+ default:
+ Perl_croak(aTHX_ "panic: unknown regstclass %d", (int)OP(c));
+ break;
+ }
+ return 0;
+ got_it:
+ return s;
+}
+
+/*
+ - regexec_flags - match a regexp against a string
+ */
+I32
+Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char *strend,
+ char *strbeg, I32 minend, SV *sv, void *data, U32 flags)
+/* strend: pointer to null at end of string */
+/* strbeg: real beginning of string */
+/* minend: end of match must be >=minend after stringarg. */
+/* data: May be used for some additional optimizations. */
+/* nosave: For optimizations. */
+{
+ dTHR;
+ register char *s;
+ register regnode *c;
+ register char *startpos = stringarg;
+ register I32 tmp;
+ I32 minlen; /* must match at least this many chars */
+ I32 dontbother = 0; /* how many characters not to try at end */
+ I32 start_shift = 0; /* Offset of the start to find
+ constant substr. */ /* CC */
+ I32 end_shift = 0; /* Same for the end. */ /* CC */
+ I32 scream_pos = -1; /* Internal iterator of scream. */
+ char *scream_olds;
+ SV* oreplsv = GvSV(PL_replgv);
+
+ PL_regcc = 0;
+
+ cache_re(prog);
+#ifdef DEBUGGING
+ PL_regnarrate = PL_debug & 512;
+#endif
+
+ /* Be paranoid... */
+ if (prog == NULL || startpos == NULL) {
+ Perl_croak(aTHX_ "NULL regexp parameter");
+ return 0;
+ }
+
+ minlen = prog->minlen;
+ if (strend - startpos < minlen) goto phooey;
+
+ if (startpos == strbeg) /* is ^ valid at stringarg? */
+ PL_regprev = '\n';
+ else {
+ PL_regprev = (U32)stringarg[-1];
+ if (!PL_multiline && PL_regprev == '\n')
+ PL_regprev = '\0'; /* force ^ to NOT match */
+ }
+
+ /* Check validity of program. */
+ if (UCHARAT(prog->program) != REG_MAGIC) {
+ Perl_croak(aTHX_ "corrupted regexp program");
+ }
+
+ PL_reg_flags = 0;
+ PL_reg_eval_set = 0;
+ PL_reg_maxiter = 0;
+
+ if (prog->reganch & ROPT_UTF8)
+ PL_reg_flags |= RF_utf8;
+
+ /* Mark beginning of line for ^ and lookbehind. */
+ PL_regbol = startpos;
+ PL_bostr = strbeg;
+ PL_reg_sv = sv;
+
+ /* Mark end of line for $ (and such) */
+ PL_regeol = strend;
+
+ /* see how far we have to get to not match where we matched before */
+ PL_regtill = startpos+minend;
+
+ /* We start without call_cc context. */
+ PL_reg_call_cc = 0;
+
+ /* If there is a "must appear" string, look for it. */
+ s = startpos;
+
+ if (prog->reganch & ROPT_GPOS_SEEN) { /* Need to have PL_reg_ganch */
+ MAGIC *mg;
+
+ if (flags & REXEC_IGNOREPOS) /* Means: check only at start */
+ PL_reg_ganch = startpos;
+ else if (sv && SvTYPE(sv) >= SVt_PVMG
+ && SvMAGIC(sv)
+ && (mg = mg_find(sv, 'g')) && mg->mg_len >= 0) {
+ PL_reg_ganch = strbeg + mg->mg_len; /* Defined pos() */
+ if (prog->reganch & ROPT_ANCH_GPOS) {
+ if (s > PL_reg_ganch)
+ goto phooey;
+ s = PL_reg_ganch;
+ }
+ }
+ else /* pos() not defined */
+ PL_reg_ganch = strbeg;
+ }
+
+ if (!(flags & REXEC_CHECKED) && prog->check_substr != Nullsv) {
+ re_scream_pos_data d;
+
+ d.scream_olds = &scream_olds;
+ d.scream_pos = &scream_pos;
+ s = re_intuit_start(prog, sv, s, strend, flags, &d);
+ if (!s)
+ goto phooey; /* not present */
+ }
+
+ DEBUG_r( if (!PL_colorset) reginitcolors() );
+ DEBUG_r(PerlIO_printf(Perl_debug_log,
+ "%sMatching REx%s `%s%.60s%s%s' against `%s%.*s%s%s'\n",
+ PL_colors[4],PL_colors[5],PL_colors[0],
+ prog->precomp,
+ PL_colors[1],
+ (strlen(prog->precomp) > 60 ? "..." : ""),
+ PL_colors[0],
+ (int)(strend - startpos > 60 ? 60 : strend - startpos),
+ startpos, PL_colors[1],
+ (strend - startpos > 60 ? "..." : ""))
+ );
+
+ /* Simplest case: anchored match need be tried only once. */
+ /* [unless only anchor is BOL and multiline is set] */
+ if (prog->reganch & (ROPT_ANCH & ~ROPT_ANCH_GPOS)) {
+ if (s == startpos && regtry(prog, startpos))
+ goto got_it;
+ else if (PL_multiline || (prog->reganch & ROPT_IMPLICIT)
+ || (prog->reganch & ROPT_ANCH_MBOL)) /* XXXX SBOL? */
+ {
+ char *end;
+
+ if (minlen)
+ dontbother = minlen - 1;
+ end = HOPc(strend, -dontbother) - 1;
+ /* for multiline we only have to try after newlines */
+ if (prog->check_substr) {
+ if (s == startpos)
+ goto after_try;
+ while (1) {
+ if (regtry(prog, s))
+ goto got_it;
+ after_try:
+ if (s >= end)
+ goto phooey;
+ if (prog->reganch & RE_USE_INTUIT) {
+ s = re_intuit_start(prog, sv, s + 1, strend, flags, NULL);
+ if (!s)
+ goto phooey;
+ }
+ else
+ s++;
+ }
+ } else {
+ if (s > startpos)
+ s--;
+ while (s < end) {
+ if (*s++ == '\n') { /* don't need PL_utf8skip here */
+ if (regtry(prog, s))
+ goto got_it;
+ }
+ }
+ }
+ }
+ goto phooey;
+ } else if (prog->reganch & ROPT_ANCH_GPOS) {
+ if (regtry(prog, PL_reg_ganch))
+ goto got_it;
+ goto phooey;
+ }
+
+ /* Messy cases: unanchored match. */
+ if (prog->anchored_substr && prog->reganch & ROPT_SKIP) {
+ /* we have /x+whatever/ */
+ /* it must be a one character string (XXXX Except UTF?) */
+ char ch = SvPVX(prog->anchored_substr)[0];
+ if (UTF) {
+ while (s < strend) {
+ if (*s == ch) {
+ if (regtry(prog, s)) goto got_it;
+ s += UTF8SKIP(s);
+ while (s < strend && *s == ch)
+ s += UTF8SKIP(s);
+ }
+ s += UTF8SKIP(s);
+ }
+ }
+ else {
+ while (s < strend) {
+ if (*s == ch) {
+ if (regtry(prog, s)) goto got_it;
+ s++;
+ while (s < strend && *s == ch)
+ s++;
+ }
+ s++;
+ }
+ }
+ }
+ /*SUPPRESS 560*/
+ else if (prog->anchored_substr != Nullsv
+ || (prog->float_substr != Nullsv
+ && prog->float_max_offset < strend - s)) {
+ SV *must = prog->anchored_substr
+ ? prog->anchored_substr : prog->float_substr;
+ I32 back_max =
+ prog->anchored_substr ? prog->anchored_offset : prog->float_max_offset;
+ I32 back_min =
+ prog->anchored_substr ? prog->anchored_offset : prog->float_min_offset;
+ I32 delta = back_max - back_min;
+ char *last = HOPc(strend, /* Cannot start after this */
+ -(I32)(CHR_SVLEN(must)
+ - (SvTAIL(must) != 0) + back_min));
+ char *last1; /* Last position checked before */
+
+ if (s > PL_bostr)
+ last1 = HOPc(s, -1);
+ else
+ last1 = s - 1; /* bogus */
+
+ /* XXXX check_substr already used to find `s', can optimize if
+ check_substr==must. */
+ scream_pos = -1;
+ dontbother = end_shift;
+ strend = HOPc(strend, -dontbother);
+ while ( (s <= last) &&
+ ((flags & REXEC_SCREAM)
+ ? (s = screaminstr(sv, must, HOPc(s, back_min) - strbeg,
+ end_shift, &scream_pos, 0))
+ : (s = fbm_instr((unsigned char*)HOP(s, back_min),
+ (unsigned char*)strend, must,
+ PL_multiline ? FBMrf_MULTILINE : 0))) ) {
+ if (HOPc(s, -back_max) > last1) {
+ last1 = HOPc(s, -back_min);
+ s = HOPc(s, -back_max);
+ }
+ else {
+ char *t = (last1 >= PL_bostr) ? HOPc(last1, 1) : last1 + 1;
+
+ last1 = HOPc(s, -back_min);
+ s = t;
+ }
+ if (UTF) {
+ while (s <= last1) {
+ if (regtry(prog, s))
+ goto got_it;
+ s += UTF8SKIP(s);
+ }
+ }
+ else {
+ while (s <= last1) {
+ if (regtry(prog, s))
+ goto got_it;
+ s++;
+ }
+ }
https://perl5.git.perl.org / perl5.git / blobdiff