5 * One Ring to rule them all, One Ring to find them
7 * [p.v of _The Lord of the Rings_, opening poem]
8 * [p.50 of _The Lord of the Rings_, I/iii: "The Shadow of the Past"]
9 * [p.254 of _The Lord of the Rings_, II/ii: "The Council of Elrond"]
12 /* This file contains functions for executing a regular expression. See
13 * also regcomp.c which funnily enough, contains functions for compiling
14 * a regular expression.
16 * This file is also copied at build time to ext/re/re_exec.c, where
17 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
18 * This causes the main functions to be compiled under new names and with
19 * debugging support added, which makes "use re 'debug'" work.
22 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
23 * confused with the original package (see point 3 below). Thanks, Henry!
26 /* Additional note: this code is very heavily munged from Henry's version
27 * in places. In some spots I've traded clarity for efficiency, so don't
28 * blame Henry for some of the lack of readability.
31 /* The names of the functions have been changed from regcomp and
32 * regexec to pregcomp and pregexec in order to avoid conflicts
33 * with the POSIX routines of the same names.
36 #ifdef PERL_EXT_RE_BUILD
41 * pregcomp and pregexec -- regsub and regerror are not used in perl
43 * Copyright (c) 1986 by University of Toronto.
44 * Written by Henry Spencer. Not derived from licensed software.
46 * Permission is granted to anyone to use this software for any
47 * purpose on any computer system, and to redistribute it freely,
48 * subject to the following restrictions:
50 * 1. The author is not responsible for the consequences of use of
51 * this software, no matter how awful, even if they arise
54 * 2. The origin of this software must not be misrepresented, either
55 * by explicit claim or by omission.
57 * 3. Altered versions must be plainly marked as such, and must not
58 * be misrepresented as being the original software.
60 **** Alterations to Henry's code are...
62 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
63 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
64 **** by Larry Wall and others
66 **** You may distribute under the terms of either the GNU General Public
67 **** License or the Artistic License, as specified in the README file.
69 * Beware that some of this code is subtly aware of the way operator
70 * precedence is structured in regular expressions. Serious changes in
71 * regular-expression syntax might require a total rethink.
74 #define PERL_IN_REGEXEC_C
77 #ifdef PERL_IN_XSUB_RE
83 #define RF_tainted 1 /* tainted information used? */
84 #define RF_warned 2 /* warned about big count? */
86 #define RF_utf8 8 /* Pattern contains multibyte chars? */
88 #define UTF ((PL_reg_flags & RF_utf8) != 0)
90 #define RS_init 1 /* eval environment created */
91 #define RS_set 2 /* replsv value is set */
97 #define REGINCLASS(prog,p,c) (ANYOF_FLAGS(p) ? reginclass(prog,p,c,0,0) : ANYOF_BITMAP_TEST(p,*(c)))
103 #define CHR_SVLEN(sv) (do_utf8 ? sv_len_utf8(sv) : SvCUR(sv))
104 #define CHR_DIST(a,b) (PL_reg_match_utf8 ? utf8_distance(a,b) : a - b)
106 #define HOPc(pos,off) \
107 (char *)(PL_reg_match_utf8 \
108 ? reghop3((U8*)pos, off, (U8*)(off >= 0 ? PL_regeol : PL_bostr)) \
110 #define HOPBACKc(pos, off) \
111 (char*)(PL_reg_match_utf8\
112 ? reghopmaybe3((U8*)pos, -off, (U8*)PL_bostr) \
113 : (pos - off >= PL_bostr) \
117 #define HOP3(pos,off,lim) (PL_reg_match_utf8 ? reghop3((U8*)(pos), off, (U8*)(lim)) : (U8*)(pos + off))
118 #define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim))
120 #define LOAD_UTF8_CHARCLASS(class,str) STMT_START { \
121 if (!CAT2(PL_utf8_,class)) { bool ok; ENTER; save_re_context(); ok=CAT2(is_utf8_,class)((const U8*)str); assert(ok); LEAVE; } } STMT_END
122 #define LOAD_UTF8_CHARCLASS_ALNUM() LOAD_UTF8_CHARCLASS(alnum,"a")
123 #define LOAD_UTF8_CHARCLASS_DIGIT() LOAD_UTF8_CHARCLASS(digit,"0")
124 #define LOAD_UTF8_CHARCLASS_SPACE() LOAD_UTF8_CHARCLASS(space," ")
125 #define LOAD_UTF8_CHARCLASS_MARK() LOAD_UTF8_CHARCLASS(mark, "\xcd\x86")
127 /* TODO: Combine JUMPABLE and HAS_TEXT to cache OP(rn) */
129 /* for use after a quantifier and before an EXACT-like node -- japhy */
130 /* it would be nice to rework regcomp.sym to generate this stuff. sigh */
131 #define JUMPABLE(rn) ( \
133 (OP(rn) == CLOSE && (!cur_eval || cur_eval->u.eval.close_paren != ARG(rn))) || \
135 OP(rn) == SUSPEND || OP(rn) == IFMATCH || \
136 OP(rn) == PLUS || OP(rn) == MINMOD || \
137 OP(rn) == KEEPS || (PL_regkind[OP(rn)] == VERB) || \
138 (PL_regkind[OP(rn)] == CURLY && ARG1(rn) > 0) \
140 #define IS_EXACT(rn) (PL_regkind[OP(rn)] == EXACT)
142 #define HAS_TEXT(rn) ( IS_EXACT(rn) || PL_regkind[OP(rn)] == REF )
145 /* Currently these are only used when PL_regkind[OP(rn)] == EXACT so
146 we don't need this definition. */
147 #define IS_TEXT(rn) ( OP(rn)==EXACT || OP(rn)==REF || OP(rn)==NREF )
148 #define IS_TEXTF(rn) ( OP(rn)==EXACTF || OP(rn)==REFF || OP(rn)==NREFF )
149 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL || OP(rn)==REFFL || OP(rn)==NREFFL )
152 /* ... so we use this as its faster. */
153 #define IS_TEXT(rn) ( OP(rn)==EXACT )
154 #define IS_TEXTF(rn) ( OP(rn)==EXACTF )
155 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL )
160 Search for mandatory following text node; for lookahead, the text must
161 follow but for lookbehind (rn->flags != 0) we skip to the next step.
163 #define FIND_NEXT_IMPT(rn) STMT_START { \
164 while (JUMPABLE(rn)) { \
165 const OPCODE type = OP(rn); \
166 if (type == SUSPEND || PL_regkind[type] == CURLY) \
167 rn = NEXTOPER(NEXTOPER(rn)); \
168 else if (type == PLUS) \
170 else if (type == IFMATCH) \
171 rn = (rn->flags == 0) ? NEXTOPER(NEXTOPER(rn)) : rn + ARG(rn); \
172 else rn += NEXT_OFF(rn); \
177 static void restore_pos(pTHX_ void *arg);
180 S_regcppush(pTHX_ I32 parenfloor)
183 const int retval = PL_savestack_ix;
184 #define REGCP_PAREN_ELEMS 4
185 const int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS;
187 GET_RE_DEBUG_FLAGS_DECL;
189 if (paren_elems_to_push < 0)
190 Perl_croak(aTHX_ "panic: paren_elems_to_push < 0");
192 #define REGCP_OTHER_ELEMS 7
193 SSGROW(paren_elems_to_push + REGCP_OTHER_ELEMS);
195 for (p = PL_regsize; p > parenfloor; p--) {
196 /* REGCP_PARENS_ELEMS are pushed per pairs of parentheses. */
197 SSPUSHINT(PL_regoffs[p].end);
198 SSPUSHINT(PL_regoffs[p].start);
199 SSPUSHPTR(PL_reg_start_tmp[p]);
201 DEBUG_BUFFERS_r(PerlIO_printf(Perl_debug_log,
202 " saving \\%"UVuf" %"IVdf"(%"IVdf")..%"IVdf"\n",
203 (UV)p, (IV)PL_regoffs[p].start,
204 (IV)(PL_reg_start_tmp[p] - PL_bostr),
205 (IV)PL_regoffs[p].end
208 /* REGCP_OTHER_ELEMS are pushed in any case, parentheses or no. */
209 SSPUSHPTR(PL_regoffs);
210 SSPUSHINT(PL_regsize);
211 SSPUSHINT(*PL_reglastparen);
212 SSPUSHINT(*PL_reglastcloseparen);
213 SSPUSHPTR(PL_reginput);
214 #define REGCP_FRAME_ELEMS 2
215 /* REGCP_FRAME_ELEMS are part of the REGCP_OTHER_ELEMS and
216 * are needed for the regexp context stack bookkeeping. */
217 SSPUSHINT(paren_elems_to_push + REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS);
218 SSPUSHINT(SAVEt_REGCONTEXT); /* Magic cookie. */
223 /* These are needed since we do not localize EVAL nodes: */
224 #define REGCP_SET(cp) \
226 PerlIO_printf(Perl_debug_log, \
227 " Setting an EVAL scope, savestack=%"IVdf"\n", \
228 (IV)PL_savestack_ix)); \
231 #define REGCP_UNWIND(cp) \
233 if (cp != PL_savestack_ix) \
234 PerlIO_printf(Perl_debug_log, \
235 " Clearing an EVAL scope, savestack=%"IVdf"..%"IVdf"\n", \
236 (IV)(cp), (IV)PL_savestack_ix)); \
240 S_regcppop(pTHX_ const regexp *rex)
245 GET_RE_DEBUG_FLAGS_DECL;
247 PERL_ARGS_ASSERT_REGCPPOP;
249 /* Pop REGCP_OTHER_ELEMS before the parentheses loop starts. */
251 assert(i == SAVEt_REGCONTEXT); /* Check that the magic cookie is there. */
252 i = SSPOPINT; /* Parentheses elements to pop. */
253 input = (char *) SSPOPPTR;
254 *PL_reglastcloseparen = SSPOPINT;
255 *PL_reglastparen = SSPOPINT;
256 PL_regsize = SSPOPINT;
257 PL_regoffs=(regexp_paren_pair *) SSPOPPTR;
260 /* Now restore the parentheses context. */
261 for (i -= (REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS);
262 i > 0; i -= REGCP_PAREN_ELEMS) {
264 U32 paren = (U32)SSPOPINT;
265 PL_reg_start_tmp[paren] = (char *) SSPOPPTR;
266 PL_regoffs[paren].start = SSPOPINT;
268 if (paren <= *PL_reglastparen)
269 PL_regoffs[paren].end = tmps;
271 PerlIO_printf(Perl_debug_log,
272 " restoring \\%"UVuf" to %"IVdf"(%"IVdf")..%"IVdf"%s\n",
273 (UV)paren, (IV)PL_regoffs[paren].start,
274 (IV)(PL_reg_start_tmp[paren] - PL_bostr),
275 (IV)PL_regoffs[paren].end,
276 (paren > *PL_reglastparen ? "(no)" : ""));
280 if (*PL_reglastparen + 1 <= rex->nparens) {
281 PerlIO_printf(Perl_debug_log,
282 " restoring \\%"IVdf"..\\%"IVdf" to undef\n",
283 (IV)(*PL_reglastparen + 1), (IV)rex->nparens);
287 /* It would seem that the similar code in regtry()
288 * already takes care of this, and in fact it is in
289 * a better location to since this code can #if 0-ed out
290 * but the code in regtry() is needed or otherwise tests
291 * requiring null fields (pat.t#187 and split.t#{13,14}
292 * (as of patchlevel 7877) will fail. Then again,
293 * this code seems to be necessary or otherwise
294 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
295 * --jhi updated by dapm */
296 for (i = *PL_reglastparen + 1; i <= rex->nparens; i++) {
298 PL_regoffs[i].start = -1;
299 PL_regoffs[i].end = -1;
305 #define regcpblow(cp) LEAVE_SCOPE(cp) /* Ignores regcppush()ed data. */
308 * pregexec and friends
311 #ifndef PERL_IN_XSUB_RE
313 - pregexec - match a regexp against a string
316 Perl_pregexec(pTHX_ REGEXP * const prog, char* stringarg, register char *strend,
317 char *strbeg, I32 minend, SV *screamer, U32 nosave)
318 /* strend: pointer to null at end of string */
319 /* strbeg: real beginning of string */
320 /* minend: end of match must be >=minend after stringarg. */
321 /* nosave: For optimizations. */
323 PERL_ARGS_ASSERT_PREGEXEC;
326 regexec_flags(prog, stringarg, strend, strbeg, minend, screamer, NULL,
327 nosave ? 0 : REXEC_COPY_STR);
332 * Need to implement the following flags for reg_anch:
334 * USE_INTUIT_NOML - Useful to call re_intuit_start() first
336 * INTUIT_AUTORITATIVE_NOML - Can trust a positive answer
337 * INTUIT_AUTORITATIVE_ML
338 * INTUIT_ONCE_NOML - Intuit can match in one location only.
341 * Another flag for this function: SECOND_TIME (so that float substrs
342 * with giant delta may be not rechecked).
345 /* Assumptions: if ANCH_GPOS, then strpos is anchored. XXXX Check GPOS logic */
347 /* If SCREAM, then SvPVX_const(sv) should be compatible with strpos and strend.
348 Otherwise, only SvCUR(sv) is used to get strbeg. */
350 /* XXXX We assume that strpos is strbeg unless sv. */
352 /* XXXX Some places assume that there is a fixed substring.
353 An update may be needed if optimizer marks as "INTUITable"
354 RExen without fixed substrings. Similarly, it is assumed that
355 lengths of all the strings are no more than minlen, thus they
356 cannot come from lookahead.
357 (Or minlen should take into account lookahead.)
358 NOTE: Some of this comment is not correct. minlen does now take account
359 of lookahead/behind. Further research is required. -- demerphq
363 /* A failure to find a constant substring means that there is no need to make
364 an expensive call to REx engine, thus we celebrate a failure. Similarly,
365 finding a substring too deep into the string means that less calls to
366 regtry() should be needed.
368 REx compiler's optimizer found 4 possible hints:
369 a) Anchored substring;
371 c) Whether we are anchored (beginning-of-line or \G);
372 d) First node (of those at offset 0) which may distingush positions;
373 We use a)b)d) and multiline-part of c), and try to find a position in the
374 string which does not contradict any of them.
377 /* Most of decisions we do here should have been done at compile time.
378 The nodes of the REx which we used for the search should have been
379 deleted from the finite automaton. */
382 Perl_re_intuit_start(pTHX_ REGEXP * const rx, SV *sv, char *strpos,
383 char *strend, const U32 flags, re_scream_pos_data *data)
386 struct regexp *const prog = (struct regexp *)SvANY(rx);
387 register I32 start_shift = 0;
388 /* Should be nonnegative! */
389 register I32 end_shift = 0;
394 const bool do_utf8 = (sv && SvUTF8(sv)) ? 1 : 0; /* if no sv we have to assume bytes */
396 register char *other_last = NULL; /* other substr checked before this */
397 char *check_at = NULL; /* check substr found at this pos */
398 const I32 multiline = prog->extflags & RXf_PMf_MULTILINE;
399 RXi_GET_DECL(prog,progi);
401 const char * const i_strpos = strpos;
403 GET_RE_DEBUG_FLAGS_DECL;
405 PERL_ARGS_ASSERT_RE_INTUIT_START;
407 RX_MATCH_UTF8_set(rx,do_utf8);
410 PL_reg_flags |= RF_utf8;
413 debug_start_match(rx, do_utf8, strpos, strend,
414 sv ? "Guessing start of match in sv for"
415 : "Guessing start of match in string for");
418 /* CHR_DIST() would be more correct here but it makes things slow. */
419 if (prog->minlen > strend - strpos) {
420 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
421 "String too short... [re_intuit_start]\n"));
425 strbeg = (sv && SvPOK(sv)) ? strend - SvCUR(sv) : strpos;
428 if (!prog->check_utf8 && prog->check_substr)
429 to_utf8_substr(prog);
430 check = prog->check_utf8;
432 if (!prog->check_substr && prog->check_utf8)
433 to_byte_substr(prog);
434 check = prog->check_substr;
436 if (check == &PL_sv_undef) {
437 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
438 "Non-utf8 string cannot match utf8 check string\n"));
441 if (prog->extflags & RXf_ANCH) { /* Match at beg-of-str or after \n */
442 ml_anch = !( (prog->extflags & RXf_ANCH_SINGLE)
443 || ( (prog->extflags & RXf_ANCH_BOL)
444 && !multiline ) ); /* Check after \n? */
447 if ( !(prog->extflags & RXf_ANCH_GPOS) /* Checked by the caller */
448 && !(prog->intflags & PREGf_IMPLICIT) /* not a real BOL */
449 /* SvCUR is not set on references: SvRV and SvPVX_const overlap */
451 && (strpos != strbeg)) {
452 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not at start...\n"));
455 if (prog->check_offset_min == prog->check_offset_max &&
456 !(prog->extflags & RXf_CANY_SEEN)) {
457 /* Substring at constant offset from beg-of-str... */
460 s = HOP3c(strpos, prog->check_offset_min, strend);
463 slen = SvCUR(check); /* >= 1 */
465 if ( strend - s > slen || strend - s < slen - 1
466 || (strend - s == slen && strend[-1] != '\n')) {
467 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String too long...\n"));
470 /* Now should match s[0..slen-2] */
472 if (slen && (*SvPVX_const(check) != *s
474 && memNE(SvPVX_const(check), s, slen)))) {
476 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String not equal...\n"));
480 else if (*SvPVX_const(check) != *s
481 || ((slen = SvCUR(check)) > 1
482 && memNE(SvPVX_const(check), s, slen)))
485 goto success_at_start;
488 /* Match is anchored, but substr is not anchored wrt beg-of-str. */
490 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
491 end_shift = prog->check_end_shift;
494 const I32 end = prog->check_offset_max + CHR_SVLEN(check)
495 - (SvTAIL(check) != 0);
496 const I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end;
498 if (end_shift < eshift)
502 else { /* Can match at random position */
505 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
506 end_shift = prog->check_end_shift;
508 /* end shift should be non negative here */
511 #ifdef QDEBUGGING /* 7/99: reports of failure (with the older version) */
513 Perl_croak(aTHX_ "panic: end_shift: %"IVdf" pattern:\n%s\n ",
514 (IV)end_shift, RX_PRECOMP(prog));
518 /* Find a possible match in the region s..strend by looking for
519 the "check" substring in the region corrected by start/end_shift. */
522 I32 srch_start_shift = start_shift;
523 I32 srch_end_shift = end_shift;
524 if (srch_start_shift < 0 && strbeg - s > srch_start_shift) {
525 srch_end_shift -= ((strbeg - s) - srch_start_shift);
526 srch_start_shift = strbeg - s;
528 DEBUG_OPTIMISE_MORE_r({
529 PerlIO_printf(Perl_debug_log, "Check offset min: %"IVdf" Start shift: %"IVdf" End shift %"IVdf" Real End Shift: %"IVdf"\n",
530 (IV)prog->check_offset_min,
531 (IV)srch_start_shift,
533 (IV)prog->check_end_shift);
536 if (flags & REXEC_SCREAM) {
537 I32 p = -1; /* Internal iterator of scream. */
538 I32 * const pp = data ? data->scream_pos : &p;
540 if (PL_screamfirst[BmRARE(check)] >= 0
541 || ( BmRARE(check) == '\n'
542 && (BmPREVIOUS(check) == SvCUR(check) - 1)
544 s = screaminstr(sv, check,
545 srch_start_shift + (s - strbeg), srch_end_shift, pp, 0);
548 /* we may be pointing at the wrong string */
549 if (s && RXp_MATCH_COPIED(prog))
550 s = strbeg + (s - SvPVX_const(sv));
552 *data->scream_olds = s;
557 if (prog->extflags & RXf_CANY_SEEN) {
558 start_point= (U8*)(s + srch_start_shift);
559 end_point= (U8*)(strend - srch_end_shift);
561 start_point= HOP3(s, srch_start_shift, srch_start_shift < 0 ? strbeg : strend);
562 end_point= HOP3(strend, -srch_end_shift, strbeg);
564 DEBUG_OPTIMISE_MORE_r({
565 PerlIO_printf(Perl_debug_log, "fbm_instr len=%d str=<%.*s>\n",
566 (int)(end_point - start_point),
567 (int)(end_point - start_point) > 20 ? 20 : (int)(end_point - start_point),
571 s = fbm_instr( start_point, end_point,
572 check, multiline ? FBMrf_MULTILINE : 0);
575 /* Update the count-of-usability, remove useless subpatterns,
579 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
580 SvPVX_const(check), RE_SV_DUMPLEN(check), 30);
581 PerlIO_printf(Perl_debug_log, "%s %s substr %s%s%s",
582 (s ? "Found" : "Did not find"),
583 (check == (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)
584 ? "anchored" : "floating"),
587 (s ? " at offset " : "...\n") );
592 /* Finish the diagnostic message */
593 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%ld...\n", (long)(s - i_strpos)) );
595 /* XXX dmq: first branch is for positive lookbehind...
596 Our check string is offset from the beginning of the pattern.
597 So we need to do any stclass tests offset forward from that
606 /* Got a candidate. Check MBOL anchoring, and the *other* substr.
607 Start with the other substr.
608 XXXX no SCREAM optimization yet - and a very coarse implementation
609 XXXX /ttx+/ results in anchored="ttx", floating="x". floating will
610 *always* match. Probably should be marked during compile...
611 Probably it is right to do no SCREAM here...
614 if (do_utf8 ? (prog->float_utf8 && prog->anchored_utf8)
615 : (prog->float_substr && prog->anchored_substr))
617 /* Take into account the "other" substring. */
618 /* XXXX May be hopelessly wrong for UTF... */
621 if (check == (do_utf8 ? prog->float_utf8 : prog->float_substr)) {
624 char * const last = HOP3c(s, -start_shift, strbeg);
626 char * const saved_s = s;
629 t = s - prog->check_offset_max;
630 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
632 || ((t = (char*)reghopmaybe3((U8*)s, -(prog->check_offset_max), (U8*)strpos))
637 t = HOP3c(t, prog->anchored_offset, strend);
638 if (t < other_last) /* These positions already checked */
640 last2 = last1 = HOP3c(strend, -prog->minlen, strbeg);
643 /* XXXX It is not documented what units *_offsets are in.
644 We assume bytes, but this is clearly wrong.
645 Meaning this code needs to be carefully reviewed for errors.
649 /* On end-of-str: see comment below. */
650 must = do_utf8 ? prog->anchored_utf8 : prog->anchored_substr;
651 if (must == &PL_sv_undef) {
653 DEBUG_r(must = prog->anchored_utf8); /* for debug */
658 HOP3(HOP3(last1, prog->anchored_offset, strend)
659 + SvCUR(must), -(SvTAIL(must)!=0), strbeg),
661 multiline ? FBMrf_MULTILINE : 0
664 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
665 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
666 PerlIO_printf(Perl_debug_log, "%s anchored substr %s%s",
667 (s ? "Found" : "Contradicts"),
668 quoted, RE_SV_TAIL(must));
673 if (last1 >= last2) {
674 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
675 ", giving up...\n"));
678 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
679 ", trying floating at offset %ld...\n",
680 (long)(HOP3c(saved_s, 1, strend) - i_strpos)));
681 other_last = HOP3c(last1, prog->anchored_offset+1, strend);
682 s = HOP3c(last, 1, strend);
686 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
687 (long)(s - i_strpos)));
688 t = HOP3c(s, -prog->anchored_offset, strbeg);
689 other_last = HOP3c(s, 1, strend);
697 else { /* Take into account the floating substring. */
699 char * const saved_s = s;
702 t = HOP3c(s, -start_shift, strbeg);
704 HOP3c(strend, -prog->minlen + prog->float_min_offset, strbeg);
705 if (CHR_DIST((U8*)last, (U8*)t) > prog->float_max_offset)
706 last = HOP3c(t, prog->float_max_offset, strend);
707 s = HOP3c(t, prog->float_min_offset, strend);
710 /* XXXX It is not documented what units *_offsets are in. Assume bytes. */
711 must = do_utf8 ? prog->float_utf8 : prog->float_substr;
712 /* fbm_instr() takes into account exact value of end-of-str
713 if the check is SvTAIL(ed). Since false positives are OK,
714 and end-of-str is not later than strend we are OK. */
715 if (must == &PL_sv_undef) {
717 DEBUG_r(must = prog->float_utf8); /* for debug message */
720 s = fbm_instr((unsigned char*)s,
721 (unsigned char*)last + SvCUR(must)
723 must, multiline ? FBMrf_MULTILINE : 0);
725 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
726 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
727 PerlIO_printf(Perl_debug_log, "%s floating substr %s%s",
728 (s ? "Found" : "Contradicts"),
729 quoted, RE_SV_TAIL(must));
733 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
734 ", giving up...\n"));
737 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
738 ", trying anchored starting at offset %ld...\n",
739 (long)(saved_s + 1 - i_strpos)));
741 s = HOP3c(t, 1, strend);
745 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
746 (long)(s - i_strpos)));
747 other_last = s; /* Fix this later. --Hugo */
757 t= (char*)HOP3( s, -prog->check_offset_max, (prog->check_offset_max<0) ? strend : strpos);
759 DEBUG_OPTIMISE_MORE_r(
760 PerlIO_printf(Perl_debug_log,
761 "Check offset min:%"IVdf" max:%"IVdf" S:%"IVdf" t:%"IVdf" D:%"IVdf" end:%"IVdf"\n",
762 (IV)prog->check_offset_min,
763 (IV)prog->check_offset_max,
771 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
773 || ((t = (char*)reghopmaybe3((U8*)s, -prog->check_offset_max, (U8*) ((prog->check_offset_max<0) ? strend : strpos)))
776 /* Fixed substring is found far enough so that the match
777 cannot start at strpos. */
779 if (ml_anch && t[-1] != '\n') {
780 /* Eventually fbm_*() should handle this, but often
781 anchored_offset is not 0, so this check will not be wasted. */
782 /* XXXX In the code below we prefer to look for "^" even in
783 presence of anchored substrings. And we search even
784 beyond the found float position. These pessimizations
785 are historical artefacts only. */
787 while (t < strend - prog->minlen) {
789 if (t < check_at - prog->check_offset_min) {
790 if (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) {
791 /* Since we moved from the found position,
792 we definitely contradict the found anchored
793 substr. Due to the above check we do not
794 contradict "check" substr.
795 Thus we can arrive here only if check substr
796 is float. Redo checking for "other"=="fixed".
799 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld, rescanning for anchored from offset %ld...\n",
800 PL_colors[0], PL_colors[1], (long)(strpos - i_strpos), (long)(strpos - i_strpos + prog->anchored_offset)));
801 goto do_other_anchored;
803 /* We don't contradict the found floating substring. */
804 /* XXXX Why not check for STCLASS? */
806 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld...\n",
807 PL_colors[0], PL_colors[1], (long)(s - i_strpos)));
810 /* Position contradicts check-string */
811 /* XXXX probably better to look for check-string
812 than for "\n", so one should lower the limit for t? */
813 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m, restarting lookup for check-string at offset %ld...\n",
814 PL_colors[0], PL_colors[1], (long)(t + 1 - i_strpos)));
815 other_last = strpos = s = t + 1;
820 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Did not find /%s^%s/m...\n",
821 PL_colors[0], PL_colors[1]));
825 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Starting position does not contradict /%s^%s/m...\n",
826 PL_colors[0], PL_colors[1]));
830 ++BmUSEFUL(do_utf8 ? prog->check_utf8 : prog->check_substr); /* hooray/5 */
833 /* The found string does not prohibit matching at strpos,
834 - no optimization of calling REx engine can be performed,
835 unless it was an MBOL and we are not after MBOL,
836 or a future STCLASS check will fail this. */
838 /* Even in this situation we may use MBOL flag if strpos is offset
839 wrt the start of the string. */
840 if (ml_anch && sv && !SvROK(sv) /* See prev comment on SvROK */
841 && (strpos != strbeg) && strpos[-1] != '\n'
842 /* May be due to an implicit anchor of m{.*foo} */
843 && !(prog->intflags & PREGf_IMPLICIT))
848 DEBUG_EXECUTE_r( if (ml_anch)
849 PerlIO_printf(Perl_debug_log, "Position at offset %ld does not contradict /%s^%s/m...\n",
850 (long)(strpos - i_strpos), PL_colors[0], PL_colors[1]);
853 if (!(prog->intflags & PREGf_NAUGHTY) /* XXXX If strpos moved? */
855 prog->check_utf8 /* Could be deleted already */
856 && --BmUSEFUL(prog->check_utf8) < 0
857 && (prog->check_utf8 == prog->float_utf8)
859 prog->check_substr /* Could be deleted already */
860 && --BmUSEFUL(prog->check_substr) < 0
861 && (prog->check_substr == prog->float_substr)
864 /* If flags & SOMETHING - do not do it many times on the same match */
865 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "... Disabling check substring...\n"));
866 SvREFCNT_dec(do_utf8 ? prog->check_utf8 : prog->check_substr);
867 if (do_utf8 ? prog->check_substr : prog->check_utf8)
868 SvREFCNT_dec(do_utf8 ? prog->check_substr : prog->check_utf8);
869 prog->check_substr = prog->check_utf8 = NULL; /* disable */
870 prog->float_substr = prog->float_utf8 = NULL; /* clear */
871 check = NULL; /* abort */
873 /* XXXX This is a remnant of the old implementation. It
874 looks wasteful, since now INTUIT can use many
876 prog->extflags &= ~RXf_USE_INTUIT;
883 /* XXXX BmUSEFUL already changed, maybe multiple change is meaningful... */
884 /* trie stclasses are too expensive to use here, we are better off to
885 leave it to regmatch itself */
886 if (progi->regstclass && PL_regkind[OP(progi->regstclass)]!=TRIE) {
887 /* minlen == 0 is possible if regstclass is \b or \B,
888 and the fixed substr is ''$.
889 Since minlen is already taken into account, s+1 is before strend;
890 accidentally, minlen >= 1 guaranties no false positives at s + 1
891 even for \b or \B. But (minlen? 1 : 0) below assumes that
892 regstclass does not come from lookahead... */
893 /* If regstclass takes bytelength more than 1: If charlength==1, OK.
894 This leaves EXACTF only, which is dealt with in find_byclass(). */
895 const U8* const str = (U8*)STRING(progi->regstclass);
896 const int cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
897 ? CHR_DIST(str+STR_LEN(progi->regstclass), str)
900 if (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
901 endpos= HOP3c(s, (prog->minlen ? cl_l : 0), strend);
902 else if (prog->float_substr || prog->float_utf8)
903 endpos= HOP3c(HOP3c(check_at, -start_shift, strbeg), cl_l, strend);
907 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "start_shift: %"IVdf" check_at: %"IVdf" s: %"IVdf" endpos: %"IVdf"\n",
908 (IV)start_shift, (IV)(check_at - strbeg), (IV)(s - strbeg), (IV)(endpos - strbeg)));
911 s = find_byclass(prog, progi->regstclass, s, endpos, NULL);
914 const char *what = NULL;
916 if (endpos == strend) {
917 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
918 "Could not match STCLASS...\n") );
921 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
922 "This position contradicts STCLASS...\n") );
923 if ((prog->extflags & RXf_ANCH) && !ml_anch)
925 /* Contradict one of substrings */
926 if (prog->anchored_substr || prog->anchored_utf8) {
927 if ((do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) == check) {
928 DEBUG_EXECUTE_r( what = "anchored" );
930 s = HOP3c(t, 1, strend);
931 if (s + start_shift + end_shift > strend) {
932 /* XXXX Should be taken into account earlier? */
933 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
934 "Could not match STCLASS...\n") );
939 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
940 "Looking for %s substr starting at offset %ld...\n",
941 what, (long)(s + start_shift - i_strpos)) );
944 /* Have both, check_string is floating */
945 if (t + start_shift >= check_at) /* Contradicts floating=check */
946 goto retry_floating_check;
947 /* Recheck anchored substring, but not floating... */
951 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
952 "Looking for anchored substr starting at offset %ld...\n",
953 (long)(other_last - i_strpos)) );
954 goto do_other_anchored;
956 /* Another way we could have checked stclass at the
957 current position only: */
962 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
963 "Looking for /%s^%s/m starting at offset %ld...\n",
964 PL_colors[0], PL_colors[1], (long)(t - i_strpos)) );
967 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr)) /* Could have been deleted */
969 /* Check is floating subtring. */
970 retry_floating_check:
971 t = check_at - start_shift;
972 DEBUG_EXECUTE_r( what = "floating" );
973 goto hop_and_restart;
976 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
977 "By STCLASS: moving %ld --> %ld\n",
978 (long)(t - i_strpos), (long)(s - i_strpos))
982 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
983 "Does not contradict STCLASS...\n");
988 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s%s:%s match at offset %ld\n",
989 PL_colors[4], (check ? "Guessed" : "Giving up"),
990 PL_colors[5], (long)(s - i_strpos)) );
993 fail_finish: /* Substring not found */
994 if (prog->check_substr || prog->check_utf8) /* could be removed already */
995 BmUSEFUL(do_utf8 ? prog->check_utf8 : prog->check_substr) += 5; /* hooray */
997 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch rejected by optimizer%s\n",
998 PL_colors[4], PL_colors[5]));
1002 #define DECL_TRIE_TYPE(scan) \
1003 const enum { trie_plain, trie_utf8, trie_utf8_fold, trie_latin_utf8_fold } \
1004 trie_type = (scan->flags != EXACT) \
1005 ? (do_utf8 ? trie_utf8_fold : (UTF ? trie_latin_utf8_fold : trie_plain)) \
1006 : (do_utf8 ? trie_utf8 : trie_plain)
1008 #define REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc, uscan, len, \
1009 uvc, charid, foldlen, foldbuf, uniflags) STMT_START { \
1010 switch (trie_type) { \
1011 case trie_utf8_fold: \
1012 if ( foldlen>0 ) { \
1013 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1018 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1019 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1020 foldlen -= UNISKIP( uvc ); \
1021 uscan = foldbuf + UNISKIP( uvc ); \
1024 case trie_latin_utf8_fold: \
1025 if ( foldlen>0 ) { \
1026 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1032 uvc = to_uni_fold( *(U8*)uc, foldbuf, &foldlen ); \
1033 foldlen -= UNISKIP( uvc ); \
1034 uscan = foldbuf + UNISKIP( uvc ); \
1038 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1046 charid = trie->charmap[ uvc ]; \
1050 if (widecharmap) { \
1051 SV** const svpp = hv_fetch(widecharmap, \
1052 (char*)&uvc, sizeof(UV), 0); \
1054 charid = (U16)SvIV(*svpp); \
1059 #define REXEC_FBC_EXACTISH_CHECK(CoNd) \
1061 char *my_strend= (char *)strend; \
1064 !ibcmp_utf8(s, &my_strend, 0, do_utf8, \
1065 m, NULL, ln, (bool)UTF)) \
1066 && (!reginfo || regtry(reginfo, &s)) ) \
1069 U8 foldbuf[UTF8_MAXBYTES_CASE+1]; \
1070 uvchr_to_utf8(tmpbuf, c); \
1071 f = to_utf8_fold(tmpbuf, foldbuf, &foldlen); \
1073 && (f == c1 || f == c2) \
1075 !ibcmp_utf8(s, &my_strend, 0, do_utf8,\
1076 m, NULL, ln, (bool)UTF)) \
1077 && (!reginfo || regtry(reginfo, &s)) ) \
1083 #define REXEC_FBC_EXACTISH_SCAN(CoNd) \
1087 && (ln == 1 || !(OP(c) == EXACTF \
1089 : ibcmp_locale(s, m, ln))) \
1090 && (!reginfo || regtry(reginfo, &s)) ) \
1096 #define REXEC_FBC_UTF8_SCAN(CoDe) \
1098 while (s + (uskip = UTF8SKIP(s)) <= strend) { \
1104 #define REXEC_FBC_SCAN(CoDe) \
1106 while (s < strend) { \
1112 #define REXEC_FBC_UTF8_CLASS_SCAN(CoNd) \
1113 REXEC_FBC_UTF8_SCAN( \
1115 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1124 #define REXEC_FBC_CLASS_SCAN(CoNd) \
1127 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1136 #define REXEC_FBC_TRYIT \
1137 if ((!reginfo || regtry(reginfo, &s))) \
1140 #define REXEC_FBC_CSCAN(CoNdUtF8,CoNd) \
1142 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1145 REXEC_FBC_CLASS_SCAN(CoNd); \
1149 #define REXEC_FBC_CSCAN_PRELOAD(UtFpReLoAd,CoNdUtF8,CoNd) \
1152 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1155 REXEC_FBC_CLASS_SCAN(CoNd); \
1159 #define REXEC_FBC_CSCAN_TAINT(CoNdUtF8,CoNd) \
1160 PL_reg_flags |= RF_tainted; \
1162 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1165 REXEC_FBC_CLASS_SCAN(CoNd); \
1169 #define DUMP_EXEC_POS(li,s,doutf8) \
1170 dump_exec_pos(li,s,(PL_regeol),(PL_bostr),(PL_reg_starttry),doutf8)
1172 /* We know what class REx starts with. Try to find this position... */
1173 /* if reginfo is NULL, its a dryrun */
1174 /* annoyingly all the vars in this routine have different names from their counterparts
1175 in regmatch. /grrr */
1178 S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s,
1179 const char *strend, regmatch_info *reginfo)
1182 const I32 doevery = (prog->intflags & PREGf_SKIP) == 0;
1186 register STRLEN uskip;
1190 register I32 tmp = 1; /* Scratch variable? */
1191 register const bool do_utf8 = PL_reg_match_utf8;
1192 RXi_GET_DECL(prog,progi);
1194 PERL_ARGS_ASSERT_FIND_BYCLASS;
1196 /* We know what class it must start with. */
1200 REXEC_FBC_UTF8_CLASS_SCAN((ANYOF_FLAGS(c) & ANYOF_UNICODE) ||
1201 !UTF8_IS_INVARIANT((U8)s[0]) ?
1202 reginclass(prog, c, (U8*)s, 0, do_utf8) :
1203 REGINCLASS(prog, c, (U8*)s));
1206 while (s < strend) {
1209 if (REGINCLASS(prog, c, (U8*)s) ||
1210 (ANYOF_FOLD_SHARP_S(c, s, strend) &&
1211 /* The assignment of 2 is intentional:
1212 * for the folded sharp s, the skip is 2. */
1213 (skip = SHARP_S_SKIP))) {
1214 if (tmp && (!reginfo || regtry(reginfo, &s)))
1227 if (tmp && (!reginfo || regtry(reginfo, &s)))
1235 ln = STR_LEN(c); /* length to match in octets/bytes */
1236 lnc = (I32) ln; /* length to match in characters */
1238 STRLEN ulen1, ulen2;
1240 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
1241 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
1242 /* used by commented-out code below */
1243 /*const U32 uniflags = UTF8_ALLOW_DEFAULT;*/
1245 /* XXX: Since the node will be case folded at compile
1246 time this logic is a little odd, although im not
1247 sure that its actually wrong. --dmq */
1249 c1 = to_utf8_lower((U8*)m, tmpbuf1, &ulen1);
1250 c2 = to_utf8_upper((U8*)m, tmpbuf2, &ulen2);
1252 /* XXX: This is kinda strange. to_utf8_XYZ returns the
1253 codepoint of the first character in the converted
1254 form, yet originally we did the extra step.
1255 No tests fail by commenting this code out however
1256 so Ive left it out. -- dmq.
1258 c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXBYTES_CASE,
1260 c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXBYTES_CASE,
1265 while (sm < ((U8 *) m + ln)) {
1280 c2 = PL_fold_locale[c1];
1282 e = HOP3c(strend, -((I32)lnc), s);
1284 if (!reginfo && e < s)
1285 e = s; /* Due to minlen logic of intuit() */
1287 /* The idea in the EXACTF* cases is to first find the
1288 * first character of the EXACTF* node and then, if
1289 * necessary, case-insensitively compare the full
1290 * text of the node. The c1 and c2 are the first
1291 * characters (though in Unicode it gets a bit
1292 * more complicated because there are more cases
1293 * than just upper and lower: one needs to use
1294 * the so-called folding case for case-insensitive
1295 * matching (called "loose matching" in Unicode).
1296 * ibcmp_utf8() will do just that. */
1298 if (do_utf8 || UTF) {
1300 U8 tmpbuf [UTF8_MAXBYTES+1];
1303 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1305 /* Upper and lower of 1st char are equal -
1306 * probably not a "letter". */
1309 c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
1314 REXEC_FBC_EXACTISH_CHECK(c == c1);
1320 c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
1326 /* Handle some of the three Greek sigmas cases.
1327 * Note that not all the possible combinations
1328 * are handled here: some of them are handled
1329 * by the standard folding rules, and some of
1330 * them (the character class or ANYOF cases)
1331 * are handled during compiletime in
1332 * regexec.c:S_regclass(). */
1333 if (c == (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA ||
1334 c == (UV)UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA)
1335 c = (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA;
1337 REXEC_FBC_EXACTISH_CHECK(c == c1 || c == c2);
1342 /* Neither pattern nor string are UTF8 */
1344 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1);
1346 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1 || *(U8*)s == c2);
1350 PL_reg_flags |= RF_tainted;
1357 U8 * const r = reghop3((U8*)s, -1, (U8*)PL_bostr);
1358 tmp = utf8n_to_uvchr(r, UTF8SKIP(r), 0, UTF8_ALLOW_DEFAULT);
1360 tmp = ((OP(c) == BOUND ?
1361 isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
1362 LOAD_UTF8_CHARCLASS_ALNUM();
1363 REXEC_FBC_UTF8_SCAN(
1364 if (tmp == !(OP(c) == BOUND ?
1365 (bool)swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
1366 isALNUM_LC_utf8((U8*)s)))
1374 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n';
1375 tmp = ((OP(c) == BOUND ? isALNUM(tmp) : isALNUM_LC(tmp)) != 0);
1378 !(OP(c) == BOUND ? isALNUM(*s) : isALNUM_LC(*s))) {
1384 if ((!prog->minlen && tmp) && (!reginfo || regtry(reginfo, &s)))
1388 PL_reg_flags |= RF_tainted;
1395 U8 * const r = reghop3((U8*)s, -1, (U8*)PL_bostr);
1396 tmp = utf8n_to_uvchr(r, UTF8SKIP(r), 0, UTF8_ALLOW_DEFAULT);
1398 tmp = ((OP(c) == NBOUND ?
1399 isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
1400 LOAD_UTF8_CHARCLASS_ALNUM();
1401 REXEC_FBC_UTF8_SCAN(
1402 if (tmp == !(OP(c) == NBOUND ?
1403 (bool)swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
1404 isALNUM_LC_utf8((U8*)s)))
1406 else REXEC_FBC_TRYIT;
1410 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n';
1411 tmp = ((OP(c) == NBOUND ?
1412 isALNUM(tmp) : isALNUM_LC(tmp)) != 0);
1415 !(OP(c) == NBOUND ? isALNUM(*s) : isALNUM_LC(*s)))
1417 else REXEC_FBC_TRYIT;
1420 if ((!prog->minlen && !tmp) && (!reginfo || regtry(reginfo, &s)))
1424 REXEC_FBC_CSCAN_PRELOAD(
1425 LOAD_UTF8_CHARCLASS_ALNUM(),
1426 swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8),
1430 REXEC_FBC_CSCAN_TAINT(
1431 isALNUM_LC_utf8((U8*)s),
1435 REXEC_FBC_CSCAN_PRELOAD(
1436 LOAD_UTF8_CHARCLASS_ALNUM(),
1437 !swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8),
1441 REXEC_FBC_CSCAN_TAINT(
1442 !isALNUM_LC_utf8((U8*)s),
1446 REXEC_FBC_CSCAN_PRELOAD(
1447 LOAD_UTF8_CHARCLASS_SPACE(),
1448 *s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8),
1452 REXEC_FBC_CSCAN_TAINT(
1453 *s == ' ' || isSPACE_LC_utf8((U8*)s),
1457 REXEC_FBC_CSCAN_PRELOAD(
1458 LOAD_UTF8_CHARCLASS_SPACE(),
1459 !(*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8)),
1463 REXEC_FBC_CSCAN_TAINT(
1464 !(*s == ' ' || isSPACE_LC_utf8((U8*)s)),
1468 REXEC_FBC_CSCAN_PRELOAD(
1469 LOAD_UTF8_CHARCLASS_DIGIT(),
1470 swash_fetch(PL_utf8_digit,(U8*)s, do_utf8),
1474 REXEC_FBC_CSCAN_TAINT(
1475 isDIGIT_LC_utf8((U8*)s),
1479 REXEC_FBC_CSCAN_PRELOAD(
1480 LOAD_UTF8_CHARCLASS_DIGIT(),
1481 !swash_fetch(PL_utf8_digit,(U8*)s, do_utf8),
1485 REXEC_FBC_CSCAN_TAINT(
1486 !isDIGIT_LC_utf8((U8*)s),
1492 is_LNBREAK_latin1(s)
1502 !is_VERTWS_latin1(s)
1507 is_HORIZWS_latin1(s)
1511 !is_HORIZWS_utf8(s),
1512 !is_HORIZWS_latin1(s)
1518 /* what trie are we using right now */
1520 = (reg_ac_data*)progi->data->data[ ARG( c ) ];
1522 = (reg_trie_data*)progi->data->data[ aho->trie ];
1523 HV *widecharmap = MUTABLE_HV(progi->data->data[ aho->trie + 1 ]);
1525 const char *last_start = strend - trie->minlen;
1527 const char *real_start = s;
1529 STRLEN maxlen = trie->maxlen;
1531 U8 **points; /* map of where we were in the input string
1532 when reading a given char. For ASCII this
1533 is unnecessary overhead as the relationship
1534 is always 1:1, but for Unicode, especially
1535 case folded Unicode this is not true. */
1536 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1540 GET_RE_DEBUG_FLAGS_DECL;
1542 /* We can't just allocate points here. We need to wrap it in
1543 * an SV so it gets freed properly if there is a croak while
1544 * running the match */
1547 sv_points=newSV(maxlen * sizeof(U8 *));
1548 SvCUR_set(sv_points,
1549 maxlen * sizeof(U8 *));
1550 SvPOK_on(sv_points);
1551 sv_2mortal(sv_points);
1552 points=(U8**)SvPV_nolen(sv_points );
1553 if ( trie_type != trie_utf8_fold
1554 && (trie->bitmap || OP(c)==AHOCORASICKC) )
1557 bitmap=(U8*)trie->bitmap;
1559 bitmap=(U8*)ANYOF_BITMAP(c);
1561 /* this is the Aho-Corasick algorithm modified a touch
1562 to include special handling for long "unknown char"
1563 sequences. The basic idea being that we use AC as long
1564 as we are dealing with a possible matching char, when
1565 we encounter an unknown char (and we have not encountered
1566 an accepting state) we scan forward until we find a legal
1568 AC matching is basically that of trie matching, except
1569 that when we encounter a failing transition, we fall back
1570 to the current states "fail state", and try the current char
1571 again, a process we repeat until we reach the root state,
1572 state 1, or a legal transition. If we fail on the root state
1573 then we can either terminate if we have reached an accepting
1574 state previously, or restart the entire process from the beginning
1578 while (s <= last_start) {
1579 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1587 U8 *uscan = (U8*)NULL;
1588 U8 *leftmost = NULL;
1590 U32 accepted_word= 0;
1594 while ( state && uc <= (U8*)strend ) {
1596 U32 word = aho->states[ state ].wordnum;
1600 DEBUG_TRIE_EXECUTE_r(
1601 if ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1602 dump_exec_pos( (char *)uc, c, strend, real_start,
1603 (char *)uc, do_utf8 );
1604 PerlIO_printf( Perl_debug_log,
1605 " Scanning for legal start char...\n");
1608 while ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1613 if (uc >(U8*)last_start) break;
1617 U8 *lpos= points[ (pointpos - trie->wordlen[word-1] ) % maxlen ];
1618 if (!leftmost || lpos < leftmost) {
1619 DEBUG_r(accepted_word=word);
1625 points[pointpos++ % maxlen]= uc;
1626 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
1627 uscan, len, uvc, charid, foldlen,
1629 DEBUG_TRIE_EXECUTE_r({
1630 dump_exec_pos( (char *)uc, c, strend, real_start,
1632 PerlIO_printf(Perl_debug_log,
1633 " Charid:%3u CP:%4"UVxf" ",
1639 word = aho->states[ state ].wordnum;
1641 base = aho->states[ state ].trans.base;
1643 DEBUG_TRIE_EXECUTE_r({
1645 dump_exec_pos( (char *)uc, c, strend, real_start,
1647 PerlIO_printf( Perl_debug_log,
1648 "%sState: %4"UVxf", word=%"UVxf,
1649 failed ? " Fail transition to " : "",
1650 (UV)state, (UV)word);
1655 (base + charid > trie->uniquecharcount )
1656 && (base + charid - 1 - trie->uniquecharcount
1658 && trie->trans[base + charid - 1 -
1659 trie->uniquecharcount].check == state
1660 && (tmp=trie->trans[base + charid - 1 -
1661 trie->uniquecharcount ].next))
1663 DEBUG_TRIE_EXECUTE_r(
1664 PerlIO_printf( Perl_debug_log," - legal\n"));
1669 DEBUG_TRIE_EXECUTE_r(
1670 PerlIO_printf( Perl_debug_log," - fail\n"));
1672 state = aho->fail[state];
1676 /* we must be accepting here */
1677 DEBUG_TRIE_EXECUTE_r(
1678 PerlIO_printf( Perl_debug_log," - accepting\n"));
1687 if (!state) state = 1;
1690 if ( aho->states[ state ].wordnum ) {
1691 U8 *lpos = points[ (pointpos - trie->wordlen[aho->states[ state ].wordnum-1]) % maxlen ];
1692 if (!leftmost || lpos < leftmost) {
1693 DEBUG_r(accepted_word=aho->states[ state ].wordnum);
1698 s = (char*)leftmost;
1699 DEBUG_TRIE_EXECUTE_r({
1701 Perl_debug_log,"Matches word #%"UVxf" at position %"IVdf". Trying full pattern...\n",
1702 (UV)accepted_word, (IV)(s - real_start)
1705 if (!reginfo || regtry(reginfo, &s)) {
1711 DEBUG_TRIE_EXECUTE_r({
1712 PerlIO_printf( Perl_debug_log,"Pattern failed. Looking for new start point...\n");
1715 DEBUG_TRIE_EXECUTE_r(
1716 PerlIO_printf( Perl_debug_log,"No match.\n"));
1725 Perl_croak(aTHX_ "panic: unknown regstclass %d", (int)OP(c));
1734 S_swap_match_buff (pTHX_ regexp *prog)
1736 regexp_paren_pair *t;
1738 PERL_ARGS_ASSERT_SWAP_MATCH_BUFF;
1741 /* We have to be careful. If the previous successful match
1742 was from this regex we don't want a subsequent paritally
1743 successful match to clobber the old results.
1744 So when we detect this possibility we add a swap buffer
1745 to the re, and switch the buffer each match. If we fail
1746 we switch it back, otherwise we leave it swapped.
1748 Newxz(prog->swap, (prog->nparens + 1), regexp_paren_pair);
1751 prog->swap = prog->offs;
1757 - regexec_flags - match a regexp against a string
1760 Perl_regexec_flags(pTHX_ REGEXP * const rx, char *stringarg, register char *strend,
1761 char *strbeg, I32 minend, SV *sv, void *data, U32 flags)
1762 /* strend: pointer to null at end of string */
1763 /* strbeg: real beginning of string */
1764 /* minend: end of match must be >=minend after stringarg. */
1765 /* data: May be used for some additional optimizations.
1766 Currently its only used, with a U32 cast, for transmitting
1767 the ganch offset when doing a /g match. This will change */
1768 /* nosave: For optimizations. */
1771 struct regexp *const prog = (struct regexp *)SvANY(rx);
1772 /*register*/ char *s;
1773 register regnode *c;
1774 /*register*/ char *startpos = stringarg;
1775 I32 minlen; /* must match at least this many chars */
1776 I32 dontbother = 0; /* how many characters not to try at end */
1777 I32 end_shift = 0; /* Same for the end. */ /* CC */
1778 I32 scream_pos = -1; /* Internal iterator of scream. */
1779 char *scream_olds = NULL;
1780 const bool do_utf8 = (bool)DO_UTF8(sv);
1782 RXi_GET_DECL(prog,progi);
1783 regmatch_info reginfo; /* create some info to pass to regtry etc */
1784 bool swap_on_fail = 0;
1785 GET_RE_DEBUG_FLAGS_DECL;
1787 PERL_ARGS_ASSERT_REGEXEC_FLAGS;
1788 PERL_UNUSED_ARG(data);
1790 /* Be paranoid... */
1791 if (prog == NULL || startpos == NULL) {
1792 Perl_croak(aTHX_ "NULL regexp parameter");
1796 multiline = prog->extflags & RXf_PMf_MULTILINE;
1797 reginfo.prog = rx; /* Yes, sorry that this is confusing. */
1799 RX_MATCH_UTF8_set(rx, do_utf8);
1801 debug_start_match(rx, do_utf8, startpos, strend,
1805 minlen = prog->minlen;
1807 if (strend - startpos < (minlen+(prog->check_offset_min<0?prog->check_offset_min:0))) {
1808 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
1809 "String too short [regexec_flags]...\n"));
1814 /* Check validity of program. */
1815 if (UCHARAT(progi->program) != REG_MAGIC) {
1816 Perl_croak(aTHX_ "corrupted regexp program");
1820 PL_reg_eval_set = 0;
1824 PL_reg_flags |= RF_utf8;
1826 /* Mark beginning of line for ^ and lookbehind. */
1827 reginfo.bol = startpos; /* XXX not used ??? */
1831 /* Mark end of line for $ (and such) */
1834 /* see how far we have to get to not match where we matched before */
1835 reginfo.till = startpos+minend;
1837 /* If there is a "must appear" string, look for it. */
1840 if (prog->extflags & RXf_GPOS_SEEN) { /* Need to set reginfo->ganch */
1843 if (flags & REXEC_IGNOREPOS) /* Means: check only at start */
1844 reginfo.ganch = startpos + prog->gofs;
1845 else if (sv && SvTYPE(sv) >= SVt_PVMG
1847 && (mg = mg_find(sv, PERL_MAGIC_regex_global))
1848 && mg->mg_len >= 0) {
1849 reginfo.ganch = strbeg + mg->mg_len; /* Defined pos() */
1850 if (prog->extflags & RXf_ANCH_GPOS) {
1851 if (s > reginfo.ganch)
1853 s = reginfo.ganch - prog->gofs;
1857 reginfo.ganch = strbeg + PTR2UV(data);
1858 } else /* pos() not defined */
1859 reginfo.ganch = strbeg;
1861 if (PL_curpm && (PM_GETRE(PL_curpm) == rx)) {
1863 swap_match_buff(prog); /* do we need a save destructor here for
1866 if (!(flags & REXEC_CHECKED) && (prog->check_substr != NULL || prog->check_utf8 != NULL)) {
1867 re_scream_pos_data d;
1869 d.scream_olds = &scream_olds;
1870 d.scream_pos = &scream_pos;
1871 s = re_intuit_start(rx, sv, s, strend, flags, &d);
1873 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not present...\n"));
1874 goto phooey; /* not present */
1880 /* Simplest case: anchored match need be tried only once. */
1881 /* [unless only anchor is BOL and multiline is set] */
1882 if (prog->extflags & (RXf_ANCH & ~RXf_ANCH_GPOS)) {
1883 if (s == startpos && regtry(®info, &startpos))
1885 else if (multiline || (prog->intflags & PREGf_IMPLICIT)
1886 || (prog->extflags & RXf_ANCH_MBOL)) /* XXXX SBOL? */
1891 dontbother = minlen - 1;
1892 end = HOP3c(strend, -dontbother, strbeg) - 1;
1893 /* for multiline we only have to try after newlines */
1894 if (prog->check_substr || prog->check_utf8) {
1898 if (regtry(®info, &s))
1903 if (prog->extflags & RXf_USE_INTUIT) {
1904 s = re_intuit_start(rx, sv, s + 1, strend, flags, NULL);
1915 if (*s++ == '\n') { /* don't need PL_utf8skip here */
1916 if (regtry(®info, &s))
1923 } else if (RXf_GPOS_CHECK == (prog->extflags & RXf_GPOS_CHECK))
1925 /* the warning about reginfo.ganch being used without intialization
1926 is bogus -- we set it above, when prog->extflags & RXf_GPOS_SEEN
1927 and we only enter this block when the same bit is set. */
1928 char *tmp_s = reginfo.ganch - prog->gofs;
1929 if (regtry(®info, &tmp_s))
1934 /* Messy cases: unanchored match. */
1935 if ((prog->anchored_substr || prog->anchored_utf8) && prog->intflags & PREGf_SKIP) {
1936 /* we have /x+whatever/ */
1937 /* it must be a one character string (XXXX Except UTF?) */
1942 if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr))
1943 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1944 ch = SvPVX_const(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0];
1949 DEBUG_EXECUTE_r( did_match = 1 );
1950 if (regtry(®info, &s)) goto got_it;
1952 while (s < strend && *s == ch)
1960 DEBUG_EXECUTE_r( did_match = 1 );
1961 if (regtry(®info, &s)) goto got_it;
1963 while (s < strend && *s == ch)
1968 DEBUG_EXECUTE_r(if (!did_match)
1969 PerlIO_printf(Perl_debug_log,
1970 "Did not find anchored character...\n")
1973 else if (prog->anchored_substr != NULL
1974 || prog->anchored_utf8 != NULL
1975 || ((prog->float_substr != NULL || prog->float_utf8 != NULL)
1976 && prog->float_max_offset < strend - s)) {
1981 char *last1; /* Last position checked before */
1985 if (prog->anchored_substr || prog->anchored_utf8) {
1986 if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr))
1987 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1988 must = do_utf8 ? prog->anchored_utf8 : prog->anchored_substr;
1989 back_max = back_min = prog->anchored_offset;
1991 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr))
1992 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
1993 must = do_utf8 ? prog->float_utf8 : prog->float_substr;
1994 back_max = prog->float_max_offset;
1995 back_min = prog->float_min_offset;
1999 if (must == &PL_sv_undef)
2000 /* could not downgrade utf8 check substring, so must fail */
2006 last = HOP3c(strend, /* Cannot start after this */
2007 -(I32)(CHR_SVLEN(must)
2008 - (SvTAIL(must) != 0) + back_min), strbeg);
2011 last1 = HOPc(s, -1);
2013 last1 = s - 1; /* bogus */
2015 /* XXXX check_substr already used to find "s", can optimize if
2016 check_substr==must. */
2018 dontbother = end_shift;
2019 strend = HOPc(strend, -dontbother);
2020 while ( (s <= last) &&
2021 ((flags & REXEC_SCREAM)
2022 ? (s = screaminstr(sv, must, HOP3c(s, back_min, (back_min<0 ? strbeg : strend)) - strbeg,
2023 end_shift, &scream_pos, 0))
2024 : (s = fbm_instr((unsigned char*)HOP3(s, back_min, (back_min<0 ? strbeg : strend)),
2025 (unsigned char*)strend, must,
2026 multiline ? FBMrf_MULTILINE : 0))) ) {
2027 /* we may be pointing at the wrong string */
2028 if ((flags & REXEC_SCREAM) && RXp_MATCH_COPIED(prog))
2029 s = strbeg + (s - SvPVX_const(sv));
2030 DEBUG_EXECUTE_r( did_match = 1 );
2031 if (HOPc(s, -back_max) > last1) {
2032 last1 = HOPc(s, -back_min);
2033 s = HOPc(s, -back_max);
2036 char * const t = (last1 >= PL_bostr) ? HOPc(last1, 1) : last1 + 1;
2038 last1 = HOPc(s, -back_min);
2042 while (s <= last1) {
2043 if (regtry(®info, &s))
2049 while (s <= last1) {
2050 if (regtry(®info, &s))
2056 DEBUG_EXECUTE_r(if (!did_match) {
2057 RE_PV_QUOTED_DECL(quoted, do_utf8, PERL_DEBUG_PAD_ZERO(0),
2058 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
2059 PerlIO_printf(Perl_debug_log, "Did not find %s substr %s%s...\n",
2060 ((must == prog->anchored_substr || must == prog->anchored_utf8)
2061 ? "anchored" : "floating"),
2062 quoted, RE_SV_TAIL(must));
2066 else if ( (c = progi->regstclass) ) {
2068 const OPCODE op = OP(progi->regstclass);
2069 /* don't bother with what can't match */
2070 if (PL_regkind[op] != EXACT && op != CANY && PL_regkind[op] != TRIE)
2071 strend = HOPc(strend, -(minlen - 1));
2074 SV * const prop = sv_newmortal();
2075 regprop(prog, prop, c);
2077 RE_PV_QUOTED_DECL(quoted,do_utf8,PERL_DEBUG_PAD_ZERO(1),
2079 PerlIO_printf(Perl_debug_log,
2080 "Matching stclass %.*s against %s (%d chars)\n",
2081 (int)SvCUR(prop), SvPVX_const(prop),
2082 quoted, (int)(strend - s));
2085 if (find_byclass(prog, c, s, strend, ®info))
2087 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass... [regexec_flags]\n"));
2091 if (prog->float_substr != NULL || prog->float_utf8 != NULL) {
2096 if (!(do_utf8 ? prog->float_utf8 : prog->float_substr))
2097 do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
2098 float_real = do_utf8 ? prog->float_utf8 : prog->float_substr;
2100 if (flags & REXEC_SCREAM) {
2101 last = screaminstr(sv, float_real, s - strbeg,
2102 end_shift, &scream_pos, 1); /* last one */
2104 last = scream_olds; /* Only one occurrence. */
2105 /* we may be pointing at the wrong string */
2106 else if (RXp_MATCH_COPIED(prog))
2107 s = strbeg + (s - SvPVX_const(sv));
2111 const char * const little = SvPV_const(float_real, len);
2113 if (SvTAIL(float_real)) {
2114 if (memEQ(strend - len + 1, little, len - 1))
2115 last = strend - len + 1;
2116 else if (!multiline)
2117 last = memEQ(strend - len, little, len)
2118 ? strend - len : NULL;
2124 last = rninstr(s, strend, little, little + len);
2126 last = strend; /* matching "$" */
2131 PerlIO_printf(Perl_debug_log,
2132 "%sCan't trim the tail, match fails (should not happen)%s\n",
2133 PL_colors[4], PL_colors[5]));
2134 goto phooey; /* Should not happen! */
2136 dontbother = strend - last + prog->float_min_offset;
2138 if (minlen && (dontbother < minlen))
2139 dontbother = minlen - 1;
2140 strend -= dontbother; /* this one's always in bytes! */
2141 /* We don't know much -- general case. */
2144 if (regtry(®info, &s))
2153 if (regtry(®info, &s))
2155 } while (s++ < strend);
2163 RX_MATCH_TAINTED_set(rx, PL_reg_flags & RF_tainted);
2165 if (PL_reg_eval_set)
2166 restore_pos(aTHX_ prog);
2167 if (RXp_PAREN_NAMES(prog))
2168 (void)hv_iterinit(RXp_PAREN_NAMES(prog));
2170 /* make sure $`, $&, $', and $digit will work later */
2171 if ( !(flags & REXEC_NOT_FIRST) ) {
2172 RX_MATCH_COPY_FREE(rx);
2173 if (flags & REXEC_COPY_STR) {
2174 const I32 i = PL_regeol - startpos + (stringarg - strbeg);
2175 #ifdef PERL_OLD_COPY_ON_WRITE
2177 || (SvFLAGS(sv) & CAN_COW_MASK) == CAN_COW_FLAGS)) {
2179 PerlIO_printf(Perl_debug_log,
2180 "Copy on write: regexp capture, type %d\n",
2183 prog->saved_copy = sv_setsv_cow(prog->saved_copy, sv);
2184 prog->subbeg = (char *)SvPVX_const(prog->saved_copy);
2185 assert (SvPOKp(prog->saved_copy));
2189 RX_MATCH_COPIED_on(rx);
2190 s = savepvn(strbeg, i);
2196 prog->subbeg = strbeg;
2197 prog->sublen = PL_regeol - strbeg; /* strend may have been modified */
2204 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch failed%s\n",
2205 PL_colors[4], PL_colors[5]));
2206 if (PL_reg_eval_set)
2207 restore_pos(aTHX_ prog);
2209 /* we failed :-( roll it back */
2210 swap_match_buff(prog);
2217 - regtry - try match at specific point
2219 STATIC I32 /* 0 failure, 1 success */
2220 S_regtry(pTHX_ regmatch_info *reginfo, char **startpos)
2224 REGEXP *const rx = reginfo->prog;
2225 regexp *const prog = (struct regexp *)SvANY(rx);
2226 RXi_GET_DECL(prog,progi);
2227 GET_RE_DEBUG_FLAGS_DECL;
2229 PERL_ARGS_ASSERT_REGTRY;
2231 reginfo->cutpoint=NULL;
2233 if ((prog->extflags & RXf_EVAL_SEEN) && !PL_reg_eval_set) {
2236 PL_reg_eval_set = RS_init;
2237 DEBUG_EXECUTE_r(DEBUG_s(
2238 PerlIO_printf(Perl_debug_log, " setting stack tmpbase at %"IVdf"\n",
2239 (IV)(PL_stack_sp - PL_stack_base));
2242 cxstack[cxstack_ix].blk_oldsp = PL_stack_sp - PL_stack_base;
2243 /* Otherwise OP_NEXTSTATE will free whatever on stack now. */
2245 /* Apparently this is not needed, judging by wantarray. */
2246 /* SAVEI8(cxstack[cxstack_ix].blk_gimme);
2247 cxstack[cxstack_ix].blk_gimme = G_SCALAR; */
2250 /* Make $_ available to executed code. */
2251 if (reginfo->sv != DEFSV) {
2253 DEFSV_set(reginfo->sv);
2256 if (!(SvTYPE(reginfo->sv) >= SVt_PVMG && SvMAGIC(reginfo->sv)
2257 && (mg = mg_find(reginfo->sv, PERL_MAGIC_regex_global)))) {
2258 /* prepare for quick setting of pos */
2259 #ifdef PERL_OLD_COPY_ON_WRITE
2260 if (SvIsCOW(reginfo->sv))
2261 sv_force_normal_flags(reginfo->sv, 0);
2263 mg = sv_magicext(reginfo->sv, NULL, PERL_MAGIC_regex_global,
2264 &PL_vtbl_mglob, NULL, 0);
2268 PL_reg_oldpos = mg->mg_len;
2269 SAVEDESTRUCTOR_X(restore_pos, prog);
2271 if (!PL_reg_curpm) {
2272 Newxz(PL_reg_curpm, 1, PMOP);
2275 SV* const repointer = &PL_sv_undef;
2276 /* this regexp is also owned by the new PL_reg_curpm, which
2277 will try to free it. */
2278 av_push(PL_regex_padav, repointer);
2279 PL_reg_curpm->op_pmoffset = av_len(PL_regex_padav);
2280 PL_regex_pad = AvARRAY(PL_regex_padav);
2285 /* It seems that non-ithreads works both with and without this code.
2286 So for efficiency reasons it seems best not to have the code
2287 compiled when it is not needed. */
2288 /* This is safe against NULLs: */
2289 ReREFCNT_dec(PM_GETRE(PL_reg_curpm));
2290 /* PM_reg_curpm owns a reference to this regexp. */
2293 PM_SETRE(PL_reg_curpm, rx);
2294 PL_reg_oldcurpm = PL_curpm;
2295 PL_curpm = PL_reg_curpm;
2296 if (RXp_MATCH_COPIED(prog)) {
2297 /* Here is a serious problem: we cannot rewrite subbeg,
2298 since it may be needed if this match fails. Thus
2299 $` inside (?{}) could fail... */
2300 PL_reg_oldsaved = prog->subbeg;
2301 PL_reg_oldsavedlen = prog->sublen;
2302 #ifdef PERL_OLD_COPY_ON_WRITE
2303 PL_nrs = prog->saved_copy;
2305 RXp_MATCH_COPIED_off(prog);
2308 PL_reg_oldsaved = NULL;
2309 prog->subbeg = PL_bostr;
2310 prog->sublen = PL_regeol - PL_bostr; /* strend may have been modified */
2312 DEBUG_EXECUTE_r(PL_reg_starttry = *startpos);
2313 prog->offs[0].start = *startpos - PL_bostr;
2314 PL_reginput = *startpos;
2315 PL_reglastparen = &prog->lastparen;
2316 PL_reglastcloseparen = &prog->lastcloseparen;
2317 prog->lastparen = 0;
2318 prog->lastcloseparen = 0;
2320 PL_regoffs = prog->offs;
2321 if (PL_reg_start_tmpl <= prog->nparens) {
2322 PL_reg_start_tmpl = prog->nparens*3/2 + 3;
2323 if(PL_reg_start_tmp)
2324 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2326 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2329 /* XXXX What this code is doing here?!!! There should be no need
2330 to do this again and again, PL_reglastparen should take care of
2333 /* Tests pat.t#187 and split.t#{13,14} seem to depend on this code.
2334 * Actually, the code in regcppop() (which Ilya may be meaning by
2335 * PL_reglastparen), is not needed at all by the test suite
2336 * (op/regexp, op/pat, op/split), but that code is needed otherwise
2337 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
2338 * Meanwhile, this code *is* needed for the
2339 * above-mentioned test suite tests to succeed. The common theme
2340 * on those tests seems to be returning null fields from matches.
2341 * --jhi updated by dapm */
2343 if (prog->nparens) {
2344 regexp_paren_pair *pp = PL_regoffs;
2346 for (i = prog->nparens; i > (I32)*PL_reglastparen; i--) {
2354 if (regmatch(reginfo, progi->program + 1)) {
2355 PL_regoffs[0].end = PL_reginput - PL_bostr;
2358 if (reginfo->cutpoint)
2359 *startpos= reginfo->cutpoint;
2360 REGCP_UNWIND(lastcp);
2365 #define sayYES goto yes
2366 #define sayNO goto no
2367 #define sayNO_SILENT goto no_silent
2369 /* we dont use STMT_START/END here because it leads to
2370 "unreachable code" warnings, which are bogus, but distracting. */
2371 #define CACHEsayNO \
2372 if (ST.cache_mask) \
2373 PL_reg_poscache[ST.cache_offset] |= ST.cache_mask; \
2376 /* this is used to determine how far from the left messages like
2377 'failed...' are printed. It should be set such that messages
2378 are inline with the regop output that created them.
2380 #define REPORT_CODE_OFF 32
2383 /* Make sure there is a test for this +1 options in re_tests */
2384 #define TRIE_INITAL_ACCEPT_BUFFLEN 4;
2386 #define CHRTEST_UNINIT -1001 /* c1/c2 haven't been calculated yet */
2387 #define CHRTEST_VOID -1000 /* the c1/c2 "next char" test should be skipped */
2389 #define SLAB_FIRST(s) (&(s)->states[0])
2390 #define SLAB_LAST(s) (&(s)->states[PERL_REGMATCH_SLAB_SLOTS-1])
2392 /* grab a new slab and return the first slot in it */
2394 STATIC regmatch_state *
2397 #if PERL_VERSION < 9 && !defined(PERL_CORE)
2400 regmatch_slab *s = PL_regmatch_slab->next;
2402 Newx(s, 1, regmatch_slab);
2403 s->prev = PL_regmatch_slab;
2405 PL_regmatch_slab->next = s;
2407 PL_regmatch_slab = s;
2408 return SLAB_FIRST(s);
2412 /* push a new state then goto it */
2414 #define PUSH_STATE_GOTO(state, node) \
2416 st->resume_state = state; \
2419 /* push a new state with success backtracking, then goto it */
2421 #define PUSH_YES_STATE_GOTO(state, node) \
2423 st->resume_state = state; \
2424 goto push_yes_state;
2430 regmatch() - main matching routine
2432 This is basically one big switch statement in a loop. We execute an op,
2433 set 'next' to point the next op, and continue. If we come to a point which
2434 we may need to backtrack to on failure such as (A|B|C), we push a
2435 backtrack state onto the backtrack stack. On failure, we pop the top
2436 state, and re-enter the loop at the state indicated. If there are no more
2437 states to pop, we return failure.
2439 Sometimes we also need to backtrack on success; for example /A+/, where
2440 after successfully matching one A, we need to go back and try to
2441 match another one; similarly for lookahead assertions: if the assertion
2442 completes successfully, we backtrack to the state just before the assertion
2443 and then carry on. In these cases, the pushed state is marked as
2444 'backtrack on success too'. This marking is in fact done by a chain of
2445 pointers, each pointing to the previous 'yes' state. On success, we pop to
2446 the nearest yes state, discarding any intermediate failure-only states.
2447 Sometimes a yes state is pushed just to force some cleanup code to be
2448 called at the end of a successful match or submatch; e.g. (??{$re}) uses
2449 it to free the inner regex.
2451 Note that failure backtracking rewinds the cursor position, while
2452 success backtracking leaves it alone.
2454 A pattern is complete when the END op is executed, while a subpattern
2455 such as (?=foo) is complete when the SUCCESS op is executed. Both of these
2456 ops trigger the "pop to last yes state if any, otherwise return true"
2459 A common convention in this function is to use A and B to refer to the two
2460 subpatterns (or to the first nodes thereof) in patterns like /A*B/: so A is
2461 the subpattern to be matched possibly multiple times, while B is the entire
2462 rest of the pattern. Variable and state names reflect this convention.
2464 The states in the main switch are the union of ops and failure/success of
2465 substates associated with with that op. For example, IFMATCH is the op
2466 that does lookahead assertions /(?=A)B/ and so the IFMATCH state means
2467 'execute IFMATCH'; while IFMATCH_A is a state saying that we have just
2468 successfully matched A and IFMATCH_A_fail is a state saying that we have
2469 just failed to match A. Resume states always come in pairs. The backtrack
2470 state we push is marked as 'IFMATCH_A', but when that is popped, we resume
2471 at IFMATCH_A or IFMATCH_A_fail, depending on whether we are backtracking
2472 on success or failure.
2474 The struct that holds a backtracking state is actually a big union, with
2475 one variant for each major type of op. The variable st points to the
2476 top-most backtrack struct. To make the code clearer, within each
2477 block of code we #define ST to alias the relevant union.
2479 Here's a concrete example of a (vastly oversimplified) IFMATCH
2485 #define ST st->u.ifmatch
2487 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2488 ST.foo = ...; // some state we wish to save
2490 // push a yes backtrack state with a resume value of
2491 // IFMATCH_A/IFMATCH_A_fail, then continue execution at the
2493 PUSH_YES_STATE_GOTO(IFMATCH_A, A);
2496 case IFMATCH_A: // we have successfully executed A; now continue with B
2498 bar = ST.foo; // do something with the preserved value
2501 case IFMATCH_A_fail: // A failed, so the assertion failed
2502 ...; // do some housekeeping, then ...
2503 sayNO; // propagate the failure
2510 For any old-timers reading this who are familiar with the old recursive
2511 approach, the code above is equivalent to:
2513 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2522 ...; // do some housekeeping, then ...
2523 sayNO; // propagate the failure
2526 The topmost backtrack state, pointed to by st, is usually free. If you
2527 want to claim it, populate any ST.foo fields in it with values you wish to
2528 save, then do one of
2530 PUSH_STATE_GOTO(resume_state, node);
2531 PUSH_YES_STATE_GOTO(resume_state, node);
2533 which sets that backtrack state's resume value to 'resume_state', pushes a
2534 new free entry to the top of the backtrack stack, then goes to 'node'.
2535 On backtracking, the free slot is popped, and the saved state becomes the
2536 new free state. An ST.foo field in this new top state can be temporarily
2537 accessed to retrieve values, but once the main loop is re-entered, it
2538 becomes available for reuse.
2540 Note that the depth of the backtrack stack constantly increases during the
2541 left-to-right execution of the pattern, rather than going up and down with
2542 the pattern nesting. For example the stack is at its maximum at Z at the
2543 end of the pattern, rather than at X in the following:
2545 /(((X)+)+)+....(Y)+....Z/
2547 The only exceptions to this are lookahead/behind assertions and the cut,
2548 (?>A), which pop all the backtrack states associated with A before
2551 Bascktrack state structs are allocated in slabs of about 4K in size.
2552 PL_regmatch_state and st always point to the currently active state,
2553 and PL_regmatch_slab points to the slab currently containing
2554 PL_regmatch_state. The first time regmatch() is called, the first slab is
2555 allocated, and is never freed until interpreter destruction. When the slab
2556 is full, a new one is allocated and chained to the end. At exit from
2557 regmatch(), slabs allocated since entry are freed.
2562 #define DEBUG_STATE_pp(pp) \
2564 DUMP_EXEC_POS(locinput, scan, do_utf8); \
2565 PerlIO_printf(Perl_debug_log, \
2566 " %*s"pp" %s%s%s%s%s\n", \
2568 PL_reg_name[st->resume_state], \
2569 ((st==yes_state||st==mark_state) ? "[" : ""), \
2570 ((st==yes_state) ? "Y" : ""), \
2571 ((st==mark_state) ? "M" : ""), \
2572 ((st==yes_state||st==mark_state) ? "]" : "") \
2577 #define REG_NODE_NUM(x) ((x) ? (int)((x)-prog) : -1)
2582 S_debug_start_match(pTHX_ const REGEXP *prog, const bool do_utf8,
2583 const char *start, const char *end, const char *blurb)
2585 const bool utf8_pat = RX_UTF8(prog) ? 1 : 0;
2587 PERL_ARGS_ASSERT_DEBUG_START_MATCH;
2592 RE_PV_QUOTED_DECL(s0, utf8_pat, PERL_DEBUG_PAD_ZERO(0),
2593 RX_PRECOMP_const(prog), RX_PRELEN(prog), 60);
2595 RE_PV_QUOTED_DECL(s1, do_utf8, PERL_DEBUG_PAD_ZERO(1),
2596 start, end - start, 60);
2598 PerlIO_printf(Perl_debug_log,
2599 "%s%s REx%s %s against %s\n",
2600 PL_colors[4], blurb, PL_colors[5], s0, s1);
2602 if (do_utf8||utf8_pat)
2603 PerlIO_printf(Perl_debug_log, "UTF-8 %s%s%s...\n",
2604 utf8_pat ? "pattern" : "",
2605 utf8_pat && do_utf8 ? " and " : "",
2606 do_utf8 ? "string" : ""
2612 S_dump_exec_pos(pTHX_ const char *locinput,
2613 const regnode *scan,
2614 const char *loc_regeol,
2615 const char *loc_bostr,
2616 const char *loc_reg_starttry,
2619 const int docolor = *PL_colors[0] || *PL_colors[2] || *PL_colors[4];
2620 const int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */
2621 int l = (loc_regeol - locinput) > taill ? taill : (loc_regeol - locinput);
2622 /* The part of the string before starttry has one color
2623 (pref0_len chars), between starttry and current
2624 position another one (pref_len - pref0_len chars),
2625 after the current position the third one.
2626 We assume that pref0_len <= pref_len, otherwise we
2627 decrease pref0_len. */
2628 int pref_len = (locinput - loc_bostr) > (5 + taill) - l
2629 ? (5 + taill) - l : locinput - loc_bostr;
2632 PERL_ARGS_ASSERT_DUMP_EXEC_POS;
2634 while (do_utf8 && UTF8_IS_CONTINUATION(*(U8*)(locinput - pref_len)))
2636 pref0_len = pref_len - (locinput - loc_reg_starttry);
2637 if (l + pref_len < (5 + taill) && l < loc_regeol - locinput)
2638 l = ( loc_regeol - locinput > (5 + taill) - pref_len
2639 ? (5 + taill) - pref_len : loc_regeol - locinput);
2640 while (do_utf8 && UTF8_IS_CONTINUATION(*(U8*)(locinput + l)))
2644 if (pref0_len > pref_len)
2645 pref0_len = pref_len;
2647 const int is_uni = (do_utf8 && OP(scan) != CANY) ? 1 : 0;
2649 RE_PV_COLOR_DECL(s0,len0,is_uni,PERL_DEBUG_PAD(0),
2650 (locinput - pref_len),pref0_len, 60, 4, 5);
2652 RE_PV_COLOR_DECL(s1,len1,is_uni,PERL_DEBUG_PAD(1),
2653 (locinput - pref_len + pref0_len),
2654 pref_len - pref0_len, 60, 2, 3);
2656 RE_PV_COLOR_DECL(s2,len2,is_uni,PERL_DEBUG_PAD(2),
2657 locinput, loc_regeol - locinput, 10, 0, 1);
2659 const STRLEN tlen=len0+len1+len2;
2660 PerlIO_printf(Perl_debug_log,
2661 "%4"IVdf" <%.*s%.*s%s%.*s>%*s|",
2662 (IV)(locinput - loc_bostr),
2665 (docolor ? "" : "> <"),
2667 (int)(tlen > 19 ? 0 : 19 - tlen),
2674 /* reg_check_named_buff_matched()
2675 * Checks to see if a named buffer has matched. The data array of
2676 * buffer numbers corresponding to the buffer is expected to reside
2677 * in the regexp->data->data array in the slot stored in the ARG() of
2678 * node involved. Note that this routine doesn't actually care about the
2679 * name, that information is not preserved from compilation to execution.
2680 * Returns the index of the leftmost defined buffer with the given name
2681 * or 0 if non of the buffers matched.
2684 S_reg_check_named_buff_matched(pTHX_ const regexp *rex, const regnode *scan)
2687 RXi_GET_DECL(rex,rexi);
2688 SV *sv_dat= MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
2689 I32 *nums=(I32*)SvPVX(sv_dat);
2691 PERL_ARGS_ASSERT_REG_CHECK_NAMED_BUFF_MATCHED;
2693 for ( n=0; n<SvIVX(sv_dat); n++ ) {
2694 if ((I32)*PL_reglastparen >= nums[n] &&
2695 PL_regoffs[nums[n]].end != -1)
2704 /* free all slabs above current one - called during LEAVE_SCOPE */
2707 S_clear_backtrack_stack(pTHX_ void *p)
2709 regmatch_slab *s = PL_regmatch_slab->next;
2714 PL_regmatch_slab->next = NULL;
2716 regmatch_slab * const osl = s;
2723 #define SETREX(Re1,Re2) \
2724 if (PL_reg_eval_set) PM_SETRE((PL_reg_curpm), (Re2)); \
2727 STATIC I32 /* 0 failure, 1 success */
2728 S_regmatch(pTHX_ regmatch_info *reginfo, regnode *prog)
2730 #if PERL_VERSION < 9 && !defined(PERL_CORE)
2734 register const bool do_utf8 = PL_reg_match_utf8;
2735 const U32 uniflags = UTF8_ALLOW_DEFAULT;
2736 REGEXP *rex_sv = reginfo->prog;
2737 regexp *rex = (struct regexp *)SvANY(rex_sv);
2738 RXi_GET_DECL(rex,rexi);
2740 /* the current state. This is a cached copy of PL_regmatch_state */
2741 register regmatch_state *st;
2742 /* cache heavy used fields of st in registers */
2743 register regnode *scan;
2744 register regnode *next;
2745 register U32 n = 0; /* general value; init to avoid compiler warning */
2746 register I32 ln = 0; /* len or last; init to avoid compiler warning */
2747 register char *locinput = PL_reginput;
2748 register I32 nextchr; /* is always set to UCHARAT(locinput) */
2750 bool result = 0; /* return value of S_regmatch */
2751 int depth = 0; /* depth of backtrack stack */
2752 U32 nochange_depth = 0; /* depth of GOSUB recursion with nochange */
2753 const U32 max_nochange_depth =
2754 (3 * rex->nparens > MAX_RECURSE_EVAL_NOCHANGE_DEPTH) ?
2755 3 * rex->nparens : MAX_RECURSE_EVAL_NOCHANGE_DEPTH;
2756 regmatch_state *yes_state = NULL; /* state to pop to on success of
2758 /* mark_state piggy backs on the yes_state logic so that when we unwind
2759 the stack on success we can update the mark_state as we go */
2760 regmatch_state *mark_state = NULL; /* last mark state we have seen */
2761 regmatch_state *cur_eval = NULL; /* most recent EVAL_AB state */
2762 struct regmatch_state *cur_curlyx = NULL; /* most recent curlyx */
2764 bool no_final = 0; /* prevent failure from backtracking? */
2765 bool do_cutgroup = 0; /* no_final only until next branch/trie entry */
2766 char *startpoint = PL_reginput;
2767 SV *popmark = NULL; /* are we looking for a mark? */
2768 SV *sv_commit = NULL; /* last mark name seen in failure */
2769 SV *sv_yes_mark = NULL; /* last mark name we have seen
2770 during a successfull match */
2771 U32 lastopen = 0; /* last open we saw */
2772 bool has_cutgroup = RX_HAS_CUTGROUP(rex) ? 1 : 0;
2773 SV* const oreplsv = GvSV(PL_replgv);
2774 /* these three flags are set by various ops to signal information to
2775 * the very next op. They have a useful lifetime of exactly one loop
2776 * iteration, and are not preserved or restored by state pushes/pops
2778 bool sw = 0; /* the condition value in (?(cond)a|b) */
2779 bool minmod = 0; /* the next "{n,m}" is a "{n,m}?" */
2780 int logical = 0; /* the following EVAL is:
2784 or the following IFMATCH/UNLESSM is:
2785 false: plain (?=foo)
2786 true: used as a condition: (?(?=foo))
2789 GET_RE_DEBUG_FLAGS_DECL;
2792 PERL_ARGS_ASSERT_REGMATCH;
2794 DEBUG_OPTIMISE_r( DEBUG_EXECUTE_r({
2795 PerlIO_printf(Perl_debug_log,"regmatch start\n");
2797 /* on first ever call to regmatch, allocate first slab */
2798 if (!PL_regmatch_slab) {
2799 Newx(PL_regmatch_slab, 1, regmatch_slab);
2800 PL_regmatch_slab->prev = NULL;
2801 PL_regmatch_slab->next = NULL;
2802 PL_regmatch_state = SLAB_FIRST(PL_regmatch_slab);
2805 oldsave = PL_savestack_ix;
2806 SAVEDESTRUCTOR_X(S_clear_backtrack_stack, NULL);
2807 SAVEVPTR(PL_regmatch_slab);
2808 SAVEVPTR(PL_regmatch_state);
2810 /* grab next free state slot */
2811 st = ++PL_regmatch_state;
2812 if (st > SLAB_LAST(PL_regmatch_slab))
2813 st = PL_regmatch_state = S_push_slab(aTHX);
2815 /* Note that nextchr is a byte even in UTF */
2816 nextchr = UCHARAT(locinput);
2818 while (scan != NULL) {
2821 SV * const prop = sv_newmortal();
2822 regnode *rnext=regnext(scan);
2823 DUMP_EXEC_POS( locinput, scan, do_utf8 );
2824 regprop(rex, prop, scan);
2826 PerlIO_printf(Perl_debug_log,
2827 "%3"IVdf":%*s%s(%"IVdf")\n",
2828 (IV)(scan - rexi->program), depth*2, "",
2830 (PL_regkind[OP(scan)] == END || !rnext) ?
2831 0 : (IV)(rnext - rexi->program));
2834 next = scan + NEXT_OFF(scan);
2837 state_num = OP(scan);
2840 switch (state_num) {
2842 if (locinput == PL_bostr)
2844 /* reginfo->till = reginfo->bol; */
2849 if (locinput == PL_bostr ||
2850 ((nextchr || locinput < PL_regeol) && locinput[-1] == '\n'))
2856 if (locinput == PL_bostr)
2860 if (locinput == reginfo->ganch)
2865 /* update the startpoint */
2866 st->u.keeper.val = PL_regoffs[0].start;
2867 PL_reginput = locinput;
2868 PL_regoffs[0].start = locinput - PL_bostr;
2869 PUSH_STATE_GOTO(KEEPS_next, next);
2871 case KEEPS_next_fail:
2872 /* rollback the start point change */
2873 PL_regoffs[0].start = st->u.keeper.val;
2879 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
2884 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
2886 if (PL_regeol - locinput > 1)
2890 if (PL_regeol != locinput)
2894 if (!nextchr && locinput >= PL_regeol)
2897 locinput += PL_utf8skip[nextchr];
2898 if (locinput > PL_regeol)
2900 nextchr = UCHARAT(locinput);
2903 nextchr = UCHARAT(++locinput);
2906 if (!nextchr && locinput >= PL_regeol)
2908 nextchr = UCHARAT(++locinput);
2911 if ((!nextchr && locinput >= PL_regeol) || nextchr == '\n')
2914 locinput += PL_utf8skip[nextchr];
2915 if (locinput > PL_regeol)
2917 nextchr = UCHARAT(locinput);
2920 nextchr = UCHARAT(++locinput);
2924 #define ST st->u.trie
2926 /* In this case the charclass data is available inline so
2927 we can fail fast without a lot of extra overhead.
2929 if (scan->flags == EXACT || !do_utf8) {
2930 if(!ANYOF_BITMAP_TEST(scan, *locinput)) {
2932 PerlIO_printf(Perl_debug_log,
2933 "%*s %sfailed to match trie start class...%s\n",
2934 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2943 /* what type of TRIE am I? (utf8 makes this contextual) */
2944 DECL_TRIE_TYPE(scan);
2946 /* what trie are we using right now */
2947 reg_trie_data * const trie
2948 = (reg_trie_data*)rexi->data->data[ ARG( scan ) ];
2949 HV * widecharmap = MUTABLE_HV(rexi->data->data[ ARG( scan ) + 1 ]);
2950 U32 state = trie->startstate;
2952 if (trie->bitmap && trie_type != trie_utf8_fold &&
2953 !TRIE_BITMAP_TEST(trie,*locinput)
2955 if (trie->states[ state ].wordnum) {
2957 PerlIO_printf(Perl_debug_log,
2958 "%*s %smatched empty string...%s\n",
2959 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2964 PerlIO_printf(Perl_debug_log,
2965 "%*s %sfailed to match trie start class...%s\n",
2966 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
2973 U8 *uc = ( U8* )locinput;
2977 U8 *uscan = (U8*)NULL;
2979 SV *sv_accept_buff = NULL;
2980 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
2982 ST.accepted = 0; /* how many accepting states we have seen */
2984 ST.jump = trie->jump;
2987 traverse the TRIE keeping track of all accepting states
2988 we transition through until we get to a failing node.
2991 while ( state && uc <= (U8*)PL_regeol ) {
2992 U32 base = trie->states[ state ].trans.base;
2995 /* We use charid to hold the wordnum as we don't use it
2996 for charid until after we have done the wordnum logic.
2997 We define an alias just so that the wordnum logic reads
3000 #define got_wordnum charid
3001 got_wordnum = trie->states[ state ].wordnum;
3003 if ( got_wordnum ) {
3004 if ( ! ST.accepted ) {
3006 SAVETMPS; /* XXX is this necessary? dmq */
3007 bufflen = TRIE_INITAL_ACCEPT_BUFFLEN;
3008 sv_accept_buff=newSV(bufflen *
3009 sizeof(reg_trie_accepted) - 1);
3010 SvCUR_set(sv_accept_buff, 0);
3011 SvPOK_on(sv_accept_buff);
3012 sv_2mortal(sv_accept_buff);
3015 (reg_trie_accepted*)SvPV_nolen(sv_accept_buff );
3018 if (ST.accepted >= bufflen) {
3020 ST.accept_buff =(reg_trie_accepted*)
3021 SvGROW(sv_accept_buff,
3022 bufflen * sizeof(reg_trie_accepted));
3024 SvCUR_set(sv_accept_buff,SvCUR(sv_accept_buff)
3025 + sizeof(reg_trie_accepted));
3028 ST.accept_buff[ST.accepted].wordnum = got_wordnum;
3029 ST.accept_buff[ST.accepted].endpos = uc;
3031 } while (trie->nextword && (got_wordnum= trie->nextword[got_wordnum]));
3035 DEBUG_TRIE_EXECUTE_r({
3036 DUMP_EXEC_POS( (char *)uc, scan, do_utf8 );
3037 PerlIO_printf( Perl_debug_log,
3038 "%*s %sState: %4"UVxf" Accepted: %4"UVxf" ",
3039 2+depth * 2, "", PL_colors[4],
3040 (UV)state, (UV)ST.accepted );
3044 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
3045 uscan, len, uvc, charid, foldlen,
3049 (base + charid > trie->uniquecharcount )
3050 && (base + charid - 1 - trie->uniquecharcount
3052 && trie->trans[base + charid - 1 -
3053 trie->uniquecharcount].check == state)
3055 state = trie->trans[base + charid - 1 -
3056 trie->uniquecharcount ].next;
3067 DEBUG_TRIE_EXECUTE_r(
3068 PerlIO_printf( Perl_debug_log,
3069 "Charid:%3x CP:%4"UVxf" After State: %4"UVxf"%s\n",
3070 charid, uvc, (UV)state, PL_colors[5] );
3077 PerlIO_printf( Perl_debug_log,
3078 "%*s %sgot %"IVdf" possible matches%s\n",
3079 REPORT_CODE_OFF + depth * 2, "",
3080 PL_colors[4], (IV)ST.accepted, PL_colors[5] );
3083 goto trie_first_try; /* jump into the fail handler */
3085 case TRIE_next_fail: /* we failed - try next alterative */
3087 REGCP_UNWIND(ST.cp);
3088 for (n = *PL_reglastparen; n > ST.lastparen; n--)
3089 PL_regoffs[n].end = -1;
3090 *PL_reglastparen = n;
3099 ST.lastparen = *PL_reglastparen;
3102 if ( ST.accepted == 1 ) {
3103 /* only one choice left - just continue */
3105 AV *const trie_words
3106 = MUTABLE_AV(rexi->data->data[ARG(ST.me)+TRIE_WORDS_OFFSET]);
3107 SV ** const tmp = av_fetch( trie_words,
3108 ST.accept_buff[ 0 ].wordnum-1, 0 );
3109 SV *sv= tmp ? sv_newmortal() : NULL;
3111 PerlIO_printf( Perl_debug_log,
3112 "%*s %sonly one match left: #%d <%s>%s\n",
3113 REPORT_CODE_OFF+depth*2, "", PL_colors[4],
3114 ST.accept_buff[ 0 ].wordnum,
3115 tmp ? pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 0,
3116 PL_colors[0], PL_colors[1],
3117 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0)
3119 : "not compiled under -Dr",
3122 PL_reginput = (char *)ST.accept_buff[ 0 ].endpos;
3123 /* in this case we free tmps/leave before we call regmatch
3124 as we wont be using accept_buff again. */
3126 locinput = PL_reginput;
3127 nextchr = UCHARAT(locinput);
3128 if ( !ST.jump || !ST.jump[ST.accept_buff[0].wordnum])
3131 scan = ST.me + ST.jump[ST.accept_buff[0].wordnum];
3132 if (!has_cutgroup) {
3137 PUSH_YES_STATE_GOTO(TRIE_next, scan);
3140 continue; /* execute rest of RE */
3143 if ( !ST.accepted-- ) {
3145 PerlIO_printf( Perl_debug_log,
3146 "%*s %sTRIE failed...%s\n",
3147 REPORT_CODE_OFF+depth*2, "",
3158 There are at least two accepting states left. Presumably
3159 the number of accepting states is going to be low,
3160 typically two. So we simply scan through to find the one
3161 with lowest wordnum. Once we find it, we swap the last
3162 state into its place and decrement the size. We then try to
3163 match the rest of the pattern at the point where the word
3164 ends. If we succeed, control just continues along the
3165 regex; if we fail we return here to try the next accepting
3172 for( cur = 1 ; cur <= ST.accepted ; cur++ ) {
3173 DEBUG_TRIE_EXECUTE_r(
3174 PerlIO_printf( Perl_debug_log,
3175 "%*s %sgot %"IVdf" (%d) as best, looking at %"IVdf" (%d)%s\n",
3176 REPORT_CODE_OFF + depth * 2, "", PL_colors[4],
3177 (IV)best, ST.accept_buff[ best ].wordnum, (IV)cur,
3178 ST.accept_buff[ cur ].wordnum, PL_colors[5] );
3181 if (ST.accept_buff[cur].wordnum <
3182 ST.accept_buff[best].wordnum)
3187 AV *const trie_words
3188 = MUTABLE_AV(rexi->data->data[ARG(ST.me)+TRIE_WORDS_OFFSET]);
3189 SV ** const tmp = av_fetch( trie_words,
3190 ST.accept_buff[ best ].wordnum - 1, 0 );
3191 regnode *nextop=(!ST.jump || !ST.jump[ST.accept_buff[best].wordnum]) ?
3193 ST.me + ST.jump[ST.accept_buff[best].wordnum];
3194 SV *sv= tmp ? sv_newmortal() : NULL;
3196 PerlIO_printf( Perl_debug_log,
3197 "%*s %strying alternation #%d <%s> at node #%d %s\n",
3198 REPORT_CODE_OFF+depth*2, "", PL_colors[4],
3199 ST.accept_buff[best].wordnum,
3200 tmp ? pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 0,
3201 PL_colors[0], PL_colors[1],
3202 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0)
3203 ) : "not compiled under -Dr",
3204 REG_NODE_NUM(nextop),
3208 if ( best<ST.accepted ) {
3209 reg_trie_accepted tmp = ST.accept_buff[ best ];
3210 ST.accept_buff[ best ] = ST.accept_buff[ ST.accepted ];
3211 ST.accept_buff[ ST.accepted ] = tmp;
3214 PL_reginput = (char *)ST.accept_buff[ best ].endpos;
3215 if ( !ST.jump || !ST.jump[ST.accept_buff[best].wordnum]) {
3218 scan = ST.me + ST.jump[ST.accept_buff[best].wordnum];
3220 PUSH_YES_STATE_GOTO(TRIE_next, scan);
3225 /* we dont want to throw this away, see bug 57042*/
3226 if (oreplsv != GvSV(PL_replgv))
3227 sv_setsv(oreplsv, GvSV(PL_replgv));
3234 char *s = STRING(scan);
3236 if (do_utf8 != UTF) {
3237 /* The target and the pattern have differing utf8ness. */
3239 const char * const e = s + ln;
3242 /* The target is utf8, the pattern is not utf8. */
3247 if (NATIVE_TO_UNI(*(U8*)s) !=
3248 utf8n_to_uvuni((U8*)l, UTF8_MAXBYTES, &ulen,
3256 /* The target is not utf8, the pattern is utf8. */
3261 if (NATIVE_TO_UNI(*((U8*)l)) !=
3262 utf8n_to_uvuni((U8*)s, UTF8_MAXBYTES, &ulen,
3270 nextchr = UCHARAT(locinput);
3273 /* The target and the pattern have the same utf8ness. */
3274 /* Inline the first character, for speed. */
3275 if (UCHARAT(s) != nextchr)
3277 if (PL_regeol - locinput < ln)
3279 if (ln > 1 && memNE(s, locinput, ln))
3282 nextchr = UCHARAT(locinput);
3286 PL_reg_flags |= RF_tainted;
3289 char * const s = STRING(scan);
3292 if (do_utf8 || UTF) {
3293 /* Either target or the pattern are utf8. */
3294 const char * const l = locinput;
3295 char *e = PL_regeol;
3297 if (ibcmp_utf8(s, 0, ln, (bool)UTF,
3298 l, &e, 0, do_utf8)) {
3299 /* One more case for the sharp s:
3300 * pack("U0U*", 0xDF) =~ /ss/i,
3301 * the 0xC3 0x9F are the UTF-8
3302 * byte sequence for the U+00DF. */
3305 toLOWER(s[0]) == 's' &&
3307 toLOWER(s[1]) == 's' &&
3314 nextchr = UCHARAT(locinput);
3318 /* Neither the target and the pattern are utf8. */
3320 /* Inline the first character, for speed. */
3321 if (UCHARAT(s) != nextchr &&
3322 UCHARAT(s) != ((OP(scan) == EXACTF)
3323 ? PL_fold : PL_fold_locale)[nextchr])
3325 if (PL_regeol - locinput < ln)
3327 if (ln > 1 && (OP(scan) == EXACTF
3328 ? ibcmp(s, locinput, ln)
3329 : ibcmp_locale(s, locinput, ln)))
3332 nextchr = UCHARAT(locinput);
3337 STRLEN inclasslen = PL_regeol - locinput;
3339 if (!reginclass(rex, scan, (U8*)locinput, &inclasslen, do_utf8))
3341 if (locinput >= PL_regeol)
3343 locinput += inclasslen ? inclasslen : UTF8SKIP(locinput);
3344 nextchr = UCHARAT(locinput);
3349 nextchr = UCHARAT(locinput);
3350 if (!REGINCLASS(rex, scan, (U8*)locinput))
3352 if (!nextchr && locinput >= PL_regeol)
3354 nextchr = UCHARAT(++locinput);
3358 /* If we might have the case of the German sharp s
3359 * in a casefolding Unicode character class. */
3361 if (ANYOF_FOLD_SHARP_S(scan, locinput, PL_regeol)) {
3362 locinput += SHARP_S_SKIP;
3363 nextchr = UCHARAT(locinput);
3369 PL_reg_flags |= RF_tainted;
3375 LOAD_UTF8_CHARCLASS_ALNUM();
3376 if (!(OP(scan) == ALNUM
3377 ? (bool)swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
3378 : isALNUM_LC_utf8((U8*)locinput)))
3382 locinput += PL_utf8skip[nextchr];
3383 nextchr = UCHARAT(locinput);
3386 if (!(OP(scan) == ALNUM
3387 ? isALNUM(nextchr) : isALNUM_LC(nextchr)))
3389 nextchr = UCHARAT(++locinput);
3392 PL_reg_flags |= RF_tainted;
3395 if (!nextchr && locinput >= PL_regeol)
3398 LOAD_UTF8_CHARCLASS_ALNUM();
3399 if (OP(scan) == NALNUM
3400 ? (bool)swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
3401 : isALNUM_LC_utf8((U8*)locinput))
3405 locinput += PL_utf8skip[nextchr];
3406 nextchr = UCHARAT(locinput);
3409 if (OP(scan) == NALNUM
3410 ? isALNUM(nextchr) : isALNUM_LC(nextchr))
3412 nextchr = UCHARAT(++locinput);
3416 PL_reg_flags |= RF_tainted;
3420 /* was last char in word? */
3422 if (locinput == PL_bostr)
3425 const U8 * const r = reghop3((U8*)locinput, -1, (U8*)PL_bostr);
3427 ln = utf8n_to_uvchr(r, UTF8SKIP(r), 0, uniflags);
3429 if (OP(scan) == BOUND || OP(scan) == NBOUND) {
3430 ln = isALNUM_uni(ln);
3431 LOAD_UTF8_CHARCLASS_ALNUM();
3432 n = swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8);
3435 ln = isALNUM_LC_uvchr(UNI_TO_NATIVE(ln));
3436 n = isALNUM_LC_utf8((U8*)locinput);
3440 ln = (locinput != PL_bostr) ?
3441 UCHARAT(locinput - 1) : '\n';
3442 if (OP(scan) == BOUND || OP(scan) == NBOUND) {
3444 n = isALNUM(nextchr);
3447 ln = isALNUM_LC(ln);
3448 n = isALNUM_LC(nextchr);
3451 if (((!ln) == (!n)) == (OP(scan) == BOUND ||
3452 OP(scan) == BOUNDL))
3456 PL_reg_flags |= RF_tainted;
3462 if (UTF8_IS_CONTINUED(nextchr)) {
3463 LOAD_UTF8_CHARCLASS_SPACE();
3464 if (!(OP(scan) == SPACE
3465 ? (bool)swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
3466 : isSPACE_LC_utf8((U8*)locinput)))
3470 locinput += PL_utf8skip[nextchr];
3471 nextchr = UCHARAT(locinput);
3474 if (!(OP(scan) == SPACE
3475 ? isSPACE(nextchr) : isSPACE_LC(nextchr)))
3477 nextchr = UCHARAT(++locinput);
3480 if (!(OP(scan) == SPACE
3481 ? isSPACE(nextchr) : isSPACE_LC(nextchr)))
3483 nextchr = UCHARAT(++locinput);
3487 PL_reg_flags |= RF_tainted;
3490 if (!nextchr && locinput >= PL_regeol)
3493 LOAD_UTF8_CHARCLASS_SPACE();
3494 if (OP(scan) == NSPACE
3495 ? (bool)swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
3496 : isSPACE_LC_utf8((U8*)locinput))
3500 locinput += PL_utf8skip[nextchr];
3501 nextchr = UCHARAT(locinput);
3504 if (OP(scan) == NSPACE
3505 ? isSPACE(nextchr) : isSPACE_LC(nextchr))
3507 nextchr = UCHARAT(++locinput);
3510 PL_reg_flags |= RF_tainted;
3516 LOAD_UTF8_CHARCLASS_DIGIT();
3517 if (!(OP(scan) == DIGIT
3518 ? (bool)swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
3519 : isDIGIT_LC_utf8((U8*)locinput)))
3523 locinput += PL_utf8skip[nextchr];
3524 nextchr = UCHARAT(locinput);
3527 if (!(OP(scan) == DIGIT
3528 ? isDIGIT(nextchr) : isDIGIT_LC(nextchr)))
3530 nextchr = UCHARAT(++locinput);
3533 PL_reg_flags |= RF_tainted;
3536 if (!nextchr && locinput >= PL_regeol)
3539 LOAD_UTF8_CHARCLASS_DIGIT();
3540 if (OP(scan) == NDIGIT
3541 ? (bool)swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
3542 : isDIGIT_LC_utf8((U8*)locinput))
3546 locinput += PL_utf8skip[nextchr];
3547 nextchr = UCHARAT(locinput);
3550 if (OP(scan) == NDIGIT
3551 ? isDIGIT(nextchr) : isDIGIT_LC(nextchr))
3553 nextchr = UCHARAT(++locinput);
3556 if (locinput >= PL_regeol)
3559 LOAD_UTF8_CHARCLASS_MARK();
3560 if (swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8))
3562 locinput += PL_utf8skip[nextchr];
3563 while (locinput < PL_regeol &&
3564 swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8))
3565 locinput += UTF8SKIP(locinput);
3566 if (locinput > PL_regeol)
3571 nextchr = UCHARAT(locinput);
3578 PL_reg_flags |= RF_tainted;
3583 n = reg_check_named_buff_matched(rex,scan);
3586 type = REF + ( type - NREF );
3593 PL_reg_flags |= RF_tainted;
3597 n = ARG(scan); /* which paren pair */
3600 ln = PL_regoffs[n].start;
3601 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
3602 if (*PL_reglastparen < n || ln == -1)
3603 sayNO; /* Do not match unless seen CLOSEn. */
3604 if (ln == PL_regoffs[n].end)
3608 if (do_utf8 && type != REF) { /* REF can do byte comparison */
3610 const char *e = PL_bostr + PL_regoffs[n].end;
3612 * Note that we can't do the "other character" lookup trick as
3613 * in the 8-bit case (no pun intended) because in Unicode we
3614 * have to map both upper and title case to lower case.
3618 STRLEN ulen1, ulen2;
3619 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
3620 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
3624 toLOWER_utf8((U8*)s, tmpbuf1, &ulen1);
3625 toLOWER_utf8((U8*)l, tmpbuf2, &ulen2);
3626 if (ulen1 != ulen2 || memNE((char *)tmpbuf1, (char *)tmpbuf2, ulen1))
3633 nextchr = UCHARAT(locinput);
3637 /* Inline the first character, for speed. */
3638 if (UCHARAT(s) != nextchr &&
3640 (UCHARAT(s) != (type == REFF
3641 ? PL_fold : PL_fold_locale)[nextchr])))
3643 ln = PL_regoffs[n].end - ln;
3644 if (locinput + ln > PL_regeol)
3646 if (ln > 1 && (type == REF
3647 ? memNE(s, locinput, ln)
3649 ? ibcmp(s, locinput, ln)
3650 : ibcmp_locale(s, locinput, ln))))
3653 nextchr = UCHARAT(locinput);
3663 #define ST st->u.eval
3668 regexp_internal *rei;
3669 regnode *startpoint;
3672 case GOSUB: /* /(...(?1))/ /(...(?&foo))/ */
3673 if (cur_eval && cur_eval->locinput==locinput) {
3674 if (cur_eval->u.eval.close_paren == (U32)ARG(scan))
3675 Perl_croak(aTHX_ "Infinite recursion in regex");
3676 if ( ++nochange_depth > max_nochange_depth )
3678 "Pattern subroutine nesting without pos change"
3679 " exceeded limit in regex");
3686 (void)ReREFCNT_inc(rex_sv);
3687 if (OP(scan)==GOSUB) {
3688 startpoint = scan + ARG2L(scan);
3689 ST.close_paren = ARG(scan);
3691 startpoint = rei->program+1;
3694 goto eval_recurse_doit;
3696 case EVAL: /* /(?{A})B/ /(??{A})B/ and /(?(?{A})X|Y)B/ */
3697 if (cur_eval && cur_eval->locinput==locinput) {
3698 if ( ++nochange_depth > max_nochange_depth )
3699 Perl_croak(aTHX_ "EVAL without pos change exceeded limit in regex");
3704 /* execute the code in the {...} */
3706 SV ** const before = SP;
3707 OP_4tree * const oop = PL_op;
3708 COP * const ocurcop = PL_curcop;
3712 PL_op = (OP_4tree*)rexi->data->data[n];
3713 DEBUG_STATE_r( PerlIO_printf(Perl_debug_log,
3714 " re_eval 0x%"UVxf"\n", PTR2UV(PL_op)) );
3715 PAD_SAVE_LOCAL(old_comppad, (PAD*)rexi->data->data[n + 2]);
3716 PL_regoffs[0].end = PL_reg_magic->mg_len = locinput - PL_bostr;
3719 SV *sv_mrk = get_sv("REGMARK", 1);
3720 sv_setsv(sv_mrk, sv_yes_mark);
3723 CALLRUNOPS(aTHX); /* Scalar context. */
3726 ret = &PL_sv_undef; /* protect against empty (?{}) blocks. */
3733 PAD_RESTORE_LOCAL(old_comppad);
3734 PL_curcop = ocurcop;
3737 sv_setsv(save_scalar(PL_replgv), ret);
3741 if (logical == 2) { /* Postponed subexpression: /(??{...})/ */
3744 /* extract RE object from returned value; compiling if
3750 SV *const sv = SvRV(ret);
3752 if (SvTYPE(sv) == SVt_REGEXP) {
3754 } else if (SvSMAGICAL(sv)) {
3755 mg = mg_find(sv, PERL_MAGIC_qr);
3758 } else if (SvTYPE(ret) == SVt_REGEXP) {
3760 } else if (SvSMAGICAL(ret)) {
3761 if (SvGMAGICAL(ret)) {
3762 /* I don't believe that there is ever qr magic
3764 assert(!mg_find(ret, PERL_MAGIC_qr));
3765 sv_unmagic(ret, PERL_MAGIC_qr);
3768 mg = mg_find(ret, PERL_MAGIC_qr);
3769 /* testing suggests mg only ends up non-NULL for
3770 scalars who were upgraded and compiled in the
3771 else block below. In turn, this is only
3772 triggered in the "postponed utf8 string" tests
3778 rx = (REGEXP *) mg->mg_obj; /*XXX:dmq*/
3782 rx = reg_temp_copy(rx);
3786 const I32 osize = PL_regsize;
3789 assert (SvUTF8(ret));
3790 } else if (SvUTF8(ret)) {
3791 /* Not doing UTF-8, despite what the SV says. Is
3792 this only if we're trapped in use 'bytes'? */
3793 /* Make a copy of the octet sequence, but without
3794 the flag on, as the compiler now honours the
3795 SvUTF8 flag on ret. */
3797 const char *const p = SvPV(ret, len);
3798 ret = newSVpvn_flags(p, len, SVs_TEMP);
3800 rx = CALLREGCOMP(ret, pm_flags);
3802 & (SVs_TEMP | SVs_PADTMP | SVf_READONLY
3804 /* This isn't a first class regexp. Instead, it's
3805 caching a regexp onto an existing, Perl visible
3807 sv_magic(ret, MUTABLE_SV(rx), PERL_MAGIC_qr, 0, 0);
3812 re = (struct regexp *)SvANY(rx);
3814 RXp_MATCH_COPIED_off(re);
3815 re->subbeg = rex->subbeg;
3816 re->sublen = rex->sublen;
3819 debug_start_match(re_sv, do_utf8, locinput, PL_regeol,
3820 "Matching embedded");
3822 startpoint = rei->program + 1;
3823 ST.close_paren = 0; /* only used for GOSUB */
3824 /* borrowed from regtry */
3825 if (PL_reg_start_tmpl <= re->nparens) {
3826 PL_reg_start_tmpl = re->nparens*3/2 + 3;
3827 if(PL_reg_start_tmp)
3828 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
3830 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
3833 eval_recurse_doit: /* Share code with GOSUB below this line */
3834 /* run the pattern returned from (??{...}) */
3835 ST.cp = regcppush(0); /* Save *all* the positions. */
3836 REGCP_SET(ST.lastcp);
3838 PL_regoffs = re->offs; /* essentially NOOP on GOSUB */
3840 /* see regtry, specifically PL_reglast(?:close)?paren is a pointer! (i dont know why) :dmq */
3841 PL_reglastparen = &re->lastparen;
3842 PL_reglastcloseparen = &re->lastcloseparen;
3844 re->lastcloseparen = 0;
3846 PL_reginput = locinput;
3849 /* XXXX This is too dramatic a measure... */
3852 ST.toggle_reg_flags = PL_reg_flags;
3854 PL_reg_flags |= RF_utf8;
3856 PL_reg_flags &= ~RF_utf8;
3857 ST.toggle_reg_flags ^= PL_reg_flags; /* diff of old and new */
3859 ST.prev_rex = rex_sv;
3860 ST.prev_curlyx = cur_curlyx;
3861 SETREX(rex_sv,re_sv);
3866 ST.prev_eval = cur_eval;
3868 /* now continue from first node in postoned RE */
3869 PUSH_YES_STATE_GOTO(EVAL_AB, startpoint);
3872 /* logical is 1, /(?(?{...})X|Y)/ */
3873 sw = (bool)SvTRUE(ret);
3878 case EVAL_AB: /* cleanup after a successful (??{A})B */
3879 /* note: this is called twice; first after popping B, then A */
3880 PL_reg_flags ^= ST.toggle_reg_flags;
3881 ReREFCNT_dec(rex_sv);
3882 SETREX(rex_sv,ST.prev_rex);
3883 rex = (struct regexp *)SvANY(rex_sv);
3884 rexi = RXi_GET(rex);
3886 cur_eval = ST.prev_eval;
3887 cur_curlyx = ST.prev_curlyx;
3889 PL_reglastparen = &rex->lastparen;
3890 PL_reglastcloseparen = &rex->lastcloseparen;
3892 /* XXXX This is too dramatic a measure... */
3894 if ( nochange_depth )
3899 case EVAL_AB_fail: /* unsuccessfully ran A or B in (??{A})B */
3900 /* note: this is called twice; first after popping B, then A */
3901 PL_reg_flags ^= ST.toggle_reg_flags;
3902 ReREFCNT_dec(rex_sv);
3903 SETREX(rex_sv,ST.prev_rex);
3904 rex = (struct regexp *)SvANY(rex_sv);
3905 rexi = RXi_GET(rex);
3906 PL_reglastparen = &rex->lastparen;
3907 PL_reglastcloseparen = &rex->lastcloseparen;
3909 PL_reginput = locinput;
3910 REGCP_UNWIND(ST.lastcp);
3912 cur_eval = ST.prev_eval;
3913 cur_curlyx = ST.prev_curlyx;
3914 /* XXXX This is too dramatic a measure... */
3916 if ( nochange_depth )
3922 n = ARG(scan); /* which paren pair */
3923 PL_reg_start_tmp[n] = locinput;
3929 n = ARG(scan); /* which paren pair */
3930 PL_regoffs[n].start = PL_reg_start_tmp[n] - PL_bostr;
3931 PL_regoffs[n].end = locinput - PL_bostr;
3932 /*if (n > PL_regsize)
3934 if (n > *PL_reglastparen)
3935 *PL_reglastparen = n;
3936 *PL_reglastcloseparen = n;
3937 if (cur_eval && cur_eval->u.eval.close_paren == n) {
3945 cursor && OP(cursor)!=END;
3946 cursor=regnext(cursor))
3948 if ( OP(cursor)==CLOSE ){
3950 if ( n <= lastopen ) {
3952 = PL_reg_start_tmp[n] - PL_bostr;
3953 PL_regoffs[n].end = locinput - PL_bostr;
3954 /*if (n > PL_regsize)
3956 if (n > *PL_reglastparen)
3957 *PL_reglastparen = n;
3958 *PL_reglastcloseparen = n;
3959 if ( n == ARG(scan) || (cur_eval &&
3960 cur_eval->u.eval.close_paren == n))
3969 n = ARG(scan); /* which paren pair */
3970 sw = (bool)(*PL_reglastparen >= n && PL_regoffs[n].end != -1);
3973 /* reg_check_named_buff_matched returns 0 for no match */
3974 sw = (bool)(0 < reg_check_named_buff_matched(rex,scan));
3978 sw = (cur_eval && (!n || cur_eval->u.eval.close_paren == n));
3984 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
3986 next = NEXTOPER(NEXTOPER(scan));
3988 next = scan + ARG(scan);
3989 if (OP(next) == IFTHEN) /* Fake one. */
3990 next = NEXTOPER(NEXTOPER(next));
3994 logical = scan->flags;
3997 /*******************************************************************
3999 The CURLYX/WHILEM pair of ops handle the most generic case of the /A*B/
4000 pattern, where A and B are subpatterns. (For simple A, CURLYM or
4001 STAR/PLUS/CURLY/CURLYN are used instead.)
4003 A*B is compiled as <CURLYX><A><WHILEM><B>
4005 On entry to the subpattern, CURLYX is called. This pushes a CURLYX
4006 state, which contains the current count, initialised to -1. It also sets
4007 cur_curlyx to point to this state, with any previous value saved in the
4010 CURLYX then jumps straight to the WHILEM op, rather than executing A,
4011 since the pattern may possibly match zero times (i.e. it's a while {} loop
4012 rather than a do {} while loop).
4014 Each entry to WHILEM represents a successful match of A. The count in the
4015 CURLYX block is incremented, another WHILEM state is pushed, and execution
4016 passes to A or B depending on greediness and the current count.
4018 For example, if matching against the string a1a2a3b (where the aN are
4019 substrings that match /A/), then the match progresses as follows: (the
4020 pushed states are interspersed with the bits of strings matched so far):
4023 <CURLYX cnt=0><WHILEM>
4024 <CURLYX cnt=1><WHILEM> a1 <WHILEM>
4025 <CURLYX cnt=2><WHILEM> a1 <WHILEM> a2 <WHILEM>
4026 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM>
4027 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM> b
4029 (Contrast this with something like CURLYM, which maintains only a single
4033 a1 <CURLYM cnt=1> a2
4034 a1 a2 <CURLYM cnt=2> a3
4035 a1 a2 a3 <CURLYM cnt=3> b
4038 Each WHILEM state block marks a point to backtrack to upon partial failure
4039 of A or B, and also contains some minor state data related to that
4040 iteration. The CURLYX block, pointed to by cur_curlyx, contains the
4041 overall state, such as the count, and pointers to the A and B ops.
4043 This is complicated slightly by nested CURLYX/WHILEM's. Since cur_curlyx
4044 must always point to the *current* CURLYX block, the rules are:
4046 When executing CURLYX, save the old cur_curlyx in the CURLYX state block,
4047 and set cur_curlyx to point the new block.
4049 When popping the CURLYX block after a successful or unsuccessful match,
4050 restore the previous cur_curlyx.
4052 When WHILEM is about to execute B, save the current cur_curlyx, and set it
4053 to the outer one saved in the CURLYX block.
4055 When popping the WHILEM block after a successful or unsuccessful B match,
4056 restore the previous cur_curlyx.
4058 Here's an example for the pattern (AI* BI)*BO
4059 I and O refer to inner and outer, C and W refer to CURLYX and WHILEM:
4062 curlyx backtrack stack
4063 ------ ---------------
4065 CO <CO prev=NULL> <WO>
4066 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
4067 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
4068 NULL <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi <WO prev=CO> bo
4070 At this point the pattern succeeds, and we work back down the stack to
4071 clean up, restoring as we go:
4073 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
4074 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
4075 CO <CO prev=NULL> <WO>
4078 *******************************************************************/
4080 #define ST st->u.curlyx
4082 case CURLYX: /* start of /A*B/ (for complex A) */
4084 /* No need to save/restore up to this paren */
4085 I32 parenfloor = scan->flags;
4087 assert(next); /* keep Coverity happy */
4088 if (OP(PREVOPER(next)) == NOTHING) /* LONGJMP */
4091 /* XXXX Probably it is better to teach regpush to support
4092 parenfloor > PL_regsize... */
4093 if (parenfloor > (I32)*PL_reglastparen)
4094 parenfloor = *PL_reglastparen; /* Pessimization... */
4096 ST.prev_curlyx= cur_curlyx;
4098 ST.cp = PL_savestack_ix;
4100 /* these fields contain the state of the current curly.
4101 * they are accessed by subsequent WHILEMs */
4102 ST.parenfloor = parenfloor;
4103 ST.min = ARG1(scan);
4104 ST.max = ARG2(scan);
4105 ST.A = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
4109 ST.count = -1; /* this will be updated by WHILEM */
4110 ST.lastloc = NULL; /* this will be updated by WHILEM */
4112 PL_reginput = locinput;
4113 PUSH_YES_STATE_GOTO(CURLYX_end, PREVOPER(next));
4117 case CURLYX_end: /* just finished matching all of A*B */
4118 cur_curlyx = ST.prev_curlyx;
4122 case CURLYX_end_fail: /* just failed to match all of A*B */
4124 cur_curlyx = ST.prev_curlyx;
4130 #define ST st->u.whilem
4132 case WHILEM: /* just matched an A in /A*B/ (for complex A) */
4134 /* see the discussion above about CURLYX/WHILEM */
4136 assert(cur_curlyx); /* keep Coverity happy */
4137 n = ++cur_curlyx->u.curlyx.count; /* how many A's matched */
4138 ST.save_lastloc = cur_curlyx->u.curlyx.lastloc;
4139 ST.cache_offset = 0;
4142 PL_reginput = locinput;
4144 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4145 "%*s whilem: matched %ld out of %ld..%ld\n",
4146 REPORT_CODE_OFF+depth*2, "", (long)n,
4147 (long)cur_curlyx->u.curlyx.min,
4148 (long)cur_curlyx->u.curlyx.max)
4151 /* First just match a string of min A's. */
4153 if (n < cur_curlyx->u.curlyx.min) {
4154 cur_curlyx->u.curlyx.lastloc = locinput;
4155 PUSH_STATE_GOTO(WHILEM_A_pre, cur_curlyx->u.curlyx.A);
4159 /* If degenerate A matches "", assume A done. */
4161 if (locinput == cur_curlyx->u.curlyx.lastloc) {
4162 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4163 "%*s whilem: empty match detected, trying continuation...\n",
4164 REPORT_CODE_OFF+depth*2, "")
4166 goto do_whilem_B_max;
4169 /* super-linear cache processing */
4173 if (!PL_reg_maxiter) {
4174 /* start the countdown: Postpone detection until we
4175 * know the match is not *that* much linear. */
4176 PL_reg_maxiter = (PL_regeol - PL_bostr + 1) * (scan->flags>>4);
4177 /* possible overflow for long strings and many CURLYX's */
4178 if (PL_reg_maxiter < 0)
4179 PL_reg_maxiter = I32_MAX;
4180 PL_reg_leftiter = PL_reg_maxiter;
4183 if (PL_reg_leftiter-- == 0) {
4184 /* initialise cache */
4185 const I32 size = (PL_reg_maxiter + 7)/8;
4186 if (PL_reg_poscache) {
4187 if ((I32)PL_reg_poscache_size < size) {
4188 Renew(PL_reg_poscache, size, char);
4189 PL_reg_poscache_size = size;
4191 Zero(PL_reg_poscache, size, char);
4194 PL_reg_poscache_size = size;
4195 Newxz(PL_reg_poscache, size, char);
4197 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4198 "%swhilem: Detected a super-linear match, switching on caching%s...\n",
4199 PL_colors[4], PL_colors[5])
4203 if (PL_reg_leftiter < 0) {
4204 /* have we already failed at this position? */
4206 offset = (scan->flags & 0xf) - 1
4207 + (locinput - PL_bostr) * (scan->flags>>4);
4208 mask = 1 << (offset % 8);
4210 if (PL_reg_poscache[offset] & mask) {
4211 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4212 "%*s whilem: (cache) already tried at this position...\n",
4213 REPORT_CODE_OFF+depth*2, "")
4215 sayNO; /* cache records failure */
4217 ST.cache_offset = offset;
4218 ST.cache_mask = mask;
4222 /* Prefer B over A for minimal matching. */
4224 if (cur_curlyx->u.curlyx.minmod) {
4225 ST.save_curlyx = cur_curlyx;
4226 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4227 ST.cp = regcppush(ST.save_curlyx->u.curlyx.parenfloor);
4228 REGCP_SET(ST.lastcp);
4229 PUSH_YES_STATE_GOTO(WHILEM_B_min, ST.save_curlyx->u.curlyx.B);
4233 /* Prefer A over B for maximal matching. */
4235 if (n < cur_curlyx->u.curlyx.max) { /* More greed allowed? */
4236 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4237 cur_curlyx->u.curlyx.lastloc = locinput;
4238 REGCP_SET(ST.lastcp);
4239 PUSH_STATE_GOTO(WHILEM_A_max, cur_curlyx->u.curlyx.A);
4242 goto do_whilem_B_max;
4246 case WHILEM_B_min: /* just matched B in a minimal match */
4247 case WHILEM_B_max: /* just matched B in a maximal match */
4248 cur_curlyx = ST.save_curlyx;
4252 case WHILEM_B_max_fail: /* just failed to match B in a maximal match */
4253 cur_curlyx = ST.save_curlyx;
4254 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4255 cur_curlyx->u.curlyx.count--;
4259 case WHILEM_A_min_fail: /* just failed to match A in a minimal match */
4260 REGCP_UNWIND(ST.lastcp);
4263 case WHILEM_A_pre_fail: /* just failed to match even minimal A */
4264 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4265 cur_curlyx->u.curlyx.count--;
4269 case WHILEM_A_max_fail: /* just failed to match A in a maximal match */
4270 REGCP_UNWIND(ST.lastcp);
4271 regcppop(rex); /* Restore some previous $<digit>s? */
4272 PL_reginput = locinput;
4273 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4274 "%*s whilem: failed, trying continuation...\n",
4275 REPORT_CODE_OFF+depth*2, "")
4278 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4279 && ckWARN(WARN_REGEXP)
4280 && !(PL_reg_flags & RF_warned))
4282 PL_reg_flags |= RF_warned;
4283 Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s limit (%d) exceeded",
4284 "Complex regular subexpression recursion",
4289 ST.save_curlyx = cur_curlyx;
4290 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4291 PUSH_YES_STATE_GOTO(WHILEM_B_max, ST.save_curlyx->u.curlyx.B);
4294 case WHILEM_B_min_fail: /* just failed to match B in a minimal match */
4295 cur_curlyx = ST.save_curlyx;
4296 REGCP_UNWIND(ST.lastcp);
4299 if (cur_curlyx->u.curlyx.count >= cur_curlyx->u.curlyx.max) {
4300 /* Maximum greed exceeded */
4301 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4302 && ckWARN(WARN_REGEXP)
4303 && !(PL_reg_flags & RF_warned))
4305 PL_reg_flags |= RF_warned;
4306 Perl_warner(aTHX_ packWARN(WARN_REGEXP),
4307 "%s limit (%d) exceeded",
4308 "Complex regular subexpression recursion",
4311 cur_curlyx->u.curlyx.count--;
4315 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4316 "%*s trying longer...\n", REPORT_CODE_OFF+depth*2, "")
4318 /* Try grabbing another A and see if it helps. */
4319 PL_reginput = locinput;
4320 cur_curlyx->u.curlyx.lastloc = locinput;
4321 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4322 REGCP_SET(ST.lastcp);
4323 PUSH_STATE_GOTO(WHILEM_A_min, ST.save_curlyx->u.curlyx.A);
4327 #define ST st->u.branch
4329 case BRANCHJ: /* /(...|A|...)/ with long next pointer */
4330 next = scan + ARG(scan);
4333 scan = NEXTOPER(scan);
4336 case BRANCH: /* /(...|A|...)/ */
4337 scan = NEXTOPER(scan); /* scan now points to inner node */
4338 ST.lastparen = *PL_reglastparen;
4339 ST.next_branch = next;
4341 PL_reginput = locinput;
4343 /* Now go into the branch */
4345 PUSH_YES_STATE_GOTO(BRANCH_next, scan);
4347 PUSH_STATE_GOTO(BRANCH_next, scan);
4351 PL_reginput = locinput;
4352 sv_yes_mark = st->u.mark.mark_name = scan->flags ? NULL :
4353 MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
4354 PUSH_STATE_GOTO(CUTGROUP_next,next);
4356 case CUTGROUP_next_fail:
4359 if (st->u.mark.mark_name)
4360 sv_commit = st->u.mark.mark_name;
4366 case BRANCH_next_fail: /* that branch failed; try the next, if any */
4371 REGCP_UNWIND(ST.cp);
4372 for (n = *PL_reglastparen; n > ST.lastparen; n--)
4373 PL_regoffs[n].end = -1;
4374 *PL_reglastparen = n;
4375 /*dmq: *PL_reglastcloseparen = n; */
4376 scan = ST.next_branch;
4377 /* no more branches? */
4378 if (!scan || (OP(scan) != BRANCH && OP(scan) != BRANCHJ)) {
4380 PerlIO_printf( Perl_debug_log,
4381 "%*s %sBRANCH failed...%s\n",
4382 REPORT_CODE_OFF+depth*2, "",
4388 continue; /* execute next BRANCH[J] op */
4396 #define ST st->u.curlym
4398 case CURLYM: /* /A{m,n}B/ where A is fixed-length */
4400 /* This is an optimisation of CURLYX that enables us to push
4401 * only a single backtracking state, no matter now many matches
4402 * there are in {m,n}. It relies on the pattern being constant
4403 * length, with no parens to influence future backrefs
4407 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
4409 /* if paren positive, emulate an OPEN/CLOSE around A */
4411 U32 paren = ST.me->flags;
4412 if (paren > PL_regsize)
4414 if (paren > *PL_reglastparen)
4415 *PL_reglastparen = paren;
4416 scan += NEXT_OFF(scan); /* Skip former OPEN. */
4424 ST.c1 = CHRTEST_UNINIT;
4427 if (!(ST.minmod ? ARG1(ST.me) : ARG2(ST.me))) /* min/max */
4430 curlym_do_A: /* execute the A in /A{m,n}B/ */
4431 PL_reginput = locinput;
4432 PUSH_YES_STATE_GOTO(CURLYM_A, ST.A); /* match A */
4435 case CURLYM_A: /* we've just matched an A */
4436 locinput = st->locinput;
4437 nextchr = UCHARAT(locinput);
4440 /* after first match, determine A's length: u.curlym.alen */
4441 if (ST.count == 1) {
4442 if (PL_reg_match_utf8) {
4444 while (s < PL_reginput) {
4450 ST.alen = PL_reginput - locinput;
4453 ST.count = ST.minmod ? ARG1(ST.me) : ARG2(ST.me);
4456 PerlIO_printf(Perl_debug_log,
4457 "%*s CURLYM now matched %"IVdf" times, len=%"IVdf"...\n",
4458 (int)(REPORT_CODE_OFF+(depth*2)), "",
4459 (IV) ST.count, (IV)ST.alen)
4462 locinput = PL_reginput;
4464 if (cur_eval && cur_eval->u.eval.close_paren &&
4465 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
4468 if ( ST.count < (ST.minmod ? ARG1(ST.me) : ARG2(ST.me)) )
4469 goto curlym_do_A; /* try to match another A */
4470 goto curlym_do_B; /* try to match B */
4472 case CURLYM_A_fail: /* just failed to match an A */
4473 REGCP_UNWIND(ST.cp);
4475 if (ST.minmod || ST.count < ARG1(ST.me) /* min*/
4476 || (cur_eval && cur_eval->u.eval.close_paren &&
4477 cur_eval->u.eval.close_paren == (U32)ST.me->flags))
4480 curlym_do_B: /* execute the B in /A{m,n}B/ */
4481 PL_reginput = locinput;
4482 if (ST.c1 == CHRTEST_UNINIT) {
4483 /* calculate c1 and c2 for possible match of 1st char
4484 * following curly */
4485 ST.c1 = ST.c2 = CHRTEST_VOID;
4486 if (HAS_TEXT(ST.B) || JUMPABLE(ST.B)) {
4487 regnode *text_node = ST.B;
4488 if (! HAS_TEXT(text_node))
4489 FIND_NEXT_IMPT(text_node);
4492 (HAS_TEXT(text_node) && PL_regkind[OP(text_node)] == EXACT)
4494 But the former is redundant in light of the latter.
4496 if this changes back then the macro for
4497 IS_TEXT and friends need to change.
4499 if (PL_regkind[OP(text_node)] == EXACT)
4502 ST.c1 = (U8)*STRING(text_node);
4504 (IS_TEXTF(text_node))
4506 : (IS_TEXTFL(text_node))
4507 ? PL_fold_locale[ST.c1]
4514 PerlIO_printf(Perl_debug_log,
4515 "%*s CURLYM trying tail with matches=%"IVdf"...\n",
4516 (int)(REPORT_CODE_OFF+(depth*2)),
4519 if (ST.c1 != CHRTEST_VOID
4520 && UCHARAT(PL_reginput) != ST.c1
4521 && UCHARAT(PL_reginput) != ST.c2)
4523 /* simulate B failing */
4525 PerlIO_printf(Perl_debug_log,
4526 "%*s CURLYM Fast bail c1=%"IVdf" c2=%"IVdf"\n",
4527 (int)(REPORT_CODE_OFF+(depth*2)),"",
4530 state_num = CURLYM_B_fail;
4531 goto reenter_switch;
4535 /* mark current A as captured */
4536 I32 paren = ST.me->flags;
4538 PL_regoffs[paren].start
4539 = HOPc(PL_reginput, -ST.alen) - PL_bostr;
4540 PL_regoffs[paren].end = PL_reginput - PL_bostr;
4541 /*dmq: *PL_reglastcloseparen = paren; */
4544 PL_regoffs[paren].end = -1;
4545 if (cur_eval && cur_eval->u.eval.close_paren &&
4546 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
4555 PUSH_STATE_GOTO(CURLYM_B, ST.B); /* match B */
4558 case CURLYM_B_fail: /* just failed to match a B */
4559 REGCP_UNWIND(ST.cp);
4561 if (ST.count == ARG2(ST.me) /* max */)
4563 goto curlym_do_A; /* try to match a further A */
4565 /* backtrack one A */
4566 if (ST.count == ARG1(ST.me) /* min */)
4569 locinput = HOPc(locinput, -ST.alen);
4570 goto curlym_do_B; /* try to match B */
4573 #define ST st->u.curly
4575 #define CURLY_SETPAREN(paren, success) \
4578 PL_regoffs[paren].start = HOPc(locinput, -1) - PL_bostr; \
4579 PL_regoffs[paren].end = locinput - PL_bostr; \
4580 *PL_reglastcloseparen = paren; \
4583 PL_regoffs[paren].end = -1; \
4586 case STAR: /* /A*B/ where A is width 1 */
4590 scan = NEXTOPER(scan);
4592 case PLUS: /* /A+B/ where A is width 1 */
4596 scan = NEXTOPER(scan);
4598 case CURLYN: /* /(A){m,n}B/ where A is width 1 */
4599 ST.paren = scan->flags; /* Which paren to set */
4600 if (ST.paren > PL_regsize)
4601 PL_regsize = ST.paren;
4602 if (ST.paren > *PL_reglastparen)
4603 *PL_reglastparen = ST.paren;
4604 ST.min = ARG1(scan); /* min to match */
4605 ST.max = ARG2(scan); /* max to match */
4606 if (cur_eval && cur_eval->u.eval.close_paren &&
4607 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4611 scan = regnext(NEXTOPER(scan) + NODE_STEP_REGNODE);
4613 case CURLY: /* /A{m,n}B/ where A is width 1 */
4615 ST.min = ARG1(scan); /* min to match */
4616 ST.max = ARG2(scan); /* max to match */
4617 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
4620 * Lookahead to avoid useless match attempts
4621 * when we know what character comes next.
4623 * Used to only do .*x and .*?x, but now it allows
4624 * for )'s, ('s and (?{ ... })'s to be in the way
4625 * of the quantifier and the EXACT-like node. -- japhy
4628 if (ST.min > ST.max) /* XXX make this a compile-time check? */
4630 if (HAS_TEXT(next) || JUMPABLE(next)) {
4632 regnode *text_node = next;
4634 if (! HAS_TEXT(text_node))
4635 FIND_NEXT_IMPT(text_node);
4637 if (! HAS_TEXT(text_node))
4638 ST.c1 = ST.c2 = CHRTEST_VOID;
4640 if ( PL_regkind[OP(text_node)] != EXACT ) {
4641 ST.c1 = ST.c2 = CHRTEST_VOID;
4642 goto assume_ok_easy;
4645 s = (U8*)STRING(text_node);
4647 /* Currently we only get here when
4649 PL_rekind[OP(text_node)] == EXACT
4651 if this changes back then the macro for IS_TEXT and
4652 friends need to change. */
4655 if (IS_TEXTF(text_node))
4656 ST.c2 = PL_fold[ST.c1];
4657 else if (IS_TEXTFL(text_node))
4658 ST.c2 = PL_fold_locale[ST.c1];
4661 if (IS_TEXTF(text_node)) {
4662 STRLEN ulen1, ulen2;
4663 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
4664 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
4666 to_utf8_lower((U8*)s, tmpbuf1, &ulen1);
4667 to_utf8_upper((U8*)s, tmpbuf2, &ulen2);
4669 ST.c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXLEN, 0,
4671 0 : UTF8_ALLOW_ANY);
4672 ST.c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXLEN, 0,
4674 0 : UTF8_ALLOW_ANY);
4676 ST.c1 = utf8n_to_uvuni(tmpbuf1, UTF8_MAXBYTES, 0,
4678 ST.c2 = utf8n_to_uvuni(tmpbuf2, UTF8_MAXBYTES, 0,
4683 ST.c2 = ST.c1 = utf8n_to_uvchr(s, UTF8_MAXBYTES, 0,
4690 ST.c1 = ST.c2 = CHRTEST_VOID;
4695 PL_reginput = locinput;
4698 if (ST.min && regrepeat(rex, ST.A, ST.min, depth) < ST.min)
4701 locinput = PL_reginput;
4703 if (ST.c1 == CHRTEST_VOID)
4704 goto curly_try_B_min;
4706 ST.oldloc = locinput;
4708 /* set ST.maxpos to the furthest point along the
4709 * string that could possibly match */
4710 if (ST.max == REG_INFTY) {
4711 ST.maxpos = PL_regeol - 1;
4713 while (UTF8_IS_CONTINUATION(*(U8*)ST.maxpos))
4717 int m = ST.max - ST.min;
4718 for (ST.maxpos = locinput;
4719 m >0 && ST.maxpos + UTF8SKIP(ST.maxpos) <= PL_regeol; m--)
4720 ST.maxpos += UTF8SKIP(ST.maxpos);
4723 ST.maxpos = locinput + ST.max - ST.min;
4724 if (ST.maxpos >= PL_regeol)
4725 ST.maxpos = PL_regeol - 1;
4727 goto curly_try_B_min_known;
4731 ST.count = regrepeat(rex, ST.A, ST.max, depth);
4732 locinput = PL_reginput;
4733 if (ST.count < ST.min)
4735 if ((ST.count > ST.min)
4736 && (PL_regkind[OP(ST.B)] == EOL) && (OP(ST.B) != MEOL))
4738 /* A{m,n} must come at the end of the string, there's
4739 * no point in backing off ... */
4741 /* ...except that $ and \Z can match before *and* after
4742 newline at the end. Consider "\n\n" =~ /\n+\Z\n/.
4743 We may back off by one in this case. */
4744 if (UCHARAT(PL_reginput - 1) == '\n' && OP(ST.B) != EOS)
4748 goto curly_try_B_max;
4753 case CURLY_B_min_known_fail:
4754 /* failed to find B in a non-greedy match where c1,c2 valid */
4755 if (ST.paren && ST.count)
4756 PL_regoffs[ST.paren].end = -1;
4758 PL_reginput = locinput; /* Could be reset... */
4759 REGCP_UNWIND(ST.cp);
4760 /* Couldn't or didn't -- move forward. */
4761 ST.oldloc = locinput;
4763 locinput += UTF8SKIP(locinput);
4767 curly_try_B_min_known:
4768 /* find the next place where 'B' could work, then call B */
4772 n = (ST.oldloc == locinput) ? 0 : 1;
4773 if (ST.c1 == ST.c2) {
4775 /* set n to utf8_distance(oldloc, locinput) */
4776 while (locinput <= ST.maxpos &&
4777 utf8n_to_uvchr((U8*)locinput,
4778 UTF8_MAXBYTES, &len,
4779 uniflags) != (UV)ST.c1) {
4785 /* set n to utf8_distance(oldloc, locinput) */
4786 while (locinput <= ST.maxpos) {
4788 const UV c = utf8n_to_uvchr((U8*)locinput,
4789 UTF8_MAXBYTES, &len,
4791 if (c == (UV)ST.c1 || c == (UV)ST.c2)
4799 if (ST.c1 == ST.c2) {
4800 while (locinput <= ST.maxpos &&
4801 UCHARAT(locinput) != ST.c1)
4805 while (locinput <= ST.maxpos
4806 && UCHARAT(locinput) != ST.c1
4807 && UCHARAT(locinput) != ST.c2)
4810 n = locinput - ST.oldloc;
4812 if (locinput > ST.maxpos)
4814 /* PL_reginput == oldloc now */
4817 if (regrepeat(rex, ST.A, n, depth) < n)
4820 PL_reginput = locinput;
4821 CURLY_SETPAREN(ST.paren, ST.count);
4822 if (cur_eval && cur_eval->u.eval.close_paren &&
4823 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4826 PUSH_STATE_GOTO(CURLY_B_min_known, ST.B);
4831 case CURLY_B_min_fail:
4832 /* failed to find B in a non-greedy match where c1,c2 invalid */
4833 if (ST.paren && ST.count)
4834 PL_regoffs[ST.paren].end = -1;
4836 REGCP_UNWIND(ST.cp);
4837 /* failed -- move forward one */
4838 PL_reginput = locinput;
4839 if (regrepeat(rex, ST.A, 1, depth)) {
4841 locinput = PL_reginput;
4842 if (ST.count <= ST.max || (ST.max == REG_INFTY &&
4843 ST.count > 0)) /* count overflow ? */
4846 CURLY_SETPAREN(ST.paren, ST.count);
4847 if (cur_eval && cur_eval->u.eval.close_paren &&
4848 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4851 PUSH_STATE_GOTO(CURLY_B_min, ST.B);
4859 /* a successful greedy match: now try to match B */
4860 if (cur_eval && cur_eval->u.eval.close_paren &&
4861 cur_eval->u.eval.close_paren == (U32)ST.paren) {
4866 if (ST.c1 != CHRTEST_VOID)
4867 c = do_utf8 ? utf8n_to_uvchr((U8*)PL_reginput,
4868 UTF8_MAXBYTES, 0, uniflags)
4869 : (UV) UCHARAT(PL_reginput);
4870 /* If it could work, try it. */
4871 if (ST.c1 == CHRTEST_VOID || c == (UV)ST.c1 || c == (UV)ST.c2) {
4872 CURLY_SETPAREN(ST.paren, ST.count);
4873 PUSH_STATE_GOTO(CURLY_B_max, ST.B);
4878 case CURLY_B_max_fail:
4879 /* failed to find B in a greedy match */
4880 if (ST.paren && ST.count)
4881 PL_regoffs[ST.paren].end = -1;
4883 REGCP_UNWIND(ST.cp);
4885 if (--ST.count < ST.min)
4887 PL_reginput = locinput = HOPc(locinput, -1);
4888 goto curly_try_B_max;
4895 /* we've just finished A in /(??{A})B/; now continue with B */
4897 st->u.eval.toggle_reg_flags
4898 = cur_eval->u.eval.toggle_reg_flags;
4899 PL_reg_flags ^= st->u.eval.toggle_reg_flags;
4901 st->u.eval.prev_rex = rex_sv; /* inner */
4902 SETREX(rex_sv,cur_eval->u.eval.prev_rex);
4903 rex = (struct regexp *)SvANY(rex_sv);
4904 rexi = RXi_GET(rex);
4905 cur_curlyx = cur_eval->u.eval.prev_curlyx;
4906 ReREFCNT_inc(rex_sv);
4907 st->u.eval.cp = regcppush(0); /* Save *all* the positions. */
4908 REGCP_SET(st->u.eval.lastcp);
4909 PL_reginput = locinput;
4911 /* Restore parens of the outer rex without popping the
4913 tmpix = PL_savestack_ix;
4914 PL_savestack_ix = cur_eval->u.eval.lastcp;
4916 PL_savestack_ix = tmpix;
4918 st->u.eval.prev_eval = cur_eval;
4919 cur_eval = cur_eval->u.eval.prev_eval;
4921 PerlIO_printf(Perl_debug_log, "%*s EVAL trying tail ... %"UVxf"\n",
4922 REPORT_CODE_OFF+depth*2, "",PTR2UV(cur_eval)););
4923 if ( nochange_depth )
4926 PUSH_YES_STATE_GOTO(EVAL_AB,
4927 st->u.eval.prev_eval->u.eval.B); /* match B */
4930 if (locinput < reginfo->till) {
4931 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4932 "%sMatch possible, but length=%ld is smaller than requested=%ld, failing!%s\n",
4934 (long)(locinput - PL_reg_starttry),
4935 (long)(reginfo->till - PL_reg_starttry),
4938 sayNO_SILENT; /* Cannot match: too short. */
4940 PL_reginput = locinput; /* put where regtry can find it */
4941 sayYES; /* Success! */
4943 case SUCCEED: /* successful SUSPEND/UNLESSM/IFMATCH/CURLYM */
4945 PerlIO_printf(Perl_debug_log,
4946 "%*s %ssubpattern success...%s\n",
4947 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5]));
4948 PL_reginput = locinput; /* put where regtry can find it */
4949 sayYES; /* Success! */
4952 #define ST st->u.ifmatch
4954 case SUSPEND: /* (?>A) */
4956 PL_reginput = locinput;
4959 case UNLESSM: /* -ve lookaround: (?!A), or with flags, (?<!A) */
4961 goto ifmatch_trivial_fail_test;
4963 case IFMATCH: /* +ve lookaround: (?=A), or with flags, (?<=A) */
4965 ifmatch_trivial_fail_test:
4967 char * const s = HOPBACKc(locinput, scan->flags);
4972 sw = 1 - (bool)ST.wanted;
4976 next = scan + ARG(scan);
4984 PL_reginput = locinput;
4988 ST.logical = logical;
4989 logical = 0; /* XXX: reset state of logical once it has been saved into ST */
4991 /* execute body of (?...A) */
4992 PUSH_YES_STATE_GOTO(IFMATCH_A, NEXTOPER(NEXTOPER(scan)));
4995 case IFMATCH_A_fail: /* body of (?...A) failed */
4996 ST.wanted = !ST.wanted;
4999 case IFMATCH_A: /* body of (?...A) succeeded */
5001 sw = (bool)ST.wanted;
5003 else if (!ST.wanted)
5006 if (OP(ST.me) == SUSPEND)
5007 locinput = PL_reginput;
5009 locinput = PL_reginput = st->locinput;
5010 nextchr = UCHARAT(locinput);
5012 scan = ST.me + ARG(ST.me);
5015 continue; /* execute B */
5020 next = scan + ARG(scan);
5025 reginfo->cutpoint = PL_regeol;
5028 PL_reginput = locinput;
5030 sv_yes_mark = sv_commit = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5031 PUSH_STATE_GOTO(COMMIT_next,next);
5033 case COMMIT_next_fail:
5040 #define ST st->u.mark
5042 ST.prev_mark = mark_state;
5043 ST.mark_name = sv_commit = sv_yes_mark
5044 = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5046 ST.mark_loc = PL_reginput = locinput;
5047 PUSH_YES_STATE_GOTO(MARKPOINT_next,next);
5049 case MARKPOINT_next:
5050 mark_state = ST.prev_mark;
5053 case MARKPOINT_next_fail:
5054 if (popmark && sv_eq(ST.mark_name,popmark))
5056 if (ST.mark_loc > startpoint)
5057 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
5058 popmark = NULL; /* we found our mark */
5059 sv_commit = ST.mark_name;
5062 PerlIO_printf(Perl_debug_log,
5063 "%*s %ssetting cutpoint to mark:%"SVf"...%s\n",
5064 REPORT_CODE_OFF+depth*2, "",
5065 PL_colors[4], SVfARG(sv_commit), PL_colors[5]);
5068 mark_state = ST.prev_mark;
5069 sv_yes_mark = mark_state ?
5070 mark_state->u.mark.mark_name : NULL;
5074 PL_reginput = locinput;
5076 /* (*SKIP) : if we fail we cut here*/
5077 ST.mark_name = NULL;
5078 ST.mark_loc = locinput;
5079 PUSH_STATE_GOTO(SKIP_next,next);
5081 /* (*SKIP:NAME) : if there is a (*MARK:NAME) fail where it was,
5082 otherwise do nothing. Meaning we need to scan
5084 regmatch_state *cur = mark_state;
5085 SV *find = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5088 if ( sv_eq( cur->u.mark.mark_name,
5091 ST.mark_name = find;
5092 PUSH_STATE_GOTO( SKIP_next, next );
5094 cur = cur->u.mark.prev_mark;
5097 /* Didn't find our (*MARK:NAME) so ignore this (*SKIP:NAME) */
5099 case SKIP_next_fail:
5101 /* (*CUT:NAME) - Set up to search for the name as we
5102 collapse the stack*/
5103 popmark = ST.mark_name;
5105 /* (*CUT) - No name, we cut here.*/
5106 if (ST.mark_loc > startpoint)
5107 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
5108 /* but we set sv_commit to latest mark_name if there
5109 is one so they can test to see how things lead to this
5112 sv_commit=mark_state->u.mark.mark_name;
5120 if ( n == (U32)what_len_TRICKYFOLD(locinput,do_utf8,ln) ) {
5122 } else if ( 0xDF == n && !do_utf8 && !UTF ) {
5125 U8 folded[UTF8_MAXBYTES_CASE+1];
5127 const char * const l = locinput;
5128 char *e = PL_regeol;
5129 to_uni_fold(n, folded, &foldlen);
5131 if (ibcmp_utf8((const char*) folded, 0, foldlen, 1,
5132 l, &e, 0, do_utf8)) {
5137 nextchr = UCHARAT(locinput);
5140 if ((n=is_LNBREAK(locinput,do_utf8))) {
5142 nextchr = UCHARAT(locinput);
5147 #define CASE_CLASS(nAmE) \
5149 if ((n=is_##nAmE(locinput,do_utf8))) { \
5151 nextchr = UCHARAT(locinput); \
5156 if ((n=is_##nAmE(locinput,do_utf8))) { \
5159 locinput += UTF8SKIP(locinput); \
5160 nextchr = UCHARAT(locinput); \
5165 CASE_CLASS(HORIZWS);
5169 PerlIO_printf(Perl_error_log, "%"UVxf" %d\n",
5170 PTR2UV(scan), OP(scan));
5171 Perl_croak(aTHX_ "regexp memory corruption");
5175 /* switch break jumps here */
5176 scan = next; /* prepare to execute the next op and ... */
5177 continue; /* ... jump back to the top, reusing st */
5181 /* push a state that backtracks on success */
5182 st->u.yes.prev_yes_state = yes_state;
5186 /* push a new regex state, then continue at scan */
5188 regmatch_state *newst;
5191 regmatch_state *cur = st;
5192 regmatch_state *curyes = yes_state;
5194 regmatch_slab *slab = PL_regmatch_slab;
5195 for (;curd > -1;cur--,curd--) {
5196 if (cur < SLAB_FIRST(slab)) {
5198 cur = SLAB_LAST(slab);
5200 PerlIO_printf(Perl_error_log, "%*s#%-3d %-10s %s\n",
5201 REPORT_CODE_OFF + 2 + depth * 2,"",
5202 curd, PL_reg_name[cur->resume_state],
5203 (curyes == cur) ? "yes" : ""
5206 curyes = cur->u.yes.prev_yes_state;
5209 DEBUG_STATE_pp("push")
5212 st->locinput = locinput;
5214 if (newst > SLAB_LAST(PL_regmatch_slab))
5215 newst = S_push_slab(aTHX);
5216 PL_regmatch_state = newst;
5218 locinput = PL_reginput;
5219 nextchr = UCHARAT(locinput);
5227 * We get here only if there's trouble -- normally "case END" is
5228 * the terminating point.
5230 Perl_croak(aTHX_ "corrupted regexp pointers");
5236 /* we have successfully completed a subexpression, but we must now
5237 * pop to the state marked by yes_state and continue from there */
5238 assert(st != yes_state);
5240 while (st != yes_state) {
5242 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5243 PL_regmatch_slab = PL_regmatch_slab->prev;
5244 st = SLAB_LAST(PL_regmatch_slab);
5248 DEBUG_STATE_pp("pop (no final)");
5250 DEBUG_STATE_pp("pop (yes)");
5256 while (yes_state < SLAB_FIRST(PL_regmatch_slab)
5257 || yes_state > SLAB_LAST(PL_regmatch_slab))
5259 /* not in this slab, pop slab */
5260 depth -= (st - SLAB_FIRST(PL_regmatch_slab) + 1);
5261 PL_regmatch_slab = PL_regmatch_slab->prev;
5262 st = SLAB_LAST(PL_regmatch_slab);
5264 depth -= (st - yes_state);
5267 yes_state = st->u.yes.prev_yes_state;
5268 PL_regmatch_state = st;
5271 locinput= st->locinput;
5272 nextchr = UCHARAT(locinput);
5274 state_num = st->resume_state + no_final;
5275 goto reenter_switch;
5278 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch successful!%s\n",
5279 PL_colors[4], PL_colors[5]));
5281 if (PL_reg_eval_set) {
5282 /* each successfully executed (?{...}) block does the equivalent of
5283 * local $^R = do {...}
5284 * When popping the save stack, all these locals would be undone;
5285 * bypass this by setting the outermost saved $^R to the latest
5287 if (oreplsv != GvSV(PL_replgv))
5288 sv_setsv(oreplsv, GvSV(PL_replgv));
5295 PerlIO_printf(Perl_debug_log,
5296 "%*s %sfailed...%s\n",
5297 REPORT_CODE_OFF+depth*2, "",
5298 PL_colors[4], PL_colors[5])
5310 /* there's a previous state to backtrack to */
5312 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5313 PL_regmatch_slab = PL_regmatch_slab->prev;
5314 st = SLAB_LAST(PL_regmatch_slab);
5316 PL_regmatch_state = st;
5317 locinput= st->locinput;
5318 nextchr = UCHARAT(locinput);
5320 DEBUG_STATE_pp("pop");
5322 if (yes_state == st)
5323 yes_state = st->u.yes.prev_yes_state;
5325 state_num = st->resume_state + 1; /* failure = success + 1 */
5326 goto reenter_switch;
5331 if (rex->intflags & PREGf_VERBARG_SEEN) {
5332 SV *sv_err = get_sv("REGERROR", 1);
5333 SV *sv_mrk = get_sv("REGMARK", 1);
5335 sv_commit = &PL_sv_no;
5337 sv_yes_mark = &PL_sv_yes;
5340 sv_commit = &PL_sv_yes;
5341 sv_yes_mark = &PL_sv_no;
5343 sv_setsv(sv_err, sv_commit);
5344 sv_setsv(sv_mrk, sv_yes_mark);
5347 /* clean up; in particular, free all slabs above current one */
5348 LEAVE_SCOPE(oldsave);
5354 - regrepeat - repeatedly match something simple, report how many
5357 * [This routine now assumes that it will only match on things of length 1.
5358 * That was true before, but now we assume scan - reginput is the count,
5359 * rather than incrementing count on every character. [Er, except utf8.]]
5362 S_regrepeat(pTHX_ const regexp *prog, const regnode *p, I32 max, int depth)
5365 register char *scan;
5367 register char *loceol = PL_regeol;
5368 register I32 hardcount = 0;
5369 register bool do_utf8 = PL_reg_match_utf8;
5371 PERL_UNUSED_ARG(depth);
5374 PERL_ARGS_ASSERT_REGREPEAT;
5377 if (max == REG_INFTY)
5379 else if (max < loceol - scan)
5380 loceol = scan + max;
5385 while (scan < loceol && hardcount < max && *scan != '\n') {
5386 scan += UTF8SKIP(scan);
5390 while (scan < loceol && *scan != '\n')
5397 while (scan < loceol && hardcount < max) {
5398 scan += UTF8SKIP(scan);
5408 case EXACT: /* length of string is 1 */
5410 while (scan < loceol && UCHARAT(scan) == c)
5413 case EXACTF: /* length of string is 1 */
5415 while (scan < loceol &&
5416 (UCHARAT(scan) == c || UCHARAT(scan) == PL_fold[c]))
5419 case EXACTFL: /* length of string is 1 */
5420 PL_reg_flags |= RF_tainted;
5422 while (scan < loceol &&
5423 (UCHARAT(scan) == c || UCHARAT(scan) == PL_fold_locale[c]))
5429 while (hardcount < max && scan < loceol &&
5430 reginclass(prog, p, (U8*)scan, 0, do_utf8)) {
5431 scan += UTF8SKIP(scan);
5435 while (scan < loceol && REGINCLASS(prog, p, (U8*)scan))
5442 LOAD_UTF8_CHARCLASS_ALNUM();
5443 while (hardcount < max && scan < loceol &&
5444 swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
5445 scan += UTF8SKIP(scan);
5449 while (scan < loceol && isALNUM(*scan))
5454 PL_reg_flags |= RF_tainted;
5457 while (hardcount < max && scan < loceol &&
5458 isALNUM_LC_utf8((U8*)scan)) {
5459 scan += UTF8SKIP(scan);
5463 while (scan < loceol && isALNUM_LC(*scan))
5470 LOAD_UTF8_CHARCLASS_ALNUM();
5471 while (hardcount < max && scan < loceol &&
5472 !swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
5473 scan += UTF8SKIP(scan);
5477 while (scan < loceol && !isALNUM(*scan))
5482 PL_reg_flags |= RF_tainted;
5485 while (hardcount < max && scan < loceol &&
5486 !isALNUM_LC_utf8((U8*)scan)) {
5487 scan += UTF8SKIP(scan);
5491 while (scan < loceol && !isALNUM_LC(*scan))
5498 LOAD_UTF8_CHARCLASS_SPACE();
5499 while (hardcount < max && scan < loceol &&
5501 swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
5502 scan += UTF8SKIP(scan);
5506 while (scan < loceol && isSPACE(*scan))
5511 PL_reg_flags |= RF_tainted;
5514 while (hardcount < max && scan < loceol &&
5515 (*scan == ' ' || isSPACE_LC_utf8((U8*)scan))) {
5516 scan += UTF8SKIP(scan);
5520 while (scan < loceol && isSPACE_LC(*scan))
5527 LOAD_UTF8_CHARCLASS_SPACE();
5528 while (hardcount < max && scan < loceol &&
5530 swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
5531 scan += UTF8SKIP(scan);
5535 while (scan < loceol && !isSPACE(*scan))
5540 PL_reg_flags |= RF_tainted;
5543 while (hardcount < max && scan < loceol &&
5544 !(*scan == ' ' || isSPACE_LC_utf8((U8*)scan))) {
5545 scan += UTF8SKIP(scan);
5549 while (scan < loceol && !isSPACE_LC(*scan))
5556 LOAD_UTF8_CHARCLASS_DIGIT();
5557 while (hardcount < max && scan < loceol &&
5558 swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
5559 scan += UTF8SKIP(scan);
5563 while (scan < loceol && isDIGIT(*scan))
5570 LOAD_UTF8_CHARCLASS_DIGIT();
5571 while (hardcount < max && scan < loceol &&
5572 !swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
5573 scan += UTF8SKIP(scan);
5577 while (scan < loceol && !isDIGIT(*scan))
5583 while (hardcount < max && scan < loceol && (c=is_LNBREAK_utf8(scan))) {
5589 LNBREAK can match two latin chars, which is ok,
5590 because we have a null terminated string, but we
5591 have to use hardcount in this situation
5593 while (scan < loceol && (c=is_LNBREAK_latin1(scan))) {
5602 while (hardcount < max && scan < loceol && (c=is_HORIZWS_utf8(scan))) {
5607 while (scan < loceol && is_HORIZWS_latin1(scan))
5614 while (hardcount < max && scan < loceol && !is_HORIZWS_utf8(scan)) {
5615 scan += UTF8SKIP(scan);
5619 while (scan < loceol && !is_HORIZWS_latin1(scan))
5627 while (hardcount < max && scan < loceol && (c=is_VERTWS_utf8(scan))) {
5632 while (scan < loceol && is_VERTWS_latin1(scan))
5640 while (hardcount < max && scan < loceol && !is_VERTWS_utf8(scan)) {
5641 scan += UTF8SKIP(scan);
5645 while (scan < loceol && !is_VERTWS_latin1(scan))
5651 default: /* Called on something of 0 width. */
5652 break; /* So match right here or not at all. */
5658 c = scan - PL_reginput;
5662 GET_RE_DEBUG_FLAGS_DECL;
5664 SV * const prop = sv_newmortal();
5665 regprop(prog, prop, p);
5666 PerlIO_printf(Perl_debug_log,
5667 "%*s %s can match %"IVdf" times out of %"IVdf"...\n",
5668 REPORT_CODE_OFF + depth*2, "", SvPVX_const(prop),(IV)c,(IV)max);
5676 #if !defined(PERL_IN_XSUB_RE) || defined(PLUGGABLE_RE_EXTENSION)
5678 - regclass_swash - prepare the utf8 swash
5682 Perl_regclass_swash(pTHX_ const regexp *prog, register const regnode* node, bool doinit, SV** listsvp, SV **altsvp)
5688 RXi_GET_DECL(prog,progi);
5689 const struct reg_data * const data = prog ? progi->data : NULL;
5691 PERL_ARGS_ASSERT_REGCLASS_SWASH;
5693 if (data && data->count) {
5694 const U32 n = ARG(node);
5696 if (data->what[n] == 's') {
5697 SV * const rv = MUTABLE_SV(data->data[n]);
5698 AV * const av = MUTABLE_AV(SvRV(rv));
5699 SV **const ary = AvARRAY(av);
5702 /* See the end of regcomp.c:S_regclass() for
5703 * documentation of these array elements. */
5706 a = SvROK(ary[1]) ? &ary[1] : NULL;
5707 b = SvTYPE(ary[2]) == SVt_PVAV ? &ary[2] : NULL;
5711 else if (si && doinit) {
5712 sw = swash_init("utf8", "", si, 1, 0);
5713 (void)av_store(av, 1, sw);
5730 - reginclass - determine if a character falls into a character class
5732 The n is the ANYOF regnode, the p is the target string, lenp
5733 is pointer to the maximum length of how far to go in the p
5734 (if the lenp is zero, UTF8SKIP(p) is used),
5735 do_utf8 tells whether the target string is in UTF-8.
5740 S_reginclass(pTHX_ const regexp *prog, register const regnode *n, register const U8* p, STRLEN* lenp, register bool do_utf8)
5743 const char flags = ANYOF_FLAGS(n);
5749 PERL_ARGS_ASSERT_REGINCLASS;
5751 if (do_utf8 && !UTF8_IS_INVARIANT(c)) {
5752 c = utf8n_to_uvchr(p, UTF8_MAXBYTES, &len,
5753 (UTF8_ALLOW_DEFAULT & UTF8_ALLOW_ANYUV) | UTF8_CHECK_ONLY);
5754 /* see [perl #37836] for UTF8_ALLOW_ANYUV */
5755 if (len == (STRLEN)-1)
5756 Perl_croak(aTHX_ "Malformed UTF-8 character (fatal)");
5759 plen = lenp ? *lenp : UNISKIP(NATIVE_TO_UNI(c));
5760 if (do_utf8 || (flags & ANYOF_UNICODE)) {
5763 if (do_utf8 && !ANYOF_RUNTIME(n)) {
5764 if (len != (STRLEN)-1 && c < 256 && ANYOF_BITMAP_TEST(n, c))
5767 if (!match && do_utf8 && (flags & ANYOF_UNICODE_ALL) && c >= 256)
5771 SV * const sw = regclass_swash(prog, n, TRUE, 0, (SV**)&av);
5774 if (swash_fetch(sw, p, do_utf8))
5776 else if (flags & ANYOF_FOLD) {
5777 if (!match && lenp && av) {
5779 for (i = 0; i <= av_len(av); i++) {
5780 SV* const sv = *av_fetch(av, i, FALSE);
5782 const char * const s = SvPV_const(sv, len);
5784 if (len <= plen && memEQ(s, (char*)p, len)) {
5792 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
5795 to_utf8_fold(p, tmpbuf, &tmplen);
5796 if (swash_fetch(sw, tmpbuf, do_utf8))
5802 if (match && lenp && *lenp == 0)
5803 *lenp = UNISKIP(NATIVE_TO_UNI(c));
5805 if (!match && c < 256) {
5806 if (ANYOF_BITMAP_TEST(n, c))
5808 else if (flags & ANYOF_FOLD) {
5811 if (flags & ANYOF_LOCALE) {
5812 PL_reg_flags |= RF_tainted;
5813 f = PL_fold_locale[c];
5817 if (f != c && ANYOF_BITMAP_TEST(n, f))
5821 if (!match && (flags & ANYOF_CLASS)) {
5822 PL_reg_flags |= RF_tainted;
5824 (ANYOF_CLASS_TEST(n, ANYOF_ALNUM) && isALNUM_LC(c)) ||
5825 (ANYOF_CLASS_TEST(n, ANYOF_NALNUM) && !isALNUM_LC(c)) ||
5826 (ANYOF_CLASS_TEST(n, ANYOF_SPACE) && isSPACE_LC(c)) ||
5827 (ANYOF_CLASS_TEST(n, ANYOF_NSPACE) && !isSPACE_LC(c)) ||
5828 (ANYOF_CLASS_TEST(n, ANYOF_DIGIT) && isDIGIT_LC(c)) ||
5829 (ANYOF_CLASS_TEST(n, ANYOF_NDIGIT) && !isDIGIT_LC(c)) ||
5830 (ANYOF_CLASS_TEST(n, ANYOF_ALNUMC) && isALNUMC_LC(c)) ||
5831 (ANYOF_CLASS_TEST(n, ANYOF_NALNUMC) && !isALNUMC_LC(c)) ||
5832 (ANYOF_CLASS_TEST(n, ANYOF_ALPHA) && isALPHA_LC(c)) ||
5833 (ANYOF_CLASS_TEST(n, ANYOF_NALPHA) && !isALPHA_LC(c)) ||
5834 (ANYOF_CLASS_TEST(n, ANYOF_ASCII) && isASCII(c)) ||
5835 (ANYOF_CLASS_TEST(n, ANYOF_NASCII) && !isASCII(c)) ||
5836 (ANYOF_CLASS_TEST(n, ANYOF_CNTRL) && isCNTRL_LC(c)) ||
5837 (ANYOF_CLASS_TEST(n, ANYOF_NCNTRL) && !isCNTRL_LC(c)) ||
5838 (ANYOF_CLASS_TEST(n, ANYOF_GRAPH) && isGRAPH_LC(c)) ||
5839 (ANYOF_CLASS_TEST(n, ANYOF_NGRAPH) && !isGRAPH_LC(c)) ||
5840 (ANYOF_CLASS_TEST(n, ANYOF_LOWER) && isLOWER_LC(c)) ||
5841 (ANYOF_CLASS_TEST(n, ANYOF_NLOWER) && !isLOWER_LC(c)) ||
5842 (ANYOF_CLASS_TEST(n, ANYOF_PRINT) && isPRINT_LC(c)) ||
5843 (ANYOF_CLASS_TEST(n, ANYOF_NPRINT) && !isPRINT_LC(c)) ||
5844 (ANYOF_CLASS_TEST(n, ANYOF_PUNCT) && isPUNCT_LC(c)) ||
5845 (ANYOF_CLASS_TEST(n, ANYOF_NPUNCT) && !isPUNCT_LC(c)) ||
5846 (ANYOF_CLASS_TEST(n, ANYOF_UPPER) && isUPPER_LC(c)) ||
5847 (ANYOF_CLASS_TEST(n, ANYOF_NUPPER) && !isUPPER_LC(c)) ||
5848 (ANYOF_CLASS_TEST(n, ANYOF_XDIGIT) && isXDIGIT(c)) ||
5849 (ANYOF_CLASS_TEST(n, ANYOF_NXDIGIT) && !isXDIGIT(c)) ||
5850 (ANYOF_CLASS_TEST(n, ANYOF_PSXSPC) && isPSXSPC(c)) ||
5851 (ANYOF_CLASS_TEST(n, ANYOF_NPSXSPC) && !isPSXSPC(c)) ||
5852 (ANYOF_CLASS_TEST(n, ANYOF_BLANK) && isBLANK(c)) ||
5853 (ANYOF_CLASS_TEST(n, ANYOF_NBLANK) && !isBLANK(c))
5854 ) /* How's that for a conditional? */
5861 return (flags & ANYOF_INVERT) ? !match : match;
5865 S_reghop3(U8 *s, I32 off, const U8* lim)
5869 PERL_ARGS_ASSERT_REGHOP3;
5872 while (off-- && s < lim) {
5873 /* XXX could check well-formedness here */
5878 while (off++ && s > lim) {
5880 if (UTF8_IS_CONTINUED(*s)) {
5881 while (s > lim && UTF8_IS_CONTINUATION(*s))
5884 /* XXX could check well-formedness here */
5891 /* there are a bunch of places where we use two reghop3's that should
5892 be replaced with this routine. but since thats not done yet
5893 we ifdef it out - dmq
5896 S_reghop4(U8 *s, I32 off, const U8* llim, const U8* rlim)
5900 PERL_ARGS_ASSERT_REGHOP4;
5903 while (off-- && s < rlim) {
5904 /* XXX could check well-formedness here */
5909 while (off++ && s > llim) {
5911 if (UTF8_IS_CONTINUED(*s)) {
5912 while (s > llim && UTF8_IS_CONTINUATION(*s))
5915 /* XXX could check well-formedness here */
5923 S_reghopmaybe3(U8* s, I32 off, const U8* lim)
5927 PERL_ARGS_ASSERT_REGHOPMAYBE3;
5930 while (off-- && s < lim) {
5931 /* XXX could check well-formedness here */
5938 while (off++ && s > lim) {
5940 if (UTF8_IS_CONTINUED(*s)) {
5941 while (s > lim && UTF8_IS_CONTINUATION(*s))
5944 /* XXX could check well-formedness here */
5953 restore_pos(pTHX_ void *arg)
5956 regexp * const rex = (regexp *)arg;
5957 if (PL_reg_eval_set) {
5958 if (PL_reg_oldsaved) {
5959 rex->subbeg = PL_reg_oldsaved;
5960 rex->sublen = PL_reg_oldsavedlen;
5961 #ifdef PERL_OLD_COPY_ON_WRITE
5962 rex->saved_copy = PL_nrs;
5964 RXp_MATCH_COPIED_on(rex);
5966 PL_reg_magic->mg_len = PL_reg_oldpos;
5967 PL_reg_eval_set = 0;
5968 PL_curpm = PL_reg_oldcurpm;
5973 S_to_utf8_substr(pTHX_ register regexp *prog)
5977 PERL_ARGS_ASSERT_TO_UTF8_SUBSTR;
5980 if (prog->substrs->data[i].substr
5981 && !prog->substrs->data[i].utf8_substr) {
5982 SV* const sv = newSVsv(prog->substrs->data[i].substr);
5983 prog->substrs->data[i].utf8_substr = sv;
5984 sv_utf8_upgrade(sv);
5985 if (SvVALID(prog->substrs->data[i].substr)) {
5986 const U8 flags = BmFLAGS(prog->substrs->data[i].substr);
5987 if (flags & FBMcf_TAIL) {
5988 /* Trim the trailing \n that fbm_compile added last
5990 SvCUR_set(sv, SvCUR(sv) - 1);
5991 /* Whilst this makes the SV technically "invalid" (as its
5992 buffer is no longer followed by "\0") when fbm_compile()
5993 adds the "\n" back, a "\0" is restored. */
5995 fbm_compile(sv, flags);
5997 if (prog->substrs->data[i].substr == prog->check_substr)
5998 prog->check_utf8 = sv;
6004 S_to_byte_substr(pTHX_ register regexp *prog)
6009 PERL_ARGS_ASSERT_TO_BYTE_SUBSTR;
6012 if (prog->substrs->data[i].utf8_substr
6013 && !prog->substrs->data[i].substr) {
6014 SV* sv = newSVsv(prog->substrs->data[i].utf8_substr);
6015 if (sv_utf8_downgrade(sv, TRUE)) {
6016 if (SvVALID(prog->substrs->data[i].utf8_substr)) {
6018 = BmFLAGS(prog->substrs->data[i].utf8_substr);
6019 if (flags & FBMcf_TAIL) {
6020 /* Trim the trailing \n that fbm_compile added last
6022 SvCUR_set(sv, SvCUR(sv) - 1);
6024 fbm_compile(sv, flags);
6030 prog->substrs->data[i].substr = sv;
6031 if (prog->substrs->data[i].utf8_substr == prog->check_utf8)
6032 prog->check_substr = sv;
6039 * c-indentation-style: bsd
6041 * indent-tabs-mode: t
6044 * ex: set ts=8 sts=4 sw=4 noet: