* "One Ring to rule them all, One Ring to find them..."
*/
+/* This file contains functions for executing a regular expression. See
+ * also regcomp.c which funnily enough, contains functions for compiling
+ * a regular expression.
+ *
+ * This file is also copied at build time to ext/re/re_exec.c, where
+ * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
+ * This causes the main functions to be compiled under new names and with
+ * debugging support added, which makes "use re 'debug'" work.
+
+ */
+
/* NOTE: this is derived from Henry Spencer's regexp code, and should not
* confused with the original package (see point 3 below). Thanks, Henry!
*/
# define PERL_NO_GET_CONTEXT
#endif
-/*SUPPRESS 112*/
/*
* pregcomp and pregexec -- regsub and regerror are not used in perl
*
*
**** Alterations to Henry's code are...
****
- **** Copyright (c) 1991-2002, Larry Wall
+ **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
+ **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, by Larry Wall and others
****
**** You may distribute under the terms of either the GNU General Public
**** License or the Artistic License, as specified in the README file.
#define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim))
#define HOPMAYBE3c(pos,off,lim) ((char*)HOPMAYBE3(pos,off,lim))
-#define LOAD_UTF8_CHARCLASS(a,b) STMT_START { if (!CAT2(PL_utf8_,a)) { ENTER; save_re_context(); (void)CAT2(is_utf8_, a)((U8*)b); LEAVE; } } STMT_END
+#define LOAD_UTF8_CHARCLASS(class,str) STMT_START { \
+ if (!CAT2(PL_utf8_,class)) { bool ok; ENTER; save_re_context(); ok=CAT2(is_utf8_,class)((U8*)str); assert(ok); LEAVE; } } STMT_END
+#define LOAD_UTF8_CHARCLASS_ALNUM() LOAD_UTF8_CHARCLASS(alnum,"a")
+#define LOAD_UTF8_CHARCLASS_DIGIT() LOAD_UTF8_CHARCLASS(digit,"0")
+#define LOAD_UTF8_CHARCLASS_SPACE() LOAD_UTF8_CHARCLASS(space," ")
+#define LOAD_UTF8_CHARCLASS_MARK() LOAD_UTF8_CHARCLASS(mark, "\xcd\x86")
/* for use after a quantifier and before an EXACT-like node -- japhy */
#define JUMPABLE(rn) ( \
STATIC CHECKPOINT
S_regcppush(pTHX_ I32 parenfloor)
{
- int retval = PL_savestack_ix;
+ const int retval = PL_savestack_ix;
#define REGCP_PAREN_ELEMS 4
- int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS;
+ const int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS;
int p;
if (paren_elems_to_push < 0)
Perl_croak(aTHX_ "panic: paren_elems_to_push < 0");
#define REGCP_OTHER_ELEMS 6
- SSCHECK(paren_elems_to_push + REGCP_OTHER_ELEMS);
+ SSGROW(paren_elems_to_push + REGCP_OTHER_ELEMS);
for (p = PL_regsize; p > parenfloor; p--) {
/* REGCP_PARENS_ELEMS are pushed per pairs of parentheses. */
SSPUSHINT(PL_regendp[p]);
I32 i;
U32 paren = 0;
char *input;
- I32 tmps;
/* Pop REGCP_OTHER_ELEMS before the parentheses loop starts. */
i = SSPOPINT;
/* Now restore the parentheses context. */
for (i -= (REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS);
i > 0; i -= REGCP_PAREN_ELEMS) {
+ I32 tmps;
paren = (U32)SSPOPINT;
PL_reg_start_tmp[paren] = (char *) SSPOPPTR;
PL_regstartp[paren] = SSPOPINT;
STATIC char *
S_regcp_set_to(pTHX_ I32 ss)
{
- I32 tmp = PL_savestack_ix;
+ const I32 tmp = PL_savestack_ix;
PL_savestack_ix = ss;
regcppop();
/* Assumptions: if ANCH_GPOS, then strpos is anchored. XXXX Check GPOS logic */
-/* If SCREAM, then SvPVX(sv) should be compatible with strpos and strend.
+/* If SCREAM, then SvPVX_const(sv) should be compatible with strpos and strend.
Otherwise, only SvCUR(sv) is used to get strbeg. */
/* XXXX We assume that strpos is strbeg unless sv. */
register SV *check;
char *strbeg;
char *t;
- int do_utf8 = sv ? SvUTF8(sv) : 0; /* if no sv we have to assume bytes */
+ const int do_utf8 = sv ? SvUTF8(sv) : 0; /* if no sv we have to assume bytes */
I32 ml_anch;
register char *other_last = Nullch; /* other substr checked before this */
char *check_at = Nullch; /* check substr found at this pos */
#ifdef DEBUGGING
- char *i_strpos = strpos;
- SV *dsv = PERL_DEBUG_PAD_ZERO(0);
+ const char * const i_strpos = strpos;
+ SV * const dsv = PERL_DEBUG_PAD_ZERO(0);
#endif
RX_MATCH_UTF8_set(prog,do_utf8);
}
DEBUG_r({
- char *s = PL_reg_match_utf8 ?
+ const char *s = PL_reg_match_utf8 ?
sv_uni_display(dsv, sv, 60, UNI_DISPLAY_REGEX) :
strpos;
- int len = PL_reg_match_utf8 ?
+ const int len = PL_reg_match_utf8 ?
strlen(s) : strend - strpos;
if (!PL_colorset)
reginitcolors();
DEBUG_r(PerlIO_printf(Perl_debug_log,
"UTF-8 target...\n"));
PerlIO_printf(Perl_debug_log,
- "%sGuessing start of match, REx%s `%s%.60s%s%s' against `%s%.*s%s%s'...\n",
+ "%sGuessing start of match, 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],
if (!ml_anch) {
if ( !(prog->reganch & (ROPT_ANCH_GPOS /* Checked by the caller */
| ROPT_IMPLICIT)) /* not a real BOL */
- /* SvCUR is not set on references: SvRV and SvPVX overlap */
+ /* SvCUR is not set on references: SvRV and SvPVX_const overlap */
&& sv && !SvROK(sv)
&& (strpos != strbeg)) {
DEBUG_r(PerlIO_printf(Perl_debug_log, "Not at start...\n"));
}
/* Now should match s[0..slen-2] */
slen--;
- if (slen && (*SvPVX(check) != *s
+ if (slen && (*SvPVX_const(check) != *s
|| (slen > 1
- && memNE(SvPVX(check), s, slen)))) {
+ && memNE(SvPVX_const(check), s, slen)))) {
report_neq:
DEBUG_r(PerlIO_printf(Perl_debug_log, "String not equal...\n"));
goto fail_finish;
}
}
- else if (*SvPVX(check) != *s
+ else if (*SvPVX_const(check) != *s
|| ((slen = SvCUR(check)) > 1
- && memNE(SvPVX(check), s, slen)))
+ && memNE(SvPVX_const(check), s, slen)))
goto report_neq;
+ check_at = s;
goto success_at_start;
}
}
end_shift = prog->minlen - start_shift -
CHR_SVLEN(check) + (SvTAIL(check) != 0);
if (!ml_anch) {
- I32 end = prog->check_offset_max + CHR_SVLEN(check)
+ const I32 end = prog->check_offset_max + CHR_SVLEN(check)
- (SvTAIL(check) != 0);
- I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end;
+ const I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end;
if (end_shift < eshift)
end_shift = eshift;
the "check" substring in the region corrected by start/end_shift. */
if (flags & REXEC_SCREAM) {
I32 p = -1; /* Internal iterator of scream. */
- I32 *pp = data ? data->scream_pos : &p;
+ I32 * const pp = data ? data->scream_pos : &p;
if (PL_screamfirst[BmRARE(check)] >= 0
|| ( BmRARE(check) == '\n'
goto fail_finish;
/* we may be pointing at the wrong string */
if (s && RX_MATCH_COPIED(prog))
- s = prog->subbeg + (s - SvPVX(sv));
+ s = strbeg + (s - SvPVX_const(sv));
if (data)
*data->scream_olds = s;
}
/* Update the count-of-usability, remove useless subpatterns,
unshift s. */
- DEBUG_r(PerlIO_printf(Perl_debug_log, "%s %s substr `%s%.*s%s'%s%s",
+ /* FIXME - DEBUG_EXECUTE_r if that is merged to maint. */
+ DEBUG_r(PerlIO_printf(Perl_debug_log, "%s %s substr \"%s%.*s%s\"%s%s",
(s ? "Found" : "Did not find"),
(check == (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) ? "anchored" : "floating"),
PL_colors[0],
(int)(SvCUR(check) - (SvTAIL(check)!=0)),
- SvPVX(check),
+ SvPVX_const(check),
PL_colors[1], (SvTAIL(check) ? "$" : ""),
(s ? " at offset " : "...\n") ) );
/* Got a candidate. Check MBOL anchoring, and the *other* substr.
Start with the other substr.
XXXX no SCREAM optimization yet - and a very coarse implementation
- XXXX /ttx+/ results in anchored=`ttx', floating=`x'. floating will
+ XXXX /ttx+/ results in anchored="ttx", floating="x". floating will
*always* match. Probably should be marked during compile...
Probably it is right to do no SCREAM here...
*/
if (check == (do_utf8 ? prog->float_utf8 : prog->float_substr)) {
do_other_anchored:
{
- char *last = HOP3c(s, -start_shift, strbeg), *last1, *last2;
+ char * const last = HOP3c(s, -start_shift, strbeg);
+ char *last1, *last2;
char *s1 = s;
SV* must;
PL_multiline ? FBMrf_MULTILINE : 0
);
DEBUG_r(PerlIO_printf(Perl_debug_log,
- "%s anchored substr `%s%.*s%s'%s",
+ "%s anchored substr \"%s%.*s%s\"%s",
(s ? "Found" : "Contradicts"),
PL_colors[0],
(int)(SvCUR(must)
- (SvTAIL(must)!=0)),
- SvPVX(must),
+ SvPVX_const(must),
PL_colors[1], (SvTAIL(must) ? "$" : "")));
if (!s) {
if (last1 >= last2) {
(unsigned char*)last + SvCUR(must)
- (SvTAIL(must)!=0),
must, PL_multiline ? FBMrf_MULTILINE : 0);
- DEBUG_r(PerlIO_printf(Perl_debug_log, "%s floating substr `%s%.*s%s'%s",
+ /* FIXME - DEBUG_EXECUTE_r if that is merged to maint */
+ DEBUG_r(PerlIO_printf(Perl_debug_log, "%s floating substr \"%s%.*s%s\"%s",
(s ? "Found" : "Contradicts"),
PL_colors[0],
(int)(SvCUR(must) - (SvTAIL(must)!=0)),
- SvPVX(must),
+ SvPVX_const(must),
PL_colors[1], (SvTAIL(must) ? "$" : "")));
if (!s) {
if (last1 == last) {
regstclass does not come from lookahead... */
/* If regstclass takes bytelength more than 1: If charlength==1, OK.
This leaves EXACTF only, which is dealt with in find_byclass(). */
- U8* str = (U8*)STRING(prog->regstclass);
- int cl_l = (PL_regkind[(U8)OP(prog->regstclass)] == EXACT
- ? CHR_DIST(str+STR_LEN(prog->regstclass), str)
+ const U8* const str = (U8*)STRING(prog->regstclass);
+ const int cl_l = (PL_regkind[(U8)OP(prog->regstclass)] == EXACT
+ ? CHR_DIST((U8 *)str+STR_LEN(prog->regstclass),
+ (U8 *)str)
: 1);
- char *endpos = (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
+ const char * const endpos = (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
? HOP3c(s, (prog->minlen ? cl_l : 0), strend)
: (prog->float_substr || prog->float_utf8
? HOP3c(HOP3c(check_at, -start_shift, strbeg),
cl_l, strend)
: strend);
- char *startpos = strbeg;
t = s;
- if (prog->reganch & ROPT_UTF8) {
- PL_regdata = prog->data;
- PL_bostr = startpos;
- }
- S_cache_re(prog);
- s = find_byclass(prog, prog->regstclass, s, endpos, startpos, 1);
+ cache_re(prog);
+ s = find_byclass(prog, prog->regstclass, s, endpos, 1);
if (!s) {
#ifdef DEBUGGING
- char *what = 0;
+ const char *what = 0;
#endif
if (endpos == strend) {
DEBUG_r( PerlIO_printf(Perl_debug_log,
/* We know what class REx starts with. Try to find this position... */
STATIC char *
-S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, char *startpos, I32 norun)
+S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, const char *strend, I32 norun)
{
- I32 doevery = (prog->reganch & ROPT_SKIP) == 0;
+ const I32 doevery = (prog->reganch & ROPT_SKIP) == 0;
char *m;
STRLEN ln;
+ STRLEN lnc;
+ register STRLEN uskip;
unsigned int c1;
unsigned int c2;
char *e;
register I32 tmp = 1; /* Scratch variable? */
- register bool do_utf8 = PL_reg_match_utf8;
+ register const bool do_utf8 = PL_reg_match_utf8;
/* We know what class it must start with. */
switch (OP(c)) {
case ANYOF:
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if ((ANYOF_FLAGS(c) & ANYOF_UNICODE) ||
!UTF8_IS_INVARIANT((U8)s[0]) ?
reginclass(c, (U8*)s, 0, do_utf8) :
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
}
break;
case EXACTF:
- m = STRING(c);
- ln = STR_LEN(c);
+ m = STRING(c);
+ ln = STR_LEN(c); /* length to match in octets/bytes */
+ lnc = (I32) ln; /* length to match in characters */
if (UTF) {
STRLEN ulen1, ulen2;
- U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
- U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
+ U8 *sm = (U8 *) m;
+ U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
+ U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
+ const U32 uniflags = ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY;
to_utf8_lower((U8*)m, tmpbuf1, &ulen1);
to_utf8_upper((U8*)m, tmpbuf2, &ulen2);
- c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXLEN_UCLC,
- 0, ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
- c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXLEN_UCLC,
- 0, ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
+ c1 = utf8n_to_uvchr(tmpbuf1, UTF8_MAXBYTES_CASE,
+ 0, uniflags);
+ c2 = utf8n_to_uvchr(tmpbuf2, UTF8_MAXBYTES_CASE,
+ 0, uniflags);
+ lnc = 0;
+ while (sm < ((U8 *) m + ln)) {
+ lnc++;
+ sm += UTF8SKIP(sm);
+ }
}
else {
c1 = *(U8*)m;
}
goto do_exactf;
case EXACTFL:
- m = STRING(c);
- ln = STR_LEN(c);
+ m = STRING(c);
+ ln = STR_LEN(c);
+ lnc = (I32) ln;
c1 = *(U8*)m;
c2 = PL_fold_locale[c1];
do_exactf:
- e = HOP3c(strend, -(I32)ln, s);
+ e = HOP3c(strend, -((I32)lnc), s);
if (norun && e < s)
e = s; /* Due to minlen logic of intuit() */
if (do_utf8) {
UV c, f;
- U8 tmpbuf [UTF8_MAXLEN+1];
- U8 foldbuf[UTF8_MAXLEN_FOLD+1];
+ U8 tmpbuf [UTF8_MAXBYTES+1];
STRLEN len, foldlen;
-
+ const U32 uniflags = ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY;
if (c1 == c2) {
+ /* Upper and lower of 1st char are equal -
+ * probably not a "letter". */
while (s <= e) {
- c = utf8n_to_uvchr((U8*)s, UTF8_MAXLEN, &len,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
+ uniflags);
if ( c == c1
&& (ln == len ||
ibcmp_utf8(s, (char **)0, 0, do_utf8,
&& (norun || regtry(prog, s)) )
goto got_it;
else {
+ U8 foldbuf[UTF8_MAXBYTES_CASE+1];
uvchr_to_utf8(tmpbuf, c);
f = to_utf8_fold(tmpbuf, foldbuf, &foldlen);
if ( f != c
}
else {
while (s <= e) {
- c = utf8n_to_uvchr((U8*)s, UTF8_MAXLEN, &len,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ c = utf8n_to_uvchr((U8*)s, UTF8_MAXBYTES, &len,
+ uniflags);
/* Handle some of the three Greek sigmas cases.
* Note that not all the possible combinations
&& (norun || regtry(prog, s)) )
goto got_it;
else {
+ U8 foldbuf[UTF8_MAXBYTES_CASE+1];
uvchr_to_utf8(tmpbuf, c);
f = to_utf8_fold(tmpbuf, foldbuf, &foldlen);
if ( f != c
}
tmp = ((OP(c) == BOUND ?
isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
- LOAD_UTF8_CHARCLASS(alnum,"a");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_ALNUM();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (tmp == !(OP(c) == BOUND ?
swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
isALNUM_LC_utf8((U8*)s)))
if ((norun || regtry(prog, s)))
goto got_it;
}
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
}
tmp = ((OP(c) == NBOUND ?
isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0);
- LOAD_UTF8_CHARCLASS(alnum,"a");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_ALNUM();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (tmp == !(OP(c) == NBOUND ?
swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
isALNUM_LC_utf8((U8*)s)))
tmp = !tmp;
else if ((norun || regtry(prog, s)))
goto got_it;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case ALNUM:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(alnum,"a");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_ALNUM();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case ALNUML:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (isALNUM_LC_utf8((U8*)s)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case NALNUM:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(alnum,"a");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_ALNUM();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case NALNUML:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!isALNUM_LC_utf8((U8*)s)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case SPACE:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(space," ");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_SPACE();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case SPACEL:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (*s == ' ' || isSPACE_LC_utf8((U8*)s)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case NSPACE:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(space," ");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_SPACE();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!(*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8))) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case NSPACEL:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!(*s == ' ' || isSPACE_LC_utf8((U8*)s))) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case DIGIT:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(digit,"0");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_DIGIT();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (swash_fetch(PL_utf8_digit,(U8*)s, do_utf8)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case DIGITL:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (isDIGIT_LC_utf8((U8*)s)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
break;
case NDIGIT:
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(digit,"0");
- while (s < strend) {
+ LOAD_UTF8_CHARCLASS_DIGIT();
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!swash_fetch(PL_utf8_digit,(U8*)s, do_utf8)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
case NDIGITL:
PL_reg_flags |= RF_tainted;
if (do_utf8) {
- while (s < strend) {
+ while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!isDIGIT_LC_utf8((U8*)s)) {
if (tmp && (norun || regtry(prog, s)))
goto got_it;
}
else
tmp = 1;
- s += UTF8SKIP(s);
+ s += uskip;
}
}
else {
register char *startpos = stringarg;
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);
- bool do_utf8 = DO_UTF8(sv);
+ const bool do_utf8 = DO_UTF8(sv);
#ifdef DEBUGGING
SV *dsv0 = PERL_DEBUG_PAD_ZERO(0);
SV *dsv1 = PERL_DEBUG_PAD_ZERO(1);
#endif
+ PERL_UNUSED_ARG(data);
RX_MATCH_UTF8_set(prog,do_utf8);
PL_regcc = 0;
}
DEBUG_r({
- char *s0 = UTF ?
- pv_uni_display(dsv0, (U8*)prog->precomp, prog->prelen, 60,
- UNI_DISPLAY_REGEX) :
- prog->precomp;
- int len0 = UTF ? SvCUR(dsv0) : prog->prelen;
- char *s1 = do_utf8 ? sv_uni_display(dsv1, sv, 60,
+ const char * const s0 = UTF
+ ? pv_uni_display(dsv0, (U8*)prog->precomp, prog->prelen, 60,
+ UNI_DISPLAY_REGEX)
+ : prog->precomp;
+ const int len0 = UTF ? SvCUR(dsv0) : prog->prelen;
+ const char * const s1 = do_utf8 ? sv_uni_display(dsv1, sv, 60,
UNI_DISPLAY_REGEX) : startpos;
- int len1 = do_utf8 ? SvCUR(dsv1) : strend - startpos;
+ const int len1 = do_utf8 ? SvCUR(dsv1) : strend - startpos;
if (!PL_colorset)
reginitcolors();
PerlIO_printf(Perl_debug_log,
- "%sMatching REx%s `%s%*.*s%s%s' against `%s%.*s%s%s'\n",
+ "%sMatching REx%s \"%s%*.*s%s%s\" against \"%s%.*s%s%s\"\n",
PL_colors[4],PL_colors[5],PL_colors[0],
len0, len0, s0,
PL_colors[1],
#endif
if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr))
do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog);
- ch = SvPVX(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0];
+ ch = SvPVX_const(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0];
if (do_utf8) {
while (s < strend) {
"Did not find anchored character...\n")
);
}
- /*SUPPRESS 560*/
else if (prog->anchored_substr != Nullsv
|| prog->anchored_utf8 != Nullsv
|| ((prog->float_substr != Nullsv || prog->float_utf8 != Nullsv)
else
last1 = s - 1; /* bogus */
- /* XXXX check_substr already used to find `s', can optimize if
+ /* XXXX check_substr already used to find "s", can optimize if
check_substr==must. */
scream_pos = -1;
dontbother = end_shift;
PL_multiline ? FBMrf_MULTILINE : 0))) ) {
/* we may be pointing at the wrong string */
if ((flags & REXEC_SCREAM) && RX_MATCH_COPIED(prog))
- s = prog->subbeg + (s - SvPVX(sv));
+ s = strbeg + (s - SvPVX_const(sv));
DEBUG_r( did_match = 1 );
if (HOPc(s, -back_max) > last1) {
last1 = HOPc(s, -back_min);
}
DEBUG_r(if (!did_match)
PerlIO_printf(Perl_debug_log,
- "Did not find %s substr `%s%.*s%s'%s...\n",
+ "Did not find %s substr \"%s%.*s%s\"%s...\n",
((must == prog->anchored_substr || must == prog->anchored_utf8)
? "anchored" : "floating"),
PL_colors[0],
(int)(SvCUR(must) - (SvTAIL(must)!=0)),
- SvPVX(must),
+ SvPVX_const(must),
PL_colors[1], (SvTAIL(must) ? "$" : ""))
);
goto phooey;
}
DEBUG_r({
SV *prop = sv_newmortal();
- char *s0;
- char *s1;
+ const char *s0;
+ const char *s1;
int len0;
int len1;
regprop(prop, c);
s0 = UTF ?
- pv_uni_display(dsv0, (U8*)SvPVX(prop), SvCUR(prop), 60,
+ pv_uni_display(dsv0, (U8*)SvPVX_const(prop), SvCUR(prop), 60,
UNI_DISPLAY_REGEX) :
- SvPVX(prop);
+ SvPVX_const(prop);
len0 = UTF ? SvCUR(dsv0) : SvCUR(prop);
s1 = UTF ?
sv_uni_display(dsv1, sv, 60, UNI_DISPLAY_REGEX) : s;
len1 = UTF ? SvCUR(dsv1) : strend - s;
PerlIO_printf(Perl_debug_log,
- "Matching stclass `%*.*s' against `%*.*s'\n",
+ "Matching stclass \"%*.*s\" against \"%*.*s\"\n",
len0, len0, s0,
len1, len1, s1);
});
- if (find_byclass(prog, c, s, strend, startpos, 0))
+ if (find_byclass(prog, c, s, strend, 0))
goto got_it;
DEBUG_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass...\n"));
}
last = scream_olds; /* Only one occurrence. */
/* we may be pointing at the wrong string */
else if (RX_MATCH_COPIED(prog))
- s = prog->subbeg + (s - SvPVX(sv));
+ s = strbeg + (s - SvPVX_const(sv));
}
else {
STRLEN len;
- char *little = SvPV(float_real, len);
+ const char * const little = SvPV_const(float_real, len);
if (SvTAIL(float_real)) {
if (memEQ(strend - len + 1, little, len - 1))
if (len)
last = rninstr(s, strend, little, little + len);
else
- last = strend; /* matching `$' */
+ last = strend; /* matching "$" */
}
}
if (last == NULL) {
SAVEDESTRUCTOR_X(restore_pos, 0);
}
if (!PL_reg_curpm) {
- Newz(22,PL_reg_curpm, 1, PMOP);
+ Newxz(PL_reg_curpm, 1, PMOP);
#ifdef USE_ITHREADS
{
SV* repointer = newSViv(0);
if(PL_reg_start_tmp)
Renew(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
else
- New(22,PL_reg_start_tmp, PL_reg_start_tmpl, char*);
+ Newx(PL_reg_start_tmp, PL_reg_start_tmpl, char*);
}
/* XXXX What this code is doing here?!!! There should be no need
#define sayNO_SILENT goto do_no
#define saySAME(x) if (x) goto yes; else goto no
+#define POSCACHE_SUCCESS 0 /* caching success rather than failure */
+#define POSCACHE_SEEN 1 /* we know what we're caching */
+#define POSCACHE_START 2 /* the real cache: this bit maps to pos 0 */
+#define CACHEsayYES STMT_START { \
+ if (cache_offset | cache_bit) { \
+ if (!(PL_reg_poscache[0] & (1<<POSCACHE_SEEN))) \
+ PL_reg_poscache[0] |= (1<<POSCACHE_SUCCESS) || (1<<POSCACHE_SEEN); \
+ else if (!(PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS))) { \
+ /* cache records failure, but this is success */ \
+ DEBUG_r( \
+ PerlIO_printf(Perl_debug_log, \
+ "%*s (remove success from failure cache)\n", \
+ REPORT_CODE_OFF+PL_regindent*2, "") \
+ ); \
+ PL_reg_poscache[cache_offset] &= ~(1<<cache_bit); \
+ } \
+ } \
+ sayYES; \
+} STMT_END
+#define CACHEsayNO STMT_START { \
+ if (cache_offset | cache_bit) { \
+ if (!(PL_reg_poscache[0] & (1<<POSCACHE_SEEN))) \
+ PL_reg_poscache[0] |= (1<<POSCACHE_SEEN); \
+ else if ((PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS))) { \
+ /* cache records success, but this is failure */ \
+ DEBUG_r( \
+ PerlIO_printf(Perl_debug_log, \
+ "%*s (remove failure from success cache)\n", \
+ REPORT_CODE_OFF+PL_regindent*2, "") \
+ ); \
+ PL_reg_poscache[cache_offset] &= ~(1<<cache_bit); \
+ } \
+ } \
+ sayNO; \
+} STMT_END
+
#define REPORT_CODE_OFF 24
/*
#if 0
I32 firstcp = PL_savestack_ix;
#endif
- register bool do_utf8 = PL_reg_match_utf8;
+ register const bool do_utf8 = PL_reg_match_utf8;
#ifdef DEBUGGING
SV *dsv0 = PERL_DEBUG_PAD_ZERO(0);
SV *dsv1 = PERL_DEBUG_PAD_ZERO(1);
SV *dsv2 = PERL_DEBUG_PAD_ZERO(2);
#endif
+ U32 uniflags = ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY;
#ifdef DEBUGGING
PL_regindent++;
DEBUG_r( {
SV *prop = sv_newmortal();
- int docolor = *PL_colors[0];
- int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */
+ const int docolor = *PL_colors[0];
+ const int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */
int l = (PL_regeol - locinput) > taill ? taill : (PL_regeol - locinput);
/* The part of the string before starttry has one color
(pref0_len chars), between starttry and current
pref0_len = pref_len;
regprop(prop, scan);
{
- char *s0 =
+ const char * const s0 =
do_utf8 && OP(scan) != CANY ?
pv_uni_display(dsv0, (U8*)(locinput - pref_len),
pref0_len, 60, UNI_DISPLAY_REGEX) :
locinput - pref_len;
- int len0 = do_utf8 ? strlen(s0) : pref0_len;
- char *s1 = do_utf8 && OP(scan) != CANY ?
+ const int len0 = do_utf8 ? strlen(s0) : pref0_len;
+ const char * const s1 = do_utf8 && OP(scan) != CANY ?
pv_uni_display(dsv1, (U8*)(locinput - pref_len + pref0_len),
pref_len - pref0_len, 60, UNI_DISPLAY_REGEX) :
locinput - pref_len + pref0_len;
- int len1 = do_utf8 ? strlen(s1) : pref_len - pref0_len;
- char *s2 = do_utf8 && OP(scan) != CANY ?
+ const int len1 = do_utf8 ? strlen(s1) : pref_len - pref0_len;
+ const char * const s2 = do_utf8 && OP(scan) != CANY ?
pv_uni_display(dsv2, (U8*)locinput,
PL_regeol - locinput, 60, UNI_DISPLAY_REGEX) :
locinput;
- int len2 = do_utf8 ? strlen(s2) : l;
+ const int len2 = do_utf8 ? strlen(s2) : l;
PerlIO_printf(Perl_debug_log,
"%4"IVdf" <%s%.*s%s%s%.*s%s%s%s%.*s%s>%*s|%3"IVdf":%*s%s\n",
(IV)(locinput - PL_bostr),
15 - l - pref_len + 1,
"",
(IV)(scan - PL_regprogram), PL_regindent*2, "",
- SvPVX(prop));
+ SvPVX_const(prop));
}
});
if (do_utf8 != UTF) {
/* The target and the pattern have differing utf8ness. */
char *l = locinput;
- char *e = s + ln;
- STRLEN ulen;
+ const char *e = s + ln;
if (do_utf8) {
/* The target is utf8, the pattern is not utf8. */
while (s < e) {
+ STRLEN ulen;
if (l >= PL_regeol)
sayNO;
if (NATIVE_TO_UNI(*(U8*)s) !=
- utf8n_to_uvuni((U8*)l, UTF8_MAXLEN, &ulen,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY))
+ utf8n_to_uvuni((U8*)l, UTF8_MAXBYTES, &ulen,
+ uniflags))
sayNO;
l += ulen;
s ++;
else {
/* The target is not utf8, the pattern is utf8. */
while (s < e) {
+ STRLEN ulen;
if (l >= PL_regeol)
sayNO;
if (NATIVE_TO_UNI(*((U8*)l)) !=
- utf8n_to_uvuni((U8*)s, UTF8_MAXLEN, &ulen,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY))
+ utf8n_to_uvuni((U8*)s, UTF8_MAXBYTES, &ulen,
+ uniflags))
sayNO;
s += ulen;
l ++;
if (!nextchr)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(alnum,"a");
+ LOAD_UTF8_CHARCLASS_ALNUM();
if (!(OP(scan) == ALNUM
? swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
: isALNUM_LC_utf8((U8*)locinput)))
if (!nextchr && locinput >= PL_regeol)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(alnum,"a");
+ LOAD_UTF8_CHARCLASS_ALNUM();
if (OP(scan) == NALNUM
? swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8)
: isALNUM_LC_utf8((U8*)locinput))
if (locinput == PL_bostr)
ln = '\n';
else {
- U8 *r = reghop3((U8*)locinput, -1, (U8*)PL_bostr);
+ const U8 * const r = reghop3((U8*)locinput, -1, (U8*)PL_bostr);
- ln = utf8n_to_uvchr(r, UTF8SKIP(r), 0, 0);
+ ln = utf8n_to_uvchr((U8 *)r, UTF8SKIP(r), 0, 0);
}
if (OP(scan) == BOUND || OP(scan) == NBOUND) {
ln = isALNUM_uni(ln);
- LOAD_UTF8_CHARCLASS(alnum,"a");
+ LOAD_UTF8_CHARCLASS_ALNUM();
n = swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8);
}
else {
sayNO;
if (do_utf8) {
if (UTF8_IS_CONTINUED(nextchr)) {
- LOAD_UTF8_CHARCLASS(space," ");
+ LOAD_UTF8_CHARCLASS_SPACE();
if (!(OP(scan) == SPACE
? swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
: isSPACE_LC_utf8((U8*)locinput)))
if (!nextchr && locinput >= PL_regeol)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(space," ");
+ LOAD_UTF8_CHARCLASS_SPACE();
if (OP(scan) == NSPACE
? swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8)
: isSPACE_LC_utf8((U8*)locinput))
if (!nextchr)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(digit,"0");
+ LOAD_UTF8_CHARCLASS_DIGIT();
if (!(OP(scan) == DIGIT
? swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
: isDIGIT_LC_utf8((U8*)locinput)))
if (!nextchr && locinput >= PL_regeol)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(digit,"0");
+ LOAD_UTF8_CHARCLASS_DIGIT();
if (OP(scan) == NDIGIT
? swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8)
: isDIGIT_LC_utf8((U8*)locinput))
if (locinput >= PL_regeol)
sayNO;
if (do_utf8) {
- LOAD_UTF8_CHARCLASS(mark,"~");
+ LOAD_UTF8_CHARCLASS_MARK();
if (swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8))
sayNO;
locinput += PL_utf8skip[nextchr];
s = PL_bostr + ln;
if (do_utf8 && OP(scan) != REF) { /* REF can do byte comparison */
char *l = locinput;
- char *e = PL_bostr + PL_regendp[n];
+ const char *e = PL_bostr + PL_regendp[n];
/*
* Note that we can't do the "other character" lookup trick as
* in the 8-bit case (no pun intended) because in Unicode we
* have to map both upper and title case to lower case.
*/
if (OP(scan) == REFF) {
- STRLEN ulen1, ulen2;
- U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
- U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
while (s < e) {
+ STRLEN ulen1, ulen2;
+ U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
+ U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
+
if (l >= PL_regeol)
sayNO;
toLOWER_utf8((U8*)s, tmpbuf1, &ulen1);
COP *ocurcop = PL_curcop;
PAD *old_comppad;
SV *ret;
+ struct regexp *oreg = PL_reg_re;
n = ARG(scan);
PL_op = (OP_4tree*)PL_regdata->data[n];
re_cc_state state;
CHECKPOINT cp, lastcp;
int toggleutf;
+ register SV *sv;
- if(SvROK(ret) || SvRMAGICAL(ret)) {
- SV *sv = SvROK(ret) ? SvRV(ret) : ret;
-
- if(SvMAGICAL(sv))
- mg = mg_find(sv, PERL_MAGIC_qr);
+ if(SvROK(ret) && SvSMAGICAL(sv = SvRV(ret)))
+ mg = mg_find(sv, PERL_MAGIC_qr);
+ else if (SvSMAGICAL(ret)) {
+ if (SvGMAGICAL(ret))
+ sv_unmagic(ret, PERL_MAGIC_qr);
+ else
+ mg = mg_find(ret, PERL_MAGIC_qr);
}
+
if (mg) {
re = (regexp *)mg->mg_obj;
(void)ReREFCNT_inc(re);
}
else {
STRLEN len;
- char *t = SvPV(ret, len);
+ const char *t = SvPV_const(ret, len);
PMOP pm;
- char *oprecomp = PL_regprecomp;
- I32 osize = PL_regsize;
- I32 onpar = PL_regnpar;
+ char * const oprecomp = PL_regprecomp;
+ const I32 osize = PL_regsize;
+ const I32 onpar = PL_regnpar;
Zero(&pm, 1, PMOP);
if (DO_UTF8(ret)) pm.op_pmdynflags |= PMdf_DYN_UTF8;
- re = CALLREGCOMP(aTHX_ t, t + len, &pm);
+ re = CALLREGCOMP(aTHX_ (char*)t, (char*)t + len, &pm);
if (!(SvFLAGS(ret)
- & (SVs_TEMP | SVs_PADTMP | SVf_READONLY)))
+ & (SVs_TEMP | SVs_PADTMP | SVf_READONLY
+ | SVs_GMG)))
sv_magic(ret,(SV*)ReREFCNT_inc(re),
PERL_MAGIC_qr,0,0);
PL_regprecomp = oprecomp;
}
DEBUG_r(
PerlIO_printf(Perl_debug_log,
- "Entering embedded `%s%.60s%s%s'\n",
+ "Entering embedded \"%s%.60s%s%s\"\n",
PL_colors[0],
re->precomp,
PL_colors[1],
sw = SvTRUE(ret);
logical = 0;
}
- else
+ else {
sv_setsv(save_scalar(PL_replgv), ret);
+ cache_re(oreg);
+ }
break;
}
case OPEN:
CHECKPOINT cp, lastcp;
CURCUR* cc = PL_regcc;
char *lastloc = cc->lastloc; /* Detection of 0-len. */
+ I32 cache_offset = 0, cache_bit = 0;
n = cc->cur + 1; /* how many we know we matched */
PL_reginput = locinput;
PL_reg_leftiter = PL_reg_maxiter;
}
if (PL_reg_leftiter-- == 0) {
- I32 size = (PL_reg_maxiter + 7)/8;
+ const I32 size = (PL_reg_maxiter + 7 + POSCACHE_START)/8;
if (PL_reg_poscache) {
if ((I32)PL_reg_poscache_size < size) {
Renew(PL_reg_poscache, size, char);
}
else {
PL_reg_poscache_size = size;
- Newz(29, PL_reg_poscache, size, char);
+ Newxz(PL_reg_poscache, size, char);
}
DEBUG_r(
PerlIO_printf(Perl_debug_log,
);
}
if (PL_reg_leftiter < 0) {
- I32 o = locinput - PL_bostr, b;
+ cache_offset = locinput - PL_bostr;
- o = (scan->flags & 0xf) - 1 + o * (scan->flags>>4);
- b = o % 8;
- o /= 8;
- if (PL_reg_poscache[o] & (1<<b)) {
+ cache_offset = (scan->flags & 0xf) - 1 + POSCACHE_START
+ + cache_offset * (scan->flags>>4);
+ cache_bit = cache_offset % 8;
+ cache_offset /= 8;
+ if (PL_reg_poscache[cache_offset] & (1<<cache_bit)) {
DEBUG_r(
PerlIO_printf(Perl_debug_log,
"%*s already tried at this position...\n",
REPORT_CODE_OFF+PL_regindent*2, "")
);
- sayNO_SILENT;
+ if (PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS))
+ /* cache records success */
+ sayYES;
+ else
+ /* cache records failure */
+ sayNO_SILENT;
}
- PL_reg_poscache[o] |= (1<<b);
+ PL_reg_poscache[cache_offset] |= (1<<cache_bit);
}
}
REGCP_SET(lastcp);
if (regmatch(cc->next)) {
regcpblow(cp);
- sayYES; /* All done. */
+ CACHEsayYES; /* All done. */
}
REGCP_UNWIND(lastcp);
regcppop();
"Complex regular subexpression recursion",
REG_INFTY - 1);
}
- sayNO;
+ CACHEsayNO;
}
DEBUG_r(
REGCP_SET(lastcp);
if (regmatch(cc->scan)) {
regcpblow(cp);
- sayYES;
+ CACHEsayYES;
}
REGCP_UNWIND(lastcp);
regcppop();
cc->cur = n - 1;
cc->lastloc = lastloc;
- sayNO;
+ CACHEsayNO;
}
/* Prefer scan over next for maximal matching. */
REGCP_SET(lastcp);
if (regmatch(cc->scan)) {
regcpblow(cp);
- sayYES;
+ CACHEsayYES;
}
REGCP_UNWIND(lastcp);
regcppop(); /* Restore some previous $<digit>s? */
if (PL_regcc)
ln = PL_regcc->cur;
if (regmatch(cc->next))
- sayYES;
+ CACHEsayYES;
if (PL_regcc)
PL_regcc->cur = ln;
PL_regcc = cc;
cc->cur = n - 1;
cc->lastloc = lastloc;
- sayNO;
+ CACHEsayNO;
}
/* NOT REACHED */
case BRANCHJ:
if (OP(next) != c1) /* No choice. */
next = inner; /* Avoid recursion. */
else {
- I32 lastparen = *PL_reglastparen;
+ const I32 lastparen = *PL_reglastparen;
I32 unwind1;
re_unwind_branch_t *uw;
CHECKPOINT lastcp;
/* We suppose that the next guy does not need
- backtracking: in particular, it is of constant length,
+ backtracking: in particular, it is of constant non-zero length,
and has no parenths to influence future backrefs. */
ln = ARG1(scan); /* min to match */
n = ARG2(scan); /* max to match */
minmod = 0;
if (ln && regrepeat_hard(scan, ln, &l) < ln)
sayNO;
- /* if we matched something zero-length we don't need to
- backtrack - capturing parens are already defined, so
- the caveat in the maximal case doesn't apply
-
- XXXX if ln == 0, we can redo this check first time
- through the following loop
- */
- if (ln && l == 0)
- n = ln; /* don't backtrack */
locinput = PL_reginput;
if (HAS_TEXT(next) || JUMPABLE(next)) {
regnode *text_node = next;
if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
if (PL_regkind[(U8)OP(text_node)] == REF) {
- I32 n, ln;
- n = ARG(text_node); /* which paren pair */
- ln = PL_regstartp[n];
- /* assume yes if we haven't seen CLOSEn */
- if (
- (I32)*PL_reglastparen < n ||
- ln == -1 ||
- ln == PL_regendp[n]
- ) {
- c1 = c2 = -1000;
- goto assume_ok_MM;
- }
- c1 = *(PL_bostr + ln);
+ c1 = c2 = -1000;
+ goto assume_ok_MM;
}
else { c1 = (U8)*STRING(text_node); }
if (OP(text_node) == EXACTF || OP(text_node) == REFF)
c1 = c2 = -1000;
assume_ok_MM:
REGCP_SET(lastcp);
- /* This may be improved if l == 0. */
- while (n >= ln || (n == REG_INFTY && ln > 0 && l)) { /* ln overflow ? */
+ while (n >= ln || (n == REG_INFTY && ln > 0)) { /* ln overflow ? */
/* If it could work, try it. */
if (c1 == -1000 ||
UCHARAT(PL_reginput) == c1 ||
}
else {
n = regrepeat_hard(scan, n, &l);
- /* if we matched something zero-length we don't need to
- backtrack, unless the minimum count is zero and we
- are capturing the result - in that case the capture
- being defined or not may affect later execution
- */
- if (n != 0 && l == 0 && !(paren && ln == 0))
- ln = n; /* don't backtrack */
locinput = PL_reginput;
DEBUG_r(
PerlIO_printf(Perl_debug_log,
if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
if (PL_regkind[(U8)OP(text_node)] == REF) {
- I32 n, ln;
- n = ARG(text_node); /* which paren pair */
- ln = PL_regstartp[n];
- /* assume yes if we haven't seen CLOSEn */
- if (
- (I32)*PL_reglastparen < n ||
- ln == -1 ||
- ln == PL_regendp[n]
- ) {
- c1 = c2 = -1000;
- goto assume_ok_REG;
- }
- c1 = *(PL_bostr + ln);
+ c1 = c2 = -1000;
+ goto assume_ok_REG;
}
else { c1 = (U8)*STRING(text_node); }
if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
if (PL_regkind[(U8)OP(text_node)] == REF) {
- I32 n, ln;
- n = ARG(text_node); /* which paren pair */
- ln = PL_regstartp[n];
- /* assume yes if we haven't seen CLOSEn */
- if (
- (I32)*PL_reglastparen < n ||
- ln == -1 ||
- ln == PL_regendp[n]
- ) {
- c1 = c2 = -1000;
- goto assume_ok_easy;
- }
- s = (U8*)PL_bostr + ln;
+ c1 = c2 = -1000;
+ goto assume_ok_easy;
}
else { s = (U8*)STRING(text_node); }
else { /* UTF */
if (OP(text_node) == EXACTF || OP(text_node) == REFF) {
STRLEN ulen1, ulen2;
- U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
- U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
+ U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
+ U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
to_utf8_lower((U8*)s, tmpbuf1, &ulen1);
to_utf8_upper((U8*)s, tmpbuf2, &ulen2);
- c1 = utf8n_to_uvuni(tmpbuf1, UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
- c2 = utf8n_to_uvuni(tmpbuf2, UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ c1 = utf8n_to_uvuni(tmpbuf1, UTF8_MAXBYTES, 0,
+ uniflags);
+ c2 = utf8n_to_uvuni(tmpbuf2, UTF8_MAXBYTES, 0,
+ uniflags);
}
else {
- c2 = c1 = utf8n_to_uvchr(s, UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ c2 = c1 = utf8n_to_uvchr(s, UTF8_MAXBYTES, 0,
+ uniflags);
}
}
}
count = locinput - old;
}
else {
- STRLEN len;
if (c1 == c2) {
+ STRLEN len;
/* count initialised to
* utf8_distance(old, locinput) */
while (locinput <= e &&
utf8n_to_uvchr((U8*)locinput,
- UTF8_MAXLEN, &len,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY) != (UV)c1) {
+ UTF8_MAXBYTES, &len,
+ uniflags) != (UV)c1) {
locinput += len;
count++;
}
} else {
+ STRLEN len;
/* count initialised to
* utf8_distance(old, locinput) */
while (locinput <= e) {
UV c = utf8n_to_uvchr((U8*)locinput,
- UTF8_MAXLEN, &len,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ UTF8_MAXBYTES, &len,
+ uniflags);
if (c == (UV)c1 || c == (UV)c2)
break;
locinput += len;
if (c1 != -1000) {
if (do_utf8)
c = utf8n_to_uvchr((U8*)PL_reginput,
- UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ UTF8_MAXBYTES, 0,
+ uniflags);
else
c = UCHARAT(PL_reginput);
/* If it could work, try it. */
if (c1 != -1000) {
if (do_utf8)
c = utf8n_to_uvchr((U8*)PL_reginput,
- UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ UTF8_MAXBYTES, 0,
+ uniflags);
else
c = UCHARAT(PL_reginput);
}
if (c1 != -1000) {
if (do_utf8)
c = utf8n_to_uvchr((U8*)PL_reginput,
- UTF8_MAXLEN, 0,
- ckWARN(WARN_UTF8) ?
- 0 : UTF8_ALLOW_ANY);
+ UTF8_MAXBYTES, 0,
+ uniflags);
else
c = UCHARAT(PL_reginput);
}
case RE_UNWIND_BRANCHJ:
{
re_unwind_branch_t *uwb = &(uw->branch);
- I32 lastparen = uwb->lastparen;
+ const I32 lastparen = uwb->lastparen;
REGCP_UNWIND(uwb->lastcp);
for (n = *PL_reglastparen; n > lastparen; n--)
goto do_no;
}
/* Have more choice yet. Reuse the same uwb. */
- /*SUPPRESS 560*/
if ((n = (uwb->type == RE_UNWIND_BRANCH
? NEXT_OFF(next) : ARG(next))))
next += n;
* rather than incrementing count on every character. [Er, except utf8.]]
*/
STATIC I32
-S_regrepeat(pTHX_ regnode *p, I32 max)
+S_regrepeat(pTHX_ const regnode *p, I32 max)
{
register char *scan;
register I32 c;
case ALNUM:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(alnum,"a");
+ LOAD_UTF8_CHARCLASS_ALNUM();
while (hardcount < max && scan < loceol &&
swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
scan += UTF8SKIP(scan);
case NALNUM:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(alnum,"a");
+ LOAD_UTF8_CHARCLASS_ALNUM();
while (hardcount < max && scan < loceol &&
!swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) {
scan += UTF8SKIP(scan);
case SPACE:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(space," ");
+ LOAD_UTF8_CHARCLASS_SPACE();
while (hardcount < max && scan < loceol &&
(*scan == ' ' ||
swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
case NSPACE:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(space," ");
+ LOAD_UTF8_CHARCLASS_SPACE();
while (hardcount < max && scan < loceol &&
!(*scan == ' ' ||
swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) {
case DIGIT:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(digit,"0");
+ LOAD_UTF8_CHARCLASS_DIGIT();
while (hardcount < max && scan < loceol &&
swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
scan += UTF8SKIP(scan);
case NDIGIT:
if (do_utf8) {
loceol = PL_regeol;
- LOAD_UTF8_CHARCLASS(digit,"0");
+ LOAD_UTF8_CHARCLASS_DIGIT();
while (hardcount < max && scan < loceol &&
!swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) {
scan += UTF8SKIP(scan);
{
SV *prop = sv_newmortal();
- regprop(prop, p);
+ regprop(prop, (regnode *)p);
PerlIO_printf(Perl_debug_log,
"%*s %s can match %"IVdf" times out of %"IVdf"...\n",
- REPORT_CODE_OFF+1, "", SvPVX(prop),(IV)c,(IV)max);
+ REPORT_CODE_OFF+1, "", SvPVX_const(prop),(IV)c,(IV)max);
});
return(c);
/*
- regrepeat_hard - repeatedly match something, report total lenth and length
*
- * The repeater is supposed to have constant length.
+ * The repeater is supposed to have constant non-zero length.
*/
STATIC I32
SV *alt = NULL;
if (PL_regdata && PL_regdata->count) {
- U32 n = ARG(node);
+ const U32 n = ARG(node);
if (PL_regdata->what[n] == 's') {
- SV *rv = (SV*)PL_regdata->data[n];
- AV *av = (AV*)SvRV((SV*)rv);
- SV **ary = AvARRAY(av);
+ SV * const rv = (SV*)PL_regdata->data[n];
+ AV * const av = (AV*)SvRV((SV*)rv);
+ SV **const ary = AvARRAY(av);
SV **a, **b;
/* See the end of regcomp.c:S_reglass() for
* documentation of these array elements. */
si = *ary;
- a = SvTYPE(ary[1]) == SVt_RV ? &ary[1] : 0;
+ a = SvROK(ary[1]) ? &ary[1] : 0;
b = SvTYPE(ary[2]) == SVt_PVAV ? &ary[2] : 0;
if (a)
*/
STATIC bool
-S_reginclass(pTHX_ register regnode *n, register U8* p, STRLEN* lenp, register bool do_utf8)
+S_reginclass(pTHX_ register const regnode *n, register const U8* p, STRLEN* lenp, register bool do_utf8)
{
- char flags = ANYOF_FLAGS(n);
+ const char flags = ANYOF_FLAGS(n);
bool match = FALSE;
UV c = *p;
STRLEN len = 0;
STRLEN plen;
- if (do_utf8 && !UTF8_IS_INVARIANT(c))
- c = utf8n_to_uvchr(p, UTF8_MAXLEN, &len,
- ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
+ if (do_utf8 && !UTF8_IS_INVARIANT(c)) {
+ c = utf8n_to_uvchr((U8 *)p, UTF8_MAXBYTES, &len,
+ ckWARN(WARN_UTF8) ? UTF8_CHECK_ONLY :
+ UTF8_ALLOW_ANYUV|UTF8_CHECK_ONLY);
+ if (len == (STRLEN)-1)
+ Perl_croak(aTHX_ "Malformed UTF-8 character (fatal)");
+ }
plen = lenp ? *lenp : UNISKIP(NATIVE_TO_UNI(c));
if (do_utf8 || (flags & ANYOF_UNICODE)) {
match = TRUE;
if (!match) {
AV *av;
- SV *sw = regclass_swash(n, TRUE, 0, (SV**)&av);
+ SV * const sw = regclass_swash((regnode *)n, TRUE, 0, (SV**)&av);
if (sw) {
- if (swash_fetch(sw, p, do_utf8))
+ if (swash_fetch(sw, (U8 *)p, do_utf8))
match = TRUE;
else if (flags & ANYOF_FOLD) {
if (!match && lenp && av) {
I32 i;
-
for (i = 0; i <= av_len(av); i++) {
- SV* sv = *av_fetch(av, i, FALSE);
+ SV* const sv = *av_fetch(av, i, FALSE);
STRLEN len;
- char *s = SvPV(sv, len);
+ const char * const s = SvPV_const(sv, len);
if (len <= plen && memEQ(s, (char*)p, len)) {
*lenp = len;
}
}
if (!match) {
- U8 tmpbuf[UTF8_MAXLEN_FOLD+1];
+ U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
STRLEN tmplen;
- to_utf8_fold(p, tmpbuf, &tmplen);
+ to_utf8_fold((U8 *)p, tmpbuf, &tmplen);
if (swash_fetch(sw, tmpbuf, do_utf8))
match = TRUE;
}
static void
restore_pos(pTHX_ void *arg)
{
+ PERL_UNUSED_ARG(arg);
if (PL_reg_eval_set) {
if (PL_reg_oldsaved) {
PL_reg_re->subbeg = PL_reg_oldsaved;
STATIC void
S_to_utf8_substr(pTHX_ register regexp *prog)
{
- SV* sv;
if (prog->float_substr && !prog->float_utf8) {
- prog->float_utf8 = sv = NEWSV(117, 0);
- SvSetSV(sv, prog->float_substr);
+ SV* sv;
+ prog->float_utf8 = sv = newSVsv(prog->float_substr);
sv_utf8_upgrade(sv);
if (SvTAIL(prog->float_substr))
SvTAIL_on(sv);
prog->check_utf8 = sv;
}
if (prog->anchored_substr && !prog->anchored_utf8) {
- prog->anchored_utf8 = sv = NEWSV(118, 0);
- SvSetSV(sv, prog->anchored_substr);
+ SV* sv;
+ prog->anchored_utf8 = sv = newSVsv(prog->anchored_substr);
sv_utf8_upgrade(sv);
if (SvTAIL(prog->anchored_substr))
SvTAIL_on(sv);
STATIC void
S_to_byte_substr(pTHX_ register regexp *prog)
{
- SV* sv;
if (prog->float_utf8 && !prog->float_substr) {
- prog->float_substr = sv = NEWSV(117, 0);
- SvSetSV(sv, prog->float_utf8);
+ SV* sv;
+ prog->float_substr = sv = newSVsv(prog->float_utf8);
if (sv_utf8_downgrade(sv, TRUE)) {
if (SvTAIL(prog->float_utf8))
SvTAIL_on(sv);
prog->check_substr = sv;
}
if (prog->anchored_utf8 && !prog->anchored_substr) {
- prog->anchored_substr = sv = NEWSV(118, 0);
- SvSetSV(sv, prog->anchored_utf8);
+ SV* sv;
+ prog->anchored_substr = sv = newSVsv(prog->anchored_utf8);
if (sv_utf8_downgrade(sv, TRUE)) {
if (SvTAIL(prog->anchored_utf8))
SvTAIL_on(sv);
prog->check_substr = sv;
}
}
+
+/*
+ * Local variables:
+ * c-indentation-style: bsd
+ * c-basic-offset: 4
+ * indent-tabs-mode: t
+ * End:
+ *
+ * ex: set ts=8 sts=4 sw=4 noet:
+ */
https://perl5.git.perl.org / perl5.git / blobdiff