*/
#define CHR_SVLEN(sv) (utf8_target ? sv_len_utf8(sv) : SvCUR(sv))
-#define CHR_DIST(a,b) (reginfo->is_utf8_target ? utf8_distance(a,b) : a - b)
#define HOPc(pos,off) \
(char *)(reginfo->is_utf8_target \
(U8*)(off >= 0 ? reginfo->strend : reginfo->strbeg)) \
: (U8*)(pos + off))
-#define HOPBACKc(pos, off) \
- (char*)(reginfo->is_utf8_target \
- ? reghopmaybe3((U8*)pos, (SSize_t)0-off, (U8*)(reginfo->strbeg)) \
- : (pos - off >= reginfo->strbeg) \
- ? (U8*)pos - off \
+/* like HOPMAYBE3 but backwards. lim must be +ve. Returns NULL on overshoot */
+#define HOPBACK3(pos, off, lim) \
+ (reginfo->is_utf8_target \
+ ? reghopmaybe3((U8*)pos, (SSize_t)0-off, (U8*)(lim)) \
+ : (pos - off >= lim) \
+ ? (U8*)pos - off \
: NULL)
+#define HOPBACKc(pos, off) ((char*)HOPBACK3(pos, off, reginfo->strbeg))
+
#define HOP3(pos,off,lim) (reginfo->is_utf8_target ? reghop3((U8*)(pos), off, (U8*)(lim)) : (U8*)(pos + off))
#define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim))
#define HOP3lim(pos,off,lim) (reginfo->is_utf8_target \
? reghop3((U8*)(pos), off, (U8*)(lim)) \
: (U8*)((pos + off) > lim ? lim : (pos + off)))
+#define HOP3clim(pos,off,lim) ((char*)HOP3lim(pos,off,lim))
#define HOP4(pos,off,llim, rlim) (reginfo->is_utf8_target \
? reghop4((U8*)(pos), off, (U8*)(llim), (U8*)(rlim)) \
(int)parenfloor, (unsigned)REGCP_PAREN_ELEMS);
if ((elems_shifted >> SAVE_TIGHT_SHIFT) != total_elems)
- Perl_croak(aTHX_ "panic: paren_elems_to_push offset %"UVuf
+ Perl_croak(aTHX_ "panic: paren_elems_to_push offset %" UVuf
" out of range (%lu-%ld)",
total_elems,
(unsigned long)maxopenparen,
DEBUG_BUFFERS_r(
if ((int)maxopenparen > (int)parenfloor)
Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" offs=0x%"UVxf": saving capture indices:\n",
+ "rex=0x%" UVxf " offs=0x%" UVxf ": saving capture indices:\n",
depth,
PTR2UV(rex),
PTR2UV(rex->offs)
SSPUSHIV(rex->offs[p].start);
SSPUSHINT(rex->offs[p].start_tmp);
DEBUG_BUFFERS_r(Perl_re_exec_indentf( aTHX_
- " \\%"UVuf": %"IVdf"(%"IVdf")..%"IVdf"\n",
+ " \\%" UVuf ": %" IVdf "(%" IVdf ")..%" IVdf "\n",
depth,
(UV)p,
(IV)rex->offs[p].start,
/* These are needed since we do not localize EVAL nodes: */
#define REGCP_SET(cp) \
DEBUG_STATE_r( \
- Perl_re_exec_indentf( aTHX_ \
- "Setting an EVAL scope, savestack=%"IVdf",\n", \
+ Perl_re_exec_indentf( aTHX_ \
+ "Setting an EVAL scope, savestack=%" IVdf ",\n", \
depth, (IV)PL_savestack_ix \
) \
); \
#define REGCP_UNWIND(cp) \
DEBUG_STATE_r( \
if (cp != PL_savestack_ix) \
- Perl_re_exec_indentf( aTHX_ \
- "Clearing an EVAL scope, savestack=%"IVdf"..%"IVdf"\n",\
+ Perl_re_exec_indentf( aTHX_ \
+ "Clearing an EVAL scope, savestack=%" \
+ IVdf "..%" IVdf "\n", \
depth, (IV)(cp), (IV)PL_savestack_ix \
) \
); \
DEBUG_BUFFERS_r(
if (i || rex->lastparen + 1 <= rex->nparens)
Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" offs=0x%"UVxf": restoring capture indices to:\n",
+ "rex=0x%" UVxf " offs=0x%" UVxf ": restoring capture indices to:\n",
depth,
PTR2UV(rex),
PTR2UV(rex->offs)
if (paren <= rex->lastparen)
rex->offs[paren].end = tmps;
DEBUG_BUFFERS_r( Perl_re_exec_indentf( aTHX_
- " \\%"UVuf": %"IVdf"(%"IVdf")..%"IVdf"%s\n",
+ " \\%" UVuf ": %" IVdf "(%" IVdf ")..%" IVdf "%s\n",
depth,
(UV)paren,
(IV)rex->offs[paren].start,
rex->offs[i].start = -1;
rex->offs[i].end = -1;
DEBUG_BUFFERS_r( Perl_re_exec_indentf( aTHX_
- " \\%"UVuf": %s ..-1 undeffing\n",
+ " \\%" UVuf ": %s ..-1 undeffing\n",
depth,
(UV)i,
(i > *maxopenparen_p) ? "-1" : " "
#define regcpblow(cp) LEAVE_SCOPE(cp) /* Ignores regcppush()ed data. */
-STATIC bool
-S_isFOO_lc(pTHX_ const U8 classnum, const U8 character)
+#ifndef PERL_IN_XSUB_RE
+
+bool
+Perl_isFOO_lc(pTHX_ const U8 classnum, const U8 character)
{
/* Returns a boolean as to whether or not 'character' is a member of the
* Posix character class given by 'classnum' that should be equivalent to a
return FALSE;
}
+#endif
+
STATIC bool
S_isFOO_utf8_lc(pTHX_ const U8 classnum, const U8* character)
{
* /\d+abc/: "abc" is anchored to neither the pattern nor the string;
* /^\d+abc/: "abc" is anchored to neither the pattern nor the string,
* but the pattern is anchored to the string.
+ *
+ * Note the variables names suffixed with _c represent character counts
+ * while _b represent byte counts
*/
char *
re_scream_pos_data *data)
{
struct regexp *const prog = ReANY(rx);
- SSize_t start_shift = prog->check_offset_min;
- /* Should be nonnegative! */
- SSize_t end_shift = 0;
+
+ /* Minimum number of chars which *must* precede the check substring to
+ * be capable of matching, e.g. 2 in /[ab]cd?substring/. */
+ SSize_t check_min_offset_c = prog->check_offset_min;
+
+ /* Minimum number of chars which *must* follow the check substring to
+ * be capable of matching, e.g. 2 in /substring[ab]cd?/.
+ * Should be nonnegative! */
+
+ SSize_t check_end_shift_c = 0;
/* current lowest pos in string where the regex can start matching */
char *rx_origin = strpos;
SV *check;
goto fail;
}
- RX_MATCH_UTF8_set(rx,utf8_target);
+ RXp_MATCH_UTF8_set(prog, utf8_target);
reginfo->is_utf8_target = cBOOL(utf8_target);
reginfo->info_aux = NULL;
reginfo->strbeg = strbeg;
continue;
Perl_re_printf( aTHX_
- " substrs[%d]: min=%"IVdf" max=%"IVdf" end shift=%"IVdf
- " useful=%"IVdf" utf8=%d [%s]\n",
+ " substrs[%d]: min=%" IVdf " max=%" IVdf " end shift=%" IVdf
+ " useful=%" IVdf " utf8=%d [%s]\n",
i,
(IV)prog->substrs->data[i].min_offset,
(IV)prog->substrs->data[i].max_offset,
if (prog->check_offset_min == prog->check_offset_max) {
/* Substring at constant offset from beg-of-str... */
- SSize_t slen = SvCUR(check);
+ SSize_t slen_b = SvCUR(check);
char *s = HOP3c(strpos, prog->check_offset_min, strend);
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " Looking for check substr at fixed offset %"IVdf"...\n",
+ " Looking for check substr at fixed offset %" IVdf "...\n",
(IV)prog->check_offset_min));
if (SvTAIL(check)) {
/* In this case, the regex is anchored at the end too.
* Unless it's a multiline match, the lengths must match
- * exactly, give or take a \n. NB: slen >= 1 since
+ * exactly, give or take a \n. NB: slen_b >= 1 since
* the last char of check is \n */
if (!multiline
- && ( strend - s > slen
- || strend - s < slen - 1
- || (strend - s == slen && strend[-1] != '\n')))
+ && ( strend - s > slen_b
+ || strend - s < slen_b - 1
+ || (strend - s == slen_b && strend[-1] != '\n')))
{
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
" String too long...\n"));
goto fail_finish;
}
- /* Now should match s[0..slen-2] */
- slen--;
+ /* Now should match s[0..slen_b-2] */
+ slen_b--;
}
- if (slen && (strend - s < slen
+ if (slen_b && (strend - s < slen_b
|| *SvPVX_const(check) != *s
- || (slen > 1 && (memNE(SvPVX_const(check), s, slen)))))
+ || (slen_b > 1 && (memNE(SvPVX_const(check), s, slen_b)))))
{
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
" String not equal...\n"));
}
}
- end_shift = prog->check_end_shift;
+ check_end_shift_c = prog->check_end_shift;
#ifdef DEBUGGING /* 7/99: reports of failure (with the older version) */
- if (end_shift < 0)
- Perl_croak(aTHX_ "panic: end_shift: %"IVdf" pattern:\n%s\n ",
- (IV)end_shift, RX_PRECOMP(prog));
+ if (check_end_shift_c < 0)
+ Perl_croak(aTHX_ "panic: end_shift: %" IVdf " pattern:\n%s\n ",
+ (IV)check_end_shift_c, RX_PRECOMP(rx));
#endif
restart:
/* This is the (re)entry point of the main loop in this function.
* The goal of this loop is to:
* 1) find the "check" substring in the region rx_origin..strend
- * (adjusted by start_shift / end_shift). If not found, reject
+ * (adjusted by check_min_offset_c / check_end_shift_c). If not found, reject
* immediately.
* 2) If it exists, look for the "other" substr too if defined; for
* example, if the check substr maps to the anchored substr, then
DEBUG_OPTIMISE_MORE_r({
Perl_re_printf( aTHX_
- " At restart: rx_origin=%"IVdf" Check offset min: %"IVdf
- " Start shift: %"IVdf" End shift %"IVdf
- " Real end Shift: %"IVdf"\n",
+ " At restart: rx_origin=%" IVdf " Check offset min: %" IVdf
+ " Start shift: %" IVdf " End shift %" IVdf
+ " Real end Shift: %" IVdf "\n",
(IV)(rx_origin - strbeg),
(IV)prog->check_offset_min,
- (IV)start_shift,
- (IV)end_shift,
+ (IV)check_min_offset_c,
+ (IV)check_end_shift_c,
(IV)prog->check_end_shift);
});
- end_point = HOP3(strend, -end_shift, strbeg);
- start_point = HOPMAYBE3(rx_origin, start_shift, end_point);
+ end_point = HOPBACK3(strend, check_end_shift_c, rx_origin);
+ if (!end_point)
+ goto fail_finish;
+ start_point = HOPMAYBE3(rx_origin, check_min_offset_c, end_point);
if (!start_point)
goto fail_finish;
&& prog->intflags & PREGf_ANCH
&& prog->check_offset_max != SSize_t_MAX)
{
- SSize_t len = SvCUR(check) - !!SvTAIL(check);
+ SSize_t check_len_b = SvCUR(check) - !!SvTAIL(check);
const char * const anchor =
(prog->intflags & PREGf_ANCH_GPOS ? strpos : strbeg);
+ SSize_t targ_len_b = (char*)end_point - anchor;
+
+ if (check_len_b > targ_len_b) {
+ DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
+ "Anchored string too short...\n"));
+ goto fail_finish;
+ }
/* do a bytes rather than chars comparison. It's conservative;
* so it skips doing the HOP if the result can't possibly end
* up earlier than the old value of end_point.
*/
- if ((char*)end_point - anchor > prog->check_offset_max) {
+ assert(anchor + check_len_b <= (char *)end_point);
+ if (prog->check_offset_max + check_len_b < targ_len_b) {
end_point = HOP3lim((U8*)anchor,
prog->check_offset_max,
- end_point -len)
- + len;
+ end_point - check_len_b
+ )
+ + check_len_b;
}
}
check, multiline ? FBMrf_MULTILINE : 0);
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " doing 'check' fbm scan, [%"IVdf"..%"IVdf"] gave %"IVdf"\n",
+ " doing 'check' fbm scan, [%" IVdf "..%" IVdf "] gave %" IVdf "\n",
(IV)((char*)start_point - strbeg),
(IV)((char*)end_point - strbeg),
(IV)(check_at ? check_at - strbeg : -1)
rx_origin = HOP3c(check_at, -prog->check_offset_max, rx_origin);
/* Finish the diagnostic message */
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- "%ld (rx_origin now %"IVdf")...\n",
+ "%ld (rx_origin now %" IVdf ")...\n",
(long)(check_at - strbeg),
(IV)(rx_origin - strbeg)
));
: (char*)HOP3lim(rx_origin, other->max_offset, last1);
}
else {
- assert(strpos + start_shift <= check_at);
- last = HOP4c(check_at, other->min_offset - start_shift,
+ assert(strpos + check_min_offset_c <= check_at);
+ last = HOP4c(check_at, other->min_offset - check_min_offset_c,
strbeg, strend);
}
if (from > to) {
s = NULL;
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " skipping 'other' fbm scan: %"IVdf" > %"IVdf"\n",
+ " skipping 'other' fbm scan: %" IVdf " > %" IVdf "\n",
(IV)(from - strbeg),
(IV)(to - strbeg)
));
multiline ? FBMrf_MULTILINE : 0
);
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " doing 'other' fbm scan, [%"IVdf"..%"IVdf"] gave %"IVdf"\n",
+ " doing 'other' fbm scan, [%" IVdf "..%" IVdf "] gave %" IVdf "\n",
(IV)(from - strbeg),
(IV)(to - strbeg),
(IV)(s ? s - strbeg : -1)
? HOP3c(rx_origin, 1, strend)
: HOP4c(last, 1 - other->min_offset, strbeg, strend);
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- "; about to retry %s at offset %ld (rx_origin now %"IVdf")...\n",
+ "; about to retry %s at offset %ld (rx_origin now %" IVdf ")...\n",
(other_ix ? "floating" : "anchored"),
(long)(HOP3c(check_at, 1, strend) - strbeg),
(IV)(rx_origin - strbeg)
other_last = HOP3c(s, 1, strend);
}
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " at offset %ld (rx_origin now %"IVdf")...\n",
+ " at offset %ld (rx_origin now %" IVdf ")...\n",
(long)(s - strbeg),
(IV)(rx_origin - strbeg)
));
else {
DEBUG_OPTIMISE_MORE_r(
Perl_re_printf( aTHX_
- " Check-only match: offset min:%"IVdf" max:%"IVdf
- " check_at:%"IVdf" rx_origin:%"IVdf" rx_origin-check_at:%"IVdf
- " strend:%"IVdf"\n",
+ " Check-only match: offset min:%" IVdf " max:%" IVdf
+ " check_at:%" IVdf " rx_origin:%" IVdf " rx_origin-check_at:%" IVdf
+ " strend:%" IVdf "\n",
(IV)prog->check_offset_min,
(IV)prog->check_offset_max,
(IV)(check_at-strbeg),
* didn't contradict, so just retry the anchored "other"
* substr */
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " Found /%s^%s/m, rescanning for anchored from offset %"IVdf" (rx_origin now %"IVdf")...\n",
+ " Found /%s^%s/m, rescanning for anchored from offset %" IVdf " (rx_origin now %" IVdf ")...\n",
PL_colors[0], PL_colors[1],
(IV)(rx_origin - strbeg + prog->anchored_offset),
(IV)(rx_origin - strbeg)
if (progi->regstclass && PL_regkind[OP(progi->regstclass)]!=TRIE) {
const U8* const str = (U8*)STRING(progi->regstclass);
- /* XXX this value could be pre-computed */
- const int cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
+ /* The length (in chars) if the character class. This is almost
+ * always 1.
+ * XXX this value could be pre-computed */
+ const int class_len_c = (PL_regkind[OP(progi->regstclass)] == EXACT
? (reginfo->is_utf8_pat
? utf8_distance(str + STR_LEN(progi->regstclass), str)
: STR_LEN(progi->regstclass))
*/
if (prog->anchored_substr || prog->anchored_utf8 || ml_anch)
- endpos= HOP3c(rx_origin, (prog->minlen ? cl_l : 0), strend);
+ endpos = HOP3clim(rx_origin,
+ (prog->minlen ? class_len_c : 0),
+ strend);
else if (prog->float_substr || prog->float_utf8) {
- rx_max_float = HOP3c(check_at, -start_shift, strbeg);
- endpos= HOP3c(rx_max_float, cl_l, strend);
+ rx_max_float = HOP3c(check_at, -check_min_offset_c, strbeg);
+ endpos = HOP3clim(rx_max_float, class_len_c, strend);
}
else
endpos= strend;
DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
- " looking for class: start_shift: %"IVdf" check_at: %"IVdf
- " rx_origin: %"IVdf" endpos: %"IVdf"\n",
- (IV)start_shift, (IV)(check_at - strbeg),
+ " looking for class: check_min_offset: %" IVdf " check_at: %" IVdf
+ " rx_origin: %" IVdf " endpos: %" IVdf "\n",
+ (IV)check_min_offset_c, (IV)(check_at - strbeg),
(IV)(rx_origin - strbeg), (IV)(endpos - strbeg)));
s = find_byclass(prog, progi->regstclass, rx_origin, endpos,
if (prog->anchored_substr || prog->anchored_utf8) {
if (prog->substrs->check_ix == 1) { /* check is float */
/* Have both, check_string is floating */
- assert(rx_origin + start_shift <= check_at);
- if (rx_origin + start_shift != check_at) {
+ assert(rx_origin + check_min_offset_c <= check_at);
+ if (rx_origin + check_min_offset_c != check_at) {
/* not at latest position float substr could match:
* Recheck anchored substring, but not floating.
* The condition above is in bytes rather than
* practice the extra fbm_instr() is likely to
* get skipped anyway. */
DEBUG_EXECUTE_r( Perl_re_printf( aTHX_
- " about to retry anchored at offset %ld (rx_origin now %"IVdf")...\n",
+ " about to retry anchored at offset %ld (rx_origin now %" IVdf ")...\n",
(long)(other_last - strbeg),
(IV)(rx_origin - strbeg)
));
/* uses bytes rather than char calculations for efficiency.
* It's conservative: it errs on the side of doing 'goto restart',
* where there is code that does a proper char-based test */
- if (rx_origin + start_shift + end_shift > strend) {
+ if (rx_origin + check_min_offset_c + check_end_shift_c > strend) {
DEBUG_EXECUTE_r( Perl_re_printf( aTHX_
" Could not match STCLASS...\n") );
goto fail;
}
DEBUG_EXECUTE_r( Perl_re_printf( aTHX_
- " about to look for %s substr starting at offset %ld (rx_origin now %"IVdf")...\n",
+ " about to look for %s substr starting at offset %ld (rx_origin now %" IVdf ")...\n",
(prog->substrs->check_ix ? "floating" : "anchored"),
- (long)(rx_origin + start_shift - strbeg),
+ (long)(rx_origin + check_min_offset_c - strbeg),
(IV)(rx_origin - strbeg)
));
goto restart;
uscan += len; \
len=0; \
} else { \
- uvc = _to_utf8_fold_flags( (const U8*) uc, foldbuf, &foldlen, flags); \
len = UTF8SKIP(uc); \
+ uvc = _toFOLD_utf8_flags( (const U8*) uc, uc + len, foldbuf, &foldlen, \
+ flags); \
skiplen = UVCHR_SKIP( uvc ); \
foldlen -= skiplen; \
uscan = foldbuf + skiplen; \
tmp = TEST_UV(tmp); \
LOAD_UTF8_CHARCLASS_ALNUM(); \
REXEC_FBC_UTF8_SCAN( /* advances s while s < strend */ \
- if (tmp == ! (TEST_UTF8((U8 *) s))) { \
+ if (tmp == ! (TEST_UTF8((U8 *) s, (U8 *) reginfo->strend))) { \
tmp = !tmp; \
IF_SUCCESS; \
} \
REXEC_FBC_UTF8_CLASS_SCAN(
reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target));
}
+ else if (ANYOF_FLAGS(c)) {
+ REXEC_FBC_CLASS_SCAN(reginclass(prog,c, (U8*)s, (U8*)s+1, 0));
+ }
else {
- REXEC_FBC_CLASS_SCAN(REGINCLASS(prog, c, (U8*)s, 0));
+ REXEC_FBC_CLASS_SCAN(ANYOF_BITMAP_TEST(c, *((U8*)s)));
}
break;
* trying that it will fail; so don't start a match past the
* required minimum number from the far end */
e = HOP3c(strend, -((SSize_t)ln), s);
-
- if (reginfo->intuit && e < s) {
- e = s; /* Due to minlen logic of intuit() */
- }
+ if (e < s)
+ break;
c1 = *pat_string;
c2 = fold_array[c1];
*/
e = HOP3c(strend, -((SSize_t)lnc), s);
- if (reginfo->intuit && e < s) {
- e = s; /* Due to minlen logic of intuit() */
- }
-
/* XXX Note that we could recalculate e to stop the loop earlier,
* as the worst case expansion above will rarely be met, and as we
* go along we would usually find that e moves further to the left.
goto do_boundu;
}
- FBC_BOUND(isWORDCHAR_LC, isWORDCHAR_LC_uvchr, isWORDCHAR_LC_utf8);
+ FBC_BOUND(isWORDCHAR_LC, isWORDCHAR_LC_uvchr, isWORDCHAR_LC_utf8_safe);
break;
case NBOUNDL:
goto do_nboundu;
}
- FBC_NBOUND(isWORDCHAR_LC, isWORDCHAR_LC_uvchr, isWORDCHAR_LC_utf8);
+ FBC_NBOUND(isWORDCHAR_LC, isWORDCHAR_LC_uvchr, isWORDCHAR_LC_utf8_safe);
break;
case BOUND: /* regcomp.c makes sure that this only has the traditional \b
meaning */
assert(FLAGS(c) == TRADITIONAL_BOUND);
- FBC_BOUND(isWORDCHAR, isWORDCHAR_uni, isWORDCHAR_utf8);
+ FBC_BOUND(isWORDCHAR, isWORDCHAR_uni, isWORDCHAR_utf8_safe);
break;
case BOUNDA: /* regcomp.c makes sure that this only has the traditional \b
meaning */
assert(FLAGS(c) == TRADITIONAL_BOUND);
- FBC_NBOUND(isWORDCHAR, isWORDCHAR_uni, isWORDCHAR_utf8);
+ FBC_NBOUND(isWORDCHAR, isWORDCHAR_uni, isWORDCHAR_utf8_safe);
break;
case NBOUNDA: /* regcomp.c makes sure that this only has the traditional \b
case NBOUNDU:
if ((bound_type) FLAGS(c) == TRADITIONAL_BOUND) {
- FBC_NBOUND(isWORDCHAR_L1, isWORDCHAR_uni, isWORDCHAR_utf8);
+ FBC_NBOUND(isWORDCHAR_L1, isWORDCHAR_uni, isWORDCHAR_utf8_safe);
break;
}
do_boundu:
switch((bound_type) FLAGS(c)) {
case TRADITIONAL_BOUND:
- FBC_BOUND(isWORDCHAR_L1, isWORDCHAR_uni, isWORDCHAR_utf8);
+ FBC_BOUND(isWORDCHAR_L1, isWORDCHAR_uni, isWORDCHAR_utf8_safe);
break;
case GCB_BOUND:
if (s == reginfo->strbeg) {
if (utf8_target) {
/* The complement of something that matches only ASCII matches all
* non-ASCII, plus everything in ASCII that isn't in the class. */
- REXEC_FBC_UTF8_CLASS_SCAN(! isASCII_utf8(s)
+ REXEC_FBC_UTF8_CLASS_SCAN( ! isASCII_utf8_safe(s, strend)
|| ! _generic_isCC_A(*s, FLAGS(c)));
break;
}
if ((UTF8_IS_INVARIANT(*s)
&& to_complement ^ cBOOL(_generic_isCC((U8) *s,
classnum)))
- || (UTF8_IS_DOWNGRADEABLE_START(*s)
+ || ( UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(s, strend)
&& to_complement ^ cBOOL(
_generic_isCC(EIGHT_BIT_UTF8_TO_NATIVE(*s,
*(s + 1)),
macros */
case _CC_ENUM_SPACE:
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(isSPACE_utf8(s)));
+ to_complement ^ cBOOL(isSPACE_utf8_safe(s, strend)));
break;
case _CC_ENUM_BLANK:
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(isBLANK_utf8(s)));
+ to_complement ^ cBOOL(isBLANK_utf8_safe(s, strend)));
break;
case _CC_ENUM_XDIGIT:
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(isXDIGIT_utf8(s)));
+ to_complement ^ cBOOL(isXDIGIT_utf8_safe(s, strend)));
break;
case _CC_ENUM_VERTSPACE:
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(isVERTWS_utf8(s)));
+ to_complement ^ cBOOL(isVERTWS_utf8_safe(s, strend)));
break;
case _CC_ENUM_CNTRL:
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(isCNTRL_utf8(s)));
+ to_complement ^ cBOOL(isCNTRL_utf8_safe(s, strend)));
break;
default:
* FBC macro instead of being expanded out. Since we've loaded the
* swash, we don't have to check for that each time through the loop */
REXEC_FBC_UTF8_CLASS_SCAN(
- to_complement ^ cBOOL(_generic_utf8(
+ to_complement ^ cBOOL(_generic_utf8_safe(
classnum,
s,
+ strend,
swash_fetch(PL_utf8_swash_ptrs[classnum],
(U8 *) s, TRUE))));
break;
dump_exec_pos( (char *)uc, c, strend,
real_start, s, utf8_target, 0);
Perl_re_printf( aTHX_
- " Charid:%3u CP:%4"UVxf" ",
+ " Charid:%3u CP:%4" UVxf " ",
charid, uvc);
});
}
dump_exec_pos( (char *)uc, c, strend, real_start,
s, utf8_target, 0 );
Perl_re_printf( aTHX_
- "%sState: %4"UVxf", word=%"UVxf,
+ "%sState: %4" UVxf ", word=%" UVxf,
failed ? " Fail transition to " : "",
(UV)state, (UV)word);
});
if (leftmost) {
s = (char*)leftmost;
DEBUG_TRIE_EXECUTE_r({
- Perl_re_printf( aTHX_ "Matches word #%"UVxf" at position %"IVdf". Trying full pattern...\n",
+ Perl_re_printf( aTHX_ "Matches word #%" UVxf " at position %" IVdf ". Trying full pattern...\n",
(UV)accepted_word, (IV)(s - real_start)
);
});
}
else {
/* create new COW SV to share string */
- RX_MATCH_COPY_FREE(rx);
+ RXp_MATCH_COPY_FREE(prog);
prog->saved_copy = sv_setsv_cow(prog->saved_copy, sv);
}
prog->subbeg = (char *)SvPVX_const(prog->saved_copy);
assert(min >= 0 && min <= max && min <= strend - strbeg);
sublen = max - min;
- if (RX_MATCH_COPIED(rx)) {
+ if (RXp_MATCH_COPIED(prog)) {
if (sublen > prog->sublen)
prog->subbeg =
(char*)saferealloc(prog->subbeg, sublen+1);
prog->subbeg[sublen] = '\0';
prog->suboffset = min;
prog->sublen = sublen;
- RX_MATCH_COPIED_on(rx);
+ RXp_MATCH_COPIED_on(prog);
}
prog->subcoffset = prog->suboffset;
if (prog->suboffset && utf8_target) {
}
}
else {
- RX_MATCH_COPY_FREE(rx);
+ RXp_MATCH_COPY_FREE(prog);
prog->subbeg = strbeg;
prog->suboffset = 0;
prog->subcoffset = 0;
: strbeg; /* pos() not defined; use start of string */
DEBUG_GPOS_r(Perl_re_printf( aTHX_
- "GPOS ganch set to strbeg[%"IVdf"]\n", (IV)(reginfo->ganch - strbeg)));
+ "GPOS ganch set to strbeg[%" IVdf "]\n", (IV)(reginfo->ganch - strbeg)));
/* in the presence of \G, we may need to start looking earlier in
* the string than the suggested start point of stringarg:
/* match via INTUIT shouldn't have any captures.
* Let @-, @+, $^N know */
prog->lastparen = prog->lastcloseparen = 0;
- RX_MATCH_UTF8_set(rx, utf8_target);
+ RXp_MATCH_UTF8_set(prog, utf8_target);
prog->offs[0].start = s - strbeg;
prog->offs[0].end = utf8_target
? (char*)utf8_hop((U8*)s, prog->minlenret) - strbeg
Perl_croak(aTHX_ "corrupted regexp program");
}
- RX_MATCH_TAINTED_off(rx);
- RX_MATCH_UTF8_set(rx, utf8_target);
+ RXp_MATCH_TAINTED_off(prog);
+ RXp_MATCH_UTF8_set(prog, utf8_target);
reginfo->prog = rx; /* Yes, sorry that this is confusing. */
reginfo->intuit = 0;
/* do we need a save destructor here for eval dies? */
Newxz(prog->offs, (prog->nparens + 1), regexp_paren_pair);
DEBUG_BUFFERS_r(Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" saving offs: orig=0x%"UVxf" new=0x%"UVxf"\n",
+ "rex=0x%" UVxf " saving offs: orig=0x%" UVxf " new=0x%" UVxf "\n",
0,
PTR2UV(prog),
PTR2UV(swap),
regprop(prog, prop, c, reginfo, NULL);
{
RE_PV_QUOTED_DECL(quoted,utf8_target,PERL_DEBUG_PAD_ZERO(1),
- s,strend-s,60);
+ s,strend-s,PL_dump_re_max_len);
Perl_re_printf( aTHX_
"Matching stclass %.*s against %s (%d bytes)\n",
(int)SvCUR(prop), SvPVX_const(prop),
DEBUG_BUFFERS_r(
if (swap)
Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" freeing offs: 0x%"UVxf"\n",
+ "rex=0x%" UVxf " freeing offs: 0x%" UVxf "\n",
0,
PTR2UV(prog),
PTR2UV(swap)
if (swap) {
/* we failed :-( roll it back */
DEBUG_BUFFERS_r(Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" rolling back offs: freeing=0x%"UVxf" restoring=0x%"UVxf"\n",
+ "rex=0x%" UVxf " rolling back offs: freeing=0x%" UVxf " restoring=0x%" UVxf "\n",
0,
PTR2UV(prog),
PTR2UV(prog->offs),
PerlIO *f= Perl_debug_log;
PERL_ARGS_ASSERT_RE_EXEC_INDENTF;
va_start(ap, depth);
- PerlIO_printf(f, "%*s|%4"UVuf"| %*s", REPORT_CODE_OFF, "", (UV)depth, INDENT_CHARS(depth), "" );
+ PerlIO_printf(f, "%*s|%4" UVuf "| %*s", REPORT_CODE_OFF, "", (UV)depth, INDENT_CHARS(depth), "" );
result = PerlIO_vprintf(f, fmt, ap);
va_end(ap);
return result;
reginitcolors();
{
RE_PV_QUOTED_DECL(s0, utf8_pat, PERL_DEBUG_PAD_ZERO(0),
- RX_PRECOMP_const(prog), RX_PRELEN(prog), 60);
+ RX_PRECOMP_const(prog), RX_PRELEN(prog), PL_dump_re_max_len);
RE_PV_QUOTED_DECL(s1, utf8_target, PERL_DEBUG_PAD_ZERO(1),
- start, end - start, 60);
+ start, end - start, PL_dump_re_max_len);
Perl_re_printf( aTHX_
"%s%s REx%s %s against %s\n",
const int is_uni = utf8_target ? 1 : 0;
RE_PV_COLOR_DECL(s0,len0,is_uni,PERL_DEBUG_PAD(0),
- (locinput - pref_len),pref0_len, 60, 4, 5);
+ (locinput - pref_len),pref0_len, PL_dump_re_max_len, 4, 5);
RE_PV_COLOR_DECL(s1,len1,is_uni,PERL_DEBUG_PAD(1),
(locinput - pref_len + pref0_len),
- pref_len - pref0_len, 60, 2, 3);
+ pref_len - pref0_len, PL_dump_re_max_len, 2, 3);
RE_PV_COLOR_DECL(s2,len2,is_uni,PERL_DEBUG_PAD(2),
locinput, loc_regeol - locinput, 10, 0, 1);
const STRLEN tlen=len0+len1+len2;
Perl_re_printf( aTHX_
- "%4"IVdf" <%.*s%.*s%s%.*s>%*s|%4u| ",
+ "%4" IVdf " <%.*s%.*s%s%.*s>%*s|%4u| ",
(IV)(locinput - loc_bostr),
len0, s0,
len1, s1,
}
else {
STRLEN len;
- _to_utf8_fold_flags(s,
- d,
- &len,
- FOLD_FLAGS_FULL | FOLD_FLAGS_LOCALE);
+ _toFOLD_utf8_flags(s,
+ pat_end,
+ d,
+ &len,
+ FOLD_FLAGS_FULL | FOLD_FLAGS_LOCALE);
d += len;
s += UTF8SKIP(s);
}
}
else { /* Does participate in folds */
AV* list = (AV*) *listp;
- if (av_tindex_nomg(list) != 1) {
+ if (av_tindex_skip_len_mg(list) != 1) {
/* If there aren't exactly two folds to this, it is
* outside the scope of this function */
/* Do not break within emoji flag sequences. That is, do not
* break between regional indicator (RI) symbols if there is an
* odd number of RI characters before the break point.
- * GB12 ^ (RI RI)* RI × RI
+ * GB12 sot (RI RI)* RI × RI
* GB13 [^RI] (RI RI)* RI × RI */
while (backup_one_GCB(strbeg,
* only if there are an even number of regional indicators
* preceding the position of the break.
*
- * sot (RI RI)* RI × RI
+ * sot (RI RI)* RI × RI
* [^RI] (RI RI)* RI × RI */
while (backup_one_LB(strbeg,
* odd number of RI characters before the potential break
* point.
*
- * WB15 ^ (RI RI)* RI × RI
+ * WB15 sot (RI RI)* RI × RI
* WB16 [^RI] (RI RI)* RI × RI */
while (backup_one_WB(&previous,
regnode *next;
U32 n = 0; /* general value; init to avoid compiler warning */
SSize_t ln = 0; /* len or last; init to avoid compiler warning */
+ SSize_t endref = 0; /* offset of end of backref when ln is start */
char *locinput = startpos;
char *pushinput; /* where to continue after a PUSH */
I32 nextchr; /* is always set to UCHARAT(locinput), or -1 at EOS */
SV *sv_yes_mark = NULL; /* last mark name we have seen
during a successful match */
U32 lastopen = 0; /* last open we saw */
- bool has_cutgroup = RX_HAS_CUTGROUP(rex) ? 1 : 0;
+ bool has_cutgroup = RXp_HAS_CUTGROUP(rex) ? 1 : 0;
SV* const oreplsv = GvSVn(PL_replgv);
/* these three flags are set by various ops to signal information to
* the very next op. They have a useful lifetime of exactly one loop
U8 gimme = G_SCALAR;
CV *caller_cv = NULL; /* who called us */
CV *last_pushed_cv = NULL; /* most recently called (?{}) CV */
- CHECKPOINT runops_cp; /* savestack position before executing EVAL */
U32 maxopenparen = 0; /* max '(' index seen so far */
int to_complement; /* Invert the result? */
_char_class_number classnum;
bool is_utf8_pat = reginfo->is_utf8_pat;
bool match = FALSE;
+ I32 orig_savestack_ix = PL_savestack_ix;
/* Solaris Studio 12.3 messes up fetching PL_charclass['\n'] */
#if (defined(__SUNPRO_C) && (__SUNPRO_C == 0x5120) && defined(__x86_64) && defined(USE_64_BIT_ALL))
DUMP_EXEC_POS( locinput, scan, utf8_target, depth );
regprop(rex, prop, scan, reginfo, NULL);
Perl_re_printf( aTHX_
- "%*s%"IVdf":%s(%"IVdf")\n",
+ "%*s%" IVdf ":%s(%" IVdf ")\n",
INDENT_CHARS(depth), "",
(IV)(scan - rexi->program),
SvPVX_const(prop),
if (scan->flags == EXACTL || scan->flags == EXACTFLU8) {
_CHECK_AND_WARN_PROBLEMATIC_LOCALE;
if (utf8_target
+ && nextchr >= 0 /* guard against negative EOS value in nextchr */
&& UTF8_IS_ABOVE_LATIN1(nextchr)
&& scan->flags == EXACTL)
{
DUMP_EXEC_POS( (char *)uc, scan, utf8_target, depth );
/* HERE */
PerlIO_printf( Perl_debug_log,
- "%*s%sState: %4"UVxf" Accepted: %c ",
+ "%*s%sState: %4" UVxf " Accepted: %c ",
INDENT_CHARS(depth), "", PL_colors[4],
(UV)state, (accepted ? 'Y' : 'N'));
});
}
DEBUG_TRIE_EXECUTE_r(
Perl_re_printf( aTHX_
- "Charid:%3x CP:%4"UVxf" After State: %4"UVxf"%s\n",
+ "Charid:%3x CP:%4" UVxf " After State: %4" UVxf "%s\n",
charid, uvc, (UV)state, PL_colors[5] );
);
}
}
DEBUG_EXECUTE_r(
- Perl_re_exec_indentf( aTHX_ "%sgot %"IVdf" possible matches%s\n",
+ Perl_re_exec_indentf( aTHX_ "%sgot %" IVdf " possible matches%s\n",
depth,
PL_colors[4], (IV)ST.accepted, PL_colors[5] );
);
{
U8 *uc;
if ( ST.jump ) {
+ /* undo any captures done in the tail part of a branch,
+ * e.g.
+ * /(?:X(.)(.)|Y(.)).../
+ * where the trie just matches X then calls out to do the
+ * rest of the branch */
REGCP_UNWIND(ST.cp);
UNWIND_PAREN(ST.lastparen, ST.lastcloseparen);
}
if (locinput == reginfo->strbeg)
b1 = isWORDCHAR_LC('\n');
else {
- b1 = isWORDCHAR_LC_utf8(reghop3((U8*)locinput, -1,
- (U8*)(reginfo->strbeg)));
+ b1 = isWORDCHAR_LC_utf8_safe(reghop3((U8*)locinput, -1,
+ (U8*)(reginfo->strbeg)),
+ (U8*)(reginfo->strend));
}
b2 = (NEXTCHR_IS_EOS)
? isWORDCHAR_LC('\n')
- : isWORDCHAR_LC_utf8((U8*)locinput);
+ : isWORDCHAR_LC_utf8_safe((U8*) locinput,
+ (U8*) reginfo->strend);
}
else { /* Here the string isn't utf8 */
b1 = (locinput == reginfo->strbeg)
bool b1, b2;
b1 = (locinput == reginfo->strbeg)
? 0 /* isWORDCHAR_L1('\n') */
- : isWORDCHAR_utf8(reghop3((U8*)locinput, -1,
- (U8*)(reginfo->strbeg)));
+ : isWORDCHAR_utf8_safe(
+ reghop3((U8*)locinput,
+ -1,
+ (U8*)(reginfo->strbeg)),
+ (U8*) reginfo->strend);
b2 = (NEXTCHR_IS_EOS)
? 0 /* isWORDCHAR_L1('\n') */
- : isWORDCHAR_utf8((U8*)locinput);
+ : isWORDCHAR_utf8_safe((U8*)locinput,
+ (U8*) reginfo->strend);
match = cBOOL(b1 != b2);
break;
}
break;
}
- if (! UTF8_IS_DOWNGRADEABLE_START(nextchr)) { /* An above Latin-1 code point */
- _CHECK_AND_OUTPUT_WIDE_LOCALE_UTF8_MSG(locinput, reginfo->strend);
+ if (! UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(locinput, reginfo->strend)) {
+ /* An above Latin-1 code point, or malformed */
+ _CHECK_AND_OUTPUT_WIDE_LOCALE_UTF8_MSG(locinput,
+ reginfo->strend);
goto utf8_posix_above_latin1;
}
}
locinput++;
}
- else if (UTF8_IS_DOWNGRADEABLE_START(nextchr)) {
+ else if (UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(locinput, reginfo->strend)) {
if (! (to_complement
^ cBOOL(_generic_isCC(EIGHT_BIT_UTF8_TO_NATIVE(nextchr,
*(locinput + 1)),
do_nref_ref_common:
ln = rex->offs[n].start;
+ endref = rex->offs[n].end;
reginfo->poscache_iter = reginfo->poscache_maxiter; /* Void cache */
- if (rex->lastparen < n || ln == -1)
+ if (rex->lastparen < n || ln == -1 || endref == -1)
sayNO; /* Do not match unless seen CLOSEn. */
- if (ln == rex->offs[n].end)
+ if (ln == endref)
break;
s = reginfo->strbeg + ln;
* not going off the end given by reginfo->strend, and
* returns in <limit> upon success, how much of the
* current input was matched */
- if (! foldEQ_utf8_flags(s, NULL, rex->offs[n].end - ln, utf8_target,
+ if (! foldEQ_utf8_flags(s, NULL, endref - ln, utf8_target,
locinput, &limit, 0, utf8_target, utf8_fold_flags))
{
sayNO;
(type == REF ||
UCHARAT(s) != fold_array[nextchr]))
sayNO;
- ln = rex->offs[n].end - ln;
+ ln = endref - ln;
if (locinput + ln > reginfo->strend)
sayNO;
if (ln > 1 && (type == REF
goto eval_recurse_doit;
/* NOTREACHED */
- case EVAL: /* /(?{A})B/ /(??{A})B/ and /(?(?{A})X|Y)B/ */
+ case EVAL: /* /(?{...})B/ /(??{A})B/ and /(?(?{...})X|Y)B/ */
if (cur_eval && cur_eval->locinput==locinput) {
if ( ++nochange_depth > max_nochange_depth )
Perl_croak(aTHX_ "EVAL without pos change exceeded limit in regex");
/* save *all* paren positions */
regcppush(rex, 0, maxopenparen);
- REGCP_SET(runops_cp);
+ REGCP_SET(ST.lastcp);
if (!caller_cv)
caller_cv = find_runcv(NULL);
nop = (OP*)rexi->data->data[n];
}
- /* normally if we're about to execute code from the same
- * CV that we used previously, we just use the existing
- * CX stack entry. However, its possible that in the
- * meantime we may have backtracked, popped from the save
- * stack, and undone the SAVECOMPPAD(s) associated with
- * PUSH_MULTICALL; in which case PL_comppad no longer
- * points to newcv's pad. */
+ /* Some notes about MULTICALL and the context and save stacks.
+ *
+ * In something like
+ * /...(?{ my $x)}...(?{ my $y)}...(?{ my $z)}.../
+ * since codeblocks don't introduce a new scope (so that
+ * local() etc accumulate), at the end of a successful
+ * match there will be a SAVEt_CLEARSV on the savestack
+ * for each of $x, $y, $z. If the three code blocks above
+ * happen to have come from different CVs (e.g. via
+ * embedded qr//s), then we must ensure that during any
+ * savestack unwinding, PL_comppad always points to the
+ * right pad at each moment. We achieve this by
+ * interleaving SAVEt_COMPPAD's on the savestack whenever
+ * there is a change of pad.
+ * In theory whenever we call a code block, we should
+ * push a CXt_SUB context, then pop it on return from
+ * that code block. This causes a bit of an issue in that
+ * normally popping a context also clears the savestack
+ * back to cx->blk_oldsaveix, but here we specifically
+ * don't want to clear the save stack on exit from the
+ * code block.
+ * Also for efficiency we don't want to keep pushing and
+ * popping the single SUB context as we backtrack etc.
+ * So instead, we push a single context the first time
+ * we need, it, then hang onto it until the end of this
+ * function. Whenever we encounter a new code block, we
+ * update the CV etc if that's changed. During the times
+ * in this function where we're not executing a code
+ * block, having the SUB context still there is a bit
+ * naughty - but we hope that no-one notices.
+ * When the SUB context is initially pushed, we fake up
+ * cx->blk_oldsaveix to be as if we'd pushed this context
+ * on first entry to S_regmatch rather than at some random
+ * point during the regexe execution. That way if we
+ * croak, popping the context stack will ensure that
+ * *everything* SAVEd by this function is undone and then
+ * the context popped, rather than e.g., popping the
+ * context (and restoring the original PL_comppad) then
+ * popping more of the savestack and restoring a bad
+ * PL_comppad.
+ */
+
+ /* If this is the first EVAL, push a MULTICALL. On
+ * subsequent calls, if we're executing a different CV, or
+ * if PL_comppad has got messed up from backtracking
+ * through SAVECOMPPADs, then refresh the context.
+ */
if (newcv != last_pushed_cv || PL_comppad != last_pad)
{
U8 flags = (CXp_SUB_RE |
((newcv == caller_cv) ? CXp_SUB_RE_FAKE : 0));
+ SAVECOMPPAD();
if (last_pushed_cv) {
- /* PUSH/POP_MULTICALL save and restore the
- * caller's PL_comppad; if we call multiple subs
- * using the same CX block, we have to save and
- * unwind the varying PL_comppad's ourselves,
- * especially restoring the right PL_comppad on
- * backtrack - so save it on the save stack */
- SAVECOMPPAD();
CHANGE_MULTICALL_FLAGS(newcv, flags);
}
else {
PUSH_MULTICALL_FLAGS(newcv, flags);
}
+ /* see notes above */
+ CX_CUR()->blk_oldsaveix = orig_savestack_ix;
+
last_pushed_cv = newcv;
}
else {
nop = nop->op_next;
DEBUG_STATE_r( Perl_re_printf( aTHX_
- " re EVAL PL_op=0x%"UVxf"\n", PTR2UV(nop)) );
+ " re EVAL PL_op=0x%" UVxf "\n", PTR2UV(nop)) );
rex->offs[0].end = locinput - reginfo->strbeg;
if (reginfo->info_aux_eval->pos_magic)
if (logical == 0) /* (?{})/ */
sv_setsv(save_scalar(PL_replgv), ret); /* $^R */
else if (logical == 1) { /* /(?(?{...})X|Y)/ */
- sw = cBOOL(SvTRUE(ret));
+ sw = cBOOL(SvTRUE_NN(ret));
logical = 0;
}
else { /* /(??{}) */
* in the regexp code uses the pad ! */
PL_op = oop;
PL_curcop = ocurcop;
- regcp_restore(rex, runops_cp, &maxopenparen);
+ regcp_restore(rex, ST.lastcp, &maxopenparen);
PL_curpm_under = PL_curpm;
PL_curpm = PL_reg_curpm;
- if (logical != 2)
- break;
+ if (logical != 2) {
+ PUSH_STATE_GOTO(EVAL_B, next, locinput);
+ /* NOTREACHED */
+ }
}
/* only /(??{})/ from now on */
ST.prev_eval = cur_eval;
cur_eval = st;
/* now continue from first node in postoned RE */
- PUSH_YES_STATE_GOTO(EVAL_AB, startpoint, locinput);
+ PUSH_YES_STATE_GOTO(EVAL_postponed_AB, startpoint, locinput);
NOT_REACHED; /* NOTREACHED */
}
- case EVAL_AB: /* cleanup after a successful (??{A})B */
+ case EVAL_postponed_AB: /* cleanup after a successful (??{A})B */
/* note: this is called twice; first after popping B, then A */
DEBUG_STACK_r({
Perl_re_exec_indentf( aTHX_ "EVAL_AB cur_eval=%p prev_eval=%p\n",
sayYES;
- case EVAL_AB_fail: /* unsuccessfully ran A or B in (??{A})B */
+ case EVAL_B_fail: /* unsuccessful B in (?{...})B */
+ REGCP_UNWIND(ST.lastcp);
+ sayNO;
+
+ case EVAL_postponed_AB_fail: /* unsuccessfully ran A or B in (??{A})B */
/* note: this is called twice; first after popping B, then A */
DEBUG_STACK_r({
Perl_re_exec_indentf( aTHX_ "EVAL_AB_fail cur_eval=%p prev_eval=%p\n",
if (n > maxopenparen)
maxopenparen = n;
DEBUG_BUFFERS_r(Perl_re_exec_indentf( aTHX_
- "rex=0x%"UVxf" offs=0x%"UVxf": \\%"UVuf": set %"IVdf" tmp; maxopenparen=%"UVuf"\n",
+ "rex=0x%" UVxf " offs=0x%" UVxf ": \\%" UVuf ": set %" IVdf " tmp; maxopenparen=%" UVuf "\n",
depth,
PTR2UV(rex),
PTR2UV(rex->offs),
rex->offs[n].start = rex->offs[n].start_tmp; \
rex->offs[n].end = locinput - reginfo->strbeg; \
DEBUG_BUFFERS_r(Perl_re_exec_indentf( aTHX_ \
- "rex=0x%"UVxf" offs=0x%"UVxf": \\%"UVuf": set %"IVdf"..%"IVdf"\n", \
+ "rex=0x%" UVxf " offs=0x%" UVxf ": \\%" UVuf ": set %" IVdf "..%" IVdf "\n", \
depth, \
PTR2UV(rex), \
PTR2UV(rex->offs), \
DEBUG_EXECUTE_r( Perl_re_exec_indentf( aTHX_ "whilem: (cache) already tried at this position...\n",
depth)
);
+ cur_curlyx->u.curlyx.count--;
sayNO; /* cache records failure */
}
ST.cache_offset = offset;
if (cur_curlyx->u.curlyx.minmod) {
ST.save_curlyx = cur_curlyx;
cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
- ST.cp = regcppush(rex, ST.save_curlyx->u.curlyx.parenfloor,
- maxopenparen);
- REGCP_SET(ST.lastcp);
PUSH_YES_STATE_GOTO(WHILEM_B_min, ST.save_curlyx->u.curlyx.B,
locinput);
NOT_REACHED; /* NOTREACHED */
CACHEsayNO;
NOT_REACHED; /* NOTREACHED */
- case WHILEM_A_min_fail: /* just failed to match A in a minimal match */
- /* FALLTHROUGH */
case WHILEM_A_pre_fail: /* just failed to match even minimal A */
REGCP_UNWIND(ST.lastcp);
regcppop(rex, &maxopenparen);
+ /* FALLTHROUGH */
+ case WHILEM_A_min_fail: /* just failed to match A in a minimal match */
cur_curlyx->u.curlyx.lastloc = ST.save_lastloc;
cur_curlyx->u.curlyx.count--;
CACHEsayNO;
case WHILEM_B_min_fail: /* just failed to match B in a minimal match */
cur_curlyx = ST.save_curlyx;
- REGCP_UNWIND(ST.lastcp);
- regcppop(rex, &maxopenparen);
if (cur_curlyx->u.curlyx.count >= /*max*/ARG2(cur_curlyx->u.curlyx.me)) {
/* Maximum greed exceeded */
);
/* Try grabbing another A and see if it helps. */
cur_curlyx->u.curlyx.lastloc = locinput;
- ST.cp = regcppush(rex, cur_curlyx->u.curlyx.parenfloor,
- maxopenparen);
- REGCP_SET(ST.lastcp);
PUSH_STATE_GOTO(WHILEM_A_min,
/*A*/ NEXTOPER(ST.save_curlyx->u.curlyx.me) + EXTRA_STEP_2ARGS,
locinput);
ST.count = ST.minmod ? ARG1(ST.me) : ARG2(ST.me);
}
DEBUG_EXECUTE_r(
- Perl_re_exec_indentf( aTHX_ "CURLYM now matched %"IVdf" times, len=%"IVdf"...\n",
+ Perl_re_exec_indentf( aTHX_ "CURLYM now matched %" IVdf " times, len=%" IVdf "...\n",
depth, (IV) ST.count, (IV)ST.alen)
);
}
DEBUG_EXECUTE_r(
- Perl_re_exec_indentf( aTHX_ "CURLYM trying tail with matches=%"IVdf"...\n",
+ Perl_re_exec_indentf( aTHX_ "CURLYM trying tail with matches=%" IVdf "...\n",
depth, (IV)ST.count)
);
if (! NEXTCHR_IS_EOS && ST.c1 != CHRTEST_VOID) {
{
/* simulate B failing */
DEBUG_OPTIMISE_r(
- Perl_re_exec_indentf( aTHX_ "CURLYM Fast bail next target=0x%"UVXf" c1=0x%"UVXf" c2=0x%"UVXf"\n",
+ Perl_re_exec_indentf( aTHX_ "CURLYM Fast bail next target=0x%" UVXf " c1=0x%" UVXf " c2=0x%" UVXf "\n",
depth,
valid_utf8_to_uvchr((U8 *) locinput, NULL),
valid_utf8_to_uvchr(ST.c1_utf8, NULL),
SET_RECURSE_LOCINPUT("FAKE-END[after]", cur_eval->locinput);
- PUSH_YES_STATE_GOTO(EVAL_AB, st->u.eval.prev_eval->u.eval.B,
+ PUSH_YES_STATE_GOTO(EVAL_postponed_AB, st->u.eval.prev_eval->u.eval.B,
locinput); /* match B */
}
sv_commit = ST.mark_name;
DEBUG_EXECUTE_r({
- Perl_re_exec_indentf( aTHX_ "%ssetting cutpoint to mark:%"SVf"...%s\n",
+ Perl_re_exec_indentf( aTHX_ "%ssetting cutpoint to mark:%" SVf "...%s\n",
depth,
PL_colors[4], SVfARG(sv_commit), PL_colors[5]);
});
break;
default:
- PerlIO_printf(Perl_error_log, "%"UVxf" %d\n",
+ PerlIO_printf(Perl_error_log, "%" UVxf " %d\n",
PTR2UV(scan), OP(scan));
Perl_croak(aTHX_ "regexp memory corruption");
assert(!NEXTCHR_IS_EOS);
if (utf8_target) {
locinput += PL_utf8skip[nextchr];
- /* locinput is allowed to go 1 char off the end, but not 2+ */
+ /* locinput is allowed to go 1 char off the end (signifying
+ * EOS), but not 2+ */
if (locinput > reginfo->strend)
sayNO;
}
DEBUG_STACK_r({
regmatch_state *cur = st;
regmatch_state *curyes = yes_state;
- int curd = depth;
+ U32 i;
regmatch_slab *slab = PL_regmatch_slab;
- for (;curd > -1 && (depth-curd < 3);cur--,curd--) {
+ for (i = 0; i < 3 && i <= depth; cur--,i++) {
if (cur < SLAB_FIRST(slab)) {
slab = slab->prev;
cur = SLAB_LAST(slab);
}
- Perl_re_exec_indentf( aTHX_ "#%-3d %-10s %s\n",
+ Perl_re_exec_indentf( aTHX_ "%4s #%-3d %-10s %s\n",
depth,
- curd, PL_reg_name[cur->resume_state],
+ i ? " " : "push",
+ depth - i, PL_reg_name[cur->resume_state],
(curyes == cur) ? "yes" : ""
);
if (curyes == cur)
if (last_pushed_cv) {
dSP;
+ /* see "Some notes about MULTICALL" above */
POP_MULTICALL;
PERL_UNUSED_VAR(SP);
}
+ else
+ LEAVE_SCOPE(orig_savestack_ix);
assert(!result || locinput - reginfo->strbeg >= 0);
return result ? locinput - reginfo->strbeg : -1;
scan += UTF8SKIP(scan);
hardcount++;
}
- } else {
- while (scan < loceol && REGINCLASS(prog, p, (U8*)scan, 0))
+ }
+ else if (ANYOF_FLAGS(p)) {
+ while (scan < loceol
+ && reginclass(prog, p, (U8*)scan, (U8*)scan+1, 0))
+ scan++;
+ }
+ else {
+ while (scan < loceol && ANYOF_BITMAP_TEST(p, *((U8*)scan)))
scan++;
}
break;
/* The complement of something that matches only ASCII matches all
* non-ASCII, plus everything in ASCII that isn't in the class. */
while (hardcount < max && scan < loceol
- && (! isASCII_utf8(scan)
+ && ( ! isASCII_utf8_safe(scan, reginfo->strend)
|| ! _generic_isCC_A((U8) *scan, FLAGS(p))))
{
scan += UTF8SKIP(scan);
case _CC_ENUM_SPACE:
while (hardcount < max
&& scan < loceol
- && (to_complement ^ cBOOL(isSPACE_utf8(scan))))
+ && (to_complement
+ ^ cBOOL(isSPACE_utf8_safe(scan, loceol))))
{
scan += UTF8SKIP(scan);
hardcount++;
case _CC_ENUM_BLANK:
while (hardcount < max
&& scan < loceol
- && (to_complement ^ cBOOL(isBLANK_utf8(scan))))
+ && (to_complement
+ ^ cBOOL(isBLANK_utf8_safe(scan, loceol))))
{
scan += UTF8SKIP(scan);
hardcount++;
case _CC_ENUM_XDIGIT:
while (hardcount < max
&& scan < loceol
- && (to_complement ^ cBOOL(isXDIGIT_utf8(scan))))
+ && (to_complement
+ ^ cBOOL(isXDIGIT_utf8_safe(scan, loceol))))
{
scan += UTF8SKIP(scan);
hardcount++;
case _CC_ENUM_VERTSPACE:
while (hardcount < max
&& scan < loceol
- && (to_complement ^ cBOOL(isVERTWS_utf8(scan))))
+ && (to_complement
+ ^ cBOOL(isVERTWS_utf8_safe(scan, loceol))))
{
scan += UTF8SKIP(scan);
hardcount++;
case _CC_ENUM_CNTRL:
while (hardcount < max
&& scan < loceol
- && (to_complement ^ cBOOL(isCNTRL_utf8(scan))))
+ && (to_complement
+ ^ cBOOL(isCNTRL_utf8_safe(scan, loceol))))
{
scan += UTF8SKIP(scan);
hardcount++;
}
while (hardcount < max && scan < loceol
- && to_complement ^ cBOOL(_generic_utf8(
+ && to_complement ^ cBOOL(_generic_utf8_safe(
classnum,
scan,
+ loceol,
swash_fetch(PL_utf8_swash_ptrs[classnum],
(U8 *) scan,
TRUE))))
DEBUG_EXECUTE_r({
SV * const prop = sv_newmortal();
regprop(prog, prop, p, reginfo, NULL);
- Perl_re_exec_indentf( aTHX_ "%s can match %"IVdf" times out of %"IVdf"...\n",
+ Perl_re_exec_indentf( aTHX_ "%s can match %" IVdf " times out of %" IVdf "...\n",
depth, SvPVX_const(prop),(IV)c,(IV)max);
});
});
* UTF8_IS_INVARIANT() works even if not in UTF-8 */
if (! UTF8_IS_INVARIANT(c) && utf8_target) {
STRLEN c_len = 0;
- c = utf8n_to_uvchr(p, p_end - p, &c_len,
- (UTF8_ALLOW_DEFAULT & UTF8_ALLOW_ANYUV)
- | UTF8_ALLOW_FFFF | UTF8_CHECK_ONLY);
- /* see [perl #37836] for UTF8_ALLOW_ANYUV; [perl #38293] for
- * UTF8_ALLOW_FFFF */
- if (c_len == (STRLEN)-1)
- Perl_croak(aTHX_ "Malformed UTF-8 character (fatal)");
+ const U32 utf8n_flags = UTF8_ALLOW_DEFAULT;
+ c = utf8n_to_uvchr(p, p_end - p, &c_len, utf8n_flags | UTF8_CHECK_ONLY);
+ if (c_len == (STRLEN)-1) {
+ _force_out_malformed_utf8_message(p, p_end,
+ utf8n_flags,
+ 1 /* 1 means die */ );
+ NOT_REACHED; /* NOTREACHED */
+ }
if (c > 255 && OP(n) == ANYOFL && ! ANYOFL_UTF8_LOCALE_REQD(flags)) {
_CHECK_AND_OUTPUT_WIDE_LOCALE_CP_MSG(c);
}
&& ckWARN_d(WARN_NON_UNICODE))
{
Perl_warner(aTHX_ packWARN(WARN_NON_UNICODE),
- "Matched non-Unicode code point 0x%04"UVXf" against Unicode property; may not be portable", c);
+ "Matched non-Unicode code point 0x%04" UVXf " against Unicode property; may not be portable", c);
}
}
return TRUE;
}
+#ifndef PERL_IN_XSUB_RE
+
+bool
+Perl__is_grapheme(pTHX_ const U8 * strbeg, const U8 * s, const U8 * strend, const UV cp)
+{
+ /* Temporary helper function for toke.c. Verify that the code point 'cp'
+ * is a stand-alone grapheme. The UTF-8 for 'cp' begins at position 's' in
+ * the larger string bounded by 'strbeg' and 'strend'.
+ *
+ * 'cp' needs to be assigned (if not a future version of the Unicode
+ * Standard could make it something that combines with adjacent characters,
+ * so code using it would then break), and there has to be a GCB break
+ * before and after the character. */
+
+ GCB_enum cp_gcb_val, prev_cp_gcb_val, next_cp_gcb_val;
+ const U8 * prev_cp_start;
+
+ PERL_ARGS_ASSERT__IS_GRAPHEME;
+
+ /* Unassigned code points are forbidden */
+ if (UNLIKELY(! ELEMENT_RANGE_MATCHES_INVLIST(
+ _invlist_search(PL_Assigned_invlist, cp))))
+ {
+ return FALSE;
+ }
+
+ cp_gcb_val = getGCB_VAL_CP(cp);
+
+ /* Find the GCB value of the previous code point in the input */
+ prev_cp_start = utf8_hop_back(s, -1, strbeg);
+ if (UNLIKELY(prev_cp_start == s)) {
+ prev_cp_gcb_val = GCB_EDGE;
+ }
+ else {
+ prev_cp_gcb_val = getGCB_VAL_UTF8(prev_cp_start, strend);
+ }
+
+ /* And check that is a grapheme boundary */
+ if (! isGCB(prev_cp_gcb_val, cp_gcb_val, strbeg, s,
+ TRUE /* is UTF-8 encoded */ ))
+ {
+ return FALSE;
+ }
+
+ /* Similarly verify there is a break between the current character and the
+ * following one */
+ s += UTF8SKIP(s);
+ if (s >= strend) {
+ next_cp_gcb_val = GCB_EDGE;
+ }
+ else {
+ next_cp_gcb_val = getGCB_VAL_UTF8(s, strend);
+ }
+
+ return isGCB(cp_gcb_val, next_cp_gcb_val, strbeg, s, TRUE);
+}
+
+#endif
+
+
+
/*
* ex: set ts=8 sts=4 sw=4 et:
*/