# define PERL_NO_GET_CONTEXT
#endif
-/*SUPPRESS 112*/
/*
* pregcomp and pregexec -- regsub and regerror are not used in perl
*
**** Alterations to Henry's code are...
****
**** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
- **** 2000, 2001, 2002, 2003, 2004, 2005, by Larry Wall and others
+ **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, by Larry Wall and others
****
**** You may distribute under the terms of either the GNU General Public
**** License or the Artistic License, as specified in the README file.
* Forward declarations for pregcomp()'s friends.
*/
-static scan_data_t zero_scan_data = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0};
+static const scan_data_t zero_scan_data =
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
#define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
#define SF_BEFORE_SEOL 0x1
} STMT_END
/*
- * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
- * args. Show regex, up to a maximum length. If it's too long, chop and add
- * "...".
- */
-#define FAIL2(pat,msg) STMT_START { \
- const char *ellipses = ""; \
- IV len = RExC_end - RExC_precomp; \
- \
- if (!SIZE_ONLY) \
- SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx); \
- if (len > RegexLengthToShowInErrorMessages) { \
- /* chop 10 shorter than the max, to ensure meaning of "..." */ \
- len = RegexLengthToShowInErrorMessages - 10; \
- ellipses = "..."; \
- } \
- S_re_croak2(aTHX_ pat, " in regex m/%.*s%s/", \
- msg, (int)len, RExC_precomp, ellipses); \
-} STMT_END
-
-
-/*
* Simple_vFAIL -- like FAIL, but marks the current location in the scan
*/
#define Simple_vFAIL(m) STMT_START { \
- IV offset = RExC_parse - RExC_precomp; \
+ const IV offset = RExC_parse - RExC_precomp; \
Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
m, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
* Like Simple_vFAIL(), but accepts two arguments.
*/
#define Simple_vFAIL2(m,a1) STMT_START { \
- IV offset = RExC_parse - RExC_precomp; \
+ const IV offset = RExC_parse - RExC_precomp; \
S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
(int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
* Like Simple_vFAIL(), but accepts three arguments.
*/
#define Simple_vFAIL3(m, a1, a2) STMT_START { \
- IV offset = RExC_parse - RExC_precomp; \
+ const IV offset = RExC_parse - RExC_precomp; \
S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
(int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
* Like Simple_vFAIL(), but accepts four arguments.
*/
#define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
- IV offset = RExC_parse - RExC_precomp; \
+ const IV offset = RExC_parse - RExC_precomp; \
S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
(int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
-/*
- * Like Simple_vFAIL(), but accepts five arguments.
- */
-#define Simple_vFAIL5(m, a1, a2, a3, a4) STMT_START { \
- IV offset = RExC_parse - RExC_precomp; \
- S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, a4, \
- (int)offset, RExC_precomp, RExC_precomp + offset); \
-} STMT_END
-
-
#define vWARN(loc,m) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s" REPORT_LOCATION, \
m, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
#define vWARNdep(loc,m) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
"%s" REPORT_LOCATION, \
m, (int)offset, RExC_precomp, RExC_precomp + offset); \
#define vWARN2(loc, m, a1) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
#define vWARN3(loc, m, a1, a2) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
#define vWARN4(loc, m, a1, a2, a3) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
#define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
- IV offset = loc - RExC_precomp; \
+ const IV offset = loc - RExC_precomp; \
Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
} STMT_END
MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
__LINE__, (node), (byte))); \
if((node) < 0) { \
- Perl_croak(aTHX_ "value of node is %d in Offset macro", node); \
+ Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
} else { \
RExC_offsets[2*(node)-1] = (byte); \
} \
#define Set_Node_Length_To_R(node,len) STMT_START { \
if (! SIZE_ONLY) { \
MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
- __LINE__, (node), (len))); \
+ __LINE__, (int)(node), (int)(len))); \
if((node) < 0) { \
- Perl_croak(aTHX_ "value of node is %d in Length macro", node); \
+ Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
} else { \
RExC_offsets[2*(node)] = (len); \
} \
}
SvCUR_set(data->last_found, 0);
{
- SV * sv = data->last_found;
- MAGIC *mg =
+ SV * const sv = data->last_found;
+ MAGIC * const mg =
SvUTF8(sv) && SvMAGICAL(sv) ? mg_find(sv, PERL_MAGIC_utf8) : NULL;
- if (mg && mg->mg_len > 0)
+ if (mg)
mg->mg_len = 0;
}
data->last_end = -1;
/* Can match anything (initialization) */
STATIC int
-S_cl_is_anything(pTHX_ struct regnode_charclass_class *cl)
+S_cl_is_anything(pTHX_ const struct regnode_charclass_class *cl)
{
int value;
/* We assume that cl is not inverted */
STATIC void
S_cl_and(pTHX_ struct regnode_charclass_class *cl,
- struct regnode_charclass_class *and_with)
+ const struct regnode_charclass_class *and_with)
{
if (!(and_with->flags & ANYOF_CLASS)
&& !(cl->flags & ANYOF_CLASS)
/* 'OR' a given class with another one. Can create false positives */
/* We assume that cl is not inverted */
STATIC void
-S_cl_or(pTHX_ RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, struct regnode_charclass_class *or_with)
+S_cl_or(pTHX_ RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
{
if (or_with->flags & ANYOF_INVERT) {
/* We do not use
/* REx optimizer. Converts nodes into quickier variants "in place".
Finds fixed substrings. */
-/* Stops at toplevel WHILEM as well as at `last'. At end *scanp is set
+/* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
to the position after last scanned or to NULL. */
STATIC I32
n = regnext(n);
}
else if (stringok) {
- int oldl = STR_LEN(scan);
- regnode *nnext = regnext(n);
+ const int oldl = STR_LEN(scan);
+ regnode * const nnext = regnext(n);
if (oldl + STR_LEN(n) > U8_MAX)
break;
another valid sequence of UTF-8 bytes.
*/
- char *s0 = STRING(scan), *s, *t;
- char *s1 = s0 + STR_LEN(scan) - 1, *s2 = s1 - 4;
- const char *t0 = "\xcc\x88\xcc\x81";
- const char *t1 = t0 + 3;
-
+ char * const s0 = STRING(scan), *s, *t;
+ char * const s1 = s0 + STR_LEN(scan) - 1;
+ char * const s2 = s1 - 4;
+ const char * const t0 = "\xcc\x88\xcc\x81";
+ const char * const t1 = t0 + 3;
+
for (s = s0 + 2;
s < s2 && (t = ninstr(s, s1, t0, t1));
s = t + 4) {
/* Follow the next-chain of the current node and optimize
away all the NOTHINGs from it. */
if (OP(scan) != CURLYX) {
- int max = (reg_off_by_arg[OP(scan)]
+ const int max = (reg_off_by_arg[OP(scan)]
? I32_MAX
/* I32 may be smaller than U16 on CRAYs! */
: (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
data->start_class->flags |= ANYOF_EOS;
}
}
+
}
else if (code == BRANCHJ) /* single branch is optimized. */
scan = NEXTOPER(NEXTOPER(scan));
}
else if (OP(scan) == EXACT) {
I32 l = STR_LEN(scan);
- UV uc = *((U8*)STRING(scan));
+ UV uc;
if (UTF) {
- U8 *s = (U8*)STRING(scan);
- l = utf8_length(s, s + l);
- uc = utf8_to_uvchr(s, NULL);
+ const U8 * const s = (U8*)STRING(scan);
+ l = utf8_length((U8 *)s, (U8 *)s + l);
+ uc = utf8_to_uvchr((U8 *)s, NULL);
+ } else {
+ uc = *((U8*)STRING(scan));
}
min += l;
if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
? I32_MAX : data->pos_min + data->pos_delta;
}
sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
+ if (UTF)
+ SvUTF8_on(data->last_found);
{
- SV * sv = data->last_found;
- MAGIC *mg = SvUTF8(sv) && SvMAGICAL(sv) ?
+ SV * const sv = data->last_found;
+ MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
mg_find(sv, PERL_MAGIC_utf8) : NULL;
if (mg && mg->mg_len >= 0)
mg->mg_len += utf8_length((U8*)STRING(scan),
(U8*)STRING(scan)+STR_LEN(scan));
}
- if (UTF)
- SvUTF8_on(data->last_found);
data->last_end = data->pos_min + l;
data->pos_min += l; /* As in the first entry. */
data->flags &= ~SF_BEFORE_EOL;
if (flags & SCF_DO_SUBSTR)
scan_commit(pRExC_state, data);
if (UTF) {
- U8 *s = (U8 *)STRING(scan);
+ U8 * const s = (U8 *)STRING(scan);
l = utf8_length(s, s + l);
uc = utf8_to_uvchr(s, NULL);
}
}
if (!scan) /* It was not CURLYX, but CURLY. */
scan = next;
- if (ckWARN(WARN_REGEXP)
- /* ? quantifier ok, except for (?{ ... }) */
- && (next_is_eval || !(mincount == 0 && maxcount == 1))
+ if ( /* ? quantifier ok, except for (?{ ... }) */
+ (next_is_eval || !(mincount == 0 && maxcount == 1))
&& (minnext == 0) && (deltanext == 0)
&& data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
- && maxcount <= REG_INFTY/3) /* Complement check for big count */
+ && maxcount <= REG_INFTY/3 /* Complement check for big count */
+ && ckWARN(WARN_REGEXP))
{
vWARN(RExC_parse,
"Quantifier unexpected on zero-length expression");
if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
pars++;
if (flags & SCF_DO_SUBSTR) {
- SV *last_str = Nullsv;
+ SV *last_str = NULL;
int counted = mincount != 0;
if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
#if defined(SPARC64_GCC_WORKAROUND)
I32 b = 0;
STRLEN l = 0;
- char *s = NULL;
+ const char *s = NULL;
I32 old = 0;
if (pos_before >= data->last_start_min)
b = data->last_start_min;
l = 0;
- s = SvPV(data->last_found, l);
+ s = SvPV_const(data->last_found, l);
old = b - data->last_start_min;
#else
I32 b = pos_before >= data->last_start_min
? pos_before : data->last_start_min;
STRLEN l;
- char *s = SvPV(data->last_found, l);
+ const char *s = SvPV_const(data->last_found, l);
I32 old = b - data->last_start_min;
#endif
if (mincount > 1) {
SvGROW(last_str, (mincount * l) + 1);
repeatcpy(SvPVX(last_str) + l,
- SvPVX(last_str), l, mincount - 1);
+ SvPVX_const(last_str), l, mincount - 1);
SvCUR_set(last_str, SvCUR(last_str) * mincount);
/* Add additional parts. */
SvCUR_set(data->last_found,
the group. */
scan_commit(pRExC_state,data);
if (mincount && last_str) {
- sv_setsv(data->last_found, last_str);
+ SV *sv = data->last_found;
+ MAGIC *mg = SvUTF8(sv) && SvMAGICAL(sv) ?
+ mg_find(sv, PERL_MAGIC_utf8) : NULL;
+
+ if (mg)
+ mg->mg_len = -1;
+ sv_setsv(sv, last_str);
data->last_end = data->pos_min;
data->last_start_min =
data->pos_min - CHR_SVLEN(last_str);
if (data)
data->whilem_c = data_fake.whilem_c;
if (f & SCF_DO_STCLASS_AND) {
- int was = (data->start_class->flags & ANYOF_EOS);
+ const int was = (data->start_class->flags & ANYOF_EOS);
cl_and(data->start_class, &intrnl);
if (was)
RExC_rx->data->count += n;
}
else {
- Newc(1207, RExC_rx->data, sizeof(*RExC_rx->data) + sizeof(void*) * (n - 1),
+ Newxc(RExC_rx->data, sizeof(*RExC_rx->data) + sizeof(void*) * (n - 1),
char, struct reg_data);
- New(1208, RExC_rx->data->what, n, U8);
+ Newx(RExC_rx->data->what, n, U8);
RExC_rx->data->count = n;
}
Copy(s, RExC_rx->data->what + RExC_rx->data->count - n, n, U8);
void
Perl_reginitcolors(pTHX)
{
- int i = 0;
- char *s = PerlEnv_getenv("PERL_RE_COLORS");
-
+ const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
if (s) {
- PL_colors[0] = s = savepv(s);
+ char *t = savepv(s);
+ int i = 0;
+ PL_colors[0] = t;
while (++i < 6) {
- s = strchr(s, '\t');
- if (s) {
- *s = '\0';
- PL_colors[i] = ++s;
+ t = strchr(t, '\t');
+ if (t) {
+ *t = '\0';
+ PL_colors[i] = ++t;
}
else
- PL_colors[i] = s = (char *)"";
+ PL_colors[i] = t = (char *)"";
}
} else {
+ int i = 0;
while (i < 6)
- PL_colors[i++] = "";
+ PL_colors[i++] = (char *)"";
}
PL_colorset = 1;
}
REGC((U8)REG_MAGIC, (char*)RExC_emit);
#endif
if (reg(pRExC_state, 0, &flags) == NULL) {
- RExC_precomp = Nullch;
+ RExC_precomp = NULL;
return(NULL);
}
DEBUG_r(PerlIO_printf(Perl_debug_log, "size %"IVdf" ", (IV)RExC_size));
RExC_whilem_seen = 15;
/* Allocate space and initialize. */
- Newc(1001, r, sizeof(regexp) + (unsigned)RExC_size * sizeof(regnode),
+ Newxc(r, sizeof(regexp) + (unsigned)RExC_size * sizeof(regnode),
char, regexp);
if (r == NULL)
FAIL("Regexp out of space");
r->subbeg = NULL;
r->reganch = pm->op_pmflags & PMf_COMPILETIME;
r->nparens = RExC_npar - 1; /* set early to validate backrefs */
+ r->lastparen = 0; /* mg.c reads this. */
r->substrs = 0; /* Useful during FAIL. */
r->startp = 0; /* Useful during FAIL. */
r->endp = 0; /* Useful during FAIL. */
- Newz(1304, r->offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
+ Newxz(r->offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
if (r->offsets) {
- r->offsets[0] = RExC_size;
+ r->offsets[0] = RExC_size;
}
DEBUG_r(PerlIO_printf(Perl_debug_log,
- "%s %"UVuf" bytes for offset annotations.\n",
- r->offsets ? "Got" : "Couldn't get",
+ "%s %"UVuf" bytes for offset annotations.\n",
+ r->offsets ? "Got" : "Couldn't get",
(UV)((2*RExC_size+1) * sizeof(U32))));
RExC_rx = r;
/* XXXX To minimize changes to RE engine we always allocate
3-units-long substrs field. */
- Newz(1004, r->substrs, 1, struct reg_substr_data);
+ Newxz(r->substrs, 1, struct reg_substr_data);
StructCopy(&zero_scan_data, &data, scan_data_t);
/* XXXX Should not we check for something else? Usually it is OPEN1... */
!(r->reganch & ROPT_ANCH) )
{
/* turn .* into ^.* with an implied $*=1 */
- int type = OP(NEXTOPER(first));
-
- if (type == REG_ANY)
- type = ROPT_ANCH_MBOL;
- else
- type = ROPT_ANCH_SBOL;
-
+ const int type =
+ (OP(NEXTOPER(first)) == REG_ANY)
+ ? ROPT_ANCH_MBOL
+ : ROPT_ANCH_SBOL;
r->reganch |= type | ROPT_IMPLICIT;
first = NEXTOPER(first);
goto again;
*/
minlen = 0;
- data.longest_fixed = newSVpvn("",0);
- data.longest_float = newSVpvn("",0);
- data.last_found = newSVpvn("",0);
+ data.longest_fixed = newSVpvs("");
+ data.longest_float = newSVpvs("");
+ data.last_found = newSVpvs("");
data.longest = &(data.longest_fixed);
first = scan;
if (!r->regstclass) {
if (SvUTF8(data.longest_float)) {
r->float_utf8 = data.longest_float;
- r->float_substr = Nullsv;
+ r->float_substr = NULL;
} else {
r->float_substr = data.longest_float;
- r->float_utf8 = Nullsv;
+ r->float_utf8 = NULL;
}
r->float_min_offset = data.offset_float_min;
r->float_max_offset = data.offset_float_max;
}
else {
remove_float:
- r->float_substr = r->float_utf8 = Nullsv;
+ r->float_substr = r->float_utf8 = NULL;
SvREFCNT_dec(data.longest_float);
longest_float_length = 0;
}
if (SvUTF8(data.longest_fixed)) {
r->anchored_utf8 = data.longest_fixed;
- r->anchored_substr = Nullsv;
+ r->anchored_substr = NULL;
} else {
r->anchored_substr = data.longest_fixed;
- r->anchored_utf8 = Nullsv;
+ r->anchored_utf8 = NULL;
}
r->anchored_offset = data.offset_fixed;
t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
}
else {
- r->anchored_substr = r->anchored_utf8 = Nullsv;
+ r->anchored_substr = r->anchored_utf8 = NULL;
SvREFCNT_dec(data.longest_fixed);
longest_fixed_length = 0;
}
&& !(data.start_class->flags & ANYOF_EOS)
&& !cl_is_anything(data.start_class))
{
- I32 n = add_data(pRExC_state, 1, "f");
+ const I32 n = add_data(pRExC_state, 1, "f");
- New(1006, RExC_rx->data->data[n], 1,
+ Newx(RExC_rx->data->data[n], 1,
struct regnode_charclass_class);
StructCopy(data.start_class,
(struct regnode_charclass_class*)RExC_rx->data->data[n],
DEBUG_r({ SV *sv = sv_newmortal();
regprop(sv, (regnode*)data.start_class);
PerlIO_printf(Perl_debug_log,
- "synthetic stclass `%s'.\n",
- SvPVX(sv));});
+ "synthetic stclass \"%s\".\n",
+ SvPVX_const(sv));});
}
/* A temporary algorithm prefers floated substr to fixed one to dig more info. */
data.last_closep = &last_close;
minlen = study_chunk(pRExC_state, &scan, &fake, scan + RExC_size, &data, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS);
r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
- = r->float_substr = r->float_utf8 = Nullsv;
+ = r->float_substr = r->float_utf8 = NULL;
if (!(data.start_class->flags & ANYOF_EOS)
&& !cl_is_anything(data.start_class))
{
- I32 n = add_data(pRExC_state, 1, "f");
+ const I32 n = add_data(pRExC_state, 1, "f");
- New(1006, RExC_rx->data->data[n], 1,
+ Newx(RExC_rx->data->data[n], 1,
struct regnode_charclass_class);
StructCopy(data.start_class,
(struct regnode_charclass_class*)RExC_rx->data->data[n],
DEBUG_r({ SV* sv = sv_newmortal();
regprop(sv, (regnode*)data.start_class);
PerlIO_printf(Perl_debug_log,
- "synthetic stclass `%s'.\n",
- SvPVX(sv));});
+ "synthetic stclass \"%s\".\n",
+ SvPVX_const(sv));});
}
}
r->reganch |= ROPT_EVAL_SEEN;
if (RExC_seen & REG_SEEN_CANY)
r->reganch |= ROPT_CANY_SEEN;
- Newz(1002, r->startp, RExC_npar, I32);
- Newz(1002, r->endp, RExC_npar, I32);
+ Newxz(r->startp, RExC_npar, I32);
+ Newxz(r->endp, RExC_npar, I32);
PL_regdata = r->data; /* for regprop() */
DEBUG_r(regdump(r));
return(r);
register regnode *ret; /* Will be the head of the group. */
register regnode *br;
register regnode *lastbr;
- register regnode *ender = 0;
+ register regnode *ender = NULL;
register I32 parno = 0;
- I32 flags, oregflags = RExC_flags, have_branch = 0, open = 0;
+ I32 flags;
+ const I32 oregflags = RExC_flags;
+ bool have_branch = 0;
+ bool is_open = 0;
/* for (?g), (?gc), and (?o) warnings; warning
about (?c) will warn about (?g) -- japhy */
- I32 wastedflags = 0x00,
- wasted_o = 0x01,
- wasted_g = 0x02,
- wasted_gc = 0x02 | 0x04,
- wasted_c = 0x04;
+#define WASTED_O 0x01
+#define WASTED_G 0x02
+#define WASTED_C 0x04
+#define WASTED_GC (0x02|0x04)
+ I32 wastedflags = 0x00;
char * parse_start = RExC_parse; /* MJD */
- char *oregcomp_parse = RExC_parse;
- char c;
+ char * const oregcomp_parse = RExC_parse;
*flagp = 0; /* Tentatively. */
if (*RExC_parse == '?') { /* (?...) */
U32 posflags = 0, negflags = 0;
U32 *flagsp = &posflags;
- int logical = 0;
- char *seqstart = RExC_parse;
+ bool is_logical = 0;
+ const char * const seqstart = RExC_parse;
RExC_parse++;
paren = *RExC_parse++;
vWARNdep(RExC_parse, "(?p{}) is deprecated - use (??{})");
/* FALL THROUGH*/
case '?': /* (??...) */
- logical = 1;
+ is_logical = 1;
if (*RExC_parse != '{')
goto unknown;
paren = *RExC_parse++;
I32 count = 1, n = 0;
char c;
char *s = RExC_parse;
- SV *sv;
- OP_4tree *sop, *rop;
RExC_seen_zerolen++;
RExC_seen |= REG_SEEN_EVAL;
while (count && (c = *RExC_parse)) {
- if (c == '\\' && RExC_parse[1])
- RExC_parse++;
+ if (c == '\\') {
+ if (RExC_parse[1])
+ RExC_parse++;
+ }
else if (c == '{')
count++;
else if (c == '}')
count--;
RExC_parse++;
}
- if (*RExC_parse != ')')
- {
+ if (*RExC_parse != ')') {
RExC_parse = s;
vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
}
if (!SIZE_ONLY) {
PAD *pad;
-
- if (RExC_parse - 1 - s)
- sv = newSVpvn(s, RExC_parse - 1 - s);
- else
- sv = newSVpvn("", 0);
+ OP_4tree *sop, *rop;
+ SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
ENTER;
Perl_save_re_context(aTHX);
}
nextchar(pRExC_state);
- if (logical) {
+ if (is_logical) {
ret = reg_node(pRExC_state, LOGICAL);
if (!SIZE_ONLY)
ret->flags = 2;
}
else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
/* (?(1)...) */
+ char c;
parno = atoi(RExC_parse++);
while (isDIGIT(*RExC_parse))
RExC_parse++;
ret = reganode(pRExC_state, GROUPP, parno);
-
+
if ((c = *nextchar(pRExC_state)) != ')')
vFAIL("Switch condition not recognized");
insert_if:
if (*RExC_parse == 'o' || *RExC_parse == 'g') {
if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
- I32 wflagbit = *RExC_parse == 'o' ? wasted_o : wasted_g;
+ const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
if (! (wastedflags & wflagbit) ) {
wastedflags |= wflagbit;
vWARN5(
}
else if (*RExC_parse == 'c') {
if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
- if (! (wastedflags & wasted_c) ) {
- wastedflags |= wasted_gc;
+ if (! (wastedflags & WASTED_C) ) {
+ wastedflags |= WASTED_GC;
vWARN3(
RExC_parse + 1,
"Useless (%sc) - %suse /gc modifier",
ret = reganode(pRExC_state, OPEN, parno);
Set_Node_Length(ret, 1); /* MJD */
Set_Node_Offset(ret, RExC_parse); /* MJD */
- open = 1;
+ is_open = 1;
}
}
else /* ! paren */
parse_start = RExC_parse; /* MJD */
br = regbranch(pRExC_state, &flags, 1);
/* branch_len = (paren != 0); */
-
+
if (br == NULL)
return(NULL);
if (*RExC_parse == '|') {
else if (paren == ':') {
*flagp |= flags&SIMPLE;
}
- if (open) { /* Starts with OPEN. */
+ if (is_open) { /* Starts with OPEN. */
regtail(pRExC_state, ret, br); /* OPEN -> first. */
}
else if (paren != '?') /* Not Conditional */
RExC_extralen += 2; /* Account for LONGJMP. */
nextchar(pRExC_state);
br = regbranch(pRExC_state, &flags, 0);
-
+
if (br == NULL)
return(NULL);
regtail(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
register char op;
register char *next;
I32 flags;
- char *origparse = RExC_parse;
+ const char * const origparse = RExC_parse;
char *maxpos;
I32 min;
I32 max = REG_INFTY;
if (op == '{' && regcurly(RExC_parse)) {
parse_start = RExC_parse; /* MJD */
next = RExC_parse + 1;
- maxpos = Nullch;
+ maxpos = NULL;
while (isDIGIT(*next) || *next == ',') {
if (*next == ',') {
if (maxpos)
reginsert(pRExC_state, CURLYX,ret);
/* MJD hk */
Set_Node_Offset(ret, parse_start+1);
- Set_Node_Length(ret,
+ Set_Node_Length(ret,
op == '{' ? (RExC_parse - parse_start) : 1);
-
+
if (!SIZE_ONLY && RExC_extralen)
NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
regtail(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
goto do_curly;
}
nest_check:
- if (ckWARN(WARN_REGEXP) && !SIZE_ONLY && !(flags&HASWIDTH) && max > REG_INFTY/3) {
+ if (!SIZE_ONLY && !(flags&HASWIDTH) && max > REG_INFTY/3 && ckWARN(WARN_REGEXP)) {
vWARN3(RExC_parse,
"%.*s matches null string many times",
- RExC_parse - origparse,
+ (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
origparse);
}
STATIC regnode *
S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp)
{
- register regnode *ret = 0;
+ register regnode *ret = NULL;
I32 flags;
char *parse_start = RExC_parse;
case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
{
- I32 num = atoi(RExC_parse);
+ const I32 num = atoi(RExC_parse);
if (num > 9 && num >= RExC_npar)
goto defchar;
(U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
num);
*flagp |= HASWIDTH;
-
+
/* override incorrect value set in reganode MJD */
- Set_Node_Offset(ret, parse_start+1);
+ Set_Node_Offset(ret, parse_start+1);
Set_Node_Cur_Length(ret); /* MJD */
RExC_parse--;
nextchar(pRExC_state);
FAIL("Trailing \\");
/* FALL THROUGH */
default:
- /* Do not generate `unrecognized' warnings here, we fall
+ /* Do not generate "unrecognized" warnings here, we fall
back into the quick-grab loop below */
parse_start--;
goto defchar;
register UV ender;
register char *p;
char *oldp, *s;
- STRLEN numlen;
STRLEN foldlen;
U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
break;
case 'x':
if (*++p == '{') {
- char* e = strchr(p, '}');
+ char* const e = strchr(p, '}');
if (!e) {
RExC_parse = p + 1;
else {
I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
| PERL_SCAN_DISALLOW_PREFIX;
- numlen = e - p - 1;
+ STRLEN numlen = e - p - 1;
ender = grok_hex(p + 1, &numlen, &flags, NULL);
if (ender > 0xff)
RExC_utf8 = 1;
}
else {
I32 flags = PERL_SCAN_DISALLOW_PREFIX;
- numlen = 2;
+ STRLEN numlen = 2;
ender = grok_hex(p, &numlen, &flags, NULL);
p += numlen;
}
if (*p == '0' ||
(isDIGIT(p[1]) && atoi(p) >= RExC_npar) ) {
I32 flags = 0;
- numlen = 3;
+ STRLEN numlen = 3;
ender = grok_oct(p, &numlen, &flags, NULL);
p += numlen;
}
FAIL("Trailing \\");
/* FALL THROUGH */
default:
- if (!SIZE_ONLY && ckWARN(WARN_REGEXP) && isALPHA(*p))
+ if (!SIZE_ONLY&& isALPHA(*p) && ckWARN(WARN_REGEXP))
vWARN2(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
goto normal_default;
}
default:
normal_default:
if (UTF8_IS_START(*p) && UTF) {
+ STRLEN numlen;
ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
&numlen, 0);
p += numlen;
if (FOLD) {
/* Emit all the Unicode characters. */
+ STRLEN numlen;
for (foldbuf = tmpbuf;
foldlen;
foldlen -= numlen) {
if (FOLD) {
/* Emit all the Unicode characters. */
+ STRLEN numlen;
for (foldbuf = tmpbuf;
foldlen;
foldlen -= numlen) {
if (RExC_utf8)
SvUTF8_on(sv);
if (sv_utf8_downgrade(sv, TRUE)) {
- char *s = sv_recode_to_utf8(sv, PL_encoding);
- STRLEN newlen = SvCUR(sv);
+ const char * const s = sv_recode_to_utf8(sv, PL_encoding);
+ const STRLEN newlen = SvCUR(sv);
if (SvUTF8(sv))
RExC_utf8 = 1;
}
STATIC char *
-S_regwhite(pTHX_ char *p, char *e)
+S_regwhite(pTHX_ char *p, const char *e)
{
while (p < e) {
if (isSPACE(*p))
STATIC I32
S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
{
- char *posixcc = 0;
I32 namedclass = OOB_NAMEDCLASS;
if (value == '[' && RExC_parse + 1 < RExC_end &&
/* I smell either [: or [= or [. -- POSIX has been here, right? */
POSIXCC(UCHARAT(RExC_parse))) {
- char c = UCHARAT(RExC_parse);
+ const char c = UCHARAT(RExC_parse);
char* s = RExC_parse++;
while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
/* Grandfather lone [:, [=, [. */
RExC_parse = s;
else {
- char* t = RExC_parse++; /* skip over the c */
+ const char* t = RExC_parse++; /* skip over the c */
+ const char *posixcc;
assert(*t == c);
RExC_parse++; /* skip over the ending ] */
posixcc = s + 1;
if (*s == ':') {
- I32 complement = *posixcc == '^' ? *posixcc++ : 0;
- I32 skip = t - posixcc;
+ const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
+ const I32 skip = t - posixcc;
/* Initially switch on the length of the name. */
switch (skip) {
S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
{
if (!SIZE_ONLY && POSIXCC(UCHARAT(RExC_parse))) {
- char *s = RExC_parse;
- char c = *s++;
+ const char *s = RExC_parse;
+ const char c = *s++;
while(*s && isALNUM(*s))
s++;
register regnode *ret;
STRLEN numlen;
IV namedclass;
- char *rangebegin = 0;
+ char *rangebegin = NULL;
bool need_class = 0;
- SV *listsv = Nullsv;
+ SV *listsv = NULL;
register char *e;
UV n;
bool optimize_invert = TRUE;
- AV* unicode_alternate = 0;
+ AV* unicode_alternate = NULL;
#ifdef EBCDIC
UV literal_endpoint = 0;
#endif
if (LOC)
ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
ANYOF_BITMAP_ZERO(ret);
- listsv = newSVpvn("# comment\n", 10);
+ listsv = newSVpvs("# comment\n");
}
nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
if (RExC_parse >= RExC_end)
vFAIL2("Empty \\%c{}", (U8)value);
if (*RExC_parse == '{') {
- U8 c = (U8)value;
+ const U8 c = (U8)value;
e = strchr(RExC_parse++, '}');
if (!e)
vFAIL2("Missing right brace on \\%c{}", c);
break;
}
default:
- if (!SIZE_ONLY && ckWARN(WARN_REGEXP) && isALPHA(value))
+ if (!SIZE_ONLY && isALPHA(value) && ckWARN(WARN_REGEXP))
vWARN2(RExC_parse,
"Unrecognized escape \\%c in character class passed through",
(int)value);
/* a bad range like a-\d, a-[:digit:] ? */
if (range) {
if (!SIZE_ONLY) {
- if (ckWARN(WARN_REGEXP))
+ if (ckWARN(WARN_REGEXP)) {
+ int w =
+ RExC_parse >= rangebegin ?
+ RExC_parse - rangebegin : 0;
vWARN4(RExC_parse,
"False [] range \"%*.*s\"",
- RExC_parse - rangebegin,
- RExC_parse - rangebegin,
+ w,
+ w,
rangebegin);
+ }
if (prevvalue < 256) {
ANYOF_BITMAP_SET(ret, prevvalue);
ANYOF_BITMAP_SET(ret, '-');
/* a bad range like \w-, [:word:]- ? */
if (namedclass > OOB_NAMEDCLASS) {
- if (ckWARN(WARN_REGEXP))
+ if (ckWARN(WARN_REGEXP)) {
+ int w =
+ RExC_parse >= rangebegin ?
+ RExC_parse - rangebegin : 0;
vWARN4(RExC_parse,
"False [] range \"%*.*s\"",
- RExC_parse - rangebegin,
- RExC_parse - rangebegin,
+ w,
+ w,
rangebegin);
+ }
if (!SIZE_ONLY)
ANYOF_BITMAP_SET(ret, '-');
} else
IV i;
if (prevvalue < 256) {
- IV ceilvalue = value < 256 ? value : 255;
+ const IV ceilvalue = value < 256 ? value : 255;
#ifdef EBCDIC
/* In EBCDIC [\x89-\x91] should include
ANYOF_BITMAP_SET(ret, i);
}
if (value > 255 || UTF) {
- UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
- UV natvalue = NATIVE_TO_UNI(value);
+ const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
+ const UV natvalue = NATIVE_TO_UNI(value);
ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
if (prevnatvalue < natvalue) { /* what about > ? */
if (FOLD) {
U8 foldbuf[UTF8_MAXBYTES_CASE+1];
STRLEN foldlen;
- UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
+ const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
/* If folding and foldable and a single
* character, insert also the folded version
STATIC regnode * /* Location. */
S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
{
- register regnode *ret;
register regnode *ptr;
+ regnode * const ret = RExC_emit;
- ret = RExC_emit;
if (SIZE_ONLY) {
SIZE_ALIGN(RExC_size);
RExC_size += 1;
STATIC regnode * /* Location. */
S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
{
- register regnode *ret;
register regnode *ptr;
+ regnode * const ret = RExC_emit;
- ret = RExC_emit;
if (SIZE_ONLY) {
SIZE_ALIGN(RExC_size);
RExC_size += 2;
- reguni - emit (if appropriate) a Unicode character
*/
STATIC void
-S_reguni(pTHX_ RExC_state_t *pRExC_state, UV uv, char* s, STRLEN* lenp)
+S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s, STRLEN* lenp)
{
*lenp = SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
}
register regnode *src;
register regnode *dst;
register regnode *place;
- register int offset = regarglen[(U8)op];
+ const int offset = regarglen[(U8)op];
/* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, regnode *val)
{
register regnode *scan;
- register regnode *temp;
if (SIZE_ONLY)
return;
/* Find last node. */
scan = p;
for (;;) {
- temp = regnext(scan);
+ regnode * const temp = regnext(scan);
if (temp == NULL)
break;
scan = temp;
- regcurly - a little FSA that accepts {\d+,?\d*}
*/
STATIC I32
-S_regcurly(pTHX_ register char *s)
+S_regcurly(pTHX_ register const char *s)
{
if (*s++ != '{')
return FALSE;
return TRUE;
}
-
-#ifdef DEBUGGING
-
-STATIC regnode *
-S_dumpuntil(pTHX_ regnode *start, regnode *node, regnode *last, SV* sv, I32 l)
-{
- register U8 op = EXACT; /* Arbitrary non-END op. */
- register regnode *next;
-
- while (op != END && (!last || node < last)) {
- /* While that wasn't END last time... */
-
- NODE_ALIGN(node);
- op = OP(node);
- if (op == CLOSE)
- l--;
- next = regnext(node);
- /* Where, what. */
- if (OP(node) == OPTIMIZED)
- goto after_print;
- regprop(sv, node);
- PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
- (int)(2*l + 1), "", SvPVX(sv));
- if (next == NULL) /* Next ptr. */
- PerlIO_printf(Perl_debug_log, "(0)");
- else
- PerlIO_printf(Perl_debug_log, "(%"IVdf")", (IV)(next - start));
- (void)PerlIO_putc(Perl_debug_log, '\n');
- after_print:
- if (PL_regkind[(U8)op] == BRANCHJ) {
- register regnode *nnode = (OP(next) == LONGJMP
- ? regnext(next)
- : next);
- if (last && nnode > last)
- nnode = last;
- node = dumpuntil(start, NEXTOPER(NEXTOPER(node)), nnode, sv, l + 1);
- }
- else if (PL_regkind[(U8)op] == BRANCH) {
- node = dumpuntil(start, NEXTOPER(node), next, sv, l + 1);
- }
- else if ( op == CURLY) { /* `next' might be very big: optimizer */
- node = dumpuntil(start, NEXTOPER(node) + EXTRA_STEP_2ARGS,
- NEXTOPER(node) + EXTRA_STEP_2ARGS + 1, sv, l + 1);
- }
- else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
- node = dumpuntil(start, NEXTOPER(node) + EXTRA_STEP_2ARGS,
- next, sv, l + 1);
- }
- else if ( op == PLUS || op == STAR) {
- node = dumpuntil(start, NEXTOPER(node), NEXTOPER(node) + 1, sv, l + 1);
- }
- else if (op == ANYOF) {
- /* 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. */
- node += NODE_SZ_STR(node) - 1;
- node = NEXTOPER(node);
- }
- else {
- node = NEXTOPER(node);
- node += regarglen[(U8)op];
- }
- if (op == CURLYX || op == OPEN)
- l++;
- else if (op == WHILEM)
- l--;
- }
- return node;
-}
-
-#endif /* DEBUGGING */
-
/*
- regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
*/
Perl_regdump(pTHX_ regexp *r)
{
#ifdef DEBUGGING
- SV *sv = sv_newmortal();
+ SV * const sv = sv_newmortal();
(void)dumpuntil(r->program, r->program + 1, NULL, sv, 0);
/* Header fields of interest. */
if (r->anchored_substr)
PerlIO_printf(Perl_debug_log,
- "anchored `%s%.*s%s'%s at %"IVdf" ",
+ "anchored \"%s%.*s%s\"%s at %"IVdf" ",
PL_colors[0],
(int)(SvCUR(r->anchored_substr) - (SvTAIL(r->anchored_substr)!=0)),
- SvPVX(r->anchored_substr),
+ SvPVX_const(r->anchored_substr),
PL_colors[1],
SvTAIL(r->anchored_substr) ? "$" : "",
(IV)r->anchored_offset);
else if (r->anchored_utf8)
PerlIO_printf(Perl_debug_log,
- "anchored utf8 `%s%.*s%s'%s at %"IVdf" ",
+ "anchored utf8 \"%s%.*s%s\"%s at %"IVdf" ",
PL_colors[0],
(int)(SvCUR(r->anchored_utf8) - (SvTAIL(r->anchored_utf8)!=0)),
- SvPVX(r->anchored_utf8),
+ SvPVX_const(r->anchored_utf8),
PL_colors[1],
SvTAIL(r->anchored_utf8) ? "$" : "",
(IV)r->anchored_offset);
if (r->float_substr)
PerlIO_printf(Perl_debug_log,
- "floating `%s%.*s%s'%s at %"IVdf"..%"UVuf" ",
+ "floating \"%s%.*s%s\"%s at %"IVdf"..%"UVuf" ",
PL_colors[0],
(int)(SvCUR(r->float_substr) - (SvTAIL(r->float_substr)!=0)),
- SvPVX(r->float_substr),
+ SvPVX_const(r->float_substr),
PL_colors[1],
SvTAIL(r->float_substr) ? "$" : "",
(IV)r->float_min_offset, (UV)r->float_max_offset);
else if (r->float_utf8)
PerlIO_printf(Perl_debug_log,
- "floating utf8 `%s%.*s%s'%s at %"IVdf"..%"UVuf" ",
+ "floating utf8 \"%s%.*s%s\"%s at %"IVdf"..%"UVuf" ",
PL_colors[0],
(int)(SvCUR(r->float_utf8) - (SvTAIL(r->float_utf8)!=0)),
- SvPVX(r->float_utf8),
+ SvPVX_const(r->float_utf8),
PL_colors[1],
SvTAIL(r->float_utf8) ? "$" : "",
(IV)r->float_min_offset, (UV)r->float_max_offset);
if (r->regstclass) {
regprop(sv, r->regstclass);
- PerlIO_printf(Perl_debug_log, "stclass `%s' ", SvPVX(sv));
+ PerlIO_printf(Perl_debug_log, "stclass \"%s\" ", SvPVX_const(sv));
}
if (r->reganch & ROPT_ANCH) {
PerlIO_printf(Perl_debug_log, "anchored");
(UV)r->offsets[i*2]);
PerlIO_printf(Perl_debug_log, "\n");
}
+#else
+ PERL_UNUSED_ARG(r);
#endif /* DEBUGGING */
}
-#ifdef DEBUGGING
-
-STATIC void
-S_put_byte(pTHX_ SV *sv, int c)
-{
- if (isCNTRL(c) || c == 255 || !isPRINT(c))
- Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
- else if (c == '-' || c == ']' || c == '\\' || c == '^')
- Perl_sv_catpvf(aTHX_ sv, "\\%c", c);
- else
- Perl_sv_catpvf(aTHX_ sv, "%c", c);
-}
-
-#endif /* DEBUGGING */
-
/*
- regprop - printable representation of opcode
*/
k = PL_regkind[(U8)OP(o)];
if (k == EXACT) {
- SV *dsv = sv_2mortal(newSVpvn("", 0));
+ SV * const dsv = sv_2mortal(newSVpvs(""));
/* Using is_utf8_string() is a crude hack but it may
* be the best for now since we have no flag "this EXACTish
* node was UTF-8" --jhi */
- bool do_utf8 = is_utf8_string((U8*)STRING(o), STR_LEN(o));
- char *s = do_utf8 ?
+ const bool do_utf8 = is_utf8_string((U8*)STRING(o), STR_LEN(o));
+ const char * const s = do_utf8 ?
pv_uni_display(dsv, (U8*)STRING(o), STR_LEN(o), 60,
UNI_DISPLAY_REGEX) :
STRING(o);
Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
else if (k == ANYOF) {
int i, rangestart = -1;
- U8 flags = ANYOF_FLAGS(o);
- const char * const anyofs[] = { /* Should be synchronized with
- * ANYOF_ #xdefines in regcomp.h */
+ const U8 flags = ANYOF_FLAGS(o);
+
+ /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
+ static const char * const anyofs[] = {
"\\w",
"\\W",
"\\s",
};
if (flags & ANYOF_LOCALE)
- sv_catpv(sv, "{loc}");
+ sv_catpvs(sv, "{loc}");
if (flags & ANYOF_FOLD)
- sv_catpv(sv, "{i}");
+ sv_catpvs(sv, "{i}");
Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
if (flags & ANYOF_INVERT)
- sv_catpv(sv, "^");
+ sv_catpvs(sv, "^");
for (i = 0; i <= 256; i++) {
if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
if (rangestart == -1)
put_byte(sv, rangestart);
else {
put_byte(sv, rangestart);
- sv_catpv(sv, "-");
+ sv_catpvs(sv, "-");
put_byte(sv, i - 1);
}
rangestart = -1;
sv_catpv(sv, anyofs[i]);
if (flags & ANYOF_UNICODE)
- sv_catpv(sv, "{unicode}");
+ sv_catpvs(sv, "{unicode}");
else if (flags & ANYOF_UNICODE_ALL)
- sv_catpv(sv, "{unicode_all}");
+ sv_catpvs(sv, "{unicode_all}");
{
SV *lv;
- SV *sw = regclass_swash(o, FALSE, &lv, 0);
+ SV * const sw = regclass_swash(o, FALSE, &lv, 0);
if (lv) {
if (sw) {
U8 s[UTF8_MAXBYTES_CASE+1];
for (i = 0; i <= 256; i++) { /* just the first 256 */
- U8 *e = uvchr_to_utf8(s, i);
+ uvchr_to_utf8(s, i);
if (i < 256 && swash_fetch(sw, s, TRUE)) {
if (rangestart == -1)
rangestart = i;
} else if (rangestart != -1) {
- U8 *p;
-
if (i <= rangestart + 3)
for (; rangestart < i; rangestart++) {
- for(e = uvchr_to_utf8(s, rangestart), p = s; p < e; p++)
+ const U8 * const e = uvchr_to_utf8(s,rangestart);
+ U8 *p;
+ for(p = s; p < e; p++)
put_byte(sv, *p);
}
else {
- for (e = uvchr_to_utf8(s, rangestart), p = s; p < e; p++)
+ const U8 *e = uvchr_to_utf8(s,rangestart);
+ U8 *p;
+ for (p = s; p < e; p++)
+ put_byte(sv, *p);
+ sv_catpvs(sv, "-");
+ e = uvchr_to_utf8(s, i-1);
+ for (p = s; p < e; p++)
put_byte(sv, *p);
- sv_catpv(sv, "-");
- for (e = uvchr_to_utf8(s, i - 1), p = s; p < e; p++)
- put_byte(sv, *p);
}
rangestart = -1;
}
}
- sv_catpv(sv, "..."); /* et cetera */
+ sv_catpvs(sv, "..."); /* et cetera */
}
{
char *s = savesvpv(lv);
- char *origs = s;
+ char * const origs = s;
while(*s && *s != '\n') s++;
if (*s == '\n') {
- char *t = ++s;
+ const char * const t = ++s;
while (*s) {
if (*s == '\n')
}
else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
Perl_sv_catpvf(aTHX_ sv, "[-%d]", o->flags);
+#else
+ PERL_UNUSED_ARG(sv);
+ PERL_UNUSED_ARG(o);
#endif /* DEBUGGING */
}
Perl_re_intuit_string(pTHX_ regexp *prog)
{ /* Assume that RE_INTUIT is set */
DEBUG_r(
- { STRLEN n_a;
- char *s = SvPV(prog->check_substr
- ? prog->check_substr : prog->check_utf8, n_a);
+ {
+ const char * const s = SvPV_nolen_const(prog->check_substr
+ ? prog->check_substr : prog->check_utf8);
if (!PL_colorset) reginitcolors();
PerlIO_printf(Perl_debug_log,
- "%sUsing REx %ssubstr:%s `%s%.60s%s%s'\n",
+ "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
PL_colors[4],
prog->check_substr ? "" : "utf8 ",
PL_colors[5],PL_colors[0],
Perl_pregfree(pTHX_ struct regexp *r)
{
#ifdef DEBUGGING
- SV *dsv = PERL_DEBUG_PAD_ZERO(0);
+ SV * const dsv = PERL_DEBUG_PAD_ZERO(0);
#endif
if (!r || (--r->refcnt > 0))
return;
DEBUG_r({
- const char *s = (r->reganch & ROPT_UTF8)
+ const char * const s = (r->reganch & ROPT_UTF8)
? pv_uni_display(dsv, (U8*)r->precomp, r->prelen, 60, UNI_DISPLAY_REGEX)
: pv_display(dsv, r->precomp, r->prelen, 0, 60);
const int len = SvCUR(dsv);
len > 60 ? "..." : "");
});
- if (r->precomp)
- Safefree(r->precomp);
- if (r->offsets) /* 20010421 MJD */
- Safefree(r->offsets);
+ /* gcov results gave these as non-null 100% of the time, so there's no
+ optimisation in checking them before calling Safefree */
+ Safefree(r->precomp);
+ Safefree(r->offsets); /* 20010421 MJD */
if (RX_MATCH_COPIED(r))
Safefree(r->subbeg);
if (r->substrs) {
Perl_croak(aTHX_ "panic: pregfree comppad");
PAD_SAVE_LOCAL(old_comppad,
/* Watch out for global destruction's random ordering. */
- (SvTYPE(new_comppad) == SVt_PVAV) ?
- new_comppad : Null(PAD *)
+ (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
);
OP_REFCNT_LOCK;
refcnt = OpREFCNT_dec((OP_4tree*)r->data->data[n]);
#endif
msv = vmess(buf, &args);
va_end(args);
- message = SvPV(msv,l1);
+ message = SvPV_const(msv,l1);
if (l1 > 512)
l1 = 512;
Copy(message, buf, l1 , char);
void
Perl_save_re_context(pTHX)
{
- SAVEI32(PL_reg_flags); /* from regexec.c */
- SAVEPPTR(PL_bostr);
- SAVEPPTR(PL_reginput); /* String-input pointer. */
- SAVEPPTR(PL_regbol); /* Beginning of input, for ^ check. */
- SAVEPPTR(PL_regeol); /* End of input, for $ check. */
- SAVEVPTR(PL_regstartp); /* Pointer to startp array. */
- SAVEVPTR(PL_regendp); /* Ditto for endp. */
- SAVEVPTR(PL_reglastparen); /* Similarly for lastparen. */
- SAVEVPTR(PL_reglastcloseparen); /* Similarly for lastcloseparen. */
- SAVEPPTR(PL_regtill); /* How far we are required to go. */
- SAVEGENERICPV(PL_reg_start_tmp); /* from regexec.c */
+
+ struct re_save_state *state;
+
+ SAVEVPTR(PL_curcop);
+ SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
+
+ state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
+ PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
+ SSPUSHINT(SAVEt_RE_STATE);
+
+ state->re_state_reg_flags = PL_reg_flags;
+ state->re_state_bostr = PL_bostr;
+ state->re_state_reginput = PL_reginput;
+ state->re_state_regbol = PL_regbol;
+ state->re_state_regeol = PL_regeol;
+ state->re_state_regstartp = PL_regstartp;
+ state->re_state_regendp = PL_regendp;
+ state->re_state_reglastparen = PL_reglastparen;
+ state->re_state_reglastcloseparen = PL_reglastcloseparen;
+ state->re_state_regtill = PL_regtill;
+ state->re_state_reg_start_tmp = PL_reg_start_tmp;
+ state->re_state_reg_start_tmpl = PL_reg_start_tmpl;
+ state->re_state_reg_eval_set = PL_reg_eval_set;
+ state->re_state_regnarrate = PL_regnarrate;
+ state->re_state_regindent = PL_regindent;
+ state->re_state_reg_call_cc = PL_reg_call_cc;
+ state->re_state_reg_re = PL_reg_re;
+ state->re_state_reg_ganch = PL_reg_ganch;
+ state->re_state_reg_sv = PL_reg_sv;
+ state->re_state_reg_match_utf8 = PL_reg_match_utf8;
+ state->re_state_reg_magic = PL_reg_magic;
+ state->re_state_reg_oldpos = PL_reg_oldpos;
+ state->re_state_reg_oldcurpm = PL_reg_oldcurpm;
+ state->re_state_reg_curpm = PL_reg_curpm;
+ state->re_state_reg_oldsaved = PL_reg_oldsaved;
+ state->re_state_reg_oldsavedlen = PL_reg_oldsavedlen;
+ state->re_state_reg_maxiter = PL_reg_maxiter;
+ state->re_state_reg_leftiter = PL_reg_leftiter;
+ state->re_state_reg_poscache = PL_reg_poscache;
+ state->re_state_reg_poscache_size = PL_reg_poscache_size;
+ state->re_state_regsize = PL_regsize;
+ state->re_state_reg_starttry = PL_reg_starttry;
+
+ /* These variables have been eliminated from 5.10: */
+ state->re_state_regdata = PL_regdata;
+ state->re_state_regprogram = PL_regprogram;
+ state->re_state_regcc = PL_regcc;
+ state->re_state_regprecomp = PL_regprecomp;
+ state->re_state_regnpar = PL_regnpar;
+
PL_reg_start_tmp = 0;
- SAVEI32(PL_reg_start_tmpl); /* from regexec.c */
PL_reg_start_tmpl = 0;
- SAVEVPTR(PL_regdata);
- SAVEI32(PL_reg_eval_set); /* from regexec.c */
- SAVEI32(PL_regnarrate); /* from regexec.c */
- SAVEVPTR(PL_regprogram); /* from regexec.c */
- SAVEINT(PL_regindent); /* from regexec.c */
- SAVEVPTR(PL_regcc); /* from regexec.c */
- SAVEVPTR(PL_curcop);
- SAVEVPTR(PL_reg_call_cc); /* from regexec.c */
- SAVEVPTR(PL_reg_re); /* from regexec.c */
- SAVEPPTR(PL_reg_ganch); /* from regexec.c */
- SAVESPTR(PL_reg_sv); /* from regexec.c */
- SAVEBOOL(PL_reg_match_utf8); /* from regexec.c */
- SAVEVPTR(PL_reg_magic); /* from regexec.c */
- SAVEI32(PL_reg_oldpos); /* from regexec.c */
- SAVEVPTR(PL_reg_oldcurpm); /* from regexec.c */
- SAVEVPTR(PL_reg_curpm); /* from regexec.c */
- SAVEPPTR(PL_reg_oldsaved); /* old saved substr during match */
- PL_reg_oldsaved = Nullch;
- SAVEI32(PL_reg_oldsavedlen); /* old length of saved substr during match */
+ PL_reg_oldsaved = NULL;
PL_reg_oldsavedlen = 0;
- SAVEI32(PL_reg_maxiter); /* max wait until caching pos */
PL_reg_maxiter = 0;
- SAVEI32(PL_reg_leftiter); /* wait until caching pos */
PL_reg_leftiter = 0;
- SAVEGENERICPV(PL_reg_poscache); /* cache of pos of WHILEM */
- PL_reg_poscache = Nullch;
- SAVEI32(PL_reg_poscache_size); /* size of pos cache of WHILEM */
+ PL_reg_poscache = NULL;
PL_reg_poscache_size = 0;
- SAVEPPTR(PL_regprecomp); /* uncompiled string. */
- SAVEI32(PL_regnpar); /* () count. */
- SAVEI32(PL_regsize); /* from regexec.c */
- {
- /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
- U32 i;
- GV *mgv;
- REGEXP *rx;
- char digits[TYPE_CHARS(long)];
-
- if (PL_curpm && (rx = PM_GETRE(PL_curpm))) {
+ /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
+ if (PL_curpm) {
+ const REGEXP * const rx = PM_GETRE(PL_curpm);
+ if (rx) {
+ U32 i;
for (i = 1; i <= rx->nparens; i++) {
- sprintf(digits, "%lu", (long)i);
- if ((mgv = gv_fetchpv(digits, FALSE, SVt_PV)))
+ char digits[TYPE_CHARS(long)];
+#ifdef USE_SNPRINTF
+ const STRLEN len = snprintf(digits, sizeof(digits), "%lu", (long)i);
+#else
+ const STRLEN len = my_sprintf(digits, "%lu", (long)i);
+#endif /* #ifdef USE_SNPRINTF */
+ GV * const mgv = gv_fetchpvn_flags(digits, len, 0, SVt_PV);
+ if (mgv)
save_scalar(mgv);
}
}
}
-
-#ifdef DEBUGGING
- SAVEPPTR(PL_reg_starttry); /* from regexec.c */
-#endif
}
static void
ReREFCNT_dec((regexp *)r);
}
+#ifdef DEBUGGING
+
+STATIC void
+S_put_byte(pTHX_ SV *sv, int c)
+{
+ if (isCNTRL(c) || c == 255 || !isPRINT(c))
+ Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
+ else if (c == '-' || c == ']' || c == '\\' || c == '^')
+ Perl_sv_catpvf(aTHX_ sv, "\\%c", c);
+ else
+ Perl_sv_catpvf(aTHX_ sv, "%c", c);
+}
+
+
+STATIC regnode *
+S_dumpuntil(pTHX_ regnode *start, regnode *node, regnode *last, SV* sv, I32 l)
+{
+ register U8 op = EXACT; /* Arbitrary non-END op. */
+ register regnode *next;
+
+ while (op != END && (!last || node < last)) {
+ /* While that wasn't END last time... */
+
+ NODE_ALIGN(node);
+ op = OP(node);
+ if (op == CLOSE)
+ l--;
+ next = regnext(node);
+ /* Where, what. */
+ if (OP(node) == OPTIMIZED)
+ goto after_print;
+ regprop(sv, node);
+ PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
+ (int)(2*l + 1), "", SvPVX_const(sv));
+ if (next == NULL) /* Next ptr. */
+ PerlIO_printf(Perl_debug_log, "(0)");
+ else
+ PerlIO_printf(Perl_debug_log, "(%"IVdf")", (IV)(next - start));
+ (void)PerlIO_putc(Perl_debug_log, '\n');
+ after_print:
+ if (PL_regkind[(U8)op] == BRANCHJ) {
+ register regnode *nnode = (OP(next) == LONGJMP
+ ? regnext(next)
+ : next);
+ if (last && nnode > last)
+ nnode = last;
+ node = dumpuntil(start, NEXTOPER(NEXTOPER(node)), nnode, sv, l + 1);
+ }
+ else if (PL_regkind[(U8)op] == BRANCH) {
+ node = dumpuntil(start, NEXTOPER(node), next, sv, l + 1);
+ }
+ else if ( op == CURLY) { /* "next" might be very big: optimizer */
+ node = dumpuntil(start, NEXTOPER(node) + EXTRA_STEP_2ARGS,
+ NEXTOPER(node) + EXTRA_STEP_2ARGS + 1, sv, l + 1);
+ }
+ else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
+ node = dumpuntil(start, NEXTOPER(node) + EXTRA_STEP_2ARGS,
+ next, sv, l + 1);
+ }
+ else if ( op == PLUS || op == STAR) {
+ node = dumpuntil(start, NEXTOPER(node), NEXTOPER(node) + 1, sv, l + 1);
+ }
+ else if (op == ANYOF) {
+ /* 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. */
+ node += NODE_SZ_STR(node) - 1;
+ node = NEXTOPER(node);
+ }
+ else {
+ node = NEXTOPER(node);
+ node += regarglen[(U8)op];
+ }
+ if (op == CURLYX || op == OPEN)
+ l++;
+ else if (op == WHILEM)
+ l--;
+ }
+ return node;
+}
+
+#endif /* DEBUGGING */
+
/*
* Local variables:
* c-indentation-style: bsd
* indent-tabs-mode: t
* End:
*
- * vim: shiftwidth=4:
-*/
+ * ex: set ts=8 sts=4 sw=4 noet:
+ */
https://perl5.git.perl.org / perl5.git / blobdiff