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
75 #undef PERL_IN_XSUB_RE
76 #define PERL_IN_XSUB_RE 1
79 #undef PERL_IN_XSUB_RE
81 #ifdef PERL_IN_XSUB_RE
87 #define RF_tainted 1 /* tainted information used? e.g. locale */
88 #define RF_warned 2 /* warned about big count? */
90 #define RF_utf8 8 /* Pattern contains multibyte chars? */
92 #define UTF_PATTERN ((PL_reg_flags & RF_utf8) != 0)
94 #define RS_init 1 /* eval environment created */
95 #define RS_set 2 /* replsv value is set */
101 /* Valid for non-utf8 strings, non-ANYOFV nodes only: avoids the reginclass
102 * call if there are no complications: i.e., if everything matchable is
103 * straight forward in the bitmap */
104 #define REGINCLASS(prog,p,c) (ANYOF_FLAGS(p) ? reginclass(prog,p,c,0,0) \
105 : ANYOF_BITMAP_TEST(p,*(c)))
111 #define CHR_SVLEN(sv) (utf8_target ? sv_len_utf8(sv) : SvCUR(sv))
112 #define CHR_DIST(a,b) (PL_reg_match_utf8 ? utf8_distance(a,b) : a - b)
114 #define HOPc(pos,off) \
115 (char *)(PL_reg_match_utf8 \
116 ? reghop3((U8*)pos, off, (U8*)(off >= 0 ? PL_regeol : PL_bostr)) \
118 #define HOPBACKc(pos, off) \
119 (char*)(PL_reg_match_utf8\
120 ? reghopmaybe3((U8*)pos, -off, (U8*)PL_bostr) \
121 : (pos - off >= PL_bostr) \
125 #define HOP3(pos,off,lim) (PL_reg_match_utf8 ? reghop3((U8*)(pos), off, (U8*)(lim)) : (U8*)(pos + off))
126 #define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim))
128 /* these are unrolled below in the CCC_TRY_XXX defined */
129 #define LOAD_UTF8_CHARCLASS(class,str) STMT_START { \
130 if (!CAT2(PL_utf8_,class)) { bool ok; ENTER; save_re_context(); ok=CAT2(is_utf8_,class)((const U8*)str); assert(ok); LEAVE; } } STMT_END
132 /* Doesn't do an assert to verify that is correct */
133 #define LOAD_UTF8_CHARCLASS_NO_CHECK(class) STMT_START { \
134 if (!CAT2(PL_utf8_,class)) { bool throw_away; ENTER; save_re_context(); throw_away = CAT2(is_utf8_,class)((const U8*)" "); LEAVE; } } STMT_END
136 #define LOAD_UTF8_CHARCLASS_ALNUM() LOAD_UTF8_CHARCLASS(alnum,"a")
137 #define LOAD_UTF8_CHARCLASS_DIGIT() LOAD_UTF8_CHARCLASS(digit,"0")
138 #define LOAD_UTF8_CHARCLASS_SPACE() LOAD_UTF8_CHARCLASS(space," ")
140 #define LOAD_UTF8_CHARCLASS_GCB() /* Grapheme cluster boundaries */ \
141 LOAD_UTF8_CHARCLASS(X_begin, " "); \
142 LOAD_UTF8_CHARCLASS(X_non_hangul, "A"); \
143 /* These are utf8 constants, and not utf-ebcdic constants, so the \
144 * assert should likely and hopefully fail on an EBCDIC machine */ \
145 LOAD_UTF8_CHARCLASS(X_extend, "\xcc\x80"); /* U+0300 */ \
147 /* No asserts are done for these, in case called on an early \
148 * Unicode version in which they map to nothing */ \
149 LOAD_UTF8_CHARCLASS_NO_CHECK(X_prepend);/* U+0E40 "\xe0\xb9\x80" */ \
150 LOAD_UTF8_CHARCLASS_NO_CHECK(X_L); /* U+1100 "\xe1\x84\x80" */ \
151 LOAD_UTF8_CHARCLASS_NO_CHECK(X_LV); /* U+AC00 "\xea\xb0\x80" */ \
152 LOAD_UTF8_CHARCLASS_NO_CHECK(X_LVT); /* U+AC01 "\xea\xb0\x81" */ \
153 LOAD_UTF8_CHARCLASS_NO_CHECK(X_LV_LVT_V);/* U+AC01 "\xea\xb0\x81" */\
154 LOAD_UTF8_CHARCLASS_NO_CHECK(X_T); /* U+11A8 "\xe1\x86\xa8" */ \
155 LOAD_UTF8_CHARCLASS_NO_CHECK(X_V) /* U+1160 "\xe1\x85\xa0" */
157 #define PLACEHOLDER /* Something for the preprocessor to grab onto */
159 /* The actual code for CCC_TRY, which uses several variables from the routine
160 * it's callable from. It is designed to be the bulk of a case statement.
161 * FUNC is the macro or function to call on non-utf8 targets that indicate if
162 * nextchr matches the class.
163 * UTF8_TEST is the whole test string to use for utf8 targets
164 * LOAD is what to use to test, and if not present to load in the swash for the
166 * POS_OR_NEG is either empty or ! to complement the results of FUNC or
168 * The logic is: Fail if we're at the end-of-string; otherwise if the target is
169 * utf8 and a variant, load the swash if necessary and test using the utf8
170 * test. Advance to the next character if test is ok, otherwise fail; If not
171 * utf8 or an invariant under utf8, use the non-utf8 test, and fail if it
172 * fails, or advance to the next character */
174 #define _CCC_TRY_CODE(POS_OR_NEG, FUNC, UTF8_TEST, CLASS, STR) \
175 if (locinput >= PL_regeol) { \
178 if (utf8_target && UTF8_IS_CONTINUED(nextchr)) { \
179 LOAD_UTF8_CHARCLASS(CLASS, STR); \
180 if (POS_OR_NEG (UTF8_TEST)) { \
183 locinput += PL_utf8skip[nextchr]; \
184 nextchr = UCHARAT(locinput); \
187 if (POS_OR_NEG (FUNC(nextchr))) { \
190 nextchr = UCHARAT(++locinput); \
193 /* Handle the non-locale cases for a character class and its complement. It
194 * calls _CCC_TRY_CODE with a ! to complement the test for the character class.
195 * This is because that code fails when the test succeeds, so we want to have
196 * the test fail so that the code succeeds. The swash is stored in a
197 * predictable PL_ place */
198 #define _CCC_TRY_NONLOCALE(NAME, NNAME, FUNC, \
201 _CCC_TRY_CODE( !, FUNC, \
202 cBOOL(swash_fetch(CAT2(PL_utf8_,CLASS), \
203 (U8*)locinput, TRUE)), \
206 _CCC_TRY_CODE( PLACEHOLDER , FUNC, \
207 cBOOL(swash_fetch(CAT2(PL_utf8_,CLASS), \
208 (U8*)locinput, TRUE)), \
211 /* Generate the case statements for both locale and non-locale character
212 * classes in regmatch for classes that don't have special unicode semantics.
213 * Locales don't use an immediate swash, but an intermediary special locale
214 * function that is called on the pointer to the current place in the input
215 * string. That function will resolve to needing the same swash. One might
216 * think that because we don't know what the locale will match, we shouldn't
217 * check with the swash loading function that it loaded properly; ie, that we
218 * should use LOAD_UTF8_CHARCLASS_NO_CHECK for those, but what is passed to the
219 * regular LOAD_UTF8_CHARCLASS is in non-locale terms, and so locale is
221 #define CCC_TRY(NAME, NNAME, FUNC, \
222 NAMEL, NNAMEL, LCFUNC, LCFUNC_utf8, \
223 NAMEA, NNAMEA, FUNCA, \
226 PL_reg_flags |= RF_tainted; \
227 _CCC_TRY_CODE( !, LCFUNC, LCFUNC_utf8((U8*)locinput), CLASS, STR) \
229 PL_reg_flags |= RF_tainted; \
230 _CCC_TRY_CODE( PLACEHOLDER, LCFUNC, LCFUNC_utf8((U8*)locinput), \
233 if (locinput >= PL_regeol || ! FUNCA(nextchr)) { \
236 /* Matched a utf8-invariant, so don't have to worry about utf8 */ \
237 nextchr = UCHARAT(++locinput); \
240 if (locinput >= PL_regeol || FUNCA(nextchr)) { \
244 locinput += PL_utf8skip[nextchr]; \
245 nextchr = UCHARAT(locinput); \
248 nextchr = UCHARAT(++locinput); \
251 /* Generate the non-locale cases */ \
252 _CCC_TRY_NONLOCALE(NAME, NNAME, FUNC, CLASS, STR)
254 /* This is like CCC_TRY, but has an extra set of parameters for generating case
255 * statements to handle separate Unicode semantics nodes */
256 #define CCC_TRY_U(NAME, NNAME, FUNC, \
257 NAMEL, NNAMEL, LCFUNC, LCFUNC_utf8, \
258 NAMEU, NNAMEU, FUNCU, \
259 NAMEA, NNAMEA, FUNCA, \
261 CCC_TRY(NAME, NNAME, FUNC, \
262 NAMEL, NNAMEL, LCFUNC, LCFUNC_utf8, \
263 NAMEA, NNAMEA, FUNCA, \
265 _CCC_TRY_NONLOCALE(NAMEU, NNAMEU, FUNCU, CLASS, STR)
267 /* TODO: Combine JUMPABLE and HAS_TEXT to cache OP(rn) */
269 /* for use after a quantifier and before an EXACT-like node -- japhy */
270 /* it would be nice to rework regcomp.sym to generate this stuff. sigh
272 * NOTE that *nothing* that affects backtracking should be in here, specifically
273 * VERBS must NOT be included. JUMPABLE is used to determine if we can ignore a
274 * node that is in between two EXACT like nodes when ascertaining what the required
275 * "follow" character is. This should probably be moved to regex compile time
276 * although it may be done at run time beause of the REF possibility - more
277 * investigation required. -- demerphq
279 #define JUMPABLE(rn) ( \
281 (OP(rn) == CLOSE && (!cur_eval || cur_eval->u.eval.close_paren != ARG(rn))) || \
283 OP(rn) == SUSPEND || OP(rn) == IFMATCH || \
284 OP(rn) == PLUS || OP(rn) == MINMOD || \
286 (PL_regkind[OP(rn)] == CURLY && ARG1(rn) > 0) \
288 #define IS_EXACT(rn) (PL_regkind[OP(rn)] == EXACT)
290 #define HAS_TEXT(rn) ( IS_EXACT(rn) || PL_regkind[OP(rn)] == REF )
293 /* Currently these are only used when PL_regkind[OP(rn)] == EXACT so
294 we don't need this definition. */
295 #define IS_TEXT(rn) ( OP(rn)==EXACT || OP(rn)==REF || OP(rn)==NREF )
296 #define IS_TEXTF(rn) ( (OP(rn)==EXACTFU || OP(rn)==EXACTFA || OP(rn)==EXACTF) || OP(rn)==REFF || OP(rn)==NREFF )
297 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL || OP(rn)==REFFL || OP(rn)==NREFFL )
300 /* ... so we use this as its faster. */
301 #define IS_TEXT(rn) ( OP(rn)==EXACT )
302 #define IS_TEXTFU(rn) ( OP(rn)==EXACTFU || OP(rn) == EXACTFA)
303 #define IS_TEXTF(rn) ( OP(rn)==EXACTF )
304 #define IS_TEXTFL(rn) ( OP(rn)==EXACTFL )
309 Search for mandatory following text node; for lookahead, the text must
310 follow but for lookbehind (rn->flags != 0) we skip to the next step.
312 #define FIND_NEXT_IMPT(rn) STMT_START { \
313 while (JUMPABLE(rn)) { \
314 const OPCODE type = OP(rn); \
315 if (type == SUSPEND || PL_regkind[type] == CURLY) \
316 rn = NEXTOPER(NEXTOPER(rn)); \
317 else if (type == PLUS) \
319 else if (type == IFMATCH) \
320 rn = (rn->flags == 0) ? NEXTOPER(NEXTOPER(rn)) : rn + ARG(rn); \
321 else rn += NEXT_OFF(rn); \
326 static void restore_pos(pTHX_ void *arg);
328 #define REGCP_PAREN_ELEMS 4
329 #define REGCP_OTHER_ELEMS 5
330 #define REGCP_FRAME_ELEMS 1
331 /* REGCP_FRAME_ELEMS are not part of the REGCP_OTHER_ELEMS and
332 * are needed for the regexp context stack bookkeeping. */
335 S_regcppush(pTHX_ I32 parenfloor)
338 const int retval = PL_savestack_ix;
339 const int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS;
340 const UV total_elems = paren_elems_to_push + REGCP_OTHER_ELEMS;
341 const UV elems_shifted = total_elems << SAVE_TIGHT_SHIFT;
343 GET_RE_DEBUG_FLAGS_DECL;
345 if (paren_elems_to_push < 0)
346 Perl_croak(aTHX_ "panic: paren_elems_to_push < 0");
348 if ((elems_shifted >> SAVE_TIGHT_SHIFT) != total_elems)
349 Perl_croak(aTHX_ "panic: paren_elems_to_push offset %"UVuf
350 " out of range (%lu-%ld)",
351 total_elems, (unsigned long)PL_regsize, (long)parenfloor);
353 SSGROW(total_elems + REGCP_FRAME_ELEMS);
355 for (p = PL_regsize; p > parenfloor; p--) {
356 /* REGCP_PARENS_ELEMS are pushed per pairs of parentheses. */
357 SSPUSHINT(PL_regoffs[p].end);
358 SSPUSHINT(PL_regoffs[p].start);
359 SSPUSHPTR(PL_reg_start_tmp[p]);
361 DEBUG_BUFFERS_r(PerlIO_printf(Perl_debug_log,
362 " saving \\%"UVuf" %"IVdf"(%"IVdf")..%"IVdf"\n",
363 (UV)p, (IV)PL_regoffs[p].start,
364 (IV)(PL_reg_start_tmp[p] - PL_bostr),
365 (IV)PL_regoffs[p].end
368 /* REGCP_OTHER_ELEMS are pushed in any case, parentheses or no. */
369 SSPUSHPTR(PL_regoffs);
370 SSPUSHINT(PL_regsize);
371 SSPUSHINT(*PL_reglastparen);
372 SSPUSHINT(*PL_reglastcloseparen);
373 SSPUSHPTR(PL_reginput);
374 SSPUSHUV(SAVEt_REGCONTEXT | elems_shifted); /* Magic cookie. */
379 /* These are needed since we do not localize EVAL nodes: */
380 #define REGCP_SET(cp) \
382 PerlIO_printf(Perl_debug_log, \
383 " Setting an EVAL scope, savestack=%"IVdf"\n", \
384 (IV)PL_savestack_ix)); \
387 #define REGCP_UNWIND(cp) \
389 if (cp != PL_savestack_ix) \
390 PerlIO_printf(Perl_debug_log, \
391 " Clearing an EVAL scope, savestack=%"IVdf"..%"IVdf"\n", \
392 (IV)(cp), (IV)PL_savestack_ix)); \
396 S_regcppop(pTHX_ const regexp *rex)
401 GET_RE_DEBUG_FLAGS_DECL;
403 PERL_ARGS_ASSERT_REGCPPOP;
405 /* Pop REGCP_OTHER_ELEMS before the parentheses loop starts. */
407 assert((i & SAVE_MASK) == SAVEt_REGCONTEXT); /* Check that the magic cookie is there. */
408 i >>= SAVE_TIGHT_SHIFT; /* Parentheses elements to pop. */
409 input = (char *) SSPOPPTR;
410 *PL_reglastcloseparen = SSPOPINT;
411 *PL_reglastparen = SSPOPINT;
412 PL_regsize = SSPOPINT;
413 PL_regoffs=(regexp_paren_pair *) SSPOPPTR;
415 i -= REGCP_OTHER_ELEMS;
416 /* Now restore the parentheses context. */
417 for ( ; i > 0; i -= REGCP_PAREN_ELEMS) {
419 U32 paren = (U32)SSPOPINT;
420 PL_reg_start_tmp[paren] = (char *) SSPOPPTR;
421 PL_regoffs[paren].start = SSPOPINT;
423 if (paren <= *PL_reglastparen)
424 PL_regoffs[paren].end = tmps;
426 PerlIO_printf(Perl_debug_log,
427 " restoring \\%"UVuf" to %"IVdf"(%"IVdf")..%"IVdf"%s\n",
428 (UV)paren, (IV)PL_regoffs[paren].start,
429 (IV)(PL_reg_start_tmp[paren] - PL_bostr),
430 (IV)PL_regoffs[paren].end,
431 (paren > *PL_reglastparen ? "(no)" : ""));
435 if (*PL_reglastparen + 1 <= rex->nparens) {
436 PerlIO_printf(Perl_debug_log,
437 " restoring \\%"IVdf"..\\%"IVdf" to undef\n",
438 (IV)(*PL_reglastparen + 1), (IV)rex->nparens);
442 /* It would seem that the similar code in regtry()
443 * already takes care of this, and in fact it is in
444 * a better location to since this code can #if 0-ed out
445 * but the code in regtry() is needed or otherwise tests
446 * requiring null fields (pat.t#187 and split.t#{13,14}
447 * (as of patchlevel 7877) will fail. Then again,
448 * this code seems to be necessary or otherwise
449 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
450 * --jhi updated by dapm */
451 for (i = *PL_reglastparen + 1; i <= rex->nparens; i++) {
453 PL_regoffs[i].start = -1;
454 PL_regoffs[i].end = -1;
460 #define regcpblow(cp) LEAVE_SCOPE(cp) /* Ignores regcppush()ed data. */
463 * pregexec and friends
466 #ifndef PERL_IN_XSUB_RE
468 - pregexec - match a regexp against a string
471 Perl_pregexec(pTHX_ REGEXP * const prog, char* stringarg, register char *strend,
472 char *strbeg, I32 minend, SV *screamer, U32 nosave)
473 /* strend: pointer to null at end of string */
474 /* strbeg: real beginning of string */
475 /* minend: end of match must be >=minend after stringarg. */
476 /* nosave: For optimizations. */
478 PERL_ARGS_ASSERT_PREGEXEC;
481 regexec_flags(prog, stringarg, strend, strbeg, minend, screamer, NULL,
482 nosave ? 0 : REXEC_COPY_STR);
487 * Need to implement the following flags for reg_anch:
489 * USE_INTUIT_NOML - Useful to call re_intuit_start() first
491 * INTUIT_AUTORITATIVE_NOML - Can trust a positive answer
492 * INTUIT_AUTORITATIVE_ML
493 * INTUIT_ONCE_NOML - Intuit can match in one location only.
496 * Another flag for this function: SECOND_TIME (so that float substrs
497 * with giant delta may be not rechecked).
500 /* Assumptions: if ANCH_GPOS, then strpos is anchored. XXXX Check GPOS logic */
502 /* If SCREAM, then SvPVX_const(sv) should be compatible with strpos and strend.
503 Otherwise, only SvCUR(sv) is used to get strbeg. */
505 /* XXXX We assume that strpos is strbeg unless sv. */
507 /* XXXX Some places assume that there is a fixed substring.
508 An update may be needed if optimizer marks as "INTUITable"
509 RExen without fixed substrings. Similarly, it is assumed that
510 lengths of all the strings are no more than minlen, thus they
511 cannot come from lookahead.
512 (Or minlen should take into account lookahead.)
513 NOTE: Some of this comment is not correct. minlen does now take account
514 of lookahead/behind. Further research is required. -- demerphq
518 /* A failure to find a constant substring means that there is no need to make
519 an expensive call to REx engine, thus we celebrate a failure. Similarly,
520 finding a substring too deep into the string means that less calls to
521 regtry() should be needed.
523 REx compiler's optimizer found 4 possible hints:
524 a) Anchored substring;
526 c) Whether we are anchored (beginning-of-line or \G);
527 d) First node (of those at offset 0) which may distinguish positions;
528 We use a)b)d) and multiline-part of c), and try to find a position in the
529 string which does not contradict any of them.
532 /* Most of decisions we do here should have been done at compile time.
533 The nodes of the REx which we used for the search should have been
534 deleted from the finite automaton. */
537 Perl_re_intuit_start(pTHX_ REGEXP * const rx, SV *sv, char *strpos,
538 char *strend, const U32 flags, re_scream_pos_data *data)
541 struct regexp *const prog = (struct regexp *)SvANY(rx);
542 register I32 start_shift = 0;
543 /* Should be nonnegative! */
544 register I32 end_shift = 0;
549 const bool utf8_target = (sv && SvUTF8(sv)) ? 1 : 0; /* if no sv we have to assume bytes */
551 register char *other_last = NULL; /* other substr checked before this */
552 char *check_at = NULL; /* check substr found at this pos */
553 const I32 multiline = prog->extflags & RXf_PMf_MULTILINE;
554 RXi_GET_DECL(prog,progi);
556 const char * const i_strpos = strpos;
558 GET_RE_DEBUG_FLAGS_DECL;
560 PERL_ARGS_ASSERT_RE_INTUIT_START;
562 RX_MATCH_UTF8_set(rx,utf8_target);
565 PL_reg_flags |= RF_utf8;
568 debug_start_match(rx, utf8_target, strpos, strend,
569 sv ? "Guessing start of match in sv for"
570 : "Guessing start of match in string for");
573 /* CHR_DIST() would be more correct here but it makes things slow. */
574 if (prog->minlen > strend - strpos) {
575 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
576 "String too short... [re_intuit_start]\n"));
580 strbeg = (sv && SvPOK(sv)) ? strend - SvCUR(sv) : strpos;
583 if (!prog->check_utf8 && prog->check_substr)
584 to_utf8_substr(prog);
585 check = prog->check_utf8;
587 if (!prog->check_substr && prog->check_utf8)
588 to_byte_substr(prog);
589 check = prog->check_substr;
591 if (check == &PL_sv_undef) {
592 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
593 "Non-utf8 string cannot match utf8 check string\n"));
596 if (prog->extflags & RXf_ANCH) { /* Match at beg-of-str or after \n */
597 ml_anch = !( (prog->extflags & RXf_ANCH_SINGLE)
598 || ( (prog->extflags & RXf_ANCH_BOL)
599 && !multiline ) ); /* Check after \n? */
602 if ( !(prog->extflags & RXf_ANCH_GPOS) /* Checked by the caller */
603 && !(prog->intflags & PREGf_IMPLICIT) /* not a real BOL */
604 /* SvCUR is not set on references: SvRV and SvPVX_const overlap */
606 && (strpos != strbeg)) {
607 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not at start...\n"));
610 if (prog->check_offset_min == prog->check_offset_max &&
611 !(prog->extflags & RXf_CANY_SEEN)) {
612 /* Substring at constant offset from beg-of-str... */
615 s = HOP3c(strpos, prog->check_offset_min, strend);
618 slen = SvCUR(check); /* >= 1 */
620 if ( strend - s > slen || strend - s < slen - 1
621 || (strend - s == slen && strend[-1] != '\n')) {
622 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String too long...\n"));
625 /* Now should match s[0..slen-2] */
627 if (slen && (*SvPVX_const(check) != *s
629 && memNE(SvPVX_const(check), s, slen)))) {
631 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String not equal...\n"));
635 else if (*SvPVX_const(check) != *s
636 || ((slen = SvCUR(check)) > 1
637 && memNE(SvPVX_const(check), s, slen)))
640 goto success_at_start;
643 /* Match is anchored, but substr is not anchored wrt beg-of-str. */
645 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
646 end_shift = prog->check_end_shift;
649 const I32 end = prog->check_offset_max + CHR_SVLEN(check)
650 - (SvTAIL(check) != 0);
651 const I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end;
653 if (end_shift < eshift)
657 else { /* Can match at random position */
660 start_shift = prog->check_offset_min; /* okay to underestimate on CC */
661 end_shift = prog->check_end_shift;
663 /* end shift should be non negative here */
666 #ifdef QDEBUGGING /* 7/99: reports of failure (with the older version) */
668 Perl_croak(aTHX_ "panic: end_shift: %"IVdf" pattern:\n%s\n ",
669 (IV)end_shift, RX_PRECOMP(prog));
673 /* Find a possible match in the region s..strend by looking for
674 the "check" substring in the region corrected by start/end_shift. */
677 I32 srch_start_shift = start_shift;
678 I32 srch_end_shift = end_shift;
679 if (srch_start_shift < 0 && strbeg - s > srch_start_shift) {
680 srch_end_shift -= ((strbeg - s) - srch_start_shift);
681 srch_start_shift = strbeg - s;
683 DEBUG_OPTIMISE_MORE_r({
684 PerlIO_printf(Perl_debug_log, "Check offset min: %"IVdf" Start shift: %"IVdf" End shift %"IVdf" Real End Shift: %"IVdf"\n",
685 (IV)prog->check_offset_min,
686 (IV)srch_start_shift,
688 (IV)prog->check_end_shift);
691 if (flags & REXEC_SCREAM) {
692 I32 p = -1; /* Internal iterator of scream. */
693 I32 * const pp = data ? data->scream_pos : &p;
695 if (PL_screamfirst[BmRARE(check)] >= 0
696 || ( BmRARE(check) == '\n'
697 && (BmPREVIOUS(check) == SvCUR(check) - 1)
699 s = screaminstr(sv, check,
700 srch_start_shift + (s - strbeg), srch_end_shift, pp, 0);
703 /* we may be pointing at the wrong string */
704 if (s && RXp_MATCH_COPIED(prog))
705 s = strbeg + (s - SvPVX_const(sv));
707 *data->scream_olds = s;
712 if (prog->extflags & RXf_CANY_SEEN) {
713 start_point= (U8*)(s + srch_start_shift);
714 end_point= (U8*)(strend - srch_end_shift);
716 start_point= HOP3(s, srch_start_shift, srch_start_shift < 0 ? strbeg : strend);
717 end_point= HOP3(strend, -srch_end_shift, strbeg);
719 DEBUG_OPTIMISE_MORE_r({
720 PerlIO_printf(Perl_debug_log, "fbm_instr len=%d str=<%.*s>\n",
721 (int)(end_point - start_point),
722 (int)(end_point - start_point) > 20 ? 20 : (int)(end_point - start_point),
726 s = fbm_instr( start_point, end_point,
727 check, multiline ? FBMrf_MULTILINE : 0);
730 /* Update the count-of-usability, remove useless subpatterns,
734 RE_PV_QUOTED_DECL(quoted, utf8_target, PERL_DEBUG_PAD_ZERO(0),
735 SvPVX_const(check), RE_SV_DUMPLEN(check), 30);
736 PerlIO_printf(Perl_debug_log, "%s %s substr %s%s%s",
737 (s ? "Found" : "Did not find"),
738 (check == (utf8_target ? prog->anchored_utf8 : prog->anchored_substr)
739 ? "anchored" : "floating"),
742 (s ? " at offset " : "...\n") );
747 /* Finish the diagnostic message */
748 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%ld...\n", (long)(s - i_strpos)) );
750 /* XXX dmq: first branch is for positive lookbehind...
751 Our check string is offset from the beginning of the pattern.
752 So we need to do any stclass tests offset forward from that
761 /* Got a candidate. Check MBOL anchoring, and the *other* substr.
762 Start with the other substr.
763 XXXX no SCREAM optimization yet - and a very coarse implementation
764 XXXX /ttx+/ results in anchored="ttx", floating="x". floating will
765 *always* match. Probably should be marked during compile...
766 Probably it is right to do no SCREAM here...
769 if (utf8_target ? (prog->float_utf8 && prog->anchored_utf8)
770 : (prog->float_substr && prog->anchored_substr))
772 /* Take into account the "other" substring. */
773 /* XXXX May be hopelessly wrong for UTF... */
776 if (check == (utf8_target ? prog->float_utf8 : prog->float_substr)) {
779 char * const last = HOP3c(s, -start_shift, strbeg);
781 char * const saved_s = s;
784 t = s - prog->check_offset_max;
785 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
787 || ((t = (char*)reghopmaybe3((U8*)s, -(prog->check_offset_max), (U8*)strpos))
792 t = HOP3c(t, prog->anchored_offset, strend);
793 if (t < other_last) /* These positions already checked */
795 last2 = last1 = HOP3c(strend, -prog->minlen, strbeg);
798 /* XXXX It is not documented what units *_offsets are in.
799 We assume bytes, but this is clearly wrong.
800 Meaning this code needs to be carefully reviewed for errors.
804 /* On end-of-str: see comment below. */
805 must = utf8_target ? prog->anchored_utf8 : prog->anchored_substr;
806 if (must == &PL_sv_undef) {
808 DEBUG_r(must = prog->anchored_utf8); /* for debug */
813 HOP3(HOP3(last1, prog->anchored_offset, strend)
814 + SvCUR(must), -(SvTAIL(must)!=0), strbeg),
816 multiline ? FBMrf_MULTILINE : 0
819 RE_PV_QUOTED_DECL(quoted, utf8_target, PERL_DEBUG_PAD_ZERO(0),
820 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
821 PerlIO_printf(Perl_debug_log, "%s anchored substr %s%s",
822 (s ? "Found" : "Contradicts"),
823 quoted, RE_SV_TAIL(must));
828 if (last1 >= last2) {
829 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
830 ", giving up...\n"));
833 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
834 ", trying floating at offset %ld...\n",
835 (long)(HOP3c(saved_s, 1, strend) - i_strpos)));
836 other_last = HOP3c(last1, prog->anchored_offset+1, strend);
837 s = HOP3c(last, 1, strend);
841 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
842 (long)(s - i_strpos)));
843 t = HOP3c(s, -prog->anchored_offset, strbeg);
844 other_last = HOP3c(s, 1, strend);
852 else { /* Take into account the floating substring. */
854 char * const saved_s = s;
857 t = HOP3c(s, -start_shift, strbeg);
859 HOP3c(strend, -prog->minlen + prog->float_min_offset, strbeg);
860 if (CHR_DIST((U8*)last, (U8*)t) > prog->float_max_offset)
861 last = HOP3c(t, prog->float_max_offset, strend);
862 s = HOP3c(t, prog->float_min_offset, strend);
865 /* XXXX It is not documented what units *_offsets are in. Assume bytes. */
866 must = utf8_target ? prog->float_utf8 : prog->float_substr;
867 /* fbm_instr() takes into account exact value of end-of-str
868 if the check is SvTAIL(ed). Since false positives are OK,
869 and end-of-str is not later than strend we are OK. */
870 if (must == &PL_sv_undef) {
872 DEBUG_r(must = prog->float_utf8); /* for debug message */
875 s = fbm_instr((unsigned char*)s,
876 (unsigned char*)last + SvCUR(must)
878 must, multiline ? FBMrf_MULTILINE : 0);
880 RE_PV_QUOTED_DECL(quoted, utf8_target, PERL_DEBUG_PAD_ZERO(0),
881 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
882 PerlIO_printf(Perl_debug_log, "%s floating substr %s%s",
883 (s ? "Found" : "Contradicts"),
884 quoted, RE_SV_TAIL(must));
888 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
889 ", giving up...\n"));
892 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
893 ", trying anchored starting at offset %ld...\n",
894 (long)(saved_s + 1 - i_strpos)));
896 s = HOP3c(t, 1, strend);
900 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, " at offset %ld...\n",
901 (long)(s - i_strpos)));
902 other_last = s; /* Fix this later. --Hugo */
912 t= (char*)HOP3( s, -prog->check_offset_max, (prog->check_offset_max<0) ? strend : strpos);
914 DEBUG_OPTIMISE_MORE_r(
915 PerlIO_printf(Perl_debug_log,
916 "Check offset min:%"IVdf" max:%"IVdf" S:%"IVdf" t:%"IVdf" D:%"IVdf" end:%"IVdf"\n",
917 (IV)prog->check_offset_min,
918 (IV)prog->check_offset_max,
926 if (s - strpos > prog->check_offset_max /* signed-corrected t > strpos */
928 || ((t = (char*)reghopmaybe3((U8*)s, -prog->check_offset_max, (U8*) ((prog->check_offset_max<0) ? strend : strpos)))
931 /* Fixed substring is found far enough so that the match
932 cannot start at strpos. */
934 if (ml_anch && t[-1] != '\n') {
935 /* Eventually fbm_*() should handle this, but often
936 anchored_offset is not 0, so this check will not be wasted. */
937 /* XXXX In the code below we prefer to look for "^" even in
938 presence of anchored substrings. And we search even
939 beyond the found float position. These pessimizations
940 are historical artefacts only. */
942 while (t < strend - prog->minlen) {
944 if (t < check_at - prog->check_offset_min) {
945 if (utf8_target ? prog->anchored_utf8 : prog->anchored_substr) {
946 /* Since we moved from the found position,
947 we definitely contradict the found anchored
948 substr. Due to the above check we do not
949 contradict "check" substr.
950 Thus we can arrive here only if check substr
951 is float. Redo checking for "other"=="fixed".
954 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld, rescanning for anchored from offset %ld...\n",
955 PL_colors[0], PL_colors[1], (long)(strpos - i_strpos), (long)(strpos - i_strpos + prog->anchored_offset)));
956 goto do_other_anchored;
958 /* We don't contradict the found floating substring. */
959 /* XXXX Why not check for STCLASS? */
961 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld...\n",
962 PL_colors[0], PL_colors[1], (long)(s - i_strpos)));
965 /* Position contradicts check-string */
966 /* XXXX probably better to look for check-string
967 than for "\n", so one should lower the limit for t? */
968 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m, restarting lookup for check-string at offset %ld...\n",
969 PL_colors[0], PL_colors[1], (long)(t + 1 - i_strpos)));
970 other_last = strpos = s = t + 1;
975 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Did not find /%s^%s/m...\n",
976 PL_colors[0], PL_colors[1]));
980 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Starting position does not contradict /%s^%s/m...\n",
981 PL_colors[0], PL_colors[1]));
985 ++BmUSEFUL(utf8_target ? prog->check_utf8 : prog->check_substr); /* hooray/5 */
988 /* The found string does not prohibit matching at strpos,
989 - no optimization of calling REx engine can be performed,
990 unless it was an MBOL and we are not after MBOL,
991 or a future STCLASS check will fail this. */
993 /* Even in this situation we may use MBOL flag if strpos is offset
994 wrt the start of the string. */
995 if (ml_anch && sv && !SvROK(sv) /* See prev comment on SvROK */
996 && (strpos != strbeg) && strpos[-1] != '\n'
997 /* May be due to an implicit anchor of m{.*foo} */
998 && !(prog->intflags & PREGf_IMPLICIT))
1003 DEBUG_EXECUTE_r( if (ml_anch)
1004 PerlIO_printf(Perl_debug_log, "Position at offset %ld does not contradict /%s^%s/m...\n",
1005 (long)(strpos - i_strpos), PL_colors[0], PL_colors[1]);
1008 if (!(prog->intflags & PREGf_NAUGHTY) /* XXXX If strpos moved? */
1010 prog->check_utf8 /* Could be deleted already */
1011 && --BmUSEFUL(prog->check_utf8) < 0
1012 && (prog->check_utf8 == prog->float_utf8)
1014 prog->check_substr /* Could be deleted already */
1015 && --BmUSEFUL(prog->check_substr) < 0
1016 && (prog->check_substr == prog->float_substr)
1019 /* If flags & SOMETHING - do not do it many times on the same match */
1020 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "... Disabling check substring...\n"));
1021 /* XXX Does the destruction order has to change with utf8_target? */
1022 SvREFCNT_dec(utf8_target ? prog->check_utf8 : prog->check_substr);
1023 SvREFCNT_dec(utf8_target ? prog->check_substr : prog->check_utf8);
1024 prog->check_substr = prog->check_utf8 = NULL; /* disable */
1025 prog->float_substr = prog->float_utf8 = NULL; /* clear */
1026 check = NULL; /* abort */
1028 /* XXXX If the check string was an implicit check MBOL, then we need to unset the relevant flag
1029 see http://bugs.activestate.com/show_bug.cgi?id=87173 */
1030 if (prog->intflags & PREGf_IMPLICIT)
1031 prog->extflags &= ~RXf_ANCH_MBOL;
1032 /* XXXX This is a remnant of the old implementation. It
1033 looks wasteful, since now INTUIT can use many
1034 other heuristics. */
1035 prog->extflags &= ~RXf_USE_INTUIT;
1036 /* XXXX What other flags might need to be cleared in this branch? */
1042 /* Last resort... */
1043 /* XXXX BmUSEFUL already changed, maybe multiple change is meaningful... */
1044 /* trie stclasses are too expensive to use here, we are better off to
1045 leave it to regmatch itself */
1046 if (progi->regstclass && PL_regkind[OP(progi->regstclass)]!=TRIE) {
1047 /* minlen == 0 is possible if regstclass is \b or \B,
1048 and the fixed substr is ''$.
1049 Since minlen is already taken into account, s+1 is before strend;
1050 accidentally, minlen >= 1 guaranties no false positives at s + 1
1051 even for \b or \B. But (minlen? 1 : 0) below assumes that
1052 regstclass does not come from lookahead... */
1053 /* If regstclass takes bytelength more than 1: If charlength==1, OK.
1054 This leaves EXACTF-ish only, which are dealt with in find_byclass(). */
1055 const U8* const str = (U8*)STRING(progi->regstclass);
1056 const int cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
1057 ? CHR_DIST(str+STR_LEN(progi->regstclass), str)
1060 if (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
1061 endpos= HOP3c(s, (prog->minlen ? cl_l : 0), strend);
1062 else if (prog->float_substr || prog->float_utf8)
1063 endpos= HOP3c(HOP3c(check_at, -start_shift, strbeg), cl_l, strend);
1067 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "start_shift: %"IVdf" check_at: %"IVdf" s: %"IVdf" endpos: %"IVdf"\n",
1068 (IV)start_shift, (IV)(check_at - strbeg), (IV)(s - strbeg), (IV)(endpos - strbeg)));
1071 s = find_byclass(prog, progi->regstclass, s, endpos, NULL);
1074 const char *what = NULL;
1076 if (endpos == strend) {
1077 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1078 "Could not match STCLASS...\n") );
1081 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1082 "This position contradicts STCLASS...\n") );
1083 if ((prog->extflags & RXf_ANCH) && !ml_anch)
1085 /* Contradict one of substrings */
1086 if (prog->anchored_substr || prog->anchored_utf8) {
1087 if ((utf8_target ? prog->anchored_utf8 : prog->anchored_substr) == check) {
1088 DEBUG_EXECUTE_r( what = "anchored" );
1090 s = HOP3c(t, 1, strend);
1091 if (s + start_shift + end_shift > strend) {
1092 /* XXXX Should be taken into account earlier? */
1093 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1094 "Could not match STCLASS...\n") );
1099 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1100 "Looking for %s substr starting at offset %ld...\n",
1101 what, (long)(s + start_shift - i_strpos)) );
1104 /* Have both, check_string is floating */
1105 if (t + start_shift >= check_at) /* Contradicts floating=check */
1106 goto retry_floating_check;
1107 /* Recheck anchored substring, but not floating... */
1111 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1112 "Looking for anchored substr starting at offset %ld...\n",
1113 (long)(other_last - i_strpos)) );
1114 goto do_other_anchored;
1116 /* Another way we could have checked stclass at the
1117 current position only: */
1122 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
1123 "Looking for /%s^%s/m starting at offset %ld...\n",
1124 PL_colors[0], PL_colors[1], (long)(t - i_strpos)) );
1127 if (!(utf8_target ? prog->float_utf8 : prog->float_substr)) /* Could have been deleted */
1129 /* Check is floating substring. */
1130 retry_floating_check:
1131 t = check_at - start_shift;
1132 DEBUG_EXECUTE_r( what = "floating" );
1133 goto hop_and_restart;
1136 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
1137 "By STCLASS: moving %ld --> %ld\n",
1138 (long)(t - i_strpos), (long)(s - i_strpos))
1142 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
1143 "Does not contradict STCLASS...\n");
1148 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s%s:%s match at offset %ld\n",
1149 PL_colors[4], (check ? "Guessed" : "Giving up"),
1150 PL_colors[5], (long)(s - i_strpos)) );
1153 fail_finish: /* Substring not found */
1154 if (prog->check_substr || prog->check_utf8) /* could be removed already */
1155 BmUSEFUL(utf8_target ? prog->check_utf8 : prog->check_substr) += 5; /* hooray */
1157 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch rejected by optimizer%s\n",
1158 PL_colors[4], PL_colors[5]));
1162 #define DECL_TRIE_TYPE(scan) \
1163 const enum { trie_plain, trie_utf8, trie_utf8_fold, trie_latin_utf8_fold } \
1164 trie_type = (scan->flags != EXACT) \
1165 ? (utf8_target ? trie_utf8_fold : (UTF_PATTERN ? trie_latin_utf8_fold : trie_plain)) \
1166 : (utf8_target ? trie_utf8 : trie_plain)
1168 #define REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc, uscan, len, \
1169 uvc, charid, foldlen, foldbuf, uniflags) STMT_START { \
1170 switch (trie_type) { \
1171 case trie_utf8_fold: \
1172 if ( foldlen>0 ) { \
1173 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1178 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1179 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1180 foldlen -= UNISKIP( uvc ); \
1181 uscan = foldbuf + UNISKIP( uvc ); \
1184 case trie_latin_utf8_fold: \
1185 if ( foldlen>0 ) { \
1186 uvc = utf8n_to_uvuni( uscan, UTF8_MAXLEN, &len, uniflags ); \
1192 uvc = to_uni_fold( *(U8*)uc, foldbuf, &foldlen ); \
1193 foldlen -= UNISKIP( uvc ); \
1194 uscan = foldbuf + UNISKIP( uvc ); \
1198 uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); \
1205 charid = trie->charmap[ uvc ]; \
1209 if (widecharmap) { \
1210 SV** const svpp = hv_fetch(widecharmap, \
1211 (char*)&uvc, sizeof(UV), 0); \
1213 charid = (U16)SvIV(*svpp); \
1218 #define REXEC_FBC_EXACTISH_SCAN(CoNd) \
1222 && (ln == 1 || folder(s, pat_string, ln)) \
1223 && (!reginfo || regtry(reginfo, &s)) ) \
1229 #define REXEC_FBC_UTF8_SCAN(CoDe) \
1231 while (s + (uskip = UTF8SKIP(s)) <= strend) { \
1237 #define REXEC_FBC_SCAN(CoDe) \
1239 while (s < strend) { \
1245 #define REXEC_FBC_UTF8_CLASS_SCAN(CoNd) \
1246 REXEC_FBC_UTF8_SCAN( \
1248 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1257 #define REXEC_FBC_CLASS_SCAN(CoNd) \
1260 if (tmp && (!reginfo || regtry(reginfo, &s))) \
1269 #define REXEC_FBC_TRYIT \
1270 if ((!reginfo || regtry(reginfo, &s))) \
1273 #define REXEC_FBC_CSCAN(CoNdUtF8,CoNd) \
1274 if (utf8_target) { \
1275 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1278 REXEC_FBC_CLASS_SCAN(CoNd); \
1281 #define REXEC_FBC_CSCAN_PRELOAD(UtFpReLoAd,CoNdUtF8,CoNd) \
1282 if (utf8_target) { \
1284 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1287 REXEC_FBC_CLASS_SCAN(CoNd); \
1290 #define REXEC_FBC_CSCAN_TAINT(CoNdUtF8,CoNd) \
1291 PL_reg_flags |= RF_tainted; \
1292 if (utf8_target) { \
1293 REXEC_FBC_UTF8_CLASS_SCAN(CoNdUtF8); \
1296 REXEC_FBC_CLASS_SCAN(CoNd); \
1299 #define DUMP_EXEC_POS(li,s,doutf8) \
1300 dump_exec_pos(li,s,(PL_regeol),(PL_bostr),(PL_reg_starttry),doutf8)
1303 #define UTF8_NOLOAD(TEST_NON_UTF8, IF_SUCCESS, IF_FAIL) \
1304 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n'; \
1305 tmp = TEST_NON_UTF8(tmp); \
1306 REXEC_FBC_UTF8_SCAN( \
1307 if (tmp == ! TEST_NON_UTF8((U8) *s)) { \
1316 #define UTF8_LOAD(TeSt1_UtF8, TeSt2_UtF8, IF_SUCCESS, IF_FAIL) \
1317 if (s == PL_bostr) { \
1321 U8 * const r = reghop3((U8*)s, -1, (U8*)PL_bostr); \
1322 tmp = utf8n_to_uvchr(r, UTF8SKIP(r), 0, UTF8_ALLOW_DEFAULT); \
1325 LOAD_UTF8_CHARCLASS_ALNUM(); \
1326 REXEC_FBC_UTF8_SCAN( \
1327 if (tmp == ! (TeSt2_UtF8)) { \
1336 /* The only difference between the BOUND and NBOUND cases is that
1337 * REXEC_FBC_TRYIT is called when matched in BOUND, and when non-matched in
1338 * NBOUND. This is accomplished by passing it in either the if or else clause,
1339 * with the other one being empty */
1340 #define FBC_BOUND(TEST_NON_UTF8, TEST1_UTF8, TEST2_UTF8) \
1341 FBC_BOUND_COMMON(UTF8_LOAD(TEST1_UTF8, TEST2_UTF8, REXEC_FBC_TRYIT, PLACEHOLDER), TEST_NON_UTF8, REXEC_FBC_TRYIT, PLACEHOLDER)
1343 #define FBC_BOUND_NOLOAD(TEST_NON_UTF8, TEST1_UTF8, TEST2_UTF8) \
1344 FBC_BOUND_COMMON(UTF8_NOLOAD(TEST_NON_UTF8, REXEC_FBC_TRYIT, PLACEHOLDER), TEST_NON_UTF8, REXEC_FBC_TRYIT, PLACEHOLDER)
1346 #define FBC_NBOUND(TEST_NON_UTF8, TEST1_UTF8, TEST2_UTF8) \
1347 FBC_BOUND_COMMON(UTF8_LOAD(TEST1_UTF8, TEST2_UTF8, PLACEHOLDER, REXEC_FBC_TRYIT), TEST_NON_UTF8, PLACEHOLDER, REXEC_FBC_TRYIT)
1349 #define FBC_NBOUND_NOLOAD(TEST_NON_UTF8, TEST1_UTF8, TEST2_UTF8) \
1350 FBC_BOUND_COMMON(UTF8_NOLOAD(TEST_NON_UTF8, PLACEHOLDER, REXEC_FBC_TRYIT), TEST_NON_UTF8, PLACEHOLDER, REXEC_FBC_TRYIT)
1353 /* Common to the BOUND and NBOUND cases. Unfortunately the UTF8 tests need to
1354 * be passed in completely with the variable name being tested, which isn't
1355 * such a clean interface, but this is easier to read than it was before. We
1356 * are looking for the boundary (or non-boundary between a word and non-word
1357 * character. The utf8 and non-utf8 cases have the same logic, but the details
1358 * must be different. Find the "wordness" of the character just prior to this
1359 * one, and compare it with the wordness of this one. If they differ, we have
1360 * a boundary. At the beginning of the string, pretend that the previous
1361 * character was a new-line */
1362 #define FBC_BOUND_COMMON(UTF8_CODE, TEST_NON_UTF8, IF_SUCCESS, IF_FAIL) \
1363 if (utf8_target) { \
1366 else { /* Not utf8 */ \
1367 tmp = (s != PL_bostr) ? UCHARAT(s - 1) : '\n'; \
1368 tmp = TEST_NON_UTF8(tmp); \
1370 if (tmp == ! TEST_NON_UTF8((U8) *s)) { \
1379 if ((!prog->minlen && tmp) && (!reginfo || regtry(reginfo, &s))) \
1382 /* We know what class REx starts with. Try to find this position... */
1383 /* if reginfo is NULL, its a dryrun */
1384 /* annoyingly all the vars in this routine have different names from their counterparts
1385 in regmatch. /grrr */
1388 S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s,
1389 const char *strend, regmatch_info *reginfo)
1392 const I32 doevery = (prog->intflags & PREGf_SKIP) == 0;
1393 char *pat_string; /* The pattern's exactish string */
1394 char *pat_end; /* ptr to end char of pat_string */
1395 re_fold_t folder; /* Function for computing non-utf8 folds */
1396 const U8 *fold_array; /* array for folding ords < 256 */
1399 register STRLEN uskip;
1403 register I32 tmp = 1; /* Scratch variable? */
1404 register const bool utf8_target = PL_reg_match_utf8;
1405 UV utf8_fold_flags = 0;
1406 RXi_GET_DECL(prog,progi);
1408 PERL_ARGS_ASSERT_FIND_BYCLASS;
1410 /* We know what class it must start with. */
1414 if (utf8_target || OP(c) == ANYOFV) {
1415 STRLEN inclasslen = strend - s;
1416 REXEC_FBC_UTF8_CLASS_SCAN(
1417 reginclass(prog, c, (U8*)s, &inclasslen, utf8_target));
1420 REXEC_FBC_CLASS_SCAN(REGINCLASS(prog, c, (U8*)s));
1425 if (tmp && (!reginfo || regtry(reginfo, &s)))
1433 if (UTF_PATTERN || utf8_target) {
1434 utf8_fold_flags = FOLDEQ_UTF8_NOMIX_ASCII;
1435 goto do_exactf_utf8;
1437 fold_array = PL_fold_latin1; /* Latin1 folds are not affected by */
1438 folder = foldEQ_latin1; /* /a, except the sharp s one which */
1439 goto do_exactf_non_utf8; /* isn't dealt with by these */
1442 if (UTF_PATTERN || utf8_target) {
1443 utf8_fold_flags = 0;
1444 goto do_exactf_utf8;
1446 fold_array = PL_fold_latin1;
1447 folder = foldEQ_latin1;
1448 /* XXX This uses the full utf8 fold because if the pattern contains
1449 * 'ss' it could match LATIN_SMALL_LETTER SHARP_S in the string.
1450 * There could be a new node type, say EXACTFU_SS, which is
1451 * generated by regcomp only if there is an 'ss', and then every
1452 * other case could goto do_exactf_non_utf8;*/
1453 goto do_exactf_utf8;
1456 if (UTF_PATTERN || utf8_target) {
1457 utf8_fold_flags = 0;
1458 goto do_exactf_utf8;
1460 fold_array = PL_fold;
1462 goto do_exactf_non_utf8;
1465 if (UTF_PATTERN || utf8_target) {
1466 utf8_fold_flags = FOLDEQ_UTF8_LOCALE;
1467 goto do_exactf_utf8;
1469 fold_array = PL_fold_locale;
1470 folder = foldEQ_locale;
1474 do_exactf_non_utf8: /* Neither pattern nor string are UTF8 */
1476 /* The idea in the non-utf8 EXACTF* cases is to first find the
1477 * first character of the EXACTF* node and then, if necessary,
1478 * case-insensitively compare the full text of the node. c1 is the
1479 * first character. c2 is its fold. This logic will not work for
1480 * Unicode semantics and the german sharp ss, which hence should
1481 * not be compiled into a node that gets here. */
1482 pat_string = STRING(c);
1483 ln = STR_LEN(c); /* length to match in octets/bytes */
1485 e = HOP3c(strend, -((I32)ln), s);
1487 if (!reginfo && e < s) {
1488 e = s; /* Due to minlen logic of intuit() */
1492 c2 = fold_array[c1];
1493 if (c1 == c2) { /* If char and fold are the same */
1494 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1);
1497 REXEC_FBC_EXACTISH_SCAN(*(U8*)s == c1 || *(U8*)s == c2);
1503 /* If one of the operands is in utf8, we can't use the simpler
1504 * folding above, due to the fact that many different characters
1505 * can have the same fold, or portion of a fold, or different-
1507 pat_string = STRING(c);
1508 ln = STR_LEN(c); /* length to match in octets/bytes */
1509 pat_end = pat_string + ln;
1510 lnc = (UTF_PATTERN) /* length to match in characters */
1511 ? utf8_length((U8 *) pat_string, (U8 *) pat_end)
1514 /* Set the end position to the final character available */
1515 e = HOP3c(strend, -1, s);
1517 if (!reginfo && e < s) {
1518 e = s; /* Due to minlen logic of intuit() */
1522 char *my_strend= (char *)strend;
1523 if (foldEQ_utf8_flags(s, &my_strend, 0, utf8_target,
1524 pat_string, NULL, ln, cBOOL(UTF_PATTERN), utf8_fold_flags)
1525 && (!reginfo || regtry(reginfo, &s)) )
1529 s += (utf8_target) ? UTF8SKIP(s) : 1;
1533 PL_reg_flags |= RF_tainted;
1534 FBC_BOUND(isALNUM_LC,
1535 isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp)),
1536 isALNUM_LC_utf8((U8*)s));
1539 PL_reg_flags |= RF_tainted;
1540 FBC_NBOUND(isALNUM_LC,
1541 isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp)),
1542 isALNUM_LC_utf8((U8*)s));
1545 FBC_BOUND(isWORDCHAR,
1547 cBOOL(swash_fetch(PL_utf8_alnum, (U8*)s, utf8_target)));
1550 FBC_BOUND_NOLOAD(isWORDCHAR_A,
1552 isWORDCHAR_A((U8*)s));
1555 FBC_NBOUND(isWORDCHAR,
1557 cBOOL(swash_fetch(PL_utf8_alnum, (U8*)s, utf8_target)));
1560 FBC_NBOUND_NOLOAD(isWORDCHAR_A,
1562 isWORDCHAR_A((U8*)s));
1565 FBC_BOUND(isWORDCHAR_L1,
1567 cBOOL(swash_fetch(PL_utf8_alnum, (U8*)s, utf8_target)));
1570 FBC_NBOUND(isWORDCHAR_L1,
1572 cBOOL(swash_fetch(PL_utf8_alnum, (U8*)s, utf8_target)));
1575 REXEC_FBC_CSCAN_TAINT(
1576 isALNUM_LC_utf8((U8*)s),
1581 REXEC_FBC_CSCAN_PRELOAD(
1582 LOAD_UTF8_CHARCLASS_ALNUM(),
1583 swash_fetch(PL_utf8_alnum,(U8*)s, utf8_target),
1584 isWORDCHAR_L1((U8) *s)
1588 REXEC_FBC_CSCAN_PRELOAD(
1589 LOAD_UTF8_CHARCLASS_ALNUM(),
1590 swash_fetch(PL_utf8_alnum,(U8*)s, utf8_target),
1595 /* Don't need to worry about utf8, as it can match only a single
1596 * byte invariant character */
1597 REXEC_FBC_CLASS_SCAN( isWORDCHAR_A(*s));
1600 REXEC_FBC_CSCAN_PRELOAD(
1601 LOAD_UTF8_CHARCLASS_ALNUM(),
1602 !swash_fetch(PL_utf8_alnum,(U8*)s, utf8_target),
1603 ! isWORDCHAR_L1((U8) *s)
1607 REXEC_FBC_CSCAN_PRELOAD(
1608 LOAD_UTF8_CHARCLASS_ALNUM(),
1609 !swash_fetch(PL_utf8_alnum, (U8*)s, utf8_target),
1620 REXEC_FBC_CSCAN_TAINT(
1621 !isALNUM_LC_utf8((U8*)s),
1626 REXEC_FBC_CSCAN_PRELOAD(
1627 LOAD_UTF8_CHARCLASS_SPACE(),
1628 *s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, utf8_target),
1633 REXEC_FBC_CSCAN_PRELOAD(
1634 LOAD_UTF8_CHARCLASS_SPACE(),
1635 *s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, utf8_target),
1640 /* Don't need to worry about utf8, as it can match only a single
1641 * byte invariant character */
1642 REXEC_FBC_CLASS_SCAN( isSPACE_A(*s));
1645 REXEC_FBC_CSCAN_TAINT(
1646 isSPACE_LC_utf8((U8*)s),
1651 REXEC_FBC_CSCAN_PRELOAD(
1652 LOAD_UTF8_CHARCLASS_SPACE(),
1653 !( *s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, utf8_target)),
1654 ! isSPACE_L1((U8) *s)
1658 REXEC_FBC_CSCAN_PRELOAD(
1659 LOAD_UTF8_CHARCLASS_SPACE(),
1660 !(*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, utf8_target)),
1671 REXEC_FBC_CSCAN_TAINT(
1672 !isSPACE_LC_utf8((U8*)s),
1677 REXEC_FBC_CSCAN_PRELOAD(
1678 LOAD_UTF8_CHARCLASS_DIGIT(),
1679 swash_fetch(PL_utf8_digit,(U8*)s, utf8_target),
1684 /* Don't need to worry about utf8, as it can match only a single
1685 * byte invariant character */
1686 REXEC_FBC_CLASS_SCAN( isDIGIT_A(*s));
1689 REXEC_FBC_CSCAN_TAINT(
1690 isDIGIT_LC_utf8((U8*)s),
1695 REXEC_FBC_CSCAN_PRELOAD(
1696 LOAD_UTF8_CHARCLASS_DIGIT(),
1697 !swash_fetch(PL_utf8_digit,(U8*)s, utf8_target),
1708 REXEC_FBC_CSCAN_TAINT(
1709 !isDIGIT_LC_utf8((U8*)s),
1716 is_LNBREAK_latin1(s)
1728 !is_VERTWS_latin1(s)
1734 is_HORIZWS_latin1(s)
1739 !is_HORIZWS_utf8(s),
1740 !is_HORIZWS_latin1(s)
1747 /* what trie are we using right now */
1749 = (reg_ac_data*)progi->data->data[ ARG( c ) ];
1751 = (reg_trie_data*)progi->data->data[ aho->trie ];
1752 HV *widecharmap = MUTABLE_HV(progi->data->data[ aho->trie + 1 ]);
1754 const char *last_start = strend - trie->minlen;
1756 const char *real_start = s;
1758 STRLEN maxlen = trie->maxlen;
1760 U8 **points; /* map of where we were in the input string
1761 when reading a given char. For ASCII this
1762 is unnecessary overhead as the relationship
1763 is always 1:1, but for Unicode, especially
1764 case folded Unicode this is not true. */
1765 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1769 GET_RE_DEBUG_FLAGS_DECL;
1771 /* We can't just allocate points here. We need to wrap it in
1772 * an SV so it gets freed properly if there is a croak while
1773 * running the match */
1776 sv_points=newSV(maxlen * sizeof(U8 *));
1777 SvCUR_set(sv_points,
1778 maxlen * sizeof(U8 *));
1779 SvPOK_on(sv_points);
1780 sv_2mortal(sv_points);
1781 points=(U8**)SvPV_nolen(sv_points );
1782 if ( trie_type != trie_utf8_fold
1783 && (trie->bitmap || OP(c)==AHOCORASICKC) )
1786 bitmap=(U8*)trie->bitmap;
1788 bitmap=(U8*)ANYOF_BITMAP(c);
1790 /* this is the Aho-Corasick algorithm modified a touch
1791 to include special handling for long "unknown char"
1792 sequences. The basic idea being that we use AC as long
1793 as we are dealing with a possible matching char, when
1794 we encounter an unknown char (and we have not encountered
1795 an accepting state) we scan forward until we find a legal
1797 AC matching is basically that of trie matching, except
1798 that when we encounter a failing transition, we fall back
1799 to the current states "fail state", and try the current char
1800 again, a process we repeat until we reach the root state,
1801 state 1, or a legal transition. If we fail on the root state
1802 then we can either terminate if we have reached an accepting
1803 state previously, or restart the entire process from the beginning
1807 while (s <= last_start) {
1808 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1816 U8 *uscan = (U8*)NULL;
1817 U8 *leftmost = NULL;
1819 U32 accepted_word= 0;
1823 while ( state && uc <= (U8*)strend ) {
1825 U32 word = aho->states[ state ].wordnum;
1829 DEBUG_TRIE_EXECUTE_r(
1830 if ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1831 dump_exec_pos( (char *)uc, c, strend, real_start,
1832 (char *)uc, utf8_target );
1833 PerlIO_printf( Perl_debug_log,
1834 " Scanning for legal start char...\n");
1838 while ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1842 while ( uc <= (U8*)last_start && !BITMAP_TEST(bitmap,*uc) ) {
1848 if (uc >(U8*)last_start) break;
1852 U8 *lpos= points[ (pointpos - trie->wordinfo[word].len) % maxlen ];
1853 if (!leftmost || lpos < leftmost) {
1854 DEBUG_r(accepted_word=word);
1860 points[pointpos++ % maxlen]= uc;
1861 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
1862 uscan, len, uvc, charid, foldlen,
1864 DEBUG_TRIE_EXECUTE_r({
1865 dump_exec_pos( (char *)uc, c, strend, real_start,
1867 PerlIO_printf(Perl_debug_log,
1868 " Charid:%3u CP:%4"UVxf" ",
1874 word = aho->states[ state ].wordnum;
1876 base = aho->states[ state ].trans.base;
1878 DEBUG_TRIE_EXECUTE_r({
1880 dump_exec_pos( (char *)uc, c, strend, real_start,
1882 PerlIO_printf( Perl_debug_log,
1883 "%sState: %4"UVxf", word=%"UVxf,
1884 failed ? " Fail transition to " : "",
1885 (UV)state, (UV)word);
1891 ( ((offset = base + charid
1892 - 1 - trie->uniquecharcount)) >= 0)
1893 && ((U32)offset < trie->lasttrans)
1894 && trie->trans[offset].check == state
1895 && (tmp=trie->trans[offset].next))
1897 DEBUG_TRIE_EXECUTE_r(
1898 PerlIO_printf( Perl_debug_log," - legal\n"));
1903 DEBUG_TRIE_EXECUTE_r(
1904 PerlIO_printf( Perl_debug_log," - fail\n"));
1906 state = aho->fail[state];
1910 /* we must be accepting here */
1911 DEBUG_TRIE_EXECUTE_r(
1912 PerlIO_printf( Perl_debug_log," - accepting\n"));
1921 if (!state) state = 1;
1924 if ( aho->states[ state ].wordnum ) {
1925 U8 *lpos = points[ (pointpos - trie->wordinfo[aho->states[ state ].wordnum].len) % maxlen ];
1926 if (!leftmost || lpos < leftmost) {
1927 DEBUG_r(accepted_word=aho->states[ state ].wordnum);
1932 s = (char*)leftmost;
1933 DEBUG_TRIE_EXECUTE_r({
1935 Perl_debug_log,"Matches word #%"UVxf" at position %"IVdf". Trying full pattern...\n",
1936 (UV)accepted_word, (IV)(s - real_start)
1939 if (!reginfo || regtry(reginfo, &s)) {
1945 DEBUG_TRIE_EXECUTE_r({
1946 PerlIO_printf( Perl_debug_log,"Pattern failed. Looking for new start point...\n");
1949 DEBUG_TRIE_EXECUTE_r(
1950 PerlIO_printf( Perl_debug_log,"No match.\n"));
1959 Perl_croak(aTHX_ "panic: unknown regstclass %d", (int)OP(c));
1969 - regexec_flags - match a regexp against a string
1972 Perl_regexec_flags(pTHX_ REGEXP * const rx, char *stringarg, register char *strend,
1973 char *strbeg, I32 minend, SV *sv, void *data, U32 flags)
1974 /* strend: pointer to null at end of string */
1975 /* strbeg: real beginning of string */
1976 /* minend: end of match must be >=minend after stringarg. */
1977 /* data: May be used for some additional optimizations.
1978 Currently its only used, with a U32 cast, for transmitting
1979 the ganch offset when doing a /g match. This will change */
1980 /* nosave: For optimizations. */
1983 struct regexp *const prog = (struct regexp *)SvANY(rx);
1984 /*register*/ char *s;
1985 register regnode *c;
1986 /*register*/ char *startpos = stringarg;
1987 I32 minlen; /* must match at least this many chars */
1988 I32 dontbother = 0; /* how many characters not to try at end */
1989 I32 end_shift = 0; /* Same for the end. */ /* CC */
1990 I32 scream_pos = -1; /* Internal iterator of scream. */
1991 char *scream_olds = NULL;
1992 const bool utf8_target = cBOOL(DO_UTF8(sv));
1994 RXi_GET_DECL(prog,progi);
1995 regmatch_info reginfo; /* create some info to pass to regtry etc */
1996 regexp_paren_pair *swap = NULL;
1997 GET_RE_DEBUG_FLAGS_DECL;
1999 PERL_ARGS_ASSERT_REGEXEC_FLAGS;
2000 PERL_UNUSED_ARG(data);
2002 /* Be paranoid... */
2003 if (prog == NULL || startpos == NULL) {
2004 Perl_croak(aTHX_ "NULL regexp parameter");
2008 multiline = prog->extflags & RXf_PMf_MULTILINE;
2009 reginfo.prog = rx; /* Yes, sorry that this is confusing. */
2011 RX_MATCH_UTF8_set(rx, utf8_target);
2013 debug_start_match(rx, utf8_target, startpos, strend,
2017 minlen = prog->minlen;
2019 if (strend - startpos < (minlen+(prog->check_offset_min<0?prog->check_offset_min:0))) {
2020 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
2021 "String too short [regexec_flags]...\n"));
2026 /* Check validity of program. */
2027 if (UCHARAT(progi->program) != REG_MAGIC) {
2028 Perl_croak(aTHX_ "corrupted regexp program");
2032 PL_reg_eval_set = 0;
2036 PL_reg_flags |= RF_utf8;
2038 /* Mark beginning of line for ^ and lookbehind. */
2039 reginfo.bol = startpos; /* XXX not used ??? */
2043 /* Mark end of line for $ (and such) */
2046 /* see how far we have to get to not match where we matched before */
2047 reginfo.till = startpos+minend;
2049 /* If there is a "must appear" string, look for it. */
2052 if (prog->extflags & RXf_GPOS_SEEN) { /* Need to set reginfo->ganch */
2054 if (flags & REXEC_IGNOREPOS){ /* Means: check only at start */
2055 reginfo.ganch = startpos + prog->gofs;
2056 DEBUG_GPOS_r(PerlIO_printf(Perl_debug_log,
2057 "GPOS IGNOREPOS: reginfo.ganch = startpos + %"UVxf"\n",(UV)prog->gofs));
2058 } else if (sv && SvTYPE(sv) >= SVt_PVMG
2060 && (mg = mg_find(sv, PERL_MAGIC_regex_global))
2061 && mg->mg_len >= 0) {
2062 reginfo.ganch = strbeg + mg->mg_len; /* Defined pos() */
2063 DEBUG_GPOS_r(PerlIO_printf(Perl_debug_log,
2064 "GPOS MAGIC: reginfo.ganch = strbeg + %"IVdf"\n",(IV)mg->mg_len));
2066 if (prog->extflags & RXf_ANCH_GPOS) {
2067 if (s > reginfo.ganch)
2069 s = reginfo.ganch - prog->gofs;
2070 DEBUG_GPOS_r(PerlIO_printf(Perl_debug_log,
2071 "GPOS ANCH_GPOS: s = ganch - %"UVxf"\n",(UV)prog->gofs));
2077 reginfo.ganch = strbeg + PTR2UV(data);
2078 DEBUG_GPOS_r(PerlIO_printf(Perl_debug_log,
2079 "GPOS DATA: reginfo.ganch= strbeg + %"UVxf"\n",PTR2UV(data)));
2081 } else { /* pos() not defined */
2082 reginfo.ganch = strbeg;
2083 DEBUG_GPOS_r(PerlIO_printf(Perl_debug_log,
2084 "GPOS: reginfo.ganch = strbeg\n"));
2087 if (PL_curpm && (PM_GETRE(PL_curpm) == rx)) {
2088 /* We have to be careful. If the previous successful match
2089 was from this regex we don't want a subsequent partially
2090 successful match to clobber the old results.
2091 So when we detect this possibility we add a swap buffer
2092 to the re, and switch the buffer each match. If we fail
2093 we switch it back, otherwise we leave it swapped.
2096 /* do we need a save destructor here for eval dies? */
2097 Newxz(prog->offs, (prog->nparens + 1), regexp_paren_pair);
2099 if (!(flags & REXEC_CHECKED) && (prog->check_substr != NULL || prog->check_utf8 != NULL)) {
2100 re_scream_pos_data d;
2102 d.scream_olds = &scream_olds;
2103 d.scream_pos = &scream_pos;
2104 s = re_intuit_start(rx, sv, s, strend, flags, &d);
2106 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not present...\n"));
2107 goto phooey; /* not present */
2113 /* Simplest case: anchored match need be tried only once. */
2114 /* [unless only anchor is BOL and multiline is set] */
2115 if (prog->extflags & (RXf_ANCH & ~RXf_ANCH_GPOS)) {
2116 if (s == startpos && regtry(®info, &startpos))
2118 else if (multiline || (prog->intflags & PREGf_IMPLICIT)
2119 || (prog->extflags & RXf_ANCH_MBOL)) /* XXXX SBOL? */
2124 dontbother = minlen - 1;
2125 end = HOP3c(strend, -dontbother, strbeg) - 1;
2126 /* for multiline we only have to try after newlines */
2127 if (prog->check_substr || prog->check_utf8) {
2128 /* because of the goto we can not easily reuse the macros for bifurcating the
2129 unicode/non-unicode match modes here like we do elsewhere - demerphq */
2132 goto after_try_utf8;
2134 if (regtry(®info, &s)) {
2141 if (prog->extflags & RXf_USE_INTUIT) {
2142 s = re_intuit_start(rx, sv, s + UTF8SKIP(s), strend, flags, NULL);
2151 } /* end search for check string in unicode */
2153 if (s == startpos) {
2154 goto after_try_latin;
2157 if (regtry(®info, &s)) {
2164 if (prog->extflags & RXf_USE_INTUIT) {
2165 s = re_intuit_start(rx, sv, s + 1, strend, flags, NULL);
2174 } /* end search for check string in latin*/
2175 } /* end search for check string */
2176 else { /* search for newline */
2178 /*XXX: The s-- is almost definitely wrong here under unicode - demeprhq*/
2181 /* We can use a more efficient search as newlines are the same in unicode as they are in latin */
2183 if (*s++ == '\n') { /* don't need PL_utf8skip here */
2184 if (regtry(®info, &s))
2188 } /* end search for newline */
2189 } /* end anchored/multiline check string search */
2191 } else if (RXf_GPOS_CHECK == (prog->extflags & RXf_GPOS_CHECK))
2193 /* the warning about reginfo.ganch being used without initialization
2194 is bogus -- we set it above, when prog->extflags & RXf_GPOS_SEEN
2195 and we only enter this block when the same bit is set. */
2196 char *tmp_s = reginfo.ganch - prog->gofs;
2198 if (tmp_s >= strbeg && regtry(®info, &tmp_s))
2203 /* Messy cases: unanchored match. */
2204 if ((prog->anchored_substr || prog->anchored_utf8) && prog->intflags & PREGf_SKIP) {
2205 /* we have /x+whatever/ */
2206 /* it must be a one character string (XXXX Except UTF_PATTERN?) */
2211 if (!(utf8_target ? prog->anchored_utf8 : prog->anchored_substr))
2212 utf8_target ? to_utf8_substr(prog) : to_byte_substr(prog);
2213 ch = SvPVX_const(utf8_target ? prog->anchored_utf8 : prog->anchored_substr)[0];
2218 DEBUG_EXECUTE_r( did_match = 1 );
2219 if (regtry(®info, &s)) goto got_it;
2221 while (s < strend && *s == ch)
2229 DEBUG_EXECUTE_r( did_match = 1 );
2230 if (regtry(®info, &s)) goto got_it;
2232 while (s < strend && *s == ch)
2237 DEBUG_EXECUTE_r(if (!did_match)
2238 PerlIO_printf(Perl_debug_log,
2239 "Did not find anchored character...\n")
2242 else if (prog->anchored_substr != NULL
2243 || prog->anchored_utf8 != NULL
2244 || ((prog->float_substr != NULL || prog->float_utf8 != NULL)
2245 && prog->float_max_offset < strend - s)) {
2250 char *last1; /* Last position checked before */
2254 if (prog->anchored_substr || prog->anchored_utf8) {
2255 if (!(utf8_target ? prog->anchored_utf8 : prog->anchored_substr))
2256 utf8_target ? to_utf8_substr(prog) : to_byte_substr(prog);
2257 must = utf8_target ? prog->anchored_utf8 : prog->anchored_substr;
2258 back_max = back_min = prog->anchored_offset;
2260 if (!(utf8_target ? prog->float_utf8 : prog->float_substr))
2261 utf8_target ? to_utf8_substr(prog) : to_byte_substr(prog);
2262 must = utf8_target ? prog->float_utf8 : prog->float_substr;
2263 back_max = prog->float_max_offset;
2264 back_min = prog->float_min_offset;
2268 if (must == &PL_sv_undef)
2269 /* could not downgrade utf8 check substring, so must fail */
2275 last = HOP3c(strend, /* Cannot start after this */
2276 -(I32)(CHR_SVLEN(must)
2277 - (SvTAIL(must) != 0) + back_min), strbeg);
2280 last1 = HOPc(s, -1);
2282 last1 = s - 1; /* bogus */
2284 /* XXXX check_substr already used to find "s", can optimize if
2285 check_substr==must. */
2287 dontbother = end_shift;
2288 strend = HOPc(strend, -dontbother);
2289 while ( (s <= last) &&
2290 ((flags & REXEC_SCREAM)
2291 ? (s = screaminstr(sv, must, HOP3c(s, back_min, (back_min<0 ? strbeg : strend)) - strbeg,
2292 end_shift, &scream_pos, 0))
2293 : (s = fbm_instr((unsigned char*)HOP3(s, back_min, (back_min<0 ? strbeg : strend)),
2294 (unsigned char*)strend, must,
2295 multiline ? FBMrf_MULTILINE : 0))) ) {
2296 /* we may be pointing at the wrong string */
2297 if ((flags & REXEC_SCREAM) && RXp_MATCH_COPIED(prog))
2298 s = strbeg + (s - SvPVX_const(sv));
2299 DEBUG_EXECUTE_r( did_match = 1 );
2300 if (HOPc(s, -back_max) > last1) {
2301 last1 = HOPc(s, -back_min);
2302 s = HOPc(s, -back_max);
2305 char * const t = (last1 >= PL_bostr) ? HOPc(last1, 1) : last1 + 1;
2307 last1 = HOPc(s, -back_min);
2311 while (s <= last1) {
2312 if (regtry(®info, &s))
2318 while (s <= last1) {
2319 if (regtry(®info, &s))
2325 DEBUG_EXECUTE_r(if (!did_match) {
2326 RE_PV_QUOTED_DECL(quoted, utf8_target, PERL_DEBUG_PAD_ZERO(0),
2327 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
2328 PerlIO_printf(Perl_debug_log, "Did not find %s substr %s%s...\n",
2329 ((must == prog->anchored_substr || must == prog->anchored_utf8)
2330 ? "anchored" : "floating"),
2331 quoted, RE_SV_TAIL(must));
2335 else if ( (c = progi->regstclass) ) {
2337 const OPCODE op = OP(progi->regstclass);
2338 /* don't bother with what can't match */
2339 if (PL_regkind[op] != EXACT && op != CANY && PL_regkind[op] != TRIE)
2340 strend = HOPc(strend, -(minlen - 1));
2343 SV * const prop = sv_newmortal();
2344 regprop(prog, prop, c);
2346 RE_PV_QUOTED_DECL(quoted,utf8_target,PERL_DEBUG_PAD_ZERO(1),
2348 PerlIO_printf(Perl_debug_log,
2349 "Matching stclass %.*s against %s (%d bytes)\n",
2350 (int)SvCUR(prop), SvPVX_const(prop),
2351 quoted, (int)(strend - s));
2354 if (find_byclass(prog, c, s, strend, ®info))
2356 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass... [regexec_flags]\n"));
2360 if (prog->float_substr != NULL || prog->float_utf8 != NULL) {
2365 if (!(utf8_target ? prog->float_utf8 : prog->float_substr))
2366 utf8_target ? to_utf8_substr(prog) : to_byte_substr(prog);
2367 float_real = utf8_target ? prog->float_utf8 : prog->float_substr;
2369 if (flags & REXEC_SCREAM) {
2370 last = screaminstr(sv, float_real, s - strbeg,
2371 end_shift, &scream_pos, 1); /* last one */
2373 last = scream_olds; /* Only one occurrence. */
2374 /* we may be pointing at the wrong string */
2375 else if (RXp_MATCH_COPIED(prog))
2376 s = strbeg + (s - SvPVX_const(sv));
2380 const char * const little = SvPV_const(float_real, len);
2382 if (SvTAIL(float_real)) {
2383 if (memEQ(strend - len + 1, little, len - 1))
2384 last = strend - len + 1;
2385 else if (!multiline)
2386 last = memEQ(strend - len, little, len)
2387 ? strend - len : NULL;
2393 last = rninstr(s, strend, little, little + len);
2395 last = strend; /* matching "$" */
2400 PerlIO_printf(Perl_debug_log,
2401 "%sCan't trim the tail, match fails (should not happen)%s\n",
2402 PL_colors[4], PL_colors[5]));
2403 goto phooey; /* Should not happen! */
2405 dontbother = strend - last + prog->float_min_offset;
2407 if (minlen && (dontbother < minlen))
2408 dontbother = minlen - 1;
2409 strend -= dontbother; /* this one's always in bytes! */
2410 /* We don't know much -- general case. */
2413 if (regtry(®info, &s))
2422 if (regtry(®info, &s))
2424 } while (s++ < strend);
2433 RX_MATCH_TAINTED_set(rx, PL_reg_flags & RF_tainted);
2435 if (PL_reg_eval_set)
2436 restore_pos(aTHX_ prog);
2437 if (RXp_PAREN_NAMES(prog))
2438 (void)hv_iterinit(RXp_PAREN_NAMES(prog));
2440 /* make sure $`, $&, $', and $digit will work later */
2441 if ( !(flags & REXEC_NOT_FIRST) ) {
2442 RX_MATCH_COPY_FREE(rx);
2443 if (flags & REXEC_COPY_STR) {
2444 const I32 i = PL_regeol - startpos + (stringarg - strbeg);
2445 #ifdef PERL_OLD_COPY_ON_WRITE
2447 || (SvFLAGS(sv) & CAN_COW_MASK) == CAN_COW_FLAGS)) {
2449 PerlIO_printf(Perl_debug_log,
2450 "Copy on write: regexp capture, type %d\n",
2453 prog->saved_copy = sv_setsv_cow(prog->saved_copy, sv);
2454 prog->subbeg = (char *)SvPVX_const(prog->saved_copy);
2455 assert (SvPOKp(prog->saved_copy));
2459 RX_MATCH_COPIED_on(rx);
2460 s = savepvn(strbeg, i);
2466 prog->subbeg = strbeg;
2467 prog->sublen = PL_regeol - strbeg; /* strend may have been modified */
2474 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch failed%s\n",
2475 PL_colors[4], PL_colors[5]));
2476 if (PL_reg_eval_set)
2477 restore_pos(aTHX_ prog);
2479 /* we failed :-( roll it back */
2480 Safefree(prog->offs);
2489 - regtry - try match at specific point
2491 STATIC I32 /* 0 failure, 1 success */
2492 S_regtry(pTHX_ regmatch_info *reginfo, char **startpos)
2496 REGEXP *const rx = reginfo->prog;
2497 regexp *const prog = (struct regexp *)SvANY(rx);
2498 RXi_GET_DECL(prog,progi);
2499 GET_RE_DEBUG_FLAGS_DECL;
2501 PERL_ARGS_ASSERT_REGTRY;
2503 reginfo->cutpoint=NULL;
2505 if ((prog->extflags & RXf_EVAL_SEEN) && !PL_reg_eval_set) {
2508 PL_reg_eval_set = RS_init;
2509 DEBUG_EXECUTE_r(DEBUG_s(
2510 PerlIO_printf(Perl_debug_log, " setting stack tmpbase at %"IVdf"\n",
2511 (IV)(PL_stack_sp - PL_stack_base));
2514 cxstack[cxstack_ix].blk_oldsp = PL_stack_sp - PL_stack_base;
2515 /* Otherwise OP_NEXTSTATE will free whatever on stack now. */
2517 /* Apparently this is not needed, judging by wantarray. */
2518 /* SAVEI8(cxstack[cxstack_ix].blk_gimme);
2519 cxstack[cxstack_ix].blk_gimme = G_SCALAR; */
2522 /* Make $_ available to executed code. */
2523 if (reginfo->sv != DEFSV) {
2525 DEFSV_set(reginfo->sv);
2528 if (!(SvTYPE(reginfo->sv) >= SVt_PVMG && SvMAGIC(reginfo->sv)
2529 && (mg = mg_find(reginfo->sv, PERL_MAGIC_regex_global)))) {
2530 /* prepare for quick setting of pos */
2531 #ifdef PERL_OLD_COPY_ON_WRITE
2532 if (SvIsCOW(reginfo->sv))
2533 sv_force_normal_flags(reginfo->sv, 0);
2535 mg = sv_magicext(reginfo->sv, NULL, PERL_MAGIC_regex_global,
2536 &PL_vtbl_mglob, NULL, 0);
2540 PL_reg_oldpos = mg->mg_len;
2541 SAVEDESTRUCTOR_X(restore_pos, prog);
2543 if (!PL_reg_curpm) {
2544 Newxz(PL_reg_curpm, 1, PMOP);
2547 SV* const repointer = &PL_sv_undef;
2548 /* this regexp is also owned by the new PL_reg_curpm, which
2549 will try to free it. */
2550 av_push(PL_regex_padav, repointer);
2551 PL_reg_curpm->op_pmoffset = av_len(PL_regex_padav);
2552 PL_regex_pad = AvARRAY(PL_regex_padav);
2557 /* It seems that non-ithreads works both with and without this code.
2558 So for efficiency reasons it seems best not to have the code
2559 compiled when it is not needed. */
2560 /* This is safe against NULLs: */
2561 ReREFCNT_dec(PM_GETRE(PL_reg_curpm));
2562 /* PM_reg_curpm owns a reference to this regexp. */
2563 (void)ReREFCNT_inc(rx);
2565 PM_SETRE(PL_reg_curpm, rx);
2566 PL_reg_oldcurpm = PL_curpm;
2567 PL_curpm = PL_reg_curpm;
2568 if (RXp_MATCH_COPIED(prog)) {
2569 /* Here is a serious problem: we cannot rewrite subbeg,
2570 since it may be needed if this match fails. Thus
2571 $` inside (?{}) could fail... */
2572 PL_reg_oldsaved = prog->subbeg;
2573 PL_reg_oldsavedlen = prog->sublen;
2574 #ifdef PERL_OLD_COPY_ON_WRITE
2575 PL_nrs = prog->saved_copy;
2577 RXp_MATCH_COPIED_off(prog);
2580 PL_reg_oldsaved = NULL;
2581 prog->subbeg = PL_bostr;
2582 prog->sublen = PL_regeol - PL_bostr; /* strend may have been modified */
2584 DEBUG_EXECUTE_r(PL_reg_starttry = *startpos);
2585 prog->offs[0].start = *startpos - PL_bostr;
2586 PL_reginput = *startpos;
2587 PL_reglastparen = &prog->lastparen;
2588 PL_reglastcloseparen = &prog->lastcloseparen;
2589 prog->lastparen = 0;
2590 prog->lastcloseparen = 0;
2592 PL_regoffs = prog->offs;
2593 if (PL_reg_start_tmpl <= prog->nparens) {
2594 PL_reg_start_tmpl = prog->nparens*3/2 + 3;
2595 if(PL_reg_start_tmp)
2596 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2598 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
2601 /* XXXX What this code is doing here?!!! There should be no need
2602 to do this again and again, PL_reglastparen should take care of
2605 /* Tests pat.t#187 and split.t#{13,14} seem to depend on this code.
2606 * Actually, the code in regcppop() (which Ilya may be meaning by
2607 * PL_reglastparen), is not needed at all by the test suite
2608 * (op/regexp, op/pat, op/split), but that code is needed otherwise
2609 * this erroneously leaves $1 defined: "1" =~ /^(?:(\d)x)?\d$/
2610 * Meanwhile, this code *is* needed for the
2611 * above-mentioned test suite tests to succeed. The common theme
2612 * on those tests seems to be returning null fields from matches.
2613 * --jhi updated by dapm */
2615 if (prog->nparens) {
2616 regexp_paren_pair *pp = PL_regoffs;
2618 for (i = prog->nparens; i > (I32)*PL_reglastparen; i--) {
2626 if (regmatch(reginfo, progi->program + 1)) {
2627 PL_regoffs[0].end = PL_reginput - PL_bostr;
2630 if (reginfo->cutpoint)
2631 *startpos= reginfo->cutpoint;
2632 REGCP_UNWIND(lastcp);
2637 #define sayYES goto yes
2638 #define sayNO goto no
2639 #define sayNO_SILENT goto no_silent
2641 /* we dont use STMT_START/END here because it leads to
2642 "unreachable code" warnings, which are bogus, but distracting. */
2643 #define CACHEsayNO \
2644 if (ST.cache_mask) \
2645 PL_reg_poscache[ST.cache_offset] |= ST.cache_mask; \
2648 /* this is used to determine how far from the left messages like
2649 'failed...' are printed. It should be set such that messages
2650 are inline with the regop output that created them.
2652 #define REPORT_CODE_OFF 32
2655 #define CHRTEST_UNINIT -1001 /* c1/c2 haven't been calculated yet */
2656 #define CHRTEST_VOID -1000 /* the c1/c2 "next char" test should be skipped */
2658 #define SLAB_FIRST(s) (&(s)->states[0])
2659 #define SLAB_LAST(s) (&(s)->states[PERL_REGMATCH_SLAB_SLOTS-1])
2661 /* grab a new slab and return the first slot in it */
2663 STATIC regmatch_state *
2666 #if PERL_VERSION < 9 && !defined(PERL_CORE)
2669 regmatch_slab *s = PL_regmatch_slab->next;
2671 Newx(s, 1, regmatch_slab);
2672 s->prev = PL_regmatch_slab;
2674 PL_regmatch_slab->next = s;
2676 PL_regmatch_slab = s;
2677 return SLAB_FIRST(s);
2681 /* push a new state then goto it */
2683 #define PUSH_STATE_GOTO(state, node) \
2685 st->resume_state = state; \
2688 /* push a new state with success backtracking, then goto it */
2690 #define PUSH_YES_STATE_GOTO(state, node) \
2692 st->resume_state = state; \
2693 goto push_yes_state;
2699 regmatch() - main matching routine
2701 This is basically one big switch statement in a loop. We execute an op,
2702 set 'next' to point the next op, and continue. If we come to a point which
2703 we may need to backtrack to on failure such as (A|B|C), we push a
2704 backtrack state onto the backtrack stack. On failure, we pop the top
2705 state, and re-enter the loop at the state indicated. If there are no more
2706 states to pop, we return failure.
2708 Sometimes we also need to backtrack on success; for example /A+/, where
2709 after successfully matching one A, we need to go back and try to
2710 match another one; similarly for lookahead assertions: if the assertion
2711 completes successfully, we backtrack to the state just before the assertion
2712 and then carry on. In these cases, the pushed state is marked as
2713 'backtrack on success too'. This marking is in fact done by a chain of
2714 pointers, each pointing to the previous 'yes' state. On success, we pop to
2715 the nearest yes state, discarding any intermediate failure-only states.
2716 Sometimes a yes state is pushed just to force some cleanup code to be
2717 called at the end of a successful match or submatch; e.g. (??{$re}) uses
2718 it to free the inner regex.
2720 Note that failure backtracking rewinds the cursor position, while
2721 success backtracking leaves it alone.
2723 A pattern is complete when the END op is executed, while a subpattern
2724 such as (?=foo) is complete when the SUCCESS op is executed. Both of these
2725 ops trigger the "pop to last yes state if any, otherwise return true"
2728 A common convention in this function is to use A and B to refer to the two
2729 subpatterns (or to the first nodes thereof) in patterns like /A*B/: so A is
2730 the subpattern to be matched possibly multiple times, while B is the entire
2731 rest of the pattern. Variable and state names reflect this convention.
2733 The states in the main switch are the union of ops and failure/success of
2734 substates associated with with that op. For example, IFMATCH is the op
2735 that does lookahead assertions /(?=A)B/ and so the IFMATCH state means
2736 'execute IFMATCH'; while IFMATCH_A is a state saying that we have just
2737 successfully matched A and IFMATCH_A_fail is a state saying that we have
2738 just failed to match A. Resume states always come in pairs. The backtrack
2739 state we push is marked as 'IFMATCH_A', but when that is popped, we resume
2740 at IFMATCH_A or IFMATCH_A_fail, depending on whether we are backtracking
2741 on success or failure.
2743 The struct that holds a backtracking state is actually a big union, with
2744 one variant for each major type of op. The variable st points to the
2745 top-most backtrack struct. To make the code clearer, within each
2746 block of code we #define ST to alias the relevant union.
2748 Here's a concrete example of a (vastly oversimplified) IFMATCH
2754 #define ST st->u.ifmatch
2756 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2757 ST.foo = ...; // some state we wish to save
2759 // push a yes backtrack state with a resume value of
2760 // IFMATCH_A/IFMATCH_A_fail, then continue execution at the
2762 PUSH_YES_STATE_GOTO(IFMATCH_A, A);
2765 case IFMATCH_A: // we have successfully executed A; now continue with B
2767 bar = ST.foo; // do something with the preserved value
2770 case IFMATCH_A_fail: // A failed, so the assertion failed
2771 ...; // do some housekeeping, then ...
2772 sayNO; // propagate the failure
2779 For any old-timers reading this who are familiar with the old recursive
2780 approach, the code above is equivalent to:
2782 case IFMATCH: // we are executing the IFMATCH op, (?=A)B
2791 ...; // do some housekeeping, then ...
2792 sayNO; // propagate the failure
2795 The topmost backtrack state, pointed to by st, is usually free. If you
2796 want to claim it, populate any ST.foo fields in it with values you wish to
2797 save, then do one of
2799 PUSH_STATE_GOTO(resume_state, node);
2800 PUSH_YES_STATE_GOTO(resume_state, node);
2802 which sets that backtrack state's resume value to 'resume_state', pushes a
2803 new free entry to the top of the backtrack stack, then goes to 'node'.
2804 On backtracking, the free slot is popped, and the saved state becomes the
2805 new free state. An ST.foo field in this new top state can be temporarily
2806 accessed to retrieve values, but once the main loop is re-entered, it
2807 becomes available for reuse.
2809 Note that the depth of the backtrack stack constantly increases during the
2810 left-to-right execution of the pattern, rather than going up and down with
2811 the pattern nesting. For example the stack is at its maximum at Z at the
2812 end of the pattern, rather than at X in the following:
2814 /(((X)+)+)+....(Y)+....Z/
2816 The only exceptions to this are lookahead/behind assertions and the cut,
2817 (?>A), which pop all the backtrack states associated with A before
2820 Backtrack state structs are allocated in slabs of about 4K in size.
2821 PL_regmatch_state and st always point to the currently active state,
2822 and PL_regmatch_slab points to the slab currently containing
2823 PL_regmatch_state. The first time regmatch() is called, the first slab is
2824 allocated, and is never freed until interpreter destruction. When the slab
2825 is full, a new one is allocated and chained to the end. At exit from
2826 regmatch(), slabs allocated since entry are freed.
2831 #define DEBUG_STATE_pp(pp) \
2833 DUMP_EXEC_POS(locinput, scan, utf8_target); \
2834 PerlIO_printf(Perl_debug_log, \
2835 " %*s"pp" %s%s%s%s%s\n", \
2837 PL_reg_name[st->resume_state], \
2838 ((st==yes_state||st==mark_state) ? "[" : ""), \
2839 ((st==yes_state) ? "Y" : ""), \
2840 ((st==mark_state) ? "M" : ""), \
2841 ((st==yes_state||st==mark_state) ? "]" : "") \
2846 #define REG_NODE_NUM(x) ((x) ? (int)((x)-prog) : -1)
2851 S_debug_start_match(pTHX_ const REGEXP *prog, const bool utf8_target,
2852 const char *start, const char *end, const char *blurb)
2854 const bool utf8_pat = RX_UTF8(prog) ? 1 : 0;
2856 PERL_ARGS_ASSERT_DEBUG_START_MATCH;
2861 RE_PV_QUOTED_DECL(s0, utf8_pat, PERL_DEBUG_PAD_ZERO(0),
2862 RX_PRECOMP_const(prog), RX_PRELEN(prog), 60);
2864 RE_PV_QUOTED_DECL(s1, utf8_target, PERL_DEBUG_PAD_ZERO(1),
2865 start, end - start, 60);
2867 PerlIO_printf(Perl_debug_log,
2868 "%s%s REx%s %s against %s\n",
2869 PL_colors[4], blurb, PL_colors[5], s0, s1);
2871 if (utf8_target||utf8_pat)
2872 PerlIO_printf(Perl_debug_log, "UTF-8 %s%s%s...\n",
2873 utf8_pat ? "pattern" : "",
2874 utf8_pat && utf8_target ? " and " : "",
2875 utf8_target ? "string" : ""
2881 S_dump_exec_pos(pTHX_ const char *locinput,
2882 const regnode *scan,
2883 const char *loc_regeol,
2884 const char *loc_bostr,
2885 const char *loc_reg_starttry,
2886 const bool utf8_target)
2888 const int docolor = *PL_colors[0] || *PL_colors[2] || *PL_colors[4];
2889 const int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */
2890 int l = (loc_regeol - locinput) > taill ? taill : (loc_regeol - locinput);
2891 /* The part of the string before starttry has one color
2892 (pref0_len chars), between starttry and current
2893 position another one (pref_len - pref0_len chars),
2894 after the current position the third one.
2895 We assume that pref0_len <= pref_len, otherwise we
2896 decrease pref0_len. */
2897 int pref_len = (locinput - loc_bostr) > (5 + taill) - l
2898 ? (5 + taill) - l : locinput - loc_bostr;
2901 PERL_ARGS_ASSERT_DUMP_EXEC_POS;
2903 while (utf8_target && UTF8_IS_CONTINUATION(*(U8*)(locinput - pref_len)))
2905 pref0_len = pref_len - (locinput - loc_reg_starttry);
2906 if (l + pref_len < (5 + taill) && l < loc_regeol - locinput)
2907 l = ( loc_regeol - locinput > (5 + taill) - pref_len
2908 ? (5 + taill) - pref_len : loc_regeol - locinput);
2909 while (utf8_target && UTF8_IS_CONTINUATION(*(U8*)(locinput + l)))
2913 if (pref0_len > pref_len)
2914 pref0_len = pref_len;
2916 const int is_uni = (utf8_target && OP(scan) != CANY) ? 1 : 0;
2918 RE_PV_COLOR_DECL(s0,len0,is_uni,PERL_DEBUG_PAD(0),
2919 (locinput - pref_len),pref0_len, 60, 4, 5);
2921 RE_PV_COLOR_DECL(s1,len1,is_uni,PERL_DEBUG_PAD(1),
2922 (locinput - pref_len + pref0_len),
2923 pref_len - pref0_len, 60, 2, 3);
2925 RE_PV_COLOR_DECL(s2,len2,is_uni,PERL_DEBUG_PAD(2),
2926 locinput, loc_regeol - locinput, 10, 0, 1);
2928 const STRLEN tlen=len0+len1+len2;
2929 PerlIO_printf(Perl_debug_log,
2930 "%4"IVdf" <%.*s%.*s%s%.*s>%*s|",
2931 (IV)(locinput - loc_bostr),
2934 (docolor ? "" : "> <"),
2936 (int)(tlen > 19 ? 0 : 19 - tlen),
2943 /* reg_check_named_buff_matched()
2944 * Checks to see if a named buffer has matched. The data array of
2945 * buffer numbers corresponding to the buffer is expected to reside
2946 * in the regexp->data->data array in the slot stored in the ARG() of
2947 * node involved. Note that this routine doesn't actually care about the
2948 * name, that information is not preserved from compilation to execution.
2949 * Returns the index of the leftmost defined buffer with the given name
2950 * or 0 if non of the buffers matched.
2953 S_reg_check_named_buff_matched(pTHX_ const regexp *rex, const regnode *scan)
2956 RXi_GET_DECL(rex,rexi);
2957 SV *sv_dat= MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
2958 I32 *nums=(I32*)SvPVX(sv_dat);
2960 PERL_ARGS_ASSERT_REG_CHECK_NAMED_BUFF_MATCHED;
2962 for ( n=0; n<SvIVX(sv_dat); n++ ) {
2963 if ((I32)*PL_reglastparen >= nums[n] &&
2964 PL_regoffs[nums[n]].end != -1)
2973 /* free all slabs above current one - called during LEAVE_SCOPE */
2976 S_clear_backtrack_stack(pTHX_ void *p)
2978 regmatch_slab *s = PL_regmatch_slab->next;
2983 PL_regmatch_slab->next = NULL;
2985 regmatch_slab * const osl = s;
2992 #define SETREX(Re1,Re2) \
2993 if (PL_reg_eval_set) PM_SETRE((PL_reg_curpm), (Re2)); \
2996 STATIC I32 /* 0 failure, 1 success */
2997 S_regmatch(pTHX_ regmatch_info *reginfo, regnode *prog)
2999 #if PERL_VERSION < 9 && !defined(PERL_CORE)
3003 register const bool utf8_target = PL_reg_match_utf8;
3004 const U32 uniflags = UTF8_ALLOW_DEFAULT;
3005 REGEXP *rex_sv = reginfo->prog;
3006 regexp *rex = (struct regexp *)SvANY(rex_sv);
3007 RXi_GET_DECL(rex,rexi);
3009 /* the current state. This is a cached copy of PL_regmatch_state */
3010 register regmatch_state *st;
3011 /* cache heavy used fields of st in registers */
3012 register regnode *scan;
3013 register regnode *next;
3014 register U32 n = 0; /* general value; init to avoid compiler warning */
3015 register I32 ln = 0; /* len or last; init to avoid compiler warning */
3016 register char *locinput = PL_reginput;
3017 register I32 nextchr; /* is always set to UCHARAT(locinput) */
3019 bool result = 0; /* return value of S_regmatch */
3020 int depth = 0; /* depth of backtrack stack */
3021 U32 nochange_depth = 0; /* depth of GOSUB recursion with nochange */
3022 const U32 max_nochange_depth =
3023 (3 * rex->nparens > MAX_RECURSE_EVAL_NOCHANGE_DEPTH) ?
3024 3 * rex->nparens : MAX_RECURSE_EVAL_NOCHANGE_DEPTH;
3025 regmatch_state *yes_state = NULL; /* state to pop to on success of
3027 /* mark_state piggy backs on the yes_state logic so that when we unwind
3028 the stack on success we can update the mark_state as we go */
3029 regmatch_state *mark_state = NULL; /* last mark state we have seen */
3030 regmatch_state *cur_eval = NULL; /* most recent EVAL_AB state */
3031 struct regmatch_state *cur_curlyx = NULL; /* most recent curlyx */
3033 bool no_final = 0; /* prevent failure from backtracking? */
3034 bool do_cutgroup = 0; /* no_final only until next branch/trie entry */
3035 char *startpoint = PL_reginput;
3036 SV *popmark = NULL; /* are we looking for a mark? */
3037 SV *sv_commit = NULL; /* last mark name seen in failure */
3038 SV *sv_yes_mark = NULL; /* last mark name we have seen
3039 during a successful match */
3040 U32 lastopen = 0; /* last open we saw */
3041 bool has_cutgroup = RX_HAS_CUTGROUP(rex) ? 1 : 0;
3042 SV* const oreplsv = GvSV(PL_replgv);
3043 /* these three flags are set by various ops to signal information to
3044 * the very next op. They have a useful lifetime of exactly one loop
3045 * iteration, and are not preserved or restored by state pushes/pops
3047 bool sw = 0; /* the condition value in (?(cond)a|b) */
3048 bool minmod = 0; /* the next "{n,m}" is a "{n,m}?" */
3049 int logical = 0; /* the following EVAL is:
3053 or the following IFMATCH/UNLESSM is:
3054 false: plain (?=foo)
3055 true: used as a condition: (?(?=foo))
3058 GET_RE_DEBUG_FLAGS_DECL;
3061 PERL_ARGS_ASSERT_REGMATCH;
3063 DEBUG_OPTIMISE_r( DEBUG_EXECUTE_r({
3064 PerlIO_printf(Perl_debug_log,"regmatch start\n");
3066 /* on first ever call to regmatch, allocate first slab */
3067 if (!PL_regmatch_slab) {
3068 Newx(PL_regmatch_slab, 1, regmatch_slab);
3069 PL_regmatch_slab->prev = NULL;
3070 PL_regmatch_slab->next = NULL;
3071 PL_regmatch_state = SLAB_FIRST(PL_regmatch_slab);
3074 oldsave = PL_savestack_ix;
3075 SAVEDESTRUCTOR_X(S_clear_backtrack_stack, NULL);
3076 SAVEVPTR(PL_regmatch_slab);
3077 SAVEVPTR(PL_regmatch_state);
3079 /* grab next free state slot */
3080 st = ++PL_regmatch_state;
3081 if (st > SLAB_LAST(PL_regmatch_slab))
3082 st = PL_regmatch_state = S_push_slab(aTHX);
3084 /* Note that nextchr is a byte even in UTF */
3085 nextchr = UCHARAT(locinput);
3087 while (scan != NULL) {
3090 SV * const prop = sv_newmortal();
3091 regnode *rnext=regnext(scan);
3092 DUMP_EXEC_POS( locinput, scan, utf8_target );
3093 regprop(rex, prop, scan);
3095 PerlIO_printf(Perl_debug_log,
3096 "%3"IVdf":%*s%s(%"IVdf")\n",
3097 (IV)(scan - rexi->program), depth*2, "",
3099 (PL_regkind[OP(scan)] == END || !rnext) ?
3100 0 : (IV)(rnext - rexi->program));
3103 next = scan + NEXT_OFF(scan);
3106 state_num = OP(scan);
3108 REH_CALL_EXEC_NODE_HOOK(rex, scan, reginfo, st);
3111 assert(PL_reglastparen == &rex->lastparen);
3112 assert(PL_reglastcloseparen == &rex->lastcloseparen);
3113 assert(PL_regoffs == rex->offs);
3115 switch (state_num) {
3117 if (locinput == PL_bostr)
3119 /* reginfo->till = reginfo->bol; */
3124 if (locinput == PL_bostr ||
3125 ((nextchr || locinput < PL_regeol) && locinput[-1] == '\n'))
3131 if (locinput == PL_bostr)
3135 if (locinput == reginfo->ganch)
3140 /* update the startpoint */
3141 st->u.keeper.val = PL_regoffs[0].start;
3142 PL_reginput = locinput;
3143 PL_regoffs[0].start = locinput - PL_bostr;
3144 PUSH_STATE_GOTO(KEEPS_next, next);
3146 case KEEPS_next_fail:
3147 /* rollback the start point change */
3148 PL_regoffs[0].start = st->u.keeper.val;
3154 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
3159 if ((nextchr || locinput < PL_regeol) && nextchr != '\n')
3161 if (PL_regeol - locinput > 1)
3165 if (PL_regeol != locinput)
3169 if (!nextchr && locinput >= PL_regeol)
3172 locinput += PL_utf8skip[nextchr];
3173 if (locinput > PL_regeol)
3175 nextchr = UCHARAT(locinput);
3178 nextchr = UCHARAT(++locinput);
3181 if (!nextchr && locinput >= PL_regeol)
3183 nextchr = UCHARAT(++locinput);
3186 if ((!nextchr && locinput >= PL_regeol) || nextchr == '\n')
3189 locinput += PL_utf8skip[nextchr];
3190 if (locinput > PL_regeol)
3192 nextchr = UCHARAT(locinput);
3195 nextchr = UCHARAT(++locinput);
3199 #define ST st->u.trie
3201 /* In this case the charclass data is available inline so
3202 we can fail fast without a lot of extra overhead.
3204 if (scan->flags == EXACT || !utf8_target) {
3205 if(!ANYOF_BITMAP_TEST(scan, *locinput)) {
3207 PerlIO_printf(Perl_debug_log,
3208 "%*s %sfailed to match trie start class...%s\n",
3209 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
3217 /* the basic plan of execution of the trie is:
3218 * At the beginning, run though all the states, and
3219 * find the longest-matching word. Also remember the position
3220 * of the shortest matching word. For example, this pattern:
3223 * when matched against the string "abcde", will generate
3224 * accept states for all words except 3, with the longest
3225 * matching word being 4, and the shortest being 1 (with
3226 * the position being after char 1 of the string).
3228 * Then for each matching word, in word order (i.e. 1,2,4,5),
3229 * we run the remainder of the pattern; on each try setting
3230 * the current position to the character following the word,
3231 * returning to try the next word on failure.
3233 * We avoid having to build a list of words at runtime by
3234 * using a compile-time structure, wordinfo[].prev, which
3235 * gives, for each word, the previous accepting word (if any).
3236 * In the case above it would contain the mappings 1->2, 2->0,
3237 * 3->0, 4->5, 5->1. We can use this table to generate, from
3238 * the longest word (4 above), a list of all words, by
3239 * following the list of prev pointers; this gives us the
3240 * unordered list 4,5,1,2. Then given the current word we have
3241 * just tried, we can go through the list and find the
3242 * next-biggest word to try (so if we just failed on word 2,
3243 * the next in the list is 4).
3245 * Since at runtime we don't record the matching position in
3246 * the string for each word, we have to work that out for
3247 * each word we're about to process. The wordinfo table holds
3248 * the character length of each word; given that we recorded
3249 * at the start: the position of the shortest word and its
3250 * length in chars, we just need to move the pointer the
3251 * difference between the two char lengths. Depending on
3252 * Unicode status and folding, that's cheap or expensive.
3254 * This algorithm is optimised for the case where are only a
3255 * small number of accept states, i.e. 0,1, or maybe 2.
3256 * With lots of accepts states, and having to try all of them,
3257 * it becomes quadratic on number of accept states to find all
3262 /* what type of TRIE am I? (utf8 makes this contextual) */
3263 DECL_TRIE_TYPE(scan);
3265 /* what trie are we using right now */
3266 reg_trie_data * const trie
3267 = (reg_trie_data*)rexi->data->data[ ARG( scan ) ];
3268 HV * widecharmap = MUTABLE_HV(rexi->data->data[ ARG( scan ) + 1 ]);
3269 U32 state = trie->startstate;
3271 if (trie->bitmap && trie_type != trie_utf8_fold &&
3272 !TRIE_BITMAP_TEST(trie,*locinput)
3274 if (trie->states[ state ].wordnum) {
3276 PerlIO_printf(Perl_debug_log,
3277 "%*s %smatched empty string...%s\n",
3278 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
3284 PerlIO_printf(Perl_debug_log,
3285 "%*s %sfailed to match trie start class...%s\n",
3286 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5])
3293 U8 *uc = ( U8* )locinput;
3297 U8 *uscan = (U8*)NULL;
3298 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
3299 U32 charcount = 0; /* how many input chars we have matched */
3300 U32 accepted = 0; /* have we seen any accepting states? */
3303 ST.jump = trie->jump;
3306 ST.longfold = FALSE; /* char longer if folded => it's harder */
3309 /* fully traverse the TRIE; note the position of the
3310 shortest accept state and the wordnum of the longest
3313 while ( state && uc <= (U8*)PL_regeol ) {
3314 U32 base = trie->states[ state ].trans.base;
3318 wordnum = trie->states[ state ].wordnum;
3320 if (wordnum) { /* it's an accept state */
3323 /* record first match position */
3325 ST.firstpos = (U8*)locinput;
3330 ST.firstchars = charcount;
3333 if (!ST.nextword || wordnum < ST.nextword)
3334 ST.nextword = wordnum;
3335 ST.topword = wordnum;
3338 DEBUG_TRIE_EXECUTE_r({
3339 DUMP_EXEC_POS( (char *)uc, scan, utf8_target );
3340 PerlIO_printf( Perl_debug_log,
3341 "%*s %sState: %4"UVxf" Accepted: %c ",
3342 2+depth * 2, "", PL_colors[4],
3343 (UV)state, (accepted ? 'Y' : 'N'));
3346 /* read a char and goto next state */
3349 REXEC_TRIE_READ_CHAR(trie_type, trie, widecharmap, uc,
3350 uscan, len, uvc, charid, foldlen,
3357 base + charid - 1 - trie->uniquecharcount)) >= 0)
3359 && ((U32)offset < trie->lasttrans)
3360 && trie->trans[offset].check == state)
3362 state = trie->trans[offset].next;
3373 DEBUG_TRIE_EXECUTE_r(
3374 PerlIO_printf( Perl_debug_log,
3375 "Charid:%3x CP:%4"UVxf" After State: %4"UVxf"%s\n",
3376 charid, uvc, (UV)state, PL_colors[5] );
3382 /* calculate total number of accept states */
3387 w = trie->wordinfo[w].prev;
3390 ST.accepted = accepted;
3394 PerlIO_printf( Perl_debug_log,
3395 "%*s %sgot %"IVdf" possible matches%s\n",
3396 REPORT_CODE_OFF + depth * 2, "",
3397 PL_colors[4], (IV)ST.accepted, PL_colors[5] );
3399 goto trie_first_try; /* jump into the fail handler */
3403 case TRIE_next_fail: /* we failed - try next alternative */
3405 REGCP_UNWIND(ST.cp);
3406 for (n = *PL_reglastparen; n > ST.lastparen; n--)
3407 PL_regoffs[n].end = -1;
3408 *PL_reglastparen = n;
3410 if (!--ST.accepted) {
3412 PerlIO_printf( Perl_debug_log,
3413 "%*s %sTRIE failed...%s\n",
3414 REPORT_CODE_OFF+depth*2, "",
3421 /* Find next-highest word to process. Note that this code
3422 * is O(N^2) per trie run (O(N) per branch), so keep tight */
3423 register U16 min = 0;
3425 register U16 const nextword = ST.nextword;
3426 register reg_trie_wordinfo * const wordinfo
3427 = ((reg_trie_data*)rexi->data->data[ARG(ST.me)])->wordinfo;
3428 for (word=ST.topword; word; word=wordinfo[word].prev) {
3429 if (word > nextword && (!min || word < min))
3442 ST.lastparen = *PL_reglastparen;
3446 /* find start char of end of current word */
3448 U32 chars; /* how many chars to skip */
3449 U8 *uc = ST.firstpos;
3450 reg_trie_data * const trie
3451 = (reg_trie_data*)rexi->data->data[ARG(ST.me)];
3453 assert((trie->wordinfo[ST.nextword].len - trie->prefixlen)
3455 chars = (trie->wordinfo[ST.nextword].len - trie->prefixlen)
3459 /* the hard option - fold each char in turn and find
3460 * its folded length (which may be different */
3461 U8 foldbuf[UTF8_MAXBYTES_CASE + 1];
3469 uvc = utf8n_to_uvuni((U8*)uc, UTF8_MAXLEN, &len,
3477 uvc = to_uni_fold(uvc, foldbuf, &foldlen);
3482 uvc = utf8n_to_uvuni(uscan, UTF8_MAXLEN, &len,
3496 PL_reginput = (char *)uc;
3499 scan = (ST.jump && ST.jump[ST.nextword])
3500 ? ST.me + ST.jump[ST.nextword]
3504 PerlIO_printf( Perl_debug_log,
3505 "%*s %sTRIE matched word #%d, continuing%s\n",
3506 REPORT_CODE_OFF+depth*2, "",
3513 if (ST.accepted > 1 || has_cutgroup) {
3514 PUSH_STATE_GOTO(TRIE_next, scan);
3517 /* only one choice left - just continue */
3519 AV *const trie_words
3520 = MUTABLE_AV(rexi->data->data[ARG(ST.me)+TRIE_WORDS_OFFSET]);
3521 SV ** const tmp = av_fetch( trie_words,
3523 SV *sv= tmp ? sv_newmortal() : NULL;
3525 PerlIO_printf( Perl_debug_log,
3526 "%*s %sonly one match left, short-circuiting: #%d <%s>%s\n",
3527 REPORT_CODE_OFF+depth*2, "", PL_colors[4],
3529 tmp ? pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 0,
3530 PL_colors[0], PL_colors[1],
3531 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0)|PERL_PV_ESCAPE_NONASCII
3533 : "not compiled under -Dr",
3537 locinput = PL_reginput;
3538 nextchr = UCHARAT(locinput);
3539 continue; /* execute rest of RE */
3544 char *s = STRING(scan);
3546 if (utf8_target != UTF_PATTERN) {
3547 /* The target and the pattern have differing utf8ness. */
3549 const char * const e = s + ln;
3552 /* The target is utf8, the pattern is not utf8. */
3557 if (NATIVE_TO_UNI(*(U8*)s) !=
3558 utf8n_to_uvuni((U8*)l, UTF8_MAXBYTES, &ulen,
3566 /* The target is not utf8, the pattern is utf8. */
3571 if (NATIVE_TO_UNI(*((U8*)l)) !=
3572 utf8n_to_uvuni((U8*)s, UTF8_MAXBYTES, &ulen,
3580 nextchr = UCHARAT(locinput);
3583 /* The target and the pattern have the same utf8ness. */
3584 /* Inline the first character, for speed. */
3585 if (UCHARAT(s) != nextchr)
3587 if (PL_regeol - locinput < ln)
3589 if (ln > 1 && memNE(s, locinput, ln))
3592 nextchr = UCHARAT(locinput);
3597 const U8 * fold_array;
3599 U32 fold_utf8_flags;
3601 PL_reg_flags |= RF_tainted;
3602 folder = foldEQ_locale;
3603 fold_array = PL_fold_locale;
3604 fold_utf8_flags = FOLDEQ_UTF8_LOCALE;
3608 folder = foldEQ_latin1;
3609 fold_array = PL_fold_latin1;
3610 fold_utf8_flags = 0;
3614 folder = foldEQ_latin1;
3615 fold_array = PL_fold_latin1;
3616 fold_utf8_flags = FOLDEQ_UTF8_NOMIX_ASCII;
3621 fold_array = PL_fold;
3622 fold_utf8_flags = 0;
3628 if (utf8_target || UTF_PATTERN) {
3629 /* Either target or the pattern are utf8. */
3630 const char * const l = locinput;
3631 char *e = PL_regeol;
3633 if (! foldEQ_utf8_flags(s, 0, ln, cBOOL(UTF_PATTERN),
3634 l, &e, 0, utf8_target, fold_utf8_flags))
3639 nextchr = UCHARAT(locinput);
3643 /* Neither the target nor the pattern are utf8 */
3644 if (UCHARAT(s) != nextchr &&
3645 UCHARAT(s) != fold_array[nextchr])
3649 if (PL_regeol - locinput < ln)
3651 if (ln > 1 && ! folder(s, locinput, ln))
3654 nextchr = UCHARAT(locinput);
3658 /* XXX Could improve efficiency by separating these all out using a
3659 * macro or in-line function. At that point regcomp.c would no longer
3660 * have to set the FLAGS fields of these */
3663 PL_reg_flags |= RF_tainted;
3671 /* was last char in word? */
3673 && FLAGS(scan) != REGEX_ASCII_RESTRICTED_CHARSET
3674 && FLAGS(scan) != REGEX_ASCII_MORE_RESTRICTED_CHARSET)
3676 if (locinput == PL_bostr)
3679 const U8 * const r = reghop3((U8*)locinput, -1, (U8*)PL_bostr);
3681 ln = utf8n_to_uvchr(r, UTF8SKIP(r), 0, uniflags);
3683 if (FLAGS(scan) != REGEX_LOCALE_CHARSET) {
3684 ln = isALNUM_uni(ln);
3685 LOAD_UTF8_CHARCLASS_ALNUM();
3686 n = swash_fetch(PL_utf8_alnum, (U8*)locinput, utf8_target);
3689 ln = isALNUM_LC_uvchr(UNI_TO_NATIVE(ln));
3690 n = isALNUM_LC_utf8((U8*)locinput);
3695 /* Here the string isn't utf8, or is utf8 and only ascii
3696 * characters are to match \w. In the latter case looking at
3697 * the byte just prior to the current one may be just the final
3698 * byte of a multi-byte character. This is ok. There are two
3700 * 1) it is a single byte character, and then the test is doing
3701 * just what it's supposed to.
3702 * 2) it is a multi-byte character, in which case the final
3703 * byte is never mistakable for ASCII, and so the test
3704 * will say it is not a word character, which is the
3705 * correct answer. */
3706 ln = (locinput != PL_bostr) ?
3707 UCHARAT(locinput - 1) : '\n';
3708 switch (FLAGS(scan)) {
3709 case REGEX_UNICODE_CHARSET:
3710 ln = isWORDCHAR_L1(ln);
3711 n = isWORDCHAR_L1(nextchr);
3713 case REGEX_LOCALE_CHARSET:
3714 ln = isALNUM_LC(ln);
3715 n = isALNUM_LC(nextchr);
3717 case REGEX_DEPENDS_CHARSET:
3719 n = isALNUM(nextchr);
3721 case REGEX_ASCII_RESTRICTED_CHARSET:
3722 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
3723 ln = isWORDCHAR_A(ln);
3724 n = isWORDCHAR_A(nextchr);
3727 Perl_croak(aTHX_ "panic: Unexpected FLAGS %u in op %u", FLAGS(scan), OP(scan));
3731 /* Note requires that all BOUNDs be lower than all NBOUNDs in
3733 if (((!ln) == (!n)) == (OP(scan) < NBOUND))
3738 if (utf8_target || state_num == ANYOFV) {
3739 STRLEN inclasslen = PL_regeol - locinput;
3740 if (locinput >= PL_regeol)
3743 if (!reginclass(rex, scan, (U8*)locinput, &inclasslen, utf8_target))
3745 locinput += inclasslen;
3746 nextchr = UCHARAT(locinput);
3751 nextchr = UCHARAT(locinput);
3752 if (!nextchr && locinput >= PL_regeol)
3754 if (!REGINCLASS(rex, scan, (U8*)locinput))
3756 nextchr = UCHARAT(++locinput);
3760 /* Special char classes - The defines start on line 129 or so */
3761 CCC_TRY_U(ALNUM, NALNUM, isWORDCHAR,
3762 ALNUML, NALNUML, isALNUM_LC, isALNUM_LC_utf8,
3763 ALNUMU, NALNUMU, isWORDCHAR_L1,
3764 ALNUMA, NALNUMA, isWORDCHAR_A,
3767 CCC_TRY_U(SPACE, NSPACE, isSPACE,
3768 SPACEL, NSPACEL, isSPACE_LC, isSPACE_LC_utf8,
3769 SPACEU, NSPACEU, isSPACE_L1,
3770 SPACEA, NSPACEA, isSPACE_A,
3773 CCC_TRY(DIGIT, NDIGIT, isDIGIT,
3774 DIGITL, NDIGITL, isDIGIT_LC, isDIGIT_LC_utf8,
3775 DIGITA, NDIGITA, isDIGIT_A,
3778 case CLUMP: /* Match \X: logical Unicode character. This is defined as
3779 a Unicode extended Grapheme Cluster */
3780 /* From http://www.unicode.org/reports/tr29 (5.2 version). An
3781 extended Grapheme Cluster is:
3784 | Prepend* Begin Extend*
3787 Begin is (Hangul-syllable | ! Control)
3788 Extend is (Grapheme_Extend | Spacing_Mark)
3789 Control is [ GCB_Control CR LF ]
3791 The discussion below shows how the code for CLUMP is derived
3792 from this regex. Note that most of these concepts are from
3793 property values of the Grapheme Cluster Boundary (GCB) property.
3794 No code point can have multiple property values for a given
3795 property. Thus a code point in Prepend can't be in Control, but
3796 it must be in !Control. This is why Control above includes
3797 GCB_Control plus CR plus LF. The latter two are used in the GCB
3798 property separately, and so can't be in GCB_Control, even though
3799 they logically are controls. Control is not the same as gc=cc,
3800 but includes format and other characters as well.
3802 The Unicode definition of Hangul-syllable is:
3804 | (L* ( ( V | LV ) V* | LVT ) T*)
3807 Each of these is a value for the GCB property, and hence must be
3808 disjoint, so the order they are tested is immaterial, so the
3809 above can safely be changed to
3812 | (L* ( LVT | ( V | LV ) V*) T*)
3814 The last two terms can be combined like this:
3816 | (( LVT | ( V | LV ) V*) T*))
3818 And refactored into this:
3819 L* (L | LVT T* | V V* T* | LV V* T*)
3821 That means that if we have seen any L's at all we can quit
3822 there, but if the next character is a LVT, a V or and LV we
3825 There is a subtlety with Prepend* which showed up in testing.
3826 Note that the Begin, and only the Begin is required in:
3827 | Prepend* Begin Extend*
3828 Also, Begin contains '! Control'. A Prepend must be a '!
3829 Control', which means it must be a Begin. What it comes down to
3830 is that if we match Prepend* and then find no suitable Begin
3831 afterwards, that if we backtrack the last Prepend, that one will
3832 be a suitable Begin.
3835 if (locinput >= PL_regeol)
3837 if (! utf8_target) {
3839 /* Match either CR LF or '.', as all the other possibilities
3841 locinput++; /* Match the . or CR */
3843 && locinput < PL_regeol
3844 && UCHARAT(locinput) == '\n') locinput++;
3848 /* Utf8: See if is ( CR LF ); already know that locinput <
3849 * PL_regeol, so locinput+1 is in bounds */
3850 if (nextchr == '\r' && UCHARAT(locinput + 1) == '\n') {
3854 /* In case have to backtrack to beginning, then match '.' */
3855 char *starting = locinput;
3857 /* In case have to backtrack the last prepend */
3858 char *previous_prepend = 0;
3860 LOAD_UTF8_CHARCLASS_GCB();
3862 /* Match (prepend)* */
3863 while (locinput < PL_regeol
3864 && swash_fetch(PL_utf8_X_prepend,
3865 (U8*)locinput, utf8_target))
3867 previous_prepend = locinput;
3868 locinput += UTF8SKIP(locinput);
3871 /* As noted above, if we matched a prepend character, but
3872 * the next thing won't match, back off the last prepend we
3873 * matched, as it is guaranteed to match the begin */
3874 if (previous_prepend
3875 && (locinput >= PL_regeol
3876 || ! swash_fetch(PL_utf8_X_begin,
3877 (U8*)locinput, utf8_target)))
3879 locinput = previous_prepend;
3882 /* Note that here we know PL_regeol > locinput, as we
3883 * tested that upon input to this switch case, and if we
3884 * moved locinput forward, we tested the result just above
3885 * and it either passed, or we backed off so that it will
3887 if (! swash_fetch(PL_utf8_X_begin, (U8*)locinput, utf8_target)) {
3889 /* Here did not match the required 'Begin' in the
3890 * second term. So just match the very first
3891 * character, the '.' of the final term of the regex */
3892 locinput = starting + UTF8SKIP(starting);
3895 /* Here is the beginning of a character that can have
3896 * an extender. It is either a hangul syllable, or a
3898 if (swash_fetch(PL_utf8_X_non_hangul,
3899 (U8*)locinput, utf8_target))
3902 /* Here not a Hangul syllable, must be a
3903 * ('! * Control') */
3904 locinput += UTF8SKIP(locinput);
3907 /* Here is a Hangul syllable. It can be composed
3908 * of several individual characters. One
3909 * possibility is T+ */
3910 if (swash_fetch(PL_utf8_X_T,
3911 (U8*)locinput, utf8_target))
3913 while (locinput < PL_regeol
3914 && swash_fetch(PL_utf8_X_T,
3915 (U8*)locinput, utf8_target))
3917 locinput += UTF8SKIP(locinput);
3921 /* Here, not T+, but is a Hangul. That means
3922 * it is one of the others: L, LV, LVT or V,
3924 * L* (L | LVT T* | V V* T* | LV V* T*) */
3927 while (locinput < PL_regeol
3928 && swash_fetch(PL_utf8_X_L,
3929 (U8*)locinput, utf8_target))
3931 locinput += UTF8SKIP(locinput);
3934 /* Here, have exhausted L*. If the next
3935 * character is not an LV, LVT nor V, it means
3936 * we had to have at least one L, so matches L+
3937 * in the original equation, we have a complete
3938 * hangul syllable. Are done. */
3940 if (locinput < PL_regeol
3941 && swash_fetch(PL_utf8_X_LV_LVT_V,
3942 (U8*)locinput, utf8_target))
3945 /* Otherwise keep going. Must be LV, LVT
3946 * or V. See if LVT */
3947 if (swash_fetch(PL_utf8_X_LVT,
3948 (U8*)locinput, utf8_target))
3950 locinput += UTF8SKIP(locinput);
3953 /* Must be V or LV. Take it, then
3955 locinput += UTF8SKIP(locinput);
3956 while (locinput < PL_regeol
3957 && swash_fetch(PL_utf8_X_V,
3958 (U8*)locinput, utf8_target))
3960 locinput += UTF8SKIP(locinput);
3964 /* And any of LV, LVT, or V can be followed
3966 while (locinput < PL_regeol
3967 && swash_fetch(PL_utf8_X_T,
3971 locinput += UTF8SKIP(locinput);
3977 /* Match any extender */
3978 while (locinput < PL_regeol
3979 && swash_fetch(PL_utf8_X_extend,
3980 (U8*)locinput, utf8_target))
3982 locinput += UTF8SKIP(locinput);
3986 if (locinput > PL_regeol) sayNO;
3988 nextchr = UCHARAT(locinput);
3992 { /* The capture buffer cases. The ones beginning with N for the
3993 named buffers just convert to the equivalent numbered and
3994 pretend they were called as the corresponding numbered buffer
3996 /* don't initialize these in the declaration, it makes C++
4001 const U8 *fold_array;
4004 PL_reg_flags |= RF_tainted;
4005 folder = foldEQ_locale;
4006 fold_array = PL_fold_locale;
4008 utf8_fold_flags = FOLDEQ_UTF8_LOCALE;
4012 folder = foldEQ_latin1;
4013 fold_array = PL_fold_latin1;
4015 utf8_fold_flags = FOLDEQ_UTF8_NOMIX_ASCII;
4019 folder = foldEQ_latin1;
4020 fold_array = PL_fold_latin1;
4022 utf8_fold_flags = 0;
4027 fold_array = PL_fold;
4029 utf8_fold_flags = 0;
4036 utf8_fold_flags = 0;
4039 /* For the named back references, find the corresponding buffer
4041 n = reg_check_named_buff_matched(rex,scan);
4046 goto do_nref_ref_common;
4049 PL_reg_flags |= RF_tainted;
4050 folder = foldEQ_locale;
4051 fold_array = PL_fold_locale;
4052 utf8_fold_flags = FOLDEQ_UTF8_LOCALE;
4056 folder = foldEQ_latin1;
4057 fold_array = PL_fold_latin1;
4058 utf8_fold_flags = FOLDEQ_UTF8_NOMIX_ASCII;
4062 folder = foldEQ_latin1;
4063 fold_array = PL_fold_latin1;
4064 utf8_fold_flags = 0;
4069 fold_array = PL_fold;
4070 utf8_fold_flags = 0;
4076 utf8_fold_flags = 0;
4080 n = ARG(scan); /* which paren pair */
4083 ln = PL_regoffs[n].start;
4084 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
4085 if (*PL_reglastparen < n || ln == -1)
4086 sayNO; /* Do not match unless seen CLOSEn. */
4087 if (ln == PL_regoffs[n].end)
4091 if (type != REF /* REF can do byte comparison */
4092 && (utf8_target || type == REFFU))
4093 { /* XXX handle REFFL better */
4094 char * limit = PL_regeol;
4096 /* This call case insensitively compares the entire buffer
4097 * at s, with the current input starting at locinput, but
4098 * not going off the end given by PL_regeol, and returns in
4099 * limit upon success, how much of the current input was
4101 if (! foldEQ_utf8_flags(s, NULL, PL_regoffs[n].end - ln, utf8_target,
4102 locinput, &limit, 0, utf8_target, utf8_fold_flags))
4107 nextchr = UCHARAT(locinput);
4111 /* Not utf8: Inline the first character, for speed. */
4112 if (UCHARAT(s) != nextchr &&
4114 UCHARAT(s) != fold_array[nextchr]))
4116 ln = PL_regoffs[n].end - ln;
4117 if (locinput + ln > PL_regeol)
4119 if (ln > 1 && (type == REF
4120 ? memNE(s, locinput, ln)
4121 : ! folder(s, locinput, ln)))
4124 nextchr = UCHARAT(locinput);
4134 #define ST st->u.eval
4139 regexp_internal *rei;
4140 regnode *startpoint;
4143 case GOSUB: /* /(...(?1))/ /(...(?&foo))/ */
4144 if (cur_eval && cur_eval->locinput==locinput) {
4145 if (cur_eval->u.eval.close_paren == (U32)ARG(scan))
4146 Perl_croak(aTHX_ "Infinite recursion in regex");
4147 if ( ++nochange_depth > max_nochange_depth )
4149 "Pattern subroutine nesting without pos change"
4150 " exceeded limit in regex");
4157 (void)ReREFCNT_inc(rex_sv);
4158 if (OP(scan)==GOSUB) {
4159 startpoint = scan + ARG2L(scan);
4160 ST.close_paren = ARG(scan);
4162 startpoint = rei->program+1;
4165 goto eval_recurse_doit;
4167 case EVAL: /* /(?{A})B/ /(??{A})B/ and /(?(?{A})X|Y)B/ */
4168 if (cur_eval && cur_eval->locinput==locinput) {
4169 if ( ++nochange_depth > max_nochange_depth )
4170 Perl_croak(aTHX_ "EVAL without pos change exceeded limit in regex");
4175 /* execute the code in the {...} */
4177 SV ** const before = SP;
4178 OP_4tree * const oop = PL_op;
4179 COP * const ocurcop = PL_curcop;
4181 char *saved_regeol = PL_regeol;
4182 struct re_save_state saved_state;
4184 /* To not corrupt the existing regex state while executing the
4185 * eval we would normally put it on the save stack, like with
4186 * save_re_context. However, re-evals have a weird scoping so we
4187 * can't just add ENTER/LEAVE here. With that, things like
4189 * (?{$a=2})(a(?{local$a=$a+1}))*aak*c(?{$b=$a})
4191 * would break, as they expect the localisation to be unwound
4192 * only when the re-engine backtracks through the bit that
4195 * What we do instead is just saving the state in a local c
4198 Copy(&PL_reg_state, &saved_state, 1, struct re_save_state);
4201 PL_op = (OP_4tree*)rexi->data->data[n];
4202 DEBUG_STATE_r( PerlIO_printf(Perl_debug_log,
4203 " re_eval 0x%"UVxf"\n", PTR2UV(PL_op)) );
4204 PAD_SAVE_LOCAL(old_comppad, (PAD*)rexi->data->data[n + 2]);
4205 PL_regoffs[0].end = PL_reg_magic->mg_len = locinput - PL_bostr;
4208 SV *sv_mrk = get_sv("REGMARK", 1);
4209 sv_setsv(sv_mrk, sv_yes_mark);
4212 CALLRUNOPS(aTHX); /* Scalar context. */
4215 ret = &PL_sv_undef; /* protect against empty (?{}) blocks. */
4221 Copy(&saved_state, &PL_reg_state, 1, struct re_save_state);
4224 PAD_RESTORE_LOCAL(old_comppad);
4225 PL_curcop = ocurcop;
4226 PL_regeol = saved_regeol;
4229 sv_setsv(save_scalar(PL_replgv), ret);
4233 if (logical == 2) { /* Postponed subexpression: /(??{...})/ */
4236 /* extract RE object from returned value; compiling if
4242 SV *const sv = SvRV(ret);
4244 if (SvTYPE(sv) == SVt_REGEXP) {
4246 } else if (SvSMAGICAL(sv)) {
4247 mg = mg_find(sv, PERL_MAGIC_qr);
4250 } else if (SvTYPE(ret) == SVt_REGEXP) {
4252 } else if (SvSMAGICAL(ret)) {
4253 if (SvGMAGICAL(ret)) {
4254 /* I don't believe that there is ever qr magic
4256 assert(!mg_find(ret, PERL_MAGIC_qr));
4257 sv_unmagic(ret, PERL_MAGIC_qr);
4260 mg = mg_find(ret, PERL_MAGIC_qr);
4261 /* testing suggests mg only ends up non-NULL for
4262 scalars who were upgraded and compiled in the
4263 else block below. In turn, this is only
4264 triggered in the "postponed utf8 string" tests
4270 rx = (REGEXP *) mg->mg_obj; /*XXX:dmq*/
4274 rx = reg_temp_copy(NULL, rx);
4278 const I32 osize = PL_regsize;
4281 assert (SvUTF8(ret));
4282 } else if (SvUTF8(ret)) {
4283 /* Not doing UTF-8, despite what the SV says. Is
4284 this only if we're trapped in use 'bytes'? */
4285 /* Make a copy of the octet sequence, but without
4286 the flag on, as the compiler now honours the
4287 SvUTF8 flag on ret. */
4289 const char *const p = SvPV(ret, len);
4290 ret = newSVpvn_flags(p, len, SVs_TEMP);
4292 rx = CALLREGCOMP(ret, pm_flags);
4294 & (SVs_TEMP | SVs_PADTMP | SVf_READONLY
4296 /* This isn't a first class regexp. Instead, it's
4297 caching a regexp onto an existing, Perl visible
4299 sv_magic(ret, MUTABLE_SV(rx), PERL_MAGIC_qr, 0, 0);
4304 re = (struct regexp *)SvANY(rx);
4306 RXp_MATCH_COPIED_off(re);
4307 re->subbeg = rex->subbeg;
4308 re->sublen = rex->sublen;
4311 debug_start_match(re_sv, utf8_target, locinput, PL_regeol,
4312 "Matching embedded");
4314 startpoint = rei->program + 1;
4315 ST.close_paren = 0; /* only used for GOSUB */
4316 /* borrowed from regtry */
4317 if (PL_reg_start_tmpl <= re->nparens) {
4318 PL_reg_start_tmpl = re->nparens*3/2 + 3;
4319 if(PL_reg_start_tmp)
4320 Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
4322 Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
4325 eval_recurse_doit: /* Share code with GOSUB below this line */
4326 /* run the pattern returned from (??{...}) */
4327 ST.cp = regcppush(0); /* Save *all* the positions. */
4328 REGCP_SET(ST.lastcp);
4330 PL_regoffs = re->offs; /* essentially NOOP on GOSUB */
4332 /* see regtry, specifically PL_reglast(?:close)?paren is a pointer! (i dont know why) :dmq */
4333 PL_reglastparen = &re->lastparen;
4334 PL_reglastcloseparen = &re->lastcloseparen;
4336 re->lastcloseparen = 0;
4338 PL_reginput = locinput;
4341 /* XXXX This is too dramatic a measure... */
4344 ST.toggle_reg_flags = PL_reg_flags;
4346 PL_reg_flags |= RF_utf8;
4348 PL_reg_flags &= ~RF_utf8;
4349 ST.toggle_reg_flags ^= PL_reg_flags; /* diff of old and new */
4351 ST.prev_rex = rex_sv;
4352 ST.prev_curlyx = cur_curlyx;
4353 SETREX(rex_sv,re_sv);
4358 ST.prev_eval = cur_eval;
4360 /* now continue from first node in postoned RE */
4361 PUSH_YES_STATE_GOTO(EVAL_AB, startpoint);
4364 /* logical is 1, /(?(?{...})X|Y)/ */
4365 sw = cBOOL(SvTRUE(ret));
4370 case EVAL_AB: /* cleanup after a successful (??{A})B */
4371 /* note: this is called twice; first after popping B, then A */
4372 PL_reg_flags ^= ST.toggle_reg_flags;
4373 ReREFCNT_dec(rex_sv);
4374 SETREX(rex_sv,ST.prev_rex);
4375 rex = (struct regexp *)SvANY(rex_sv);
4376 rexi = RXi_GET(rex);
4378 cur_eval = ST.prev_eval;
4379 cur_curlyx = ST.prev_curlyx;
4381 /* rex was changed so update the pointer in PL_reglastparen and PL_reglastcloseparen */
4382 PL_reglastparen = &rex->lastparen;
4383 PL_reglastcloseparen = &rex->lastcloseparen;
4384 /* also update PL_regoffs */
4385 PL_regoffs = rex->offs;
4387 /* XXXX This is too dramatic a measure... */
4389 if ( nochange_depth )
4394 case EVAL_AB_fail: /* unsuccessfully ran A or B in (??{A})B */
4395 /* note: this is called twice; first after popping B, then A */
4396 PL_reg_flags ^= ST.toggle_reg_flags;
4397 ReREFCNT_dec(rex_sv);
4398 SETREX(rex_sv,ST.prev_rex);
4399 rex = (struct regexp *)SvANY(rex_sv);
4400 rexi = RXi_GET(rex);
4401 /* rex was changed so update the pointer in PL_reglastparen and PL_reglastcloseparen */
4402 PL_reglastparen = &rex->lastparen;
4403 PL_reglastcloseparen = &rex->lastcloseparen;
4405 PL_reginput = locinput;
4406 REGCP_UNWIND(ST.lastcp);
4408 cur_eval = ST.prev_eval;
4409 cur_curlyx = ST.prev_curlyx;
4410 /* XXXX This is too dramatic a measure... */
4412 if ( nochange_depth )
4418 n = ARG(scan); /* which paren pair */
4419 PL_reg_start_tmp[n] = locinput;
4425 n = ARG(scan); /* which paren pair */
4426 PL_regoffs[n].start = PL_reg_start_tmp[n] - PL_bostr;
4427 PL_regoffs[n].end = locinput - PL_bostr;
4428 /*if (n > PL_regsize)
4430 if (n > *PL_reglastparen)
4431 *PL_reglastparen = n;
4432 *PL_reglastcloseparen = n;
4433 if (cur_eval && cur_eval->u.eval.close_paren == n) {
4441 cursor && OP(cursor)!=END;
4442 cursor=regnext(cursor))
4444 if ( OP(cursor)==CLOSE ){
4446 if ( n <= lastopen ) {
4448 = PL_reg_start_tmp[n] - PL_bostr;
4449 PL_regoffs[n].end = locinput - PL_bostr;
4450 /*if (n > PL_regsize)
4452 if (n > *PL_reglastparen)
4453 *PL_reglastparen = n;
4454 *PL_reglastcloseparen = n;
4455 if ( n == ARG(scan) || (cur_eval &&
4456 cur_eval->u.eval.close_paren == n))
4465 n = ARG(scan); /* which paren pair */
4466 sw = cBOOL(*PL_reglastparen >= n && PL_regoffs[n].end != -1);
4469 /* reg_check_named_buff_matched returns 0 for no match */
4470 sw = cBOOL(0 < reg_check_named_buff_matched(rex,scan));
4474 sw = (cur_eval && (!n || cur_eval->u.eval.close_paren == n));
4480 PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
4482 next = NEXTOPER(NEXTOPER(scan));
4484 next = scan + ARG(scan);
4485 if (OP(next) == IFTHEN) /* Fake one. */
4486 next = NEXTOPER(NEXTOPER(next));
4490 logical = scan->flags;
4493 /*******************************************************************
4495 The CURLYX/WHILEM pair of ops handle the most generic case of the /A*B/
4496 pattern, where A and B are subpatterns. (For simple A, CURLYM or
4497 STAR/PLUS/CURLY/CURLYN are used instead.)
4499 A*B is compiled as <CURLYX><A><WHILEM><B>
4501 On entry to the subpattern, CURLYX is called. This pushes a CURLYX
4502 state, which contains the current count, initialised to -1. It also sets
4503 cur_curlyx to point to this state, with any previous value saved in the
4506 CURLYX then jumps straight to the WHILEM op, rather than executing A,
4507 since the pattern may possibly match zero times (i.e. it's a while {} loop
4508 rather than a do {} while loop).
4510 Each entry to WHILEM represents a successful match of A. The count in the
4511 CURLYX block is incremented, another WHILEM state is pushed, and execution
4512 passes to A or B depending on greediness and the current count.
4514 For example, if matching against the string a1a2a3b (where the aN are
4515 substrings that match /A/), then the match progresses as follows: (the
4516 pushed states are interspersed with the bits of strings matched so far):
4519 <CURLYX cnt=0><WHILEM>
4520 <CURLYX cnt=1><WHILEM> a1 <WHILEM>
4521 <CURLYX cnt=2><WHILEM> a1 <WHILEM> a2 <WHILEM>
4522 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM>
4523 <CURLYX cnt=3><WHILEM> a1 <WHILEM> a2 <WHILEM> a3 <WHILEM> b
4525 (Contrast this with something like CURLYM, which maintains only a single
4529 a1 <CURLYM cnt=1> a2
4530 a1 a2 <CURLYM cnt=2> a3
4531 a1 a2 a3 <CURLYM cnt=3> b
4534 Each WHILEM state block marks a point to backtrack to upon partial failure
4535 of A or B, and also contains some minor state data related to that
4536 iteration. The CURLYX block, pointed to by cur_curlyx, contains the
4537 overall state, such as the count, and pointers to the A and B ops.
4539 This is complicated slightly by nested CURLYX/WHILEM's. Since cur_curlyx
4540 must always point to the *current* CURLYX block, the rules are:
4542 When executing CURLYX, save the old cur_curlyx in the CURLYX state block,
4543 and set cur_curlyx to point the new block.
4545 When popping the CURLYX block after a successful or unsuccessful match,
4546 restore the previous cur_curlyx.
4548 When WHILEM is about to execute B, save the current cur_curlyx, and set it
4549 to the outer one saved in the CURLYX block.
4551 When popping the WHILEM block after a successful or unsuccessful B match,
4552 restore the previous cur_curlyx.
4554 Here's an example for the pattern (AI* BI)*BO
4555 I and O refer to inner and outer, C and W refer to CURLYX and WHILEM:
4558 curlyx backtrack stack
4559 ------ ---------------
4561 CO <CO prev=NULL> <WO>
4562 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
4563 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
4564 NULL <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi <WO prev=CO> bo
4566 At this point the pattern succeeds, and we work back down the stack to
4567 clean up, restoring as we go:
4569 CO <CO prev=NULL> <WO> <CI prev=CO> <WI> ai <WI prev=CI> bi
4570 CI <CO prev=NULL> <WO> <CI prev=CO> <WI> ai
4571 CO <CO prev=NULL> <WO>
4574 *******************************************************************/
4576 #define ST st->u.curlyx
4578 case CURLYX: /* start of /A*B/ (for complex A) */
4580 /* No need to save/restore up to this paren */
4581 I32 parenfloor = scan->flags;
4583 assert(next); /* keep Coverity happy */
4584 if (OP(PREVOPER(next)) == NOTHING) /* LONGJMP */
4587 /* XXXX Probably it is better to teach regpush to support
4588 parenfloor > PL_regsize... */
4589 if (parenfloor > (I32)*PL_reglastparen)
4590 parenfloor = *PL_reglastparen; /* Pessimization... */
4592 ST.prev_curlyx= cur_curlyx;
4594 ST.cp = PL_savestack_ix;
4596 /* these fields contain the state of the current curly.
4597 * they are accessed by subsequent WHILEMs */
4598 ST.parenfloor = parenfloor;
4603 ST.count = -1; /* this will be updated by WHILEM */
4604 ST.lastloc = NULL; /* this will be updated by WHILEM */
4606 PL_reginput = locinput;
4607 PUSH_YES_STATE_GOTO(CURLYX_end, PREVOPER(next));
4611 case CURLYX_end: /* just finished matching all of A*B */
4612 cur_curlyx = ST.prev_curlyx;
4616 case CURLYX_end_fail: /* just failed to match all of A*B */
4618 cur_curlyx = ST.prev_curlyx;
4624 #define ST st->u.whilem
4626 case WHILEM: /* just matched an A in /A*B/ (for complex A) */
4628 /* see the discussion above about CURLYX/WHILEM */
4630 int min = ARG1(cur_curlyx->u.curlyx.me);
4631 int max = ARG2(cur_curlyx->u.curlyx.me);
4632 regnode *A = NEXTOPER(cur_curlyx->u.curlyx.me) + EXTRA_STEP_2ARGS;
4634 assert(cur_curlyx); /* keep Coverity happy */
4635 n = ++cur_curlyx->u.curlyx.count; /* how many A's matched */
4636 ST.save_lastloc = cur_curlyx->u.curlyx.lastloc;
4637 ST.cache_offset = 0;
4640 PL_reginput = locinput;
4642 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4643 "%*s whilem: matched %ld out of %d..%d\n",
4644 REPORT_CODE_OFF+depth*2, "", (long)n, min, max)
4647 /* First just match a string of min A's. */
4650 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4651 cur_curlyx->u.curlyx.lastloc = locinput;
4652 REGCP_SET(ST.lastcp);
4654 PUSH_STATE_GOTO(WHILEM_A_pre, A);
4658 /* If degenerate A matches "", assume A done. */
4660 if (locinput == cur_curlyx->u.curlyx.lastloc) {
4661 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4662 "%*s whilem: empty match detected, trying continuation...\n",
4663 REPORT_CODE_OFF+depth*2, "")
4665 goto do_whilem_B_max;
4668 /* super-linear cache processing */
4672 if (!PL_reg_maxiter) {
4673 /* start the countdown: Postpone detection until we
4674 * know the match is not *that* much linear. */
4675 PL_reg_maxiter = (PL_regeol - PL_bostr + 1) * (scan->flags>>4);
4676 /* possible overflow for long strings and many CURLYX's */
4677 if (PL_reg_maxiter < 0)
4678 PL_reg_maxiter = I32_MAX;
4679 PL_reg_leftiter = PL_reg_maxiter;
4682 if (PL_reg_leftiter-- == 0) {
4683 /* initialise cache */
4684 const I32 size = (PL_reg_maxiter + 7)/8;
4685 if (PL_reg_poscache) {
4686 if ((I32)PL_reg_poscache_size < size) {
4687 Renew(PL_reg_poscache, size, char);
4688 PL_reg_poscache_size = size;
4690 Zero(PL_reg_poscache, size, char);
4693 PL_reg_poscache_size = size;
4694 Newxz(PL_reg_poscache, size, char);
4696 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4697 "%swhilem: Detected a super-linear match, switching on caching%s...\n",
4698 PL_colors[4], PL_colors[5])
4702 if (PL_reg_leftiter < 0) {
4703 /* have we already failed at this position? */
4705 offset = (scan->flags & 0xf) - 1
4706 + (locinput - PL_bostr) * (scan->flags>>4);
4707 mask = 1 << (offset % 8);
4709 if (PL_reg_poscache[offset] & mask) {
4710 DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log,
4711 "%*s whilem: (cache) already tried at this position...\n",
4712 REPORT_CODE_OFF+depth*2, "")
4714 sayNO; /* cache records failure */
4716 ST.cache_offset = offset;
4717 ST.cache_mask = mask;
4721 /* Prefer B over A for minimal matching. */
4723 if (cur_curlyx->u.curlyx.minmod) {
4724 ST.save_curlyx = cur_curlyx;
4725 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4726 ST.cp = regcppush(ST.save_curlyx->u.curlyx.parenfloor);
4727 REGCP_SET(ST.lastcp);
4728 PUSH_YES_STATE_GOTO(WHILEM_B_min, ST.save_curlyx->u.curlyx.B);
4732 /* Prefer A over B for maximal matching. */
4734 if (n < max) { /* More greed allowed? */
4735 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4736 cur_curlyx->u.curlyx.lastloc = locinput;
4737 REGCP_SET(ST.lastcp);
4738 PUSH_STATE_GOTO(WHILEM_A_max, A);
4741 goto do_whilem_B_max;
4745 case WHILEM_B_min: /* just matched B in a minimal match */
4746 case WHILEM_B_max: /* just matched B in a maximal match */
4747 cur_curlyx = ST.save_curlyx;
4751 case WHILEM_B_max_fail: /* just failed to match B in a maximal match */
4752 cur_curlyx = ST.save_curlyx;
4753 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4754 cur_curlyx->u.curlyx.count--;
4758 case WHILEM_A_min_fail: /* just failed to match A in a minimal match */
4760 case WHILEM_A_pre_fail: /* just failed to match even minimal A */
4761 REGCP_UNWIND(ST.lastcp);
4763 cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
4764 cur_curlyx->u.curlyx.count--;
4768 case WHILEM_A_max_fail: /* just failed to match A in a maximal match */
4769 REGCP_UNWIND(ST.lastcp);
4770 regcppop(rex); /* Restore some previous $<digit>s? */
4771 PL_reginput = locinput;
4772 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4773 "%*s whilem: failed, trying continuation...\n",
4774 REPORT_CODE_OFF+depth*2, "")
4777 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4778 && ckWARN(WARN_REGEXP)
4779 && !(PL_reg_flags & RF_warned))
4781 PL_reg_flags |= RF_warned;
4782 Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s limit (%d) exceeded",
4783 "Complex regular subexpression recursion",
4788 ST.save_curlyx = cur_curlyx;
4789 cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
4790 PUSH_YES_STATE_GOTO(WHILEM_B_max, ST.save_curlyx->u.curlyx.B);
4793 case WHILEM_B_min_fail: /* just failed to match B in a minimal match */
4794 cur_curlyx = ST.save_curlyx;
4795 REGCP_UNWIND(ST.lastcp);
4798 if (cur_curlyx->u.curlyx.count >= /*max*/ARG2(cur_curlyx->u.curlyx.me)) {
4799 /* Maximum greed exceeded */
4800 if (cur_curlyx->u.curlyx.count >= REG_INFTY
4801 && ckWARN(WARN_REGEXP)
4802 && !(PL_reg_flags & RF_warned))
4804 PL_reg_flags |= RF_warned;
4805 Perl_warner(aTHX_ packWARN(WARN_REGEXP),
4806 "%s limit (%d) exceeded",
4807 "Complex regular subexpression recursion",
4810 cur_curlyx->u.curlyx.count--;
4814 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
4815 "%*s trying longer...\n", REPORT_CODE_OFF+depth*2, "")
4817 /* Try grabbing another A and see if it helps. */
4818 PL_reginput = locinput;
4819 cur_curlyx->u.curlyx.lastloc = locinput;
4820 ST.cp = regcppush(cur_curlyx->u.curlyx.parenfloor);
4821 REGCP_SET(ST.lastcp);
4822 PUSH_STATE_GOTO(WHILEM_A_min,
4823 /*A*/ NEXTOPER(ST.save_curlyx->u.curlyx.me) + EXTRA_STEP_2ARGS);
4827 #define ST st->u.branch
4829 case BRANCHJ: /* /(...|A|...)/ with long next pointer */
4830 next = scan + ARG(scan);
4833 scan = NEXTOPER(scan);
4836 case BRANCH: /* /(...|A|...)/ */
4837 scan = NEXTOPER(scan); /* scan now points to inner node */
4838 ST.lastparen = *PL_reglastparen;
4839 ST.next_branch = next;
4841 PL_reginput = locinput;
4843 /* Now go into the branch */
4845 PUSH_YES_STATE_GOTO(BRANCH_next, scan);
4847 PUSH_STATE_GOTO(BRANCH_next, scan);
4851 PL_reginput = locinput;
4852 sv_yes_mark = st->u.mark.mark_name = scan->flags ? NULL :
4853 MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
4854 PUSH_STATE_GOTO(CUTGROUP_next,next);
4856 case CUTGROUP_next_fail:
4859 if (st->u.mark.mark_name)
4860 sv_commit = st->u.mark.mark_name;
4866 case BRANCH_next_fail: /* that branch failed; try the next, if any */
4871 REGCP_UNWIND(ST.cp);
4872 for (n = *PL_reglastparen; n > ST.lastparen; n--)
4873 PL_regoffs[n].end = -1;
4874 *PL_reglastparen = n;
4875 /*dmq: *PL_reglastcloseparen = n; */
4876 scan = ST.next_branch;
4877 /* no more branches? */
4878 if (!scan || (OP(scan) != BRANCH && OP(scan) != BRANCHJ)) {
4880 PerlIO_printf( Perl_debug_log,
4881 "%*s %sBRANCH failed...%s\n",
4882 REPORT_CODE_OFF+depth*2, "",
4888 continue; /* execute next BRANCH[J] op */
4896 #define ST st->u.curlym
4898 case CURLYM: /* /A{m,n}B/ where A is fixed-length */
4900 /* This is an optimisation of CURLYX that enables us to push
4901 * only a single backtracking state, no matter how many matches
4902 * there are in {m,n}. It relies on the pattern being constant
4903 * length, with no parens to influence future backrefs
4907 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
4909 /* if paren positive, emulate an OPEN/CLOSE around A */
4911 U32 paren = ST.me->flags;
4912 if (paren > PL_regsize)
4914 if (paren > *PL_reglastparen)
4915 *PL_reglastparen = paren;
4916 scan += NEXT_OFF(scan); /* Skip former OPEN. */
4924 ST.c1 = CHRTEST_UNINIT;
4927 if (!(ST.minmod ? ARG1(ST.me) : ARG2(ST.me))) /* min/max */
4930 curlym_do_A: /* execute the A in /A{m,n}B/ */
4931 PL_reginput = locinput;
4932 PUSH_YES_STATE_GOTO(CURLYM_A, ST.A); /* match A */
4935 case CURLYM_A: /* we've just matched an A */
4936 locinput = st->locinput;
4937 nextchr = UCHARAT(locinput);
4940 /* after first match, determine A's length: u.curlym.alen */
4941 if (ST.count == 1) {
4942 if (PL_reg_match_utf8) {
4944 while (s < PL_reginput) {
4950 ST.alen = PL_reginput - locinput;
4953 ST.count = ST.minmod ? ARG1(ST.me) : ARG2(ST.me);
4956 PerlIO_printf(Perl_debug_log,
4957 "%*s CURLYM now matched %"IVdf" times, len=%"IVdf"...\n",
4958 (int)(REPORT_CODE_OFF+(depth*2)), "",
4959 (IV) ST.count, (IV)ST.alen)
4962 locinput = PL_reginput;
4964 if (cur_eval && cur_eval->u.eval.close_paren &&
4965 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
4969 I32 max = (ST.minmod ? ARG1(ST.me) : ARG2(ST.me));
4970 if ( max == REG_INFTY || ST.count < max )
4971 goto curlym_do_A; /* try to match another A */
4973 goto curlym_do_B; /* try to match B */
4975 case CURLYM_A_fail: /* just failed to match an A */
4976 REGCP_UNWIND(ST.cp);
4978 if (ST.minmod || ST.count < ARG1(ST.me) /* min*/
4979 || (cur_eval && cur_eval->u.eval.close_paren &&
4980 cur_eval->u.eval.close_paren == (U32)ST.me->flags))
4983 curlym_do_B: /* execute the B in /A{m,n}B/ */
4984 PL_reginput = locinput;
4985 if (ST.c1 == CHRTEST_UNINIT) {
4986 /* calculate c1 and c2 for possible match of 1st char
4987 * following curly */
4988 ST.c1 = ST.c2 = CHRTEST_VOID;
4989 if (HAS_TEXT(ST.B) || JUMPABLE(ST.B)) {
4990 regnode *text_node = ST.B;
4991 if (! HAS_TEXT(text_node))
4992 FIND_NEXT_IMPT(text_node);
4995 (HAS_TEXT(text_node) && PL_regkind[OP(text_node)] == EXACT)
4997 But the former is redundant in light of the latter.
4999 if this changes back then the macro for
5000 IS_TEXT and friends need to change.
5002 if (PL_regkind[OP(text_node)] == EXACT)
5005 ST.c1 = (U8)*STRING(text_node);
5006 switch (OP(text_node)) {
5007 case EXACTF: ST.c2 = PL_fold[ST.c1]; break;
5009 case EXACTFU: ST.c2 = PL_fold_latin1[ST.c1]; break;
5010 case EXACTFL: ST.c2 = PL_fold_locale[ST.c1]; break;
5011 default: ST.c2 = ST.c1;
5018 PerlIO_printf(Perl_debug_log,
5019 "%*s CURLYM trying tail with matches=%"IVdf"...\n",
5020 (int)(REPORT_CODE_OFF+(depth*2)),
5023 if (ST.c1 != CHRTEST_VOID
5024 && UCHARAT(PL_reginput) != ST.c1
5025 && UCHARAT(PL_reginput) != ST.c2)
5027 /* simulate B failing */
5029 PerlIO_printf(Perl_debug_log,
5030 "%*s CURLYM Fast bail c1=%"IVdf" c2=%"IVdf"\n",
5031 (int)(REPORT_CODE_OFF+(depth*2)),"",
5034 state_num = CURLYM_B_fail;
5035 goto reenter_switch;
5039 /* mark current A as captured */
5040 I32 paren = ST.me->flags;
5042 PL_regoffs[paren].start
5043 = HOPc(PL_reginput, -ST.alen) - PL_bostr;
5044 PL_regoffs[paren].end = PL_reginput - PL_bostr;
5045 /*dmq: *PL_reglastcloseparen = paren; */
5048 PL_regoffs[paren].end = -1;
5049 if (cur_eval && cur_eval->u.eval.close_paren &&
5050 cur_eval->u.eval.close_paren == (U32)ST.me->flags)
5059 PUSH_STATE_GOTO(CURLYM_B, ST.B); /* match B */
5062 case CURLYM_B_fail: /* just failed to match a B */
5063 REGCP_UNWIND(ST.cp);
5065 I32 max = ARG2(ST.me);
5066 if (max != REG_INFTY && ST.count == max)
5068 goto curlym_do_A; /* try to match a further A */
5070 /* backtrack one A */
5071 if (ST.count == ARG1(ST.me) /* min */)
5074 locinput = HOPc(locinput, -ST.alen);
5075 goto curlym_do_B; /* try to match B */
5078 #define ST st->u.curly
5080 #define CURLY_SETPAREN(paren, success) \
5083 PL_regoffs[paren].start = HOPc(locinput, -1) - PL_bostr; \
5084 PL_regoffs[paren].end = locinput - PL_bostr; \
5085 *PL_reglastcloseparen = paren; \
5088 PL_regoffs[paren].end = -1; \
5091 case STAR: /* /A*B/ where A is width 1 */
5095 scan = NEXTOPER(scan);
5097 case PLUS: /* /A+B/ where A is width 1 */
5101 scan = NEXTOPER(scan);
5103 case CURLYN: /* /(A){m,n}B/ where A is width 1 */
5104 ST.paren = scan->flags; /* Which paren to set */
5105 if (ST.paren > PL_regsize)
5106 PL_regsize = ST.paren;
5107 if (ST.paren > *PL_reglastparen)
5108 *PL_reglastparen = ST.paren;
5109 ST.min = ARG1(scan); /* min to match */
5110 ST.max = ARG2(scan); /* max to match */
5111 if (cur_eval && cur_eval->u.eval.close_paren &&
5112 cur_eval->u.eval.close_paren == (U32)ST.paren) {
5116 scan = regnext(NEXTOPER(scan) + NODE_STEP_REGNODE);
5118 case CURLY: /* /A{m,n}B/ where A is width 1 */
5120 ST.min = ARG1(scan); /* min to match */
5121 ST.max = ARG2(scan); /* max to match */
5122 scan = NEXTOPER(scan) + NODE_STEP_REGNODE;
5125 * Lookahead to avoid useless match attempts
5126 * when we know what character comes next.
5128 * Used to only do .*x and .*?x, but now it allows
5129 * for )'s, ('s and (?{ ... })'s to be in the way
5130 * of the quantifier and the EXACT-like node. -- japhy
5133 if (ST.min > ST.max) /* XXX make this a compile-time check? */
5135 if (HAS_TEXT(next) || JUMPABLE(next)) {
5137 regnode *text_node = next;
5139 if (! HAS_TEXT(text_node))
5140 FIND_NEXT_IMPT(text_node);
5142 if (! HAS_TEXT(text_node))
5143 ST.c1 = ST.c2 = CHRTEST_VOID;
5145 if ( PL_regkind[OP(text_node)] != EXACT ) {
5146 ST.c1 = ST.c2 = CHRTEST_VOID;
5147 goto assume_ok_easy;
5150 s = (U8*)STRING(text_node);
5152 /* Currently we only get here when
5154 PL_rekind[OP(text_node)] == EXACT
5156 if this changes back then the macro for IS_TEXT and
5157 friends need to change. */
5160 switch (OP(text_node)) {
5161 case EXACTF: ST.c2 = PL_fold[ST.c1]; break;
5163 case EXACTFU: ST.c2 = PL_fold_latin1[ST.c1]; break;
5164 case EXACTFL: ST.c2 = PL_fold_locale[ST.c1]; break;
5165 default: ST.c2 = ST.c1; break;
5168 else { /* UTF_PATTERN */
5169 if (IS_TEXTFU(text_node) || IS_TEXTF(text_node)) {
5170 STRLEN ulen1, ulen2;
5171 U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
5172 U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
5174 to_utf8_lower((U8*)s, tmpbuf1, &ulen1);
5175 to_utf8_upper((U8*)s, tmpbuf2, &ulen2);
5177 ST.c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXLEN, 0,
5179 0 : UTF8_ALLOW_ANY);
5180 ST.c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXLEN, 0,
5182 0 : UTF8_ALLOW_ANY);
5184 ST.c1 = utf8n_to_uvuni(tmpbuf1, UTF8_MAXBYTES, 0,
5186 ST.c2 = utf8n_to_uvuni(tmpbuf2, UTF8_MAXBYTES, 0,
5191 ST.c2 = ST.c1 = utf8n_to_uvchr(s, UTF8_MAXBYTES, 0,
5198 ST.c1 = ST.c2 = CHRTEST_VOID;
5203 PL_reginput = locinput;
5206 if (ST.min && regrepeat(rex, ST.A, ST.min, depth) < ST.min)
5209 locinput = PL_reginput;
5211 if (ST.c1 == CHRTEST_VOID)
5212 goto curly_try_B_min;
5214 ST.oldloc = locinput;
5216 /* set ST.maxpos to the furthest point along the
5217 * string that could possibly match */
5218 if (ST.max == REG_INFTY) {
5219 ST.maxpos = PL_regeol - 1;
5221 while (UTF8_IS_CONTINUATION(*(U8*)ST.maxpos))
5224 else if (utf8_target) {
5225 int m = ST.max - ST.min;
5226 for (ST.maxpos = locinput;
5227 m >0 && ST.maxpos + UTF8SKIP(ST.maxpos) <= PL_regeol; m--)
5228 ST.maxpos += UTF8SKIP(ST.maxpos);
5231 ST.maxpos = locinput + ST.max - ST.min;
5232 if (ST.maxpos >= PL_regeol)
5233 ST.maxpos = PL_regeol - 1;
5235 goto curly_try_B_min_known;
5239 ST.count = regrepeat(rex, ST.A, ST.max, depth);
5240 locinput = PL_reginput;
5241 if (ST.count < ST.min)
5243 if ((ST.count > ST.min)
5244 && (PL_regkind[OP(ST.B)] == EOL) && (OP(ST.B) != MEOL))
5246 /* A{m,n} must come at the end of the string, there's
5247 * no point in backing off ... */
5249 /* ...except that $ and \Z can match before *and* after
5250 newline at the end. Consider "\n\n" =~ /\n+\Z\n/.
5251 We may back off by one in this case. */
5252 if (UCHARAT(PL_reginput - 1) == '\n' && OP(ST.B) != EOS)
5256 goto curly_try_B_max;
5261 case CURLY_B_min_known_fail:
5262 /* failed to find B in a non-greedy match where c1,c2 valid */
5263 if (ST.paren && ST.count)
5264 PL_regoffs[ST.paren].end = -1;
5266 PL_reginput = locinput; /* Could be reset... */
5267 REGCP_UNWIND(ST.cp);
5268 /* Couldn't or didn't -- move forward. */
5269 ST.oldloc = locinput;
5271 locinput += UTF8SKIP(locinput);
5275 curly_try_B_min_known:
5276 /* find the next place where 'B' could work, then call B */
5280 n = (ST.oldloc == locinput) ? 0 : 1;
5281 if (ST.c1 == ST.c2) {
5283 /* set n to utf8_distance(oldloc, locinput) */
5284 while (locinput <= ST.maxpos &&
5285 utf8n_to_uvchr((U8*)locinput,
5286 UTF8_MAXBYTES, &len,
5287 uniflags) != (UV)ST.c1) {
5293 /* set n to utf8_distance(oldloc, locinput) */
5294 while (locinput <= ST.maxpos) {
5296 const UV c = utf8n_to_uvchr((U8*)locinput,
5297 UTF8_MAXBYTES, &len,
5299 if (c == (UV)ST.c1 || c == (UV)ST.c2)
5307 if (ST.c1 == ST.c2) {
5308 while (locinput <= ST.maxpos &&
5309 UCHARAT(locinput) != ST.c1)
5313 while (locinput <= ST.maxpos
5314 && UCHARAT(locinput) != ST.c1
5315 && UCHARAT(locinput) != ST.c2)
5318 n = locinput - ST.oldloc;
5320 if (locinput > ST.maxpos)
5322 /* PL_reginput == oldloc now */
5325 if (regrepeat(rex, ST.A, n, depth) < n)
5328 PL_reginput = locinput;
5329 CURLY_SETPAREN(ST.paren, ST.count);
5330 if (cur_eval && cur_eval->u.eval.close_paren &&
5331 cur_eval->u.eval.close_paren == (U32)ST.paren) {
5334 PUSH_STATE_GOTO(CURLY_B_min_known, ST.B);
5339 case CURLY_B_min_fail:
5340 /* failed to find B in a non-greedy match where c1,c2 invalid */
5341 if (ST.paren && ST.count)
5342 PL_regoffs[ST.paren].end = -1;
5344 REGCP_UNWIND(ST.cp);
5345 /* failed -- move forward one */
5346 PL_reginput = locinput;
5347 if (regrepeat(rex, ST.A, 1, depth)) {
5349 locinput = PL_reginput;
5350 if (ST.count <= ST.max || (ST.max == REG_INFTY &&
5351 ST.count > 0)) /* count overflow ? */
5354 CURLY_SETPAREN(ST.paren, ST.count);
5355 if (cur_eval && cur_eval->u.eval.close_paren &&
5356 cur_eval->u.eval.close_paren == (U32)ST.paren) {
5359 PUSH_STATE_GOTO(CURLY_B_min, ST.B);
5367 /* a successful greedy match: now try to match B */
5368 if (cur_eval && cur_eval->u.eval.close_paren &&
5369 cur_eval->u.eval.close_paren == (U32)ST.paren) {
5374 if (ST.c1 != CHRTEST_VOID)
5375 c = utf8_target ? utf8n_to_uvchr((U8*)PL_reginput,
5376 UTF8_MAXBYTES, 0, uniflags)
5377 : (UV) UCHARAT(PL_reginput);
5378 /* If it could work, try it. */
5379 if (ST.c1 == CHRTEST_VOID || c == (UV)ST.c1 || c == (UV)ST.c2) {
5380 CURLY_SETPAREN(ST.paren, ST.count);
5381 PUSH_STATE_GOTO(CURLY_B_max, ST.B);
5386 case CURLY_B_max_fail:
5387 /* failed to find B in a greedy match */
5388 if (ST.paren && ST.count)
5389 PL_regoffs[ST.paren].end = -1;
5391 REGCP_UNWIND(ST.cp);
5393 if (--ST.count < ST.min)
5395 PL_reginput = locinput = HOPc(locinput, -1);
5396 goto curly_try_B_max;
5403 /* we've just finished A in /(??{A})B/; now continue with B */
5405 st->u.eval.toggle_reg_flags
5406 = cur_eval->u.eval.toggle_reg_flags;
5407 PL_reg_flags ^= st->u.eval.toggle_reg_flags;
5409 st->u.eval.prev_rex = rex_sv; /* inner */
5410 SETREX(rex_sv,cur_eval->u.eval.prev_rex);
5411 rex = (struct regexp *)SvANY(rex_sv);
5412 rexi = RXi_GET(rex);
5413 cur_curlyx = cur_eval->u.eval.prev_curlyx;
5414 (void)ReREFCNT_inc(rex_sv);
5415 st->u.eval.cp = regcppush(0); /* Save *all* the positions. */
5417 /* rex was changed so update the pointer in PL_reglastparen and PL_reglastcloseparen */
5418 PL_reglastparen = &rex->lastparen;
5419 PL_reglastcloseparen = &rex->lastcloseparen;
5421 REGCP_SET(st->u.eval.lastcp);
5422 PL_reginput = locinput;
5424 /* Restore parens of the outer rex without popping the
5426 tmpix = PL_savestack_ix;
5427 PL_savestack_ix = cur_eval->u.eval.lastcp;
5429 PL_savestack_ix = tmpix;
5431 st->u.eval.prev_eval = cur_eval;
5432 cur_eval = cur_eval->u.eval.prev_eval;
5434 PerlIO_printf(Perl_debug_log, "%*s EVAL trying tail ... %"UVxf"\n",
5435 REPORT_CODE_OFF+depth*2, "",PTR2UV(cur_eval)););
5436 if ( nochange_depth )
5439 PUSH_YES_STATE_GOTO(EVAL_AB,
5440 st->u.eval.prev_eval->u.eval.B); /* match B */
5443 if (locinput < reginfo->till) {
5444 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
5445 "%sMatch possible, but length=%ld is smaller than requested=%ld, failing!%s\n",
5447 (long)(locinput - PL_reg_starttry),
5448 (long)(reginfo->till - PL_reg_starttry),
5451 sayNO_SILENT; /* Cannot match: too short. */
5453 PL_reginput = locinput; /* put where regtry can find it */
5454 sayYES; /* Success! */
5456 case SUCCEED: /* successful SUSPEND/UNLESSM/IFMATCH/CURLYM */
5458 PerlIO_printf(Perl_debug_log,
5459 "%*s %ssubpattern success...%s\n",
5460 REPORT_CODE_OFF+depth*2, "", PL_colors[4], PL_colors[5]));
5461 PL_reginput = locinput; /* put where regtry can find it */
5462 sayYES; /* Success! */
5465 #define ST st->u.ifmatch
5467 case SUSPEND: /* (?>A) */
5469 PL_reginput = locinput;
5472 case UNLESSM: /* -ve lookaround: (?!A), or with flags, (?<!A) */
5474 goto ifmatch_trivial_fail_test;
5476 case IFMATCH: /* +ve lookaround: (?=A), or with flags, (?<=A) */
5478 ifmatch_trivial_fail_test:
5480 char * const s = HOPBACKc(locinput, scan->flags);
5485 sw = 1 - cBOOL(ST.wanted);
5489 next = scan + ARG(scan);
5497 PL_reginput = locinput;
5501 ST.logical = logical;
5502 logical = 0; /* XXX: reset state of logical once it has been saved into ST */
5504 /* execute body of (?...A) */
5505 PUSH_YES_STATE_GOTO(IFMATCH_A, NEXTOPER(NEXTOPER(scan)));
5508 case IFMATCH_A_fail: /* body of (?...A) failed */
5509 ST.wanted = !ST.wanted;
5512 case IFMATCH_A: /* body of (?...A) succeeded */
5514 sw = cBOOL(ST.wanted);
5516 else if (!ST.wanted)
5519 if (OP(ST.me) == SUSPEND)
5520 locinput = PL_reginput;
5522 locinput = PL_reginput = st->locinput;
5523 nextchr = UCHARAT(locinput);
5525 scan = ST.me + ARG(ST.me);
5528 continue; /* execute B */
5533 next = scan + ARG(scan);
5538 reginfo->cutpoint = PL_regeol;
5541 PL_reginput = locinput;
5543 sv_yes_mark = sv_commit = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5544 PUSH_STATE_GOTO(COMMIT_next,next);
5546 case COMMIT_next_fail:
5553 #define ST st->u.mark
5555 ST.prev_mark = mark_state;
5556 ST.mark_name = sv_commit = sv_yes_mark
5557 = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5559 ST.mark_loc = PL_reginput = locinput;
5560 PUSH_YES_STATE_GOTO(MARKPOINT_next,next);
5562 case MARKPOINT_next:
5563 mark_state = ST.prev_mark;
5566 case MARKPOINT_next_fail:
5567 if (popmark && sv_eq(ST.mark_name,popmark))
5569 if (ST.mark_loc > startpoint)
5570 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
5571 popmark = NULL; /* we found our mark */
5572 sv_commit = ST.mark_name;
5575 PerlIO_printf(Perl_debug_log,
5576 "%*s %ssetting cutpoint to mark:%"SVf"...%s\n",
5577 REPORT_CODE_OFF+depth*2, "",
5578 PL_colors[4], SVfARG(sv_commit), PL_colors[5]);
5581 mark_state = ST.prev_mark;
5582 sv_yes_mark = mark_state ?
5583 mark_state->u.mark.mark_name : NULL;
5587 PL_reginput = locinput;
5589 /* (*SKIP) : if we fail we cut here*/
5590 ST.mark_name = NULL;
5591 ST.mark_loc = locinput;
5592 PUSH_STATE_GOTO(SKIP_next,next);
5594 /* (*SKIP:NAME) : if there is a (*MARK:NAME) fail where it was,
5595 otherwise do nothing. Meaning we need to scan
5597 regmatch_state *cur = mark_state;
5598 SV *find = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
5601 if ( sv_eq( cur->u.mark.mark_name,
5604 ST.mark_name = find;
5605 PUSH_STATE_GOTO( SKIP_next, next );
5607 cur = cur->u.mark.prev_mark;
5610 /* Didn't find our (*MARK:NAME) so ignore this (*SKIP:NAME) */
5612 case SKIP_next_fail:
5614 /* (*CUT:NAME) - Set up to search for the name as we
5615 collapse the stack*/
5616 popmark = ST.mark_name;
5618 /* (*CUT) - No name, we cut here.*/
5619 if (ST.mark_loc > startpoint)
5620 reginfo->cutpoint = HOPBACKc(ST.mark_loc, 1);
5621 /* but we set sv_commit to latest mark_name if there
5622 is one so they can test to see how things lead to this
5625 sv_commit=mark_state->u.mark.mark_name;
5633 if ( n == (U32)what_len_TRICKYFOLD(locinput,utf8_target,ln) ) {
5635 } else if ( LATIN_SMALL_LETTER_SHARP_S == n && !utf8_target && !UTF_PATTERN ) {
5638 U8 folded[UTF8_MAXBYTES_CASE+1];
5640 const char * const l = locinput;
5641 char *e = PL_regeol;
5642 to_uni_fold(n, folded, &foldlen);
5644 if (! foldEQ_utf8((const char*) folded, 0, foldlen, 1,
5645 l, &e, 0, utf8_target)) {
5650 nextchr = UCHARAT(locinput);
5653 if ((n=is_LNBREAK(locinput,utf8_target))) {
5655 nextchr = UCHARAT(locinput);
5660 #define CASE_CLASS(nAmE) \
5662 if ((n=is_##nAmE(locinput,utf8_target))) { \
5664 nextchr = UCHARAT(locinput); \
5669 if ((n=is_##nAmE(locinput,utf8_target))) { \
5672 locinput += UTF8SKIP(locinput); \
5673 nextchr = UCHARAT(locinput); \
5678 CASE_CLASS(HORIZWS);
5682 PerlIO_printf(Perl_error_log, "%"UVxf" %d\n",
5683 PTR2UV(scan), OP(scan));
5684 Perl_croak(aTHX_ "regexp memory corruption");
5688 /* switch break jumps here */
5689 scan = next; /* prepare to execute the next op and ... */
5690 continue; /* ... jump back to the top, reusing st */
5694 /* push a state that backtracks on success */
5695 st->u.yes.prev_yes_state = yes_state;
5699 /* push a new regex state, then continue at scan */
5701 regmatch_state *newst;
5704 regmatch_state *cur = st;
5705 regmatch_state *curyes = yes_state;
5707 regmatch_slab *slab = PL_regmatch_slab;
5708 for (;curd > -1;cur--,curd--) {
5709 if (cur < SLAB_FIRST(slab)) {
5711 cur = SLAB_LAST(slab);
5713 PerlIO_printf(Perl_error_log, "%*s#%-3d %-10s %s\n",
5714 REPORT_CODE_OFF + 2 + depth * 2,"",
5715 curd, PL_reg_name[cur->resume_state],
5716 (curyes == cur) ? "yes" : ""
5719 curyes = cur->u.yes.prev_yes_state;
5722 DEBUG_STATE_pp("push")
5725 st->locinput = locinput;
5727 if (newst > SLAB_LAST(PL_regmatch_slab))
5728 newst = S_push_slab(aTHX);
5729 PL_regmatch_state = newst;
5731 locinput = PL_reginput;
5732 nextchr = UCHARAT(locinput);
5740 * We get here only if there's trouble -- normally "case END" is
5741 * the terminating point.
5743 Perl_croak(aTHX_ "corrupted regexp pointers");
5749 /* we have successfully completed a subexpression, but we must now
5750 * pop to the state marked by yes_state and continue from there */
5751 assert(st != yes_state);
5753 while (st != yes_state) {
5755 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5756 PL_regmatch_slab = PL_regmatch_slab->prev;
5757 st = SLAB_LAST(PL_regmatch_slab);
5761 DEBUG_STATE_pp("pop (no final)");
5763 DEBUG_STATE_pp("pop (yes)");
5769 while (yes_state < SLAB_FIRST(PL_regmatch_slab)
5770 || yes_state > SLAB_LAST(PL_regmatch_slab))
5772 /* not in this slab, pop slab */
5773 depth -= (st - SLAB_FIRST(PL_regmatch_slab) + 1);
5774 PL_regmatch_slab = PL_regmatch_slab->prev;
5775 st = SLAB_LAST(PL_regmatch_slab);
5777 depth -= (st - yes_state);
5780 yes_state = st->u.yes.prev_yes_state;
5781 PL_regmatch_state = st;
5784 locinput= st->locinput;
5785 nextchr = UCHARAT(locinput);
5787 state_num = st->resume_state + no_final;
5788 goto reenter_switch;
5791 DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch successful!%s\n",
5792 PL_colors[4], PL_colors[5]));
5794 if (PL_reg_eval_set) {
5795 /* each successfully executed (?{...}) block does the equivalent of
5796 * local $^R = do {...}
5797 * When popping the save stack, all these locals would be undone;
5798 * bypass this by setting the outermost saved $^R to the latest
5800 if (oreplsv != GvSV(PL_replgv))
5801 sv_setsv(oreplsv, GvSV(PL_replgv));
5808 PerlIO_printf(Perl_debug_log,
5809 "%*s %sfailed...%s\n",
5810 REPORT_CODE_OFF+depth*2, "",
5811 PL_colors[4], PL_colors[5])
5823 /* there's a previous state to backtrack to */
5825 if (st < SLAB_FIRST(PL_regmatch_slab)) {
5826 PL_regmatch_slab = PL_regmatch_slab->prev;
5827 st = SLAB_LAST(PL_regmatch_slab);
5829 PL_regmatch_state = st;
5830 locinput= st->locinput;
5831 nextchr = UCHARAT(locinput);
5833 DEBUG_STATE_pp("pop");
5835 if (yes_state == st)
5836 yes_state = st->u.yes.prev_yes_state;
5838 state_num = st->resume_state + 1; /* failure = success + 1 */
5839 goto reenter_switch;
5844 if (rex->intflags & PREGf_VERBARG_SEEN) {
5845 SV *sv_err = get_sv("REGERROR", 1);
5846 SV *sv_mrk = get_sv("REGMARK", 1);
5848 sv_commit = &PL_sv_no;
5850 sv_yes_mark = &PL_sv_yes;
5853 sv_commit = &PL_sv_yes;
5854 sv_yes_mark = &PL_sv_no;
5856 sv_setsv(sv_err, sv_commit);
5857 sv_setsv(sv_mrk, sv_yes_mark);
5860 /* clean up; in particular, free all slabs above current one */
5861 LEAVE_SCOPE(oldsave);
5867 - regrepeat - repeatedly match something simple, report how many
5870 * [This routine now assumes that it will only match on things of length 1.
5871 * That was true before, but now we assume scan - reginput is the count,
5872 * rather than incrementing count on every character. [Er, except utf8.]]
5875 S_regrepeat(pTHX_ const regexp *prog, const regnode *p, I32 max, int depth)
5878 register char *scan;
5880 register char *loceol = PL_regeol;
5881 register I32 hardcount = 0;
5882 register bool utf8_target = PL_reg_match_utf8;
5885 PERL_UNUSED_ARG(depth);
5888 PERL_ARGS_ASSERT_REGREPEAT;
5891 if (max == REG_INFTY)
5893 else if (max < loceol - scan)
5894 loceol = scan + max;
5899 while (scan < loceol && hardcount < max && *scan != '\n') {
5900 scan += UTF8SKIP(scan);
5904 while (scan < loceol && *scan != '\n')
5911 while (scan < loceol && hardcount < max) {
5912 scan += UTF8SKIP(scan);
5923 /* To get here, EXACTish nodes must have *byte* length == 1. That
5924 * means they match only characters in the string that can be expressed
5925 * as a single byte. For non-utf8 strings, that means a simple match.
5926 * For utf8 strings, the character matched must be an invariant, or
5927 * downgradable to a single byte. The pattern's utf8ness is
5928 * irrelevant, as since it's a single byte, it either isn't utf8, or if
5929 * it is, it's an invariant */
5932 assert(! UTF_PATTERN || UNI_IS_INVARIANT(c));
5934 if (! utf8_target || UNI_IS_INVARIANT(c)) {
5935 while (scan < loceol && UCHARAT(scan) == c) {
5941 /* Here, the string is utf8, and the pattern char is different
5942 * in utf8 than not, so can't compare them directly. Outside the
5943 * loop, find find the two utf8 bytes that represent c, and then
5944 * look for those in sequence in the utf8 string */
5945 U8 high = UTF8_TWO_BYTE_HI(c);
5946 U8 low = UTF8_TWO_BYTE_LO(c);
5949 while (hardcount < max
5950 && scan + 1 < loceol
5951 && UCHARAT(scan) == high
5952 && UCHARAT(scan + 1) == low)
5960 utf8_flags = FOLDEQ_UTF8_NOMIX_ASCII;
5964 PL_reg_flags |= RF_tainted;
5965 utf8_flags = FOLDEQ_UTF8_LOCALE;
5972 /* The comments for the EXACT case above apply as well to these fold
5977 assert(! UTF_PATTERN || UNI_IS_INVARIANT(c));
5979 if (utf8_target) { /* Use full Unicode fold matching */
5980 char *tmpeol = loceol;
5981 while (hardcount < max
5982 && foldEQ_utf8_flags(scan, &tmpeol, 0, utf8_target,
5983 STRING(p), NULL, 1, cBOOL(UTF_PATTERN), utf8_flags))
5990 /* XXX Note that the above handles properly the German sharp s in
5991 * the pattern matching ss in the string. But it doesn't handle
5992 * properly cases where the string contains say 'LIGATURE ff' and
5993 * the pattern is 'f+'. This would require, say, a new function or
5994 * revised interface to foldEQ_utf8(), in which the maximum number
5995 * of characters to match could be passed and it would return how
5996 * many actually did. This is just one of many cases where
5997 * multi-char folds don't work properly, and so the fix is being
6003 /* Here, the string isn't utf8 and c is a single byte; and either
6004 * the pattern isn't utf8 or c is an invariant, so its utf8ness
6005 * doesn't affect c. Can just do simple comparisons for exact or
6008 case EXACTF: folded = PL_fold[c]; break;
6010 case EXACTFU: folded = PL_fold_latin1[c]; break;
6011 case EXACTFL: folded = PL_fold_locale[c]; break;
6012 default: Perl_croak(aTHX_ "panic: Unexpected op %u", OP(p));
6014 while (scan < loceol &&
6015 (UCHARAT(scan) == c || UCHARAT(scan) == folded))
6023 if (utf8_target || OP(p) == ANYOFV) {
6026 inclasslen = loceol - scan;
6027 while (hardcount < max
6028 && ((inclasslen = loceol - scan) > 0)
6029 && reginclass(prog, p, (U8*)scan, &inclasslen, utf8_target))
6035 while (scan < loceol && REGINCLASS(prog, p, (U8*)scan))
6043 LOAD_UTF8_CHARCLASS_ALNUM();
6044 while (hardcount < max && scan < loceol &&
6045 swash_fetch(PL_utf8_alnum, (U8*)scan, utf8_target))
6047 scan += UTF8SKIP(scan);
6051 while (scan < loceol && isWORDCHAR_L1((U8) *scan)) {
6059 while (scan < loceol && isALNUM((U8) *scan)) {
6064 while (scan < loceol && isWORDCHAR_A((U8) *scan)) {
6069 PL_reg_flags |= RF_tainted;
6072 while (hardcount < max && scan < loceol &&
6073 isALNUM_LC_utf8((U8*)scan)) {
6074 scan += UTF8SKIP(scan);
6078 while (scan < loceol && isALNUM_LC(*scan))
6088 LOAD_UTF8_CHARCLASS_ALNUM();
6089 while (hardcount < max && scan < loceol &&
6090 ! swash_fetch(PL_utf8_alnum, (U8*)scan, utf8_target))
6092 scan += UTF8SKIP(scan);
6096 while (scan < loceol && ! isWORDCHAR_L1((U8) *scan)) {
6103 goto utf8_Nwordchar;
6104 while (scan < loceol && ! isALNUM((U8) *scan)) {
6110 while (scan < loceol && ! isWORDCHAR_A((U8) *scan)) {
6111 scan += UTF8SKIP(scan);
6115 while (scan < loceol && ! isWORDCHAR_A((U8) *scan)) {
6121 PL_reg_flags |= RF_tainted;
6124 while (hardcount < max && scan < loceol &&
6125 !isALNUM_LC_utf8((U8*)scan)) {
6126 scan += UTF8SKIP(scan);
6130 while (scan < loceol && !isALNUM_LC(*scan))
6140 LOAD_UTF8_CHARCLASS_SPACE();
6141 while (hardcount < max && scan < loceol &&
6143 swash_fetch(PL_utf8_space,(U8*)scan, utf8_target)))
6145 scan += UTF8SKIP(scan);
6151 while (scan < loceol && isSPACE_L1((U8) *scan)) {
6160 while (scan < loceol && isSPACE((U8) *scan)) {
6165 while (scan < loceol && isSPACE_A((U8) *scan)) {
6170 PL_reg_flags |= RF_tainted;
6173 while (hardcount < max && scan < loceol &&
6174 isSPACE_LC_utf8((U8*)scan)) {
6175 scan += UTF8SKIP(scan);
6179 while (scan < loceol && isSPACE_LC(*scan))
6189 LOAD_UTF8_CHARCLASS_SPACE();
6190 while (hardcount < max && scan < loceol &&
6192 swash_fetch(PL_utf8_space,(U8*)scan, utf8_target)))
6194 scan += UTF8SKIP(scan);
6200 while (scan < loceol && ! isSPACE_L1((U8) *scan)) {
6209 while (scan < loceol && ! isSPACE((U8) *scan)) {
6215 while (scan < loceol && ! isSPACE_A((U8) *scan)) {
6216 scan += UTF8SKIP(scan);
6220 while (scan < loceol && ! isSPACE_A((U8) *scan)) {
6226 PL_reg_flags |= RF_tainted;
6229 while (hardcount < max && scan < loceol &&
6230 !isSPACE_LC_utf8((U8*)scan)) {
6231 scan += UTF8SKIP(scan);
6235 while (scan < loceol && !isSPACE_LC(*scan))
6242 LOAD_UTF8_CHARCLASS_DIGIT();
6243 while (hardcount < max && scan < loceol &&
6244 swash_fetch(PL_utf8_digit, (U8*)scan, utf8_target)) {
6245 scan += UTF8SKIP(scan);
6249 while (scan < loceol && isDIGIT(*scan))
6254 while (scan < loceol && isDIGIT_A((U8) *scan)) {
6259 PL_reg_flags |= RF_tainted;
6262 while (hardcount < max && scan < loceol &&
6263 isDIGIT_LC_utf8((U8*)scan)) {
6264 scan += UTF8SKIP(scan);
6268 while (scan < loceol && isDIGIT_LC(*scan))
6275 LOAD_UTF8_CHARCLASS_DIGIT();
6276 while (hardcount < max && scan < loceol &&
6277 !swash_fetch(PL_utf8_digit, (U8*)scan, utf8_target)) {
6278 scan += UTF8SKIP(scan);
6282 while (scan < loceol && !isDIGIT(*scan))
6288 while (scan < loceol && ! isDIGIT_A((U8) *scan)) {
6289 scan += UTF8SKIP(scan);
6293 while (scan < loceol && ! isDIGIT_A((U8) *scan)) {
6299 PL_reg_flags |= RF_tainted;
6302 while (hardcount < max && scan < loceol &&
6303 !isDIGIT_LC_utf8((U8*)scan)) {
6304 scan += UTF8SKIP(scan);
6308 while (scan < loceol && !isDIGIT_LC(*scan))
6315 while (hardcount < max && scan < loceol && (c=is_LNBREAK_utf8(scan))) {
6321 LNBREAK can match two latin chars, which is ok,
6322 because we have a null terminated string, but we
6323 have to use hardcount in this situation
6325 while (scan < loceol && (c=is_LNBREAK_latin1(scan))) {
6334 while (hardcount < max && scan < loceol && (c=is_HORIZWS_utf8(scan))) {
6339 while (scan < loceol && is_HORIZWS_latin1(scan))
6346 while (hardcount < max && scan < loceol && !is_HORIZWS_utf8(scan)) {
6347 scan += UTF8SKIP(scan);
6351 while (scan < loceol && !is_HORIZWS_latin1(scan))
6359 while (hardcount < max && scan < loceol && (c=is_VERTWS_utf8(scan))) {
6364 while (scan < loceol && is_VERTWS_latin1(scan))
6372 while (hardcount < max && scan < loceol && !is_VERTWS_utf8(scan)) {
6373 scan += UTF8SKIP(scan);
6377 while (scan < loceol && !is_VERTWS_latin1(scan))
6383 default: /* Called on something of 0 width. */
6384 break; /* So match right here or not at all. */
6390 c = scan - PL_reginput;
6394 GET_RE_DEBUG_FLAGS_DECL;
6396 SV * const prop = sv_newmortal();
6397 regprop(prog, prop, p);
6398 PerlIO_printf(Perl_debug_log,
6399 "%*s %s can match %"IVdf" times out of %"IVdf"...\n",
6400 REPORT_CODE_OFF + depth*2, "", SvPVX_const(prop),(IV)c,(IV)max);
6408 #if !defined(PERL_IN_XSUB_RE) || defined(PLUGGABLE_RE_EXTENSION)
6410 - regclass_swash - prepare the utf8 swash
6414 Perl_regclass_swash(pTHX_ const regexp *prog, register const regnode* node, bool doinit, SV** listsvp, SV **altsvp)
6420 RXi_GET_DECL(prog,progi);
6421 const struct reg_data * const data = prog ? progi->data : NULL;
6423 PERL_ARGS_ASSERT_REGCLASS_SWASH;
6425 assert(ANYOF_NONBITMAP(node));
6427 if (data && data->count) {
6428 const U32 n = ARG(node);
6430 if (data->what[n] == 's') {
6431 SV * const rv = MUTABLE_SV(data->data[n]);
6432 AV * const av = MUTABLE_AV(SvRV(rv));
6433 SV **const ary = AvARRAY(av);
6436 /* See the end of regcomp.c:S_regclass() for
6437 * documentation of these array elements. */
6440 a = SvROK(ary[1]) ? &ary[1] : NULL;
6441 b = SvTYPE(ary[2]) == SVt_PVAV ? &ary[2] : NULL;
6445 else if (si && doinit) {
6446 sw = swash_init("utf8", "", si, 1, 0);
6447 (void)av_store(av, 1, sw);
6464 - reginclass - determine if a character falls into a character class
6466 n is the ANYOF regnode
6467 p is the target string
6468 lenp is pointer to the maximum number of bytes of how far to go in p
6469 (This is assumed wthout checking to always be at least the current
6471 utf8_target tells whether p is in UTF-8.
6473 Returns true if matched; false otherwise. If lenp is not NULL, on return
6474 from a successful match, the value it points to will be updated to how many
6475 bytes in p were matched. If there was no match, the value is undefined,
6476 possibly changed from the input.
6478 Note that this can be a synthetic start class, a combination of various
6479 nodes, so things you think might be mutually exclusive, such as locale,
6480 aren't. It can match both locale and non-locale
6485 S_reginclass(pTHX_ const regexp * const prog, register const regnode * const n, register const U8* const p, STRLEN* lenp, register const bool utf8_target)
6488 const char flags = ANYOF_FLAGS(n);
6494 PERL_ARGS_ASSERT_REGINCLASS;
6496 /* If c is not already the code point, get it */
6497 if (utf8_target && !UTF8_IS_INVARIANT(c)) {
6498 c = utf8n_to_uvchr(p, UTF8_MAXBYTES, &c_len,
6499 (UTF8_ALLOW_DEFAULT & UTF8_ALLOW_ANYUV)
6500 | UTF8_ALLOW_FFFF | UTF8_CHECK_ONLY);
6501 /* see [perl #37836] for UTF8_ALLOW_ANYUV; [perl #38293] for
6502 * UTF8_ALLOW_FFFF */
6503 if (c_len == (STRLEN)-1)
6504 Perl_croak(aTHX_ "Malformed UTF-8 character (fatal)");
6510 /* Use passed in max length, or one character if none passed in or less
6511 * than one character. And assume will match just one character. This is
6512 * overwritten later if matched more. */
6514 maxlen = (*lenp > c_len) ? *lenp : c_len;
6522 /* If this character is potentially in the bitmap, check it */
6524 if (ANYOF_BITMAP_TEST(n, c))
6526 else if (flags & ANYOF_NON_UTF8_LATIN1_ALL
6533 else if (flags & ANYOF_LOCALE) {
6534 PL_reg_flags |= RF_tainted;
6536 if ((flags & ANYOF_LOC_NONBITMAP_FOLD)
6537 && ANYOF_BITMAP_TEST(n, PL_fold_locale[c]))
6541 else if (ANYOF_CLASS_TEST_ANY_SET(n) &&
6542 ((ANYOF_CLASS_TEST(n, ANYOF_ALNUM) && isALNUM_LC(c)) ||
6543 (ANYOF_CLASS_TEST(n, ANYOF_NALNUM) && !isALNUM_LC(c)) ||
6544 (ANYOF_CLASS_TEST(n, ANYOF_SPACE) && isSPACE_LC(c)) ||
6545 (ANYOF_CLASS_TEST(n, ANYOF_NSPACE) && !isSPACE_LC(c)) ||
6546 (ANYOF_CLASS_TEST(n, ANYOF_DIGIT) && isDIGIT_LC(c)) ||
6547 (ANYOF_CLASS_TEST(n, ANYOF_NDIGIT) && !isDIGIT_LC(c)) ||
6548 (ANYOF_CLASS_TEST(n, ANYOF_ALNUMC) && isALNUMC_LC(c)) ||
6549 (ANYOF_CLASS_TEST(n, ANYOF_NALNUMC) && !isALNUMC_LC(c)) ||
6550 (ANYOF_CLASS_TEST(n, ANYOF_ALPHA) && isALPHA_LC(c)) ||
6551 (ANYOF_CLASS_TEST(n, ANYOF_NALPHA) && !isALPHA_LC(c)) ||
6552 (ANYOF_CLASS_TEST(n, ANYOF_ASCII) && isASCII(c)) ||
6553 (ANYOF_CLASS_TEST(n, ANYOF_NASCII) && !isASCII(c)) ||
6554 (ANYOF_CLASS_TEST(n, ANYOF_CNTRL) && isCNTRL_LC(c)) ||
6555 (ANYOF_CLASS_TEST(n, ANYOF_NCNTRL) && !isCNTRL_LC(c)) ||
6556 (ANYOF_CLASS_TEST(n, ANYOF_GRAPH) && isGRAPH_LC(c)) ||
6557 (ANYOF_CLASS_TEST(n, ANYOF_NGRAPH) && !isGRAPH_LC(c)) ||
6558 (ANYOF_CLASS_TEST(n, ANYOF_LOWER) && isLOWER_LC(c)) ||
6559 (ANYOF_CLASS_TEST(n, ANYOF_NLOWER) && !isLOWER_LC(c)) ||
6560 (ANYOF_CLASS_TEST(n, ANYOF_PRINT) && isPRINT_LC(c)) ||
6561 (ANYOF_CLASS_TEST(n, ANYOF_NPRINT) && !isPRINT_LC(c)) ||
6562 (ANYOF_CLASS_TEST(n, ANYOF_PUNCT) && isPUNCT_LC(c)) ||
6563 (ANYOF_CLASS_TEST(n, ANYOF_NPUNCT) && !isPUNCT_LC(c)) ||
6564 (ANYOF_CLASS_TEST(n, ANYOF_UPPER) && isUPPER_LC(c)) ||
6565 (ANYOF_CLASS_TEST(n, ANYOF_NUPPER) && !isUPPER_LC(c)) ||
6566 (ANYOF_CLASS_TEST(n, ANYOF_XDIGIT) && isXDIGIT(c)) ||
6567 (ANYOF_CLASS_TEST(n, ANYOF_NXDIGIT) && !isXDIGIT(c)) ||
6568 (ANYOF_CLASS_TEST(n, ANYOF_PSXSPC) && isPSXSPC(c)) ||
6569 (ANYOF_CLASS_TEST(n, ANYOF_NPSXSPC) && !isPSXSPC(c)) ||
6570 (ANYOF_CLASS_TEST(n, ANYOF_BLANK) && isBLANK(c)) ||
6571 (ANYOF_CLASS_TEST(n, ANYOF_NBLANK) && !isBLANK(c))
6572 ) /* How's that for a conditional? */
6579 /* If the bitmap didn't (or couldn't) match, and something outside the
6580 * bitmap could match, try that. Locale nodes specifiy completely the
6581 * behavior of code points in the bit map (otherwise, a utf8 target would
6582 * cause them to be treated as Unicode and not locale), except in
6583 * the very unlikely event when this node is a synthetic start class, which
6584 * could be a combination of locale and non-locale nodes. So allow locale
6585 * to match for the synthetic start class, which will give a false
6586 * positive that will be resolved when the match is done again as not part
6587 * of the synthetic start class */
6589 if (utf8_target && (flags & ANYOF_UNICODE_ALL) && c >= 256) {
6590 match = TRUE; /* Everything above 255 matches */
6592 else if (ANYOF_NONBITMAP(n)
6593 && ((flags & ANYOF_NONBITMAP_NON_UTF8)
6596 || (! (flags & ANYOF_LOCALE))
6597 || (flags & ANYOF_IS_SYNTHETIC)))))
6600 SV * const sw = regclass_swash(prog, n, TRUE, 0, (SV**)&av);
6608 /* Not utf8. Convert as much of the string as available up
6609 * to the limit of how far the (single) character in the
6610 * pattern can possibly match (no need to go further). If
6611 * the node is a straight ANYOF or not folding, it can't
6612 * match more than one. Otherwise, It can match up to how
6613 * far a single char can fold to. Since not utf8, each
6614 * character is a single byte, so the max it can be in
6615 * bytes is the same as the max it can be in characters */
6616 STRLEN len = (OP(n) == ANYOF
6617 || ! (flags & ANYOF_LOC_NONBITMAP_FOLD))
6619 : (maxlen < UTF8_MAX_FOLD_CHAR_EXPAND)
6621 : UTF8_MAX_FOLD_CHAR_EXPAND;
6622 utf8_p = bytes_to_utf8(p, &len);
6625 if (swash_fetch(sw, utf8_p, TRUE))
6627 else if (flags & ANYOF_LOC_NONBITMAP_FOLD) {
6629 /* Here, we need to test if the fold of the target string
6630 * matches. The non-multi char folds have all been moved to
6631 * the compilation phase, and the multi-char folds have
6632 * been stored by regcomp into 'av'; we linearly check to
6633 * see if any match the target string (folded). We know
6634 * that the originals were each one character, but we don't
6635 * currently know how many characters/bytes each folded to,
6636 * except we do know that there are small limits imposed by
6637 * Unicode. XXX A performance enhancement would be to have
6638 * regcomp.c store the max number of chars/bytes that are
6639 * in an av entry, as, say the 0th element. Even better
6640 * would be to have a hash of the few characters that can
6641 * start a multi-char fold to the max number of chars of
6644 * If there is a match, we will need to advance (if lenp is
6645 * specified) the match pointer in the target string. But
6646 * what we are comparing here isn't that string directly,
6647 * but its fold, whose length may differ from the original.
6648 * As we go along in constructing the fold, therefore, we
6649 * create a map so that we know how many bytes in the
6650 * source to advance given that we have matched a certain
6651 * number of bytes in the fold. This map is stored in
6652 * 'map_fold_len_back'. Let n mean the number of bytes in
6653 * the fold of the first character that we are folding.
6654 * Then map_fold_len_back[n] is set to the number of bytes
6655 * in that first character. Similarly let m be the
6656 * corresponding number for the second character to be
6657 * folded. Then map_fold_len_back[n+m] is set to the
6658 * number of bytes occupied by the first two source
6659 * characters. ... */
6660 U8 map_fold_len_back[UTF8_MAXBYTES_CASE+1] = { 0 };
6661 U8 folded[UTF8_MAXBYTES_CASE+1];
6662 STRLEN foldlen = 0; /* num bytes in fold of 1st char */
6663 STRLEN total_foldlen = 0; /* num bytes in fold of all
6666 if (OP(n) == ANYOF || maxlen == 1 || ! lenp || ! av) {
6668 /* Here, only need to fold the first char of the target
6669 * string. It the source wasn't utf8, is 1 byte long */
6670 to_utf8_fold(utf8_p, folded, &foldlen);
6671 total_foldlen = foldlen;
6672 map_fold_len_back[foldlen] = (utf8_target)
6678 /* Here, need to fold more than the first char. Do so
6679 * up to the limits */
6680 U8* source_ptr = utf8_p; /* The source for the fold
6683 U8* folded_ptr = folded;
6684 U8* e = utf8_p + maxlen; /* Can't go beyond last
6685 available byte in the
6689 i < UTF8_MAX_FOLD_CHAR_EXPAND && source_ptr < e;
6693 /* Fold the next character */
6694 U8 this_char_folded[UTF8_MAXBYTES_CASE+1];
6695 STRLEN this_char_foldlen;
6696 to_utf8_fold(source_ptr,
6698 &this_char_foldlen);
6700 /* Bail if it would exceed the byte limit for
6701 * folding a single char. */
6702 if (this_char_foldlen + folded_ptr - folded >
6708 /* Add the fold of this character */
6709 Copy(this_char_folded,
6713 source_ptr += UTF8SKIP(source_ptr);
6714 folded_ptr += this_char_foldlen;
6715 total_foldlen = folded_ptr - folded;
6717 /* Create map from the number of bytes in the fold
6718 * back to the number of bytes in the source. If
6719 * the source isn't utf8, the byte count is just
6720 * the number of characters so far */
6721 map_fold_len_back[total_foldlen]
6723 ? source_ptr - utf8_p
6730 /* Do the linear search to see if the fold is in the list
6731 * of multi-char folds. */
6734 for (i = 0; i <= av_len(av); i++) {
6735 SV* const sv = *av_fetch(av, i, FALSE);
6737 const char * const s = SvPV_const(sv, len);
6739 if (len <= total_foldlen
6740 && memEQ(s, (char*)folded, len)
6742 /* If 0, means matched a partial char. See
6744 && map_fold_len_back[len])
6747 /* Advance the target string ptr to account for
6748 * this fold, but have to translate from the
6749 * folded length to the corresponding source
6752 *lenp = map_fold_len_back[len];
6761 /* If we allocated a string above, free it */
6762 if (! utf8_target) Safefree(utf8_p);
6767 return (flags & ANYOF_INVERT) ? !match : match;
6771 S_reghop3(U8 *s, I32 off, const U8* lim)
6775 PERL_ARGS_ASSERT_REGHOP3;
6778 while (off-- && s < lim) {
6779 /* XXX could check well-formedness here */
6784 while (off++ && s > lim) {
6786 if (UTF8_IS_CONTINUED(*s)) {
6787 while (s > lim && UTF8_IS_CONTINUATION(*s))
6790 /* XXX could check well-formedness here */
6797 /* there are a bunch of places where we use two reghop3's that should
6798 be replaced with this routine. but since thats not done yet
6799 we ifdef it out - dmq
6802 S_reghop4(U8 *s, I32 off, const U8* llim, const U8* rlim)
6806 PERL_ARGS_ASSERT_REGHOP4;
6809 while (off-- && s < rlim) {
6810 /* XXX could check well-formedness here */
6815 while (off++ && s > llim) {
6817 if (UTF8_IS_CONTINUED(*s)) {
6818 while (s > llim && UTF8_IS_CONTINUATION(*s))
6821 /* XXX could check well-formedness here */
6829 S_reghopmaybe3(U8* s, I32 off, const U8* lim)
6833 PERL_ARGS_ASSERT_REGHOPMAYBE3;
6836 while (off-- && s < lim) {
6837 /* XXX could check well-formedness here */
6844 while (off++ && s > lim) {
6846 if (UTF8_IS_CONTINUED(*s)) {
6847 while (s > lim && UTF8_IS_CONTINUATION(*s))
6850 /* XXX could check well-formedness here */
6859 restore_pos(pTHX_ void *arg)
6862 regexp * const rex = (regexp *)arg;
6863 if (PL_reg_eval_set) {
6864 if (PL_reg_oldsaved) {
6865 rex->subbeg = PL_reg_oldsaved;
6866 rex->sublen = PL_reg_oldsavedlen;
6867 #ifdef PERL_OLD_COPY_ON_WRITE
6868 rex->saved_copy = PL_nrs;
6870 RXp_MATCH_COPIED_on(rex);
6872 PL_reg_magic->mg_len = PL_reg_oldpos;
6873 PL_reg_eval_set = 0;
6874 PL_curpm = PL_reg_oldcurpm;
6879 S_to_utf8_substr(pTHX_ register regexp *prog)
6883 PERL_ARGS_ASSERT_TO_UTF8_SUBSTR;
6886 if (prog->substrs->data[i].substr
6887 && !prog->substrs->data[i].utf8_substr) {
6888 SV* const sv = newSVsv(prog->substrs->data[i].substr);
6889 prog->substrs->data[i].utf8_substr = sv;
6890 sv_utf8_upgrade(sv);
6891 if (SvVALID(prog->substrs->data[i].substr)) {
6892 const U8 flags = BmFLAGS(prog->substrs->data[i].substr);
6893 if (flags & FBMcf_TAIL) {
6894 /* Trim the trailing \n that fbm_compile added last
6896 SvCUR_set(sv, SvCUR(sv) - 1);
6897 /* Whilst this makes the SV technically "invalid" (as its
6898 buffer is no longer followed by "\0") when fbm_compile()
6899 adds the "\n" back, a "\0" is restored. */
6901 fbm_compile(sv, flags);
6903 if (prog->substrs->data[i].substr == prog->check_substr)
6904 prog->check_utf8 = sv;
6910 S_to_byte_substr(pTHX_ register regexp *prog)
6915 PERL_ARGS_ASSERT_TO_BYTE_SUBSTR;
6918 if (prog->substrs->data[i].utf8_substr
6919 && !prog->substrs->data[i].substr) {
6920 SV* sv = newSVsv(prog->substrs->data[i].utf8_substr);
6921 if (sv_utf8_downgrade(sv, TRUE)) {
6922 if (SvVALID(prog->substrs->data[i].utf8_substr)) {
6924 = BmFLAGS(prog->substrs->data[i].utf8_substr);
6925 if (flags & FBMcf_TAIL) {
6926 /* Trim the trailing \n that fbm_compile added last
6928 SvCUR_set(sv, SvCUR(sv) - 1);
6930 fbm_compile(sv, flags);
6936 prog->substrs->data[i].substr = sv;
6937 if (prog->substrs->data[i].utf8_substr == prog->check_utf8)
6938 prog->check_substr = sv;
6945 * c-indentation-style: bsd
6947 * indent-tabs-mode: t
6950 * ex: set ts=8 sts=4 sw=4 noet: