/* for use after a quantifier and before an EXACT-like node -- japhy */
#define JUMPABLE(rn) ( \
OP(rn) == OPEN || OP(rn) == CLOSE || OP(rn) == EVAL || \
- OP(rn) == SUSPEND || OP(rn) == IFMATCH \
+ OP(rn) == SUSPEND || OP(rn) == IFMATCH || \
+ OP(rn) == PLUS || OP(rn) == MINMOD || \
+ (PL_regkind[(U8)OP(rn)] == CURLY && ARG1(rn) > 0) \
)
-#define NEAR_EXACT(rn) (PL_regkind[(U8)OP(rn)] == EXACT || JUMPABLE(rn))
+#define HAS_TEXT(rn) ( \
+ PL_regkind[(U8)OP(rn)] == EXACT || PL_regkind[(U8)OP(rn)] == REF \
+)
-#define NEXT_IMPT(rn) STMT_START { \
+#define FIND_NEXT_IMPT(rn) STMT_START { \
while (JUMPABLE(rn)) \
- if (OP(rn) == SUSPEND || OP(rn) == IFMATCH) \
+ if (OP(rn) == SUSPEND || OP(rn) == IFMATCH || \
+ PL_regkind[(U8)OP(rn)] == CURLY) \
rn = NEXTOPER(NEXTOPER(rn)); \
+ else if (OP(rn) == PLUS) \
+ rn = NEXTOPER(rn); \
else rn += NEXT_OFF(rn); \
} STMT_END
PL_reg_flags |= RF_utf8;
if (prog->minlen > CHR_DIST((U8*)strend, (U8*)strpos)) {
- DEBUG_r(PerlIO_printf(Perl_debug_log, "String too short...\n"));
+ DEBUG_r(PerlIO_printf(Perl_debug_log,
+ "String too short... [re_intuit_start]\n"));
goto fail;
}
strbeg = (sv && SvPOK(sv)) ? strend - SvCUR(sv) : strpos;
}
minlen = prog->minlen;
- if (do_utf8 && !(prog->reganch & ROPT_CANY_SEEN)) {
- if (utf8_distance((U8*)strend, (U8*)startpos) < minlen) {
- DEBUG_r(PerlIO_printf(Perl_debug_log,
- "Too short (in characters)...\n"));
- goto phooey;
- }
- }
- else {
- if (strend - startpos < minlen) {
- DEBUG_r(PerlIO_printf(Perl_debug_log,
- "Too short (in bytes)...\n"));
- goto phooey;
- }
+ if (strend - startpos < minlen) {
+ DEBUG_r(PerlIO_printf(Perl_debug_log,
+ "String too short [regexec_flags]...\n"));
+ goto phooey;
}
/* Check validity of program. */
s = STRING(scan);
ln = STR_LEN(scan);
if (do_utf8 != (UTF!=0)) {
+ /* The target and the pattern have differing "utf8ness". */
char *l = locinput;
char *e = s + ln;
STRLEN len;
- if (do_utf8)
+
+ if (do_utf8) {
+ /* The target is utf8, the pattern is not utf8. */
while (s < e) {
if (l >= PL_regeol)
- sayNO;
- if (*((U8*)s) != utf8_to_uvchr((U8*)l, &len))
- sayNO;
- s += len;
+ sayNO;
+ if (NATIVE_TO_UNI(*(U8*)s) !=
+ utf8_to_uvchr((U8*)l, &len))
+ sayNO;
l += len;
+ s ++;
}
- else
+ }
+ else {
+ /* The target is not utf8, the pattern is utf8. */
while (s < e) {
if (l >= PL_regeol)
sayNO;
- if (*((U8*)l) != utf8_to_uvchr((U8*)s, &len))
+ if (NATIVE_TO_UNI(*((U8*)l)) !=
+ utf8_to_uvchr((U8*)s, &len))
sayNO;
- s+=len;
- l++;
+ s += len;
+ l ++;
}
+ }
locinput = l;
nextchr = UCHARAT(locinput);
break;
}
+ /* The target and the pattern have the same "utf8ness". */
/* Inline the first character, for speed. */
if (UCHARAT(s) != nextchr)
sayNO;
if (ln && l == 0)
n = ln; /* don't backtrack */
locinput = PL_reginput;
- if (NEAR_EXACT(next)) {
+ if (HAS_TEXT(next) || JUMPABLE(next)) {
regnode *text_node = next;
- if (PL_regkind[(U8)OP(next)] != EXACT)
- NEXT_IMPT(text_node);
+ if (! HAS_TEXT(text_node)) FIND_NEXT_IMPT(text_node);
- if (PL_regkind[(U8)OP(text_node)] != EXACT) {
- c1 = c2 = -1000;
- }
+ if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
- c1 = (U8)*STRING(text_node);
+ 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 (
+ *PL_reglastparen < n ||
+ ln == -1 ||
+ ln == PL_regendp[n]
+ ) {
+ c1 = c2 = -1000;
+ goto assume_ok_MM;
+ }
+ c1 = *(PL_bostr + ln);
+ }
+ else { c1 = (U8)*STRING(text_node); }
if (OP(next) == EXACTF)
c2 = PL_fold[c1];
else if (OP(text_node) == EXACTFL)
}
else
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 ? */
(IV) n, (IV)l)
);
if (n >= ln) {
- if (NEAR_EXACT(next)) {
+ if (HAS_TEXT(next) || JUMPABLE(next)) {
regnode *text_node = next;
- if (PL_regkind[(U8)OP(next)] != EXACT)
- NEXT_IMPT(text_node);
+ if (! HAS_TEXT(text_node)) FIND_NEXT_IMPT(text_node);
- if (PL_regkind[(U8)OP(text_node)] != EXACT) {
- c1 = c2 = -1000;
- }
+ if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
- c1 = (U8)*STRING(text_node);
+ 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 (
+ *PL_reglastparen < n ||
+ ln == -1 ||
+ ln == PL_regendp[n]
+ ) {
+ c1 = c2 = -1000;
+ goto assume_ok_REG;
+ }
+ c1 = *(PL_bostr + ln);
+ }
+ else { c1 = (U8)*STRING(text_node); }
+
if (OP(text_node) == EXACTF)
c2 = PL_fold[c1];
else if (OP(text_node) == EXACTFL)
else
c1 = c2 = -1000;
}
+ assume_ok_REG:
REGCP_SET(lastcp);
while (n >= ln) {
/* If it could work, try it. */
* of the quantifier and the EXACT-like node. -- japhy
*/
- if (NEAR_EXACT(next)) {
+ if (HAS_TEXT(next) || JUMPABLE(next)) {
U8 *s;
regnode *text_node = next;
- if (PL_regkind[(U8)OP(next)] != EXACT)
- NEXT_IMPT(text_node);
+ if (! HAS_TEXT(text_node)) FIND_NEXT_IMPT(text_node);
- if (PL_regkind[(U8)OP(text_node)] != EXACT) {
- c1 = c2 = -1000;
- }
+ if (! HAS_TEXT(text_node)) c1 = c2 = -1000;
else {
- s = (U8*)STRING(text_node);
+ 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 (
+ *PL_reglastparen < n ||
+ ln == -1 ||
+ ln == PL_regendp[n]
+ ) {
+ c1 = c2 = -1000;
+ goto assume_ok_easy;
+ }
+ s = PL_bostr + ln;
+ }
+ else { s = (U8*)STRING(text_node); }
if (!UTF) {
c2 = c1 = *s;
}
else
c1 = c2 = -1000;
+ assume_ok_easy:
PL_reginput = locinput;
if (minmod) {
CHECKPOINT lastcp;
projects / p5sagit/p5-mst-13.2.git / blobdiff