* of t/op/regmesg.t, the tests in t/op/re_tests, and those in
* op/pragma/warn/regcomp.
*/
-#define MARKER1 "HERE" /* marker as it appears in the description */
-#define MARKER2 " << HERE " /* marker as it appears within the regex */
+#define MARKER1 "<-- HERE" /* marker as it appears in the description */
+#define MARKER2 " <-- HERE " /* marker as it appears within the regex */
-#define REPORT_LOCATION " before " MARKER1 " mark in regex m/%.*s" MARKER2 "%s/"
+#define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
/*
* Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
r->offsets[0] = RExC_size;
}
DEBUG_r(PerlIO_printf(Perl_debug_log,
- "%s %u bytes for offset annotations.\n",
+ "%s %"UVuf" bytes for offset annotations.\n",
r->offsets ? "Got" : "Couldn't get",
- (2*RExC_size+1) * sizeof(U32)));
+ (UV)((2*RExC_size+1) * sizeof(U32))));
RExC_rx = r;
/* No compiled RE interpolated, has runtime
components ===> unsafe. */
FAIL("Eval-group not allowed at runtime, use re 'eval'");
- if (PL_tainted)
+ if (PL_tainting && PL_tainted)
FAIL("Eval-group in insecure regular expression");
}
break;
case 'p':
case 'P':
- { /* a lovely hack--pretend we saw [\pX] instead */
+ {
char* oldregxend = RExC_end;
char* parse_start = RExC_parse;
if (RExC_parse[1] == '{') {
+ /* a lovely hack--pretend we saw [\pX] instead */
RExC_end = strchr(RExC_parse, '}');
if (!RExC_end) {
RExC_parse += 2;
S_regclass(pTHX_ RExC_state_t *pRExC_state)
{
register UV value;
- register IV lastvalue = OOB_UNICODE;
+ register IV prevvalue = OOB_UNICODE;
register IV range = 0;
register regnode *ret;
STRLEN numlen;
register char *e;
char *parse_start = RExC_parse; /* MJD */
UV n;
- bool dont_optimize_invert = FALSE;
+ bool optimize_invert = TRUE;
ret = reganode(pRExC_state, ANYOF, 0);
rangebegin = RExC_parse;
if (UTF) {
value = utf8n_to_uvchr((U8*)RExC_parse,
- RExC_end - RExC_parse,
- &numlen, 0);
+ RExC_end - RExC_parse,
+ &numlen, 0);
RExC_parse += numlen;
}
else
RExC_parse - rangebegin,
RExC_parse - rangebegin,
rangebegin);
- if (lastvalue < 256) {
- ANYOF_BITMAP_SET(ret, lastvalue);
+ if (prevvalue < 256) {
+ ANYOF_BITMAP_SET(ret, prevvalue);
ANYOF_BITMAP_SET(ret, '-');
}
else {
ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
Perl_sv_catpvf(aTHX_ listsv,
- "%04"UVxf"\n%04"UVxf"\n", (UV)lastvalue, (UV) '-');
+ "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
}
}
}
if (!SIZE_ONLY) {
+ if (namedclass > OOB_NAMEDCLASS)
+ optimize_invert = FALSE;
/* Possible truncation here but in some 64-bit environments
* the compiler gets heartburn about switch on 64-bit values.
* A similar issue a little earlier when switching on value.
if (isALNUM(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsWord\n");
break;
case ANYOF_NALNUM:
if (!isALNUM(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsWord\n");
break;
case ANYOF_ALNUMC:
if (isALNUMC(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsAlnum\n");
break;
case ANYOF_NALNUMC:
if (!isALNUMC(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsAlnum\n");
break;
case ANYOF_ALPHA:
if (isALPHA(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsAlpha\n");
break;
case ANYOF_NALPHA:
if (!isALPHA(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsAlpha\n");
break;
case ANYOF_ASCII:
ANYOF_BITMAP_SET(ret, value);
#else /* EBCDIC */
for (value = 0; value < 256; value++) {
- if (PL_hints & HINT_RE_ASCIIR) {
- if (NATIVE_TO_ASCII(value) < 128)
- ANYOF_BITMAP_SET(ret, value);
- }
- else {
- if (isASCII(value))
- ANYOF_BITMAP_SET(ret, value);
- }
+ if (isASCII(value))
+ ANYOF_BITMAP_SET(ret, value);
}
#endif /* EBCDIC */
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsASCII\n");
break;
case ANYOF_NASCII:
ANYOF_BITMAP_SET(ret, value);
#else /* EBCDIC */
for (value = 0; value < 256; value++) {
- if (PL_hints & HINT_RE_ASCIIR) {
- if (NATIVE_TO_ASCII(value) >= 128)
- ANYOF_BITMAP_SET(ret, value);
- }
- else {
- if (!isASCII(value))
- ANYOF_BITMAP_SET(ret, value);
- }
+ if (!isASCII(value))
+ ANYOF_BITMAP_SET(ret, value);
}
#endif /* EBCDIC */
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsASCII\n");
break;
case ANYOF_BLANK:
if (isBLANK(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsBlank\n");
break;
case ANYOF_NBLANK:
if (!isBLANK(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsBlank\n");
break;
case ANYOF_CNTRL:
if (isCNTRL(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsCntrl\n");
break;
case ANYOF_NCNTRL:
if (!isCNTRL(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsCntrl\n");
break;
case ANYOF_DIGIT:
for (value = '0'; value <= '9'; value++)
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsDigit\n");
break;
case ANYOF_NDIGIT:
for (value = '9' + 1; value < 256; value++)
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsDigit\n");
break;
case ANYOF_GRAPH:
if (isGRAPH(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsGraph\n");
break;
case ANYOF_NGRAPH:
if (!isGRAPH(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsGraph\n");
break;
case ANYOF_LOWER:
if (isLOWER(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsLower\n");
break;
case ANYOF_NLOWER:
if (!isLOWER(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsLower\n");
break;
case ANYOF_PRINT:
if (isPRINT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsPrint\n");
break;
case ANYOF_NPRINT:
if (!isPRINT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsPrint\n");
break;
case ANYOF_PSXSPC:
if (isPSXSPC(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsSpace\n");
break;
case ANYOF_NPSXSPC:
if (!isPSXSPC(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsSpace\n");
break;
case ANYOF_PUNCT:
if (isPUNCT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsPunct\n");
break;
case ANYOF_NPUNCT:
if (!isPUNCT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsPunct\n");
break;
case ANYOF_SPACE:
if (isSPACE(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsSpacePerl\n");
break;
case ANYOF_NSPACE:
if (!isSPACE(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsSpacePerl\n");
break;
case ANYOF_UPPER:
if (isUPPER(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsUpper\n");
break;
case ANYOF_NUPPER:
if (!isUPPER(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsUpper\n");
break;
case ANYOF_XDIGIT:
if (isXDIGIT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "+utf8::IsXDigit\n");
break;
case ANYOF_NXDIGIT:
if (!isXDIGIT(value))
ANYOF_BITMAP_SET(ret, value);
}
- dont_optimize_invert = TRUE;
Perl_sv_catpvf(aTHX_ listsv, "!utf8::IsXDigit\n");
break;
default:
} /* end of namedclass \blah */
if (range) {
- if (((lastvalue > value) && !(PL_hints & HINT_RE_ASCIIR)) ||
- ((NATIVE_TO_UNI(lastvalue) > NATIVE_TO_UNI(value)) && (PL_hints & HINT_RE_ASCIIR))) /* b-a */ {
+ if (prevvalue > value) /* b-a */ {
Simple_vFAIL4("Invalid [] range \"%*.*s\"",
RExC_parse - rangebegin,
RExC_parse - rangebegin,
rangebegin);
+ range = 0; /* not a valid range */
}
- range = 0; /* not a true range */
}
else {
- lastvalue = value; /* save the beginning of the range */
+ prevvalue = value; /* save the beginning of the range */
if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
RExC_parse[1] != ']') {
RExC_parse++;
/* now is the next time */
if (!SIZE_ONLY) {
- if (lastvalue < 256 && value < 256) {
-#ifdef EBCDIC /* EBCDIC, for example. */
- if (PL_hints & HINT_RE_ASCIIR) {
- IV i;
- /* New style scheme for ranges:
- * after :
- * use re 'asciir';
- * do ranges in ASCII/Unicode space
- */
- for (i = NATIVE_TO_ASCII(lastvalue) ; i <= NATIVE_TO_ASCII(value); i++)
- ANYOF_BITMAP_SET(ret, ASCII_TO_NATIVE(i));
- }
- else if ((isLOWER(lastvalue) && isLOWER(value)) ||
- (isUPPER(lastvalue) && isUPPER(value)))
+ IV i;
+
+ if (prevvalue < 256) {
+ IV ceilvalue = value < 256 ? value : 255;
+
+#ifdef EBCDIC
+ if ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
+ (isUPPER(prevvalue) && isUPPER(ceilvalue)))
{
- IV i;
- if (isLOWER(lastvalue)) {
- for (i = lastvalue; i <= value; i++)
+ if (isLOWER(prevvalue)) {
+ for (i = prevvalue; i <= ceilvalue; i++)
if (isLOWER(i))
ANYOF_BITMAP_SET(ret, i);
} else {
- for (i = lastvalue; i <= value; i++)
+ for (i = prevvalue; i <= ceilvalue; i++)
if (isUPPER(i))
ANYOF_BITMAP_SET(ret, i);
}
}
else
#endif
- for ( ; lastvalue <= value; lastvalue++)
- ANYOF_BITMAP_SET(ret, lastvalue);
- } else {
+ for (i = prevvalue; i <= ceilvalue; i++)
+ ANYOF_BITMAP_SET(ret, i);
+ }
+ if (value > 255) {
ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
- if (lastvalue < value)
+ if (prevvalue < value)
Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
- (UV)lastvalue, (UV)value);
- else
+ (UV)prevvalue, (UV)value);
+ else if (prevvalue == value)
Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
(UV)value);
}
}
if (need_class) {
+ ANYOF_FLAGS(ret) |= ANYOF_LARGE;
if (SIZE_ONLY)
RExC_size += ANYOF_CLASS_ADD_SKIP;
else
}
/* optimize inverted simple patterns (e.g. [^a-z]) */
- if (!SIZE_ONLY && !dont_optimize_invert &&
+ if (!SIZE_ONLY && optimize_invert &&
/* If the only flag is inversion. */
(ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
n = add_data(pRExC_state, 1, "s");
RExC_rx->data->data[n] = (void*)rv;
ARG_SET(ret, n);
- NEXT_OFF(ret) = RExC_emit - ret;
}
return ret;
node = dumpuntil(start, NEXTOPER(node), NEXTOPER(node) + 1, sv, l + 1);
}
else if (op == ANYOF) {
- node = next;
+ /* arglen 1 + class block */
+ node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
+ ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
+ node = NEXTOPER(node);
}
else if (PL_regkind[(U8)op] == EXACT) {
/* Literal string, where present. */
if (r->offsets) {
U32 i;
U32 len = r->offsets[0];
- PerlIO_printf(Perl_debug_log, "Offsets: [%u]\n\t", r->offsets[0]);
+ PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)r->offsets[0]);
for (i = 1; i <= len; i++)
- PerlIO_printf(Perl_debug_log, "%u[%u] ",
- r->offsets[i*2-1],
- r->offsets[i*2]);
+ PerlIO_printf(Perl_debug_log, "%"UVuf"[%"UVuf"] ",
+ (UV)r->offsets[i*2-1],
+ (UV)r->offsets[i*2]);
PerlIO_printf(Perl_debug_log, "\n");
}
#endif /* DEBUGGING */
for (i = 0; i <= 256; i++) { /* just the first 256 */
U8 *e = uvchr_to_utf8(s, i);
- if (i < 256 && swash_fetch(sw, s)) {
+ if (i < 256 && swash_fetch(sw, s, TRUE)) {
if (rangestart == -1)
rangestart = i;
} else if (rangestart != -1) {
SAVEVPTR(PL_regendp); /* Ditto for endp. */
SAVEVPTR(PL_reglastparen); /* Similarly for lastparen. */
SAVEPPTR(PL_regtill); /* How far we are required to go. */
- SAVEI8(PL_regprev); /* char before regbol, \n if none */
SAVEGENERICPV(PL_reg_start_tmp); /* from regexec.c */
PL_reg_start_tmp = 0;
SAVEI32(PL_reg_start_tmpl); /* from regexec.c */
https://perl5.git.perl.org / perl5.git / blobdiff