X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=regexec.c;h=fa3eb1ba5770a216846cc7c85c5b323723ea671a;hb=901017d64784f408c305bb1efe061b7ae9d47fc0;hp=17ee6af0f722bc257686d2a46371ce39ab118466;hpb=e45843364c6c14961f5f9e7f270e051a7c0891dc;p=p5sagit%2Fp5-mst-13.2.git diff --git a/regexec.c b/regexec.c index 17ee6af..fa3eb1b 100644 --- a/regexec.c +++ b/regexec.c @@ -55,7 +55,6 @@ # define PERL_NO_GET_CONTEXT #endif -/*SUPPRESS 112*/ /* * pregcomp and pregexec -- regsub and regerror are not used in perl * @@ -140,7 +139,12 @@ #define HOP3c(pos,off,lim) ((char*)HOP3(pos,off,lim)) #define HOPMAYBE3c(pos,off,lim) ((char*)HOPMAYBE3(pos,off,lim)) -#define LOAD_UTF8_CHARCLASS(a,b) STMT_START { if (!CAT2(PL_utf8_,a)) { ENTER; save_re_context(); (void)CAT2(is_utf8_, a)((const U8*)b); LEAVE; } } STMT_END +#define LOAD_UTF8_CHARCLASS(class,str) STMT_START { \ + if (!CAT2(PL_utf8_,class)) { bool ok; ENTER; save_re_context(); ok=CAT2(is_utf8_,class)((const U8*)str); assert(ok); LEAVE; } } STMT_END +#define LOAD_UTF8_CHARCLASS_ALNUM() LOAD_UTF8_CHARCLASS(alnum,"a") +#define LOAD_UTF8_CHARCLASS_DIGIT() LOAD_UTF8_CHARCLASS(digit,"0") +#define LOAD_UTF8_CHARCLASS_SPACE() LOAD_UTF8_CHARCLASS(space," ") +#define LOAD_UTF8_CHARCLASS_MARK() LOAD_UTF8_CHARCLASS(mark, "\xcd\x86") /* for use after a quantifier and before an EXACT-like node -- japhy */ #define JUMPABLE(rn) ( \ @@ -174,9 +178,9 @@ static void restore_pos(pTHX_ void *arg); STATIC CHECKPOINT S_regcppush(pTHX_ I32 parenfloor) { - int retval = PL_savestack_ix; + const int retval = PL_savestack_ix; #define REGCP_PAREN_ELEMS 4 - int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS; + const int paren_elems_to_push = (PL_regsize - parenfloor) * REGCP_PAREN_ELEMS; int p; if (paren_elems_to_push < 0) @@ -221,7 +225,6 @@ S_regcppop(pTHX) I32 i; U32 paren = 0; char *input; - I32 tmps; GET_RE_DEBUG_FLAGS_DECL; @@ -237,6 +240,7 @@ S_regcppop(pTHX) /* Now restore the parentheses context. */ for (i -= (REGCP_OTHER_ELEMS - REGCP_FRAME_ELEMS); i > 0; i -= REGCP_PAREN_ELEMS) { + I32 tmps; paren = (U32)SSPOPINT; PL_reg_start_tmp[paren] = (char *) SSPOPPTR; PL_regstartp[paren] = SSPOPINT; @@ -282,7 +286,7 @@ S_regcppop(pTHX) STATIC char * S_regcp_set_to(pTHX_ I32 ss) { - I32 tmp = PL_savestack_ix; + const I32 tmp = PL_savestack_ix; PL_savestack_ix = ss; regcppop(); @@ -365,7 +369,7 @@ S_cache_re(pTHX_ regexp *prog) /* Assumptions: if ANCH_GPOS, then strpos is anchored. XXXX Check GPOS logic */ -/* If SCREAM, then SvPVX(sv) should be compatible with strpos and strend. +/* If SCREAM, then SvPVX_const(sv) should be compatible with strpos and strend. Otherwise, only SvCUR(sv) is used to get strbeg. */ /* XXXX We assume that strpos is strbeg unless sv. */ @@ -406,11 +410,11 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, register SV *check; char *strbeg; char *t; - int do_utf8 = sv ? SvUTF8(sv) : 0; /* if no sv we have to assume bytes */ + const int do_utf8 = sv ? SvUTF8(sv) : 0; /* if no sv we have to assume bytes */ I32 ml_anch; register char *other_last = Nullch; /* other substr checked before this */ char *check_at = Nullch; /* check substr found at this pos */ - I32 multiline = prog->reganch & PMf_MULTILINE; + const I32 multiline = prog->reganch & PMf_MULTILINE; #ifdef DEBUGGING char *i_strpos = strpos; SV *dsv = PERL_DEBUG_PAD_ZERO(0); @@ -427,10 +431,10 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, } DEBUG_EXECUTE_r({ - char *s = PL_reg_match_utf8 ? + const char *s = PL_reg_match_utf8 ? sv_uni_display(dsv, sv, 60, UNI_DISPLAY_REGEX) : strpos; - int len = PL_reg_match_utf8 ? + const int len = PL_reg_match_utf8 ? strlen(s) : strend - strpos; if (!PL_colorset) reginitcolors(); @@ -438,7 +442,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "UTF-8 target...\n")); PerlIO_printf(Perl_debug_log, - "%sGuessing start of match, REx%s `%s%.60s%s%s' against `%s%.*s%s%s'...\n", + "%sGuessing start of match, REx%s \"%s%.60s%s%s\" against \"%s%.*s%s%s\"...\n", PL_colors[4], PL_colors[5], PL_colors[0], prog->precomp, PL_colors[1], @@ -480,7 +484,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, if (!ml_anch) { if ( !(prog->reganch & (ROPT_ANCH_GPOS /* Checked by the caller */ | ROPT_IMPLICIT)) /* not a real BOL */ - /* SvCUR is not set on references: SvRV and SvPVX overlap */ + /* SvCUR is not set on references: SvRV and SvPVX_const overlap */ && sv && !SvROK(sv) && (strpos != strbeg)) { DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "Not at start...\n")); @@ -502,17 +506,17 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, } /* Now should match s[0..slen-2] */ slen--; - if (slen && (*SvPVX(check) != *s + if (slen && (*SvPVX_const(check) != *s || (slen > 1 - && memNE(SvPVX(check), s, slen)))) { + && memNE(SvPVX_const(check), s, slen)))) { report_neq: DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "String not equal...\n")); goto fail_finish; } } - else if (*SvPVX(check) != *s + else if (*SvPVX_const(check) != *s || ((slen = SvCUR(check)) > 1 - && memNE(SvPVX(check), s, slen))) + && memNE(SvPVX_const(check), s, slen))) goto report_neq; goto success_at_start; } @@ -523,9 +527,9 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, end_shift = prog->minlen - start_shift - CHR_SVLEN(check) + (SvTAIL(check) != 0); if (!ml_anch) { - I32 end = prog->check_offset_max + CHR_SVLEN(check) + const I32 end = prog->check_offset_max + CHR_SVLEN(check) - (SvTAIL(check) != 0); - I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end; + const I32 eshift = CHR_DIST((U8*)strend, (U8*)s) - end; if (end_shift < eshift) end_shift = eshift; @@ -550,7 +554,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, the "check" substring in the region corrected by start/end_shift. */ if (flags & REXEC_SCREAM) { I32 p = -1; /* Internal iterator of scream. */ - I32 *pp = data ? data->scream_pos : &p; + I32 * const pp = data ? data->scream_pos : &p; if (PL_screamfirst[BmRARE(check)] >= 0 || ( BmRARE(check) == '\n' @@ -562,7 +566,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, goto fail_finish; /* we may be pointing at the wrong string */ if (s && RX_MATCH_COPIED(prog)) - s = strbeg + (s - SvPVX(sv)); + s = strbeg + (s - SvPVX_const(sv)); if (data) *data->scream_olds = s; } @@ -578,12 +582,12 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, /* Update the count-of-usability, remove useless subpatterns, unshift s. */ - DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s %s substr `%s%.*s%s'%s%s", + DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s %s substr \"%s%.*s%s\"%s%s", (s ? "Found" : "Did not find"), (check == (do_utf8 ? prog->anchored_utf8 : prog->anchored_substr) ? "anchored" : "floating"), PL_colors[0], (int)(SvCUR(check) - (SvTAIL(check)!=0)), - SvPVX(check), + SvPVX_const(check), PL_colors[1], (SvTAIL(check) ? "$" : ""), (s ? " at offset " : "...\n") ) ); @@ -598,7 +602,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, /* Got a candidate. Check MBOL anchoring, and the *other* substr. Start with the other substr. XXXX no SCREAM optimization yet - and a very coarse implementation - XXXX /ttx+/ results in anchored=`ttx', floating=`x'. floating will + XXXX /ttx+/ results in anchored="ttx", floating="x". floating will *always* match. Probably should be marked during compile... Probably it is right to do no SCREAM here... */ @@ -645,12 +649,12 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, multiline ? FBMrf_MULTILINE : 0 ); DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, - "%s anchored substr `%s%.*s%s'%s", + "%s anchored substr \"%s%.*s%s\"%s", (s ? "Found" : "Contradicts"), PL_colors[0], (int)(SvCUR(must) - (SvTAIL(must)!=0)), - SvPVX(must), + SvPVX_const(must), PL_colors[1], (SvTAIL(must) ? "$" : ""))); if (!s) { if (last1 >= last2) { @@ -704,11 +708,11 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, (unsigned char*)last + SvCUR(must) - (SvTAIL(must)!=0), must, multiline ? FBMrf_MULTILINE : 0); - DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s floating substr `%s%.*s%s'%s", + DEBUG_EXECUTE_r(PerlIO_printf(Perl_debug_log, "%s floating substr \"%s%.*s%s\"%s", (s ? "Found" : "Contradicts"), PL_colors[0], (int)(SvCUR(must) - (SvTAIL(must)!=0)), - SvPVX(must), + SvPVX_const(must), PL_colors[1], (SvTAIL(must) ? "$" : ""))); if (!s) { if (last1 == last) { @@ -861,7 +865,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, const int cl_l = (PL_regkind[(U8)OP(prog->regstclass)] == EXACT ? CHR_DIST(str+STR_LEN(prog->regstclass), str) : 1); - char *endpos = (prog->anchored_substr || prog->anchored_utf8 || ml_anch) + const char * const endpos = (prog->anchored_substr || prog->anchored_utf8 || ml_anch) ? HOP3c(s, (prog->minlen ? cl_l : 0), strend) : (prog->float_substr || prog->float_utf8 ? HOP3c(HOP3c(check_at, -start_shift, strbeg), @@ -963,9 +967,10 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, /* We know what class REx starts with. Try to find this position... */ STATIC char * -S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun) +S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, const char *strend, I32 norun) { - I32 doevery = (prog->reganch & ROPT_SKIP) == 0; + dVAR; + const I32 doevery = (prog->reganch & ROPT_SKIP) == 0; char *m; STRLEN ln; STRLEN lnc; @@ -974,7 +979,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun unsigned int c2; char *e; register I32 tmp = 1; /* Scratch variable? */ - register bool do_utf8 = PL_reg_match_utf8; + register const bool do_utf8 = PL_reg_match_utf8; /* We know what class it must start with. */ switch (OP(c)) { @@ -1078,7 +1083,6 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun if (do_utf8) { UV c, f; U8 tmpbuf [UTF8_MAXBYTES+1]; - U8 foldbuf[UTF8_MAXBYTES_CASE+1]; STRLEN len, foldlen; if (c1 == c2) { @@ -1095,6 +1099,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun && (norun || regtry(prog, s)) ) goto got_it; else { + U8 foldbuf[UTF8_MAXBYTES_CASE+1]; uvchr_to_utf8(tmpbuf, c); f = to_utf8_fold(tmpbuf, foldbuf, &foldlen); if ( f != c @@ -1134,6 +1139,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun && (norun || regtry(prog, s)) ) goto got_it; else { + U8 foldbuf[UTF8_MAXBYTES_CASE+1]; uvchr_to_utf8(tmpbuf, c); f = to_utf8_fold(tmpbuf, foldbuf, &foldlen); if ( f != c @@ -1187,7 +1193,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun } tmp = ((OP(c) == BOUND ? isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0); - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (tmp == !(OP(c) == BOUND ? swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) : @@ -1230,7 +1236,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun } tmp = ((OP(c) == NBOUND ? isALNUM_uni(tmp) : isALNUM_LC_uvchr(UNI_TO_NATIVE(tmp))) != 0); - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (tmp == !(OP(c) == NBOUND ? swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8) : @@ -1259,7 +1265,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case ALNUM: if (do_utf8) { - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8)) { if (tmp && (norun || regtry(prog, s))) @@ -1317,7 +1323,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case NALNUM: if (do_utf8) { - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (!swash_fetch(PL_utf8_alnum, (U8*)s, do_utf8)) { if (tmp && (norun || regtry(prog, s))) @@ -1375,7 +1381,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case SPACE: if (do_utf8) { - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8)) { if (tmp && (norun || regtry(prog, s))) @@ -1433,7 +1439,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case NSPACE: if (do_utf8) { - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (!(*s == ' ' || swash_fetch(PL_utf8_space,(U8*)s, do_utf8))) { if (tmp && (norun || regtry(prog, s))) @@ -1491,7 +1497,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case DIGIT: if (do_utf8) { - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (swash_fetch(PL_utf8_digit,(U8*)s, do_utf8)) { if (tmp && (norun || regtry(prog, s))) @@ -1549,7 +1555,7 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, I32 norun break; case NDIGIT: if (do_utf8) { - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); while (s + (uskip = UTF8SKIP(s)) <= strend) { if (!swash_fetch(PL_utf8_digit,(U8*)s, do_utf8)) { if (tmp && (norun || regtry(prog, s))) @@ -1631,14 +1637,12 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * register char *startpos = stringarg; I32 minlen; /* must match at least this many chars */ I32 dontbother = 0; /* how many characters not to try at end */ - /* I32 start_shift = 0; */ /* Offset of the start to find - constant substr. */ /* CC */ I32 end_shift = 0; /* Same for the end. */ /* CC */ I32 scream_pos = -1; /* Internal iterator of scream. */ char *scream_olds; SV* oreplsv = GvSV(PL_replgv); - bool do_utf8 = DO_UTF8(sv); - I32 multiline = prog->reganch & PMf_MULTILINE; + const bool do_utf8 = DO_UTF8(sv); + const I32 multiline = prog->reganch & PMf_MULTILINE; #ifdef DEBUGGING SV *dsv0 = PERL_DEBUG_PAD_ZERO(0); SV *dsv1 = PERL_DEBUG_PAD_ZERO(1); @@ -1646,6 +1650,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * GET_RE_DEBUG_FLAGS_DECL; + (void)data; /* Currently unused */ RX_MATCH_UTF8_set(prog,do_utf8); PL_regcc = 0; @@ -1730,18 +1735,18 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * } DEBUG_EXECUTE_r({ - char *s0 = UTF ? - pv_uni_display(dsv0, (U8*)prog->precomp, prog->prelen, 60, - UNI_DISPLAY_REGEX) : - prog->precomp; - int len0 = UTF ? SvCUR(dsv0) : prog->prelen; - char *s1 = do_utf8 ? sv_uni_display(dsv1, sv, 60, + const char * const s0 = UTF + ? pv_uni_display(dsv0, (U8*)prog->precomp, prog->prelen, 60, + UNI_DISPLAY_REGEX) + : prog->precomp; + const int len0 = UTF ? SvCUR(dsv0) : prog->prelen; + const char * const s1 = do_utf8 ? sv_uni_display(dsv1, sv, 60, UNI_DISPLAY_REGEX) : startpos; - int len1 = do_utf8 ? SvCUR(dsv1) : strend - startpos; + const int len1 = do_utf8 ? SvCUR(dsv1) : strend - startpos; if (!PL_colorset) reginitcolors(); PerlIO_printf(Perl_debug_log, - "%sMatching REx%s `%s%*.*s%s%s' against `%s%.*s%s%s'\n", + "%sMatching REx%s \"%s%*.*s%s%s\" against \"%s%.*s%s%s\"\n", PL_colors[4], PL_colors[5], PL_colors[0], len0, len0, s0, PL_colors[1], @@ -1812,7 +1817,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * #endif if (!(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)) do_utf8 ? to_utf8_substr(prog) : to_byte_substr(prog); - ch = SvPVX(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0]; + ch = SvPVX_const(do_utf8 ? prog->anchored_utf8 : prog->anchored_substr)[0]; if (do_utf8) { while (s < strend) { @@ -1843,7 +1848,6 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * "Did not find anchored character...\n") ); } - /*SUPPRESS 560*/ else if (prog->anchored_substr != Nullsv || prog->anchored_utf8 != Nullsv || ((prog->float_substr != Nullsv || prog->float_utf8 != Nullsv) @@ -1881,7 +1885,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * else last1 = s - 1; /* bogus */ - /* XXXX check_substr already used to find `s', can optimize if + /* XXXX check_substr already used to find "s", can optimize if check_substr==must. */ scream_pos = -1; dontbother = end_shift; @@ -1895,7 +1899,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * multiline ? FBMrf_MULTILINE : 0))) ) { /* we may be pointing at the wrong string */ if ((flags & REXEC_SCREAM) && RX_MATCH_COPIED(prog)) - s = strbeg + (s - SvPVX(sv)); + s = strbeg + (s - SvPVX_const(sv)); DEBUG_EXECUTE_r( did_match = 1 ); if (HOPc(s, -back_max) > last1) { last1 = HOPc(s, -back_min); @@ -1924,12 +1928,12 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * } DEBUG_EXECUTE_r(if (!did_match) PerlIO_printf(Perl_debug_log, - "Did not find %s substr `%s%.*s%s'%s...\n", + "Did not find %s substr \"%s%.*s%s\"%s...\n", ((must == prog->anchored_substr || must == prog->anchored_utf8) ? "anchored" : "floating"), PL_colors[0], (int)(SvCUR(must) - (SvTAIL(must)!=0)), - SvPVX(must), + SvPVX_const(must), PL_colors[1], (SvTAIL(must) ? "$" : "")) ); goto phooey; @@ -1943,22 +1947,22 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * } DEBUG_EXECUTE_r({ SV *prop = sv_newmortal(); - char *s0; - char *s1; + const char *s0; + const char *s1; int len0; int len1; regprop(prop, c); s0 = UTF ? - pv_uni_display(dsv0, (U8*)SvPVX(prop), SvCUR(prop), 60, + pv_uni_display(dsv0, (U8*)SvPVX_const(prop), SvCUR(prop), 60, UNI_DISPLAY_REGEX) : - SvPVX(prop); + SvPVX_const(prop); len0 = UTF ? SvCUR(dsv0) : SvCUR(prop); s1 = UTF ? sv_uni_display(dsv1, sv, 60, UNI_DISPLAY_REGEX) : s; len1 = UTF ? SvCUR(dsv1) : strend - s; PerlIO_printf(Perl_debug_log, - "Matching stclass `%*.*s' against `%*.*s'\n", + "Matching stclass \"%*.*s\" against \"%*.*s\"\n", len0, len0, s0, len1, len1, s1); }); @@ -1984,11 +1988,11 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * last = scream_olds; /* Only one occurrence. */ /* we may be pointing at the wrong string */ else if (RX_MATCH_COPIED(prog)) - s = strbeg + (s - SvPVX(sv)); + s = strbeg + (s - SvPVX_const(sv)); } else { STRLEN len; - const char * const little = SvPV(float_real, len); + const char * const little = SvPV_const(float_real, len); if (SvTAIL(float_real)) { if (memEQ(strend - len + 1, little, len - 1)) @@ -2003,7 +2007,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * if (len) last = rninstr(s, strend, little, little + len); else - last = strend; /* matching `$' */ + last = strend; /* matching "$" */ } } if (last == NULL) { @@ -2055,7 +2059,7 @@ got_it: RX_MATCH_COPY_FREE(prog); if (flags & REXEC_COPY_STR) { I32 i = PL_regeol - startpos + (stringarg - strbeg); -#ifdef PERL_COPY_ON_WRITE +#ifdef PERL_OLD_COPY_ON_WRITE if ((SvIsCOW(sv) || (SvFLAGS(sv) & CAN_COW_MASK) == CAN_COW_FLAGS)) { if (DEBUG_C_TEST) { @@ -2064,7 +2068,7 @@ got_it: (int) SvTYPE(sv)); } prog->saved_copy = sv_setsv_cow(prog->saved_copy, sv); - prog->subbeg = SvPVX(prog->saved_copy); + prog->subbeg = (char *)SvPVX_const(prog->saved_copy); assert (SvPOKp(prog->saved_copy)); } else #endif @@ -2163,7 +2167,7 @@ S_regtry(pTHX_ regexp *prog, char *startpos) $` inside (?{}) could fail... */ PL_reg_oldsaved = prog->subbeg; PL_reg_oldsavedlen = prog->sublen; -#ifdef PERL_COPY_ON_WRITE +#ifdef PERL_OLD_COPY_ON_WRITE PL_nrs = prog->saved_copy; #endif RX_MATCH_COPIED_off(prog); @@ -2380,6 +2384,7 @@ typedef union re_unwind_t { STATIC I32 /* 0 failure, 1 success */ S_regmatch(pTHX_ regnode *prog) { + dVAR; register regnode *scan; /* Current node. */ regnode *next; /* Next node. */ regnode *inner; /* Next node in internal branch. */ @@ -2402,7 +2407,7 @@ S_regmatch(pTHX_ regnode *prog) #if 0 I32 firstcp = PL_savestack_ix; #endif - register bool do_utf8 = PL_reg_match_utf8; + const register bool do_utf8 = PL_reg_match_utf8; #ifdef DEBUGGING SV *dsv0 = PERL_DEBUG_PAD_ZERO(0); SV *dsv1 = PERL_DEBUG_PAD_ZERO(1); @@ -2425,8 +2430,8 @@ S_regmatch(pTHX_ regnode *prog) DEBUG_EXECUTE_r( { SV *prop = sv_newmortal(); - int docolor = *PL_colors[0]; - int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */ + const int docolor = *PL_colors[0]; + const int taill = (docolor ? 10 : 7); /* 3 chars for "> <" */ int l = (PL_regeol - locinput) > taill ? taill : (PL_regeol - locinput); /* The part of the string before starttry has one color (pref0_len chars), between starttry and current @@ -2452,22 +2457,22 @@ S_regmatch(pTHX_ regnode *prog) pref0_len = pref_len; regprop(prop, scan); { - char *s0 = + const char * const s0 = do_utf8 && OP(scan) != CANY ? pv_uni_display(dsv0, (U8*)(locinput - pref_len), pref0_len, 60, UNI_DISPLAY_REGEX) : locinput - pref_len; - int len0 = do_utf8 ? strlen(s0) : pref0_len; - char *s1 = do_utf8 && OP(scan) != CANY ? + const int len0 = do_utf8 ? strlen(s0) : pref0_len; + const char * const s1 = do_utf8 && OP(scan) != CANY ? pv_uni_display(dsv1, (U8*)(locinput - pref_len + pref0_len), pref_len - pref0_len, 60, UNI_DISPLAY_REGEX) : locinput - pref_len + pref0_len; - int len1 = do_utf8 ? strlen(s1) : pref_len - pref0_len; - char *s2 = do_utf8 && OP(scan) != CANY ? + const int len1 = do_utf8 ? strlen(s1) : pref_len - pref0_len; + const char * const s2 = do_utf8 && OP(scan) != CANY ? pv_uni_display(dsv2, (U8*)locinput, PL_regeol - locinput, 60, UNI_DISPLAY_REGEX) : locinput; - int len2 = do_utf8 ? strlen(s2) : l; + const int len2 = do_utf8 ? strlen(s2) : l; PerlIO_printf(Perl_debug_log, "%4"IVdf" <%s%.*s%s%s%.*s%s%s%s%.*s%s>%*s|%3"IVdf":%*s%s\n", (IV)(locinput - PL_bostr), @@ -2484,7 +2489,7 @@ S_regmatch(pTHX_ regnode *prog) 15 - l - pref_len + 1, "", (IV)(scan - PL_regprogram), PL_regindent*2, "", - SvPVX(prop)); + SvPVX_const(prop)); } }); @@ -2578,7 +2583,7 @@ S_regmatch(pTHX_ regnode *prog) case TRIEFL: { - U32 uniflags = ckWARN( WARN_UTF8 ) ? 0 : UTF8_ALLOW_ANY; + const U32 uniflags = ckWARN( WARN_UTF8 ) ? 0 : UTF8_ALLOW_ANY; U8 *uc = ( U8* )locinput; U32 state = 1; U16 charid = 0; @@ -2586,7 +2591,6 @@ S_regmatch(pTHX_ regnode *prog) UV uvc = 0; STRLEN len = 0; STRLEN foldlen = 0; - U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ]; U8 *uscan = (U8*)NULL; STRLEN bufflen=0; accepted = 0; @@ -2615,6 +2619,7 @@ S_regmatch(pTHX_ regnode *prog) uscan += len; len=0; } else { + U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ]; uvc = utf8n_to_uvuni( (U8*)uc, UTF8_MAXLEN, &len, uniflags ); uvc = to_uni_fold( uvc, foldbuf, &foldlen ); foldlen -= UNISKIP( uvc ); @@ -2646,7 +2651,7 @@ S_regmatch(pTHX_ regnode *prog) from previous if blocks */ case TRIE: { - U32 uniflags = ckWARN( WARN_UTF8 ) ? 0 : UTF8_ALLOW_ANY; + const U32 uniflags = ckWARN( WARN_UTF8 ) ? 0 : UTF8_ALLOW_ANY; U8 *uc = (U8*)locinput; U32 state = 1; U16 charid = 0; @@ -2720,7 +2725,7 @@ S_regmatch(pTHX_ regnode *prog) "%*s %sonly one match : #%d <%s>%s\n", REPORT_CODE_OFF+PL_regindent*2, "", PL_colors[4], accept_buff[ 0 ].wordnum, - tmp ? SvPV_nolen( *tmp ) : "not compiled under -Dr", + tmp ? SvPV_nolen_const( *tmp ) : "not compiled under -Dr", PL_colors[5] ); }); PL_reginput = (char *)accept_buff[ 0 ].endpos; @@ -2755,7 +2760,7 @@ S_regmatch(pTHX_ regnode *prog) PerlIO_printf( Perl_debug_log, "%*s %strying alternation #%d <%s> at 0x%p%s\n", REPORT_CODE_OFF+PL_regindent*2, "", PL_colors[4], accept_buff[best].wordnum, - tmp ? SvPV_nolen( *tmp ) : "not compiled under -Dr",scan, + tmp ? SvPV_nolen_const( *tmp ) : "not compiled under -Dr",scan, PL_colors[5] ); }); if ( best= PL_regeol) sayNO; if (NATIVE_TO_UNI(*(U8*)s) != @@ -2812,6 +2817,7 @@ S_regmatch(pTHX_ regnode *prog) else { /* The target is not utf8, the pattern is utf8. */ while (s < e) { + STRLEN ulen; if (l >= PL_regeol) sayNO; if (NATIVE_TO_UNI(*((U8*)l)) != @@ -2926,7 +2932,7 @@ S_regmatch(pTHX_ regnode *prog) if (!nextchr) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); if (!(OP(scan) == ALNUM ? swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8) : isALNUM_LC_utf8((U8*)locinput))) @@ -2949,7 +2955,7 @@ S_regmatch(pTHX_ regnode *prog) if (!nextchr && locinput >= PL_regeol) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); if (OP(scan) == NALNUM ? swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8) : isALNUM_LC_utf8((U8*)locinput)) @@ -2976,13 +2982,13 @@ S_regmatch(pTHX_ regnode *prog) if (locinput == PL_bostr) ln = '\n'; else { - U8 *r = reghop3((U8*)locinput, -1, (U8*)PL_bostr); + const U8 * const r = reghop3((U8*)locinput, -1, (U8*)PL_bostr); ln = utf8n_to_uvchr(r, UTF8SKIP(r), 0, 0); } if (OP(scan) == BOUND || OP(scan) == NBOUND) { ln = isALNUM_uni(ln); - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); n = swash_fetch(PL_utf8_alnum, (U8*)locinput, do_utf8); } else { @@ -3014,7 +3020,7 @@ S_regmatch(pTHX_ regnode *prog) sayNO; if (do_utf8) { if (UTF8_IS_CONTINUED(nextchr)) { - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); if (!(OP(scan) == SPACE ? swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8) : isSPACE_LC_utf8((U8*)locinput))) @@ -3044,7 +3050,7 @@ S_regmatch(pTHX_ regnode *prog) if (!nextchr && locinput >= PL_regeol) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); if (OP(scan) == NSPACE ? swash_fetch(PL_utf8_space, (U8*)locinput, do_utf8) : isSPACE_LC_utf8((U8*)locinput)) @@ -3067,7 +3073,7 @@ S_regmatch(pTHX_ regnode *prog) if (!nextchr) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); if (!(OP(scan) == DIGIT ? swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8) : isDIGIT_LC_utf8((U8*)locinput))) @@ -3090,7 +3096,7 @@ S_regmatch(pTHX_ regnode *prog) if (!nextchr && locinput >= PL_regeol) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); if (OP(scan) == NDIGIT ? swash_fetch(PL_utf8_digit, (U8*)locinput, do_utf8) : isDIGIT_LC_utf8((U8*)locinput)) @@ -3110,7 +3116,7 @@ S_regmatch(pTHX_ regnode *prog) if (locinput >= PL_regeol) sayNO; if (do_utf8) { - LOAD_UTF8_CHARCLASS(mark,"~"); + LOAD_UTF8_CHARCLASS_MARK(); if (swash_fetch(PL_utf8_mark,(U8*)locinput, do_utf8)) sayNO; locinput += PL_utf8skip[nextchr]; @@ -3140,17 +3146,18 @@ S_regmatch(pTHX_ regnode *prog) s = PL_bostr + ln; if (do_utf8 && OP(scan) != REF) { /* REF can do byte comparison */ char *l = locinput; - char *e = PL_bostr + PL_regendp[n]; + const char *e = PL_bostr + PL_regendp[n]; /* * Note that we can't do the "other character" lookup trick as * in the 8-bit case (no pun intended) because in Unicode we * have to map both upper and title case to lower case. */ if (OP(scan) == REFF) { - STRLEN ulen1, ulen2; - U8 tmpbuf1[UTF8_MAXBYTES_CASE+1]; - U8 tmpbuf2[UTF8_MAXBYTES_CASE+1]; while (s < e) { + STRLEN ulen1, ulen2; + U8 tmpbuf1[UTF8_MAXBYTES_CASE+1]; + U8 tmpbuf2[UTF8_MAXBYTES_CASE+1]; + if (l >= PL_regeol) sayNO; toLOWER_utf8((U8*)s, tmpbuf1, &ulen1); @@ -3244,15 +3251,15 @@ S_regmatch(pTHX_ regnode *prog) } else { STRLEN len; - char *t = SvPV(ret, len); + const char *t = SvPV_const(ret, len); PMOP pm; - char *oprecomp = PL_regprecomp; - I32 osize = PL_regsize; - I32 onpar = PL_regnpar; + char * const oprecomp = PL_regprecomp; + const I32 osize = PL_regsize; + const I32 onpar = PL_regnpar; Zero(&pm, 1, PMOP); if (DO_UTF8(ret)) pm.op_pmdynflags |= PMdf_DYN_UTF8; - re = CALLREGCOMP(aTHX_ t, t + len, &pm); + re = CALLREGCOMP(aTHX_ (char*)t, (char*)t + len, &pm); if (!(SvFLAGS(ret) & (SVs_TEMP | SVs_PADTMP | SVf_READONLY | SVs_GMG))) @@ -3264,7 +3271,7 @@ S_regmatch(pTHX_ regnode *prog) } DEBUG_EXECUTE_r( PerlIO_printf(Perl_debug_log, - "Entering embedded `%s%.60s%s%s'\n", + "Entering embedded \"%s%.60s%s%s\"\n", PL_colors[0], re->precomp, PL_colors[1], @@ -3537,7 +3544,7 @@ S_regmatch(pTHX_ regnode *prog) PL_reg_leftiter = PL_reg_maxiter; } if (PL_reg_leftiter-- == 0) { - I32 size = (PL_reg_maxiter + 7 + POSCACHE_START)/8; + const I32 size = (PL_reg_maxiter + 7 + POSCACHE_START)/8; if (PL_reg_poscache) { if ((I32)PL_reg_poscache_size < size) { Renew(PL_reg_poscache, size, char); @@ -3686,7 +3693,7 @@ S_regmatch(pTHX_ regnode *prog) if (OP(next) != c1) /* No choice. */ next = inner; /* Avoid recursion. */ else { - I32 lastparen = *PL_reglastparen; + const I32 lastparen = *PL_reglastparen; I32 unwind1; re_unwind_branch_t *uw; @@ -3996,8 +4003,8 @@ S_regmatch(pTHX_ regnode *prog) count = locinput - old; } else { - STRLEN len; if (c1 == c2) { + STRLEN len; /* count initialised to * utf8_distance(old, locinput) */ while (locinput <= e && @@ -4009,6 +4016,7 @@ S_regmatch(pTHX_ regnode *prog) count++; } } else { + STRLEN len; /* count initialised to * utf8_distance(old, locinput) */ while (locinput <= e) { @@ -4301,7 +4309,7 @@ do_no: case RE_UNWIND_BRANCHJ: { re_unwind_branch_t *uwb = &(uw->branch); - I32 lastparen = uwb->lastparen; + const I32 lastparen = uwb->lastparen; REGCP_UNWIND(uwb->lastcp); for (n = *PL_reglastparen; n > lastparen; n--) @@ -4318,7 +4326,6 @@ do_no: goto do_no; } /* Have more choice yet. Reuse the same uwb. */ - /*SUPPRESS 560*/ if ((n = (uwb->type == RE_UNWIND_BRANCH ? NEXT_OFF(next) : ARG(next)))) next += n; @@ -4357,8 +4364,9 @@ do_no: * rather than incrementing count on every character. [Er, except utf8.]] */ STATIC I32 -S_regrepeat(pTHX_ regnode *p, I32 max) +S_regrepeat(pTHX_ const regnode *p, I32 max) { + dVAR; register char *scan; register I32 c; register char *loceol = PL_regeol; @@ -4431,7 +4439,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case ALNUM: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (hardcount < max && scan < loceol && swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) { scan += UTF8SKIP(scan); @@ -4459,7 +4467,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case NALNUM: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(alnum,"a"); + LOAD_UTF8_CHARCLASS_ALNUM(); while (hardcount < max && scan < loceol && !swash_fetch(PL_utf8_alnum, (U8*)scan, do_utf8)) { scan += UTF8SKIP(scan); @@ -4487,7 +4495,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case SPACE: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); while (hardcount < max && scan < loceol && (*scan == ' ' || swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) { @@ -4516,7 +4524,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case NSPACE: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(space," "); + LOAD_UTF8_CHARCLASS_SPACE(); while (hardcount < max && scan < loceol && !(*scan == ' ' || swash_fetch(PL_utf8_space,(U8*)scan, do_utf8))) { @@ -4545,7 +4553,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case DIGIT: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); while (hardcount < max && scan < loceol && swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) { scan += UTF8SKIP(scan); @@ -4559,7 +4567,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) case NDIGIT: if (do_utf8) { loceol = PL_regeol; - LOAD_UTF8_CHARCLASS(digit,"0"); + LOAD_UTF8_CHARCLASS_DIGIT(); while (hardcount < max && scan < loceol && !swash_fetch(PL_utf8_digit, (U8*)scan, do_utf8)) { scan += UTF8SKIP(scan); @@ -4588,7 +4596,7 @@ S_regrepeat(pTHX_ regnode *p, I32 max) regprop(prop, p); PerlIO_printf(Perl_debug_log, "%*s %s can match %"IVdf" times out of %"IVdf"...\n", - REPORT_CODE_OFF+1, "", SvPVX(prop),(IV)c,(IV)max); + REPORT_CODE_OFF+1, "", SvPVX_const(prop),(IV)c,(IV)max); }); }); @@ -4652,14 +4660,14 @@ S_regrepeat_hard(pTHX_ regnode *p, I32 max, I32 *lp) */ SV * -Perl_regclass_swash(pTHX_ register regnode* node, bool doinit, SV** listsvp, SV **altsvp) +Perl_regclass_swash(pTHX_ register const regnode* node, bool doinit, SV** listsvp, SV **altsvp) { SV *sw = NULL; SV *si = NULL; SV *alt = NULL; if (PL_regdata && PL_regdata->count) { - U32 n = ARG(node); + const U32 n = ARG(node); if (PL_regdata->what[n] == 's') { SV *rv = (SV*)PL_regdata->data[n]; @@ -4704,9 +4712,10 @@ Perl_regclass_swash(pTHX_ register regnode* node, bool doinit, SV** listsvp, SV */ STATIC bool -S_reginclass(pTHX_ register regnode *n, register U8* p, STRLEN* lenp, register bool do_utf8) +S_reginclass(pTHX_ register const regnode *n, register const U8* p, STRLEN* lenp, register bool do_utf8) { - char flags = ANYOF_FLAGS(n); + dVAR; + const char flags = ANYOF_FLAGS(n); bool match = FALSE; UV c = *p; STRLEN len = 0; @@ -4740,7 +4749,7 @@ S_reginclass(pTHX_ register regnode *n, register U8* p, STRLEN* lenp, register b for (i = 0; i <= av_len(av); i++) { SV* sv = *av_fetch(av, i, FALSE); STRLEN len; - char *s = SvPV(sv, len); + const char *s = SvPV_const(sv, len); if (len <= plen && memEQ(s, (char*)p, len)) { *lenp = len; @@ -4896,7 +4905,7 @@ restore_pos(pTHX_ void *arg) if (PL_reg_oldsaved) { PL_reg_re->subbeg = PL_reg_oldsaved; PL_reg_re->sublen = PL_reg_oldsavedlen; -#ifdef PERL_COPY_ON_WRITE +#ifdef PERL_OLD_COPY_ON_WRITE PL_reg_re->saved_copy = PL_nrs; #endif RX_MATCH_COPIED_on(PL_reg_re); @@ -4958,3 +4967,13 @@ S_to_byte_substr(pTHX_ register regexp *prog) prog->check_substr = sv; } } + +/* + * Local variables: + * c-indentation-style: bsd + * c-basic-offset: 4 + * indent-tabs-mode: t + * End: + * + * ex: set ts=8 sts=4 sw=4 noet: + */