#ifdef DEBUGGING
/* At least one required character in the target string is expressible only in
* UTF-8. */
-static const char* const non_utf8_target_but_utf8_required
+static const char non_utf8_target_but_utf8_required[]
= "Can't match, because target string needs to be in UTF-8\n";
#endif
const UV total_elems = paren_elems_to_push + REGCP_OTHER_ELEMS;
const UV elems_shifted = total_elems << SAVE_TIGHT_SHIFT;
I32 p;
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
PERL_ARGS_ASSERT_REGCPPUSH;
{
UV i;
U32 paren;
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
PERL_ARGS_ASSERT_REGCPPOP;
* Ideally this could be replaced by a just an array of function pointers
* to the C library functions that implement the macros this calls.
* However, to compile, the precise function signatures are required, and
- * these may vary from platform to to platform. To avoid having to figure
+ * these may vary from platform to platform. To avoid having to figure
* out what those all are on each platform, I (khw) am using this method,
* which adds an extra layer of function call overhead (unless the C
* optimizer strips it away). But we don't particularly care about
* rules, ignoring any locale. So use the Unicode function if this class
* requires an inversion list, and use the Unicode macro otherwise. */
- dVAR;
PERL_ARGS_ASSERT_ISFOO_UTF8_LC;
span_word |= span_word << 4;
/* That reduces the problem to what this function solves */
- return s + _variant_byte_number(span_word);
+ return s + variant_byte_number(span_word);
#endif
masked &= PERL_VARIANTS_WORD_MASK;
/* This reduces the problem to that solved by this function */
- s += _variant_byte_number(masked);
+ s += variant_byte_number(masked);
return s;
} while (s + PERL_WORDSIZE <= send);
masked |= masked << 1;
masked |= masked << 2;
masked |= masked << 4;
- return s + _variant_byte_number(masked);
+ return s + variant_byte_number(masked);
#endif
RXi_GET_DECL(prog,progi);
regmatch_info reginfo_buf; /* create some info to pass to find_byclass */
regmatch_info *const reginfo = ®info_buf;
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
PERL_ARGS_ASSERT_RE_INTUIT_START;
PERL_UNUSED_ARG(flags);
/* now look for the 'other' substring if defined */
- if (utf8_target ? prog->substrs->data[other_ix].utf8_substr
- : prog->substrs->data[other_ix].substr)
+ if (prog->substrs->data[other_ix].utf8_substr
+ || prog->substrs->data[other_ix].substr)
{
/* Take into account the "other" substring. */
char *last, *last1;
do_other_substr:
other = &prog->substrs->data[other_ix];
+ if (!utf8_target && !other->substr) {
+ if (!to_byte_substr(prog)) {
+ NON_UTF8_TARGET_BUT_UTF8_REQUIRED(fail);
+ }
+ }
/* if "other" is anchored:
* we've previously found a floating substr starting at check_at.
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
+ const SSize_t cl_l = (PL_regkind[OP(progi->regstclass)] == EXACT
? (reginfo->is_utf8_pat
- ? utf8_distance(str + STR_LEN(progi->regstclass), str)
- : STR_LEN(progi->regstclass))
+ ? (SSize_t)utf8_distance(str + STR_LEN(progi->regstclass), str)
+ : (SSize_t)STR_LEN(progi->regstclass))
: 1);
char * endpos;
char *s;
} else { \
uvc = _toFOLD_utf8_flags( (const U8*) uc, uc_end, foldbuf, &foldlen, \
flags); \
- len = UTF8SKIP(uc); \
+ len = UTF8_SAFE_SKIP(uc, uc_end); \
skiplen = UVCHR_SKIP( uvc ); \
foldlen -= skiplen; \
uscan = foldbuf + skiplen; \
STMT_START { \
while (s < strend) { \
CODE \
- s += ((UTF8) ? UTF8SKIP(s) : 1); \
+ s += ((UTF8) \
+ ? UTF8_SAFE_SKIP(s, reginfo->strend) \
+ : 1); \
} \
} STMT_END
#define REXEC_FBC_CLASS_SCAN_GUTS(UTF8, COND) \
if (COND) { \
FBC_CHECK_AND_TRY \
- s += ((UTF8) ? UTF8SKIP(s) : 1); \
+ s += ((UTF8) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1);\
previous_occurrence_end = s; \
} \
else { \
* of the one we're looking for. Knowing that, we can see right away if the
* next occurrence is adjacent to the previous. When 'doevery' is FALSE, we
* don't accept the 2nd and succeeding adjacent occurrences */
-#define FBC_CHECK_AND_TRY \
- if ( ( doevery \
- || s != previous_occurrence_end) \
- && (reginfo->intuit || regtry(reginfo, &s))) \
- { \
- goto got_it; \
+#define FBC_CHECK_AND_TRY \
+ if ( ( doevery \
+ || s != previous_occurrence_end) \
+ && ( reginfo->intuit \
+ || (s <= reginfo->strend && regtry(reginfo, &s)))) \
+ { \
+ goto got_it; \
}
previous_occurrence_end = s; \
}
+/* This differs from the above macros in that it is passed a single byte that
+ * is known to begin the next occurrence of the thing being looked for in 's'.
+ * It does a memchr to find the next occurrence of 'byte', before trying 'COND'
+ * at that position. */
+#define REXEC_FBC_FIND_NEXT_UTF8_BYTE_SCAN(byte, COND) \
+ while (s < strend) { \
+ s = (char *) memchr(s, byte, strend -s); \
+ if (s == NULL) { \
+ s = (char *) strend; \
+ break; \
+ } \
+ \
+ if (COND) { \
+ FBC_CHECK_AND_TRY \
+ s += UTF8_SAFE_SKIP(s, reginfo->strend); \
+ previous_occurrence_end = s; \
+ } \
+ else { \
+ s += UTF8SKIP(s); \
+ } \
+ }
+
/* The three macros below are slightly different versions of the same logic.
*
* The first is for /a and /aa when the target string is UTF-8. This can only
/* Like FBC_UTF8_A, but TEST_UV is a macro which takes a UV as its input, and
* TEST_UTF8 is a macro that for the same input code points returns identically
- * to TEST_UV, but takes a pointer to a UTF-8 encoded string instead */
+ * to TEST_UV, but takes a pointer to a UTF-8 encoded string instead (and an
+ * end pointer as well) */
#define FBC_UTF8(TEST_UV, TEST_UTF8, IF_SUCCESS, IF_FAIL) \
if (s == reginfo->strbeg) { \
tmp = '\n'; \
}
/* This is the macro to use when we want to see if something that looks like it
- * could match, actually does, and if so exits the loop */
-#define REXEC_FBC_TRYIT \
- if ((reginfo->intuit || regtry(reginfo, &s))) \
+ * could match, actually does, and if so exits the loop. It needs to be used
+ * only for bounds checking macros, as it allows for matching beyond the end of
+ * string (which should be zero length without having to look at the string
+ * contents) */
+#define REXEC_FBC_TRYIT \
+ if (reginfo->intuit || (s <= reginfo->strend && regtry(reginfo, &s))) \
goto got_it
/* The only difference between the BOUND and NBOUND cases is that
S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s,
const char *strend, regmatch_info *reginfo)
{
- dVAR;
/* TRUE if x+ need not match at just the 1st pos of run of x's */
const I32 doevery = (prog->intflags & PREGf_SKIP) == 0;
break;
case ANYOFM: /* ARG() is the base byte; FLAGS() the mask byte */
- /* UTF-8ness doesn't matter, so use 0 */
+ /* UTF-8ness doesn't matter because only matches UTF-8 invariants, so
+ * use 0 */
REXEC_FBC_FIND_NEXT_SCAN(0,
(char *) find_next_masked((U8 *) s, (U8 *) strend,
(U8) ARG(c), FLAGS(c)));
break;
- case NANYOFM:
- REXEC_FBC_FIND_NEXT_SCAN(0,
+ case NANYOFM: /* UTF-8ness does matter because can match UTF-8 variants.
+ */
+ REXEC_FBC_FIND_NEXT_SCAN(utf8_target,
(char *) find_span_end_mask((U8 *) s, (U8 *) strend,
(U8) ARG(c), FLAGS(c)));
break;
case ANYOFH:
- if (utf8_target) REXEC_FBC_CLASS_SCAN(TRUE,
+ if (utf8_target) { /* Can't possibly match a non-UTF-8 target */
+ REXEC_FBC_CLASS_SCAN(TRUE,
+ ( (U8) NATIVE_UTF8_TO_I8(*s) >= ANYOF_FLAGS(c)
+ && reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target)));
+ }
+ break;
+
+ case ANYOFHb:
+ if (utf8_target) { /* Can't possibly match a non-UTF-8 target */
+
+ /* We know what the first byte of any matched string should be */
+ U8 first_byte = FLAGS(c);
+
+ REXEC_FBC_FIND_NEXT_UTF8_BYTE_SCAN(first_byte,
reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target));
+ }
+ break;
+
+ case ANYOFHr:
+ if (utf8_target) { /* Can't possibly match a non-UTF-8 target */
+ REXEC_FBC_CLASS_SCAN(TRUE,
+ ( inRANGE(NATIVE_UTF8_TO_I8(*s),
+ LOWEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(c)),
+ HIGHEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(c)))
+ && reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target)));
+ }
+ break;
+
+ case ANYOFHs:
+ if (utf8_target) { /* Can't possibly match a non-UTF-8 target */
+ REXEC_FBC_CLASS_SCAN(TRUE,
+ ( strend -s >= FLAGS(c)
+ && memEQ(s, ((struct regnode_anyofhs *) c)->string, FLAGS(c))
+ && reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target)));
+ }
+ break;
+
+ case ANYOFR:
+ if (utf8_target) {
+ REXEC_FBC_CLASS_SCAN(TRUE,
+ ( NATIVE_UTF8_TO_I8(*s) >= ANYOF_FLAGS(c)
+ && withinCOUNT(utf8_to_uvchr_buf((U8 *) s,
+ (U8 *) strend,
+ NULL),
+ ANYOFRbase(c), ANYOFRdelta(c))));
+ }
+ else {
+ REXEC_FBC_CLASS_SCAN(0, withinCOUNT((U8) *s,
+ ANYOFRbase(c), ANYOFRdelta(c)));
+ }
+ break;
+
+ case ANYOFRb:
+ if (utf8_target) {
+
+ /* We know what the first byte of any matched string should be */
+ U8 first_byte = FLAGS(c);
+
+ REXEC_FBC_FIND_NEXT_UTF8_BYTE_SCAN(first_byte,
+ withinCOUNT(utf8_to_uvchr_buf((U8 *) s,
+ (U8 *) strend,
+ NULL),
+ ANYOFRbase(c), ANYOFRdelta(c)));
+ }
+ else {
+ REXEC_FBC_CLASS_SCAN(0, withinCOUNT((U8) *s,
+ ANYOFRbase(c), ANYOFRdelta(c)));
+ }
break;
case EXACTFAA_NO_TRIE: /* This node only generated for non-utf8 patterns */
| FOLDEQ_S2_FOLDS_SANE;
goto do_exactf_utf8;
- case EXACTFU_ONLY8:
+ case EXACTFU_REQ8:
if (! utf8_target) {
break;
}
* first character. c2 is its fold. This logic will not work for
* Unicode semantics and the german sharp ss, which hence should
* not be compiled into a node that gets here. */
- pat_string = STRING(c);
- ln = STR_LEN(c); /* length to match in octets/bytes */
+ pat_string = STRINGs(c);
+ ln = STR_LENs(c); /* length to match in octets/bytes */
/* We know that we have to match at least 'ln' bytes (which is the
* same as characters, since not utf8). If we have to match 3
/* If one of the operands is in utf8, we can't use the simpler folding
* above, due to the fact that many different characters can have the
* same fold, or portion of a fold, or different- length fold */
- pat_string = STRING(c);
- ln = STR_LEN(c); /* length to match in octets/bytes */
+ pat_string = STRINGs(c);
+ ln = STR_LENs(c); /* length to match in octets/bytes */
pat_end = pat_string + ln;
lnc = is_utf8_pat /* length to match in characters */
? utf8_length((U8 *) pat_string, (U8 *) pat_end)
{
goto got_it;
}
- s += (utf8_target) ? UTF8SKIP(s) : 1;
+ s += (utf8_target) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1;
}
break;
}
}
/* Didn't match. Try at the next position (if there is one) */
- s += (utf8_target) ? UTF8SKIP(s) : 1;
+ s += (utf8_target) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1;
if (UNLIKELY(s >= reginfo->strend)) {
break;
}
goto got_it;
}
before = after;
- s += UTF8SKIP(s);
+ s += UTF8_SAFE_SKIP(s, reginfo->strend);
}
}
else { /* Not utf8. Everything is a GCB except between CR and
/* And, since this is a bound, it can match after the final
* character in the string */
- if ((reginfo->intuit || regtry(reginfo, &s))) {
+ if ( reginfo->intuit
+ || (s <= reginfo->strend && regtry(reginfo, &s)))
+ {
goto got_it;
}
break;
if (reginfo->intuit || regtry(reginfo, &s)) {
goto got_it;
}
- s += (utf8_target) ? UTF8SKIP(s) : 1;
+ s += (utf8_target) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1;
if (UNLIKELY(s >= reginfo->strend)) {
break;
}
goto got_it;
}
before = after;
- s += UTF8SKIP(s);
+ s += UTF8_SAFE_SKIP(s, reginfo->strend);
}
}
else { /* Not utf8. */
}
}
- if (reginfo->intuit || regtry(reginfo, &s)) {
+ if ( reginfo->intuit
+ || (s <= reginfo->strend && regtry(reginfo, &s)))
+ {
goto got_it;
}
if (reginfo->intuit || regtry(reginfo, &s)) {
goto got_it;
}
- s += (utf8_target) ? UTF8SKIP(s) : 1;
+ s += (utf8_target) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1;
if (UNLIKELY(s >= reginfo->strend)) {
break;
}
goto got_it;
}
before = after;
- s += UTF8SKIP(s);
+ s += UTF8_SAFE_SKIP(s, reginfo->strend);
}
}
else { /* Not utf8. */
/* Here are at the final position in the target string. The SB
* value is always true here, so matches, depending on other
* constraints */
- if (reginfo->intuit || regtry(reginfo, &s)) {
+ if ( reginfo->intuit
+ || (s <= reginfo->strend && regtry(reginfo, &s)))
+ {
goto got_it;
}
if (reginfo->intuit || regtry(reginfo, &s)) {
goto got_it;
}
- s += (utf8_target) ? UTF8SKIP(s) : 1;
+ s += (utf8_target) ? UTF8_SAFE_SKIP(s, reginfo->strend) : 1;
if (UNLIKELY(s >= reginfo->strend)) {
break;
}
}
previous = before;
before = after;
- s += UTF8SKIP(s);
+ s += UTF8_SAFE_SKIP(s, reginfo->strend);
}
}
else { /* Not utf8. */
}
}
- if (reginfo->intuit || regtry(reginfo, &s)) {
+ if ( reginfo->intuit
+ || (s <= reginfo->strend && regtry(reginfo, &s)))
+ {
goto got_it;
}
}
U8 *bitmap=NULL;
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
/* We can't just allocate points here. We need to wrap it in
* an SV so it gets freed properly if there is a croak while
LEAVE;
goto got_it;
}
- s = HOPc(s,1);
+ if (s < reginfo->strend) {
+ s = HOPc(s,1);
+ }
DEBUG_TRIE_EXECUTE_r({
Perl_re_printf( aTHX_ "Pattern failed. Looking for new start point...\n");
});
regmatch_info *const reginfo = ®info_buf;
regexp_paren_pair *swap = NULL;
I32 oldsave;
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
PERL_ARGS_ASSERT_REGEXEC_FLAGS;
PERL_UNUSED_ARG(data);
if (!startpos ||
((flags & REXEC_FAIL_ON_UNDERFLOW) && startpos < stringarg))
{
- DEBUG_r(Perl_re_printf( aTHX_
+ DEBUG_GPOS_r(Perl_re_printf( aTHX_
"fail: ganch-gofs before earliest possible start\n"));
return 0;
}
minlen = prog->minlen;
if ((startpos + minlen) > strend || startpos < strbeg) {
- DEBUG_r(Perl_re_printf( aTHX_
- "Regex match can't succeed, so not even tried\n"));
+ DEBUG_EXECUTE_r(Perl_re_printf( aTHX_
+ "Regex match can't succeed, so not even tried\n"));
return 0;
}
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
+ ? (char*)utf8_hop_forward((U8*)s, prog->minlenret, (U8 *) strend) - strbeg
: s - strbeg + prog->minlenret;
if ( !(flags & REXEC_NOT_FIRST) )
S_reg_set_capture_string(aTHX_ rx,
to_utf8_substr(prog);
}
ch = SvPVX_const(prog->anchored_utf8)[0];
- REXEC_FBC_SCAN(0, /* 0=>not-utf8 */
+ REXEC_FBC_SCAN(1, /* 1=>utf8 */
if (*s == ch) {
DEBUG_EXECUTE_r( did_match = 1 );
if (regtry(reginfo, &s)) goto got_it;
- s += UTF8SKIP(s);
+ s += UTF8_SAFE_SKIP(s, strend);
while (s < strend && *s == ch)
s += UTF8SKIP(s);
}
U32 depth = 0; /* used by REGCP_SET */
#endif
RXi_GET_DECL(prog,progi);
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
PERL_ARGS_ASSERT_REGTRY;
* to/from code points */
bool utf8_has_been_setup = FALSE;
- dVAR;
U8 *pat = (U8*)STRING(text_node);
U8 folded[UTF8_MAX_FOLD_CHAR_EXPAND * UTF8_MAXBYTES_CASE + 1] = { '\0' };
if ( OP(text_node) == EXACT
- || OP(text_node) == EXACT_ONLY8
+ || OP(text_node) == LEXACT
+ || OP(text_node) == EXACT_REQ8
+ || OP(text_node) == LEXACT_REQ8
|| OP(text_node) == EXACTL)
{
* copy the input to the output, avoiding finding the code point of
* that character */
if (!is_utf8_pat) {
- assert(OP(text_node) != EXACT_ONLY8);
+ assert( OP(text_node) != EXACT_REQ8
+ && OP(text_node) != LEXACT_REQ8);
c2 = c1 = *pat;
}
else if (utf8_target) {
Copy(pat, c2_utf8, UTF8SKIP(pat), U8);
utf8_has_been_setup = TRUE;
}
- else if (OP(text_node) == EXACT_ONLY8) {
+ else if ( OP(text_node) == EXACT_REQ8
+ || OP(text_node) == LEXACT_REQ8)
+ {
return FALSE; /* Can only match UTF-8 target */
}
else {
}
}
else { /* an EXACTFish node */
- U8 *pat_end = pat + STR_LEN(text_node);
+ U8 *pat_end = pat + STR_LENs(text_node);
/* An EXACTFL node has at least some characters unfolded, because what
* they match is not known until now. So, now is the time to fold
}
}
else if (c1 > 255) {
- const unsigned int * remaining_folds;
- unsigned int first_fold;
+ const U32 * remaining_folds;
+ U32 first_fold;
/* Look up what code points (besides c1) fold to c1; e.g.,
* [ 'K', KELVIN_SIGN ] both fold to 'k'. */
case EXACTFU:
c2 = PL_fold_latin1[c1];
break;
- case EXACTFU_ONLY8:
+ case EXACTFU_REQ8:
return FALSE;
NOT_REACHED; /* NOTREACHED */
S_isGCB(pTHX_ const GCB_enum before, const GCB_enum after, const U8 * const strbeg, const U8 * const curpos, const bool utf8_target)
{
/* returns a boolean indicating if there is a Grapheme Cluster Boundary
- * between the inputs. See http://www.unicode.org/reports/tr29/. */
+ * between the inputs. See https://www.unicode.org/reports/tr29/. */
PERL_ARGS_ASSERT_ISGCB;
}
while (prev == GCB_Extend);
- return prev != GCB_XPG_XX;
+ return prev != GCB_ExtPict_XX;
}
default:
STATIC GCB_enum
S_backup_one_GCB(pTHX_ const U8 * const strbeg, U8 ** curpos, const bool utf8_target)
{
- dVAR;
GCB_enum gcb;
PERL_ARGS_ASSERT_BACKUP_ONE_GCB;
STATIC LB_enum
S_advance_one_LB(pTHX_ U8 ** curpos, const U8 * const strend, const bool utf8_target)
{
- dVAR;
LB_enum lb;
STATIC LB_enum
S_backup_one_LB(pTHX_ const U8 * const strbeg, U8 ** curpos, const bool utf8_target)
{
- dVAR;
LB_enum lb;
PERL_ARGS_ASSERT_BACKUP_ONE_LB;
const bool utf8_target)
{
/* returns a boolean indicating if there is a Sentence Boundary Break
- * between the inputs. See http://www.unicode.org/reports/tr29/ */
+ * between the inputs. See https://www.unicode.org/reports/tr29/ */
U8 * lpos = (U8 *) curpos;
bool has_para_sep = FALSE;
STATIC SB_enum
S_advance_one_SB(pTHX_ U8 ** curpos, const U8 * const strend, const bool utf8_target)
{
- dVAR;
SB_enum sb;
PERL_ARGS_ASSERT_ADVANCE_ONE_SB;
STATIC SB_enum
S_backup_one_SB(pTHX_ const U8 * const strbeg, U8 ** curpos, const bool utf8_target)
{
- dVAR;
SB_enum sb;
PERL_ARGS_ASSERT_BACKUP_ONE_SB;
const bool utf8_target,
const bool skip_Extend_Format)
{
- dVAR;
WB_enum wb;
PERL_ARGS_ASSERT_ADVANCE_ONE_WB;
STATIC WB_enum
S_backup_one_WB(pTHX_ WB_enum * previous, const U8 * const strbeg, U8 ** curpos, const bool utf8_target)
{
- dVAR;
WB_enum wb;
PERL_ARGS_ASSERT_BACKUP_ONE_WB;
/* push a new state then goto it */
-#define PUSH_STATE_GOTO(state, node, input, eol) \
+#define PUSH_STATE_GOTO(state, node, input, eol, sr0) \
pushinput = input; \
pusheol = eol; \
+ pushsr0 = sr0; \
scan = node; \
st->resume_state = state; \
goto push_state;
/* push a new state with success backtracking, then goto it */
-#define PUSH_YES_STATE_GOTO(state, node, input, eol) \
+#define PUSH_YES_STATE_GOTO(state, node, input, eol, sr0) \
pushinput = input; \
pusheol = eol; \
+ pushsr0 = sr0; \
scan = node; \
st->resume_state = state; \
goto push_yes_state;
rest of the pattern. Variable and state names reflect this convention.
The states in the main switch are the union of ops and failure/success of
-substates associated with with that op. For example, IFMATCH is the op
+substates associated with that op. For example, IFMATCH is the op
that does lookahead assertions /(?=A)B/ and so the IFMATCH state means
'execute IFMATCH'; while IFMATCH_A is a state saying that we have just
successfully matched A and IFMATCH_A_fail is a state saying that we have
STATIC SSize_t
S_regmatch(pTHX_ regmatch_info *reginfo, char *startpos, regnode *prog)
{
- dVAR;
const bool utf8_target = reginfo->is_utf8_target;
const U32 uniflags = UTF8_ALLOW_DEFAULT;
REGEXP *rex_sv = reginfo->prog;
char *loceol = reginfo->strend;
char *pushinput; /* where to continue after a PUSH */
char *pusheol; /* where to stop matching (loceol) after a PUSH */
+ U8 *pushsr0; /* save starting pos of script run */
I32 nextchr; /* is always set to UCHARAT(locinput), or -1 at EOS */
bool result = 0; /* return value of S_regmatch */
#endif
#ifdef DEBUGGING
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
#endif
/* protect against undef(*^R) */
/* update the startpoint */
st->u.keeper.val = rex->offs[0].start;
rex->offs[0].start = locinput - reginfo->strbeg;
- PUSH_STATE_GOTO(KEEPS_next, next, locinput, loceol);
+ PUSH_STATE_GOTO(KEEPS_next, next, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case KEEPS_next_fail:
while (chars) {
if (utf8_target) {
+ /* XXX This assumes the length is well-formed, as
+ * does the UTF8SKIP below */
uvc = utf8n_to_uvchr((U8*)uc, UTF8_MAXLEN, &len,
uniflags);
uc += len;
});
if ( ST.accepted > 1 || has_cutgroup || ST.jump ) {
- PUSH_STATE_GOTO(TRIE_next, scan, (char*)uc, loceol);
+ PUSH_STATE_GOTO(TRIE_next, scan, (char*)uc, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
/* only one choice left - just continue */
}
#undef ST
+ case LEXACT_REQ8:
+ if (! utf8_target) {
+ sayNO;
+ }
+ /* FALLTHROUGH */
+
+ case LEXACT:
+ {
+ char *s;
+
+ s = STRINGl(scan);
+ ln = STR_LENl(scan);
+ goto join_short_long_exact;
+
case EXACTL: /* /abc/l */
_CHECK_AND_WARN_PROBLEMATIC_LOCALE;
_CHECK_AND_OUTPUT_WIDE_LOCALE_UTF8_MSG(locinput, reginfo->strend);
}
goto do_exact;
- case EXACT_ONLY8:
+ case EXACT_REQ8:
if (! utf8_target) {
sayNO;
}
/* FALLTHROUGH */
- case EXACT: { /* /abc/ */
- char *s;
+
+ case EXACT: /* /abc/ */
do_exact:
- s = STRING(scan);
- ln = STR_LEN(scan);
+ s = STRINGs(scan);
+ ln = STR_LENs(scan);
+
+ join_short_long_exact:
if (utf8_target != is_utf8_pat) {
/* The target and the pattern have differing utf8ness. */
char *l = locinput;
fold_array = PL_fold_latin1;
goto do_exactf;
- case EXACTFU_ONLY8: /* /abc/iu with something in /abc/ > 255 */
+ case EXACTFU_REQ8: /* /abc/iu with something in /abc/ > 255 */
if (! utf8_target) {
sayNO;
}
fold_utf8_flags = 0;
do_exactf:
- s = STRING(scan);
- ln = STR_LEN(scan);
+ s = STRINGs(scan);
+ ln = STR_LENs(scan);
if ( utf8_target
|| is_utf8_pat
if (locinput == reginfo->strbeg)
b1 = isWORDCHAR_LC('\n');
else {
- b1 = isWORDCHAR_LC_utf8_safe(reghop3((U8*)locinput, -1,
- (U8*)(reginfo->strbeg)),
- (U8*)(reginfo->strend));
+ U8 *p = reghop3((U8*)locinput, -1,
+ (U8*)(reginfo->strbeg));
+ b1 = isWORDCHAR_LC_utf8_safe(p, (U8*)(reginfo->strend));
}
b2 = (NEXTCHR_IS_EOS)
? isWORDCHAR_LC('\n')
case TRADITIONAL_BOUND:
{
bool b1, b2;
- b1 = (locinput == reginfo->strbeg)
- ? 0 /* isWORDCHAR_L1('\n') */
- : isWORDCHAR_utf8_safe(
- reghop3((U8*)locinput,
- -1,
- (U8*)(reginfo->strbeg)),
- (U8*) reginfo->strend);
+ if (locinput == reginfo->strbeg) {
+ b1 = 0 /* isWORDCHAR_L1('\n') */;
+ }
+ else {
+ U8 *p = reghop3((U8*)locinput, -1,
+ (U8*)(reginfo->strbeg));
+
+ b1 = isWORDCHAR_utf8_safe(p, (U8*) reginfo->strend);
+ }
b2 = (NEXTCHR_IS_EOS)
? 0 /* isWORDCHAR_L1('\n') */
: isWORDCHAR_utf8_safe((U8*)locinput,
case ANYOFH:
if ( ! utf8_target
|| NEXTCHR_IS_EOS
+ || ANYOF_FLAGS(scan) > NATIVE_UTF8_TO_I8(*locinput)
+ || ! reginclass(rex, scan, (U8*)locinput, (U8*) loceol,
+ utf8_target))
+ {
+ sayNO;
+ }
+ goto increment_locinput;
+ break;
+
+ case ANYOFHb:
+ if ( ! utf8_target
+ || NEXTCHR_IS_EOS
+ || ANYOF_FLAGS(scan) != (U8) *locinput
+ || ! reginclass(rex, scan, (U8*)locinput, (U8*) loceol,
+ utf8_target))
+ {
+ sayNO;
+ }
+ goto increment_locinput;
+ break;
+
+ case ANYOFHr:
+ if ( ! utf8_target
+ || NEXTCHR_IS_EOS
+ || ! inRANGE((U8) NATIVE_UTF8_TO_I8(*locinput),
+ LOWEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(scan)),
+ HIGHEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(scan)))
+ || ! reginclass(rex, scan, (U8*)locinput, (U8*) loceol,
+ utf8_target))
+ {
+ sayNO;
+ }
+ goto increment_locinput;
+ break;
+
+ case ANYOFHs:
+ if ( ! utf8_target
+ || NEXTCHR_IS_EOS
+ || loceol - locinput < FLAGS(scan)
+ || memNE(locinput, ((struct regnode_anyofhs *) scan)->string, FLAGS(scan))
|| ! reginclass(rex, scan, (U8*)locinput, (U8*) loceol,
utf8_target))
{
goto increment_locinput;
break;
+ case ANYOFR:
+ if (NEXTCHR_IS_EOS) {
+ sayNO;
+ }
+
+ if (utf8_target) {
+ if ( ANYOF_FLAGS(scan) > NATIVE_UTF8_TO_I8(*locinput)
+ || ! withinCOUNT(utf8_to_uvchr_buf((U8 *) locinput,
+ (U8 *) reginfo->strend,
+ NULL),
+ ANYOFRbase(scan), ANYOFRdelta(scan)))
+ {
+ sayNO;
+ }
+ }
+ else {
+ if (! withinCOUNT((U8) *locinput,
+ ANYOFRbase(scan), ANYOFRdelta(scan)))
+ {
+ sayNO;
+ }
+ }
+ goto increment_locinput;
+ break;
+
+ case ANYOFRb:
+ if (NEXTCHR_IS_EOS) {
+ sayNO;
+ }
+
+ if (utf8_target) {
+ if ( ANYOF_FLAGS(scan) != (U8) *locinput
+ || ! withinCOUNT(utf8_to_uvchr_buf((U8 *) locinput,
+ (U8 *) reginfo->strend,
+ NULL),
+ ANYOFRbase(scan), ANYOFRdelta(scan)))
+ {
+ sayNO;
+ }
+ }
+ else {
+ if (! withinCOUNT((U8) *locinput,
+ ANYOFRbase(scan), ANYOFRdelta(scan)))
+ {
+ sayNO;
+ }
+ }
+ goto increment_locinput;
+ break;
+
/* The argument (FLAGS) to all the POSIX node types is the class number
* */
}
break;
}
- locinput += UTF8SKIP(locinput);
+ locinput += UTF8_SAFE_SKIP(locinput, reginfo->strend);
}
break;
}
break;
- case NREFFL: /* /\g{name}/il */
+ case REFFLN: /* /\g{name}/il */
{ /* 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
utf8_fold_flags = FOLDEQ_LOCALE;
goto do_nref;
- case NREFFA: /* /\g{name}/iaa */
+ case REFFAN: /* /\g{name}/iaa */
folder = foldEQ_latin1;
fold_array = PL_fold_latin1;
type = REFFA;
utf8_fold_flags = FOLDEQ_UTF8_NOMIX_ASCII;
goto do_nref;
- case NREFFU: /* /\g{name}/iu */
+ case REFFUN: /* /\g{name}/iu */
folder = foldEQ_latin1;
fold_array = PL_fold_latin1;
type = REFFU;
utf8_fold_flags = 0;
goto do_nref;
- case NREFF: /* /\g{name}/i */
+ case REFFN: /* /\g{name}/i */
folder = foldEQ;
fold_array = PL_fold;
type = REFF;
utf8_fold_flags = 0;
goto do_nref;
- case NREF: /* /\g{name}/ */
+ case REFN: /* /\g{name}/ */
type = REF;
folder = NULL;
fold_array = NULL;
rex->recurse_locinput[arg]= locinput;
DEBUG_r({
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
DEBUG_STACK_r({
Perl_re_exec_indentf( aTHX_
"entering GOSUB, prev_recurse_locinput=%p recurse_locinput[%d]=%p\n",
/* NOTREACHED */
case EVAL: /* /(?{...})B/ /(??{A})B/ and /(?(?{...})X|Y)B/ */
- if (cur_eval && cur_eval->locinput==locinput) {
+ if (logical == 2 && cur_eval && cur_eval->locinput==locinput) {
if ( ++nochange_depth > max_nochange_depth )
Perl_croak(aTHX_ "EVAL without pos change exceeded limit in regex");
} else {
PL_curpm = PL_reg_curpm;
if (logical != 2) {
- PUSH_STATE_GOTO(EVAL_B, next, locinput, loceol);
+ PUSH_STATE_GOTO(EVAL_B, next, locinput, loceol,
+ script_run_begin);
/* NOTREACHED */
}
}
ST.prev_eval = cur_eval;
cur_eval = st;
/* now continue from first node in postoned RE */
- PUSH_YES_STATE_GOTO(EVAL_postponed_AB, startpoint, locinput, loceol);
+ PUSH_YES_STATE_GOTO(EVAL_postponed_AB, startpoint, locinput,
+ loceol, script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
sw = cBOOL(rex->lastparen >= n && rex->offs[n].end != -1);
break;
- case NGROUPP: /* (?(<name>)) */
+ case GROUPPN: /* (?(<name>)) */
/* reg_check_named_buff_matched returns 0 for no match */
sw = cBOOL(0 < reg_check_named_buff_matched(rex,scan));
break;
ST.count = -1; /* this will be updated by WHILEM */
ST.lastloc = NULL; /* this will be updated by WHILEM */
- PUSH_YES_STATE_GOTO(CURLYX_end, PREVOPER(next), locinput, loceol);
+ PUSH_YES_STATE_GOTO(CURLYX_end, PREVOPER(next), locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
cur_curlyx->u.curlyx.lastloc = locinput;
REGCP_SET(ST.lastcp);
- PUSH_STATE_GOTO(WHILEM_A_pre, A, locinput, loceol);
+ PUSH_STATE_GOTO(WHILEM_A_pre, A, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
ST.save_curlyx = cur_curlyx;
cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
PUSH_YES_STATE_GOTO(WHILEM_B_min, ST.save_curlyx->u.curlyx.B,
- locinput, loceol);
+ locinput, loceol, script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
maxopenparen);
cur_curlyx->u.curlyx.lastloc = locinput;
REGCP_SET(ST.lastcp);
- PUSH_STATE_GOTO(WHILEM_A_max, A, locinput, loceol);
+ PUSH_STATE_GOTO(WHILEM_A_max, A, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
goto do_whilem_B_max;
ST.save_curlyx = cur_curlyx;
cur_curlyx = cur_curlyx->u.curlyx.prev_curlyx;
PUSH_YES_STATE_GOTO(WHILEM_B_max, ST.save_curlyx->u.curlyx.B,
- locinput, loceol);
+ locinput, loceol, script_run_begin);
NOT_REACHED; /* NOTREACHED */
case WHILEM_B_min_fail: /* just failed to match B in a minimal match */
REGCP_SET(ST.lastcp);
PUSH_STATE_GOTO(WHILEM_A_min,
/*A*/ NEXTOPER(ST.save_curlyx->u.curlyx.me) + EXTRA_STEP_2ARGS,
- locinput, loceol);
+ locinput, loceol, script_run_begin);
NOT_REACHED; /* NOTREACHED */
#undef ST
/* Now go into the branch */
if (has_cutgroup) {
- PUSH_YES_STATE_GOTO(BRANCH_next, scan, locinput, loceol);
+ PUSH_YES_STATE_GOTO(BRANCH_next, scan, locinput, loceol,
+ script_run_begin);
} else {
- PUSH_STATE_GOTO(BRANCH_next, scan, locinput, loceol);
+ PUSH_STATE_GOTO(BRANCH_next, scan, locinput, loceol,
+ script_run_begin);
}
NOT_REACHED; /* NOTREACHED */
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, loceol);
+ PUSH_STATE_GOTO(CUTGROUP_next, next, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case CUTGROUP_next_fail:
goto curlym_do_B;
curlym_do_A: /* execute the A in /A{m,n}B/ */
- PUSH_YES_STATE_GOTO(CURLYM_A, ST.A, locinput, loceol); /* match A */
+ PUSH_YES_STATE_GOTO(CURLYM_A, ST.A, locinput, loceol, /* match A */
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case CURLYM_A: /* we've just matched an A */
);
if (! NEXTCHR_IS_EOS && ST.c1 != CHRTEST_VOID) {
if (! UTF8_IS_INVARIANT(nextchr) && utf8_target) {
- if (memNE(locinput, ST.c1_utf8, UTF8SKIP(locinput))
- && memNE(locinput, ST.c2_utf8, UTF8SKIP(locinput)))
+
+ /* (We can use memEQ and memNE in this file without
+ * having to worry about one being shorter than the
+ * other, since the first byte of each gives the
+ * length of the character) */
+ if ( memNE(locinput, ST.c1_utf8, UTF8_SAFE_SKIP(locinput,
+ reginfo->strend))
+ && memNE(locinput, ST.c2_utf8, UTF8_SAFE_SKIP(locinput,
+ reginfo->strend)))
{
/* simulate B failing */
DEBUG_OPTIMISE_r(
}
}
- PUSH_STATE_GOTO(CURLYM_B, ST.B, locinput, loceol); /* match B */
+ PUSH_STATE_GOTO(CURLYM_B, ST.B, locinput, loceol, /* match B */
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case CURLYM_B_fail: /* just failed to match a B */
n = (ST.oldloc == locinput) ? 0 : 1;
if (ST.c1 == ST.c2) {
/* set n to utf8_distance(oldloc, locinput) */
- while (locinput <= ST.maxpos
- && memNE(locinput, ST.c1_utf8, UTF8SKIP(locinput)))
+ while ( locinput <= ST.maxpos
+ && locinput < loceol
+ && memNE(locinput, ST.c1_utf8,
+ UTF8_SAFE_SKIP(locinput, reginfo->strend)))
{
- locinput += UTF8SKIP(locinput);
+ locinput += UTF8_SAFE_SKIP(locinput,
+ reginfo->strend);
n++;
}
}
else {
/* set n to utf8_distance(oldloc, locinput) */
- while (locinput <= ST.maxpos
- && memNE(locinput, ST.c1_utf8, UTF8SKIP(locinput))
- && memNE(locinput, ST.c2_utf8, UTF8SKIP(locinput)))
+ while ( locinput <= ST.maxpos
+ && locinput < loceol
+ && memNE(locinput, ST.c1_utf8,
+ UTF8_SAFE_SKIP(locinput, reginfo->strend))
+ && memNE(locinput, ST.c2_utf8,
+ UTF8_SAFE_SKIP(locinput, reginfo->strend)))
{
- locinput += UTF8SKIP(locinput);
+ locinput += UTF8_SAFE_SKIP(locinput, reginfo->strend);
n++;
}
}
curly_try_B_min:
CURLY_SETPAREN(ST.paren, ST.count);
- PUSH_STATE_GOTO(CURLY_B_min, ST.B, locinput, loceol);
+ PUSH_STATE_GOTO(CURLY_B_min, ST.B, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
if (ST.c1 != CHRTEST_VOID && could_match) {
if (! UTF8_IS_INVARIANT(UCHARAT(locinput)) && utf8_target)
{
- could_match = memEQ(locinput,
- ST.c1_utf8,
- UTF8SKIP(locinput))
- || memEQ(locinput,
- ST.c2_utf8,
- UTF8SKIP(locinput));
+ could_match = memEQ(locinput, ST.c1_utf8,
+ UTF8_SAFE_SKIP(locinput,
+ reginfo->strend))
+ || memEQ(locinput, ST.c2_utf8,
+ UTF8_SAFE_SKIP(locinput,
+ reginfo->strend));
}
else {
- could_match = UCHARAT(locinput) == ST.c1
- || UCHARAT(locinput) == ST.c2;
+ could_match = UCHARAT(locinput) == ST.c1
+ || UCHARAT(locinput) == ST.c2;
}
}
if (ST.c1 == CHRTEST_VOID || could_match) {
CURLY_SETPAREN(ST.paren, ST.count);
- PUSH_STATE_GOTO(CURLY_B_max, ST.B, locinput, loceol);
+ PUSH_STATE_GOTO(CURLY_B_max, ST.B, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
}
}
SET_RECURSE_LOCINPUT("FAKE-END[after]", cur_eval->locinput);
- PUSH_YES_STATE_GOTO(EVAL_postponed_AB, st->u.eval.prev_eval->u.eval.B,
- locinput, loceol); /* match B */
+ PUSH_YES_STATE_GOTO(EVAL_postponed_AB, /* match B */
+ st->u.eval.prev_eval->u.eval.B,
+ locinput, loceol, script_run_begin);
}
if (locinput < reginfo->till) {
PERL_UINT_FAST8_T back_count = scan->flags;
char * s;
- /* Lookbehind ends here */
- ST.end = locinput;
+ /* Lookbehind can look beyond the current position */
+ ST.end = loceol;
/* ... and starts at the first place in the input that is in
* the range of the possible start positions */
logical = 0; /* XXX: reset state of logical once it has been saved into ST */
/* execute body of (?...A) */
- PUSH_YES_STATE_GOTO(IFMATCH_A, NEXTOPER(NEXTOPER(scan)), ST.start, ST.end);
+ PUSH_YES_STATE_GOTO(IFMATCH_A, NEXTOPER(NEXTOPER(scan)), ST.start,
+ ST.end, script_run_begin);
NOT_REACHED; /* NOTREACHED */
{
/* restore old position except for (?>...) */
locinput = st->locinput;
loceol = st->loceol;
+ script_run_begin = st->sr0;
}
scan = ST.me + ARG(ST.me);
if (scan == ST.me)
case PRUNE: /* (*PRUNE) */
if (scan->flags)
sv_yes_mark = sv_commit = MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
- PUSH_STATE_GOTO(COMMIT_next, next, locinput, loceol);
+ PUSH_STATE_GOTO(COMMIT_next, next, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case COMMIT_next_fail:
= MUTABLE_SV(rexi->data->data[ ARG( scan ) ]);
mark_state = st;
ST.mark_loc = locinput;
- PUSH_YES_STATE_GOTO(MARKPOINT_next, next, locinput, loceol);
+ PUSH_YES_STATE_GOTO(MARKPOINT_next, next, locinput, loceol,
+ script_run_begin);
NOT_REACHED; /* NOTREACHED */
case MARKPOINT_next:
/* (*SKIP) : if we fail we cut here*/
ST.mark_name = NULL;
ST.mark_loc = locinput;
- PUSH_STATE_GOTO(SKIP_next,next, locinput, loceol);
+ PUSH_STATE_GOTO(SKIP_next,next, locinput, loceol,
+ script_run_begin);
} else {
/* (*SKIP:NAME) : if there is a (*MARK:NAME) fail where it was,
otherwise do nothing. Meaning we need to scan
find ) )
{
ST.mark_name = find;
- PUSH_STATE_GOTO( SKIP_next, next, locinput, loceol);
+ PUSH_STATE_GOTO( SKIP_next, next, locinput, loceol,
+ script_run_begin);
}
cur = cur->u.mark.prev_mark;
}
/* push a new regex state, then continue at scan */
{
regmatch_state *newst;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
- DEBUG_STACK_r({
+ DEBUG_r( /* DEBUG_STACK_r */
+ if (DEBUG_v_TEST || RE_DEBUG_FLAG(RE_DEBUG_EXTRA_STACK)) {
regmatch_state *cur = st;
regmatch_state *curyes = yes_state;
U32 i;
if (curyes == cur)
curyes = cur->u.yes.prev_yes_state;
}
- } else
+ } else {
DEBUG_STATE_pp("push")
- );
+ });
depth++;
st->locinput = locinput;
st->loceol = loceol;
+ st->sr0 = script_run_begin;
newst = st+1;
if (newst > SLAB_LAST(PL_regmatch_slab))
newst = S_push_slab(aTHX);
locinput = pushinput;
loceol = pusheol;
+ script_run_begin = pushsr0;
st = newst;
continue;
/* NOTREACHED */
if (no_final) {
locinput= st->locinput;
loceol= st->loceol;
+ script_run_begin = st->sr0;
}
state_num = st->resume_state + no_final;
goto reenter_switch;
PL_regmatch_state = st;
locinput= st->locinput;
loceol= st->loceol;
+ script_run_begin = st->sr0;
DEBUG_STATE_pp("pop");
depth--;
S_regrepeat(pTHX_ regexp *prog, char **startposp, const regnode *p,
char * loceol, regmatch_info *const reginfo, I32 max _pDEPTH)
{
- dVAR;
char *scan; /* Pointer to current position in target string */
I32 c;
char *this_eol = loceol; /* potentially adjusted version. */
else
scan = this_eol;
break;
+
+ case LEXACT_REQ8:
+ if (! utf8_target) {
+ break;
+ }
+ /* FALLTHROUGH */
+
+ case LEXACT:
+ {
+ U8 * string;
+ Size_t str_len;
+
+ string = (U8 *) STRINGl(p);
+ str_len = STR_LENl(p);
+ goto join_short_long_exact;
+
case EXACTL:
_CHECK_AND_WARN_PROBLEMATIC_LOCALE;
if (utf8_target && UTF8_IS_ABOVE_LATIN1(*scan)) {
}
goto do_exact;
- case EXACT_ONLY8:
+ case EXACT_REQ8:
if (! utf8_target) {
break;
}
/* FALLTHROUGH */
case EXACT:
do_exact:
- assert(STR_LEN(p) == reginfo->is_utf8_pat ? UTF8SKIP(STRING(p)) : 1);
+ string = (U8 *) STRINGs(p);
+ str_len = STR_LENs(p);
- c = (U8)*STRING(p);
+ join_short_long_exact:
+ assert(str_len == reginfo->is_utf8_pat ? UTF8SKIP(string) : 1);
+
+ c = *string;
/* Can use a simple find if the pattern char to match on is invariant
* under UTF-8, or both target and pattern aren't UTF-8. Note that we
* string EQ */
while (hardcount < max
&& scan < this_eol
- && (scan_char_len = UTF8SKIP(scan)) <= STR_LEN(p)
- && memEQ(scan, STRING(p), scan_char_len))
+ && (scan_char_len = UTF8SKIP(scan)) <= str_len
+ && memEQ(scan, string, scan_char_len))
{
scan += scan_char_len;
hardcount++;
/* Target isn't utf8; convert the character in the UTF-8
* pattern to non-UTF8, and do a simple find */
- c = EIGHT_BIT_UTF8_TO_NATIVE(c, *(STRING(p) + 1));
+ c = EIGHT_BIT_UTF8_TO_NATIVE(c, *(string + 1));
scan = (char *) find_span_end((U8 *) scan, (U8 *) this_eol, (U8) c);
} /* else pattern char is above Latin1, can't possibly match the
non-UTF-8 target */
}
}
break;
+ }
case EXACTFAA_NO_TRIE: /* This node only generated for non-utf8 patterns */
assert(! reginfo->is_utf8_pat);
| FOLDEQ_S2_FOLDS_SANE;
goto do_exactf;
- case EXACTFU_ONLY8:
+ case EXACTFU_REQ8:
if (! utf8_target) {
break;
}
int c1, c2;
U8 c1_utf8[UTF8_MAXBYTES+1], c2_utf8[UTF8_MAXBYTES+1];
- assert(STR_LEN(p) == reginfo->is_utf8_pat ? UTF8SKIP(STRING(p)) : 1);
+ assert(STR_LENs(p) == reginfo->is_utf8_pat ? UTF8SKIP(STRINGs(p)) : 1);
if (S_setup_EXACTISH_ST_c1_c2(aTHX_ p, &c1, c1_utf8, &c2, c2_utf8,
reginfo))
if (c1 == CHRTEST_VOID) {
/* Use full Unicode fold matching */
char *tmpeol = loceol;
- STRLEN pat_len = reginfo->is_utf8_pat ? UTF8SKIP(STRING(p)) : 1;
+ STRLEN pat_len = reginfo->is_utf8_pat ? UTF8SKIP(STRINGs(p)) : 1;
while (hardcount < max
&& foldEQ_utf8_flags(scan, &tmpeol, 0, utf8_target,
- STRING(p), NULL, pat_len,
+ STRINGs(p), NULL, pat_len,
reginfo->is_utf8_pat, utf8_flags))
{
scan = tmpeol;
if (c1 == c2) {
while (scan < this_eol
&& hardcount < max
- && memEQ(scan, c1_utf8, UTF8SKIP(scan)))
+ && memEQ(scan, c1_utf8, UTF8_SAFE_SKIP(scan,
+ loceol)))
{
- scan += UTF8SKIP(scan);
+ scan += UTF8SKIP(c1_utf8);
hardcount++;
}
}
else {
while (scan < this_eol
&& hardcount < max
- && (memEQ(scan, c1_utf8, UTF8SKIP(scan))
- || memEQ(scan, c2_utf8, UTF8SKIP(scan))))
+ && ( memEQ(scan, c1_utf8, UTF8_SAFE_SKIP(scan,
+ loceol))
+ || memEQ(scan, c2_utf8, UTF8_SAFE_SKIP(scan,
+ loceol))))
{
- scan += UTF8SKIP(scan);
+ scan += UTF8_SAFE_SKIP(scan, loceol);
hardcount++;
}
}
break;
case ANYOFH:
- if (utf8_target) while ( hardcount < max
- && scan < this_eol
- && reginclass(prog, p, (U8*)scan, (U8*) this_eol,
- TRUE))
- {
- scan += UTF8SKIP(scan);
- hardcount++;
+ if (utf8_target) { /* ANYOFH only can match UTF-8 targets */
+ while ( hardcount < max
+ && scan < this_eol
+ && NATIVE_UTF8_TO_I8(*scan) >= ANYOF_FLAGS(p)
+ && reginclass(prog, p, (U8*)scan, (U8*) this_eol, TRUE))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ break;
+
+ case ANYOFHb:
+ if (utf8_target) { /* ANYOFHb only can match UTF-8 targets */
+
+ /* we know the first byte must be the FLAGS field */
+ while ( hardcount < max
+ && scan < this_eol
+ && (U8) *scan == ANYOF_FLAGS(p)
+ && reginclass(prog, p, (U8*)scan, (U8*) this_eol,
+ TRUE))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ break;
+
+ case ANYOFHr:
+ if (utf8_target) { /* ANYOFH only can match UTF-8 targets */
+ while ( hardcount < max
+ && scan < this_eol
+ && inRANGE(NATIVE_UTF8_TO_I8(*scan),
+ LOWEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(p)),
+ HIGHEST_ANYOF_HRx_BYTE(ANYOF_FLAGS(p)))
+ && NATIVE_UTF8_TO_I8(*scan) >= ANYOF_FLAGS(p)
+ && reginclass(prog, p, (U8*)scan, (U8*) this_eol, TRUE))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ break;
+
+ case ANYOFHs:
+ if (utf8_target) { /* ANYOFH only can match UTF-8 targets */
+ while ( hardcount < max
+ && scan + FLAGS(p) < this_eol
+ && memEQ(scan, ((struct regnode_anyofhs *) p)->string, FLAGS(p))
+ && reginclass(prog, p, (U8*)scan, (U8*) this_eol, TRUE))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ break;
+
+ case ANYOFR:
+ if (utf8_target) {
+ while ( hardcount < max
+ && scan < this_eol
+ && NATIVE_UTF8_TO_I8(*scan) >= ANYOF_FLAGS(p)
+ && withinCOUNT(utf8_to_uvchr_buf((U8 *) scan,
+ (U8 *) this_eol,
+ NULL),
+ ANYOFRbase(p), ANYOFRdelta(p)))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ else {
+ while ( hardcount < max
+ && scan < this_eol
+ && withinCOUNT((U8) *scan, ANYOFRbase(p), ANYOFRdelta(p)))
+ {
+ scan++;
+ hardcount++;
+ }
+ }
+ break;
+
+ case ANYOFRb:
+ if (utf8_target) {
+ while ( hardcount < max
+ && scan < this_eol
+ && (U8) *scan == ANYOF_FLAGS(p)
+ && withinCOUNT(utf8_to_uvchr_buf((U8 *) scan,
+ (U8 *) this_eol,
+ NULL),
+ ANYOFRbase(p), ANYOFRdelta(p)))
+ {
+ scan += UTF8SKIP(scan);
+ hardcount++;
+ }
+ }
+ else {
+ while ( hardcount < max
+ && scan < this_eol
+ && withinCOUNT((U8) *scan, ANYOFRbase(p), ANYOFRdelta(p)))
+ {
+ scan++;
+ hardcount++;
+ }
}
break;
*startposp = scan;
DEBUG_r({
- GET_RE_DEBUG_FLAGS_DECL;
+ DECLARE_AND_GET_RE_DEBUG_FLAGS;
DEBUG_EXECUTE_r({
SV * const prop = sv_newmortal();
regprop(prog, prop, p, reginfo, NULL);
STATIC bool
S_reginclass(pTHX_ regexp * const prog, const regnode * const n, const U8* const p, const U8* const p_end, const bool utf8_target)
{
- dVAR;
- const char flags = ANYOF_FLAGS(n);
+ const char flags = (inRANGE(OP(n), ANYOFH, ANYOFHs))
+ ? 0
+ : ANYOF_FLAGS(n);
bool match = FALSE;
UV c = *p;
}
/* If this character is potentially in the bitmap, check it */
- if (c < NUM_ANYOF_CODE_POINTS && OP(n) != ANYOFH) {
+ if (c < NUM_ANYOF_CODE_POINTS && ! inRANGE(OP(n), ANYOFH, ANYOFHb)) {
if (ANYOF_BITMAP_TEST(n, c))
match = TRUE;
else if ((flags
}
else if (flags & ANYOF_LOCALE_FLAGS) {
if ( (flags & ANYOFL_FOLD)
- && c < sizeof(PL_fold_locale)
+ && c < 256
&& ANYOF_BITMAP_TEST(n, PL_fold_locale[c]))
{
match = TRUE;
&& IN_UTF8_CTYPE_LOCALE)))
{
SV* only_utf8_locale = NULL;
- SV * const definition = _get_regclass_nonbitmap_data(prog, n, TRUE,
- 0, &only_utf8_locale, NULL);
+ SV * const definition =
+#if !defined(PERL_IN_XSUB_RE) || defined(PLUGGABLE_RE_EXTENSION)
+ get_regclass_nonbitmap_data(prog, n, TRUE, 0,
+ &only_utf8_locale, NULL);
+#else
+ get_re_gclass_nonbitmap_data(prog, n, TRUE, 0,
+ &only_utf8_locale, NULL);
+#endif
if (definition) {
U8 utf8_buffer[2];
U8 * utf8_p;
regmatch_info_aux_eval *eval_state = reginfo->info_aux_eval;
eval_state->rex = rex;
+ eval_state->sv = reginfo->sv;
if (reginfo->sv) {
/* Make $_ available to executed code. */
SAVE_DEFSV;
DEFSV_set(reginfo->sv);
}
+ /* will be dec'd by S_cleanup_regmatch_info_aux */
+ SvREFCNT_inc_NN(reginfo->sv);
if (!(mg = mg_find_mglob(reginfo->sv))) {
/* prepare for quick setting of pos */
/* this regexp is also owned by the new PL_reg_curpm, which
will try to free it. */
av_push(PL_regex_padav, repointer);
- PL_reg_curpm->op_pmoffset = av_tindex(PL_regex_padav);
+ PL_reg_curpm->op_pmoffset = av_top_index(PL_regex_padav);
PL_regex_pad = AvARRAY(PL_regex_padav);
}
#endif
}
PL_curpm = eval_state->curpm;
+ SvREFCNT_dec(eval_state->sv);
}
PL_regmatch_state = aux->old_regmatch_state;
&& !prog->substrs->data[i].substr) {
SV* sv = newSVsv(prog->substrs->data[i].utf8_substr);
if (! sv_utf8_downgrade(sv, TRUE)) {
+ SvREFCNT_dec_NN(sv);
return FALSE;
}
if (SvVALID(prog->substrs->data[i].utf8_substr)) {
#ifndef PERL_IN_XSUB_RE
bool
-Perl__is_grapheme(pTHX_ const U8 * strbeg, const U8 * s, const U8 * strend, const UV cp)
+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
+ * '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. */
- dVAR;
GCB_enum cp_gcb_val, prev_cp_gcb_val, next_cp_gcb_val;
const U8 * prev_cp_start;
- PERL_ARGS_ASSERT__IS_GRAPHEME;
+ PERL_ARGS_ASSERT_IS_GRAPHEME;
if ( UNLIKELY(UNICODE_IS_SUPER(cp))
|| UNLIKELY(UNICODE_IS_NONCHAR(cp)))
}
/*
-=head1 Unicode Support
+=for apidoc_section Unicode Support
=for apidoc isSCRIPT_RUN
* characters for at least one language in the Unicode Common Locale Data
* Repository [CLDR]. */
- dVAR;
/* Things that match /\d/u */
SV * decimals_invlist = PL_XPosix_ptrs[_CC_DIGIT];
/* If is within the range [+0 .. +9] of the script's zero, it also is a
* digit in that script. We can skip the rest of this code for this
* character. */
- if (UNLIKELY( zero_of_run
- && cp >= zero_of_run
- && cp - zero_of_run <= 9))
- {
+ if (UNLIKELY(zero_of_run && withinCOUNT(cp, zero_of_run, 9))) {
continue;
}
* several scripts, and the intersection is not empty. However, if the
* character is a decimal digit, it could still mean failure if it is
* from the wrong sequence of 10. So, we need to look at if it's a
- * digit. We've already handled the 10 decimal digits, and the next
+ * digit. We've already handled the 10 digits [0-9], and the next
* lowest one is this one: */
if (cp < FIRST_NON_ASCII_DECIMAL_DIGIT) {
continue; /* Not a digit; this character is part of the run */
if ( script_of_char >= 0
&& (zero_of_char = script_zeros[script_of_char]))
{
- if ( cp < zero_of_char
- || cp > zero_of_char + 9)
- {
+ if (! withinCOUNT(cp, zero_of_char, 9)) {
continue; /* Not a digit; this character is part of the run
*/
}
https://perl5.git.perl.org / perl5.git / blobdiff