: strend);
t = s;
- s = find_byclass(prog, prog->regstclass, s, endpos, 1);
+ s = find_byclass(prog, prog->regstclass, s, endpos, NULL);
if (!s) {
#ifdef DEBUGGING
const char *what = NULL;
}
/* We know what class REx starts with. Try to find this position... */
+/* if reginfo is NULL, its a dryrun */
+
STATIC char *
-S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s, const char *strend, I32 norun)
+S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s, const char
+*strend, const regmatch_info *reginfo)
{
dVAR;
const I32 doevery = (prog->reganch & ROPT_SKIP) == 0;
!UTF8_IS_INVARIANT((U8)s[0]) ?
reginclass(prog, c, (U8*)s, 0, do_utf8) :
REGINCLASS(prog, c, (U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
/* The assignment of 2 is intentional:
* for the folded sharp s, the skip is 2. */
(skip = SHARP_S_SKIP))) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
break;
case CANY:
while (s < strend) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
do_exactf:
e = HOP3c(strend, -((I32)lnc), s);
- if (norun && e < s)
+ if (!reginfo && e < s)
e = s; /* Due to minlen logic of intuit() */
/* The idea in the EXACTF* cases is to first find the
&& (ln == len ||
ibcmp_utf8(s, (char **)0, 0, do_utf8,
m, (char **)0, ln, (bool)UTF))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
else {
U8 foldbuf[UTF8_MAXBYTES_CASE+1];
(char **)0, foldlen, do_utf8,
m,
(char **)0, ln, (bool)UTF))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
}
s += len;
&& (ln == len ||
ibcmp_utf8(s, (char **)0, 0, do_utf8,
m, (char **)0, ln, (bool)UTF))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
else {
U8 foldbuf[UTF8_MAXBYTES_CASE+1];
(char **)0, foldlen, do_utf8,
m,
(char **)0, ln, (bool)UTF))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
}
s += len;
&& (ln == 1 || !(OP(c) == EXACTF
? ibcmp(s, m, ln)
: ibcmp_locale(s, m, ln)))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
s++;
}
&& (ln == 1 || !(OP(c) == EXACTF
? ibcmp(s, m, ln)
: ibcmp_locale(s, m, ln)))
- && (norun || regtry(prog, s)) )
+ && (!reginfo || regtry(reginfo, s)) )
goto got_it;
s++;
}
isALNUM_LC_utf8((U8*)s)))
{
tmp = !tmp;
- if ((norun || regtry(prog, s)))
+ if ((!reginfo || regtry(reginfo, s)))
goto got_it;
}
s += uskip;
if (tmp ==
!(OP(c) == BOUND ? isALNUM(*s) : isALNUM_LC(*s))) {
tmp = !tmp;
- if ((norun || regtry(prog, s)))
+ if ((!reginfo || regtry(reginfo, s)))
goto got_it;
}
s++;
}
}
- if ((!prog->minlen && tmp) && (norun || regtry(prog, s)))
+ if ((!prog->minlen && tmp) && (!reginfo || regtry(reginfo, s)))
goto got_it;
break;
case NBOUNDL:
swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) :
isALNUM_LC_utf8((U8*)s)))
tmp = !tmp;
- else if ((norun || regtry(prog, s)))
+ else if ((!reginfo || regtry(reginfo, s)))
goto got_it;
s += uskip;
}
if (tmp ==
!(OP(c) == NBOUND ? isALNUM(*s) : isALNUM_LC(*s)))
tmp = !tmp;
- else if ((norun || regtry(prog, s)))
+ else if ((!reginfo || regtry(reginfo, s)))
goto got_it;
s++;
}
}
- if ((!prog->minlen && !tmp) && (norun || regtry(prog, s)))
+ if ((!prog->minlen && !tmp) && (!reginfo || regtry(reginfo, s)))
goto got_it;
break;
case ALNUM:
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isALNUM(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (isALNUM_LC_utf8((U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isALNUM_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isALNUM(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!isALNUM_LC_utf8((U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isALNUM_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isSPACE(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (*s == ' ' || isSPACE_LC_utf8((U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isSPACE_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isSPACE(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!(*s == ' ' || isSPACE_LC_utf8((U8*)s))) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isSPACE_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isDIGIT(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (isDIGIT_LC_utf8((U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (isDIGIT_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
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)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isDIGIT(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
if (do_utf8) {
while (s + (uskip = UTF8SKIP(s)) <= strend) {
if (!isDIGIT_LC_utf8((U8*)s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
else {
while (s < strend) {
if (!isDIGIT_LC(*s)) {
- if (tmp && (norun || regtry(prog, s)))
+ if (tmp && (!reginfo || regtry(reginfo, s)))
goto got_it;
else
tmp = doevery;
SV* dsv0;
SV* dsv1;
#endif
+ regmatch_info reginfo; /* create some info to pass to regtry etc */
GET_RE_DEBUG_FLAGS_DECL;
}
multiline = prog->reganch & PMf_MULTILINE;
+ reginfo.prog = prog;
#ifdef DEBUGGING
dsv0 = PERL_DEBUG_PAD_ZERO(0);
dsv1 = PERL_DEBUG_PAD_ZERO(1);
#endif
-#ifdef DEBUGGING
- PL_regnarrate = DEBUG_r_TEST;
-#endif
-
RX_MATCH_UTF8_set(prog, do_utf8);
minlen = prog->minlen;
PL_reg_flags |= RF_utf8;
/* Mark beginning of line for ^ and lookbehind. */
- PL_regbol = startpos;
+ reginfo.bol = startpos; /* XXX not used ??? */
PL_bostr = strbeg;
- PL_reg_sv = sv;
+ reginfo.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;
+ reginfo.till = startpos+minend;
/* If there is a "must appear" string, look for it. */
s = startpos;
- if (prog->reganch & ROPT_GPOS_SEEN) { /* Need to have PL_reg_ganch */
+ if (prog->reganch & ROPT_GPOS_SEEN) { /* Need to set reginfo->ganch */
MAGIC *mg;
if (flags & REXEC_IGNOREPOS) /* Means: check only at start */
- PL_reg_ganch = startpos;
+ reginfo.ganch = startpos;
else if (sv && SvTYPE(sv) >= SVt_PVMG
&& SvMAGIC(sv)
&& (mg = mg_find(sv, PERL_MAGIC_regex_global))
&& mg->mg_len >= 0) {
- PL_reg_ganch = strbeg + mg->mg_len; /* Defined pos() */
+ reginfo.ganch = strbeg + mg->mg_len; /* Defined pos() */
if (prog->reganch & ROPT_ANCH_GPOS) {
- if (s > PL_reg_ganch)
+ if (s > reginfo.ganch)
goto phooey;
- s = PL_reg_ganch;
+ s = reginfo.ganch;
}
}
else /* pos() not defined */
- PL_reg_ganch = strbeg;
+ reginfo.ganch = strbeg;
}
if (!(flags & REXEC_CHECKED) && (prog->check_substr != NULL || prog->check_utf8 != NULL)) {
/* 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))
+ if (s == startpos && regtry(®info, startpos))
goto got_it;
else if (multiline || (prog->reganch & ROPT_IMPLICIT)
|| (prog->reganch & ROPT_ANCH_MBOL)) /* XXXX SBOL? */
if (s == startpos)
goto after_try;
while (1) {
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
after_try:
if (s >= end)
s--;
while (s < end) {
if (*s++ == '\n') { /* don't need PL_utf8skip here */
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
}
}
}
goto phooey;
} else if (prog->reganch & ROPT_ANCH_GPOS) {
- if (regtry(prog, PL_reg_ganch))
+ if (regtry(®info, reginfo.ganch))
goto got_it;
goto phooey;
}
while (s < strend) {
if (*s == ch) {
DEBUG_EXECUTE_r( did_match = 1 );
- if (regtry(prog, s)) goto got_it;
+ if (regtry(®info, s)) goto got_it;
s += UTF8SKIP(s);
while (s < strend && *s == ch)
s += UTF8SKIP(s);
while (s < strend) {
if (*s == ch) {
DEBUG_EXECUTE_r( did_match = 1 );
- if (regtry(prog, s)) goto got_it;
+ if (regtry(®info, s)) goto got_it;
s++;
while (s < strend && *s == ch)
s++;
}
if (do_utf8) {
while (s <= last1) {
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
s += UTF8SKIP(s);
}
}
else {
while (s <= last1) {
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
s++;
}
len0, len0, s0,
len1, len1, s1);
});
- if (find_byclass(prog, c, s, strend, 0))
+ if (find_byclass(prog, c, s, strend, ®info))
goto got_it;
DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass...\n"));
}
/* We don't know much -- general case. */
if (do_utf8) {
for (;;) {
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
if (s >= strend)
break;
}
else {
do {
- if (regtry(prog, s))
+ if (regtry(®info, s))
goto got_it;
} while (s++ < strend);
}
- regtry - try match at specific point
*/
STATIC I32 /* 0 failure, 1 success */
-S_regtry(pTHX_ regexp *prog, char *startpos)
+S_regtry(pTHX_ const regmatch_info *reginfo, char *startpos)
{
dVAR;
register I32 *sp;
register I32 *ep;
CHECKPOINT lastcp;
+ regexp *prog = reginfo->prog;
GET_RE_DEBUG_FLAGS_DECL;
#ifdef DEBUGGING
/* SAVEI8(cxstack[cxstack_ix].blk_gimme);
cxstack[cxstack_ix].blk_gimme = G_SCALAR; */
- if (PL_reg_sv) {
+ if (reginfo->sv) {
/* Make $_ available to executed code. */
- if (PL_reg_sv != DEFSV) {
+ if (reginfo->sv != DEFSV) {
SAVE_DEFSV;
- DEFSV = PL_reg_sv;
+ DEFSV = reginfo->sv;
}
- if (!(SvTYPE(PL_reg_sv) >= SVt_PVMG && SvMAGIC(PL_reg_sv)
- && (mg = mg_find(PL_reg_sv, PERL_MAGIC_regex_global)))) {
+ if (!(SvTYPE(reginfo->sv) >= SVt_PVMG && SvMAGIC(reginfo->sv)
+ && (mg = mg_find(reginfo->sv, PERL_MAGIC_regex_global)))) {
/* prepare for quick setting of pos */
#ifdef PERL_OLD_COPY_ON_WRITE
if (SvIsCOW(sv))
sv_force_normal_flags(sv, 0);
#endif
- mg = sv_magicext(PL_reg_sv, (SV*)0, PERL_MAGIC_regex_global,
+ mg = sv_magicext(reginfo->sv, (SV*)0, PERL_MAGIC_regex_global,
&PL_vtbl_mglob, NULL, 0);
mg->mg_len = -1;
}
}
#endif
REGCP_SET(lastcp);
- if (regmatch(prog, prog->program + 1)) {
+ if (regmatch(reginfo, prog->program + 1)) {
prog->endp[0] = PL_reginput - PL_bostr;
return 1;
}
#define sayNO goto no
#define sayNO_ANYOF goto no_anyof
#define sayYES_FINAL goto yes_final
-#define sayYES_LOUD goto yes_loud
#define sayNO_FINAL goto no_final
#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 (st->u.whilem.cache_offset | st->u.whilem.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))) { \
+ if (!(PL_reg_poscache[0] & (1<<POSCACHE_SEEN))) { \
+ PL_reg_poscache[0] |= (1<<POSCACHE_SUCCESS) | (1<<POSCACHE_SEEN); \
+ PL_reg_poscache[st->u.whilem.cache_offset] |= (1<<st->u.whilem.cache_bit); \
+ } \
+ else if (PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS)) { \
+ PL_reg_poscache[st->u.whilem.cache_offset] |= (1<<st->u.whilem.cache_bit); \
+ } \
+ else { \
/* cache records failure, but this is success */ \
DEBUG_r( \
PerlIO_printf(Perl_debug_log, \
} \
sayYES; \
} STMT_END
+
#define CACHEsayNO STMT_START { \
if (st->u.whilem.cache_offset | st->u.whilem.cache_bit) { \
- if (!(PL_reg_poscache[0] & (1<<POSCACHE_SEEN))) \
+ if (!(PL_reg_poscache[0] & (1<<POSCACHE_SEEN))) { \
PL_reg_poscache[0] |= (1<<POSCACHE_SEEN); \
- else if ((PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS))) { \
+ PL_reg_poscache[st->u.whilem.cache_offset] |= (1<<st->u.whilem.cache_bit); \
+ } \
+ else if (!(PL_reg_poscache[0] & (1<<POSCACHE_SUCCESS))) { \
+ PL_reg_poscache[st->u.whilem.cache_offset] |= (1<<st->u.whilem.cache_bit); \
+ } \
+ else { \
/* cache records success, but this is failure */ \
DEBUG_r( \
PerlIO_printf(Perl_debug_log, \
/* Make sure there is a test for this +1 options in re_tests */
#define TRIE_INITAL_ACCEPT_BUFFLEN 4;
+#define SLAB_FIRST(s) (&(s)->states[0])
+#define SLAB_LAST(s) (&(s)->states[PERL_REGMATCH_SLAB_SLOTS-1])
+
/* grab a new slab and return the first slot in it */
STATIC regmatch_state *
PL_regmatch_slab->next = s;
}
PL_regmatch_slab = s;
- return &s->states[0];
+ return SLAB_FIRST(s);
}
/* simulate a recursive call to regmatch */
st->locinput = locinput; \
st->resume_state = resume; \
newst = st+1; \
- if (newst > &(PL_regmatch_slab->states[PERL_REGMATCH_SLAB_SLOTS-1])) \
+ if (newst > SLAB_LAST(PL_regmatch_slab)) \
newst = S_push_slab(aTHX); \
PL_regmatch_state = newst; \
newst->cc = 0; \
DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "POP STATE(%d)\n", depth)); \
depth--; \
st--; \
- if (st < &PL_regmatch_slab->states[0]) { \
+ if (st < SLAB_FIRST(PL_regmatch_slab)) { \
PL_regmatch_slab = PL_regmatch_slab->prev; \
- st = &PL_regmatch_slab->states[PERL_REGMATCH_SLAB_SLOTS-1]; \
+ st = SLAB_LAST(PL_regmatch_slab); \
} \
PL_regmatch_state = st; \
scan = st->scan; \
STATIC I32 /* 0 failure, 1 success */
-S_regmatch(pTHX_ regexp *rex, regnode *prog)
+S_regmatch(pTHX_ const regmatch_info *reginfo, regnode *prog)
{
dVAR;
register const bool do_utf8 = PL_reg_match_utf8;
const U32 uniflags = UTF8_ALLOW_DEFAULT;
+ regexp *rex = reginfo->prog;
+
regmatch_slab *orig_slab;
regmatch_state *orig_state;
regnode *inner; /* Next node in internal branch. */
int depth = 0; /* depth of recursion */
regmatch_state *newst; /* when pushing a state, this is the new one */
- regmatch_state *cur_eval = NULL; /* most recent (??{}) state */
+ regmatch_state *yes_state = NULL; /* state to pop to on success of
+ subpattern */
#ifdef DEBUGGING
SV *re_debug_flags = NULL;
Newx(PL_regmatch_slab, 1, regmatch_slab);
PL_regmatch_slab->prev = NULL;
PL_regmatch_slab->next = NULL;
- PL_regmatch_state = &PL_regmatch_slab->states[0] - 1;
+ PL_regmatch_state = SLAB_FIRST(PL_regmatch_slab);
}
/* remember current high-water mark for exit */
/* grab next free state slot */
st = ++PL_regmatch_state;
- if (st > &(PL_regmatch_slab->states[PERL_REGMATCH_SLAB_SLOTS-1]))
+ if (st > SLAB_LAST(PL_regmatch_slab))
st = PL_regmatch_state = S_push_slab(aTHX);
st->minmod = 0;
case BOL:
if (locinput == PL_bostr)
{
- /* regtill = regbol; */
+ /* reginfo->till = reginfo->bol; */
break;
}
sayNO;
break;
sayNO;
case GPOS:
- if (locinput == PL_reg_ganch)
+ if (locinput == reginfo->ganch)
break;
sayNO;
case EOL:
st->u.eval.prev_rex = rex;
rex = re;
- st->u.eval.prev_eval = cur_eval;
- st->u.eval.prev_slab = PL_regmatch_slab;
- st->u.eval.depth = depth;
- cur_eval = st;
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
PUSH_STATE(newst, resume_EVAL);
st = newst;
/* No need to save/restore up to this paren */
I32 parenfloor = scan->flags;
- if (next && (OP(PREVOPER(next)) == NOTHING)) /* LONGJMP */
+ /* Dave says:
+
+ CURLYX and WHILEM are always paired: they're the moral
+ equivalent of pp_enteriter anbd pp_iter.
+
+ The only time next could be null is if the node tree is
+ corrupt. This was mentioned on p5p a few days ago.
+
+ See http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2006-04/msg00556.html
+ So we'll assert that this is true:
+ */
+ assert(next);
+ if (OP(PREVOPER(next)) == NOTHING) /* LONGJMP */
next += ARG(next);
/* XXXX Probably it is better to teach regpush to support
parenfloor > PL_regsize... */
* that we can try again after backing off.
*/
+ /* Dave says:
+
+ st->cc gets initialised by CURLYX ready for use by WHILEM.
+ So again, unless somethings been corrupted, st->cc cannot
+ be null at that point in WHILEM.
+
+ See http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2006-04/msg00556.html
+ So we'll assert that this is true:
+ */
+ assert(st->cc);
st->u.whilem.lastloc = st->cc->u.curlyx.lastloc; /* Detection of 0-len. */
st->u.whilem.cache_offset = 0;
st->u.whilem.cache_bit = 0;
/* cache records failure */
sayNO_SILENT;
}
- PL_reg_poscache[st->u.whilem.cache_offset] |= (1<<st->u.whilem.cache_bit);
}
}
st->u.curlym.maxwanted = st->minmod ? st->ln : n;
if (st->u.curlym.maxwanted) {
while (PL_reginput < PL_regeol && st->u.curlym.matches < st->u.curlym.maxwanted) {
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
REGMATCH(scan, CURLYM1);
+ yes_state = st->u.yes.prev_yes_state;
/*** all unsaved local vars undefined at this point */
if (!result)
break;
else
PL_regendp[st->u.curlym.paren] = -1;
}
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
REGMATCH(next, CURLYM2);
+ yes_state = st->u.yes.prev_yes_state;
/*** all unsaved local vars undefined at this point */
if (result)
- sayYES;
+ /* XXX tmp sayYES; */
+ sayYES_FINAL;
REGCP_UNWIND(st->u.curlym.lastcp);
}
/* Couldn't or didn't -- move forward. */
PL_reginput = locinput;
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
REGMATCH(scan, CURLYM3);
+ yes_state = st->u.yes.prev_yes_state;
/*** all unsaved local vars undefined at this point */
if (result) {
st->ln++;
else
PL_regendp[st->u.curlym.paren] = -1;
}
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
REGMATCH(next, CURLYM4);
+ yes_state = st->u.yes.prev_yes_state;
/*** all unsaved local vars undefined at this point */
if (result)
- sayYES;
+ /* XXX tmp sayYES; */
+ sayYES_FINAL;
REGCP_UNWIND(st->u.curlym.lastcp);
}
/* Couldn't or didn't -- back up. */
sayNO;
break;
case END:
- if (cur_eval) {
- /* we have successfully completed the execution of a
- * postponed re. Pop all states back to the last EVAL
- * then continue with the node following the (??{...})
- */
-
- /* this simulates a POP_STATE, except that it pops several
- * levels, and doesn't restore locinput */
-
- st = cur_eval;
- PL_regmatch_slab = st->u.eval.prev_slab;
- cur_eval = st->u.eval.prev_eval;
- depth = st->u.eval.depth;
-
- PL_regmatch_state = st;
- scan = st->scan;
- next = st->next;
- n = st->n;
-
- if (st->u.eval.toggleutf)
- PL_reg_flags ^= RF_utf8;
- ReREFCNT_dec(rex);
- rex = st->u.eval.prev_rex;
- /* XXXX This is too dramatic a measure... */
- PL_reg_maxiter = 0;
-
- /* Restore parens of the caller without popping the
- * savestack */
- {
- I32 tmp = PL_savestack_ix;
- PL_savestack_ix = st->u.eval.lastcp;
- regcppop(rex);
- PL_savestack_ix = tmp;
- }
-
-
- PL_reginput = locinput;
- /* resume at node following the (??{...}) */
- break;
-
- }
-
- if (locinput < PL_regtill) {
+ if (locinput < reginfo->till) {
DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
"%sMatch possible, but length=%ld is smaller than requested=%ld, failing!%s\n",
PL_colors[4],
(long)(locinput - PL_reg_starttry),
- (long)(PL_regtill - PL_reg_starttry),
+ (long)(reginfo->till - PL_reg_starttry),
PL_colors[5]));
sayNO_FINAL; /* Cannot match: too short. */
}
PL_reginput = locinput; /* put where regtry can find it */
sayYES_FINAL; /* Success! */
- case SUCCEED:
+
+ case SUCCEED: /* successful SUSPEND/UNLESSM/IFMATCH/CURLYM */
+ DEBUG_EXECUTE_r(
+ PerlIO_printf(Perl_debug_log,
+ "%*s %ssubpattern success...%s\n",
+ REPORT_CODE_OFF+PL_regindent*2, "", PL_colors[4], PL_colors[5]));
PL_reginput = locinput; /* put where regtry can find it */
- sayYES_LOUD; /* Success! */
- case SUSPEND:
- n = 1;
+ sayYES_FINAL; /* Success! */
+
+ case SUSPEND: /* (?>FOO) */
+ st->u.ifmatch.wanted = 1;
PL_reginput = locinput;
goto do_ifmatch;
- case UNLESSM:
- n = 0;
- if (scan->flags) {
- char * const s = HOPBACKc(locinput, scan->flags);
- if (!s)
- goto say_yes;
- PL_reginput = s;
- }
- else
- PL_reginput = locinput;
- goto do_ifmatch;
- case IFMATCH:
- n = 1;
+
+ case UNLESSM: /* -ve lookaround: (?!FOO), or with flags, (?<!foo) */
+ st->u.ifmatch.wanted = 0;
+ goto ifmatch_trivial_fail_test;
+
+ case IFMATCH: /* +ve lookaround: (?=FOO), or with flags, (?<=foo) */
+ st->u.ifmatch.wanted = 1;
+ ifmatch_trivial_fail_test:
if (scan->flags) {
char * const s = HOPBACKc(locinput, scan->flags);
- if (!s)
- goto say_no;
+ if (!s) {
+ /* trivial fail */
+ if (st->logical) {
+ st->logical = 0;
+ st->sw = 1 - st->u.ifmatch.wanted;
+ }
+ else if (st->u.ifmatch.wanted)
+ sayNO;
+ next = scan + ARG(scan);
+ if (next == scan)
+ next = NULL;
+ break;
+ }
PL_reginput = s;
}
else
PL_reginput = locinput;
do_ifmatch:
- REGMATCH(NEXTOPER(NEXTOPER(scan)), IFMATCH);
- /*** all unsaved local vars undefined at this point */
- if (result != n) {
- say_no:
- if (st->logical) {
- st->logical = 0;
- st->sw = 0;
- goto do_longjump;
- }
- else
- sayNO;
- }
- say_yes:
- if (st->logical) {
- st->logical = 0;
- st->sw = 1;
- }
- if (OP(scan) == SUSPEND) {
- locinput = PL_reginput;
- nextchr = UCHARAT(locinput);
- }
- /* FALL THROUGH. */
+ /* resume to current state on success */
+ st->u.yes.prev_yes_state = yes_state;
+ yes_state = st;
+ PUSH_STATE(newst, resume_IFMATCH);
+ st = newst;
+ next = NEXTOPER(NEXTOPER(scan));
+ break;
+
case LONGJMP:
- do_longjump:
next = scan + ARG(scan);
if (next == scan)
next = NULL;
/* grab the next free state slot */
st++;
- if (st > &(PL_regmatch_slab->states[PERL_REGMATCH_SLAB_SLOTS-1]))
+ if (st > SLAB_LAST(PL_regmatch_slab))
st = S_push_slab(aTHX);
PL_regmatch_state = st;
/*NOTREACHED*/
sayNO;
-yes_loud:
- DEBUG_EXECUTE_r(
- PerlIO_printf(Perl_debug_log,
- "%*s %scould match...%s\n",
- REPORT_CODE_OFF+PL_regindent*2, "", PL_colors[4], PL_colors[5])
- );
- goto yes;
yes_final:
+
+ if (yes_state) {
+ /* we have successfully completed a subexpression, but we must now
+ * pop to the state marked by yes_state and continue from there */
+
+ /*XXX tmp for CURLYM*/
+ regmatch_slab *oslab = PL_regmatch_slab;
+ regmatch_state *ost = st, *oys=yes_state;
+ int odepth = depth;
+
+ assert(st != yes_state);
+ while (yes_state < SLAB_FIRST(PL_regmatch_slab)
+ || yes_state > SLAB_LAST(PL_regmatch_slab))
+ {
+ /* not in this slab, pop slab */
+ depth -= (st - SLAB_FIRST(PL_regmatch_slab) + 1);
+ PL_regmatch_slab = PL_regmatch_slab->prev;
+ st = SLAB_LAST(PL_regmatch_slab);
+ }
+ depth -= (st - yes_state);
+ DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "POP STATE TO (%d)\n", depth));
+ st = yes_state;
+ yes_state = st->u.yes.prev_yes_state;
+ PL_regmatch_state = st;
+
+ switch (st->resume_state) {
+ case resume_EVAL:
+ if (st->u.eval.toggleutf)
+ PL_reg_flags ^= RF_utf8;
+ ReREFCNT_dec(rex);
+ rex = st->u.eval.prev_rex;
+ /* XXXX This is too dramatic a measure... */
+ PL_reg_maxiter = 0;
+ /* Restore parens of the caller without popping the
+ * savestack */
+ {
+ I32 tmp = PL_savestack_ix;
+ PL_savestack_ix = st->u.eval.lastcp;
+ regcppop(rex);
+ PL_savestack_ix = tmp;
+ }
+ PL_reginput = locinput;
+ /* continue at the node following the (??{...}) */
+ next = st->next;
+ goto reenter;
+
+ case resume_IFMATCH:
+ if (st->logical) {
+ st->logical = 0;
+ st->sw = st->u.ifmatch.wanted;
+ }
+ else if (!st->u.ifmatch.wanted)
+ sayNO;
+
+ if (OP(st->scan) == SUSPEND)
+ locinput = PL_reginput;
+ else {
+ locinput = PL_reginput = st->locinput;
+ nextchr = UCHARAT(locinput);
+ }
+ next = st->scan + ARG(st->scan);
+ if (next == st->scan)
+ next = NULL;
+ goto reenter;
+
+ /* XXX tmp don't handle yes_state yet */
+ case resume_CURLYM1:
+ case resume_CURLYM2:
+ case resume_CURLYM3:
+ case resume_CURLYM4:
+ PL_regmatch_slab =oslab;
+ st = ost;
+ PL_regmatch_state = st;
+ depth = odepth;
+ yes_state = oys;
+ DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "XXX revering a CURLYM\n"));
+ goto yes;
+
+ default:
+ Perl_croak(aTHX_ "unexpected yes reume state");
+ }
+ }
+
DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%sMatch successful!%s\n",
PL_colors[4], PL_colors[5]));
yes:
result = 1;
/* XXX this is duplicate(ish) code to that in the do_no section.
- * eventually a yes should just pop the whole stack */
+ * eventually a yes should just pop the stack back to the current
+ * yes_state */
if (depth) {
/* restore previous state and re-enter */
POP_STATE;
goto resume_point_TRIE1;
case resume_TRIE2:
goto resume_point_TRIE2;
- case resume_EVAL:
- break;
case resume_CURLYX:
goto resume_point_CURLYX;
case resume_WHILEM1:
goto resume_point_CURLYM3;
case resume_CURLYM4:
goto resume_point_CURLYM4;
- case resume_IFMATCH:
- goto resume_point_IFMATCH;
case resume_PLUS1:
goto resume_point_PLUS1;
case resume_PLUS2:
goto resume_point_PLUS3;
case resume_PLUS4:
goto resume_point_PLUS4;
+
+ case resume_IFMATCH:
+ case resume_EVAL:
default:
Perl_croak(aTHX_ "regexp resume memory corruption");
}
PL_reg_flags ^= RF_utf8;
ReREFCNT_dec(rex);
rex = st->u.eval.prev_rex;
- cur_eval = st->u.eval.prev_eval;
+ yes_state = st->u.yes.prev_yes_state;
/* XXXX This is too dramatic a measure... */
PL_reg_maxiter = 0;
case resume_CURLYM4:
goto resume_point_CURLYM4;
case resume_IFMATCH:
- goto resume_point_IFMATCH;
+ yes_state = st->u.yes.prev_yes_state;
+ if (st->logical) {
+ st->logical = 0;
+ st->sw = !st->u.ifmatch.wanted;
+ }
+ else if (st->u.ifmatch.wanted)
+ sayNO;
+
+ assert(OP(scan) != SUSPEND); /* XXX DAPM tmp */
+ locinput = PL_reginput = st->locinput;
+ nextchr = UCHARAT(locinput);
+ next = scan + ARG(scan);
+ if (next == scan)
+ next = NULL;
+ goto reenter;
+
case resume_PLUS1:
goto resume_point_PLUS1;
case resume_PLUS2:
https://perl5.git.perl.org / perl5.git / blobdiff