# define PL_nextval (PL_parser->nextval)
#endif
-static const char ident_too_long[] = "Identifier too long";
+static const char* const ident_too_long = "Identifier too long";
#ifdef PERL_MAD
# define CURMAD(slot,sv) if (PL_madskills) { curmad(slot,sv); sv = 0; }
void
Perl_parser_free(pTHX_ const yy_parser *parser)
{
-#ifdef PERL_MAD
- I32 nexttoke = parser->lasttoke;
-#else
- I32 nexttoke = parser->nexttoke;
-#endif
-
PERL_ARGS_ASSERT_PARSER_FREE;
PL_curcop = parser->saved_curcop;
SvREFCNT_dec(parser->rsfp_filters);
SvREFCNT_dec(parser->lex_stuff);
SvREFCNT_dec(parser->sublex_info.repl);
+
+ Safefree(parser->lex_brackstack);
+ Safefree(parser->lex_casestack);
+ Safefree(parser->lex_shared);
+ PL_parser = parser->old_parser;
+ Safefree(parser);
+}
+
+void
+Perl_parser_free_nexttoke_ops(pTHX_ yy_parser *parser, OPSLAB *slab)
+{
+#ifdef PERL_MAD
+ I32 nexttoke = parser->lasttoke;
+#else
+ I32 nexttoke = parser->nexttoke;
+#endif
+ PERL_ARGS_ASSERT_PARSER_FREE_NEXTTOKE_OPS;
while (nexttoke--) {
#ifdef PERL_MAD
if (S_is_opval_token(parser->nexttoke[nexttoke].next_type
- & 0xffff))
- op_free(parser->nexttoke[nexttoke].next_val.opval);
+ & 0xffff)
+ && parser->nexttoke[nexttoke].next_val.opval
+ && parser->nexttoke[nexttoke].next_val.opval->op_slabbed
+ && OpSLAB(parser->nexttoke[nexttoke].next_val.opval) == slab) {
+ op_free(parser->nexttoke[nexttoke].next_val.opval);
+ parser->nexttoke[nexttoke].next_val.opval = NULL;
+ }
#else
- if (S_is_opval_token(parser->nexttype[nexttoke] & 0xffff))
+ if (S_is_opval_token(parser->nexttype[nexttoke] & 0xffff)
+ && parser->nextval[nexttoke].opval
+ && parser->nextval[nexttoke].opval->op_slabbed
+ && OpSLAB(parser->nextval[nexttoke].opval) == slab) {
op_free(parser->nextval[nexttoke].opval);
+ parser->nextval[nexttoke].opval = NULL;
+ }
#endif
}
-
- Safefree(parser->lex_brackstack);
- Safefree(parser->lex_casestack);
- Safefree(parser->lex_shared);
- PL_parser = parser->old_parser;
- Safefree(parser);
}
if (flags & LEX_STUFF_UTF8) {
goto plain_copy;
} else {
- STRLEN highhalf = 0;
+ STRLEN highhalf = 0; /* Count of variants */
const char *p, *e = pv+len;
- for (p = pv; p != e; p++)
- highhalf += !!(((U8)*p) & 0x80);
+ for (p = pv; p != e; p++) {
+ if (! UTF8_IS_INVARIANT(*p)) {
+ highhalf++;
+ }
+ }
if (!highhalf)
goto plain_copy;
lex_grow_linestr(SvCUR(PL_parser->linestr)+1+len+highhalf);
PL_parser->bufend += len+highhalf;
for (p = pv; p != e; p++) {
U8 c = (U8)*p;
- if (c & 0x80) {
- *bufptr++ = (char)(0xc0 | (c >> 6));
- *bufptr++ = (char)(0x80 | (c & 0x3f));
+ if (! UTF8_IS_INVARIANT(c)) {
+ *bufptr++ = UTF8_TWO_BYTE_HI(c);
+ *bufptr++ = UTF8_TWO_BYTE_LO(c);
} else {
*bufptr++ = (char)c;
}
const char *p, *e = pv+len;
for (p = pv; p != e; p++) {
U8 c = (U8)*p;
- if (c >= 0xc4) {
+ if (UTF8_IS_ABOVE_LATIN1(c)) {
Perl_croak(aTHX_ "Lexing code attempted to stuff "
"non-Latin-1 character into Latin-1 input");
- } else if (c >= 0xc2 && p+1 != e &&
- (((U8)p[1]) & 0xc0) == 0x80) {
+ } else if (UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(p, e)) {
p++;
highhalf++;
- } else if (c >= 0x80) {
+ } else if (! UTF8_IS_INVARIANT(c)) {
/* malformed UTF-8 */
ENTER;
SAVESPTR(PL_warnhook);
SvCUR_set(PL_parser->linestr,
SvCUR(PL_parser->linestr) + len-highhalf);
PL_parser->bufend += len-highhalf;
- for (p = pv; p != e; p++) {
- U8 c = (U8)*p;
- if (c & 0x80) {
- *bufptr++ = (char)(((c & 0x3) << 6) | (p[1] & 0x3f));
- p++;
- } else {
- *bufptr++ = (char)c;
+ p = pv;
+ while (p < e) {
+ if (UTF8_IS_INVARIANT(*p)) {
+ *bufptr++ = *p;
+ p++;
}
+ else {
+ assert(p < e -1 );
+ *bufptr++ = TWO_BYTE_UTF8_TO_UNI(*p, *(p+1));
+ p += 2;
+ }
}
} else {
- plain_copy:
+ plain_copy:
lex_grow_linestr(SvCUR(PL_parser->linestr)+1+len);
bufptr = PL_parser->bufptr;
Move(bufptr, bufptr+len, PL_parser->bufend+1-bufptr, char);
bufend = PL_parser->bufend;
}
head = (U8)*s;
- if (!(head & 0x80))
+ if (UTF8_IS_INVARIANT(head))
return head;
- if (head & 0x40) {
- len = PL_utf8skip[head];
+ if (UTF8_IS_START(head)) {
+ len = UTF8SKIP(&head);
while ((STRLEN)(bufend-s) < len) {
if (!lex_next_chunk(flags | LEX_KEEP_PREVIOUS))
break;
/* skip space before PL_thistoken */
STATIC char *
-S_skipspace0(pTHX_ register char *s)
+S_skipspace0(pTHX_ char *s)
{
PERL_ARGS_ASSERT_SKIPSPACE0;
/* skip space after PL_thistoken */
STATIC char *
-S_skipspace1(pTHX_ register char *s)
+S_skipspace1(pTHX_ char *s)
{
const char *start = s;
I32 startoff = start - SvPVX(PL_linestr);
}
STATIC char *
-S_skipspace2(pTHX_ register char *s, SV **svp)
+S_skipspace2(pTHX_ char *s, SV **svp)
{
char *start;
const I32 bufptroff = PL_bufptr - SvPVX(PL_linestr);
*/
STATIC char *
-S_skipspace(pTHX_ register char *s)
+S_skipspace(pTHX_ char *s)
{
#ifdef PERL_MAD
char *start = s;
tokereport(type, &NEXTVAL_NEXTTOKE);
}
#endif
- /* Don’t let opslab_force_free snatch it */
- if (S_is_opval_token(type & 0xffff) && NEXTVAL_NEXTTOKE.opval) {
- assert(!NEXTVAL_NEXTTOKE.opval->op_savefree);
- NEXTVAL_NEXTTOKE.opval->op_savefree = 1;
- }
#ifdef PERL_MAD
if (PL_curforce < 0)
start_force(PL_lasttoke);
*/
STATIC char *
-S_force_word(pTHX_ register char *start, int token, int check_keyword, int allow_pack, int allow_initial_tick)
+S_force_word(pTHX_ char *start, int token, int check_keyword, int allow_pack, int allow_initial_tick)
{
dVAR;
char *s;
*/
STATIC void
-S_force_ident(pTHX_ register const char *s, int kind)
+S_force_ident(pTHX_ const char *s, int kind)
{
dVAR;
PERL_ARGS_ASSERT_FORCE_IDENT;
- if (*s) {
- const STRLEN len = strlen(s);
+ if (s[0]) {
+ const STRLEN len = s[1] ? strlen(s) : 1; /* s = "\"" see yylex */
OP* const o = (OP*)newSVOP(OP_CONST, 0, newSVpvn_flags(s, len,
UTF ? SVf_UTF8 : 0));
start_force(PL_curforce);
PERL_STATIC_INLINE SV*
S_get_and_check_backslash_N_name(pTHX_ const char* s, const char* const e)
{
- /* Get the value for NAME */
- STRLEN len;
- const char *str;
- SV* res = newSVpvn(s, e - s);
+ /* <s> points to first character of interior of \N{}, <e> to one beyond the
+ * interior, hence to the "}". Finds what the name resolves to, returning
+ * an SV* containing it; NULL if no valid one found */
+
+ SV* res = newSVpvn_flags(s, e - s, UTF ? SVf_UTF8 : 0);
+
+ HV * table;
+ SV **cvp;
+ SV *cv;
+ SV *rv;
+ HV *stash;
+ const U8* first_bad_char_loc;
+ const char* backslash_ptr = s - 3; /* Points to the <\> of \N{... */
PERL_ARGS_ASSERT_GET_AND_CHECK_BACKSLASH_N_NAME;
- res = new_constant( NULL, 0, "charnames",
- /* includes all of: \N{...} */
- res, NULL, s - 3, e - s + 4 );
+ if (UTF && ! is_utf8_string_loc((U8 *) backslash_ptr,
+ e - backslash_ptr,
+ &first_bad_char_loc))
+ {
+ /* If warnings are on, this will print a more detailed analysis of what
+ * is wrong than the error message below */
+ utf8n_to_uvuni(first_bad_char_loc,
+ e - ((char *) first_bad_char_loc),
+ NULL, 0);
+
+ /* We deliberately don't try to print the malformed character, which
+ * might not print very well; it also may be just the first of many
+ * malformations, so don't print what comes after it */
+ yyerror(Perl_form(aTHX_
+ "Malformed UTF-8 character immediately after '%.*s'",
+ (int) (first_bad_char_loc - (U8 *) backslash_ptr), backslash_ptr));
+ return NULL;
+ }
+
+ res = new_constant( NULL, 0, "charnames", res, NULL, backslash_ptr,
+ /* include the <}> */
+ e - backslash_ptr + 1);
if (! SvPOK(res)) {
+ SvREFCNT_dec_NN(res);
return NULL;
}
- /* Most likely res will be in utf8 already since the standard charnames
- * uses pack U, but a custom translator can leave it otherwise, so make
- * sure. XXX This can be revisited to not have charnames use utf8 for
- * characters that don't need it when regexes don't have to be in utf8 for
- * Unicode semantics. If doing so, remember EBCDIC */
- sv_utf8_upgrade(res);
-
- /* Don't accept malformed input */
- str = SvPV_const(res, len);
- if (! is_utf8_string((U8 *) str, len)) {
- yyerror("Malformed UTF-8 returned by \\N");
- return NULL;
+ /* See if the charnames handler is the Perl core's, and if so, we can skip
+ * the validation needed for a user-supplied one, as Perl's does its own
+ * validation. */
+ table = GvHV(PL_hintgv); /* ^H */
+ cvp = hv_fetchs(table, "charnames", FALSE);
+ if (cvp && (cv = *cvp) && SvROK(cv) && ((rv = SvRV(cv)) != NULL)
+ && SvTYPE(rv) == SVt_PVCV && ((stash = CvSTASH(rv)) != NULL))
+ {
+ const char * const name = HvNAME(stash);
+ if strEQ(name, "_charnames") {
+ return res;
+ }
}
- /* Deprecate non-approved name syntax */
- if (ckWARN_d(WARN_DEPRECATED)) {
- bool problematic = FALSE;
- const char* i = s;
+ /* Here, it isn't Perl's charname handler. We can't rely on a
+ * user-supplied handler to validate the input name. For non-ut8 input,
+ * look to see that the first character is legal. Then loop through the
+ * rest checking that each is a continuation */
- /* For non-ut8 input, look to see that the first character is an alpha,
- * then loop through the rest checking that each is a continuation */
- if (! UTF) {
- if (! isALPHAU(*i)) problematic = TRUE;
- else for (i = s + 1; i < e; i++) {
- if (isCHARNAME_CONT(*i)) continue;
- problematic = TRUE;
- break;
+ /* This code needs to be sync'ed with a regex in _charnames.pm which does
+ * the same thing */
+
+ if (! UTF) {
+ if (! isALPHAU(*s)) {
+ goto bad_charname;
+ }
+ s++;
+ while (s < e) {
+ if (! isCHARNAME_CONT(*s)) {
+ goto bad_charname;
+ }
+ s++;
+ }
+ }
+ else {
+ /* Similarly for utf8. For invariants can check directly; for other
+ * Latin1, can calculate their code point and check; otherwise use a
+ * swash */
+ if (UTF8_IS_INVARIANT(*s)) {
+ if (! isALPHAU(*s)) {
+ goto bad_charname;
+ }
+ s++;
+ } else if (UTF8_IS_DOWNGRADEABLE_START(*s)) {
+ if (! isALPHAU(UNI_TO_NATIVE(TWO_BYTE_UTF8_TO_UNI(*s, *(s+1))))) {
+ goto bad_charname;
}
+ s += 2;
}
else {
- /* Similarly for utf8. For invariants can check directly. We
- * accept anything above the latin1 range because it is immaterial
- * to Perl if it is correct or not, and is expensive to check. But
- * it is fairly easy in the latin1 range to convert the variants
- * into a single character and check those */
- if (UTF8_IS_INVARIANT(*i)) {
- if (! isALPHAU(*i)) problematic = TRUE;
- } else if (UTF8_IS_DOWNGRADEABLE_START(*i)) {
- if (! isALPHAU(UNI_TO_NATIVE(TWO_BYTE_UTF8_TO_UNI(*i,
- *(i+1)))))
- {
- problematic = TRUE;
+ if (! PL_utf8_charname_begin) {
+ U8 flags = _CORE_SWASH_INIT_ACCEPT_INVLIST;
+ PL_utf8_charname_begin = _core_swash_init("utf8",
+ "_Perl_Charname_Begin",
+ &PL_sv_undef,
+ 1, 0, NULL, &flags);
+ }
+ if (! swash_fetch(PL_utf8_charname_begin, (U8 *) s, TRUE)) {
+ goto bad_charname;
+ }
+ s += UTF8SKIP(s);
+ }
+
+ while (s < e) {
+ if (UTF8_IS_INVARIANT(*s)) {
+ if (! isCHARNAME_CONT(*s)) {
+ goto bad_charname;
}
+ s++;
}
- if (! problematic) for (i = s + UTF8SKIP(s);
- i < e;
- i+= UTF8SKIP(i))
- {
- if (UTF8_IS_INVARIANT(*i)) {
- if (isCHARNAME_CONT(*i)) continue;
- } else if (! UTF8_IS_DOWNGRADEABLE_START(*i)) {
- continue;
- } else if (isCHARNAME_CONT(
- UNI_TO_NATIVE(
- TWO_BYTE_UTF8_TO_UNI(*i, *(i+1)))))
+ else if (UTF8_IS_DOWNGRADEABLE_START(*s)) {
+ if (! isCHARNAME_CONT(UNI_TO_NATIVE(TWO_BYTE_UTF8_TO_UNI(*s,
+ *(s+1)))))
{
- continue;
+ goto bad_charname;
+ }
+ s += 2;
+ }
+ else {
+ if (! PL_utf8_charname_continue) {
+ U8 flags = _CORE_SWASH_INIT_ACCEPT_INVLIST;
+ PL_utf8_charname_continue = _core_swash_init("utf8",
+ "_Perl_Charname_Continue",
+ &PL_sv_undef,
+ 1, 0, NULL, &flags);
+ }
+ if (! swash_fetch(PL_utf8_charname_continue, (U8 *) s, TRUE)) {
+ goto bad_charname;
}
- problematic = TRUE;
- break;
+ s += UTF8SKIP(s);
}
}
- if (problematic) {
- /* The e-i passed to the final %.*s makes sure that should the
- * trailing NUL be missing that this print won't run off the end of
- * the string */
- Perl_warner(aTHX_ packWARN(WARN_DEPRECATED),
- "Deprecated character in \\N{...}; marked by <-- HERE in \\N{%.*s<-- HERE %.*s",
- (int)(i - s + 1), s, (int)(e - i), i + 1);
+ }
+
+ if (SvUTF8(res)) { /* Don't accept malformed input */
+ const U8* first_bad_char_loc;
+ STRLEN len;
+ const char* const str = SvPV_const(res, len);
+ if (! is_utf8_string_loc((U8 *) str, len, &first_bad_char_loc)) {
+ /* If warnings are on, this will print a more detailed analysis of
+ * what is wrong than the error message below */
+ utf8n_to_uvuni(first_bad_char_loc,
+ (char *) first_bad_char_loc - str,
+ NULL, 0);
+
+ /* We deliberately don't try to print the malformed character,
+ * which might not print very well; it also may be just the first
+ * of many malformations, so don't print what comes after it */
+ yyerror_pv(
+ Perl_form(aTHX_
+ "Malformed UTF-8 returned by %.*s immediately after '%.*s'",
+ (int) (e - backslash_ptr + 1), backslash_ptr,
+ (int) ((char *) first_bad_char_loc - str), str
+ ),
+ SVf_UTF8);
+ return NULL;
}
}
return res;
+
+ bad_charname: {
+ int bad_char_size = ((UTF) ? UTF8SKIP(s) : 1);
+
+ /* The final %.*s makes sure that should the trailing NUL be missing
+ * that this print won't run off the end of the string */
+ yyerror_pv(
+ Perl_form(aTHX_
+ "Invalid character in \\N{...}; marked by <-- HERE in %.*s<-- HERE %.*s",
+ (int)(s - backslash_ptr + bad_char_size), backslash_ptr,
+ (int)(e - s + bad_char_size), s + bad_char_size
+ ),
+ UTF ? SVf_UTF8 : 0);
+ return NULL;
+ }
}
/*
this_utf8 = PL_sublex_info.sub_op->op_private & (PL_lex_repl ? OPpTRANS_FROM_UTF : OPpTRANS_TO_UTF);
}
+ /* Protect sv from errors and fatal warnings. */
+ ENTER_with_name("scan_const");
+ SAVEFREESV(sv);
while (s < send || dorange) {
#ifdef EBCDIC
&& !native_range
#endif
- ) {
+ ) {
char * const c = (char*)utf8_hop((U8*)d, -1);
char *e = d++;
while (e-- > c)
#endif
if (min > max) {
- SvREFCNT_dec(sv);
Perl_croak(aTHX_
"Invalid range \"%c-%c\" in transliteration operator",
(char)min, (char)max);
/* range begins (ignore - as first or last char) */
else if (*s == '-' && s+1 < send && s != start) {
if (didrange) {
- SvREFCNT_dec(sv);
Perl_croak(aTHX_ "Ambiguous range in transliteration operator");
}
if (has_utf8
/* FALL THROUGH */
default:
{
- if ((isALNUMC(*s)))
+ if ((isALPHANUMERIC(*s)))
Perl_ck_warner(aTHX_ packWARN(WARN_MISC),
"Unrecognized escape \\%c passed through",
*s);
/* Here it looks like a named character */
- if (PL_lex_inpat) {
-
- /* XXX This block is temporary code. \N{} implies that the
- * pattern is to have Unicode semantics, and therefore
- * currently has to be encoded in utf8. By putting it in
- * utf8 now, we save a whole pass in the regular expression
- * compiler. Once that code is changed so Unicode
- * semantics doesn't necessarily have to be in utf8, this
- * block should be removed. However, the code that parses
- * the output of this would have to be changed to not
- * necessarily expect utf8 */
- if (!has_utf8) {
- SvCUR_set(sv, d - SvPVX_const(sv));
- SvPOK_on(sv);
- *d = '\0';
- /* See Note on sizing above. */
- sv_utf8_upgrade_flags_grow(sv,
- SV_GMAGIC|SV_FORCE_UTF8_UPGRADE,
- /* 5 = '\N{' + cur char + NUL */
- (STRLEN)(send - s) + 5);
- d = SvPVX(sv) + SvCUR(sv);
- has_utf8 = TRUE;
- }
- }
-
if (*s == 'U' && s[1] == '+') { /* \N{U+...} */
I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
| PERL_SCAN_DISALLOW_PREFIX;
* returned by charnames */
const char *str_end = str + len;
- STRLEN char_length; /* cur char's byte length */
- STRLEN output_length; /* and the number of bytes
- after this is translated
- into hex digits */
const STRLEN off = d - SvPVX_const(sv);
- /* 2 hex per byte; 2 chars for '\N'; 2 chars for
- * max('U+', '.'); and 1 for NUL */
- char hex_string[2 * UTF8_MAXBYTES + 5];
-
- /* Get the first character of the result. */
- U32 uv = utf8n_to_uvuni((U8 *) str,
- len,
- &char_length,
- UTF8_ALLOW_ANYUV);
-
- /* The call to is_utf8_string() above hopefully
- * guarantees that there won't be an error. But
- * it's easy here to make sure. The function just
- * above warns and returns 0 if invalid utf8, but
- * it can also return 0 if the input is validly a
- * NUL. Disambiguate */
- if (uv == 0 && NATIVE_TO_ASCII(*str) != '\0') {
- uv = UNICODE_REPLACEMENT;
- }
-
- /* Convert first code point to hex, including the
- * boiler plate before it. For all these, we
- * convert to native format so that downstream code
- * can continue to assume the input is native */
- output_length =
- my_snprintf(hex_string, sizeof(hex_string),
- "\\N{U+%X",
- (unsigned int) UNI_TO_NATIVE(uv));
-
- /* Make sure there is enough space to hold it */
- d = off + SvGROW(sv, off
- + output_length
- + (STRLEN)(send - e)
- + 2); /* '}' + NUL */
- /* And output it */
- Copy(hex_string, d, output_length, char);
- d += output_length;
-
- /* For each subsequent character, append dot and
- * its ordinal in hex */
- while ((str += char_length) < str_end) {
- const STRLEN off = d - SvPVX_const(sv);
- U32 uv = utf8n_to_uvuni((U8 *) str,
- str_end - str,
- &char_length,
- UTF8_ALLOW_ANYUV);
- if (uv == 0 && NATIVE_TO_ASCII(*str) != '\0') {
- uv = UNICODE_REPLACEMENT;
- }
-
- output_length =
- my_snprintf(hex_string, sizeof(hex_string),
- ".%X",
- (unsigned int) UNI_TO_NATIVE(uv));
-
- d = off + SvGROW(sv, off
- + output_length
- + (STRLEN)(send - e)
- + 2); /* '}' + NUL */
- Copy(hex_string, d, output_length, char);
- d += output_length;
+ if (! SvUTF8(res)) {
+ /* For the non-UTF-8 case, we can determine the
+ * exact length needed without having to parse
+ * through the string. Each character takes up
+ * 2 hex digits plus either a trailing dot or
+ * the "}" */
+ d = off + SvGROW(sv, off
+ + 3 * len
+ + 6 /* For the "\N{U+", and
+ trailing NUL */
+ + (STRLEN)(send - e));
+ Copy("\\N{U+", d, 5, char);
+ d += 5;
+ while (str < str_end) {
+ char hex_string[4];
+ my_snprintf(hex_string, sizeof(hex_string),
+ "%02X.", (U8) *str);
+ Copy(hex_string, d, 3, char);
+ d += 3;
+ str++;
+ }
+ d--; /* We will overwrite below the final
+ dot with a right brace */
+ }
+ else {
+ STRLEN char_length; /* cur char's byte length */
+
+ /* and the number of bytes after this is
+ * translated into hex digits */
+ STRLEN output_length;
+
+ /* 2 hex per byte; 2 chars for '\N'; 2 chars
+ * for max('U+', '.'); and 1 for NUL */
+ char hex_string[2 * UTF8_MAXBYTES + 5];
+
+ /* Get the first character of the result. */
+ U32 uv = utf8n_to_uvuni((U8 *) str,
+ len,
+ &char_length,
+ UTF8_ALLOW_ANYUV);
+ /* Convert first code point to hex, including
+ * the boiler plate before it. For all these,
+ * we convert to native format so that
+ * downstream code can continue to assume the
+ * input is native */
+ output_length =
+ my_snprintf(hex_string, sizeof(hex_string),
+ "\\N{U+%X",
+ (unsigned int) UNI_TO_NATIVE(uv));
+
+ /* Make sure there is enough space to hold it */
+ d = off + SvGROW(sv, off
+ + output_length
+ + (STRLEN)(send - e)
+ + 2); /* '}' + NUL */
+ /* And output it */
+ Copy(hex_string, d, output_length, char);
+ d += output_length;
+
+ /* For each subsequent character, append dot and
+ * its ordinal in hex */
+ while ((str += char_length) < str_end) {
+ const STRLEN off = d - SvPVX_const(sv);
+ U32 uv = utf8n_to_uvuni((U8 *) str,
+ str_end - str,
+ &char_length,
+ UTF8_ALLOW_ANYUV);
+ output_length =
+ my_snprintf(hex_string,
+ sizeof(hex_string),
+ ".%X",
+ (unsigned int) UNI_TO_NATIVE(uv));
+
+ d = off + SvGROW(sv, off
+ + output_length
+ + (STRLEN)(send - e)
+ + 2); /* '}' + NUL */
+ Copy(hex_string, d, output_length, char);
+ d += output_length;
+ }
}
*d++ = '}'; /* Done. Add the trailing brace */
/* return the substring (via pl_yylval) only if we parsed anything */
if (s > PL_bufptr) {
+ SvREFCNT_inc_simple_void_NN(sv);
if ( PL_hints & ( PL_lex_inpat ? HINT_NEW_RE : HINT_NEW_STRING ) ) {
const char *const key = PL_lex_inpat ? "qr" : "q";
const STRLEN keylen = PL_lex_inpat ? 2 : 1;
type, typelen);
}
pl_yylval.opval = (OP*)newSVOP(OP_CONST, 0, sv);
- } else
- SvREFCNT_dec(sv);
+ }
+ LEAVE_with_name("scan_const");
return s;
}
/* This is the one truly awful dwimmer necessary to conflate C and sed. */
STATIC int
-S_intuit_more(pTHX_ register char *s)
+S_intuit_more(pTHX_ char *s)
{
dVAR;
}
STATIC char *
-S_filter_gets(pTHX_ register SV *sv, STRLEN append)
+S_filter_gets(pTHX_ SV *sv, STRLEN append)
{
dVAR;
PL_lex_allbrackets--;
next_type &= 0xffff;
}
- if (S_is_opval_token(next_type) && pl_yylval.opval)
- pl_yylval.opval->op_savefree = 0; /* release */
return REPORT(next_type == 'p' ? pending_ident() : next_type);
}
#ifdef PERL_MAD
while (PL_bufptr != PL_bufend &&
PL_bufptr[0] == '\\' && PL_bufptr[1] == 'E') {
- if (!PL_thiswhite)
+ if (PL_madskills) {
+ if (!PL_thiswhite)
PL_thiswhite = newSVpvs("");
- sv_catpvn(PL_thiswhite, PL_bufptr, 2);
+ sv_catpvn(PL_thiswhite, PL_bufptr, 2);
+ }
PL_bufptr += 2;
}
#else
s = PL_bufptr + 1;
if (s[1] == '\\' && s[2] == 'E') {
#ifdef PERL_MAD
- if (!PL_thiswhite)
+ if (PL_madskills) {
+ if (!PL_thiswhite)
PL_thiswhite = newSVpvs("");
- sv_catpvn(PL_thiswhite, PL_bufptr, 4);
+ sv_catpvn(PL_thiswhite, PL_bufptr, 4);
+ }
#endif
PL_bufptr = s + 3;
PL_lex_state = LEX_INTERPCONCAT;
case ' ': case '\t': case '\f': case 013:
#ifdef PERL_MAD
PL_realtokenstart = -1;
- if (!PL_thiswhite)
+ if (PL_madskills) {
+ if (!PL_thiswhite)
PL_thiswhite = newSVpvs("");
- sv_catpvn(PL_thiswhite, s, 1);
+ sv_catpvn(PL_thiswhite, s, 1);
+ }
#endif
s++;
goto retry;
force_next(formbrack ? '.' : '}');
if (formbrack) LEAVE;
#ifdef PERL_MAD
- if (!PL_thistoken)
+ if (PL_madskills && !PL_thistoken)
PL_thistoken = newSVpvs("");
#endif
if (formbrack == 2) { /* means . where arguments were expected */
"Experimental \"%s\" subs not enabled",
tmp == KEY_my ? "my" :
tmp == KEY_state ? "state" : "our");
+ Perl_ck_warner_d(aTHX_
+ packWARN(WARN_EXPERIMENTAL__LEXICAL_SUBS),
+ "The lexical_subs feature is experimental");
goto really_sub;
}
PL_in_my_stash = find_in_my_stash(PL_tokenbuf, len);
for (d = s; isALNUM_lazy_if(d,UTF);) {
d += UTF ? UTF8SKIP(d) : 1;
if (UTF) {
- while (UTF8_IS_CONTINUED(*d) && is_utf8_mark((U8*)d)) {
+ while (UTF8_IS_CONTINUED(*d) && _is_utf8_mark((U8*)d)) {
d += UTF ? UTF8SKIP(d) : 1;
}
}
SV *tmpwhite = 0;
char *tstart = SvPVX(PL_linestr) + PL_realtokenstart;
- SV *subtoken = newSVpvn_flags(tstart, s - tstart, SvUTF8(PL_linestr));
+ SV *subtoken = PL_madskills
+ ? newSVpvn_flags(tstart, s - tstart, SvUTF8(PL_linestr))
+ : NULL;
PL_thistoken = 0;
d = s;
}
}
-/* Either returns sv, or mortalizes sv and returns a new SV*.
+/* Either returns sv, or mortalizes/frees sv and returns a new SV*.
Best used as sv=new_constant(..., sv, ...).
If s, pv are NULL, calls subroutine with one argument,
- and type is used with error messages only. */
+ and <type> is used with error messages only.
+ <type> is assumed to be well formed UTF-8 */
STATIC SV *
S_new_constant(pTHX_ const char *s, STRLEN len, const char *key, STRLEN keylen,
dVAR; dSP;
HV * table = GvHV(PL_hintgv); /* ^H */
SV *res;
+ SV *errsv = NULL;
SV **cvp;
SV *cv, *typesv;
const char *why1 = "", *why2 = "", *why3 = "";
PERL_ARGS_ASSERT_NEW_CONSTANT;
+ /* We assume that this is true: */
+ if (*key == 'c') { assert (strEQ(key, "charnames")); }
+ assert(type || s);
/* charnames doesn't work well if there have been errors found */
- if (PL_error_count > 0 && strEQ(key,"charnames"))
+ if (PL_error_count > 0 && *key == 'c')
+ {
+ SvREFCNT_dec_NN(sv);
return &PL_sv_undef;
+ }
+ sv_2mortal(sv); /* Parent created it permanently */
if (!table
|| ! (PL_hints & HINT_LOCALIZE_HH)
|| ! (cvp = hv_fetch(table, key, keylen, FALSE))
|| ! SvOK(*cvp))
{
- SV *msg;
+ char *msg;
/* Here haven't found what we're looking for. If it is charnames,
* perhaps it needs to be loaded. Try doing that before giving up */
- if (strEQ(key,"charnames")) {
+ if (*key == 'c') {
Perl_load_module(aTHX_
0,
newSVpvs("_charnames"),
}
}
if (!table || !(PL_hints & HINT_LOCALIZE_HH)) {
- msg = Perl_newSVpvf(aTHX_
- "Constant(%s) unknown", (type ? type: "undef"));
+ msg = Perl_form(aTHX_
+ "Constant(%.*s) unknown",
+ (int)(type ? typelen : len),
+ (type ? type: s));
}
else {
why1 = "$^H{";
why2 = key;
why3 = "} is not defined";
report:
- if (strEQ(key,"charnames")) {
- msg = Perl_newSVpvf(aTHX_
- /* The +3 is for '\N{'; -4 for that, plus '}' */
- "Unknown charname '%.*s'", (int)typelen - 4, type + 3);
+ if (*key == 'c') {
+ msg = Perl_form(aTHX_
+ /* The +3 is for '\N{'; -4 for that, plus '}' */
+ "Unknown charname '%.*s'", (int)typelen - 4, type + 3
+ );
}
else {
- msg = Perl_newSVpvf(aTHX_ "Constant(%s): %s%s%s",
- (type ? type: "undef"), why1, why2, why3);
+ msg = Perl_form(aTHX_ "Constant(%.*s): %s%s%s",
+ (int)(type ? typelen : len),
+ (type ? type: s), why1, why2, why3);
}
}
- yyerror(SvPVX_const(msg));
- SvREFCNT_dec(msg);
- return sv;
+ yyerror_pv(msg, UTF ? SVf_UTF8 : 0);
+ return SvREFCNT_inc_simple_NN(sv);
}
now_ok:
- sv_2mortal(sv); /* Parent created it permanently */
cv = *cvp;
if (!pv && s)
pv = newSVpvn_flags(s, len, SVs_TEMP);
SPAGAIN ;
/* Check the eval first */
- if (!PL_in_eval && SvTRUE(ERRSV)) {
+ if (!PL_in_eval && ((errsv = ERRSV), SvTRUE_NN(errsv))) {
STRLEN errlen;
const char * errstr;
- sv_catpvs(ERRSV, "Propagated");
- errstr = SvPV_const(ERRSV, errlen);
+ sv_catpvs(errsv, "Propagated");
+ errstr = SvPV_const(errsv, errlen);
yyerror_pvn(errstr, errlen, 0); /* Duplicates the message inside eval */
(void)POPs;
- res = SvREFCNT_inc_simple(sv);
+ res = SvREFCNT_inc_simple_NN(sv);
}
else {
res = POPs;
- SvREFCNT_inc_simple_void(res);
+ SvREFCNT_inc_simple_void_NN(res);
}
PUTBACK ;
why2 = key;
why3 = "}} did not return a defined value";
sv = res;
+ (void)sv_2mortal(sv);
goto report;
}
*slp
*/
STATIC char *
-S_scan_word(pTHX_ register char *s, char *dest, STRLEN destlen, int allow_package, STRLEN *slp)
+S_scan_word(pTHX_ char *s, char *dest, STRLEN destlen, int allow_package, STRLEN *slp)
{
dVAR;
char *d = dest;
for (;;) {
if (d >= e)
Perl_croak(aTHX_ ident_too_long);
- if (isALNUM(*s) || (!UTF && isALNUMC_L1(*s))) /* UTF handled below */
+ if (isALNUM(*s)
+ || (!UTF && isALPHANUMERIC_L1(*s))) /* UTF handled below */
+ {
*d++ = *s++;
+ }
else if (allow_package && (*s == '\'') && isIDFIRST_lazy_if(s+1,UTF)) {
*d++ = ':';
*d++ = ':';
else if (UTF && UTF8_IS_START(*s) && isALNUM_utf8((U8*)s)) {
char *t = s + UTF8SKIP(s);
size_t len;
- while (UTF8_IS_CONTINUED(*t) && is_utf8_mark((U8*)t))
+ while (UTF8_IS_CONTINUED(*t) && _is_utf8_mark((U8*)t))
t += UTF8SKIP(t);
len = t - s;
if (d + len > e)
}
STATIC char *
-S_scan_ident(pTHX_ register char *s, register const char *send, char *dest, STRLEN destlen, I32 ck_uni)
+S_scan_ident(pTHX_ char *s, const char *send, char *dest, STRLEN destlen, I32 ck_uni)
{
dVAR;
char *bracket = NULL;
}
else if (UTF && UTF8_IS_START(*s) && isALNUM_utf8((U8*)s)) {
char *t = s + UTF8SKIP(s);
- while (UTF8_IS_CONTINUED(*t) && is_utf8_mark((U8*)t))
+ while (UTF8_IS_CONTINUED(*t) && _is_utf8_mark((U8*)t))
t += UTF8SKIP(t);
if (d + (t - s) > e)
Perl_croak(aTHX_ ident_too_long);
char *end = s;
while ((end < send && isALNUM_lazy_if(end,UTF)) || *end == ':') {
end += UTF8SKIP(end);
- while (end < send && UTF8_IS_CONTINUED(*end) && is_utf8_mark((U8*)end))
+ while (end < send && UTF8_IS_CONTINUED(*end) && _is_utf8_mark((U8*)end))
end += UTF8SKIP(end);
}
Copy(s, d, end - s, char);
*/
STATIC char *
-S_scan_heredoc(pTHX_ register char *s)
+S_scan_heredoc(pTHX_ char *s)
{
dVAR;
I32 op_type = OP_SCALAR;
s += termlen;
#ifdef PERL_MAD
tstart = SvPVX(PL_linestr) + stuffstart;
- if (!PL_thisopen && !keep_delims) {
+ if (PL_madskills && !PL_thisopen && !keep_delims) {
PL_thisopen = newSVpvn(tstart, s - tstart);
stuffstart = s - SvPVX(PL_linestr);
}
SV *sv = NULL; /* place to put the converted number */
bool floatit; /* boolean: int or float? */
const char *lastub = NULL; /* position of last underbar */
- static char const number_too_long[] = "Number too long";
+ static const char* const number_too_long = "Number too long";
PERL_ARGS_ASSERT_SCAN_NUM;
case 'v':
vstring:
sv = newSV(5); /* preallocate storage space */
+ ENTER_with_name("scan_vstring");
+ SAVEFREESV(sv);
s = scan_vstring(s, PL_bufend, sv);
+ SvREFCNT_inc_simple_void_NN(sv);
+ LEAVE_with_name("scan_vstring");
break;
}
}
STATIC char *
-S_scan_formline(pTHX_ register char *s)
+S_scan_formline(pTHX_ char *s)
{
dVAR;
char *eol;
else
qerror(msg);
if (PL_error_count >= 10) {
- if (PL_in_eval && SvCUR(ERRSV))
+ SV * errsv;
+ if (PL_in_eval && ((errsv = ERRSV), SvCUR(errsv)))
Perl_croak(aTHX_ "%"SVf"%s has too many errors.\n",
- SVfARG(ERRSV), OutCopFILE(PL_curcop));
+ SVfARG(errsv), OutCopFILE(PL_curcop));
else
Perl_croak(aTHX_ "%s has too many errors.\n",
OutCopFILE(PL_curcop));
Function must be called like
- sv = newSV(5);
+ sv = sv_2mortal(newSV(5));
s = scan_vstring(s,e,sv);
where s and e are the start and end of the string.
The sv should already be large enough to store the vstring
passed in, for performance reasons.
+This function may croak if fatal warnings are enabled in the
+calling scope, hence the sv_2mortal in the example (to prevent
+a leak). Make sure to do SvREFCNT_inc afterwards if you use
+sv_2mortal.
+
*/
char *