* didn't contradict, so just retry the anchored "other"
* substr */
DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log,
- " Found /%s^%s/m, rescanning for anchored from offset %ld (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],
- (long)(rx_origin - strbeg + prog->anchored_offset),
- (long)(rx_origin - strbeg)
+ (IV)(rx_origin - strbeg + prog->anchored_offset),
+ (IV)(rx_origin - strbeg)
));
goto do_other_substr;
}
/* FALLTHROUGH */
case BOUNDL: /* /\b/l */
+ {
+ bool b1, b2;
_CHECK_AND_WARN_PROBLEMATIC_LOCALE;
if (FLAGS(scan) != TRADITIONAL_BOUND) {
if (utf8_target) {
if (locinput == reginfo->strbeg)
- ln = isWORDCHAR_LC('\n');
+ b1 = isWORDCHAR_LC('\n');
else {
- ln = isWORDCHAR_LC_utf8(reghop3((U8*)locinput, -1,
+ b1 = isWORDCHAR_LC_utf8(reghop3((U8*)locinput, -1,
(U8*)(reginfo->strbeg)));
}
- n = (NEXTCHR_IS_EOS)
+ b2 = (NEXTCHR_IS_EOS)
? isWORDCHAR_LC('\n')
: isWORDCHAR_LC_utf8((U8*)locinput);
}
else { /* Here the string isn't utf8 */
- ln = (locinput == reginfo->strbeg)
+ b1 = (locinput == reginfo->strbeg)
? isWORDCHAR_LC('\n')
: isWORDCHAR_LC(UCHARAT(locinput - 1));
- n = (NEXTCHR_IS_EOS)
+ b2 = (NEXTCHR_IS_EOS)
? isWORDCHAR_LC('\n')
: isWORDCHAR_LC(nextchr);
}
- if (to_complement ^ (ln == n)) {
+ if (to_complement ^ (b1 == b2)) {
sayNO;
}
break;
+ }
case NBOUND: /* /\B/ */
to_complement = 1;
/* FALLTHROUGH */
case BOUNDA: /* /\b/a */
+ {
+ bool b1, b2;
bound_ascii_match_only:
/* Here the string isn't utf8, or is utf8 and only ascii characters
* 2) it is a multi-byte character, in which case the final byte is
* never mistakable for ASCII, and so the test will say it is
* not a word character, which is the correct answer. */
- ln = (locinput == reginfo->strbeg)
+ b1 = (locinput == reginfo->strbeg)
? isWORDCHAR_A('\n')
: isWORDCHAR_A(UCHARAT(locinput - 1));
- n = (NEXTCHR_IS_EOS)
+ b2 = (NEXTCHR_IS_EOS)
? isWORDCHAR_A('\n')
: isWORDCHAR_A(nextchr);
- if (to_complement ^ (ln == n)) {
+ if (to_complement ^ (b1 == b2)) {
sayNO;
}
break;
+ }
case NBOUNDU: /* /\B/u */
to_complement = 1;
bound_utf8:
switch((bound_type) FLAGS(scan)) {
case TRADITIONAL_BOUND:
- ln = (locinput == reginfo->strbeg)
+ {
+ bool b1, b2;
+ b1 = (locinput == reginfo->strbeg)
? 0 /* isWORDCHAR_L1('\n') */
: isWORDCHAR_utf8(reghop3((U8*)locinput, -1,
(U8*)(reginfo->strbeg)));
- n = (NEXTCHR_IS_EOS)
+ b2 = (NEXTCHR_IS_EOS)
? 0 /* isWORDCHAR_L1('\n') */
: isWORDCHAR_utf8((U8*)locinput);
- match = cBOOL(ln != n);
+ match = cBOOL(b1 != b2);
break;
+ }
case GCB_BOUND:
if (locinput == reginfo->strbeg || NEXTCHR_IS_EOS) {
match = TRUE; /* GCB always matches at begin and
else { /* Not utf8 target */
switch((bound_type) FLAGS(scan)) {
case TRADITIONAL_BOUND:
- ln = (locinput == reginfo->strbeg)
+ {
+ bool b1, b2;
+ b1 = (locinput == reginfo->strbeg)
? 0 /* isWORDCHAR_L1('\n') */
: isWORDCHAR_L1(UCHARAT(locinput - 1));
- n = (NEXTCHR_IS_EOS)
+ b2 = (NEXTCHR_IS_EOS)
? 0 /* isWORDCHAR_L1('\n') */
: isWORDCHAR_L1(nextchr);
- match = cBOOL(ln != n);
+ match = cBOOL(b1 != b2);
break;
+ }
case GCB_BOUND:
if (locinput == reginfo->strbeg || NEXTCHR_IS_EOS) {
break;
case ACCEPT: /* (*ACCEPT) */
- if (ARG(scan)){
+ if (scan->flags)
+ sv_yes_mark = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
+ if (ARG2L(scan)){
regnode *cursor;
for (cursor=scan;
cursor && OP(cursor)!=END;
NOT_REACHED; /* NOTREACHED */
case CUTGROUP: /* /(*THEN)/ */
- sv_yes_mark = st->u.mark.mark_name = scan->flags ? NULL :
- MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
+ sv_yes_mark = st->u.mark.mark_name = scan->flags
+ ? MUTABLE_SV(rexi->data->data[ ARG( scan ) ])
+ : NULL;
PUSH_STATE_GOTO(CUTGROUP_next, next, locinput);
/* NOTREACHED */
NOT_REACHED; /* NOTREACHED */
/* FALLTHROUGH */
case PRUNE: /* (*PRUNE) */
- if (!scan->flags)
+ if (scan->flags)
sv_yes_mark = sv_commit = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
PUSH_STATE_GOTO(COMMIT_next, next, locinput);
/* NOTREACHED */
case COMMIT_next_fail:
no_final = 1;
/* FALLTHROUGH */
+ sayNO;
+ NOT_REACHED; /* NOTREACHED */
case OPFAIL: /* (*FAIL) */
- sayNO;
+ if (scan->flags)
+ sv_commit = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
+ if (logical) {
+ /* deal with (?(?!)X|Y) properly,
+ * make sure we trigger the no branch
+ * of the trailing IFTHEN structure*/
+ sw= 0;
+ break;
+ } else {
+ sayNO;
+ }
/* NOTREACHED */
NOT_REACHED; /* NOTREACHED */
NOT_REACHED; /* NOTREACHED */
case SKIP: /* (*SKIP) */
- if (scan->flags) {
+ if (!scan->flags) {
/* (*SKIP) : if we fail we cut here*/
ST.mark_name = NULL;
ST.mark_loc = locinput;
if (UTF8_IS_CONTINUED(*s)) {
while (s > lim && UTF8_IS_CONTINUATION(*s))
s--;
+ if (! UTF8_IS_START(*s)) {
+ Perl_croak_nocontext("Malformed UTF-8 character (fatal)");
+ }
}
/* XXX could check well-formedness here */
}
if (UTF8_IS_CONTINUED(*s)) {
while (s > llim && UTF8_IS_CONTINUATION(*s))
s--;
+ if (! UTF8_IS_START(*s)) {
+ Perl_croak_nocontext("Malformed UTF-8 character (fatal)");
+ }
}
/* XXX could check well-formedness here */
}
if (UTF8_IS_CONTINUED(*s)) {
while (s > lim && UTF8_IS_CONTINUATION(*s))
s--;
+ if (! UTF8_IS_START(*s)) {
+ Perl_croak_nocontext("Malformed UTF-8 character (fatal)");
+ }
}
/* XXX could check well-formedness here */
}