "UTF-8 %s%s%s...\n",
2572 utf8_pat ? "pattern" : "",
2573 utf8_pat && do_utf8 ? " and " : "",
2574 do_utf8 ? "string" : ""
2580 S_dump_exec_pos(pTHX_ const char *locinput,
2581 const regnode *scan,
2582 const char *loc_regeol,
2583 const char *loc_bostr,
2584 const char *loc_reg_starttry,
2587 const int docolor = *PL_colors[0] || *PL_colors[2] || *PL_colors[4];
2588 const int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */
2589 int l = (loc_regeol - locinput) > taill ? taill : (loc_regeol - locinput);
2590 /* The part of the string before starttry has one color
2591 (pref0_len chars), between starttry and current
2592 position another one (pref_len - pref0_len chars),
2593 after the current position the third one.
2594 We assume that pref0_len <= pref_len, otherwise we
2595 decrease pref0_len. */
2596 int pref_len = (locinput - loc_bostr) > (5 + taill) - l
2597 ? (5 + taill) - l : locinput - loc_bostr;
2600 while (do_utf8 && UTF8_IS_CONTINUATION(*(U8*)(locinput - pref_len)))
2602 pref0_len = pref_len - (locinput - loc_reg_starttry);
2603 if (l + pref_len < (5 + taill) && l < loc_regeol - locinput)
2604 l = ( loc_regeol - locinput > (5 + taill) - pref_len
2605 ? (5 + taill) - pref_len : loc_regeol - locinput);
2606 while (do_utf8 && UTF8_IS_CONTINUATION(*(U8*)(locinput + l)))
2610 if (pref0_len > pref_len)
2611 pref0_len = pref_len;
2613 const int is_uni = (do_utf8 && OP(scan) != CANY) ? 1 : 0;
2615 RE_PV_COLOR_DECL(s0,len0,is_uni,PERL_DEBUG_PAD(0),
2616 (locinput - pref_len),pref0_len, 60, 4, 5);
2618 RE_PV_COLOR_DECL(s1,len1,is_uni,PERL_DEBUG_PAD(1),
2619 (locinput - pref_len + pref0_len),
2620 pref_len - pref0_len, 60, 2, 3);
2622 RE_PV_COLOR_DECL(s2,len2,is_uni,PERL_DEBUG_PAD(2),
2623 locinput, loc_regeol - locinput, 10, 0, 1);
2625 const STRLEN tlen=len0+len1+len2;
2626 PerlIO_printf(Perl_debug_log,
2627 "%4"IVdf" <%.*s%.*s%s%.*s>%*s|",
2628 (IV)(locinput - loc_bostr),
2631 (docolor ? "" : "> <"),
2633 (int)(tlen > 19 ? 0 : 19 - tlen),
2640 /* reg_check_named_buff_matched()
2641 * Checks to see if a named buffer has matched. The data array of
2642 * buffer numbers corresponding to the buffer is expected to reside
2643 * in the regexp->data->data array in the slot stored in the ARG() of
2644 * node involved. Note that this routine doesn't actually care about the
2645 * name, that information is not preserved from compilation to execution.
2646 * Returns the index of the leftmost defined buffer with the given name
2647 * or 0 if non of the buffers matched.
2650 S_reg_check_named_buff_matched(pTHX_ const regexp *rex, const regnode *scan) {
2652 RXi_GET_DECL(rex,rexi);
2653 SV *sv_dat=(SV*)rexi->data->data[ ARG( scan ) ];
2654 I32 *nums=(I32*)SvPVX(sv_dat);
2655 for ( n=0; n<SvIVX(sv_dat); n++ ) {
2656 if ((I32)*PL_reglastparen >= nums[n] &&
2657 PL_regoffs[nums[n]].end != -1)
2666 /* free all slabs above current one - called during LEAVE_SCOPE */
2669 S_clear_backtrack_stack(pTHX_ void *p)
2671 regmatch_slab *s = PL_regmatch_slab->next;
2676 PL_regmatch_slab->next = NULL;
2678 regmatch_slab * const osl = s;
2685 #define SETREX(Re1,Re2) \
2686 if (PL_reg_eval_set) PM_SETRE((PL_reg_curpm), (Re2)); \
2689 STATIC I32 /* 0 failure, 1 success */
2690 S_regmatch(pTHX_ regmatch_info *reginfo, regnode *prog)
2692 #if PERL_VERSION < 9 && !defined(PERL_CORE)
2696 register const bool do_utf8 = PL_reg_match_utf8;
2697 const U32 uniflags = UTF8_ALLOW_DEFAULT;
2699 regexp *rex = reginfo->prog;
2700 RXi_GET_DECL(rex,rexi);
2704 /* the current state. This is a cached copy of PL_regmatch_state */
2705 register regmatch_state *st;
2707 /* cache heavy used fields of st in registers */
2708 register regnode *scan;
2709 register regnode *next;
2710 register U32 n = 0; /* general value; init to avoid compiler warning */
2711 register I32 ln = 0; /* len or last; init to avoid compiler warning */
2712 register char *locinput = PL_reginput;
2713 register I32 nextchr; /* is always set to UCHARAT(locinput) */
2715 bool result = 0; /* return value of S_regmatch */
2716 int depth = 0; /* depth of backtrack stack */
2717 U32 nochange_depth = 0; /* depth of GOSUB recursion with nochange */
2718 const U32 max_nochange_depth =
2719 (3 * rex->nparens > MAX_RECURSE_EVAL_NOCHANGE_DEPTH) ?
2720 3 * rex->nparens : MAX_RECURSE_EVAL_NOCHANGE_DEPTH;
2722 regmatch_state *yes_state = NULL; /* state to pop to on success of
2724 /* mark_state piggy backs on the yes_state logic so that when we unwind
2725 the stack on success we can update the mark_state as we go */
2726 regmatch_state *mark_state = NULL; /* last mark state we have seen */
2728 regmatch_state *cur_eval = NULL; /* most recent EVAL_AB state */
2729 struct regmatch_state *cur_curlyx = NULL; /* most recent curlyx */
2731 bool no_final = 0; /* prevent failure from backtracking? */
2732 bool do_cutgroup = 0; /* no_final only until next branch/trie entry */
2733 char *startpoint = PL_reginput;
2734 SV *popmark = NULL; /* are we looking for a mark? */
2735 SV *sv_commit = NULL; /* last mark name seen in failure */
2736 SV *sv_yes_mark = NULL; /* last mark name we have seen
2737 during a successfull match */
2738 U32 lastopen = 0; /* last open we saw */
2739 bool has_cutgroup = RX_HAS_CUTGROUP(rex) ? 1 : 0;
2741 SV* const oreplsv = GvSV(PL_replgv);
2744 /* these three flags are set by various ops to signal information to
2745 * the very next op. They have a useful lifetime of exactly one loop
2746 * iteration, and are not preserved or restored by state pushes/pops
2748 bool sw = 0; /* the condition value in (?(cond)a|b) */
2749 bool minmod = 0; /* the next "{n,m}" is a "{n,m}?" */
2750 int logical = 0; /* the following EVAL is:
2754 or the following IFMATCH/UNLESSM is:
2755 false: plain (?=foo)
2756 true: used as a condition: (?(?=foo))
2760 GET_RE_DEBUG_FLAGS_DECL;
2763 DEBUG_OPTIMISE_r( DEBUG_EXECUTE_r({
2764 PerlIO_printf(Perl_debug_log,"regmatch start\n");
2766 /* on first ever call to regmatch, allocate first slab */
2767 if (!PL_regmatch_slab) {
2768 Newx(PL_regmatch_slab, 1, regmatch_slab);
2769 PL_regmatch_slab->prev = NULL;
2770 PL_regmatch_slab->next = NULL;
2771 PL_regmatch_state = SLAB_FIRST(PL_regmatch_slab);
2774 oldsave = PL_savestack_ix;
2775 SAVEDESTRUCTOR_X(S_clear_backtrack_stack, NULL);
2776 SAVEVPTR(PL_regmatch_slab);
2777 SAVEVPTR(PL_regmatch_state);
2779 /* grab next free state slot */
2780 st = ++PL_regmatch_state;
2781 if (st > SLAB_LAST(PL_regmatch_slab))
2782 st = PL_regmatch_state = S_push_slab(aTHX);
2784 /* Note that nextchr is a byte even in UTF */
2785 nextchr = UCHARAT(locinput);
2787 while (scan != NULL) {
2790 SV * const prop = sv_newmortal();
2791 regnode *rnext=regnext(scan);
2792 DUMP_EXEC_POS( locinput, scan, do_utf8 );
2793 regprop(rex, prop, scan);
2795 PerlIO_printf(Perl_debug_log,
2796 "%3"IVdf":%*s%s(%"IVdf")\n",
2797 (IV)(scan - rexi->program), depth*2, "",
2799 (PL_regkind[OP(scan)] == END || !rnext) ?
2800 0 : (IV)(rnext - rexi->program));
2803 next = scan + NEXT_OFF(scan);
2806 state_num = OP(scan);
2809 switch (state_num) {
2811 if (locinput == PL_bostr)
2813 /* reginfo->till = reginfo->bol; */
2818 if (locinput == PL_bostr ||
2819 ((nextchr || locinput < PL_regeol) && locinput[-1] == '\n'))
2825 if (locinput == PL_bostr)
2829 if (locinput == reginfo->ganch)
2834 /* update the startpoint */
2835 st->u.keeper.val = PL_regoffs[0].start;
2836 PL_reginput = locinput;
2837 PL_regoffs[0].start = locinput - PL_bostr;
2838 PUSH_STATE_GOTO(KEEPS_next, next);
2840 case KEEPS_next_fail:
2841 /* rollback the start point change */
2842 PL_regoffs[0].start = st->u.keeper.val;
2848 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
2853 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
2855 if (PL_regeol - locinput > 1)
2859 if (PL_regeol != locinput)
2863 if (!nextchr && locinput >= PL_regeol)
2866 locinput += PL_utf8skip[nextchr];
2867 if (locinput > PL_regeol)
2869 nextchr = UCHARAT(locinput);
2872 nextchr = UCHARAT(++locinput);
2875 if (!nextchr && locinput >= PL_regeol)
2877 nextchr = UCHARAT(++locinput);
2880 if ((!nextchr && locinput >= PL_regeol) || nextchr == '\n')
2883 locinput += PL_utf8skip[nextchr];
2884 if (locinput > PL_regeol)
2886 nextchr = UCHARAT(locinput);
2889 nextchr = UCHARAT(++locinput);
2893 #define ST st->u.trie
2895 /* In this case the charclass data is available inline so
2896 we can fail fast without a lot of extra overhead.
2898 if (scan->flags == EXACT || !do_utf8) {
2899 if(!ANYOF_BITMAP_TEST(scan, *locinput)) {
2901 PerlIO_printf(Perl_debug_log,
2902 "%*s %sfailed to match trie start class...%s\n",
2903 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2912 /* what type of TRIE am I? (utf8 makes this contextual) */
2913 DECL_TRIE_TYPE(scan);
2915 /* what trie are we using right now */
2916 reg_trie_data * const trie
2917 = (reg_trie_data*)rexi->data->data[ ARG( scan ) ];
2918 HV * widecharmap = (HV *)rexi->data->data[ ARG( scan ) + 1 ];
2919 U32 state = trie->startstate;
2921 if (trie->bitmap && trie_type != trie_utf8_fold &&
2922 !TRIE_BITMAP_TEST(trie,*locinput)
2924 if (trie->states[ state ].wordnum) {
2926 PerlIO_printf(Perl_debug_log,
2927 "%*s %smatched empty string...%s\n",
2928 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2933 PerlIO_printf(Perl_debug_log,
2934 "%*s %sfailed to match trie start class...%s\n",
2935 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2942 U8 *uc = ( U8* )locinput;
2946 U8 *uscan = (U8*)NULL;
2948 SV *sv_accept_buff = NULL;
2949 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
2951 ST.accepted = 0; /* how many accepting states we have seen */
2953 ST.jump = trie->jump;
2956 traverse the TRIE keeping track of all accepting states
2957 we transition through until we get to a failing node.
2960 while ( state && uc <= (U8*)PL_regeol ) {
2961 U32 base = trie->states[ state ].trans.base;
2964 /* We use charid to hold the wordnum as we don't use it
2965 for charid until after we have done the wordnum logic.
2966 We define an alias just so that the wordnum logic reads
2969 #define got_wordnum charid
2970 got_wordnum = trie->states[ state ].wordnum;
2972 if ( got_wordnum ) {
2973 if ( ! ST.accepted ) {
2975 /* SAVETMPS; */ /* XXX is this necessary? dmq */
2976 bufflen = TRIE_INITAL_ACCEPT_BUFFLEN;
2977 sv_accept_buff=newSV(bufflen *
2978 sizeof(reg_trie_accepted) - 1);
2979 SvCUR_set(sv_accept_buff, 0);
2980 SvPOK_on(sv_accept_buff);
2981 sv_2mortal(sv_accept_buff);
2984 (reg_trie_accepted*)SvPV_nolen(sv_accept_buff );
2987 if (ST.accepted >= bufflen) {
2989 ST.accept_buff =(reg_trie_accepted*)
2990 SvGROW(sv_accept_buff,
2991 bufflen * sizeof(reg_trie_accepted));
2993 SvCUR_set(sv_accept_buff,SvCUR(sv_accept_buff)
2994 + sizeof(reg_trie_accepted));
2997 ST.accept_buff[ST.accepted].wordnum = got_wordnum;
2998 ST.accept_buff[ST.accepted].endpos = uc;
3000 } while (trie->nextword && (got_wordnum= trie->nextword[got_wordnum]));
3004 DEBUG_TRIE_EXECUTE_r({
3005 DUMP_EXEC_POS( (char *)uc, scan, do_utf8 );
3006 PerlIO_printf( Perl_debug_log,
3007 "%*s %sState: %4"UVxf" Accepted: %4"UVxf" ",
3008 2+depth * 2, "", PL_colors[4],
3009 (UV)state, (UV)ST.accepted );
3013 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
3014 uscan, len, uvc, charid, foldlen,
3018 (base + charid > trie->uniquecharcount )
3019 && (base + charid - 1 - trie->uniquecharcount
3021 && trie->trans[base + charid - 1 -
3022 trie->uniquecharcount].check == state)
3024 state = trie->trans[base + charid - 1 -
3025 trie->uniquecharcount ].next;
3036 DEBUG_TRIE_EXECUTE_r(
3037 PerlIO_printf( Perl_debug_log,
3038 "Charid:%3x CP:%4"UVxf" After State: %4"UVxf"%s\n",
3039 charid, uvc, (UV)state, PL_colors[5] );
3046 PerlIO_printf( Perl_debug_log,
3047 "%*s %sgot %"IVdf" possible matches%s\n",
3048 REPORT_CODE_OFF + depth * 2, "",
3049 PL_colors[4], (IV)ST.accepted, PL_colors[5] );
3052 goto trie_first_try; /* jump into the fail handler */
3054 case TRIE_next_fail: /* we failed - try next alterative */
3056 REGCP_UNWIND(ST.cp);
3057 for (n = *PL_reglastparen; n > ST.lastparen; n--)
3058 PL_regoffs[n].end = -1;
3059 *PL_reglastparen = n;
3068 ST.lastparen = *PL_reglastparen;
3071 if ( ST.accepted == 1 ) {
3072 /* only one choice left - just continue */
3074 AV *const trie_words
3075 = (AV *) rexi->data->data[ARG(ST.me)+TRIE_WORDS_OFFSET];
3076 SV ** const tmp = av_fetch( trie_words,
3077 ST.accept_buff[ 0 ].wordnum-1, 0 );
3078 SV *sv= tmp ? sv_newmortal() : NULL;
3080 PerlIO_printf( Perl_debug_log,
3081 "%*s %sonly one match left: #%d <%s>%s\n",
3082 REPORT_CODE_OFF+depth*2, "", PL_colors[4],
3083 ST.accept_buff[ 0 ].wordnum,
3084 tmp ? pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 0,
3085 PL_colors[0], PL_colors[1],
3086 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0)
3088 : "not compiled under -Dr",
3091 PL_reginput = (char *)ST.accept_buff[ 0 ].endpos;
3092 /* in this case we free tmps/leave before we call regmatch
3093 as we wont be using accept_buff again. */
3095 locinput = PL_reginput;
3096 nextchr = UCHARAT(locinput);
3097 if ( !ST.jump || !ST.jump[ST.accept_buff[0].wordnum])
3100 scan = ST.me + ST.jump[ST.accept_buff[0].wordnum];
3101 if (!has_cutgroup) {
3106 PUSH_YES_STATE_GOTO(TRIE_next, scan);
3109 continue; /* execute rest of RE */
3112 if ( !ST.accepted-- ) {
3114 PerlIO_printf( Perl_debug_log,
3115 "%*s %sTRIE failed...%s\n",
3116 REPORT_CODE_OFF+depth*2, "",
3127 There are at least two accepting states left. Presumably
3128 the number of accepting states is going to be low,
3129 typically two. So we simply scan through to find the one
3130 with lowest wordnum. Once we find it, we swap the last
3131 state into its place and decrement the size. We then try to
3132 match the rest of the pattern at the point where the word
3133 ends. If we succeed, control just continues along the
3134 regex; if we fail we return here to try the next accepting
3141 for( cur = 1 ; cur <= ST.accepted ; cur++ ) {
3142 DEBUG_TRIE_EXECUTE_r(
3143 PerlIO_printf( Perl_debug_log,
3144 "%*s %sgot %"IVdf" (%d) as best, looking at %"IVdf" (%d)%s\n",
3145 REPORT_CODE_OFF + depth * 2, "", PL_colors[4],
3146 (IV)best, ST.accept_buff[ best ].wordnum, (IV)cur,
3147 ST.accept_buff[ cur ].wordnum, PL_colors[5] );
3150 if (ST.accept_buff[cur].wordnum <
3151 ST.accept_buff[best].wordnum)
3156 AV *const trie_words
3157 = (AV *) rexi->data->data[ARG(ST.me)+TRIE_WORDS_OFFSET];
3158 SV ** const tmp = av_fetch( trie_words,
3159 ST.accept_buff[ best ].wordnum - 1, 0 );
3160 regnode *nextop=(!ST.jump || !ST.jump[ST.accept_buff[best].wordnum]) ?
3162 ST.me + ST.jump[ST.accept_buff[best].wordnum];
3163 SV *sv= tmp ? sv_newmortal() : NULL;
3165 PerlIO_printf( Perl_debug_log,
3166 "%*s %strying alternation #%d <%s> at node #%d %s\n",
3167 REPORT_CODE_OFF+depth*2, "", PL_colors[4],
3168 ST.accept_buff[best].wordnum,
3169 tmp ? pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 0,
3170 PL_colors[0], PL_colors[1],
3171 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0)
3172 ) : "not compiled under -Dr",
3173 REG_NODE_NUM(nextop),
3177 if ( best<ST.accepted ) {
3178 reg_trie_accepted tmp = ST.accept_buff[ best ];
3179 ST.accept_buff[ best ] = ST.accept_buff[ ST.accepted ];
3180 ST.accept_buff[ ST.accepted ] = tmp;
3183 PL_reginput = (char *)ST.accept_buff[ best ].endpos;
3184 if ( !ST.jump || !ST.jump[ST.accept_buff[best].wordnum]) {
3187 scan = ST.me + ST.jump[ST.accept_buff[best].wordnum];
3189 PUSH_YES_STATE_GOTO(TRIE_next, scan);
3200 char *s = STRING(scan);
3202 if (do_utf8 != UTF) {
3203 /* The target and the pattern have differing utf8ness. */
3205 const char * const e = s + ln;
3208 /* The target is utf8, the pattern is not utf8. */
3213 if (NATIVE_TO_UNI(*(U8*)s) !=
3214 utf8n_to_uvuni((U8*)l, UTF8_MAXBYTES, &ulen,
3222 /* The target is not utf8, the pattern is utf8. */
3227 if (NATIVE_TO_UNI(*((U8*)l)) !=
3228 utf8n_to_uvuni((U8*)s, UTF8_MAXBYTES, &ulen,
3236 nextchr = UCHARAT(locinput);
3239 /* The target and the pattern have the same utf8ness. */
3240 /* Inline the first character, for speed. */
3241 if (UCHARAT(s) != nextchr)
3243 if (PL_regeol - locinput < ln)
3245 if (ln > 1 && memNE(s, locinput, ln))
3248 nextchr = UCHARAT(locinput);
3252 PL_reg_flags |= RF_tainted;
3255 char * const s = STRING(scan);
3258 if (do_utf8 || UTF) {
3259 /* Either target or the pattern are utf8. */
3260 const char * const l = locinput;
3261 char *e = PL_regeol;
3263 if (ibcmp_utf8(s, 0, ln, (bool)UTF,
3264 l, &e, 0, do_utf8)) {
3265 /* One more case for the sharp s:
3266 * pack("U0U*", 0xDF) =~ /ss/i,
3267 * the 0xC3 0x9F are the UTF-8
3268 * byte sequence for the U+00DF. */
3271 toLOWER(s[0]) == 's' &&
3273 toLOWER(s[1]) == 's' &&
3280 nextchr = UCHARAT(locinput);
3284 /* Neither the target and the pattern are utf8. */
3286 /* Inline the first character, for speed. */
3287 if (UCHARAT(s) != nextchr &&
3288 UCHARAT(s) != ((OP(scan) == EXACTF)
3289 ? PL_fold : PL_fold_locale)[nextchr])
3291 if (PL_regeol - locinput < ln)
3293 if (ln > 1 && (OP(scan) == EXACTF
3294 ? ibcmp(s, locinput, ln)
3295 : ibcmp_locale(s, locinput, ln)))
3298 nextchr = UCHARAT(locinput);
3303 STRLEN inclasslen = PL_regeol - locinput;
3305 if (!reginclass(rex, scan, (U8*)locinput, &inclasslen, do_utf8))
3307 if (locinput >= PL_regeol)
3309 locinput += inclasslen ? inclasslen : UTF8SKIP(locinput);
3310 nextchr = UCHARAT(locinput);
3315 nextchr = UCHARAT(locinput);
3316 if (!REGINCLASS(rex, scan, (U8*)locinput))
3318 if (!nextchr && locinput >= PL_regeol)
3320 nextchr = UCHARAT(++locinput);
3324 /* If we might have the case of the German sharp s
3325 * in a casefolding Unicode character class. */
3327 if (ANYOF_FOLD_SHARP_S(scan, locinput, PL_regeol)) {
3328 locinput += SHARP_S_SKIP;
3329 nextchr = UCHARAT(locinput);
3335 PL_reg_flags |= RF_tainted;
3341 LOAD_UTF8_CHARCLASS_ALNUM();
3342 if (!(OP(scan) == ALNUM
3343 ? (bool)swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
3344 : isALNUM_LC_utf8((U8*)locinput)))
3348 locinput += PL_utf8skip[nextchr];
3349 nextchr = UCHARAT(locinput);
3352 if (!(OP(scan) == ALNUM
3353 ? isALNUM(nextchr) : isALNUM_LC(nextchr)))
3355 nextchr = UCHARAT(++locinput);
3358 PL_reg_flags |= RF_tainted;
3361 if (!nextchr && locinput >= PL_regeol)
3364 LOAD_UTF8_CHARCLASS_ALNUM();
3365 if (OP(scan) == NALNUM
3366 ? (bool)swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
3367 : isALNUM_LC_utf8((U8*)locinput))
3371 locinput += PL_utf8skip[nextchr];
3372 nextchr = UCHARAT(locinput);
3375 if (OP(scan) == NALNUM
3376 ? isALNUM(nextchr) : isALNUM_LC(nextchr))
3378 nextchr = UCHARAT(++locinput);
3382 PL_reg_flags |= RF_tainted;
3386 /* was last char in word? */
3388 if (locinput == PL_bostr)
3391 const U8 * const r = reghop3((U8*)locinput, -1, (U8*)PL_bostr);
3393 ln = utf8n_to_uvchr(r, UTF8SKIP(r), 0, uniflags);
3395 if (OP(scan) == BOUND || OP(scan) == NBOUND) {
3396 ln = isALNUM_uni(ln);
3397 LOAD_UTF8_CHARCLASS_ALNUM();
3398 n = swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8);
3401 ln = isALNUM_LC_uvchr(UNI_TO_NATIVE(ln));
3402 n = isALNUM_LC_utf8((U8*)locinput);
3406 ln = (locinput != PL_bostr) ?
3407 UCHARAT(locinput - 1) : '\n';
3408 if (OP(scan) == BOUND || OP(scan) == NBOUND) {
3410 n = isALNUM(nextchr);
3413 ln = isALNUM_LC(ln);
3414 n = isALNUM_LC(nextchr);
3417 if (((!ln) == (!n)) == (OP(scan) == BOUND ||
3418 OP(scan) == BOUNDL))
3422 PL_reg_flags |= RF_tainted;
3428 if (UTF8_IS_CONTINUED(nextchr)) {
3429 LOAD_UTF8_CHARCLASS_SPACE();
3430 if (!(OP(scan) == SPACE
3431 ? (bool)swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
3432 : isSPACE_LC_utf8((U8*)locinput)))
3436 locinput += PL_utf8skip[nextchr];
3437 nextchr = UCHARAT(locinput);
3440 if (!(OP(scan) == SPACE
3441 ? isSPACE(nextchr) : isSPACE_LC(nextchr)))
3443 nextchr = UCHARAT(++locinput);
3446 if (!(OP(scan) == SPACE
3447 ? isSPACE(nextchr) : isSPACE_LC(nextchr)))
3449 nextchr = UCHARAT(++locinput);
3453 PL_reg_flags |= RF_tainted;
3456 if (!nextchr && locinput >= PL_regeol)
3459 LOAD_UTF8_CHARCLASS_SPACE();
3460 if (OP(scan) == NSPACE
3461 ? (bool)swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
3462 : isSPACE_LC_utf8((U8*)locinput))
3466 locinput += PL_utf8skip[nextchr];
3467 nextchr = UCHARAT(locinput);
3470 if (OP(scan) == NSPACE
3471 ? isSPACE(nextchr) : isSPACE_LC(nextchr))
3473 nextchr = UCHARAT(++locinput);
3476 PL_reg_flags |= RF_tainted;
3482 LOAD_UTF8_CHARCLASS_DIGIT();
3483 if (!(OP(scan) == DIGIT
3484 ? (bool)swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
3485 : isDIGIT_LC_utf8((U8*)locinput)))
3489 locinput += PL_utf8skip[nextchr];
3490 nextchr = UCHARAT(locinput);
3493 if (!(OP(scan) == DIGIT
3494 ? isDIGIT(nextchr) : isDIGIT_LC(nextchr)))
3496 nextchr = UCHARAT(++locinput);
3499 PL_reg_flags |= RF_tainted;
3502 if (!nextchr && locinput >= PL_regeol)
3505 LOAD_UTF8_CHARCLASS_DIGIT();
3506 if (OP(scan) == NDIGIT
3507 ? (bool)swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
3508 : isDIGIT_LC_utf8((U8*)locinput))
3512 locinput += PL_utf8skip[nextchr];
3513 nextchr = UCHARAT(locinput);
3516 if (OP(scan) == NDIGIT
3517 ? isDIGIT(nextchr) : isDIGIT_LC(nextchr))
3519 nextchr = UCHARAT(++locinput);
3522 if (locinput >= PL_regeol)
3525 LOAD_UTF8_CHARCLASS_MARK();
3526 if (swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8))
3528 locinput += PL_utf8skip[nextchr];
3529 while (locinput < PL_regeol &&
3530 swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8))
3531 locinput += UTF8SKIP(locinput);
3532 if (locinput > PL_regeol)
3537 nextchr = UCHARAT(locinput);
3544 PL_reg_flags |= RF_tainted;
3549 n = reg_check_named_buff_matched(rex,scan);
3552 type = REF + ( type - NREF );
3559 PL_reg_flags |= RF_tainted;
3563 n = ARG(scan); /* which paren pair */
3566 ln = PL_regoffs[n].start;
3567 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
3568 if (*PL_reglastparen < n || ln == -1)
3569 sayNO; /* Do not match unless seen CLOSEn. */
3570 if (ln == PL_regoffs[n].end)
3574 if (do_utf8 && type != REF) { /* REF can do byte comparison */
3576 const char *e = PL_bostr + PL_regoffs[n].end;
3578 * Note that we can't do the "other character" lookup trick as
3579 * in the 8-bit case (no pun intended) because in Unicode we
3580 * have to map both upper and title case to lower case.
3584 STRLEN ulen1, ulen2;
3585 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
3586 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
3590 toLOWER_utf8((U8*)s, tmpbuf1, &ulen1);
3591 toLOWER_utf8((U8*)l, tmpbuf2, &ulen2);
3592 if (ulen1 != ulen2 || memNE((char *)tmpbuf1, (char *)tmpbuf2, ulen1))
3599 nextchr = UCHARAT(locinput);
3603 /* Inline the first character, for speed. */
3604 if (UCHARAT(s) != nextchr &&
3606 (UCHARAT(s) != (type == REFF
3607 ? PL_fold : PL_fold_locale)[nextchr])))
3609 ln = PL_regoffs[n].end - ln;
3610 if (locinput + ln > PL_regeol)
3612 if (ln > 1 && (type == REF
3613 ? memNE(s, locinput, ln)
3615 ? ibcmp(s, locinput, ln)
3616 : ibcmp_locale(s, locinput, ln))))
3619 nextchr = UCHARAT(locinput);
3629 #define ST st->u.eval
3633 regexp_internal *rei;
3634 regnode *startpoint;
3637 case GOSUB: /* /(...(?1))/ /(...(?&foo))/ */
3638 if (cur_eval && cur_eval->locinput==locinput) {
3639 if (cur_eval->u.eval.close_paren == (U32)ARG(scan))
3640 Perl_croak(aTHX_ "Infinite recursion in regex");
3641 if ( ++nochange_depth > max_nochange_depth )
3643 "Pattern subroutine nesting without pos change"
3644 " exceeded limit in regex");
3650 (void)ReREFCNT_inc(rex);
3651 if (OP(scan)==GOSUB) {
3652 startpoint = scan + ARG2L(scan);
3653 ST.close_paren = ARG(scan);
3655 startpoint = rei->program+1;
3658 goto eval_recurse_doit;
3660 case EVAL: /* /(?{A})B/ /(??{A})B/ and /(?(?{A})X|Y)B/ */
3661 if (cur_eval && cur_eval->locinput==locinput) {
3662 if ( ++nochange_depth > max_nochange_depth )
3663 Perl_croak(aTHX_ "EVAL without pos change exceeded limit in regex");
3668 /* execute the code in the {...} */
3670 SV ** const before = SP;
3671 OP_4tree * const oop = PL_op;
3672 COP * const ocurcop = PL_curcop;
3676 PL_op = (OP_4tree*)rexi->data->data[n];
3677 DEBUG_STATE_r( PerlIO_printf(Perl_debug_log,
3678 " re_eval 0x%"UVxf"\n", PTR2UV(PL_op)) );
3679 PAD_SAVE_LOCAL(old_comppad, (PAD*)rexi->data->data[n + 2]);
3680 PL_regoffs[0].end = PL_reg_magic->mg_len = locinput - PL_bostr;
3683 SV *sv_mrk = get_sv("REGMARK", 1);
3684 sv_setsv(sv_mrk, sv_yes_mark);
3687 CALLRUNOPS(aTHX); /* Scalar context. */
3690 ret = &PL_sv_undef; /* protect against empty (?{}) blocks. */
3697 PAD_RESTORE_LOCAL(old_comppad);
3698 PL_curcop = ocurcop;
3701 sv_setsv(save_scalar(PL_replgv), ret);
3705 if (logical == 2) { /* Postponed subexpression: /(??{...})/ */
3708 /* extract RE object from returned value; compiling if
3713 if(SvROK(ret) && SvSMAGICAL(sv = SvRV(ret)))
3714 mg = mg_find(sv, PERL_MAGIC_qr);
3715 else if (SvSMAGICAL(ret)) {
3716 if (SvGMAGICAL(ret))
3717 sv_unmagic(ret, PERL_MAGIC_qr);
3719 mg = mg_find(ret, PERL_MAGIC_qr);
3723 re = reg_temp_copy((regexp *)mg->mg_obj); /*XXX:dmq*/
3727 const I32 osize = PL_regsize;
3729 if (DO_UTF8(ret)) pm_flags |= RXf_UTF8;
3730 re = CALLREGCOMP(ret, pm_flags);
3732 & (SVs_TEMP | SVs_PADTMP | SVf_READONLY
3734 sv_magic(ret,(SV*)ReREFCNT_inc(re),
3739 RX_MATCH_COPIED_off(re);
3740 re->subbeg = rex->subbeg;
3741 re->sublen = rex->sublen;
3744 debug_start_match(re, do_utf8, locinput, PL_regeol,
3745 "Matching embedded");
3747 startpoint = rei->program + 1;
3748 ST.close_paren = 0; /* only used for GOSUB */
3749 /* borrowed from regtry */
3750 if (PL_reg_start_tmpl <= re->nparens) {
3751 PL_reg_start_tmpl = re->nparens*3/2 + 3;
3752 if(PL_reg_start_tmp)
3753 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
3755 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
3758 eval_recurse_doit: /* Share code with GOSUB below this line */
3759 /* run the pattern returned from (??{...}) */
3760 ST.cp = regcppush(0); /* Save *all* the positions. */
3761 REGCP_SET(ST.lastcp);
3763 PL_regoffs = re->offs; /* essentially NOOP on GOSUB */
3765 *PL_reglastparen = 0;
3766 *PL_reglastcloseparen = 0;
3767 PL_reginput = locinput;
3770 /* XXXX This is too dramatic a measure... */
3773 ST.toggle_reg_flags = PL_reg_flags;
3774 if (re->extflags & RXf_UTF8)
3775 PL_reg_flags |= RF_utf8;
3777 PL_reg_flags &= ~RF_utf8;
3778 ST.toggle_reg_flags ^= PL_reg_flags; /* diff of old and new */
3781 ST.prev_curlyx = cur_curlyx;
3786 ST.prev_eval = cur_eval;
3788 /* now continue from first node in postoned RE */
3789 PUSH_YES_STATE_GOTO(EVAL_AB, startpoint);
3792 /* logical is 1, /(?(?{...})X|Y)/ */
3793 sw = (bool)SvTRUE(ret);
3798 case EVAL_AB: /* cleanup after a successful (??{A})B */
3799 /* note: this is called twice; first after popping B, then A */
3800 PL_reg_flags ^= ST.toggle_reg_flags;
3802 SETREX(rex,ST.prev_rex);
3803 rexi = RXi_GET(rex);
3805 cur_eval = ST.prev_eval;
3806 cur_curlyx = ST.prev_curlyx;
3807 /* XXXX This is too dramatic a measure... */
3809 if ( nochange_depth )
3814 case EVAL_AB_fail: /* unsuccessfully ran A or B in (??{A})B */
3815 /* note: this is called twice; first after popping B, then A */
3816 PL_reg_flags ^= ST.toggle_reg_flags;
3818 SETREX(rex,ST.prev_rex);
3819 rexi = RXi_GET(rex);
3820 PL_reginput = locinput;
3821 REGCP_UNWIND(ST.lastcp);
3823 cur_eval = ST.prev_eval;
3824 cur_curlyx = ST.prev_curlyx;
3825 /* XXXX This is too dramatic a measure... */
3827 if ( nochange_depth )
3833 n = ARG(scan); /* which paren pair */
3834 PL_reg_start_tmp[n] = locinput;
3840 n = ARG(scan); /* which paren pair */
3841 PL_regoffs[n].start = PL_reg_start_tmp[n] - PL_bostr;
3842 PL_regoffs[n].end = locinput - PL_bostr;
3843 /*if (n > PL_regsize)
3845 if (n > *PL_reglastparen)
3846 *PL_reglastparen = n;
3847 *PL_reglastcloseparen = n;
3848 if (cur_eval && cur_eval->u.eval.close_paren == n) {
3856 cursor && OP(cursor)!=END;
3857 cursor=regnext(cursor))
3859 if ( OP(cursor)==CLOSE ){
3861 if ( n <= lastopen ) {
3863 = PL_reg_start_tmp[n] - PL_bostr;
3864 PL_regoffs[n].end = locinput - PL_bostr;
3865 /*if (n > PL_regsize)
3867 if (n > *PL_reglastparen)
3868 *PL_reglastparen = n;
3869 *PL_reglastcloseparen = n;
3870 if ( n == ARG(scan) || (cur_eval &&
3871 cur_eval->u.eval.close_paren == n))
3880 n = ARG(scan); /* which paren pair */
3881 sw = (bool)(*PL_reglastparen >= n && PL_regoffs[n].end != -1);
3884 /* reg_check_named_buff_matched returns 0 for no match */
3885 sw = (bool)(0 < reg_check_named_buff_matched(rex,scan));
3889 sw = (cur_eval && (!n || cur_eval->u.eval.close_paren == n));
3895 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
3897 next = NEXTOPER(NEXTOPER(scan));
3899 next = scan + ARG(scan);
3900 if (OP(next) == IFTHEN) /* Fake one. */
3901 next = NEXTOPER(NEXTOPER(next));
3905 logical = scan->flags;
3908 /*******************************************************************
3910 The CURLYX/WHILEM pair of ops handle the most generic case of the /A*B/
3911 pattern, where A and B are subpatterns. (For simple A, CURLYM or
3912 STAR/PLUS/CURLY/CURLYN are used instead.)
3914 A*B is compiled as <CURLYX><A><WHILEM><B>
3916 On entry to the subpattern, CURLYX is called. This pushes a CURLYX
3917 state, which contains the current count, initialised to -1. It also sets
3918 cur_curlyx to point to this state, with any previous value saved in the
3921 CURLYX then jumps straight to the WHILEM op, rather than executing A,
3922 since the pattern may possibly match zero times (i.e. it's a while {} loop
3923 rather than a do {} while loop).
3925 Each entry to WHILEM represents a successful match of A. The count in the
3926 CURLYX block is incremented, another WHILEM state is pushed, and execution
3927 passes to A or B depending on greediness and the current count.
3929 For example, if matching against the string a1a2a3b (where the aN are
3930 substrings that match /A/), then the match progresses as follows: (the
3931 pushed states are interspersed with the bits of strings matched so far):
3934 <CURLYX cnt=0><WHILEM>
3935 <CURLYX cnt=1><WHILEM> a1 <WHILEM>
3936 <CURLYX cnt=2><WHILEM> a1 <WHILEM> a2 <WHILEM>
3937 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM>
3938 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM> b
3940 (Contrast this with something like CURLYM, which maintains only a single
3944 a1 <CURLYM cnt=1> a2
3945 a1 a2 <CURLYM cnt=2> a3
3946 a1 a2 a3 <CURLYM cnt=3> b
3949 Each WHILEM state block marks a point to backtrack to upon partial failure
3950 of A or B, and also contains some minor state data related to that
3951 iteration. The CURLYX block, pointed to by cur_curlyx, contains the
3952 overall state, such as the count, and pointers to the A and B ops.
3954 This is complicated slightly by nested CURLYX/WHILEM's. Since cur_curlyx
3955 must always point to the *current* CURLYX block, the rules are:
3957 When executing CURLYX, save the old cur_curlyx in the CURLYX state block,
3958 and set cur_curlyx to point the new block.
3960 When popping the CURLYX block after a successful or unsuccessful match,
3961 restore the previous cur_curlyx.
3963 When WHILEM is about to execute B, save the current cur_curlyx, and set it
3964 to the outer one saved in the CURLYX block.
3966 When popping the WHILEM block after a successful or unsuccessful B match,
3967 restore the previous cur_curlyx.
3969 Here's an example for the pattern (AI* BI)*BO
3970 I and O refer to inner and outer, C and W refer to CURLYX and WHILEM:
3973 curlyx backtrack stack
3974 ------ ---------------
3976 CO <CO prev=NULL> <WO>
3977 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
3978 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
3979 NULL <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi <WO prev=CO> bo
3981 At this point the pattern succeeds, and we work back down the stack to
3982 clean up, restoring as we go:
3984 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
3985 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
3986 CO <CO prev=NULL> <WO>
3989 *******************************************************************/
3991 #define ST st->u.curlyx
3993 case CURLYX: /* start of /A*B/ (for complex A) */
3995 /* No need to save/restore up to this paren */
3996 I32 parenfloor = scan->flags;
3998 assert(next); /* keep Coverity happy */
3999 if (OP(PREVOPER(next)) == NOTHING) /* LONGJMP */
4002 /* XXXX Probably it is better to teach regpush to support
4003 parenfloor > PL_regsize... */
4004 if (parenfloor > (I32)*PL_reglastparen)
4005 parenfloor = *PL_reglastparen; /* Pessimization... */
4007 ST.prev_curlyx= cur_curlyx;
4009 ST.cp = PL_savestack_ix;
4011 /* these fields contain the state of the current curly.
4012 * they are accessed by subsequent WHILEMs */
4013 ST.parenfloor = parenfloor;
4014 ST.min = ARG1(scan);
4015 ST.max = ARG2(scan);
4016 ST.A = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
4020 ST.count = -1; /* this will be updated by WHILEM */
4021 ST.lastloc = NULL; /* this will be updated by WHILEM */
4023 PL_reginput = locinput;
4024 PUSH_YES_STATE_GOTO(CURLYX_end, PREVOPER(next));
4028 case CURLYX_end: /* just finished matching all of A*B */
4029 cur_curlyx = ST.prev_curlyx;
4033 case CURLYX_end_fail: /* just failed to match all of A*B */
4035 cur_curlyx = ST.prev_curlyx;
4041 #define ST st->u.whilem
4043 case WHILEM: /* just matched an A in /A*B/ (for complex A) */
4045 /* see the discussion above about CURLYX/WHILEM */
4047 assert(cur_curlyx); /* keep Coverity happy */
4048 n = ++cur_curlyx->u.curlyx.count; /* how many A's matched */
4049 ST.save_lastloc = cur_curlyx->u.curlyx.lastloc;
4050 ST.cache_offset = 0;
4053 PL_reginput = locinput;
4055 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4056 "%*s whilem: matched %ld out of %ld..%ld\n",
4057 REPORT_CODE_OFF+depth*2, "", (long)n,
4058 (long)cur_curlyx->u.curlyx.min,
4059 (long)cur_curlyx->u.curlyx.max)
4062 /* First just match a string of min A's. */
4064 if (n < cur_curlyx->u.curlyx.min) {
4065 cur_curlyx->u.curlyx.lastloc = locinput;
4066 PUSH_STATE_GOTO(WHILEM_A_pre, cur_curlyx->u.curlyx.A);
4070 /* If degenerate A matches "", assume A done. */
4072 if (locinput == cur_curlyx->u.curlyx.lastloc) {
4073 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4074 "%*s whilem: empty match detected, trying continuation...\n",
4075 REPORT_CODE_OFF+depth*2, "")
4077 goto do_whilem_B_max;
4080 /* super-linear cache processing */
4084 if (!PL_reg_maxiter) {
4085 /* start the countdown: Postpone detection until we
4086 * know the match is not *that* much linear. */
4087 PL_reg_maxiter = (PL_regeol - PL_bostr + 1) * (scan->flags>>4);
4088 /* possible overflow for long strings and many CURLYX's */
4089 if (PL_reg_maxiter < 0)
4090 PL_reg_maxiter = I32_MAX;
4091 PL_reg_leftiter = PL_reg_maxiter;
4094 if (PL_reg_leftiter-- == 0) {
4095 /* initialise cache */
4096 const I32 size = (PL_reg_maxiter + 7)/8;
4097 if (PL_reg_poscache) {
4098 if ((I32)PL_reg_poscache_size < size) {
4099 Renew(PL_reg_poscache, size, char);
4100 PL_reg_poscache_size = size;
4102 Zero(PL_reg_poscache, size, char);
4105 PL_reg_poscache_size = size;
4106 Newxz(PL_reg_poscache, size, char);
4108 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4109 "%swhilem: Detected a super-linear match, switching on caching%s...\n",
4110 PL_colors[4], PL_colors[5])
4114 if (PL_reg_leftiter < 0) {
4115 /* have we already failed at this position? */
4117 offset = (scan->flags & 0xf) - 1
4118 + (locinput - PL_bostr) * (scan->flags>>4);
4119 mask = 1 << (offset % 8);
4121 if (PL_reg_poscache[offset] & mask) {
4122 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4123 "%*s whilem: (cache) already tried at this position...\n",
4124 REPORT_CODE_OFF+depth*2, "")
4126 sayNO; /* cache records failure */
4128 ST.cache_offset = offset;
4129 ST.cache_mask = mask;
4133 /* Prefer B over A for minimal matching. */
4135 if (cur_curlyx->u.curlyx.minmod) {
4136 ST.save_curlyx = cur_curlyx;
4137 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4138 ST.cp = regcppush(ST.save_curlyx->u.curlyx.parenfloor);
4139 REGCP_SET(ST.lastcp);
4140 PUSH_YES_STATE_GOTO(WHILEM_B_min, ST.save_curlyx->u.curlyx.B);
4144 /* Prefer A over B for maximal matching. */
4146 if (n < cur_curlyx->u.curlyx.max) { /* More greed allowed? */
4147 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4148 cur_curlyx->u.curlyx.lastloc = locinput;
4149 REGCP_SET(ST.lastcp);
4150 PUSH_STATE_GOTO(WHILEM_A_max, cur_curlyx->u.curlyx.A);
4153 goto do_whilem_B_max;
4157 case WHILEM_B_min: /* just matched B in a minimal match */
4158 case WHILEM_B_max: /* just matched B in a maximal match */
4159 cur_curlyx = ST.save_curlyx;
4163 case WHILEM_B_max_fail: /* just failed to match B in a maximal match */
4164 cur_curlyx = ST.save_curlyx;
4165 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4166 cur_curlyx->u.curlyx.count--;
4170 case WHILEM_A_min_fail: /* just failed to match A in a minimal match */
4171 REGCP_UNWIND(ST.lastcp);
4174 case WHILEM_A_pre_fail: /* just failed to match even minimal A */
4175 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4176 cur_curlyx->u.curlyx.count--;
4180 case WHILEM_A_max_fail: /* just failed to match A in a maximal match */
4181 REGCP_UNWIND(ST.lastcp);
4182 regcppop(rex); /* Restore some previous $<digit>s? */
4183 PL_reginput = locinput;
4184 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4185 "%*s whilem: failed, trying continuation...\n",
4186 REPORT_CODE_OFF+depth*2, "")
4189 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4190 && ckWARN(WARN_REGEXP)
4191 && !(PL_reg_flags & RF_warned))
4193 PL_reg_flags |= RF_warned;
4194 Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s limit (%d) exceeded",
4195 "Complex regular subexpression recursion",
4200 ST.save_curlyx = cur_curlyx;
4201 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4202 PUSH_YES_STATE_GOTO(WHILEM_B_max, ST.save_curlyx->u.curlyx.B);
4205 case WHILEM_B_min_fail: /* just failed to match B in a minimal match */
4206 cur_curlyx = ST.save_curlyx;
4207 REGCP_UNWIND(ST.lastcp);
4210 if (cur_curlyx->u.curlyx.count >= cur_curlyx->u.curlyx.max) {
4211 /* Maximum greed exceeded */
4212 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4213 && ckWARN(WARN_REGEXP)
4214 && !(PL_reg_flags & RF_warned))
4216 PL_reg_flags |= RF_warned;
4217 Perl_warner(aTHX_ packWARN(WARN_REGEXP),
4218 "%s limit (%d) exceeded",
4219 "Complex regular subexpression recursion",
4222 cur_curlyx->u.curlyx.count--;
4226 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4227 "%*s trying longer...\n", REPORT_CODE_OFF+depth*2, "")
4229 /* Try grabbing another A and see if it helps. */
4230 PL_reginput = locinput;
4231 cur_curlyx->u.curlyx.lastloc = locinput;
4232 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4233 REGCP_SET(ST.lastcp);
4234 PUSH_STATE_GOTO(WHILEM_A_min, ST.save_curlyx->u.curlyx.A);
4238 #define ST st->u.branch
4240 case BRANCHJ: /* /(...|A|...)/ with long next pointer */
4241 next = scan + ARG(scan);
4244 scan = NEXTOPER(scan);
4247 case BRANCH: /* /(...|A|...)/ */
4248 scan = NEXTOPER(scan); /* scan now points to inner node */
4249 ST.lastparen = *PL_reglastparen;
4250 ST.next_branch = next;
4252 PL_reginput = locinput;
4254 /* Now go into the branch */
4256 PUSH_YES_STATE_GOTO(BRANCH_next, scan);
4258 PUSH_STATE_GOTO(BRANCH_next, scan);
4262 PL_reginput = locinput;
4263 sv_yes_mark = st->u.mark.mark_name = scan->flags ? NULL :
4264 (SV*)rexi->data->data[ ARG( scan ) ];
4265 PUSH_STATE_GOTO(CUTGROUP_next,next);
4267 case CUTGROUP_next_fail:
4270 if (st->u.mark.mark_name)
4271 sv_commit = st->u.mark.mark_name;
4277 case BRANCH_next_fail: /* that branch failed; try the next, if any */
4282 REGCP_UNWIND(ST.cp);
4283 for (n = *PL_reglastparen; n > ST.lastparen; n--)
4284 PL_regoffs[n].end = -1;
4285 *PL_reglastparen = n;
4286 /*dmq: *PL_reglastcloseparen = n; */
4287 scan = ST.next_branch;
4288 /* no more branches? */
4289 if (!scan || (OP(scan) != BRANCH && OP(scan) != BRANCHJ)) {
4291 PerlIO_printf( Perl_debug_log,
4292 "%*s %sBRANCH failed...%s\n",
4293 REPORT_CODE_OFF+depth*2, "",
4299 continue; /* execute next BRANCH[J] op */
4307 #define ST st->u.curlym
4309 case CURLYM: /* /A{m,n}B/ where A is fixed-length */
4311 /* This is an optimisation of CURLYX that enables us to push
4312 * only a single backtracking state, no matter now many matches
4313 * there are in {m,n}. It relies on the pattern being constant
4314 * length, with no parens to influence future backrefs
4318 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
4320 /* if paren positive, emulate an OPEN/CLOSE around A */
4322 U32 paren = ST.me->flags;
4323 if (paren > PL_regsize)
4325 if (paren > *PL_reglastparen)
4326 *PL_reglastparen = paren;
4327 scan += NEXT_OFF(scan); /* Skip former OPEN. */
4335 ST.c1 = CHRTEST_UNINIT;
4338 if (!(ST.minmod ? ARG1(ST.me) : ARG2(ST.me))) /* min/max */
4341 curlym_do_A: /* execute the A in /A{m,n}B/ */
4342 PL_reginput = locinput;
4343 PUSH_YES_STATE_GOTO(CURLYM_A, ST.A); /* match A */
4346 case CURLYM_A: /* we've just matched an A */
4347 locinput = st->locinput;
4348 nextchr = UCHARAT(locinput);
4351 /* after first match, determine A's length: u.curlym.alen */
4352 if (ST.count == 1) {
4353 if (PL_reg_match_utf8) {
4355 while (s < PL_reginput) {
4361 ST.alen = PL_reginput - locinput;
4364 ST.count = ST.minmod ? ARG1(ST.me) : ARG2(ST.me);
4367 PerlIO_printf(Perl_debug_log,
4368 "%*s CURLYM now matched %"IVdf" times, len=%"IVdf"...\n",
4369 (int)(REPORT_CODE_OFF+(depth*2)), "",
4370 (IV) ST.count, (IV)ST.alen)
4373 locinput = PL_reginput;
4375 if (cur_eval && cur_eval->u.eval.close_paren &&
4376 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
4379 if ( ST.count < (ST.minmod ? ARG1(ST.me) : ARG2(ST.me)) )
4380 goto curlym_do_A; /* try to match another A */
4381 goto curlym_do_B; /* try to match B */
4383 case CURLYM_A_fail: /* just failed to match an A */
4384 REGCP_UNWIND(ST.cp);
4386 if (ST.minmod || ST.count < ARG1(ST.me) /* min*/
4387 || (cur_eval && cur_eval->u.eval.close_paren &&
4388 cur_eval->u.eval.close_paren == (U32)ST.me->flags))
4391 curlym_do_B: /* execute the B in /A{m,n}B/ */
4392 PL_reginput = locinput;
4393 if (ST.c1 == CHRTEST_UNINIT) {
4394 /* calculate c1 and c2 for possible match of 1st char
4395 * following curly */
4396 ST.c1 = ST.c2 = CHRTEST_VOID;
4397 if (HAS_TEXT(ST.B) || JUMPABLE(ST.B)) {
4398 regnode *text_node = ST.B;
4399 if (! HAS_TEXT(text_node))
4400 FIND_NEXT_IMPT(text_node);
4403 (HAS_TEXT(text_node) && PL_regkind[OP(text_node)] == EXACT)
4405 But the former is redundant in light of the latter.
4407 if this changes back then the macro for
4408 IS_TEXT and friends need to change.
4410 if (PL_regkind[OP(text_node)] == EXACT)
4413 ST.c1 = (U8)*STRING(text_node);
4415 (IS_TEXTF(text_node))
4417 : (IS_TEXTFL(text_node))
4418 ? PL_fold_locale[ST.c1]
4425 PerlIO_printf(Perl_debug_log,
4426 "%*s CURLYM trying tail with matches=%"IVdf"...\n",
4427 (int)(REPORT_CODE_OFF+(depth*2)),
4430 if (ST.c1 != CHRTEST_VOID
4431 && UCHARAT(PL_reginput) != ST.c1
4432 && UCHARAT(PL_reginput) != ST.c2)
4434 /* simulate B failing */
4436 PerlIO_printf(Perl_debug_log,
4437 "%*s CURLYM Fast bail c1=%"IVdf" c2=%"IVdf"\n",
4438 (int)(REPORT_CODE_OFF+(depth*2)),"",
4441 state_num = CURLYM_B_fail;
4442 goto reenter_switch;
4446 /* mark current A as captured */
4447 I32 paren = ST.me->flags;
4449 PL_regoffs[paren].start
4450 = HOPc(PL_reginput, -ST.alen) - PL_bostr;
4451 PL_regoffs[paren].end = PL_reginput - PL_bostr;
4452 /*dmq: *PL_reglastcloseparen = paren; */
4455 PL_regoffs[paren].end = -1;
4456 if (cur_eval && cur_eval->u.eval.close_paren &&
4457 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
4466 PUSH_STATE_GOTO(CURLYM_B, ST.B); /* match B */
4469 case CURLYM_B_fail: /* just failed to match a B */
4470 REGCP_UNWIND(ST.cp);
4472 if (ST.count == ARG2(ST.me) /* max */)
4474 goto curlym_do_A; /* try to match a further A */
4476 /* backtrack one A */
4477 if (ST.count == ARG1(ST.me) /* min */)
4480 locinput = HOPc(locinput, -ST.alen);
4481 goto curlym_do_B; /* try to match B */
4484 #define ST st->u.curly
4486 #define CURLY_SETPAREN(paren, success) \
4489 PL_regoffs[paren].start = HOPc(locinput, -1) - PL_bostr; \
4490 PL_regoffs[paren].end = locinput - PL_bostr; \
4491 *PL_reglastcloseparen = paren; \
4494 PL_regoffs[paren].end = -1; \
4497 case STAR: /* /A*B/ where A is width 1 */
4501 scan = NEXTOPER(scan);
4503 case PLUS: /* /A+B/ where A is width 1 */
4507 scan = NEXTOPER(scan);
4509 case CURLYN: /* /(A){m,n}B/ where A is width 1 */
4510 ST.paren = scan->flags; /* Which paren to set */
4511 if (ST.paren > PL_regsize)
4512 PL_regsize = ST.paren;
4513 if (ST.paren > *PL_reglastparen)
4514 *PL_reglastparen = ST.paren;
4515 ST.min = ARG1(scan); /* min to match */
4516 ST.max = ARG2(scan); /* max to match */
4517 if (cur_eval && cur_eval->u.eval.close_paren &&
4518 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4522 scan = regnext(NEXTOPER(scan) + NODE_STEP_REGNODE);
4524 case CURLY: /* /A{m,n}B/ where A is width 1 */
4526 ST.min = ARG1(scan); /* min to match */
4527 ST.max = ARG2(scan); /* max to match */
4528 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
4531 * Lookahead to avoid useless match attempts
4532 * when we know what character comes next.
4534 * Used to only do .*x and .*?x, but now it allows
4535 * for )'s, ('s and (?{ ... })'s to be in the way
4536 * of the quantifier and the EXACT-like node. -- japhy
4539 if (ST.min > ST.max) /* XXX make this a compile-time check? */
4541 if (HAS_TEXT(next) || JUMPABLE(next)) {
4543 regnode *text_node = next;
4545 if (! HAS_TEXT(text_node))
4546 FIND_NEXT_IMPT(text_node);
4548 if (! HAS_TEXT(text_node))
4549 ST.c1 = ST.c2 = CHRTEST_VOID;
4551 if ( PL_regkind[OP(text_node)] != EXACT ) {
4552 ST.c1 = ST.c2 = CHRTEST_VOID;
4553 goto assume_ok_easy;
4556 s = (U8*)STRING(text_node);
4558 /* Currently we only get here when
4560 PL_rekind[OP(text_node)] == EXACT
4562 if this changes back then the macro for IS_TEXT and
4563 friends need to change. */
4566 if (IS_TEXTF(text_node))
4567 ST.c2 = PL_fold[ST.c1];
4568 else if (IS_TEXTFL(text_node))
4569 ST.c2 = PL_fold_locale[ST.c1];
4572 if (IS_TEXTF(text_node)) {
4573 STRLEN ulen1, ulen2;
4574 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
4575 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
4577 to_utf8_lower((U8*)s, tmpbuf1, &ulen1);
4578 to_utf8_upper((U8*)s, tmpbuf2, &ulen2);
4580 ST.c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXLEN, 0,
4582 0 : UTF8_ALLOW_ANY);
4583 ST.c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXLEN, 0,
4585 0 : UTF8_ALLOW_ANY);
4587 ST.c1 = utf8n_to_uvuni(tmpbuf1, UTF8_MAXBYTES, 0,
4589 ST.c2 = utf8n_to_uvuni(tmpbuf2, UTF8_MAXBYTES, 0,
4594 ST.c2 = ST.c1 = utf8n_to_uvchr(s, UTF8_MAXBYTES, 0,
4601 ST.c1 = ST.c2 = CHRTEST_VOID;
4606 PL_reginput = locinput;
4609 if (ST.min && regrepeat(rex, ST.A, ST.min, depth) < ST.min)
4612 locinput = PL_reginput;
4614 if (ST.c1 == CHRTEST_VOID)
4615 goto curly_try_B_min;
4617 ST.oldloc = locinput;
4619 /* set ST.maxpos to the furthest point along the
4620 * string that could possibly match */
4621 if (ST.max == REG_INFTY) {
4622 ST.maxpos = PL_regeol - 1;
4624 while (UTF8_IS_CONTINUATION(*(U8*)ST.maxpos))
4628 int m = ST.max - ST.min;
4629 for (ST.maxpos = locinput;
4630 m >0 && ST.maxpos + UTF8SKIP(ST.maxpos) <= PL_regeol; m--)
4631 ST.maxpos += UTF8SKIP(ST.maxpos);
4634 ST.maxpos = locinput + ST.max - ST.min;
4635 if (ST.maxpos >= PL_regeol)
4636 ST.maxpos = PL_regeol - 1;
4638 goto curly_try_B_min_known;
4642 ST.count = regrepeat(rex, ST.A, ST.max, depth);
4643 locinput = PL_reginput;
4644 if (ST.count < ST.min)
4646 if ((ST.count > ST.min)
4647 && (PL_regkind[OP(ST.B)] == EOL) && (OP(ST.B) != MEOL))
4649 /* A{m,n} must come at the end of the string, there's
4650 * no point in backing off ... */
4652 /* ...except that $ and \Z can match before *and* after
4653 newline at the end. Consider "\n\n" =~ /\n+\Z\n/.
4654 We may back off by one in this case. */
4655 if (UCHARAT(PL_reginput - 1) == '\n' && OP(ST.B) != EOS)
4659 goto curly_try_B_max;
4664 case CURLY_B_min_known_fail:
4665 /* failed to find B in a non-greedy match where c1,c2 valid */
4666 if (ST.paren && ST.count)
4667 PL_regoffs[ST.paren].end = -1;
4669 PL_reginput = locinput; /* Could be reset... */
4670 REGCP_UNWIND(ST.cp);
4671 /* Couldn't or didn't -- move forward. */
4672 ST.oldloc = locinput;
4674 locinput += UTF8SKIP(locinput);
4678 curly_try_B_min_known:
4679 /* find the next place where 'B' could work, then call B */
4683 n = (ST.oldloc == locinput) ? 0 : 1;
4684 if (ST.c1 == ST.c2) {
4686 /* set n to utf8_distance(oldloc, locinput) */
4687 while (locinput <= ST.maxpos &&
4688 utf8n_to_uvchr((U8*)locinput,
4689 UTF8_MAXBYTES, &len,
4690 uniflags) != (UV)ST.c1) {
4696 /* set n to utf8_distance(oldloc, locinput) */
4697 while (locinput <= ST.maxpos) {
4699 const UV c = utf8n_to_uvchr((U8*)locinput,
4700 UTF8_MAXBYTES, &len,
4702 if (c == (UV)ST.c1 || c == (UV)ST.c2)
4710 if (ST.c1 == ST.c2) {
4711 while (locinput <= ST.maxpos &&
4712 UCHARAT(locinput) != ST.c1)
4716 while (locinput <= ST.maxpos
4717 && UCHARAT(locinput) != ST.c1
4718 && UCHARAT(locinput) != ST.c2)
4721 n = locinput - ST.oldloc;
4723 if (locinput > ST.maxpos)
4725 /* PL_reginput == oldloc now */
4728 if (regrepeat(rex, ST.A, n, depth) < n)
4731 PL_reginput = locinput;
4732 CURLY_SETPAREN(ST.paren, ST.count);
4733 if (cur_eval && cur_eval->u.eval.close_paren &&
4734 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4737 PUSH_STATE_GOTO(CURLY_B_min_known, ST.B);
4742 case CURLY_B_min_fail:
4743 /* failed to find B in a non-greedy match where c1,c2 invalid */
4744 if (ST.paren && ST.count)
4745 PL_regoffs[ST.paren].end = -1;
4747 REGCP_UNWIND(ST.cp);
4748 /* failed -- move forward one */
4749 PL_reginput = locinput;
4750 if (regrepeat(rex, ST.A, 1, depth)) {
4752 locinput = PL_reginput;
4753 if (ST.count <= ST.max || (ST.max == REG_INFTY &&
4754 ST.count > 0)) /* count overflow ? */
4757 CURLY_SETPAREN(ST.paren, ST.count);
4758 if (cur_eval && cur_eval->u.eval.close_paren &&
4759 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4762 PUSH_STATE_GOTO(CURLY_B_min, ST.B);
4770 /* a successful greedy match: now try to match B */
4771 if (cur_eval && cur_eval->u.eval.close_paren &&
4772 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4777 if (ST.c1 != CHRTEST_VOID)
4778 c = do_utf8 ? utf8n_to_uvchr((U8*)PL_reginput,
4779 UTF8_MAXBYTES, 0, uniflags)
4780 : (UV) UCHARAT(PL_reginput);
4781 /* If it could work, try it. */
4782 if (ST.c1 == CHRTEST_VOID || c == (UV)ST.c1 || c == (UV)ST.c2) {
4783 CURLY_SETPAREN(ST.paren, ST.count);
4784 PUSH_STATE_GOTO(CURLY_B_max, ST.B);
4789 case CURLY_B_max_fail:
4790 /* failed to find B in a greedy match */
4791 if (ST.paren && ST.count)
4792 PL_regoffs[ST.paren].end = -1;
4794 REGCP_UNWIND(ST.cp);
4796 if (--ST.count < ST.min)
4798 PL_reginput = locinput = HOPc(locinput, -1);
4799 goto curly_try_B_max;
4806 /* we've just finished A in /(??{A})B/; now continue with B */
4808 st->u.eval.toggle_reg_flags
4809 = cur_eval->u.eval.toggle_reg_flags;
4810 PL_reg_flags ^= st->u.eval.toggle_reg_flags;
4812 st->u.eval.prev_rex = rex; /* inner */
4813 SETREX(rex,cur_eval->u.eval.prev_rex);
4814 rexi = RXi_GET(rex);
4815 cur_curlyx = cur_eval->u.eval.prev_curlyx;
4817 st->u.eval.cp = regcppush(0); /* Save *all* the positions. */
4818 REGCP_SET(st->u.eval.lastcp);
4819 PL_reginput = locinput;
4821 /* Restore parens of the outer rex without popping the
4823 tmpix = PL_savestack_ix;
4824 PL_savestack_ix = cur_eval->u.eval.lastcp;
4826 PL_savestack_ix = tmpix;
4828 st->u.eval.prev_eval = cur_eval;
4829 cur_eval = cur_eval->u.eval.prev_eval;
4831 PerlIO_printf(Perl_debug_log, "%*s EVAL trying tail ... %"UVxf"\n",
4832 REPORT_CODE_OFF+depth*2, "",PTR2UV(cur_eval)););
4833 if ( nochange_depth )
4836 PUSH_YES_STATE_GOTO(EVAL_AB,
4837 st->u.eval.prev_eval->u.eval.B); /* match B */
4840 if (locinput < reginfo->till) {
4841 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4842 "%sMatch possible, but length=%ld is smaller than requested=%ld, failing!%s\n",
4844 (long)(locinput - PL_reg_starttry),
4845 (long)(reginfo->till - PL_reg_starttry),
4848 sayNO_SILENT; /* Cannot match: too short. */
4850 PL_reginput = locinput; /* put where regtry can find it */
4851 sayYES; /* Success! */
4853 case SUCCEED: /* successful SUSPEND/UNLESSM/IFMATCH/CURLYM */
4855 PerlIO_printf(Perl_debug_log,
4856 "%*s %ssubpattern success...%s\n",
4857 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5]));
4858 PL_reginput = locinput; /* put where regtry can find it */
4859 sayYES; /* Success! */
4862 #define ST st->u.ifmatch
4864 case SUSPEND: /* (?>A) */
4866 PL_reginput = locinput;
4869 case UNLESSM: /* -ve lookaround: (?!A), or with flags, (?<!A) */
4871 goto ifmatch_trivial_fail_test;
4873 case IFMATCH: /* +ve lookaround: (?=A), or with flags, (?<=A) */
4875 ifmatch_trivial_fail_test:
4877 char * const s = HOPBACKc(locinput, scan->flags);
4882 sw = 1 - (bool)ST.wanted;
4886 next = scan + ARG(scan);
4894 PL_reginput = locinput;
4898 ST.logical = logical;
4899 /* execute body of (?...A) */
4900 PUSH_YES_STATE_GOTO(IFMATCH_A, NEXTOPER(NEXTOPER(scan)));
4903 case IFMATCH_A_fail: /* body of (?...A) failed */
4904 ST.wanted = !ST.wanted;
4907 case IFMATCH_A: /* body of (?...A) succeeded */
4909 sw = (bool)ST.wanted;
4911 else if (!ST.wanted)
4914 if (OP(ST.me) == SUSPEND)
4915 locinput = PL_reginput;
4917 locinput = PL_reginput = st->locinput;
4918 nextchr = UCHARAT(locinput);
4920 scan = ST.me + ARG(ST.me);
4923 continue; /* execute B */
4928 next = scan + ARG(scan);
4933 reginfo->cutpoint = PL_regeol;
4936 PL_reginput = locinput;
4938 sv_yes_mark = sv_commit = (SV*)rexi->data->data[ ARG( scan ) ];
4939 PUSH_STATE_GOTO(COMMIT_next,next);
4941 case COMMIT_next_fail:
4948 #define ST st->u.mark
4950 ST.prev_mark = mark_state;
4951 ST.mark_name = sv_commit = sv_yes_mark
4952 = (SV*)rexi->data->data[ ARG( scan ) ];
4954 ST.mark_loc = PL_reginput = locinput;
4955 PUSH_YES_STATE_GOTO(MARKPOINT_next,next);
4957 case MARKPOINT_next:
4958 mark_state = ST.prev_mark;
4961 case MARKPOINT_next_fail:
4962 if (popmark && sv_eq(ST.mark_name,popmark))
4964 if (ST.mark_loc > startpoint)
4965 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
4966 popmark = NULL; /* we found our mark */
4967 sv_commit = ST.mark_name;
4970 PerlIO_printf(Perl_debug_log,
4971 "%*s %ssetting cutpoint to mark:%"SVf"...%s\n",
4972 REPORT_CODE_OFF+depth*2, "",
4973 PL_colors[4], SVfARG(sv_commit), PL_colors[5]);
4976 mark_state = ST.prev_mark;
4977 sv_yes_mark = mark_state ?
4978 mark_state->u.mark.mark_name : NULL;
4982 PL_reginput = locinput;
4984 /* (*SKIP) : if we fail we cut here*/
4985 ST.mark_name = NULL;
4986 ST.mark_loc = locinput;
4987 PUSH_STATE_GOTO(SKIP_next,next);
4989 /* (*SKIP:NAME) : if there is a (*MARK:NAME) fail where it was,
4990 otherwise do nothing. Meaning we need to scan
4992 regmatch_state *cur = mark_state;
4993 SV *find = (SV*)rexi->data->data[ ARG( scan ) ];
4996 if ( sv_eq( cur->u.mark.mark_name,
4999 ST.mark_name = find;
5000 PUSH_STATE_GOTO( SKIP_next, next );
5002 cur = cur->u.mark.prev_mark;
5005 /* Didn't find our (*MARK:NAME) so ignore this (*SKIP:NAME) */
5007 case SKIP_next_fail:
5009 /* (*CUT:NAME) - Set up to search for the name as we
5010 collapse the stack*/
5011 popmark = ST.mark_name;
5013 /* (*CUT) - No name, we cut here.*/
5014 if (ST.mark_loc > startpoint)
5015 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
5016 /* but we set sv_commit to latest mark_name if there
5017 is one so they can test to see how things lead to this
5020 sv_commit=mark_state->u.mark.mark_name;
5028 if ( n == (U32)what_len_TRICKYFOLD(locinput,do_utf8,ln) ) {
5030 } else if ( 0xDF == n && !do_utf8 && !UTF ) {
5033 U8 folded[UTF8_MAXBYTES_CASE+1];
5035 const char * const l = locinput;
5036 char *e = PL_regeol;
5037 to_uni_fold(n, folded, &foldlen);
5039 if (ibcmp_utf8((const char*) folded, 0, foldlen, 1,
5040 l, &e, 0, do_utf8)) {
5045 nextchr = UCHARAT(locinput);
5048 if ((n=is_LNBREAK(locinput,do_utf8))) {
5050 nextchr = UCHARAT(locinput);
5055 #define CASE_CLASS(nAmE) \
5057 if ((n=is_##nAmE(locinput,do_utf8))) { \
5059 nextchr = UCHARAT(locinput); \
5064 if ((n=is_##nAmE(locinput,do_utf8))) { \
5067 locinput += UTF8SKIP(locinput); \
5068 nextchr = UCHARAT(locinput); \
5073 CASE_CLASS(HORIZWS);
5077 PerlIO_printf(Perl_error_log, "%"UVxf" %d\n",
5078 PTR2UV(scan), OP(scan));
5079 Perl_croak(aTHX_ "regexp memory corruption");
5083 /* switch break jumps here */
5084 scan = next; /* prepare to execute the next op and ... */
5085 continue; /* ... jump back to the top, reusing st */
5089 /* push a state that backtracks on success */
5090 st->u.yes.prev_yes_state = yes_state;
5094 /* push a new regex state, then continue at scan */
5096 regmatch_state *newst;
5099 regmatch_state *cur = st;
5100 regmatch_state *curyes = yes_state;
5102 regmatch_slab *slab = PL_regmatch_slab;
5103 for (;curd > -1;cur--,curd--) {
5104 if (cur < SLAB_FIRST(slab)) {
5106 cur = SLAB_LAST(slab);
5108 PerlIO_printf(Perl_error_log, "%*s#%-3d %-10s %s\n",
5109 REPORT_CODE_OFF + 2 + depth * 2,"",
5110 curd, PL_reg_name[cur->resume_state],
5111 (curyes == cur) ? "yes" : ""
5114 curyes = cur->u.yes.prev_yes_state;
5117 DEBUG_STATE_pp("push")
5120 st->locinput = locinput;
5122 if (newst > SLAB_LAST(PL_regmatch_slab))
5123 newst = S_push_slab(aTHX);
5124 PL_regmatch_state = newst;
5126 locinput = PL_reginput;
5127 nextchr = UCHARAT(locinput);
5135 * We get here only if there's trouble -- normally "case END" is
5136 * the terminating point.
5138 Perl_croak(aTHX_ "corrupted regexp pointers");
5144 /* we have successfully completed a subexpression, but we must now
5145 * pop to the state marked by yes_state and continue from there */
5146 assert(st != yes_state);
5148 while (st != yes_state) {
5150 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5151 PL_regmatch_slab = PL_regmatch_slab->prev;
5152 st = SLAB_LAST(PL_regmatch_slab);
5156 DEBUG_STATE_pp("pop (no final)");
5158 DEBUG_STATE_pp("pop (yes)");
5164 while (yes_state < SLAB_FIRST(PL_regmatch_slab)
5165 || yes_state > SLAB_LAST(PL_regmatch_slab))
5167 /* not in this slab, pop slab */
5168 depth -= (st - SLAB_FIRST(PL_regmatch_slab) + 1);
5169 PL_regmatch_slab = PL_regmatch_slab->prev;
5170 st = SLAB_LAST(PL_regmatch_slab);
5172 depth -= (st - yes_state);
5175 yes_state = st->u.yes.prev_yes_state;
5176 PL_regmatch_state = st;
5179 locinput= st->locinput;
5180 nextchr = UCHARAT(locinput);
5182 state_num = st->resume_state + no_final;
5183 goto reenter_switch;
5186 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch successful!%s\n",
5187 PL_colors[4], PL_colors[5]));
5189 if (PL_reg_eval_set) {
5190 /* each successfully executed (?{...}) block does the equivalent of
5191 * local $^R = do {...}
5192 * When popping the save stack, all these locals would be undone;
5193 * bypass this by setting the outermost saved $^R to the latest
5195 if (oreplsv != GvSV(PL_replgv))
5196 sv_setsv(oreplsv, GvSV(PL_replgv));
5203 PerlIO_printf(Perl_debug_log,
5204 "%*s %sfailed...%s\n",
5205 REPORT_CODE_OFF+depth*2, "",
5206 PL_colors[4], PL_colors[5])
5218 /* there's a previous state to backtrack to */
5220 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5221 PL_regmatch_slab = PL_regmatch_slab->prev;
5222 st = SLAB_LAST(PL_regmatch_slab);
5224 PL_regmatch_state = st;
5225 locinput= st->locinput;
5226 nextchr = UCHARAT(locinput);
5228 DEBUG_STATE_pp("pop");
5230 if (yes_state == st)
5231 yes_state = st->u.yes.prev_yes_state;
5233 state_num = st->resume_state + 1; /* failure = success + 1 */
5234 goto reenter_switch;
5239 if (rex->intflags & PREGf_VERBARG_SEEN) {
5240 SV *sv_err = get_sv("REGERROR", 1);
5241 SV *sv_mrk = get_sv("REGMARK", 1);
5243 sv_commit = &PL_sv_no;
5245 sv_yes_mark = &PL_sv_yes;
5248 sv_commit = &PL_sv_yes;
5249 sv_yes_mark = &PL_sv_no;
5251 sv_setsv(sv_err, sv_commit);
5252 sv_setsv(sv_mrk, sv_yes_mark);
5255 /* clean up; in particular, free all slabs above current one */
5256 LEAVE_SCOPE(oldsave);
5262 - regrepeat - repeatedly match something simple, report how many
5265 * [This routine now assumes that it will only match on things of length 1.
5266 * That was true before, but now we assume scan - reginput is the count,
5267 * rather than incrementing count on every character. [Er, except utf8.]]
5270 S_regrepeat(pTHX_ const regexp *prog, const regnode *p, I32 max, int depth)
5273 register char *scan;
5275 register char *loceol = PL_regeol;
5276 register I32 hardcount = 0;
5277 register bool do_utf8 = PL_reg_match_utf8;
5279 PERL_UNUSED_ARG(depth);
5283 if (max == REG_INFTY)
5285 else if (max < loceol - scan)
5286 loceol = scan + max;
5291 while (scan < loceol && hardcount < max && *scan != '\n') {
5292 scan += UTF8SKIP(scan);
5296 while (scan < loceol && *scan != '\n')
5303 while (scan < loceol && hardcount < max) {
5304 scan += UTF8SKIP(scan);
5314 case EXACT: /* length of string is 1 */
5316 while (scan < loceol && UCHARAT(scan) == c)
5319 case EXACTF: /* length of string is 1 */
5321 while (scan < loceol &&
5322 (UCHARAT(scan) == c || UCHARAT(scan) == PL_fold[c]))
5325 case EXACTFL: /* length of string is 1 */
5326 PL_reg_flags |= RF_tainted;
5328 while (scan < loceol &&
5329 (UCHARAT(scan) == c || UCHARAT(scan) == PL_fold_locale[c]))
5335 while (hardcount < max && scan < loceol &&
5336 reginclass(prog, p, (U8*)scan, 0, do_utf8)) {
5337 scan += UTF8SKIP(scan);
5341 while (scan < loceol && REGINCLASS(prog, p, (U8*)scan))
5348 LOAD_UTF8_CHARCLASS_ALNUM();
5349 while (hardcount < max && scan < loceol &&
5350 swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
5351 scan += UTF8SKIP(scan);
5355 while (scan < loceol && isALNUM(*scan))
5360 PL_reg_flags |= RF_tainted;
5363 while (hardcount < max && scan < loceol &&
5364 isALNUM_LC_utf8((U8*)scan)) {
5365 scan += UTF8SKIP(scan);
5369 while (scan < loceol && isALNUM_LC(*scan))
5376 LOAD_UTF8_CHARCLASS_ALNUM();
5377 while (hardcount < max && scan < loceol &&
5378 !swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
5379 scan += UTF8SKIP(scan);
5383 while (scan < loceol && !isALNUM(*scan))
5388 PL_reg_flags |= RF_tainted;
5391 while (hardcount < max && scan < loceol &&
5392 !isALNUM_LC_utf8((U8*)scan)) {
5393 scan += UTF8SKIP(scan);
5397 while (scan < loceol && !isALNUM_LC(*scan))
5404 LOAD_UTF8_CHARCLASS_SPACE();
5405 while (hardcount < max && scan < loceol &&
5407 swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
5408 scan += UTF8SKIP(scan);
5412 while (scan < loceol && isSPACE(*scan))
5417 PL_reg_flags |= RF_tainted;
5420 while (hardcount < max && scan < loceol &&
5421 (*scan == ' ' || isSPACE_LC_utf8((U8*)scan))) {
5422 scan += UTF8SKIP(scan);
5426 while (scan < loceol && isSPACE_LC(*scan))
5433 LOAD_UTF8_CHARCLASS_SPACE();
5434 while (hardcount < max && scan < loceol &&
5436 swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
5437 scan += UTF8SKIP(scan);
5441 while (scan < loceol && !isSPACE(*scan))
5446 PL_reg_flags |= RF_tainted;
5449 while (hardcount < max && scan < loceol &&
5450 !(*scan == ' ' || isSPACE_LC_utf8((U8*)scan))) {
5451 scan += UTF8SKIP(scan);
5455 while (scan < loceol && !isSPACE_LC(*scan))
5462 LOAD_UTF8_CHARCLASS_DIGIT();
5463 while (hardcount < max && scan < loceol &&
5464 swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
5465 scan += UTF8SKIP(scan);
5469 while (scan < loceol && isDIGIT(*scan))
5476 LOAD_UTF8_CHARCLASS_DIGIT();
5477 while (hardcount < max && scan < loceol &&
5478 !swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
5479 scan += UTF8SKIP(scan);
5483 while (scan < loceol && !isDIGIT(*scan))
5489 while (hardcount < max && scan < loceol && (c=is_LNBREAK_utf8(scan))) {
5495 LNBREAK can match two latin chars, which is ok,
5496 because we have a null terminated string, but we
5497 have to use hardcount in this situation
5499 while (scan < loceol && (c=is_LNBREAK_latin1(scan))) {
5508 while (hardcount < max && scan < loceol && (c=is_HORIZWS_utf8(scan))) {
5513 while (scan < loceol && is_HORIZWS_latin1(scan))
5520 while (hardcount < max && scan < loceol && !is_HORIZWS_utf8(scan)) {
5521 scan += UTF8SKIP(scan);
5525 while (scan < loceol && !is_HORIZWS_latin1(scan))
5533 while (hardcount < max && scan < loceol && (c=is_VERTWS_utf8(scan))) {
5538 while (scan < loceol && is_VERTWS_latin1(scan))
5546 while (hardcount < max && scan < loceol && !is_VERTWS_utf8(scan)) {
5547 scan += UTF8SKIP(scan);
5551 while (scan < loceol && !is_VERTWS_latin1(scan))
5557 default: /* Called on something of 0 width. */
5558 break; /* So match right here or not at all. */
5564 c = scan - PL_reginput;
5568 GET_RE_DEBUG_FLAGS_DECL;
5570 SV * const prop = sv_newmortal();
5571 regprop(prog, prop, p);
5572 PerlIO_printf(Perl_debug_log,
5573 "%*s %s can match %"IVdf" times out of %"IVdf"...\n",
5574 REPORT_CODE_OFF + depth*2, "", SvPVX_const(prop),(IV)c,(IV)max);
5582 #if !defined(PERL_IN_XSUB_RE) || defined(PLUGGABLE_RE_EXTENSION)
5584 - regclass_swash - prepare the utf8 swash
5588 Perl_regclass_swash(pTHX_ const regexp *prog, register const regnode* node, bool doinit, SV** listsvp, SV **altsvp)
5594 RXi_GET_DECL(prog,progi);
5595 const struct reg_data * const data = prog ? progi->data : NULL;
5597 if (data && data->count) {
5598 const U32 n = ARG(node);
5600 if (data->what[n] == 's') {
5601 SV * const rv = (SV*)data->data[n];
5602 AV * const av = (AV*)SvRV((SV*)rv);
5603 SV **const ary = AvARRAY(av);
5606 /* See the end of regcomp.c:S_regclass() for
5607 * documentation of these array elements. */
5610 a = SvROK(ary[1]) ? &ary[1] : NULL;
5611 b = SvTYPE(ary[2]) == SVt_PVAV ? &ary[2] : NULL;
5615 else if (si && doinit) {
5616 sw = swash_init("utf8", "", si, 1, 0);
5617 (void)av_store(av, 1, sw);
5634 - reginclass - determine if a character falls into a character class
5636 The n is the ANYOF regnode, the p is the target string, lenp
5637 is pointer to the maximum length of how far to go in the p
5638 (if the lenp is zero, UTF8SKIP(p) is used),
5639 do_utf8 tells whether the target string is in UTF-8.
5644 S_reginclass(pTHX_ const regexp *prog, register const regnode *n, register const U8* p, STRLEN* lenp, register bool do_utf8)
5647 const char flags = ANYOF_FLAGS(n);
5653 if (do_utf8 && !UTF8_IS_INVARIANT(c)) {
5654 c = utf8n_to_uvchr(p, UTF8_MAXBYTES, &len,
5655 (UTF8_ALLOW_DEFAULT & UTF8_ALLOW_ANYUV) | UTF8_CHECK_ONLY);
5656 /* see [perl #37836] for UTF8_ALLOW_ANYUV */
5657 if (len == (STRLEN)-1)
5658 Perl_croak(aTHX_ "Malformed UTF-8 character (fatal)");
5661 plen = lenp ? *lenp : UNISKIP(NATIVE_TO_UNI(c));
5662 if (do_utf8 || (flags & ANYOF_UNICODE)) {
5665 if (do_utf8 && !ANYOF_RUNTIME(n)) {
5666 if (len != (STRLEN)-1 && c < 256 && ANYOF_BITMAP_TEST(n, c))
5669 if (!match && do_utf8 && (flags & ANYOF_UNICODE_ALL) && c >= 256)
5673 SV * const sw = regclass_swash(prog, n, TRUE, 0, (SV**)&av);
5676 if (swash_fetch(sw, p, do_utf8))
5678 else if (flags & ANYOF_FOLD) {
5679 if (!match && lenp && av) {
5681 for (i = 0; i <= av_len(av); i++) {
5682 SV* const sv = *av_fetch(av, i, FALSE);
5684 const char * const s = SvPV_const(sv, len);
5686 if (len <= plen && memEQ(s, (char*)p, len)) {
5694 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
5697 to_utf8_fold(p, tmpbuf, &tmplen);
5698 if (swash_fetch(sw, tmpbuf, do_utf8))
5704 if (match && lenp && *lenp == 0)
5705 *lenp = UNISKIP(NATIVE_TO_UNI(c));
5707 if (!match && c < 256) {
5708 if (ANYOF_BITMAP_TEST(n, c))
5710 else if (flags & ANYOF_FOLD) {
5713 if (flags & ANYOF_LOCALE) {
5714 PL_reg_flags |= RF_tainted;
5715 f = PL_fold_locale[c];
5719 if (f != c && ANYOF_BITMAP_TEST(n, f))
5723 if (!match && (flags & ANYOF_CLASS)) {
5724 PL_reg_flags |= RF_tainted;
5726 (ANYOF_CLASS_TEST(n, ANYOF_ALNUM) && isALNUM_LC(c)) ||
5727 (ANYOF_CLASS_TEST(n, ANYOF_NALNUM) && !isALNUM_LC(c)) ||
5728 (ANYOF_CLASS_TEST(n, ANYOF_SPACE) && isSPACE_LC(c)) ||
5729 (ANYOF_CLASS_TEST(n, ANYOF_NSPACE) && !isSPACE_LC(c)) ||
5730 (ANYOF_CLASS_TEST(n, ANYOF_DIGIT) && isDIGIT_LC(c)) ||
5731 (ANYOF_CLASS_TEST(n, ANYOF_NDIGIT) && !isDIGIT_LC(c)) ||
5732 (ANYOF_CLASS_TEST(n, ANYOF_ALNUMC) && isALNUMC_LC(c)) ||
5733 (ANYOF_CLASS_TEST(n, ANYOF_NALNUMC) && !isALNUMC_LC(c)) ||
5734 (ANYOF_CLASS_TEST(n, ANYOF_ALPHA) && isALPHA_LC(c)) ||
5735 (ANYOF_CLASS_TEST(n, ANYOF_NALPHA) && !isALPHA_LC(c)) ||
5736 (ANYOF_CLASS_TEST(n, ANYOF_ASCII) && isASCII(c)) ||
5737 (ANYOF_CLASS_TEST(n, ANYOF_NASCII) && !isASCII(c)) ||
5738 (ANYOF_CLASS_TEST(n, ANYOF_CNTRL) && isCNTRL_LC(c)) ||
5739 (ANYOF_CLASS_TEST(n, ANYOF_NCNTRL) && !isCNTRL_LC(c)) ||
5740 (ANYOF_CLASS_TEST(n, ANYOF_GRAPH) && isGRAPH_LC(c)) ||
5741 (ANYOF_CLASS_TEST(n, ANYOF_NGRAPH) && !isGRAPH_LC(c)) ||
5742 (ANYOF_CLASS_TEST(n, ANYOF_LOWER) && isLOWER_LC(c)) ||
5743 (ANYOF_CLASS_TEST(n, ANYOF_NLOWER) && !isLOWER_LC(c)) ||
5744 (ANYOF_CLASS_TEST(n, ANYOF_PRINT) && isPRINT_LC(c)) ||
5745 (ANYOF_CLASS_TEST(n, ANYOF_NPRINT) && !isPRINT_LC(c)) ||
5746 (ANYOF_CLASS_TEST(n, ANYOF_PUNCT) && isPUNCT_LC(c)) ||
5747 (ANYOF_CLASS_TEST(n, ANYOF_NPUNCT) && !isPUNCT_LC(c)) ||
5748 (ANYOF_CLASS_TEST(n, ANYOF_UPPER) && isUPPER_LC(c)) ||
5749 (ANYOF_CLASS_TEST(n, ANYOF_NUPPER) && !isUPPER_LC(c)) ||
5750 (ANYOF_CLASS_TEST(n, ANYOF_XDIGIT) && isXDIGIT(c)) ||
5751 (ANYOF_CLASS_TEST(n, ANYOF_NXDIGIT) && !isXDIGIT(c)) ||
5752 (ANYOF_CLASS_TEST(n, ANYOF_PSXSPC) && isPSXSPC(c)) ||
5753 (ANYOF_CLASS_TEST(n, ANYOF_NPSXSPC) && !isPSXSPC(c)) ||
5754 (ANYOF_CLASS_TEST(n, ANYOF_BLANK) && isBLANK(c)) ||
5755 (ANYOF_CLASS_TEST(n, ANYOF_NBLANK) && !isBLANK(c))
5756 ) /* How's that for a conditional? */
5763 return (flags & ANYOF_INVERT) ? !match : match;
5767 S_reghop3(U8 *s, I32 off, const U8* lim)
5771 while (off-- && s < lim) {
5772 /* XXX could check well-formedness here */
5777 while (off++ && s > lim) {
5779 if (UTF8_IS_CONTINUED(*s)) {
5780 while (s > lim && UTF8_IS_CONTINUATION(*s))
5783 /* XXX could check well-formedness here */
5790 /* there are a bunch of places where we use two reghop3's that should
5791 be replaced with this routine. but since thats not done yet
5792 we ifdef it out - dmq
5795 S_reghop4(U8 *s, I32 off, const U8* llim, const U8* rlim)
5799 while (off-- && s < rlim) {
5800 /* XXX could check well-formedness here */
5805 while (off++ && s > llim) {
5807 if (UTF8_IS_CONTINUED(*s)) {
5808 while (s > llim && UTF8_IS_CONTINUATION(*s))
5811 /* XXX could check well-formedness here */
5819 S_reghopmaybe3(U8* s, I32 off, const U8* lim)
5823 while (off-- && s < lim) {
5824 /* XXX could check well-formedness here */
5831 while (off++ && s > lim) {
5833 if (UTF8_IS_CONTINUED(*s)) {
5834 while (s > lim && UTF8_IS_CONTINUATION(*s))
5837 /* XXX could check well-formedness here */
5846 restore_pos(pTHX_ void *arg)
5849 regexp * const rex = (regexp *)arg;
5850 if (PL_reg_eval_set) {
5851 if (PL_reg_oldsaved) {
5852 rex->subbeg = PL_reg_oldsaved;
5853 rex->sublen = PL_reg_oldsavedlen;
5854 #ifdef PERL_OLD_COPY_ON_WRITE
5855 rex->saved_copy = PL_nrs;
5857 RX_MATCH_COPIED_on(rex);
5859 PL_reg_magic->mg_len = PL_reg_oldpos;
5860 PL_reg_eval_set = 0;
5861 PL_curpm = PL_reg_oldcurpm;
5866 S_to_utf8_substr(pTHX_ register regexp *prog)
5870 if (prog->substrs->data[i].substr
5871 && !prog->substrs->data[i].utf8_substr) {
5872 SV* const sv = newSVsv(prog->substrs->data[i].substr);
5873 prog->substrs->data[i].utf8_substr = sv;
5874 sv_utf8_upgrade(sv);
5875 if (SvVALID(prog->substrs->data[i].substr)) {
5876 const U8 flags = BmFLAGS(prog->substrs->data[i].substr);
5877 if (flags & FBMcf_TAIL) {
5878 /* Trim the trailing \n that fbm_compile added last
5880 SvCUR_set(sv, SvCUR(sv) - 1);
5881 /* Whilst this makes the SV technically "invalid" (as its
5882 buffer is no longer followed by "\0") when fbm_compile()
5883 adds the "\n" back, a "\0" is restored. */
5885 fbm_compile(sv, flags);
5887 if (prog->substrs->data[i].substr == prog->check_substr)
5888 prog->check_utf8 = sv;
5894 S_to_byte_substr(pTHX_ register regexp *prog)
5899 if (prog->substrs->data[i].utf8_substr
5900 && !prog->substrs->data[i].substr) {
5901 SV* sv = newSVsv(prog->substrs->data[i].utf8_substr);
5902 if (sv_utf8_downgrade(sv, TRUE)) {
5903 if (SvVALID(prog->substrs->data[i].utf8_substr)) {
5905 = BmFLAGS(prog->substrs->data[i].utf8_substr);
5906 if (flags & FBMcf_TAIL) {
5907 /* Trim the trailing \n that fbm_compile added last
5909 SvCUR_set(sv, SvCUR(sv) - 1);
5911 fbm_compile(sv, flags);
5917 prog->substrs->data[i].substr = sv;
5918 if (prog->substrs->data[i].utf8_substr == prog->check_utf8)
5919 prog->check_substr = sv;
5926 * c-indentation-style: bsd
5928 * indent-tabs-mode: t
5931 * ex: set ts=8 sts=4 sw=4 noet: