/* Currently these are only used when PL_regkind[OP(rn)] == EXACT so
we don't need this definition. */
#define IS_TEXT(rn) ( OP(rn)==EXACT || OP(rn)==REF || OP(rn)==NREF )
-#define IS_TEXTF(rn) ( OP(rn)==EXACTF || OP(rn)==REFF || OP(rn)==NREFF )
+#define IS_TEXTF(rn) ( (OP(rn)==EXACTFU || OP(rn)==EXACTF) || OP(rn)==REFF || OP(rn)==NREFF )
#define IS_TEXTFL(rn) ( OP(rn)==EXACTFL || OP(rn)==REFFL || OP(rn)==NREFFL )
#else
/* ... so we use this as its faster. */
#define IS_TEXT(rn) ( OP(rn)==EXACT )
+#define IS_TEXTFU(rn) ( OP(rn)==EXACTFU )
#define IS_TEXTF(rn) ( OP(rn)==EXACTF )
#define IS_TEXTFL(rn) ( OP(rn)==EXACTFL )
even for \b or \B. But (minlen? 1 : 0) below assumes that
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(). */
+ This leaves EXACTF, EXACTFU only, which are dealt with in find_byclass(). */
const U8* const str = (U8*)STRING(progi->regstclass);
const int cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
? CHR_DIST(str+STR_LEN(progi->regstclass), str)
#define REXEC_FBC_EXACTISH_SCAN(CoNd) \
STMT_START { \
+ re_fold_t folder; \
+ switch (OP(c)) { \
+ case EXACTFU: folder = foldEQ_latin1; break; \
+ case EXACTFL: folder = foldEQ_locale; break; \
+ case EXACTF: folder = foldEQ; break; \
+ default: \
+ Perl_croak(aTHX_ "panic: Unexpected op %u", OP(c)); \
+ } \
while (s <= e) { \
if ( (CoNd) \
- && (ln == 1 || (OP(c) == EXACTF \
- ? foldEQ(s, m, ln) \
- : foldEQ_locale(s, m, ln))) \
- && (!reginfo || regtry(reginfo, &s)) ) \
+ && (ln == 1 || folder(s, m, ln)) \
+ && (!reginfo || regtry(reginfo, &s)) ) \
goto got_it; \
s++; \
} \
tmp = doevery;
);
break;
+ case EXACTFU:
case EXACTF:
m = STRING(c);
ln = STR_LEN(c); /* length to match in octets/bytes */
}
else {
c1 = *(U8*)m;
- c2 = PL_fold[c1];
+ if (utf8_target || OP(c) == EXACTFU) {
+
+ /* Micro sign folds to GREEK SMALL LETTER MU;
+ LATIN_SMALL_LETTER_SHARP_S folds to 'ss', and this sets
+ c2 to the first 's' of the pair, and the code below will
+ look for others */
+ c2 = (c1 == MICRO_SIGN)
+ ? GREEK_SMALL_LETTER_MU
+ : (c1 == LATIN_SMALL_LETTER_SHARP_S)
+ ? 's'
+ : PL_fold_latin1[c1];
+ } else c2 = PL_fold[c1];
}
goto do_exactf;
case EXACTFL:
nextchr = UCHARAT(locinput);
break;
}
- case EXACTFL:
+ case EXACTFL: {
+ re_fold_t folder;
+ const U8 * fold_array;
+ const char * s;
+
PL_reg_flags |= RF_tainted;
- /* FALL THROUGH */
- case EXACTF: {
- char * const s = STRING(scan);
+ folder = foldEQ_locale;
+ fold_array = PL_fold_locale;
+ goto do_exactf;
+
+ case EXACTFU:
+ folder = foldEQ_latin1;
+ fold_array = PL_fold_latin1;
+ goto do_exactf;
+
+ case EXACTF:
+ folder = foldEQ;
+ fold_array = PL_fold;
+
+ do_exactf:
+ s = STRING(scan);
ln = STR_LEN(scan);
if (utf8_target || UTF_PATTERN) {
/* Inline the first character, for speed. */
if (UCHARAT(s) != nextchr &&
- UCHARAT(s) != ((OP(scan) == EXACTF)
- ? PL_fold : PL_fold_locale)[nextchr])
+ UCHARAT(s) != fold_array[nextchr])
+ {
sayNO;
+ }
if (PL_regeol - locinput < ln)
sayNO;
- if (ln > 1 && (OP(scan) == EXACTF
- ? ! foldEQ(s, locinput, ln)
- : ! foldEQ_locale(s, locinput, ln)))
+ if (ln > 1 && ! folder(s, locinput, ln))
sayNO;
locinput += ln;
nextchr = UCHARAT(locinput);
break;
- }
+ }
case BOUNDL:
case NBOUNDL:
PL_reg_flags |= RF_tainted;
break;
case NREFFL:
- {
+ { /* The capture buffer cases. The ones beginning with N for the
+ named buffers just convert to the equivalent numbered and
+ pretend they were called as the corresponding numbered buffer
+ op. */
+ /* don't initialize these, it makes C++ unhappy */
char *s;
char type;
+ re_fold_t folder;
+ const U8 *fold_array;
+
+ folder = NULL; /* NULL assumes will be NREF, REF: no
+ folding */
+ fold_array = NULL;
+
PL_reg_flags |= RF_tainted;
- /* FALL THROUGH */
- case NREF:
+ folder = foldEQ_locale;
+ fold_array = PL_fold_locale;
+ type = REFFL;
+ goto do_nref;
+
+ case NREFFU:
+ folder = foldEQ_latin1;
+ fold_array = PL_fold_latin1;
+ type = REFFU;
+ goto do_nref;
+
case NREFF:
- type = OP(scan);
+ folder = foldEQ;
+ fold_array = PL_fold;
+ type = REFF;
+ goto do_nref;
+
+ case NREF:
+ type = REF;
+ do_nref:
+
+ /* For the named back references, find the corresponding buffer
+ * number */
n = reg_check_named_buff_matched(rex,scan);
- if ( n ) {
- type = REF + ( type - NREF );
- goto do_ref;
- } else {
+ if ( ! n ) {
sayNO;
- }
- /* unreached */
+ }
+ goto do_nref_ref_common;
+
case REFFL:
PL_reg_flags |= RF_tainted;
+ folder = foldEQ_locale;
+ fold_array = PL_fold_locale;
+ goto do_ref;
+
+ case REFFU:
+ folder = foldEQ_latin1;
+ fold_array = PL_fold_latin1;
+ goto do_ref;
+
+ case REFF:
+ folder = foldEQ;
+ fold_array = PL_fold;
/* FALL THROUGH */
+
case REF:
- case REFF:
- n = ARG(scan); /* which paren pair */
+ do_ref:
type = OP(scan);
- do_ref:
+ n = ARG(scan); /* which paren pair */
+
+ do_nref_ref_common:
ln = PL_regoffs[n].start;
PL_reg_leftiter = PL_reg_maxiter; /* Void cache */
if (*PL_reglastparen < n || ln == -1)
break;
s = PL_bostr + ln;
- if (utf8_target && type != REF) { /* REF can do byte comparison */
- char *l = locinput;
- const char *e = PL_bostr + PL_regoffs[n].end;
- /*
- * 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 (type == REFF) {
- 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);
- toLOWER_utf8((U8*)l, tmpbuf2, &ulen2);
- if (ulen1 != ulen2 || memNE((char *)tmpbuf1, (char *)tmpbuf2, ulen1))
- sayNO;
- s += ulen1;
- l += ulen2;
- }
+ if (type != REF /* REF can do byte comparison */
+ && (utf8_target
+ || (type == REFFU
+ && (*s == (char) LATIN_SMALL_LETTER_SHARP_S
+ || *locinput == (char) LATIN_SMALL_LETTER_SHARP_S))))
+ { /* XXX handle REFFL better */
+ char * limit = PL_regeol;
+
+ /* This call case insensitively compares the entire buffer
+ * at s, with the current input starting at locinput, but
+ * not going off the end given by PL_regeol, and returns in
+ * limit upon success, how much of the current input was
+ * matched */
+ if (! foldEQ_utf8(s, NULL, PL_regoffs[n].end - ln, utf8_target,
+ locinput, &limit, 0, utf8_target))
+ {
+ sayNO;
}
- locinput = l;
+ locinput = limit;
nextchr = UCHARAT(locinput);
break;
}
- /* Inline the first character, for speed. */
+ /* Not utf8: Inline the first character, for speed. */
if (UCHARAT(s) != nextchr &&
(type == REF ||
- (UCHARAT(s) != (type == REFF
- ? PL_fold : PL_fold_locale)[nextchr])))
+ UCHARAT(s) != fold_array[nextchr]))
sayNO;
ln = PL_regoffs[n].end - ln;
if (locinput + ln > PL_regeol)
sayNO;
if (ln > 1 && (type == REF
? memNE(s, locinput, ln)
- : (type == REFF
- ? ! foldEQ(s, locinput, ln)
- : ! foldEQ_locale(s, locinput, ln))))
+ : ! folder(s, locinput, ln)))
sayNO;
locinput += ln;
nextchr = UCHARAT(locinput);
{
ST.c1 = (U8)*STRING(text_node);
- ST.c2 =
- (IS_TEXTF(text_node))
- ? PL_fold[ST.c1]
- : (IS_TEXTFL(text_node))
- ? PL_fold_locale[ST.c1]
- : ST.c1;
+ switch (OP(text_node)) {
+ case EXACTF: ST.c2 = PL_fold[ST.c1]; break;
+ case EXACTFU: ST.c2 = PL_fold_latin1[ST.c1]; break;
+ case EXACTFL: ST.c2 = PL_fold_locale[ST.c1]; break;
+ default: ST.c2 = ST.c1;
+ }
}
}
}
if this changes back then the macro for IS_TEXT and
friends need to change. */
if (!UTF_PATTERN) {
- ST.c2 = ST.c1 = *s;
- if (IS_TEXTF(text_node))
- ST.c2 = PL_fold[ST.c1];
- else if (IS_TEXTFL(text_node))
- ST.c2 = PL_fold_locale[ST.c1];
+ ST.c1 = *s;
+ switch (OP(text_node)) {
+ case EXACTF: ST.c2 = PL_fold[ST.c1]; break;
+ case EXACTFU: ST.c2 = PL_fold_latin1[ST.c1]; break;
+ case EXACTFL: ST.c2 = PL_fold_locale[ST.c1]; break;
+ default: ST.c2 = ST.c1; break;
+ }
}
else { /* UTF_PATTERN */
- if (IS_TEXTF(text_node)) {
+ if (IS_TEXTFU(text_node) || IS_TEXTF(text_node)) {
STRLEN ulen1, ulen2;
U8 tmpbuf1[UTF8_MAXBYTES_CASE+1];
U8 tmpbuf2[UTF8_MAXBYTES_CASE+1];
n = ARG(scan);
if ( n == (U32)what_len_TRICKYFOLD(locinput,utf8_target,ln) ) {
locinput += ln;
- } else if ( 0xDF == n && !utf8_target && !UTF_PATTERN ) {
+ } else if ( LATIN_SMALL_LETTER_SHARP_S == n && !utf8_target && !UTF_PATTERN ) {
sayNO;
} else {
U8 folded[UTF8_MAXBYTES_CASE+1];
PL_reg_flags |= RF_tainted;
/* FALL THROUGH */
case EXACTF:
+ case EXACTFU:
/* The comments for the EXACT case above apply as well to these fold
* ones */
* fold matching. */
switch (OP(p)) {
case EXACTF: folded = PL_fold[c]; break;
+ case EXACTFU: folded = PL_fold_latin1[c]; break;
case EXACTFL: folded = PL_fold_locale[c]; break;
default: Perl_croak(aTHX_ "panic: Unexpected op %u", OP(p));
}
https://perl5.git.perl.org / perl5.git / blobdiff